Preliminary Assessment of Adverse Health
Effects from Carbon Monoxide and Implications
for Possible Modifications of the Standard
Strategies and Air Standards Division
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
June 1979
-------
DRAFT
1 June 79
PRELIMINARY ASSESSMENT OF ADVERSE HEALTH EFFECTS FROM
CARBON MONOXIDE AND IMPLICATIONS FOR POSSIBLE MODIFICATIONS
OF THE STANDARD
STAFF PAPER
I. PURPOSE
The purpose of this paper is to evaluate the key studies in
the EPA document "Air Quality Criteria for Carbon Monoxide" and identify
the critical elements to be considered in the possible revision of the
primary carbon monoxide (CO) National Ambient Air Quality Standard
(NAAQS). The paper also identifies critical factors that must be con-
sidered in selecting an adequate margin of safety for the CO air quality
standard.
II. BACKGROUND
The Clean Air Act Amendments of 1977 provide authority and
guidance for setting and revising NAAQS, where appropriate. Primary
standards must be based on health effects criteria and provide an adequate
margin of safety to ensure protection of public health. Economic or
related, impacts cannot be considered in the selection of the standard
2
level. Further guidance provided in the legislative history of the
Clean Air Act indicates that margins of safety should be defined such
that standards are set at "the maximum permissible ambient air level ...
which will protect the health of any [sensitive] group of the population."
Also, margins of safety are to be defined such that the standards will
provide "a reasonable degree of protection ... against hazards which
2
research has not yet identified." In the final analysis, the primary
standard is set by the EPA Administrator based on his judgment of the
implications of all the health effects evidence, and the need for an
adequate margin of safety.
-------
The primary (health-based) and secondary (welfare-based)
NAAQS for CO are both presently set at 9 parts per million (ppm) and 35
ppm for 8-hour and 1-hour averaging times, respectively, not to be
^
exceeded more than once per year. This paper considers only the primary
NAAQS revision since there is no data to support setting a secondary
standard more stringent than the primary NAAQS.
III. APPROACH
The approach used in this paper is to identify the critical
factors to be considered in the standard-setting process for carbon
monoxide, specifically those points where judgments or decisions
must be made and where careful interpretation of incomplete or uncer-
tain evidence is required.. Where possible, the paper states our under-
standing of the evidence as it relates to a particular judgment or,
in some cases, proposes alternative choices that might be made. The
essential elements that are addressed in this process include the
following:
(1) the most probable mechanism(s) of toxicity,
(2) a description of the adverse effects attributed to carbon
monoxide and a judgment on the critical effect of concern
for standard-setting,
(3) a description of the most sensitive population groups,
(4) the level at which the indicator of adverse effects (blood
carboxyhemoglobin) signals a danger to public health in
the sensitive population,
(5) the CO exposure which could give rise to a critical
carboxyhemoglobin level,
(6) other aspects of the standard, and
-------
(7) a discussion of the uncertainties in the medical evidence
and other factors which should be considered in selecting
an adequate margin of safety and a final standard level.
IV. CRITICAL ELEMENTS IN THE STANDARD REVIEW
A. Mechanism of CO Toxlcity
2a
We interpret the existing health effects evidence to indicate
that the principal mechanism of CO toxicity is through hypoxemia
(deficient oxygenation of the blood). This mechanism suggests that
adverse effects on the body result from the strong affinity of blood
hemoglobin for CO (over 200 times greater than for oxygen), which
results in the formation of carboxyhemoglobin (COHb). Thus, the oxygen-
carrying capacity of the blood is. reduced since hemoglobin that has
combined with CO in this manner is not available to transport oxygen;
furthermore, the presence of COHb inhibits the release of oxygen from
the remaining hemoglobin. Effects on the cardiovascular, central
nervous, pulmonary, and. other systans are directly related to this reduction
in the ability of the blood to deliver oxygen to these systems.
The COHb in an individual's blood stream re-fleets input from
endogenous (produced by the metabolic breakdown of hemoglobin and other
heme-containing materials) and exogenous (derived from the external
environment) sources. The physiologic norm for endogenous COHb levels
has been estimated to be in the range of 0.3 - 0.7 percent, but endogenous
»n
production may be significantly increased V persons with nemolytic
4 5
anemias , in women during pregnancy and during the menstrual cycle , and
in persons taking certain types of drugs. Our judgment is that increments
above the physiologic norm, with the above-mentioned exceptions, result
-------
from exogenous sources such as cigarette smoking (which can result in COHb
levels ranging from 2 or 3 percent up to as high as 15 to 17 percent) a
or community air pollution.
Some researchers have suggested an alternative mechanism for
carbon monoxide toxicity that results from a blocking of the energy
flow at the cellular level through the cytochrome system . Dis-
cussion of some of these studies by the CO subcommittee of the Clean
Air Scientific Advisory Committee (CASAC) raised the possibility
that the results found by these researchers could be more readily
explained on the basis of the test protocols than as an alternative
mechanism of toxicity. Several of these studies exposed animals to
very high CO concentrations (greater than 100,000 ppm) for short time
periods, with the result that the animals died at a total-body
COHb level lower than that which would be required if the CO dose had
been administered as a lower concentration given over a longer period
12
of time. The CASAC discussion indicated that the hypoxemia mechanism
probably provides an adequate explanation of this phenomenon
as a manifestation of the "bolus effect" wherein a high concentration inhaled
over a short period of time results in a portion of the blood supply
that is essentially devoid of oxygen reaching the heart, with life-
threatening consequences. Since the CASAC questioned the sufficiency
of the information base on which the cytochrome system toxicity mechanism
has been proposed, we have decided to focus this NAAQS review on the
hypoxemia mechanism and related COHb levels associated with
observed adverse health effects.
B. Description of Adverse Effects
1. Cardiovascular System Effects. Angina pectoris is a cardio-
-------
vascular disease in which mild exercise or excitement produces symptoms of
pressure and pain in the chest because of insufficient oxygen supply to the
heart muscle. Angina patients have been reported to experience heart
pains earlier during exercise after a resting exposure to low levels of
13 16
CO. We consider aggravation of angina to be an adverse health
effect because it may result in cardiovascular damage. Aggravation
of angina has been convincingly demonstrated at COHb levels of 2.5 -
13 14
3 percent, ' which are lower than those associated with any other
measurable adverse effect of CO exposure. A 1978 study by Aronow
has reported angina aggravation at CQHb levels in the range of about 1.8 -
2.3 percent, but these COHb levels were obtained through a passive
smoking exposure regime,, with possible confounding factors. The
appropriate utilization of this study in the standard-setting process
will be discussed in a subsequent section.
