United States
Environmental Protection
Agency
Environmental Criteria and
Assessment Office
Research Triangle Park, NC 27711
EPA-600/8-83/028cF
June 1986
xvEPA
Research and Development
Air Quality
Criteria for Lead
Volume III of IV
-------
ABSTRACT
The document evaluates and assesses scientific information on the health
and welfare effects associated with exposure to various concentrations of lead
in ambient air. The literature through 1985 has been reviewed thoroughly for
information relevant to air quality criteria, although the document is not
intended as a complete and detailed review of all literature pertaining to
lead. An attempt has been made to identify the major discrepancies in our
current knowledge and understanding of the effects of these pollutants.
Although this document is principally concerned with the health and
welfare effects of lead, other scientific data are presented and evaluated in
order to provide a better understanding of this pollutant in the environment.
To this end, the document includes chapters that discuss the chemistry and
physics of the pollutant; analytical techniques; sources, and types of
emissions; environmental concentrations and exposure levels; atmospheric
chemistry and dispersion modeling; effects on vegetation; and respiratory,
physiological, toxicological, clinical, and epidemiological aspects of human
exposure.
m
-------
CONTENTS
VOLUME I
Chapter 1. Executive Summary and Conclusions 1"!
VOLUME II
Chapter 2. Introduction 2-1
Chapter 3. Chemical and Physical Properties 3-1
Chapter 4. Sampling and Analytical Methods for Environmental Lead 4-1
Chapter 5. Sources and Emissions 5-1
Chapter 6. Transport and Transformation 6-1
Chapter 7. Environmental Concentrations and Potential Pathways to Human Exposure .. 7-1
Chapter 8. Effects of Lead on Ecosystems 8-1
VOLUME III
Chapter 9. Quantitative Evaluation of Lead and Biochemical Indices of Lead
Exposure in Physiological Media 9~1
Chapter 10. Metabolism of Lead 10"1
Chapter 11. Assessment of Lead Exposures and Absorption in Human Populations 11-1
Volume IV
Chapter 12. Biological Effects of Lead Exposure l"\'"*
Chapter 13. Evaluation of Human Health Risk Associated with Exposure to Leaa
and Its Compounds
-------
TABLE OF CONTENTS
9 QUANTITATIVE EVALUATION OF LEAD AND BIOCHEMICAL INDICES OF LEAD EXPOSURE
IN PHYSIOLOGICAL MEDIA 9-1
9.1 INTRODUCTION 9-1
9.2 DETERMINATIONS OF LEAD IN BIOLOGICAL MEDIA 9-2
9.2.1 Sampling and Sample Handling Procedures for Lead
in Biological Media 9-2
9.2.1.1 Blood Sampling 9-3
9.2.1.2 Urine Sampling 9-4
9.2.1.3 Hair Sampling 9-4
9.2.1.4 Mineralized Tissue 9-5
9.2.1.5 Sample Handling in the Laboratory 9-5
9.2.2 Methods of Lead Analysis 9-6
9.2.2.1 Lead Analysis in Whole Blood 9-7
9.2.2.2 Lead in Plasma 9-11
9.2.2.3 Lead in Teeth 9-12
9.2.2.4 Lead in Hair 9-13
9.2.2.5 Lead in Urine 9-14
9.2.2.6 Lead in Other Tissues 9-15
9.2.3 Quality Assurance Procedures in Lead Analysis 9-16
9.3 DETERMINATION OF ERYTHROCYTE PORPHYRIN (FREE ERYTHROCYTE
PROTOPORPHYRIN, ZINC PROTOPORPHYRIN) 9-20
9.3.1 Methods of Erythrocyte Porphyrin Analysis 9-20
9.3.2 Interlaboratory Testing of Accuracy and Precision in
EP Measurement 9-23
9.4 MEASUREMENT OF URINARY COPROPORPHYRIN 9-25
9.5 MEASUREMENT OF DELTA-AMINOLEVULINIC ACID DEHYDRASE ACTIVITY 9-25
9.6 MEASUREMENT OF DELTA-AMINOLEVULINIC ACID IN URINE AND OTHER MEDIA 9-27
9.7 MEASUREMENT OF PYRIMIDINE-5'-NUCLEOTIDASE ACTIVITY 9-29
9.8 MEASUREMENT OF PLASMA 1,25-DIHYDROXYVITAMIN D 9-30
9.9 SUMMARY 9-31
9.9.1 Determinations of Lead in Biological Media 9-32
9.9.2 Determination of Erythrocyte Porphyrin (Free Erythrocyte
Protoporphyrin, Zinc Protoporphyrin) 9-35
9.9.3 Measurement of Urinary Coproporphyrin 9-36
9.9.4 Measurement of Delta-Ami no!evulinic Acid Dehydrase Activity 9-36
9.9.5 Measurement of Delta-Aminolevulinic Acid in Urine and Other Media ... 9-37
9.9.6 Measurement of Pyrimidine-5'-Nucleotidase Activity 9-38
9.9.7 Measurement of Plasma 1,25-Dihydroxyvitamin D 9-38
9.10 REFERENCES 9-39
10. METABOLISM OF LEAD 10-1
10.1 INTRODUCTION 10-1
10. 2 LEAD ABSORPTION IN HUMANS AND ANIMALS 10-1
10.2.1 Respiratory Absorption of Lead 10-1
10.2.1.1 Human Studies 10-2
10.2.1.2 Animal Studies 10-6
10.2.2 Gastrointestinal Absorption of Lead 10-6
10.2.2.1 Human Studies 10-6
10.2.2.2 Animal Studies 10-10
10.2.3 Percutaneous Absorption of Lead 10-13
10.2.4 Transplacental Transfer of Lead 10-14
-------
TABLE OF CONTENTS (continued).
10.3 DISTRIBUTION OF LEAD IN HUMANS AND ANIMALS 10-14
10.3.1 Lead in Blood 10-15
10.3.2 Lead Levels in Tissues 10-19
10.3.2.1 Soft Tissues 10-20
10.3.2.2 Mineralizing Tissue 10-23
10.3.3 Chelatable Lead 10-24
10.3.4 Mathematical Descriptions of Physiological Lead Kinetics 10-26
10.3.5 Animal Studies 10-31
10.4 LEAD EXCRETION AND RETENTION IN HUMANS AND ANIMALS 10-32
10.4.1 Human Studies 10-32
10.4.2 Animal Studies 10-38
10.5 INTERACTIONS OF LEAD WITH ESSENTIAL METALS AND OTHER FACTORS 10-41
10.5.1 Human Studies 10-41
10.5.2 Animal Studies 10-44
10.5.2.1 Interactions of Lead with Calcium 10-44
10.5.2.2 Interactions of Lead with Iron 10-48
10.5.2.3 Lead Interactions with Phosphate 10-48
10.5.2.4 Interactions of Lead with Vitamin D 10-49
10. 5. 2. 5 Interactions of Lead with Lipids 10-49
10.5.2.6 Lead Interaction with Protein 10-50
10.5.2.7 Interactions of Lead with Milk Components 10-50
10.5.2.8 Lead Interactions with Zinc and Copper 10-50
10.6 INTERRELATIONSHIPS OF LEAD EXPOSURE, EXPOSURE INDICATORS,
AND TISSUE LEAD BURDENS 10-51
10.6.1 Temporal Characteristics of Internal Indicators
of Lead Exposure 10-52
10.6.2 Biological Aspects of External Exposure/Internal
Indicator Relationships 10-53
10.6.3 Internal Indicator/Tissue Lead Relationships 10-54
10. 7 METABOLISM OF LEAD ALKYLS 10-57
10.7.1 Absorption of Lead Alkyls in Humans and Animals 10-57
10.7.1.1 Gastrointestinal Absorption 10-57
10.7.1.2 Percutaneous Absorption of Lead Alkyls 10-58
10.7.2 Biotransformation and Tissue Distribution of Lead Alkyls 10-58
10.7.3 Excreti on of Lead Alky1s 10-59
10.8 SUMMARY 10-60
10.8.1 Lead Absorption in Humans and Animals 10-60
10.8.1.1 Respiratory Absorption of Lead 10-60
10.8.1.2 Gastrointestinal Absorption of Lead 10-61
10.8.1.3 Percutaneous Absorption of Lead 10-62
10.8.1.4 Transplacental Transfer of Lead 10-62
10.8.2 Distribution of Lead in Humans and Animals 10-62
10.8.2.1 Lead in Blood 10-62
10.8.2.2 Lead Levels in Tissues 10-63
10.8.2.2.1 Soft Tissues !. 10-63
10.8.2.2.2 Mineralizing Tissue 10-64
10.8.2.2.3 Chelatable Lead 10-65
10.8.2.2.4 Animal Studies 10-65
10.8.3 Lead Excretion and Retention in Humans and Animals 10-66
10.8.3.1 Human Studies 10-66
-------
TABLE OF CONTENTS (continued).
10.8.3.2 Animal Studies 10-67
10.8.4 Interactions of Lead with Essential Metals and Other Factors 10-67
10.8.4.1 Human Studies 10-67
10.8.4.2 Animal Studies 10-67
10.8.5 Interrelationships of Lead Exposure with Exposure Indicators
and Tissue Lead Burdens 10-68
10.8.5.1 Temporal Characteristics of Internal Indicators of
Lead Exposure 10-69
10.8.5.2 Biological Aspects of External Exposure/Internal
Indicator Relationships 10-69
10.8.5.3 Internal Indicator/Tissue Lead Relationships 10-69
10.8.6 Metabolism of Lead Alkyls 10-70
10.8.6.1 Absorption of Lead Alkyls in Humans and Animals 10-70
10.8.6.2 Biotransformation and Tissue Distribution of
Lead Alkyls 10-71
10.8.6.3 Excretion of Lead Alkyls 10-71
10. 9 REFERENCES 10-72
11. ASSESSMENT OF LEAD EXPOSURES AND ABSORPTION IN HUMAN POPULATIONS 11-1
11.1 INTRODUCTION 11-1
11.2 METHODOLOGICAL CONSIDERATIONS 11-4
11.2.1 Analytical Problems 11-4
11.2.2 Statistical Approaches 11-5
11.2.3 Confounding of Relevant Variables 11-6
11. 3 LEAD IN HUMAN POPULATIONS 11-8
11.3.1 Introduction 11-8
11.3.2 Ancient and Remote Populations 11-8
11.3.2.1 Ancient Populations 11-10
11.3.2.2 Remote Populations 11-13
11.3.3 Levels of Lead and Demographic Covariates in U.S. and Other
Populations 11-14
11.3.3.1 The NHANES II Study 11-14
11.3.3.2 The Childhood Blood Lead Screening Programs 11-20
11.3.3.3 Levels of Lead and Demographic Covariates Worldwide 11-24
11.3.4 Distributional Aspects of Population Blood Lead Levels 11-24
11 3 5 Time Trends in Blood Lead Levels Since 1970 11-31
11.3.5.1 Time Trends in NHANES II Study Data 11-31
11.3.5.2 Time Trends in the Childhood Lead Poisoning Screening
Programs 11-34
11.3.5.3 Newark 11-37
11.3.5.4 Boston 11-37
11.3.5.5 Lead Studies in the United Kingdom 11-40
11.3.5.6 Other Studies 11-41
11 3.6 Gasoline Lead as an Important Determinant of Trends in Blood
Lead Levels 11-42
11.3.6.1 NHANES II Study Data 11-42
11.3.6.2 Isotope Studies 11-45
11.3.6.2.1 Italy 11-45
11.3.6.2.2 United States 11-52
11.3.6.3 Studies of Childhood Blood Lead Poisoning Control
Programs 11-55
11.3.6.4 Frankfurt, West Germany 11-60
vii
-------
TABLE OF CONTENTS (continued).
11.4 STUDIES RELATING EXTERNAL DOSE TO INTERNAL EXPOSURE 11-63
11.4.1 Air Studies 11-66
11.4.1.1 The Griffin et al. Study 11-67
11.4.1.2 The Rabinowitz et al. Study 11-71
11.4.1.3 The Chamberlain et al. Study 11-74
11.4.1.4 The Kehoe Study 11-76
11.4.1.5 The Azar et al. Study 11-78
11.4.1.6 Silver Valley/Kellogg, Idaho Study 11-81
11.4.1.7 Omaha, Nebraska Studies 11-89
11.4.1.8 Roels et al. Studies 11-91
11.4.1.9 Other Studies Relating Blood Lead Levels to
Ai r Exposure 11-94
11.4.1.10 Summary of Blood Lead versus Inhaled Air Lead Relations .. 11-99
11.4.2 Dietary Lead Exposures Including Water 11-106
11.4.2.1 Lead Ingestion from Typical Diets 11-108
11.4.2.1.1 Ryu Study on Infants and Toddlers 11-108
11.4.2.1.2 Rabinowitz Infant Study 11-110
11.4.2.1.3 Rabinowitz Adult Study 11-111
11.4.2.1.4 Hubermont Study 11-111
11.4.2.1.5 Sherlock Studies 11-111
11.4.2.1.6 Central Directorate on Environmental
Pollution Study 11-114
11.4.2.1.7 Pocock Study 11-115
11.4.2.1.8 Thomas Study 11-119
11.4.2.1.9 Elwood Study 11-119
11.4.2.2 Lead Ingestion from Experimental Dietary Supplements 11-119
11.4.2.2.1 Kehoe Study '" 11-119
11.4.2.2.2 Stuik Study 11-120
11.4.2.2.3 Cools Study 11-122
11.4.2.2.4 Schlegel Study ' 11-122
11.4.2.2.5 Chamberlain Study '/. 11-122
11.4.2.3 Inadvertent Lead Ingestion From Lead Plumbing 11-122
11.4.2.3.1 Early Studies ''"' H-122
11.4.2.3.2 Moore Studies '.['/. H-124
11.4.2.3.3 Thomas Study 11-126
11.4.2.3.4 Worth Study ','.'. 11-127
11.4.2.4 Summary of Dietary Lead Exposures, Including Water 11-127
11.4.3 Studies Relating Lead in Soil and Dust to Blood Lead 11-134
11.4.3.1 Omaha, Nebraska Studies 11-134
11.4.3.2 Stark Study 11-134
11.4.3.3 The Silver Valley/Kellogg Idaho Study 11-137
11.4.3.4 Blood Lead Levels of Dutch City Children 11-137
11.4.3.5 Charney Study 11-138
11.4.3.6 Charleston Studies 11-141
11.4.3.7 Barltrop Studies 11-142
11.4.3.8 The British Columbia Studies 11-143
11.4.3.9 The Baltimore Charney Study: A Controlled Trial of
Household Dust Lead Reduction 11-145
11.4.3.10 Gallacher Study 11-146
11.4.3.11 Other Studies of Soil and Dusts 11-147
11.4. 3.12 Summary of Soi 1 and Dust Lead 11-151
11.4.4 Paint Lead Exposures 11-151
viii
-------
TABLE OF CONTENTS (continued).
11.5 SPECIFIC SOURCE STUDIES 11-161
11 5.1 Primary Smelter Populations 11-161
11.5.1.1 El Paso, Texas 11-161
11.5.1.2 CDC-EPA Study 11-163
11.5.1.3 Meza Valley, Yugoslavia 11-163
11.5.1.4 Kosovo Province, Yugoslavia 11-165
11.5.1.5 The Cavalleri Study 11-165
11.5.1.6 Hartwel1 Study 11-166
11.5.2 Battery Plants 11-166
11.5.3 Secondary Smelters 11-166
11.5.4 Secondary Exposure of Children 11-170
11.5.5 Miscellaneous Studies 11-177
11.5.5.1 Studies Using Indirect Measures of Air Exposure 11-177
11.5.5.1.1 Studies in the United States 11-177
11.5.5.1.2 British Studies 11-179
11.5.5.2 Miscellaneous Sources of Lead 11-181
11.6 SUMMARY AND CONCLUSIONS 11-183
11.7 REFERENCES 11-193
APPENDIX 11A 11A-1
APPENDIX 11B 11B-1
APPENDIX 11C 11C-1
IX
-------
LIST OF FIGURES
Figure Page
10-1 Effect of particle size on lead deposition rate in the lung 10-4
10-2 The curvilinear relationship of serum lead to blood lead 10-18
10-3 Schematic model of lead metabolism in infant baboons, with compartmental
transfer coefficients 10-28
10-4 A compartmental model for lead biokinetics with multiple pools for blood
lead 10-29
10-5 Fitting of nonlinear blood lead model to data of DeSilva (1981). Broken
line incorporates an intercept term of 0.25; solid line does not
incorporate intercept term 10-30
10-6 Renal clearance (ratio of urinary lead to blood lead) from (A) King et al.,
1979; (B) Williams et al., 1969; (C) Gross, 1981; (D) DeVoto and
Spinazzola, 1973; (E) Azar et al., 1975; (G) Chamberlain et al., 1978 10-35
11-1 Pathways of lead from the environment to and within man 11-3
11-2 Estimated lead concentrations in bones (ug/g) from 5500 years before
present (BP) to the present time 11-12
11-3 Geometric mean blood lead levels by race and age for younger children in
the NHANES II study. EPA calculations from data furnished by the National
Center for Health Statistics 11-19
11-4 Geometric mean blood lead values by race and age for younger children in
the New York City screening program (1970-1976) 11-23
11-5 Unweighted geometric mean blood lead level for male and female nonsmoking
teachers (ug/dl) for several countries 11-25
11-6 Histograms of blood lead levels with fitted lognormal curves for the
NHANES II study. All subgroups are white non-SMSA residents with family
incomes over $6000/year 11-28
11-7 Average blood lead levels of U.S. population aged 6 months-74 years,
United States February 1976-February 1980, based on dates of examination
of NHANES II examinees with blood lead determinations 11-32
11-8 Reduction in mean blood lead levels, according to race, sex, and age.
Data on sex and age are for whites 11-33
11-9 Time-dependence of blood lead levels for blacks, aged 25 to 36 months, in
New York City and Chicago 11-35
11-10 Modeled umbilical cord blood lead levels by date of sample collection for
i nfants i n Boston 11-38
11-11 Parallel decreases in blood lead values observed in the NHANES II study
and amounts of lead used in gasoline during 1976-1980 11-43
11-12 Change in 206Pb/207Pb ratios in petrol, airborne particulate and blood
from 1974 to 1984 1;L_47
11-13 Estimated direct and indirect contributions of lead in gasoline to blood
lead in Italian men based on EPA analysis of ILE data (Table 11-16) 11-51
11-14 Geometric mean blood lead levels of New York City children (aged 25-36
months) by ethnic group, and ambient air lead concentration versus
quarterly sampling period, 1970-1976 11-58
11-15 Geometric mean blood lead levels of New York City children (aged 25-36
months) by ethnic group, and estimated amount of lead present in gasoline
sold in New York, New Jersey and Connecticut versus quarterly samplinq
period, 1970-1976 11-59
11-16 Geometric mean blood levels for blacks and Hispanics in the 25- to 36-month
age group and rooftop quarterly averages for ambient city-wide lead
levels ' 11-61
-------
LIST OF FIGURES (continued).
Page
11-17 Time dependence of blood lead and gas lead for blacks, aged 25 to 36
months, in New York 11-62
11-18 Data plots for individual subjects as a function of time for Kehoe
subjects, as presented by Gross (1979) 11-77
11-19 Blood lead versus air lead relationships derived from Kehoe inhalation
studies: Linear relationship holds for low exposures, quadratic for high
exposures. 95 percent confidence bands are also shown 11-79
11-20 Monthly ambient air lead concentrations in Kellogg, Idaho, 1971
through 1975 11-83
11-21 Fitted equations to the Kellogg Idaho/Silver Valley, adjusted blood lead
data 11-88
11-22 Blood lead concentrations versus weekly lead intake for bottle-fed
infants 11-116
11-23 Mean blood lead for men grouped by first draw water concentration 11-118
11-24 Average blood lead levels, Phase I 11-121
11-25 Average blood lead levels, Phase II 11-121
11-26 Lead in blood (mean values and range) in volunteers. In the lower curve
the average daily lead dose of the exposed group is shown 11-123
11-27 Cube root regression of blood lead on first flush water lead. This shows
mean ± S.D. of blood lead for pregnant women grouped in 7 intervals of
first flush water lead 11-125
11-28 Relation of blood lead (adult female) to first flush water lead in combined
estates. (Numbers are coincidental points; 9 = 9 or more.) Curve a,
present data; curve b, data of Moore et al. (1979) 11-128
11-29 Cumulative distribution of lead levels in dwelling units 11-155
11-30 Correlations of children's blood lead levels with fractions of surfaces
within a dwelling having lead concentrations J2 mg/cm2 11-157
11-31 Arithmetic mean air lead levels by traffic volume, Dallas, 1976 11-178
11-32 Blood lead concentration and traffic density by sex and age, Dallas,
1976 11-180
11-33 Geometric mean blood lead levels by race and age for younger children
in the NHANES II study, and the Kellogg/Silver Valley and New York
Childhood Screening Studies 11-184
11B-1 Residual sum of squares for nonlinear regression models for Azar data
(N=149) 11B-2
11C-1 Individual values of blood Pb-206/207 ratio for subjects follow-up in Turin
(12 subjects) 11C-2
11C-2 Individual values of blood Pb-206/207 ratio for subjects follow-up in
Costagneto (4 subjects) 11C-3
11C-3 Individual values of blood Pb-206/207 ratio for subjects follow-up in
Duento and Fiano (6 subjects) 11C-3
11C-4 Individual values of blood Pb-206/207 ratio for subjects follow-up in Nole
and Santeno (9 subjects) 11C-4
11C-5 Individual values of blood Pb-206/207 ratio for subjects follow-up in Viu
(4 subjects) 11C-4
-------
LIST OF TABLES
Table Page
10-1 Deposition of lead in the human respiratory tract 10-3
10-2 Distribution of lead in brain regions of humans and animals 10-21
10-3 Daily lead excretion and retention data for adults and infants 10-34
10-4 Effect of nutritional factors on lead uptake in animals 10-45
11-1 Summary of Representative Studies of Past Exposures to Lead 11-11
11-2 NHANES II blood lead levels of persons 6 months-74 years, with weighted
arithmetic mean, standard error of the mean, weighted geometric mean,
median, and percent distribution, by race and age, United States,
1976-80 11-16
11-3 NHANES II blood lead levels of males 6 months-74 years, with weighted
arithmetic mean, standard error of the mean, weighted geometric mean,
median, and percent distribution, by race and age, United States,
1976-80 11-17
11-4 NHANES II blood lead levels of females 6 months-74 years, with weighted
arithmetic mean, standard error of the mean, weighted geometric mean,
median, and percent distribution, by race and age, United States,
1976-80 11-18
11-5 Weighted geometric mean blood lead levels from NHANES II survey by
degree of urbanization of place of residence in the U.S. by age
and race, United States 1976-80 11-21
11-6 Annual geometric mean blood lead levels from the New York blood lead
screening studies of Billick et al. (1979). Annual geometric means
are calculated from quarterly geometric means estimated by the method of
Hasselblad et al. (1980) 11-22
11-7 Summary of unweighted blood lead levels in whites not living in an
SMSA, with family income greater than $6,000 11-26
11-8 Summary of fits to NHANES II blood lead levels of whites not
living in an SMSA, with income greater than $6,000, for five
different two-parameter distributions 11-27
11-9 Estimated mean square errors resulting from analysis of variance on
various subpopulations of the NHANES II data using unweighted data 11-30
11-10 Characteristics of childhood lead poisoning screening data 11-36
11-11 Distribution of blood lead levels for 13- to 48-month-old blacks
by season and year for New York screening data 11-36
11-12 Comparison of median blood lead levels (ug/dl) in several countries from
studies of Goldwater and Hoover (1967) and Friberg and Vahter (1983) 11-42
11-13 Pearson correlation coefficients between the average blood lead levels for
six-month periods and the total lead used in gasoline production per six
months, according to race, sex, and age 11-44
11-14 Estimated contribution of leaded gasoline to blood lead by inhalation and
non-inhalation pathways 11-49
11-15 Assumed air lead concentrations for model 11-50
11-16 Regression model for blood lead attributable to gasoline 11-51
11-17 Rate of change of 206pb/204P|? and 2oePb/207pb in air and blood, and
percentage of airborne lead in blood of subjects 1, 3, 5, 6 and 9 11-54
11-18 Calculated blood lead uptake from air lead using Manton isotope study 11-54
11-19 Respired and other inputs of airborne Pb to blood for some Dallas residents
i n 1975 H'56
11-20 Mean air lead concentrations during the various blood sampling periods at
the measurement sites described in the text (ug/m3) 11-63
11-21 Griffin et al. (1975) experiment inhalation slope estimates 11-70
11-22 Griffin et al. (1975) experiment mean residence time in blood 11-70
xii
-------
LIST OF TABLES (continued).
Table
11-28 foomtrlc mean blood lead levels by age and area for subjects living near
11-37 cenan
11-23 Air lead concentrations (pgA.3) for two subjects in the Rabinowitz studies ... 11-72
11-24 Estimates of inhalation slope, p, for Rabinowitz studies ........... ....... 11-73
11-25 Linear sloSe for blood lead versus air lead at low air lead exposures in
11-29 Agtspedf ,. R,
11-30 E^iiateS coe> nctn^anT sSffiri' errors' for' the' Idaho' s.el ter' study' ! ! ! ! ! .' ! ll^
U-31 A r d slfan anS Mood lead concentrations in Omaha NE study 1970-1977 . . . 11-90
ii-5? SI airborne and blood lead levels recorded during five distinct surveys
11 32 ??S?4 Jo 1978) for study populations of 11-year old children living less
than l km or 2 5 km from a lead smelter, or living in a rural or urban area .. 11-93
11-33 Geomelic mean air lead and adjusted blood lead levels for 11 communities
of Tepper and Levin (1975) as reported by Hasselblad and
n 14 pirndooea. 11-96
11-35 6?ood !ead-air lead slopes for several population studies as calculated
11-36 cLracLrUtics'of'studUs^n'the' relationship between'air'Uad'and'biood'''^
11-38 Cross-sectional observational 'study with measured individual air lead ^^
11-39 c^-sectiona'l'observationai' studies' on' chi idren' with' estimated ....... '""" ^^
11-40 LoSgifSdinal 'experimental 'studies with^measured individual air lead ^^
n-41 HoEsehold consumption' of 'canned foods, pounds per week ....................... 11-109
llll Slood lead ?!IelS and lead intake values for infants in the study ^^
11-43 Inf?uence of level of lead in water on blood lead level in biood and
^ *^ .»,*«***«**** i. j. H.c.
11-44 Distributions' of 'observed blood lead values in Ayr ........................... 11-113
11-45 Blood lead and kettle water lead concentrations for adult women living
. * ..... .....* ........... * ........... .......,*....... ........ 1.1. XJ.J
11-46 Relationship of blood lead and water lead in 910 men aged 40-59 from
24 British towns [[[ 11-117
11-47 Dose-response analysis for blood lead levels in the Kehoe study as
analyzed by Gross (1981) ........ : ....... . ...... ...................... ........ 11-120
11-48 Blood lead levels of 771 persons in relation to lead content of drinking
water Boston, MA [[[ 11-129
-------
LIST OF TABLES (continued).
Table
11-50 Studies involving blood lead levels (pg/dl) and experimental dietary
i ntakes 11-131
11-51 Studies relating blood lead levels (ug/dl) to first-flush water lead (pg/1) .. 11-132
11-52 Studies relating blood lead levels (pg/dl) to running water lead (yg/1) 11-133
11-53 Coefficients and standard errors for Omaha study model 11-135
11-54 Multiple regression models for blood lead of children in New Haven,
Connecticut, September 1974 - February 1977 11-136
11-55 Air Lead Levels in the Rotterdam Area 11-139
11-56 Blood lead levels in ug/lOG ml for children who participated in blood
survey and environmental survey 11-139
11-57 School variables (arithmetic means) for measured lead concentrations 11-139
11-58 Results of lead measurements reported by Brunekreef et al. (1983) 11-140
11-59 Coefficients and standard errors from model of Charleston study 11-142
11-60 Mean blood and soil lead concentrations in English study 11-143
11-61 Lead concentration of surface soil and children's blood by residential
area of trail, British Columbia 11-145
11-62 Analysis of relationship between soil lead and blood lead in children 11-150
11-63 Estimates of the contribution of soil lead to blood lead 11-152
11-64 Estimates of the contribution of housedust to blood lead in children 11-153
11-65 Results of screening and housing inspection in childhood lead poisoning
control project by fi seal year - 11-161
11-66 Mean blood lead levels in selected Yugoslavian populations, by estimated
weekly time-weighted air lead exposure 11-164
11-67 Levels of lead recorded in Hartwell et al. (1983) study 11-167
11-68 Spearman correlations of lead in air, water, dust, soil, and paint with
lead levels in blood: by site and age groups, 1978-1979 11-167
11-69 Environmental parameters and methods: Arnhem lead study, 1978 11-169
11-70 Geometric mean blood lead levels for children based on reported
occupation of father, history of pica, and distance of residence
from smelter (micrograms per deciliter) 11-171
11-71 Sources of lead -.: '*""*'. 11-182
11-72 Summary of blood lead pooled geometric standard deviations and estimated
analytic errors .; 11-185
11-73 Estimated contribution of leaded gasoline to blood lead by inhalation
and non-inhalation pathways 11-187
11-74 Summary of blood inhalation slopes, (p) ug/dl per pg/m-* 11-188
xiv
-------
tSST OF ABBREVIATIONS
AAS
Ach
ACTH
A0CC
ADP/0 ratio
AIOS
All
ALA
AU-D
ALA-S
ALA-U
APDC
APHA
ASTM
ASV
ATP
B-cells
8a
BAL
BAP
6SA
BUN
BW
C.V.
Ca8P
CaEDTA
CaNa,EOTA
CBD *
Cd
COC
C£C
CEH
CFR
CMP
CNS
CO
COHb
CP8
CP-U
cBal)
D.F.
DA
6-ALA
OCMU
DPP
ONft
DTH
E£
E£G
EMC
Atomic absorption spectrometry
Acetylcholine
Adrenocorticctrophic hormone
Antibody-dependent eel 1 -medi ated cytotox1C1 ty
Adenosine diphosphate/oxygen «**'«
Acquired immune deficiency syndrome
American Industrial Hygiene Association
Angiotensin II
Amim>1evuHnic acio
AmiRolevulinic acid dehydrase
AminolevuHnic acid synthetase
ulinic acid in urine
pyrrol idine-dithiocarbamate
Wlc Health Association _
can Society for Testing and Hatenals
Anodic stripping voltawetry
Adenosine triphosphate
Bone marrow-derived ly^hocytes
BrHish anti-lewisite (AKA dimercaprol)
benio{a)pyrene
Bovine serum albw"
glood serum urea nitrogen
Body weight
Coefficient of vacation
Calciuw binding protein
fal ci urn ethy lenedi ami netetraacetate
Calcitffl! sodiuffl ethy lenedi ami netetraacetate
Central business district
aflt ,, . ,
Centers for Disease Control
Cation exchange capacity
Center for Environmental Health
reference method
Cytidine monophosphate
Central nervous system
Carbon monoxide
CarboxyhemogloiJi n
Cawpetitfve protein binding
Urinary coproporphyrin
plasma clearance of p-asinohippuric acid
Copper
Degrees of freedom
Dopamine
delta-aminolevulinic acid
[3-(3,4-dichlorophenyl)-l,l-dimethylurea
Differential pulse polarography
Oeoxyribonucleic acid
Oelayed-type hypersensitivity
European Economic Community
E 1 ec troencepha 1 ograro
Encephalomyocardi ti s
-------
LIST OF ABBREVIATIONS (continued).
EP Erythrocyte protoporphyrin
EPA U.S. Environmental Protection Agency
FA Fulvic acid
FDA Food and Drug Administration
Fe Iron
FEP Free erythrocyte protoporphyrin
FY Fiscal year
G.M. Grand mean
G-6-PD Glucose-6-phosphate dehydrogenase
GABA Gamma-aminobutyric acid
GALT Gut-associated lymphoid tissue
GC Gas chromatography
GFR Glomerular filtration rate
GI Gastrointestinal
HA Humic acid
HANES I Health Assessment and Nutrition Evaluation Survey
Hb Hemoglobin
Hg Mercury
hi-vol High-volume air sampler
HPLC High-performance liquid chromatography
i.m. Intramuscular (method of injection)
i.p. Intraperitoneally (method of injection)
i.v. Intravenously (method of injection)
IAA Indol-3-ylacetic acid
IARC International Agency for Research on Cancer
ICD International classification of diseases
ICP Inductively coupled plasma emission spectroscopy
IDMS Isotope dilution mass spectrometry
IP Interferon
Isotopic Lead Experiment (Italy)
International Radiological Protection Commission
Potassium
Lactate dehydrogenase isoenzyme x
Lethyl concentration (50 percent)
Lethal dose (50 percent)
Luteinizing hormone
Laboratory Improvement Program Office
In Natural logarithm
Lipopolysaccharide
Long range transport
Messenger ribonucleic acid
ME Mercaptoethanol
Miniature end-plate potential
Maximal electroshock seizure
Mega-electron volts
MLC Mixed lymphocyte culture
Mass median diameter
Mass median aerodynamic diameter
Manganese
Motor neuron disease
Moloney sarcoma virus
Maximum tolerated dose
xvi
-------
LIST OF ABBREVIATIONS (continued).
n
N/A
NA
NAAQS
NAD
NADB
NAMS
NAS
NASN
NBS
NE
NFAN
NFR-82
NHANES II
Ni
NTA
OSHA
P
P
PAH
Pb
PBA
Pb(Ac)2
PbB
PbBrCl
PBG
PFC
pH
PHA
PHZ
PIXE
PMN
PND
PNS
P.O.
ppm
PRA
PRS
PWM
Py5N
RBC
RBF
RCR
redox
RES
RLV
RNA
S-HT
SA-7
S.C.
son
S.D.
Number of subjects or observations
Not Available
Not Applicable
National ambient air quality standards
Nicotinamide Adenine Dinucleotide
National Aerometric Data Bank
National Air Monitoring Station
National Academy of Sciences
National Air Surveillance Network
National Bureau of Standards
Norepinephrine
National Filter Analysis Network
Nutrition Foundation Report of 1982
National Health Assessment and Nutritional Evaluation Survey II
Nickel
Nitrilotriacetonitrile
Occupational Safety and Health Administration
Phosphorus
Significance symbol
Para-aminohippuric acid
Lead
Air lead
Lead acetate
concentration of lead in blood
Lead (II) bromochloride
Porphobilinogen
Plaque-forming cells
Measure of acidity
Phytohemagglutinin
Polyacrylamide-hydrous-zirconia
Proton-induced X-ray emissions
Polymorphonuclear leukocytes
Post-natal day
Peripheral nervous system
Per os (orally)
Parts per million
Plasma renin activity
Plasma renin substrate
Pokeweed mitogen
Pyrimide-5'-nuc1eotidase
Red blood cell; erythrocyte
Renal blood flow
Respiratory control ratios/rates
Oxidation-reduction potential
Reticuloendothelial system
Rauscher leukemia virus
Ribonucleic acid
Serotonin
Simian adenovirus
Subcutaneously (method of injection)
Standard cubic meter
Standard deviation
xv ii
-------
LIST OF ABBREVIATIONS (continued).
SOS
S.E.M.
SES
SCOT
slg
SLAMS
SMR
Sr
SRBC
SRMs
STEL
SW voltage
T-cells
t-tests
TBL
TEA
TEL
TIBC
TML
TMLC
TSH
TSP
U.K.
UMP
USPHS
VA
V^R
WHO
XRF
X^
Zn
ZPP
Sodium dodecyl sulfate
Standard error of the mean
Socioeconomic status
Serum glutamic oxaloacetic transaminase
Surface immunoglobulin
State and local air monitoring stations
Standardized mortality ratio
Strontium
Sheep red blood cells
Standard reference materials
Short-term exposure limit
Slow-wave voltage
Thymus-derived lymphocytes
Tests of significance
Tri-n-butyl lead
Tetraethyl-ammonium
Tetraethyllead
Total iron binding capacity
Tetramethyllead
Tetramethyllead chloride
Thyroid-stimulating hormone
Total suspended particulate
United Kingdom
Uridine monophosphate
U.S. Public Health Service
Veterans Administration
Deposition velocity
Visual evoked response
World Health Organization
X-Ray fluorescence
Chi squared
Zinc
Erythrocyte zinc protoporphyrin
MEASUREMENT ABBREVIATIONS
dl
ft
g
g/gal
g/ha-mo
km/hr
1/min
mg/km
|jg/m3
mm
)jm
(jmol
ng/cm2
nm
deciliter
feet
gram
gram/gallon
gram/hectare-month
kilometer/hour
liter/minute
milligram/kilometer
microgram/cubic meter
millimeter
micrometer
micromole
nanograms/square centimeter
nanometer
xvm
-------
LIST OF ABBREVIATIONS (continued).
nM nanomole
sec second
t tons
xix
-------
GLOSSARY VOLUME III
aerosol - a suspension of liquid or solid particles in a gas
BAL (British Anti-Lewi site) - a chelating agent often used in the treatment of
metal toxicity
biliary clearance - an excretion route involving movement of an aqent throuah
bile into the GI tract y
Brownian diffusion - the random movement of microscopic particles
"chelatable" or systemically active zinc - fraction of body's zinc store
available or accessible to
removal by a zinc-binding agent
chi-square goodness-of-fit tests - made to determine how well the observed
data fit a specified model, these tests
usually are approximately distributed as a
chi-square variable
first-order kinetics - a kinetic process whose rate is proportional to the
concentration of the species undergoing change
geochronometry - determination of the age of geological materials
hematocrit - the percentage of the volume of a blood sample occupied by cells
intraperitoneal - within the body cavity
likelihood function - a relative measure of the fit of observed data to a
specified model. In some special cases it is equivalent
to the sum of squares function used in least squares
analysis.
mass median aerodynamic diameter (MMAD) - the aerodynamic diameter (in urn) at
which half the mass of particles in
an aerosol is associated with values
below and half above
multiple regression analysis - the fitting of a single dependent variable to a
linear combination of independent variables using
least squares analysis
plumburesis - lead excreted in urine
R2 - this statistic, often called the multiple R squared, measures the proportion
of total variation explained. A value near 1 means that nearly all of the
variation is explained, whereas a value near zero means that almost none of
the variation is explained.
xx
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
Chapter 9: Quantitative Evaluation of Lead and Biochemical Indices of Lead
Exposure in Physiological Media
Principal Author
Or. Paul Mushak
Department of Pathology
School of Medicine
University of North Carolina
Chapel Hill, NC 27514
The following persons reviewed this chapter at EPA's request. The evaluations
and conclusions contained herein, however, are not necessarily those of the
reviewers.
Dr. Carol Angle
Department of Pediatrics
University of Nebraska
College of Medicine
Omaha, NE 68105
Dr. Lee Annest
Division of Health Examin. Statistics
National Center for Health Statistics
3700 East-West Highway
Hyattsville, MD 20782
Dr. Donald Barltrop
Department of Child Health
Westminister Children's Hospital
London SW1P 2NS
England
Dr. Irv Billick
Gas Research Institute
8600 West Bryn Mawr Avenue
Chicago, IL 60631
Dr. Joe Boone
Clinical Chemistry and
Toxicology Section
Centers for Disease Control
Atlanta, GA 30333
Dr. Robert Bornschein
University of Cincinnati
Kettering Laboratory
Cincinnati, OH 45267
Dr. A. C. Chamberlain
Environmental and Medical
Sciences Division
Atomic Energy Research
Establishment
Harwell 0X11
England
Dr. Neil Chernoff
Division of Developmental Biology
MD-67
U.S. Environmental Protection
Agency
Research Triangle Park, NC 27711
Dr. Julian Chisolm
Baltimore City Hospital
4940 Eastern Avenue
Baltimore, MD 21224
Mr. Jerry Cole
International Lead-Zinc Research
Organization
292 Madison Avenue
New York, NY 10017
Dr. Max Costa
Department of Pharmacology
University of Texas Medical
School
Houston, TX 77025
Dr. Anita Curran
Commissioner of Health
Westchester County
White Plains, NY 10607
xxi
-------
Dr. Jack Dean
Immunobiology Program and
Immunotoxicology/Cell Biology program
CUT
P.O. Box 12137
Research Triangle Park, NC 27709
Dr. H. T. Delves
Chemical Pathology and Human
Metabolism
Southampton General Hospital
Southampton S09 4XY
England
Dr. Fred deSerres
Assoc. Director for Genetics
NIEHS
P.O. Box 12233
Research Triangle Park, NC 27709
Dr. Robert Dixon
Laboratory of Reproductive and
Developmental Toxicology
NIEHS
P.O. Box 12233
Research Triangle Park, NC 27709
Dr. Claire Ernhart
Department of Psychiatry
Cleveland Metropolitan General Hospital
Cleveland, OH 44109
Dr. Sergio Fachetti
Section Head - Isotope Analysis
Chemistry Division
Joint Research Center
121020 Ispra
Varese, Italy
Dr. Virgil Ferm
Department of Anatomy and Cytology
Dartmouth Medical School
Hanover, NH 03755
Dr. Alf Fischbein
Environmental Sciences Laboratory
Mt. Sinai School of Medicine
New York, NY 10029
Dr. Jack Fowle
Reproductive Effects Assessment Group
U.S. Environmental Protection Agency
RD-689
Washington, DC 20460
Dr. Bruce Fowler
Laboratory of Pharmacology
NIEHS
P.O. Box 12233
Research Triangle Park, NC 27709
Dr. Warren Galke
Department of Biostatistics
and Epidemiology
School of Allied Health
East Carolina University
Greenville, NC 27834
Mr. Eric Goldstein
Natural Resources Defense
Council, Inc.
122 E. 42nd Street
New York, NY 10168
Dr. Harvey Gonick
1033 Gayley Avenue
Suite 116
Los Angeles, CA 90024
Dr. Robert Goyer
Deputy Director
NIEHS
P.O. Box 12233
Research Triangle Park, NC 27709
Dr. Stanley Gross
Hazard Evaluation Division
Toxicology Branch
U.S. Environmental Protection
Agency
Washington, DC 20460
Dr. Paul Hammond
University of Cincinnati
Kettering Laboratory
Cincinnati, OH 45267
Dr. Ronald D. Hood
Department of Biology
The University of Alabama
University, AL 35486
Dr. V. Houk
Centers for Disease Control
1600 Clifton Road, NE
Atlanta, GA 30333
xxi i
-------
Dr. Loren D. Roller
School of Veterinary Medicine
University of Idaho
Moscow, ID 83843
Dr. Kristal Kostial
Institute for Medical Research
and Occupational Health
Yu-4100 Zagreb
Yugoslavia
Dr. Lawrence Kupper
Department of Biostatistics
UNC School of Public Health
Chapel Hill, NC 27514
Dr. Phillip Landrigan
Division of Surveillance,
Hazard Evaluation and Field Studies
Taft Laboratories - NIOSH
Cincinnati, OH 45226
Dr. David Lawrence
Microbiology and Immunology Dept,
Albany Medical College of Union
University
Albany, NY 12208
Dr. Jane Lin-Fu
Office of Maternal and Child Health
Department of Health and Human Services
Rockville, MD 20857
Dr. Don Lynam
Air Conservation
Ethyl Corporation
451 Florida Boulevard
Baton Rouge, LA 70801
Dr. Kathryn Mahaffey
Division of Nutrition
Food and Drug Administration
1090 Tusculum Avenue
Cincinnati, OH 45226
Dr. Ed McCabe
Department of Pediatrics
University of Wisconsin
Madison, WI 53706
Dr. Chuck Nauman
Exposure Assessment Group
U.S. Environmental Protection
Agency
Washington, DC 20460
Dr. Herbert L. Needleman
Department of Psychiatry
Children's Hospital of Pittsburgh
Pittsburgh, PA 15213
Dr. H. Mitchell Perry
V.A. Medical Center
St. Louis, MO 63131
Dr. Jack Pierrard
E.I. duPont de Nemours and
Company, Inc.
