United States
                 Environmental Protection
                 Agency
                  Office of Health and
                  Environmental Assessment
                  Washington DC 20460
                 Research and Development
&EPA
Air Quality
Criteria for
Carbon Monoxide
EPA/600/8-90/045 A
March 1990
External Review Draft
 Draft
 (Do Not
 Cite or Quote)
                                 NOTICE

                 This document is an external review draft.  It has not been
                 formally released by EPA and should not at this stage be
                 construed to represent Agency policy. It is being circulated for
                 comment on its technical accuracy and policy implications

-------
    DRAFT:
DO  NOT CITE
 OR QUOTE
EPA 600-8/90/045A
March 1990
External Review Draft
        Air  Quality Criteria
                    for
         Carbon  Monoxide
                      NOTICE
     This document is a preliminary draft It has not been formally
     released by EPA and should not at this stage be construed to
     represent Agency policy. It 1« being circulated for comment on
     tts technical accuracy and policy Implication*.
  Environmental Criteria  and Assessment Office
 Office of Health and Environmental Assessment
      Office of Research and Development
      U.S. Environmental Protection Agency
       Research Triangle Park, NC 27711

         U.S. Environmental Protection Agency
         Region 5, Library (PL-12J)
         77 West Jackson Boulevard, 12th Floor
         Chicago, IL  60604-3590

-------
                                  DISCLAIMER








     This document is an external draft for review purposes only and does not constitute



Agency policy. Mention of trade names or commercial products does not constitute



endorsement or recommendation for use.
                                                        .'  "\

-------
                              CONTENTS
                                                                  Page

TABLES 	       xiii
FIGURES   	       xix
AUTHORS, CONTRIBUTORS, AND REVIEWERS  	       xxv

1.   SUMMARY AND CONCLUSIONS	       1-1

2.   INTRODUCTION  	       2-1
    2.1 ORGANIZATION AND CONTENT OF THIS DOCUMENT ...       2-1
    2.2 LEGISLATIVE HISTORY OF NAAQS	       2-3
    2.3 REGULATORY BACKGROUND FOR CARBON MONOXIDE
        NAAQS	       2-4
    2.4 SCIENTIFIC BACKGROUND FOR THE CURRENT
        CARBON MONOXIDE NAAQS   	       2-6
        2.4.1   Mechanisms of Action  	       2-6
        2.4.2   Carbon Monoxide Exposure Levels 	       2-8
        2.4.3   Health Effects of Low-Level Carbon Monoxide
               Exposures	       2-8
               2.4.3.1   Cardiovascular Effects  	       2-8
               2.4.3.2   Neurobehavioral Effects 	       2-10
               2.4.3.3   Other Health Effects   	       2-11
    2.5 CRITICAL ISSUES IN REVIEW OF THE NAAQS
        FOR CARBON MONOXIDE	       2-11
        2.5.1   Exposure Assessment in the Population	       2-11
        2.5.2   Mechanisms of Action of Carbon Monoxide	       2-14
        2.5.3   Health Effects from Exposure to Carbon Monoxide  ....       2-16
               2.5.3.1   Effects on the Cardiovascular System	       2-16
               2.5.3.2   Neurobehavioral Effects 	       2-18
               2.5.3.3   Perinatal Effects	       2-19
        2.5.4   Population Groups at Greatest Risk for Ambient
               CO Exposure Effects	       2-19
    2.6 CARBON MONOXIDE POISONING	       2-20
    2.7 REFERENCES	       2-23

3.   PROPERTIES AND PRINCIPLES OF FORMATION OF  	
    CARBON MONOXIDE  	       3-1
    3.1 INTRODUCTION	       3-1
    3.2 PHYSICAL PROPERTIES	       3-2
    3.3 GASEOUS CHEMICAL REACTIONS OF
        CARBON MONOXIDE	       3-2
    3.4 PRINCIPLES OF FORMATION BY SOURCE CATEGORY ...       3-7
        3.4.1   General Combustion Processes	       3-8
                                  in

-------
                                 CONTENTS (cont'd)

                                                                      Page

         3.4.2   Combustion Engines  	       3-11
         3.4.3   Other Sources  	       3-14
     3.5  REFERENCES 	       3-15

4.    THE GLOBAL CYCLE OF CARBON MONOXIDE:
     TRENDS AND MASS BALANCE  	       4-1
     4.1  INTRODUCTION	       4-1
     4.2  GLOBAL SOURCES, SINKS, AND LIFETIME	       4-1
         4.2.1   Sources 	       4-2
         4.2.2   Sinks  	       4-3
         4.2.3   Atmospheric Lifetime  	       4-5
         4.2.4   Latitudinal Distribution of Sources  	       4-5
         4.2.5   Uncertainties and Consistencies   	       4-6
     4.3  GLOBAL DISTRIBUTIONS   	       4-9
         4.3.1   Seasonal Variations	       4-9
         4.3.2   Latitudinal Variation  	       4-10
         4.3.3   Variations with Altitude 	       4-10
         4.3.4   Other Variations  	       4-12
     4.4  GLOBAL TRENDS  	       4-12
     4.5  SUMMARY   	       4-15
     4.6  REFERENCES 	       4-16

5.    MEASUREMENT METHODS FOR CARBON MONOXIDE   	       5-1
     5.1  INTRODUCTION	       5-1
         5.1.1   Overview of Techniques for Measurement
                of Ambient Carbon Monoxide  	       5-2
         5.1.2   Calibration Requirements   	       5-5
     5.2  PREPARATION OF STANDARD REFERENCE MATERIALS  .       5-5
         5.2.1   Gas Standards  	       5-5
         5.2.2   Gravimetric Method  	       5-6
         5.2.3   Volumetric Gas Dilution Methods	       5-8
         5.2.4   Other Methods  	       5-8
     5.3  MEASUREMENT IN AMBIENT AIR  	       5-9
         5.3.1   Sampling System Components  	       5-9
         5.3.2   Quality Assurance Procedures for Sampling	       5-11
         5.3.3   Sampling Schedules   	       5-13
         5.3.4   Continuous Analysis  	       5-14
                5.3.4.1  Nondispersive Infrared Photometry  	       5-14
                5.3.4.2  Gas Chromatography - Flame lonization   ...       5-18
                5.3.4.3  Other Analyzers  	       5-19
         5.3.5   Intermittent Analysis  	       5-23
                5.3.5.1  Colorimetric Analysis  	       5-23
                                     IV

-------
                                 CONTENTS (cont'd)
     5.4  MEASUREMENT USING PERSONAL MONITORS	       5-25
     5.5  REFERENCES	       5-27

6.    AMBIENT CARBON MONOXIDE SOURCES, EMISSIONS, AND
     CONCENTRATIONS   	       6-1
     6.1  ESTIMATING NATIONAL EMISSION FACTORS  	       6-1
     6.2  EMISSION SOURCES AND EMISSION FACTORS BY
         SOURCE CATEGORY	       6-1
         6.2.1   Transportation Sources   	       6-3
                 6.2.1.1   Motor Vehicles  	       6-3
                 6.2.1.2   Aircraft  	       6-4
                 6.2.1.3   Railroads  	       6-4
                 6.2.1.4   Vessels   	       6-4
                 6.2.1.5   Nonhighway Use of Motor Fuels  	       6-4
         6.2.2   Stationary Source Fuel Combustion	       6-5
         6.2.3   Industrial Processes   	       6-5
         6.2.4   Solid Waste Disposal	       6-6
         6.2.5   Miscellaneous Combustion Sources  	       6-6
                 6.2.5.1   Forest Fires   	       6-6
                 6.2.5.2   Agricultural Burning  	       6-6
                 6.2.5.3   Coal Refuse   	       6-7
                 6.2.5.4   Structural Fires  	       6-7
     6.3  NATIONAL CO EMISSIONS ESTIMATES 1970-1988   	       6-7
     6.4  OUTDOOR AIR CONCENTRATIONS  	       6-11
         6.4.1   Introduction   	       6-11
         6.4.2   Site Selection  	       6-11
         6.4.3   United States Data Base   	       6-14
         6.4.4   Techniques of Data Analysis  	       6-15
                 6.4.4.1   Frequency Analysis	,	       6-16
                 6.4.4.2   Trend Analyses  	       6-17
                 6.4.4.3   Special Analyses   	       6-17
         6.4.5   Urban Levels of Carbon Monoxide  	       6-19
                 6.4.5.1   Ten-year CO  Trends 1979-1988  	       6-19
                 6.4.5.2   Five-year CO Trends 1984-1988  	       6-22
                 6.4.5.3   Air Quality Levels in Metropolitan Statistical
                         Areas  	       6-24
         6.4.6   Effects of Meteorology and Topography	       6-24
     6.5  CARBON MONOXIDE DISPERSION MODELS  	       6-32
         6.5.1   Line Source Modeling   	       6-32
                 6.5.1.1   CALINE3	       6-33
                 6.5.1.2   GMLINE  	       6-33
                 6.5.1.3   HIWAY-2   	       6-33
                 6.5.1.4   PAL  	       6-34

-------
                                 CONTENTS (cont'd)

                                                                       Page

                6.5.1.5   Model Evaluation  	       6-34
         6.5.2   Intersection Modeling  	       6-35
                6.5.2.1   "Volume 9"   	       6-35
                6.5.2.2   Intersection Midblock Model  	       6-36
                6.5.2.3   Georgia Intersection Model  	       6-37
                6.5.2.4   TEXIN2	       6-37
                6.5.2.5   CAL3Q  	       6-39
                6.5.2.6   CALINE4	       6-39
                6.5.2.7   Comparison of Intersection Models  	       6-40
         6.5.3   Urban Area Modeling  	       6-43
                6.5.3.1   APRAC-3	       6-44
                6.5.3.2   Urban Airshed Model	       6-45
                6.5.3.3   RAM   	       6-45
     6.6  REFERENCES 	       6-46

7.    INDOOR CARBON MONOXIDE SOURCES, EMISSIONS, AND
     CONCENTRATIONS   	       7-1
     7.1  INTRODUCTION	       7-1
     7.2  EMISSIONS FROM INDOOR SOURCES  	       7-4
         7.2.1   Emissions from Gas Cooking Ranges, Gas
                Ovens, and Gas Appliances	; .       7-5
         7.2.2   Emissions from Unvented Space Heaters   	       7-12
         7.2.3   Emissions from Wood Stoves and Tobacco Combustion  .       7-15
         7.2.4   Summary of Emission Data	       7-17
     7.3  CONCENTRATIONS IN INDOOR ENVIRONMENTS  	       7-20
         7.3.1   Indoor Concentrations Recorded in Personal Exposure
                Studies   	       7-20
         7.3.2   Targeted  Microenvironmental  Studies  	       7-25
                7.3.2.1   Indoor Microenvironmental Concentrations  ..       7-25
                7.3.2.2   Concentrations Associated with Indoor
                         Sources  	       7-29
         7.3.3   Spatial Concentration Variations	       7-39
         7.3.4   Summary of Indoor Concentrations   	       7-42
     7.4  REFERENCES 	       7-44

8.    POPULATION EXPOSURE TO CARBON MONOXIDE   	       8-1
     8.1  INTRODUCTION	       8-1
     8.2  EXPOSURE MONITORING IN THE POPULATION  	       8-3
         8.2.1   Personal  Monitoring  	       8-4
         8.2.2   Carbon Monoxide Exposures Indoors	       8-8
         8.2.3   Carbon Monoxide Exposures Inside Vehicles	       8-12
         8.2.4   Carbon Monoxide Exposures Outdoors	       8-14
                                      VI

-------
                                  CONTENTS (cont'd)
                                                                         Page
     8.3  ESTIMATING POPULATION EXPOSURE TO
          CARBON MONOXIDE  	        8-15
          8.3.1   Defining Concentration, Exposure, and Dose	        8-16
          8.3.2   Components of Exposure  	        8-17
          8.3.3   Relationship to Fixed-Site Monitors  	        8-20
          8.3.4   Alternative Approaches to Exposure Estimation  	        8-21
          8.3.5   Statistical Models Based on Personal Monitoring Data  . .        8-23
          8.3.6   Physical and Physical-Stochastic Models	        8-28
     8.4  OCCUPATIONAL EXPOSURE TO CARBON MONOXIDE  . . .        8-44
          8.4.1   Historical Perspective  	        8-45
          8.4.2   Exposure  Monitoring Techniques  	        8-46
          8.4.3   Occupational Exposures   	        8-51
     8.5  BIOLOGICAL MONITORING   	        8-61
          8.5.1   Blood Carboxyhemoglobin Measurement   	        8-61
                 8.5.1.1   Measurement Methods  	        8-61
                 8.5.1.2   Carboxyhemoglobin Measurements in
                          Populations 	        8-75
          8.5.2   Carbon Monoxide in Expired Breath  	        8-80
                 8.5.2.1   Measurement Methods  	        8-82
                 8.5.2.2   Breath Measurements in Populations  	        8-89
          8.5.3   Potential Limitations  	        8-98
                 8.5.3.1   Pulmonary  Disease  	        8-98
                 8.5.3.2   Subject Age   	        8-99
                 8.5.3.3   Effects of Smoking  	        8-99
     8.6  SUMMARY AND CONCLUSIONS   	        8-100
     8.7  REFERENCES	        8-103

9.    PHARMACOKINETICS AND MECHANISMS OF
     ACTION OF CARBON MONOXIDE   	        9-1
     9.1   ABSORPTION, DISTRIBUTION, AND
          PULMONARY ELIMINATION   	        9-1
          9.1.1   Introduction   	        9-1
          9.1.2   Pulmonary Uptake	        9-1
                 9.1.2.1   Mass Transfer of Carbon Monoxide	        9-1
                 9.1.2.2   Effects of Dead Space and Uneven  Distribution
                          of Ventilation and Perfusion   	        9-2
                 9.1.2.3   Alveolo-Capillary Membrane and
                          Blood-Phase Diffusion  	        9-4
          9.1.3   Tissue Uptake  	        9-5
                 9.1.3.1   The Blood   	        9-6
                 9.1.3.2   The Lung  	        9-8
                 9.1.3.3   Heart and Skeletal Muscles  	        9-8
                 9.1.3.4   Brain and Other Tissues 	        9-9
          9.1.4   Pulmonary and Tissue Elimination  	        9-10

                                      vii

-------
                                 CONTENTS (cont'd)

                                                                       Page

    9.2  TISSUE PRODUCTION AND METABOLISM OF
         CARBON MONOXIDE  	       9-11
    9.3  MODELING CARBOXYHEMOGLOBIN FORMATION  	       9-12
         9.3.1   Introduction   	       9-12
         9.3.2   Regression Models   	       9-12
         9.3.3   The Coburn-Forster-Kane Differential Equations  	       9-14
                9.3.3.1   Linear and Nonlinear CFK Differential
                         Equations  	       9-14
                9.3.3.2   Confirmation Studies of the CFK Model  ....       9-16
                9.3.3.3   Modified CFK Models 	       9-18
                9.3.3.4   Application of the CFK Model  	       9-20
         9.3.4   Summary  	       9-21
    9.4  INTRACELLULAR EFFECTS OF CARBON MONOXIDE  ....       9-22
         9.4.1   Introduction   	       9-22
         9.4.2   Carbon Monoxide Binding to Myoglobin   	       9-26
         9.4.3   Carbon Monoxide Uptake by Cytochrome P-450  	       9-27
         9.4.4   Carbon Monoxide and Cytochrome c Oxidase   	       9-28
    9.5  REFERENCES	       9-32

10.  HEALTH EFFECTS OF CARBON MONOXIDE  	       10-1
    10.1 INTRODUCTION	       10-1
    10.2 ACUTE PULMONARY EFFECTS OF CARBON MONOXIDE  .       10-4
         10.2.1  Introduction   	       10-4
         10.2.2  Effects on Lung Morphology	       10-4
                10.2.2.1  Studies in Laboratory Animals  	       10-5
                10.2.2.2  Studies in Humans   	       10-7
         10.2.3  Effects on Lung Function  	       10-7
                10.2.3.1  Lung Function in Laboratory Animals  	       10-7
                10.2.3.2  Lung Function in Humans	       10-9
         10.2.4  Summary 	       10-13
    10.3 CARDIOVASCULAR EFFECTS OF CARBON MONOXIDE  .  .       10-14
         10.3.1  Introduction   	       10-14
         10.3.2  Experimental Studies in Humans  	       10-15
                10.3.2.1  Cardiorespiratory Response to Exercise   ....       10-15
                10.3.2.2  Arrhythmogenic Effects  	       10-35
                10.3.2.3  Effects on Coronary Blood Flow	       10-37
         10.3.3  Relationship between CO Exposure and Risk of
                Cardiovascular Disease in Man   	       10-38
                10.3.3.1  Risk of Ischemic Heart Disease  	       10-38
                10.3.3.2  Risk of Hypertension	       10-42
         10.3.4  Studies in Laboratory Animals  	       10-42
                10.3.4.1   Introduction   	       10-42
                10.3.4.2   Ventricular Fibrillation Studies	       10-43

                                     viii

-------
                              CONTENTS (cont'd)

                                                                      Page

             10.3.4.3  Hemodynamic Studies  	        10-47
             10.3.4.4  Cardiomegaly  	        10-52
             10.3.4.5  Hematology Studies   	        10-58
             10.3.4.6  Atherosclerosis and Thrombosis  	        10-60
     10.3.5   Summary and Conclusions  	        10-69
10.4 CEREBROVASCULAR AND BEHAVIORAL EFFECTS OF
     CARBON MONOXIDE  	        10-71
     10.4.1   Control of Cerebral Blood Flow and Tissue PO2
             with Carbon Monoxide and Hypoxic Hypoxia   	        10-71
             10.4.1.1  Introduction   	        10-71
             10.4.1.2  Effects on Global Cerebral  Blood Flow   ....        10-72
             10.4.1.3  Effects on Regional Cerebral Blood Flow  ...        10-83
             10.4.1.4  Effect of Low Levels of Carbon Monoxide on
                      Cerebral Blood Flow  	        10-87
             10.4.1.5  Synergistic Effects of Carbon Monoxide  ....        10-93
             10.4.1.6  Mechanism of Regulation of Cerebral Blood
                      Flow in Hypoxia   	        10-99
             10.4.1.7  Summary  	        10-102
     10.4.2   Behavioral Effects of Carbon Monoxide  	        10-103
             10.4.2.1  Introduction   	        10-103
             10.4.2.2  Sensory Effects  	        10-105
             10.4.2.3  Motor and Sensorimotor Performance   	        10-113
             10.4.2.4  Vigilance  	        10-122
             10.4.2.5  Miscellaneous Measures of Performance  ....        10-124
             10.4.2.6  Automobile Driving   	        10-131
             10.4.2.7  Brain Electrical Activity  	        10-132
             10.4.2.8  Schedule-Controlled Behavior   	        10-137
             10.4.2.9  Summary and Discussion of Behavioral
                      Literature  	        10-139
             10.4.2.10  Hypotheses	        10-146
             10.4.2.11  Conclusions   	        10-147
10.5 DEVELOPMENTAL TOXICITY OF CARBON MONOXIDE   . .        10-148
     10.5.1   Introduction  	        10-148
     10.5.2   Theoretical Basis for Fetal Exposure to Excessive
             Carbon Monoxide and for Excess Fetal Toxicity	        10-151
             10.5.2.1  Evidence for Elevated Fetal
                      Carboxyhemoglobin Relative to
                      Maternal Hemoglobin   	        10-151
             10.5.2.2  Effect of Maternal Carboxyhemoglobin on
                      Placental O2 Transport   	        10-152
     10.5.3   Measurement of Carboxyhemoglobin Content in
             Fetal Blood	        10-153
     10.5.4   Consequences of Carbon Monoxide in Development  ...        10-154

                                   ix

-------
                                CONTENTS (cont'd)

                                                                      Page

                10.5.4.1  Fetotoxic and Teratogenic Consequence of
                        Prenatal Carbon Monoxide Exposure  	        10-155
                10.5.4.2  Carbon Monoxide and Body Weight   	        10-159
                10.5.4.3  Alteration in Cardiovascular Development
                        following Early Carbon Monoxide Exposure  .        10-161
                10.5.4.4  Neurobehavioral Consequences of Perinatal
                        Carbon Monoxide Exposure  	        10-167
                10.5.4.5  Neurochemical Consequences of Prenatal and
                        Perinatal Carbon Monoxide Exposure   	        10-174
                10.5.4.6  Morphological Consequences of Acute Prenatal
                        Carbon Monoxide	        10-177
         10.5.5  Summary  	        10-178
    10.6 OTHER SYSTEMIC EFFECTS OF CARBON MONOXIDE   . . .        10-178
    10.7 ADAPTATION, HABITUATION,  AND COMPENSATORY
         RESPONSES TO CARBON MONOXIDE
         EXPOSURE  	        10-182
         10.7.1  Short-Term Habituation  	        10-183
         10.7.2  Long-Term Adaptation   	        10-184
         10.7.3  Summary  	        10-187
    10.8 REFERENCES 	        10-188

11.  COMBINED EXPOSURE OF CARBON MONOXIDE WITH
    OTHER POLLUTANTS, DRUGS, AND ENVIRONMENTAL
    FACTORS  	        11-1
    11.1 HIGH ALTITUDE EFFECTS OF CARBON MONOXIDE  	        11-1
         11.1.1  Introduction   	        11-1
         11.1.2  Carboxyhemoglobin Formation  	        11-3
         11.1.3  Cardiovascular Effects   	        11-4
         11.1.4  Chronic Studies  	        11-10
         11.1.5  Neurobehavioral Effects	        11-16
         11.1.6  Compartmental Shifts   	        11-17
         11.1.7  Conclusions   	        11-18
    11.2 CARBON MONOXIDE INTERACTIONS WITH DRUGS  ....        11-18
         11.2.1  Introduction   	        11-18
         11.2.2  Alcohol  	        11-20
         11.2.3  Barbiturates   	        11-23
         11.2.4  Other Psychoactive Drugs	        11-24
    11.3 COMBINED EXPOSURE TO CARBON MONOXIDE AND
         OTHER AIR POLLUTANTS AND ENVIRONMENTAL
         FACTORS  	        11-24
         11.3.1  Exposure in Ambient Air  	        11-25
         11.3.2  Exposure to Combustion Products	        11-30
         11.3.3  Exposure to Other Environmental  Factors	        11-35

-------
                                 CONTENTS (cont'd)
                11.3.3.1  Environmental Heat   	       11-35
                11.3.3.2  Environmental Noise  	       11-36
         11.3.4  Summary 	       11-37
     11.4 ENVIRONMENTAL TOBACCO SMOKE	       11-38
     11.5 REFERENCES  	       11-40

12.   EVALUATION OF SUBPOPULATIONS POTENTIALLY AT RISK
     TO CARBON MONOXIDE EXPOSURE	       12-1
     12.1 INTRODUCTION	       12-1
     12.2 AGE AND GENDER AS RISK FACTORS  	       12-2
     12.3 RISK OF CARBON MONOXIDE EXPOSURE IN
         INDIVIDUALS WITH PREEXISTING DISEASE 	       12-4
         12.3.1  Subjects with Coronary Artery Disease	       12-4
         12.3.2  Subjects with Congestive Heart Failure	       12-5
         12.3.3  Subjects with Other Vascular Diseases  	       12-5
         12.3.4  Subjects with Anemia and Other Hematologic
                Disorders 	       12-6
         12.3.5  Subjects with Obstructive Lung Disease  	       12-7
     12.4 SUBPOPULATIONS AT RISK FROM COMBINED
         EXPOSURE TO CARBON MONOXIDE AND OTHER
         CHEMICAL SUBSTANCES   	       12-8
         12.4.1  Interactions with Psychoactive Drugs	       12-8
         12.4.2  Interactions with Cardiovascular Drugs   	       12-9
         12.4.3  Mechanisms of Carbon Monoxide Interactions with
                Drugs:  Need for Further Research  	       12-10
                12.4.3.1   Metabolic Effects  	,	       12-10
                12.4.3.2  Central Nervous System Depression	       12-10
                12.4.3.3  Alteration in Cerebral Blood Flow	       12-11
         12.4.4  Interactions with Other Chemical  Substances
                in the Environment	       12-13
     12.5 SUBPOPULATIONS EXPOSED TO CARBON MONOXIDE
         AT HIGH ALTITUDES  	       12-14
     12.6 REFERENCES  	       12-17

APPENDIX A: GLOSSARY OF TERMS AND SYMBOLS	       A-l
                                     XI

-------
                                      TABLES


Number                                                                      Page

2-1       National Ambient Air Quality Standards for Carbon Monoxide ....       2-6

2-2       Lowest Observed Effect Levels for Human Health Effects
          Associated with Low-Level Carbon Monoxide Exposure   	       2-9

3-1       Physical Properties of Carbon Monoxide 	       3-3

3-2       Reported Room Temperature Rate Constants for the Reaction
          of OH* Radicals with CO	       3-5

3-3       Summary of Light-Duty Vehicle (LDV) Emissions Standards   ....       3-9

4-1       Sources of Carbon Monoxide   	       4-4

5-1       Performance Specifications for Automated Analytical Methods
          for Carbon Monoxide (Code of Federal Regulations, 1977a)    ....       5-3

6-1       Carbon Monoxide National Emission Estimates (teragrams/year)  .  .       6-2

6-2       Carbon Monoxide Emissions from Transportation
          (gigagrams/year)  	       6-8

6-3       Specific Probe Exposure Criteria  	       6-13

6-4       National Carbon Monoxide Emission Estimates, 1979-1988   	       6-22

6-5       Distribution of Population in Metropolitan Statistical Areas   	       6-25

6-6       Parameters for Intersection Scenarios   	       6-41

6-7       Results of Model Comparisons for the Undercapacity
          Intersection Scenario   	       6-42

6-8       Results of Model Comparisons for the Near Capacity
          Intersection Scenario   	       6-42

6-9       Results of Model Comparisons for the Overcapacity Intersection
          Scenario  	       6-43

7-1       Carbon Monoxide-Emission Rates for 12 Range-Top Burners
          Operating with Blue and Yellow-Tipping Flames by the Direct
          Sampling Method   	       7-7

                                         xiii

-------
                                     TABLES (cont'd)


Number                                                                     Page
7-2       Carbon Monoxide-Emission Rates for Gas Range Ovens,
          Gas Range Pilot Lights, and Gas Dryers  	       7-8

7-3       Carbon Monoxide-Emission Rates from 18 Gas Ranges,
          Gas Ovens, and Gas Pilot Lights for Blue Flame and
          Yellow-Tipping Flame by the Direct-Sampling Method  	       7-9

7-4       Carbon Monoxide Emissions from Gas Ranges for Studies
          of Small Sample Size  	       7-12

7-5       Carbon Monoxide Emissions from Unvented Gas Space Heaters   . .       7-14

7-6       Carbon Monoxide Emissions from Unvented Kerosene Space
          Heaters	       7-16

7-7       Summary of CO Exposure Levels and Time Spent Per Day in
          Selected Microenvironments	       7-22

7-8       Indoor Microenvironments Listed in Descending Order of
          Weighted Mean CO Concentration  	       7-23

7-9       Weighted Means of Residential Exposure Grouped According to
          the Presence or Absence of Selected Indoor Carbon
          Monoxide Sources  	       7-24

7-10      Average Residential CO Exposures (ppm):  Impact of
          Combustion Appliance Use and Tobacco Smoking  	       7-25

7-11      Carbon Monoxide Concentrations Measured in Various
          Indoor Environments as a Function of Microenvironments  	       7-26

7-12      Weighted Summary Statistics for CO Concentrations (ppm)
          in the Main Living Area by Use for Selected Sources
          by County   	       7-31

7-13      Summary of Continuous  CO Monitoring Results by
          Heating Equipment	       7-34

7-14      Peak CO Concentrations by Indoor Source Measured in
          Field Studies   	       7-35
                                        xiv

-------
                                      TABLES (cont'd)
Number                                                                     Page

7-15      Measured Concentrations of Carbon Monoxide in Environmental
          Tobacco Smoke	      7-40

8-1       Carbon Monoxide Concentrations in In-Transit
          Microenvironments - Denver, Colorado	      8-8

8-2       Carbon Monoxide Concentrations in Outdoor
          Microenvironments - Denver, Colorado	      8-9

8-3       Carbon Monoxide Concentrations in Indoor
          Microenvironments - Denver Colorado  	      8-10

8-4       Comparison of Different Approaches to Air Pollution Exposure
          Modeling	      8-24

8-5       Models Which Have Been Used to Estimate CO Exposure
          by Model Type	      8-25

8-6       Results of Weighted Linear Regression Analysis with Nontransit
          PEM Value as Dependent Variable and Simultaneous Value
          at Nearest Denver Fixed-Site as Independent Variable	      8-27

8-7       Results of Weighted Linear Regression Analyses with In-Transit
          PEM Value as Dependent Variable and Simultaneous Value
          from Denver Composite Data Set as Independent Variable   	      8-29

8-8       Results of Weighted Linear Regression Analyses with Nontransit
          PEM Value as Dependent Variable and Simultaneous Value at
          Nearest Fixed-Site in Washington, DC as Independent Variable  . .  .      8-30

8-9       Results of Weighted Linear Regression Analyses with In-Transit
          PEM Value as Dependent Variable and Simultaneous Value from
          Composite Washington, DC Data Set as Independent Variable  ....      8-31

8-10      Diagnostic Criteria for CO Intoxication  	      8-50

8-11      Comparison of Representative Methods for Analysis of
          Carbon Monoxide in Blood   	      8-63

8-12      Evaluation of the Ability  of Co-Oximeters to Measure Low Levels
          of COHB as Compared to Proposed Reference Methods   	      8-73
                                         xv

-------
                                      TABLES (cont'd)


Number                                                                      Page

8-13      Regression Parameters for the Relationship Between COHb
          and Eight-Hour CO Averages for 20 Cities  	       8-79

8-14      Summary of Studies Comparing End-Expired Breath CO with
          COHB Levels   	       8-83

9-1       In Vitro Inhibition Ratios for Hemoproteins that Bind
          Carbon Monoxide  	       9-24

10-1      Summary of Effects of Carbon Monoxide on Maximal and
          Submaximal Exercise Performance	       10-16

10-2      Summary of Effects of Carbon Monoxide Exposure in Patients
          with Angina	       10-22

10-3      Comparison of Subjects in Studies of the Effect of
          Carbon Monoxide Exposure on Occurrence of Angina
          During Exercise   	       10-30

10-4      Ventricular Fibrillation and Hemodynamic Studies in
          Laboratory Animals    	       10-44

10-5      Cardiac Hypertrophy Studies in Laboratory Animals  	       10-53

10-6      Hematology Studies in Laboratory Animals  	       10-59

10-7      Atherosclerotic Studies in Laboratory Animals   	       10-62

10-8      Brain Regions Ranked from Greatest to Least in Response
          to Hypoxia	       10-84

10-9      Effects of COHb on Absolute Visual Threshold	       10-106

10-10     Effects of COHb on Critical Flicker Fusion  	       10-108

10-11     Effects of COHb on Miscellaneous Visual Functions	       10-111
                                        xvi

-------
                               TABLES (cont'd)





Number
10-12
10-13
10-14
10-15
10-16
10-17
10-18
10-19
10-20
10-21
10-22
10-23
10-24
10-25
10-26
10-27
10-28

10-29

Effects of COHb on Miscellaneous Auditory Functions 	
Effects of COHb on Fine Motor Skills 	
Effects of COHb on Reaction Time 	
Effects of COHb on Tracking 	
Effects of COHb on Vigilance 	
Effects of COHb on Continuous Performance 	 ,
Effects of COHb on Time Estimation 	
Effects of COHb on Miscellaneous Cognitive Tasks 	 ,
Effects of COHb on Automobile Driving Tasks 	
Effects of COHb on Brain Electrical Activity 	
Effects of COHb on Schedule-Controlled Behavior 	
Effect of Blind Conditions 	
Effect of Statistical Methodology 	
Probability of Effects of COHb 	
Effect of Single vs. Multiple Task Performance 	 ,
Effect of Rate of COHb Formation 	
Teratogenic Consequences of Prenatal Carbon Monoxide
Exposure in Laboratory Animals 	
Consequences of Prenatal Carbon Monoxide Exposure on
Cardiovascular Development in Laboratory Rats 	
10-114
10-115
10-117
10-120
10-123
10-125
10-128
10-129
10-133
10-135
10-138
10-140
10-141
10-142
10-145
10-145

10-156

10-163
                                    XVll

-------
                                     TABLES (cont'd)


Number                                                                    Page

10-30     Neurobehavioral Consequences of Prenatal
          Carbon Monoxide Exposure in Laboratory Animals	      10-169

10-31     Consequences of Human Carbon Monoxide Intoxication During
          Early Development   	      10-171

10-32     Other Systemic Effects of Carbon Monoxide	      10-179

11-1      Calculated Equilibrium Values of Percent COHb and Percent
          02Hb in Humans Exposed to Ambient CO at Various
          Altitudes	      11-4

11-2      Summary of Effects of Carbon Monoxide at Altitude	      11-5

11-3      Chronic Effects of Altitude and Carbon Monoxide Exposure	      11-15

11-4      Combined Exposure to Carbon Monoxide and Other Pollutants ....      11-26

11-5      Combined Exposure to Carbon Monoxide and Combustion
          Products  	      11-31
                                       XVlll

-------
                                      FIGURES


Number

2-1        Relationship between carbon monoxide exposure and
           carboxyhemoglobin levels in the blood	         2-7

2-2        Currently accepted or proposed mechanisms of action of
           carbon monoxide resulting from external exposure sources
           can interfere with cellular respiration and cause
           tissue hypoxia   	         2-15

3-1        Effect of air-fuel ratio on exhaust gas carbon monoxide
           concentrations from three  gasoline-fueled test engines 	         3-13

4-1        The estimated sources of CO as a function of latitude 	         4-7

4-2        The global seasonal variations of CO	         4-11

4-3        The global concentrations  and trends of CO  	         4-13

5-1        Loss of carbon monoxide with time in mild steel cylinders  ....         5-7

5-2        Carbon monoxide monitoring system  	         5-10

5-3        Schematic diagram of gas  filter correlation (GFC) monitor
           for CO	         5-17

6-1        Estimated emissions of carbon monoxide from highway
           vehicles, 1970-1988	         6-9

6-2        CO pollution rose for St. Louis, MO	         6-18

6-3        National trend in the composite average of the second
           highest nonoverlapping 8-hour average carbon monoxide
           concentration 1979-1988  	         6-20

6-4        Boxplot comparisons of trends in second highest
           nonoverlapping 8-hour average carbon monoxide
           concentrations at 248 sites, 1979-1988	         6-20
                                         xix

-------
                                      FIGURES (cont'd)
Number                                                                        Page

6-5        National trend in the composite average of the estimated
           number of exceedances of the 8-hour carbon monoxide
           NAAQS, 1979-1988   	          6-21

6-6        Boxplot comparisons of trends in second highest
           nonoverlapping 8-hour average carbon monoxide
           concentrations at 359 sites,  1984-1988	          6-23

6-7        Regional comparisons of the 1986,  1987,  1988 composite
           averages of the second highest nonoverlapping 8-hour
           average carbon monoxide concentration	          6-24

6-8        United States map of the highest second maximum
           nonoverlapping 8-hour average carbon monoxide
           concentration by Metropolitan  Statistical Area
           for 1988  	          6-26

6-9        Effect of terrain roughness on the wind speed profile	          6-28

6-10       Schematic representation of an elevated inversion	          6-30

6-11       Hourly variations in inversion height and wind speed
           for Los Angeles in summer	          6-31

7-1        Cumulative frequency distributions  and summary statistics
           for indoor CO concentrations in three groups of
           monitored homes	          7-33

7-2        A time history of CO concentrations, 2-hour averages,
           winter of 1974   	          7-38

8-1        Frequency distributions of maximum eight-hour carbon monoxide
           population exposures and fixed-site monitor values in
           Denver,  CO and Washington, DC;  November 1982 -
           February 1983   	          8-6

8-2        Typical individual exposure as a function of time	          8-18

8-3        Logarithmic-probability plot of cumulative frequency
           distribution of maximum one-hour average exposure of
           CO predicted by SHAPE, plus an observed frequency
           distribution for Day 2 in Denver  	          8-37

                                         xx

-------
                                     FIGURES (cont'd)
Number
8-4        Logarithmic-probability plot of cumulative frequency
           distribution of maximum moving average eight-hour
           exposure of CO predicted by SHAPE, plus an observed
           frequency distribution for Day 2 in Denver	          8-38

8-5        Frequency distributions of carboxyhemoglobin levels
           in the U.S. population, by smoking habit	          8-78

8-6        Changes in alveolar CO of nonsmoking basement office
           workers compared to nonsmoking workers in other offices
           between Friday afternoon, Monday morning, and
           Monday afternoon  	          8-91

8-7        Eight-hour average CO concentrations in basement office
           before and after corrective action	          8-91

8-8        Distributions of CO in breath of adult nonsmokers in
           Denver and Washington	          8-94

8-9        Percent of Washington sample population with eight-hour
           average CO exposures exceeding the concentrations shown  ....          8-95

9-1        Oxyhemoglobin dissociation curves of normal human blood,
           of blood containing 50% carboxyhemoglobin, and of
           blood with a 50% normal hemoglobin concentration
           due to anemia	          9-7

9-2        Measured and predicted COHb concentrations from six
           intermittently exercising subjects  	          9-19

10-1       Relationship between carboxyhemoglobin level (COHb) and
           decrement in maximal oxygen uptake (VO2 max) for
           healthy nonsmokers  	          10-19

10-2       The effect of CO exposure on time to onset of angina	          10-32

10-3       Effect of hypoxic hypoxia and CO hypoxia on cerebral
           blood flow in 13 control and 9 chemodenervated dogs	          10-74
                                        xxi

-------
                                     FIGURES (cont'd)
Number

10-4       Effects of hypoxic and carbon monoxide hypoxia on cerebral
           blood flow, mean arterial blood pressure, and cerebral
           vascular resistance in control, carotid baroreceptor-,
           and chemoreceptor-denervated animals	

10-5       Effects of hypoxic and carbon monoxide hypoxia on cerebral
           blood flow, mean arterial blood pressure, and cerebral
           vascular resistance in control and vagotomized animals  ....

10-6       Cerebral blood flow as function of fractional arterial O2
           saturation   	

10-7       Comparison of newborn and adult responses of the reciprocal
           of the cerebral arteriovenous O2 content difference
           (CaO2 - QOj)-! to a reduction in arterial O2 content
           (C.O2) during hypoxic hypoxia (HH)	

10-8       Profiles of slopes of regional blood flow responses to
           hypoxic hypoxia (HH, solid lines) and CO hypoxia
           (COH, dashed lines) in adults (top) and newborns (bottom) . .

10-9       Effect of hypoxic hypoxia and carbon monoxide (CO) hypoxia
           on neurohypophyseal and regional cerebral blood flow (rCBF)

10-10      Effect of complete chemoreceptor denervation on total
           cerebral and neurohypophyseal blood flow	

10-11      Effect of increasing carboxyhemoglobin levels on cerebral
           blood flow, with special reference to low-level
           administration (below 20% COHb)	

10-12      Effect of cyanide (CN) and CO hypoxia, alone and in
           combination, on cerebral blood flow	

10-13      Effect of CN and CO hypoxia, alone and in combination,
           on cerebral oxygen consumption  	

10-14      Relationship of CBF to cerebral O2 consumption
           during CN and CO hypoxia
10-76
10-77
10-79
10-81



10-85


10-88


10-89



10-91


10-96


10-97


10-98
                                         xxu

-------
                                      FIGURES (cont'd)
Number                                                                       Pa;
11-1       Increment in percent carboxyhemoglobin (A% HbCO) over basal
           (control) levels at the end of a maximum aerobic
           capacity test and at the 5th min of recovery from a test
           in a typical  (A) male and (B) female subject	         11-9

11-2       Change in carboxyhemoglobin concentration (% COHb) during
           eight-hour exposures to 0 to  9 ppm CO for (A) resting and
           (B) exercising subjects	         11-11

11-3       The effects of altitude and ambient CO exposure on COHb
           in Fischer 344 rats	         11-13

11-4       Higher concentrations of COHb observed at the end of
           a five-minute recovery period after attainment of the
           subject's maximum aerobic capacity indicate that
           liberation of CO from tissue  stores is linearly
           related to COHb concentration present at exhaustion	         11-19
                                        XXlll

-------
                 AUTHORS, CONTRIBUTORS, AND REVIEWERS
CHAPTER 3.  PROPERTIES AND PRINCIPLES OF FORMATION OF CARBON
             MONOXIDE
Principal Authors

Dr. Marcia C. Dodge
Atmospheric Research and Exposure Assessment Laboratory (MD-84)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711

Dr. Harold G. Richter
Environmental Sciences
NSI Technology Services Corporation
P.O. Box 12313
Research Triangle Park, NC 27709
Contributors and Reviewers

Dr. Aubrey P. Altshuller
Atmospheric Research and Exposure Assessment Laboratory (MD-59)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711

Dr. Joseph J. Bufalini
Atmospheric Research and Exposure Assessment Laboratory (MD-84)
U.S Environmental Protection Agency
Research Triangle Park, NC 27711
CHAPTER 4. THE GLOBAL CYCLE OF CARBON MONOXIDE:  TRENDS AND MASS
            BALANCE
Principal Author

Dr. Aslam K. Khalil
Institute of Atmospheric Sciences
Oregon Graduate Center
19600 N.W. Von Neumann Drive
Beaverton, OR 97006
                                    xxv

-------
Contributors and Reviewers

Dr. Aubrey P. Altshuller
Atmospheric Research and Exposure Assessment Laboratory (MD-59)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711

Dr. Joseph J. Bufalini
Atmospheric Research and Exposure Assessment Laboratory (MD-84)
U.S Environmental Protection Agency
Research Triangle Park, NC 27711

Dr. Jack Fishman
Atmospheric Sciences Division
NASA - Langley Research Center
Mail Stop 401A
Hampton, VA 23665
CHAPTER 5. MEASUREMENT METHODS FOR CARBON MONOXIDE
Principal Author

Mr. Gerald G. Akland
Atmospheric Research and Exposure Assessment Laboratory (MD-75)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711

Contributors and Reviewers

Dr. William McClenny
Atmospheric Research and Exposure Assessment Laboratory (MD-44)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711

Dr. Joseph R. Stetter
Transducer Research, Inc.
1228 Olympus Drive
Naperville, IL 60540
                                      xxvi

-------
 CHAPTER 6.  AMBIENT SOURCES, EMISSIONS, AND CONCENTRATIONS
 Principal Authors

 Dr. James N. Braddock
 Atmospheric Research and Exposure Assessment Laboratory (MD-46)
 U.S. Environmental Protection Agency
 Research Triangle Park, NC 27711

 Mr. Thomas N. Braverman
 Office of Air Quality Planning and Standards (MD-14)
 U.S. Environmental Protection Agency
 Research Triangle Park, NC 27711

 Mr. William B. Petersen
 Atmospheric Research and Exposure Assessment Laboratory (MD-80)
 U.S. Environmental Protection Agency
 Research Triangle Park, NC 27711

 Mr. Thomas B. McMullen
 Environmental Criteria and Assessment Office (MD-52)
 U.S. Environmental Protection Agency
 Research Triangle Park, NC 27711

 Contributors and Reviewers

 Mr. Paul Benson
 Transportation Laboratory
 5900 Folsom Boulevard
 Sacramento, CA 95819

Dr. Thomas C. Curran
Office of Air Quality Planning and Standards (MD-14)
U.S.  Environmental Protection Agency
Research Triangle Park, NC  27711

Mr. Tom Hansen
U.S.  Environmental Protection Agency
(4AT Air Program Branch)
345 Courtland Street, NE
Atlanta,  GA 30365
                                     XXVll

-------
Mr. George Schewe
PEI Associates, Inc.
11499 Chester Road
Cincinnati, OH 45246-0100
CHAPTER 7.  INDOOR SOURCES, EMISSIONS, AND CONCENTRATIONS
Principal Author

Dr. Brian Leaderer
John B. Pierce Foundation Laboratory
290 Congress Avenue
New Haven, CT 06519

Contributors and Reviewers

Dr. Irwin H. Billick
Gas Research Institute
8600 West Bryn Mawr Avenue
Chicago, IL 60631

Mr. Tom  Hansen
U.S. Environmental Protection Agency
(4AT Air Program Branch)
345 Courtland Street, NE
Atlanta, GA 30365

Dr. P. Barry Ryan
Department of Environmental Sciences and Physics
Harvard University School of Public Health
665 Huntington Avenue
Boston, MA 02115
CHAPTER 8. POPULATION EXPOSURE TO CARBON MONOXIDE
Principal Authors

Mr. Gerald G. Akland
Atmospheric Research and Exposure Assessment Laboratory (MD-75)
U.S. Environmental Protection Agency
Research Triangle Park, NC  27711
                                     XXVlll

-------
Dr. Steven D. Colome
Integrated Environmental Services
University Tower, Suite 1090
4199 Campus Drive
Irvine, CA  92715

Dr. Thomas E. Dahms
Department of Anesthesiology
St. Louis University School of Medicine
3635 Vista Avenue
St. Louis, MO  63110

Mr. Ted Johnson
PEI Associates, Inc.
505 S. Duke Street
Durham,  NC 27701

Dr. Brian Leaderer
John B. Pierce Foundation Laboratory
290 Congress Avenue
New Haven, CT  06519

Dr. Wayne  Ott
Office of Research and Development (RD-680)
U.S. Environmental Protection Agency
Washington, DC  20460

Dr. Lance Wallace
U.S. Environmental Protection Agency
Building  166
Vint Hill  Farms Station
Bicher Road
Warrenton,  VA  22186

Contributors and Reviewers

Dr. William F. Biller
68 Yorktown Road
East Brunswick, NJ 08816

Mr. N. O. Gerald
Office of Air Quality Planning and Standards (MD-14)
U.S. Environmental Protection Agency
Research  Triangle Park, NC 27711
                                       XXIX

-------
Dr. Steven M. Horvath
Environmental Stress Laboratory
Neuroscience Research Institute
University of California
Santa Barbara, CA  93106

Mr. Thomas R. McCurdy
Office of Air Quality Planning and Standards (MD-12)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711

Mr. Harvey M. Richmond
Office of Air Quality Planning and Standards (MD-12)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711

Dr. P. Barry Ryan
Department of Environmental Sciences and Physics
Harvard University School of Public Health
665 Huntington Avenue
Boston, MA  02115

Dr. Carr J. Smith
Bowman  Gray Technical Center
R.J. Reynolds Tobacco Company
Winston-Salem, NC 27102

Dr. John Spengler
Harvard School of Public Health
665 Huntington Avenue
Boston, MA  02115

Dr. David K. Stevenson
Department of Pediatrics
Laboratory for Neonatal Metabolism
Stanford University School of Medicine
Stanford, CA 94305

Dr. HenkJ. Vreman
Laboratory for Neonatal Metabolism
Department of Pediatrics (S214)
Stanford University School of Medicine
Stanford, CA 94305
                                        XXX

-------
CHAPTER 9. PHARMACOKINETICS AND MECHANISMS OF ACTION OF CARBON
             MONOXIDE
Principal Authors

Dr. Milan J. Hazucha
Pulmonary Division
Department of Medicine
Center for Environmental Medicine and Lung Biology
The University of North Carolina
Trailer #4, Medical Building C 224H
Chapel Hill, NC 27599

Dr. Marjolein V. Smith
2501 Anne Carol Court
Raleigh, NC  27603

Dr. Claude A. Piantadosi
Division of Allergy, Critical Care and Respiratory Medicine
Department of Medicine, Box 3315
Duke University Medical Center
Durham, NC 27710

Contributors and Reviewers

Dr. Clyde H. Barlow
Laboratory I
Evergreen State College
Olympia, WA 98505

Dr. William F.  Biller
68 Yorktown Road
East Brunswick, NJ 08816

Dr. Henry J.  Forman
Department of Pediatrics
Cell Biology Group (Box 83)
University of Southern California
Childrens Hospital of Los Angeles
4650 Sunset Boulevard
Los Angeles, CA  90054
                                      xxxi

-------
Dr. Steven M, Horvath
Environmental Stress Laboratory
Neuroscience Research Institute
University of California
Santa Barbara, CA 93106

Dr. Robert Jensen
LDS Hospital
University of Utah
Salt Lake City, UT  84143

Maj. David Farmer
U.S. Army
Biomedical Research and Development Lab
Fort Detrick, Building 568
Frederick, MD 21701

Dr. Peter Tikuisis
Defence and Civil Institute of Environmental Medicine
1133 Sheppard Avenue, W
Downsview, Ontario CANADA M3M3B9
CHAPTER 10. HEALTH EFFECTS OF CARBON MONOXIDE
Principal Authors

Dr. Vernon A. Benignus
Health Effects Research Laboratory (MD-58)
U.S. Environmental Protection Agency
Research Triangle Park, NC  27711

Dr. Lars-Goran Ekelund
Department of Biostatistics
School of Public Health
The University of North Carolina
Suite 203, NCNB Plaza 322A
Chapel Hill, NC 27514
                                      XXXll

-------
Dr. Laurence D. Fechter
Department of Environmental Health Sciences
The Johns Hopkins University School of Hygiene
 and Public Health
615 N. Wolfe Street
Baltimore, MD  21205

Dr. Thomas R.  Griggs
Division of Cardiology
School of Medicine
The University of North Carolina
349 Clinical Sciences Building 229H
Chapel Hill, NC 27599

Dr. Steven M. Horvath
Environmental Stress Laboratory
Neuroscience Research Institute
University of California
Santa Barbara, CA  93106

Dr. James J. McGrath
Department of Physiology
School of Medicine
Texas Tech University Health Sciences Center
Lubbock, TX 79430

Mr. James A. Raub
Environmental Criteria and Assessment Office (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711

Dr. David S.  Sheps
Division of Cardiology
School of Medicine
The University of North Carolina
338 Clinical Sciences Building 229H
Chapel Hill, NC 27599

Dr. Richard J. Traystman
Department of Anesthesiology and Critical Care Medicine
The Johns Hopkins Medical Institutions
600 N. Wolfe Street
Baltimore, MD  21205
                                       XXXlll

-------
Contributors and Reviewers

Dr. Zoltan Annau
Department of Environmental Health Sciences
School of Hygiene and Public Health
The Johns Hopkins University
615 N. Wolfe Street
Baltimore, MD 21205

Dr. Steven M. Ayres
Medical College of Virginia, Box 565
Virginia Commonwealth University
Richmond, VA 23298

Dr. Robert L. Balster
Department of Pharmacology and Toxicology
Medical College of Virginia
Virginia Commonwealth University
Richmond, VA  23298

Dr. Thomas Clarkson
Department of Comparative Medicine
Bowman Gray School of Medicine
300 South Hawthorne Road
Winston-Salem, NC  27103

Dr. Steven D. Colome
Integrated Environmental Services
University Tower, Suite 1090
4199 Campus Drive
Irvine, CA 92715

Dr. Thomas E. Dahms
Department of Anesthesiology
St. Louis University School of Medicine
3635 Vista Avenue
St. Louis, MO  63110

Dr. Samuel Fox
Georgetown University
3800 Reservoir Road, NW
Washington, DC  20007
                                       XXXIV

-------
Dr. Donald Heistad
Cardiology Division
Department of Internal Medicine
College of Medicine
University of Iowa
Iowa City, IA  52242

Dr. Gregory L. Hirsch
Escondido Pulmonary Medical Group
215 South Hickory
Escondido, CA 92025

Dr. John R. Holmes
State of California
Air Resources Board
1102 Q Street
P.O. Box 2815
Sacramento,  CA 95812

Dr. Steven M.  Horvath
Environmental  Stress Laboratory
Neuroscience Research Institute
University of California
Santa Barbara,  CA 93106

Dr. Michael T.  Kleinman
Department of  Community and Environmental Medicine
California College of Medicine
University of California
Irvine, CA 92717

Dr. Victor G. Laties
Department of Radiation  Biology and Biophysics
School of Medicine and Dentistry
University of Rochester
Rochester, NY  14642

Dr. Lawrence D. Longo
School of Medicine
Department of Physiology
Division of Perinatal Biology
Loma Linda University
Loma Linda, CA 92350
                                        xxxv

-------
Dr. George Malindzak
NIEHS North Campus (MD-3-03)
104 Alexander Drive
Research Triangle Park, NC 27709

Dr. Steve McFaul
Letterman Army Institute of Research
Presidio in San Francisco
Blood Research Division
San Francisco, CA  94129

Dr. James J. McGrath
Department of Physiology
School of Medicine
Texas Tech University Health Sciences Center
Lubbock, TX 79430

Dr. Fathy Messiha
Department of Pharmacology
University of North Dakota
Medical Science North 501
North Columbia Road
Grand Forks, ND  58201

Maj. David Farmer
U.S. Army
Biomedical Research and Development Lab
Fort Detrick, Building 568
Frederick, MD 21701

Dr. David G. Penney
Department of Physiology
Wayne State University School of Medicine
Scott Hall, Room 5374
540 East Canfield
Detroit,  MI  48201

Mr. Harvey M. Richmond
Office of Air Quality Planning and Standards (MD-12)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
                                       XXXVI

-------
Dr. Carr J. Smith
Bowman Gray Technical Center
RJ. Reynolds Tobacco Company
Winston-Salem, NC  27102

Dr. David K. Stevenson
Department of Pediatrics
Laboratory for Neonatal Metabolism
Stanford University School of Medicine
Stanford, CA 94305

Dr. Jane Warren
Health Effects Institute
215 First Street
Cambridge, MA 02142

Dr. Robert Winslow
Combat Trauma Management Directorate
Blood Research Division
Letterman Army Institute of Research
Presidio in San Francisco
San Francisco, CA 94129-6800

Dr. Ronald W. Wood
Department of Environmental Medicine
New York University Medical Center
Lanza Laboratory
Long Meadow Road
Tuxedo, NY 10987

Dr. Henk J. Vreman
Laboratory for Neonatal Metabolism
Department of Pediatrics (S214)
Stanford University School of Medicine
Stanford, CA 94305
                                      XXXVll

-------
CHAPTER 11. COMBINED EXPOSURE OF CARBON MONOXIDE WITH OTHER
              POLLUTANTS, DRUGS, AND ENVIRONMENTAL FACTORS
Principal Authors

Dr. Robert L. Balster
Department of Pharmacology and Toxicology
Medical College of Virginia
Virginia Commonwealth University
Richmond, VA  23298

Dr. Steven M. Horvath
Environmental Stress Laboratory
Neuroscience Research Institute
University of California
Santa Barbara, CA 93106

Dr. James J. McGrath
Department of Physiology
School of Medicine
Texas Tech University Health Sciences Center
Lubbock,  TX 79430

Mr. James A. Raub
Environmental Criteria and Assessment Office (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park,  NC  27711

Contributors and Reviewers

Dr. Zoltan Annau
Department of Environmental Health Sciences
School of Hygiene and Public Health
The Johns Hopkins University
615 N. Wolfe Street
Baltimore, MD  21205

Dr. Thomas E.  Dahms
Department of Anesthesiology
St. Louis  University School of Medicine
3635 Vista Avenue
St. Louis, MO  63110

Dr. Robert F. Grover
216 Mariposa Circle
Arroyo Grande, CA 93420

                                    xxxviii

-------
Dr. Donald H. Horstman
Health Effects Research Laboratory (MD-58)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711

Dr. Victor G. Laties
Department of Radiation Biology and Biophysics
School of Medicine and Dentistry
University of Rochester
Rochester, NY  14642

Dr. Barbara C. Levin
Center for Fire Research
National Institute for Standards and Technology
Building 224, Room A363
Gaithersburg, MD  20899

Dr. Lawrence D. Longo
School of Medicine
Department of Physiology
Division of Perinatal Biology
Loma Linda University
Loma Linda, CA  92350

Dr. Robert Winslow
Combat Trauma Management Directorate
Blood Research Division
Letterman Army Institute of Research
Presidio in San Francisco
San Francisco, CA 94129-6800

Dr. Ronald W. Wood
Department of Environmental Medicine
New York University Medical Center
Lanza Laboratory
Long Meadow Road
Tuxedo, NY 10987
                                       XXXIX

-------
CHAPTER 12. EVALUATION OF SUBPOPULATIONS POTENTIALLY AT RISK
Principal Authors

Dr. Robert L. Balster
Department of Pharmacology and Toxicology
Medical College of Virginia
Virginia Commonwealth University
Richmond, VA 23298

Dr. Lars-Goran Ekelund
Department of Biostatistics
School of Public Health
The University of North Carolina
Suite 203, NCNB Plaza 322A
Chapel Hill, NC 27514

Dr. Robert F. Grover
216 Mariposa Circle
Arroyo Grande, CA 93420

Dr. Steven M. Horvath
Environmental Stress Laboratory
Neuroscience Research Institute
University of California
Santa Barbara, CA 93106

Dr. David S. Sheps
Division of Cardiology
School of Medicine
The University of North Carolina
338 Clinical Sciences Building 229H
Chapel Hill, NC 27599

Contributors and Reviewers

Dr. Zoltan Annau
Department of Environmental Health Sciences
School of Hygiene and Public Health
The Johns Hopkins University
615 N. Wolfe Street
Baltimore, MD 21205
                                        xl

-------
Dr. Thomas Clarkson
Department of Comparative Medicine
Bowman Gray School of Medicine
300 South Hawthorne Road
Winston-Salem, NC 27103

Dr. Thomas E. Dahms
Department of Anesthesiology
St. Louis University School of Medicine
3635 Vista Avenue
St. Louis, MO 63110

Dr. Samuel Fox
Georgetown University
3800 Reservoir Road, NW
Washington, DC 20007

Dr. Carr J. Smith
Bowman Gray Technical Center
RJ. Reynolds Tobacco Company
Winston-Salem, NC 27102

Dr. Ronald W. Wood
Department of Environmental Medicine
New York University Medical Center
Lanza Laboratory
Long Meadow Road
Tuxedo, NY  10987
                                       xli

-------
                    PROJECT TEAM FOR DEVELOPMENT OF
               AIR QUALITY CRITERIA FOR CARBON MONOXIDE
Scientific Staff

Mr. James A. Raub, Project Manager and Health Effects Coordinator
Environmental Criteria and Assessment Office (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711

Mr. Thomas B. McMullen, Air Quality Coordinator
Environmental Criteria and Assessment Office (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711

Ms. Ellie Speh, Secretary
Environmental Criteria and Assessment Office (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711

Technical Support Staff

Ms. Frances P. Bradow, Support and Operations Management
Environmental Criteria and Assessment Office (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711

Mr. Douglas B. Fennell, Technical Information Retrieval
Environmental Criteria and Assessment Office (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711

Mr. Allen G. Hoyt, Technical Editing and Graphics
Environmental Criteria and Assessment Office (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711

Ms. Diane H. Ray, Docket Information (Public Comments)
Environmental Criteria and Assessment Office (MD-52)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
                                      xliii

-------
                      1. SUMMARY AND CONCLUSIONS
            Carbon monoxide (CO) is a colorless, odorless gas. It is a trace constituent of the
 5     troposphere, produced by both natural processes and human activities. The major source of
       CO in urban areas is incompletely combusted fuels; two-thirds of total emissions are
       attributed to vehicle exhaust.  The principal effect on humans  inhaling CO-contaminated air is
       reduced transport of oxygen by the blood stream, a consequence of CO displacing oxygen in
       hemoglobin.
10
       The Global Cycle of Carbon Monoxide
            Limited data on global trends in tropospheric CO concentrations indicate a 1 to 2%
       annual increase over the last several decades.  Global background concentrations fall in the
       range of 50 to  120 ppb; higher levels are found in the northern hemisphere, lower levels in
15     the southern hemisphere. Average background concentrations fluctuate seasonally; higher
       levels occur in the winter months, lower levels in the summer months.  About 60%  of the CO
       in the non-urban troposphere is attributed to human activities,  both directly from combustion
       processes, and indirectly through the oxidation of hydrocarbons and ammonia that, in turn,
       arise from agricultural activities, landfills, etc.  Atmospheric reactions involving CO can
20     produce O3 in the troposphere. Other reactions may deplete concentrations of the hydroxyl
       radical (OH*),  a key participant in the global removal cycles of many other natural and
       anthropogenic trace gases, thus possibly contributing to changes in atmospheric chemistry
       and, ultimately, global climate change.

25     Measurement Methods for Carbon Monoxide
            The non-dispersive infrared (NDIR) optical transmission technique, the  technique on
       which the EPA designated reference analytical method is based, is the only technique being
       used for compliance monitoring of CO.  One category of NDIR monitor, the gas filter
       correlation (GFC) monitor, is currently the single most widely used NDIR-type analyzer for
30     fixed-site monitoring stations.  In general, NDIR monitors have significant advantages
       including small size, good sensitivity and specificity for CO, and reliability of operation under

       March 14, 1990                          1-1     DRAFT - DO NOT QUOTE OR CITE

-------
       typical network monitoring conditions.  An associated recorder compiles and stores hourly
       averages for subsequent computer storage and analysis.
            The recent development of small, portable electrochemical monitors has made possible
       the measurement of CO concentrations incurred by individuals as they move through
 5     numerous diverse indoor and outdoor microenvironments that cannot be monitored by
       fixed-site ambient stations. Such measurements show that some individuals can receive
       exposures significantly higher than would be inferred from a simple interpretation of data
       from local fixed-site stations.

10     Ambient Sources. Emissions and Concentrations of Carbon Monoxide
            Current air quality standards define 1-hour and 8-hour average concentrations that
       should not be exceeded more than once per year.  The 1-hour standard of 35 ppm is almost
       never exceeded in data reported from fixed-site monitoring stations.  Concentrations of CO
       exceeding the 8-hour standard of 9 ppm have declined over the 10-year period  1979 to 1988
15     from an average of about ten per station per year to about two per year (U.S. Environmental
       Protection Agency, 1990). This decline reflects the efficacy of emission control systems on
       newer vehicles. In 1979, vehicular emissions of CO  accounted for about 72%  of total U. S.
       emissions; in 1988, it was  67%.  During this same period, there was a 33% increase in
       highway vehicle miles traveled. The other categories of CO emissions are other fuel
20     combustion sources, such as steam boilers (12%), industrial processes (8%), Solid waste
       disposal (3%), and miscellaneous other sources (10%).

       Indoor Sources. Emissions, and Concentrations of Carbon Monoxide
            EPA's mandate is to monitor and regulate pollutants in the ambient, i.e. outdoor, air;
25     however, the great majority of people spend most of  their time indoors, one place or another.
       A realistic assessment of ambient exposure to CO, therefore, must be set in the context of
       total exposure, a major component of which is exposure while indoors.
            Indoor concentrations of CO are a function of outdoor concentrations, indoor sources,
       infiltration, ventilation and air mixing between  and within rooms. In residences without
30     sources, average CO concentrations are approximately equal to average outdoor levels.
       March 14, 1990                          1-2     DRAFT - DO NOT QUOTE OR CITE

-------
       Proximity to outdoor sources (e.g., heavily traveled roadways, attached garages, or parking
       garages) can have a major impact on indoor CO concentrations.
            The extensive total personal CO exposure studies conducted by EPA in Washington, DC
       and Denver, CO (Akland et al., 1985; Whitmore et al., 1984; Hartwell et al., 1984; Johnson,
 5     1984) have shown  that the highest CO concentrations occur in indoor microenvironments
       associated with transportation sources. Concentrations in these environments frequently were
       found to exceed 9 ppm.  Studies targeted toward specific indoor microenvironments have also
       identified the indoor commuting microenvironment as one in which CO concentrations
       frequently exceed 9 ppm and occasionally exceed 35 ppm.  Similar concentrations have been
10     reported in special environments or accompanying unusual occurrences (indoor ice skating
       rinks, offices where emissions from parking garages migrate indoors, etc.).
            Unvented gas and kerosene space heaters that are used for substantial periods of time
       appear to be the major contributors to residential CO concentrations. Peak concentrations of
       CO in such  residences often exceed an 8-hour average of 9 ppm and a 1-hour average of
15     35 ppm. One extensive study (Koontz and Nagda, 1988) of unvented gas space heaters
       indicated that 12 % of the homes had 15-hour average CO concentrations greater than 8 ppm;
       the highest recorded concentration was 36.6 ppm.
            Intermittent sources such as gas cooking ranges can result in high peak CO
       concentrations (in excess of 9 ppm), while long-term average concentrations (i.e., 24 hours)
20     associated with gas ranges are on the order of 1 ppm. The contribution of tobacco
       combustion  to indoor CO levels is variable.  One study suggested that its contribution to
       residential CO concentrations was on the order of 1  ppm; another study showed no significant
       increase.
            Very limited  data on CO levels in residences with wood burning stoves or fireplaces
25     indicate non-airtight stoves can contribute substantially to residential CO concentrations;
       airtight stoves can  contribute small increases. The available data indicate that fireplaces do
       not contribute measurably to average indoor concentrations.  No information is available on
       residences with leaky flues or on the impact of attached garages.
            The available data on the spatial and temporal  variability of indoor CO concentrations as
30     a function of microenvironments and associated sources are not adequate to assess exposures
       March 14, 1990                           1-3      DRAFT - DO NOT QUOTE OR CITE

-------
       in these environments. These indoor microenvironments represent the most important CO
       exposures for the majority of individuals and as such need to be better characterized.

       Population Exposure to Carbon Monoxide
 5          The current NAAQS for CO (9 ppm for 8 hours, 35  ppm for 1 hour) are designed to
       protect against actual and potential human exposures in outdoor air that would cause adverse
       health effects. Compliance with the NAAQS is determined by measurements taken at
       fixed-site ambient monitors, the use of which is intended to provide some measure of the
       general level of exposure of the population represented by the CO monitors.  Results of both
10     exposure monitoring in the field, and modeling studies, summarized in this document indicate
       that individual personal exposure does not directly correlate with CO concentrations
       determined by using fixed-site monitors alone. This observation is due to the mobility of
       people and to the spatial and temporal variability of CO concentrations. While  failing to
       show a correlation between individual personal monitor exposures and simultaneous nearest
15     fixed-site monitor concentrations, studies do suggest that aggregate personal exposures are
       lower  on days of lower ambient CO levels as determined by the fixed-site monitors and
       higher on days of higher ambient levels.
            A unique feature of carbon monoxide exposure is that there is a biological marker of the
       dose that the individual has received: the blood level of CO. This level may  be calculated by
20     measuring blood carboxyhemoglobin (COHb) or by measuring CO in exhaled breath.
            The use of CO-Oximeters to measure low levels of COHb can provide useful
       information regarding mean values, provided a reference technique is used to properly
       calibrate the instrument.  It has been shown,  however, that the range of values obtained with
       this optical method will be greater than that obtained with a reference method.  For example,
25     in a group of subjects with cardiovascular disease, the standard deviation of the percent
       COHb values for non-smoking, resting  subjects was 2 to 2.5 times greater for the CO-
       Oximeter values than for the gas chromatograph  values on paired samples (Allred et al.,
       1989b).  Therefore, the potential exists with the  CO-Oximeter for having an incorrect
       absolute  value for COHb as well as an incorrectly broadened range of values.
30          In addition, it is not clear exactly  how sensitive the CO-Oximeter techniques are to
       small changes in COHb at the low end of the CO dissociation curve.  Allred  et al. (1989b)

       March 14, 1990                           1-4     DRAFT - DO NOT QUOTE OR CITE

-------
       have noted that the interference from changing O2 saturation can have a very significant
       influence on the apparent COHb reading in a sample. The interaction between hemoglobin
       species was also reported by Dennis and Valeri (1980).  This suggests nonlinearity or a
       disproportionality in the absorption spectrum of these two species of hemoglobin. It is also a
 5     potential source of considerable error in the estimation of COHb by optical methods.
            The measurement of exhaled breath has the advantages of ease, speed, precision,  and
       greater subject acceptance than measurement of blood COHb.  However, the accuracy of the
       breath measurement procedure and the validity of the Haldane relationship between breath and
       blood at  low environmental CO concentrations remains in question.  There appears to be a
10     clear research need to validate the breath  method at low CO exposures.  In view of the
       possible problems with the CO-Oximeter, such validation should be done using gas
       chromatography for the blood COHb measurements.
            Cigarette consumption represents a special case of CO exposure; for the smoker it
       almost always dominates over personal exposure from other sources.  Studies by Radford and
15     Drizd (1982) show that COHb levels of cigarette smokers average 4% while those of
       nonsmokers average 1 %. Therefore, this document  focuses on environmental  exposure of
       nonsmokers to CO.
            People encounter  CO in  a variety of environments that include travelling in motor
       vehicles, working at their jobs, visiting urban locations associated with combustion sources,
20     or cooking over a gas range.  Studies of human exposure have shown that among these
       settings the motor vehicle is the most important for regularly encountered  elevations of CO.
       Studies by Flachsbart et al. (1987) indicated that CO exposures while commuting in
       Washington, D.C. average 9 to 14 ppm at the same  time that fixed station monitors record
       concentrations of 2.7 to 3.1 ppm.  Similar studies conducted by EPA in Denver and
25     Washington, D.C., have demonstrated that the motor vehicle interior has the highest average
       CO concentrations (averaging  7 to 10 ppm) of all microenvironments (Johnson, 1984).  In
       these studies, 8% of all commuters experienced 8-hour exposures greater than  9 ppm while
       only 1 %  of noncommuters received exposures over that level.  Furthermore, commuting
       exposures have  been shown to be highly variable with some commuters breathing CO in
30     excess of 25 ppm.
       March 14,  1990                          1-5     DRAFT - DO NOT QUOTE OR CITE

-------
            Another important setting for CO exposure is the workplace.  In general, exposures at
       work exceed CO exposures during nonwork periods, apart from commuting to and from
       work.  Average concentrations may be elevated during this period since workplaces are often
       located in congested areas that have higher background CO concentrations than do many
 5     residential neighborhoods.  Occupational and nonoccupational exposures may overlay one
       another and result in a higher concentration of CO in the blood.  Certain occupations also
       increase the risk of high CO  exposure (e.g., those occupations involved directly with vehicle
       driving, maintenance, and parking).  Occupational groups exposed to CO by vehicle exhaust
       include auto mechanics; parking garage and gas station attendants; bus,  truck or taxi drivers;
10     police; and warehouse workers. Other industrial processes produce CO directly or as a by-
       product, including steel production, coke ovens, carbon black production, and petroleum
       refining.   Firefighters, cooks, and construction workers also may be exposed at work to
       higher CO levels. Occupational exposure in industries or setting with CO production also
       represent some of the highest individual exposures observed in field monitoring studies.  For
15     example, in EPA's CO exposure study in Washington, DC,  of the approximately  4% (29 of
       712) of subjects working in jobs classified as having a high potential for CO exposure, seven
       subjects (or approximately 25%) experienced 8-hour CO exposures in excess of 9 ppm.
            The  highest indoor nonoccupational CO exposures are associated with combustion
       sources and include enclosed parking garages, service stations, restaurants and stores.  The
20     lowest indoor CO concentrations are found in homes, churches, and health care facilities.
       EPA's Denver Study showed that passive cigarette smoke is associated with increasing a
       nonsmoker's exposure by an  average of about 1.5 ppm and that use of a gas range is
       associated with about 2.5 ppm increase at home. Other sources which may contribute to CO
       in the home include combustion space heaters and wood burning stoves.
25          As noted above, people encounter different and often higher exposures than  predicted
       from fixed-site monitoring data because of the highly localized nature of CO sources.  For
       example, during the winter sampling period, 10% of Denver volunteers and 4% of
       Washington, DC volunteers recorded personal exposures in excess of 9 ppm for 8 hours.
       Breath measurements from the Washington volunteers indicated that as much as 9% of the
30     population could have experienced a 9 ppm, 8-hour average. In contrast, during the entire
       winter period of 1982-1983,  the two ambient CO monitors in Washington reported only one

       March 14, 1990                          1-6     DRAFT - DO NOT QUOTE OR CITE

-------
       exceedance of the 9-ppm level.  In another study, using data from analyses of COHb in
       blood, Wallace and Ziegenfus (1985) report that CO in blood is uncorrelated with CO
       measured by ambient monitors.  These findings point out the necessity of having personal CO
       measurements to augment fixed-site ambient monitoring data when total human exposure is to
 5     be evaluated.  Data from these field  studies can be used to construct and test models of
       human exposure that account for time and activity patterns known to affect exposure to CO.
       Models developed to date tend to underpredict the variability of CO exposures observed in
       field studies and have not been able to successfully predict individual exposures.  The models
       may be modified and adjusted using  information from field monitoring studies in order to
10     capture the observed distribution of CO exposures, including the higher exposures found in
       the tail of the exposure distribution.  The models also are useful for evaluating alternative
       pollutant control strategies.

       Pharmacokinetics and Mechanisms of Action of Carbon Monoxide
15          This section of the document reviews the basic relationships of O2 and CO with Hb and
       other O2 binding proteins, outlines the fundamentals of CO uptake and elimination from blood
       and extravascular space, examines the mechanisms of CO toxicity, and evaluates the models
       of COHb formation.
            The exchange of CO between the atmosphere (airway opening) and RBC is controlled
20     by physical (mass transport, diffusion) and physiological (alveolar ventilation, cardiac output,
       etc.) processes.  The final step of  the competitive binding between CO and O2 to Hb to form
       COHb and O2Hb, respectively, is a complex sequence of reversible reactions, the kinetics of
       which are still not fully understood.  The toxic effects of CO are due to its high affinity for
       Hb. The presence of CO reduces  O2-carrying capacity of blood and impairs release of O2
25     from O2Hb to  extravascular tissues.  Brain and heart tissues are particularly sensitive to the
       resultant drop  in PO2 and CO hypoxia.  Because of tight binding of CO to Hb, the elimination
       half-time is quite long.  This might lead to accumulation of COHb and even relatively low
       concentrations of CO might produce substantial blood levels of COHb.
            The rate of rise of blood COHb levels as well as the  time to equilibrium saturation
30     depends on the inhaled concentration of CO, alveolar ventilation, lung diffusion, cardiac
       March 14, 1990                           1-7     DRAFT - DO NOT QUOTE OR CITE

-------
       output, and duration of exposure.  These factors have been integrated into many empirical and
       mathematical models of COHb formation under static and dynamic conditions.
            The best all around model for COHb prediction is still the equation developed by
       Coburn, Forster, and Kane (1965). The linear solution is useful for examining air pollution
 5     data leading to relatively low COHb levels, whereas the nonlinear solution shows good
       predictive power even for high CO exposures. The two regression models might be useful
       only when the conditions of application closely approximate those under which the parameters
       were estimated.
            It is important to remember that almost all of the available modeling studies assumed a
10     constant rate of CO uptake and elimination, which is rarely true.  A number of physiological
       factors, particularly changes in ventilation, will affect both rates.  The predicted COHb values
       also will differ from individual to individual due to smoking, age, or lung disease.  There
       does not appear to be a single optimal averaging time period for ambient CO; however, the
       shorter the period the greater the precision.  In general, the averaging time period should be
15     well within the [COHb] half-life, which decreases with increased activity.
            Although the principal cause of CO toxicity is tissue hypoxia due to CO binding to Hb,
       certain physiological aspects of CO exposure are not  explained well by decreases in
       intracellular PO2 related to the presence of COHb. For many years, it has been known that
       CO is distributed to extravascular sites such as skeletal muscle (Coburn et al., 1971; Coburn
20     et al.,  1973) and that 10 to 50% of the total body store of CO may be extravascular
       (Luomanmaki and Coburn, 1969).  Furthermore, extravascular CO is metabolized slowly  to
       CO2 in vivo (Fenn, 1970).  Consequently, secondary mechanisms  of CO toxicity related to
       intracellular uptake of CO have been the focus of a great deal of research interest.  CO
       binding to many intracellular compounds has been well documented both in vitro and in vivo;
25     however, it is still uncertain whether or not intracellular uptake of CO in the presence of Hb
       is sufficient to cause either acute organ system dysfunction or long-term health effects.  The
       virtual absence of  sensitive techniques capable of assessing intracellular CO binding under
       physiological conditions has resulted in a variety of indirect approaches to the problem as well
       as many negative studies.
30          Current knowledge pertaining to intracellular CO-binding proteins suggests that the most
       likely ones to be inhibited functionally at relevant levels of COHb are myoglobin (Mb), found

       March 14, 1990                            1-8     DRAFT -  DO NOT QUOTE OR CITE

-------
       predominantly in heart and skeletal muscle, and cytochrome oxidase.  The physiological
       significance of CO uptake by Mb is uncertain at this time but sufficient concentrations of
       carboxymyoglobin (COMb) could potentially limit maximal O2 uptake of exercising muscle.
       Although there is suggestive evidence for significant binding of CO to cytochrome oxidase in
 5     heart and brain tissue, it is unlikely that any significant CO binding would occur at low
       COHb levels.  Therefore,  further research is needed to determine if secondary, intracellular
       mechanisms will occur at exposure concentrations found in ambient air.

       Health Effects of Carbon Monoxide
10          Concerns about the potential health effects of exposure to carbon  monoxide have been
       addressed in extensive studies with various animal species as subjects.  Under varied
       experimental protocols, considerable information has been obtained on  the toxicity of CO, its
       direct effects on  the blood and other tissues, and the manifestations of these effects in the
       form of changes in organ function.  Many of these studies, however, have been conducted at
15     extremely high levels of CO (i.e., levels not found in ambient air). Although  severe effects
       from exposure to these high levels of CO are not directly germane to the problems from
       exposure to current ambient levels of CO, they can provide valuable information about
       potential effects of accidental exposure to CO, particularly those exposures occurring indoors.

20     Acute Pulmonary Effects
            Currently available studies on the effects of CO exposures producing COHb
       concentrations of up to 39%  fail to find any consistent effects on lung parenchyma and
       vasculature (Hugod, 1980; Fisher et al.,  1969) or on alveolar macrophages (Chen et al.,
       1982; Weissbecker et al., 1969). The lack of significant changes in lung tissue is consistent
25     with the  lack of histologic changes in the pulmonary and coronary arteries. Alveolar
       epithelial permeability to 51Cr-EDTA increased in rabbits (Fein et al., 1980) exposed to high
       concentrations of CO (63% COHb), and increased capillary endothelial permeability to 131I-
       labeled human serum albumin was reported in early human studies (Parving, 1972) following
       acute, high-level CO exposure (23% COHb); however, no accumulation of lung water was
30     found in  dogs (Halebian et al., 1984a,b) with COHb levels of 59% and no edema was found
       in the lungs of rats chronically exposed to CO concentrations as high as 1300 ppm (Penney

       March 14, 1990                           1-9      DRAFT - DO NOT QUOTE OR CITE

-------
       et al.,  1988).  In addition, no changes in diffusing capacity of the lung were found in dogs
       with COHb levels up to 18%  (Fisher et al., 1969). It is unlikely, therefore, that CO has any
       direct effect on lung tissue except at extremely high concentrations. The capillary endothelial
       and alveolar epithelial edema  found with high levels of CO exposure in victims of CO
 5     poisoning may be secondary to cardiac failure produced by myocardial hypoxia (Fisher et al.,
       1969) or may  be due to acute cerebral anoxia (Naeije et al., 1980).
            Ventilatory responses to CO are related to the CO concentration as well as to the
       experimental conditions and the animal species being studied.  In conscious goats (Chapman
       et al.,  1980; Doblar et al., 1977; Santiago and Edelman, 1976) and cats (Gautier and Bonora,
10     1983), after an initial depression, ventilation suddenly increases, particularly at high CO
       concentrations (>2000 ppm).  This response may result from the direct effects of hypoxia
       (and possibly central acidosis) and/or a specific CNS effect of CO.  No effects on ventilation
       and perfusion  distribution were found, however, in dogs exposed to 1 % CO for 10 min,
       resulting in COHb levels of 59% (Robinson et al.,  1985). At very high concentrations of CO
15     (COHb >60%) total pulmonary resistance, measured indirectly by trachea! pressure, was
       reported to increase (Mordelet-Dambrine et al., 1978; Mordelet-Dambrine and Stupfel, 1979).
            Human studies on the pulmonary function effects of CO are complicated by the lack of
       adequate exposure information,  the small number of subjects studied,  and the short exposures
       explored. Occupational or accidental exposure to the products of combustion and pyrolysis,
20     particularly indoors, may lead to acute decrements in lung function if the COHb levels are
        > 17% (Sheppard et al., 1986) but not at concentrations <2% (Evans et al.,  1988; Hagberg
       et al.,  1985; Cooper and Alberti, 1984). It is difficult, however, to separate the potential
       effects of CO  from those due to other respiratory irritants in the smoke and exhaust.
       Community population studies on CO in ambient air have not found any relationships with
25     pulmonary function, symptomatology, and disease (Lebowitz et al., 1987; Robertson and
       Lebowitz, 1984; Lutz, 1983).

       Cardiovascular Effects
            The 1984 Addendum to the 1979 Air Quality Criteria Document for Carbon Monoxide
30     (U.S. Environmental Protection Agency, 1984) reported what appears to be a linear
       relationship between level of COHb and decrements in human exercise performance,

       March 14, 1990                          1-10     DRAFT - DO NOT QUOTE OR CITE

-------
        measured as maximal O2 uptake.  Exercise performance consistently decreases at a blood level
        of about 5.0% COHb in young, healthy, nonsmoking individuals (Klein et al., 1980; Stewart
        et al., 1978; Weiser et al., 1978).  Some studies have even observed a decrease in
        performance at levels as low as 2.3 to 4.3% COHb (Horvath et al., 1975; Drinkwater et al.,
 5      1974; Raven et al., 1974a); however, this decrease is so small as to be of concern mainly for
        competing athletes rather than for ordinary people conducting the activities of daily life.
        Cigarette smoking has a similar effect on cardiorespiratory response to exercise in nonathletic
        human subjects indicating a reduced ability for sustained work (Hirsch et al., 1985; Klausen
        etal., 1983).
10           Since the 1979 Air Quality Criteria Document (U.S. Environmental Protection Agency,
        1979), several important studies appearing in the literature have expanded the cardiovascular
        data base.  Adverse effects in patients with reproducible exercise-induced angina (Allred
        et al., 1989a,b) have been noted with postexposure COHb levels (CO-Oximeter measurement)
        as low as 3.2% (corresponding to an increase of 2.0% from baseline).  Sheps et al. (1987)
15      also found a similar effect in a group of patients with angina at COHb levels of 3.8%
        (representing an increase of 2.2% from baseline).  Kleinman et al.  (1989) studied subjects
        with angina and found an effect at 3% COHb representing an increase of 1.5% from baseline.
        Thus, the lowest observed adverse effect level in patients with stable angina is somewhere
        between 3 and 4% COHb (CO-Oximeter measurement), representing an increase from
20      baseline of from 1.5 to  2.2%.  Effects on silent ischemia episodes, which represent the
        majority of episodes in these patients, have not been studied.
             Exposure sufficient to achieve 6% COHb recently has been shown to adversely affect
        exercise-related arrhythmia in patients with coronary artery disease (Sheps et al., 1989).  This
        finding combined with the epidemiologic work of Stern et al. (1988) in tunnel workers is
25      suggestive but not conclusive that CO exposure  may provide an increased risk of sudden death
        from arrhythmia in patients with coronary artery disease.
             There is also strong evidence from both theoretical considerations and experimental
        studies in animals that carbon monoxide can adversely affect  the cardiovascular system.
        Results from animal  studies suggest that inhaled CO can cause disturbances in cardiac rhythm
30      and conduction in healthy as well as cardiac-impaired animals. The lowest observed effect
        level varies, depending upon the exposure regime used and species tested. Values reported in

        March 14, 1990                          1-11     DRAFT - DO NOT QUOTE OR CITE

-------
       the literature for 6 to 24 week exposures range from 2,6 to 12.0% COHb (50 and 100 ppm
       CO) in dogs to 12.4% (100 ppm CO) in monkeys.  For shorter duration exposures of 0.6 to
       16 h, disturbances in cardiac rhythm were found at 4.9 to 17.0% COHb (500 ppm CO) in
       dogs and 9.3% COHb (100 ppm CO) in monkeys.  Results from animal studies also indicate
 5     that inhaled CO can increase hemoglobin concentration and hematocrit ratio at COHb levels
       of 9.26% (100 ppm CO).  Small increases in hemoglobin and hematocrit probably represent a
       compensation for the reduction in oxygen transport caused by CO.  At higher CO
       concentrations, excessive increases in hemoglobin and hematocrit may impose an additional
       workload on the heart and compromise blood flow to the tissues.  For example,  cardiomegaly
10     has been reported in adult animals at COHb levels of 12% (200 ppm CO).  The oxygen
       transport system of the fetus, however,  is especially sensitive to CO inhaled by the mother,
       and may be affected by CO at concentrations as low as 60 ppm.
            There is conflicting evidence that CO exposure will  enhance development of
       atherosclerosis in laboratory animals; and most studies show no measurable effect. Similarly,
15     the possibility that CO will promote significant changes in lipid metabolism that might
       accelerate atherosclerosis is suggested in only a few studies.  Any such effect must be subtle
       at most.  Finally, CO probably inhibits rather than promotes  platelet aggregation.  Except for
       the studies by Rogers et al. (1980,  1988) on baboons, the CO exposures used in the studies
       on atherosclerosis created COHb levels of 7%  or higher;  sometimes much higher. While
20     occupational exposures in some workplace situations might regularly lead to COHb levels of
       10% or more, such high exposure levels are almost never encountered in the
       nonoccupationally exposed general public.  In  this general population, exposures are rarely as
       much as 25 to 50 ppm, and COHb levels typically are below 3% in nonsmokers. When
       examined in this context,  this document therefore, provides little data to indicate that an
25     atherogenic effect of exposure would be likely to occur in human populations at commonly
       encountered levels of ambient  CO.

       Cerebrovascular and Behavioral Effects
            The data reviewed in this document indicate that carbon monoxide hypoxia increases
30     cerebral blood flow, even at very low exposure levels.  Cerebral O2 consumption is well
       maintained until levels of COHb reach upwards of 30%.  The overall responses of the

       March 14, 1990                         1-12     DRAFT - DO NOT QUOTE OR CITE

-------
        cerebrovasculature are similar in the fetus, newborn, and adult animal; however, the
        mechanism of the increase in cerebral blood flow is still unclear.  In fact, several mechanisms
        working simultaneously to increase cerebral blood flow appear likely and these may involve
        metabolic and neural aspects as well as the oxyhemoglobin dissociation curve, tissue O2
 5      levels, and even a histotoxic effect of CO. These potential mechanisms of CO-induced
        alterations in  the cerebral circulation need to be investigated  further.  The interaction of CO
        with cyanide (additive and synergistic) on the cerebral vasculature is clear, however the
        interaction of CO with other agents and their combined effects on brain blood vessels is
        unknown.  This also is true for the long term (chronic) effects of CO alone, or in
10      combination with other agents in low- or high-dose levels on the cerebral vasculature.
        Finally, under normal circumstances the brain can increase its blood flow or its O2 extraction
        in order to compensate for a reduced O2 environment. Whether these compensatory
        mechanisms continue to operate successfully in a variety of conditions where the brain, or its
        vasculature are compromised (i.e., stroke, head injury, atherosclerosis, hypertension) is
15      unknown and requires further investigation.
             Behaviors that require sustained attention and/or sustained performance are most
        sensitive to disruption by COHb.  Therefore, the group of studies of tracking, vigilance,  and
        continuous performance offer the most consistent  and defensible evidence of COHb effects on
        behavior.  The results across studies is,  however,  far from consistent.  Further examination of
20      the three areas seems appropriate.
             Compensatory tracking was studied by two groups of investigators using virtually
        identical task  parameters and equipment (Putz et al.,  1976, 1979; Benignus et al.,  1987,
        1989).  Both of the studies by Putz et al. (1976, 1979) found significant and moderately large
        effects of 5%  COHb.  Benignus et al. (1987) reported similar but smaller significant effects in
25      a nearly identical experiment to Putz et al. (1976). However, in a dose-effects study
        including another direct replication group, Benignus et al. (1989a) found no significant
        effects, even for COHb levels of 17%.  In the latter study, the means were nearly doseordinal
        but too small  to be statistically significant. It is particularly  puzzling why the latter study,
        using a large number of subjects on an identical task, should find no significant effects for
30      even 17% COHb when three other studies found effects at lower levels.  Three other double-
        blind tracking studies of various methods found no effects of COHb levels of 12% or greater.

        March 14, 1990                           1-13     DRAFT - DO NOT QUOTE OR CITE

-------
       There is a similar disunity among studies on the effects of COHb on vigilance.  Because of
       the many failed attempts at direct replication, the conclusions seem weaker than for tracking.
             Of the five double-blind experiments in which continuous performance was measured,
       three were mentioned earlier in the discussion of tracking. In these studies (direct
 5     replications), continuous performance was measured simultaneously with tracking (Putz et al.,
       1976, 1979; Benignus et al., 1987). The latter of the three found no effects. A small study
       reported continuous performance effects that were disordinal in COHb (O'Donnell et al.,
       1971).  The remaining study (Benignus et al., 1977) used a different task and obtained no
       COHb effects.
10           As discussed above, a proportional vasodilation occurs in the brain in response to  COHb
       elevation. This vasodilation is sufficient, on the average, to keep the cerebral O2 consumption
       from being reduced even though  the COHb has reduced the blood's O2-carrying capacity 20  to
       30%  and the presence of COHb has shifted the  oxyhemoglobin dissociation curve to the left.
       The cerebral vasodilation may be viewed, Ideologically, as a closed-loop compensatory
15     mechanism to assure adequate oxygenation of the brain in the presence of elevated COHb.  If
       the cerebral vasodilation is adequate in any individual and if the vasodilation is homogeneous
       for all cerebral tissue, then that individual should not be behaviorally impaired by COHb
       elevation. This statement assumes that the sole mechanism for CO toxicity is the hypoxic
       effect of COHb.
20           The agreement between the behavioral literature and the compensatory  mechanism
       hypothesis is noteworthy.  According to the compensatory mechanism data, O2 consumption
       in the brain does not begin to decrease until COHb exceeds 20  to 30%.  Data from behavioral
       studies in laboratory animals demonstrate that significant effects in schedule-controlled
       behavior do not occur below 20 to 30% COHb. Behavioral effects in healthy humans have
25     not been clearly demonstrated below 20 to 30% COHb. It seems unwise,  however, to totally
       ignore the evidence suggesting effects below these levels. Even if effects are small or
       occasional, they might be important to the performance of critical tasks  in some individuals.
             Some of the differences among studies of the effect of COHb on the behavior of humans
       are due apparently to technical problems in the execution of experiments, because single-
30     blind or nonblind experiments tend to yield a much higher rate of significant effects than do
       double-blind studies. Even when non-double-blind  experiments are eliminated from

       March 14, 1990                           1-14     DRAFT - DO NOT QUOTE OR CITE

-------
        consideration, however, a substantial amount of disparity remains among results of studies.  It
        is possible that such residual disagreement is due to the action of an unsuspected variable that
        is not being controlled across experiments.
            If the compensatory CNS blood flow hypothesis has validity, it is possible that there
 5      exist groups that are at higher risk to COHb elevation than the usual  subjects who were
        studied in the behavioral experiments. Disease or injury might either impair the
        compensatory mechanism or reduce the non-exposed O2 delivery. Aging increases the
        probability of such injury and disease. It also is possible that there exist individual
        differences with regard to COHb sensitivity and/or compensatory mechanisms.  Too little is
10      known about the compensatory process to make conjectures, but the matters seem important
        to investigate.
            The literature on the behavioral effects of COHb elevation has grown considerably since
        the last Criteria Document was written (U.S. Environmental Protection Agency, 1979). It
        seems safe to state that the effect of the new information did not increase the certainty about
15      COHb effects.  Unless some key piece of information is uncovered by new research, there
        does not seem to be much hope of gaining clarification in the conflicting findings. The
        solution to the puzzle would seem to lie in the conduct of more research into mechanisms of
        action of CO rather than in further attempts to show reliable behavioral effects. The latter
        approach, which has not been successful in the past, should be resumed only when
20      mechanisms of toxicity are understood better.  More findings of behavioral effects of COHb
        would not appreciably alter the conclusions of this document unless future studies were to
        show an unusual unanimity.

        Developmental Toxicity
25          The data reviewed in this document provide strong evidence that CO exposures of 150
        to 200 ppm produce reductions in birthweight, cardiomegaly, delays in behavioral
        development, and disruption in cognitive function in laboratory animals of several species.
        Isolated experiments suggest that some of these effects  may be present at doses as low as 60
        to 65 ppm maintained throughout gestation. The current data from human children suggesting
30      a link between environmental CO exposures and sudden infant death syndrome (SIDS)  are
        weak, but further study  should be encouraged.  Human data from cases of accidental high

        March 14, 1990                          1-15     DRAFT - DO NOT QUOTE OR CITE

-------
       dose CO exposures are difficult to use in identifying a lowest observed effect level (LOEL) or
       a no observed effects level (NOEL) for this agent because of the small numbers of cases
       reviewed and problems in documenting levels of exposure.  However, such data if
       systematically gathered and reported could be useful in identifying possible ages of special
 5     sensitivity to CO and cofactors or other risk factors that might identify sensitive
       subpopulations.

       Other Systemic Effects of Carbon Monoxide
            Laboratory animal studies reviewed in the previous criteria document (U.S.
10     Environmental Protection Agency, 1979) and again in this document suggest that enzyme
       metabolism and the P-450-mediated metabolism of xenobiotic compounds may be affected by
       CO exposure (Montgomery and Rubin, 1971; Kustov et al., 1972; Pankow and Ponsold,
       1972, 1974; Martynjuk and Dacenko, 1973; Swiecicki,  1973; Pankow et al., 1974; Roth and
       Rubin, 1976a,b).  Most of the authors have concluded, however, that effects on metabolism
15     at low COHb levels (<15%) are attributable entirely to tissue hypoxia produced by increased
       levels of COHb because they are no greater than the effects produced by comparable levels of
       hypoxic hypoxia.  At higher levels of exposure, where COHb concentrations exceed 15 to
       20%, there may be direct inhibitory effects of CO on the activity of mixed-function oxidases
       but more basic research is needed.  The decreases in xenobiotic metabolism shown with CO
20     exposure might be important to individuals receiving treatment with drugs.
            The effects of CO on tissue metabolism noted above may partially explain the body
       weight changes associated with exposure to high concentrations of CO. Short-term exposure
       to 250-1000 ppm for 24 h was reported previously to cause weight loss in laboratory rats
       (Koob et al., 1974) but no significant body weight effects were reported in long-term
25     exposure studies in laboratory animals at CO concentrations ranging from 50 ppm for 3 mo to
       3000 ppm for 300 days (Theodore et al., 1971; Musselman et al., 1959; Campbell, 1934;
       Stupfel and Bouley, 1970).
            Inhalation of high levels of CO, leading to COHb concentrations greater than 10 to
       15%, have been reported to cause a number of systemic effects in laboratory animals as well
30     as effects in humans suffering from acute CO poisoning. Tissues of highly active oxygen
       metabolism, such as heart, brain, liver, kidney, and muscle, may be particularly  sensitive to

       March 14, 1990                          1-16     DRAFT - DO NOT QUOTE OR CITE

-------
        CO poisoning.  The impairment of function in the heart and brain caused by CO exposure is
        well known and has been described above.  Other systemic effects of CO poisoning are not as
        well known and are, therefore, less certain.  There are reports in the literature of effects on
        liver (Katsumata et al., 1980), kidney (Kuska et al., 1980), and bone (Zebro et al., 1983).
 5      Results from one additional study in adult guinea pigs suggest that immune capacity in the
        lung and spleen was affected by intermittent exposure to high levels of CO for 3 to 4 weeks
        (Snella and Rylander,  1979). It generally is agreed that these systemic effects are caused by
        the severe tissue damage occurring during acute CO poisoning due to (1) ischemia resulting
        from the formation of COHb, (2) inhibition of O2 release from HbO2,  (3) inhibition of cellular
10      cytochrome function (e.g., cytochrome oxidases), and (4) metabolic acidosis.

        Adaptation
            The only evidence for short- or long-term COHb compensation in man is indirect.
        Experimental animal data indicate that COHb levels produce physiological responses that tend
15      to offset other deleterious effects of CO exposure. Such responses are (1) increased coronary
        blood flow, (2) increased cerebral blood flow, (3) increased hemoglobin through increased
        hemopoiesis, and (4) increased O2 consumption in muscle.
            Short-term compensatory responses in blood flow or O2 consumption may not be
        complete or might even be lacking in certain persons.  For example, from laboratory animal
20      studies it is known that coronary blood flow is increased with COHb, and from human
        clinical studies it is known  that subjects with ischemic heart disease respond to the lowest
        levels of COHb (5%, or less).  The implication is that in some cases of cardiac impairment,
        the short-term compensatory mechanism is impaired.
            From neurobehavioral studies, it is apparent that decrements due to CO have not
25      occurred consistently in all subjects, or even in the same studies, and have not demonstrated a
        dose-response relationship with increasing COHb levels.  The implication from this data
        suggests that there might be some threshold or time lag in a compensatory mechanism such as
        increased cerebral blood flow.  Without direct physiological evidence in  either laboratory
        animals or, preferably humans, this concept only can be hypothesized. The observed results
30      from the neurobehavioral studies could be explained by differences or  problems in
        experimental protocols or due to possible nonrandom sampling.

        March 14, 1990                          1-17    DRAFT - DO NOT QUOTE OR CITE

-------
            The idea of a threshold or a time lag in compensatory mechanisms should not be
       rejected entirely, however.  There simply is no direct evidence.  Studies need to be performed
       to (1) measure cerebral blood flow and tissue PO2 with low COHb levels at various ambient
       concentrations of CO to determine early and low level effects accurately, and (2) design
 5     behavioral studies where threshold effects or time lags are factors in the experimental
       protocols that can be explicitly studied.
            The mechanism by which long-term adaptation would occur, if it could be demonstrated
       in humans, is assumed to be an increased Hb concentration via a several-day increase in
       hemopoiesis. This alteration in Hb production has been demonstrated repeatedly in animal
10     studies but no recent studies have been conducted indicating or suggesting that some
       adaptational benefit has or would occur.  Furthermore, even  if the Hb increase is a signature
       of adaptation, it has not been  demonstrated to  occur at low ambient concentrations of CO.
       The human studies of the 1940s have not been replicated, so  the question of adaptation
       remains unresolved.
15
       Combined Exposure of Carbon Monoxide with Other Pollutants. Drugs, and Environmental
       Factors

       High Altitude Effects
20          While  there are many studies comparing  and contrasting inhaling CO with exposure to
       altitude,  there are relatively few reports on the effects of inhaling CO at altitude.  There are
       data to support the possibility that the effects of these two hypoxia episodes  are at least
       additive. These data were obtained at CO concentrations that are too high to have much
       meaning for regulatory concerns.  There also are data that indicate decrements in visual
25     sensitivity and flicker-fusion frequency in subjects exposed to CO (COHb = 5 to  10%) at
       higher altitudes. These data,  however, are somewhat controversial.
            There are even fewer studies of the long-term effects of CO at high altitude.  These
       studies generally indicate few changes at CO concentrations below 100 ppm and altitudes
       below 4572 m (15,000 ft).  A provocative study by McDonagh  et al. (1986) suggests that the
30     increase  in ventricular capillarity seen with altitude exposure may be blocked by CO.  The
       fetus may be particularly sensitive to the effects of CO at altitude; this is especially true with
       the high  levels of CO associated with maternal smoking.
       March 14, 1990                          1-18     DRAFT  - DO NOT  QUOTE OR CITE

-------
       Carbon Monoxide Interaction with Drugs
            There remains little direct information on the possible enhancement of CO toxicity by
       concomitant drug use or abuse; however, there are some data suggesting cause for concern.
       There is some evidence that interactions of drug effects with CO exposure can occur in both
 5     directions, that is, CO toxicity may be enhanced by drug use and the toxic or other effects of
       drugs may be altered by CO exposure. Nearly all the published data that are available on CO
       combinations with drugs concern psychoactive drugs.
            The use and abuse of psychoactive drugs and alcohol is ubiquitous in society.  Because
       of CO's effects on brain functioning, interactions between CO and psychoactive drugs could
10     be anticipated. Unfortunately, very little systematic research has addressed this question.  In
       addition, very little of the research that has been done has  utilized models for expected effects
       for treatment combinations. Thus, often it is not possible to assess whether the combined
       effects of drugs and CO exposure are additive or differ from additivity. It is important to
       recognize that even additive effects of combinations can be of clinical significance, especially
15     when the individual is unaware of the  combined hazard.

       Combined Exposure of Carbon Monoxide with Other Air Pollutants
       and Environmental Factors
            Much of the data concerning the combined effects  of CO and other pollutants found in
20     the ambient air are based on animal experiments. Only  a few human studies are available.
       Early studies in healthy human  subjects by Hackney et al.  (1975a,b), Raven et al. (1974a,b),
       Gliner et al. (1975), and Drinkwater et al. (1974) on common air pollutants such as NO2, O3,
       or PAN and more recent work on CO + O3 by DeLucia et al. (1983) failed to show any
       interaction from combined exposure.
25          In animal studies, no interaction  was observed following combined exposure of CO and
       pollutants such as HCN, NO2, SO2, or PbClBr (Hugod,  1979; Busey, 1972; Murray et al.,
       1978). However, an additive effect was observed following combined exposure  of high levels
       of CO + NO (Groll-Knapp et al., 1988), and a synergistic effect was observed after
       combined exposure to  CO and O3 (Murphy, 1964).
30          Toxicological interactions of combustion products, primarily CO, CO2, and HCN, from
       indoor and outdoor fires, have shown  a synergistic effect following CO + CO2 exposure
       (Rodkey and Collison, 1979; Levin et al., 1987a) and an additive effect with CO + HCN
       March 14, 1990                          1-19    DRAFT - DO NOT QUOTE OR CITE

-------
       (Levin et al., 19875). Additional studies are needed, however, to evaluate the effects of CO
       under conditions of hypoxic hypoxia.
            Finally, laboratory animal studies (Yang et al., 1988; Fechter et al., 1988; Young
       et al.,  1987) suggest that the combination of environmental factors such as heat stress and
 5     noise may be important determinants of health effects occurring in combination with exposure
       to CO.  Of the effects described, the one potentially most relevant to typical human exposures
       is a greater decrement in exercise performance seen when heat stress is combined with 50
       ppm CO (Drinkwater et al., 1974; Raven et al., 1974a,b; Gliner et, al., 1975).

10     Environmental Tobacco Smoke
            Although tobacco smoke is another source of CO for smokers as well as nonsmokers, it
       is also a source of other chemicals with which environmental CO levels could interact.
       Available data strongly suggest that  acute and chronic CO exposure attributed to tobacco
       smoke can affect the cardiopulmonary system, but the potential interaction of CO with other
15     products of tobacco smoke confounds the results.  In addition, it is not clear if incremental
       increases in COHb caused by environmental exposure would actually be additive to
       chronically elevated COHb levels due to tobacco smoke, because some physiological
       adaptation may take place.  There is, therefore, a need for further research to describe these
       relationships better.
20
       Evaluation of Subpopulations Potentially At Risk to Carbon Monoxide
       Exposure
            Most of the information on the human health effects of carbon monoxide discussed in
       this document has concentrated on two carefully defined population groups - young healthy,
25     predominantly male adults  and patients with diagnosed coronary artery disease. On the basis
       of the  known effects described, patients with reproducible exercise-induced angina  appear to
       be best established as a sensitive group within the general population that is at increased risk
       for experiencing health effects (i.e., decreased exercise duration due to exacerbation of
       cardiovascular symptoms) of concern at ambient or near-ambient CO-exposure concentrations
30     that result in COHb levels  of <5%.  A  smaller sensitive group of healthy individuals
       experience decreased exercise duration at similar levels of CO exposure, but only during
       short-term maximal exercise. Decrements in exercise duration in the healthy  population,
       March 14, 1990                           1-20     DRAFT - DO NOT QUOTE OR CITE

-------
        therefore, would be mainly of concern to competing athletes rather than for nonathletic people
        carrying out the common activities of daily life.
             It is known, however, from both theoretical work and from experimental research in
        laboratory animals that certain other groups in the population are at potential risk to exposure
 5      from CO. Another purpose of this document is to explore the potential effects of CO in
        population groups that have not been studied adequately, but which could be expected to be
        susceptible to CO because of underlying physiological status either due to gender differences,
        aging, preexisting disease, or because of the use of medications or alterations in their
        environment. These probable risk groups include (1) fetuses and young infants;  (2) pregnant
10      women; (3) the elderly, especially those with compromised cardiopulmonary or
        cerebrovascular functions; (4) individuals with obstructed coronary arteries, but not yet
        manifesting overt symptomatology of coronary artery disease;  (5) individuals with congestive
        heart failure; (6) individuals with peripheral vascular or cerebrovascular disease;
        (7) individuals with hematological diseases  (e.g., anemia) that affect oxygen-carrying capacity
15      or transport in the blood; (8) individuals with genetically unusual forms of hemoglobin
        associated with reduced oxygen-carrying capacity; (9) individuals with chronic obstructive
        lung diseases; (10) individuals using medicinal or recreational drugs having CNS depressant
        properties; (11) individuals exposed to other pollutants (e.g.,  methylene chloride) that
        increase endogenous formation of CO; and  (12) individuals who have not been adapted to
20      high altitude and are exposed to a combination of high altitude and CO. Unfortunately, little
        empirical evidence currently is available by which to specify health effects associated with
        ambient or near-ambient CO exposures in these probable risk groups.
25      References

       Akland, G. G.; Hartwell, T. D.; Johnson, T. R.; Whitmore, R. W. (1985) Measuring human exposure to carbon
             monoxide in Washington, D.C., and Denver, Colorado, during the winter of 1982-1983. Environ. Sci.
             Technol. 19: 911-918.
30
       Allred, E. N.; Bleecker, E. R.; Chaitman, B. R.; Dahms, T. E.; Gottlieb, S. O.; Hackney, J. D.; Pagano, M.;
             Selvester, R. H.; Walden, S. M.; Warren, J. (1989a) Short-term effects of carbon monoxide exposure on
             the exercise performance of subjects with coronary artery disease. N. Engl. J. Med. 321: 1426-1432.
35

        March 14,  1990                           1-21     DRAFT - DO NOT QUOTE OR CITE

-------
        Allred, E. N.; Bleecker, E. R.; Chairman, B. R.; Dahms, T. E.; Gottlieb, S. O.; Hackney, J. D.; Hayes, D.;
               Pagano, M. Selvester, R. H.; Walden, S. M.; Warren, J. (1989b) Acute effects of carbon monoxide
               exposure on individuals with coronary artery disease. Cambridge, MA: Health Effects Institute; research
               report no. 25.
 5
        Benignus, V. A.; Otto, D. A.; Prah, J. D.; Benignus, G. (1977) Lack of effects of carbon monoxide on human
               vigilance. Percept. Mot. Skills 45: 1007-1014.

        Benignus, V. A.; Muller, K. E.; Barton, C. N.; Prah, J.  D. (1987) Effect of low level carbon monoxide on
10            compensatory tracking and event monitoring. Neurotoxicol. Teratol. 9: 227-234.

        Benignus, V. A.; Muller, K. E.; Pieper, K. S.; Prah, J. D. (1989) Compensatory tracking in humans with
               elevated carboxyhemoglobin. Neurotoxicol. Teratol.: in press.

15     Busey, W. M. (1972) Chronic exposure of albino rats to certain airborne pollutants [unpublished material].
               Vienna,  VA: Hazleton Laboratories, Inc.

        Campbell, J. A. (1934) Growth, fertility, etc. in animals during attempted acclimatization to carbon monoxide.
               Exp. Physiol. 24: 271-281.
20
        Chapman, R. W.; Santiago, T. V.; Edelman, N. H. (1980) Brain hypoxia and control of breathing:
               neuromechanical control. J. Appl. Physiol.:  Respir. Environ. Exercise Physiol. 49: 497-505.

        Chen, S.; Weller, M. A.; Penney, D. G. (1982) A study  of free lung cells from young rats chronically exposed
25            to carbon monoxide from birth. Scanning Electron Microsc. 2: 859-867.

        Coburn, R.  F.; Forster, R. E.; Kane, P. B. (1965) Considerations of the physiological variables that determine
               the blood carboxyhemoglobin concentration in man. J. Clin. Invest. 44:  1899-1910.

30     Cobum, R.  F.; Wallace, H. W.; Abboud, R. (1971) Redistribution of body carbon monoxide after hemorrhage.
               Am. J. Physiol. 220: 868-873.

        Coburn, R.  F.; Ploegmakers, F.; Gondrie, P.; Abboud, R. (1973) Myocardial myoglobin oxygen tension. Am. J.
               Physiol. 224: 870-876.
35
        Cooper, K.  R.; Alberti, R. R. (1984) Effect of kerosene heater emissions on indoor air quality and pulmonary
               function. Am. Rev. Respir. Dis. 129: 629-631.

        DeLucia, A. J.;  Whitaker, J. H.; Bryant, L. R. (1983) Effects of combined exposure to ozone and carbon
40            monoxide (CO) in humans. In: Lee, S. D.; Mustafa, M. G.; Mehlman, M. A., eds. International
               symposium on the biomedical effects of ozone and related photochemical oxidants; March; Pinehurst,
               NC. Princeton, NJ: Princeton Scientific Publishers, Inc.; pp. 145-159. (Advances in modern
               environmental toxicology: v. 5).

45     Dennis, R. C.; Valeri, C. R. (1980) Measuring percent oxygen saturation of hemoglobin, percent
               carboxyhemoglobin and methemoglobin, and concentrations of total hemoglobin and oxygen in blood of
               man, dog, and baboon. Clin. Chem. (Winston-Salem,  NC) 26: 1304-1308.

        Doblar, D.  D.; Santiago, T. V.; Edelman, N. H. (1977) Correlation between ventilatory and cerebrovascular
50            responses to inhalation of CO. J. Appl. Physiol.:  Respir. Environ. Exercise Physiol. 43: 455-462.

        Drinkwater, B. L.; Raven, P. B.; Horvath, S. M.; Gliner, J. A.; Ruhling, R. O.; Bolduan,  N. W.; Taguchi, S.
               (1974) Air pollution, exercise, and heat stress. Arch. Environ. Health 28:  177-181.
         March 14, 1990                               1-22     DRAFT - DO NOT QUOTE OR CITE

-------
        Evans, R. G.; Webb, K.; Homan, S.; Ayres, S. M. (1988) Cross-sectional and longitudinal changes in
               pulmonary function associated with automobile pollution among bridge and tunnel officers. Am. J. Ind.
               Med. 14: 25-36.

 5     Fechter, L.  D.; Young, J. S.; Carlisle, L. (1988) Potentiation of noise induced threshold shifts and hair cell loss
               by carbon monoxide. Hear. Res. 34: 39-47.

        Fein, A.; Grossman, R. F.; Jones, J. G.; Hoeffel, J.; McKay,  D.  (1980) Carbon monoxide effect on alveolar
               epithelial permeability. Chest 78: 726-731.
10
        Fenn, W. O. (1970) The burning of CO  in the tissues. In: Cobum, R. F., ed. Biological effects of carbon
               monoxide. Ann. N. Y. Acad. Sci. 174: 64-71.

        Fisher, A. B.; Hyde, R. W.; Baue, A. E.; Reif, J. S.; Kelly, D. F. (1969) Effect of carbon monoxide on
15            function and structure of the lung. J. Appl. Physiol. 26: 4-12.

        Flachsbart, P. G.; Mack, G. A.; Howes, J. E.; Rodes, C. E. (1987) Carbon monoxide exposures of Washington
               commuters. J. Air Pollut.  Control Assoc. 37:  135-142.

20     Gautier, H.; Bonora, M. (1983) Ventilatory response of intact cats to carbon monoxide hypoxia.  J. Appl.
               Physiol.:  Respir. Environ. Exercise Physiol. 55: 1064-1071.

        Gliner, J. A.; Raven,  P. B.; Horvath, S. M.; Drinkwater, B. L.; Sutton, J. C. (1975) Man's physiologic
               response to  long-term work during thermal and pollutant stress. J. Appl. Physiol. 39: 628-632.
25
        Groll-Knapp, E.; Haider, M.; Kienzl, K.; Handler, A.; Trimmel,  M. (1988) Changes in discrimination learning
               and brain activity (ERP's) due to combined exposure to NO and CO in rats. Toxicology 49: 441-447.

        Hackney, J. D.; Linn, W. S.; Mohler, J. G.; Pedersen, E.  E.;  Breisacher, P.; Russo, A. (1975a) Experimental
30            studies on human health effects of air pollutants: II. four-hour exposure to ozone alone and in
               combination with other pollutant gases. Arch.  Environ. Health 30: 379-384.

        Hackney, J. D.; Linn, W. S.; Law,  D. C.; Karuza, S. K.; Greenberg, H.; Buckley, R. D.; Pedersen, E. E.
               (1975b) Experimental studies on  human health effects of air pollutants: III. two-hour exposure to ozone
35            alone and in combination with other pollutant  gases. Arch. Environ. Health 30: 385-390.

        Hagberg, M.; Kolmodin-Hedman, B.; Lindahl, R.; Nilsson, C.-A.; Norstrom, A. (1985) Irritative complaints,
               carboxyhemoglobin increase and  minor ventilatory function changes due to exposure to chain-saw
               exhaust. Eur. J. Respir. Dis. 66: 240-247.
40
        Halebian, P. H.; Barie, P. S.; Robinson, N.; Shires, G. T. (1984a) Effects of carbon monoxide on pulmonary
               fluid accumulation. Curr. Surg. 41: 369-371.

        Halebian, P.; Sicilia, C.; Hariri, R.; Inamdar, R.; Shires, G. T. (1984b) A safe and reproducible model of
45            carbon monoxide poisoning. Ann. N. Y. Acad. Sci. 435: 425-428.

        Hartwell, T. D.; Clayton, C. A.; Ritchie, R. M.; Whitmore, R. W.;  Zelon, H. S.; Jones,  S. M.; Whitehurst, D.
               A. (1984) Study of carbon monoxide exposure of residents of Washington, DC and Denver, Colorado.
               Research Triangle Park, NC: U.  S. Environmental Protection Agency, Environmental Monitoring
50            Systems Laboratory; EPA report no. 600/4-84-031. Available from: NTIS, Springfield, VA;
               PB84-183516.

        Hirsch, G. L.; Sue, D. Y.; Wassennan,  K.; Robinson, T. E.; Hansen, J. E. (1985) Immediate effects of cigarette
               smoking on cardiorespiratory responses to exercise. J. Appl. Physiol, 58: 1975-1981.


         March 14,  1990                               1-23      DRAFT - DO NOT QUOTE OR CITE

-------
       Horvath, S. M.; Raven, P. B.; Dahms, T. E.; Gray, D. J. (1975) Maximal aerobic capacity at different levels of
              carboxyhemoglobin. J. Appl. Physiol. 38: 300-303.

       Hugod, C. (1979) Effect of exposure to 0.5 ppm hydrogen cyanide singly or combined with 200 ppm carbon
 S            monoxide and/or 5 ppm nitric oxide on coronary arteries, aorta, pulmonary artery, and lungs in the
              rabbit. Int. Arch. Occup. Environ. Health 44: 13-23.

       Hugod, C. (1980) The effect of carbon monoxide exposure on morphology of lungs and pulmonary arteries in
              rabbits: a light- and electron-microscopic study.  Arch. Toxicol. 43: 273-281.
10
       Johnson, T. (1984) A study of personal exposure to carbon monoxide in Denver, Colorado.  Research Triangle
              Park, NC: U. S.  Environmental Protection Agency, Environmental Monitoring Systems Laboratory; EPA
              report no. EPA-600/4-84-014. Available from: NTIS, Springfield, VA; PB84-146125.

15     Katsumata, Y.; Aoki, M.; Oya, M.; Yada, S.; Suzuki, O. (1980) Liver damage in rats during acute carbon
              monoxide poisoning. Forensic Sci. Int. 16: 119-123.

       Klausen, K.; Andersen, C.; Nandrup, S. (1983) Acute effects of cigarette smoking and inhalation of carbon
              monoxide during maximal exercise. Eur. J. Appl. Physiol. Occup. Physiol. 51: 371-379.
20
       Klein, J. P.; Forster, H.  V.; Stewart, R. D.; Wu, A. (1980) Hemoglobin affinity for oxygen during short-term
              exhaustive exercise. J. Appl.  Physiol.: Respir. Environ. Exercise Physiol. 48:  236-242.

       Kleinman, M. T.; Davidson, D. M.; Vandagriff, R.  B.; Caiozzo, V. J.; Whittenberger, J. L.  (1989) Effects of
25            short-term exposure to carbon monoxide in subjects with coronary artery disease. Arch. Environ. Health
              44: 361-369.

       Koob, G. F.; Annau, Z.; Rubin, R. J.; Montgomery, M. R. (1974) Effect of hypoxic hypoxia and carbon
              monoxide on food intake, water intake, and body weight in two strains of rats. Life Sci. 14:  1511-1520.
30
       Koontz, M. D.; Nagda, N. L. (1987) Survey effectors affecting NO2 concentrations. In: Seifert, B.; Esdorn, H.;
              Fischer, M.; Rueden, H.; Wegner, J., eds. Indoor air '87: proceedings of the  4th international conference
              on indoor air quality and climate, v. 1, volatile organic compounds, combustion gases, particles and
              fibres, microbiological agents; August; Berlin, Federal Republic of Germany. Berlin, Federal Republic of
35            Germany: Institute for Water, Soil, and Air Hygiene; pp. 430-434.

       Koontz, M. D.; Nagda, N. L. (1988) A topical report on a field monitoring study of homes with unvented gas
              space heaters: v.  III. methodology and results. Gas Research Institute final report; contract no.
              5083-251-0941.
40
       Kuska, J.; Kokot, F.; Wnuk, R. (1980) Acute renal failure after exposure to carbon monoxide. Mater. Med. Pol.
              (Engl. Ed.) 12: 236-238.

       Kustov, V. V.; Belkin, V. I.; Abidin, B.  L; Ostapenko, O. F.;  Malkuta, A. N.; Poddubnaja, L. T. (1972)
45            Aspects of chronic carbon monoxide poisoning in young animals. Gig. Tr. Prof.  Zabol.  5: 50-52.

       Lebowitz, M. D.; Collins, L.; Holberg, C. J. (1987) Time series analyses of respiratory responses to indoor and
              outdoor environmental phenomena. Environ.  Res. 43: 332-341.

50     Levin, B. C.; Paabo, M.; Gurman, J. L.; Harris, S.  E.; Braun, E. (1987a) Toxicological interactions between
              carbon monoxide and carbon dioxide.  Toxicology 47: 135-164.
         March  14,  1990                              1-24      DRAFT - DO NOT QUOTE OR CITE

-------
        Levin, B. C.; Paabo, M.; Gurman, J. L.; Harris, S. E. (1987b) Effects of exposure to single or multiple
               combinations of the predominant toxic gases and low oxygen atmospheres produced in fires. Fundam.
               Appl. Toxicol. 9: 236-250.

  5     Luomanmaki, K.; Coburn, R. F. (1969) Effects of metabolism and distribution of carbon monoxide on blood and
               body stores. Am. J. Physiol. 217: 354-363.

        Lutz, L. J. (1983) Health effects of air pollution measured by outpatient visits. J. Fam. Pract. 16: 307-313.

 10     Martynjuk, V. C.; Dacenko, I. I. (1973) Aktivnost'  transaminaz v usloviyakh khronicheskoi intoksikatsii okis'yu
               ugleroda [Aminotransferase activity in chronic carbon monoxide poisoning]. Gig. Naselennykh. Mest. 12:
               53-56.

        McDonagh, P. F.; Reynolds, J. M.; McGrath, J. J.  (1986) Chronic altitude plus carbon monoxide exposure
 15            causes left ventricular hypertrophy but an attenuation of coronary capillarity. Presented at: 70th annual
               meeting of the Federation of American Societies for Experimental Biology; April; St. Louis, MO. Fed.
               Proc.  Fed. Am. Soc. Exp. Biol. 45: 883.

        Montgomery,  M. R.; Rubin, R. J. (1971) The effect of carbon monoxide inhalation on in vivo drug metabolism
20            in the rat. J. Pharmacol. Exp. Ther. 179: 465-473.

        Mordelet-Dambrine, M.; Stupfel, M. (1979) Comparison in guinea-pigs and in rats of the effects of vagotomy
               and of atropine on respiratory resistance modifications induced by an acute carbon monoxide or nitrogen
               hypoxia. Comp. Biochem. Physiol. A 63: 555-559.
25
        Mordelet-Dambrine, M.; Stupfel, M.; Duriez, M. (1978) Comparison of tracheal pressure and circulatory
               modifications induced in guinea pigs and in rats by carbon monoxide inhalation.  Comp. Biochem.
               Physiol. A 59: 65-68.

30     Murphy, S. D. (1964) A review of effects on animals of exposure to auto exhaust and some of its components.  J.
               Air Pollut. Control Assoc. 14: 303-308.

        Murray, F. J.; Schwetz, B. A.; Crawford, A. A.; Henck, J. W.; Staples, R. E. (1978) Teratogenic potential of
               sulfur dioxide and carbon monoxide in mice and rabbits. In: Mahlum,  D. D.; Sikov, M. R.; Hackett, P.
35            L.; Andrew, F. D.,  eds. Developmental toxicology of energy-related pollutants: proceedings of the
               seventeenth  annual Hanford  biology symposium; October 1977; Richland, WA. Oak Ridge, TN: U. S.
               Department  of Energy, Technical Information Center; pp.  469-478. Available from: NTIS, Springfield,
               VA; CONF-771017.

40     Mussehnan, N. P.; Groff, W. A.; Yevich, P. P.; Wilinski, F. T.; Weeks, M.  H.; Oberst, F. W.  (1959)
               Continuous exposure of laboratory animals to low concentration of carbon monoxide. Aerosp. Med. 30:
               524-529.

        Naeije, R.; Peretz,  A.; Cornil, A. (1980) Acute pulmonary edema following carbon monoxide poisoning.
45           Intensive Care Med. 6:  189-191.

        O'Donnell, R. D.; Mikulka, P.; Heinig, P.; Theodore, J. (1971) Low level carbon monoxide exposure and
              human psychomotor performance. Toxicol. Appl. Pharmacol. 18: 593-602.

50     Pankow, D.; Ponsold, W. (1972) Leucine aminopeptidase activity  in plasma of normal and carbon monoxide
              poisoned rats. Arch. Toxicol. 29: 279-285.
        March 14, 1990                               1-25      DRAFT - DO NOT QUOTE OR CITE

-------
        Pankow, D.; Ponsold, W. (1974) Kombinationswirkungen von Kohlenmonoxid mit anderen biologischaktiven
               Schadfaktoren auf den Organismus [The combined effects of carbon monoxide and other biologically
               active detrimental factors on the organism]. Z. Gesamte Hyg. Ihre Grenzgeb. 20: 561-571.

 5      Pankow, D.; Ponsold, W.; Fritz, H. (1974) Combined effects of carbon monoxide and ethanol on the activities of
               leucine aminopeptidase and glutamic-pyruvic  transaminase in the plasma of rats.  Arch. Toxicol. 32:
               331-340.

        Parving, H.-H. (1972) The effect of hypoxia and carbon monoxide exposure on plasma volume and capillary
10             permeability to albumin. Scand. J. Clin. Lab. Invest. 30: 49-56.

        Penney, D. G.; Davidson, S. B.; Gargulinski, R. B.; Caldwell-Ayre, T. M. (1988)  Heart and lung hypertrophy,
               changes in blood volume, hematocrit and plasma renin  activity in rats chronically exposed in increasing
               carbon monoxide concentrations. JAT J. Appl. Toxicol. 8: 171-178.
15
        Penney, D. G.; Gargulinski, R. B.; Hawkins,  B. J.;  Santini, R.; Caldwell-Ayre, T.  M.; Davidson, S. B. (1988)
               The effects of carbon monoxide on persistent changes in young rat heart: cardiomegaly,  tachycardia and
               altered DNA content. JAT J. Appl. Toxicol.  8: 275-283.

20      Putz, V. R.;  Johnson, B. L.; Setzer, J. V. (1976) Effects of CO on vigilance performance: effects of low level
               carbon monoxide on divided attention, pitch discrimination, and the auditory evoked potential.
               Cincinnati, OH: U. S. Department of Health, Education, and Welfare, National Institute for Occupational
               Safety and Health. Available from: NTIS, Springfield,  VA; PB-274219.

25      Putz, V. R.;  Johnson, B. L.; Setzer, J. V. (1979) A  comparative study of the effects of carbon monoxide and
               methylene chloride on human performance. J. Environ. Pathol. Toxicol. 2: 97-112.

        Radford, E. P.; Drizd, T. A. (1982) Blood carbon monoxide levels in persons 3-74  years of age: United States,
               1976-80.  Hyattsville, Md: U. S. Department of Health and Human Services, National Center for Health
30             Statistics; DHHS publication no. (PHS) 82-1250. (Advance data from vital and health statistics, no. 76).

        Raven, P. B.; Drinkwater, B. L.; Horvath, S. M.; Ruhling, R. O.; Gliner, J. A.; Sutton, J. C.; Bolduan, N. W.
               (1974a) Age, smoking habits, heat stress, and their interactive effects with carbon monoxide and
               peroxyacetylnitrate on man's aerobic power. Int. J. Biometeorol. 18: 222-232.
35
        Raven, P. B.; Drinkwater, B. L.; Ruhling, R. O.; Bolduan, N.; Taguchi, S.; Gliner, J.; Horvath, S. M. (1974b)
               Effect of carbon monoxide  and peroxyacetyl nitrate on  man's maximal aerobic capacity.  J. Appl. Physiol.
               36: 288-293.

40      Robertson, G.; Lebowitz, M. D. (1984) Analysis of  relationships  between symptoms and environmental factors
               over time. Environ. Res. 33:  130-143.

        Robinson, N. B.; Barie, P. S.; Halebian, P. H.; Shires, G. T.  (1985) Distribution of ventilation and perfusion
               following acute carbon monoxide poisoning.  In:  41st annual  forum on fundamental surgical problems held
45             at the 71st annual clinical congress of the American College  of Surgeons;  October; Chicago, IL. Surg.
               Forum 36: 115-118.

        Rodkey, F. L.; Collison, H. A. (1979) Effects of oxygen and carbon dioxide on carbon  monoxide toxicity. J.
               Combust. Toxicol. 6: 208-212.
50
        Rogers, W. R.; Bass, R. L., Ill; Johnson, D.  E.; Kruski,  A. W.;  McMahan, C. A.; Montiel, M. M.; Mott, G.
               E.; Wilbur, R. L.; McGill, H. C., Jr. (1980) Atherosclerosis-related responses to cigarette  smoking in
               the baboon. Circulation 61: 1188-1193.
         March 14, 1990                               1-26      DRAFT - DO NOT QUOTE OR CITE

-------
       Rogers, W. R.; Carey, K. D.; McMahan, C. A.; Montiel, M, M.; Mott, G. E.; Wigodsky, H. S.; McGill, H.
               C., Jr. (1988) Cigarette smoking, dietary hyperlipidemia, and experimental atherosclerosis in the baboon.
               Exp. Mol. Pathol. 48: 135-151.

 5     Roth, R. A., Jr.; Rubin, R. J. (1976a) Role of blood flow in carbon monoxide- and hypoxic hypoxia-induced
               alterations in hexobarbital metabolism in rats. Drug Metab. Dispos. 4: 460-467.

       Roth, R. A., Jr.; Rubin, R. J. (1976b) Comparison of the effect of carbon monoxide and of hypoxic hypoxia.  II.
               Hexobarbital metabolism in the isolated,  perfused rat liver. J. Pharmacol. Exp. Ther. 199: 61-66.
10
       Santiago, T. V.; Edelman, N. H. (1976) Mechanism of the ventilatory response to carbon monoxide. J. Clin.
               Invest. 57: 977-986.

       Sheppard, D.; Distefano, S.; Morse, L.; Becker, C. (1986) Acute effects of routine firefighting on lung function.
15             Am. J. Ind.  Med. 9:  333-340.

       Sheps, D. S.; Adams, K. F., Jr.; Bromberg,  P. A.; Goldstein, G. M.; O'Neil, J. J.; Horstman, D.; Koch, G.
               (1987) Lack of effect of low levels of carboxyhemoglobin on cardiovascular function in patients with
               ischemic heart disease. Arch.  Environ. Health 42: 108-116.
20
       Sheps, D. S.; Herbst, M. C.; Hinderliter, A. L.; Adams, K. F.;  Ekelund, L. G.; O'Neil, J. J.; Goldstein, G.
               M.; Bromberg, P. A.; Herdt, J.; Ballenger, M.; Davis, S. M.; Koch, G. (1989) Effects of 4% and 6%
               carboxyhemoglobin on arrhythmia production in patients with coronary artery disease. Submitted for
               publication.
25
       Snella, M.-C.; Rylander, R. (1979) Alteration in local and systemic immune capacity  after exposure to bursts of
               CO. Environ. Res. 20: 74-79.

       Stern, F. B.; Halperin, W. E.; Hornung,  R. W.; Ringenburg, V. L.; McCammon, C. S. (1988) Heart disease
30             mortality among bridge and tunnel officers exposed to carbon monoxide. Am. J. Epidemiol. 128:
               1276-1288.

       Stewart, R. D.; Newton, P. E.; Kaufman, J.; Forster, H. V.; Klein, J. P.;  Keelen, M. H., Jr.; Stewart, D. J.;
               Wu, A.; Hake, C. L. (1978) The effect of a rapid 4% carboxyhemoglobin saturation increase on maximal
35             treadmill exercise. New York, NY: Coordinating Research Council, Inc.; report no.
               CRC-APRAC-CAPM-22-75. Available from: NTIS, Springfield, VA; PB-296627.

       Stupfel, M.; Bouley, G. (1970) Physiological and biochemical effects on rats and mice exposed to small
               concentrations of carbon monoxide for long periods. Ann. N. Y. Acad. Sci. 174: 342-368.
40
       Swiecicki, W. (1973) Wplyw wibracji i treningu fizycznego na przemiane weglowodanowa u szczurow zatrutych
               tlenkiem wegla [The effect of vibration and physical training on carbohydrate metabolism in rats
               intoxicated with carbon monoxide]. Med. Pr. 34: 399-405.

45     Theodore, J.; O'Donnell, R.  D.; Back, K. C. (1971) Toxicological evaluation of carbon monoxide in humans and
               other mammalian species. JOM J. Occup. Med.  13: 242-255.

       U. S. Environmental Protection Agency. (1979) Air quality criteria for carbon monoxide. Research Triangle
               Park,  NC: Office of Health and Environmental Assessment, Environmental Criteria and Assessment
50             Office; EPA report no. EPA-600/8-79-022. Available from: NTIS, Springfield, VA; PB81-244840.
         March 14,  1990                               1-27      DRAFT - DO NOT QUOTE OR CITE

-------
       U. S. Environmental Protection Agency. (1984) Revised evaluation of health effects associated with carbon
               monoxide exposure: an addendum to the 1979 EPA air quality criteria document for carbon monoxide.
               Research Triangle Park, NC: Office of Health and Environmental Assessment, Environmental Criteria
               and Assessment Office; EPA report no. EPA-600/9-83-033F. Available from: NTIS, Springfield, VA;
 5             PB85-103471.

       U. S. Environmental Protection Agency. (1990) National air quality and emissions trends report, 1988. Research
               Triangle Park, NC: Office of Air Quality Planning and Standards; EPA report no. EPA-450/4-90-002.

10     Wallace, L. A.; Ziegenfus, R. C. (1985) Comparison of carboxyhemoglobin concentrations in adult nonsmokers
               with ambient carbon monoxide levels.  J. Air Pollut. Control Assoc. 35: 944-949.

       Weiser, P. C.; Morrill, C. G.; Dickey, D. W.; Kurt, T. L.; Cropp, G. J. A. (1978) Effects of low-level carbon
               monoxide exposure on the adaptation of healthy young men to aerobic work at an altitude of 1,610
15             meters. In: Folinsbee, L. J.; Wagner, J. A.; Borgia, J. F.; Drinkwater, B. L.; Gliner, J. A.; Bedi, J. F.,
               eds. Environmental stress: individual human adaptations. New York, NY: Academic Press, Inc.; pp.
               101-110.

       Weissbecker,  L.; Carpenter, R. D.; Luchsinger, P. C.; Osdene, T. S. (1969) In vitro alveolar macrophage
20             viability: effect of gases. Arch. Environ. Health 18: 756-759.

       Whitmore, R. W.; Jones, S. M.; Rosenzweig, M. S. (1984) Final sampling report for the study of personal CO
               (carbon monoxide) exposure. Research Triangle park,  NC:  U. S. Environmnetal Protection Agency,
               Environmental Monitoring Systems Laboratory; EPA report no. EPA-600/4-84-034. Available from:
25             NTIS, Springfield, VA; PB84-181957.

       Yang, L.; Zhang, W.; He, H.; Zhang, G.  (1988) Experimental studies on combined effects of high temperature
               and carbon monoxide. J. Tongji Med.  Univ.  8: 60-65.

30     Young, J. S.; Upchurch, M. B.; Kaufman, M. J.; Fechter, L. D. (1987)  Carbon monoxide exposure potentiates
               high-frequency auditory threshold shifts induced by noise. Hear.  Res. 26: 37-43.

       Zebro, T.; Wright, E. A.; Littleton,  R. J.; Prentice, A.  I. D.  (1983) Bone changes in mice after prolonged
               continuous exposure to a high concentration of carbon monoxide.  Exp. Pathol. 24: 51-67.
35
         March 14,  1990                              1-28      DRAFT - DO NOT QUOTE OR CITE

-------
            Most of the scientific information selected for review and comment in this document
       comes from the more recent literature published since completion of the previous criteria
       document (U.S. Environmental Protection Agency, 1979). Some of the these newer studies
       were reviewed briefly in the addendum to that document (U.S. Environmental Protection
 5     Agency,  1984a). Emphasis has been placed on studies conducted at or near CO
       concentrations found in ambient air.  Other studies, however, were included if they contained
       unique data, such as the documentation of a previously unreported effect or a mechanism of
       an effect; or if they were multiple-concentration studies designed to provide exposure-
       response relationships relevant to total human exposure to CO. Studies that were presented in
10     the previous criteria document and whose data are still considered relevant are summarized in
       tables or reviewed briefly in the text.  Older studies were considered for discussion in the
       document if they were (1) judged to be significant because of their usefulness in deriving the
       current NAAQS, (2) open to reinterpretation because of newer data, or (3) potentially useful
       in deriving revised standards for CO.  Generally, only published information that has
15     undergone scientific peer review is included in this criteria document.  Some newer studies
       not published in the open literature but meeting high standards of scientific reporting also are
       included.
20     2.2 LEGISLATIVE HISTORY OF NAAQS
            Two sections of the Clean Air Act (CAA) govern the establishment, review, and
       revision of NAAQS.  Section 108 (U.S. Code, 1982) directs the Administrator of the U.S.
       EPA to identify pollutants that reasonably may be anticipated to endanger public health or
       welfare and to issue air quality criteria for them.  These air quality criteria are to reflect the
25     latest scientific information useful in indicating the kind and extent of all identifiable effects
       on public health or welfare that may be expected  from the presence of the pollutant in
       ambient air.
            Section 109(a) of the CAA (U.S. Code,  1982) directs the Administrator of EPA to
       propose and promulgate primary and secondary NAAQS for pollutants identified under
30     Section 108.  Section 109(b)(l) defines a primary standard as one the attainment and
       maintenance of which in the judgment of the Administrator, based on the criteria and

       March 12,  1990                          2-3     DRAFT - DO NOT QUOTE OR  CITE

-------
       allowing for an adequate margin of safety, is requisite to protect the public health. The
       secondary standard, as defined in Section 109(b)(2), must specify a level of air quality the
       attainment and maintenance of which in the judgment of the Administrator, based on the
       criteria, is requisite to protect the public welfare from any known or anticipated adverse
 5     effects associated with the presence of the pollutant in ambient air. Section 109(d) of the
       CAA (U.S. Code, 1982) requires periodic review and, if appropriate, revision of existing
       criteria and standards.  If, in the Administrator's judgment, the Agency's review and revision
       of criteria make appropriate the proposal of new or revised standards, such standards are to be
       revised and promulgated in accordance with Section 109(b).  Alternately, the Administrator
10     may find that revision of the standards is inappropriate and may conclude the review by
       leaving the  existing standards unchanged.
            In keeping with the requirements of the CAA,  the Environmental Criteria and
       Assessment Office of EPA's Office of Health and Environmental Assessment has started to
       review and  revise once again the criteria for CO.  New data on the health and air quality
15     aspects of CO exposure have become available since completion of the previous Air Quality
       Criteria Document (U.S. Environmental Protection Agency, 1979) and an addendum to that
       document (U.S. Environmental Protection Agency,  1984a).
20     2.3 REGULATORY BACKGROUND FOR CARBON MONOXIDE
            NAAQS*
            On April 30, 1971, EPA promulgated identical primary and secondary NAAQS for CO
       at levels of 9 ppm for an 8-h average and 35 ppm for a 1-h average, not to be exceeded more
       than once per year.  The scientific basis for the primary standard, as described in the first
25     criteria document (National Air Pollution Control Administration, 1970), was a study
       suggesting that low levels of CO exposure resulting in carboxyhemoglobin (COHb)
       concentrations of 2 to 3% were associated with neurobehavioral effects in exposed subjects
       (Beard and Wertheim,  1967).
          "This text is excerpted and adapted from "Review of the National Ambient Air Quality Standards for Carbon
30     Monoxide; Final Rule" (Federal Register, 1985).
       March 12, 1990                          2-4      DRAFT - DO NOT QUOTE OR CITE

-------
            In accordance with Sections 108 and 109 of the CAA, EPA has reviewed and revised
       the criteria upon which the existing NAAQS for CO (Table 2-1) are based.  On August 18,
       1980, EPA proposed certain changes in the standards (Federal Register, 1980) based on
       scientific evidence reported in the revised criteria document for CO (U.S. Environmental
 5     Protection Agency, 1979).  Such evidence indicated that the Beard and Wertheim (1967)
       study was no longer considered to be a sound scientific basis for the standard.  Additional
       medical evidence accumulated since 1970, however, indicated that aggravation of angina
       pectoris and  other cardiovascular diseases would occur at COHb levels as low as 2.7 to 2.9%.
       The proposed changes included (1) retaining the 8-h primary standard level of 9 ppm,
10     (2) revising the 1-h primary standard level from 35 ppm to 25 ppm,  (3) revoking the existing
       secondary CO standards (because no adverse welfare effects have been reported at or near
       ambient CO  levels), (4) changing the form of the primary standards from deterministic to
       statistical, and (5) adopting a daily interpretation  for exceedances of the primary  standards, so
       that exceedances would be determined on the basis of the number of days on which the 8- or
15     1-h average concentrations are above the standard levels.
            The 1980 proposal was based in part on health studies conducted by Dr. Wilbert
       Aronow.   In March of 1983 EPA learned that the Food  and Drug  Administration (FDA) had
       raised serious questions regarding the technical adequacy of several studies conducted by Dr.
       Aronow on experimental drugs, leading FDA to reject use of the Aronow drug study data.
20     Therefore, EPA convened an expert committee to examine the Aronow CO studies before any
       final decisions were made on the NAAQS for CO.  The committee concluded that EPA
       should not rely on Dr. Aronow's data due to concerns regarding the research, which
       substantially  limited the validity and usefulness of the results.
            An addendum to the 1979 criteria document for CO (U.S. Environmental Protection
25     Agency, 1984a) reevaluated the scientific data concerning health effects associated with
       exposure  to CO at or near ambient exposure levels in light of the committee recommendations
       and taking into account new findings reported beyond those previously reviewed.  (These data
       are summarized in the following section.)  On September 13,  1985, EPA issued a final notice
       (Federal Register,  1985) announcing retention of the existing primary NAAQS for CO and
30     rescinding the secondary NAAQS for CO.
       March 12, 1990                          2-5     DRAFT - DO NOT QUOTE OR CITE

-------
                 TABLE 2-1.  NATIONAL AMBIENT AIR QUALITY STANDARDS
                                  FOR CARBON MONOXIDE

 5     Date of Promulgation                 Primary NAAQS                Averaging Time
       September 13, 1985                 9 ppm' (10 mg/m3)                   ITh1
                                        35 ppm' (40 mg/m3)                   l-hb
10
       *  1 ppm = 1.145 mg/m3, 1 mg/m3 = 0.873 ppm @ 25°C, 760 mm Hg.
       b  Not to be exceeded more than once per year.
       See glossary of terms and symbols for abbreviations and acronyms.
15
       2.4  SCIENTIFIC BACKGROUND FOR THE CURRENT CARBON
20          MONOXIDE NAAQS
            The following is a summary of the scientific basis for the current CO NAAQS. These
       key points were derived from a revised evaluation of the health effects of CO that was
       released as an addendum (U.S. Environmental Protection Agency, 1984a) to the previous air
       quality criteria document for CO (U.S. Environmental Protection Agency, 1979).
25
       2.4.1  Mechanisms of Action
            The binding of CO to hemoglobin, producing COHb and decreasing the oxygen-
       carrying capacity of blood,  appears to be the principal mechanism of action underlying the
       induction  of toxic effects of low-level CO exposures.  The precise mechanisms by which toxic
30     effects are induced via COHb formation are not understood fully, but likely include the
       induction  of a hypoxic state in many tissues of diverse organ systems.  Alternative or
       secondary mechanisms of CO-induced toxicity (besides COHb) have been hypothesized, but
       none have been demonstrated to operate at relatively low (near-ambient) CO-exposure levels.
       Blood COHb levels, then, currently are accepted as representing a useful physiological marker
35     by which  to estimate internal CO burdens due to the combined contribution of
       (1) endogenously derived CO and (2) exogenously derived CO resulting from exposure to
       external sources of CO.  COHb levels likely to result from particular patterns (concentrations,
       durations, etc.) of external CO exposure can be estimated reasonably well from equations
       developed by Coburn et al.  (1965) as demonstrated in Figure 2-1.
       March 12, 1990                          2-6     DRAFT - DO NOT QUOTE OR CITE

-------
  0)
  V
  a
 .a
 O
 o
14
13-
12-
11 -
10-
 9-
 8-
 7-
 6-
 5-
 4-
 3-
 2-
 1 -
 0
                                          1  h, 10 t/min
                      20
                                40             60
                         CARBON MONOXIDE, ppm
80
100
Figure 2-1. Relationship between carbon monoxide exposure and carboxyhemoglobin levels
in the blood. Predicted COHb levels resulting from 1- and 8-h exposures to carbon
monoxide at rest (10 L/min) and with light exercise (20 L/min) are based on the Coburn-
Forster-Kane equation (Coburn et al., 1965) using the following assumed parameters for
nonsmoking adults:  altitude = 0 ft; initial COHb level = 0.5%; Haldane constant = 218;
blood volume = 5.5 L; hemoglobin level = 15 g/100 ml; lung diffusivity = 30 ml/torr/min;
endogenous rate = 0.007 ml/min. See glossary of terms and symbols for abbreviations and
acronyms.
March 12, 1990
                                  2-7     DRAFT - DO NOT QUOTE OR CITE

-------
       2.4.2  Carbon Monoxide Exposure Levels
            Evaluation of human CO-exposure situations indicates that occupational exposures in
       some workplaces or exposures in homes with faulty combustion appliances can exceed
       100 ppm CO, often leading to COHb levels of 10% or more with continued exposure. In
 5     contrast,  such high exposure levels are encountered much less commonly by the general
                                                          /
       public exposed under ambient conditions.  More frequently, exposures to less than 25 to
                                                       V
       50 ppm CO for any extended period of time occur among the general population and, at the
       low exercise levels usually engaged in under such circumstances, the resulting COHb levels
       most typically remain 2 to 3 % among nonsmokers. Those levels can be compared to the
10     physiologic norm for nonsmokers, which is estimated to be in the range of 0.3 to 0.7%
       COHb.  Baseline COHb concentrations in smokers, however, average 4% with a usual range
       of 3 to 8%, reflecting absorption of CO from inhaled smoke.

       2.4.3  Health Effects of Low-Level Carbon  Monoxide Exposures
15          Four types of health effects reported or hypothesized to be associated with CO
       exposures (especially those producing COHb levels below 10%) were evaluated in the last
       review of the CO NAAQS (U.S. Environmental Protection Agency, 1984a):  (1) cardio-
       vascular effects, (2) neurobehavioral effects, (3) fibrinolysis  effects, and (4) perinatal effects.
       Data available at that time (Table 2-2) demonstrated an association between cardiovascular
20     and neurobehavioral effects at relatively low-level CO exposures.  Much less clear evidence
       existed to indicate that other types of health effects were associated with low-level CO
       exposures.

       2.4.3.1  Cardiovascular Effects
25          In regard to cardiovascular effects, decreased oxygen uptake and resultant decreased
       work capacity under maximal exercise conditions clearly have been shown to occur in healthy
       young adults starting at 5.0% COHb; and several studies observed small decreases in work
       capacity at COHb levels as low as 2.3 to 4.3%.  These cardiovascular effects may have health
       implications for the general population in terms of potential curtailment of certain physically
30     demanding occupational or recreational activities under circumstances of sufficiently high CO-
       exposure. However, of greater concern at more typical ambient CO-exposure levels were

       March 12, 1990                           2-8     DRAFT - DO NOT QUOTE OR CITE

-------
            TABLE 2-2.  LOWEST OBSERVED EFFECT LEVELS FOR HUMAN HEALTH
           EFFECTS ASSOCIATED WITH LOW-LEVEL CARBON MONOXIDE EXPOSURE
              Effects
                                         COHb
                                     concentration,
                                        percent*
     References
10
15
20
25
30
35
40
45
50
55
Statistically significant
decreased (3-7%) work time
to exhaustion in exercising,
young, healthy men

Statistically significant
decreased exercise capacity
(i.e., shortened duration of
exercise before onset of
pain) in patients with angina
pectoris and increased
duration of angina attacks

No statistically significant
vigilance decrements after
exposure to CO
Statistically significant
decreased maximal oxygen
consumption and exercise time
during strenuous exercise in
young, healthy men

Statistically significant
diminution of visual percep-
tion, manual dexterity,
ability to learn, or
performance in complex
sensorimotor tasks (such
as driving)
        Statistically significant
        decreased maximal oxygen
        consumption during strenuous
        exercise in young, healthy men
                                                2.3-4.3
                                                2.9-4.5
                                                Below 5
                                                5-5.5
                                                5-17
                                         7-20
Horvath et al. (1975)
Drinkwater et al. (1974)
Anderson et al. (1973)
Haider et al. (1976)
Winneke (1974)
Christensen et al. (1977)
Benignus et al. (1977)
Putz et al. (1976)

Klein et al. (1980)
Stewart et al. (1978)
Weiser et al. (1978)
Bender et al. (1971)
Schulte (1973)
O'Donnell et al. (1971)
McFarland et al. (1944)
McFarland (1973)
Putz et al. (1976)
Salvatore (1974)
Wright et al. (1973)
Rockwell and Weir (1975)
Rummo and Sarlanis (1974)
Putz et al. (1979)
Putz (1979)

Ekblom and Huot (1972)
Pirnay et al. (1971)
Vogel and Gleser (1972)
'The physiologic norm (i.e., COHb levels resulting from the normal catabolism of hemoglobin and other heme-
 containing materials) has been estimated to be in the range of 0.3 to 0.7 % (Coburn et al., 1963). See glossary
 of terms and symbols for abbreviations and acronyms.

Source: U.S. Environmental Protection Agency (1984b).
  March 12, 1990
                                      2-9      DRAFT - DO NOT QUOTE OR CITE

-------
        certain cardiovascular effects (i.e., aggravation of angina symptoms during exercise) likely to
        occur in a smaller, but sizeable, segment of the general population.  This group, chronic
        angina patients, is presently viewed as the most sensitive risk group for CO-exposure effects,
        based on evidence for aggravation of angina occurring  in patients at COHb levels of 2.9 to
 5      4.5%. Such aggravation of angina is thought to represent an adverse health effect for several
        reasons articulated in the 1980 proposal preamble (Federal Register, 1980), and the Clean Air
        Scientific Advisory Committee (CASAC) concurred with EPA's judgment on this matter.
        Dose-response relationships for cardiovascular effects in coronary artery disease patients
        remain to be defined more conclusively, and the possibility cannot be ruled out at this time
10      that such effects may occur at levels below 2.9% COHb (as hinted at by the results  of the
        now-questioned Aronow studies).  Therefore, new studies published since the last review
        cycle are evaluated in this revised criteria document to  determine the effects of CO on
        aggravation of angina at levels in the range of 2 to 6%  COHb.

15      2.4.3.2  Neurobehavioral Effects
             No reliable evidence demonstrating decrements in neurobehavioral function in healthy,
        young adults has been reported at COHb levels below 5%.  Results of studies conducted at or
        above 5% COHb are equivocal.  Much of the research  at 5% COHb did not show any effect
        even when behaviors similar to those affected in other studies at higher COHb levels were
20      involved. However, investigators failing to find CO decrements at 5% or higher COHb
        levels may have utilized tests not sufficiently sensitive to reliably detect small effects of CO.
        From the empirical evidence, then, it can be said that COHb levels >5% do produce
        decrements in neurobehavioral function. It cannot be said confidently, however, that COHb
        levels lower than 5% would be without effect.  One important point made in the 1979 criteria
25      document should be reiterated here. Only young, healthy adults have been studied using
        demonstrably sensitive tests and COHb levels at 5% or greater.  The question of groups at
        special risk for neurobehavioral effects of CO, therefore, has not been explored.  Of special
        note are  those individuals who are taking drugs that have primary or secondary depressant
        effects which would be expected to exacerbate CO-related neurobehavioral decrements.  Other
30      groups at possibly increased risk for CO-induced neurobehavioral effects are the aged and ill,
        but these groups have not been evaluated for such risk.

        March 12, 1990                          2-10     DRAFT - DO NOT QUOTE OR CITE

-------
       2.4.3.3 Other Health Effects
            Only relatively weak evidence points toward possible CO effects on fibrinolytic activity,
       generally only at rather high CO-exposure levels. Similarly, whereas certain data also suggest
       that perinatal effects (e.g., reduced birth weight, slowed postnatal development, Sudden
       Infant Death Syndrome) are associated with CO exposure, insufficient evidence presently
       exists by which to either qualitatively confirm such an association in humans or to establish
       any pertinent exposure-effect relationships.
10     2.5 CRITICAL ISSUES IN REVIEW OF THE NAAQS FOR CARBON
            MONOXIDE
            Based on the scientific evidence currently evaluated in air quality criteria documents
       (U.S. Environmental Protection Agency, 1979, 1984a), potentially adverse health effects of
       CO have been demonstrated to occur at COHb levels in the range of 2.3 to 20% (see
15     Table 2-2).  However, several critical issues have developed during the current review of the
       scientific criteria for CO air quality standards that will need to be resolved in order to
       determine the extent to which adverse effects are occurring in the population, particularly at
       the lower COHb levels of greatest interest to standard-setting (<5 percent). The following
       section will focus on  specific issues pertaining to (1) exposure assessment in the general
20     population, including the measurement of CO in ambient air and in blood; (2) mechanisms of
       action of CO; (3) health effects from exposure to CO; and (4) groups of individuals
       considered to be at greatest risk to CO at ambient or near-ambient exposure levels.

       2.5.1  Exposure Assessment in the Population
25          The 1986 and 1987 trends in ambient air quality reported by EPA (U.S. Environmental
       Protection Agency, 1988, 1989a) summarize fixed-site monitoring data for carbon  monoxide
       but only focus on 8-h averages. The rationale for this approach is that the 8-h standard
       (9 ppm) is typically the controlling standard and the 1-h  standard (35 ppm) rarely is exceeded.
       For example, in  1987 there were only four exceedances of the 1-h standard for the entire
30     United States and in each case the 8-h standard was exceeded by a greater percentage on that
       day. In 1988, only two areas (Denver, CO, and Steubenville, OH) exceeded the 1-h CO

       March 12, 1990                          2-11     DRAFT - DO NOT QUOTE OR CITE

-------
       standard, while 44 areas failed to meet the 8-h standard (U.S. Environmental Protection
       Agency, 1989b).
            Ambient CO-concentration data from fixed-site monitors alone will not necessarily give
       a good estimate of potential total exposure to the population, based on experience from the
 5     Denver and Washington, DC human-exposure field studies using personal monitors.  It is
       estimated that over 10% of the residents in Denver and 4% of the residents in Washington,
       DC, were exposed to CO levels above 9 ppm for 8 h during the winter of 1982-83 (Akland
       et al.,  1985). The effects of personal activity, indoor sources, and time spent commuting
       contribute greatly to an individual's total exposure to CO.  Available 1-h CO concentrations
10     taken at fixed-site monitors in these field studies did not correlate well (0.14 < r < 0.27) with
       measurements made by personal monitors.
            The best available study for determining relevant exposure to the most susceptible target
       population, that is,  individuals with ischemic heart disease (IHD), is the work of Lambert  and
       Colome (1988).  A total of 36 nonsmoking men with IHD were followed during personal-
15     exposure monitoring.  A wide range of peak exposures to CO were measured.  The highest
       CO exposures were found while the subjects were commuting and when the subjects were
       near internal combustion engines.  For example, CO exposures on freeways in Los Angeles
       averaged 10 to 12 ppm. The average personal exposure for all  time spent in automobiles was
       8.6 ppm with a maximum 1-min average of 239 ppm.  Concentrations of CO, averaging 7.9
20     ppm, also were found in parking lots, parking structures,  service stations, and motor repair
       facilities.  Residential CO exposure was much lower, averaging 2.0 ppm.  In typical outdoor
       residential activities, transient peaks as high as 134 ppm were observed for woodcutting with
       a gas-powered chain saw and 226 ppm for gardening activity where a two-stroke, gasoline-
       powered engine was utilized. Exposures under these conditions would be expected, based on
25     equations developed by Coburn et al. (1965), to cause COHb levels in excess  of 2.5%.
            The best indicator of exposure to CO continues to be the direct  measurement of COHb
       in blood.  There are, however, several issues regarding measurement techniques for COHb
       that have been raised during the current review of the CO air quality  criteria.  For many
       years,  routine clinical laboratory measurements of COHb commonly have been made using
30     the IL  182 and it successor,  the IL 282 CO-Oximeter (Instrumentation Laboratory, Inc.,
       Lexington, MA) which is a spectrophotometric instrument.  Optical methods of COHb

       March 12, 1990                          2-12     DRAFT - DO NOT QUOTE OR CITE

-------
        measurement, however, are limited in sensitivity, particularly in the range of 0 to 5% where
        the lowest observed health effects associated with CO exposure have been described.  Other,
        more sensitive techniques require the release of CO from hemoglobin into a gas phase that
        can be detected directly.  One method for COHb measurement that has become more widely
 5      used in laboratory settings is gas chromatography.  Recent efforts to compare COHb
        measurement by spectrophotometry versus gas chromatography have indicated that the high
        correlation over a wide range of concentrations (0 to >20%) becomes much worse at COHb
        levels  <5% because of an apparent instrument offset or potential error.  Thus, there has been
        concern about the relative accuracy and precision of the COHb measurements at levels that
10      are of particular concern to the CO NAAQS review. Further, ongoing work is needed in
        order to determine (1) which method should be used to accurately quantify low levels of
        COHb; (2) if there is a scientifically acceptable way to compare COHb measurements made
        by different instruments across different laboratories; and (3) the relationship of measured
        COHb values to those derived from modeling efforts based on actual CO exposures in the
15      general population.
            Most "real-life" exposures to CO are to concentrations that vary with time and those
        exposures are experienced by people with differing physiological attributes and at varying
        exercise levels.  Direct measurements of COHb are not readily available in the general
        population exposed to  CO under these conditions.  Mathematical models, therefore, have been
20      developed to predict COHb levels from known CO exposures under a variety of
        circumstances.  The most used model for COHb formation is still the Coburn-Forster-Kane
        equation (CFKE) developed by Coburn et al. (1965).  The COHb levels predicted by this
        equation generally have been accepted as the best available estimates of COHb levels likely to
        result from varying CO concentrations, exposure durations, and  exercise levels.  Further
25      research, however, is needed to evaluate the predictive capabilities of the CFKE in individuals
        exposed to low  concentrations  of CO leading to COHb levels of less than 10%.  Of particular
        interest is the variation of predicted COHb in a population whose pattern of CO exposure
        involves frequent concentration variations.  In addition, the CFKE needs to be evaluated for
        applicability to CO-susceptible subjects, such as patients with cardiovascular or pulmonary
30      disease. Clinical evaluation of CO uptake by these individuals should be considered.
       March 12, 1990                          2-13     DRAFT - DO NOT QUOTE OR CITE

-------
            Epidemiology studies have suggested the possibility that increased mortality from heart
       attacks and increased cardiovascular complaints may be associated with elevated ambient
       concentrations of CO.  Unfortunately, due to inadequate characterization of exposure as well
       as other limitations, inconclusive results have been obtained from existing studies.  The
 5     availability of both personal-exposure monitors for CO and ambulatory EKG monitoring
       techniques have made it possible to design epidemiology studies to determine whether ambient
       CO exposures are related to serious or irreversible cardiovascular effects. It would be
       desirable, therefore, to obtain CO exposure data on CO-susceptible individuals in order to
       characterize their risk from elevated levels of COHb. Potentially susceptible individuals
10     include infants, the elderly, and patients with known cardiovascular diseases.

       2.5.2 Mechanisms of Action of Carbon Monoxide
            The accepted mechanisms of action underlying the potentially toxic effects of low-level
       CO exposure continue to be the decreased oxygen-carrying capacity of blood and subsequent
15     interference of oxygen release at the tissue level that is caused by the binding of CO with
       hemoglobin, producing COHb (Figure 2-2).  The resulting impaired delivery of oxygen can
       interfere with cellular respiration and cause tissue hypoxia.
            Review of the newer information on mechanisms of action of CO has focused on the
       possibility that secondary mechanisms that also can impair cellular respiration may be
20     occurring at relatively low (near-ambient) CO-exposure levels.  Approximately 10 to 50% of
       the total-body burden of CO can be distributed to extravascular sites, suggesting that intra-
       cellular uptake of CO may contribute to CO-induced toxicity.  It is uncertain, however, if
       intracellular uptake of CO occurs at low levels of COHb or if it would be likely to contribute
       to the physiological effects of CO.
25          Carbon monoxide will bind to  intracellular hemoproteins such as myoglobin  (Mb),
       cytochrome oxidase, mixed function oxidases (e.g., cytochrome P-450), tryptophan
       oxygenase, and dopamine nydroxylase. Binding to CO would be favorable under  conditions
       of low intracellular partial pressure of oxygen (PO2), particularly in brain and myocardial
       tissue where intracellular PO2 decreases with increasing COHb levels. The most likely
30     hemoproteins to be inhibited functionally at relevant levels of COHb  are Mb, found
       predominantly in heart and skeletal muscle, and cytochrome oxidase.  The physiological

       March 12, 1990                          2-14      DRAFT - DO NOT QUOTE OR CITE

-------
       MECHANISMS  OF  ACTION  OF  CARBON MONOXIDE
      Source
      Internal
      storage
      compartment
      Target
      organ
                           External Exposure
                       Carbon
                       Monoxide
Ambient
Indoor
Occupational
             Halogenated
             Hydrocarbons
                                   e.g.
                                    Internal
Endogenous
Production
of CO
                                                           1
                                 Extravascular
                                                     CO + Mb
                                       CO + Cyt
                      Decreased 02 Delivery
Decreased  Og-carrying capacity
L-shift  Hb02 dissociation curve
                          I
                      Compensatory
                      vasodilation
                      and increased
                      blood  flow
                      to maintain
                      Og consumption
                 Decreased cellular respiration
                     i
                 Tissue hypoxia
                                       Ischemia cascade
Figure 2-2.  Currently accepted or proposed mechanisms of action of carbon monoxide
resulting from external exposure sources can interfere with cellular respiration and cause
tissue hypoxia (see text for details). See glossary of terms and symbols for abbreviations and
acronyms.
March 12, 1990
               2-15     DRAFT - DO NOT QUOTE OR CITE

-------
       significance of CO uptake by Mb is uncertain at this time but sufficient concentrations of
       carboxymyoglobin (COMb) could potentially limit maximal oxygen uptake of exercising
       muscle.  Although there is suggestive evidence for significant binding of CO to cytochrome
       oxidase in heart and brain tissue, it is unlikely that any significant CO binding would occur at
 5     low COHb levels.  Therefore, further research still is needed to determine if secondary,
       intracellular mechanisms will occur at exposure concentrations found in ambient air.

       2.5.3  Health Effects from Exposure to Carbon Monoxide
       2.5.3.1  Effects on the Cardiovascular System
10          Scientific support for the current NAAQS  for CO is based primarily  on studies of
       patients with stable angina pectoris (chest pain)  from coronary artery disease.   Although it is
       assumed that the development of angina reflects adverse effects of CO on myocardial
       metabolism, more  specific research supporting the validity of this assumption is needed.  For
       example, little is known about the reproducibility  or reoccurrence of this disease.  Time to
15     onset of angina and the duration of angina are measurable outcomes that need to be defined
       more precisely.  Research also is needed on more objective measures of myocardial ischemia,
       such as continuous EKG tracing for ST depression and arrhythmias, and on measurement of
       ventricular function using a gamma camera or thallium scan.
            In view of questions concerning the validity  of angina studies by Aronow et al.  reviewed
20     in the previous criteria document, additional data  were clearly needed in order to (1)  provide
       more reliable dose-response information in individuals with stable angina,  (2) allow a better
       determination of the level of COHb necessary to cause adverse effects in the sensitive
       population, and (3) ultimately set an appropriate level  for the CO NAAQS. In response to
       this need, additional studies recently have been  completed by a number of independent
25     laboratories to identify the relationship between COHb and aggravation of preexisting chronic
       heart disease.  Four of these studies now have been published (Sheps et al., 1987; Adams
       et al., 1988; Kleinman et al.,  1989; Allred et al., 1989).  Collectively, all four studies
       provide new information on the likelihood that patients exposed to CO will experience angina
       earlier during exercise when compared to clean air exposure.  Levels of COHb across the
30     studies range from 2.9 to 5.9%, as measured by the spectrophotometric method (CO-
       Oximeter).  An evaluation of these data is provided in this document as part of the overall

       March 12, 1990                          2-16    DRAFT - DO NOT QUOTE OR CITE

-------
        review of the scientific basis for the CO NAAQS. Any potential differences in the results
        between these studies primarily will be due to either the patient population studied or to the
        experimental design of the study itself.
             Heart attack is the leading cause of death in the United States. In 1985 alone, over
 5      540,000 deaths were attributed to coronary artery disease and over 4.8 million people alive at
        that time were estimated to have a history of heart attack, angina, or both (American Heart
        Association,  1988).  Today that estimate may be as high as 6 to 7 million individuals inflicted
        with coronary artery disease.  A major question that will become important in the evaluation
        of all the clinical studies involving subjects with coronary heart disease is whether the study
10      population is representative of this broad group of patients with angina and, therefore, is
        applicable  to the subpopulation of potentially susceptible individuals that are exposed routinely
        to ambient levels of CO. The possibility of studying the effects of CO in a more
        representative group  of patients with coronary heart disease should be  investigated. Recent
        changes in the treatment of coronary artery disease indicate that the sensitive subpopulation of
15      angina patients may be changing from one of untreated patients to one of angina patients who
        have had coronary artery bypass or balloon angioplasty. The susceptibility of this new
        population to CO may not be the same.  In addition, there is a greater likelihood of increased
        risk to CO exposure in a virtually unknown group of individuals who have silent ischemia (no
        symptomatic episodes of chest pain).
20           Additional research is needed to  determine dose-response relationships for the acute
        effects of CO in other potentially  susceptible groups.  Patients with arteriosclerosis of the
        arteries of  the lower limbs who develop intermittent claudication are analogous to patients
        with angina and could be studied in  a  similar manner.  Research is needed to determine dose-
        response relationships for cardiovascular effects in individuals with ventricular arrhythmias.
25      Patients with anemia may be susceptible to increased levels of COHb,  because CO would
        further reduce the already compromised arterial oxygen content of the blood.  Patients with
        chronic obstructive pulmonary disease and those with congestive heart failure also should be
        studied to determine if they are at increased risk to low levels of CO exposure.
             Other cardiovascular effects  of low-level CO exposure, particularly with prolonged or
30      chronic exposure, have not been demonstrated.  Previous studies on laboratory animals that
        were reviewed in the last criteria document failed to clearly link CO exposure with

        March 12,  1990                          2-17     DRAFT - DO NOT  QUOTE OR CITE

-------
        atherogenesis and the development of arteriosclerosis. Newer data published since then still
        fail to prove conclusively an atherogenic effect of exposure to low concentrations of CO
        despite strong evidence from epidemiology studies showing an association between cigarette
        smoke and increased risk for arteriosclerosis.  Other components of cigarette smoke (e.g.,
 5      nicotine) as well as other risk factors (e.g., diet) also may promote atherogenesis, making it
        difficult to attribute the atherogenic effects of cigarette smoke to CO alone.

        2.5.3.2  Neurobehavioral Effects
             Neurobehavioral effects of CO exposure, such as changes in (1) hand-eye coordination
10      (compensatory tracking), (2) detection of infrequent events (vigilance), and (3) visual system
        sensitivity have been reported in healthy young adults at COHb levels as low as 5%.  These
        effects at low CO-exposure concentrations, however, have been very small and somewhat
        controversial.  The newer data on neurobehavioral effects of CO discussed in this document
        apparently have provided little help in resolving this controversy. Nevertheless, the potential
15      consequences of a lapse of coordination, vigilance, and visual sensitivity in the performance
        of critical tasks by operators of machinery such as public transportation vehicles could be
        serious.  Therefore, additional research  is necessary to provide a better understanding of the
        mechanisms of action of CO and compensatory changes in the vascular bed that may act to
        maintain an adequate oxygen supply to the brain.
20           Certain subgroups of the population are at increased risk from the neural and behavioral
        effects of elevated COHb.  For example, any condition that would reduce oxygen supply to
        the brain also would potentially exacerbate the effects of CO exposure.  A very large
        subgroup that is known to have a reduced oxygen supply to the brain is the aged. Therefore,
        it is important to determine COHb dose-response functions for neurobehavioral variables in
25      older subjects.   Other conditions that might reduce oxygen supply to the brain include certain
        cerebrovascular, cardiovascular, and pulmonary disease states mentioned above.
             Another large subgroup that may be at increased risk from  neurobehavioral effects of
        CO exposure are those people who take prescription or over-the- counter medications such as
        antihistamines, sedatives, antipsychotics, antiseizure drugs, antiemetics, and analgesics, that
30      reduce alertness or motor abilities.  The effects of ethanol, caffeine, nicotine, and other
        nonprescription drugs should not be overlooked. Such individuals already would be affected

        March 12, 1990                          2-18     DRAFT - DO NOT QUOTE OR CITE

-------
        behaviorally so that any further impairment due to elevated COHb might have serious
        consequences.

        2.5.3.3  Perinatal Effects
 5           The fetus and newborn infant are theoretically susceptible to CO exposure for several
        reasons.  Fetal circulation is likely to have a higher COHb level than the maternal circulation
        due to differences in uptake and elimination of CO from fetal hemoglobin.  Since the fetus
        also has a lower oxygen tension in the blood than adults, any further drop in fetal oxygen
        tension due to the presence of COHb could have a potentially serious effect. The newborn
10      infant with a comparatively high rate of oxygen consumption and lower hemoglobin blood
        oxygen transport capacity than most  adults also would be potentially susceptible to the
        hypoxic effects of increased COHb.  Newer data from laboratory animal studies on the
        developmental toxicity of CO suggest that prolonged exposure to high levels (> 100 ppm)  of
        CO during gestation may produce a reduction in birthweight, cardiomegaly, and delayed
15      behavioral development.  Human data are scant and more difficult to evaluate, but further
        research is warranted. Therefore, additional studies are needed in order to determine if
        chronic exposure to CO, particularly at low, near-ambient levels, can compromise the already
        marginal conditions existing in the fetus and newborn infant.
             The effects of CO on maternal-fetal relationships are not understood well.  In addition to
20      fetuses and newborn infants, pregnant women also represent a susceptible group because
        pregnancy is associated with increased alveolar ventilation and an increased rate of oxygen
        consumption that serves to increase the rate of CO uptake from inspired air.  Perhaps a more
        important factor is that pregnant women experience hemodilution due to the disproportionate
        increase in plasma volume as compared to erythrocyte volume.  This group, therefore, should
25      be studied to evaluate the effects of CO exposure and elevated COHb levels.

        2.5.4 Population Groups at Greatest Risk for Ambient CO Exposure
              Effects
             Angina patients or others with obstructed coronary arteries, but not yet manifesting overt
30      symptomatology of coronary artery disease, appear to be best established as a sensitive group
        within the general population that is  at increased risk for experiencing health effects (i.e.,

        March 12, 1990                          2-19     DRAFT - DO NOT QUOTE OR CITE

-------
       exacerbation of cardiovascular symptoms) of concern at ambient or near-ambient CO-
       exposure levels. Several other probable risk groups were identified: (1) fetuses and young
       infants; (2) pregnant women;  (3) the elderly, especially those with compromised
       cardiopulmonary or cerebrovascular functions; (4) individuals with obstructed coronary
 5     arteries, but not yet manifesting overt symptomatology of coronary artery disease;
       (5) individuals with congestive heart failure; (6) individuals with peripheral vascular or
       cerebrovascular disease; (7) individuals with hematological diseases (e.g., anemia)  that affect
       oxygen-carrying capacity or transport in the blood; (8) individuals with genetically unusual
       forms of hemoglobin associated with reduced oxygen-carrying capacity; (9) individuals with
10     chronic obstructive lung diseases; (10) individuals using medicinal or recreational drugs
       having central nervous system (CNS) depressant properties; (11) individuals exposed to other
       pollutants (e.g., methylene chloride) that increase endogenous formation of CO; and
       (12) individuals who have  not been adapted to  high altitude and are exposed to a combination
       of high altitude and CO. However, little empirical evidence currently is available by which
15     to specify health effects associated with ambient or near-ambient CO exposures in these
       probable risk groups.
        2.6  CARBON MONOXIDE POISONING
20           The majority of this document deals with the relatively low concentrations of CO that
        induce effects in humans at or near the lower margin of detection by current medical
        technology.  Yet, the health effects associated with exposure to this pollutant range from the
        more subtle cardiovascular and neurobehavioral effects at low-ambient concentrations, as
        identified in the preceding sections, to unconsciousness and death after prolonged chronic
25      exposure or after acute exposure to high concentrations of carbon monoxide.  The morbidity
        and mortality resulting from the latter exposures are described briefly here to complete the
        picture of CO exposure in present-day society.
             Carbon monoxide is responsible for more than half of the fatal poisonings that are
        reported in the United States each year (National Safety Council, 1982).  At sublethal levels,
30      CO poisoning occurs in a small but important fraction of the population.  Certain conditions
        exist  in both the indoor and outdoor ambient environments that cause a small percentage of

        March 12, 1990                          2-20     DRAFT - DO NOT QUOTE OR CITE

-------
        the population to become exposed to dangerous levels of CO.  Outdoors, concentrations of
        CO are highest near intersections, in congested traffic, near exhaust gases from internal
        combustion engines and from industrial combustion sources, and in poorly ventilated areas
        such as parking garages and tunnels. Indoors, CO concentrations in the workplace or in
 5      homes that have faulty appliances or downdrafts and backdrafts have been measured in excess
        of 100 ppm, resulting in COHb levels of greater than 10% for 8 h of exposure.  In addition,
        CO is found in the smoke produced by all types of fires.  Of the 6000 deaths from burns in
        the United States each year, more than half are related to inhalation injuries where victims die
        from CO poisoning, hypoxia, and smoke inhalation (Heimbach and Waeckerle, 1988).
 10           Carbon monoxide poisoning is not new, although more attention to this problem has
        been addressed recently in the scientific literature as well as in the popular media.  The first
        scientific studies of the hypoxic effects of CO were described by Claude Bernard (1865).  The
        attachment of CO to hemoglobin, producing carboxyhemoglobin, was evaluated by Douglas
        et al. (1912), providing the necessary tools for studying man's response to CO.  During the
 15      next half century, numerous studies were conducted with the principal emphasis being on high
        concentrations of COHb.  Carbon monoxide poisoning as an occupational hazard (Grut, 1949)
        received the greatest attention due to the increased use of natural gas and the potential for
        leakage of exhaust fumes in homes and industry.  Other sources of CO have become more
        important and more insidious.  The clinical picture of CO poisoning, as described  by Grut
20      (1949), relate primarily to the alterations in cardiac and central nervous system function due
        to the extreme hypoxia induced.
            Mortality from carbon monoxide exposure is high.  In 1985, 1365 deaths due to CO
        exposure were reported in England and Wales (Meredith and Vale, 1988).  In the United
        States, more than 3800 people die annually from CO (accidental and intentional), and more
25      than 10,000 individuals seek medical attention or miss at least one day of work because of a
        sublethal exposure  (U.S. Centers  for Disease Control, 1982). The per capita mortality and
        morbidity statistics for CO are surprisingly similar for the Scandinavian countries and for
        Canada, as well. However, not all instances of CO poisoning are reported and complete up-
        to-date data are difficult to obtain. Often the individuals suffering from CO poisoning are
30      unaware of their exposure because symptoms are similar to those associated with the flu or
        with clinical depression. This may result in a significant number of misdiagnoses by medical

        March 12,  1990                          2-21     DRAFT - DO NOT QUOTE OR CITE

-------
       professionals (Heckerling et al., 1988, 1987; Kirkpatrick, 1987; Dolan et al., 1987; Barrett
       et al., 1985; Fisher and Rubin, 1982; Grace and Platt, 1981).  The precise number of
       individuals who have suffered from CO intoxication, therefore, is not known but it is
       certainly larger than the mortality figures indicate.  Nonetheless, the reported literature
 5     available for review indicates the seriousness of this problem.
             The symptoms, signs, and prognosis of acute poisoning correlates poorly with the level
       of COHb measured at the time of arrival at the hospital (Meredith and Vale, 1988).
       Carboxyhemoglobin levels below  10% usually are not associated with symptoms.  At the
       higher COHb saturations of 10 to 30%, neurological symptoms of CO poisoning can occur,
10     such as headache, dizziness, weakness, nausea, confusion, disorientation, and visual
       disturbances. Exertional dyspnea, increases in pulse and respiratory rates, and syncope are
       observed with continuous exposure producing COHb levels in excess of 30 to 50%. When
       COHb levels are higher than 50%, coma, convulsions, and cardiorespiratory arrest may
       occur.
15           Different individuals experience very different clinical manifestations of CO poisoning
       and, therefore,  have different outcomes even under similar exposure conditions.  Norkool and
       Kirkpatrick (1985) found that COHb levels in individuals who had never lost consciousness
       ranged from 5 to 47%.  In  individuals who were found unconscious but regained
       consciousness at hospital arrival, the range was 10 to 64%; for those remaining unconscious,
20     COHb levels varied from 1 to 53%. These data clearly indicate that COHb saturations
       correlate poorly with clinical status and,  furthermore, have little prognostic  significance.
             The level of CO in the tissues may have an equal or greater impact on the clinical
       status of the patient than the blood level of CO (Broome  et al., 1988).  For example a short
       exposure to CO at high ambient concentrations may allow insufficient time for significant
25     increases in tissue levels of CO to occur.  The syncope observed in these individuals may be
       the result of simple hypoxia with rapid recovery despite high COHb levels.  A prolonged
       exposure to low concentrations of CO prior to hospital arrival  may allow sufficient uptake of
       CO by tissues to inhibit the function of intracellular compounds such as myoglobin. This
       effect, in combination with hypoxia, may cause irreversible central nervous system or cardiac
30     damage.
        March 12, 1990                           2-22    DRAFT - DO NOT QUOTE OR CITE

-------
             Patients with CO poisoning respond to treatment with 100% oxygen (Pace et al., 1950).
        If available, treatment with hyperbaric oxygen (HBO) at 2.5 to 3 times atmospheric pressure
        for 90 min is preferable (Myers, 1986), but the precise conditions requiring treatment have
        been a topic of debate in the literature (Thorn and Keim, 1989; Roy et al., 1989; Raphael
 5      et al., 1989; Brown et al., 1989; James, 1989; Van Hoesen et al., 1989; Broome et al.,
        1988; Norkool and Kirkpatrick, 1985; Mathieu et al., 1985). It has been suggested that if
        COHb is above 25%, HBO treatment should be initiated (Norkool and Kirkpatrick, 1985),
        although treatment plans based on specific COHb saturations is not  well founded (Thorn  and
        Keim, 1989).  Most hyperbaric centers treat patients with CO intoxication when they manifest
10      loss of consciousness or other neurological signs and symptoms (excluding headace)
        regardless of the COHb  saturation at presentation (Piantadosi, 1990).  The halftime
        elimination of CO while breathing air is approximately 320 min; when breathing 100%
        oxygen it is 80 min, and when breathing oxygen at 3 atmospheres it is 23 min (Penney et al.,
        1983; Myers etal., 1985).
15           Successful removal of CO from the blood does not ensure an uneventful recovery with
        no further clinical signs  or symptoms.  Neurological problems may  develop insidiously weeks
        after recovery from the acute episode of CO poisoning (Meredith and Vale, 1988). These
        problems include  intellectual deterioration;  memory impairment; and cerebral, cerebellar, and
        mid-brain damage. Up to two-fifths of patients develop memory impairment and a third
20      suffer late deterioration of personality. Arrhythmias are a common complication of CO
        poisoning.  Conduction defects also are found, possibly from cardiomyopathies, but the
        precise mechanisms by which these occur are not understood. Other systemic complications,
        such as skeletal muscle necrosis, renal failure, blood dyscrasias, pulmonary edema, and
        hemorrhage in various tissues also can occur as a result of CO poisoning.
25

      2.7  REFERENCES

30    Adams, K. F.; Koch, G.; Chatterjee, B.; Goldstein, G. M.; O'Neil, J. J.; Bromberg, P. A.; Sheps, D. S.;
             McAllister, S.; Price,  C. J.; Bissette, J. (1988) Acute elevation of blood carboxyhemoglobin to 6%
             impairs exercise performance and aggravates symptoms in patients with ischemic heart disease.  J. Am.
             Coll. Cardiol.  12: 900-909.

        March 12, 1990                          2-23     DRAFT - DO  NOT QUOTE OR CITE

-------
       Akland, G. G.; Hartwell, T. D.; Johnson, T. R.; Whitmore, R. W. (1985) Measuring human exposure to carbon
              monoxide in Washington, D.C., and Denver, Colorado, during the winter of 1982-1983. Environ. Sci.
              Technol. 19: 911-918.

 5     Allred, E. N.; Bleecker, E. R.; Chaitman, B. R.; Dahms, T. E.; Gottlieb, S. O.; Hackney, J. D.; Hayes, D.;
              Pagano, M. Selvester, R. H.; Walden, S. M.; Warren, J. (1989) Short-term effects of carbon monoxide
              exposure on the exercise performance of subjects with coronary artery disease. N. Engl. J. Med.
              321: 1426-1432.

10     American Heart Association. (1988)  1988 heart facts. Dallas, TX: American Heart Association.

       Anderson, E. W.; Andelman, R. J.;  Strauch, J. M.; Fortuin, N. J.; Knelson, J. H. (1973) Effect of low-level
              carbon monoxide exposure on onset and duration of angina pectoris: a study in ten patients with ischemic.
              heart disease. Ann. Intern. Med. 79: 46-50.
15
       Barret, L.;  Danel, V.; Faure, J. (1985) Carbon monoxide poisoning, a diagnosis frequently overlooked. J.
              Toxicol. Clin. Toxicol. 23: 309-313.

       Beard, R. R.; Wertheim, G. A. (1967) Behavioral impairment associated with small doses of carbon monoxide.
20            Am. J. Public Health 57: 2012-2022.

       Bender, E.; Cothert, M.; Malomy, G.; Sebbesse, P. (1971) The effects of low concentrations of carbon
              monoxide in man. Arch. Toxicol. 27: 142-158.

25     Benignus, V. A.; Otto, D. A.; Prah, J. D.; Benignus, G. (1977) Lack of effects of carbon monoxide on human
              vigilance. Percept. Mot. Skills 45: 1007-1014.

       Bernard, C. (1865) An introduction to the study of experimental medicine. 1957 reprint. Greene,  H. C., trans.
              New York, NY: Dover.
30
       Broome, J. R.; Skrine, H.; Pearson, R. R. (1988) Carbon monoxide poisoning: forgotten not gone! Br. J. Hosp.
              Med. 39: 298, 300, 302, 304-305.

       Brown, S. D.; Piantadosi,  C. A.; Gorman, D. F.; Gilligan, J. E.  F.; Clayton, D.  G.; Neubauer, R. A.;
35            Gottlieb, S. F.; Raphael, J.-C.; Elkharrat, D.; Jars-Guincestre, M.-C.; Chastang, C.; Chasles, V.;
              Vercken, J.-B.; Gajdos, P. (1989) Hyperbaric oxygen for carbon monoxide poisoning [letters]. Lancet
              (8670): 1032-1033.

       Christensen, C. L.; Gliner, J. A.; Horvath, S. M.; Wagner, J. A.  (1977) Effects of three kinds of hypoxias on
40            vigilance performance.  Aviat. Space Environ. Med.  48: 491-496.

       Coburn,  R.  F.; Blakemore, W. S.; Forster, R. E. (1963) Endogenous carbon monoxide production in man. J.
              Clin. Invest. 42: 1172-1178.

45     Coburn,  R. F.; Forster, R. E.; Kane, P. B.  (1965) Considerations of the physiological variables that determine
              the blood carboxyhemoglobin concentration in man. J. Clin. Invest. 44: 1899-1910.

       Dolan, M. C.; Haltom, T. L.; Barrows, G.  H.; Short, C. S.; Ferriell, K. M. (1987) Carboxyhemoglobin levels
              in patients with flu-like symptoms. Ann.  Emerg. Med. 16: 782-786.
50
       Douglas, C. G.; Haldane, J. S.; Haldane, J. B. S. (1912) The laws of combination of haemoglobin with carbon
               monoxide and oxygen. J. Physiol. (London) 44:  275-304.
         March 12, 1990                               2-24      DRAFT - DO NOT QUOTE OR CITE

-------
        Drinkwater, B. L.; Raven, P. B.; Horvath, S. M.; Gliner, J. A.; Ruhling, R. O.; Bolduan, N. W.; Taguchi, S.
               (1974) Air pollution, exercise, and heat stress. Arch. Environ. Health 28: 177-181.

        Ekblom, B.; Huot, R. (1972) Response to submaximal and maximal exercise at different levels of
 5            carboxyhemoglobin. Acta Physiol. Scand. 86: 474-482.

        Federal Register. (1980) Carbon monoxide; proposed revisions to the national ambient air quality standards:
               proposed rule. F. R. (August 18) 45: 55066-55084.

10     Federal Register. (1985) Review of the national ambient air quality standards for carbon monoxide; final rule.
               F. R. ( September 13) 50: 37484-37501.

        Fisher, J.; Rubin, K. P. (1982) Occult carbon monoxide poisoning [editorial]. Arch. Int. Med. 142: 1270-1271.

15     Grace, T. W.; Platt, F. W. (1981) Subacute carbon monoxide poisoning: another great imitator. JAMA J. Am.
               Med.  Assoc. 246: 1698-1700.

        Grut,  A. (1949) Chronic carbon monoxide poisoning: a study in occupational medicine. Copenhagen,  Denmark:
               Ejnar Munksgaard.
20
        Haider, M.; Groll-Knapp, E.; Hoeller, H.; Neuberger, M.; Stidl, H. (1976) Effects of moderate carbon
               monoxide dose on the central nervous system—electrophysiological and behaviour data and clinical
               relevance. In: Finkel, A. J.; Duel, W. C., eds. Clinical implications of air pollution research: air
               pollution medical research conference; December 1974; San Francisco, CA.  Acton, MA: Publishing
25            Sciences Group, Inc.; pp. 217-232.

        Heckerling, P. S.; Leikin, J. B.; Maturen, A.;  Perkins, J. T. (1987) Predictors of occult carbon monoxide
               poisoning in patients with headache and dizziness. Ann. Intern. Med.  107: 174-176.

30     Heckerling, P. S.; Leikin, J. B.; Maturen, A. (1988) Occult carbon monoxide poisoning: validation of a
               prediction model. Am. J. Med.  84: 251-256.

        Heimbach, D. M.; Waeckerle, J. F. (1988) Inhalation Injuries. Presented at: ACEP winter symposium clinical
               advances track; March;  San Diego, CA. Ann. Emerg. Med. 17: 1316-1320.
35
        Horvath, S. M.; Raven, P. B.;  Dahms, T. E.; Gray, D. J. (1975) Maximal aerobic capacity at different levels of
               carboxyhemoglobin. J. Appl. Physiol. 38: 300-303.

        James, P. B. (1989) Hyperbaric and normobaric oxygen in acute carbon monoxide poisoning [letter]. Lancet
40            (8666): 799-780.

        Kirkpatrick, J. N. (1987) Occult carbon monoxide poisoning. West. J. Med. 146: 52-56.

        Klein, J. P.;  Forster,  H. V.; Stewart, R. D.; Wu, A. (1980) Hemoglobin affinity for oxygen during short-term
45            exhaustive exercise. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 48: 236-242.

        Kleinman, M. T.; Davidson, D. M.; Vandagriff, R.  B.; Caiozzo, V. J.; Whittenberger, J. L.  (1989) Effects of
               short-term exposure to carbon monoxide in subjects with coronary artery disease. Arch. Environ. Health
               44: 361-369.
50
        Lambert, W. E.;  Colome, S. D. (1988) Distribution  of carbon monoxide exposures within indoor residential
               locations. Presented at:  81st annual meeting of the Air Pollution Control Association; June; Dallas, TX.
               Pittsburgh, PA: Air Pollution Control Association; paper no. 88-90.1.
         March 12, 1990                               2-25      DRAFT - DO NOT QUOTE OR CITE

-------
       Mathieu, D.; Nolf, M.; Durocher, A.; Saulnier, F.; Frimat, P.; Furon, D.; Wattel, F. (1985) Acute carbon
               monoxide poisoning risk of late sequelae and treatment by hyperbaric oxygen. J.  Toxicol. Clin. Toxicol.
               24: 315-324.

 5     McFarland, R. A. (1973) Low level exposure to carbon monoxide and driving performance. Arch. Environ.
               Health 27: 355-359.

       McFarland, R. A.; Roughton, F. J. W.; Halperin, M. H.; Niven, J. I. (1944) The effects of carbon monoxide
               and altitude on visual thresholds. J. Aviat. Med. 15: 381-394.
10
       Meredith, T.; Vale,  A. (1988) Carbon monoxide poisoning. Br. Med. J.  296: 77-79.

       Myers,  R. A. M., chairman. (1986) Hyperbaric oxygen therapy: a committee report. Bethesda, MD:  Undersea
               and Hyperbaric Medical Society; pp. 33-36.
15
       Myers,  R. A. M.; Snyder, S. K.; Emhoff, T. A.  (1985) Subacute sequelae of carbon monoxide poisoning. Ann.
               Emerg. Med. 14: 1163-1167.

       National Air Pollution Control Administration. (1970) Air quality criteria for carbon monoxide. Washington,
20             DC: U. S. Department of Health, Education, and Welfare, Public Health Service; report no.
               NAPCA-PUB-AP-62. Available from: NTIS, Springfield, VA; PB-190261.

       National Safety Council. (1982) Accident facts. National Safety Council;  pp. 80-84.

25     Norkool, D. M.; Kirkpatrick, J. N. (1985) Treatment of acute carbon monoxide poisoning with hyperbaric
               oxygen: a review of 115 cases. Ann. Emerg. Med. 14: 1168-1171.

       O'Donnell, R. D.; Mikulka, P.; Heinig, P.; Theodore, J. (1971) Low level carbon monoxide exposure and
               human psychomotor performance. Toxicol. Appl. Pharmacol. 18:  593-602.
30
       Pace, N.;  Strajman,  E.; Walker, E. L. (1950) Acceleration of carbon monoxide elimination in man by high
               pressure oxygen. Science (Washington, DC)  111: 652-654.

       Penney, D. G.; Zak, R.;  Aschenbrenner, V. (1983) Carbon monoxide  inhalation: effect on heart cytochrome c in
35             the neonatal and adult rat. J. Toxicol. Environ. Health 12: 395-406.

       Piantadosi, C. A. (1990) Carbon monoxide intoxication.  In:  Update in intensive care and emergency medicine,
               Vol. 10 (in press).

40     Pirnay,  J.; DuJardin, J.; DeRoanne, R.; Petit, J.  M. (1971) Muscular exercise during intoxication by carbon
               monoxide. J. Appl. Physiol. 31: 573.

       Putz, V. R. (1979) The effects of carbon monoxide on dual-task performance. Human Factors 21: 13-24.

45     Putz, V. R.; Johnson, B.  L.; Setzer, J. V. (1976) Effects of CO on vigilance performance: effects of low level
               carbon monoxide  on divided attention, pitch discrimination, and the auditory evoked potential.
               Cincinnati, OH: U. S. Department of Health, Education, and Welfare, National Institute for Occupational
               Safety and Health. Available from: NTIS, Springfield, VA; PB-274219.

50     Putz, V. R.; Johnson, B.  L.; Setzer, J. V. (1979) A comparative study of the effects of carbon monoxide and
               methylene chloride on human  performance. J. Environ. Pathol. Toxicol. 2: 97-112.
         March  12,  1990                              2-26      DRAFT - DO NOT QUOTE OR CITE

-------
       Raphael, J. C.; Elkharrat, D.; Jars-Guincestre, M.-C.; Chastang, C.; Chasles, V.; Vercken, J. B.; Gajdos, P.
               (1989) Trial of nonnobaric and hyperbaric oxygen for acute carbon monoxide intoxication. Lancet
               (8660): 414-419.

 5     Rockwell, T. J.; Weir, F. W. (1975) The interactive effects of carbon monoxide and alcohol on driving skills.
               Columbus, OH: The Ohio State University Research Foundation; CRC-APRAC project CAPM-9-69.
               Available from: NTIS, Springfield, VA; PB-242266.

       Roy, T. M.; Mendieta, J. M.; Ossorio, M. A.; Walker, J. F. (1989) Perceptions and utilization of hyperbaric
10             oxygen therapy for carbon monoxide poison-ing  in an academic setting. J. Ky. Med. Assoc. 87: 223-
               226.

       Rummo, N.;  Sarlanis, K. (1974) The effect of carbon monoxide on several measures of vigilance in a simulated
               driving task. J. Saf. Res. 6: 126-130.
15
       Salvatore, S.  (1974) Performance decrement caused by mild carbon monoxide levels on two visual functions. J.
               Saf. Res.  6: 131-134.

       Schulte, J.  H. (1973) Effects of mild carbon  monoxide intoxication. Arch. Environ. Health 7: 524-530.
20
       Sheps, D. S.; Adams, K. F., Jr.; Bromberg,  P. A.; Goldstein, G. M.; O'Neil, J. J.; Horstman, D.; Koch, G.
               (1987) Lack of effect of low levels of carboxyhemoglobin on cardiovascular function in patients with
               ischemic heart disease. Arch. Environ. Health 42:  108-116.

25     Statutes-at-Large. (1986) Superfund Amendments and Reauthorization Act of 1986, PL 99-499, October 17,
               1986, title IV, §403: radon gas and indoor air quality research program. Stat. 100: 1758.

       Stewart, R. D.; Newton, P. E.; Kaufman, J.; Forster, H. V.; Klein, J. P.; Keelen, M. H., Jr.; Stewart, D. J.;
               Wu, A.; Hake, C. L. (1978) The effect of a rapid 4% carboxyhemoglobin saturation increase on maximal
30             treadmill exercise. New York, NY: Coordinating Research Council, Inc.; report no.
               CRC-APRAC-CAPM-22-75. Available from: NTIS, Springfield,  VA; PB-296627.

       Thorn, S. R.; Keim, L. W.  (1989) Carbon monoxide poisoning: a review epidemiology, pathophysiology, clinical
               findings, and treatment options including hyperbaric oxygen therapy. J. Toxicol. Clin. Toxicol.
35             27: 141-156.

       U.  S. Centers for Disease Control. (1982) Carbon monoxide intoxication—a preventable environmental health
               hazard. Morb. Mortal. Wkly. Rep. 31: 529-531.

40     U.  S. Code. (1982) Clean Air Act: air quality criteria and control techniques; national primary and secondary
               ambient air quality standards. U. S. C. 42: §7408-7409.

       U.  S. Environmental Protection Agency. (1979)  Air quality criteria for carbon monoxide.  Research Triangle
               Park, NC: Office of Health and Environmental Assessment, Environmental Criteria and Assessment
45             Office;  EPA report no. EPA-600/8-79-022. Available from: NTIS, Springfield, VA; PB81-244840.

       U.  S. Environmental Protection Agency. (1984a) Revised evaluation of health effects associated with carbon
               monoxide exposure: an addendum to  the 1979 EPA air quality criteria document for carbon monoxide.
               Research Triangle Park, NC: Office of Health and Environmental Assessment, Environmental Criteria
50             and Assessment Office; EPA report no. EPA-600/9-83-033F. Available from: NTIS, Springfield, VA;
               PB85-103471.
         March 12,  1990                               2-27     DRAFT - DO NOT QUOTE OR CITE

-------
       U. S. Environmental Protection Agency. (1984b) Review of the NAAQS for carbon monoxide: reassessment of
              scientific and technical information. Research Triangle Park, NC: Office of Air Quality Planning and
              Standards; EPA report no. EPA-450/5-84-004. Available from: NTIS, Springfield, VA; PB84-231315.

 5     U. S. Environmental Protection Agency. (1988) National air quality and emissions trends report, 1986. Research
              Triangle Park,  NC: Office of Air Quality Planning and Standards; EPA report no. EPA/450/4-88/001.
              Available from NTIS, Springfield, VA; PB88-172473/XAB.

       U. S. Environmental Protection Agency. (1989a) National air quality and emissions  trends report, 1987. Research
10            Triangle Park,  NC: Office of Air Quality Planning and Standards; report no. EPA/450/4-89/001.

       U. S. Environmental Protection Agency. (1989b) Ozone and carbon monoxide 1986-1988 air quality data.
              Environmental News July 27.

15     Van Hoesen, K. B.; Camporesi, E. M.; Moon, R. E.; Hage,  M. L.; Piantadosi, C.  A. (1989) Should hyperbaric
              oxygen be used to treat the pregnant patient for acute carbon monoxide poisoning? A case report and
              literature review. JAMA J. Am. Med.  Assoc. 261: 1039-1043.

       Vogel, J. A.; Gleser, M. A. (1972) Effect of carbon monoxide on oxygen transport during exercise. J. Appl.
20            Physiol. 32: 234-239.

       Weiser, P. C.; Morrill, C. G.; Dickey, D. W.; Kurt, T. L.; Cropp, G. J. A. (1978) Effects of low-level carbon
              monoxide exposure on the adaptation of healthy young men to aerobic work at an altitude of 1,610
              meters.  In:  Folinsbee, L. J.; Wagner, J. A.; Borgia, J. F.;  Drinkwater, B. L.; Gliner, J. A.; Bedi,  J. F.,
25            eds. Environmental stress: individual human adaptations. New York, NY: Academic Press,  Inc.;
              pp. 101-110.

       Winneke, G. (1974) Behavioral effects of methylene chloride and carbon monoxide as  assessed by sensory and
              psychomotor performance. In: Xintaras, C.; Johnson, B. L.; de Groot, I., eds. Behavioral toxicology:
30           early detection of occupational hazards [proceedings of a workshop; June 1973; Cincinnati,  OH].
              Cincinnati, OH: U. S. Department of Health, Education, and Welfare, National Institute for Occupational
              Safety and Health; pp. 130-144; DHEW publication no. (NIOSH) 74-126. Available from: NTIS,
              Springfield, VA; PB-259322.

35    Wright, G.; Randell, P.; Shephard, R. J. (1973) Carbon monoxide and driving skills.  Arch. Environ. Health
              27: 349-354.
         March  12,  1990                              2-28      DRAFT - DO NOT QUOTE OR CITE

-------
           3. PROPERTIES AND  PRINCIPLES OF FORMATION
                              OF  CARBON MONOXIDE

         3.1  INTRODUCTION
  5             Carbon monoxide was first discovered to be a minor constituent of the earth's
         atmosphere by Migeotte (1949) in 1948. While taking measurements of the solar spectrum,
         he observed a strong absorption band in the infrared region at 4.7 ^m, which he attributed to
         CO (Lagemann et al., 1947).  On the twin bases of the belief that the solar contribution to
         that band was negligible and his observation of a strong day-to-day variability in absorption,
 10       Migeotte concluded that an appreciable  amount of CO was present in the terrestrial
         atmosphere of Columbus, Ohio.  In the 1950s many more observations (Benesch et al., 1953;
         Faith et al., 1959; Locke and Herzberg, 1953; Migeotte and Neven, 1952; Robbins et al.,
         1968; Sie et al., 1976) of CO were made, with measured concentrations ranging from 0.08 to
         100 ppm. On the basis of these and other measurements available in 1963, Junge (1963)
 15       stated that CO appeared  to be the most abundant trace gas, other than carbon dioxide, in the
         atmosphere.  The studies of Sie et al. (1976) indicated higher mixing ratios near the ground
         than in the upper atmosphere, implying  a source in the biosphere, but Junge emphasized that
         knowledge of the sources and sinks of atmospheric CO was extremely poor. It  was not until
         the late 1960s that concerted efforts were made to determine the various production and
20       destruction  mechanisms for CO in the atmosphere.
               Even far from human habitation, carbon monoxide occurs in air at an average
         background concentration of 0.05 mg/m3, primarily as a result of natural processes such as
         forest fires and the oxidation of methane.  Much higher concentrations occur in  cities  from
         technological sources such  as automobiles and the production of heat and power. Carbon
25       monoxide emissions are increased when the fuel is burned in an incomplete or inefficient
         way.  The physical and chemical properties of CO suggest that its atmospheric removal occurs
         primarily by reaction of CO with hydroxyl (OH*) radicals.
              The  remainder of this chapter focuses on the physical properties and formation
         principles of CO that contribute to its release into the atmosphere. In Chapter 4, other source
30       and sink estimates as well as the global cycle of CO are described; in Chapter 6, the various
         factors that  determine the technological emission source strengths are discussed.
         February 8, 1990                        3-1      DRAFT-DO NOT QUOTE OR CITE

-------
        3.2  PHYSICAL PROPERTIES
               Carbon monoxide is a tasteless, odorless, colorless diatomic molecule that exists as a
        gas in the earth's atmosphere.  Radiation in the visible and near ultraviolet regions of the
        electromagnetic spectrum is not absorbed by CO, although the molecule does have weak
 5      absorption bands between 125 and 155 nm.  It absorbs radiation in the infrared region
        corresponding to the vibrational excitation of its electronic ground state.  Carbon monoxide
        has a low electric dipole moment (0.10 debye), short interatomic distance (1.23A), and high
        heat of formation from atoms, or bond strength (2072 kJ/mol). These observations suggest
        that the molecule is a resonance hybrid of three structures (Perry et al., 1977), all of which
10      contribute nearly equally to the normal ground state.  General physical properties of CO are
        given in Table 3-1.
        3.3 GASEOUS CHEMICAL REACTIONS OF CARBON MONOXIDE
15             In the atmosphere, carbon monoxide reacts with OH* radicals to produce carbon
        dioxide and (CO2) hydrogen (H*) atoms.

                                       OH* + CO -> CO2 + H*                          (3-1)

20      The H' atoms formed in this process react very rapidly with oxygen (OJ to produce
        hydroperoxyl radicals (HOj ).
30
                                    H* + O2 (+M) -> HO; (+M)                        (3-2)

25       The liberated HO; radicals can react with nitric oxide (NO) to form nitrogen dioxide (NO2)
         and regenerate OH* radicals.
                                        ; + NO -> NO2 + OH*                         (3-3)
         February 8, 1990                        3-2      DRAFT-DO NOT QUOTE OR CITE

-------
           TABLE 3-1. PHYSICAL PROPERTIES OF CARBON MONOXIDE"
       Molecular weight

       Critical point

       Melting point

       Boiling point

       Density
          at 0°C,  1 atm
          at 25°C, 1 atm

       Specific gravity relative to air

       Solubility in water1"
          atO°C
          at 20°C
          at 25 °C

       Explosive limits in air

       Fundamental vibration transition
       Conversion factors
          at 0°C, 1 atm
          at25°C, 1 atm
     28.01
     -140°C at 34.5 atm
     -199°C
     -191.5°C
     1.250g/L
     1.145 g/L

     0.967
     3.54 mL/100 mL (44.3 ppmm)c
     2.32 mL/100 mL (29.0 ppmm)
     2.14 mL/100 mL (26.8 ppmm)

     12.5-74.2%

     2,143.3cm-1

     CO(X'Zg+, v1 =  1 Ev"0)(4.67
     1 mg/m3 = 0.800 ppmd
     1 ppm = 1.250 mg/m3

     1 mg/m3 = 0.873 ppm
     1 ppm = 1.145 mg/m3
" National Research Council (1977).
 Volume of carbon monoxide is at 0°C, 1 atm (atmospheric pressure at sea
 level = 760 ton).
c Parts per million by mass (ppmm = ftg/g).
d Parts per million by volume (ppm = mg/L).
 See glossary of terms and symbols for abbreviations and acronyms.
February 8, 1990
3-3     DRAFT - DO NOT QUOTE OR CITE

-------
         The photolysis of NO2 leads to the formation of ozone (O3); hence, CO can contribute to the
         production of photochemical smog in the lower troposphere. Other radicals besides OH* also
         can react with CO.

 5                                       CO + HO; •*  CO2 + OH*                          (3-4)
                                         CO + NO; ->  CO2 + NO2                          (3-5)
                                         CO + CH3O; •* product                            (3-6)

         The rates of these reactions, however, are so slow that their contribution to the overall
10       chemistry occurring in the atmosphere is expected to be very slight. Hampson and Garvin
         (1978), in their review of chemical kinetics data,  recommended a rate constant of < 10"" cm3
         molecule"1 s"1 for reaction  (3-4).  DeMore et al. (1987), based on their analysis of rate data,
         suggested a rate constant of <4.0 x 10~19 cm3 molecule"1 s"1 for reaction (3-5) and Heicklen
         (1973) recommended a value of 4 x 10~17 for reaction (3-6). In contrast, the rate constant for
15       the CO + OH*  reaction is  of the order of 10"13 cm3 molecule"1 s"1, a factor of at least 104 to
         10* greater than other known reactions between CO and atmospheric constituents. Thus, the
         reaction with OH* is the only reaction involving CO that is expected to be of any consequence
         in the atmosphere.
                The reaction of CO with OH* is one of the most studied of all atmospheric reactions.
20       Table 3-2 summarizes the  results  obtained in a few of these studies. More complete reviews
         of the kinetics of this reaction can be found in Hampson and Garvin (1978), Baulch et al.
         (1980), and DeMore et al. (1987).  As seen in Table 3-2, the rate constants  obtained in the
         late 1960s and early 1970s agreed fairly well and led the National Bureau of Standards
         (Hampson and Garvin,  1975) to recommend a value of 1.4  x 10~13 cm3 molecule"1 s"1 for the
25       rate constant. At that time there did not appear to be either a substantial temperature or
         pressure dependency for this reaction.  In the mid 1970s, however, Cox et al.  (1976)  studied
         the reaction at 700 torr using a mixture of N2 and O2 as the diluent gas. They obtained a rate
         constant of 2.7  x 10"13 cm3 molecule"1 s"1 and  suggested that the reaction might be subject to
         a pressure effect. At approximately the same time, Sie et al. (1976) studied the CO + OH"
30       reaction as a function of pressure.  When molecular hydrogen was used as the diluent gas,
         they found that the rate constant increased from 0.9 x  10"13 cm3 molecule"1 s~' at a pressure

         February 8,  1990                         3-4      DRAFT-DO NOT QUOTE OR CITE

-------
      TABLE 3-2. REPORTED ROOM TEMPERATURE RATE CONSTANTS FOR THE
                      REACTION OF OH* RADICALS WITH CO
Reference
No observed pressure dependence
Greiner (1969)
Stuhl and Niki (1972)
Westenberg and deHaas (1973)
Smith and Zellner (1973)
Howard and Evenson (1974)
Davis et al. (1974)
Gordon and Mulac (1975)
Atkinson et al. (1976)
Pressure dependence observed*
Cox et al. (1976)
Sie et al. (1976)
Perry et al. (1977)
Chan et al. (1977)
Biermann et al. (1978)
Paraskevopoulos and Irwin (1984)
DeMore (1984)
Hofzumahaus and Stuhl (1984)
Hynes et al. (1986)
Niki et al. (1984)
Hynes et al. (1986)
Pressure
(torr)

100
20
1-3
10-20
0.3-6
20
730
25-650

700
20-700
25-600
100-700
25-750
20-700
200-730
20-700
50-700
700
50-700
Diluent

He
He
He or Ar
He or N2O + H2
He, Ar, or N2
He or N2
Ar
Ar

N2 + O2
H2
SF6
Air
N2
N2
N2
N2
N2
Air
Air
Rate constant x 10 "
(cm3 molecule'1 s ')

1.4 + 0.2
1.3 ± 0.2
1.3
1.4
1.6 ± 0.2
1.6
1.5 + 0.1
1.5 ± 0.2

2.7 ± 0.2
3.3 + 0.2
3.4 ± 0.3
3.0 + 0.2
2.8 + 0.3
2.2 + 0.1
2.1 ± 0.4
2.3 + 0.2
2.1 ±0.2
2.4 ± 0.1
2.3 + 0.2
* Rate constants listed are the values obtained at the highest pressure used in each study.