Patients having peripheral vascular disease may also have this
18
condition aggravated by CO exposure. The one clinical study examining
such an effect involved the exposure of 10 persons with occlusive
arterial disease to 50 pom CO for 2. hours followed by exercise until
leg pain occurred. The COHb levels produced by this exposure (about
2.3 percent) significantly decreased the time to onset of pain and
cessation of activity.
Another cardiovascular system effect of concern is the
possible detrimental effect of increased blood flow that occurs
19 2Q
as a compensatory response to CO exposures- ' This response could
result in coronary damage or other vascular effects due to the cardio-
vascular system being pushed beyond its capabilities.
Cardiovascular damage and electrocardiogram abnormalities have been
reported in persons who have experienced acute non-fatal CO poisoning
-------
21
episodes (20 percent COHb) or who have been chronically exposed to CO
79 2"\
in the workplace. ' Chronic exposure to average CO levels of 70 ppm,
with peak levels of about 300 ppm (no COHb levels reported), has been
associated with increased mortality from heart failure in a Japanese
24 25-27
population. Some epidemiological studies in Los Angeles
have suggested the possibility that increased mortality
from myocardial infarction (death of heart tissue) is associated with
high ambient air concentrations of CO (sufficient to produce COHb levels
26 28
in the range of 8-17 percent), but a similar study in Baltimore
failed to find such a correlation with considerably lower ambient CO
levels (sufficient to produce COHb levels in the range of 1-10 percent).
29
Another study evaluated patients admitted to the myocardial infarction
research unit at the Johns Hopkins University. While the investigator's.
diagnoses were consistent with both acute and chronic effects on the
myocardium of long-term exposure to CO, the effects observed could not
be clearly related to that factor. Therefore, the possibility of an
association between CO levels in the ambient air and incidence of
myocardial infarction or of sudden deaths due to arteriosclerotic heart
disease remains in question; more research is needed to clarify this issue.
A recent epidemiology study reported an increase in the frequency
of cardlorespiratory complaints by patients at the emergency
room of a Denver hospital on "high CO days" when the maximum 1-hour
mean ambient CO concentrations at a nearby monitoring site averaged
27 ppm as compared to "low CO days" when the- corresponding concentration
was 12 ppm. However, the CASAC expressed a need for caution in the in-
terpretation of this study because (1) the cardiorespiratory complaints
evaluated in this study are inadequate indicators of cardiovascular
-------
disease aggravation, (2) the authors did not report any COHb levels for
the patients evaluated and (3) the single monitoring site near the hospital
is inadequate to determine the exposures sustained by the patients.
32-35
2. Central Nervous System Effects. Some studies have re-
ported that CO exposures resulting in CQHb levels of 1.3-7.5 percent
have produced decrements in vigilance (the individual's ability to
detect small changes in his environment that take place at unpredic-
table times). The effects of CO on vigilance are in considerable
dispute, however, since similar studies have failed to observe
similar effects. Our judgment of the evidence is that, if relevant
variables are controlled, CO exposures at a threshold level of about
4-6 percent CQHb may produce decrements in vigilance. Decrements in
visual function and sensitivity have been reported at COHb levels as
low as 4> to 5 percent.42'43
We consider vigilance and visual function effects to be important
since these functions are components of more complex tasks, such as
driving, and reduced alertness or visual sensitivity could lead to
IT 44-59
increased accidents. Several studies *have suggested that
elevated COHb levels adversely affect the performance of complex
tasks, and suggestive (but not conclusive) evidence has been reported
indicating that a greater proportion of drivers in fatal accidents have
eg
COHb levels above 5 percent. In this respect, the possibility of an
interactive effect between alcohol and CO that has been suggested by
some experimentation seems to be of particular concern.
3. Pulmonary Function and Exercise Effects. In studies using
submaximal exercfse for short periods (5 to 60 minutes), oxygen uptake
-------
8
during work does not appear to be affected by COHb levels as high as
10 to 20 percent, although several of these studies have shown
that these COHb levels produce increased heart rates. In maximal
exercise protocols of several minutes duration, COHb levels in the range
of 5-33 percent have been demonstrated to produce a linear decline in
maximal oxygen uptake (and hence work capacity). One study has
reported that competitive swimmers have impaired performance when events
are held in atmospheres containing 30 ppm CO originating from traffic
(COHb level not reported), but these results may reflect interaction
with other pollutants since ambient air exposures were examined in this
study rather than systematic,, controlled exposures to CO.
Although few studies have been conducted, to examine the effects of
CO exposure on persons with chronic obstructive pulmonary disease (e.g.,
asthma, emphysema, and chronic bronchitis),. this group is presumably at
high risk to CO exposures. A reduction in oxygen supply due to increased
COHb levels could exacerbate existing effects of low oxygen levels
caused by impaired respiratory system functioning. However, such persons
may absorb less CO due to their disease and may have compensated for
their respiratory deficiencies by increased production of red blood
cells and by other adaptations. One study exposed ten persons with
chronic obstructive pulmonary disease for 1 hour to sufficient CO to
produce 4.1 percent COHb. A 33 percent reduction in time to onset
of marked dyspnea (difficulty in breathing) occurred during exercise as
compared to COHb levels of 1.5 percent. The investigators concluded that the
limited exercise performance after CO exposure was probably a cardio-
vascular limitation rather than a respiratory one.
-------
yjj (iA
*. Fetal Development. Several experimental animal studies
have exposed pregnant females to CO and in general have shown deleterious
effects in the offspring even when the mothers were not affected. For
84
example, one study exposed pregnant rabbits to 90 ppm CO continuously
for 30 days, with resultant maternal COHb levels of 9 to 10 percent.
Birth weights decreased IT percent and the newborn mortality rate increased
81
to 10 percent from a control value of 4.5 percent. In another study,
the offspring of rats exposed to 150 ppm CO throughout gestation (maternal
COHb levels, of 15 percent) weighed slightly (3 percent) less at birth and
failed to gain weight as rapidly after birth. Reduced brain protein levels
at birth and lower behavioral activity levels through the preweaning period
were observed. In many cases the deleterious effects have been shown to
disappear by adulthood, but the inference that such effects may occur in
humans during maturation is of concern with respect to possible impacts
on learning and social behavior development.
78 82.
Experimental animal studies ' have shown that short-term maternal
CO exposure results in lower COHb levels in the fetus than in the mother,
but may have greater detrimental effects on the fetus than on the mother. One
gc
study examining long-term maternal CO exposures has shown that fetal
COHb levels exceeded maternal values after an uptake lag of a few hours
following initiation of exposure. At equilibrium, fetal COHb levels
significantly exceeded maternal values, and fetal elimination of CO after
cessation of exposure was slower than maternal elimination. The COHb
concentration in human fetal blood has been reported to vary from 0.7 to
2.5 percent (for non-smoking mothers)", with the ratio of fetal to maternal
86
CQHb levels varying from O.S to 1.5.