Petroleum Laboratory
Wilmington, DE 19898
Dr. Sergio Piomelli
Columbia University Medical School
Division of Pediatric Hematology
and Oncology
New York, NY 10032
Dr. Magnus Piscator
Department of Environmental Hygiene
The Karolinska Institute 104 01
Stockholm
Sweden
Dr. Robert Putnam
International Lead-Zinc
Research Organization
292 Madison Avenue
New York, NY 10017
Dr. Michael Rabinowitz
Children's Hospital Medical
Center
300 Longwood Avenue
Boston, MA 02115
XX111
-------
Dr. Harry Roels
Unite de Toxicologic
Industrielle et Medicale
Universite de Louvain
Brussels, Belgium
Dr. John Rosen
Division of Pediatric Metabolism
Albert Einstein College of Medicine
Montefiore Hospital and Medical Center
111 East 210 Street
Bronx, NY 10467
Dr. Michael Rutter
Department of Psychology
Institute of Psychiatry
DeCrespigny Park
London SE5 SAL
England
Dr. Stephen R. Schroeder
Division for Disorders
of Development and Learning
Biological Sciences Research Center
University of North Carolina
Chapel Hill, NC 27514
Dr. Anna-Maria Seppalainen
Institutes of Occupational Health
Tyoterveyslaitos
Haartmaninkatu I
00290 Helsinki 29
Finland
Or. Ellen Silbergeld
Environmental Defense Fund
1525 18th Street, NW
Washington, DC 20036
Dr Ron Snee
E.I. duPont Nemours and
Company, Inc.
Engineering Department L3167
Wilmington, OE 19898
Dr. Gary Ter Haar
Toxicology and Industrial
Hygiene
Ethyl Corporation
451 Florida Boulevard
Baton Rouge, LA 70801
Dr. Ian von Lindern
Department of Chemical Engineering
University of Idaho
Moscow, Idaho 83843
Or. Richard P. Wedeen
V.A. Medical Center
Tremont Avenue
East Orange, MJ 07019
xxiv
-------
Chapter 10: Metabolism of Lead
Principal Author
Or. Paul Mushak
Department of Pathology
School of Medicine
University of North Carolina
Chapel Hill, NC 27514
Contributing Author
Dr. Alan Marcus
Department of Mathematics
Washington State University
Pullman, WA 99164-2930
Th
persons reviewed this chapter at EPA's request. The evaluations
. ~nn+9-inari hprein. however, are not necessarily those of the
) rmriii^nns contained herein, however^
reviewers?
Dr. Carol Angle
Department of Pediatrics
University of Nebraska
College of Medicine
Omaha, NE 68105
Dr. Lee Annest
Division of Health Examin. Statistics
National Center for Health Statistics
3700 East-West Highway
Hyattsville, MD 20782
Dr. Donald Barltrop
Department of Child Health
Westminister Children's Hospital
London SW1P 2NS
England
Dr. Irv Billick
Gas Research Institute
8600 West 8ryn Mawr Avenue
Chicago, IL 60631
Or. Joe Boone
Clinical Chemistry and
Toxicology Section
Centers for Disease Control
Atlanta, GA 30333
Dr. Robert Bornschein
University of Cincinnati
Kettering Laboratory
Cincinnati, OH 45267
Dr. A. C. Chamberlain
Environmental and Medical
Sciences Division
Atomic Energy Research
Establishment
Harwell 0X11
England
Dr. Neil Chernoff
Division of Developmental Biology
MD-67
U.S. Environmental Protection
Agency
Research Triangle Park, NC 27711
Dr. Julian Chisolm
Baltimore City Hospital
4940 Eastern Avenue
Baltimore, MD 21224
Mr. Jerry Cole
International Lead-Zinc Research
Organization
292 Madison Avenue
New York, NY 10017
XXV
-------
Dr. Max Costa
Department of Pharmacology
University of Texas Medical School
Houston, TX 77025
Dr. Anita Curran
Commissioner of Health
Westchester County
White Plains, NY 10607
Dr. Jack Dean
Immunobiology Program and
Immunotoxioology/Cell Biology program
CUT
P.O. Box 12137
Research Triangle Park, NC 27709
Dr. H.T. Delves
Chemical Pathology and Human Metabolism
Southampton General Hospital
Southampton S09 4XY
England
Or. Fred deSerres
Assoc. Director for Genetics
NIEHS
P.O. Box 12233
Research Triangle Park, NC 27709
Dr. Robert Dixon
Laboratory of Reproductive and
Developmental Toxicology
NIEHS
P.O. Box 12233
Research Triangle Park, NC 27709
Dr. Claire Ernhart
Department of Psychiatry
Cleveland Metropolitan General Hospital
Cleveland, OH 44109
Dr. Sergio Fachetti
Section Head - Isotope Analysis
Chemistry Division
Joint Research Center
121020 Ispra
Varese, Italy
Dr. Virgil Ferm
Department of Anatomy and Cytology
Dartmouth Medical School
Hanover, NH 03755
Dr. Alf Fischbein
Environmental Sciences Laboratory
Mt. Sinai School of Medicine
New York, NY 10029
Dr. Jack Fowle
Reproductive Effects Assessment
Group
U.S. Environmental Protection
Agency
RD-689
Washington, DC 20460
Dr. Bruce Fowler
L?ruoatory of Pharmacology
NIEHS
P.O. Box 12233
Research Triangle Park, NC 27709
Dr. Warren Galke
Department of Biostatistics
and Epidemiology
School of Allied Health
East Carolina University
Greenville, NC 27834
Mr. Eric Goldstein
Natural Resources Defense
Council, Inc.
122 E. 42nd Street
New York, NY 10168
Dr. Harvey Gonick
1033 Gayley Avenue
Suite 116
Los Angeles, CA 90024
Dr. Robert Goyer
Deputy Director
NIEHS
P.O. Box 12233
Research Triangle Park, NC
27709
Dr. Stanley Gross
Hazard Evaluation Division
Toxicology Branch
U.S. Environmental Protection
Agency
Washington, DC 20460
Dr. Paul Hammond
University of Cincinnati
Kettering Laboratory
Cincinnati, OH 45267
xxvi
-------
Dr. Ronald D. Hood
Department of Biology
The University of Alabama
University, AL 35486
Dr. V. Houk
Centers for Disease Control
1600 Clifton Road, NE
Atlanta, GA 30333
Dr. Loren 0. Keller .
School of Veterinary Medicine
University of Idaho
Moscow, ID 83843
Dr. Kristal Kostial
Institute for Medical Research
and Occupational Health
Yu-41QO Zagreb
Yugoslavia
Dr. Lawrence Kupper
Department of Biostatistics
UNC School of Public Health
Chapel Hill, NC 27514
Dr Phillip Landrigan
Division of Surveillance,
Hazard Evaluation and Field Studies
Taft Laboratories - NIOMi
Cincinnati, OH 45226
Dr. David Lawrence
Microbiology and Immunology Dept.
Albany Medical College of Union
University
Albany, NY 12208
co and Child Health
artment of Health and Human Services
Rockville,
MD 20857
Or. Don Lynam
Air Conservation
Ethyl Corporation
451 Florida Boulevard
Baton Rouge, LA 70801
Dr, Kathryn Mahaffey
Division of Nutrition
Food and Drug Administration
1090 Tusculum Avenue
Cincinnati, OH 45226
xxvii
Dr. Ed McCabe
Department of Pediatrics
University of Wisconsin
Madison, WI 53706
Dr. Chuck Nauman
Exposure Assessment Group
U.S. Environmental Protection Agency
Washington, DC 20460
Dr. Herbert L. Neddleman
Department of Psychiatry
Children's Hospital of Pittsburgh
Pittsburgh, PA 15213
Dr. H. Mitchell Perry
V.A. Medical Center
St. Louis, MO 63131
Dr. Jack Pierrard
E.I. duPont de Nemours and
Company, Inc.
Petroleum Laboratory
Wilmington, DE 19898
Or. Sergio Piomelli
Columbia University Medical School
Division of Pediatric Hematology
and Oncology
New York, NY 10032
Dr. Magnus Piscator
Department of Environmental Hygiene
The Karolinska Institute 104 01
Stockholm
Sweden
Dr. Robert Putnam
International Lead-Zinc
Research Organization
292 Madison Avenue
New York, NY 10017
Dr. Harry Roels
Unite de Toxicologie
Industrie!le et Medicale
Universite de Louvain
Brussels, Belgium
Dr. John Rosen
Division of Pediatric Metabolism
Albert Einstein College of Medicine
Montefiore Hospital and Medical Center
111 East 210 Street
Bronx, NY 10467
-------
Dr. Michael Rutter
Department of Psychology
Institute of Psychiatry
DeCrespigny Park
London SE5 SAL
England
Dr. Stephen R. Schroeder
Division for Disorders
of Development and Learning
Biological Sciences Research Center
University of North Carolina
Chapel Hill, NC 27514
Dr. Anna-Maria Seppalainen
Institutes of Occupational Health
Tyoterveys1ai tos
Haartmaninkatu 1
00290 Helsinki 29
Finland
Dr. Ellen Silbergeld
Environmental Defense Fund
1525 18th Street, NW
Washington, DC 20036
Dr. Ron Snee
E.I. duPont Nemours and
Company, Inc.
Engineering Department L3167
Wilmington, DE 19898
Dr. Gary Ter Haar
Toxicology and Industrial
Hygiene
Ethyl Corporation
451 Florida Boulevard
Baton Rouge, LA 70801
Dr. Ian von Lindern
Department of Chemical
Engineering
University of Idaho
Moscow, ID 83843
Dr. Richard P. Wedeen
V.A. Medical Center
Tremont Avenue
East Orange, NJ 07019
xxviii
-------
Chapter 11: Assessment of Lead Exposures and Absorption in Human Populations
Principal Authors
Dr. Warren Galke
Department of Biostatistics and Epidemiology
School of Allied Health
East Carolina University
Greenville, NC 27834
Dr. Alan Marcus
Department of Mathematics
Washington State University
Pullman, WA 99164-2930
Contributing Author:
Dr. Dennis Kotchmar
Environmental Criteria and Assessment Office
MD-52
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Dr. Vic Hasselblad
Biometry Division
MD-55
U.S. Environmental Protection
Agency
Research Triangle Park, NC 27711
The following persons reviewed this chapter at EPA's request. The evaluations
and conclusions contained herein, however, are not necessarily those of the
reviewers.
Dr. Carol Angle
Department of Pediatrics
University of Nebraska
College of Medicine
Omaha, NE 68105
Dr. Lee Annest
Division of Health Examin. Statistics
National Center for Health Statistics
3700 East-West Highway
Hyattsville, MD 20782
Dr. Donald Barltrop
Department of Child Health
Westminister Children's Hospital
London SW1P 2NS
England
Dr. Irv Billick
Gas Research Institute
8600 West Bryn Mawr Avenue
Chicago, IL 60631
Dr. Joe Boone
Clinical Chemistry and
Toxicology Section
Centers for Disease Control
Atlanta, GA 30333
Dr. Robert Bornschein
University of Cincinnati
Kettering Laboratory
Cincinnati, OH 45267
Dr. A. C. Chamberlain
Environmental and Medical
Sciences Division
Atomic Energy Research
Establishment
Harwell 0X11
England
Dr. Neil Chernoff
Division of Developmental Biology
MD-67
U.S. Environmental Protection
Agnecy
Research Triangle Park, NC 27711
xxix
-------
Dr. Julian Chisolm
Baltimore City Hospital
4940 Eastern Avenue
Baltimore, MD 21224
Mr. Jerry Cole
International Lead-Zinc Research Organization
292 Madison Avenue
New York, NY 10017
Dr. Max Costa
Department of Pharmacology
University of Texas Medical School
Houston, TX 77025
Dr. Anita Curran
Commissioner of Health
Westchester County
White Plains, NY 10607
Dr. Jack Dean
Immunobiology Program and
Immunotoxicology/Cell Biology Program
CUT
P.O. Box 12137
Research Triangle Park, NC 27709
Dr. Fred deSerres
Assoc. Director for Genetics
NIEHS
P.O. Box 12233
Research Triangle Park, NC 27709
Dr. Robert Dixon
Laboratory of Reproductive and
Developmental Toxicology
NIEHS
P.O. Box 12233
Research Triangle Park, NC 27709
Dr. Claire Ernhart
Department of Psychiatry
Cleveland Metropolitan General Hospital
Cleveland, OH 44109
Dr. Sergio Fachetti
Section Head - Isotope Analysis
Chemistry Division
Joint Research Center
121020 Ispra
Varese, Italy
Dr. Virgil Ferm
Department of Anatomy and Cytology
Dartmouth Medical School
Hanover, NH 03755
Dr. Alf Fischbein
Environmental Sciences Laboratory
Mt. Sinai School of Medicine
New York, NY 10029
Dr. Jack Fowle
Reproductive Effects Assessment
Group
U.S. Environmental Protection
Agency
RD-689
Washington, DC 20460
Dr. Bruce Fowler
Laboratory of Pharmocology
NIEHS
P.O. Box 12233
Research Triangle Park, NC 27709
Mr. Eric Goldstein
Natural Resources Defense
Council, Inc.
School of Allied Health
122 E. 42nd Street
New York, NY 10168
Dr. Harvey Gonick
1033 Gayley Avenue
Suite 116
Los Angeles, CA 90024
Dr. Robert Goyer
Deputy Director
NIEHS
P.O. Box 12233
Research Triangle Park, NC 27709
Dr. Stanley Gross
Hazard Evaluation Division
Toxicology Branch
U.S. Environmental Protection Agency
Washington, DC 20460
Dr. Paul Hammond
University of Cincinnati
Kettering Laboratory
3223 Eden Avenue
Cincinnati, OH 45267
XXX
-------
Dr. Ronald D. Hood
Department of Biology
The University of Alabama
University, AL 35486
Dr. V. Houk
Centers for Disease Control
1600 Clifton Road, NE
Atlanta, GA 30333
Dr. Loren Koller
School of Veterinary Medicine
University of Idaho
Moscow, ID 83843
Dr. Kristal Kostial
Institute for Medical Research
and Occupational Health
Yu-4100 Zagreb
Yugoslavia
Dr. Lawrence Kupper
Department of Biostatisties
UNC School of Public Health
Chapel Hill, NC 27514
Dr. Phillip Landrigan
Division of Surveillance,
Hazard Evaluation and Field Studies
Taft Laboratories - NIOSH
Cincinnati, OH 45226
Dr. David Lawrence
Microbiology and Immunology Dept.
Albany Medical College of Union
University
Albany, NY 12208
Dr. Jane Lin-Fu
Office of Maternal and Child Health
Department of Health and Human Services
Rockville, MD 20857
Dr. Don Lynam
Air Conservation
Ethyl Corporation
451 Florida Boulevard
Baton Rouge, LA 70801
Dr. Kathryn Mahaffey
Division of Nutrition
Food and Drug Administration
1090 Tusculum Avenue
Cincinnati, OH 45226
Dr. Ed McCabe
Department of Pediatrics
University of Wisconsin
Madison, WI 53706
Dr. Paul Mushak
Department of Pathology
UNC School of Medicine
Chapel Hill, NC 27514
Dr. Chuck Nauman
Exposure Assessment Group
U.S. Environmental Protection
Agency
Washington, DC 20460
Dr. Herbert L. Needleman
Children's Hospital of Pittsburgh
Pittsburgh, PA 15213
Dr. H. Mitchell Perry
V.A. Medical Center
St. Louis, MO 63131
Dr. Charles G. Pfieffer
Engineering Department
Engineering Services Division
E I. duPont, Incorporated
Wilmington, DE 19898
Dr. Jack Pierrard
E.I. duPont de Nemours and
Company, Inc.
Petroleum Laboratory
Wilmington, DE 19898
Dr. Sergio Piomelli
Columbia University Medical School
Division of Pediatric Hematology
and Oncology
New York, NY 10032
XXXI
-------
Dr. Magnus Piscator
Department of Environmental Hygiene
The Karolinska Institute 104 01
Stockholm
Sweden
Dr. Robert Putnam
International Lead-Zinc
Research Organization
292 Madison Avenue
New York, NY 10017
Dr. Michael Rabinowitz
Children's Hospital Medical Center
300 Longwood Avenue
Boston, MA 02115
Dr. Harry Roels
Unite de Toxicologie
Industrielle et Medicale
Universite de Louvain
Brussels, Belgium
Dr. John Rosen
Division of Pediatric Metabolism
Albert Einstein College of Medicine
Montefiore Hospital and Medical Center
111 East 210 Street
Bronx, NY 10467
Dr. Stephen R. Schroeder
Division for Disorders
of Development and Learning
Biological Sciences Research Center
University of North Carolina
Chapel Hill, NC 27514
Dr. Anna-Maria Seppalainen
Institutes of Occupational Health
Tyoterveyslaitos
Haartmaninkatu 1
00290 Helsinki 29
Finland
Dr. Ellen Silbergeld
Environmental Defense Fund
1525 18th Street, NW
Washington, DC 20036
Dr. Ron Snee
E.I. duPont Nemours and
Company, Inc.
Engineering Department L3267
Wilmington, DE 19898
Dr. Gary Ter Haar
Toxicology and Industrial
Hygiene
Ethyl Corporation
451 Florida Boulevard
Baton Rouge, LA 70801
Dr. Ian von Lindern
Department of Chemical Engineering
University of Idaho
Moscow, ID 83843
Dr. Richard P. Weeden
V.A. Medical Center
Tremont Avenue
East Orange, NJ 07019
xxxn
-------
QUANTITATIVE EVALUATION OF LEAD AND BIOCHEMICAL INDICES
OF LEAD EXPOSURE IN PHYSIOLOGICAL MEDIA
9.1 INTRODUCTION
To understand the effects of an agent on an organism and, in particular, to formulate
statements of dose-effect relationships, one must be able to assess quantitatively the organ-
ism's degree of exposure to the substance. In the case of lead, internal biologically based
measures provide a more accurate indication of exposure than do external measures such as am-
bient air concentrations. Internal measures may be either directe.g., the level of lead 1n
a biological medium such as blood, calcified tissue, etc.or indirecte.g., the level of
some biochemical parameter or "indicator" closely associated with internal lead exposure.
This chapter examines the merits and weaknesses of various measurement methods as they are
currently used to assess lead exposure.
Quantitative analysis involves a number of discrete steps, all of which are important
contributors to the quality of the final result: (1) sample collection and transmission to
the laboratory; (2) laboratory manipulation of samples, physically and chemically, before ana-
lysis by instruments; (3) instrumental analysis and quantitative measurement; and (4) esta-
blishment of relevant criteria for accuracy and precision, namely, Internal and external qua-
lity assurance checks. Each of these steps Is discussed in this chapter in relation to the
measurement of lead exposure.
Clearly, the definition of "satisfactory analytical method" for lead has changed over the
years, paralleling (1) the evolution of more sophisticated instrumentation and procedures, (2)
a greater awareness of such factors as background contamination and loss of the element from
samples, and (3) development of new statistical methods to analyze data. For example, current
methods of lead analysis, such as anodic stripping voltammetry, background-corrected atomic
absorption spectrometry, and particularly isotope-dilution mass spectrometry, are more sensi-
tive and specific than the older classical approaches. Increasing use of the newer methods
would tend to result in lower lead values being reported for a given sample. Whether this
trend in analytical improvement can be isolated from other variables such as temporal changes
in exposure is another matter.
Because lead is ubiquitously distributed as a contaminant, the constraints (i.e., ultra-
clean, ultra-trace analysis) placed upon a laboratory attempting analysis of geochemical sam-
ples of pristine origin, or of extremely low lead levels in biological samples such as plasma,
are quite severe (Patterson, 1980). Very few laboratories can credibly claim such capability.
9-1
-------
Ideally, similar standards of quality should be adhered to across the rest of the analytical
spectrum. With many clinical, epldemiological, and experimental studies, however, these
standards may be unrealistic given the practical limitations and objectives of the studies.
Laboratory performance is but one part of the quality equation; the problems of sampling are
equally important but less subject to tight control. The necessity of rapidly obtaining a
blood sample in cases of suspected lead poisoning, or of collecting hundreds or thousands of
blood samples in urban populations, limits the number of sampling safeguards that can be rea-
listically achieved. Sampling in this context will always be accompanied by a certain amount
of analytical "suspicion." Furthermore, a certain amount of biological lead analysis data is
employed for comparative purposes, as in experimental studies concerned with the relative in-
crease in tissue burden of lead associated with increases in doses or severity of effects In
addition, any major compromise of an analytical protocol may be statistically discernible.
Thus, analysis of biological media for lead must be done under protocols that minimize the
risk of inaccuracy. Specific accuracy and precision characteristics of a method in a parti-
cular report should be noted to permit some judgment on the part of the reader about the in-
fluence of methodology on the reported results.
The choice of measurement method and medium for analysis is dictated both by the type of
information desired and by technical or logistical considerations. As noted elsewhere in this
document, whole blood lead reflects recent or continuing exposure, whereas lead in mineralized
tissue, such as deciduous teeth, reflects an exposure period of months and years. While urine
lead values are not particularly good correlates of lead exposure under steady-state condi-
tions in populations at large, such measurements may be of considerable clinical value. In ac-
quiring blood samples, the choice of venipuncture or finger puncture will be governed by such
factors as cost and feasibility, contamination risk, and the biological quality of the sample.
The use of biological indicators that strongly correlate with lead burden may be more desira-
ble, since they provide evidence of actual response and, together with blood lead data, pro-
vide a less risky diagnostic tool for assessing lead exposure.
9.2 DETERMINATIONS OF LEAD IN BIOLOGICAL MEDIA
9.2.1 Sampling and Sample Handling Procedures for Lead in Biological Media
Lead analysis in biological media requires careful sample collection and handling for two
reasons: (1) lead occurs at trace levels in most indicators of subject exposure, even under
conditions of high lead exposure; and (2) such samples must be obtained against a backdrop of
9-2
-------
pervasive contamination, the full extent of which may still be unrecognized by many laborato-
ries.
The reports of Speecke et a!. (1976), Patterson and Settle (1976), Murphy (1976), Berman
(1976), and Settle and Patterson (1980) review detailed aspects of the problems of sampling
and subsequent sample handling in the laboratory. These reports indicate that the normal pre-
cautions taken during sampling (detailed below for clinical and epidemiological studies)
should not be considered absolute, but rather as what is practical and feasible. They further
indicate that the inherent sensitivity or accuracy of a given method or instrument may be less
of a determining factor in the overall analysis than the quality of sample collection and
handling.
9.2.1.1 Blood Sampling. Samples for blood lead determination may be collected by venipunc-
ture (venous blood) or fingertip puncture (capillary blood). Collection of capillary versus
venous blood is usually decided by a number of factors, including the feasibility of obtaining
samples during the screening of many subjects and the difficulty of securing subject com-
pliance, particularly in the case of children and their parents. Furthermore, capillary blood
may be collected as discrete quantities in small-volume capillary tubes or as spots on filter
paper disks. With capillary tubes, obtaining good mixing with anticoagulant to avoid clotting
is important, as is the problem of lead contamination of the tube. The use of filter paper
requires the selection of paper with uniform composition, low lead content, and uniform blood
dispersal characteristics.
Whether venous or capillary blood is collected, much care must be exercised in cleaning
the site before puncture as well as in selecting lead-free receiving containers. Cooke et al.
(1974) employed vigorous scrubbing with a low-lead soap solution and rinsing with deionized
water, while Marcus et al. (1975) carried out preliminary cleaning with an ethanolic citric
acid solution followed by rinsing with 70-percent ethanol. Vigor in cleaning the puncture
site is probably as important as the choice of any particular cleaning agent. Marcus et al.
(1977) have noted that in one procedure for puncture site preparation, where the site is
covered with wet paper towels, contamination will occur if the paper towels are made from re-
cycled paper. Recycled paper retains a significant amount of lead.
In theory, capillary and venous blood lead levels should be virtually identical. However,
the literature indicates that some differences, which mainly reflect sampling problems, do
arise in the case of capillary blood. A given amount of contaminant has a greater impact on a
100-ul fingerstick sample than on a 5-ml sample of venous blood. Finger-coating techniques
may reduce some of the contamination (Mitchell et al., 1974). An additional problem is the
presence of lead in the anticoagulants used to coat capillary tubes. Also, lower values of
capillary versus venous blood lead may reflect "dilution" of the sample by extracellular fluid
9-3
-------
from excessive compression of the puncture site. When Joselow and Bogden (1972) compared a
method using finger puncture and spotting onto filter paper with a procedure using venous
blood and Hessel's procedure (1968) for flame atomic absorption spectrometry (see Section
9.2.2.1), they obtained a correlation coefficient of r = 0.9 (range, 20-46 pg/dl). Similarly,
Cooke et al. (1974) found an r value of 0.8 (no range given), while Mitchell et al. (1974) ob-
tained a value of 0.92 (10-92 ug/dl). Mahaffey et al. (1979) found that capillary blood
levels in a comparison test were approximately 20 percent higher than corresponding venous
blood levels in the same subjects, presumably reflecting sample contamination. Similar eleva-
tions have been described by DeSUva and Donnan (1980). Carter (1978) has found that blood
samples with lower hemoglobin levels may spread onto filter paper differently from normal
hemoglobin samples, requiring correction in quantification to obtain reliable values. This
complication should be kept in mind when considering children, who are frequently prone to
iron-deficiency anemia.
The relative freedom of the blood container from interior surface lead and the presence
of lead in the anticoagulant to be added to the blood are important considerations in venous
sampling. For studies focusing on "normal" ranges, such tubes may add some lead to blood and
still meet certification requirements. The "low-lead" heparinized blood tubes commercially
available (blue stopper Vacutainer, Becton-Dickinson) were found to contribute less than 0.2
ug/dl to whole blood samples (Rabinowitz and Needleman, 1982). Nackowski et al. (1977) sur-
veyed a large variety of commercially available blood tubes for lead and other metal contami-
nation. Lead uptake by blood over time from the various tubes was minimal with the "low-lead"
Vacutainer tubes and with all but four of the other tube types. In the large survey of
Mahaffey et al. (1979), 5-ml Monoject (Sherwood) or 7-ml lavender-top Vacutainer (Becton-
Dickinson) tubes were satisfactory. However, when more precision is needed, tubes are best
recleaned in the laboratory and lead-free anticoagulant added (although this would be less
convenient for sampling efficiency than the commercial tubes). In addition, blank levels for
every batch of samples should be verified.
9.2.1.2 Urine Sampling. Urine samples require collection using lead-free containers and caps
as well as the addition of a low-lead bactericide if samples are to be stored. While not
always feasible, 24-hr samples should be obtained because they level out any effect of vari-
ation in excretion over time. If spot sampling is done, lead levels should be expressed per
unit creatinine, or corrected for a constant specific gravity, if greater than 1.010.
9.2.1.3 Hair Sampling. The usefulness of hair lead analysis depends on the manner of samp-
ling. Hair samples should be removed from subjects by a consistent method, either by a pre-
determined length measured from the skin or by using the entire hair. Hair should be placed
in air-tight containers for shipment or storage. For segmental analysis, the entire hair
length 1s required.
9-4
-------
9.2.1.4 Mineralized Tissue. An important consideration in deciduous tooth collection is con-
sistency in the type of teeth collected from various subjects. Fosse and Justesen (1978) re-
ported no difference in lead content between molars and incisors, and Chatman and Wilson
(1975) reported comparable whole tooth levels for cuspids, incisors, and molars. On the other
hand, Mackie et al. (1977) and Lockeretz (1975) noted levels varying with tooth type, with a
statistically significant difference (Mackie et al., 1977) between second molars (lowest
levels) and incisors (highest levels). That the former two studies found rather low overall
lead levels across groups, while Mackie et al. (1977) reported higher values, suggests that
dentition differences in lead content may be magnified at relatively higher levels of ex-
posure. Delves et al, (1982), comparing pairs of central incisors or pairs of central and
lateral incisors from the same child, found that lead content may even vary within a specific
type of tooth. These data suggest the desirability of acquiring two teeth per subject to get
an average lead value.
Teeth containing fillings or extensive decay are best eliminated from analysis. Mackie
et al. (1977) discarded decayed teeth if the extent of decay exceeded approximately 30 per-
cent.
9.2.1.5 Sample Handling in the Laboratory. The effect of storage on lead content is a poten-
tial problem with blood samples. During storage, dilute aqueous solutions of lead surrender a
sizable portion of the lead content to the container surface, whether glass or plastic, unless
the sample is acidified (Issaq and Zielinskl, 1974; Unger and Green, 1977). Whether there is
a comparable effect, or comparable extent of such an effect, with blood is not clear. Unger
and Green (1977) claim that lead loss from blood to containers parallels that seen with aque-
ous solutions, but their data do not support this assertion. Moore and Meredith (1977) used
isotopic lead spiking (203Pb) with and without carrier in various containers at differing tem-
peratures to monitor lead stability in blood over time. The only material loss occurred with
soda glass at room temperature after 16 days. Nackowski et al. (1977) found that "low-lead"
blood tubes, while quite satisfactory in terms of sample contamination, began to show transfer
of lead to the container wall after 4 days. Meranger et al. (1981) studied movement of lead,
spiked to various levels, to containers of various composition as a function of temperature
and time. In all cases, reported lead loss to containers was significant. However, problems
exist with the above reports. Spiked samples probably are not incorporated into the same bio-
chemical environment as lead Inserted 1_n vivo. Also, Nackowski et al. (1977) did not indicate
whether the blood samples were kept frozen or refrigerated between testing intervals.
Mitchell et al. (1972) found that the effect of blood storage depends on the method of analy-
sis, with lower recoveries of lead from aged blood using the Hessel (1968) method.
9-5
-------
Lerner (1975) collected blood samples (35 originally) from a single subject into lead-
free tubes and, after freezing, forwarded them in blind fashion to a certified testing labor-
atory over a period of 9 months. Four samples were lost, and one was rejected as grossly con-
taminated (4 standard deviations from mean). Of the remaining 30 samples, the mean was 18.3
(jg/dl with a standard deviation (S.D.) of 3.9. The analytical method had a precision of ±3.5
pg/dl (S.D. = 1) at normal levels of lead, suggesting that the overall stability of the sam-
ples' lead content was good. Boone et al. (1979) reported that samples frozen for periods of
less than 1 year showed no effect of storage, while Piscator (1982) noted no change in low
levels (<10 pg/dl) when samples were stored at -20°C for 6 months. Based on the above data,
blood samples to be stored for any period of time should be frozen rather than refrigerated,
with care taken to prevent breaking the tube during freezing. Teeth and hair samples, when
stored in containers to minimize contamination, are indefinitely stable.
The actual site of analysis should be as free from lead as possible. Given the limited
availability of an "ultra-clean" facility such as that described by Patterson and Settle
(1976), the next desirable level of laboratory is the "Class 100" facility, in which fewer
than 100 airborne particles are greater than 0.5 urn in diameter. These facilities employ high-
efficiency particulate air filtering and laminar air flow (with movement away from sample
handling areas). Totally inert surfaces in the working area and an antechamber for removing
contaminated clothes, appliance cleaning, etc., are other necessary features.
All plastic and glass ware coming into contact with samples should be cleaned rigorously
and stored away from dust contact, and materials such as ashing vessels should permit minimal
lead leaching. In this regard, Teflon or quartz ware is preferable to other plastics or boro-
silicate glass (Patterson and Settle, 1976).
Reagents, particularly for chemical degradation of biological samples, should be both
certified and periodically tested for quality. Several commercial grades of reagents are
available, although precise work may require doubly purified materials from the National
Bureau of Standards (NBS). These reagents should be stored with a minimum of surface contami-
nation around the top of the containers.
For a more detailed discussion of appropriate laboratory practices, the reader may con-
sult LaFleur (1976).
9.2.2 Methods of Lead Analysis
Detailed technical discussion of the array of instruments available to measure lead in
blood and other media is outside the scope of this chapter (see Chapter 4). This discussion
is structured more appropriately to those aspects of methodology dealing with relative sensi-
tivity, specificity, accuracy, and precision. While acceptance of international standardized
9-6
-------
(SI) units for expressing lead levels in various media is increasing, units familiar to clini-
cians and epidemiologists will be used here. (To convert ug Pb/dl blood to SI units [umoles/
liter], multiply by 0.048.)
Many reports over the years have purported to offer satisfactory analysis of lead in bio-
logical media, but in fact have shown rather meager adherence to criteria for accuracy and
precision or have shown a lack of demonstrable utility across a wide spectrum of analytical
applications. Therefore, discussion in this section is confined to "definitive" and reference
methods for lead analysis, except for a brief treatment of the traditional but now widely sup-
planted colorimetric method.
Using the definition of Cali and Reed (1976), a definitive method is one in which all
major or significant parameters are related by solid evidence to the absolute mass of the ele-
ment with a high degree of confidence. A reference method, by contrast, is one of demonstra-
ted accuracy, validated by a definitive method, and arrived at by consensus through perfor-
mance testing by a number of different laboratories. In the case of lead in biological media,
the definitive method is isotope-dilution mass spectrometry (IDMS). IOMS is so accurate be-
cause all manipulations are on a weight basis involving simple procedures. The measurements
entail only ratios and not the absolute determinations of the isotopes involved, which greatly
reduces instrumental corrections or errors. No interferences occur from sample matrix or
other elements, and the method does not depend on recovery. Reproducible results to a pre-
cision of one part in 104 or 105 are routine with specially designed instruments.
In terms of reference methods for lead in biological media, such a label is commonly
attached to atomic absorption spectrometry (AAS) in its various instrumentation/ methodology
configurations and to the electrochemical technique, anodic stripping voltammetry (ASV).
These have been termed reference methods insofar as their precision and accuracy can be veri-
fied or calibrated against IDMS.
Other methods that are recognized for general trace-metal analysis are not fully applica-
ble to biological lead or have inherent shortcomings. X-ray fluorescence analysis lacks the
requisite sensitivity for media with low lead content, and the associated sample preparation
may present a high contamination risk. A notable exception may be X-ray fluorescence analysis
of teeth or bone i_n sjtu as discussed below. Neutron-activation analysis is the method of
choice with many elements, but it is not technically feasible for lead analysis because of the
absence of long-lived isotopes.
9.2.2.1 Lead Analysis in Whole Blood. The first generally accepted technique for quantifying
lead in whole blood and other biological media was a colorimetric method that involved spec-
trophotometric measurement based on the binding of lead to a chromogenic agent to yield a
chromophoric complex. The complexing agent has typically been dithizone, 1,5-diphenylthio-
carbazone, yielding a lead complex that is spectrally measured at 510 nm.
9-7
-------
Two variations of the spectrophotometric technique used when measuring low levels of lead
have been the procedures of the U.S. Public Health Service (USPHS) (National Academy of
Sciences, 1972) and of the American Public Health Association (APHA) (1955). In both, venous
blood or urine is wet ashed using concentrated nitric acid of low lead content followed by ad-
justment of the ash with hydroxylamine and sodium citrate to a pH of 9-10. Cyanide ion is
added and the solution extracted with dithizone in chloroform. Back extraction removes the
lead into dilute nitric acid; the acid layer is treated with ammonia, then cyanide, and re-
extracted with dithizone in chloroform. The extracts are read in a spectrophotometer at 510
nm. Bismuth interference is handled (APHA variation) by removal with dithizone at pH 3.4.
According to Lerner (1975), the analytical precision in the "normal" range is about ±3.5 ug/dl
(S.D. = 1), using 5 ml of sample.
The most accurate and precise method for lead measurement in blood is IDMS. As typified
by the report of Machlan et al. (1976), whole blood samples are accurately weighed, and a
weighed aliquot of 206Pb-enr1ched isotope solution is added. After sample decomposition with
ultra-pure nitric and perchloric acids, samples are evaporated, residues are taken up in di-
lute lead-free hydrochloric acid (HC1), and lead is isolated using anion-exchange columns.
Column eluates are evaporated with the above acids, and lead is deposited onto high-purity
platinum wire from dilute perchloric acid. The 206Pb/208Pb ratio is then determined by ther-
mal ionization mass spectrometry. Samples without added isotope and reagent blanks are also
carried through the procedure. In terms of precision, the 95-percent confidence level for
lead samples overall is within 0.15 percent. Because of the expense incurred by the require-
ments for operator expertise, the amount of time involved, and the high standard of laboratory
cleanliness, IDMS is mainly of practical value in the development of standard reference ma-
terials and for the verification of other analytical methods.
AAS is widely used for lead measurements in whole blood, with sample analysis involving
analysis of venous blood with chemical degradation, analysis of liquid samples with or without
degradation, and samples applied to filter paper. It is thus the most flexible for samples
already collected or subject to manipulation. By means of flame or electrothermal excitation,
ionic lead In a matrix is first vaporized and then converted to the atomic state, followed by
resonance absorption from either a hollow cathode or electrodeless discharge lamp generating
lead absorption lines at 217.0 and 283.3 nm. After monochrometer separation and photomulti-
plier enhancement of the differential signal, lead content is measured electronically.
The earliest methods of AAS analysis involved the aspiration of ashed blood samples into
a flame, usually subsequent to extraction into an organic solvent, to enhance sensitivity by
preconcentration. Some methods did not involve digestion steps prior to solvent extraction
9-8
-------
(KopHo et al., 1974). Of these various flame AAS methods, Hessel's (1968) technique con-
tinues to be used with some frequency.
Currently, lead measurement in blood by AAS employs several different methods that permit
greater sensitivity, precision, and economy of sample and time. The flame method of Delves
(1970), called the "Delves cup" procedure, usually involves delivery of discrete small samples
(^100 jjl) of unmodified whole blood to nickel cups, with subsequent drying and peroxide decom-
position of organic content before positioning in the flame. The marked enhancement of sen-
sitivity over conventional flame aspiration results from immediate, total consumption of the
sample and generation of a localized population of atoms. In addition to discrete blood vol-
umes, blood-containing filter paper disks have been used (Joselow and Bogden, 1972; Cernik and
Sayers, 1971; Piomelli et al., 1980). Among the several modifications of the Delves method
are that of Ediger and Coleman (1972), in which dried blood samples in the cups are pre-
ignited to destroy organic matter by placement near the flame in a precise, repeatable manner,
and the variation of Barthel et al. (1973), in which blood samples are mixed with dilute
nitric acid in the cups followed by drying in an oven at 200°C and charring at 450°C on a hot
plate. A number of laboratories eschew even these modifications and follow dispensing and
drying with direct placement of the cup into the flame (e.g., Mitchell et al., 1974). The
Delves cup procedure may require correction for background spectral interference. This cor-
rection is usually achieved using instrumentation equipped at a nonresonance absorption line.
While the 217.0-nm line of lead is less subject to such interference, precise work is best
done with correction. This method as applied to whole blood lead appears to have an oper-
ational sensitivity down to 1.0 ug Pb/dl, or somewhat below when competently employed, and a
relative precision of approximately 5 percent in the range of levels encountered in the United
States.
AAS methods using electrothermal (furnace) excitation in lieu of a flame can be approxi-
mately tenfold more sensitive than the Delves procedure. A number of reports describing whole
blood lead analysis have appeared in the literature (Lawrence, 1982, 1983). Because of in-
creased sensitivity, the "fTameless" AAS technique permits the use of small blood volumes
(1-5 ul) with samples undergoing drying and dry ashing in situ. Physicochemical and spectra!
interferences are inherently severe with this approach, requiring careful background cor-
rection. In one flameless AAS configuration, background correction exploits the Zeeman ef-
fect, where correction is made at the specific absorption line of the element and not over a
band-pass region, as is the case with the deuterium arc. While control of background inter-
ference up to 1.5 molecular absorbance is claimed with the Zeeman system (Koizumi and Yasuda,
1976), employing charring before atomization is technically preferable. Hinderberger et al.
(1981) used dilute ammonium phosphate solution to minimize chemical interference in their fur-
nace AAS method.
9-9
-------
Precision can be a problem in the flameless technique unless careful attention is paid to
the problem of sample diffusibility over and into the graphite matrix of the receiving recep-
tacle (tube, cup, or rod). With the use of diluted samples and larger applied volumes, the
relative precision of this method can approach that of the Delves technique (Delves, 1977).
In addition to the various AAS methods noted above, electrochemical techniques have been
applied to blood lead analysis. Electrochemical methods, in theory, differ from AAS methods
in that the latter are "concentration" methods regardless of sample volumes available, while
electrochemical analysis involves bulk consumption of sample and hence would have infinite
sensitivity, given an infinite sample volume. This intrinsic property is of little practical
advantage given usual limits of sample volume, instrumentation design, and blanks.
The most widely used electrochemical method for lead measurement in whole blood and other
biological media is ASV, which is also probably the most sensitive because it involves an elec-
trochemical preconcentration (deposition) step in the analysis (Matson and Roe, 1966; Matson
et al., 1971). In this method, samples such as whole blood (50-100 (jl) are preferably, but
not commonly, wet ashed and reconstituted in dilute acid or made electro-available with metal
exchange reagents. Using freshly prepared composite electrodes of mercury film deposited on
carbon, lead is plated out from the solution for a specific amount of time and at a selected
negative voltage. The plated lead is then reoxidized in the course of anodic sweeping, gene-
rating a current peak that may be recorded on a chart or displayed on commercial instruments
as units of concentration (ug/dl).
One alternative to the time and space demands of wet ashing blood samples is the use of
metal exchange reagents that displace lead from binding sites in blood by competitive binding
(Morrell and Giridhar, 1976; Lee and Meranger, 1980). In one commercial preparation, this re-
agent consists of a solution of calcium, chromium, and mercuric ions. Use of the metal ex-
change reagent adds a chemical step that must be carefully controlled for full recovery of
lead from the sample.
The working detection limit of ASV for blood is comparable to that of the AAS flameless
methods, while the relative precision is best with prior sample degradation, approximately 5
percent. The precision is less when the blood samples are run directly with the ion exchange
reagents (Morrell and Giridhar, 1976), particularly at the low end of "normal" blood lead
values. While AAS methods require attention to various spectral interferences to achieve
satisfactory performance, electrochemical methods such as ASV require consideration of such
factors as the effects of co-reducible metals and agents that complex lead and alter its re-
duction-oxidation (redox) potential properties. Chelants used in therapy, particularly peni-
cillamine, may interfere, as does blood copper, which may be elevated in pregnancy and during
such disease states as leukemia, lymphoma, and hyperthyroidism (Berman, 1981).
9-10
-------
Correction of whole blood lead values for hematocrit, although carried out in the past,
is probably not appropriate and not commonly done at present. While the erythrocyte is the
carrier for virtually all lead in blood, the saturation capacity of the red blood cell (RBC)
for lead is so high that it can still carry lead even at highly toxic levels (Kochen and
Greener, 1973). Kochen and Greener (1973) also showed that acute or chronic dosing at a given
lead level in rats with a wide range of hematocrits (induced by bleeding) gave similar blood
lead values. Rosen et al. (1974), based on studies of hematocrit, plasma, and whole blood
lead in children, noted hematocrit correction was not necessary, a view supported by Chisolro
(1974).
9.2.2.2 Lead in Plasma. While virtually all of the lead present in whole blood is bound to
the erythrocyte (Robinson et al., 1958; Kochen and Greener, 1973), lead in plasma is trans-
ported to affected tissues. Therefore, every precaution must be taken to use nonhemolyzed
blood samples for plasma isolation. The very low levels of lead in plasma require that more
attention be paid to "ultra-clean" methods.
Rosen et al. (1974) used fTameless AAS and microliter samples of plasma to measure plasma
lead, with background correction for the smoke signal generated for the unmodified sample.
Cavalleri et al. (1978) used a combination of solvent extraction of modified plasma with pre-
concentrating and flameless AAS. These authors noted that the method used by Rosen et al.
(1974) permitted less precision and accuracy than did their technique, because a significantly
smaller amount of lead was delivered to the furnace accessory.