See glossary of terms and symbols for abbreviations and acronyms.
    February 8, 1990
3-5     DRAFT - DO NOT QUOTE OR CITE

-------
         of 20 torr to 3.3 X 10~13 cm3 molecule"1 s"1 at 700 torr. When argon was used as the diluent
         gas, however, the rate of the reaction was found to be insensitive to pressure changes.
         Subsequent research has supported this finding.  In general, it appears that there is no
         pressure effect if noble gases (for example, helium or argon) are used as the carrier gas, but
 5       when other gases are used, such as N2 or O2, which are more representative of the
         atmosphere, the CO + OH* reaction rate exhibits a strong pressure dependency. Table 3-2
         summarizes the more  recent pressure dependency studies. In all cases, the rate constant listed
         for the pressure dependent studies is the value obtained at the highest pressure used in each
         study. Excellent agreement is noted for the studies conducted in 1984 and later years.
10              The National Aeronautics and Space Administration (NASA) Data Evaluation Panel
         (DeMore,  1987) recently examined the kinetics  data for the CO + OH* reaction.  They first
         analyzed all of the direct, low-pressure determinations to derive a zero pressure value for the
         rate constant.  They then performed a weighted  squares analysis of all the pressure-dependent
         data obtained since 1984 and fitted it to the expression
15
                                          K = K° x (1 + CPalm)                            (3-7)

         where
             K°  = zero pressure value for the rate constant,
20           c  = constant,
             P  = pressure in atmospheres.

         They found that the data were best fit using the  expression

25                               K  = (1.50 ± 0.45) X 10'13 (1 + 0.6Patm )                   (3-8)

         At a pressure of 760 torr, this corresponds to a  rate constant of 2.4 + 0.7 x  10~13 cm3
         molecule"1 s"1, independent of temperature. Thus,  the rate constant at atmospheric pressure is
         substantially larger than the value that previously had been assumed for this reaction.  The
30       larger value has led to important changes in our understanding of the global CO cycle.  This
         point is discussed in more detail in Chapter 4.

         February 8, 1990                           3-6       DRAFT-DO NOT QUOTE OR CITE

-------
         3.4  PRINCIPLES OF FORMATION BY SOURCE CATEGORY
              Carbon monoxide is produced at the earth's surface during the combustion of fuels and
         in the atmosphere during the oxidation of anthropogenic and biogenic hydrocarbons.  The role
         of man-made and natural hydrocarbons in CO production is discussed in Chapter 6 and only
 5       production of CO from combustion sources is addressed here.
              The burning of any carbonaceous fuel produces two primary products: CO2 and CO.
         The production of CO2 predominates when the air or oxygen supply is in excess of the
         stoichiometric needs for complete combustion. If burning occurs under fuel-rich conditions,
         with less air or oxygen than is needed, CO will be produced in abundance. Most of the CO
10       and CO2 formed in past years simply was emitted into the atmosphere.
              In recent years, concerted efforts have been made to reduce concentrations of potentially
         harmful materials in ambient air. Today, the CO in urban air originates almost entirely from
         local combustion processes.  The background concentration of CO contributes less than
         0.23 mg/m3 (0.20 ppm) to the ambient air concentration at a given urban location.  As a
15       result of natural processes such as forest fires,  oxidation of methane, and biological activity.
         the background level of CO is estimated to be about 0.05 mg/m3 (0.04 ppm) (Seiler and
         Junge, 1970). See Chapter 4 for a discussion of global background  concentrations.
              Considerable effort has been made to reduce emissions of CO and other pollutants to the
         atmosphere.  Since the automobile engine is recognized to be the major source of CO in most
20       urban areas, special attention is given to the control of automotive emissions. Generally the
         approach has been technological: reduction of CO emissions to the  atmosphere either by
         improving the efficiency of the combustion processes, thereby increasing the yield of CO2 and
         decreasing the yield of CO; or by applying secondary catalytic combustion reactors to the
         waste gas stream to convert CO to CO2.
25            The development and application of control technology to reduce emissions of CO from
         combustion processes generally have been successful and are continuing to receive deserved
         attention.  The reduction of CO emissions from 7.0 to 3.4 g/mi, scheduled for the  1981
         model year, was delayed two years, reflecting in part the apparent difficulty encountered by
         the automobile industry in developing and supplying the required control technology. The
30       CO emission limit for light-duty vehicles (LDVs) at low altitude has been 3.4 g/mi since
         1983; since 1984, this limit applied to LDVs at all altitudes.

         February 8, 1990                        3-7      DRAFT-DO NOT QUOTE OR CITE

-------
              Table 3-3 shows the automobile emissions control schedules that have resulted from the
         1970 Clean Air Act and subsequent amendments, notably the 1977 and 1981 CAA
         Amendments.
              The problems encountered in mass producing and marketing effective control technology
 5       for automobile engines are complex because a number of simultaneous requirements are
         involved (i.e., control of multiple air pollutants, fuel economy and efficiency, durability and
         quality control of components, and maintenance).  Emission factor program testing conducted
         by EPA during the 1980s indicates that the durability of emission control systems continues to
         present a problem for in-use vehicles intended to comply with the 1983 and later requirement
10       that CO emissions be limited to not more than 3.4 g/mi through the 50,000-mile, useful-life
         compliance period. (CO emissions are less than 3.4 g/mi for new, low-mileage automobiles.)
              The following subsections present a brief discussion of the general principles and
         mechanisms of CO formation and control of emissions associated with the many combustion
         processes. The processes commonly are classified in two broad types, mobile sources and
15       stationary sources, because this division generally does separate distinct types of major
         combustion devices.  Control techniques for CO emissions from mobile and stationary sources
         are detailed in two National Air Pollution Control Administration publications (National Air
         Pollution Control Administration, 1970a,b).

20       3.4.1  General Combustion Processes
              Incomplete combustion of carbon-containing compounds creates varying amounts of CO.
         The chemical and physical processes that occur during combustion are complex because they
         depend not only on the type of carbon compound reacting with oxygen but also on the
         conditions existing in the combustion chamber (Mellor, 1972; Pauling, 1960).  Despite the
25       complexity of the combustion process, certain general principles regarding the formation of
         CO from the combustion of hydrocarbon fuels are accepted widely.
              Gaseous or liquid hydrocarbon fuel reacts with molecular oxygen in a chain of reactions
         that result in CO. Carbon monoxide then reacts with OH* radicals to form CO2.  The second
         reaction is approximately ten times slower than the first. In coal combustion, too, the
30       reaction of carbon and oxygen to form CO is one of the primary reactions, and a large
         February 8,  1990                         3-8       DRAFT-DO NOT QUOTE OR CITE

-------
  TABLE 3-3.  SUMMARY OF UGHT-DUTY VEHICLE (LDV) EMISSIONS STANDARDS1
Year
Prior to
controls
1968-69

1970
1971
1972
1973-74

1975-76
1977*
1978-79
1980
1981
198210
198310
1984-8612
1987 &
later12
Teat
Procedure2
7-raode
7-mode
CVS-75
7-mode
50-100 CID
101-140 CID
> 140 OD
7-mode
7-mode
CVS-72
CVS-72

CVS-75
CVS-75
CVS-75
CVS-75
CVS-75
CVS-75
CVS-75
CVS-75
CVS-75
Hydro-
carbom
850 ppm
11 g/mi
8.8 g/mi
410 ppm
350 ppm
275 ppm
2.2 g/mi
2.2 g/mi
3.4 g/mi
3.4 g/mi

1.5 g/mi
1.5 g/mi
1.5 g/mi
0.41 g/mi
0.41 g/mi
0.41 g/mi
(0.57)
0.41 g/mi
(0.57)
0.41 g/mi
0.42 g/mi
(0.41)
Carbon Oxides of
Monoxide Nitrogen
Partial- Evaporative
Isles3 Hydrocarbons4
Gtsofine-fociedLDVs
3.4 % 1000 ppm
80 g/mi 4 g/mi
87.0 g/mi 3. 6 g/mi
2.3%
2.0%
1.5*
23 g/mi
23 g/mi
39 g/mi
39 g/mi
Ganfioe-fiiekd and dioei LDV.
15 g/mi
15 g/mi
15 g/mi
7.0 g/mi
3.4 g/mi7
3.4 g/mi7
(7.8)"
3.4 g/mi
(7.8)
3.4 g/mi
3.4 g/mi
(3.4)
-
-
-
-
-

3.1 g/mi
2.0 g/mi
2.0 g/mi
2.0 g/mi
1.0 g/mi8-9
1.0 g/mi8-9
(1.0)8-9
1.0 g/mi8
(l.O)8
1.0 g/mi8
1.0 g/mi
(1.0)
-
-
6.0 g/test3
2.0 g/test
2.0 g/test

2.0 g/test
2.0 g/test
6.0 g/test
6.0 g/test
2.0 g/test
0.60 g/mi 2.0 g/test
(-) (2.6)
0.60 g/mi 2.0 g/test
(0.60) (2-6)
0.60 g/mi 2.0 g/test
0.20 g/mi13 2.0 g/test
(0.20)13 (2-0)
February 8, 1990
3-9
DRAFT-DO NOT QUOTE OR CITE

-------
TABLE 3-3. (cont'd) SUMMARY OF LIGHT-DUTY VEHICLE (LDV) EMISSIONS STANDARDS


1  Standards do not apply to LDVs with engines less than SO CID from 1968 through 1974.

2 Different test procedures, which vary in stringency, have been used since the early years of emission control.  The appearance that the standards were relaxed
  from 1971 to 1972 is incorrect; the 1972 standards actually are more stringent because of the 1972 test procedure.

3 Applies only to diesel LDVs.

4 Evaporative emissions determined by carbon-trap method through 1977, SHED procedure beginning in 1978.  Applies only to gasoline-fueled LDVs.

5 Evaporative standard does not apply to off-road utility LDVs for 1971.

* LDVs sold in specified high-altitude counties are required to meet these standards at high altitude.

7 Carbon monoxide standard is waived to 7.0 g/mi for 1981-82 for certain LDVs.

8 Oxides of nitrogen standard is waived to 7.0 g/mi for 1981-82 for certain LDVs.

9 Oxides of nitrogen standard for 1981-82 is 2.0 g/mi for American Motors Corporation LDVs.

10 Standards in parentheses apply to LDVs sold in specified high-altitude counties.

11 LDVs eligible for a carbon monoxide waiver to 7.0 g/mi at low altitude are eligible for a waiver to 11 g/mi at high altitude.

12 The same numerical standards apply to LDVs sold in high-altitude areas.  Exemptions from compliance at high-altitude are provided for qualifying low-
  performance vehicles.

13 Emissions averaging  may be used to meet this standard, provided that emissions from LDVs produced for sale in California or in designated high-altitude areas
  may be averaged only within each of those areas.
 CID    - cubic inch displacement
 CVS-72 - constant volume sample cold start test
 CVS-75 - constant volume sample test including cold and hot starts
 g/mi    - grams per mile
 ppm    - parts per million
 7-mode - 137 second driving cycle test
 SHED  - sealed housing for evaporative determination
 Source: AP-42 Supplement (1989); (in press)
   February 8, 1990                                    3-10         DRAFT-DO  NOT QUOTE OR CITE

-------
        fraction of carbon atoms go through the monoxide form.  Again, the reaction of monoxide to
        dioxide is much slower.
             Four basic variables control the concentration of CO in the combustion of all
        hydrocarbon gases.  These  are (1) O2 concentration, (2) flame temperature, (3) gas residence
 5      time at high temperatures, and (4) combustion chamber turbulence.  Oxygen concentration
        affects the formation of both CO and CO2 because O2 is required in the initial reactions with
        the fuel molecule and in the formation of the OH* radical. As the availability of O2 increases,
        more complete conversion of monoxide to dioxide results. Flame and gas temperatures affect
        both the formation of monoxide and the conversion of monoxide to dioxide because both
10      reaction rates increase exponentially with increasing temperature.  Also, the OH* radical
        concentration in the combustion chamber is very temperature-dependent. The conversion of
        CO to CO2 also is enhanced by longer residence time, because this is a relatively slow
        reaction in comparison with CO formation.  Increased gas turbulence in the  combustion zones
        increases the actual reaction rates by increasing the mixing of the reactants and assisting the
15      relatively slower gaseous diffusion process, thereby resulting in more complete combustion.

        3.4.2  Combustion Engines
        Mobile Combustion Engines—Most mobile sources of CO are internal combustion engines of
        two types: (1) gasoline-fueled, spark-ignition, reciprocating engines (carbureted or fuel-
20      injected); and (2) diesel-fueled reciprocating engines.  The CO emitted from any given engine
        is the product of the following factors:  (1) the concentration of CO in the exhaust gases,
        (2) the flow rate of exhaust gases, and (3) the duration of operation.
        InternaLCQrnbustiQn Engines (Gasoline-Fueled. Spark-Ignition EnginesV-Exhaust
        concentrations of CO  increase with lower (richer) air-to-fuel (A/F) ratios, and decrease with
25      higher (leaner) A/F ratios,  but remain relatively constant with ratios above the stoichiometric
        ratio of about 15:1 (Hagen  and Holiday, 1964).  The behavior of gasoline automobile engines
        before and after the installation of pollutant control devices differs considerably.  Depending
        on the mode of driving, the average uncontrolled  engine operates at A/F ratios ranging from
        about 11:1 to a point slightly above the stoichiometric ratio.  During the idling mode, at low
30      speeds with light load (such as low-speed cruise), during the full-open throttle mode until
        speed picks up, and during  deceleration,  the A/F  ratio is low in uncontrolled cars and CO

        February 8, 1990                         3-11      DRAFT-DO NOT QUOTE OR CITE

-------
        emissions are high.  At higher speed cruise and during moderate acceleration, the reverse is
        true. Cars with exhaust controls generally remain much closer to stoichiometric A/F ratios in
        all modes, and thus the CO emissions are kept lower.  The relationship between CO
        concentrations in engine exhaust and A/F ratios is shown in Figure 3-1.  The exhaust flow
 5      rate increases with increasing engine power output.
             The decrease in available oxygen with increasing altitude has the effect of enriching the
        A/F mixture and increasing CO emissions from carbureted engines.  Fuel-injected gasoline
        engines, which predominate in the vehicle fleet today, have more closely controlled A/F
        ratios and are designed and certified to comply with applicable emission standards regardless
10      of elevation (U.S. Environmental Protection Agency, 1983).
             Correlations between total emissions of CO in grams per vehicle mile and average route
        speed show a decrease in emissions  with increasing average speed (Simonaitis and Heicklen,
        1972; Stuhl and Niki, 1972; U.S. Environmental Protection Agency, 1985).  During low
        speed conditions (below 32 km/h or 20 mi/h average route speed), the greater emissions per
15      unit of distance traveled are attributable to (1) an increased frequency of acceleration,
        deceleration, and idling encountered in heavy traffic; and (2) the consequent increase in the
        operating time per mile driven.
             The CO and the unburned hydrocarbon exhaust emissions from an uncontrolled engine
        result from incomplete combustion of the fuel-air mixture. Emission control on new vehicles
20      is being achieved by engine modifications, improvements in engine design, and changes in
        engine operating conditions.  Substantial reductions in CO and other pollutant emissions result
        from consideration of design and operating factors such as leaner, uniform mixing of fuel and
        air during carburetion, controlled heating of intake air, increased idle speed, retarded spark
        timing, improved cylinder head design, exhaust thermal reactors, oxidizing and reducing
25      catalysts, secondary  air systems, exhaust recycle systems, electronic fuel injection, A/F ratio
        feedback controls, and modified ignition systems (National Research Council, 1973).

        Internal Combustion Engines (Diesel Engines)—Diesel engines in use are primarily the  heavy-
        duty type that power trucks and buses. Diesel engines allow more complete combustion and
30      use less volatile fuels than do spark-ignition engines. The operating principles are
        February 8, 1990                         3-12      DRAFT-DO NOT QUOTE OR CITE

-------
                                                                   £   <
                                                                   ^   e
                                                                       j
                                                                       UJ

                                                                       Ik
                                                                       I
                                                                       a
                                                                       5
                  o>    eo
                               % »(oiu '3QIXONOW N09HVO
Figure 3-1. Effect of air-fuel ratio on exhaust gas carbon monoxide concentrations from three

gasoline-fueled test engines.


Source:  Hagen and Holiday,  1964.
February 8, 1990
3-13      DRAFT-DO NOT QUOTE OR CITE

-------
       significantly different from those of the gasoline engine.  In diesel combustion, CO
       concentrations in the exhaust are relatively low because high temperature and large excesses
       of oxygen are involved in normal operation.  The exhaust emissions from diesel engines have
       the same general composition as gasoline engine emissions, though the concentrations of
 5     different pollutants vary considerably.  For example, the diesel emits larger quantities of
       nitrogen oxides (NOJ and polycyclic organic particulates than gasoline engines; it emits less
       CO.

       Stationary Combustion Sources (Steam Boilers)--This section refers to fuel-burning
10     installations such as coal-, gas-, or oil-fired heating or power generating plants (external
       combustion boilers).
            In these combustion systems, the formation of CO is lowest at a ratio near or slightly
       above  the stoichiometric ratio of air to fuel.  At lower than stoichiometric A/F ratios, high
       CO concentrations reflect the relatively low O2 concentration and the possibility of poor
15     reactant mixing from low turbulence. These two factors can increase emissions even though
       flame temperatures and residence time are high. At higher than stoichiometric A/F ratios,
       increased CO emissions result from decreased flame temperatures and shorter residence time.
       These two factors remain predominant even when O2 concentrations and turbulence increase.
       Minimal CO emissions and maximum thermal efficiency, therefore, require combustor
20     designs that provide high turbulence, sufficient residence time, high temperatures, and near
       stoichiometric A/F ratios.  Combustor design dictates the actual approach to that minimum.
            The measurement of CO in effluent  gas is used as an indication of improper and
       inefficient operating practice for any given combustor,  or of inefficient combustion.

25     3.4.3  Other Sources
            There are numerous industrial activities that result in the emission of CO at one or more
       stages of the process (Walsh and Nussbaum,  1978).  Manufacturing pig iron can produce as
       much as 700 to 1,050 kg CO/metric ton of pig iron. Other methods of producing iron and
       steel can produce CO at a rate of 9 to  118.5  kg/metric ton.  However,  most of the CO
30     generated is normally recovered  and used as  fuel.  Conditions such as "slips", abrupt
       collapses of cavities in the coke-ore mixture, can cause instantaneous emissions of CO that

       February 8,  1990                          3-14      DRAFT-DO NOT QUOTE OR CITE

-------
        temporarily exceed the capacity of the control equipment. Slips have been reduced greatly
        with modern equipment.  Grey-iron foundries can produce 72.5 kg CO/metric ton of product,
        but an efficient afterburner can reduce the CO emission to 4.5 kg/metric ton.
             Charcoal production results in CO emissions of 160 kg/metric ton with or without the
 5      installation of chemical recovery equipment. Emissions from carbon black manufacture can
        range from 5 to 3,200 kg CO/metric ton depending on the efficiency and quality of the
        emission control systems.
             Some chemical processes such as phthalic anhydride production give off as little as 6 kg
        CO/metric ton with proper controls or as much as 200 kg CO/metric ton if no controls are
10      installed.  There are numerous other chemical processes that produce relatively small CO
        emissions per metric ton  of product, such as sulfate pulping for paper at 1 to 30 kg CO/metric
        ton; lime  manufacturing normally runs 1 to 4 kg CO/metric ton; and CO from adipic acid
        production is zero or slight with proper controls.  Other industrial chemical processes that
        cause CO emissions are the manufacture of terephthalic acid and the synthesis of methanol
15      and higher alcohols.  As  a rule, most industries find it economically desirable  to install
        suitable controls to reduce CO emissions.
             Even though some of these CO emission rates seem excessively high, they are, in fact,
        only a small part of the total pollutant load. Mention of these industries is made to emphasize
        the concern for localized pollution problems when accidents occur or proper controls are not
20      used.
             While the estimated CO emissions resulting from forest wildfires in the United  States for
        1971 was 4 X 106 metric tons, the estimated total industrial process CO emission of the U.S.
        for 1971 was 10.3 X  10* metric tons.

25
      3.5 REFERENCES
      Atkinson, R.; Perry, R. A.; Pitts, J. N., Jr. (1976) Kinetics of the reactions of OH radicals with CO and N2O.
            Chem. Phys. Lett. 44: 204-208.
30
      Baulch, D. L.; Cox, R. A.; Hampson, R. F., Jr.; Kerr, J. A.; Troe, J.; Watson, R. T., eds. (1980) Evaluated
            inetic and photochemical data for atmospheric chemistry. J. Phys. Chem. Ref. Data 9: 295-471.
       February 8, 1990                          3-15       DRAFT-DO NOT QUOTE OR CITE

-------
       Benesch, W.; Migeotte, M.; Neven, L. (1953) Investigations of atmospheric CO at the Jungfraujoch. J. Opt. Soc.
              Am. 43:  1119-1123.

       Biermann, H. W.; Zetsch, C.; Stuhl, F. (1978) On the pressure dependence of the reaction of HO with CO. Ber.
 5            Bunsenges. Phys. Chem. 82: 633-639.

       Chan, W. H.; Uselman, W. M.; Calvert, J. G.; Shaw, J. H. (1977) The pressure dependence of the rate constant
              for the reaction: OH + CO — > H + CO2. Chem. Phys. Lett. 45: 240-244.

10     Cox, R. A.; Derwent, R. G.; Holt, P. M. (1976) Relative rate constants for the reactions of OH radicals with
              H2, CH4, CO, NO, and HONO at atmospheric pressure and 296 K. J.  Chem. Soc.  Faraday Trans. 1
              72: 2031-2043.

       Davis, D. D.; Fischer, S.; Schiff, R. (1974) Flash photolysis-resonance fluorescence kinetics study: temperature
15            dependence of the reactions OH +  CO = CO2 + H and OH + CH4 = H2O + CH3. J. Chem. Phys.
              61: 2213-2219.

       DeMore, W. B. (1984) Rate constant for the OH +  CO  reaction: pressure dependence and the effect of oxygen.
              Int. J. Chem. Kinet. 16: 1187-1900.
20
       DeMore, W. B.; Molina, M. J.; Sander, S. P.; Golden,  D. M.; Hampson, R. F.;  Kurylo, M.  J.;  Howard, C. J.;
              Ravishankara, A. R. (1987) Chemical kinetics and photochemical data for use in stratospheric modeling:
              evaluation number 8. Washington, DC: National Aeronautics and Space Administration;; report no.
              NASA-CR-182919. Available from: NTIS, Springfield, VA; N88-24012.
25
       Faith, W. L.; Renzetti, N.  A.; Rogers, L. H. (1959) Fifth technical progress report. San Marino,  CA: Air
              Pollution Foundation.

       Gordon,  S.; Mulac, W. A.  (1975) Reaction of the OH(^ II) radical produced by the pulse radiolysis of water
30            vapor. In: Benson,  S. W.; Golden,  D. M.; Barker, J. R., eds. Chemical kinetics data for the upper and
              lower atmosphere: proceedings of the symposium; September 1974; Warrenton, VA. Int. J. Chem.  Kinet.
              Symp. (1): 289-299.

       Greiner,  N. R. (1969) Hydroxyl radical kinetics by kinetic spectroscopy: V. reactions with H2 and CO in the
35            range 300-500° K.  J. Chem. Phys.  51: 5049-5051.

       Hagen, D. F.; Holiday, G. W. (1964) The effects of engine operating and design variables on exhaust emissions.
              In: Vehicle emissions (selected SAE papers).  New York, NY:  Society  of Automotive Engineers;
              pp. 206-223.  (Technical progress series volume 6).
40
       Hampson, R. F., Jr.; Garvin, D., eds. (1975) Chemical  kinetic and photochemical data for modelling
              atmospheric chemistry. Washington, DC: National Bureau of Standards; NBS  technical note 866.

       Hampson, R. F., Jr.; Garvin, D., eds. (1978) Reaction rate and pyhotochemical data for atmospheric chemistry -
45             1977. Washington,  DC: National Bureau of Standards;  NBS special publication 513.

       Heicklen, J. (1973) Photochemical and rate data for methyl nitrite, methoxy and methylperoxy. In: Garvin, D.,
              ed. Chemical kinetics data survey V. Washington, DC: National Bureau of Standards; NBSIR 73-206.

50     Hofzumahaus, A.; Stuhl, F. (1984) Rate constant of the  reaction HO + CO in the presence of N2  and O2. Ber.
              Bunsenges. Phys. Chem. 88: 557-561.

       Howard, C. J.; Evenson, K. M. (1974) Laser magnetic resonance study of the gas phase reactions of OH with
              CO, NO, and NO2. J. Chem. Phys. 61: 1943-1952.


         February 8, 1990                             3-16       DRAFT-DO NOT QUOTE OR  CITE

-------
        Hynes, A. J.; Wine, P. H.; Ravishankara, A. R. (1986) Kinetics of the OH  + CO reaction under atmospheric
               conditions. JGR J. Geophys. Res. 91: 11,815-11,820.

        Junge, C. E. (1963) Air chemistry and radioactivity: v. 4. New York, NY: Academic Press. (Mieghem, J.;
  5            Hales, A. L., eds. International geophysics series).

        Lagemann, R. T.; Nielsen, A. H.; Dickey, F. P. (1947) The infra-red spectrum and molecular constants of
               C12O16 and C13O16. Phys. Rev. 72: 284-289.

 10     Locke, J. L.; Herzberg, L. (1953) The absorption due to carbon monoxide in the infrared solar spectrum. Can. J.
               Phys. 31: 504-516.

        Mellor, A. (1972) Current kinetic modeling techniques for continuous flow combustors. In: Cornelius, W.;
               Agnew, W. G., eds. Emissions from continuous combustion systems: proceedings of the symposium;
 15            September 1971; Warren, MI. New York, NY: Plenum Press; pp. 23-53.

        Migeotte, M. V. (1949) The fundamental band of carbon monoxide at 4.7 jw,  in the solar spectrum. Phys. Rev.
               75: 1108-1109.

20     Migeotte, M.; Neven, L. (1952) Recents progres dans 1'observation du spectre infra-rouge du soleil a la station
               scientifique du Jungfraujoch (Suisse) [Recent progress in observing the infrared solar spectrum at the
               scientific station at Jungfraujoch, Switzerland]. Mem. Soc. R. Sci. Liege 12: 165-178.

        National Air Pollution Control Administration. (1970a) Control techniques for carbon monoxide emissions from
25            stationary sources. Washington, DC: U. S.  Department of Health, Education, and Welfare, Public Health
               Service; National Air Pollution Control Administration publication no. AP-65. Available from:  NTIS,
               Springfield, VA; PB-190263.

        National Air Pollution Control Administration. (1970b) Control techniques for carbon monoxide, nitrogen oxide,
30            and hydrocarbon emissions from mobile sources. Washington, DC: U. S. Department of Health,
               Education, and Welfare; publication no. AP-66.

        National Research Council. (1973) Automotive spark ignition engine emission control systems to meet the
               requirements of the 1970 clean air amendments. Washington, DC: National Academy of Sciences.
35            Available from: NTIS, Springfield,  VA; PB-224862.

        National Research Council. (1977) Carbon monoxide. Washington, DC: National Academy of Sciences. (Medical
               and biologic effects of environmental pollutants).

40     Niki, H.; Maker,  P. D.; Savage, C. M.;  Breitenbach, L. P. (1984) Fourier transform infrared spectroscopic
               study of the kinetics for the HO radical reaction of 13C"O and 12C18O. J. Phys. Chem. 88: 2116-2119.

        Paraskevopoulos, G.; Invin, R. S. (1984) The pressure dependence of the rate constant of the reaction of OH
               radicals with CO. J. Chem. Phys. 80: 259-266.
45
        Pauling, L. (1960) The nature of the chemical bond and the structure of molecules and crystals:  an introduction
               to modem structural chemistry. 3rd ed. Ithaca, NY: Cornell University Press; pp.  194-195.

        Perry, R. A.; Atkinson, R.; Pitts, J. N.,  Jr. (1977) Kinetics of the reactions  of OH radicals with C2H2 and CO.
50            J. Chem. Phys. 67: 5577-5584.

        Robbins, R. C.; Borg, K. M.; Robinson, E. (1968) Carbon monoxide in the  atmosphere. J. Air Pollut.  Control
               Assoc. 18: 106-110.
         February 8, 1990                             3-17       DRAFT-DO NOT QUOTE OR CITE

-------
       Seiler, W.; Junge, C. (1970) Carbon monoxide in the atmosphere. JGR J. Geophys. Res. 75: 2217-2226.

       Sie, B. K. T.; Simonaitis, R.; Heicklen, J. (1976) The reaction of OH with CO. Int. J. Chem. Kinet. 8: 85-98.

 5     Simonaitis, R.; Heicklen, J. (1972) Kinetics and mechanism of the reaction of O(jP) with carbon monoxide. J.
              Chem. Phys. 56: 2004-2011.

       Smith, I. W. M.; Zellner, R. (1973) Rate measurements of reactions of OH by resonance absorption: part 2.
              reactions of OH with CO,  C2H4, and C2H2. J. Chem. Soc. Faraday Trans. 2 69:  1617-1627.
10
       Stuhl, F.; Niki,  H. (1972) Pulsed vacuum-uv photochemical study of reactions of OH with H2, D2, and CO using
              a resonance-fluorescent detection method. J. Chem. Phys. 57: 3671-3677.

       U. S. Code. (1970-1981) Clean Air Act as amended by PL 91-604, 95-95, 97-23. U. S.  C. 42: §7401-7626.
15
       U. S. Environmental Protection Agency. (1983) Controlling emissions from light-duty vehicles at higher
              elevations. Ann Arbor, MI: Office of Mobile Sources; EPA report no. EPA-460/3-83-001. Available
              from: NTIS, Springfield, VA; PB83-204883.

20     U. S. Environmental Protection Agency. (1985) Compilation of air pollutant emission factors: v.  1, stationary
              point and the area sources, v. 2, mobile sources. 4th ed. Research Triangle Park, NC: Office of Air
              Quality Planning and Standards; EPA report nos. AP-42-ED-4-VOL-1 and AP-42-ED-4-VOL-2.
              Available from: NTIS, Springfield, VA; PB86-124906 and PB87-205266.

25     U. S. Environmental Protection Agency. (1989) Supplement to compilation of air pollutant emission factors.
              Research Triangle Park, NC: Office of Air Quality Planning Standards; EPA report no. AP-42 (in press).

       Walsh, M. P.; Nussbaum, B.  D. (1978) Who's responsible for emissions after 50,000 miles? Automot. Eng.
              86: 32-35.
30
       Westenberg, A.  A.; deHaas, N. (1973) Rates of CO  + OH and H2 + OH over an extended temperature range. J.
              Chem. Phys. 58: 4061-4065.
         February 8, 1990                            3-18       DRAFT-DO NOT QUOTE OR CITE

-------
        4.  THE  GLOBAL CYCLE OF CARBON MONOXIDE:
                        TRENDS AND MASS BALANCE
 5     4.1 INTRODUCTION
           In the troposphere CO may control the removal, and therefore the concentrations, of
       OH" radicals (Crutzen, 1974;  Khalil and Rasmussen, 1985; Levine et al., 1985; Sze, 1977;
       Thompson and Cicerone,  1986).  The chemical reactions of CO also may produce substantial
       amounts of 03 in the troposphere (Conrad and Seiler, 1982; Fishman and Crutzen, 1978;
10     Fishman et al., 1980; Fishman and Seiler, 1983; Seiler and Fishman, 1981).  If the
       concentrations of CO increase, O3 may increase; concomitantly, OH* may be depleted thus
       affecting the global cycles of  many natural and anthropogenic trace gases that are removed
       from the atmosphere by reacting  with OH*.  Therefore increasing CO may indirectly affect
       the global climate and contribute to widespread changes in atmospheric chemistry.
15         The purpose of this chapter is to review the present understanding of the sources and
       sinks of CO and the resulting global distributions and trends. The first section is a review of
       the global sources and sinks and the estimated atmospheric lifetime of CO.  The next section
       deals with the global distribution of CO resulting from the sources and sinks,  including
       variations with seasons, altitude,  and latitude.  Next, there is an analysis of the current
20     evidence for global trends that reflect an imbalance of the sources and sinks probably caused
       by steadily increasing emissions from anthropogenic sources. The chapter is concluded with a
       summary.
25     4.2  GLOBAL SOURCES, SINKS, AND LIFETIME
            The mass balance of a trace gas in the atmosphere can be described as a balance
       between the rate of change of the global burden added to the annual rate of loss on the one
       side and balanced by the global emissions on the other side (d concentration/dt + loss rate =
       source emissions). In steady state the atmospheric lifetime (T) is the ratio of the global
30     burden to the loss rate. The global burden is the total number of molecules of a trace gas in
       the atmosphere or its total mass.  The concentration of a trace gas can vary (dC/dt is not 0)
       March 5, 1990                          4-1      DRAFT-DO NOT QUOTE OR CITE

-------
       when either the loss rate or the emissions vary cyclically in time representing seasonal
       variations or vary over a long time often representing trends of human industrial activities and
       increasing population. For CO both types of trends exist.  There are large seasonal cycles
       mostly driven by seasonal variations in the loss rate but also affected by seasonal variations of
 5     emissions, and there are indications of long-term trends probably caused by increasing
       anthropogenic emissions.
            It appears that the largest sources of CO in the global atmosphere are combustion
       processes and the oxidation of hydrocarbons. CO is produced in the atmosphere by reactions
       of OH* with methane (CH4) and other hydrocarbons, both man-made and natural, and also
10     from the reactions of alkenes with O3 and of isoprene and terpenes with both OH* and O3.
       Most of the CO is removed from the atmosphere by reacting with tropospheric OH* radicals.

       4.2.1 Sources
            Carbon monoxide comes from  both natural and anthropogenic processes.  About half of
15     the CO is released at the earth's surface while the rest is produced in the atmosphere. Many
       papers on the global sources of CO have been published over the la^t 15 years; whether most
       of the CO in the atmosphere is from human  activities or from natural processes has been
       debated for nearly as long.  Before 1970 it was believed that CO in the troposphere was
       almost all man-made (Jaffe, 1968, 1973).  Later, based on the theory that oxidation of CH4
20     produces large amounts of CO, it was suggested that much of the CO in the non-urban
       troposphere was of natural origin (Levy, 1971, 1973; McConnell et al., 1971;  National
       Research Council, 1977; Weinstock and Niki, 1972; Wofsy, 1976; Wofsy et al., 1972).
       However, this view was controversial (Newell, 1977; Stevens et al., 1972).  At that time, it
       commonly was believed that CH4 came  from natural processes, although the existing
25     tabulation of the sources suggested otherwise (Ehhalt and Schmidt, 1978). Even now the
       source of CO from the oxidation of  CH4 often is regarded as natural as opposed to the direct
       emissions of CO from combustion processes.  However, there is good evidence that about
       half the CH4 in the atmosphere is from human activities, particularly rice paddies, cattle,
       urban areas, landfills, and other sources (Khalil and Rasmussen,  1983). Therefore, regardless
30     of how much CO is estimated to come from the  oxidation of CH4 about half of it could be
       considered to come indirectly from anthropogenic activities.  In recent years the estimates of

       March 5, 1990                            4-2      DRAFT-DO NOT QUOTE OR CITE

-------
       the average level of OH* radicals have been revised downward so that the production of CO
       from CH4 and other hydrocarbons no longer is thought to be the dominant source (Hameed
       and Stewart,  1979; Logan et al., 1981; Pinto et al., 1983).
            The recent budgets that take into account previously published data, suggest that human
 5     activities are responsible for about 60% of the CO in the non-urban troposphere while natural
       processes account for the remaining 40%.  It also appears that combustion processes directly
       produce about 40% of the annual emissions of CO (Jaffe, 1968, 1973; Robinson and Robbins,
       1969, 1970; Swinnerton et al.,  1971), while oxidation of hydrocarbons make up most of the
       remainder (Greenberg et al., 1985; Hanst et al.,  1980; Rasmussen and Went, 1965; Went,
10     1960, 1966; Zimmerman et al., 1978) (about 50%) along with other sources such as the
       oceans (Bauer et al.,  1980; DeMore et al., 1985; Lamontagne et al., 1971; Logan et al.,
       1981; Linnenbom et al., 1973; Liss and Slater, 1974; National Research Council,  1977;
       Seiler, 1974;  Seiler and Junge,  1970; Seller and  Schmidt, 1974; Swinnerton et al., 1969,
       1974) and vegetation (Bauer et al., 1980;  Bidwell and Fraser, 1972;  DeMore et al., 1985;
15     Krall and Tolbert, 1957; Logan et al., 1981; National Research  Council,  1977;  Seiler, 1974;
       Seiler and Giehl, 1977; Seiler et al., 1978; Seiler and Junge, 1970; Siegel et al., 1962;
       Wilks, 1959). Some of the hydrocarbons that eventually end up as CO also are produced  by
       combustion processes constituting an indirect source of CO from combustion. These
       conclusions are summarized in Table 4-1 which is adapted from the 1981 budget of Logan
20     et al. (1981) in  which most of the previous work was incorporated (Logan et al., 1981;  World
       Meteorological  Organization, 1986), including the CO budget of Seiler (1974).  The total
       emissions of CO are about 2,600 Tg/year. Other budgets by Volz et al. (1981) and by Seiler
       and Conrad (1987) are reviewed by Warneck (1988).  Global emissions between 2,000 and
       3,000 Tg/year are consistent with these budgets.
25
       4.2.2  Sinks
            It is believed that reaction with OH* radicals is the major sink for removing CO from
       the atmosphere.  The cycle of OH* itself cannot be uncoupled from the cycles of CO,  CH4,
       water (H2O),  and O3. In the troposphere OH* radicals are produced by the photolysis of O3
30     (hu + O3 —>  O('D) + O2) followed by the reaction of the excited  oxygen atoms with H2O
       March 5, 1990                            4-3       DRAFT-DO NOT QUOTE OR CITE

-------
                          TABLE 4-1.  SOURCES OF CARBON MONOXIDE
                                         (Teragrams per year)
10
15
20
25
30
35
40
45
50

I.

II.

in.


Directly from Combustion
Fossil Fuels
Forest Clearing
Savanna Burning
Wood Burning
Forest Fires
Oxidation of Hydrocarbons
Methane
Non-methane HCs
Other Sources
Plants
Oceans
TOTALS (Rounded)
Notes:
1.

Table adapted from Logan el
Anthropogenic

500
400
200
50

300
90

-
1,500

t al. (1981) and revisions reporte
Natural Global

500
400
200
50
30 30

300 600
600 690

100 100
40 40
1,100 2,600

d by World Meteorological Organization (198
Range

400 -
200 -
100 -
25 -
10 -

400 -
300 -

50 -
20 -
2,000 -

6). All estimates


1,000
800
400
150
50

1,000
1,400

200
80
3,000

are in
     Tg/year of CO. Tg/year = megatons/year = 10 g/year.

2.    All estimates are expressed to one significant figure. The sums are rounded to two significant digits.

3.    Half the production of CO from the oxidation of CH4 is attributed to anthropogenic sources and the other half to natural sources
     based on the budget of CH4 from Khalil and Rasmussen (1984c).
vapor to produce two hydroxyl radicals (O(1D) + H2O —> OH* + OH").  The production
of OH* radicals is balanced by their removal principally by reactions with CO and CH4.  On a
global scale CO removes many more OH* radicals than CH4, however CH4 becomes more
important in the southern hemisphere where there is much less CO than in the northern
hemisphere but the amount of CH4 is only slightly less.
     The amount of CO that is removed by reactions with OH* radicals can be estimated by
calculating the loss as  = KJ^OH'LJCO]™.  The reaction rate constant of CO +  OH* is K =
(1.5 x 10'13)(1 + 0.6 PaJ cmVmolecules-s  (DeMore et al.,  1987).  K^ describes the effective
reaction rate constant taking into account the decreasing atmospheric pressure and decreasing
CO concentrations with height. Estimating  K^ to be 2 X 10'13 cmVmolecules-s and taking
[OH*]ave to be 8 x 105  molecules/cm3 and [CO].W to be 90 ppbv, the annual loss of CO from
reactions with OH* is about 2,200 Tg/year.  The values adopted for [OH*]>ve and  [CO]m are
discussed in more detail later in this chapter.
       March 5, 1990
                                         4-4      DRAFT - DO NOT QUOTE OR CITE

-------
            Uptake of CO by soils has been documented and may amount to about 250 Tg/year or
       about 10% of the total emitted into the atmosphere (Bartholomew and Alexander, 1981;
       Ingersoll et al., 1974; Inman et al.,  1971; Seiler and Schmidt, 1974), although arid soils may
       release CO into the atmosphere (Conrad and Seiler, 1982).  Another 100 Tg (5%) or so are
 5     probably removed annually in the stratosphere (Seiler, 1974).

       4.2.3 Atmospheric Lifetime
            Based on the global sources and sinks described above, the average atmospheric lifetime
       of CO can be calculated to be about 2 months with a range of between 1 and 4 months which
10     reflects the uncertainty in the annual emissions of CO (T =  C/S).  The lifetime, however, can
       vary enormously with latitude and season compared to its global average value.  During
       winters at high and middle latitudes CO has a lifetime of more than a year but during
       summers at mid latitudes the lifetime may be closer to the average global lifetime of about
       2 months. Moreover, in the tropics the average lifetime of CO is probably about 1  month.
15     These calculated variations reflect the seasonal cycles of OH* at various latitudes.

       4.2.4 Latitudinal Distribution of Sources
            When the sources, sinks, transport, and observed concentrations are combined into a
       mass balance model it is possible to calculate any one of these four components if the others
20     are known.  In the case of CO, the sources can be estimated assuming that the sinks (OH*
       reaction and soils), transport, and concentrations are known. The latitudinal distribution of
       sources can be described in a one-dimensional model as  follows:
25     "	^	   —•—    -  —  ™*   H(/i)7}-C(M,t)          (4-1)
30     where
             C = tropospheric mixing ratio
             K = zonally and height averaged transport coefficient
             T = the lifetime
35           S = emissions
             £ = a factor to account for the lower concentrations of CO in the stratosphere
             H = a factor to account for the variation of the tropopause height with latitude
             H = the sine of latitude
             R = the radius of the earth
       March 5, 1990                            4-5      DRAFT-DO NOT QUOTE OR CITE

-------
       This model is similar to that described by Czeplak and Junge (1974) and Fink and Klais
       (1978).  A time-averaged version was applied to the CO budget by Hameed and Stewart
       (1979) and a somewhat modified and time-dependant version shown above was applied by
       Khalil and Rasmussen (1988c) to derive the latitudinal distribution of CO shown in
 5     Figure 4-1.  Calculations by Khalil and Rasmussen (1988c) also suggest that emissions are
       higher in spring and summer compared with the other seasons, particularly in the middle
       northern latitudes.  This is expected for two reasons: (1) oxidation of CH4 and other
       hydrocarbons is faster during the summer because of the seasonal variation of OH", and
       (2) other direct emissions are also greater during spring and summer.
10          From Figure 4-1  the emissions from the northern and southern tropical latitudes sum up
       to 480 Tg/year and 330 Tg/year, respectively; from the northern and southern middle
       latitudes the emissions  are 960 Tg/year and 210 Tg/year; from the Arctic some 50 Tg are
       emitted  each year and some 10 Tg/year come from the Antarctic. The largest fluxes of CO
       are from the industrial  band of latitudes between 30° to 50° North.  From this region some
15     620 Tg/year are emitted representing about 30% of the total emissions of 2050 Tg/year.  The
       model does not distinguish between the anthropogenic or natural sources nor does it
       distinguish between direct emissions and photochemical production of CO from the oxidation
       of hydrocarbons.  Much of estimated fluxes from the mid-northern latitudes and from tropical
       regions  are likely to be of anthropogenic origin. The latitudinal distribution in Figure 4-1 is
20     compatible with the estimate (Table 4-1) that about 60% of the total emissions are from
       anthropogenic activities.

       4.2.5  Uncertainties and Consistencies
             The first consistency one notes is that the total emissions of CO estimated from the
25     various  sources are balanced by the estimated removal of CO.  The approximate balance
       between sources and sinks is expected because the trends, to be discussed later, are only about
       4 to 8 Tg/year compared to the total global emission rate of more than 2000 Tg/year.  On the
       other hand, there are many uncertainties in the sources and sinks.
        March 5, 1990                           4-6      DRAFT-DO NOT QUOTE OR CITE

-------
GO
o
o
IH
2
o
CO

O
O
                               Sine  of  Latitude
         -1.0
                -0.6
     -0.2
            0.2
     0.6
            1.0
       50
      40  -
       30  -
20   -
10  -
         0
                 CO
                 to
                  I
                 00
                  I
CO
CM
 I
-    Eq     •:
 I
                                      Latitude
CO
CM
10
CO
                                                               CO
                                                                     N
       Figure 4-1. The estimated sources of CO as a function of latitude.  The sources are, in
       teragrams per year in each latitude band 0.02 units in sine of latitude. The solid lines are
       estimates of uncertainties as OH* concentrations and the rate of dispersion are varied
       simultaneously so that the maximum values of each of these parameters are twice the
       minimum values.

       Source: Khalil and Rasmussen (1988c).
  March 5, 1990
                              4-7
              DRAFT-DO NOT QUOTE OR CITE

-------
     There are large uncertainties in the estimates of emissions from individual sources as
expressed in Table 4-1.  In most cases the stated uncertainty is a qualitative expression of the
likely range of emissions and it cannot be interpreted statistically.  Therefore the resulting
uncertainty in the total emissions, obtained by adding up the uncertainties in individual
sources, appears to be large.
     There are two difficulties in improving the estimates of CO from individual sources.
First, although many critical experiments to determine the production and emissions of CO
from individual sources are yet to be done, there is a limit to the accuracy with which
laboratory data can be extrapolated to the global scale.  Second, the cycle of CO may be so
intimately tied up with the cycles of hydrocarbons that accurate global estimates of CO
emissions may not be possible until the cycles of the hydrocarbons are better understood.
     Whereas the global distribution and seasonal variations of OH* can be calculated, there
are no direct measurements of OH" that can be used to estimate the removal of CO. The
effective average concentration of OH* that acts on trace gases can be estimated indirectly
from the cycles of other trace gases with known global emissions.  Therefore the total
emissions of CO are constrained by the budgets of other trace gases even though the estimates
of emissions from individual sources may  remain uncertain. The most notable constraint may
be the budget of methylchloroform (CH3CC13).  Methylchloroform is a degreasing solvent that
has been emitted into the atmosphere in substantial quantities for more than 20 years. It is
thought to be removed principally by reacting with OH* radicals and to a lesser extent by
photodissociation in the  stratosphere.  Because industry records on CH3CC13 production and
sales have been kept for a long time, it can be used to estimate the average amount of OH*
radicals needed to explain the observed concentrations compared to the emissions.  The
accuracy of the source estimates of CH3CC13 is improved by the patterns of its  uses; most of it
tends to be released shortly after purchase, so that large unknown or unqualified  reservoirs
probably do not exist. The recent budgets of CH3CC13 suggest that on an average there are
about 8 X  10s molecules of OH* per cubic centimeter although significant  uncertainties remain
(see for example Khalil  and Rasmussen, 1984c; Prinn  et al.,  1987).  This  is the value used
earlier in estimating the  loss of CO from reaction with OH*.  The same average value of OH*
also explains the CH4 concentrations compared to estimated sources,  lending more support to
the accuracy of the estimated OH* concentrations.  Neither of these constraints is very

March 5, 1990                           4-8       DRAFT-DO NOT QUOTE OR CITE

-------
stringent; however, if the total global emissions of CO from all sources are much different
from the estimated 2600 Tg/year then revisions of the budgets of both CH4 and CH3CC13 may
be required.
     There are other sources and sinks of CO, believed to be of lesser importance on a global
scale, which are reviewed in the previous EPA criteria document on CO (Chan et al., 1977;
Swinnerton et al., 1971).
4.3 GLOBAL DISTRIBUTIONS
     Atmospheric concentrations and thus the global distribution generally are the most
accurately known components of a global mass balance of a trace gas because direct
atmospheric measurements can be taken (Dianov-Klokov and Yurganov, 1981; Ehhalt and
Schmidt, 1978; Fraser et al., 1986; Heidt et al., 1980; Hoell et al., 1984; Khalil and
Rasmussen,  1988a,b; Pratt and Falconer, 1979; Rasmussen and Khalil, 1982; Reichle et al.,
1982; Seiler, 1974; Seiler and Fishman,  1981; Wilkness et al., 1973).  Much has been
learned about the global distribution of CO over the last decade. The experiments leading to
the present understanding range from systematic global observations at ground level for the
last 8 to 10 years reported by Khalil and Rasmussen (Khalil and Rasmussen,  1988a,b) and
Seiler (Seiler, 1974; Seiler and Junge, 1970), to finding the instantaneous global distribution
of CO  from  remote-sensing instruments on board NASA's space shuttle as reported by
Reichle et al. (1982, 1989a).

4.3.1  Seasonal Variations
     The seasonal variations of CO are well established (Dianov-Klokov and Yurganov,
1981; Fraser et al., 1986;  Khalil and Rasmussen, 1988b; Seiler et al., 1984).  High
concentrations are observed during the winters in each hemisphere and lowest concentrations
are seen in late summer.  The amplitude  of the cycle is largest at high northern latitudes and
diminishes as one moves towards the equator until it is reversed in the southern hemisphere
reflecting the reversal of the seasons. The seasonal variations are small in the equatorial
region. These patterns are expected from the seasonal variations of OH* concentrations. At
mid and high latitudes, diminished  solar radiation, water vapor, and O3 during winters cause

March  5, 1990                           4-9       DRAFT-DO NOT QUOTE OR CITE

-------
the concentrations of OH* to be much lower than during summer. The removal of CO is
slowed down and its concentrations build up.  In summer the opposite effect exists causing
the large seasonal variations of CO.  These variations are apparent in the observed global
seasonal cycles shown in Figure 4-2a.
     On the hemispherical scale the seasonal variation of CO is approximately proportional to
the concentration. Therefore, because there is much more CO in the northern hemisphere
than in the southern hemisphere, the  decline of concentrations during the summer of the
northern hemisphere is not balanced by the rise of concentrations in the southern hemisphere.
This causes a global seasonal variation.  The total amount of CO in the earth's atmosphere
undergoes a remarkably large seasonal variation with highest global burden during northern
winters and lowest during northern summers.  This feature is shown in Figure 4-2b.

4.3.2  Latitudinal  Variation
     The global  seasonal variation of CO in the earth's atmosphere also creates a seasonal
variation in the latitudinal distribution (Khalil and Rasmussen, 1988a,b; Newell et al.,  1974;
Reichle et al., 1982, 1986; Seiler, 1974).  During northern winters, CO levels are at their
highest in the northern hemisphere whereas southern hemisphere concentrations are at a
minimum.  The interhemispheric gradient, defined as the ratio of the amount of CO in the
northern and southern hemisphere, is at its maximum of about 3.2 during northern
hemisphere winters and falls  to about 1.8 during northern hemisphere summers, which is
about half of the winter value. The average latitudinal gradient is about 2.5, which means
that on an average there is about 2.5 times as much CO in the northern hemisphere as in the
southern hemisphere.  Earlier data on the latitudinal variations did not account for the
seasonal variations.

4.3.3   Variations with Altitude
     In the northern hemisphere troposphere, the concentrations of CO generally decline with
altitude but in the southern hemisphere the vertical gradient  may be reversed due to the
transport of CO from  the northern hemisphere into the southern hemisphere. Above the
tropopause concentrations decline rapidly so that there is very little  CO between 20 km and


March 5, 1990                            4-10     DRAFT - DO NOT QUOTE OR CITE

-------
            a
            P.
            O
            U
            S  -20
                -40  -
            ft
            o,
            O
            U
            o
            •O
                                  Time  of year
-10
                -20
                                                      o   AK

                                                      A   OR

                                                      a   HA

                                                      O   SA

                                                      V   TA

                                                      +   SP
                                                          GL
Figure 4-2.  The global seasonal variations of CO.  Figure 4-2a shows the seasonal cycle at
6 sites in polar, middle, and tropical latitudes of both hemispheres (AK = Alaska, OR =
Oregon, HA = Hawaii, SA = Samoa, TA = Tasmania, SP = South Pole). Figure 4-2b
shows the seasonal variation of the global burden of atmospheric CO.  The atmospheric
content of CO is much higher during northern hemisphere winters compared to summers.

Sources: Khalil and Rasmussen (1988b).
March 5,  1990
                      4-11      DRAFT-DO NOT QUOTE OR CITE

-------
       40 km; at still higher altitudes the mixing ratio may increase again (Fabian et al., 1981; Seller
       and Junge, 1969; Seiler and Warneck, 1972).

       4.3.4 Other Variations
 5          The concentration of CO generally is higher over populated continental areas compared
       to the air over oceans,  even though oceans release CO into the atmosphere.  Other regions,
       such as tropical forests also may be a source of isoprene and other hydrocarbons that may
       form CO in the atmosphere.  Such sources produce shifting patterns of high CO
       concentrations over regional and perhaps even larger spatial scales.  Concentrations are
10     representative of the middle troposphere and were measured during the 1984 flights of the
       space shuttle and reported by Reichle et al. (1989a). Eventually, CO in the lower troposphere
       may be measured from space using the techniques described by Reichle et al. (1989b).  The
       new method uses gas correlation filter radiometry at 2.3 /Ltm in addition to the 4.67 pun line
       used earlier to obtain mid-tropospheric concentrations of CO.
15          Occasionally in some locations significant diurnal variations of CO also may occur.  For
       instance, diurnal variations have been observed over some parts of the oceans with high
       concentrations during the day and low concentrations at night.  Because similar patterns also
       exist in the surface sea water, the  diurnal variations in the air can be explained by emissions
       from the oceans.
20          Finally, after the repeating cycles and other trends are subtracted, considerable random
       fluctuations still remain in time  series of measurements. These fluctuations reflect the short
       lifetime of CO and the vicinity of the sources and complicate the detection of long-term trends
       (see Figure 4-3).

25
       4.4  GLOBAL TRENDS
            Because some 60%  of the  global emissions of CO  are believed to come from
       anthropogenic sources  with increasing emissions, it stands to reason that the global
        March 5, 1990                            4-12     DRAFT-DO NOT QUOTE OR CITE

-------
             .0
             ex
             p.
             O
             U
170
160
150
140
130
120
110
100
 90
 SO
 70
 60
 50
 40
 30
 20
 10
                        1979
              1982        1985
                Time  (months)
                 1988
                      •—-Alaska   ±—* Oregon   •—• Hawaii
                      T—»Samoa   *—• Tasmania  *   "Antarctic
Figure 4-3. The global concentrations and trends of CO.  The seasonal cycles have been
subtracted from the time series of measurements at various latitudes ranging from inside the
Arctic Circle to the South Pole. The latitudinal variation of CO also is apparent in the figure.
Sources:  Khalil and Rasmussen (1988a).
March 5, 1990
              4-13
DRAFT-DO NOT QUOTE OR CITE

-------
       concentration of CO should be increasing. At present, there are several independent pieces of
       evidence for an increasing trend although none are definitive.  First, direct atmospheric
       observations reported by Khalil and Rasmussen (1984a) showed a detectable trend at Cape
       Meares in Oregon between 1979 and 1982. Over these 3 years the rate of increase was about
 5     5%/year.  Subsequent data from the same site showed that the rate was not sustained for long
       and a much smaller trend of a somewhat less  than 2%/year emerged over the longer period of
       1970 to 1987 (Khalil and Rasmussen, 1988a). Similar data from other sites distributed
       worldwide now show that there is evidence for a global increase of about 1 % per year as
       shown in Figure 4-3 (Khalil and Rasmussen,  1988a). This is the only study in which trends
10     from different parts of the world are evaluated.  It shows that the trends are strongest in the
       mid-northern latitudes where most of the sources are located and become smaller and weakei
       in the southern hemisphere.  At the mid-southern latitude site the trends persist but are not
       statistically significant (Khalil and  Rasmussen, 1988a).  Second, Rinsland  and Levine (1985)
       have reported estimates of CO concentrations from spectroscopic plates from Europe that
15     show that between 1950 and  1984  CO increased at about 2% per year. Finally, spectroscopic
       measurements of CO taken by Dvoryashina et al.  and Dianov-Klokov et al. in the Soviet
       Union also suggest an increase of about 2%/year between 1974 and 1982 (Dianov-Klokov
       et al., 1978;  Dianov-Klokov and Yurganov,  1981; Dvoryashina et al., 1982, 1984; Khalil and
       Rasmussen, 1988a; Khalil and Rasmussen, 1984b).
20          There is good evidence that the concentrations of CO are increasing in the non-urban
       troposphere,  however the rate of increase still is not well known and may  vary considerably
       over time. The random variability is so large and the trends so small that there are just
       enough data to detect the increase  but not enough to estimate the long-term rate of increase
       with confidence.  Such increases of tropospheric CO can cause a reduction of OH*
25     concentrations and thus reduce the oxidizing capacity of the atmosphere causing other trace
       gases, including CH4, to build up more rapidly in the atmosphere and reach higher levels.
       This occurrence could add to the greenhouse  effect and deplete the stratospheric O3 layer.  In
       the troposphere, increased CO in the presence of NOX also could result in  an increase of O3
       concentrations.
30           All the studies show increases of 1 to 2% per year over the last several decades.  These
       trends and the sources shown in Table 4-1 suggest that the anthropogenic  sources, both direct

       March 5, 1990                            4-14      DRAFT-DO NOT QUOTE OR CITE

-------
        and indirect, probably were very small until this century.  Therefore, the average CO
        concentration may have doubled over the last 50 years or so.  This change could be a
        significant contributing factor to increasing levels of O3 in the non-urban troposphere.
        Because the influence of CO on tropospheric O3 is not understood fully, the role of increasing
 5      CO on tropospheric O3 also remains uncertain.
        4.5  SUMMARY
             The annual global emissions of CO are estimated to be about 2600 ± 600 Tg, of which
 10      about 60% are from human activities including combustion of fossil fuels and oxidation of
        hydrocarbons including CH4.  The remaining 40% of the emissions are from natural
        processes, mostly from the oxidation of hydrocarbons, but also from plants and the oceans.
        Almost all the CO emitted into the atmosphere each year is removed by reactions with OH*
        radicals (85%), by soils (10%), and by diffusion into the stratosphere.  There is a small
 15      imbalance between annual emissions and removal causing an increase of about 1 % per year.
        It is very likely that the imbalance is due to increasing emissions from anthropogenic
        activities. The average concentration of CO is about 90 ppbv, which amounts to about
        400 Tg in the atmosphere and the average lifetime is about 2 mo. This view of the global
        cycle of CO is consistent with the present estimates of average OH* concentrations and the
20      budgets of other trace gases including CH4 and CH3CC13.
             There are large remaining uncertainties that in the future may upset the apparently
        cohesive present budget of CO.  Although the patterns of the global distribution are becoming
        established, there still are uncertainties about the absolute concentrations. Estimates of
        emissions from individual sources are very uncertain, however the total annual emissions are
25      likely to be more accurate.
            There are just sufficient data on the trends to suggest that CO is increasing, but the rates
        are not certain.   However, if the present view  of the global cycle of CO is correct, then it is
        likely that, in time, increasing levels of CO will contribute to widespread changes in
        atmospheric chemistry and the global climate.
30
       March 5, 1990                            4-15      DRAFT-DO NOT QUOTE OR CITE

-------
       4.6  REFERENCES
 5     Bartholomew, G. W.; Alexander, M. (1981) Soil as a sink for atmospheric carbon monoxide. Science
              (Washington, DC) 212: 1389-1391.

       Bauer, K.; Conrad, R.; Seiler, W. (1980) Photooxidative production of carbon monoxide by phototrophic
              microorganisms. Biochim. Biophys. Acta 589: 46-55.
10
       Bidwell, R. G. S.; Fraser, D. E.  (1972) Carbon monoxide uptake and metabolism by leaves. Can. J. Bot. 50:
              1435-1439.

       Chan, W. H.; Uselman, W. M.; Calvert, J. G.; Shaw, J. H. (1977) The pressure dependence of the rate constant
15            for the reaction: OH* + CO — > H' +  CO2. Chem. Phys. Lett. 45: 240-244.

       Conrad, R.; Seiler, W. (1982) Arid soils as a source of atmospheric carbon monoxide. Geophys. Res. Lett.
              9: 1353-1356.

20     Crutzen, P. J. (1974) Photochemical reactions initiated by and influencing O3 in unpolluted tropospheric air.
              Tellus 26: 47-56.

       Czeplak, G.; Junge, C. (1974) Studies of interhemispheric exchange in the troposphere by a diffusion model. In:
              Frenkiel, F. N.; Munn, R. E., eds. Turbulent diffusion in environmental pollution: proceedings of a
25            symposium; April 1973; Charlottesville, VA. New York,  NY: Academic Press; pp. 57-72.

       DeMore, W. B.; Margitan, J. J.; Molina, M. J.; Watson, R. T.; Golden, D. M.; Hampson, R. F.; Kurylo,
              M. J.;  Howard, C. J.; Ravishankara, A. R. (1985) Chemical  kinetics and photochemical data for use in
              stratospheric modeling: evaluation number 7. Washington, DC:  National Aeronautics and Space
30            Administration; report no. NASA-CR-176198. Available from: NTIS, Springfield, VA; N86-10707.

       DeMore, W. B.; Molina, M. J.; Sander, S. P.; Golden, D. M.; Hampson, R. F.; Kurylo, M. J.;  Howard, C.  J.;
              Ravishankara, A. R.  (1987) Chemical kinetics and photochemical data for use in stratospheric modeling:
              evaluation number 8. Washington, DC: National Aeronautics  and Space Administration;; report
35            no.  NASA-CR-182919. Available from: NTIS, Springfield, VA; N88-24012.

       Dianov-Klokov, V. I.; Yurganov, L. N. (1981) A spectroscopic study of the global space-time distribution of
              atmospheric CO. Tellus 33: 262-273.

40     Dianov-Klokov, V. I.; Fokeyeva, YE. V.; Yurganov, L. N. (1978) A study of the carbon monoxide content of
              the atmosphere. Izv.  Acad. Sci.  USSR Atmos. Oceanic Phys.  (Engl. Transl.) 14: 263-269.

       Dianov-Klokov, V. I.; Yurganov, L. N. (1982) Izmereniya integral'nogo soderzhaiya prime sei CO, CH4 i N2O v
              atmosfere [The measurements of integral atmospheric CO, CH4 and N2O abundances in the atmosphere].
45            Izv. Akad. Nauk SSSR Fiz. Atmos. Okeana 18:  738-743.

       Dvoryashina, E. V.; Dianov-Klokov, V. I.; Yurganov, Y.  L.  (1982)  Results of carbon monoxide  abundance
              measurements at Zvenigorod, 1970-1982. Moscow, USSR: USSR Academy of Sciences (preprint).

50     Dvoryashina, E. V.; Dianvo-Klokov, V. I.; Yurganov, Y.  L.  (1984)  Variations of the carbon monoxide content
              of the atmosphere for 1970-1982. Izv. 20: Akad. Nauk SSSR Fiz.  Atmos. Okeana 40-47.

       Ehhalt, D. H.; Schmidt, U. (1978) Sources and sinks of atmospheric  methane. Pure Appl. Geophys.
              116: 452-464.

         March 5, 1990                               4-16       DRAFT-DO NOT QUOTE OR CITE

-------
       Fabian, P.; Borchers, R.; Flentje, G.; Matthews, W. A.; Seiler, W.; Giehl, H.; Bunse, K.; Mueller, F.;
               Schmidt, U.; Volz, A.; Khedim, A.; Johnen, F. J. (1981) The vertical distribution of stable trace gases at
               mid-latitudes. JGR J. Geophys. Res. 86: 5179-5184.

 5     Fink, H. J.; Klais, O. (1978) Global distribution of fluorocarbons. Ber. Bunsen Ges. Phys. Chem.
               82: 1147-1150.

       Fishman, J.; Crutzen, P. J. (1978) The origin of O3 in the troposphere. Nature (London) 274: 855-858.

10     Fishman, J.; Seiler, W. (1983) Correlative nature of O3 and carbon monoxide in the troposphere: implications for
               the tropospheric O3 budget. JGR J. Geophys. Res. 88: 3662-3670.

       Fishman, J.; Seiler, W.; Haagenson, P. (1980) Simultaneous presence of O3 and CO bands in the troposphere.
               Tellus 32: 456-463.
15
       Fraser, P. J.; Hyson, P.; Rasmussen, R. A.; Crawford, A. J.; Khalil, M. A. K. (1986) Methane, carbon
               monoxide and methylchloroform in the Southern Hemisphere. J. Atmos. Chem. 4: 3-42.

       Greenberg, J. P.; Zimmerman, P. R.; Chatfield, R. B. (1985) Hydrocarbons and carbon monoxide in African
20             savannah air. Geophys. Res.  Lett.  12: 113-116.

       Hameed, S.; Stewart, R. W. (1979) Latitudinal distribution of the sources of carbon monoxide in the
               troposphere. Geophys. Res. Lett. 6: 841-844.

25     Hanst, P. L.; Spence, J. W.; Edney, E.  O. (1980) Carbon monoxide production in photooxidation of organic
               molecules in the air. Atmos.  Environ. 14:  1077-1088.

       Heidt, L. E.; Krasnec, J. P.; Lueb, R. A.; Pollock, W. H.; Henry,  B. E.; Crutzen, P. J. (1980) Latitudinal
               distributions of CO and CH4 over the Pacific. JGR J. Geophys.  Res. 85: 7329-7336.
30
       Hoell, J. M.; Gregory, G. L.; Carroll, M. A.; McFarland, M.; Ridley, B. A.;  Davis, D. D.; Bradshaw, J.;
               Rodgers, M. O.; Torres, A.  L.; Sachse, G. W.; Hill, G. F.; Condon, E. P.; Rasmussen, R.  A.;
               Campbell, M. C.; Farmer, J. C.; Sheppard, J.  C.; Wang, C. C. ; Davis, L. I. (1984) An inter-
               comparison of carbon monoxide, nitric oxide, and hydroxyl measurement techniques: overview of results.
35             JGRJ. Geophys. Res. 89:  11,819-11,825.

       Ingersoll, R. B.; Inman, R. E.; Fisher, W.  R. (1974) Soil's potential as a sink for atmospheric carbon monoxide.
               Tellus 26:  151-159.

40     Inman, R. E.; Ingersoll, R. B.; Levy, E. A. (1971) Soil: a natural sink for carbon monoxide. Science
               (Washington, DC) 172: 1229-1231.

       Jaffe, L. S. (1968) Ambient carbon monoxide and its fate in the atmosphere. J.  Air Pollut. Control Assoc.
               18: 534-540.
45
       Jaffe, L. S. (1973) Carbon monoxide in the biosphere:  sources, distribution, and concentrations. JGR J.  Geophys.
               Res. 78: 5293-5305.

       Khalil, M. A.  K.; Rasmussen, R. A. (1983) Sources, sinks, and seasonal cycles of atmospheric methane. JGR J.
50             Geophys. Res. 88: 5131-5144.

       Khalil, M. A.  K.; Rasmussen, R. A. (1984a) Carbon monoxide in the earth's atmosphere: increasing trend.
               Science (Washington, DC) 224: 54-56.
         March 5, 1990                                4-17       DRAFT-DO NOT QUOTE OR CITE

-------
       Khalil, M. A. K.; Rasmussen, R. A. (1984b) The global increase of carbon monoxide. In: Aneja, V. P., ed.
              Environmental impact of natural emissions: proceedings of an Air Pollution Control Association specialty
              conference; March. Pittsburgh, PA: Air Pollution Control Association; pp. 403-414.

 5     Khalil, M. A. K.; Rasmussen, R. A. (1984c) The atmospheric lifetime of methylchloroform (CH3CC13). Tellus
              Ser. B 36: 317-332.

       Khalil, M. A. K.; Rasmussen, R. A. (1985) Causes of increasing atmospheric methane: depletion of hydroxyl
              radicals and the rise of emissions. Atmos. Environ. 19: 397-407.
10
       Khalil, M. A. K.; Rasmussen, R. A. (1988a) Carbon monoxide in the earth's atmosphere: indications of a global
              increase. Nature (London) 332: 242-245.

       Khalil, M. A. K.; Rasmussen, R. A. (1988b) The global distribution of carbon monoxide.  Oregon Graduate
15            Center preprint.

       Khalil, M. A. K.; Rasmussen, R. A. (1989) Atmospheric carbon monoxide: latitudinal distribution of sources.
              Geophys. Res. Lett.: submitted.

20     Krall, A.  R.; Tolbert, N. E.  (1957) A comparison of the light dependent metabolism of carbon monoxide by
              barley leaves with that of formaldehyde, formate and carbon dioxide.  Plant Physiol. 32: 321-326.

       Lamontagne, R.  A.; Swinnerton, J. W.; Linnenbom, V. J. (1971) Nonequilibrium of carbon monoxide and
              methane at the air-sea interface. JGR J. Geophys. Res. 76: 5117-5121.
25
       Levine, J. S.; Rinsland,  C. P.; Tennille, G. M. (1985) The photochemistry of methane and carbon monoxide in
              the troposphere in 1950  and 1985. Nature (London) 318: 254-257.

       Levy, H., II. (1971) Normal atmosphere: large radical and formaldehyde concentrations predicted. Science
30            (Washington, DC) 173:  141-143.

       Levy, H., II. (1973) Tropospheric budgets for methane, carbon monoxide,  and related species. JGR J. Geophys.
              Res. 78: 5325-5332.

35     Linnenbom, V. J.; Swinnerton,  J. W.; Lamontagne,  R. A. (1973) The ocean as a source for atmospheric carbon
              monoxide. JGR J. Geophys. Res. 78: 5333-5340.

       Liss, P. S.; Slater, P. G. (1974) Flux of gases across the air-sea interface. Nature (London) 247:  181-184.

40     Logan, J. A.; Prather, M. J.; Wofsy, S. C.; McElroy, M. B.  (1981) Tropospheric chemistry: a global
              perspective. JGR J. Geophys. Res.  C: Oceans 86: 7210-7254.

       McConnell, J. C.; McElroy, M. B.; Wofsy, S. C. (1971) Natural sources of atmospheric CO. Nature (London)
              233:  187-188.
45
       National Research Council. (1977) Carbon monoxide. Washington,  DC: National Academy of Sciences. (Medical
              and biologic effects of environmental pollutants).

       Newell, R. E. (1977) One-dimensional models: a critical comment, and their application to carbon monoxide.
50            JGR J. Geophys. Res. 82: 1449-1450.

       Newell, R. E.; Boer, G.  J., Jr.; Kidson, J. W. (1974) An estimate of the interhemispheric transfer of carbon
              monoxide from tropical  general circulation data. Tellus 26:  103-107.
         March 5,  1990                                4-18       DRAFT-DO NOT QUOTE OR CITE

-------
        Pinto, J. P.; Yung, Y. L.; Rind, D.; Russell, G. L.; Lerner, J. A.; Hansen, J. E.; Hameed, S. (1983) A general
               circulation model study of atmospheric carbon monoxide.  JGR J. Geophys. Res.  88: 3691-3702.

        Pratt, R.; Falconer, P. (1979) Circumpolar measurements of O3, particles,  and carbon monoxide from a
  5            commercial airliner. JGR J. Geophys. Res. 84: 7876-7882.

        Prinn, R.; Cunnold, D.; Rasmussen, R.; Simmonds, P.; Alyea, F.; Crawford, A.; Fraser, P.;  Rosen, R. (1987)
               Atmospheric trends in methylchloroform and the global average for the hydroxyl radical. Science
               (Washington, DC) 238: 945-950.
 10
        Rasmussen, R. A.; Khalil, M.  A. K. (1982) Latitudinal distributions of trace gases in and above the boundary
               layer. Chemosphere 11: 227-235.

        Rasmussen, R. A.; Went, F. W. (1965) Volatile organic material  of plant origin in the atmosphere. Proc. Natl.
 15            Acad. Sci. U. S. A. 53: 215-220.

        Reichle, H. G., Jr.; Beck, S. M.; Haynes, R. E.;  Hesketh, W. D.;  Holland, J. A.; Hypes,  W. D.; Orr,  H. D.,
               Ill; Sherrill,  R. T.; Wallio, H. A.; Casas, J. C.; Saylor, M. S.; Gormsen, B. B. (1982) Carbon
               monoxide measurements in the troposphere. Science (Washington, DC) 218: 1024-1026.
20
        Reichle, H. G., Jr.; Connors, V. S.; Holland, J. A.; Hypes, W. D.; Wallio, H. A.; Casas, J. C.; Gormsen,
               B. B.; Saylor, M. S.; Hesketh, W. D. (1986) Middle and upper tropospheric carbon monoxide mixing
               ratios as measured by a satellite-borne remote sensor during November 1981. JGR J. Geophys. Res.
               91: 10,865-10,887.
25
        Reichle, H. G., Jr.; Connors, V. S.; Wallio, H. A.; Holland,  J. A.; Sherrill,  R. T.; Casas, J.  C.; Gormsen,
               B. B. (1989a) The distribution of middle tropospheric carbon monoxide during early October 1984. JGR
               J. Geophys. Res.: submitted.

30     Reichle, H. G., Jr.; Wallio, H. A.; Gormsen, B. B. (1989b) Feasibility of determining the vertical profile of
               carbon monoxide from  a space platform. Appl. Opt. 28: 2104-2110.

        Rinsland, C. P.; Levine, J. S. (1985) Free tropospheric carbon monoxide concentrations in  1950 and 1951
               deduced from infrared total column amount measurements. Nature (London) 318: 250-254.
35
        Robinson, E.; Robbins, R. C. (1969) Sources, abundance, and fate of gaseous atmospheric pollutants:
               supplement. Menlo Park, CA: Stanford Research Institute.

        Robinson, E.; Robbins, R. C. (1970) Atmospheric background concentrations of carbon monoxide. In: Coburn,
40            R. F., ed. Biological effects of carbon monoxide. Ann. N. Y. Acad. Sci.  174: 89-95.

        Seiler, W.  (1974) The cycle of atmospheric CO. Tellus 26: 116-135.

        Seiler, W.; Conrad, R. (1987) Contribution of tropical ecosystems to the global budget of trace gases especially
45            CH4, H2,  CO and N2O. In: Dickenson, R., ed. Geophysiology of Amazonia. New York, NY: John
               Wiley.

        Seiler, W.; Fishman, J. (1981)  The distribution of carbon monoxide and O3 in the free troposphere. JGR  J.
               Geophys.  Res. 86: 7255-7265.
50
        Seiler, W.; Giehl, H. (1977) Influence of plants on the atmospheric carbon monoxide. Geophys. Res. Lett.
               4: 329-332.
        March 5, 1990                                4-19       DRAFT-DO NOT QUOTE OR CITE

-------
       Seller, W.; Junge, C. (1969) Decrease of carbon monoxide mixing ratio above the polar tropopause. Tellus
              21: 447-449.

       Seller, W.; Junge, C. (1970) Carbon monoxide in the atmosphere. JGR J. Geophys. Res. 75: 2217-2226.
 5
       Seller, W.; Schmidt, U. (1974) Dissolved nonconservative gases in seawater. In: Goldberg, E. D., ed. The sea:
              volume 5, marine chemistry. New York, NY: John Wiley & Sons, Inc.; pp. 219-243.

       Seiler, W.; Warneck, P.  (1972) Decrease of the carbon monoxide mixing ratio at the tropopause. JGR J.
10            Geophys. Res. 77: 3204-3214.

       Seiler, W.; Giehl, H.; Bunse, G. (1978) The influence of plants on atmospheric carbon monoxide and dinitrogen
              oxide.  Pure Appl. Geophys. 116: 439-451.

15     Seiler, W.; Giehl, H.; Brunke, E.-G.; Halliday, E. (1984) The seasonally of CO abundance in the Southern
              Hemisphere. Tellus Ser. B 36: 219-231.

       Siegel, S.  M.;  Renwick,  G.; Rosen, L.  A. (1962) Formation of carbon monoxide during seed germination and
              seedling growth.  Science (Washington, DC) 137: 683-684.
20
       Stevens, C. M.; Krout, L.; Walling, D.; Venters, A.;  Engelkemeir, A.; Ross, L. E. (1972) The isotopic
              composition of atmospheric carbon monoxide. Earth Planet. Sci. Lett.  16: 147-165.

       Swinnerton, J.  W.; Lamontagne, R. A.; Linnenbom, V. J.  (1971) Carbon monoxide in rainwater. Science
25            (Washington, DC) 172: 943-945.

       Swinnerton, J.  W.; Lamontagne, R. A.  (1974) Carbon monoxide in the South  Pacific Ocean. Tellus
              26: 136-142.

30     Swinnerton, J.  W.; Linnenbom, V. J.; Cheek, C.  H. (1969) Distribution of methane and carbon monoxide
              between the atmosphere and natural waters. Environ. Sci.  Technol. 3:  836-838.

       Swinnerton, J.  W.; Lamontagne, R. A.; Linnenbom, V. J.  (1971) Carbon monoxide in rainwater. Science
              (Washington, DC) 172: 943-945.
35
       Sze, N. D. (1977) Anthropogenic CO emissions: implications for the atmospheric CO-OH*-CH4 cycle. Science
              (Washington, DC) 195: 673-675.

       Thompson, A.  M.; Cicerone, R.  J. (1986) Possible perturbations to atmospheric CO, CH4, and OH*. JGR J.
40            Geophys. Res. D: Atmos. 91: 10853-10864.

       Volz, A.;  Ehhalt, D. H.; Derwent, R. G. (1981) Seasonal and latitudinal variation of HCO and the tropospheric
              concentration of  OH* radicals. JGR J. Geophys. Res. 86: 5163-5171.

45     Warneck,  P. (1988) Chemistry of the natural atmosphere. New York, NY: Academic Press, Inc.

       Weinstock, B.; Niki, H.  (1972) Carbon monoxide balance in nature. Science (Washington, DC)  176: 290-292.

       Went, F. W. (1960) Organic matter in the atmosphere, and its possible relation to petroleum formation. Proc.
50            Natl. Acad. Sci.  U. S. A. 46: 212-221.

       Went, F. W. (1966) On  the nature of the Aitken condensation nuclei. Tellus 18: 549-556.
         March 5,  1990                                4-20      DRAFT-DO NOT QUOTE OR CITE

-------
       Wilkness, P. E.; Lamontagne, R. A.; Larson, R. E.; Swinnerton, J. W.; Dickson, C. R.; Thompson, T. (1973)
              Atmospheric trace gases in the Southern Hemisphere. Nature (London) 245: 45-47.

       Wilks, S. S. (1959) Carbon monoxide in green plants. Science (Washington, DC) 129: 964-966.
 5
       Wofsy, S. C. (1976) Interactions of CH4 and CO in the earth's atmosphere. Annu. Rev. Earth Planet. Sci. 4:
              441-469.

       Wofsy, S. C.; McConnell, J. C.;  McElroy, M. B.  (1972) Atmospheric CH4, CO, and C02. JGR J. Geophys.
10            Res. 77: 4477-4493.

       World Meteorological Organization. (1986) Atmospheric O3 1985: assessment of our understanding of the
              processes controlling its present distribution and change. Geneva, Switzerland: World Meteorological
              Organization; Global O3 Research and Monitoring Project report no. 16. 3v.
15
       Zimmerman, P. R.; Chatfield, R. B.; Fishman, J.; Crutzen, P. J.; Hanst, P. L. (1978) Estimates on  the
              production of CO and H2  from the oxidation of hydrocarbon emissions from vegetation. Geophys. Res.
              Lett. 5: 679-682.
        March 5,  1990                                4-21       DRAFT-DO NOT QUOTE OR CITE

-------
    5.  MEASUREMENT METHODS FOR CARBON MONOXIDE

       5.1  INTRODUCTION
            To promote uniform enforcement of the air quality standards set forth under the Clean
 5     Air Act as amended (U.S. Senate Committee, 1977), the EPA has established provisions
       under which analytical methods can be designated as "reference" or "equivalent" methods
       (Code of Federal Regulations, 1977a).  A reference method or equivalent method for air
       quality measurements is required for acceptance of measurement data. An "equivalent
       method" for monitoring CO can be so designated when the method is shown to produce
10     results equivalent to the approved reference monitoring method based on absorption of
       infrared  radiation from a nondispersed beam.
            EPA-designated reference methods are automated, continuous methods utilizing the
       nondispersive infrared (NDIR) technique, which generally is accepted as being the most
       reliable method for the measurement of CO in ambient air. The official EPA reference
15     methods are described in Code of Federal Regulations, 1988. Eleven reference methods for
       CO have been designated for use in determining compliance and all methods employ the
       NDIR technique (Code of Federal Regulations, 1988).  Before a particular NDIR instrument
       can be used in a reference method, it must be designated  by the EPA as approved in terms of
       manufacturer,  model number, components, operating range,  etc.  Several NDIR instruments
20     have been so designated (Code of Federal Regulations, 1988), including the gas filter
       correlation (GFC) technique which was developed through EPA-sponsored research (Burch
       et al., 1976).  No equivalent methods that use a principle other than  NDIR have as of January
       1988 been designated for measuring CO in ambient air. The performance specifications for
       automated CO analyzers are shown in Table 5-1 (Code of Federal  Regulations, 1977a).
25          The normal full scale operating range for reference  methods is 0 to 50 ppm (0 to
       58 mg/m3).  Some instruments offer higher ranges, typically 0 to 100 ppm (0 to  116 mg/m3)
       or lower ranges such as 0 to 20 ppm (0 to 23 mg/m3).  A narrower range up to 1 ppm may be
       needed to measure background levels in unpolluted atmospheres.  Higher ranges up to
       1000 ppm (1150 mg/m3) are used to measure CO concentrations in vehicular tunnels and
30     parking garages.

       March 12, 1990                        5-1      DRAFT - DO NOT QUOTE OR CITE

-------
       5.1.1  Overview of Techniques for Measurement of Ambient Carbon
              Monoxide
            There have been several excellent reviews on the measurement of CO in the atmosphere
       (National Research Council, 1977; Driscoll and Berger, 1971; Harrison, 1975; American
 5     Industrial Hygiene Association, 1972; Leithe, 1971; Repp, 1977; Schnakenberg, 1976;
       Stevens and Herget,  1974; National Air Pollution Control Association, 1970; National
       Institute for Occupational Safety and Health, 1972; Verdin,  1973). The nondispersive
       infrared (NDIR) method is discussed widely in the literature (Dailey and Fertig, 1977;
       Houben, 1976; McKee and Childers,  1972; McKee et al., 1973; Perez et al.,  1975; Pierce
10     and Collins, 1971; Schunck, 1976; Scott, 1975; Smith and Nelson, 1973; Smith, 1969, and
       Luft, 1975). Currently, the most commonly-used measurement technique is the type of NDIR
       method referred to as gas filter correlation (Acton et al., 1973; Bartie and Hall, 1977; Burch
       and Gryvnak, 1974; Burch et al.,  1976; Chaney and McClenny, 1977; Goldstein et al., 1976;
       Gryvanak and Burch, 1976a,b; Herget et al., 1976; Ward and Zwick, 1975).  This technique
15     was developed to a commercial prototype stage through EPA sponsored research (Burch
       etal.,  1976).
            The NDIR method is an automated, continuous method that generally is accepted as
       being the most reliable method for the measurement of CO in ambient air.  NDIR analyzers
       are based on the specific absorption of infrared radiation by  the CO molecule (Feldstein,
20     1967).  Most commercially available analyzers incorporate a gas filter to minimize
       interferences from other gases; they operate at atmospheric pressure and the most sensitive
       analyzers are able to detect minimum CO concentrations of about 0.05 mg/m3.  Interferences
       due to  CO2 and water vapor can be dealt with so as not to affect the data quality. NDIR
       analyzers with Luft type detectors are relatively insensitive to flow rate, require no wet
25     chemicals,  are sensitive over wide concentration ranges, and have short response times.
       NDIR  analyzers of the newer GFC type have overcome zero and span problems and minor
       problems due to vibrations.
            A more sensitive method for measuring low, background levels is gas chromatography
       (Bergman et al., 1975; Bruner et al.,  1973; Dagnall et al., 1973; Porter and Volman, 1962;
30     Feldstein, 1967; Smith et al.,  1975a; Swinnerton et al., 1968; Tesarik and Krejci, 1974).
       This technique is an automated, semicontinuous method where CO is separated from water,

       March 12,  1990                         5-2     DRAFT - DO NOT QUOTE OR CITE

-------
                TABLE 5-1.  PERFORMANCE SPECIFICATIONS FOR AUTOMATED
                      ANALYTICAL METHODS FOR CARBON MONOXIDE
                          (CODE OF FEDERAL REGULATIONS, 1977a)

 5     Range                                              0 to 57 mg/m3 (0 to 50 ppm)
       Noise                                                0.6 mg/m3 (0.50 ppm)
       Lower detectable limit                                   1.2 mg/m3 (1.0 ppm)
       Interference equivalent
        Each interfering substance                              ±1.2 mg/m3 (±1.0 ppm)
10      Total interfering substances                              1.7 mg/m3 (1.5 ppm)
       Zero drift
        12 h                                                ±1.2 mg/m3 (±1.0 ppm)
        24 h                                                ±1.2 mg/m3 (±1.0 ppm)
       Span drift, 24 h
15      20% of upper range limit                                ± 10.0%
        80% of upper range limit                                ±2.5%
       Lag time                                              10 min
       Rise  time                                               5 min
       Fall time                                               5 min
20     Precision
        20% of upper range limit                               0.6 mg/m3 (0.5 ppm)
        80% of upper range limit                               0.6 mg/m3 (0.5 ppm)

       Definitions:
25
        Range:  Nominal minimum and maximum concentrations that a method is capable of
        measuring.
        Noise:  The standard deviation about the mean of short duration deviations in output that are
        not caused by input concentration changes.
30      Lower detectable limit: The minimum pollutant concentration that produces a signal of
        twice the noise level.
        Interference equivalent:  Positive or negative response caused by a substance other than the
        one being measured.
        Zero drift: The change in response to zero pollutant concentration during continuous
35      unadjusted operation.
        Span drift:  The percent change in response to an upscale pollutant concentration during
        continuous unadjusted operation.
        Lag time:  The time interval between a step change in input concentration and the first
        observable corresponding change in  response.
40      Rise time: The time interval between initial response and 95% of final response.
        Fall time:  The time interval between initial response to a step decrease in concentration and
        95% of final response.
        Precision: Variation about the  mean of repeated measurements of the same pollutant
        concentration expressed as one  standard deviation about the mean.
45
       March 12, 1990                          5-3     DRAFT - DO NOT QUOTE OR CITE

-------
       CO2, and hydrocarbons other than methane by a stripper column. Carbon monoxide and CH4
       then are separated on an analytical column and the CO is passed through a catalytic reduction
       tube where it is converted to CH4. The CO (converted to CH4) passes through a flame
       ionization detector (FID), and the resulting signal is proportional to the concentration of CO
 5     in the air.  This method has been used throughout the world. It has no known interferences
        and can be used to measure levels from 0.03 to 50 mg/m3.  These analyzers are expensive
       and require continuous attendance by a highly trained operator to produce valid results. For
       high levels, a useful technique is catalytic oxidation of the CO by Hopcalite or other catalysts
       (Stetter and Blurton, 1976), either with temperature-rise sensors (Naumann, 1975; Poli et al.,
10     1976; Schnakenberg, 1976) or with electrochemical sensors (Bay et al., 1974,  1972; Bergman
       et al.,  1975; Dempsey et al., 1975; Repp, 1977; Schnakenberg, 1975).
            Other analytical schemes used for CO in air include dual-isotope infrared fluorescence,
       another technique derived from NDIR (Link et al., 1971; McClatchie, 1972; McClatchie
       et al.,  1972); reaction with hot mercuric oxide to give elemental mercury vapor (Beckman
15     et al.,  1948; McCullough et al.,  1947; Mueller,  1954; Palanos,  1972; Robbins et al., 1968);
       reaction with heated iodine pentoxide to give elemental  iodine (Adams and Simmons, 1951;
       Moore et al., 1973; Newton and Morss, 1974; van Dijk and Falkenburg, 1976; Vol'berg  and
       Pochina, 1974); and color reactions (Allen and Root, 1955; Bell et al., 1975; Feldstein,
       1965; Jones,  1977; Lambert and Wiens, 1974; Levaggi and Feldstein, 1964; Ray et al.,
20     1975; Simonescu et al., 1975; Smith et al., 1975b), as with palladium salts or the silver salt
       of/?-sulfamoylbenzoate. Many of these methods are described in Section 5.3,  Measurement
       in Ambient Air.  A classical procedure for many decades was to use gasometric apparatus
       such as the Orsat or Haldane (Cormack, 1972), in which the CO present in a gas sample is
       absorbed by cuprous chloride solution and the decrease in volume or pressure is measured.
25     This method, however, is not sensitive enough for trace amounts.
            Microwave rotational spectroscopy is an analytical technique with high specificity
       (Hrubesh,  1973; Morgan and Morris, 1977).  Other possible ways to determine CO include
       chemiluminescent reaction with  ozone (v. Heusden and Hoogeveen, 1976), X-ray excited
       optical fluorescence (Goldstein et al., 1974),  radiorelease of 85Kr from the kryptonates of
30     mercuric oxide or iodine pentoxide (I2O5) (Goodman, 1972; Naoum et al., 1974), and
        March 12, 1990                           5-4      DRAFT - DO NOT QUOTE OR CITE

-------
       utilization of narrow-band infrared laser sources (Chancy et al.,  1979; Optical Society of
       America, 1975; Golden and Yeung, 1975).

       5.1.2  Calibration Requirements
 5          Whichever method or instrument is used, it is essential that the results be validated by
       frequent calibration with samples of known composition similar to the unknowns (Commins
       etal., 1977; Goldstein, 1977; National Bureau of Standards,  1975).  Chemical analyses can
       be relied on only after the analyst has achieved acceptable accuracy  in the analysis of such
       standard samples through an audit program.
10

       5.2  PREPARATION OF STANDARD REFERENCE MATERIALS
       5.2.1  Gas Standards
            A set of reliable gas standards for CO in air, certified at levels of approximately 12, 23,
15     and 46 mg/cm3 (10, 20, and 40 ppm) is obtainable from the National Institute of Standards
       and Technology (formerly National Bureau of Standards), Washington, DC (National Bureau
       of Standards, 1975). These Standard Reference Materials (SRMs) are supplied as compressed
       gas (at about 1700 psi) in high-strength aluminum cylinders containing 31 ft3 of gas at dry
       standard temperature and pressure and are accurate to better than 1 % of the stated values.
20     Because of the  time and effort required in their preparation, SRMs are not intended for use as
       daily working standards, but rather as primary standards against which transfer standards can
       be calibrated.

       5.2.2  Gravimetric Method
25          The gravimetric method  used by NBS for preparing primary standards of CO (Hughes,
       1975, 1976) is  as follows. An empty gas cylinder is tared on  an analytical balance; then 2 g
       of pure CO, weighed accurately to ±2 mg, is added  from a high-pressure tank.  Next, 100 g
       of pure air (accurately weighed) is added from a pressure tank, and  the concentration of CO is
       calculated from the sum of the respective weights added to the molecular weights of the two
30     gases. Not only the average "molecular weight" of the air, but also the requisite careful
       check of purity, is obtained by mass spectrometry and gas chromatography analyses of the air
       March 12, 1990                          5-5     DRAFT - DO NOT QUOTE OR CITE

-------
       and the CO. Lower-concentration primary standards are prepared by serial dilutions (not
       more than a factor of 100 for each step) by the same technique.
            The commercial suppliers of compressed gases are another source of air samples
       containing CO in the milligram per cubic meter or parts-per-million range. However, the
 5     nominal values for CO concentration supplied by the vendor should be verified by
       intercomparison with an SRM or other validated standard sample.  A three-way
       intercomparison has been  made among the NBS SRMs, commercial gas blends, and an
       extensive set of standard gas mixtures prepared by gravimetric blending at the Environmental
       Protection Agency (Paulsell, 1976).  Results of the comparison showed that commercial gas
10     blends are within  +2% of the true value represented by a primary standard.  Another study
       on commercial blends (Elwood, 1976) found poorer accuracy.  To achieve compatible results
       in sample analyses, different laboratories  should interchange and compare their respective
       working standards frequently.
            In making and using standards, many precautions are needed (Hughes, 1975): One
15     deserves special mention.  Large but unpredictable decreases in CO concentration occur
       within a few months in mixtures prepared in ordinary mild steel gas cylinders, as shown in
       Figure 5-1.  This  may be  due  to carbonyl formation or oxidation and CO to CO2. The
       difficulty can be avoided by the use of gas cylinders made of stainless steel or aluminum. A
       special treatment for aluminum, which includes enhancement of the aluminum oxide surface
20     layer, has been recommended  (Wechter,  1976).
            In addition to the set of SRMs for CO in  air, another set of SRMs is available from
       NBS for CO in nitrogen.  This second set covers concentrations from 10 to 957 ppm.

       5.2.3 Volumetric  Gas Dilution Methods
25          Standard samples of CO in air also  can be prepared by volumetric gas dilution
       techniques.  In a versatile system designed for  this purpose  (Hughes et al., 1973), air at a
       pressure of 10 to  100 psi  is first purified  and dried by passage through cartridges of charcoal
       and  silica  gel, then passed through a sintered metal filter into a flow control and flowmeter
       system. The CO  (or a mixture of CO in  air that is to be diluted further), also under pressure,
       is passed through  a similar flow control and  flowmeter system.
        March 12, 1990                          5-6     DRAFT - DO NOT QUOTE OR CITE

-------
I
o
a
cw
VO
                                                   CONCENTRATION OF CARBON MONOXIDE , ppm
O
o
n

-------
     Both gas streams are fed into a mixing chamber, which is designed to mix the gas
streams rapidly and completely before passage into the sampling manifold from which the
standard samples will be withdrawn.  From the air flow rate, FA, and the CO flow rate, Fco,
the concentration of CO in the sample, Cco, is readily calculated by the expression
                             Cco=     Fc°                                      (5-1)
                                    F
                                     co
For samples prepared by dilution of a more concentrated bulk mixture, the concentration is
given by
                                       F"
                                             (Q                                (5-2)
where Fb and Cb are the values of flow rate and concentration of CO, respectively, for the
bulk mixture.

5.2.4  Other Methods
     Permeation tubes have been used for preparing standard mixtures of such pollutant gases
as SO2 and NO2 (O'Keeffe and Ortman, 1966; Scaringelli et al., 1970). Permeation tubes are
not used routinely in the United States for making CO standard samples.  In the permeation
tube techniques, a sample of the pure gas under pressure is allowed to diffuse through a
calibrated partition at a defined rate into a diluent gas stream to give a standard sample of
known composition.
     Another possible way to liberate known amounts of CO into a diluent gas is by thermal
decomposition of nickel tetracarbonyl.  However, an attempt to use this as a gravimetric
calibration source showed that the relation between CO output and weight loss of the Ni(CO)4
is nonstoichiometric (Stedman et al.,  1976).
5.3  MEASUREMENT IN AMBIENT AIR
     Ambient CO monitoring is an expensive and time-consuming task, requiring skilled
personnel and sophisticated analytical equipment. This section discusses several important

March 12, 1990                           5-8     DRAFT - DO NOT QUOTE OR CITE

-------
aspects of the continuous and intermittent measurement of CO in the atmosphere, including
sampling techniques, sampling schedules, and recommended analytical methods for CO
measurement.

5.3.1  Sampling System Components
     Carbon monoxide monitoring requires a sample introduction system, an analyzer system,
and a data recording system, as illustrated in Figure 5-2 (Smith and Nelson, 1973). While
the "heart" of any air pollution monitoring system is the air pollution analyzer, Figure 5-2
shows that there is a considerable amount of supportive equipment necessary for continuous
air monitoring.
     A sample introduction system consists of a sampling probe, an intake manifold,  tubing,
and air movers. This system is needed to collect the air sample from the atmosphere  and to
transport it to the analyzer without altering the original concentration.  It also may be used to
introduce known gas concentrations in order to check periodically the reliability of the
analyzer output.  Construction materials for the sampling probe, intake manifold, and tubing
should be tested to demonstrate that the test atmosphere composition or concentration  is not
altered significantly.  It is recommended that sample introduction systems be fabricated from
borosilicate glass or PEP Teflon® (Code of Federal Regulations, 1977b)  if several pollutants
are to be monitored.  However, in monitoring for CO  only, it has been reported (Wohlers
et al., 1967) that no measurable pollutant losses were observed at the high (> 1 L/min)
sampling flow rates when sampling systems were constructed of tygon, polypropylene,
polyvinylchloride, aluminum, or stainless steel piping.  The sample introduction system
should be constructed so that it presents no pressure drop to the analyzer. At low flow and
low concentrations, such operation may require validation.
     The analyzer system consists of the analyzer itself and any sample preconditioning
components that may be necessary.  Sample preconditioning might require a moisture control
system to help minimize the false positive response of  the analyzer (e.g., the NDIR analyzer)
to water vapor,  and a particulate filter to help protect the analyzer from clogging and  possible
chemical interference due to particulate buildup in the  sample lines or analyzer inlet.  The
sample preconditioning system also may include a flow metering and flow control device to
control the sampling rate to the analyzer. As for the analyzer, there are  several analytical

March 12, 1990                           5-9       DRAFT - DO NOT QUOTE OR CITE

-------
ar
>—»
K)
 I
§
I
I
a
§
n
                  SAMPLE INTRODUCTION SYSTEM
                        SAMPLE INTAKE PORT
             BLOWER
  FIRST STAGE
  PRESSURE V_
  GAUGE    Y?)

CYLINDER
PRESSURE
VALVE
    SECOND STAGE
\   PRESSURE GAUGE
SECOND STAGE
PRESSURE VALVE
     ZERO GAS
                             SPAN GAS
                                                                                                           DATA RECORDING
                                                                                                               AND
                                                                                                           DISPLAY SYSTEM
                                                                    CARBON MONOXIDE ANALYZER
                 Figure 5-2.  Carbon monoxide monitoring system (Smith and Nelson, 1973).

-------
        methods for the continuous measurement of CO.  These are described in Section 5.3.4,
        Continuous Analysis.
             A data recording system is needed to record the output of the analyzer.  Data recording
        systems range from simple strip chart recorders to digital magnetic tape recorders to
 5      computerized telemetry systems that transfer data from remote stations to a central location
        via telephone lines or radio waves.

        5.3.2  Quality Assurance  Procedures for Sampling
             The accuracy and validity of data collected  from a CO monitoring system must be
10      ensured through a quality assurance program.  Such a program  consists of procedures for
        calibration, operational and preventive maintenance, data handling, and auditing; and the
        procedures are documented fully in a quality assurance program manual maintained by the
        monitoring organization.
             Calibration procedures consist of periodic multipoint primary calibration and secondary
15      calibration, both of which are prescribed to minimize systematic error. Primary calibration
        involves the introduction of test atmospheres of known concentration to an instrument in its
        normal mode of operation for the purpose of producing a calibration curve.
             A calibration curve is derived from the analyzer response  obtained by introducing
        several successive test atmospheres of different known concentrations.  One recommended
20      method for generating CO test atmospheres is to  use zero air (containing no CO) along with
        several known concentrations of CO in air or nitrogen contained in high-pressure gas
        cylinders and verified by NBS-certified SRMs wherever possible (Code of Federal
        Regulations, 1977a).  The number of standard gas mixtures (cylinders) necessary to establish
        a calibration curve depends on the nature of the analyzer output. A multipoint calibration at
25      five or six different CO concentrations covering the operating range of the analyzer is
        recommended by EPA (Code of Federal Regulations,  1977b; Federal Register, 1978).
        Alternatively,  the multipoint calibration is accomplished by diluting a known high-
        concentration CO standard gas with zero gas in a calibrated flow dilution  system.
             Primary calibrations should be performed when  the analyzer is first purchased and every
30      30 days thereafter (Smith and Nelson, 1973).  Primary calibration also is  recommended after
        the analyzer has had maintenance that could affect its  response characteristics or when results

        March 12, 1990                           5-11    DRAFT  - DO NOT QUOTE OR CITE

-------
        from auditing show that the desired performance standards are not being met (Smith and
        Nelson, 1973).
             Secondary calibration consists of a zero and upscale span of the analyzer.  This is
        recommended to be performed daily (Federal Register,  1978).  If the analyzer response
 5      differs by more than 2% from the certified concentrations, then the analyzer is adjusted
        accordingly. Complete records of secondary calibrations should be kept to aid in data
        reduction and for use in auditing.
             Operational and preventive maintenance procedures consist of operational checks to
        ensure proper operation of the analyzer and a preventive maintenance schedule necessary to
10      prevent unexpected analyzer failure and the associated loss of data (PEDCo Environmental
        Specialists, Inc., 1971).  Operational checks include checks of zero and span control settings,
        sample flow rate, gas cylinder pressures, sample cell pressure, shelter temperature, water
        vapor control, the paniculate filter, the sample introduction system, the recording system, and
        the strip chart record.  These checks may indicate the need for corrective/remedial action.
15      They usually are performed in conjunction with secondary calibrations.  In addition  to
        operational checks, a routine schedule of preventive maintenance should be developed.
        Maintenance requirements for the  analyzer usually are specified in the manufacturer's
        instrument manual.  Routine maintenance of supportive equipment (i.e., the sample
        introduction system and the data recording system) also is required.  This may include sample
20      line filter changes, water vapor control changes,  sample line cleaning, leak checks, and chart
        paper supply changes.
             Data handling procedures consist of data generation, reduction, validation, recording,
        and analysis and interpretation.  Data generation is the process of generating raw,
        unprocessed, and unvalidated observations as recorded on a strip  chart record. Data reduction
25      is the conversion, by use of calibration records, of raw  data to concentration units.  Data
        validation involves final screening of data before recording.  Then, questionable data
        "flagged" by the monitoring technician are reviewed with the aid of daily calibration and
        operation records to assess their validity.  Specific criteria for data selection and several
        instrument checks are available (Smith and Nelson, 1973).  Data recording involves recording
30      in a standard format for data storage, interchange of data with other agencies, and/or data
        analysis.  Data analysis and interpretation usually include a mathematical or statistical analysis

        March  12, 1990                          5-12    DRAFT - DO NOT QUOTE OR CITE

-------
        of air quality data and a subsequent effort to interpret results in terms of exposure patterns,
        meteorological conditions, characteristics of emission sources, and geographic and
        topographic conditions.
             Auditing procedures consist of several quality control checks and subsequent error
 5      analyses to estimate the accuracy and precision of air quality measurements. The quality
        control checks for CO include a data processing check, a control sample check, and a water
        vapor interference check, which should be performed by a qualified individual independent of
        the regular operator.  The error analysis is a statistical evaluation of the accuracy and
        precision of air quality data.  Guidelines have been published by EPA (Smith and Nelson,
10      1973) for calculating an overall bias and standard deviation of errors associated with data
        processing, measurement of control samples, and water vapor interference, from which the
        accuracy and precision of CO measurements can be determined. Since January 1, 1983, all
        state and local agencies submitting data to EPA must provide estimates of accuracy and
        precision of the CO measurements based on primary and secondary calibration records
15      (Federal Register, 1978).  The precision and accuracy  audit results through 1985 indicate that
        the 95% national probability limits for precision are +9% and the 95%  national probability
        limits for accuracy are within +1.5% for all audit levels from 3 to 8 ppm to 80 to 90 ppm.
        The results for CO are better than comparable results for the other pollutants with national air
        quality  standards (Rhodes and Evans, 1987).
20
        5.3.3  Sampling Schedules
             Carbon monoxide concentrations in the atmosphere exhibit large temporal variations due
        to changes in the time and rate that CO is emitted by different sources and due to changes in
        meteorological conditions that govern the amounts of transport and dilution that take place.
25      During  a one-year period an urban CO station may monitor hourly concentrations of CO
        ranging from 0 to as high as 50 mg/m3 (45 ppm).  The NAAQS for CO are based on the
        second highest one- and eight-hour average concentrations; violations represent extreme
        events when compared to the 8760 hours that constitute a year. In order to measure the
        highest  two values from the distribution of 8760 hourly values, the "best" sampling schedule
30      to employ is continuous monitoring 24 hours per day, 365 days per year.  Even so,
        continuous monitors rarely operate for long periods without data losses due to malfunctions,

        March 12, 1990                          5-13      DRAFT - DO NOT QUOTE OR CITE

-------
       upsets, and routine maintenance.  Data losses of 5 to 10% (438 to 876 hours per year) are not
       uncommon. Consequently, the data must be interpreted in terms of the "likelihood" that the
       NAAQS were attained or violated.  Statistical methods can be employed to interpret the
       results (Garbarz et al., 1977; Larsen,  1971).
 5          Compliance with one- and eight-hour NAAQS requires continuous monitoring.
       Statistically valid sampling could be performed on random or systematic schedules, however,
       if annual averages or relative concentration levels were of importance.  Most investigations of
       various sampling schedules have been conducted for paniculate air pollution data (Hunt,
       1972; Ott and Mage, 1975; Phinney and Newman, 1972), but the same schedules also could
10     be used for CO monitoring.  However, most instruments do not perform reliably in
       intermittent sampling.

       5.3.4  Continuous Analysis
       5.3.4.1 Nondispersive Infrared Photometry
15          Carbon monoxide has a characteristic infrared absorption near 4.6 urn:  The absorption
       of infrared radiation by the CO molecule therefore can be used to measure CO concentration
       in the presence  of other gases.  The NDIR method is based on this principle.
            Nondispersive infrared systems have several advantages.  They are not sensitive to flow
       rate, they require no wet chemicals, they are reasonably independent of ambient air
20     temperature changes, they are sensitive over wide  concentration ranges, and they have short
       response times.  Further, NDIR systems may be operated by nontechnical personnel. NDIR
       analyzers using  Luft-type detectors were widely used in the 1970s while GFC analyzers are
       most commonly used now in documenting compliance with ambient air standards.

25     NDIR Using Luft-Type Detectors
            The Luft-type detector is the primary distinguishing feature for the type of NDIR
       monitor that was widely used in the 1970s.  This type of analyzer contains a hot filament
       source of infrared radiation, a rotating sector (chopper), a sample cell, and reference cell, and
       a detector. The reference cell contains a non-infrared-absorbing gas,  while the sample cell is
30     continuously flushed with the sample atmosphere.  The detector consists of a two-
       compartment gas cell (both filled with CO under pressure) separated by a  diaphragm whose

       March 12, 1990                          5-14     DRAFT - DO NOT QUOTE OR CITE

-------
       movement causes a change of electrical capacitance in an external circuit and ultimately an
       amplified electrical signal suitable for input to a servo-type recorder.
            During analyzer operation a mechanical chopper alternately exposes the reference and
       sample cells to the infrared sources. At the frequency imposed by the chopper, infrared
 5     energy passes unattended through the reference cell to one compartment of the detector cell.
       Transmission through the sample cell is adjusted with no CO present so that the two beams
       are matched. Subsequently,  when sample is introduced into the sample cell, infrared energy
       is attenuated by CO absorption, causing an imbalance in the energy reaching the two
       compartments of the detector cell.  These unequal amounts of infrared energy differentially
10     heat the absorbing gas in the detector cell and the resulting pressure difference inside the cells
       causes movement of the diaphragm that forms their common wall.  A signal is generated at
       the chopping frequency with an amplitude related to the concentration of CO in the sample.
       This in turn produces the electrical signal previously discussed.
            Because water vapor is the principal interfering substance in determining CO by NDIR
15     techniques, a moisture control or compensation system is particularly important. Water vapor
       can be removed by absorption using in-line drying agents or by removal of condensate in a
       cooled inlet line.  Alternatively, the water vapor concentration can be measured independently
       and its contribution subtracted from the total  signal.

20     Gas-Filter Correlation Spectroscopy
            A GFC monitor (Burch et al., 1976) is  in essence a modern NDIR monitor.  It has all
       the advantages of an NDIR instrument and the additional advantages of smaller size,  no
       interference from CO2, and very small interference from water vapor.  It is not sensitive to
       flow rate, requires no wet chemicals, has a very fast response, and is relatively independent
25     of normal ambient temperature changes.
            A top view of the GFC monitor is presented  schematically in Figure 5-3A, showing the
       components of the optical path for CO detection. During operation, sample air is
       continuously pushed through the sample cell.  Radiation from the source is directed by optical
       transfer elements through the two main optical subsystems:  the rotating gas filter (designated
30     as correlation cell in Figure 5-3A) and the optical multipass  (sample) cell.  The beam exits the
       sample cell through interference filter FC,  which limits  the spectral passband to a few of the

       March 12, 1990                           5-15     DRAFT - DO NOT QUOTE OR CITE

-------
        strongest CO absorption lines in the 4.6-fj.m region.  Detection of the transmitted radiation
        occurs at the infrared detector, C.
             Although the passbound of filter FC is chosen to minimize interference from other
        gases, some residual H2O interference occurs. This residual interference is not significant at
 5      criteria pollutant levels, but can be corrected by independent measurement of H2O in the same
        cell.
             The gas correlation cell is constructed with two compartments (Figure 5-3B):  one
        compartment (gas cell 1) is filled with one-half atmosphere of CO, and the other compartment
        (gas cell 2) is filled with pure N2.  Radiation transmitted through cell 1 is completely
10      attenuated at spectral positions where CO absorbs strongly. The radiation  transmitted by cell
        2 is reduced by coating the exit window of the cell with a neutral attenuator. In this way, the
        amounts of radiation transmitted by the two cells are made approximately equal in the spectral
        passband that reached detector C through filter FC.
             In operation,  radiation passes alternately through the two cells as they are rotated by a
15      synchronous motor drive.  This establishes a  signal modulation frequency. Transmission to
        the detector is constant if no absorption by the ambient sample occurs.  If  CO is present in
        the sample, the radiation transmitted through cell  1 is not appreciably changed, while that
        through cell 2 is changed.  This imbalance is linearly related to CO concentration for small
        concentrations.  Other gas  species  absorb the radiation transmitted by cells 1 and 2 in
20      approximately equal amounts since their absorption structure does not correlate with that of
        CO.
             Superimposed on the entrance window of the cell is a typical light chopper pattern
        (Figure 5-3B) that creates a carrier frequency 12 times the signal modulation frequency, i.e.,
        a carrier frequency of 400  Hz. The detector output from the CO channel is fed to two phase-
25      sensitive amplifiers that separate the detector  response at the signal frequency from the
        detector response at the reference (carrier) frequency.  The signal due to CO is divided by the
        reference signal to substantially reduce many  of the causes of sensitivity change, such as
        accumulation of material on optical components and variation in detector sensitivity.

30
       March 12, 1990                           5-16     DRAFT - DO NOT QUOTE OR CITE

-------
      DETC
         SOURCE
                                             SAMPLE CELL
                                                      ELECTRONICS
                                                                                  Ms
        PLATED PATTERN
         CHOPPER
                                             CELL CONTAINING
                                                                NEUTRAL
                                                               ATTENUATOR
                             CELL CONTAINING
                                 CO
Figure 5-3. Schematic diagram of gas filter correlation (GFC) monitor for CO.  A:
Optical layout (M denotes mirror reflector; L denotes lens); B: Detail of correlation
cell.
Source:  Chancy and McClenny (1977).

March 12, 1990
5-17     DRAFT - DO NOT QUOTE OR CITE

-------
       5.3.4.2 Gas Chromatography - Flame lonization
            In this type of system, CO is separated from other trace gases by gas chromatography
       and catalytically converted to CH4 prior to detection.  A gas sampling valve, a back flush
       valve, a precolumn, a gas chromatographic column, a catalytic reactor, and an FID comprise
 5     the gas chromatography-flame ionization system.  In operation, measured volumes of air are
       delivered 4 to 12 times per hour to a hydrogen FID that measures the total hydrocarbon
       content (THC).  A portion of the same air sample, injected into a hydrogen carrier gas
       stream, is passed through the precolumn where it is separated from water, CO2,  and
       hydrocarbons other than CH4. Methane then is separated from CO on a second gas
10     chromatographic column.  The CH4, which is eluted first, is unchanged after passing through
       a catalytic reduction tube into the FID.  The CO eluted into a catalytic reduction tube is
       reduced to CH4 before passing through the FID (Porter and Volman, 1962).  Between
       analyses the precolumn is flushed out.  Nonmethane hydrocarbon concentrations also can be
       determined by subtracting the CH4 value from the total hydrocarbon (TH) value.
15          There are two possible modes of operation.  One of these is a complete chromatographic
       analysis showing the continuous output from the  detector for each sample injection. In the
       other, the system is programmed for both automatic zero and span settings to display selected
       elution peaks as bar graphs. The peak height is then  the measure of the concentration.  The
       first operation is referred to as the chromatographic or "spectro" mode and the second as the
20     barographic or "normal" mode.
            Because measuring CO entails only small increases in cost, instrument complexity, and
       analysis time, these instruments customarily are used to measure three pollutants:  CH4, THs,
       and CO.
            The instrumental sensitivity for each of these three components is 0.023 mg/m3
25     (0.02 ppm). The lowest full-scale range available is usually 2.3 mg/m3 (2 ppm) to 5.7 mg/m3
       (5 ppm), although at least one instrument has a 1.2 mg/m3 (1 ppm) range. Because of the
       complexity of these instruments, continuous maintenance by skilled technicians is required to
       minimize downtime.  This maintenance requirement may be considered a possible
       disadvantage of the system.  Depending on the frequency of analysis and the temporal
30     variability of CO, the representativeness over short averaging times may not be accurate
       (Chancy and McClenny,  1977).

       March 12,  1990                         5-18     DRAFT - DO NOT QUOTE OR CITE

-------
       5.3.4.3  Other Analyzers
       Controlled-Potential Electrochemical Analysis
             Carbon monoxide is measured by means of the current produced in aqueous solution by
       its electro-oxidation by an electro-catalytically active noble metal.  The concentration of CO
 5     reaching the electrode is controlled by its rate of diffusion through a membrane. This is
       dependent on its concentration in the sampled atmosphere (Bay et al.,  1974; 1972).  Proper
       selection of both the membrane and such cell characteristics as the nature of the electrodes,
       the electrode potential, and the solution make the technique selective for various pollutants.
       A similar technique has been reported by Yamate and Inoue (1973).
10           The generated current is linearly proportional to the CO concentration from 0 to
       115 mg/m3 (0 to 100 ppm).  A sensitivity of 1.2 mg/m3 (1 ppm) and a 10-s response time (to
       reach 90% of full scale) are claimed for currently available commercial instruments.
             Acetylene and ethylene are the chief interfering substances:  one part acetylene responds
       as 11 parts CO, and one part ethylene as 0.25 part CO.  For hydrogen, ammonia,  hydrogen
15     sulfide, NO, NO2, SO2, natural gas, and gasoline vapor, interference is less than 0.03 part CO
       per one part interfering substance.

       Galvanic Analyzer
             Galvanic cells employed in the manner described by Hersch (1966; 1964) can be used to
20     measure atmospheric CO continuously.  When an air stream containing CO is passed into a
       chamber packed with I2O5 and heated to 150°C, the following  reaction takes place:

                                    SCO + I2O5	* 5CO2 + I2                        (5-3)

25     The liberated iodine is absorbed by an electrolyte and transferred to the cathode of a galvanic
       cell.  At the cathode, the iodine is reduced and the resulting current is measured by a
       galvanometer. Instruments with this detection system have been used  successfully to measure
       CO levels in traffic along freeways (Haagen-Smit, 1966).
             Mercaptans, hydrogen sulfide, hydrogen, olefins, acetylenes, and water vapor interfere.
30     Water may be removed by sampling through a drying column; hydrogen, hydrogen sulfide,
       March 12, 1990                           5-19    DRAFT - DO NOT QUOTE OR CITE

-------
       acetylene, and olefin interferences can be minimized by sampling through an absorption tube
       containing mercuric sulfate on silica gel.

       Coulometric Analyzer
 5          A coulometric method employing a modified Hersch-type cell has been used for
       continuous measurement of CO in ambient air (Dubois et al., 1966). The reaction of I2O3
       with CO liberates iodine, which then is passed into a Ditte cell, and the current generated is
       measured by an electrometer-recorder combination.  Interferences are the same as those
       discussed above for the galvanic analyzer.
10          This technique may be used for a minimum detectable concentration of 1.2 mg/m3
       (1 ppm) with good reproducibility and accuracy if flow rates and temperatures are controlled
       well. This method requires careful column preparation and use of filters to remove
       interferences.  Its relatively slow response time may be an  added disadvantage in some work.

15     Mercury Replacement
            Mercury vapor formed by the reduction of mercuric oxide (HgO) by CO is detected
       photometrically by its absorption of ultraviolet light at 253.7 nm.  The reaction involved is as
       follows:
20                                             (2io° c>
                                   CO 4- HgO	> CO2 + Hg                       (5-4)
       This is potentially a much more sensitive method than infrared absorption because the
25     oscillator strength of Hg at 253.7 nm is 2000 times greater than that of CO at 4.6 jwm.
       Hydrogen and hydrocarbons also reduce HgO to Hg, and there is some thermal decomposition
       of the oxide. Operation of the detector at constant temperature results in a regular
       background concentration of Hg from thermal decomposition. The instrument is portable and
       can analyze CO concentrations of 0.025 to 12 mg/m3 (0.020 to 10.0 ppm). Changes of
30     0.002 mg/m3 (0.002 ppm) are detectable.  For this reason, this instrument has been used to
       determine global CO levels.  McCullough et al. (1947) and Beckman et al. (1948)
       recommended a detector operating temperature of 175° C to minimize hydrogen interference.
       A commercial instrument employing these principles was made and used during the middle
       March 12, 1990                          5-20     DRAFT - DO NOT QUOTE OR CITE

-------
        1950s (Mueller, 1954).  The technique recently has been used for measuring background CO
        concentrations.  Robbins et al. (1968) have described an instrument in which the HgO
        chamber is operated at 210°C, and the amount of hydrogen interference is assessed by
        periodic introduction of a tube of silver oxide into the intake air stream. At room temperature
 5      silver oxide quantitatively oxidizes CO, but not hydrogen. Thus,  the baseline hydrogen
        concentration can be determined.  Additional minor improvements are discussed by Seiler and
        Junge (1970), who gave the detection limit for CO as 0.003 mg/m3 (0.003 ppm).
             More recently,  Palanos (1972) described a less sensitive model of this instrument
        intended for use in urban monitoring. It has a range of 0 to 23 mg/m3 (20 ppm), a sensitivity
10      of about 0.58 mg/m3 (0.5 ppm), and a span and zero drift of less  than 2%  per day. As in
        other similar instruments, specificity is achieved by removal of the potentially interfering
        substances (which is  less than 10%) other than hydrogen.
             With all of these instruments, a constant geophysical hydrogen concentration is assumed.
        In unpolluted atmospheres, the hydrogen concentration is roughly 46.5 /xg/m3 (0.56 ppm).
15      However, the automobile not only is a source of CO but also of hydrogen.  Therefore, if this
        technique is used in polluted areas, it will be necessary to measure the hydrogen concentration
        frequently.

        Dual Isotope Fluorescence
20           This instrumental method utilizes the slight difference in the infrared spectra of isotopes.
        The sample is alternately illuminated with the characteristic infrared wavelengths of carbon
        monoxide-16 (12C16O) and carbon monoxide-18 (12C18O).  The  CO  in the sample that has the
        normal isotope ratio, nearly 100%  12C16O, absorbs only the 12C16O  wavelengths.  Therefore,
        there is a cyclic variation in the intensity of the fluorescent light that is dependent on the
25      12C16O content of the sample (Link et al., 1971; McClatchie, 1972; McClatchie et al., 1972).
             Full-scale ranges of 0 to 23 mg/m3 (0 to 20 ppm) and up to  0 to 230 mg/m3 (0 to
        200 ppm) with a claimed sensitivity of 0.23 mg/m3 (0.2 ppm) are available in this instrument.
        The response time (to reach 90% of full scale) is 25 s, but a 1-s response time also is
        available.  An advantage of this technique is that it minimizes the effects of interfering
30      substances.
        March 12, 1990                           5-21      DRAFT - DO NOT QUOTE OR CITE

-------
        Catalytic Combustion - Thermal Detection
             Determination of CO by this method is based on measuring the temperature rise
        resulting from catalytic oxidation of the CO in the sample air.
             The sample air is pumped first into a furnace that brings it to a preset, regulated
 5      temperature and then over the catalyst bed in the furnace.  A thermopile assembly measures
        the temperature difference between the air leaving the catalyst bed and the air entering it.
        The output of the thermopile, which is calibrated  with known concentrations of CO in air, is
        read on a strip chart recorder as parts of CO per million parts of air. The sensitivity is about
        1.2 mg/m3 (1 ppm). Most HCs are oxidized by the same catalyst, and will interfere unless
10      removed.  These systems are widely used in enclosed spaces; their applicability for ambient
        air monitoring is limited because they function best at high CO concentrations.

        Second-Derivative Spectrometry
             A second-derivative spectrometer processes the transmission-versus-wavelength function
15      of an ordinary spectrometer to produce an output  signal proportional to the second derivative
        of this function. Ultraviolet light of continuous wavelength is collected and focused onto an
        oscillating entrance slit of a grating spectrometer.  When the grating orientation is changed
        slowly, a slowly scanning center wavelength with sinusoidal wavelength modulation is  created
        in the existing light by the oscillating  entrance slit.  This radiation passes through a gas
20      sample and is detected with a photomultiplier tube.  The signal then is electronically
        processed to produce a second-derivative spectrum (Lawrence Berkeley Laboratory, 1973).
        This method has the advantage that it  can be used to measure other pollutants as well as CO.

        Fourier-Transform Spectroscopy
25           Fourier-transform spectroscopy is an extremely powerful infrared spectroscopic
        technique (Bell, 1972) that has developed in the past 20 years and has been applied in the last
        10 years to air pollution measurement problems (Hanst et al., 1973; Lawrence Berkeley
        Laboratory, 1973). The advantage of this technique over a standard grating or prism
        spectrometer is that it has a higher throughput, which means that the available energy is used
30      more effectively and that a much higher resolving power is obtainable.  In air pollution
        measurements individual absorption lines can be resolved.

        March 12, 1990                           5-22     DRAFT - DO NOT QUOTE OR CITE

-------
            A special advantage for air pollution measurements is that all the data required to
       reconstruct the entire absorption spectrum are acquired at the same time.  The spectrum as a
       function of wavelength is generated by a built-in computer.  This means that several gases can
       be measured simultaneously.  Several commercial instruments now are available with
 5     resolutions of 0.06/cm or better.  These instruments are capable of clearly defining the
       spectrum of any gaseous pollutant, including CO and currently are being used for special air
       pollution studies.

       5.3.5  Intermittent Analysis
10          Intermittent samples may be collected in the field and later analyzed in the laboratory by
       the continuous analyzing techniques described above. Sample containers may be rigid (glass
       cylinders or stainless steel tanks) or they may be nonrigid (plastic bags). Because of location
       or cost, intermittent  sampling at times may be the only practical method for air monitoring.
       Samples can be taken over a few minutes or accumulated intermittently to obtain, after
15     analysis, either "spot" or "integrated" results. Additional techniques for analyzing
       intermittent samples  are described below.

       5.3.5.1  Colorimetric Analysis
       Colored Silver Sol Method
20          Carbon monoxide reacts in an alkaline solution with the silver salt of p-sulfamoyl-
       benzoate to  form a colored silver sol. Concentrations of 12 to 23,000 mg/m3 (10 to
       20,000 ppm) CO may be measured by this method (Ciuhandu, 1958, 1957, 1955; Ciuhandu
       and Krall, 1960; Ciuhandu et al., 1965; Levaggi and Feldstein,  1964). The method has been
       modified to determine CO concentrations in incinerator effluents. Samples are collected in an
25     evacuated flask and reacted.  The absorbance of the resulting colloidal solution is measured
       spectrophotometrically. Acetylene and formaldehyde interfere, but can be removed by
       passing the sample through mercuric sulfate on silica gel. Carbon monoxide concentrations of
       5.8 to 20,700 mg/m3 (5 to 18,000 ppm) may be measured with an accuracy of 90 to 100% of
       the true value.
30
       March 12, 1990                           5-23     DRAFT - DO NOT QUOTE OR CITE

-------
       National Bureau of Standards Colorimetric Indicating Gel
            An NBS colorimetric-indicating gel (incorporating palladium and molybdenum salts) has
       been devised to measure CO in the laboratory and in the field (Shepherd, 1947; Shepherd
       et al., 1955). The laboratory method involves colorimetric comparison with freshly prepared
 5     indicating gels exposed to known concentrations of CO. The method has an accuracy range
       of 5 to 10% of the amount of CO involved, and the minimum detectable concentration is
       1.2 mg/m3 (1 ppm). This technique requires relatively simple and inexpensive equipment;
       however, oxidizing and reducing gases interfere, and the preparation of the indicator tube is a
       tedious and time-consuming task.
10
       Length-of-Stain Indicator Tube
            An indicator tube that uses potassium palladosulfite is a commonly employed manual
       method (Silverman and Gardner, 1965). Carbon monoxide reacts with the contents of the
       tube and produces a discoloration.
15          The length of discoloration is an exponential function of the CO concentration.  This
       method and other indicator tube manual methods are estimated to be accurate to within ±25%
       of the amount present, particularly at CO concentrations of about 115 mg/m3 (100 ppm).
       Such indicator tube manual methods have been used frequently in air pollution studies.
       Ramsey (1966) used the technique to measure CO at traffic intersections, and Brice and
20     Roesler (1966) estimated CO concentrations with an accuracy of ± 15% by means of color-
       shade detector tubes.
            Colorimetric  techniques and length-of-stain discoloration methods are recommended for
       use only when other physicochemical monitoring systems are not available.  They may be
       used in the field for gross mapping where accuracy is not required and may possibly be of
25     great value during  emergencies.

       Frontal Analysis
            Air is passed over an adsorbent until equilibrium is established between the
       concentration of CO in the air and the concentration of CO on the adsorbent.  The CO then is
30     eluted with hydrogen, reduced  to CH4 on a nickel catalyst at 250°C, and determined by flame
       ionization as CH4.

       March 12, 1990                          5-24     DRAFT - DO NOT QUOTE OR CITE

-------
            Concentrations of CO as low as 0.12 mg/m3 (0.10 ppm) can be measured.  This method
        does not give instantaneous concentrations, but does give averages over a six-minute or longer
        sampling period (Dubois and Monkman, 1972; 1970).
        5.4 MEASUREMENT USING PERSONAL MONITORS
            Until the 1960s, most of the data available on ambient CO concentrations came from
        fixed monitoring stations operated routinely in urban areas. The accepted measurement
        technique was by NDIR spectrometry, but the instruments were large and cumbersome, often
10      requiring vibration-free, air-conditioned enclosures.  Without a portable, convenient monitor
        for CO, it was extremely difficult to measure CO concentrations accurately in the
        microenvironments that people usually visited.  In the late 1960s, studies were initiated to
        investigate the CO concentrations within vehicles (Brice and Roesler, 1966; Lynn et al.,
        1967). In 1971, an investigator walked on congested downtown streets alongside pedestrians
15      to measure their exposures (Ott, 1971).  With a portable pump, the investigator filled
        sampling bags in various locations, then transported them to the laboratory where the contents
        were analyzed by NDIR spectrometry.
            In the early 1970s, portable electrochemical monitors about the size of a shoe box
        became available. Using the Ecolyzer monitor, CO concentrations were measured in traffic
20      in Boston, MA. (Cortese and Spengler, 1976).  In the late 1970s, smaller personal monitors
        using electrochemical sensing systems became available and were deployed in specialized field
        surveys involving a few people (Jabara  et al., 1980).
            As CO monitors continued to evolve, they were used in studies of indoor
        microenvironments.  Many of the microenvironmental CO data on indoor concentrations were
25      collected as an integral part of multipollutant indoor health or dosage studies in homes,
        offices, or rooms (Berglund et al.,  1982; Hoffman et al., 1984; Hugod, 1984), or as more
        narrowly focused multipollutant exposure field  studies in homes (Quackenboss et al., 1984;
        Koontz and Nagda, 1984; Traynor  et al., 1984) and in buildings (Konopinski, 1984;
        Malaspinaetal., 1984;  Clarkson, 1984).
30          Although the CO personal monitors evolved rapidly,  they were not used in large-scale
        field surveys of indoor microenvironments until the early 1980s.  Personal monitors have

        March  12,  1990                          5-25     DRAFT - DO NOT QUOTE OR CITE

-------
        been used in studies of CO concentrations in sustained-use vehicles (Ziskind et al.,  1981) and
        in passenger compartments of vehicles traveling on highways (Ott and Willits, 1981;
        Flachsbart and Ah Yo, 1986).
             Ultimately, small personal exposure monitors were developed that could measure CO
  5     concentrations continuously over time and store the readings automatically on internal digital
        memories (Ott et al., 1986).  These small personal exposure monitors (PEMs) made possible
        the large-scale CO human exposure field studies in Denver, CO, and Washington, DC, in the
        winter of 1982-1983 (Akland et al., 1985).  The PEM employed in these studies uses an
        aqueous solid polymer ion exchange material as the electrolyte in which CO is converted to
 10     CO2 by an electrochemical reaction at a noble metal electrode,  thereby generating an electrical
        current.  The signal (current) is proportional to the quantity of CO present in the gas stream,
        and the continuous electrical signal is recorded in internal memory.  A small pump operates
        continuously to send air into the sensing cell, and chemical filters in the intake stream remove
        interfering chemicals, such as ethanol.  The pump operates on batteries for up to 40 h with a
 15     precision of 2 ppm, with zero and span checks required before and after field service. Other
        studies have employed the CO detector and combined it with small computers such as the HP-
        41CV to enhance the utility of the detector  for studying factors that affect CO concentration
        variability (Fitz-Simons and Sauls,  1984).  These  monitors proved effective in generating
        24-h CO exposure profiles on more than 1600 persons. By breaking up the profiles into the
20     microenvironments visited by these people, it was possible to develop CO concentration
        readings on more than 40 indoor and in-transit microenvironments (see Chapter 8).
             Detector tubes also can be used in studies where high concentrations occur (above
        5 mg/m3) or long exposure times are possible and  only cumulative exposures, are required.
        Air is drawn  through specifically manufactured tubes containing an absorbent impregnated
25     with a chemical reagent that changes color if CO is present (Jacobs,  1949).  The length of the
        stain produced in the tube after exposure is  read on a chart to give the concentration of CO.
        Unfortunately, interferences also may  produce color changes, unless additional precautions are
        taken to filter out particles and to absorb interfering gases such as oxides of nitrogen, SO2,
        HCs, ammonia, hydrogen sulfide, and water vapor.  Techniques such as the detector tube,
30     may have the greatest utility to the researcher by providing inexpensive approximate value for
       March 12, 1990                           5-26    DRAFT - DO NOT QUOTE OR CITE

-------
         screening purposes, which would require confirmation found about some predetermined
         "action" level.
        5.5  REFERENCES
 5


        Acton, L. L.; Griggs, M.; Hall, G. D.; Ludwig, C. B.; Malkmus, W.; Hesketh, W. D.; Reichle, H. (1973)
              Remote measurement of carbon monoxide by a gas filter correlation instrument. AIAA J.  11: 899-900.
10
        Adams, E. G.; Simmons, N. T. (1951) The determination of carbon monoxide by means of iodine pentoxide. J.
              Appl. Chem. l(suppl. 1): S20-S40.

        Akland, G. G. (1972) Design of sampling schedules. J. Air Pollut. Control Assoc. 22: 264-266.
15
        Akland, G. G.; Hartwell, T. D.; Johnson, T. R.; Whitmore, R. W. (1985) Measuring human exposure to carbon
              monoxide in Washington, D.C., and Denver, Colorado, during the winter of 1982-1983. Environ. Sci.
              Technol. 19: 911-918.

20     Allen, T. H.; Root, W. S. (1955) An improved palladium chloride method for the determination of carbon
              monoxide in blood. J. Biol. Chem. 216: 319-323.

        American Industrial Hygiene Association. (1972) Intersociety Committee methods for ambient air sampling and
              analysis: report II. Am. Ind. Hyg. Assoc. J. 33: 353-359.
25
        Bartle, E. R.; Hall, G. (1977) Airborne HC1 - CO sensing system: final report. La Jolla, CA: Science
              Applications, Inc.; report no. SAI-76-717-LJ. Available from: NTIS, Springfield, VA; N77-21733.

        Bay, H. W.; Blurton, K. F.; Lieb, H. C.; Oswin, H. G.  (1972) Electrochemical measurement of carbon
30           monoxide. Am. Lab. 4: 57-58, 60-61.

        Bay, H. W.; Blurton, K. F.; Sedlak, J. M.; Valentine, A. M. (1974) Electrochemical technique for the
              measurement of carbon monoxide. Anal. Chem. 46: 1837-1839.

35     Beckman, A. O.; McCullough, J. D.; Crane,  R. A. (1948) Microdetermination of carbon monoxide in air: a
              portable instrument. Anal. Chem. 20: 674-677.

        Bell, D. R.; Reiszner, K. D.; West, P. W. (1975) A permeation method for the determination of average
              concentrations of carbon monoxide in the atmosphere. Anal. Chim. Acta 77: 245-254.
40
        Bell, R. J. (1972) Introductory Fourier transform spectroscopy. New York, NY: Academic Press, Inc.

        Berglund, B.; Johansson, L; Lindvall, T. (1982) A longitudinal study of air contaminants in a newly built
              preschool. Environ. Int. 8: 111-115.
45
        Bergman, I.; Coleman, J. E.; Evans, D. (1975) A simple gas chromatograph with an electrochemical detector for
              the measurement of hydrogen and carbon monoxide in the parts per million range, applied to exhaled air.
              Chromatographia 8: 581-583.

50     Brice, R. M.; Roesler, J. F. (1966) The exposure to carbon monoxide of occupants of vehicles moving in heavy
              traffic. J. Air Pollut. Control Assoc. 16: 597-600.
        March 12, 1990                              5-27     DRAFT - DO NOT QUOTE OR CITE

-------
        Bruner, F.; Ciccioli, P.; Rastelli, R. (1973) The determination of carbon monoxide in air in the parts per billion
               range by means of a helium detector. J. Chromatogr. 77: 125-129.

        Burch, D.  E.; Gates, F. J.; Pembrook, J. D. (1976) Ambient carbon monoxide monitor. Research Triangle Park,
  5            NC: U. S. Environmental Protection Agency, Environmental Sciences Research Laboratory; EPA report
               no. EPA-600/2-76-210. Available from: NTIS, Springfield,  VA; PB-259577.

        Burch, D.  E.; Gryvnak, D. A. (1974) Cross-stack measurement of pollutant concentrations using gas-cell
               correlation spectroscopy. In: Stevens, R. K.; Herget, W. F., eds. Analytical methods applied  to air
10            pollution measurements. Ann Arbor, MI: Ann Arbor Science Publishers, Inc.; pp. 193-231.

        Chaney, L. W.; McClenny, W. A. (1977) Unique ambient carbon monoxide monitor based on gas filter
               correlation: performance and application. Environ.  Sci. Technol. 11: 1186-1190.

15     Chaney, L. W.; Rickel, D. G.; Russwurm, G. M.; McClenny, W. A. (1979) Long-path laser monitor of carbon
               monoxide: system improvements. Appl. Opt.  18: 3004-3009.

        Ciuhandu,  Gh. (1955) O noua metoda de determinare a oxidului de carbon in aer [New method for the
               determination of carbon monoxide in air]. Acad. Repub. Pop. Rom. Baza Cercet. Stiint. Timisoara Stud.
20            Cercet. Stiint. Ser. 1:  133-142.

        Ciuhandu,  G. (1957) Photometrische Bestimmung von Kohlenmonoxyd in der Luft [Photometric determination of
               carbon monoxide in air]. Fresenius' Z. Anal.  Chem. 155:  321-327.

25     Ciuhandu,  G. (1958) Mikromethode fuer die photometrische Kohlenmonoxydbestimmung in der Luft
               [Micromethod for the photometric determination of carbon monoxide in air]. Fresenius' Z. Anal. Chem.
               161: 345-348.

        Ciuhandu,  G.; Krall, G. (1960) Photometrische Bestimmung von Kohlenmonoxydspuren in Wasserstoff
30            [Photometric determination of traces of carbon monoxide in hydrogen]. Fresenius' Z. Anal. Chem.
               172: 81-87.

        Ciuhandu,  G.; Rusu, V.; Diaconovici, M. (1965) Bestimmung von Kohlenmonoxidspuren in Sauerstoff
               [Determination of traces of carbon monoxide in oxygen]. Fresenius' Z. Anal. Chem. 208: 81-86.
35
        Clarkson, M. (1984) Indoor air quality as a part of total building performance. In: Berglund, B.; Lindvall, T.;
               Sundell, J., eds. Indoor air: proceedings of the 3rd international conference on indoor air quality and
               climate, v. 5, buildings, ventilation and thermal climate; August; Stockholm, Sweden. Stockholm,
               Sweden: Swedish Council for Building Research; pp. 493-498. Available from: NTIS, Springfield, VA;
40            PB85-104222.

        Code of Federal Regulations. (1977a) Ambient air monitoring reference and equivalent methods. C. F. R.
               40: 812-839.

45     Code of Federal Regulations. (1977b) National primary and secondary ambient air quality standards. C.  F. R.
               40(50) (July 1): 3-33.

        Code of Federal Regulations. (1988) List of designated reference and equivalent methods. C. F. R. 53: 812-839.

50     Commins,  B. T.; Berlin, A.; Langevin,  M.; Peal, J. A. (1977)  Intercomparison of measurement of
               carboxyhaemoglobin in different European laboratories and establishment of the methodology for the
               assessment of COHb levels in exposed populations. Luxembourg, Sweden: Commission of the European
               Communities, Health and Safety Directorate; doc. V/F/1315/77e.
         March  12,  1990                              5-28      DRAFT - DO NOT QUOTE OR CITE

-------
        Cormack, R. S. (1972) Eliminating two sources of error in the Lloyd-Haldane apparatus. Respir. Physiol.
               14: 382-390.

        Cortese, A. D.; Spengler, J. D. (1976) Ability of fixed monitoring stations to represent personal carbon
  5            monoxide exposure. J. Air Pollut. Control Assoc.  26: 1144-1150.

        Dagnall, R. M.; Johnson, D. J.; West, T. S. (1973) A method for the determination of carbon monoxide, carbon
               dioxide, nitrous oxide and sulphur dioxide in air by gas chromatography using an emissive helium plasma
               detector.  Spectrosc. Lett. 6: 87-95.
 10
        Dailey, W. V.; Fertig, G. H. (1977) A novel NDIR analyzer for NO, SO2 and CO analysis. Anal. Instrum.
               15: 79-82.

        Dempsey, R. M.; LaConti, A. B.; Nolan, M. E.; Torkildsen, R. A.; Schnakenberg, G.; Chilton, E. (1975)
 15            Development of fuel cell CO detection instruments for use in a mine atmosphere. Washington, DC: U.S.
               Department of the Interior, Bureau of Mines; report no. 77-76. Available from:  NTIS, Springfield, VA;
               PB-254823.

        Driscoll, J. N.; Berger, A. W. (1971) Improved chemical methods for sampling and analysis of gaseous
20            pollutants from the combustion of fossil fuels. Volume III. Carbon monoxide. Cincinnati, OH:  U.S.
               Environmental Protection Agency, Office of Air Programs; report no.  APTD-1109. Available from:
               NTIS,  Springfield, VA; PB-209269.

        Dubois, L.; Monkman, J. L. (1970) L'emploi de Panalyse frontale pour 1'echantillonage et le  dosage de 1'oxyde
25            de carbone dans 1'air [Sampling by frontal analysis and determination of carbon monoxide]. Mikrochim.
               Acta27: 313-320.

        Dubois, L.; Monkman, J. L. (1972) Continuous determination of carbon monoxide by frontal  analysis. Anal.
               Chem.  44: 74-76.
30
        Dubois, L.; Zdrojewski,  A.; Monkman, J.  L. (1966) The analysis of carbon monoxide  in urban air at the ppm
               level, and the normal carbon monoxide value. J. Air Pollut.  Control Assoc.  16:  135-139.

        Elwood, J. H.  (1976) A calibration procedural system and facility for gas turbine engine exhaust emission
35            analysis. In: Calibration in air monitoring, a symposium; August 1975; Boulder, CO.  Philadelphia, PA:
               American Society for Testing and Materials;  ASTM special technical publication no. 598; pp. 255-272.

        Federal Register. (1978) Air quality surveillance and data reporting:  proposed  regulatory revisions. F. R.
               (August 7) 43: 34892-34934.
40
        Feldstein, M. (1965) The colorimetric determination of blood and breath carbon monoxide. J.  Forensic Sci.
               10:  35-42.

        Feldstein, M. (1967) Methods for the determination of carbon monoxide. Prog. Chem.  Toxicol. 3: 99-119.
45
        Fitz-Simons, T.; Sauls, H. B. (1984) Using the HP-41CV  calculator as a data acquisition system for personal
               carbon  monoxide exposure monitors.  J. Air Pollut. Control Assoc. 34: 954-956.

        Flachsbart, P. G.; Ah Yo, C. (1986) Test of a theoretical commuter exposure model to  vehicle exhaust in traffic.
50            Presented at: 79th annual meeting of the Air Pollution Control Association; June; Minneapolis, MN.
               Pittsburgh, PA: Air Pollution Control Association; paper no.  86-79.4.

        Garbarz, J.-J.;  Sperling, R. B.; Peache, M. A.  (1977) Time-integrated urban carbon monoxide measurements.
               San Francisco, CA: Environmental  Measurements,  Inc.


         March 12,  1990                              5-29      DRAFT - DO  NOT QUOTE OR CITE

-------
        Golden, B. M.; Yeung, E. S. (1975) Analytical lines for long-path infrared absorption spectrometry of air
              pollutants. Anal. Chem. 47: 2132-2135.

        Goldstein, G. M. (1977) [Letter to Mr. Jack Thompson concerning carboxyhemoglobin measurements]. Research
 5            Triangle Park, NC: U.S. Environmental Protection Agency; September 23, 1977.

        Goldstein, S. A.; D'Silva, A. P.; Fassel, V. A. (1974) X-ray excited optical fluorescence of gaseous atmospheric
              pollutants: analytical feasibility study. Radial. Res. 59:  422-437.

10      Goldstein, H. W.; Bortner, M. H.; Grenda, R. N.; Dick, R.; Barringer, A. R. (1976) Correlation interferometric
              measurement of carbon monoxide and methane from the Canada Centre for Remote Sensing Falcon
              Fan-Jet Aircraft. Can. J. Remote Sens. 2:  30-41.

        Goodman, P. (1972) Measurement of automobile exhaust pollutant concentrations by use of solid kryptonates.
15            Highw. Res. Rec. (412): 9-12.

        Gryvnak, D. A.; Burch, D. E. (1976a) Monitoring of pollutant gases in aircraft exhausts by gas- filter correlation
              methods. In: Proceedings of the AIAA 14th aerospace sciences meeting; January; Washington, DC.
              New York, NY: American Institute of Aeronautics and  Astronautics; AIAA paper no. 76-110.
20
        Gryvnak, D. A.; Burch, D. E. (1976b) Monitoring NO and CO in aircraft jet exhaust by a gas-filter correlation
              technique.  Wright-Patterson AFB,  OH: U. S. Air Force, Aero-propulsion Laboratory; report
              no. AFAPL-TR-75-101. Available from: NTIS, Springfield, VA; AD-A022353.

25      Haagen-Smit, A. J. (1966) Carbon monoxide levels in city driving. Arch. Environ. Health  12: 548-551.

        Hanst, P. L.; Lefohn, A. S.; Gay, B. W., Jr. (1973) Detection of atmospheric pollutants at parts-per-billion
              levels by infrared spectroscopy. Appl. Spectrosc. 27: 188-198.

30      Harrison, N. (1975) A review of techniques for the measurement of carbon monoxide in the atmosphere. Ann.
              Occup.  Hyg. 18: 37-44.

        Herget, W. F.; Jahnke, J. A.; Burch, D. E.; Gryvnak, D. A. (1976) Infrared gas-filter correlation instrument for
              in situ measurement of gaseous pollutant concentrations. Appl. Opt. 15: 1222-1228.
35
        Hersch, P.  (1964)  Galvanic analysis. In: Reilley, C. N., ed. Advances in analytical chemistry and
              instrumentation, v. 3. New York, NY: Interscience Publishers;  pp. 183-249.

        Hersch, P.  A.  (1966) Method and apparatus for measuring the carbon monoxide content of a gas stream.
40            Washington, DC: U. S. Patent Office; patent no. 3,258,411; June 28, 1966.

        Hoffmann, D.;  Brunnemann, K. D.; Adams, J. D.; Haley, N. J. (1984) Indoor air pollution by tobacco smoke:
              model studies on the uptake by nonsmokers. In: Berglund, B.; Lindvall, T.; Sundell, J., eds. Indoor  air:
              proceedings of the 3rd international conference on indoor air quality and climate, v. 2, radon, passive
45            smoking, particulates and housing epidemiology; August; Stockholm, Sweden.  Stockholm, Sweden:
              Swedish Council for Building Research; pp. 313-318.

        Houben, W. P. (1976) Continuous monitoring of carbon monoxide in the atmosphere by an improved NDIR
              method. In: APCA specialty conference on air pollution measurement accuracy as it relates to regulation
50            compliance; October 1975; New Orleans,  LA. Pittsburgh, PA: Air Pollution Control Association;
              pp. 55-64.

        Hrubesh, L. W. (1973) Microwave rotational spectroscopy: a technique for specific pollutant monitoring. Radio
              Sci. 8:  167-175.


         March 12,  1990                              5-30      DRAFT - DO NOT QUOTE OR CITE

-------
        Hughes, E. E. (1975) Development of standard reference materials for air quality measurement. ISA Trans.
               14: 281-291.

        Hughes, E. E. (1976) Role of the National Bureau of Standards in calibration problems associated with air
 5             pollution measurements. In: Calibration in air monitoring, a symposium; August 1975; Boulder, CO.
               Philadelphia, PA: American Society for Testing and Materials; ASTM special technical publication
               no. 598; pp. 223-231.

        Hughes, E. E.; Dorko, W.  D.; Scheide, E. P.; Hall, L. C.; Beilby, A. L.; Taylor, J. K. (1973) Gas generation
10             systems for the evaluation of gas detecting devices. Washington, DC: U. S. Department of Commerce,
               National Bureau of Standards; report no. NBSIR 73-292.

        Hugod, C. (1984) Passive smoking - a source of indoor air pollution. In: Berglund, B.; Lindvall, T.; Sundell, J.,
               eds. Indoor air: proceedings of the 3rd international conference on indoor air quality and climate, v. 2,
15             radon, passive smoking, particulates and housing  epidemiology; August; Stockholm, Sweden. Stockholm,
               Sweden: Swedish Council for Building Research;  pp. 319-325.

        Hunt, W. F., Jr. (1972) The precision associated with the sampling frequency of log-normally distributed air
               pollutant measurements. J. Air Pollut.  Control Assoc. 22: 687-691.
20
        Jabara,  J. W.; Beaulieu, H. J.; Buchan, R. M.; Keefe,  T. J. (1980) Carbon monoxide: dosimetry in occupational
               exposures in Denver,  Colorado. Arch. Environ. Health 35: 198-204.

        Jacobs,  M. B. (1949) The analytical chemistry of industrial poisons,  hazards, and solvents. 2nd ed. New York,
25             NY: Wiley  Interscience.

        Jones, J. K. (1977) Determination of carbon monoxide by means of a palladium-promazine complex. J. Anal.
               Toxicol. 1:  54-56.

30      Konopinski, V. J. (1984) Residential formaldehyde and carbon dioxide. In: Berglund,  B.; Lindvall, T.; Sundell,
               J., eds. Indoor air:  proceedings of the 3rd international conference on indoor air quality and climate,
               v. 3, sensory and hyperreactivity reactions to sick buildings; August; Stockholm, Sweden.  Stockholm,
               Sweden: Swedish Council for Building Research; pp. 329-334. Available from: NTIS, Springfield, VA;
               PB85-104206.
35
        Koontz, M. D.; Nagda, N. L. (1984) Infiltration and air quality in well-insulated homes: 3. measurement and
               modeling of pollutant levels. In: Berglund, B.;  Lindvall,  T.; Sundell, J., eds.  Indoor air: proceedings of
               the 3rd international conference on indoor air quality and climate, v.  5, buildings, ventilation and thermal
               climate; August; Stockholm, Sweden. Stockholm, Sweden: Swedish Council for Building Research;
40             pp. 511-516.  Available from: NTIS, Springfield, VA; PB85-104222.

        Lambert, J. L.; Wiens, R. E. (1974) Induced colorimetric method for carbon monoxide. Anal. Chem.
               46: 929-930.

45      Larsen, R. I. (1971) A mathematical model for relating air quality measurements to air quality standards.
               Research Triangle Park, NC: U. S. Environmental Protection Agency, Office of Air Programs;
               report AP-89.

        Lawrence Berkeley  Laboratory. (1973) CO monitoring  methods and instrumentation. In: Instrumentation for
50             environmental monitoring air: v. 1. Berkeley, CA: University of California, Lawrence Berkeley
               Laboratory.
         March 12, 1990                               5-31      DRAFT - DO NOT QUOTE OR CITE

-------
       Leithe, W. (1971) The analysis of air pollutants. Ann Arbor, MI: Ann Arbor Science Publishers, Inc. Levaggi,
              D. A.; Feldstein, M. (1964) The colorimetric determination of low concentrations of carbon monoxide.
              Am. Ind. Hyg. Assoc. J. 25: 64-66.

 5     Link, W. T.; McClatchie, E. A.; Watson, D. A.; Compher, A. B. (1971) A fluorescent source NDIR carbon
              monoxide analyzer. Presented at: the joint conference on sensing of environmental pollutants; November;
              Palo Alto, CA. New York, NY: American Institute of Aeronautics and Astronautics; AIAA paper
              no. 71-1047.

10     Luft, K.  F. (1975) Infrared techniques for the measurement of carbon monoxide. Ann. Occup. Hyg. 18: 45-51.

       Lynn, D. A.; Ott, W.; Tabor, E. C.; Smith, R. (1967) Present and future commuter exposure to carbon
              monoxide. Presented at: 60th annual meeting of the Air Pollution Control Association; June; Cleveland,
              OH. Pittsburgh, PA: Air Pollution Control Association; paper no. 67-5.
15
       Malaspina, J.-P.; Bodilis, H.; Giacomoni, L.; Marble, G. (1984) Indoor air pollution: study of two buildings in
              the Paris area. In: Berglund, B.; Lindvall, T.;  Sundell, J., eds. Indoor air: proceedings of the 3rd
              international conference on indoor air quality and climate, v. 5, buildings, ventilation and thermal
              climate;  August; Stockholm, Sweden. Stockholm, Sweden: Swedish Council for Building Research;
20            pp. 499-504. Available from: NTIS, Springfield, VA; PB85-104222.

       McClatchie, E. A. (1972) Development of an infrared  fluorescent gas analyzer. Research Triangle Park, NC: U.
              S. Environmental Protection Agency, Office of Air Quality Planning and Standards; EPA report
              no. EPA-R2-72-121. Available from: NTIS, Springfield, VA; PB-213846.
25
       McClatchie, E. A.;  Compher, A. B.; Williams, K. G.  (1972) A high specificity carbon monoxide analyzer. Anal.
              Instrum.  10: 67-69.

       McCullough, J.  D.; Crane, R. A.; Beckman, A. O. (1947) Determination of carbon monoxide in air  by use of
30            red mercuric oxide. Anal. Chem. 19: 999-1002.

       McKee,  H. C.; Childers, R. E. (1972) Collaborative study of reference method for the continuous  measurement
              of carbon monoxide in the atmosphere (nondispersive infrared spectrometry).  Houston, TX: Southwest
              Research Institute. Available from: NTIS, Springfield, VA; PB-211265.
35
       McKee,  H. C.; Margeson, J. H.; Stanley, T. W. (1973) Collaborative testing of methods to measure air
              pollutants. II.  The non-dispersive infrared method for carbon monoxide. J. Air Pollut. Control Assoc.
              23: 870-875.

40     Moore, J.; West, T. S.; Dagnall, R. M. (1973) A microwave plasma detector system for the measurement of
              trace levels  of carbon monoxide in air. Proc. Soc. Anal. Chem. 10: 197.

       Morgan, M. G.; Morris, S. C. (1977) Needed:  a national R&D effort to  develop individual air pollution monitor
              instrumentation. J. Air Pollut.  Control Assoc.  27: 670-673.
45
       Mueller, R. H. (1954) A supersensitive gas detector permits accurate detection of toxic or combustible gases in
              extremely low concentrations. Anal. Chem. 26: 39A-42A.

       Naoum,  M. M.;  Pruzinec, J.; Toelgyessy, J.; Klehr, E. H. (1974) Contribution to the determination  of carbon
50            monoxide in air by the radio-release method. Radiochem. Radioanal. Lett. 17: 87-93.

       National Air Pollution Control Administration.  (1970) Air quality criteria for carbon  monoxide. Washington,
              DC: U.  S. Department of Health, Education, and Welfare, Public Health Service; report
              no. NAPCA-PUB-AP-62. Available from: NTIS, Springfield, VA; PB-190261.


         March 12, 1990                               5-32       DRAFT - DO NOT QUOTE OR CITE

-------
        National Bureau of Standards. (1975) Catalog of NBS standard reference materials, 1975-76 edition. Washington,
               DC: U. S. Department of Commerce, National Bureau of Standards; NBS special publication no. 260.

        National Institute  for Occupational Safety and Health. (1972) Criteria for a recommended standard....occupational
  5            exposure to carbon monoxide. Washington, DC: U. S. Department of Health, Education, and Welfare;
               report no. NIOSH-TR-007-72. Available from: NTIS, Springfield, VA; PB-212629.

        National Research Council. (1977) Carbon monoxide. Washington, DC:  National Academy of Sciences. (Medical
               and biologic effects of environmental pollutants).
 10
        Naumann, R. J. (1975) Carbon monoxide monitor. Washington, DC: U. S. Patent Office; patent no. 3,895,912;
               July 22, 1975.

        Newton, C.; Morss, L. R. (1974) Portable apparatus for determining atmospheric carbon monoxide. Chemistry
 15            47: 27-29.

        O'Keeffe, A. E.;  Ortman, G.  C. (1966) Primary standards for trace gas analysis. Anal. Chem.  38: 760-763.

        Optical Society of America. (1975) Applications of laser spectroscopy. In: Proceedings of the Spring conference;
20            March; Anaheim, CA. Washington, DC: Optical Society of America.

        Ott, W. R. (1971) An urban survey technique for measuring the spatial variation of carbon monoxide
               concentrations in cities. Stanford, CA: Stanford University, Dept. of Civil  Engineering.

25     Ott, W. R.; Mage, D. T. (1975) Random sampling as an inexpensive means for measuring average annual air
               pollutant concentrations in urban areas. Presented at: 68th annual meeting of the Air Pollution Control
               Association; June; Boston, MA. Pittsburgh, PA: Air Pollution Control Association; paper no. 75-14.3.

        Ott, W. R.; Willits, N. H. (1981) CO exposures of occupants of motor vehicles: modeling the  dynamic response
30            of the vehicle. Stanford, CA: Stanford University, Dept.  of Statistics; SIMS technical report no. 48.

        Ott, W. R.; Rodes, C. E.; Drago, R. J.; Williams, C.; Burmann, F. J. (1986) Automated data-logging personal
               exposure monitors for carbon monoxide. J. Air Pollut. Control Assoc. 36:  883-887.

35     Palanos, P. N. (1972) A practical design for an ambient carbon monoxide mercury replacement analyzer. Anal.
               Instrum. 10: 117-125.

        Paulsell, C. D.  (1976) Use of the National Bureau of Standards standard  reference  gases in mobile source
               emissions testing. In: Calibration in air monitoring, a symposium; August 1975; Boulder, CO.
40            Philadelphia, PA: American Society for Testing and Materials; ASTM special technical publication 598;
               pp. 232-245.

        PEDCo Environmental Specialists, Inc. (1971) Field operations guide for automatic air monitoring equipment.
               Research Triangle Park, NC: U. S. Environmental Protection Agency; report no. APTD-0736. Available
45            from: NTIS, Springfield, VA; PB-202249.

        Perez, J. M.; Broering, L. C.; Johnson, J. H. (1975) Cooperative evaluation of techniques for  measuring nitric
               oxide and carbon monoxide (Phase IV tests). Presented at: the Automotive  Engineering Congress and
               Exposition; February; Detroit, MI. Warrendale, PA: Society of Automotive Engineers, Inc.; SAE
50            technical paper no. 750204.

        Phinney, D. E.; Newman, J.  E. (1972) The precision associated with the sampling frequencies  of total particulate
              at Indianapolis, Indiana. J. Air Pollut. Control Assoc. 22: 692-695.
         March 12,  1990                               5-33      DRAFT - DO NOT QUOTE OR CITE

-------
        Pierce, J. O.; Collins, R. J. (1971) Calibration of an infrared analyzer for continuous measurement of carbon
               monoxide. Am. Ind. Hyg. Assoc. J. 32: 457-462.

        Poli, A. A.; Dessert, C. J.; Benzie, T. P. (1976) Carbon monoxide detection apparatus and method.  Washington,
  5            DC: U. S. Patent Office; patent no. 4,030,887; June 18, 1976.

        Porter, K.;  Volman, D. H. (1962) Flame ionization detection of carbon monoxide for gas chromatographic
               analysis. Anal. Chem. 34: 748-749.

 10     Quackenboss, J. J.; Kanarek, M. S.; Kaarakka, P.; Duffy, C. P.; Flickinger, J.; Turner, W. A. (1984)
               Residential indoor air quality, structural leakage and occupant activities for 50 Wisconsin homes. In:
               Berglund, B.;  Lindvall, T.; Sundell, J., eds.  Indoor air: proceedings of the 3rd international  conference
               on indoor air quality and climate, v. 5, buildings, ventilation and thermal climate; August; Stockholm,
               Sweden. Stockholm, Sweden: Swedish Council for Building Research; pp. 411-420.  Available from:
 15            NTIS, Springfield, VA;  PB85-104222.

        Ramsey, J.  M. (1966) Concentrations of carbon monoxide at traffic intersections in Dayton, Ohio. Arch.
               Environ. Health  13: 44-46.

 20     Ray, R. M.; Carroll, H. B.; Armstrong, F. E.  (1975) Evaluation of small, color-changing carbon monoxide
               dosimeters.  In: Bureau of Mines report of investigations 8051. Washington, DC: U. S. Department of the
               Interior, Bureau of Mines.

        Repp, M. (1977) Evaluation of continuous monitors for carbon monoxide in stationary sources. Research Triangle
 25            Park, NC: U.  S. Environmental Protection Agency, Environmental Science Research Laboratory; EPA
               report no. EPA-600/2-77-063. Available from: NTIS, Springfield, VA; PB-268861.

        Rhodes, R.  C.; Evans, E. G. (1987) Precision and accuracy assessments for state and local air monitoring
               networks, 1985. Research Triangle Park, NC: U. S. Environmental Protection Agency, Environmental
 30            Monitoring Systems Laboratory; EPA report no. EPA-600/4-87-003. Available from:  NTIS, Springfield,
               VA; PB87-145447.

        Robbins, R.  C.; Borg, K. M.; Robinson, E. (1968) Carbon monoxide in the atmosphere. J. Air Pollut. Control
               Assoc. 18:  106-110.
 35
        Saltzman, B. E. (1972) Simplified methods for  statistical interpretation of monitoring data. J. Air Pollut. Control
               Assoc. 22:  90-95.

        Scaringelli,  F. P.; Rosenberg, E.; Rehme, K. A. (1970) Comparison of permeation devices and nitrite ion  as
40            standards for the colorimetric determination of nitrogen dioxide. Environ. Sci. Technol. 4: 924-929.

        Schnakenberg, G. H. (1975) Gas detection instrumentation...what's new and what's to come. Coal Age
               80:  84-92.

45     Schnakenberg, G. H. (1976) Improvements in coal mine gas detection instrumentation. Pap. Symp. Underground
               Min. 2: 206-216.

        Schunck, G.  (1976) Nichtdispersive Infrarot-Gasanalysatoren fuer Industrieprozesse und Umweltschutz
               [Nondispersive infrared gas analyzer for industrial processes and protection of the environment].
 50            DECHEMA Monogr. 80: 753-761.

        Scott, B. (1975) Development of an optical carbon monoxide detector.  Washington, DC: U. S. Department of
               the Interior, Bureau of Mines; report no. 93-75. Available from: NTIS, Springfield, VA; PB-246570.
         March 12,  1990                               5-34      DRAFT - DO NOT QUOTE OR CITE

-------
        Seller, W.; Junge, C. (1970) Carbon monoxide in the atmosphere. JGR J. Geophys. Res. 75: 2217-2226.

        Shepherd, M. (1947) Rapid determination of small amounts of carbon monoxide: preliminary report on the NBS
               colorimetric indicating gel. Anal. Chem.  19: 77-81.
 5
        Shepherd, M.; Schuhmann, S.; Kilday, M. V. (1955) Determination of carbon monoxide in air pollution studies.
               Anal. Chem. 27: 380-383.

        Silverman, L.; Gardner, G. R. (1965) Potassium pallado sulfite method for carbon monoxide detection. Am. Ind.
10            Hyg. Assoc.  J. 26: 97-105.

        Simonescu, T.; Rusu, V.; Kiss, L. (1975) Noi aplicatii analitice ale unor compusi sulfonamidici: metoda cinetica
               de determinare a oxidului de carbon din aer [New analytical applications of some sulfonamide
               compounds: kinetic method for the determination of carbon monoxide in air]. Rev. Chim. (Bucharest)
15            26: 75-78.

        Smith, F.; Nelson, A. C., Jr. (1973) Guidelines for development of a quality assurance program: reference
               method for the continuous measurement of carbon monoxide in the atmosphere. Research Triangle Park,
               NC:  U. S. Environmental  Protection Agency, Quality Assurance and Environmental Monitoring
20            Laboratory; EPA report no. EPA-R4-73-028a. Available from: NTIS, Springfield, VA; PB-222512.

        Smith, R. G. (1969)  Air quality standards for carbon monoxide. New York, NY: American Petroleum Institute,
               Division of Environmental Affairs; air quality monograph no. 69-9.

25     Smith, R. G.; Bryan, R. J.; Feldstein, M.; Locke, D. C.; Warner, P. O. (1975a) Tentative method for constant
               pressure volumetric gas analysis for O2, CO2, CO, N2, hydrocarbons (ORSAT). Health Lab. Sci.
               12: 177-181.

        Smith, R. G.; Bryan, R. J.; Feldstein, M.; Locke, D. C.; Warner, P. O. (1975b) Tentative method for gas
30            chromatographic analysis of O2,  N2, CO,  CO2, and CH4. Health Lab. Sci. 12:  173-176.

        Stedman, D.  H.; Kok, G.; Delumyea, R.; Alvord, H. H.  (1976) Redundant calibration of nitric oxide, carbon
               monoxide, nitrogen dioxide, and ozone air pollution monitors by chemical and gravimetric techniques. In:
               Calibration in air monitoring, a symposium; August 1975; Boulder, CO. Philadelphia, PA: American
35            Society for Testing and Materials; ASTM  special technical publication 598; pp. 337-344.

        Stetter,  J. R.; Blurton, K.  F. (1976) Portable high-temperature catalytic reactor: application to air pollution
               monitoring instrumentation. Rev. Sci. Instrum. 47: 691-694.

40     Stevens, R. K.; Herget, W. F. (1974) Analytical  methods applied to air pollution measurements. Ann Arbor, MI:
               Ann Arbor Science Publishers, Inc.

        Swinnerton, J. W.; Linnenbom, V. J.; Cheek,  C.  H. (1968) A sensitive gas chromatographic method for
               determining carbon monoxide in seawater. Limnol. Oceanogr. 13: 193-195.
45
        Tesarik, K.; Krejci, M.  (1974) Chromatographic determination of carbon monoxide below the 1  ppm level. J.
               Chromatogr.  91: 539-544.

        Traynor, G. W.; Apte, M. G.; Carruthers, A.  R.; Dillworth, J. F.; Grimsrud, D. T.; Thompson, W. T. (1984)
50            Indoor air pollution and inter-room transport due to unvented kerosene-fired space heaters. Berkeley, CA:
               Lawrence Berkeley Laboratory; report no. LBL-17600. Available from: NTIS, Springfield, VA;
               DE84015949.
         March 12,  1990                              5-35      DRAFT - DO NOT QUOTE OR CITE

-------
        U. S. Senate Committee on Environment and Public Works. (1977) The Clean Air Act as amended August 1977.
               Washington, DC: U.S. Government Printing Office; serial no. 95-11.

        v. Heusden, S.; Hoogeveen, L. P. J. (1976) Chemiluminescent determination of reactive hydrocarbons. Z. Anal.
 5             Chem. 282: 307-313.

        van Dijk, J. F. M.; Falkenburg, R. A. (1976) A high sensitivity carbon monoxide monitor for ambient air. In:
               International conference on environmental sensing and assessment. Volume 2. A joint conference
               comprising the International Symposium on Environmental Monitoring and Third Joint Conference on
10             Sensing of Environmental Pollutants; September 1975; Las Vegas, NV. New York, NY: Institute of
               Electrical & Electronics Engineers, Inc.; paper no. 35-5.

        Verdin, A. (1973) Gas analysis instrumentation. New York, NY: John Wiley & Sons, Inc.

15      Vol'berg, N. Sh.; Pochina, I.  I. (1974) Continuous determination of the concentration of carbon monoxide in the
               atmosphere by the coulometric method. In: Voprosy atmosfernoy diffuzii i zagryazneniya vozdukha
               [Atmospheric diffusion and air pollution problems]. Leningrad, USSR: Gidrometeoizdat Press;
               pp. 146-157.

20      Ward, T. V.; Zwick, H. H. (1975) Gas cell correlation spectrometer: GASPEC. Appl. Opt.  14: 2896-2904.

        Wechter, S. G. (1976) Preparation of stable pollution gas  standards using treated aluminum cylinders. In:
               Calibration in air monitoring, a symposium;  August 1975; Boulder,  CO. Philadelphia,  PA: American
               Society for Testing and Materials; ASTM special technical publication 598; pp. 40-54.
25
        Wohlers, H.  C.; Newstein, H.; Daunis, D.  (1967) Carbon monoxide and sulfur dioxide adsorption on- and
               desorption from glass, plastic, and metal tubings. J. Air Pollut. Control  Assoc. 17: 753-756.

        Yamate, N.;  Inoue, A. (1973) Continuous analyzer of carbon monoxide in ambient air using electrochemical
30             analysis. Kogai to Taisaku 9: 292-296.

        Ziskind, R. A.; Rogozen,  M.  B.; Carlin, T.; Drago, R. (1981) Carbon monoxide intrusion into sustained-use
               vehicles. Environ. Int.  5: 109-123.
         March 12, 1990                               5-36      DRAFT - DO NOT QUOTE OR CITE

-------
            6.   AMBIENT CARBON MONOXIDE SOURCES,
                  EMISSIONS, AND  CONCENTRATIONS
 5     6.1 ESTIMATING NATIONAL EMISSION FACTORS
           The national carbon monoxide emission estimates presented herein are taken from two
       U.S. Environmental Protection Agency (EPA) reports: National Air Quality and Emissions
       Trends Report, 1988 (U.S. Environmental Protection Agency, 1990a) and National Air
       Pollutant Emission Estimates, 1940-1988 (U.S. Environmental Protection Agency, 1990b).
10     These data are most useful as indicators of overall emission trends, since national totals or
       averages are not the best guide for estimating or predicting specific trends in local areas.  The
       emission data represent calculated estimates based on standard emission inventory procedures
       developed by  the Office of Air Quality Planning and Standards of the U.S. Environmental
       Protection Agency (1990b). These procedures either estimate the emissions directly or
15     estimate the magnitude of other variables that can then be related to emissions. For CO,
       these indicators include fuel consumption, vehicle population, vehicle miles traveled (VMT),
       sales of new vehicles,  tons of refuse burned, raw materials processed, etc., which are then
       multiplied by  the appropriate CO emission factor(s) to obtain the CO emission estimate(s).  It
       should be noted that emission factors have specific limitations and applicability.  Emission
20     factors, in general, are not precise indicators of emissions from a single source;  rather, they
       are quantitative estimates of the average rate of pollutant released as a result of some activity.
       They are most valid when applied to a large number of sources and processes. Emission
       factors thus relate quantity of pollutants emitted to indicators such as those noted above, and
       are EPA's approach for determining national estimates of emissions from various source
25     categories.
       6.2  EMISSION SOURCES AND EMISSION FACTORS BY SOURCE
            CATEGORY
30          Emission source categories, as presented in Table 6-1, are divided into five individual
       categories: transportation, stationary source fuel combustion, industrial processes, solid waste

       March 12, 1990                         6-1      DRAFT-DO NOT QUOTE OR CITE

-------
                                TABLE 6-1.  CARBON MONOXIDE NATIONAL EMISSION ESTIMATES (TERAGRAMS/YEAR)
K>

VO
£
to
Source Category
Transportation
Highway Vehicles
Aircraft
Railroads
Vessels
Other-Off Highway
Transportation Total
Stationary Source Fuel Combustion
Electric Utilities
Industrial
Commercial-Institutional
Residential
Fuel Combustion Total
Industrial Processes
Solid Waste Disposal
Incineration
Open Burning
Solid Waste Total
Miscellaneous
Forest Fires
Other Burning
Miscellaneous Organic Solvent
Miscellaneous Total
Total of All Sources
1970

65.3
0.9
0.3
1.2
6.8
74.4

0.2
0.7
0.1
3.5
4.5
8.9

2.7
3.7
6.4

5.1
2.1
0.0
7.2
101.4
1975

57.2
0.9
0.2
1.4
5.4
65.0

0.3
0.7
0.1
3.3
4.3
6.9

1.8
1.3
3.1

4.0
0.8
0.0
4.8
84.1
1978

55.6
1.0
0.3
1.5
4.8
63.1

0.3
0.8
0.1
4.8
5.9
7.2

1.4
1.1
2.5

5.0
0.7
0.0
5.7
84.4
1979

51.9
1.0
0.3
1.4
4.5
59.1

0.3
0.7
0.1
5.7
6.7
7.1

1.3
1.0
2.3

5.8
0.7
^Q
6.5
81.7
1980

48.7
1.0
0.3
1.4
4.7
56.1

0.3
0.7
0.1
6.4
7.4
6.3

1.2
1.0
2.2

6.9
0.7
0.0
7.6
79.6
1981

48.0
1.0
0.3
1.4
4.7
55.4

0.3
0.6
0.1
6.7
7.7
5.9

1.2
0.9
2.1

5.8
0.6
0.0
6.4
77.4
1982

45.9
1.0
0.2
1.4
4.4
52.9

0.3
0.6
0.1
7.3
8.2
4.3

1.1
0.9
2.0

4.3
0.6
0.0
4.9
72.4
1983

45.9
1.0
0.2
1.4
3.9
52.4

0.3-
0.6
0.1
i2
8.2
4.3

1.0
.09
1.9

7.1
0.6
0.0
7.7
74.5
1984

43.5
1.0
0.2
1.7
4.2
50.6

0.3
0.6
0.1
JL3
8.3
4.7

1.0
.02
1.9

5.7
0.6
0.0
6.3
71.8
1985

40.7
1.1
0.2
1.4
4.5
47.9

0.3
0.6
0.1
^5
7.4
4.4

1.1
.09
2.0

4.7
0.6
0.0
5.3
67.0
1986

37.5
1.1
0.2
1.5
4.4
44.6

0.3
0.6
0.1
A6
7.5
4.3

0.9
.08
1.7

4.4
0.6
0.0
5.0
63.1
1987

36.0
1.1
0.2
1.6
4.4
43.2

0.3
0.6
0.1
^6
7.6
4.5

0.9
0.8
1.7

6.5
0.6
0.0
7.1
64.1
1988

34.1
1.1
0.2
1.6
4.2
41.2

0.3
0.6
0.1
^6
7.6
4.7

0.9
0.8
1.7

5.4
0.6
_OQ
6.0
61.2
           Source: U.S. Environmental Protection Agency (1990b).

-------
        disposal, and miscellaneous.  The methodology used in the generation of emission estimates
        for the individual source categories is summarized below.

        6.2.1  Transportation Sources
 5           Transportation sources include emissions from all mobile sources including highway and
        other off-highway motor vehicles. Highway motor vehicles include passenger cars, trucks,
        buses and motorcycles. Off-highway vehicles include aircraft, railroads, vessels, and
        miscellaneous engines such as farm equipment, industrial and construction machinery,
        lawnmowers, and snowmobiles.
10
        6.2.1.1  Motor Vehicles
             Emission estimates from gasoline- and diesel-powered motor vehicles are based upon
        vehicle-mile tabulations and emission factors.  Eight vehicle categories are considered; light
        duty gasoline (mostly passenger cars), light duty diesel passenger cars, light duty gasoline
15      trucks (weighing less than 6000 pounds) light duty gasoline trucks (weighing 6001 to
        8500 pounds),  light duty diesel trucks, heavy duty gasoline trucks and buses, and heavy duty
        diesel trucks and buses, and motorcycles.  The emission factors used are based on EPA's
        MOBILE4 mobile source emission factor model, which uses the latest available data to
        estimate average in-use emissions from highway vehicles (U.S. Environmental Protection
20      Agency, 1989a,b).  The MOBILE4 model, developed  by the EPA Office of Mobile Sources,
        was used to calculate emission factors for each year.  The emission factors are weighted to
        consider the approximate amount of motor vehicle travel in low- and high-altitude areas to
        obtain overall national average emission factors. For each area a representative average
        annual temperature (low altitude average annual temperature  = 57°F, high altitude = 54°F,
25      California  = 65 °F), together with national averages for motor vehicle registration
        distributions and annual mileage accumulation rates by age and hot/cold start vehicle
        operation percentages were used to calculate the emission factors.  Average speed is taken
        into account according to the published distribution of vehicle-miles travelled (VMT) (U.S.
        Department of  Transportation, 1988). The published VMT are divided into three road
30      categories corresponding to roads with assumed average  speeds of 55 miles per hour for
       March 12, 1990                           6-3       DRAFT-DO NOT QUOTE OR CITE

-------
       interstates and other primary highways, 45 miles per hour for rural roads, and 19.6 miles per
       hour for urban streets.

       6.2.1.2 Aircraft
 5          Aircraft emissions are based on emission factors and aircraft activity statistics reported
       by the Federal Aviation Administration (1988). Emissions are based on the number of
       landing-takeoff (LTO) cycles.  Any emissions in cruise mode, which is defined to be above
       3000 feet (1000 meters) are ignored.  Average emission factors for each year, which take into
       account the national mix of aircraft types for general aviation, military, and commercial
10     aircraft, are used to compute the emissions.

       6.2.1.3 Railroads
            The Department of Energy reports consumption  of diesel fuel and residual fuel oil by
       railroads (U.S. Department of Energy, 1988a). Average emission factors applicable to diesel
15     fuel consumption were used to calculate emissions.

       6.2.1.4  Vessels
            Vessel use of diesel fuel, residual oil, and coal is reported by the Department of Energy
       (U.S.  Department  of Energy, 1988a,b).  Gasoline use is based on national boat and motor
20     registrations, coupled with a use factor (gallons/motor/year) (U.S. Environmental Protection
       Agency, 1973) and marine gasoline sales (U.S. Department of Transportation, 1988).
       Emission factors from AP-42 are used to compute emissions (U.S. Environmental Protection
       Agency, 1985).

25     6.2.1.5  Nonhighway Use  of Motor Fuels
             Gasoline and diesel fuel are consumed by off-highway vehicles in substantial quantities.
       The fuel consumption is divided into several categories including  farm tractors, other farm
       machinery, construction equipment, industrial machinery, snowmobiles, and small general
       utility engines such as lawnmowers and snowthrowers.  Fuel use is estimated for each
30     category from estimated equipment population and an annual use  factor of gallons per unit per
       year (Hare and Springer, 1974), together with reported off-highway diesel fuel deliveries

       March 12, 1990                            6-4       DRAFT-DO NOT QUOTE OR CITE

-------
       (U.S. Department of Energy, 1988a) and off-highway gasoline sales (U.S. Department of
       Transportation, 1988).

       6.2.2  Stationary Source Fuel Combustion
 5          Stationary combustion equipment, such as coal-, gas-, or oil-fired heating or power
       generating plants, generate CO as a result of improper or inefficient operating practices or of
       inefficient combustion techniques. The specific emission factors for stationary fuel
       combustors vary according to the type and size of the installation and the fuel used, as well as
       the mode of operation.  The EPA compilation of air pollutant emission factors provides
10     emission data obtained from source tests,  material balance studies, engineering estimates, etc.,
       for the various common emission categories. For example, coal-fired electricity-generating
       plants report coal use to the Department of Energy (U.S. Department of Energy, 1988b,c).
       Distillate oil,  residual oil, kerosene, and natural gas consumed by stationary combustors are
       also reported by user category to the U.S. Department of Energy (1988a, 1989a,b).  Average
15     emission factors from AP-42 (U.S. Environmental Protection Agency,  1985) were used to
       calculate the emission estimates. The consumption of wood in residential wood stoves has
       likewise been estimated by the U.S. Department of Energy (1982, 1984).

       6.2.3  Industrial Processes
20          In addition to fuel combustion, certain other industrial processes generate and emit
       varying quantities of CO into the air.  The lack of published national data on production, type
       of equipment, and controls,  as well  as an absence of emission factors, makes it impossible to
       include estimates of emissions from all industrial process sources.
            Production data for industries that produce the great majority of emissions were derived
25     from literature data.  Generally, the Minerals Yearbook, (U.S. Department of the Interior,
       1987) published by the Bureau of Mines,  and Current Industrial Reports, (U.S.  Department of
       Commerce, annual) published by the Bureau of the Census, provide adequate data for most
       industries. Average emission factors were applied to production data to obtain emissions.
       Control efficiencies applicable to various  processes were estimated on the basis of published
30     reports and from NEDS data (U.S.  Environmental Protection Agency, 1971).
        March 12, 1990                          6-5      DRAFT-DO NOT QUOTE OR CITE

-------
       6.2.4  Solid Waste Disposal
            Solid waste CO emissions result from the combustion of wastes in municipal and other
       incinerators, and also from the open burning of domestic and municipal refuse.  Specific
       emission estimates for the various waste combustion procedures in use were taken from a
 5     study conducted in  1968 concerning solid waste collection and disposal practices (U.S.
       Department of Health, Education, and Welfare, 1968).  Results of this study indicate that the
       average collection rate of solid waste is about 5.5 pounds per capita per day in the United
       States.  It has been stated that a conservative estimate of the total generation rate is 10 pounds
       per capita per day.  The results of this survey were updated based on data reported in NEDS
10     and used to estimate, by disposal method, the quantities of solid waste generated (NEDS,
       National Emissions Data System,  n.d.).  Average emission factors were applied to these totals
       to obtain estimates of total emissions from the disposal of solid  wastes.

       6.2.5  Miscellaneous Combustion Sources
15          Miscellaneous CO emissions results from the burning of forest and agricultural
       materials, smoldering coal refuse  materials, and structural fires.

       6.2.5.1 Forest Fires
            The Forest Service of the Department of Agriculture publishes information on the
20     number of forest fires and the acreage burned (U.S. Forest Service, 1988).  Estimates of the
       amount of material burned per acre are made to estimate the total amount of material burned.
       Similar estimates are made to account for managed burning of forest areas.  Average emission
       factors were applied to the quantities of materials burned to calculate emissions.

25     6.2.5.2 Agricultural Burning
            A study was conducted by EPA (Yamate, 1974) to obtain  from local agricultural and
       pollution control agencies estimates of the number of acres and  estimated quantity of material
       burned per acre in agricultural burning operations.  These data have been updated and used to
       estimate agricultural burning emissions, based on average emission factors.
30
       March 12, 1990                          6-6      DRAFT-DO NOT QUOTE OR CITE

-------
       6.2.5.3 Coal Refuse Burning
            Estimates of the number of burning coal-refuse piles existing in the United States are
       made in reports by the Bureau of Mines (U.S. Department of the Interior, 1971).  This
       publication presents a detailed discussion of the nature, origin, and extent of this source of
 5     pollution.  Rough estimates of the quantity of emissions were obtained using this information
       by applying average emission factors for coal combustion. It was assumed that the number of
       burning refuse piles decreased to a negligible amount by 1975.

       6.2.5.4 Structural Fires
10          The United States Department of Commerce publishes, in their statistical abstracts,
       information on the number and types of structures damaged by fire (U.S. Department of
       Commerce, 1988). Emissions were estimated by applying average emission  factors for wood
       combustion to these totals.

15
       6.3  NATIONAL CO EMISSIONS ESTIMATES 1970-1988
            Table 6-1 displays the total annual 1970-1988 CO emissions from the various source
       categories. The CO estimations cited herein are the result of current methodology and refined
       emission factors and should not be compared with data reported earlier.  These data indicate
20     that CO from all man-made sources in the U.S. declined from 101.4 teragrams in  1970 (one
       teragram equals 1012 grams, 103 gigagrams,  106 metric tons, or approximately 1.1 x 10* short
       tons) to 61.2 teragrams in 1988.  The majority, about 70 percent, of the CO emissions total
       comes from transportation sources, 12 percent comes from fuel combustion processes,
       7 percent comes from industrial processes, and 11 percent comes from miscellaneous sources.
25     Table 6-2  contains a more detailed listing of CO emissions from the dominant category,
       transportation sources.
            The single largest contributing source of CO emissions is highway vehicles, which emit
       60 percent of the total national CO estimate. Because of the implementation of the Federal
       Motor Vehicle Control Program (FMVCP),  CO emissions from highway vehicles have
30     declined 48 percent, from 65.3 teragrams in 1970 to  34.1 teragrams in 1988. Figure 6-1
       displays how CO emissions from the major highway vehicle categories have changed from

       March 12,  1990                           6-7       DRAFT-DO NOT QUOTE OR CITE

-------
1
Ki
t — i
VO
'O
o







TABLE 6-2.

Source Category


Highway vehicles
Gasoline-powered
Passenger cars
Light trucks - 1
Light trucks - 2
Heavy duty vehicles
Motorcycles

1970




49,090
5,800
2,070
7,810
260
CARBON MONOXIDE EMISSIONS FROM TRANSPORTATION (GIGAGRAMS/YEAR)

1975




41,430
5,730
2,450
6,610
540

1978




37,840
6,100
4,120
6,500
550

1979




34,450
5,960
4,340
6,170
490

1980




31,850
5,810
4,210
5,870
370

1981




30,160
6,370
4,700
5,780
280

1982




30,150
5,760
4,220
4,910
200

1983




29,510
6,190
4,610
4,720
190

1984




27,790
6,050
4,450
4,380
170

1985




25,410
6,280
4,330
3,750
130

1986




23,650
6,060
4,040
2,920
120

1987




22,530
6,100
3,740
2,800
130

1988




21,220
5,770
3,540
2,730
120
          Total - Gasoline
                           65,030  56,760   55,110   51,410  48,110   47,290   45,240   45,220   42,840   39,900   36,790  35,300  33,380
oo
Diesel-powered
  Passenger cars
  Light trucks
  Heavy duty vehicles

    Total - Diesel
  0
  0
300

300
  0
  0
390

390
  2
  1
500

503
  5
  1
530

536
  8
  3
610

621
 10
  6
700

716
 10
  6
680

696
 20
  5
650

675
 20
  3
650

673
 20
  4
770

794
 20
  4
670

694
 20
  3
680

703
 10
  4
720

734
          Highway Vehicle Total    65,330
                                             55,613   51,
                                         48,006
                                                          43,513
                                                                    37,484  36,003   34,114
O
o
25
3
o
d
3
w
o
5*3
O
3

Aircraft
Railroads
Vessels
Farm Machinery
Construction Machinery
Industrial Machinery
Other Off-highway Vehicles

Transportation Total


900
250
1,150
3,570
580
1,780
840

74,400


880
240
1,360
2,930
370
1,060
990

64,980


960
260
1,470
2,370
340
1,070
1.050

63,133


990
270
1,420
2,240
370
820
1.080

59,136


990
270
1,380
2,040
460
1,110
1.100

56,081


960
250
1,440
1,880
370
1,330
1.150

55,386


950
240
1,390
1,780
320
1,190
1.130

52,936


980
190
1,410
1,470
260
1,040
1.140

52,385


1,010
200
1,700
1,900
250
900
1.130

50,603


1,090
190
1,400
2,120
410
850
1.150

47,904


1,080
180
1,500
1,910
450
840
1.170

44,614


1,060
190
1,560
1,830
520
880
1.190

43,233


1,050
190
1,620
1,630
530
880
1.200

41,214

      Source:  U.S. Environmental Protection Agency (1990b).

-------
                                                                          Heavy Duty Gasoline
                                                                          Vehicles
                                                                          Light Duty Trucks
                                                                          1&2
                                                                          Light Duty Gasoline
                                                                          Vehicles
        1975   1980   1981
1982  1983

      Year
1984   1985   1986   1987   1988
 Figure 6-1.  Estimated emissions of carbon monoxide from highway vehicles, 1970-1988.

 Source: U.S. Environmental Protection Agency (1990b).
March 12, 1990
             6-9
         DRAFT-DO NOT QUOTE OR CITE

-------
       1970-1988 as a result of the FMVCP.  Although VMT increased 38 percent from 1970 to
       1978, total CO emissions from highway vehicles decreased 15 percent, because of the
       installation of FMVCP-mandated air pollution control devices on new vehicles.  Petroleum
       crises contributed to a 1.7 percent decrease in VMT from 1978 to 1980.  This lack of growth
 5     in vehicle travel together with an increased degree of pollution control because of stricter
       emission standards for new vehicles, coupled with the gradual disappearance of older
       uncontrolled vehicles from the vehicle fleet, produced an estimated 12 percent drop in
       highway vehicle CO emissions in this two year period from 1978 to 1980.  Since 1980, VMT
       have grown each year.  From 1980 to 1988, VMT increased by 33 percent. However, due to
10     the FMVCP controls, CO emissions from highway vehicles actually decreased 30 percent
       during this period. Overall from 1970 to 1988, without the implementation of FMVCP,
       highway vehicle emissions would have increased 61 percent; with FMVCP implementation,
       emissions are estimated to have decreased 48 percent.
            CO emissions from other sources have also generally decreased. In 1970,  emissions
15     from burning of agricultural crop residues were greater than in  more recent years.  Solid
       waste disposal emissions have also decreased as the result of implementation of regulations
       limiting or prohibiting burning of solid waste in many areas.  Emissions  of CO from
       stationary source fuel combustion occur mainly from the residential sector. These emissions
       were reduced somewhat through the mid-1970's as residential consumers converted  to natural
20     gas, oil, or electric heating equipment.  Recent growth in the use of residential wood stoves
       has reversed this trend, but increased CO emissions from residential sources continue to be
       small compared to highway vehicle emissions.  Nevertheless, in 1988 residential wood
       combustion accounted for about 10 percent of national CO emissions, more than any source
       category except highway vehicles. CO emissions from industrial processes have generally
25     been declining since 1970 as the result of the obsolescence of a few high-polluting processes
       such as manufacture of carbon black by the channel process and installation of controls on
       other processes.  However, industrial process emissions increased slightly (4 percent) from
       1987  to 1988 due to increased industrial activity.
30
       March 12, 1990                          6-10      DRAFT-DO NOT QUOTE OR CITE

-------
       6.4  OUTDOOR AIR CONCENTRATIONS
       6.4.1  Introduction
            Ambient concentrations of carbon monoxide (CO) in urban communities vary widely
       with time and space.  Actual human exposure to CO in various indoor and outdoor activities
 5     is affected by highly localized microenvironments which are influenced by nearness to
       sources, including vehicles, occupations, and by personal activities, such as smoking. Indoor
       sources and concentrations are summarized in Chapter 7. Exposure information is presented
       in Chapter 8. This section presents information about ambient concentrations.  It will
       describe observed diurnal, seasonal,  and annual patterns of ambient urban CO levels and will
10     explain the importance of air monitoring site selection, of meteorological and geographic
       effects on CO exposures, techniques of CO trend analyses, and special CO exposure
       situations. An overview of meteorological diffusion models is also provided.

       6.4.2  Site Selection
15          Site selection is one of the most complex and critical elements in the design of CO air
       monitoring programs.  This is especially important for CO monitoring because the proximity
       of the monitor to traffic will influence the magnitude of CO concentrations.  Naturally, the
       choice of monitoring sites depends greatly on the objective of the monitoring to be performed.
       The U.S. Environmental Protection Agency (1977a) recognizes the following as general
20     objectives for monitoring:

            1.     To judge compliance with and/or progress made toward meeting ambient air
                  quality standards.
25          2.     To activate emergency control procedures to ameliorate air pollution episodes.
            3.     To observe pollution trends throughout the region, including the nonurban areas.
                  (Information from nonurban areas is needed to evaluate whether air quality in the
                  cleaner portions of a region is deteriorating significantly and to gain knowledge
30                about background pollutant levels.)
       March 12, 1990                          6-11      DRAFT-DO NOT QUOTE OR CITE

-------
             4.     To provide a data base for application in the evaluation of effects; in urban, land
                   use, and transportation planning; in development and evaluation of abatement
                   strategies; and in development and validation of diffusion models.
             In addition to these general objectives, site selection is also based on the scale of
        representativeness that will meet the objectives.  Data representativeness, like measures of
        concentration at a site, is dependent on the proximity of the monitor to the CO source but,
10      further, is influenced by the intended use to which the data will be put.  Ground level
        concentrations of CO within an urban area vary widely because the principal source of CO in
        cities is automobiles which, obviously, move and are more concentrated in some areas at
        some times than at other times. Monitoring sites at the edge of a highway will measure CO
        concentrations representative of a fairly small area.  Sites well removed from highways can be
15      representative of a fairly large-scale area.  The EPA has defined six scales of spatial
        representativeness for CO monitoring sites:  microscale, middle scale, neighborhood scale,
        urban scale, regional scale, and national and global  scale (Federal Register, 1979).
             Most CO  monitoring conducted in the United States is for the purpose of determining
        attainment or nonattainment of air quality  standards. Since monitoring resources have been
20      and continue to be severely limited, monitoring sites are usually selected by a "worst case"
        principle; that is,  they are set up where maximum CO concentrations are expected because the
        NAAQSs focus on peaks.  As a result, many CO sites are located in close proximity to major
        highways, arterials, and downtown street canyons.  This means that they are situated  where
        maximum CO levels occur, but that their scale of representativeness is small. Monitoring
25      results may thus relate primarily to pedestrian exposure near the monitor.  Sites located away
        from the major roadways, but within highly populated neighborhoods with high traffic
        densities, may be more representative of the maximum CO concentrations to which a large
        portion of the population of a city may be exposed.
             The EPA  has published guidelines (Federal Register, 1979) for CO monitor siting
30      (Table 6-3). EPA guidelines (U.S. Environmental Protection Agency, n.d.) give the highest
        priority to microscale sites within street canyons and to neighborhood sites where maximum
        concentrations are expected.
       March 12, 1990                          6-12      DRAFT-DO NOT QUOTE OR CITE

-------
                                                 TABLE 6-3.  SPECIFIC PROBE EXPOSURE CRITERIA
 I
 I—>
 K)
 vo
 O
O
O
3
9
a
O
X)
n

Site Type

Height
Above
Ground
Expected
Concentration
Gradient
with Height
(1-hr Average)

Separation
of Monitor from
Influencing Sources

General Remarks
Street Canyon
 Peak Concentration             3 ± 1/2 m
 Average Concentration          3 ± 112m
                                                               ~.5 ppm/m
                                                               ~3 ppni/m
Mid-sidewalk or 2 m
from side of building.
On leeward side of
street.

Mid-sidewalk or 2 m
from side of building.
Central Business District.
High density, slow-moving
traffic.
Dense multiple-story buildings
lining both sides of street.
Neighborhood
Peak Concentration 3 ± 1/2 m 5%/m
Average Concentration 3 ± 1/2 m 5%/m
<.3 ppm/m
Background 3 to 10 m .2%m
New Source Review
Preconstruction 3 ± 1/2 m 5%/m
Postconstruction 3 ± 1/2 m > 5%/m to
< .Sppm
Setback VPD
3.5 km 100,000
1.5 km 50,000
200 m 10,000
100 m 5,000
35 m 1,000
25 m any
Dependent on traffic
volume, road config-
uration and setback
distance of commercial
or residential activity.
5 to 6 km; > 3,000 VPH
maximum.
400 m; > 100 VPD.
Usually the same as
neighborhood.
Usually the same as
corridor or street canyon.
Commericial or residential
neighborhood. This separation
criteria limits the effect of
these streets to =1 ppm.
Stop and go or limited access
traffic > 50,000 VPD or greatest
in area.
35 km downwind in least
frequent wind direction from
city, limit effects to .2 ppm.
Area of lowest concentration
in proposed indirect source
location for background.
Area of maximum concentration
in area of complete area source.
            Source:  Federal Register (1979).

-------
            The variability of CO concentration with height in the vicinity of a highway is
       sufficiently large that the representativeness of measurements will be strongly affected by
       variability of the inlet probe height.  It is, therefore, necessary to standardize the height of the
       inlet probe so that data collected at one air monitoring station is comparable to data collected
 5     at others.  In an effort to characterize typical human exposure, the sample inlet probe height
       should ideally be at breathing level.  However, as a compromise between representation of
       breathing height and practical considerations, such as prevention of vandalism, it is
       recommended that inlets for most kinds of sampling be at a height of 3 ±  0.5 m (Altshuller
       et al., 1966; Bach et al.,  1973). A one-meter minimum separation of the probe from adjacent
10     structures is also recommended to avoid the frictional effects of surfaces on the movement of
       air (Anonymous, 1976).
            Site selection for monitors used for purposes other than trend analysis and determination
       of compliance with air quality standards may not follow the specific criteria that apply to
       continuous monitoring sites.  In fact, special purpose studies in which CO concentrations are
15     measured at many locations provide information about the spatial variations of ambient CO
       that form the basis for setting site-selection criteria.  Among the principal  types of special
       purpose monitoring are research studies for diffusion model development and improvement
       and for source surveillance studies.

20     6.4.3 United States  Data Base
            Monitoring stations reporting data to EPA's Aerometric Information Retrieval System
       (AIRS) fall into two major categories: the National Air Monitoring Stations (NAMS) and the
       State and Local Air Monitoring Stations (SLAMS).  The NAMS were established through
       monitoring regulations promulgated in May 1979 (Federal Register,  1979) to provide EPA
25     with accurate and  timely  data on a national scale. The NAMS are located at  sites expected to
       incur high pollutant concentrations and to typify areas with the potential for high population
       exposure.  These stations meet uniform criteria for site location, quality assurance, and
       equivalent analytical methodology, sampling intervals, and instrument selection to assure
       consistent data reporting  nationwide. The SLAMS, in general, meet the same rigid criteria
30     but,  in addition to the above siting criteria for highest concentrations and population exposure
       potential, they may be located to monitor a greater diversity of urban neighborhoods.

       March 12, 1990                          6-14      DRAFT-DO NOT QUOTE OR CITE

-------
            In accordance with requirements of the Clean Air Act and EPA regulations for State
       Implementation Plans (SIPs) (Code of Federal Regulations, 1977) ambient CO data from
       Federal networks must be reported each calendar quarter to AIRS.  State and local agencies
       report most of the data from their SLAMS stations as well. As a result, continuous
 5     measurements of ambient CO concentrations from numerous cities throughout the United
       States are available from the U.S. Environmental Protection Agency.
             Computer retrievals of raw data submitted to the EPA's AIRS data bank and published
       data summaries such as the National Air Quality and Emission Trends Report (U.S.
       Environmental Protection Agency, 1988) and Air Quality Data - Annual Statistics are
10     available (U.S. Environmental Protection Agency, 1974b,c, 1976d,e,f,g,h, 1977b).
       However, state and local  air pollution control agencies are not required to submit all CO data
       collected from their monitoring network.  These agencies may also conduct special studies for
       certain "in-house" purposes.  State departments of transportation and local metropolitan
       planning commissions are sources of CO data for the preparation of environmental impact
15     statements for proposed transportation projects and/or in the preparation of SIP revisions.  Air
       quality impact research sponsored by the EPA, the Federal Highway Administration,
       universities, and private industries also are sources of CO data.

       6.4.4  Techniques of Data Analysis
20           Air quality surveys inherently involve taking a limited number of samples from a highly
       variable and uncontrolled population (i.e., the environment).  For this reason, air quality data
       should be analyzed through statistical methods, which can be used to describe the behavior of
       the total population on the basis of a finite number of samples.  In particular, statistical
       parameters can be calculated to describe the typical values observed, the maximum or peak
25     values observed, and the range of values observed.
             Although intermittent sampling is an important research tool for conducting special
       studies, the majority of CO monitoring instruments in use today are intended to operate
       continuously and to yield successive hourly averages. These data are applied for two
       principal uses:  (1) characterizing environmental conditions by describing short-term (hourly,
30     daily, seasonal) and long-term (year-to-year) urban CO concentration patterns, and
        March 12, 1990                           6-15      DRAFT-DO NOT QUOTE OR CITE

-------
       (2) evaluating, for statutory purposes, an area's status with respect to the 1-hour and 8-hour
       average NAAQS for CO.
             At a minimum, an analysis of CO air quality data should include a comparison of the
       highest (or second highest) observed pollution concentration to established air quality
 5     standards.  In addition, an analysis of CO data may include calculation of population
       statistics, frequency analyses, averaging time analyses, trend analyses, and case analysis.

       6.4.4.1 Frequency Analysis
             In most areas of air pollution monitoring modeling, we do not have enough knowledge
10     about the generation, dispersion, and transport of air pollutants to formulate a convincing
       theoretical model for air quality data.  Air pollutant concentrations are often generated by
       autocorrelated stochastic processes.  In most situations we never know which of several
       hypothetical  models are "correct." Fortunately, it is usually possible to identify time periods
       and pollutant averaging times when observations are approximately stationary and independent
15     sequences of concentrations.  Horowitz and Barakat (1979) showed that autocorrelation does
       not significantly affect the validity of the usual methods for estimating the parameters of the
       maximum pollutant concentration distribution.  The most widely used model has been the
       two-parameter lognormal distribution which  has played a major role  in the formulation of air
       quality standards for many pollutants Georgopoulos and Seinfeld (1982).  However, there are
20     many data sets for which some other distribution fits better. These candidate models include
       the three-parameter lognormal, Weibull, exponential, and gamma distributions (Bencala and
       Seinfeld, 1976; Ott et al., 1979; Pollack, 1975; Simpson et al., 1984). This variety of
       distribution types probably reflects the phenomenon that an air pollution concentration is the
       superposition of a random number of point,  line, and area sources of different emission
25     strengths.
             The NAAQSs for CO are currently based on a 1-hour and an 8-hour averaging time.
       Carbon monoxide data are most frequently collected in time averages of one hour. Evaluating
       compliance with the 1-hour standard simply  requires rank-ordering 1-hour values for a year
       and comparing the second highest value with the 1-hour standard, which is currently
30     40 mg/m3 (35 ppm), not to be exceeded more than once per year.  If the second highest
       1-hour value is less than 40 mg/m3, the standard has been met.

       March 12, 1990                           6-16      DRAFT-DO NOT QUOTE OR CITE

-------
             Evaluating compliance with the 8-hour standard involves the calculation of moving
        8-hour averages from the 1-hour data set.  These 8-hour averages are also rank-ordered to
        obtain the second highest nonoverlapping value for comparison with the 8-hour standard,
        which is currently 10 mg/m3 (9 ppm). For enforcement purposes, only nonoverlapping
 5      8-hour intervals are counted as violations, as discussed in the Guidelines  for the Interpretation
        of Air Quality Standards (U.S. Environmental Protection Agency, 1977c).  It has been
        shown, however,  that the full set of moving 8-hour averages should be examined for
        excessive values.  Proposed simplifications, such as calculating only three consecutive
        nonoverlapping 8-hour averages per day, can easily result in missing peak 8-hour intervals
10      and may not afford equitable comparisons among stations with differing diurnal patterns.

        6.4.4.2 Trend Analyses
             Carbon monoxide ambient concentrations vary considerably from hour to hour, day to
        day, season to season, and year to year.   These variations are usually not random but follow
15      fairly predictable  temporal patterns according to season of the year, day of the week, and
        hour of the day.  Long-term, statistical patterns in CO concentrations are referred to as
        trends. Carbon monoxide trends are best illustrated by graphs that can show diurnal, daily,
        seasonal, or yearly CO concentration comparisons. Examples of the different ways trends can
        be shown are provided later in this section.  Carbon monoxide concentrations also follow
20      fairly predictable  spatial patterns.  Spatial distributions of CO concentrations can be illustrated
        by the use of isopleth maps.

        6.4.4.3 Special Analyses
             A useful  analysis technique is the "pollution rose," as illustrated in  Figure 6-2. The
25      pollution rose presents the joint frequency distribution of wind direction versus ambient CO
        concentration.  The pollution rose is very helpful in determining the wind direction associated
        with the highest ambient CO concentrations and, intuitively, the location of high CO
        emissions sources.
             Another analysis technique is case analysis, which can be used to characterize the
30      meteorological and/or emission conditions associated with observed CO concentrations.  For
        example, in order to characterize the meteorological conditions associated with the occurrence

        March  12, 1990                          6-17       DRAFT-DO NOT QUOTE OR CITE

-------
                                    NORTH
                                                               10%
                                    LEGEND
                                CO CONC, mg/m3

                         0-6.8       6.8-9.1      >91
Figure 6-2. CO pollution rose for St. Louis, MO.
March 12, 1990
6-18      DRAFT-DO NOT QUOTE OR CITE

-------
       of high CO levels, meteorological records can be evaluated for the days when highest CO
       concentrations were observed concurrently at several monitoring sites throughout an urban
       area.  The results of the analysis can then be used to develop a meteorological scenario for
       input to a mathematical model for the purpose of modeling "worst-case" CO concentrations.
 5
       6.4.5  Urban Levels of Carbon Monoxide
            The ambient CO data cited in this document were obtained from EPA's Air Quality and
       Emissions Trend Report (1990) and directly from  the Aerometric Information Retrieval
       System (AIRS, no date).  To be included in the 10-year trend analyses, a given station had  to
10     report data for at least 8 of the 10 years in the period 1979-1988; 248 stations qualified.  For
       the 5-year (1984-1988) trend and urbanized area trend analyses, a station had to report data
       for at least 4 of those five years; 359 stations qualified. The shorter time period was used in
       the urbanized area analyses to expand the number of stations available for analysis.