The biologic affects of CO exposure on fetal tissues during intra-
uterine development require clarification. The ability of CO to decrease
-------
10
the oxygen transport capacity of maternal and fetal hemoglobin may result
in interference in fetal tissue oxygenation during important developmental
stages. Whereas a normal adult has reserve capacity and compensatory
responses that enable him to handle moderately high COHb levels without
irreversible consequences, the fetus may under normal situations be
operating close to the critical levels in terms of tissue oxygen supply.
Thus, even moderate CO exposures may have a deleterious effect on fetal
86
development.
Some verification of this hypothesis has been suggested in studies
examining the impact of maternal smoking and altitude on the unborn child.
Several studies have demonstrated that babies born to smoking mothers have
ft7.Q4 Q5 Qfi
reduced birth weight, ** as do children born at higher altitudes. °'*D
Some question remains as to the relationship of fetal deaths to maternal
92-94 97-102
smoking, but several studies ' using data from large population
samples have concluded that perinatal deaths do increase in the infants of
mothers who smoke when these data are corrected for other factors affecting
perinatal death rates (e.g., maternal age, the number of children previous-
ly borne, race, and social status). The causes of this increased perinatal
mortality have not been adequately identified.
o-f C* *
The/pffesfc of maternal smoking on surviving children is not well under-
stood. One study has reported almost a two-fold increase in the incidence
of congenital heart disease in the infants of mothers who smoke. British
studies ' of large population groups have found highly significant
differences in reading attainment at seven years of age between the children
of mothers who smoked and those who did not. A follow-up study of these
children at 11 years of age found several months retardation in general
-------
n
ability, reading, and mathematics between the children of mothers who
smoked (0.5 pack of cigarettes or more per day; average COHb levels,
though not reported, may have been 3 to 4 percent or more) and those
who did not. While the fact that cigarette smoke contains substances
other than CO prevents a direct application of the results of these
studies in setting the CO standard, the studies do suggest the need for
caution in protecting unborn children from such potentially deleterious
effects of CO exposure.
C. Population Groups Most Sensitive to Low Levels of CO
On the basis of the previous section's description of the principal
adverse effects associated with low levels of CO exposure, we have concluded
that the following groups may be particularly sensitive to exposures of CO:
angina patients, individuals with other types of cardiovascular disease, per-
sons with chronic obstructive pulmonary disease, anemic individuals, fetuses,
and pregnant women. While there is no evidence in the criteria document
that healthy children are particularly sensitive to CO, concern exists that
they may also be at increased risk ta CO exposure because of the increased
oxygen requirements that result from their higher metabolism rates.
In our judgment* the available health effects data identify per-
sons with angina and those with other types of cardiovascular disease as the
groups at greatest risk from low-level, ambient exposures to CO. Aggrava-
tion of angina has been convincingly demonstrated to occur at COHb levels
1314
(about 2.5-3 percent) ' which are lower than those associated with any
other measurable adverse effect of CO exposure. The low threshold to
CO effects results from the fact that the angina condition is due to an
insufficient oxygen supply to cardiac tissue, so that such persons have an
inadequate reserve capacity and an impaired ability to compensate for
the effects of CO.
-------
12
The second group of prime concern consists of individuals suffering
from other cardiovascular diseases, such as peripheral vascular disease.
While much less information is available concerning the effects of CO on
this group, one clinical study has shown significantly decreased exercise
time to onset of leg pain at COHb levels of about 2.8% for persons with
peripheral vascular disease.
A wide variation exists in the estimates of the number of persons
who have angina and other types of cardiovascular disease. It has been
estimated that 2.1% of the population has stable angina, while
some cardiologists believe the total number of angina patients may range
108
up to 25% of the national population. The U.S. National Health Survey
Examination reported that of the population aged 18 to 79 years, 3.1 million
persons have definite heart disease and another 2.4 million are suspected
109
to have heart disease.. Another National Health Survey estimated that 12% of
the population has arteriosclerotic disease. In addition, data from
autopsies has indicated that nearly 25% of persons dying from coronary
disease have had no prior recognized symptoms of heart disease.
On the basis of the available effects data, we are focusing on
angina patients and those with other types of cardiovascular disease as
the most sensitive groups. Other groups such as fetuses, anemics, and
persons with chronic obstructive pulmonary disease can reasonably be
projected to be affected at CO levels possible in ambient air exposures.
Because of the lack of human data for these groups, however, the potential
effects on such persons will be considered in determining the margin of
safety for the CO standard.
D. Critical COHb Levels
Table 1 provides a summary of the key health effects studies
-------
13
TABLE 1
ESTIMATED HEALTH EFFECTS LEVELS FOR CARBON MONOXIDE EXPOSURE
Effects
COHb concen-
tration,^
References
Passive smoking aggravates
angina pectoris
Decreased exercise capacity
in patients with angina pec-
toris, intermittent claudica-
tion, or peripheral arterio-
sclerosis
Impairment of vigilance tasks
in healthy experimental subjects
Decreased exercise performance
in normal persons and in
patients with chronic obstructive
pulmonary disease
Increased angina attacks for
freeway travel
Changes in heart functioning
and possible impairment
Linear relationship between COHb
and decreasing maximal oxygen
consumption during strenuous
exercise in young healthy men
Statistically significant diminu-
tion of visual perception* manual
dexterity, ability to learn, or
performance in complex sensor!motor
tasks (such as driving)
1.3-3.3 17. Aronow, 1978
2.5-3.0 13. Anderson et al., 1973
14. Aronow and Isfaell, 1973
18. Aronow et al., 1974
3.0-7.6 35. Horvath et al., 1971
33. Groll-Knapp et al., 1972
32. Fodor and Winneke, 1972
3.0-4.9 74. Aronow and Cassidy, 1975
77. Aronow et al., 1977
3.3-8.0 16. Aronow et al., 1972
3.9 131. Aronow et al., 1974
5-20 67. Efcblom and Huot, 1972
73. Horvath, 1975
70. Danms et al., 1975
75. Seppanen, 1977
5-17 45. Bender, et al., 1971
46. Schulte, 1973
47. O'Oonnell et al., 1971
48. McFarland, 1973
41. Putz et al., 1975
50. Salvatore, 1974
53. Wright et al., 1973
55. Rockwell and Weir, 1975
58. Rummo and Sari am's, 1974
physiologic norm (i.a., COHb levels resulting from the normal metabolic
breakdown of hemoglobin and other name-cantaining materials) has been
estimated to be in the range of 0.3 to 0.7 percent.
-------
14
reporting effects due to elevated blood COHb levels. While the table in-
cludes studies reporting effects in three of the basic categories described
earlier, the selection of a critical COHb level will be based primarily on
the cardiovascular effect category, as discussed previously.