DeSilva (1981), using a technique similar to that of Cavalleri et al. (1978), but col-
lecting samples in heparinized tubes, claimed that the use of ethylenediaminetetraacetic acid
(EDTA) as anticoagulant disturbs the cell-plasma distribution of lead enough to yield errone-
ous data. Much more care was given in this procedure to background contamination. In both
cases, increasing levels of plasma lead were measured with increasing whole blood lead, sug-
gesting an equilibrium ratio that contradicts the data of Rosen et al. (1974). They found a
fixed level of 2-3 ug/dl plasma over a wide range of blood lead values. However, the actual
levels of lead in plasma in the DeSilva (1981) study were much lower than those reported by
Cavalleri et al. (1978).
Using IDMS and sample collection/manipulation in an "ultra-clean" facility, Everson and
Patterson (1980) measured the plasma lead levels in two subjects, a control and a lead-exposed
worker. The control had a plasma lead level of 0.002 (jg/dl, several orders of magnitude lower
than that seen with studies using less precise analytical approaches. The lead-exposed worker
had a plasma level of 0.2 ug/dl. Several other reports in the literature using IDMS noted
somewhat higher values of plasma lead (Manton and Cook, 1979; Rabinowitz et al., 1974), which
Everson and Patterson (1980) have ascribed to problems of laboratory contamination.
9-11
-------
Using tracer lead to minimize the impact of contamination results in a value of 0.15
(Rablnowitz et al., 1974).
With appropriate plasma lead methodology, reported lead levels are extremely low, the de-
gree varying with the methods used to measure such concentrations. While the data of Everson
and Patterson (1980) were obtained from only two subjects, it seems unlikely that using more
subjects would result in a plasma lead range extending upward to the levels seen with ordinary
methodology in ordinary laboratory surroundings. The above considerations are important when
discussing appropriate methodology for plasma analysis, and the Everson and Patterson (1980)
report indicates that some doubt surrounds results obtained with conventional methods. Al-
though not the primary focus of their study, the values obtained by Everson and Patterson
(1980) for whole blood lead, unlike the data for plasma, are within the ranges for unexposed
(11 ug/dl) and exposed (80 ug/dl) subjects generally reported with other methods. This agree-
ment would suggest that, for the most part, reported values do actually reflect HI vivo blood
lead levels rather than sampling problems or inaccurate methods.
9.2.2.3 Lead in Teeth. When analyzing shed deciduous or extracted permanent teeth, some in-
vestigators have used the whole tooth after surface cleaning to remove contaminating lead
(e.g., Moore et al., 1978; Fosse and Justesen, 1978; Mackie et al. , 1977), while others have
measured lead in dentine (e.g., Shapiro et al., 1973; Needleman et al., 1979; Al-Naimi et al.,
1980). Several reports (Grandjean et al., 1979; Shapiro et al., 1973) have also described the
analysis of circumpulpal dentine, that portion of the tooth found to have the highest relative
fraction of lead. Needleman et al. (1979) separated dentine by embedding the tooth in wax,
followed by thin central sagittal sectioning. The dentine was then isolated from the sawed
sections by careful chiseling.
Determining mineral and organic composition of teeth and their components requires the
use of thorough chemical decomposition techniques, including wet ashing and dry ashing steps
and sample pulverizing or grinding. In the procedure of Steenhout and Pourtois (1981), teeth
are dry ashed at 450°C, powdered, and dry ashed again. The powder is then dissolved in nitric
acid. Fosse and Justesen (1978) reduced tooth samples to a coarse powder by crushing in a
vise, followed by acid dissolution. Oehme and Lund (1978) crushed samples to a fine powder in
an agate mortar and dissolved the samples in nitric acid. Mackie et al. (1977) and Moore et
al. (1978) dissolved samples directly in concentrated acids. Chatman and Wilson (1975) and
Needleman et al. (1974) carried out wet ashing with nitric acid followed by dry ashing at
450°C. Oehme and Lund (1978) found that acid wet ashing of tooth samples yielded better re-
sults If carried out in a heated Teflon bomb at 200°C.
With regard to methods of measuring lead in teeth, AAS and ASV have been employed most
often. With the AAS methods, the high mineral content of teeth tends to argue for isolating
9-12
-------
lead from this matrix before analysis. In the methods of Needleman et al. (1974) and Chatman
and Wilson (1975), ashed residues 1n nitric add were treated with ammonium nitrate and ammo-
nium hydroxide to a pH of 2.8, followed by dilution and extraction with a methylIsobutylketone
solution of ammonium pyrrol1d1necarbod1th1oate. Analysis was by flame AAS, using the 217.0-nm
lead-abs'orption line. A similar procedure was employed by Fosse and Justesen (1978).
ASV has been successfully used 1n tooth lead measurement (Shapiro et al., 1973; Needleman
et al., 1979; Oehme and Lund, 1978). As typified by the method of Shapiro et al. (1973), sam-
ples of dentine were dissolved 1n a small volume of low-lead concentrated perchloric acid and
diluted (5.0 ml) with lead-free sodium acetate solution. With deoxygenation, samples were
analyzed 1n a commercial ASV unit, using a plating time of 10 min at a plating potential of
-1.05 V. Anodic sweeping was at a rate of 60 mV/sec with a variable current of 100-500 pA.
Since lead content of teeth 1s higher than 1n most samples of biological media, the relative
precision of analysis with appropriate accommodation of the matrix effect, such as the use of
matrix-matched standards, 1n the better studies indicates a value of approximately 5-7 per-
cent.
In an analysis of lead levels 1n permanent teeth of Swedish subjects, Moller et al.
(1982) used particle-Induced X-ray emission (PIXE). While this method permits analysis with
minimal contamination risk, 1t measures only coronal dentine, which is relatively less re-
vealing about cumulative exposure than secondary or circumpulpal dentine.
All of the above methods involve shed or extracted teeth and consequently provide a ret-
rospective determination of lead exposure. In Bloch et al.'s (1976) procedure, tooth lead 1s
measured i_n situ using an X-ray fluorescence technique. A collimated beam of radiation from
"Co was allowed to irradiate the upper central incisor teeth of the subject. Using a rela-
tively safe 100-sec irradiation time and measurement of KQi and Ka2 lead lines via a germanium
diode and a pulse-height analyzer for signal processing, lead levels of 15 ppm or higher could
be measured. Multiple measurement by this method would be very useful 1n prospective studies
because it would show the "ongoing" rate of increase in body lead burden. Furthermore, when
combined with serial blood sampling, it would provide data for blood lead-tooth lead relation-
ships.
9224 Lead in Hair. Hair constitutes a noninvasive sampling source with virtually no prob-
lems with sample stability on extended storage. However, the advantages of accessibility and
stability are offset by the problem of assessing external contamination of the hair surface by
atmospheric fallout, hand dirt, lead 1n hair preparations, etc. Thus, such samples are prob-
ably of less value overall than those from other media.
The various methods that have been employed for removal of external lead have been re-
viewed (Chatt et al., 1980; Gibson, 1980; Chattopadhyay et al., 1977). Cleaning techniques
obviously should be vigorous enough to remove surface lead but not so vigorous as to remove
9-13
-------
the endogenous fraction. To date, no published cleaning procedure has been proven reliable
enough to permit acceptance of reported levels of lead in hair. Such a demonstration would
have to use lead isotopic studies with both surface and endogenous isotopic lead removal moni-
tored as a function of a particular cleaning technique.
9.2.2.5 Lead in Urine. Analysis of lead in urine is complicated by its relatively low con-
centrations (lower than in blood in many cases) as well as by the complex mixture of mineral
elements present. Lead levels are higher, of course, in cases where lead mobilization or
therapy with chelants is in progress, but in these cases samples must be analyzed to account
for lead bound to chelants such as EDTA. Such analysis requires either sample ashing or the
use of standards containing the chelant. Although analytical methods have been published for
the direct analysis of lead in urine, samples are probably best wet ashed before analysis,
using the usual mixtures of nitric plus sulfuric and/or perchloric acids.
Both AAS and ASV methods have been applied to urine lead analyses, the former employing
either direct analysis of ashed residues or a preliminary chelation-extraction step. With
flame AAS, ashed urine samples must invariably be extracted with a chelant such as ammonium
pyrrolidinecarbodithioate in methylisobutylketone to achieve reasonably satisfactory results.
Furthermore, direct analysis creates mechanical problems with burner operation, due to the
high mineral content of urine, and results in considerable maintenance problems with equip-
ment. The procedure of Lauwerys et al. (1975) is typical of flame AAS methods with prelimin-
ary lead separation. Because of the relatively greater sensitivity of graphite furnace
(flameless) AAS, this variation of the method has been applied to urine analysis. In scat-
tered reports of such analyses, adequate performance for direct sample analysis seems to
require steps to minimize matrix interference. A typical example of one of the better direct
analysis methods is that of Hodges and Skelding (1981). Urine samples were mixed with iodine
solution and heated, then diluted with a special reagent containing ammonium molybdate, phos-
phoric acid, and ascorbic acid. Small aliquots (5 (jl) were delivered to the furnace accessory
of an AAS unit containing a graphite tube pretreated with ammonium molybdate. The relative
standard deviation of the method is reported to be about 6 percent. In the method of Legotte
et al. (1980), such tube treatment and sample modifications were not employed and the average
precision figure was 13 percent.
Compared with various AAS methods, ASV has been less frequently employed for urine lead
analysis. From a survey of available electrochemical methods in general, such techniques
applied to urine appear to require further development. Franke and de Zeeuw (1977) used dif-
ferential-pulse ASV as a screening tool for lead and other elements in urine. Jagner et al.
(1979) described analysis of urine lead using potentiometric stripping. In their procedure the
element was preconcentrated at a thin-film mercury electrode as in conventional ASV, but
9-14
-------
deoxygenated samples were reoxidized with either oxygen or mercuric ions after the circuitry
was disconnected.
As noted in Section 9.1.1.2, if collection of 24-hr samples is not possible, spot sam-
pling of lead in urine can be conducted, and results should be expressed per unit creatinine.
9.2.2.6 Lead in Other Tissues. Bone samples of experimental animal or human autopsy origin
require preliminary cleaning procedures for removal of muscle and connective tissue, with care
being taken to minimize sample contamination. As is the case with teeth, samples must be
chemically decomposed before analysis. Satisfactory instrumental methods for bone lead analy-
sis comprise a much smaller literature than is the case for other media.
Wittmers et al. (1981) have described the measurement of lead in dry ashed (450°C) bone
samples using flameless AAS. Ashed samples were weighed and dissolved in dilute nitric acid
containing lanthanum ion, the latter being used to suppress interference from bone elements.
Small volumes (20 |jl) and high calcium content required that atomization be done at 2400°C to
avoid condensation of calcium within the furnace. Quantification was by the method of addi-
tions. Relative precision was 6-8 percent at relatively high lead content (60 (jg/g ash) and
10-12 percent at levels of 14 ug/g ash or less.
Ahlgren et al. (1980) described the application of X-ray fluorescence analysis to jn vivo
lead measurement in the human skeleton, using tibia and phalanges. In this technique, irra-
diation is carried out with a dual 57Co gamma ray source. The generated Kal and KQ2 lead
lines are detected with a lithium-drifted germanium detector. The detection limit is 20 ppm.
Soft organs differ from other biological media in the extent of anatomic heterogeneity as
well as lead distribution, e.g., brain versus kidney. Hence, sample analysis involves either
discrete regional sampling or the homogenizing of an organ. The efficiency of the latter can
vary considerably, depending on the density of the homogenate, the efficiency of rupture of
the formed elements, and other factors. Glass-on-glass homogenizing should be avoided because
lead is liberated from the glass matrix with abrasion.
AAS in its flame or flameless variations, is the method of choice in many studies. In
the procedure of Slavin et al. (1975), tissues were wet ashed and the residues taken up in di-
lute acid and analyzed with the furnace accessory of an AAS unit. A large number of reports
representing slight variations of this basic technique have appeared over the years (Lawrence,
1982 1983). Flame procedures, being less sensitive than the graphite furnace method, require
more'sample than may be available or are restricted to measurement in tissues where levels are
relatively high, e.g., Mdney. In the method of Farris et al. (1978), samples of brain,
liver lung or' spleen (as discrete segments) were lyophilized and then solubilized at room
temperature 'with nitric acid. Following neutralization, lead was extracted into methyliso-
butylketone with ammonium pyrrolidinecarbodithioate and aspirated into the flame of an AAS
unit. The reported relative precision was 8 percent.
9-15
-------
9.2.3 Quality Assurance Procedures In Lead Analysis
Regardless of technical differences among the different methodologies for lead analysis,
one can define the quality of such techniques as being of certain categories: (1) poor accu-
racy and poor precision; (2) poor accuracy and good precision; or (3) good accuracy and good
precision. In terms of available information, the major focus in assessing quality has been
on blood lead determinations.
According to Boutwell (1976), the use of quality control testing for lead measurement
rests on four assumptions: (1) that the validity of the specific procedure for lead in some
matrix has been established; (2) that the stability of the factors making up the method has
been both established and manageable; (3) that the validity of the calibration process .and the
calibrators with respect to the media being analyzed has been established; and (4) that surro-
gate quality control materials of reliably determined analyte content can be provided. These
assumptions, when translated into practice, revolve around steps employed within the labora-
tory, using a battery of "internal checks" and a further reliance on "external checks" such as
a formal, well-organized, multi-laboratory proficiency testing program.
Analytical quality protocols can be further divided into start-up and routine procedures,
the former entailing the establishment of detection limits, "within-run" and "between-run"
precision, and recovery of analyte. When a new method is adopted for some specific analytical
advantage, the procedure is usually tested Inside or outside the laboratory for comparative
performance. For example, Hicks et al. (1973) and Kubasik et al. (1972) reported that flame-
less techniques for measuring lead in whole blood had a satisfactory correlation with results
using conventional flame procedures. Matson et al. (1971) noted a good agreement between ASV
and both AAS and dithizone colorimetric techniques. The problem with such comparisons is that
the reference method is assumed to be accurate for the particular level of lead in a given
matrix. High correlations obtained in this manner may simply indicate that two inaccurate
methods are simultaneously performing with the same level of precision.
Preferable approaches for assessing accuracy are the use of certified samples determined
by a definitive method or direct comparison of different techniques with a definitive proce-
dure. For example, Eller and Hartz (1977) compared the precision and accuracy of five availa-
ble methods for measuring lead in blood: dithizone spectrometry, extraction and tantalum boat
AAS, extraction and flame aspiration AAS, direct aspiration AAS, and graphite furnace AAS
techniques. Porcine whole blood certified by NBS using IDMS at 1.00 ug/g (±0.023) was tested
and all methods were found to be equally accurate. The tantalum boat technique was the least
precise. The obvious limitation of data from this technique is that they relate to a high
blood lead content, suitable for use in measuring the exposure of lead workers or in some
other occupational context, but less appropriate for clinical or epidemiological investi-
gations.
9-16
-------
Boone et al. (1979) compared the analytical performance of 113 laboratories using various
methods and 12 whole blood samples (blood from cows fed a lead salt) certified as to lead con-
tent using IDMS at the NBS. Lead content ranged from 13 to 102 (jg/dl, determined by ASV and
five variations of AAS. The order of agreement with NBS values, i.e., relative accuracy, was
as follows: extraction > ASV > tantalum strip > graphite furnace > Delves cup > carbon rod.
The AAS methods all showed bias, having positive values at less than 40 ug/dl and negative
values at levels greater than 50 ug/dl. ASV showed less of a positive bias problem, although
it was not bias free within either of the blood lead ranges. In terms of relative precision,
the ranking was: ASV > Delves cup > tantalum strip > graphite furnace > extraction > carbon
rod. The overall ranking 1n accuracy and precision Indicated: ASV > Delves cup > extraction
> tantalum strip > graphite furnace > carbon rod. As the authors cautioned, the above data
should not be taken to Indicate that any established laboratory using one particular technique
would not perform better; rather, 1t should be used as a guide for newer facilities choosing
among methods.
A number of steps in quality assurance pertinent to the routine measurement of lead are
necessary 1n an ongoing program. With respect to Internal checks of routine performance,
these steps Include calibration and precision and accuracy testing. With biological matrices,
the use of matrix-matched standards 1s quite Important, as is an understanding of the range of
linearity and variation of calibration curve slopes from day to day. Analyzing a given sample
1n duplicate 1s common practice, with further replication carried out if the first two deter-
minations vary beyond a predetermined range. A second desirable step is the analysis of sam-
ples collected in duplicate but analyzed "blind" to avoid bias.
Monitoring accuracy within the laboratory 1s limited to the availability of control sam-
ples having a certified lead content in the same medium as the samples being analyzed. Con-
trols should be as physically close to the media being analyzed as possible. Standard refer-
ence materials (SRMs), such as orchard leaves and lyophlUzed bovine liver, are of help in
some cases, but NBS-certified blood samples are needed for the general laboratory community.
Whole blood samples, prepared and certified by the marketing facility (TOX-EL, A.R. Smith Co.,
Los Angeles, CA; Kaulson Laboratories, Caldwell, NJ; Behringwerke AG, Marburg, W. Germany; and
Health Research Institute, Albany, NY) are available commercially. With these samples, atten-
tion must be paid to the reliability of the methods used by reference laboratories. The use
of such materials, from whatever source, must minimize bias; for example, the attention given
control specimens should be the same as that given routine samples.
Finally, the most important form of quality assurance 1s the ongoing assessment of lab-
oratory performance by proficiency testing programs using externally provided specimens for
analysis. Earlier Interlaboratory surveys of lead measurement in blood and in urine indicated
9-17
-------
that a number of laboratories had performed unsatisfactorily, even when dealing with high con-
centrations of lead (Keppler et al., 1970; Donovan et al., 1971; Berlin et al., 1973), al-
though some of the problems may have originated in the preparation and status of the blood
samples during and after distribution (World Health Organization, 1977). These earlier tests
for proficiency indicated the following: (1) many laboratories were able to achieve a good
degree of precision within their own facilities; (2) the greater the number of samples rou-
tinely analyzed by a facility, the better the performance; and (3) 30 percent of the labora-
tories routinely analyzing blood lead reported values differing by more than 15 percent from
the true level (Pierce et al., 1976).
In the more recent, but very limited, study of Paulev et al. (1978), five facilities par-
ticipated 1n a survey, using samples to which known amounts of lead had been added. For lead
in both whole blood and urine, the interlaboratory coefficient of variation was reported to be
satisfactory, ranging from 12.3 to 17.2 percent. Aside from Its limited scope, this study
used "spiked" instead of jfn vivo lead, so that extraction techniques used in most of the labo-
ratories surveyed would have given misleadlngly better results in terms of actual recovery.
Maher et al. (1979) described the outcome of a proficiency study involving up to 38 lab-
oratories that analyzed whole blood pooled from a large number of samples submitted for blood
lead testing. The Delves cup technique was the most heavily represented, followed by the che-
lation-extraction plus flame AAS method and the graphite furnace AAS method. ASV was used by
only approximately 10 percent of the laboratories, so that the results basically portray AAS
methods. All laboratories had about the same degree of accuracy, with no evidence of consis-
tent bias, while the interlaboratory coefficient of variation was approximately 15 percent. A
subset of this group, certified by the American Industrial Hygiene Association (AIHA) for air
lead, showed a corresponding precision figure of approximately 7 percent. Over time, the sub-
set of AIHA-certified laboratories remained about the same in proficiency, while the other
facilities showed continued improvement in both accuracy and precision. This study Indicates
that program participation does help the performance of a laboratory doing blood lead determi-
nations.
The most comprehensive proficiency testing program 1s that carried out by the Centers for
Disease Control (CDC) of the U.S. Public Health Service (USPHS). This testing program con-
sists of two operationally and administratively distinct subprograms, one conducted by the
Center for Environmental Health (CEH) and the other by the L1censure and Proficiency Testing
Division, Laboratory Improvement Program Office (LIPO). The CEH program is directed at fa-
cilities involved in lead poisoning prevention and screening, while LIPO is concerned with
laboratories seeking certification under the Clinical Laboratories Improvement Act of 1967 as
well as under regulations of the Occupational Safety and Health Administration (OSHA). Both
9-18
-------
the CEH and LIPO protocols Involve the use of bovine whole blood certified as to content by
reference laboratories (6 in the CEH program, 20-23 in LIPO) with an ad hoc target range of ±6
(jg/dl for values of 40 ug/dl or less and ±15 percent for higher levels. Three samples are
provided monthly from CEH, for a total of 36 yearly, while LIPO participants receive three
samples quarterly (12 samples yearly). Use of a fixed range rather than a standard deviation
has the advantage of allowing the monitoring of overall laboratory Improvement.
For fiscal year (FY) 1981, 114 facilities were in the CEH program, 92 of them participa-
ting for the entire year. Of these, 57 percent each month reported all three samples within
the target range, and 85 percent on average reported two out of three samples correctly. Of
the facilities reporting throughout the year, 95 percent had a 50 percent or better perfor-
mance, i.e., 18 blood samples or better. Comparing the summary data for FY 1981 with earlier
annual reports, one sees considerable improvement in the number of laboratories achieving
higher levels of proficiency. For the interval FY 1977-79, there was a 20 percent increase in
the number correctly analyzing more than 80 percent of all samples and a 33 percent decrease
in those reporting less than 50 percent correct. In the last several years, FY 1979-81, over-
all performance has more or less stabilized.
With the LIPO program for 1981 (Dudley, 1982), the overall laboratory performance aver-
aged across all quarters was 65 percent of the laboratories analyzing all samples correctly
and approximately 80 percent performing well with two of three samples. Over the 4 years of
this program, an increasing ability to analyze lead in blood correctly has been demonstrated.
Dudley's (1982) survey also Indicates that reference laboratories in the LIPO program are be-
coming more accurate relative to IDMS values, i.e., bias over the blood lead range is con-
tracting.
Current OSHA criteria for certification of laboratories measuring occupational blood lead
levels require that eight of nine samples, 89 percent, be within 6 pg/dl or 15 percent of re-
ference laboratory means for samples sent over the three previous quarters (U.S. Occupational
Safety and Health Administration, 1982). These criteria reflect the ability of a number of
laboratories to perform at this level.
Note that most proficiency programs, including the CEH and LIPO surveys, are appropriate-
ly concerned with blood lead levels encountered in such cases as pedlatric screening for ex-
cessive exposure to lead or in occupational exposures. As a consequence, underrepresentation
of lead values in the low end of the "normal" range occurs. In the CEH distribution for FY
1981, four samples (11 percent) were below 25 ug/dl. The relative performance of the 114
facilities with these samples indicates outcomes much better than with the whole sample range.
This relative distribution of low blood lead samples appears to have continued to the present.
9-19
-------
The National Bureau of Standards has' recently made available certified porcine blood lead
standard reference material (SRM 955) at two levels of blood lead. Certified urine lead
samples are also being offered.
9.3 DETERMINATION OF ERYTHROCYTE PORPHYRIN (FREE ERYTHROCYTE PROTOPORPHYRIN,
ZINC PROTOPORPHYRIN)
9.3.1 Methods of Erythrocyte Porphyrln Analysis
Lead exposure results 1n inhibition of the final step in heme biosynthesis, the Insertion
of iron into protoporphyrln IX to form heme. Inhibition of this step leads to an accumulation
of the porphyrin, with zinc (II) occupying the position normally filled by Iron. Depending on
the particular method of analysis, zinc protoporphyrln (ZPP) Itself or the metal-free form,
free erythrocyte protoporphyrln (FEP), is measured. FEP generated as a consequence of chemi-
cal manipulation should be kept distinct from the metal-free form biochemically produced in
the disease, erythropoietic protoporphyria. The chemical or "wet" methods measure FEP or ZPP,
depending upon the relative acidity of the extraction medium. The hematofluorometer in its
commercially available form measures ZPP.
Porphyrins are labile due to photochemical decomposition; hence, samples must be protect-
ed from light during collection and handling and analyzed as soon as possible. Hematocrits
must also be obtained to adjust for anemic subjects.
In terms of methodological approaches for erythrocyte porphyrin (EP) analysis, virtually
all methods now in use exploit the ability of porphyrlns to undergo intense fluorescence when
excited at the appropriate wavelength of light. Such fluorometric techniques can be further
classified as wet chemical micromethods or as mlcromethods using a recently developed Instru-
ment, the hematofluorometer. The latter Involves direct measurement In whole blood. Because
the mammalian erythrocyte contains all of the EP in whole blood, either packed cells or whole
blood may be used, although the latter 1s more expedient.
Because of the relatively high sensitivity of fluorometric measurement for FEP or ZPP,
laboratory methods for spectrofluorometHc analysis require a relatively small sample of
blood; hence, microtechniques are currently the most popular in most laboratories. These in-
volve either liquid samples or blood collected on filter paper, the latter used particularly
in field sampling.
As noted above, chemical methods for EP analysis measure either FEP, where zinc 1s chemi-
cally removed, or ZPP, where zinc is retained. The procedures of Plomelli and Davidow (1972),
Granick et al., (1972), and Chisholm and Brown (1975) typify "free" EP methods, while those of
Lamola et al. (1975), Joselow and Flores (1977), and Chisolm and Brown (1979) involve mea-
surement of zinc EP.
9-20
-------
In Piomelli and Davidow's (1972) microprocedure, small volumes of whole blood, analyzed
either directly or after collection on filter paper, were treated with a suspension of Celite
in saline followed by a 4:1 mixture of ethyl acetate to glacial acetic acid. After agitation
and centrifugation, the supernatant was extracted with 1.5N HC1. The acid layer was analyzed
fluorometrically using an excitation wavelength of 405 nm and measurement at 615 nm. Blood
collected on filter paper discs was first eluted with 0.2 ml H20. The filter paper method was
found to work just as well as liquid samples of whole blood. Protoporphyrin IX was employed
as a quantitative standard. Granick et al. (1972) used a similar microprocedure, but it dif-
fered 1n the concentration of acid employed and the use of a ratio of maxima.
In Chisolm and Brown's (1975) variation, volumes of 20 ^1 of whole blood were treated
with ethyl acetate/acetic add (3:1) and briefly mixed. The acid-extraction step was done
with 3N HC1, followed by a further dilution step with more acid if the value was beyond the
range of the calibration curve. In this procedure, protoporphyrin IX was used as the working
standard, with coproporphyrin (a precursor to protoporphyrin) used to monitor the calibration
of the fluorometer and any variance with the protoporphyrin standard.
Lamola et al. (1975) analyzed the ZPP as such in their procedure. Small volumes of blood
(20 pi) were worked up in a detergent (dimethyl dodecylamine oxide) and phosphate buffer solu-
tion, and fluorescence was measured at 594 nm with excitation at 424 nm. In the variation of
Joselow and Flores (1977), 10 ul of whole blood was diluted 1000-fold, along with protoporphy-
rin (Zn) standards, with the detergent-buffer solution. Note that the ZPP standard is virtu-
ally impossible to obtain in pure form. Chisolm and Brown (1979) reported the use of proto-
porphyrin IX plus very pure zinc salt for such standards.
In the single-extraction variation of Orfanos et al. (1977), liquid samples of whole
blood (40 ul) or blood on filter paper were treated with acidified ethanol. The mixtures were
agitated and centrlfuged, and the supernatants analyzed directly in fluorometer cuvettes. For
blood samples on filter paper, blood was first leached from the paper with saline by soaking
for 60 min. Coproporphyrin was used as the quantitative standard. The correlation coeffici-
ent with the Piomelli and Davidow (1972) procedure (see above) over the range 40-650 ug EP/dl
RBCs was r = 0.98. As in the above methods, ZPP itself is measured.
Regardless of the extraction methods used, some instrumental parameters are important,
including the variation between cut-offs in secondary emission filters and variation among
photomultipller tubes in the red region of the spectrum. Hanna et al. (1976) compared four
micromethods for EP analysis: double extraction with ethyl acetate/acetic acid and with HC1
(Piomelli and Oavidow, 1972), single extraction with either ethanol or acetone (Chisolm et
al., 1974), and direct solubilization with detergent (Lamola et al., 1975). Of these, the
ethyl acetate and ethanol procedures were satisfactory; complete extraction occurred only with
9-21
-------
the ethyl acetate/acetic add method. In the method of Chlsholm et al. (1974), the choice of
add and its concentration appears to be more significant than the choice of organic solvent.
The levels of precision with these wet micromethods differ with the specifics of analy-
sis. Plomelli (1973) reported a coefficient of variation (C.V.) of 5 percent, compared to
Herber's (1980) observation of 2-4 percent for the methods per se and 6-11 percent total C.V.,
which Included precision of samples, standards, and day-to-day variation. The Lamola et al.
(1975) method for ZPP measurement was found to have a C.V. of 10 percent (same day, presuma-
bly), whereas Herber (1980) reported a day-to-day C.V. of 9.3-44.6 percent. Herber (1980)
also found that the wet chemical micromethod of PiomelH (1973) had a detection limit of 20 (jg
EP/dl whole blood, while that of Lamola et al. (1975) was sensitive to 50 ug EP/dl whole
blood.
The recent development of direct Instrumental measurement of ZPP with the hematofluoro-
meter has made 1t possible to use EP measurement in field screening for lead exposure in large
groups of subjects. However, hematofluorometers were developed for and remain most useful for
lead screening programs; they were not meant to be laboratory substitutes for the chemical
methods of EP analysis. (See Section 9.3.2 for a comparative discussion.) As originally de-
veloped by Bell Laboratories (Blumberg et al., 1977) and now produced commercially, the appa-
ratus employs front-face optics, 1n which excitation of the fluorophore 1s at an acute angle
to the sample surface, with emitted light emerging from the same surface and thus being de-
tected. Routine calibration requires a stable fluorescing material with spectra comparable to
ZPP; the triphenylmethane dye Rhodamlne B 1s used for this purpose. Absolute calibration re-
quires adjusting the microprocessor-controlled readout system to read the known concentration
of ZPP in reference blood samples, the latter calibration performed as frequently as possible.
Hematofluorometers are designed for measuring EP 1n samples containing oxyhemoglobln,
I.e., capillary blood. Venous blood, therefore, must first be oxygenated, usually by moderate
shaking for approximately 10 m1n (Blumberg et al., 1977; Grandjean and Llntrup, 1978). A
second problem with hematofluorometer use, in contrast to wet chemical methods, is Interfer-
ence by bilirubln (Karadc et al., 1980; Grandjean and Llntrup, 1978). This Interference oc-
curs with relatively low levels of EP. At levels normally encountered 1n lead workers or sub-
jects with anemia or nonoccupatlonal lead exposure, the degree of such Interference 1s not
considered significant (Grandjean and Llntrup, 1978). Karadc et al. (1980) have found that
carboxyhemoglobln (COHb) may pose a potential problem, but its relevance to EP levels of sub-
jects exposed to lead has not been fully elucidated. Background fluorescence 1n cover glass
may be a problem and should be tested 1n advance. Finally, the accuracy of the hematofluoro-
meter appears to be affected by hemolyzed blood.
Competently employed, the hematofluorometer appears to be reasonably precise, but Its ac-
curacy may still be biased (see below). Blumberg et al. (1977) reported a C.V. of 3 percent
9-22
-------
over the entire range of ZPP values measured when using a prototype apparatus. Karacic et al.
(1980) found the relative standard deviation to vary from 1 percent (0.92 mM ZPP/M Hb) to 5
percent (0.41 mM ZPP/M Hb) depending on concentration. Grandjean and Lintrup (1978) obtained
a day-to-day C.V. of 5 percent using blood samples refrigerated for up to 9 weeks. Herber
(1980) obtained a total C.V. of 4.1-11.5 percent.
A number of Investigators have compared EP measured by the hematofluorometer with EP mea-
sured by the laboratory or wet chemical techniques, ranging from a single, intralaboratory
comparison to Interlaboratory performance testing. The latter Included the EP proficiency
testing program of the USPHS1 CDC. Working with prototype instrumentation, Blumberg et al.
(1977) obtained correlation coefficients of r = 0.98 (range: 50-800 pg EP/dl RBCs) and 0.99
(range: up to 1000 pg EP/dl RBCs) for comparisons with the Granick and Piomelli methods,
respectively. Grandjean and Lintrup (1978), Castoldl et al. (1979) and Karacic et al. (1980)
have achieved equally good correlation results.
Several reports (Culbreth et al., 1979; Scoble et al., 1981; Smith et al., 1980) have
described the application of high-performance liquid chromatography (HPLC) to the analysis of
either FEP or ZPP 1n whole blood. In one of the studies (Scoble et al., 1981), the protopor-
phyrlns as well as coproporphyrln and mesoporphyrin IX were reported to be determined on-line
fluorometrically in less than 6 m1n using 0.1 ml of blood sample. The HPLC approach remains
to be tested 1n Interlaboratory proficiency programs.
9.3.2 Interlaboratory Testing of Accuracy and Precision 1n EP Measurement
In a relatively early attempt to assess interlaboratory proficiency 1n EP measurement,
Jackson (1978) reported results of a survey of 65 facilities that analyzed 10 whole blood sam-
ples by direct measurement with the hematofluorometer or by one of the wet chemical methods.
In this survey, the instrumental methods had a low bias compared to the extraction techniques
but tended to show better Interlaboratory correlation.
At present, CDC's ongoing EP proficiency testing program constitutes the most comprehen-
sive assessment of laboratory performance (U.S. Centers for Disease Control, 1981). Every
month, three samples of whole blood prepared at the University of Wisconsin Laboratory of
Hygiene are forwarded to participants. Reference means are determined by a group of reference
laboratories with a target range of ±15 percent across the whole range of EP values. For FY
1981, of the 198 laboratories participating, 139 facilities were Involved for the entire year.
Three of the 36 samples 1n the year were not Included. Of the 139 year-long participants,
93.5 percent had better than half of the samples within the target range, 84.2 percent per-
formed satisfactorily with 70 percent or more of the samples within range, and 50.4 percent of
all laboratories had 90 percent or more of the samples yielding the correct results. The par-
ticipants as a whole showed greater proficiency than in the previous year. Of the various
9-23
-------
methods currently used, the hematofluorometer direct measurement technique was most heavily
represented. For example, in the January 1982 survey of the three major techniques, 154 par-
ticipants used the hematofluorometer, 30 used the Piomelli method, and 7 used the Chlsolm/
Brown method.
A recent survey by Balamut et al. (1982) raises the troublesome observation that the use
of commercially available hematofluorometers may yield satisfactory proficiency results but
still be inaccurate when compared to the wet chemical method using freshly drawn whole blood.
Two hematofluorometers In wide use performed well in proficiency testing but showed an appro-
ximately 30 percent negative bias with clinical samples analyzed by both instrument and chemi-
cal microtechniques. This bias leads to false negatives when used in screening. Periodic
testing of split samples by both fluorometer and chemical means Is necessary to monitor, and
correct for, instrument negative bias. The basis of the bias is much more than can be ex-
plained by the difference between FEP and ZPP. This survey points out precautions noted
earlier on the restrictive use of the hematofluorometer to screening situations.
Mitchell and Doran (1985) compared EP values measured in their laboratory by the chemical
extraction technique with results obtained by the hematofluorometer in 21 other laboratories.
These workers found that (a) hematofluorometer results were 11-28 percent lower than the cor-
responding chemical method values, (b) hematofluorometers demonstrated mean error of up to 3
percent for proficiency samples, and (c) hematofluorometers showed a negative bias of 20 per-
cent at EP levels of 50 M9/dl and would miss about one third (false negatives) of children at
or somewhat above this level.
One factor that can be important in the relative accuracy of the hematofluorometer versus
wet chemical methods is the relative stability of ZPP levels as a proportion of total EP
across that age range in childhood of most interest in screening. Hammond and coworkers
(1984) have observed that the fraction of ZPP versus total EP was at a relative minimum at 3
months of age in 165 children serially tested, and that 1t increased to 1.0 by around 33
months of age. These observations suggest that this variation of proportionality with age
should be taken into account when screening children under approximately 30 months of age and
when the hematofluorometer is the chief means of EP quantification.
The technical basis for this age-related change 1n proportionality may be spectroscoplc,
i.e., changes in erythrocyte size over this age range would lead to differences in cell
packing, which in turn would affect fluorescence yield during front-face Irradiation 1n the
hematofluorometer. A second factor noted by the authors may have to do with relative availa-
bility of zinc. Since zinc deficiency is common at this stage of development (see Chapter
10), bioavailability of zinc for a nonessential complexing with FEP would be restricted by
homeostatic sparing of the element for physiological needs. However, since the work of
9-24
-------
Chisolm and Brown (1979), using a chemical method, did not reveal any disparity between the
two forms in subjects of the same age range, there is probably an instrumental artifact
operating here.
9.4 MEASUREMENT OF URINARY COPROPORPHYRIN
The elevation of urinary coproporphyrin (CP-U) with lead intoxication served as a useful
indicator of such intoxication in children and lead workers for many years. Although analysis
of CP-U has declined considerably in recent times with the development of other testing
methods, such as measurement of EP, it still has the advantage of showing active intoxication
(Piomelli and Graziano, 1980).
The standard method of CP-U determination is the fluorometric procedure described by
Schwartz et al. (1951). Urine samples are treated with acetate buffer and aqueous iodine, the
latter converting coproporphyrinogen to coproporphyrin (CP). The porphyrin is partitioned
into ethyl acetate and back extracted (4 times) with 1.5N HC1. Coproporphyrin is employed as
the quantitative standard. Working curves are linear below 5 pg CP/1 urine.
In the absorption spectrometric technique of Haeger-Aronsen (1960), iodine is also used
to convert coproporphyrinogen to CP. The extractant is ethyl ether, from which the CP is re-
moved with 0.1N HC1. Absorption is read at three wavelengths, 380, 430, and the Soret maximum
at 402 nm. Quantification is carried out using an equation involving the three wavelengths.
9.5 MEASUREMENT OF DELTA-AMINOLEVULINIC ACID DEHYDRASE ACTIVITY
Delta-aminolevulinic acid dehydrase (5-aminolevulinate hydrolase; porphobilinogen synthe-
tase; E.C. 4.2.1.24; i.e., ALA-D) is an allosteric sulfhydryl enzyme that mediates the con-
version of two units of 6-aminolevulinic acid (6-ALA) to porphobilinogen, a precursor in the
heme biosynthetic pathway to the porphyrins. Lead's inhibition of the activity of this enzyme
is the enzymological basis of ALA-D's diagnostic utility in assessing lead exposure using
erythrocytes.
A number of sampling precautions are necessary when measuring this enzyme's activity.
ALA-D activity is modified by the presence of zinc as well as lead. Consequently, blood col-
lection tubes that have high background zinc content, mainly in the rubber stoppers, must be
avoided completely or care must be taken to avoid stopper contact with blood. Nackowski et
al. (1977) observed that the presence of zinc in blood collection tubes is a pervasive prob-
lem, and plastic-cup tubes appear the only practical means to avoid it. To guard against zinc
in the tube itself, one should determine the extent of zinc Teachability by blood and use one
9-25
-------
tube lot, if possible. Heparin is the anticoagulant of choice, because the lead binding
agent, EDTA, or other chelants would affect the lead-enzyme interaction. The relative in-
stability of the enzyme in blood makes rapid determinations of activity necessary, preferably
as soon after collection as possible. Even with refrigeration, analysis of activity should be
done within 24 hr (Berlin and Schaller, 1974). Furthermore, porphobilinogen is light labile,
which requires that the assay be done under restricted light.
Various procedures for ALA-D activity measurement are chemically based on measurement of
porphobilinogen generated from the substrate. Delta-ALA porphobilinogen is condensed with p-
dimethylaminobenzaldehyde (Ehrlich's reagent) to yield a chromophore measured at 553 nm in a
spectrophotometer. In the European Standardized Method for ALA-D activity measurement (Berlin
and Schaller, 1974), developed with the collaboration of nine laboratories for use with blood
samples having relatively low lead content, triplicate blood samples (0.2 ml) are hemolyzed,
along with a blood blank, with water for 10 min at 37°C. Samples are then mixed with 6-ALA
solution and incubated for 60 min. The enzyme reaction is terminated by addition of a solu-
tion of mercury (II) in trichloroacetic acid, followed by centrifugation and filtration. Fil-
trates are mixed with modified Ehrlich's reagent (p-d1methylaminobenzalehyde in trichloro-
acetic/perchloric acid mixture) and allowed to react for 5 m1n, followed by chromophore
measurement in a spectrophotometer at 555 nm. Activity is quantified in terms of |jM 6-ALA/
min«l erythrocytes. Note that the amount of phosphate for Solution A in Berlin & Schaller's
(1974) report should be 1.78 g, not the 1.38 g stated. In a microscale variation, Granick et
al. (1973) used only 5 pi of blood and terminated the assay by trichloroacetic acid.
In comparing various reports concerning the relationship between lead exposure and ALA-D
inhibition, attention should be paid to the units of activity measurement employed with the
different techniques. Berlin and Schaller's (1974) procedure expresses activity as uM 6-ALA/
min*l cells, while Tomokuni's (1974) method expresses activity as pM porphobilinogen/hr/ml
cells. Similarly, when comparing the Bonsignore et al. (1965) procedure to that of Berlin and
Schaller (1974), a conversion factor of 3.8 is necessary when converting from Bonsignore to
European Standard Method units (Trevisan et al., 1981).
Several factors have been shown to affect ALA-D activity. Rather than measuring enzyme
activity 1n blood once, Granick et al. (1973) measured activity before and after treatment
with dlthiothreitol, an agent that reactivates the enzyme by complexing lead. The ratio of
activated to unactivated enzymes versus blood lead levels accommodates Inherent differences in
enzyme activity among individuals due to genetic factors and other reasons. Other agents for
such activation include zinc (FinelH et al., 1975) and zinc plus glutathione (Mitchell et
al., 1977). In the Mitchell et al. (1977) study, nonphyslological levels of zinc were used.
Wigfield and Farant (1979) found that enzyme activity is related to assay pH; thus, reduced
9-26
-------
activity from such a pH-act1v1ty relationship could be misinterpreted as lead Inhibition.
These researchers find that pH shifts away from optimal, 1n terms of activity, as blood lead
content Increases and the Incubation step proceeds.
9.6 MEASUREMENT OF DELTA-AMINOLEVULINIC ACID IN URINE AND OTHER MEDIA
Delta-ara1nolevul1n1c add (6-ALA) levels increase with elevated lead exposure, because of
the Inhibitory effect of lead on the activity of ALA dehydrase and/or the increase of ALA syn-
thetase activity by feedback derepression. The result is that this intermediate in heme bio-
synthesis rises in the body and eventually results 1n Increased urinary excretion. The meas-
urement of this metabolite in urine provides an indication of the level of lead exposure.
The ALA content of urine samples (ALA-U)is stable for approximately 2 weeks or more if
urine samples are acidified with tartaric or acetic acid and kept refrigerated. Values of
ALA-U are adjusted for urine density if concentration is expressed in mg/1 or is measured per
gram creatinine. As noted in the case of urinary lead measurement, 24-hr collection is more
desirable than spot sampling.
Five manual procedures and one automated procedure for urinary ALA measurement are most
widely used. Mauzerall and Granick (1956) and Davis and Andelman (1967) described the most
involved procedures, requiring the initial chromatographic separation of ALA. The approach of
Grabecki et al. (1967) omitted chromatographic Isolation, whereas the automated variation of
Lauwerys et al. (1972) omitted prechronatography but included the use of an internal standard.
Tomokunl and Ogata (1972) omitted chromatography but employed solvent extraction to isolate
the pyrrole intermediate.
Mauzerall and Granick (1956) condensed ALA with a p-dlcarbonyl compound, acetylacetone,
at pH 4.6 to yield a pyrrole intermediate (Knorr condensation reaction), which was further re-
acted with p-dimethylaminobenzaldehyde in perchloric/acetic acid. The samples were then read
1n a spectrophotometer at 553 nm 15 min after mixing. In this method, both porphobilinogen
and ALA are separated from urine by means of a dual-column configuration of cation and anion
exchange resins. The latter retains the porphobilinogen and the former separates ALA from
urea. The detection limit is 3 umol/1 urine. In the modification of this method by Davis and
Andelman (1967), disposable catlon/anion resin cartridges were used, in a sequential configu-
ration, to expedite chromatographic separation and Increase the sample analysis rate. Commer-
cial (Bio-Rad) disposable columns based on this design are now available and appear satis-
factory.