15     6.4.5.1   Ten-year CO Trends 1979-1988
            Figure 6-3 illustrates the national 1979-1988 composite average trend for the second
       highest nonoverlapping 8-hour CO value for the 248 long-term sites and the subset of
       72 NAMS sites (U.S.  Environmental Protection Agency, 1990). The national average for all
       248 stations and for the NAMS subset of 72 stations decreased both by 28 percent.
20          A Box plot of the data for all stations (Figure 6-4) provides a measure of the distribution
       changes (U.S. Environmental Protection Agency,  1990). Each horizontal line of a Box plot
       represents a percentile value.  Starting at the top,  each line represents the 95th, 90th, 75th,
       50th (median), 25th, 10th, and 5th percentile values.  The composite average is represented
       by an x near the median value. Although certain percentiles fluctuate from year to year,  the
25     general long-term improvement is clear.
            The 10-year trend of the composite average of the estimated number of nonoverlapping
       8-hr CO average concentrations that exceed the 8-hr NAAQS across all stations  is shown in
       Figure 6-5 (U.S. Environmental Protection Agency, 1990).   The trend is clearly decreasing
       with an 88 percent improvement for the 248 long-term stations.  Note that these percentage
30     improvements for exceedances are typically much larger than those found for the second
       March 12, 1990                          6-19      DRAFT-DO NOT QUOTE OR CITE

-------
            CONCENTRATION, PPM
                 I	1	l	1	1	1	1	1	1	1
                1979  1980  1981  1982 1983  1984 1985  1986  1987  1988
Figure 6-3. National trend in the composite average of the second highest nonoverlapping
8-hour average carbon monoxide concentration 1979-1988. Bars show 95 percent confidence
intervals.

Source:  (U.S. Environmental Protection Agency, 1990).
      20
           CONCENTRATION, PPM
       15-
       10-
        5-
                                                            248 SITES
                                                                     MSi
              1979 1980  1981 1982  1983 1984  1985 1986  1987 1988

Figure 6-4. Boxplot comparisons of trends in second highest nonoverlapping 8-hour average
carbon monoxide concentrations at 248 sites, 1979-1988.

Source: U.S. Environmental Protection Agency (1990).
March 12, 1990
6-20     DRAFT-DO NOT QUOTE OR CITE

-------
       20
           EST. 8-HR EXCEEDANCES
       15-
       10-
        5-
• NAMS SfTES (72)     ° AU-_SITESJ[248l
               1979  1980 1981 1982  1983 1984  1985 1986  1987  1988
Figure 6-5. National trend in the composite average of the estimated number of exceedances
of the 8-hour carbon monoxide NAAQS, 1979-1988. Bars show 95 percent confidence
intervals.
Source:  U.S. Environmental Protection Agency (1990).
maximum 8-hour concentrations depicted in Figure 6-4.  The concentrations data are more
likely to reflect percentage change in emission levels.
     National CO emission estimates (Table 6-4) show a 25 percent decrease over the
10-year period (U.S. Environmental Protection Agency, 1990). The predominant CO
emission source, transportation, accounted for about 72 percent of total CO emissions in
March 12, 1990
             6-21
DRAFT-DO NOT QUOTE OR CITE

-------
         TABLE 64. NATIONAL CARBON MONOXIDE EMISSION ESTIMATES, 1979-1988
                                          (million metric tons/year)
10
1979
Source Category
Transportation
Fuel Combustion
Industrial
Processes
Solid Waste
Miscellaneous
Total

59
6.

7.
2.
6.
81

.1
7

1
3
5
.7
1980

56.1
7.4

6.3
2.2
7.6
79.6
1981

55.4
7.7

5.9
2.1
6.4
77.4
1982

52.9
8.2

4.3
2.0
4.9
72.4
1983

52.4
8.2

4.3
1.9
7.7
74.5
1984

50.6
8.3

4.7
1.9
6.3
71.8
1985

47.9
7.4

4.4
2.0
5.3
67.0
1986

44.3
7.5

4.3
1.7
5.0
63.1
1987

43.2
7.6

4.5
1.7
7.1
64.1
1988

41.2
7.6

4.7
1.7
6.0
61.2
15

       Note: The sums of sub-category may not equal total due to rounding.
20     Source:  U.S. Environmental Protection Agency (1990).
       1979, but had decreased to about 67 percent in 1988. This result provides further evidence
25     that the Federal Motor Vehicle Control Program has been effective on a national scale with
       controls more than offsetting the growth during the period. It should be noted that CO
       monitors are typically located to identify potential problems and are placed in areas of high
       traffic densities that may not experience significant increases in traffic.  Thus, CO levels at
       these locations may improve at a faster rate than the nationwide reduction in emissions.
30
       6.4.5.2 Five-year CO Trends 1984-1988
            Evaluation of five-year trends allows use of the expanded data base (359 stations).  The
       composite averages (Figure 6-6) indicate  16 percent improvement between 1984 and  1988.
       Total estimated CO emissions (Table 6-4) decreased 15 percent across the five-year period;
35     emissions from transportation sources decreased 19 percent.
            Composite Regional  averages for 1986 through 1988 of the second highest
       nonoverlapping 8-hour CO averages (Figure 6-7) show the largest declines occurring in
       Regions I, II, VI, VII, and VIII.  Smaller, sometimes fluctuating declines occurred in
       Regions III, IV, V, and IX; Region X ended 1988 on a level with 1986.

       March 12, 1990                         6-22     DRAFT-DO NOT QUOTE  OR CITE

-------
             CONCENTRATION, PPM
                     1984      1985      1986      1987
                     1988
  Figure 6-6.  Boxplot comparisons of trends in second highest nonoverlapping 8-hour average
  carbon monoxide concentrations at 359 sites, 1984-1988.

  Source: U.S. Environmental Protection Agency (1990).
March 12, 1990
6-23     DRAFT-DO NOT QUOTE OR CITE

-------
15
                U
                   CONCENTRATION, PPM
               12-

               10-

                8-

                6-

                4-
                o _

                0
                                 COMPOSITE AVERAGE
                                 E3 1986  • 1987  O 1988
           EPA REGION    I
           NO. OF SITES  15
                      II     III    IV     V    VI    VII    VIII    IX    X
                      27   48    50   49    29   18    17   83    23
Figure 6-7. Regional comparisons of the 1986, 1987, 1988 composite averages of the second
highest nonoverlapping 8-hour average carbon monoxide concentration.
Source: U.S. Environmental Protection Agency (1990).
20
       6.4.5.3  Air Quality Levels in Metropolitan Statistical Areas
            Metropolitan Statistical Areas (MSAs) consist of a central urban county or counties and
       any adjacent counties with at least 50% of their population within the urban perimeter. The
       nation's 339 MSAs,  grouped by population range in Table 6-5, include 77% of the U.S.
 5     population. Figure 6-8 compares the highest second-high nonoverlapping 8-hour value
       recorded during 1988 for the 90 largest MSAs in the continental U.S. (not shown: Honolulu,
       HI, and San Juan, PR), containing approximately 55% of the U.S. population. Nineteen of
       these MSAs exceeded the current 8-hour standard of 9 ppm in 1988.

10     6.4.6  Effects of  Meteorology and Topography
            Meteorology governs the transport and dispersion of CO emissions in the atmosphere
       and thus has a strong influence on the ground level CO concentrations detected at receptor
       points downwind of emission sources.  Meteorological parameters that determine CO
       March 12, 1990
                                        6-24
DRAFT-DO NOT QUOTE OR CITE

-------
         TABLE 6-5.  DISTRIBUTION OF POPULATION IN METROPOLITAN STATISTICAL
                                  AREAS (Based on 1987 estimates)
10
Population Range
< 100,000
100,000 < 250,000
250,000 < 500,000
500,000 < 1,000,000
1,000,001 < 2,000,000
> 2,000,000
Total
No. of MSAs
27
147
73
48
26
18
339
Total Population
2,274,000
23^,372,000
25,218,000
34,367,000
38,685,000
65,747,000
189,663,000
15
       Source: U.S. Environmental Protection Agency (1990).
20
       transport and dispersion patterns include wind speed, wind direction, atmospheric stability,
       and mixing depth.  The relative importance of each parameter depends upon the scale of the
       analysis. For example, concentration patterns around an intersection would not be greatly
25     influenced by mixing depth.  However, concentration patterns over the whole urban area
       would be.
            Wind direction determines the direction of horizontal transport of CO emissions and
       consequently the impact that CO emissions from one are will have on air quality  in another
       area.  If emissions are uniform across the urban area, as air flows across the whole urban
30     area, the additive effect of the CO emissions being transported downwind will result in higher
       CO concentrations at the downwind edge of the urban area.  However, concentrations  at a
       particular location may be dominated by local emissions. Concentrations adjacent to a
       highway will be higher than urban levels away from large highways.  Wind directions nearly
       parallel to a highway will allow for an accumulation of CO emissions in the downwind
35     direction, resulting in CO concentrations higher than would be expected for winds
       perpendicular to the highway under the same conditions.
            Low wind speeds provide little atmospheric dilution, allowing CO emissions to build up,
       resulting in higher CO concentrations.  Conversely, high wind speeds aid in the dispersion of

       March 12, 1990                          6-25       DRAFT-DO NOT QUOTE OR CITE

-------

-------
        CO emissions by increasing the amount of dilution that takes place, thus decreasing CO
        concentrations.  The effect of surface roughness (i.e., mountains, buildings, etc.) on the wind
        speed profile over several types of topographic features is illustrated in Figure 6-9 (Benson,
        1979). With increased surface roughness, either natural or man-made, the depth of the
 5      affected layer is increased.  The winds affected by frictional drag are reduced, but the
        turbulence induced by mechanical effects in increased. The net effect of increased surface
        roughness over an urban area is to mix the CO emissions through a larger depth in the
        atmosphere, which aids in the dispersion of CO emissions. Thermal forces in the atmosphere
        either enhance or suppress the production of turbulent motion in the atmosphere.  The
10      dispersive properties of the atmosphere are correlated with atmospheric stability, which is
        generally easier to characterized.
             Radiation and thermal properties of topographic features influence the heating and
        cooling of the atmosphere near the ground surface. The most notable of these effects is the
        urban "heat island" effect. Heat sources, including the asphalt and concrete associated with
15      an urban area, tend to radiate heat, causing a "heat island" compared to the cooler
        surrounding terrain. The buoyant effect of warmer air over the city tends to induce thermal
        turbulence (i.e., more unstable atmospheric conditions), which tend to aid in the dispersion of
        CO emissions, thus lowering ambient CO concentrations.
             Local wind circulations, such as sea-land breezes, lake-land breezes, or mountain-valley
20      winds, are caused by the differential heating of topographic forms.  These circulations
        generally flow in one direction during the day,  and in the opposite direction at night.  As a
        result, an urban area can experience "blow-back" of CO emissions emitted during the day;
        these will be experienced as higher CO concentrations at night.   The boundary region between
        the local circulation winds and the prevailing synoptic flow sometimes remains nearly
25      stationary, or slowly oscillates back and forth for periods up to several hours, and can be the
        site of nearly calmwind conditions. These characteristics result in slow  net transport of CO
        emissions which then accumulate and result in higher ambient CO concentrations.
             The depth through which pollutants  are routinely mixed affects the total ventilation
        capacity of the atmosphere.  When the potential temperature lapse rate is positive, the
30      resulting increase in temperature with increase in height procedures an inversion or inversion
        lid that limits vertical mixing, and thereby limits the dilution capacity of the atmosphere.

        March 12, 1990                           6-27      DRAFT-DO NOT QUOTE OR CITE

-------
to
QO
8
O
                     600
                     500  -
                     400
                  t  300  -
                  o

                  Ul
                  X
                     200  -
                     100  -
                                                     to
                                                                     0       6



                                                                        WIND SPEED, m/uc
                                       Figure 6-9.  Effect of terrain roughness on the wind speed profile.


                                       Source:  Benson (1979).

-------
            An important form of inversion for CO dispersion is the surface or radiation inversion.
       This usually occurs at night with light winds and clear skies, when the loss of heat by
       longwave radiation from the ground surface cools the surface and subsequently the air
       adjacent to it.  With the proper relative humidity, these same conditions will lead to the
 5     formation of radiation fog.  The presence of early morning fog is often associated with a
       surface based temperature inversion. The surface inversion usually persists for hours, and
       because it typifies stable atmospheric conditions, it tends  to result in high microscale and
       mesoscale CO concentrations.
            Another type of inversion is the subsidence inversion.  It is caused by a gradual descent
10     of air aloft, that results in adiabatic wanning of the descending layer. The resulting
       subsidence inversion is illustrated in Figure 6-10, which shows the temperature decreased with
       height and  the capping by a subsidence inversion layer, above which there is a normal
       decrease of temperature with height. The subsidence inversion usually persists for days and
       tends to contribute to high urban background CO concentrations.  The subsidence inversion is
15     usually more persistent during summer and fall than in winter or spring.
            The shape of typical plots of hourly CO concentrations can be attributed in large part to
       the effect of changing wind speeds, atmospheric stability, and inversion height during the
       course of a day. Figure 6-11 shows average hourly wind speeds and inversion heights
       occurring in Los Angeles during summer (Tiao et al.,  1975). The higher wind speeds and
20     inversion height during early afternoon are typical throughout the continental United States
       and play a significant role in lowering urban CO concentrations at midday.  Traffic volumes,
       and subsequently CO emissions from cars, would still be expected to be high at this time of
       day.  Around midnight, when traffic volumes are relatively low, the effects of low wind
       speeds and low inversion heights tend to cause  increases in CO concentrations.  Many
25     monitoring stations in the United Sates observe these relatively high CO concentrations late at
       night.
            Ambient surface temperature also has a unique effect on the production rate of CO
       emissions from automobiles.  Using a variety of automobiles tested at artificially controlled
       ambient temperatures of 20°, 50°,  75°, and 110°F, the EPA (Bullin et al., 1986) found that
30     lowest CO emissions were produced at 75 °F and tend  to increase with colder temperatures.
       Colder temperatures coupled with a strong surface based radiative inversion are generally

       March 12, 1990                           6-29      DRAFT-DO NOT QUOTE OR CITE

-------
               1000
            9)
            **
            a




            H

            I

            o
               500
        2nd MIXING LAYER   —
         INVERSION "LID"
                                             lit MIXING LAYER  —
                             I   I   I    I   I   \  I  " I   I
                  0                10                20



                                  TEMPERATURE, °C




            Figure 6-10. Schematic representation of an elevated inversion.
March 12, 1990
6-30     DRAFT-DO NOT QUOTE OR CITE

-------
  10
£
a



a"
UJ
Ul

8s
a

2  5

i
         I      I     I      I       I      I     I
                                         riii
                                                               WIND SPEED
J	I	\	L
                                                             J	L
                                      12
                                                       18
                                                                        24
                                         TtME.hourt
                                                                                    20
                                                                         o

                                                                         iij
                                                                         X
                                                                      15
     Figure 6-11.  Hourly variations in inversion height and wind speed for Los Angeles in

     summer.



     Source: Tiao et al. (1975).
   March 12, 1990
                            6-31      DRAFT-DO NOT QUOTE OR CITE

-------
30
       associated with poor dispersion in the atmosphere.  The combined effect of higher emission
       rates and poor dispersive conditions results in higher ambient CO concentrations than would
       be expected for warmer temperatures.
       6.5  CARBON MONOXIDE DISPERSION MODELS
            A dispersion model relates pollutant emissions to ambient air quality by providing a
       mathematical description of the transport, dispersion, and chemical transformations that occur
       in the atmosphere.  This ability to relate source emissions to receptor air quality is very
10     important to air quality maintenance planning and environmental impact assessment.
            Dispersion models vary in complexity from simple empirical or statistical relationships
       to sophisticated multi-source models that describe the transport and dispersion of CO
       throughout an urban area. For estimates of ambient CO concentrations, a line source model
       is needed to estimate the CO levels near a highway, an intersection model is needed to
15     estimate CO levels near an intersection, and an urban model is needed  to estimate CO levels
       that result from the cumulative effects of urban sources such as roadways and wood stoves.
       The types of models used will depend mainly on  the source configuration to be modeled (i.e.,
       highway intersection, or urban area).

20     6.5.1  Line Source Modeling
            Several models have been used to estimate CO concentrations from line sources. The
       guideline on Air Quality Models (Revised) (U.S. Environmental Protection Agency, 1986)
       makes specific recommendations on procedures to utilize for line source modeling.  Refer to
       the latest version of this document for these recommendations.  Available line source models
25     are CALINE-3 (Benson, 1979), GMLINE (Chock, 1978), HIWAY-2 (Petersen,  1980), and
       PAL (Petersen, 1978).  A brief description of each of the line source models excerpted from
       "Evaluation of Mobile Source Air Quality Simulation Models" (Wackter and Bodner, 1986)
       follows:
       March 12,  1990                         6-32      DRAFT-DO NOT QUOTE OR CITE

-------
       6.5.1.1  CALINE3
            The CALINE3 model was developed by the California Department of Transportation. It
       simulates dispersion of highway emissions by dividing individual roadway links into a series
       of elements from which incremental concentrations are computed using a finite line source
 5     equation. The incremental concentrations are summed to obtain a total concentration estimate
       at a particular receptor location.
            CALINE3  simulates the region directly over the roadway as a zone of uniform
       emissions and turbulence called the "mixing zone."  This zone experiences increased
       dispersion due to mechanical turbulence created by moving vehicles as well as thermal
10     turbulence created by hot vehicle exhaust.  CALINE3 adjusts the level of turbulence as a
       function of wind speed.  At low wind speeds residence time of an air parcel within the mixing
       zone is increased, resulting in turbulence enhancement through the use of a larger initial
       vertical sigma value.
            The CALINE3 model includes options for simulating dispersion  from four types of
15     roadways: at grade, elevated filled sections, elevated bridges, and cut or depressed sections.
       Multiple lanes, links and orientations can be simulated.

       6.5.1.2  GMLINE
            GMLINE (Chock, 1978) was developed by General Motors Research Laboratories to
20     describe dispersion near straight-line, at-grade highways.  Multiple parallel or crossing
       roadway links can be simulated and the model allows for a variable emissions height.  The
       model  was not designed to treat cut-sections.
            GMLINE simulates dispersion of vehicle emissions by dividing the roadway into
       separate, straight-line sources, each with a uniform emission rate.  Downwind concentrations
25     at a receptor are calculated for each infinite line source, then summed  to obtain a total
       concentration. The model accounts for plume rise due to heated exhaust and includes a wind
       speed correction  to account for increased turbulence created by traffic  wakes.

       6.5.1.3  HIWAY-2
30          HIWAY-2  (Petersen, 1980) was developed by EPA to replace the HIWAY model
       (Zimmerman and Thompson, 1975) for estimating roadway pollutant impacts. The model

       March 12, 1990                          6-33     DRAFT-DO NOT QUOTE OR CITE

-------
       was designed to determine concentrations at receptors downwind of at-grade roadways and cut
       sections (outside of the cut only).
            HIWAY-2 simulates dispersion by treating highway emissions as a series of finite line
       sources, each with a uniform emission rate. Concentrations downwind are calculated by
 5     numerically integrating a Gaussian point-source plume along each line segment.  The primary
       differences between HIWAY-2 and HIWAY are that HIWAY-2 includes a new set of
       dispersion curves and an aerodynamic drag factor to account for dispersion due to vehicle
       motion under low wind speed conditions.

10     6.5.1.4 PAL
            The PAL model (Petersen, 1978)  was developed by EPA to estimate pollutant dispersion
       from point, area and line sources.  It was designed to simulate dispersion  from several types
       of roadway geometries including straight or curved horizontal lines and straight or  curved
       elevated lines with variable emissions along each line segment.  Model documentation
15     specifies that treatment of elevated line sources is appropriate for open bridge type road
       segments but not for elevated filled roadways.  Cut or depressed roadway sections  are not
       treated by PAL.
            PAL determines concentrations at a receptor due to a line source by  numerically
       integrating the Gaussian point source equation.  Calculations are made for a number of points
20     along the finite line, assuming a linear change in concentration between these points.
       Subsequent estimates of concentrations  are made by including additional points along the line.
       When the difference between succeeding estimates becomes smaller than a prescribed value,
       the calculations are considered complete.

25     6.5.1.5 Model Evaluation
            A comprehensive evaluation of CALINE3, GMLINE, HIWAY-2, and PAL was
       undertaken using 5 field measurement programs and is described in "Evaluation of Mobile
       Source Air Quality Simulation Models" (Wackier and Bodner, 1986).  This  report contains
       numerous tabulations of each model's performance in terms of statistical measures
30     recommended by the American Meteorological Society.  The results indicate that the
       GMLINE model performed the best most often, while the PAL model ranked lowest most

       March 12, 1990                          6-34       DRAFT-DO NOT  QUOTE OR CITE

-------
       frequently.  All the models tended to overpredict for light wind speeds and near parallel
       wind/road angles, while underpredictions occurred for high wind speeds.

       6.5.2  Intersection Modeling
 5          Several models have been used to estimate CO concentrations from intersections. The
       Guideline on Air Quality Models (Revised) (U.S.  Environmental Protection Agency, 1986)
       makes specific recommendations on procedures to utilize for intersection modeling. Refer to
       the latest version of this document for these recommendations.  Available intersection models
       are: "Volume 9" (U.S.  Environmental Protection Agency, 1978), CAL3Q (Smith, 1985),
10     CALINE4 (Benson, 1984), GIM (EMI Consultants,  1985), IMM (New York State
       Department of Transportation, 1980), and TEXIN2 (Bullin et al., 1986) a brief description of
       each of these models excerpted from the above references follows.

       6.5.2.1 "Volume 9"
15          Carbon monoxide concentrations are calculated in a three-step process.  In the first step,
       the network description and traffic demand volume are used to estimate the traffic flow
       characteristics.  Emissions are then computed as the sum of two parts: cruise emissions
       produced by non-stopping vehicles and excess emissions emitted by stopping/starting vehicles.
       Lastly, the effect of atmospheric dispersion on actual concentrations at the specified receptor
20     locations is estimated.
            Excess emissions consist of deceleration, idle, and acceleration emissions due to vehicles
       stopping and starting at intersections.  Idle emissions rates are determined using MOBILES.
       Acceleration, deceleration, and cruise emission rates are determined using modal emission
       factors based on the updated (December 1977) version of the Modal Emissions Model
25     (Kunselman et al., 1974). MOBILES correction factors (Wolcott, 1986) to the modal
       emission factors can then be utilized to adjust for calendar year, cold starts, hot starts, speed,
       temperature, and vehicle mix.
            The traffic model contained in "Volume 9" calculates the  length over which excess
       emissions apply.  This calculation is based on  the proportion of vehicles that stop and the
30     number of vehicles subject to queueing delay.  It should be noted that the traffic model
       March 12, 1990                          6-35      DRAFT-DO NOT QUOTE OR CITE

-------
        contained in "Volume 9" is not applicable for overcapacity intersections thus, "Volume 9"
        cannot be utilized for such intersections scenarios.

        6.5.2.2  Intersection Midblock Model
 5           The Intersection Midblock Model (IMM) is a combination of signalization and vehicle
        queueing estimation procedures using accepted traffic engineering principles.  It also predicts
        emissions using the Modal Analysis Model and the MOBILE-2 program, and models
        dispersion with the HIWAY-2 model.
             The IMM first calculates various traffic parameters. Once the traffic calculations have
10      been performed, the estimation of emission rates is carried out.  Using the input parameters
        of speed into the queue, speed out of the queue, deceleration into the queue and acceleration
        out of the queue, the IMM utilizes the Modal Analysis Model as a subroutine to calculate
        cruise and acceleration/deceleration emissions for all approaches. Idle emissions are
        calculated by use of the MOBILE-2 program.  Based on the previously calculated queue
15      lengths,  a set of pseudolinks is constructed.  These pseudolinks lie along the actual links with
        the same termination points and center lines as the actual links, but each has a length equal to
        the calculated queue length for that approach.  The only emissions assigned to the actual links
        are the cruise emissions (calculated with the Modal Analysis Model).  The emissions assigned
        to the pseudolinks are the excess emissions due to accelerating, decelerating and idling.
20           A correction factor is applied to the emissions calculated from the Modal Analysis
        Model since these apply only for 1977 emission rates from stabilized light-duty vehicles.  The
        correction factor used is the ratio of the MOBILE-2 composite emission estimate for the
        specified scenario to the MOBILE-2 composite emission estimate for 1977 stabilized light-
        duty vehicles.
25           Once the traffic calculations have been performed and emission rates assigned to each
        lane, the HIWAY-2 model is employed as a  subroutine to calculate carbon monoxide
        concentrations at selected receptors. For the special case of a "street canyon" intersection
        between tall buildings in a highly urban area, a special dispersion routine is used.
30
        March 12, 1990                           6-36      DRAFT-DO NOT QUOTE OR CITE

-------
       6.5.2.3  Georgia Intersection Model
            The Georgia Intersection Model (GIM) uses a computer program to calculate the
       average vehicle delay, the average route speed, and the emission rate of carbon monoxide of
       vehicles traveling through the intersection over a distance called the "effective length" where
 5     speeds are lower due to the effect of vehicles slowing and stopping during red light cycles.  It
       eliminates the need for using modal emission factors  and allows for the analysis of
       overcapacity intersections.
            GIM uses many of the same assumptions and equations as in the  "Volume 9" approach,
       with few modifications. The procedure can be summarized briefly as follows.  GIM
10     calculates the effected length of roadway upstream of the intersection where vehicle speeds
       are reduced due to delays caused by vehicle slowing  and stopping.  It calculates the  CO
       emissions for vehicles traversing the effected length,  based on average speed over the length
       and MOBILES emission factors. Using this approach,  modal emission factors (i.e., accelera-
       tion mode emissions, deceleration mode emissions, idle emissions,  and cruise emissions) are
15     not utilized.  The output of GM defines finite line source segments with their associated CO
       emission rates which can be input to the CALENE3 line source dispersion model.

       6.5.2.4 TEXIN2
            The TEXIN2 Model follows a general three-step process:
20
            (1)   Estimation of traffic parameters.
            (2)   Estimation and distribution of vehicle emissions.
25          (3)   Modeling downwind dispersion of pollutants.


            Traffic parameters are calculated using either the modified Planning or Operations and
30     Design procedures of the Critical Movement Analysis (CMA) (National Cooperative Highway
       Research Program, 1979) for signalized intersections.  Basically, the difference between the
       two traffic algorithms concerns the different adjustment factors present in the CMA
       Operations and Design algorithm.  These adjustment factors tend to decreased the capacity of
       a given intersection. Therefore, the Operations and  Design technique  will occasionally

       March 12, 1990                          6-37      DRAFT-DO NOT QUOTE OR CITE

-------
        calculate that an intersection is over capacity while the Planning procedure indicates that the
        intersection is below full capacity.
             Research has provided adjustment factors for a number of elements that affect traffic
        flow and hence modify critical volumes. These elements are:  (1) left turns; (2) bus and
 5      truck volume; (3) peaking characteristics; (4) lane width; (5) bus stop operations; (6) right
        turns with pedestrian activity; and (7) parking activity. In the TEXIN2 Model, the CMA
        Planning procedure utilizes only the left turn adjustment factor, while the CMA Operations
        and Design procedure uses the first four adjustment factors listed above with no additional
        user input.  In both algorithms, left turns are treated in detail for the simple reason that left
 10      turns have a large impact on intersection capacity. This effect is created using passenger car
        equivalency (PCE) values.  PCE values are multiplicative adjustment  factors applied to the
        left turning traffic volumes.
             The second function performed by TEXIN2 is the estimation of vehicle emissions. The
        emissions are modeled as the sum of two components: cruise and excess emissions. Cruise
 15      emissions and excess emissions are released by free-flowing and delayed vehicles,
        respectively. Initially, cruise emissions are assumed to be released along the entire length of
        each intersection leg.  The emissions are subsequently redistributed to better reflect actual
        traffic movement.  A modified version of the MOBILES program is used to estimate cruise
        emissions and an idle emission factor, while acceleration and deceleration emissions are
20      calculated using modal emission factors as suggested by Ismart (1981).  As an alternative, a
        shortcut method combining the MOBILE3 estimation of the idle emission factor with values
        for individual vehicle emission rates based on speed, temperature, percent hot/cold starts, and
        the vehicle scenario is available to the user (Federal Highway Administration, 19  ).
             As  used in  TEXIN2,  the MOBILES program provides inspection/maintenance (I/M) and
25      antitampering program (ATP) options.   To conserve computer time, several sizable portions
        of the extremely large MOBILES program were deleted, namely the nitrogen oxide and
        hydrocarbon emission factors modeling and user- supplied corrections to the emission rates.
        Since the MOBILE-2 program does not allow for California scenarios, the California data and
        options from the MOBILE-2 program were added to the emission routine.
30
       March 12, 1990                           6-38     DRAFT-DO NOT QUOTE OR CITE

-------
       6.5.2.5  CAL3Q
            The CAL3Q model utilizes the Connecticut Department of Transportation queueing
       model to calculate traffic parameters including queue length. The average speed of vehicles
       through the intersection is estimated so a composite MOBILES emission factor can be applied
 5     over the length of the queue.  In addition, the MOBILES idle emission rate is applied over
       the queue length.  No modal emission factors are utilized in CAL3Q and the model cannot
       handle overcapacity intersections.  The emissions, and queue length are input to the
       CALINE3 dispersion model to calculate carbon  monoxide concentrations at selected receptors.

10     6.5.2.6  CALINE4
            The CALINE4 intersection model focuses  on a rather complex concept of spatially
       resolved modal emissions over links. A  CALINE4 intersection link encompasses the
       acceleration and deceleration zones created by the presence of the intersection.   Each link can
       treat only one direction of traffic flow, so that four links are required to model a full
15     intersection.
            Four cumulative modal emission profiles representing the deceleration, idle, acceleration
       and cruise modes of operation are constructed for each intersection link.  These profiles are
       determined using the following input variables:
20          SP      =    Cruise speed (mph)
            ACCT  =    Acceleration time (seconds)
            DCLT  =    Deceleration time (seconds)
            IDT1   =    Maximum idle time (seconds)
            IDT2   =    Minimum idle time (seconds)
25          NCYC  =    Total number of vehicles per cycle per lane
            NDLA  =    Number of vehicles delayed per cycle per lane.
30          NCYC and NDLA are chosen to represent the dominant movement for the link.  The
       model assumes a uniform vehicle arrival rate, constant acceleration and deceleration rates, and
       full stops for all delayed vehicles.  Acceleration and deceleration rates (ACCR, DCLR) and
       acceleration and deceleration lengths (LACC, LDCL) are determined using the input values
       for SPD, ACCT and DCLT. By assuming an "at rest" vehicle spacing (VSP) of 7 meters,
35     the average queue length (LQU) is also determined. IDT1 represents the delay at full stop
       March 12,  1990                          6-39      DRAFT-DO NOT QUOTE OR CITE

-------
       experienced by the first vehicle in the queue. Similarly, IDT2 represents this same measure
       for the last vehicle.  IDT2 is used to model a platooned arrival and should be assigned a value
       of zero for non-platooned applications.
            The time rate modal emission factors over the link are computed by a rather complex
 5     method. To develop these factors, the model must be provided with composite emission rates
       for average route speeds of 0 (idle) and 16 mph. The resulting time rate factors are denoted
       as EFA (acceleration), EFD (deceleration), EFC (cruise) and EFI (idle).
            The cumulative emission profile for a given mode is developed by determining the time
       in mode for each vehicle as a function of distance from link endpoint 1 (ZD), multiplying the
10     time by the respective modal emission rate and summing the results over the number of
       vehicles per cycle per lane (NCYC).  The elementary equations of motion are used to relate
       time to ZD for each mode.  The assumed vehicle spacing (VSP) is used to specify the
       positional  distribution of the vehicles.  The total cumulative emissions per cycle per lane at
       distance ZD from XLl, YLl are denoted as ECUMk(ZD) in the CALINE4 coding, where the
15     subscript signifies the mode (1 =accel., 2=decel., 3=cruise, 4=idle).
            The CALINE4 model handles atmospheric dispersion, somewhat similarly to that in the
       CALINE3 model. The most significant difference is that CALINE4, unlike CALINE3,
       requires the input of site specific wind direction, fluctuation, and sigma-theta data.

20     6.5.2.7 Comparison of Intersection Models
            A study was conducted by Braverman (1987) to compare the emission rates and
       distances over which acceleration, cruise, deceleration,  and idle emissions occur as generated
       by different intersection models. The models included in the study were CAL3Q,  GIM,
       IMM, TEXIN2, and "Volume 9".  CALINE4 was not included in this study because it does
25     not contain an explicit traffic model and does not print out emission rates.
            In this study, scenarios for a simple undercapacity, near capacity, and overcapacity
       intersection were modeled.  All of the parameters for each of the intersection scenarios
       modeled were held constant except for the approach volumes which were 408,  608, and 768
       vehicles/hour for the under capacity,  near capacity, and overcapacity scenarios, respectively.
30     The parameters for the intersection scenarios modeled are given in Table 6-6.
       March 12,  1990                         6-40      DRAFT-DO NOT QUOTE OR CITE

-------
             The results of the emission rates and distances generated by the models are given in
        Tables 6-7, 6-8, and 6-9 for the undercapacity, near capacity, and overcapacity intersection
        scenarios, respectively. The traffic models contained in "Volume 9" and CAL3Q cannot
        handle overcapacity intersections, so no results are reported for these models for the
        overcapacity scenario.  In addition, the traffic model contained in CAL3Q calculates the near
        capacity scenario as slightly overcapacity, so no results are reported for CAL3Q for the near
        capacity scenario.
                   TABLE 6-6.  PARAMETERS FOR INTERSECTION SCENARIOS
 10     	

        Approach Volumes (Vehicles/Hour)                             480*, 608*, and 768*
        Approach Capacity (Vehicles/Hour)                             640
        Capacity Service Volume (Vehicles/Hour of Green Time)          1600
 15     Cycle Length (Seconds)                                        90
        Percent Green Time                                           40
        Uninterrupted Speed (Miles/Hour)                              30
        Number of Lanes                                             1
        Temperature (°F)                                             40
 20     Year                                                        1985
        Percent Cold Non-Catalyst                                     20.6
        Percent Hot Catalyst                                           27.3
        Percent Cold Catalyst                                          20.6
        Altitude                                                     Low
 25     Vehicle Mix Fractions

           Light Duty Gas Vehicle                                    0.652
           Light Duty Gas Truck 1                                    0.128
           Light Duty Gas Truck 2                                    0.087
 30        Heavy Duty Gas Vehicle                                    0.040
           Light Duty Diesel Vehicle                                  0.023
           Light Duty Diesel Truck                                    0.008
           Heavy Duty Diesel Vehicle                                 0.054
           Motorcycle                                               0.007
 35      	
40
       *Undercapacity, near capacity, and overcapacity, respectively.

       Source: Braverman (1987).
       March 12, 1990                          6-41      DRAFT-DO NOT QUOTE OR CITE

-------
10
         TABLE 6-7.  RESULTS OF MODEL COMPARISONS FOR THE UNDERCAPACITY
                                   INTERSECTION SCENARIO
                         Emission Rate (g/m-s)
        Model              Unit Distance           Distance (m)         Emission Rate (g/s)
Volume 9
IMM
TEXIN2
GIM
CAL3Q
0.0327
0.0119*
0.0236*
0.0102
0.0218
106.3
220.1
81.7
162.1
54.0
3.476
2.615
1.930
1.653
1.177
15
       "'Emission Rate was obtained by multiplying each component (acceleration, United Distance deceleration, and
        idle) of emissions times the length over which they occur individually, summing the products, and dividing by
        the entire length of emissions.

20     Source:  Braverman (1987).


          TABLE 6-8. RESULTS OF MODEL COMPARISONS FOR THE NEAR CAPACITY
                                   INTERSECTION SCENARIO
25     	;	

                         Emission Rate (g/m-s)
         Model              Unit Distance           Distance (m)         Emission Rate (g/s)

30
Volume 9
IMM
TEXIN2
GIM
0.0452
0.0229*
0.0413*
0.0124
269.7
294
141.5
193.4
12.19
6.74
5.85
2.40
35

       *Emission Rate was obtained by multiplying each component (acceleration, United Distance deceleration, and
        idle) of emissions times the length over which they occur individually, summing the products, and dividing by
        the entire length of emissions.
40

       Source: Braverman (1987).
       March 12, 1990                         6-42      DRAFT-DO NOT QUOTE OR CITE

-------
           TABLE 6-9. RESULTS OF MODEL COMPARISONS FOR THE OVERCAPACITY
                                    INTERSECTION SCENARIO
                         Emission Rate (g/m-s)
         Model              Unit Distance           Distance (m)         Emission Rate (g/s)
IMM
GIM
TEXIN2
0.0413*
0.0228
0.0426*
611
733.3
176.6
25.24
16.7
7.53
 10
        *Emission Rate was obtained by multiplying each component (acceleration, United Distance deceleration, and
15      idle) of emissions times the length over which they occur individually, summing the products, and dividing by
        the entire length of emissions.
        Source: Braverman (1987).
20
            The results indicate that under- to near-capacity situations the order of model predicted
        emission rates  from highest to lowest is "Volume 9", IMM, TEXIN2, GIM, CAL3Q. For
        overcapacity situations, the order of model predicted emission rates from highest to lowest is
        IMM,  GIM, TEXIN2. Note that "Volume 9", IMM, and TEXIN2 utilize modal emission
25      factors.  GIM and CAL3Q do not utilize modal emission factors; instead they utilize
        composite MOBILES emission factors.  Since "Volume 9", IMM, and TEXIN2 give higher
        emission rates  for the under to near capacity situations than GIM and CAL3Q, it can be
        inferred that models that utilize modal emission factors generate higher emission rates than
        models that utilize composite MOBILES emission factors. The reason that GIM gives higher
30      emissions than TEXIN2 for the overcapacity scenario is that TEXIN2 simply extrapolates
        from the undercapacity situation to generate overcapacity emission rates, whereas GIM treats
        overcapacity as an entirely different situation.

        6.5.3  Urban Area Modeling
35          Several models have been used to model the cumulative effects of urban CO sources
        such as roadways and wood stoves. The Guideline on Air Quality Models (Revised) (U.S.
       Environmental Protection Agency, 1986) does not recommend one specific model for
       March 12, 1990                          6-43     DRAFT-DO NOT QUOTE OR CITE

-------
        urbanwide CO analysis; instead, it recommends these analyses be considered on a case-by-
        case basis.
             Urban exceedances of the 8-hour CO NAAQS result primarily from the cumulative
        effects of motor vehicle emissions throughout the urban area.  The APRAC-3 model (Simmon
 5      et al., 1981), which is briefly described below, was developed to handle this situation.

        6.5,3.1  APRAC-3
             APRAC-3 is a Gaussian-plume diffusion model which computes hourly average carbon
        monoxide concentrations for any urban location.  The model calculates contributions from
10      dispersion on various scales: extraurban, mainly from sources upwind of the city of interest;
        intraurban, from freeway, arterial, and feeder street sources; and local,  from dispersion
        within a street canyon.  APRAC-3 requires an extensive traffic inventory for the city of
        interest.
             Traffic links may have arbitrary length and orientation. Off-link traffic is allocated to
15      two-mile square grids. Link traffic emissions are aggregated into a receptor oriented area
        source array.  The boundaries of the area sources actually treated are (1) arcs at radial
        distances from the receptor which increase in geometric progression, (2) the sides of a 22.5°
        sector oriented upwind for distances greater than 1,000 m,  and (3) the sides of a 45° sector
        oriented upwind for distances less than 1,000 m. A similar area source array is established
20      for each receptor.  Up to 625 receptors are accepted for a single-hour.
             Meteorological data requirements are hourly wind direction (nearest 10 degrees), hourly
        wind speed, and hourly cloud cover for stability calculations.  Constant, uniform (steady-
        state) wind is assumed within each hour.  The model can interpolate winds  at receptors if
        more than one wind is provided.  Mixing height is  ignored until the concentration equals that
25      calculated using a box model. A box model the (uniform vertical distribution) is used beyond
        that distance.
             A secondary contributor to some urban exceedances of the 8-hour  CO NAAQS is the
        cumulative effect of wood stove emissions throughout the urban area. These emissions are
        trapped under the nighttime radiation inversion along with evening traffic emission on cold
30      clear nights with light and variable winds. This situation can be handled best by either a
        Gaussian or a numerical model.  A numerical model provides a better treatment than a

        March 12, 1990                           6-44      DRAFT-DO NOT QUOTE OR CITE

-------
       Gaussian model of the time dependent changes in meteorology under these conditions.
       However, use of numerical models is extremely data and resource intensive. Thus, numerical
       models have only been applied in very large cities. One numerical model that has been used
       in a few cases is the Urban Airshed Model (Ames et ah, 1985).  An acceptable Gaussian
 5     model for urban area applications is RAM (Catalano et ah, 1987).  A brief description of the
       Urban Airshed Model and RAM follows.

       6.5.3.2 Urban Airshed Model
            The Urban Airshed Model (Ames et ah,  1985) simulates the major physical and
10     chemical processes in the polluted troposphere. These include gas phase chemistry, advective
       transport, and turbulent diffusion.  The modeling domain is divided into a large array of grid
       cells. Horizontally the cells are uniformly sized squares 3 to 5 kilometers on a side.
       Typically, four or five layers of cells represent the vertical domain.  The depth of the layers
       is scaled by the height of the mixed layer and the  height of the top of the modeling domain
15     (region top). The latter typically ranges from  500 meters in the morning hours to
       1,000 meters or more in the afternoon.  Emissions are injected into individual cells depending
       on the location of the sources, their height of release, and the buoyant rise of individual stack
       gas plumes.
            The theoretical basis for the Urban  Airshed Model rests of the conservation of mass
20     equation for atmosphere diffusion.  Primary inputs to the Urban Airshed Model are point and
       area source emissions, initial and boundary concentrations both at the surface and aloft, and a
       variety of meteorological data.   These include a three-dimensional wind field, mixing depths,
       surface temperature, and exposure class,  the latter an indicator of thermal instability.

25     6.5.3.3 RAM
            RAM (Catalano et ah,  1987) provides a  readily available computer program based on
       the assumptions of steady-state Gaussian dispersion for short-term (one-hour to one-day)
       determination  of urban air quality resulting from pollutants released from point and/or area
       sources.
30          RAM is applicable for locations with level or gently rolling terrain where a single wind
       vector for each hour is a reasonable approximation of the flow over the source area

       March 12, 1990                          6-45     DRAFT-DO NOT QUOTE OR  CITE

-------
        considered.  Calculations are performed for each hour.  Hourly meteorological data required
        are wind direction, wind speed, temperature,  stability class, and mixing height.
             Computations are performed hour by hour as if the atmosphere had achieved a steady-
        state condition. Therefore, errors will occur where there is a gradual buildup (or decrease) in
 5      concentrations from hour to hour, such as with light wind conditions.  Also under light wind
        conditions the definition of wind direction is likely to be inaccurate, and variations in the
        wind flow from location to location in the area are quite probable.
             Considerable time is saved in calculating concentrations from area sources by using a
        narrow plume simplification which considers sources at various distances on a line directly
10      upwind from the receptor to be representative in the crosswind direction of the sources at
        those distances affecting the receptor.  Area source  sizes are used as given in the emission
        inventory in lieu of creating an internal inventory of uniform elements.
15     6.6  REFERENCES
       AIRS, Aerometric Information Retrieval System [data base], (n. d.) [Standard computer retrievals]. Unpublished
20           computer reports available from: U. S. Environmnetal Protection Agency, Office of Air Quality Planning
             and Standards, Research Triangle Park, NC 27711.

       Altshuller, A. P.; Ortman, G. C.; Saltzman, B. E.; Neligan, R. E. (1966) Continuous monitoring of methane and
             other hydrocarbons in urban atmospheres. J. Air Pollut. Control Assoc. 16: 87-91.
25
       Ames, J.; Myers, T. C.; Reid, L. E.; Whitney, D. C.; Golding, S. H. (1985) SAI Airshed Model operations
             manuals. Volume I-user's manual. Research Triangle Park, NC: U. S. Environmental Protection
             Agency, Atmospheric Sciences Research Laboratory; EPA  report no. EPA-600/8-85-007A. Available
             from:  NTIS, Springfield,  VA; PB85-191567.
30
       Anonymous. (1976) Report of the air monitoring siting workshop.  Las Vegas,  NV.

       Bach, W. D.;  Crissman, B. W.; Decker, C. E.; Minear, J. W.; Rasberry, P. P.; Tommerdahl, J. B. (1973)
             Carbon monoxide measurements in the vicinity of sports stadiums. Research Triangle Park, NC: U. S.
35           Environmental Protection  Agency, Office of Air Quality Planning and Standards; EPA report no.
             EPA-450/3-74-049. Available from: NTIS, Springfield, VA; PB-250850.

       Bencala,  K. E.; Seinfeld, J. H. (1976)  On frequency distributions of air pollutant concentrations. Atmos.
             Environ. 10: 941-950.
40
       Benson, P. E. (1979) CALINE3 - a versatile dispersion model  for  predicting air pollutant levels near highways
             and arterial streets. Washington, DC: Federal Highway Administration; report no. FHWA/CA/TL-79/23.
        March 12, 1990                             6-46      DRAFT-DO NOT QUOTE OR CITE

-------
        Benson, P. E. (1984) CALJNE4 - a dispersion model for predicting air pollutant concentrations near roadways.
               Washington, DC: Federal Highway Administration; report no. FHWA/CA/TL-84/15.

        Braverman, T. N. (1987) Intersection model comparison. Research Triangle Park, NC: U. S. Environmental
  5            Protection Agency.

        Bullin, J. A.; Korpics, J. J.; Hlavinka, M. W. (1986) User's guide to TEXIN2 Model~a model for predicting
               carbon monoxide concentrations near intersections. College Station, TX: Texas Transportation Institute.

 10     Catalano, J. A.; Turner, D. B.; Novak, J. H. (1987) User's guide for RAM. 2nd ed. Research Triangle Park,
               NC: U. S. Environmental Protection Agency; EPA report no. EPA-600/8-87-046.

        Chock, D. P. (1978) A simple line-source model for dispersion near roadways. Atmos. Environ. 12: 823-829.

 15     Code of Federal Regulations. (1977) Requirements for preparation, adoption, and submission of implementation
               plans. C. F. R. 40:  sect.  51.

        EMI Consultants. (1985) The Georgia intersection model for air quality analysis. Knoxville, TN: EMI
               Consultants.
20
        Federal Aviation Administration.  (1988) FAA air traffic activity. Washington, DC: U. S. Department of
               Transportation, Office of Management Systems.

        Federal Highway Administration, (n.d.) Factors in highway-project analysis. Washington, DC: Federal Highway
25            Administration; FHWA technical advisory T6640.3.

        Federal Register. (1979) Ambient air quality monitoring, data reporting and surveillance provisions. F. R. (May
               10) 44:  27558-27604.

30     Georgopoulos, P. G.; Seinfeld, J. H. (1982) Statistical distributions of air pollutant concentrations. Environ. Sci.
               Technol. 16: 401A-416A.

        Hare, C. T.; Springer, K. J. (1974) Exhaust emissions from uncontrolled vehicles and related equipment using
               internal combustion  engines: part 4, small air-cooled spark ignition  utility engines; part 5, heavy-duty
35            farm, construction, and industrial engines; part 7, snowmobiles. Research Triangle Park, NC: U. S.
               Environmental Protection Agency; EPA  contract no.  EHS 70-108. Available from: NTIS, Springfield,
               VA; PB-224885, PB-232507, and PB-238295.

        Horowitz, J.; Barakat, S. (1979) Statistical analysis of the maximum concentration of an air pollutant: effects of
40            autocorrelation and nonstationarity. Atmos. Environ. 13:  811-812.

        Ismart, D. (1981) Mobile source emissions and energy analysis at an isolated intersection. Washington, DC:
               Federal Highway Administration.

45     Kunselman,  P.; McAdams, H. T.; Domke, C. J.; Williams, M.  (1974) Automobile exhaust emission modal
               analysis model. Ann Arbor, MI: U. S. Environmental Protection Agency, Office of Mobile Source Air
               Pollution Control; EPA report no. EPA-460/3-74-005. Available from: NTIS, Springfield, VA;
               PB-229635.

50     National Cooperative Highway Research Program. (1979) Development of an improved highway capacity manual.
               National Cooperative Highway Research Program report 3-28.
         March 12,  1990                               6-47       DRAFT-DO NOT QUOTE OR CITE

-------
       NEDS, National Emissions Data System [data base], (n.d.) [Standard computer retrievals]. Unpublished computer
              report available from: U. S. Environmental Protection Agency, Office of Air Quality Planning and
              Standards, Research Triangle Park, NC.

 5     New York State Department of Transportation. (1980) Intersection Midblock Model user's guide. Albany, NY:
              New York State Department of Transportation.

       Ott, W. R.; Mage, D. T.; Randecker, V. W. (1979) Testing the validity of the lognormal probability model:
              computer analysis of carbon monoxide data from U. S. cities. Washington, DC: U. S. Environmental
10            Protection Agency, Office of Monitoring and Technical Support; EPA report no. EPA-600/4-79-040.

       Petersen, W.  B. (1978) User's guide for PAL - a Gaussian-plume algorithm for point, area, and line sources.
              Research Triangle Park, NC: U. S. Environmental Protection Agency, Environmental Sciences Research
              Laboratory; EPA report no. EPA-600/4-78-013. Available from: NTIS, Springfield, VA; PB80-227564.
15
       Petersen, W.  B. (1980) User's guide for HIWAY-2: a highway air pollution model. Research Triangle Park, NC:
              U. S. Environmental Protection Agency,  Environmental Sciences Research Laboratory; EPA report no.
              EPA-600/8-80-018. Available from:  NTIS, Springfield, VA; PB80-227556.

20     Pollack, R. I. (1975) Studies of pollutant concentration frequency distributions. Research Triangle Park, NC:
              U. S. Environmental Protection Agency,  National Environmental Research Center; EPA report no.
              EPA-650/4-75-004. Available from:  NTIS, Springfield, VA; PB-242549.

       Simmon, P. B.; Patterson, R. M.; Ludwig, F. L.; Jones, L. B. (1981) The APRAC-3/Mobile 1 emissions and
25            diffusion modeling package. San Francisco, CA: U. S. Environmental Protection Agency, Region IX;
              EPA  report no. EPA-909/9-81-002. Available from: NTIS, Springfield, VA; PB82-103763.

       Simpson, R. W.; Butt, J.; Takeman, A. J. (1984) An averaging time model of SO2 frequency distributions from
              a single point source. Atmos.  Environ. 18: 115-118.
30
       Smith, W. A. (1985) Updated methodology to assess carbon monoxide (CO) impact at intersections
              [memorandum to Richard G. Rhoads]. Atlanta, GA: U. S. Environmental Protection Agency; January 23.

       Tiao, G. C.;  Box, G. E. P.; Hamming, W. J. (1975) A statistical analysis of the Los Angeles ambient carbon
35            monoxide data 1955-1972. J. Air Pollut.  Control Assoc. 25: 1129-1136.

       U. S. Department of Commerce. (1988) Statistical abstract of the United States. 108th ed. Washington, DC:
              Bureau  of the Census.

40     U. S. Department of Commerce, (n.d.) Current industrial reports. Washington, DC: Bureau of the Census.

       U. S. Department of Energy. (1982) Estimates of U.S. wood energy consumption from 1949 to 1981.
              Washington, DC: Energy Information Adminstration; publication no. DOE/EIA-0341.

45     U. S. Department of Energy. (1984) Estimates of U.S. wood energy consumption 1980-1983. Washington, DC:
              U. S. Department of Energy,  Office of Coal, Nuclear, Electric and Alternate Fuels; publication no.
              DOE/EIA-0341(83).

       U. S. Department of Energy. (1988a) Petroleum marketing monthly. Washington, DC: Energy Information
50            Administration; publication no. DOE/EIA-0380(88/06).

       U. S. Department of Energy. (1988b) Coal distribution January-December. Washington, DC: Energy Information
              Administration; publication no. DOE/EIA-25(88/4Q).
         March  12, 1990                              6-48       DRAFT-DO NOT QUOTE OR CITE

-------
        U. S. Department of Energy. (1988c) Electric power annual. Washington, DC: Energy Information
               Administration; publication no. DOE/EIA-0348(87).

        U. S. Department of Energy. (1989a) Natural gas annual. Washington, DC: Energy Information Administration;
  5            publication no. DOE/EIA-0131(88)/1.

        U. S. Department of Energy. (1989b) Cost and quality of fuels for electric utility plants 1988. Washington, DC:
               Office of Coal, Nuclear, Electric and Alternative Fuels; publication no. DOE/EIA-0191(88).

 10     U. S. Department of Health, Education, and Welfare. (1968) 1968 national survey of community solid waste
               practices. Cincinnati, OH: Public Health Service; PHS publication no. 1867.

        U. S. Department of the Interior. (1971)  Coal refuse fires: an environmental hazard. Washington, DC: Bureau of
               Mines; information circular 8515.
 15
        U. S. Department of the Interior. (1987)  Minerals yearbook. Washington, DC: Bureau of Mines.

        U. S. Department of Transportation. (1988) Highway statistics. Washington, DC: Federal Highway
               Administration.
20
        U. S. Environmental Protection Agency.  (1971) Particulate pollutant system study. Research Triangle Park, NC:
               National Air Pollution Control Administration; contract no. CPA 22-69-104. Available from: NTIS,
               Springfield, VA; PB-203522, PB-203521 and PB-203128.

25     U. S. Environmental Protection Agency.  (1974a) Air quality data - 1972 annual statistics. Research Triangle
               Park, NC: Office of Air Quality Planning and Standards; EPA report no.  EPA-450/2-74-001. Available
               from: NTIS, Springfield, VA; PB-232588.

        U. S. Environmental Protection Agency.  (1974b) Air quality data - 1973 annual statistics. Research Triangle
30            Park, NC: Office of Air Quality Planning and Standards; EPA report no.  EPA-450/2-74-015.

        U. S. Environmental Protection Agency.  (1976a) Air quality data - 1969 annual statistics. Research Triangle
               Park, NC: Office of Air Quality Planning and Standards; EPA report no.  EPA-450/2-76-018.

35     U. S. Environmental Protection Agency.  (1976b) Air quality data - 1970 annual statistics. Research Triangle
               Park, NC: Office of Air Quality Planning and Standards; EPA report no.  EPA-450/2-76-019.

        U. S. Environmental Protection Agency.  (1976c) Air quality data - 1971 annual statistics. Research Triangle
               Park, NC: Office of Air Quality Planning and Standards; EPA report no.  EPA-450/2-76-020.
40
        U. S. Environmental Protection Agency.  (1976d) Air quality data - 1974 annual statistics. Research Triangle
               Park, NC: Office of Air Quality Planning and Standards; EPA report no.  EPA-450/2-76-011.

        U. S. Environmental Protection Agency.  (1976e) National air quality and emissions trends report, 1975. Research
45            Triangle Park, NC: Office of Air Quality Planning and Standards; EPA report no. EPA-450/1-76-002.
               Available from: NTIS,  Springfield, VA; PB-263922.

        U. S. Environmental Protection Agency. (1977a) Guidelines: air quality surveillance networks.  Research Triangle
               Park, NC: Office of Air Programs; publication no. AP-98.
50
        U. S. Environmental Protection Agency. (1977b) Air quality data - 1975 annual statistics including summaries
               with reference to standards. Research Triangle Park, NC: Office of Air Quality Planning and Standards;
               EPA report no. EPA-450/2-77-002. Available from: NTIS, Springfield, VA; PB83-127712.
         March 12,  1990                              6-49       DRAFT-DO NOT QUOTE OR CITE

-------
       U. S. Environmental Protection Agency. (1977c) Guidelines for the interpretation of air quality standards.
              Research Triangle Park, NC: Office of Air Quality Planning and Standards; report no 1.2-008. (Guideline
              series).

 5     U. S. Environmental Protection Agency. (1978) Guidelines for air quality maintenance planning and analysis.
              Volume 9 (revised): evaluating indirect sources. Research Triangle Park, NC: Office of Air Quality
              Planning and Standards; EPA report no. EPA-450/4-78-001. Available from: NTIS, Springfield, VA;
              PB-288206.

10     U. S. Environmental Protection Agency. (1985) Compilation of air pollutant emission factors: v. 1, stationary
              point and the area sources, v. 2, mobile sources. 4th ed. Research Triangle Park, NC: Office of Air
              Quality Planning and Standards; EPA report nos. AP-42-ED-4-VOL-1 and AP-42-ED-4-VOL-2.
              Available from: NTIS, Springfield, VA; PB86-124906 and PB87-205266.

15     U. S. Environmental Protection Agency. (1986) Guideline on air quality models (revised). Research Triangle
              Park, NC: Office of Air Quality Planning and Standards; EPA report no. EPA-450/2-78-027R. Available
              from: NTIS, Springfield, VA; PB86-245248.

       U. S. Environmental Protection Agency. (1987a) National air quality and emissions trends report, 1985. Research
20            Triangle Park, NC: Office of Air Quality Planning and Standards; EPA report no. EPA-450/4-87-001.
              Available from: NTIS, Springfield, VA; PB87-180352.

       U. S. Environmental Protection Agency. (1987b) National air pollutant estimates, 1940-1985. Research Triangle
              Park, NC: Office of Air Quality Planning and Standards; EPA report no. EPA-450/4-86-018. Available
25            from: NTIS, Springfield, VA; PB87-181236.

       U. S. Environmental Protection Agency. (1988) National air quality and emissions trends report, 1986. Research
              Triangle Park, NC: Office of Air Quality Planning and Standards; EPA report no. EPA/450/4-88/001.
              Available from NTIS, Springfield,  VA; PB88-172473/XAB.
30
       U. S. Environmental Protection Agency. (1989a) Supplement to compilation of air pollutant emission factors.
              Research Triangle Park, NC: Office of Air Quality Planning Standards; EPA report no. AP-42 (in press).

       U. S. Environmental Protection Agency. (1989b) User's guide to MOBILE4 (Mobile Source Emissions Factor
35            Model). Ann Arbor, MI:  Office of Mobile Sources; EPA report no. EPA-AA-TEB-89-01.  Available
              from: NTIS, Springfield,  VA; PB89-164271.

       U. S. Environmental Protection Agency. (1990a) National air quality and emissions trends report,  1988. Research
              Triangle Park, NC: Office of Air Quality Planning and Standards; EPA report no.  EPA-450/4-90-002.
40
       U. S. Environmental Protection Agency. (1990b) National air pollutant emissions estimates, 1940-1988. Research
              Triangle Park, NC: Office of Air Quality Planning and Standards; EPA report no.  EPA-450/4-90-001.

       U. S. Forest Service. (1988) Wildfire statistics. Washington, DC: U. S. Department of Agriculture, State and
45            Private Forestry.

       Wackter, D. J.; Bodner, P. (1986) Evaluation of mobile source air quality simulation models. Research Triangle
              Park, NC: U.  S. Environmental Protection Agency, Office of Air Quality Planning and Standards; EPA
              report no. EPA-450/4-86-002. Available from:  NTIS, Springfield, VA; PB86-167293.
50
       Wolcott, M. (1986) Volume 9 update [memorandum to Raymond Vogel]. Ann  Arbor, MI: U. S. Environmental
              Protection Agency, Office of Mobile Sources; January 14.
         March 12, 1990                              6-50       DRAFT-DO NOT QUOTE OR CITE

-------
Yamate, G. (1974) Emissions inventory from forest wildfires, forest managed burns, and agricultural bums.
       Research Triangle Park, NC: U. S. Environmental Protection Agency; EPA report no.
       EPA-450/3-74-062. Available from: NTIS, Springfield, VA; PB-238766.

Zimmerman, J. R.; Thompson, R. S. (1975) User's guide for HIWAY: a highway air pollution model. Research
       Triangle Park, NC: U. S. Environmental Protection Agency, National Environmental Research Center;
       EPA report no. EPA-650/4-74-008. Available from: NTIS,  Springfield, VA; PB-239944.
March 12, 1990                             6-51       DRAFT-DO NOT QUOTE OR CITE

-------
             7.  INDOOR CARBON MONOXIDE SOURCES,
                   EMISSIONS, AND CONCENTRATIONS
 5     7.1  INTRODUCTION
            The activities of individuals are the most important determinants of their exposure to air
       borne contaminants.  In the course of a day individuals spend varying amounts of time in a
       variety of microenvironments (residences, industrial and nonindustrial workplaces,
       automobiles, public access buildings, outdoors, etc.).  Exposures across all
10     microenvironments need to be assessed in evaluating adverse health or comfort effects and in
       formulating cost-effective mitigation measures to reduce or minimize the risks associated with
       exposure.  Exposure can be assessed by direct methods (personal monitoring or by measuring
       biomarkers of exposure) or by indirect methods (microenvironmental monitoring and
       questionnaires combined with an appropriate human exposure model).
15          In recent years there has been a growing recognition of the importance of nonindustrial
       indoor microenvironments in assessing exposures to a wide range of air contaminants
       (National Research Council,  1981; World Health Organization,  1985). This recognition
       reflects the fact that concentrations for many important air contaminants are higher in many
       indoor microenvironments than outdoors and that most individuals spend little time outdoors.
20          Carbon monoxide is introduced to indoor environments through emissions from a variety
       of combustion sources and in  the infiltration or ventilation air from outdoors.  The resulting
       indoor concentration, both average and peak, is dependent on a complex interaction of several
       interrelated factors affecting the introduction, dispersion, and removal of CO.  These factors
       include, for example, such variables as:  (1) the type, nature (factors affecting the generating
25     rate of CO), and number of sources; (2) source-use characteristics; (3) building
       characteristics; (4) infiltration or ventilation rates; (5) air mixing between and within
       compartments in an indoor space; (6) removal rates and potential remission or generation by
       indoor surfaces and chemical transformations;  (7) existence and effectiveness of air
       contaminant removal  systems; and (8) outdoor concentrations; etc.  The interaction of these
30     factors to produce the resulting indoor concentrations usually is considered within the
       framework of the mass balance principle.

       March 12, 1990                         7-1      DRAFT - DO NOT QUOTE OR CITE

-------
            In its simplest form, where steady-state or equilibrium conditions are assumed for a
       single compartment with complete mixing and no air cleaner, the mass-balance model can be
       represented by the following equation.

 5                                        Q = C, + C2                                (7-1)
       where:
                        C, = PAQ/A  + K = outdoor air contribution                     (7-2)
10
                        C2 = S/V/A + K = indoor source contribution                     (7-3)

15     and where:
                  Q = steady-state indoor concentration of CO (jig/m3)
                  Cj =  contribution to indoor CO from outdoor air 0*g/m3)
20
                  C2 =  contribution to indoor CO from indoor sources (/*g/m3)
                  C0 =  outdoor CO (jtg/m3)
25                 P =  fraction of outdoor CO that penetrates the building shell
                   A = air-exchange rate in air changes per hour - ACH per hour
                   S =  generation rate or source strength of CO  (jwg/hr)
30
                   V = volume of the indoor space (m3)
                   K = removal rate of CO by indoor surfaces or chemical transformations -
                        equivalent ACH
35

       This simplified form of the model generally is used to evaluate CO levels indoors.  In
       actuality, however, indoor spaces often are multicompartments with incomplete mixing where
40     the source-generation and contaminant-removal rates and air-contaminant concentrations vary
       considerably in time.  Equation 7-1 is useful particularly for determining the impact on indoor
       air-contaminant concentrations from sources that are used over relatively long periods of time

       March 12, 1990                          7-2     DRAFT - DO NOT QUOTE OR CITE

-------
        (e.g., unvented kerosene or gas space heaters) where steady-state or equilibrium conditions
        are reached.  When applied to sources that are intermittent in their use (e.g., gas range or
        tobacco combustion), Equation 7-1 averages over the off/on periods of the sources to
        determine average input parameters for the model.  Short-term indoor concentrations of air
  5     contaminants associated with sources whose use varies considerably with time can be modeled
        with the differential version of Equation 7-1,  when detailed information on the time
        variability of the source use, mixing, and removal terms are available.  Field data on short-
        term variability of contaminant concentrations and associated variables, however, are not
        available.
 10          Carbon monoxide generally is assumed have low reactivity indoors (Yamanaka, 1984;
        Leaderer et al.,  1986; Traynor et al.,  1982; Borrazzo et al., 1987; Caceres et al., 1983), that
        is, CO removal by indoor surfaces or chemical transformations is approximately equal  to zero
        for typical indoor CO residence times (K=0 in Equations 7-2 and 7-3).  There are no
        chamber or field studies that have measured the penetration factor for CO. Given the low CO
 15     reactivity, P generally is assumed equal to one (P= 1) for conditions where outdoor levels of
        CO do not vary  rapidly.  Under these typical  assumptions (K=0 and P= 1), indoor
        concentrations of CO can be represented by the following simplified form of Equations 7-1,
        7-2, and 7-3.

20                                       Q = C0 + S/V/A                              (7-4)


        In the case where there are no indoor sources, then the average indoor concentration  is equal
25      to the average outdoor concentration.  With short variations in outdoor concentrations,
        however, indoor CO concentrations will lag outdoor concentrations and will be dependent on
        the air-exchange rate in a space. When an indoor source exists, the indoor CO concentration
        will be equal to the outdoor concentration plus the contribution of the indoor source,  which is
        a function of the source strength (CO emission rate), volume of the indoor environment and
30      air-exchange rate.
             This chapter will summarize the data currently available on emissions of CO from
        sources commonly  found indoors and on levels of CO measured in a variety of indoor
        microenvironments.
        March 12, 1990                           7-3     DRAFT - DO NOT QUOTE OR CITE

-------
       7.2  EMISSIONS FROM INDOOR SOURCES
            Carbon monoxide emitted directly into the indoor environment is one of several air
       contaminants resulting from combustion sources.  Such emissions into occupied spaces can be
       unintentional or the result of accepted use of unvented or partially vented combustion sources.
 5     Faulty or leaky flue pipes, backdrafting and spillage from combustion appliances that draw
       their air from indoors, (i.e., Moffatt, 1986), improper use of combustion sources (i.e., use of
       a poorly maintained kerosene heater), and air intake into a building from attached parking
       garages are all examples of unintentional or accidental indoor sources of CO. The U.S.
       Departmental of Health and Human Services (1986) estimates between 700 and 1,000 deaths
10     per year in the U.S.  alone are due to accidental carbon monoxide poisoning.  The number of
       individuals experiencing severe adverse health effects at sublethal carbon monoxide
       concentrations from accidental indoor sources is no doubt many times the estimated deaths.
       While the unintentional or accidental indoor sources of CO represent a serious health hazard,
       little is known about the extent of the problem. Such sources cannot be characterized for CO
15     emissions in any standard way that would make the results extendable to the general
       population. Unintentional or accidental sources are not considered in this review of emissions
       of CO from indoor sources.
            The major indoor sources of CO emissions that result from the accepted use of unvented
       or partially vented combustion sources include gas cooking ranges and ovens, gas appliances,
20     unvented gas space heaters, unvented kerosene space heaters, cigarette combustion, and wood
       burning stoves. This section of the chapter will summarize the available CO-emission data
       for the major indoor-sources  of CO.  The experimental design, measurement methods, and
       results of studies of indoor-source CO emissions will be discussed.
            The characterization of CO emissions from indoor sources is essential in providing
25     source strength input data for indoor modeling, understanding  fundamental processes
       influencing emissions, guiding field study designs aimed at assessing indoor CO exposures,
       identifying and ranking important sources, and in developing cost-effective mitigation
       measures that will minimize exposures.
30
        March 12, 1990                          7-4      DRAFT - DO NOT QUOTE OR CITE

-------
        7.2.1  Emissions from Gas Cooking Ranges, Gas Ovens, and Gas
               Appliances
             Estimates indicate that gas (natural gas and liquid propane) is used for cooking, heating
        water, and drying clothes in approximately 45.1% of all homes in the United States (U.S.
 5      Bureau of the Census, 1982) and near 100% of the homes in other countries (e.g., The
        Netherlands). Unvented, partially vented, and improperly vented gas appliances, particularly
        the gas cooking range and oven, represent an  important source category of CO emissions into
        the indoor residential environment. Emissions of CO from these gas appliances (the source
        term, S, in Equation 7-4) are a function of a number of variables relating the source type
10      (range top or oven, water heater, dryer, number of pilot lights, burner design, etc.), source
        condition (age, maintenance, combustion efficiency, etc.), source use (number of burners
        used, frequency of use, fuel consumption rate, length of use, improper use, etc.) and venting
        of emissions (existence and use of outside vents over ranges, efficiency of vents, venting of
        gas dryers, etc.).
15           A number of chamber studies have investigated CO emissions from gas cooking ranges,
        ovens, and appliances. The studies have used two basic  approaches.  The first is called  the
        direct or sampling-hood approach (Moschandreas et al.,  1985; Himmel and DeWerth, 1974)
        and the second is the mass-balance or chamber approach (Moschandreas et al., 1985; Traynor
        et al., 1982; Leaderer, 1982).  In the direct approach, emissions are  sampled using a quartz
20      hood through which the combustion emissions pass and are sampled.   Because the effluent
        gases measured by this method contain substantial amounts of excess  air that can vary
        considerably from run to run, the measured concentrations are converted to a hypothetical
        undiluted or "air-free" basis for calculating CO emission rates.  This  method is used in both
        chamber and field evaluations of emission rates (Moschandreas et al., 1985; Borrazzo et al.,
25      1987).  The mass-balance approach utilizes a  well-mixed environmental chamber where  the
        relationship between changes in concentration of CO over time in relation to outdoor
        concentration, source-emission rate, air-exchange rate, and removal rate are evaluated (mass-
        balance equation).  Both approaches yield source emission rates for CO, usually expressed as
        micrograms per kilojoule (/xg/kJ).
30           Emissions from gas range-top burners typically are evaluated using a standardized  water
        load in a cooking pot (American National Standards Institute Z21.1-1974, 1982) or a

        March 12,  1990                           7-5     DRAFT - DO NOT QUOTE OR CITE

-------
       modification of it (Borrazzo et al., 1987).  The cooking pot has a top that is sealed, except
       for a 3/4 in. pipe that extends from the center of the top to allow steam to escape.
            Using both the direct-sampling hood and  mass-balance approach and a standardized
       water load, Moschandreas et al. (1985) evaluated CO emissions from three new gas ranges
 5     (including pilot and non-pilot light and self-cleaning oven and conventional oven) with six
       hours of conditioning before use in actual testing. The data of Cole and Zawacki (1985) are
       incorporated into the Moschandreas et al. (1985) report.  The range CO emissions were
       evaluated for appliance type, the conditions of blue-flame operation (burner air shutter set at
       the manufacturer's recommended level), and yellow-tipping (air shutters are closed - worst
10     condition). Two natural gas mixes were evaluated:  lean mix (983 Btu/scf) and rich mix
       (1022 Btu/scf).
            A total of 116 direct-sampling experiments were conducted in which CO emissions were
       measured for all 12 burners (four burners per stove) using rich and lean fuels. Thirty direct-
       sampling experiments were conducted with the burners operating with yellow tipping. No
15     significant differences (criterion: p<0.05) were found by fuel type.  The results of the tests
       using blue and yellow flame setting are shown in Table 7-1.  The improper operation of the
       burners (yellow-tipping flame)  resulted in an approximately two- to fivefold increase in CO
       emissions. Considerable variation in CO emissions from burner to burner within and between
       ranges were noted by the authors for blue-flame operation although overall CO emission
20     averages for burners by each range were within a factor of two.
             A total  of 144 mass-balance experiments on the three ranges were conducted in a 33-m3
       chamber to evaluate the impact of fuel composition, range type, primary aeration level (blue
       or yellow flame), fuel-consumption rate (high-9149 Btu/h, medium-7673 Btu/h,
       low-1492 Btu/h, and warm-807 Btu/h), test chamber air-exchange rate, temperature and
25     humidity (range of 15 to 50%), and temporal effects on CO-emission rates. CO emissions
       showed little variation by fuel composition (lean vs. rich), range type, test chamber air-
       exchange rate, and time.  Yellow-tipping flame conditions resulted in a two- to tenfold
       increase in CO emissions over blue-flame operation conditions.  No changes in emissions
       were noted as a function of changes in chamber temperature and relative humidity.  Little
30     change in CO emissions were noted for the high, medium, and low fuel-consumption rates
       while a sevenfold  increase in emissions were observed for the warm setting.  In a comparison

       March 12, 1990                          7-6     DRAFT - DO NOT QUOTE OR CITE

-------
       TABLE 7-1. CARBON MONOXIDE-EMISSION RATES' FOR 12 RANGE-TOP BURNERS
           OPERATING WITH BLUE AND YELLOW-TIPPING FLAMES BY THE DIRECT
                  SAMPLING METHOD (calculated from Moschandreas et al., 1985)
10

Number of
Tests
25
33
58
11
9
12

Gas
Range
1
2
3
1
2
3

Flame
Condition
Blue"
Blue
Blue
Yellow0
Yellow
Yellow

Emission

Rate 0*g/kJ)
Average
50.7
34.3
70.9
190.0
196.9
108.4
SD
9.46
4.93
12.9
5.8
3.8
5.6
Range
17.2 - 107.5
15.1 - 71.4
15.1-215.0
53.8 - 344.0
60.2 - 344.0
94.6 - 227.9
15
       'Lean and rich fuel mixtures combined.
20     bBlue-flame condition - well tuned.
       "Yellow-tipping flame - improperly tuned.
25     of CO emission results obtained from both the direct-sampling and mass-balance experiments
       (Moschandreas et al.,  1985), for blue-flame and yellow-tipping flame conditions, differences
       were observed. No clear trend emerged, however, because CO emissions varied among
       experimental runs by as much as an order of magnitude.
            As part of the above study CO emissions from gas ovens, gas range pilot lights, and a
30     gas dryer were evaluated for blue-flame operating conditions. The results of these
       experiments are shown in Table 7-2.  Pilot light emissions are comparable to those of gas
       range burners (Table 7-1).  Variability by oven use was observed but the limited data is not
       sufficient to draw any  conclusions. Gas dryer CO emissions appear to be comparable to those
       from gas range burners.
35          Eighteen different gas ranges, representing greater than 90% of the gas stoves in use at
       the time, were tested for CO emissions by the American Gas Association Laboratories
       (Himmel and Dewerth, 1974). CO emissions were evaluated for top burners, ovens and
       broilers, burner pilot lights, and oven pilot lights. The protocol utilized the direct
       measurement method for both blue-flame (well-adjusted flame) and yellow-tipping flame
40     (poorly adjusted flame) operating conditions.  Range-top burner evaluations, a total of 72,
       March 12, 1990                         7-7      DRAFT - DO NOT QUOTE OR CITE

-------
10
15
20
25
          TABLE 7-2. CARBON MONOXIDE-EMISSION RATES' FOR GAS RANGE OVENS,
                          GAS RANGE PILOT LIGHTS, AND GAS DRYERS
                              (calculated from Moschandreas et al., 1985)

Gas
Range
1
1
2
3
3
3
1
1
2
Gas dryer"


Burner
Bake
Broil
Broil
Bake
Broil
Self-clean
Oven door openb
Oven door closedb
Pilot light0

Number
of
Tests
4
4
9
2
7
3
3
3
20
4/3
Average Emission Rates
G*g/kJ)
Direct Method
19.1 (6.7)
13.8 (3.9)
13.2(2.1)
68.5(6.1)
21.5 (1.0)
16.0 (2.2)



40.4 (3.0)
(SD)
Mass Balance






54.6 (2.6)
127.3 (3.5)
40.4 (5.2)
68.8 (14.2)
"Combined and rich and lean fuel and no oven load.
bBroiler operated.
"Results are on a per pilot light basis, experiments covered various pilot light combinations.
dUsing Association of Home Appliance Manufacturers Standard HLD-2EC, four tests are by the
 direct method and three tests by the mass-balance method.
30     were made with a standard water pot load centered on the grate (American National Standards
       Institute, 1974). Oven tests, 27 in all, were conducted without a cooking load because the
       substantial thermal mass of the cavity itself makes a load unnecessary.
            A summary of the CO-emission rates measured in the Himmel and Dewerth (1974)
       study are shown in Table 7-3.  Yellow-tipping flame operating conditions resulted in higher
35     CO emissions for burners and ovens than the blue-flame conditions.  Considerable variability
       existed from burner to burner within and between gas ranges and among ovens,  yet the
       average emission rates for blue-flame operation among top burners, ovens, and burner pilot
       lights were generally within a factor of four or five of each other.  The authors noted that the
       CO emission-rate distributions were skewed, leading them to average the emission data using
40     a log-normal transform and to present the average concentration and an interval  representing
       66% of the measurements.  The infrared burner, pyrolytic self-cleaning oven, and oven pilot
       light emission rates were considerably higher than the typical range top burner.
       March 12, 1990                           7-8     DRAFT - DO NOT QUOTE OR CITE

-------
           TABLE 7-3. CARBON MONOXIDE-EMISSION RATES FROM 18 GAS RANGES,
                  GAS OVENS, AND GAS PILOT LIGHTS FOR BLUE FLAME AND
                 YELLOW-TIPPING FLAME BY THE DmECT-S AMPLING METHOD
           Burner Type
                                                Average and Range of Emissions'
   Blue Flame
                                                                Yellow-Tipping
                                                                    Flame
10
15
20
25
30
35
Top burners

Ovens and broilers

Top burners with thermostat

Top burners (142 kJ/min)

Top burners (190 kJ/min)

Infrared burners

Ovens and broilers with
 catalytic clean

Pyrolytic self-clean oven

Pilot lights-burner
 a. Free standing
 b. Baffle around flame
 c. Baffle around flame and
    shield above flame

Pilot lights-oven
 a. Constant horizontal
 b. Constant horizontal
    operates in two modes
 22.6 (8.0-64.2)'

 15.7 (6.3-38.8)

 51.8 (11.9-228.8)

 15.3 (6.6-35.4)

 26.1 (9.4-72.0)

 77.4

 11.9 (4.9-29.1)


 87.7

 44.0 (30.4-63.8)
 35.7
 28.3
 56.1 (39.6-69.7)b
248.3 (146.5-420.0)
322.2 (158.5-491.2)"
208.8 (135.8-281.8)"
                                                                      156.6  (58.0-421.0)

                                                                       62.0 (11.1-349.2)
                                                                      53.5  (11.8-243.4)
40     "Values are the range of emission rates that contain two-thirds of the measured values.
       "Values at the low and high measured level.

       Source:  Himmel and Dewerth (1974).
       March 12, 1990
                                       7-9     DRAFT - DO NOT QUOTE OR CITE

-------
            Himmel and Dewerth (1974) noted that significant differences (p^O.05) for the range
        averages for all four burners existed for three of the eighteen ranges tested. Emissions from
        front burners were found to average 13% higher than back burners. Significant CO emission
        differences (p>0.05) were noted between oven burners.  Type of cooking pot (material), size
 5      of cooking pot, physical properties of the individual pots (density, thermal conductivity, etc.)
        did not have a pronounced impact on emissions.  Burner cap design was found to influence
        CO emissions.
            Using the indirect measurement method and the standard water pot load (American
        National Standards Institute, 1982), Tikalsky et al.  (1987) reported the results of CO-
10      emission measurements made on gas ranges in 10 homes with each home having a different
        range make, spanning a use age of from 7 to 30 years.  The sample of homes was drawn
        from a sample of 50 homes from  which house NO2 measurements (Dames and  Moore, 1986)
        were available. This is the only study reported where field measurements were made on gas
        cooking ranges actually in use. Five of the 10 homes were resampled after routine service
15      adjustments.  Emissions were measured for top burners and  ovens (without a cooking load) by
        two different research groups using similar protocols but different air-sampling equipment.
        Emissions measurements were made for two top burners and for bake and broil oven use on
        each range. The top-burner emissions were measured for high, medium, and low settings.
            Results of the Tikalsky et al. (1987) study showed CO-emission  rates for the pre-
20      serviced top burners (across all burners, burner settings,  and both measurement groups) to
        range from 9.5 to 1746 /ig/kJ.  CO emissions for baking ranged from 6.9 to 413 ftg/kJ,
        whereas emissions for broiling ranged from 4 to 310 ^g/kJ.  Low top-burner settings resulted
        in significantly higher CO emissions than the medium or high settings (p>0.05).  Comparison
        of the  emission-rate measurements between the two measurement teams indicated that one
25      team measured higher rates.  CO  emission rates showed a significant reduction after routine
        service adjustments (p^O.05).  The authors noted that the field measurements of CO emissions
        in their study sometimes exceeded those previously measured by other investigators by a
        factor  of four, while oven emissions for both the bake and broil were  within the range of
        those reported by others.
30          A number of studies of CO  emission rates from gas cooking ranges have been conducted
        in which a limited number of samples (ranges, burners, and  number of experiments) have

        March 12,  1990                          7-10    DRAFT - DO  NOT QUOTE OR CITE

-------
        been collected.  These studies have utilized the direct and mass-balance methods.  The results
        of these studies for both top burners and ovens are shown in Table 7-4. As seen in the
        previous studies the CO-emission rates for both the top burners and ovens showed
        considerable variation with the yellow-tipping flame condition resulting in higher emissions
 5      than the blue-flame condition. Borrazzo et al.  (1987) measured CO-emission rates for the
        pilot lights of the gas stove in their test house as well as for fugitive emissions from other
        vented combustion appliances (gas dryer, water heater, furnace, etc.).  A CO-emission rate
        for a pilot light flow of 7.6 kJ/min was measured  at 91 +  16 pig/U, whereas emissions from
        other vented sources were negligible.
10           Natural gas is used for domestic water heating in approximately 55 million residences in
        the United States.  Cole and Zawacki  (1985) summarized the available data on CO emissions
        from domestic hot water heaters.  They reported an average CO-emission rate for a total of
        18 gas water heaters tested in three studies (Belles et al., 1979; Thrasher and DeWerth, 1977;
        A.G.A.L., 1983) of 12.0 /*g/kJ.  Thrasher and DeWerth (1977).  In comparing emissions
15      from 13 water heaters for blue-flame and yellow-tipping flame conditions, they found that
        yellow-tipping flame operation conditions of the heaters resulted in a fivefold increase in
        emissions.
             The available literature for gas appliances indicates that CO emissions are (1) highly
        variable for range-top burners on  a single range and between ranges and for ovens  for blue-
20      flame conditions (properly adjusted), varying by as much as an order of magnitude or more;
        (2) much higher for range top and oven burners operated under yellow-tipping flame
        conditions (improperly adjusted) than for blue-flame conditions; (3) not different for rich  or
        lean fuels; (4) higher for top burners when  they are operated under very low fuel consumption
        rates;  (5) comparable for top burners, ovens, pilot lights, and unvented gas dryers; and
25      (6) roughly comparable when obtained by either the direct or mass-balance method.
             The data base on CO emissions from gas cooking ranges is largely based upon
        laboratory studies in which a relatively few ranges were tested. The one field study in which
        CO  emissions were measured, for a small number of gas cooking ranges used in private
        residences, indicates that for top burners, the laboratory data may underestimate actual CO
30      emission rates. More extensive field CO emission data for cooking ranges  is needed to
        determine how representative the laboratory derived data is.

        March 12, 1990                           7-11      DRAFT - DO NOT QUOTE OR CITE

-------
 10
15
20
25
30
               TABLE 7-4.  CARBON MONOXIDE EMISSIONS FROM GAS RANGES
                            FOR STUDIES OF SMALL SAMPLE SIZE
Study
MIT/AGA
(1976)
Fortman et al.
(1984)
Borrazzo et al.
(1987)
Cote et al.
(1974)
Traynor et al.
(1982)
Goto&
Tammura
(1984)
Burner
1 top burner
4 top burners
2 ovens
4 top burners
1 oven-300 U/min
-150 kJ/min
-160 kJ/min
2 ranges/top burner
2 ranges/top burner
2 ovens
2 top burners
1 oven
1 top burner

Test"
Method
Db
D
D
D
D
D
D
C
C
C
C
C
C

Flame
Condition
Blue
Blue
Blue
Blue
Blue
Blue
Blue
Blue
Yellow
Blue
Blue
Blue
Blue

Number
of
Tests
1
11
4
16
2
6
2
2
2
2
5
2
1

Emission
Rate
(Mg/kJ)
11.6
110 ± 40
25.8 ± 4
98 ± 18
150 ± 18
33 + 1.3
33 ± 6.5
56
92.5
257
200 + 34
226 ± 17
86.9

35
       *D = direct method, C = mass-balance chamber method.
       bDid not use standard pot water load.
       7.2.2  Emissions from Unvented Space Heaters
40          Unvented kerosene and gas space heaters are used in the colder climates to supplement
       central heating systems or in more moderate climates as the primary source of heat.  During
       the heating season, space heaters generally will be used for a number of hours during the day
       resulting in emissions over relatively long periods of time.
       March 12, 1990
7-12    DRAFT - DO NOT QUOTE OR CITE

-------
             Over the last several years there has been a dramatic increase in the use of unvented
        kerosene space heaters in residential and commercial establishments primarily as a
        supplemental heat source.  The U.S. Consumer Product Safety Commission estimates that a
        total of 16.1 million such heaters have been sold through 1986 (Womble, 1988).  A
 5      residential energy survey conducted by the U.S. Department of Housing and Urban
        Development (1980) estimated that three million residences use unvented gas space heaters
        (fueled by natural gas or propane), with their use more prevalent in the South Census region
        of the United States.  The large number of unvented space heaters sold in the United States
        and the potential for their high use, particularly during periods when energy costs rise
10      quickly, make them an important source of CO indoors.
             CO emissions from unvented kerosene and gas space heaters can vary considerably and
        are a function of heater design (convective, radiant, combination, etc.), condition of heater,
        and manor of operation (e.g., flame setting).
             Unvented gas space heaters (UVGSHs) range in size from 7000 to 40,000 Btu/h and
15      vary in design and operation. Design characteristics of different heaters include the burners
        (cast iron,  steel, ceramic tile, catalytic surfaces, screen, etc.), ignition, heat exchanger, and
        auxiliary equipment (i.e., oxygen depletion sensor).  Operation characteristics include type of
        fuel (natural or LPG), input modulation, primary air-shutter, flame type (blue, yellow-
        tipping, infrared, etc.), and flame discharge temperature (blue flame or convective - 2800°F,
20      infrared -  1800°F, and catalytic - 1200°F).  UVGSHs often are distinguished by their flame
        discharge temperature and type of fuel used, when characterizing CO emissions. Both the
        direct (hood over the heater from which gases are sampled) and the mass balance or chamber
        methods have been used to evaluate emissions from invented gas space heaters.  Many of the
        studies were parametric in nature, seeking to evaluate the impact of some of the design and
25      operational features on emissions. Table 7-5 presents a general summary of the CO emission
        data from UVGSHs that were obtained by operating the heaters in a well-adjusted and
        typically full-input mode.
            The summary data in Table 7-5 indicates that there is considerable variability of CO
        emissions from UVGSHs.  Infrared heaters produce higher emissions than the convective or
30      catalytic heaters.  The mass-balance or chamber method seems to result in somewhat higher
        emissions than the direct method.  The Traynor et al. (1984, 1985) data indicate, that for the

        March 12, 1990                          7-13     DRAFT - DO NOT QUOTE OR CITE

-------
                        TABLE 7-5. CARBON MONOXIDE EMISSIONS FROM UNVENTED GAS SPACE HEATERS
VO
71


Study
Tray nor et al.
(1984, 1985)




Moschandreas
et al. (1985)




Thrasher and
DeWerth (1979)
Zawacki et al.
(1984)
Private11
Communication
Number
Type of of
Heater Heaters
Convective 9
3
12
Infrared 4
1
5
Convective 1

Infrared 1

Catalytic 1

Convective 2
3
Convective 1

Convective 1
Infrared 1
Number of
Tests/
Heaters
1
1
1
1
1
1
16
4
7
3
6
3
2
3





Fuel"
Type
NG
P
Both
NG
P
Both
NG

NG

NG

NG
NG





Testb
Method
C
C
C
C
C
C
D
C
D
C
D
C
D
D
D

U
U

Fuel Consumption
Rate (kJ/min)
177 - 784
353 - 660

277 - 368
258

186
186
260
260
207
207
131
381
263

211
ND


Flame
Blue
Blue

Infrared
Infrared

Blue
Blue
Infrared
Infrared
Blue
Blue
Blue
Blue
Blue



Type(s)c
of
Burner
SP, DP, R
SP, DP, R

CT
CT

R
R
SP
SP
CT, R
CT, R
DP
DP
R

RS
RS
Suspended
Radiating
Tiles
Yes
Yes

No
No



No
No
No
No
No
Yes
Yes

No
Yes
Emission Rate
(us/kJ)
Average
33
16
29
47
45
47
9.7
16.8
69
54.6
9.0
14.2
3.1
5.2
16.3

32.7
0.22
SD
26
1.9
2.5
1.6

1.5
0.4
1.7
0.9
8.2
0.9
2.6






          "NG = natural gas, P = propane.
          bC = chamber mass balance method, D = direct method, U = unknown.
          CSP = slotted port burner, DP = drilled port burner, R = pressed metal ribbon burner, CT = ceramic tile, RS = retention screen.
          dData obtained and reported by Moschandreas et al. (1985).

-------
       subsample of convective heaters tested for the impact of fuel (natural gas vs. propane),
       natural gas use results in higher emissions.
            In a series of tests (mass-balance methods applied in a chamber and test house) on
       subsamples of heaters (Traynor et al.,  1984, 1986) CO emissions were found to (1) be lower
 5     for partial input heater operation, (2) increase at lower chamber oxygen concentrations,
       (3) be comparable between chamber and test house studies and show some decrease in time in
       test house studies, and (4) be increased substantially for some maltuned heaters.
       Moschandreas et al. (1985) found no difference in emissions for lean versus rich fuel.
            Unvented kerosene space heaters fall into four basic design categories: convective,
10     radiant, two-stage, and wickless.  The convective heaters operate at a relative high
       combustion temperature and, depending upon burner design, can be a blue or white flame.
       The radiant or infrared heaters utilize perforated ceramic or metal cylinders  which become red
       hot. The two-stage heaters (newer design kerosene heaters) are very similar to the radiant
       heaters in design except they have a second combustion chamber above the first.  The
15     wickless heaters have a chamber where the fuel and air are mixed and combustion occurs with
       the resultant heat  distributed via a fan. CO emissions from unvented kerosene heaters of the
       convective, radiant, and two-stage design have been evaluated by both the direct and mass-
       balance method for conditions of wick height (fuel consumption rate) for well-tuned heaters.
       Few emission data are available for the wickless heaters.
20          A summary  of CO-emission data for unvented kerosene space heaters is shown in
       Table 7-6. There is considerable variability in emissions between and among heater types
       and heater wick settings.  Radiant heaters, when operated under normal wick settings,
       produce considerably more CO than normal wick settings for the convective and two-stage
       heaters.  Two-stage heater emissions, however, jump considerably at lower  wick settings.
25     Low-wick settings can increase CO emissions for convective heaters.

       7.2.3  Emissions from Wood Stoves and Tobacco Combustion
            Use of wood-burning stoves has been a popular cost savings alternative to conventional
       central heating systems using gas or oil.  CO and other combustion by-products enter the
30     indoor environment during fire start-up, fire-tending functions, or through leaks in the stove
       March 12, 1990                          7-15     DRAFT - DO NOT QUOTE OR CITE

-------
            TABLE 7-6. CARBON MONOXIDE EMISSIONS FROM UNVENTED KEROSENE SPACE HEATERS
o
^^
^3
»— *
^^5
^CJ







^
H-L


o
5
5
i
o
o
^-\
3
*§
g
3
o
5*
O
HH
a


Study
Leaderer (1982)





Tray nor et al.
(1983)






Jones et al.
(1983)c

Moschandreas et al.
(1985)


Number
Type of of
Heater Heaters
C 1


R 1


C 2


R 3


2S 2

C
R

C 1
R 1



Test"
Method
C


C


C


C


C

D
D

C
C



Fuel Consumption
Rate (kJ/min)
37.3
97.9
158.0
84.4
113
144
130
193

113
148

132
182
202
168d

100
129


Number
of
Tests
3
3
3
3
3
3











4
4




Emission Rate f^g/kJ)
Average
25.8
22.3
10.1
72.9
58.2
42.6
60
12

173
68

54
9
4.7
27.5

35.3
64.1


SD
4.7
1.5
4.1
2.6
5.0
2.5
1.1
1.1

1.4
1.5

2.5
1.2



8.2
5.2


"C = convective, R = radiant, 2S = two-stage.
bC = chamber/mass balance
°As reported by Michandreas
dManufacturer's rating.
method, D = direct/load method.
et al. (1985).

















-------
       or venting system.  Hence it is difficult to evaluate indoor CO-emission rates for wood-
       burning stoves.  Traynor et al. (1987) evaluated indoor CO levels from four wood-burning
       stoves (three airtight and one non-airtight stove) in a residence.  The non-airtight stove
       emitted substantial amounts of CO to the residence, particularly when operated with a large
 5     fire. The airtight stoves contributed considerably less.  The average CO source strengths
       during stove operation reported for the airtight stoves ranged from 10 to 140 cm3/h, whereas
       levels for the non-airtight stove source strengths ranged from 220 to 1800 cmVh.
            In 1980, 32% of the U.S. adult population was reported to be smokers (U.S.  Depart-
       ment of Commerce, 1984).  The combustion of tobacco represents an important source of
10     indoor air contaminants. CO is emitted indoors from tobacco combustion through the exhaled
       mainstream smoke (MS) and from the smoldering end of the cigarette (sidestream smoke -
       SS). MS and SS CO-emission rates have been evaluated  extensively in small chambers (less
       than a liter in volume) using a standardized smoking machine protocol. The results of these
       studies have been summarized and evaluated in the Surgeon General's reports (e.g., 1986)
15     and the National Academy of Science Report on environmental tobacco smoke (1986).  These
       results indicate considerable variability in total (MS+SS) CO emissions, with a typical range
       of from 40 to 67 mg/cig. A small chamber study of 15 brands of Canadian cigarettes
       (Rickert et al., 1984) found the average CO-emission rate (MS+SS) to be 65 mg/cig.  A
       more  limited number of studies have been done using large chambers with the occupants
20     smoking or using smoking machines. Girman et al. (1982) reported a CO-emission rate of
       94.6 mg/cig for a large chamber study in which one cigarette was evaluated. A CO-emission
       factor of 88.3 mg/cig was reported by Moschandreas et al.  (1985) for a large chamber study
       of one reference cigarette.

25     7.2.4  Summary of Emission Data
            Indoor sources of CO can be considered as  unintentional or accidental (leaky  flue pipes,
       backdrafting, etc.) and intentional (emissions from unvented combustion sources).  Emissions
       from unintentional sources can result in indoor concentrations associated with serious acute
       health effects and result in several hundred deaths per year  in the U.S. CO emissions from
30     these  unintentional sources,  despite their importance, cannot be characterized in any
       standardized way.  Unvented or partially vented gas cooking ranges and ovens, gas

       March 12, 1990                          7-17    DRAFT - DO NOT QUOTE OR CITE

-------
       appliances, unvented gas space heaters, unvented kerosene space heaters, cigarette
       combustion, and wood-burning stoves are all notable "intentional" indoor sources of CO
       emissions.
            Unvented or partially vented sources of CO have been evaluated for CO emissions by
 5     either the direct method or the mass-balance approach. The direct method samples the
       emitted combustion gases as they pass through a sampling hood above the source. The mass-
       balance approach measures the changes in CO over time in an environmental chamber or test
       house in relation to changes  in outdoor CO concentration, source emission rates, and CO
       removal rates. For gas range top burners, emissions typically are evaluated using a
10     standardized water load in a cooking pot.
            Emissions from unvented, partially vented, or improperly vented gas cooking ranges and
       ovens and gas appliances represent an important source of CO in the residential environment
       in the United States due to the high percentage of homes  (approximately 45%) using gas to
       cook.  CO emissions from these sources are a function of a number of variables relating to
15     the source type (range top or oven, burner design, pilot light, etc.), source condition (age,
       maintenance, etc.), source use (number of burners used, fuel consumption rate, etc.) and use
       of outside vents.  The source emission studies typically have been conducted in the laboratory
       setting and involved relatively few gas ranges and gas appliances. The reported studies
       indicate that CO emissions are highly variable among burners on a single gas cooking range
20     and between gas cooking ranges and ovens,  varying by as much as an order of magnitude.
       Operating a gas cooking range or oven under improperly adjusted flame conditions (yellow-
       tipping) can result in greater than a fivefold increase in emissions when compared to properly
       operating flame conditions (blue flame). Use of a rich or lean fuel appeared to have little
       effect on CO emissions. In  general, CO emissions were  roughly, on an average, comparable
25     for top burners, ovens, pilot lights, and unvented gas dryers when corrected for fuel
       consumption rate. The emissions rates gathered by either the direct or mass-balance method
       were comparable. Only one study attempted to evaluate gas stove emissions in the field for a
       small number (10) of residences.  This study found CO emissions to be as much as a factor of
       four higher than chamber studies.  Given the prevalence of the source, limited field
30     measurements and poor agreement between existing laboratory and field derived CO emission
       March 12, 1990                          7-18     DRAFT - DO NOT QUOTE OR CITE

-------
       data, there is a need to establish a better CO emission data base for gas cooking ranges in
       residential settings.
            CO emissions from unvented gas space heaters were found to be variable from heater to
       heater but roughly comparable to those for gas cooking ranges.  Infrared gas space heaters
 5     produced higher emissions than the convective or catalytic heaters. Emissions of CO for
       these heaters were higher for maltuned heaters and for the mass-balance versus direct method
       of testing.  No differences for rich or lean fuel were found, but use of natural gas resulted in
       higher emissions than propane.  Lower fuel consumption settings resulted in lower CO
       emissions.  Emissions were observed to vary in time during a heater run and increase when
10     room or chamber oxygen levels  decreased.
            Among the three principal unvented kerosene space heater designs (radiant, convective
       and two-stage burners), radiant heaters produced the highest CO emissions with the
       convective heaters producing the lowest emissions. Wick setting (low, normal, or high) had a
       major impact on emissions with the low-wick setting resulting in the highest CO emissions.
15     Data from different laboratories are in good agreement for this source.
            CO emissions into indoor spaces from wood-burning stoves occur during fire start-up,
       fire-tending,  or through leaks in the stove or venting  system. Few data are available
       characterizing CO emissions for normal wood stove or fireplace operation.  The available data
       indicate that the non-airtight, wood-burning stoves can contribute substantial amounts of CO
20     directly to the indoor environment while the airtight stoves contribute little or none.
            Tobacco combustion represents an important indoor source of CO based upon the
       number of cigarettes smoked. In comparison to other unvented combustion sources, CO
       emissions into indoor spaces from tobacco combustion are relatively low and show little
       variation from brand to brand.
25          The available data on CO emissions from unvented combustion sources are based largely
       upon chamber or test house studies using the mass-balance or direct-measurement method for
       a small sample of sources (i.e.,  a small number of gas cooking ranges). Given the high
       variability of CO emissions observed from these sources in the available studies, additional
       data are needed to better understand the factors impacting those emissions.  Little or nothing
30     is known about CO emissions from unvented combustion  sources actually in use in residences.
        March 12, 1990                          7-19     DRAFT - DO NOT QUOTE OR CITE

-------
       The few data available indicate that CO emissions from sources in the field are considerably
       more variable and typically higher than those observed in the chamber or test house studies.
 5     7.3  CONCENTRATIONS IN INDOOR ENVIRONMENTS
            CO concentrations in an enclosed environment are affected by a number of factors in
       addition to the source factors discussed in the previous sections.  These factors include
       outdoor concentrations, proximity to outdoor sources (i.e., parking garages or traffic),
       volume of the space, and mixing within and between indoor spaces.
10          Outdoor CO concentrations have been measured in a number of locations across the
       United States utilizing continuous CO monitoring based upon NDIR spectroscopic detection.
       The NDIR instruments, however, are too bulky and complicated for either indoor or personal
       monitoring.  Over the past decade small, lightweight, and portable CO monitors have been
       developed. These monitors are based primarily on electrochemical detection (see Chapter 5).
15     These highly versatile CO monitors, when equipped with internal or external data loggers,
       have permitted the measurement of personal exposures to CO as well as CO concentrations  in
       a number of indoor environments.
            CO measurements in enclosed spaces have been made either in support of total personal
       exposure studies or in targeted indoor studies. In the personal exposure studies, individuals
20     wear the monitors in the course of their daily activities, taking them through a number of
       different microenvironments. In targeted studies CO measurements are taken  in indoor spaces
       independent of the activities of occupants of those spaces.

       7.3.1  Indoor Concentrations Recorded in Personal Exposure Studies
25          Three studies have reported CO concentrations  in various microenvironments as part of
       an effort to measure total human exposure to CO and to assess the accuracy of exposure
       estimates calculated from fixed-site monitoring data.  In each study, subjects wore personal
       CO exposure monitors for one or more 24-h periods.  CO concentrations were recorded on
       data loggers at varying time intervals as a function of time spent in various
30     microenvironments. A activity diary was kept by participants where they were asked to
       provide information such as time, activity (i.e., cooking), location (microenvironment type),

       March 12, 1990                         7-20     DRAFT - DO NOT QUOTE OR CITE

-------
       presence and use of sources (smokers present, gas stove, etc.), etc.  CO concentrations by
       microenvironment were extracted from the measured concentrations by use of the activity
       diaries. This section will discuss the results of those studies as they relate to the
       concentrations measured in different microenvironments. A discussion of the results as they
 5     relate to total exposure to CO are discussed in Chapter 8.
             Two of the studies, conducted in Denver, CO, and Washington, DC, by EPA (Akland
       et al., 1985; Whitmore et al., 1984; Hartwell et al.,  1984; Johnson, 1984), measured the
       frequency distribution of CO exposure in a representative sample of the urban population.
       The study populations were selected using a multistage sampling strategy.  The third study,
10     also conducted in Washington, DC, (Nagda and Koontz, 1985), utilized a convenience
       sample.
             The first-mentioned Washington study obtained a total of 814 person-day samples for
       1161 participants while the Denver study obtained 899 person-day samples for
       485 participants. The Denver study obtained consecutive 24-h samples for each participant
15     while the Washington study obtained one 24-h sample for each participant.  Both studies  were
       conducted during the winter of 1982-1983.
             A comparison of CO concentrations measured in the Washington and Denver studies is
       shown in  Table 7-7 (from Akland et al., 1985).  Concentrations measured in all
       microenvironments for the Denver  study were higher than those for the Washington study.
20     This is consistent with the finding that daily maximum one- and eight-hour CO concentrations
       at outdoor fixed monitoring sites were about a factor of two higher in the Denver area than
       the Washington area during the course of the studies  (Akland et al., 1985).  The highest
       concentrations in both studies were associated with commuting while the lowest levels were
       measured in indoor environments.  Concentrations associated with commuting are no doubt
25     higher due to the proximity to and density of outside CO sources (cars, buses, and trucks)
       particularly during commuting hours when traffic is heaviest. Indoor levels, especially
       residential levels in the absence of indoor sources, are lower primarily due to the time of day
       of sampling (noncommuting hours with lower outdoor levels). A more detailed breakdown of
       CO concentrations by microenvironments for the Denver study is shown in Table 7-8
30     (Johnson, 1984). Microenvironments associated with motor vehicles result in the highest
       March 12, 1990                          7-21     DRAFT - DO NOT QUOTE OR CITE

-------
                  TABLE 7-7. SUMMARY OF CO EXPOSURE LEVELS AND TIME SPENT PER DAY
KT
i— '
JO
^O
&
o
IN SELECTED MICROENVIRONMENTS
Location



Concentration





§

g
1
i
O
o
55
O
^^
1
0
Microenvironment
Indoors, parking garage
In transit, car
In transit, other (bus, truck, etc.)
Outdoors, near roadway
In transit, walking
Indoors, restaurant

Indoors, office
Indoors, store/shopping mall


Indoors, residence
Indoor, total
"n = number of person-days with nonzero
n
31
643
107
188
171
205

283
243


776
776
durations, x =
X
18.8
8.0
7.9
3.9
4.2
4.2

3.0
3.0


1.7
2.1
mean, SE
Denver. CO
,' ppm
SE
4.96
0.32
0.61
0.36
0.45
0.29

0.20
0.22


0.10
0.09
= standard error.

Median
time, min
14
71
66
33
28
58

478
50


975
1,243


Washington. DC
Concentration,' ppm
n
59
592
130
164
226
170

349
225


705
705

X
10.4
5.0
3.6
2.6
2.4
2.1

1.9
2.5


1.2
1.4

SE
4.43
0.14
0.30
0.20
0.29
0.32

0.27
0.49


0.10
0.08

Median
time, min
11
79
49
20
32
45

428
36


1,048
1,332

^   Source: Akland et al. (1985).

-------
10
15
20
25
30
         TABLE 7-8.  INDOOR MICROENVIRONMENTS LISTED IN DESCENDING ORDER
                         OF WEIGHTED MEAN CO CONCENTRATION
Category
Public garage
Service station or motor vehicle
repair facility
Other location
Other repair shop
Shopping mall
Residential garage
Restaurant
Office
Auditorium, sports arena,
concert hall, etc.
Store
Health care facility
Other public buildings
Manufacturing facility
Residence
School
Church
Number
of Subjects
116

125
427
55
58
66
524
2,287

100
734
351
115
42
21,543
426
179
CO Concentration
Mean
13.46

9.17
7.40
5.64
4.90
4.35
3.71
3.59

3.37
3.23
2.22
2.15
2.04
2.04
1.64
1.56
T ppm
SD
18.14

9.33
17.97
7.67
6.50
7.06
4.35
4.18

4.76
5.56
4.25
3.26
2.55
4.06
2.76
3.35
       Source: Johnson et al. (1984).
       concentrations, with concentrations reaching or exceeding the 9 ppm reference level
       (NAAQS).
35          No statistical difference (p>0.05) in CO concentrations were found for residences with
       and without gas ranges in the Washington study.  The results of a similar analysis on the
       Denver data, according to the presence or absence of selected indoor sources, is shown in
       Table 7-9 (Johnson, 1984).
            Attached garages, use of gas ranges, and presence of smokers were all shown to result
40     in higher indoor CO concentrations. Concentrations were well below the 9 ppm reference
       level (NAAQS), but substantially above concentrations in residences  without the sources.
       March 12,  1990
7-23    DRAFT - DO NOT QUOTE OR CITE

-------
10
15
20
            TABLE 7-9. WEIGHTED MEANS OF RESIDENTIAL EXPOSURE GROUPED
             ACCORDING TO THE PRESENCE OR ABSENCE OF SELECTED INDOOR
                                CARBON MONOXIDE SOURCES

CO Source
Attached garage
Operating gas
stove
Smokers
Carbon
Source
Mean
2.29
4.52
3.48
Monoxide Concentration, oom
Present
SD
5.34
6.10
6.58
Source
Mean
1.88
1.93
1.89
Absent
SD
3.00
3.92
3.69
Difference
in Means
0.41
2.59
1.59
Significance
Level of
ttesf
p< 0.0005
p< 0.0005
p< 0.0005
"Student t test was performed on logarithms of PEM values.
Source: Johnson (1984).
25
30
35
     In the second Washington study (Nagda and Koontz, 1985), a total of 197 person-days
of samples were collected from 58 subjects, representing three population subgroups
(housewives, office workers, and construction workers). A comparison of residential CO
concentrations from that study as a function of combustion sources and whether smoking was
reported is shown in Table 7-10.  Use of gas ranges and kerosene space heaters were found to
result in higher indoor CO concentrations. The statistical significance of the differences was
not given.  Concentrations were highest in microenvironments associated with commuting.
The data collected in this study were consistent with the data collected in the EPA
Washington study discussed above.
     It is difficult for all three studies to assess the contribution to indoor CO concentrations
from either outdoor or indoor sources because concentrations outside each indoor
microenvironment were not measured.
40
       March 12, 1990
                                       7-24     DRAFT - DO NOT QUOTE OR CITE

-------
10
                  TABLE 7-10. AVERAGE RESIDENTIAL CO EXPOSURES (ppm):
             IMPACT OF COMBUSTION APPLIANCE USE AND TOBACCO SMOKING1
                                                      Reported Tobacco Smoking
Appliances
None
Gas stove
Kerosene space heater
Wood burning
Multiple appliances
All cases
No
1.2 (66)
2.2 (15)
5.1 (3)
0.7 (2)
1.0 (1)
1.5 (87)
Yes
1.5 (12)
1.3 (1)
NDb
NDb
NDb
1.5 (13)
All
Cases
1.2
2.2
5.1
0.7
1.0
1.5
(78)
(16)
(3)
(2)
(1)
(100)
15
       "Percentage of subjects' time in their own residences indicated in parentheses for each category of appliance use
        and tobacco smoking.
20     bNo data available.
       Source:  Nagda and Koontz (1985).

25
       7.3.2  Targeted  Microenvironmental Studies
            As demonstrated from the personal exposure studies discussed above, individuals, in the
       course of their daily activities, can encounter a wide range of CO concentrations as a function
30     of the microenvironments in which they spend time.  A number of studies have been
       conducted over the last decade to investigate concentrations of CO in indoor
       microenvironments.  These  "targeted" studies have either focused on indoor CO
       concentrations as a function of the microenvironment or sources in specific
       microenvironments.
35
       7.3.2.1  Indoor Microenvironmental Concentrations
            A summary of the results of the larger studies that have investigated CO levels in
       various indoor environments, independent of the existence of specific indoor sources, is
       shown in Table 7-11. Major foci of these studies are microenvironments associated with
40     commuting.  A wide range of concentrations were recorded in these studies with the highest

       March  12, 1990                         7-25     DRAFT - DO NOT QUOTE OR CITE

-------
                 TABLE 7-11. CARBON MONOXIDE CONCENTRATIONS' MEASURED IN VARIOUS
                     INDOOR ENVIRONMENTS AS A FUNCTION OF MICROENVIRONMENTS
g
ON


Study
Cortege and
Spengler
(1976)


Spengler
et al. (1978)
Colwill and
Hickman
(1980)
Ziskind et al.
(1981)






Wallace (1983)


Holland (1983)







Locations Microenvironment
Boston, MA Autos
Transit
Split
All
Outside*
Boston, MA Seven skating
rinks
London, Autos
England Outside0

Denver, CO Buses
and Taxis
Boston, MA Police cars





One office


Stamford, CT Commercial
Commuting
Residential
Los Angeles, Commercial
CA Commuting
Residential
Average
Time-
Frame of
Sampling
(min)
40-70
40-70
40-70
40-70
40-70
40-160

65-90


-540
-540
-540





Hourly


10-30
10-30
10-30
10-30
10-30
10-30

Number of
Observations
248
28
70
346
1,076
17

11


75
38
19





80


659
1,341
577
1,938
96
807


Inside
Mean
1.34
7.4
8.3
11.9

53.6

25.2


4-36d
lO-lT*
0-46d





19.0


5.8
6.2
2.9
3.3
16.1
7.6
SD
5.4
3.7
2.8
5.5

18.0

7.0










5.9


8.0
4.7
3.9
2.5
5.8
5.0

CO - ppm
Max



>35

192

40


84
48
59





50


61
38
39
61
42
38

Outside Source
Mean SD Identified
Traffic
Traffic
Traffic
Traffic
6.0 4.0 Ambient
fee

Traffic
47.0 13.1

Traffic
Traffic
Traffic





Leakage
from
garage
4.2" 3.0
5.5e 4.1 Traffic
4.3° 3.1
4.0 3.1
5.8 4.2 Traffic
3.9 2.8


Comments
66 volunteers used-some
levels (4%) related to
faulty exhaust


Ventilation measures-
cleaning from CO decay
1 1 driven over a 35 km
route

Only data gathered by
electrochemical
monitors presented.
Passive dosimeter
data not included;
58% of values for
rides >8 h were
>9ppm
65 employees affected;
corrective action taken








-------
£
8*
TABLE 7-11 (cont'd). CARBON MONOXIDE CONCENTRATIONS- MEASURED IN VARIOUS
       INDOOR ENVIRONMENTS AS A FUNCTION OF MICROENVIRONMENTS
i







^
to
"^

V
^
3
I
o

*-/
o
H
O
Q
w
0
»
n

Study



Flachsbart
al. (1987)


Yocom
et al. (1987)


Peterson and
Sabersky (1975)
Chancy
(1978)

Ziskind
et. a. (1982)


Amendola and
Hanes (1984)
Flachsbart and
Ott (1984)



Locations
Phoenix, AZ
Denver, CO


DC


Hartford
CT


Los Angeles
CA
Several
U.S. Cities

Los Angeles
CA


New England

5 California
cities


Average
Time-
Frame of
Sampling
Microenvironment (min)
Commercial 10-30
Commuting 10-30
Residential 10-30
Commercial 10-30
Commuting 10-30
Residential 10-30
Autos 34-69
Bus 82-115
Rail 27-48
2 Garages 3
Public Building
Office Building
Private Home

Autos 3

Autos


Home
Work
Commute

Service Station 480
Car dealership
Enclosed parking 2-5
Bldg. attached to 2-5
enclosed parking
Commercial settings 2-5
Number of
Inside CO - ppm
Observations Mean SD Max
380
839
48
1,949
3,634
528
213
35
8
47




-




564
557
461

81

10
7

202
2.2 2.2 17
6.8 4.9 50
5.8 3.6 17
5.9 4.3 30
11.0 7.7 54
5.6 4.4 45
9 i_22 3 2-9
3.7-10.2 1-7
2.2-5.2 0.5-5
21-94 10-56
1.8-22.7 -
2.1-22.9
1.8-21.9

<2.5 - 45

2-50


4-4.6
2.2-4.3
6.7-10.0 -

2.2-110.8 - 110.8

27.7 12.5
6.1 2.9

2.1 1.6
Outside Source
Mean SD Identified
2.8 2.5
3.9 3.3 Traffic
2.4 2.1
5.0 3.3
5.8 3.7 Traffic
3.1 23
Traffic
— — Traffic
- - Traffic
- - Traffic
Very similar Traffic
to indoor
concentrations

Similar to Traffic
auto levels
Traffic






Autos

3.0-2.6 Autos
Autos

Autos

Comments



Measurements made et
during commuting hours


Two week avenges day
and night over a summer,
fall, and winter period



Slower the traffic
the higher the CO





Higher in Winter
than Summer
Indoor Values have
outdoor concentrations
subtracted


-------
I
I—«
N>
\»

SO
 00
 H
 I
 o
 o
                  TABLE 7-11 (cont'd). CARBON MONOXIDE CONCENTRATIONS' MEASURED IN VARIOUS

                           INDOOR ENVIRONMENTS AS A FUNCTION OF MICROENVIRONMENTS


Study
Sisoric and
Fugas (1985)


Average
Time-
Frame of
* Sampling Number of Inside CO - ppm Outside Source
Locations Microenvironment (min) Observations Mean SD Max Mean SD Identified Comments
Zagreb 8 institutions Winter - 1.1-6.0 0.6-13.7 - - Traffic
Yugoslavia and
summer
periods
"All measurements made with electrochemical devices.

bFixed central station sites.

'Measurement made outside auto.

^9S% confidence limits.

"Average of two fixed sites.
 8
 n

-------
       CO concentrations found in the indoor commuting microenvironments. These concentrations
       frequently are higher than concentrations recorded at fixed-site monitors but lower than
       concentrations measured immediately outside the vehicles.  Concentrations generally are
       higher in automobiles than in public transportation microenvironments. A number of the
 5     studies noted that CO concentrations in commuting vehicles can exceed both the eight-hour,
       9-ppm level and the one-hour, 35-ppm level specified in the NAAQS (1970).  Flachsbart
       et al. (1987) noted that the most important factors influencing CO concentrations inside
       automobiles were such factors as link-to-link variability (a proxy for traffic density, vehicle
       mix, and roadway setting), day-to-day variability (proxy for variations in meteorological
10     factors and ambient CO concentrations), and time of day.  This study  noted that with
       increased automobile speed,  interior CO concentrations decreased.
            Service stations, car dealerships, parking garages and office space that have attached
       garages can exhibit high concentrations of CO due to automobile exhaust.  In one case
       (Wallace, 1983), corrective measures reduced office space CO concentrations originating from
15     an attached parking garage from 19 ppm to approximately 4 ppm.  In  an investigation of
       seven ice skating rinks in the Boston area, one study (Spengler et al.,  1978) reported
       exceptionally high average CO concentrations (53.6 ppm) with a high  reading of 192 ppm.
       Ice-cleaning machines and poor ventilation were found to be responsible.
            Residential and commercial levels generally were found to have low concentrations, but
20     no information was provided on the presence of indoor sources or outdoor levels.

       7.3.2.2  Concentrations Associated with Indoor Sources
            As noted earlier, the major indoor sources of CO in residences are gas ranges and
       unvented kerosene and gas space heaters, with properly operating wood-burning stoves and
25     fireplaces (non-leaky venting system) and tobacco combustion of secondary importance.
       Properly used gas ranges (ranges used for cooking and not  space heating) are used
       intermittently and thus would contribute to short-term peak CO levels  indoors but likely
       would not result in substantial increases in longer-term average concentrations.  Unvented
       kerosene and gas space heaters typically are used for several hours at a time and thus are
30     likely to result in sustained higher levels of CO.  The improper operation of gas ranges or
       unvented gas or kerosene space heaters (e.g.,  low-wick setting for  kerosene heaters or yellow-

       March 12, 1990                          7-29     DRAFT - DO  NOT QUOTE OR CITE

-------
       tipping operation of gas ranges) could result in substantial increases in indoor CO levels.  CO
       levels indoors associated with tobacco combustion are, based upon source emission data,
       expected to be low unless there is a very high smoking density and low ventilation.  In the
       absence of a leaky flue or leaky fire box, indoor CO levels from fire places or stoves should
 5     be low with short peaks associated with charging the fire when some back draft might occur.
            The majority of studies investigating CO concentrations in residences, as a function of
       the presence or absence of a known CO source, typically have measured CO concentrations
       associated with the source's use over short time periods (on the order of a few minutes to a
       few hours). These studies typically have involved fewer than 10 residences and have reported
10     peak CO levels (on the order of minutes). Only two studies (Hartwell et al.,  1988; Koontz
       and Nagda, 1987) have reported long-term average CO concentrations (over several hours) as
       a function of the presence of a CO source for large residential sample sizes, while one study
       (McCarthy et al., 1987) reported longer term average indoor CO concentrations for a small
       sample.
15
       Average Indoor Source Related Concentrations
            As part of a study to determine the impact of combustion sources on indoor air quality,
       a sample of 382 homes in New York State (172 in Onondaga County and 174 in Suffolk
       County) were monitored for CO concentrations during the winter of 1986 (Hartwell et al.,
20     1988). In this study four combustion sources were examined: gas cooking appliances,
       unvented kerosene space heaters, wood-burning stoves and fireplaces, and tobacco products.
       A factorial sample that included all sixteen combinations of combustion sources was utilized.
       CO concentrations were monitored in the main living area (e.g., family room) and source
       area (e.g., in the kitchen for homes with gas ranges) for each home over a three-day period
25     using an electrochemical monitor with the data stored on a data logger.  Outdoor CO levels
       were not recorded for these homes. CO concentrations were reported as averages for the full
       three-day  period of measurement.
            Average CO concentrations measured in the main living area as a function of county and
       the presence or absence of a combustion source are shown  in Table 7-12.  Gas ranges and
30     kerosene heaters were found to result in small increases in average CO levels.  Use of a
       wood-burning stove or fireplace resulted in lower average CO levels, presumably due to

       March 12, 1990                          7-30     DRAFT - DO NOT QUOTE OR CITE

-------
10
15
20
25
30
35
40
                   TABLE 7-12. WEIGHTED SUMMARY STATISTICS FOR CO
              CONCENTRATIONS (ppm) IN THE MAIN LIVING AREA BY USE FOR
                              SELECTED SOURCES BY COUNTY
Source
and
County
Source
Present
Sample Percent
Size Detected
Arith.
Mean
SE
Geo.
Mean
Geo.
SE
KEROSENE HEATER
Onondaga
Suffolk
Yes
No
Yes
No
10
198
16
158
89.3*
60.0
100.0'
72.1
3.33
1.72
3.86*
2.03
1.34
0.15
0.73
0.15
2.20
1.29
3.35*
1.62
1.33
1.06
1.22
1.07
WOOD-BURNING STOVE/FIREPLACE
Onondaga
Suffolk
GAS STOVE
Onondaga
Suffolk
Yes
No
Yes
No

Yes
No
Yes
No
39
169
33
141

90
118
86
88
44.5
62.9
82.7
72.7

77.4*
47.0
82.8
68.2
1.04
1.86*
1.93
2.24

2.29*
1.33
2.55*
1.91
0.09
0.16
0.23
0.17

0.24
0.17
0.21
0.19
0.93
1.37*
1.72
1.72

1.74*
1.04
2.04*
1.51
1.09
1.06
1.14
1.07

1.08
1.07
1.10
1.09
45
       "Significantly different at .05 level.

       Source: Hartwell et al. (1988).
increased air-exchange rates associated with use. The study found no effect on average CO
levels with tobacco combustion and no difference by location in the residence. The data base

has not yet been analyzed for differences in short-term CO concentrations (eight-hour, one-
hour, or less than one-hour concentrations) as a function of sources and source use. When
       March 12, 1990
                                     7-31     DRAFT - DO NOT QUOTE OR CITE

-------
       such an analysis is made available the impact of the combustion sources on residential CO
       levels will be much more pronounced.
            Koontz and Nagda (1987 and 1988) utilizing Census data for sample selection,
       monitored 157 homes in 16 neighborhoods in North Central Texas over a nine-week period
 5     between January and March 1985.  Unvented gas space heaters were used as the primary
       means of heating in 82 residences (13 had one UVGSH, 36 had two UVGSHs, and 33 had
       three or more UVGSHs) and as a secondary heat source in 29 (17 had one UVGSH and 12
       had two or more UVGSHs). There was no gas space heater present or used in 41 of the
       homes and 5 of the homes were not included for various reasons (e.g., air sample lost). A
10     majority of all the homes in each UVGSH use category had gas ranges and gas water heaters
       (typically greater than 80%). Air samples were collected for all residences on two separate
       occasions over integrating periods of approximately 15 h using Collectaire samplers.  CO
       concentrations in each sampler then were measured using a electrochemical monitor.  In 30%
       of the residences (46 residences) CO was monitored continuously, consisting of sequential 15-
15     min averages over an average monitoring period of about five days with an electrochemical
       monitor.  Measurements were made close to the geometric center of the house.
            The cumulative frequency distributions for the first integrated CO measurements by
       source category are shown in Figure 7-1.  Residences where UVGSHs are the primary heat
       source exhibited the highest CO concentrations. CO concentrations were greater than or
20     equal to 9 ppm in 12% of the homes with the highest concentration measured at 36.6 ppm.
       No values were measured above 9 ppm for residences where a UVGSH was not used at all or
       was used as a secondary heat source.  The second CO sample produced summary statistics
       virtually identical to the first.  Table 7-13 presents a comparison of the CO concentrations
       measured in the continuously monitored residences (15-min average concentrations summed
25     by one- and eight-hour time periods) with the one-hour, 35 ppm and eight-hour, 9 ppm CO
       standard by source category. The table also presents the mean concentrations measured in
       these home over the full five-day periods.  Five of the residences exceeded the one-hour,
       35 ppm level while seven of the residences exceeded the eight-hour, 9 ppm level. Higher CO
       levels were associated with maltuned unvented gas appliances and the use of multiple
30     unvented gas appliances.
       March 12,  1990                         7-32     DRAFT - DO NOT QUOTE OR CITE

-------
  0
  o
  z
  o
  a.
100-

 90-

 80-

 70-

 60-

 50-

 40-

 30-

 20-

 10-

 0
Meant Standard Deviation:

O Non-UVQ8H—2.211.7
O Secondary UVQSH—2.9±1.6
A Primary UVQSH—5.5± 8.0
                                 T-
                                  8
                                  i	
                                 10

                              CO.ppm
         —r-
         12
—r—
 14
—r~
 16
—T~
 18
                                                                     20
Figure 7-1. Cumulative frequency distributions and summary statistics for indoor CO
concentrations in three groups of monitored homes.

Source: Koontz and Nagda (1988).
March 12, 1990
                            7-33    DRAFT - DO NOT QUOTE OR CITE

-------
            TABLE 7-13. SUMMARY OF CONTINUOUS CO MONITORING RESULTS
                                  BY HEATING EQUIPMENT
5


10


15

Heating
Equipment
Primary UVGSH
Secondary UVGSH
Non-UVGSH

Number
of
Homes
26
11
9


Number
1-h, 34 ppm
4
1
0


Exceeding
8-h, 9 ppm
5
0
2

CO
Concentration.
Mean
6.2
2.3
2.2


ppm
SD
7.6
1.1
1.2

       Source: Koontz and Nagda (1988).

20
            In a study of 14 homes with one or more unvented gas space heaters (primary source of
       heat) in the Atlanta, GA, area, McCarthy et al. (1987) measured CO levels by continuous
       NDIR monitors in two locations in the homes (room with the heater and a remote room in the
       house) and outdoors. Measurements were taken over five-minute periods in turn from each of
25     the three sampling points for each house over 96-h sampling periods. The authors reported
       only the summary statistics for CO (average 96-h concentrations) based on the continuously
       collected data in the room with the heater and outdoors.  One out of the 14 UVGSH homes
       exceeded 9 ppm during the sampling period.   Mean indoor values ranged 0.26 ppm to
       9.49 ppm and varied as a function of the use pattern of the heater. Only one of the homes
30     used more than one heater during the air sampling.  Outdoor concentrations varied from
       0.3 ppm to 1.6 ppm.

       Peak Indoor Source-Related Concentrations
            Short-term or peak  CO concentrations indoors associated with specific sources were
35     obtained for a few field studies. The peak CO concentrations measured in these studies, by
       location in the house and presence of specific sources, are shown  in Table 7-14. A wide
       range of peak CO concentrations were observed in these studies between and among
       residences with different  indoor CO sources.  The highest concentrations measured

       March 12, 1990                         7-34    DRAFT -  DO NOT  QUOTE OR CITE

-------
sr
to
H- »









&

u
§>
H
I
O
O
§
H
O
C
o
a
g
1AJ


Reference
Hartwell
etal.
(1988)





Koontz
etal.
(1987)


Leaderer
etal.
(1984)





BLJ1 /-14. FilAJk Ut

Indoor
Source"
GR





K

UVGSHP, GR

UVGSHS, GR
GR

GR


CK


RK

) UUINUtUNlK

Number of
Residents
12





1

26

11
9

1


8


5

A11UINS JSI


Location1"
K

LR, D, B



K
D
C

C
C

A

B
LR
B

LR
B
JUNUUUKJS
Averaging
Time
(min)
30

30



30
30
15

15
15

15

15
5
5

5
5
iUUKUtl JVUVASUKJtJJ JLN 1*11

CO Concentration (ppm)
Peaks Outside
1.8 -> 100 0.7-10

1.8- 17



5.7 5.0
9.7 5.0
?-69

?-69
?-26

3.5

3.0
0-3.2
0-3.4

2.1 -21.1
4.8-8.2
1LJJ MULUca


Comments
Excluding one house with
> 100 range in kitchen is
1.8 - 15, wood stoves
and smokers were present
in same houses but no
effect was seen


Houses may contain more
than one heater



Outdoor levels subtracted


Outdoor levels subtracted


Outdoor levels subtracted


-------
            TABLE 7-14 (cont'd).  PEAK CO CONCENTRATIONS BY INDOOR SOURCE MEASURED IN FIELD STUDIES
»
o
o
1
o
§
w
g
Indoor
Reference Source"
Lebret GA
etal.
(1987)

Brunekreef GA
etal.
(1982)
Moschandreas GR
and
Zabransky
(1982)
Sterling and GR
Sterling
(1979)

Number of
Residents Location1"
12 K

LR
B
254 K


8 K

LR
B
9 K



Averaging
Time CO Concentration (ppm)
(min)
1

1
1
15


60

60
60
2



Peaks Outside
4.0-

3.3-
3.3-
<10


7.2-

1.0-
1.0-
29-



90

23
40
- >600


11.3

12.6
13.0
120 3.0 - 8.5



Comments
Sample of Dutch homes



Sample of Dutch homes,
breathing zone samples,
levels related to geisers




Measurements were taken
under a variety of gas
range operating and
ventilation conditions
"GR = gas range, K = unvented kerosene space heater, CK = unvented convective kerosene space heater, RK = radiant unvented kerosene space heater,
 UVGSHP = unvented gas space heater used as primary heat source, UVGSHS = UVGSH as a secondary heat source, GA = gas appliances includes geisers
 (water heaters).
bK = kitchen, LR = living room, D = den, B = bedroom.

-------
        (>600 ppm) were associated with emissions from geisers (water heaters), found in a large
        study conducted in The Netherlands (Brunekreef et al.,  1982).  Peak levels of CO associated
        with gas ranges were from 1.0 to over 100 ppm. This broad range is somewhat consistent
        with the range of CO emissions observed in studies evaluating CO emissions from gas ranges
  5     (i.e., Table 7-1).  The variability is in part due to number of burners used, flame condition,
        condition of the burners, etc.  As might be expected radiant kerosene heaters produced higher
        CO concentrations than convective heaters. UVGSHs generally were associated with higher
        CO peaks than gas ranges or kerosene heaters. As noted earlier the peaks associated with gas
        or kerosene heaters are likely to be sustained over longer periods of time because of the long
 10     source use times.
             Test houses have been used by investigators to evaluate the impact of specific sources,
        modifications to sources, and variations in their use on residential peak CO concentrations.
             In  one of the earliest investigations of indoor air quality, Wade et al. (1975) measured
        indoor and outdoor CO levels in four houses that had gas stoves. Using an NDIR analyzer,
 15     indoor concentrations were found to range from 1.7 to 3.8 times higher than the outdoor
        levels.  CO levels in one house exceeded 9 ppm, the NAAQS reference level. A time history
        of CO measured in one house is shown in Figure 7-2. For this house and for the time
        averaging period used, CO was well mixed through the house. As part of a modeling study of
        emissions from a gas range, Davidson et al. (1987)  measured CO concentrations in three
20     residences.  Peak CO levels in excess  of 5 ppm were measured in one town house.
             Indoor CO levels associated with wood-burning stoves were measured in two test house
        studies.  In one study (Humphreys et al. 1986) indoor CO levels associated with the use of
        both airtight (conventional and catalytic) and non-airtight wood heaters were evaluated in a
        337 m3 weatherized home.  Indoor CO concentrations were higher than outdoor levels for all
25      tests. Conventional airtight stoves produced indoor CO levels typically about 1 to 2 ppm
        above background level with a peak concentration of 9.1 ppm.  Use of non-airtight  stoves
        resulted in average indoor CO concentrations 2 to 3 ppm above outdoors with peak
        concentrations as high as 29.6 ppm. In a 236 m3 house (Traynor et al., 1984), four wood-
        burning stoves (three airtight and one non-airtight) were tested.  The airtight stoves  generally
30      resulted in small contributions to both  average and peak indoor CO levels (0.1 to 1 ppm for
       March 12, 1990                          7-37     DRAFT - DO NOT QUOTE OR CITE

-------
                                	Kitchen over stove
                                	Kitchen i meter from stove
                                	Living room
                                	Outside
                     on  i
                    14 rmnj
                   1 burner on |
                    3 mm -
                                      1 burner on 1
                                        7mm.— J
jurrer on
3 mm. —^
t b
 6 mm
                                                              1 burner on 1
                                                               5 mm — .
                      1 burner onj
                        3 mm -*-4
                         1 burner on I    _J L2«n OP
              13000V-      11 mm.—1      ' '  55 mm.
              12000 |
              11000^
             .20000
             M 9000
             JJ 8000
            |  7000
             2  6000
             §  5000-
             §  4000 -
            8  3000-
               2000
               1000-
                    0400  0800  1200  1600  2000  2400  0400  0800  1200
                               J/31     Time, hrs       2/01

Figure 7-2.  A time history of CO concentrations, 2-hour averages, winter of 1974.

Source: Wade et al. (1975).
March 12, 1990
                                      7-38      DRAFT - DO NOT QUOTE OR CITE

-------
       the average and 0.2 to 2.7 ppm for the peak). The non-airtight stove contributed as much as
       9.1 ppm to the average indoor level and 43 ppm to the peak.

       Indoor Concentrations Related to Environmental Tobacco Smoke
 5          Carbon monoxide has been measured extensively in chamber studies as a surrogate for
       environmental tobacco smoke (e.g., Bridge and Corn, 1972; Hoegg, 1972; Penkala and
       De Oliveira, 1975; Weber et al., 1976, 1979a,b; Weber, 1984; Leaderer et al., 1984;
       Clausen et al., 1985).  Under steady-state conditions in chamber studies, where outdoor CO
       levels are monitored and the tobacco brands and smoking rates are controlled, CO can be a
10     reasonably good indicator of environmental tobacco smoke (ETS) and is used as such.  Under
       such chamber conditions CO concentrations typically range from less than 1 to over 10 ppm.
            A number of field studies have monitored CO in different indoor environments with and
       without smoking occupants.  A summary of the results of these studies is shown in
       Table 7-15 (National Research Council, 1986, Table 2-4).  Although CO concentrations
15     generally were higher in indoor spaces when smoking occurred, the concentrations were
       highly variable. The variability of CO production from  tobacco combustion, number of
       cigarettes smoked, and differences in ventilation and variability of outdoor concentrations
       make it difficult to assess the contribution of tobacco combustion in indoor CO
       concentrations.  The chamber studies and field studies conducted do indicate that under
20     typical smoking conditions encountered in residences or  offices, CO concentrations can be
       expected to be above background outdoor levels, but lower than the levels resulting from
       other unvented combustion sources. In indoor spaces where heavy smoking occurs and in
       small indoor spaces CO emissions from tobacco combustion will be an important contributor
       to CO concentrations.
25
       7.3.3  Spatial Concentration Variations
            Spatial variations of CO concentrations within a space are a function of mixing within
       and between spaces. Spatial variations of CO in a space are likely to be minimized if a
       continuous or nearly continuous source of CO exists (i.e., unvented kerosene or gas space
30     heater) due to the  strong convective currents  which enhance rapid mixing. Intermittent
       sources (i.e., gas burner use or tobacco combustion) are likely to produce a more pronounced

       March 12, 1990                          7-39     DRAFT - DO NOT QUOTE OR CITE

-------
TABLE 7-15. MEASURED CONCENTRATIONS OF CARBON MONOXIDE
           IN ENVIRONMENTAL TOBACCO SMOKE"
i_« Indoor
•O
°













^j
j^.
O


o
^m*t
5*
3
i
i





*Q
Q
H
tn
2
Jo
O

Location
Rooms
Train

Submarines
(66m3)
18 military
aircraft
8 commercial
aircraft
Rooms
14 public places
Ferry boat
Theater foyer
Intercity bus

2 conference rooms
Office

Automobile

9 night clubs
14 restaurants
45 restaurants
33 stores
3 hospital lobbies
6 coffee houses
Room
Hospital lobby
2 train
compartments
Automobile

10 offices

15 restaurants

14 night clubs
and taverns
Tobacco Burned
-
1-18 smokers

157 cigarettes/day
94 - 103 cigarettes/day
-

—

—
-
-
-
23 cigarettes
3 cigarettes
-
-
-
2 smokers
(4 cigarettes)
-
-
-
-
-
Varied
18 smokers
12-30 smokers
2-3 smokers

3 smokers
2 smokers
—

—

-

Ventilation
-
Natural

Yes
Yes
Yes

Yes

-
-
-
-
15 changes/h
15 changes/h
8 changes/h
236m3/h
Natural
Natural
Mechanical
Varied
-
-
-
-
-
-
-
-

Natural, open
Natural, closed
-

-

-

Mean
—
-

<40
<40
<2-5

<2

-
<10
18.4 ± 8.7
3.4 ± 0.8
32
18
-
-
-
-
-
13.4
9.9 ± 5.5
8.2 ± 2.2
10.0 ± 4.2
-
2-23
50
5
-

14
20
2.5 ± 10

4.0 ± 2.5

13.0 ± 7.0

Range
4.3 -9
0-40

-
-
-

-

5-25
-
-
—
-
—
8 (peak)
<2.5 - 4.6
<2.9 - 9.0
42 (peak)
32 (peak)
6.5-41.9
-
7.1 ± 1.7
11.5 ± 6.5
4.8
-
—
-
4-5

-
-
1.5 - 1.0

1.0-9.5

3.0 - 29.0

Outdoor
Mean
2.2 ± 0.98
-

-
-
-

-

-
-
3.0 ± 2.4
1.4 ± 0.8
-
—
1 -2
-
—
—
-
-
9.2 (outdoor)
—
11.5 ±6.5
-
-
_
-
-

-
-
2.5 ± 1.0

2.5 ± 1.5

3.0 ± 2.0

Range References
0.4 - 4.5 Coburn et al. (1965)
— Harmsenand Eflenberger
(1957)
Cano et al. (1970)
—
U.S. Department of
Transportation (1971)
- U.S. Department of
Transportation (1971)
Porthein (1971)
Perry (1973)
- Godin et al. (1972)
Godin et al. (1972)
Seiff(1973)
_
Slavin and Hertz (1975)
Harke (1974)
— _
13.5 (peak) HarkeandPeters(1974)
15.0 (peak)
Sebben et al. (1977)
3.0-35 Sebben et al. (1977)
- Sebben et al. (1977)
Sebben et al. (1977)
Sebben et al. (1977)
Badre et al. (1978)
Badre et al. (1978)
Badre et al. (1978)
Badre et al. (1978)

Badre et al. (1978)
- Badre et al. (1978)
1 .5 ± 4.5 Chappell and Parker
(1977)
1 .0 - 5 .0 Chappell and Parker
(1977)
1 .0 - 5 .0 Chappell and Parker
(1977)

-------
I
 i
O
O
1
                           TABLE 7-15 (cont'd). MEASURED CONCENTRATIONS OF CARBON MONOXIDE
                                               IN ENVIRONMENTAL TOBACCO SMOKE'
Indoor
Location
Tavern



Office

Restaurant
Restaurant
Bar
Cafeteria
44 offices
25 offices

Tavem
Tavern
Tobacco Burned Ventilation
- Artifical

— None

— Natural, open

— Mechanical
- Natural
— Natural, open
— 11 changes/h
— —
_ —

v- 6 changes/h
— 1-2 changes/h
Mean
8.5

—

1.0

5.1
2.6
4.8
1.2
1.1
2.78 ± 1.42

11.5
12.0
Range
—

35 (peak)

10.0 (peak)

2.1-9.9
1.4-3.4
2.4 - 9.6
0.7 - 1.7
6.5 (max)
—

10- 12
3 -22
Outdoor
Mean Range
_ _

- -

- -

4.8 (outdoors) -
1.5 (outdoors) —
1.7 (outdoors) -
0.4 (outdoors) -
- -
2.59 ± 2.33

2 (outdoors) -
— —
References
Chappell and Parker
(1977)
Chappell and Parker
(1977)
Chappell and Parker
(1977)
Fischer et al. (1978)
Fischer et al. (1978)
Weber et al. (1976)
Weber et al. (1976)
Weber (1984)
Szadkowski and Harke
(1976)
Cuddebacketal. (1976)
Cuddebacketal.(1976)
Time-weighted average (TWA) of carbon monoxide, 50 ppm (55 mg/m3). TWA = average concentration to which worker may be exposed continuously for 8 hr without damage to health (National
Institute for Occupational Safety and Health, 1971).

Source: NRC (1986) Table 2-4.
I
r
5

-------
        spatial gradient. The within-home spatial variations are related to such variables as air-
        exchange rates among rooms, air mixing within a room, volume of a house, and location and
        use of the source.  The question of the spatial variability of CO indoors as a function of
        different indoor sources has not been evaluated in any detail in any field study.
 5
        7.3.4  Summary of Indoor Concentrations
            Indoor concentrations of CO are a function of outdoor concentrations, indoor sources
        (source type, source condition, source use, etc.), infiltration/ventilation, and air mixing
        between and within rooms. In residences without sources, average CO concentrations are
10      approximately equal to average outdoor levels.  Proximity to outdoor sources (i.e., structures
        near heavily traveled roadways or with attached garages or parking garages) can have a major
        impact on indoor CO concentrations.
            The development of small lightweight and portable electrochemical CO monitors over
        the past decade has permitted the measurement of personal CO exposures and CO
15      concentrations in a number of indoor environments. The  available data on indoor CO
        concentrations have been obtained from total personal exposure studies or studies where
        various indoor environments have been targeted for measurements.
            The extensive total personal CO exposure studies conducted by EPA in Washington,
        DC, and Denver, CO, have shown that the highest  CO concentrations occur in indoor
20      microenvironments associated with transportation sources  (parking garages, cars, buses, etc.).
        Concentrations  in these environments were found to frequently exceed 9 ppm.  Studies
        targeted toward specific indoor microenvironments  also have identified the indoor commuting
        microenvironment as an environment in which CO  concentrations frequently exceed 9 ppm
        and occasionally exceed 35 ppm. Special environments or occurrences (indoor ice skating
25      rinks, offices where emissions from parking garages migrate indoors, etc.) have been reported
        where indoor CO levels can exceed the current ambient one- and eight-hour standards (9 and
        35 ppm, respectively).
            A majority of the targeted field studies monitored indoor CO levels as a function of the
        presence or absence of combustion sources (gas ranges, unvented gas and kerosene space
30      heaters, wood burning stoves and fireplaces, and tobacco combustion).  The results of these
        studies indicate  that the presence and use of a unvented combustion source results in indoor

        March 12,  1990                          7-42    DRAFT - DO NOT QUOTE OR CITE

-------
        CO levels above those found outdoors. The associated increase in CO concentrations can
        vary considerably as a function of the source, source use, condition of the source, and
        averaging time of the measurement.  Intermittent sources such as gas cooking ranges can
        result in high peak CO concentrations (in excess of 9 ppm), while long-term average
 5      concentrations (i.e.,  24-h) associated with gas ranges are considerably lower (on the order of
        1 ppm).  The contribution of tobacco combustion to indoor CO levels is variable.  Under
        conditions of high smoking and low ventilation the contribution can be on the order of a few
        parts per million. One study suggested that the contribution to residential CO concentrations
        of tobacco combustion is on the order of 1 ppm while another study showed no significant
10      increase in residential CO levels.
            Unvented combustion sources that are used for substantial periods of time (i.e.,
        unvented gas and kerosene space heaters) appear to be the major contributors to residential
        CO concentrations.   One extensive study of unvented gas space heaters indicated that 12% of
        the homes had  15-h  average CO concentrations greater than 8 ppm  with the highest
15      concentration at 36.6 ppm.  Only very limited data are available on the contribution of
        kerosene heaters to the average CO concentrations in residences and these data indicate a
        much lower contribution than gas heaters. Peak CO concentrations associated with both
        unvented gas and kerosene space heaters often exceed the current ambient one- and eight-
        hour, standards (9 and 35 ppm, respectively) in residences, and due to the nature of the
20      source (continuous) those peaks tend to be sustained for several hours.
            Very limited data on CO levels in residences with wood-burning stoves or fireplaces is
        available. Non-airtight stoves can  contribute substantially to residential CO concentrations,
        while airtight stoves  can result in small increases.  The available data indicate that fireplaces
        do not contribute measurably to average indoor concentrations. No information is available
25      for samples of residences with leaky  flues.  In addition, there is no  information available on
        short-term indoor CO levels associated with these sources nor are there studies that examine
        the impact of attached garages on residential CO concentrations.
            The available data on short-term (one-hour) and long-term (eight-hour) indoor CO
        concentrations as a function of microenvironments and sources in those microenvironments
30      are not adequate to assess exposures in those environments.  In addition, little is known about
        March 12, 1990                          7-43     DRAFT - DO NOT QUOTE OR CITE

-------
         the spatial variability of CO indoors.  These indoor microenvironments represent the most
         important CO exposures for individuals and as such need to be characterized better.
       7.4  REFERENCES
       Akland, G. G.; Hartwell, T. D.; Johnson, T. R.; Whitmore, R. W. (1985) Measuring human exposure to carbon
10            monoxide in Washington, D.C., and Denver, Colorado, during the winter of 1982-1983. Environ. Sci.
              Technol.  19: 911-918.

       Amendola, A. A.; Hanes, N. B. (1984) Characterization of indoor carbon monoxide levels produced by the
              automobile. In: Berglund, B.; Lindvall, T.; Sundell, J., eds. Indoor air: proceedings of the 3rd
15            international conference on indoor air quality and climate, v. 4, chemical characterization and personal
              exposure; August; Stockholm, Sweden. Stockholm, Sweden: Swedish Council for Building Research;
              pp. 97-102. Available from: NTIS, Springfield, VA;  PB85-104214.

       American National Standards Institute.  (1982) American national standard for household cooking gas appliances.
20            ANSI Z21-1982.

       Badre, R.; Guillerme, R.; Abram, N.;  Bourdin, M.; Dumas, C. (1978) Pollution atmospherique par la fumee de
              -tabac [Atmospheric pollution by smoking]. Ann.  Pharm. Fr. 36: 443-452.

25     Belles, F.  E.; DeWerth, D. W.; Himmel, R. L. (1979) Emissions of furnace and water heaters predominant to
              the California market. Presented at: 72nd annual meeting and exhibition of the Air  Pollution Control
              Association; June; Cincinnati, OH. A.G.A.L.

       Borrazzo, J. E.; Osborn, J.  F.; Fortmann, R. C.; Keefer, R.  L.; Davidson,  C. I. (1987) Modeling and
30            monitoring of CO, NO and NO2 in a modern townhouse. Atmos. Environ.  21: 299-311.

       Bridge, D. P.; Corn, M. (1972) Contribution to the assessment of exposure  of nonsmokers to air pollution from
              cigarette and cigar smoke in occupied spaces. Environ.  Res. 5: 192-209.

35     Brunekreef, B.; Smit, H. A.; Biersteker, K.; Boleij, J. S. M.; Lebret, E. (1982) Indoor carbon monoxide
              pollution in The Netherlands. Environ.  Int. 8: 193-196.

       Caceres, T.; Soto, H.; Lissi, E.; Cisternas, R. (1983) Indoor house pollution: appliance emissions and indoor
              ambient concentrations. Atmos. Environ. 17: 1009-1014.
40
       Cano, J. P.; Catalin, J.; Badre, R.; Dumas, C.; Viala, A.; Guillerme, R.  (1970) Determination de la nicotine par
              chromatographic en phase gazeuse. II. Applications. Ann. Pharm.  Fr. 28: 633-640.

       Chaney, L. W. (1978) Carbon monoxide automobile emissions  measured from the interior of a traveling
45            automobile. Science (Washington, DC)  199: 1203-1204.
50
       Chappell, S. B.; Parker, R. J. (1977) Smoking and carbon monoxide levels in enclosed public places in New
              Brunswick. Can. J. Public Health 68:  159-161.
        March 12, 1990                              7-44      DRAFT - DO NOT QUOTE OR CITE

-------
        Clausen, G. H.; Fanger, P. O.; Cain, W. S.; Leaderer, B. P. (1985) The influence of aging, particle filtration
               and humidity on tobacco smoke odor. In: Fanger, P. O., ed. CLIMA 2000: proceedings of the world
               congress on heating, ventilating and air-conditioning, v. 4, indoor climate; August, Copenhagen,
               Denmark. Copenhagen, Denmark: VVS Kongres; pp. 345-350.
 5
        Cobum, R. F.; Forster, R. E.; Kane, P. B. (1965) Considerations of the physiological variables that determine
               the blood carboxyhemoglobin concentration in man. J.  Clin. Invest. 44: 1899-1910.

        Cole, J. T.; Zawacki, T. S. (1985) Emissions from residential  gas-fired appliances. Chicago, IL:  Institute of Gas
10            Technology; report no. 84/0164.

        Colwill, D. M.; Hickman, A. J. (1980) Exposure of drivers to carbon monoxide. J. Air Pollut. Control Assoc.
               30: 1316-1319.

15     Cortese, A. D.; Spengler, J. D. (1976) Ability of fixed monitoring stations to represent personal  carbon
               monoxide exposures. J. Air Pollut. Control Assoc. 26: 1144-1150.

        Cote, W. A.; Wade, W. A., HI; Yocom, J. E. (1974) A study of indoor air quality. Washington, DC: U. S.
               Environmental Protection Agency, Office of Research and Development; EPA report no.
20            EPA-650/4-74-042. Available from: NTIS, Springfield, VA; PB-238556.

        Cuddeback, J. E.; Donovan, J. R.; Burg, W. R. (1976) Occupational aspects of passive smoking. Am. Ind. Hyg.
               Assoc.  J. 35: 263-267.

25     Dames; Moore. (1986) 50-home weatherization and indoor air  quality study - summary report. Final report to
               Wisconsin Power and  Light Company.

        Fischer, T.; Weber, A.; Grandjean, E. (1978) Luftverunreinigung durch Tabakrauch in Gaststaetten [Air
               pollution due to tobacco smoke in restaurants]. Int. Arch.  Occup. Environ. Health 41: 267-280.
30
        Flachsbart, P. G.; Ott, W. R. (1984) Field surveys of carbon monoxide  in commercial settings using personal
               exposure monitors. Washington, DC: U. S. Environmental Protection Agency, Office of Monitoring
               Systems and Quality Assurance; EPA report no. EPA-600/4-84-019. Available from: NTIS, Springfield,
               VA; PB84-211291.
35
        Flachsbart, P. G.; Mack, G. A.; Howes, J. E.;  Rodes, C. E. (1987) Carbon monoxide exposures of Washington
               commuters.  J. Air  Pollut. Control Assoc. 37: 135-142.

        Fortmann,  R. C.; Borrazzo, J. E.; Davidson, C. I. (1984) Characterization of parameters influencing indoor
40            pollutant concentrations. In: Berglund, B.; Lindvall, T.; Sundell, J., eds.  Indoor air: proceedings of the
               3rd international conference on indoor air quality and climate, v. 4, chemical characterization and
               personal exposure;  August; Stockholm, Sweden. Stockholm, Sweden: Swedish Council for Building
               Research; pp.  259-264. Available from:  NTIS, Springfield, VA;  PB85-104214.

45     Girman, J. R.;  Apte, M. G.; Traynor, G. W.; Allen, J. R.; Hollowell, C. D.  (1982) Pollutant emission rates
               from indoor combustion appliances and sidestream cigarette smoke. Environ. Int.  8: 213-221.

        Godin,  G.; Wright,  G.; Shephard, R. J. (1972) Urban exposure to carbon monoxide. Arch. Environ. Health
               25: 305-313.
50
        Goto, Y.; Tamura, G. T. (1984) Measurement of combustion products from a gas cooking stove in a two-storey
               house. Presented at: 77th annual meeting of the Air Pollution Control Association; San Francisco, CA.
               Pittsburgh, PA: Air Pollution Control Association; paper no. 84-32.5.
         March 12,  1990                              7-45      DRAFT - DO NOT QUOTE OR CITE

-------
        Harke, H.-P. (1974) Zum Problem des Passivrauchens: I. ueber den Einfluss des Rauchens auf die
              CO-Konzentration in Buroraumen [The problem of passive smoking: I. the influence of smoking on the
              CO concentration in office rooms]. Int. Arch. Arbeitsmed. 33: 199-206.

 5      Harke, H.-P.; Peters, H.  (1974) Zum Problem des Passivrauchens:  IE. ueber den Einfluss des Rauchens auf die
              CO-Konzentration im Kraftfahrzeug bei Fahrten im Stadtgebiet [The problem of passive smoking: III. the
              influence of smoking on the CO concentration in driving automobiles]. Int. Arch. Arbeitsmed.
              33: 221-229.

10      Harmsen, H.; Effenberger, E. (1957) Tabakrauch in Verkehrsmitteln, Wohn- und Arbeitsraumen. Arch. Hyg.
              Bakteriol. 141:  383-400.

        Hartwell, T.  D.; Clayton, C. A.; Ritchie, R. M.; Whitmore, R. W.;  Zelon, H. S.; Jones, S. M.; Whitehurst,
              D. A. (1984) Study of carbon monoxide exposure of residents of Washington, DC and Denver, Colorado.
15            Research Triangle Park, NC: U. S.  Environmental Protection Agency,  Environmental Monitoring
              Systems Laboratory; EPA report no. 600/4-84-031. Available from: NTIS, Springfield, VA;
              PB84-183516.

        Hartwell, T., et al. (1988) An investigation of infiltration and indoor  air quality: determination of the impact of
20            indoor combustion sources on indoor air quality - final report. Submitted to the New York State Energy
              Research and Development Authority.

        Himmel, R. L.; DeWerth, D. W. (1974) Evaluation of the pollutant emissions from gas-fired ranges. American
              Gas Association Laboratories; report no. 1492.
25
        Hoegg, U. R. (1972) Cigarette smoke in closed spaces. EHP Environ. Health Perspect.  2: 117-128.

        Holland, D. M. (1983) Carbon monoxide levels in microenvironment types of  four U.S. cities. Environ. Int.
              9: 369-378.
30
        Humphreys, M. P.; Knight, C. V.; Pinnix, J. C. (1986) Residential wood combustion impacts on indoor carbon
              monoxide and suspended participates. In: Proceedings from the 1986 EPA/APCA symposium on the
              measurement of toxic air pollutants; April; Raleigh, NC. Research Triangle Park, NC: U. S.
              Environmental Protection Agency, Environment Monitoring Systems Laboratory; pp. 736-747;  EPA
35            report no. EPA-600/9-86-013. Available from: NTIS, Springfield, VA; PB87-182713.

        Johnson, T. (1984) A study of personal exposure to carbon monoxide in Denver, Colorado.  Research Triangle
              Park, NC: U. S. Environmental Protection Agency, Environmental Monitoring Systems Laboratory; EPA
              report no. EPA-600/4-84-014. Available from: NTIS, Springfield, VA; PB84-146125.
40
        Johnson, D.; Billick, I.; Moschandreas, D.; Relwani, S. (1984) Emission rates from unvented gas appliances. In:
              Berglund, B.; Lindvall, T.; Sundell, J., eds. Indoor  air: proceedings of the 3rd  international conference
              on indoor air quality and climate, v. 4, chemical characterization and personal exposure; August;
              Stockholm, Sweden. Stockholm, Sweden: Swedish Council  for Building Research; pp.  367-373. Available
45            from: NTIS, Springfield, VA; PB85-104214.

        Jones, K. H., et al. (1983) Kerosene heater indoor air quality study -  monitoring protocol and preliminary results.
              Weston report prepared for Kerosun, Inc.

50      Koontz, M. D.; Nagda, N. L. (1987) Survey of factors affecting NO2 concentrations. In: Seifert, B.; Esdorn, H.;
              Fischer, M.; Rueden, H.; Wegner, J.,  eds.  Indoor air '87: proceedings of the 4th international  conference
              on indoor air quality and climate, v. 1, volatile organic compounds, combustion gases, particles and
              fibres, microbiological agents; August; Berlin, Federal Republic of Germany. Berlin, Federal Republic of
              Germany: Institute for Water, Soil, and Air Hygiene; pp. 430-434.


         March 12, 1990                              7-46     DRAFT - DO NOT QUOTE OR CITE

-------
        Koontz, M. D.; Nagda, N. L. (1988) A topical report on a field monitoring study of homes with unvented gas
               space heaters: v. III. methodology and results. Gas Research Institute final report; contract
               no. 5083-251-0941.

 5      Leaderer, B. P. (1982) Air pollutant emissions from kerosene space heaters. Science (Washington, DC)
               218:  1113-1115.

        Leaderer, B. P.; Cain, W. S.; Isseroff, R.; Berglund, L. G. (1984) Ventilation requirements in buildings-n.
               particulate matter and carbon monoxide from cigarette smoking. Atmos. Environ. 18: 99-106.
10
        Leaderer, B. P.; Stolwijk, J. A. J.; Zagraniski, R. T.; Qing-Shan, M. (1984) A field study of indoor air
               contaminant levels associated with unvented combustion sources. Presented at: 77th annual meeting of the
               Air Pollution Control Association; June; San Francisco, CA. Pittsburgh, PA: Air Pollution Control
               Association; paper no. 84-33.3.
15
        Leaderer, B. P.; Zagraniski, R. T.; Berwick, M.; Stolwijk, J. A. J.; Qing-Shan, M. (1984) Residential exposures
               to NO2, SO2 and HCHO associated with unvented kerosene space heaters, gas appliances and sidestream
               tobacco smoke. In: Berglund, B.; Lindvall,  T.; Sundell, J., eds. Indoor air: proceedings of the 3rd
               international conference on indoor air quality and climate, v. 4, chemical characterization and personal
20             exposure; August; Stockholm,  Sweden. Stockholm, Sweden: Swedish Council for Building Research;
               pp. 151-156.  Available from: NTIS, Springfield, VA; PB85-104214.

        Leaderer, B. P.; Zagraniski, R. T.; Holford, T. R.; Berwick, M.; Stolwijk, J.  A. J. (1984) Multivariate model
               for predicting NO2 levels in residences based upon sources and source use. In: Berglund, B.; Lindvall,
25             T.; Sundell, J., eds. Indoor air: proceedings of the 3rd international conference on indoor air quality and
               climate, v. 4, chemical characterization and personal exposure; August; Stockholm, Sweden. Stockholm,
               Sweden: Swedish Council for Building Research; pp. 43-48. Available from: NTIS, Springfield, VA;
               PB85-104214.

30      Leaderer, B. P.; Zagraniski, R. T.; Berwick, M.; Stolwijk, J. A. J.  (1986) Assessment of exposure to indoor air
               contaminants  from combustion sources: methodology and application. Am. J. Epidemiol.  124:  275-289.

        Lebret, E.; Noy, D.;  Boley, J.; Brunekreef, B. (1987) Real-time concentration  measurements of CO and NO2 in
               twelve homes. In:  Seifert, B.; Esdorn, H.; Fischer, M.; Rueden, H.; Wegner, J., eds. Indoor air '87:
35             proceedings of the 4th international conference on  indoor air quality and climate, v. 1, volatile organic
               compounds, combustion gases, particles and fibres, microbiological agents; August; Berlin, Federal
               Republic of Germany. Berlin, Federal Republic of Germany: Institute for Water, Soil, and Air Hygiene;
               pp. 435-439.

40      Massachusetts Institute of Technology. (1976) Experimental evaluation of range-top burner modification to  reduce
               NOX  formation. American Gas Association;  no. M40677.

        McCarthy, J.; Spengler, J.; Chang, B.-H.; Coultas, D.; Samet, J. (1987) A personal monitoring study to assess
               exposure to environmental tobacco smoke. In: Seifert, B.; Esdorn, H.; Fischer, M.; Rueden, H.;
45             Wegner, J., eds. Indoor air '87: proceedings of the 4th international conference on indoor air quality and
               climate, v. 2, environmental tobacco smoke, multicomponent studies, radon, sick buildings, odours and
               irritants, hyperreactivities and allergies; August; Berlin, Federal Republic of Germany. Berlin, Federal
               Republic of Germany: Institute for Water, Soil and Air Hygiene; pp. 142-146.

50
         March 12, 1990                               7-47      DRAFT - DO NOT QUOTE OR CITE

-------
        McCarthy, S. M.; Yarmac, R. F.; Yocom, J. E. (1987) Indoor nitrogen dioxide exposure: the contribution from
               unvented gas space heaters. In: Seifert, B.; Esdorn, H.; Fischer, M.; Rueden, H.; Wegner, J., eds.
               Indoor air '87: proceedings of the 4th international conference on indoor air quality and climate, v. 1,
               volatile organic compounds, combustion gases, particles and fibres, microbiological agents; August;
 5             Berlin, Federal Republic of Germany. Berlin, Federal Republic of Germany: Institute for Water, Soil,
               and Air Hygiene; pp. 478-482.

        Moffatt, S.  (1986) Backdrafting woes. Prog.  Builder (Dec.): 25-36.

10      Moschandreas, D. J.; Zabransky, J., Jr. (1982) Spatial variation of carbon monoxide and oxides of nitrogen
               concentrations inside residences. Environ. Int. 8: 177-183.

        Moschandreas, D. J.; Relwani, S. M.; O'Neill, H. J.; Cole, J. T.; Elkins, R. H.; Macriss, R. A. (1985)
               Characterization of emission rates from indoor combustion  sources. Chicago, IL: Gas Research Institute;
15             report no. GRI 85/0075. Available from: NTIS, Springfield, VA; PB86-103900.

        Nagda, N. L.; Koontz, M. D. (1985) Microenvironmental and total exposures to carbon monoxide for three
               population subgroups. J. Air Pollut. Control Assoc. 35: 134-137.

20      National Center for Health Statistics. (1986) Vital statistics of the United States 1982: v. n - mortality, part A.
               Hyattsville, MD: U. S.  Department of Health and Human Services, Public Health Service.

        National Institute for Occupational Safety and Health/National Institute of Environmental Health Sciences. (1971)
               Radon daughter exposure and respiratory cancer - quantitative and temporal aspects. Washington, DC:
25             National Institute for Occupational Safety and Health/National Institute of Environmental Health Sciences;
               joint monograph no. 1. Available from: NTIS, Springfield, VA.

        National Research Council. (1981) Indoor pollutants. Washington,  DC: National Academy Press.

30      National Research Council. (1986) Environmental tobacco smoke: measuring exposures and assessing health
               effects. Washington, DC: National Academy Press.

        Perry, J. (1973) Fasten your seatbelts: non-smoking. B. C. Med. J. 15: 304-305.

35      Petersen, G. A.;  Sabersky, R. H. (1975) Measurements of pollutants inside an automobile. J. Air. Pollut.
               Control Assoc. 25: 1028-1032.

        Porthein, F. (1971) Zum Problem des "Passiv-rauchens." Munch. Med. Wochenschr.  118: 707-709.

40      Rickert, W. S.; Robinson, J.  C.; Collishaw,  N. (1984) Yields of tar, nicotine, and carbon monoxide in the
               sidestream smoke for 15 brands of Canadian cigarettes. Am. J.  Public Health 74: 228-231.

        Sebben, J.;  Pimm, P.; Shephard, R. J. (1977) Cigarette smoke in enclosed public facilities. Arch. Environ.
               Health 32: 53-58.
45
        Seiff, H. E. (1973) Carbon monoxide  as an indicator of cigarette-caused pollution levels in intercity buses.
               Washington, DC: U. S. Department of Transportation, Bureau of Motor Carrier Safety; publication no.
               BMCS-IHS-73-1.

50      Sisovic, A.; Fugas, M. (1985) Indoor concentrations of carbon monoxide in selected urban microenvironments.
               Environ.  Monit. Assess. 5: 199-204.
         March 12, 1990                               7-48      DRAFT - DO NOT QUOTE OR CITE

-------
       Slavin, R. G.; Hertz, M. (1975) Indoor air pollution: a study of the thirtieth annual meeting of the American
              Academy of Allergy. 30th annual meeting of the American Academy of Allergy; February; San Diego,
              CA.

 5     Spengler, J. D.; Stone, K.  R.; Lilley, F.  W. (1978) High carbon monoxide levels measured in enclosed skating
              rinks. J. Air Pollut. Control Assoc. 28: 776-779.

       Sterling, T. D.; Sterling, E. (1979) Carbon monoxide levels in kitchens and homes with gas cookers. J. Air
              Pollut. Control Assoc. 29: 238-241.
10
       Surgeon General of the United States. (1986) The health consequences of involuntary smoking: a report of the
              Surgeon General. Rockville, MD: U. S. Department of Health and Human Services, Office on Smoking
              and Health.

15     Szadkowski, D.; Harke, J. (1976) Body burden of carbon monoxide from passive smoking in offices. Inn. Med.
              3: 310-313.

       Thrasher, W. H.; Dewerth, D.  W. (1977) Evaluation of the pollutant emissions from gas-fired water heaters.
              Cleveland, OH: American Gas Association Laboratories; research  report no. 1507.
20
       Tikalsky, S.; Reisdorf, K.; Flickinger, J.; Totzke, D.; Haywood,  J. (1987) Gas range/oven emisssions impact
              analysis. Final report July 1985-December 1987.  Chicago, IL: Gas Research Institute; report no.
              D/M-7683-062. Available from: NTIS, Springfield, VA; PB88-232756/XAB.

25     Traynor, G. W.; Anthon, D. W.; Hollowell, C. D.  (1982) Technique for determining pollutant emissions from a
              gas-fired range. Atmos. Environ. 16: 2979-2987.

       Traynor, G. W.; Allen, J.  R.; Apte, M. G.; Girman, J. R.; Hollowell, C. D. (1983) Pollutant emissions from
              portable kerosene-fired  space heaters. Environ. Sci. Technol.  17: 369-371.
30
       Traynor, G. W.; Apte, M. G.; Carruthers, A. R.; Dillworth, J. F.; Grimsrud, D. T. (1984) Pollutant emission
              rates from unvented infrared and  convective gas-fired space heaters. Berkeley, CA: U. S. Department of
              Energy, Lawrence Berkeley Laboratory; report no.  LBL-18258; Available from: NTIS, Springfield, VA;
              DE85010647/XAB.
35
       Traynor, G. W.; Girman,  J. R.; Apte, M. G.; Dillworth, J. F.; White, P. D. (1985) Indoor air pollution due to
              emissions from unvented gas-fired space heaters.  J. Air Pollut. Control Assoc. 35: 231-237.

       Traynor, G. W.; Apte, M. G.; Sokol, H. A.; Chuang, J. C.; Mumfbrd, J. L. (1986) Selected organic pollutant
40           emissions from unvented kerosene heaters. Presented at: 79th annual meeting of the Air Pollution Control
              Association; June; Minneapolis, MN. Pittsburgh, PA: Air Pollution Control Association; paper
              no. 86-52.5.

       Traynor, G. W.; Apte, M. G.; Carruthers, A. R.; Dillworth, J. F.; Grimsrud, D. T.; Gundel, L. A. (1986)
45            Indoor air pollution due to emission from woodburning stoves. Berkeley,  CA: Lawrence Berkeley
              Laboratories; report no. LBL-17854.

       Traynor, G. W.; Apte, M. G.; Carruthers, A. R.; Dillworth, J. F.; Grimsrud, D. T.; Gundel, L. A. (1987)
              Indoor air pollution due to emissions from wood-burning  stoves. Environ. Sci. Technol.  21:  691-697.
50
       U. S. Bureau of the Census. (1982) Census of population and housing. Supplementary report: provisional
              estimates of social, economic and housing characteristics - states and selected standard metropolitan
              statistical areas. Washington,  DC: U. S. Government Printing Office.
         March 12, 1990                               7-49      DRAFT - DO NOT QUOTE OR CITE

-------
       U. S. Department of Transportation; U. S. Department of Health, Education, and Welfare. (1971) Health aspects
              of smoking in transport aircraft. Washington, DC: U. S. Department of Transportation, Federal Aviation
              Administration, U. S. Department of Health, Education, and Welfare, NIOSH.

 5     Wade, W. A., Ill; Cote, W. A.; Yocom,  J. E. (1975) A study of indoor air quality. J. Air Pollut. Control
              Assoc. 25: 933-939.

       Wallace, L.  A. (1983) Carbon monoxide in air and breath of employees in an underground office. J. Air Pollut.
              Control Assoc.  33: 678-682.
10
       Weber, A. (1984) Annoyance and irritation by passive smoking. Prev. Med. 13: 618-625.

       Weber, A.; Jermini, C.; Grandjean, E. (1976) Irritating effects on man of air pollution due to cigarette smoke.
              Am. J. Public Health 66: 672-676.
15
       Weber, A.; Fischer, T.; Grandjean, E. (1979a) Passive smoking in experimental and field conditions. Environ.
              Res. 20: 205-216.

       Weber, A.; Fischer, T.; Grandjean, E. (1979b) Passive smoking: irritating effects of the total smoke and the gas
20            phase. Int. Arch. Occup. Environ. Health 43: 183-193.

       Whitmore, R. W.; Jones, S. M.; Rosenzweig, M.  S. (1984) Final sampling  report for the study of personal CO
              (carbon monoxide) exposure. Research Triangle park, NC: U. S. Environmnetal Protection Agency,
              Environmental Monitoring Systems Laboratory; EPA report no. EPA-600/4-84-034. Available from:
25            NTIS, Springfield, VA; PB84-181957.

       Womble, S. E. (1988) Personal communication. U. S. Consumer Product Safety Commission.

       World Health Organization. (1985) Air quality guidelines: indoor air pollutants.  Geneva, Switzerland: World
30           Health Organization, Regional Office for Europe.

       Yamanaka, S. (1984) Decay rates of nitrogen oxides in a typical Japanese living room. Environ. Sci. Technol.
               18: 566-570.

35    Yocom,  J. E.; Clink, W. L.; Cote, W. A. (1971)  Indoor/outdoor air quality relationships. J. Air Pollut.  Control
               Assoc. 21: 251-259.

       Zawacki, T. S.; Cole, J. T.; Huang, V.;  Banasiuk, H.; Macriss, R. A. (1984) Efficiency and emissions
               improvement of gas-fired space heaters. Task 2. Unvented space heater emission reduction.  Chicago, IL:
40            Gas Research Institute; report no. GRI-84/0021. Available from: NTIS,  Springfield, VA; PB84-237734.

       Ziskind, R.  A.;  Rogozen, M.  B.; Carlin, T.; Drago, R. (1981) Carbon monoxide intrusion into sustained-use
               vehicles. Environ. Int.  5: 109-123.

45    Ziskind, R.  A.;  Fite, K.; Mage, D. T. (1982) Pilot field study: carbon monoxide exposure monitoring in the
               general population. Environ. Int.  8: 283-293.
         March 12, 1990                               7-50      DRAFT - DO NOT QUOTE OR CITE

-------
                8.  POPULATION EXPOSURE TO CARBON
                                        MONOXIDE
 5     8.1  INTRODUCTION
            A fundamental purpose of the Clean Air Act is to protect public health.  The NAAQS
       are set at levels which attempt to provide a margin of safety to protect the health of the
       populace from adverse effects of air pollutants.  In setting a pollutant standard, population
       exposure is an important consideration, because public health can be affected adversely by an
10     air pollutant only if the following two conditions coincide.
            (1)   Persons actually are or potentially would be exposed in daily living to levels of
                  the pollutant occurring in ambient air locations at or above undesirable
                  concentrations.
15
            (2)   The air pollutant causes adverse effects on human health in sensitive population
                  groups at these concentrations.
20
            This chapter focuses on the degree to which the population actually is exposed to
       outdoor, in-transit, or indoor concentrations of CO that might produce adverse health effects.
       Chapter 10 deals with the second topic, the physiological and other health effects associated
       with a person's exposure to various concentrations of CO.
25          Carbon monoxide is emitted from sources of incomplete combustion of HC fuels.1
       Consequently,  ambient concentrations of CO can reach high levels close to emission  sources.
       The strong source-dependence of CO leads to highly variable spatial and temporal
       concentrations  in urban environments (see Chapter 6).  Because of the variable concentration
       patterns exhibited by CO, it is necessary to address the relationships between ambient
30     concentrations  and human exposures to evaluate the potential health risk associated with actual
       population exposures.
          'Hydrocarbon fuels are burned incompletely by internal combustion engines (e.g., automobiles, trucks, and
       small utility engines), and by sources such as cigarettes, forest fires, and poorly adjusted gas burners.
       March 12, 1990                           8-1       DRAFT-DO NOT QUOTE OR CITE

-------
            Compliance with the NAAQS currently is judged by comparing the standards to data
        from fixed, ambient-air monitoring stations. The sites for such stations are chosen, however,
        to monitor ambient conditions in a number of neighborhood types (Ott, 1977) rather than
        actual human exposure.  In fact, a number of studies have shown that ambient-air monitoring
 5      stations do not necessarily reflect the concentrations to which people are actually exposed.
        This difference between fixed stations and exposures occurs because of the spatial and      -
        temporal variability of carbon monoxide. In contrast to the stationary location of an ambient
        monitor, people are usually moving through a succession of microenvironments (e.g., homes,
        sidewalks, buses, automobiles, shopping malls, downtown street canyons, restaurants, offices,
10      factories, and garages) where they may spend time in closer proximity to CO sources and in
        more enclosed spaces than the outdoors.  The result is that existing ambient monitoring
        stations often do not reflect individual exposure patterns, nor  do they necessarily reflect the
        highest concentrations to which those people are exposed. However, fixed monitors do give
        some general information on  the overall level  of exposure of a population to  CO and are
15      useful for a variety of other purposes (see Chapter 6).
            Among all major air pollutants, CO has  one of the clearest measures available of
        biological dose (see Section 8.5). The concentration of CO circulating in the blood,
        expressed as the percentage of Hb bound with CO,  or COHb, is a useful measure of dose for
        relating this pollutant to deleterious health effects (see Chapter 10).  Blood COHb is in turn a
20      function of inhaled CO, breathing rate, blood volume, and other physiological factors (see
        Chapter 9: Pharmacokinetic modeling).  Because the relationship between ambient CO and
        blood COHb is dynamic,  it is necessary to know  the timing and duration of an exposure series
        in order to predict resulting COHb.
            Because of the spatial and temporal variability of CO as well as the known functional
25      relationships between concentration and a measure of dose, CO is a model pollutant for
        development and evaluation of improved approaches for assessing human exposure.  A
        number of field studies now have been undertaken that provide a quantitative assessment of
        the disparity between fixed monitoring stations and actual exposures.  In addition, field
        measurements of body burden (for example, blood COHb, and breath CO) are available for
30      comparison with fixed station monitoring data.  Several personal-monitoring CO field studies
        have employed representative statistical sampling procedures,  allowing inferences to be made

        March 12, 1990                           8-2      DRAFT-DO NOT QUOTE OR CITE

-------
       about the CO exposures (or COHb levels) of an entire population of a city or a region.
       Finally, models of population exposure and activity patterns have been developed to bridge
       the gap between ambient, fixed-station measurements, and actual personal exposures.
       Additional data need to be developed from personal monitoring field studies for use in
 5     validating these models. The models are important for improving our understanding of
       human exposure to CO and for evaluating different control strategies. The models can
       indicate locations and sources of most significant exposure and therefore may suggest control
       strategies to reduce human exposure and the resulting deleterious health effects.

10
       8.2  EXPOSURE MONITORING IN THE POPULATION
             In recent years, researchers have focused on the problem of determining actual
       population exposures to CO.  There are two alternative approaches for estimating the
       exposures of a population to air pollution:  the "direct approach" using field measurement of a
15     representative population carrying PEMs; and the "indirect approach" involving computation
       from field data of activity patterns and measured concentration levels within
       microenvironments (Ott, 1982).
             In the direct approach, as study participants engage in regular daily activities, they are
       responsible for recording their exposures to the pollutant of interest. Subjects can record their
20     exposures in a diary, the method used in  a study in Los Angeles (Ziskind et al., 1982), or
       they can automatically store exposure data in a data logger,  the method used in studies in
       Denver (Johnson, 1984) and Washington, DC (Hartwell et al., 1984), which are summarized
       by Akland et al. (1985).  In all of these studies, subjects recorded the time and nature of their
       activities while they monitored personal exposures to CO.  The direct approach - sometimes
25     called the Total Exposure Assessment Methodology (TEAM) - is useful to obtain an exposure
       inventory of a representative sample from either the general population, or from a specific
       subpopulation,  which can be defined by many demographic, occupational, and health factors.
       The inventory can cover a range of microenvironments encountered over a period of interest
       (e.g., a day), or it can focus on one particular microenvironment. With this flexibility,
30     policy analysts can assess the problem that emission sources pose to a particular subgroup
       (e.g., commuters) active in a specific microenvironment (e.g., automobiles).

       March 12,  1990                          8-3      DRAFT-DO NOT QUOTE OR CITE

-------
            The indirect approach to estimating personal exposure is to use PEMs or
       microenvironmental monitors (MEMs) to monitor microenvironments rather than individuals.
       Combined with additional data on human activities that occur in these microenvironments,
       data from the indirect approach can be used to estimate the percentage of a subpopulation that
 5     is at risk to pollutant concentrations that exceed national or state air quality standards.
       Flachsbart and Brown (1989) conducted this type of study to estimate merchant exposure to
       CO from motor vehicle exhaust at Honolulu's Ala Moana Shopping Center.

       8.2.1  Personal  Monitoring
10          The development of small PEM, as discussed in Chapter 5, made possible the large-
       scale CO human exposure field studies in Denver, CO, and Washington, DC, in the winter of
       1982-1983 (Akland et al., 1985). These monitors proved effective in generating 24-h CO
       exposure profiles on 450 persons in Denver and 800 in Washington.  Because personal
       monitoring techniques are new, and few field studies have been done, the science of
15     measuring the exposures to chemicals in human populations is in an early stage of
       development.  The use of PEMs and  concurrent diaries in large-scale population studies
       requires rigorous quality control and  introduces many new problems not present in ambient
       monitoring studies. The PEMs must be rugged, self-powered, lightweight, and free of drift
       while being carried and exposed to temperature variations; associated data loggers are
20     required to store the continuous PEM readings.  The diary format must be clear, easy for the
       subject to complete, and easy for the researcher to interpret.  With good calibration practices,
       the CO PEMs can  provide a precision of less than +1 ppm.  The Denver-Washington, DC
       study is the only large-scale population exposure field study  that  yet has been undertaken.
       Despite the complexity of such a study, the large probability sample and high time resolution
25     of the PEMs yielded a rich data base for characterizing the exposures of the population to CO
       in two major U.S.  cities. The findings have greatly increased the understanding of the
       causes,  severity, and variability of the exposures of human populations to CO.
            Results from  the Denver-Washington DC study  (Akland et  al., 1985) show that over
        10% of the residents in Denver and 4% of the Washington DC residents were exposed to
30     eight-hour average CO levels above 9 ppm during the winter study period. This degree of
       population exposure could not accurately be deduced  from simultaneous data collected by the

       March 12, 1990                          8-4       DRAFT-DO NOT QUOTE OR CITE

-------
        fixed-site monitors without taking into account other factors such as contributions from indoor
        sources, elevated levels within vehicles, and individuals' activity patterns.  In Denver, for
        example, the fixed-site monitors exceeded the 9 ppm level only 3.1 % of the time.  These
        results indicate that the effects of personal activity, indoor sources, and especially time spent
 5      commuting, greatly contribute to a person's CO exposure.
             This study emphasizes that additional strategies are required to augment data from fixed
        site monitoring networks in order to evaluate actual human CO exposures and health risks
        within a community.  The cumulative frequency distributions of CO data for both Denver and
        Washington DC in Figure 8-1  show that personal monitors often measure higher concentration
10      than do fixed stations.  As part of this study, comparisons were made of exposure to one-
        hour  CO concentrations as determined by personal monitors and of measured ambient
        concentrations at fixed monitor sites. Correlations between personal monitor data and fixed-
        site data were consistently poor;  the fixed-site data usually explained less than 10% of the
        observed variation in personal exposure.  For example, one-hour CO measurements taken at
15      the nearest fixed stations only were weakly correlated (0.14 < r < 0.27) with office or
        residential measurements taken with personal monitors (Akland et al., 1985).
             The conclusion that exposure of persons to ambient CO and other pollutants does not
        directly correlate with concentrations determined at fixed-site monitors is supported by the
        work of others (Ott and  Eliassen, 1973; Cortese and Spengler, 1976; Dockery and Spengler,
20      1981; Wallace and Ott,  1982).
             In view of the high degree of variability of ambient CO concentrations over both space
        and time, (see Chapter 6) the reported results are not surprising.  A given fixed monitor is
        unable to track the exposure of individuals to ambient CO as they go about their daily
        activities, moving from one location to another,  and seldom in the immediate vicinity of the
25      monitor.  This does not necessarily mean, however, that fixed monitors do not give some
        general information on the overall level of exposure of a population to CO.  The Akland data,
        although failing to show a correlation between individual personal monitor exposures and
        simultaneous nearest fixed-site monitor concentrations did suggest that, in Denver, aggregate
        personal exposures were lower on days of lower ambient CO levels as determined by fixed
30      site monitors and higher on days  of higher ambient levels. Also, both fixed-site and personal
        exposures were higher in Denver than in Washington.

        March 12, 1990                           8-5     DRAFT-DO NOT QUOTE OR CITE

-------
           99.99
         100	
         50-
E 20
0.
a.
c
.2 10.
c
o
g  5
O
O
o
•O
'x
i  2-
O
      c
      o
      n
      CO
      O
    1 -
         0.5-
        0.2-
        0.1
             Population Above Concentration Shown (%)
                99     90       50       10      1
                I   I   !  I	I    I   I	I  I  I   I
                    9 ppm NAAQS
          Denver:
           	   Personal Exposure
           	Fixed Stations
          Washington. DC:
           _	   Personal Exposure
           	   Fixed Stations
0.01
                             10
I
50
                                         M   I  i
                                         90     99
                                                         0.01
                                                           -100
                                                           -50
                                                                -20  E
                                                                      Q.
                                                                 c
                                                          Mo   .2
                                                          -5
                                                          -2
                                                          -1
                                                                      0)
                                                                      o
                                                                      o
                                                                      O
                                                                      o
                                                                      p
                                                                      I
                                                                      o
                                 O
                                n
                                                          -    o
                                                          -0.5
                                                          -0.2
                                                        99
 0.1
99
                  Population Below Concentration Shown (%)
Figure 8-1. Frequency distributions of maximum eight-hour carbon monoxide population
          exposures and fixed-ale monitor values in Denver, CO and Washington, DC;
          November 1982 - February 1983.
Source: Based on Akland, et al. (1984).
March 12, 1990
                             8-6     DRAFT-DO NOT QUOTE OR CITE

-------
            The PEMs have shown themselves to be powerful tools for quantifying air quality levels
       in in-transit, outdoor, and indoor microenvironments.  A great number of microenvironments
       can be compared in one study.  For example, Table 8-1 shows in-transit microenvironments
       in Denver, ranked from highest to lowest concentration by arithmetic mean.  The in-transit
 5     microenvironment with the highest estimated CO concentration is the motor vehicle, whereas
       walking and bicycling have the lowest CO concentrations. Outdoor microenvironments also
       can be ranked (Table 8-2) for these data. Outdoor public garages and outdoor residential
       garages and carports had the highest CO concentrations; outdoor service stations, vehicle
       repair facilities, and parking lots had intermediate concentrations. In contrast, school grounds
10     and residential grounds had relatively low concentrations, whereas extremely low CO
       concentrations were found in outdoor sports arenas, amphitheaters, parks, and golf courses.
       Finally, a wide range of concentrations was found in Denver within indoor micro-
       environments  (Table 8-3).  The highest indoor CO concentrations occurred in service stations,
       vehicle repair facilities and public parking garages;  intermediate concentrations were found in
15     shopping malls, residential garages, restaurants, offices, auditoriums, sports arenas, concert
       halls, and stores; the lowest concentrations were found in health care facilities, public
       buildings, manufacturing facilities, homes, schools, and churches.
            One activity that influences personal exposure is commuting. An estimated 1 % of the
       noncommuters in Washington were exposed to concentrations above 9 ppm for eight hours.
20     By comparison, an estimated 8% of persons reporting that they commuted more than 16 h per
       week had CO exposures above the  9-ppm, eight-hour level.   Finally, certain occupational
       groups whose work brings them in close proximity to the internal combustion engine had a
       potential for elevated CO exposures.  These include automobile mechanics, parking garage or
       gas station attendants, crane deck operators, cooks,  taxi,  bus, and truck drivers, firemen,
25     policemen, and warehouse and construction workers.  Of the 712 CO exposure profiles
       obtained in Washington 29 persons fell into this "high- exposure" category.  Of these, 25%
       had eight-hour CO exposures above the 9-ppm level.
            Several field  studies also have been conducted by the U.S. Environmental Protection
       Agency to determine the feasibility and effectiveness of monitoring selected micro-
30     environments  for use in estimating  exposure profiles indirectly.  One study (Flachsbart et al.,
       March 12, 1990                           8-7      DRAFT-DO NOT QUOTE OR CITE

-------
              TABLE 8-1.  CARBON MONOXIDE CONCENTRATIONS IN IN-TRANSIT
                        MICROENVIRONMENTS - DENVER, COLORADO
                        (Listed in descending order of mean CO concentration)
10

15

20
Microenvironment
Motorcycle
Bus
Car
Truck
Walking
Bicycling
n
22
76
3,632
405
619
9
Mean'
(ppm)
9.79
8.52
8.10
7.03
3.88
1.34
Standard
Error
(ppm)
1.74
0.81
0.16
0.49
0.27
1.20
       "An observation was recorded whenever a person changed a microenvironment, and on every clock hour; thus,
        each observation had an averaging time of 60 min or less.
25
       Source: Johnson (1984).
30     1987) conducted in Washington in 1982-1983 concentrated on the commuting micro-
       environment, because earlier studies identified this microenvironment type as the single most
       important nonoccupational microenvironment relative to total CO population exposure. It was
       observed that for the typical automobile commuter the time-weighted average CO exposure
       while commuting ranged from 9 to 14 ppm.  The corresponding rush-hour (7 to 9 am, 4  to
35     6 pm) averages at fixed-site monitors were 2.7 to 3.1 ppm.

       8.2.2  Carbon Monoxide Exposures Indoors
            The majority of people in the United States spend a majority of their time indoors;
       therefore, a comprehensive depiction of exposure to CO must include this setting. The indoor
40     sources, emissions, and concentrations are sufficiently diverse, however, that only a few
       example studies can be cited here; a thorough discussion of CO in homes, offices, and similar
       environments, is presented  in Chapter 7.  Although a number of these studies report on

       March  12, 1990                         8-8      DRAFT-DO NOT QUOTE OR  CITE

-------
10
15
20
25
30
               TABLE 8-2.  CARBON MONOXIDE CONCENTRATIONS IN OUTDOOR
                        MICROENVIRONMENrS - DENVER, COLORADO
                        (Listed in descending order of mean CO concentration)
Microenvironment
Public garages
Residential garages
or carports
Service stations or
vehicle repair facilities
Parking lots
Other locations
School grounds
Residential grounds
Sports arenas,
amphitheaters
Parks, golf courses
n
29
22
12
61
126
16
74
29
21
Mean*
(ppm)
8.20
7.53
3.68
3.45
3.17
1.99
1.36
0.97
0.69
Standard
Error
(ppm)
0.99
1.90
1.10
0.54
0.49
0.85
0.26
0.52
0.24
       "An observation was recorded whenever a person changed a microenvironment, and on every clock hour; thus,
        each observation had an averaging time of 60 min or less.
35     Source:  Johnson (1984).

       microenvironmental concentrations, they do not specifically address human exposure while
40     indoors.
            Early studies date back to before 1970 when it was found that indoor and outdoor levels
       do not necessarily agree.  For example, one study determined indoor-outdoor relationships for
       CO over two-week periods during summer, winter, and fall in 1969-1970 in buildings in
       Hartford, CT (Yocom et al., 1971). With the exceptions of the private homes, which were
45     essentially equal, there was a day to night effect in the fall and winter seasons; days were

       March  12, 1990                          8-9      DRAFT-DO NOT QUOTE OR CITE

-------
              TABLE 8-3. CARBON MONOXIDE CONCENTRATIONS IN INDOOR
                      MICROENVIRONMENTS - DENVER, COLORADO

                      (Listed in descending order of mean CO concentration)
10

15

20

25

30

35

40

Microenvironment
Public garages
Service stations or
vehicle repair facilities
Other locations
Other repair shops
Shopping malls
Residential garages
Restaurants
Offices
Auditoriums, sports arenas,
concert halls
Stores
Health care facilities
Other public buildings
Manufacturing facilities
Homes
Schools
Churches
n
116
125
427
55
58
66
524
2,287
100
734
351
115
42
21,543
426
179
Mean"
(ppm)
13.46
9.17
7.40
5.64
4.90
4.35
3.71
3.59
3.37
3.23
2.22
2.15
2.04
2.04
1.64
1.56
Standard
Error
(ppm)
1.68
0.83
0.87
1.03
0.85
0.87
0.19
0.002
0.48
0.21
0.23
0.30
0.39
0.02
0.13
0.25
       "An observation was recorded whenever a person changed a microenvironment, and on every clock hour; thus,
45     each observation had an averaging time of 60 min or less.

       Source:  Johnson (1984).
       March 12, 1990                        8-10     DRAFT-DO NOT QUOTE OR CITE

-------
       higher by about a factor of 2. These differences are consistent with higher traffic-related CO
       levels outdoors in the daytime.
            Indoor and outdoor CO concentrations were measured in four homes also in the
       Hartford, CT area in 1973-1974 (Wade et al., 1975).  All used gas-fired cooking stoves.
 5     Concentrations were measured in the kitchen, living room, and bedroom.  Stove use, as
       determined by activity diaries, correlated directly with CO concentrations.  Peak CO
       concentrations in several of the kitchens exceeded 9 ppm, but average concentrations ranged
       from 2 to 3 ppm to  about 8 ppm.  These results are in general agreement with results
       obtained in Boston,  MA (Moschandreas and Zabransky, 1982).  In this study, they found
10     significant differences between rooms in homes where there were gas appliances.
            Effects of portable kerosene-fired space heaters on indoor air quality were measured in
       an environmental chamber and a house (Traynor et al., 1982).  CO emissions from white
       flame (WF) and blue flame (BF) heaters were compared. The WF convective heater emitted
       less CO than the BF radiant heater.  Concentrations in the residence were  < 2 ppm and 2 to
15     7 ppm, respectively. The authors conclude that high levels might  occur when kerosene
       heaters are used in small spaces and/or when air exchange rates are low.
            A rapid method using an electrochemical PEM to survey CO was applied in nine high-
       rise buildings in the San Francisco and Los Angeles areas during 1980 and 1984 (Flachsbart
       and Ott,  1986).  One building had exceptionally  high CO levels compared to the other
20     buildings; average concentrations on various floors ranged from 5  to 36 ppm. The highest
       levels were in  the underground parking garage, which was found to be the source of elevated
       CO within the building.
            The effect of residential wood combustion and specific heater type on indoor CO has
       been investigated (Humphreys et al., 1986). Airtight and non-airtight heaters were compared
25     in a research home in Tennessee.  CO emissions from the non-airtight heaters was generally
       higher than from airtight heaters.  Peak indoor CO concentration (ranging  from  1.3 to
       29.6 ppm, depending on heater type) was related to fuel reloadings.
            CO levels in 254 Netherland homes with unvented gas-fired water heaters  were
       investigated during the winter of 1980 (Brunekreef et al., 1982).  Concentrations (ppm) at
30     breathing  height fell into the following categories:  < 10 (n = 154), 11-50 (n=50), 51-100
       March 12, 1990                          8-11      DRAFT-DO NOT QUOTE OR CITE

-------
       (n=25),  > 100 (n= 17).  They found that a heater vent reduced indoor CO concentrations,
       and the type of burner affected CO levels.
            Air pollution in Dutch homes was investigated by Lebret (1985). CO concentrations
       (ppm) in various locations were kitchen, 0 to 17.5; living room, 0 to 8.7; and bedroom, 0 to
 5     3.5.  CO levels were elevated in homes with gas cookers and unvented geysers (water
       heaters).  Kitchen CO levels were higher than those in other locations due to peaks from the
       use of gas appliances. Living room CO values were slightly higher in houses with smokers.
       The overall mean CO level indoors was 0 to 2.7 ppm above outdoor levels.
            In Zagreb,  Yugoslavia, CO was measured in eight urban institutions housing sensitive
10     populations, including kindergartens, a children's hospital, and homes for the elderly (Sisovic
       and Fugas, 1985). Winter CO concentrations ranged from  1.1 to  13.7 ppm, and summer
       concentrations ranged from 0.6 to 6.9 ppm.  The authors attributed indoor CO concentrations
       to nearby traffic  density, general urban pollution, seasonal differences, and day-to-day
       weather conditions.  Indoor sources were not reported.
15          Toxic levels of  CO also were found in measurements at six ice skating rinks (Johnson
       et al.,  1975b). This  study was prompted by the reporting of symptoms of headache and
       nausea among 15 children who patronized one of the rinks.  CO concentrations were found to
       be as high as 304 ppm during operation of a propane-powered, ice-resurfacing machine.
       Depending on skating activity levels, the ice-resurfacing operation was performed for 10 min
20     every one to two hours.  As this machine was found to be the main source of CO, use of
       catalytic converters and properly tuning the engine greatly reduced emissions of CO. Similar
       findings have been reported by Spengler et al. (1978).

       8.2.3  Carbon Monoxide Exposures Inside Vehicles
25          Studies of CO concentrations inside automobiles also have been reported over the past
       decade.  Petersen and Sabersky (1975) measured pollutants inside an automobile under typical
       driving conditions. CO concentrations were generally less than 25 ppm, with one three-
       minute peak of 45 ppm.  Average concentrations inside the vehicle were similar to those
       outside.  No in-vehicle CO  sources were noted.
30          Drowsiness, headache, and nausea were reported by eight children  who had ridden in
       school buses for about two hours while traveling on a ski trip (Johnson et al.,  1975a).  The

       March 12, 1990                          8-12     DRAFT-DO NOT  QUOTE OR CITE

-------
       students reporting symptoms were seated in the rear of the bus, which had a rear-mounted
       engine and a leaky exhaust. The exhaust system subsequently was repaired. During a later
       ski outing for students, CO concentrations  also were monitored for a group of 66 school buses
       in the parking lot.  The investigators found five buses with CO concentrations of 5 to 25 ppm
 5     (mean  15 ppm), 24 buses showing concentrations in excess of 9 ppm for short periods, and
       two buses showing up to 3 times the 9-ppm level for short periods. Drivers were advised to
       park so that exhausts from one bus would not be adjacent to the fresh air intake for another
       bus.
            During a cross-country trip in  the spring of 1977, Chaney (1978) measured in-vehicle
10     CO concentrations.  The CO levels  varied  depending on traffic speed.  On expressways in
       Chicago, San Diego, and Los  Angeles when traffic speed was less than  10 mph, CO exceeded
       15 ppm.  Levels increased to 45 ppm when traffic stopped.  In addition, it was observed that
       heavily loaded vehicles (e.g.,  trucks) produced high CO concentrations inside nearby
       vehicles, especially when the trucks were ascending a grade.
15          Colwill and Hickman (1980) measured CO concentrations in 11  new cars as they were
       driven on a heavily trafficked  route  in and  around London.  The inside mean CO level for the
       11 cars was 25.2 ppm vs. 47.0 ppm for the outside mean.
            In a study mandated by Congress in the 1977 Clean Air Act Amendments, the EPA
       studied CO intrusion into vehicles (Ziskind et al., 1981). The  objective was to determine
20     whether CO was leaking into the passenger compartments of school buses, police cars, and
       taxis, and,  if so, how prevalent the  situation was.  The study involved 1164 vehicles in
       Boston, and Denver. All vehicles were in  use in a working fleet at the time of testing.  The
       results indicated that all three  types  of vehicles often have multiple (an average of 4 to 5)
       points of CO intrusion - worn gaskets, accelerator pedals, rust spots in the  trunk, and such.
25     In 58% of the rides lasting longer than eight hours, CO levels exceeded 9 ppm.  Thus the
       study provided evidence that maintenance and possibly design of vehicles may be an
       important factor in human exposure to CO.
            Flachsbart (1989) investigated  the effectiveness of priority lanes on a Honolulu arterial
       in reducing commuter travel time and exposure to CO. The CO concentrations and exposure
30     of commuters in these lanes was substantially lower than in the nonpriority  lanes. CO
       exposure was reduced approximately 61%  for express buses, 28% for high-occupancy

       March 12,  1990                          8-13      DRAFT-DO  NOT QUOTE OR CITE

-------
       vehicles, and 18% for carpools when compared to that for regular automobiles. The higher
       speed associated with priority lanes helped reduce CO exposure.  These observations
       demonstrate that CO concentrations have a high degree of spatial variability on roadways.
            Additional findings on CO levels inside vehicles are summarized in a literature review
 5     by Flachsbart and Ah Yo (1989).  In general, a wide variation in CO exposures has been
       observed in in-transit microenvironments.

       8.2.4  Carbon Monoxide Exposures Outdoors
            Carbon monoxide concentrations in outdoor settings (besides those measured at fixed
10     monitoring stations) also show considerable variability, as is evident from the eight Denver
       microenvironmental groupings listed in Table 8-2. Ott (1971) made 1128 CO measurements
       at outdoor locations in San Jose at breathing height over a six-month period and compared
       these results with the official fixed monitoring station data. This study included the
       measurements of the outdoor CO  exposures of pedestrians in downtown San Jose by requiring
15     them to carry personal monitoring pumps and bags while walking standardized routes on
       congested sidewalks.  If an outdoor measurement was made more than 100 m away from any
       major street, its CO concentration was similar, suggesting the existence of a generalized urban
       background concentration in San Jose that was spatially uniform over the city (within a 33-
       km2 grid) when one is sufficiently far away from mobile  sources.  Because the San Jose
20     monitoring station then was located near a street with heavy traf^.c, it recorded concentrations
       approximately 100%  higher than this background value.  In contrast, outdoor CO levels from
       personal monitoring studies of downtown pedestrians were 60% above the corresponding
       monitoring station values and the correlation coefficient was low (r = 0.20). By collecting
       the pedestrian personal exposures over eight-hour periods, it was possible to compare the
25     levels with the NAAQS concentration level. On  two of seven days for which data were
       available, the pedestrian concentrations were particularly high (13 and 14.2 ppm)  and were
       two to three times the corresponding levels recorded at the same time (4.4 and 6.2 ppm) at
       the air monitoring station (Ott and Eliassen, 1973; Ott and Mage, 1974).  These results show
       that concentrations to which pedestrians are exposed on downtown streets can exceed a
30     9 ppm, 8-hour average while the official air monitoring station records values significantly
       less than that. It can be argued, however, that not many pedestrians spend eight hours

       March 12, 1990                          8-14     DRAFT-DO NOT QUOTE OR CITE

-------
        outdoors walking along downtown sidewalks, and that is one of the important reasons for
        including realistic human activity patterns in exposure assessments, as indicated in
        Section 8.3.
             Godin et al. (1972) conducted similar studies in downtown Toronto using 100-mL glass
  5     syringes in conjunction with nondispersive spectrometry. They measured CO concentrations
        along streets, inside passenger vehicles, and at a variety of other locations.  Like other
        investigators, they found that CO concentrations were determined by very localized
        phenomena.  In general, CO concentrations in traffic and along streets were much higher than
        those observed at conventional fixed air monitoring stations.  In a subsequent study in
 10     Toronto, Wright et al. (1975) used Ecolyzers to measure four to six minute average CO
        concentrations encountered by pedestrians and street workers and obtained similar results.
        Levels ranged from 10 to 50 ppm, varying with wind speed and direction, atmospheric
        stability, traffic density, and height of buildings. He also measured CO concentrations on the
        sidewalks of a street that subsequently was closed to traffic to become a pedestrian mall.
 15     Before the street was closed, the average concentrations at two  intersections were 9.4  +
        4.0 ppm (st. dev.) and 7.9 ± 1.9 ppm; after the street was closed, the averages dropped to
        3.7 ± 0.5 ppm and 4.0 + 1.0 ppm, respectively, which were equivalent to the background
        level.
             A large-scale field investigation was undertaken of CO concentrations in indoor and
20     outdoor locations in five California cities using personal  monitors (Ott and Flachsbart, 1982).
        For outdoor commercial settings, the average CO concentration was 4 ppm.  This CO level
        was statistically, but not substantially, greater than the average CO concentration of 1.98 ppm
        recorded simultaneously at nearby fixed-monitoring stations.  The final report of this field
        study (Flachsbart and Ott,  1984) contains an extensive literature review of CO exposures
25      found in indoor, outdoor, and in-transit microenvironments.
       8.3 ESTIMATING POPULATION EXPOSURE TO CARBON
            MONOXIDE
30          Accurate estimates of human exposure to CO are a prerequisite for a realistic appraisal
       of both the risks posed by the pollutant and the design and implementation of effective control

       March 12,  1990                         8-15      DRAFT-DO NOT QUOTE OR CITE

-------
       strategies. This section discusses the general concepts on which exposure assessment is
       based, the limitations of using ambient fixed-site monitoring data alone for estimating
       exposure, alternative approaches which have been proposed for estimating population
       exposure to air pollution, and specific applications of these approaches to estimating CO.
 5     Because of problems in estimating population exposure solely from fixed station data, several
       formal human exposure models have been developed. Some of these models include
       information on human activity patterns:  the microenvironments people visit and the times
       they spend there.  These models also contain submodels depicting the sources and
       concentrations likely to be found in each microenvironment, including indoor, outdoor, and
10     in-transit settings.

       8.3.1 Defining Concentration, Exposure, and Dose
            In evaluating models for estimating CO, it is important to understand the basic concepts
       of concentration, exposure, and dose. Sexton and Ryan (1988) provide the following
15     definitions.
            The "concentration" of a specific air pollutant is the amount of that material per unit
       volume of air.  Air pollution monitors measure pollutant concentrations, which may or may
       not provide accurate exposure estimates.
            The term "exposure" is defined as any contact between air contaminant and the outer
20     (e.g., skin) or inner (e.g., respiratory tract epithelium) surface of the human body. Exposure
       implies the simultaneous occurrence of two events (Ott, 1982):

             (1)   A pollutant concentration, C, is present at location x,y,z at time t.
25           (2)   A person, i, is present at location x,y,z at time  t.


             A key distinction is apparent between a concentration and an exposure.  The
30     concentration of an airborne contaminant measured in an empty room is just that, a
       concentration.  A concentration measured in a room with people present is a measurement of
       exposure.  A measured concentration is a surrogate for exposure only to the degree to which
       it represents  concentrations actually experienced by individuals.

       March 12, 1990                          8-16      DRAFT-DO NOT QUOTE OR CITE

-------
             A more important distinction exists between "exposure" and "dose." Whereas exposure
        is the pollutant concentration at the point of contact between the body and the external
        environment, dose is defined as the amount of pollutant that actually crosses one of the body's
        boundaries and reaches the target tissue. Among the factors that affect the magnitude of the
  5     dose received are respiration rate, respiration mode (e.g., mouth breathing versus nose
        breathing), uptake, metabolism, and clearance.

        8.3.2  Components of Exposure
             Two aspects of exposure bear directly on the related health consequences.
 10
             (1)   Magnitude:  What is the pollutant concentration?
             (2)   Duration: How long does the exposure last?
 15
             The magnitude is an important exposure parameter, since concentration typically is
        assumed to be directly proportional to dose, and ultimately, to the health outcome. But
        exposure implies a time component,  and it is essential to specify the duration of an exposure.
20     The health risks of exposure to a specific concentration for five minutes are likely to be
        different, all other factors being equal, than exposure to the same concentration for an hour.
             The magnitude and duration of exposure can be determined by  plotting an individual's
        air pollution exposure over time (Figure 8-2). The function Q(t) describes the air pollutant
        concentration to which an individual is exposed at any point in time t.  The shaded area under
25     the graph represents the accumulation of instantaneous exposures over some period of time (tr
        O.  This area also is equal to the integral of the air pollutant concentration function, C;(t),
        between to and t,.  Ott (1982) defines the quantity represented by this area as the integrated
        exposure.
             By dividing the integrated exposure by the  period of integration (t,-to),  the average
30     exposure represents the average air pollutant concentration that  an individual was exposed to
        over the defined period of exposure. To facilitate comparison with established air quality
        standards, an averaging period is chosen to equal the averaging period of the standard (t9).  In
        this case, the average exposure is referred to as a standardized exposure.