The present NAAQS for CO is not based on the cardiovascular effect
category but on central nervous system effects. The Federal Register
.113
112
notice promulgating the existing standard identified 2 percent as
the critical COHb level on the basis of a study by Beard and Wertheim
which reported an impairment in discrimination of time intervals in sub-
jects having estimated COHb levels of 2 to 3 percent. The revised criteria
document states that considerable questions, have been raised as to the
validity of these observations and points out that attempts to replicate
these findings have been less than satisfactory. ' As stated
previously, we have concluded that other central nervous system effects,
i.e. vigilance and visual function decrements, may occur in the range of
4 to 6 percent COHb.
In selecting a standard level, we propose to identify a critical
COHb blood concentration which we judge to represent most accurately
the lowest concentration that credible studies have convincing-
ly associated with human health effects of concern for sensitive persons.
This element of the standard-setting rationale does not include margin of
safety considerations, and does not reflect those uncertainties in
the medical evidence which must eventually be considered in the margin of
safety for the standard.
Selection of the critical COHb blood level is a key element in the
standard decision and has a direct effect on the final standard level.
-------
15
Based on medical evidence presented in the criteria document and summarized
in Table 1, it appears that two basic options exist regarding the appropri-
ate COHb levels. These options include:
(1) a COHb level in the range of 2.5 - 3.0 percent or,
(2) a COHb level of approximately 1.3 percent.
Selection between these options is primarily a function of the weight placed
on various studies reporting effects and related COHb concentrations. The
revised criteria document appears to endorse a range of 2.5 to 3 percent:
"It still seems safe to conclude that cardiovascular effects can be
demonstrated with CO exposures as low as...(15-18 ppm CO for an
119
3-hour exposure; 2.5 - 3.0 percent COHb)".
13 11
Two studies * have reported that aggravation of angina pectoris
resulted from CO exposures sufficient to produce COHb levels in the range
of 2.5 to 3 percent. These studies, reported a decrease in the amount of
exercise required to induce angina attacks. As stated earlier, we consider
this effect to be serious because it may result in heart damage.
A lower COHb level might be selected if more weight were given to
the 1978 Aronow passive smoking study. This study suggests that a
COHb level of approximately 1.8 percent is the point where aggravation of
angina starts to occur in angina patients. The criteria document does include
a cautionary statement regarding Interpretation of this study because
"it is possible that in addition to carton monoxide and nicotine, other
components of tobacco smoke, including oxides-of nitrogen and hydrogen
cyanide, and possibly psychological factors, may have contributed to the
12Q
decrease in exercise performance".
-------
16
121
The Clean Air Scientific Advisory Committee (CASAC) debated this issue
and characterized the situation as one where there was a preponderance of
evidence to support a COHb value of approximately 2.5 to 3 percent and
that one new study (Aronow, 1978) reported effects at approximately 1.8
ppm. The issue is essentially whether the 1978 Aronow study should form
the basis for selecting a critical COHb level of about 1.8 percent or
whether that study should receive less weight and a COHb level of 2.5
percent be selected as a critical value. If the higher COHb level is
selected, the 1978 Aronow study would be considered in selecting an
adequate margin of safety in the final air quality standards.
E. CO Exposure and Resulting COHb Levels
The primary factors determining the final level of COHb in individuals
are inspired CO concentrations, alveolar ventilation rates (which depend on
the level of exercise), endogenous CO production, red cell volume in the
blood, barometric pressure, and the relative diffusive capability of the
lungs. For tobacco smokers, the primary source of CO and the resulting
122
COHb levels is from the intake of tobacco smoke. In the following
discussion on the uptake of CO by individuals, smokers have not been
considered since "smokers generally are excreting CO into the air rather
123
than inhaling it from the ambient environment."
An approximate relationship between CO exposures and equilibrium
124
COHb levels was stated by Haldane and co-workers in 1912. The Haldane
equation shows that the ratio of the concentrations of COHb and oxyhemo-
globin (02Hb) is proportional to the ratio of the partial pressures of CO
and oxygen. For practical purposes, the Haldane equation can be used to
-------
17
estimate the level of COHb achieved at equilibrium upon exposure to various
concentrations of CO from the ambient environment.
The time required to reach equilibrium is influenced by a number of
factors, the most important for normal individuals being the level of
exercise as measured by the alveolar ventilation rate (VA). At low levels
of activity (resting, VA * 5-10 L/min) approximately 8-12 hours are needed
to achieve equilibrium. For a moderate walk (3 miles per hour, VA a 20 L/min),
the equilibrium level of COHb may be reached in half that time, or around
4-6 hours.
The Haldane equation cannot be used to predict COHb levels that are
achieved prior to equilibrium. The non-equilibrium levels of COHb are
required to evaluate the COHb levels associated with alternative 1- and 8-
-------
18
experimental verification is needed to demonstrate that the Coburn equation
accurately predicts uptake and excretion of CO under a variety of conditions,
it is the best tool available for estimating the COHb levels that will result
from short-term (1- to 8-hour exposures to ambient CO concentrations.
Table 2 shows the relationship for non-smokers between percent COHb
and various exposures of CO as estimated by the Coburn and Haldane equations.
A moderate level of exercise, equivalent to a 3 miles-per-hour walk, has
been selected as a reasonable estimate of the maximum exercise level achieved
by most individuals with angina or cardiovascular heart disease. These
estimates would indicate that CO concentrations at the current standard could
lead to COHb levels of about 2.1 and 1.5 percent for the 1-hour (35 ppm) and
8-hour (9 ppm) averages, respectively. For resting individuals, the
current 1- and 8-hour standards are reasonably consistent in that both are
expected to result in COHb levels of about 1.4 to 1.6 percent.
The impact of exercise level on rates of COHb accumulation is most
apparent for shorter duration exposures. As illustrated in Table 2,
very little difference exists in the COHb levels for resting and
moderately exercising persons for 8-hour exposures to CO concentrations
near the current standard level. Consequently, assumptions regarding
exercise levels at the 8-hour averaging time are not as critical in
estimating COHb levels as for the 1-hour averaging time.