In these two approaches (Mauzerall and Granick, 1956; Davis and Andelman, 1967), the pro-
blem of interference due to aminoacetone, a metabolite occurring in urine, is not taken into
account. However, Marver et al. (1966) used Dowex-1 in a chromatographic step subsequent to
9-27
-------
the condensation reaction to form the pyrrole. This step separates the ALA derivative from
that of the aminoacetone. Similarly, Schlenker et al. (1964) used a cation-exchange column to
retain aminoacetone.
Tomokuni and Ogata (1972) condensed ALA with ethylacetoacetate and extracted the re-
sulting pyrrole with ethyl acetate. The extract was then treated with Ehrlich's reagent and
the resulting chromophore measured spectrophotometrically. Lauwerys et al. (1972) developed
an automated ALA analysis method for lead worker screening in which ALA was added 1n known
amount as an internal standard and the prechromatography was avoided. They reported a high
correlation (r = 0.98, no range available) with the procedure of Mauzerall and Granick (1956).
Roels et al. (1974) compared the relative proficiency of four methods--those of
Mauzerall and Granick (1956), Davis and Andelman (1967), the Lauwerys et al. (1972) automated
version, and the Grabecki et al. (1967) method, which omits chromatographic separation and is
normally used with occupational screening. The chromatographic methods gave identical results
over the range of 0-60 mg ALA/1 urine, while the automated method showed a positive bias at <6
mg/1. The Grabecki et al. (1967) technique was the least satisfactory of the procedures com-
pared. Roels et al. (1974) also noted that commercial ion-exchange columns resulted in low
variability (<10 percent).
Delia Florentine et al. (1979) combined the Tomokuni and Ogata (1972) extraction method
with a correction equation for urine density. Up to 25 mg ALA/1, the C.V. was £4 percent
along with a good correlation (r = 0.937) with the Davis and Andelman (1967) technique. While
avoiding prechromatography saves time, one must prepare a curve relating urine density to a
correction factor for quantitative measurement.
Although ALA analysis is normally done with urine as the indicator medium, Haeger-Aronsen
(1960) reported a similar colorimetric method for blood and MacGee et al. (1977) described a
gas-liquid chromatographic (GLC) method for ALA in plasma as well as urine. Levels of ALA in
plasma are much lower than those in urine. In the latter method, ALA was isolated from
plasma, reacted with acetyl-acetone, and partitioned into a solvent (trimethylphenyl-
hydroxide), which also served for pyrolytic methylation in the injection port of the gas-
liquid chromatograph; the methylated pyrrole was more amenable to chromatographic isolation
than the more polar precursor. For quantification, an internal standard, 6-amino-5-oxohexa-
noic acid, was used. The sample requirement is 3 ml plasma. Measured levels ranged from 6.3
to 73.5 ng ALA/ml plasma, and yielded values that were approximately tenfold lower than the
colorimetric techniques (O'Flaherty et al., 1980).
In comparing the Haeger-Aronsen (1960) and MacGee et al. (1977) methods, a number of dif-
ferences should be pointed out. First, the colorimetric approach of Haegar-Aronsen does not
employ chromatographic steps to separate the ALA from other aminoketones, specifically amino-
acetone and porphobilinogen. While these other aminoketones are not known to be positively
9-28
-------
correlated with blood lead, they would add a positive bias to the accuracy of the levels ob-
tained. The GLC method of MacGee and coworkers does not measure simultaneously these amino-
ketones in either plasma or urine, and a reading of the published methodology and its applica-
tion (O'Flaherty et al., 1980) indicates the procedure is acceptable for urinary ALA and
levels of ALA in plasma associated with relatively high blood lead values, i.e., >40 ug/dl.
The suitability of the GLC approach for relatively low levels of plasma ALA, i.e., at blood
lead levels below 40 pg/dl, remains to be fully evaluated in the field. A careful reading of
the MacGee et al. report suggests potential interferences with low levels of ALA measurement,
while the methodology has not had wide use or multi-laboratory evaluation. Despite its added
cost, a good overall method for assessing the relationship of plasma ALA to blood lead levels
below 40 (jg/dl, now an issue of controversy (see Chapter 12.3), would be use of the MacGee
method in tandem with computerized multiple-ion monitoring in a mass spectrometer. This
method is an absolute means of ALA identification as well as a sensitive means of quantifica-
tion.
9.7 MEASUREMENT OF PYRIMIDINE-5'-NUCLEOTIDASE ACTIVITY
Erythrocyte pyrimidine-5'-nuc1eotidase (5'-ribonucleotide phosphohydrolase, E.G. 3.1.3.5,
i.e., Py5N) catalyzes the hydrolytic dephosphorylation of the pyrimidine nucleotides uridine
monophosphate (UMP) and cytidinemonophosphate (CMP) to uridine and cytidine (Paglia and Valen-
tine, 1975). Enzyme inhibition by lead in humans and animals results in incomplete degrad
ation of reticulocyte ribonucleic acid (RNA) fragments, accumulation of the nucleotides, and
increased cell hemolysis (Paglia et al., 1975; Paglia and Valentine, 1975; Angle and Mclntire,
1978; George and Duncan, 1982).
Two methods are available for measurement of Py5N activity. One is quite laborious in
terms of time and manipulation, while the other is shorter but requires the use of radioiso-
topes and radiometric measurement. In Paglia and Valentine's (1975) method, heparinized
venous blood was filtered through cotton or a commercial cellulose preparation to separate
erythrocytes from platelets and leukocytes. Cells were given multiple saline washings, packed
lightly, and subjected to freeze hemolysis. The hemolysates were dialyzed against a saline-
Tris buffer containing MgCl2 and EDTA to remove nucleotides and other phosphates. The assay
system consists of dialyzed hemolysate, MgCl2, Tris buffer at pH 8.0, and either UMP or CMP;
incubation is for 2 hr at 37°C. Activity is terminated by treatment with 20 percent trichlo-
roacetic acid, followed by centrifugation. The supernatant inorganic phosphate, P^ is meas-
ured by the classic method of Fiske and Subbarow (1925), and the phosphomolybdic acid complex
is measured spectrophotometrically at 660 nm. A unit of enzyme activity is expressed as
9-29
-------
umol P.j/hr/g hemoglobin. Hemolysates appear to be stable (90 percent) with refrigeration at
4°C for up to 6 days, provided that mercaptoethanol is added at the time of assay. Like the
other method, activity measurement requires the determination of hemoglobin.
In the simpler approach of Torrance et al. (1977), which can be feasibly applied to much
larger numbers of samples, erythrocytes were separated from leukocytes and platelets with a
1:1 mixture of microcrystalline and alphacellulose, followed by saline washing and hemolysis
with a solution of mercaptoethanol and EDTA. Hemolysates were incubated with a medium con-
taining purified 14C-CMP and MgCl2 for 30 min at 37°C. The reaction was terminated by sequen-
tial addition of barium hydroxide and zinc sulfate solution. Proteins and unreacted nucleo-
tide were precipitated, leaving the labeled cytidine in the supernatant. Aliquots were
measured for 14C-activity in a liquid scintillation counter. Enzyme activity was expressed as
nM CMP/min/g hemoglobin. The blank activity was determined for each sample by carrying out
the precipitation step as soon as the hemolysate was mixed with the labeled CMP, i.e., t = 0.
This procedure shows a good correlation (r = 0.94; range: 135-189 enzyme units) with the
method of Paglia and Valentine (1975). The two methods express units of enzyme activity dif-
ferently, so that one must know which method is used when comparing enzyme activity.
9.8 MEASUREMENT OF PLASMA 1,25-DIHYDROXYVITAMIN D
The active form of vitamin D in bone mineral metabolism, including absorption of calcium
and phosphorus as well as bone resorption of these minerals, is the hormonal metabolite,
1,25-dihydroxyvitamin D (1,25-(OH)2D). Given the growing interest in the adverse effects of
lead on the biosynthesis of this crucial metabolite (see Chapters 10, 12 and 13), a brief
discussion of the quantitative measurement of this metabolite is merited. Techniques for
measurement of 1,25-(OH)2D are all of recent vintage, are all rather lengthy procedurally, and
all require a rather high level of laboratory expertise and proficiency.
Reported methodology, whatever the differences in specific details, can be broken down
into three discrete steps: (1) isolation of the metabolite from plasma or serum by liquid-
liquid extraction using solvents common in lipid analysis; (2) preconcentration of the ex-
tracts and chromatographic purification using Sephadex LH-20 or Lipidex 5000 columns along
with, in some cases, HPLC; and (3) subsequent quantitation by either of two radiometric bind-
ing techniques: the more common competitive protein binding (CPB) assay or radioimmunoassay
(RIA). The CPB assay normally involves the use of a receptor protein in the intestinal cyto-
sol of chicks made vitamin D-deficient.
Most illustrative of 1,25-(OH)2D measurement is the technique of Shepard et al. (1979),
which also includes steps for the analysis of other metabolites not discussed here. Human
9-30
-------
plasma, 3-5 ml, to which Initiated metabolite is added as tracer internal standard, is ex-
tracted with a mixture of methanol and methylene chloride, followed by separation of the
(OH)2D fraction (to include the 24,25- and 25,26-(OH)2 metabolites) from other metabolites
using a Sephadex LH-20 column. Subsequent use of HPLC (straight phase, Zorbax-SIL) separates
the 1,25-(OH)2 metabolite from the other two dihydroxylated products. Quantification is by
CPB assay. In human adults, the mean metabolite level is 31 picograms/ml. Limit of detection
is 5 picograms/analytical tube, mean recovery is 58.4 percent, and the within-run and between-
run coefficients of variation are 17 and 26 percent, respectively.
Two interlaboratory surveys of methodology for vitamin D metabolite analysis have recent-
ly been described (Jongen et a!., 1982; Jongen et al., 1984). In the more recent and compre-
hensive of the two (Jongen et al., 1984), 15 laboratories carried out analyses of eight plasma
samples and two standards for 1,25-(OH)2D. Mean interlaboratory coefficient of variation for
analysis of 1,25-(OH)2D in the plasma samples was 52 percent. In this survey, nine labora-
tories used the CPB assay, with six using RIA for quantitation. The major reason, however,
for the variance appeared to be differences in methods of purification. The upshot of this
survey is that results for a given sample will vary with specifics of procedure. Thus each
laboratory should establish its own reference values.
9.9 SUMMARY
A complete understanding of a toxic agent's biological effects (including any statement
of dose-effect relationships) requires quantitative measurement of either that agent in some
biological medium or a physiological parameter associated with exposure to the agent. Quanti-
tative analysis Involves a number of discrete steps, all of which contribute to the overall
reliability of the final analytical result: sample collection and shipment, laboratory hand-
ling, instrumental analysis, and criteria for internal and external quality control.
From a historical perspective, the definition of "satisfactory analytical method" for
lead has been changing steadily as new and more sophisticated equipment has become available
and understanding of the hazards of pervasive contamination along the analytical course has
Increased. The best example of this change is the current use of the definitive method for
lead analysis, isotope-dilution mass spectrometry (IDMS) 1n tandem with "ultra-clean" facili-
ties and sampling methods, to demonstrate conclusively not only the true extent of anthropo-
aenic Input of lead to the environment over the years but also the relative limitations of
most of the methods used today for lead measurement.
9-31
-------
9.9.1 Determinations of Lead in Biological Media
The low levels of lead in biological media, even in the face of excessive exposure, and
the fact that sampling of such media must be done against a backdrop of pervasive lead contam-
ination necessitates that samples be collected and handled carefully. Blood lead sampling is
best done by venous puncture and collection into low-lead tubes after careful cleaning of the
puncture site. The use of finger puncture as an alternative method of sampling should be
avoided, 1f feasible, given the risk of contamination associated with the practice In indus-
trialized areas. While collection of blood onto filter paper enjoyed some popularity in the
past, paper deposition of blood requires special correction for hematocrlt/hemoglobin level.
Urine sample collection requires the use of lead-free containers as well as addition of a
bactericide. If feasible, 24-hr sampling is preferred to spot collection. Deciduous teeth
vary in lead content both within and across type of dentition. Thus, a specific tooth type
should be uniformly obtained for all study subjects and, if possible, more than a single sam-
ple should be obtained from each subject.
Measurements of Lead in Blood. Many reports over the years have purported to offer
satisfactory analysis of lead in blood and other biological media, often with severe inherent
limitations on accuracy and precision, meager adherence to criteria for accuracy and pre-
cision, and a limited utility across a spectrum of analytical applications. Therefore, it is
only useful to discuss "definitive" and, comparatively speaking, "reference" methods currently
in use.
In the case of lead in biological media, the definitive method is isotope-dilution mass
spectrometry (IDMS). The accuracy and unique precision of IDMS arise from the fact that all
manipulations are on a weight basis involving simple procedures, and measurements entail only
lead isotope ratios and not the absolute determinations of the isotopes Involved, which
greatly reduces instrumental corrections and errors. Reproducible results to a precision of
one part in 104-105 are routine with appropriately designed and competently operated instru-
mentation. Although this methodology is still not recognized in many laboratories, it was the
first breakthrough, 1n tandem with "ultra-clean" procedures and facilities, in definitive
methods for Indexing the progressive increase in lead contamination of the environment over
the centuries. Given the expense, required level of operator expertise, and time and effort
involved for measurements by IOMS, this method mainly serves for analyses that either require
extreme accuracy and precision, e.g., geochronometry, or for the establishment of analytical
reference material for general testing purposes or the validation of other methodologies.
While the term "reference method" for lead in biological media cannot be rigorously ap-
plied to any procedures in popular use, the technique of atomic absorption spectrometry (AAS)
in its various configurations, or the electrochemical method, anodic stripping voltammetry
(ASV), come closest to meriting the designation. Other methods that are generally applied in
9-32
-------
metal analyses are either limited in sensitivity or are not feasible for use on theoretical
grounds for lead analysis.
MS as applied to analysis of whole blood, generally involves flame or flameless micro-
methods. One macromethod, the Hessel procedure, still enjoys some popularity. Flame micro-
nalvsis the Delves cup procedure, applied to blood lead appears to have an operational sen-
sitivity of about 10 ug/dl blood and a relative precision of approximately 5 percent in the
range of blood lead seen in populations in industrialized areas. The flameless, or electro-
thermal, method of AAS enhances sensitivity about tenfold, but precision can be more proble-
matic because of chemical and spectral interferences.
The most widely used and sensitive electrochemical method for lead in blood is ASV. For
the most accurate results, chemical wet ashing of samples must be carried out, although this
cess -s time-consuming and requires the use of lead-free reagents. The use of metal ex-
change reagents has been employed in lieu of the ashing step to liberate lead from binding
sites although this substitution is associated with less precision. For the ashing method,
relative precision is approximately 5 percent. In terms of accuracy and sensitivity, problems
noear at low levels, e.g., 5 pg/dl or below, particularly if samples contain elevated copper
levels.
Lead in Plasma. Since lead in whole blood is virtually all confined to the erythrocyte,
lasma levels are quite low and extreme care must be employed to measure plasma levels relia-
The best method for such measurement is IDMS, in tandem with ultra-clean facility use.
is satisfactory for comparative analyses across a range of relatively high whole blood
values.
Lead in Teeth. Lead measurement in teeth has involved either whole tooth sampling or
nalvsis of specific regions, such as dentine or circumpulpal dentine. In either case, sam-
nles must be solubilized after careful surface cleaning to remove contamination; solubili-
zation is usually accompanied by either wet ashing directly or ashing subsequent to a dry
ashing step.
AAS and anodic stripping have been employed more frequently for such determinations than
nv other method. With AAS, the high mineral content of teeth argues for preliminary isola-
. of 1eacj via chelation/extraction. The relative precision of analysis for within-run mea-
ement is around 5-7 percent, with the main determinant of variance in regional assay being
initial isolation step. One change from the usual methods for such measurement is the j_n
itu measurement of lead by X-ray fluorescence spectrometry in children. Lead measured in
tiTs fashion allows observation of ongoing lead accumulation, rather than waiting for exfolia-
tion.
Lead in Hair. Hair as an exposure indicator for lead offers the advantages of being non-
sive and a medium of indefinite stability. However, the crucial problem of external
9-33
-------
surface contamination is such that it is still not possible to state that any cleaning
protocol reliably differentiates between externally and internally deposited lead.
Studies that demonstrate a correlation between increasing hair lead and increasing sever-
ity of a measured effect tend to support arguments for using hair as an external indicator of
exposure. Probably, then, such measurement using cleaning protocols that have not been inde-
pendently validated will overstate the relative accumulation of "internal" hair lead in terms
of some endpoint and will also underestimate the relative sensitivity of changes in internal
lead content with exposure. One consequence of this would be, for example, an apparent
threshold for a given effect in terms of hair lead which is significantly above the actual
threshold. Because of these concerns, hair is best used with the simultaneous measurement of
blood lead.
Lead in Urine. Analysis of lead in urine is complicated by the relatively low levels of
the element in this medium as well as the complex mixture of mineral elements present. Urine
lead levels are most useful and also somewhat easier to determine in cases of chelation mobil-
ization or chelation therapy, where levels are high enough to permit good precision and dilu-
tion of matrix interference.
Samples are probably best analyzed by prior chemical wet ashing, using the usual mixture
of acids. Both ASV and AAS have been applied to urine analysis, with the latter more routine-
ly used and usually with a chelation/extraction step.
Lead in Other Tissues. Bone samples require cleaning procedures for removal of muscle
and connective tissue and chemical solubilization prior to analysis. Methods of analysis are
comparatively limited and flameless AAS is the technique of choice.
|n vivo lead measurements in bone of lead workers have been reported using X-ray fluores-
cence analysis and a radioisotopic source for excitation. One problem with this approach with
moderate lead exposure is the detection limit, approximately 20 ppm. Soft organ analysis
poses a problem in terms of heterogeneity in lead distribution within an organ (e.g., brain
and kidney). In such cases, regional sampling or homogenization must be carried out. Both
flame and flameless AAS appear to be satisfactory for soft tissue analysis and are the most
widely used.
Quality Assurance Procedures in Lead Analyses. In terms of available information, the
major focus in establishing quality control protocols for lead has involved whole blood meas-
urements. Translated into practice, quality control revolves around steps employed within the
laboratory, using a variety of internal checks, and the further reliance on external checks,
such as a formal continuing muHi-laboratory proficiency testing program.
Within the laboratory, quality assurance protocols can be divided into start-up and rou-
tine procedures, the former involving establishment of detection limits, within-run and
between-run precision, analytical recovery, and comparison with some reference technique
9-34
-------
within or outside the laboratory. The reference method is assumed to be accurate for th
ticular level of lead In some matrix at a particular point in time. Correlation with such a
method at a satisfactory level, however, may simply indicate that both methods are equal!
Inaccurate but performing with the same level of precision proficiency. More preferable is
the use of certified samples having lead at a level established by the definitive method
For blood lead, the Centers for Disease Control (CDC) periodically survey overall accu-
racy and precision of methods used by reporting laboratories. In terms of overall
and precision, one such survey found that ASV as well as the Delves cup and extraction var1C-
tlons of AAS performed better than other procedures. These results do not mean th
laboratory cannot perform better with a particular technique; rather, such data ar
ance for new facilities choosing among methods.
Of particular value to laboratories carrying out blood lead analysis are the external
quality assurance programs at both the State and Federal levels. The most comprehensive
proficiency testing program is that carried out by the CDC. This program actually consists of
two subprograms, one directed at facilities involved in lead poisoning prevention and screen-
ing (Center for Environmental Health) and the other concerned with laboratories
fication under the Clinical Laboratories Improvement Act of 1967 as well as u d
of the Occupational Safety and Health Administration's (OSHA) Laboratory ImprovTme^nTprogram
Office. Judging from the relative overall improvements in reporting laboratories over the
years of the programs' existence, the proficiency testing programs have served their purpose
well. In this regard, OSHA criteria for laboratory certification require that eight of nin
samples be analyzed correctly for the previous quarter. This level of required proficiency
reflects the ability of a number of laboratories to actually perform at this level
9-9-2 Determination of Erythrocyte Porphyrin (Free Ervthrocyt.p
Protoporphyrin)
With lead exposure, erythrocyte protoporphyrin IX accumulates because of impaired Dlac
ment of divalent iron to form heme. Divalent zinc occupies the place of the native i
Depending upon the method of analysis, either metal-free erythrocyte porphyrin (EP) or T
protoporphyrin (ZPP) Is measured, the former arising from loss of zinc in the chemical m*nT
pulation. Virtually all methods now in use for EP analysis exploit the abiHty of theT~
phyrin to undergo intense fluorescence when excited by ultraviolet light. Such fluorometHc
methods can be further classified as wet chemical micromethods or direct measuring fluoromet y
using the hematofluorometer. Because of the high sensitivity of such measurement relative^
small blood samples are required, with liquid samples or blood collected on finer JL?
The most common laboratory or wet chemical procedures now in use represent variation of
several common chemical procedures: (1) treatment of blood samples with a mixture of e hyl
-------
acetate/acetic acid followed by a repartitioning into an inorganic acid medium or (2) solu-
bilization of a blood sample directly into a detergent/buffer solution at a high dilution.
Quantification has been done using protoporphyrin, coproporphyrin, or zinc protoporphyrin IX
plus pure zinc ion. The levels of precision for these laboratory techniques vary somewhat
with the specifics of analysis. The Piomelli method has a coefficient of variation of 5
percent, while the direct ZPP method using buffered detergent solution is higher and more
variable.
The recent development of the hematofluorometer has made it possible to carry out EP
measurements in high numbers, thereby making population screening feasible. Absolute calibra-
tion is necessary and requires periodic adjustment of the system using known concentrations of
EP in reference blood samples. Since these units are designed for oxygenated blood (i.e.,
capillary blood), use of venous blood requires an oxygenation step, usually a moderate shaking
for several minutes. Measurement of low or moderate levels of EP can be affected by interfer-
ence with bilirubin. Competently employed, the hematofluorometer is reasonably precise, show-
ing a total coefficient of variation of 4.11-11.5 percent. While the comparative accuracy of
the unit has been reported to be good relative to the reference wet chemical technique, a very
recent study has shown that commercial units carry with them a significant negative bias
which may lead to false negatives in subjects having only moderate EP elevation. Such a bias
in accuracy has been difficult to detect in existing EP proficiency testing programs. By com-
parison to wet methods, the hematofluorometer should be restricted to field use rather than
becoming a substitute in the laboratory for chemical measurement, and this field use should
involve periodic split-sample comparison testing with the wet method.
9.9.3 Measurement of Urinary Coproporphyrin
Although EP measurement has largely supplanted the use of urinary coproporphyrin (CP-U)
analysis to monitor excessive lead exposure in humans, this measurement is still of value in
that it reflects active intoxication. The standard analysis is a fluorometric technique,
whereby urine samples are treated with buffer, and an oxldant (iodine) is added to generate CP
from its precursor. The CP-U is then partitioned into ethyl acetate and re-extracted with
dilute hydrochloric acid. The working curve is linear below 5 ug CP/dl urine.
9.9.4 Measurement of Delta-Aminolevulinic Acid Dehydrase Activity
Inhibition of the activity of the erythrocyte enzyme delta-aminolevulinic acid dehydrase
(ALA-D) by lead is the basis for using such activity in screening for excessive lead exposure.
A number of sampling and sample handling precautions attend such analysis. Since zinc (II)
ion will offset the degree of activity inhibition by lead, blood collecting tubes must have
extremely low zinc content, which essentially rules out the use of rubber-stoppered blood
9-36
-------
tubes. Enzyme instability necessitates that the activity measurement be carried out within 24
hr of blood collection. Porphobilinogen, the product of enzyme action, is light labile and
requires the assay be done in restricted light. Various procedures for ALA-D measurement are
based on measurement of the level of the chromophoric pyrrole (approximately 555 nm) formed by
condensation of the porphobilinogen with p-dimethylaminobenzaldehyde.
In the European Standardized Method for ALA-D activity determination, blood samples are
hemolyzed with water, ALA solution added, followed by incubation at 37°C, and the reaction
terminated by a solution of mercury (II) in trichloroacetic acid. Filtrates are treated with
modified Ehrlich's reagent (p-dimethylaminobenzaldehyde) in trichloroacetic/perchloroacetic
acid mixture. Activity is quantified in terms of micromoles 6-ALA/min-l erythrocytes
One variation in the above procedure 1s the initial use of a thiol agent, such as dithio-
threotol, to reactivate the enzyme, giving a measure of the full native activity of the
zyme. The ratio of activated/unactivated activity versus blood lead levels accommod t
genetic differences between individuals.
9.9.5 Measurement of Delta-Aminolevulinic Acid in Urine and Other Media
Levels of delta-aminolevulinic add (6-ALA) in urine and plasma increase with elevated
lead exposure. Thus, measurement of this metabolite, generally in urine, provides an index of
the level of lead exposure. ALA content of urine samples (ALA-U) is stable for about 2 weeks
or more with sample acidification and refrigeration. Levels of ALA-U are adjusted for urine
density or expressed per unit creatinine. If feasible, 24-hr collection is more desirable
than spot sampling.
Virtually all the various procedures for ALA-U measurement employ preliminary isolation
of ALA from the balance of urine constituents. In one method, further separation of ALA from
the metabolite aminoacetone 1s done. Aminoacetone can interfere with colorimetric measure
ent. ALA is recovered, condensed with a beta-dicarbonyl compound, e.g., acety1 acetone to
yield a pyrrole intermediate. This intermediate is then reacted with p-dimethylaminobenzal-
dehyde in perchloric/acetic add, followed by colorimetric reading at 553 nm In one vari-
ation of the basic methodology, ALA Is condensed with ethyl acetoacetate directly and the re-
sulting pyrrole extracted with ethyl acetate. Ehrlich's reagent is then added as in other
procedures and the resulting chromophore Is measured spectrophotometrically
Measurement of ALA in plasma is much more difficult than in urine, since plasma ALA is at
nanogram/milliter levels. In one gas-liquid chromatographic procedure, ALA is isolated from
plasma, reacted with acetyl acetone and partitioned into a solvent that also serves for pyro-
lytic mediation of the involatile pyrrole In the injector port of the chromatograph making
the derivative more volatile. For quantification, an internal standard, 6-amino-5-oxonexanoic
9-37
-------
acid, is used. While the method is more involved, it is more specific than the older colori-
metric technique.
9.9.6 Measurement of Pyrimidine-S'-Nucleotidase Activity
Erythrocyte pyrimidine-5'-nucleotidase (Py5N) activity is inhibited with lead exposure.
Currently, two different methods are used for assaying the activity of this enzyme. The older
method is quite laborious in time and effort, whereas the more recent approach is shorter but
uses radioisotopes and radiometric measurement.
In the older method, heparinized venous blood is filtered through cellulose to separate
erythrocytes from platelets and leukocytes. Cells are then freeze-fractured and the hemo-
lysates dialyzed to remove nucleotides and other phosphates. This dialysate is then incubated
in the presence of a nucleoside monophosphate and cofactors, the enzyme reaction being termi-
nated by treatment with trichloroacetic acid. The inorganic phosphate isolated from added
substrate is measured colorimetrically as the phosphomolybdic acid complex.
In the radiometric assay, hemolysates obtained as before are incubated with pure 14C-CMP.
By addition of a barium hydroxide/zinc sulfate solution, proteins and unreacted nucleotide are
precipitated, leaving labeled cytidine in the supernatant. Aliquots are measured for 14C ac-
tivity in a liquid scintillation counter. This method shows a good correlation with the ear-
lier technique.
9.9.7 Measurement of Plasma 1,25-Dihydroxyvitamin D
Measurement techniques for this vitamin D metabolite, all of recent vintage, consist of
three main parts: (1) isolation from plasma or serum by liquid-liquid extraction, (2) precon-
centration of the extract and chromatographic purification using Sephadex LH-20 or Lipidex
5000 columns, as well as high performance liquid chromatography (HPLC) in some cases, and (3)
quantification by either of two radiometric binding techniques, the more common competitive
protein binding (CPB) assay or radioimmunoassay (RIA). The CPB assay uses a receptor protein
in intestinal cytosol of chicks made vitamin D-deficient.
In one typical study, human adults had a mean level of 31 picograms/ml. The limit of
detection was 5 picograms/analytical tube, and within-run and between-run coefficients of
variation were 17 and 26 percent, respectively. In a recent interlaboratory survey involving
15 laboratories, the level of variance was such that it was recommended that each laboratory
should establish its own reference values.
9-38
-------
9.10 REFERENCES
Ahlgren, L. ; Haeger-Aronsen, B. ; Mattson, S. ; Schutz, A. (1980) In-vivo determination of lead
in the skeleton after occupational exposure to lead. Br. J. TndTUed. 37: 109-113.
Al-Naimi, T. ; Edmonds, M. I.; Fremlin, J. H. (1980) The distribution of lead in human teeth,
using charged particle activation analysis. Phys. Med. Biol. 25: 719-726.
American Public Health Association. (1955) Methods for determining lead in air and in bio-
logical materials. New York, NY: American Public Health Association.
Angle, C. R. ; Mclntire, M. S. (1978) Low level lead and inhibition of erythrocyte pyrimidine
nucleotidase. Environ. Res. 17: 296-302.
Balamut, R. ; Doran, D. ; Giridhar, G. ; Mitchell, D. ; Soule, S. (1982) Systematic error between
erythrocyte protoporphyrin in proficiency test samples and patients' samples as measured
with two hematofluorometers. Clin. Chem. (Winston-Salem, NC) 28: 2421-2422.
Barthel, W. F. ; Smrek, A. L. ; Angel, G. P.; Liddle, J. A.; Landrigan, P. J. ; Gehlbach, S. H
Chisolm, J. J. (1973) Modified Delves cup atomic absorption determination of lead in
blood. J. Assoc. Off. Anal. Chem. 56: 1252-1256.
Berlin, A.; Schaller, K. H. (1974) European standardized method for the determination of
6-aminolevulinic acid dehydratase activity in blood. Z. Klin. Chem. Klin. Biochem 12-
389-390.
Berlin, A.; Del Castilho, P.; Smeets, J. (1973) European intercomparison programmes. In:
Barth, D. ; Berlin, A.; Engel , R. ; Recht, P.; Smeets, J. , eds. Environmental health as-
pects of lead: proceedings, international symposium; October 1972; Amsterdam, The Nether-
lands. Luxembourg: Commission of the European Communities; pp. 1033-1049.
Berman, E. (1976) The challenge of getting the lead out. In: LaFleur, P. D. , ed. Accuracy in
trace analysis: sampling, sample handling, analysis - volume 2. Proceedings of the 7th
materials research symposium; October 1974; Gaithersburg, MD. Washington DC- U S
Department of Commerce, National Bureau of Standards; NBS special publication no' 4?2-
pp. 715-719. Available from: NTIS, Springfield, VA; PB-258092. »-«ion no. <^,
Berman, E. (1981) Heavy metals. Lab. Med. 12: 677-684.
Bloch, P.; Garavaglia G.; Mitchell, G.; Shapiro, I. M. (1976) Measurement of lead content of
children's teeth jn situ by X-ray fluorescence. Phys. Med. Biol. 20: 56-63. conuent OT
Blumberg, W. E. ; Eisinger, J. ; Lamola, A. A.; Zuckerman D. M. (1977) Zinc DrotoDorohvrin levpl
Bonsignore D. ; Calissano, P. ; Cartasegna C. (1965) Un semplice metodo per la determinazione
denaS-a/mno-levulinico-deidratasi nel sangue: comportamento dell 'enzima nelT intossica-
zione saturnina [A .simple method for determining 6-aminolevulinic dehydratase in the
blood: behavior of the enzyme in lead poisoning]. Med. Lav. 56: 199-205.
Boone, J. ; Hearn, T. ; Lewis, S. (1979) Comparison of interlaboratory results for blood lead
with results from a definitive method. Clin. Chem. (Winston-Salem, NC) 25: 389-393
9-39
-------
Boutwell, J. H. (1976) Accuracy and quality control in trace element analysis. In: LaFleur,
P. D. , ed. Accuracy in trace analysis: sampling, sample handling, analysis - volume 1.
Proceedings of the 7th materials research symposium; October 1974; Gaithersburg, MD.
Washington, DC: U.S. Department of Commerce, National Bureau of Standards; NBS special
publication no. 422; pp. 35-40. Available from: NTIS, Springfield, VA; PB-258092.
Cali, S. P.; Reed, W. P. (1976) The role of the National Bureau of Standards: standard refer-
ence materials in accurate trace analysis. In: LaFleur, P. D., ed. Accuracy in trace
analysis: sampling, sample handling, analysis - volume 1. Proceedings of the 7th
materials research symposium; October 1974; Gaithersburg, MD. Washington, DC: U.S.
Department of Commerce National Bureau of Standards; NBS special publication no. 422; pp.
41-63. Available from: NTIS, Springfield, VA; PB-258092.
Carter, G. F. (1978) The paper punched disc technique for lead in blood samples with abnormal
haemoglobin values. Br. J. Ind. Med. 35: 235-240.
Castoldi, M. R.; Odone, P.; Buratti, M.; Alessio, L. (1979) Determination of erythrocyte zinc
protoporphyrin: methodological problems. In: International conference: management and
control of heavy metals in the environment; September; London, United Kingdom. Edin-
burgh, United Kingdom: CEP Consultants, Ltd.; pp. 113-117.
Cavalleri, A.; Minoia, C.; Pozzoli, L.; Baruffini, A. (1978) Determination of plasma lead
levels in normal subjects and in lead-exposed workers. Br. J. Ind. Med. 35: 21-26.
Cernik, A. A.; Sayers, M. P. H. (1971) Determination of lead in capillary blood using a paper
punched disc atomic absorption technique: application to the supervision of lead workers.
Br. J. Ind. Med. 28: 392-398.
Chatman, T.; Wilson, D. J. (1975) Lead levels in human deciduous teeth in Tennessee. Environ.
Lett. 8: 173-183.
Chatt, A.; Secord, C. A.; Tiefenbach, B.; Jervis, R. E. (1980) Scalp hair as a monitor of
community exposure to environmental pollutants. In: Brown, A. C.; Crounse, R. C., eds.
Hair, trace elements and human illness. New York, NY: Praeger; pp. 46-73.
Chattopadhyay, A.; Roberts, T. M.; Jervis, R. E. (1977) Scalp hair as a monitor of community
exposure to lead. Arch. Environ. Health 32: 226-236.
Chisolm, J. J., Jr. (1974) Lead in red blood cells and plasma. J. Pediatr. (St. Louis) 84:
163-164.
Chisolm, J. J., Jr.; Brown, D. H. (1975) Micro-scale photofluorometric determination of
"free erythrocyte porphyrin" (protoporphyrin IX). Clin. Chem. (Winston-Salem, NC) 21:
1669-1682.
Chisolm, J. J., Jr.; Brown, D. H. (1979) Micromethod for zinc protoporphyrin in erythrpcytes:
including new data on the absorptivity of zinc protoporphyrin and new observations in
neonates and sickle cell disease. Biochem. Med. 22: 214-237.
Chisolm, J. J., Jr.; Hastings, C. W.; Cheung, D. K. K. (1974) Microphoto fluorometric assay
for protoporphyrin in acidified acetone extracts of whole blood. Biochem. Med. 9:
113-135.
9-40
-------
Cooke, R. E.; Glynn, K. L.; UHman, W. W.; Lun'e, N.; Lepow, M. (1974) Comparative study of
a micro-scale test for lead in blood, for use in mass screening programs. Clin. Chem.
(Winston-Salem, NC) 20: 582-585.
Culbreth, P.; Walter, G.; Carter, R.; Burtis, C. (1979) Separation of proto porphyrins and
related compounds by reversed-phase liquid chromatography. Clin. Chem. (Winston-Salem,
NC) 25: 605-610.
Davis, J. R.; Andelman, S. L. (1967) Urinary delta-aminolevulinic acid (ALA) levels in lead
poisoning. Arch. Environ. Health 15: 53-59.
Delia Fiorentina, H.; Grogna, M.; Dewiest, F. (1979) Simplified determination of urinary
6-aminolevulinic acid in a wide range of concentrations. Clin. Chem. (Winston-Salem, NC)
25: 581-583.
Delves, H. T. (1970) A micro-sampling method for the rapid determination of lead in blood by
atomic-absorption spectrophotometry. Analyst (London) 95: 431-438.
Delves, H. T. (1977) Analytical techniques for blood-lead measurements. J. Anal. Toxicol. 1:
261-264.
Delves, H. T.; Clayton, B. E.; Carmichael, A.; Bubear, M.; Smith M. (1982) An appraisal of the
analytical significance of tooth-lead measurements as possible indices of environmental
exposure of children to lead. Ann. Clin. Biochem. 19: 329-337.
DeSilva, P. E. (1981) Determination of lead in plasma and studies on its relationship to lead
in erythrocytes. Br. J. Ind. Med. 38: 209-217.
DeSilva, P. E.; Donnan, M. B. (1980) Blood lead levels in Victorian children. Med. J. Aust. 2:
315-318.
Donovan, D. T.; Vought, V. M.; Rakow, A. B. (1971) Laboratories which conduct lead analyses on
biologic specimens. Arch. Environ. Health 23: 111-113.
Dudley, D. M. T. (1982) Critique: blood lead analyses 1981. Atlanta, GA: U.S. Centers for
Disease Control.
Ediger, R. D.; Coleman, R. L. (1972) A modified Delves cup atomic absorption procedure for the
determination of lead in blood. At. Absorpt. Newsl. 11: 33-36.
Eller, P. M.; Hartz, J. C. (1977) A study of methods for the determination of lead and cad-
mium. Am. Ind. Hyg. Assoc. J. 38: 116-124.
Everson, J.; Patterson, C. C. (1980) "Ultra-clean" isotope dilution/mass spectrometric anal-
yses for lead in human blood plasma indicate that most reported values are artificially
high. Clin. Chem. (Winston-Salem, NC) 26: 1603-1607.
Farris, F. F.; Poklis, A.; Griesmann, G. E. (1978) Atomic absorption spectroscopic deter-
mination of lead extracted from acid-sol utilized tissues. J. Assoc. Off. Anal. Chem. 61:
660-663.
Finelli, V. N.; Klauder, D. S.; Karaffa, M. A.; Petering, H. G. (1975) Interaction of zinc and
lead on 6-aminolevulinate dehydratase. Biochem. Biophys. Res. Commun. 65: 303-311.
9-41
-------
/,nn Y' (1925) The Co1orimetric determination of phosphorus. J. Biol. Chem.
ob: "
Fosse, G. ; Justesen, N. P. B. (1978) Lead in deciduous teeth of Norwegian children Arch
Environ. Health 33: 166-175. Norwegian children, arcn.
Franke, J. P. ; de Zeeuw, R. A. (1977) Toxic metal analysis by differential pulse anodic strip-
ping voltammetry in clinical and forensic toxicology. J. Anal. Toxicol. 1- 291-295.
George, J. W. ; Duncan, J. R. (1982) Pyrimidine-specific 5' nucleotidase activity in bovine
erythrocytes: effect of phlebotomy and lead poisoning. Am. J. Vet. Res 43- 17-20.
Gibson, R. S. (1980) Hair as a biopsy material for the assessment of trace element status in
infancy: a review. J. Human Nutr. 34: 405-416.
Grabecki, J. ; Haduch, T. ; Urbanowicz, H. (1967) Die einfachen Bestimmungsmethoden der 6-amino-
lavulinsaure im Harn [Simple methods for the determination of 6-aminolevulimV acid in
urine]. Int. Arch. Gewerbepathol . Gewerbehyg. 23: 226-240. isvunmc aciu
Grandjean, P.; Lintrup, J. (1978) Erythrocyte-Zn-protoporphyrin as an indicator of lead ex-
posure. Scand. J. Clin. Lab. Invest. 38: 669-675.
Grandjean, P.; Nielsen, 0. V.; Shapiro, I. M. (1979) Lead retention in ancient Nubian and
contemporary populations. J. Environ. Pathol. Toxicol. 2: 781-787.
Granick, S.; Sassa, S. ; Granick, J. L. ; Levere, R. 0.; Kappas, A. (1972) Assays for por-
phyrins, 6-aminolevul ink-acid dehydratase, and porphyrinogen synthetase in microliter
samples of whole blood: applications to metabolic defects involving the heme oathwav
Proc. Natl. Acad. Sci. U.S.A. 69: 2381-2385. M y'
Granick, J. L. ; Sassa, S. ; Granick, S. ; Levere, R. D. ; Kappas, A. (1973) Studies in lead poi-
soning: II. correlation between the ratio of activated to inactivated 6-aminolevulinic
acid dehydratase of whole blood and the blood lead level. Biochem. Med. 8: 149-159.
Haeger-Aronsen, B. (1960) Studies on urinary excretion of 6-aminolaevulic acid and other haem
precursors in lead workers and lead-intoxicated rabbits. Scand. J. Clin. Lab Invest
12(suppl. 47): 1-128.
Hammond, P. B. ; Borschein, R. L. ; Zemick, H. (1984) Toxicological consideration in the assess-
ment of lead exposure. Neurotoxicology 5: 53-66.
Hammond, P. B. ; Bornschein, R. L. ; Succop, P. (1985) Dose-effect and dose-response relation-
ships of blood lead to erythrocytic protoporphyrin in young children. Environ. Res. 38:
in press.
Hanna, T. L. ; Dietzler, D. N. ; Smith, C. H. ; Gupta, S. ; Zarkowsky, H. S. (1976) Erythrocyte
porphyrin analysis in the detection of lead poisoning in children: evaluation of four
micromethods. Clin. Chem. (Winston-Sal em, NC) 22: 161-168.
Herber, R. F. M. (1980) Estimation of blood lead values from blood porphyrin and urinary
5-aminolevulinic acid levels in workers. Int. Arch. Occup. Environ. Health 45: 169-179.
Hessel , D. W. (1968) A simple and rapid quantitative determination of lead in blood. At.
Absorp. News!. 7: 55-56.
9-42
-------
1 M Gutierrez, A. N.; Worthy, B. E. (1973) Evaluation of the Delves micro system for
blood lead analysis. Clin. Chem. (Winston-Salem, NC) 19: 322-325.
Hinderberger E. J.; Kaiser, M. L.; Koirtyohann, S. R. (1981) Furnace atomic absorption anal-
ysis of'biological samples using the L'vov platform and matrix modification. At. Spect-
rosc. 2: 1-7.
Hodqes D. J.; Skelding, D. (1981) Determination of lead in urine by atomic-absorption spec-
troscopy with electrothermal atomisation. Analyst (London) 106: 299-304.
Issaq, H. J.; Zielinski, W. L., Jr. (1974) Loss of lead from aqueous solutions during storage.
Anal. Chem. 46: 1328-1329.
Jackson, K. W. (1978) Interlaboratory comparison of results of erythrocyte protoporphyrin anal-
ysis. Clin. Chem. (Winston-Salem, NC) 24: 2135-2138.
Jagner, D.; Danielsson, L. G.; Aren, K. (1979) Potentiometric stripping analysis for lead in
urine. Anal. Chim. Acta 106: 15-21.
Jongen, M. J. M; Van der Vijgh, W. J. F.; Van Beresteyn, E. C. H. (1982) Inter laboratory
variation of vitamin D metabolite measurements. Clin. Chem. Clin. Biochem. 20: 753-756.
Jongen, M. J. M.; Van Ginkel, F. C.; Van der Vijgh, W. J. F.; Kuiper, S.; Netelenbos, J. C ;
Lips, P. (1984) An international comparsion of vitamin D metabolite measurements Clin
Chem. (Winston-Salem, NC) 30: 399-403.
Joselow, M. M.; Bogden, J. D. (1972) A simplified micro method for collection and determina-
tion of lead in blood using a paper disk-in-Delves cup technique. At. Absorpt. Newsl 11-
99-101.