        March 12, 1990                           8-17      DRAFT-DO NOT QUOTE OR CITE

-------
  o

  z
  <
  cr
  UJ
  u


  o
  u
                            TWECO
        Figure 8-2.  Typical individual exposure as a function of time.




        Source: Ott (1982).
March 12, 1990
8-18      DRAFT-DO NOT QUOTE OR CITE

-------
             As previously discussed, exposure represents the joint occurrence of an individual being
        located at point (x,y,z) during time t, with the simultaneous presence of an air pollutant at
        concentration C^t).  Consequently, an individual's exposure to an air pollutant is a function
        of location as well as time.  If a volume at a location can be defined such that air pollutant
 5      concentrations within it are homogeneous yet potentially different from other locations, the
        volume may be considered a "microenvironment" (Duan, 1982). Microenvironments may be
        aggregated by location (i.e., indoor or outdoor) or activity performed at a location (i.e.,
        residential, commercial) to form microenvironment types.
             It is important to distinguish between individual exposures and population exposures.
10      Sexton and Ryan (1988) define the pollutant concentrations experienced by a specific
        individual during normal daily activities as "personal" or "individual" exposures.  A personal
        exposure depends on the air pollutant concentrations that are present in the locations through
        which the person moves, as well as on the time spent at each location.  Because time-
        activity patterns can vary substantially from person to person, individual exposures exhibit
15      wide variability (Dockery and Spengler, 1981; Quackenboss et al., 1982; Sexton et al., 1984;
        Spengler et al.,  1985; Stock et al.,  1985; Wallace et al., 1985). Thus, although it
        is a relatively straightforward procedure to measure any one person's exposure, many such
        measurements may be needed to quantify exposures for a defined group.  The daily activities
        of a person in time and space define his or her activity pattern.  Accurate estimates of air
20      pollution exposure generally require that an exposure model account for the activity patterns
        of the population of interest.  The activity patterns may be determined through  "time budget"
        studies of the population.  Studies of this type have been performed by Szalai (1972), Chapin
        (1974), Robinson (1977),  Michelson and Reed (1975), Johnson (1987) and Schwab et al.
        (1989).  The earlier studies may now be dated and were not designed to investigate human
25      exposure questions. Ongoing exposure studies have adopted the diary methods that were
        developed for sociological investigations and applied them to current exposure and time
        budget investigations.  A few of these studies have been reported (e.g., Schwab et al., 1989;
        Johnson, 1987).
             From a public health perspective, it is important to determine the "population
30      exposure," which is the aggregate exposure for a specified group of people (e.g., a
        community, an identified occupational cohort).  Because exposures are likely to vary

        March 12, 1990                           8-19      DRAFT-DO NOT QUOTE OR CITE

-------
       substantially between individuals, specification of the distribution of personal exposures
       within a population, including the average value and the associated variance, is often the
       focus of exposure assessment studies. The upper tail of the distribution, which represents
       those individuals exposed to the highest concentrations, is frequently of special interest
 5     because the determination of the number of individuals who experience elevated pollutant
       levels can be critical for health risk assessments. This is especially true for pollutants  for
       which the relationship between dose and response is highly nonlinear.

       8.3.3 Relationship to Fixed-Site Monitors
10           Many early attempts to estimate exposure of human population used ambient air  quality
       from fixed stations. An example of such an analysis can be found in the 1980 Annual Report
       of the President's Council on Environmental Quality (CEQ, 1980).  In this analysis, a
       county's  exposure to an air pollutant was estimated as the product of the number of days that
       violations of the primary NAAQS were observed at county monitoring sites multiplied by the
15     county's  population. Exposure was expressed in units of person-days.  National exposure to
       an air pollutant was estimated by the sum of all county exposures.
             The methodology employed by CEQ provides a relatively crude estimate of exposure
       and is limited by four assumptions.

20           (1)    The exposed populations do not travel outside areas represented by fixed-site
                   monitors.
             (2)    The air pollutant concentrations measured with the network of fixed-site monitors
                   are representative of the concentrations breathed by the population throughout the
25                 area.
             (3)    The air quality in any one area was only as good as that at the location that had
                   the worst air quality.
30           (4)    There were no violations in areas of the county not monitored.
             Many studies cast doubt on the validity of this assumption for CO. Reviews of these
35      studies are provided by Ott (1982) and by Spengler and Soczek (1984).  Doubts over the

        March 12, 1990                           8-20      DRAFT-DO NOT QUOTE OR CITE

-------
10
       ability of fixed-site monitors alone to accurately depict air pollutant exposures are based on
       two major findings on fixed-site monitor representativeness.
            (1)   Indoor and in-transit concentrations of CO may be significantly different from
                  ambient CO concentrations.
            (2)   Ambient outdoor concentrations of CO that people come in contact with may vary
                  significantly from CO concentrations measured at fixed-site monitors.
            In estimating exposure, CEQ also assumed that each person in the population spends
       24 h at home. This assumption permitted the use of readily available demographic data from
       the U.S. Census Bureau.  Data collected 20 years ago indicate that people spend a substantial
15     portion of their time away from home. In a study of metropolitan Washington residents
       during 1968, Chapin (1974) found that people spent an average 6.3 h away from home on
       Sunday and 10.6 h away from home on Friday.  This translates to between 26.4 and 44.3%
       of the day spent away from home. More recent personal exposure and time budget studies
       (e.g., Schwab et al., 1989; Johnson, 1987) also indicate that a substantial portion of time is
20     spent away from home.
            Fixed-site monitors measure concentrations of pollutants in ambient air.  Ambient air
       has been defined by EPA in the Code of Federal Regulations (1977) as air that is "external to
       buildings, to which the general public has access." But the nature of modern urban life-
       styles in  many countries, including the United States, indicates that people spend an average
25     of over 20 h per day indoors (Meyer,  1983).  Reviews of studies on this subject by  Yocom
       (1982), Meyer (1983) and Spengler and Soczek (1984) show that indoor CO concentration
       measurements vary significantly from  simultaneous measurements in ambient air. The
       difference between indoor and outdoor air quality and the amount of time people spend
       indoors reinforces the conclusion that  using ambient air quality measurements alone will not
30     provide accurate estimates of population exposure.

       8.3.4  Alternative Approaches to Exposure Estimation
            In recent years, the limitations of using fixed-site monitors alone to estimate public
       exposure to air pollutants have stimulated interest in using portable monitors to measure

       March 12,  1990                          8-21      DRAFT-DO NOT QUOTE OR CITE

-------
       personal exposure. These instruments, which were developed for CO in the late 1970s by
       Energetics Science Incorporated and by General Electric, are called PEMs.  Wallace and Ott
       (1982) surveyed PEMs available then for CO and other air pollutants.  (See Chapter 5,
       Section 5.4, for a more complete description of PEMs.)
 5          The availability of these monitors has facilitated use of the direct and indirect
       approaches to assessing personal exposure (see Section 8.2).  Whether the direct or indirect
       approach is followed, the estimation of population exposure requires a "model"; that is, a
       mathematical or computerized approach of some kind.  Sexton and Ryan  (1988) suggest that
       most exposure models can be classified as one of three types:  statistical,  physical, or
10     physical-stochastic.
            The statistical approach requires the collection of data on human exposures and the
       factors thought to be determinants of exposure. These data are combined in a statistical
       model, normally a regression equation or an analysis of variance (ANOVA), to investigate the
       relationship between air pollution exposure (dependent variable) and the factors contributing
15     to the measured exposure (independent variables).  An example of a statistical model is the
       regression model developed by Johnson et al. (1986) for estimating CO exposures in Denver
       based on data obtained from the Denver Personal Monitoring Study.
            If the study group constitutes a representative sample, the derived statistical model may
       be extrapolated to the population defined by the sampling frame. It also  should be noted that
20     selection of factors thought to influence exposure has a substantial effect  on the outcome of
       the analysis.  Spurious conclusions can be drawn, for example, from statistical models that
       include parameters that are correlated with, but not causally related to, air pollution exposure.
            In the physical modeling approach, the investigator makes an a priori assumption about
       the underlying physical processes that determine air pollution exposure and then attempts to
25     approximate these processes through a mathematical formulation.  Because the model is
       chosen by the  investigator, it may produce biased results because of the inadvertent inclusion
       of inappropriate parameters or the improper exclusion of critical components.  The NAAQS
       Exposure Model (NEM) as originally applied to CO by Johnson and Paul (1983) is an
       example of a physical model.
30          The physical-stochastic approach combines elements of both the physical and statistical
       modeling approaches.  The investigator begins by constructing a mathematical model that

       March 12, 1990                          8-22      DRAFT-DO NOT QUOTE OR CITE

-------
       describes the physical basis for air pollution exposure.  Then a random or stochastic
       component that takes into account the imperfect knowledge of the physical parameters that
       determine exposure is introduced into the model.  The physical-stochastic approach limits the
       effect of investigator-induced bias by the inclusion of the random component, and allows for
 5     estimates of population distributions for air pollution exposure. Misleading results still may
       be produced, however, because of poor selection of model parameters. In  addition, the
       required knowledge about distributional characteristics may be difficult to obtain.  Examples
       of models based on this approach which have been applied to CO include the simulation of
       human activity and pollutant exposure (SHAPE) model (Ott,  1984; Ott et al., 1988) and
10     two NEM-derived models developed by Johnson (1988) and by Johnson and Wijnberg (1988).
            Table 8-4 provides a summary of the three model types. Table 8-5  lists exposure
       models which have been applied to CO by model type. These models are described in the
       following sections. General reviews of the exposure modeling literature have been provided
       by Repace et al. (1980), Ott (1985), Fugas (1986), Ott et al.  (1986), Sexton and Ryan
15     (1988), and Pandian (1987). EPA has developed a computerized bibliographic literature
       information system (BLIS) to facilitate access to literature concerned with total human
       exposure monitoring.  Included in the BLIS  data base is an extensive bibliography on human
       exposure modeling (Dellarco et al.,  1988; Shackelford et al., 1988).

20     8.3.5  Statistical Models Based on Personal Monitoring Data
            As discussed above, fixed-site  monitoring data may not provide an accurate indication of
       personal exposure within an urban population, which is a function of both geographic location
       (e.g., downtown vs. suburbia)  and immediate physical surroundings (e.g., indoors vs.
       outdoors).  Better estimates of personal exposure can be developed by equipping a large
25     number of subjects with portable monitors and activity diaries.  If the subjects are
       properly selected, their exposures can be extrapolated to a larger "target" population.
            Two such studies were conducted during the winter of 1982-1983 in Denver and
       Washington.  In the Denver study, each of 454 subjects carried a PEM and completed an
       activity diary for two consecutive 24-h sampling periods and provided a breath sample at the
30     end of each sampling period (Johnson, 1984).  Each participant also was requested to
       complete a detailed background questionnaire. The questionnaire results and approximately

       March 12, 1990                          8-23      DRAFT-DO NOT QUOTE OR CITE

-------
                   TABLE 8-4.  COMPARISON OF DIFFERENT APPROACHES TO
                             AIR POLLUTION EXPOSURE MODELING
       Parameter
                  Statistical
                     Physical
                     Physical-stochastic
10
15
20
Method of
formulation

Required
input
       Advantages
25     Disadvantages
Hypothesis
testing

Collected data
on human expo-
posure
                  Makes use of
                  real data in the
                  model building
                  process
30
                  Requires data on
                  hand for model
                  building;
                  extrapolation
                  beyond data base
                  is difficult
Physical
laws

Knowledge of
important parameters
and their values in
the system to be
modeled

True model
developed from
a priori consid-
erations
                     Includes
                     researcher's
                     biases; must
                     be validated
Physical Laws and
Statistics

Knowledge of
important parameters
and their distributions
in the systems to be
modeled.

Model developed
from a priori
considerations.
Stochastic part allows
uncertainty to contribute,
which reduces
importance of research
biases.

Requires much know-
ledge of system.
Must be validated.
35
40
Source:  Sexton and Ryan (1988).



900 subject-days of PEM and activity diary data collected between 1 November 1982, and

28 February 1983, were analyzed to determine if factors such as microenvironment and the

presence of indoor CO sources significantly affect personal CO exposure.  In addition, the

exposure of a defined target population was extrapolated from exposures recorded by the

study participants.  Detailed descriptions of the Denver study design and data collection

procedures, together with results of initial data analyses, are available in a report by Johnson

(1984).
        March 12, 1990
                                       8-24
                               DRAFT-DO NOT QUOTE OR CITE

-------
                 TABLE 8-5. MODELS WHICH HAVE BEEN USED TO ESTIMATE
                               CO EXPOSURE BY MODEL TYPE
       Model type
                Model
                                      References
10
20
25
30
35
       Statistical
15     Physical
Physical/
Stochastic
Regression models based on
statistical analyses of data
obtained from Denver and Washington
Personal Monitoring Studies

Results of ANOVA of data obtained
from Washington Commuter Study

NAAQS Exposure Model (NEM)

Ott-Willits Commuter Model

Simmon-Patterson Commuter Model

Davidson Indoor Mass-Balance
Models

Pierce Integrated Exposure Model

Duan Convolution Model

Duan Hybrid Model

Flachsbart Prototypical Commuter
Models

Flachsbart-Ah Yo Commuter Model

SHAPE
Probabilistic NEM

REHEX
                                                      Johnson et al. (1986)
Flachsbart et al. (1987)


Johnson and Paul (1983)

Ott and Willits (1981)

Simmon and Patterson (1983)

Davidson et al. (1984)


Pierce et al. (1984)

Duan (1985)

Duan (1985)

Flachsbart (1985)


Flachsbart and Ah Yo (1989)

Ott (1984)
Johnson and Wijnberg (1988)

Lurmann et al.  (1989)
40
Source: Adapted from Sexton and Ryan (1988).
       March 12, 1990
                                      8-25
                               DRAFT-DO NOT QUOTE OR CITE

-------
            The Washington study has been described in detail by Hartwell et al. (1984). It differs
       from the Denver study in that (1) twice as many subjects were used in the Washington study,
       and (2) each subject carried a PEM and a diary for a single 24-h period.  Results of analyses
       of the Washington data base are provided by Settergren et al. (1984), Clayton et al. (1985),
 5     and Johnson et al. (1986).
            A primary goal of the Denver and Washington personal monitoring studies was  to
       investigate whether personal exposures could be predicted by fixed-site ambient monitoring
       data. This investigation was conducted by performing linear regression analyses that  used
       PEM values grouped by microenvironment as the dependent variable and simultaneously-
10     recorded fixed-site values as the independent variable.
            To perform these analyses,  each PEM value had to be paired with a value reported by a
       single fixed-site monitor.  Because the census tract of each nontransit PEM value was known,
       it was possible to use census tracts as a means of linking PEM and fixed-site values.
       Whenever a PEM value was reported for a given census tract, it was paired with the
15     simultaneous value of the fixed-site monitor assigned to that census tract.
            This analysis suggested that a linear regression analysis that pairs each PEM value
       reported for a nontransit microenvironment with the simultaneous value reported at the nearest
       fixed-site might be appropriate for the Denver study data. Weighted linear regression
       analyses were performed with the data grouped by selected codes related to
20     microenvironment. Results for nontransit  microenvironments are listed in Table 8-6.  Values
       of R2 range from 0.00  to 0.46.  As might be expected, many of the microenvironments with
       small R2 values are associated with local CO sources that tend to reduce the correlation
       between PEM value and nearest fixed-site  value; however, other microenvironments not
       associated with local CO sources have relatively larger R2 values (e.g., park or golf course, or
25     "other locations").
            Table 8-6 does not list any  in-transit  microenvironments because of the difficulty in
       pairing in-transit PEM values with a  " nearest" fixed-site monitor value.  In the Denver data
       base, each in-transit PEM value has two census tract listings, one associated with  the start
       address and the other with  the end address. Neither was considered a good indicator of the
30     CO conditions encountered during the trip. An alternative procedure consisted of pairing
       in-transit PEM values  with simultaneous values from a composite data set created by

       March  12, 1990                           8-26       DRAFT-DO NOT QUOTE OR CITE

-------
rO
               TABLE 8-6. RESULTS OF WEIGHTED LINEAR REGRESSION ANALYSIS WITH NONTRANSIT
                  PEM VALUE AS DEPENDENT VARIABLE AND SIMULTANEOUS VALUE AT NEAREST
                                DENVER FIXED-SITE AS INDEPENDENT VARIABLE
Microenvironment
Category
Outdoors
Outdoors
Outdoors
Indoors

Indoors
Outdoors


Outdoors
Indoors
Outdoors
Indoors
Outdoors

Indoors
Indoors
Indoors
Indoors
Indoors
Indoors
Indoors
Outdoors
Not specified
Outdoors

Indoors

Indoors
Indoors
Indoors
Subcategory
Other location
Park or golf course
School grounds
Service station or motor
vehicle repair facility
Restaurant
Service station or
motor vehicle repair
facility
Within 10 yards of road
Church
Parking lot
Other repair shop
Sports arena, amphitheater,
etc.
Other public building
Shopping mall
Store
Health care facility
Residence
School
Office
Residential garage or carport
Not specified
Residential grounds
Public garage
Auditorium, sports arena,
concert hall, etc.
Manufacturing facility
Residential garage
Other location
n
115
18
15

112
486


11
468
178
51
46

16
111
55
675
333
20,969
342
2,090
22
583
70
139

94
41
66
381
Intercept
0.35
-0.09
-0.37

4.18
1.69


1.61
1.58
0.09
2.26
3.69

3.05
0.74
1.24
1.67
0.97
1.00
0.97
2.53
5.67
2.07
0.84
8.44

2.25
1.41
4.98
7.94
Linear regression*
Slope
1.11
0.39
1.15

1.68
0.76


1.21
0.89
0.70
0.60
0.88

-1.76
0.42
1.43
0.56
0.45
0.43
0.32
0.34
0.61
0.63
0.30
0.72

0.38
0.18
0.14
0.07

R2
0.46
0.44
0.27

0.27
0.25


0.23
0.21
0.21
0.21
0.18

0.15
0.14
0.14
0.09
0.09
0.07
0.07
0.05
0.05
0.05
0.04
0.04

0.04
0.03
0.00
0.00

Pb
0.000
0.003
0.049

0.000
0.000


0.134
0.000
0.000
0.000
0.003

0.128
0.000
0.005
0.000
0.000
0.000
0.000
0.000
0.304
0.000
0.099
0.019

0.060
0.246
0.662
0.791
      Listed in order of R value.

     b Probability that slope = 0.

     Source: Johnson et al. (1986).

-------
       averaging the data from the 15 fixed-site monitors.  The composite data set was found to
       exhibit relatively high correlations with most of the fixed-site data sets. Consequently, the
       composite site was assumed to provide an indication of the average ambient CO level in the
       study area.  Table 8-7 lists the results of linear regression analyses pairing in-transit PEM
 5     values with  simultaneous values from the composite data set.  Values of R2 range from 0.04
       (car) to 0.58 (motorcycle).
            The linear regression analyses described above suggested that the correlation between
       PEM values and fixed-site CO values is weak for most microenvironments.  A statistical
       analysis was subsequently performed to investigate whether the one-hour CO values reported
10     by a particular fixed-site monitor or groups of fixed-site monitors were better correlated with
       PEM values.  Again, the correlations were low with R2 values ranging from approximately
       0.01 to 0.05 (Johnson et al., 1986).
            Similar regression analyses were performed on the Washington CO data, and are shown
       in Tables 8-8 and 8-9.  Values of R2 range from 0.00 to 0.66. Several of the
15     microenvironments with small R2 values are associated with local CO sources that tend to
       reduce the correlation between PEM value and nearest fixed-site value.
            Only two nontransit microenvironments have R2 values exceeding 0.20: hospital (R2 =
       0.66) and church (R2 = 0.60).  The R2 value for office is 0.06; the R2 value for residence is
       0.02. The in-transit microenvironments also tend to have low R2 values (e.g., the R2 value
20     for car is 0.08).
           The analyses discussed above suggest that individual PEM readings are not highly
       correlated with simultaneous fixed-site readings.  Also it was learned that composite fixed-
       site daily maximum values are poor predictors of daily maximum exposures.  However, the
       magnitude of daily maximum eight-hour exposures  among the Denver study participants on
25     days when violations of the eight-hour NAAQS occurred (median exposure of 5.6 ppm)
       versus exposures on days when violations did not occur (median exposure of 3.2 ppm) was
       statistically  significant at the p< 0.001.

       8.3.6  Physical and Physical-Stochastic Models
30           In applying physical and physical-stochastic models,  the analyst constructs a
       mathematical model that describes the physical basis for air pollution exposure.  As discussed

       March 12, 1990                           8-28      DRAFT-DO NOT QUOTE OR CITE

-------
10
          TABLE 8-7. RESULTS OF WEIGHTED LINEAR REGRESSION ANALYSES WITH
                   IN-TRANSIT PEM VALUE AS DEPENDENT VARIABLE AND
                SIMULTANEOUS VALUE FROM DENVER COMPOSITE DATA SET
                                 AS INDEPENDENT VARIABLE
In-Transit
Subcategory
Motorcycle
Bus
Walking
Track
Car
All
Linear Regression
n
22
76
619
405
3,632
4,763
Intercept
4.50
3.17
0.06
3.27
6.01
5.15
Slope
2.14
2.02
1.47
1.54
0.78
0.92
R2
0.58
0.36
0.23
0.11
0.04
0.05
P'
0.000
0.010
0.000
0.000
0.000
0.000
15
       'Probability that slope = 0.
20     Source:  Johnson et al. (1986).

       above, physical-stochastic models differ from physical models in that the former include a
25     random component that reflects the analyst's imperfect knowledge concerning the physical
       parameters in the model.
            The Convolution and Hybrid Models - Duan (1985) evaluated two methods for
       estimating CO exposures which combine activity pattern data obtained from one source with
       data on CO levels measured in microenvironments obtained from another source.  Duan used
30     the Washington Personal Monitoring Study (Hartwell et al., 1984) as the source of activity
       pattern data and the Washington Commuter Study  (Flachsbart et al., 1987) as the source of
       the CO data. Each of 705 subjects in the former study completed a 24-h activity diary from
       which could be determined the sequence of microenvironments occupied by the subject. The
       latter study  measured CO levels in a variety of microenvironments on each of 43 days.  In the
35     first method - referred to as the Convolution Method - each of the 43 sets  of
       microenvironmental CO data was paired with each of the 705 person-days of activity diary
       data to yield (43)  (705) = 30,315 "convoluted" person-days of CO exposure.  The CO levels
       for all microenvironments occupied by a  subject on a given convoluted person-day are
       obtained from a single day of microenvironmental  monitoring data.
40          In the  second method - referred to as the hybrid approach - the average CO level across
       all 43 days was determined for each microenvironment and was used as the estimate of CO

       March  12, 1990                          8-29     DRAFT-DO NOT QUOTE OR CITE

-------
oo

u>
o
      TABLE 8-8. RESULTS OF WEIGHTED LINEAR REGRESSION ANALYSES WITH NONTRANSIT PEM VALUE

               AS DEPENDENT VARIABLE AND SIMULTANEOUS VALUE AT NEAREST FIXED-SITE

                             IN WASHINGTON, DC AS INDEPENDENT VARIABLE

Category
Indoors
Indoors
Indoors
Outdoors
Indoors
Outdoors
Indoors
Outdoors
Indoors
Indoors
Outdoors
Indoors
Indoors
Indoors
Indoors
Microenvironment
Subcategory"
Hospital
Church
Garage
Park, sports arena
Laboratories
Residential area
Office
Within 10 yards of road or street
Store
Residence
Garage, parking lot
Not specified
School, school gym
Restaurant
Other indoor

n
46
44
70
11
23
82
1,741
224
178
14,962
38
57
239
120
129

Intercept
-0.05
-0.04
4.02
0.06
0.30
0.53
0.94
1.33
1.25
1.21
5.05
3.52
1.01
2.88
5.07
Linear regression
Slope
0.63
0.58
3.43
-0.01
0.26
0.52
0.45
0.50
0.33
0.18
-0.42
-0.16
0.06
-0.03
0.09

R2
0.66
0.60
0.19
0.15
0.22
0.10
0.10
0.04
0.02
0.02
0.00
0.00
0.00
0.00
0.00

Pb
0.000
0.000
0.000
0.239
0.132
0.003
0.003
0.002
0.047
0.000
0.709
0.751
0.555
0.848
0.900
    "Listed in order of R2 value.



    •"Probability that slope = 0.



    Source: Johnson et al. (1986).

-------
10
15
20
25
30
35
40
45
         TABLE 8-9.  RESULTS OF WEIGHTED LINEAR REGRESSION ANALYSES WITH
                  IN-TRANSIT PEM VALUE AS DEPENDENT VARIABLE AND
          SIMULTANEOUS VALUE FROM COMPOSITE WASHINGTON, DC DATA SET
                               AS INDEPENDENT VARIABLE
In-Transit
Subcategory*
Train/subway
Jogging
Multiple response
Missing
Car
Truck
Bus
Walking
Van
Bicycle
Linear Regression
n
38
11
20
22
2,646
85
67
510
21
16
Intercept
0.05
0.43
-0.98
-0.21
1.51
2.16
1.01
1.21
1.91
3.62
Slope
1.09
0.67
2.58
1.83
1.74
2.00
2.45
0.94
0.33
-0.08
R2
0.61
0.25
0.20
0.13
0.08
0.07
0.05
0.03
0.03
0.01
P'
0.000
0.118
0.050
0.100
0.000
0.014
0.066
0.000
0.478
0.721
"Listed in order of R2 value.
""Probability that slope = 0.
Source: Johnson et al. (1986).

exposure whenever a diary-derived activity pattern indicated a subject was in the
microenvironment.  This method yielded 705 person-days of CO exposure.  The exposures
estimated by each of the two methods were compared to exposures indicated by the PEMs
carried by the Washington subjects.  The convolution and hybrid methods produced exposure
estimates that were, on average, approximately 40% higher than the PEM-derived exposure
estimates. Despite this discrepancy,  Duan (1985) found that the two methods were powerful
predictors of PEM-derived exposure  estimates, in that the correlations between model
estimates and PEM-derived estimates were relatively high.
     NAAQS Exposure Model (NEM) - In assessing the health risks associated with alternative
forms of NAAQS, EPA routinely uses the NEM to estimate the pollutant exposures of
       March 12, 1990
                                     8-31
DRAFT-DO NOT QUOTE OR CITE

-------
       sensitive population groups.  NEM itself is a general modeling framework that can be applied
       to estimate the exposures of the population to individual criteria air pollutants (Biller et al.,
       1981).  The general NEM framework, which continues to evolve over time, can be tailored to
       reflect the characteristics of particular air pollutants. NEM is designed to estimate population
 5     exposures under alternative values of the NAAQS.
            NEM divides time into finite intervals (e.g., 1 min,  10 min, 1 h) over which pollutant
       concentrations are assumed to be constant.  Geographic locations can be as small as tiny
       microenvironments (e.g., homes, automobiles) or they can be aggregated into larger physical
       areas (e.g., neighborhoods).  The population is divided into cohorts and their activity patterns
10     (movement into successive microenvironments over time)  are based on census data and
       information from transporation agencies (e.g., commuter travel times).  Human activities can
       be represented either as deterministic or stochastic variables, as can pollutant concentrations.
            In the initial application of NEM to CO (Johnson and Paul, 1983), four cities (Chicago,
       Los Angeles,  Philadelphia, and St. Louis) were selected as representative study areas.  In
15     applications of NEM to CO, the assumption was made that the CO concentration reflects
       (1) ambient CO levels as reported by outdoor fixed-site monitors and (2) sources and sinks
       specific to a microenvironment.  In the initial version of CO NEM, the CO exposure
       associated with an event occurring at time t in microenvironment m was estimated by a first
       order approximation which can be stated in general terms  as:
20
                             CO(m,t) = MULT(m) * MON(t) + ADD(m)                  (8-1)

       where MON(t) is the CO concentration expected to occur  at a fixed-site monitor at time t,
       MULT(m) is  a multiplicative constant specific to m, and ADD(m)  is an additive constant
25     specific to m  (Johnson and Paul, 1983).  This deterministic approximation  does not capture
       the findings of PEM studies  which point to relatively low  correlations between
       microenvironment exposure concentrations and fixed-site monitor concentrations.  Further,  it
       captures neither the stochastic nature of any relationship that might exist between exposure
       concentrations and fixed-site monitor concentrations nor the stochastic nature of source/sink
30     contributions  within a microenvironment.
       March 12, 1990                          8-32      DRAFT-DO NOT QUOTE OR CITE

-------
            A modified version treated the term ADD(m) as an independent, identically-distributed
        stochastic variable which could be characterized by the Box-Cox distribution (Johnson et al.,
        1988).  This change resulted in reduced levels of correlation between CO(m,t) and MON(t)
        that were in agreement with correlations observed in a personal monitoring study conducted in
 5      Denver, Colorado (Akland et al., 1985). A further refinement incorporates serial correlation
        (Johnson and Wijnberg,  1988).
            Simulation of Human Activity and Pollutant Exposure (SHAPE) - SHAPE simulates the
        activity patterns and CO exposures of a sample of urban commuters during their daily
        routines (Ott, 1984).  The simulation is over a fixed period for all individuals in the sample,
10      usually a 24-h period. The model uses the following equation (Duan, 1981, 1982):
                                                  J
                                            E, = Z 
-------
             A fundamental assumption about microenvironmental pollutant concentrations for inert
        pollutants such as CO in the SHAPE model is the "superposition hypothesis."  According to
        this hypothesis, the total concentration Cj(t) as a function of time encountered in
        microenvironment j is treated as the sum of two concentration components: (1) a
        microenvironmental component concentration cm(t) resulting from the sources of CO within
        the microenvironment, and, (2) an ambient (background) component concentration cu(t)
        assumed to be free of any microenvironmental source influences; that is,
                                        c,(t) = [cm(t) + c.(t)],,                            (8-4)
10
       The basis for this hypothesis is the interpretation of the spatial variability of CO
       concentrations from field studies (Ott, 1971; Ott and Eliassen, 1973).
            In the SHAPE model, the microenvironmental component depends only on the sources
       of CO within the microenvironment and is independent of location in the urban area or of
15     conditions in the metropolitan area.  An example is the CO concentrations contributed by
       motor vehicles inside an indoor parking garage.  In contrast, the background concentration
       component is the CO concentration that would be present if there were no specific sources of
       CO.  For example, in a house or building, the background component would be the CO in the
       outdoor air entering through the ventilation system  or the windows, which depends primarily
20     on seasonal and daily changes in meteorological conditions.
            Because data on true ambient background concentrations of CO are generally unavailable
       for the many microenvironments that an urban population regularly visits on a  daily basis, Ott
       et al. (1988) investigated the use an "overall surrogate" ambient CO concentration thought to
       be associated with all the microenvironments  of an urban area.  Usually, the only data
25     available to serve as an overall surrogate measurement are CO concentrations measured by
       fixed monitoring stations located in metropolitan areas. These data may yield unrealistically
       high estimates of ambient CO levels  as most air monitoring  stations are placed  near streets
       with heavy traffic. Ott recommended using the ambient component given by the hourly CO
       readings from fixed-site monitors located away from streets, but the hourly average of all
30     fixed stations in Denver performed satisfactorily.
       March 12, 1990                          8-34      DRAFT-DO NOT QUOTE OR CITE

-------
            The original version of SHAPE (Ott, 1981, 1984) assumed that pollutant concentrations
       in microenvironments behave stochastically.  This assumption was based partly on a study by
       Ott and Willits (1981) in which CO concentrations inside an automobile passenger
       compartment were found to show considerable random fluctuation from minute to minute.
 5     The CO data were collected on drives during a one-year study of an urban arterial highway,
       El Camino Real in California. Statistical analysis indicated that the one-minute average CO
       concentration [Cj(t)] could be treated as independent, lognormally distributed random variables
       during the length of a car trip (one hour or less). Ott and Willits (1981) developed these
       conclusions for the exposures incurred by the occupants of vehicles free of CO intrusion from
10     the vehicle's own exhaust system.
            The SHAPE model (Ott, 1981, 1984) was  designed with these findings in mind.  All
       microenvironmental CO component concentrations were represented by stationary two-
       parameter lognormal  distributions with cu(t) held constant. Thus, the computer treated the
       microenvironmental component as the random variable [cjj whose mean and variance for
15     each microenvironment j where specified by the user and were held constant.  The values of
       the mean and variance usually were based on CO field studies in various microenvironments
       reported in the literature (inside moving  automobiles, buses, trucks; on bicycles in traffic; in
       indoor parking garages, houses, and similar environments) and on the judgment of the user.
            The SHAPE model (Ott, 1981, 1984) originally sampled microenvironmental CO
20     concentrations on a minute-by-minute basis.  Fourteen microenvironments were defined for
       this purpose.  Associated with each  was  a lognormal distribution of one-minute CO values
       from which one-minute CO exposures were drawn.
            The original SHAPE model  simulated activity patterns for each individual by sampling
       from probability distributions representing the chance of entry, the time of entry, and  time
25     spent in specific activities or microenvironments (Ott, 1981).  For example, the probability
       distributions for the starting times of home-to-work trips, trip times, and travel modes (the
       proportion of commuters traveling to work by car, bus, truck, and such) were based on data
       provided by Svercl and  Asin (1973). Reliable data were not available for some activities;  in
       these cases the probability distributions were  assumed by the user.
30          In an attempt to validate SHAPE, Ott et al. (1988) compared measured personal CO
       exposures obtained from the Denver personal monitoring study to CO exposures estimated by

       March 12,  1990                          8-35      DRAFT-DO NOT QUOTE OR CITE

-------
       SHAPE. Microenvironmental CO concentrations for the model were generated by Monte
       Carlo simulation based on Denver PEM data reported for 22 microenvironments.  The
       activity simulation portions of the model were modified to accommodate actual activity data
       obtained from the diaries carried by Denver subjects.
 5          A total of 899 24-h responses from the Denver study yielded 772 usable profiles after
       invalid responses were eliminated, giving 33 paired days of observations (CO exposure
       profiles from two successive days for the same respondent).  From these data,
       22 microenvironments were identified with at least 10 measurements on each of the two days.
       Microenvironmental CO concentrations were calculated by subtracting hourly ambient
10     background CO concentrations.  Ambient background CO concentrations were estimated by
       three different approaches.  All three yielded similar results, with the average value from all
       fixed monitoring sites performing slightly better than the nearest fixed monitoring site.  For
       nearly every microenvironment, the study found negligible differences between the
       microenvironmental CO frequency distributions on the two days, showing the statistical
15     stability of the microenvironmental concentrations.
            In the SHAPE validation project (Ott et al., 1988), the microenvironmental CO
       frequency distributions for Day 1 provided the  basis for SHAPE model estimates of Day 2
       exposure profiles, and the activity patterns were based on the Denver diaries  for Day 2 (the
       observed times at which people entered and left each microenvironment).  The CO exposure
20     profiles were calculated using Monte Carlo sampling from the Day 1  microenvironmental CO
       concentration distributions and adding the estimated ambient background components.
            The arithmetic means of the predicted one- and eight-hour maximum average CO
       exposures agreed well with the corresponding observed arithmetic means. The variability of
       the observed values, however, exceeded the variability of the predicted values by a significant
25     amount (Figures 8-3 and 8-4).  Ott et al. (1988) suggested that the lack of agreement may be
       caused by use of a histogram rather than a continuous distribution in implementing the Monte
       Carlo simulation or the model's implicit assumption that the successive exposures of a subject
       are uncorrelated.  Ott et al.  (1988) suggested that better estimates would result if an
       autoregressive process was used to model successive exposures.
30          Commuter Exposure Models - Ott and Willits (1981) conducted a study in which the CO
       exposures of occupants of a motor vehicle were measured by weekly drives on an urban

       March 12,  1990                          8-36      DRAFT-DO NOT QUOTE OR CITE

-------
     2
     LU
     o
     o
     o
     o
     u
          00' 0 OS 01 0 2  o.S  1  2
        100
         90
         8(1
         70
         60

         bO
         40

         30
         20
                        CUMULATIVE FREQUENCY, X
                      10   20  30 40 SO SO 70  10   10
                                                                 M  99 99.S99J99J  99.99
_  I I
         1
        09
        08
        07
        06

        0.5

        04

        0.3


        02
        01
             I  I
                   I    I
                           I   I
I  I   I
                                               I
I    III
OBSERVED
Tl    _
                                 OBSERVED:
                                   (n-336)
                                 PREDICTED:
                                    Composite
                                   (n-336)
            Mean:
            S.D.:
            Max:

            Mean:

            S.D.:
            Max:
              111   III
                                                            PREDICTED
     10.2 ppm
      8.9 ppm
     70.7 ppm


     10.6 ppm

      6.0 ppm
     42.7 ppm
                                 I
                                     I
                                        I
                                                I	I
                                                     J	I
                                                                      I
         0.01  O.OS 0.1 0.2 0.5 1
                                 10   20  30 40 SO 60  70  10   90
                                  CUMJLATIVE FREQUENCY, %
100
 99
 10
 70
 (0
 SO
 40

 30
                                                                                   20
                                                                                   10
                                                                                   9
                                                                                   7 —
                                                                                   6 §
          1
         0.9
         OJ
         0.7
         0.6
         O.S

         0.4

         0.3


         0.2
                                                                                  0.1
                                                                91 H 99i S9J 91.}   99.99
                                                                          O
Figure 8-3. Logarithmic-probability plot of cumulative frequency distribution of maximum
one-hour average exposure of CO predicted by SHAPE, plus an observed frequency
distribution for Day 2 in Denver.

Source: Ott et al. (1988).
March 12, 1990
                                8-37
        DRAFT-DO NOT QUOTE OR CITE

-------
         2
cc
t-
2
UJ
u
O
u
o
o
            100
              001  005 01 0.2 0.5  I   2
 CUMULATIVE FREQUENCY, %

10   20  30  40  50 60  70  10   90
                                                                   95
                                                                       M  99 99.5 99J 99.9  99.99
            90
            80 -
            70 -
            60 —

            59 —

            40 —

            30 -


            20 —
             1
            09
            0.8
            o;
            06
            05
            02
            0 1
                                      I
                                           r
             r i
   i
                                        OBSERVED
                                                                                100
                                                                                90
                                                                                10
                                                                              —1 70
                                                                                10
                                                                                58
                                                                                40
                            Mean:
                            S.D.:
                            Max.:
               PREDICTED:
                (n=336)
               1. Composite of fixed stations
                            Mean:
                            S.D.:
                            Max:
               2. Nearest fixed station
                            Mean:
                            S.D.:
                            Max:
               3. No-Source microenvironment
                            Mean:     3.8ppm
                            S.D.:     1.9ppm
                            Max.:     11 3 ppm
                                                                           4.9 ppm
                                                                           4.2 ppm
                                                                          38.7 ppm
                    4.8 ppm
                    2.4 ppm
                   12.4 ppm

                    4.4 ppm
                    2.7 ppm
                   15.4 ppm
                                                   I
                                                        I
                                                               I
                                                                   I
                                                                       I
                                                                            I
                                                                               I  I
              OCt  0.050.102  05  1  2
                                      10   20  30  40 50 SO 70 10   90
                                       CUMULATIVE FREQUENCY, %
                                                                                        20
                                                                         99 99.5 99.1 99.9   99.99
                                                                                             a
                                                                                         4  S
                                                                                1
                                                                                0.9
                                                                                0.1
                                                                                0.7
                                                                                0.6
                                                                                0.5

                                                                                0.4

                                                                                0.3
                                                                                        0.2
                                     z
                                     UJ
                                     
-------
       artery in California.  The study consisted of 93 repeated drives over exactly the same route -
       5.9 miles in each direction for an 11.8 mile total distance - on El Camino Real with a 1974
       VW test vehicle.  Data were collected on CO levels inside the passenger compartment of the
       motor vehicle, traffic counts and time spent waiting at each traffic light, meteorological
 5     factors, as well as other variables.  Measurements in the vehicle showed that the passenger
       compartment was free of self-generated CO intrusion. Ott and Willits developed a theoretical
       model for estimating diffusion of CO into a motor vehicle and then applied the model to data
       collected during the study.  The model incorporates a time constant that was found to vary
       according to the position of the windows (closed, partially open,  completely open).
10           Simmon and Patterson (1983) developed a model for simulating commuter exposures
       individually and collectively based on traffic flow, emissions, and atmospheric dispersion.
       This model consists of two programs that are run separately on the computer.  The first
       program is an emissions preprocessor, which has been separated from the main model to
       facilitate updating of the model package when emission factors are revised by EPA. The
15     second program is the main portion of the commuter exposure model, which simulates traffic
       flow, computes the emission rates resulting from the traffic (using the emission factors
       calculated by the preprocessor), simulates the dispersive effects of the atmosphere, and
       computes statistics describing commuter exposure.  Because the model treats the spatial
       variation of exposure, regions of the city in which commuters experience high exposures can
20     be identified from model output.  If a single commute pathway is of interest, that pathway
       can be examined in detail.  Dispersion modeling is performed by the CALINE 3 model.  In-
       vehicle CO concentrations are assumed  to equal roadway CO  concentrations. To date, the
       Simmon-Patterson model has not been used in a modeling analysis.
             Petersen and Sabersky (1975) conducted experiments to  measure the CO concentrations
25     inside a vehicle under typical driving conditions during the summer in Los Angeles, CA.
       They observed that the average CO concentrations inside the vehicle were about equal to the
       outside air concentrations.
             Petersen and Allen (1982) conducted a similar experiment in Los Angeles over 5 days in
       October  1979.  They found that the average ratio of interior to exterior  CO concentrations
30     was 0.92. However, the hourly average interior CO concentrations were 3.9 times higher
       than the  fixed-site measurements. In their analysis of the factors that influence interior CO

       March 12,  1990                          8-39      DRAFT-DO NOT QUOTE OR CITE

-------
       levels, they observed that traffic flow, and traffic congestion (stop-and-go) are important, but
       "comfort state" (i.e., car windows open/closed, fan on/off, etc.) and meteorological
       parameters (i.e., wind speed, wind direction) have little influence on incremental exposures.
            Flachsbart (1985) developed three empirical models for predicting commuter exposure
 5     inside a well-ventilated vehicle on  a congested Honolulu artery during morning rush hour
       under neutral atmospheric stability. PEMs were used to collect exposure data for commuting
       trips on 12 days between November 1981 and April 1982. Model A assumed that commuter
       CO exposure was a function of the roadway's source strength and the ambient CO level, as
       expressed in the following equation.
10
                     Commuter CO = (0.00012728)(CO emissions/mi)1-06 + ambient CO     (8-5)

       Model B assumed that the roadway's source strength was diluted by windspeed, as expressed
       in the following equation.
15
                    Commuter CO = (0.0001972) (CO emissions/mil1-039 + ambient CO     (8-6)
                                               (windspeed)0083

20
       Model C assumed that commuter CO exposure was simply a function of the emission factor
       and ambient CO level.  This model was developed  for situations for which the analyst does
       not have access to traffic counts.

25                  Commuter CO = (0.0713358)(CO emission factor)1-289 + ambient CO     (8-7)
            Flachsbart considered these models to be prototypes because they were the first such
30     models to link commuter exposure, inside a vehicle, directly to automotive emission factors.
       Each model assumed that exposure is an additive function of a background CO level and a
       roadway CO contribution, as affected by meteorological and traffic characteristics.
        March 12, 1990                         8-40      DRAFT-DO NOT QUOTE OR CITE

-------
            Flachsbart compared the observed exposure values with exposures predicted by each of
       the three models. Correlations between observed and predicted values, expressed as R2, were
       0.78 for Models A and B, and 0.64 for Model C.
            Flachsbart's prototypical models (1985) served as model templates in subsequent efforts
 5     by Flachsbart and Ah Yo (1989) to develop a general model of commuter exposure based on
       data obtained from a study of commuter exposures in Washington.  Their approach described
       commuter exposure on  a specified commuting link with an expression that superimposes a
       microenvironment component upon a background concentration:

10                                         E, = B; + M,                                (8-8)
       where:
               E, = commuter exposure on link i,
15             B; = background concentration on link i,
               M  = microenvironment concentration on link i.
20          Ideally, the background concentration should be measured near the link and should
       reflect concentrations that would exist on the roadway if there were no traffic.  Flachsbart
       (1985) approximated this value with CO ambient air quality readings from the nearest fixed-
       site station away from heavy traffic. The microenvironment component mathematically
       describes the air pollutant emission and dispersion processes over the roadway.  This
25     component also considers how the air pollutant infiltrates the vehicle's interior.  An air
       pollutant infiltration factor, however, was not included in Flachsbart's prototypal models,
       since there was a free exchange of air between the vehicle and the ambient environment.
            Using the format of the Honolulu prototypal models, Flachsbart and Ah Yo (1989)
       developed 33 commuter exposure models from the Washington survey data base. Of these
30     models, only five were considered unsatisfactory based on the statistical significance of the
       model  or an illogical sign for the emission coefficient. However, the explanatory power of
       the best of these models (R2 =  0.12) did not approach that of the worst Honolulu model (R2
       = 0.63).

       March  12,  1990                          8-41     DRAFT-DO  NOT QUOTE OR CITE

-------
            Flachsbart and Ah Yo (1989) found use of the Honolulu prototypal models for
       characterizing the Washington data to be overly simplistic.  For morning trips originating in
       low density suburbs, the Washington data showed that a commuter's average exposure to CO
       was less than the ambient concentration measured at the fixed-site station providing the
 5     background concentration. In addition, commuters, who began their homeward evening trips
       from highly polluted parking garages, had unusually high concentrations in their cars as they
       traveled along downtown streets.  Tests of each vehicle at the beginning, middle, and end of
       the study indicated that the high CO levels were not caused by leaks from the exhaust system
       into the passenger compartment.
10          These observations suggested that vehicle occupants were, to some degree,
       "encapsulated" from the ambient environment such that their exposure on the early links of a
       trip had more to do with the concentration inside the vehicle (prior to the trip) than with any
       traffic or meteorological factors on these links.  Statistical analysis supported this hypothesis.
       For the evening commute from downtown Washington on Route 1,  the average CO exposure
15     on the link was significantly correlated with pretrip interior CO concentrations.  For the
       morning commute into downtown Washington on Route 2,  the average link exposure was well
       correlated with pretrip interior CO concentrations.
            Given winter temperatures and closed windows and vents on the test vehicles, Flachsbart
       and Ah Yo (1989) decided to treat the roadway setting and the vehicular passenger
20     compartment as separate microenvironments.  Each microenvironment was modeled separately
       and then combined into a two-stage model.
            The data base available for development of a roadway CO empirical model was limited
       to 150 measurements of roadway CO.  Of 43 different models applied to this data set, the
       best model was a loglinear relationship between predicted roadway CO concentrations and the
25     density of CO emissions. This density was the product of the CO emission factor and the
       average 15-min traffic count divided by the test vehicle's average link speed. This model had
       an R2 value equal to 0.26; the F statistic was significant at p< 0.0001.  The final step of the
       regression left two independent variables in the equation: the CO emission factor and the
       average 15-min traffic count.
30
        March 12, 1990                          8-42      DRAFT-DO NOT QUOTE OR CITE

-------
            The equation for this model was:
                                      = (0.7906252[(Fe)(QT)/Uv]036fiW2                      (8-9)
       where:
                  = predicted CO concentration within the roadway microenvironment (ppm),
            Fe    = MOBILE3 emission factor estimated using observed traffic speeds, ambient
                    temperatures, the percentages of five vehicle types, and default values for other
10                  required inputs (g/veh-mi),
            QT   = observed average 15-min traffic count (veh/15 min),
            Uv   = test vehicle's average link speed.
15
            Flachsbart and Ah Yo (1989) assumed that commuters are exposed to CO from three
       major sources within the passenger compartment:  passenger smoking, vehicle exhaust system
       leaks, and emissions from traffic.  Flachsbart and Ah Yo (1989) further assumed that the in-
20     vehicle CO concentrations created by these three sources can be described by box or cell
       models.  Such models are based on the principle of conservation of mass:  The total mass of
       an air pollutant within a volume is equal to the balance of the mass entered, exited, emitted,
       and reacted within that volume.  Using this principle, Flachsbart and Yo (1989) derived a
       theoretical commuter exposure model for the passenger compartment:
25
                               E = COR + (T/tR)[COv - COJtl - e(lR>/T]                   (8-10)
       where:
30            E  =  average CO exposure of the commuter (ppm),
            COR =  observed CO concentration within the roadway microenvironment (ppm),
              T  =  time constant for the vehicle (s),
             tR   =  time vehicle spends within the roadway microenvironment (s),
            COV =  CO concentration within the vehicle when it enters the microenvironment (ppm),
35            e  =  base of a natural logarithm (2.72).
       March 12, 1990                          8-43      DRAFT-DO NOT QUOTE OR CITE

-------
            This model predicts commuter exposure to CO inside a vehicle by exponentially
       diffusing observed roadway concentrations and by exponentially decaying initial
       compartmental concentrations that exist when the vehicle enters a new link on the roadway.
       The plot of the observed average CO exposure with average CO exposures estimated with
 5     Equation 8-10 suggested a linear relationship.  These data had an R2 = 0.75 and the
       significance of the F statistic was p< 0.001.
            Predicted values CO^ generated by the roadway microenvironmental model
       (Equation 8-9) were substituted for the observed roadway concentrations COR in the passenger
       compartment model (Equation 8-10).  The values estimated by the resulting two-stage model
10     correlated well with the observed values (R =  0.737); the coefficient of determination (R2)
       indicated that the estimates explained approximately  54% of the variation in observed average
       exposures. Although the two-stage model did not have the predictive power of the passenger
       compartment model which used observed roadway CO concentrations, Flachsbart and Ah Yo
       (1989) considered the performance of the two-stage model to be respectable and far better
15     than any of the 33 models initially developed.
            Other Exposure Models - Davidson et al. (1984) developed one- and two-compartment
       mass-balance models for estimating indoor pollutant concentrations. They compared
       measured levels of NO, NO2, and CO in a new townhouse residence with estimates provided
       by the one-compartment model.  The townhouse was constructed according to rigid energy-
20     conservation guidelines.  Reasonable agreement between estimated and measured
       concentrations was observed, although the measured CO levels decayed  somewhat faster than
       predicted.
            Pierce et al. (1984) presented a model for estimating integrated (i.e., cumulative) and
       average exposures based on an activity pattern listing a sequence of indoor and outdoor
25     locations and estimates of the pollutant concentration at each  location. The model was used to
       estimate CO exposures for a hypothetical 24-h activity pattern.  The CO level assigned to
       each location was derived from microenvironmental monitoring data obtained from other
       researchers.
30
        March 12, 1990                          8-44      DRAFT-DO NOT QUOTE OR CITE

-------
       8.4 OCCUPATIONAL EXPOSURE TO CARBON MONOXIDE
            Carbon monoxide is a ubiquitous contaminant occurring in a variety of settings.
       Exposures, both acute and chronic, that occur in the occupational environment represent only
       one of several sources that may contribute to a potential body burden for carbon monoxide.
 5     Two main sources for background exposures in both occupational and nonoccupational
       settings appear to be smoking and the internal combustion engine (National Academy of
       Sciences, 1969).  Smoking is a personal habit that must be considered in evaluating exposures
       in general,  as well as those occurring in work places.
            In addition, work environments are often located in densely populated areas, and such
10     areas frequently have a higher background concentration of CO compared to less densely
       populated residential areas. Thus, background exposures during work hours may be greater
       than during nonwork hours. There are several sources other than smoking and the internal
       combustion engine that contribute to exposure during work hours. These include
       contributions to background by combustion of organic materials in the geographic area of the
15     work place, work in specific industrial processes that produce CO, and work in environments
       that result in accumulations of CO, such as garages, toll booths, and confined spaces.

       8.4.1   Historical Perspective
            Production of CO results from incomplete combustion of organic substances such as
20     natural gas, coal, wood, petroleum, coal, coke, vegetation, explosives, and manufactured gas.
       A rich fuel mixture favors generation of CO. Carbon monoxide also can be produced when
       rapid cooling or submersion of the flame is used to quench the combustion process.
            Dangerous concentrations of carbon monoxide can occur in numerous settings, including
       environmental background, the home, and the street - at work or play.  Sources include
25     exhaust gases from  internal combustion engines, gas manufacturing plants, blast furnaces in
       iron and steel manufacturing, coke ovens, coal mines, incinerators, and numerous other
       processes that involve combustion of organics.  CO is used in the manufacture of metal
       carbonyls, and CO is produced in industrial  quantities by the partial oxidation of
       hydrocarbons or natural gas, and by gasification of coal or coke (Lundgren,  1971).  (See
30     Chapter 6 for a more complete discussion of sources and emissions of CO.)
       March 12, 1990                          8-45     DRAFT-DO NOT QUOTE OR CITE

-------
            Both chronic and acute CO intoxication in a variety of occupations and processes is
       discussed by Grut (1949).  Acute effects related to production of anoxia from exposures to
       CO historically have been a basis for concern.  In recent years, however, this concern has
       grown to include concerns for potential effects from chronic exposures as well (Rosenstock
 5     and Cullen, 1986a, 1986b; Sammons and Coleman, 1974).
            With regard to the occupational environment, the National Institute for Occupational
       Safety and Health (1972) published "Criteria for a Recommended Standard...Occupational
       Exposure to Carbon Monoxide."  NIOSH observed that"... the potential for exposure to
       carbon monoxide for employees in the work place is greater than for any other chemical or
10     physical agent." NIOSH recommended that exposure to CO be limited to a concentration no
       greater than 35 ppm, expressed as TWA for a normal eight-hour workday, 40  hours per
       week.  A ceiling concentration was also recommended at a limit of 200 ppm, not to exceed an
       exposure time greater than 30 min.  Occupational exposures at the proposed concentrations
       and conditions underlying the basis of the standard were considered to maintain COHb in
15     blood below 5 %.  The Occupational Safety and Health Administration has recently adopted
       these exposure limits in order to substantially reduce the risk of deleterious health effects
       among American workers (Federal Register, 1989).
            Although it was not stated, the basis of the recommended NIOSH standard (i.e.,
       maintaining COHb below 5%  in blood), assumes that contributions from other
20     nonoccupational sources would also be less than a TWA concentration of 35 ppm.  It was
       recognized that such a standard may not provide the same degree of protection to smokers,
       for example.  Although recognizing that biologic changes might occur at the low level of
       exposure recommended in the proposed standard, NIOSH concluded that subtle aberrations in
       the nervous system with exposures producing COHb concentrations in blood at or below 5%
25     did not demonstrate significant impairments which would cause concern for the health and
       safety of workers.  In addition, NIOSH observed that individuals with impairments that
       interfere with normal O2 delivery to tissues (e.g., emphysema, anemia, coronary heart
       disease)  may not have the same degree of protection as for less impaired individuals. It also
       was recognized that work at higher altitudes (e.g., 5000 to 8000 feet above sea level) would
30     necessitate decreasing the exposure limit below 35 ppm, to compensate for a decrease in the
       oxygen partial pressure as a result of high altitude environments and a corresponding decrease

       March 12, 1990                          8-46       DRAFT-DO NOT QUOTE OR CITE

-------
       in oxygenation of the blood.  High altitude environments of concern include airline cabins at a
       pressure altitude of 5000 feet or greater (National Research Council, 1986) or work in high
       mountain tunnels (Miranda et al., 1967).

 5     8.4.2  Exposure  Monitoring Techniques
            Exposures to CO in air can be manifested in a variety of ways.  At low levels,
       manifestations include development and reporting of symptoms. In the work place,
       environmental monitoring and inventory of sources for the presence of CO may occur.
       Additionally, biologic tests, medical surveillance, diagnosis and treatment may be conducted
10     on individuals who show signs and symptoms of exposure.  Finally,  mathematical models
       may be used to predict exposures, doses, and responses to CO inhalation.
            Acute and chronic CO intoxication (Grut,  1949) may be indicated by a range of signs
       and symptoms from headache, dizziness, weakness, and nausea at low levels and short
       durations of exposure to unconsciousness, coma, and death at high levels and durations of
15     exposure.  Headache and nausea resulting from CO intoxication has been described in a study
       of tollbooth collectors (Johnson et al., 1974) exposed at low concentrations of CO from
       exhaust gases.
            A medical study of the occupational hazards of fire fighting demonstrates the signs and
       symptoms of CO, as well as other associated exposures (Gordon and Rogers, 1969). A group
20     of 35 fire fighters were evaluated in a medical study for heart, lung, liver, and kidney
       diseases, and were also provided neurologic examinations.  Half of the study group were
       smokers. Baseline tests including enzyme tests, EKG,  COHb, and other measurements were
       conducted at the start of the study. The fire fighters were in normal ranges for COHb and the
       enzyme tests performed. They were followed through  31 fires of less than five minutes,
25     4  fires of more than five minutes, and 6 staged fires; they were also subjected to exercise
       tests. Occasionally, there were substantial exposures to CO, and changes in blood enzyme
       levels were greater when fighting longer fires.  These changes were not associated with
       exercise, and they were reversible when not fighting fires.  EKG tests did not reflect changes
       related to enzyme levels.  Masks were found to provide substantial protection.
30           Occupational exposure and associated signs and symptoms for fire fighters also have
       been described in a study using age-matched controls (Sammons and Coleman,  1974).  Blood

       March 12, 1990                          8-47      DRAFT-DO NOT QUOTE OR CITE

-------
       samples were collected from a group of 27 fire fighters and a group of 27 control subjects
       every 28 days for five months.  Differences between the cardiac enzyme levels found in fire
       fighters versus those of the matched control suggested that chronic low level exposures to CO
       have a deleterious effect on the body and myocardium.
 5          Environmental monitoring for CO is often carried out in studies that are primarily
       concerned with potential exposures to other substances, such as exhaust gases, environmental
       tobacco smoke and combustion processes. CO analyses also are used to screen for the
       presence of other gaseous pollutants.  Exposures to CO therefore are often associated with
       exposures to other substances as well, including lead, paniculate matter containing
10     polyaromatic hydrocarbons (PAHs), NOs, and SO2.
            Monitoring for exposures to CO has included peak and TWA sampling of ambient or
       breathing zone air, collection and analysis of expired air, analysis of blood gases by gas
       chromatographic methods, and use of empirical relationships to estimate CO in air from
       determinations of percent COHb in blood.  Measurements techniques for CO include infrared,
15     volumetric, colorimetric tubes, electrolytic detection, and gas chromatographic methods.
       Samples are collected to represent the breathing zone or environmental air; these may be grab
       samples or periodic or continuous samples.
            Several investigators have proposed approaches to medical surveillance of workers who
       are potentially exposed to carbon monoxide.  Medical surveillance activity is usually
20     precipitated by complaints that are associated with a source of potential exposure to CO.    A
       recent study of stevedores who loaded and unloaded cars and diesel trucks in a ferrying
       operation (Purdham et al., 1987), assessed medical conditions by administering a
       questionnaire and conducting pulmonary function tests.  The questionnaire included questions
       on work history; smoking history; respiratory symptoms; and nose, eye, and skin complaints.
25     Questions on respiratory symptoms included details on cough, sputum,  wheeze, chest
       tightness, and shortness of breath.  Pulmonary function tests were conducted for forced vital
       capacity (FVC) and forced expiratory volume at one minute (FEV,).  The subjects were
       seated and their noses were clipped closed for the tests. A minimum of three and as many as
       six efforts were required for each subject.
30          The focus of the study was on characterizing adverse responses  to exhaust fumes,
       primarily from diesel trucks, and secondarily from gasoline-powered  vehicles transported in

       March 12, 1990                          8-48     DRAFT-DO NOT QUOTE OR CITE

-------
        garages on the ferry. The medical findings were that the stevedores had significantly lower
        values for all lung function tests except for FVC, as compared to unexposed controls, and to
        normal values for a general population.  Environmental sampling for CO and other substances
        in exhaust fumes then were conducted. There was no direct correlation offered by the authors
  5     between exposure to CO and the differences found in the medical assessments of the
        stevedores versus the controls.  The authors suggested use of percent COHb to assess CO
        exposure, and they recommended that a larger group of longshoreman should be assessed for
        chronic obstructive pulmonary disease.
             Rosenstock and Cullen (1986a) have linked cardiovascular diseases occasioned by angina
 10     at the end of a workday with high percent COHb when this phenomenon is associated with
        exposure to CO in the work place. Consequences of chronic low level exposure are not well
        established; however, in workers with underlying coronary artery heart disease, a level of
        3 to 5% COHb has been associated with increasing frequency of angina and decrease in
        exercise tolerance (See Chapter 10). When levels approaching 25% COHb are reached, there
 15     are manifestations of ischemia, dysrhythmias and EKG abnormalities in otherwise healthy
        workers.
             Miranda et al.  (1967) feel that medical surveillance for high altitude work should
        include screening for cardiopulmonary abnormalities and blood dyscrasias (sickle-cell
        anemia).  They also  recommend acclimatization before the start of work.
20          This study classified the onset of CO intoxication into three groups: fulminating (a
        decrease in  O2 to tissues within seconds), acute (a decrease in O2 occurring in minutes), and
        chronic (a decrease in O2 oxygen to tissues over days, months, or years).
             Miranda et al. (1967) listed the concerns for evaluation of CO exposures at high
        altitudes as:  decreased oxygen in the air,  percent COHb due to smoking, and accumulation of
25      fumes, particularly in vehicular tunnels.  Altitude tolerance is lowered by about 335 ft for
        each percentage point increase in COHb.  The average percent COHb for smokers who smoke
        20 to 30 cigarettes per day is 5%, with a range of 3 to 10%. To decrease from 20 to
        5%  COHb requires breathing fresh air at sea level for three to five hours. Inhalation of CO
        at a concentration of 100 ppm for two hours at 11,000 feet results in 18 ± 5% COHb.  This
30      level does not threaten survival, but it may impair visual threshold  (see Chapter 10).
       March 12, 1990                          8-49     DRAFT-DO NOT QUOTE OR CITE

-------
            Empirical relationships have been proposed for use as diagnostic criteria for CO
       intoxication (Castellino, 1984).  The criteria proposed are shown in Table 8-10.
            Blankart et al. (1986) found hyperbaric O2 to be the best form of treatment for
       decreasing the percent COHb in blood when it was administered to traffic policemen in a
 5     clinical study.  The study compared cycling ergometry, administration of pure O2 at
       atmospheric pressure, and administration of pure O2 under hyperbaric conditions (2.8 atm).
       The authors recommended the hyperbaric treatment approach for both acute and chronic CO
       poisoning.
            A study of toll bridge authority workers investigated normal red cell adaptation to
10     anemia as a measure of CO effects on tollbooth collectors and maintenance personnel
       (Goldstein et al., 1975).  Diphosphoglycerides (DPG) increase the release of oxygen to tissues
       as an adaptive mechanism in anemia.   Results of the studies were inconclusive, in that they
       considered increased DPG to be a response to hypoxia from increased percent COHb.
       However, formation of methoxyhemoglobin from exposures to NO was independent  of the
15     COHb reaction, and the hypoxic effects of CO and NO exposures were considered to be
       additive.
            Use of percent COHb in blood has been proposed for use as a biological exposure index
       (BEI), as a supplement to the threshold limit value (TLV) value for CO exposure
       recommended by the American Conference of Governmental Industrial Hygienists (ACGIH)
20     (Lowry,  1986). The proposed BEI is  intended to be an index of exposure.  It is not
       necessarily an indication of an adverse response.
            Finally, Hickey et al. (1975) expressed the need to consider genetic and other factors
       resulting in differences in hemoglobin and other individual characteristics that could  influence
       the extent of COHb formation on exposures to CO in air or cigarette smoke.
25
       8.4.3 Occupational Exposures
            The number of persons potentially exposed to CO in the work environment is greater
       than that for any other physical or chemical agent (Hosey, 1970), with estimates as high as
       975,000  occupationally exposed at high levels (National Institute for  Occupational Safety and
30     Health, 1972).
        March 12, 1990                         8-50      DRAFT-DO NOT QUOTE OR CITE

-------
                  TABLE 8-10.  DIAGNOSTIC CRITERIA FOR CO INTOXICATION


                  Normal                                               Abnormal
 5
       Nonsmokers              Smokers                             Increased Surveillance

       COHb <3%             COHb <8%                          COHb 8 to 12%

10     SCN/blood <40mg/L      SCN/blood <200 mg/L
                                                                 Increased Risk

                                                                 COHb 12 to 15%

15                                                               Medical Treatment
                                                                 COHb > 15%


       Source: Castellino (1984).
20


            The contribution of occupational exposures can be separated from other sources of

       exposure, but there are at least two conditions to consider.

25-

            (1)   When CO concentrations at work are higher than the CO equilibrium
                  concentration associated with the percent COHb at the start of the work shift,
                  there will be a net absorption of CO and an increase in percent COHb.
                  Nonsmokers will show an increase that is greater than  that for smokers because
30                they start from a lower baseline COHb level. In some cases, nonsmokers may
                  show an increase, and smokers a decrease in percent COHb.

            (2)   When CO concentrations at work are lower than the equilibrium concentration
                  necessary to produce the worker's current level of COHb, then the percent COHb
35                will show a decrease. There will be a net loss of  CO at work.
            Occupational exposures can stem from two sources:  (1) through background
       concentrations of CO obtained by working in a densely populated area (as compared to the
40     residential environment), or (2) through work in industrial processes that produce CO as a

       product or by-product.  In addition, work in environments that tend to accumulate CO
       concentrations may result in occupational exposures.  Rosenman (1984) lists a number of
       occupations where the workers may be exposed to high CO concentrations.  This list includes
       acetylene workers, blast furnace workers,  coke oven workers, diesel engine operators, garage


       March 12, 1990                          8-51     DRAFT-DO NOT QUOTE OR CITE

-------
       mechanics, steel workers, metal oxide reducers, miners, mond process (nickel refining)
       workers, organic chemical synthesizers, petroleum refinery workers, pulp and paper workers,
       and water gas workers.  In addition, because methylene chloride is metabolized to CO in the
       body, aerosol packagers, anesthetic makers, bitumen makers, degreasers, fat extractors,
 5     flavoring makers, leather finish workers, oil processors, paint remover makers, resin makers,
       solvent workers, and stain removers also can have high COHb levels.
            Background sources are generally a result of combustion of organic materials.  With rich
       fuel mixtures, decreased amounts of O2 are available, and therefore production of CO as a
       product of incomplete combustion is favored.  There are numerous sources for CO
10     background exposures, and there is considerable variation and uncertainty in identifying the
       CO exposure resulting from specific sources.  Traffic patterns and emissions from mobile
       sources, as well as an overlay of emissions from stationary sources along with wind and
       weather conditions, make predictions difficult. (See Chapters 6 and 7 for a complete
       discussion of the mobile, stationary, and indoor sources and emissions of CO.)
15          Investigations and analyses of exhaust gas in Paris (Chovin, 1967)  showed that CO air
       concentrations were correlated with the activity of and  distribution pattern for traffic in Paris.
       The average CO concentrations for the years 1965 and 1966 were 16.0 and 16.6 ppm, based
       on 15,187 samples for 1965 and  15,203 samples for 1966,  respectively.  The maps of
       pollutant distribution indicate that the areas  of high and low concentrations were similar for
20     each year and were closely associated  with the volumes and patterns of vehicular traffic.
       When  the measured concentration of CO in the air exceeded 100 ppm, the sample was
       automatically diluted 10-fold for analysis, thereby introducing dilution and scale factors as
       possible sources of error at high  concentrations of CO. The variations in the measurements of
       CO were closely linked to variations in the volume of traffic at each sampling location.
25     Carbon Monoxide concentrations in the blood were determined for 331 traffic policemen
       during 5 hours of duty. Blood samples were collected at the beginning and end of the five-
       hour shifts.  Carbon monoxide in blood was determined by heating samples and measuring the
       evolved gas by an infrared method.  The values obtained were compared with the average
       concentrations of CO found for the air breathed.  The correlation was good between CO in
30     blood  and CO in the air breathed for nonsmokers, but with smokers, initial concentrations  of
       CO in blood were high, and there was often a decrease in the CO in blood over an exposure

       March 12, 1990                           8-52     DRAFT-DO NOT QUOTE OR CITE

-------
       period.  This was observed with sampling of smokers and nonsmokers at the same locations
       with similar concentrations and durations of exposure. Car drivers showed increases in CO
       concentrations in blood, as did traffic policemen who were nonsmokers. Carbon Monoxide
       concentrations in blood for smoking and nonsmoking drivers involved in traffic accidents
 5     were greater than those for traffic policemen and others in the population considered to be
       accidentally exposed to CO.
            Aircraft accidents involving 113 aircraft, 184 crew members, and 207 passengers were
       investigated to characterize accident toxicology and to aid in search for causation of a crash
       (Blackmore, 1974). Determinations of percent COHb in blood  samples obtained from victims
10     enabled differentiation of a variety of accident sequences involving fires.  For example,
       percent COHb determinations combined with passenger seating  information and crew
       assignments can allow differentiating between fire in flight or after the crash, survivability of
       crash with death due to smoke inhalation, specific equipment malfunctions in equipment
       operated by a particular crew member,  or defects in space heating in the crew cabin or
15     passenger compartment.  One accident in the series was associated with a defective space
       heater in the crew compartment.  Another accident also was suspicious  with regard to a space
       heater.
            Contributions to background CO concentrations from industrial processes may be
       determined by an inventory of sources and locations for the processes, as well as by emission
20     rates for CO as a function of production, and the air pollution distribution pattern for the
       region (see Chapter 6).  The types and distributions of industrial and community activities
       contributing to CO concentrations in air depend on identification of the various sources and
       volumes of production involved.  Production schedules are dynamic; it  is therefore difficult to
       model sources and predict levels.
25          Carbon monoxide concentrations measured in the air were used to classify workers from
       20 foundries into three groups: those with definite occupational exposure,  those with slight
       exposure, and controls (Hernberg et al., 1976).  Angina pectoris, EKG findings, and blood
       pressures of foundry workers were evaluated in terms of CO exposure for the 1000 workers
       who had the longest occupational exposures for the 20 foundries.  Angina showed  a clear dose
30     response with exposure to CO either from occupational sources  or from smoking, but there
       was no  such trend in EKG findings.  The systolic and diastolic pressures of CO-exposed

       March 12, 1990                          8-53      DRAFT-DO NOT QUOTE OR CITE

-------
       workers were higher than those for other workers, when age and smoking habits were
       considered.
            Carboxyhemoglobin and smoking habits were studied for a population of steel workers
       and compared to blast furnace workers, as well as to employees not exposed at work (Jones
 5     and Walters, 1962). Carbon monoxide is produced in coke ovens, blast furnaces, and in
       sintering operations.  Exhaust gases from these operations are often used for heating and as
       fuels for other processes. Fifty-seven volunteers working in the blast furnace area were
       studied for smoking habits, symptoms of CO exposure, and estimations of COHb levels by an
       expired air technique.  The main increase in COHb for blast furnace personnel was 2.0% for
10     both smokers and nonsmokers in the group. For smokers in the unexposed control group,
       there was a  decrease in percent COHb.  A follow-up study found similar results (Butt et al.,
       1974). Virtamo and Tossavainen (1976) report a study of CO measurements in air of 67
       iron, steel, or copper alloy foundries.  Blood COHb of ironworkers exceeded 6% in 26% of
       the nonsmokers and 71 % of the smokers studied.
15          Poulton (1987) found that a medical helicopter with engine running in a narrowed or
       enclosed helipad was found to be a source of potential exposure to CO, JP-4 fuel and possibly
       other combustion products for flight crews, medical personnel, bystanders and patients being
       evacuated. Measurements were made by means of a portable infrared analyzer.  Carbon
       monoxide concentrations were found to be greatest near the heated exhaust.  Concentrations
20     ranged from 8 ppm to 43 ppm.
            Exhaust from seven most commonly used chain  saws (Nilsson et al., 1987) were
       analyzed under laboratory conditions to characterize emissions.  The investigators conducted
       field studies on exposures of loggers using chain saws in felling operations, and also in
       limbing and bucking into lengths.  In response to an inquiry, 34% of the loggers responded
25     that they often experienced discomfort from the exhaust fumes of chainsaws, and another 50%
       complained of occasional problems. Sampling for CO exposures was carried out for five days
       during a two-week work period in a sparse pine stand at an average wind speed of 0 to
       3 m/s, a temperature range of 16 to 1°C, and a snow depth of 50  to 90 cm. Carbon
       monoxide concentrations ranged from 10 to 23 mg/m3 (9 to 20 ppm) with a mean value of
30     20.0 mg/m3.  Carbon monoxide concentrations measured under similar but snow free
       conditions ranged from 24 to 44 mg/m3 with a mean value of 34.0. In another study, CO

       March 12, 1990                          8-54     DRAFT-DO NOT QUOTE OR CITE

-------
        exposures were monitored for nonsmoking chain saw operators with average exposures
        recorded from 20 to 55 ppm with COHb levels ranging from 1.5 to 3.0% (Van Netten et al.,
        1987).
             Fork lift operators, stevedores and winch operators were monitored for CO in expired
 5      air to calculate percent COHb, using an MSA analyzer (Breysse and Bovee, 1969).  Periodic
        blood samples were collected to validate the calculations. Bull operators and stevedores work
        in the holds of ships; winch operators do not work in the holds. The ships to be evaluated
        were selected on  the basis of their use of gasoline-powered fork lifts for operations. To
        evaluate seasonal variations in percent COHb, analyses were performed for one five-day
 10      period per month for a full year.  Efforts were made to select a variety of ships for
        evaluation.  A total of 689 determinations of percent COHb were made from blood samples to
        compare with values from expired air samples. The samples were collected on 51  separate
        days involving 26 different ships. Two hundred men were available before work, while only
        147 were available at the end of the work day. Men lost to follow-up either left before the
 15      end of shift or were transferred to other work.  Smoking was found to be a major
        contributing factor to percent COHb levels found.  Carboxyhemoglobin values for
        nonsmokers indicated that the use of gasoline powered lifts in the holds of the  ships did not
        produce a CO concentration in excess of 50 ppm for up to eight hours as a TWA under the
        work rules and operating conditions in practice during the study.  Smoking behavior
20      confounded exposure evaluations. The exposure conditions may not provide the same degree
        of protection for smokers as they do for nonsmokers.
             Carbon monoxide concentrations have been measured in a variety of work places where
        potential exists for accumulation from outside sources. Exposure conditions in work places,
        however, are substantially different. There is no standard approach that applies in all
25      situations requiring evaluation and study.  The methods to be applied, group characteristics,
       jobs being performed, smoking habits, and physical characteristics of the  facilities themselves
        introduce considerable variety in the approaches used.  Typical studies are discussed below.
             Wallace  (1983) investigated CO in air and breath of employees working in an office
        constructed in an  underground parking garage at various times over a one-month period.
30      Carbon  monoxide levels were determined by use of a device containing a proprietary solid
        polymer electrolyte to detect electrons emitted in the oxidation of CO to CO2.  The  device

        March 12, 1990                          8-55     DRAFT-DO NOT QUOTE OR CITE

-------
       was certified by the Mine Safety and Health Administration (MSHA) to be accurate within
       15%. A data logger was attached to provide readings each second, and to provide one-hour
       averages from the CO monitors placed on desks.  Variation in CO measurements in ambient
       air showed a strong correlation with traffic activity in the parking garage. Initially, the office
 5     CO levels were found to be at an average of 18 ppm per day with the average from 12:00 to
       4:00 p.m. at 22 ppm and from 4:00 to 5:00 p.m. at 36 ppm.  Analyses of expired air
       collected from a group of 20 nonsmokers working in the office showed a strong correlation
       with ambient air concentrations for  CO and traffic activity. For example, the average CO in
       expired air for one series of measurements was 23.4 ppm, as compared to simultaneous
10     measurements of air concentrations  of CO at 22 to 26 ppm. After a weekend, CO
       concentrations in breath on Monday morning were substantially decreased (around 7 ppm) but
       rose again on Monday afternoon to  equal the air levels of 12 ppm.  Closing fire doors, and
       the use of existing garage fans decreased CO concentrations in the garage offices to 2 ppm or
       less, concentrations similar to those for other offices in the complex that were located away
15     from the garage area.
            Carboxyhemoglobin levels (Ramsey, 1967) were determined over a three-month period
       during winter months for 38 parking garage attendants, and the values for COHb were
       compared  with values from a group of 27 control subjects.  Blood samples were collected by
       finger stick on Monday mornings at the start of the work week, at the end of the work shift
20     on Mondays and at the end of the work  week on Friday afternoons.  Hourly analyses were
       carried out on three different weekdays using potassium palado sulfite  indicator tubes for the
       concentrations of CO at three of the six  garages in the study.  Hourly values ranged from 7 to
       240 ppm,  and the composite mean of the 18 daily averages was 58.9 + 24.9 ppm. While the
       Monday versus Friday afternoon values  for COHb were not significantly different, there were
25     significant differences between Monday morning and Monday afternoon values.  Smokers
       showed higher starting baseline values, but there was no apparent difference in net increase in
       COHb body burden between smokers and nonsmokers.  COHb values for nonsmokers ranged
       from a mean of 1.5 + 0.83% for the a.m. samples to 7.3 ± 3.46% for the p.m.  samples.
       For smokers these values were 2.9  + 1.88% for the a.m. and 9.3 + 3.16% for the p.m.
30     The authors  observed a crude correlation between daily average for CO in air and COHb
       values observed for a two-day sampling period.

       March 12, 1990                         8-56      DRAFT-DO NOT QUOTE OR CITE

-------
            In a study of motor vehicle examiners conducted by NIOSH (Stern et al., 1981) CO
       levels were recorded in six outdoor motor vehicle inspection stations with TWA levels of 4 to
       21 ppm. In contrast, the semi-open and enclosed stations had levels of 10 to 40 ppm TWA.
       The levels exceeded the recommended NIOSH standard of 35 ppm TWA on 10% of the days
 5     sampled. In addition, all stations experienced peak short-term levels above 200 ppm.
            Carboxyhemoglobin levels were measured for 22 employees of an automobile dealership
       during the winter months when garage doors were closed and ceiling exhaust fans were turned
       off (Andrecs et al., 1979).  Employees subjected to testing included garage mechanics,
       secretaries, and sales  personnel.  This included 17 males aged 21 to 37 and five females aged
10     19 to 36.  Blood samples were collected on a Monday morning before start of work, and on
       Friday at the end of the work week. Analysis for COHb was by addition of sodium dithionite
       and tris aminomethane, and COHb was measured in duplicate samples using a
       spectrophotometer. Smokers working in the garage area showed a Monday mean value for
       COHb of 4.87 + 3.64% and a Friday mean value of 12.9  + 0.83%.  Nonsmokers in the
15     garage showed a corresponding increase in COHb, with a Monday mean value of
       1.50 ± 1.37% and a  Friday afternoon mean value of 8.71  +2.95%.  Nonsmokers working
       in areas other than the garage had a Friday mean value of 2.38 + 2.32%, which was
       significantly lower than the mean values for smokers and nonsmokers in the garage area.
       Environmental concentrations or  breathing  zone samples for CO were not collected.  The
20     authors concluded that smokers have a higher baseline level of COHb than do nonsmokers,
       but both groups show similar increases in COHb during the work week while working in the
       garage area. The authors observed that the concentrations of COHb found in garage workers
       were at levels reported to produce neurologic impairment.  These results are consistent with
       those reported by Amendola and  Hanes (1984).  They reported some of the highest indoor
25     levels collected at automobile service stations and dealerships.  Concentrations ranged from
       16.2 to 110.8 ppm on cold weather to 2.2 to 21.6 ppm in warm weather.
            A group of 34 employees, 30 men and 4 women, working multi-story garages, were
       evaluated for exposures to exhaust fumes (Fristedt and Akesson, 1971).  Thirteen were
       service employees working at street level, and 21 were shop employees working either one
30     story above or one story below street level. Six facilities were  included in the study. Blood
       samples were collected on a Friday at four facilities, on Thursday and Friday at another, and

       March 12, 1990                          8-57     DRAFT-DO NOT QUOTE OR CITE

-------
       on a Thursday only at a sixth facility.  The blood samples were evaluated for red blood cell
       (RBC), and white blood cell (WBC) counts, COHb, lead and delta-ALA. Work histories,
       medical case histories, and smoking habits were recorded.  Among the employees evaluated,
       11 of 24 smokers and 3 of 10 nonsmokers complained of discomfort from exhaust fumes.
 5     Smokers complaining of discomfort averaged 6.6% COHb and nonsmokers complaining
       averaged 2.2% COHb. The corresponding values for non-complaining workers averaged
       4.2% and 1.1%, respectively.
           Air pollution by CO in underground garages was investigated as part of a larger study of
       traffic pollutants in Paris (Chovin, 1967). Work conducted between the hours of 8:00 a.m.
10     and 10:00 p.m. resulted in exposures in excess of 50 ppm and up to 75 ppm, on a TWA
       basis.
           As part of a larger study of CO concentrations and traffic patterns in Paris (Chovin,
       1967),  samples were taken in road tunnels.  There was good correlation between traffic
       volumes combined with the lengths of the tunnels and CO concentrations found. None of the
15     tunnels studied had mechanical ventilation.  The average CO concentrations in the tunnels
       were 27 and 30 ppm for 1965 and 1966, respectively, as compared  to an average of 24 ppm
       CO in the streets for both years.  The "real  average risk" for a man working or walking in a
       street or tunnel was considered by the authors to be 3 to 4  times less than the maximal risk
       indicated by values for CO from instantaneous air sample measurements. In the United
20     States,  Evans et al. (1988) studied bridge and tunnel workers in metropolitan New York City.
       The average COHb concentration over the 11 years of study averaged 1.73% for nonsmoking
       bridge  workers and 1.96% for tunnel workers.
            In a discussion of factors to consider in CO control of high altitude highway tunnels,
       Miranda et al. (1967) reviewed the histories of several tunnels.  Motor vehicles were
25     estimated to emit about 0.1 Ib of CO/mi at  sea level.  At 11,000 feet and a grade of 1.64%,
       emissions were estimated at 0.4 Ib/mi (for vehicles moving upgrade). Tunnels with
       ventilation are generally designed to control CO concentrations at or below 100 ppm. The
       Holland Tunnel in New York was reported  to average 65 ppm, with a recorded maximum of
       365 ppm due to a fire. For the Sumner Tunnel in Boston, ventilation is started at CO
30     concentrations of 100 ppm and additional fans are turned on with the sounding of an alarm  at
       250 ppm.  The average value for CO concentration is 50 ppm.  The Mont Blanc Tunnel is

       March 12,  1990                          8-58     DRAFT-DO NOT QUOTE OR CITE

-------
       7.2 mi long at an average elevation of 4179 ft. This tunnel is designed to maintain CO
       concentrations at or below 100 ppm. The Grand Saint Bernard tunnel is 3.5 mi long at an
       average elevation of 6000 feet. The tunnel is designed to maintain CO concentrations at or
       below 200 ppm. For the tunnel at 11,000 feet, the authors recommended maintaining CO
 5     concentrations at or below 25 ppm for long-term exposures, and no greater than 50 ppm for
       peaks of one-hour exposure.  The recommendations are based on considerations of a
       combination of hypoxia from lack of O2 due to the altitude and stress of CO exposures of
       workers and motorists.  The authors recommend  that warning signs and notices be posted to
       warn susceptible individuals to take another route.
10          Carbon monoxide exposures of tollbooth operators were studied along the New  Jersey
       Turnpike. The results reported by Heinold et al. (1987), indicated peak exposures for one-
       hour ranged from 12 to 24 ppm with peak eight-hour exposures of 6 to  15 ppm.
            Carboxyhemoglobin levels were determined for 15 nonsmokers at the start, middle, and
       end of a 40-day submarine patrol (Bondi et al., 1978).  Values found were 2.1 %, 1.7%, and
15     1.7%, respectively. The average ambient air concentration for CO was 7 ppm. The authors
       observed that the levels of percent COHb found would not cause  significant impairment of the
       submariners.
            In contrast, Iglewicz et. al. (1984) found in a 1981  study that CO  concentrations inside
       ambulances in New Jersey were often above the EPA eight-hour standard of 9  ppm.  For
20     example, measurements made at the head of the stretcher exceeded 9 ppm on nearly 27% of
       the 690 vehicles tested, with 4.2%  (29 vehicles) exceeding 35 ppm.
            Environmental tobacco smoke (ETS) has been reviewed (National  Research Council,
       1986) for contributions to air contaminants in airliner cabins, and to potential exposures for
       passengers and flight crew members. ETS is  described as a complex  mixture containing
25     many components. Analyses of CO content and paniculate matter in cabin air were used as
       surrogates for the vapor phases and solid components of ETS, respectively.  A mathematical
       model was developed and used to calculate the dilution of contaminants by outside make-up
       air.  Total emissions for CO in mainstream smoke range from 10,000 to 23,000 mg per
       cigarette.  More CO is emitted in sidestream  smoke; the ratio of sidestream smoke to
30     mainstream smoke ranges from 2.5:1 to 4.7:1. This ratio depends on the length of time a
       cigarette is held without active smoking compared to the total inhalation and smoking time.

       March 12, 1990                         8-59     DRAFT-DO NOT QUOTE OR CITE

-------
       The amount of CO in the cabin environment depends on the rate and number of cigarettes
       smoked, and on the rate of dilution by outside make-up air. An additional factor to consider
       is the influence of pressure altitude on the absorption of CO and other gases.  The legal limit
       for pressure altitude is 8000 feet.  The partial pressure of O2 is 120 mm Hg assuming
 5     20% O2 in the cabin air, compared to 152 mmHg at sea level.  It is possible that the
       absorption rate for CO would be increased under hypobaric conditions.
            An examination of CO hazard in city traffic for policemen in three Swedish towns
       (Gome et al., 1969) showed that the increases observed in the percent COHb in blood for a
       group of 28 policemen were associated with exposures to exhaust fumes from heavy traffic.
10     Conversely, results from studies of 28 traffic policemen who were smokers and had relatively
       high percent COHb in blood at the beginning of a work period either showed no change, or
       showed a decrease in percent COHb while exposed to exhaust fumes in directing traffic.
       Exposures were higher in a larger, more congested city, as compared to two smaller cities in
       the study.
15          Carbon monoxide levels in city driving in Los Angeles were measured using a prototype
       CO measuring device mounted in the passenger seat (Haagen-Smit, 1966).  Carbon monoxide
       concentrations were continuously monitored and were sampled by means of a glass tube
       projecting through the window.  Typical commuting trips were made throughout the
       downtown Los Angeles area during commuting hours. The distance traveled was  about
20     30 mi. The shortest time was 40 min and the longest time was one-hour and 55 min.
       Concentrations of CO averaged 37 ppm for the best trips,  with an average of 54 ppm in
       heavy traffic moving at 20 mph; peak CO concentrations reached as high as 120 ppm.
            A study of municipal bus drivers in the San Francisco Bay area by Quinlan et al.  (1985)
       showed a TWA of 1 to 23 ppm, with mean TWA = 5.5 ppm and standard deviation of
25     4.9 ppm.  The peak exposures ranged from 7 to 47 ppm with mean 25.3 ppm and standard
       deviation of 12.5 ppm.
            Cooke (1986) reports finding no significant increases outside normal ranges, as
       compared to the general population, for levels of blood lead and COHb in a group of
       13 roadside workers. Samples were collected in the afternoon of a workday. Among the
30     subjects, 7 to 13  were smokers and showed percent COHb in blood ranging from  3.0 to 8.8
       (mean of 5.5%). Each nonsmoker percent COHb ranged  from 0.5 to  1.4 (mean of 1.2%).

       March 12, 1990                          8-60     DRAFT-DO NOT QUOTE OR CITE

-------
       Each smoker had smoked at least one cigarette in the four hours preceding collection of blood
       samples. No samples were collected before start of work, and no measurements of CO in air
       at the work sites were presented.
       8.5 BIOLOGICAL MONITORING
            A unique feature of carbon monoxide exposure is that there is a biological marker of the
       dose that the individual has received: the blood level of CO. This level may be calculated by
       measuring blood COHb or by measuring CO in exhaled breath.
10
       8.5.1  Blood Carboxyhemoglobin Measurement
            Carbon monoxide in the inspired air is rapidly transferred to the blood in the alveoli at a
       rate that is dependent upon  several physiological variables.  The blood level of CO is
       conventionally represented as a percentage of the total Hb available (i.e., the percentage of
15     Hb that is in the form of percent COHb or simply COHb).  The high affinity of CO for Hb
       has the effect of retaining the bulk (90 to 95%) of the absorbed gas in the vascular space and
       at the same time amplifying the exposure. This latter phenomenon occurs because the affinity
       of Hb for CO is 200 to 250 times that for O2 (Douglas et al., 1912) resulting in a relatively
       high COHb level at very low partial pressures of CO in the alveolar gas phase.  The primary
20     physical and physiological variables that determine the relationship between ambient exposure
       and blood levels of CO are  presented in detail in  Chapter 9.

       8.5.1.1 Measurement Methods
            Any technique for the measurement of CO in blood must  be specific for CO and have
25     sufficient sensitivity and accuracy for the purpose of the values obtained. The majority of
       technical methods that have been published on measurement of CO in blood have been for
       forensic purposes. These methods are less accurate than generally required for the
       measurement of low levels of COHb (<5% COHb).  Blood levels of CO resulting from
       exposure to existing NAAQS levels of CO would not be expected to exceed 5% COHb in
30     nonsmoking subjects.  The focus of the forensic methods  has been the reliability of
       measurements over the entire range of possible values: from less than 1 % to 100% COHb.

       March 12, 1990                         8-61      DRAFT-DO NOT QUOTE OR CITE

-------
       These forensically oriented methods are adequate for the intended use of the values and the
       nonideal storage conditions of the samples being analyzed.
            In the areas of exposure assessment and low level health effects of CO, it is more
       important to know the accuracy of any method in the low level range of 0 to 5% COHb.
 5     There is little agreement upon acceptable reference methods in this range  nor are there
       accurate reference standards available in this range.  The use of techniques that have
       unsubstantiated accuracy in the low range of COHb levels can lead to considerable differences
       in estimations of exposure conditions.  Measurement of low levels of CO in blood demands
       careful evaluation because of the implications based upon this data for the setting of air
10     quality standards.  Therefore this section will focus on the methods that have been evaluated
       at levels below 10% COHb and methods that have been extensively used  in assessing
       exposure to CO.
            The measurement of CO in blood can be  accomplished by a variety of techniques that
       have been divided into destructive  and nondestructive methods (U.S.  Environmental
15     Protection Agency,  1979). Carboxyhemoglobin can be determined nondestructively by
       observing  the change in the absorption spectrum in either the Soret or visible region brought
       about the combination of CO with  Hb. With present optical sensing techniques, however, all
       optical methods are limited in sensitivity to approximately 1% of the range of expected
       values.  If attempts are made to expand the lower range of absorbances, sensitivity is lost on
20     the upper end where, in the case of COHb, total Hb is measured. For example, in the
       spectrophotometric method described by  Small et al. (1971), a change in  absorbance equal to
       the limits of resolution of 0.01 units can result in a difference in 0.6%  COHb. Therefore,
       optical techniques can not be expected to obtain the resolution that is possible with other
       means of detection of CO (Table 8-11).  The more sensitive (higher resolution) techniques
25     require the release of the CO from the Hb into a gas phase that can be  detected directly by:
       (1) infrared absorption (Coburn et al., 1964; Maas et al., 1970)  following separation using
       gas chromatography; (2) the difference in thermal conductivity between CO and the carrier
       gas (Allred et al., 1989; Ayres et al., 1966; Dahms and Horvath, 1974; Goldbaum et al.,
       1986; Horvath et al., 1988; McCredie and Jose, 1967); (3) the amount of ionization
30     following quantitative conversion of CO to CH4 and ionization of the CH4 (Clerbaux et al.,
       1984; Collison et al., 1968;  Constantino et al., 1986;  Dennis and Valeri, 1980; Guillot et al.,

       March 12, 1990                          8-62      DRAFT-DO NOT QUOTE OR CITE

-------
10
15
20
25
30
          TABLE 8-11.  COMPARISON OF REPRESENTATIVE METHODS FOR ANALYSIS
                               OF CARBON MONOXIDE IN BLOOD
Source
Gasometric Detection
Scholander and Roughton
(1943)
Horvath and Roughton
(1942)
Spectrophotometric Detection
Coburn et al.
(1964)
Small et al.
(1971)
Maas et al.
(1970)
Brown (1980)

Gas Chromatographv
Ayres et al.
(1966)
Goldbaum et al.
(1986)
McCredie and Jose
(1967)
Dahms and Horvath
(1974)
Collison et al.
(1968)
Kane (1985)

Vreman et al. (1984)
Method

Syringe-
capillary
Van Slyke


Infared

Spectro-
photometry
CO-Oximeter
(IL 182)
CO-Oximeter
(IL-282)

Thermal
conductivity
Thermal
conductivity
Thermal
conductivity
Thermal
conductivity
Flame
ionization
Flame
ionization
Mercury vapor
Resolution*
ml/dl

0.02

0.03


0.006

0.12

0.21

0.2


0.001

ND

0.005

0.006

0.002

ND

0.002
CV%b

2 to 4%

6%


1.8%

ND

5%

5%


2.0%

1.35%

1.8%

1.7%

1.8%

6.2%

2.2%
Reference Method

Van Slyke

Van Slyke-Neill


Van Slyke-Syringe

Flame Ionization

Spectrophotometric

Flame Ionization


ND

Flame Ionization

ND

Van Slyke

Van Slyke

CO-Oximeter

ND
r"

ND"

ND


ND

ND

ND

0.999


ND

0.996

ND

0.983

ND

1.00

ND
35
40
45
       'The resolution is the smallest detectable amount of CO or the smallest detectable difference between
        samples.

50     ""Coefficient of variation was computed on samples containing less than 15 % COHb, where possible.
55
The r value is the correlation coefficient between the technique reported and the reference method used to
 verify its accuracy.

dlndicates no data were available.
       March 12, 1990
                                        8-63     DRAFT-DO NOT QUOTE OR CITE

-------
       1981; Kane, 1985; Katsumata et al., 1985); or (4) the release of Hg vapor due to the
       combination of CO with mercuric oxide (Vreman et al., 1984).

       Sample Handling
 5          Carbon monoxide bound to Hb is a relatively stable compound which can be dissociated
       by exposure to O2 or ultraviolet (UV) radiation (Chace et al.,  1986; Horvath and Roughton,
       1942). Ultraviolet light has not been shown by Goldstein et al.  (1986) to affect COHb levels
       in glass vials exposed to room lighting conditions, however the ability of UV light to
       penetrate these tubes was not demonstrated. If the blood sample is maintained in  the dark
10     under cool, sterile conditions, the CO content will remain stable for a long period of time.
       Various investigators have reported no loss of percent COHb over 10 days (Collison et al.,
       1968), three weeks (Dahms and Horvath, 1974), four months (Ocak et al., 1985) and six
       months (Vreman et al., 1984).  The anticoagulant system used appears to influence the CO
       level since some EDTA vacutainer tube stoppers contain CO (Vreman et al.,  1984).  The
15     increased levels of COHb due to this amount of CO have been demonstrated (Goldstein et al.,
       1986; Vreman et al.,  1984).  The stability of the CO content in  properly stored samples does
       not indicate that constant values will be obtained by all techniques of analysis. The
       spectrophotometric methods are particularly susceptible to changes in optical qualities of the
       sample which results in small changes in COHb with storage (Allred et al., 1989).
20          Carboxyhemoglobin values obtained with the IL 282 CO-Oximeter have been shown to
       decrease over the first three days following  collection (Allred et al.,  1989; Goldstein et al.,
       1986). This decrease occurs within the first 24 h (Allred et al., 1989) and does not fall
       further over the next 14 days (Goldstein et al., 1986).  Storage of blood samples can result in
       the formation of methemoglobin (Goldstein et al., 1986) and under some conditions
25     sulfhemoglobin (Rai and Minty, 1987).  Both species of Hb can result in the optical methods
       of COHb detection being incorrect depending upon the specific wavelengths utilized.
            Therefore the care needed to make a COHb determination  depends upon the technique
       that is being utilized.   It appears as though measurement of low  levels of COHb with  optical
       techniques should be conducted out as soon as possible following collection of the samples.
30
        March 12, 1990                          8-64     DRAFT-DO NOT QUOTE OR CITE

-------
        Potential Reference Methods
             Exposure to CO at equilibrium conditions results in COHb levels of between 0.1 and
        0.2% COHb for each part per million of CO in the atmosphere.  A reference technique for
        the measurement of COHb should be able to discriminate between two blood samples with a
 5      difference of 0.1 % COHb (approximately 0.02 mL/dL).  To accomplish this task the
        coefficient of variation (standard deviation of repeated measures on any given sample divided
        by the mean of the values times 100) of the method should be less than 5% so that the two
        values that are 0.1 %  COHb different can be statistically proven to be distinct. In practical
        terms a reference method should have the sensitivity to detect approximately 0.025% COHb
10      to provide this level of confidence in the values obtained.
            The accurate measurement of CO in a blood sample requires the quantitation of the
        content of CO in blood.  Optically based techniques have limitations of resolution and
        specificity due to the  potential interference from many sources.  The techniques that can be
        used as reference methods involve the quantitative release of CO from the Hb followed by the
15      measurement of the amount of CO released.  Classically this quantitation was measured
        manometrically with a Van Slyke apparatus (Horvath and Roughton, 1942) or a Roughton-
        Scholander syringe (Roughton and Root, 1945).  These techniques have served as the "Gold
        Standard" in this field for almost 50 years.  However, there are limitations of resolution with
        these techniques at the lower ranges of COHb.  The gasometric standard methodology has
20      been replaced with head-space extraction followed by the use of solid phase gas
        chromatographic separation with several different types of detection:  thermal conductivity,
        flame ionization, and mercury vapor reduction.  With the use of National Institute of
        Standards and Technology (NIST) standard gas mixtures of CO, the gas chromatographic
        techniques can be standardized when proper consideration is given to potential sources of  loss
25      of standard.  The CO in  the headspace can also be quantitated by infrared detection which can
        be calibrated with gas standards.

            Flame Ionization Detection.  This technique requires the separation of CO from the
        other headspace gases and the  reduction of the CO to CH4 by catalytic reduction.  Collison
30      et al. (1968) reported that the results from their method correlated with the Van Slyke
        gasometric method at high levels of CO (8 to 13 mL/dL) where the error in the gasometric

        March 12, 1990                          8-65      DRAFT-DO NOT QUOTE OR CITE

-------
       method was minimal.  The values obtained from the two independent techniques were highly
       significantly correlated (p< 0.0001) with a linear regression r = 0.992. The limit of
       detection was reported to be 0.01% COHb using 100 pL of blood.  The coefficient of
       variation was 1.08% on a sample containing approximately 50% COHb and 1.80% on a
 5     sample containing approximately 0.8% COHb. The basic technique of Collison et al. (1968)
       using headspace analysis and flame ionization  detection is the most sensitive method that has
       been compared with the gasometric methods.  Modifications of this method have been widely
       used by other investigators for evaluating technically simpler methods of CO analysis
       (Clerbaux et al., 1984; Collison et al., 1968; Dennis and Valeri,  1980; Guillot et al., 1981;
10     Kane,  1985; Katsumata et al., 1985). This method conforms to all the requirements of a
       reference method.

            Thermal  Conductivity Detection. Ayres et al. (1966) reported a method for using
       vacuum extraction of CO from blood in a Van Slyke apparatus for gas chromatographic
15     separation with thermal conductivity analysis of the CO.  The gas phase of the reaction
       chamber was eluted onto a 5A molecular-sieve column for separation of the components.
       This technique was reported to have a lower limit of detectability of 0.001 mL/dL or
       approximately  0.005% COHb.  The coefficient of variation was reported to be 1.95% on a
       sample of unspecified percent COHb.  The analysis system was calibrated  using a gas sample
20     of known  CO content injected directly onto the column.  No comparisons were performed
       with other standard techniques.  McCredie and Jose (1967) also reported results from
       chromatographic separation of vacuum extracted gas.  Thermal conductivity detection enabled
       the limit of detection to be 0.005 mL/dL or approximately 0.025% COHb.  This system was
       also calibrated with standard gas mixtures injected directly onto the column. A coefficient of
25     variation was not presented but a standard deviation of 0.004 mL/dL on a  series of repeat
       analyses on an average blood sample indicates that this method is sufficiently reproducible.
       This would represent a coefficient of variation on the blood CO content measured from  the
       average nonsmoker of 2.5 %. Dahms and Horvath (1974) described  a technique of headspace
       analysis of CO using thermal conductivity detection. The CO was released from the blood
30     while the mixture was  stirred to produce a vortex,  using Van Slyke reagents in a sealed
       reaction vial.  The extraction occurred into the headspace of a sealed vial pressurized to the

       March 12, 1990                         8-66      DRAFT-DO NOT QUOTE OR CITE

-------
        head pressure on the column.  The limit of detection with this technique was 0.006 mL/dL or
        0.03% COHb with a coefficient of variation of 1.7% on a sample containing 6.5% COHb.
        This method used standard gases injected into the reaction vial to calibrate the system.  The
        results of this technique were compared to the standard Van Slyke method (Horvath and
 5      Roughton,  1942) over the whole range of values and more specifically on 90 blood samples
        containing less than 10% COHb.  The correlation coefficient between the gas chromatography
        and the Van Slyke method was 0.984. Linear regression analysis demonstrated essentially a
        zero intercept (0.009 mL/dL) between the two techniques.  This close agreement between
        values obtained with these independent methods provides a basis for the use of standard gases
10      to calibrate gas chromatographic techniques.  All of the above mentioned gas chromatographic
        methods for determination of CO in blood are acceptable as reference methods.

            Infrared Detection.  Coburn et al. (1964) described a method for extracting CO from
        blood under normal atmospheric conditions and then injecting the headspace gas into an
15      infrared analyzer. This technique has a reported limit of detectability of 0.007 mL/dL or
        0.035% COHb.  The coefficient of variation was 1.8% on an average COHb of 1.67%. The
        results of this technique were compared with the gasometric technique of Roughton and Root
        (1945) on five samples; there was no difference between the two techniques.  This method is
        acceptable as  a reference method for the measurement of CO in blood.
20
            Hemoglobin Measurement. The conventional means of representing the quantity of CO
        in a blood sample is the percent COHb:  the percentage of the total CO combining capacity
        that is in the form of COHb.  This is conventionally determined by the use of the following
        formula:
25
                          %COHb = [CO content/(hemoglobin x 1.389)] X 100          (8-11)
       March 12, 1990                          8-67     DRAFT-DO NOT QUOTE OR CITE

-------
       where:
            CO content =  cc/dL blood STPD,
            hemoglobin =  g/dL blood, and
                1.389  =  the stoichiometric combining capacity of hemoglobin for CO in units of
                          mL/g STPD.
10     The analytical methods that quantify the CO content in blood require the conversion of these
       quantities to percent COHb.  The product of the Hb and the theoretical combining capacity
       (1.389 according to International Committee for Standardization in Haemotology, 1978)
       yields the CO capacity. With the use of capacity and the measured content, the percent of
       CO capacity (percent COHb) is calculated.  To be absolutely certain of the accuracy of the
15     Hb measurement, the theoretical value should be routinely substantiated by direct
       measurement (internal validation) of the Hb CO combining capacity.  The total CO combining
       capacity should be determined as accurately as the content of CO. The error of the
       techniques that measure CO content are dependent on the error in Hb  analysis for the final
       form of the data, percent COHb.  Therefore the actual CO combining capacity should be
20     measured and compared with the calculated value based upon the reference method for Hb
       measurement. The measurement of CO combining  capacity can be routinely carried out by
       equilibration of a blood sample with CO (Allred et al., 1989).
            The standard methods for Hb determination involve the conversion of all species of Hb
       to cyanmethemoglobin with the use of a mixture of potassium ferricyanide, potassium
25     cyanide, and sodium bicarbonate.  Three combinations of similar reagents have been routinely
       used for the quantification  of Hb.  Drabkin's solution contains 0.6 mM K4Fe(CN)6, 0.8 mM
       KCN,  and 12 mM NaHCO3 (Drabkin and Austin, 1932).  Van Kampen and Zijlstra (1961)
       substituted 0.7 mM K2HPO4 for the bicarbonate in the reagent mixture.  A third reagent for
       producing cyanmethemoglobin is that of Taylor and Miller (1965) who increased the
30     concentration of potassium ferricyanide to 3 mM in Van Kampen and Zijlstra's mixture to
       decrease the reaction time with COHb.  The presence of high levels of COHb slows the rate
       of conversion to cyanmethemoglobin so that the use of the conventional Drabkin's reagent
       requires a reaction time of at least 180 min (Allred  et al., 1989; Kane, 1985) as opposed to
       the recommended time of 20 to 30 min. This increased reaction time is essential for the
       March 12,  1990                          8-68     DRAFT-DO NOT QUOTE OR CITE

-------
       accurate comparison of cyanmethemoglobin values with CO combining capacity
       measurements.  If the reaction is not permitted to go to completion the spectrophotometric
       method will underestimate the amount of Hb present in the sample.

 5     Other Methods of Measurement
            There are a wide variety of techniques that have been described for the analysis of CO
       in blood.  These methods have been reviewed  previously (U.S. Environmental Protection
       Agency,  1979) and include UV-visible spectrophotometry (Brown, 1980; Small et al.,  1971;
       Zwart et al., 1984; 1986), magnetic circular dichroism spectroscopy (Wigfield et al., 1981),
10     photochemistry (Sawicki and Gibson,  1979), gasometric methods (Horvath and Roughton,
       1942; Roughton and Root, 1945), and a calorimetric method (Sjostrand, 1948).  Not all of
       these methods have been as well characterized for the measurement of low levels of COHb as
       those listed above as potential reference methods.

15          Spectrophotometric Methods. The majority of the techniques are based upon optical
       detection of COHb which is more rapid than the reference techniques because these methods
       do not involve extraction of the CO from the blood sample.  These direct measurements also
       enable the simultaneous measurement of several  species of hemoglobin including  reduced Hb,
       O2Hb,  and COHb.  The limitations of the spectrophotometric techniques have been reviewed
20     by Kane  (1985).  The optical methods utilizing ultraviolet wave lengths require dilution of the
       blood sample which can lead to the loss of CO due to the competition with the dissolved O2in
       solution.  Removing the dissolved O2 with dithionite can lead to the formation of
       sulfhemoglobin which interferes with the measurement of COHb (Rai and Minty, 1987).
       Another limitation is that the absorption maxima (and spectral curves) are not precisely
25     consistent between individuals.  This may be due to slight variations in types of hemoglobin
       in subjects. For these reasons the techniques using fixed wave length measurement points
       would  not be expected to be as precise, accurate, or specific  as the proposed reference
       methods  mentioned above.
            Rodkey et al. (1979) reported a modification of the spectrophotometric technique for
30     measuring COHb.  This method converts all the hemoglobin  species in a blood sample to
       either COHb or Hb by the quantitative addition of the reducing agent sodium hydrosulfite.

       March 12, 1990                          8-69      DRAFT-DO NOT QUOTE OR CITE

-------
       The absorbance at 420 nm was used for the determination of COHb and 432 nm for Hb.  The
       optically based values were compared with those obtained by gas chromatography on the same
       28 samples. Twenty-five of these values were below 6% COHb and linear regression analysis
       demonstrated a slope of 1.038 with an intercept of -0.154 and an r = 0.994. The number of
 5     samples studied was relatively small and no error term was presented for the relationship.
       Visual inspection of the data, however, indicates a wide scatter of optical values for any given
       gas chromatograph value when the levels were at or below 1 % COHb (normal range of values
       for unexposed, nonsmoking individuals).
            A multicomponent spectrophotometric technique for the measurement of hemoglobin
10     derivatives was reported by Zwart et al. (1984).  This technique employs a multiwavelength
       spectrophotometer that uses reversed optics to enable the rapid collection of the absorbance
       spectrum from an array of photomultiplier tubes that detect transmission of light at intervals
       of 2 nm.  This method offers the possibility of instantaneous absorption data over the entire
       spectrum rather than the collection of data at a few selected wavelengths.  This optical system
15     offers the potential for correcting for individual  variability in absorption characteristics of
       hemoglobin.  The COHb data produced with this technique has not been compared with any
       of the proposed reference methods but has been compared with that obtained from the
       mercury vapor detector. The correlation coefficient of the optical data  with the gas
       chromatographic-mercury vapor technique was only 0.87; linear regression analysis resulted
20     in the following relationship:  GC = 0.65 (MCA) + 0.24 (Vreman et al., 1987).

            Mercury Vapor Detection. The most sensitive detector for the measurement of CO is the
       UV-photometer that senses mercury vapor produced by the reaction of CO with HgO (Trace
       Analytical). This unit has the reported ability to resolve 1 ppb. The use of such a sensitive
25     detector for blood determinations requires that measurements be carried out on only 1 to
       10 n\ quantities of blood.  Vreman et al. (1984) reported the use of this detector following
       gas chromatographic separation of the CO from other gases in blood.  Mercuric oxide will
       react with other gases so the chromatographic separation is an essential step in the use of this
       detector.  Values for COHb obtained with this technique have not been compared those
30     obtained with any of the proposed reference methods.  The COHb analysis method of Vreman
       et al.  (1984) was used in a parallel with a gas chromatography method  using thermal

       March 12, 1990                          8-70      DRAFT-DO NOT QUOTE OR CITE

-------
       conductivity detection for the routine measurement of COHb in a series of samples from
       subjects exposed to CO to produce levels of COHb up to 6% (Allred et al., 1989). Paired
       data from samples analyzed by both techniques were obtained on 108 samples. The values
       were significantly correlated r = 0.987; however, the reduction gas analyzer results were not
 5     corrected to STPD conditions, so an absolute comparison was not possible. This technique
       needs further validation by comparison with other methods to assure that the levels measured
       are accurate.

       CO-Oximeter Measurements of Carboxyhemoglobin
10          The speed of measurement and relative accuracy of spectrophotometric measurements
       over the entire range of expected values led to the development of CO-Oximeters.  These
       instruments utilize from two to seven wavelengths in the visible region for the determination
       of proportions  of oxyhemoglobin, carboxyhemoglobin,  reduced hemoglobin and
       methemoglobin.  The proportion of each species of Hb  is determined from the absorbance and
15     molar extinction coefficients at present wavelengths.  All of the commercially available
       instruments provide  rapid results for all the species of Hb being measured.  In general, the
       manufacturers' listed limit of accuracy for  COHb for all of the instruments is  1% COHb.
       However, this  level  of accuracy is not suitable for measurements associated with background
       CO levels (<2% COHb) because it corresponds to errors exceeding 50%. The precision of
20     measurement for these instruments is excellent and has  misled users regarding the accuracy of
       the instrument.  The relatively modest level of accuracy is adequate for the design purposes of
       the instrument; however, at low  levels of COHb the ability of the instrument to measure the
       percent COHb accurately is limited.
            The commercially available instruments for the measurement of COHb all utilize the
25     same basic principles:  hemolysis, constant temperature, and the measurement of absorbance
       at several wave lengths.  These instruments have been designed to provide information
       regarding COHb measurement that is ±1% COHb.   However, these instruments are all very
       precise so that  the coefficient of error between repeat measurements (standard deviation of
       repetitions/mean of the repetitions) is very  low.  Unfortunately, very few studies  have
30     evaluated  the accuracy of the measurements made with  these instruments as a routine aspect
       of quality control. The concern  regarding  the accuracy of any optical measurement on a

       March 12, 1990                         8-71      DRAFT-DO NOT QUOTE OR CITE

-------
       diluted blood sample should be of greatest concern due to the wide variety of substances that
       can subtly alter the absorption spectrum of Hb and the optical quality of the blood sample
       itself.  Because of the widespread use of these instruments, the evaluations of this instrument
       will be carefully reviewed.  These instruments consist of the Instrumentation Laboratories
 5     CO-Oximeters known as the IL 182 and the IL 282, the Radiometer Oximeter OSM-3, and
       the Coming Oximeter 2500.  There is very little information regarding the accuracy of the
       OSM-3 in the low range of COHb values compared to the data obtained from paired analysis
       with a reference method.
            The instruments that have been used to the greatest extent in the health effects of CO
10     have been the IL 182 and IL 282 (Instrumentation Laboratory, Inc., Lexington, MA).  The
       IL 282 instrument uses absorbances at four wavelengths in the visible region and a matrix of
       molar extinction coefficients to calculate each species of hemoglobin.  This  method is
       susceptible to interference from high concentrations of methemoglobin and sulfhemoglobin.
       The IL 282 CO-Oximeter has been shown to provide accurate data when the range of 0 to
15     100% COHb is considered (Brown, 1980).  However, comparison of the  results from this
       method with the proposed reference methods indicates that at low levels of COHb the results
       from this instrument are not sufficiently accurate to warrant their use alone  for low level
       COHb investigation. Resting levels of percent COHb have been shown to be below 0.9% for
       non-smokers by all the proposed reference methods (Ayres et al., 1966; Coburn et al., 1964;
20     Collison et al., 1968; Dahms and Horvath,  1974; McCredie and Jose, 1967).  The limit of
       accuracy for the IL 282 CO-oximeter for percent COHb is 1%, which has raised concern over
       the capability of all CO-Oximeters in the low range of COHb levels.  Therefore the accuracy
       of these instruments has been determined by paired observations on blood samples with the
       CO-Oximeter and a reference method.  The results are shown in Table 8-12 below.
25          The results from the linear regression analyses of all these comparisons indicate that
       there is considerable difference between instruments of the same model type.  The slope of
       the relationship between the optical methods are sufficiently close to  unity that there is no
       difference between instruments in the linearity of the  measurements.   Confidence intervals for
       the regressions are not given so this comparison can not be made.  The intercept values vary
30     widely relative to the purpose of accurately measuring low levels of COHb. These
       differences probably reflect the difference between  instruments.  In order to use these

       March 12, 1990                          8-72      DRAFT-DO NOT QUOTE OR CITE

-------
        TABLE 8-12. EVALUAITON OF THE ABILITY OF CO-OXIMETERS TO MEASURE LOW LEVELS
                    OF COHb AS COMPARED TO PROPOSED REFERENCE METHODS
t— »
i





oo

o
£
>
HH
3
6
o
1
H
O
CJ
B
a
o
5*
n
^^

Instrument
IL 182
IL 182
IL182
IL282
IL282
IL282
IL282
IL282




Corning
2500


Corning
2500


"Abbreviations:
Reference
Method"
GC-FID
GC-FID
Infrared
GC-FID
GC-FID
GC-TCD
GC-FID
CG-TCD




GC-FID



GC-FID



GC-TCD is gas c

Slope
0.690*GC
1.049*GC
0.977*IR
0.990*GC
1.122*GC
0.919*GC
0.895*GC
1.0069*GC
1.05*GC
1.05*GC
1.05*GC

0.92*GC



1.013*GC



:hromatography thermal

Intercept
+3.59
-0.54
+3.33
+0.45
-0.907
-0.068
+0.66
-0.01
+0.79
+0.55
+0.47

+ 1.17



-1.279



conductivity

R
0.59
NDb
ND
0.997
0.993
0.961
0.856
0.99
0.99



0.979



0989



r detection; GC-FID

n
16
275
12
39
13
20
16
ND
203
192
162

50



286



is gas chrc
COHb
Range Reference
< 15% Constantino et al. (1986)
<15% Guillot et al. (1981)
< 100% Maas et al. (1970)
< 100% Dennis and Valeri (1980)
< 17% Dennis and Valeri (1980)
< 8 % Goldbaum et al. (1986)
< 15% Constantino et al. (1986)
<9% Horvath et al. (1988)
<6% Allred et al. (1989)
<6% Allred et al. (1989)
<6% Allred et al. (1989)

<20% Kane (1985)



< 15% Tikuisis et al. (1987)



imatography with flame ionization detection.
blndicates no data were available.

-------
       instruments for the measurement of low levels of COHb they must be individually and
       routinely calibrated with a reference method.  The linearity of the response of the instruments
       implies that a standard correction can be applied to the value for COHb with the result that
       the average value of COHb obtained with these instruments will be correct.
 5          The interaction of Hb species with the measurement of COHb below 10% has been
       evaluated by Allred et al. (1989).  In freshly drawn blood samples, levels of COHb were
       maintained constant, as measured by GC, while levels of metHb, total Hb and O2Hb were
       varied.  Only the level of O2Hb interacted significantly with the COHb value.  In a series of
       46 subjects the effect of O2Hb was measured to determine its role in routine measurements of
10     COHb.  The effect of O2Hb varied considerably between individuals, with the average change
       being approximately 0.1% COHb  for every 10% change in O2Hb.  Almost all COHb
       measurements were made on venous blood which can vary considerably in O2Hb
       concentration and consequently affect the measurement of low levels of COHb.
            Hydrogen ion concentration was shown to have an effect on the measured percent COHb
15     in blood  stored in acid citrate dextrose (ACD) solution for two days.  However the effect of
       pH on percent COHb in freshly drawn samples has not been clearly demonstrated (Allred
       et al., 1989). Hydrogen ion concentration has been  demonstrated to change the absorbance
       spectrum of oxyhemoglobin and therefore may be expected to have an effect on the ability of
       CO-Oximeters to measure COHb.  Plasma lipid,  triglyceride and cholesterol levels were
20     found to not have any effect on the ability of the IL282 CO-Oximeter to measure COHb as
       determined by the difference between in instrument value and the reference value obtained by
       gas chromatography.
            The content of CO in blood stored in a tightly  capped syringe at 4°C in the dark has
       been shown to remain stable for up to four months.  Measurement of COHb by IL 282 CO-
25     Oximeter on blood samples (COHb range of 4.3 to 1.3%) within 15 min of collection,
       followed by storage  for 24 and 48 h as described above, resulted in a decrease in the detected
       percent COHb.  The apparent loss of COHb occurred in the first 24 h and averaged 16%
       (Allred et al., 1989). There was no change in percent COHb as determined by GS.  It is not
       clear when in the first 24-h period this change occurred.
30          The use of CO-Oximeters to measure low levels of COHb can provide useful
       information regarding mean values, provided  a reference technique is used to properly

       March 12, 1990                          8-74      DRAFT-DO NOT QUOTE OR CITE

-------
        calibrate the instrument.  It has been shown, however, that the range of values obtained with
        the optical method will be greater than that obtained with a reference method.  In a group of
        subjects with cardiovascular disease the standard deviation of the percent COHb values for
        non-smoking, resting subjects was 2 to 2.5 times greater for the CO-Oximeter values than for
 5      the GC values on paired samples (Allred et al., 1989).  Therefore, the potential exists with
        the CO-Oximeter for having an incorrect absolute value for COHb as well as an incorrectly
        broadened range of values.
             In addition, it is not clear exactly how sensitive the CO-Oximeter techniques are to
        small changes in COHb at the low end of the CO dissociation curve. Allred et al. (1989)
10      have noted that the interference from changing O2 saturation can have a very significant
        influence on the apparent COHb reading in a sample.  The interaction between Hb species
        was also reported by Dennis and Valeri (1980).  This suggests nonlinearity or a
        disproportionality in the absorption spectrum of these two species of Hb. It is also a potential
        source of considerable error in the estimation of COHb by optical methods.
15
        8.5.1.2  Carboxyhemoglobin Measurements in Populations
             Numerous studies have used the above described methodologies to characterize the
        levels of COHb for the general population. These studies have been designed to determine
        frequency distributions of COHb levels in the populations being  studied. In general  the
20      higher the frequency of COHb levels above baseline in nonsmoking subjects the greater the
        incidence of significant CO exposure.
             Carboxyhemoglobin levels in blood donors have been studied for various urban
        populations in the United States.  Included have been studies of blood donors and sources of
        CO in the metropolitan St. Louis population (Kahn et al., 1974), evaluation of smoking and
25      COHb in  the St. Louis metropolitan population (Wallace et al., 1974), analyses of
        16,649 blood samples for COHb provided by the Red Cross Missouri-Illinois blood donor
        program (Davis and Gantner, 1974), a survey of blood donors for percent COHb in Chicago,
        Milwaukee, New York and Los Angeles (Stewart et al., 1976), a national survey for COHb
        in American blood donors from urban, suburban, and rural communities across the United
30      States (Stewart et al., 1974), and the trend for percent COHb associated with vehicular traffic
        in Chicago blood donors (Stewart et al., 1976).  These extensive studies of volunteer blood

        March 12, 1990                         8-75     DRAFT-DO NOT QUOTE  OR CITE

-------
       donor populations show three main sources of exposure to CO in urban environments.  These
       are smoking, general activities (usually associated with internal combustion engines), and
       occupational exposures. For comparisons of sources, the populations are divided into two
       main groups - smokers and nonsmokers.  The main groups often are divided further into
 5     subgroups consisting of industrial workers, drivers, pedestrians, and others, for example.
       Among the two main groups, smokers show an average of 4% COHb with a usual range of 3
       to 8%; nonsmokers average about 1% COHb (Radford and Drizd,  1982).  Smoking behavior
       generally occurs as an intermittent diurnal pattern, but in some individuals who chain smoke,
       COHb levels can rise to a maximum of about 15%.
10          In addition to tobacco smoke, the most significant sources of other potential exposure to
       CO in the population are community air pollution, occupational exposures, and household
       exposures (Goldsmith,  1970).  Community air pollution comes mainly from auto exhaust and
       has a typical intermittent diurnal pattern (see Chapter 6).  Occupational exposures occur for
       up to eight hours for 5 days a week, producing COHb levels generally less than 10%.
15     However, exposures to high concentrations of CO in occupational settings have caused death
       from CO intoxication.  Household exposures usually result in less than 2% COHb, but high
       concentrations, occurring particularly during nighttime hours, have been known to cause
       death.  For example, during the winter, a number of people die as a result of using a variety
       of space heating devices in poorly ventilated spaces (Goldsmith, 1970).  Poorly vented floor
20     heaters are also a source of CO intoxications, with many such exposures occurring at night.
            More recent studies characterizing COHb levels in  the population have appeared in the
       literature. Turner et al. (1986) used an IL 182 CO-Oximeter to determine percent COHb in
       venous blood of a study group consisting of both  smoking and nonsmoking hospital staff,
       inpatients and outpatients. Blood samples were collected for 3487 subjects
25     (1255 nonsmokers) during morning hours over a five-year period.  A detailed smoking history
       was obtained at the time of blood collection. Secondary pipe or secondary cigar smokers
       were considered to be those who were initially cigarette smokers but subsequently switched to
       cigars or pipes. Primary cigar or pipe smokers were those who had never smoked cigarettes
       and were not in the habit of inhaling  large amounts of tobacco smoke as is the  observed
30     custom with cigarette smoking.  Using 1.7% COHb as a normal cutoff value, the distribution
       for the population studied showed above normal results for 94.7%  of cigarette  smokers,

       March 12, 1990                         8-76     DRAFT-DO NOT QUOTE OR CITE

-------
        10.3% of primary cigar smokers, 97.4% of secondary cigar smokers, and 94.7% of
        secondary pipe smokers.
            Zwart and Van Kampen (1985) tested a blood supply using a routine spectrophotometric
        method for total Hb and for COHb in 3022 samples of blood for transfusion in hospital
 5      patients in the Netherlands. For surgery patients over a one-year period, the distribution of
        percent COHb in samples collected as a part of the surgical protocol showed 65% below
        1.5% COHb, 26.5% between 1.5 and 5%  COHb, 6.7% between 5 and 10% COHb, and
        0.3% in excess of 10% COHb. This distribution of percent COHb was homogeneous across
        the entire blood supply, resulting in 1 in 12 patients having blood transfusions at 75%
10      available Hb capacity.
            Radford and Drizd (1982) have analyzed blood COHb in approximately 8400 samples
        obtained from respondents in the 65 geographic areas of the nationwide Health and Nutrition
        Examination Survey (HANES) during the period 1976-1980. When the frequency
        distributions of blood COHb levels are plotted on logarithmic-probability paper (Figure 8-5)
15      to facilitate comparison of the results for different age groups and smoking habits, it is
        evident that adult smokers in the U.S. have COHb levels considerably higher than those of
        nonsmokers, with 79% of the smokers' blood samples above 2% COHb and 27% of the
        observations above 5% COHb.  The nationwide distributions of persons aged 12 to 74  who
        have never smoked and ex-smokers were similar, with 5.8% of the ex-smokers and 6.4% of
20      the never-smokers above 2% COHb.   It is  evident that a significant proportion of the
        nonsmoking United States population had blood levels above 2% COHb. For these two
        nonsmoking groups, blood levels above 5% were found in  0.7% of the never-smokers  and
        1.5% of the ex-smokers. It is possible that these high blood levels could be due, in part, to
        misclassification of some smokers as either ex- or nonsmokers.  Children aged 2 to 11  had
25      lower COHb levels than the other groups, with only 2.3%  of the children's samples above
        2% COHb and 0.2% above 5% COHb.
            Wallace and Ziegenfus (1985) utilized the Radford HANES data to analyze the
        relationship between the measured COHb levels and the associated eight-hour CO
       concentration at nearby fixed monitors. COHb data were available for a total of 1658
30     nonsmokers in 20 cities. The day and hour the blood samples were drawn for each individual
       were obtained from the NHANES II data, and the preceding one-hour and eight-hour running

       March 12, 1990                         8-77      DRAFT-DO NOT QUOTE OR CITE

-------
                                   % 'PHOQ
 Figure 8-5. Frequency distributions of carboxyhemoglobin levels in the U.S. population, by
 smoking habits.

 Source: Adapted from Radford and Drizd (1982); data for NHANES II.

March 12,  1990                         8-78      DRAFT-DO NOT QUOTE OR CITE

-------
       average ambient CO levels at each fixed station in the city were calculated using the U.S.
       EPA centralized data base (SAROAD). For each of the 20 cities the station with the highest
       Spearman correlation between COHb concentrations and the preceding eight-hour CO
       averages was selected for a linear regression.  The results (Table 8-13) show that 17 of the
 5     20 stations had /f2 values ranging from 0.00 (6 cities) to 0.10.
            Finally, because the participants were part of a nationwide probability sample, all COHb
       data were merged with the CO data from the station within each city that showed the
       strongest correlation with the COHb values and a linear regression was run.  The /J2 value for
       the 1528 paired measurements was 0.031 (i.e., only 3% of the variance in the COHb
10     concentrations was explained by the ambient CO data).  The authors concluded that fixed
       outdoor CO monitors alone are, in general, not providing useful estimates of carbon
       monoxide exposure of urban residents.

       8.5.2   Carbon  Monoxide in Expired Breath
15          Carbon monoxide levels in expired breath can be used to estimate the levels of carbon
       monoxide in the subject's blood.  The basic determinants of CO levels in the alveolar air have
       been described by Douglas et al. (1912), indicating that there  are predictable equilibrium
       conditions that exist between CO bound to the Hb and the partial pressure of the CO in the
       blood.   The equilibrium relationship for CO between blood and the gas phase to which the
20     blood is exposed can  be described as follows:

                                   PCO/P02 = M (%COHb/%O2Hb)                      (8-12)
       where:
25          Pco      = partial pressure of CO in the blood,
            P02      = partial pressure of 02  in the blood,
            M       = Haldane constant (reflecting the relative affinity of the  hemoglobin for O2
                       and CO),
            %COHb = percent of total Hb combining capacity bound with CO, and
30          %O2Hb  = percent of total Hb combining capacity bound with O2.
35

       March 12, 1990                          8-79      DRAFT-DO NOT QUOTE OR CITE

-------
10
15
20
25
30
35
TABLE 8-13. REGRESSION PARAMETERS FOR THE RELATIONSHIP BETWEEN
COHb AND EIGHT-HOUR CO AVERAGES FOR 20 CITIES'
City
Atlanta
Bronx
Cincinnati
Chicago
Dayton
Des Moines
Washington
Hampton
Honolulu
Houston
Indianapolis
Los Angeles
Manhattan
Pittsburgh
Racine
Rock Hill
San Diego
San Jose
Tacoma
Washington
Wichita
All cities
n
63
65
93
78
91
90
73
89
65
71
93
66
71
55
91
85
67
59
82
73
81
1528
Slopeb
0.12(±0.03)
0.18(±0.10)
-0.02(±0.21)
-0.02(+0.06)
-0.03(+0.08)
0.003(±0.03)
0.06(±0.04)
0.16(+0.10)
0.39(+0.14)
0.24(+0.12)
0.0005(±0.019)
0.12(±0.03)
0.09(±0.03)
0.03(±0.02)
-0.12(±0.13)
0.23(±0.11)
0.01(±0.08)
0.08(±0.04)
0.04(±0.06)
0.06(±0.04)
-0.11(+0.28)
0.066(±0.009)
Intercept"
0.41(±0.09)
0.60(±0.29)
0.94(±0.22)
1.21(±0.18)
0.93(+0.15)
0.52(±0.12)
1.38(±0.18)
0.50(±0.09)
0.44(±0.18)
0.68(±0.15)
0.79(±0.07)
0.99(±0.18)
0.84(+0.08)
0.77(±0.13)
0.75(±0.14)
0.61(±0.24)
0.84(±0.13)
0.87(±0.11)
0.76(±0.14)
1.38(±0.18)
0.84(±0.35)
0.77(±0.03)
I?
0.27
0.05
0.00
0.00
0.00
0.00
0.03
0.03
0.11
0.06
0.00
0.16
0.10
0.05
0.01
0.05
0.00
0.06
0.01
0.03
0.00
0.03
"For cities with multiple CO stations, the station with the strongest Spearman correlation was chosen for the
 regression.
Percent COHb per mg/m3.
"Percent COHb.

Source: Wallace and Ziegenfus (1985).
        March 12, 1990
                                          8-80
DRAFT-DO NOT QUOTE OR CITE

-------
             The partial pressure of CO in the arterial blood will reach a steady state value relative to
        the partial pressure of CO in the alveolar gas.  Therefore, by measuring the end-expired
        breath from a subject's lungs, one can measure the end-expired CO partial pressure and, with
        the use of the Haldane relationship, estimate the blood level of COHb.  This measurement
  5     will always be an estimate because the Haldane relationship is based upon attainment of an
        equilibrium which does not occur under physiological conditions.
             The measurement of CO levels in expired breath to estimate blood levels is based upon
        application of the Haldane relationship to gas transfer hi the lung (Equation 8-12).  For
        example, when the O2 partial pressure is increased in the alveolar gas, it is possible to predict
 10     the extent to which the partial pressure of CO will increase in the alveolar gas.  This
        approach is limited, however, because of the uncertainty associated with variables that are
        known to influence gas transfer in  the lung and that mediate the direct relationship  between
        liquid phase gas partial pressures and air phase partial pressures.
             The basic mechanisms that are known to influence CO transfer in the lung have been
 15     identified through the establishment of the techniques to measure pulmonary diffusion capacity
        for CO (Forster,  1964).  Some of the factors that can result in decreased diffusion capacity
        for CO (altering the relationship between expired CO pressures and COHb levels) are
        increased membrane resistance, intravascular resistance, age, alveolar volume, pulmonary
        vascular blood volume, pulmonary blood flow, and ventilation/perfusion inequality (Forster,
 20     1964).  The extent to which each of these variables actually contributes to the variability in
        the relationship has not been experimentally demonstrated.  There are very few experiments
        that focus on the factors leading to variability in the relationship between alveolar CO and
        percent COHb at the levels of COHb currently  deemed to be of regulatory importance.  This
        may be due in part to the difficulties in working with analytical techniques, particularly the
25     blood techniques, that are very close to their limits of reproducibility. For example, a change
        of approximately 6 ppm of CO in the alveolar gas occurs for every change of 1 % COHb
        (Jones et al., 1958).  Therefore, in order to reliably measure COHb levels  to better than 0.1 %
        COHb the analytical method must be able to resolve at least 0.5 ppm CO.  This is well within
        the range of precision of the electrochemical methods (Lambert et al., 1988; Wallace et al.,
30     1988). Without the use of a well-established method for  the measurement of CO levels in
       March 12,  1990                          8-81      DRAFT-DO NOT QUOTE OR CITE

-------
       blood, the influence of all the physiological variables on the accuracy of this method remain
       undetermined.
            The expired breath method for obtaining estimates of blood levels of CO has a distinct
       advantage for monitoring large numbers of subjects because of the non-invasive nature of the
 5     method.  Other advantages include the ability to obtain an instantaneous reading and the
       ability to take an immediate replicate sample for internal standardization. The breathholding
       technique for enhancing the normal CO concentration in exhaled breath has been widely used;
       however, it should be noted that the absolute relationship between breathhold CO pressures
       and blood CO pressures has not been thoroughly established for percent COHb levels below
10     5%. The breathholding method allows time (20 seconds) for diffusion of CO into the
       alveolar air so that CO levels are higher than following normal tidal breathing.
            Partial pressures of CO in expired breath are highly correlated with percent COHb
       levels over a wide range of COHb levels (Table 8-14).  The accuracy of the breathhold
       method is unknown due to the lack of paired sample analyses of CO partial pressures in
15     exhaled breath and concurrent COHb levels in blood  utilizing a sensitive reference method
       (see Section 8.5.1). No one has attempted to determine the error of estimate involved in
       applying group  average regression relationships to the accurate determination of COHb.
       Therefore, the extrapolation of breathhold CO partial pressures to actual COHb levels must be
       made with reservation until the accuracy of this method is better understood.
20
       8.5.2.1 Measurement Methods
            Ventilation in healthy individuals involves air movement through areas in the pulmonary
       system that are  either primarily involved in conduction  of gas or in gas exchange in  the
       alveoli. In a normal breath (tidal volume) the proportion of the volume in the non-gas
25     exchanging area is termed the dead space. In the  measurement of CO in the exhaled air, the
       dead space gas volume serves to dilute the alveolar CO concentration.  Several methods have
       been developed to account for the dead space dilution.
        March 12, 1990                          8-82      DRAFT-DO NOT QUOTE OR CITE

-------
                      TABLE 8-14.  SUMMARY OF STUDIES COMPARING END-EXPIRED BREATH CO WITH COHb LEVELS
JO

to
     Thesis
                              Methods
 Sample
population
  (n)
                                                                               % COHb
                                                                               range
                                       Expired
                                       CO range
                                       (ppm)
Blood-breath
relationship
                                                                                                                                              Reference
Developed rebrealhing
method to estimate
COHb from alveolar air
CO concentration

Relationship between
alveolar breath CO and
blood COHb levels
                              Rebreathing into Douglas bag
                              Rebreathing method of
                              Sjostrand (1948)
                              Blood: venous; van Slyke
23 (sex not                  5-35
reported)
55 (men and women;          0-6
smokers and nonsmokers)
                                                                                                                         COHb
                                                                                                                         O2Hb
                                                     No regression equation
                                                     reported; line of fit
                                                     as predicted by Haldane
                                                     equation
                                                                                                                                              Sjostrand (1948)
                                                                                                                                              Carlsten et al. (19S4)



CO
1
oo


G
?0
P**
H
i
6
o


H
O
H
a
o
«

1
Using lungs as aero-
tonometers, sampling
of alveolar air allows
estimation of COHb

Verify method of
Jones et al. (1958);
apply to community
exposure survey



End-expired breath
measurements can be
used as an indicator
of exposure to ciga-
rette smoking and
community air pollu-
tion
Experimental exposure
study correlating
alveolar breath CO
with venous blood
COHb

20-s breath-hold; save end-
expired sample
Breath: NDIR corrected for CO2
Blood: Venous; NDIR

20-s breath-hold; first few
hundred ml volume discarded;
save end-expired sample
Breath: IR (CO2 scrubbed by
Ascarite)
Blood: Venous; NDIR

Not described





20-s breath-hold; discard first
half expired; save end-
expired
Breath: GC and long path IR
Blood: Venous; GC

13 (men and
women)



4 (men; 2
smokers, 2
nonsmokers)




209 (men, long
shoremen, smokers
and nonsmokers)



14 (men, white,
ages 24 to 42
year)



0.7-26.0 2-185 Line of fit as predicted
by Haldane equation:


«COHb= °-206fCOPpJ
1.2-20.0 3-100 %COHb = O.aiCO^J+O.S






0.2-19.0 0-82 Forrespondents(N=130)
with cardiorespiratory
conditions:
%COHb = l.W+O.HICOppJ
r2 = 0.56

0-32 4-250 %COHb = 109.08
+7.60[COp|DI-11.89
SE = 1.06% COHb
r = 0.976


Jones et al. (1958)




Ringold et al. (1962)






Goldsmith (1965)





Peterson (1970)






-------
s
ar
H-k
K)
                 TABLE 8-14 (cont'd).  SUMMARY OF STUDIES COMPARING END-EXPIRED BREATH CO WITH COHb LEVELS











00
oo




Thesis
Epidemiologic research
investigating tobacco
smoking behavior and
blood COHb levels


Developed practical
method to rapidly
estimate COHb from
breath samples in
field firefighting
situation


Methods
20-s breath-hold; discard first
300 ml; save next 500 ml
expired air
Breath: IR (CO2 scrubbed by
soda lime)
Blood: Venous; Spcctrophotomclric
20-s breath-hold; discard first
portion; save remainder
expired breath
Breath: Electrochemical
(Ecolyzer 2100) and GC
Blood: not described
Sample
population
(n)
59 (men and
women, smokers
and nonsmokers)



56 (men, fire-
fighters)




Expired
% COHb CO range Blood-breath
range (ppm) relationship
0.3-8.1 2-41 No regression equation
reported; estimated
regression from
bivariate plot:
«COHb = 0.21 [COJ

0.8-33 1-239 Line of fit as predicted
by Haldane equation
(without correction for
water vapor pressure):

*COHh -•197*COfpnl


Reference
Rea et al. (1973)





Stewart et al. (1976)





End-expired air
analysis may be used
to distinguish
between populations of
smokers and non-smokers
Breath: IR (CO2 scrubbed by
 soda lime)
Blood: Venous; spcclropholo-
 inetric (Tietz and Fiereck,
 (1973)
14 (sex not
reported)
                                                                             0.3-8.0
                                                                                         4-46
%COHb = O.lSICO^J-0.26
r2 = 0.92;
95% confidence limits
= ±1% COHb
                                                                                                                                 Rawbone et al. (1976)
O
o
2
0
H
O
a
o
H
W
0

Ambient CO levels
during time of breath-
holding maneuver bias
% COHb estimate






20-s breath-hold; discard first 46 (sex not
portion; save end-expired reported)
Breath: Electrochemical
(Ecolyzer 2000)
Blood: Venous; IL 192 (veri-
fied by unspecified spectro-
photometric technique)



0.4-11.5 2-64 For constant, low
ambient CO environment:
%COHb = 0.18[CO ]
r2 = 0.94

For fluctuating, high
ambient CO environment:
«COHb = O.HfCO^J
r2 = 0.48


Smith (1977)









-------
                TABLE 8-14 (cont'd).  SUMMARY OF STUDIES COMPARING END-EXPIRED BREATH CO WITH COHb LEVELS
M3
00
oo
Tl
H
6
O
n
i—i
a


Thesis
Mixed-expired air
samples are equiva-
lent to 30-s end-
expired air sample
collection method




Methods
30-s breath-hold and
rebreathing methods
Breath: K (CO2 scrubbed by
soda lime)
Blood: Venous, IL282; verified
by spcctropholometric method
of Tietz and Fiereck (1973)
Sample Expired
population % COHb CO range Blood-breath
(n) range (ppm) relationship Reference
29 (sex not 0.8-10.4 8-62 %COHb = 0.395(00^] Rees et al. (1980)
reported)
(4 non -0.0032([COBJ)2-2.4
smokers, 25
smokers)


End-expired breath
analysis is useful
for estimating %COHb
in traffic control
personnel
                           Breath: Electrochemical,
                            Ecolyzer 2000
                           Blood: Venous; IL282
                                                   ND
1.1 -12.5      5-60         Cites Stewart and (1980)
                       Stewart (1978):
                       %COHb = 0.202[CO_]
                       +0.0365
Jabara et al. (1980)
In subjects with
emphysema, decreased
end-expired [CO] is
attributed to impaired
diffusion







End-expired breath
analysis may be used
to distinguish between
smokers and non-smokers

20-8 breath-hold; expire to bag
Breath: Electrochemical,
Ecolyzer 2000
Blood: Venous; IL282








20-s breath-hold; expire to
collection tube
Breath: Electrochemical,
Ecolyzer 2000
Blood: Venous, 1L182
182 smokers



35 emphysema
patients


(sex not reported)



187 (men; 162
smokers, 25 non
smokers)


0.3-14.5 4-90 For normal smokers:
%COHb =
-0.28 + 0. 175 [COppJ

r2 = 0.98;SE = 0.76%COHb
For emphysema patients:
«COHb=
-0.12+0.21 UCO^J
r2 = 0.92
Slopes of two regres-
sion lines were signi-
ficantly different
0.4-13 3-65 *COHb = O.ISICO^J
-0.14
r = 0.97


larvis et al. (1980)











Wald et al. (1981)





-------
i
               TABLE 8-14 (cont'd).  SUMMARY OF STUDIES COMPARING END-EXPIRED BREATH CO WITH COHb LEVELS
OO
oo
Thesis
To most accurately
estimate % COHb, end-
expired breath samples
require a correction
for inspired ambient
CO at time of sampling
The correction for
inspired air may vary
between persons
Cigarette smoking interferes
with alveolar sampling
Methods
20-s breath-hold; discard first
portion; save end-expired
Breath: Electrochemical,
COED-1 (GE)
Blood: Not sampled
20-s breath-hold at room air
CO level, and al 10, 30,
and 50 ppm CO
20-s breath-hold
Breath: Dt (CO2 scrubbed
before analysis)
Blood: Venous; OSM2
spectrophotomctcr
Sample
population
(n)
1 (male, non
smoker)
7 (sex not
reported)
101 smokers
(42 men;
59 women)
Expired
% COHb CO range Blood-breath
range (ppm) relationship Reference
	 	 tCOBJnMMM
-------
25
        Mixed Expired Gas Using the Bohr Equation
             This technique involves the measurement of the mixed expired CO concentration from
        which the alveolar CO concentration is calculated.  The Bohr equation used to determine the
        physiological dead space is:
                                      Ex
* VE = F^ * VA + Fh * VD                       (8-13)
10      where:
             F& = the fractional concentration of a gas in the mixed expired air,
             VE = the minute volume of ventilation or a tidal volume,
             F^ = the fractional concentration of the gas in the alveolar space,
15           VA = the volume of alveolar gas,
             Ffc = the fractional concentration of gas in the inspired air,
             VD = the volume of dead space gas.
20
        Solving this equation for CO concentration in the alveolar gas results in:
                                      = (VE* FE  - VD* F,  )/(VE - VD)                   (8-14)
                                    CO          CO       CO
            This equation has been used by Rawbone et al. (1976) to describe the relationship
       between alveolar CO concentrations and COHb levels. These investigators measured inspired
30     ventilatory volume, inspired CO concentration and estimated dead space from anatomical
       correlations.  Carbon monoxide concentration must be converted to partial pressure in order
       to relate alveolar gas tension to percent COHb.  However, the transfer of CO from blood to
       the alveolar gas phase is not in equilibrium so the alveolar gas is a reflection of the PO, in the
       capillary blood.  This is demonstrated by the increase in alveolar CO with breathholding.
35     The relationship between alveolar levels determined from mixed expired CO concentrations
       and percent COHb is comparable to that of other methods (Table 8-14).
       March 12, 1990                          8-87      DRAFT-DO NOT QUOTE OR CITE

-------
       Breathhold
            Early methods of measurement of CO concentration in air samples by non-
       chromatographic techniques required a relatively large sample of gas, usually larger than 1 L.
       Therefore, end-tidal (alveolar) gas samples from normal respiration would not provide a
 5     sample of sufficient volume for analysis.  Jones et al. (1958) developed a method of
       inspiration to total lung capacity followed by a breathhold period of various durations.  A
       breathhold times of 20 s was found to provide near maximal values for CO pressures.  The
       breathhold period allows more time for diffusion of CO from the blood into the alveolar
       space.
10          The precision of the method has been found to be of the order of 0.1 to 0.2 ppm by
       several investigators (Hartwell et al., 1984; Wallace et al., 1988; Lambert et al., 1988).  This
       is the theoretical equivalent of 0.02  to 0.04% COHb. Physiologically, however, the breath-
       hold gas is not normal alveolar gas, since this breathhold maneuver results in the CO2
       concentration being below normal, with presumably an elevated O2 tension (Jones et al.,
15     1958; Guyattetal., 1988).
            The blood-breathhold alveolar air CO relationship is influenced by the inspired pressure
       level of CO.  Several investigators (Smith,  1977; Wallace, 1983; Wallace et al., 1988) have
       found that a correction is required in the CO pressure found in the breathhold sample.  This is
       an important consideration when this method is used to assess the exposure of subjects in their
20     normal environment (see Section 8.5.2.2).

       Rebreathing
            The earliest approach to obtaining a sufficient volume of exhaled air was rebreathing
       5 L of O2 for two to three minutes while removing the CO2 (Hackney et al.,  1962;  Carlsten
25     et al., 1954).  Hackney et al. (1962) reported that the O2 content in the rebreathing system
       fell due to dilution over the first minute after which time the decline  in O2 was related to the
       O2 consumption of the subject.  The CO concentration in the system reached its peak value at
       one minute of rebreathing in healthy subjects.  Hackney et al. also reported that the CO
       concentration in the system was related to the O2 tension in the system. The advantage to
30     using a rebreathing system is that the  ratio of change in percent COHb to change in expired
       CO is between 27 (Hackney et al.,  1962) and 30 ppm/percent COHb (Carlsten et al.,  1954).

       March 12, 1990                          8-88       DRAFT-DO NOT QUOTE OR CITE

-------
        This a gain of fivefold over the breathhold method of Jones et al. (1958).  The disadvantage
        is the time required for the measurement and the need to measure O2 in the system.

        Summary of the Methods
  5          Kirkham et al.  (1988) compared all three techniques for measuring expired CO to
        predict percent COHb. The rebreathing and breathhold methods both yield approximately
        20% high levels of "alveolar" CO than does the Bohr computation from mixed expired gas.
        Subjects rebreathed from a system that contained 20% O2 for the three minutes of the
        rebreathing.  Kirkham et al. (1988) also carried out an experiment to determine if these
10     techniques had reached a steady state between alveolar gas and blood levels. If a steady state
        existed, then changes in ventilation/perfusion and capillary blood volume would not effect the
        relationship.  Ventilation/perfusion was altered by changing  body position from lying to
        standing.  Both the mixed expired and breathholding techniques showed a significant decline
        in the alveolar CO tension when standing. Therefore, measurements of expired CO must be
15     made in the same body position relative to control measurements or reference measurements.
             The conventional relationship between blood and expired CO is assumed to be linear
        (Table 8-14).  Data collected by Rees et al. (1980), however, indicates that the relationship is
        not linear.  A second order polynomial equation proved to be the best fit of the data where:
        percent COHb = 3.95 (CO) - 0.32 (CO)2 - 2.4.  Guyatt et al. (1988) also reported a
20     nonlinear relationship where:  %COHb = -0.47 + 0.217 (CO) - 0.0006 (CO)2. Peterson
        (1970) also found that a quadratic equation described  the relationship between FA(CO) and
        percent COHb over the range of 0 to 30% COHb.  Without  more precise data,  the
        relationship between FA(CO) and COHb for under 5% COHb appears to be sufficiently linear to
        justify the use of a  linear  expression to predict percent COHb from FA(CO) measurements.
25
        8.5.2.2 Breath Measurements in Populations
             There are numerous approaches described in the literature utilizing the above methods
        for the collection and analysis of CO in expired air. In addition, many of the investigators
        have also provided  data demonstrating a relationship between the concentration of CO in
30      ambient or expired  air samples and the percent COHb in blood.  All of these approaches show
        internally consistent results and are based  on the assumption  that the air collection

        March 12, 1990                          8-89      DRAFT-DO NOT QUOTE OR CITE

-------
       methodology represents expired alveolar air. In making comparisons, differences in
       collection methods, analytical techniques, smoking history, and types of subjects being studied
       must be considered. For example, possible subjects include hospital patients with certain
       types of medical histories, joggers, and the general population. Sampling locations vary as
 5     well, ranging from outdoors to indoors, and from clinics to living rooms.
            A study in which breath measurements of CO were used to detect an indoor air problem
       has been reported (Wallace, 1983).  Sixty-five workers in an office had been complaining for
       some months of late-afternoon sleepiness and other symptoms, which they attributed to the
       new carpet.  About 40 of the workers had their breath tested for CO on a Friday afternoon
10     and again on a Monday morning. The average breath CO levels decreased from 23 ppm on
       Friday to 7 ppm on Monday morning (Figure 8-6), indicating a work-related condition.  Non-
       working fans in the parking garage and broken  fire doors were identified as the cause of the
       problem.  In this case, the ease with which the  breath measurements were taken contributed to
       the swiftness with which the problem was identified and rectified (Figure 8-7).  All
15     measurements were taken in a period of less than two hours, without the necessity for
       drawing blood,  sterilizing needles, or using a trained phlebotomist.
            Wald et al. (1981) obtained measurements of percent COHb for 11,749 men, ages 35 to
       64, who attended a medical center in London for comprehensive health screening
       examinations between 11am and  5pm. The time of smoking for each cigarette, cigar, or pipe
20     since waking was recorded at the time of collection of a venous blood sample.  COHb was
       determined using an IL 181 CO-Oximeter.  Using 2% COHb as a normal cutoff value, 81%
       of cigarette smokers, 35 % of cigar and pipe smokers, and 1 % of nonsmokers were found to
       be above normal. An investigation of COHb and alveolar CO was conducted on a subgroup
       of 187 men  (162 smokers and 25 nonsmokers).  Three samples of alveolar air were collected
25     at two-minute intervals within five minutes of collecting venous blood for COHb estimation.
       Alveolar air was collected by having  the subject hold his breath for 20 s and then
       exhale through a one-meter glass tube with an internal diameter of 17 mm and  fitted with a
       three-liter anesthetic bag at the distal end.  Air  at the proximal end of the tube  was considered
       to be alveolar air, and a sample was removed by a small side tube located at 5  mm from the
30     mouthpiece. CO content was measured using an Ecolyzer.  The instrumental measurement is
       based on detection of the oxidation of CO to CO2 by a catalytically active electrode in an

       March 12, 1990                         8-90      DRAFT-DO NOT QUOTE OR CITE

-------
                  15
                8
                  10





1
i
11



Basement office I
I

P
I
•1
f



1
m
Legend
Tfift Alveolar
T3 Ambient

l
n
Control office
J,
I
n
                     *<*
'x'X
Figure 8-6. Changes in alveolar CO of nonsmoking basement office workers compared to
nonsmoking workers in other offices between Friday afternoon, Monday morning, and
Monday afternoon.

Source: Wallace (1983).
JU
25-
20-

10-
5-
0














Before





































After

i
I



EPA 8-hour
CO standard

Illll ,
                             Period of record, February 8 to March 15
Figure 8-7. Eight-hour average CO concentrations in basement office before and after
corrective action.
Source: Wallace (1983).


March 12, 1990
            8-91     DRAFT-DO NOT QUOTE OR CITE

-------
      aqueous electrolyte.  The mean of the last two readings to the nearest 0.25 ppm was recorded
      as the alveolar CO.  Subjects reporting recent alcohol consumption were excluded because
      ethanol in the breath affects the response of the Ecolyzer.  A linear regression equation of
      percent COHb on alveolar CO (see Table 8-14) had a correlation coefficient of 0.97,
 5    indicating that a COHb level could be estimated reliably from an alveolar CO level.
            Honigman et al. (1982) determined alveolar CO concentrations by end-expired breath
      analysis for athletes (joggers).  The group included 36 nonsmoking males and 7 nonsmoking
      females, all conditioned joggers,  covering at least 21 mi per week for the previous
      six months in the Denver area. The participants exercised for a 40-min period each day over
10    one of three defined courses in the Denver urban environment (elevation 1610 meters).
      Expired air samples were collected and analyzed before start of exercise, after 20 min and
      again at the end of the 40-min exercise period.  Heart rate measurements at 20 min and
      40 min were 84 and 82% of mean age predicted maxima, respectively, indicating exercise in
      the aerobic range.  Relative changes in expired air CO concentrations were plotted and
15    compared to ambient air concentrations for CO measured at the time of collecting breath
      samples.  Air and breath samples were analyzed using an MSA model 70.  Relative  changes
      in expired end air CO based on the concentration of CO in breath before the start of exercise
      were plotted in terms of the ambient air concentrations measured during  the exercise period,
      at both 20 and 40 min of exercise.  For ambient concentrations of CO below 6 ppm, the
20    aerobic exercise served to decrease the relative amount of end air expired CO as compared to
      the concentration measured before the start of exercise. For ambient concentrations in the
      range of 6 to 7 ppm, there was no net change in the CO concentrations in  the expired air.
      For ambient air concentrations in excess of 7 ppm, the aerobic exercise resulted in relative
      increases of expired CO, with the increases after 40 min being greater than similar increases
25    observed at the 20 min measurements.  Sedentary controls at the measurement stations
      showed no relative changes. Thus, aerobic exercise, as predicted by the physiologic models
      of uptake and elimination, is shown to enhance transport of CO,  thereby decreasing  the time
      to reach equilibrium conditions.
            Verhoeff et al. (1983) surveyed fifteen identical residences  which used  natural gas for
30    cooking and geyser units for water heating.  CO concentrations in the flue gases were
      measured using an Ecolyzer (2000 series).  The flue gases were diluted to the dynamic range

      March 12, 1990                          8-92      DRAFT-DO NOT QUOTE OR CITE

-------
       of the instrument for CO (determined by Draeger tube analyses for CO2 dilution to 2-2.5%).
       Theoretical concentrations for CO2 in the flue gasses is 11.70% CO2 under conditions of zero
       excess air for the natural gas to air mixture used.  Breath samples were collected from
       29 inhabitants by having each participant hold a deep breath for 20 seconds and exhale
 5     completely through a glass sampling tube (225 mL volume).  The sampling tube was
       stoppered and taken to a laboratory for analysis of CO content using a gas-liquid
       chromatograph (Hewlett Packard, 5880A).  The overall coefficient of variation for sampling
       and analysis was 7%,  based on results of previous measurements.  No significant differences
       were observed for nonsmokers as a result of their cooking or dishwashing activities using the
10     natural gas fixtures. There was a slight increase in expired air CO for smokers, but this may
       be due to the possibility of increased smoking during the dinner hour.
            Wallace et al. (1984) report data on measurements of end expired air CO and
       comparisons with predicted values based on personal CO measurements for populations in
       Denver and Washington, D.C. Correlations between breath CO and preceding eight-hour
15     average CO exposures were high (0.6 to 0.7) in both cities. However, breath CO levels
       showed no relationship with ambient CO measurements at the nearest fixed-station monitor.
       Correlation coefficients were calculated for one-hour, two-hour, ... up to 10-h average
       personal CO exposures;  the highest correlations occurred at seven- to nine-hours, providing
       support for the EPA choice of 8 h as an averaging time for the NAAQS.
20          The major large-scale study employing breath measurements of CO was carried out by
       EPA in Washington and Denver in the winter of 1982-1983 (Johnson 1984; Hartwell et al.,
       1984; Akland et al., 1985; Wallace et al., 1984, 1988).  In Washington, 870 breath samples
       were collected from 812 participants; 895 breath samples were collected from 454 Denver
       participants (two breath  samples on two consecutive days in Denver).  All participants also
25     carried personal monitors to measure their exposures over a 24-h period in Washington or a
       48-h period in Denver.  The subjects in each city formed a probability sample representing
       1.2 million adult nonsmokers in Washington and 500,000 adult nonsmokers in Denver.  The
       distribution of breath levels in the two cities is shown for the subjects themselves
       ("unweighted" curves) and the larger populations they represented  ("weighted" curves) in
30     Figure 8-8.
      March 12, 1990                          8-93      DRAFT-DO NOT QUOTE OR CITE

-------
10*
» **** ****?'
•

t
f

1




•
*
7
4
1
a
i


. 'T
- ' i

	 1-rt
:!'
: i yK-
! ': i
• *(-
. , . .
• t ' •
1 ( •"
1 . .,.
1 ' '"


'•
•4 "i- r
\ I 1
•+ :|
': i :i
DENVER
NASHIN
. 1 ..
• i •.
• • ••
001 008
1
:
• - ' 1 :

! •
i •
•ii i:
: 1 ! ,1
; -j- -7— fr
', ! I
• ' ::

.
i : i


i—trJi


'I i "
'. 	 rr •-
l :
(N-M9)
QTON. DC (
. 1
	
W W
r irfr
! : 1
44 TTf
li-W
• 1

i: : :

:: ::

i. .1



T — r-
I* • :
•r — "+-
:. r
•*ili

: j |

i:
N"«2
* > • i
• i -i
... !"

»4 . 90 W TO
| 1 ! 1 ! 1 I'M. I"1 1 . : 1 ' 1. '











«> •
•



— r~

—r-]





:
.
6)

r^
t 1
y


01 02 OJ 1 2
\\i






':


!


Ill'
: t




. ^^
£ !



|~T


^tr





:: i
....





f4
: l :
• i :


*X^







/•










TT*

ii
'is*
: 1 1 ;








4







<-?

&'
*:
ili*
; "
	 |....
.'. cu


MUU
1 : . .
tO 50 40 30 70
.l.i; Ti. — 1. 1 I1 . ! . I !l" !
fr

















"ITT


n


-rr

;;;•
kTIVI
5 10 MX
i i
•1::
















r^
d
i: :

•rrr

• ;;
EFR
•"-r
















T*
^



i
. i .
EQU
lir



T^*~













^
4-\


i

: ::
ENC
i i
-tt~


-r^-











j^K
H
: f :


= ;.
: i:
:::.
1:1 i














V
'.^


w
10













^
yf*


7
.r «
[/

1

: L' i
EIOI
•~rrj
i 2 S



• i '


...*-'


; :

k
Jrt
• • : .
-U-
•fj-r
: 4 ;. .


4TEO



I




^
• y.

V


-4-i—

i i
UNWEIOHTED :
....
.::
1"'
'. 1 .

..
40 90 w » n


. :
A
..
«0 '



U5

i ' :
. . I
~- T^^"


02010





:':< rt/i
,-n
fj

El


:::; :
f. . . {*•&

«


^ '••
....',....
•^TTI
• i


-t— i



M i
— - - T
'

'
I

I



:
'::
.L.
-





'


• i
fi M t*
r •


^
r4-
/-y
y
/
/
















,


; ;







yJ

;;l



'::

»
001


T]



— i — i —



: i :



: i :

; ! j

l :
' !

* ) |
1 '
• 1 •

**l*9»
1111
*
/
t
»
4
J
2
1
                                        to«

                                        »•

                                        M

                                        7*
                                        I
                                        8
                                        II
8
       Figure 8-8. Distributions of CO in breath of adult nonsmokers in Denver and Washington.


       Source: Wallace et al. (1988).
       March 12, 1990
8-94      DRAFT-DO NOT QUOTE OR CITE

-------
           These distributions appear to be roughly lognormal, with geometric means of 5.2 ppm
      CO for Denver and 4.4 ppm for Washington. Geometric standard deviations were about 1.6
      for each city.  Arithmetic means were 7.1 ppm for Denver and 5.2 ppm for Washington.
           Of greater regulatory significance is the number of people whose COHb levels exceeded
 5    the value of 2.1 %, because EPA has determined that the current 9 ppm, eight-hour average
      standard would keep more than 99.9% of the most sensitive nonsmoking adult population
      below this level of protection (Federal Register, 1985). An alveolar CO value of about
      10 ppm would correspond to a COHb level of 2%.  The percent of people with measured
      breath values exceeding this level was about 6% in Washington. This percentage was
10    increased to 10% when the correction for the effect of room air was applied (Figure 8-9).  Of
      course, since the breath samples were taken on days and at times when they were not
      necessarily at their highest level during the year, these percentages are lower limits of the
      estimated number of people who may have incurred  COHb levels above 2%. Yet the two
      central stations in Washington recorded a total of one exceedance of the 9 ppm standard
15    during the winter of 1982-1983. Models based on fixed-station readings would have
      predicted that an exceedingly tiny proportion of the Washington population received exposures
      exceeding the standard.  Therefore, the results from the breath  measurements indicated that a
      much larger portion of both Denver and Washington residents were exceeding 2% COHb than
      was predicted by models based on fixed-station measurements.
20         It also should be noted that the number of people with measured maximum eight-hour
      exposures exceeding the EPA outdoor standard of 9  ppm was only about 3.5% of the
      Washington subjects.  This value appears to disagree with the value of 10% obtained from the
      corrected breath samples.  However, the personal  monitors used in the study were shown to
      experience several different problems, including a loss of response associated with battery
25    discharge toward the end of the 24-h monitoring period, which caused them to read low.
      Therefore, Wallace et al. (1988) concluded that the breath measurements were correct and the
      personal air measurements were biased low. The  importance of including breath
      measurements in future exposure and epidemiology studies is indicated by this study.
           Hwang et al. (1984) describe the use of expired air analysis for CO in an emergency
30    clinical setting to diagnose the presence and extent of CO intoxication.  The subjects were
      47 Korean patients brought for emergency treatment showing various levels of

      March 12, 1990                           8-95       DRAFT-DO NOT QUOTE OR CITE

-------
      PERCENT OF SAMPLE POPULATION (N?625) EXCEEDING
                      CONCENTRATION SHOWN
                      95 90 80  70 8050 40 30 20 10 5  21
       50
       40
       30

       20
    10
?  8
£  6
         _  EPA 8-h STANDARD
    I-   3

    H   2h
    ~Z.
    UJ
    8  °-8
       0.6
      0.3

      0.2
       ai
                                             CORRECTED-
                                                  OBSERVED
                  12   5  10  20  3040506070  80  90 95  9899
                   CUMULATIVE  FREQUENCY (%)
Figure 8-9. Percent of Washington sample population with eight-hour average CO exposures
exceeding the concentrations shown. The eight-hour period ended at the time the breath
sample was taken. The curve marked "OBSERVED" contains the actual readings of the
personal monitors; these readings were corrected using the measured bream values.
Source: Wallace et al. (1988).

March 12, 1990
                            8-96    DRAFT-DO NOT QUOTE OR CITE

-------
       consciousness ranging from alertness (11), drowsiness (21), stupor (7), semicoma (5), coma
       (1), and unknown (2). The study group included 16 males, ages 16 to 57, and 31 females,
       ages 11 to 62.  Exposure durations ranged from two to ten hours, with all exposures
       occurring in the evening and nighttime hours.  The source of CO was mainly from use of
  5    charcoal fires for cooking and heating.  In order to estimate expired air CO concentrations, a
       detector tube (Gastec ILa containing potassium palado sulfite as both a reactant and color
       change indicator for the presence of CO on silica gel) was fitted to a Gastec  manual sampling
       pump.  One stroke of the sampling plunger represents 100 cc of  air.  A 100-cc expired air
       sample was collected by inserting a detector tube at a nostril  and slowly pulling back the
 10    plunger for one full stroke for expired air. A 10-cc sample of venous blood  also was
       collected at this time for determination of percent COHb using a CO-Oximeter. The subjects
       showed signs of acute intoxication, and two relationships were found: a low CO (less the
       100 ppm) and a high CO,  (greater than  100 ppm) between expired air CO (CO^) and percent
       COHb.
 15         Cox and Whichelow (1985) analyzed end-exhaled air (collected over approximately the
       last half of the exhalation cycle) for CO concentrations  for a  random population of 168 adults
       - 69 smokers and 99 nonsmokers.  The results were used to evaluate the influence of home
       heating systems on exposures  to and adsorption of CO.  Ambient indoor concentrations of CO
       were measured in the homes of study subjects.  The subjects  included 86 men and 82 women,
 20    ranging in age from 18 to 74. Interviews  were conducted usually in the living room of the
       subject's home. The type of heating system in use was noted and indoor air  concentration of
       CO was measured using an Ecolyzer. After the ambient indoor CO was determined, a breath
       sample was collected from the subject. The subject was asked to hold a deep breath for 20 s,
       and then to exhale completely into a trilaminate plastic bag.   The bag was fitted to the port of
25     the Ecolyzer and the CO content of the exhaled air was measured. For smokers, the time
       since smoking their last cigarette and the number of cigarettes per day were  noted. For
       nonsmokers there was a strong correlation between ambient CO and expired air CO. With
       smokers, the correlation was strongest with the number  of cigarettes per day.  The data also
       supported the supposition that smokers are a further source of ambient CO in the indoor
30     environment.
      March 12,  1990                          8-97      DRAFT-DO NOT QUOTE OR CITE

-------
           Lambert et al. (1988) compared breath CO levels to blood COHb levels in 28 subjects
      (including two smokers).  Breath CO was collected using the standard technique developed by
      Jones et al. (1958):  maximal inspiration followed by a 20-s breath hold and discarding the
      first portion of the expired breath.  One-liter bags were used to collect the breath samples,
 5    which were measured on an Ecolyzer 2000 monitor equipped with Purafil® and activated
      charcoal filters to scrub interferences such as alcohol.  Excellent precision (±0.2 ppm) was
      obtained in 35 duplicate samples.  Blood samples were collected within 15 min of the breath
      samples using a gas-tight plastic syringe rinsed with sodium heparin.  Carboxyhemoglobin
      was measured using an IL 282 CO-Oximeter.  Some samples also were measured using a GC.
10    The CO-Oximeter appeared to be reading high, particularly in the <2% COHb range of
      interest.  A reading of 0.5  %COHb on the particular CO-Oximeter used in this study would
      be only 0.3% using the GC and a reading of 1% COHb  on the CO-Oximeter would be only
      0.7% on the GC.
           The results showed poor correlation between the pooled nonsmokers' breath CO and
15    blood COHb levels (n = 104 measurements, r2 = 0.19). However, better correlation was
      observed for three individual nonsmoking subjects, who appeared to have roughly parallel
      slopes (0.13 to 0.27) but widely differing blood COHb intercepts (0.1, 0.4, and
      1.0 %COHb). The authors interpreted these findings  as suggesting that the CO-Oximeter
      may be sensitive to an unidentified factor in the blood of individuals.  Possible factors
20    suggested by the authors include triglycerides and hemochromagens (a group of compounds
      formed  when heme combines with  organic nitrogen species), which are known to absorb light
      in the 550 to 555 nm wavelength used by the CO-Oximeter. Another concern regarding the
      CO-Oximeter is the calibration method, which uses saturated (98%COHb) bovine serum as
      the only span calibration point. This is far above the  0.5 to 3 %COHb range of interest for
25    nonsmoking subjects.
            In view of the great dependence in laboratory studies on the CO-Oximeter, the authors
      concluded that there was "an important and immediate need to further investigate the
      instrument's performance at COHb levels resulting from typical ambient  exposures." Such
      studies should include a comprehensive side-by-side study with other reference methods,
30    including gas chromatography, manometry, and other spectrophotometric methods.  Full
      spectral scans should be performed to quantify  light absorbance and scattering effects on

      March 12, 1990                         8-98       DRAFT-DO NOT QUOTE OR CITE

-------
      COHb measurement. Also, an improved calibration method should be developed, including
      whole blood and dye standards and the use of multiple calibration points in the 0 to
      3% COHb range of interest.

 5    8.5.3 Potential Limitations
      8.5.3.1 Pulmonary Disease
           A major potential influence on the relationship between blood and alveolar partial
      pressures of CO is the presence of significant lung disease.  Hackney et al. (1962)
      demonstrated the slow increase in exhaled CO concentration in a rebreathing system peaked
10    after 1.5 minutes in healthy subjects but required 4 minutes in a subject with lung disease.
      These findings have been substantiated by Guyatt et al. (1988) who reported that patients with
      pulmonary disease did not have the same relationship between percent COHb and breathhold
      CO concentrations.  The group with pulmonary disease had a FEV/FVC percentage of
       <71.5% compared to the healthy subjects with a FEV/FVC percentage of > 86%. The
15    linear regression for the healthy group was COHb =  0.629 + 0.158(ppm CO); and for the
      pulmonary disease group was COHb = 0.369 + 0.185(ppm CO).  This means that at low
      CO levels, individuals with obstructive pulmonary disease would have a lower "alveolar" CO
      level for any given percent COHb level than would the healthy subjects.

20    8.5.3.2 Subject Age
           The relationship between age and COHb level is not well established.  Kahn et al.
      (1974) reported that nonsmoking subjects under the age of 19 years had a significantly lower
      percent COHb than older subjects but there was no difference in COHb between the ages of
      20 and 59 years.  Kahn et al. also reported that there was a slight decrease in the  COHb
25    levels in nonsmoking subjects over the age of 60 years.  Radford and Drizd (1982) also
      reported that younger subjects, 3 to 11 years' old, had lower levels of COHb than did the
      older age group of 12 to 74 years.  Goldsmith (1970) reported that expired CO levels  were
      unchanged with age in nonsmokers; however, there was a steady decline in the expired CO
      levels with age in smokers.  The decrease in expired CO is disproportionately large for the
30    decrease in COHb levels measured by  Kahn et al. (1974) in older subjects.  Therefore, by
      comparison of the data from these two studies, it would appear that older subjects have higher

      March 12, 1990                         8-99      DRAFT-DO NOT QUOTE OR CITE

-------
      levels of COHb than predicted from the expired CO levels.  It is not known how much of this
      effect is due to aging of the pulmonary system, resulting in a condition similar to the subjects
      with obstructive pulmonary disease.

 5    8.5.3.3 Effects of Smoking
           Studies evaluating the effect of cigarette smoking on end-expired CO have found a
      phasic response that depends on smoking behavior (Woodman et al., 1987; Henningfield
      et al.,  1980). There is an initial rapid increase in the CO concentration of expired air as a
      result of smoking.  This is followed by a rapid (five-minute) decrease after cessation of
10    smoking and a slow decrease over the 5- to 60-min period after smoking.  A comparison of
      the results from one study (Tsukamoto and Matsuda, 1985) showed that the CO concentration
      in expired air increases by approximately 5 ppm by  smoking one cigarette.  This corresponds
      to an increase of 0.67% COHb based on blood-breath relationships developed by the authors.
      Use of cigarettes with different tar and nicotine yields or the use of filter tip cigarettes showed
15    no apparent effect on end-expired CO concentrations (Castelli et al.. 1982).  However,
      knowledge of the breath sample results does.  King et al.  (1983) were able to show that
      immediate feedback on CO concentrations promoted behavioral changes in cigarette smokers
      which subsequently  resulted in lower CO concentrations in expired air for  return visits.
      Furthermore, reported rates of smoking were lower for the second visit than those reported
20    for the first visit.
           The relationship between breathhold CO and blood CO is apparently altered due to
      smoking making the detection of small changes difficult.  Guyatt et al. (1988) have shown
      that smoking one cigarette results in a variable response in the relationship between breath-
      hold CO and COHb levels. The range of FA(CO) values for a 1 % increase in COHb was from -
25    5 ppm to +5 ppm.  The correlation between  the change in FA(CO) and the change in  COHb in
      500 subjects was only 0.705.  This r value indicates that only 50% of the change in FA(CO) was
      due to changes in COHb.  It is not known how much of this residual error is due to subject
      compliance or to error in the method.  Therefore, the results obtained with breathholding in
      smoking subjects should be viewed with caution unless large differences in FA(CO) are reported
30    (i.e., considerable cigarette consumption is being evaluated).
       March 12, 1990                          8-100     DRAFT-DO NOT QUOTE OR CITE

-------
           In summary, the measurement of exhaled breath has the advantages of ease, speed,
      precision, and greater subject acceptance than measurement of blood COHb.  However, the
      accuracy of the breath measurement procedure and the validity of the Haldane relationship
      between breath and blood at low environmental CO concentrations remains in question.
      There appears to be a clear research need to validate the breath method at low CO exposures.
      In view of the possible problems with the CO-Oximeter, such validation should be done using
      gas chromatography for the blood COHb measurements.
10    8.6 SUMMARY AND CONCLUSIONS
           The current NAAQS for CO (9 ppm for eight hours, 35 ppm for one hour) are designed
      to protect against actual and potential human exposures in outdoor air that would cause
      adverse health effects.  Compliance with the NAAQS is determined by measurements taken at
      fixed-site ambient monitors, the use of which is intended to provide some measure of the
15    general level of exposure of the population represented by the CO monitors.  Results of both
      exposure monitoring in the field, and modeling studies, summarized in this Chapter indicate
      that individual personal exposure does not directly correlate with CO concentrations
      determined by using fixed-site monitors alone.  This observation is due to the mobility of
      people and to the spatial and temporal variability of CO concentrations.  While failing to
20    show a correlation between individual personal monitor exposures and simultaneous nearest
      fixed-site monitor concentrations, studies do suggest that aggregate personal exposures are
      lower on days of lower ambient CO levels as determined by the fixed-site monitors and
      higher on days of higher ambient levels.
           Cigarette consumption represents a special case of CO exposure; for the smoker it
25    almost always dominates over personal exposure from other sources. Studies by Radford and
      Dridz (1982)  show that COHb levels of cigarette smokers average 4% while those of
      nonsmokers average 1 %.  Therefore, this summary focuses on environmental exposure of
      nonsmokers to CO.
           People encounter CO in a variety of environments that include travelling in motor
30    vehicles, working at their jobs, visiting urban locations associated with combustion sources,
      or cooking over a gas range. Studies of human exposure have shown that among these

      March 12,  1990                         8-101     DRAFT-DO NOT QUOTE OR CITE

-------
       settings the motor vehicle is the most important for regularly encountered elevations of CO.
       Studies by Flachsbart et al. (1987) indicated that CO exposures while commuting in
       Washington, D.C. average 9 to 14 ppm at the same time that fixed station monitors record
       concentrations of 2.7 to 3.1 ppm.  Similar studies conducted by EPA in Denver and
 5     Washington have demonstrated that the motor vehicle interior has the highest average CO
       concentrations (averaging 7 to 10 ppm) of all microenvironments (Johnson,  1984). In these
       studies, 8% of all commuters experienced eight-hour exposures greater than 9 ppm while only
       1 % of noncommuters received exposures over that level. Furthermore, commuting exposures
       have been shown to be highly variable with some commuters breathing CO in excess of
10     25 ppm.
           Another important setting for CO exposure is the workplace.  In general, exposures at
       work exceed CO exposures during nonwork periods, apart from commuting to and from
       work.  Average concentrations  may be elevated during this period since workplaces are often
       located in congested areas  that have higher background CO concentrations than do many
15     residential neighborhoods.  Occupational and nonoccupational exposures may overlay one
       another and result in a higher concentration of CO in the blood. Certain occupations also
       increase the risk of high CO exposure (e.g., those occupations involved directly with vehicle
       driving, maintenance, and  parking).  Occupational groups exposed to CO by vehicle exhaust
       include auto mechanics; parking garage and gas station attendants; bus, truck or taxi drivers;
20     police  and warehouse workers.   Other industrial processes produce CO directly or as a by-
       product, including steel production, coke ovens, carbon black production, and petroleum
       refining. Firefighters,  cooks, and construction workers also may be exposed at work to
       higher CO levels.  Occupational exposure in industries or setting with CO production also
       represent some of the highest individual exposures observed in field monitoring studies.  For
25     example, in EPA's CO exposure study in Washington, of the approximately 4% (29 of 712)
       of subjects working in jobs classified as having a high potential for CO exposure, seven
       subjects (or approximately 25%) experienced eight-hour CO exposures in excess of 9 ppm.
           The highest indoor nonoccupational CO exposures are associated with combustion
       sources and include enclosed parking garages, service stations, restaurants and stores. The
30     lowest indoor CO concentrations are found in homes, churches, and health care facilities.
       EPA's Denver Study showed that passive cigarette smoke is associated with increasing a

       March 12, 1990                          8-102    DRAFT-DO NOT QUOTE OR CITE

-------
       nonsmoker's exposure by an average of about 1.5 ppm and that use of a gas range is
       associated with about 2.5 ppm increase at home. Other sources which may contribute to CO
       in the home include combustion space heaters and wood burning stoves.
            As noted above, people encounter different and often higher exposures than predicted
  5    from fixed-site monitoring data, because of the highly  localized nature of CO sources.  For
       example, during the winter sampling period, 10% of Denver volunteers  and 4% of
       Washington volunteers recorded personal exposures in  excess of 9 ppm for eight hours.
       Breath measurements  from the Washington volunteers indicated that as much as 9% of the
       population could have experienced a 9 ppm, eight-hour average. In contrast, during the
 10    entire winter period of 1982-1983, the two ambient CO monitors in Washington reported only
       one exceedance of the 9-ppm level.  In another study, using data from analyses of COHb in
       blood,  Wallace and Ziegenfus (1985) report that CO in blood is uncorrelated with CO
       measured by ambient  monitors. These findings  point out the necessity of having personal CO
       measurements augment fixed-site ambient monitoring data when total human exposure is to be
 15    evaluated.  Data from these field studies can be  used to construct and test models of human
       exposure that account for time and activity patterns known to affect exposure to CO.  Models
       developed to date tend to underpredict the variability of CO  exposures observed in field
       studies and have not been able to successfully predict individual exposures. The models may
       be modified and adjusted using information  from field  monitoring studies in order to capture
20    the observed distribution of CO exposures, including the higher exposures found in the tail of
       the exposure distribution. The models also are useful for evaluating alternative pollutant
       control strategies.
25    8.7  REFERENCES
      Akland, G. G.; Ott, W. R.; Wallace, L. A. (1984) Human exposure assessment: background concepts, purpose,
30          and overview of the Washington, DC - Denver, Colorado field studies. Presented at: 77th annual meeting
            of the Air Pollution Control Association; June; San Francisco, CA. Pittsburgh, PA: Air Pollution Control
            Association; paper no. 84-121.1.
      Akland, G. G.; Hartwell, T. D.; Johnson, T. R.; Whitmore, R. W. (1985) Measuring human exposure to carbon
35          monoxide in Washington, D.C., and Denver, Colorado, during the winter of 1982-1983. Environ. Sci.
            Technol. 19: 911-918.

       March 12, 1990                          8-103     DRAFT-DO NOT QUOTE OR CITE

-------
      Allred, E. N.; Bleecker, E. R.; Chairman, B. R.; Dahms, T. E.; Gottlieb, S. O.; Hackney, J. D.; Hayes, D.;
             Pagano, M. Selvester, R. H.; Walden, S. M.; Warren, J. (1989) Acute effects of carbon monoxide
             exposure on individuals with coronary artery disease. Cambridge, MA:  Health Effects Institute; research
             report no. 25.
 5
      Amendola, A. A.; Hanes, N. B. (1984) Characterization of indoor carbon monoxide levels produced by the
             automobile. In: Berglund, B.; Lindvall, T.; Sundell, J., eds. Indoor air: proceedings of the 3rd
             international conference on indoor air quality and climate, v. 4, chemical characterization and personal
             exposure; August; Stockholm, Sweden.  Stockholm, Sweden: Swedish Council for Building Research;
10           pp. 97-102. Available from: NTIS, Springfield, VA; PB85-104214.

      Andrecs, L.;  Stenzler, A.; Steinberg, H.; Johnson, T, (1979) Carboxyhemoglobin levels in garage workers.  Am.
             Rev.  Respir. Dis. 119: 199.

15    Ayres, S. M.; Criscitiello, A.; Giannelli, S., Jr. (1966) Determination of blood carbon monoxide content by gas
             chromatography. J. Appl.  Physiol. 21: 1368-1370.

      Biller, W. F.; Johnson, T. R.; Paul, R.; Feagans, T. B.; Duggan, G. M.;  McCurdy, T. R.; Thomas, H. C.
             (1981) A general model for estimating exposure associated with alternative NAAQS.  Presented at: 74th
20           annual meeting of the Air Pollution Control Association; June. Pittsburgh,  PA: Air Pollution Control
             Association; paper no. 81-18.4.

      Blackmore, D. J. (1974) Aircraft addicent toxicology: U.K.  experience 1967-1972. Aerosp. Med. 45: 987-994.

25    Blankart, R.; Koller,  E.; Habegger, S. (1986) Occupational exposure to carbon monoxide and treatment with
             hyperbaric oxygenation. J. Clin. Chem. Clin. Biochem. 24: 806-807.

      Bondi, K. R.; Very, K. R.;  Schaefer, K.  E. (1978) Carboxyhemoglobin levels during a submarine patrol.
             Undersea Biomed. Res. 5:  17-18.
30
      Breysse,  P. A.; Bovee, H. H. (1969) Use of expired air-carbon monoxide for Carboxyhemoglobin determinations
             in evaluating carbon monoxide exposures resulting from the operation of gasoline fork lift trucks in holds
             of ships. Am.  Ind. Hyg. Assoc. J. 30: 477-483.

35    Brown, L. J. (1980) A new instrument for the simultaneous measurement of total hemoglobin,  %
             oxyhemoglobin, % Carboxyhemoglobin, % methemoglobin, and oxygen content in whole blood.  IEEE
             Trans. Biomed. Eng. 27:  132-138.

      Brunekreef, B.; Smit,  H. A.; Biersteker,  K.; Boleij, J. S. M.;  Lebret, E. (1982) Indoor carbon  monoxide
40           pollution in The Netherlands. Environ.  Int. 8: 193-196.

      Butt, J.;  Davies, G. M.; Jones, J.  G.; Sinclair, A. (1974) Carboxyhaemoglobin levels in blast furnace workers.
             Ann. Occup.  Hyg. 17: 57-63.

45    Carlsten, A.; Holmgren, A.; Linroth, K.; Sjostrand, T.;  Strom, G. (1954) Relationship between low values of
             alveolar carbon monoxide concentration and Carboxyhemoglobin percentage in human blood. Acta
             Physiol. Scand. 31:  62-74.

      Castelli,  W. P.; Garrison, R. J.; McNamara, P. M.; Feinleib, M.; Faden, E. (1982) The relationship of carbon
50           monoxide in expired air of cigarette smokers to the tar nicotine content of the cigarette they smoke:  the
             Framingham study.  In: 55th scientific sessions of the American Heart Association: November; Dallas TX.
             Circulation 66: 11-315. (American Heart Association monograph no. 91).
        March 12, 1990                              8-104      DRAFT-DO NOT QUOTE OR CITE

-------
       Castellino, N. (1984) Evaluation des tests de dose et d'effet dans 1'exposition professionnelle au monoxyde de
              carbone [Evaluation of dose and effect tests for occupational exposure to carbon monoxide]. In:
              Hommage Professeur Rene Truhaut; pp. 191-193.

  5    Chace, D. H.; Goldbaum, L. R.; Lappas, N. T. (1986) Factors affecting the loss of carbon monoxide from
              stored blood samples. J. Anal. Toxicol. 10: 181-189.

       Chaney, L. W. (1978) Carbon monoxide automobile emissions measured from the interior of a traveling
              automobile. Science (Washington, DC) 199: 1203-1204.
 10
       Chapin, F. S., Jr. (1974) Human activity patterns in the city. New York, NY: Wiley-Interscience Publishers.

       Chovin, P. (1967) Carbon monoxide: analysis of exhaust gas investigations in Paris. Environ. Res. 1: 198-216.

 15    Clayton, A. C.; White, S. B.; Settergren, S. K. (1985) Carbon monoxide exposure of residents of Washington,
              D. C.: comparative analyses. Research Triangle Park, NC: U. S. Environmental Protection Agency,
              Environmental Monitoring Systems Laboratory; contract no. 68-02-3679.

       Clerbaux, T.; Willems, E.; Brasseur, L. (1984) Evaluation d'une methode chromatographique de reference pour
 20           la determination du taux sanguin de carboxyhemoglobine [Assessment of a reference chromatographic
              method for the determination of the carboxyhemoglobin level in blood].  Pathol. Biol. 32:  813-816.

       Coburn, R. F.; Danielson, G. K.; Blakemore,  W. S.; Forster, R. E., II. (1964) Carbon monoxide in blood:
              analytical method and sources of error. J. Appl. Physiol. 19: 510-515.
 25
       Code of Federal Regulations. (1977) National primary and secondary ambient air quality standards. C. F. R.
              40(50) (July 1): 3-33.

       Collison, H. A.; Rodkey, F. L.;  O'Neal, J. D. (1968) Determination of carbon  monoxide in blood by gas
 30           chromatography. Clin. Chem. (Winston-Salem, NC) 14:  162-171.

       Colwill, D. M.; Hickman, A. J. (1980) Exposure of drivers to carbon monoxide. J. Air Pollut. Control Assoc.
              30: 1316-1319.

 35    Constantino, A. G.; Park, J.; Caplan, Y. H. (1986) Carbon monoxide analysis:  a comparison of two
              co-oximeters and headspace gas chromatography. J. Anal. Toxicol. 10: 190-193.

       Cooke, R. A. (1986) Blood lead and carboxyhaemoglobin levels in roadside workers. J. Soc. Occup. Med.
              36: 102-103.
 40
       Cortese, A. D.; Spengler, J,  D. (1976) Ability of fixed monitoring stations to represent personal carbon
              monoxide exposures. J. Air Pollut. Control Assoc. 26: 1144-1150.

       Council on Environmental Quality. (1980) The eleventh annual report of the Council on Environmental Quality.
45            Washington, DC: Council on Environmental Quality.

       Cox, B. D.; Whichelow, M.  J. (1985)  Carbon monoxide levels in the breath of smokers and nonsmokers: effect
              of domestic heating systems. J.  Epidemiol. Commun. Health 39: 75-78.

50     Dahms, T. E.; Horvath, S. M. (1974) Rapid, accurate technique for determination of carbon monoxide in blood.
              Clin. Chem. (Winston-Salem, NC) 20: 533-537.
       March 12,  1990                              8-105       DRAFT-DO NOT QUOTE OR CITE

-------
      Davidson, C. I.; Osborn, J. F.; Fortmann, R. C. (1984) Modeling and measurement of pollutants inside houses
             in Pittsburgh, Pennsylvania. In: Berglund, B.; Lindvall, T.;  Sundell, J., eds. Indoor air: proceedings of
             the 3rd international conference on indoor air quality and climate, v. 4, chemical characterization and
             personal exposure; August; Stockholm, Sweden. Stockholm, Sweden: Swedish Council for Building
 5           Research; pp. 69-74. Available from: NTIS, Springfield, VA; PB85-104214.

      Davis, G. L.; Gantner, G. E., Jr. (1974) Carboxyhemoglobin in volunteer blood donors. JAMA J. Am. Med.
             Assoc. 230: 996-997.

10    Dellarco, M. J.; Wallace, L. A.; Ott, W. R. (1988) A computerized bibliographic literature information system
             for total human exposure monitoring research. Presented at:  81st annual meeting of the Air Pollution
             Control Association;  June; Dallas, TX. Pittsburgh, PA: Air  Pollution Control Association; paper
             no. 88-115.2.

15    Dennis, R. C.; Valeri, C. R. (1980)  Measuring percent oxygen saturation of hemoglobin, percent
             carboxyhemoglobin and methemoglobin, and concentrations  of total hemoglobin and oxygen in blood of
             man, dog, and baboon. Clin. Chem. (Winston-Salem, NC) 26: 1304-1308.

      Dockery, D. W.; Spengler, J. D. (1981) Personal exposure to respirable particulates and sulfates. J. Air Pollut.
20           Control Assoc. 31: 153-159.

      Douglas, C. G.; Haldane, J.  S.; Haldane, J. B. S. (1912)  The laws  of combination of haemoglobin with carbon
             monoxide and oxygen. J. Physiol. (London) 44: 275-304.

25    Drabkin, D. L.; Austin, J. H. (1932) Spectrophotometric  constants for common haemoglobin derivatives in
             human, dog and rabbit blood. J. Biol. Chem. 98:  719.

      Duan, N. (1981) Microenvironment types: a model for human exposure to air pollution. Stanford, CA: Stanford
             University, Dept. of Statistics; SIMS technical report no. 47.
30
      Duan, N. (1982) Models for human exposure to air pollution. Environ. Int.  8:  305-309.

      Duan, N. (1985) Application of the microenvironment monitoring approach  to assess human exposure to carbon
             monoxide. Research Triangle Park, NC: U. S. Environmental Protection Agency, Environmental
35           Monitoring Systems Laboratory; EPA report no. EPA-600/4-85-046. Available from: NTIS, Springfield,
             VA; PB85-228955.

      Evans, R. G.; Webb, K.; Homan, S.; Ayres, S. M. (1988) Cross-sectional and longitudinal changes in
             pulmonary function associated with automobile pollution among bridge and tunnel  officers. Am. J. Ind.
40           Med. 14: 25-36.

      Federal Register. (1985) Review of the national ambient air quality  standards for carbon monoxide; final rule.
             F. R. (September 13) 50: 37484-37501.

45    Federal Register. (1989) Air contaminants; final rule, carbon monoxide. F. R.  (January 19) 54: 2651-2652.

      Flachsbart, P. G. (1985) Prototypal models of commuter exposure to CO from motor vehicle exhaust.  Presented
             at: 78th annual meeting of the Air Pollution Control Association; June; Detroit, MI. Pittsburgh, PA: Air
             Pollution Control Association; paper no.  85-39.6.
50
      Flachsbart, P. G. (1989) Effectiveness of priority lanes in reducing  travel time and carbon monoxide exposure.
             Inst. Transport.  Engin. J. 59: 41-45.
        March 12, 1990                              8-106      DRAFT-DO NOT QUOTE OR CITE

-------
       Flachsbart, P. G.; Ah Yo, C. (1989) Microenvironmental models of commuter exposure to carbon monoxide
              from motor vehicle exhaust. Research Triangle Park, NC: U. S. Environmental Protection Agency,
              Atmospheric Research and Exposure Assessment Laboratory; in press.

  5    Flachsbart, P. G.; Brown, D. E. (1989) Employee exposure to motor vehicle exhaust at a Honolulu shopping
              center. J. Architect.  Plan. Res. 6: 19-33.

       Flachsbart, P. G.; Ott, W. R. (1984) Field surveys of carbon monoxide in commercial settings using personal
              exposure monitors. Washington, DC: U. S. Environmental Protection Agency, Office of Monitoring
 10           Systems and Quality Assurance; EPA report no. EPA-600/4-84-019. Available from: NTIS, Springfield,
              VA;PB84-211291.

       Flachsbart, P. G.; Ott, W. R. (1986) A rapid method for surveying CO concentrations in high-rise buildings. In:
              Berglund, B.; Berglund, U.; Lindvall, T.;  Spengler, J.; Sundell, J.,  eds. Indoor air quality: papers from
 15           the third international conference on indoor air quality and climate; August 1984; Stockholm, Sweden.
              Environ. Int.  12: 255-264.

       Flachsbart, P. G.; Mack, G. A.; Howes, J.  E.; Rodes, C. E.  (1987) Carbon monoxide  exposures of Washington
              commuters. J. Air Pollut.  Control Assoc. 37:  135-142.
 20
       Forster, R. E. (1964) Diffusion of gases. In: Fenn, W. O.; Rahn, H., eds. Handbook of physiology: a critical,
              comprehensive presentation of physiological knowledge and concepts. Section 3: Respiration, volume I.
              Washington, DC: American Physiological Society; pp. 839-872.

 25    Fristedt, B.; Akesson, B. (1971) Haelsorisker av bilavgaser i parkeringshusens serviceanlaeggningar [Health
              hazards from automobile exhausts at service facilities of multistory garages]. Hyg. Revy 60: 112-118.

       Fugas,  M. (1986) Assessment of true human exposure to air pollution. In: Berglund, B.; Berglund, U.; Lindvall,
              T.; Spengler, J.; Sundell, J., eds. Indoor air quality: papers from the third international conference on
 30           indoor air quality and climate; August 1984; Stockholm, Sweden. Environ.  Int.  12: 363-367.

       Godin,  G.; Wright, G.; Shephard, R. J. (1972) Urban exposure to carbon monoxide. Arch. Environ. Health
              25: 305-313.

 35    Goldbaum, L. R.; Chace, D. H.; Lappas, N. T. (1986) Determination of carbon monoxide in blood by gas
              chromatography using a thermal conductivity detector. J. Forensic Sci. 31: 133-142.

       Goldsmith, J. R. (1965) Discussion: epidemiologic studies of chronic respiratory diseases. In: Seventh annual air
              pollution medical research conference; February 1964; Los Angeles,  CA. Arch.  Environ. Health
40           10: 386-388.

       Goldsmith, J. R. (1970) Contribution of motor vehicle exhaust, industry, and cigarette smoking to community
              carbon monoxide exposures. In: Coburn, R. F., ed. Biological effects of carbon monoxide. Ann. N. Y.
              Acad. Sci. 174: 122-134.
45
       Goldstein, B. D.; Goldring, R. M.; Amorosi, E. L. (1975)  Investigation of mechanisms of oxygen delivery  to
              the tissues in individuals with chronic low-grade carbon monoxide exposure. New York, NY:
              Coordinating Research Council; report no. CRC-APRAC-CAPM-8-68-4. Available from: NTIS,
              Springfield, VA; PB-244153.
50
       Goldstein, G. M.; Raggio, L.; House, D. (1986) Factors influencing carboxyhemoglobin stability. Research
              Triangle Park, NC: U. S. Environmental Protection Agency, Health Effects Research Laboratory; EPA
              technical report no. TR-1811.  Available from:  NTIS, Springfield, VA; AD-A165032.
        March 12,  1990                              8-107      DRAFT-DO NOT QUOTE OR CITE

-------
      Gordon, G. S.; Rogers, R. L. (1969) Project monoxide: a medical study of an occupational hazard of fire
              fighters. Washington, DC: International Association of Fire Fighters.

      Gothe, C.-J.; Fristedt, B.; Sundell, L.; Kolmodin, B.; Ehrner-Samuel, H.; Gothe, K. (1969) Carbon monoxide
 5            hazard in city traffic: an examination of traffic policemen in three Swedish towns. Arch. Environ. Health
              19: 310-314.

      Grut, A. (1949) Chronic carbon monoxide poisoning: a study in occupational medicine. Copenhagen,  Denmark:
              Ejnar Munksgaard.
10
      Guillot, J. G.; Weber, J.  P.; Savoie, J. Y. (1981) Quantitative determination of carbon monoxide in blood by
              head-space gas chromatography. J. Anal. Toxicol. 5: 264-266.

      Guyatt, A. R.;  Kirkham,  A. J. T.; Mariner, D. C.; Gumming, G. (1988) Is alveolar carbon monoxide an
15            unreliable index of carboxyhaemoglobin changes during smoking in man? Clin. Sci. 74: 29-36.

      Haagen-Smit, A. J. (1966) Carbon monoxide levels in city driving. Arch. Environ. Health 12: 548-551.

      Hackney, J. D.; Kaufman, G. A.; Lashier, H.; Lynn,  K. (1962) Rebreathing estimate of carbon monoxide
20            hemoglobin. Arch. Environ. Health 5:  300-307.

      Hartwell, T. D.; Clayton, C. A.; Ritchie, R. M.; Whitmore, R. W.; Zelon, H. S.; Jones, S.  M.; Whitehurst,
              D. A.  (1984) Study of carbon monoxide exposure of residents of Washington, DC and Denver, Colorado.
              Research Triangle Park, NC: U. S. Environmental Protection Agency, Environmental Monitoring
25            Systems Laboratory; EPA report no. 600/4-84-031. Available from: NTIS, Springfield, VA;
              PB84-183516.

      Heinold, D. W.; Sacco, A. M.; Insley, E. M»(1987)  Toolbooth operator exposure on the New Jersey Turnpike.
              Presented at: 80th annual meeting of the Air Pollution Control Association; June; New York, NY.
30            Pittsburgh, PA: Air Pollution Control Association; paper no. 87-84A. 12.

      Henningfield, J. E.; Stitzer, M. L.; Griffiths, R. R. (1980) Expired air carbon monoxide accumulation and
              elimination as a function of number of cigarettes smoked. Addict. Behav. 5:  265-272.

35    Hernberg, S.;  Karava, R.; Koskela, R.-S.;  Luoma, K. (1976) Angina pectoris, ECG findings and blood pressure
              of foundry workers in relation to carbon monoxide exposure. Scand. J. Work Environ. Health 2(suppl.
              1): 54-63.

      Hickey, R. J.;  Clelland, R. C.; Boyce, D. E.; Bowers, E. J. (1975) Carboxyhemoglobin levels. JAMA J. Med.
40           Assoc.  232: 486-488.

      Honigman, B.; Cromer, R.; Kurt, T. L. (1982) Carbon monoxide levels in athletes during exercise in an urban
              environment. J. Air Pollut. Control Assoc. 32: 77-79.

45   Horvath, S. M.; Roughton, F. J. W. (1942) Improvements in the gasometric estimation of carbon monoxide in
              blood.  J. Biol. Chem.  144: 747-755.

       Horvath, S. M.; Agnew, J. W.; Wagner, J. A.; Bedi, J. F. (1988) Maximal aerobic capacity at several ambient
              concentrations of carbon monoxide at several altitudes. Health Effects Institute research report no. 21.
50
       Hosey, A. D. (1970) Priorities in developing criteria for "breathing air" standards. J. Occup. Med. 12: 43-46.
        March 12, 1990                               8-108      DRAFT-DO NOT QUOTE OR CITE

-------
       Humphreys, M. P.; Knight, C. V.; Pinnix, J. C. (1986) Residential wood combustion impacts on indoor carbon
              monoxide and suspended particulates. In: Proceedings from the 1986 EPA/APCA symposium on the
              measurement of toxic air pollutants; April; Raleigh, NC. Research Triangle Park, NC: U. S.
              Environmental Protection Agency, Environment Monitoring Systems Laboratory; pp. 736-747; EPA
  5           report no. EPA-600/9-86-013. Available from: NTIS, Springfield, VA; PB87-182713.

       Hwang, S. J.; Cho, S. H.; Yun,  D. R. (1984) [Postexposure relationship between carboxyhemoglobin in blood
              and carbon monoxide in expired air]. Seoul J. Med. 25: 511-516.

 10    Iglewicz, R.; Rosenman,  K. D.;  Iglewicz, B.; O'Leary, K.; Hockemeier, R. (1984) Elevated levels of carbon
              monoxide in the patient compartment of ambulances. Am. J. Public  Health 74: 511-512.

       International Committee for Standardization in Haemotology. (1978) Recommendations for reference method for
              haemoglobinometry in human blood (ICSH standard EP 6/2: 1977) and specifications for international
 15           haemiglobincyanide reference preparation (ICSH standard EP 6/3: 1977). J. Clin.  Pathol. 31: 139-143.

       Jabara, J. W.; Keefe, T. J.; Beaulieu, H. J.; Buchan, R.  M. (1980)  Carbon  monoxide: dosimetry in occupational
              exposures in Denver, Colorado.  Arch. Environ. Health 35: 198-204.

 20    Jarvis, M. J.;  Russell, M. A. H.; Saloojee, Y. (1980) Expired air carbon monoxide: a simple breath test of
              tobacco smoke intake. Br. Med.  J. 281: 484-485.

       Johnson, T. (1984) A study of personal exposure to carbon monoxide in Denver, Colorado. Research Triangle
              Park, NC: U.  S. Environmental  Protection Agency, Environmental Monitoring Systems Laboratory; EPA
 25           report no. EPA-600/4-84-014. Available from: NTIS, Springfield, VA; PB84-146125.

       Johnson, T. (1987) A study of human activity patterns in Cincinnati, OH. Palo Alto, CA: Electric Power
              Research  Institute; report no. RP940-06.

 30    Johnson, T. (1988) A methodology for estimating carbon monoxide exposure and resulting carboxyhemoglobin
              levels in Denver,  Colorado. Presented at: Research planning conference on human activity patterns; May;
              Las Vegas, NV. Las Vegas, NV: University of Nevada, Las Vegas,  Environmental Research Center.

       Johnson, T.; Paul, R.  A.  (1983) The NAAQS exposure model (NEM) applied  to carbon monoxide. Research
 35           Triangle Park, NC:  U.S.  Environmental Protection Agency,  Office of Air Quality Planning and
              Standards; report no. EPA-450/5-83-003. Available from:  NTIS, Springfield, VA; PB84-242551.

       Johnson, T.; Wijnberg, L. (1988) A model for simulating personal exposure to carbon monoxide which
              incorporates serial correlation. PEI Associates, Inc; report no. 68-02-4606 WA6,  PN3766-6.
 40
       Johnson, B. L.; Cohen, H. H.; Struble,  R.; Setzer, J. V.; Anger,  W. K.; Gutnik, B. D.; McDonough, T.;
              Hauser, P. (1974) Field evaluation of carbon monoxide exposed toll  collectors. In:  Xintaras, C.; Johnson,
              B. L.; de Groot, I.,  eds. Behavioral toxicology: early detection of occupational hazards: proceedings of a
              workshop; National  Institute for  Occupational Safety and Health and University of Cincinnati; June 1973;
45           Cincinnati, OH. Cincinnati, OH: U. S. Department of Health, Education, and Welfare, Robert A. Taft
              Laboratories; National Institute for Occupational Safety and Health report no. 74-126; pp. 306-328.
              Available from: NTIS, Springfield, VA; PB-259322.

       Johnson, C. J.; Moran, J.; Pekich, R. (1975a) Carbon monoxide in school buses. Am. J. Public Health
50           65: 1327-1329.

       Johnson, C. J.; Moran, J. C.; Paine,  S. C.; Anderson, H. W.;  Breysse, P. A.  (1975b) Abatement of toxic levels
              of carbon monoxide  in Seattle ice -skating rinks. Am. J. Public Health  65: 1087-1090.
       March 12,  1990                              8-109      DRAFT-DO NOT QUOTE OR CITE

-------
      Johnson, T.; Capel, J.; Wijnberg, L. (1986) Selected data analyses relating to studies of personal carbon
             monoxide exposure in Denver and Washington, D.C. Research Triangle Park, NC: U. S. Environmental
             Protection Agency, Environmental Monitoring Systems Laboratory; contract no. 68-02-3496.

 5    Johnson, T.; Paul,  R. A.; McCurdy, T. (1988) Advancements in estimating urban population exposure. Presented
             at: 81st annual meeting of the Air Pollution Control Association; June; Dallas,  TX. Pittsburgh, PA: Air
             Pollution Control Association; paper no. 88-127.1.

      Jones, J. G.; Walters, D. H.  (1962) A study of carboxyhaemoglobin levels in employees at an integrated
10           steelworks. Ann. Occup. Hyg. 5: 221-230.

      Jones, R. H.; Ellicott, M. F.; Cadigan, J. B.; Gaensler, E. A.  (1958) The relationship between alveolar and
             blood carbon monoxide concentrations during breathholding. J. Lab.  Clin. Med. 51: 553-564.

15    Kahn, A.; Rutledge, R. B.; Davis, G. L.; Altes, J. A.; Gantner, G. E.; Thornton, C. A.; Wallace, N. D. (1974)
             Carboxyhemoglobin sources in the metropolitan St. Louis population. Arch. Environ. Health
             29: 127-135.

      Kane, D. M. (1985) Investigation of the method to determine carboxyhaemoglobin in blood. Ontario, Canada:
20           Department of National Defence - Canada, Defence and Civil Institute of Environmental Medicine;
             DCIEM no. 85-R-32.

      Katsumata, Y.; Sato, K.; Yada, S. (1985) A simple and high-sensitive method for determination of carbon
             monoxide in blood by gas chromatography. Hanzaigaku Zasshi 51: 139-144.
25
      King, A. C.; Scott, R. R.; Prue, D. M. (1983) Reactive effects of assessing  reported rates and alveolar carbon
             monoxide levels  on smoking behavior. Addict. Behav. 8: 323-327.

      Kirkham, A. J. T.; Guyatt, A. R.; Gumming, G. (1988) Alveolar carbon monoxide:  a  comparison of methods of
30           measurement and a study of the effect of change in body posture. Clin. Sci. 74: 23-28.

      Lambert, W. E.; Colome, S.  D. (1988) Distribution of carbon  monoxide exposures within indoor residential
             locations. Presented at: 81st annual meeting of the Air Pollution Control Association; June; Dallas, TX.
             Pittsburgh, PA: Air Pollution Control Association; paper no. 88-90.1.
35
      Lambert, W. E.; Colome, S.  D.; Wojciechowski, S.  L. (1988) Application of end-expired breath sampling to
             estimate carboxyhemoglobin levels in community air pollution exposure assessments. Atmos. Environ.
             22: 2171-2181.