-------
19
TABLE 2
PERCENT COHb AS A FUNCTION OF CO EXPOSURE
CO
(ppm)
5.0
9.0
15.0
20.0
25.0
35.0
50.0
Resting
0.6
0.7
1.0
1.1
1.3
1.5
2.2
% COHb Based on Coburn Equation
Exposure Time (Hours)
1
Moderate
Exercise
0.6
0.3
1.1
1.4
1.6
2.1
2.9
2 4
Resting
0.7
0.9
1.3
1.6
1.9
2.5
3.5
Moderate
Exercise
0.7
1.0
1.5
2.0
2.4
3.2
4.5
Resting
0.3
1.2
1.3
2.3
2.8
3.3
5.2
Moderate
Exercise
0.3
1.3
2.0
2.6
3.2
4.5
6.3
% COHb at
Equilibrium
Based on
Haldane
Equation
3
Resting
0.9
1.4
2.2
2.9
3.6
4.9
7.0
Moderate
Exerci se
0.9
1.5
2.4
3.1
3.3
5.3
7.5
0.9
1.6
-
3.5
-
-
3.2
Assumed conditions: Alveolar ventilation rates: resting = 10 L/min,
moderate exercise » 20 L/min (equivalent to 3
mpn walk on level ground or light industry or housework);
hemoglobin » 15 g/TOO mL (normal); altitude » sea level;
endogenous COHb level a 0.5 percent.
Accumulation rates and related COHb levels could be higher than
indicated in Table 2 for individuals with anemia, pregnant women, fetuses,
and individuals taking cartain types of medication or drugs. For example,
because of their reduced hemoglobin (Hb) levels, anemic individuals approach
equilibrium levels of COHb mor« rapidly than those with normal Hb levels,
and consequently attain a higher COHb level for a given exposure to CO. The
Coburn equation predicts that for moderately exercising individuals, exposure
to 35 ppm CO for 1 hour would result in 3.3 percent COHb in a person with
severe anemia (Hb 3 7 g/100 mL), compared to an anticipated level of 2.1
percent for normal individuals. The preceding calculations are based
-------
20
on persons with a normal endogenous COHb level (0.5 percent); however,
persons with hemolytic anemias may attain even higher COHb levels for a
given exogenous CO exposure due to their significantly (2- to 9-fold)
126 127
increased endogenous CO production. *
Table 2 portrays the COHb levels that may be achieved upon exposure
to various CO concentrations for 1 or more hours. Exposure to much
higher CO concentrations (50-200 ppm) for very short periods of time
(1-60 minutes) may also result in significant localized COHb blood levels.
(bolus effect). For instance, moderately exercising individuals exposed
to 50 ppm CO for 40 minutes or 200 ppm for 10 minutes may reach COHb
levels in excess of 2.5 percent. These unusually high levels of CO may
128
result from any of the, following scenarios: (1) in heavy traffic
that has come to a halt, the ambient CO level may exceed 40-50 ppm; (2)
inside a closed auto where cigarettes^are being smoked, CO concentrations
may exceed 87 ppm; (3) in enclosed, unventiTated garages, CO levels in
excess of 100 ppm have been found; (4) in a heavily-traveled vehicular
tunnel, a 1-hour maximum of 218 ppm CO was recorded; and (5) for certain
occupational exposures, such as those encountered by firefighters,
foundry workers, miners, toll collectors, and taxi drivers, CO concentrations
of 200 ppm for short periods and 60 ppm for 8 hours have been reported.
F. Other Aspects of the CO Standard
Selection of the averaging time, form of the standard, and allowed
number of exceedances are all key elements in the standard decision and
have a direct effect on the required level of control. The NMQS for CO
is presently 9 ppm and 35 ppm for 8-hour and 1-hour averaging times, res-
pectively, not to be exceeded more than once per year.
-------
21
The original 8-hour averaging time was selected primarily because
most individuals achieve equilibrium or near-equilibrium levels of CQHb
after 8 hours of exposure. As mentioned previously, approximately 4-12
hours are required to achieve an equilibrium level of COHb upon con-
tinuous exposure to CO. The time to reach equilibrium is influenced pri-
marily by the exercise level of the individual, with shorter times required
for greater exercise* Another basis for the 3-hour averaging time is that
most people are exposed in approximately 3-hour blocks of time (e.g.,
work, sleep). With respect to the 1-hour averaging time, the health
effects rationale is now stronger than in 1971. Aronow has conducted
several studies14'17'13'77 which have reported health effects for
persons with cardiovascular disease or chronic obstructive pulmonary
disease after 1- to 2-hr exposures. In light of the above considerations
we see no need to change the current 1-hour and 3-hour averaging times
for the CO NAAQS.
As was noted in the preceding section, short-term (1-60 minute)
peak CO concentrations (50-200 pom) that have been observed in ambient
situations may result in CQHb levels of concern due to the bolus effect.
However, analyses of existing air quality data suggest that attainment of a
longer averaging time standard will limit the magnitude of short-term
peak concentrations. For example, air quality data obtained in 1974-1976
at a site in Los Angeles that is 5 meters (m) high and 3 m from the curb
of a highway bearing an average traffic load of 25,000 vehicles per day
were analyzed to determine for various averaging times the peak concentra-
-------
22
tions that would be expected if the present NAAQS were just attained. The
maximum values observed were 46 and 30 ppm for 1- and 8-hour averaging
times. Using a 2-parameter averaging time model, we have calculated
that attainment of the 8-hour standard would result in peak 5- and 10-
minute concentrations of 36 and 29 ppm, respectively. Although this
site might not represent the worst possible situation, this analysis
does seem to indicate that attainment of longer averaging time standards
would tend to limit short-term peaks.
No decision has been reached on a recommendation regarding the need
for, or nature of, any modification of the standard. However, in the
event that a recommendation were made to retain the current 8-hour standard
level, a question would remain as to the appropriateness of retaining the
current 1-hour standard level. As indicated by the COHb levels anticipated
for moderately exercising individuals for the current 1- and 8-hour standard
levels, arguments could be made for some change in the 1-hour standard level.
On the other hand, a case could be made for leaving the standard essentially
unchanged on the basis that attainment of the existing 9 ppm 8-hour
standard level also protects against shorter averaging times of concern.
The latter choice might be attractive because it would limit disruptions
in the existing air quality management program without any apparent
health liabilities.
The current CO standard has a deterministic form, allowing only one
exceedance per year. This deterministic approach has several limitations,
one of which is that it does not adequately take into account the random
nature of meteorological variations. The original purpose of permitting a
single exceedance was to allow for unique meteorological conditions that
were not representative of air quality problems in a given area. However,
-------
23
since the present form of the standard specifies, in effect, that there be
zero probability that the second-highest concentration measured in a year
exceed the standard level, it does not achieve this purpose because when
a single exceedance of the standard is permitted, a definite possibility
exists that a second or third exceedance will also occur. The only way
to be certain that subsequent exceedances will not occur is to permit no
initial exceedances of the standard. The limitation of the current
deterministic form means that compliance with the standard, and consequently
pollutant emission control requirements, would be determined on the
basis of exceedingly rare weather conditions. These are the same arguments
that prompted EPA to change the ozone standard from a deterministic to a
129
statistical form. Because of these arguments, we believe that the CO
standard also should be statad in a statistical form. This would mean
that the allowable number of excsedancas of the standard would be expressed
as an average or expected number per year. The emission reductions to
be achieved in the required control implementation program would be
based on a statistical analysis of monitoring data over the preceding 3-
year period.