Joselow, M. M.; Flores, J. (1977) Application of the zinc protoporphyrin (ZP) test as a mon-
itor of occupational exposure to lead. Am. Ind. Hyg. Assoc. J. 38: 63-66.
Karacic, V.; Prpfc-Majic, D.; Telisman, S. (1980) The relationship between zinc protoporphyrin
(ZPP) and "free" erythrocyte protoporphyrin (FEP) in lead-exposed individuals. Int.
Arch. Occup. Environ. Health 47: 165-177.
Keppler, J. F.; Maxfield, M. E.; Moss, W. D.; Tietjen, G.; Linen, A. L. (1970) Interlaborator
evaluation of the reliability of blood lead analyses. J. Am. Ind. Hyq Assoc 31-
412-429.
Kochen, J. A.; Greener, Y. (1973) Levels of lead in blood and hematocrit: implications for the
evaluation of the newborn and anemic patient. Pediatr. Res. 7: 937-944.
Koizumi, H.; Yasuda, K. (1976) Determination of lead, cadmium, and zinc using the Zeeman ef-
fect in atomic absorption spectrometry. Anal. Chem. 48: 1178-1182.
Kopito, L. E.; Davis, M. A.; Shwachman, H. (1974) Sources of error in determining lead in
blood by atomic absorption spectrophotometry. Clin. Chem. (Winston-Salem NC) 20-
205-211.
Kubasik, N. P.; Volosin, M. T.; Murray, M. H. (1972) Carbon rod atomizer applied to measure-
ment of lead in whole blood by atomic absorption spectrophotometry. Clin. Chem (Winston
Salem, NC) 18: 410-412.
9-43
-------
LaFleur, P. D., ed. (1976) Accuracy in trace analysis: sampling, sample handling, analysis; v.
1 and 2. proceedings of the 7th materials research symposium; October 1974; Gaither-
sburg, MD. Washington, DC: U.S. Department of Commerce, National Bureau of
Standards; NBS special publication no. 422. Available from: NTIS, Springfield, VA;
PB-258092.
Lamola, A.-A.; Joselow, M.; Yamane, T. (1975) Zinc protoporphyrin (ZPP): a simple, sensitive,
fluorometric screening test for lead poisoning. Clin. Chem. (Winston-Sal em, NC) 21: 93-97.
Lauwerys, R.; Delbroeck, R.; Vens, M. 0. (1972) Automated analysis of delta- aminolaevulinic
acid in urine. Clin. Chiro. Acta 40: 443-447.
Lauwerys, R.; Buchet, J.-P.; Roels, H.; Berlin, A.; Smeets, J. (1975) Intercomparison program
of lead, mercury, and cadmium analysis in blood, urine, and aqueous solutions. Clin.
Chem. (Winston Salem, NC) 21: 551-557.
Lawrence, D. M. (1982) An atomic spectroscopy bibliography for January-June 1982. At. Spectr.
3: 93-115.
Lawrence, D. M. (1983) An atomic spectroscopy bibliography for July-December 1982. At. Spectr.
4: 10-33.
Lee, S. W.; Meranger, J. C. (1980) Direct methods for the determination of lead in whole blood
by anodic stripping voltammetry. Am. J. Med. Technol. 46: 853-857.
Legotte, P. A.; Rosa, W. C.; Sutton, D. C. (1980) Determination of cadmium and lead in urine
and other biological samples by graphite-furnace atomic-absorption spectrometry. Talanta
27: 39-44.
Lerner, S. (1975) Blood lead analysisprecision and stability. J. Occup. Med. 17: 153-154.
Lockeretz, W. (1975) Lead content of deciduous teeth of children in different environments.
Arch. Environ. Health 30: 583-587.
MacGee, J.; Roda, S. M. B.; Elias, S. V.; Lington, E. A.; Tabor, M. W.; Hammond, P. B. (1977)
Determination of 6-aminolevulinic acid in blood plasma and urine by gas-liquid chromato-
graphy. Biochem. Med. 17: 31-44.
Machlan, L. A.; Gramlich, J. W. ; Murphy, T. J.; Barnes, I. L. (1976) The accurate determina-
tion of lead in biological and environmental samples by isotope dilution mass spectro-
metry. In: LaFleur, P. D., ed. Accuracy in trace analysis: sampling, sample handling,
analysis - volume 2. Proceedings of the 7th materials research symposium; October 1974;
Gaithersburg, MO. Washington, DC: National Bureau of Standards; NBS special publication
no. 422; pp. 929-935. Available from: NTIS, Springfield, VA; PB-258092.
Mackie, A. C.; Stephens, R.; Townsend, A.; Waldron, H. A. (1977) Tooth lead levels in Birming-
ham children. Arch. Environ. Health 32: 178-185.
Mahaffey, K. R.; Annest, J. L.; Barbano, H. E.; Murphy, R. S. (1979) Preliminary analysis of
blood lead concentrations for children and adults: HANES II, 1976-1978. In: Hemphill, D.
D., ed. Trace substances in environmental health - XIII: [proceedings of University of
Missouri's 13th annual conference on trace substances in environmental health]; June;
Columbia, MO. Columbia, MO: University of Missouri-Columbia; pp. 37-51.
9-44
-------
Maher, C. C.; Roettgers, D. M.; Conlon, H. J. (1979) Interlaboratory comparison of blood lead
determinations. Am. Ind. Hyg. Assoc. J. 40: 230-237.
Manton, W. I.; Cook, J. 0. (1979) Lead content of cerebrospinal fluid and other tissue in amyo-
trophic lateral sclerosis (ALS). Neurology 29: 611-612.
Marcus, M.; Hollander, M.; Lucas, R. E.; Pfeiffer, N. C. (1975) Micro-scale blood lead determi-
nations in screening: evaluation of factors affecting results. Clin. Chem. (Winston-
Salem, NC) 21: 533-536.
Marcus, S. M.; Joselow, M. M.; Kemp, F.; Ziering, R.; Mihalovic, D.; Anderson, L. (1977)
Warning: spurious elevations of blood lead in micro puncture techniques [letter]. J.
Pediatr. (St. Louis) 91: 164.
Marver, H. S.; Tschudy, D. P.; Perlroth, M. G.; Collins, A.; Hunter, G., Jr. (1966) The determ-
ination of aminoketones in biological fluids. Anal. Biochem. 14: 53-60.
Matson, W. R.; Roe, D. K. (1966) Trace metal analysis of natural media by anodic stripping vol-
tammetry. Anal. Instruro. 4: 19-22.
Matson, W. R.; Griffin, R. M.; Schreiber, G. B. (1971) Rapid sub-nanogram simultaneous anal-
ysis of Zn, Cd, Pb, Cu, Bi and Tl. In: Hemphill, 0. D., ed. Trace substances in environ-
mental health - IV: [proceedings of University of Missouri's 4th annual conference on
trace substances in environmental health]; June 1970; Columbia, MO. Columbia, MO: Uni-
versity of Missouri-Columbia; pp. 396-406.
Mauzerall, D.; Granick, S. (1956) The occurrence and determination of 6-aminolevulinic acid
and porphobilinogen in urine. J. Biol. Chem. 219: 435-446.
Meranger, J. C.; Hollebone, B. R.; Blanchette, G. A. (1981) The effects of storage times,
temperatures and container types on the accuracy of atomic absorption determinations of
Cd, Cu, Hg, Pb and Zn in whole heparinized blood. J. Anal. Toxicol. 5: 33-41.
Mitchell, D. G.; Doran, D. (1985) Effect of bias in hematofluorometer measurements of proto-
porphyrin in screening programs for lead poisoning. Clin. Chem. (Winston-Salem, NC) 30:
386-390.
Mitchell, 0. G.; Ryan, F. J.; Aldous, K. M. (1972) The precise determination of lead in whole
blood by solvent extraction-atomic absorption spectrometry. At. Absorpt. Newsl. 11:
120-121.
Mitchell, D. G.; Aldous, K. M.; Ryan, F. J. (1974) Mass screening for lead poisoning: capil-
lary blood sampling and automated Delves-cup atomic-absorption analysis. N.Y. State J.
Med. 74: 1599-1603.
Mitchell, R. A.; Drake, J. E.; Wittlin, L. A.; Rejent, T. A. (1977) Erytnrocyte porphobi-
linogen synthase (delta-aminolaevulinate dehydratase) activity: a reliable and quanti-
tative indicator of lead exposure in humans. Clin. Chem. (Winston-Salem, NC) 23: 105-111.
Moller, B.; Carlsson, L.-E; Johannsson, G. I.; Malmqvist, K. G.; Hammarstrom, L,; Berlin, M.
(1982) Lead levels determined in Swedish permanent teeth by particle-induced X-ray em-
mission. Scand. J. Work Environ. Health 8: 267-272.
9-45
-------
Moore, M. R.; Meredith, P. A. (1977) The storage of samples for blood and water lead analysis
Clin. Chim. Acta 75: 167-170. J
Moore, M. R.; Campbell, B. C.; Meredith, P. A.; Beattie, A. D.; Goldberg, A Campbell D
(1978) The association between lead concentrations in teeth and domestic water l*>ari rnn-
centrations. Clin. Chim. Acta 87: 77-83.
Morrell, G.; Giridhar, G. (1976) Rapid micromethod for blood lead analysis by anodic striooina
voltammetry. Clin. Chem. (Winston-Salem, NC) 22: 221-223. 'HH"-a
Murphy, T. J. (1976) The role of the analytical blank in accurate trace analysis. In- LaFleur
P. D. , ed. Accuracy in trace analysis: sampling, sample handling, analysis - volume 1.'
Proceedings of the 7th materials research symposium; October 1974; Gaithersburg, MD.
Washington, DC: U.S. Department of Commerce, National Bureau of Standards; NBS special
publication no. 422; pp. 509-539. Available from: NTIS, Springfield, VA; PB-258092.
Nackowski, S. B.; Putnam, R. D.; Robbins, D. A.; Varner, M. 0.; White, L. D.; Nelson, K. W.
Trace metal contamination of evacuated blood collection tubes. Am. Ind. Hyg Ass'oc J
38: 503-508.
National Academy of Sciences. (1972) Lead: airborne lead in perspective. Washington, DC:
National Academy of Sciences. (Biologic effects of atmospheric pollutants).
Needleman, H. L.; Davidson, I.; Sewell, E. M.; Shapiro, I. M. (1974) Subclinical lead exposure
in Philadelphia schoolchildren: identification by dentine lead analysis. N. Engl. J. Med.
290: 245-248.
Needleman, H. L.; Gunnoe, C.; Leviton, A.; Reed, R.; Peresie, H.; Maher, C.; Barrett, P.
(1979) Deficits in psychologic and classroom performance of children with elevated den-
tine lead levels. N. Engl. J. Med. 300: 689-695.
Oehme, M.; Lund, W. (1978) The determination of copper, lead, cadmium and zinc in human teeth
by anodic stripping voltammetry. Anal. Chim. Acta 100: 389-398.
O'Flaherty, E. J. ; Hammond, P. B. ; Lerner, S. I.; Hanenson, I. B.; Roda, S. M. B. (1980) The
renal handling of 6-aminolevulinic acid in the rat and in the human. Toxicol. Appl.
Pharmacol. 55: 423-432.
Orfanos, A. P.; Murphey, W. H.; Guthrie, R. (1977) A simple fluorometric assay of protopor-
phyrin in erythrocytes (EPP) as a screening test for lead poisoning. J. Lab. Clin. Med.
89: 659-665.
Paglia, D. E.; Valentine, W. N. (1975) Characteristics of a pyrimidine-specific 5'nucleoti-
dase in human erythrocytes. J. Biol. Chem. 250: 7973-7979.
Paglia, D. E.; Valentine, W. N.; Dahlgren, J. G. (1975) Effects of low-level lead exposure on
pyrimidine 5'-nucleotidase and other erythrocyte enzymes: possible role of pyrimidine
5'-nucleotidase in the pathogenesis of lead-induced anemia. J. Clin. Invest. 56:
1164-1169.
Patterson, C. C. (1980) An alternative perspective - lead pollution in the human environment:
origin, extent and significance. In: National Academy of Sciences, Committee on Lead in
the Human Environment. Lead in the human environment. Washington, DC: National Academy of
Sciences; pp. 265-349.
9-46
-------
Patterson, C. C.; Settle, D. M. (1976) The reduction of orders of magnitude errors in lead ana-
lyses of biological materials and natural waters by evaluating and controlling the extent
and sources of industrial lead contamination introduced during sample collecting, handl-
ing, and analyses. In: LaFleur, P. D., ed. Accuracy in trace analysis: sampling, sample
handling, and analysis - volume 1. Proceedings of the 7th materials research symposium;
October 1974; Gaithersburg, MD. Washington, DC: National Bureau of Standards; NBS special
publication no. 422; pp. 321-352. Available from: NTIS, Springfield, VA; PB-258092.
Paulev, P.-E.; Solgaard, P.; Tjell, J. C. (1978) Interlaboratory comparison of lead and cad-
mium in blood, urine, and aqueous solutions. Clin. Chem. (Winston-Salem, NC) 24:
1797-1800.
Pierce, J. 0.; Koirtyohann, S. R.; Clevenger, T. E. ; Lichte, F. E. (1976) The determination
of lead in blood: a review and critique of the state of the art, 1975. New York, NY: In-
ternational Lead Zinc Research Organization, Inc.
Piomelli, S. (1973) A micromethod for free erythrocyte porphyrins: the FEP test. J. Lab. Clin.
Med. 81: 932-940.
Piomelli, S.; Davidow, B. (1972) Free erythrocyte protoporphyrin concentration: a promising
screening test for lead poisoning. Pediatr. Res. 6: 366.
Piomelli, S.; Graziano, J. (1980) Laboratory diagnosis of lead poisoning. Pediatr. Clin. North
Am. 27: 843-853.
Piomelli, S.; Corash, L.; Corash, M. B.; Seaman, C.; Mushak, P.; Glover, B.; Padgett, R.
(1980) Blood lead concentrations in a remote Himalayan population. Science (Washington,
DC) 210: 1135-1137.
Piscator, M. (1982) The importance of quality control for estimating dose-effect and dose-
response relationships. In: Schramel, P.; Bratter, P., eds. Trace element analytical
chemistry in medicine and biology. New York, NY: deGruyter and Co.: in press.
Rabinowitz, M. B.; Needleman, H. L. (1982) Temporal trends in the lead concentrations of umbil-
ical cord blood. Science (Washington, DC) 216: 1429-1431.
Rabinowitz, M.; Wetherill, G. W.; Kopple, J. D. (1974) Studies of human lead metabolism by use
of stable isotope tracers. Environ. Health Perspect. 7: 145-153.
Robinson, M. J.; Karpinski, F. E., Jr.; Brieger, H. (1958) The concentration of lead in
plasma, whole blood and erythroctyes of infants and children. Pediatrics 21: 793-796.
Roels, H.; Lauwerys, R.; Buchet, J.-P.; Berlin, A.; Smeets, J. (1974) Comparison of four
methods for determination of 6-aminolevulinic acid in urine, and evaluation of clinical
factors. Clin. Chem. (Winston-Salem, NC) 20: 753-760.
J F Zarate-Salvador, C.; Trinidad, E. E. (1974) Plasma lead levels in normal and
lead-intoxicated children. J. Pediatr. (St. Louis) 84: 45-48.
srhlenker F. S.; Taylor, N. A.; Kiehn, B. P. (1964) The chromatographic separation, deter-
mination and daily excretion of urinary porphobilinogen, amino acetone, and 6-aminolevu-
linic acid. Am. J. Clin. Pathol. 42: 349-354.
9-47
-------
Schwartz, S. ; Zieve, L. ; Watson, C. J. (1951) An improved method for the determination f
urinary coproporphyrin and an evaluation of factors influencing the analysis i V k
Clin. Med. 37: 843-859. y ana'ysis. J. Lab.
Scoble H. A.; McKeag, M. ; Brown, P. R. ; Kavarnos, G. J. (1981) The rapid determination of erv-
throcyte porphynns using reversed-phase high performance liquid chromatography Clin
Cnl ni« nCT.3 lloltbo~£Ob.
Settle, D. M. ; Patterson, C. C. (1980) Lead in albacore: guide to lead pollution in
Science (Washington, DC) 207: 1167-1176. v n in
Shapiro, I. M. ; Dobkin, B. ; Tuncay, 0. C. ; Needleman, H. L. (1973) Lead levels in dentin« an
circumpulpal dentine of deciduous teeth of normal and lead poisoned childrpn riV rk-
Acta 46: 119-123. U1n- Chlm-
Shepard, R. M. ; Horst, R. L. ; Hamstra, A. L. ; DeLuca, H. F. (1979) Determination of vitamin n
and its metabolites from normal and anephric man. Biochem. J. 182: 55-69. vitamin u
Slavin, S. ; Peterson, G. E. ; Lindahl, P. C. (1975) Determination of heavv metal c , * L.
atomic absorption spectroscopy. At. Absorp. News!. 14: 57-59. meats by
Smith, R. M. ; Doran, D. ; Mazur, M. ; Bush, B. (1980) High-performance liquid chromatn k-
determination of protoporphyrin and zinc protoporphyrin in blood. J Chroma*,!^9 ,«}c
319-327 ' "uina''0yr. 181:
Speecke, A.; Hoste, J. ; Versieck, J. (1976) Sampling of biological materials In- LaFl
D. , ed. Accuracy in trace analysis: sampling, sample handling, analysis - volume I^P
ceedings of the 7th materials research symposium; October 1974; Gaithersburn Mn~
Washington, DC: U.S. Department of Commerce, National Bureau of Standards- wnc ;
publication no>. 422; pp. 299-310. Available from: NTIS, Springfield, VA; PB- 258092
Steenhout, A.; Pourtois, M. (1981) Lead accumulation in teeth as a function of ano -+k j-
ferent exposures. Br. J. Ind. Med. 38: 297-303. ge W1th dlf~
Tomokuni, K. (1974) 6-aminolevulinic acid dehydratase test for lead exposure Arrh c
Health 29: 274-281. ' tnviron.
Tomokuni, K. ; Ogata, M. (1972) Simple method for determination of urinary 6-aminolevulin'
acid as an index of lead exposure. Clin. Chem. (Winston-Salem, NC) 18: 1534-1536 inic
Torrance, J. ; West, C. ; Beutler, E. (1977) A simple rapid radiometric assay for pvrimirtino ci
-nucleotidase. J. Lab. Clin. Med. 90: 563-568. y 'uine-u
Trevisan, A.; Buzzo, A,; Scarpa, F. M. (1981) Studio comparative delle ntetodiche di determin-
azione dell'attivita amino levulinico deidratasi eritrocitaria [Comparative study of"
methods for measurements of erythrocyte aminolevulinic acid dehydratase activitvl MDH
Lav. 72: 113-117. yj< nea"
U.S. Centers for Disease Control. (1981) Erythrocyte protoporphyrin proficiency testing- 1931
data summary. Atlanta, GA: U.S. Department of Health and Human Services, Centers for
Disease Control .
U.S. Occupational Safety and Health Administration. (1982) OSHA criteria for laboratory pro-
ficiency in blood lead analysis. Arch. Environ. Health 37: 58-60.
9-48
-------
Unger, B. C. ; Green. V. A. (1977) Blood lead analysislead loss to storage containers cifn
Toxicol. 11: 237-243. y ' uainers- nin.
Wigfield, D. C.; Farant, J.-P. (1979) Factors influencing the pH-activity relationshin of
6-aminolevulinic acid dehydratase from human blood, and their relevance to blood lead
assay. J. Anal. Toxicol. 3: 161-168.
Wittmers, L. E., Jr.; Alien, A.; Aufderheide, A. C. (1981) Lead in bone. 1 Direct analysis
for lead in milligram quantities of bone ash by graphite furnace atomic absorption soec-
troscopy. Am. J. Clin. Pathol. 75: 80-85. *<»pnon spec
World Health Organization, United Nations Environmental Programme. (1977) Lead Geneva
Switzerland: World Health Organization. (Environmental health criteria 3). '
9-49
-------
10. METABOLISM OF LEAD
10.1 INTRODUCTION
This chapter examines the absorption, distribution, retention, and excretion of lead in
humans and animals and the various factors that mediate the extent of the toxicokinetic pro-
cesses of lead. While inorganic lead is the form of the element that has been most heavily
studied, organolead compounds are also emitted into the environment and, because they are
quite toxic, they are also included in the discussion. Since the preparation of the 1977 Air
Quality Criteria Document for Lead (U.S. Environmental Protection Agency, 1977), a number of
reports have appeared that have proven particularly helpful in both quantifying the various
processes to be discussed in this chapter and assessing the interactive impact of factors such
as nutritional status in determining Internal exposure risk.
10.2 LEAD ABSORPTION IN HUMANS AND ANIMALS
The amounts of lead entering the bloodstream from various routes of absorption are deter-
mined not only by the levels of the element in the particular media, but also by the various
physical and chemical parameters that characterize lead. Furthermore, specific host factors
such as age and nutritional status are important, as is interindividual variability. Addi-
tionally, to assess absorption rates, one must know whether or not the subject is in "equili-
brium" with respect to a given level of lead exposure.
10.2.1 Respiratory Absorption of Lead
The movement of lead from ambient air to the bloodstream is a two-part process: a frac-
tion of air lead 1s deposited in the respiratory tract and, of this deposited amount, some
fraction is subsequently absorbed directly into the bloodstream or otherwise cleared from the
respiratory tract. At present, enough data exist to make some quantitative statements about
both of these components of respiratory absorption of lead.
The 1977 Air Quality Criteria Document for Lead described the model of the International
Radiological Protection Commission (IRPC) for the deposition and removal of lead from the
lungs and the upper respiratory tract (International Radiological Protection Commission,,
1966). Briefly, the model predicts that 35 percent of lead inhaled from ambient air by humans
is deposited in the respiratory tract, with most of the lead going to the parenchyma and air-
ways. The IRPC model predicts a total deposition of 40-50 percent for particles with a mass
median aerodynamic diameter (MMAD) of 0.5 urn and indicates that the absorption rate would vary
10-1
-------
depending on the solubility of the particular form. More recent data on lead deposition
modeling, however, provide a more precise picture (see next section).
10.2.1.1 Human Studies. Table 10-1 tabulates the various studies of human subjects that pro-
vide data on the deposition of inorganic lead in the respiratory tract. Studies of this type
have used diverse methodologies to characterize the inhaled particles in terms of both size
(and size ranges) and fractional distribution. The use of radioactive or stable lead isotopes
to directly or indirectly measure lead deposition and uptake into the bloodstream has been
particularly helpful in quantifying these processes.
From the studies of Kehoe (1961a,b,c) and their update by Gross (1981), as well as data
from Chamberlain et al. (1978), Morrow et al. (1980), and Nozaki (1966), the respiratory depo-
sition of airborne lead as encountered in the general population appears to be approximately
30-50 percent, depending on particle size and ventilation rates. Ventilation rate is parti-
cularly important with submicrometer particles, where Brownian diffusion governs deposition,
because a slower breathing rate enhances the frequency of collisions of particles with the
alveolar wall.
Figure 10-1 (Chamberlain et al., 1978) compares data, both calculated and experimentally
measured, on the relationship of percentage deposition to particle size. As particle size
increases, deposition rate decreases to a minimum over the range where Brownian diffusion pre-
dominates. Subsequently, deposition increases with size (>0.5 urn MMAD) as impaction and sedi-
mentation become the main deposition factors.
In contrast to the ambient air or chamber data tabulated in Table 10-1, higher deposition
rates in some occupational settings are associated with relatively large particles. However,
much of this deposition is in the upper respiratory tract, with eventual movement to the gas-
trointestinal tract by ciliary action and swallowing. Mehani (1966) measured total deposition
rates of 28-70 percent in battery workers and workers in marine scrap yards. Chamberlain and
Heard (1981) calculated an absorption rate of approximately 47 percent for particle sizes en-.
countered in workplace air.
Systemic absorption of lead from the lower respiratory tract occurs directly, while much
of the absorption from the upper tract involves swallowing and some uptake in the gut. From
the radioactive isotope data of Chamberlain et al. (1978) and Morrow et al. (1980), and the
stable isotope studies of Rabinowitz et al. (1977), one can conclude that lead deposited in
the lower respiratory tract is totally absorbed.
Chamberlain et al. (1978) used 203Pb in engine exhaust, lead oxide, or lead nitrate
aerosols in experiments where human subjects inhaled the lead from a chamber through a mouth-
piece or in wind-tunnel aerosols. By 14 days, approximately 90 percent of the label was re-
moved from the lung. Lead movement into the bloodstream could not be described by a simple
exponential function; 20 percent was absorbed within 1 hr and 70 percent within 10 hr.
10-2
-------
TABLE 10-1. DEPOSITION OF LEAD IN THE HUMAN RESPIRATORY TRACT
Form
Particle size
Lead Exposure
Percent deposition
Reference
00
Pb203 aerosols
from engine
exhaust
Lead "fumes"
made in Induc-
tion furnace
203Pb203
aerosol
Ambient air
lead near
motorway and
other urban
areas in U.K.
203Pb(OH)2 or
203PbCl2
aerosols
Lead In work-
place air;
battery
factory and
shipbreaking
operations
0.05 urn median
count diameter
in 38 studies;
5 subjects ex-
posed to average
of 0.9 urn
0.05-1.0 pin mean
diameter
Mean densities
of 0.02, 0.04,
0.09 urn
Mainly 0.1 urn
Both forms at
0.25 urn MMAD
Not determined;
defined as fumes,
fine dust, or
coarse dust
Chamber studies; 10, 20,
or 150 ug/m3; 3 hr on
alternate days;
12 subjects
Mouthpi ece/aerosol
chamber; 10 mg/m3;
adult subjects
Mouthpiece/aerosol
chamber; adult
subjects
2-10 ug/m3; adult
subjects
0.2 ud/liter for 5 min
or ^50 liters air;
adult subjects
3 adult groups:
23 |jg/m3 - controls
86 ug/m3 - battery
workers
180 ug/m3 - scrap yard
workers
30-70% (mean: 48%) Kehoe (1961a,b,c);
for mainly Gross (1981)
0.05-um particles
42% 0.05 M
63% 1.0 urn
80% 0.02
45% 0.04
30% 0.09
Nozaki (1966)
Chamberlain et al.
(1978)
60% fresh exhaust; Chamberlain et al.
50% other urban (1978)
area
23% chloride;
26% hydroxide
Morrow et al. (1980)
47% battery workers; Mehani (1966)
39% shipyard and
controls
-------
o
80
70
Z 60
Z
o
1 50
O
0.
UJ
O
UJ
O
cr
40
30
20
10
0.01
(V) Chamberlain et al. (1978)
(T) Heyder et al. (1975)
(7) Mitchell (1977)
(4a) James (1978)
(4b) James (1978)
f 5 } Yu and Taulbee (1977)
0.02
0.05
0.1
0.2
0.5
1.0
DIFFUSION MEAN
EQUIVALENT DIAMETER,
MASS MEDIAN
EQUIVALENT DIAMETER, pm
Figure 10-1. Effect of particle size on lead deposition rate in the lung. Broken lines
derived by calculation from reported data. Tidal volume equals 1000 cm3 except for
line 4b, where it equals 500 cm3. Breathing cycle equals 4 sec.
Source: Chamberlain et al. (1978).
-------
Rabinowitz et al. (1977) administered 204Pb tracer to adult volunteers and determined (by
isotope tracer and balance data) that 14 (jg lead was absorbed by these subjects daily at am-
bient air lead levels of 1-2 |jg/m3. Assuming a daily ventilation rate of 20 m3, a deposition
rate of 50 percent of ambient air (Chamberlain et al., 1978), and a mean air lead level of 1.5
(jg/m3 (2.0 |jg/m3 outside the study unit, 1.0 ug/m3 inside, as determined by the authors), then
15 ug lead was available for absorption. Hence, better than 90 percent of deposited lead was
absorbed dally.
Morrow et al. (1980) followed the systemic uptake of 203Pb in 17 adult subjects using
either lead chloride or lead hydroxide aerosols with an average size of 0.25 (±0.1) urn MMAD.
Half of the deposited fraction of either aerosol was absorbed in 14 hr or less. The radio-
label data described above are consistent with the data of Hursh and Mercer (1970), who
studied the systemic uptake of 212Pb on a carrier aerosol.
Given the apparent 1nvar1ance of absorption rate for deposited lead in the above studies
as a function of the chemical form of the element (Chamberlain et al., 1978; Morrow et al.,
1980), inhaled lead lodging deep in the respiratory tract seems to be absorbed equally, re-
gardless of form. Supporting evidence for total human systemic uptake of lead comes from
autopsy tissue analysis for lead content. Barry (1975) found that lead was not accumulated in
the lungs of lead workers. This observation is corroborated by the data of Gross et al.
(1975) for nonoccupatlonally exposed subjects.
Dependence of the respiratory absorption rate for lead in humans on the level of lead in
air has not been extensively studied, although the data of Chamberlain and coworkers (1978),
using human volunteers, show that the lung clearance rate in the adult for single lead pulses
dtd not vary over a lung burden range of 0.3 to 450 ug. In occupational settings, a curvi-
linear relationship between workplace airborne lead and blood lead results at least partly
from particle size changes, I.e., with Increasing dust concentration, particle aggregation
rate increases and the effective fraction of submicron particles (those penetrating to the
lung) compared to total particles steadily lessens (Chamberlain, 1983).
All of the available data for lead deposition and uptake from the respiratory tract in
humans have been obtained with adults, and quantitative comparisons with the same exposures in
children are not possible. Although children 2 years of age weigh one-sixth as much as an
adult, they Inhale 40 percent as much air lead as adults (Barltrop, 1972). James (1978) has
taken into account differences 1n airway dimensions 1n adults versus children, and has esti-
mated that, after controlling for weight, the 10-year-old child has a deposition rate 1.6- to
2.7-fold higher than the adult.
Recent studies support the above estimates of James (1978). Hofmann and coworkers
(Hofmann, 1982; Hofmann et al., 1979) reported dose calculations for the respiratory tract as
a function of age using airway length estimates from the literature and determined that intake
10-5
-------
of radioactive nuclides into both the tracheobronchial and pulmonary regions was highly age-
dependent, with maximal intake occurring at about age six.
10.2.1.2 Animal Studies. Experimental animal data for quantitative assessment of lead depo-
sition and absorption for the lung and upper respiratory tract are limited. The available in-
formation does, however, support the finding that respired lead is extensively and rapidly
absorbed.
Morgan and Holmes (1978) exposed adult rats, by nose-only technique, to a 203Pb-labeled
engine exhaust aerosol generated in the same manner as by Chamberlain et al. (1978) over a
period of 8 days. Exposure was at a level of 21.9 to 23.6 nCi label/liter chamber air. Ad-
justing for deposition on the animal pelt, 20-25 percent of the label was deposited in the
lungs. Deposited lead was taken up extensively in blood (50 percent within 1 hr and 98 per-
cent within 7 days). The absorption-rate kinetic profile was similar to that reported for
humans (Chamberlain et al., 1978).
Boudene et al. (1977) exposed rats to 210Pb-labeled aerosols at a level of 1 pg label/m3
and 10 (jg label/m3, the majority of the particles being 0.1-0.5 nm in size. At 1 hr, 30 per-
cent of the label had left the lung; by 48 hr, 90 percent was gone.
Bianco et al. (1974) used 212Pb aerosol (£0.2 urn) inhaled briefly by dogs and found a
clearance half-time from the lung of approximately 14 hr. Greenhalgh et al. (1979) found that
direct instillation of 203Pb-labeled lead nitrate solution into the lungs of rats led to an
uptake of approximately 42 percent within 30 min, compared with an uptake rate of 15 percent
within 15 min in the rabbit. These instillation data are consistent with the report of Pott
and Brockhaus (1971), who noted that intratracheal instillation of lead in solution (as bro-
mide) or in suspension (as oxide) serially over 8 days resulted in systemic lead levels in
tissues indistinguishable from injected lead levels. Rendall et al. (1975) found that the
movement of lead into blood of baboons inhaling a lead oxide (Pb304) was more rapid and resul-
ted in higher blood lead levels when coarse (1.6 urn mean diameter) rather than fine (0.8 urn
mean diameter) particles were used.
10.2.2 Gastrointestinal Absorption of Lead
Gastrointestinal (GI) absorption of lead mainly involves uptake from food and beverages,
as well as lead deposited in the upper respiratory tract that is eventually swallowed. It
also includes ingestion of nonfood material, primarily in children via normal mouthing activ-
ity and pica. Two issues of concern with lead uptake from the gut are the comparative rates
of such absorption in developing versus adult organisms, including humans, and how the bio-
availability of lead affects such uptake.
10.2.2.1 Human Studies. Based on long-term metabolic studies with adult volunteers, Kehoe
(1961a,b,c) estimated that approximately 10 percent of dietary lead is absorbed from the human
10-6
-------
gut. According to Gross (1981), various balance parameters can vary considerably among sub-
jects. These studies (Kehoe, 1961a,b,c) did not take into account the contribution of biliary
clearance of lead into the gut, which would have affected measurements for both absorption and
total excretion. Chamberlain et al. (1978) determined that the level of endogenous fecal lead
is approximately 50 percent of urinary lead values. They have estimated that 15 percent of
dietary lead is absorbed, if the amount of endogenous fecal lead is taken into account.
Following the Kehoe studies, a number of reports determined GI absorption using both sta-
ble and radloisotopic labeling of dietary lead. Generally, these reports support the observa-
tion that in the adult human the absorption of lead is limited when taken with food. Harrison
et al. (1969) determined a mean absorption rate of 14 percent for three adult subjects ingest-
ing 203Pb in diet, a figure in accord with the results of Hursh and Suomela (1968).
Chamberlain et al. (1978) studied the absorption of 203Pb in two forms (as the chloride and as
the sulfide) taken with food. The corresponding absorption rates were 6 percent (sulfide) and
7 percent (chloride), taking into account endogenous fecal excretion. Using adult subjects who
ingested the stable isotope 204Pb in their diet, Rabinowitz et al. (1974) reported an average
gut absorption of 7.7 percent. In a later study, Rabinowitz et al. (1980) measured an ab-
sorption rate of 10.3 percent.
A number of recent studies indicate that lead ingested under fasting conditions is absor-
bed to a much greater extent than lead taken with or incorporated into food. For example,
Blake (1976) measured a mean absorption rate of 21 percent when 11 adult subjects ingested
203Pb-labeled lead chloride several hours after breakfast. Chamberlain et al. (1978) found
that lead uptake in six subjects fed 203Pb as the chloride was 45 percent after a fasting
period, compared to 6 percent with food. Heard and Chamberlain (1982) obtained a rate of 63.3
percent using a similar procedure with eight subjects. Rabinowitz et al. (1980) reported an
absorption rate of 35 percent in five subjects when 204Pb was ingested after 16 hr of fasting.
These isotope studies support the observations of Barltrop (1975) and Garber and Wei (1974)
that lead in between-meal beverages is absorbed to a greater extent than is lead in food.
Dependence of the lead absorption rate from the human GI tract on the concentration of
lead in diet or water has not been well studied. Recent data from the reports of Blake
(1980), Flanagan et al. (1982), and Heard and Chamberlain (1983), however, indicate little
concentration dependency across the range of dietary lead content encountered by the general
population. For example, Flanagan et al. (1982) found that human volunteers absorbed 4, 40,
and 400 ug of ingested lead at about the same rate.
The relationship of lead bioavailability in the human gut to the chemical/biochemical
form of lead can be determined from available data, although interpretation is complicated by
the relatively small amounts administered and the presence of various components of food
10-7
-------
already present in the gut. Harrison et al. (1969) found no difference 1n lead absorption
from the human gut when lead isotope was given either as the chloride or incorporated into al-
qinate Chamberlain et al. (1978) found that labeled lead as the chloride or sulfide was ab-
sorbed to the same extent when ingested with food, but the sulfide form was absorbed at a rate
of 12 percent compared with 45 percent for the chloride under fasting conditions. Rabinowitz
et al. (1980) obtained similar absorption rates for the chloride, sulfide, or cysteine complex
forms when administered with food or under fasting conditions. Heard and Chamberlain (1982)
found no difference in absorption rate when isotopic lead (203Pb) was ingested with unlabeled
meat (sheep's liver and kidney) or when the label was incorporated into the food prior to
slaughter.
The data of Moore et al. (1979) are of interest with respect to relative GI uptake of
lead in adult males and females. Human volunteers (seven males, four females) were given
203Pb in water and whole-body counting was carried out at time points. It appeared that
females absorbed somewhat more of the label than males, but the difference did not reach sta-
tistical significance.
Two reports have focused on the question of differences in GI absorption rates between
adults and children. Alexander et al. (1973) carried out 11 balance studies with eight chil-
dren, aged 3 months to 8 years. Dally intake averaged 10.6 ug Pb/kg body weight (range 5-17).
The mean absorption rate determined from metabolic balance studies was 53 percent. A two-part
investigation by Ziegler et al. (1978) comprised a total of 89 metabolic balance studies with
12 normal infants aged 2 weeks to 2 years. In the first part, 51 balance studies using 9
children furnished a mean absorption rate of 42.7 percent. In the second, six children were
involved in 38 balance studies involving dietary lead intake at 3 levels. Diets were closely
controlled and lead content was measured. For all daily intakes of 5 ug Pb/kg or higher, the
mean absorption rate was 42 percent. At low levels of lead intake the data were variable,
with some children apparently in negative balance, probably because of the difficulty 1n con-
trolling low lead intake.
In contrast to these reports, Barltrop and Strehlow (1978) found that the results for
children hospitalized as orthopedic or "social" admissions were highly variable. A total of
104 balance studies were carried out in 29 children ranging in age from 3 weeks to 14 years.
Fifteen of the subjects were in net negative balance, with an average dietary absorption of
-40 percent or, when weighted by number of balance studies, -16 percent. Closely comparing
these data with those of Ziegler et al. (1978) is difficult. Subjects were Inpatients, repre-
sented a much greater age range, and were not classified 1n terms of mineral nutrition or
weight-change status. As an urban pedlatric group, the children 1n this study may have had
higher prior lead exposure so that the "washout" phenomenon (Kehoe, 1961a,b,c; Gross, 1981)
may have contributed to the highly variable results. The calculated mean daily lead Intake in
10-8
-------
the Barltrop and Strehlow group (6.5 ug/kg) was lower than that for all but one study group
described by Ziegler et al. (1978). In the latter study, data for absorption became more
variable as the daily lead intake was lowered. Finally, in those children classified as or-
thopedic admissions, whether skeletal trauma was without effect on lead equilibrium between
bone and other body compartments is unclear.
As typified by the results of the second National Health Assessment and Nutritional Eval-
uation Survey (NHANES II) (Mahaffey et al., 1979), children at 2-3 years of age show a small
peak in blood lead. The question arises whether this peak indicates an intrinsic biological
factor, such as increased absorption or retention when compared with older children, or whe-
ther this age group is exposed to lead in some special way. Several studies are relevant to
the question. Zielhuis et al. (1978) reported data for blood lead levels in 48 hospitalized
Dutch children, who ranged in age from 2 months to 6 years. Children up to 3 years old had a
mean blood lead level of 11.9 ug/dl versus a level of 15.5 in children aged 4-6 years. A sig-
nificant positive relationship between child age and blood lead was calculated (r = 0.44,
p <0.05). In the Danish survey by Nygaard et al. (1977), a subset of 126 children represent-
ing various geographical areas and age groups yielded the following blood lead values by mean
age group: children (N = 8) with a mean age of 1.8 years had a mean blood lead level of 4.3
ug/dl; those with a mean age of 3.7-3.9 years had values ranging from 5.6 to 8.3 ug/dl; and
children 4.6-4.8 years of age had a range of 9.2 to 10 ug/dl. These authors note that the
youngest group was kept at a nursery, whereas the older kindergarten children had more inter-
action with the outside environment. Sartor and Rondia (1981) surveyed two population groups
in Belgium, one of which consisted of groups of children aged 1-4, 5-8, and 9-14 years.
Children under the age of 1 year had a mean blood lead level of 10.7 M9/dl. The 1- to 4-year
and 5- to 8-year age groups were comparable, 13.9 and 13.7 ug/dl, respectively, while those
9-14 years old had a blood lead level of 17.2 ug/dl. All of the children in this study were
hospital patients. While these European studies suggest that any significant restriction of
children in terms of environmental interaction, e.g., in hospitals or nurseries, is associated
with an apparently different age-blood lead relationship than the U.S. NHANES II subjects,
whether European children in the 2- to 3-year age group show a similar peak remains to be
demonstrated. The issue merits further study.
The normal mouthing activity of young children, as well as the actual ingestion of non-
food items (i.e., pica), is a major concern in pediatric lead exposure, particularly in urban
areas with deteriorating housing stock and high automobile density and in nonurban areas con-
tiguous to lead-production facilities. The magnitude of such potential exposures is discussed
in Chapter 7, and an integrated assessment of impact on human intake appears in Chapter 13.
Such intake is intensified for children with pica and would include paint, dust, and dirt.
10-9
-------
Drill et al. (1979), using data from Day et al. (1975) and Lepow et al. (1974), have at-
tempted to quantify the dally intake of soil/dust in young children from such mouthing activi-
ties as thumb sucking and finger licking. A total of 100 mg/day was obtained for children 2-3
years old, but the amount of lead in this ingested quantity varied considerably from site to
site. In the report, a GI absorption rate of 30 percent was estimated for lead in soil and
dust. Of relevance to this estimate are the animal data discussed in the next section, which
show that lead of variable chemical forms in soil or dust is as available for absorption as
lead in food. The iji vitro studies relating lead solubility in street dusts with acidity
clearly demonstrate that the acidity of the human stomach is adequate to extensively solubi-
lize lead assimilated from soil and dust. To the extent that ingestion of such material by
children occurs other than at mealtime, the fasting factor in enhancing lead absorption from
the human GI tract (vide supra) must also be considered. Hence, a factor of 30 percent for
lead absorption from dusts and soils is not an unreasonable value.
A National Academy of Sciences (MAS) report on lead poisoning in children has estimated
that paint chip Ingestion by children with pica occurs with considerable frequency (National
Academy of Sciences, 1976). In the case of paint chips, Drill et al. (1979) estimated an ab-
sorption rate as high as 17 percent. This value may be compared with the animal data in Sec-
tion 10.2.2.2, which indicate that lead in old paint films can undergo significant absorption
in animals.
10.2.2.2 Animal Studies. Lead absorption via the gut of various adult experimental animal
species appears to resemble that for the adult human, on the order of 1-15 percent in most
cases. Kostial and her coworkers (Kostial and Kello, 1979; Kostial et al., 1978, 1971) re-
ported a value of 1 percent or less in adult rats maintained on commercial rat chow. These
studies were carried out using radioisotopic tracers. Similarly, Barltrop and Meek (1975)
reported an absorption rate of 4 percent in control diets, while Aungst et al. (1981) found
the value to range from 0.9 to 6.9 percent, depending on the level of lead given in the diet.
In these rat studies, lead was ingested with food. Quarterman and Morrison (1978) admini-
stered 203Pb label in small amounts of food to adult rats and found an uptake rate of appro-
ximately 2 percent at 4 months of age. Pounds et al. (1978) obtained a value of 26.4 percent
with four adult Rhesus monkeys given 210Pb by gastric intubation. The higher rate, relative
to the rat, may reflect various states of fasting at time of Intubation or differences in
dietary composition (vide infra), two factors that affect rates of absorption.
As seen above with human subjects, fasting appears to enhance the rate of lead uptake in
experimental animals. Garber and Wei (1974) found that fasting markedly enhanced gut uptake
of lead in rats. Forbes and Reina (1972) found that lead dosing by gastric intubation of rats
yielded an absorption rate of 16 percent, which is higher than other data for the rat indi-
cate. Intubation was likely done when little food was in the gut. The data of Pounds et al.