40    Lebret, E. (1985) Air pollution in Dutch homes: an exploratory study in environmental epidemiology.
             Wageningen, The Netherlands: Department of Air Pollution, Department of Environmental and Tropical
             Health; report R-138; report 1985-221.

      Lowry, L. K.  (1986) Biological exposure index as a complement to the TLV. JOM J. Occup. Med. 28: 578-582.
45
      Lundgren, G.  R. (1971) Carbon monoxide. In: Encyclopedia of occupational health and safety, v. 1, A-K. New
              York,  NY: McGraw-Hill Book Company; pp. 253-256.

      Lurmann, F. W.; Coyner, L.; Winer, A. M.; Colome, S. (1989) Development of a new regional human exposure
50            (REHEX) model and its application to the California south  coast air  basin. Presented at: 60th annual
              meeting of the Air and Waste Management Association; June; Anaheim, CA. Pittsburgh, PA: Air and
              Waste Management Association; paper no. 89-27.5.
        March 12, 1990                              8-110      DRAFT-DO NOT QUOTE OR CITE

-------
       Maas, A. H. J.; Hamelink, M. L.; De Leeuw, R. J. M. (1970) An evaluation of the spectrophotometric
              determination of HbO2, HbCO and Hb in blood with the CO-Oximeter IL 182. Clin. Chim. Acta
              29: 303-309.

 5    McCredie, R. M.; Jose, A. D. (1967) Analysis of blood carbon monoxide and oxygen by gas chromatography. J.
              Appl. Physiol. 22: 863-866.

       Meyer, B. (1983) Indoor air quality. Reading, MA: Addison-Wesley Publishing Co., Inc.

10    Michelson, W.; Reed, P. (1975) The time budget. In: Michelson, W., ed. Behavioral research methods in
              environmental design. Stroudsburg, PA: Dowden, Hutchinson, and Ross; pp.  180-234.

       Miranda, J. M.; Konopinski, V. J.; Larsen, R. I. (1967) Carbon monoxide control in a high highway tunnel.
              Arch. Environ. Health 15: 16-25.
15
       Moschandreas, D.  J.; Zabransky, J., Jr. (1982) Spatial variation of carbon monoxide and oxides of nitrogen
              concentrations inside residences. Environ. Int. 8: 177-183.

       National Academy of Sciences. (1969) Effects of chronic exposure to low levels of carbon monoxide on human
20           health, behavior, and performance. Washington, DC: National Academy of Sciences and National
              Academy of Engineering.

       National Institute for Occupational Safety and Health. (1972) Criteria for a recommended standard....occupational
              exposure to carbon monoxide. Washington, DC: U. S. Department of Health, Education, and Welfare;
25           report no.  NIOSH-TR-007-72. Available from:  NTIS, Springfield, VA; PB-212629.

       National Research  Council. (1986) The airliner cabin environment: air quality and safety. Washington,  DC:
              National Academy Press.

30    Nilsson, C.-A.; Lindahl, R.; Norstrom, A. (1987) Occupational exposure to chain saw exhausts in logging
              operations. Am. Ind. Hyg. Assoc. J. 48: 99-105.

       Ocak, A.; Valentour, J. C.; Blanke, R. V. (1985) The effects of storage conditions on the stability of carbon
              monoxide in  postmortem blood. J. Anal. Toxicol. 9: 202-206.
35
       Ott, W. R. (1971)  An urban survey technique for measuring the spatial variation of carbon monoxide
              concentrations in cities. Stanford, CA: Stanford University, Dept. of Civil Engineering.

       Ott, W. R. (1977)  Development of criteria for siting air monitoring stations. J.  Air Pollut. Control Assoc.
40           27: 543-547.

       Ott, W. R. (1981)  Exposure estimates based on computer generated activity patterns. Presented at: 74th annual
              meeting of the Air Pollution Control Association; June; Philadelphia, PA. Pittsburgh, PA: Air Pollution
              Control Association; paper no. 81-57.6.
45
       Ott, W. R. (1982)  Concepts of human exposure to air pollution. Environ. Int. 7: 179-196.

       Ott, W. R. (1984)  Exposure estimates based on computer generated activity patterns. J. Toxicol. Clin.  Toxicol.
              21: 97-128.
50
       Ott, W. R. (1985)  Total human exposure:  an  emerging science focuses on humans as receptors of environmental
              pollution. Environ. Sci. Technol. 19:  880-886.
        March  12, 1990                              8-111      DRAFT-DO NOT QUOTE OR CITE

-------
      Ott, W.; Eliassen, R. (1973) A survey technique for determining the representativeness of urban air monitoring
             stations with respect to carbon monoxide. J. Air Pollut. Control Assoc. 23: 685-690.

      Ott, W.; Flachsbart, P. (1982) Measurement of carbon monoxide concentrations in indoor and outdoor locations
 5           using personal exposure monitors. Environ. Int. 8: 295-304.

      Ott, W.; Mage, D. (1974) A method for simulating the true human exposure of critical population groups to air
             pollutants. In: proceedings [of the] international symposium, recent advances in assessing the health
             effects of environmental pollution, v. IV; June; Paris, France. Luxembourg; Commission of the European
10           Communities; pp. 2097-2107.

      Ott, W. R.; Willits, N. H. (1981) CO exposures of occupants of motor vehicles: modeling the dynamic response
             of the vehicle. Stanford, CA: Stanford University, Dept. of Statistics; SIMS technical report no. 48.

15    Ott, W. R.; Rodes,  C. E.; Drago, R. J.; Williams, C.; Burmann, F. J. (1986) Automated data-logging personal
             exposure monitors for carbon monoxide. J. Air Pollut. Control Assoc. 36: 883-887.

      Ott, W.; Thomas, J.; Mage,  D.; Wallace, L. (1988) Validation of the simulation of human activity and pollutant
             exposure (SHAPE) model using paired days from the Denver, CO, carbon monoxide field study. Atmos.
20           Environ. 22: 2101-2113.

      Pandian, M. D. (1987) Evaluation of existing total human exposure  models. Las Vegas, NV:  U. S.
             Environmental Protection Agency, Environmental Monitoring Systems Laboratory; EPA report no.
             EPA/600/4-87/044. Available from: NTIS, Springfield, VA; PB88-146840.
25
      Petersen, W. B.; Allen, R. (1982) Carbon monoxide exposures to Los Angeles area commuters. J. Air Pollut.
              Control Assoc. 32: 826-833.

      Petersen, G. A.; Sabersky, R. H. (1975) Measurements of pollutants inside an automobile. J. Air Pollut. Control
30            Assoc. 25: 1028-1032.

      Peterson, J. E.  (1970) Postexposure relationship of carbon monoxide in blood and expired air. Arch. Environ.
             Health 21: 172-173.

35    Pierce, R. C.; Louie, A. H.; Sheffer, M. G.; Woodbury, N. L. (1984) The estimation of total human exposure
              to pollutants: integrated models for indoor and outdoor exposure to air pollutants. In: Berglund, B.;
              Lindvall, T.; Sundell, J., eds. Indoor air: proceedings of the 3rd international conference on indoor air
              quality and  climate,  v. 4, chemical characterization and personal exposure; August; Stockholm,  Sweden.
              Stockholm,  Sweden:  Swedish Council for Building Research; pp. 397-402. Available  from: NTIS,
40            Springfield, VA; PB85-104214.

      Poulton, T. J. (1987) Medical helicopters: carbon monoxide risk? Aviat. Space Environ. Med. 58: 166-168.

      Purdham, J. T.; Holness, D. L.; Pilger, C. W. (1987) Environmental and medical assessment of stevedores
45            employed in ferry operations. Appl. Ind. Hyg. 2: 133-139.

      Quackenboss, J. J.; Kanarek, M.  S.; Spengler, J. D.; Letz, R. (1982) Personal monitoring for nitrogen dioxide
              exposure: methodological considerations for a community  study. Environ.  Int. 8: 249-258.

50    Quinlan, P.; Connor, M.; Waters, M. (1986) Environmental evaluation of stress and hypertension in municipal
              bus drivers. Cincinnati, OH: National Institute for Occupational Safety and Health. Available from:
              NTIS, Springfield, VA; PB86-144763.
        March 12, 1990                              8-112      DRAFT-DO NOT QUOTE OR CITE

-------
       Radfoid, E. P. (1981) Sensitivity of health endpoints: effect on conclusions of studies. EHP Environ. Health
              Perspect. 42: 45-50.

       Radford, E. P.; Drizd, T. A. (1982) Blood carbon monoxide levels in persons 3-74 years of age: United States,
  5           1976-80. Hyattsville, Md: U. S. Department of Health and Human Services, National Center for Health
              Statistics; DHHS publication no. (PHS) 82-1250. (Advance data from vital and health statistics, no. 76).

       Rai, V. S.; Minty,  P. S. B.  (1987) The determination of carboxyhaemoglobin in the presence of
              sulfhaemoglobin. Forensic Sci. Int. 33: 1-6.
 10
       Ramsey, J. M. (1967) Carboxyhemoglobinemia in parking garage employees. Arch. Environ. Health
              15: 580-583.

       Rawbone, R. G.; Coppin, C. A.; Guz, A. (1976) Carbon monoxide in alveolar air as an index of exposure to
 15           cigarette smoke. Clin. Sci. Mol. Med. 51: 495-501.

       Rea, J. N.; Tyrer, P. J.; Kasap, H. S.; Beresford, S. A. A. (1973) Expired air, carbon monoxide, smoking, and
              other variables: a community study. Br. J. Prev. Soc. Med. 27: 114-120.

 20    Rees, P. J.; Chilvers, C.; Clark, T. J. H. (1980) Evaluation of methods used to estimate inhaled dose of carbon
              monoxide. Thorax 35: 47-51.

       Repace, J. L.; Ott,  W. R.; Wallace, L. A.  (1980) Total human exposure to air pollution. Presented at: 73rd
              annual meeting of the Air Pollution Control Association; June; Montreal, QB, Canada. Pittsburgh, PA:
 25           Air Pollution Control Association; paper no. 80-61.6.

       Ringold, A.; Goldsmith, J. R.; Helwig, H. L.; Finn, R.; Schuette, F. (1962) Estimating recent carbon monoxide
              exposures: a rapid method. Arch. Environ.  Health 5: 308-318.

 30    Robinson, J. P.  (1977) How Americans use their time: a social psychological analysis of everyday behavior. New
              York, NY:  Praeger Publishers.

       Rodkey, F. L.; Hill, T. A.; Pitts, L. L.; Robertson, R. F.  (1979) Spectrophotometric measurement of
              caroxyhemoglobin and methemoglobin in blood. Clin. Chem. (Winston-Salem, NC) 25:  1388-1393.
 35
       Rosenman, K. D. (1984)  Cardiovascular disease and work place exposures. Arch. Environ. Health 39: 218-224.

       Rosenstock, L.;  Cullen, M. R. (1986a) Cardiovascular disease. In: Clinical occupational medicine. Philadelphia,
              PA: W. B. Saunders; pp. 71-80.
40
       Rosenstock, L.; Cullen, M. R. (1986b) Neurologic disease. In: Clinical occupational medicine. Philadelphia, PA:
              W. B. Saunders; pp.  118-134.

       Roughton, F. J.  W.; Root, W. S. (1945) The estimation of small amounts of carbon monoxide in air. J. Biol.
45           Chem. 160:  135-148.

       Sammons, J. H.; Coleman, R. L. (1974) Firefighters' occupational exposure to carbon monoxide. J.  Occup.
              Med. 16: 543-546.

50    Sawicki, C. A.; Gibson, Q. H. (1979) A photochemical method for rapid and precise determination of carbon
              monoxide levels in blood. Anal. Biochem. 94: 440-449.

       Scholander, P. F.; Roughton, F. J. W.  (1943) Micro gasometric estimation of the blood gases: II. carbon
              monoxide. J. Biol. Chem. 148:  551-563.


       March 12, 1990                              8-113       DRAFT-DO NOT QUOTE OR CITE

-------
      Schwab, M.; Colome, S. D.; Spengler, J. D.; Ryan, P. B.; Billick, I. H.  (1989) J. Ind. Toxicol.: submitted.

      Settergren, S. K.; Hartwell, T. D.; Clayton, C. A. (1984) Study of carbon monoxide exposure of residents of
             Washington, D. C. - additional analysis. Research Triangle Park, NC: U. S. Environmental Protection
 5           Agency, Environmental Monitoring Systems Laboratory; contract no. 68-02-3679.

      Sexton, K.; Ryan, P. B. (1988) Assessment of human exposure to air pollution: methods, measurements, and
             models. In: Watson, A. Y.; Bates, R. R.; Kennedy, D., eds. Air pollution, the automobile, and public
             health. Washington, DC: National Academy Press; pp. 207-238.
10
      Sexton, K.; Spengler, J. D.; Treitman, R. D. (1984) Personal exposure to respirable particles: a case study in
             Waterbury, Vermont. Atmos. Environ. 18: 1385-1398.

      Shackelford, J.; Ott, W.; Wallace, L.  (1988) Total human exposure and indoor air quality: an automated
15           bibliography (BLIS) with summary abstracts. Research Triangle Park, NC: U.  S. Environmental
             Protection Agency, Office of Research and Development; EPA report no. EPA-600/9-88-011.

      Simmon, P. B.; Patterson,  R.  M. (1983) Commuter exposure model: description of model methodology and
             code. Research Triangle Park, NC: U. S. Environmental Protection Agency, Environmental Monitoring
20           Systems Laboratory; EPA report no. EPA-600/8-83-022. Available from: NTIS, Springfield, VA;
             PB83-215566.

      Sisovic, A.; Fugas, M. (1985) Indoor concentrations of carbon monoxide in selected urban microenvironments.
             Environ. Monit. Assess. 5:  199-204.
25
      Sjostrand, T. (1948) A method for the determination of carboxyhaemoglobin concentrations by analysis of the
             alveolar air. Acta Physiol. Scand. 16: 201-210.

      Small, K. A.; Radford, E.  P.; Frazier, J. M.; Rodkey, F. L.; Collison, H. A. (1971) A rapid method for
30           simultaneous measurement of carboxy- and methemoglobin in blood. J. Appl. Physiol. 31: 154-160.

      Smith, N. J. (1977) End-expired air technic for determining occupational carbon monoxide exposure. J. Occup.
             Med. 19: 766-769.

35    Spengler, J. D.; Soczek, M. L. (1984) Evidence for improved ambient air quality and the need for personal
             exposure research.  Environ. Sci. Technol. 18:  268A-280A.

      Spengler, J. D.; Stone, K.  R.; Lilley, F. W. (1978) High carbon monoxide levels measured in enclosed skating
             rinks. J. Air Pollut. Control Assoc. 28: 776-779.
40
      Spengler, J. D.; Treitman, R. D.; Tosteson, T. D.; Mage, D. T.; Soczek, M. L. (1985) Personal exposures to
             respirable particulates and implications for air pollution epidemiology. Environ. Sci. Technol.
             19: 700-707.

45    Stem, F. B.; Lemen, R. A.; Curtis, R. A. (1981) Exposure of motor vehicle examiners to carbon monoxide: a
             historical prospective mortality study. Arch. Environ. Health 36: 59-66.

      Stewart,  R. D.; Baretta, E. D.; Platte, L. R.; Stewart, E. B.; Kalbfleisch, J. H.; Van Yserloo, B.; Rimm, A. A.
             (1974) Carboxyhemoglobin levels in American blood donors. JAMA J. Am. Med. Assoc. 229:
50           1187-1195.

      Stewart,  R. D.; Hake, C. L.;  Wu, A.; Stewart, T. A.; Kalbfleisch, J. H.  (1976) Carboxyhemoglobin trend in
             Chicago blood donors, 1970-1974. Arch. Environ. Health 31: 280-286.
        March 12, 1990                              8-114      DRAFT-DO NOT QUOTE OR CITE

-------
       Stock, T. H.; Kotchmar, D. J.; Contant, C. F.; Buffler, P. A.; Holguin, A. H.; Gehan, B. M.; Noel, L. M.
              (1985) The estimation of personal exposures to air pollutants for a community-based study of health
              effects in asthmatics - design and results of air monitoring. J. Air Pollut. Control Assoc. 35: 1266-1273.

  5    Svercl, P. V.; Asin, R. H. (1973) Home-to-work trips and travel: report no. 8 prepared for the Nationwide
              Personal Transportation Study. Washington, DC:  U. S. Department of Transportation, Federal Highway
              Administration.

       Szalai, A., ed. (1972) The use of time: daily activities of urban and suburban populations in 12 countries. The
 10           Hague, The Netherlands: Mouton and Co.

       Taylor, J. D.; Miller, J. D.  M. (1965) A source of error  in the cyanmethaemoglobin method of determination of
              haemoglobin concentration in blood containing carbon monoxide. Am. J. Clin. Pathol. 43: 265.

 15    Tietz, N. W.; Fiereck, E. A. (1973) The spectrophotometric measurement of carboxyhemoglobin. Ann. Clin.
              Lab. Sci. 3: 36-42.

       Tikuisis, P.; Buick, F.; Kane, D. M. (1987) Percent carboxyhemoglobin in resting humans exposed repeatedly to
              1,500 and 7,500 ppm CO. J. Appl. Physiol. 63: 820-827.
 20
       Traynor, G. W.; Allen, J. R.; Apte, M. G.;  Dillworth, J. F.; Girman, J. R.; Hollowell, C.  D.; Koonce, J. F.,
              Jr. (1982)  Indoor air pollution from portable kerosene-fired space heaters, wood-burning stoves, and
              wood-burning furnaces.  In: Proceedings of the Air Pollution Control Association specialty conference on
              residential wood and coal combustion; March; Louisville, KY. Pittsburgh, PA: Air Pollution Control
 25           Association; pp. 253-263.

       Tsukamoto, H.; Matsuda, Y. (1985) [Relationship between smoking and both the carbon monoxide concentration
              in expired  air and the carboxyhemoglobin content]. Kotsu Igaku 39: 367-376.

 30    Turner, J. A.  M.;  McNicol, M. W.; Sillett, R.  W. (1986) Distribution of carboxyhaemoglobin concentrations in
              smokers and non-smokers. Thorax 41: 25-27.

       U. S. Environmental Protection Agency. (1979) Air quality criteria for carbon monoxide. Research Triangle
              Park, NC:  Environmental Criteria and Assessment Office; EPA report no. EPA-600/8-79-022. Available
 35           from:  NTIS, Springfield, VA; PB81-244840.

       van Kampen, E. J.; Zijlstra, W. G. (1961) Standardization of hemoglobinometry. II. The hemiglobincyanide
              method. Clin. Chim. Acta 6: 538-544.

 40    Van Netten, C.; Brubaker, R. L.; Mackenzie, C. J. G.; Godolphin, W. J. (1987) Blood lead and
             carboxyhemoglobin levels in chainsaw operators. Environ. Res. 43: 244-250.

       Verhoeff, A. P.; van der Velde, H. C. M.; Boleij, J. S. M.; Lebret, E.; Brunekreef, B. (1983) Detecting indoor
             CO exposure by measuring CO in exhaled breath.  Int. Arch. Occup. Environ.  Health 53: 167-174
45
       Virtamo, M.; Tossavainen, A. (1976) Carbon monoxide in foundry air. Scand. J. Work Environ. Health 2(suppl.
              1): 37-41.

       Vreman, H. J.; Kwong, L. K.; Stevenson, D. K. (1984) Carbon monoxide in blood: an improved microliter
50          blood-sample collection system, with rapid analysis by gas chromatography. Clin. Chem.
             (Winston-Salem, NC) 30: 1382-1386.

       Vreman, H. J.; Stevenson, D. K.; Zwart, A.  (1987) Analysis for carboxyhemoglobin by gas chromatography and
             multicomponent spectrophotometry compared. Clin. Chem. (Winston-Salem, NC) 33: 694-697.


       March 12,  1990                             8-115     DRAFT-DO NOT QUOTE OR CITE

-------
      Wade, W. A., Ill; Cote, W. A.; Yocom, J. E. (1975) A study of indoor air quality. J. Air Pollut. Control
              Assoc. 25: 933-939.

      Wald, N. J.; Idle, M.; Boreham, J.; Bailey, A. (1981) Carbon monoxide in breath in relation to smoking and
 5            carboxyhaemoglobin levels. Thorax 36: 366-369.

      Wallace, L. A. (1983) Carbon monoxide in air and breath of employees in an underground office. J. Air Pollut.
              Control Assoc. 33: 678-682.

10    Wallace, L. A.; Ott, W. R. (1982) Personal monitors: a state-of-the-art survey. J. Air Pollut. Control Assoc.
              32: 601-610.

      Wallace, L. A.; Ziegenfus,  R. C. (1985) Comparison of carboxyhemoglobin concentrations in adult nonsmokers
              with ambient carbon monoxide levels. J. Air Pollut. Control Assoc. 35: 944-949.
15
      Wallace, N. D.; Davis, G. L.; Rutledge, R. B.; Kahn. A. (1974) Smoking and carboxyhemoglobin in the St.
              Louis metropolitan population: theoretical and empirical considerations. Arch. Environ.  Health
              29:  136-142.

20    Wallace, L. A.; Thomas, J.; Mage, D. T.  (1984)  Comparison of end-tidal  breath CO estimates of COHb with
              estimates based on exposure profiles of individuals in the Denver, Colorado and Washington, DC area.
              Research Triangle Park,  NC: U. S. Environmental Protection Agency, Environmnetal Monitoring
              Systems Laboratory; EPA report no. EPA-600/D-84-194.  Available from: NTIS, Springfield, VA;
              PB84-229822.
25
      Wallace, L. A.; Pellizzari, E. D.; Hartwell, T.  D.; Sparacino, C. M.; Sheldon, L. S.; Zelon, H. S. (1985)
              Results from the first three seasons of the  TEAM Study: personal exposures, indoor-outdoor
              relationships, and breath levels of toxic air pollutants measured for 355 persons in New  Jersey. Presented
              at:  78th annual meeting of the Air Pollution Control Association; June; Detroit, MI. Pittsburgh, PA:  Air
30            Pollution  Control Association; paper no. 85-31.6.

      Wallace, L.; Thomas, J.; Mage, D.; Ott, W. (1988) Comparison of breath CO, CO exposure, and Coburn model
              predictions in the U.S. EPA Washington-Denver (CO) study. Atmos. Environ. 22: 2183-2193.

35    Wigfield, D. C.; Hollebone, B. R.; Mackeen, J. E.; Selwin, J. C. (1981) Assessment of the methods available
              for the determination of carbon monoxide  in blood. J. Anal. Toxicol. 5:  122-125.

      Woodman, G.; Wintoniuk, D. M.; Taylor, R.  G.; Clarke, S. W.  (1987) Time course of end-expired carbon
              monoxide concentration is important hi studies of cigarette smoking. Clin. Sci. 73: 553-555.
40
      Wright, G. R.; Jewczyk, S.; Onrot, J.; Tomlinson, P.; Shephard, R. J.  (1975) Carbon monoxide in the urban
              atmosphere: hazards to the pedistrian and street worker. Environ. Health 30: 123-129.

      Yocom, J.  E. (1982) Indoor-outdoor air quality relationships: a critical review. J. Air Pollut. Control Assoc.
45            32: 500-520.

      Yocom, J.  E.; Clink, W. L.; Cote, W. A. (1971) Indoor/outdoor air quality relationships. J. Air Pollut. Control
              Assoc. 21: 251-259.

50    Ziskind, R. A.; Rogozen, M. B.; Carlin, T.; Drago,  R. (1981) Carbon monoxide intrusion into sustained-use
              vehicles. Environ. Int. 5: 109-123.

      Ziskind, R. A.; File,  K.; Mage, D. T. (1982)  Pilot field study: carbon monoxide exposure monitoring in the
              general population.  Environ. Int. 8: 283-293.


        March 12, 1990                              8-116      DRAFT-DO NOT QUOTE OR CITE

-------
      Zwart, A.; van Kampen, E. J. (1985) Dyshaemologlobins, especially carboxyhaemoglobin, levels in hospitalized
            patients. Clin. Chem. (Winston-Salem, NC) 31: 945.

      Zwart, A.; Buursma, A.; Zijlstra, W. G. (1984) Multicomponent analysis of hemoglobin derivatives by means of
 5          a reversed optics spectrophotometer. Clin. Chem. (Winston-Salem, NC) 30: 373-379.
      Zwart, A.; Van Kampen, E. J.; Zijlstra, W. G. (1986) Results of routine determination of clinically significant
            hemoglobin derivatives by multicomponent analysis. Clin. Chem. (Winston-Salem, NC) 32: 972-978.
10
       March 12, 1990                            8-117      DRAFT-DO NOT QUOTE OR CITE

-------
   9.  PHARMACOKINETICS AND MECHANISMS OF ACTION
                           OF CARBON MONOXIDE

           Pharmacokinetics in the classical sense has been concerned primarily with the
 5     determination of blood levels for various dosage regimens of pharmacological agents. More
       modern approaches tend to extend the definition to include other aspects of substance kinetics.
       For example, it may include membrane diffusion, substance binding and release
       characteristics, modeling, metabolic pathways, and other processes. The general tendency is
       to depart from anatomically or physiologically defined region(s) to a more encompassing and
10     unifying concept of compartment(s) comprised of real as well as abstract constructs (Bischoff,
       1986). It will be in this sense that the CO pharmacokinetics will be approached and presented
       in this chapter.
15     9.1  ABSORPTION, DISTRIBUTION, AND PULMONARY
           ELIMINATION
       9.1.1  Introduction
           The scope of this chapter and generally of this criteria document does not allow for an
       extensive review of the mechanisms and factors involved in CO uptake and elimination. The
20     review will concentrate on fundamental processes and the key factors affecting CO
       metabolism and resultant effects.  For a more in-depth explanation of certain facets of CO
       toxicity the reader is referred to other chapters of this document and to other review material
       (Fishmanetal., 1987).

25     9.1.2  Pulmonary  Uptake
       9.1.2.1  Mass Transfer of Carbon Monoxide
           Although CO is not one of the respiratory gases, the similarity of the physicochemical
       properties of CO and O2 permit an extension of the findings of studies on the kinetics of
       transport of O2 to that of CO.
30


       March 15, 1990                        9-1     DRAFT - DO NOT QUOTE OR CITE

-------
            The rate of formation and elimination of COHb, its concentration in blood, as well as its
       catabolism is controlled by numerous physical and physiological mechanisms. The relative
       contribution of these mechanisms to the overall COHb kinetics will depend on the
       environmental conditions (ambient CO concentration, altitude, etc.), physical activity of an
 5     individual, and many other physiological processes, some of which are complex and still
       poorly understood.
            The mass transport of CO between the airway opening (mouth and nose) and red blood
       cell (hemoglobin) is predominantly controlled by physical processes.  The CO transfer to the
       Hb-binding sites is accomplished in two sequential steps: (1) transfer of CO in a gas phase,
10     between the airway opening and the alveoli, and (2) transfer in a "liquid" phase, across  air-
       blood interface including the red blood cell (RBC). While the mechanical action of the
       respiratory system and the molecular diffusion within the alveoli are the key mechanisms of
       transport in the gas phase, the diffusion of CO across the alveolo-capillary barrier, plasma,
       and RBC is the virtual mechanism of the liquid phase.
15
       9.1.2.2 Effects of Dead Space and Uneven Distribution of Ventilation and Perfusion
            Ideally, the optimal transfer of gases across alveolo-capillary membrane can be achieved
       only if regional distribution of ventilation is uniform and matches regional blood flow.
       Numerous studies have shown that in the upright subject ventilation is preferentially
20     distributed to the lower lung zones (Milic-Emili et al., 1966). Besides posture (Clarke et al.,
       1969), changes in resting lung volume (Sutherland et al., 1968), airway resistance (Hughes
       et al.,  1972), and lung compliance (Glaister et al., 1973) by either exogenous factors or
       pathophysiological conditions will aggravate maldistribution of ventilation. The unevenness is
       further affected by inspiratory (Anthonisen et al., 1970) and expiratory flow  rates (Millette
25     et al.,  1969), which influence sequential filling and emptying of the lung regions.
            Even in perfectly healthy subjects the homogeneity of ventilation, perfusion,  and
       consequent VA/Q ratio of unity is unattainable because of a right-to-left shunt. Normally, only
       a small amount of mixed venous blood (2 to 4%) bypasses the alveoli and reaches systemic
       circulation without oxygenation. Any increase in the alveolo-arterial O2 gradient (A-a DOj)
30     will contribute to hypoxemia, thus enhancing CO loading (Riley and  Permutt, 1973). It
       follows that any imbalance in the distribution patterns of these two compartments must result

       March 15, 1990                           9-2     DRAFT - DO NOT QUOTE OR CITE

-------
       in a decrease in the efficiency of gas exchange (Scrimshire,  1977) including CO.  The
       average ventilation to perfusion ratio of about 0.9 reported in the upright  subjects indicates
       that overall perfusion (Q) exceeds ventilation (V) ; regional nommiformity, however, is
       considerably greater  (the VA/Q ratios range from 0.6 to 3.0;  Inkley and Maclntyre, 1973).
 5     Consequently, in the underventilated but overperfused regions of the lung  the amount of CO
       available for diffusion will be less than if the ventilation and perfusion were matched, while
       in the overventilated  but underperfused regions the amount of CO that could diffuse would be
       the same as if the distributions were matched.  On exercise, when the distribution of both
       ventilation and perfusion becomes more uniform, the ratios approach unity and the rate of
10     COHb formation will accelerate (Harf et al., 1978).
             Besides regional inhomogeneity of distribution, the bulk movement of inhaled air will be
       influenced by factors related to inspiratory flow and subsequent mixing with residual air. At
       rest,  mixing of gases is almost complete and no discernible stratification of concentration
       between the large airways and the alveoli occurs.  However, any changes  in ventilation or
15     pattern of breathing (e.g., during exercise) will aggravate stratified inhomogeneity and
       increase a concentration gradient between central and peripheral airways.  The relative effects
       of ventilation and perfusion inhomogeneities on convectional and diffusional transport of CO
       will very much depend on the rate of change and concentration of CO in inspired air.  The
       higher the concentration and the shorter the rise time of CO in the inspired air, the greater the
20     effects these factors will have on the CO uptake and ultimately COHb concentration in blood.
             The ventilation-perfusion unevenness will not only contribute to hypoxemia, but the
       mismatch will influence the size of the physiological dead space (VD) (Standfuss, 1970) and
       ultimately alveolar ventilation, which is one of the principal, but seldom-measured
       determinants of the rate of uptake of CO (see Section 9.3 on Coburn-Forster-Kane [CFK]
25     modeling).  Any increase in a dead space to tidal volume ratio (VD /VT) will decrease VA and
       vice versa.  In normal healthy subjects at rest VD comprises  about 25 to 45%  of tidal volume;
       in older subjects or in patients with pulmonary disease the percentage might be as high as
       70%  (Martin etal., 1979).
30
        March 15, 1990                          9.3      DRAFT - DO NOT QUOTE OR CITE

-------
        9.1.2.3 Alveolo-Capillary Membrane and Blood-Phase Diffusion
             While the above mechanisms controlling the rate of formation of blood COHb are
        predominantly active processes, the second key mechanism, a diffusion of gases across the
        alveolar air-hemoglobin barrier, is an entirely passive process. In order to reach the
 5      Hb-binding sites, the CO and other gas molecules have to pass across the alveolo-capillary
        membrane, diffuse through the plasma, pass across the RBC membrane, and finally the RBC
        stroma before reaction between CO and Hb can take place. The molecular transfer across the
        membrane and the blood phase is governed by general physicochemical laws, particularly by
        the Pick's first law of diffusion.  The exchange and equilibration of gases between the two
10      compartments (air and blood) is very rapid.  The dominant driving force is a partial pressure
        differential of CO across this membrane.  For example, inhalation of a bolus of air containing
        high levels of CO will rapidly increase blood COHb; by immediate and tight binding of CO
        to Hb the partial pressure of CO within the RBC is kept low,  thus maintaining a high pressure
        differential between air and blood, and consequent diffusion of CO into blood.  Subsequent
15      inhalation of CO-free air progressively decreases the gradient  to the point of its reversal
        (higher CO pressure on the blood side than alveolar air) and CO will be released into alveolar
        air.  Because binding of CO to Hb is a much stronger and considerably faster reaction (half-
        time  <0.07) than clearance of CO by ventilation, the air-blood pressure gradient is usually
        higher than the blood-air gradient, and the CO uptake will be  a proportionally faster process
20      than CO elimination.  The rate of CO release will be further affected by the products of tissue
        metabolism.  Under pathologic conditions, where one or several components of the air-blood
        interface might be severely  affected,  as in emphysema, fibrosis, or edema,  both the uptake
        and elimination of CO will be affected.
             The rate of diffusion of gases might be altered considerably by many physiological
25      factors acting concomitantly.  Diurnal variations in CO diffusion related to variations in Hb
        have been reported in normal healthy subjects (Frey et al., 1987). Others found the changes
        to be related also to physiological factors such as O2Hb, COHb, partial pressure of alveolar
        CO2,  ventilatory pattern,  oxygen consumption (VOJ, blood flow, functional residual capacity,
        etc. (Forster, 1987).  It has been confirmed repeatedly that diffusion is body-position and
30      ventilation dependent. In a supine position at rest, CO diffusion has been significantly higher
        than that at rest in a sitting position.  In both positions CO diffusion during exercise has been

        March 15,  1990                          9-4     DRAFT - DO NOT QUOTE OR CITE

-------
       greater than at rest (McClean et al.,  1981).  CO diffusion will increase with exercise, and at
       maximum work rates the diffusion will be maximal regardless of position.  This increase is
       attained by increases in both the membrane-diffusing component and the pulmonary capillary
       blood flow (Stokes et al., 1981). Diffusion seems to be relatively independent of lung
 5     volume within the midrange of vital  capacity.  However, at extreme volumes the differences
       in diffusion rates could be significant; at total lung capacity the diffusion is higher, while at
       residual volume it is lower than the average (McClean et al., 1981). Smokers showed on the
       average lower diffusion rates than nonsmokers (Knudson et al., 1989).
            The above physiological processes will affect minimally COHb formation in healthy
10     individuals exposed to low and relatively uniform levels of CO.  Under such ambient
       conditions these factors will be the most influential during  the initial period of CO distribution
       and exchange.  If sufficient time is allowed for equilibration, the sole determinant of COHb
       concentration in blood will be the ratio of CO to O2.  However,  the shorter the half-time for
       equilibration (e.g., due to hyperventilation, high concentration of CO, increased cardiac
15     output, etc.) the more involved these mechanisms will become in modulating the rate of CO
       uptake (Pace et al.,  1950; Coburn et al., 1965).  At high transient CO exposures of resting
       individuals both the cardiac and the lung function mechanisms will control the rate of CO
       uptake. Incomplete mixing of blood might result in a  substantial difference between the
       arterial and venous COHb concentrations (Godin and Shephard,  1972).  In chronic
20     bronchitics, asthmatics, and other subpopulations at risk (pregnant women,  the elderly, etc.)
       the kinetics of COHb formation will be even more complex, because any abnormalities of
       ventilation  and perfusion and gas diffusion will aggravate CO exchange (see Chapter 12 for
       details on subpopulations at risk).

25     9.1.3 Tissue Uptake
            Distribution of CO within the tissue(s) will be determined primarily by exchange and
       chemical reaction kinetics.  In order  to facilitate understanding of these well integrated
       processes it would be helpful to consider CO uptake by the most involved physiological
       compartments/organs.
30
       March 15, 1990                          9-5      DRAFT - DO NOT QUOTE OR CITE

-------
       9.1.3.1 The Blood
            Although the rate of CO binding with Hb is about 1/5 slower and the rate of dissociation
       from Hb is an order of magnitude slower than the respective rates for O2, the CO chemical
       affinity (M) for Hb is about 245 (240 to 250) times greater than that of O2 (Roughton, 1970).
 5     One part of CO and 245 parts of O2 would form equal parts of O2Hb  and COHb (50% of
       each) which would be achieved by breathing air containing 21% oxygen and 570 ppm CO.
       Moreover,  under steady-state conditions (gas exchange between blood and atmosphere
       remains constant), the ratio of COHb to OzHb is proportional to the ratio of their respective
       partial pressures.  The relationship between the affinity constant M and PO2 and PCO first
10     expressed by Haldane (1898), has the following form.

                                   COHb/O2Hb = M * (PCO/PO,)                        (9-1)

       At equilibrium, when Hb is maximally saturated by O2 and CO at their respective gas
15     tensions, the M value for all practical purposes is independent of pH and 2,3-DPG over a
       wide range of PCO/PO2 ratios.  The M, however, is temperature dependent (Wyman et al.,
       1982).
            Under dynamic conditions competitive binding of O2 and CO to Hb is complex; simply
       said, the greater the number of hemes bound to CO, the greater is the affinity of free hemes
20     for O2.  Any decrease in the amount of  available Hb for O2 transport  (CO poisoning,
       bleeding, anemia, blood diseases, etc.) will reduce the quantity of O2 carried by blood to  the
       tissue.  However, CO not only occupies O2-binding sites, molecule for molecule, thus
       reducing the amount of available O2, but also alters characteristic relationships between O2Hb
       and PO2 which in normal blood is S-shaped.  With increasing concentration of COHb in
25     blood, the dissociation curve  is shifted gradually to the left and its shape is transformed into
       that of a rectangular hyperbola (Figure 9-1).  Because the shift occurs over a critical
       saturation range for release of O2 to tissues, a reduction in O2Hb by CO poisoning will have
       more severe effects on the release of O2 than the equivalent reduction in Hb due to anemia.
       Thus, in an anemic patient (50%) at the tissue PO2 of 26 torr (v',), 5  vol % of O2 (50%
30     desaturation) might be extracted from blood, the amount sufficient to sustain tissue
       metabolism.  In contrast, in a person poisoned with CO (50% COHb), the tissue PO2 will

       March 15,  1990                          9-6      DRAFT - DO NOT QUOTE OR CITE

-------
                                  VOLUME PERCENT OXYGEN. mL/100 ml blood
v
                                        PERCENT Hb SATURATION

-------
       have to drop to 16 ton (v'2; severe hypoxia) to release the same, 5 vol % O2 (Figure 9-1).
       Any higher demand on oxygen under these conditions (e.g., by exercise) might result in coma
       of the CO-poisoned individual.

 5     9.1.3.2  The Lung
            Although the lung in its function as a transport system for gases is exposed continuously
       to CO,  very little CO actually diffuses and is stored in the lung tissue itself, except for the
       alveolar region.  The epithelium of the conductive zone (nasopharynx and large airways)
       presents a significant barrier to diffusion of CO (Guyatt et al., 1981).  Therefore, diffusion
10     and gas uptake by the tissue, even at very high CO concentrations, will be very slow; most of
       this small amount of CO will be dissolved in the mucosa of the airways. Diffusion into the
       submucosal layers and interstitium will depend very much on the concentration of CO and
       duration of exposure. Experimental exposures of the oronasal cavity of monkeys to very high
       concentrations of CO for a very short period of time increased their blood COHb level to only
15     1.5%.  Comparative exposures of the whole lung, however, elevated COHb to almost 60%
       (Schoenfisch et al., 1980).  Thus diffusion of CO across the airway mucosa will contribute
       very little if at all to overall COHb concentration.  In the transitional zone (<20th generation)
       where both conductive  and diffusive transport take place, diffusion of CO into lung
       interstitium will be much easier, and at times more complete.  In the respiratory zone
20     (alveoli), which is the most effective interface for CO transfer, diffusion into lung interstitium
       will be  complete.  Because the total lung tissue mass is rather small compared to other CO
       compartments, relatively small amount of CO (primarily as dissolved CO) will be distributed
       within the  lung structures.

25     9.1.3.3  Heart and Skeletal Muscles
             The role of myoglobin in O2 transport is not yet fully understood. Myoglobin (Mb)  as a
       respiratory hemoprotein of muscular tissue will undergo a reversible reaction with CO in a
       manner similar to O2.  The greater affinity of O2 for myoglobin than hemoglobin (hyperbolic
       versus S-shaped dissociation curve) is in this instance physiologically beneficial because a
30     small drop in tissue PO2 will release a large amount of O2 from oxymyoglobin (O2Mb). The
        March 15, 1990                           9-8      DRAFT - DO NOT QUOTE OR CITE

-------
        main function of Mb is thought to serve as a temporary store of O2 and act as a diffusion
        facilitator between hemoglobin and the tissues (for details see Section 9.4.2).
             Myoglobin has an affinity constant approximately eight times lower than hemoglobin
        (M=20 to 40 vs.  245, respectively).  As with Hb, the combination velocity constant between
 5      CO and Mb is only slightly lower than for O2, but the dissociation velocity constant is much
        lower than for O2. The combination of greater affinity (Mb is 90% saturated at PO2 of
        20 mmHg) and lower dissociation velocity constant for CO favors retention of CO in the
        muscular tissue. Thus, a considerable amount of CO potentially can be stored in the skeletal
        muscle. The ratio of carboxymyoglobin (COMb)  to COHb saturation for skeletal muscle of a
10      resting dog and cat has been determined to be 0.4 to 0.9; for cardiac muscle the ratio is
        slightly higher (0.8 to 1.2) (Coburn et al., 1973; Sokal et al., 1986). Prolonged exposures
        did not change this ratio in either muscle indicating that certain, not yet identified,
        mechanisms prevented equilibrium between the vascular and extravascular compartments
        (Sokal et al., 1984). During exercise the relative  rate of CO binding increases more for Mb
15      than for Hb and CO will diffuse from blood to skeletal muscle (Werner and Lindahl, 1980);
        consequently, the COMb/COHb will increase for both skeletal and cardiac muscles (Sokal
        et al., 1986). A similar shift in CO has been observed under hypoxic conditions because a
        fall in intracellular PO2 below a critical level will  increase the relative affinity of Mb to CO
        (Coburn et al.,  1971). Consequent reduction in Mb-carrying capacity of O2 might have a
20      profound effect on the supply of O2 to the tissue (see Section 11.1).

        9.1.3.4 Brain  and Other Tissues
             Apart from Hb and Mb,  which are the largest stores of CO, other hemoproteins will
        react with CO.  However, the exact role of such compounds on O2-CO kinetics still needs to
25      be ascertained (see Section 9.4).  Concentration of CO in brain tissue has been found to be
        about 30 to 50 times lower than that in blood. During the elimination of CO from brain the
        above ratio of concentrations was still maintained  (Sokal et al., 1984).  (For a more in-depth
        discussion see Chapter 10, Section 10.4.)
30
       March 15, 1990                          9-9      DRAFT - DO NOT QUOTE OR CITE

-------
       9.1.4 Pulmonary and Tissue Elimination
            An extensive amount of data available on the rate of CO uptake and the formation of
       COHb contrast sharply with the limited information available on the dynamics of CO washout
       from body stores and blood.  Although the same factors that govern CO uptake will affect CO
 5     elimination, the relative importance of these factors might not be the same (Landaw, 1973;
       Petersen  and Stewart, 1970). Both the formation as well as the decline of COHb fit a second-
       order function best, increasing during the uptake period and decreasing during the elimination
       period.  Hence, an initial rapid decay will gradually slow down (Landaw, 1973; Wagner
       et al., 1975; Stewart et al., 1970).  The elimination rate of CO from an equilibrium state will
10     follow a  monotonically decreasing second-order (logarithmic or exponential) function (Pace
       et al., 1950).  The rate, however, might not be constant following transient exposures to CO,
       whereas at the end of exposure the steady-state conditions were not reached yet. In this
       situation, particularly after very short and high CO exposures, it is possible that COHb
       decline could be biphasic and it can be approximated best by a double-exponential function:
15     The initial rate of decline or "distribution" might be considerably faster than the later
       "elimination" phase (Wagner et al., 1975).  Reported divergence of COHb decline rate in
       blood and in exhaled air  suggests that CO elimination rate(s) from extravascular pool(s) is
       (are) slower than that reported for blood (Landaw, 1973). Although the absolute elimination
       rates are associated positively with the initial concentration of COHb, the relative elimination
20     rates appear to be independent of the initial concentration of COHb (Wagner et al., 1975).
            The half-time of CO disappearance from blood under normal recovery conditions while
       breathing air showed considerable between-individual variance.  For COHb concentration of 2
       to 10%, the half-time ranged from 3 to 5 h (Landaw, 1973); others reported the range to be 2
       to 6.5 h for slightly higher initial concentrations of COHb (Petersen and Stewart, 1970).
25     Increased inhaled concentration of oxygen accelerated elimination of CO; by breathing 100%
       oxygen the half-time was shortened by almost 75% (Petersen and Stewart, 1970).  The
       elevation of partial pressure of oxygen to three atmospheres reduced the half-time to about
       20 min, which is approximately a 14-fold decrease over that seen when breathing room air
       (Britten and Myers, 1985; Landaw, 1973).  Although the washout of CO can be somewhat
30     accelerated by an admixture of 5% CO2 in O2, hyperbaric O2 treatment is more effective in
       facilitating displacement  of CO.

       March 15, 1990                         9-10     DRAFT - DO NOT QUOTE OR CITE

-------
       9.2  TISSUE PRODUCTION AND METABOLISM OF CARBON
            MONOXIDE
            In the process of natural degradation of hemoglobin to bile pigments, a carbon atom (a-
       bridge C) is separated from the porphyrin nucleus and subsequently is catabolized by
 5     microsomal heme oxygenase into CO. The major site of heme breakdown and therefore the
       major production organ of endogenous CO is the liver (Berk et al., 1976).  The spleen and
       the erythropoietic system are other important catabolic generators of CO.  Because the amount
       of porphyrin breakdown is stoichiometrically related to the amount of endogenously formed
       CO, the blood level of COHb or the concentration of CO in the alveolar air has been used
10     with mixed success as quantitative indices of the rate of heme catabolism (Landaw and
       Callahan, 1970; Solanki et al.,  1988). Not all of endogenous CO comes from RBC
       degradation.  Other hemoproteins,  such as myoglobin, cytochromes, peroxidases, and catalase
       contribute approximately 20 to 25% to the total amount of generated CO (Berk et al., 1976).
       Approximately 0.4 mL/h of CO is  formed by hemoglobin catabolism and about 0.1 mL/h
15     originates from nonhemoglobin sources (Coburn et al., 1964).  Metabolic processes other than
       heme catabolism contribute only a very small amount of CO (Miyahara and Takahashi,  1971).
       In both males and females, week-to-week variations of CO production are greater than day-
       to-day or within-day variations.  Moreover, in females COHb levels fluctuated with the
       menstrual cycle; the mean  rate of CO production in the premenstrual, progesterone phase
20     almost doubled (Lynch and Moede, 1972; Delivoria-Papadopoulos et al., 1970). Neonates
       and pregnant women also showed a significant increase in endogenous CO production related
       to increased breakdown of RBC.
            Any disturbance leading to increased destruction of RBC  and accelerated breakdown of
       other hemoproteins would lead to increased production of CO. Hematomas, intravascular
25     hemolysis of RBC, blood transfusion, and ineffective erythropoiesis  all will elevate CO
       concentration in blood.  Degradation of RBC under pathologic conditions such as anemias
       (hemolytic, sideroblastic, sickle cell), thalassemia, Gilbert's syndrome with hemolysis, and
       other hematological diseases also will accelerate CO production (Berk et al., 1974;  Solanki
       et al., 1988).  In patients with hemolytic anemia the CO production  rate was 2 to 8 times
30     higher, and blood COHb concentration 2 to 3 times higher than in normals (Coburn et al.,
       1966). Increased CO-production rates have been reported after administration of

       March 15, 1990                         9-11      DRAFT - DO NOT QUOTE OR CITE

-------
        phenobarbital, diphenylhydantoin (Coburn, 1970), and progesterone (Delivoria-Papadopoulos
        etal., 1970).
 5      9.3 MODELING CARBOXYHEMOGLOBIN FORMATION
        9.3.1  Introduction
            The NAAQS for CO were designed to establish ambient levels of CO which would
        protect sensitive individuals from experiencing adverse health effects.  In retaining the current
        CO primary standards, both EPA and the Clean Air Scientific Advisory Committee concluded
 10      that the critical effects level for NAAQS-setting purposes was approximately 3% COHb
        without including  a margin of safety (Federal Register, 1985).  Using exposure modeling and
        available monitoring data, EPA estimated that the current 9 ppm, eight-hour average standard
        would keep more than 99.9% of the adult population with cardiovascular disease below 2.1%
        COHb. Considering uncertainties regarding the lowest level at  which adverse health effects
 15      may occur, as well as uncertainties about the exposure estimates, EPA concluded that this
        level of protection would provide an adequate margin of safety  for this  sensitive group.
        Because of the variability of ambient CO concentration profiles, and other exogenous and
        endogenous factors affecting formation of COHb in an individual, it is obvious that the only
        practical approach to evaluate the protection provided by these standards is to continue to use
20      mathematical models. The COHb formation modeling, however, has much wider application
        because the quantification of the relationship between exogenous CO and blood COHb is also
        of clinical and occupational interest.

        9.3.2  Regression Models
25          The most direct approach to establishing a prediction equation for COHb is to regress
        observed COHb values against the level and duration of exogenous CO  exposure. Inclusion
        of other predictor variables such as initial COHb level and alveolar ventilation generally will
        improve the precision of the predictions. All regression models are purely empirical and have
        no physiological basis. Their applicability therefore is limited to the exact conditions that
30      were used to collect the data on which they are based. So far, the most viable models have
        been tested and used to estimate COHb levels for healthy subjects only.  No validation studies

        March 15, 1990                          9-12     DRAFT  - DO NOT QUOTE OR CITE

-------
        have been reported on potentially at-risk subpopulations (see Chapter 12), such as patients
        with cardiovascular or hematologic dysfunction.
            Peterson and Stewart (1970) developed regression Equation 9-2 for percent COHb after
        exposure to moderate CO levels, where CO refers to the concentration of CO in ambient air
 5      inhaled parts per million, t is the exposure duration in minutes, and t' is the postexposure
        time in minutes. The final term (-.000941') reflects CO elimination and was computed using

            Log10 %COHb  =  .85753 Log,0 CO + .62995 Log,01 - 2.29519 -.000941'          (9-2)

10      the average COHb half-life found in the study.  The percent COHb in the blood samples was
        determined twice, using an IL CO-Oximeter and a gas chromatograph.  The percent COHb
        values that were used to estimate the equation were themselves averages over observations on
        2 to 10 subjects (r = 0.985). The range of CO concentrations used was 25 to 523 ppm CO,
        and  the exposures lasted from 15 min to eight hours. The subjects were 18 healthy males that
15      did not smoke during the duration of the study.  More recently Equation 9-2, without its final
        term, was modified by Zankl (1981) to correct the time, t,  in the equation for altitude and
        subject activity level. No justification, however, nor reference was cited for these changes.
            Another regression equation (9-3) developed by Stewart et al. (1973) applies to briefer
        exposures of considerably higher levels of CO.  In this study the exposures  ranged from
20      1000 ppm (for 10 min) to 35,600 ppm (for 45 sec).  The regression equation was based on
                      Log10[%COHb(t)] = Log10[%COHb(t)] + Log10[%COHb(0]           (9-3)
                                          + 1.036 Log10 CO-4.4793
25                                        + Log10 (liters inhaled)
        13 experimental exposures but only on six different subjects (r = 0.995). The subjects
        remained sedentary throughout the study.  Possible correlations between readings on the same
30      subject were not taken into account.  The predicted quantity is the logarithm of the "increase
        in percent COHb saturation in venous blood per liter of CO mixture inhaled." The percent
        COHb in the blood samples was determined twice, using automated blood analyzing system
        and a gas chromatograph.  The increase in COHb saturation was computed using the peak

        March 15, 1990                          9-13    DRAFT - DO NOT QUOTE OR CITE

-------
       COHb concentration occurring approximately two minutes after CO exposure stopped.
       However, the immediate postexposure inhalation of pure O2 almost certainly lowered the peak
       COHb values and influenced subsequent estimates.

 5     9.3.3  The Coburn-Forster-Kane Differential Equations
            In 1965, Coburn, Forster, and Kane developed a differential equation to describe the
       major physiological variables that determine blood [COHb] for the examination of the
       endogenous production of CO.  The equation, referred to as the CFK model, is still much in
       use today for the prediction of [COHb] consequent to inhalation of CO for two reasons.
10     First, the model is quite robust to challenges to the original assumptions.  Second, the model
       can be relatively easily adapted to more specialized applications.

       9.3.3.1  Linear and Nonlinear CFK Differential Equations
            Equation 9-4 represents the linear CFK model, with constant O2Hb level:
15
          VB d[COHb]/dt = Vco - [COHb]PcO2 / MB[02Hb] + P.CO/B                      (9-4)
       where:
20               [COHb] = milliliters of CO per milliliter of blood, maximum O2 capacity of
                           blood
                  [O2Hb] = milliliters of O2 per milliliter of blood (=0.2)
25                    Hb = grams hemoglobin per milliliter of blood, hemoglobin concentration in
                           blood (=1.38)
                       B =  1/DLCO + PL/VA
30                 DLCO = milliliters per minute per millimeter of Hg, pulmonary diffusing
                           capacity for CO (=30)
                      PL = millilmeters of Hg, pressure dry gases in the lungs (=713)
35                    VA =  milliliters per minute alveolar ventilation rate (=6000)
                      M = Haldane affinity ratio (=218)
                      VB =  millimeters of blood volume (=5500)
       March 15, 1990                         9-14     DRAFT - DO NOT QUOTE OR CITE

-------
                      v
                       co = milliliters per minute, endogenous CO production (=0.007)
                    PIcO = millimeters of Hg, partial pressure CO in air inhaled
                    PCO2 = millimeters of Hg, average partial pressure of O2 in lung capillaries
                            (=100)
10     The values in parentheses are the values given in Peterson and Stewart (1970), although it is
       not clear whether a consistent set of conditions (i.e., BTPS or STPD) was used.   Restricting
       the conditions to low CO exposures allows the mathematical assumption of instant equili-
       bration of (1) the gases in the lungs, (2) COHb concentrations between venous and arterial
       blood, and (3) COHb concentrations between the blood and CO stores in nonvascular tissues.
15     In addition, the washout  time becomes unimportant, and the inhaled and exhaled volumes
       could be presumed equal. In solving Equation 9-4, Coburn, Forster, and Kane (1965) further
       assumed that [O2Hb] is constant and not dependent on [COHb]. The resulting linear
       differential equation is restricted to relatively low COHb levels. For high ambient CO levels,
       it may erroneously predict equilibrium values greater than 100% COHb.
20          The advantage to using the linear differential equation (where applicable) is that the
       solution can be written explicitly as.

                                [COHb] (t) = [COHb]0e'At + C/A (1  - e'*)                (9-5)
       where
25
                                         A = PCO2 /V, MB [O, Hb]
                                         C = VCO/VB + P, CO /V, B
       From this solution, we see that for small t, the formation of COHb proceeds linearly, as
30
                                         A [COHb] * P, CO (t) /V, B                    (9-6)
       March 15, 1990                          9-15    DRAFT - DO NOT QUOTE OR CITE

-------
            Since O2 and CO combine with Hb from the same pool, higher COHb values do affect
       the amount of Hb available for bonding with O2. Such interdependence can be modeled by
       substituting (1.38 Hb - [COHb]) for [OjHb]) (see e.g., Tikuisis et al., 1987b).  The CFK
       differential equation then becomes nonlinear and iterative methods or numerical integration
 5     must be used to solve the equation (Muller and Barton, 1987).  Solutions of either CFK
       equation require that the volumes of all gases be adjusted to the same conditions.  Coburn,
       Forster, and Kane (1965) use STPD conditions, but the equation can be solved under any
       conditions if consistently used (Tikuisis et al., 1987a,b).
            The equilibrium value predicted by the nonlinear differential equation will always be
10     less than 100% COHb, and is given by the following expression.
                   [COHb] = 1.38 Hb M(P,CO + BVo,) / PCO2 + M (P,CO +BVCO          (9-7)

            A sensitivity analysis has been done on the parameters of both the linear and nonlinear
15     CFK equation at five different work levels (McCartney, 1990). The author shows that a 1%
       error in any one of the parameters produces  no more than a 1 % error in COHb prediction by
       the nonlinear model.
            The nonlinear CFK model is more accurate physiologically,  but has no explicit solution.
       It is reasonable, therefore, to ask under what conditions the solutions to the linear and
20     nonlinear equations are "close" together.  Because both solutions are generated by known
       differential equations, the question is a purely  mathematical one.  The precise answer is
       complex  and depends on the ambient CO level. In general, the linear CFK differential
       equation  is a better approximation to the  nonlinear equation during the uptake of CO than
       during the elimination of CO. The approximation also is better for COHb levels further from
25     the equilibrium predicted by the nonlinear model. In particular, it can be shown that as long
       as the linear CFK equation predicts COHb levels at or below 6%  COHb, the solution to the
       nonlinear CFK model will be no more than 0.5% COHb away (Smith et al., 1990).

       9.3.3.2  Confirmation Studies of the CFK Model
30          Since the publication of the original paper (Coburn et al., 1965), other investigators
       have tested the fit of the CFK model to experimental data by using different exposure profiles

       March 15, 1990                         9-16     DRAFT - DO NOT QUOTE OR CITE

-------
        and different approaches to evaluating the parameters of the model.  Stewart et al. (1970)
        and Peterson and Stewart (1970) tested the CFK linear differential equation on  18 resting
        subjects exposed to 25 CO exposure profiles for periods of one-half to 24 hours and to CO
        concentrations ranging from 1 to 1000 ppm. All physiological coefficients were assumed (see
 5      p. 9-14).  The percent COHb in the blood samples was determined twice, using an IL CO-
        Oximeter and a gas chromatograph.  It is important to note that in this experiment the
        predictions were compared to individual observations instead of averages.  The predicted
        values yielded COHb values quite close to the measured values.  The greatest discrepancy
        (4.9%) was observed in the experiment with steadily rising inhaled CO concentration over a
10      two-hour period, which is not surprising because the authors assumed a constant inhaled
        concentration of CO in solving the CFK equation.
            In 1975, Peterson and Stewart presented a second series of experiments testing the
        nonlinear CFK model.  Three women were included among the 22 subjects, and three
        different levels of exercise were used.  The parameter values of PCO2, DLCO, VB, and VA
15      were estimated for each subject. The percent COHb in the blood samples was determined by
        a CO-Oximeter that was continually compared to a gas chromatograph. Based on summary
        data, they concluded that the predicted and measured values were very close for both males
        and females under conditions at rest and exercise as well.
            In 1981 Joumard et al. tested both the linear and nonlinear CFK models for CO uptake
20      and elimination in pedestrians and car passengers exposed to ambient CO levels in the city of
        Lyon,  France. The cohort, consisting of 37 male and 36 female nonsmoking subjects who
        were 18 to 60 years old, was divided into two groups. One group was driven around the city
        in cars while the second group walked on the street at a nearly uniform pace. Each journey
        lasted about two hours. Blood COHb readings were taken at the beginning and end of each
25     journey. The percent COHb in the blood samples was determined by infrared spectroscopy.
        All other physiological parameters were estimated. As might be expected at these COHb
        levels (-2.3%), the authors found no significant difference between the linear and nonlinear
        CFK equations.  No significant difference (a =  0.05) was found between the final predicted
        and observed COHb values except for male pedestrians. The unspecified difference for that
30      group was attributed to an underestimate of the alveolar ventilation.
       March 15, 1990                         9-17     DRAFT - DO NOT QUOTE OR CITE

-------
            In 1984, Hauck and Neuberger ran a series of experiments testing the predictive ability
       of the CFK model on four subjects exposed to a total of 10 different CO exposure profiles
       combined with a variety of exercise (bicycle ergometer) patterns so that each exposure was a
       unique combination of CO concentration and exercise pattern.  The group, all nonsmokers,
 5     included three adult males and one ten-year-old female. The COHb values were calculated at
       1-min intervals using a numerical solution of the CFK model; all but ventilation-derived
       parameters, which were updated every minute, were kept constant. The percent COHb in the
       blood was determined by an improved van Slyke method.  The maximal differences within
       each experimental run (expressed as percent of a maximal predicted value) ranged from 4.2 to
10     11.1%.
            The most recent validation of the nonlinear CFK  model was reported by Tikuisis et al.
       (1987a,b).  Experiments were completed on 6 to 11 nonsmoking middle-aged males.  All of
       the CFK parameters but DLCO and VA were estimated;  DLCO and VA were measured for each
       subject. The percent COHb in the blood samples was determined by gas chromatography.
15     Several transient intermittent CO exposure profiles were tested:  1500 ppm for 5 min, and
       7500 ppm for 1 min at rest along with stepwise symmetric profiles of 500 to 4000 ppm for
       4.5 min and 4000 ppm for 75 during rest and intermittent exercise (VA « 30L/min;
       Figure 9-2). On an average, the predicted and measured values at rest were very close, with
       the CFK model slightly overpredictive (<0.5% COHb).  This overprediction was greater
20     during exercise, reaching almost 3% COHb in one of the  subjects (Figure 9-2). It is of
       interest to note that predicted values based on a current National Institute for Occupational
       Safety and Health (NIOSH) solution of the CFK model are even higher, overpredicting by as
       much as 6% COHb. The model appeared to be most sensitive to VA; thus errors in
       conversion of gas volumes (e.g., from ATPS to BTPS) will affect the predicted values.
25
       9.3.3.3 Modified CFK Models
             Bernard and Duker (1981) simplified the linear form of the CFK model in a unique
       way. Using regression equations from the literature, they were able to relate physiological
       parameters to the O2 uptake by the body  (VOj), which  in turn related to an activity level.  A
30     linear relationship was assumed between the rate of O2 uptake  and the maximum COHb level
        March 15, 1990                         9-18     DRAFT - DO NOT QUOTE OR CITE

-------
       %COHb
20^


15-


10-


 5


 0


20-


15-


10-


 5-


 0


20-


15-


10


 5


 0
                   Subject RV
                   Subject DH
                   Subject JG
                                Subject RP
                                            4487
Subject MB
5036     4956
                                            J
Subject RE
4945       4923
-4000


-3000


-2000


-1000


 0


-4000


-3000
   ppmCO
-2000


-1000


 0
                                                                       -4000


                                                                       -3000


                                                                        2000


                                                                       1-1000
                   0 5 10 15 20 25 30 35 40 45 50 0  5 10 15 20 25 30 35 40 45  50

                                        Time, mm.
Figure 9-2. Measured and predicted COHb concentrations from six intermittently exercising
subjects.  The solid lines represent the measured percent COHb; the short-dashed lines are the
solutions to the nonlinear CFK equation; and the long-dashed lines are predicted values based
on the CFK model adapted by NIOSH.

Source: Tikuisis et al. (1987b).
March 15, 1990
                         9-19     DRAFT - DO NOT QUOTE OR CITE

-------
       under which that rate could be sustained. A summary of predictive relationships between
       pairs of variables were developed, but none were experimentally tested.
            A more fundamental modification of the CFK model was made by Hill et al. (1977) to
       study the effect of CO inspired by the mother on the level of fetal COHb. The Hill
 5     equation (9-8) combines the CFK equation (for maternal COHb), with a term denoting COHb
       transfer from the placenta into the fetus (the subscripts m and f denote maternal and fetal
       quantities, respectively).

                        Vta d[COHbJ/dt = Vco* - [COHbJ PcO2/(Mm[O2Hb]B)             (9-8)
10                                     + P,CO/B - DPCO (PmCO - P£0)
       Thus, Equation 9-8 is the same as Equation 9-4, except for the final term on the right.
15     DPCO is the CO diffusion coefficient across the placenta.  PmCO and P£O are the partial
       pressures of CO in the maternal and fetal placenta! capillaries, respectively.  The latter two
       quantities are estimated using the Haldane relationship and separate models for the lungs and
       placenta. The level of fetal COHb is predicted from a similar equation.  Comparative
       evaluation of predicted and measured fetal COHb concentrations under time-varying and
20     steady-state conditions in both men  and animals showed acceptable agreement only under
       steady-state conditions (Hill et al., 1977; Longo and Hill, 1977).

       9.3.3.4 Application of the CFK Model
            Ott and Mage (1978), using a linear differential equation model that was patterned after
25     the linear CFK differential equation, examined the dynamics of blood COHb concentration
       fluctuation as a function of ambient CO concentration for a one-year period.  Other
       parameters of the model were estimated and kept constant The calculated COHb levels
       exceeded 2% on  25 occasions without violating the one-hour standard, whereas the eight-
       hour standard was violated 23 times.  During the same year 29  violations of the CO standard
30     occurred but the  2% COHb level was exceeded in only 23 instances. Besides evaluation of
       the averaged CO concentrations, the authors examined the effects of peak, transient CO
       concentrations on the target COHb. They showed that the presence of such spikes in CO data
       averaged over hourly intervals may lead to underestimating the COHb level (due to  exogenous

       March 15,  1990                         9-20     DRAFT - DO NOT QUOTE OR CITE

-------
        CO) by as much as 21%. Consequently, they recommended that monitored CO be averaged
        over shorter periods, such as 10 to 15 min.  (See Chapter 8 for a more complete description
        of population exposure to CO.)
            Venkatram and Louch (1979) extended the above application to more dynamic
 5      conditions by fitting interpolated values of the ambient one-hour CO averages from Toronto,
        Canada into the CFK model. In addition, they reexpressed the solution of the model from
        units of percent COHb to parts per million of CO. Such a transformation allows the
        examination of a variety of CO concentration profiles, while keeping a simple preselected
        target COHb as a constant.  They calculated that a 2% COHb level in blood very likely
10      would be exceeded on numerous occasions without ever  violating the standard.  By including
        transients their approach appears to predict COHb more  accurately, particularly  in response to
        eight-hour running averages.
            Biller and Richmond (1982) investigated the effects of inhaling various patterns of
        hourly-averaged CO concentrations that just attained alternative 1-hour and 8-hour CO
15      NAAQS using the CFK equation. Their analysis also estimated the distributions of various
        physiological parameters that are inputs to the CFK equation for individuals with
        cardiovascular disease.  The authors found that depending on which air quality pattern was
        used, the percentage of the population exceeding 2.1% COHb ranged from less than 0.01 % to
        10%.
20          More recently, Saltzman and Fox (1986) investigated the effect of inhaling oscillating
        levels of CO on the COHb  level of rabbits using the linear CFK equation simplified by
        combining the original parameters. They concluded that ambient CO values could be
        averaged safely over any time period less than or equal to the half-life of blood COHb.

25      9.3.4 Summary
            The best  all around model for COHb prediction is still the equation  developed by
        Coburn,  Forster, and Kane (1965). The linear solution is useful for examining air pollution
        data leading to relatively low COHb levels,  whereas the  nonlinear solution shows good
        predictive power even for high CO exposures.  The two  regression models might be useful
30      only when the  conditions of application closely approximate those under which the parameters
        were estimated.

        March 15, 1990                          9-21     DRAFT -  DO NOT QUOTE OR CITE

-------
            It is important to remember that almost all of the above studies assumed a constant rate
       of CO uptake and elimination, which is rarely true.  A number of physiological factors,
       particularly changes in ventilation,  will affect both rates.  The predicted COHb values also
       will differ from individual to individual due to smoking, age, or lung disease.  There does not
       appear to be a single optimal averaging time period  for ambient CO; however, the shorter the
       period the greater the precision.  In general, the averaging time period should be well within
       the [COHb] half-life, which decreases with increased activity.
10     9.4  INTRACELLULAR EFFECTS OF CARBON MONOXIDE
       9.4.1  Introduction
            The principal cause of CO toxicity is tissue hypoxia due to CO binding to Hb, yet
       certain physiological aspects of CO exposure are not explained well by decreases in
       intracellular PO2 related to the presence of COHb.  For many years, it has been known that
15     CO is distributed to extravascular sites such as skeletal muscle (Coburn et al.,  1971; Coburn
       et al.,  1973) and that 10 to 50%  of the total body store of CO may be extravascular
       (Luomanmaki and Coburn, 1969). Furthermore, extravascular CO is metabolized slowly to
       CO2 in vivo (Fenn, 1970).  Consequently,  secondary mechanisms of CO  toxicity related to
       intracellular uptake of CO have been the focus of a great deal of research interest.  CO
20     binding to many intracellular compounds has been well documented both in vitro and in vivo,
       however, it is still uncertain whether or  not intracellular uptake of CO in the presence of Hb
       is sufficient to cause either acute organ system dysfunction or long-term  health effects.  The
       virtual absence of sensitive techniques capable of assessing intracellular CO binding under
       physiological conditions has resulted in a variety of indirect approaches to the problem as well
25     as many negative studies.  The purposes of this  section of the document  are to summarize
       current knowledge pertaining to intracellular CO-binding proteins and to evaluate the potential
       contribution of intracellular CO uptake to the overall physiological effects of CO exposure.
       Selected aspects of this topic have been  reviewed previously (Forster, 1970; Coburn et al.,
        1977;  Coburn,  1979; Piantadosi, 1987;  Coburn and Forman, 1987).
30          CO is known to react with a variety of metal-containing proteins found in nature.  CO-
       binding metalloproteins present in mammalian tissues include O2-carrier  proteins such as

       March 15, 1990                          9-22     DRAFT - DO NOT QUOTE OR CITE

-------
        hemoglobin (Douglas et al., 1912) and myoglobin (Antonini and Brunori, 1971) and
        metalloenzymes (oxidoreductases) such as cytochrome c oxidase (Keilin and Hartree, 1939),
        cytochromes of the P-450 type (Omura and Sato, 1964), tryptophan oxygenase (Tanaka and
        Knox, 1959), and dopamine hydroxylase (Kaufman, 1966).  These metalloproteins contain
 5      iron and/or copper centers at their active sites that form metal-ligand complexes with CO in
        competition with molecular oxygen.  CO and O2 form complexes with metalloenzymes only
        when the iron and copper are in their reduced forms (Fe II, Cu I).  Caughey (1970) has
        reviewed the similarities and differences in the physicochemical characteristics of CO and O2
        binding to these transition metal ions.  The competitive relationship between CO and O2 for
10      the active site of intracellular hemoproteins usually is described by  the Warburg partition
        coefficient (K), which is the CO/O2 ratio that produces 50%  inhibition of the O2 uptake of the
        enzyme or, in the case of myoglobin, a 50% decrease in the number of available O2-binding
        sites.
            The measured Warburg coefficients of various mammalian CO-binding proteins have
15      been tabulated recently by Coburn and Forman (1987) (see Table 9-1).  These K values range
        from approximately 0.025 for myoglobin to 0.1 to 12 for cytochromes P-450.  K values of 2
        to 28 have been reported for cytochrome c oxidase (Keilin and Hartree, 1939; Wohlrab and
        Ogunmola, 1971; Wharton and Gibson 1976). By comparison, the K value for human
        hemoglobin of 0.005 is some three orders of magnitude less than that of cytochrome c
20      oxidase.   This means, for example, that CO would bind to cytochrome oxidase in vivo only if
        O2 gradients from RBCs in the capillary to the mitochondria were quite steep.  Application of
        K values for intracellular hemoproteins in this way, however, needs to be used with caution
        because most measurements of CO binding have not been made at physiological temperatures
        or at relevant rates of electron transport.
25          Apart from questions about the relevance of extrapolating in vitro partition coefficients
        to physiological conditions, experimental problems arise that are related to determining actual
        CO/O2 ratios in intact tissues. Reasonably good estimates of tissue  Pco may be obtained by
        calculating the value in mean capillary blood from the Haldane relationship (Coburn et al.,
        1977), neglecting the low rate of CO metabolism by the tissue.  Experimental estimates of the
30      Pa, in animal tissues have been found to be in close agreement with these calculations and
        average slightly less than alveolar P^ (Goethert et al., 1970; Goethert, 1972). In general,

        March 15, 1990                          9-23     DRAFT - DO NOT QUOTE OR CITE

-------
TABLE 9-1.  IN VTTRO INHIBITION RATIOS FOR HEMOPROTEINS THAT BIND CARBON MONOXIDE
Hemoprotein
Hemoglobin
Myoglobin
Cytochrome c oxidase
VO
K> Cytochrome P-450
Dopamine ft hydroxylase
O
j> Tryptophan oxygenase
S'R = CO/O2 at 50% inhibition
bM = 1/R
1
-~ Source: Adapted from Cobum and
•§
s
o
Source R' M" Temperature (°C)
Human RBC 0.0045 218 37
Sperm whale 0.025-0.040 25-40 25
Bovine heart 5-15 0.1 - 0.2 25
Rat liver 0.1-12 10-0.1 30-37
Bovine adrenal 2 0.5 —
Pseudomonas 0.55 1.8 25
Fonnan (1987).


-------
        steady state estimates for tissue PCQ range from 0.02 to 0.5 torr at COHb concentrations of 5
        to 50%. Therefore, at 50% COHb, a CO/O2 ratio of 5 may be achieved at sites of
        intracellular O2 uptake only if tissue PO2 in the vicinity of the CO-binding proteins is
        approximately 0.1 torr.
 5           Whether such low intracellular PO2 values exist in target tissues such as brain and heart
        during CO exposure is difficult to determine from the existing scientific literature.
        Experimental measurements of tissue PO2 using polarographic  microelectrodes indicate
        significant differences in PO2 in different tissues and regional differences in PO2 within a
        given tissue. This normal variability in tissue PO2 is related to differences in capillary
10      perfusion,  red blood cell spacing, velocity and path length, and local requirements for
        O2.  Normal PO2 values obtained from such recordings are generally in the range of 0 to
        30 torr (Leniger-Follert et al., 1975).  These PO2 values usually represent average interstitial
        values, although it is often difficult to determine the exact location of the electrode and the
        effect of O2 consumption by the electrode on the PO2 measurement. Furthermore, the
15      gradient between the capillary and the intracellular sites of O2 utilization are thought  to be
        quite steep (Sies, 1977).  A  major component of the gradient arises between the red blood cell
        and interstitium (Heliums, 1977) but the PO2 gradient between the cell membrane and
        respiring mitochondria and other O2-requiring organelles remains undetermined in intact
        normal tissues. Even less is known about intracellular PO2 in the presence of COHb. It has
20      been  determined, however, that both PO2 in brain tissue (Zorn, 1972) and cerebrovenous PO2
        (Koehler et al., 1984) decrease linearly as a function of COHb concentration. Presumably
        then, intracellular PO2 declines with increasing COHb concentration, and at certain locations,
        CO forms ligands with the O2-dependent, intracellular hemoproteins. As the intracellular PO2
        decreases,  the CO/O2 ratio in the tissue increases at constant Pco and an increasing fraction of
25      the available intracellular O2-binding sites become occupied by CO.
             The intracellular uptake of CO behaves according to the preceding general principles;
        most of the experimental evidence for this line of reasoning was derived from in  vivo studies
        of COHb formation by Coburn and colleagues (1965) at the University of Pennsylvania.  For
        all intracellular hemoproteins, however, two crucial quantitative unknowns remain.  These are
30      (1) the fraction of intracellular-binding sites in discrete tissues inhibited by CO at any level of
        COHb saturation, and (2)  the critical fraction of inhibited  sites necessary to amplify or initiate

        March 15,  1990                           9-25     DRAFT - DO NOT QUOTE OR CITE

-------
       a deleterious physiological effect, or trigger biochemical responses with long-term health
       effects. In general then, the activities of certain intracelliilar hemoproteins may be altered at
       physiologically tolerable levels of carboxyhemoglobin. The problem is in determining what
       level of intracellular reserve is available during CO hypoxia.  In view of this general
 5     conclusion, recent literature for the candidate hemoproteins has been evaluated to obtain
       positive evidence for intracellular CO binding and corroboration of functional consequences of
       the intracellular CO effects at  specific COHb levels.

       9.4.2  Carbon Monoxide Binding to  Myoglobin
10          The red protein myoglobin is involved in the transport of O2 from capillaries to
       mitochondria in red muscles.  The binding of CO to Mb in heart and skeletal muscle in vivo
       has been demonstrated at levels of COHb below 2% in heart and 1% in skeletal muscle
       (Coburn and Mayers, 1971; Coburn et al., 1973).  The ratio of COMb/COHb saturation has
       been found to be approximately one in cardiac muscle and less than one in skeletal muscle.
15     These ratios did not increase with increases in COHb up to 50% saturation.  In the presence
       of hypoxemia and hypoperfusion, the amount of CO uptake by Mb has been measured and
       was shown to increase (Coburn et al., 1973;  Coburn et al., 1971).  A similar conclusion has
       been reached  during maximal exercise in humans,  where CO shifts from Hb to the
       intracellular compartment (i.e., Mb, at COHb levels of 2 to 2.5%) (Clark and Coburn,
20     1975).  The significance of CO uptake by Mb is uncertain because our understanding of the
       functional role of Mb in working muscle is incomplete.  Myoglobin  undoubtedly enhances the
       uptake of O2 by muscle cells so that the continuous O2 demand of working muscle is satisfied
       (Wittenberg et al.,  1975). Myoglobin may contribute to muscle function by serving as an O2
       store, by enhancing intracellular diffusion of O2, or by acting as an O2 buffer to maintain a
25     constant mitochondria! PO2 during changes in O2 supply.  Functional Mb has been found to be
       necessary for maintenance of maximum O2 uptake and mechanical tension in exercising
       skeletal muscle (Cole,  1982).  The binding of CO  to Mb would therefore be expected to limit
       O2 availability to mitochondria in working muscle.  This possibility has been verified theoreti-
       cally by computer simulations  of Hoofd and Kreuzer (1978) and Agostoni et al. (1980).  The
30     three-compartment (arterial and venous capillary blood, and Mb) computer model of Agostoni
       et al.  (1980) predicted that COMb formation in low PO2 regions of the heart (e.g.,

       March 15,  1990                           9-26     DRAFT - DO NOT QUOTE OR CITE

-------
       subendocardium) could be sufficient to impair intracellular O2 transport to mitochondria at
       COHb saturations of 5 to 10%.  The [COMb] also was predicted to increase during conditions
       of hypoxia, ischemia, and increased O2 demand.
            The direct effects of CO on cardiac function also have been evaluated in the absence of
 5     Hb in fluorocarbon-perfused rabbits (Takano et al., 1981).  Exposure of these animals to high
       concentrations of CO (CO/02  = 0.05-0.25) significantly decreased the heart rate-systolic
       pressure product in the absence of COHb formation.  Cardiac output and [COMb], however,
       were not determined.  Increases in cardiac [COMb] have been measured after heavy work
       loads in CO-exposed rats, independent, of changes in [COHb] (Sokal et al., 1986).  These
10     investigators reported that exercise significantly increased cardiac [COMb] at COHb
       saturations of approximately 10, 20, and 50%, although metabolic acidosis worsened only at
       50% COHb. It remains unknown,  however, whether or not low [MbCO] could be
       responsible for decreases in maximal O2 uptake during exercise reported at COHb levels of 4
       to 5% (see Chapter 10, Section  10.3).
15
       9.4.3  Carbon Monoxide Uptake by Cytochrome P-450
            Mixed-function oxidases (cytochrome P-450) are involved in the detoxification of a
       number of drugs and steroids by "oxidation." These enzymes are distributed widely through-
       out mammalian tissues; the highest concentrations are found in the microsomes of liver,
20     adrenal gland, and the lungs of some species (Estabrook et al., 1970). These oxidases also
       are present in low concentrations in kidney and brain tissues. Mixed-function oxidases
       catalyze a variety of hydroxylation reactions involving the uptake of a pair of electrons from
       NADPH with reduction of one atom of O2 to H2O and incorporation of the other into
       substrates (White and Coon, 1980). These enzymes bind CO, and their Warburg binding
25     coefficients range from 0.1 to 12 in vitro (see Coburn and Forman, 1987). The sensitivity of
       cytochrome P-450 to CO is increased under conditions of rapid electron transport (Estabrook
       et al., 1970), however, previous calculations have indicated that tissue Pco is too low to
       inhibit the function of these hemoproteins in vivo at less than 15 to 20% COHb (Coburn and
       Forman, 1987). There have been few attempts to measure CO-binding coefficients in intact
30     tissues.  In isolated rabbit lung, the effects of CO on mixed-function oxidase are consistent
       with a Warburg coefficient of approximately 0.5 (Fisher et al.,  1979).  CO exposure

       March  15, 1990                          9-27    DRAFT - DO NOT QUOTE OR CITE

-------
       decreases the rate of hepatic metabolism of hexobarbital and other drugs in experimental
       animals (Montgomery and Rubin, 1973; Roth and Rubin, 1976a,b). Most authors have
       concluded that these effects of CO on xenobiotic metabolism are attributable entirely to
       COHb-related tissue hypoxia because they are no greater than the effects of "equivalent"
 5     levels of hypoxic hypoxia. Three optical studies of rat liver perfused in situ with Hb-free
       buffers have demonstrated uptake of CO by cytochrome P-450 at CO/O2 ratios of 0.03 to
       0.10 (Sies and Brauser, 1970; lyanagi et al., 1981; Takano et al., 1985). In the study by
       Takano et al. (1985) significant inhibition of hexobarbital metabolism was found at a CO/O2
       of about 0.1. This CO/O2 ratio, if translated directly to [COHb], would produce [COHb] that
10     are incompatible with survival (-95%).  At present, there is no scientific evidence that CO
       significantly inhibits the activity of mixed-function oxidases at COHb saturations below 15 to
       20%.  Although most studies do not indicate effects of CO on cytochrome P-450 activity at
       physiologically  relevant CO concentrations, specific P-450 isoenzymes may have higher
       affinities for CO. Also, the rate of substrate metabolism and substrate type may increase CO
15     binding by P-450 enzymes.  More basic research is needed in this area because of the
       important role of these enzymes in living organisms.

       9.4.4   Carbon Monoxide and Cytochrome c Oxidase
            Cytochrome c oxidase, a.k.a. cytochrome a a3, is the terminal enzyme in the
20     mitochondrial electron transport chain that catalyzes the reduction of molecular O2 to water.
       Although the enzyme complex binds CO, three reasons are often cited for why this should
       occur only under conditions of severe hypoxia.  First, the Warburg binding constant for
       cytochrome oxidase is unfavorable for CO uptake relative to the other candidate
       hemoproteins.  Second, the enzyme has an in vitro Michaelis-Menten constant  (KJ for O2 of
25     less than 1 torr  (Chance and Williams, 1955).  Because intracellular PO2 is probably higher
       than this, the oxidase should remain oxidized until severe tissue hypoxia is present.  The
       above arguments, however rational, are not supported well by in vivo observations and may
       not be  valid for the conditions encountered in living systems. The reasons for  this difficulty
       center around differences in the redox behavior of cytochrome oxidase in vivo relative to its
30     in vitro behavior. The enzyme has a high resting reduction level at normal PO2 in brain
       (Jobsis et al., 1977) and other tissues, and its oxidation state varies directly  with PO2 in vitro

       March 15, 1990                         9-28     DRAFT - DO NOT  QUOTE OR CITE

-------
        (Kreisman et al., 1981).  These findings may indicate that the oxidase operates near its
        effective K^ratios in vivo or that the availability of O2 to each mitochondrion or respiratory
        chain is not continuous under most physiological circumstances.  There also may be
        differences in or regulation of the K,,, for O2 of the enzyme according to regional metabolic
 5      conditions.  For example, the apparent B^ for O2 of cytochrome oxidase increases several
        times during rapid respiration (Oshino et al., 1974), and in isolated cells it varies as a
        function of the cytosolic phosphorylation potential  (Erecinska and Wilson,  1982). Conditions
        of high respiration and/or high cytosolic phosphorylation potential in vitro increase the
        concentration of CO-cytochrome oxidase at any CO/O2 ratio.  This concept is particularly
10      relevant for tissues like the heart and brain.
            Enhanced sensitivity of cytochrome oxidase to CO has been demonstrated in uncoupled
        mitochondria, where CO/O2 ratios as low as 0.2 delay the oxidation  of reduced cytochrome
        oxidase in transit from anoxia to normoxia (Chance et al., 1970). Several  studies of respiring
        tissues, however, have found  CO/O2 ratios of 12 to 20 to be necessary for 50% inhibition of
15      O2 uptake (Coburn et al.,  1979; Fisher and Dodia, 1981; Kidder, 1980).  In this context, it is
        important to note that in a given tissue,  the CO/O2 ratio necessary to inhibit one half of the
        O2 uptake does not necessarily correspond to CO binding to one half of the oxidase
        molecules.  This is because unblocked cytochrome oxidase molecules may oxidize respiratory
        complexes of blocked chains, thus causing the O2 consumption to fall more slowly than
20      predicted for strictly linear systems.  The capacity  of tissues  to compensate for electron
        transport inhibition by branching has  not been investigated systematically as a function of
        PO2, CO/O2 ratio, cytosolic phosphorylation potential, or rate of electron transport in vivo.
            The contention that intracellular CO uptake by cytochrome  oxidase occurs is supported
        by a few experiments.  It has been known for many years, primarily through the work of
25      Fenn (Fenn and  Cobb, 1932;  Fenn, 1970), that CO is slowly oxidized in the body to CO2.
        This oxidation occurs normally at a much lower rate than the endogenous rate  of CO
        production,  however, the rate of oxidation of CO increases in proportion to the CO body
        store (Luomanmaki and Coburn, 1969).  The oxidation of CO to CO2 was shown in 1965 by
        Tzagoloff and Wharton to be catalyzed by reduced cytochrome oxidase. More recently,
30      Young et al. (1979) demonstrated that oxidized cytochrome oxidase  promotes CO oxidation,
        and subsequently, that cytochrome oxidase in intact heart and brain mitochondria was capable

        March 15, 1990                          9-29     DRAFT - DO NOT QUOTE  OR CITE

-------
       of catalyzing the reaction at a CO/O2 ratio of approximately 4 (Young and Caughey, 1986).
       The physiological significance of this reaction is unknown.
             Other studies indicating possible direct effects of CO on cytochrome oxidase include a
       photoreversible effect of 500 to 1000 ppm CO on spontaneous electrical activity of cerebellar
 5     Purkinje cells in tissue culture (Raybourn et al., 1978).  These CO concentrations would be
       expected to produce [COHb] in the range of 33 to 50%.  At 7.5% COHb, inhibition of the fa-
       wave of the electroretinogram has been reported in the cat (Ingenito and Durlacher, 1979).
       Persistent changes in the retinogram were reminiscent of the "remnant effect" of CO on visual
       thresholds in humans reported by Halperin et al. (1959).  Other optical evidence suggesting
10     that cytochrome oxidase is sensitive to CO in vivo comes from studies of the effects of CO on
       cerebrocortical cytochromes in fluorocarbon-perfused rats (Piantadosi et al., 1985, 1987).  In
       these studies, CO/O2 ratios of 0.006 to 0.06 were associated with spectral evidence of CO
       binding to reduced cytochrome oxidase.  The spectral data also indicated that the  intracellular
       uptake of CO produced increases in the reduction level of b-type cytochromes in the brain
15     cortex.  At CO/O2 ratios of 0.06, most (> 80%) of the cytochrome b became reduced in the
       cerebral cortex.  The cytochrome b response is not understood well;  it is thought  to represent
       an indirect (e.g., energy-dependent) response of mitochondrial fr-cytochromes to CO because
       these cytochromes are not known to bind CO in situ.  The CO/O2 ratios used in the studies of
       Piantadosi et al. (1985, 1987) would produce [COHb] in the range of 50 to  90%. The
20     venous  PO2  in those experiments, however, was about 100 torr; thus at tissue PO2s that are
       significantly lower, this effect should occur at lower COHb saturations. It is unlikely,
       however, that cerebral uptake of CO is significant at COHb below 5 % because tissue Pco is so
       low in the presence of Hb.  The physiological significance of these effects of CO have not yet
       been determined.
25           Direct effects of CO on mitochondrial function have been suggested by several recent
       studies  which indicate decreases in cytochrome oxidase activity by histochemistry in brain and
       heart after severe CO intoxication in experimental animals (Pankow and Ponsold, 1984;
       Savolainen et al., 1980; Somogyi et al.,  1981).  The magnitude of the decrease in cytochrome
       oxidase activity may exceed that associated with severe hypoxia, although problems of deter-
30     mining  "equivalent" levels of CO hypoxia and hypoxic hypoxia have not been addressed
       adequately by these studies.  The effects of passive cigarette smoking on oxidative

       March  15, 1990                          9-30     DRAFT - DO NOT  QUOTE OR CITE

-------
       phosphorylation in myocardial mitochondria have been studied in rabbits (Gvozdjakova et al.,
       1984).  Mitochondrial respiratory rate (State 3 and State 4) and rates of oxidative
       phosphorylation were found to be decreased significantly by [COHb] of 6 to 7%.  These data,
       however, are not definitive with respect to CO because they include effects of nicotine, which
 5     reached concentrations of 5.7 jig/L in blood. A recent study by Snow et al. (1988) in dogs
       with prior experimental myocardial infarction indicated that a COHb of 9.4% increased the
       resting reduction level of cytochrome oxidase in the heart.  The CO exposures also were
       accompanied by more rapid cytochrome oxidase reductions after coronary artery occlusion
       and less rapid reoxidation of the enzyme after release of the occlusion.  The authors
10     concluded that CO trapped the oxidase in the reduced state during transient cardiac ischemia.
       There is also evidence that formation of the CO-cytochrome oxidase ligand occurs in the brain
       of the rat at COHb saturations of 40 to 50% (Brown and Piantadosi, 1990).  This binding
       appears to be related to hypotension and probable cerebral hypoperfusion during CO
       exposure. This effect is in concert with experimental evidence that CO  produces direct
15     vasorelaxation of smooth muscle.  This vasodilation occurs in rabbit aorta (Coburn et al.,
       1979), in the coronary circulation of isolated perfused rat heart (McFaul and McGrath, 1987),
       and in the cerebral circulation of the fluorocarbon-perfused rat (Piantadosi et al., 1987).  The
       mechanism of this vasodilator effect is unclear, although it appears to be related to decreased
       calcium concentrations in vascular smooth muscle (Lin and McGrath, 1988) and elevation of
20     cellular cyclic guanosine monophosphate (GMP) levels (Ramos et al., 1989).  The stimulus
       does not require hypoxia, adenosine or prostaglandins and it is possible  that it represents  a
       direct toxic effect of CO on the cytochrome  system in vascular smooth muscle.  The
       physiological significance of this phenomenon is undetermined.
             In summary, there is evidence to suggest that CO binds to cytochrome oxidase in
25     mammalian heart and brain tissues at a range of systemic PO2 values.  The only experimental
       evidence at present that this effect occurs at COHb levels less than 10% is the slow oxidation
       of CO to CO2, which has been shown to occur in vivo and in isolated  mitochondria in vitro.
       Experimental evidence indicates that CO binding to cytochrome oxidase does occur during
       tissue hypoxia produced by overtly toxic COHb concentrations.  The physiological
30     significance of these effects beyond those of tissue hypoxia remains unknown.  Once CO
       binding to cytochrome oxidase occurs, however, the small rate constant for CO  dissociation

       March 15, 1990                          9-31      DRAFT - DO NOT QUOTE OR  CITE

-------
         from the enzyme yields an apparent rate-dependent inhibition constant for CO under

         nonequilibrium conditions.  This means that at high rates of respiration and low O2

         concentrations, recovery of enzymatic function by the oxidase is relatively slow in comparison

         to simple O2 deprivation.

 5



       9.5  REFERENCES


10
       Agostoni, A.; Stabilini, R.; Viggiano, G.; Luzzana, M.; Samaja, M. (1980) influence of capillary and tissue PQ
              on carbon monoxide binding to myoglobin: a theoretical evaluation. Microvasc. Res.  20: 81-87.

       Anthonisen, N. R.; Robertson, P. C.; Ross, W. R. D. (1970) Gravity-dependent sequential emptying of lung
15            regions. J.  Appl. Physiol. 28: 589-595.

       Antonini, E.; Brunori, M.  (1971) The partition constant between two ligands. In: Hemoglobin and myoglobin in
              their reactions with ligands. Amsterdam, The Netherlands: North-Holland Publishing Co.; pp. 174-175.

20     Berk,  P. D.; Rodkey, F. L.; Blaschke, T. F.; Collison, H. A.; Waggoner, J.  G. (1974) Comparison of plasma
              bilirubin turnover and carbon monoxide production in man. J. Lab. Clin. Med. 83: 29-37.

       Berk,  P. D.; Blaschke, T. F.; Scharschmidt, B. F.; Waggoner, J. G.; Berlin,  N. I. (1976) A new approach to
              quantitation of the various sources of bilirubin in man. J. Lab. Clin. Med. 87: 767-780.
25
       Bernard, T. E.; Duker, J. (1981) Modeling carbon monoxide uptake during work. Am. Ind. Hyg. Assoc. J.
              42: 361-364.

       Biller, W.  F.; Richmond, H. M. (1982) Sensitivity analysis on Coburn model  predictions of COHb levels
30            associated with alternative CO standards. Research Triangle Park, NC: U.  S. Environmental Protection
              Agency, Office of Air Quality Planning and Standards; EPA contract no. 68-02-3600.

       Bischoff, K.  B. (1986) Physiological pharmacokinetics. Bull. Math. Biol. 48:  309-322.

35     Bosisio, E.; Grisetti, G. C.; Panzuti,  F.; Sergi, M. (1980)  Pulmonary diffusing capacity and its components (DM
              and Vc) in young, healthy smokers.  Respiration 40: 307-310.

       Britten, J.  S.; Myers, R. A. M. (1985) Effects of hyperbaric treatment on carbon monoxide elimination in
              humans. Undersea Biomed. Res. 12: 431-438.
40
       Brown, S.  D.; Piantadosi, C. A. (1990) In vivo binding of carbon monoxide to cytochrome c oxidase hi rat
              brain. J. Appl. Physiol.: in press.

       Buerk, D.  G.; Bridges, E.  W.  (1986) A simplified algorith for computing the variation in oxyhemoglobin
45            saturation with pH, PCO2, T and DPG. Chem. Eng. Commun. 47: 113-124.

       Caughey, W. S. (1970) Carbon monoxide bonding in hemeproteins. In:  Coburn, R. F., ed. Biological effects of
              carbon monoxide. Ann. N. Y. Acad. Sci.  174: 148-153.
        March 15, 1990                             9-32     DRAFT - DO NOT QUOTE OR CITE

-------
        Chance, B.; Williams, G. R. (1955) Respiratory enzymes in oxidative phosphorylation: HI. the steady state J.
               Biol. Chem. 217: 409-427. Clark, B. J.; Cobum, R. F. (1975) Mean myoglobin oxygen tension during
               exercise at maximal oxygen uptake. J. Appl. Physiol. 39: 135-144.

  5     Chance, B.; Erecinska, M.; Wagner, M. (1970) Mitochondrial responses to carbon monoxide toxicity. In:
               Coburn, R. F., ed. Biological effects of carbon monoxide. Ann. N. Y. Acad. Sci. 174: 193-204.

        Clarke, S. W.; Jones, J. G.; Glaister, D. H. (1969) Change in pulmonary ventilation in different postures. Gin.
               Sci. 37: 357-369.
10
        Coburn, R. F.  (1970) Enhancement by phenobarbital and diphenylhydantoin of carbon monoxide production in
               normal man. N. Engl. J. Med. 283: 512-515.

        Coburn, R. F.  (1979) Mechanisms of carbon monoxide toxicity. Prev. Med. 8: 310-322.
15
        Coburn, R. F.; Forman, H. J. (1987) Carbon monoxide toxicity.  In: Geiger, S. R.; Farhi, L. E.; Tenney, S.  M.;
               Fishman, A. P., eds. Handbook of physiology: a critical, comprehensive presentation of physiological
               knowledge and concepts. Section 3: the respiratory system, volume IV. gas exchange. Bethesda, MD:
               American Physiological  Society; pp. 439-456.
20
        Coburn, R. F.; Mayers, L. B. (1971) Myoglobin O2 tension determined from measurements of carboxymyoglobin
               in skeletal muscle. Am.  J. Physiol. 220: 66-74.

        Cobum, R. F.; Williams, W. J.; Forster, R. E. (1964) Effect of erythrocyte destruction on carbon monoxide
25            production in man. J. Clin. Invest. 43: 1098-1103.

        Coburn, R. F.; Forster, R. E.; Kane, P. B. (1965) Considerations of the physiological variables that determine
               the blood carboxyhemoglobin concentration in man. J. Clin. Invest. 44: 1899-1910.

30     Coburn, R. F.; Williams,  W. J.; Kahn, S. B. (1966) Endogenous carbon monoxide production in patients with
               hemolytic anemia. J. Clin. Invest. 45: 460-468.

        Cobum, R. F.; Wallace, H. W.; Abboud, R. (1971) Redistribution of body carbon monoxide after hemorrhage.
               Am. J. Physiol. 220: 868-873.
35
        Coburn, R. F.; Ploegmakers, F.; Gondrie, P.; Abboud, R. (1973) Myocardial myoglobin oxygen tension. Am. J.
               Physiol. 224: 870-876.

        Coburn, R. F.; Allen, E. R.; Ayres, S.; Bartlett, D., Jr.; Horvath, S. M.; Kuller, L. H.; Laties, V.; Longo,
40            L. D.; Radford, E. P., Jr. (1977) Carbon monoxide.  Washington, DC: National Academy of Sciences.

        Coburn, R. F.; Grubb, B.; Aronson, R. D. (1979) Effect of cyanide on oxygen tension-dependent mechanical
               tension in rabbit aorta. Circ. Res. 44: 368-378.

45     Cole, R. P. (1982) Myoglobin function in exercising skeletal muscle. Science (Washington, DC) 216: 523-525.

        Delivoria-Papadopoulos, M.; Coburn, R. F.; Forster, R. E. (1970) Cyclical variation of rate of heme destruction
               and carbon monoxide production (Vco) in normal women. Physiologist 13: 178.

50     Douglas, C. G.; Haldane, J. S.; Haldane, J. B. S. (1912) The laws of combination of haemoglobin with carbon
               monoxide and oxygen. J. Physiol. (London) 44: 275-304.

        Erecinska, M.; Wilson, D. F. (1982) Regulation of cellular energy  metabolism. J. Membr. Biol. 70: 1-14.
         March  15, 1990                               9-33      DRAFT - DO NOT QUOTE OR CITE

-------
        Estebrook, R. W.; Franklin, M. R.; Hildebrandt, A. G. (1970) Factors influencing the inhibitory effect of
               carbon monoxide on cytochrome P-450-catalyzed mixed function oxidation reactions. In: Coburn, R. F.,
               ed. Biological effects of carbon monoxide. Ann. N. Y. Acad. Sci. 174: 218-232.

 5      Federal Register. (1985) Review of the national ambient air quality standards for carbon monoxide; final rule.
               F. R. (September 13) 50: 37484-37501.

        Fenn, W. O. (1970) The burning of CO in the tissues. In:  Coburn, R. F., ed. Biological effects of carbon
               monoxide. Ann. N. Y. Acad. Sci. 174: 64-71.
10
        Fenn, W. O.; Cobb, D. M. (1932) The burning of carbon monoxide by heart and skeletal muscle. Am. J.
               Physiol. 102: 393-401.

        Fisher,  A. B.; Dodia, C. (1981) Lung as a model for evaluation of critical intracellular PQ  and P^. Am.  J.
15             Physiol. 241: E47-E50.                                                     2

        Fisher,  G. L.; Chrisp, C. E.; Raabe, O. G. (1979) Physical factors affecting the mutagenicity of fly ash from a
               coal-fired power plant.  Science (Washington, DC) 204: 879-881.

20      Fishman, A.  P.; Farhi, L. E.; Tenney, S. M., eds. (1987) Handbook of physiology section 3: the respiratory
               system, v. 4, gas exchange. Bethesda, MD: American Physiological  Society.

        Forster, R. E. (1970) Carbon monoxide and the partial pressure of oxygen in  tissue.  In: Coburn, R. F., ed.
               Biological effects of carbon monoxide. Ann. N. Y. Acad. Sci. 174: 233-241.
25
        Forster, R. E. (1987) Diffusion of gases across the alveolar membrane. In: Fishman,  A. P.; Farhi, L. E.;
               Tenney, S. M., eds. Handbook of physiology section 3: the respiratory system, v. 4, gas exchange.
               Bethesda, MD: American Physiological Society; pp. 71-88.

30      Frey, T. M.; Crapo, R. O.; Jensen,  R. L.; Elliott, C. G. (1987) Diurnal variation of the diffusing capacity of the
               lung: is it real? Am. Rev.  Respir. Dis.  136: 1381-1384.

        Glaister, D. H.; Schroter, R. C.; Sudlow, M. F.; Milic-Emili, J. (1973) Transpulmonary pressure gradient and
               ventilation distribution  in excised lungs. Respir.  Physiol.  17: 365-385.
35
        Godin,  G.; Shephard, R. J. (1972) On the course of carbon monoxide uptake and release. Respiration
               29: 317-329.

        Goethert, M. (1972) Factors influencing the CO content of tissues.  Staub Reinhalt. Luft 32: 15-20.
40
        Goethert, M.; Lutz, F.; Malorny,  G. (1970) Carbon monoxide partial pressure in tissue of different animals.
               Environ. Res. 3: 303-309.

        Guyatt, A. R.; Holmes, M. A.; Gumming, G. (1981) Can carbon monoxide be absorbed from the upper
45             respiratory tract in man? Eur. J. Respir. Dis. 62: 383-390.

        Gvozdjakova, A.; Bada, V.; Sany, L.; Kucharska, J.;  Kruty, F.; Bozek, P.; Trstansky, L.; Gvozdjak, J. (1984)
               Smoke cardiomyopathy: disturbance of oxidative processes in myocardial mitochondria. Cardiovasc. Res.
               18: 229-232.
50
        Haldane, J. (1898) Some improved methods of gas analysis. J. Physiol. (London) 22: 465-480.

        Halperin, M. H.; McFarland, R. A.; Niven,  J. I.; Roughton,  F. J. W. (1959) The time course of the effects of
               carbon monoxide on visual thresholds. J. Physiol. (London) 146: 583-593.


         March 15, 1990                               9-34      DRAFT - DO NOT QUOTE OR CITE

-------
       Harf, A.; Pratt, T.; Hughes, J. M. B. (1978) Regional distribution ofV  •
                            e                      6                    a/Q in man at rest and with exercise
              measured with krypton-Sim. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 44: 115-123.

       Hauck, H.; Neuberger, M. (1984) Carbon monoxide uptake and the resulting carboxyhemoglobin in man. Eur. J.
 5            Appl. Physiol. 53: 186-190.

       Heliums, J. D. (1977) The resistance to oxygen transport in the capillaries relative to that in the surrounding
              tissue. Microvasc. Res. 13:  131-136.

10     Hill, E. P.; Hill, J. R.; Power, G. G.; Longo, L. D. (1977) Carbon monoxide exchanges between the human
              fetus and mother: a mathematical model. Am. J. Physiol. 232: H311-H323.

       Hoofd, L.; Kreuzer, F. (1978) Calculation of the facilitation of O2 or CO transport by Hb or Mb by means of a
              new method for solving the carrier-diffusion problem.  In:  Silver, I. A.; Erecinska, M.; Bicher, H. L,
15            eds. Oxygen transport to tissue - III. New York, NY: Plenum Press; pp.  163-168. (Advances in
              experimental medicine and biology: v. 94).

       Hughes, J. M. B.; Grant, B. J. B.;  Greene, R. E.; Iliff, L. D.; Milic-Emili, J. (1972) Inspiratory flow rate and
              ventilation distribution in normal subjects and in patients with simple chronic bronchitis.  Clin. Sci.
20            43: 583-595.

       Ingenito, A. J.; Durlacher, L. (1979) Effects of carbon monoxide on the b-wave of the cat electroretinogram:
              comparisons with nitrogen hypoxia, epinephrine, vasodilator drugs and changes in respiratory tidal
              volume. J. Pharmacol. Exp. Ther. 211:  638-646.
25
       Inkley, S. R.; Maclntyre, W. J. (1973) Dynamic measurement of ventilation-perfusion with xenon-133 at resting
              lung volumes. Am. Rev. Respir.  Dis. 107: 429-441.

       lyanagi, T.; Suzaki, T.; Kobayashi, S. (1981) Oxidation-reduction states of pyridine nucleotide and cytochrome
30            P-450 during mixed-function oxidation in perfused rat liver. J. Biol. Chem. 256: 12933-12939.

       Jobsis, F. F.; Keizer,  J. H.; LaManna, J. C.; Rosenthal, M. (1977) Reflectance spectrophotometry of
              cytochrome aa3 in vivo. J. Appl.  Physiol.: Respir. Environ. Exercise Physiol. 43: 858-872.

35     Joumard, R.; Chiron,  M.; Vidon, R.; Maurin, M.; Rouzioux, J.-M. (1981) Mathematical models of the uptake
              of carbon monoxide on hemoglobin at low carbon monoxide levels. EHP Environ. Health Perspect.
              41: 277-289.

       Kaufman, S. (1966) D. Coenzymes and hydroxylases: ascorbate and dopamine-/3-hydroxylase;
40            tetrahydropteridines and phenylalaniue and tyrosine hydroxylases. Pharmacol. Rev.  18: 61-69.

       Keilin, D.; Hartree, E. F. (1939) Cytochrome and cytochrome oxidase. Proc.  R. Soc. London B 127: 167-191.

       Kidder, G. W., III.  (1980) Carbon  monoxide insensitivity of gastric acid  secretion. Am. J. Physiol.
45            238: G197-G202.

       Knudson, R. J.; Kaltenborn, W. T.; Burrows, B. (1989) The effects of cigarette smoking and smoking cessation
              on the carbon monoxide diffusing capacity of the lung in asymptomatic subjects. Am. Rev. Respir. Dis.
               140: 645-651.
50
       Koehler, R. C.; Traystman, R. J.; Zeger, S.; Rogers, M.  C.;  Jones, M. D., Jr. (1984) Comparison of
              cerebrovascular responses to hypoxic and carbon monoxide hypoxia in newborn and adult sheep. J.
              Cereb. Blood Flow Metab.  4: 115-122.
         March 15, 1990                               9-35      DRAFT - DO NOT QUOTE OR CITE

-------
       Kreisman, N. R.; Sick, T. J.; LaManna, J. C.; Rosenthal, M. (1981) Local tissue oxygen tension - cytochrome
              8,% redox relationships in rat cerebral cortex in vivo. Brain Res. 218:  161-174.

       Landaw, S. A. (1973) The effects of cigarette smoking on total body burden and excretion rates of carbon
 5            monoxide. JOM J. Occup. Med. 15:  231-235.

       Landaw, S. A.; Callahan, E. W., Jr.;  Schmid, R. (1970) Catabolism of heme in vivo: comparison of the
              simultaneous production of bilirubin and carbon monoxide. J. Clin.  Invest. 49: 914-925.

10     Leniger-Follert, E.; Luebbers, D. W.; Wrabetz, W. (1975) Regulation of local tissue Po2 of the brain cortex at
              different arterial O2 pressures.  Pfluegers Arch.  359: 81-95.

       Lin, H.; McGrath, J. J. (1988) Carbon monoxide effects on calcium levels in vascular smooth muscle. Life Sci.
              43: 1813-1816.
15
       Longo, L. D.; Hill, E. P. (1977) Carbon monoxide uptake and elimination in  fetal and maternal sheep. Am. J.
              Physiol. 232: H324-H330.

       Luomanmaki, K.; Coburn, R.  F. (1969) Effects of metabolism and distribution of carbon monoxide on blood and
20            body stores. Am. J. Physiol. 217: 354-363.

       Lynch, S. R.; Moede, A. L. (1972) Variation in the rate of endogenous carbon monoxide production in normal
              human beings. J. Lab. Clin. Med. 79: 85-95.

25     Marcus,  A. H.  (1980a) Mathematical models for carboxyhemoglobin. Atmos. Environ. 14:  841-844.

       Marcus,  A. H.  (1980b) Author's reply. Atmos. Environ. 14:  1777-1781.

       Martin, C. J.; Das, S.; Young, A. C.  (1979) Measurements of the dead space volume. J. Appl. Physiol.: Respir.
30             Environ. Exercise Physiol. 47: 319-324.

       McCartney, M. L. (1990) Sensitivity analysis applied to Coburn-Forster-Kane models of carboxyhemoglobin
               formation.  Am. Ind. Hyg. Assoc. J.: in press.

35     McClean, P. A.; Duguid, N. J.; Griffin, P. M.; Newth, C. J. L.; Zamel, N.  (1981) Changes in  exhaled
               pulmonary diffusing capacity at rest and exercise in individuals with impaired positional diffusion. Bull.
               Eur. Physiopathol. Respir. 17: 179-186.

       McFaul, S. J.;  McGrath, J. J. (1987)  Studies on the mechanism of carbon monoxide-induced vasodilation in the
40            isolated perfused rat heart. Toxicol. Appl. Pharmacol. 87: 464-473.

       Milic-Emili, J.; Henderson, J. A. M.; Dolovich, M. B.; Trop, D.; Kaneko, K. (1966) Regional  distribution of
               inspired gas in the lung J.  Appl. Physiol. 21: 749-759.

45     Millette, B.; Robertson, P.  C.; Ross,  W. R. D.; Anthonisen, N. R. (1969) Effect of expiratory flow rate on
               emptying of lung regions. J. Appl. Physiol. 27: 587-591.

       Miyahara, S.; Takahashi, H. (1971) Biological CO evolution: carbon monoxide evolution during auto- and
               enzymatic oxidation of phenols. J. Biochem. (Tokyo) 69: 231-233.
50
       Montgomery, M. R.;  Rubin, R. J. (1973) Oxygenation during inhibition of drug metabolism by carbon monoxide
               or hypoxic hypoxia. J. Appl. Physiol. 35: 505-509.
         March 15, 1990                               9-36      DRAFT - DO NOT QUOTE OR CITE

-------
        Muller, K. E.; Barton, C. N. (1987) A nonlinear version of the Coburn, Forster and Kane model of blood
               carboxyhemoglobin. Atmos. Environ. 21: 1963-1968.

        Omura, T.; Sato, R. (1964) The carbon monoxide-binding pigment of liver microsomes: I. Evidence for its
  5            hemoprotein nature. J. Biol. Chem. 239: 2370-2378.

        Oshino, N.; Sugano, T.; Oshino, R.; Chance, B. (1974) Mitochondrial function under hypoxic conditions: the
               steady states of cytochrome a+a3 and their relation to mitochondria! energy states. Biochim. Biophys.
               Acta 368: 298-310.
 10
        Ott, W. R.; Mage, D.  T. (1978) Interpreting urban carbon monoxide concentrations by means of a computerized
               blood COHb model. J. Air Pollut. Control Assoc. 28: 911-916.

        Ott, W.; Mage, D. (1980) Evaluation of CO air quality criteria using a COHb model. Atmos. Environ.
 15            14: 979-980.

        Pace, N.; Strajman, E.; Walker, E. L. (1950) Acceleration of carbon monoxide elimination in man by high
               pressure oxygen. Science (Washington,  DC) 111: 652-654.

20     Pankow, D.; Ponsold,  W. (1984) Effect of carbon monoxide  exposure on heart cytochrome c oxidase activity of
               rats. Biomed. Biochim. Acta 43: 1185-1189.

        Peterson, J. E.; Stewart, R. D. (1970) Absorption and elimination of carbon monoxide by inactive young  men.
               Arch. Environ. Health 21: 165-171.
25
        Peterson, J. E.; Stewart, R. D. (1975) Predicting the carboxyhemoglobin levels resulting from carbon monoxide
               exposures. J. Appl. Physiol. 39: 633-638.

        Piantadosi, C. A. (1987) Carbon monoxide, oxygen transport, and oxygen metabolism. J. Hyperbaric Med.
30            2: 27-44.

        Piantadosi, C. A.; Sylvia, A. L.; Saltzman, H.  A.; Jobsis-Vandervliet, F. F. (1985) Carbon
               monoxide-cytochrome interactions in the brain of the  fluorocarbon-perfused rat. J. Appl.  Physiol.
               58: 665-672.
35
        Piantadosi, C. A.; Sylvia, A. L.; Jobsis-Vandervliet,  F.  F. (1987) Differences in brain cytochrome responses to
               carbon  monoxide and cyanide in vivo. J. Appl. Physiol. 62:  1277-1284.

        Rahn, H.; Fenn, W.  O. (1955) A graphical analysis of the respiratory gas exchange: the O2-CO2  diagram.
40           Washington,  DC: American Physiological Society; p. 37.

        Raybourn, M. S.; Cork, C.; Schimmerling, W.; Tobias, C. A.  (1978) An in vitro  electrophysiological assessment
              of the direct cellular toxicity of carbon monoxide. Toxicol. Appl. Pharmacol. 46: 769-779.

45     Riley, R. L.; Permutt,  S. (1973) Venous admixture component of the AaPg gradient. J.  Appl. Physiol.
              35: 430-431.                                                  2

        Roth, R. A., Jr.; Rubin, R. J. (1976a) Role of blood flow in carbon monoxide- and hypoxic hypoxia-induced
              alterations in hexobarbital metabolism in rats. Drug Metab. Dispos. 4: 460-467.
50
        Roth, R. A., Jr.; Rubin, R. J. (1976b) Comparison of the effect of carbon monoxide and of hypoxic hypoxia. I.
              In vivo  metabolism, distribution and action of hexobarbital. J. Pharmacol. Exp. Ther. 199: 53-60.
         March 15,  1990                              9-37      DRAFT - DO NOT QUOTE OR CITE

-------
       Roughton, F. J. W. (1970) The equilibrium of carbon monoxide with human hemoglobin in whole blood. In:
              Biological effects of carbon monoxide, proceedings of a conference; January; New York, NY. Ann.
              N. Y. Acad. Sci. 174(art.  1): 177-188.

 5     Roughton, F. J. W.; Darling, R. C. (1944) The effect of carbon monoxide on the oxyhemoglobin dissociation
              curve. Am.  J.  Physiol. 141: 17-31.

       Saltzman, B. E.; Fox, S. H.  (1986) Biological significance of fluctuating concentrations of carbon monoxide.
              Environ. Sci. Technol. 20: 916-923.
10
       Savolainen,  H.; Kurppa, K.; Tenhunen, R.; Kivisto, H. (1980) Biochemical effects of carbon monoxide
              poisoning in rat brain with special reference to blood carboxyhemoglobin and cerebral cytochrome
              oxidase activity. Neurosci.  Lett. 19: 319-323.

15     Schoenfisch, W. H.; Hoop, K. A.; Struelens, B. S. (1980) Carbon monoxide absorption through the oral and
              nasal mucosae of cynomolgus monkeys. Arch. Environ. Health 35: 152-154.
                                                               •      •
       Scrimshire,  D. A. (1977) Theoretical analysis of independent V. and Q inequalities upon pulmonary gas exchange.
              Respir.  Physiol. 29:  163-178.
20
       Severinghaus, J. W.; Stupfel, M. (1957) Alveolar dead space as an index of distribution of blood flow in
              pulmonary capillaries. J. Appl. Physiol.  10: 335-348.

       Sies, H.  (1977) Oxygen gradients  during hypoxic steady states in liver: urate oxidase and cytochrome oxidase as
25            intracellular O2 indicators. Hoppe Seylers Z. Physiol. Chem. 358: 1021-1032.

       Sies, H.; Brauser, B. (1970) Interaction of mixed function oxidase with its substrates and associated redox
              transitions of cytochrome P-450 and pyridine nucleotides hi perfused rat liver. Eur. J. Biochem.
               15:  531-540.
30
       Smith, M. V. (1990) Comparing solutions to the linear and nonlinear CFK equations for predicting COHb
               formation. Math. Biosci.:  in press.

       Smith, R. G. (1968) Discussion of paper by Dr. Dinman. JOM J. Occup. Med. 10: 456-462.
35
       Snow, T. R.; Vanoli, E.; De Ferrari, G.; Stramba-Badiale, M.; Dickey, D. T. (1988) Response of cytochrome
               a.aj to carbon monoxide in canine hearts with prior infarcts. Life Sci. 42: 927-931.

        Sokal, J. A.; Majka, J.; Palus, J.  (1984) The content of carbon monoxide in the tissues of rats intoxicated with
40            carbon monoxide in various conditions of acute exposure. Arch. Toxicol. 56: 106-108.

        Sokal, J.; Majka, J.; Palus,  J. (1986) Effect of work load on the content of carboxymyoglobin in the heart and
               skeletal muscles of rats exposed to carbon monoxide. J. Hyg. Epidemiol.  Microbiol. Immunol.
               30:  57-62.
45
        Solanki, D. L.; McCurdy, P. R.;  Cuttitta, F.  F.; Schechter, G. P. (1988) Hemolysis in sickle  cell disease as
               measured by endogenous carbon monoxide production: a preliminary report. Am. J. Clin. Pathol.
               89:  221-225.

50     Somogyi, E.; Balogh, L; Rubanyi, G.; Sotonyi, P.; Szegedi, L. (1981) New findings concerning the
               pathogenesis of acute carbon monoxide (CO) poisoning. Am. J. Forensic Med.  Pathol. 2:  31-39.
          March 15,  1990                               9-38      DRAFT - DO NOT QUOTE OR CITE

-------
        Standfuss, K. (1970) Die Auswirkung der physiologischen Aenderungen des Ventilations-Perfusionsverhaeltnisses
               in der Zeit auf den funktionellen Totraum [Variations of the ventilation-perfusion ratio during the
               respiratory cycle and their effect upon physiologic dead space].  Pfluegers Arch. 317: 198-227.

  5     Stewart, R. D.; Peterson, J. E.; Baretta, E. D.; Bachand, R. T.; Hosko, M. J.;  Herrmann, A. A. (1970)
               Experimental human exposure to carbon monoxide. Arch. Environ. Health 21: 154-164.

        Stewart, R. D.; Peterson, J. E.; Fisher, T. N.; Hosko, M.  J.; Baretta, E. D.; Dodd, H. C.; Herrmann, A. A.
               (1973) Experimental human exposure to high concentrations of carbon monoxide. Arch. Environ. Health
 10            26: 1-7.

        Stokes, D. L.; Maclntyre, N. R.; Nadel,  J. A. (1981) Nonlinear increases in diffusing capacity during exercise
               by seated and supine subjects. J. Appl. Physiol.:  Respir. Environ. Exercise Physiol. 51: 858-863.

 15     Sutherland, P. W.; Katsura, T.; Milic-Emili, J.  (1968) Previous volume history of the lung and regional
               distribution of gas. J. Appl. Physiol. 25: 566-574.

        Takano, T.; Miyazaki, Y.; Shimoyama, H.; Maeda, H.; Okeda, R.; Funata, N. (1981) Direct effects of carbon
               monoxide on cardiac function. Int. Arch. Occup.  Environ. Health 49: 35-40.
 20
        Takano, T.; Motohashi, Y.; Miyazaki, Y.; Okeda, R. (1985) Direct effect of carbon monoxide on hexobarbital
               metabolism in the isolated perfused liver in the absence of hemoglobin. J. Toxicol. Environ.  Health
               15: 847-854.

 25     Tanaka, T.; Knox, W. E. (1959) The nature and mechanism of the tryptophan pyrrolase (peroxidase-oxidase)
               reaction of Pseudomonas and of rat liver. J. Biol. Chem. 234: 1162-1170.

        Tikuisis, P.; Buick,  F.; Kane, D. M. (1987a) Percent carboxyhemoglobin in resting humans exposed repeatedly
               to  1,500 and 7,500 ppm CO. J. Appl. Physiol.  63:  820-827.
 30
        Tikuisis, P.; Madill, H. D.; Gill, B. J.; Lewis, W. F.; Cox, K. M.; Kane, D. M. (1987b) A critical analysis of
               the use of the CFK equation in predicting COHb formation. Am. Ind. Hyg. Assoc. J. 48: 208-213.

        Tyuma, I.; Ueda, Y.; Imaizumi, K.; Kosaka, H. (1981) Prediction of the carbonmonoxyhemoglobin  levels during
 35            and after carbon monoxide exposures in various animal species.  Jpn. J. Physiol. 31: 131-143.

        Tzagoloff,  A.; Wharton, D. C.  (1965) Studies on the electron transfer system: LXII. the reaction of cytochrome
               oxidase with carbon monoxide. J.  Biol. Chem. 240: 2628-2633.

 40     Venkatram, A.; Louch, R. (1979) Evaluation of CO air quality criteria using a COHb model. Atmos. Environ.
               13: 869-872.

        Wagner, J. A.; Horvath, S. M.; Dahms, T.  E. (1975) Carbon monoxide elimination. Respir. Physiol. 23: 41-47.

45     Werner, B.; Lindahl, J.  (1980) Endogeneous carbon monoxide production after bicycle exercise in healthy
               subjects and in patients with hereditary spherocytosis. J. Clin. Lab. Invest. 38: 319-325.

        Wharton, D. C.; Gibson, Q. H. (1976) Cytochrome oxidase from Pseudomonas aeruginosa: IV. reaction with
               oxygen and carbon monoxide. Biochim. Biophys.  Acta 430: 445-453.
50
        White, R. E.; Coon,  M. J. (1980) Oxygen activation by cytochrome P-450. Annu. Rev. Biochem. 49: 315-356.

        Wittenberg, B. A.; Wittenberg,  J. B.; Caldwell,  P. R. B. (1975) Role of myoglobin in the oxygen supply to red
               skeletal muscle. J. Biol. Chem. 250: 9038-9043.


        March 15, 1990                               9-39     DRAFT - DO NOT QUOTE OR CITE

-------
       Wohlrab, H.; Ogunmola, B. G. (1971) Carbon monoxide binding studies of cytochrome a^ hemes in intact rat
              liver mitochondria.  Biochemistry 10: 1103-1106.

       Wyman, J.; Bishop, G.; Richey, B.; Spokane, R.; Gill, S. (1982) Examination of Haldane's first law for the
 5            partition of CO and O2 to hemoglobin A,,. Biopolymers 21: 1735-1747.

       Young, L. J.; Caughey, W. S. (1986) Mitochondrial oxygenation of carbon monoxide.  Biochem. J.
              239: 225-227.

10     Young, L. J.; Choc, M. G.; Caughey, W. S. (1979) Role of oxygen and cytochrome c oxidase in the
              detoxification of CO by oxidation to CO2. In: Caughey, W. S.; Caughey, H., eds. Biochemical and
              clinical aspects of oxygen: proceedings of a symposium; September 1975; Fort Collins, CO. New York,
              NY: Academic;  pp. 355-361.

15     Zankl, J. G. (1981) Berechnung der CO-Aufnahme des menschlichen Blutes [Calculation of the uptake of CO into
              human blood]. In: Mitteilungen des Institutes fuer Verbrennungskraftmaschinen und Thermodynamik.
              Graz, V/est Germany:  Technische University. Available from:  NTIS, Springfield, VA; DE83-901060.

       Zorn, H. (1972) The partial oxygen pressure in the brain  and liver at subtoxic concentrations of carbon
20            monoxide. Staub Reinhalt. Luft 32: 24-29.
         March 15,  1990                              9-40     DRAFT - DO NOT QUOTE OR CITE

-------
            10.   HEALTH EFFECTS OF CARBON MONOXIDE
        10.1  INTRODUCTION
 5          Concerns about the potential health effects of exposure to CO have been addressed in
        extensive studies with various animal species as subjects. Under varied experimental
        protocols, considerable information has been obtained on the toxicity of CO, its direct effects
        on the blood and other tissues, and the manifestations of these effects in the form of changes
        in organ function. Many of these studies, however, have been conducted at extremely high
10      levels of CO (i.e., levels not found in ambient air). Although severe effects from exposure to
        these high levels of CO are not directly germane to the problems from exposure to current
        ambient levels of CO, they can provide valuable information about potential effects of
        accidental exposure to CO, particularly those exposures occurring indoors. These higher level
        studies, therefore, are being considered in this chapter only if they extend dose-response
15      information or if they provide clues to  other potential  health effects of CO that have not been
        identified already. Emphasis has been  placed on studies conducted at ambient or near-
        ambient concentrations of CO that have been published in the more recent peer-reviewed
        literature since completion of the previous criteria document (U.S. Environmental Protection
        Agency, 1979) and an addendum to that document (U.S. Environmental Protection Agency,
20      1984). Where appropriate, information available from older studies either has been
        summarized in the text or placed in tables.
            The effects observed from nonhuman experimental studies have provided  some insight
        into the role CO plays in cellular metabolism.  Caution must be exercised, however, in
        extrapolating the results obtained from  these data to man. Not only are there questions
25      related to species differences, but exposure conditions differ markedly in the studies
        conducted by different investigators. Although these studies must be interpreted with caution,
        they do serve the valuable ends of (1) suggesting studies to be verified in man,  (2) exploring
        the properties and principles of an effect much more thoroughly and extensively than is
        possible in man, (3) protecting human subjects from unwarranted exposure, (4) permitting a
30      compression of exposure  duration in relation to aging as a result of the  shorter life

        March 12, 1990                         10-1      DRAFT-DO NOT QUOTE OR CITE

-------
        expectancies of laboratory animals, and (5) providing tissues, organs, and cellular material
        more readily, allowing more precise observation of specific functions.
             Fortunately, our knowledge of the influence of CO on biological systems is not limited
        to studies on nonhuman animals. Many direct experiments on humans have been conducted
 5      during the last century.  Although many reports describe inadvertent exposures to various
        levels of CO, there are a considerable number of precise and delineated studies utilizing
        human subjects.  Most of these have been conducted by exposing young adult males to
        concentrations of CO equivalent to those frequently or occasionally detected during ambient
        monitoring.  Research on human subjects,  however, also can be limited by methodological
10      problems.  As with the literature on experimental laboratory animals, many methodological
        and reporting problems make the data difficult to interpret. These problems include
        (1) failure to measure blood COHb levels;  (2) failure to distinguish between the physiological
        effects from a CO dose of high concentration  (i.e.,  bolus effect) and the slow, insidious
        increment in COHb over time from lower inhaled CO concentrations; (3)  failure to
15      distinguish between normal blood flow and blood flow increased in response to hypoxia
        (compensatory responses); and (4) the use of small numbers of experimental subjects. Other
        factors involve failure to provide control measures (e.g., double-blind conditions) for
        experimenter bias and experimenter effects; control periods so that task-learning effects do not
        mask negative results; homogeneity in the subject pool, particularly  in groups labeled
20      "smokers"; control of possible boredom and fatigue effects; and poor or inadequate statistical
        treatment of the data. In this chapter, an effort will be made to account for such
        methodological and reporting problems whenever possible  by making appropriate comments
        in the text.  Contributors to this chapter are limited, however, by the data provided in the
        reports published in the peer-reviewed literature.  For example, information on the COHb
25      levels achieved and the duration of exposure utilized in the studies will be provided in the text
        or tables if they were available in the original manuscript.  Where this information is lacking,
        only the  CO levels (parts per million) will be reported.
            An almost universal problem in research on both humans and laboratory animals is the
        use of inappropriate statistical techniques for data analysis.  Experimenters commonly use
30      tests designed for simple two-group designs when analysis  of variance is required, or use of
        several univariate tests when more than one dependent variable is measured and multivariate

        March 12,  1990                          10-2      DRAFT-DO NOT QUOTE OR CITE

-------
        tests are inappropriate.  Such statistical problems usually yield results in which the p- value is
        too small, so that possibility exists that too many results were falsely declared to be
        statistically significant.  Possible consequences of such errors will be discussed in the text or
        appropriate corrections will be made.  Unless actual p values are given, all statements of
 5      effects reported in the text or tables are statistically significant at p<0.05.
             One problem that is particularly unique to human research is that only low levels of CO
        exposure are commonly used. In such instances of low-level exposure, research findings
        necessarily deal with near-threshold effects.  When research, by necessity, is restricted to
        such barely noticeable effects it  may be expected that (1) results will be more variable because
10      of statistical sampling fluctuations, and (2) other uncontrolled variables that also affect the
        dependent variable in question will be of major importance and will increase the variability of
        results.  For these reasons, data on human subjects, although being of prime interest, also will
        be of highest variability.  Such high variability must be resolved with (1) large groups of
        subjects,  (2) theoretical interpretation of results relying on knowledge gained from
15      experimental laboratory animal data, and (3) consideration of consistency of the data within
        and across experiments.
             This chapter is intended to review available data from published studies in which both
        humans and laboratory animals have been exposed to low levels of CO. The chapter is
                                                   4
        divided according to specific health effects starting with pulmonary and cardiovascular effects.
20      The neurobehavioral effects of CO are described next, followed by developmental toxicity and
        other systemic effects of CO.  Finally, adaptation to CO exposure is discussed. An
        introduction and summary is provided for each major section of the chapter in order to set the
        tone for a clearer understanding  of the health effects of CO.  Although human and laboratory
        animal data may be presented separately under each effect category, the summary and
25      conclusion of these sections makes an attempt to integrate the relevant material from each of
        these types of studies.
        March 12, 1990                           10-3      DRAFT-DO NOT QUOTE OR CITE

-------
        10.2  ACUTE PULMONARY EFFECTS OF CARBON MONOXIDE
        10.2.1  Introduction
            The binding of carbon monoxide to hemoglobin, producing COHb, decreases the
        O2-carrying capacity of blood and interferes with O2 release at the tissue level; these two main
 5      mechanisms of action underly the potentially toxic effects of low-level CO exposure (see
        Chapter 9).  Impaired delivery of O2 can interfere with cellular respiration and result in tissue
        hypoxia. Hypoxia of sensitive tissues, in turn, can affect the function of many organs
        including the lungs.  The effects would be expected to be more pronounced under conditions
        of stress, as with exercise, for example.  Although the physiological mechanism by which
10      adverse effects of COHb formation are well known, CO-induced toxicity at the cellular level
        and its related biochemical effects still are not fully understood.   Other mechanisms of CO-
        induced toxicity have been hypothesized, but none have been demonstrated to operate at
        relatively low (near-ambient) CO exposure levels.  The effect of CO on cytochromes involved
        in cellular oxidative pathways is just one of the possible mechanisms of action of CO.
15      Mitochrondia, the principal site of oxygen utilization, are present in parenchymal lung cells
        and the highest concentrations are found in the type 2 epithelial cell.  Prolonged exposure to
        low levels of CO, therefore, may potentially interfere with cell function and cause loss of
        alveolar epithelial integrity.
            This section will review the available literature on morphological effects of CO and
20      determine if it is likely that CO can  cause direct toxicity to cells lining the respiratory tract
        through an effect on O2 transport or  cellular metabolism.  In addition, this section will review
        a predominately newer data base on  the effects of CO on pulmonary function.

        10.2.2  Effects on Lung Morphology
25          Reports appearing in the published literature have investigated the  histotoxic effects of
        CO on lung parenchyma and vasculature, an area not reviewed in the previous criteria
        document (U.S. Environmental  Protection Agency, 1979). Results from human autopsies
        have indicated that severe pulmonary congestion and edema was produced in the lungs of
        individuals who died from acute smoke inhalation resulting from fires (Burns et al., 1986;
30      Fein et al.,  1980). These individuals,  however, were exposed to relatively high
        concentrations of CO as well as other combustion components of smoke, such as carbon

        March 12, 1990                          10-4     DRAFT-DO NOT QUOTE OR CITE

-------
        dioxide, hydrogen cyanide, various aldehydes (e.g., acrolein), hydrochloric acid, phosgene,
        and ammonia (see Section 11.3.2).  If CO, contained in relatively high concentrations in the
        inhaled smoke, was responsible for the pathological sequelae described in fire victims, then to
        what extent can edema be attributed to the primary injury of capillary endothelial or alveolar
 5      epithelial cells?

        10.2.2.1  Studies in Laboratory Animals
            Laboratory animal studies by Niden and Schulz (1965) and Fein et al. (1980) found that
        very high levels of CO (5000 to 10,000 ppm) for 15 to 45 min were capable of producing
10      capillary endothelial and alveolar epithelial edema in rats and rabbits, respectively. Evidence
        of increased capillary permeability to protein also was reported in early studies on human
        subjects by Siggaard-Andersen et al. (1968) and Parving (1972) following acute, high-level
        CO exposure. These effects of CO have not been reported, however, at lower levels of CO
        exposure.
15          In a small number (n  = 5) of New Zealand white rabbits, Fein et al. (1980) reported a
        significant increase in the permeability of 51Cr-EDTA from alveoli to arterial blood within
        15 min after the start of exposure to 0.8% (8000 ppm) CO.  Passage of this labeled marker
        persisted and increased throughout the remaining 30 min of the study.  The mean COHb level
        after exposure was 63 + 4  (SEM) percent.  Although morphometric examination was not
20      performed, transmission electron microscopy (TEM) showed evidence of capillary endothelial
        and alveolar epithelial swelling and edema along with detachment of the endothelium from the
        basement membrane.  Mitochondria were disintegrated and alveolar type 2 cells were depleted
        of lamellar bodies.  None of these effects were found in four control animals exposed to air.
            Despite an increase in gross lung weight, Penney et al. (1988a) were unable to
25      demonstrate any evidence of edema in the lungs of male albino rats after 7.5 weeks of
        exposure to incrementally increasing concentrations of CO ranging from 250 to 1300 ppm.
        The authors also reported that this effect was not due to increased blood volume in the lung
        nor due to fibrosis, as measured by lung hydroxyproline content.  There was, therefore, no
        obvious explanation for the lung hypertrophy reported in this study after chronic exposure to
30      high concentrations of CO.
       March 12,  1990                          10-5      DRAFT-DO NOT QUOTE OR CITE

-------
            Fisher et al. (1969) failed to find any histologic changes in the lungs of mongrel dogs
       exposed to CO concentrations of 8000 to 14,000 ppm for 14 to 20 min (up to 18% COHb).
       Similarly, no morphological changes were found by Hugod (1980) in the lungs of adult
       rabbits continuously exposed to 200 ppm CO for up to six weeks (range of 11.9 to 19%
 5     COHb) or to  1900 ppm CO for five hours (range of 31 to 39% COHb).
            Niden (1971) speculated about possible effects of low levels of CO on cellular oxidative
       pathways when he reported that exposure of mice to concentrations of CO from 50 to 90 ppm
       for one to five days, resulting in COHb levels of < 10%, produced increased cristae in the
       mitochondria and dilation of the smooth endoplasmic reticulum in the nonciliated bronchiolar
10     (Clara) cell.  Minimal changes, consisting of fragmentation of lamellar bodies, were found in
       the type 2 epithelial cell.  Morphological appearance of the remaining cells of the terminal
       airways was normal. The results of this study were not presented in detail, however, and
       have not been confirmed at low concentrations of CO.  Thus, the significance, if any, of
       changes in the structure of cells lining the terminal airways is unknown.
15          Weissbecker et al. (1969) found no significant changes in the viability of alveolar
       macrophages  exposed in vitro to high concentrations of CO (up to 190,000 ppm). These
       results were later confirmed in more extensive in vivo exposure studies by Chen et al. (1982).
       They obtained alveolar macrophages by bronchoalveolar lavage from rats exposed to
       500 ppm (41  to 42% COHb) from birth through 33 days of age.  Morphological and
20     functional changes in the exposed cells were minimal.  There were no statistically significant
       differences in cell number, viability, maximal diameter, surface area, or acid phosphatase
       activity.  The phagocytic ability of alveolar macrophages was enhanced by CO exposure, as
       determined by a statistically significant (p<0.05) increase in the percentage of spread forms
       and cells containing increased numbers of retained latex particles.  The  biological
25     significance of this finding is questionable, however, because very few (n =  5) animals were
       evaluated and no follow-up studies have been performed.
       March 12, 1990                          10-6     DRAFT-DO NOT QUOTE OR CITE

-------
       10.2.2.2 Studies in Humans
            In a study by Parving (1972) on 16 human subjects, transcapillary permeability to
       131I-labeled human serum albumin increased from an average 5.6% per hour in controls to
       7.5% per hour following exposure to CO.  The subjects were exposed for three to five hours
 5     to 0.43% (4300 ppm) CO, producing approximately 23% COHb.  There were no associated
       changes in plasma volume, hematocrit, or total protein concentration.
            The only other relevant permeability studies were conducted on cigarette smoke.  Mason
       et al. (1983) showed rapidly reversible alterations in pulmonary epithelial permeability
       induced by smoking using "TcDTPA as a marker.  This increased permeability reverted to
10     normal fairly rapidly when subjects stopped smoking (Minty et al., 1981).  Using a rat
       model, the permeability changes associated with cigarette smoke were demonstrated later by
       Minty and Royston (1985) to be due to the paniculate matter contained in the smoke. The
       increase in ""TcDTPA clearance observed after exposure to dilute whole smoke did not occur
       when the particles were removed, suggesting that the CO contained in the gaseous phase does
15     not alter permeability of the alveolar-capillary membrane.

       10.2.3  Effects on Lung Function
       10.2.3.1 Lung Function in Laboratory Animals
            Laboratory animal studies of lung function changes associated with  CO exposure parallel
20     the morphology studies previously described (see Section 10.2.2) because high concentrations
       (1500-10,000 ppm) of CO were utilized.
            Fisher et al. (1969) ventilated the left lung of seven dogs  with 8 to  14% CO for 14  to
       20 min.  Femoral artery blood COHb levels ranged from 8 to 18% at the end of CO
       breathing.  No changes in the diffusing capacity or pressure-volume characteristics of the lung
25     were found.
            Fein et al. (1980) measured lung function in the same study discussed in
       Section  10.2.2. Nine New Zealand white rabbits were exposed for 45 min to either 0.8% CO
       or air. After CO exposure, COHb levels reached 63%.  Dynamic lung compliance
       significantly decreased and airway resistance significantly increased at 15 and 30 min after the
30     start of CO exposure, respectively. Mean blood pressure fell to 62% of the baseline value by
       March 12, 1990                          10-7      DRAFT-DO NOT QUOTE OR CITE

-------
       the end of exposure; heart rate was not changed.  Arterial pH decreased progressively
       throughout exposure although there were no changes in the alveolar-arterial PO2 difference.
            Robinson et al. (1985), also interested in the effects of acute CO poisoning in humans,
       used mongrel dogs to examine ventilation (VJ and perfusion (Q) distribution during and
 5     following CO exposure. A small number (n = 5) of mongrel dogs were exposed to 1 % CO
       (10,000 ppm) for 10 min,  resulting in peak COHb levels of 59 ± 5.4%. Inert gas
       distributions were measured at peak exposure and 2,4, and 24 h after exposure.  No changes
       in VA/Q  were found.  Previous studies were unable to show accumulation of lung water in
       the same model (Halebian  et al.,  1984a,b).  The authors concluded that other constituents of
10     smoke, besides CO, were responsible for the pulmonary edema and VA/Q mismatching
       found in victims exposed to smoke in closed-space fires.
            Very little is known about the effects of CO on ventilation in laboratory animals and the
       few studies available are contradictory.  No effects of CO on ventilation were found in
       unanesthetized rabbits (Korner, 1965) or cats (Neubauer et al., 1981), while large increases
15     were reported in conscious goats  (Chapman  et al., 1980; Doblar et al., 1977; Santiago and
       Edelman,  1976).  In anesthetized cats, high  concentrations of CO (10,000 ppm) increased
       ventilation (Lahiri and Delaney, 1976).  Gautier and Bonora (1983) used cats to compare the
       central effects of hypoxia on control of ventilation under conscious and anesthetized
       conditions. The cats were exposed for 60 min to either low inspired O2 fraction
20     (FIO2 = O.l 15) or CO diluted in  air.  In conscious cats, 1500 ppm CO caused a decreased
       ventilation, while higher concentrations  (2000 ppm) induced first a small decrease followed
       by tachypnea that is typical of hypoxic hypoxia in carotid-denervated conscious animals. In
       anesthetized cats, however, CO caused only mild changes in ventilation.
            Other respiratory effects of CO hypoxia, such  as the increased total pulmonary
25     resistance estimated by trachea! pressure, have been reported in anesthetized laboratory rats
       and guinea pigs (Mordelet-Dambrine et  al.,  1978; Mordelet-Dambrine and Stupfel, 1979).
       The significance of this effect is unknown,  however, particularly under the extremely high
       CO exposure conditions utilized in these studies (4 min inhalation of 2.84% CO) that
       produced  COHb concentrations >60%  (Stupfel et al., 1981). Similar increases in tracheal
30     pressure also were seen with hypoxic hypoxia (FIO2  = 0.89), suggesting a possible general
       mechanism associated with severe tissue hypoxia.

       March 12, 1990                          10-8      DRAFT-DO NOT QUOTE OR CITE

-------
        10.2.3.2  Lung Function in Humans
             Human studies of pulmonary function mostly are devoted to the identification of effects
        occurring in the lungs of individuals exposed to relatively high concentrations of CO. Older
        studies in the literature describe the effects of brief, controlled experiments with high CO-air
 5      mixtures. Chevalier et al. (1966) exposed 10 subjects to 5000 ppm CO for 2 to 3 min until
        COHb levels reached 4%. Measurements of pulmonary function and exercise studies were
        performed before and after exposure. Inspiratory capacity and total lung capacity decreased
        7.5 (p<0.05) and 2.1% (p<0.02), respectively, while maximum breathing capacity increased
        5.7% (p<0.05) following exposure.  Mean resting diffusing capacity of the lungs decreased
10      7.6% (p<0.05) compared to air-exposed controls. Fisher et al. (1969) exposed a small
        number (n = 4) of male subjects, aged 23 to 36 years,  to 6000 ppm CO for 6 s, resulting in
        estimated COHb concentrations of 17 to 19%. There were no significant changes in lung
        volume,  mechanics, or diffusing capacity. Neither of these studies was definitive, however,
        and no follow-up studies were reported.
15           More recent studies in the literature describing effects of CO on pulmonary function
        have been concerned with exposure to the products of combustion and pyrolysis from such
        sources as tobacco, fires, or gas- and kerosene-fueled appliances and engines. One group of
        individuals, representing the largest proportion of the population exposed to CO, is tobacco
        smokers.  The reader is referred to Section 11.4 for a discussion on environmental tobacco
20      smoke and to other reviews on the direct effects of smoking.
             A second group evaluated for potential changes in acute ventilatory function includes
        occupations where individuals are exposed to variable, and often unknown, concentrations of
        CO in both indoor and outdoor environments (see Section 8.4  for a more complete discussion
        of occupational exposure to CO).  Firefighters, tunnel workers, and loggers are typical
25      examples of individuals at possible risk. Unfortunately, as described above in
        Section 10.2.2, these individuals also are exposed to high concentrations of other combustion
        components of smoke and exhaust.  It is very difficult to separate the potential effects of CO
        from  those due to other respiratory irritants (see Section 11.3.2 for more complete discussion
        of exposure to combustion products).
30          Firefighters previously have been shown to have a greater loss of lung function
        associated with acute and chronic exposure to smoke inhalation (as reviewed by Sparrow

        March 12,  1990                          10-9       DRAFT-DO NOT QUOTE OR CITE

-------
       et al.,  1982).  None of these earlier studies, however, characterized the exposure variables,
       particularly the concentrations of CO found in smoke, nor did they report the COHb levels
       found in firefighters after exposure. Most reports of lung function loss associated with other
       occupational exposures also fail to characterize exposure to CO.  The following studies have
 5     attempted to monitor, or at least estimate, the CO and COHb levels found in occupational
       settings where lung function also was measured.
            Sheppard et al. (1986) reported that acute decrements in lung function were associated
       with routine firefighting.  Baseline airway responsiveness to methacholine was measured in
       29 firefighters from one fire station in San Francisco, CA, who were monitored over an
10     eight-week period.  Spirometry was measured before and after each 24-h workshift and after
       each fire.  Exhaled gas was sampled 55 times from 21 firefighters immediately after each fire
       and analyzed for CO.  Despite the use of personal respiratory protection, exhaled CO levels
       exceeded 100 ppm on four occasions, with a maximum of 132 ppm, corresponding to
       calculated COHb values of 17 to 22%.  Of the 76 spirometry measurements obtained within
15     two hours after a fire,  18 showed a greater fall in forced expiratory volume (FEV,) and/or
       forced vital capacity (FVC) compared to routine workshifts without fires. Decrements in lung
       function persisted for as long as 18  h in some of the individuals, but they did not appear to
       occur selectively in those individuals with preexisting airway hyperresponsiveness.
            Evans et al. (1988) reported on changes in lung function and respiratory  symptoms
20     associated with exposure to automobile exhaust among bridge and tunnel officers. Spirometry
       and symptom questionnaires were administered on a voluntary basis to 944 officers of the
       Triborough Bridge and Tunnel Authority in New York City over an 11-year period between
       1970 and  1981.  Regression analyses were performed on 466 individuals (49%) who had been
       tested at least three times during that period. Carboxyhemoglobin levels were calculated from
25     expired-air breath samples.  Small,  but significant differences were found between the bridge
       and tunnel officers.  Estimated levels of COHb were consistently higher in tunnel workers
       compared to bridge workers for both nonsmoking individuals (1.96 and 1.73%, respectively)
       and smoking individuals (4.47 and 4.25%, respectively).  Lung function measures of FEV,
       and FVC were reduced,  on an average, in tunnel versus bridge workers. There were no
30     reported differences in respiratory symptoms except for a  slightly higher symptom prevalence
       in tunnel workers who smoked.  Because differences in lung function between the two groups

       March 12, 1990                         10-10     DRAFT-DO NOT QUOTE OR CITE

-------
       were small, it is questionable if the results are clinically significant or if they were even
       related to CO exposure.
            Hagberg et al. (1985) evaluated the complaints of 211 loggers reporting dyspnea and
       irritative symptoms in their eyes, nose, and throat after chain-saw use.  Measurements of lung
 5     spirometry, COHb, and exposure to CO, HCs, and aldehydes were conducted on 23 loggers
       over 36 work periods lasting two hour each.  Ventilation levels during tree felling averaged
       41 L/min. Carboxyhemoglobin  levels increased after chain-saw use (p<0.05) but were
       weakly correlated (r  = 0.63) with  mean CO concentrations of 17 ppm (4 to 73 ppm range) in
       nonsmokers.  Corresponding COHb levels were apparently  <2%;  unfortunately, the absolute
10     values before and after exposure were not reported. Peripheral bronchoconstriction, measured
       by a decreased FEV/FVC (p<0.03) and forced expiratory  flow (FEF)^^ (p< 0.005), was
       found after the work periods but no correlations were obtained between lung function, COHb
       levels, and exposure variables. There were no reported changes in FEV,  or FVC.
            High CO concentrations  also  can be found indoors near unvented space heaters (see
15     Section 7.2).  The potential effects on lung function by indoor combustion products of
       kerosene space heaters was evaluated by Cooper and Alberti (1984). Carbon monoxide and
       SO2 concentrations were monitored in 14 suburban homes in Richmond, VA, during January
       and February of 1983 while modern kerosene heaters were in operation.  Spirometry was
       measured in 29 subjects over a two-day period, randomizing exposures between days with and
20     without the heater on. During heater operation, CO concentration was 6.8 ± 5.9 ppm (0 to
       14 ppm range), and SO2 concentration was 0.4 ± 0.4 ppm (0 to 1  ppm range).  On control
       days, indoor CO concentration was 0.14 + 0.53 ppm, whereas SO2 was undetectable.  Six of
       the homes had CO concentrations exceeding the NAAQS primary eight-hour standard of
       9 ppm.  Corresponding outdoor  CO concentrations were 0 to 3 ppm. Carboxyhemoglobin
25     levels significantly increased from  0.82 ± 0.43% on control days to 1.11 ± 0.52% on days
       when kerosene heaters were used.  Exposure to heater emissions, however, had no effect on
       FVC, FEV,, or maximum mid-expiratory flow rate (MMFR).
            Most of the published community population studies on CO have investigated the
       relationship between  ambient CO levels and hospital admissions, deaths, or symptoms
30     attributed to cardiovascular diseases (see Section 10.3). Little epidemiological information is
       March 12, 1990                         10-11     DRAFT-DO NOT QUOTE OR CITE

-------
       available on the relationship between CO and pulmonary function, symptomatology, and
       disease.
            One study by Lutz (1983) attempted to relate levels of ambient pollution to pulmonary
       diseases seen in a family practice clinic in Salt Lake City,  UT, during the winter of 1980-
 5     1981, when heavy smog conditions prevailed. Data on patient diagnoses; local climatological
       conditions; and levels of CO, O3, and paniculate matter were obtained over a 13-week period.
       Pollutant levels were measured daily and then averaged for each week of the study; absolute
       values were not reported.  For each week, weighted simple linear regression and correlation
       analyses were performed.  Significant correlations (p = 0.01) between pollution-related
10     diseases and the environmental variables were found for paniculate (r = 0.79), O3
       (r = -0.67), percent of smoke and fog (r = 0.79), but not for CO (r = 0.43) or percent of
       cloud cover (r = 0.33).  The lack of a significant correlation with CO was explained by a
       small fraction (2%) of diagnoses for ischemic heart disease compared to a predominance of
       respiratory tract diseases such as asthma, bronchitis, bronchiolitis, and emphysema.
15          Daily lung function in a large community population exposed to indoor and outdoor air
       pollution was measured in Tucson, AZ, by Lebowitz et al. (1983a,b, 1984, 1985, 1987),
       Lebowitz (1984), and Robertson and Lebowitz (1984).  Subsets of both healthy subjects and
       subjects with asthma, allergies, and airway obstructive disease were drawn from a symptom-
       stratified, geographic sample of 117  middle-class households.  Symptoms, medication use,
20     and peak flow measurements were recorded daily over a two-year period.  Indoor and  outdoor
       monitoring was conducted in a random sample of 41 representative houses.  Maximum one-
       hour concentrations of O3, CO, and NO2 and daily levels of TSP, allergens, and meteoro-
       logical variables were monitored at central stations within one-half mile of each population
       subset.  Indoor pollutant measurements were made for particles and CO, indicating that gas
25     stoves and tobacco smoking were the predominant indoor  sources. Levels of CO were low,
       averaging less than 2.4 ppm indoors and 3.8 to 4.9 ppm outdoors. Spectral time series
       analysis was used to evaluate relationships between environmental exposure and pulmonary
       effects over time (Lebowitz et al., 1987; Robertson and Lebowitz, 1984). Asthmatics were
       the most responsive while healthy subjects showed no significant responses.  Outdoor O3,
30     NO2, allergens, meteorology, and indoor gas stoves were significantly related to symptoms
       and peak flow.

       March 12, 1990                         10-12      DRAFT-DO NOT QUOTE OR CITE

-------
        10.2.4  Summary
             Currently available studies on the effects of CO exposures producing COHb
        concentrations of up to 39% fail to find any consistent effects on lung parenchyma and
        vasculature (Hugod, 1980; Fisher et al., 1969) or on alveolar macrophages (Chen et al.,
  5     1982; Weissbecker et al., 1969).  The lack of significant changes in lung tissue is consistent
        with the lack of histologic changes in the pulmonary and coronary arteries (see
        Section 10.3.4).  Alveolar epithelial permeability to 51Cr-EDTA increased in rabbits (Fein
        et al., 1980) exposed to high concentrations of CO  (63% COHb), and increased capillary
        endothelial permeability to 131I-labeled human  serum albumin was reported in early human
 10     studies (Parving, 1972) following acute, high-level  CO exposure (23% COHb); however, no
        accumulation of lung water was found in dogs (Halebian et al., 1984a,b)  with COHb levels of
        59% and no edema was found in the lungs of rats chronically exposed to  CO concentrations
        as high as 1300 ppm (Penney et al., 1988a).  In addition, no changes in diffusing capacity of
        the lung were found in dogs with COHb levels up to  18% (Fisher et al., 1969). It is
 15     unlikely,  therefore, that CO has any direct effect on lung tissue except at  extremely high
        concentrations.  The capillary endothelial and alveolar epithelial edema found with high levels
        of CO exposure in victims of CO poisoning may be secondary to cardiac failure produced by
        myocardial hypoxia (Fisher et al.,  1969) or may be due to acute cerebral anoxia (Naeije
        etal., 1980).
20          Ventilatory responses to CO are related to the CO concentration as well as to the
        experimental conditions and the animal  species being studied. In conscious goats (Chapman
        et al., 1980; Doblar et al., 1977; Santiago  and Edelman, 1976) and cats (Gautier and Bonora,
        1983), after an initial depression, ventilation suddenly increases, particularly at high CO
        concentrations (>2000 ppm).  This response may result from the direct effects of hypoxia
25      (and possibly central acidosis) and/or a  specific CNS effect of CO (see Section  10.3).  No
        effects on ventilation and perfusion distribution were found, however,  in dogs exposed to 1 %
        CO for 10 min, resulting in COHb levels of 59% (Robinson et al., 1985).  At very high
        concentrations of CO (COHb >60%) total pulmonary resistance, measured indirectly by
        trachea! pressure, was reported to increase  (Mordelet-Dambrine et al.,  1978; Mordelet-
30      Dambrine and Stupfel, 1979).
       March 12, 1990                          10-13      DRAFT-DO NOT QUOTE OR CITE

-------
            Human studies on the pulmonary function effects of CO are complicated by the lack of
       adequate exposure information, the small number of subjects studied, and the short exposures
       explored. Occupational or accidental exposure to the products of combustion and pyrolysis,
       particularly indoors, may lead to acute decrements in lung function if the COHb levels are
 5      > 17% (Sheppard et al., 1986) but not at concentrations <2% (Evans et al., 1988; Hagberg
       et al., 1985; Cooper and Alberti, 1984). It is difficult, however, to separate the potential
       effects of CO from those due to other respiratory irritants in the smoke and exhaust.
       Community population studies on CO in ambient air have not found any relationships with
       pulmonary function, symptomatology, and disease (Lebowitz et al., 1987;  Robertson and
10     Lebowitz, 1984; Lutz, 1983).
       10.3  CARDIOVASCULAR EFFECTS OF CARBON MONOXIDE
       10.3.1  Introduction
15          Carbon monoxide exposure exerts deleterious effects in humans by several known
       mechanisms. Carbon monoxide combines with Hb to form COHb, which directly decreases
       the O2 content of blood.  In addition, CO shifts the oxyhemoglobin dissociation curve to the
       left, providing less O2 to the tissues at a given tissue PO2. Although no clinical studies have
       been done, in vitro studies suggest that CO also may exert a deleterious effect in man with
20     coronary artery disease (injured vascular endothelium) by inhibiting the effects of oxyhemo-
       globin on the action of acetylcholine (Ignarro et al., 1987).  Acetylcholine causes a release of
       endothelium-derived relaxing factor (EDRF).  Carbon monoxide exposure in patients with
       diseased endothelium could accentuate acetylcholine-induced vasospasm and aggravate silent
       ischemia.
25          This section will  discuss studies in man dealing with the effects in healthy individuals, in
       patients with heart disease, and in other susceptible populations.
       March 12, 1990                          10-14     DRAFT-DO NOT QUOTE OR CITE

-------
        10.3.2  Experimental Studies in Humans
        10.3.2.1  Cardiorespiratory Response to Exercise
        Effects in Healthy Individuals
             The most extensive human studies on the Cardiorespiratory effects of CO are those
  5      involving the measurement of O2 uptake during exercise. Tnese studies were discussed in the
        previous CO criteria document (U.S. Environmental Protection Agency, 1979), an addendum
        to that document (U.S. Environmental Protection Agency, 1984), and in other published
        reviews (Horvath 1981; Shephard, 1983,1984).
             Healthy young individuals were used in most of the studies evaluating the effects of CO
 10      on exercise performance (see Table 10-1); healthy older individuals were studied in only two
        (Raven et al.,  1974a; Aronow and Cassidy, 1975).  In these studies, O2 uptake during
        submaximal exercise for short durations (5 to 60 min) was not affected by COHb levels as
        high as 15 to 20% (Table 10-1). Under conditions of short-term maximal exercise, however,
        statistically significant decreases (3 to 23%) in  maximal O2 uptake (VO2 max) were found at
 15      COHb levels ranging from 5 to 20% (Klein et al., 1980; Stewart et al., 1978; Weiser et al.,
        1978; Ekblom and Huot, 1972; Vogel and Gleser, 1972; Pirnay et al., 1971). In another
        study by Horvath et al. (1975), the critical level at which COHb marginally influenced  VO2
        max (p<0.10) was approximately 4.3%. The  data obtained by Horvath's group and others
        are summarized in Figure 10-1.  There is a linear relationship between decline in VO2 max
20      and increase in COHb that can be expressed as percent decrease in VO2 max = 0.91 (%
        COHb) +  2.2 (U.S. Environmental Protection  Agency, 1979;  Horvath (1981).  Short-term
        maximal exercise duration also has been shown to be reduced (3 to 38%) at COHb levels
        ranging from 2.3 to 7% (Horvath et al., 1975;  Drinkwater et al.,  1974; Raven et al., 1974
        a,b; Weiser et al., 1978; Ekblom and Huot, 1972).  (See Table 10-1.)
25          Acute effects of cigarette smoke on maximal exercise performance are apparently  similar
        to those described above in subjects exposed to CO.  Hirsch  et al. (1985) studied the acute
        effect of smoking on the Cardiorespiratory function during exercise in nine healthy male
        subjects who were current smokers. They were tested twice-once after smoking three
        cigarettes per hour for five hours and once after not having smoked. The exercise tests were
30      done on a bicycle ergometer with analysis of gas exchange and intra-arterial blood gases and
       March 12, 1990                          10-15     DRAFT-DO NOT QUOTE OR CITE

-------
I
TABLE 10-1. SUMMARY OF EFFECTS OF CARBON MONOXIDE ON MAXIMAL AND
                SUBMAXIMAL EXERCISE PERFORMANCE
Exposure*'
50 and 100 ppm CO
4-h
treadmill exercise
at 85 % maximal
heart rate (HR)
50 ppm CO, 25 and
35 °C
5-min
treadmill exercise
to exhaustion
50 ppm CO, 35 °C
20-min
treadmill exercise
to exhaustion
50 ppm CO, 25 °C
5-min
treadmill exercise
to exhaustion
75 and 100 ppm CO
1 5-min
treadmill exercise
to exhaustion
100 ppm CO
Ih
treadmill exercise
to exhaustion
0.5% CO
2.5-3.5 min
5-min submaximal
exercise at
1.84L/minVO2
COHbc
2.17% (50 ppm)
4.15% (100 ppm)
2.3%
(nonsmokers)
5.1% (smokers)
2.5%
(nonsmokers)
4.1% (smokers)
2.7
(nonsmokers)
4.5% (smokers)
3.3-4.3%
3.95%
3.95%
Subjects(s)
23 males
20-38 years
(8 smokers)
16 males
40-57 years
(7 smokers)
20 young males
equally divided
by smoking
history
20 males
21-30 years
equally divided
by smoking
history
4 males
24-33 years
(1 smoker)
9 male
1 female
nonsmokers
44-55 years
10 nonsmokers
x = 30 years
Observed Effects'1
Mean exercise duration was
19 s shorter on CO days;
coagulation variables,
cholesterol, and triglycerides
were not significantly changed
•
No change in VO2 max; total
work time decreased at 25 °C in
older nonsmokers
•
No change in VO2 max; exercise
duration decreased in non
smokers; change in respiratory
pattern in both smokers and
nonsmokers
•
No change in VO2 max or work
time; no smoking effect
VO2 max decreased (p< 0.10) at
4.3% COHb; lower work times
and ventilatory volumes at all
COHb levels (p< 0.05)
Mean exercise time until
exhaustion decreased 5%
(p<0.001)
•
No change in mean VO2; O2 debt
per VO2 increased 14%
Conclusions
Submaximal exercise duration
decreased significantly at
4% COHb
No significant decrease in
VO2 max in older men exposed
to CO but work time to
exhaustion decreased in
nonsmokers at 2.3% COHb
Work time decreased in
nonsmokers at 2.5% COHb
No significant decrease in
maximal exercise performance
Maximal exercise performance
decreased at COHb >4%
Exercise time decreased
in older nonsmokers at
3.95% COHb
Work at 4% COHb was per-
formed with greater metabolic
cost
Reference*
Brinkhouse (1977)
Raven et al. (1974a)
Drinkwater et al. (1974)
Raven et al. (1974b)
Horvath et al. (1975)
Aronow and Cassidy
(1975)
Chevalier et al. (1966)

-------

0
1
e6




Conclusions
Observed Effects'1


Subjects(s

y
ffi
0
u





"a
g
1
.•s1
g
•a
«s
1
5

k o
"2 °
S 2 -S
No major change in i
respiratory response
submaximal work wi
levels <7%
Stroke volume decreased with
higher ambient temperature; HR
increased with CO exposure but
no change in cardiac output or
stroke volume


|l
G up
Ov 00


^
OO
vo
vq
Tf


C
CO
S *
CJ « x
S 8 |
&P 'g r.
|
1
X
|
g
o
3
3
O
II
Maximal exercise pei
and VO2 max decrea
increasing COHb
During maximal exercise, work
time and VO2 max significantly
decreased at 7-20% COHb; no
change in VO2 with submaximal
exercise


S $
8 !
?*
CO CO


*S
*"*?
£!
oo
Tf
X
£
e? « <5*
'•s E §•>
? a 'S *
.a 2 o S
H g S2

i?
•3
S
•§
3

£
Maximal O2 uptake
decreased at 5% COI
• *
VO2 max decreased while VE
and HR both increased


6 male non
nonsmokers
25-39 years




^
o
V)


o
30 ppm CO
5-h
treadmill exercis
until exhaustion
So1
2
•3
i,
'3
i
8
•gls
Maximal exercise pe:
in Denver, CO, (161
decreased at 5% CO!
Total exercise time decreased
3.8%; total work performed
decreased 10%, and VO2 max
decreased 2.8%
2
S
>>
|2 °8
1 1 «": | i
<* § S e (S




ss

»n


9
20-min rebreathi
to achieve target
COHb; treadmill
exercise until
exhaustion

Sj
•5
•s

«
4
i
li
Maximal exercise pe
decreased at 5.5% C
•
Maximal exercise time and VO2
decreased while HR and VE
increased


6 male
nonsmokers
25-39 years




&
"5
VI


4)
yi
O a U, 'g .-3
a » s B 1
i?
c
•a
.s
J
*
tf
Its
CO did not affect pu
function, subjective i
or exercise metabolU
k
W H
o „

1
ill
?
M
§ c
c S5 S3
S% -S *£
P) c V
r~ S oi



O 'S «
O u O
C O* ^
a -s *
>»°
*-t -1« j5 '«

B
*«3
o

.a
6

,
Maximal O2 uptake
decreased at 8% CO
•
VE and breathing
frequency (f^ jn-
creased while VO2 max
and (A-a) O2 difference
decreased with exercise


9 male
nonsmokers




g*
00
r~

_3
"o
?*-
if!
2" ll*
March 12, 1990
10-17    DRAFT-DO NOT QUOTE OR CITE

-------
cr
o
9
i— »
oo
                     TABLE 10-1 (cont'd). SUMMARY OF EFFECTS OF CARBON MONOXIDE ON MAXIMAL AND
                                                 SUBMAXIMAL EXERCISE PERFORMANCE
Exposure'-1" COHb'
15-min rebreathing 12.8-15.8%
to achieve target
COHb; exercise at
30, 70, and 100%
VO2 max
0.05% CO 15.4%
5-min
moderate exercise
for 15 min at
4 km/hr
0.15-0.35% CO 16-52%
>70 min
225 ppm CO 18-20%
1-h
bicycle exercise
aj 50, 75, and 100%
VO2max
225 ppm CO 20.3%
1-h
bicycle exercise
at 45, 75, and 100%
VO2max
Subjects(s)
9 males
23-34 years
5 males
24-35 years
4 males
21-33 years
8 males
20-23 years
(3 smokers)
16 males
(6 smokers)
Observed Effects'1
VO2 max decreased 14.2% with
maximal exercjse; no change in
ventilation or VO2 with sub-
maximal exercise
Increased HR but no change in
VO2 or ventilation with sub-
maximal exercise; VO2 max
decreased 15.1%
No hyperpnea at rest; arterial
PCO2 increased and pH
decreased; cardiac output
increased with increasing COHb
VO2 max decreased 23%
(p<0.001); with submaximal
exercise HR increased (p<0.05)
while VO2 was unchanged
VO2 max decreased 24%
(p<0.001); no change in work
efficiency or VO2 with
submaximal exercise
Conclusions
Maximal exercise performance
decreased after CO exposure
Maximal O2 uptake
decreased at 15% COHb
CO has a depressive action on
the respiratory center
Maximal O2 uptake
decreased at >18% COHb
Maximal O2 uptake
decreased at >20% COHb
Reference"
Ekblom et al. (1975)
Pirnay et al. (1971)
Chiodi et al. (1941)
Vogel and Gleser (1972)
Vogel et al. (1972)
       'Exposure concentration, duration, and activity level.
       bl ppm = 1.145 mg/m3; 1 mg/m3 = 0.873 ppm at 25°C, 760 mm Hg; \%
       'Estimated or measured blood carboxyhemoglobin (COHb) levels.
       dSee glossary of terms and symbols for abbreviations and acronyms.
       'Cited in U.S. Environmental Protection Agency (1979; 1984).
10,000 ppm.

-------
X
o



 IN
     40
     35-
     30-
 Z   25-1

 LJ

 §20-1
o:
O
     is-
     le-
      s'
                         10      15      20       25

                                PERCENT COHb
                                                         30
35
40
    Figure 10-1. Relationship^ between carboxyhemoglobin level (COHb) and decrement in

    maximal oxygen uptake (VO2 max) for healthy nonsmokers.


    Source: Adapted from U.S. Environmental Protection Agency (1979) and Horvath (1981).
March 12, 1990
                                  10-19     DRAFT-DO NOT QUOTE OR CITE

-------
       pressures. On the smoking day, the maximal O2 uptake was significantly decreased by 4%
       and the anaerobic threshold was decreased by 14%.  The rate-pressure product was a
       significant 12% higher at comparable work loads of 100 watts on the smoking day compared
       to the nonsmoking day. There were no changes due to smoking,  however, on the duration of
 5     exercise or on the mean work rate during maximal exercise testing. The blood COHb level
       before exercise was 1.8% on the nonsmoking and 6.6% on the smoking day.  At peak
       exercise the COHb was 0.9% and 4.8%, respectively, on the nonsmoking and smoking day.
       The authors concluded that the main adverse effect of smoking was due to CO, although the
       increase in rate-pressure product also might be the result of the simultaneous inhalation of
10     nicotine. They felt that the magnitude of change in performance indicators corresponded well
       with earlier reports.
            It would be interesting, therefore, to determine if smokers and nonsmokers had different
       responses to CO exposure.  Unfortunately, smokers  and nonsmokers were not always
       identified in many of the studies on exercise performance, making it difficult  to interpret the
15     available data. Information derived  from studies on cigarette smoke is also sparse.   As a
       result, attempts to sort out the acute effects of CO from those due to other components of
       cigarette smoke have been equivocal.  Seppanen (1977) reported that the physical work
       capacities of cigarette smokers decreased at 9.1% COHb levels after breathing either boluses
       of 1100 ppm CO or after smoking cigarettes. The greatest decrease in maximal work,
20     however, was observed after CO inhalation.
            Klausen et al. (1983) compared the acute effects of cigarette smoking and inhalation of
       CO on maximal exercise performance.  They studied 16 male smokers under  three different
       conditions: after eight hours without smoking (control), after inhalation of the smoke of three
       cigarettes, and after CO inhalation.  Just before maximal exercise testing the arterial CO
25     saturation reached 4.51 and 5.26% after cigarette smoke and CO inhalation, respectively,
       compared to 1.51% for controls.  Average maximal O2 uptake decreased by about 7% with
       both smoke and CO.  Exercise time, however, decreased 20% with smoke but only  10% with
       CO, suggesting that nicotine, smoke particles, or other components of tobacco smoke may
       contribute to the observed effects. The authors, therefore, concluded  that a specified COHb
30     level induced by either smoke or CO decreased maximal work performance to the same
        March 12, 1990                         10-20     DRAFT-DO NOT QUOTE OR CITE

-------
        degree. Of note is the more marked decrease in work time compared to maximal O2 uptake
        induced by CO, a finding that agrees with the Ekblom and Huot (1972) results (see
        Table 10-1).
            If the magnitude of the effect of CO exposure is due only to a decrease in O2-carrying
 5      capacity proportional to the COHb concentration, the magnitude should be roughly the same
        as if the O2 capacity is decreased by anemia. Celsing et al. (1987) found in a series of very
        carefully performed studies in normal subjects that maximal O2 uptake decreased by
        19 mL/min/kg per gram per liter change in Hb over a wide range of Hb concentrations from
        137 to 170 g/L.  This decrease corresponds to a 2%  decrease in maximal O2 uptake for every
10      3% decrease in Hb concentration in a well-trained subject.  The decrease also corresponds to
        the decrease in VO2 max reported by Ekblom and Huot (1972) and Horvath et al. (1975).
        However, Ekblom and Huot found a much more marked effect on maximal work time (i.e.,
        work on a constant load until exhaustion with a duration of about 6 min).  An explanation for
        the marked decrease in maximal work time could be that CO has a negative effect on the
15      oxidative enzymatic system wheareas the decrease in work time is due to a combination of a
        decrease in O2 capacity and a less efficient oxidative enzymatic system.  If the data are
        extrapolated to lower COHb values, a 3%  level of COHb should decrease the maximal work
        time by about 20%. However, this decrease is more than the 10% average decrease reported
        by Klausen et al. (1983)  where they also found more marked effects in less well-trained
20      subjects compared to well-trained subjects.

        Effects in Individuals with Heart Disease
            The previous criteria document (U.S. Environmental Protection Agency,  1979)
        concluded that patients with heart disease are especially at risk to CO exposure sufficient to
25      produce 2.5 to 3% COHb.  This statement was based primarily on studies initiated by
        Aronow et al. (1972) and Aronow and Isbell (1973) demonstrating that patients with angina
        pectoris, when exposed to low levels of CO, experienced reduced time to onset of exercise-
        induced chest pain as a result of insufficient O2 supply to the heart muscle.  A study by
        Anderson et al. (1973) reported similar results at mean COHb levels of 2.9 and 4.5% (see
30      Table 10-2).
       March 12, 1990                          10-21      DRAFT-DO NOT QUOTE OR CITE

-------

S
to
100 ppm CO for
60 min; postexposure
incremental exercise
at 48.6 L/min on
a cycle ergometer
        117 or 253 ppm CO
        for 50-70 min;
        pre- and post-
        exposure incremental
        exercise at -6 METS
        on a treadmill
        (modified Naughton
        protocol)
                                TABLE 10-2.  SUMMARY OF EFFECTS OF CARBON MONOXIDE EXPOSURE IN
                                                                   PATIENTS WITH ANGINA
Exposure"'1" COHbc COHbd
50 or 100 ppm CO 2.9% (SP) ND
for 50 min of 4.5% (SP)
each hour x 4 h;
postexposure
exercise on a
treadmill



ACOHb" Subject(s)
1.6% 10 males, 5 smokers
3.2% and 5 nonsmokers,
with reproducible
exercise-induced
angina; 49.9 years




Observed Effects
Duration of exercise before onset
of angina was significantly
shortened at 2.9 and 4.5% COHb
(p<0.005); duration of angina was
significantly prolonged at 4.5%
(p <0.01) but not at 2.9% COHb.
The response of smokers was not
significantly different from that
of nonsmokers.
Reference
Anderson et at.
(1973)f







3.0% (CO-Ox)      2.8% (CO-Ox)       1.5%         24 male nonsmokers
                                                  with reproducible
                                                  exercise-induced
                                                  angina; 59 ± 1 years
                                                  (49-66 years)
                        3.2% (CO-Ox)      2.7% (CO-Ox)       2.0%         63 male nonsmokers
                        5.6% (CO-Ox)      4.7% (CO-Ox)       4.4%         with reproducible
                                                                          exercise-induced
                        2.4% (GC)         2.0% (GC)          1.8%         angina; 62 ± 8 years
                        4.7% (GC)         3.9% (GC)          4.0%         (41-75 years)
Time to onset of angina decreased
5.9% (p = 0.046); no significant
effect on the duration of angina.
O2 uptake at angina was
reduced about 3% (p = 0.04).
There were no significant changes
in heart rate or systolic blood
pressure at angina.

Earlier onset of myocardial
ischemia was found with CO exposure:
time to ST endpoint decreased 5.1 and
12.1 % (p <0.05) and time to angina
onset decreased 4.2 and 7.1%

-------
I
                     TABLE 10-2 (cont'd).   SUMMARY OF EFFECTS OF CARBON MONOXIDE EXPOSURE IN
                                                               PATIENTS WITH ANGINA
              Exposure"'11
                             COHbc
COHb"
                                                                ACOHb"
                                                                                    Subjects)
                                                                   Observed Effects
                                                                                                      Reference
         100-200 ppm CO
         for 60 min; postex-
         posure incremental
         exercise at 317 KPM
         on a cycle ergometer
                         3.8% (CO-Ox)       3.6% (CO-Ox)       2.2%          25 male and 5 female
                                                                              nonsmokers with
                                                                              evidence of exercise-
                                                                              induced angina on at
                                                                              least one day; 58  ±
                                                                              11 years (36-75 years)
                                                            No significant difference in time
                                                            to onset or duration of angina. No
                                                            significant difference in maximal
                                                            exercise time, maximal ST segment
                                                            depression, or time to significant
                                                            ST segment depression during
                                                            exercise.  No significant difference
                                                            in maximal ejection fraction; small
                                                            decreases in blood pressure
                                                            (p = 0.031) and change in ejection
                                                            fraction (p = 0.049)  during CO
                                                            exposure requires further evaluation.
                                                            Actuarial analysis (Bissette et al.,
                                                            1986) including 3/30 subjects
                                                            experiencing angina only on the CO
                                                            exposure day showed a significant
                                                            decrease in time to onset of angina
                                                            after CO exposure.
Sheps et al.
(1987)
O
o
z
3
I
s
o
         100-200 ppm CO
         for 60 min; postex-
         posure incremental
         exercise @300 KPM
         on a cycle ergometer
                         5.9% (CO-Ox)       5.2% (CO-Ox)       4.2%          22 male and 8 female
                                                                               nonsmokers with evidence
                                                                               of exercise-induced
                                                                               angina on at least one
                                                                               day; 58 ±  11 years
                                                                               (36-75 years)
                                                            Earlier onset of ventricular
                                                            dysfunction, angina, and poorer
                                                            exercise performance was found with
                                                            CO exposure; mean duration of
                                                            exercise was shorter (p<0.05);
                                                            subjects were likely to experience
                                                            angina earlier during exercise with
                                                            CO (p<0.05) using actuarial analysis.
                                                            Both the level (p = 0.05) and  change
                                                            in left ventricular ejection fraction
                                                            at submaximal exercise (p = 0.05)
                                                            were less on CO-exposure compared
                                                            to air-exposure day.  There was no
                                                            significant difference in the peak
                                                            exercise left ventricular ejection
                                                            fraction.
Adams etal. (1988)
"Exposure concentration, duration, and peak activity level. NOTE: Because oxygen consumption was not measured, it is not possible to determine the actual level of exercise at which angina occurred.
 1    = 1.145 mg/m3; 1 mg/m3 = 0.873 ppm at 25° C, 760 mm Hg; 1% = 10,000 ppm.
'Measured blood carboxyhemoglobin (COHb) level after CO exposure; GC = gas chromatograph; CO-Ox = CO-Oximeter; SP =  spectrophotometric method of Buchwald (1969).
dMeasured blood carboxyhemoglobin (COHb) level after exercise stress test; GC = gas chromatograph; CO-Ox = CO-Oximeter; ND = not determined.
"Postexposure increase in COHb over baseline.
'Cited in U.S. Environmental Protection Agency (1979; 1984).

-------
            In 1981, Aronow reported an effect of 2% COHb on time to onset of angina levels in
       15 patients.  The protocol was similar to previously reported studies, with patients exercising
       until onset of angina.  Only 8 of the 15 subjects developed more than 1 mm ischemic ST
       segment depression at the onset of angina during the control periods.  This was not
 5     significantly affected by CO. It is questionable, therefore, as to whether the remaining
       patients truly met adequate criteria for ischemia despite angiographically documented cardiac
       disease.  After breathing 50 ppm of CO for one hour, the patients' times to onset of angina
       significantly decreased from a mean of 321.7 ± 96 s to 289.2 ± 88 s.
            In 1983, the studies by Aronow and his colleagues were reevaluated by an ad hoc
10     committee to the EPA (especially the 1981 study).  The committee concluded that the results
       of Aronow's studies did not meet a reasonable standard of scientific quality and,  therefore,
       should not be used by the Agency in defining the critical COHb level at which adverse health
       effects of CO are occurring.  A summary of the committee findings and a reevaluation of the
       key health effects information reported to be associated with relatively low level  CO exposure
15     can be found in an addendum to the 1979 criteria document for CO (U.S. Environmental
       Protection Agency, 1984).
            The experimental design used in the Aronow studies on  CO exposure effects in patients
       with angina set the stage for subsequent, more precisely designed studies. Aronow and his
       colleagues used the subjective measure of time to onset of angina as their only significant
20     variable of CO effect. In an attempt to improve upon these earlier preliminary studies, the
       more recent studies employed electrocardiogram (EKG) changes as objective measures of
       ischemia. Another consideration in the conduct of the newer  studies on angina was to
       establish better the dose response relationships for low levels of CO exposure. While the
       COHb level is accepted as the best measure of the effective dose of CO,  the reporting of low
25     level effects is problematic because of inconsistencies in the rigor with which the devices for
       measuring COHb have been validated.  The most frequently used technique for measuring
       COHb has been the optical method found in the IL series of CO-Oximeters (CO-Ox).  Not
       only is there a lot of individual variability in these machines,  but recent comparisons with the
       gas chromatographic (GC) technique of measuring COHb have suggested that the optical
30     method may not be a suitable reference technique for measuring low levels of COHb.  (See
       Chapter 8,  Section 8.5 for more details.) Several additional studies have appeared in the

       March 12,  1990                         10-24     DRAFT-DO NOT QUOTE OR CITE

-------
        literature to help define better the precise COHb levels at which cardiovascular effects occur
        in angina patients (see Table 10-2).
            Sheps et al.  (1987) studied 30 patients with ischemic heart disease age 38 to 75 years
        and assessed not only symptoms during exercise, but radionuclide evidence of ischemia (left
 5      ventricular ejection fraction changes). This study was designed to be representative of a
        broad group of patients with myocardial ischemia.  Patients were nonsmokers with ischemia,
        defined either by exercise-induced ST segment depression, angina, or abnormal ejection
        fraction response (i.e., all patients had documented evidence of ischemia).
            Patients were exposed to CO (100 ppm) or air during a three-day, randomized double-
10      blind protocol to achieve a postexposure level of 4% COHb (CO-Ox measurement).  Resting
        preexposure levels were 1.7%, and postexposure levels averaged 3.8% on the CO exposure
        day, thus the study examined acute elevation of COHb levels from 1.7% to  an average of
        3.8%, or an average increase of 2.2% from resting values. Comparing exposure to CO to
        exposure to air, there was no significant difference in time to onset of angina, maximal
15      exercise time, maximal ST segment depression (1.5 mm for both), or time to significant ST
        segment depression.  The conclusion of this study was that 3.8% COHb produces no
        clinically significant effects on this patient population.
            Interestingly, further analysis of the time to onset of angina data in this paper
        demonstrated that 3 of the 30 patients experienced angina on the CO exposure day but not on
20      the air control day.  These patients had to be deleted from the classical analysis of differences
        between time to onset of angina that was reported in the publication. However, actuarial
        analysis of time to onset of angina including these patients revealed a statistically significant
        difference in time to onset of angina favoring an earlier time under the CO-exposure
        conditions (Bissette et al., 1986).  None of the patients had angina only on the air exposure
25      day.
            Subsequent work from these  same investigators (Adams et al., 1988) focused on
        repeating the study at 6% COHb (CO-Ox measurement) using an identical protocol and a
        similar patient population.  Postexposure COHb levels averaged 5.9 + 0.1% compared to
        1.6 + 0.1% after air exposure, representing an increase of 4.3%. The mean duration of
30      exercise was significantly longer after air compared to CO exposure (626 + 50 s for air vs.
        585 + 49 s for CO, p<0.05). Actuarial methods suggested that subjects experienced angina

        March 12,  1990                          10-25     DRAFT-DO NOT  QUOTE OR CITE

-------
       earlier during exercise on the day of CO exposure (p<0.05). In addition this study showed
       that, at a slightly higher level of CO exposure, both the level and change in ejection fraction
       at submaximal exercise were greater on the air day than on the CO day.  The peak exercise
       left ventricular ejection fraction, however, was not different for the two exposures.
 5          These results demonstrated earlier onset of ventricular dysfunction and angina and
       poorer exercise performance in patients with ischemic heart disease after acute CO exposure
       sufficient to increase COHb to 6%.  It is of interest that in both the 4% study and the 6%
       study reported by this group, some of the patients experienced angina on the CO day, but not
       on the air exposure day. There were no patients who experienced angina in the reverse
10     sequence.
            Kleinman and Whittenberger (1985) and Kleinman et al. (1989) studied nonsmoking
       male subjects with a history of stable angina pectoris and positive exercise tests.  All but two
       of the 26 subjects had additional confirmation of ischemic heart disease, such as previous
       myocardial infarction,  positive angiogram, positive thallium  scan, prior angioplasty, or prior
15     bypass surgery.  Subjects were exposed for one hour in a randomized double-blind crossover
       fashion to either 100 ppm CO or to clean air on two separate days.  Subjects performed an
       incremental exercise test on a cycle ergometer to the point at which they noticed the onset of
       angina.  For the study group, the one-hour exposure to 100 ppm CO resulted in an increase
       of COHb from 1.4% after clean air to 3% (CO-Ox measurement) after CO.  For the entire
20     study group (n = 26), the one-hour exposure to 100 ppm resulted in a decrease of the time to
       onset of angina by 6.9% from 6.5 to 6.05 min (Kleinman and Whittenberger, 1985). This
       difference was significant in a one-tailed paired T-test (p = 0.03).  When using a two-tailed
       test the difference loses statistical significance at the p = 0.05 level.
             In a subsequent publication of results from this study (Kleinman et al.,  1989), the two
25     subjects with inconsistencies in their medical records and histories were dropped from the
       analysis.  For this study group (n = 24), the one-hour exposure to 100 ppm CO (3% COHb
       by CO-Ox measurement) resulted in a decrease of time to onset of angina by 5.9%
       (p  = 0.046).  There was no significant effect on the duration of angina but O2 uptake at
       angina point was reduced 2.7% (p = 0.04). Only eight of the subjects exhibited depression
30     in the ST segment of their EKG traces during exercise. For this subgroup, there was a 10%
        March 12, 1990                          10-26      DRAFT-DO NOT QUOTE OR CITE

-------
        reduction (p< 0.036) in time to onset of angina and a 19% reduction (p< 0.044) in the time
        to onset of 0.1 mV ST segment depression.
             A multicenter study of effects of low levels of COHb has been conducted on a relatively
        large sample (n = 63) of individuals with coronary artery disease from three different cities
 5      (Allred et al., 1989a,b).  The purpose of this study was to determine the effects of 2.0% and
        3.9% COHb (GC measurement) on time of onset of significant ischemia during a standard
        treadmill exercise test.  Significant ischemia was measured subjectively by the time of
        exercise required for the development of angina (time of onset of angina) and objectively by
        the time required to demonstrate a 1-mm change in the ST segment of the ECG (time to ST).
10      Male subjects, ages 41 to 75 (mean = 62.1 years)  with stable exertional angina pectoris and
        positive stress test, as measured by > 1-mm ST segment change,  were studied.  Further
        evidence that these subjects had coronary artery disease was provided by the presence of at
        least one of the following criteria:  angiographic evidence of >70%  narrowing of at least one
        coronary artery, documented prior myocardial infarction, or a positive stress thallium
15      demonstrating an unequivocal perfusion defect.  Thus, as opposed to some previous studies
        reported, this study critically identified patients with documented  coronary artery disease.
            The protocol for this study was  similar to that used in the Aronow studies because two
        exercise tests were performed on the  same day.  The two tests were separated by a recovery
        period and double-blind exposure period.  On each of the three exposure days the subject
20      performed a symptom-limited exercise test on a treadmill, then was exposed for 50 to 70 min
        to the test environment  (clean air, 117 ppm CO, or 253 ppm CO), and then performed a
        second symptom-limited exercise test. The time to the onset of angina and the time to 1-mm
        ST change was determined for each test. The percent change following exposure at both 2.0
        and 3.9% COHb (GC measurement) then were compared to the same subject's response to the
25      randomized exposure to room air (less than 2 ppm CO.)
            When potential exacerbation of the exercise-induced ischemia by exposure  to CO was
        tested using the objective measure of  time to 1-mm ST segment change,  exposure to 2.0%
        COHb resulted in a 5.1 %  decrease (p = 0.01) in the time to attain this level of ischemia. At
        3.9% COHb the decrease in time to the ST criterion was  12.1% (p^O.OOOl) relative to the
30      air-day results.  At the 3.9% COHb level this reduction in time to ST depression was
        accompanied by a significant (p = 0.03) reduction  in the heart rate-blood pressure product

        March 12, 1990                          10-27      DRAFT-DO NOT QUOTE OR CITE

-------
       (double product), an index of myocardial work.  The maximal amplitude of the ST segment
       change also was significantly affected by the CO exposures:  at 2% COHb the increase was
       11% (p = 0.002) and at 3.9% COHb the increase was 17% relative to the air day
       (p 
-------
        drug interaction with the effects of CO). The major medications being used in this group
        were betablockers (38/63), nitrates (36/63), and calcium-channel blockers (40/63).  The other
        major concern was the influence of the severity of the disease. The simplest approach to this
        was to evaluate the influence of the duration of the exercise because the subjects with more
 5      severe disease were limited in their exercise performance. No significant correlation was
        found between duration of exercise and the change in the time to angina or ST criterion.
        There also was no relationship between the average time of exercise until the onset of angina
        and either of the endpoints. There also was no relationship between the presence of a
        previous myocardial infarction and the study endpoints.
10          The duration of exercise was significantly shortened by the 3.9%  COHb but not by the
        2.0%  level.  This finding must be used cautiously because these subjects were not exercised
        to their maximum capacity in the usual  sense.  The major reason for termination of the
        exercise was the progression of the angina (306/376 exercise tests.)  The subjects were to
        grade  their angina on a four-point scale, and when the exercise progressed beyond level two
15      they were stopped. Therefore this significant decrease in exercise time of 40 s at the 3.9%
        COHb level is undoubtedly due to the earlier onset of angina followed by the normal rate of
        progression of the severity of the angina.
            The individual center data provides insight into the interpretation of other studies that
        have been conducted in this area.  Each of the centers enrolled the numbers of subjects that
20      have been reported by  other investigators.  The findings reported above were not substantiated
        in all instances at each center.  When one considers the responses of the group to even the
        3.9%  COHb, it is clear as to why one might not find significance in one parameter or
        another.  For the decrease in ST segment at 3.9%, only 49 of 62 subjects demonstrated this
        effect  on the day tested.  The potential for finding significance at this effect rate with a
25      smaller sample size is reduced.  Random sampling of this population with a smaller sample
        easily  could provide subjects that would not show significant effects of these low levels of CO
        on the test day.
            The recent report (Allred et al., 1989b) of the multicenter study conducted by the Health
        Effects Institute  (HEI)  discusses some reasons for differences between the results of the
30      studies cited above (also see Table 10-2 and Table 10-3). The studies have different designs,
        types of exercise tests, inclusion criteria (and,  therefore, patient populations), exposure

        March 12, 1990                          10-29     DRAFT-DO NOT QUOTE OR CITE

-------
s
er
9
s
                   TABLE 10-3.  COMPARISON OF SUBJECTS IN STUDIES OF THE EFFECT OF CARBON MONOXIDE
                                         EXPOSURE ON OCCURRENCE OF ANGINA DURING EXERCISE
Subiect Characteristics
Number of
Study Subjects Gender
Anderson et al. (1973) 10 male
Kleinman et al. (1989) 24 male

Allied et al. 63 male
(1989a,b)
Medication
1 subject took
digitalis; drug
therapy basis
for exclusion
14 on betablockers;
19 on nitrates;
8 on Ca-channel
blockers
38 on betablockers;
36 on nitrates; 40
Smoking
History
5 smokers
(refrained for
12 h prior
to exposure)
No current
smokers

No current
smokers
Description of
Disease
Stable angina pectoris,
positive exercise test
(ST changes); reproducible
angina on treadmill
Ischemic heart disease,
stable exertional angina
pectoris

Stable exertional angina
and positive exercise test
Age
(years)
(mean = 49.9)
49-66
(mean = 59)

41-75
(mean - 62.1)
                                                               on calcium antagonists
        Sheps et al. (1987)
                                  30
                                  (23 with
                                  angina)
25 male
5 female
        Adams et al. (1988)
                                 30
                                                  22 male
                                                  8 female
26 subjects on
medication; 19 on beta
blockers; 11 on
Ca-channel blockers; 1
on long-acting nitrates
               25 subjects on
               medication; 13 on beta
               blockers + Ca-channel
               blockers; 6 on beta
               blockers; 5 on
               Ca-channel blockers;
               1 on long-acting nitrates
No current
smokers
                          No current
                          smokers
(ST changes) plus 1 or more
of the following: (1) >70%
lesion by angiography in 1
or more major vessels,
(2) prior Ml, (3) positive
exercise thallium test

Ischemia during exercise
(ST changes or abnormal
ejection fraction response)
and 1 or more of the
following: (1) angic-
graphically proven CAD,
2) prior MI, (3) typical
angina

Ischemia during exercise
(ST changes or abnormal
ejection fraction response)
and 1 or more of the
following: (1) angio-
graphically proven CAD,
(2) prior MI, (3) typical
angina
36-75
(mem = 58.2)
                                                 36-75
                                                 (mean = 58)
         Source:  Adapted from Allred et al. (1989b)

-------
       conditions, and means of measuring COHb.  All of the studies have shown an effect of COHb
       elevation on the time to onset of angina (see Figure 10-2). Results form the Kleinman et al.
       (1989) study showed a 6% decrease in exercise time to angina at 3.0% COHb (CO-Ox
       measurement) measured at the end of exposure.  Allred et al. (1989a,b) reported a 5 and 7%
 5     decrease in time to onset of angina after increasing COHb levels to 3.2 and 5.6% (CO-Ox
       measurement), respectively, at the end of exposure. Both Sheps et al. (1987) and Adams
       et al. (1988) reported a significant decrease in the time to onset of angina on days when
       COHb levels at the end of exposure were 3.8 and 5.9% (CO-Ox measurement), respectively,
       if the data analysis by actuarial method included subjects who experienced angina on the CO
10     but not the air day.
            The multicenter study (Allred et al., 1989a,b) demonstrated a dose-response effect of
       COHb on time to onset of angina.  The only other single  study that investigated more than a
       single target level of COHb was Anderson et al. (1973) and their results did not show a dose
       response for angina.
15          The time to onset of significant ECG ST-segment changes, which are indicative of
       myocardial ischemia in patients with documented cornary artery  disease (CAD), is a more
       objective indicator of ischemia than  angina. Allred et al.  (1989a,b) reported a 5.1 and
       12.1 % decrease in time to ST depression at COHb levels  of 2.0  and 3.9% (GC
       measurement), respectively, measured at the end of exercise.  An additional  measurement of
20     the ST change was made by Allred et al. (1989b) to confirm this response - all the leads
       showing ST segment changes were summed.  This summed ST score also was significantly
       affected by both levels of COHb. The significant finding for the summed ST score indicates
       that the effect reported for time to ST was not dependent upon changes observed in a single
       EKG lead.
25          The differences between the results of these five studies on exercise-induced angina can
       largely be explained by differences in experimental methodology and analysis of data and, to
       some extent, by differences in subject populations and sample size. For example, the
       Kleinman study and the Allred study used one-tailed p values. The Sheps and Adams studies
       used two-tailed p values.  If a two-sided p value was utilized on  the time to onset of angina
30     variable observed at the lowest COHb  level in the Allred study, it would become 0.054 rather
       than 0.027, a result that would be considered borderline significant. If a two-sided p value

       March 12,  1990                          10-31      DRAFT-DO NOT QUOTE OR CITE

-------
 o
 UJ
     30
     25 -
     20 -
     15 -
     10 -
O   5 -
UJ   ^ ^
a



ui   0 -
o
a:
LJ
a.
    -5 -
   -10
                               ANDERSON
              KLEINMAN
111
                             ALLRED
                                                       ALLRED
                                          SHEPS
                                                                     ADAMS
                          PERCENT COHb BY OPTICAL METHODS
Figure 10-2. The effect of CO exposure on time to onset of angina. For comparison across

studies, data are presented as the mean percent differences between air and CO exposure days

for individual subjects that were calculated from each of the studies.  Bars indicate calculated

standard errors of the mean.  (See text and Table 10-2 and Table 10-3 for more detailed

information).



Source:  Adapted from Allred et al. (1989b).
 March 12, 1990
                10-32     DRAFT-DO NOT QUOTE OR CITE

-------
        were used in the Kleinman study, the difference in time to onset of angina would lose
        significance at the p = 0.05 level.
             The entry criteria in the Allred study were more rigorous than in the other studies.  All
        subjects were required to have stable exertional angina and reproducible exercise-induced ST
 5      depression and angina.  Besides these criteria, all subjects were required to have either a
        previous myocardial infarction (MI), angiographic disease, or a positive thallium stress test.
        In addition, only men were studied.  These strict entry criteria were helpful in allowing the
        investigators to measure more precisely an adverse effect of CO exposure. However, because
        of the difficulty the investigators had in recruiting subjects, some questions remain about  how
10      representative the study population is of all patients in the United States with exercise-
        induced ischemia.  In addition, the protocol for the multicenter study was slightly different
        from some of the protocols previously reported.  On each test day, the subject performed a
        symptom-limited exercise test on a treadmill, then was exposed for approximately one hour to
        air or one of two levels of CO in air, and then underwent a second exercise test.  Time to the
15      onset of ischemic EKG changes and time to the onset of angina were determined for each
        exercise test.  The percent difference for  these endpoints from  the pre- and postexposure test
        then was determined. The results on the  2% target day and then the 4% target day were
        compared to those on the control day.
             The statistical significance reported  at the low-level CO exposure is only present when
20      the differences between pre- and postexposure exercise tests are analyzed.  Analysis of only
        the postexposure test results in a loss of statistical significance  for the 2%  COHb level. Some
        of the differences between the results of this multicenter study  and previous studies may be
        related to the fact that the exposure was conducted  shortly after patients exercised to angina.
        Because the effects of ischemia may have a variable duration (radionuclide studies have shown
25      metabolic effects of ischemia to last for over an hour), differences between pre- and
        postexposure tests may have been due to  effects of CO exposure on recovery from a previous
        episode of exercise-induced ischemia rather than detrimental effects only during exercise.
             In conclusion, five key studies have investigated the potential for CO exposure to
        enhance the development of myocardial ischemia at <6% COHb during progressive exercise
30      tests.  Despite differences between these  studies, it is impressive that Figure 10-2 shows a
        consistent relationship in percent decrease in time to onset of angina across multiple studies.

        March 12, 1990                          10-33      DRAFT-DO NOT QUOTE OR CITE

-------
        Therefore, there are clearly demonstrable effects of low-level CO exposure in patients with
        ischemic heart disease. The decrements in performance that have been described at the lowest
        levels (<3% COHb) are probably in the range of reproducibility of the test and would not be
        alarming to most physicians.  The adverse health consequences of these types of effects,
 5      however, are very difficult to project.

        Effects in Individuals with Chronic Obstructive lung Disease
             Aronow et al. (1977) studied the effects of a one-hour exposure to 100 ppm CO on
        exercise performance in 10 men, aged 53 to 67 years, with chronic obstructive lung disease.
10      The resting mean COHb levels increased from baseline levels of 1.4% to 4.1% after
        breathing CO.  The mean exercise time until marked dyspnea significantly decreased  (33%)
        from 218 s in the air-control period to 147 s after breathing CO. The authors speculated that
        the reduction in exercise performance was due to a cardiovascular limitation rather than
        respiratory impairment.
15           Only one  other study in the literature by Calverley et al. (1981) looked at the effects of
        CO on exercise performance in older subjects with chronic lung disease.  They evaluated
        15 patients with severe reversible airway obstruction due to chronic bronchitis and
        emphysema. Six of the patients were current smokers but they were asked to stop smoking
        for 12 h before each study. The distance walked within 12  min was measured before and
20      after each subject breathed 0.02%  CO in air from a mouthpiece for 20 to 30 min until COHb
        levels were 8 to 12% above their initial levels.  A significant decrease in walking distance
        was reported when the mean COHb concentration reached 12.3%,  a level much higher than
        most of those reported in the studies on healthy subjects.
             Thus,  while it is possible that individuals with hypoxia due to chronic lung diseases such
25      as bronchitis and emphysema may be  susceptible to CO during submaximal exercise typically
        found during normal daily exercise, these effects have not been studied adequately at relevant
        COHb concentrations of <5%.

        Effects in Individuals with  Chronic Anemia
30           An additional study by  Aronow et al. (1984) on the  effect of CO on exercise
        performance in anemic subjects found a highly significant decrease in work time (16%)

        March 12, 1990                          10-34      DRAFT-DO NOT QUOTE OR CITE

-------
       induced by a 1.24% increase in COHb.  The magnitude of change seems to be very unlikely,
       however, even considering the report by Ekblom and Huot (1972).  The study was double-
       blind and randomized, but with only 10 subjects.  The exercise tests were done on a bicycle
       in the upright position with an increase in workload of 25 watt every 3 min.  However, no
 5     measure of maximal performance such as blood lactate was used. The mean maximal heart
       rate was only 139 to 146 beats/min compared to a predicted maximal heart rate of
       170 beats/min for the mean  age of the subjects. A subject repeating a test within the same
       day, which was the case in the Aronow et al.  (1984) study, often will remember the time and
       work load and try to do the  same in the second test. Normally, however, some subjects will
10     increase while others will decrease the time. This situation was apparent on the air-control
       day, with an increase demonstrated in 6 out of 10 subjects, despite the high reproducibility
       for such a soft, subjective endpoint.  Also,  comparing the control tests on the air day with the
       CO day,  7 out of 10 subjects increased their work time.  After CO exposure,  however, every
       subject decreased their time between 29 and 65 s. These data appear to be implausible given
15     the soft endpoint used, when two to three of the subjects would be expected to increase their
       time even if there was a true effect of CO.

       10.3.2.2 Arrhythmogenic  Effects
            The literature until recent years has been confusing with regard to potential
20     arrhythmogenic effects of CO.
            Davies and Smith (1980) studied the effects of moderate CO exposure on healthy
       individuals.  Six matched groups of human  subjects lived in a closed, environmental-exposure
       chamber for 18 days and were exposed to varying levels of CO.  Standard 12-lead
       electrocardiograms were recorded during five control, eight exposure, and five recovery days,
25     respectively.  P-wave changes of at least 0.1 mV were seen in the electrocardiograms during
       the CO exposure period in 3 of 15 subjects  at 2.4% COHb and in 6 of 15 at 7.1 % COHB
       compared to none of 14 at 0.5% COHb.  The authors felt that CO had a  specific toxic effect
       on the myocardium in addition to producing a generalized decrease in O2 transport to tissue.
            Several  methodological problems create difficulties of interpretation for this study.  The
30     study design did not use each subject as his  own control.  Thus, only one exposure was
       conducted out for each subject.  Half of the subjects were tobacco smokers who were required

       March 12, 1990                         10-35     DRAFT-DO NOT QUOTE OR CITE

-------
        to stop and certainly some of the ECG changes could have been due to the effects of nicotine
        withdrawal. Although the subjects were deemed to be normal, no screening stress tests were
        performed to uncover latent ischemic heart disease or propensity to arrhythmia. Most
        importantly, no sustained arrhythmias or measurable effects on the conduction  system were
 5      noted by the authors.  If p-wave changes of clinical significance are representative of a toxic
        effect of CO on the atrium, then an effect on conduction of arrhythmias should be
        demonstrated.
             Knelson (1972) reported that 7 of 26 individuals, aged 41 to 60 years, had abnormal
        electrocardiograms after exposure to 100 ppm CO for four hours (COHb levels of 5 to 9%).
10      Two of them developed arrhythmias.  No further details were given regarding specifics of
        these abnormalities.  Among 12 younger subjects aged 25  to 36 years, all electrocardiograms
        were normal.
             Hinderliter et al. (1989) reported on effects of low-level CO exposure on resting and
        exercise-induced ventricular arrhythmias in patients with CAD and no baseline ectopy. They
15      studied 10 patients with ischemic heart disease and no ectopy according to baseline
        monitoring.  After an initial training session, patients were exposed to air, 100 ppm CO, or
        200 ppm CO on successive days in a randomized, double-blinded crossover fashion. Venous
        COHb levels after exposure to 100 and 200 ppm CO averaged 4 and 6%,  respectively.
        Symptom-limited supine exercise was performed after exposure.  Eight of the ten patients had
20      evidence of exercise-induced ischemia - either angina,  ST segment depression, or abnormal
        left ventricular ejection fraction response - during one or more exposure days.  Ambulatory
        electrocardiograms were obtained for each day and analyzed for arrhythmia frequency and
        severity.  On air- and CO-exposure days, each patient had only zero to one ventricular
        premature beat per hour in the two hours prior to exposure, during the exposure period,
25      during the subsequent exercise test, and in the five hours following exercise. The authors
        concluded that low-level CO exposure is not arrhythmogenic in patients with CAD and no
        ventricular ectopy at baseline.
             When patients with other levels of ventricular ectopy were studied (Sheps et al., 1989),
        there was an increase in exercise-related arrhythmia (both simple  and  complex). Although no
30      definite evidence exists to date relating effects of CO exposure and lethal arrhythmias, the
        recent epidemiologic study of Stern and colleagues (Stern et al., 1988) indicates that an excess

        March 12, 1990                          10-36      DRAFT-DO NOT QUOTE OR CITE

-------
        of cardiovascular mortality in tunnel workers could be due to exposure to high levels of CO
        (see Section 10.3.3.1).  Their findings that risk decreased after job cessation and that risk was
        not related to length of exposure suggest an acute effect of CO exposure maybe the causative
        factor (perhaps due to arrhythmia production).  These findings are consistent with the general
 5      lack of effect of CO exposure on the development or progression of atherosclerosis.