We are also considering changing the form of the CO standard to per-
mit one calendar day in which the 1-hour standard cauld be exceeded. No
health effects rationale exists to support such a change; however, as in
the case of the ozone standard revision, this form of the standard has
several advantages, such as (1) requiring less interpretation of data in
calculating attainment or non-attainment and (2) achieving greater stability
in the design statistics needed for control strategy development. We are'
still studying alternative forms for the 3-hour standard before making a
decision on how to handle the problem of overlapping or running averages.
-------
24
V. FACTORS TO BE CONSIDERED IN SELECTING A MARGIN OF SAFETY AND A
STANDARD LEVEL
Selecting an ambient air quality standard with an adequate margin
of safety requires that the Administrator consider and account for
uncertainties in the health effects evidence in arriving at the standard.
In the case of CO, these effects are principally associated with the
evidence leading to a critical COHb level and with the relationship
between ambient CO exposures and the resultant COHb levels in selected
population groups under various environmental conditions and levels of
stress. These factors include: (1) the relevance of the 1978 Aronow
study, (2) the implications for humans from animal study findings, (3) the
predictability of the relationship between ambient CO concentrations and
elevated COHb levels, (4) the altitude effect, (5) the increased risk of
adverse, effects in persons with anemia, and (6) the bolus effect.
A. Role of the 1978 Aronow Study in Selecting a Critical COHb Level
The preponderance of evidence indicates that adverse effects in angina
patients are associated with COHb levels in the range of 2.5 - 3.0 percent.
The 1978 Aronow study (previously discussed) is an exception and suggests
that effects can occur in angina patients at COHb levels of approximately
1.8 percent. If this study is given full weight and is interpreted as
showing adverse effects down to COHb levels of 1.8 percent, a more
stringent 1-hour standard must be considered. Such a standard would
need to protect against COHb levels below 1.3 percent in order to provide
an adequate margin of safety.
The fact that the 1978 Aronow study has'not been replicated, as well
as the possibility that effects may have been enhanced by exposure to
toxic agents other than CO, argues for giving this study less weight
in selecting the final air quality standard. In that case, 2.5 - 3.0 percent
would be considered the critical COHb level and the 1978 Aronow study
-------
25
would serve as an element for consideration in selecting an adequate
margin of safety but not in Identifying the specific critical COHb level.
3. Animal Studies
Although the findings of animal studies currently cannot be
extrapolated directly to identify a CO or COHb concentration that will
cause an effect in man, much of the evidence compiled from animal studies tends
to reinforce the findings obtained in human studies. One important area of
concern still exists, however, where little human data are available. This
evidence relates primarily to animal studies showing that the developing
fetus is exposed to COHb concentrations considerably higher than the
pregnant mother for long-term CO exposures. Because the COHb levels
may well be elevated in the fetus, and because the fetus is probably more
susceptible to the adverse impacts of impaired oxygen delivery, findings
from these animal studies denote a need for caution in assessing possible
human effects and in establishing an adequate margin of safety for the
standard. Even moderate CO exposures may have a harmful effect on fetal
development, as has been suggested in studies of smoking mothers.
C. Uncertainty Regarding the Relationship Between Ambient CO
Exposure and Resulting COHb Levels
No simple model 1s available that can provide a foolproof method of
predicting COHb levels that result from alternative CO exposure concentra-
tions and patterns. Not only do we lack a perfect model to provide this
information, but numerous confounding factors, such as altitude, smoking
habits, exercise levels, and individual health status, make the task even
more difficult. While projections must be made of ambient CO concentra-
tions that could produce critical CQHb levels, the uncertainty in these
projections must be accounted for in the margin of safety.
-------
26
D. Altitude
Hypoxemia can occur at higher altitudes due to the reduced oxygen
pressure in the atmosphere; in addition, altitude can increase the rate
of accumulation of COHb in the blood. However, normal residents of
high-altitude locations appear to have adjusted to the elevation and do
not seem to exhibit an interaction between the effects of altitude and
CO exposures. Still of concern are impaired visitors from
lower altitude locations who may be adversely affected by altitude
hypoxemia and by COHb concentrations which are higher than those reached
at lower elevations for the same ambient CO concentration. The possible
adverse effect on impaired visitors to high-altitude areas should be
considered in selecting an adequate margin of safety for the CO standard.
E. Individuals with Anemia
Little quantitative data are available on the COHb concentrations
that result in anemic individuals for CO exposures near or below the
current standards. The Coburn equation predicts little difference in the
final COHb level achieved by anemics and normal individuals for 8-hour
exposures to 10 ppm, but a more significant difference for shorter-term
exposures at higher concentrations. For example, persons with severe
anemia (Hb = 7 g/100 ml) who are moderately exercising and are exposed
to 35 ppm CO for 1 hour may reach 3.3 percent COHb, compared to an
anticipated level of 2.1 percent for normal individuals. In addition to
their more rapid equilibration with a given exogenous CO exposure, persons
with hemolytic anemia may be expected to attain even higher COHb levels
due to their higher initial COHb values (due to increased endogenous CO
-------
27
production). While dietary anemia is no longer as widespread a disease
as it was some years ago, anemias resulting from various pathological
conditions are still major health problems and low levels of CO exposure
could pose a health threat to such groups. Anemic individuals who also
suffer from angina or who are. pregnant would seem to be particularly
high-risk categories. However, sines very little data exist regarding
the effects of CO exposure on anemic Individuals, we must base our
assessment of the critical COHb level on the available evidence for persons
with angina and other cardiovascular diseases, in whom the lowest CO
effect level has been observed. Nevertheless, the increased risk for
anemics, and other Individuals whose uptake of CO is greater, should be
considered in determining an adequate margin of safety.
F. Smokers
Little or no evidence exists that would suggest the need for a more
restrictive national ambient air quality standard to protect smokers
from a possible incremental COHb burden from the air. In fact, even in
a pristine environment, smokers will have COHb levels ranging from 2
to 17 percent. Furthermore, the existing evidence suggests that cigarette
smokers exposed to an environment where CO is at the ambient standard
level are generally excreting more CO into the air than they are inhaling
from that environment. While these individuals are at a definite risk
from accelerated incidence of heart and related disease, these risks are
principally associated with cigarette smoking and not with incremental
CO levels at or near the standard.