10-10
-------
(1978), as described above, may also suggest a problem with administering lead by gastric in-
tubation or mixed with water as opposed to food.
The bioavailability of lead in the GI tract of experimental animals has been the subject
of a number of reports. The designs of these studies differ in regard to how "bioavailabi-
lity" is defined. In some cases, the dietary matrix was kept constant, or nearly so, while
the chemical or physical form of the lead was varied. By contrast, other data described the
effect of changes in bioavailability as the basic diet matrix was changed. The latter case is
complicated by the simultaneous operation of lead-nutrient interactive relationships (de-
scribed in Section 10.5.2).
Allcroft (1950) observed comparable effects when calves were fed lead in the form of the
phosphate, oxide, or basic carbonate (PbC03*Pb(OH)2), or incorporated into wet or dry paint.
By contrast, lead sulfide in the form of finely ground galena ore was less toxic. Criteria
for relative toxicity included kidney and blood lead levels and survival rate over time.
In the rat, Barltrop and Meek (1975) carried out a comparative absorption study using
lead in the form of the acetate as the reference substance. The carbonate and thai late were
absorbed to the greatest extent, while absorption of the sulfide, chromate, napthenate, and
octoate was 44-67 percent of the reference agent. Barltrop and Meek (1979) also studied the
relationship of the size of lead particles (as the metal or as lead octoate or chromate in
powdered paint films) to the amount of gut absorption in the rat; they found an inverse rela-
tionship between uptake and particle size for both forms.
Gage and Litchfield (1968, 1969) found that lead napthenate and chromate can undergo con-
siderable absorption from the rat gut when incorporated Into dried paint films, although less
than when given with other vehicles. Ku et al. (1978) found that lead in the form of the ace-
tate or as a phosphollpid complex was equally absorbed from the GI tract of both adult and
young rats at a level of 300 ppm. Uptake was assessed by weight change, tissue levels of
lead, and urinary aminolevulinlc add (ALA) levels.
In a study relevant to the problem of lead bloavailability in soils and dusts, particu-
larly in exposed children, Dacre and Ter Haar (1977) compared the effects of lead as acetate
with lead contained 1n roadside soil and in house paint soil, at a level of approximately 50
ppm, in commercial rat chow. Uptake of lead was Indexed by weight change, tissue lead con-
tent, and Inhibition of aminolevulinlc add dehydrase (ALA-D) activity. None of these para-
meters differed significantly across the three groups, suggesting that neither the geochemlcal
matrix 1n the soils nor the various chemical forms (basic carbonate in paint soil, and the
oxide, carbonate, and basic carbonate in roadside soil) affect lead uptake.
These data are consistent with the behavior of lead in dusts upon acid extraction as re-
ported by Day et al. (1979), Harrison (1979), and Duggan and Williams (1977). In the Day et
al. study, street dust samples from England and New Zealand were extracted with hydrochloric
10-11
-------
acfd (HC1) over the pH range of 0-5. At an acidity that may be equalled by gastric secre-
tions, i.e., pH of 1, approximately 90 percent of the dust lead was solubilized. Harrison
(1979) noted that at this same acidity, up to 77 percent of Lancaster, England, street-dust
lead was soluble, while an average 60 percent solubility was seen in London dust samples
(Duggan and Williams, 1977). Because gastric solubilization must occur for lead in these
media to be absorbed, the above data are useful in determining relative risk.
Kostial and Kello (1979) compared the absorption of 203Pb from the gut of rats maintained
on commercial rat chow versus rats fed such "human" diets as baby foods, porcine liver, bread,
and cow's milk. Absorption in the latter cases varied from 3 to 20 percent, compared with
<1.0 percent with rat chow. This range of uptake for the nonchow diet compares closely with
that reported for human subjects (vide supra). Similarly, Jugo et al. (1975a) observed that
rats maintained on fruit diets had an absorption rate of 18-20 percent. The generally ob-
served lower absorption of lead in the adult rat compared to the adult human appears, then,
less reflective of a species difference than of a dietary difference.
A number of studies have documented that the developing animal absorbs a relatively
greater fraction of ingested lead than does the adult, thus supporting studies showing this
age dependency in humans. For example, the adult rat absorbs approximately 1 percent lead or
less via diet versus a corresponding value 40-50 times greater in the rat pup (Kostial et al.,
1971, 1978; Forbes and Reina, 1972). In the rat, this difference persists through weaning
(Forbes and Reina, 1972), at which point uptake resembles that of adults. Part of this dif-
ference can be ascribed to the nature of the diet (mother's milk versus regular diet), al-
though the extent of absorption enhancement with milk versus rat chow in the adult rat found
by Kello and Kostial (1973) fell short of what is seen in the neonate. An undeveloped, less
selective intestinal barrier may also exist in the rat neonate. In nonhuman primates, Munro
et al. (1975) observed that infant monkeys absorbed 65-85 percent via the gut versus 4 percent
in adults. Similarly, Pounds et al. (1978) noted that juvenile rhesus monkeys absorbed appro-
ximately 50 percent more lead than adults.
The question of the relationship of level of lead intake through the GI tract and rate of
lead absorption was addressed by Aungst et al. (1981), who exposed adult and suckling rats to
doses of lead by intubation over the range 1-100 mg/kg or by variable concentrations 1n drink-
ing water. With both age groups and both forms of oral exposure, lead absorption as a percent-
age of dose decreased, suggesting a saturation phenomenon for lead transport across the gut
wall.
Similar data were obtained by Bushnell and DeLuca (1983) for weanling rats given 203Pb by
intubation along with carrier doses of 1, 10, 100, or 1000 ppm 1n diet. The GI absorption
rate was observed to decrease significantly between 10 and 100 ppm carrier lead. Using Iso-
lated duodenal loop preparations, Conrad and Barton (1978) reported that lead uptake across
10-12
-------
the gut wall was constant from 0.001 to 10 ppm lead, but fell off to 40 percent of the 10-ppm
level at the 100-ppm dosing.
The above concentration dependency is consistent with a saturable, active transport pro-
cess for lead in the mammalian gut, based on the kinetic data of Aungst and Fung (1981).
Mykkanen and Wasserman (1981) also noted that lead uptake by chick intestine occurs in two
kinetic phases; a rapid uptake is followed by a rate-limiting slow transfer of lead. These
kinetic observations agree with an increasingly retarded active transport process as lead con-
tent increases in the gut; i.e., lead affects its own transport, manifested as an increasingly
lower absorption rate at higher lead intake.
Of interest here is the comparison of the kinetic behavior of blood lead as a function of
oral versus parenteral dosing. With single intravenous injections of 0.5, 1, 5, 10, and 15 mg
Pb/kg lead in the rat, Aungst et al. (1981) did not observe any dose dependency of the kinetic
rate coefficients governing lead in blood. Integrated exposure, i.e., area under the blood
lead curves, increased linearly with dose. On the other hand, injection of lead into rabbits
at levels of 5, 10, 25, 50, and 500 ug/kg, by single dally injections for 6 days, resulted in
clear curvilinearity to the dose-blood lead curve (Prpic-Majic et al., 1973). The differences
in these two reports probably reflect dosing regimen differences: Aungst et al. (1981) used a
higher dosing level as single exposures.
The implication of these experimental findings for human oral lead exposure is not clear.
As noted earlier, lead intake orally by human subjects up to 400 ug is associated with a
rather fixed absorption rate. Direct extrapolation of the animal data described above indi-
cates that humans would have to ingest 20 to 200 mg lead per day (assuming a 2-kg diet/day at
lead contents of 10 or 100 ppm) to have a lowered absorption rate. This value is up to 4500-
fold above the upper oral intake guideline for lead (National Academy of Sciences, 1980).
10.2.3 Percutaneous Absorption of Lead
Absorption of inorganic lead compounds through the skin appears to be considerably less
significant than uptake through the respiratory and GI routes. This observation contrasts
with observations for lead alkyls and other organic derivatives (see Section 10.7). Rastogi
and Clausen (1976) found that cutaneous or subcutaneous administration of lead napthenate in
rat skin was associated with higher lead tissue levels and more severe toxic effects than was
the case for lead acetate. Laug and Kunze (1948) applied lead as the acetate, orthoarsenate,
oleate, and ethyl lead to rat skin and determined that the greatest levels of kidney lead were
associated with the alkyl contact.
Moore et al. (1980) studied the percutaneous absorption of 203Pb-labeled lead acetate 1n
cosmetic preparations using eight adult volunteers. Applied in wet or dry forms, absorption
was indexed by blood, urine, and whole body counting. Absorption rates ranged from 0 to 0.3
10-13
-------
percent, with the highest values obtained when the application sites were scratched. These
researchers estimated that the normal use of such preparations would result in an absorption
of approximately 0.06 percent.
10.2.4 Transplacental Transfer of Lead
Lead uptake by the human and animal fetus occurs readily, based on such indices as fetal
tissue lead measurements and, in the human, cord blood lead levels. Barltrop (1969) and
Horiuchi et al. (1959) demonstrated by fetal tissue analysis that placental transfer in the
human occurs by the 12th week of gestation, with fetal lead uptake Increasing throughout
development. The highest lead levels occur in bone, kidney, and liver, followed by blood,
brain, and heart. Cord blood contains significant amounts of lead, which generally correlate
with maternal blood values and are slightly but significantly lower in concentration than the
mother's (Scanlon, 1971; Harris and Holley, 1972; Gershanlk et al., 1974; Buchet et al., 1978;
Alexander and Delves, 1981; Rabinowitz and Needleman, 1982).
A cross-sectional study of maternal blood lead levels carried out by Alexander and Delves
(1981) showed that a significant decrease in maternal blood lead occurs throughout pregnancy,
a decrease greater than the dilution effect of the concurrent increase in plasma volume.
Hence, during pregnancy there is either an increasing deposition of lead in placental or fetal
tissue or an increased loss of body lead via other routes. Increasing absorption by the fetus
during gestation, as demonstrated by Barltrop (1969), implies that the former explanation is
likely. Hunter (1978) found that summer-born children showed a trend toward higher blood lead
levels than those born in the spring, suggesting increased fetal uptake in the summer result-
ing from increases in circulating maternal lead. This observation was confirmed in the report
of Rabinowitz and Needleman (1982). Ryu et al. (1978) and Singh et al. (1978) both reported
that infants born to women having a history of lead exposure had significantly elevated blood
lead values at birth.
10.3 DISTRIBUTION OF LEAD IN HUMANS AND ANIMALS
A quantitative understanding of the sequence of changes in lead levels in various body
pools and tissues is essential in interpreting measured lead levels with respect to past expo-
sure as well as present and future risks of toxicity. This section discusses the distribution
kinetics of lead in various portions of the body (blood, soft tissues, calcified tissues, and
the "chelatable" or toxicologically active body burden) as a function of such parameters as
exposure history and age.
10-14
-------
A given quantity of lead taken up from the GI tract or the respiratory tract into the
bloodstream is initially distributed according to the rate of delivery by blood to the various
organs and systems. Lead is then redistributed to organs and systems in proportion to their
respective affinities for the element. With consistent exposure for an extended period, a
near steady state of intercompartmental distribution is achieved.
Fluctuations in the near steady state will occur whenever short-term lead exposures are
superimposed on a long-term uptake pattern. Furthermore, the steady-state description is im-
perfect because, on a very short (hourly) time scale, intake is not constant. Lead intake
with meals and changes in ambient air lead (outside to inside and vice versa) cause quick
changes in exposure levels that may be viewed as short-term alterations in the small, labile
lead pool. Metabolic stress could remobilize and redistribute body stores, although documen-
tation of the extent to which this happens is very limited (Chisolm and Harrison, 1956).
10.3.1 Lead in Blood
Viewed from different time scales, lead in whole blood may be seen as residing in several
distinct, interconnected pools. More than 99 percent of blood lead is associated with the
erythrocytes (DeSilva, 1981; Everson and Patterson, 1980; Manton and Cook, 1979) under typical
conditions, but it is the very small fraction of lead transported in plasma and extracellular
fluid that provides lead to the various body organs (Baloh, 1974).
Although the toxicity of lead to the erythrocyte (Raghavan et al., 1981) is mainly asso-
ciated with membrane lead content, most of the erythrocyte lead is bound within the cell.
Within erythrocytes from nonexposed subjects, lead is primarily bound to hemoglobin, in par-
ticular HbA2, which binds approximately 50 percent of cell lead while constituting only 1-2
percent of total hemoglobin (Bruenger et al., 1973). A further 5 percent is bound to a
10,000-dalton molecular-weight fraction, about 20 percent to a much heavier molecule, and
about 25 percent is considered "free" or bound to lower-weight molecules (Ong and Lee, 1980a;
Raghavan and Gonick, 1977). Raghavan et al. (1980) have observed that, among workers exposed
to lead, those who develop signs of toxicity at relatively low blood lead levels seem to have
a diminished binding of intracellular lead with the 10,000-dalton fraction. This reduction in
binding suggests an impaired biosynthesis of a protective species. According to Ong and Lee
(1980b), fetal hemoglobin has a higher affinity for lead than adult hemoglobin.
Whole blood lead 1n daily equilibrium with other compartments was found to have a mean
life of 35 days (25-day half-life) and a total lead content of 1.9 mg, based on studies with a
small number of subjects (Rabinowitz et al., 1976). Chamberlain et al. (1978) established a
similar half-life for 203Pb in blood when volunteers were given the label by 1ngest1on, Inha-
lation, or injection. The lead Inhalation studies 1n adults described by Griffin et al.
10-15
-------
(1975) permit calculation of half-lives of 28 and 26 days for inhalation of 10.4 and 3.1 ug
Pb/m3, respectively. These estimates of biological half-life, based as they are on isotopic
study, do not reflect the impact of mobile body burden on half-life. The higher the mobiliza-
ble lead burden, the greater will be the length of the half-life, as clearly seen in the
report of O'Flaherty et al. (1982), where half-life in lead workers was a function of cumula-
tive occupational exposure.
Alterations in blood lead levels in response to abrupt changes in exposure apparently oc-
cur over somewhat different periods, depending on whether the direction of change is greater
or smaller. With increased lead intake, blood lead level achieves a new value in approxi-
mately 60 days (Griffin et al., 1975; Tola et al., 1973). A decrease may involve a longer
period of time, depending on the magnitude of the past higher exposure (O'Flaherty et al.,
1982; Rabinowitz et al. 1977; Gross, 1981).
In adulthood, the human's blood lead level appears to increase moderately. Awad et al.
(1981) reported an increase of 1 ug for each 14 years of age. However, in the NHANES II sur-
vey (see Chapter 11), white adults showed increasing blood lead until 35-44 years of age, fol-
lowed by a decrease. By contrast, blacks showed increasing blood lead after 44. In the case
of reduced exposure, particularly occupational exposure, the time for re-establishing near
steady state depended more upon the extent of lead resorption from bone and the total quanti-
ty deposited, either of which can extend the "washout" interval.
Lead levels in newborn children are similar to but somewhat lower than those of their
mothers: 8.3 versus 10.4 ug/dl (Buchet et al., 1978) and 11.0 versus 12.4 ug/dl (Alexander
and Delves, 1981). Maternal blood lead levels decrease throughout pregnancy, the decrease
being greater than the expected dilution via the concurrent increase in plasma volume
(Alexander and Delves, 1981). This decrease in maternal blood lead levels suggests increased
fetal uptake during gestation (Barltrop, 1969). Increased tissue retention of lead by the
child may also be a factor.
Levels of lead in blood are sex related; adult women invariably show lower levels than
adult males (e.g., Mahaffey et al., 1979). Of interest in this regard is the study of Stuik
(1974) showing lower blood lead response in women than in men for an equivalent level of lead
intake.
The small but biologically significant lead pool in blood plasma has proven technically
difficult to measure, and reliable values have become available only recently (see Chapter 9).
Chamberlain et al. (1978) found that injected 203Pb was removed from plasma (and, by infer-
ence, from extracellular fluid) with a half-life of less than 1 hr. These data support the
observation of DeSilva (1981) that lead is rapidly cleared from plasma. Ong and Lee (1980a),
in their jm vitro studies, found that 203Pb is virtually all bound to albumin and that only
10-16
-------
trace amounts are bound to high-weight globulins. To state which binding form constitutes an
"active" fraction for movement to tissues is not possible.
Although Rosen et al. (1974) reported that plasma lead did not vary across a range of
whole blood levels, the findings of Everson and Patterson (1980), DeSilva (1981), and
Cavalleri et al. (1978) indicate that there is an equilibrium between red blood cells (RBCs)
and plasma, such that levels in plasma rise with levels in whole blood. This observation is
consistent with the data of Clarkson and Kench (1958), who found that lead in the RBC is rela-
tively labile to exchange and a logical prerequisite for a dose-effect relationship in various
organs. Ong and Lee (1980c), furthermore, found that plasma calcium is capable of displacing
RBC membrane lead, suggesting that plasma calcium is a factor in the cell-plasma lead equilib-
rium.
Several studies concerning the relative distribution of lead between erythrocytes and
plasma or serum indicate that the relative percentage of blood lead in plasma versus erythro-
cytes is relatively constant up to a blood lead concentration of about 50-60 pg/dl, but
becomes increasingly greater above this level, i.e., the overall blood lead/plasma lead rela-
tionship is curvilinear upward.
DeSilva (1981) found that the relative fraction of plasma lead versus erythrocytes in 105
Australian lead workers increased at -v-60 |jg/dl. Similarly, Manton and Malloy (1983) observed
that a subject having lead intoxication had serum lead values ranging from 1.6 to 0.3 percent
as blood concentration changed from 116 to 31 ug/dl. More recently, Manton and Cook (1984)
demonstrated a curvilinear relationship between serum and whole blood lead levels. As de-
picted in Figure 10-2, the curve indicates that there is a linear segment up to ~50 pg/dl,
followed by rather steep increases in relative serum lead content at higher levels.
Measurement of lead in plasma by these investigators was carefully carried out, and the
Manton reports involved the definitive lead analysis technique of isotope-dilution mass spec-
trometry (IDMS, see Chapter 9). Given the increased erythrocyte fragility with increasing
blood lead content (see Section 12.3), slight hemolysis during sampling might contaminate
plasma or serum with high erythrocyte lead and complicate such analyses; however, the reports
did not indicate that hemolysis was considered a problem.
The biological basis for higher levels of plasma versus whole blood lead with increasing
blood lead burden may be related to marked changes in the binding capacity of the erythrocyte
at high lead content. These changes may result from alterations in binding sites or in the
efficiency of lead movement from membrane to erythrocyte interior. Fukumoto et al. (1983)
have demonstrated changes (in the form of a decrease) in lead-worker erythrocyte-membrane pro-
teins that may have a role in lead transport. Perhaps more important are the long-known ef-
fects of lead exposure on erythrocyte morphology and destruction rate (see Section 12.3).
10-17
-------
30
3
O>
2.0
flC
01
(A
1.0
0.0
I I I I I 1 I I I I II I I /I
I I I I I I I I I I I
50 100
BLOOD LEAD. M9/dl
150
Figure 10-2. The curvilinear relationship of serum lead to blood lead.
Cross-hatched area represents several overlapping points.
Source: Manton and Cook (1984).
10-18
-------
Changes in cell morphology with Increasing blood lead may alter accessibility to binding sites
or the relative stability of these sites. Increased cell destruction may Increase protein-
bound cell lead in plasma, which is only slowly transferred back to cell membrane.
In vitro data concerning the concentration dependency of lead partitioning between ery-
throcytes and plasma are of Interest. Keep 1n mind, however, that such jj± vitro data have em-
ployed normal erythrocytes. Clarkson and Kench (1958) showed that the relative partitioning
between normal erythrocytes and plasma is relatively constant up to the highest level tested,
equivalent to 100 ug Pb/dl. In the related study of Kochen and Greener (1973), tracer plus
carrier lead was added to blood of varying hematocrlt up to a maximum addition of 1000 ^g/dl.
At a normal hematocrit and a higher value (0.65), the percent uptake of lead label by the
cells diminished at around 100 ug/dl, consistent with the Clarkson and Kench (1958) data.
Onset of curviUnearlty at a lower blood lead level ^ri vivo in lead-exposed subjects below the
IB vitro value of ^100 ug/dl probably reflects in part altered cell morphology and stability
(DeSilva, 1981; Manton and Malloy, 1983; Manton and Cook, 1984).
The curvilinear relationship of plasma to whole blood lead may well be a factor in
Chamberlain's (1983) observation that the relative rate of urinary excretion of lead in human
adults Increases with blood lead content, as determined from various published reports provid-
ing both blood and urinary lead data (see Section 10.4). It may also figure in the apparently
better proportionality of tissue lead burdens to dose than blood lead (vide infra) and,
equally Important, the curvilinear relationship of chelatable lead to blood lead. That Is, at
Increasing blood lead, the higher relative rate of plasma lead movement to soft tissues and
bone is greater than would be anticipated from simple Inspection of blood lead content, the
latter rising at a slower rate relative to the increase in plasma lead.
10.3.2 Lead Levels in Tissues
Of necessity, various relationships of tissue lead to exposure and toxicity in humans
generally must be obtained from autopsy samples, although in some studies biopsy data have
been described. The inherent question then is whether such samples adequately represent the
behavior of lead in the living population, particularly in cases where death was preceded by
prolonged illness or disease states. Also, victims of fatal accidents are not well character-
ized as to exposure status and are usually described as having no "known" lead exposure.
Finally, these studies are necessarily cross-sectional in design, and, in the case of body ac-
cumulation of lead, different age groups are assumed to have been similarly exposed. Some im-
portant aspects of the available data include the distribution of lead between soft and cal-
cifying tissue, the effect of age and development on lead content of soft and mineral tissue,
and the relationship between total and "active" lead burdens in the body.
10-19
-------
10.3.2.1 Soft Tissues. In humans over age 20 most soft tissues do not show age-related
changes 1n lead levels, in contrast to the case with bone (Barry and Mossman, 1970; Barry,
1975, 1981; Schroeder and Upton, 1968; Butt et al., 1964). Kidney cortex also shows In-
creases in lead with age that may be associated with formation of lead nuclear inclusion
bodies (Indraprasit et al., 1974). Based on these rates of accumulation, the total body bur-
den may be divided into pools that behave differently. The largest and kinetlcally slowest
pool is the skeleton, which accumulates lead with age. The much more labile lead pool 1s in
soft tissue.
Soft-tissue lead levels generally stabilize 1n early adult life and show a turnover rate
similar to that for blood. This turnover is sufficient to prevent accumulation except in the
renal cortex, which may reflect formation of lead-containing nuclear inclusion bodies (Cramer
et al., 1974; Indraprasit et al., 1974). The data of Gross et al. (1975) and Barry (1975)
indicate that aortic levels rise with age, although this rise may only reflect entrapment of
lead in atherosclerotic deposits. Biliary and pancreatic secretions, while presumably re-
flecting some of the organ levels, have tracer lead concentrations distinct from either blood
or bone pools (Rabinowitz et al., 1973).
For levels of lead in soft tissue, the reports of Barry (1975, 1981), Gross et al. (1975),
and Horiuchi et al. (1959) indicate that soft-tissue content generally is below 0.5 ug/g
wet weight, with higher values for aorta and kidney cortex. The higher values in aorta may or
may not reflect lead in plaque deposits, while higher kidney levels may be associated with the
presence of lead-accumulating tubular cell nuclear inclusions. The relatively constant lead
concentration in lung tissue across age groups suggests no accumulation of respired lead and
is consistent with data for deposition and absorption (see Section 10.2.1). Brain tissue was
generally under 0.2 ppm wet weight and appeared to show no change with increasing age. Since
these data were collected by cross-sectional study, age-related changes in the low levels of
lead in brain would have been difficult to discern. Barry (1975) found that tissues in a
small group of samples from subjects with known or suspected occupational exposure showed
higher lead levels in aorta, liver, brain, skin, pancreas, and prostate.
Analysis of lead levels in whole brain is less illuminating than regional analysis to the
issue of sensitivity of certain regions within the organ to toxic effects of lead. The dis-
tribution of lead across brain regions has been reported by various laboratories. The rele-
vant data for humans and animals are set forth 1n Table 10-2. The data of Grandjean (1978)
and Niklowitz and Mandybur (1975) for human adults, and those of Okazaki et al. (1963) for
autopsy samples from young children who died of lead poisoning, are consistent 1n showing that
lead is selectively accumulated in the hippocampus. The correlation of lead level with potas-
sium level suggests that uptake of lead is greater in cellulated areas. The involvement of
10-20
-------
TABLE 10-2. DISTRIBUTION OF LEAD IN BRAIN REGIONS OF HUMANS AND ANIMALS
Subjects
Exposure status
Relative distribution
Reference
o
rss
Humans
Adult males
Children
"Unexposed"
Fatal lead poisoning
Child, 2 yr old Fatal lead poisoning
Adults
Animals
Adult rats
Adult rats
Neonatal rats
Young dogs
3 subjects "unexposed";
1 subject with lead
poisoning as child
"Unexposed"
"Unexposed"
Controls and
dally i.p. injection,
5.0 or 7.5 mg/kg
Controls and dietary
exposure, 100 ppm;
12 weeks of exposure
Hippocampus = amygdala > medulla
oblongata > half brain > optic tract
= corpus callosum. Pb correlated
with potassium.
Hippocampus > frontal cortex »
occipital white matter, pons
Cortical gray matter > basal ganglia >
cortical white matter
Hippocampus > cerebellum = temporal
lobes > frontal cortex in 3 unexposed
subjects; temporal lobes > frontal
cortex > hippocampus > cerebellum in
case with prior exposure
Hippocampus > amygdala » whole brain
Hippocampus had 50% of brain lead
with a 4:1 ratio of hippocampus
to whole brain concentrations
In both treated and control animals
cerebellum > cerebral cortex >
brainstem + hippocampus
Controls: cerebellum = medulla >
caudate > occipital gray > frontal
gray
Exposed: occipital gray > frontal
gray = caudate > occipital
white = thalamus > medulla > cerebellum
Grandjean (1978)
Okazaki et al. (1963)
Klein et al. (1970)
Niklowitz and
Mandybur (1975)
Danscher et al. (1975)
Fjerdingstad et al.
(1974)
Klein and Koch (1981)
Stowe et al. (1973)
-------
the cerebellum 1n lead encephalopathy 1n children (see Section 12.4) and 1n adult Intoxication
from occupational exposure Indicates that the sensitivity of various brain regions to lead as
well as their relative uptake characteristics are factors in lead neuropathology.
In adult rats, selective uptake of lead 1s shown by the hippocampus (Fjerdingstad et al.,
1974; Danscher et al., 1975) and the amygdala (Oanscher et al., 1975). By contrast, lead-
exposed neonate rats show greatest uptake of lead into cerebellum, followed by cerebral cor-
tex, then brainstem plus hippocampus. Hence, there 1s a developmental difference in lead dis-
tribution in the rat with or without increased lead exposure (Klein and Koch, 1981).
In studies of young dogs, "unexposed" animals showed highest levels 1n the cerebellum.
Increased lead exposure was associated with selective uptake Into gray matter, while cerebel-
lar levels were relatively low. Unlike the young rat, then, the distribution of lead in brain
regions of dogs appears dose-dependent (Stowe et al., 1973).
The relationship of lead distribution to various tissues with changes in lead exposure
has not been well researched. Available information does suggest that the nature of lead ex-
posure in experimental animals influences the relationship of tissue lead level to both blood
lead level and level of Intake. Long-term oral exposure of experimental animals at relatively
moderate dosing would appear to result 1n tissue values that show more proportionality to dose
than do blood lead values, although tissue versus blood lead relationships still appear to be
curvilinear. Such is the case with dogs exposed to dietary lead for 2 years (Azar et al.,
1973) and rats exposed in utero and postnatally up to 9 months of age (Grant et al., 1980).
By contrast, short-term exposure at various dosing levels yields highly variable data
(see Section 12.4.3.5 and Table 12-8). Bull et al. (1979) have reported brain and blood lead
data for dam-exposed suckling rats that show marked deviation from linear response to dose
when lead was administered in drinking water at 0.0005 to 0.02 percent lead. Over this 40-
fold oral dosing range, brain lead levels increased only approximately threefold at 21 days of
age. Whether this low absorption of lead by brain reflects tissue distribution curvilinearity
in the pups or reflects a function of nonlinear milk lead versus maternal dosing relationships
cannot be determined. Collins et al. (1982) reported that rats orally exposed to lead from 3
days of age for 4-8 weeks showed a two- to threefold increase in brain regions when the dosing
level was increased to 1.0 mg/kg from 0.1 mg/kg. Blood lead at these two dosing levels showed
a concentration ratio of -^2.5, indicating that both brain tissue and blood showed similar non-
linear response over this 10-fold change in oral exposure.
Barry (1975, 1981) compared lead levels in soft tissues of children and adults. Tissue
lead of infants under 1 year old was generally lower than in older children, while children
aged 1-16 years had values that were comparable to those for adult women. In Barry's (1981)
10-22
-------
study, the absolute concentration of lead In brain cortex or the ratios of brain cortex to
blood lead levels did not appear to be different 1n Infants or older children compared to
adults. Such direct comparisons do not account for relative tissue mass changes with age, but
this factor is comparatively less with soft tissue than with the skeletal system (see Section
10.4).
Subcellular distribution of lead in soft tissue is not uniform, with high amounts of lead
being sequestered in the mitochondria and nucleus. Cramer et al. (1974) studied renal biopsy
tissue in lead workers having exposures of variable duration. They observed lead-binding
nuclear Inclusion bodies in the renal proximal tubules of subjects having short exposure, with
all showing mitochondrlal changes. A considerable body of animal data (see Section 10.3.5)
documents the selective uptake of lead into these organelles. Pounds et al. (1982) describe
these organellar pools in kinetic terms as having comparatively short half-lives 1n cultured
rat hepatocytes, while McLachUn et al. (1980) found that rat kidney epithelial cells form
lead-sequestering nuclear Inclusions within 24 hr.
10.3.2.2 Mineralizing Tissue. Biopsy and autopsy data have shown that lead becomes localized
and accumulates in human calcified tissues, i.e., bones and teeth. The accumulation begins
with fetal development (Barltrop, 1969; Horiuchi et al., 1959).
Total lead content in bone may exceed 200 mg 1n men aged 60 to 70 years, but in women the
accumulation is somewhat lower. Various investigators (Barry, 1975; Horiguchi and Utsunomiya,
1973; Schroeder and Tlpton, 1968; Horiuchi et al., 1959) have documented that approximately 95
percent of total body lead is lodged in bone. These reports not only establish the affinity
of bone for lead, but also provide evidence that lead increases in bone until 50-60 years of
age, the later fall-off reflecting some combination of diet and mineral metabolism changes.
Tracer data show accumulation in both trabecular and compact bone (Rablnowitz et al., 1976).
In adults, bone lead is the most inert pool as well as the largest, and accumulation can
serve to maintain elevated blood lead levels years after past, particularly occupational, ex-
posure has ended. This fact accounts for the observation that duration of exposure correlates
with the rate of reduction of blood lead after termination of exposure (O'Flaherty et al.,
1982). The proportion of body lead lodged in bone is reported to be lower in children than 1n
adults, although concentrations of lead in bone increase more rapidly than 1n soft tissue
during childhood (Barry, 1975, 1981). In 23 children, bone lead was 9 mg, or 73 percent of
total body burden, versus 94 percent in adults. Expression of lead 1n bone in terms of con-
centration across age groups, however, does not accommodate the "dilution" factor, which is
quite large for the skeletal system in children (see Section 10.4).
The isotope kinetic data of Rabinowitz et al. (1976) and Holtzman (1978) Indicate biolog-
ical half-lives of lead 1n bone on the order of several decades, although it appears that
10-23
-------
there are two bone compartments, one of which is a repository for relatively labile lead
(Rabinowitz et al., 1977).
Tooth lead levels also increase with age at a rate proportional to exposure (Steenhout
and Pourtois, 1981), and are also roughly proportional to blood lead levels in man (Winneke et
al., 1981; Shapiro et al., 1978) and experimental animals (Kaplan et al., 1980). Dentine lead
is perhaps the most responsive component of teeth to lead exposure because it 1s laid down
from the time of eruption until the tooth is shed. Needleman and Shapiro (1974) have docu-
mented the usefulness of dentine lead as an indicator of the degree of subject exposure.
Fremlin and Edmonds (1980), using alpha-particle excitation and microautoradiography, have
shown dentine zones of lead enrichment related to abrupt changes in exposure. The rate of
lead deposition in teeth appears to vary with the type of tooth. Deposition is highest in the
central incisors and lowest in the molars, a difference that must be taken into account when
using tooth lead data for exposure assessment, particularly for low levels of lead exposure
(Mackie et al., 1977; Delves et al., 1982).
10.3.3 Chelatable Lead
Mobile lead in organs and systems is potentially more "active" toxicologically in terms
of being available to sites of action. Hence, the presence of diffusible, mobilizable, or ex-
changeable lead may be a more significant predictor of imminent toxicity or recent exposure
than total body or whole blood burdens. In reality, however, assays for mobile lead would be
quite difficult.
In this regard, chelatable urinary lead has been shown to provide an index of this
mobile portion of total body burden. Note that "chelatable" lead refers here to the use of
calcium disodium ethylenediaminetetraacetic acid (CaNa2EDTA) and body compartments accessible
to this chelant. Based mainly on the relationship of chelatable lead to indices of heme bio-
synthesis impairment, chelation challenge is now viewed as the most useful probe of undue body
burden in children and adults (U.S. Centers for Disease Control, 1978; World Health Organiza-
tion, 1977; Chisolm and Barltrop, 1979; Chisolm et al., 1976; Saenger et al., 1982; Hansen et
al., 1981). In adults, chelation challenge is the most reliable diagnostic test for assess-
ment of lead nephropathy, particularly when exposure is remote in time (Emerson, 1963; Wedeen
et al., 1979) or unrecognized (Batuman et al., 1981, 1983).
A quantitative description of inputs to the fraction of body lead that 1s chelatable from
various body compartments 1s difficult to define fully, but it very likely includes a sizable,
fairly mobile compartment within bone as well as within soft tissues. This assertion is based
on several factors. First, the amount of lead mobilized by chelation is age-dependent in non-
exposed adults (Araki, 1973; Araki and Ushio, 1982), while blood and soft-tissue lead levels
10-24
-------
are not (Barry, 1975). This difference indicates a lead pool labile to chelatlon but kineti-
cally distinct from soft tissue. Second, studies of chelatable lead in animals (Hammond,
1971, 1973) suggest removal of some bone lead fraction, as does the response of explanted
fetal rat bone lead to chelants (Rosen and Markowitz, 1980). Third, the tracer modeling esti-
mates of Rabinowitz et al. (1977) suggest a mobile bone compartment, and fourth, there is a
complex, nonlinear relationship of lead intake by air, food, and water (see Chapter 11) to
blood lead, and an exponential relationship of chelatable lead to blood lead (Chisolm et al.,
1976).
The logarithmic relationship of chelatable lead to blood lead in children (Chisolm et
al., 1976) is consistent with the studies of Saenger et al. (1982), who reported that levels
of mobilizable lead in "asymptomatic" children with moderate elevations in blood lead were
quite similar in many cases to those values obtained in children with signs of overt toxicity.
Hansen et al. (1981) reported that lead workers challenged with CaNa2EDTA showed 24-hr urine
lead levels that in many cases exceeded the accepted limits even though blood lead was only
moderately elevated in many of those workers. The action level corresponded, on the regres-
sion curve, to a blood lead value of 35 pg/dl.
Several reports provide insight into the behavior of labile lead pools in children treat-
ed with chelating agents over varying periods of time. Treatment regimens using CaNa2EDTA or
CaNa2EDTA + BAL (British anti-Lewi site, or dimercaprol) for up to 5 days have been invariably
associated with a "rebound" in blood lead, ascribed to a redistribution of lead among mobile
lead compartments (Chisolm and Barltrop, 1979). Marcus (1982) reported that 41 children given
oral D-penicillamine for 3 months showed a significant drop in blood lead by 2 weeks (mean
initial value of 53.2 ug/dl), then a slight rise that was within measurement error with a peak
at 4 weeks, and a fall at 6 weeks, followed by no further change at a blood lead level of
36 ug/dl. Hence, there was a near steady state at an elevated level for 10 of the 12 weeks
with continued treatment. This observation could have indicated that re-exposure was occur-
ring, with oral penicillamine and ingested lead leading to increased lead uptake, as seen by
Jugo et al. (1975a). However, Marcus (1982) stated that an effort was made to limit further
lead intake as much as possible. From these reports, a re-equilibration does appear to occur,
varying in characteristics with type and duration of chelation. The rebound seen in short-
term treatment with CaNa2EDTA or CaNa2EDTA + BAL, although attributed to soft tissue, could
well include a shift of lead from a larger mobile bone compartment to soft tissues and blood.
The apparent steady state between the blood lead pool and other compartments that is achieved
in the face of plumburesis, induced by D-penicillamine (Marcus, 1982), suggests a rather siza-
ble labile body pool which, in quantitative terms, would appear to exceed that of soft tissue
alone.
10-25
-------
Several studies of EDTA mobilization of lead 1n children (Saenger et al., 1982; PlomelH
et al., 1984) Indicate the relative merit of assessing chelatable lead burden 1n children
otherwise characterized as having mild or moderate lead exposure as Indicated by blood lead
levels. Saenger et al. (1982) noted that significant percentages of children having mild or
moderate lead exposure as commonly Indexed were found after EDTA challenge to have levels of
plumburesis that would qualify them for chelatlon therapy under U.S. Centers for Disease
Control (CDC) guidelines.
In the most comprehensive evaluation of this Issue to date (PlomelH et al., 1984), 210
children from four different urban lead-poisoning treatment centers were evaluated by EDTA
provocation testing. The results showed that at a blood lead level of 30-39 ug/dl, 12 percent
(6/52) of children exceed the ratio of 0.6 for ug Pb excreted per mg EDTA per 8 hr. This
ratio was selected by the study clinicians as differentiating children with mobile lead bur-
dens who require further evaluation and/or treatment. Thirty-eight percent of children with
blood lead levels of 40-49 ug/dl exceeded the action ratio of 0.6.
As indicated 1n Section 10.3.1, one basis for the curvilinear relationship between chela-
table lead and blood lead may be the curvilinear relationship of plasma lead to blood lead.
The former increases at a faster rate with exposure increases than blood lead, permitting an
increasingly greater rate of lead transfer to the chelatable lead compartment.
10.3.4 Mathematical Descriptions of Physiological Lead Kinetics
To account for observed kinetic data and make predictive statements, a variety of mathe-
matical models have been suggested, including those describing "steady-state" conditions.
Tracer experiments have suggested compartmental models of lead turnover based on a central
blood pool (Holtzman, 1978; Rabinowitz et al., 1976; Batschelet et al., 1979). These experi-
ments have hypothesized well-mixed, interconnected pools and have used coupled differential
equations with linear exponential solutions to predict blood and tissue lead exchange rates.
Were lead to be retained in these pools in accordance with a power-law distribution of resi-
dence times, rather than being uniform, a semi-Markov model would be more appropriate (Marcus,
1979).
In the model proposed by Rabinowitz et al. (1976), based on the use of stable lead Iso-
tope tracer 1n adult volunteers, lead biokinetics is envisioned in terms of three body com-
partments. These compartments, consisting of a central blood compartment as well as soft-
tissue and bone compartments, differ as to biological half-lives or mean-lives (half life =
mean-life x 0.693). Blood shows the shortest biological half-life, followed by soft tissue
and then the bone compartment. Bone contains most of total body lead burden.
10-26
-------
A more recent approach has been that of Kneip et al. (1983) for multi-organ compartmen-
talization of lead, based on data obtained with infant and juvenile baboons administered sin-
gle and chronic lead doses orally. The model proposed for infant baboons is depicted in
Figure 10-3. Figure 10-3 acknowledges differences in certain features of lead biokinetics
that differ in the developing versus adult organism. One of these differences is the lead
transfer rate from blood to bone. In addition, an extracellular space-gut (ECS-Gut) compart-
ment is included in Figure 10-3. The emphasis is on lead intake through the gut, and a
respiratory intake component is not included. In common with other attempts at modeling, the
blood compartment in the approach of Kneip et al. (1983) is not further characterized kineti-
cally, which is a limitation in view of the data base concerning such relationships as the
curvilinear one between plasma and blood lead (see Section 10.3.2).
Most extant steady-state models are deficient because they are based on small numbers of
subjects and neglect a dose dependency for some of the interpool transfer coefficients. In
this case, a nonlinear dose-indicator response model would be more appropriate when consid-
ering changes in blood lead levels. For example, the relationship between blood lead and air
lead (Hammond et al., 1981; Brunekreef, 1984) as well as that between diet (United Kingdom
Central Directorate on Environmental Pollution, 1982) and tap drinking water (Sherlock et al.,
1982) are all nonlinear in mathematical form. In addition, alterations in nutritional status
or the onset of metabolic stresses can complicate steady-state relationships.
In a series of papers, Marcus (1985a,b,c,d) has discussed linear and nonlinear multicom-
partmental models of lead kinetics and has addressed in particular the relationship between
plasma lead and blood lead and the relationship between blood lead and total lead intake. As
shown in Figure 10-4, Marcus (1985d) differentiated four discrete pools within the blood com-
partment: diffusible lead in plasma, protein-bound lead in plasma, a "shallow" red blood cell
pool (possibly the erythrocyte membrane), and a "deep" red blood cell pool (probably within
the erythrocyte). This model was based on previously published data from a volunteer subject
who ingested lead under controlled experimental conditions (DeSilva, 1981). Different ver-
sions of the model, all assuming steady-state conditions for lead in all tissues, were ana-
lyzed in terms of three possible mechanisms that might underlie nonlinear blood kinetics:
site-limited lead uptake, saturated active absorption, and increased urinary elimination
(Marcus, 1985c). The site-limited absorption model provided the best description of a non-
linear relationship between plasma lead and blood lead. Figure 10-5 shows the fit of the
model to data from 103 subjects studied by DeSilva (1981). At relatively high blood lead
levels, the fit appears quite satisfactory, but plasma lead is underestimated below 30 ug/dl
blood lead (see solid line in Figure 10-5). Adding an intercept term of 0.25 (see broken line
in Figure 10-5) improves the fit at low blood lead values. The need for an intercept term can
10-27
-------
INTAKE
GUT
EXCRETION
BONE
2
URINE EXCRETION
S
A12 =0.34 (INFANT) = 0.11 (JUVENILE)
A,, = 1.73 x 10'1
A,, = 0.10
A,, = 0.03
A,4 = 0.03
A,, = 0.07
A,. = 0.08
AM = 0.01
A., = 0.23
Figure 10-3. Schematic model of lead metabolism in infant baboons,
with compartments! transfer coefficients.
Source: Kneip et al. (1983).
10-28
-------
o
ro
LEAD IN
RBC DEEP
POOL
LEAD IN «
RBC SHALLOW
POOL
DIFFUSIBLE
LEAD IN
PLASMA
PROTEIN- l
BOUND LEAD
IN PLASMA
LEAD IN
SOR TISSUES
LEAD IN
EXTRACELLULAR
FLUID
LEAD IN
HARD TISSUES
I I
, x
V
BLOOD LEAD
Figure 10-4. A compartmental model for lead biokinetics with multiple pools for blood lead.
Source: Marcus <1985d).
-------
o
s
2
O
cc
O
O
o
<
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
1 I T
o
o
0 10 20 30 40 50 60 70 80 90 100 110 120
BLOOD LEAD CONCENTRATION,
Figure 10-5. Fitting of nonlinear blood lead model to data of
DeSilva (1981). Broken line incorporates an intercept term of
0.25; solid line does not incorporate intercept term.
Source: Marcus (1985c).
10-30
-------
be attributed to possible analytic error due to contamination of the plasma samples or to
transient fluctuations 1n plasma lead due to lead exposure just prior to sampling (Marcus,
1985c). In any event, curvllinearity 1s modest below 30 M9/dl. F°r Individuals without oc-
cupational or other excessive exposure to lead (>30 ^g/cll blood lead), 1t is not possible to
distinguish linear and nonlinear kinetic models (Marcus, 1985c).