        10.3.2.3  Effects on Coronary Blood Flow
             The effects of breathing CO on myocardial function in patients with and without
        coronary heart disease have been examined by Ayres et al. (1969; 1970).  Acute elevation of
10      COHb from 0.98 to 8.96% by a bolus exposure using either  1000 ppm CO for 8 to 15  min or
        5000 ppm for 30 to 45 s caused a 20% average decrease in coronary sinus O2 tension without
        a concomitant increase in coronary blood flow in the patients with coronary artery disease.
        Observations in patients with coronary disease revealed that acute elevation of COHb to
        approximately 9% decreased  the extraction of O2 by the myocardium. However, overall
15      myocardial  O2 consumption did not change significantly because an increase in coronary blood
        flow  served as a mechanism to compensate for a lower overall myocardial O2 extraction.  In
        contrast, patients with noncoronary disease increased their coronary blood flow with an
        insignificant decrease in coronary sinus O2 tension as a response to increased COHb.  The
        coronary patients also switched from lactate extraction to lactate production. Thus, a
20      potential threat exists for patients with coronary heart disease who inhale CO because of their
        inability to increase coronary blood flow to compensate for the effects of increased COHb.
             Although in this study the coronary sinus PO2 dropped only slightly (reflecting average
        coronary venous O2 tension),  it is certainly possible that, in areas beyond a significant
        coronary arterial stenosis, tissue hypoxia might be precipitated by very low tissue PO2 values.
25      Tissue hypoxia might be further exacerbated by a coronary-steal phenomena whereby
        increased  overall coronary flow diverts flow from areas beyond a stenosis to other normal
        areas. Therefore the substrate for the worsening of ischemia and consequent precipitation of
        arrhythmias is present with CO exposure.
30
       March 12, 1990                          10-37      DRAFT-DO NOT QUOTE OR CITE

-------
       10.3.3  Relationship between CO Exposure and Risk of Cardiovascular
                Disease in Man
       10.3.3.1 Risk of bchemic Heart Disease
            Epidemiologic studies on the relation between CO exposure and ischemic heart disease
 5     are scarce.  Earlier epidemiological data were summarized by Kuller and Radford (1983).
       They concluded that mortality and morbidity studies have been negative or equivocal in
       relating CO levels to health effects, but studies in human subjects with compromised coronary
       circulation  support an effect of acute exposure to CO at blood levels equivalent to about
       20 ppm over several hours. They calculate that based on health surveys, probably over
10     10 million subjects in the United States are exposed to potentially deleterious levels of CO
       and that perhaps 1250 excess deaths related to low-dose environmental CO exposure occur
       each year.
            Stern et al. (1981) reported a study performed by the NIOSH.  They investigated the
       health effects of chronic exposure to low concentrations of CO by conducting a historical
15     prospective cohort study  of mortality patterns among 1558 white, male motor vehicle
       examiners in New Jersey.  The examiners were exposed to 10 to 24 ppm. The COHb levels
       were determined in 27 volunteers. The average COHb level before a work shift was 3.3%
       and the post-shift level was 4.7% in the whole group and 2.1 and 3.7%, respectively, in
       nonsmokers only. The death rates were compared to the rates in the U.S. population based
20     on vital statistics. The cohort demonstrated a slight overall increase in cardiovascular deaths
       but a more pronounced excess was observed  within the first 10 years following employment.
       The study has several important limitations, however, including the use of historical controls.
       A second limitation is the lack of knowledge about smoking habits.  A third is that the
       individuals' values of COHb  were not known.
25          Stern et al. (1988) published coronary heart disease mortality data among bridge and
       tunnel officers exposed to CO.  They investigated  the effect of occupational exposure to CO
       on mortality from arteriosclerotic heart disease in a retrospective study of 5529 New York
       City bridge and tunnel officers.  There were 4317 bridge officers and 1212 tunnel officers.
       Among former tunnel officers, the standardized mortality ratio was 1.35 (90% confidence
30     interval 1.09  to 1.68) compared to the New York City population.  Using the proportional
       hazards model,  the authors compared the risk of mortality from arteriosclerotic heart disease

       March 12,  1990                          10-38     DRAFT-DO NOT QUOTE OR CITE

-------
        among tunnel workers with that of the less-exposed bridge officers.  They found an elevated
        risk in the tunnel workers that declined within as little as five years after cessation of
        exposure.  The 24-h average CO level in the tunnel was around 50 ppm in 1961 and around
        40 ppm in 1968. However, higher values were recorded during rush hours.  In 1971 the
 5      ventilation was further improved and the officers were allowed clean-air breaks. Although
        the authors concluded that CO exposure may play an important role in the pathophysiology of
        cardiovascular mortality, other factors must be taken into consideration.  Mortality from
        arteriosclerotic heart disease has a complex multifactor etiology.  The presence of other risk
        factors, such as cigarette smoke,  hypertension, hyperlipidemia, family history of heart
 10      disease, obesity, socioeconomic status, and sedentary living all can increase the risk of
        developing coronary heart disease.  In addition, detailed exposure monitoring was not done in
        this study. The bridge and tunnel workers were not only exposed to CO  but also were
        exposed to other compounds emitted from motor vehicle exhaust and  to the noise and stress of
        their environment.  These other factors could have contributed to the  findings.
 15          Intoxication with CO that induces COHb levels around 50 to 60% is often lethal.
        However, even levels around 20%  COHb have been associated with death, mainly coronary
        events, in patients with severe  coronary artery disease.  Balraj (1984) reported on 38 cases of
        individuals dying immediately  or within a few days following exposure to 10 to 50% COHb,
        usually nonlethal levels of CO. All of the subjects had coronary artery disease, and 29 of
20      them had severe cases. The author concluded that  the CO exposure, between 10 to 30%
        COHb in 24 cases, triggered the lethal event in subjects with a restricted coronary flow
        reserve. Similar associations between CO exposure and death or myocardial infarction have
        been reported by several other  authors.  Atkins and Baker (1985) reported two cases with 23
        and 30% COHb; McMeekin and Finegan (1987) reported one case with 45%  COHb; Minor
25      and Seidler (1986) reported one case with 19% COHb; and Ebisuno et al. (1986) reported
        one case with 21% COHb.
            Forycki et al. (1980) described electrocardiographic changes  in 880  patients treated for
        acute poisoning. Effects were  observed in 279 cases, with the most marked changes in cases
        with CO poisoning.   In those, the most common change was a T-wave abnormality and in
30      six cases a pattern of acute myocardial infarction was present. Conduction disturbances also
        were common in CO poisoning but arrhythmias were less common.

        March 12,  1990                          10-39     DRAFT-DO  NOT QUOTE OR CITE

-------
            The association between smoking and cardiovascular disease (CVD) is fully established.
       Although little is known about the relative importance of CO compared to nicotine, most
       researchers consider them to be equally important.  The nicotine component clearly aggravates
       the decrease in O2 capacity induced by CO through an increase in the O2 demand of the heart.
 5     This is exemplified in the study by Deanfield et al. (1986) using positron emission
       tomography.  They found that smoking one cigarette induced perfusion abnormalities in six
       out of eight patients with CAD and exercise-induced angina. However, the smoke-induced
       ischemia was without angina or silent ischemia in all of the patients and without ST
       depression in seven of the patients.  This raises an important question regarding analyses of
10     the effect of CO.  Ischemia is not always associated with angina and/or ST depression.
       However, most of the reports  used to develop the guidelines for CO exposure have used
       angina and/or ST depression as a sign of ischemia. Only the two studies by Sheps et al.
       (1987) and  Adams et al. (1988) used additional techniques to diagnose ischemia (see
       Section 10.3.2).  Both used gated radionuclide angiography to measure changes in ejection
15     fraction and wall motion induced by exercise, allowing the detection for signs of ischemia in
       the absence of angina and/or ST depression.  Positron emission tomography is even more
       sensitive, however.  Future studies on the effects of CO in patients with CVD,  therefore, will
       need to include more sensitive measures of ischemia than angina and/or ST depression.
            Passive smoking exposes an individual to all  components in the cigarette smoke, but the
20     CO component dominates heavily because  only 1 % or less of the nicotine is absorbed from
       sidestream  smoke compared to 100% in  an active smoker (Wall et al., 1988; Jarvis, 1987).
       Therefore,  exposure to  sidestream smoke will be the closest to pure CO exposure even if the
       resultant levels of COHb are low (about 2  to 3%).  Two recent studies report on the
       relationship between passive smoking and risk of coronary heart disease (CHD).  Svendsen
25     et al. (1987) report from the multiple risk factor intervention trial (MRFIT) study on
       1245 married, never-smoking  men and 286 men married to women who smoke. The relative
       risk for exposure to sidestream smoke was 1.48 (p = 0.13, 95% confidence interval [CI] =
       0.89 to 2.47) for nonfatal and fatal coronary events and 1.96 (p = 0.08, 95% CI = 0.93 to
       4.11) for all causes of mortality.  Even more significant results were reported by Helsing
30     et al. (1988) who studied 4162 men and 14,873 women that were nonsmokers,  some of which
       had been exposed to sidestream smoke.  The relative risk for exposure to passive smoke was

       March 12,  1990                          10-40      DRAFT-DO NOT QUOTE OR CITE

-------
        1.31 (95% CI =  1.1 to 1.6) in men and 1.24 (95% CI = 1.1 to 1.4) in women for
        arteriosclerotic heart disease death. Even though it is impossible to rule out an effect of the
        other component in sidestream smoke, the data suggest an increase in risk of CHD associated
        with a prolonged exposure to low levels of CO. In the United States, a population study by
 5      Cohen et al. (1969) suggested an association between atmospheric levels of CO and increased
        mortality from myocardial infarction in Los Angeles, but potential confounders were not
        effectively controlled.  In contrast, a similar study in Baltimore (Kuller et al., 1975) showed
        no association between ambient CO levels and myocardial infarction or sudden death. In a
        Finnish study (Hernberg et al. 1976), the prevalence of angina among foundry workers
10      showed an exposure-response relationship with regard to CO exposure,  but no such result was
        found for ischemic EKG findings during exercise.
            In a cross-sectional study of 625 smokers, age 30 to 69, Wald et al. (1984) reported that
        the incidence of CVD was higher in subjects with COHb >5% compared to subjects below
        3%, a relative risk of 21.2 (95%  CI = 3.3 to 34.3).  Even if all of the subjects were
15      smokers, the association between COHb and CVD might be due to the fact that percent
        COHb is a measure of smoke exposure.
            Low to intermediate levels of COHb might interfere with the early course of an acute
        myocardial infarction.  The increase in COHb can be due to recent smoking or environmental
        exposure.  Mall et al. (1985) reported on a prospective study in smoking and nonsmoking
20      patients with acute myocardial infarction who were separated by their baseline COHb levels.
        A total of 66 patients were studied. Thirty-one patients presented with a COHb level of 1.5%
        and 35 with a level of 4.5%.  In the group with elevated COHb, more patients developed
        transmural infarction, but the difference was not significant.  Patients with transmural
        infarction had higher maximum creatine phosphokinase values when COHb was over 2%.
25      During the first six hour after admission to the hospital, these patients needed an
        antiarrhythmic treatment significantly more frequently.  Differences in rhythm disorders were
        still present at a time when nicotine, due to its short half-life, was already eliminated. The
        authors concluded that moderately elevated levels of COHb may aggravate the course of an
        acute myocardial infarction.
30
       March 12,  1990                         10-41     DRAFT-DO NOT QUOTE OR CITE

-------
       10.3.3.2 Risk of Hypertension
            Because short-term exposure to CO does not change the arterial blood pressure, it is
       unlikely that a prolonged CO exposure will induce hypertension. However, in a study by
       Ahmad and Ahmad (1980), a higher prevalence of hypertension in the population was
 5     observed in a large city compared to a small village where COHb levels were 16.4 and 5.6%,
       respectively.  The main difference between the two locations was that greater exposures to
       noise and CO in the large city are known to induce hypertension, which might explain the
       findings. On the other hand, laboratory animal studies (see Section 10.3.4) indicate that
       prolonged CO exposure induces cardiac hypertrophy.  If a similar effect occurs in humans, it
10     may be significant because hypertension increases the risk of developing left ventricular
       hypertrophy (Frohlich, 1987).  Therefore, it is important to study further the association
       between CO exposure and risk of hypertension, which should be one of the research tasks
       identified for future investigation.

15     10.3.4  Studies in Laboratory Animals
       10.3.4.1 Introduction
            The mechanisms by which CO exerts its toxic effects are detailed in Chapter 9. In
       brief, CO combines with blood Hb to form COHb; this  decreases the O2-carrying capacity of
       the blood and shifts the O2Hb dissociation  curve to the left. The cardiovascular system is
20     sensitive to alterations in O2 supply, and because  inhaled CO limits O2 supply,  it might be
       expected to adversely affect the cardiovascular system; the degree of hypoxia and the extent
       of tissue injury will be determined by the dose of CO. The effect of CO on the
       cardiovascular system has been the subject of several recent reviews  (Turino, 1981; McGrath,
       1982; Penney, 1988). This section  will discuss studies in animals that have evaluated the
25     effects of CO on ventricular fibrillation, hemodynamics, cardiomegaly, hematology, and
       atherosclerosis. In this review, CO concentrations, times of exposure, and COHb levels are
       provided whenever they  were mentioned in the original  manuscript.  An attempt has been
       made to focus, where possible, on those studies that have used the most relevant concentra-
       tions of CO.  For a more detailed treatment of the effects of higher concentrations  of CO the
30     reader is referred to the review by Penney (1988).
       March 12, 1990                         10-42     DRAFT-DO NOT QUOTE OR CITE

-------
       10.3.4.2 Ventricular Fibrillation Studies
            Data obtained from animal studies suggest that CO can disturb cardiac conduction and
       cause cardiac arrhythmias (see Table 10-4).  In dogs exposed for six weeks in environmental
       chambers to CO (50 and 100 ppm; COHb = 2.6 to 12.0%) intermittently and continuously,
 5     Preziosi et al. (1970) reported abnormal electrocardiograms; the changes appeared during the
       second week and continued throughout the exposure. The blood cytology, hemoglobin, and
       hematocrit values were unchanged from control values. DeBias et al.  (1973) studied the
       effects of breathing CO (96 to 102 ppm; COHb = 12.4%) continuously (23 h/day; for
       24 weeks) on the electrocardiograms of healthy monkeys and monkeys with myocardial
10     infarcts induced by injecting microspheres into the coronary circulation.  The authors
       observed higher P-wave amplitudes in both the infarcted and non-infarcted monkeys and a
       higher incidence of T-wave inversion in the infarcted monkeys. The authors concluded there
       was a greater degree of ischemia in the infarcted animals breathing CO.  Although there was
       a greater incidence of T-wave inversion in the infarcted monkeys the effects were transient
15     and of such low magnitude that accurate measurements of amplitude were not possible.
            In other long-term studies, however, several groups have reported no effects of CO
       either on the EKG or on cardiac arrhythmias.  Musselman et al. (1959) observed no changes
       in the EKG of dogs exposed continuously to CO (50 ppm COHb  = 7.3%) for three months.
       These observations were confirmed by Malinow et al. (1976) who reported no effects on the
20     EKG in cynomolgus monkeys exposed to CO (500 ppm-pulsed; COHb = 21.6%) for 14 mo.
            Several research groups have investigated the effects of CO on the vulnerability of the
       heart to induced ventricular fibrillation. DeBias et al. (1976) reported that CO (100 ppm
       inhaled for 16 h; COHb = 9.3%) reduced the threshold for ventricular fibrillation induced by
       an electrical stimulus applied to the myocardium of monkeys during the final stage of
25     ventricular repolarization.  The voltage required to induce fibrillation  was highest in normal
       animals breathing air and lowest in infarcted animals breathing CO. Infarction alone and CO
       alone each required significantly less voltage for fibrillation; when the two were combined the
       effects on the myocardium were additive. These observations were confirmed in both
       anesthetized, open-chested dogs with acute myocardial injury (Aronow et al., 1978) and in
30     normal dogs (Aronow et al., 1979) breathing CO (100 ppm; COHb = 6.3 to 6.5%) for
       two hours.  However, Kaul et al. (1974) reported that anesthetized dogs inhaling 500 ppm

       March 12, 1990                         10-43     DRAFT-DO  NOT QUOTE OR CITE

-------
S TABLE 1O4. VENTRICULAR FIBRILLATION AND HEMODYNAMIC
1 STUDIES IN LABORATORY ANIMALS
sy
Exposurea>b COHb0 Animal Dependent Variable4 Results
VO CO = 50 ppm - Dog (n = 4) EKG, heart rate No effects
® continuously for 3 mo Rabbit (n = 4)
Rat (n = 100)
Comments Reference
Musselman et al. (1959)
CO = 50-100 ppm for
6 weeks intermittently or
continuously
2.6-12%
CO = 100 ppm for 24 wk,   12.4%;
23 h/day; CO =             9.3%
100 ppm for 6 h
CO = 500 ppm, pulsed      21.6%
12 h/day for 14 mo
C0= 100 ppm, 2 h         6.3%
Coronary artery
ligated; normal

CO = 5,000 ppm            4.9-17.0%
5 sequential exposures
to produce desired
COHb;  Coronary artery
ligated

CO = 100 ppm             6.8-14.6%
Coronary artery
occluded briefly
                Dog (n = 28)
                      EKG and pathology
                Cynomolgus monkey    EKG and susceptibility
                (n = 52; 20)           to induced fibrillation
                Cynomolgus monkey    EKG, arterial pressure,
                (n = 26)              left ventricular
                                      pressure, dP/dt, Vamli

                Dog (n = 21, 20)       Ventricular fibrilla-
                                      tion threshold (VFT)
                Dog(n=  11)
                Dog (n =  14)
                      EKG, coronary blood
                      flow
                      Arrhythmia; conduction
                      slowing in ischemic
                      myocardium
Abnormal EKG, heart
dilation, myocardial
thinning, some subjects
showed scarring and
degeneration in heart
muscle

Abnormal EKG and
increased sensitivity
to fibrillation
voltage

No effects
                                                 Reduced VFT in normal
                                                 and ligated dogs
Elevated ST segment;
increased flow to
non-ischemic myocar-
dium
No changes
                         Preziosi et al. (1970)
                                                                           Infarcted animals
                                                                           showed greatest effect
                                                                           of COHb on both
                                                                           dependent variables
                         DeBias et al. (1973)
                                                                           Subjects on normal and    Malinow et al. (1976)
                                                                           high cholesterol diets
                          Studies were conducted    Aronow et al. (1978,
                          blind                     1979)
CO can augment
ischemia in acute
myocardial infarction
Concluded CO is not
arrhythmogenic during
early minutes of
infarction
Becker and Haak (1979)
Foster (1981)
CO = 200 ppm for 60
and 90 min; Paced
hearts and introduced
premature stimulus
1.64-6.30%
Dog (n = ?)
                                      Threshold for ventri-
                                      cular arrhythmias and
                                      refractory period
No effects
                                                   Hutcheon et al. (1983)

-------





w
EJ
2
<
s
HEMOD
§3
^ ^
IBRILLATION
ATORYANIM
£ o
«? CQ





















i"

£• 2 "o
f- § a
O C 
?
03

O
3
1
a.
T3 2
1 i
o a
1 °
C. (S
3 c
Arrhythmia woi
at 20% COHb i
9 animals




ta
CO
'S
>^
Ji
<


CK
II
c
I


tR
o
""


CO = ?; previous
myocardial infarction
and 2 min coronary
artery occlusion

C

0
1








No changes
.2
f*»
S
g" tt

ill
.« o 5
O > o


«-s §
pgl
s2|
l'|S
e e -S.J
^Ift
1-l-f
S-k'l:
'Exposure concentrati
bl ppm = 1.145 mg/:
"Measured blood carb
dSee glossary of term
March 12, 1990
10-45    DRAFT-DO NOT QUOTE OR CITE

-------
       CO (COHb = 20 to 35%) for 90 min were resistant to direct electrocardiographic changes.
       At 20% COHb there was evidence of enhanced sensitivity to digitalis-induced ventricular
       tachycardia but there was no increase in vulnerability of the ventricles to
       hydrocarbon/epinephrine or to digitalis-induced fibrillation following exposure to 35% COHb.
 5          Several workers have investigated the effect of breathing CO shortly after cardiac injury
       on the electrical activity of the heart. Becker and Haak (1979) evaluated the effects of CO
       (five sequential exposures to 5,000 ppm, producing COHb = 4.9 to 17.0%) on the
       electrocardiograms of anesthetized dogs one hour after coronary artery ligation. Myocardial
       ischemia, as judged by the amount of ST-segment elevation in epicardial electrocardiograms,
10     increased significantly at the lowest COHb levels (4.9%) and increased further with increasing
       CO exposure;  there were no changes in heart rate, blood pressure, left atrial pressure, cardiac
       output, or blood  flow to the ischemic myocardium. Similar results were noted by Sekiya
       et al.  (1983) who investigated the influence of CO (3,000 ppm for 15 min followed by
       130 ppm CO for one hour, COHb = 13 to 15%) on the extent and severity of myocardial
15     ischemia in dogs. This dose of CO inhaled prior to coronary artery ligation increased the
       severity and extent of ischemic injury, and the magnitude of ST segment elevation, more than
       did ligation alone.  There were no changes in heart rate or arterial pressure.  Vanoli et al.
       (1986) reported that CO (COHb =  10 and 20%) worsened the arrhythmia, following a
       two minute coronary artery occlusion, in two of nine dogs with one-month-old anterior
20     myocardial infarcts. They were unable, however, to reproduce their results.  The authors
       conclude that while acute exposure to CO is unlikely to produce detrimental effects in post-
       Mi dogs, the CO-dependent tachycardia may enhance the risks and consequences of any
       ischemic episode.
            On the other hand, several groups have reported no effects of CO on the EKG or on
25     cardiac arrhythmias.  Musselman et al. (1959) observed no changes in the EKG of dogs
       exposed continuously to CO (500 ppm) for three months.  Their observations were confirmed
       by Malinow et al. (1976) who reported no effects on the EKG  in cynomolgus monkeys
       exposed to CO for 14 mo (500 ppm-pulsed; COHb = 21.6%).  Foster (1981) concludes that
       CO (100 ppm  CO for six to nine minutes; COHb = 10.4%) is not arrhythmogenic in dogs
30     during the early  minutes of acute myocardial infarction following occlusion of the left anterior
       descending coronary artery.  This level of CO did not effect either slowing of conduction

       March 12,  1990                          10-46     DRAFT-DO NOT QUOTE OR CITE

-------
        through the ischemic myocardium or the incidence of spontaneous ventricular tachycardia.
        These results were confirmed by Hutcheon et al. (1983) in their investigation of the effects of
        CO on the electrical threshold for ventricular arrhythmias and the effective refractory period
        of the heart. They conclude that CO (200 ppm CO for 60 and 90 min; COHb = 5.1 to
  5      6.3%) does not alter the effective refractory period or the electrical threshold for ventricular
        arrhythmias in dogs. These results are consistent with those of Mills et al. (1987), who
        studied the effects of 0 to 20% COHb on the electrical stability of the heart in chloralose-
        anesthetized dogs during coronary occulsion. There were no major effects on heart rate,
        mean arterial blood pressure,  effective refractory period, vulnerable period, or ventricular
 10      fibrillation threshold.
            The effects of CO (1500 ppm, COHb = 15%) during acute myocardial ischemia on
        arrhythmias in dogs with a healed anterior myocardial infarction and at low or high risk for
        ventricular fibrillation were investigated by Vanoli et al.  (1989). Following a two minute
        coronary artery occlusion, malignant arrhythmias occurred in two dogs at low risk but in none
 15      of the dogs at high risk for ventricular fibrillation.  The authors conclude that, in dogs at high
        risk for ventricular fibrillation, arrhythmogenic effects seldom can be expected from acute
        exposure to CO.
            On balance the results from animal studies suggest that mhaled CO can cause
        disturbances in cardiac rhythm in both healthy and compromised hearts.  Depending on the
20      exposure regime and species tested, the threshold for this response varies between 50 and
        100 ppm CO (COHb = 2.6 to 12%) in dogs and 100 ppm (COHb = 12.4%) in  monkeys
        inhaling CO for 6 to 24 weeks; and 500 ppm CO (COHb  =  4.9 to 17.0%) in dogs and
        100 ppm (COHb =  9.3%) in  monkeys inhaling CO for 0.6 to 16 h.

25      10.3.4.3  Hemodynamic Studies
            The effects of CO on coronary flow, heart rate, blood pressure,  cardiac output,
        myocardial O2 consumption, and blood flow to various organs have been investigated in
        laboratory animals.  The results are somewhat contradictory (partly because exposure regimes
        differed);  however, most workers agree that CO in sufficiently high doses can affect many
30      hemodynamic variables.
       March 12, 1990                          10-47     DRAFT-DO NOT QUOTE OR CITE

-------
            Adams et al. (1973) described increased coronary flow and heart rate and decreased
       myocardial O2 consumption in anesthetized dogs breathing 1500 ppm CO for 30 min
       (COHb  = 23.1%).  The decreased O2 consumption indicates that the coronary flow response
       was not great enough to compensate for the decreased O2 availability.  The authors noted that
 5     although there was a positive chronotropic response, there was no positive inotropic response.
       The authors speculated that (1) the CO may have caused an increase in the endogenous
       rhythm or blocked the positive inotropic response or (2) the response to CO was mediated
       reflexly through the cardiac afferent receptors to give a chronotropic response without the
       concomitant inotropic response.  When they used /3-adrenergic blocking agents, the heart-rate
10     response to CO disappeared,  suggesting possible reflex mediation by the sympathetic nervous
       system.
            In a later study in chronically-instrumented, awake dogs exposed to 1000 ppm CO
       producing COHb levels of 30%, Young and Stone (1976) reported an  increase in coronary
       flow with no  change in myocardial O2 consumption. The increased coronary flow occurred in
15     animals with  hearts paced at  150 beats/min, as well as in nonpaced animals, and in animals
       with propranolol and atropine blockade. Because the changes in coronary flow with arterial
       O2 saturation  were similar whether the animals were paced or not, these workers conclude that
       the increase in coronary  flow is independent of changes in heart rate.  Furthermore, the
       authors reasoned that if the coronary vasodilation was caused entirely by the release of a
20     metabolic vasodilator, associated with decreased arterial O2 saturation, the change in coronary
       flow in animals with both /J-adrenergic and parasympathetic blockade should be the same as in
       control dogs. Young and Stone conclude that coronary vasodilation observed with an arterial
       O2 saturation reduced by CO is mediated partially through an active neurogenic process.
            Increased myocardial blood flow after CO inhalation in dogs was confirmed by Einzig
25     et al. (1980), who also demonstrated the regional nature of the blood flow response. Using
       labeled microspheres, these workers demonstrated that whereas both right and left ventricular
       beds were dilated maximally at COHb levels of 41% (CO = 15,000 to 20,000 ppm for
        10 min), subendocardial/subepicardial blood flow ratios were reduced.  The authors conclude
       that in addition to the global hypoxia associated with CO  poisoning, there is also an
30     underperfusion of the subendocardial layer, which is most pronounced in the left ventricle.
        March 12, 1990                          10-48     DRAFT-DO NOT QUOTE OR CITE

-------
            These results were confirmed by Kleinert et al. (1980) who investigated the effects of
       lowering O2 content by about 30% with low O2 or CO gas mixtures (CO = 10,000 ppm for
       three minutes, COHb = 21 to 28%) on regional myocardial relative tissue PO2, perfusion and
       small vessel blood content.  In anesthetized, thoracotomized rabbits, both hypoxic conditions
 5     increased regional blood flow to the myocardium, but to a lesser extent in the endocardium.
       Relative endocardial PO2 fell more markedly than epicardial PO2 in both conditions. Small
       vessel blood content increased more with CO than with low PO2, whereas regional 02
       consumption increased under both conditions. The authors conclude that the response to
       lowered O2 content (whether by inhaling low O2 or CO gas mixtures) is increased flow,
10     metabolic rate, and the number of open capillaries, and the effects of both types of hypoxia
       appear more severe in the endocardium.
            A decrease in tissue PO2 with  CO exposure also has been reported by Weiss and Cohen
       (1974). These workers exposed anesthetized rats to 80 and 160 ppm  CO for 20-min periods
       and measured tissue O2 tension as well as heart rate.   A  statistically significant decrease in
15     brain PO2 occurred with inhalation of 160 ppm CO, but there was no change in heart rate.
            Horvath (1975) investigated the coronary flow response in dogs with COHb levels of
       6.2 to 35.6% produced by  continuous administration of precisely measured volumes of CO.
       Coronary flow increased progressively as blood COHb increased and was maintained for the
       duration of the experiment. However, when animals  with complete atrioventricular block
20     were maintained by cardiac pacemakers and exposed to COHb levels  of 6 to 7%, there was
       no longer an increase in coronary blood flow.  These results are provocative, because they
       suggest an increased danger from low COHb levels in cardiac-disabled individuals.
            Patajan et al. (1976) exposed unanesthetized rats to 1500 ppm CO for 80 min to achieve
       COHb levels of 60 to 70%. After a slight transient increase, heart rate as well as blood
25     pressure decreased throughout the exposure.  The authors interpret their data as indicating that
       the lowering of blood pressure was  more important than the degree of hypoxia to the
       neurological impairment seen in their studies.
            The effects of CO hypoxia and hypoxic hypoxia on arterial blood pressure and other
       vascular parameters also were studied in carotid baroreceptor and chemoreceptor-denervated
30     dogs (Traystman and Fitzgerald, 1977).  Arterial blood pressure was  unchanged by CO
       hypoxia but increased with hypoxic hypoxia.  Similar results were seen in carotid

       March 12, 1990                         10-49     DRAFT-DO NOT QUOTE OR CITE

-------
       baroreceptor-denervated animals with intact chemoreceptors.  Following carotid
       chemodenervation, arterial blood pressure decreased equally with both types of hypoxia.
            In a subsequent report from the same laboratory (Sylvester et al., 1979), the effects of
       CO hypoxia (CO = 10,000 ppm followed by 1000 ppm for 15 to 20 min; COHb = 61 to
 5     67%) and hypoxic hypoxia, were compared in anesthetized, paralyzed dogs.  Cardiac output
       and stroke volume increased during both CO and hypoxic hypoxia whereas heart rate was
       variable. Mean arterial pressure decreased during CO hypoxia, but increased during hypoxic
       hypoxia. Total peripheral resistance fell during both hypoxias, but the decrease was greater
       during the CO hypoxia.  After resection of the carotid body, the circulatory effects of hypoxic
10     and CO hypoxia were the same and were characterized by decreases in mean arterial pressure
       and total peripheral resistance.  In a second series of closed-chest dogs, hypoxic and CO
       hypoxia caused equal catecholamine secretion before carotid body resection.  After carotid
       body resection, the magnitude of the catecholamine response was doubled with both hypoxias.
       These workers conclude that the responses to hypoxic and CO hypoxia are different and that
15     the difference is dependent on intact chemo- and baroreflexes and on differences in arterial  O2
       tension, but not on differences in catecholamine secretion or ventilatory response.
            In cynomolgus monkeys exposed to 500 ppm intermittently for 12 h a day for 14 mo
       (COHb = 21.6%), Malinow et al. (1976) reported no changes in arterial pressure, left
       ventricular pressure, dp/dt, and Vmax.  On the other hand, Kanten et al. (1983) studied the
20     effects of CO (150 ppm, COHb up to 16%)  for 0.5-2 h on hemodynamic parameters in
       open-chest, anesthetized rats, and reported that heart rate,  cardiac output, cardiac index,
       dF/dtmax (aortic), and  stroke volume increased significantly, whereas mean arterial pressure,
       total peripheral resistance, and left ventricular systolic pressure decreased.  These effects were
       evident at COHb levels as low as 7.5%  (0.05 h).  There were no changes in stroke work, left
25     ventricular dp/dtmax, and  stroke power.
            The effects of CO on blood flow to various vascular beds has been investigated in
       several animal models,  and most of the  studies have been conducted at rather high CO or
       COHb levels. In general, CO increases cerebral blood flow. However, the effects of CO on
       the cerebral circulation are discussed in  detail in Section 10.4.1.
30          In recent studies Oremus et al. (1988) reported that in the anesthetized rat breathing CO
       (500 ppm; COHb =  23%) for one hour that CO reduces mean  arterial pressure through

       March 12, 1990                          10-50     DRAFT-DO NOT QUOTE OR CITE

-------
       peripheral vasodilation predominantly in the skeletal muscle vasculature.  There were no
       differences in heart rate or mesenteric or renal resistances between the CO-exposed and
       control groups.  This was confirmed by Gannon et al. (1988) who reported that in the
       anesthetized rat breathing CO (500 ppm; COHb = 24%) for one hour that CO increased
 5     inside vessel diameter (36 to 40%), increased flow rate (38 to 54%), and decreased mean
       arterial pressure to 79% of control in the cremaster muscle. There was no change in the
       response of 3A vessels to topical applications of phenylephrine as a result of CO exposure.
            King et al. (1984, 1985) compared whole-body and hindlimb blood flow responses in
       anesthetized dogs exposed to CO or anemic hypoxia.  Arterial O2 content was reduced by
10     moderate (50%) or severe (65%) CO-hypoxia (produced by dialysis with 100%  CO) or
       anemic hypoxia (produced by hemodilution). These workers noted that cardiac  output was
       elevated in all groups at 30 min and in the severe CO group at 60 min.  Hindlimb blood flow
       remained unchanged during CO hypoxia in the  animals with intact hindlimb innervation but
       was greater in animals with denervated  hindlimbs. There was a decrease in mean arterial
15     pressure in all groups associated with a fall in total peripheral resistance. Hindlimb resistance
       remained unchanged during moderate CO hypoxia in the intact groups but was increased in
       the denervated group.  The authors concluded that the increase in cardiac output during CO
       was directed to nonmuscle areas of the body and that intact sympathetic  innervation was
       required to achieve this redistribution.   However, aortic chemoreceptor input was not
20     necessary for the increase in cardiac output during severe CO  hypoxia, nor for the diversion
       of the increased flow to nonmuscle tissues.
            King et al. (1987) investigated the effects of severe CO (1000 to 10,000 ppm to lower
       arterial O2 content to 5 to 6 vol) and hypoxic hypoxia on the contracting gastrocnemius
       muscle of anesthetized dogs.  Oxygen uptake decreased from the normoxic level in the CO
25     group but not in the hypoxic hypoxia group. Blood flow increased in both groups during
       hypoxia but more so in the CO group.  Oxygen extraction increased  further during
       contractions in the hypoxic group but fell in the CO group. The authors observed that the O2
       uptake limitation occurring during CO hypoxia and isometric contractions was associated with
       a reduced O2 extraction and concluded that the leftward shift in the O2Hb dissociation curve
30     during CO hypoxia may have impeded O2 extraction.
       March 12,  1990                         10-51     DRAFT-DO NOT QUOTE OR CITE

-------
             Melinyshyn et al. (1988) investigated the role of /J-adrenoreceptors in the circulatory
        responses to severe CO (about a 63% decrease in arterial O2 content obtained by dialyzing
        with  100% CO) of anesthetized dogs.  One group was 0 blocked with propanolol (0, and ft
        blockade), a second with ICI 118,551 (ft blockade), and a third was a time control. Cardiac
 5      output increased in all groups during CO hypoxia with the increase being greatest in the
        unblockaded group. Hindlimb blood flow rose during CO hypoxia only in the unblockaded
        group.  The authors conclude that 35%  of the rise in cardiac output occurring during CO
        hypoxia (COH) depended on peripheral vasodilation mediated through ft-adrenoreceptors.
             Thus, the results from animal studies indicate that inhaled CO can adversely affect
10      several hemodynamic parameters. The threshold for these effects may be near 150 ppm CO
        (COHb = 7.5%).

        10.3.4.4  Cardiomegaly
             The early investigations of cardiac enlargement following prolonged exposure to CO
15      have been confirmed in different animal models and extended  to characterize the development
        and regression of the cardiomegaly (see Table 10-5).  Theodore et al.  (1971) reported cardiac
        hypertrophy in rats breathing 500 ppm CO (COHb = 32 to 38%) for  168 days, but not in
        dogs, baboons, or monkeys.  Penney et al. (1974a) also  noted cardiomegaly in rats breathing
        500 ppm CO; heart weights were one-third greater than predicted within 14 days of exposure,
20      and 140 to 153% of controls after 42 days of exposure.  The cardiomegaly was accompanied
        by changes in cardiac lactate dehydrogenase (LDH) isoenzyme composition that were similar
        to those reported in other conditions that cause cardiac hypertrophy (e.g., aortic and
        pulmonary artery constriction, coronary artery disease, altitude acclimation,  severe anemia).
             To further characterize the hypertrophy and determine its threshold, Penney et al.
25      (1974b) measured heart weights in rats exposed continuously to 100, 200, and 500 ppm CO
        (COHb = 9.26, 15.82, and 41.14%), for various times (20 to 46 days); they noted
        significant increases in heart weights at 200 and 500 ppm CO, with changes occurring in the
        left ventricle and septum, right ventricle, and atria especially.  The authors concluded that
        whereas the threshold for the Hb response is 100 ppm CO (COHb = 9.26%), the threshold
       March 12, 1990                          10-52      DRAFT-DO NOT QUOTE OR CITE

-------
I
                          TABLE 10-5.  CARDIAC HYPERTROPHY STUDIES IN LABORATORY ANIMALS
Exposure"'1"
400-500 ppm for
168 days
100 ppm, 46 days
200 ppm, 30 days
500 ppm, 20-42 days
CO = 60, 125, 250,
500 ppm for 21 days
COHb°
32-38%
(dogs and
monkeys
only)
9.2%
15.8%
41.12%
-
Animal
Monkey (n = 9)
Baboon (n = 3)
Dog (n = 16)
Rat (n = 136)
Mouse (n = 80)
Rat (n = 32)
Fetal rats
(n = 75)
Dependent Variable*
Cardiovascular damage
in rat heart
Heart size; LDH
Hb, Hot, HW
Results
No changes except slight
hypertrophy
Hypertrophy of both left
and right ventricles;
LDH increases
Hb and Hot depressed with
60 ppm and elevated by
Comments

Threshold for cardiac
enlargement near 200 ppm
HW increase probably not
due to increased viscosity
Reference
Theodore et al. (1971)
Penney et al. (1974a,b)
Prigge and Hochraincr (1977)
      gestation
250 and 500 ppm; HW
increased at all concen-
trations
or pulmonary hypertension
o
o
c
I
8
n
CO = 400 ppm, or 35-58%
500 ppm increased
to 1,100 ppm

CO = 500 ppm 38-42%
until 50 days of age
CO = 500 ppm for 38-42%
1-42 days. Open-
chest, anesthetized
preparation.
CO = 150 ppm (15% in
throughout gestation adult rats)
Rat (n = 30)

Rat (n > 200)
5 and 25 days
old
Rat
(n = 25)
Rat
(n = 88)
Cardiomegaly (HW/BW)
and LDH

HW
Right ventricle (RV)
Left ventricle (LV)
Stoke index (SI),
Mean stroke power (SP),
Mean cardiac output (CO),
Systemic resistance (SR),
Pulmonary resistance (PR).
BW
Wet-Heart Weight (WHW)
HW/BW and %M LDH
subunits increased with
low and high CO; after
removal of CO HW/BW and
%M LDH remained high
HW, LV, and RV increased
response greater in
younger group
SI, SP, CO increased;
SR and PR decreased
BW depressed;
WHW increased


Potential for cardiac DNA
synthesis and hyperplasia
ends between 5-25 days
postnatal life
Concluded that increased
CO via increased stroke
volume is compensation for
CO intoxication; increased
work may cause cardiomegaly
Increased HW due
to increased water content
Styka and Penney (1978)

Penney and Weeks (1979)
Penney et al. (1979)
Fechter et al. (1980)

-------
vo
^o
O
                            TABLE 10-5 (cont'd).  CARDIAC HYPERTROPHY STUDIES IN LABORATORY ANIMALS
Exposure"'11
CO = 500 ppm
for 32 days
(1982)
COHbc
38-42%
Animal
Rat
(n = 140)
Dependent Variable"*
Cardiomegaly
Results
HW/BW higher after
70 days of exposure
and after 30 days
Comments
Cannot be explained by
changes in DNA or
hydroxyproline
Reference
Penney et al. (1982)
                                                                                   of recovery; both RV
                                                                                   and LV were affected
        CO = 157-200 ppm    21.8-33.5%    Rat
        last 17 days                        (n = 96)
        gestation
                              RBC count, HW,
                              placenta! weight
                              (PW), cardiac
                              LDH M subunit,
                              Mb
                                          Depressed RBC; HW and
                                          PW increased; LDH(M)
                                          increased; Mb increased
                                                    Cardiomegaly not due to
                                                    elevated water content
                                                    (Disagrees with Fechter
                                                    et al., 1980)
                                                    Penney et al. (1983)
        CO = 500 ppm
        for 38-47 days
38-40%
Rat (n = 25)
Cardiac compliance and
dimensions
No change in compliance;
LV length and outside
diameter increased
Chronic COHb produces
eccentric Cardiomegaly
with no change in wall
stiffness
Penney et aJ. (1984a)
        CO = 200 ppm from
        Day 7 of pregnancy
        until parturition,
        and for 28 days fol-
        lowing parturition
             Rat (n >  180)     HW, RV, and LV weight
                                          RV increased with CO
                                          during fetal period,
                                          HW and LV increased
                                          with CO during
                                          postnatal period
                                                    Hemodynamic load caused
                                                    by CO during fetal period
                                                    results in Cardiomegaly
                                                    due to myocyte hyperplasia
                                                    Clubb et al. (1986)
        'Exposure concentration and duration.
        bl ppm = 1.145 mg/m3; 1 mg/m3 = 0.873 ppm at 25 °C, 760 mm Hg; 1%
        "Measured blood carboxyhemoglobin (COHb) levels.
        dSee glossary of terms and symbols for abbreviations and acronyms.
                                           10,000 ppm.

-------
        for cardiac enlargement is near 200 ppm CO (COHb = 12.03%), and unlike cardiac
        hypertrophy caused by altitude, which primarily involves the right ventricle, cardiac
        hypertrophy caused by CO involves the whole heart.
            The regression of cardiac hypertrophy in rats exposed continuously to moderate
 5      (400 ppm;  COHb = 35%) or severe (500-1100 ppm; COHb = 58%) CO for six weeks was
        followed by Styka and Penney (1978).  Heart weight to body weight ratio (HW/BW)
        increased from 2.65 in controls, to 3.52 and 4.01 with moderate and severe CO exposure,
        respectively.  Myocardial LDH M subunits (M LDH) were elevated 5 to 6% by moderate and
        12 to 14%  by severe CO exposure. Forty-one to 48 days after terminating the CO exposure,
10      Hb concentrations among groups did not differ significantly; HW/BW ratios were similar in
        the control and moderately exposed animals, but remained significantly elevated in the
        severely exposed animals.
            In addition to cardiomegaly,  Kjeldsen et al. (1972) has reported ultrastructual changes in
        the myocardium of rabbits breathing 180 ppm CO (COHb = 16.7%)  for two weeks. The
15      changes included focal areas of necrosis of myofibrils and degenerative changes of the
        mitochondria. In addition, varying degrees of injury were noted in the blood vessels.  These
        included edema in the capillaries; stasis and perivascular hemorrhages on the venous side; and
        endothelial swelling, subendothelial edema, and degenerative changes in myocytes on the
        arterial side.
20          The hemodynamic consequences of prolonged CO exposure have been examined in rats
        breathing 500 ppm CO (COHb =  38 to 42%) for 1 to 42 days, (Penney et al.,  1979) and in
        goats breathing 160 to 220 ppm CO (COHb = 20%) for two weeks (James et al., 1979).  In
        rats, cardiomegaly developed and stroke index, stroke power, and cardiac index increased;
        total systemic and pulmonary resistances decreased.  Left and right ventricular systolic
25      pressures, mean aortic pressure, maximum left ventricular dp/dt, and heart rate did not
        change significantly.  Penney et al. concluded that enhanced cardiac output, via an increased
        stroke  volume, is  a compensatory mechanism to provide tissue oxygenation during CO
        intoxication and that increased cardiac work is the major factor responsible for the
        development of cardiomegaly.  In chronically-instrumented  goats, James et al. noted that
30      cardiac index, stroke volume,  left ventricular contractility, and heart rate were all unchanged
        during exposure to CO, but  were depressed significantly during the first week following

        March 12,  1990                          10-55     DRAFT-DO  NOT QUOTE OR CITE

-------
       termination of the exposure.  Discrepancies between the Penney and James studies may be
       caused by differences in the CO concentrations or in the species used.
            Penney et al. (1984a) studied the compliance and measured the dimensions of
       hypertrophied hearts from rats breathing 500 ppm CO (COHb = 38 to 40%) for 38 to
 5     47 days. Heart weight to body weight ratios increased from 2.69 to 3.34.  Although
       compliance of the right and left ventricles was higher in the CO group, the differences
       disappeared  when the heart weight was normalized by body weight.  Left ventricular apex-to-
       base length and left ventricular outside diameter increased 6.4% and 7.3%, respectively; there
       were no changes in left ventricle, right ventricle, or septum thickness.  The authors  conclude
10     that chronic  CO exposure produces eccentric cardiomegaly with no intrinsic change  in wall
       stiffness.
            The consequences of breathing CO also have been investigated in perinatal animals.
       Prigge and Hochrainer (1977) reported elevated heart weights in fetuses from pregnant rats
       exposed for  21 days to CO concentrations as low as 60 ppm.  Because these animals
15     developed anemia rather than polycythemia, these workers discounted increased blood
       viscosity as a cause of the cardiomegaly.  Penney and Weeks (1979) examined the effects of
       inhaling 500 ppm CO  (COHb = 38 to 42%) until 50 days of age on cardiac growth in young
       (5 days) and old (25 days) rats.  They observed that the younger rats experienced the greatest
       change in heart weight and DNA synthesis and concluded that the potential for cardiac DNA
20     synthesis and muscle cell hyperplasia ends in rats during the 5th through 25th days of
       postnatal development.
            Fechter et al. (1980) reported elevated wet-heart weights at birth in neonatal rats from
       dams exposed throughout gestation to 150 ppm CO (COHb = 15%). There were no
       differences in dry-heart weight, total protein, or RNA or DNA levels; the differences between
25     groups in wet-heart weight disappeared after four days.  These workers concluded that the
       increased heart weight seen at birth in the CO-exposed rats is caused by cardiac edema.
            These results were not verified by Penney et al. (1983) in offspring of pregnant rats
       exposed to 157,  166, and 200 ppm CO (COHb  = 21.8 to 33.5%) for the last 17 of 22  days
       gestation. These workers observed that wet- and dry-heart weights increase proportionately
30     and conclude that cardiomegaly, present at birth, is not due to elevated myocardial water
       content. They also determined that cardiac LDH M subunit composition and Mb

       March 12, 1990                         10-56     DRAFT-DO NOT QUOTE OR CITE

-------
       concentration were elevated at 200 ppm CO.  They conclude that maternal CO inhalation
       exerts significant effects on fetal body and placenta! weights, heart weight, enzyme
       constituents, and composition.  Moreover, in newborn rats inhaling 500 ppm CO (COHb =
       38 to 42%) for 32 days and then developing in air, Penney et al. (1982) observed that
 5     HW/BW ratio increased sharply after birth, peaked at 14 days of age, and then fell
       progressively; it remained higher in rats exposed prenatally to CO than in control rats for up
       to 107 days of age.  The persistent cardiomegaly could not be explained by changes in DNA
       or hydroxyproline.
             Ventricular weights (wet and dry) and myocyte size and volume were measured in
10     perinatal rats exposed to 200 ppm CO by Clubb et al. (1986).  Pregnant rats were exposed to
       air or CO, and, at birth, pups from these two groups were subdivided into four groups:
       (1) control group (air/air), maintained in air in utero and postpartum; (2) air/CO group,
       received CO only postpartum; (3) CO/CO group; received CO in utero and postpartum;  and
       (4) CO/air group, received CO in utero, but in air postpartum.  Right ventricle weights were
15     increased in animals exposed to CO during the fetal period, but left ventricular weights were
       increased by CO during the neonatal period. Although HW/BW ratios increased to that of the
       CO/CO group by 12  days of age in animals exposed to CO postnatally only (air/CO),
       HW/BW ratios decreased to that of controls (air/air) by 28 days of age in animals exposed to
       air postnatally following fetal CO exposure (CO/air).  There was no difference in myocyte
20     volume between groups at birth. Left ventricle plus septum and right ventricle cell volumes
       of the CO/CO group were smaller than the controls at 28 days of age despite the heavier wet
       and dry weights of the CO/CO neonates. At birth, the CO-exposed animals had more
       myocytes in the right ventricle than the air-exposed controls; CO exposure after birth resulted
       in left ventricular hyperplasia.
25           Clubb et al. (1986) concluded that the increased hemodynamic load caused by CO
       during the fetal period results in cardiomegaly, characterized by myocyte hyperplasia, and this
       cellular response is sustained throughout the early neonatal period in animals exposed to  CO
       postpartum.
             Thus, results from animal studies indicate that inhaled CO can cause cardiomegaly, and
30     that the threshold for this response is near 200 ppm (COHb =  12%) in adult rats, and
       60 ppm in fetal rats.

       March 12, 1990                         10-57     DRAFT-DO NOT QUOTE OR CITE

-------
       10.3.4.5 Hematology Studies
            Increase in Hb concentration, as well as hematocrit ratio, is a well-documented response
       to hypoxia, which serves to increase the O2 carrying-capacity of the blood.  Guyton and
       Richardson (1961) and Smith and Crowall (1967) however, suggest that changes in hematocrit
 5     ratio not only affect the O2-carrying capacity of the blood, but blood flow as well. Therefore,
       when hematocrit ratios increase much above normal, O2 delivery to the tissues may be
       reduced because the resultant decrease in blood flow can more than offset the increased O2
       carrying capacity of the blood. Smith and Crowall conclude that there is an optimum
       hematocrit ratio at sea level that shifts to a higher value with altitude acclimation.
10     Presumably a similar compensation would occur when O2 transport is reduced by CO.
            Changes in Hb concentration and hematocrit ratio have been reported in numerous
       animal studies (see Table 10-6).  In dogs exposed to 50 ppm CO (COHb = 7.3%) for
       three months, Musselman et al. (1959) reported a slight increase in Hb concentration (12%),
       hematocrit ratio (10%), and in red blood cells (10%). These observations were extended by
15     Jones et al. (1971) to include several species of animals exposed to 51 ppm or more CO
       (COHb = 3.2 to 20.2%), intermittently or continuously, for up to 90 days. There were no
       significant increases in the Hb and hematocrit values observed in any of the species at
       51 ppm CO (COHb = 3.2 to 6.2%).  At 96 ppm CO (COHb = 4.9 to 12.7%), significant
       increases were noted in the hematocrit value for monkeys (from 43 to 47%) and in the Hb
20     (from 14.0 to 16.49%) and hematocrit values (from 46 to 52%) for rats.  Hemoglobin and
       hematocrit values were elevated in rats (14 and 10%, respectively) guinea pigs  (8 and 10%,
       respectively), and monkeys (34 and 26%, respectively) exposed to 200 ppm CO
       (COHb = 9.4 to 12.0%); they also were elevated in dogs, but there were too few animals to
       determine statistical significance. However,  in dogs exposed to CO (195 ppm;
25     COHb = 30%) for 72 h, Syvertsen and  Harris (1973) reported that hematocrit and Hb
       increased from 50.3 to 57.8% and 15.0  to 16.2 g%, respectively.  The differences in
       hematocrit and Hb occurred after 72 h exposure and were attributed to increased
       erythropoiesis. Penney et al. (1974b) observed significant increases in Hb (from 15.6 to
       16.7 g%) in rats exposed to  100 ppm  CO over several weeks and conclude that the threshold
30     for the Hb response is close to 100 ppm (COHb = 9.26%).
       March 12, 1990                         10-58     DRAFT-DO NOT QUOTE OR CITE

-------
                        TABLE 10-6. HEMATOLOGY STUDIES IN LABORATORY ANIMALS
CJ
1— »
ro
\f
VO
s




£
VO
O
»
Exposure*'1"
CO = 50 ppra
continuously for 3 mo
CO = 51,96, or 200 ppm
for 90 days

CO = 67.5 ppm
22 h/day, 7 day/
for 2 years
CO = 195 for 72 h
CO = 100 ppm, 46 days;
200 ppm, 30 days;
500 ppm, 20-42 days
CO = 200 ppm last
18 days gestation
COHbc
7.3%
3.296
1.8%
3.2-6.2%
4.9-12.7%
9.4-20.2%
depending
upon species
1.9-5.5%
and
2.8-10.2%
-30%
9.20%
15.82%
41.12%
27%
Animal Dependent Variable1*
Dog (n = 4) Hb, Hct, and RBC,
Rabbit (n = 40) EKG
Rat (n = 100)
Rat (n = 35) Hb
Guinea pig (n = 35)
Monkey (n = 9)
Dog (n = 6)

Cynomolgus Hct, Hb RBC counts
monkey (n = 27)
Dog (n = 12) Hct and Hb
Rat (n = 32) Hb
Rat Hb, Hct, and RBC
Results
Hb, Hct, RBC increased
in dogs and rabbit; no
change in EKG in dog
Increases in rats at 96,
106, and 200 ppm;
increases in all animals
at 200 ppm

No effects
Both increased
Increased at all levels
Hb, Hct, and RBC ail
lower
Comments
No toxic signs in dogs,
rabbits or rats


Unusual variation in
COHb
Increase due to
erythropoiesis
About 30 days until Hb
approached asymptotic
values

Reference
Musselman et at. (1959)
Jones et al. (1971)

Eckhardt et al. (1972)
Syvertsen and Harris (1973)
Penney et al. (1974b)
Penney et al. (1980)
50 and 100 ppm
for 6 weeks on various
2.6-12%
Dog (n = 46)
                         Hb
                                         No effects
                                                                           Preziosi et al. (1970)
o
o
§
3
o
c
o
a
2
so
n
intermittent daily
schedules
CO = 50 ppm, 95 h/week
whole natural life
expectancy up to
2 years (also short-term)


Rat (n = 336) EKG, organ weights, No effects Also showed no effects Stupfel and Bouley (1970)
Mouse (n = 767) Hb, Hct, and RBC on other variables



"Exposure concentration and duration.
bl ppm = 1.145 mg/m3; 1 mg/m3 =
°Measured blood carboxy hemoglobin
0.873 ppm at 25°C, 760 mm Hg; 1 % = 10,000 ppm.
(COHb) levels.
dSee glossary of terms and symbols for abbreviations and acronyms.

-------
            Several groups have reported no change in Hb or hematocrit following CO exposure.
       Thus, Preziosi et al. (1970) observed no significant change in Hb concentration in dogs
       exposed to 50 and 100 ppm CO (COHb = 2.6 to 12.0%) for six weeks.  In monkeys,
       exposed to 20 and 65 ppm CO (COHb = 1.9 to 10.2%) for two years, Eckardt et al. (1972)
 5     noted no compensatory increases in Hb concentration or hematocrit ratio.  In mice exposed
       5 days a week to 50 ppm CO for one to three months, Stupfel and Bouley (1970) observed no
       signfiicant increase in Hb.
            Interestingly, in fetuses removed from pregnant rats after 21  days exposure to CO,
       Prigge and Hochrainer (1977) reported a significant increase in fetal hematocrit (from 33.3 to
10     34.5%) at 60 ppm and a significant decrease in Hb and hematocrit at 250 ppm (from 9.1 to
       8.0 g% and from 33.3 to 28.4%, respectively)  and 500 ppm (from 9.1 to 6.5 g% and from
       33.3 to 21.9%, respectively). These results were confirmed by Penney et al. (1980) who
       reported significantly lower Hb (12.6 vs. 15.8 g%), hematocrit (46.2 vs. 54.4%), and RBC
       counts (27.2 vs. 29.1%) in newborns from pregnant rats exposed to 200 ppm CO (COHb =
15     27.8%) for the final 18 days of development than in controls.  However, in a later study,
       Penney et al. (1983) reported that although RBC count was depressed in neonates from
       pregnant rats exposed to 157, 166, and 200 ppm CO (COHb = 21.8 to 33.5%) for the last
       17 out of 22 days gestation, mean corpuscular Hb and volume were elevated.
            The results from animal studies indicate inhaled CO can increase Hb concentration and
20     hematocrit ratio and that the threshold for this response, at least in rats, appears to be near
       100 ppm (COHb = 9.26%).  Small increases in Hb and hematocrit probably represent a
       compensation for the reduction in O2-transport caused by CO.  At higher CO concentrations,
       excessive increases in Hb and hematocrit may impose an additional workload on the heart and
       compromise blood flow to the tissue.  The O2 transport system of the fetus is especially
25     sensitive to CO inhaled by the mother, and it may be affected by CO concentrations as low as
       60 ppm.

       10.3.4.6 Atherosclerosis and Thrombosis
            The section dealing with cholesterol and atherosclerosis in the previous air quality
30     criteria document for CO (U.S. Environmental Protection Agency, 1979) described about
       March 12, 1990                         10-60     DRAFT-DO NOT QUOTE OR CITE

-------
        12 publications. These studies generally utilized animal models of atherosclerosis or animal
        models describing arterial wall cholesterol uptake in response to COHb concentrations ranging
        from 4.5 to 41.1% (see Table 10-7). The conclusion was that the evidence failed to support
        conclusively a relationship between CO exposure and atherosclerosis in animal models.  Since
 5      completion of the  1979 air quality criteria document,  a number of additional studies have
        been published (Table 10-7). However, taken in aggregate, the studies still fail to
        conclusively prove an atherogenic effect of exposure to low doses of CO.
             Astrup et al. (1967) described atheromatosis as well as increased cholesterol
        accumulation in aortas of rabbits fed cholesterol and exposed to CO (170 to 350 ppm;
10      COHb  = 17 to 33%) for 10 weeks. These observations were not verified, however, by
        Webster et al. (1970) who observed no changes in the aorta or carotid arteries or in serum
        cholesterol levels in squirrel monkeys fed cholesterol  and exposed intermittently to CO (100
        to 300 ppm; COHb = 9  to 26%) for seven months; they did note enhanced atherosclerosis in
        the coronary arteries.  Davies et al.  (1976) confirmed that coronary artery atherosclerosis was
15      significantly higher in rabbits fed cholesterol and exposed intermittently to CO for 10 weeks
        (250 ppm; COHb  = 20%); but they also reported no  significant differences between groups
        in aortic concentrations of triglycerides, cholesterol, phospholipids,  or plasma cholesterol.
             In cynomolgus monkeys fed cholesterol and exposed intermittently to CO for 14 mo (50
        to 500 ppm; COHb = 21.6%),  Malinow et al.  (1976) observed no differences in plasma
20      cholesterol levels or in coronary or aortic atherosclerosis.  Armitage et al. (1976) confirmed
        that intermittent CO (150 ppm;  COHb = 10%  for 52 and  84 weeks) did not enhance the
        extent or severity of atherosclerosis in  the normal White Carneau pigeon. While CO
        exposure did increase the severity of coronary artery atherosclerosis in birds fed cholesterol;
        the difference between groups, noted at 52 weeks,  was not present after 84 weeks.
25           Stender et al. (1977) exposed rabbits that  were fed high levels of cholestrol  to CO for
        six weeks continuously and intermittently (200 ppm; COHb = 17%).  In the cholesterol-fed
        group, CO had no effect  on free- and esterified-cholesterol concentrations in the inner layer of
        the aortic wall.  In the normal group, CO increased the concentration of cholesterol in the
        aortic arch, but there was no difference in the cholesterol content of the total aorta.
30           Hugod et al.  (1978), using a blind technique and the same criteria to assess  intimal
        damage as was used in earlier studies (Kjeldsen et al., 1972; Thomsen, 1974; Thomsen and

        March 12, 1990                          10-61      DRAFT-DO NOT QUOTE OR CITE

-------
1
cr
to
                                      TABLE 10-7.  ATHEROSCLEROTIC STUDIES IN LABORATORY ANIMALS
Exposure"'1"
COHbc
Animal
Dependent Variable"1
                                           Results
                                                                                 Comments
                                                                                                                Reference
        CO = 170 ppm for
        8 week, then 350 ppm
        for last 2 week, fed
        cholesterol
                   17-33%
           Rabbit (a = 24)
                 Atherosclerotic
                 changes
                    Increased aortic ather-
                    omatosis and cholesterol;
                    Local degenerative
                    signs and hemorrhages in
                    hearts
Not verified in subsequent
studies
Astrup et al. (1967)
        CO = 100-300 ppm
        4 h/day, 5 days/week
        for 7 mo,
        fed cholesterol
                   9-26%
           Squirrel monkey
                 Atherosclerosis
                 in various blood
                 vessels plus serum
                 cholesterol
                    Increased coronary athero-
                    sclerosis but no effects
                    on aorta or carotid
                    arteries or serum
                    cholesterol
                           Webster et al. (1970)
        CO = 250 ppm            20.6%
        continuously for 2 week
        CO = 50, 100, and         4.5%
        180 ppm for periods        9.0%
        ranging from 30 min -
        24 h, and from
        2-11 days
                               Cynomolgus monkey    Coronary artery
                               (n = 20)              pathology
                               Rabbit (n = 61)        Aortic damage
                                                     Subcndothelial edema,
                                                     gaps between endothelial
                                                     cells, infiltration cells
                                                     containing lipid droplets

                                                     Increased aortic intimal
                                                     lesions at 180 ppm CO
                                                     for 4 h or more
                                                              Lipid-laden cell findings
                                                              suggest greater sensitivity
                                                              of monkeys than of rabbits
                                                              Postulates 180 ppm CO for
                                                              4 h is threshold for
                                                              injury
                                                                        Thomsen (1974)
                                                                        Thomsen and Kjeldsen (1975)
CO = 150 ppm 6 h/ 10%
day, 5 day/week for
52 and 84 weeks, fed
cholesterol
CO = 250 ppm 4 h/ 20%
day, 7 day/week,
10 week


White carncau Severity of
pigeon (n = 180) alherosclerosis
Rabbit (n = 24) Blood cholesterol;
coronary artery
atherosclerosis;
aortic cholesterol
content
No effect in normocholes-
terolemic birds; coronary
artery atherosclerosis
significantly enhanced in
hypercholesterolemic
birds at 52 weeks
Increased atherosclerosis
in coronary arteries but
no differences in aortic
or plasma cholesterol

No significant changes in
coronary arteries after
84 weeks
Study disagrees with Astrup
et al. (1967)


Armitage et al. (1976)
Davies et al. (1976)



-------
                              TABLE 10-7 (cont'd).  ATHEROSCLEROTIC STUDIES IN LABORATORY ANIMALS
o\
Exposure"-1" COHb°
CO = 50-500 ppm, 21.6%
12 h/day for 14 mo


CO = 200 ppm, 17%
continuously or
12 h/day for 6 week




CO = 200 ppm for -
5-12 weeks, 2,000 ppm
for 320 min; 4,000 ppm
for 205 min

CO = 400 ppm for 23%
10 alternate half-
hours of each day
for 12 mo
Animal Dependent Variable'1
Cynomolgus monkey Aortic and coronary
(n = 26) atherosclerosis


Rabbit (n = 30) Cardiovascular
pathology





Rabbit (n = 150) Coronary artery
and aortic damage



Cynomolgus monkey Cholesterol content
(n = 11) of vessels and
plasma

Results
No effects



No differences in
atherosclerosis but CO
produced higher serum
cholesterol levels



No effect




No effect on plasma-free
cholesterol, cholesterol
ester, tri- and diglycerides,
and phospholipids; no
Comments
Subjects on high- and low-
cholesterol diets;
disagrees with Astrup
et al. (1967).
Serum cholesterol was con-
trolled by adjusting
individual diets; apparently
coronary atherosclerosis
in Astrup et al. (1967) was
caused by increased serum
cholesterol
Inability to reproduce
earlier results may be due
to lack of blind technique
and smaller number of
animals in earlier studies
Agrees with Malinow et al.
(1976)


Reference
Malinow et al. (1976)



Slender et al. (1977)






Hugod et al. (1978)




Bing et al. (1980)



       Smoked 43 cigarettes
       per day for 14-19
       mo; fed choles-
       terol
       CO = 200-300 ppm
       continuously for
0.6-1.9%    Baboons (n = 36)
Serum cholesterol
           Rabbits (n = 140)
Myocardial mor-
phology using
electron microscopy
significant increase in
cholesterol content of
aorta; no histologic
damage and no fat
deposition

No significant differences
in serum total cholesterol,
VLDL + LDL cholesterol,
HDL cholesterol, or
triglyceride concentrations

No histotoxic effects
Rogers et al. (1980)
Hugod (1981)

-------
i
H-*
to
TABLE 10-7 (cont'd).  ATHEROSCLEROTIC STUDIES IN LABORATORY ANIMALS
Exposure*'11
Smoked 43 cigarettes
per day for up to
33 mo; fed
cholesterol



COHb0 Animal
0.64-2.0% Male baboons
(n = 36)
0.35-1.13% Female baboons
(n = 25)


Dependent Variable*1 Results Comments
Serum cholesterol No significant differences
in serum total cholesterol,
VLDL -f LDL cholesterol,
HDL cholesterol, or
triglyceride concentrations;
slightly enhanced plague
formation in carotid artery;
no difference in lesions or
vascular content of lipid
or prostaglandin in aorta
or coronary arteries
Reference
Rogers el al. (1988)




        CO = 100 ppm
        8 h/day, 5 day/week
        for 4 mo,
        fed cholesterol
                              6.8-7.6%
           Pigs (n = 38)
           (normal or
           homozygous and
           heterozygous
           for von
           Willebrand's
           disease) with
           balloon-catheter
           injury of coro-
           nary arteries
Coronary artery
and aortic lesions
No significant changes
Sultzer et al. (1982)
 O
 o
        "Exposure concentration and duration.
        bl ppm = 1.145 mg/m3; 1 mg/m3 = 0.873 ppm at 25°C, 760 mm Hg; 1% = 10,000 ppm.
        "Measured blood carboxyhemoglobin (COHb) levels.
        dSee glossary of terms and symbols for abbreviations and acronyms.


-------
        Kjeldsen, 1975), noted no histologic changes in the coronary arteries or aorta in rabbits
        exposed to CO (200, 2000,  or 4000 ppm) for 0.5 to 12 weeks. These workers suggested that
        the positive results obtained earlier were due to the non-blind evaluation techniques and the
        small number of animals used in the earlier studies.  Later, Hugod (1981) confirmed these
  5     negative results using electron microscopy.
             Only a few of the studies published since completion of the  1979 criteria document have
        demonstrated a significant atherogenic effect of low-level CO exposure.  Turner et al. (1979)
        showed that CO enhanced the development of coronary artery lesions in White Carneau
        pigeons that were fed a diet of 0.5 and 1 %, but not 2%, cholesterol. The exposure was to
 10     150 ppm for six hours, five days each week  for 52 weeks (COHb = 10%-20%).  Plasma
        cholesterol levels may have  been increased slightly by the CO, but this was significant
        (p<0.5) only at Week 11.  Marshall and Hess (1981) exposed minipigs to 160, 185, and
        420 ppm CO for four hours per day for 1 to 16 days (COHb = 5 to 30%).  The higher
        concentrations were associated with adhesion of platelets to arterial endothelium and to fossae
 15     of degenerated endothelial cells.  Additional  changes at the higher concentration included an
        increased hematocrit, an increase in blood viscosity, and an increase in platelet aggregation.
             Alcindor et al.  (1984)  studied rabbits with induced hypercholesterolemia. Three sets of
        rabbits were studied.  The first was a control group receiving a normal diet and breathing air.
        The second group was given a 2%  cholesterol diet. The third group was given the same diet
20     and exposed to 150 ppm CO.  COHb levels were not  reported. Low-density lipoprotein
        particles in the CO-exposed  rabbits were richer in cholesterol and had a higher cholesterol-to-
        phospholipid molar ratio than did the particles from the nonexposed rabbits after 45 days
             Other animal studies have given generally negative results.  Bing et al. (1980) studied
25      cynomolgus monkeys (Macacafascicularis). Four animals were used as controls.  Seven
        were exposed to CO at a level of 400 ppm for 10 alternate half-hours of each day during
        12 mo. Carboxyhemoglobin levels showed a gradual increase to a peak at five hours of 20%.
        The monkeys had no histologic evidence of atherosclerosis, vessel wall  damage, or fat deposi-
        tion in the arterial wall.  There was no significant change in cholesterol or in lipoprotein
30      levels.  High density to total cholesterol ratios did  not differ between the CO-exposed and air-
        exposed animals. These  animals were on a normal diet with no augmentation of cholesterol

        March 12, 1990                          10-65      DRAFT-DO NOT QUOTE OR CITE

-------
       or fat content. The study demonstrated that even high levels of CO exposure are not
       invariably followed by arterial injury or abnormal lipid accumulation.
            Similar negative results were reported by Sultzer et al. (1982) who studied swine. Pigs
       with and without von Willebrand's disease were divided into groups that were exposed to
 5     intermittent, low-level CO or to air.  CO was delivered at 100 ppm for eight hours each week
       day for four months.  COHb levels averaged 7% after five hours of exposure. The degree of
       coronary and aortic atherosclerotic lesion development in response to a 2% cholesterol diet
       was similar in the two exposure groups. There was no effect of the ambient CO on the
       degree of hypercholesterolemia induced by the diet.  The findings showed no obvious effect
10     of CO on atherogenesis in hypercholesterolemic pigs.
            A number of studies have examined the contribution of CO in cigarette smoke to the
       purported effects of smoking on atherogenesis and thrombosis. Rogers et al. (1980) fed a
       high-cholesterol diet to 36 baboons for up to 81  weeks.  The animals were taught to puff
       either cigarette smoke or air by operant conditioning using a water reward.  Half of the
15     baboons smoked 43 cigarettes each day. The baboons were given a cigarette or sham every
       15 min during a 12-h day except during times of blood drawing.  Average COHb levels in
       smokers were about 1.9%.  Only slight differences in the very low-density lipoprotein
       (VLDL), LDL and high-density lipoprotein (HDL) levels were noted between the smokers
       and nonsmokers.  Additionally, platelet aggregation with adenosine 5'-phosphate (ADP) and
20     collagen was similar in the two groups.
            Rogers et al. (1988) extended their previous study of male baboons for an additional
       1.2 years of diet and smoking (total diet, 3.2 years; total smoking,  2.8 years).  They also
       studied a separate group of 25 female baboons that received the diet for 2.6 years and were
       exposed to cigarette smoke for 1.6 years.  Blood levels of COHb were determined by GC and
25     reported both as total concentration in milligrams per deciliter and as percent saturation of
       Hb, as calculated by a validated linear regression equation. Levels of COHb in the male
       baboons averaged 0.64%  at baseline, whereas COHb was on an average of 0.35% in female
       baboons at baseline.  The weekly averages of COHb levels determined after 57 weeks were
       2.01 and 1.13% in male and female baboons, respectively. The baseline cholesterol levels
30     were 105 mg/dL and 88 mg/dL in the two groups of baboons. Levels at 16 weeks were
       226 mg/dL in males and 291 mg/dL in females.  There were no significant differences in total

       March 12, 1990                          10-66      DRAFT-DO NOT QUOTE OR CITE

-------
        cholesterol, HDL cholesterol, or LDL cholesterol between smokers and controls.  There were
        slightly more fatty streaks and fibrous lesions in the carotid arteries of smokers than in
        controls.  No differences in lesion prevalence, vascular content of lipids, or prostaglandins
        were seen in aorta or coronary arteries.
 5           The results reported by Rogers et al. (1980; 1988) suggest little if any effect of cigarette
        smoking on atherosclerotic lesion development in baboons. How these findings can be
        extrapolated to effects of smoking in humans is difficult to know.  The COHb levels attained
        in the experimental animals were barely 2%.  Levels in human smokers are probably 4% or
        more during the waking hours of the day.  On the other hand, the findings are consistent with
10      most studies of the effects of low levels of CO on atherogenesis.
             A study of cockerels by Penn et al.  (1983) has shown negative results as well. Three
        groups of cockerels, each including seven animals, were used to determine if the atherogenic
        effect of cigarette smoke could be separated from an effect due solely to CO. Cockerels
        develop aortic fibromuscular atherosclerotic lesions spontaneously. Various agents, including
15      some carcinogens, have been shown to accelerate the growth in thickness and extent of these
        lesions.  The authors used this model by exposing one group of animals to the smoke from
        40 cigarettes each day for five days each week.  The cockerels were exposed  from about six
        weeks of age until about 22 weeks of age. A similar group of animals was exposed to CO
        calibrated to give a similar COHb level to that achieved in the animals  exposed to cigarette
20      smoke.  The third group of animals was exposed to filtered air.  Carboxyhemoglobin levels
        following an exposure session were measured at 6, 9, 12, and 15 weeks.  The average level
        in the air-exposed animals was 1.6%.  Levels in the cigarette smoke and CO groups were 6.7
        and  7.2%, respectively.  Atherosclerosis was quantified both by the extent of the aorta
        involved and by the cross-sectional area of the intimal thickening.  The cigarette smoke-
25      exposed group had more aortic lesions and lesions with greater cross-sectional area than did
        either the CO-exposed group or the air-exposed group.  This difference was significant at
        p<0.05 in a one-tailed chi-square test. The data suggest that  atherogenic effects of cigarette
        smoke are not solely attributable to CO.
             It has  been postulated that a possible atherogenic effect of CO may be mediated through
30      an ability of CO to enhance platelet aggregation or some other component of thrombosis.
        This possibility was raised in the study by Marshall and Hess (1981) noted above.  Other

        March 12, 1990                          10-67     DRAFT-DO NOT QUOTE OR CITE

-------
        studies, however, have demonstrated that the effect of CO is to depress platelet aggregation.
        In one study (Mansouri and Perry, 1982), platelet aggregation to epinephrine and arachidonic
        acid was reduced in in vitro experiments in which CO was bubbled through platelet-rich
        plasma.  Similarly, platelets from smokers aggregated less well than platelets from
 5      nonsmokers, although this inhibition of aggregation was not correlated with the level of
        COHb.
            Madsen and Dyerberg (1984) extended these observations by studying effects of CO and
        nicotine on bleeding time in humans.  Smoke from high-nicotine cigarettes caused a
        significant shortening of the bleeding time.  Smoke from low-nicotine cigarettes caused no
10      significant change in bleeding time. CO inhalation sufficient to raise the COHb to 15% was
        followed by a shortening of the bleeding time (6.0 minutes to 4.8 min), but for a short period
        of time (< 1.5 h).  After administration of aspirin, neither nicotine nor CO affected bleeding
        times or platelet aggregation. The findings suggest that the proaggregating effects of cigarette
        smoke are mediated through an inhibitory effect of nicotine on prostacyclin (PGIj) production.
15      Effects of CO in the smoke seem to be minor and short lived.
            These findings were corroborated by Renaud et al. (1984). The effects of smoking
        cigarettes of varying nicotine content on plasma clotting times and on aggregation of platelets
        with thrombin,  ADP, collagen, and epinephrine were studied in 10 human subjects.  Both the
        clotting functions and platelet aggregation were increased with increasing nicotine content in
20      cigarettes.  There was no correlation of these parameters, however, with COHb levels.
        COHb levels, reported as percent increase from baseline, achieved about a 60% increase.
            Effeney (1987) has provided convincing evidence that these effects of nicotine and CO
        on platelet function are mediated through opposing effects on PGI2 production.  Four rabbits
        were exposed to CO in an exposure chamber at 400 ppm for 7 to  10 days. Carboxy-
25      hemoglobin levels averaged about 20%.   Ten rabbits received an infusion of nicotine for 7 to
        10 days.   Full-thickness  samples of atrial and ventricular myocardium were incubated with
        arachidonic acid for determination of PGI2 production by radioimmunoassay of 6-keto-PGF,a
        and by inhibition of platelet aggregation.  Carbon monoxide exposure increased PGI2
        production which was significant in ventricular myocardium.   Nicotine exposure reduced
30      PGI2 production in all tissues examined.  The combination of nicotine and CO caused a net
        March 12, 1990                         10-68      DRAFT-DO NOT QUOTE OR CITE

-------
       increase in PGI2 production.  The effect of CO may be to induce hypoxemia, a known
       stimulant of PGI2 production.  This effect of CO would serve to reduce aggregation.
            Another explanation for the antiaggregatory effect of CO exposure recently has been
       provided by Brune and Ullrich (1987).  These investigators bubbled CO through platelet-rich
 5     plasma and then challenged the platelets with various agonists.  The CO exposure was much
       greater than that encountered in physiologic or even toxic states.  The results, however,
       indicated that inhibition of aggregation was related to enhancement of guanylate cyclase action
       and associated increased cyclic guanosine monophosphate (cGMP) levels.

10     10.3.5  Summary and Conclusions
            The 1984 Addendum to the 1979 Air Quality Criteria Document for Carbon Monoxide
       (U.S. Environmental Protection Agency, 1984) reported what appears to be a linear
       relationship between level of COHb and decrements in human exercise performance,
       measured as maximal O2 uptake.  Exercise performance consistently decreases at a blood level
15     of about 5.0% COHb in young, healthy, nonsmoking individuals  (Klein et al., 1980;  Stewart
       et al., 1978; Weiser et  al., 1978).  Some studies have even observed a decrease in
       performance at levels as low as 2.3 to 4.3% COHb (Horvath et al., 1975;  Drinkwater et al.,
       1974; Raven et al., 1974a); however, this decrease is so small as  to be of concern mainly for
       competing athletes rather than for ordinary people conducting the activities of daily life.
20     Cigarette smoking  has a similar effect on cardiorespiratory response to exercise in nonathletic
       human subjects indicating a reduced ability for sustained work (Hirsch et al., 1985; Klausen
       etal., 1983).
            Since the 1979 Air Quality Criteria Document, several important studies appearing in
       the literature have expanded the cardiovascular data base. Adverse effects in patients with
25     reproducible exercise-induced angina (Allred et al., 1989a,b) have been  noted with
       postexposure COHb levels (CO-Ox measurement) as low as  3.2% (corresponding to an
       increase of 2.0% from baseline).  Sheps et al. (1987) also found a similar effect in a group of
       patients with angina at COHb levels of 3.8%  (representing an increase of 2.2% from
       baseline). Kleinman et al. (1989) studied subjects with angina and found an effect at  3%
30     COHb representing an increase of 1.5% from baseline.  Thus, the lowest observed adverse
       effect level in patients with stable angina is somewhere between 3 and 4%  COHb (CO-Ox

       March 12,  1990                         10-69     DRAFT-DO NOT QUOTE OR CITE

-------
       measurement), representing an increase from baseline of from 1.5 to 2.2%. Effects on silent
       ischemia episodes, which represent the majority of episodes in these patients, have not been
       studied (see Chapter 12).
            Exposure sufficient to achieve 6% COHb recently has been shown to adversely affect
 5     exercise-related arrhythmia in patients with CAD (Sheps et al., 1989).  This finding combined
       with the epidemiologic work of Stern et al. (1988) in tunnel workers is suggestive but not
       conclusive that CO exposure may provide an increased risk of sudden death from arrhythmia
       in patients with CAD.
            There is also strong evidence from both theoretical considerations and experimental
10     studies in animals that CO can adversely affect the cardiovascular system.  Tables 10-4
       through 10-7 are summaries of the data pertinent to the effects of CO on the cardiovascular
       systems of experimental animals. Accordingly, disturbances in cardiac rhythm and con-
       duction have been  noted in healthy and cardiac-impaired animals at CO concentrations of 50
       to 100 ppm (COHb = 2.6 to  12%); alterations in various hemodynamic parameters have been
15     observed at CO concentrations of 150 ppm (COHb = 7.5%); cardiomegaly has been reported
       at CO  concentrations of 200 ppm (COHb = 12%) and 60 ppm in adult and fetal animals,
       respectively;  changes in Hb concentrations have been reported at CO concentrations of
       100 ppm (COHb = 9.26%) and 60 ppm in adult and fetal animals, respectively.
            There is conflicting evidence that CO exposure will enhance development of
20     atherosclerosis in laboratory animals; and most studies show no measurable effect.  Similarly,
       the possibility that CO will promote significant changes in lipid metabolism that might
       accelerate atherosclerosis is suggested in only a few studies. Any such effect must be subtle
       at most.  Finally, CO probably inhibits rather than promotes platelet aggregation.  Except for
       the studies by Rogers et al. (1980, 1988) on baboons, the CO exposures used in the studies
25     on atherosclerosis created COHb levels of 7% or higher; sometimes much higher. While
       occupational  exposures in some workplace situations might regularly lead to COHb levels of
       10% or more, such high-exposure levels are almost never encountered in the
       nonoccupationally  exposed general public.  In this general population, exposures are rarely as
       much as 25 to 50 ppm, and COHb levels typically are below 3% in nonsmokers (see
30     Chapter 8).  When examined in this context, this review, therefore, provides little data to
        March 12, 1990                         10-70      DRAFT-DO NOT QUOTE OR CITE

-------
       indicate that an atherogenic effect of exposure would be likely to occur in human populations
       at commonly encountered levels of ambient CO.
 5     10.4 CEREBROVASCULAR AND BEHAVIORAL EFFECTS OF
             CARBON MONOXIDE
       10.4.1   Control of Cerebral Blood Flow and Tissue PO2 with Carbon
                Monoxide and Hypoxic Hypoxia
       10.4.1.1 Introduction
10          The effect of CO on cerebral blood flow (CBF) and cerebral O2 consumption (CMROj)
       is complicated by the relationship between CBF and cerebral O2 delivery or availability.
       Alterations in cerebral neurological function, as evaluated by neurological symptoms or
       changes in evoked potential responses, are particularly difficult to correlate with changes in
       CBF or  cerebral O2 delivery.  One of the most fundamental challenges to the organism is to
15     obtain O2 from its environment and deliver it to the tissues. However, each tissue or organ
       may have regulatory mechanisms to obtain O2 which differ from other tissues or organs.
       Literature concerning the cerebrovascular effects of CO is incomplete and in many cases
       conflicting, and, despite the enormous literature concerning hypoxia and the cerebrovascula-
       ture, the mechanisms that regulate the cerebral vessels during hypoxia are unclear.
20          Kety and Schmidt (1948) demonstrated that CMRO2 is about 3.5 mL of O2 per 100 g of
       brain (cerebral hemispheres) per minute in normal adult man.  This consumption of O2 is
       virtually unchanged under a variety of conditions.  About one-half of CMRO2 is dedicated to
       synaptic transmission (Astrup, 1982; Donegan et al., 1985) and this remains relatively
       constant under all conditions.  Half of the remaining, vegetative O2 consumption, a quarter of
25     the overall value, maintains resting neuronal membrane potentials.  The remaining quarter is
       consumed by a variety of unidentified, but presumably irreducible, processes (Astrup, 1982).
       In order for the brain to maintain its CMRO2 it has but two adaptations:  (1) the brain could
       extract more O2 from the blood or (2) CBF could increase.  In fact, the brain generally relies
       largely on increasing CBF for its major adaptability mechanism to provide more O2 to the
30     tissue.  Thus, the following discussion concerns the regulation of CBF with hypoxia, with
       little discussion of the regulation of O2 extraction.

       March 12, 1990                         10-71     DRAFT-DO NOT QUOTE OR CITE

-------
            The cerebral vasculature responds to decreases in O2 availability by increasing CBF in
       order to maintain cerebral O2 delivery and/or by increasing O2 extraction in order to maintain
       cerebral O2 utilization when cerebral O2 delivery is limited.  Compared with other forms of
       cerebral O2 deprivation, such as hypoxic hypoxia (lowered inspired O2 concentration) and
 5     anemia, CO hypoxia may interfere with O2 delivery and cerebral O2 utilization through effects
       on the Hb dissociation curve and on the cytochrome oxidase system.  During O2 deprivation
       (hypoxia) CBF and cerebral O2 delivery may be altered by hemodynamic responses,
       specifically changes in cerebral perfusion pressure, as well as the absolute amount of O2
       limitation (arterial O2 content). Because hypoxia adversely effects cerebral autoregulation,
10     hypertension during hypoxia may result in an increased CBF and, hence, cerebral O2
       availability. Conversely, hypotension during hypoxia may decrease CBF and cerebral O2
       delivery. In the following sections, the effects of hypoxia (hypoxic and CO) on the cerebro-
       vasculature will be demonstrated and the potential mechanisms of action of hypoxia on
       cerebral blood vessels will be examined.  The effects of CO on global and regional CBF, and
15     the effects of both high and low levels of CO on CBF and CMRO2 also will be examined.
       An attempt will be made to demonstrate the  potential mechanisms of action of hypoxia on the
       cerebrovasculature, and the synergistic effects of CO and cyanide hypoxia on the cerebral
       circulation will be examined.

20     10.4.1.2 Effects on Global Cerebral Blood Flow
            At the present time, there is conflicting information concerning whether the
       cerebrovascular response to CO is similar to other forms of cerebral hypoxia, such  as hypoxic
       hypoxia and anemic hypoxia.  Few studies are available in which other types of hypoxia have
       been compared to CO hypoxia, especially at similar levels of O2 deprivation.  In addition,
25     comparison of cerebrovascular effects of CO and other types of hypoxia from laboratories of
       different investigators is difficult because of differences in anesthetic techniques, use of
       different animal  species, and use of different CBF techniques.  An important point  to
       emphasize when comparing CO hypoxia to hypoxic hypoxia is that although arterial O2
       content is reduced with both hypoxic and CO hypoxia, there is no reduction  in arterial O2
30     tension with CO hypoxia.  Comparisons of the equivalent effects of both CO and hypoxic
       hypoxia on  CBF and CMRO2 in the same animal preparations have been made by Traystman's

       March 12, 1990                          10-72     DRAFT-DO NOT QUOTE OR CITE

-------
        laboratory in several studies (Traystman and Fitzgerald, 1977; Traystman et al., 1978;
        Fitzgerald and Traystman, 1980; Traystman and Fitzgerald, 1981; Koehler et al., 1982;
        Koehler et al., 1984; Koehler et al.,  1983; Koehler et al., 1985). The concept of equivalent
        effects of both types of hypoxia (hypoxic and CO) has been described by Permutt and Farhi
 5      (1969) and involves the comparison of physiologic effects of elevated COHb and low O2 at
        equal reductions in Hb, arterial O2 content,  arterial or venous O2 tension, or blood flow.  The
        focus of investigation of several laboratories, including Traystman's, concerning the effects of
        hypoxia on the cerebral vasculature is not merely to describe the well-known vasodilation that
        occurs, but to examine the mechanism that produces this vasodilation. This is where much of
10      the controversy lies. Here issues focus on the importance of local mechanisms in controlling
        CBF such as tissue acidosis (Kety and Schmidt, 1948; Molnar and Szanto, 1964) and the
        direct effect of hypoxia on vascular smooth  muscle (Detar and Bohr, 1968).  Some  years ago
        it also had been postulated that O2-sensitive  carotid arterial chemoreceptors might play a role
        in the CBF response during hypoxia (Ponte  and Purves, 1974). Other groups of investigators
15      additionally suggested that carotid baroreceptor stimulation produces cerebral vasoconstriction
        (James and MacDonnell,  1975; Ponte and Purves, 1974).  This vasoconstrictor response to an
        elevation  in blood pressure could contribute to the autoregulatory responses of the cerebral
        vessels.  Because the carotid and aortic chemoreceptors are stimulated by certain forms of
        hypoxia, as is systemic arterial blood pressure, the underlying mechanism of the cerebral
20      vasodilator response to hypoxia is complicated.
            Traystman et al. (1978) previously reported that the increase in CBF in dogs during a
        reduction in arterial O2 content, produced by breathing the animal with a low O2 gas mixture
        (hypoxic hypoxia), was not different  from the increase in CBF when O2  content was
        decreased by adding CO to the breathing gas mixture (CO hypoxia)  (Figure 10-3).  This was
25      true both before and after carotid sinus chemodenervation.  This study also demonstrated that
        with hypoxic hypoxia mean arterial blood pressure increased, whereas it decreased with CO
        hypoxia.  Because the CBF increase with CO and hypoxic hypoxia was similar,
        cerebrovascular resistance actually decreased more with CO hypoxia.  This represents the
        effect of the carotid chemoreceptors on systemic blood pressure during each type of hypoxia.
30      When the carotid chemoreceptors were denervated, cerebrovascular  resistance decreased to the
       March 12, 1990                          10-73     DRAFT-DO NOT QUOTE OR CITE

-------
     O
     O
     o
     -J
     CD
     
-------
        same level as with CO hypoxia. They concluded that the carotid chemoreceptors do not play
        an important role in the global cerebral vasodilator response to either CO or hypoxic hypoxia.
             There were two important limitations of that study, however.  One involved the
        possibility that the aortic chemoreceptors (aortic bodies) might have a role in controlling
 5     CBF, and this was not considered at that time. The aortic bodies have been reported to have
        a role in the control of pulmonary blood flow (Stern et al., 1964). The other involved the
        fact that sectioning  of the carotid sinus nerves denervated not only the carotid
        chemoreceptors, but also the carotid sinus baroreceptors. Because hypoxic hypoxia and CO
        hypoxia affect blood pressure, they therefore could modulate chemoreceptor input by
 10     baroreceptor input.  Traystman and Fitzgerald (1981) demonstrated that the carotid and aortic
        chemoreceptors are not necessary for the increase in CBF with hypoxia and that the increase
        in CBF is not modified by the carotid and aortic baroreceptors (Figure 10-4).  They also
        showed that the cerebral vasodilation to hypoxia in carotid chemoreceptor-denervated animals
        and in carotid sinus nerve sectioned and vagotomized animals resembles that occurring in
 15     animals exposed to  CO hypoxia, with intact chemoreceptors in which both the arterial O2
        tension is high  and the chemoreceptors may not be activated (Figure 10-5). CMRO2 remained
        at control values under both hypoxic hypoxia and CO hypoxia conditions and  was unchanged
        by any denervation  condition.  These data support the notion that  the brain increases its blood
        flow in response to its O2 needs with both hypoxic and CO hypoxia in control or baroreceptor
 20     and chemoreceptor denervated dogs in order to maintain CMRO2 constant. The cerebral
        blood vessels appear to be relatively unresponsive to reflex stimuli (Heymans  and Bouckaert,
        1932; Heistad and Marcus, 1978; Heistad et al., 1976), and the CBF responses to low-
        inspired O2 or elevated CO are not dependent on either the carotid or aortic chemoreceptors.
        These responses also are  not modified by either the carotid or aortic baroreceptors.  These
25     findings would be most compatible with the idea that control of the cerebral vasculature
        during hypoxia is mediated locally, however it remains possible that central (brainstem)
        mechanisms are involved. These brainstem mechanisms concerning the CBF responses to
        hypoxic hypoxia have been studied little, and their possible involvement in CBF responses to
        CO have not been examined at all.
30
       March 12, 1990                          10-75     DRAFT-DO NOT QUOTE OR CITE

-------
         •Jo _

         i*i
         UIQ X
         KO e
         "o 1
         < o
          o
         e
         <
 60


 40


 20


  0

140


120


100


 BO


 60
       0

      80


      60


      40


      20
S"l
(E  —
UJ
U
                     CONTROL
                     ntzi
                   »   « HTPO'IC HTPQXIA

                   O •- • O CO HYPOKIA

                    ( ) % el CONTROL
                    "5   To  Ft  5b
                                 BARO
                             DENERVATED
                                      . 
-------
                           CONTROL
                                                  VAGOTOMY
£
-! o _
ecu. -f
co E
UJQ V
CCO "c
o° -
CO

UJ
cc
3
_l V)
< CO —
cc <£ x
UJQ. £
cc o 3
< 0
o
CO
cc
<
3 uj !s
<=M
j'S i
< io x
CC yj c

2 * E
f^ **•*
cc
UJ
u
60

40


20

160


1 40

120

100
4
4
O1
•
•"••'.i,
. .d.u'K^
^s**v».
•IP*
•



• (112) i
^^\^
(98) J 	 ?^


» • HYPOXIC HYPOXIA
• 0 	 0 CO HYPOXIA
5 ( ) •/. of CONTROL
4
•
t(l78)»
•(I"»'*S.
^«s«^
^T
•

.,M»I
^^^.^
^^V^^
^^i
X-T
x»''
• (83) H(,''

J
i . « •


6.0

4.0
2.0



°C

j£
• (63) t*f^''
• (50) •¥



• 1 « ^^^^J


/S
*<59) JS .•'
T ^x
. .(46)*'



I i i J
) 5 10 15 20 0 5 10 15 20


ARTERIAL 02 CONTENT (vol %)
Figure 10-5.  Effects of hypoxic and carbon monoxide hypoxia on cerebral blood flow, mean
arterial blood pressure, and cerebral vascular resistance in control and vagotomized animals.
Data points and bars represent means ± SE of five animals. Numbers in parentheses are
percent of control.  *p < 0.05.
Source: Traystman and Fitzgerald (1981).

March 12, 1990
                                    10-71     DRAFT-DO NOT QUOTE OR CITE

-------
            The idea, however, that CBF responses to CO are always similar to those of hypoxic
       hypoxia is not universal.  Studies on fetal (Jones et al., 1978) and newborn lambs (Jones
       et al., 1981) have demonstrated that CBF increases during hypoxic hypoxia may correlate
       better with decreased arterial O2 content than with decreased arterial O2 partial pressure. The
 5     description of hypoxia in terms of arterial O2 content rather than arterial O2 partial pressure is
       more than simply an arbitrary choice between two similar variables.  When one considers
       hypoxic hypoxia as a fall in arterial O2 content, this emphasizes the importance of cerebral O2
       delivery to the microvascular exchange site, whereas O2 partial pressure emphasizes diffusion
       from the exchange site to the parenchyma. The studies previously mentioned (Jones et al.,
10     1978; Jones et al., 1981) demonstrated that the relationship between CBF and arterial  O2
       content is such that the product of CBF and arterial O2 content, which equals cerebral  O2
       delivery, is essentially constant as arterial O2 partial pressure falls. The study in newborn
       lambs (Jones  et al., 1978) demonstrated that arterial fractional O2 extraction (CMRO2 per
       amount of O2 delivered)  was well maintained in both anemic and hypoxic hypoxia conditions.
15     The maintenance of cerebral O2 delivery  and fractional O2 extraction during anemic and
       hypoxic hypoxia is not unique to  the lamb and does apply to adults of other species (Jones
       etal., 1981).
            Koehler et al. (1982), working with a newborn-lamb model, tested the hypothesis that
       CBF and CMRO2 bear relationships to arterial O2 content during CO hypoxia that are  not
20     different from those occurring during hypoxic hypoxia.  They reasoned that if these
       relationships differ between hypoxic hypoxia and CO hypoxia, then other effects of CO
       exposure, such as the shift in the O2Hb dissociation curve or histotoxic effects, need to be
       considered. Koehler et al. (1982) found that CO hypoxia causes a 47% greater increase in
       CBF compared with hypoxic hypoxia for a similar reduction in arterial O2 content
25     (Figure 10-6).  A greater CBF response to CO hypoxia than to anemic hypoxia also has been
       reported in humans (Paulson et al., 1973).  In the study of Koehler et al. (1982), CMRO2  and
       O2 delivery were constant during hypoxic hypoxia.  Thus, fractional O2 extraction, which
       equals O2 consumption divided by O2 delivery, remained constant with hypoxic hypoxia.
       During CO hypoxia, although CMRO2 remained constant, O2 delivery  increased and fractional
30     O2 extraction decreased. This decline in fractional O2 extraction was correlated with the
       leftward shift of the O2Hb dissociation curve that accompanied CO hypoxia.  In this situation,

       March 12, 1990                          10-78     DRAFT-DO NOT QUOTE OR CITE

-------
              250
              200

   Cerebrol
     Blood     ,50
     Flow
(ml/min/lOOg)
               100
                50
                                                  •— Control ond HH
                                                  0— COM
                               0.4           0.6           0.8
                               Fractional Arterial O  Saturation
                                       1.0
   Figure 10-6.  Cerebral blood flow as a function of fractional arterial O2 saturation. Circles
   represent control or hypoxic hypoxia; squares represent CO hypoxia.

   Source: Koehler et al. (1982).
March 12, 1990
10-79     DRAFT-DO NOT QUOTE OR CITE

-------
       the additional increase in CBF with CO hypoxia could be explained because the shift in the
       curve lessens the O2 diffusion gradient into the tissue and further lowers the arterial O2 partial
       pressure.  In other words, while increases in CBF maintain O2 availability at the micro-
       vascular exchange site, overall O2 transport to the cells becomes relatively more diffusion-
 5     dependent with CO hypoxia.  Although these investigators believe that the best explanation
       for the difference in the CBF response to hypoxic hypoxia and CO hypoxia is the leftward
       shift of the O2Hb dissociation curve, they cannot completely rule out a potential histotoxic
       effect of CO resulting from the competition of CO and O2 for cytochrome aa3 oxidase.  This
       effect generally is considered to be insignificant at low CO levels because in vitro cytochrome
10     oxidase remains completely oxidized until very low tissue O2 partial pressure levels are
       reached. However, Hempel et al. (1977) has shown  that cerebral cytochrome aa3 in vivo is in
       a substantially reduced state, raising the possibility that CO may readily compete with O2 at
       relatively low CO levels.  This binding relationship between CO and the oxidase also has
       been demonstrated by Piantadosi et al. (1985; 1987).  On the  other hand, if CO were exerting
15     a histotoxic effect, CMRO2 would be expected to fall, and this was not observed in Koehler's
       experiments. It also is possible, however, that CO could have a direct histotoxic effect on
       cerebral vascular smooth muscle, independently of brain tissue metabolism.
            In a subsequent study, Koehler et al. (1984) compared the effect of hypoxic hypoxia and
       CO hypoxia on CBF in adult and newborn sheep in which arterial O2 content was reduced to
20     50 to 60% of control with both types of hypoxia.  During hypoxic hypoxia, CBF increased to
       maintain cerebral O2 delivery in both adults and newborns; however CMRO2 did not change.
       Although CMRO2 was higher in newborns, the responses of CBF/CMRO2 to hypoxic hypoxia
       was not different in newborns and adults.  In newborns and adults, CBF increased to a greater
       extent with  CO hypoxia than with hypoxic hypoxia for similar reductions in arterial O2
25     content (Figure 10-7).  This resulted in an increase in cerebral O2 delivery  with CO hypoxia.
       As discussed previously, the degree to which CO hypoxic differed from hypoxic hypoxia
       correlated with the magnitude of the leftward shift of the O2Hb dissociation curve that
       accompanies CO hypoxia. In the adult animals with  CO hypoxia, CMRO2 was reduced by
       16%, however, CMRO2 was maintained in the newborns.  These data allowed for the
30     conclusion that maintenance of cerebral O2 delivery during hypoxic hypoxia is a property of
       CBF regulation common to both newborn and adult sheep. During CO hypoxia, the position

       March 12, 1990                          10-80     DRAFT-DO NOT QUOTE OR CITE

-------
        .50
  I
   **   .40
     •
   "o
   >
   ~   .30
   ^.
     N
     O
     >    20
    I
     CSJ
    O JO
    o
  o
                                               8
                                               (vol. %)
                          12
16
Figure 10-7. Comparison of newborn and adult responses of the reciprocal of the cerebral
arteriovenous O2 content difference (C.O2 - QO^-l to a reduction in arterial O2 content
during hypoxic hypoxia (HH).  Open circles, room-air control in newborns; solid circles, HH
in newborns; open triangles, room-air control in adults; solid triangles, HH in adults.
Regression lines were fitted to the reciprocal of C.O2. For newborns (solid line), (C.O2 -
CAH = 1.74 C.O2-1 + 0.01 (r = 0.91). For adults (dashed line), (C.O2 - CvO2)-l =
1.64 C.O2 + 0.02 (r = 0.67).  Responses of blood flow per unit O2 consumption are not
significantly different between newborns and adults.

Source:  Koehler et al. (1985).
March  12, 1990
10-81     DRAFT-DO NOT QUOTE OR CITE

-------
        of the OzHb dissociation curve is an additional factor that sets the level of O2 delivery.  The
        fetal conditions of low arterial-O2 content and a left-shifted OaHb dissociation curve may have
        provided the newborn with a microcirculation better suited for maintaining CMRO2 during CO
        hypoxia.
 5           Some points need to be considered when comparing the cerebrovascular effects of
        hypoxic hypoxia and CO hypoxia as described in the dog studies (Traystman and Fitzgerald,
        1981; Traystman et al., 1978) with those described in the newborn and adult sheep.  In the
        studies concerning the anesthetized dogs, the CBF response to hypoxic hypoxia and CO
        hypoxia was not statistically different although the mean  blood flows tended to be higher with
10      moderate levels of CO. This response was statistically significant in the newborn and adult
        sheep experiments.  One likely explanation for the different result is that arterial blood
        pressure declined during CO hypoxia in the anesthetized  dog, whereas it was well maintained
        in the unanesthetized sheep. Because cerebral autoregulation is impaired during severe
        hypoxia (Haggendal and Johannsson, 1965), a drop in perfusion pressure during CO hypoxia
15      may have limited the increase in CBF in the dog.  A reexamination of data from  the dog
        study indicated that cerebral O2 delivery increased and fractional O2 extraction decreased
        during CO hypoxia. Thus, the dog study is consistent with the data obtained in the sheep
        study.  Similar results have been reported in humans (Paulson et al., 1973) and in goats
        (Doblar et al., 1977).  Another possible explanation for the differences in CBF responses to
20      hypoxia in the dogs versus the sheep is that the dogs were anesthetized with sodium
        pentobarbital, whereas the sheep were studied in the unanesthetized state. Pentobarbital
        anesthesia reduces CBF and CMRO2 so that differences in flow would be minimized in the
        dog studies. Finally, the Px of sheep Hb is considerably higher than in the dog (44 mmHg
        for sheep vs. 27 mmHg for dogs) so that the leftward shift of the O2Hb dissociation curve
25      would be larger in sheep and therefore result in a greater increase in CBF with CO hypoxia
        versus hypoxic hypoxia.  Evidence supporting a role for  Px in the CBF response  to CO was
        obtained by Koehler et al. (1983) in experiments in which lambs were first exchange
        transfused  with high-Pjo donor blood, which resulted in an  increase in cerebral fractional O2
        extraction.  With the induction of CO hypoxia to return Px to the pretransfusion level,
30      cerebral O2 delivery and O2 extraction also returned to pretransfusion levels.  These
        investigators suggested that since PK can affect capillary  and tissue O2 partial pressure

        March 12, 1990                          10-82     DRAFT-DO NOT QUOTE OR CITE

-------
       independent of arterial O2 content, the position of the O2Hb dissociation curve appears to set
       the level of cerebral O2 delivery about which CBF is regulated when arterial O2 content is
       reduced.  These data, taken together with the previously mentioned work from this group, are
       consistent with the existence of a  tissue O2 tension-dependent mechanism controlling the
 5     cerebral vasculature in which tissue O2 tension is a function of CMRO2, cerebral O2 delivery
       to the microcirculation, the position of the O2Hb dissociation curve, and microcirculatory
       morphology.

       10.4.1.3  Effects on Regional Cerebral Blood Plow
10           Both human and animal histopathology studies have suggested that there are regional
       differences in tissue injury following severe CO exposure.  One potential source of these
       differences is regional differences in the CBF response to CO exposure.  Two logical
       comparisons of the regional CBF  response to CO hypoxia are (1) anatomical, (i.e.,  rostral to
       caudal [cortex to brainstem] comparison) and (2) physiological (i.e. brain areas with a func-
15     tional blood brain barrier versus brain areas without an intact blood brain barrier). Koehler
       et al. (1984) observed interesting  regional CBF responses to hypoxic and CO hypoxia in
       newborn  lambs and adult sheep (Table 10-8; Figure 10-8).  In adults, regions with high
       normoxic blood flows such as the caudate nucleus and midbrain showed a large response to
       hypoxia,  whereas lower blood flow regions with large white matter tracts,  such as the cervical
20     spinal cord, pons, diencephalon, and piriform lobe, showed a relatively lower response.
       Other brain regions were essentially homogeneous in their responses. CO  hypoxia increased
       CBF to a greater extent than hypoxic hypoxia in all brain regions, but the overall pattern of
       regional CBF was similar for the  two types of hypoxia in the adults.  In Table 10-8 the
       regions are listed in order from highest to lowest responsivity, and the particular groups of
25     regions that are significantly different are separated by pairs of vertical  brackets.  In the
       newborns, regional responses differed for each type of hypoxia.  With hypoxic hypoxia in the
       newborns, the brainstem regions had a significantly greater response than all other regions
       except the caudate nucleus, whereas all cerebral lobes responded significantly less than all
       other regions.  With CO hypoxia, the difference between brainstem responses and those of
30     other regions was less marked. In the adults, in contrast, there was no  significant interactive
       effect between the type of hypoxia and the pattern of regional response.  With both types of

       March  12, 1990                           10-83      DRAFT-DO NOT QUOTE OR CITE

-------
                                        jEMiar JBEI •'fipa® %jeq^n^a^g HE

                                           nn
E
                                                   r
                                                r
                                           4-
                                           I
                                              r r
                                              ,j  ,i
                                              I    I
                                                                         fit

-------
I
   i
   i
 I
,t
n

I
P
 ,

it
      41
         I
      lit
   "i.1
  flri
                  i    11  r
                              IRINAU

                              G^MIHILLUM

                                MIOUUUA

                                 P3NS

                               MiDlRAIN
                               CAUDATE N.
                                      I,


                                      L-:

                               F NONTAX I
                                                            1
                                                           >   E

-------
       reported.  Okeda et al. (1987) also demonstrated CO-induced regional CBF differences in
       cats.  This group attempted to demonstrate that there is a selective vulnerability of the
       pallidum and cerebral white matter and showed low CBF values for these brain areas. In the
       newborn lambs (Koehler et al.,  1984), unlike the adult sheep, the patterns of regional CBF
 5     responses were not similar with the two types of hypoxia. Brainstem regions, especially the
       medulla, had marked responses relative to the cerebrum during hypoxic hypoxia.  Peelers
       et al. (1979) also have made this observation in unanesthetized fetal lambs.  Rosenberg et al.
       (1982) found in the fetal and neonatal lamb, but not in the adult sheep, that the brainstem also
       had a greater CBF response to arterial CO2 than other regions, and Cavazzuti and Duffy
10     (1982), in newborn dogs, observed results consistent with those in the lamb in that brainstem
       regions displayed a greater blood flow response to hypoxic hypoxia and to hypercapnia.  The
       most likely explanation in the puppy for the higher brainstem response to hypoxia and
       hypercapnia is that this region has a higher CMRO2.  Normoxic glucose consumption in vivo
       (Cavazzuti and Duffy, 1982) and O2 consumption in vitro (Himwich and Fazekas, 1941) are
15     relatively high in the brainstem of the neonatal puppy.  This also may be true in the neonatal
       lamb.  However, if this were the only explanation, there should be a much greater CBF
       response in brainstem regions relative to the rest of the brain during CO  hypoxia, as  well as
       hypoxic hypoxia, but this has not been observed.  The larger CBF response of caudate
       nucleus with hypoxia compared with cortical lobes may be explained by  the  high fraction of
20     grey matter in the caudate nucleus. The large response of the brainstem region also may  be
       partly explained by a relatively high proportion of grey matter.  It also is likely that increased
       activation of cardiovascular and respiratory centers in the brainstem during hypoxia produces
       local increases in metabolism and O2 demand, which in turn would produce an additional
       increase in blood flow in this area. The capability of the central  nervous system to increase
25     CMRO2 during hypoxic hypoxia has been demonstrated in certain strains of rats (Berntman
       et al.,  1979).  An alternative explanation for the regional differences in sensitivity to hypoxia
       is that stimulation of the peripheral chemoreceptors by hypoxia produced sympathetic
       vasoconstriction preferentially in the cerebral hemispheres.  This explanation is considered
       unlikely because previous studies have shown that neither carotid nor aortic chemoreceptor
30     denervation alters CBF from cortical regions during hypoxic or CO hypoxia (Traystman and
       Fitzgerald, 1981; Traystman et al., 1978).

       March  12, 1990                          10-86     DRAFT-DO NOT QUOTE  OR CITE

-------
            There is at least one area of the brain that does not respond to alterations in arterial O2
       content and partial pressure as do other brain areas, the neurohypophysis. The
       neurohypophysis is an anatomically unique region of the brain, and the regulation of blood
       flow to this area appears to be different from other areas of the brain.  Hanley et al. (1986)
 5     demonstrated that when arterial O2 content was reduced equivalently with hypoxic hypoxia
       and CO hypoxia, global CBF increased by 239 and 300%, respectively.  Regional CBF also
       showed similar responses for all brain areas except the neurohypophysis. With hypoxic
       hypoxia, neurohypophysis blood flow increased markedly (320%), but it was unchanged with
       CO (Figure 10-9). These blood flow responses of the neurohypophysis occur independently
10     of alterations in blood pressure.
            Wilson et al. (1987) determined the role of the chemoreceptors in the neurohypophyseal
       response to hypoxia and found that chemoreceptor denervation completely inhibited the
       increase in neurohypophyseal blood flow associated with hypoxia. The response to CO was
       unaltered (Figure 10-10).  These data (Hanley et al., 1986; Wilson et al., 1987) demonstrated
15     that the mechanism responsible  for the increase in neurohypophyseal blood flow with hypoxia
       is unique when compared to other brain regions.  The only animals in which
       neurohypophyseal blood flow did not respond were the denervated animals and those given
       CO. Both of these conditions are ones in which the chemoreceptors have been shown to be
       inactive (Traystman and Fitzgerald, 1981; Comroe, 1974).  Although for most brain regions
20     the blood flow response to hypoxia does not involve the peripheral chemoreceptors, this is not
       true for the neurohypophysis. Thus,  here is an example of one regional brain area that does
       not respond to CO hypoxia (i.e., a change in arterial O2 content) but does respond to a change
       in arterial O2 tension. This suggests that the chemoreceptor represents the mechanism
       involved in the neurohypophyseal response to hypoxic hypoxia and that local changes in
25     arterial O2  content are not involved in this response, because the neurohypophysis does not
       respond to CO.  It is unclear whether other  regional brain areas have similar responses and
       mechanisms to hypoxic and CO hypoxia.

       10.4.1.4 Effect of Low Levels of Carbon Monoxide on Cerebral Blood Flow
30          Little information is available concerning the effects of low levels of CO on the cerebral
       vasculature. This is particularly unfortunate because many investigators have shown

       March 12,  1990                          10-87     DRAFT-DO NOT QUOTE OR CITE

-------
        CONTROL I      CD CONTROL 2
                                               HYPOXIC HYPOXIA
                           CO-HYPOXIA
  O
  O
  z
  2
  13
  u.
  CD
  U
200
160

160
140

120
100
 60
 60
 40

 20
  0
           CEftCBMAl
           HEMISPMCJU
                      CAUOATC
                           WMITC
                           MATTER
KTPOTHALAMUS  COVKUJLM
                       MCOIAN
                       CMINtMX
Figure 10-9.  Effect of hypoxic hypoxia and carbon monoxide (CO) hypoxia on
neurohypophyseal and regional cerebral blood flow (rCBF). Each bar represents mean ± SE
of five dogs.  Both types of hypoxia (diagonal and cross-hatched bars) produced significant
increases from control (open and dark bars) in blood flow to all regions except
neurohypophysis.  Both parts of neurohypophysis, median eminence and neural lobe, showed
no change from control with CO hypoxia but did not have significant flow responses to
hypoxic hypoxia.  Note changes in vertical axis at right for median eminence and neural lobe
blood flow.

Source: Hanley et al. (1986).
March 12, 1990
                                  10-88     DRAFT-DO NOT QUOTE OR CITE

-------
8
CD
125


CO


75


50


25
                 KNCRVATO

                 INTACT
                                                               2000

                                                                     o
                                                                     CD
                                                               1600  3?
                                                               1200
                                                               600
                                                               400
                                             5
                                             i
                                            *>
                                             §
         NORMOXIA
HYPOXIC
HYPOXIA
                                  NORMOXIA
HYPOXIC
HYPOXIA
       CEREBRAL HEMISPHERES
                NEUROHYPOPHYS1S
 Figure 10-10. Effect of complete chemoreceptor denervation on total cerebral and
 neurohypophyseal blood flow. Each line represents mean ± SEM of six dogs.  (Note change
 in y axis for neurohypophyseal blood flow.)

 Source: Wilson et al. (1987).
 March 12, 1990
          10-89    DRAFT-DO NOT QUOTE OR CITE

-------
       behavioral and electrophysiological abnormalities with various levels of CO exposure
       (Xintaras et al.,  1966; Beard and Wertheim, 1967; Fodor and Winneke, 1972; Horvath et al.,
       1971), and it is conceivable that these effects could result from abnormalities of CBF.
       Carbon monoxide hypoxia results in an increase in CBF, and this has been demonstrated by a
 5     number of investigators (Traystman and Fitzgerald, 1981; Traystman et al., 1978; Sjostrand,
       1948; Haggendal and Norbeck, 1966; Paulson et al., 1973).  However, many difficulties have
       been encountered in these experiments such as extracranial contamination of the measured
       CBF, surgical trauma to the cerebral vasculature, inadequate control of blood gases, and
       failure to measure COHb concentrations.
10          Traystman (1978) examined the CBF responses to CO hypoxia in anesthetized dogs,
       particularly in the range of COHb less than 20% (Figure 10-11). A COHb level as low as
       2.5% resulted in a small, but significant,  increase in CBF to 102%  of control. With
       reductions in O2-carrying capacity of 5, 10, 20,  and 30% (COHb 5, 10, 20, and 30%) CBF
       increased to approximately 105, 110, 120, and 130% of control, respectively.  At each of
15     these levels, CMRO2 remained unchanged.  At COHb levels above  30%, CBF increased out
       of proportion to the decrease in O2-carrying capacity, but the brain  could no longer maintain
       CMRO2 constant.  At a COHb level of 50%, CBF increased to about 200% of control. These
       findings are in general agreement with those of MacMillan (1975) who demonstrated that as
       COHb increased to 20, 50, and 65%, CBF increased to 200, 300, and then 400%,
20     respectively, in  cats. These CBF increases at 20% COHb are higher than those reported in
       Traystman's (1978) study but the reason for this is not known.  Haggendal and Norbeck
       (1966) demonstrated a 50 to 150% of control increase in CBF with COHbs of 30 to 70%, and
       Paulson et al. (1973) showed a 26% increase in CBF with a COHb of 20%.  These findings
       also indicate that CBF increases progressively with increasing COHb concentrations, and that
25     CMRO2 is maintained constant even at a COHb level of 30%. This has important
       implications regarding the behavioral and electrophysiological consequences of CO exposure.
       These findings also would be consistent with those of Dyer and Annau (1978) who found that
       superior colliculus-evoked potential latencies are not affected by COHb levels up to 40%. At
       levels above this, the brain cannot increase blood flow enough to compensate for decreased
30     tissue O2 delivery.  At these high COHb  levels, then, behavioral and neurophysiological
       abnormalities should be quite evident. At lower COHb levels, these abnormalities should not

       March 12, 1990                         10-90     DRAFT-DO  NOT QUOTE OR CITE

-------
  c
  O
  u
 Q
 O
 O
 m
 m
 LU
 o:
 Ld
 O
220
210 -
200 -
190 -
180 -
170 -
160 -
150 -
140 -
130 -
120 -
110 -
100
                        10           20           30
                              CARBOXYHEMOGLOBIN, %
                                                         40
50
Figure 10-11. Effect of increasing carboxyhemoglobin levels on cerebral blood flow, with
special reference to low-level administration (below 20% COHb).  Each point represents the
mean ± SE of 10 dog preparations.
Source: Traystman (1978).
March 12, 1990
                            10-91     DRAFT-DO NOT QUOTE OR CITE

-------
                      OdiPMOHiiSy., m x oanpnoMBBd timriliiji IHH irialanr fLc.,, nations wife
                     |miijyit cmiaB ant •«.««•«?. ifc Hnnd Ham or rmttaeA^
                                                               CadtGKtiOB tD I
                                               clfal* of hjyuua at these Icwds CLci.,, vpto
                                   r «'Jm MiiijAjMii J
20
               iiKjibp^
              addfeakiH. the idea of a Ifltao&nld teiidl bdov vlodi donees M CGHb varid ant
                                        T 197g|
                 I978|.  A JftErzataA! fcicB
15     a^dtoiJiq|iiyairinEpk^dBaoi^^
                                 BeaBnd and Weriham, 1967).  Beczne Tn^tmn (1978|
                      CfflF and ititaMe CMRO, lavtoCXHblevdbof 30%,tiUssflgge^aliiii»t
                                  M^                                       Sevoal
20     alimillin^iaiCTF^Alreiiomiiypa^  McDowan (1966) reported a tinesfaoM xnciial OL
             rf SP mmUjg m samto&far^ Ajfp.  CBF bpgan to increase as aiterial Qz tension
                      H|u ^                                         KognrcetaL
             oaofinnBd MdXwnalS''s ^'"'^1^c
-------
                      tcfQK. Aho, becaanecf Inssrf fp^ftaniliiiflirii artcrinies(DBiiagctaL,
        1979) and fesjgn«rt slope of
                                   •jpednfic finKtaan. flat applies owr a. wide aange of Oj,
 10
                                                                      absorption is 2

                            in fine deadis (Radfanl et >L, 19H6). Cyanide also has been
2©
                 «*r  - -
                       1959; Meyer, 1963; RriedyetaL, 1976).  Gonatose stales and deep
               i of dectncal activity have been obsetvod in a vaiiety of y*"»f following, cyanide
                   (WanlandWheafley, 1947; Brierty, 1975).  Cyanide appeared to sdectmahj
                        te matter (South et aL, 1963; Levine, 1967), but it is unclear wteHief
25     dfonennipalholo^iRas due to diiect effects of cyanide on n^
                       i or ischemia. Uris neuropathology aho may have been due to n^analty
                      Ldradati
           The action of cyanide on ffle cendbnl vasodatnie and on CMRO^ has not been studied in
       any peat Asaffl nnder cimiiiJlrf cnMBiiMif  Hin^iffiH-it in meihuds, anhnal species.,,
       dosages, finhne to nxasnie blood or tissue cyanides, lack of control in regard toother
                   as respiration and consequently CO^ and diffiailty in securing pore cerebral
       March 12, 1990                        1O^3     DRAFT-DO NOT QUOTE OR CHE

-------
       venous samples, have led to some inconsistencies in observations and interpretations of
       cyanide and the cerebral circulation.  This has been so for CO as well.  McGinty (1929)
       observed an increase in the outflow from the sagittal sinus of anesthetized dogs following a
       small dose of sodium cyanide. Paulet (1958) observed an increase in cerebrospinal fluid
 5     pressure following a bolus injection of cyanide, and concluded that CBF must have increased.
       Studies in cats and rabbits showed large increases in blood flow through the sagittal sinus
       without any alteration in the arterio-venous O2 content difference across the blood brain
       barrier (Russek et al., 1963). Russek observed an increase in CMRO2 to 300% of control in
       these experiments.  He concluded that stimulation of the carotid chemoreceptors caused the
10     increase in CMRO2 and that the increase in CBF was  secondary to the metabolic change.
       Brierly et al. (1976) speculated that CBF decreased during cyanide administration due to
       cyanide's depressive effect on the myocardium.  However, he never actually measured CBF,
       but rather he made speculations on CBF that were based upon vascular pressure changes in
       the sagittal sinus of monkeys.  Aliukhin et al. (1974)  also observed increases in CBF
15     following acute cyanide poisoning in rats.
             The cerebral metabolic response to cyanide has been studied in vitro and in vivo.
       McGinty (1929) observed an outpouring of lactic acid into the cerebral venous drainage
       during moderate cyanide  intoxication, and this was the first indication of an altered cerebral
       aerobic metabolism in cyanide hypoxia. Fazekas et al.  (1939) reported a decreased CMRO2
20     in dogs following administration of potassium cyanide, however, these workers did not
       measure CBF.  Doses of cyanide sufficient to decrease cerebral cytochrome oxidase activity
       by 50% in rats led to increases in lactate, inorganic phosphate, and triphosphate (Albaum
       et al., 1946).  Olsen and Klein (1947) reported similar metabolic findings in rats and
       calculated that glucose consumption must have increased to account for the rise in lactic acid.
25     They also discussed how the addition of cyanide to brain slices, in vitro, failed to decrease O2
       consumption until high cyanide levels were achieved.  Gasteva and Raize (1975) demon-
       strated a reduction in CMRO2 with cyanide.
             Early studies of the combined effects of CO and cyanide were inconclusive in describing
       any interaction of these two agents (Hofer, 1926; Moss et al., 1951).  Two inhalation
30     toxicological studies have reported no interaction between CO and cyanide with respect to
       lethal-dose levels (Higgins et al., 1972; Yamamoto, 1976).  Smith et al. (1976b) using

       March 12, 1990                           10-94      DRAFT-DO NOT QUOTE OR CITE

-------
       behavioral assessments reported at least an additive effect of cyanide and CO on the time to
       cause incapacitation and death in exercising rats.  However, few studies have ever examined
       the combined effects of these agents on the cerebrovasculature.  Pitt et al. (1979) examined
       individual and combined effects of cyanide and CO on CBF and CMRO2 because many of the
 5     deleterious effects of the fire environment may be due to altered cerebral nervous system
       function.  They found that CBF increases during  CO hypoxia and it also increases with
       cyanide hypoxia.  CMRO2 remained unchanged until the higher levels of either cyanide or CO
       were reached. These data are consistent with those previously presented for CO. When CO
       and cyanide were administered simultaneously, CBF increased in an additive manner
10     (Figure 10-12), but significant decreases in CMRO2 occurred at the combination of the lower
       concentrations (Figure 10-13).  These data suggest that CO and cyanide are physiologically
       additive in producing changes in CBF, but may act synergistically on CMRO2.
            Figure 10-14 (from Pitt et al., 1979) demonstrates the relationship between CBF and
       CMRO2 (VEOz) in CO and cyanide hypoxia.  Three aspects of this figure suggest that cyanide
15     and CO may act through similar mechanisms with respect to changes  in CBF and CMRO2.
       First, low doses of either agent alone produce increases in CBF that maintain CMRO2
       constant.  Second,  higher doses of CO or cyanide increase CBF to 200% of control while
       CMRO2 decreased  to around 80%  of control.  Finally, combinations of cyanide and CO
       hypoxia result in an increased CBF with a decreased CMRO2 that would be predicted on the
20     basis of an additive effect.  Although CO binds to nonhemoglobin proteins, including
       cytochrome  oxidase in vitro, it is unlikely that in vivo this binding contributes to the hypoxic
       effect of CO (Root, 1965). Because we demonstrated previously that the effect of CO
       hypoxia on the brain was similar to an equivalent reduction in arterial O2 content with hypoxic
       hypoxia, it may be that the mechanism that mediates the increase in CBF to maintain CMRO2
25     relatively constant  with CO and hypoxic hypoxia, may be similarly affected by blocking
       cellular respiration with cyanide hypoxia.  The nonspecificity of the hypoxic response
       suggests a common mechanism of cerebral vasodilation with hypoxia, and it lends support to
       a metabolic control of cerebral vessels although the precise mediator is unknown.
            As Pitt et al.  (1979) explains, there are several explanations for the loss of maintaining
30     CMR02 constant when CBF increased to 200% of control, or more, for combinations of CO
       and cyanide hypoxia.  First, it is possible that the cortical cells themselves were damaged by

       March 12, 1990                         10-95     DRAFT-DO NOT QUOTE OR CITE

-------
     -I  400
     O
     o:

     t;  300

     o
     o
     „.  200-
     o

     ^   100-

     u.
     CQ       o
     o
                             • Q 1.5 ^.g/ml CN

                             ,5 1.0 pg/ml CN


                                 No CN
10    20    30    40    50


      %  COHb
    Figure 10-12. Effect of cyanide (CN) and CO hypoxia, alone and in combination, on

    cerebral blood flow. Each point represents mean + SE.  Closed circles = CO alone

    (19 animals); open circles =1.0 /tg/mL blood CN (12 animals); and open squares =

    1.5 Mg/inL blood CN (seven animals).
    Source: Pitt et al. (1979).
March 12, 1990
          10-96    DRAFT-DO NOT QUOTE OR CITE

-------
o
cc
H
z
o
o
     CM
    o
        125
        100
          75
      50
          25
f^T
                                                      No CN
                                                      1.0 pg/ml  CN

                                                      1.5 /ig/ml  CN
                      10    20     30   40    50


                           %  COHb
Figure 10-13. Effect of CN and CO hypoxia, alone and in combination, on cerebral oxygen

consumption. Each point represents mean ± SE. Closed circles = CO alone (19 animals);

open circles =1.0 /ig/mL blood CN (12 animals); and open squares = 1.5 /tg/mL blood CN

(seven animals).


Source: Pitt et al. (1979).
March 12, 1990
                             10-97    DRAFT-DO NOT QUOTE OR CITE

-------
            300
         o  250
         §200
            150
            100
        CD
        O
50
                               •  CO ALONE
                               o  CN ALONE
                               o  CN/CO
                               4  CONTROL
                      25   50   75   100  125
                    V 02  (% of  Control)
Figure 10-14. Relationship of CBF to cerebral O2 consumption (VOj) during CN and CO
hypoxia.  The mean ± SEM of CBF (percent of control) and VO2 (percent of control) is
plotted for the groups previously described. A straight line was fitted by regression for the
effects of CO or CN alone, and extended up to a CBF of 300% of control.
Source: Pitt et al. (1979).

March 12, 1990
               10-98    DRAFT-DO NOT QUOTE OR CITE

-------
       combination of cyanide and CO and were rendered incapable of utilizing the O2 and substrate
       delivered.  Because the observed physiological parameters returned to control when blood
       cyanide and COHb levels were returned to control, the cortical cell damage was temporary
       and reversible. Second, it is possible that within the brain, regional inequalities existed
 5     between the increase in CBF and the metabolic demands of the tissue.  Third, it is possible
       that in severe hypoxia, although the energy state of the brain may be normal and CBF
       increases in an apparent adequate manner, a compensatory decrease in neuronal activity is
       produced (Duffy et al., 1972). Finally, a fourth mechanism of how cyanide and CO  hypoxia
       could act to synergistically  reduce CMRO2 is based on the characteristics of the multi-enzyme
10     cellular respiratory chain (Chance et al.,  1970). Under normal metabolic conditions,
       cerebral mitochondria! respiration is zero order with respect to O2, until mitochondria! arterial
       O2 tension  is less than  1 mmHg (Chance et al., 1962). During low levels of cyanide  hypoxia,
       the abundance of cytochrome oxidase (Lubbers, 1968), the ability of unblocked respiratory
       chains to branch out and oxidize  cyanide blocked chains, and the ability of a multi-enzyme
15     system to maintain a constant electron flow for a wide range of steady-state changes in the
       terminal oxidase (Chance et al., 1970) enable the brain to maintain CMRO2 constant by
       increasing  CBF. However, further reduction in O2 supply with the addition of CO  result in a
       further reduction in the enzyme system.

20     10.4.1.6 Mechanism of Regulation of Cerebral Blood Flow in Hypoxia
            Although it is clear that hypoxia produces  cerebral vasodilation and an increase  in CBF,
       the precise mechanism by which  this occurs is not clear.  Hypotheses to explain this
       mechanism include direct effects of O2, neurogenic (which was referred to earlier in this
       chapter), and  chemical or metabolic theories.  Little evidence exists concerning the direct
25     effects of O2 on cerebral vessels,  however, there is some evidence that O2 may act directly on
       the smooth muscle of cerebral vessels, with a high O2 tension leading to vasodilation.  Garry
       (1928) showed that spirals of carotid artery of sheep contract with high O2 tension.  This
       response was  confirmed later by Smith  and Vane (1966) and Detar and Bohr (1968),  and in
       addition, Detar and Bohr (1968) showed that isolated  rabbit aortas dilated when perfused with
30     blood or saline with low O2 tension.  The dependence of the contractile response to O2 tension
       is explained if one assumes that O2 plays a metabolic role within the mitochondria of  smooth

       March 12, 1990                          10-99     DRAFT-DO NOT QUOTE OR CITE

-------
                  ft fa
                             OE fc «•• (pi ^mij|» tamaxr, 'Kittam aad Mag (lyo| fantr
                                         1Q,
                                         da
                      -      --.-  »_ ^» -        »•  - -
                     qpaon atauaiifc •> 1% »••• OGMB enat

               mmrMarteriBJB
               rfi
      to lit Ih ••iliMiim nf ii inlail i imiliHiiai imift fcjjnrii QTrtr, I"T" Tiiiimjj ITrfi)

                                                           1971).
                               i 
-------
15

                                                             1L^»^ ^u^^—^-       "•• J. __ 
-------
       increased CBF with hypercapnia has been demonstrated (Shalit et al., 1967).  Other
       investigators also have demonstrated the potential involvement of higher brain centers in the
       regulation of CBF (Langfitt and Kassell, 1968; Molnar and Szanto, 1964; Stavraky, 1936),
       and it remains possible that these central neurogenic centers may be involved in the CBF
 5     response to hypoxia.
            Carbon monoxide can compromise tissue oxygenation in three ways:  a fall in arterial O2
       content, an increase in OjHb affinity, and theoretically, a direct cellular effect (Coburn,
       1979). Arterial O2 content falls as CO occupies O2-binding sites, but when CO occupies
       binding sites, the O2 affinity of the remaining sites increases (Paulson et al., 1973; Coburn,
10     1979; Roughton and Darling, 1944).  As a result, cerebral venous O2 tension, and presumably
       tissue O2 tension, decrease.  Both  types of hypoxia, hypoxic and CO, produce essentially
       identical decreases in cerebral venous O2 tension. This reduction in cerebral venous O2
       tension could be translated into an increased CBF via some of the mechanisms previously
       described (i.e., adenosine, O2 receptors, neurogenic mechanisms, etc).  Because arterial O2
15     tension falls only with hypoxic hypoxia, one has difficulty in ascribing changes in CBF to
       alterations in arterial O2 tension.  In CO hypoxia, arterial O2 content is reduced, not O2
       tension, thus one must consider the possibility of O2 content-type receptors, which is unlikely.
       Another controversial potential effect of CO is an identifiable direct effect of CO on cellular
       metabolism.  If the CO/O2 tension ratio in the mitochondrion is sufficiently high, CO can
20     combine with cytochrome a3 (Coburn,  1979).  This would prevent oxidation at the terminal
       electron  transport chain and would be equivalent to a lack of molecular O2.  Whether this
       potential mechanism operates in vivo is unclear at the present time.

       10.4.1.7 Summary
25          The data reviewed indicate that CO hypoxia increases cerebral blood flow, even at very
       low exposure levels.  Cerebral O2 consumption is well maintained until levels of COHb reach
       upwards of 30%.  The overall responses of the cerebrovasculature are similar in the fetus,
       newborn, and adult animal; however, the mechanism of the increase in cerebral blood flow is
       still unclear. In fact, several mechanisms working  simultaneously to increase cerebral blood
30     flow appear likely and these may involve metabolic and neural aspects as well as the O2Hb
       dissociation curve, tissue  O2 levels, and even a histotoxic effect of CO. These potential

       March 12, 1990                         10-102     DRAFT-DO NOT QUOTE OR CITE

-------
        mechanisms of CO-induced alterations in the cerebral circulation need to be investigated
        further.  The interaction of CO with cyanide (additive and synergistic) on the cerebral
        vasculature is clear, however the interaction of CO with other agents and their combined
        effects on brain blood vessels is unknown. This also is true for the long-term (chronic)
 5      effects of CO alone, or in combination with other agents in low- or high-dose levels on the
        cerebral vasculature. Finally, under normal circumstances the brain can increase its blood
        flow or its O2 extraction in order to compensate for a reduced O2 environment.  Whether these
        compensatory mechanisms continue to operate successfully in a variety of conditions where
        the brain, or its vasculature are compromised (i.e., stroke, head injury, atherosclerosis,
10      hypertension) is unknown and requires further investigation.

        10.4.2  Behavioral Effects of Carbon Monoxide
        10.4.2.1  Introduction
            The following is an evaluative review of the literature concerning the behavioral and
15      nervous system effects of elevated COHb. An effort was made to organize the  findings by
        subject matter,  devoting a section of the review to each of several subtopics.  Such an
        organizational scheme is always arbitrary and therefore occasionally strained. The
        organization of the material is, however, a benefit which outweighs the disadvantages.
            Extensive use is made of tables in each subtopic to help summarize the findings and give
20      a critique of each study.  For each published report the following information is given:
        duration of CO exposure,  range of COHb achieved,  number of subjects studied (n), the
        dependent variable studied, the authors' conclusions  about effects, a comment about the
        features of the study, and technical critique notes.
            The technical critique notes refer to technique problems that frequently exist in studies
25      in the CO literature concerned with behavior. A  critique code has been devised to facilitate
        reference to two of the most common problems -  blinding and multiple statistical tests
        conducted on the same data base.  The following  paragraphs describe the problems and a table
        is given to define the codes used.
            One of the features of a study is the so-called blinding of subjects and experimenters to
30      the exposure conditions.  To avoid the effects of suggestion and expectation on  the part of the
        subject, the subject should not be informed about his own exposure condition until after the

        March 12, 1990                         10-103     DRAFT-DO NOT QUOTE OR CITE

-------
       completion of the study (i.e., the subject should be kept 'blind1 regarding exposure.  To avoid
       unintentional bias in the handling of subjects or making unintentional suggestions to the
       subjects, experimenters who deal directly with subjects also should be blind. All subjective
       scoring of data should be performed blindly. An experiment in which both subjects and
 5     experimenters were blind is called 'double blind.'  When only subjects are blind, a single-
       blind condition is said to exist.  When no blinding was used, the study will be called
       nonblind.
             The other technical aspect of an experiment that was included in the summary tables
       involves the statistical significance test methodology that was employed. When many
10     individual tests are conducted on the same data set the probability is increased that at least one
       of them will be significant by chance alone. Thus, the 'experiment-wise' Type I error rate is
       increased (Muller et al., 1984). Studies in which data were analyzed in the above manner
       will have an increased  probability of reporting a significant effect even when no effect exists
       in the population.  This bias toward significant effect worsens  with the number of tests
15     conducted. Statistical methods exist that can be used  to test multiple hypotheses for each  data
       set without increased probability of false significant effects.
             The two technical problems, blinding and statistical methods, are noted in the summary
       tables by use of the following code letters in the column labeled 'Technical Critique. 1'  If
       nothing appears in the  Technical Critique column, the experiment was conducted double-
20     blind and multiple significance testing was not done.  The following list defines the code.

             A -  No or unspecified statistical test
             B -  Multiple-significance tests on the same data set
             C -  Single-blind study
25           D -  Nonblind study

             Examination of the summary tables reveals that many of the same references appear
       repeatedly.  These were instances in which the experimenters measured several variables in
       the same study. Thus, the reader should recognize that the number of experiments reported in
30     the literature is considerably smaller than the number of summary table entries.
        March 12, 1990                          10-104     DRAFT-DO NOT QUOTE OR CITE

-------
        10.4.2.2  Sensory Effects
        Vision
             Absolute Threshold.  Studies of the effects of COHb on absolute visual threshold are
        summarized in Table 10-9.  In an experiment using four well-trained young subjects it was
 5      demonstrated that visual sensitivity was decreased in a dose-related manner by COHb levels
        of 4.5,  9.4, 15.8, and 19.7% (McFarland et al., 1944).  Various aspects of these data were
        subsequently reported by Halperin et al. (1959) and McFarland (1970). COHb elevations
        were accomplished by inhalation of boluses of high-concentration CO.  Visual thresholds were
        measured repeatedly over a five-minute period at each COHb level. Experimenters were not
 10      blind to the exposure conditions and the subjects could have easily deduced the conditions
        from the experimental design, because no air-only condition was included to control for the
        effects of the testing scheme itself.  Data from only one typical subject were presented.
        Thresholds were measured at only one level of dark adaptation (0.002 foot candles).
             The McFarland et al. (1944) study (above) stands in disagreement with several other
 15      studies of absolute threshold effects of COHb. An early study of dark adaptation was
        reported by Abramson and Heyman (1944) in which effects were inconsistent and not
        statistically significant.  Nine subjects were tested and the COHb level ranged up to 30%.
        Documentation of the study was, however, very sparse so that it was difficult to consider the
        study critically but the power was apparently quite low. McFarland (1973), in a scantly
20      documented article,  reported that similar threshold shifts occurred at the end of a CO-
        exposure period (17% COHb) and an air-only session. Thus it is possible that the effects
        reported by McFarland et al. (1944)  were due to fatigue or some other time-on-task related
        variable.  Von Restorff and Hebisch  (1988) found no dark adaptation effects on subjects with
        COHb levels ranging from 9 to 17%. Luria and McKay (1979) found no effect of 9% COHb
25      on scotopic visual threshold.
            The effect of 17% COHb (bolus administration, followed by maintenance CO level for
        135 min) on the entire dark adaptation curve was studied by Hudnell and Benignus (1989)
        using 21 young men in a double-blind study.  No difference between CO and air groups was
        observed.  A power of 0.7 was calculated for  the test employed so that the conclusions are
30      reasonably defensible. From the above evidence, it appears that if COHb elevation affects
        visual sensitivity, it remains to be demonstrated.

        March 12, 1990                         10-105     DRAFT-DO NOT QUOTE OR CITE

-------
o
tr
                                  TABLE 10-9.  EFFECTS OF COHb ON ABSOLUTE VISUAL THRESHOLD
o
o
o
o
cj
O
»
n
Exposure
Duration,
min
7
Bolus

18
Elevated
COHb Range
% n
10.0 - 30.0 9
17.0 21
+ 135
9.0 18
CO
Effect Comment
No Stimuli, methods not well specified.
No statistics given. Only one subject
at 10% COHb.
No Power for McFarland-size effect was ca.
0.7. Entire dark adaptation curve was
measured. Stimulus was 10.0 deg visual
angle p31 phosphor CRT with neutral
density filters.
No Measured only scotopic sensitivity.
Technical
Critique* Reference
A,D Abramson
and Heyman (1944)
Hudnell and
Benignus (1989)
B,C Luria and
McKay (1979)
       0.5
              4.5 - 19.7
        Various
              6.0 - 17.0
Bolus
9.0 17.0
+60
 4       Yes        No control group for air only. Only
                    typical data for one subject.  Tested only
                    at adaptation to 0.002 ft candle.  Effect
                    was COHb-ordinal, beginning at ca. 5%.

27       No         Smokers and nonsmokers tested,  n not
                    given. Few methods, specifications, or
                    statistics. Minutes of exposure (not
                    given) adjusted to target COHb.

 5       No         Dependent variables were time to
                    adaptation and sensitivity after
                    adaptation.
                                                                                     A,C
                                                                                     A.C
McFarland et al.
(1944); Halperin
et al. (1959)
McFarland (1970)

McFarland (1973)
Von Restorff
and Hebisch (1988)
"Technical problems: A=No or unspecified statistical tests; B = Multiple-significance tests on the same data; C=Single-blind study; D=Nonblind study.  If
 no technical problems are noted, the experiment was conducted under double-blind conditions and multiple-significance testing was not done.

-------
            Temporal Resolution.  The temporal resolution of the visual system has been studied in
       the form of critical flicker fusion (CFF). In the CFF paradigm, subjects report the frequency
       at which light flashes begin to appear as a continuous light.  Studies of the effects of COHb
       on CFF are summarized in Table 10-10.
 5          Seppanen et al. (1977) reported dose-ordinal decreases of CFF for COHb values of ca.
       4.0, 6.1, 8.4, 10.7, and 12.7%.  The experiment was conducted with 22 healthy subjects
       whose age ranged  from 20 to 62 yrs.  COHb was induced by breathing high concentrations of
       CO from a Douglas bag.  Subjects were blind as to the condition but apparently
       experimenters were informed.  Appropriate controls for fatigue were included and the
10     exposure levels were randomized.
            A  study was  reported by Von Post-Lingen (1964) in which COHb levels ranged up to
       23%.  COHb was  induced in 100 subjects by breathing CO-contaminated air from a
       spirometer for about seven minutes in a single-blind procedure.  One group of subjects was
       given an injection  of Evipan (sodium hexobarbitone, see Reynolds, 1982), a drug previously
15     shown to have produced decreases in CFF only if patients had demonstrable brain damage.
       In the nondrug group, CFF was unaffected until ca. 14% COHb.  In the drug group,
       however, effects began at COHb levels as low as 6% and were dose proportional up to the
       highest COHb value.  When the drug + CO study was repeated in a small (n=15)
       double-blind replication, no effects were seen. The latter replication study was given only
20     one paragraph in the report and thus  it is not clear exactly what was done.
            Beard and Grandstaff (1970) reported significant effects on CFF in an earlier study in
       which four subjects had been exposed to CO level of 50, 150, or 250 ppm for one hour.
       COHb was estimated to have reached 3.0, 5.0, and 7.5%, respectively, by the end of the
       exposure.  Documentation was extremely sparse and with only four subjects, power was
25     probably low.  Even though the elevated COHb groups had decreased CFF, the results were
       not dose ordinal.   There is a comparatively large literature published before the Seppanen
       et al. (1977) article, in which the effect of elevated COHb on CFF was tested. In none of the
       earlier studies was CFF found to be affected, even though much higher levels of COHb were
       tried.  The studies and their maximum COHb levels were Fodor and Winneke (1972) -
30     7.5%, Guest et al. (1970) - 8.9%, Lilienthal and Fugitt (1946) - 15.4%, O'Donnell et al.
       March 12, 1990                         10-107    DRAFT-DO NOT QUOTE OR CITE

-------
                                    TA!til040s             OP eOHb ON
       Ikpiure
       Buration,      6§HI Range
        Bin             %
              Comment
                                                Te§hai§al
                                                                                    Refereaee
       lelus
i
41
1
11,§«17,§



1,9 --11,7




7,1-11,1

f,0 - 11,7




7J*17J
4       Yes
II      No
                                                 No
                                  i       Ne
m
17
                                                 No
                                                           Test speiifieatioa, methods, aad statisties
                                                           net given, Iffetts disgfdinal in
                                                           estimated from ereath sample,
Few test spe@ifleatiens, m\y grand name sf
test devige,  gQHk estimated frem exposure by
eriginal authers,

Tested with red ae§8 limp, 0.7 deg visual
      viewed binaeularly, Luntinanee net
Tested with nesn lamp, 1 deg visual angle!
viewed mgneeularty;  Luminanee ml given
Minutes 3f exposure adjusted t§ target
                                                           Tested with red light, 1J deg visual angle,
                                                           viewed 6ino@ularly;  Lumiaanee not given,
                                                           Tested in neisy eavir§ament alter 69 exposure
No speeiflgations for 6FF test,

Tested with white light, IJ deg visual angle,
apparently viewed toeularlv; l
-------
•us
o
S
IS
      oe
      (XQ
     as
        £
        55
        tC
            fjP   W
          as 1Z.-S «=•£ ffi
          3K S. S 'f a *

                           in
                           os
                            ar
                            us:
                            S'.JJ
 *3
                            -
                           « gj« SB «l
                           S _
                            S "*
                           55 £ ffi
                           2 ~
                           ..an.
                               £
^^ fR •'IB"*
3i;S
Ha|8
M   Jab ueu

-------
       (1971a) - 12.7%, Ramsey (1973) -11.2%, Vollmer et al. (1946) - 17.5%, and Winneke
       (1974) - 10.0%. To be sure, there was much variation in size of the subject group, method,
       and experimental design among the above studies, but no pattern emerges as to why the
       Seppanen et al. (1977), Beard and Grandstaff (1970), and Von Post-Lingen (1964) studies
 5     found significant effects when the others did not.  It is noteworthy that the studies reporting
       significant effects were all conducted in a  single- or nonblind manner.

            Miscellaneous Visual Functions.  A number of researchers reported the results of
       experiments in which visual parameters other than absolute threshold or CFF were measured
10     as part of a battery of tests. Many of these experiments studied a large group of subjects.
       Table 10-11 summarizes these studies.
            Beard and Grandstaff (1970) reported a study in which four subjects were exposed to
       CO sufficient to produce estimated COHb levels of 3.0, 5.0, and 7.5%.  The measurements
       made were CFF (see above), brightness-difference thresholds, visual acuity, and absolute
15     threshold.  Data for the latter variable were unreliable and not reported.  Dose-related
       impairments in acuity and brightness-difference sensitivity were reported.  The scant
       documentation of methods, plus the few subjects, make the results difficult to evaluate.
            Five other reports of significant visual function effects by COHb elevation are extant.
       Two of the studies (Bender et al., 1972; and Fodor and Winneke, 1972) reported that
20     tachistoscopic pattern detection was impaired by COHb levels of 7.3 and 5.3%, respectively.
       Weir et al. (1973), Ramsey (1972), and Salvatore (1974)  reported that brightness
       discrimination was adversely affected by COHb levels of  6 to 20%.
             Tests of visual function after COHb elevation conducted by other authors have been
       uniformly nonsignificant.  Table 10-11 summarizes these  data.   Especially noteworthy are
25     studies by Hudnell and Benignus (1989) and Stewart et al. (1972) both of which found no
       acuity effects as reported by Beard and Grandstaff (1970). Brightness discrimination was
       similarly not found to be affected (Ramsey, 1973), in contradiction with the reports of others.
       The latter study is especially interesting in that it represents a failure to replicate an earlier
       study by the same author (Ramsey, 1972).  The first of the pair of studies by Ramsey was
30     conducted in a single-blind manner, the second was double blind.
        March 12, 1990                          10-110    DRAFT-DO NOT QUOTE OR CITE

-------
2 TABLE 10-11. EFFECTS OF COHb ON MISCELLANEOUS VISUAL FUNCTIONS
K
s-
i_> Exposure Elevated
1° Duration, COHb Range
i_* min %
0 60 1.8-6.7b



150 7.3°

780 5.3C


60 27.0-41.0


9
£ Bolus 17.0

Various 6.0 - 17.0
O
i£
H 45 5.0
6
^ 45 7.6-11.2
2!
s
*Q 390 5.4
O
W 30 4.0 - 12.7
Q
Dependent CO
n Variable Effect
4 Brightness Yes



42 Pattern Yes

12 Pattern Yes


5 See No
comment


21 Acuity and No
motion
27 Peripheral No
vision %

20 Brightness Yes
and depth
60 Brightness No
and depth

6 Brightness Yes
discrimination
22 Perceptual No
speed
Technical
Comment Critique* Reference
Test specification, methods, and statistics A,D Beard and Grandstaff (1970)
not given. Effects and acuity disordinal
in COHb (both variables). COHb estimated
from breath sample.
Pattern displayed for unspecified short B,C Bender et al. (1972)
time. COHb estimated from breath sample.
Pattern displayed for 0.1 s. COHb B.C Fodor and Winneke (1972)
estimated by original authors from
exposure.
Tested detection of dim objects in glare A,D Forbes et al. (1937)
and approach/recession comments of
objects. No specifications or statistics
given.
Stimulus was x x deg p 31 phosphor CRT. Hudnell and Benignus (1989)
Tested both photopic + 135 and scotopic.
Few methods, specifications, or statistics. A,C McFariand (1973)
Minutes of exposure (not given), adjusted
to target COHb.
No specification for tests. Only B,C Ramsey (1972)
brightness discrimination affected.
No specifications for tests. Results did B Ramsey (1973)
not support previous study by Ramsey
(1972).
None. D Salvatore (1974)

Test not well specified. B,C Seppanen et al. (1977)

n

-------
I
h— *
O
8
                TABLE 10-11 (cont'd). EFFECTS OF COHb ON MISCELLANEOUS VISUAL FUNCTIONS









H1 '
O
5


o

K>
?
6
0
25
o
H
0
Exposure
Duration,
min
Var up
to 1440

150-
300

5

90-120


Bolus



Elevated
COHb Range
%
Continuous
distribution

Continuous
distribution
up to 20.0
7.5 - 17.5

2.0


5.6°





n
11


27


17

15


50




Dependent CO
Variable Effect
See No
comment

Defect No
detection

Red field No
size
Brightness Yes
discrimination

See No
comment



Technical
Comment Critique. Reference
Little documentation. Tested acuity, B Stewart et al. (1970)
depth, color, and phoria using clinical
instruments.
Subject inspected small parts. Not Stewart et al. (1972)
well specified.

Tested with perimeter bar and red B,C Vollmer et al. (1946)
sample patch.
Tested intensity matching with red, green, B Weir et al. (1973)
and white.

Poor documentation. Tested target detection B Wright et al. (1973)
in "dim" light, during glare, recovery after
glare, and depth.

Technical problems: A=No or unspecified statistical tests; B= Multiple-significance tests on the same data; C= Single-blind study; D=Nonblind study. If no technical problems are noted, Ihe
experiment was conducted under double-blind conditions and multiple-significance testing was not done.
''Values of COHb were not reported
The original

by the original
authors estimated COHb from expired


authors. Values given in the table
air.

were estimated by the present author using exposure parameters and the method of Cobum et al. (1965).


 o

-------
            The most thorough modern tests of visual function was performed by Hudnell and
       Benignus (1989) who tested absolute threshold (see above), acuity, and motion detection with
       COHb levels of 17% and found no effects due to COHb.  The acuity and motion detection
       were tested at both scotopic and photopic levels.
 5          It would  appear that the results of studies of the effects of COHb elevation on
       miscellaneous visual function are not supportive of significant effects. Results that were
       significant in two studies (Beard and Grandstaff,  1970;  Weir et al., 1973) were contradicted
       by other reports using relatively large groups of subjects.  One author (Ramsey, 1972, 1973)
       failed to confirm his own findings.
10
       Audition
            Surprisingly little work has been done concerning the effects of COHb on auditory
       processes.  Table 10-12 is a summary of the studies. Stewart et al. (1970) reported that the
       audiogram of subjects exposed to as high as 12.0% COHb was not affected. Haider et al.
15     (1976) exposed subjects to a 105 dB, one-octave bandwidth  random noise (center frequency of
       2 kHz) for 15  min while COHb level was elevated to 13%.  Under continued COHb
       elevation, the temporary threshold shifts (ITS) were measured after noise cessation. No
       effects of COHb on TTS was observed. Guest et al. (1970) tested the effects of elevated
       COHb (8.9%) on auditory flutter fusion and found no significant effect. The flutter fusion
20     test is analogous to CFF in vision  and was tested by having  the subject judge the rate at which
       an interrupted  white noise became apparently continuous.  From these data  it would appear
       that  the functioning of the auditory system is not comparatively sensitive to COHb elevation,
       but little research has been done.

25     10.4.2.3 Motor and Sensorimotor Performance
       Fine Motor Skills
            Bender et al. (1972) found that manual dexterity and precision (Purdure pegboard) were
       impaired by 7% COHb. Winneke (1974) reported that hand steadiness was affected by 10%
       COHb, but no supportive statistical test was presented.
30          Similar motor functions were evaluated by a number of other investigators and found
       not to be affected, even at higher COHb levels. Table  10-13 summarizes the literature.

       March 12,  1990                         10-113     DRAFT-DO NOT QUOTE OR CITE

-------
                            TABLE 10-12.  EFFECTS OF COHb ON MISCELLANEOUS AUDITORY FUNCTIONS
Exposure Elevated
Duration, COHb Range
min %
65 8.9


H-»


-------
                                TABLE 10-13.  EFFECTS OF COHb ON FINE MOTOR SKILLS

K
o
to

v^5
10
o








>— »
O
i— »
>— k
L/l


O
>
H
i
0
0
o
H
*8
g
a
I.AJ
o
Exposure
Duration,
min
150


780



150

180

30
Varup
to 480


Varup
to 1440
up to 20.0
5


300

90 - 120

Bolus


Technical
exnerimer
Elevated
COHb Range
%
7.3"


5.3b



5.5

3.0- 12.4

4.0 - 12.7
Continuous
distribution
up to 12.0

Continuous
distribution

7.5 - 17.5


10.0b

20.0

5.6b


problems: A=No or unspecified
it was conducted under double-blii

Dependent
n Variable
42 Tapping
and
pegboard
12 Tapping,
pegboard,
and
steadiness
16 Tapping

9 Ataxia

22 Tapping
11 See
comments


27 See
comment

17 Postural
stability

18 See
comments
15 See
comment
50 Steadiness



CO
Effect
Yes


No



No

No

No
No



No


No


Yes

No

No


statistical tests; B=Multiple-significance
id conditions and multiDle-siffnificance te

Technical
Comment Critique*
Some aspect of each task declared affected. B,C
COHb estimated from breath sample.

COHb estimated by original authors from B,C
exposure.


Tested tapping with and without simultaneous C
arithmetic task.
No data above 6.6% COHb shown, only results B
of significance tests. Noisy environment.
None. B,C
Little documentation of procedure or results. B
Tested collar/pin, screw, Flanagan
coordination, tapping, and hand steadiness.

Tested collar/pin, spiral drawing, and hand
steadiness.

Tested both eyes open and closed. B,C


Tested tapping, steadiness, and Purdue hand A,C
precision. Only steadiness declared affected.
Tested tapping, star tracing, and rail walking. B

None. B


tests on the same data; C=Single-bIind study; D=Nonblind study. If no
stine was not done.


Reference
Bender et al. (1972)


Fodor and Winneke (1972)



Mihevic et al. (1983)

O'Donnell (1971b)

Seppanen et al. (1977)
Stewart et al. (1970)



Stewart et al. (1972)


Vollmer et al. (1946)


Winneke (1974)

Weir et al. (1973)

Wright et al. (1973)


technical problems are noted, the
o
      Originial authors estimated COHb from expired air.

-------
       Vollmer et al. (1946) reported that 20% COHb did not affect postural stability.  O'Donnell
       et al. (1971b) used the Pensacola Ataxia Battery to measure various aspects of locomotion and
       postural stability.  Subjects with 6.6% COHb were not affected.  Stewart et al. (1970, 1972)
       tested the ability of subjects to manipulate small parts using the Crawford collar and pin test
 5     and screw test, the AAA hand steadiness test and the Flanagan coordination test. COHb
       levels up to 15% had no effect on any of the measures. Two subjects were taken to 33% and
       40% COHb, however, and in these subjects, the collar and pin performance was impaired and
       the subjects reported hand fatigue.  Manual dexterity (Purdue Pegboard), rapid precision
       movement (Purdue hand precision), and static hand steadiness (pen in hole) and tapping tests
10     were not affected by COHb levels of ca. 5.3% (Fodor and Winneke, 1972). Wright et al.
       (1973) reported that hand steadiness was not affected by COHb levels of 5.6%.  Weir et al.
       (1973) found no effects of 14% COHb on tapping, star tracing, and rail walking.  Mihevic
       et al. (1983) discovered no effect on tapping when the task was performed alone or
       simultaneously with an arithmetic task.  Finally, Seppannen et al. (1977) demonstrated that
15     tapping speed was unaffected by  12.7% COHb. Most of the above nonsignificant studies
       used a moderate-to-large number of subjects. The overwhelming evidence in the area of fine
       motor control indicates that COHb levels below ca. 20% (the highest level tested) do not
       produce effects.

20     Reaction Time
            Table 10-14 summarizes the literature  with respect to the effects of elevated COHb on
       reaction time.  Of the 11 experiments that studied reaction time, only one reported a
       significant result (Weir et al., 1973), and that effect occurred only at 20% COHb.  A number
       of the nonsignificant effects were from studies using a large number of subjects. The
25     pervasive finding that COHb elevation does not affect reaction time is especially impressive
       because of the wide range of COHb levels employed (5.3 to 27.8%).

       Tracking
            Tracking is a special form of fine motor behavior and hand-eye coordination that
30     requires a subject to either follow a moving target or compensate for a moving target's
       motion by manipulation of a lever, for example.  The literature on tracking is summarized in

       March 12, 1990                         10-116     DRAFT-DO NOT QUOTE OR CITE

-------





1
fc
O
u
^•j
w
§
a
i
u
s
1
I— 1
6
T"H
H













Reference
11
H O





g
o
CJ
8l
w

£

c
U)
c
15
-> MM *JQ
Jj ^M Q^

«0
2 B"
| -2 B
|| B
|
JS

Fodor and Wi
0
«




1
•'§• •
° §
COHb estimated
authors from exf
o
JS o
9- •-
111

~^




CO



1

f^1 So
C*l OO
vv CTv
^* *— i
•a -a
1$ IS
1 1
£ =
q
<


—
u
2 H v. % o
o o
Z Z
u u
1 1
C/5 O

NO to

'q

o

ts >o


§ 1
ON
C-
& p
s s;
" C
^ T>
•a s
S -G
•s %
3 2
u o
« <




ca *^ ^>
u O u
•^ C W)
'I "g S
« 5 2
b o -0
a X*l
•g15 t
4> w
1 ill
< 1
JH u 8
!,-§ 1
1 S •§ u

2 R


q
i
q q
Ot tO
CO
S
2 £
                                                            R
                                                            ON

                                                                  •a

                                                                  .b
                                              O
                                              cd
                                                 CO
                  O
                  «"
                                                            a
                                                                  CQ
                                              1
              1
                                                    o —
                                                    S 8
                                                     <
CB J) >,
2 5.3
S 2 '3
.S S Sf d
•o * " 3
2 j< T3 -3
ss-s s
H H a a
                                                                  sr
                                                                  o E
                                              0   O  0
                                              2   2  Z
                                              O   U  bo
i
o
                                                            o
                                                            '

                                                            
-------
I
to
 d
 O
 O
                                             TABLE 10-14 (cont'd).   EFFECTS OF COHb ON REACTION TIME
Exposure Elevated
Duration, COHb Range
min %
300 10.0C
O
,L
°° Bolus 5.6C

CO
n Type Effect Comment
18 Simple No COHb estimated from exposure
and by original authors.
choice
50 Simple No None.

Technical
Critique* Reference
A,C Winneke (1974)


B Wright et al. (1973)
         Technical problems: A = No or unspecified statistical tests; B = Multiple-significance tests on the same data; C = Single-blind study; D = Nonblind study. If no technical problems are noted, the


          experiment was conducted under double-blind conditions and multiple-significance testing was not done.


         bOriginal authors estimated COHb from expired air.

         °COHb estimated from exposure by original authors.
 8
 n

-------
        Table 10-15.  Of the 11 studies on the topic, four reported significant effects and one of those
        found effects only at 20% COHb. The matter is more complicated, however, and the
        literature in the area offers some clues to the reasons for the diversity among the reports.
             O'Donnell et al. (1971a,b) used critical instability compensatory tracking in which the
 5      task was to keep a meter needle centered. Simultaneous performance of detection tasks also
        was required in one of the studies. No effects were demonstrated for COHb levels of
        12 to 13%. The critical instability tracking  task also was used by Gliner et al. (1983) in
        conjunction with peripheral light detection.  COHb levels up to 5.8% had no effect on
        performance.  Pursuit rotor tracking also was reported to be unaffected at 5.3%  COHB
10      (Fodor and Winneke, 1972).  Weir et al. (1973) reported that pursuit rotor performance was
        slightly affected beginning at 20% COHb.  In a 1988 study, Bunnell and Horvath used a two-
        dimensional tracking task in which the stimulus was presented on a CRT and controlled with a
        joystick. No effect of COHb or exercise or combination of the two was seen for COHb
        levels up to 10.2%.  Schaad et al. (1983) reported that pursuit and compensatory tracking
15      were not affected by COHb of 20%  even during simultaneous performance of monitoring
        tasks.
             In a pair of careful  studies of different design, Putz et al. (1976, 1979) studied
        compensatory tracking by having the subject try to keep a vertically moving spot in the center
        of an oscilloscope screen. The tracking was performed while simultaneously performing a
20      light-brightness detection task. In both studies, tracking was significantly affected by COHb
        levels of 5%.  The fact that both studies demonstrated significant results despite  differences in
        experimental design lends credibility to the finding.   Additional credibility was gained when
        the Putz et al. (1976) study was replicated with similar results by Benignus et al. (1987).
        The consistency of the compensatory tracking results in the Putz et al. (1976) protocol was
25      not continued when Benignus et al. (1989a)  attempted to demonstrate a dose-effect
        relationship using the same experimental design. In the latter study, independent groups were
        exposed to CO sufficient to produce COHb levels of control, 5, 12, and  17%  COHb. CO
        was administered via Douglas bag breathing and then COHb was maintained by low-level CO
        in room air.  A fifth group was exposed to CO  in the chamber only and this group served as a
30      positive control because it was treated in exactly the same ways as the subjects in Putz et al.
        (1976) and in Beningus et al. (1987). No significant effects were demonstrated on tracking in

        March 12,  1990                         10-119     DRAFT-DO NOT QUOTE OR CITE

-------
I
^ J
1 — »
JsJ
£
0










9
H-ft
to
O



•art
£
?
O'
O
g
s

O
d
tn
^^
s
O
TABLE 10-15. EFFECTS OF COHb ON TRACKING

Exposure Elevated
Duration, COHb Range
min %
240 8.2


Bolus + 5.6 - 17.0
240

Bolus + 7.0 - 10.0
55

780 5.3b

150 5.8


540 5.9 - 12.7




180 3.0 - 12.4


240 3.0-5.1


240 3.5 - 4.6






CO
n Type Effect
22 Compensatory Yes


74 Compensatory No


15 Compensatory No


12 Pursuit No
rotor
15 Compensatory No


4 Compensatory No




9 Compensatory No


30 Compensatory Yes


30 Compensatory Yes







Comment
CRT display with simultaneous
monitoring, same task as Putz
et al. (1976, 1979).
CRT display with simultaneous
monitoring, same task as Putz
et al. (1976, 1979).
Two-dimensional tracking test
performed for 9 min using
CRT and joystick.
COHb estimated from exposure
by original authors.
Critical instability method
using a line on a CRT screen.

Task was to keep meter needle
centered while monitoring
other meters and lights.
Tested in noisy environment.
CO exposure during sleep.
Critical instability method
using a meter needle
centering. Noisy environment.
CRT display with simultaneous
monitoring. Significant at
5.1% COHb only.
CRT display with simultaneous
monitoring. Significant at
4.6% COHb only.




Technical
Critique* Reference
Benignus et al. (1987)


Benignu. et al. (1989a)


C Bunnell and Horvath (1988)


B,C Fodor and Winneke (1972)

C Gliner et al. (1983)


B O'Donnell et al. (1971a)




B O'Donnell et al. (1971b)


Putz et al. (1976)


Putz et al. (1979)





-------
***

5
6
o
                                              TABLE 10-15 (cont'd).  EFFECTS OF COHb ON TRACKING
Exposure Elevated
Duration, COHb Range
min %
270 20.0

90 - 120 7.0 - 20.0



n
10

15-
25

CO
Type Effect
Pursuit and No
compensatory
Pursuit Yes



Comment
Light-monitoring and arithmetic
tasks performed simultaneously.
No consistent effects until
20% COHb.

Technical
Critique* Reference
B,D Schaad et at. (1983)

B Weir et al. (1973)

Technical problems:  A=No or unspecified statistical tests; B=Multiple-significance tests on the same data; C=Single-blind study; D = Nonblind study.  If no technical problems are noted, the

experiment was conducted under double-blind conditions and multiple-significance testing was not done.


 Original authors estimated COHb from expired air.
O
c!



I

O
W

r>

-------
       any group. The means for the tracking error were elevated in a nearly dose-ordinal manner
       but not to a statistically significant extent.
            At present there is no apparent reason for the lack of consistency among the reports of
       tracking performance. The largest study, with the widest dose range (Benignus et al., 1989a),
 5     appears to be the strongest indicator of no significant effects of COHb elevation.  However, it
       is difficult to ignore the several other studies which were controlled well and did demonstrate
       significant effects.  At this point the best summary seems to be that COHb elevation produces
       small decrements in tracking which are sometimes significant.  The possible reasons for such
       high variability are unclear.  Benignus et al. (1989b) discussed the issues in a speculative
10     manner.  The latter article will be reviewed later in the present document.

       10.4.2.4  Vigilance
            A dependent variable, which is possibly affected by elevated COHb, is  the performance
       of extended, low-demand tasks characterized as vigilance tasks. Because of the low-demand
15     characteristic of vigilance tasks, they are always of a single-task type.  Table 10-16 is a
       summary of the literature on the subject. Of the eight reports, four reported  significant
       effects. Despite the seemingly greater unanimity  in this area, it is noteworthy that for each
       report of significant effects, there exists a failed attempt at direct replication.
            Horvath et al. (1971) reported a significant vigilance effect at 6.6%  COHb.  A second
20     study, conducted in the same laboratory (Christensen et al., 1977), failed to find significant
       effects of 4.8% COHb on the same task. To be sure, the second study used  slightly lower
       COHb levels, but the means left no  suggestion of an effect.  Roche et al.  (1981) reported that
       performance of the same task using bolus exposure to produce 5% COHb was not affected.
            Fodor and Winneke (1972) reported a study in which 5.3% COHb  significantly impaired
25     performance of a vigilance task.  The same task and protocol were tried again in the same
       laboratory (Winneke, 1974).  In the replication attempt, no significant effects were found,
       even for COHb levels up to  10%.
            Groll-Knapp et al. (1972) reported dose-related significant effects of COHb  ranging
       from estimated values of 3 to 7.6%.  Effects were large but apparently the study was not
30     blind.  Haider et al. (1976) reported similar effects at low COHb levels but not at higher
       levels. The authors have mentioned  twice failures to replicate the results (Haider et al.,  1976;

       March 12, 1990                         10-122    DRAFT-DO NOT QUOTE OR CITE

-------
TABLE 10-16. EFFECTS OF COHb ON VIGILANCE
O
1— *
1°
1— «
O








9
NJ

O
§
H
6
O
Exposure
Duration,
min
120
780


120


210
120 - 240


135
Bolus
+60
300
Technical
Elevated
COHb Range
% n
4.8 10
5.3b 12


3.0 - 7.6b 20


6.0 - 12.0 20
3.0-13.0 20


2.3 - 6.6 15
5.0 18

10.0° 18

CO Technical
Type Effect Comment Critique* Reference
Light No None. Christensen et al. (1977)
White Yes Effect disordinal in COHb. B,C Fodor and Winneke (1972)
noise COHb estimated from exposure
by original authors.
Tone Yes Effects were significant at A,D Groll-Knapp et al. (1972)
3 % COHb and increased with
dose.
Click No None. B Groll-Knapp et al. (1978)
Tone Yes Two experiments were reported, A Haider et al. (1976)
one gave effect at ca. 7.6%
COHb and the other gave no
effect at 13%. No data
presented, only conclusions.
Light Yes No effect at 2.356 COHb. C Horvath et al. (1971)
Light No None. Roche et al. (1981)

White No COHb estimated from exposure A,C Winneke (1974)
noise by original authors.
problems: A=No or unspecified statistical tests; B=MuItiple-significance tests on the same data; C= Single-blind study; D= Nonblind study. If no technical problems are noted, the
*y experiment was conducted under double-blind conditions
A '"Original authors estimated COHb from expired air.
and multiple-significance testing was not done.
t-3 cCOHb estimated from exposure by original authors.
O
n







-------
       Groll-Knapp et al., 1978). A similar experiment using a different stimulus failed to produce
       significant effects at 12% COHb (Groll-Knapp et al., 1978).
            The fact that all replication attempts for each of the reported significant effects of COHb
       on vigilance have failed to verify the original reports is evidence for some unreliability or the
 5     operation of unknown and uncontrolled variables.  That the non-verifications were conducted
       by the original researchers, as well as by others, makes the case for unreliability even more
       binding.  If vigilance is affected by COHb elevation, a convincing demonstration remains to
       be made.  Perhaps a case may be made for small effects similar to the argument advanced in
       Benignus et al. (1989b).
10
       10.4.2.5 Miscellaneous Measures of Performance
       Continuous Performance
            Continuous performance is a category of behavior which is related to vigilance.  The
       difference is that many tasks that are performed over a long period of time are more
15     demanding and involve more than simple vigilance.  Sometimes the continuous performance
       tasks are not performed for a sufficiently long period of time to involve decrements in
       vigilance or are interrupted too frequently. Table 10-17 is a summary of the literature
       regarding the effects of COHb on continuous performance.
            Putz et al. (1976,  1979) reported that monitoring performed simultaneously with
20     tracking  was impaired at COHb values as low as 5%. In a replication attempt of the Putz
       et al. studies, Benignus et al. (1987) failed to find any effects of ca. 8% COHb.  O'Donnell
       et al. (1971a) also failed to find effects of COHb on a monitoring task performed
       simultaneously with tracking. Schaad et al. (1983) found no effects on  monitoring
       simultaneously with tracking even when COHb was 20%.   Gliner et al.  (1983)  reported that
25     signal detection was affected by 5.8% COHb when performed singly, but not when performed
       simultaneously with tracking. The latter results are in conflict with Putz et al. (1976, 1979).
             Isogna and Warren (1984) reported that the total game score on the performance of a
       multi-task video game was reduced by COHb levels of 4.2%. Separate task scores were not
       collected.  Schulte (1963) reported that letter, word, and color detection tasks were dose-
30     ordinally impaired by COHb levels of as low as 5% and ranging up to 20%. Reported COHb
        March 12, 1990                         10-124    DRAFT-DO NOT QUOTE OR CITE

-------
                       TABLE 10-17.  EFFECTS OF COHb ON CONTINUOUS PERFORMANCE
r>
sr
ts)
s
9

to
o
o
1
g
n
Exposure Elevated
Duration, COHb Range
min %
240 8.2

200 4.6 - 12.6


150 5.8




120 2.1 - 4.2


540 5.9 - 12.7


240 3.0 - 5.1




240 3.5 - 4.6





270 20.0


CO
n Type Effect
22 Light No

52 Numeric No
display

15 Light Yes




9 Complex Yes
target
detection
4 Meters No
and
lights
30 Light Yes
and
tone
tasks

30 Light Yes
and
tone
tasks


10 Light No
moni-
toring
Comment
Light monitoring simulta-
neously with tracking.
Task was to detect numerals
of unmatched parity in a
series of three.
Same task as Putz et al. (1976,
1979) but under one condition
performed without tracking.
Only the latter condition was
affected.
Video game. Targets tracked
and "shot down."

Monitoring meters and lights
while tracking.

Light monitoring simulta-
neously with tracking. Tone
monitoring as separate task.
Only light tasks affected. No
effect at 3.0% COHb.
Light monitoring simulta-
neously with tracking. Tone
monitoring as separate task.
Both light and tone task
affected. No effect at 3.5%
COHb.
Pefonned simultaneously with
tracking.

Technical
Critique* Reference
Benignus et al. (1987)

Benignus et al. (1977)


C Gliner et al. (1983)




C Insogna and Warren (1984)


B O'Donnell et al. (1971a)


Putz et al. (1976)




Putz et al. (1979)





B,D Schaad et al. (1983)



-------
o
9

to
 i
 O
 2
O
 c
                                TABLE 10-17 (cont'd).  EFFECTS OF COHb ON CONTINUOUS PERFORMANCE
Exposure Elevated
Duration, COHb Range CO
min % n Type Effect
? 0 - 20.0 49 Letter, Yes
word,
and
color
detection


Comment
COHb values larger than expected
asymptotic value. All values
affected even for low COHb.



Technical
Critique" Reference
B,C Schulte (1963)




        Technical problems: A=No or unspecified statistical tests; B=Multiple-significance tests on the same data; C=Single-blind study; D=Nonblind study.  If no technical problems are noted, the


        experiment was conducted under double-blind conditions and multiple-significance testing was not done.
 s
 n

-------
       levels were at considerable variance with values expected from the exposure parameters
       (Laties and Merigan, 1979).  Benignus et al. (1977) reported that 8.2% COHb did not effect
       a numeric monitoring task.
            Again, there is disturbing lack of replicability in the literature.  The two most credible
 5     studies showing effects of COHb on continuous performance (Putz et al., 1976, 1979) were
       not verified by Benignus et al. (1987).  In the latter study, the tracking effects of the Putz
       et al. work were verified. Similar studies of monitoring during tracking (O'Donnell et al.,
       1971a; Schaad et al., 1983; Gliner et al., 1983) also reported no effects of COHb, even with
       levels of up to 20%. It seems necessary to suspend judgment regarding the continuous
10     performance results until further data and understanding are available.  Perhaps the best
       judgment is to hypothesize small effects.

       Time Estimation
            In 1967 Beard and Wertheim reported that COHb produced a dose-related decrement in
15     single-task time estimation accuracy beginning at 2.7%. Various versions of the same task
       were tested by others (see Table 10-18) with COHb levels ranging up to 20% without effects
       being demonstrated (Stewart et al., 1970, 1972, 1973b; O'Donnell et al., 1971b; Weir and
       Rockwell, 1973, Wright and Shephard, 1978b). An exact replication was conducted by Otto
       et al. (1979) which also did not find significant results. It seems safe to assume that time
20     estimation is not affected by COHb elevation.
            Cognitive Effects. Table 10-19 is a summary of the literature concerning  the effects of
       COHb elevation on the performance of cognitive tasks. Five of the 11  experiments that have
       been reported found cognitive effects of COHb. Bender et al. (1972) reported effects of
       7.3% COHb on a variety of tasks.  Groll-Knapp et al. (1978) reported memory to be affected
25     after exposure to CO during sleep (11 % COHb), but a very similar study performed by the
       same group later found no effects of 10% COHb (Groll-Knapp et al., 1982). Arithmetic
       performance was affected slightly in a nondose-ordinal  manner when a simultaneous tapping
       task was performed (Mihevic et al., 1983).  Schulte (1963) reported a dose-ordinal effect on
       arithmetic performance beginning at 5% and ranging to 20% COHb.  COHb levels in the
30     latter study were at considerable variance with expected values from the exposure parameters
       (Laties and Merigan, 1979).  Similar variables were tested by others, sometimes at higher

       March 12, 1990                         10-127     DRAFT-DO NOT QUOTE OR CITE

-------
g
1
-p
t— »
VO
o









o
t— k
to
oo

O
^
^
1
0
0
2J
9
O
a
9
w
0
TABLE 10-18. EFFECTS OF COHb ON TIME ESTIMATION
Exposure Elevated
Duration, COHb Range
min %
120 2.7 - 12.5b

180 3.0 - 12.4


120 3.7 - 7.8

Var up Continuous
to 1440 distribu-
tion up
to 12.0
150-300 Continuous
distribu-
tion up
to 20.0
90 20.0


2 2.0 - 8.0°




n
18

9


13

11



27



15


13



CO
Type Effect
Duration Yes
discrimination
Duration No
and lime

Duration No
discrimination
Duration No
discrimination


See No
comment


Duration No
estimation

Duration No
discrimination

"Technical problems: A=No or unspecified statistical tests; B= Multiple-significance tests on

Technical
Comment Critique*
Effects were COHb ordinal B,C
beginning at ca. 2.7% COHb.
Tone duration discrimination B
and time interval estimation.
Noisy environment.
Replication of Beard and
Wertheim (1967).
Tone duration compared to B
light duration.


Used duration discrimination,
time estimation, and Marquette
test.

Various tone duration judg- B
ments were used.

Results of three experiments. D


the same data; C = Single-blind study; D=Nonblind study.


Reference
Beard and Wertheim (1967)

O'Donnell et al. (1971b)


Otto et al. (1979)

Stewart et al. (1970)



Stewart et al. (1972, 1973b)



Weir et al. (1973)


Wright and Shephard (1978b)


If no technical problems are noted, the
experiment was conducted under double-blind conditions and multiple-significance testing was not done.
bCOHb values were not reported by the
°Original authors estimated COHb from
originial authors.
expired air.
Values given in the table were estimated by the present authors using exposure parameters and


the method of Coburn et al. (1965).

O

-------
                    TABLE 10-19. EFFECTS OF COHb ON MISCELLANEOUS COGNITIVE TASKS
i— » Exposure Elevated
J° Duration, COHb Range
h— > min %
NT5
VO v.
0 150 - 300 7.3b



>55 7.0 - 10.0




210 6.0 - 12.0


O 410 11. Oc
t— »
to
480 10.0


ffl 120-240 3.0-13.0
h-^
i
i
o 15° 5-5
.,


Q 540 5.9 - 12.7
O
H
tn 270 20.0
0
CO
n Task Effect

42 Digit span, Yes
nonsense syllables,
intelligence test

15 Memory, stroop Yes
test, visual
search, and
arithmetic

20 Arithmetic, No
nonsense syllables,
mood scale
10 Memory, mood Yes


20 Verbal learning No
and memory

20 Attention, memory, No
arithmetic.


16 Arithmetic Yes



4 Arithmetic No


10 Arithmetic No

Comment

Some aspect of each
declared affected. COHb
estimated from breath
sample.
Stroop test affected in
nondose-ordinal manner.
Visual search affected
interactively by CO and
exercise.
None.


Only memory declared
affected. Exposure
during sleep.
Tested before and after
exposure. Was exposed
during sleep.
None.



Effect was present only
during multiple-task
performance and was not
dose ordinal.
Tested in noisy
environment.

Performed simultaneously
with tracking.
Technical
Critique" Reference

B,C Bender et al. (1972)



C Bunnell and Horvath (1988)




B Groll-Knapp et al. (1978)


B Groll-Knapp et al. (1978)


B Groll-Knapp et al. (1982)


A Haider et al. (1976)



C Mihevic et al. (1983)



B O'Donnell et al. (1971a)


B,D Shaad et al. (1983)

n

-------
                           TABLE 10-19 (cont'd).  EFFECTS OF COHb ON MISCELLANEOUS COGNITIVE TASKS
o

o
O
O
2,
Exposure Elevated
Duration, COHb Range
min %
? 0.0 - 20.0



150 - 300 Continuous




n Task
49 Arithmetic



27 Arithmetic
distribution
up to 20.0

CO
Effect Comment
Yes COHb values larger than
expected asymptotic value.
Significant effects even
at low COHb.
No None.



Technical
Critique" Reference
B,C Schulte (1963)



Stewart et al. (1972)


"A=No or unspecified statistical tests; B=Multiple-significance tests on the same data; C=Single-blind study; D=Nonblind study.
bOriginal authors estimated COHb from expired air.
"Original authors estimated COHb via Coburn, Forster, and Kane equation.
1
g
n

-------
        levels of COHb and with relatively large groups of subjects, without finding effects
        (O'Donnell et al., 1971a; Stewart et al., 1972; Haider et al., 1976; GroU-Knapp et al., 1978;
        Schaad et al., 1983).  The conclusions are, at best, equivocal.
             A recent study by Bunnell and Horvath (1988) utilized a wide range of cognitive effects
 5      involving short-term memory, manikin rotation,  stroop word-color tests, visual search, and
        arithmetic problems (the latter as part of a divided attention task performed simultaneously
        with tracking). COHb was formed by bag breathing followed by a CO level in room air
        designed to maintain a constant COHb level. Subjects were exercised at either 0, 35, or 60%
        of VO2 max before cognitive tests were performed.  The stroop test performance was slightly
10      but significantly affected by either 7 or 10% COHb by the same amount but exercise had no
        effect.  The authors suggested that negative transfer effects  (difficultly in reversing
        instructional sets) were responsible for the decrement.  Visual searching improved for both
        COHb levels at rest and at medium exercise. At the high exercise level, however, COHb
        produced dose-ordinal impairments in performance.  The authors conjectured that hypoxic
15      depression of cortical function interacted with hypoxic stress and exercise stress to produce
        the effects.
             Most of the data on cognitive effects of COHb elevation are not sufficiently consistent to
        consider. The study by Bunnell and Horvath (1988), however, is suggestive of potentially
        important effects of interactions of COHb and exercise.  Before any conclusions may be
20      drawn about the results, the study must be expanded to unravel the mechanisms by which the
        interactions were produced.

        10.4.2.6  Automobile Driving
             Complex behavior, in the form of automobile driving, has been tested a number of times
25      for effects of COHb elevation.  Not only is automobile driving potentially more sensitive to
        disruption because of its complexity, but it is an inherently interesting variable because of its
        direct applicability to nonlaboratory situations.  The well-practiced nature of the behavior, on
        the other hand, may make performances more resistant to disruption.  The complexity of the
        behavior also leads to methodological difficulties.  To exhaustively measure the complex
30      behaviors usually  leads investigators to measure many dependent variables. Statistically
       March 12, 1990                         10-131     DRAFT-DO NOT QUOTE OR CITE

-------
       analyzing a large number of variables in a defensible way requires many subjects and leads to
       greater expense.
            Table 10-20 is a summary of the studies of automobile driving as affected by COHb.  In
       an early study by Forbes et al. (1937), using only five subjects, steering accuracy in a
 5     simulator was investigated with COHb levels of up to 27.8%. No effects were demonstrated.
       A sparsely documented experiment by Wright et al. (1973), using 50 subjects with 5.6%
       COHb, tested a number of functions of simulator performance but found no effects.  Weir
       et al. (1973) performed  an experiment with actual automobile driving on a highway in  which
       a great number of variables were measured and tested.  None of the variables were reliably
10     affected until COHb exceeded ca. 20%.  Wright and Shephard (1978a) failed to find effects
       of 7% COHb on driving.  In the latter study, the authors reported effects but only after
       misapplication of the chi-square test.  The only effect of COHb on driving at a lower level
       (7.6%) was reported by Rummo and Sarlanis (1974), who found that the ability to follow
       another car at a fixed distance was impaired.
15          The difference between the experiments of Rummo and Sarlanis (1974) and Weir et al.
       (1973) is troubling.  Both measured following distance but only the experiment employing the
       lower-level COHb found effects. If automobile driving is affected by COHb elevation, it
       remains to be demonstrated in a conclusive manner.

20     10.4.2.7 Brain Electrical Activity
            Electrical activity  of the brain (see review by Benignus, 1984) offers the possibility of
       testing the effects of COHb without the problem of selecting the most sensitive behavioral
       dependent variable.  It is less dependent upon  subject cooperation and effort and may be a
       more general screening  method. The major disadvantage of the measures is the lack of
25     functional interpretability.  The area has been  plagued with poor quantification and,
       frequently, a lack of statistical significance testing.
            The electroencephalogram (EEG) is a recording of the continuous voltage fluctuations
       emitted by the intact brain. The slow-evoked  potential originally called the contingent
       negative variation (CNV) is computed by averaging over trials and was  linked to (among
30     other things) cognitive processes or expectancy (Donchin et al., 1977).  The evoked potential
       (EP) is the electrical activity in the brain resulting from sensory stimulation, either auditory

       March 12, 1990                         10-132     DRAFT-DO NOT QUOTE OR CITE

-------
tr
to
                                 TABLE 10-20.   EFFECTS  OF COHb ON AUTOMOBILE DRIVING TASKS
8
Exposure
Duration,
  min
 Elevated
COHb Range
                                                              Task
CO
Effect
                                                                                                Comment
Technical
Critique*
                                                                                                                                                       Reference
         60
         20
                           27.0-41.0
                           7.6
                                                           Steering accuracy   No
                                                           Steering wheel
                                                           reversals and
                                                           following
                                                           distance
                                                                             Yes
                                                                                  In auto simulator
                                                                                  (unspecified).

                                                                                  Only following distance
                                                                                  was affected.  Tested
                                                                                  in auto simulator.
                                                                                                        A,D
                                                                                                        B,C
                                                                      Forbes et al. (1937)
                                                                      Rummo and Sarlanis (1974)
9
UJ
O
O
         90 - 120
         Bolus
         Bolus
                           7.0 - 20.0
                  5.6b
                           7.0
                                                 12
                                                 50
                                                 10
                                                           See comments
                                                           See comment
                                                  Brake-reaction
                                                  time
                                                                             Yes
                                                                             No
                                                                              No
                                                                Instrumented automobile
                                                                driven on highway. Many
                                                                measures of control
                                                                functions plus measures
                                                                of driving stability and
                                                                information processing
                                                                load were utilized.
                                                                The latter was affected
                                                                at 20% but  not reliably
                                                                below that.  Authors
                                                                debated functional
                                                                importance  of findings.

                                                                Poor documentation.
                                                                Tested automobile
                                                                simulator performance-
                                                                brake, accelerator,
                                                                steering, and signals.

                                                                None.
                                                                                                                                                   Weir et al. (1973)
                                                                                                                                                   Wright etal.  (1973)
                                                                                                                                                    Wright and Shephard (1978a)
s
n
"Technical problems: A=No or unspecified statistical tests; B=MultipIe-significance tests on the same data; C=Single-blind study; D=Nonblind study. If no technical problems are noted, the
 experiment was conducted under double-blind conditions and multiple-significance testing was not done.
''Original authors estimated COHb from expired air.

-------
       (AEP) or visual (VEP).  EEC, CNV, and EPs have been studied with COHb elevation.
       Table 10-21 is a summary of results from these studies.
            Groll-Knapp et al. (1972) reported that the CNV was decreased in amplitude in a dose-
       related manner for COHb values ranging from 3 to 7.6%. In a second study, Groll-Knapp
 5     et al.  (1978) again reported that CNV amplitude was reduced by  12% COHb when subjects
       missed a signal in a vigilance task. More evidence is required before the functional
       significance of such an effect can be deduced but it is a potentially important finding.
            Clinical EEGs were analyzed by visual inspection by Stewart et al. (1970, 1973a) and
       Hosko (1970) after exposure to sufficient CO to produce COHb levels ranging up to 33%.
10     No effects were noticed. Groll-Knapp et al. (1978) reported similar results using spectrum
       analysis on EEGs from  subjects with 12% COHb.  Haider et al. (1976) reported slight
       changes in the EEG spectrum for COHb levels of 13%,  but no tests of significance were
       conducted.  In view of the above studies, it seems  reasonable to assume that no EEG effects
       of COHb levels below at least 10% should be expected.
15          O'Donnell et al. (1971b) reported that sleep stages (as determined from the EEG) were
       not distributed by COHb levels up to 12.4%.  Groll-Knapp et al. (1978) and Haider et al.
       (1976),  however,  both reported distributed sleep stages at similar COHb levels using EEG
       spectra. Groll-Knapp et al. (1982) repeated their earlier study and found essentially the same
       effects.
20          The VEP was consistently not affected by COHb elevation below ca. 22% and usually
       the lowest level for effects was higher (see Table 10-21). At higher levels the effects were
       dose-related (Dyer and Annau, 1977; Stewart et al.,  1970; Hosko, 1970). The above is true
       for both rats and man.
            A single study of visual electrophysiology has reported low-level effects of COHb
25     (Ingenito and Durlacher, 1979).  The electroretinogram  (ERG) of anesthetized cats was
       reported to have exhibited a reduced /3-wave amplitude beginning at 7.5%.   Effects were
       dose-ordinal up to 42% COHb.  The contribution of the anesthesia (chloralose) to the effect
       as a possible potentiator of COHb effects was not tested in the study. The authors reported
       that the effect outlasted the COHb elevation and possibly was due to direct cellular CO
30     toxicity.
        March 12, 1990                         10-134    DRAFT-DO NOT QUOTE OR CITE

-------
TABLE 10-21. EFFECTS OF COHb ON BRAIN ELECTRICAL ACTIVITY
t— > Exposure
* Duration,
•— » min
vn
O 120


Injection


120

210


o

8
410

O
^
H 480
6
o
^ 420
g
H
O
§ 21°
H
W
0
W 120
n
M 	
Elevated
COHb Range Dependent CO
% n Variable Species Effect
6.0-55.0 15 VEP Rat Yes


10.0-75.0 10 Tone AEP Rat Yes


3.0 - 17.6b 20 CNV Human Yes

6.0 - 12.0 20 VEP, Human Yes
click
AEP,
CNV, and
EEG
spectra
11.0 10 Click, Human Yes
AEP, and
EEG sleep
stages
10.0 20 EEG sleep Human Yes
stages

12.0 20 Sleep Human Yes
stages
and EEG
spectra
5.3 55 VEP Human No



7.5 - 42.0 6 ERG Cat Yes

Comment
First significant effect increased amplitude
at 22% COHb in cortex (at 38% in superior
colliculus). Changes were dose related.
COHb-ordinal effects beginning at ca. 45%
in rat. Small effect possible near 25% when
CO was injected ip.
Significant differences at all COHb levels
above endogenous.
CNV only declared affected. No data
given, only conclusions stated.




Both affected.



Same conclusions as Groll-Knapp et al.
(1978)

Both changed slightly (no significance
test).


Both young (n = 33) and elderly (n = 22)
were tested. Mean age for young: 22.8, and
for old: 68.7.

Decreased /?-wave amplitude. Dose-related
effect beginning at 7.5% COHb.
Technical
Critique* Reference
Dyer and Annau (1977)


Fechter et al. (1987)


A,D Groll-Knapp et al. (1972)

B Groll-Knapp et al. (1978)





B Groll-Knapp et al. (1978)



B Groll-Knapp et al. (1982)


A Haider et al. (1976)



Harbin et al. (1988)



Ingenito and Durlacher (1979)


-------
TABLE 10-21 (cont'd). EFFECTS OF COHb ON BRAIN ELECTRICAL ACTIVITY
o
1— >
1°
I— I
o



0
0
6
0
1
o
1
o
n
tn
Exposure Elevated
Duration, COHb Range
min % n
18 9.0 18
180 3.0-12.4 9
Var up Continuous 1 1
to 1440 distribution
up to 33.0
Var 3.2 - 15.2 6
240 3.0-5.1 30
120 7.0 - 62.0C 6
Dependent CO
Variable Species Effect
VEP Human No
Sleep Human No
stages
VEP and Human Yes
EEC
VEP and Human No
EEC
Tone AEP Human Yes
VEP Rat Yes
Technical
Comment Critique* Reference
None. B,C Luna and McKay (1974)
No data above 6.6% COHb shown only results B O'Donnell et al. (1971b)
of signficance tests. Noisy environment.
No effects until COHb ca.2\%. Only 2 subjects B Stewart et al. (1970); Hosko (1970)
were tested in the high range. Effect was an
increased amplitude of peaks Nl, P2, and N2.
No statistical tests.
None. B,D Stewart et al. (1973a)
p-p Amplitude of N1-P1 peak increased in COHb- Putz et al. (1976)
ordinal manner beginning at 3%.
Increased amplitude at both cortex and superior Xintaras et al. (1966)
colliculus at 62% COHb. Late component amplitude
decreased at 7% COHb in superior colliculus. No
statistics, only typical data given.
Technical problems: A=No or unspecified statistical tests; B=Multiple-significance tests on the same data; C= Single-blind study; D=Nonblind study. If no technical problems are noted, the
experiment was conducted under double-blind conditions and multiple-significance testing was not done.
bCOHb was estimated by comparison to a separate series of animals.
°COHb was estimated from exposure by published data (Montgomery and Rubin, 1971) by present author.














i




-------
             Groll-Knapp et al. (1978) found no effect of COHb (8.6%) on auditory (click)-evoked
        potentials during waking, but reported increased positive-peak amplitudes when subjects were
        tested during sleep at ca. 11 % COHb. The finding was verified by Groll-Knapp et al.
        (1982).  The fact that the data were collected during sleep is potentially important.
 5           Putz et al. (1976) conducted a double-blind study in which 30 persons were exposed to
        70 ppm CO for 240 min (5% COHb at the end of the session).  Among other variables, the
        auditory-evoked potential was measured.  The peak-to-peak amplitude of the N1-P1
        components was increased in a dose-ordinal manner beginning at ca. 3% COHb.
             Ten millisecond tone bursts were used by Fechter et al. (1987) to produce AEPs in rats
 10      exposed to graded doses of injected CO.  COHb levels ranged up to 75%. Slight evaluations
        of the mean threshold began at about 25% COHb. Effects were first seen at higher
        frequencies.  All effects were reversible.  During the exposure,  normal body temperature was
        maintained to avoid hypothermia (Annau  and Dyer, 1977).
             Many of the brain electrical activity measures seem to be altered by COHb elevation.
 15      The functional significance of these changes is not clear.  Sometimes an alteration is not an
        indication of a deleterious effect but merely implies some change in processing. When
        induced by low levels of COHb, however, any change should be viewed as potentially
        serious.

20      10.4.2.8  Schedule-Controlled Behavior
             Because of the high levels of COHb that can be employed in studies of laboratory
        animals,  using schedule-controlled behavior,  effects of COHb are reported in all articles on
        the subject.  Table 10-22 is a summary of the literature. There are a number of problems
        with the published literature,  however, as seen in Table 10-22.  Only a few investigators
25      measured COHb;  instead, they simply specified the exposure parameters. Another problem is
        that of hypothermia, which occurs in rats  when COHb levels rise (Annau and Dyer, 1977;
        Mullin and Krivanek, 1982).  If hypothermia develops as a consequence of COHb elevation,
        behavioral effects may be secondary to the hypothermia, not the COHb directly.  None of the
        experimenters attempted to control for hypothermia effects. Thus behavioral effects of COHb
30      may be overestimated in the rat.
       March 12, 1990                         10-137     DRAFT-DO NOT QUOTE OR CITE

-------
                   TABLE 10-22. EFFECTS OF COHb ON SCHEDULE-CONTROLLED BEHAVIOR
\^ Exposure
M" Duration,
O min
P 12°
\o 1>44°
£
75


90

48


Injection

9 240
Co
oo


a
» 30
p>
H
a 90
o
S90

0
O Variable
3
o
^ «/-T>UK ..,<.<,
Elevated
COHb Range
%
9.0 - 58.0"
9.0 - 58.8'

35.0-55.0'


8.0 - 54.0«

15.0-55.0"


9.0 - 58.0

12.2 - 54.9





Continuous
distribution
up to 32.0
34.0 - 53.0

40.0 - 66.0


15.0 - 40.01



CO
n Scheds. Species Effect
5 CRF Rat Yes
8 Body Rat Yes
weight
15 MULTcombi- Rat Yes
nations of
FI3 and FR30
4 DRL21 Rat Yes

? FB, FR25 Rat Yes
VE5, VR15
VR25, DRL
22 CRF brain Rat Yes
stimulation
6 Behavioral Rat Yes
screen




3 Appetitive Monkey Yes
shuttling

3 MULT FR 30, Rat Yes
DRL 18
3 Multiple Rat Yes
sequential
responses
4 FCN Rat Yes




Comment
Rates fell inversely at COHb beginning at ca.20%
Weight fell inversely at COHb beginning at 22%. Food and
water consumption also fell.
Rates fell inversely at COHb beginning between 32 and
48% for schedules.

Rates fell inversely at COHb beginning at ca. 37%.
Temporal discrimination was undisturbed.
Effects were COHb ordinal beginning at ca. 14%. DRL was
affected at COHb < 1 %. Methods were poorly described

CO was injected ip. Effects were first noted near 45%
COHb.
Behavioral screen included reflexes, grasping, and
conditioned avoidance. Lowest level effect was on
conditioned avoidance at 12.2%. COHb levels measured
by authors were much lower than predicted from data of
Montgomery and Rubin (1971).

Shuttling velocity decreased as COHb beginning at 16 - 22%.


Rates fell beginning ca. 45% COHb.

Rats repeated releamed response chain after extinction.
More time to relearn was required beginning ca. 50% COHb.

Rates fell inversely at COHb beginning between 20 and 28%




Reference
Annau (1975)
Annau (1975)

Ator (1982)


Ator et al.
(1976)
Beard and
Wcrtheim (1967)

Fountain et al.
(1986)
Mullin and
Krivanek (1982)




Purser and
Berrill (1983)

Schrot and
Thomas (1986)
Schrot et al.
(1984)

Smith et al.
(1976a)

O

-------
            The level of COHb may be estimated for studies in which it was not given by use of the
       data from Montgomery and Rubin (1971). The latter-published normative curves can be used
       for such estimates.  Schrot and Thomas (1986) and Schrot et al. (1984) have published
       corroborating curves.  For all rat studies in which COHb was not measured, the present
 5     reviewer estimated the levels from exposure parameters.
            With one exception (Mullin and Krivanek, 1982),  effects of COHb did not occur on
       schedule-controlled behavior until COHb exceeded ca. 20%. In some studies no effect was
       observed until even higher levels.  It is possible, however, that a number of the studies were
       insensitive because of the small numbers of subjects employed.  In the study by Mullin and
10     Krivanek (1982), it was reported that conditioned-avoidance behavior was affected at COHb
       levels as low as 12.2%. The COHb level reported in the latter study is, however, about half
       of the value that would be estimated from the Montgomery and Rubin (1971) data.  It seems
       likely that either exposure or COHb values were erroneous in the report.  If the exposure data
       were correct, the effects threshold would fit the other data in the literature.  It thus appears
15     that COHb does not affect schedule-controlled data in laboratory animals until levels exceed
       20%.
            When there were frank effects on schedule-controlled behavior, they seemed all to be in
       the direction of a slowing of rate or speed of response.  Schedule-produced patterns of
       behavior were not disrupted, in general.  Thus,  it appears that the effect of elevated COHb is
20     on some general aspect of behavioral control having to do with the rate of processing.

       10.4.2.9 Summary and Discussion of Behavioral Literature
            The literature regarding the effects of COHb on behavior, as seen in the above review,
       does not allow clear-cut conclusions.  Results of studies frequently were not replicable or
25     were not supported by related studies.  In  this section an attempt will be made to discover
       what, if any, general conclusions can be made.

       Analysis of Technical Problems
            It  is possible that the many technical problems that were noted in the summary tables in
30     the 'technical critique' column may account for some of the lack of agreement among experi-
       mental results.  In the following analysis, tabulations were made of all of the studies in which

       March 12, 1990                        10-139     DRAFT-DO NOT QUOTE OR CITE

-------
       either blinding or statistical analysis problems occurred. Studies were cast into 2x2 tables
       according to the presence or absence of a particular condition and according to the occurrence
       (or not) of a COHb effect.  If multiple-dependent variables were measured in a particular
       experiment, that experiment was tallied as having reported a significant effect if any one or
 5     more of the variables was reported as affected.  A study was tallied as having reported a
       significant effect only if the effect occurred below 10% COHb to avoid the inclusion of frank
       effects that will occur if COHb is  made sufficiently high.  All studies were included,
       regardless of what dependent variables were studied.  Only human studies were included. To
       decide whether the technical problem in question can be inferred to have influenced the
10     results, a  Fisher's exact test was conducted on each table.  Two such tests were conducted;
       therefore, the alpha level selected  for each test was 0.05/2 or 0.025.  Given signficant results,
       exploratory tests also were conducted.
            Table 10-23 is a tabulation of studies according to their blinding practices.  Non-blind
       and single-blind studies were pooled because of the few non-blind studies.  The Fisher test
15     yielded significant results, p = 0.0154.  It is impressive that the rate of reported COHb
       effects was about 2.24 times as high for studies not using a double-blind design.  When the
       five non-blind studies were dropped and the data reanalyzed in an exploratory manner, the
       rate of finding significant effects for single-blind studies was 2.275 times as high as  for
       double-blind  studies (p = 0.0172).
20

                          TABLE 10-23. EFFECT OF BLIND CONDITIONS

25                                     Non-Double Blind                     Double Blind
       Effects                               16                                   6
       No Effects                            9                                   15
30
        Fisher's exact test, p = 0.0154.

35


        March 12, 1990                          10-140    DRAFT-DO NOT QUOTE OR CITE

-------
             Table 10-24 is a tabulation of studies according to the employment of multiple-
        significance tests on the same data set.  The Fisher test was not significant (p = 0.221).
 5                   TABLE 10-24.  EFFECT OF STATISTICAL METHODOLOGY

                                           Multiple-                           Conservative-
                                          Significance                         Significance
 10      	Test Methods	Test Methods
        Effects                                 13                                   9
        No Effects                              13                                  11
 15
        Fisher's exact test, p = 0.221.
20           From the above analyses it may be concluded that studies that were not conducted in a
        double-blind manner tend to demonstrate more apparent COHb effects.  It may be argued that
        the bias thus introduced into the study is added to whatever COHb effect may be present.  No
        evidence was demonstrated, however, for the hypothesis that multiple statistical tests tended to
        produce an inflated Type I experiment-wise error rate.
25
        Evaluation of the Literature
             In the evaluation of the literature on the behavioral effects of COHb it is not clear how
        to treat the results from studies not conducted in a double-blind manner.  Although they are
        biased toward reporting COHb effects, they clearly contain useful information. In the
30      following summary, results from double-blind studies will be tabulated separately from results
        of studies not conducted in a double-blind manner.
             Table 10-25 summarizes literature on  the behavioral effects of COHb. Both double-
        blind and non-double-blind studies are tabulated. For each family of dependent variables, the
        table gives the number of studies in the  double-blind and non-double-blind categories.
35      Finally, the proportion of studies reporting  a COHb effect, p(E), is given for both double-
        blind and non-double-blind studies. Several conclusions may be drawn from Table 10-25.

        March 12, 1990                          10-141    DRAFT-DO NOT QUOTE OR CITE

-------
                       TABLE 10-25. PROBABILITY OF EFFECTS OF COHbm
10
15
20
25



Dependent Variable
Absolute visual threshold
Critical flicker fusion
Misc. visual functions
Misc. auditory functions
Fine motor skills
Reaction time
Tracking
Vigilance
Continuous performance
Time estimation
Misc. cognitive function
Automobile driving
Brain electrical activity
Non-
Double
Blind
(n)
4
7
9
0
6
7
4
4
4
2
5
3
3

Double
Blind
(n)
1
3
5
3
4
5
7
4
5
4
5
2
6
Non-
Double
Blind,
P(E)
.20
.33
.55
N/A
.33
.00
.00
.75
.75
.50
.80
.33
.33

Double
Blind
P(E)
.00
.00
.00
.00
.00
.00
.43
.25
.40
.00
.00
.00
.50
       'Based on numbers of studies in each category.
        Those conclusions are as follows:
30
35
40
A.    Non-double-blind studies produce a greater proportion of reported effects, even
      for the individual dependent variables.  This is true with the exception of tracking
      and brain electrical activity. This observation supports the previous inference that
      non-double-blind studies are biased toward finding more effects than justified.
B.    Studies using double-blind procedures found effects on four of the 14 dependent
      variable families: tracking, vigilance, continuous performance, and brain
      electrical activity.
C.    In the five dependent variable categories where COHb was found to affect
      behavior, usually less than half of the reported studies found effects in double-
      blind studies.
D.    Sensory and cognitive effects were not found to be affected by COHb in double-
      blind studies.
        March 12, 1990
                                   10-142     DRAFT-DO NOT QUOTE OR CITE

-------
             E.    In most instances, the rate of finding COHb effects in non-double-blind studies
                   was high in the same dependent variables where the double-blind studies reported
                   effects.  This observation lends further support to the findings in the double-
                   blind studies.
             Continuous Performance.  It may be argued that the dependent variables of tracking,
        vigilance, and continuous performance are related functionally.  Each of the dependent
        variables in the three categories require the continuous performance of some sort of behavior.
10      The response rate and/or attention demand in vigilance behavior is low compared to the other
        two groups, otherwise, the behaviors are similar. Tracking is clearly a particular form of
        continuous performance which was categorized separately simply because of the homogeneity
        of a group of studies that existed in the literature.  It also may be argued that, despite the
        high response rates in the tracking and continuous performance  studies, both of these
15      behaviors require a strong component of sustained attention.  It  seems fair to conjecture,
        therefore, that behaviors that require sustained attention and/or sustained performance are
        most sensitive to disruption by COHb.
             The group of studies of tracking, vigilance, and continuous performance offer the most
        consistent and defensible evidence of COHb effects on behavior. The results across studies
20      is, however, far from consistent. Further examination of the three areas seems appropriate.
             Compensatory tracking was studied by two groups of investigators using virtually
        identical task parameters and equipment (Putz et al.,  1976,  1979; Benignus  et al., 1987,
        1989a).  Both of the studies by Putz et al. (1976, 1979) found significant and  moderately
        large effects of 5% COHb.  Benignus et al. (1987) reported similar but smaller significant
25      effects in a nearly identical experiment to Putz et al.  (1976).  However, in a dose-effects
        study including another direct replication group, Benignus et al. (1989a) found no significant
        effects, even for COHb levels of 17%. In the latter study, the means were  nearly dose
        ordinal but too small to be statistically significant.  It is particularly puzzling why the latter
        study, using a large number of subjects on an identical task, should find no  significant effects
30      for even 17% COHb when three other studies found effects at lower levels.  Three other
       March 12, 1990                          10-143     DRAFT-DO NOT QUOTE OR CITE

-------
        double-blind tracking studies of various methods found no effects of COHb levels of 12% or
        greater.
            As discussed in the above literature review, there is a similar disunity among studies on
        the effects of COHb on vigilance.  Because of the many failed attempts at direct replication,
 5      the conclusions seem weaker than for tracking.
            Of the five double-blind experiments in which continuous performance was measured,
        three were mentioned earlier in the discussion  of tracking. In these studies (direct
        replications), continuous performance was measured simultaneously with tracking (Putz et al.,
        1976,  1979; Benignus et al., 1987).  The latter of the three found no effects.  A small study
10      reported continuous performance effects that were disordinal in COHb (O'Donnel et al.,
        1971a).  The remaining study (Benignus et al., 1977) used a different task and obtained no
        COHb effects.

            Multiple Performance.  It is possible that COHb impairs task performance more when
15      multiple tasks are performed simultaneously, thus decreasing the amount of behavioral reserve
        capacity. To test this exploratory hypothesis, behavioral studies (sensory and
        electrophysiological excluded) were tabulated according to single/multiple task performance
        and significant/nonsignificant effects of COHb. If any given study used both single-task and
        multiple-task behaviors, the study was tabulated according to the multiple-task results only.
20      For studies in which only single tasks were required, the study was classed as having  shown
        significant results if one or more variables were significant, otherwise as nonsignificant.  The
        above rules prevented any study from being tabulated more than once.
            Table 10-26 is  the result of the above tabulation.  The exploratory Fisher's exact test
        yielded p = 0.081.  While the result would have been nonsignificant by a priori  rules, the
25      table shows a slight tendency toward more multitask studies showing  a significant effect,
        whereas more of the single-task studies found  no effect.
            The question of multitask vs. single-task  performance sensitivity to COHb disruption
        would be answered best by experiments in which the tasks were performed singly and
        together within the same study. In the literature there were only two cases in which this was
30      done.  In both of these, one of the single tasks alone was affected more than multitask
        performance (Bender et al., 1972; Gliner et al., 1983).  Experiments designed to specifically

        March 12, 1990                          10-144     DRAFT-DO NOT QUOTE OR CITE

-------
10
                        TABLE 10-26. EFFECT OF SINGLE VS. MULTIPLE
                                     TASK PERFORMANCE
         Single Task                      Multiple Task
       Effects                               7
       No Effects                           14
       Fisher's exact test, p = 0.081.