3. Bolus Effect
A factor that should be considered in selecting an adequate margin
of safety is the uncertainty relating to adverse health impacts from
short duration (5-10 nrinutas) high-level CO exposures. While the data
-------
28
base is incomplete regarding the mechanism or effect of the peak short-
term exposures, the criteria document and the CASAC have underlined the
potential seriousness of CO delivered to the body in this manner. The
consequence may well be adverse effects separate and, perhaps, more
intense than those associated with elevated COHb concentrations, a
phenomenon referred to as the bolus effect. While this factor should be
a consideration in establishing an adequate margin of safety for the
standard, existing air quality data does indicate that attainment of a
longer averaging time standard will limit the magnitude of short-term
peak concentrations, as discussed previously. Consequently, the bolus
effect, while of concern in selecting a margin of safety, does not
appear to be an over-riding consideration in the final standard decision.
-------
29
REFERENCES
1. U.S. EPA. Air Quality Criteria for Carbon Monoxide. External
Review Draft, April 1979. (Hereinafter referred to as Criteria
Document.) Environmental Criteria and Assessment Office,
Office of Research and Development, U.S. EPA, Research Triangle
Park, N.C.
2. A Legislative 'History of the Clean Air Act Amendments of 1970,
p. 410.
2a. Criteria Document, pp. 9-1 to 9-12.
3. Ibid. Chp. 11 Ref. 222 (Coburn et al., 1969).
4. Ibid. Chp. 9 Ref. 15 (Cobum et al., 1966).
5. Ibid. Chp. 11 Ref. 53 (Delivoria-Papadopoulos et al., 1970).
5. Ibid. Chp. 9 Ref. 9 (Cobum, 1970).
6a. Ibid. pp. 11-51 and 11-52.
7. Ibid. Chp. 9 Ref. 53 (Ramirez et al., 1974).
3. Ibid. Chp. 9 Ref. 25 (Soldbaum, 1977).
9. Ibid. Chp. 9 Ref. 27 (SoTdbaum et al., 1975).
10.. Ibid. Chp. 9 Ref. 28 (Goldbaum et al., 1976).