10.3.5 Animal Studies
The relevant questions to be asked of animal data are those that cannot be readily or
fully satisfied by data from human subjects. What is the effect of exposure level on distri-
bution within the body at specific time points? What 1s the relationship of age or develop-
mental stage on the distribution of lead 1n organs and systems, particularly the nervous sys-
tem? What are the relationships of physiological stress and nutritional status to the redis-
bution kinetics? Can the relationship of chelatable lead to such indicator lead pools as
blood be defined better?
Administration of a single dose of lead to rats produces high initial lead concentrations
in soft tissues, which then fall rapidly as the result of excretion and transfer to bone
(Hammond, 1971), while the distribution of lead appears to be Independent of the dose.
X
Castellino and Aloj (1964) reported that single-dose exposure of rats to lead was associated
with a fairly constant ratio of erythrocyte lead to plasma lead, a rapid distribution to
tissues, and relatively higher uptake 1n liver, kidney, and particularly bone. Lead loss from
oTgans and tissues follows first-order kinetics except from bone. The data of Morgan et al.
(1977), Castellino and Aloj (1964), and Keller and Doherty (1980a) document that the skeletal
system 1n rats and mice is the kinetically rate-limiting step in whole-body lead clearance.
Subcellular distribution studies involving either tissue fractionation after jn vivo lead
exposure or j[n vitro data document that lead is preferentially sequestered in the nucleus
(Castellino and Aloj, 1964; Goyer et al., 1970) and mitochondria! fractions (Castellino and
Aloj, 1964; Barltrop et al., 1974) of cells from lead-exposed animals. Lead enrichment in the
mitochondrion is consistent with the high sensitivity of this organelle to the toxic effects
of lead.
The neonatal animal seems to retain proportionately higher levels of tissue lead compared
with the adult (Goldstein et al., 1974; Momcilovlc and Kostial, 1974; Mykkanen et al., 1979;
Klein and Koch, 1981) and shows slow decay of brain lead levels while other tissue levels sig-
nificantly decrease over time. This decay appears to result from enhanced entry by lead due
to a poorly developed brain barrier system in the developing animals, as well as enhanced body
retention in the young animals. The effects of such changes as metabolic stress and nutri-
tional status have been noted 1n the literature. Keller and Doherty (1980b) have documented
10-31
-------
that tissue redistribution of lead, specifically bone lead mobilization, occurs 1n
lactatlng female mice, with both lead and calcium transfer occurring from mother to pups
(Keller and Doherty, 1980c). Changes 1n lead movement from body compartments, particularly
bone, with changes 1n nutrition are described in Section 10.5.
In animal studies that are relevant both to the Issue of chelatable lead versus lead in-
dicators in humans and to the relative lability of lead 1n the young versus the adult, Jugo et
al. (1975b) and Jugo (1980) studied the chelatabUHy of lead 1n neonate versus adult rats and
its lability in the erythrocyte. Challenging young rats with metal chelants yielded propor-
tionately lower levels of urinary lead than in the adult, a finding that has been ascribed to
tighter binding of lead 1n the young animal (Jugo et al., 1975b). In a related observation,
the chelatable fraction of lead bound to erythrocytes of young animals given 203Pb was approx-
imately threefold greater than in the adult rat (Jugo, 1980), although the fraction of dose 1n
the cells was higher in the suckling rat. The difference in the suckling rat erythrocyte re-
garding the binding of lead and relative content compared with the adult may be compared with
Ong and Lee's (1980b) observation that human fetal hemoglobin binds lead more avidly than does
mature hemoglobin.
10.4 LEAD EXCRETION AND RETENTION IN HUMANS AND ANIMALS
Dietary lead that is not absorbed in humans and animals passes through the, GI tract and
is eliminated with feces, as is the deposited fraction of air lead that is swallowed and not
absorbed. Lead absorbed into the blood stream and not retained is excreted through the renal
and GI tracts, the latter by biliary clearance. The amounts appearing in urine and feces
appear to be a function of such factors as species, age, and differences 1n dosing.
10.4.1 Human Studies
Booker et al. (1969) found that 212Pb injected into two adult volunteers led to initial
appearance of the label in urine (4.4 percent of dose in 24 hr), then in both urine and feces
in approximately equal amounts. By use of the stable isotope 204Pb, Rab1now1tz et al. (1973)
reported that urinary and fecal excretion of the label amounted to 38 and 8 ug/day in adult
subjects, accounting for 76 and 16 percent, respectively, of the measured recovery. Fecal ex-
cretion was thus approximately twice that of all the remaining modes of excretion: hair,
sweat, and nails (8 percent).
Perhaps the most detailed study of lead excretion in adult humans was done by Chamberlain
et al. (1978), who administered 203Pb by Injection, Inhalation, and Ingestion. After Injec-
tion or oral intake, the amounts in urine (Pb-U) and feces (Pb-Fe, endogenous fecal lead) were
10-32
-------
compared for the two administration routes. Endogenous fecal lead was 50 percent of that in
urine, or a 2:1 ratio of urinary to fecal lead. (Increased transit time was allowed for fecal
lead to pass through the GI tract.)
Based on the metabolic balance and isotope excretion data of Kehoe (1961a,b,c), Rabino-
witz et al. (1976), and Chamberlain et al. (1978), as well as on some recalculations of the
Kehoe and Rabinowitz data by Chamberlain et al. (1978), short-term lead excretion amounts to
50-60 percent of the absorbed fraction, the balance moving primarily to bone with some sub-
sequent fraction (approximately half) of this stored amount eventually being excreted. The
rapidly excreted fraction was determined by Chamberlain et al. (1978) to have an excretion
half-life of about 19 days. This value is consistent with the estimates of Rabinowitz et al.
(1976), who expressed clearance in terms of mean-lives. Mean-lives are multiplied by In 2
(0.693) to arrive at half-lives. The similarity of the blood 203Pb half-life with that of
body excretion noted by Chamberlain et al. (1978) indicates a steady rate of clearance from
the body.
The age dependency of lead excretion rates in humans has not been well studied; all of
the above lead excretion data involved only adults. Table 10-3 combines available data from
adults (Rabinowitz et al., 1977; Thompson, 1971; Chamberlain et al., 1978) and infants
(Ziegler et al., 1978) for purposes of comparison. Intake, urine, fecal, and endogenous fecal
lead data from two studies on adults and one report on infants are used. For consistency in
the adult data, 70 kg is used as an average adult weight, and a Pb-Fe:Pb-U ratio of 0.5 is
used. Daily lead intake, absorption, and excretion values are expressed as ug/kg body weight.
For the infant data, daily endogenous fecal lead excretion is calculated using the adult ratio
as well as the extrapolated value of 1.5 ug/kg. The respiratory lead intake value for the
infants is an upper value (0.2 ug/m3), since Ziegler et al. (1978) found air lead to be <0.2
ug/m3. Compared to the two representative adult groups, infants appear to have a lower total
excretion rate, although the excretion of endogenous fecal lead may be higher than for adults.
In humans, the dependence of lead excretion rate on level of exposure has been studied in
some detail by Chamberlain (1983), who used data from the published reports of King et al.
(1979), Williams et al. (1969), Gross (1981), Devoto and Spinazzola (1973), Azar et al.
(1975), and Chamberlain et al. (1978). Figure 10-6 reproduces Chamberlain's plots of urinary
excretion rate for lead versus blood lead as provided in the various studies. Renal clearance
of lead appears to increase as blood lead increases from 25 to 80 ug/dl, the highest blood
value reported. Given the earlier discussion concerning the increased fractional partitioning
of blood lead into plasma with increasing blood lead burden (see Section 10.3.1), one would
anticipate an increasing renal excretion rate for lead over a broad range of blood lead.
10-33
-------
TABLE 10-3. DAILY LEAD EXCRETION AND RETENTION DATA FOR ADULTS AND INFANTS
Dietary Intake (|jg/kg)
Fraction of intake absorbed
Diet lead absorbed (ug/kg)
A1r lead absorbed ((jg/kg)
Total absorbed lead (ug/kg)
Urinary lead excreted (ug/kg)
Ratio: urinary/absorbed lead
Endogenous fecal lead (ng/kg)
Total excreted lead (ug/kg)
Ratio: total excreted/absorbed
lead
Fraction of Intake retained
Children3
10.76
0.46 (0.55)d
4.95 (5.92)
0.20
5.15 (6.12)
1.00
0.19 (0.16)
0.5 (1.56)f
1.50 (2.56)
0.29 (0.42)
0.34 (0.33)
Adult b
group A
3.63
0.15e
0.54
0.21
0.75
0.47
0.62 '
0.249
0.71
0.92
0.01
Adult
group B
3.86
0.15e
0.58
0.11
0.68
0.34
0.50
0.179
0.51
0.75
0.04
aZ1egler et al. (1978).
bRab1now1tz et al. (1977).
cThompson (1971) and estimates of Chamberlain et al. (1978).
dEach of the values 1n parentheses 1n this column 1s corrected for endogenous fecal
lead at extrapolated value from Zlegler et al. (1978).
Corrected for endogenous fecal lead (Pb-Fe = 0.5 x Pb-U).
Extrapolated value of 1.56 for endogenous fecal Pb.
gPb-Fe = 0.5 x Pb-U.
10-34
-------
0.3
ui
U
1
U
UJ
E
0.2
0.1
iii\i
i
10 20 30 40 50 60
BLOOD LEAD, w/100 g
70
80
Figure 10-6. Renal clearance (ratio of urinary lead to
blood lead) from (A) King et al., 1979; (B) Williams et
al., 1969; (C) Gross, 1981; (D) DeVoto and Spinazzola,
1973; (E) Azar et al.. 1975; (G) Chamberlain et al.,
1978.
Source: Chamberlain (1983).
10-35
-------
Data in Figure 10-6 indicate increased renal excretion of lead only. How the correspon-
ding biliary excretion rate changes in the face of increasing lead absorption is not known.
Hence, the overall impact of increasing exposure on total body clearance of the toxicant is
difficult to assess. In experimental animals, the relative partitioning of lead between renal
and biliary excretion routes has been shown to be dose- and species-dependent (see Section
10.4.2).
Lead accumulates in the human body with age, mainly in bone, up to approximately 60 years
of age, when a decrease occurs with changes in intake as well as in bone mineral metabolism.
Total accumulation by 60 years of age ranges up to approximately 200 mg (see review by Barry,
1978), although occupational exposure can raise this figure several-fold (Barry, 1975).
Holtzman (1978) has reviewed the available literature on studies of lead retention in bone.
In normally exposed humans a biological half-life of approximately 17 years has been calcula-
ted, while data for uranium miners yield a range of 1320-7000 days (4-19 years). Chamberlain
et al. (1978) have estimated lifetime averaged daily retention at 9.5 ug using data of Barry
(1975). Within shorter time frames, however, retention can vary considerably due to such fac-
tors as disruption of the individual's equilibrium with changes in level of exposure, the dif-
ferences between children and adults, and, in elderly subjects, the presence of osteoporosis
(Gross and Pfitzer, 1974).
Lead labeling experiments, such as those of Chamberlain et al. (1978), indicate a short-
term or initial retention of approximately 40-50 percent of the fraction absorbed. Much of
this retention is by bone. Determining how much lead resorption from bone will eventually
occur using labeled lead is difficult, given the extremely small fraction of labeled to
unlabeled lead (i.e., label dilution) that would exist. Based on the estimates of Kehoe
(1961a,b,c), the Gross (1981) evaluation of the Kehoe studies, the RabinowHz et al. (1976)
study, the Chamberlain et al. (1978) assessments of the aforementioned reports, and the data
of Thompson (1971), one can estimate that approximately 25 percent of the lead absorbed daily
undergoes long-term bone storage.
The above estimates relate either to adults or to long-term retention over most of an in-
dividual's lifetime. Studies with children and developing animals (see Section 10.4.2) indi-
cate lead retention in childhood can be higher than in adulthood. By means of metabolic
balance studies, Ziegler et al. (1978) obtained a retention figure (as percentage of total in-
take) of 31.5 percent for infants, while Alexander et al. (1973) provided an estimate of 18
percent. Corrected retention data for both total and absorbed intake for the pediatric sub-
jects of Ziegler et al. (1978) were shown in Table 10-3, using the two values for endogenous
fecal excretion as noted. Barltrop and Strehlow (1978) calculated a net negative lead reten-
tion in their subjects, but problems in comparing this report with the others were noted
10-36
-------
earlier. Given the increased retention of lead in children relative to adults, as well as the
greater rate of lead intake on a body-weight basis, increased uptake in soft tissues and/or
bone is indicated.
Barry (1975, 1981) measured the lead content of soft and mineral tissues in a small group
of autopsy samples from children 16 years of age and under, and noted that average soft-tissue
values were comparable to those in female adults, while mean bone lead values were lower than
in adults. These results suggest that bone in children has less retention capacity for lead
than bone in adults. Note, however, that "dilution" of bone lead will occur because of the
significant growth rate of the skeletal system through childhood. Trotter and Hixon (1974)
studied changes in skeletal mass, density, and mineral content as a function of age, and noted
that skeletal mass increases exponentially in children until the early teens, increases less
up to the early 20s, levels off in adulthood, and then slowly decreases. From infancy to the
late teens, bone mass increases up to 40-fold. Barry (1975) noted an approximate doubling in
bone lead concentration over this interval, indicating that total skeletal lead had actually
increased 80-fold. He also obtained a mean total bone lead content of approximately 8 mg for
children up to 16 years old, compared with a value of approximately 18 mg estimated from both
the bone concentrations in his study of children at different ages and the bone growth data of
Trotter and Hixon (1974). In a later study (Barry, 1981), autopsy samples from infants and
children between 1 and 9 years old showed an approximately 3.5-fold increase in mean bone con-
centrations across the three bone types studied, compared with a skeletal mass increase from
0-6 months to 3-13 years old of greater than 10-fold, for an estimated increase in total lead
of approximately 35-fold. Five reports (see Barry, 1981) noted age versus tissue lead rela-
tionships indicating that overall bone lead levels in infants and children were less than in
adults, whereas four reports observed comparable levels in children and adults.
If one estimates total daily retention of lead in the infants studied by Ziegler et al.
(1978), using a mean body weight of approximately 10 kg and the corrected retention rate in
Table 10-3, one obtains a total daily retention of approximately 40 pg. By contrast, the
total reported or estimated skeletal lead accumulated between 2 and 14 years is 8-18 mg (vide
supra), which averages out to a daily long-term retention of 2.0 to 4.5 ug/day or 6-13 percent
of total retention. Lead retention may be highest in infants up to about 2 years of age (the
subjects of the Ziegler et al. study), then decreases in older children. The mean retention
in the Alexander et al. (1973) study was 18 percent, about half that seen by Ziegler et al.
(1978). This difference may result from the greater age range in the former study.
"Normal" blood lead levels in children either parallel adult male levels or are approxi-
mately 30 percent greater than adult female levels (Chamberlain et al., 1978), indicating
(1) that the soft-tissue lead pool in very young children is not greatly elevated and thus,
10-37
-------
(2) that there is a huge labile lead pool in bone that is still kinetically quite distinct
from soft-tissue lead or (3) that in young children, blood lead is a much less reliable indi-
cator of greatly elevated soft-tissue or labile bone lead than is the case with adults. Barry
(1981) found that soft-tissue lead levels were comparable in infants SI year old and children
1-5 and 6-9 years old.
Given the implications of the above discussion--that retention of lead 1n young children
is higher than in adults and possibly older children, while at the same time their skeletal
system is less effective for long-term lead sequestrationthe very young child is at greatly
elevated risk to a toxicologically "active" lead burden. For further discussion, see Chapter
13.
Rabinowitz et al. (1976) examined the biokinetics of a stable isotope of lead (20
-------
of Lloyd et al. (1975), who observed 75 percent of the excreted lead eliminated through bili-
ary clearance. Note that the latter researchers used carrier-free label while the other in-
vestigators used injections with carrier at levels of 3.0 mg Pb/kg. In mice, Keller and
Doherty (1980a) observed that the cumulative excretion rate of 210Pb in urine was 25-50 per-
cent of that in feces. In nonhuman primates, Cohen (1970) observed that baboons excreted lead
at the rate of 40 percent in feces and 60 percent in urine. Pounds et al. (1978) noted that
the rhesus monkey lost 30 percent of lead by renal excretion and 70 percent by fecal excre-
tion. This discrepancy may also reflect a carrier-dosing difference.
The extent of total lead excretion in experimental animals given labeled lead orally or
parenterally varies, in part due to the time frames for post-exposure observation. In the
adult rat, Morgan et al. (1977) found that 62 percent of injected 203Pb was excreted by 6
days. By 8 days, 66 percent of injected 203Pb was eliminated in the adult rats studied by
Momcilovic and Kostial (1974), while the 210Pb excretion data of Castellino and Aloj (1964)
for the adult rat showed 52 percent excreted by 14 days. Similar data were obtained by
Klaassen and Shoeman (1974). Lloyd et al. (1975) found that dogs excreted 52 percent of in-
jected lead label by 21 days, 83 percent by 1 year, and 87 percent by 2 years. In adult mice
(Keller and Doherty, 1980a), 62 percent of injected lead label was eliminated by 50 days. In
nonhuman primates, Pounds et al. (1978) measured approximately 18 percent excretion in adult
rhesus monkeys by 4 days.
Kinetic studies of lead elimination in experimental animals indicate that excretion is
described by two or more components. From the elimination data of Momcilovic and Kostial
(1974), Morgan et al. (1977) estimated that in the rat the excretion curve obeys a two-compo-
nent exponential expression with half-lives of 21 and 280 hr. In dogs, Lloyd et al. (1975)
found that excretion could be described by three components, i.e., a sum of exponentials with
half-lives of 12 days, 184 days, and 4951 days. Keller and Doherty (1980a) reported that the
half-life of whole-body clearance of injected 203Pb consisted of an initial rapid and a much
slower terminal component, the latter having a half-life of 110 days in the adult mouse.
The dependency of excretion rate on dose level has been investigated in several studies.
Although Castellino and Aloj (1964) saw no difference in total excretion rate when label was
injected with 7 or 100 ug of carrier, Klaassen and Shoeman (1974) did observe that the excre-
tion rate by biliary tract was dose-dependent at 0.1, 1.0, and 3.0 mg Pb/kg (urine values were
not provided for obtaining estimates of total excretion). Momcilovic and Kostial (1974) ob-
served an increased rate of excretion into urine over the added carrier range of 0.1 to 2.0 ug
Pb/kg with no change in fecal excretion. In the report of Aungst et al. (1981), excretion
rate in the rat did not change over the injected lead dosing range of 1.0 to 15.0 mg/kg. Rat
urinary excretion rates thus seem dose-dependent over a narrow range less than 7 ug, while
10-39
-------
elimination of lead through biliary clearance is dose-dependent up to an exposure level of 3
mg/kg.
Lead movement from lactatlng animals to their offspring via milk constitutes both a route
of excretion for the mother and a route of exposure for the young. Investigations directed at
this phenomenon have examined both prior-plus-ongoing maternal lead exposure during lactation
and the effects of Immediate prior treatment. Keller and Doherty (1980b) exposed two groups
of female rats to 210Pb: one group for 105 days before mating; the second before and during
gestation and nursing. During lactation, there was an overall loss of lead from the bodies of
the lactating females compared with controls, while the femur ash weights were Inversely re-
lated to level of lead excretion, Indicating that such enhancement is related to bone mineral
metabolism. Lead transfer via milk was approximately 3 percent of maternal body burden, in-
creasing with continued lead exposure during lactation. Lorenzo et al. (1977) found that
blood lead levels 1n nursing rabbits given Injected lead peaked rather rapidly (within 1 hr),
while milk lead levels showed a continuous increase for about 8 days, at which point the con-
centration of lead was eightfold higher than in blood. This observation indicates that the
transfer of lead to milk can occur against a concentration gradient 1n blood. Momc1lov1c"
(1978) and Kostial and Momcilovic" (1974) observed that transfer of 203Pb in the late stage of
lactation occurs readily in the rat, with higher overall excretion of lead in nursing versus
control females. Furthermore, the rate of lead movement to milk appeared dose-dependent over
the added lead carrier range of 0.2 to 2.0 ug.
The comparative retention of lead 1n developing versus adult animals has been investigat-
ed in several studies using rats, mice, and nonhuman primates. Momcilovid and Kostial (1974)
compared the kinetics of lead distribution in suckling and adult rats after Injection of
203Pb. Over an 8-day interval, 85 percent of the label was retained in the suckling rat, com-
pared with 34 percent in the adult. Keller and Doherty (1980a) compared the levels of 210Pb
in 10-day-old mice and adults, noting from the clearance half-lives (vide supra) that lead
retention was greater 1n the suckling animals than 1n the adults. In both adult and young
mice, the rate of long-term retention was governed by the rate of release of lead from bone,
indicating that in the mouse, skeletal lead retention in the young is greater than in the
adult. With infant and adult monkeys orally exposed to 210f>b, Pounds et al. (1978) observed
that at 23 days the corresponding amounts of initial dose retained were 92.7 and 81.7 percent,
respectively.
The studies of Rader et al. (1981a,b) are of particular interest because they demonstrate
not only that young experimental animals continue to show greater retention of lead in tissue
when exposure occurs after weaning, but also that such retention occurs 1n terms of either
uniform exposure (Rader et al., 1981a) or uniform dosing (Rader et al., 1981b) when compared
with adult animals. With uniform exposure, 30-day-old rats given lead in drinking water
10-40
-------
showed significantly higher lead levels 1n blood and higher percentages of dose retained 1n
brain, femur, and kidney, as well as higher Indices of hematopoietlc Impairment (ALA in urine,
erythrocyte porphyrln) when compared to adult animals. As a percentage of dose retained,
levels of lead retained in the tissue of the young animals were approximately two- to three-
fold higher. In part, this difference results from a higher ingestlon rate of lead. However,
in the uniform dosing study where a higher Ingestlon rate was not the case, an increased re-
tention of lead still prevailed, the amount of lead in brain being approximately 50 percent
higher in young versus adult animals. Comparison of values in terms of percent retained is
more meaningful for such assessments, because the factor of changes In organ mass (see above)
is taken into account. Delayed excretion of lead in the young animal may reflect an Immature
excretory system or a tighter binding of lead 1n various body compartments.
10.5 INTERACTIONS OF LEAD WITH ESSENTIAL METALS AND OTHER FACTORS
Deleterious agents, particularly toxic metals such as lead, do not express their toxico-
kinetic or toxlcological behavior in a physiological vacuum, but rather are affected by inter-
actions of the agent with a variety of biochemical factors such as nutrients. Growing recog-
nition of this phenomenon and Its Implications for lead toxiclty in humans has prompted a
number of studies, many of them recent, that address both the scope and mechanistic nature of
such interactive behavior.
Taken collectively, the diverse human and animal data described in this section make 1t
clear that there is heterogeneity 1n pediatrlc populations in terms of relative risk for lead
exposure and deleterious effects depending on nutritional status. Children having multiple
nutrient deficiencies are at greater risk.
10.5.1 Human Studies
In humans, the interactive behavior of lead and various nutritional factors is appropri-
ately viewed as particularly significant for children, since this age group 1s not only parti-
cularly sensitive to lead's effects, but also experiences the greatest flux in relative nutri-
ent status. Such interactions occur against a backdrop of rather widespread deficiencies in a
number of nutritional components in children. While such deficiencies are more pronounced in
lower-income groups, they exist in all socioeconomic strata. Mahaffey and Michaelson (1980)
have summarized the three national nutritional status surveys carried out 1n the United States
for Infants and young children: the Preschool Nutrition Survey, the Ten State Nutrition Sur-
vey, and the Health Assessment and Nutrition Evaluation Survey (HANES I). The most recent
body of data of this type is the second National Health Assessment and Nutrition Evaluation
Survey (NHANES II) study (Mahaffey et al., 1979), although the dietary Information from it has
10-41
-------
yet to be reported. In the older surveys, iron deficiency was the most common nutritional
deficit in children under 2 years of age, particularly children from low-income groups. Re-
duced vitamin C intake was noted in about one-third of the children, while sizable numbers of
them had significantly reduced intakes of calcium. Owen and Lippman (1977) reviewed the
regional surveys of low-income groups within Hispanic, white, and black populations. In these
groups, iron deficiency was a common finding, and low intakes of calcium and vitamins A and C
were observed regularly. Hambidge (1977) concluded that zinc intake in low-income groups is
generally inadequate relative to recommended daily allowances.
Available data from a number of reports document the association of lead absorption with
suboptimal nutritional status. Mahaffey et al. (1976) summarized their studies showing that
children with blood lead levels greater than 40 ng/dl had significantly (p <0.01) lower intake
of phosphorus and calcium compared with a control group, while iron intake in the two groups
was comparable. This study involved children 1-4 years old from an inner-city, low-income
population, with close matching for all parameters except the blood lead level. Sorrell et
al. (1977), in their nutritional assessment of 1- to 4-year-old children with a range of blood
lead levels, observed that blood lead content was inversely correlated with calcium intake,
while children with blood lead levels >60 ug/dl had significantly (p <0.001) lower intakes of
calcium and vitamin D.
Rosen et al. (1980, 1981) found that children with elevated blood lead (33-120 ug/dl) had
significantly lower serum concentrations of the vitamin D metabolite 1,25-dihydroxyvitamin D
(1,25-(OH)2D) compared with age-matched controls (p <0.001), and showed a negative correlation
of serum 1,25-(OH)2D with lead over the range of blood lead levels measured (see Chapter 12,
Section 12.5, for further discussion). These observations and animal data (Barton et al.,
1978a; see Section 10.5.2) may suggest an increasingly adverse interactive cycle of
1,25-(OH)2D, lead, and calcium 1n which lead reduces biosynthesis of the vitamin D metabolite.
This cycle leads to reduced induction of calcium binding protein (CaBP), less absorption of
calcium from the gut, and greater uptake of lead, thus further reducing metabolite levels.
Barton et al. (1978a) Isolated two mucosal proteins in rat intestine, one of which bound
mainly lead and was not vitamin D-st1mulated. The second bound mainly calcium and was under
vitamin control. The authors suggested direct site-binding competition between lead and cal-
cium 1n these proteins. Hunter (1978) Investigated the possible Interactive role of seasonal
vitamin D biosynthesis 1n adults and children; lead poisoning occurs more often 1n summer than
in other seasons (see Hunter, 1977, for review). Seasonality accounts for 16 percent of ex-
plained variance of blood lead levels in black children, 12 percent 1n Hispanic children, and
4 percent in white children. More recently, it has been documented that there 1s no seasonal
variation in circulating levels of 1,25-(OH)2D, the metabolite that affects the rate of lead
10-42
-------
absorption from the GI tract (Chesney et al., 1981). These results suggest that seasonality
1s related to changes in exposure.
Johnson and Tenuta (1979) determined that calcium intake was negatively correlated
(r = -0.327, p <0.05) with blood lead in 43 children aged 1-6 years. The high lead group
consumed less zinc than children with lower blood levels. Yip et al. (1981) found that 43
children with elevated blood lead (>30 ug/dl) and erythrocyte protoporphyrin (EP) (>35 (jg/dl)
had an increased prevalence of iron deficiency as these two parameters increased. Children
classed in CDC categories Ib and II had a 79 percent iron deficiency rate, while those in
Class III were all iron deficient. Chisolm (1981) demonstrated an inverse relationship
between chelatable iron and chelatable body lead levels as indexed by urinary ALA levels in 66
children with elevated blood lead. Watson et al. (1980) reported that adult subjects who were
Iron deficient (determined from serum ferritin measurement) showed a lead absorption rate 2-3
times greater than subjects who were iron replete. In a group of 13 children, Markowitz and
Rosen (1981) reported that the mean serum zinc levels in children with plumbism were signifi-
cantly below the values seen in normal children. Chelation therapy reduced the mean level
even further. Chisolm (1981) reported an Inverse relationship between ALA in urine (ALA-U)
and the amount of chelatable or systemically active zinc in 66 children challenged with EDTA
and having blood lead levels ranging from 45 to 50 pg/dl. These two studies suggest that zinc
status 1s probably as Important an interactive modifier of lead toxicity as is either calcium
or iron.
The role of nutrients 1n lead absorption has been reported in several metabolic balance
studies for both adults and children. Zlegler et al. (1978), 1n their investigations of lead
absorption and retention in Infants, observed that lead retention was inversely correlated
with calcium intake, expressed either as a percentage of total Intake (r = -0.284, p <0.01) or
on a weight basis (r = -0.279, p <0.01). Interestingly, the calcium intake range measured was
within the range considered adequate for Infants and toddlers by the National Research Council
(National Academy of Sciences, National Research Council, 1974). These data also support the
premise that severe deficiency need not be present for an interactive relationship to occur.
Using adults, Heard and Chamberlain (1982) monitored the uptake of 203Pb from the gut in eight
subjects as a function of the amounts of dietary calcium and phosphorus. Without supplementa-
tion of these minerals 1n fasting subjects, the label absorption rate was approximately 60
percent, compared to 10 percent with 200 mg calcium plus 140 mg phosphorus, the amounts pres-
ent 1n an average meal. Calcium alone reduced uptake by a factor of 1.3 and phosphorus alone
by 1.2; both together yielded a reduction factor of 6. This work suggests that Insoluble cal-
cium phosphate is formed and co-precipitates any lead present. This Interpretation is sup-
ported by animal data (see Section 10.5.2).
10-43
-------
10.5.2 Animal Studies
Reports of lead-nutrient Interactions 1n experimental animals have generally described
such relationships In terms of a single nutrient, using relative absorption or tissue reten-
tion in the animal to Index the effect. Most of the recent data are concerned with the impact
of dietary levels of calcium, iron, phosphorus, and vitamin D. Furthermore, some investiga-
tors have attempted to elucidate the site(s) of interaction as well as the mecham'sm(s)
governing the interactions. Lead's interactions involve the effect of the nutrient on lead
uptake, as well as lead's effect on nutrients. The focus of this discussion is on the former.
These interaction studies are tabulated in Table 10-4.
10.5.2.1 Interactions of Lead with Calcium. The early report of Sobel et al. (1940) noted
that variation of dietary calcium and other nutrients affected the uptake of lead by bone and
blood in animals. Subsequent studies by Mahaffey-Six and Goyer (1970) in the rat have demon-
strated that a considerable reduction in dietary calcium was necessary (from 0.7 percent to
0.1 percent), at which level blood lead was increased fourfold, kidney lead content was ele-
vated 23-fold, and relative toxidty (Mahaffey et al., 1973) was increased. The changes in
calcium necessary to alter lead's effects 1n the rat appear to be greater than those seen by
Zlegler et al. (1978) in young children, which indicates a species difference in terms of sen-
sitivity to basic dietary differences as well as to levels of all interactive nutrients.
These observations in the rat have been confirmed by Kostial et al. (1971), Quarterman and
Morrison (1975), Barltrop and Khoo (1975), and Barton et al. (1978a). The inverse relation-
ship between dietary calcium and lead uptake has also been noted in the pig (Hsu et al.,
1975), horse (Wllloughby et al., 1972), lamb (Morrison et al., 1977), and domestic fowl (Berg
et al., 1980).
The mechanism(s) governing lead's interaction with calcium operate at both the gut wall
and within body compartments. Barton et al. (1978a), using everted duodenal sac preparations
in the rat, reported the following: (1) Interactions at the gut wall require the presence of
intubated calcium to affect lead label absorption (pre-existing calcium deficiency in the
animal and no added calcium had no effect on lead transport); (2) calcium-deficient animals
show increased retention of lead rather than absorption (confirmed by Quarterman et al.,
1973); and (3) lead transport may be mediated by two mucosal proteins, one of which has high
molecular weight and a high proportion of bound lead, and is affected in extent of lead bind-
ing with changes in lead uptake. The second protein binds mainly calcium and 1s vitamin D-
dependent.
Smith et al. (1978) found that lead is taken up at a different site in the duodenum of
rats than is calcium, but absorption does occur at the site of phosphate uptake, suggesting a
complex interaction of phosphorus, calcium, and lead. This observation is consistent with the
data of Barltrop and Khoo (1975) for rats and the data of Heard and Chamberlain (1982) for
10-44
-------
TABLE 10-4. EFFECT OF NUTRITIONAL FACTORS ON LEAD UPTAKE IN ANIMALS
Factor
Species
Index of effect
Interactive effect
Reference
o
I
en
Calcium
Calcium
Calcium
Calcium
Calcium
Iron
Iron
Iron
Rat Lead in tissues and
effect severity at
low levels of dietary
calcium
Pig Lead in tissues at
low levels of
dietary calcium
Horse Lead in tissues at
low levels of
dietary calcium
Lamb Lead in tissues at
low levels of
dietary calcium
Rat Lead retention
Rat Tissue levels and
relative toxicity
of lead
Rat Lead absorption in
everted duodenal
sac preparation
Mouse Lead retention
Low dietary calcium (0.13!)
increases lead absorption
and severity of effects
Increased absorption of
lead with low dietary
calcium
Increased absorption of
lead with low dietary
calcium
Increased absorption of
lead with low dietary
calcium
Retention increased in
calcium deficiency
Iron deficiency increases
lead absorption and
toxicity
Reduction in intubated
iron increases lead
absorption; increased
levels decrease lead
uptake
Iron deficiency has no
effect on lead
retention
Mahaffey-Six and Goyer
(1970); Mahaffey et al.
(1973)
Hsu et al. (1975)
Willoughby et al. (1972)
Morrison et al. (1977)
Barton et al. (1978a)
Mahaffey-Six and Goyer
(1972)
Barton et al. (1978b)
Hamilton (1978)
-------
TABLE 10-4. (continued)
Factor
Iron
Species
Rat
Index of effect
In utero or milk
Interactive effect
Iron deficiency increases
Reference
Cerklewski (1980)
o
I
CTi
Phosphorus
Phosphorus
Phosphorus
Vitamin D
Vitamin D
Lipid
Protein
Rat
Rat
Rat
Rat
Rat
Rat
Rat
transfer of lead in
pregnant or lactating
rats
Lead uptake in tissues
Lead retention
Lead retention
Lead absorption
using everted sac
techniques
Lead absorption
using everted sac
techniques
Lead absorption
Lead uptake by tissues
both in utero and milk
transfer of lead to
sucklings
Reduced phosphorus increased
203Pb uptake 2.7-fold
Low dietary phosphorus
enhances lead retention; no
effect on lead resorption
in bone
Low dietary phosphorus
enhances both lead
retention and deposition
in bone
Increasing vitamin D
increases intubated
lead absorption
Both low and excess
levels of vitamin D
increase lead uptake
by affecting motility
Increases in lipid (corn
oil) content up to
40% enhance lead
absorption
Both low and high protein
in diet increase lead
absorption
Barltrop and Khoo (1975)
Quarterman and Morrison
(1975)
Barton and Conrad (1981)
Smith et al. (1978)
Barton et al. (1980)
Barltrop and Khoo (1975)
Barltrop and Khoo (1975)
-------
TABLE 10-4. (continued)
Factor
Species
Index of effect
Interactive effect
Reference
Protein
Protein
Zinc
Zinc
Zinc
Copper
Rat
Rat
Milk components Rat
Milk components Rat
Rat
Rat
Rat
Rat
Body lead retention
Tissue levels of
lead
Lead absorption
Lead absorption
Lead absorption
Lead transfer jn
utero and in milk
during lactation
Tissue retention
Lead absorption
Low dietary protein either
reduces or does not affect
retention in various
tissues
Casein in diet increases
lead uptake compared to
soybean meal
Lactose-hydrolyzed milk
does not increase lead
absorption, but ordinary
milk does
Lactose in diet enhances
lead absorption compared
to glucose
Low zinc in diets
increases lead absorption
Low-zinc diet of mother
increases lead transfer
in utero and in maternal
mTlTE
Low zinc diet enhances
brain Pb levels
Low copper in diet
increases lead
absorption
Quarterman et al. (1978b)
Anders et al. (1982)
Bell and Spickett (1981)
Bushnell and DeLuca (1981)
Cerklewski and Forbes
(1976); El-Gazzar et al.
(1978)
Cerklewski (1979)
Bushnell and Levin (1983)
Klauder et al. (1973);
Klauder and Petering (1975)
-------
humans. Thus, the combined action of the two mineral nutrients is greater than the sum of
their individual effects.
Mykkanen and Wassermann (1981) observed that lead uptake in the intestine of the chick
occurs in two phases: a rapid uptake (within 5 min) followed by a rate-limiting slow transfer
of lead into blood. Conrad and Barton (1978) have observed a similar process in the rat.
Hence, either a saturation process occurs (i.e., carrier-mediated transport) or lead simply
precipitates in the lumen. In the former case, calcium interacts to saturate the carrier pro-
teins as Isolated by Barton et al. (1978a) or may precipitate lead in the lumen by initial
formation of calcium phosphate.
Quarterman et al. (1978a) observed that calcium supplementation of the diet above normal
also resulted In increased body retention of lead in the rat. Because both deficiency (Barton
et al., 1978a) and excess in calcium intake enhance retention, two sites of influence on
retention are suggested. Goyer (1978) has suggested that body retention of lead in calcium
deficiency, I.e., reduced excretion rate, may result from renal Impairment, while Quarterman
et al. (1978a) suggest that excess calcium suppresses calcium resorption from bone, hence also
reducing lead release.
10.5.2.2 Interactions of Lead with Iron. Mahaffey-S1x and Goyer (1972) reported that Iron-
deficient rats had increased tissue levels of lead and manifested greater toxicity compared
with control animals. This uptake change was seen with but minor alterations in hematocrit,
indicating a primary change 1n lead absorption over the time of the study. Barton et al.
(1978b) found that dietary restriction of iron, using 210Pb and everted sac preparations in
the rat, led to enhanced lead absorption, whereas iron loading suppressed the extent of lead
uptake, using normal intake levels of iron. This suppression suggests receptor-binding com-
petition at a common site, consistent with the isolation by these workers of two iron-binding
mucosa fractions. While the Iron level of diet affects lead absorption, the effect of changes
1n lead content in the gut on Iron absorption 1s not clear. Barton et al. (1978b) and Dobbins
et al. (1978) observed no effect of lead 1n the gut on iron absorption in the rat, while
Flanagan et al. (1979) reported that lead reduced iron absorption in mice.
In the mouse, Hamilton (1978) found that body retention of 203Pb was unaffected by iron
deffciency, using intraperitoneal administration of the label, while gastric Intubation did
lead to Increased retention. Animals with adequate Iron showed no changes in lead retention
at intubation levels of 0.01 to 10 nM. Cerklewskl (1980) observed that lead transfer both jhi
utero and in milk to nursing rats was enhanced compared with controls when dams were maintain-
ed from gestation through lactation on low-Iron diets.
10.5.2.3 Lead Interactions with Phosphate. The early studies of Shelling (1932), Grant et
al. (1938), and Sobel et al. (1940) documented that dietary phosphate Influenced the extent of
lead toxlcity and tissue retention of lead 1n animals. Low levels of phosphate enhanced these
10-48
-------
parameters, while excess Intake retarded the effects. More recently, Barltrop and Khoo (1975)
reported that reduced phosphate increased the uptake of 203Pb approximately 2.7-fold compared
with controls. Quarterman and Morrison (1975) found that low dietary phosphate enhanced lead
retention in rats but had no effect on skeletal lead mobilization, nor was injected lead label
affected by such restriction. In a related study, Quarterman et al. (1978a) found that dou-
bling the nutrient over normal levels resulted in lowering lead absorption by approximately
half. Barton and Conrad (1981) found that reduced dietary phosphorus increased the retention
of labeled lead and deposition in bone, in contrast to the results of Quarterman and Morrison
(1975). Increasing the intraluminal level of phosphorus reduced lead absorption, possibly by
increasing Intraluminal precipitation of lead as the mixed lead/calcium phosphate. Smith et
al. (1978) reported that lead uptake occurs at the same site as phosphate, suggesting that
lead absorption may be more related to phosphate than calcium transport.
10.5.2.4 Interactions of Lead with Vitamin D. Several studies had earlier indicated that a
positive relationship might exist between dietary vitamin D and lead uptake, resulting in
either greater manifestations of lead toxicity or a greater extent of lead uptake (Sobel et
al., 1938, 1940). Using the everted sac technique and testing with 210Pb, Smith et al. (1978)
observed that increasing levels of intubated vitamin D in the rat resulted in increased ab-
sorption of the label, with uptake occurring at the distal end of the rat duodenum, the site
of phosphorus uptake and greatest stimulation by the vitamin. Barton et al. (1980) used 21ffPb
to monitor lead absorption in the rat under conditions of normal, deficient, and excess
amounts of dietary vitamin D. Lead absorption is increased with either low or excess vitamin
D. This increased absorption apparently occurs as a result of increased retention time of
fecal mass containing the lead due to alteration of intestinal motility rather than as a re-
sult of direct enhancement of mucosal uptake rate. Hart and Smith (1981) reported that vita-
min D repletion of diet enhanced lead absorption (210Pb) in the rat, while also enhancing
femur and kidney lead uptake when the label was injected.
10.5.2.5 Interactions of Lead with Lipids. Barltrop and Khoo (1975) observed that varying
the lipid (corn oil) content of rat diet from 5 up to 40 percent resulted in an increase of
lead in blood 13.6-fold higher than the normal level. Concomitant increases were observed in
lead levels in kidney, femur, and carcass. Reduction of dietary lipid below the 5 percent
control figure did not affect the lead-absorption rate. As an extension of this earlier work,
Barltrop (1982) has noted that the chemical composition of the lipid is a significant factor
in affecting lead absorption. Study of triglycerides of saturated and unsaturated fatty acids
showed that polyunsaturated trilinolein increased lead absorption by 80 percent in rats, when
given as 5- or 10-percent loadings in diet, compared with monounsaturated triolein or any of
the saturates in the series tricaproin to tristearin.
10-49
-------
10.5.2.6 Lead Interaction with Protein. Quarter-man et al. (1978b) have drawn attention to
one of the inherent difficulties of measuring lead-protein interactions, i.e., the effect of
protein on both growth and the toxicokinetic parameters of lead. Der et al. (1974) found that
reduction of dietary protein, from 20 to 4 percent, led to increased uptake of lead in rat
tissues, but the approximately sixfold reduction in body weight over the interval of the study
makes it difficult to draw any firm conclusions. Barltrop and Khoo (1975) found that 203Pb
uptake by rat tissue could be enhanced with either suboptimal or excess levels of protein in
diet. Quarterman et al. (1978b) reported that retention of labeled lead in rats maintained on
a synthetic diet containing approximately 7 percent protein was either unaffected or reduced
compared with controls, depending on tissues taken for study.
Not only levels of protein but also the type of protein appears to affect tissue lead
levels. Anders et al. (1982) found that rats maintained on either of two synthetic diets
varying only by having casein or soybean meal as the protein source showed significantly
higher lead levels in the casein group.
10.5.2.7 Interactions of Lead with Milk Components. For many years, milk was recommended as
a counteractant for lead poisoning among lead workers (Stephens and Waldron, 1975). More recen
t data, however, pose a mixed picture. Kello and Kostial (1973) found that rats maintained on
milk diets absorbed a greater amount of 203Pb than those fed commercial rat chow. This
phenomenon was ascribed to relatively lower levels of certain nutrients in milk compared with
the rat chow. These observations were confirmed by Bell and Spickett (1981), who also
observed that lactose-hydrolyzed milk was less effective than the ordinary form in promoting
lead absorption, suggesting that lactose may be the enhancing agent. Bushnell and DeLuca
(1981) demonstrated that lactose significantly increased 210Pb absorption and tissue retention
by weanling rats when given in high doses by intubation. However, lactose levels close to
usual dietary content actually have an inhibiting effect on lead absorption (Bushnell and
DeLuca, 1983). In human studies, moreover, milk consumption is inversely related to blood
lead levels, suggesting a net protective effect (Johnson and Tenuta, 1979; Brunekreef et al.,
1983).