15
       answer the question of single-task vs. multitask performance sensitivity need to be done.

      COHb Formation Rate.  It is possible that the rate at which COHb is formed is an important
       variable in the effects of COHb on the CNS.  To explore this possibility, studies were cast
20     into a table according to their rate of COHb formation (fast or slow) and whether effects were
       found or not.  When COHb was formed to its target value in 10 min or less, the
       study was tabulated as fast. Studies in which effects were found only above 10% COHb were
       tabulated as no effects to avoid consideration of frank effects.
            Table 10-27 is the result of the above classification.  The exploratory two-sided Fisher's
25     exact test yielded p = 0.002.  It appears that studies using slow COHb formation are more
       likely to find significant effects. Experiments explicitly designed to test the hypothesis would
       be needed before firm conclusions can be drawn.
30                   TABLE 10-27. EFFECT OF RATE OF COHb FORMATION
                                           Slow                                   Fast
35 Effects
No Effects
20
12
2
9
40     Fisher's exact test, p = 0.002.

       March 12, 1990                         10-145     DRAFT-DO NOT QUOTE OR CITE

-------
       10.4.2.10  Hypotheses
       Dose-Effect Function
            An effort has been made to unify the dose-effects literature concerning CO and behavior
       (Benignus et al., 1989b). The analyses and hypotheses of the latter article will be reviewed
 5     below.  Both laboratory animal and human data were considered. Only dose-effects studies
       were considered  so that comparisons and extrapolations could be made.  The literature
       concerning dose-effects functions in humans, as above, was found to be inconsistent.  As has
       been pointed  out in the present review, effects of COHb  elevation do not become significant
       until ca. 20% in laboratory animals. The literature for such higher levels of COHb is quite
10     consistent. Nonlinear, positively-accelerated functions were fitted to the laboratory animal
       data to describe dose-effects relationships.
            To compare the human data to the laboratory animal data, it was necessary to select
       from the divergent group of human dose-effects studies.  The argument was made that the
       most accurate estimate of the dose-effects function was that the effects below ca. 20% COHb
15     were either zero  or very small in size.  This argument was based  on the observations about
       nonverification of findings in the literature as well as upon the findings of Benignus et al.
       (1989a), which demonstrated small but nonsignificant elevations in mean tracking error in a
       large study (n=74) with COHb levels up to 17%.
            The data from Benignus et al. (1989a) were fitted with the same form of function  as
20     fitted to the laboratory animal data.  Extrapolation of the curve projected the human data as
       passing through laboratory animals curves. The latter observation was used to imply that the
       human and laboratory animal findings were similar and that frank effects of COHb elevation
       in humans should not be expected below ca. 20% COHb.
            An argument was made for the possibility and importance of small effects for low
25     COHb levels but no unimpeachable evidence could be marshalled.  The possibility of small
       effects below ca. 20% COHb was supported by the observation that in many studies, using
       both laboratory animals and humans, means were usually shifted in the direction of
       deleterious effects at low levels, but not in a statistically significantly manner.  Means were
       much more rarely shifted in directions implying improvement of behavioral abilities with
30     small levels of COHb. Thus, there may exist  small effects below 20% COHb (or some
        March 12, 1990                         10-146     DRAFT-DO NOT QUOTE OR CITE

-------
       individuals are affected while most are not).  The latter possibility cannot be ignored but it
       cannot be confirmed.

       Compensatory Mechanisms
 5          As discussed above in Section 10.4.1, a proportional vasodilation occurs in the brain in
       response to COHb elevation. This vasodilation is sufficient, on the average, to keep the
       cerebral O2 consumption from being reduced even though the COHb has reduced the blood's
       O2-carrying capacity 20 to 30% and the presence of COHb has shifted the OjHb dissociation
       curve to the left.  The cerebral vasodilation may be viewed, Ideologically, as a closed-loop
10     compensatory mechanism to assure adequate oxygenation of the brain in the presence of
       elevated COHb.
            If the cerebral vasodilation is adequate in any individual and if the vasodilation is
       homogeneous for all cerebral tissue, then that individual should not be behaviorally impaired
       by COHb elevation.  This statement assumes that the sole mechanism for CO toxicity is the
15     hypoxic effect of COHb.
            The agreement between the behavioral literature and the compensatory mechanism
       hypothesis is noteworthy.  According to the compensatory mechanism data, O2 consumption
       in the brain does not begin to decrease until COHb exceeds 20 to 30%.   Data from behavioral
       studies in laboratory animals demonstrate that significant effects in schedule-controlled
20     behavior do not occur below 20 to 30%  COHb.  Behavioral effects in humans have not been
       unambiguously demonstrated below 20 to 30% COHb.

       10.4.2.11  Conclusions
            At the present stage  of evidence, it seems unfounded to conclude that COHb elevation
25     below ca. 20% is deleterious to the behavioral abilities in humans. It seems unwise,
       however,  to ignore the frequent evidence in favor of effects.  Even if effects are small or
       occasional, they might be important to the performance of critical tasks.
            Some of the differences among studies of the effect of COHb on the behavior of humans
       is due apparently to technical problems in the execution of experiments, because single-blind
30     or nonblind experiments tend to yield a much higher rate of significant effects than do double-
       blind studies. Even when non-double-blind experiments are eliminated from consideration,

       March 12, 1990                         10-147     DRAFT-DO NOT QUOTE OR CITE

-------
       however, a substantial amount of disparity remains among results of studies.  It is possible
       that such residual disagreement is due to the action of an unsuspected variable that is not
       being controlled across experiments.
            If the compensatory CNS blood flow hypothesis has validity, it is possible that there
 5     exist groups that are at higher risk to COHb elevation than the usual subjects who were
       studied in the behavioral experiments.  Disease or injury might either impair the
       compensatory mechanism or reduce the non-exposed O2 delivery.  Aging increases the
       probability of such injury and disease.  It also is possible that there exist individual
       differences with regard to COHb sensitivity and/or compensatory mechanisms.  Too little is
10     known about the compensatory process to make conjectures, but the matters seem important
       to investigate.
            The literature on the behavioral effects of COHb elevation has grown considerably since
       the last Criteria Document was written (U.S. Environmental Protection Agency, 1979). It
       seems safe to state that the effect of the new information did not increase the certainty about
15     COHb effects.  Unless some key piece of information is uncovered by new research, there
       does not seem to be much hope of gaining clarification in the conflicting findings.  The
       solution to the puzzle would seem to lie in the conduct of more research into mechanisms of
       action of CO rather than in further attempts to show reliable behavioral effects.  The latter
       approach, which has not been successful in the past, should be resumed only when
20     mechanisms of toxicity are understood better. More findings of behavioral effects of COHb
       would not appreciably alter the conclusions of the present  section unless future studies were to
       show an unusual unanimity.
25     10.5  DEVELOPMENTAL TOXICITY OF CARBON MONOXIDE
       10.5,1  Introduction
            Developmental toxicity has been described in the U.S. Environmental Protection
       Agency's Guidelines for the Health Assessment of Suspect Developmental Toxicants (Federal
       Register, 1986) as including death of the developing organism, structural abnormalities,
30     altered growth, and functional deficits resulting from toxic exposures that occur prior to the
       subject's attaining sexual maturity.  The appearance of toxic effects may occur at any time

       March 12, 1990                         10-148     DRAFT-DO NOT QUOTE OR CITE

-------
       throughout life.  Concern for special vulnerability of immature organisms to toxic compounds
       focuses on the possibilities that (1) a toxic exposure that is not sufficient to produce maternal
       toxicity or toxicity in the adult organism will adversely affect the fetus or neonate or (2) at a
       level of exposure that does produce a toxic consequence to the adult or mother that the fetus
 5     or neonate suffers a qualitatively different toxic response.  Toxic responses that occur early in
       life, but which are not permanent, may or may not be a cause of concern.  In some cases they
       truly may be transient events with no persisting consequences.  However, in other cases  such
       results may have their own consequences for development of the organism or may reappear
       under conditions of ill health produced by other toxic exposures, environmental stresses, or
10     exposure to pathogenic agents. Therefore, even seemingly transitory toxic  effects must be
       viewed as serious consequences of exposure when they occur in humans.  In this section, data
       are presented that describe the toxic consequences of CO exposure early in  development.
       These data describe the types of toxic outcomes that the immature subject may show and also
       help to identify the dosage at which such toxicity is seen.
15          There are theoretical reasons and supporting experimental data that  suggest that the fetus
       and developing organism are especially vulnerable to CO.  One reason for approaching the
       fetus as a separate entity for purposes of regulation is that the fetus is likely to experience a
       different CO exposure than the adult given identical concentrations of the gas in air. This is
       due to differences in uptake and elimination of CO from fetal hemoglobin which are
20     documented below.  Less studied is the possibility that tissue hypoxia may differ between the
       fetus and adult even at equivalent carboxyhemoglobin concentrations as a result of differences
       in perfusion of critical organs, in maturation of adaptive cardiovascular responses to hypoxia
       and as a result of tissue requirements for  oxygen.  Inferences concerning these factors  are
       obtained principally from experimentation performed in laboratory animals in which the
25     immature organism does show enhanced toxicity relative to the adult.  Concern must be
       expressed, too, for the development of sensitive and appropriate animal models. It is
       necessary to bear in mind the relative state of development of the human and laboratory
       animal in question at the time  of birth in  developing useful animal models.  For example, the
       neonatal rat is very immature relative to the neonatal human at birth, with respect to
30     development of the central nervous system, (Fechter et al., 1986) and so a  combined prenatal
       and neonatal exposure model may be more accurate in predicting consequences of prenatal

       March  12, 1990                          10-149     DRAFT-DO NOT QUOTE OR CITE

-------
        exposure in the human. Further, differences appear to exist among species in the relative
        affinity of fetal and adult hemoglobin for CO.  These data are reviewed by Longo (1970).
            There exist a variety of relevant data bases that will be reviewed.  These include
        experimental investigations conducted using laboratory animals (and these are most
 5      numerous),  case report data collected in offspring of women exposed to generally high-level,
        acute CO poisoning during pregnancy, and epidemiological data.  One large, but
        problematical, literature from the standpoint of this document concerns maternal cigarette
        smoking.
            Cigarette smoking constitutes a major source of exposure of the individual to CO and
10      this is particularly relevant for the fetus because of the high affinity of fetal hemoglobin for
        CO (Longo, 1977).  Maternal smoking has been associated with a variety of untoward
        consequences ranging from increased incidence of placenta previa, abruptio placentae,
        spontaneous abortion and subsequent fetal deaths to depressed birthweight, and increased
        numbers of hospital admissions for a broad range of complaints during at least the first
15      5 years of life, to poorer than predicted school performance during the first 11  years of life.
        This literature has been thoroughly reviewed as a report to the U.S. Surgeon General
        (National Institute of Child Health and Human Development, 1979). These outcomes should
        be  cause for significant concern. However, it is not easy to determine the extent to which CO
        is a causative factor.  Cigarette smoke contains a large number of toxic chemicals other than
20      CO and these other agents either alone or in combination may be responsible /or the untoward
        outcomes that are readily associated with maternal smoking. A few epidemiological reports
        are reviewed below in which it is concluded that CO, either from ambient sources or cigarette
        smoke, is responsible for developmental disruption.  However, these reports generally are
        deficient in  characterizing the level of CO exposure or in ruling out potential contributions by
25      other toxic agents contained in smoke.  Investigations with laboratory  animals exposed to CO
        rather than cigarette smoke early in development have demonstrated developmental anomalies
        and persisting neurobehavioral disorders that are most relevant to this  document and that are
        reviewed below.  Because such effects are seen at CO levels approaching values observed in
        the offspring of cigarette smokers,  this  must be cause for serious concern.  However, it is not
30      possible to use the cigarette smoker literature in establishing criteria for permissible CO
        exposure.

        March 12, 1990                          10-150     DRAFT-DO  NOT QUOTE OR CITE

-------
       10.5.2  Theoretical Basis for Fetal Exposure to Excessive Carbon Monoxide
                and for Excess Fetal Toxicity
            Hill et al. (1977) aptly described mathematical models for predicting fetal exposure to
       CO based upon placental transport and the differences between maternal and fetal hemoglobin
 5     affinity for CO and O2. They predicted that for any maternal CO exposure of moderate
       duration that fetal COHb levels would lag behind maternal COHb levels for several hours, but
       would, given sufficient time, surpass maternal COHb levels by as much as  10% (in humans)
       due to the higher affinity of fetal hemoglobin for CO than adult hemoglobin.  Moreover, they
       predicted a far longer wash-out period for the fetal circulation to eliminate CO following
10     termination of exposure than that found in the adult.  Data, accumulated in both laboratory
       animal and human studies, support these conclusions.

       10.5.2.1  Evidence for Elevated Fetal Carboxyhemoglobin Relative to Maternal
                Hemoglobin
15          A fairly wide range of neonatal and maternal COHb levels has been published for
       humans, probably due to wide differences in cigarette smoking patterns prior to and during
       labor.  In one recent study, measurement of fetal cord blood in the offspring of cigarette
       smokers who smoked during labor showed fetal COHb levels 2.55 times higher than in
       maternal  blood. Cord blood averaged 10.1 % COHb at delivery while maternal blood
20     averaged 5.6% on the mother's arrival at the hospital and 4.1 % at delivery (Bureau et al.,
       1982). These values for fetal COHb are fairly  high relative to other published sources
       (Longo, 1977 - Table IV). However, greater fetal COHb levels have been found in
       laboratory studies across a broad range of animal species if sufficient time was allowed for
       COHb to equilibrate in the fetal compartment.  Christensen et al.  (1986) ultimately observed
25     higher CO levels in fetal guinea pigs than in  their dams following CO exposure given near
       term.  Immediately following maternal exposure, at gestational age  62 to 65 days, to a bolus
       of CO (5 mL given over 65 seconds), they reported a faster elevation in maternal COHb
       levels than in fetal levels,  a finding consistent with the models of Hill et al. (1977).
            Anders and Sunram (1982) exposed gravid rats to 22 ppm CO for one hour on Day 21
30     of gestation and reported that fetal COHb levels averaged 12%  higher than  levels taken at the
       same time period in the dam. These results are consistent with those of Garvey and Longo
       (1978) whose study involved chronic CO exposures in rats. Dominick and  Carson (1983)
       March 12, 1990                         10-151     DRAFT-DO NOT QUOTE OR CITE

-------
       exposed pregnant sows to CO concentrations of 150 to 400 ppm for 48 to 96 h between
       gestation days (GDs) 108 to 110.  They reported fetal COHb levels that exceeded maternal
       levels by 3 to 22% using a CO-oximeter.
            Longo and Hill (1977) similarly reported that COHb levels in fetal lambs do exceed
 5     maternal levels once equilibrium is reached in the fetal compartment and that this washout
       from the fetal blood exceeds that observed for maternal blood.
            Fetal COHb kinetics may not be static throughout pregnancy. Bissonnette and Wickham
       (1977) studied transplacental CO uptake in guinea pigs at approximate gestational ages 45 to
       68 days. They report that placental diffusing capacity increases significantly with increased
10     gestational age and appears  to be correlated with fetal weight rather than placental weight.
       Longo and Ching (1977) also showed increases in CO diffusion rates across the placenta of
       the ewe during the last trimester of pregnancy.  However, they did not find a consistent
       increase when diffusion rate was corrected for fetal weight (i.e., when diffusing capacity was
       expressed on a per kilogram fetal weight basis).
15
       10.5.2.2  Effect of Maternal Carboxyhemoglobin on Placental O2 Transport
            Gurtner et al. (1982) studied the transport of O2 across the placenta in the presence of
       CO by cannulating both the maternal and fetal vessels of anesthetized sheep preparation.
       They measured the transport of O2 across the placenta compared to transport of argon, urea,
20     and tritiated water when CO was introduced. They  showed a reduction in O2 diffusing
       capacity relative to Ar that appeared to be related to the level of maternal COHb. Reduction
       of O2 transport was observed below 10% COHb and O2 transport approached zero at COHb
       values of 40 to 50%.  They interpreted these data as supporting the role of carrier-mediated
       transport for O2 and suggest that CO competitively binds to this carrier.  An alternative
25     explanation is that the introduction of CO simply reduces the amount of fetal hemoglobin
       available to bind O2. Moreover, Longo and Ching (1977), for example, were unable to alter
       CO diffusing capacity across the placenta by administration of a series of drugs that bind to
       cytochrome P-450.  Gilbert et al. (1979) have underscored the low concentration of
       cytochrome P-450 in human placenta, as compared to liver and to the very low levels found
30     in many other species.
        March 12, 1990                          10-152     DRAFT-DO NOT QUOTE OR CITE

-------
             Christensen et al. (1986) suggest that maternal CO exposure may independently impair
        O2 diffusion across the placenta due to the enhanced affinity of maternal hemoglobin for O2 in
        the presence of COHb (the Haldane effect).  Using the guinea pig, these authors demonstrated
        an initial almost instantaneous fall in fetal paO2 levels and an increase in fetal pCOj, which
 5      subsequently was followed by an increase in fetal COHb between approximately 5 to 10 min
        (the last time point studied but a time when fetal COHb values were still far below maternal
        values). They calculated that the decrease in fetal O2 transfer was due mostly to a decrease in
        maternal O2-carrying capacity, but also, perhaps up to one-third, by the increased affinity of
        hemoglobin for O2 in the presence of CO. This model assumes that uterine perfusion remains
10      constant under the experimental conditions used. Longo (1976) also showed a significant
        dose related drop in fetal O2 levels in blood taken from the fetal descending artery and fetal
        inferior vena cava after pregnant ewes were exposed to variable levels of CO for durations
        sufficient to yield COHb equilibration in both the fetal and maternal compartments. To
        summarize, it has been demonstrated that the presence of maternal COHb over a range of
15      values results in depressed O2 levels in fetal blood.  The simplest explanations for the inverse
        relationship between maternal COHb and fetal O2 levels are reduced maternal O2-carrying
        capacity, impaired dissociation of O2 from maternal hemoglobin (the Haldane effect), and
        reduced availability  of free fetal hemoglobin able to bind O2.

20      10.5.3  Measurement of Carboxyhemoglobin Content in Fetal Blood
            Zwart et al. (1981) and Huch et al. (1983) have called into question the accuracy of
        spectrophotometric measurements of COHb in fetal blood using the IL 182 and 282 CO-
        oximeters.  The CO-oximeter is effectively a spectrophotometer preset to read samples at
        specific wavelengths that correspond to absorbance maxima for oxy- carboxy- and met-
25      hemoglobin determined using adult blood samples.  Different plug-in modules (IL 182) or
        programmed absorbance values (IL 282) can be used to correct for species differences in the
        absorbance spectrum of rat, human, dog, and cow.  Some investigators have used these
        instruments for  estimating COHb levels in species for which the instrument has not been
        calibrated such as the pig and guinea pig.  Typically, individual investigators have calibrated
30      the CO-oximeter using blood standards fully saturated with CO and with O2.  The adequacy
        of such a procedure  is not certain. (See Chapter 8, Section 8.5 for more details on the

        March 12,  1990                        10-153     DRAFT-DO NOT QUOTE OR CITE

-------
       measurement of COHb.) Further, the correspondence of absorbance maxima between adult
       and fetal hemoglobin for species upon which the CO-oximeter is calibrated at the factory is
       an empirical question for which little data are published. Noting the finding  of higher
       apparent COHb levels in the venous cord blood of humans than in the uterine artery, Huch
 5     et al. (1983) examined the possibility that oxyhemoglobin in fetal blood might interfere with
       accurate measurement of COHb levels in the fetus due, presumably, to different absorbance
       maxima for fetal than adult blood.  Working in vitro, Huch et al. (1983) deoxygenated fetal
       and maternal blood by flushing a tonometer with nitrogen and 5% CO2.  They introduced a
       "small volume" of CO gas, measured the blood gases using the IL 282 CO-oximeter, and then
10     studied the effect of stepwise addition of O2 to the apparent COHb levels.  They showed little
       influence of O2 saturation upon maternal COHb, but indicate that O2 saturation did affect
       readings of fetal  COHb so as to overestimate COHb. This confound is particularly likely at
       high oxyhemoglobin concentrations.  Zwart et al. (1981) suggest an apparent elevation of
       COHb levels of approximately 2% with 40% oxyhemoglobin saturation and of approximately
15     6% with oxyhemoglobin levels of 90 to 95%.  Such errors do not invalidate  the finding that
       fetal COHb exceeds maternal values, but do bring into question the magnitude of this
       difference. Whether similar errors also occur in measuring fetal COHb levels in animal blood
       is uncertain and should be subjected to experimental test.  The calibration of
       spectrophotometers based upon fetal hemoglobin absorbance spectra rather than automated
20     analysis based  upon adult absorbance spectra is recommended to achieve greater accuracy in
       determining absolute levels of  CO in fetal blood.  Vreman et al. (1984) have described a gas
       chromotographic method for measuring COHb which has been applied to human neonates.
       Because of the very small volume of blood needed to make these measurements and because
       they eliminate  the problem of absorbance spectra of fetal hemoglobin, this may be considered
25     a useful means of accurately assessing COHb in developing organisms. There also is a new
       model of the Co-oximeter (#482) which apparently allows for use of absorbance spectra based
       on calibration of fetal blood.

       10.5.4  Consequences of  Carbon Monoxide in Development
30          This section presents the  evidence that CO exposure during early development has the
       potential of producing untoward effects.  The four types of toxic outcomes - fetotoxicity,

       March 12, 1990                          10-154     DRAFT-DO NOT QUOTE OR CITE

-------
         gross teratogenicity, altered growth, and functional deficiencies in sensitive organ systems -
         are considered in order. As is the case in adult organisms, the nervous and cardiovascular
         systems appear to be most sensitive to CO exposure.

  5      10.5.4.1  Fetotoxic and Teratogenic Consequence of Prenatal Carbon Monoxide
                  Exposure
             There is clear evidence from human and animal studies that very high levels of CO
         exposure may be fetotoxic.  However, there exists some question concerning the level of
         exposure that causes fetal death as both the duration and concentration of maternal exposure
 10      are critical values in determining fetal exposure.  More important for the setting of ambient
         air standards is  the suggestion of a causal relationship between sudden infant death syndrome
         (SIDS) and ambient CO levels.  These data are extremely limited at present and no conclusive
         correlation can be drawn between SIDS and CO.
             The lowest-observed-effect level (LOEL) for fetotoxicity in laboratory animals appears
 15      to be in the range of 500 pprn for rodents, but one experiment conducted in pigs suggests that
         this species may be especially sensitive to this effect showing significant fetotoxicity at
         250 ppm for 2 to 4 days late in gestation.  Evidence of fetotoxicity in animals also has come
         from acute, high-dose experiments that are not included here because they are not directly
        relevant to standard setting.
20          The data that suggest that prenatal CO exposure produces terata is extremely limited
        and, again, comes largely from quite high exposure levels. Table 10-28 presents the reported
        effects of prenatal CO exposure on fetotoxicity, teratogenicity, and growth abnormalities.
        Perinatal Carbon Monoxide Exposure and Sudden Infant Death Syndrome (SIDS)
             It has been suggested that CO may be a causative factor in SIDS. Hoppenbrouwers
25      et al. (1981) reported a statistical association between the frequency of SIDS and levels of
        several airborne pollutants including CO,  sulfur dioxide (SO2),  NO2, and hydrocarbons.
        SIDS was reported more commonly in the winter, at a time when the burning of fossil fuels
        for heating would be greatest.  It is interesting to note that there is a phase lag of
        approximately 7 weeks between the increase in pollutant levels and the increase of SIDS
30      cases. Whether  this phase lag  for SIDS represents a failure to identify the true cause of SIDS
        or a lag with some meaning in terms of the physiology  is uncertain.  Further correlations

        March 12,  1990                          10-155     DRAFT-DO NOT QUOTE  OR CITE

-------
cr
to
                     TABLE 10-28.  TERATOGENIC CONSEQUENCES OF PRENATAL CARBON MONOXIDE EXPOSURE
                                                                     IN LABORATORY ANIMALS
        Species (Strain)
Maternal Treatment
                              Maternal Toxicity
                                         Development Abnormality
                                                                                                                                                 References
o

ON
        Mouse (strain NR)
        Rat (Ames-Wilson)
        Rat (Sprague-Dawley)
        Rabbit (strain NR)
        Mouse (CF-1) and
        Rabbit (New Zealand)
5,900 or 15,000 ppm CO
for 5-8 min every
other day of
gestation

3,400 ppm CO for 1 h/day
for 3, 6, or 8.3 mo
750 ppm CO for 3 h/day
on GDs 7, 8, or 9
90 or 180 ppm CO from
mating to the day before
parturition
7 or 24 h/day of 250 ppm
CO on GDs 6-15 for mice
and on GDs 6-18 for rabbits
Acute effects:  unconsciousness
(no COHb levels)
Decreased body weight, appetite,
and muscle tone; lack of grooming
(COHb levels of 60-70%)
NR
(no COHb levels)
NR
(COHb levels of 8-9% and
and 16-18%)
Transient increase in body
weight for mice in 7-h/day
group; COHb levels of 10-11 %
(mice) and 13-15% (rabbits)
for 7-h/day exposures
Abortions, resorptions, and
abnormal growth of survivors
Decrease of litter size, decrease
of preweaning survival (50%
reduction of pregnancy at 3 mo,
no pregnancies induced with
longer exposures;  19% increase
of estrous cycle)

Absorptions, stillbirths, and
skeletal anomalies, decreased
fetal body weight and crown-
rump length

180 ppm: 35% mortality of
neonates,  11 % decrease in
birth weights, and increase
in malformations 90 ppm:
9.9% mortality of neonates,
13% decrease in birth weights

Mice: increase in resorptions
and body weight w/7-h/day
exposure, decrease in body
weight and crown-rump length
w/24-h day exposure; both
exposures increased skeletal
anomalies (GD 18)
Rabbits: increase in body
weight and crown-rump length
w/7-h/day exposure
Wells (1933)
William and Smith (1935)
Choi and Oh (1975)
Astrup et al. (1972)
                                                                                                                                                 Scwetz et al. (1979)

-------
»—*
to
>«

VO
               TABLE 10-28 (cont'd).  TERATOGENIC CONSEQUENCES OF PRENATAL CARBON MONOXIDE EXPOSURE
                                                                IN LABORATORY ANIMALS
Species (Strain)
Maternal Treatment
Maternal Toxicity
Development Abnormality
                                                                                                                               References
        Rat (Long-Evans)
        Mice (CD-I)
        Mice (CD-I)
        Pig
        Rabbit
                           0, 30, or 90 ppm CO or
                           13% oxygen in nitrogen
                           on CDs 3-20
                           125
                           250
                           500 ppm CO CDs 8-18
                           0, 65, 125 ppm CO
                           GDs 7-18
                           150-450 ppm for 48-
                           96 h between GDs
                           108-110
                           12 "puffs" of 2700-
                           5400 ppm CO daily from
                           GDs 6-18
                            Decrease in successful
                            pregnancies; COHb levels
                            of 4.8 and 8.8%
                                                               None
                            Decreased maternal respiration rate
                            Significant maternal death rate
                                      13% oxygen:  12% decrease in
                                      body weight
                                      90 ppm CO: 14% increase in brain
                                      weight, 24% decrease in lung
                                      weight, serotonin concentration
                                      decrease in brain

                                      Increased fetal mortality
                                      significant w/500 ppm, no
                                      effect on number of
                                      implantation sites

                                      Impaired righting reflex on
                                      PD 1 for 125-ppm group,
                                      impaired negative geotaxis
                                      on PD 10 for 125-ppm group,
                                      impaired serial righting on
                                      PD 14 for both 65- and 125-ppm
                                      CO groups

                                      Linear increase in number of
                                      stillbirths significant when
                                      maternal COHb exceeded 23%
                                      (approximately 2500 ppm)

                                      Larger number of fetal deaths
                                      No terata
                                  Garvey and Longo (1978)
                                                                                                    Singh and Scott (1984)
                                                                                                    Singh (1986)
                                                                                                    Dominick and Canon (1983)
                                   Rosenkrantz et al. (1986)
        Rat (Wistar)
                           1,000-6,000 ppm CO
                           2x/day for2h
                           40 min total from
                           GDs
                           0-6
                           7-13
                           14-20
                           0-20
                                                                                                     Decreased fetal weight
                                                                                                     at GD 21
                                                                                                    Tachi and Aoyama (1983)
                                                                                                    Tachi and Aoyama (1986)

-------
       were obtained between SIDS and the predicted level of CO and lead for the child's birth
       month and between SIDS and the level of pollution at the reporting station closest to the
       infant's home.  These correlations are not compelling without more information on the
       methods by which other possible risk factors were controlled in making the geographical
 5     correlations.  While it is technically difficult, it would be very useful to obtain COHb levels
       close to the time of death in SIDS victims as this would greatly assist in determining the
       incidence of elevated CO exposure in such cases.
            There have been several studies linking maternal cigarette smoking with SIDS (these are
       reviewed in the National Institute of Child Health and Human Development, 1979 report on
10     "Smoking and Health"), but it is uncertain what the role of CO might be in such a
       relationship.  Thus, it is clear that severe, acute CO intoxication can be fetotoxic although
       specification of maternal and fetal COHb levels is difficult because such exposures rarely
       involve the achievement of steady-state COHb levels  or permit careful and rapid
       determination of COHb levels.  More relevant to the  issue of standards for ambient exposure
15     is the possible link between CO and SIDS, but this literature currently is insufficient to
       determine whether such a relationship exists.

       Fetotoxicity in Laboratory Animals
            Working with CD-I strain mice, Singh and Scott  (1984) found significant increases in
20     the number of dead or resorbed fetuses per litter and  an increase in fetal mortality with
       continuous CO exposure of 500 ppm from Gestation  Day (GD) 8 until subjects were
       sacrificed at GD 18.  Although not statistically  significant, they found a dose-related trend in
       these measures beginning at approximately 125  ppm.  The no-observed-effect level (NOEL)
       for these measures was 250 ppm.  There was no effect of CO on the number of implantation
25     sites suggesting a fetotoxic rather than an embryopathic event.  Schwetz et al. (1979) also
       exposed mice to 250 ppm CO for 7 and 24 h per day on GDs 6 to 15.  They found no effect
       on number of implantation sites or number of live fetuses per litter but did show a significant
       elevation in resorptions with the 7-h exposure (10 to  11% COHb) but not with 24-h per day
       exposures.
30          Dominick and Carson (1983) exposed pregnant  sows to CO concentrations of 150 to
       400 ppm for  48 to 96 h between GDs 108 to 110 (average gestation was 114 days).  They

       March 12, 1990                         10-158     DRAFT-DO NOT QUOTE OR CITE

-------
        showed a significant linear increase in the number of stillbirths as a function of increasing CO
        exposure.  Stillbirths were significantly elevated above control levels when the maternal
        COHb levels exceeded 23%  saturation. These saturation levels were obtained at
        approximately 250 ppm.  COHb levels were measured using an IL 182 CO-oximeter equipped
 5      with a human blood board; pig blood fully saturated with CO and with O2 were run each day
        to calibrate the instrument. There was a very large variability among Utters at a given
        concentration level in the percentage of stillbirths that occurred.  Penney et al. (1980) found
        evidence of reduced litter size in rats exposed for the last 18 days of gestation to 200 ppm CO
        (maternal COHb levels averaged 28%).  However, Fechter et al. (1987) did not observe
10      similar effects on litter size in rats exposed to levels of CO as high as 300 ppm (maternal
        COHb levels of 24%).

        Teratogenicity
            There are very limited data (Astrup et al.,  1972) suggesting increased fetal mortality and
15      malformations among rabbits exposed to 180 ppm CO throughout gestation (COHb levels =
        16 to 18%).  The frequency of malformations reported was very small, the historical rate of
        such anomalies in the laboratory undocumented, and so these results require replication by
        other workers before they  can be considered as the basis for regulation.  Rosenkrantz et al.
        (1986) exposed rabbits to high doses (12 puffs of 2,700 to 5,400 ppm CO) for short time
20      periods daily from CDs 6 to 18.  COHb levels were estimated at 16% although animals had
        not equilibrated with the inhaled mixture. Despite a large number of fetal deaths, there was
        no evidence of terata in the CO-exposed animals.
            Choi and  Oh (1975) reported  skeletal anomalies in rats exposed to 750 ppm CO for 3 h
        per day on GDs 7, 8, or 9. They also reported an excess in fetal absorptions and stillbirths
25      and  a decrease in body length. Schwetz et al. (1979) reported no teratogenic  effects but an
        increase in minor skeletal variants in CF-1  mice exposed to 250 ppm CO for 24 h per day
        from GDs 6 to  15.

        10.5.4.2  Carbon Monoxide and Body Weight
30          One of the best studied and possibly one of the most  sensitive measures  of early CO
        exposure is a depression in birthweight. The effect seen in animals following fetal CO

        March 12, 1990                         10-159     DRAFT-DO NOT QUOTE OR CITE

-------
       exposure is generally transitory and occurs despite the fact that maternal body weight growth
       through pregnancy does not appear to be adversely affected.  The LOEL is in the range of
       150 to 200 ppm in laboratory animals. In as much as the depressed birthweight observed is a
       transient event, its significance is not clear.  However, in humans, low birthweight babies
 5     may be at particular risk for many other developmental disorders so that the effect cannot be
       disregarded casually.  Moreover, in humans there is a strong correlation between maternal
       cigarette smoking and reduced birthweight.  Whether the causative agent here is CO, nicotine,
       or a combination of these or other agents is uncertain.
            Studies relating human CO exposure from ambient sources or cigarette smoking to
10     reduced birthweight frequently have failed to take into account all sources of CO exposure.
       Alderman et al. (1987), for example, studied the relationship between birthweight and
       maternal CO exposure based upon neighborhood ambient CO data obtained from stationary
       air monitoring sites in Denver. They failed  to show a relationship between these factors, but
       failed to control for maternal cigarette smoking or possible occupational exposures to CO.
15     COHb measurements were not made either among the mothers or their offspring to estimate
       net exposure levels. A similar design problem is found in the study of Wouters et al. (1987)
       where cord blood COHb and birthweight were correlated.  The authors report a significant
       correlation between cigarette smoking and reduced birthweight, but no correlation between
       cord blood COHb and birthweight. Such data might be interpreted to mean that CO is not the
20     component in cigarette smoke responsible for reduced birthweight. Such a conclusion appears
       to be unjustified based upon Wouters et al. (1987), because COHb is a good estimate of
       recent CO exposure only.  Thus it may indicate only how recently women in this study
       smoked their last cigarette before delivery of the child rather than estimating smoking rates or
       history throughout pregnancy.
25          Other studies have related indirect exposure to smoke in pregnancy with reduced
       birthweights. Martin and Bracken (1986) showed an association between passive smoking
       (exposure to cigarette  smoke for at least 2 h per day) and reduced birthweight.
       Unfortunately, side-stream smoke contains significant nicotine as well as CO and so it is not
       possible to relate this effect to CO exposure.
30          Mochizuki et al. (1984) attempted to evaluate the role of maternal nicotine intake in
       reduced birthweight and did present evidence of possibly impaired utero-placental circulation

       March 12, 1990                         10-160     DRAFT-DO NOT QUOTE OR CITE

-------
       among smokers.  These changes were not related specifically to the nicotine content of the
       cigarettes and failed, moreover, to take into account the possible synergistic effects between
       reduced perfusion that might have resulted from the vasoconstrictive effects of nicotine and
       the reduced O2 availability that might have resulted from CO exposure. As noted in the
 5     section of this report that deals with the effects of high altitude, many of the outcomes of
       maternal CO exposure also are observed in offspring of women living at high  altitude. These
       include reduced birthweight, increased risk of perinatal mortality, and increased risk of
       placenta! abnormalities. Limited data exist on the possibility of increased risk of CO
       exposure to the fetus being carried at high altitude.  Such findings are considered in the
10     section on high altitude.
            Fechter and Annau (1980) replicated earlier data from their laboratory showing
       significantly depressed  birthweights in pigmented rats exposed throughout gestation to
       150 ppm CO. Penney  et al. (1980) also found a significant depression in birthweight among
       rats exposed for the last 18 days of gestation to 200 ppm CO.  Penney et al. (1983) showed a
15     trend toward divergence in bodyweight among fetuses exposed to 200 ppm CO, which
       developed progressively during the last 17 days of parturition suggesting that late gestational
       exposure to CO may be essential to observe the effect.   Storm et al. (1986) reported that in
       the following CO exposure from the beginning of gestation through postnatal day (PD) 10
       that body weight  was depressed in a dose-dependent fashion  at 75, 150, and 300 ppm CO.
20     Moreover, these values were all significantly lower than air-control subjects.  By age 21 days
       no significant body weight differences were seen among the  test groups.  At no time have
       Fechter and colleagues observed evidence of maternal toxicity as identified by death, reduced
       maternal weight gain, or gross physical appearance.  Morris et al. (1985b) exposed neonatal
       piglets chronically to CO for 21 days starting at approximately 28 days of age (200- and 300-
25     ppm COHb levels averaged 16 and 21%, respectively).  They reported a significant
       impairment in weight gain in pigs exposed to 300 ppm, but no effect in pigs exposed to
       200 ppm CO.

       10.5.4.3  Alteration in Cardiovascular Development following Early Carbon Monoxide
30               Exposure
            It is known  that a variety of cardiovascular and hematopoietic changes can accompany
       hypoxia in neonates and adult subjects including elevation in hemoglobin, hematocrit, and
       March 12, 1990                          10-161      DRAFT-DO NOT QUOTE OR CITE

-------
        heart weight.  Data gathered in adult laboratory animals suggest that these changes may be
        related.  Cardiomegaly resulting from hypoxia reflects the amount of work performed to
        extract an adequate supply of O2.  Whether the same processes occur in prenatal and neonatal
        CO-induced hypoxia has been the subject of several reports (these reports are summarized in
 5      Table 10-29).  For prenatal exposure, the accumulated laboratory animal data suggest that
        CO-induced cardiomegaly may be proportionately greater than in adult animals at a given
        maternal CO-exposure level. Whether this is due to higher fetal COHb levels, as a
        consequence of fetal hemoglobin's affinity for CO is not clear. While the cardiomegaly may
        resolve when the neonatal subject is placed in a normal air environment,  there is evidence for
10      a persisting increase in the number of muscle fibers. The functional significance of these
        changes, if any, is uncertain.  The LOEL for fetal cardiomegaly has not been well
        determined. One experiment has shown significant  elevation of heart weight following CO
        exposures as low as 60 ppm (Prigge and Hochrainer, 1977) and there are no published dose-
        response experiments that provide a NOEL. Chronic prenatal CO exposure at approximately
15      200 ppm results in a significant increase in the number of muscle fibers in the heart.  The
        NOEL for this change has not been determined.
             Fechter et al. (1980) measured wet-and-dry heart weight and protein and nucleic acid
        levels at several time points between birth and weaning in rats prenatally  exposed to 150 ppm
        CO or to air.  They reported that neonates had significantly elevated wet-heart weights despite
20      a slightly reduced body weight at birth. Groups did not differ at birth in  dry-heart weight,
        total protein, or RNA or DNA levels in whole heart. No significant differences between
        groups on any measure were present at PD 4 or subsequent ages studied.  The data were
        interpreted as evidence for cardiac edema rather than a change in heart muscle mass itself.
        The finding of a heavier heart at birth replicated the finding of Prigge and Hochrainer (1977),
25      who exposed rats prenatally to CO at levels as low as 60 ppm and observed a similar increase
        in heart weight.  Clubb et al. (1986) have conducted a comprehensive experiment in which
        rats were exposed to 200  ppm CO either prenatally from GD 7 (CO/air),  prenatally and
        postnatally until age 28 days (CO/CO), only postnatally for various durations (air/CO), or to
        air (air/air). Subjects were sacrificed at different ages and ventricular wet and dry weights
30      and myocyte volume and  number were measured histologically so that estimates of cell size
        and cell number could be made.  They observed increases in right ventricular weight due to

        March 12, 1990                         10-162     DRAFT-DO NOT QUOTE OR CITE

-------
TABLE 10-29. CONSEQUENCES OF PRENATAL CARBON MONOXIDE EXPOSURE ON CARDIOVASCULAR
&
jo
I— '
^£>
\O
O











^
<*


0
1
H

a
o
^
2
H
O
cj
O
w
._
M
r*»
UH V ULAJl'JVU:
Wet- Heart/
COHb Body Heart Body
Exposure (%) Weight Weight Weight
150 ppm 15 — Increased Increased
CO, CDs
1-21



230 ppm 24 Decreased ND Increased
CO, GD 24 PD 5
2-PD 21
60, 125, ND Decreased Increased Increased
250,
500 ppm
CO
157, 166, 24.9 Decreased Increased Increased
200 ppm ventricles
CO, CDs
5-22
200 ppm 27.8 Decreased Increased Increased
CO


30, 90 ppm 4.8-8.8 -
CO


200 ppm
CO from
GD 7-PD 28 ND - Increased Increased
at birth
GD 7-PD 1 Decreased Increased
at PD 28
1IN1 1IN J-.AJ5UKAHJKT KAld
Dry-
Heart Total
Weight Hematocrit Hemoglobin Other
Decreased ND ND Nucleic acid
protein
unchanged, no
significant
differences at
PDs 4-21
ND Increased Increased
PD 5 PD 5

ND Decreased, Decreased,
250-500 ppm 250-500 ppm


Increased — — Increased LDH M
subunit, increased
DNA content

ND - - No lasting effects
of prenatal
exposure

ND





ND ND Prenatal CO
increased myocytes
ND in right ventricle,
postnatal CO

References
Fechter et al.
(1980)




Hoffman and
Campbell
(1977)
Prigge and
Hochainer
(1977)

Penney et al.
(1983)


Penney et al.
(1980)


Garvey and
Longo
(1978)



Clubb et al.
(1986)



-------
TABLE 10-29 (cont'd). CONSEQUENCES OF PRENATAL CARBON MONOXIDE EXPOSURE ON CARDIOVASCULAR
                            DEVELOPMENT IN LABORATORY RATS
i










,_!
O
t— *
•^


O
"i>
3
j
i
0
0
^x
Q
H
O
COHb
Exposure (%)
PD 1-PD 28


500 ppm ND
CO PDs
1-32 (CO
gradually
increased
from
PDs 1-7)
300 and Approxi-
700 ppm mately
CO PDs 30 and
1-32 (CO 50%
gradually
increased
from
PDs 1-7)
500 ppm ND
CO PDs
1-32 (CO
gradually
increased
from
PDs 1-7)

Body
Weight
Decreased
at PD 21
and PD 28
—






Decreased
in 700-ppm
CO group
only in
adulthood



Decreased
only during
CO exposure





Wet-
Heart
Weight
Increased
at birth

Elevated in
adulthood





Increased
in 700-ppm
CO group
only in
adulthood



ND







Heart/
Body
Weight
Increased


Elevated in
adulthood





Increase
in 700-ppm
CO group
only in
adulthood



Increased
during
exposure





Dry-
Heart Total
Weight Hematocrit Hemoglobin Other References
ND increased myocytes
in left ventricle

ND Increase at ND Heart rate Penney et al.
some ages in elevated 10- (1984)
adulthood 15% in
adulthood.
No effect on
blood pressure

Decrease Increased ND No consistent Penney et al.
in dry/wet in females effect on heart (1988)
heart in adulthood rate at either
weight in exposure level.
females Elevated ventri-
cular DNA levels
in adulthood

ND ND ND Clubbetal.
(1989)







-------
!
cr
     TABLE 10-29 (cont'd). CONSEQUENCES OF PRENATAL CARBON MONOXIDE EXPOSURE ON CARDIOVASCULAR
                                DEVELOPMENT DSf LABORATORY RATS

com
Exposure (%)
500 ppm ND
CO PDs
1-32 (CO
gradually
increased
from
PDs 1-7)



Body
Weight
Decreased
during CO
exposure






Wet-
Heart
Weight
Increased
during CO
exposure






Heart/ Dry-
Body Heart
Weight Weight
Increased ND
during CO
exposure







Total
Hematocrit Hemoglobin Other
Increased ND Exercise in
during CO adulthood
exposure increased
atrium to body
weight ratio.
CO in neonatal
period
produced small
additional effect


References
Penney et al.
(1989)








-------
        fetal CO exposure, and increases in left ventricular weight following neonatal CO exposure.
        As in the case of Fechter et al. (1980), they showed a gradual return to normal heart weight
        when prenatally exposed neonates were placed in an air environment neonatally.  They
        attributed the reversal of cardiomegaly in the CO/air group to a loss in cell volume rather
 5      than a loss in cell number (which remains elevated).  Myocyte volume did not differ between
        CO and air subjects at birth. Left ventricle plus septum and right ventricle cell volumes of
        the CO/CO group were smaller than controls at 28 days of age despite the heavier wet heart
        weight shown by the CO/CO subjects.  Clubb et al. (1986) report that prenatal CO increased
        right ventricular  myocyte number and that neonatal CO exposure increased left ventricular
10      myocytes suggesting that cardiomegaly in early development is related to increased
        hemodynamic load. This possibility is supported by reports from Penney et al. (1983)
        showing that prenatal exposures to CO levels between 157 and 200 ppm during the last
        17 days of gestation did lead to a significant elevation in DNA content among  treated
        subjects. Moreover, hydroxyproline content, an indicator of collagen, also was increased
15      following the CO exposure as  was cardiac lactate dehydrogenase M (LDH M)  subunit among
        the 200-ppm CO subjects.
            Penney et al. (1980) compared the effects of prenatal CO exposure at a dose of 200 ppm
        with exposure both prenatally and neonatally until age 29 days.  Neonatal CO concentrations
        were elevated to  500 ppm.   Cardiomegaly and depressed hemoglobin, hematocrit, and red
20      blood cell counts were found following CO exposure.  In subjects allowed to survive until
        young adulthood, the heart weight to body weight ratio of subjects receiving CO  both
        prenatally and neonatally still was elevated, while those in the prenatal CO condition did not
        differ from control subjects  in this measure.  Penney and colleagues have published a series of
        papers that propose that the  neonatal period is a time when CO exposure might produce
25      persisting cardiovascular consequences. Typically their experimental protocol  involves
        exposure of rats to CO from soon after birth until PD 33.  CO levels are increased in step-
        wise fashion during the first week of neonatal life reaching the nominal CO exposure level by
        PD 7.  Spectrophotometric determination  of COHb levels were reported in one manuscript
        published by these authors to be approximately 30% for subjects exposed to peak CO values
30      of 350 ppm, 40% for subjects receiving 500 ppm, and 50% for  those exposed  to  700 ppm
        CO. (Penney et  al., 1988). The treatment produces significant reductions in body weight

        March 12, 1990                         10-166     DRAFT-DO NOT QUOTE OR CITE

-------
        (Penney et al., 1988), elevated hematocrit (Penney et al., 1988; Penney et al., 1989), and
        significant increases in heart weight (left ventricle plus septum and right ventricle) above
        control subjects at the end of CO exposure to 350 ppm (Penney et al., 1989).  The elevation
        in heart weight partially recovers as subjects mature although in some studies (e.g., Penney
 5      et al., 1984) persistent effects were observed into adulthood when neonatal CO levels were
        500 ppm.  In other studies, the elevation in heart weight or the heart weight:body weight  ratio
        resulting from 500-ppm CO exposure neonatally was no longer present in adulthood (e.g.,
        Clubb et al., 1989).  Penney et al. (1984) also suggested a 10-15%  increase in adult heart
        rates  associated with neonatal exposure to 500 ppm CO, but this effect was not replicated
10      using 350- and 750-ppm CO exposure (Penney et al., 1988).  Further, there was no evidence
        for an increase in blood pressure or other functional changes that might explain the
        tachycardia associated with 500-ppm CO exposure.  Finally,  a recent paper (Penney et al.,
        1989) suggested possible additive effects of neonatal exposure to 500 ppm CO and exercise-
        induced changes on adult heart size. Analysis of these effects are complicated by particularly
15      large effects of exercise on atrial weight rather than ventricular weight.
             To summarize, there is good evidence for the development of severe cardiomegaly
        following early life CO exposure at doses between 60 to 200 ppm.  These effects are
        transitory if exposure is prenatal and it is not clear whether they alter cardiac function or
        produce latent cardiovascular effects that may become overt later in life.  Persisting elevation
20      in heart weight results from combined prenatal CO exposure  at 200 ppm and neonatal
        exposure at 500 ppm.  The LOEL for this effect has not been determined.
             There are many published reports suggesting some residual increase in  heart weight
        associated with neonatal CO exposures of 500 ppm and greater, maintained over the first
        33 days of life. Even granting a small (10-15%) increase in  heart rate found in one study
25      among subjects exposed to 500 ppm CO neonatally, there is no evidence that neonatal CO
        exposure has functional consequences for experimental subjects.

        10.5.4.4  Neurobehavioral Consequences of Perinatal Carbon Monoxide Exposure
             The Developmental Toxicology Guidelines published by the U.S. Environmental
30      Protection Agency (Federal Register 1986) recognized the importance of neurobehavioral
        investigations as a means of assessing nervous system function.  Behavior is  an essential

        March 12, 1990                         10-167     DRAFT-DO NOT QUOTE OR CITE

-------
       function of the nervous system and abnormalities in this outcome can be diagnostic for
       particular neurological disorders or for nervous system dysfunctions.  The LOEL for such
       effects appears to be in the range of 125 to 150 ppm using a variety of behavioral tasks in
       experimental animals, though isolated studies suggest possible anomalies in the range of 60 to
 5     65 ppm.  These studies are summarized in Table 10-30.  There are a  limited number of
       human case reports that also are described here (see Table 10-31 for summary).  However,
       those reports generally are not adequate for evaluating a threshold for persisting
       neurobehavioral impairments in children.
            Crocker and Walker (1985) reported on the consequences of acute CO exposure in 28
10     children of which 16 had COHb levels over 15% and were considered to have had
       "potentially toxic" COHb levels.  These children were between the ages of 8  mo and 14
       years.  The authors report nausea, vomiting, headache, lethargy, and syncope to be the most
       common signs and symptoms. A very limited follow-up investigation was performed with
       these children and so no conclusions can be drawn from this work concerning persisting
15     effects. In addition to very large differences in the nature (dose and duration) of exposure,
       the extreme variability in patient age limits the potential value of the  data presented in this
       work.  The absence of any reports from children having COHb levels of 15% or less  (a very
       significant COHb level) is regrettable because these are the children one must study to
       develop an understanding of the relationship between dose and effect for the purpose of
20     setting standards for ambient air.
            Klees et al. (1985) conducted a more comprehensive study of the consequences of
       childhood CO exposure on subsequent behavioral development. They report that the age
       atwhich exposure occurred, its severity, and also the child's intellectual level at the time of
       exposure also play a role in the outcome.  Younger children tended to show somewhat milder
25     symptoms if they did recover fully than did children who were older  at the time of CO
       exposure.  Subjects who had higher intellectual function prior to accidental exposure also
       appeared to fare better  after CO exposure.  However, the authors stress that long-term
       perceptual and intellectual consequences of CO exposure may occur that are not identified
       well in short-term cursory examinations.  Of 14 children followed up for 2 to 11 years after
30     intoxication only 1 showed no sequelae (despite COHb levels of 42% on admission to hospital
       at the age of 9 years 10 mo). Seven children had impairment of visual memory and

       March 12, 1990                         10-168     DRAFT-DO NOT QUOTE OR CITE

-------
                TABLE 10-30.  NEUROBEHAVIORAL CONSEQUENCES OF PRENATAL CARBON MONOXIDE
                                                 EXPOSURE IN LABORATORY ANIMALS
Species (strain)
 Maternal/Neonatal Treatment
                                    Maternal/Embryonic Toxicity
                                                                                                Developmental Abnormality
                                                                              References
Rat
(Sprague-Dawley)
Rat
(Long-Evans)
Rat
(Long-Evans)
10,000 ppm CO for 2 or 3 h on
GD 15; no cross-fostering
150 ppm CO throughout
gestation; no cross-fostering
150 ppm CO throughout
gestations; no cross-
fostering
Acute effects: loss of righting
reflex followed by coma.  Litter
size normal; COHb levels of
approximately 50%

No difference in litter size or
fetal mortality; COHb levels
of 12.2-14%
Litter size normal; no
differences in neonatal
mortality; COHb levels of 15%
26% increase in exploratory
activity in figure-8 maze at PD 30
(3-h exposure)
3.3 % decreased birth weights and
decrease in preweaning weights;
decreased locomotor response to
L-dopa in open field (PD 4 and PD 14);
increased rate of habituation
(PD 14)

4.9% decreased birth weights and
decrease in preweaning weights;
decreased response to L-dopa (in
open field) at PD 1, PD 4 (also
decreased dopamine levels);
increase in rate of habituation
of activity (PD 14)
Daughtrey and Norton
 (1983)
Fechter and Annau
 (1976)
Fechter and Annau
 (1977)
Mouse
(Swiss-Webster)

Rat
(Long-Evans)




CO exposure throughout
gestation; no cross-
fostering
CO exposure throughout
gestation; no cross-
fostering



ND
Maternal COHb levels of
6-11%
Maternal weight gain,
gestation length, and
litter size normal;
COHb levels of 15.6%


Increased errors in heat-
motivated Y-nuze at PD 40

Decreased acquisition and (24-h)
retention of two-way active
avoidance (PD 30); neither multiple
measures nor use of pseudo-
conditioning controls similarly
affected
AbbatieDoandMohmann
(1979)

Mactutui and Fechter
(1984)





-------
I
                 TABLE 10-30 (cont'd).  NEUROBEHAVIORAL CONSEQUENCES OR PRENATAL CARBON MONOXIDE
                                                    EXPOSURE IN LABORATORY ANIMALS
       Species (strain)
                           Maternal/Neonatal Treatment
                                Maternal/Embryonic Toxic ity
                                   Developmental Abnormality
        Rat
        (Long-Evans)
150 ppm CO throughout
gestation; cross-
fostering for weight
measures
ND
(no COHb levels)
(PD 120), but minimal (24-h)
and pronounced (28-day)
decreased retention; decreased
acquisition and retention of
two-way avoidance (PDs 300-360)

7.6% decrease in birth weights
and decreased preweaning weights;
decreased negative geotaxis (PD 3);
decreased homing behavior (PDs 3-5)
                                                                                                                               References
Rat
(Long-Evans)
CO exposure throughout
gestation; no cross-
fostering
ND
COHb levels of 15.6%
Normal two-way avoidance
acquisition with moderate or
difficult task requirements
Mactutus and Fechter
(1985)
Fechter and Annau
 (1980)

-------
I
TABLE 10-31.   CONSIJQUHNCliS OF  HUMAN CARBON MONOXIDE INTOXICATION DURING
                                             HARLY DEVELOPMENT
        Characterization of
           Exposure1
              Approximate
               COHb Level
       Immediate
Symptoms and Their Frequency
    Persisting
  Symptoms and Their
Frequency After Hyperbaric
O2 and Normobaric O2 Therapy
                                                                                                                                                 References
%-^
£
o
o
2;
3
§
 g
 n
 a
        Light
        Medium
        Severe
        Accidental at 13 weeks
        of age

        Accidental at 21 days'
        old
         28 pediatric exposures
                                    4-27%
                                    6-36%
                                    37%
                                    60%
            6.7% 4 h after exposure
            (> 15 % at time of removal
            from CO)

            Threshold value at which
            symptom was first observed in
            any subject
             15%
             16.7%
             19.8%
             18.6%
            24.5%
            36.9%
                                           Hyperreflexia (1/3) auditive
                                           memory impairment and spatial
                                           orientation problems (1/3)

                                           ND (6/12)
                                           Coma (1/12)
                                           Unconscious (1/12)
                                           Normal (2/12)
Coma-developmental level
Regression (language and
motor)
Violent anger/nervousness
(1/1)

Convulsion, hypotonic,
unconscious (1/1)

Lethargy, vomiting (1/1)
                                                                    Asymptomatic
                                                                    Nausea/headache
                                                                    Vomiting
                                                                    Lethargy
                                                                    Visual symptoms/syncope
                                                                    Seizures
                                   Auditive and visual memory impairment
                                   (1/3)
Anxiety or emotional instability 0/12)
Memory impairment (2/12)
Spatial/temporal disorganization
and perceptual problems (3/12)
None or questionable effect (4/12)

Persistent emotional instability
(1/1)
                                                                               Recovery of minor neurologic
                                                                               deficits by 6 weeks (1/1)

                                                                               None
                                    Headaches, memory deficit, decline
                                    in school performance (3/28)
                                           Kleei et al. (1985)
                                                                              Kleet et «1. (1985)
                                                                                                                          Kleei et al. (1985)
                                            Venningetal.(1982)


                                            O'Sullivan (1983)
                                            Crocker and Walker
                                            (1985)

-------
o
           TABLE 10-31 (cont'd). CONSEQUENCES OF HUMAN CARBON MONOXIDE INTOXICATION DURING
                                      EARLY DEVELOPMENT







) .
0
5


o
jg

6
o
z
3
0
G
3
tn
O
K
n
H
W
Persisting
Symptoms and Their
Characterisation of Approximate Immediate Frequency After Hyperbaric
Exposure1 COHb Level Symptoms and Their Frequency O2 and Normobaric O2 Therapy
Light (6/14) 19-42% Somnolence (2/6) Perceptual (2/6)
Headache/nausea (3/6) Memory (3/6)
Emotional (1/16)
Psychomotor (1/16)
Cognitive (3/6)
Medium (5/14) 16-42% Incontinent (1/5) Perceptual (1/5)
Unconscious (4/5) Memory (1/5)
Emotional (2/5)
Cognitive (2/5)
Severe (3/14) ?-13% Vigil coma (3/3) Perceptual (2/3)
Cognitive (3/3)
Notes: Exposure duration and level of CO exposure are poorly defined in all studies. Exposures are grouped according to authors' descriptive characterization when
level. The latter varies widely with group.
See glossary of terms and symbols for abbreviations and acronyms.












Reference*
Klee* et al. (1985)




Klees et al. (1985)



Klees et al. (1985)

available rather than COHb












-------
        concentration, but normal IQ scores.  These children had "slight or medium" exposure
        (COHb levels in the low to mid-twenties) and no coma. The six children with serious
        learning disorders did not have more severe CO exposures as judged from thek COHb levels.
        They include several cases where exposures did occur at a young age and in children who had
 5      psychological difficulties prior to CO exposure. This study leaves some question concerning
        the relative vulnerability of children to CO as a function of their age as several of the
        youngest children did make full recovery while others did not.  It seems likely that the child's
        age may have an influence on the duration of CO exposure which is survivable and perhaps
        also on the promptness with which either hospitalization or measurement of COHb levels is
10      made.  Further study of the outcomes of childhood CO exposures will  be useful in
        determining whether there are differences with respect to vulnerability  to CO level.  Yenning
        et al. (1982) report on a case of acute CO poisoning in a 13-week-old baby who had
        profoundly elevated COHb levels (60% 2 h after removal from  the automobile in which she
        accidentally was exposed to CO). Her parents had much lower  COHb values though this may
15      reflect differences in concentration of CO inhaled. The child was reported to be unconscious
        for 48* h, to go through convulsions over the next 18 days, but,  again,  to show recovery from
        "minor neurological abnormalities"  by 6 weeks later.  There have been a series of
        experiments reported in rodents that identify both persistent neurobehavioral effects of
        prenatal CO exposure and also transient effects that may be symptomatic of functional delays
20      in development.  Fechter and Annau (1980) reported delays in  the development of negative
        geotaxis and homing in rats exposed prenatally to 150 ppm CO  (maternal COHb levels were
        not reported in this paper, but levels previously reported in this  laboratory under that
        exposure regimen are 15 to 17% [Fechter and Annau, 1977]).  These data were replicated by
        Singh (1986) using CD-I mice exposed from GD 7 to 18 to 0, 65, and 125 ppm CO.  He
25      found that exposure at 125 ppm significantly impaired the righting reflex on PD  1 and
        negative geotaxis on PD 10.  He also reported impaired aerial righting among subjects
        exposed prenatally to 65 or 125 ppm CO.  Morris et al. (1985a) studied the consequences of
        moderate CO exposure given very late in gestation. They exposed pigs to 200 and 250 ppm
        CO (COHb levels of 20 and 22%) from GD 109 until birth.  They found impairment of
30      negative-geotaxis behavior and open field activity 24 h after birth in pigs exposed to
        250 ppm.  Activity in the open field was significantly reduced in subjects exposed to both 200

        March 12, 1990                         10-173     DRAFT-DO NOT QUOTE OR CITE

-------
       and 250 ppm 48 h after birth.  The significance of these behavioral dysfunctions is that they
       point to delays in behavioral development that may themselves contribute to impairments in
       the way in which the individual interacts with its environment.
             There also are reports of impaired cognitive function produced by prenatal CO which
 5     may be related to permanent neurological damage. Mactutus and Fechter (1984) showed
       poorer acquisition and retention of a learned active avoidance task in rats of 30 to 31 days of
       age that had received continuous prenatal exposures of 150 ppm CO. This study is
       noteworthy because very careful efforts were made to distinguish cognitive deficits and
       performance deficits such as motivational, emotional, or motoric factors.  These findings
10     were replicated and extended by Mactutus and Fechter (1985). They studied the effects of
       prenatal exposures to 150 ppm CO (16% maternal COHb) on learning and retention in
       weanling, young adult, and aging  (1-year-old) rats. They found that both the weanling and
       young adult rats showed significant retention deficits, while in aging adults impairments were
       found in both learning and retention relative to control subjects. They interpreted these
15     results to mean that there are permanent neurological sequelae of prenatal CO exposure.
       They raise the important issue  that sensitivity of tests for consequences of early toxic exposure
       may reflect the developmental  status of the test subject and complexity of the task. In this
       case, a learning impairment not observed in the early adult period was detected by working
       with aged subjects.  No systematic attempts have been made to replicate these effects using
20     lower levels of CO.  One earlier study (Abbatiello and Mohrmann, 1979) suggested an
       increase in the number of errors made by mice prenatally exposed to CO throughout gestation
       (maternal COHb levels were 6 to  11%) and required to learn a maze discrimination task at
       6 weeks of age.  The absence of many details concerning the manner in which the control
       subjects were handled during pregnancy and the absence of details in the exposure protocol
25     make it difficult to draw firm conclusions from this paper.

       10.5.4.5  Neurochemical Consequences of Prenatal and Perinatal Carbon Monoxide
                 Exposure
             A significant number of studies have appeared concerning the consequence of acute and
30     chronic prenatal and perinatal CO exposure upon a variety of neurochemical parameters.
       These experiments are important because the transmission of information between nerve cells
       is based upon neurochemical processes.  Neurotransmitters can sometimes serve as markers
       March 12, 1990                         10-174     DRAFT-DO NOT QUOTE OR CITE

-------
        for the development of specific neurons in the brain, thereby serving as a sensitive alternative
        to histopathologic investigation particularly when the toxic agent selectively lesions neurons
        based upon a biochemical target.  The absence of a specific cell group identified by
        neurochemical methods may have important consequences for subsequent brain development
 5      because the absence of targets for synapse formation can have additional consequences on
        brain development. Altered neurochemical development has been observed at CO-exposure
        levels lower than those necessary to produce signs of maternal toxicity or gross  teratogenicity
        in the neonates.  Chronic prenatal and perinatal exposures to  150-300 ppm CO have been
        shown to yield persisting alterations in norepinephrine, serotonin, and GABA levels and in
10      GABA uptake in rats.  There also are a substantial number of acute exposure studies that have
        demonstrated  neurochemical effects of CO. However, these generally have been conducted at
        life-threatening levels and are not particularly relevant to setting ambient air standards for
        CO.
            Storm and Fechter (1985a) and Storm et al. (1986) have carefully studied the developing
15      cerebellum because this structure shows a rather slow developmental pattern and has been
        shown to be sensitive to hypoxic injury.  The cerebellum plays  prominent roles  in many
        diverse functions. It is a part of the extra-pyramidal motor system, and it plays an important
        role in maintaining balance.  The cerebellum also receives diverse sensory inputs and plays a
        role in sensory-motor integration.  The cerebellar cortex contains a diverse group of neurons
20      whose organization has been studied very well.  The intrinsic neurons of the cerebellum -
        those having their cell bodies and axonal processes within this structure - consist of the
        granule, pyramidal, stellate, basket, and Golgi cells.  The granule cells use the excitatory
        amino acid, glutamate,  as their neurotransmitter and synapse on the pyramidal cells.  The
        other intrinsic cells of the cerebellum appear to use the inhibitory amino acid, GABA, as their
25      neurotransmitter.  The Purkinje cells, being extremely large in  size, probably contribute
        considerably to the total GABA levels found in the cerebellum.
             The cerebellum receives several different neurotransmitter inputs from other brain
        regions. Most important of these is a noradrenergic input from the brainstem, a cholinergic
        link via mossy fibers and possibly aspartate or glutamate climbing fibers.
30           Storm and Fechter (1985a) reported that chronic prenatal  CO exposures of 150 ppm CO
        (approximately 16 to 18% COHb based upon other research in  this laboratory) decreased

        March 12, 1990                         10-175     DRAFT-DO NOT QUOTE OR CITE

-------
        cerebellar wet weight, but increased norepinephrine levels in this structure when expressed
        either in terms of concentration (nanograms per milligram wet weight) or total cerebellar
        content above control values between the ages of 14 to 42 days.  This period represented the
        duration of the experiment.  While this persisting elevation in norepinephrine cannot be
 5      considered permanent, it is the case that rats do obtain normal adult values of monoamine
        neurotransmitters at about the age of 40 to 45 days. There was no effect of CO treatment on
        norepinephrine levels in the cerebral cortex.  Since noradrenergic neurons have their cell
        bodies outside of the cerebellum and project axons  that terminate on cell bodies in this
        structure, Storm and Fechter's data may reflect an effect of increased noradrenergic
10      innervation secondary to toxic injury to target neurons in the cerebellum. Consistent with this
        hypothesis, Storm et al. (1986) reported deficits in  cerebellar weight,  but more importantly
        deficits in markers of GABA-ergic activity in the cerebellum following prenatal and perinatal
        CO exposures. GABA is thought to be an inhibitory neurotransmitter present in several
        neurons that are endogenous to the cerebellum.  Subjects in this experiment received 0, 75,
15      150,  and 300 ppm CO (corresponding maternal COHb levels were 2.5,  11.5, 18.5,  and
        26.8%) from the beginning of gestation until PD 10.  Neurochemical  measurements were
        made either on PD 10 or PD 21.  They showed reduced total GABA levels in the cerebellum
        following either 150- or 300-ppm CO exposure at both measurement times.  They also
        reported a significant reduction in total GABA uptake, but not glutamate uptake in
20      synaptosomes prepared from cerebella of 21-day-old neonates.  Glutamate is an excitatory
        neurotransmitter found within the cerebellum.  Histological markers of cerebellar toxicity also
        were obtained that were compatible with the neurochemical data and these are described
        below under histopathology.
             In a subsequent paper, Storm and Fechter (1985b) evaluated norepinephrine and
25      serotonin levels at PDs 21 and 42 in four different brain regions (pons/medulla, neocortex,
        hippocampus, and cerebellum) following chronic prenatal exposures to 75, 150, and 300 ppm
        CO.  They reported that norepinephrine and  serotonin concentrations decreased linearly with
        dose in the pons/medulla at 21 but not 42 days of age (i.e., evidence of a transient effect) the;
        LOEL was 150 ppm.  Norepinephrine increased linearly with CO dose in neocortex  at 42, but
30      not at 21 days of age.  They also showed that cerebellar weight was significantly reduced at
       March 12, 1990                          10-176    DRAFT-DO NOT QUOTE OR CITE

-------
        150 and 300 ppm CO when measured on PD 21 and for the 300-ppm-exposed rats at 42 days
        of age.

        10.5.4.6  Morphological Consequences of Acute Prenatal Carbon Monoxide
 5          Profound acute CO exposures do result in obvious neurological pathology that can be
        predicted  to be inconsistent with life or with normal neurological development. These data
        are not as relevant to setting standards for ambient air quality as they are in demonstrating
        both the danger of accidental high-level CO exposures and in providing possible insight into
        the susceptibility of the developing brain to toxic exposure. The one possible exception is a
10      study conducted in fetal pigs exposed via the mother to 300 ppm CO for 96 h (Dominick and
        Carson, 1983). The authors reported quite marked sensitivity to the CO as reflected in
        fetotoxicity, but also identified multifocal hemorrhages and vacuolation of the neuropile
        throughout the cortical white matter and brain stem.  They also observed cerebellar edema
        with swollen oligodendrocytes and astxocytes, two non-neuronal cell types that have important
15      roles in supporting neural function.
            Okeda et al. (1986) studied the effects of 2,000- to 3,000-ppm CO exposure for 76 to
        150 min in cats of different gestational ages.  They suggest a different pattern of neurological
        damage in cats exposed late versus those exposed early to the CO.  In late gestation, the
        primary changes were seen in cerebral white matter and brain stem. Basal ganglia and
20      thalamus were affected less and cerebral cortex even less affected.  Kittens exposed to CO at
        an early gestational age show less histopathology than those exposed later. Cerebral white
        matter and the basal ganglia tended to be most affected by early CO exposure.
            Daughtrey and Norton (1982) studied the effect of exposing pregnant rats (GD 15) to
        1,000 ppm CO exposure for 3 h upon central nervous system development of the fetuses on
25      GD 16. Estimated maternal COHb levels reached about 50%.  They reported 13 to 28%
        fetotoxicity (lethality) and described hemorrhagic infarcts and the most consistent damage to
        the ventral germinal matrix overlying the caudate nucleus.  Further study (Daughtrey and
        Norton, 1983) indicated damage to the dendritic branches of Golgi type II neurons in  the CO
        exposed fetuses.
30          Storm et al. (1986) showed that exposure of rats to 300 ppm CO (maternal COHb levels
        of 26.8%) throughout gestation  and until day 10 after gestation resulted in a noticeably

        March 12, 1990                         10-177    DRAFT-DO NOT QUOTE OR CITE

-------
       smaller cerebellum at PD 21. The cerebellum of exposed neonates had fewer fissures than
       normal controls.

       10.5.5  Summary
 5          The data reviewed provide strong evidence that CO exposures of 150 to 200 ppm
       produce reductions in birthweight, cardiomegaly, delays in behavioral development, and
       disruption in cognitive function in laboratory animals of several species. Isolated experiments
       suggest that some of these effects may be present at doses as low as 60 to 65 ppm maintained
       throughout gestation. The current data from human children suggesting a link between
10     environmental CO exposures and SIDS are weak, but further study should be encouraged.
       Human data from cases of accidental high dose CO exposures are difficult to use in
       identifying a LOEL or NOEL for this agent because of the small numbers of cases reviewed
       and problems in documenting levels of exposure. However, such data if systematically
       gathered and reported could be useful in identifying possible ages of special sensitivity to CO
15     and cofactors or other risk factors that might identify sensitive subpopulations.
       10.6  OTHER SYSTEMIC EFFECTS OF CARBON MONOXIDE
            Studies (see Table 10-32) reviewed in the previous criteria document (U.S.
20     Environmental Protection Agency,  1979) and again in Chapter 9 of this document suggest that
       enzyme metabolism and the P-450-mediated metabolism of xenobiotic compounds may be
       affected by CO exposure (Montgomery and Rubin, 1971; Kustov et al.,  1972; Pankow and
       Ponsold,  1972, 1974; Martynjuk and Dacenko,  1973; Swiecicki, 1973; Pankow et al., 1974;
       Roth and Rubin, 1976a,b).  Most of the authors have concluded, however, that effects on
25     metabolism at low COHb levels (<15%) are attributable entirely to tissue hypoxia produced
       by increased levels of COHb because they are no greater than the effects produced by
       comparable levels of hypoxic hypoxia.  At higher levels of exposure, where COHb
       concentrations exceed 15 to 20%, there may be direct inhibitory effects of CO on the activity
       of mixed-function oxidases but more basic research is needed (see Chapter 9, Section 9.4).
30     The decreases in xenobiotic metabolism shown with CO exposure might be important to


       March 12, 1990                        10-178    DRAFT-DO NOT QUOTE OR CITE

-------
TABLE 10-32. OTHER SYSTEMIC EFFECTS OF CARBON MONOXIDE
o
sr
J°
»—*
^o
v^5











0
i
>— »
-j
VO


O
£•
^Tl
>-3
I
O
o
z
o
H

d
O
0
n
H
w
Exposure*'1"

Increasing exposure
to 3000 ppm at
100 days
0.8 or 3.0%
until death
group

250, 500, and
1000 ppm for 24 h
Accidental
exposure
50 ppm

17 ppm


250-3000 ppm
for 90 min

50 ppm for
3 mo





Subcutaneous CO
at 7.2 and 9.6
mol/fcg; 40
injections in
S3 days





COIIb0 Subject(s)

ND Rat
n = 36-45

71.3% or Rat
79.2% n = 5 per
group

ND Rat
n = 36
3.4% to Man
32% n = 6
ND Rat
n = 92
ND Rat
n = 95

20-60% Rat
n = 10-20
per group
ND Rat
n = 100;
Rabbit
n = 40;
Dog
n = 4

50% Rat
n = 20-30








Observed Effects'1

Slightly few weight gains during first
100 days; increased weight gain in last
200 days
Increased plasma levels of leucine aminopepti-
dase. No change in state-4 respiration of
mitochondria; decreased state-3 rate in CO-
exposed rats
Decreased food and water intake; decreased
weight gain
Increased serum phosphocreatine-kinase

Decreased liver cytochrome oxidase; increased
liver succinate dehydrogenase
Increased aspartate and alanine amino
transferase activity

Prolonged response to hexobarbital at
1000 ppm and to zoxazolamine at 250 ppm

No effect on body weight






Increased leucine aminopeptidase
activity in the liver with single
and repeated injections; increased
liver weight with repeated injections






Conclusions

No significant body weight
effect

Acute CO poisoning caused
damage to liver mitochondria


Significant body weight
effect
Diffuse myolysis indicative
of acute renal failure
Tissue hypoxia

Tissue hypoxia


Decreased xenobiotic
metabolism

No significant body weight
effect















Reference

Campbell (1934)°


Katsumata et al.
(1980)


Koob et al.e
(1974)
Kuska et al.
(1980)
Kustov et al.e
(1972)
Martynjuck and0
Dacenko (1973)

Montgomery and0
Rubin (1971)

Musselman0
et al. (1959)





Pankow and
Ponsold (1972,
1974)e








-------
eren
Obse
     8
Pank
Ponso
Pank
(1974
ope
leucine aminp
the liver; enlarge
ethanol
CO exposure
combined with
1% ethanol
                c
                I
                 g
                  -s.s
.2
1
CO
o

a|
•B-s
li
X J3
£.a
og.
a£
^
"S
H
Dec
                  '
                « " S3
                A c a.
                1
e

I
1.
II
"SCO
Sa!?
18"
                      11
                      o£
ii
X *
Jjp-a
r; C
rate of
live
e h
11
                         •s
Q
                           J3

                           1
                           g

                           I§
                           e S
                      P5 It
                      ill Is
                           3*
                           "•5.
                       fe   -5

                           v.
                           £§£
                              '
                           s*
                           * — c
                           E M'"
                           3 C US
li*
o .S3 c
on  o
Q.S5  £
                           •Oc
                           li!
     •3 II  h
     o C  S.
     Q
                            IN S
                            -
                           o *o «->
                           **..>,
                           •o c .•=
                                  1
                                  I
                                  1
            e
            I
            'i
            •3 |
            Z u
             ,
            "8
            JO
                             s

                             8
                            •a

                            a;
rgic syste
I
5
i
                       n,
                       1
                                                  o
                        •o
                        1
                                .
                             -s ts
                                           S
                                         *• n
                                   00
                                  u II
                                  2 e
      § B I a   s
                             « 8
                              E
                                         e
                                         1
                    -
                   £
                        £
                        e
                                  0 0, T
                                  
-------
       individuals receiving treatment with drugs. The implications of this effect are discussed in
       Chapter 12, Section 12.3.
            The effects of CO on tissue metabolism noted above may partially explain the body
       weight changes associated with CO. Short-term exposure to 250-1000 ppm for 24 h was
 5     reported previously to cause weight loss in laboratory rats (Koob et al., 1974) but no
       significant body weight effects were reported in long-term exposure studies in laboratory
       animals at CO concentrations ranging from 50 ppm for 3 mo to 3000 ppm for 300 days
       (Theodore et al.,  1971; Musselman et al., 1959; Campbell, 1934; Stupfel and Bouley, 1970).
       It is quite probable that the initial hypoxic stress resulted in decreased weight gain followed
10     by compensation for the hypoxia with continued exposure by adaptive changes in the blood
       and circulatory system (see Section 10.3).  It is known,  however, that CO-induced hypoxia
       during gestation will cause a reduction in the birthweight of laboratory animals.  While a
       similar effect has been difficult to demonstrate  in humans exposed to CO alone, there is a
       strong correlation between maternal cigarette smoking and reduced birthweight.  (See
15     Section 10.5 for a more complete discussion of fetal  effects of CO exposure.)
            Inhalation of high levels of CO, leading to COHb concentrations greater than 10 to
       15%, have been reported to cause a number of systemic effects in laboratory animals as well
       as effects in humans suffering from acute CO poisoning. Tissues of highly active oxygen
       metabolism, such as heart, brain, liver, kidney, and muscle, may be particularly sensitive to
20     CO poisoning.  The impairment of function in  the heart and brain caused by CO exposure is
       well known and has been described in other sections  of this chapter.  Other systemic effects
       of CO poisoning are not as well known and are, therefore, less certain. There are reports in
       the literature (see Table 10.6-1) of  effects on liver (Katsumata et al., 1980), kidney (Kuska
       et al., 1980), and bone (Zebro et al., 1983).  Results from one additional study in adult
25     guinea pigs suggest that immune capacity in the lung and spleen was affected by intermittent
       exposure to high levels of CO for 3 to 4 weeks (Snella and  Rylander, 1979).   It generally is
       agreed that these systemic effects are caused by the severe tissue damage occurring during
       acute CO poisoning due to (1) ischemia resulting from the formation of COHb, (2) inhibition
       of O2 release from HbO2, (3) inhibition of cellular cytochrome function (e.g.,  cytochrome
30     oxidases), and (4) metabolic acidosis.
       March 12, 1990                         10-181     DRAFT-DO NOT QUOTE OR CITE

-------
            The effects of CO on visual acuity and dark adaptation, caused primarily by CNS
       alterations, were described previously (see Section 10.4). Besides central effects, however,
       acute CO exposure can cause damage at the sensory end-organ level.  Observed ocular effects
       from acute CO poisoning range from retinal hemorrhages (Dempsey et al., 1973; Kelley and
 5     Sophocleus, 1978) to blindness (Duncan and Gumpert, 1983; Katafuchi et al., 1985).  In
       addition, peripheral neuropathy and tortuous retinal vessels have been described after chronic,
       intermittent exposure to low  levels of CO over a 16-mo period (Trese et al., 1980). The
       authors of the latter report speculated  that increased blood flow from low-level, chronic
       exposure to CO may lead to  the development of a compensatory retinal vascular tortuosity.
10     With high-level, acute exposures to CO, the compensation will not take place and localized
       vascular hemorrhages result.
            Finally, exposure to CO has been associated with direct and indirect mutagenic activity
       (van Houdt et al.,  1987), as  measured by the Salmonella/microsome test (Ames et al., 1975).
       When tested in human populations living in the San Francisco Bay area, however, no
15     significant association was found between ambient levels of CO and site-specific cancer
       incidence (Selvin et al., 1980). A case-control study of persons with newly diagnosed
       multiple myeloma  living in four geographical locations of the United States found an
       increased risk for self-respondents reporting an exposure to combustion products including CO
       (Morris et al.,  1986). The difference in number of cases was  small when compared to
20     geographically matched controls and the exposure was not defined well. Giver; the limited
       information available at this time, it is unlikely that exposure to CO would contribute
       significantly to the development of cancer in nonoccupationally exposed individuals.
25     10.7 ADAPTATION, HABITUATION, AND COMPENSATORY
             RESPONSES TO CARBON MONOXIDE  EXPOSURE
            This section considers whether exposure to CO eventually will lead to the development
       of physiological responses that tend to offset some of the deleterious effects.  While there is
       possibly a temporal continuum in such processes, in this review the term "adaptation" will be
30     used to refer to long-term phenomena, and the term "habituation" will refer to short-term
       processes. Allusions will be made, where possible, to the physiological chain of events by

       March 12, 1990                         10-182     DRAFT-DO NOT QUOTE OR CITE

-------
       which adaptation and habituation come about, but extensive reductive explanations will be
       avoided. The term "compensatory mechanism" will be used to refer to those physiological
       responses that tend to ameliorate deleterious effects, whether in the long-term or short-term
       case.
 5
       10.7.1  Short-Term Habituation
            Arguments have been made for the possibility that there exist short-term compensatory
       mechanisms for CO exposure.  These hypothetical mechanisms have been (1) based upon
       physiological evidence, and (2) used to account for certain behavioral findings reported in the
10     literature.
            There is physiological evidence for responses that would compensate for the deleterious
       effects of CO in a very short time span. As discussed in Section 10.4,  CO exposure has been
       demonstrated to produce an increased cerebral blood flow which is apparently produced by
       cerebrovascular vasodilation.  It also has been shown (Doblar et al., 1977; Miller and Wood,
15     1974; Traystman, 1978; Zorn, 1972), however, that the tissue PO2 values for various CNS
       sites fall in proportion to COHb, despite the increased blood flow.  Apparently, the PO2
       values would fall considerably more without the increased blood flow.  Although the
       published graphs of these data do not  show very short time intervals, it appears that tissue PO2
       falls immediately and continuously as COHb rises.  Although there is no evidence  for time
20     delays or for threshold effects in these data, it is noteworthy that only very high CO levels
       were employed.  Thus, the saturation  rates were high, and time lags or thresholds would be
       difficult to detect.
            As discussed in Section 10.3, both coronary blood flow and O2 extraction in the
       peripheral musculature increase as COHb rises.  These, too, are compensatory mechanisms,
25     but mechanisms that have been shown to be only partly effective.  None of the studies present
       evidence of time lags or threshold effects,  because only terminal or near-asymptotic values
       were reported.
            The behavioral work upon which short-term habituation hypotheses have been predicated
       are mostly human studies,  where CO exposure at very low levels or at very early exposure
30     times (well before asymptotic saturation) have shown performance decrements that were not
       apparent with higher or longer exposures (Section 10.4).  Depending upon the particular

       March 12, 1990                         10-183     DRAFT-DO NOT QUOTE OR CITE

-------
       version, the habituation hypothesis holds that there might exist some threshold value of CO
       below which no compensation would be initiated, or that there might be some time lag in the
       compensatory mechanism so that the early effects of CO exposures would later subside.  The
       behavioral data to support this contention have been controversial and need to be examined
 5     closely.
             Most of the hypotheses about compensatory mechanisms were based, however, upon
       post hoc reasoning to explain empirical findings, not upon results form experiments to test the
       existence or nature of such mechanisms.  Disregarding hypothesized time lags and thresholds,
       without the compensatory mechanisms CO would apparently have even more deleterious
10     effects and the threshold  for such effects would be lower.

       10.7.2  Long-Term Adaptation
             Adaptation is an all-inclusive term that incorporates all of the acute or chronic
       adjustments of an organism to a stressor.  It does not indicate (or predict) whether the
15     adjustments are initially or eventually beneficial or detrimental.  Acclimatization is an
       adaptive process that results in reduction of the physiological strain produced by exposure to a
       stressor.  Generally, the  main effect of repeated, constant exposure to the stressor is
       considered to result in  an improvement of performance or a reduced physiological cost.  Both
       of these phenomenon tend to  exploit the reserve potential of the organism.
20           It generally is agreed that adaptation to lowered levels of oxygen tension and oxygen-
       carrying capacity can occur with continued hypoxic exposure. This is evident especially in
       healthy individuals living for lengthy periods of time at high  terrestrial altitudes. It should be
       noted, however, that there is  no assurance that individuals moving from low altitudes to
       higher ones will attain the physiological status to the higher altitudes that is observed in
25     natives (viz. natives of the Andes and Himalayas).  Prominent features of prolonged altitude
       exposures are increases in hemoglobin concentration and hematocrit.  Additional alterations
       are right ventricular hypertrophy, pulmonary artery vasoconstriction, possible changes in
       cardiac output, and increased blood volume due to increases in the red cell mass.
             Whether or not adaptation can occur in individuals chronically exposed to various
30     ambient concentrations of CO remains unresolved. Concern  for CO intoxication in England
       and Scandinavia led to the speculation that adaptational adjustments could occur in man (Grut,

       March 12, 1990                          10-184     DRAFT-DO NOT QUOTE OR CITE

-------
        1949; Killick, 1940).  These concerns were directed to situations where high ambient CO
        concentrations were present.  There are only a few available studies conducted in humans.
            Killick (1940), using herself as a subject, reported that she developed acclimatization as
        evidenced by diminished symptoms,  slower heart rate, and the attainment of a lower COHb
 5      equilibrium level following exposure to a given inspired CO concentration. Interestingly,
        Haldane and Priestly (1935) already had reported a similar finding as to the attainment of a
        different COHb equilibrium following exposure, to a fixed level of CO in the ambient air.
            Killick (1948) repeated her CO-exposure studies in an attempt to obtain  more precise
        estimations of the acclimatization effects she had noted previously.  The degree of
10      acclimatization was indicated by (1) a diminution in severity of symptoms during successive
        exposure to the same concentrations of CO, and (2) a lower COHb level after acclimatization
        than that obtained prior to acclimatization during exposure to the same concentrations of
        inhaled CO.
            Before using herself as a subject, Killick (1937) studied the effects of CO on laboratory
15      animals. Mice were exposed to successively higher concentrations of CO, which in a period
        of 6 to  15 weeks reached levels of 2300 to 3275 mg/m3 (2000 to 2850 ppm) CO and produced
        60 to 70%  COHb.  The nonadapted mice exhibited much more extreme symptoms when
        exposed to such levels. A control group was used to partially rule out effects of selection of
        CO-resistant animals.
20          Clark and Otis (1952) exposed  mice to gradually increasing CO levels over a period of
        14 days until a level of 1380 mg/m3 (1200 ppm) was reached. When exposed to a simulated
        altitude of 34,000 ft, survival of the  CO-adapted groups was much greater than controls.
        Similarly, Clark and Otis (1952) acclimatized mice to a simulated altitude of  18,000 ft and
        showed that these altitude-adapted mice survived 2875 mg/m3 (2500 ppm) CO better than
25      controls. Wilks et al. (1959) reported similar effects in dogs.
            Gorbatow and Noro (1948) showed that rats given successive daily short-term exposures
        could tolerate, without loss of consciousness, longer and longer exposures. Their CO-
        exposure levels were 2875 to 11,500 mg/m3  (2000 to 10,000 ppm). Increases in tolerance to
        CO began to be evident as early as the fourth or fifth day of exposure and still were occurring
30      as late as the 47th day. Nonexposure for several days eliminated some of the adaptation.
        Similar results were reported by Zebro et al. (1976).
                *
        March 12,  1990                         10-185    DRAFT-DO NOT QUOTE OR CITE

-------
            Chronic CO exposure of rats increases hemoglobin concentration, hematocrit, and
       erythrocyte counts via. erythropoietin production (see Section 10.3.4). Penney et al. (1974)
       concluded that the threshold for the erythropoietin response was 100 ppm (9.26% COHb).
       Cardiac enlargement, involving the entire heart during CO exposure (compared to right
 5     ventricular hypertrophy with high-altitude exposure), is induced when ambient CO is near
       200 ppm, producing COHb levels of 15.8% (Penney et al., 1974). Blood volume of the rat
       exposed for 7.5 weeks to CO exposures peaking at 1300 ppm nearly doubled while
       erythrocyte mass more than tripled (Penney et al., 1988).  After 42 days of continuous
       exposure to 500 ppm, rat blood volume almost doubled, primarily as a consequence of
10     increases in erythrocytes (Davidson and Penney, 1988).  It should be noted that all the
       demonstrated effects on tissues and fluids are induced by long-term exposures to high CO
       concentrations.  McGrath (1989) exposed rats for 6 weeks to altitudes ranging from 3,300 ft
       (ambient) to 18,000 ft and to concentrations of CO ranging from 0 to 500 ppm. At 9 and
       35 ppm CO, where COHb levels ranged from 0.9 to 3.3%, there were no significant changes
15     in body weight, right ventricular weight, hematocrit, or hemoglobin. Small but
       nonsignificant changes in these variables were measured when the CO concentration was
       100 ppm and COHb levels ranged from 9.4 to 10.2%. This is consistent with the
       observations noted above (Penney et al., 1974) - that the threshold for erythropoietin effects
       was 100 ppm.
20          Besides the level of exposure, the time course of exposure to CO also is important.  As
       discussed in Section 10.3.4, Hb increases in laboratory animals exposed to CO after about
       48 h, and continues to increase in the course of continued exposure until about 30 days,
       depending perhaps upon exposure level. This hemopoietic response to long-term CO
       exposure is similar to that shown for long-term hypoxic hypoxia, except that it is slower to
25     start and tends to  offset CO hypoxic  effects.
            Most investigators have at least implied that increased Hb level is the mechanism by
       which adaptation occurs.  Certainly this explanation is reasonable for the studies showing
       increased survival in  groups adapted for several days. Little has been done, however, to
       elucidate the extent to which such increases offset the deleterious effects of CO. The
30     probability that some adaptation occurs is supported theoretically due to Hb increases, and
       March 12, 1990                         10-186    DRAFT-DO NOT QUOTE OR CITE

-------
        empirically in the findings of laboratory animal studies measuring survival time.  But
        adaptation has not been demonstrated for specific health effects other than survival time.
             Compensatory increases in Hb are not without deleterious consequences of their own,
        such as cardiac hypertrophy (see Section 10.3.4). The Hb increases also are not entirely
 5     compensatory at all CO levels in view of the fact that deleterious effects still occur at some
        CO levels for many physiological systems.  It is possible, however, that without such
        mechanisms as Hb increases, CO effects would be worse or would occur at lower exposure
        concentrations.

 10     10.7.3  Summary
             The only evidence for short- or long-term COHb compensation in man is indirect.
        Experimental animal data indicate that COHb levels produce physiological responses that tend
        to offset other deleterious effects of CO exposure. Such responses are (1) increased coronary
        blood flow, (2) increased cerebral blood flow, (3) increased hemoglobin through increased
 15     hemopoiesis,  and (4) increased O2 consumption in muscle.
             Short-term compensatory responses in blood flow or O2 consumption may not be
        complete or might even be lacking in certain persons.  For example, from laboratory animal
        studies it is known  that coronary blood flow is increased with COHb, and from human
        clinical studies it is known that subjects with ischemic heart disease respond to the lowest
 20     levels of COHb (5%, or less).  The implication is that in some cases of cardiac impairment,
        the short-term compensatory mechanism is impaired.
             From neurobehavioral studies, it is apparent that decrements due to CO have not
        occurred consistently in all subjects, or even in the same studies, and have not demonstrated a
        dose-response relationship with increasing COHb  levels. The implication from this data
25      suggests that there might be some threshold or time lag in a compensatory mechanism such as
        increased cerebral blood flow.  Without direct physiological evidence in either laboratory
        animals or, preferably humans, this concept only can be hypothesized. The observed results
        from the neurobehavioral studies could be explained by differences or problems in
        experimental protocols or due to possible nonrandom sampling.
30           The idea of a threshold or a time lag in compensatory mechanisms should not be
        rejected entirely, however. There simply is no direct evidence. Studies need  to be performed

        March 12, 1990                          10-187    DRAFT-DO NOT QUOTE OR CITE

-------
        to (1) measure cerebral blood flow and tissue PO2 with low COHb levels at various ambient
        concentrations of CO to determine early and low level effects accurately, and (2) design
        behavioral studies where threshold effects or time lags are factors in the experimental
        protocols that can be explicitly studied.
 5           The mechanism by which long-term adaptation would occur, if it could be demonstrated
        in humans, is assumed to be an increased Hb concentration via a several-day increase in
        hemopoiesis.  This alteration in Hb production has been demonstrated repeatedly in animal
        studies but no recent studies have been conducted indicating or suggesting that some
        adaptational benefit has or would occur.  Furthermore, even if the Hb increase is a signature
10      of adaptation, it has not been demonstrated to occur at low ambient concentrations of CO.
        The human studies of the 1940s have not been replicated, so the question of adaptation
        remains  unresolved.
15      10.8  REFERENCES

       Abbatiello, E. R.; Mohrmann, K. (1979) Effects on the offspring of chronic low exposure carbon monoxide
              during mice pregnancy. Clin. Toxicol. 14: 401-406.
20     Abramson, E.; Heyman, T. (1944) Dark adaptation and inhalation of carbon monoxide. Acta Physiol. Scand.
              7: 303-305.
       Adams, J. D.; Erickson, H. H.; Stone, H. L. (1973) Myocardial metabolism during exposure to carbon
              monoxide in the conscious dog. J. Appl. Physiol. 34: 238-242.
25
       Adams, K. F.; Koch, G.; Chatterjee, B.; Goldstein, G. M.;  O'Neil, J. J.; Bromberg, P. A.; Sheps, D. S.;
              McAllister, S.; Price, C. J.; Bissette, J. (1988) Acute elevation of blood carboxyhemoglobin to 6%
              impairs exercise performance and aggravates symptoms in patients with ischemic heart disease. J. Am.
              Coll. Cardiol. 12: 900-909.
30
       Ahmad, I.; Ahmad, J. (1980) Environmental carbon monoxide and hypertension. J. Tenn. Med. Assoc.
              73: 563-566.

       Albaum, H. G.; Tepperman, J.; Bodansky, O. (1946) The in vivo inactivation by cyanide of brain cytochrome
35            oxidase and its effect on glycolysis and on the high energy phosphorus compounds in brain. J. Biol.
              Chem. 164: 45-51.

       Alcindor, L. G.; Belegaud, J.; Aalam, H.; Piot, M. C.; Heraud, C.; Boudene, C. (1984) Teneur en cholesterol
              des lipoproteines legeres chez le lapin hypercholesterolemique  intoxique ou non par 1'oxyde de carbone
40            [Low-density lipoprotein levels in hypercholesterolemic rabbits intoxicated or not by carbon monoxide].
              Ann. Nutr. Metab. 28: 117-122.
         March 12, 1990                           10-188      DRAFT-DO NOT QUOTE OR CITE

-------
       Alderman, B. W.; Baron, A. E.; Savitz, D. A. (1987) Maternal exposure to neighborhood carbon monoxide and
              risk of low infant birth weight. Public Health Rep. 102: 410-414.

       Aliukhin, Y.; Antonov, L. M.; Gasteva, S. V. (1974) Oxygen consumption in the brain in histotoxic hypoxia.
 5            Sechonov Physiol. J. USSR 60: 1376-1381.

       Allied, E. N.; Bleecker, E. R.; Chaitman, B. R.; Dahms, T. E.; Gottlieb, S.  O.; Hackney, J. D.; Pagano, M.;
              Selvester, R. H.; Walden, S. M.; Warren, J. (1989a) Short-term effects of carbon monoxide exposure on
              the exercise performance of subjects with coronary artery disease. N. Engl. J. Med. 321: 1426-1432.
10
       Allred, E. N.; Bleecker, E. R.; Chairman, B. R.; Dahms, T. E.; Gottlieb, S. O.; Hackney, J. D.; Hayes, D.;
              Pagano, M. Selvester, R. H.; Walden, S. M.; Warren, J. (1989b) Acute effects of carbon monoxide
              exposure on individuals with coronary artery disease. Cambridge, MA: Health Effects Institute; research
              report no. 25.
15
       Ames, B. N.; McCann,  J.; Yamasaki, E. (1975) Methods for detecting carcinogens and mutagens with the
              Salmonella/mammalian-microsome mutagenicity test. Mutat. Res. 31:  347-363.

       Anders, M.  W.; Sunram, J. M. (1982) Transplacental passage of dichloromethane and carbon monoxide. Toxicol.
20            Lett. 12: 231-234.

       Anderson, E. W.;  Andelman, R. J.; Strauch, J.  M.; Fortuin, N. J.; Knelson, J. H. (1973) Effect of low-level
              carbon monoxide exposure on onset and duration of angina pectoris: a study in ten patients with ischemic
              heart disease. Ann, Intern. Med. 79: 46-50.
25
       Annau, Z. (1974) The comparative effects of hypoxic and carbon monoxide hypoxia on behavior. In: Weiss, B.;
               Laties, V.  G., eds. Behavioral toxicology. New York, NY: Plenum Publishing; pp.  105-127.

       Annau, Z. (1975) The comparative effects of hypoxic and carbon monoxide hypoxia on behavior.  In: Weiss, B.;
30            Laties, V.  G., eds. Behavioral toxicology. New York, NY: Plenum Press; pp. 151-155.

       Annau, Z.;  Dyer,  R. S. (1977) Effects of environmental temperature upon body temperature in the hypoxic rat.
               Fed. Proc.  Fed. Am. Soc. Exp. Biol. 36: 599.

35     Armitage, A. K.; Davies, R. F.; Turner, D. M. (1976) The effects of carbon monoxide on  the development of
               atherosclerosis in the White Carneau pigeon. Atherosclerosis 23: 333-344.

        Aronow, W. S.  (1981)  Aggravation of angina pectoris by two percent carboxyhemoglobin. Am. Heart J. 101:
               154-157.
40
        Aronow, W. S.; Cassidy, J. (1975) Effect of carbon monoxide on maximal treadmill exercise: a study in normal
               persons. Ann. Intern. Med. 83: 496-499.

        Aronow, W. S.; Isbell, M. W. (1973) Carbon monoxide effect on exercise-induced angina pectoris. Ann. Intern.
45            Med. 79: 392-395.

        Aronow, W. S.; Harris, C. N.; Isbell, M. W.; Rokaw, S. N.; Imparato, B.  (1972) Effect of freeway travel on
               angina pectoris. Ann. Intern. Med. 77: 669-676.

 50     Aronow, W. S.; Ferlinz, J.; Glauser, F. (1977) Effect of carbon monoxide on exercise performance in chronic
               obstructive pulmonary disease. Am. J.  Med. 63: 904-908.

        Aronow, W. S.; Stemmer, E. A.; Wood, B.; Zweig, S.; Tsao, K.-P.; Raggio, L. (1978) Carbon monoxide and
               ventricular fibrillation threshold in dogs with acute myocardial injury. Am. Heart J. 95: 754-756.


          March 12, 1990                              10-189     DRAFT-DO NOT QUOTE  OR CITE

-------
        Aronow, W. S.; Stemmer, E. A.; Zweig, S. (1979) Carbon monoxide and ventricular fibrillation threshold in
               normal dogs. Arch. Environ. Health 34: 184-186.

        Aronow, W. S.; Schlueter, W. J.; Williams, M. A.; Petratis, M.; Sketch, M. H. (1984) Aggravation of exercise
 5             performance in patients with anemia by 3%  carboxyhemoglobin. Environ. Res. 35: 394-398.

        Astrup, J. (1982) Energy-requiring cell functions in the ischemic brain. J. Neurosurg. 56: 482-497.

        Astrup, P.;  Kjeldsen, K.; Wanstrup, J. (1967) Enhancing influence of carbon monoxide on the development of
10             atheromatosis in cholesterol-fed rabbits. J. Atheroscler. Res. 7: 343-354.

        Astrup, P.;  Olsen, H. M.; Trolle, D.;  Kjeldsen, K.  (1972) Effect of moderate carbon-monoxide exposure on fetal
               development. Lancet (0000): 1220-1222.

15      Atkins, E. H.; Baker, E. L.  (1985) Exacerbation of coronary artery disease by occupational carbon monoxide
               exposure: a report of two fatalities and a review of the literature. Am. J. Ind. Med. 7: 73-79.

        Ator, N. A. (1982) Modulation of the behavioral effects of carbon monoxide by reinforcement contingencies.
               Neurobehav. Toxicol. Teratol.  4: 51-61.
20
        Ator, N. A.; Merigan, W. H., Jr.; Mclntire, R. W.  (1976) The effects of brief exposures to carbon monoxide on
               temporally differentiated  responding. Environ. Res. 12: 81-91.

        Ayres, S. M.; Mueller, H. S.; Gregory,  J. J.; Giannelli, S., Jr.; Penny, J. L. (1969) Systemic and myocardial
25             hemodynamic responses to relatively small concentrations of carboxyhemoglobin (COHB).  Arch.
               Environ. Health 18:  699-709.

        Ayres, S. M.; Giannelli,  S., Jr.;  Mueller, H. (1970) Myocardial and systemic responses to carboxyhemoglobin.
               In: Coburn, R. F., ed. Biological effects of carbon monoxide. Ann.  N.  Y.  Acad. Sci.  174: 268-293.
30
        Ayres, S. M.; Evans,  R.  G.; Buehler, M. E.  (1979) The prevalence of carboxyhemoglobinemia in New Yorkers
               and its effects on the coronary and systemic circulation. Prev. Med.  8: 323-332.

        Balraj, E. K. (1984) Atherosclerotic coronary artery disease and "low" levels of carboxyhemoglobin; report of
35             fatalities and discussion of pathophysiologic mechanisms of death. J. Forensic Sci. 29: 1150-1159.

        Beard, R. R.; Grandstaff, N. (1970) Carbon monoxide exposure and cerebral function. Ann. N. Y. Acad. Sci.
               174: 385-394.

40      Beard, R. R.; Wertheim, G. A. (1967) Behavioral impairment associated with small doses of carbon monoxide.
               Am. J. Public Health 57: 2012-2022.

        Becker, L. C.; Haak, E. D., Jr.  (1979) Augmentation of myocardial ischemia by low level carbon monoxide
               exposure in dogs. Arch.  Environ. Health 34: 274-279.
45
        Bender, W.; Goethert, M.; Malorny, G.  (1972) Effect of low carbon monoxide concentrations on psychological
               functions. Staub Reinhalt. Luft 32: 54-60.

        Benignus, V. A. (1984) EEG as  a cross species indicator of neurotoxicology. Neurobehav, Toxicol. Teratol.
50             6: 473-483.

        Benignus, V. A.; Otto, D. A.; Prah, J. D.; Benignus, G. (1977) Lack of effects of carbon monoxide on human
               vigilance. Percept. Mot.  Skills 45: 1007-1014.
         March 12, 1990                              10-190      DRAFT-DO NOT QUOTE OR CITE

-------
       Benignus, V. A.; Muller, K. E.; Barton, C. N.; Prah, J. D. (1987) Effect of low level carbon monoxide on
               compensatory tracking and event monitoring. Neurotoxicol. Teratol. 9: 227-234.

       Benignus, V. A.; Muller, K. E.; Pieper, K. S.; Prah, J. D. (1989a) Compensatory tracking in humans with
 5             elevated carboxyhemoglobin. Neurotoxicol. Teratol.: in press.

       Benignus, V. A.; Muller, K. E.; Malott, C. M.  (1989b) Dose-effects functions for carboxyhemoglobin and
               behavior. Neurotoxicol. Teratol.: in press.

10     Beme, R. M.; Rubio, R.; Curnish, R. R.  (1974) Release of adenosine from ischemic brain, effect on cerebral
               vascular resistance and incorporation into cerebral adenine nucleotides. Circ. Res. 35: 262-271.

       Berntman, L.; Carlsson, C.; Siesjo, B. K. (1979) Cerebral oxygen consumption and blood flow in hypoxia:
               influence of sympathoadrenal activation.  Stroke (Dallas)  10: 20-25.
15
       Betz, E. (1972) Cerebral blood flow: its measurement and regulation. Physiol. Rev. 52: 595-630.

       Bicher, H. I.; Bruley, D. F.; Reneau, D. D.; Kinsley, M.  H. (1973) Autoregulation of oxygen supply to micro
               areas of brain tissue under hypoxic and hyperbaric conditions. Bibl.  Anat. 11: 526-531.
20
       Bing, R. J.; Sarma, J. S.; Weishaar, R.; Rackl, A.; Pawlik, G.  (1980) Biochemical and histological effects of
               intermittent carbon monoxide exposure in cynomolgus monkeys (Macaco fascicularis) in relation to
               atherosclerosis. J. Clin. Pharmacol. 20: 487-499.

25     Bissette, J.; Carr, G.; Koch, G. G.; Adams, K. F.; Sheps, D. S. (1986) Analysis of (events/time at risk) ratios
               from two-period crossover studies. In: American Statistical Association 1986 proceedings of the
               Biopharmaceutical Section; August; Chicago, IL. Washington, DC: American Statistical Association;
               pp.  104-108.

30     Bissonnette, J. M.; Wickham, W. K. (1977) Placental diffusing  capacity for carbon monoxide in unanesthetized
               guinea pigs. Respir. Physiol. 31: 161-168.

       Borgstrom, L.; Johannsson, H.; Siesjo, B. K. (1975) The relationship between arterial PQ and cerebral blood
               flow in hypoxic hypoxia. Acta Physiol. Scand. 93: 423-432.
35
       Brierly, J. B. (1975) Comparison between effects of profound arterial hypotension and cyanide on the brain of
               macaca mulatta. Adv. Neurol.  10: 213-221.

       Brierly, J. B.; Brown, A. W.; Calverly, J. (1976) Cyanide intoxication in the rat: physiological and
40             neuropathological aspects. J. Neurol. Neurosurg. Psych. 39: 129-140.

       Brinkhous, K. M. (1977) Effects of low level carbon monoxide exposure: blood lipids and coagulation
               parameters. Research Triangle Park, NC: U.  S. Environmental Protection Agency, Health Effects
               Research Laboratory;  EPA report no. EPA-600/1-77-032. Available from: NTIS,  Springfield, VA;
45             PB-269340.

       Brune, B.; Ullrich, V.  (1987) Inhibition of platelet aggregation by carbon monoxide is mediated by activation of
               guanylate cyclase. Mol. Pharmacol. 32: 497-504.

50     Buchwald, H. (1969) A rapid and sensitive method for estimating carbon monoxide in blood and its application in
               problem areas.  Am. Ind. Hyg. Assoc. J. 30:  564-569.

       Bunnell, D. E.; Horvath, S. M. (1988) Interactive effects of physical work and carbon monoxide on cognitive
               task performance. Aviat. Space Environ. Med. 59: 1133-1138.


         March 12, 1990                              10-191      DRAFT-DO NOT QUOTE OR CITE

-------
        Bureau, M. A.; Monette, J.; Shapcott, D.; Pare, C.; Mathieu, J.-L.; Lippe, J.; Blovin, D.; Berthiaume, Y.;
               Begin, R. (1982) Carboxyhemoglobin concentration in fetal cord blood and in blood of mothers who
               smoked during labor. Pediatrics 69: 371-373.

 5     Burgess, D. W.; Bean, J. W. (1970) The effect of neuronal activity on local cerebral PO2 and blood flow in the
               mammal.  In: Russell, R. W. R., ed. Brain and blood flow. London, United Kingdom: Pittman Medical
               & Scientific Publishing Co., Ltd.; pp.  120-124.

        Bums, T. R.; Greenberg, S. D.; Cartwright, J.; Jachimczyk, J. A. (1986) Smoke inhalation: an ultrastructural
10            study of reaction to injury in the human alveolar wall. Environ. Res. 41: 447-457.

        Calverley, P. M.  A.; Leggett, R. J. E.; Flenley, D. C. (1981) Carbon monoxide and exercise tolerance in
               chronic bronchitis and emphysema. Br.  Med. J. 283: 878-880.

15     Campbell, J. A. (1934) Growth, fertility, etc. in animals during attempted acclimatization to carbon monoxide.
               Exp. Physiol. 24: 271-281.

        Cavazzuti,  M.; Duffy,  T. E. (1982) Regulation of local cerebral blood flow in normal and hypoxic newborn
               dogs. Ann. Neurol. 11:  247-257.
20
        Celsing, F.;  Sverdenhag, J.; Pihlstedt, P.; Ekblom,  B. (1987) Effects of anemia and stepwise-induced
               polycythemia on maximal aerobic power in individuals with high and low haemoglobin concentration.
               Acta Physiol. Scand.  129: 47-54.

25     Chance, B.; Cohen, P.; Jobsis, F.; Schoener, B. (1962) Intracellular oxidation-reduction states in vivo. Science
               (Washington, DC) 137:  499-508.

        Chance, B.; Erecinska, M.; Wagner, M. (1970) Mitochondrial responses to carbon monoxide toxicity. In:
               Coburn, R.  F., ed.  Biological effects of carbon monoxide. Ann. N. Y. Acad. Sci. 174: 193-204.
30
        Chapman, R. W.; Santiago, T. V.; Edelman, N. H. (1980) Brain hypoxia and control of breathing:
               neuromechanical control. J. Appl. Physiol.:  Respir. Environ. Exercise Physiol.  49: 497-505.

        Chen, S.; Weller, M. A.; Penney, D. G. (1982) A study of free lung cells from young  rats chronically exposed
35            to carbon monoxide from birth. Scanning Electron Microsc. 2: 859-867.

        Chevalier, R. B.; Krumholz, R. A.; Ross, J. C. (1966) Reaction of nonsmokers to carbon monoxide inhalation:
               cardiopulmonary responses at rest and during exercise. JAMA J. Am. Med. Assoc. 198: 1061-1064.

40     Chiodi, H.; Dill,  D. B.; Consolazio, F.; Horvath, S. M. (1941) Respiratory and circulatory responses to acute
               carbon monoxide poisoning. Am. J. Physiol. 134:  683-693.

        Choi, K. D.; Oh, Y. K. (1975)  A teratological study on the effects of carbon monoxide exposure upon the fetal
               development of albino rats. Chungang Uihak 29: 209-212.
45
        Christensen, C. L.;  Gliner, J. A.; Horvath, S. M.; Wagner, J. A.  (1977) Effects of three kinds of hypoxias on
               vigilance  performance, Aviat. Space Environ.  Med. 48: 491-496.

        Christensen, P.; Gronlund, J.; Carter, A. M. (1986) Placental gas exchange in the guinea-pig: fetal blood gas
50            tensions following the reduction of maternal oxygen capacity with  carbon monoxide. J. Dev. Physiol.
               8:  1-9.

        Clark, R. T., Jr.; Otis, A. B. (1952) Comparative studies on acclimatization of mice to carbon monoxide and to
               low oxygen. Am. J.  Physiol. 169: 285-294.


         March 12, 1990                              10-192      DRAFT-DO NOT QUOTE OR CITE

-------
        Clubb, J. J., Jr.; Penney, D. G.; Baylerian, M. S.; Bishop, S. P. (1986) Cardiomegaly due to myocyte
               hyperplasia in perinatal rats exposed to 200 ppm carbon monoxide. J. Mol. Cell. Cardiol.  18: 477-486.

        Clubb, F. J., Jr.; Penney, D. G.; Bishop, S. P. (1989) Cardiomegaly in neonatal rats exposed to 500 ppm carbon
  5            monoxide. J. Mol. Cell. Cardiol. 21: 945-955.

        Cobum, R. F.  (1977) Oxygen tension sensors in vascular smooth muscle.  In: Reivich, M.; Coburn, R. F.;
               Lahiri, S.; Chance, B., eds. Tissue hypoxia and ischemia. Adv. Exp. Med. Biol. 78:  101-115.

 10     Coburn, R. F.  (1979) Mechanisms of carbon monoxide toxicity. Prev. Med. 8: 310-322.

        Coburn, R. F.; Forsters R. E.; Kane, P. B. (1965) Considerations of the physiological variables that determine
               the blood carboxyhemoglobin concentration in man, J. Clin. Invest. 44: 1899-1910.

 15     Cohen, S. I.; Deane, M.; Goldsmith, J.  R. (1969) Carbon monoxide and survival from myocardial infarction.
               Arch. Environ. Health 19:  510-517.

        Collier, C. R.; Workman, J. M.; Mohler, J.  G.; Aaronson, J.; Cabula, O. (1972) Physiological adaptations to
               carbon  monoxide levels and exercise in normal men.  Research Triangle Park, NC: U.  S. Environmental
 20            Protection Agency, Division of Health Effects Research; report no. EPA-RI-72-002. Available from:
               NTIS,  Springfield, VA; PB-213834.

        Comroe, J. H., Jr. (1974) Physiology of respiration. 2nd ed. Chicago, IL: Yearbook Medical  Publishers.

 25     Cooper, K. R.; Alberti, R. R. (1984) Effect 01 kerosene heater emissions  on indoor air quality and pulmonary
               function. Am. Rev. Respir. Dis.  129: 629-631.

        Crocker, P. J.; Walker, J. S. (1985) Pediatric carbon monoxide toxicity. J. Emerg. Med.  3: 443-448.

 30     Daughtrey, W.  C.; Norton, S. (1982) Morphological damage to the premature fetal rat brain after acute carbon
               monoxide exposure. Exp. Neurol. 78: 26-37.

        Daughtrey, W.  C.; Norton, S. (1983) Caudate morphology and behavior of rats exposed to carbon monoxide in
               utero. Exp. Neurol. 80: 265-278.
 35
        Davidson, S. B.; Penney, D. G. (1988) Time course of blood volume change with carbon monoxide inhalation
               and its  contribution to the overall cardiovascular response. Arch. Toxicol. 61: 306-313.

        Davies, D. M.; Smith, D. J. (1980) Electrocardiographic changes in healthy men during continuous low-level
40            carbon  monoxide exposure. Environ.  Res. 21: 197-206.

        Davies, R. F.;  Topping, D. L.; Turner,  D. M. (1976) The effect of intermittent carbon monoxide  exposure on
               experimental atherosclerosis in the rabbit.  Atherosclerosis 24: 527-536.

45     Deanfield, J. B.;  Shea, M. J.; Wilson, R. A.; Horlock, P.; de Landsheere, C. M.; Selwyn, A. P.  (1986) Direct
               effects  of smoking on the heart: silent ischemic disturbances of coronary flow. Am. J.  Cardiol.
               57: 1005-1009.

        DeBias, D. A.; Banerjee, C. M.; Birkhead, N. C.; Harrer, W. V.; Kazal, L. A. (1973) Carbon monoxide
50            inhalation effects following myocardial infarction in monkeys. Arch. Environ. Health 27: 161-167.

        DeBias, D. A.; Banerjee, C. M.; Birkhead, N. C.; Greene, C. H.;  Scott,  S. D.; Harrer, W. V. (1976) Effects of
               carbon  monoxide inhalation on ventricular fibrillation. Arch. Environ. Health 31: 42-46.
         March 12, 1990                              10-193     DRAFT-DO NOT QUOTE OR CITE

-------
       DeLucia, A. J.; Whitaker, J. H.; Bryant, L. R. (1983) Effects of combined exposure to ozone and carbon
              monoxide (CO) in humans. In: Lee, S. D.; Mustafa, M.  G.; Mehlrnan, M. A., eds. International
              symposium on the biomedical effects of ozone and related photochemical oxidants; March; Pinehurst,
              NC. Princeton, NJ:  Princeton Scientific Publishers, Inc.; pp.  145-159. (Advances in modern
 5            environmental toxicology: v. 5).

       Dempsey, L. C.; O'Donnell, J. J.; Hoff, J. T. (1973) Carbon monoxide retinopathy. Am. J. Ophthalmol. 82:
              691.

10     Detar, R.; Bohr, D. F. (1968) Oxygen and vascular smooth muscle contraction. Am. J. Physiol. 214: 241-244.

       Doblar, D. D.; Santiago, T. V.; Edelman, N. H. (1977) Correlation between ventilatory and cerebrovascular
              responses to inhalation of CO. J. Appl.  Physiol.: Respir. Environ. Exercise Physiol.  43: 455-462.

15     Dominick, M. A.; Carson, T. L. (1983) Effects of carbon monoxide exposure on pregnant sows and their
              fetuses. Am. J. Vet. Res. 44: 35-40.

       Donegan, J. H.; Traystman, R. J.; Koehler, R.  C.; Jones, M. D., Jr.; Rogers, M. C. (1985) Cerebrovascular
              hypoxic and autoregulatory responses during reduced  brain metabolism. Am. J. Physiol.
20            249: H421-H429.

       Drinkwater, B. L.; Raven, P. B.; Horvath, S. M.; Gliner, J.  A.; Ruhling, R. O.; Bolduan, N. W.; Taguchi, S.
              (1974) Air pollution, exercise, and heat stress. Arch.  Environ. Health 28: 177-181.

25     Dubeck,  J. K.; Penney, D.  G.; Brown, T. R.; Sharma, P. (1989) Thyroxine treatment of neonatal rats suppresses
              normal and stress-stimulated heart cell hyperplasia and abolishes persistent cardiomegaly. J. Appl.
             Cardiol. 4: 195-205.

       Duffy, T. E.; Nelson, S. R.; Lowry, O. H.  (1972) Cerebral carbohydrate metabolism during acute hypoxia and
30            recovery. J. Neurochem.  19: 959-977.

       Duling, B. R.; Kuschinsky, W.; Wahl, M. (1979) Measurements of the perivascular PO2 in  the vicinity of the
              pial vessels of the cat. Pfluegers Arch. 383: 29-34.

35     Duncan, J. S.; Gumpert, J.  (1983) A case of blindness following carbon monoxide poisoning, treated with
              dopamine [letter]. J. Neurol. Neurosurg. Psychiatry 46: 459.

       Dyer, R. S.; Annau, Z. (1977) Carbon monoxide and flash evoked potentials from rat cortex and superior
              collkiulus. Pharmacol. Biochem. Behav. 6: 461-465.
40
       Dyer, R.; Annau, Z. (1978) Carbon monoxide and superior colliculus evoked potentials.  In: Otto, D. A., ed.
               Multidisciplinary perspectives in event-related brain potential research: proceedings of the fourth
               international congress on event-related slow potentials of the brain (EPIC IV); April  1976;
               Hendersonville, NC. Washington, DC:  U.  S. Environmental  Protection Agency, Office of Research and
45             Development; pp. 417-419; EPA report no. EPA-600/9-77-043. Available from: NTIS, Springfield, VA;
               PB-297137.

        Ebisuno, S.; Yasuno, M.; Yamada, Y.; Nishino, Y.; Hori, M.;  Inoue, M.; Kamada T. (1986) Myocardial
               infarction after acute carbon monoxide poisoning: case report. Angiology 37: 621-624.
50
        Eckardt, R. E.; MacFarland, H.  N.; Alarie, Y. C. E.; Busey, W. M. (1972) The biologic effect from long-term
               exposure of primates to carbon monoxide. Arch. Environ. Health 25:  381-387.
         March 12, 1990                              10-194      DRAFT-DO NOT QUOTE OR CITE

-------
        Effeney, D. J. (1987) Prostacyclin production by the heart: effect of nicotine and carbon monoxide. J. Vase.
               Surg. 5: 237-247.

        Einzig, S.; Nicoloff, D. M.; Lucus, R. V., Jr. (1980) Myocardial perfusion abnormalities in carbon monoxide
  5            poisoned dogs. Can. J. Physiol. Pharmacol. 58: 396-405.

        Ekblom, B.; Huot, R. (1972) Response to submaximal and maximal exercise at different levels of
               carboxyhemoglobin. Acta Physiol. Scand. 86: 474-482.

 10     Ekblom, B.; Huot, R.; Stein, E. M.; Thorstensson, A. T. (1975) Effect of changes in arterial oxygen content on
               circulation and physical performance. J. Appl. Physiol. 39: 71-75.

        Evans, R. G.; Webb, K.; Homan, S.; Ayres, S. M. (1988) Cross-sectional and longitudinal changes in
               pulmonary function associated with automobile pollution among bridge and tunnel officers. Am. J. Lid.
 15            Med. 14: 25-36.

        Fazekas, J.  F.; Colyer, H.; Himwich, H. E. (1939) Effect of cyanide on cerebral metabolism. Proc. Soc. Exp.
               Biol. Med. 42: 496-498.

 20     Fechter, L.  D.; Annau, Z. (1976) Effects of prenatal carbon monoxide exposure on neonatal rats. Adverse Eff.
               Environ. Chem. Psychotropic Drugs 2: 219-227.

        Fechter, L.  D.; Annau, Z. (1977) Toxicity of mild prenatal carbon monoxide exposure. Science (Washington,
               DC) 197: 680-682.
 25
        Fechter, L.  D.; Annau, Z. (1980) Persistent neurotoxic consequences of mild prenatal carbon monoxide
               exposure. In: Di Benedetta, C.; et al, eds. Multidisciplinary approach to brain development. Dev.
               Neurosci. 9:  111-112.

 30     Fechter, L.  D.; Annau, Z. (1980) Prenatal carbon monoxide exposure alters behavioral development.
               Neurobehav. Toxicol. 2: 7-11.

        Fechter, L.  D.; Thakur, M.; Miller, B.; Annau, Z.; Srivastava, U. (1980) Effects  of prenatal carbon monoxide
               exposure on cardiac development. Toxicol. Appl. Pharmacol. 56: 370-375.
 35
        Fechter, L.  D.; Mactutus, C. F.; Storm, J. E. (1986) Carbon monoxide and brain  development. Neurotoxicology
               7: 463-473.

        Fechter, L.  D.; Karpa, M. D.; Proctor, B.; Lee, A. G.; Storm, J. E. (1987) Disruption of neostriatal
40            development in rats following perinatal exposure to mild, but chronic carbon monoxide. Neurotoxicol.
               Teratol. 9: 277-281.

        Fechter, L.  D.; Thome, P.  R.; Nuttall, A. L. (1987) Effects of carbon  monoxide on cochlear electrophysiology
               and  blood flow. Hear. Res. 27: 37-45.
45
        Federal Register. (1986) Guidelines for the health assessment of suspect developmental toxicants. F.  R.
               (September 24) 51:  34028-34040.

        Fein, A.; Grossman, R. F.; Jones, J. G.; Hoeffel, J.; McKay, D. (1980) Carbon monoxide effect on alveolar
50            epithelial permeability. Chest 78: 726-731.

        Fisher, A. B.; Hyde, R. W.; Baue, A. E.; Reif, J. S.; Kelly, D.  F. (1969) Effect of carbon monoxide on
               function and  structure of the lung. J. Appl. Physiol. 26: 4-12.
         March 12, 1990                              10-195      DRAFT-DO NOT QUOTE OR CITE

-------
       Fitzgerald, R. S.; Traystman, R. J. (1980) Peripheral chemoreceptors and the cerebral vascular response to
              hypoxemia. Fed. Proc. Fed. Am. Soc. Exp. Biol. 39: 2674-2677.

       Fodor, G. G.; Winneke, G. (1972) Effect of low CO concentrations on resistance to monotony and on
 5            psychomotor capacity. Staub Reinhalt. Luft 32: 46-54.

       Forbes, W. H.; Dill, D. B.; De Silva, H.; Van deVenter, F. M. (1937) The influence of moderate carbon
              monoxide poisoning upon the ability to drive automobiles. J. Ind. Hyg. Toxicol. 19: 598-603.

10     Forycki,  Z.; Swica, P.; Krasnowiecki, A.; Dubicki,  J.; Panow, A. (1980) Zmiany w obrazie
              elektrokardiograficznym w przebiegu ostrych zatruc [Electrocardiographic changes during acute
              poisonings]. Pol. Tyg. Lek. 35:  1941-1944.

       Foster, J. R. (1981) Arrhythmogenic effects of carbon monoxide in experimental acute myocardial ischemia: lack
15            of slowed conduction and ventricular tachycardia. Am. Heart J. 102: 876-882.

       Fountain, S. B.; Raffaele, K. C.; Annau, Z. (1986) Behavioral consequences of intraperitoneal carbon monoxide
              administration in rats. Toxicol. Appl. Pharmacol. 83: 546-555.

20     Frohlich, E. D. (1987) Potential mechanisms explaining the risk of left ventricular hypertrophy. Am. J. Cardiol.
              59: 91A-97A.

       Gannon,  B. J., et al. (1988) Effects of acute carbon monoxide exposure on precapillary  vessels in the rat
              cremaster muscle. Fed. Proc. Fed. Am.  Soc. Exp. Biol.  2: A743.
25
       Garry, R. C. (1928) The effect of oxygen lack upon surviving smooth muscle. J. Physiol. (London) 66: 235-248.

       Garvey,  D. J.; Longo, L. D. (1978) Chronic low level maternal carbon monoxide exposure and fetal growth and
              development. Biol. Reprod. 19:  8-14.
30
       Gaskell,  T. W. H. (1880) On the toxicity of the heart and blood vessel. J. Physiol. 3: 48-75.

       Gasteva, S. V.; Raize, T. E. (1975) Intensity of respiration and phospholipid metabolism in isolated rat brain
              tissue at different temperatures in the presence of KCN. Byull.  Eksp. Biol.  Med. 79: 53-55.
35
       Gautier,  H.; Bonora, M. (1983) Ventilatory response of intact cats to carbon monoxide hypoxia. J. Appl.
              Physiol.: Respir. Environ. Exercise Physiol. 55: 1064-1071.

       Gilbert,  R. D.; Cummings, L. A.; Juchau, M. R.; Longo, L. D. (1979) Placental diffusing capacity and fetal
40            development in exercising or hypoxic guinea pigs. J. Appl.  Physiol.: Respir. Environ. Exercise  Physiol.
              46: 828-834.

       Gliner, J. A.;  Raven,  P. B.; Horvath, S. M.; Drinkwater,  B. L.; Sutton, J. C. (1975) Man's physiologic
              response to long-term work during thermal and pollutant stress. J. Appl.  Physiol. 39: 628-632.
45
       Gliner, J. A.;  Horvath, S. M.; Mihevic, P. M. (1983) Carbon monoxide and human performance in a single and
              dual task methodology. Aviat. Space Environ. Med. 54: 714-717.

       Gorbatow, O.; Noro,  L. (1948) On acclimatization in connection with acute carbon monoxide poisonings. Acta
50            Physiol. Scand. 15: 77-87.

       Groll-Knapp, E.; Wagner, H.; Hauck, H.; Haider, M. (1972) Effects of low carbon monoxide concentrations on
              vigilance and  computer-analyzed brain potentials. Staub  Reinhalt. Luft 32:  64-68.
         March 12, 1990                              10-196      DRAFT-DO NOT QUOTE OR CITE

-------
        Groll-Knapp, E.; Haider, M.; Hoeller, H.; Jenkner, H.; Stidl, H. G. (1978) Neuro- and psychophysiological
               effects of moderate carbon monoxide exposure. In: Otto, D. A., ed. Multidisciplinary perspectives in
               event-related brain potential research: proceedings of the fourth international congress on event-related
               slow potentials of the brain (EPIC IV); April 1976; Hendersonville, NC. Washington, DC: U. S.
 5            Environmental Protection Agency, Office of Research and Development; pp. 424-430; EPA report no.
               EPA-600/9-77-043. Available from: NTIS, Springfield, VA; PB-297137.

        Groll-Knapp, E.; Harder, M.; Jenkner, H.; Liebich, H.; Neuberger, M.; Trimmel,  M. (1982) Moderate carbon
               monoxide exposure during sleep: neuro- and psychological effects in young  and elderly people.
10            Neurobehav. Toxicol. Teratol. 4: 709-716.

        Grut, A. (1949) Chronic carbon monoxide poisoning: a study in occupational medicine. Copenhagen, Denmark:
               Ejnar Munksgaard.

15     Guest, A. D. L.; Duncan, C.; Lawther, P. J. (1970) Carbon monoxide and phenobarbitone: a comparison of
               effects on auditory flutter fusion threshold and critical flicker fusion threshold. Ergonomics 13: 587-594.

        Gurtner, G. H.; Traystman, R. J.; Burns, B. (1982) Interactions between placental 02 and CO transfer. J. Appl.
               Physiol.: Respir. Environ. Exercise Physiol. 52: 479-487.
20
        Guyton, A. C.; Richardson, T. Q. (1961) Effect of hematocrit on venous return. Circ. Res. 9:  157-164.

        Hagberg, M.; Kolmodin-Hedman, B.; Lindahl, R.; Nilsson,  C.-A.;  Norstrom, A. (1985) Irritative complaints,
               carboxyhemoglobin increase and minor ventilatory function changes due to exposure to chain-saw
25            exhaust. Eur. J. Respir. Dis. 66: 240-247.

        Haggendal, E.; Johannsson, E. (1965) Effect of arterial carbon dioxide tension and oxygen saturation on cerebral
               blood flow autoregulation in dogs. Acta Physiol. Scand. 66: 27-53.

30     Haggendal, E.; Norbeck, B. (1966) Effect of viscosity on cerebral blood flow. Acta Cbir. Scand. Suppl.
               364: 13-22.

        Haider, M.; Groll-Knapp, E.; Hoeller, H.; Neuberger, M.; Stidl, H. (1976) Effects of moderate carbon
               monoxide dose on the central nervous system-electrophysiological and behaviour data and clinical
35            relevance. In: Finkel, A. J.; Duel, W. C., eds. Clinical implications of air pollution research:  air
               polution medical research conference; December 1974; San Francisco, CA. Acton, MA: Publishing
               Sciences Group, Inc.; pp. 217-232.

        Haldane, J. S.; Priestley, J. G. (1935) Respiration. New Haven, CT: Yale University Press.
40
        Halebian, P. H.; Barie, P. S.; Robinson, N.; Shires, G. T. (1984a)  Effects of carbon monoxide on pulmonary
               fluid accumulation. Curr.  Surg. 41: 369-371.

        Halebian, P.; Sicilia,  C.; Hariri, R.;  Inamdar, R.; Shires, G. T. (1984b) A safe and reproducible model of
45            carbon monoxide poisoning. Ann. N. Y. Acad. Sci. 435: 425-428.

        Halperin, M. H.; McFarland, R. A.; Niven, J. L; Roughton, F. J. W. (1959) The time course of the effects of
               carbon monoxide on visual thresholds. J. Physiol. (London)  146: 583-593.

50     Hanley, D.  F.; Wilson, D. A.; Traystman, R. J. (1986) Effect of hypoxia and hypercapnia on neurohypophyseal
               blood flow. Am. J. Physiol. 250:  H7-H15.

        Harbin, T. J.; Benignus, V. A.; Muller, K.  E.; Barton, C. N.  (1988) The effects of low-level carbon monoxide
               exposure upon evoked cortical potentials in young and elderly men. Neurotoxicol. Teratol.  10: 93-100.


         March 12, 1990                              10-197      DRAFT-DO  NOT QUOTE  OR CITE

-------
       Harik, S. I.; LaManna, J. C.; Light, A. I.; Rosenthal, M. (1979) Cerebral norepinephrine: influence on cortical
               oxidative metabolism in situ. Science (Washington, DC) 206: 69-71.

       Heistad, D. D.; Marcus, M. L. (1978) Controversies in cardiovascular research: evidence that neural mechanisms
 5             do not have important effects on cerebral blood flow. Circ. Res. 42: 295-302.

       Heistad, D. D.; Wheller, R. C. (1972) Effect of carbon monoxide on reflex vasoconstriction in man. J. Appl.
               Physiol.  32: 7-11.

10     Heistad, D. D.; Marcus, M. L.; Ehrhardt, J. C.; Abboud, F. M. (1976) Effect of stimulation of carotid
               chemoreceptors on total and regional cerebral blood flow. Circ. Res. 38: 20-25.

       Helsing, K. J.; Sandier, D. P.; Comstock, G. W.; Chee, E. (1988) Heart disease mortality in nonsmokers living
               with smokers. Am.  J.  Epidemiol. 127: 915-922.
15
       Hempel, F. G.; Jobsis, F. F.; LaManna, J. L.; Rosenthal, M. R.; Saltzman, H. A. (1977) Oxidation of cerebral
               cytochrome aa3 by oxygen plus carbon dioxide at hyperbaric pressures. J. Appl. Physiol.: Respir.
               Environ. Exercise Physiol.  43: 873-879.

20     Hernberg, S.; Karava, R.; Koskela, R.-S.; Luoma, K. (1976) Angina pectoris, ECG findings and blood pressure
               of foundry workers in relation to carbon monoxide exposure. Scand. J. Work Environ. Health 2(suppl.
               1): 54-63.

       Heymans, C.; Bouckaert, J. J. (1932) Sinus carotidien et regulation relfexe de la circulation arterielle
25             encephalo-bulbaire. C. B. Seances Soc. Biol. Paris 110: 996-999.

       Higgins, E. A.;  Fiorca, V.; Thomas,  A.  A.; Davis, H. V. (1972) Acute toxicity of brief exposures to HF, HCL,
               NO2, and HCN with and without CO. J. Fire Tech. 8: 12-30.

30     Hill, E. P.; Hill, J. R.; Power, G.  G.; Longo, L. D.  (1977) Carbon monoxide exchanges between the human
               fetus and mother: a mathematical model. Am. J.  Physiol. 232: H311-H323.
       Himwich, H. E.; Fazekas, J.  F. (1941) Comparative studies of the metabolism of the brain of infant and adult
               dogs.  Am. J. Physiol.  132: 454-459.

35     Hinderliter, A. L.;  Adams, K. F.,  Jr.; Price, C. J.; Herbst, M. C.; Koch, G.; Sheps, D. S. (1989) Effects of
               low-level carbon monoxide exposure on resting and exercise-induced ventricular arrhythmias in patients
               with coronary artery disease and  no baseline ectopy. Arch. Environ. Health 44: 89-93.

       Hirsch, G. L.; Sue, D. Y.; Wasserman,  K.; Robinson, T. E.; Hansen, J. E. (1985) Immediate effects of cigarette
40            smoking on cardiorespiratory responses to exercise. J. Appl. Physiol.  58: 1975-1981.

       Hofer, R. (1926) Uber die  Wirkung von Gasgemischen.  Arch. J. Exp. Path. Pharmakol. Ill: 183-205.

        Hoffman, D. J.; Campbell, K. I. (1977) Postnatal toxicity of carbon monoxide after pre- and postnatal exposure.
45            Toxicol. Lett. 1: 147-150.

        Hoppenbrouwers, T.; Calub,  M.; Arakawa, K.; Hodgman, J. (1981) Seasonal relationship of sudden infant death
               syndrome and environmental pollutants. Am. J. Epidemiol.  113: 623-635.

50     Horvath, S. M.  (1975) Influence of carbon monoxide on cardiac dynamics in  normal and cardiovascular stressed
               animals. Sacramento, CA:  Air Resources Board; final report on grant ARB-2096.

        Horvath, S. M.  (1981) Impact of air quality in exercise performance.  Exercise Sport Sci. Rev. 9: 265-296.



         March 12, 1990                               10-198      DRAFT-DO NOT QUOTE OR CITE

-------
        Horvath, S. M.; Dahms, T. E.; O'Hanlon, J. F. (1971) Carbon monoxide and human vigilance: a deleterious
               effect of present urban concentrations. Arch. Environ. Health 23: 343-347.

        Horvath, S. M.; Raven, P. B.; Dahms, T. E.; Gray, D. J. (1975) Maximal aerobic capacity at different levels of
 5             carboxyhemoglobin. J. Appl. Physiol. 38: 300-303.

        Hosko, M. J. (1970) The effect of carbon monoxide on the visual evoked response in man and the spontaneous
               electroencephalogram. Arch. Environ. Health 21:  174-180.

10      Huch, R.; Huch, A.; Tuchschmid, P.; Zijlstra, W. G.; Zwart, A. (1983) Carboxyhemoglobin concentration in
               fetal cord blood. Pediatrics 71:  461-462.

        Hudnell, H. K.; Benignus, V. A.  (1989) Carbon monoxide exposure and human visual function thresholds.
               Neurotoxicol. Teratol.  11: 363-371.
15
        Hugod, C. (1980) The effect of carbon monoxide exposure on morphology of lungs and pulmonary arteries in
               rabbits: a light- and electron-microscopic study. Arch. Toxicol. 43: 273-281.

        Hugod, C. (1981) Myocardial morphology in rabbits exposed to various gas-phase constituents of tobacco smoke:
20             an ultrastructural study. Atherosclerosis 40:  181-190.

        Hugod, C.; Astrup, P. (1980) Exposure of rabbits to carbon monoxide and other gas phase constituents of
               tobacco smoke. MMW Muench. Med. Wochenschr. 122(suppl. 1): S18-S24.

25      Hugod, C.; Hawkins, L. H.; Kjeldsen,  K.; Thomsen, H.  K.; Astrup, P. (1978) Effect of carbon monoxide
               exposure on aortic and coronary intimal morphology in the rabbit.  Atherosclerosis 30: 333-342.

        Hutcheon, D. E.; Doorley, B. M.; Oldewurtel, H. A.  (1983) Carbon monoxide inhalation and the vulnerability
               of the ventricles to electrically induced arrhythmias. Fed.  Proc. Fed. Am. Soc. Exp. Biol. 42: 1114.
30
        Ignarro, L. J.; Byrns, R. E.; Buga, G. M.; Wood, K. S.  (1987)  Endothelium-derived relaxing factor from
               pulmonary artery and vein possesses pharmacologic and chemical properties identical to those of nitric
               oxide radical. Circ. Res. 61: 866-879.

35      Ingenito, A. J.; Durlacher, L. (1979) Effects of carbon monoxide on the b-wave of the cat electroretinogram:
               comparisons with nitrogen hypoxia, epinephrine, vasodilator drugs and changes in respiratory tidal
               volume.  J. Pharmacol.  Exp.  Ther. 211: 638-646.

        Insogna, S.; Warren, C. A. (1984) The effect of carbon monoxide on psychomotor function. In: Mital, A., ed.
40             Trends in ergonomics/human factors I. Amsterdam, The Netherlands: Elsevier/North-Holland;
               pp. 331-337.

        James, I. M.; MacDonnell, L.  (1975) The role of baroreceptors and chemoreceptors in the regulation of the
               cerebral circulation. Clin.  Sci. 49: 465-471.
45
        James, W. E.; Tucker, C.  E.; Grover, R. F. (1979) Cardiac  function in goats exposed to carbon monoxide.  J.
               Appl. Physiol.: Respir.  Environ. Exercise Physiol. 47: 429-434.

        Jarvis, M. J. (1987) Uptake of environmental tobacco smoke. In:  O'Neill,  I. K.;  Brunnemann, K. D.; Dodet, B.;
50             Hoffmann, D., eds. Environmental carcinogens: methods  of analysis and  exposure measurements:  vol. 9 -
               passive smoking. Lyon, France: International Agency for  Research on Cancer; pp. 43-58. (IARC
               scientific publications: v. 81).
         March  12, 1990                              10-199     DRAFT-DO NOT QUOTE OR CITE

-------
       Jobsis, F. F.; Rosenthal, M. (1978) Behaviour of the mitochondrial respiratory chain in vivo. In: Cerebral
              vascular smooth muscle and its control. Amsterdam, The Netherlands: Elsevier;  pp. 149-169. (Ciba
              Foundation symposium series no. 56).

 5     Jones, M. D., Jr.; Traystman, R. J. (1984) Cerebral oxygenation of the fetus, newborn, and adult.  Semin.
              Perinatal. 8: 205-216.

       Jones, R. A.; Strickland, J. A.; Stunkard, J. A.; Siegel, J. (1971) Effects on experimental animals of long-term
              inhalation exposure to carbon monoxide.  Toxicol.  Appl. Pharmacol. 19: 46-53.
10
       Jones, M. D., Jr.; Sheldon, R. E.; Peelers, L. L.; Makowski, E. L.; Meschia, G. (1978) Regulation of cerebral
              blood flow in the ovine fetus. Am. J. Physiol. 235: H162-H166.

       Jones, M. D., Jr.; Traystman, R. J.; Simmons, M. A.; Molteni, R. A. (1981) Effects of changes in arterial O2
15            content  on cerebral blood flow in the lamb. Am. J. Physiol. 240:  H209-H215.

       Kanten, W. E.; Penney, D. G.; Francisco, K.; Hill, J. E. (1983) Hemodynamic responses to acute
              carboxyhemoglobinemia in the rat. Am. J. Physiol. 244: H320-H327.

20     Katafuchi, Y.; Nishimi, T.; Yamaguchi, Y.; Matsuishi, T.; Kimura, Y.;  Otaki, E.; Yamashita, Y. (1985)
              Cortical blindness in acute carbon monoxide poisoning. Brain Dev. 7: 516-519.

       Katsumata, Y.;  Aoki, M.;  Oya, M.; Yada, S.; Suzuki, O.  (1980) Liver damage in rats during acute carbon
              monoxide poisoning. Forensic Sci. Int. 16: 119-123.
25
       Kaul, B.; Calabro, J.; Hutcheon, D. E. (1974) Effects of carbon monoxide on the vulnerability of the ventricles
              to drug-induced arrhythmias. J. Clin. Pharmacol. 14: 25-31.

       Kelley, J. S.; Sophocleus,  G. J. (1978) Retinal hemorrhages in subacute carbon monoxide poisoning: exposures
30            in homes with blocked furnace flues. JAMA J. Am. Med. Assoc.  239: 1515-1517.

       Kety, S.  S.; Schmidt, C. F. (1948) The effects of altered arterial tensions of carbon dioxide and oxygen on
              cerebral blood flow and cerebral oxygen  consumption in normal young men. J. Clin. Invest. 27: 484-492.

35     Killick, E. M. (1937) The acclimatization of mice to atmospheres containing low concentrations of carbon
              monoxide. J. Physiol. (London) 91: 279-292.

       Killick, E. M. (1940) Carbon monoxide anoxemia. Physiol. Rev. 20: 313-344.

40     Killick, E. M. (1948) The nature of the acclimatization occurring during repeated exposure of the human subject
              to atmospheres containing low concentrations of carbon monoxide. J. Physiol. 107: 27-44.

       King, C. E.; Cain, S. M.; Chapler, C. K. (1984) Whole body and hindlimb cardiovascular responses of the
              anesthetized dog during CO hypoxia. Can. J. Physiol. Pharmacol. 62: 769-774.
45
       King, C. E.; Cain, S. M.; Chapler, C. K. (1985) The role of aortic chemoreceptors during severe CO hypoxia.
               Can. J. Physiol. Pharmacol 63: 509-514.

       King, C. E.; Dodd, S.  L.; Cain, S. M. (1987) O2 delivery to contracting muscle during hypoxic or CO hypoxia.
50            J. Appl. Physiol. 63: 726-732.

       Kjeldsen, K.; Astrup, P.; Wanstrup, J. (1972) Ultrastructural intimal changes in the rabbit  aorta after a moderate
               carbon  monoxide exposure.  Atherosclerosis 16: 67-82.
         March 12, 1990                              10-200      DRAFT-DO NOT QUOTE OR CITE

-------
        Klausen, K.; Andersen, C.; Nandrup, S. (1983) Acute effects of cigarette smoking and inhalation of carbon
               monoxide during maximal exercise. Eur. J. Appl. Physiol. Occup. Physiol. 51: 371-379.

        Klees, M.; Heremans, M.; Dougan, S. (1985) Psychological sequelae to carbon monoxide intoxication in the
  5            child. Sci. Total Environ. 44: 165-176.

        Klein, J. P.; Forster, H. V.; Stewart, R. D.; Wu, A.  (1980) Hemoglobin affinity for oxygen during short-term
               exhaustive exercise. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 48:  236-242.

 10     Kleinert, H.  D.; Scales, J. L.; Weiss, H. R. (1980) Effects of carbon monoxide or low oxygen gas mixture
               inhalation on regional oxygenation, blood flow, and small vessel blood content of the rabbit heart.
               Pfluegers Arch. 383:  105-111.

        Kleinman, M.; Whittenberger, J. (1985) Effects of short-term exposure to carbon monoxide in subjects with
 15            coronary artery disease. Sacramento, CA: California State Air Resources Board; report no.
               ARB-R-86/276. Available from: NTIS, Springfield, VA; PB86-217494.

        Kleinman, M. T.; Davidson, D. M.; Vandagriff,  R. B.; Caiozzo, V. J.; Whittenberger, J. L. (1989) Effects of
               short-term exposure to carbon monoxide in subjects with coronary artery disease. Arch. Environ.  Health
20            44: 361-369.

        Knelson, J, H. (1972) U.S. air quality criteria and ambient standards for carbon monoxide. Staub-Reinhalt, Luft
               32: 60-64.

25     Knelson, J. H. (1972) The carbon monoxide air quality standard: implications for inner city children [memo to
               Dr. J. F.  Finklea]. Research Triangle Park, NC: U. S. Environmental Protection Agency; March 23.

        Koehler, R. C.; Jones, M. D., Jr.; Traystman,  R. J. (1982) Cerebral circulatory response to carbon monoxide
               and hypoxic hypoxia in the lamb. Am. J.  Physiol. 243: H27-H32.
30
        Koehler, R. C.; Traystman, R. J.; Rosenberg, A. A.;  Hudak, M. L.; Jones, M. D., Jr. (1983) Role of
               O2-hemoglobin affinity on cerebrovascular response to carbon monoxide hypoxia. Am. J. Physiol. 245:
               H1019-H1023.

35     Koehler, R. C.; Traystman, R. J.; Zeger, S.; Rogers,  M. C.; Jones, M. D., Jr. (1984) Comparison of
               cerebrovascular responses to hypoxic and carbon monoxide hypoxia in newborn and adult sheep. J.
               Cereb.  Blood Flow Metab.  4: 115-122.

        Koehler, R. C.; Traystman, R. J.; Jones, M. D., Jr. (1985) Regional blood flow and O2 transport during
40            hypoxic and CO hypoxia in neonatal and adult sheep. Am. J. Physiol. 248: H118-H124.

        Kogure, K.; Scheinberg, P.; Reinmuth, O.  M.; Fujishima, M.; Busto, R. (1970) Mechanisms of cerebral
               vasodilatation in hypoxia. J. Appl.  Physiol. 29: 223-229.

45     Kontos, H. A.; Wei, E. P.; Raper, A. J.; Rosenblum, W. I.; Navari, R. M.; Patterson, J. L., Jr. (1978) Role of
               tissue hypoxia in local regulation of cerebral microcirculation.  Am. J. Physiol.  234: H582-H591.

        Koob, G.  F.; Annau,  Z.; Rubin, R. J.; Montgomery,  M. R. (1974) Effect of hypoxic hypoxia and carbon
               monoxide on food intake, water intake,  and body weight in two strains of rats.  Life Sci.  14: 1511-1520.
50
        Korner, P. I. (1965) The role of the arterial chemoreceptors and baroreceptors in the circulatory response to
               hypoxia of the rabbit. J. Physiol. 180: 279-303.
         March 12,  1990                             10-201      DRAFT-DO NOT QUOTE OR CITE

-------
        Kuller, L. H.; Radford, E. P. (1983) Epidemiological bases for the current ambient carbon monoxide standards.
               EHP Environ. Health Perspect. 52: 131-139.

        Kuller, L. H.; Radford, E. P.; Swift, D.; Perper, J. A.; Fisher, R. (1975) Carbon monoxide and heart attacks.
 5             Arch. Environ. Health 30: 477-482.

        Kuska, J.; Kokot, F.; Wnuk, R. (1980) Acute renal failure after exposure to carbon monoxide. Mater. Med. Pol.
               (Engl. Ed.) 12: 236-238.

10      Kustov, V. V.; Belkin, V. L; Abidin, B. L; Ostapenko, O. F.; Malkuta, A. N.; Poddubnaja, L. T. (1972)
               Aspects of chronic carbon monoxide poisoning in young animals. Gig. Tr. Prof. Zabol. 5: 50-52.

        Lahiri, S.; Delaney, R. G. (1976) Effect of carbon monoxide on carotid chemoreceptor activity and ventilation.
               In: Paintal, A. S., ed. Morphology and mechanisms of chemoreceptors:  proceedings of an international
15             satellite symposium;  October 1974; Srinagar, India. New Delhi, India: Navchetan Press;  pp. 340-344.

        Langfitt, T. W.; Kassell, N.  F. (1968) Cerebral vasodialatation produced by brainstem stimulation: neurogenic
               control vs. autoregulation. Am. J. Physiol. 215: 90-97.

20      Laties, V. G.; Merigan, W. H. (1979) Behavioral effects of carbon monoxide on animals and man. Annu.  Rev.
               Pharmacol. Toxicol.  19: 357-392.

        Lebowitz, M. D. (1984) The effects of environmental tobacco smoke exposure and gas stoves on daily peak flow
               rates in asthmatic and non-asthmatic families. In: Rylander, R.; Peterson, Y.; Snella, M.-C., eds. ETS -
25             environmental tobacco smoke: report from a workshop on effects and exposure levels; March 1983;
               Geneva, Switzerland. Eur. J. Respir. Dis. 65(suppl. 133): 90-97.
                                                                                             *
        Lebowitz, M. D.;  Holberg, C. J.; O'Rourke, M. K.; Corman, G.; Dodge, R. (1983a) Gas stove usage, CO and
               TSP, and respiratory effects. Research Triangle Park, NC:  U. S. Environmental Protection Agency,
30             Health Effects Research Laboratory; EPA report no. EPA-600/D-83-107. Available from: NTIS,
               Springfield, VA; PB83-250357.

        Lebowitz, M. D.;  Holberg, C. J.; Dodge, R. R. (1983b) Respiratory effects on populations from low-level
               exposures to ozone. Presented at: 76th annual meeting of the Air Pollution Control Association; June;
35             Atlanta,  GA. Pittsburgh, PA: Air Pollution Control Association; paper no. 83-12.5.

        Lebowitz, M. D.;  Corman, G.; O'Rourke, M. K.; Holberg, C. J.  (1984) Indoor-outdoor air pollution, allergen
               and meteorological monitoring hi an arid southwest area. J. Air Pollut. Control Assoc. 34: 1035-1038.

40      Lebowitz, M. D.;  Holberg, C. J.; Boyer, B.; Hayes, C. (1985) Respiratory symptoms and peak flow associated
               with indoor and outdoor air pollutants in the southwest. J.  Air Pollut. Control Assoc. 35: 1154-1158.

        Lebowitz, M. D.;  Collins, L.; Holberg, C. J.  (1987) Time series analyses of respiratory responses to indoor and
               outdoor  environmental phenomena. Environ.  Res. 43: 332-341.
45
        Levine, S.  (1967) Experimental cyanide encephalopathy: gradients  of susceptibility in the corpus callosum. J.
               Neuropathol. Exp. Neurol.  26: 214-222.

        Levine, S.;  Stypulkowski, W. (1959) Experimental cyanide encephalopathy.  Arch. Pathol. 67: 306-323.
50
        Lilienthal, J. L., Jr.; Fugitt, C. H. (1946) The effect of low concentrations of carboxyhemoglobin on the
               "altitude tolerance" of man. Am. J. Physiol.  145: 359-364.
         March 12, 1990                              10-202      DRAFT-DO NOT QUOTE OR CITE

-------
        Longo, L. D. (1970) Carbon monoxide in the pregnant mother and fetus and its exchange across the placenta. In:
               Coburn, R. F., ed. Biological effects of carbon monoxide. Ann. N. Y. Acad. Sci. 174: 313-341.

        Longo, L. D. (1976) Carbon monoxide: effects on oxygenation of the fetus in utero. Science (Washington, DC)
  5            194: 523-525.

        Longo, L. D. (1977) The biological effects of carbon monoxide on the pregnant woman,  fetus, and newborn
               infant. Am. J. Obstet. Gynecol. 129: 69-103.

 10     Longo, L. D.; Ching, K. S. (1977) Placental diffusing capacity for carbon monoxide and oxygen in
               unanesthetized sheep. J. Appl. Physiol.: Respir.  Environ. Exercise Physiol. 43: 885-893.

        Longo, L. D.; Hill,  E. P. (1977) Carbon  monoxide uptake and elimination in fetal and maternal sheep. Am. J.
               Physiol. 232: H324-H330.
 15
        Lubbers,  D. W. (1968) The oxygen pressure field of the brain and its significance for the normal and critical
               oxygen supply of the brain. In:  Lubbers, D. W.; Luft, V. C.; Thews, G., Witzleb, E., eds. Oxygen
               transport in blood and tissue. Stuttgart, Federal Republic of Germany: Verlag; pp. 124-139.

 20     Luria, S.  M.; McKay, C. L. (1979) Effects of low levels of carbon monoxide on  visions  of smokers and
               nonsmokers. Arch.  Environ. Health 34: 38-44.

        Lutz, L. J.  (1983) Health effects of air pollution measured by outpatient visits. J.  Fam. Pract. 16: 307-313.

 25     MacMillan, V. (1975) Regional cerebral blood flow of the rat in acute carbon monoxide intoxication. Can. J.
               Physiol. Pharmacol. 53: 644-650.

        Mactutus, C. F.; Fechter, L. D.  (1984) Prenatal exposure to carbon monoxide: learning and memory deficits.
               Science (Washington, DC) 223: 409-411.
 30
        Mactutus, C. F.; Fechter, L. D.  (1985) Moderate prenatal carbon monoxide exposure produces persistent, and
               apparently permanent, memory deficits in rats. Teratology 31: 1-12.

        Madsen, H.; Dyerberg, J. (1984) Cigarette smoking and its effects on the platelet-vessel wall interaction. Scand.
 35            J. Clin. Lab. Invest. 44:  203-206.

        Malinow, M. R.; McLaughlin, P.;  Dhindsa, D. S.; Metcalfe, J.; Ochsner, A. J.,  Ill; Hill, J.; McNulty, W. P.
               (1976)  Failure of carbon  monoxide to induce myocardial infarction in cholesterol-fed cynomolgus
               monkeys (Macacafascicularis).  Cardiovasc. Res. 10: 101-108.
40
        Mall, Th.; Grossenbacher, M.; Perruchoud, A. P.; Ritz, R. (1985)  Influence of moderately elevated levels of
               carboxyhemoglobin on the course of acute ischemic heart disease. Respiration 48:  237-244.

        Mansouri, A.; Perry,  C. A.  (1982) Alteration of platelet aggregation by cigarette smoke and carbon monoxide.
45            Thromb. Haemostasis 48: 286-288.

        Marshall,  M.; Hess, H. (1981) Akute Wirkungen niedriger Kohlenmonoxidkonzentrationen auf Blutrheologie,
               Thrombozytenfunktion und  Arterienwand beim Miniaturschwein [Acute effects of  low carbon monoxide
               concentrations on blood rheology,  platelet function, and the  arterial wall in the minipig]. Res. Exp. Med.
50            178: 201-210.

        Martin, T. R.;  Bracken, M.  B. (1986) Association of low birth weight with passive smoke exposure in
               pregnancy. Am. J. Epidemiol. 124: 633-642.
         March  12,  1990                             10-203      DRAFT-DO NOT QUOTE OR CITE

-------
       Martynjuk, V. C.; Dacenko, I. I. (1973) Aktivnost' transaminaz v usloviyakh khronicheskoi intoksikatsii okis'yu
              ugleroda [Aminotransferase activity in chronic carbon monoxide poisoning]. Gig. Naselennykh. Mest.
              12: 53-56.

 5     Mason, G. R.; Usder, J. M.; Effiros, R. M.; Reid, E. (1983) Rapid reversible alterations of pulmonary epithelial
              permeability induced by smoking. Chest 83: 6-11.

       McDowall, D. G. (1966) Interrelationships between blood oxygen tensions and cerebral blood flow. In:  Payne, J.
              P.; Hill, D.  W., eds. Symposium on oxygen measurements in blood and tissue and their significance.
10            London, United Kingdom: Churchill; pp.  205-214.

       McFarland, R. A. (1970) The effects of exposure to small quantities of carbon monoxide on vision. In:  Coburn,
              R. F., ed. Biological effects of carbon monoxide. Ann. N. Y. Acad. Sci. 174: 301-312.

15     McFarland, R. A. (1973) Low level exposure to carbon monoxide and driving performance. Arch. Environ.
              Health 27: 355-359.

       McFarland, R. A.; Roughton, F. J. W.; Halperin, M. H.; Niven, J. I. (1944) The effects of carbon monoxide
              and altitude on visual thresholds. J. Aviat. Med. 15: 381-394.
20
       McFarland, R. A.; Forbes, W. H.; Stoudt, H. W.; Dougherty, J. D.; Crowley, T. J.; Moore, R. C.; Nalwalk,
              T. J.  (1973) A study of the effects of low levels of carbon monoxide upon humans performing driving
              tasks. Boston, MA: Harvard School of Public Health. Available from: NTIS, Springfield, VA;
              PB-233894.
25
       McGinty, D.  (1929) The regulation of respiration: XXV. Variations  in the lactic acid metabolism in the intact
              brain. Am. J. Physiol. 88: 312-329.

       McGrath, J. J. (1982) Physiological effects of carbon monoxide.  In:  McGrath, J. J.; Barnes, C.  D., eds. Air
30            pollution - physiological effects. New York, NY: Academic Press; pp. 147-181.

       McGrath, J. J. (1989) Cardiovascular effects of chronic carbon monoxide and high-altitude exposure.  Cambridge,
              MA:  Health Effects Institute; research report number 27.

35     Mchedlishvili, G.; Nikolaischvili, L.; Antia, R. (1976) Are the pial  arterial  responses dependent on the direct
              effect of intravascular pressure and extravascular PO2, PCO2 and pH? Microvasc. Res. 10: 298-311.

       McMeekin, J. D.; Finegan, B. A. (1987) Reversible myocardial dysfunction following carbon monoxide
              poisoning. Can. J. Cardiol. 3: 118-121.
40
       Melinyshyn, M. J.;  Cain, S. M.; Villeneuve, S. M.; Chapler, C. K. (1988) Circulatory and metabolic responses
              to carbon monoxide hypoxia during /3-adrenergic blockade. Am.  J. Physiol. 255: H77-H84.

       Meyer, A. (1963) Intoxications.  In:  Blackwood, W.; McMenemey, W. H.; Meyer, A.; Norman, R. M.; Russell,
45            D. S., eds. Greenfield's  neuropathology. 2nd ed. London, United Kingdom: Edward Arnold (Publishers)
              Ltd.;  pp. 235-287.

       Mihevic, P. M.; Gliner, J. A.; Horvath, S. M. (1983) Carbon monoxide exposure and information processing
              during perceptual-motor  performance. Int. Arch. Occup.  Environ. Health 51: 355-363.
50
       Miller, A. T.; Wood, J. J. (1974) Effects of acute carbon monoxide exposure on the energy metabolism of rat
              brain and liver. Environ. Res. 8: 107-111.
         March 12, 1990                             10-204      DRAFT-DO NOT QUOTE OR CITE

-------
       Mills, A. K.; Skornik, W. A.; Valles, L. M.; O'Rourke, J. J.; Hennessey, R. M.; Verrier, R. L. (1987) Effects
              of carbon monoxide on cardiac electrical properties during acute coronary occlusion. Fed. Proc. Fed.
              Am. Soc. Exp. Biol. 46: 336.

 5     Minor, M.; Seidler, D. (1986) Myocardial infarction following carbon monoxide poisoning. W. Va. Med. J.
              82: 25-28.

       Minty, B. D.; Royston, D. (1985) Cigarette smoke induced changes in rat pulmonary clearance of "TcDTPA: a
              comparison of participate and gas phases. Am. Rev. Respir. Dis.  132: 1170-1173.
10
       Minty, B. D.; Jordan, C.; Jones, J. G. (1981) Rapid improvement in abnormal pulmonary epithelial permeability
              after stopping cigarettes. Br. Med. J. 282: 1183-1186.

       Mochizuki, M.;  Maruo, T.; Masuko, K.; Ohtsu, T. (1984) Effects of smoking on fetoplacental-matemal system
15            during pregnancy. Am. J. Obstet. Gynecol. 149: 413-420.

       Molnar,  L.; Szanto, J. (1964) Effect of electrical stimulation of the bulbar vasomotor center on the cerebral blood
              flow. Q. J. Exp. Physiol. 49: 184-193.

20     Montgomery, M. R.; Rubin, R. J. (1971) The effect of carbon monoxide inhalation on in vivo drug metabolism
              in the rat. J. Pharmacol. Exp. Ther. 179: 465-473.

       Mordelet-Dambrine, M.;  Stupfel, M. (1979) Comparison in guinea-pigs and in rats of the effects of vagotomy
              and of atropine on respiratory resistance modifications induced by an acute carbon monoxide or nitrogen
25            hypoxia. Comp. Biochem.  Physiol. A 63: 555-559.

       Mordelet-Dambrine, M.;  Stupfel, M.; Duriez, M. (1978) Comparison of tracheal pressure and circulatory
              modifications induced in guinea pigs and in rats by carbon monoxide inhalation. Comp. Biochem.
              Physiol.  A 59: 65-68.
30
       Morris, G. L.; Curtis, S. E.; Simon, J. (1985a) Perinatal piglets under sublethal concentrations of atmospheric
              carbon monoxide. J.  Anim. Sci. 61: 1070-1079.

       Morris, G. L.; Curtis, S. E.; Widowski,  T. M.  (1985b) Weanling pigs under sublethal concentrations of
35            atmospheric carbon monoxide. J.  Anim. Sci. 61: 1080-1087.

       Morris, P. D.; Koepsell,  T. D.; Baling, J.  R.; Taylor, J.  W.; Lyon, J. L.; Swanson,  G. M.; Child,  M.; Weiss,
              N. S. (1986) Toxic substance exposure and multiple myeloma: a case-control study. JNCI J. Natl.
              Cancer Inst. 76: 987-994.
40
       Moss, R. H.; Jackson, C. F.; Seiberlich, J. (1951) Toxicity of carbon monoxide and hydrogen cyanide gas
              mixtures: a preliminary report. AMA Arch. Ind. Hyg. Occup. Med. 4: 53-64.

       Muller, K. E.; Barton, C. N.; Benignus,  V. A.  (1984) Recommendations for appropriate statistical practice in
45            toxicologic experiments. Neurotoxicology 5: 113-126.

       Mullin, L. S.; Krivanek,  N.  D. (1982) Comparison of unconditioned reflex and conditioned avoidance tests in
              rats exposed by inhalation to carbon monoxide, 1,1,1-trichloroethane, toluene or ethanol.
              Neurotoxicology 3: 126-137.
50
       Musselman, N. P.; Groff, W. A.;  Yevich,  P. P.; Wilinski, F. T.; Weeks, M. H.; Oberst, F. W. (1959)
              Continuous exposure of laboratory animals to low concentration of carbon monoxide. Aerosp. Med. 30:
              524-529.
         March  12, 1990                              10-205      DRAFT-DO NOT QUOTE OR CITE

-------
       Naeije, R.; Peretz, A.; Cornil, A. (1980) Acute pulmonary edema following carbon monoxide poisoning.
               Intensive Care Med. 6: 189-191.

       National Institute of Child Health and Human Development. (1979) Pregnancy and infant health. In: Smoking
 5             and health: a report of the Surgeon General. Washington, DC: U. S. Department of Health, Education,
               and Welfare; DHEW publication no. PHS 79-50066.

       Neubauer, J. A.; Santiago, T. V.; Edelman, N. H. (1981) Hypoxic arousal in intact and carotid chemodenervated
               sleeping cats. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 51:  1294-1299.
10
       Niden, A. H. (1971) The effects of low levels of carbon monoxide on the fine structure of the terminal airways.
               Am.  Rev. Respir.  Dis. 103:  898.

       Niden, A. H.; Schulz, H.  (1965) The ultrastructural effects of carbon monoxide inhalation on the rat lung.
15             Virchows Arch. A: Pathol. Anat. 339: 283-292.

       Nilsson, B.;  Norberg, K.; Nordstrom, C.;  Siesjo, B. K. (1976) Influence of hypoxia and hypercapnia on cerebral
               blood flow in rats. In: Harper, A. M.; Jennett, W. B.; Miller, J. D.; Rowan, J. O., eds. Blood flow and
               metabolism of the brain.  Edinburgh, Scotland: Livingstone; pp. 9.19-9.23.
20
       Norberg, K.; Siesjo, B. K. (1975) Cerebral metabolism in hypoxic hypoxia. I. Pattern of activation of glycolysis:
               a re-evaluation.  Brain Res. 86: 31-44.

       O'Donnell, R. D.; Mikulka, P.;  Heinig, P.; Theodore, J. (1971a) Low level carbon monoxide exposure and
25             human psychomotor performance. Toxicol. Appl. Pharmacol. 18: 593-602.

       O'Donnell, R. D.; Chikos, P.; Theodore, J. (1971b) Effect of carbon monoxide exposure on human sleep and
               psychomotor performance. J. Appl. Physiol.  31: 513-518.

30    O'Sullivan, B. P. (1983) Carbon monoxide poisoning in an infant exposed to a kerosene heater. J.  Pediatr. 103:
               249-251.

        Okeda, R.; Matsuo, T.; Kuroiwa, T.; Tajima, T.; Takahashi, H. (1986) Experimental study on pathogenesis of
               the fetal brain damage by acute carbon monoxide intoxication of the pregnant mother.  Acta Neuropathol.
35            69: 244-252.

        Okeda, R.; Matsuo, T.; Kuroiwa, T.; Nakai, M.; Tajima, T.; Tkahashi, H. (1987) Regional cerebral blood flow
               of acute carbon monoxide poisoning in cats.  Acta Neuropathol. 72: 389-393.

40     Olsen, N.  S.; Klein, J.  R. (1947) Effect of cyanide on the concentration of lactate and phosphates in brain. J.
               Biol. Chem. 167:  739-746.

        Opitz, E.; Schneider, M.  (1950) Uber die  SauerstoffVersorgung des Gehirns und den Mechanisms von
               Mangelwirken.  Curr. Top. Pathol. 46: 126-260.
45
        Oremus, R.  A., et al. (1988) Effects of acute carbon monoxide (CO) exposure on cardiovascular (CV)
               hemodynamics in  the anesthetized rat. Fed. Proc. Fed. Am.  Soc. Exp. Biol. 2:  A1312.

        Otto, D.; Reiter, L. (1978) Neurobehavioral assessment of environmental insult. In: Otto, D.  A., ed.
50            Multidisciplinary  perspectives in event-related brain potential research: proceedings of the fourth
               international congress on event-related slow potentials of the brain (EPIC IV); April 1976;
               Hendersonville, NC. Washington,  DC: U. S. Environmental Protection Agency, Office of Research and
               Development; pp. 409-416;  EPA report no.  EPA-600/9-77-043. Available from: NTIS, Springfield, VA;
               PB-297137.


          March 12,  1990                               10-206      DRAFT-DO NOT QUOTE OR CITE

-------
        Otto, D. A.; Benigus, V. A.; Prah, J. D. (1979) Carbon monoxide and human time discrimination: failure to
               replicate Beard-Wertheim experiments. Aviat. Space Environ. Med. 50: 40-43.

        Pankow, D.; Ponsold, W. (1972) Leucine aminopeptidase activity in plasma of normal and carbon monoxide
  5            poisoned rats. Arch. Toxicol. 29: 279-285.

        Pankow, D.; Ponsold, W. (1974) Kombinationswirkungen von Kohlenmonoxid mit anderen biologischaktiven
               Schadfaktoren auf den Organismus [The combined effects of carbon monoxide and other biologically
               active detrimental factors on the organism]. Z. Gesamte Hyg. Ihre Grenzgeb. 20: 561-571.
 10
        Pankow, D.; Ponsold, W.;  Fritz, H. (1974) Combined effects of carbon monoxide and ethanol on the activities of
               leucine aminopeptidase and glutamic-pyruvic transaminase in the plasma of rats. Arch. Toxicol.
               32: 331-340.

 15     Parving,  H.-H. (1972) The effect of hypoxia and carbon monoxide exposure on plasma volume and capillary
               permeability to albumin. Scand. J. Clin.  Lab. Invest. 30: 49-56.

        Paulet, G. (1958) Sur le mechanisme de la perte  de 1'action hypertensive de 1'adrenaline au cours de 1'anoxie
               cyanhydrique. J.  Physiol. (Paris) 50: 31-52.
 20
        Paulson,  O. B.; Parving, H.-H.; Olesen, J.; Skinhoj, E. (1973) Influence of carbon monoxide and of
               hemodilution on cerebral blood flow and blood gases in man. J. Appl. Physiol. 35: 111-116.

        Peelers, L. L. H.; Sheldon, R. E.;  Jones, M. D., Jr.; Makowski, E. L.; Meschia, G. (1979) Blood flow to fetal
 25            organs as a function of arterial oxygen content. Am. J. Obstet. Gynecol. 135: 637-646.

        Penn, A.; Butler, J.; Snyder,  C.; Albert, R. E. (1983) Cigarette smoke and carbon monoxide do not have
               equivalent effects upon development of arteriosclerotic lesions.  Artery 12: 117-131.

 30     Penney, D. G.; Caldwell, T. M. (1984) Growth  of the young heart: influence of nutrition and stress. Growth
               48: 483-498.

        Penney, D. G.; Weeks, T. A. (1979) Age dependence of cardiac growth in the normal and carbon
               monoxide-exposed rat. Dev. Biol. 71: 153-162.
 35
        Penney, D.; Dunham, E.; Benjamin, M. (1974a) Chronic carbon monoxide exposure: time course of hemoglobin,
               heart weight and lactate  dehydrogenase isozyme changes. Toxicol. Appl. Pharmacol. 28: 493-497.

        Penney, D.; Benjamin, M.; Dunham, E. (1974b) Effect of carbon monoxide on cardiac weight as compared with
40            altitude effects. J. Appl. Physiol. 37: 80-84.

        Penney, D. G.; Sodt, P. C.; Cutilletta, A. (1979) Cardiodynamic changes during prolonged carbon monoxide
               exposure in the rat.  Toxicol. Appl. Pharmacol. 50: 213-218.

45     Penney, D. G.; Baylerian, M. S.; Fanning, K. E. (1980) Temporary and lasting cardiac effects of pre- and
               postnatal exposure to carbon monoxide. Toxicol. Appl. Pharmacol. 53: 271-278.

        Penney, D. G.; Baylerian, M. S.; Thill, J. E.; Fanning,  C. M.; Yedavally, S. (1982) Postnatal carbon monoxide
               exposure: immediate and lasting effects in the rat. Am. J.  Physiol.  243: H328-H339.
50
        Penney, D. G.; Baylerian, M. S.; Thill, J. E.; Yedavally, S.;  Fanning, C. M. (1983) Cardiac response of the
               fetal rat to carbon monoxide exposure. Am. J. Physiol. 244: H289-H297.
         March 12,  1990                              10-207      DRAFT-DO NOT QUOTE OR CITE

-------
       Penney, D. G.; Barthel, B. G.; Skoney, J. A. (1984a) Cardiac compliance and dimensions in carbon monoxide-
              induced cardiomegaly.  Cardiovasc. Res.  18: 270-276.

       Penney, D. G.; Stryker, A. E.; Baylerian, M. S. (1984b) Persistent cardiomegaly induced by carbon monoxide
 5            and associated tachycardia. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 56: 1045-1052.

       Penney, D. G.; Davidson, S. B.; Gargulinski, R. B.; Caldwell-Ayre, T. M. (1988a) Heart and lung hypertrophy,
              changes in blood volume,  hematocrit and plasma renin activity in rats chronically exposed in increasing
              carbon monoxide concentrations. JAT J.  Appl. Toxicol. 8:  171-178.
10
       Penney, D. G.; Gargulinski, R. B.;  Hawkins, B. J.; Santini, R.; Caldwell-Ayre, T. M.; Davidson, S. B. (1988b)
              The effects of carbon monoxide on persistent changes in young rat heart: cardiomegaly,  tachycardia and
              altered DNA content. JAT J. Appl. Toxicol. 8: 275-283.

15     Penney, D. G.; Clubb, F. J., Jr.;  Allen, R. C.;  Jen, C.; Benerjee,  S.; Hull, J. A. (1989) Persistent early heart
              changes do alter the cardiac  response to exercise training in adulthood. J.  Appl.  Cardiol. 4: 223-237.

       Permutt, S.; Farhi, L. (1969) Tissue hypoxia and carbon monoxide. In: Effects of chronic exposure to low levels
              of carbon monoxide on human health, behavior, and performance. Washington,  DC: National Academy
20            of Sciences; pp. 18-24.

       Petajan, J. H.; Packham, S. C.; Frens, D. B.; Dinger, B.  G. (1976) Sequelae of carbon monoxide-induced
              hypoxia in the rat. Arch. Neurol. (Chicago) 33: 152-157.

25     Pfluger, E. (1875) Beitrage zue lehre von der Respiration. I. Ueber die physiologische Verbrennung in der
              Lebendigeen Organismen. Pfluegers Arch. Gesamte Physiol. Menschen Tiere 10: 251-367.

       Piantadosi, C. A.; Sylvia, A. L.;  Saltzman, H. A.; Jobsis-Vandervliet, F. F. (1985) Carbon
               monoxide-cytochrome  interactions in the brain of the fluorocarbon-perfused rat.  J. Appl. Physiol.
30            58: 665-672.

       Piantadosi, C. A.; Sylvia, A. L.;  Jobsis-Vandervliet, F. F. (1987)  Differences in brain  cytochrome responses to
               carbon monoxide and cyanide in vivo. J. Appl. Physiol.  62: 1277-1284.

35    Pirnay, F.; DuJardin J.; DeRoanne, R.;  Petit, J. M. (1971) Muscular exercise during intoxication by carbon
               monoxide. J. Appl. Physiol. 31: 573-575.

        Pitt, B. R.; Radford, E. P.; Gurtner, G. H.; Traystman, R. J. (1979) Interaction of carbon monoxide and
               cyanide on cerebral circulation and metabolism. Arch. Environ. Health 34: 354-359.
40
        Pittman, R. N.; Duling, B. R. (1973) Oxygen sensitivity of vascular smooth muscle. Microvasc. Res.
               6: 202-122.

        Ponte, J.; Purves, M. J. (1974) The role of the  carotid body chemoreceptors and carotid sinus baroreceptors in
45            the control of cerebral blood vessels. J.  Physiol. (London)  237: 315-340.

        Preziosi,  T. J.; Lindenberg, R.; Levy, D.; Christenson, M. (1970) An experimental investigation in animals of
               the functional and morphologic effects of single and repeated exposures to high  and low concentrations of
               carbon monoxide. Ann. N.  Y. Acad. Sci.  174: 369-384.
50
        Prigge, E.; Hochrainer, D. (1977) Effects of carbon monoxide inhalation on erythropoiesis and cardiac
               hypertrophy in fetal rats.  Toxicol. Appl. Pharmacol. 42: 225-228.
          March  12, 1990                              10-208      DRAFT-DO NOT QUOTE OR CITE

-------
        Purser, D. A.; Berrill, K. R. (1983) Effects of carbon monoxide on behavior in monkeys in relation to human
               fire hazard. Arch. Environ. Health 38: 308-315.

        Putz, V. R.; Johnson, B. L.; Setzer, J. V. (1976) Effects of CO on vigilance performance: effects of low level
 5             carbon monoxide on divided attention, pitch discrimination, and the auditory evoked potential.
               Cincinnati, OH: U. S. Department of Health, Education, and Welfare, National Institute for Occupational
               Safety and Health. Available from: NTIS, Springfield, VA; PB-274219.

        Putz, V. R.; Johnson, B. L.; Setzer, J. V. (1979) A comparative study of the effects of carbon monoxide and
10             methylene chloride on human performance. J. Environ. Pathol. Toxicol. 2: 97-112.

        Radford, E. P.; Pitt, B.; Halpin, B.; Caplan, Y.; Fisher, R.; Schweda, P. (1976) Study of fire deaths in
               Maryland: Sept. 1971-Jan. 1974. In: Physiological and toxicological aspects of combustion products:
               international symposium; March 1974; Salt Lake City, UT. Washington, DC: National Academy of
15             Sciences;  pp. 26-35.

        Ramsey, J. M. (1972) Carbon monoxide,  tissue hypoxia, and sensory psychomotor response in hypoxaemic
               subjects. Clin.  Sci. 42: 619-625.

20      Ramsey, J. M. (1973) Effects of single exposures of carbon monoxide on sensory and psychomotor response.
               Am. Ind. Hyg. Assoc. J.: 212-216.

        Raven, P. B.; Drinkwater, B. L.; Horvath, S. M.; Ruhling, R. O.; Gliner, J. A.; Sutton, J. C.; Bolduan, N. W.
               (1974a) Age, smoking habits, heat stress, and their interactive effects with carbon monoxide and
25             peroxyacetylnitrate on man's aerobic power. Int. J. Biometeorol.  18: 222-232.

        Raven, P. B.; Drinkwater, B. L.; Ruhling, R. O.; Bolduan, N.; Taguchi, S.; Gliner, J.; Horvath, S. M. (1974b)
               Effect of carbon monoxide and peroxyacetyl nitrate on man's maximal aerobic capacity. J. Appl.
               Physiol. 36: 288-293.
30
        Renaud, S.; Blache, D.; Dumont, E.; Thevenon, C.; Wissendanger, T. (1984) Platelet function after cigarette
               smoking in relation to nicotine and carbon monoxide. Clin. Pharmacol. Ther. 36: 389-395.

        Reynolds, J. E. F., ed. (1982) Martindale's: the extra pharmacopoeia. 28th ed. London, United Kingdom: The
35             Pharmaceutical Press; p.  804.

        Robertson, G.; Lebowitz, M. D. (1984) Analysis of relationships between symptoms and environmental factors
               over time. Environ. Res. 33:  130-143.

40      Robinson, N.  B.; Barie, P.  S.; Halebian,  P. H.; Shires, G. T. (1985) Distribution of ventilation and perfusion
               following acute carbon monoxide poisoning. In: 41st annual forum on fundamental surgical problems held
               at the  71st annual clinical congress of the American College of Surgeons; October;  Chicago, IL. Surg.
               Forum 36: 115-118.

45      Roche, S.; Horvath, S.; Gliner,  J.; Wagner, J.; Borgia, J. (1981) Sustained visual attention and carbon
               monoxide: elimination of adaptation effects. Hum. Factors 23: 175-184.

        Rogers, W. R.; Bass,  R. L., Ill; Johnson, D. E.; Kruski, A. W.; McMahan, C. A.; Montiel, M. M.; Mott, G.
               E.; Wilbur, R. L.; McGill, H. C., Jr. (1980) Atherosclerosis-related  responses to cigarette smoking in
50             the baboon. Circulation 61: 1188-1193.

        Rogers, W. R.; Carey, K. D.; McMahan, C. A.; Montiel, M. M.; Mott, G.  E.; Wigodsky, H. S.; McGill, H.
               C., Jr. (1988) Cigarette smoking,  dietary hyperlipidemia, and experimental atherosclerosis in the baboon.
               Exp. Mol. Pathol. 48: 135-151.


         March 12,  1990                             10-209      DRAFT-DO NOT QUOTE OR CITE

-------
       Root, W. S. (1965) Carbon monoxide. In: Fenn, W. O.; Rahn, H., eds. Handbook of physiology: a critical,
              comprehensive presentation of physiological knowledge and concepts. Section 3: Respiration, volume II.
              Washington, DC: American Physiological Society; pp. 1087-1098.

 5     Rosenberg, A. A.; Jones, M. D., Jr.; Traystman, R. J.; Simmons, M. A.; Molteni, R. A. (1982) Response of
              cerebral blood flow to changes in P^Q  in fetal, newborn, and adult sheep.  Am. J. Physiol. 242:
              H862-H866.                       2

       Rosenkrantz, H.; Grant, R. J.; Fleischman, R. W.; Baker, J. R. (1986) Marihuana-induced embryotoxicity in the
10            rabbit. Fundam. Appl.  Toxicol. 7: 236-243.

       Rosenthal, M.; LaManna, J. C.; Jobsis, F. F.; Levasseur, J. E.; Kontos, H. A.; Patterson, J. L. (1976) Effects
              of respiratory gases on cytochrome a in intact cerebral cortex: is there a critical PQ ? Brain Res. 108:
              143-154.                                                                   2
15
       Roth, R. A., Jr.; Rubin, R. J.  (1976a) Role of blood flow in carbon monoxide- and hypoxic hypoxia-induced
              alterations in hexobarbital metabolism in rats. Drug Metab. Dispos. 4: 460-467.

       Roth, R. A., Jr.; Rubin, R. J.  (1976b) Comparison of the effect of carbon monoxide and of hypoxic hypoxia.  II.
20            Hexobarbital metabolism in the isolated, perfused rat liver. J. Pharmacol. Exp.  Ther. 199: 61-66.

       Roughton, F. J. W.; Darling,  R. C.  (1944) The effect of carbon monoxide on the oxyhemoglobin dissociation
              curve. Am. J. Physiol. 141:  17-31.

25     Roy, C.  S.; Brown, J. G. (1879) The blood pressure and its variations in the arterioles, capillaries, and smaller
              veins.  J. Physiol. 2: 323-359.

       Rubio, R.; Berne, R. M. (1969) Release of adenosine by the normal myocardium in dogs and its relationship to
              the regulation of coronary arteries. Circ. Res. 25: 407-415.
30
       Rubio, R.; Beme, R. M.; Bockman, E. L.; Curnish, R. R. (1975) Relationship between adenosine concentration
              and oxygen supply in rat brain. Am. J.  Physiol. 228: 1896-1902.

       Rummo, N.; Sarlanis, K. (1974) The effect of carbon monoxide on several measures of vigilance in a simulated
35            driving task. J. Saf. Res. 6:  126-130.

       Russek, M.; Fernandez, F.; Vega, C. (1963) Increase of cerebral blood flow produced  by low dosages of
              cyanide. Am. J. Physiol. 204: 309-312.

40     Salvatore, S. (1974) Performance decrement caused by mild carbon monoxide levels on two visual functions. J.
              Saf. Res. 6: 131-134.

       Santiago, T. V.; Edelmain, N. H. (1976) Mechanism of the ventilatory response to carbon monoxide. J. Clin.
              Invest. 57: 977-986.
45
       Schaad, G.; Kleinhanz, G.; Piekarski, C. (1983) Zum einfluss von kohlenmonoxid in der atemluft auf die
              psychophysische leistungsfahigkeit. Wehrmed. Monatsschr. 10: 423-430.

       Schrot, J.; Thomas, J. R. (1986) Multiple schedule performance changes during carbon monoxide exposure.
50            Neurobehav. Toxicol. Teratol. 8: 225-230.

       Schrot, J.; Thomas, J. R.; Robertson, R. F. (1984) Temporal changes in repeated acquisition behavior after
              carbon monoxide exposure. Neurobehav. Toxicol. Teratol. 6: 23-28.
         March 12, 1990                              10-210     DRAFT-DO NOT QUOTE OR CITE

-------
        Schulte, J. H. (1963) Effects of mild carbon monoxide intoxication. Arch. Environ. Health 7: 30-36.

        Schwetz, B. A.; Smith, F. A.; Leong, B. K. J.; Staples, R. E. (1979) Teratogenic potential of inhaled carbon
               monoxide in mice and rabbits. Teratology 19: 385-392.
 5
        Sekiya, S.; Sato, S.; Yamaguchi, H.; Harumi, K. (1983) Effects of carbon monoxide inhalation on myocardial
               infarct size following experimental coronary artery ligation. Jpn. Heart J. 24: 407-416.

        Selvin, S.; Sacks, S. T.; Merrill, D. W.; Winkelstein, W. (1980) Relationship between cancer incidence and two
10             pollutants (total suspended particulate and carbon monoxide) for the San Francisco bay area. Berkeley,
               CA: Lawrence Berkeley Laboratory. Available from: NTIS, Springfield, VA; LBL-10847.

        Seppanen, A.  (1977) Physical work capacity in relation to carbon monoxide inhalation and tobacco smoking.
               Ann. Clin. Res. 9: 269-274.
15
        Seppanen, A.; Hakkinen, J.  V.; Tenkku, M. (1977) Effect of gradually increasing carboxyhemoglobin saturation
               on visual perception and psychomotor performance of smoking and nonsmoking subjects. Ann. Clin. Res.
               9: 314-319.

20      Shalit, M.  N.; Reinmuth, O. M.; Shimojyo, S.; Scheinberg, P.  (1967)  Carbon dioxide and cerebral circulatory
               control.  III. The effects of brainstem lesions. Arch. Neurol. 17: 342-353.

        Shephard, R. J.  (1983) Carbon monoxide the silent killer. Springfield, IL: Charles C. Thomas. (American lecture
               series no. 1059).
25
        Shephard, R. J.  (1984) Athletic performance and urban air pollution. Can. Med. Assoc. J. 131: 105-109.

        Sheppard, D.; Distefano, S.; Morse, L.; Becker, C. (1986) Acute effects of routine firefighting on lung function.
               Am. J. Ind. Med. 9: 333-340.
30
        Sheps, D. S.; Adams, K. F., Jr.; Bromberg, P.  A.; Goldstein, G. M.;  O'Neil, J. J.; Horseman, D.; Koch,  G.
               (1987) Lack of effect of low levels of carboxyhemoglobin on cardiovascular function in patients with
               ischemic heart disease. Arch. Environ. Health 42: 108-116.

35      Sheps, D. S.; Herbst, M. C.; Hinderliter, A. L.; Adams, K. F.; Ekelund, L.  G.; O'Neil, J. J.; Goldstein, G.
               M.; Bromberg, P. A.;  Herdt, J.; Ballenger, M.;  Davis,  S. M.; Koch, G. (1989) Effects of 4% and  6%
               carboxyhemoglobin on arrhythmia production in patients with coronary artery disease. Submitted for
               publication.

40      Siesjo, B. K.; Nilsson, L. (1971) The influence of arterial hypoxemia upon labile phosphates and upon
               extracellular and intracellular lactate and pyruvate concentrations in the rat brain.  Scand. J. Clin. Lab.
               Invest. 27: 83-96.

        Siggaard-Andersen,  J.; Bonde-Peterson, F.; Hansen,  T. J.; Mellemgaard, K.  (1968) Plasma volume and vascular
45             permeability during hypoxia and carbon monoxide exposure. Scand. J. Clin. Lab. Invest. Suppl.  103:
               39-48.

        Singh, J. (1986) Early behavioral alterations in mice following prenatal carbon monoxide exposure.
               Neurotoxicology 7: 475-482.
50
        Singh, J.; Scott, L.  H. (1984) Threshold for carbon monoxide induced fetotoxicity. Teratology 30: 253-257.

        Sjostrand, T. L.  (1948) Brain volume, diameter of the blood vessels in the piamater,  and  intracranial pressure in
               acute carbon monoxide poisoning. Acta Physiol.  Scand.  15:  351-361.


         March 12, 1990                              10-211      DRAFT-DO NOT QUOTE OR CITE

-------
        Skinhoj, E. (1966) Regulation of cerebral blood flow as a single function of interstitial pH. Acta Neurol. Scand.
               42: 604-607.

        Smith, E. E.; Crowell, J. W. (1967) Role of an increased hematocrit in altitude acclimatization. Aerospace Med.
 5            38: 39-43.

        Smith, D. J.; Vane, J. B. (1966) Effects of oxygen on vascular and other smooth muscles. J. Physiol. (London)
               186: 284-294.

10     Smith, A. D. M.; Duckett, S.; Waters, A. H.  (1963) Neuropathological changes in chronic cyanide intoxication.
               Nature (London) 200: 179-181.

        Smith, M.  D.; Merigan, W. H.; Mclntire, R. W. (1976a) Effects of carbon monoxide on
               fixed-consecutive-number performance in rats. Pharmacol. Biochem. Behav. 5: 257-262.
15
        Smith, P. W.; Crane, C. R.; Sanders,  D. C.; Abbott, J.  K.; Endecott, B. S. (1976b) Effects of exposure to
               carbon monoxide and hydrogen cyanide. In: Physiological and toxicological aspects of combustion
               products: international symposium; March 1974;  Salt Lake City, UT. Washington, DC: National
               Academy of Sciences; pp. 75-88.
20
        Snella, M.-C.; Rylander, R. (1979) Alteration in local and systemic immune capacity after exposure to bursts of
               CO. Environ. Res. 20: 74-79.

        Sokoloff, L. (1978) Local cerebral energy metabolism: its relationships to local functional activity and blood
25            flow. In: Cerebral vascular smooth muscle and its control. Amsterdam, The Netherlands:
               Elsevier/North-Holland, Inc.; pp. 171-197. (Ciba Foundation symposium no. 56).

        Sparrow, D.; Bosse,  R.; Rosner, B.; Weiss, S. T. (1982) The effect of occupational exposure on pulmonary
               function: a longitudinal  evaluation of fire fighters and nonfire fighters. Am. Rev. Respir. Dis.  125:
30            319-322.

        Stavraky, G. W. (1936) Response of cerebral blood vessels to electrical stimulation of the thalamus and
               hypothalamus regions. AMA Arch. Neurol. Psychiatry 35: 1002-1028.

35     Stender, S.; Astrup, P.; Kjeldsen, K.  (1977) The effect of carbon monoxide on cholesterol in the aortic wall of
               rabbits. Atherosclerosis  28: 357-367.

        Stern, S.; Ferguson,  R.  E.; Rapaport,  E. (1964) Reflex pulmonary vasoconstriction due to stimulation of the
               aortic body by nicotine. Am. J. Physiol. 206:  1189-1195.
40
        Stern, F. B.; Lemen, R. A.; Curtis, R. A. (1981) Exposure of motor vehicle examiners to carbon monoxide: a
               historical prospective mortality study. Arch. Environ. Health 36: 59-66.

        Stern, F. B.; Halperin, W. E.; Hornung, R. W.; Ringenburg, V. L.; McCammon, C. S. (1988) Heart disease
45            mortality among bridge  and tunnel officers exposed to carbon monoxide. Am. J. Epidemiol. 128:
               1276-1288.

        Stewart, R. D.; Peterson, J.  E.; Baretta, E. D.; Bachand, R. T.; Hosko, M. J.; Herrmann, A. A. (1970)
               Experimental human exposure  to carbon monoxide. Arch. Environ. Health 21: 154-164.
50
        Stewart, R. D.; Newton, P. E.; Hosko, M. J.; Peterson, J. E.; Mellender, J. W. (1972) The effect of carbon
               monoxide on time perception,  manual coordination, inspection, and arithmetic. In:  Weiss, B.; Laties, V.
               G., eds. Behavioral toxicology. New York, NY:  Plenum Press; pp. 29-60.
         March 12, 1990                              10-212     DRAFT-DO NOT QUOTE OR CITE

-------
        Stewart, R. D.; Peterson, J. E.; Fisher, T. N.; Hosko, M. J.; Baretta, E. D.; Dodd, H. C.; Herrmann, A. A.
               (1973a) Experimental human exposure to high concentrations of carbon monoxide. Arch. Environ. Health
               26: 1-7.

 5     Stewart, R. D.; Newton, P. E.; Hosko, M. J.; Peterson, J. E. (1973b) Effect of carbon monoxide on time
               perception. Arch. Environ. Health 27: 155-160.

        Stewart, R. D.; Newton, P. E.; Kaufman, J.; Forster, H. V.; Klein, J. P.; Keelen, M. H., Jr.; Stewart, D. J.;
               Wu, A.; Hake, C. L. (1978) The effect of a rapid 4% carboxyhemoglobin saturation increase on maximal
10            treadmill exercise. New York, NY: Coordinating Research Council, Inc.; report no.
               CRC-APRAC-CAPM-22-75. Available from: OTIS, Springfield, VA; PB-296627.

        Storm, J. E.; Fechter, L. D. (1985a) Alteration in the postnatal ontogeny of cerebellar norepinephrine content
               following chronic prenatal carbon monoxide. J. Neurochem. 45: 965-969.
15
        Storm, J. E.; Fechter, L. D. (1985b) Prenatal carbon monoxide exposure differentially affects postnatal weight
               and monoamine concentration of rat brain regions. Toxicol. Appl. Pharmacol. 81: 139-146.

        Storm, J. E.; Valdes, J. J.; Fechter, L. D. (1986) Postnatal alterations in cerebellar GABA content, GABA
20            uptake  and morphology following exposure to carbon monoxide early in development. Dev. Neurosci. 8:
               251-261.

        Stupfel, M.; Bouley, G. (1970) Physiological and biochemical effects on rats  and mice exposed to small
               concentrations of carbon monoxide for long periods. Ann. N. Y. Acad. Sci. 174: 342-368.
25
        Stupfel, M.; Mordelet-Dambrine, M.;  Vauzelle, A. (1981) COHb formation and  acute carbon monoxide
               intoxication in adult male rats and guinea-pigs. Bull. Eur. Physiopathol. Respir. 17: 43-51.

        Styka, P. E.; Penney, D. G. (1978) Regression of carbon monoxide-induced cardiomegaly. Am. J. Physiol. 235:
30            H516-H522.

        Sultzer, D. L.; Brinkhous, K. M.; Reddick, R. L.; Griggs, T. R. (1982) Effect of carbon monoxide on
               atherogenesis in normal pigs and pigs with von Willebrand's disease.  Atherosclerosis 43: 303-319.

35     Svendsen, K. H.; Kuller, L. H.; Martin, M. J.;  Ockene, J. K. (1987) Effects of  passive smoking in the multiple
               risk factor intervention trial. Am. J. Epidemiol. 126: 783-795.

        Swiecicki, W. (1973) Wplyw wibracji  i treningu fizycznego na przemiane weglowodanowa u szczurow zatrutych
               tlenkiem wegla [The effect of vibration and physical training on carbohydrate metabolism in rats
40            intoxicated with carbon monoxide]. Med. Pr. 34: 399-405.

        Sylvester, J. T.; Scharf, S.  M.; Gilbert, R. D.; Fitzgerald, R. S.; Traystman, R. J. (1979) Hypoxic and CO
               hypoxia in dogs: hemodynamics, carotid reflexes, and catecholamines. Am. J. Physiol. 236: H22-H28.

45     Syvertsen, G. R.; Harris, J. A. (1973) Erythropoietin production in dogs exposed to high altitude and carbon
               monoxide. Am. J. Physiol. 225: 293-299.

        Tachi, N.; Aoyama, M. (1983) Effect  of cigarette smoke and carbon monoxide inhalation by gravid rats on the
               conceptus weight. Bull. Environ. Contam. Toxicol. 31: 85-92.
50
        Tachi, N.; Aoyama, M. (1986) Effect  of restricted food supply to pregnant rats inhaling carbon monoxide on
               fetal weight, compared with cigarette smoke exposure. Bull. Environ. Contam. Toxicol. 37: 877-882.
         March  12, 1990                              10-213     DRAFT-DO NOT QUOTE OR CITE

-------
       Theodore, J.; O'Donnell, R. D.; Back, K. C. (1971) Toxicological evaluation of carbon monoxide in humans and
              other mammalian species. JOM J. Occup. Med. 13: 242-255.

       Thomsen, H. K. (1974) Carbon monoxide-induced atherosclerosis in primates: an electron-microscopic  study on
 5            the coronary arteries of Macaco irus monkeys.  Atherosclerosis 20: 233-240.

       Thomsen, H. K.; Kjeldsen, K. (1975) Aortic intimal injury in rabbits: an evaluation of a threshold limit. Arch.
              Environ. Health 30: 604-607.

10     Traystman, R. J. (1978) Effect of carbon monoxide hypoxia and hypoxic hypoxia on cerebral circulation. In:
              Otto, D. A., ed. Multidisciplinary perspectives in event-related brain potential research: proceedings of
              the fourth international congress on event-related slow potentials  of the brain (EPIC IV); April  1976;
              Hendersonville, NC. Washington, DC: U. S. Environmental Protection Agency, Office of Research and
              Development;  pp. 453-458; EPA  report no. EPA-600/9-77-043.  Available from: NTIS, Springfield, VA;
15            PB-297137.

       Traystman, R. J.; Fitzgerald, R. S. (1977) Cerebral circulatory  responses to hypoxic hypoxia and carbon
              monoxide hypoxia in carotid baroreceptor and chemoreceptor denervated dogs. Acta Neurol.  Scand.
              56(suppl. 64): 294-295.
20
       Traystman, R. J.; Fitzgerald, R. S. (1981) Cerebrovascular response to hypoxia in baroreceptor- and
              chemoreceptor-denervated dogs. Am. J. Physiol.  241: H724-H731.

       Traystman, R. J.; Fitzgerald, R. S.; Loscutoff, S. C. (1978) Cerebral circulatory responses to arterial hypoxia in
25            normal and chemodenervated dogs.  Circ. Res. 42: 649-657.

       Traystman, R. J.; Gurtner, G. H.; Rogers,  M. C.; Jones, M. D., Jr.; Koehler, R. C. (1981) A possible role of
              oxygenases in the regulation of cerebral blood flow. In:  Kovach, A. G. B.; Sandor, P.; KoMai, M., eds.
              Cardiovascular physiology: natural mechanisms. Budapest, Hungary:  Pergamon Press; pp.  167-177.
30            (Adv.  Physiol. Sci. v. 9).

       Trese, M. T.; Krohel, G. B.; Hepler, R.  S. (1980) Ocular effects of chronic carbon monoxide exposure. Ann.
              Ophthalmol. 12: 536-538.

35     Turino, G. M. (1981) Effect of carbon monoxide on the  cardiorespiratory system; carbon monoxide  toxicity:
              physiology and biochemistry. Circulation 63: 253A-259A.

       Turner, D. M.; Lee, P. N.; Roe, F. J. C.;  Gough, K.  J. (1979) Atherogenesis in the White Carneau pigeon:
              further studies of the role of carbon monoxide and dietary cholesterol. Atherosclerosis 34:  407-417.
40
       U. S. Environmental Protection  Agency. (1979) Air quality criteria for carbon monoxide. Research Triangle
              Park, NC: Office of Health and Environmental Assessment, Environmental Criteria and Assessment
              Office; EPA report no. EPA-600/8-79-022. Available from:  NTIS, Springfield, VA; PB81-244840.

45     U. S. Environmental Protection  Agency. (1984) Revised evaluation of health effects associated with  carbon
              monoxide exposure: an addendum to the 1979 EPA air quality criteria document for carbon monoxide.
              Research Triangle Park, NC: Office of Health and Environmental Assessment, Environmental Criteria
              and Assessment Office; EPA report no. EPA-600/9-83-033F. Available from: NTIS, Springfield, VA;
              PB85-103471.
50
       Van Houdt, J. J.; Alink, G.  M.; Boleij, J.  S. M. (1987) Mutagenicity of airborne particles related to
              meteorological and air pollution parameters.  Sci. Total Environ.  61: 23-36.
         March 12, 1990                             10-214      DRAFT-DO NOT QUOTE OR CITE

-------
       Vanoli, E.; De Ferrari, G. M.; Stramba-Badiale, M.; Bickestaff, L. H.; Dickey, T.; Farber, J. P.; Schwartz, P.
               J. (1986) Carbon monoxide and transient ischemia in conscious dogs with a healed myocardial infarction.
               Presented at: 70th annual meeting of the Federation of American Societies for Experimental Biology;
               April; St. Louis, MO. Fed. Proc. Fed. Am. Soc. Exp. Biol. 45: 778.
 5
       Vanoli, E.; De Ferrari, G. M.; Stramba-Badiale, M.; Farber, J. P.; Schwartz, P. J. (1989) Carbon monoxide and
               lethal arrhythmias in conscious dogs with a healed myocardial infarction. Am. Heart J. 117: 348-357.

       Venning, H.; Roberton, D.;  Milner, A. D. (1982) Carbon monoxide poisoning in an infant. Br. Med. J. 284:
10             651.

       Vogel, J. A.; Gleser, M. A.  (1972) Effect of carbon monoxide on oxygen transport during exercise. J. Appl.
               Physiol. 32: 234-239.

15     Vogel, J. A.; Gleser, M. A.; Wheeler, R. C.; Whitten, B. K.  (1972) Carbon monoxide and physical work
               capacity. Arch. Environ.  Health 24: 198-203.

       Vollmer, E. P.; King, B. G.; Birren, J. E.; Fisher, M. B.  (1946) The effects of carbon monoxide on three types
               of performance, at simulated altitudes of 10,000 and 15,000 feet. J.  Exp. Psychol. 36: 244-251.
20
       Von Post-Lingen, M.-L. (1964) The significance of exposure to small concentrations of carbon monoxide. Proc.
               R. Soc. Med. 57: 1021-1029.

       Von Restorff, W.;  Hebisch, S.  (1988) Dark adaptation of the eye during carbon monoxide exposure in smokers
25             and nonsmokers. Aviat. Space Environ.  Med. 59: 928-931.

       Vreman, H. J.; Kwong, L. K.; Stevenson, D. K. (1984) Carbon monoxide in blood: an improved microliter
               blood-sample collection system, with rapid analysis by gas chromatography. Clin. Chem.
               (Winston-Salem, NC) 30: 1382-1386.
30
       Wahl, M.; Kuschinsky, W. (1976) The dilatory action of adenosine on  pial arteries of cats and its inhibition by
               theophylline. Pfluegers Arch. 362: 55-59.

       Wahl, M.; Kuschinsky, W. (1979) The dilating  effect of bistamine on pial arteries of cats and its mediation by
35             H2 receptors. Circ. Res. 44:  161-165.

       Wahl, M.; Deetjin, P.; Thuran, D.;  Ingvar, D.  H.; Lassen, N. A. (1970) Micropuncture evaluation of the
               importance of perivascular pH for the arteriolar diameter on the brain surface. Pfluegers Arch. 316:
               152-163.
40
       Wald, N. J.; Boreham, J.;  Bailey, A.;  Ritchie, C.;  Haddow, J. E.; Knight,  G.  (1984) Urinary cotinine as marker
               of breathing other people's tobacco smoke. Lancet (8370): 230-231.

       Wall, M. A.; Johnson, J.;  Jacob, P.; Benowitz, N. L. (1988) Cotine in the serum, saliva, and urine of
45             nonsmokers, passive  smokers, and active smokers. Am. J. Public Health 78: 699-701.

       Ward, A.; Wheatley, M. (1947) Sodium cyanide: sequence of changes of activity induced at various levels of the
               central nervous system. J. Neuropathol.  Exp. Neurol. 6:  292-294.

50     Webster, W. S.; Clarkson, T. B.; Lofland, H. B. (1970) Carbon monoxide-aggravated atherosclerosis in the
               squirrel monkey. Exp. Mol.  Pathol. 13: 36-50.
         March 12,  1990                             10-215      DRAFT-DO NOT QUOTE OR CITE

-------
       Weir, F. W.; Rockwell, T. H.; Mehta, M. M.; Attwood, D. A.; Johnson, D. F.; Herrin, G. D.; Anglen, D.
              M.; Safford, R. R. (1973) An investigation of the effects of carbon monoxide on humans in the driving
              task: final report. Columbus, OH: The Ohio State University Research Foundation; contract no.
              68-02-0329  and CRC-APRAC project CAPM-9-69. Available from: NTIS, Springfield, VA; PB-224646.
 5
       Weiser, P. C.; Morrill, C. G.; Dickey, D. W.; Kurt, T. L.; Cropp, G. J.  A. (1978) Effects of low-level carbon
              monoxide exposure on the adaptation of healthy young men to aerobic work at an altitude of 1,610
              meters. In: Folinsbee, L.  J.; Wagner, J. A.; Borgia, J. F.; Drinkwater, B. L.; Gliner, J. A.; Bedi, J. F.,
              eds. Environmental stress: individual human adaptations. New York, NY: Academic Press,  Inc.; pp.
10            101-110.

       Weiss, H. R.; Cohen, J. A. (1974) Effects of low levels of carbon monoxide on rat brain and muscle tissue PQ .
              Environ. Physiol. Biochem. 4: 31-39.

15     Weissbecker, L,; Carpenter, R. D.; Luchsinger, P. C.;  Osdene, T. S.  (1969) In vitro alveolar macrophage
              viability: effect of gases.  Arch. Environ. Health 18: 756-759.

       Wells, L. L. (1933) The prenatal effect of carbon monoxide on albino rats and the resulting neuropathology.
              Biologist 15: 80-81.
20
       Wilks, S. S.; Tomashefski, J. F.; Clark,  R. T., Jr. (1959) Physiological effects of chronic exposure to carbon
              monoxide. J. Appl. Physiol. 14:  305-310.