11. Ibid. Chp. 9 Ref. 23 (Soldbaum et al., 1975).
12. CASAC Transcript, Jan. 30-31, 1979, pp. 67-69.
13. Criteria Document. Chp. 11 Ref. 2 (Anderson et al., 1973).
14. Ibid. Chp. 11 Ref. 5 (Aronow•& Isfaell, 1973).
15. Ibid. Chp. 11 Ref. 9 (Aronow & Rokaw, 1971).
16. Ibid. Chp. 11 Ref. 7 (Aronow et al., 1972).
17. Ibid. Chp. 11 Ref. 2b (Aronow, 1973).
13. Ibid. Chp. 11 Ref. 6 (Aronow et al., 1974).
19. Ibid. Ghp. 11 Ref. 14 (Ayres et al., 1970).
20. Ibid. Chp. 11 Ref. 15 (Ayres et al., 1969).
21. Ibid. Chp. 11 Ref. 48 (Corya et al., 1976).
22. Ibid. Chp. 11 Ref. 221 (Zankevic, 1973).
23. Ibid. Chp. 11 Ref. 57 (Ejam-8erdyev, 1973).
-------
30
REFERENCES
24. Ibid. Chp. 11 Ref. 123a (Komatsu, 1959).
25. Ibid. Chp. 11 Ref. 42 (Cohen et al., 1969).
26. Ibid. Chp. 11 Ref. 76 (Goldsmith & Landau, 1968).
27. Ibid. Chp. 11 Ref. 103 (Hexter & Goldsmith, 1971).
28. Ibid. Chp. 11 Ref. 128 (Kuller et al., 1975).
29. Ibid. Chp. 11 Ref. 168 (Radford, 1975).
30. Ibid. Chp. 11 Ref. 128a (Kurt, 1978).
31. CASAC Transcript, pp. 141-143, 186-190.
32. Criteria Document Chp. 11 Ref. 67 (Fodor & Winneke, 1972)
33. Ibid. Chp. 11 Ref. 85 (Groll-Knapp etal., 1972).
34. Ibid. Chp. 11 Ref. 19 (Beard & Grandstaff, 1975).
35. Ibid. Chp. 11 Ref. 107 (Horvath et al., 1971).
36. Ibid.. Chp. 11 Ref. 95 (Haider et al., 1975).
37. Ibid. Chp. 11 Ref. 217 (Winneke, 1974).
38. Ibid. Chp. 11 Ref. 218 (Winneke et al., 1976).
39. Ibid. Chp. 11 Ref. 39 (Christensen et al., 1977).
40. Ibid. Chp. 11 Ref. 23 (Benignus & Otto, 1977).
41. Ibid. Chp. 11 Ref 167 (Putz et al.f 1976).
42. Ibid. Chp. 11 Ref. 1.42 (McFarland et al., 1944).
43. Ibid. Chp. 11 Ref. 97 (Halperin et al., 1959).
44. Ibid. Chp. 11 Ref. 21 (Bender et al., 1972).
45. Ibid. Chp. 11 Ref. 22 (Bender et al., 1971).
46. Ibid. Chp. 11 Ref. 186 (Schulte, 1973).
-------
31
REFERENCES
47. Ibid. Chp. 11 Ref. 153 (O'Oonnell at a!., 1971)
48. Ibid. Chp. 11 Ref. 141 (McFarland, 1973).
49. Ibid. Chp. 11 Ref. 143 (Mcfarland at al., 1972)
50. Ibid. Chp. 11 Ref. 183 (Salvatore, 1974).
51. Ibid. Chp. 11 Ref. 114 (Johnson et al., 1974).
52. Ibid. Chp. 11 Ref. 115 (Johnson et al., 1976).
53. Ibid. Chp. 11 Ref. 219 (Wright et al., 1973).
54. Ibid. Chp. 11 Ref. 175 (Ray 4 Rockwell, 1970).
55. Ibid. Chp. 11 Ref. 177 (Rockwell & Weir, 1975).
56. Ibid- Chp. 11 Ref. 173 (Rockwell & Ray, 1967).
57. Ibid. Chp. 11 Ref. 215 (Weir & Rockwell, 1973).
53. Ibid. Chp. 11 Ref. 131 (Rumno &. Sarlanis, 1974).
59. Ibid. Chp. 11 Ref. 220 (Yabroff at al., 1974).
60. Ibid. Chp. 11 Ref. 31 (Brlnkhaus, 1977).
61. Ibid. Chp. 11 Ref. 3S (Chevalier et al., 1966).
62. Ibid.- Chp. 11 Ref. 60 (Ekblora et al., 1975).
63. Ibid. Chp. 11 Ref. 73 (GUner et al., 1975).
64. Ibid. Chp. 11 Ref. 89 (Guillenn at al., 1963).
65. Ibid. Chp. 11 Ref. 207 (Vogel 4 Slesar, 1972).
66. Ibid. Chp. 11 Ref. 208 (Vogel et al., 1972).
67. Ibid. Chp. 11 Ref. 59 (Ekblom £ Huot, 1972).
68. Ibid. Chp. 11 Ref. 155 (Nielsen, 1971).
69. Ibid. Chp. 11 Ref. 163 (Pirnay at al., 1971).
-------
32
REFERENCES
70. Ibid. Chp. 11 Ref. 50 (Dahms et al., 1975).
71. Ibid. Chp. 11 Ref. 36 (Chiodi et al., 1941).
72. Ibid. Chp. 11 Ref. 45 (Collier et al., 1972).
73. Ibid. Chp. 11 Ref. 108 (Horvath et al., 1975).
74. Ibid. Chp. 11 Ref. 3 (Aronow & Cassidy, 1975).
75. Ibid. Chp. 11 Ref. 187a (Seppanen, 1977).
c?05~(Mac/"Oi;.W|. :a,^}
76. Ibid. Chp. 11 Ref. ^(GuWMiOhj 1070).
77. Ibid. Chp. 11 Ref. 4 (Aronow et al., 1977).
78. Ibid. Chp. 10 Ref. 27 (Dykyk et al., 1975).
79. Ibid. Chp. 10 Ref. 29 (Dyer et al.).
80. Ibid. Chp. 10 Ref. 32 (Fechter & Annau, 1976).
81. Ibid. Chp. 10 Ref. 33 (Fechter & Annau, 1977).
82. Ibid. Chp. 10 Ref. 36A (Ginsberg & Myers, 1974).
83. Ibid. Chp. 10 Ref. 36 (Scnwetz et al., 1975).
84. Ibid. Chp. 11 Ref. 13 (Astrup et al., 1972).
85. Ibid. Chp. 10 Ref. 55 (Longo & Hill, 1977).
86. Ibid. Chp. 11 Ref. 136 (Longo, 1977).
87. Ibid. Chp. 11 Ref. 138 (McMahon et al., 1965).
88. Ibid. Chp. 11 Ref. 136 (Longo, 1977) Ref. 161 (Mulcahy et al.,
1970).
89. Ibid. Chp. 11 Ref. 136 (Longo, 1977) Ref. 206 (Simpson, 1957).
90. Ibid. Chp. 11 Ref. 136 (Longo, 1977) Ref. 143 (Lowe, 1959).
91. Ibid. Chp. 11 Ref. 136 (Longo, 1977) Ref. 153 (Meredith, 1975).
92. Ibid. Chp. 11 Ref. 136 (Longo, 1977) Ref. 25 (Butler & Alberman,
1969)"!
-------
33
REFERENCES
93. Ibid. Chp. 11 Ref. 136 (Longo, 1977) Ref. 23 (Sutler et al.,
WT).
94. Ibid. Chp. 11 Ref. 136 (Longo, 1977) Ref. 213 (U.S. Public
Health Service, 1973).
95. Ibid. Chp. 11 Ref. 131 (Lichty et al. 1957).
96. New Mexico State Department of Health. "Sirthweight and Altitude."
New Mexico Department of Health, Albuquerque, NM, 1975, pp. 7-16.
97. Criteria Document, Chp. 11 Ref. 136 (Longo, 1977) Ref. 43 (Comstock et al.,
1971).
98. Ibid. Chp. 11 Ref. 136 (Longo, 1977) Ref. 156 (Meyer et al, 1975).
99. Ibid. Chp. 11 Ref. 136 (Longo, 1977) Ref. 157 (Meyer et al.,
15751.
100. Ibid. Chp. 11 Ref. 136 (Longo, 1977) Ref. 170 (Niswander & Gordon,
TSTZ).
101. Ibid. Chp. 11 Ref. 136 (Longo, 1977) Ref. 173 (Ontario Dept. of
Health, 1967).
102. Ibid. Chp. 11 Ref. 136 (Longo, 1977) Ref. 174 (Ontario Dept. of
Health, 1967).
103. Ibid. Chp. 11 Ref. 136 (Longo, 1977) Ref. 71 (Fedrick et al.,
T97T).
104. Ibid. Chp. 11 Ref. 136. (Longo, 1977) Ref. 55 (Davie et al., 1972).
105. Ibid. Chp. 11 Ref. 136 (Longo, 1977) Ref. 36 (Goldstein, 1972).
106. Ibid. Chp. 11 Ref. 136 (Longo, 1977) Ref. 27 (Butler & Goldstein,
T97I).
107. Knelson, John H. "General Population Morbidity Estimates from
Exacerbation of Angina Pectoris Related to Low-Level Carbon
Monoxide Exposure." EPA, Health Effects Research Laboratory,
August 1975.
108. CASAC transcript, pp. 150-151.
109. Criteria Document, p. 11-33.
110. U. S. Dept. of Health, Education, and Welfare (DHEW). Prevalence
of Chronic Circulatory Conditions, United States, 1970. OHEVf
Publication No. (HRA) 74-1511. Rockville, MO. 1973.
-------
34
REFERENCES
111. Lown, Bernard. "Sudden cardiac death: The major challenge con-
fronting contemporary cardiology." Amer. J. Cardiol. 43; 313, 1979.
112. Federal Register Vol. 36, Mo. 84, p. 8186 (April 30, 1971).
113. Criteria Document, Chp. 11 Ref. 18 (Beard & Wertheim, 1967).
114. Criteria Document, pp. 11-9 and 11-10.
115. Ibid. Chp. 11 Ref. 159 (O'Donnell et al., 1971).
116. Ibid. Chp. 11 Ref. 193 (Stewart et al., 1973).
117. Ibid. Chp. 11 Ref. 197 (Stewart et al., 1970).
118. Ibid. Chp. 11 Ref. 161a (Otto et al., 1978).
119. Ibid, p. 11-83.
120. Ibid, p. 11-28.
121. CASAC transcript, pp. 182-191.
122. Criteria Document Chp. 9 Ref. 12 (Coburn et al., 1965).
123. Ibid, p. 11-52.
124. Ibid. Chp. 9 Ref. 21 (Douglas et al., 1912).
125. Ibid. Chp. 9 Ref. 51 (Peterson & Stewart, 1975).
126. Ibid., Chp. 11 Ref. 41.
127. Ibid.. Chp. 11 Ref 134.
128. Ibid, pp. 1-5, 6-46 to 6-53, and 11-77 to 11-81.
123a. Larsen, R.I. A Mathematical Model for Relating Air Quality
Measurements to Air Quality Standards. U.S. EPA. Office of
Air Programs Publication No. AP-89. 1973.
129. Federal Register Vol. 43, No. 121, p. 26967 (June 22, 1978).
130. Criteria Document Chp. 11 Ref. 216 (Weiser et al., 1978).
131. Ibid. Chp. 11 Ref. 8 (Aronow et al., 1974).
------- |