       Williams, I. R.; Smith, E. (1935) Blood  picture, reproduction, and general condition during daily exposure to
25            illuminating gas. Am. J.  Physiol. 110: 611-615.

       Wilson, D.  F.; Erecinska, M.; Drown, C.; Silver, I. A. (1979) The oxygen dependence of cellular energy
              metabolism. Arch. Biochem. Biophys.  195: 485-493.

30     Wilson, D.  A.; Manley, D. F.; Feldman, M. A.; Traystman, R. J. (1987) Influence of chemoreceptors on
              neurohypophyseal blood flow during hypoxic hypoxia. Circ. Res. 61: II94-II101.

       Winn, H. R.;  Rubio, R.; Berne,  R. M. (1981) Brain adenosine concentration during hypoxia in rats. Am. J.
              Physiol. 241: H235-H242.
35
       Winneke, G. (1974) Behavioral effects of methylene chloride and carbon monoxide as assessed by sensory and
              psychomotor performance. In: Xintaras, C.; Johnson, B. L.; de Groot,  I.,  eds. Behavioral toxicology:
              early detection of occupational hazards [proceedings of a workshop; June 1973; Cincinnati,  OH].
              Cincinnati, OH: U. S. Department of Health, Education, and Welfare, National  Institute for Occupational
40            Safety and Health; pp. 130-144; DHEW publication no. (NIOSH) 74-126. Available from:  NTIS,
              Springfield, VA; PB-259322.

       Wouters, E. J. M.;  de Jong,  P. A.; Cornelissen, P.  J. H.; Kurver, P. H. J.; van Oel, W. C.; van Woensel, C.
              L. M. (1987) Smoking and low birth weight: absence of influence by carbon monoxide? Eur. J. Obstet.
45            Gynecol.  Reprod. Biol. 25: 35-41.

       Wright, G.  R.; Shephard, R.  J. (1978a) Brake reaction time - effects of age, sex, and carbon monoxide. Arch.
              Environ. Health 33:  141-149.

50     Wright, G.  R.; Shephard, R.  J. (1978b)  Carbon monoxide exposure and auditory duration discrimination. Arch.
               Environ. Health 33: 226-235.

       Wright, G.; Randell, P.; Shephard, R. J. (1973) Carbon monoxide and driving skills. Arch. Environ. Health 27:
               349-354.


         March 12, 1990                              10-216     DRAFT-DO NOT  QUOTE OR CITE

-------
       Xintaras, C.; Johnson, B. L.; Ulrich, C. E.; Terrill, R. E.; Sobecki, M. F. (1966) Application of the evoked
              response technique in air pollution toxicology. Toxicol. Appl. Pharmacol. 8: 77-87.

       Yamamoto, K. (1976) Acute combined effects of HCN and CO with the use of the combustion products from
 5            PAN (polyacrylonitrile) - gauze mixtures. Z. Rechtsmed. 78: 303-311.

       Young, S. H.; Stone, H. L. (1976) Effect of a reduction in arterial oxygen content (carbon monoxide) on
              coronary flow. Aviat. Space Environ. Med. 47: 142-146.

10     Zebro, T.; Littleton, R. J.; Wright,  E. A. (1976) Adaptation of mice to carbon monoxide and the effect of
              splenectomy.  Virchows Arch. A: Pathol. Anat. Histopathol. 371: 35-51.

       Zebro, T.; Wright, E. A.; Littleton, R. J.; Prentice, A. I. D. (1983) Bone changes in mice after prolonged
              continuous exposure to a high concentration of carbon monoxide. Exp. Pathol. 24: 51-67.
15
       Zorn, H. (1972) The  partial oxygen pressure in the brain and liver at subtoxic concentrations of carbon
              monoxide. Staub Reinhalt. Luft 32: 24-29.

       Zwart, A.; Buursma,  A.;  Oeseburg, B.; Zijlstra, W. G. (1981) Determination of hemoglobin derivatives with the
20            IL 282 CO-oximeter as compared with a manual spectrophotometric five-wavelength method. Clin. Chem.
              (Winston-Salem, NC) 27: 1903-1907.

       Zwart, A.; Buursma,  A.;  Van Kempen, E. J.;  Oeseburg, B.; van der Ploeg, P. H. W.; Zijlstra, W. G. (1981) A
              multi-wavelength spectrophotometric method for the simultaneous determination of five  haemoglobin
25            derivatives. J. Clin.  Chem. Clin. Biochem. 19: 457-463.
         March 12,  1990                              10-217     DRAFT-DO NOT QUOTE OR CITE

-------
       11.  COMBINED EXPOSURE OF CARBON MONOXIDE
                  WITH  OTHER POLLUTANTS, DRUGS,
                     AND ENVIRONMENTAL FACTORS
       11.1 HIGH ALTITUDE EFFECTS OF CARBON MONOXIDE
       11.1.1  Introduction
           Precise estimates of the number of people exposed to CO at high altitude are not readily
       available.  As of 1980, however, more than 4.2 (Lindsey, 1989) million people were living at
10     altitudes in excess of 1524 m (5000 ft).  Moreover, estimates obtained from several states
       with mountainous regions (i.e. California, Nevada, Hawaii, and Utah) indicate that more than
       35 million tourists may sojourn in high altitude areas during the summer and winter months.
           The potential effects on human health of inhaling CO at high altitudes are complex.
       Whenever CO binds to Hb it reduces the amount of Hb available to carry O2.  People at high
15     altitudes already live in a state of hypoxemia, however, because of the reduced partial
       pressure of oxygen (POj) in the air. Carbon monoxide, by binding to Hb, intensifies the
       hypoxemia existing at high altitudes by further reducing transport of O2 to the tissues.  Hence,
       the effects of CO and high altitude usually are considered to be additive.
           This consideration does not take into account the fact that within hours (perhaps sooner)
20     of arrival at high altitude, however, hemoconcentration occurs which increases the Hb
       concentration. The increased Hb concentration offsets the decreased O2 saturation and
       restores O2 concentration to pre-ascent levels.  Consequently, the simple additive model of
       COHb and altitude hypoxemia may be valid only during early altitude exposure.
           The visitor newly arrived to higher altitudes may be at greater risk from CO than the
25     adapted resident, however, because of a noncompensated respiratory alkalosis from
       hyperventilation, lower arterial Hb saturation without a compensatory absolute polycythemia
       (therefore greater hypoxemia) and hypoxia-induced tachycardia.  (See Chapter 12,
       Section 12.5 for further discussion of this topic.)
           Several factors tend to exacerbate ambient CO levels at high altitude (Kirkpatrick and
30     Reeser, 1976). For example, in mountain communities, automobile emissions are higher.
       Automobiles tuned for driving at 1610 m (5280 ft) emit almost 1.8 times more CO when
       March 12, 1990                        11-1     DRAFT - DO NOT QUOTE OR CITE

-------
       driven at 2438 m (8000 ft).  Automobiles tuned for driving at sea level emit almost four times
       more CO when driven at 2438 m.  Moreover, automobile emissions are increased by driving
       at reduced speeds, along steep grades, and under poor-driving conditions.  Therefore, large
       influxes of tourists driving automobiles tuned for sea-level conditions into high-altitude resort
 5     areas may drastically increase pollutant levels in general, and CO levels in particular
       (National Research Council, 1977). Newer automobile engine technologies, however, should
       significantly reduce CO emissions in general, as well as CO emissions at high altitude.
       Heating devices (space heaters and  fireplaces)  used for social effect, as well as warmth, are a
       second factor contributing to CO emissions in mountain resort areas. Finally, population
10     growth in mountain areas is concentrated along valley floors; this factor combined with the
       reduced volume of air available for pollutant dispersal in valleys causes pollutants, including
       CO, to  accumulate in mountain valleys.  As a result of these factors, the NAAQS for CO of
       9 ppm is exceeded frequently in Denver, CO, (altitude  1610 m) during the winter months
       (Haagenson, 1979).
15          Because of concern for CO exposure at high altitudes, it has been suggested that the
       NAAQS for CO set at sea level is probably too high for altitudes of 1500 m and above.  An
       example of supporting data for this opinion were studies conducted before 1950 on the
       psychophysiological effects of high altitude and CO-induced hypoxia. These studies provided
       evidence for a concept that there are physiologically equivalent altitudes dependent on the
20     ambient concentration of CO. In 1976, the states of California and Nevada adopted ambient
       standards for the Lake Tahoe air basin (1900 m; 6231 ft), which were more stringent than the
       NAAQS (i.e., 6 ppm rather than 9  ppm).
            Mitchell et al.  (1979) justified this concept by  stating that "equivalent
       carboxyhemoglobin levels observed at sea level would occur during exposure to lower
25     ambient CO concentrations at 1500 m." The high-altitude standard was calculated from the
       model developed by Coburn et al.  (1965).  This model was developed  for quasi-steady-state
       responses to low CO concentrations, such as those produced endogenously, however, and was
       not intended to apply to other, exogenous sources of CO (see Chapter 9 for a description of
       the model).  Collier and Goldsmith (1983) acknowledged that an error was made in the
30     original calculations for the California-Nevada high-altitude standard.  They expanded the
        March 12, 1990                          11-2     DRAFT - DO NOT QUOTE OR CITE

-------
10
15
20
25
30
35
       computations to include factors relating to the ambient CO concentrations and altitude and
       concluded, based on their model calculations, that the expected altitude effect would be small.


       11.1.2 Carboxyhemoglobin Formation

            The effects of high altitude on COHb formation have been considered in a theoretical
       paper by Collier and Goldsmith (1983).  Transforming and rearranging the Coburn-Foster-
       Kane equation (Coburn et al., 1965), these workers derived an equation expressing COHb in
       terms of endogenous and exogenous sources of CO. Thus,
                            FICO (PB-47) + VCO Z
where:
                        SCO  =
                SCO
                FICO
                ?B
                VCO
                K
                Z
                                       106K
                                                                                    (11-1)
                                             K
       and
       and
                       M
                       SO,  =
                       DLCO
                       (YO
COHb (%)
Fraction inspired CO (ppm)
Barometric Pressure (torr)
Rate of CO production (mL«min-' STPD)
P7O2/(M x SO,)
1/DLCO + (PB-47/CVg
                            mean partial pressure of pulmonary capillary O2 (torr)
                            Haldane constant
                            O2 Hb
                            CO diffusing capacity (mL-min-'-torr-1)
                            alveolar ventilation (mL-min-1 STPD)
     According to this relationship, a given PCO will result in a higher percent COHb at
high altitudes (where PO2 is reduced).  Thus,  Collier and Goldsmith calculate that humans

breathing 8 ppm CO will have equilibrium COHb levels of 1.4% at sea level and 1.6, 1.8,

and 1.8%, respectively, at 1530, 3050, and 3660 m (Table 11-1). Moreover, these workers

calculate an increase in COHb at altitude even in the absence of inhaled CO (due to
endogenous production of CO).
40
       March 12, 1990
                                       11-3     DRAFT - DO NOT QUOTE OR CITE

-------
10
15
             TABLE 11-1. CALCULATED EQUILIBRIUM VALUES OF PERCENT COHb
                           AND PERCENT 0^ IN HUMANS EXPOSED
                           TO AMBIENT CO AT VARIOUS ALTITUDES
Ambient CO
(ppm)
0
4
8
12
16
Sea
%
COHb
0.20
0.8
1.4
2.1
2.7
level
%
02Hb
97.3
96.8
96.2
95.6
95.1
530
%
COHb
0.26
0.9
1.6
2.3
2.9
m
%
02Hb
93.6
93.0
92.5
91.9
91.4
3050
%
COHb
0.35
1.1
1.8
2.5
3.2
m
%
02Hb
82.4
82.1
81.7
81.3
80.9
3660
%
COHb
0.37
1.1
1.8
2.5
3.2
m
%
O2Hb
73.3
73.1
72.9
72.7
72.5
20
       Notes: The table is for unacclimatized, sedentary individuals at one level of activity (v*O2 = 500 ml^min-1).
       Source: Adapted from Collier and Goldsmith (1983).
25


       11.1.3  Cardiovascular Effects
30          There are studies comparing the cardiovascular responses to CO with those to high
       altitude, but there are relatively few studies of the cardiovascular responses to CO at high
       altitude (see Table 11-2).  Forbes et al. (1945) reported that CO uptake increased during  six
       minutes of exercise of varying intensity on a bicycle ergometer at 4877 m (16,000 ft). The
       increased CO uptake was caused by altitude hyperventilation stimulated by decreased  arterial
35     O2 tension and not by diminished barometric pressure.
            Pitts and Pace (1947) reported that pulse rate increased in response to the combined
       stress of high altitude and CO. The subjects were 10 healthy men who were exposed to
       simulated altitudes of 2134, 3048 and 4572 m (7000, 10,000, and 15,000 ft) and inhaled
       3000 or 6000 ppm CO to obtain COHb levels of 6 or 13%, respectively. The mean pulse
40     rate during exercise and the mean pulse rate during the first five minutes after exercise were
       correlated with and increased with the COHb concentration and simulated altitude.  The
       authors concluded that the response to a 1 % increase in COHb level was equivalent to that

       March 12,  1990                          11-4    DRAFT - DO NOT QUOTE OR  CITE

-------
2
§•
TABLE 11-2. SUMMARY OF EFFECTS OF CARBON MONOXIDE AT ALTITUDE
Exposure COHb
Alt = 4,877 m
3,000-4,000 ppm CO
6 min exercise
Alt = 1,524- 1,848m 5-10%
CO = 1,500-2,000 ppm

Alt - 3.070-4.555 m 5-22%
CO = 2,800-5.600 ppm



Simulated alt - 2,134, 6 and 13%
3,048, 4,572 m (16, 14.
11% 0;+ Nj) 3,000 or
6,000 ppm
CO treadmill exercise
Alt = 2,134-4,877 m 1.1-20.5%
CO = 100-300 ppm

MCO in rebrealhing -
system P,O2 varied from
650 to 40 mm Hg



CO administered at 2-75 %
constant rate


All = 305-3.109 m 4.77-6.66%
smokers

Subject
Human
(n=3

Human
(n=5)

Human
(n=20)



Human
(n=10)



Human
(n=4)

Dog
(n=31)




Dog
(n=4)


Human
(n=62)

Dependent Variable
Blood CO


Flicker fusion
frequency (FFF)

Critical flicker
frequency (CFF),
body sway (BS),
red visual field
(RVF)
Pulse rate




Visual sensitivity


MCO in blood





Rale of increase in
COHb


COHb levels


Results
CO uptake increased
with altitude

FFF decreased with
CO at altitude

CFF, BS, and RVF
impaired by altitude;
no added effect of
CO

Pulse rate during and
5 min after exercise
increased with altitude and
COHb

CO decreased visual sensitivity


No change in I4CO activity
in blood when P,O2 varied
from 40 to 650 mmHg; MCO
decreased lo 50% control
when P.O2 decreased below
40 mmHg
COHb increased at
constant up to 50%; at
50%, rate of COHb forma-
tion decreased
COHb in smokers higher
at altitude than at sea
level
Comments
Caused by altitude hyperventilation.


FFF not affected by altitude or COHb
alone; 8-10% COHb reduced altitude
tolerance by 1,215 m.
No correlation of any response with
COHb. Effects of CO may be masked by
compensatory effect.


Response to 1 % COHb equivalent to
increase in altitude of 335 ft.
The effects of CO and altitude
are additive.

Recovery from the detrimental
effects of CO lagged behind elimina-
tion of CO from blood.
With severe arterial hypoxemia
(P,O2 <40 mmHg) MCO shifts into
extravascular tissue.



Suggests that at high COHb levels
CO shifts into extravascular space.





Reference
Forbes et al. (1945)


Lillianthal & Fugitt
(1946)

Vollmer et al. (1946)




Pitta & Pace (1947)




Halperin et al. (1959)


Luomanmaki & Cobum
(1969)




Luomanmaki & Coburn
(1969)


Brewer et al. (1970)



-------
I
»— *
JO
\o
8












,_,
>— *
ON


O
l>
!-H
H
I


23
0
H
O
cj
O
i—3
M
0
O






TABLE 11-2 (cont'd). SUMMARY OF EFFECTS OF CARBON MONOXIDE AT ALTITUDE
Exposure COHb


Alt = 2,438 m 5%


Alt = 4,500 m 20%
CO = 4,300 ppm every
second hour for
3-5 hours
Alt = 3,109 m 0.4-7.14%
Smokers


Alt = 3,048 m 4.2%
CO bolus followed by
40 ppm.
Bicycle exercise

Alt = 1,610m 5%
CO = 100% bolus



Alt = 1,524 m 20%
CO = 160-200 ppm
6 weeks



Alt = 3, 100m 1.8-6.2%
Smokers





Subject


Human


Human
(n=16)


Human
(n=49)


Human
(n=12)



Human
(n=9)



Goat
(n=6)




Human
(n=44)





Dependent Variable


Visual sensitivity


Capillary perme-
ability to protein
(CP)

COHb and O2 affinities



Cardiac output, stroke
volume (SV), arterial-
mixed venous O2
difference (A-V)

Work performance




Cardiac index (CO,
Stoke volume (SV),
Heart rate (HR),
Ventricular con-
tractility (V^j)

Infant birth weights






Results


Visual sensitivity
decreased by CO at
altitude
CP increased with CO
but not with altitude;
plasma volume decreased
with altitude
COHb levels and O2
affinities decreased in poly-
cythemic smokers on
cessation of smoking
At altitude, cardiac
output increased and
SV and A-V decreased
in nonsmokers; no
effect on smokers.
Increased working HR,
and shortened post-
exercise LV ejection
lime; towered anaerobic
threshold
No effect on CI, SV,
HR, and V^ during
exposure



Maternal smoking
associated with 2-3
times greater reduc-
tion in infant birth
weight than at sea
level

Comments


5 % COHb depresses visual sensitivity
as much as 2,438-3,048 m. The effects
of altitude and CO are additive.
Increase in CP appears unique to CO
(nonhypoxic effect).


O2 dropped to lower than normal sea
level values in polycythemic
smokers on cessation of
smoking.
Smokers may be partially adapted to
hypoxic environments and CO.



CO impaired exercise performance to
same degree as at KB level.



After removal from CO, both
HR and V,,^, were depressed.




COHb levels measured in mothers
were inversely related to infant
birth weight.




Reference


McFarland (1970);
McFarland et al. (1944)

Parving (1972)



Brewer et al. (1974)



Wagner et al. (1978)




Weiser et al. (1978)




lames et al. (1979)





Moore et al. (1982)







-------
1
jo
\o











,_,
^
-J



O
Tl
H
1

O
z
o
H
c
o
H
M
O
O
H
w



Exposure
Alt = 4,572 m
CO = 500 ppm
6 weeks

Alt = 5,486 m
CO = 50, 100,
500 ppm
6 weeks
Alt = 4,572 m
CO = 100 ppm
6 weeks

Alt = 55, 1,524,
2,134, and 3, 048 m;
CO = 0, 50, 100,
and 150 ppm
Alt = 55 and 2,134 m;
CO = 0 and 9 ppm for
8h















TABLE 1 1-2 (cont'd)
COHb Subject
36.2% and Rat
34.1% (n=24)


5.8, 11.1, Rat
and 4.26% (n=22)


8.4% Rats
(n=24)


2.56 - 4.42% Human
(n=23)
(11 men;
12 women)
0.2 - 0.7% Human
(n=IT)
















. SUMMARY OF
Dependent Variable
Hematocrit (Hct),
mean electrical
axis (MEA),
HW/BW ratios
Cardiac hypertrophy,
coronary capillarity


Hct ratio and weights:
BW, HW, RV, LV + S,
Pituitary (PiT)

Maximum aerobic
capacity (VO2 max)


Maximum^aero'bic
capacity (VO2 max)
















EFFECTS OF CARBON
Results
Hct increased by altitude
and CO; MEA shifted
left with CO, right
with altitude
RV hypertrophy and
coronary capillarity
increased with altitude

Alt 1BW, tHct, TRY,
tHT, TPiT; CO tHct,
TLV+S

VO2 max decreased at 2,134m
(4%) and 3,048m (8%); VO2 max
decreased slightly with
increasing ambient CO.
VO2 max decreased 7-10% with
increasing altitude at 0 ppm CO;
similar effects were found after
8-h exposure to 9 ppm CO, regard-
less of exercise level













MONOXIDE AT ALTITUDE
Comments Reference
Effects of altitude and CO on Hct, MEA, Cooper et el. (1985)
and HW/BW were additive.


Increase in coronary McDonagh et al. (1986)
capillarity was blocked by CO.


Effects produced by altitude were McGrath (1988)
not intensified by 100 ppm CO.


Altitude- and CO-hypoxia independ- Horvath et al. (1988 a,b)
ently affect VO2 max; decreased COHb
with increasing altitude was due, in
part, to decreased driving CO pressure
VO2 max was reduced in all subjects Horvath and Bed! (1989)
at altitude regardless of the ambient
CO level














-------
       obtained by raising a normal group of men 102 m (335 ft) in altitude.  This relationship was
       stated for a range of altitudes from 2134 to 3048 m (7000 to 10,000 ft) and for increases in
       COHbupto 13%.
            Weiser et al. (1978) studied the effects of CO on aerobic work at 1610 m (5280 ft) in
 5     young subjects inhaling 100%  CO until COHb levels reached 5%.  They reported that this
       level of COHb impaired exercise performance at high altitude to the same extent as that
       reported at sea level (Horvath  et al., 1975). Because these subjects were Denver residents
       and fully adapted to this altitude, however, they would have had an arterial O2 concentration
       the same as at sea level (about 20 mL  O2/dL). Hence, 5% COHb would lower arterial O2
10     concentration about the same amount at both altitudes and impair work performance at
       altitude to the same extent as at sea level. In  the Weiser study, breathing CO during
       submaximal exercise caused small but significant changes in cardiorespiratory function; the
       working heart rate increased and the postexercise left ventricular ejection time shortened, but
       not to the same extent as when filtered air was breathed.  CO lowered the anaerobic threshold
15     and, at work  rates heavier than the anaerobic  threshold, increased minute ventilation.
            Wagner et al. (1978) studied young smokers and nonsmokers who exercised at 53% of
       their VO2 max at 760 and 523  torr.  Carboxyhemoglobin levels were raised to 4.2%.  While
       at altitude with these elevated  COHb levels, nonsmokers increased their cardiac output and
       decreased their arterial-mixed  venous O2 differences.  Smokers did not respond in a similar
20     manner. Smokers, with their  initial higher Hb concentrations, may have developed some
       degree of adaptation to CO and/or high altitude.
            Horvath et  al. (1988) in  a complex study involving four altitudes (up to 3050 m) and
       four ambient  CO concentrations  (up to 150 ppm) evaluated COHb levels during a maximal
       aerobic capacity  test.  They concluded that VO2  max values determined in men were only
25     slightly diminished due to increased ambient CO. Carboxyhemoglobin concentrations attained
       at maximum were highest at 55 m (4.42%)  and  lowest at 3035 m (2.56%) while breathing
       150 ppm CO (Figure 11-1). This was attributed to the reduced partial pressure of CO at high
       altitude. No  additional effects that could be attributed to the combined exposure to high
       altitude and CO were found.  Independence of the altitude and CO hypoxia was demonstrated
30     under the condition of performing a maximum aerobic capacity test.  The reductions in VO2
        March 12, 1990                          11-8     DRAFT - DO NOT QUOTE OR CITE

-------
                                MAX
        5 MIN POST MAX
              .a
              E
              O
              CJ
              w
              o
                  -2
                                     Control COHb % = 0.67 ±0.15
                      0 50 100150 0  50 100150 0 50 100150  0  50 100150

                             CARBON  MONOXIDE LEVEL (PPM)
          B
              O
              O
              5
              w
              a
                 -2
                                 MAX
     ES3 5 MIN POST MAX
                                Control COHb 0.81 ±  0.23
                      0 50 100150 0  50 100150 0 50 100150  0  50 100150

                             CARBON MONOXIDE LEVEL (PPM)
Figure 11-1.  Increment in percent carboxyhemoglobin (A% HbCO) over basal (control) levels
at the end of a maximum aerobic capacity test and at the 5th min of recovery from a test in a
typical (A) male and (B) female subject.  Altitudes are 55, 1524, 2134, and 3048 m, whereas
exercise was conducted with ambient concentrations of 0, 50, 100, and 150 ppm CO.

Source: Horvath et al. (1988).
March 12, 1990
11-9     DRAFT - DO NOT QUOTE OR CITE

-------
       max due to high altitude and to the combined exposure of ambient CO and high altitude were
       similar.
            Horvath and Bedi (1989) studied 17 nonsmoking young men to determine the alterations
       in COHb during exposure to 0 or 9 ppm ambient CO for eight hours at sea level or
 5     an altitude of 2134 m (7000 ft). Nine subjects rested during the exposures and eight
       exercised for the last 10 minutes of each hour at a mean ventilation of 25 L (BTPS).  All
       subjects performed a maximal aerobic capacity test at the completion of their respective
       exposures.  At the low CO concentrations studied, the Coburn, Forster, and Kane (CFK)
       equation estimated COHb levels to be 1.4% (Petersen and Stewart, 1975).  Carboxy-
10     hemoglobin concentrations fell in all subjects during their exposures to 0 ppm CO at sea level
       or 2134 m. During the eight hour exposures to 9 ppm CO, COHb levels rose linearly from
       approximately 0.2 to 0.7% (Figures  11-2).  No significant differences in uptake were found
       whether the subjects were resting or intermittently exercising.  Levels of COHb were similar
       at both altitudes.  A portion of the larger estimate of COHb determined by the CFK equation
15     could be accounted for by the use of an assumed blood volume. Maximal aerobic capacity
       was reduced approximately 7 to 10% consequent to altitude exposure during 0 ppm CO.
       These values were not altered following eight-hour exposure to 9 ppm CO in either resting or
       exercising individuals.

20     11.1.4  Chronic Studies
           There have been few studies of the long-term effects of CO at altitude (see Table  11-2).
       James et al. (1979) studied cardiac function in six unsedated goats that were chronically
       instrumented and exposed to 160 to 200 ppm CO (COHb =  20%) for six weeks at 1524 m.
       Cardiac index and stroke volume were unchanged during and after the exposure.  Heart rate
25     and contractility (V^J of the left ventricular myocardium were unchanged during exposure to
       CO, but both were depressed during  the first week after removal of the CO. The authors
       concluded that if there was a decrease in intrinsic myocardial function during the CO
       exposure, it may have been  masked by increased sympathetic activity.
           McGrath (1988; 1989) studied cardiovascular, body, and organ weight changes in rats
30     exposed continuously for six weeks to (1) ambient altitude, (2) ambient altitude + CO,
       March 12, 1990                         11-10    DRAFT - DO NOT QUOTE OR CITE

-------
                1.0


                08
             .a
             x
             o 0.6
             o

             * 0.4


                0.2


                O.O
       D-D  2134m
       *-A.  55m
       O-O  0 ppm
                                      Resting Subjects
                                        \
                                        \\
                        CO 9ppm
                                   	-ex..
                          12345678    M
                                       Hours
           B
  1.0


  0.8
.o
x
00.6
o


  0.4


  0.2


  0.0
                      D-D 2134m
                      Ar-A 55m
                      o-oOppm
                       Active  Subjects
                                                            PM
Figure 11-2. Change in carboxyhemoglobin concentration (% COHb) during eight-hour
exposures to 0 to 9 ppm CO for (A) resting and (B) exercising subjects. Altitudes are sea
level (55 m) and 2134 m (7000 ft).


Source: Horvath and Bedi (1989).
March 12, 1990
                      11-11    DRAFT - DO NOT QUOTE OR CITE

-------
       (3) simulated high altitude, and (4) CO at high altitude.  Altitudes ranged from 3300 ft
       (1000 m) to 18,000 ft (5486 m ) and CO concentrations from 0 to 500 ppm.
            Carbon monoxide had no effect on body weight at any altitude.  There was a tendency
       for hematocrit to increase even at the lowest concentration of CO (9 ppm), but the increase
 5     did not become significant until 100 ppm.  At 10,000 ft, there was a tendency for the total
       heart weight to increase in rats inhaling 100 ppm CO.  Although its effects on the heart at
       high altitude are complex, CO, in concentrations of 500 ppm or less, had little effect on the
       right ventricle; it did not exacerbate any effects due to altitude. There was a tendency for the
       left ventricle weight to increase with exposure to 35 ppm carbon monoxide at high altitude,
10     but the increase was not significant until 100 ppm CO.   Heart rate, blood pressure, cardiac
       output, and peripheral resistance were unaffected by exposure to 35 ppm CO or 10,000-ft
       altitude, singly or in combination.  The author concluded that six weeks of exposure to 35
       ppm CO does not produce measurable effects in the healthy laboratory rat, nor does it
       exacerbate the effects produced by exposure to 10,000-ft altitude.
15          The data reported by McGrath (1988; 1989) are generally in agreement with findings
       reported by other investigators.  Carboxyhemoglobin obtained at the end of the six weeks of
       exposure to CO are presented in Figure 11-3.  The COHb  concentrations were (at 3300 ft)
       0.6, 0.9, 2.4, 3.7, and 8.5% for ambient  CO  levels of 0, 9, 35, 50, and 100 ppm,
       respectively.  This relationship can be expressed as:
20
                                      %COHb = 0.115 + O.OSx                          (11-2)

       where x is the CO exposure, in ppm. The correlation coefficient (r) for this relationship was
       0.99.  The changes at other altitudes were not sufficient to calculate the rate of increase.
25     Exposure of rats to 500 ppm and altitudes up to 18,000 ft resulted in COHb levels of 40 to
       42%.  An interesting, but not unexpected, finding in this study was that high altitude
       residence in the absence of exogenous CO resulted in increased basal COHb concentrations.
       These values were 0.6, 1.3, 1.7,  and 1.9% for altitudes from 3300 to  18,000  ft.  These
       increases can be expressed as:
30
                                  %COHb  = 0.0000914 + 0.26687x                      (11-3)

       March 12, 1990                          11-12     DRAFT - DO NOT QUOTE OR CITE

-------
HUD HO HXOnb JLON Od - JLdVHd    CMI
066T 'Zl
                                   Percent COHb
               o   -»
    a
   ore
   EL
   cr.

   I
   I
   I-F

   8


   I
   S*
   a

   I

-------
       where x is altitude, in feet.  The correlation coefficient (r) for this relationship was 0.99.
       Whether similar increases in basal COHb concentrations would be observed in humans
       adapted to altitude needs to be determined.  Presumably, because there are marked elevations
       in Hb and Hct in residents of high altitudes, a greater endogenous CO production might be
 5     present.  In this study, 100 ppm CO had no effect on body, right ventricle, total heart,
       adrenal, spleen, or kidney weights, but it did increase Hct ratios and left ventricle weights.
       There was no significant interaction between altitude and CO on any parameter except kidney
       weight.  The author concluded that although there was a tendency for hematocrit ratios,
       spleen weights, and total heart weights to be elevated by combined CO-altitude exposure, the
10     results were not significant and, in general, the effects produced by 4572-m altitude were not
       intensified by exposure to 100 ppm CO.
            McDonagh et al. (1986) studied cardiac hypertrophy and ventricular capillarity in rats
       exposed to 5486 m (18,000 ft) and 50,  100, and 500 ppm  CO.  Coronary capillarity
       increased after exposure to 5486 m for six weeks, but this response was blocked by CO.
15     Right ventricular thickness was increased by altitude, but was not increased further by CO.
       At 500 ppm, CO the right ventricular hypertrophy was attenuated, but the results are
       uncertain due to the high mortality in this group.  Left ventricular thickness also was
       increased at  5486  m (18,000 ft) and increased further by CO.  The authors concluded that
       because the ventricular thickness is increased while capillarity is reduced, it is possible that
20     the myocardium can be underperfused in the altitude plus CO group.
            Cooper et al. (1985) evaluated the effects of CO at altitude on EKGs and cardiac
       weights in rats exposed for six weeks to (1)  ambient (amb), (2) ambient + 500 ppm CO
       (amb+CO), (3) 4572 m (15,000 ft) (alt), and (4) 4572 m  + 500 ppm CO (alt+CO).  COHb
       values were  36.2 and 34.1 % in the amb+CO and alt+CO groups, respectively.  Hematocrits
25     were 54 ± 1, 77± 1, 68 ± 1, and 82± 1%, in the amb, amb+CO, alt, and alt+CO
       groups,  respectively.  In the amb+CO, alt, and alt+CO groups, respectively, the mean
       electrical axis shifted 33.2° left, 30° right, and 116.4° right. Heart weight to body weight
       ratios were 2.6, 3.2, 3.2, and 4.0 x  10"3 in the amb, amb+CO,  alt, and alt+CO groups,
       respectively.  Whereas CO increased left ventricular weight, and alt increased right
30     ventricular weight, alt+CO increased both.  EKG changes were consistent with changes in
       cardiac weight.

       March 12, 1990                          11-14   DRAFT - DO NOT QUOTE OR CITE

-------
            These results indicate that whereas CO inhaled at ambient altitude causes a left
       electrical axis deviation, CO inhaled at 4572 m exacerbates the well-known phenomenon of
       right electrical axis deviation.  Thus, the results from chronic animal studies indicate that
       there is little effect of CO on the cardiovascular system of rats exposed to CO concentrations
 5     of 100 ppm or less and altitudes up to 3030 m (10,000 ft).
            Exposure to CO from smoking may pose a special risk to the fetus at altitude.  Moore
       et al. (1982) reported that maternal smoking at  3100 m is associated with a two- to threefold
       greater reduction in infant birth weight than has been reported at sea level.  Moreover, COHb
       levels of 1.8 to 6.2% measured in all pregnant  subjects, were inversely related to infant birth
10     weight. Earlier, Brewer et al. (1970,  1974) reported that the mean COHb level in smokers at
       altitude is higher than smokers at sea level, and that subjects who smoked had greater O2
       affinities than nonsmokers. Moreover, cessation of smoking by polycythemic individuals at
       altitude results in a marked reduction in COHb  and a decrease in hemoglobin-O2 affinity to
       values less than those reported for normal individuals at sea level.  The chronic effects of
15     altitude and CO exposure are summarized in Table 11-3.

                 TABLE 11-3. CHRONIC EFFECTS OF ALTITUDE AND CARBON
                                     MONOXIDE EXPOSURE
20     Effect                                   Altitude                  Carbon Monoxide
       Hemoglobin                                t                            t
       Hematocrit                                 t                            T
       Pulmonary arterial                          t
25       pressure
       Cardiac hypertrophy
         Right ventricle                            t
         Both ventricles                            -                            I
       Cardiac output"                             tt                          ?
30     Blood volume                               t                            t
       Body weight                                4
       "Initial increase that later  returns to baseline value.
       March 12, 1990                          11-15     DRAFT - DO NOT QUOTE OR CITE

-------
       11.1.5  Neurobehavioral Effects
            The neurobehavioral effects following CO exposure are controversial and are described
       in detail in Chapter 10, (Section 10.4). Those neurobehavioral studies specifically concerned
       with CO exposure at altitude are reviewed briefly in this section.
 5          McFarland et al. (1944) reported changes in visual sensitivity occurring at a COHb
       concentration of 5%  or at a simulated altitude of approximately 2432 m (8000 ft).  Later,
       McFarland (1970) expanded on the original study and noted that a pilot flying at 1829 m
       (6000 ft) breathing 0.005% CO in air is at an altitude physiologically equivalent to
       approximately 3658 m (12,000 ft).  McFarland stated that sensitivity of the visual acuity test
10     was such that even the effects of small quantities of CO absorbed from cigarette smoke were
       clearly demonstrable. In subjects inhaling  smoke from three cigarettes at 2286 m (7500 ft),
       there was a combined loss of visual sensitivity equal to that occurring at 3048 to 3353 m
       (10,000 to 11,000 ft).  This report was confirmed by Halperin et al. (1959), who also
       observed that recovery  from the detrimental effects of CO on visual sensitivity lagged behind
15     elimination of CO from the blood.
            Lilienthal and Fugitt (1946) reported  that combined exposure to altitude and CO
       decreased flicker-fusion frequency (FFF) (i.e., the critical frequency in cycles per second at
       which a flickering light appears to be steady). Whereas mild hypoxia (that occurring at 2743
       to 3658 m (9000 to 12,000 ft)) alone impaired FFF, COHb levels of 5 to 10% decreased the
20     altitude threshold for onset of impairment to  1524 to 1829 m (5000 to 6000 ft).
            The psychophysiological effects of CO at altitude are a particular hazard in high-
       performance aircraft (Denniston et al., 1978). Acute ascent to altitude increases ventilation
       via the stimulating effects of a reduced PO2 on the chemoreceptors.  The increased ventilation
       causes a slight increase in blood pH and a slight leftward shift in the O2Hb dissociation curve.
25     Although such a small shift would probably have no physiological significance under normal
       conditions, it may take on physiological importance for aviators required to fly under a
       variety of operational conditions and to perform tedious tasks involving a multitude of
       cognitive processes.  The leftward shift of  the O2Hb dissociation curve may be further
       aggravated by the persisting alkalosis caused  by hyperventilation resulting from anxiety. The
30     potential for this effect has been reported by  Pettyjohn et al.  (1977), who reported that
       respiratory minute volume may be increased by 110% during final landing approaches

       March 12, 1990                          11-16    DRAFT - DO NOT QUOTE OR CITE

-------
       requiring night-vision devices.  Thus, the hypoxia-inducing effects of CO inhalation would
       accentuate the cellular hypoxia caused by stress and altitude-induced hyperventilation.

       11.1.6  Compartmental Shifts
 5          Studies by Luomanmaki and Coburn (1969) suggest that CO in very high
       concentrations may pose a special threat at higher altitudes. These workers report that during
       hypoxia, in anesthetized dogs, CO shifts out of the blood and into the tissues.  In
       experiments using 14CO, they observed that radioactivity in blood did not change when arterial
       O2 tension increased from 50 to 500 mmHg. However,  14CO activity in blood decreased to
10     50% of control levels when arterial PO2 decreased below 40 mmHg; 14CO shifted back into
       the blood when arterial  PO2 returned to normal.  Because there was no significant difference
       between splenic and central venous WCO radioactivity either before or after the MCO shift,
       these workers excluded  the possibility that the 14CO had been sequestered in the spleen.
            Luomanmaki and  Coburn (1969) also studied the shift of CO out of the blood during
15     hypoxia by administering CO into a rebreathing system and measuring the rate at which blood
       COHb increased.  They reasoned that if the partition of CO between vascular and
       extravascular stores remained constant,  the increase in blood COHb should be proportional to
       the amount of CO administered. They  found that COHb increased at a constant rate up to a
       saturation of 50%. With additional CO, there was a decrease in the rate at which COHb
20     increased; this suggests  that proportionally greater amounts of CO were entering the
       extravascular stores.  At 50% COHb (corresponding to  an arterial PO2 of 90 mmHg), the
       rate of COHb formation became nonlinear. Agostoni et al. (1980) presented a theoretical
       model supporting these  observations; they developed equations predicting that decreasing
       venous PO2 causes CO to move out of the vascular compartment and into skeletal and heart
25     muscle.  This increases  the rate at which COMb is formed in  the tissues.
            The shift of CO out of the blood has been further demonstrated in studies  (Horvath
       et al., 1988) conducted  on both men and women undergoing maximal aerobic  capacity tests
       at altitudes of 55,  1524, 2134, and 3058 m and CO concentrations of 0, 50,  100, and 150
       ppm. Carbon monoxide at maximum work shifted into extravascular spaces and returned to
       March 12,  1990                         11-17    DRAFT - DO NOT QUOTE OR CITE

-------
       the vascular space within five minutes after exercise stopped (Figure 11-4).  This liberation of
       CO was related to the concentration of COHb achieved as noted by the regression equation:

                                       y  = 0.0017 + 0.3047x                           (11-4)
 5
       where x is the COHb concentration at exhaustion.

       11.1.7  Conclusions
            While there are many studies comparing and contrasting inhaling CO with exposure to
10     altitude, there are relatively few reports on the effects of inhaling CO at altitude.  There are
       data to support the possibility that the effects of these two hypoxia episodes  are at least
       additive.  These data were obtained at CO concentrations that are too high to have much
       meaning for regulatory concerns.  There also are data that indicate decrements in visual
       sensitivity and flicker-fusion frequency in subjects exposed to CO (COHb = 5 to 10%) at
15     higher altitudes.  These data, however, are somewhat controversial.
            There are even fewer studies of the long-term effects of CO at high altitude. These
       studies generally indicate few changes at CO concentrations below 100 ppm and altitudes
       below 4572 m (15,000 ft).  A provocative study by McDonagh et al. (1986) suggests that the
       increase in ventricular capillarity seen with altitude exposure may be blocked by CO.  The
20     fetus may be particularly sensitive to  the effects of CO at altitude; this is especially true with
       the high levels of CO associated with maternal smoking.
       11.2  CARBON MONOXIDE INTERACTIONS WITH DRUGS
25     11.2.1  Introduction
            There is little direct information on the possible enhancement of CO toxicity by
       concomitant drug use or abuse; however, there are some data suggesting cause for concern.
       There is evidence that interactions of drug effects with CO exposure can occur in both
       directions, that is, CO toxicity may be enhanced by drug use and the toxic or other effects of
30     drugs may be altered by CO exposure.  Nearly all the published data that are  available on
       CO combinations with drugs concern psychoactive drugs. Possible interactions of CO with

       March 12, 1990                         11-18    DRAFT - DO NOT QUOTE OR CITE

-------
                                       Y =  0.0017  +  0.3047  X
                                     Sy   =  0.1337
                                        x
              1    1.4  1.8   2.2  2.6    3    3.4   3.8   4.2   4.6
                                    MAX COHb
  Figure 11-4. Higher concentrations of COHb observed at the end of a five-minute recovery
  period after attainment of the subject's maximum aerobic capacity indicate that liberation of
  CO from tissue stores is linearly related to COHb concentration present at exhaustion.

  Source: Horvath et al. (1988).
March 12, 1990
11-19    DRAFT - DO NOT QUOTE OR CITE

-------
       other classes of drugs (e.g., those likely to be used in patients with cardiovascular disease
       who also are at risk for CO exposure) will be discussed elsewhere in this document (see
       Chapter 12, Section 12.4).  Another related area of concern that will be reviewed elsewhere is
       interactions of CO with other toxicants (see Section 11.3).
 5          The use and abuse of psychoactive drugs and alcohol is ubiquitous in society.  Because
       of CO's well-established effects on brain functioning, interactions between CO and
       psychoactive drugs could be anticipated. Unfortunately, very little systematic research has
       addressed this question.  In addition, very little of the research that has been done has utilized
       models for expected effects for treatment combinations. Thus, often it is not possible to
10     assess whether the combined effects of drugs and CO exposure are additive or differ from
       additivity.  It is important to recognize that even additive effects of combinations can be of
       clinical significance,  especially when the individual is unaware of the combined hazard.

       11.2.2  Alcohol
15          The effects of combined CO exposure and  alcohol (ethanol) administration have been the
       most extensively studied interaction. The previous criteria document (U.S. Environmental
       Protection Agency, 1979) reviewed two human studies which examined combinations of CO
       and alcohol.  A study from the Medical College of Wisconsin (1974) found no effects of
       alcohol doses resulting in blood alcohol levels of about 0.05% and COHb levels in the general
20     range of 8 to 9%, either alone or in combination, on a number of psychomotor behavioral
       tasks. The lack of sensitivity of these measures  to alcohol doses known to affect performance
       under many other conditions, as well as other problems in the study design, raises the
       question  of the adequacy of this study to detect interactive effects.  Rockwell and Weir (1975)
       studied the interaction of CO exposures resulting in nominal 0, 2, 8, and 12% COHb levels
25     with alcohol doses resulting in nominal 0.05% blood alcohol levels for effects on actual
       driving performance  in young,  nonsmoking college students.  Dose-related effects of CO for
       perceptual narrowing and decreased eye movement were observed. In addition, effects were
       observed on some measures by this dose of alcohol alone.  An effect-addition model was used
       to evaluate the alcohol-CO interaction.  In combination, the effects of CO and alcohol were
30     often additive, and there was a supra-additive alcohol-CO interaction at 12% COHb levels.
       Although the 1979 review  highlighted the lack of an interaction effect except at high COHb

       March 12,  1990                          11-20    DRAFT - DO NOT QUOTE OR CITE

-------
        concentrations, it should be noted that interaction effects for this study were defined as effects
        greater than the sum of the effects of the treatments alone.  Thus, this extensive study in
        human subjects provides some evidence that driving-related performances already disrupted by
        alcohol could be further compromised by CO exposure.
 5           Because of a concern that persons exposed to CO may not be able to detect odors that
        would indicate a fire or other hazardous condition, especially when consuming alcohol, Engen
        (1986) conducted a carefully controlled study of combined CO-alcohol exposure in human
        subject volunteers.  The detection of a threshold concentration of the smoky odor of quaiacol
        was evaluated using signal-detection analysis.  The dose of alcohol given resulted in blood
 10      alcohol levels of about 0.04 to 0.07% and CO exposure resulted in COHb values  of 7.0 to
        7.7%.  In signal detection studies, d1 is a measure of detection threshold, with higher values
        reflecting greater detection.  The average d'  for the four treatment conditions was as follows:
        air only (1.95), CO only (2.34), alcohol + air (2.20), and alcohol +  CO (1.64).  Although
        not statistically significant, there was a tendency for both alcohol and  CO to improve odor
 15      detection compared to air only. When alcohol and CO were combined, the odor detection
        was significantly poorer than after either treatment alone, but it was not significantly poorer
        than the air control. One of the features of signal detection analysis is that it allows the
        separation of treatment effects  on  sensory sensitivity from effects on performance  that would
        influence the reporting of the signal.  Thus, in this study it was found that these changes in
20      odor detection produced by alcohol and CO occurred in the absence of an effect of any of the
        experimental treatments on reporting  bias. Thus, one could conclude  that the results of
        combined alcohol and CO exposure was to eliminate the small improvement in odor
        sensitivity produced by exposure to either treatment alone.  The relevance and importance of
        these small changes in odor detection are not readily apparent, especially because  none of the
25      treatments were significantly different from air control; however, they do suggest that a CO-
        alcohol interaction on odor thresholds may exist.  An incidental finding of this study was that
        alcohol did not alter COHb concentrations after exposure to CO; nor did CO exposure affect
        blood alcohol levels produced by a fixed dose of oral alcohol.
            There also have been a number of animal studies of combinations of alcohol and CO.
30     Although there is some evidence that  alcohol metabolism can be reduced in rat liver in situ by
       a COHb level of 20% (Toppling et al., 1981), an in vivo study in mice found no effects of

       March 12, 1990                           11-21    DRAFT - DO NOT QUOTE OR CITE

-------
       CO exposure on alcohol metabolism (Kim and Carlson, 1983).  Compared to levels in control
       mice, either 1, 3, or 5 days of eight hour per day exposure to 500 ppm CO (COHb levels
       averaged 28%) had no effect on blood alcohol levels when 2.2 g/kg of alcohol was
       administered ip after 5.5 h of exposure on each of these days. On the other hand, Pankow
 5     et al.  (1974) provide some evidence that high doses of CO associated with COHb levels of
       greater than 50% decreased blood alcohol  levels in rats 30 min after a very large dose of
       alcohol (4.8 g/kg). They also reported that this dose of alcohol significantly lowered COHb
       levels associated with a very large subcutaneous dose of CO. These high-dose combinations
       were also associated with additive effects on enzyme markers of hepatotoxicity, but no
10     interactions were observed when lower doses of CO were given.
            In contrast to the inconsistent metabolic effects seen with combinations of CO and
       alcohol, results of two behavioral studies in animals have both shown substantial effects.
       Mitchell et al.  (1978) studied the interaction of inhaled CO with two doses of alcohol
       (0.6 and 1.2 g/kg) in rats using two behavioral measures. Sensorimotor incapacitation was
15     assessed by failure to remain on a rotating rod. An additional measure of motor effects was
       the inability to withdraw the leg from a source of electric shock.  The length of exposure to
       about 2000 ppm CO before the animals failed in these performances was decreased in a dose-
       dependent manner by alcohol.  Carboxyhemoglobin determinations made at the time of
       behavioral incapacitation was inversely related to alcohol  dose.  For example, nearly  50%
20     COHb levels were required to impair rotorod performance in the absence of alcohol,  whereas
       after 1.2 g/kg ethanol, less than 45% COHb levels produced the same effect. Unfortunately,
       data was not provided on the effects of these doses of alcohol in the absence of CO exposure
       to help determine the nature and magnitude of the interaction effects.
            Knisely et al. (1989) recently  reported a large interaction of CO exposure and alcohol
25     administration on operant behavior in animals.  Mice that had been trained to lever press for
       water reinforcement were tested with 1.1 g/kg alcohol and various doses of CO, alone and in
       combination. An unusual feature of this study was that the both the alcohol and CO were
       administered by ip injection.  The authors provide evidence that this route of CO exposure
       results in COHb formation and behavioral effects comparable to those seen after inhalation
30     exposure.  The results of the study  were evaluated by comparing the effects of the
       combinations to those expected by summing the effects of each treatment alone. A dose of

       March 12,  1990                          11-22     DRAFT - DO NOT QUOTE OR CITE

-------
       alcohol that had little effect on rates of lever pressing when given alone resulted in large rate-
       decreasing effects when given in combination with doses of CO that also had no effects when
       given alone.  Supra-additive effects with alcohol were obtained by a dose of CO as low as
       7.5 mL/kg, which when given alone was associated with COHb levels of about 20%.
 5     Significant supra-additive effects also were obtained with higher doses of CO.  Typically,
       behavioral effects of CO alone were not seen under these test conditions until COHb
       saturations greater than 40 to 50%  were obtained (Knisely et al., 1989). Thus, alcohol about
       doubled the acute toxicity of CO in this study.

10     11.2.3  Barbiturates
            There has been some interest in the interaction of CO with barbiturates because
       prolongation of barbiturate  effects can reflect effects of toxicants on drug metabolism. In an
       early evaluation of the functional significance of the binding of CO to cytochrome P-450,
       Montgomery and Rubin (1971) examined the effects of CO exposure on the duration of action
15     of hexobarbital and the skeletal muscle relaxant zoxazolamine in rats.  Both drugs are largely
       deactivated by the  hepatic mixed-function oxidase (MFO) system.  Although CO was found to
       dose-dependently enhance both hexobarbital sleeping time and zoxazolamine paralysis,
       subsequent research indicated that this was probably not due to a specific inhibition of the
       MFO system by CO, but rather a nonspecific effect of hypoxia, because even greater effects
20     could be produced at a similar level of arterial O2 produced by hypoxic hypoxia (Montgomery
       and Rubin, 1973; Roth and Rubin, 1976).  In support of the lack of effects of CO on drug
       metabolism, Kim  and Carlson (1983) found  no effect of CO exposure on the plasma half-life
       for either hexobarbital or zoxazolamine in mice. This would suggest that something other
       than a  metabolic interaction may  be responsible for the enhancement of in vivo effects of these
25     drugs by CO.
            There have been two studies of the interaction of CO and pentobarbital using operant
       behavior in laboratory animals.  McMillan and Miller (1974) found that exposure of pigeons
       to 380 ppm CO, a  concentration that had little effect on behavior when given alone, reduced
       the response rate of schedule-controlled behavior, thereby increasing the effect of an
30     intermediate dose of pentobarbital.  On the other hand, the  disruptive effects of all doses of
       pentobarbital on the temporal patterning of fixed-interval responding was enhanced markedly

       March 12, 1990                          11-23     DRAFT - DO NOT QUOTE OR CITE

-------
       by 1030 ppm CO.  This concentration of CO by itself did not alter response patterning, but
       did lower overall rates of responding. In the study described more fully above in the section
       on alcohol interactions (11.2.2),  Knisely et al. (1989) found generally additive effects of ip
       CO administration with the effects of pentobarbital in mice responding under a fixed-ratio
 5     schedule.  In that study the interaction of CO with pentobarbital was not as evident as the
       interaction with alcohol, suggesting that general conclusions about CO interactions with
       central nervous system depressant drugs may not be possible.

       11.2.4  Other Psychoactive Drugs
10         Even more limited data are available on interactions of CO exposure with other
       psychoactive drugs. In the study by Knisely et al. (1989), described above (Section 11.2.2),
       of interactions of ip CO administration with psychoactive drugs on operant behavior of mice,
       d-amphetamine, chlorpromazine, nicotine, diazepam, and morphine were studied in addition
       to alcohol and pentobarbital. As with alcohol, a suggestion of greater than additive effects
15     were obtained from combinations of CO with both d-amphetamine and chlorpromazine;
       however, in these cases the differences from additivity did not reach statistical significance.
       Effects of CO in combination with nicotine, caffeine, and morphine were additive.  McMillan
       and Miller (1974) also found evidence for an interaction of CO and d-amphetamine on
       operant behavior in pigeons. In this study, CO concentrations as low as 490 and 930 ppm
20     were able to modify the behavioral effects of d-amphetamine.
       11.3  COMBINED EXPOSURE TO CARBON MONOXIDE AND OTHER
             AIR POLLUTANTS AND ENVIRONMENTAL FACTORS
25          Exposure to a single air pollutant at ambient concentrations may have no harmful
       biological effects. In real life, however, exposure occurs not only to a single agent but also
       k> multiple agents, resulting in potential interactions between them.  The result of the
       interactions may be of an additive, synergistic, or antagonistic nature.  Another possible
       interaction is potentiation, a condition in which a pollutant that is noneffective at a given
30     exposure level may enhance the toxicity of another pollutant given simultaneously. Exposure
       March 12, 1990                        11-24    DRAFT - DO NOT QUOTE OR CITE

-------
       to CO frequently occurs in the natural environment in combination with other combustion
       products and air pollutants.
            In this section both human and animal effects associated with combined exposure to CO
       and other air pollutants and environmental factors are reviewed.  Although a number of
 5     studies in the literature have tested exposure to combined pollutants, fewer studies actually
       have been designed to test specifically for interactions between CO and the other exposure
       components.  Therefore, this section emphasizes only those studies providing a combined
       treatment group where pollutant exposure levels are reported. The COHb levels resulting
       from CO exposure also are given if they were reported in the original manuscripts.  The
10     toxicity data discussed stress the newer literature published since 1979  in order to update the
       information reviewed in the previous Air Quality Criteria Document (U.S. Environmental
       Protection Agency, 1979).

       11.3.1 Exposure in Ambient Air
15          Photochemical air pollution usually is associated with two or more pollutants, consisting
       mainly of CO, sulfur oxides, ozone, nitric oxides, peroxyacetyl nitrates, and organic
       peroxides.  The gaseous compounds that constitute tobacco smoke are CO, hydrogen cyanide,
       and nitric oxide. As urban living, industrial employment, and cigarette smoking bring man
       into direct contact with CO and other pollutants, it seems appropriate to determine if
20     combined exposure to these pollutants has detrimental health effects.
            Several studies have been conducted to determine the effects resulting from combined
       exposure to CO and other pollutants.  The experimental details (e.g., concentrations and
       duration of treatment)  and the associated effects for each study are summarized in
       Table 11-4. A brief discussion of the major findings follows.
25          Murphy (1964) observed an increase in blood COHb levels in mice and rats exposed to
       CO+O3 for six hours as compared with mice exposed to CO alone.  However, another study
       (DeLucia et al., 1983) in adults exposed to CO+O3 during exercise, showed no synergistic
       effects on blood COHb levels or pulmonary or cardiorespiratory thresholds. Similarly,
       simultaneous exposure to CO-f O3+NO2 for two hours produced no consistent changes
30     (synergistic or additive) in pulmonary function indices and physiological parameters in young,
       male subjects (Hackney et al., 1975a,b).

       March 12,  1990                          11-25     DRAFT - DO NOT QUOTE OR CITE

-------
                                          TABLE 11-4.   COMBINED EXPOSURE TO  CARBON MONOXIDE AND OTHER POLLUTANTS
£3-    Pollutant
                  Concentration
                    No./sex/species
                                                                      Treatment
                                                                                                                Observed Effects
                                                                                                                                                 Reference
K)
ON
to    co
      CO
      CO
      o,
CO
      NO2
       CO
       Peroxy-
       acetyl-
       nitrate
       (PAN)
                  300 ppm
                  0.75 ppm
                  280 ppm
                  3 ppm
                  100 ppm
                  0.3 ppm
30 ppm
0.25 ppm
0.30 ppm
                   50 ppm
                   0.27 ppm
                 9-10/-"/mouse and
                 rat
                                          9/-/mice
                 24/M/human
                 24/F/human
                 (smokers and
                 nonsmokers)
                                    8/M/human
                  10/M/human
                  (smokers)
                  10/F/human
                  (nonsmokers)
Exposed to 300 ppm CO alone or
0.75 ppm O3 + 300 ppm CO for
6 h;  blood COHb levels were
determined
                                              Exposed to 280 ppm CO alone or
                                              3 ppm O3 + 280 ppm CO for 6 h
Exposed during exercise (four 1-h
rides on a bicycle to filtered
air only; 0.3 ppm O3 alone; 100 ppm
CO alone; or 0.3 ppm O3 +  100 ppm
CO); blood  COHb, pulmonary
function, cardiorespiratory per-
formance, blood laclate levels,
and subjective symptoms were
examined

Exposed to  O3 alone and in
combination with NO2 and CO for
2 h with secondary stress of
heat and intermittent light
exercise; subjective symptoms
were recorded; pulmonary
function, and physiological
studies were conducted

Subjects exposed to filtered air
only; 50 ppm CO alone; 0.27 ppm
PAN alone; or 50 ppm CO +
0.27 ppm PAN for 5 min at
two different temperatures,
25° and 35 °C (relative
humidity 30%); were tested for
maximal aerobic power, metabolic
temperature, and cardio-
respiratory  responses
Simultaneous exposure produced
higher COHb levels (30.4% in
rats and 18.9% in mice) than
exposure to CO alone (25.8% in
rats and 14.8% in mice).

COHb levels were 24.3% in mice
exposed CO + O3 compared with
19.2%  in mice exposed to CO alone.

Exposure to O3 + CO did not
elicit a  synergistic effect.
Combined exposure did not
alter the threshold(s) of any
subject for appearance of adverse
effects  due to O3 alone.  Exposure
to CO alone caused a mean increase
in COHb (5.8%) levels compared
with exposure not involving CO.

No consistent synergistic or
additive effects were observed
in subjects exposed  to 03  +
NO2 + CO in any parameter
measured, except for increases
in blood COHb (levels not
reported).
 Maximal aerobic power was not
 affected by any pollutant
 conditions. The heart rate
 was significantly (p<0.05)
 greater in the CO group
 compared with the filtered-air
 group.  Metabolic and thermo-
 regulatory responses were not
 different in the various
 pollutant environments.  Increases
 in COHb levels  of smokers during
 the CO  or CO + PAN exposures
 were observed.
                                                                                                                                                    Murphy (1964)
                                                                                                                                              Murphy (1964)
                                                                                                                                                     DeLucia et al. (1983)
                                                                                                                                              Hackney et al. (1975 a,b)
Drinkwater et al. (1974);
Gliner et al. (1975);
Raven et al. (1974a,b)

-------
                          TABLE 11-4   (cont'd).   COMBINED EXPOSURE TO  CARBON MONOXIDE AND OTHER POLLUTANTS

       Pollutant
                  Concentration
                   No./sex/species
                                   Treatment
                                                Observed Effects
                                                                                                                                                      Reference
tb
-J
 I
O
O
O
 d
 I
 §
 O
h-    CO
\o
\o
O    NO
      CO
      NO
      HCN
      CO
       NO,
       CO
       SO2
CO
so,
                  100 ppm
                  500 ppm
                  10 ppm
                  50 ppm
                  200 ppm
                   5 ppm
                  0.5 ppm
                  20 ppm
                  67.5 ppm

                  0.5 ppm
                  7.5 ppm
                  20 ppm
                  67.5 ppm
                  0.5 ppm
                  10 ppm
250 ppm
25 ppm
                 15/M/rat
                 (Long-Evans)
                 12/M/rabbit
                 (New Zealand white)
                                         24/M/rat
                                         24/F/rat
                                         (Sprague-Dawley)
                 24/M/rat
                 24/F/rat
                 (Sprague-Dawley)
32-40/F/mouse
(CF-1)
                           Exposed to clean air only; 100 or
                           500 ppm CO alone; 10 or 50 ppm
                           NO alone; 100 ppm CO +  10 ppm
                           NO; or 500 ppm CO + 50 ppm NO
                           for 3 h; changes in
                           discrimination learning and
                           brain activity were measured
                            Exposed to 0.5 ppm HCN + 5 ppm
                            NO + 200 ppm CO for 2 weeks
                                             Exposed to clean air only; 0.5 or
                                             7.5 ppm NO2 alone; 20 or
                                             67.5 ppm CO alone; 0.5 ppm
                                             NO2 + 67.5 ppm CO; or 7.5 ppm
                                             NO2 + 20 ppm CO continuously,
                                             24 h/day, 7 days/week for
                                             52 weeks; chronic toxicity was
                                             assessed
Exposed to clean air only, 0.5 ppm
or 10 ppm SO2 alone; 20 or 67.5 ppm
CO alone; 0.5 ppm SO2 + 67.5 ppm
CO; or 10 ppm SO2 + 20 ppm CO
continuously, 24 h/day,
7 days/week for 52 weeks; chronic
toxicity was assessed

Exposed to filtered air only; 25 ppm
SO2 alone; or 25 ppm SO + 250 ppm
CO for 7 h/day during Days 6 through
15 gestation; teratogenic potential
was evaluated
No significant changes were observed in
COHb levels between any treatment groups.
Exposure to 100 ppm CO +  10 ppm  NO
signinificantly (p<0.01) increased mean
metHb levels when compared to NO (10 ppm)
alone. Combined exposure caused significant
behavioral effects at both levels.  Com-
bined exposure also affected early auditory-
evoked potential components (P10 and N^; the
effect was more pronounced at higher dose
level (500 ppm CO + 50 ppm NO) than at the
lower levels, indicating a dominant role
for NO.

Combined exposure to the  three noxious
gases caused no morphological changes in
the lung, pulmonary, and coronary arteries,
or aorta.

No consistent changes in pulmonary function
indices were observed in any of the groups
exposed to the pollutants alone or in
combination with CO. Hematological and
biochemical changes were  within the
normal range.  Combined exposure to CO
+ NO2 did not  increase the severity of
the histopalhological changes observed
in the lungs of rats exposed to NO2 alone.

Combined  exposures caused no consistent
changes  in pulmonary function
indices, hematology, or biochemical
or histological parameters.
 Exposure to SO2 alone or SO2 +
 CO caused no teratogenic
 effects.
                                                                                 Groll-Knapp et al. (1988)
                                                                                 Hugod (1979)
                                                                                                             Busey (1972)
                                                                                                                                                       Busey (1972)
Murray et al. (1978)

-------
                        TABLE 11-4  (cont'd).  COMBINED  EXPOSURE TO CARBON MONOXIDE AND OTHER POLLUTANTS
i
Pollutant
CO
S02
Concentration
250 ppm
70 ppm
No ./sex/species
20/F/rabbit
(New Zealand white)
Treatment Observed Effects
Exposed to filtered air only, 70 ppm No teratogenicity was
SO2 alone, or 70 ppm SO2 + 250 ppm observed.
CO for 7 h/day during Days
6 to 18 of gestation; teratogenic
potential was evaluated
Reference
Murray et al. (1978)
    CO

    SO2
    CO
    PbClBr
N>
00
3 mg/m3
6 mg/m3
0.5 mg/m3
67.5 ppm
0.6 ppm
                      100 ppm
                      1,000 ppm
                                       3/-/human
24/M/rat
24/F/rat
(Sprague-Dawley)
                 5/M/rat
                 (Wistar)
Exposed to pure air for 5 min;
6 mg/m3 CO for 20 min; 6 mg/m3
CO + 0.5 mg/m3 SO2 for 5 min;
0.5 mg/m3 SO for 25 min; 6 mg/m3
CO + 0.5 mg/m3 for 25 min; or
3 mg/m3 CO +0.5 mg/m3 SO2 for
25 min; variations in ocular
sensitivity to light and color vision
were tested

Exposed to clean air alone, 67.5 ppm
CO alone; 0.6 ppm PbClBr alone; or
to 0.6 ppm + 67.5 ppm CO continuously
24 h/day, 7 days/week for
52 weeks; chronic toxicity was
assessed
                            Exposed to clean air only; 100 ppm
                            CO alone; 1,000 ppm CH2C12
                            only; or 100 ppm CO + 1,000 ppm
                            CH2CI2 for 3 h; mixed-
                            function oxidase activity and
                            blood COHb levels were examined
                                                                     Inhalation of 6 mg/m3 CO +
                                                                     0.5 mg/m3 SO2 for 5 min
                                                                     or CO alone for 20 min
                                                                     caused significant differences
                                                                     in light and color sensitivity
                                                                     when compared to controls.
No consistent changes in
pulmonary function, hematology,
or biochemistry were observed
between the treatment groups.
Combined exposure to CO +
PbClBr did not increase the
incidence of histological changes
observed in the kidneys of rats
exposed to PbClBr alone.

Combined exposure had an additive
effect.  Blood COHb was significantly
(p<0.001) increased in rats exposed
to CO + CH2CL2  (14.6%) compared with
with CO alone (8.8%). Combined
exposure significantly (p< 0.005)
increased the ethoxycoumarin-
0-deethylase activity in the
kidneys. Treatment had no effect
on liver microsomal oxidation.
                                        Mamatsashvili (1967)
Busey (1972)
                                                                                Kuippa et al. (1981)
                      1,500 ppm
                      2,000 ppm
                 -/-/dog
                 (Cowenose Mongrel)
                            Exposed to 1,500 ppm CO for
                            25 min followed by a 2-h
                            exposure to 2,000 ppm CH2C12 in
                            air also containing 150 ppm CO;
                            effects on the cardiovascular
                            system were evaluated
                                         Combined exposure of CO + CH2C12
                                         had no effect on the phsiologic
                                         response due to CO, instead CO
                                         antagonized the responses due to
                                       Adams (1975)
      Information was not reported in the original manuscript.

-------
            Combined exposure to CO and peroxyacetylnitrate (PAN) exerted no greater effect on
       the work capacity of healthy men (young and middle-aged smokers and nonsmokers) than did
       exposure to CO alone. Increases in blood COHb levels of smokers during the CO or
       CO+PAN exposures were observed  (Drinkwater et al., 1974; Raven et al.,  1974a,b; Gliner
 5     etal., 1975).
            Groll-Knapp et al. (1988) reported that combined exposure of rats to CO+NO for
       three hours caused a significant (p<0.01) increase in mean methemoglobin (metHb) levels
       when compared with metHb levels in rats exposed to NO alone. No significant changes were
       observed in blood COHb levels as compared with exposure to CO alone or to CO+NO.
10     Combined exposure also caused significant behavioral changes. Hugod (1979) reported that
       combined exposure to CO+NO+HCN for two weeks produced no morphological changes in
       the lungs, pulmonary arteries, coronary arteries, or aortas of rabbits.
            In a one-year inhalation toxicity study, no adverse toxic effects were seen in groups of
       rats exposed to relatively low levels of CO+NO2 or CO+SO2 as compared with rats exposed
15     to one of these pollutants alone (Busey, 1972). Murray et al.  (1978) observed no teratogenic
       effects  in offspring of mice or rabbits exposed to CO+SO2 for 7 hours/day during gestation
       Days 6 to 15 or 18, respectively.
            Halogenated hydrocarbons, such as polybrominated and polychlorinated biphenyls, are
       widely  used as organic solvents.  These chemicals are metabolized in the body to produce CO
20     which is readily bound to Hb.  Therefore, any additional exposure to CO, producing higher
       COHb  levels, could possibly cause greater health effects.  For example, up to 80% of inhaled
       methylene chloride (CHjClz) will be  metabolized to CO.  Inhalation  of 500 to 1000 ppm,
       therefore, would result in COHb levels of over 14%. This elevation in COHb can not only
       have a  significant effect when combined with CO exposure, but the  CO resulting from
25     metabolism generally requires a longer time to dissipate (Kurppa, 1984).
            In one study, combined exposure to CO+CH2C12 for three hours had an additive effect
       on blood COHb levels in rats (Kurppa et al.,  1981).  On the other hand, Adams (1975)
       reported that combined exposure to CO+CH2C12 did not have an additive effect on the
       physiologic response in the cardiovascular systems of dogs due to CO, instead, CO
30     antagonized the responses due to CH2C12.
       March 12, 1990                          11-29     DRAFT - DO NOT QUOTE OR CITE

-------
       11.3.2  Exposure to Combustion Products
            A common condition in an atmosphere produced by a fire is the presence of a rapidly
       changing combination of potentially toxic gases (primarily, CO, CO2, and HCN), reduced O2
       levels (hypoxic hypoxia), and high temperatures.  Combined exposure to these gases occurs
 5      during smoke inhalation under conditions of hypoxic hypoxia. In addition, both CO and CO2
       are common products of carbon-containing materials; consequently, accidental exposure to
       high levels of CO will rarely occur without simultaneous exposure to CO2. Exposure to CO
       and HCN is of concern because both CO and HCN produce effects by influencing tissue O2
       delivery.  Increased COHb reduces O2-carrying capacity and may interfere with tissue O2
10      release, while HCN inhibits tissue respiration.  Studies were conducted to determine the
       lexicological interactions of the combustion products with and without reduced O2.  (Also see
       Chapter 10, Section 10.4.1.5 for more discussion on CO and HCN.)
            Several studies have investigated the effects resulting from combined exposure to CO
       and combustion products from fires.  The experimental details and the associated effects for
15      each  study are summarized in Table 11-5.  The following is a brief discussion of the major
       findings.
            Rodkey and Collison (1979) reported a significant (p<0.02) decrease in mean survival
       time in mice jointly exposed until death to CO+CO2 compared with mice exposed to CO
       alone.  In contrast, Crane (1985) observed no differences in the times-to-incapacitation or
20      times-to-death in rats exposed until death to various concentrations of CO+CO2.  In a recent
       study, Levin et al. (1987a) demonstrated a synergistic effect between CO and CO2 in rats
       exposed to various concentrations of CO+CO2.  Simultaneous exposure to nonlethal levels of
       CO2 (1.7 to 17.3%) and to sublethal levels of CO (2500 to 4000 ppm) caused deaths in rats
       both during and following (up to 24 h) a 30-min exposure.  The rate of COHb formation  was
25      1.5 times greater in rats exposed to CO+C02 than in rats exposed to CO alone.
            Combined exposure to CO+HCN had an additive effect in rats as evidenced by
       increases in mortality rate and changes in COHb levels. COHb levels decreased as the CO
       level decreased and HCN level increased  (Levin et al., 1987b). HCN had a depressive effect
       on CO uptake and COHb formation, an effect that may explain the reason for the low COHb
30      levels (<50%) seen in some people who died in a fire (Levin et al., 1987b).  In contrast,
       March 12, 1990                         11-30    DRAFT - DO NOT QUOTE OR CITE

-------
1    —
ST   Combustion
H->   Product
                             TABLE 11-5.   COMBINED EXPOSURE TO  CARBON  MONOXIDE AND COMBUSTION PRODUCTS
                        Concentration
                                            No./Sex/Species
                                                     Treatment
                                                                                              Observed Effects
                                                                                                                         Reference
o
53
n
      CO
6,000 ppm
2.1,2.3,
4.5, 5.4%
                                            6/F/rat
                                            (NMRI)
      CO

      CO,
5,000-14,000
ppm
4-13%
V-/rat
d
?>    CO
H    co2
 I
d
O
                        1,470-6,000
                        ppm
                        1.7-17.3%
                     6/M/rat
                     (Fischer 344)
                         Exposed to 6,000 ppm CO alone;
                         6,000 ppm CO + 2.1 or 4.5% CO2;
                         or 6,000 ppm CO + 2.3 or 5.4%
                         CO2 until death; O2 concentrations
                         were either 14 or 21 %; mean
                         survival time (MST) and fatal blood
                         COHb level were measured
                         Exposed to concentrations ranging
                         from 5,000 to 14,000 ppm CO
                         alone or with CO2 concentrations
                         ranging from 4 to 13%
                         continuously until death;
                         synergistic effects (times-to-
                         incapacitation ([ti]) or the times-
                         to-death ([td]) were evaluated.

                         Exposed to 1,470 to 6,000 ppm
                         CO or 2,500 to 4,000 ppm CO +
                         1.7 to 17.3% CO2for30min;
                         toxicological interactions
                         (mortality and COHb formation)
                         were evaluated
The MST was significantly
(p<0.02) decreased in rats
exposed to 6,000 ppm CO + 2.1 %
CO2 (18.4 min) or 4.5% CO2
(16.8 min) compared with CO
alone (22.4 min) in the presence
of 21% O2. Exposure to 14%
O2 + 6,000 ppm CO decreased
the MST to 9.6 min; addition
of 2.3% or 5.4% CO2had no
further effect on MST.  Combined
exposure (CO + CO^ had no
effect on fatal blood COHB.

No synergistic effects were
observed; no significant CO2
changes were observed in  the
endpoints (t; or tj) for added
CO2 compared to endpoints for
CO alone.
Exposure to CO alone caused
deaths at levels of 4,600 to
6,000 ppm and at COHb levels of
> 83 %.  Deaths were primarily
due to the high COHb, low
O2Hb, and hypoxia. Combined
exposure to >2500 ppm CO +
1.7 to 17.3% CO2 caused deaths
during exposure and the follow-
up period (24 h).  No
mortality occurred in rats at
< 2,500 ppm CO alone regardless
of CO2 concentrations.  The rate
of COHb formation was 1.5 times
greater in rats exposed to
2,500 ppm CO + 5.25% CO2 than
in rats exposed to 2,500 ppm CO
alone. The COHb  equilibrium level
was the  same
Rodkey and Collison (1979)
Crane (1985)
Levin et al. (1987a)

-------
                TABLE 11-5 (cont'd).   COMBINED EXPOSURE TO  CARBON MONOXIDE AND COMBUSTION PRODUCTS
ET   Combustion
i—>   Product
rO   	
                 Concentration
                    No./Sex/Species
                                                                     Treatment
                                                                       Observed Effects
                                                                                                                                                   Reference
CO

HCN
1,242-4,600
ppm
43.2-126.4
ppm
6/M/rat
(Fischer 344)
HCN
200 ppm
0.5 ppm
12-24/M/rabbit
(albino)
                  0.63-0.66%
                  0.325-0.375%
                  4-9 mg/kg
                  1-6.35 mg/kg
                    10/M/mouse
                    (ICR)
                                                                  Exposed to 4,600 ppm CO alone or
                                                                  1,242 to 3,450 ppm CO + 43.2 to
                                                                  126.4 ppm HCN for 30 min;
                                                                  lethality and COHb formation were
                                                                  measured as toxicological
                                                                  endpoints
Exposed to 0.5 ppm HCN alone or
0.5 ppm HCN  + 200 ppm CO for 1 or
4 weeks; morphological changes
in the lung, pulmonary arteries,
coronary arteries, or aorta were
evaluated

Mice were exposed to clean air or
to atmospheric  concentrations of
0.63-0.66% CO for 3 min (pre-
treatment) and then injected
ip with 4-9 mg/kg KCN
(78%) for the combined exposure, but
was reached in 10 min in the
presence of CO2 and 20 min in the
absence of CO2.  Combined exposure
increased the concentration of
COHb, caused severe acidosis, and
prolonged the recovery of acidosis
following cessation of exposure.
Exposure to CO2 alone produced no
mortality or incapacitation.

Combined exposure to CO + HCN
had an additive effect as
evidenced by increased mortality.
As the concentrations of HCN
increased, the animals died at
lower CO concentrations and
presented lower levels of COHb
at death.  When rats were
exposed  to 1,470 ppm CO alone
or 1,450 ppm CO + 100 ppm HCN,
the initial rate of COHb
formation was the same in the
presence or absence of HCN;
however, the final COHb  level
was lower in the presence of
HCN, indicating a depressive
effect of HCN on CO uptake and
the low COHb formation.

Exposure to HCN alone or in
combination with CO produced
no morphological changes in
the lung, pulmonary arteries,
coronary arteries, or aorta.
                                                                The LDjo value was significantly
                                                                (p< 0.05) lower for KCN
                                                                (6.51 mg/kg) in CO-pretreated
                                                                mice than in air-pretreated
                                                                mice (7.90 mg/kg).
                                                                                      Levin et al. (1987b)
                                                                                                                                                   Hugod (1979)
                                                                                                                                                        Norris et al. (1986)

-------
£
8*
                       TABLE 11-5  (cont'd).   COMBINED EXPOSURE TO CARBON MONOXIDE AND COMBUSTION PRODUCTS
     Combustion
     Product
                        Concentration
                                            No./Sex/Species
                                                                             Treatment
                                                                                                                      Observed Effects
                                                                                                      Reference
Ui
LO
     CO
     KCN
                        1,000 ppm
                        2,500 ppm 7.5
                        rag/kg
20/M/mouse
(Swiss Webster)
In another experiment, mice were
pretreated with either saline
(0.1 mL/10, ip) or KCN (1 to 6.35
mg/kg, ip) and then were exposed
via inhalation to CO in the range
of 0.325 to 0.375% CO for 4
min; lethality and blood CO
and cyanide concentrations were
measured

Preexposed to 1 ,000 ppm CO for
4 h followed by a single ip
injection of 7.5 mg/kg KCN,
24 h later; effects on KCN-
induced letahlity were studied

Preexposed to 7.5 mg/kg KCN (ip)
24 h prior to exposure to
2,500  ppm CO for 2 h; effects
on KCN-induced lethality were
studied
Sublethal doses of KCN (3.5 to
6.35 mg/kg) produced a synergistic
effect in mortality:  40-100%
mortality in KCN pretreated mice
compared to 10-20% in saline
pretreated mice. There were
no differences in CO or cyanide
blood levels between  these
treatment groups.

No alterations in lethality in
CO + KCN group as compared
with control + KCN group.
                                                                                                               Pretreatment with KCN had no
                                                                                                               significant effect on lethality
                                                                                                               associated with subsequent
                                                                                                               exposure to CO.
Winston and Roberts (1975)
 I
o
o
§
     CO
     (under
     con-
     ditions
     of
     hypoxic
     hypoxia)
O   co
a     (under
     con-
O   ditions of
^   hypoxic
O   hypoxia)
                        CO: 6,000
                        ppm
                        O2: 14 or
                        21%
6/F/rat
(NMRI)
                        CO: 500,
                        1,000 or
                        2,500 ppm
                        O2: 7 or
                        10%
20/M/mouse
(Swiss Webster)
Exposed to 6,000 ppm CO until
death in the presence of either
14 or 21 % O2; MST and fatal blood
COHb levels were measured
Mice were preexposed to 500 or
1,000 ppm CO for 4 h and then
exposed to 2,500 ppm for 2 h,
24 h laterexposure to CO.

Preexposed to 500 or 1,000 ppm CO
for 4 h and then exposed to
7% 02for2h, 24 h
later
MST was significantly
(p<0.01) decreased in the
presence of low (14%) O2
(9.6 min) compared to that
of high (21%) O2 (22.4 min)
levels.  A significantly
(p <0.01) higher level of COHb
was observed in rats treated with
14% O2 (89.4%) compared with those
treated with 21% O2 (83.4%).

Preexposure to CO caused a
significant (p<0.05) decrease
in lethality during subsequent
                                                                                                               Preexposure to CO followed by
                                                                                                               exposure to O2 had no effect
                                                                                                               on lethality. Preexposure to
                                                                                                               CO had no protective effect
                                                                                                               against hypoxic hypoxia.
Rodkey and Collison (1979)
Winston and Roberts (1975)

-------
I
                      TABLE 11-5 (cont'd).  COMBINED EXPOSURE TO CARBON  MONOXIDE AND COMBUSTION PRODUCTS
      Combustion
      Product
                       Concentration
                    No./Sex/Species
                                                                           Treatment
                                                                        Observed Effects
                                                                                                                   Reference
                                                                   Preexposed to 10% O2 for 4 h
                                                                   and then exposed to 2,500 ppm CO
                                                                   for 2 h, 24 h later
                                                                   Preexposed to 1,000 ppm CO or 10%
                                                                   O2 for 4 h and then exposed
                                                                   to 2,500 ppm CO for up to 2 h,
                                                                   24 h later
                                                                                     Preexposure to O2 followed by
                                                                                     exposure to CO significantly
                                                                                     (p<0.05) decreased lethality
                                                                                     compared to controls.

                                                                                     Preexposure to either CO or O2
                                                                                     had no significant effect on
                                                                                     O2-consumption level.
                                                                                     Alterations in CO lethality
                                                                                     were not associated with
                                                                                     alterations in COHb levels.
H
 I
O
O
25
O
      CO
      (under
      condi-
      tions of
      hypoxic
      hypoxia)
CO: 500 or
1,000 ppm
O2: 6-21 % or
11.8-20.5%
*-/M/mouse
(Swiss Webster)
Exposed to reduced O2 (6-21%)
alone or reduced O2 (11.8 to
20.5%) + 500 ppm CO, or 20.2%
O2 + 1,000 ppm CO for 20 min;
animals were subjected to
behavioral tests that determined
reaction time and performance of
the animals in a mouse pole-jump
apparatus
Reaction time gradually
increased with a decrease in
O2to 10%.  At  <10% O2,
reaction time increased
dramatically and animal
performance decreased almost
immediately. At reduced O2
levels + CO, the decreases in
performance were even greater
than those seen in mice
exposed to reduced O2 levels
only. At 20% O2 (close to
ambient level) + 1,000 ppm CO,
performance was nearly
completely degraded.
Cagliostro and Islas (1982)
      'Data not provided in the published manuscript.
 n

-------
        Hugod (1979) reported that exposure to HCN alone or to CO+HCN for one to four weeks
        produced no morphological changes in the lung, pulmonary and coronary arteries, or aorta of
        rabbits.
            Combined exposures to CO+KCN have produced conflicting results.  Norris et al.
 5      (1986) reported that the LDjo  values were significantly lower in mice pretreated with CO
        prior to intraperitoneal injection of KCN.  Sublethal doses of KCN produced a synergistic
        effect on mortality.  On the other hand, Winston and Roberts (1975) observed no alterations
        in lethality in mice pretreated with CO and then treated with intraperitoneal injections of
        KCN.
 10          A number of studies examined the effects of CO administered under conditions of
        hypoxic hypoxia.  Rodkey and Collison (1979) observed a lower mean survival time and a
        higher level of COHb in mice exposed to CO in the presence of low O2 (14%) than in those
        exposed to an ambient O2 (21%) level. Winston and Roberts (1975) showed that preexposure
        of mice to CO, followed by exposure to 7% O2 24 h later, had no effect on lethality as
 15      compared with controls exposed to 7% O2 only. Thus, preexposure to CO had no protective
        effect against hypoxic hypoxia. However, preexposure to 10% O2 caused a significant
        decrease in lethality in mice exposed 24 h later to CO.  Alterations in colethality were not
        associated with alterations in COHb levels. In a behavioral study in mice, Cagliostro and
        Islas (1982) showed that reaction times gradually increased with a decrease in O2 levels to
20      10%. At < 10% O2, reaction time increased dramatically.  At reduced O2 levels and in the
        presence of CO, the decreases in performance were even greater than those observed in mice
        exposed to reduced O2 levels alone.

        11.3.3  Exposure to Other Environmental Factors
25      11.3.3.1 Environmental Heat
            Several of the studies (Drinkwater et al., 1974; Raven et al., 1974a,b;  Gliner et al.,
        1975) describing the effects of CO exposure alone and  CO combined with peroxyacetyl nitrate
        (PAN) on exercise performance in healthy adult men, reviewed in Sections 10.3.2 and
        11.3.1, also evaluated the effects of heat stress.  Subjects were exposed to 50 ppm CO and/or
30     0.27 ppm PAN in environmental exposure chamber conditions of 30% RH at 25 and 30°C.
       In these studies, O2 uptake and exercise duration were assessed during both maximal and

       March  12,  1990                          11-35     DRAFT - DO NOT QUOTE OR CITE

-------
       submaximal exercise. Heat stress was more effective in reducing maximal exercise
       performance than exposure to the polluted environments.  The combination of heat stress with
       CO exposure was found to be important, however, in producing symptom complaints during
       submaximal exercise at 35 °C that were not found at 25 °C.  Further work in the same
 5     laboratory (Bunnell and Horvath,  1989) also demonstrated that subjects experienced
       significant levels of symptoms, particularly exertion symptoms, associated with elevated
       COHb when exercising in the heat.  These studies suggest, therefore, that heat stress may be
       an important determinant of changes in exercise performance when combined with exposure
       to CO.
10          Yang et al. (1988) studied the combined effects of high temperature and CO exposure in
       laboratory mice and rats.  They were exposed one hour per  day for 23 consecutive days to
       environmental chamber temperatures of 25 and 35 °C at CO concentrations ranging from 580
       to 607 ppm.  Carboxyhemoglobin levels after one hour of exposure ranged from 31.5 to
       46.5%.  The toxicity of CO to mice, based on the LCjo and survival time, was found to be
15     three times higher at 35 °C.  High temperature also was found to enhance the effects of CO
       on the function of oxidative phosphorylation of liver mitochondria in rats. Body temperature
       regulation and heat tolerance also was affected by CO exposure.  The authors speculate that
       these effects of combined exposure to CO and high temperature are due to the production of
       higher COHb, possibly due to hyperventilation.
20
       11.3.3.2 Environmental Noise
            Fechter et al. (1987; 1988) and Fechter (1988) speculated that the cochlea would be
       particularly susceptible to injury when exposed to both CO and environmental noise. The
       rationale for this potential effect was that CO exposure could impair cochlear oxygenation at a
25     time when auditory metabolism was likely to be enhanced by noise exposure. Using
       laboratory rats exposed to high levels of CO (250 to 1200 ppm for 3.5 h) with and without
       broad-band noise (105 dBA for 120 min), the authors were  able to show that CO acts in a
       dose-dependent  manner to potentiate noise-induced auditory dysfunction. While CO or noise
       alone did not have an effect, CO combined with noise produced a more severe loss of hair
30     cells at the basal end of the cochlea.  Auditory threshold loss for the combined exposure was
       evident at all frequencies tested but was greatest for high-frequency tones. A previous pilot

       March  12, 1990                         11-36    DRAFT - DO NOT QUOTE OR CITE

-------
       study by Young et al. (1987) conducted at 1200 ppm CO also showed that combined exposure
       to noise and CO produced high-frequency shifts of greater magnitude than those produced by
       exposure to noise alone.
            Results from the toxicologic studies in rats suggest that combined exposure to noise and
 5     CO may be important in evaluating potential risk to exposed individuals.  The CO levels used
       in these studies, however, are much greater than those encountered in the typical ambient, or
       even in the typical occupational environment. Thus, it is difficult to predict how relevant
       these studies are to actual conditions of human exposure. An early epidemiologic study by
       Lumio (1948) in operators of CO-fueled vehicles found significantly greater permanent
10     hearing loss than expected after controlling for possible confounding factors.  More recently,
       Sulkowski and Bojarski (1988) studied age-matched workers with similar length of duty
       employed in foundry, cast iron, and cast steel positions of a mining devices factory where CO
       and noise exposure varied.  Careful otological and audiometric examinations were performed
       on these workers.  The group exposed to the combined effects of 95 dBA noise and a mean
15     concentration of 41 ppm CO did not experience any greater hearing loss than the groups
       exposed only to noise (96 dBA) or CO  (45 ppm).  In fact, a permanent threshold shift was
       significantly larger in workers exposed  to noise alone than those exposed to the combined
       influence of CO and noise. This study  needs to be verified, however, at similar, relevant
       exposure levels before any definitive conclusions can be made regarding the potential of
20     lower-level CO to potentiate noise-induced auditory loss in humans.

       11.3.4 Summary
            Much of the data concerning  the combined effects of CO and other pollutants found in
       the ambient air are based on animal experiments. Only a few human studies are available.
25     Early studies in healthy human subjects by Hackney et al.  (1975a,b), Raven et al. (1974a,b),
       Gliner  et al. (1975), and Drinkwater et al.  (1974) on common air pollutants such as NO2, O3,
       or PAN and more recent work on CO + O3 by DeLucia et al. (1983) failed to show any
       interaction from combined exposure.
            In animal studies, no interaction was observed following combined exposure of CO and
30     pollutants such as HCN, NO2, SO2, or PbClBr (Hugod, 1979; Busey, 1972; Murray et  al.,
       1978).   However, an additive effect was observed following combined exposure of high levels

       March  12, 1990                          11-37    DRAFT - DO NOT QUOTE OR CITE

-------
       of CO + NO (Groll-Knapp et al.,  1988), and a synergistic effect was observed after
       combined exposure to CO and O3 (Murphy, 1964).
           Toxicological interactions of combustion products, primarily CO, CO2) and HCN, from
       indoor and outdoor fires, have shown a synergistic effect following CO + CO2 exposure
 5      (Rodkey and Collison, 1979; Levin et al., 1987a) and an additive effect with CO + HCN
       (Levin et al., 1987b). Additional studies are needed, however, to evaluate the effects of CO
       under conditions of hypoxic hypoxia.
           Finally, laboratory animal studies (Yang et al., 1988; Fechter et al., 1988; Young
       et al.,  1987) suggest  that the combination of environmental factors such as heat stress and
10      noise may be important determinants of health effects occurring in combination with exposure
       to CO.  Of the effects described, the one potentially most relevant to typical human exposures
       is a greater decrement in exercise performance seen when heat stress is combined with
       50 ppm CO (Drinkwater et al., 1974; Raven et al., 1974a,b; Gliner et al., 1975).

15
       11.4  ENVIRONMENTAL TOBACCO SMOKE
           A common source of CO for the general population comes from tobacco smoke, along
       with other  primary sources arising  from the environment. Exposure to tobacco smoke not
       only affects the COHb level of the smoker himself, but under some circumstances, such as in
20      poorly ventilated spaces, tobacco smoke exposure also  can affect nonsmokers.  For example,
       acute exposure (1 to  2 h) to smoke-polluted environments has been reported to cause an
       incremental increase  in nonsmokers'  COHb of about 1% (Jarvis,  1987).  In addition to CO,
       other products inhaled by the affected individuals, such as NO2, HCN, nicotine, and potential
       carcinogens contained in tobacco smoke, may produce subtle physiological and biochemical
25      effects in both the smoker and nonsmoker.  Possible pathological changes due to the
       interaction of CO and these other constituents of tobacco smoke that may occur in the lungs
       and other tissues remain to be elucidated.
            A detailed discussion of the possible health effects due to CO emitted from tobacco
       smoke is beyond the scope of this document.  Those interested in the problems related to
30     smoking tobacco (i.e., carcinogenesis and cardiovascular and pulmonary  disease) should refer
       to review documents specifically concerned with these matters (U.S. Department of Health

       March  12, 1990                         11-38     DRAFT - DO NOT QUOTE OR CITE

-------
       and Human Services,  1983; U.S. Department of Health, Education and Welfare, 1979,
       1972).  In addition, a number of sources have reviewed the potential health effects of tobacco
       smoke on nonsmokers (Fielding and Phenow, 1988; Hulka, 1988; Mohler, 1987; Surgeon
       General of the United States, 1986; National Research Council, 1986).
 5          Tobacco smoking has been found to result in higher COHb levels  than exposure to
       ambient concentrations of CO.   The actual quantity of CO entering the lung depends on the
       form in which tobacco is smoked,  the pattern of smoking, and the depth of inhalation
       (Robinson and Forbes, 1975).  Very little CO (approximately 5%) is absorbed in the mouth
       and larynx, therefore most of the CO available for binding to Hb must reach the alveoli in
10     order to raise the level of COHb present in the blood.  The CO concentration in tobacco
       smoke is approximately 4.5% (45,000 ppm).  It has been estimated that a smoker may be
       exposed to 400 to 500 ppm CO for the approximately 6 min that it takes to smoke a typical
       cigarette, producing an average baseline COHb  of 4%, with a typical range of 3 to 8%.
       Heavy smokers can have COHb levels as high as 15%. In comparison, nonsmokers average
15     about 1% COHb in their blood. (See Chapter 8 for more information on CO exposure in the
       population.) As a result of the higher baseline COHb levels,  smokers are actually excreting
       CO into the air rather than inhaling it from the ambient environment. Smokers may even
       show an adaptive response to the elevated COHb levels, as evidenced by increased red cell
       volumes or reduced plasma volumes (Smith and Landaw, 1978a,b).  For these reasons, EPA
20     previously has not considered active smokers in determining the need for a margin of safety
       for the  CO NAAQS (Federal Register, 1980).   This position was affirmed by the Clean Air
       Scientific Advisory Committee of EPA's Science Advisory Board during review of the
       previous CO criteria document (U.S. Environmental Protection Agency, 1979).
            The effects of CO from tobacco smoke have been discussed in other sections of the
25     document.  Human experimental studies suggested that acute effects of tobacco smoke on
       maximal exercise performance  are similar to those described for healthy subjects exposed to
       CO. Prospective and retrospective epidemiological studies identified tobacco smoke as one of
       the major factors in the development of cardiovascular disease.  Tobacco smoke may
       contribute to the development and/or aggravation of effects in exposed individuals through the
30     action of several independent or complementary mechanisms, one of which is the formation
       of significant levels of COHb.  Unfortunately, attempts to separate the CO effects of tobacco

       March  12, 1990                         11-39     DRAFT - DO NOT QUOTE OR CITE

-------
        smoke from the potential effects of other substances present in the smoke have been
        unsuccessful.  (For a discussion of these studies, see Section 10.3).
             In summary, although tobacco smoke is another source of CO for smokers as well as
        nonsmokers, it is also a source of other chemicals with which environmental CO levels could
 5      interact.  Available data strongly suggest that acute and chronic CO exposure attributed to
        tobacco smoke can affect the cardiopulmonary system, but the potential interaction of CO
        with other products of tobacco smoke confounds the results.  In addition, it is not clear if
        incremental increases in COHb caused by environmental exposure would actually be additive
        to chronically elevated COHb levels due to tobacco smoke, because some physiological
10      adaptation may take place.  There is, therefore, a need for further research to describe these
        relationships better.
       11.5 REFERENCES
15


       Adams, J. D. (1975) The effects of carbon monoxide and methylene chloride on the canine heart [Ph.D. thesis].
             College Station, TX: Texas A&M University. Available from: University Microfilms, Ann Arbor, MI;
20           publication no. 75-25,077.

       Agostoni, A.; Stabilini, R.; Viggiano, G.; Luzzana, M.; Samaja, M. (1980) Influence of capillary and tissue PQ
             on carbon monoxide binding to myoglobin: a theoretical evaluation. Microvasc. Res. 20: 81-87.

25     Brewer, G. J.; Eaton, J. W.; Weil, J. V.; Grover, R. F. (1970) Studies of red cell glycolysis and interactions
             with carbon monoxide, smoking, and altitude. In: Brewer, G. J., ed. Red cell metabolism and function:
             proceedings of the first international conference on red cell metabolism and function; October 1969; Ann
             Arbor, MI.  New York, NY: Plenum Press; pp. 95-114. (Advances in experimental medicine and biology:
             v. 6).
30
       Brewer, G. J.; Sing, C. F.; Eaton, J. W.; Weil, J. V.; Brewer, L. F.; Grover, R. F. (1974) Effects on
             hemoglobin oxygen affinity of smoking in residents of intermediate altitude. J. Lab. Clin. Med.
             84: 191-205.

35     Bunnell,  D. E.; Horvath, S. M. (1989) Interactive effects of heat, physical work, and CO exposure on
             metabolism  and cognitive task performance. Aviat. Space Environ. Med. 60: 428-432.

       Busey, W. M. (1972) Chronic exposure of albino rats to certain airborne pollutants [unpublished material].
             Vienna, VA: Hazleton Laboratories, Inc.
40
       Cagliostro, D. E.; Islas, A. (1982) The effects of reduced oxygen and of carbon monoxide on performance in a
             mouse pole-jump apparatus. J. Combust. Toxicol. 9: 187-193.
        March 12, 1990                            11-40     DRAFT - DO NOT QUOTE OR CITE

-------
       Cobura, R. F. (1979) Mechanisms of carbon monoxide toxicity. Prev. Med. 8: 310-322. Coburn, R. F.; Forster,
              R. E.; Kane, P. B. (1965) Considerations of the physiological variables that determine the blood
              carboxyhemoglobin concentration in man. J. Clin. Invest. 44: 1899-1910.

 5     Collier, C. R.; Goldsmith, J. R. (1983) Interactions of carbon monoxide and hemoglobin at high altitude. Atmos.
              Environ. 17: 723-728.

       Cooper, R. L.; Dooley, B. S.;  McGrath, J. J.; McFaul, S. J.; Kopetzky, M. T. (1985) Heart weights and
              electrocardiograms in rats breathing carbon monoxide at altitude. Fed. Proc. Fed. Am. Soc. Exp. Biol.
10            44: 1048.

       Crane, C. R. (1985) Are the combined toxicities of CO and CO2 synergistic? J. Fire Sci. 3: 143-144.

       DeLucia, A. J.; Whitaker, J. H.; Bryant, L. R. (1983) Effects of combined exposure to ozone and carbon
15            monoxide (CO) in humans. In: Lee, S. D.; Mustafa, M. G.; Mehlman, M. A., eds. International
              symposium on the biomedical effects of ozone and related photochemical  oxidants; March; Pinehurst,
              NC. Princeton, NJ: Princeton Scientific Publishers,  Inc.; pp. 145-159.  (Advances in modern
              environmental toxicology: v. 5).

20     Denniston, J. C.; Pettyjohn, F. S.; Boyter, J. K.; Kelliher,  J.  C.; Hiott, B. F.; Piper, C. F.  (1978) The
              interaction of carbon monoxide and altitude on aviator performance: pathophysiology of exposure to
              carbon monoxide. Fort Rucker, AL: U. S. Army Aeromedical Research Laboratory; report no. 78-7.
              Available from: NTIS,  Springfield, VA;  AD-A055 212.

25     Drinkwater, B. L.; Raven, P. B.; Horvath, S. M.; Gliner, J. A.; Ruhling, R. O.; Bolduan, N. W.; Taguchi, S.
              (1974) Air pollution, exercise, and heat stress. Arch. Environ. Health 28: 177-181.

       Engen, T. (1986) The combined effect of carbon monoxide and alcohol on odor sensitivity. Environ. Int. 12:
              207-210.
30
       Fechter, L. D. (1988) Interactions between noise exposure and chemical asphyxiants:  evidence for potentiation of
              noise induced hearing loss. In: Berglund, B.; Berglund, U.; Karlsson, J.; Lindvall, T., eds. Noise as a
              public health problem - volume 3, performance behaviour, animal, combined  agents, and community
              responses; proceedings of the 5th international congress on noise as a public health  problem; August;
35            Stockholm, Sweden. Stockholm, Sweden: Swedish Council for Building Research; pp. 117-122.

       Fechter, L. D.; Thorne, P. R.; Nuttall, A. L. (1987) Effects of carbon monoxide on  cochlear electrophysiology
              and blood flow. Hear. Res. 27: 37-45.

40     Fechter, L. D.; Young, J. S.; Carlisle, L. (1988) Potentiation of noise induced threshold shifts and hair cell loss
              by carbon monoxide. Hear. Res. 34: 39-47.

       Federal Register. (1980) Carbon monoxide; proposed revisions to the national ambient air quality standards:
              proposed rule. F.  R. (August 18) 45: 55066-55084.
45
       Fielding, J. E.; Phenow, K. J. (1988) Health effects of involuntary  smoking. N.  Engl. J. Med. 319:  1452-1460.

       Forbes, W. H.; Sargent, F.; Roughton, F. J. W. (1945) The rate of carbon monoxide uptake by  normal men.
              Am. J. Physiol. 143: 594-608.
50
       Gliner, J. A.; Raven, P. B.; Horvath, S. M.; Drinkwater, B. L.; Sutton, J. C. (1975) Man's physiologic
              response  to long-term work during thermal and pollutant stress. J. Appl. Physiol. 39: 628-632.
        March 12,  1990                               11-41     DRAFT - DO NOT QUOTE OR CITE

-------
       Groll-Knapp, E.; Haider, M.; Kienzl, K.; Handler, A.; Trimmel, M. (1988) Changes in discrimination learning
              and brain activity (ERP's) due to combined exposure to NO and CO in rats. Toxicology 49: 441-447.

       Haagenson, P. L. (1979) Meteorological and climatological factors  affecting Denver air quality. Atmos. Environ.
 5            13: 79-85.

       Hackney, J.  D.; Linn, W. S.; Mohler, J. G.; Pedersen, E. E.; Breisacher, P.; Russo, A. (1975a) Experimental
              studies on human health effects of air pollutants: II. four-hour exposure to ozone alone and in
              combination with other  pollutant gases. Arch. Environ. Health 30: 379-384.
10
       Hackney, J.  D.; Linn, W. S.; Law, D. C.; Karuza, S. K.; Greenberg, H.; Buckley, R. D.; Pedersen, E. E.
              (1975b) Experimental studies on human health effects of air pollutants: III. two-hour exposure to ozone
              alone and in combination with other pollutant gases. Arch.  Environ. Health 30: 385-390.

15     Halperin, M. H.; McFarland, R. A.; Niven, J. L; Roughton, F.  J. W.  (1959) The time course of the effects of
              carbon monoxide on visual thresholds. J. Physiol. (London) 146: 583-593.

       Horvath, S.  M.; Bedi, J. F. (1989) Alteration in carboxyhemoglobin concentrations during exposure to  9 ppm
              carbon monoxide for 8 hours at sea level and 2134 m altitude in a hypobaric chamber.  JAPCA
20            39: 1323-1327.

       Horvath, S.  M.; Raven, P. B.;  Dahms, T. E.; Gray, D. J. (1975) Maximal aerobic capacity at different levels of
              carboxyhemoglobin. J. Appl. Physiol. 38:  300-303.

25     Horvath, S.  M.; Bedi, J. F.; Wagner, J. A.; Agnew, J. (1988) Maximal aerobic capacity at several ambient
              concentrations of CO at several altitudes. J. Appl. Physiol.  65:  2696-2708.

       Hugod, C. (1979) Effect of exposure to 0.5 ppm hydrogen cyanide singly or combined with 200 ppm carbon
              monoxide and/or 5 ppm nitric oxide on coronary arteries, aorta, pulmonary artery, and lungs in the
30            rabbit. Int.  Arch. Occup. Environ. Health  44: 13-23.

       Hulka, B. S. (1988) The health consequences of environmental tobacco  smoke. Environ. Technol. Lett.
              9: 531-538.

35     James, W. E.; Tucker, C. E.; Grover, R. F. (1979) Cardiac function in goats exposed to carbon monoxide. J.
              Appl. Physiol.: Respir.  Environ. Exercise  Physiol. 47: 429-434.

       Jarvis, M. J. (1987) Uptake of environmental tobacco smoke. In: O'Neill, I. K.; Brunnemann, K. D.; Dodet, B.;
              Hoffmann,  D., eds. Environmental carcinogens: methods of analysis and exposure measurement: volume
40            9 - passive  smoking. Lyon, France:  International Agency for Research on Cancer; pp. 43-58. (IARC
              scientific publications no.  81).

       Kim, Y. C.; Carlson, G. P. (1983) Effect of carbon monoxide inhalation exposure in mice on drug metabolism
              in vivo. Toxicol.  Lett.  19:  7-13.
45
       Kirkpatrick,  L. W.; Reeser, W. K., Jr. (1976) The air pollution carrying capacities of selected Colorado
              mountain valley ski communities. J. Air Pollut. Control Assoc.  26: 992-994.

       Knisely, J. S.; Rees, D. C.; Balster, R. L. (1989) Effects of carbon monoxide in combination with behaviorally
50            active drugs on fixed-ratio  performance in  the mouse. Neurotoxicol. Teratol.  11: 447-452.

       Kurppa, K. (1984)  Carbon monoxide. In: Aitio, A.; Riihimaki, V.; Vainio, H.,  eds. Biological monitoring and
              surveillance of workers  exposed to chemicals. New York, NY: Hemisphere Publishing Corp.;
              pp. 159-164.


         March 12, 1990                              11-42     DRAFT - DO NOT QUOTE OR CITE

-------
       Kurppa, K.; Kivisto, H.; Vainio, H. (1981) Dichloromethane and carbon monoxide inhalation:
              carboxyhemoglobin addition, and drug metabolizing enzymes in rat. Int. Arch. Occup. Environ. Health
              49: 83-87.

 5     Levin, B. C.; Paabo, M.; Gurman, J. L.; Harris, S. E.; Braun, E. (1987a) Toxicological interactions between
              carbon monoxide and carbon dioxide.  Toxicology 47: 135-164.

       Levin, B. C.; Paabo, M.; Gurman, J. L.; Harris, S. E. (1987b) Effects of exposure to single or multiple
              combinations of the predominant toxic gases and low oxygen atmospheres produced in fires. Fundam.
10            Appl. Toxicol. 9: 236-250.

       Lilienthal, J. L., Jr.; Fugitt, C. H. (1946) The effect of low concentrations of carboxyhemoglobin on the
              "altitude tolerance" of man.  Am. J. Physiol. 145: 359-364.

15     Lindsey, T. K. (1989)  [Letter to Dr. James McGrath concerning U.S. population living at 5,000 feet or more].
              Forestville, MD: July 4.

       Lumio, J. S. (1948) Hearing deficiencies caused by carbon monoxide (generator gas). Helsinki, Finland:
              Oto-Laryngological Clinic of the University.
20
       Luomanmaki, K.; Cobum, R. F. (1969) Effects of metabolism and distribution of carbon monoxide on blood and
              body stores. Am. J. Physiol. 217:  354-363.

       Mamatsashvili, M. I. (1967) Determination of the reflex effect of sulfur dioxide and carbon monoxide by
25            adaptometry and by tests of color vision. Gig. Sanit. 32: 226-230.

       McDonagh, P. F.; Reynolds, J. M.; McGrath, J. J.  (1986) Chronic altitude plus carbon monoxide exposure
              causes left ventricular hypertrophy but an attenuation of coronary capillarity. Presented at: 70th annual
              meeting of the Federation of American Societies for Experimental Biology; April; St. Louis, MO. Fed.
30            Proc. Fed. Am. Soc.  Exp. Biol. 45: 883.

       McFarland, R. A. (1970) The effects of exposure to small quantities of carbon monoxide on vision. In: Coburn,
              R. F., ed. Biological effects of carbon monoxide. Ann.  N. Y. Acad. Sci.  174: 301-312.

35     McFarland, R. A.; Roughton, F. J. W.; Halperin, M. H.; Niven, J. I.  (1944) The effects of carbon monoxide
              and altitude on visual thresholds. J. Aviat. Med. 15: 381-394.

       McGrath, J. J. (1988)  Body and organ weights of rats exposed to carbon monoxide at high altitude. J. Toxicol.
              Environ. Health 23: 303-310.
40
       McGrath, J. J. (1989)  Cardiovascular effects of chronic carbon monoxide and high-altitude exposure.  Cambridge,
              MA: Health Effects Institute; research report number 27.

       McMillan, D. E.; Miller, A.  T., Jr.  (1974) Interactions between carbon monoxide and d-amphetamine or
45            pentobarbital on schedule-controlled behavior. Environ. Res. 8: 53-63.

       Medical College of Wisconsin. (1974) Exposure of humans to carbon monoxide combined with ingestion of ethyl
              alcohol and the comparison of human  performance when exposed for varying periods of time to carbon
              monoxide.  Milwaukee, WI:  Medical College of Wisconsin; report no. MCOW-ENVM-CO-74-2.
50            Available from: NTIS, Springfield, VA; PB-242099.

       Miranda, J. M.;  Konopinski, V. J.; Larsen, R. I. (1967) Carbon monoxide control in a high highway tunnel.
              Arch. Environ. Health 15: 16-25.
        March 12,  1990                               11-43     DRAFT - DO NOT QUOTE OR CITE

-------
       Mitchell, D. S.; Packham, S. C.; Fitzgerald, W. E. (1978) Effects of ethanol and carbon monoxide on two
              measures of behavioral incapacitation of rats. Proc. West.  Pharmacol. Soc. 21: 427-431.

       Mitchell, R. S.; Judson, F. N.; Moulding, T. S.; Weiser, P.; Brock, L. L.; Kelble, D. L.; Pollard, J. (1979)
 5            Health effects of urban air pollution: special consideration of areas at 1,500 m and above. JAMA J. Am.
              Med. Assoc. 242:  1163-1168.

       Mohler, S. E. (1987) Passive smoking: a danger to children's health. J. Pediatr. Health Care 1: 298-304.

10     Montgomery, M. R.; Rubin, R.  J. (1971) The effect of carbon monoxide  inhalation on in vivo drug metabolism
              in the rat. J. Pharmacol.  Exp. Ther. 179: 465-473.

       Montgomery, M. R.; Rubin, R".  J. (1973) Oxygenation during inhibition of drug metabolism by carbon monoxide
              or hypoxic hypoxia. J. Appl. Physiol. 35:  505-509.
15
       Moore, L. G.; Rounds, S. S.; Jahnigen, D.; Grover, R. F.; Reeves, J. T. (1982) Infant birth weight is related to
              maternal arterial oxygenation at high altitude. J. Appl. Physiol.: Respir. Environ. Exercise Physiol.
              52: 695-699.

20     Murphy, S. D. (1964) A review of effects on animals of exposure to auto exhaust and some of its components. J.
              Air Pollut. Control Assoc. 14: 303-308.

       Murray, F. J.; Schwetz, B. A.; Crawford, A. A.;  Henck, J. W.;  Staples,  R. E. (1978) Teratogenic potential of
              sulfur dioxide and carbon monoxide in mice and rabbits. In: Mahlum, D. D.; Sikov, M. R.; Hackett, P.
25            L.; Andrew, F. D., eds. Developmental toxicology of energy-related pollutants: proceedings of the
              seventeenth annual Hanford biology symposium; October  1977; Richland, WA. Oak Ridge, TN: U.  S.
              Department of Energy, Technical Information Center; pp.  469-478. Available from: NTIS, Springfield,
              VA; CONF-771017.

30     National Research Council. (1977) Carbon monoxide. Washington, DC: National Academy of Sciences. (Medical
              and biologic effects of environmental pollutants).

       National Research Council. (1986) Environmental  tobacco smoke: measuring exposures and assessing health
              effects. Washington, DC: National Academy Press.
35
       Norris, J. C.; Moore, S. J.; Hume, A. S. (1986) Synergistic lethality induced by the combination of carbon
              monoxide and cyanide. Toxicology 40: 121-129.

       Pankow, D.; Ponsold, W.; Fritz, H. (1974) Combined effects of carbon monoxide and ethanol on the activities of
40            leucine aminopeptidase and glutamic-pyruvic transaminase in the plasma of rats. Arch. Toxicol.
              32: 331-340.

       Parving, H.-H. (1972) The effect of hypoxia and carbon monoxide exposure on plasma volume and capillary
              permeability to albumin. Scand. J. Clin. Lab. Invest. 30:  49-56.
45
       Peterson, J. E.; Stewart, R. D. (1975) Predicting  the carboxyhemoglobin  levels resulting from carbon monoxide
              exposures. J. Appl. Physiol. 39:  633-4538.

       Pettyjohn, F. S.; McNeil, R. J.; Akers, L.  A.;  Faber, J. M. (1977) Use of inspiratory minute volumes in
50            evaluation of rotary and  fixed wing pilot workload. Fort Rucker,  AL: U. S.  Army Aeromedical Research
               Laboratory; report no. 77-9.

       Pitts, G. C.; Pace, N. (1947) The effect  of blood  carboxyhemoglobin concentration on hypoxia tolerance. Am. J.
               Physiol. 148:  139-150.


         March  12, 1990                               11-44     DRAFT - DO  NOT QUOTE OR CITE

-------
       Raven, P. B.; Drinkwater, B. L.; Horvath, S. M.; Ruhling, R. O.; Gliner, J. A.; Sutton, J. C.; Bolduan, N. W.
              (1974a) Age, smoking habits, heat stress, and their interactive effects with carbon monoxide and
              peroxyacetylnitrate on man's aerobic power. Int. J. Biometeorol. 18: 222-232.

 5     Raven, P. B.; Drinkwater, B. L.; Ruhling, R. O.;  Bolduan, N.; Taguchi, S.; Gliner, J.; Horvath, S. M. (1974b)
              Effect of carbon monoxide and peroxyacetyl nitrate on man's maximal aerobic capacity. J. Appl. Physiol.
              36: 288-293.

       Robinson, J. C.; Forbes, W. F. (1975) The role of carbon monoxide in cigarette smoking: I. Carbon monoxide
10            yield from cigarettes. Arch.  Environ. Health 30: 425-434.

       Rockwell, T. J.; Weir, F. W. (1975) The interactive effects of carbon monoxide and alcohol on driving skills.
              Columbus, OH: The Ohio State University Research Foundation; CRC-APRAC project CAPM-9-69.
              Available from: NTIS, Springfield, VA; PB-242266.
15
       Rodkey, F. L.; Collison, H. A. (1979) Effects of oxygen and carbon dioxide on carbon monoxide toxicity. J.
              Combust. Toxicol. 6: 208-212.

       Roth, R. A., Jr.; Rubin, R. J. (1976) Role of blood flow in carbon monoxide- and hypoxic hypoxia-induced
20            alterations in hexobarbital metabolism in rats. Drug Metab. Dispos. 4: 460-467.

       Smith, J. R.; Landaw, S. A. (1978a) Smokers' polycythemia. N. Engl. J. Med. 298: 6-10.

       Smith, J. R.; Landaw, S. A. (1978b) Smokers' polycythemia [letter to the editor]. N. Engl. J.  Med. 298: 973.
25
       Sulkowski, W. J.; Bojarski, K. (1988) Hearing loss due to combined exposure to noise and carbon monoxide - a
              field study. In: Berglund, B.; Berglund,  U.; Karlsson, J.; Lindvall, T., eds. Noise as a public health
              problem - volume 1, abstract guide: proceedings of the 5th international congress on noise as a public
              health problem; August; Stockholm, Sweden. Stockholm, Sweden: Swedish Council for Building
30            Research; p. 179.

       Surgeon General of the United States. (1986) The health consequences of involuntary smoking: a report of the
              Surgeon General. Rockville, MD: U. S. Department of Health and Human Services, Office on Smoking
              and Health.
35
       Toppling, D. L.; Fishlok, R. C.;  Trimble, R. P.; Storer, G. B.; Snoswell, A. M. (1981) Carboxyhemoglobin
              inhibits the metabolism of ethanol by perfused rat liver. Biochem. Int. 3: 157-163.

       U. S. Department of Health, Education, and Welfare. (1972) The health consequences of smoking: a report of the
40            Surgeon General. Washington, DC: Public Health Service; report no. DHEW(HMS) 72-7516.

       U. S. Department of Health, Education, and Welfare. (1979) Smoking and health: a report of the Surgeon
              General. Washington, DC: Public Health Service; publication no. DHEW (PHS) 79-50066.

45     U. S. Department of Health and Human Services. (1983) The health consequences of smoking: cardiovascular
              disease - a report of the Surgeon General. Washington, DC: Public Health Service; publication no.
              DHHS (PHS) 84-50204.

       U. S. Environmental Protection Agency. (1979)  Air quality  criteria for carbon monoxide. Research Triangle
50            Park, NC: Office of Health and Environmental Assessment, Environmental Criteria and Assessment
              Office; EPA report no. EPA-600/8-79-022. Available from: NTIS, Springfield,  VA; PB81-244840.

       Vollmer, E. P.; King, B. G.; Birren, J. E.; Fisher, M. B. (1946) The effects of carbon monoxide on three types
              of performance, at simulated altitudes of 10,000 and 15,000 feet. J. Exp. Psychol. 36:  244-251.


        March 12,  1990                              11-45     DRAFT - DO NOT QUOTE OR CITE

-------
       Wagner, J. A.; Horvath, S. M.; Andrew, G. M.; Cottle, W. H.; Bedi, J. F. (1978) Hypoxia, smoking history,
              and exercise. Aviat. Space Environ. Med. 49: 785-791.

       Weiser, P. C.; Morrill, C. G.; Dickey, D. W.; Kurt, T. L.; Cropp, G. J. A. (1978) Effects of low-level carbon
 5            monoxide exposure on the adaptation of healthy young men to aerobic work at an altitude of 1,610
              meters. In: Folinsbee, L. J.; Wagner, J. A.; Borgia, J. F.; Drinkwater, B. L.; Gliner, J. A.; Bedi, J. F.,
              eds. Environmental stress: individual human adaptations. New York, NY: Academic Press, Inc.;
              pp. 101-110.

10     Winston, J. M.; Roberts, R. I.  (1975) Influence of carbon monoxide, hypoxic hypoxia or potassium cyanide
              pretreatment on acute carbon monoxide and hypoxic hypoxia lethality. J. Phannacol. Exp. Ther. 193:
              713-719.

       Yang, L.; Zhang, W.; He, H.;  Zhang, G. (1988) Experimental studies on combined effects of high temperature
15            and carbon monoxide. J. Tongji Med. Univ. 8: 60-65.

       Young, J. S.; Upchurch, M. B.; Kaufman, M. J.; Fechter, L. D. (1987) Carbon monoxide exposure potentiates
              high-frequency auditory threshold shifts induced by  noise. Hear. Res. 26: 37-43.
        March 12, 1990                              11-46     DRAFT - DO NOT QUOTE OR CITE

-------
     12.  EVALUATION OF SUBPOPULATIONS POTENTIALLY
             AT RISK TO CARBON MONOXIDE EXPOSURE
       12.1  INTRODUCTION
            Most of the information on the human health effects of carbon monoxide discussed in
       Chapter 10 of this document has concentrated on two carefully defined population groups -
       young healthy, predominantly male adults and patients with diagnosed coronary artery
10     disease. On the basis of the known effects described, patients with reproducible exercise-
       induced angina appear to be best established as a sensitive group within the general population
       that is at increased risk for experiencing health effects (i.e., decreased exercise duration due
       to exacerbation of cardiovascular symptoms) of concern at ambient or near-ambient CO-
       exposure concentrations that result in COHb levels of <5%. A smaller sensitive group of
15     healthy individuals experience decreased exercise duration at similar levels of CO exposure,
       but only during short-term maximal exercise.  Decrements in exercise duration in the healthy
       population, therefore, would be mainly of concern to competing athletes rather than for
       nonathletic people carrying out the common activities of daily life.
            It is known, however, from both theoretical  work and from experimental research in
20     laboratory animals that certain other groups in the population are at potential risk to exposure
       from CO.  The purpose of this chapter is to explore the potential effects of CO in population
       groups that have not been studied adequately, but  which could be expected to be susceptible
       to CO because of underlying physiological status either due to gender differences, aging,
       preexisting disease, or because of the use of medications or alterations in their environment.
25     These probable risk groups include (1) fetuses and young infants; (2) pregnant women; (3) the
       elderly, especially those with compromised cardiopulmonary or cerebrovascular functions;
       (4) individuals with obstructed coronary arteries, but not yet manifesting overt
       symptomatology of coronary artery disease; (5) individuals with congestive heart failure;
       (6) individuals with peripheral vascular or cerebrovascular disease; (7) individuals with
30     hematological diseases (e.g., anemia) that affect oxygen-carrying capacity or transport in the
       blood; (8) individuals with genetically unusual forms of hemoglobin associated with reduced
       March 5, 1990                           12-1     DRAFT - DO  NOT QUOTE OR CITE

-------
       oxygen-carrying capacity; (9) individuals with chronic obstructive lung diseases;
       (10) individuals using medicinal or recreational drugs having CNS depressant properties;
       (11) individuals exposed to other pollutants (e.g., methylene chloride) that increase
       endogenous formation of CO; and (12) individuals who have not been adapted to high altitude
 5     and are exposed to a combination of high altitude and CO.
            Little empirical evidence currently is  available by which to specify health effects
       associated with ambient or near-ambient CO exposures in these probable risk groups.  Where
       the previous chapters dealt with documented evidence of CO exposure through controlled or
       natural laboratory investigations, this chapter will be more speculative.  An effort will be
10     made to determine the anticipated effects of CO in special subpopulations  that form a
       significant proportion of the population at large.
       12.2  AGE AND GENDER AS RISK FACTORS
15          The fetus and newborn infant are theoretically susceptible to CO exposure for several
       reasons.  Fetal circulation is likely to have a higher COHb level than the maternal circulation
       due to differences in uptake and elimination of CO from fetal hemoglobin. Because the fetus
       also  has a lower oxygen tension in the blood than adults, any further drop in fetal oxygen
       tension due to the presence of COHb could have a potentially serious effect.  The newborn
20     infant with a comparatively high rate of oxygen consumption and lower hemoglobin blood
       oxygen-transport capacity than most adults also would be potentially susceptible to the
       hypoxic effects  of increased COHb. Newer data from laboratory animal studies on the
       developmental toxicity of CO  suggest that prolonged exposure to high levels (> 100 ppm) of
       CO during gestation may produce a reduction in birthweight, cardiomegaly, and delayed
25     behavioral development (see Chapter 10,  Section 10.5).  Human data are scant and more
       difficult to evaluate, but further research  is warranted.  Additional studies, therefore, are
       needed in order to determine if chronic exposure to CO, particularly at low, near-ambient
       levels, can compromise the already marginal conditions existing in the fetus and newborn
       infant.
30          The effects of CO on maternal-fetal relationships are not understood well. In addition  to
       fetuses and newborn infants, pregnant women also represent a susceptible group because

       March 5, 1990                            12-2     DRAFT - DO NOT QUOTE OR CITE

-------
        pregnancy is associated with increased alveolar ventilation and an increased rate of oxygen
        consumption that serves to increase the rate of CO uptake from inspired air.  A perhaps more
        important factor is that pregnant women experience hemodilution due to the disproportionate
        increase in plasma volume as compared to erythrocyte volume.  This group, therefore, should
 5      be studied to evaluate the effects of CO exposure and elevated COHb levels.
             Seventy percent of the population of the United States survive to 65 years of age and
        30%  reach 80 years of age or more (Brody et al., 1987). In 1982 about 40% were 75 years
        of age or older, corresponding to about 10.7 million subjects.  The percentage of the
        population reaching 75 years of age or older is expected to increase to 49% by the year 2000
10      and to 56%  by the year 2080, making it one of the fastest growing segments of the
        population.  Thus, the aging population represents a potentially large subgroup that may be at
        risk to CO exposure.
             Changes in metabolic capability with age may make the aging population particularly
        susceptible to CO. Maximal oxygen uptake declines steadily with age at a  rate of about
15      0.9 mL/kg/min/year on the average.  However, the rate is only 0.65 in an  active  person and
        about 1.3 in an inactive person. Since inactivity is the most prevalent condition, especially in
        the elderly,  the maximal oxygen uptake at age 65 will be about 16 to 21 mL/kg/min or 1.2-
        1.6 L/min.  At age 75 the maximal oxygen uptake will be about 10 to 15 mL/kg/min or 0.75-
        1.1 L/min.  Note that these values refer to a healthy but inactive male person.
20           A person needs about  1 L/min or 10 mL/kg/min in maximal oxygen uptake  to meet
        daily  metabolic requirements. Thus, at age  75 many healthy subjects are on the border line
        with respect to performing ordinary activities, implying that even a low level of COHb might
        be enough to critically impair oxygen delivery to the tissues and severely limit daily metabolic
        requirements.
25           The above given data refer to inactive  males.  However, the decline by age in maximal
        oxygen uptake seems to be the same in inactive females.  Because females have about 25%
        lower maximal oxygen uptake expressed in milliliters per kilogram per minute, the critical
        age for a female will be about 70 years.  If a person is physically active on a regular basis,
        the critical age with respect to maximal oxygen uptake  would be expected to increase by 15 to
30      20 years.  Because females have a longer  life expectancy than males, the aging female
       March 5,  1990                           12-3     DRAFT - DO NOT QUOTE OR CITE

-------
       population potentially at risk to CO exposure would be expected to be larger than the aging
       male population.
 5     12.3  RISK OF CARBON MONOXIDE EXPOSURE IN INDIVIDUALS
             WITH PREEXISTING DISEASE
       12.3.1  Subjects with Coronary Artery Disease
            Coronary heart disease remains the major cause of death and disability in the United
       States both in males and females. According to the most recent data compiled by the
10     American Heart Association (1988), persons with diagnosed coronary artery disease numbered
       about 5.4 million in 1985 and current estimates are as high as seven million (U.S.
       Department of Health and Human Services, 1987; Collins, 1988).  These individuals have
       myocardial ischemia, which occurs when the heart muscle receives insufficient oxygen
       delivered by the blood.  For some, chest discomfort called angina pectoris can occur.  The
15     predominant type of ischemia, as identified by ST segment depression, in all patients with
       coronary artery disease, however, is asymptomatic (i.e., silent).  Especially patients who
       experience angina usually have more ischemic episodes that are asymptomatic.  About
       10% of middle-aged men develop a positive exercise test, one of the signs of ischemia.
       Nationally,  more than one million heart disease deaths occur each year, half of them being
20     fatal.  About 20% of all myocardial infarctions are silent.  Of the 500,000 survivors of
       hospitalized  myocardial infarction, about 10% are asymptomatic but have signs of ischemia.
       Thus, many more persons than currently known are not aware that they have coronary heart
       disease and may constitute a high-risk group.
            Persons with both asymptomatic and  symptomatic coronary artery disease have a limited
25     coronary flow reserve and,  therefore, will  be sensitive to a decrease in oxygen-carrying
       capacity induced by CO exposure (see Section 10.3.2). In addition,  CO might interfere with
       different vasomotor-active substances and thereby induce a vasoconstriction or prevent a
       vasodilation in the coronary arteries which will produce ischemia  that is likely to be silent.
       Unfortunately,  studies have not been performed specifically addressing the effect of CO
30     exposure on asymptomatic ischemia.
       March 5, 1990                           12-4     DRAFT - DO NOT QUOTE OR CITE

-------
        12.3.2  Subjects with Congestive Heart Failure
             Congestive heart failure is a major and growing public health problem in the United
        States.  It has been estimated that approximately three million Americans suffer from heart
        failure, and moreover, because the prevalence of heart failure is known to increase with age,
 5      improvements in the average life expectancy of the general population would be expected to
        increase the magnitude of the problem over the next few decades.
             Today about 75% of the patients with heart failure are above the age of 60 years (Brody
        et al., 1987).  About 400,000 new cases of heart failure are diagnosed every year in the
        United States, resulting in about 1.6 million hospitalizations.  The mortality rate is high,
10      between 15 to 60% per year.  The cause of death is often sudden death and because about
        65%  of heart failure patients have serious arrhythmias,  sudden death is thought to be due to
        arrhythmia. Each year 200,000 patients die.  The mortality is highest in Class 4 patients or
        in patients with a low maximal oxygen uptake (below 10 mL/kg/min).
             Patients with congestive heart failure have a markedly reduced circulatory capacity and,
15      therefore, may be very sensitive to any limitations in oxygen-carrying capacity. Thus,
        exposure to CO  certainly will reduce their exercise capacity and even be dangerous, especially
        if CO is determined to be proarrhythmogenic (see Section 10.3.2).  The etiology of heart
        failure is diverse but the dominating disease is coronary artery disease.  The large portion of
        heart failure patients with coronary artery disease, therefore, might be even more  sensitive to
20      CO exposure.

        12.3.3  Subjects with Other Vascular Diseases
             Peripheral  vascular disease is present in about 7% of both the male and female
        population and is more prevalent above age 65 years.  Cerebrovascular disease also is present
25      in about 6.6% of both the male and female population of the same ages.  Both of these
        conditions often are found in subjects with coronary artery disease. Both conditions also are
        associated with a limited blood flow capacity  and, therefore, should be sensitive to CO
        exposure.  It is not clear, however,  how low levels of exposure to CO will affect these
        individuals. Only one study (Aronow et al.,  1974), reviewed in the previous criteria
30      document (U.S.  Environmental Protection Agency, 1979), has been reported on patients with
        peripheral vascular disease.  Ten men with diagnosed intermittent claudication experienced a

        March 5, 1990                            12-5      DRAFT - DO NOT QUOTE OR CITE

-------
        significant decrease in time to onset of leg pain when exercising on a bicycle ergometer after
        breathing 50 ppm CO for 2 h (2.8% COHb).  Further research is needed, therefore, to better
        determine the sensitivity of subjects with vascular disease to CO.

 5      12.3.4  Subjects with  Anemia and Other Hematologic Disorders
             Clinically diagnosed low values of hemoglobin, characterized as anemia, is a relatively
        prevalent condition in the United States.  If the anemia is mild to moderate,  an inactive
        person is often asymptomatic.  However, due to the limitation in the oxygen-carrying capacity
        resulting from the low hemoglobin values, an anemic person should be more sensitive to low-
10      level CO exposure than a person with normal hemoglobin levels (see Section 10.3.2).  If
        anemia is combined with other prevalent diseases, such as coronary artery disease, the
        individual also will be at an increased risk to CO exposure.  Anemia is more prevalent in
        women and in the elderly,  already two potentially "high" risk groups.  An anemic person also
        will be more sensitive to the combination of CO exposure and high altitude.  Additional
15      studies are needed, therefore, in order to  determine the susceptibility of this group to CO
        exposure.
             Individuals with hemolytic anemia often have higher baseline levels of COHb because
        the rate of endogenous CO production from heme catabolism is increased. One of the many
        causes of anemia is the presence of abnormal hemoglobin in the  blood. For  example, in
20      sickle-cell disease the average lifespan of red blood cells with abnormal hemoglobin  S
        (Hb S) is 12 days compared to an average of 88 days in healthy  individuals with normal
        hemoglobin (Hb A).  As a result, baseline COHb levels can be as high as 4% (Solanki et al.,
        1988). In subjects with hemoglobin Zurich, where affinity for CO is 65 times that of normal
        hemoglobin, COHb levels range from 4 to 7% (Zinkham et al.,  1980).
25           There are over 350 variants to normal human hemoglobin (Zinkham et  al.,  1980).  In
        the Hb S variant, sickling takes place when deoxy Hb S in the red blood cell reaches a critical
        level and causes intracellular polymerization.  Oxygenation of the Hb S molecules in the
        polymer, therefore, should lead to a change in molecular shape,  breakup of the polymer, and
        unsickling of the cell. Carbon  monoxide was considered at one time to be potentially
30      beneficial because it ultimately would reduce the concentration of deoxy Hb S by converting
       March 5, 1990                           12-6    DRAFT - DO NOT QUOTE OR CITE

-------
       part of the hemoglobin to COHb. Exposure to CO, however, was not considered to be an
       effective clinical treatment because high COHb levels (>20%) were required.
            Other hematologic disorders can cause elevated concentrations of COHb in the blood.
       Ko and Eisenberg (1987)  studied a patient with Waldenstrom's Macroglobulinemia.  Not only
 5     was the COHb saturation  elevated, but the half-life of COHb was about three times longer
       than in a normal individual.  Presumably, exogenous exposure to CO, in conjunction with
       higher endogenous CO levels, could result in critical levels of COHb. However, because CO
       also can modify the characteristics of unstable hemoglobin, as demonstrated in patients with
       Hb S it is not known how ambient or near-ambient levels of CO would affect individuals with
10     these disorders.

       12.3.5  Subjects with Obstructive Lung Disease
            Chronic obstructive  pulmonary disease (COPD) is a prevalent disease especially among
       smokers.  It is estimated (U.S. Department  of Health and Human Services, 1987; Collins,
15     1988) that 14 million persons (-6%)  suffer from COPD in the United States and that a large
       number (>50%) of these individuals have limitations in their exercise performance
       demonstrated by a decrease in oxygen saturation during mild to moderate exercise.  In  spite
       of their symptoms, many  of them (-30%) continue to smoke and already may have COHb
       levels of 4 to 8%.  Subjects with hypoxia are also  more likely to have a progression of the
20     disease resulting in severe pulmonary insufficiency, pulmonary hypertension, and right heart
       failure.  Studies by Aronow et al. (1977) and Calverley et al. (1981), reviewed in
       Chapter 10,  suggest that individuals with hypoxia due to chronic lung disease such as
       bronchitis and emphysema may be susceptible to CO during submaximal exercise typically
       found during normal daily activity.
25          The prevalence of chronic asthma in the United States is estimated to be as high as
       9 million persons or about 4% of the total population (U.S. Department of Health and  Human
       Services,  1987; Collins, 1986, 1988).  There has been evidence that hospital admissions for
       asthma have increased considerably in the past few years, particularly among individuals less
       than 18 years of age.   Because asthmatics also can experience exercise-induced airflow
30     limitation, it is likely that they also would be susceptible to hypoxia.  It is not known,
       however,  how exposure to CO would affect these individuals.

       March 5,  1990                           12-7     DRAFT - DO NOT QUOTE OR CITE

-------
       12.4  SUBPOPULATIONS AT RISK FROM COMBINED EXPOSURE TO
             CARBON MONOXIDE AND OTHER CHEMICAL SUBSTANCES
       12.4.1  Interactions with Psychoactive Drugs
            There is almost a complete lack of data on the possible toxic consequences of combined
 5     CO exposure and drug use.  The most extensively studied interaction has been combined
       exposure to CO and alcohol. The previous criteria document (U.S. Environmental Protection
       Agency, 1979) reviewed an extensive human study of alcohol-CO interactions on driving
       performance (Rockwell and Weir, 1975).  In this study of actual driving behavior, alcohol
       and CO  effects were often additive,  and at 12% COHb concentrations, combined effects were
10     observed that were greater than the sum of the effects of CO and alcohol alone.  Two animal
       studies of alcohol-CO combinations  (Mitchell et al., 1978; Knisely et al., 1989) also provide
       evidence that the effects of alcohol on behavior can be enhanced by high concentrations of
       CO exposure.
            Thus is seems prudent to tentatively  conclude that the behavioral effects of alcohol may
15     be exacerbated under some conditions of CO exposure.  What is not known is the range of
       behavioral effects for which this occurs, the quantitative nature of the interaction, the
       mechanism of the combined effects, or the minimal COHb concentrations needed to see an
       interaction. Further research on this clearly is needed.  This is particularly the case when one
       considers the role of alcohol in our society and the likelihood of frequent opportunities for
20     combined alcohol use and CO exposure.  Some statistics from the recent report to Congress
       on Alcohol and Health illustrate the  potential problem (National Institute on Alcohol Abuse
       and Alcoholism, 1987), In 1984, the estimated per capita alcohol consumption per year in the
       United States was 2.65 gal of pure alcohol per person over the age of 14.  The National
       Institute on Alcohol Abuse and Alcoholism estimates  that two-thirds of the U.S. population
25     over the age of 18 drink alcohol, and one-half of these are moderate to heavy drinkers.
       Nearly 50% of all accidental deaths  are alcohol related.  Even a small interaction of CO
       exposure with alcohol would be magnified by the high incidence of these combinations.
            Other studies of interactions of CO and drugs have been conducted; however not nearly
       enough data exist upon which one could draw conclusions concerning populations at risk.
30     Some evidence from animal research indicates that CO exposure may alter the effects of
       pentobarbital, d-amphetamine, and chlorpromazine (McMillan and Miller, 1974; Knisely

       March 5, 1990                           12-8    DRAFT - DO NOT QUOTE OR CITE

-------
        et al., 1989). Because these drugs represent diverse classes of psychoactive drugs, and many
        other classes have not been examined at all, it must be concluded that this is an area of
        concern for which it is difficult at the present time to make recommendations that will have
        an effect on air quality standards.  The lack of data on possible interactions of CO exposure
 5      and drug use was identified in both the 1979 Criteria Document and an addendum to that
        document (U.S. Environmental Protection Agency, 1979, 1984).  Little has changed since
        then.

        12.4.2 Interactions with Cardiovascular Drugs
10           There are limited data currently available to determine if there is a possible interaction
        between CO exposure and different cardiovascular drugs. Drugs  used to treat patients with
        coronary artery disease, such as betablockers, calcium channel blockers, and nitrates, should
        be tested for potential interaction with CO because those patients  already are high-risk
        subjects.  Patients with angina that were used as subjects in studies on the effects of CO
15      exposure (see Chapter 10, Section 10.3) also were treated with these classes of drugs.
        Unfortunately, drug interactions were not investigated in most of  the studies. Only Allred
        et al. (1989a,b) analyzed their data for potential medication effect and no interactions with
        CO were found.  The only other available data dealt with the interaction of CO with
        betablockers and calcium blockers in smokers.  Deanfield et al. (1984) studied 10 smoking
20      patients with stable angina in a double-blind placebo controlled study.  He studied two
        betablockers, atenolol and propranolol, and one calcium blocker,  nifedipine.  The patients
        underwent exercise tests and Holter monitoring both when they still were smoking and after
        they had stopped smoking for one month. The performance and results from Holter
        monitoring showed improvement after the patients refrained from smoking.  The difference
25      was largest for nifedipine. Blood levels of propranolol were increased when the patients
        stopped smoking; levels of nifedipine and atenolol were unchanged.  Part of the decreased
        efficacy of the drugs  while smoking might be due to lower plasma levels and part of it might
        be due to some interaction on a cellular level. However,  it currently is not known if the
        interaction was due to nicotine and/or CO.
30           Another of  the high risk groups using multiple medications are heart failure patients.
        They often use digitalis, diuretics, vasodilators, and recently, inhibitors of angiotension

        March 5, 1990                             12-9     DRAFT -  DO  NOT QUOTE OR CITE

-------
       converting enzyme. If CO exposure modifies the responses to those drugs, the patients' status
       may deteriorate when the plasma levels of a drug are lower or the patients may develop side
       effects when the plasma levels of a drug are higher.  Due to the large number of high-risk
       patients with coronary artery disease and heart failure that use often very potent and multiple
 5     mediations,  this area needs to be addressed carefully through further research.

       12.4.3  Mechanisms of Carbon Monoxide Interactions with Drugs:  Need
                for Further Research
            Because data are generally lacking on CO-drug interactions, it should be useful to
10     speculate on some of the mechanisms by which CO might be expected to alter drug effects, or
       vice versa, and discuss possible populations at risk due to these potential interaction  effects.

       12.4.3.1  Metabolic Effects
            A mechanism by which CO might be expected to interact with many drugs is through
15     the modification of drug metabolism.  CO is known to bind to cytochrome P-450 in vitro
       (Gray, 1982), but the significance of this under physiological conditions is not known.
       Another section of this document (see Section 9.4) reviews the interactions of CO with
       oxidative metabolism and concludes that clinically relevant inhibition of these systems
       probably does not occur under most conditions of exposure.  If further research provides
20     evidence that these important drug-metabolizing systems are significantly compromised as a
       result of ambient CO exposure, then drugs dependent upon these systems for activation or
       deactivation would interact with CO exposures.  If changes in drug metabolism occur as a
       result of CO exposure, this would be of considerable practical importance.  It might be
       necessary to alter prescribing practices in heavily exposed populations. More research on this
25     is needed,

       12.4.3.2  Central Nervous System (CNS) Depression
            In the  absence of systematic data on the interactions of CO with psychoactive drugs, it is
       necessary to hypothesize mechanisms by which  such interactions might occur.  In the 1984
30     Addendum (U.S. Environmental Protection Agency,  1984), it was speculated that "drugs  with
       primary or secondary CNS depressant effects should be expected to exacerbate the

       March 5,  1990                           12-10    DRAFT - DO NOT QUOTE OR CITE

-------
        neurobehavioral effects of CO," presumably because of the generally depressant effects on the
        nervous system of CO itself.  On the other hand, one might also argue the converse (i.e., that
        CNS depressant drugs, because they might reduce cerebral metabolism and hence oxygen
        utilization, could lessen the neurobehavioral  effects of CO).  It should be obvious that
  5     speculation on these matters, in the absence of data, cannot be expected to yield answers upon
        which regulatory decisions could be made.  On the other hand, because of the overall
        sensitivity of the CNS to perturbations, it is possible that interactions of these types could
        occur and may even be quite pronounced.  Clearly,  more research on this is needed because
        we cannot rely on scientific speculation.
 10
        12.4.3.3 Alteration in Cerebral Blood Flow
             Another mechanism by which CO could be speculated to interact with certain drugs is
        through modification of cerebral blood flow.  Brain  hypoxia resulting from CO exposure may
        result in compensatory increases in cerebral blood flow (Dobler  et al., 1977).  Drugs that
 15      have vasoconstrictive effects on cerebral circulation could be hypothesized  to interfere with
        this compensatory mechanism and thus exacerbate the neurobehavioral toxicity of CO.  The
        methylxanthines, such as caffeine and  theophylline, have well-established central
        vasoconstrictive effects (Rail, 1980) and thus could be hypothesized to enhance CO-induced
        brain hypoxia.  On the other hand, their vasodilatory effects in the periphery (Rail,  1980)
20      might enhance the vasodilatory effects of CO.  Because of the widespread use of
        methylxanthines, these possible interactions may be of particular significance.
             As the oxygen-carrying capacity  of the  blood decreases with CO poisoning, many
        organs, including the brain, will compensate  their blood flow to  try and maintain proper
        tissue oxygenation.  Several studies using radiolabeled microspheres to measure cerebral
25      blood flow have demonstrated autoregulation and increased blood flow in response to CO
        (Koehler et al.,  1982).  However,  if the brain oxygen supply is inadequate  despite increased
        blood flow, further metabolic changes  will undoubtedly occur. The consequences of these
        metabolic changes with respect to their effect on the regulation of brain blood flow is
        uncertain.  More simply,  damage resulting from lack of proper brain oxygenation may alter
30      the brain vasculature's ability to regulate brain blood flow.  Damage to the vasculature
        previously has been shown to alter the vasculature's response to  vasoactive  agents.  For

        March 5, 1990                            12-11    DRAFT - DO NOT QUOTE OR CITE

-------
        example, it is known that following ischemia-induced injury and other types of brain injury,
        the brain releases polyunsaturated fatty acids from its phospholipids (Gardiner et al., 1981).
        These free fatty acids may be metabolized enzymatically to compounds such as prostaglandins
        and leukotrienes with effects on the cerebral vasculature.  For example, the metabolism of
 5      arachidonic acid generates prostaglandins and oxygen-free radicals that can cause cerebral
        vasodilation in normal animals.  However, when these radicals are produced in excess, as in
        acute, extreme hypertensive episodes, the free radicals initiate peroxidation of other
        unsaturated fatty acids (Kukreja et al., 1986).  These peroxides and oxygen radicals cause
        damage to the vascular endothelium and decrease the brain's capacity to regulate blood flow
10      in response to changes in arterial CO2 (Wei et al., 1981).  Additionally, the vascular damage
        caused by these oxygen radicals alters the normal cerebral arterial response to vasoactive
        agents, including neurotransmitters.  For example, acetylcholine, which is normally a dilator
        of cerebral arterioles, produces vasoconstriction after free radical-induced damage (Wei et al.,
        1985). Potentially, therefore, in an injured brain, acetylcholine may decrease an already
15      inadequate blood flow.
             Conceptually, tissue hypoxia produced by CO may stimulate arachidonic acid
        metabolism and production of prostaglandins and free radicals in a manner similar to hypoxia
        caused by ischemia or trauma.  Production of vasodilator free radicals or vasodilator
        prostaglandins may be a mechanism by which the brain increases its blood flow in response to
20      CO. Assuming this is the case, therapeutic agents or drugs of abuse that modify the
        arachidonic acid cascade may alter the brain's capacity to increase blood flow in response to
        CO. For example, if aspirin, indomethacin, or other cyclooxygenase inhibitors are present
        during exposure to CO, the brain's capacity to increase its flow may be diminished due to
        decreased capacity to form dilator prostanoids and free radicals.  An instance where this sort
25      of possibility is known to occur is in neonatal  animals, and possibly in neonatal humans.
        Investigators recently have shown in neonatal animals that indomethacin markedly diminished
        the  brain's capacity to increase its blood flow in response to hypoxia (Leffler and Busija,
        1987).
             A possible interaction between CO and nitrite exposure also might be predicted.  Nitrites
30      can be expected  to oxidize hemoglobin to methemoglobin leaving less hemoglobin to bind
        either to  O2 or CO.  However, since CO has a greater affinity than  O2 for hemoglobin, it is

        March 5, 1990                           12-12    DRAFT - DO NOT QUOTE OR CITE

-------
        most likely to expect additive effects in reduction of oxyhemoglobin.  In addition, some
        organic nitrites such as amyl nitrite, used to relieve angina pain, and butyl nitrite, an abused
        substance, produce significant peripheral vasodilation and, consequently, an abrupt drop in
        blood pressure and subsequent tachycardia.  These potent cardiovascular effect's which result
 5      from carbon monoxide exposure, might interfere with the enhanced cardiac output,
        particularly the output to sensitive organs such as the brain.  To date there are no published
        data on the combined effect of CO and nitrite exposure. However, there are limited data
        showing reasonably parallel consequences on auditory function (Fechter et al., 1987, 1989)
        when CO and butyl nitrite are given individually to rats.
10           Additionally, recent evidence shows that acetylcholine stimulates arachidonic acid
        metabolism (Busija et al.,  1987).  Whether hypoxia increases arachidonic acid metabolism via
        stimulation of acetylcholine release is uncertain, however, exogenous acetylcholine is known
        to stimulate brain prostaglandin production.  Assuming that endogenous acetylcholine release
        in response to CO-induced hypoxia is important, other agents such as atropine or
15      scopolamine, which block muscarinic receptors, could reduce the vasodilator response to CO.
        In an opposite manner, cholinesterase inhibitors (e.g., organophosphate or carbamate
        insecticides) that penetrate the blood brain barriers may magnify the dilator response to CO.
        Other agents that may modify the capacity of the brain's blood flow to regulate in response to
        CO are agents that cause an acute, large increase in blood pressure, thus inducing excess free
20      radical production, lipid peroxidation, and abnormal vascular reactivity.  Such an agent might
        include for example, cocaine, which when administered in large doses causes acute transient
        hypertension.
             While the above points are speculative, the possibility that therapeutic agents and drugs
        of abuse may alter the brain vasculature's capacity to respond to CO is a subject that bears
25      further consideration and investigation.

        12.4.4 Interactions with Other Chemical Substances in the Environment
             Besides direct ambient exposure to CO, there are other chemical substances in the
        environment that can  lead to increased COHb saturation when inhaled.  Halogenated
30      hydrocarbons used as organic solvents undergo metabolic breakdown by cytochrome P-450 to
        form CO and inorganic halide. Possibly the greatest concern regarding  potential risk in the

        March 5, 1990                            12-13     DRAFT - DO NOT QUOTE OR CITE

-------
       population comes from exposure to one of these halogenated hydrocarbons, methylene
       chloride (CHjCy, and some of its derivatives. Almost a million kilograms are produced each
       year, making it the second highest source of CO in the environment.  Although it is present in
       ambient air emissions, the highest concentrations of CH2C12 occur from various sources such
 5     as paint removers, cleaners, propellants, and from industrial manufacturing.
            From available experimental studies (see Section 11-3), it is not clear if combined
       exposure to CO and CH2C12 would produce an additive effect in humans.  Theoretically, acute
       CH2C12 exposure can result in a steady production of endogenous CO in tissues such as the
       lung, liver, kidney, heart, and brain that contain cytochrome P-450. Any histotoxic hypoxia
10     produced at the tissue level combined with hypoxic hypoxia due to the formation of COHb
       from endogenous as well as exogenous CO exposure could place exposed individuals at risk.
       12.5  SUBPOPULATIONS EXPOSED TO CARBON MONOXIDE AT
15            HIGH ALTITUDES
            For patients with coronary artery disease, restricted coronary blood flow limits oxygen
       delivery to the myocardium.  Carbon monoxide also has the potential for compromising
       oxygen transport to the heart.  For this reason, such patients have been identified as the
       subpopulation most sensitive to the effects of CO.  A reduction in the partial pressure of
20     oxygen (POj) in the atmosphere, as at high altitude, also has the potential for compromising
       oxygen transport. Therefore, patients with coronary artery disease who visit higher elevations
       might be unusually sensitive to the added effects of atmospheric CO.
            Before considering the combined effects of CO and atmospheric hypoxia, it is important
       to distinguish between the long term resident of high altitude, as compared with the newly
25     arrived visitor from low altitude.  Specifically, the visitor will be more hypoxemic than the
       fully adapted resident for the following reasons. Initially, the visitor will exhibit relative
       hypoventilation, particularly during sleep, since ventilatory adaptation requires several days.
       The result will be a lowering of arterial P02, a fall in arterial O2 saturation, and a reduction in
       arterial O2 content.  This hypoxemia will stimulate the sympathetic nervous system to increase
30     heart rate, myocardial contractility, and systemic arterial blood pressure (Grover et al., 1986).
       These factors combine to increase cardiac work, calling for an increase in coronary blood

       March 5, 1990                           12-14    DRAFT - DO NOT QUOTE OR CITE

-------
        flow.  In addition, the initial increase in ventilation will produce a respiratory alkalosis which,
        in turn, will increase the affinity of hemoglobin for oxygen and thereby interfere with oxygen
        release to the tissues.
             Over several days following arrival at high altitude, a number of mechanism will
 5      operate to lessen the initial impact of atmospheric hypoxia.  Ventilation will increase
        progressively, and this will evaluate arterial O2 tension, saturation, and content. A decrease
        in plasma volume increases hematocrit (hemoconcentration), with an associated increase in the
        O2-carrying capacity (hemoglobin concentration) of the blood; at this point, the polycythemia
        is only relative not absolute.  Nevertheless, this will increase further the arterial O2 content.
10      Although increased  sympathetic activity persists, cardiac beta receptor responsiveness
        decreases, mitigating the initial tachycardia.  This combined with a decreased in cardiac
        stroke volume leads to a return of cardiac output to normal  (or even subnormal) levels
        (Grover et al., 1986). Compensation for the initial respiratory alkalosis returns blood pH
        towards normal.  Concurrently, there is an increase in the concentration of 2,3-
15      disphosphoglycerate (2,3-DPG) within the red cells, the net effect being not only a return of
        hemoglobin-oxygen affinity to normal but actually to levels lower than prior to ascent. This
        facilitates the release of oxygen to the tissues, an effect that more than offsets the slight
        decrease in arterial O2 saturation.  For the heart, this is particularly important, for it removes
        the demands for increased coronary blood flow at moderate altitude (Grover et al., 1976).
20           For the long-term resident at high altitude, systemic blood pressure returns to (or below)
        values normal for sea level (Marticorena et al.,  1969).  Cardiac output remains at (or below)
        levels normal for sea level (Hartley et al., 1967).  Tissue capillary density increases, thereby
        enhancing oxygen delivery. Consequently, demands on the coronary circulation are not
        increased. An absolute polycythemia develops, i.e.,  total red cell mass plateaus at levels
25      greater than at sea level.  As a consequence, the normal turn-over of this greater mass of red
        cells results in an  increase in the endogenous production of CO (Johnson, 1968).
             Based on these considerations, the population subgroup at greatest risk from CO
        exposure  would be the newly-arrived transient visitors to high altitudes.  By binding
        hemoglobin, CO would reduce further arterial O2 content, i.e., increase hypoxemia. In
30      addition,  CO would augment the effect of alkalosis by increasing further the affinity of
        hemoglobin for oxygen,  thereby impairing O2 delivery even more.  Both factors would

        March 5,  1990                            12-15     DRAFT - DO NOT QUOTE OR CITE

-------
       increase demands for greater coronary blood flow.  These initial risks would decline
       progressively if the visitor remains long enough to complete physiological adaptation.  The
       period of increased risk is probably prolonged in the elderly, since adaptation to high altitude
       proceeds more slowly with  increasing age (Dill et al.,  1963,  1985; Robinson et al., 1973).
 5          Not surprisingly, the total number of transient visitors to high altitude far exceeds the
       resident population.  Ironically, it is these same transient visitors who contribute most to
       atmospheric pollution with  CO in the mountains (automobile engines not tuned to high
       altitude,  inefficient wood burning fireplaces used for social effect in vacation cabins, etc.). In
       addition, newly arrived visitors are often unaware of the physiological effects of high  altitude
10     (plus CO), and hence are prone to over-exertion which would increase the potential hazard.
       For a variety of reasons, COHb concentrations tend to be higher in high altitude residents
       than seen at low altitude (Johnson, 1968).
            One would postulate that the combination of high altitude with CO would pose the
       greatest risk to persons newly arrived at high altitude who have underlying cardiovascular
15     disease, particularly since they are usually older individuals. Surprisingly, this hypothesis has
       never been tested adequately.  In fact, there are virtually no data on how patients with known
       or suspected coronary artery disease  responde to a sojourn at moderately high  altitude (8,000
       to 12,000 ft or higher) with or without added CO exposure.  In  two pilot studies, the  risk
       from altitude alone at least appears to be minimal (Okin,  1970; Khanna et al.,  1976).  Among
20      148,000 persons (10% over 50 years of age) trekking  in Nepal to altitudes up  to 18,000 ft,
       there were no cardiac deaths and only three  helicopter evacuations for cardiac problems
       (Shlim and Houston,  1989).  Nevertheless, the need remains for a rigorous test of the
       hypothesis.
             If the cardiovascular  effects of atmospheric hypoxia at high altitude are augmented by
25     added exposure to CO, then patients already hypoxic from chronic obstructive lung disease
        should also be at increased risk from CO at altitude.  Paradoxically, this does not appear to
       be true,  again at least for brief exposure to altitude alone, even  though hypoxemia is
        exaggerated (Graham and Houston,  1978; Schvvart et al., 1984). This may reflect the
        decrease in air density at high altitude which reduces both the work of breathing (Thoden
30      et al., 1969) as well as the effective degree  of airway obstruction in such patients (Kryger
        etal., 1978).

        March 5, 1990                            12-16    DRAFT - DO NOT QUOTE OR CITE

-------
              Although limited observations do not indicate an increased risk from exposure to
        moderate altitude (without added CO) for patients with either cardiovascular or obstructive
        airway disease, this does not imply that prolonged residence at altitude is well tolerated.
        Individuals with these disorders, while successfully living at higher altitudes initially,  tend to
  5     leave these altitudes as they reach older age.  Outward migration of older individuals with
        these disorders has been described for the higher elevations in the state of Colorado
        (Regensteiner and Moore, 1985).   These elderly residents living at altitudes above 8,000 ft
        left primarily due to poor health.  Heart disease and lung disease (each 41%) accounted for
        the majority  of reasons for leaving their high altitude homes.
10           It is known that low birth weights occur in both infants born at altitudes above 6,000 ft
        as well as infants born near sea level whose mothers had elevated COHb levels due to
        cigarette smoking (see Chapter 11, Section 11.1).  It has also been shown that COHb  levels in
        smokers at high altitude are higher than in smokers at sea level (Brewer et al., 1970).  While
        it is probable that the combination of hypoxic hypoxia and hypoxia resulting from ambient
15     exposure to CO could further reduce birth weight at high altitude and possibly modify future
        development, no data are presently available to support this hypothesis.  A study conducted in
        Colorado (Alderman et al.,  1987) failed to find a strong  relationship between risk of low birth
        weight and maternal exposure to neighborhood CO  estimated from stationary monitors.  The
        combination, however, of maternal smoking and 6,000 ft altitude did result in lower birth
20     weights than those due to altitude  alone.

       12.6  REFERENCES

25
       Alderman, B. W.; Baron, A. E.; Savitz, D. A.  (1987) Maternal exposure to neighborhood carbon monoxide and
             risk of low infant birth weight. Public Health Rep. 102: 410-414.
       Allred, E. N.; Bleecker, E.  R.; Chaitman, B. R.; Dahms, T. E.; Gottlieb, S. O. Hackney, J. D.; Pagano, M.;
30          Selvester, R. H.; Walden,  S. M.; Warren, J. (1989a) Short-term effects of carbon monoxide exposure on
             the exercise performance of subjects with coronary artery disease. N. Engl. J. Med. 321: 1426-1432.
       Allred, E. N.; Bleecker, E.  R.; Chairman, B. R.; Dahms, T. E.; Gottlieb, S. O.; Hackney, J. D.; Hayes, D.;
             Pagano, M. Selvester,  R. H.;  Walden, S. M.; Warren, J. (1989b) Acute effects of carbon monoxide
35          exposure on individuals with coronary artery disease. Cambridge, MA: Health Effects Institute; research
             report no. 25.
       American Heart Association. (1988) 1988 heart facts. Dallas, TX: American Heart  Association.

        March 5, 1990                            12-17    DRAFT - DO NOT QUOTE OR CITE

-------
        Aronow, W. S.; Stemmer, E. A.; Isbell, M. W. (1974) Effect of carbon monoxide exposure on intermittent
               claudication. Circulation 49: 415-417.

        Aronow, W. S.; Ferlinz, J.; Glauser, F. (1977) Effect of carbon monoxide on exercise performance in chronic
 5             obstructive pulmonary disease. Am. J. Med. 63:  904-908.

        Brewer, G. J.; Eaton, J. W.; Weil, J. V.; Grover, R. F. (1970) Studies of red cell glycolysis and interactions
               with carbon monoxide, smoking, and altitude. In: Brewer, G. J., ed. Red cell metabolism and function:
               proceedings of the first international conference on red cell metabolism and function; October 1969; Ann
10             Arbor, MI. New York, NY:  Plenum Press; pp.  95-114. (Advances in experimental medicine and
               biology: v. 6).

        Brody, J. A.; Brock, D. B.; Williams, T. F. (1987) Trends in the health of the elderly population.  Annu. Rev.
               Public Health 8: 211-234.
15
        Busija, D. W.; Wagerle, L.  C.; Pourcyrous, M.; Leffler, C. W. (1987) Acetylcholine dramatically increases
               prostanoid synthesis in piglet parietal cortex. Brain Res. 439:  122-126.

        Calverley, P. M. A.; Leggett, R. J. E.; Flenley, D. C. (1981)  Carbon monoxide and exercise tolerance in
20             chronic bronchitis and emphysema.  Br. Med. J.  283: 878-880.

        Collins, J. G. (1986) Prevalence of selected chronic conditions: United States,  1979-81 Hyattsville, MD: U. S.
               Department of Health and Human Services, National Center for Health Statistics; DHHS publication no.
               (PHS) 86-1583. (Data from the national health survey, series 10, no. 155).
25
        Collins, J. G. (1988) Prevalence of selected chronic conditions, United States,  1983-85. Hyattsville, MD: U. S.
               Department of Health and Human Services, National Center for Health Statistics; DHHS publication no.
               (PHS) 88-1250. (Advance data from vital and health statistics no. 155).

30      Deanfield, J.; Wright,  C.; Rrikler, S.; Ribeiro, P.; Fox, K. (1984) Cigarette smoking and the treatment of
               angina with propranolol, atenolol, and nifedipine. N. Engl. J. Med. 310: 951-954.

        Dill, D. B.; Robinson,  S.; Balke, B. (1963) Respiratory  responses to exercise as related to age.  In: Cunningham,
               D. J. C.; Lloyd, B.  B., eds. The regulation of human respiration.  Oxford, United Kingdom:  Blackwell;
35             pp. 453-460.

        Dill, D. B.; Alexander, W.  C.; Myhre, L. G.; Whinnery,  J. E.; Tucker, D. M. (1985) Aerobic capacity of
               D. B. Dill, 1928-1984. Fed.  Proc. Fed. Am. Soc.  Exp. Biol. 44:  1013.

40      Doblar, D. D.; Santiago,  T. V.; Edelman, N. H. (1977) Correlation between ventilatory and cerebrovascular
               responses to inhalation of CO. J. Appl. Physiol.: Respir.  Environ.  Exercise Physiol. 43: 455-462.

        Fechter, L. D.; Thorne, P. R.; Nuttall, A. L. (1987) Effects of carbon monoxide on cochlear electrophysiology
               and blood flow. Hear. Res. 27:  37-45.
45
        Fechter, L. D.; Richard, C. L.; Mungekar, M.; Gomez, J.; Strathern, D.  (1989) Disruption of auditory function
               by acute administration of a "room odorizer" containing butyl nitrite in rats. Fundam. Appl. Toxicol.
               12: 56-61.

50      Gardiner, M.; Nilsson, B.; Rehncrona, S.; Siesjo, B. K. (1981) Free fatty acids in the rat brain in moderate and
               severe hypoxia. J. Neurochem. 36:  1500-1505.

        Golanki, D. L.; McCurdy, P. R.; Cuttitta, F. F.;  Schechter, G. P. (1988)  Hemolysis in sickle cell disease
               measured by endogenous CO production.  Am. J. Clin. Pathol.  89: 221-225.


         March 5,  1990                                12-18      DRAFT -  DO NOT QUOTE OR CITE

-------
        Graham, W. G. B.; Houston, C. S. (1978) Short-term adaptation to moderate altitude: patients with chronic
               obstructive pulmonary disease. JAMA J. Am. Med. Assoc. 240: 1491-1494.

        Gray, R. D, (1982) Kinetics and mechanism of carbon monoxide binding to purified liver microsomal cytochrome
 5            P-450 isozymes. J. Biol. Chem.  257: 1086-1094.

        Grover, R. F.; Lufschanowski,  R.; Alexander, J. K. (1976) Alterations in the coronary circulation of man
               following ascent to 3,100 m altitude.   J. Appl. Physiol. 41: 832-838.

10     Grover, R. F.; Weil, J. V.; Reeves, J. T. (1986) Cardiovascular adaptation to exercise at high altitude. In:
               Pandolf, K. B., ed. Exercise and sport sciences reviews: v. 14.  New York, NY: Macmillan; pp. 269-302.

        Hartley, L.  H.; Alexander, J. K.; Modelski,  M.; Grover, R. F. (1967)  Subnormal cardiac output at rest and
               during exercise at 3,100 m altitude. J. Appl. Physiol. 23: 839-848.
15
        Johnson, R. L., Jr.  (1968) Rate of red cell and hemoglobin destruction after descent from high altitude. Brooks
               Air Force Base, TX: USAF School of Aerospace Medicine; contract no. AF 41609-68-C-0032.

        Khanna, P.  K.; Dham, S. K.; Hoon, R. S. (1976) Exercise in an hypoxic environment as a screening  test for
20            ischaemic heart disease. Aviat.  Space Environ. Med. 47: 1114-1117.

        Knisely, J. S.; Rees, D. C.;  Balster, R. L. (1989) Effects of carbon monoxide in combination with behaviorally
               active drugs on fixed-ratio performance in the mouse. Neurotoxicol. Teratol.  11: 447-452.

25     Ko, B. H.; Eisenberg,  R. S. (1987) Prolonged carboxyhemoglobin clearance in a patient with Waldenstrom's
               macroglobulinemia. Am. J. Emerg. Med. 5:  503-508.

        Koehler, R. C.; Jones, M. D., Jr.; Traystman, R. J. (1982) Cerebral circulatory response to carbon monoxide
               and hypoxic hypoxia in the lamb.  Am. J. Physiol. 243: H27-H32.
30
        Kryger, M.; Aldrich, F.; Reeves, J. T.; Grover, R. F.  (1978) Diagnosis of airflow obstruction at high altitude.
               Am. Rev. Respir. Dis. 117:  1055-1058.

        Kukreja, R. C.; Kontos, H. A.; Hess,  M. L.; Ellis, E. F. (1986) PGH  synthase and lipoxygenase generate
35            superoxide in the presence of NADH or NADPH. Circ. Res. 59: 612-619.

        Leffler, C. W.; Busija, D. W. (1987) Arachidonic acid metabolites and  perinatal hemodynamics. Semin.
               Perinatal. 11:  31-42.

40     Marticorena, E.; Ruiz, L.; Severino, J.; Galvez, J.; Penaloza, D. (1969) Systemic blood pressure in white men
               born at sea level:  changes after long residence at high altitudes.  Am. J. Cardiol. 23: 364-368.

        McMillan, D. E.; Miller, A. T., Jr. (1974) Interactions between carbon monoxide and a-amphetamine or
               pentobarbital on schedule-controlled behavior. Environ. Res. 8:  53-63.
45
        Mitchell, D. S.; Packham, S. C.; Fitzgerald,  W. E. (1978) Effects of ethanol and carbon monoxide on two
               measures of behavioral incapacitation of rats. Proc. West. Pharmacol. Soc. 21: 427-431.

        National Institute on Alcohol Abuse and Alcoholism. (1987)  Sixth special report to the U.S.  Congress on alcohol
50            and health from the Secretary of Health and Human Services. Rockville, MD:  U.  S. Department of
               Health and Human Services, Alcohol, Drug Abuse, and Mental  Health Administration; DHHS publication
               no.  (ADM)87-1519.
         March 5,  1990                                12-19     DRAFT - DO NOT QUOTE OR CITE

-------
       Okin, J. T. (1970) Response of patients with coronary heart disease to exercise at varying altitude. Adv. Cardiol.
              5: 92-96.

       Rail, T. W.  (1980) Central nervous system stimulants: the xanthines. In:  Oilman, A. G.; Goodman, L. S.;
 5            Oilman, A., eds. The pharmacological basis of therapeutics. 6th ed. New York, NY: Macmillan
              Publiching Co., Inc.; pp. 592-607.

       Regensteiner, J. G.; Moore, L. G. (1985) Migration of the elderly from high altitudes in Colorado. JAMA J.
              Am. Med. Assoc. 253: 3124-3128.
10
       Robinson, S.; Dill, D. B.; Ross, J. C.; Robinson, R. D. (1973) Training and physiological aging in man. Fed.
              Proc. Fed. Am. Soc. Exp.  Biol.  32:  1628-1634.

       Rockwell, T. J.; Weir, F. W. (1975) The interactive effects of carbon monoxide and alcohol on driving skills.
15            Columbus, OH: The Ohio  State University Research Foundation; CRC-APRAC project CAPM-9-69.
              Available  from: NTIS, Springfield, VA; PB-242266.

       Schwartz, J. S.; Bencowitz, H. Z.; Moser, K. M. (1984) Air travel hypoxemia with chronic obstructive
              pulmonary disease. Ann. Int. Med. 100:  473-477.
20
       Shlim, D. R.; Houston, R. (1989)  Helicopter rescues and deaths among trekkers in Nepal. JAMA J. Am. Med.
              Assoc. 261:  1017-1019.

       Solanki, D.  L.; McCurdy, P. R.; Cuttitta, F. F. (1988) Hemolysis in sickle cell disease as measured by
25            endogenous carbon monoxide production: a preliminary report. Am. J. Clin. Pathol. 89: 221-225.

       Thoden, J. S.; Dempsey, J. A.; Reddan, W.  G.;  Bimbaum, M. L.; Forster, H. V.; Grover, R. F.; Rankin, J.
              (1969) Ventilatory work during steady-state exercise. Fed. Proc. Fed. Am. Soc. Exp. Biol.
              28:  1316-1321.
30
       U. S. Department of Health and Human Services. (1987) Current estimates from the national health interview
              survey. Series 10: no. 166. Data from the National Health Interview Survey. Hyattsville,  MD: Public
              Health Service, National Center for Health Statistics. Available from: GPO, Washington, DC; GPO
              stock number 017-022-01052-6.
35
       U. S. Environmental Protection Agency. (1979) Air quality criteria for carbon monoxide. Research Triangle
              Park, NC: Office of Health and Environmental Assessment, Environmental Criteria and Assessment
              Office;  EPA report no. EPA-600/8-79-022. Available from: NTIS, Springfield, VA; PBS 1-244840.

40    U. S. Environmental Protection Agency. (1984) Revised evaluation of health effects associated with carbon
              monoxide exposure: an addendum to the 1979 EPA air quality criteria document for carbon monoxide.
              Research  Triangle Park, NC: Office of Health and Environmental Assessment, Environmental Criteria
              and Assessment Office; EPA report no. EPA-600/9-83-033F. Available from:  NTIS, Springfield, VA;
              PB85-103471.
45
       Wei, E. P.; Kontos, H. A.; Dietrich, W. D.; Povlishock, J. T.; Ellis, E. T. (1981) Inhibition by free radical
              scavengers and by cyclooxygenase inhibitors of pial arteriolar abnormalities from concussive brain injury
              in cats. Circ. Res. 48: 95-103.

50    Wei, E. P.; Kontos, H. A.; Christman, C. W.; DeWitt,  D. S.; Povlishock, J. T. (1985) Superoxide generation
              and reversal of acetylcholine-induced cerebral arteriolar dilation after acute hypertension.  Circ. Res.
               57:  781-787.
         March 5, 1990                               12-20     DRAFT - DO NOT QUOTE OR CITE

-------
Zinkham, W. H.; Houtchens, R. A.; Caughey, W. S. (1980) Carboxyhemoglobin levels in an unstable
      hemoglobin disorder (Hb Zuerich): effect on phenotypic expression. Science (Washington, DC)
      209: 406-408.
March 5, 1990                             12-21     DRAFT - DO NOT QUOTE OR CITE

-------
               APPENDIX A:  GLOSSARY OF TERMS AND SYMBOLS
Abbreviations and Acronyms
X
12C16O
12C18O
.31J
2,3-DPG
51Cr
7-mode
85Kr
9%TcDTPA
A/F
a.k.a.
A-aDO2
ACD
ACGIH
ADD (m)
AEP
AIRS
Alt
amb
ANOVA
ANSI
Ar
atm
BEI
BF
BUS
BS
BTPS
Btu
C
Ca
CAA
CAD
CALINE3  Model
CASAC
Q
CBF
cc
CL
Atmospheric lifetime
Chi
Carbon monoxide containing oxygen isotope 16
Carbon monoxide containing oxygen isotope 18
Iodine-131
2,3-diphosphoglycerate
Chromium-51
137 second driving cycle test
Krypton-85
Radiolabeled diethylene triamine pentacetic acid
Air-to-fuel ratio
Also known as
Alveolar-arterial O2 gradient
Acid citrate dextrose
American Conference of Governmental Industrial Hygienists
Additive constant specification
Auditory evoked potential
Aerometric Information Retrieval System (U.S. EPA)
Altitude above sea level
Ambient
Analysis of variance
American National Standards Institute
Argon
Atmosphere
Biological exposure index
Blue flame (heater)
Bibliographic literature information system
Body sway
Body temperature, barometric pressure, and  saturated
British  thermal unit
Celsius
Calcium
Clean Air Act
Coronary artery disease
A form of dispersion modeling
Clean Air Scientific Advisory Committee
Concentration of carbon monoxide for a bulk mixture
Cerebral blood flow
Cubic centimeter *(see also cm3)
Concentration of carbon monoxide in the sample
March 12, 1990
                 A-1     DRAFT - DO NOT QUOTE OR CITE

-------
CEQ
CFF
Cff
CFK
CFKE
cGMP
CH2C12
GH3CC13
CH4
CHD
CI
Q(t)

CID
cm
cm3
CMRO2
CN
CNS
CNV
CO
CO
CO-Ox
CO2
COM
COHb
COMb
COPD
CP
CVD
CVS-72
CVS-75
Cyt
d1
dBA
dF/dtmax
DLCO
DA
dP/dt
DpCO
DPG
EDRF
EDTA
EEC
EKG
EP
EPA
Presidents Council on Environmental Quality
Critical flicker frequency
Critical flicker fusion
Coburn-Forster-Kane
Coburn-Forster-Kane equation
Cyclic guanosine monophosphate
Methylene chloride
Methylchloroform
Methane
Coronary heart disease
Confidence interval
The air pollutant concentration to which an individual is exposed at any
   point in time t.
Cubic inch displacement
Centimeter(s)
Cubic centimeter
Cerebral O2 consumption
Cyanide
Central nervous  system
Contingent negative variation (slow-evoked potential)
Carbon monoxide
Cardiac output
CO-oximeter
Carbon dioxide
Carbon monoxide hypoxia
Carboxyhemoglobin
Carboxymyoglobin
Chronic obstructive pulmonary disease
Capillary permeability to protein
Cardiovascular disease
Constant volume sample cold start test
Constant volume sample test including cold and hot starts
Cytochrome
Measure of detection threshold
Decibels (A-scale)
Derivative of maximal force
Diffusing capacity for CO, mL min'1 ton"1 (STPD)
Diffusing capacity for O2, mL min'1  (STPD)
Derivative of pressure with time
Carbon monoxide diffusion coefficient across the placenta
Diphosphoglycerides
Endothelium-derived relaxing factor
Ethylenediaminetetraacetic acid
Electroencephalogram
Electrocardiogram (also ECG)
Evoked potential
Environmental Protection Agency
March 12, 1990
                 A-2     DRAFT - DO NOT QUOTE OR CITE

-------
ERG
ETS
f
fB
Fco
FDA
FEF
FEV,
FFF
F.CO
FID
pio2
FMVCP
FRC
FVC
g
GC
GD
GFC
GMP
h
H20
HANES
Hb
Hb5
HbA
HBO
HbO2
HCN
HCs
Hct
HDL
He
HEI
Hg
HgO
HH
HO2*
HR
HW/BW
H*
I205

K
K2HPO4
Electroretinogram
Environmental tobacco smoke
Fetal
Flow rate of carbon monoxide for a bulk mixture
Air flow rate
Breathing frequency
Carbon monoxide flow rate
Food and Drug Administration
Forced expiratory flow
Forced expiratory volume (at one minute)
Flicker-fusion frequency
Volumetric fractional concentration of CO in dry inspired air, ppm
Flame ionization detector
Fraction of inspired O2
Federal Motor Vehicle Control Program
Functional residual capacity
Forced vital capacity
Gram(s)
Gas chromatograph
Gestation day
Gas filter correlation
Guanosine monophosphate
Hour
Water
Health and Nutrition Examination Survey
Hemoglobin concentration in blood, g dL"1
Abnormal hemoglobin found in individuals with sickle-cell disease
Normal hemoglobin
Hyperbaric oxygen
Oxyhemoglobin
Hydrogen cyanide
Hydrocarbons
Hematocrit
High-density lipoprotein
Helium
Health Effects Institute
Mercury
Mercuric oxide
Hypoxic hypoxia
Hydroperoxyl radical
Heart rate
Heart weight to body weight ratio
Hydrogen atom
Iodine pentoxide
Intraperitoneal
Warburg partition coefficient
Mono-hydrogen potassium arthrophosphate
March 12, 1990
                 A-3      DRAFT - DO NOT QUOTE OR CITE

-------
K4Fe(CN)6
kg
kJ
km
K,
KPM
L
LC*
LDH
LDL
LDV
LOEL
LPG
LV
MLDH
M
m
m3
Mb
MEA
MEMs
metHb
METS
MFO
mg
mi
MI
mL
MMFR
mo
mol
MRFIT
MSA
MSHA
MST
MULT (m)
 n
N2
N2O
NAAQS
NADPH
NaHCO3
NAMS
NASA
Potassium ferrpcyanide
The effective reaction rate constant
Kilogram(s)
Kilo Joule, IxlO10 ergs, 0.948 Btu
Kilometer
Michaelis-Menten constant
Kilopondmeters per minute
Liter(s)
Concentration that is lethal to 50% of test subjects (used in inhalation
   studies)
Dose that is lethal to 50% of test subjects
Lactate dehydrogenase
Low-density lipoprotein
Light-duty vehicle
Lowest-observed-effect level
Liquefied petroleum gas
Left ventricle
Myocardial lactate dehydrogenase
Haldane constant
Maternal
Cubic meter
Myoglobin
Mean electrical axis
Microenvironmental monitors
Methemoglobin
Basal  metabolic equivalent
Mixed-function oxidase
Milligram(s)
Mile
Myocardial infarction
Milliliter(s)
Maximum mid-expiratory flow rate
Month(s)
Mole
Multiple risk factor intervention trial
Metropolitan Statistical Area
Mine  Safety and Health Administration
Mean survival time
Multiplicative constant specification
Number
Nitrogen
Nitrous oxide
National Ambient Air Quality Standards
Reduced nicotinamide adenine dinucleotide phosphate
Acid sodium carbonate
National Air Monitoring Stations
National Aeronautics and Space Administration
March 12, 1990
                 A-4     DRAFT - DO NOT QUOTE OR CITE

-------
NBS
ND
NDIR
NEDS
NEM
NG
NIOSH
NIST
nm
NO
NO2
NO3
NOEL
NOX
NR
02
O2Mb
03
OH*
P
P
P*,
P.C02
PAC02
PAH
PAN
PA
PB
PCN
"co
PA
PCO2
PD
PEMs
PfCO
PGI2
P,CO
PiT
PmCO
PMN
P02
ppbv
ppm
ppmm
PR
PW
Q
National Bureau of Standards, now NIST
Not determined
Nondispersive infrared
National Emissions Data System
NAAQS Exposure Model
Natural gas
National Institute for Occupational Safety and Health
National Institute of Standards and Technology
Nanometer
Nitric oxide
Nitrogen dioxide
Nitrate
No-observed-effect level
Nitrogen oxides
No response
Oxygen
Oxymyoglobin
Ozone
Hydroxyl radical
Pressure in atmospheres
Propane
Partial pressure of O2 at 50% saturation of hemoglobin
Partial pressure of CO2 in arterial blood, torr
Partial pressure of CO2 in alveolar gas, torr
Polyaromatic hydrocarbons
Peroxyacetyl nitrate
Partial pressure of O2 in arterial blood
Barometric pressure, torr
Potassium cyanide
Partial pressure of CO
Mean partial pressure of pulmonary capillary 02 (torr)
Partial pressure of CO2
Postnatal day
Personal exposure monitors
Partial pressures of carbon monoxide in the fetal placental capillaries
Prostacyclin
Partial pressure of CO in humidified inspired air, ton-
Pituitary
Partial pressures of CO in the maternal placental
Polymorphonuclear neutrophil leukocytes
Partial pressure of O2
Parts per billion by volume
Parts per million by volume (milligrams per liter)
Parts per million by mass (milligrams per kilogram)
Pulmonary resistance
Placental weight
Overall perfusion
March 12, 1990
                  A-5     DRAFT - DO NOT QUOTE OR CITE

-------
r
rCBF
R
R2
RBC
RV
RV
RVF
s
SAROAD
scf
SCO
SD
SE
SF,
SHAPE
SHED
SI
SIDS
SIPs
SLAMS
S02
SO2
SP
SR
SRMs
ST
STPD
SV
t
t1
td
TEAM
TEM
Tg
TH
THC
t,
TLC
torr
TSP
TTS
TV
TWA
UV
VD
VEP
Correlation coefficient
Regional cerebral blood flow
CO/O2 at 50% inhibition
Coefficient of determination
Red blood cell
Right ventricle
Residual volume
Red visual field
Second
U.S. EPA centralized data base; superceded by AIRS (q.v.)
Standard cubic foot
Percent COHb of total Hb
Standard deviation
Standard error
Sulfur hexafluoride
Simulation of Human Activity and Pollutant Exposure
Sealed housing for evaporative determination
Stroke index
Sudden infant death syndrome
State Implementation Plans
State and Local Air Monitoring Stations
%O2Hb of total Hb
Sulfur dioxide
Mean stroke power
Systemic resistance
Standard Reference Materials
Segment of the EKG (see definition of electrocardiogram)
Standard temperature and pressure, dry
Stroke volume
Time, minute
Postexposure time in minutes
Times-to-death
Total exposure assessment methodology
Transmission electron microscopy
Teragram(s); 1012 grams;  106 metric tons
Total hydrocarbon
Total hydrocarbon content
Times-to-incapacitation
Total lung capacity
A unit of pressure equal to 1/760 of an atmosphere
Total suspended particulates
Temporary threshold shifts
Tidal volume (also VT)
Time-weighted average
Ultraviolet
Physiological deadspace
Visual evoked potential
March 12, 1990
                  A-6     DRAFT - DO NOT QUOTE OR CITE

-------
 VFT
 VLDL
 " max
 VMT
 VPD
 YT
 y
 VA
 vco
 yo2
 V02max
 w/
 WBC
 WF
 WHW
 x
 [COHb]

 [C0]ave
 [OH']ave
 2
 A
Ventricular fibrillation threshold
Very low-density lipoprotein
Ventricular contractility
Vehicle-miles traveled
Vehicles per day
Tidal volume
Ventilation
Alveolar ventilation, mL min'1 (STPD)
Rate of endogenous production of CO, mL min'1 (STPD)
Minute ventilation; expired volume per minute
Oxygen uptake by the body
Oxygen consumption of tissues or cells (also,
Maximal oxygen uptake
With
White blood cell
White flame (heater)
Wet heart  weight
Mean
Concentration of COHb in blood, as milliliters of CO per milliliter of
   blood (STPD)
Average concentration of CO
Average concentration of the hydroxyl radical
Sigma (sum of terms)
Angstrom
Micrometer(s)
Definitions
Acclimatization:  The physiological and behavioral adjustments of an organism to changes in its
      environment.

Adaptation:  Changes in an organism's structure or habit that help it adjust to its surroundings.

Additivity:  A pharmacologic or toxicologic interaction in which the combined effect of two or
      more chemicals is approximately equal to the sum of the effect of each chemical  alone.
      (Compare with: antagonism,  synergism.)

Adiabatic warming:   The temperature increase produced in a descending  air mass as  pressure
      increases with decreasing altitude.
 March 12, 1990
                  A-7     DRAFT - DO NOT QUOTE OR CITE

-------
Air Pollutant:  Any substance in air that could, if in high enough concentration, harm humans,
      other animals, vegetation, or material.   Pollutants may include almost any natural or
      artificial composition of matter capable of being airborne.  They may be in the form of
      solid particles, liquid droplets, gases, or in combinations of these forms. Generally, they
      fall into two main groups: (1) those emitted directly from identifiable sources and (2) those
      produced in the air by interaction between two or more primary pollutants, or by reaction
      with normal atmospheric constituents, with or without photoactivation. Exclusive of pollen,
      fog, and dust, which are of natural origin, about 100 contaminants have been identified and
      fall into the following categories:  solids, sulfur compounds, volatile organic chemicals,
      nitrogen compounds, oxygen compounds, halogen compounds, radioactive compounds, and
      odors.

Air Pollution:  The presence of contaminant or pollutant substances in the air that do not disperse
      properly and  interfere with human health or  welfare,  or produce other harmful
      environmental effects.

Air Pollution Episode:  A period of abnormally high concentration of air pollutants, often due
      to low winds and temperature inversion,  that can cause illness and death.

Air Quality Criteria:  The levels of pollution and lengths of exposure above which adverse health
      and welfare effects may occur.

Air Quality Standards:  The level of pollutants prescribed by regulations that may not be exceeded
      during a specified time in a defined area.

Alveolar-arterial oxygen pressure  difference [P(A-a)OJ:   The difference in partial pressure of
      oxygen in the alveolar gas spaces and that in the  systemic arterial blood, measured in torr.

Alveolar carbon dioxide pressure (PACO2):  Partial pressure of carbon dioxide in the air contained
      in the lung alveoli.
Alveolar oxygen partial pressure (Ptf)J:  Partial pressure of oxygen in the air contained in the
      alveoli of the lungs.

Alveolus: Hexagonal or spherical air cells of the lungs.  The majority of alveoli arise from the
      alveolar ducts which are lined with the alveoli.  An alveolus is an ultimate respiratory unit
      where the gas exchange takes place.

Ambient Air:  Any unconfined portion of the atmosphere:  open air, surrounding air.

Ambient Air Quality  Standards:  (See:  Criteria Pollutants and National Ambient Air Quality
      Standards).

Anatomical  dead space (VDHWl):  Volume of the conducting airways down to the level  where,
      during air breathing, gas exchange with blood can occur, a region probably situated at the
      entrance of the alveolar ducts.
 March 12, 1990                           A-8      DRAFT - DO NOT QUOTE OR CITE

-------
Antagonism:  A pharmacologic or toxicologic interaction in which the combined effect of two
      chemicals is less than the sum of the effect of each chemical alone; the chemicals either
      interfere with each other's actions, or one interferes with the action of the other. (Compare
      with: additivity, synergism).

Arrhythmia: Any variation from the normal rhythm of the heartbeat.

Arterial oxygen saturation (SaOj):  Percent saturation of dissolved oxygen in arterial blood.

Arterial partial pressure of carbon dioxide (PaCOj): Partial pressure of dissolved carbon dioxide
      in arterial blood.

Arterial partial pressure of oxygen (PaOz): Partial pressure of dissolved oxygen in arterial blood.

Atmosphere [atm\:  A standard unit of pressure representing the pressure exerted by a 29.92-in
      column of mercury at sea level at 45°  latitude and equal to 1000 g/cm2. The whole mass
      of air surrounding the Earth, composed largely of oxygen and nitrogen.

ATPS condition (ATPS):  Ambient temperature and pressure, saturated with water vapor.  These
      are the conditions existing in a water spirometer.

BTPS conditions (BTPS):   Body temperature, barometric pressure, and saturated with water
      vapor.  These are the conditions existing in the  gas phase of the lungs. For humans the
      normal temperature is taken as 37°C, the pressure as the barometric pressure, and the
      partial pressure of water vapor as 47 torr.

Carbon dioxide (COJ:  A colorless, odorless, non-poisonous gas, which results from fossil fuel
      combustion and is normally a part of the ambient air.

Carbon dioxide production (VCOj): Rate of carbon dioxide production by organisms, tissues, or
      cells. Common units: mL CO2 (STPD)/kg.min.

Carbon monoxide (CO):  An odorless, colorless, toxic gas formed by incomplete combustion,
      with a strong affinity for hemoglobin and cytochrome; it reduces oxygen absorption
      capacity, transport, and utilization.

Carboxyhemoglobin (COHb): Hemoglobin in which the iron is associated with carbon monoxide.
      The affinity of hemoglobin for carbon monoxide  is about 240 to 250 times greater than for
      oxygen.

Central nervous system (CNS):  The portion of the nervous system that includes the brain and
      spinal cord, and their connecting nerves.
 March 12, 1990                          A-9     DRAFT - DO NOT QUOTE OR CITE

-------
Chronic obstructive lung disease (COLD):  This term refers to diseases of uncertain etiology
      characterized by persistent slowing of airflow during forced expiration. It is recommended
      that a more specific term, such as chronic obstructive bronchitis or  chronic obstructive
      emphysema, be used whenever possible. Synonymous with chronic obstructive pulmonary
      disease (COPD).

Combustion:  Burning, or rapid oxidation, accompanied by release of energy in the form of heat
      and light.  A basic cause of air pollution.

Combustion product: Substance produced during the burning or oxidation of a material.

Criteria: Descriptive factors taken into account by EPA in setting standards for various pollutants.
      These factors are used to  determine limits on allowable concentration levels, and to limit
      the number of violations per year.  When issued by EPA, the criteria provide guidance to
      the states on how to establish their standards.

Criteria pollutants:  The 1970 amendments to the Clean Air Act required EPA  to set National
      Ambient Air Quality Standards for certain pollutants known to be hazardous to human
      health.  EPA has identified and set standards to protect human health and  welfare for six
      pollutants:  ozone, carbon monoxide,  total suspended particulates, sulfur dioxide,  lead,
      and nitrogen oxide.  The term "criteria pollutants" derives from the requirement that EPA
      must describe the characteristics and potential health and welfare effects of these pollutants.
      It is on the basis of these  criteria that standards are set or revised.
Diffusing  capacity of the  lung  (DL, DLO2,  DLCO2, DLCO):   Amount of gas (O2,  CO,
      commonly expressed as mL gas (STPD) diffusing between alveolar gas and pulmonary
      capillary blood per  torr mean gas pressure difference per minute, that is, mL 02/(min-
      torr). Synonymous  with transfer factor and diffusion factor.

Dose-response relationship:   A  relationship between  (1)  the dose,  often actually based  on
      "administered dose" (i.e., exposure) rather than absorbed dose, and (2) the extent of toxic
      injury produced by  that chemical.  Response can be expressed either as the severity of
      injury or proportion of exposed subjects affected.

Electrocardiogram (ECG, EKG):  A graphic tracing of the variations in electrical potential caused
      by the excitation of the heart muscle and  detected at  the body surface.   The normal
      electrocardiogram shows deflections resulting from arterial and ventricular activity. The
      first deflection, P,  is due to  excitation of the atria. The QRS  deflections are due to
      excitation (depolarization) of the ventricles.  The T wave is due to recovery of the ventricles
      (repolarization). The U wave is a potential undulation  of unknown origin immediately
      following the T wave, seen in normal electrocardiograms and accentuated in hypokalemia.

Emission:  Pollution discharged into the atmosphere from smokestacks, other vents, and surface
      areas of commercial or industrial facilities; from residential chimneys; and from  motor
      vehicle, locomotive, or aircraft exhausts.
 March 12, 1990                          A-10     DRAFT - DO NOT QUOTE OR CITE

-------
Emission factor:  The relationship between the amount of pollution produced and the amount of
      raw material processed.  For example, an emission factor for a blast furnace making iron
      would be the number of pounds of particulates per ton of raw materials.

Emission inventory:  A list,  by source,  of the amount of air pollutants discharged into the
      atmosphere of a community.  It is used to establish emission standards.

Emission standard:  The maximum amount of air polluting discharge legally allowed from a
      single source, mobile or stationary.

Environment: The sum of all external conditions affecting the life, development, and survival of
      an organism.

EPA: The U.S. Environmental Protection Agency; established in 1970 by Presidential Executive
      Order, bringing together parts of various government agencies involved with the control of
      pollution.

Episode (pollution):  An air pollution incident in a given area caused  by a concentration of
      atmospheric pollution reacting with meteorological conditions that may result in a significant
      increase in illnesses or deaths. Although most commonly used in relation to air pollution,
      the term also may be used in connection with other kinds of environmental events such as
      a massive water pollution situation.

Exceedance: Violation of environmental protection standards by exceeding allowable  limits or
      concentration levels.

Exposure:  The amount of radiation  or pollutant present in an environment that represents a
      potential health threat to the living organisms in the environment.

Fetus:  The post-embryonic stage of the  developing young.  In humans, from the end of the
      second month of pregnancy up to birth.

Forced expiratory flow  (FEFx): Related to some portion of the FVC curve. Modifiers refer to
      the amount of the FVC already exhaled  when the measurement is made.

Forced expiratory volume (FEV):  Denotes the volume of gas that is exhaled in a given  time
      interval during the execution of a forced vital capacity.   Conventionally, the times  used
      are 0.5, 0.75, or  1 s, symbolized FEV0.3, FEV0.7j, FEV,.0. These values often are expressed
      as a percent of the forced vital capacity, e.g. (FEVLO/VC) X 100.

Forced vital capacity (FVC): Vital capacity performed with a maximally forced expiratory effort.

Hematocrit (Hct):  The percentage of the volume of red blood cells in whole blood.
 March 12, 1990                           A-ll     DRAFT - DO NOT QUOTE OR CITE

-------
Hemoglobin (Hb):  A hemoprotein naturally occurring in most vertebrate blood, consisting of
      four polypeptide chains (the globulin) to each of which there is attached a heme group.
      The heme is made of four pyrrole rings and a divalent iron (Fe2+-protoporphyrin) which
      combines reversibly with  molecular oxygen.

Hydrocarbons (HC): Chemical compounds that consist entirely of carbon and hydrogen.

Hypoxemia:  A state in which the oxygen pressure and/or concentration in arterial and/or venous
      blood is lower than its normal value at sea level.  Normal oxygen pressures at sea level are
      85 to 100 torr in arterial blood and 37 to 44 torr  in mixed venous blood.  In adult humans
      the normal oxygen concentration is 17 to 23 mL (yiOO mL arterial blood; in mixed venous
      blood at rest it is 13 to 18 mL O2/100 mL blood.

Hypoxia:  Any state in which the oxygen in the lung,  blood, and/or tissues is abnormally low
      compared with that of a normal resting human breathing air at sea level.  If the Pm is low
      in the environment, whether because  of decreased  barometric  pressure or  decreased
      fractional concentration of oxygen,  the  condition  is termed environmental hypoxia.
      Hypoxia when referring to the blood is termed hypoxemia.  Tissues are said to be hypoxic
      when their fm is low, even if there is no arterial hypoxemia, as in "stagnant hypoxia" which
      occurs when the local circulation is low compared to the local metabolism.

In vitro:   1. "In glass"; a test-tube culture. 2. Any laboratory test using living cells taken from
      an organism.

In vivo:  In the living body of a plant or animal. In vivo tests are those laboratory experiments
      carried out on whole animals or human volunteers.

Indoor air: The breathing air inside a habitable structure or conveyance.

Indoor air pollution:  Chemical, physical,  or biological contaminants in indoor air.

Inversion: An atmospheric condition caused by a layer of warm air preventing the rise of cooling
      air trapped beneath it. This prevents the rise of pollutants that might otherwise be dispersed
      and can cause an air pollution episode.

Isotope: A variation of an element that has the same atomic number but a different weight because
      of its neutrons.   Various isotopes  of the same  element may have different radioactive
      behaviors.

Lapse rate:  Vertical temperature gradient in the atmosphere; usually negative  (i.e.,  decreasing
      with altitude) (see "inversion").

Lowest-observed-adverse-effect level (LOAEL): The lowest dose or exposure level of a chemical
      in a study at which there is a statistically or biologically significant increase in the frequency
      or severity of an adverse effect in the exposed population as compared with an appropriate,
      unexposed control group.
 March 12,  1990                          A-12     DRAFT - DO NOT QUOTE OR CITE

-------
Lowest-observed effect level (LOEL):  In a study, the lowest dose or exposure level at which a
      statistically or biologically significant effect is observed in the exposed population compared
      with an appropriate unexposed control group.

Lung volume (VJ: Actual volume of the lung, including the volume of the conducting airways.

Maximal aerobic capacity (max VOJ:  The rate of oxygen uptake by the body during repetitive
      maximal respiratory effort.  Synonymous with maximal oxygen consumption.

Methemoglobin (MetHb):  Hemoglobin in which iron is in the ferric state.  Because the iron is
      oxidized, methemoglobin is incapable of oxygen transport.  Methemoglobins are formed
      by various drugs and occur under pathological conditions.  Many methods for hemoglobin
      measurements utilize methemoglobin (chlorhemiglobin, cyanhemiglobin).

Minute ventilation (VE):  Volume of air breathed in one minute.  It is a product of tidal volume
      (VT) and breathing frequency (fB).  (See ventilation)

Minute volume:  Synonymous with minute ventilation.

Modeling:  An investigative technique using a mathematical or physical representation of a system
      or theory that accounts for all or  some its known properties. Models are often used to test
      the effect of changes of system components on the overall performance of the system.

Monitoring: Periodic or continuous surveillance or testing to determine the level of compliance
      with statutory requirements and/or pollutant levels in various media or in humans, animals,
      and other living things.

National Ambient Air  Quality Standards  (NAAQS):  Air quality standards established by  EPA
      that apply to outside air throughout the country.  (See criteria pollutants)

Nitric oxide (NO): A gas formed by combustion under high temperature and  high pressure in
      an internal combustion engine.  It changes into nitrogen dioxide in the ambient air and
      contributes to photochemical smog.

Nitrogen dioxide (NOJ:  The result of nitric  oxide combining with oxygen in the atmosphere.
      A major component of photochemical smog.

Nitrogen oxides (NOJ: Compounds of nitrogen and oxygen in ambient air;  that is, nitric oxide
      (NO) and others with a higher oxidation state of nitrogen, of which nitrogen dioxide is the
      most important lexicologically.

No-observed-adverse-effect level (NOAEL):  The highest experimental dose at which there is no
      statistically or biologically significant increases in frequency or severity of adverse health
      effects, as seen in  the exposed population compared with an appropriate, unexposed
      population.   Effects may be produced at this level,  but  they are not considered to be
      adverse.
 March 12, 1990                          A-13     DRAFT - DO NOT QUOTE OR CITE

-------
No-observed-effect level (NOEL): The highest experimental dose at which there is no statistically
      or biologically significant increases in frequency or severity of toxic effects seen in the
      exposed compared with an appropriate, unexposed population.
Oxygen consumption  (VO2,  QOz):   Rate of oxygen uptake of organisms,  tissues, or  cells.
      Common units:  mL O2 (STPD)/(kg»min) or mL O2 (STPD)/(kg«hr).  For whole organisms
      the oxygen consumption commonly is expressed per unit surface area or some power of the
      body weight. For tissue samples or isolated cells Q02 = fj.L O2/h/mg dry weight.
Oxygen saturation (SOz):  The amount of oxygen combined with hemoglobin, expressed as a
      percentage of the oxygen capacity of that hemoglobin. In arterial blood, SaO2.

Oxygen uptake (VOj): Amount of oxygen taken up by the body from the environment, by the
      blood from the alveolar gas, or by an organ or tissue  from the blood.  When this amount
      of oxygen is expressed per unit of time one deals with an "oxygen uptake rate. "  "Oxygen
      consumption" refers more specifically to the oxygen uptake rate by all tissues of the body
      and is equal to the oxygen uptake rate of the organism only when the oxygen stores are
      constant.

Ozone (O3): Found in two layers of the atmosphere,  the stratosphere and the troposphere. In
      the stratosphere (the atmospheric layer beginning 7 to 10 miles above the Earth's surface)
      ozone is a form of oxygen found naturally which provides a  protective layer shielding the
      earth from ultraviolet radiation's harmful health effects on humans and the environment.
      In the troposphere (the layer extending up 7 to 10 miles from the Earth's  surface), ozone
      is a chemical oxidant and major component of photochemical smog. Ozone can seriously
      affect the human respiratory system and is one of the most prevalent and widespread of all
      the criteria pollutants for which the Clean Air Act required EPA to set standards.  Ozone
      in troposphere is produced through complex chemical reactions of nitrogen oxides, which
      are among the primary pollutants emitted by combustion sources; hydrocarbons, released
      into  the  atmosphere through the combustion, handling and  processing of  petroleum
      products; and sunlight.

Peroxyacetyl nitrate  (PAN):   Pollutant  created by action  of UV component of sunlight on
      hydrocarbons and  nitrogen  oxides in the air;  an ingredient of photochemical smog.

pH:  A measure of the acidity or alkalinity of a liquid  or solid material.

Photochemical smog: Air pollution  caused by chemical reactions  of various pollutants emitted
      from different sources.

Physiological dead space (VD): Calculated volume that accounts for the difference between the
      pressures of carbon dioxide in expired and alveolar gas (or arterial blood). Physiological
      dead space reflects the combination of anatomical dead space and alveolar dead space, the
      volume  of the latter  increasing  with   the  importance  of the nonuniformity of the
      ventilation/perfusion ratio in the lung.
 March 12, 1990                         A-14     DRAFT - DO NOT QUOTE OR CITE

-------
Pollutant:  Generally, any substance introduced into the environment that adversely affects the
      usefulness of a resource.

Pollution:  Generally, the presence of  matter or energy whose nature, location or quantity
      produces undesired environmental effects.  Under the Clean Water Act, for example, the
      term is defined as the man-made or man-induced alteration of the physical, biological, and
      radiological integrity of water.

Population:  A group of interbreeding organisms of the same kind occupying a particular space.
      Generically, the number of humans or other living creatures in a designated area.

Respiratory frequency (f,0:  The number of breathing cycles per unit of time.  Synonymous with
      breathing frequency (fB).

Smog: Air pollution associated with oxidants.  (See photochemical smog)

Smoke: Particles suspended in air after incomplete combustion of materials.

Sulfur dioxide (SOJ: Colorless gas with pungent odor, primarily released from burning of fossil
      fuels, such as coal, containing sulfur.

Synergism:  A pharmacologic or toxicologic interaction in which the combined effect of two or
      more chemicals is greater  than the sum of the effect of each chemical alone.  (Compare
      with: additivity, antagonism.)

STPD conditions (STPD):  Standard temperature and pressure, dry. These are the conditions of
      a volume of gas at 0°C, at 760 torr, without water vapor. A STPD volume of a given gas
      contains a known number of moles of that gas.

Tidal volume  (TV):  That volume of air inhaled or exhaled with  each breath during quiet
      breathing, only used to indicate a subdivision of lung volume.  When tidal volume is used
      in  gas exchange formulations, the symbol VT should be used.

Time-weighted average (TWA):  The average concentration to which a worker may be exposed
      continuously for 8 h without damage to health.

Torr: A unit of pressure equal to 1,333.22 dynes/cm2 or 1.33322 millibars. The torr is equal
      to  the pressure required to support a column of mercury 1 mm high when the mercury is
      of standard density and subjected to standard acceleration. These standard conditions are
      met at 0°C and 45° latitude,  where the acceleration of gravity is 980.6 cm/s2.  In reading
      a mercury barometer at other temperatures and latitudes, corrections, which commonly
      exceed 2 torr, must be introduced for these terms and for the thermal expansion of the
      measuring scale used.  The torr is  synonymous with pressure unit mm Hg.

Total lung capacity  (TLC): The sum of all volume  compartments or the volume of air in the
      lungs after maximal inspiration.  The method of measurement should be indicated, as with
      RV.

 March 12, 1990                          A-15     DRAFT - DO NOT QUOTE OR CITE

-------
Ventilation:  Physiological process by which gas is renewed in the lungs.  The word ventilation
      sometimes designates ventilatory flow rate (or ventilatory minute volume) which is the
      product of the tidal volume by the ventilatory frequency. Conditions usually are indicated
      as modifiers; that is,

                  VE =  Expired volume per minute (BTPS),
                        and
                  Vj = Inspired volume per minute (BTPS).

      Ventilation often is referred  to  as "total ventilation" to distinguish  it from  "alveolar
      ventilation".  (See ventilation, alveolar)

Ventilation, alveolar (VJ:  Physiological process by which alveolar gas is removed completely and
      replaced with fresh gas.  Alveolar ventilation is less than total ventilation because when a
      tidal volume of gas leaves the alveolar spaces, the last part does not get expelled from the
      body but occupies  the dead space,  to be reinspired with  the next inspiration.  Thus the
      volume of alveolar  gas actually expelled completely is equal to the tidal volume minus the
      volume of the dead space.  This truly complete expiration volume times the ventilatory
      frequency constitutes the alveolar ventilation.

Ventilation,  dead-space (VD):   Ventilation per minute of the physiologic dead space (wasted
      ventilation), BTPS, defined by the following equation:

            VD = VE(PaC02 - PEC02)/(PaC02 - P.CO,)

Ventilation/perfusion ratio (VA/Q): Ratio of the alveolar ventilation to the blood perfusion volume
      flow through the pulmonary parenchyma. This ratio is a  fundamental determinant of the
      oxygen and carbon dioxide pressure of the alveolar gas and of the end-capillary blood.
      Throughout the lungs the local ventilation/perfusion ratios vary, and consequently the local
      alveolar gas and end-capillary blood compositions also vary.

Vital  capacity (VC):   The maximum  volume of air  exhaled from the point of maximum
      inspiration.

Warbug partition coefficient (K): The carbon monoxide/oxygen ratio that produces 50% inhibition
      of the oxygen uptake of the enzyme or, in the case of myoglobin, a 50% decrease in the
      number of available oxygen-binding sites.

Wood-burning stove pollution:  Air pollution caused by emissions of particulate matter, carbon
      monoxide, total suspended particulates, and polycyclic organic matter from wood-burning
      stoves.
 March 12,  1990                          A-16     DRAFT - DO NOT QUOTE OR CITE

-------
References
American College of Chest Physicians - American Thoracic Society (1975).  Pulmonary terms
      and symbols:  a report of the ACCP-ATS Joint Committee on pulmonary nomenclature.
      Chest 67: 583-593.

Bartels, H.; Dejours, P.; Kellogg, R. H.; Mead, J. (1973) Glossary on  respiration   and  gas
      exchange. Journal Applied Physiology 34: 549-558.

Collier, C. R.; Goldsmith, J. R. (1983) Interactions of carbon monoxide and hemoglobin at high
      altitude. Atmospheric Environment 17: 723-728.

U. S. Environmental Protection Agency (1989) Glossary of environmental terms and acronym
      list.  Washington, DC: Office of Communications and  Public Affairs; report no. 19K-
      1002.

U. S. Environmental Protection Agency (1989) Glossary of terms  related to health,  exposure,
      and risk management. Research Triangle Park, NC:  Air Risk Information Support Center;
      report  no.  EPA/450/3-88/016.   Available from:  NTIS, Springfield,  VA; PB89-
      184584/XAB.
                                                 ftU.S.GOVERNMENTPRINTINGOFFICE:1990-7'*1'- 15? ° ° " 3 5
 March 12, 1990                         A-17     DRAFT - DO NOT QUOTE OR CITE

-------