10.5.2.8 Lead Interactions with Zinc and Copper. The studies of Cerklewski and Forbes (1976)
and El-Gazzar et al. (1978) documented that zinc-deficient diets promote lead absorption in
the rat, while repletion with zinc reduces lead uptake. The interaction continues within the
body, particularly with respect to ALA-D activity (see Chapter 12, Section 12.3.1.2). In a
study of zinc-lead interactions in female rats during gestation and lactation, Cerklewski
(1979) observed that zinc-deficient diets resulted in more transfer of lead through milk to
the pups as well as reduced litter body weights. Bushnell and Levin (1983) have shown that
10-50
-------
rats fed a 1ow-z1nc diet (2.0 ppm) containing lead at levels of 10 or 100 ppm had signi-
ficantly higher retention of lead 1n brain and calvarium compared to those fed a diet with
20 ppm zinc. Victery and coworkers (1981) evaluated the acute effects of lead on the behavior
of renal and plasma zinc in the dog. They found that lead enhanced urinary zinc excretion and
was related to both increased ultrafllterable plasma zinc and a change in renal tubular zinc
transport.
Klauder et al. (1973) reported that low dietary copper enhanced lead absorption in rats
fed a high-lead diet (5000 ppm). These observations were confirmed by Klauder and Petering
(1975) at a level of 500 ppm lead 1n diet. The same researchers subsequently observed that
reduced copper enhanced the hematological effects of lead (Klauder and Petering, 1977), and
that both copper and iron deficiencies must be corrected to restore hemoglobin levels to
normal.
10.6 INTERRELATIONSHIPS OF LEAD EXPOSURE, EXPOSURE INDICATORS, AND TISSUE LEAD BURDENS
Information presented so far in this chapter sets forth the quantitative and qualitative
aspects of lead toxicokinetics, including the compartmental modeling of lead distribution Ui
vivo, and leads up to the critical issue of the various interrelationships of lead toxico-
kinetics to lead exposure, toxicant levels 1n indicators of such exposure, and exposure-target
tissue burdens of lead.
Chapter 11 (Sections 11.4, 11.5, 11.6) discusses the various experimental and epidemi-
ological studies relating the relative impact of various routes of lead exposure on blood lead
levels 1n human subjects, and includes a description of mathematical models for such relation-
ships. In these sections, the basic question is: what 1s the mathematical relationship of
lead in air, food, water, etc., to lead 1n blood? This question is descriptive and does not
address the biological basis of the observed relationships. Nor does it consider the impli-
cations for adverse health risks in the sequence leading from external lead exposure to lead
1n some physiological indicator to lead in target tissues.
For purposes of discussion, this section separately considers (1) the temporal character-
istics of physiological indicators of lead exposure, (2) the biological aspects of the rela-
tionship of external exposure to internal indicators of exposure, and (3) Internal indicator-
tissue lead relationships, Including both steady-state lead exposure and abrupt changes in
lead exposure. The relationship of internal indicators of body lead, such as blood lead, to
biological indicators such as EP or ALA-U is discussed 1n Chapter 13.
10-51
-------
10.6.1 Temporal Characteristics of Internal Indicators of Lead Exposure
The biological half-life for blood lead or the nonretained fraction of body lead is
generally assumed to be rather short, although it in fact depends upon the mobile lead body
burden (O'Flaherty et al., 1982; also see Sections 10.3 and 10.4). Nevertheless, a given
blood or urine lead value reflects rather recent exposure compared to tooth or bone lead
values. In cases where lead exposure can be reliably assumed to have occurred at a given
level, a blood lead value is more useful than in cases where some intermittent, high level of
exposure may have occurred. The former most often occurs with occupational exposure, while
the latter is of particular relevance to young children.
Reports have appeared dealing with the stability of individuals' blood lead levels over
time under conditions of ambient exposure. David et al. (1982) followed 29 children, 4-12
years old, with monthly measurements and found the stability to be of a relatively high order
(Pearson correlation coefficients of 0.7-0.8). Rabinowitz et al. (1984) sampled more than 200
infants semi annually from birth to 2 years of age and found average changes of about 4 pg/dl.
Only 40 percent of these children tended to remain in their previous blood lead category
(quartile). Within this age range, however, there was a trend toward less fluctuation with
increasing age of the young child. Delves et al. (1984) followed 21 adults over 7-11 months
with multiple blood lead measurements and found little fluctuation over time (about 1 ug/dl or
less, on average). Hence, there appears to be Increasing stability with relatively constant
exposure as the individual Increases in age.
Accessible mineralizing tissue, such as shed teeth, extend the time frame for assessing
lead exposure from months to years (Section 10.3), since teeth accumulate lead up to the time
of shedding or extraction. Levels of lead in teeth increase with age in proportion to expo-
sure (Steenhout and Pourtois, 1981). Furthermore, tooth lead levels are correlated with blood
lead levels in humans (Shapiro et al., 1978) and animals (Kaplan et al., 1980). The technique
of Fremlln and Edmonds (1980), employing mlcroautoradlography of irradiated teeth, permits the
identification of dentine zones high in lead content, thus allowing the disclosure of past
periods of abrupt increases in lead Intake.
While levels of lead 1n shed teeth are more valuable than blood lead levels in assessing
exposure at more remote time points, such information is retrospective in nature and would not
be of use in monitoring current exposure. In this case, serial blood lead measurements must
be employed. With the development of methodology for j_n situ measurement of tooth lead in
children (described in Chapter 9), serial ijn situ tooth analysis in tandem with serial blood
lead determination would provide comparative data for determining both time-concordant blood/
tooth lead relationships as well as which measure is the better indicator of ongoing exposure.
Given the limitations of an indicator such as blood lead in reflecting lead uptake in target
organs, as discussed below, the rate of accumulation of lead in teeth measured j_n situ may
10-52
-------
well be a better Index of ongoing tissue lead uptake. This aspect merits further study, espe-
cially since Shapiro et al. (1978) were able to demonstrate the feasibility of using |n situ
tooth lead analysis 1n a large group of children screened for lead exposure.
10.6.2 Biological Aspects of External Exposure/Internal Indicator Relationships
Information provided 1n Chapter 11 as well as the critiques of Hammond et al. (1981) and
Brunekreef (1984) Indicate that the relationship of lead levels 1n air, food, and water to
lead levels 1n blood 1s curvilinear, with the result that as "baseline" blood lead rises
(I.e., as one moves up the curve), the relative change 1n the dependent variable, blood lead,
per unit change of lead 1n some Intake medium (such as air) becomes smaller. Conversely, as
one proceeds down the curve with reduction 1n "baseline" lead, the corresponding change 1n
blood lead becomes larger. One assumption 1n this "single medium" approach 1s that the base-
line 1s not Integrally related to the level of lead 1n the particular medium being studied.
This assumption 1s not necessarily appropriate for air versus food lead, nor, 1n the case of
young children, for air lead versus total oral Intake of the element. However, 1t should be
noted that Hammond et al. (1981) assigned virtually all of the body compartment lead to the
blood, giving blood lead levels 1n their modeling scheme that were too high. The authors
recognized this and later offered a qualification (Hammond et al., 1982).
Hammond et al. (1981) have also noted that the shape of the blood lead curves seen in
human subjects 1s similar to that discernible 1n certain experimental animal studies with
dogs, rats, and rabbits (Azar et al., 1973; Prpld-Majic* et al., 1973). Similarly, Klmmel et
al. (1980), after exposing adult female rats to lead at four levels 1n drinking water for 6-7
weeks, found values of blood lead that showed a curvilinear relationship to the dose levels.
Over the dosing range of 5 to 250 ppm 1n water, the blood lead range was 8.5 to 31 ug/dl. In
a related study (Grant et al., 1980) rats were exposed to lead JUi utero, through weaning, and
up to 9 months of age at the dosing range used 1n the Klmmel et al. study (0.5 to 250 ppm 1n
the dams' drinking water until weaning of pups, then the same levels 1n the weanlings' drink-
Ing water). These animals showed a blood lead range of 5 to 67 ug/dl. One may assume that 1n
all of the above studies the lead 1n the various dosing groups was near or at equilibrium
within the various body compartments.
The biological basis of the curvilinear relationship of blood lead to lead Intake, across
a broad range of blood lead values, may result from a number of factors. In lead workers, as
a specific case, Increasing workplace air lead level 1s associated with an Increased particle
aggregation rate leading to a lowering of the effective fraction of respirable, submicrometer
particles, as suggested by Chamberlain (1983). In studies with human volunteers, there
appears to be no change 1n respiratory absorption rate at lung lead burdens up to 450 ug
(Chamberlain et al., 1978). It was noted earlier that oral lead intake up to 400 ug in adults
10-53
-------
1s associated with unaltered absorption rate. However, animal data relevant to this question
Indicate that dietary levels between 10 and 100 ppm lead are associated with a decreased ab-
sorption rate (Bushnell and DeLuca, 1983). If these data were applied directly to humans, a
daily intake rate of 20-200 mg lead would be required to produce a similar decrease.
The curvilinear blood lead/diet lead relationship may or may not be Independent of GI ab-
sorption rate. The experimental animal studies of Prpic-Majic et al. (1973) indicated a cur-
vilinear relationship of blood lead to dose of lead when the toxicant was administered by in-
jection to rabbits. On the other hand, injection of higher doses into rats does show a linear
relationship (Aungst et al., 1981).
The data of DeSilva (1981), Manton and Malloy (1983), and Manton and Cook (1984) all sug-
gest that the increasingly greater fraction of lead in plasma as blood lead increases may be
significant (see Section 10.3.1). This Increase of lead in plasma would Indicate a relatively
greater movement of lead from plasma to tissues and a higher excretion rate, both of which
serve to modulate the rate of rise of the whole blood lead with increasing circulating lead.
These results are consistent with the report of Chamberlain (1983) showing an apparent in-
creased urinary excretion rate of lead with rising blood lead. They are also 1n accord with
the observations that tissue lead burdens show a better proportionality to exposure level than
does blood lead burden (see Section 10.3.1). Since an Increased movement of plasma lead to
tissues with increasing blood lead burden would also include deposition in bone, the curvilin-
ear relationship of chelatable lead to blood lead may also be influenced by the plasma/blood
relationship.
10.6.3 Internal Indicator/Tissue Lead Relationships
In living human subjects, to determine tissue lead burdens directly (or relate these
levels to adverse effects associated with target tissue) as a function of lead Intake is not
possible. Instead, measurement of lead in an accessible indicator such as blood, along with
determination of some biological indicator of impairment (e.g., ALA-U or EP), is used.
Evidence continues to accumulate in both the clinical and experimental animal literature
that the use of blood lead as an indicator can have limitations in reflecting both the amounts
of lead 1n target tissues and the temporal changes in tissue lead with changes in exposure.
Perhaps the best example of the problem Is the relationship of blood lead to chelatable lead
(see Section 10.3.3). Currently, measurement of the plumburesis associated with challenge by
a single dose of a chelating agent such as CaNa2EDTA 1s considered the best measure of the mo-
bile, potentially toxic fraction of body lead in children and adults (Vitale et al., 1975;
Wedeen et al., 1975; Chlsolm et al., 1976; U.S. Centers for Disease Control, 1978; Chlsolm and
Barltrop, 1979; Hansen et al., 1981).
10-54
-------
Chlsolm et al. (1976) have documented that the relationship of blood lead to chelatable
lead 1s curvilinear, such that a given Incremental Increase 1n blood lead is associated with
an increasingly larger Increment of mobllizable lead. The problems associated with this cur-
vilinear relationship in exposure assessment are typified by the recent reports of Saenger et
al. (1982) and PiomelH et al. (1984) concerning children and Hansen et al. (1981) concerning
adult lead workers. Saenger et al. (1982) noted that significant percentages of children
having mild to moderate lead exposure, as discernible by blood lead and EP measurements, had
urinary outputs of lead upon challenge with CaNa2EDTA that qualified them for chelatlon thera-
py under CDC guidelines. Similar data were obtained for 210 children evaluated 1n four medi-
cal centers (Piomelli et al., 1984). In adult workers, Hansen et al. (1981) observed that a
sizable fraction of subjects with only modest elevations in blood lead levels upon EDTA chal-
lenge excreted lead in amounts significantly exceeding the upper end of normal. This discrep-
ancy occurred at blood lead levels of 35 ug/dl and above.
The biological basis for the nonlinearlty of the relationship between blood lead and che-
latable lead appears, in major part, to be the existence of a sizable pool of lead in bone
that is labile to chelatlon. Evidence pointing to this explanation was summarized 1n Section
10.3.3. The question of how long any lead 1n this compartment of bone remains labile to che-
latlon has been addressed by several Investigators in studies of both children and adults.
The question is relevant to the issue of the usefulness of EOTA challenge in assessing evi-
dence for past lead exposure.
Chlsolm et al. (1976) found that a group (N = 55) of adolescent subjects 12-22 years old,
who had a clinical history of lead poisoning as young children and whose mean blood lead was
22.1 M9/d1 at tne t1me °f study, yielded chelatable lead values that placed them on the same
regression curve as a second group of young children with current elevations of blood lead.
The results with the adolescent subjects did not provide evidence that they might have had a
past history of lead poisoning. According to the authors, this failure to detect prior expo-
sure suggests that chelatable lead at the time of excessive exposure was not retained in a
pool that remained labile to chelatlon years later, but underwent subsequent excretion or
transfer to the inert compartment of bone. One problem with drawing conclusions from this
study is that all of the adolescents apparently had one or more courses of chelatlon therapy
and were removed to housing where re-exposure would be minimal as part of their clinical
management after lead poisoning was diagnosed. One must assume that chelatlon therapy removed
a significant portion of the mobile lead burden and that placement in lead-free housing re-
duced the extent of any further exposure. The obvious question is how this group of adoles-
cents would compare with subjects who had excessive chronic lead exposure as young children
but who did not require or receive chelatlon therapy.
10-55
-------
Former lead workers challenged with EDTA show chelatable lead values that are signifi-
cantly above normal years after workplace exposure ceases (e.g., Alesslo et al., 1976;
Prerovska and Telsinger, 1970). In the case of former lead workers, blood lead also remains
elevated, suggesting that the mobile lead pool 1n bone remains 1n equilibrium with lead 1n
blood.
The closer correspondence of chelatable lead with actual tissue lead burdens, compared to
blood lead, 1s also reflected In a better correlation of this parameter with such biological
Indicators of Impairment as EP, although this correlation 1s seen only 1n adults. Similarly,
Alesslo et al. (1976) found that EP In former lead workers was more significantly correlated
with chelatable lead than with blood lead.
Consideration of both the Intake versus blood lead and the blood lead versus chelatable
lead curves leads to the prediction that the level of lead exposure per se 1s more closely re-
lated to tissue lead burden than Is blood lead. This appears to be the case 1n experimental
animals. Azar et al. (1973) and Grant et al. (1980) reported that levels of lead in brain,
kidney, and femur followed more of a direct proportionality with the level of dosing than with
blood lead. These observations may relate to the fact that plasma lead rises proportionately
faster than whole blood lead.
Finally, there 1s the question of how adequately an Internal indicator such as blood lead
reflects changes in tissue burden when exposure changes abruptly. In the study of Bjbrklund
et al. (1981), lead levels in both blood and brain were monitored over a 6-week period in rats
exposed to lead through their drinking water. Blood lead rose rapidly by day 1, during which
time brain lead content was only slightly elevated. After day 1, the rate of Increase In
blood lead began to taper off, while brain lead began to rise in a nearly linear fashion up to
the end of the experiment. From day 7 to 21, blood lead increased from approximately 45 to 55
pg/dl, while brain lead increased approximately twofold.
Abrupt reduction 1n exposure similarly appears to be associated with a more rapid re-
sponse 1n blood than in soft tissues, particularly brain. Goldstein and Diamond (1974)
reported that termination of Intravenous administration of lead to 30-day-old rats resulted in
a sevenfold drop of lead in blood by day 7. At the same time, brain lead levels did not de-
crease significantly. A similar difference In brain and blood response was reported by
Momdlovlc and Kostial (1974).
In all of the above studies, blood lead was of limited value in reflecting changes in the
brain, which 1s the significant target organ for lead exposure 1n children. With abrupt in-
creases 1n exposure level, the problem concerns a much more rapid approach to steady state In
blood than 1n brain. Conversely, the biological half-time for lead clearance from blood 1n
the young rats of both the Goldstein and Diamond (1974) and Momcilovlc and Kostial (1974)
studies was much less than it appeared to be for lead movement from brain.
10-56
-------
Despite the limitations In Indexing tissue burden and exposure changes, blood lead
remains the one readily accessible measure that can demonstrate 1n a relative way the rela-
tionship of various effects to Increases 1n exposure.
10.7 METABOLISM OF LEAD ALKYLS
The lower alkyl lead compounds used as gasoline additives, tetraethyl lead (TEL) and
tetramethyl lead (TML), are much more neurotoxlc on an equivalent dose basis than inorganic
lead. These agents are emitted 1n auto exhaust, and their rate of environmental degradation
depends on such factors as sunlight, temperature, and ozone levels. There is also some con-
cern that organolead compounds may result from blomethylatlon in the environment (see Chapter
6). Finally, a problem arises with the practice among children of sniffing leaded gasoline.
The available information dealing with metabolism of lead alkyls is derived mainly from ex-
perimental animal studies, studies of workers exposed to the agents, and cases of lead alkyl
poisoning.
10.7.1 Absorption of Lead Alkyls in Humans and Animals
The respiratory Intake and absorption of TEL and TML in the vapor state was investigated
by Heard et al. (1979), who used human volunteers inhaling 203Pb-labeled TEL and TML. Initial
lung deposition rates were 37 and 51 percent for TEL and TML, respectively. Of these amounts,
40 percent of TEL was lost by exhalation within 48 hr, while the corresponding figure for TML
within 48 hr was 20 percent. The remaining fraction was absorbed. The effect of gasoline
vapor on these parameters was not investigated. In an earlier study Mortensen (1942) reported
that adult rats inhaling TEL labeled with 203Pb (0.07-7.00 mg TEL/1) absorbed 16-23 percent of
the fraction reaching the alveoli. Gasoline vapor had no effect on the absorption rates.
Respiratory absorption of organolead bound to partlculate matter has not been specifi-
cally studied as such. According to Harrison and Laxen (1978), neither TEL nor TML adheres to
partlculate matter to any significant extent, but the toxicologlcally equivalent trialkyl
derivatives, formed from photolytlc dissociation or ozonolysis in the atmosphere, may do so.
10.7.1.1 Gastrointestinal Absorption. Information on the rate of absorption of lead alkyls
through the GI tract is not available in the literature. Given the level of gastric acidity
(pH 1.0) in humans, one would expect TML and TEL to be rapidly converted to the corresponding
trialkyl forms, which are comparatively more stable (Bade and Huber, 1970). Given the simi-
larity of the chemical and biochemical behavior of trialkyl leads to their Group IV analogs,
the trialkyltlns, the report of Barnes and Stoner (1958) that triethyltin is quantitatively
absorbed from the GI tract Indicates that trlethyl and trimethyl lead would be extensively ab-
sorbed via this route.
10-57
-------
10.7.1.2 Percutaneous Absorption of Lead Alkyls. In contrast to Inorganic lead salts, both
TEL and TML are rapidly and extensively absorbed through the skin 1n rabbits and rats (Kehoe
and Thamann, 1931; Laug and Kunze, 1948), and lethal effects can be rapidly Induced 1n these
animals by merely exposing the skin. Laug and Kunze (1948) observed that systemic uptake of
TEL was still 6.5 percent even after most of the TEL had evaporated from the skin surface.
The rate of passage of TML was somewhat slower than that of TEL 1n the study of Davis et al.
(1963). Absorption of either agent was retarded somewhat when applied in gasoline.
10.7.2 BiotransformatJon and Tissue Distribution of Lead Alkyls
To understand the _1jn vivo fate of lead alkyls, one must first discuss the blotransforma-
tion processes of lead alkyls known to occur 1n mammalian systems. Tetraethyl and tetramethyl
lead both undergo oxidatlve dealkylatlon 1n mammals to the trlethyl or trimethyl metabolites,
which are now accepted as the actual toxic forms of these alkyls.
Studies of the biochemical mechanisms for these transformations, as noted by Kimmel et
al. (1977), Indicate a dealkylation mediated by a P-450 dependent mono-oxygenase system in
liver microsomes, with Intermediate hydroxylation. In addition to rats (Cremer, 1959; Stevens
et al., 1960; Bolanowska, 1968), mice (Hayakawa, 1972), and rabbits (Bolanowska and
Garczyriski, 1968), this transformation also occurs In humans accidentally poisoned with TEL
(Bolanowska et al., 1967) or workers chronically exposed to TEL (Adamlak-Ziemba and
Bolanowska, 1970).
The rate of hepatic oxidatlve de-ethylation of TEL in mammals appears to be rather rapid;
Cremer (1959) reported a maximum hourly conversion rate of approximately 200 ug TEL/g rat
liver. In comparison with TEL, TML may undergo transformation at either a slower rate (in
rats) or more rapidly (1n mice), according to Cremer and Callaway (1961) and Hayakawa (1972).
Other transformation steps involve conversion of triethyl lead to the dlethyl form, the
process appearing to be species-dependent. Bolanowska (1968) did not report the formation of
dlethyl lead in rats, while significant amounts of it are present 1n the urine of rabbits
(Aral et al., 1981) and humans (CMesura, 1970). Inorganic lead is formed in various species
treated with TEL, whether the TEL arises from degradation of the diethyl lead metabolite or
from some other direct process (Bolanowska, 1968). Degradation appears to occur in rats, since
little or no diethyl lead is found, whereas significant amounts of Inorganic lead are present.
Formation of Inorganic lead with lead alkyl exposure may account for the hematological effects
seen in humans chronically exposed to the lead alkyls (see Chapter 12, Section 12.3), includ-
ing children who inhale leaded gasoline vapor.
Partitioning of trlethyl or trimethyl lead, the corresponding neurotoxlc metabolites of
TEL and TML, between the erythrocyte and plasma appears to be species-dependent. Byington et
al. (1980) studied the partitioning of triethyl lead between cells and plasma In vitro using
10-58
-------
washed human and rat erythrocytes and found that human cells had a very low affinity for the
alkyl lead while rat cells bound the alkyl lead 1n the globln moiety at a ratio of three mole-
cules per hemoglobin tetramer. Similarly, injected trlethyl lead was found to be associated
with whole blood levels approximately 10-fold greater than 1n rat plasma. The available lit-
erature on TEL poisoning 1n humans concurs; significant plasma lead values have been routinely
reported (Boeckx et al., 1977; Goldings and Stewart, 1982). These data Indicate that the rat
1s a poor model for studying the adverse effects of lead alkyls in human subjects.
The biological half-life in blood for the lead alkyls depends on whether clearance of the
tetraalkyl or trialkyl forms is being observed. Heard et al. (1979) found that 203Pb-labeled
TML and TEL inhaled by human volunteers was rapidly cleared from the blood (by 10 hr), fol-
lowed by a reappearance of lead. The fraction of lead 1n plasma initially was quite high,
approximately 0.7, suggesting the presence of tetra/trialkyl lead. However, the subsequent
rise in blood lead showed all of it essentially present in the cell, which would indicate
inorganic or possibly diethyl lead. Trlethyl lead in rabbits was more rapidly cleared from
the blood (3-5 days) than was the trimethyl form (15 days) when administered as such
(Hayakawa, 1972).
Tissue distribution of lead in both humans and animals exposed to TEL and TML primarily
involves the trialkyl metabolites. Levels are highest in liver, followed by kidney, then
brain (Bolanowska et al., 1967; Grandjean and Nielsen, 1979). Nielsen et al. (1978) observed
measurable amounts of trialkyl lead in samples of brain tissue from subjects with no known oc-
cupational exposure.
The available studies on tissue retention of trlethyl or trimethyl lead provide variable
findings. Bolanowska (1968) noted that tissue levels of trlethyl lead 1n rats were almost
constant for 16 days after a single Injection of TEL. Hayakawa (1972) found that the half-
life of trlethyl lead 1n brain was 7-8 days for rats. The half-time for trimethyl lead was
much longer. In humans, Yamamura et al. (1975) reported two tissue compartments for trlethyl
lead having half-lives of 35 and 100 days (Yamamura et al., 1975).
10.7.3 Excretion of Lead Alkyls
The renal tract is the main route of lead excretion in various species exposed to lead
alkyls (Grandjean and Nielsen, 1979). The chemical forms of lead in urine suggest that the
differing amounts of the various forms are species-dependent. Aral et al. (1981) found that
rabbits given TEL parenterally excreted lead primarily in the form of diethyl lead (69 per-
cent) and inorganic lead (27 percent), trlethyl lead accounting for only 4 percent.
Bolanowska and Garczyfiski (1968) found that trlethyl lead levels were somewhat higher in the
urine of rats than in that of rabbits. In humans, Chiesura (1970) found that trialkyl lead
was never greater than 9 percent of total lead content in workers with heavy TEL exposure.
10-59
-------
Adam1ak-Z1emba and Bolanowska (1970) reported similar data; the fraction of triethyl lead in
the urine was approximately 10 percent of total lead.
The urinary rates of lead excretion in human subjects with known levels of TEL exposure
were also reported by Adamiak-Ziemba and Bolanowska (1970). In workers involved with the
blending and testing of leaded gasoline, workplace air levels of lead (as TEL) ranged from
0.037 to 0.289 mg/m3 and the corresponding urine lead levels ranged from 14 to 49 jjg/1,
of which approximately 10 percent was triethyl lead.
10.8 SUMMARY
Toxicokinetic parameters of lead absorption, distribution, retention, and excretion rela-
ting external environmental lead exposure to various adverse effects have been discussed in
this chapter. Also considered were various influences on these parameters, e.g., nutritional
status, age, and stage of development. A number of specific issues in lead metabolism by ani-
mals and humans were addressed, including:
1. How does the developing organism from gestation to maturity differ from the adult in
toxicokinetic response to lead intake?
2. What do these differences in lead metabolism portend for relative risk for adverse
effects?
3. What are the factors that significantly change the toxicokinetic parameters in ways
relevant to assessing health risk?
4. How do the various interrelationships among body compartments for lead translate to
assessment of internal exposure and changes in internal exposure?
10.8.1 Lead Absorption in Humans and Animals
The amounts of lead entering the bloodstream via various routes of absorption are influ-
enced not only by the levels of the element in a given medium but also by various physical and
chemical parameters and specific host factors, such as age and nutritional status.
10.8.1.1 Respiratory Absorption of Lead. The movement of lead from ambient air to the blood-
stream is a two-part process: deposition of some fraction of inhaled air lead in the deeper
part of the respiratory tract and absorption of the deposited fraction. For adult humans, the
deposition rate of particulate airborne lead as likely encountered by the general population
is around 30-50 percent, with these rates being modified by such factors as particle size and
ventilation rates. All of the lead deposited in the lower respiratory tract appears to be ab-
sorbed, so that the overall absorption rate is governed by the deposition rate, i.e., approxi-
mately 30-50 percent. Autopsy results showing no lead accumulation in the lung indicate total
absorption of deposited lead.
10-60
-------
All of the available data for lead uptake via the respiratory tract 1n humans have been
obtained with adults. Respiratory uptake of lead 1n children, while not fully quantifiable,
appears to be comparatively greater on a body-weight basis. A second factor Influencing the
relative deposition rate 1n children 1s airway dimensions. One report has estimated that the
10-year-old child has a deposition rate 1.6- to 2.7-fold higher than the adult on a weight
basis.
The chemical form of the lead compound Inhaled does not appear to be a major determinant
of the extent of alveolar absorption of lead. While experimental animal data for quantitative
assessment of lead deposition and absorption for the lung and upper respiratory tract are
limited, available Information from the rat, rabbit, dog, and nonhuman primate support the
findings that respired lead 1n humans 1s extensively and rapidly absorbed. Over the range of
air lead encountered by the general population, absorption rate does not appear to depend on
air lead level.
10.8.1.2 Gastrointestinal Absorption of Lead. Gastrointestinal (GI) absorption of lead
mainly Involves lead uptake from food and beverages as well as lead deposited 1n the upper
respiratory tract and eventually swallowed. It also Includes 1ngest1on of non-food material,
primarily 1n children via normal mouthing activity and pica. Two Issues of concern with lead
uptake from the gut are the comparative rates of such absorption 1n developing versus adult
organisms, Including humans, and how the relative bloavallability of lead affects such uptake.
By use of metabolic balance and 1sotop1c (radlolsotope or stable Isotope) studies, var-
ious laboratories have provided estimates of lead absorption 1n the human adult on the order
of 10-15 percent. This rate can be significantly Increased under fasting conditions to 45
percent, compared to lead Ingested with food. The latter figure also suggests that beverage
lead 1s absorbed to a greater degree since much beverage 1ngest1on occurs between meals.
The relationship of the chemical/biochemical form of lead 1n the gut to absorption rate
has been studied, although Interpretation 1s complicated by the relatively small amounts given
and the presence of various components 1n food already present 1n the gut. In general, how-
ever, chemical forms of lead and their Incorporation Into biological matrices seem to have a
minimal Impact on lead absorption in the human gut. Several studies have focused on the ques-
tion of differences 1n GI absorption rates for lead between children and adults. Such rates
for children are considerably higher than for adults: 10-15 percent for adults versus approx-
imately 50 percent for children. Available data for the absorption of lead from nonfood
Hems such as dust and dirt on hands are limited, but one study has estimated a figure of 30
percent. For paint chips, a value of about 17 percent has been estimated.
Experimental animal studies show that, like humans, the adult animal absorbs much less
lead from the gut than the developing animal. Adult rats maintained on ordinary rat chow ab-
sorb 1 percent or less of the dietary lead. Various animal species studies make 1t clear that
10-61
-------
the newborn absorbs a much greater amount of lead than the adult, supporting studies showing
this age dependency in humans. Compared to an absorption rate of about 1 percent in adult
rats, the rat pup has a rate 40-50 times greater. Part, but not most, of the difference can
be ascribed to a difference in dietary composition. In nonhuman primates, infant monkeys ab-
sorb 65-85 percent of lead from the gut, compared to 4 percent for the adults.
The bioavailability of lead in the GI tract as a factor 1n its absorption has been the
focus of a number of experimental studies. These data show the following: (1) lead in a
number of forms is absorbed about equally, except for lead sulfide; (2) lead in dirt and dust
and in different chemical forms is absorbed at about the same rate as pure lead salts added to
a diet; (3) lead in paint chips undergoes significant uptake from the gut; and (4) in some
cases, physical size of particulate lead can affect the rate of GI absorption. In humans, GI
absorption rate of lead appears to be independent of quantity in the gut up to a level of at
least 400 ug. In animals, dietary levels between 10 and 100 ppm result in reduced absorption.
10.8.1.3 Percutaneous Absorption of Lead. Absorption of inorganic lead compounds through the
skin is of much less significance than absorption through respiratory and GI routes. In con-
trast, absorption through skin is far more significant than through other routes for the lead
alkyls (see Section 10.7.1.2). One recent study using human volunteers and 203Pb-labeled lead
acetate showed that under normal conditions, skin absorption of lead alkyls approached 0.06
percent.
10.8.1.4 Transplacental Transfer of Lead. Lead uptake by the human and animal fetus readily
occurs, such transfer going on by the 12th week of gestation in humans, and increasing
throughout fetal development. Cord blood contains significant amounts of lead, correlating
with, but somewhat lower than, maternal blood lead levels. Evidence for such transfer, be-
sides the measured lead content of cord blood, includes fetal tissue analyses and reduction in
maternal blood lead during pregnancy. There also appears to be a seasonal effect on the
fetus, summer-born children showing a trend to higher blood lead levels than those born in the
spring.
10.8.2 Distribution of Lead 1n Humans and Animals
In this subsection, the distributional characteristics of lead in various portions of the
body (blood, soft tissue, calcified tissue, and the "chelatable" or potentially toxic body
burden) are discussed as a function of such variables as exposure history and age.
10.8.2.1 Lead in Blood. More than 99 percent of blood lead is associated with the erythro-
cytes in humans under steady-state conditions, but it is the very small fraction transported
in plasma and extracellular fluid that provides lead to the various body organs. Most (^50
percent) erythrocyte lead is bound within the cell, primarily associated with hemoglobin (par-
ticularly HbA2), with approximately 5 percent bound to a 10,000-dalton fraction, 20 percent to
10-62
-------
a heavier molecule, and 25 percent to lower-weight species. Several studies with lead workers
and patients Indicate that the fraction of lead 1n plasma versus whole blood increases above
-v50-60 ug/dl blood lead.
Whole blood lead 1n dally equilibrium with other compartments In adult humans appears to
have a biological half-life of 25-28 days and comprises about 1.9 mg In total lead content,
based on Isotope studies. Other data from lead-exposed workers indicate that half-life
depends on mobile lead burden. Human blood lead responds rather quickly to abrupt changes in
exposure. With increased lead intake, blood lead achieves a new value in approximately 40-60
days, while a decrease 1n exposure may be associated with variable new blood values, depending
upon the exposure history. This dependence presumably reflects lead resorption from bone.
With age, furthermore, a moderate increase occurs In blood lead during adulthood. Levels of
lead in blood of children tend to show a peak at 2-3 years of age (probably caused by mouthing
activity), followed by a decline. In older children and adults, levels of lead are sex-
related, females showing lower levels than males even at comparable levels of exposure.
In plasma, lead is virtually all bound to albumin and only trace amounts to high-weight
globulins. Which binding form constitutes an "active" fraction for movement to tissues is
Impossible to state. The most recent studies of the erythrocyte/plasma relationship in humans
Indicate an equilibrium between these blood compartments, such that levels in plasma rise with
levels 1n whole blood 1n fixed proportion up to approximately 50-60 ug/dl, whereupon the
relationship becomes curvilinear.
10.8-2.2 Lead Levels in Tissues. Of necessity, various relationships of tissue lead to expo-
sure and toxidty 1n humans must generally be obtained from autopsy samples. Limitations on
these data include questions of how such samples represent lead behavior in the living popula-
tion, particularly with reference to prolonged Illness and disease states. The adequate char-
acterization of exposure for victims of fatal accidents is a problem, as is the fact that such
studies are cross-sectional in nature, with different age groups assumed to have had similar
exposure in the past.
10.8.2.2.1 Soft tissues. After age 20 most soft tissues (in contrast to bone) in humans do
not show age-related changes. Kidney cortex shows an increase in lead with age, which may be
associated with the formation of nuclear Inclusion bodies. Absence of lead accumulation in
most soft tissues results from a turnover rate for lead similar to that in blood.
Based on several autopsy studies, soft-tissue lead content for individuals not occupa-
tionally exposed is generally below 0.5 ug/g wet weight, with higher values for aorta and kid-
ney cortex. Brain tissue lead level is generally below 0.2 (jg/g wet weight with no change
with increasing age, although the cross-sectional nature of these data would make changes in
low brain lead levels difficult to discern. Autopsy data for both children and adults indi
cate that lead is selectively accumulated in the hippocampus, a finding that is also consis-
tent with the regional distribution in experimental animals.
10-63
-------
Comparisons of lead levels in soft-tissue autopsy samples from children with results from
adults Indicate that such values are lower In Infants than 1n older children, while children
aged 1-16 years had levels comparable to those for adult women. In one study, lead content of
brain regions did not materially differ for infants and older children compared to adults.
Complicating these data somewhat are changes 1n tissue mass with age, although such changes
are less than for the skeletal system.
Subcellular distribution of lead in soft tissue 1s not uniform. High amounts of lead are
sequestered in the mitochondria and nucleus of the cell. Nuclear accumulation is consistent
with the existence of lead-containing nuclear Inclusions in various species, and a large body
of data demonstrate the sensitivity of mitochondria to injury by lead.
10.8.2.2.2 Mineralizing tissue. Lead becomes localized and accumulates 1n human calcified
tissues, i.e., bones and teeth. This accumulation in humans begins with fetal development and
continues to approximately 60 years of age. The extent of lead accumulation 1n bone ranges up
to 200 mg in men ages 60-70 years, while in women lower values have been measured. Based upon
various studies, approximately 95 percent of total body lead 1s lodged 1n the bones of human
adults, with uptake distributed over trabecular and compact bone. In the human adult, bone
lead Is both the most Inert and the largest body pool, and accumulation can serve to maintain
elevated blood lead levels years after exposure, particularly occupational exposure, has
ended.
By comparison to human adults, only 73 percent of body lead is lodged 1n the bones of
children, which is consistent with other information that the skeletal system of children is
more metabollcally active than that of adults. Furthermore, bone tissue 1n children 1s less
dense than in adults. While the increase 1n bone lead level across childhood is modest, about
twofold 1f expressed as concentration, the total accumulation rate 1s actually 80-fold,
taking into account a 40-fold increase 1n skeletal mass. To the extent that some significant
fraction of total bone lead 1n children and adults Is relatively labile, in terms of health
risk for the whole organism it 1s more appropriate to consider the total accumulation rather
than just changes in concentration.
The traditional view that the skeletal system was a "total" sink for body lead (and by
implication a biological safety feature to permit significant exposure in Industrialized popu-
lations) never did agree with even older information on bone physiology, e.g., bone remodel-
ing. This view is now giving way to the idea that there are at least several bone compart-
ments for lead, with different mobility profiles. Bone lead, then, may be more of an Insid-
ious source of long-term internal exposure than a sink for the element. This aspect of the
issue is summarized more fully in the next section. Available information from studies of
uranium miners and human volunteers who ingested stable isotopes indicates that there is a
relatively inert bone compartment for lead, having a half-life of several decades, as well as
a rather labile compartment that permits an equilibrium between bone and tissue lead.
10-64
-------
Tooth lead also Increases with age at a rate proportional to exposure and roughly propor-
tional to blood lead 1n humans and experimental animals. Dentine lead is perhaps the most re-
sponsive component of teeth to lead exposure since 1t 1s laid down from the time of eruption
until shedding. This characteristic underlies the usefulness of dentine lead levels In asses-
sing long-term exposure.
10.8.2.2.3 Chelatable lead. Mobile lead in organs and systems is potentially more active
toxlcologically 1n terms of being available to biological sites of action. Hence, this frac-
tion of total body lead burden is a more significant predictor of imminent toxicity. In real-
ity, direct measurement of such a fraction in human subjects would not be possible. In this
regard, chelatable lead, measured as the extent of plumburesis 1n response to administration
of a chelating agent, specifically CaNa2EDTA, is now viewed as the most useful probe of undue
body burden in children and adults.
A quantitative description of the Inputs to the body lead fraction that is chelant-mobi-
Hzable is difficult to define fully, but 1t most likely includes a labile lead compartment
within bone as well as within soft tissues. Support for this view includes the following:
(1) the age-dependency of chelatable lead, but not lead in blood or soft tissues; (2) evidence
of removal of bone lead in chelatlon studies with experimental animals; (3) jn vitro studies
of lead mobilization 1n bone organ explants under closely defined conditions; (4) tracer-
modeling estimates in human subjects; and (5) the complex nonlinear relationship of blood lead
and lead Intake through various media. Data for children and adults showing a logarithmic
relationship of chelatable lead to blood lead and the phenomenon of "rebound" in blood lead
elevation after chelatlon therapy regimens (without obvious external re-exposure) offer
further support.
10.8.2.2.4 Animal studies. Animal studies have helped to sort out some of the relationships
of lead exposure to _1n vivo distribution of the element, particularly the impact of skeletal
lead on whole body retention. In rats, lead'admlnistration results in an initial increase of
lead levels in soft tissues, followed by loss of lead from soft tissue via excretion and
transfer to bone. Lead distribution appears to be relatively independent of dose. Other
studies have shown that lead loss from organs follows first-order kinetics except for loss
from bone, and that the skeletal system in rats and mice is the kinetically rate-11 mi ting step
in whole-body lead clearance.
The neonatal animal seems to retain proportionally higher levels of tissue lead compared
to the adult and manifests slow decay of brain lead levels while showing a significant decline
over time 1n other tissues. This decay appears to result from enhanced lead entry to the
brain because of a poorly developed brain barrier system as well as from enhanced body reten-
tion of lead by young animals.
10-65
-------
The effects of such changes as metabolic stress and nutritional status on body redistri-
bution of lead have been noted. Lactatlng mice, for example, are known to demonstrate tissue
redistribution of lead, specifically bone-lead resorption with subsequent transfer of both
lead and calcium from mother to pups.
10.8.3 Lead Excretion and Retention 1n Humans and Animals
10.8.3.1 Human Studies. Dietary lead 1n humans and animals that 1s not absorbed passes
through the GI tract and 1s eliminated with feces, as 1s the fraction of air lead that 1s
swallowed and not absorbed. Lead entering the bloodstream and not retained Is excreted
through the renal and GI tracts, the latter via biliary clearance. The amounts excreted
through these routes are a function of such factors as species, age, and exposure charac-
teristics.
Based upon the human metabolic balance data and isotope excretion findings of various in-
vestigators, short-term lead excretion in adult humans amounts to 50-60 percent of the ab-
sorbed fraction, with the balance moving primarily to bone and some fraction (approximately
half) of this stored amount eventually being excreted. This estimated overall retention
figure of 25 percent necessarily assumes that isotope clearance reflects that for body lead in
all compartments. The rapidly excreted fraction has a biological half-life of 20-25 days,
similar to that for lead removal from blood, based on isotope data. This similarity indicates
a steady rate of lead clearance from the body. In terms of partitioning of excreted lead
between urine and bile, one study indicates that the biliary clearance is about 50 percent
that of renal clearance.
Lead accumulates 1n the human body with age, mainly in bone, up to around 60 years of
age, when a decrease occurs with changes in intake as well as in bone mineral metabolism. As
noted earlier, the total amount of lead 1n long-term retention can approach 200 mg, and even
much higher in the case of occupational exposure. This rate corresponds to a lifetime average
retention rate of 9-10 pg Pb/day. Within shorter time frames, however, retention will vary
considerably because of such factors as development, disruption in the Individuals' equilib-
rium with lead intake, and the onset of such states as osteoporosis.
The age-dependency of lead retention/excretion in humans has not been well studied, but
most of the available Information indicates that children, particularly infants, retain a sig-
nificantly higher amount of lead than adults. While autopsy data indicate that pediatrk sub-
jects at isolated points in time actually have a lower fraction of body lead lodged in bone,
which probably relates to the less dense bones of children as well as high bone mineral turn-
over, a full understanding of longer-term retention over childhood must consider the exponen-
tial growth rate occurring in children's skeletal systems over the time period for which bone
lead concentrations have been gathered. This parameter itself represents a 40-fold mass
10-66
-------
Increase. This significant skeletal growth rate has an Impact on an obvious question: 1f
children take 1n more lead on a body-weight basis than adults, absorb and retain more lead
than adults, and show only modest elevations in blood lead compared to adults in the face of a
more active skeletal system, where does the lead go? A second factor is the assumption that
blood lead in children relates to body lead burden in the same quantitative fashion as in
adults, an assumption that remains to be proven adequately.
10.8.3.2 Animal Studies. In rats and other experimental animals, both urinary and fecal ex-
cretion appear to be important routes of lead removal from the organism. The relative parti-
tioning between the two modes is species- and dose-dependent. With regard to species differ-
ences, biliary clearance of lead in the dog is but 2 percent of that for the rat, while such
excretion in the rabbit 1s 50 percent that of the rat.
Lead movement from laboratory animals to their offspring via milk constituents is a route
of excretion for the mother as well as a route of exposure for the young. Comparative studies
of lead retention in developing versus adult animals such as rats, mice, and nonhuman primates
make it clear that retention is significantly greater in the young animal. These observations
support those studies showing greater lead retention in children. Some recent data indicate
that a differential retention of lead 1n young rats persists Into the post-weaning period,
calculated as either uniform dosing or uniform exposure.
10.8.4 Interactions of Lead with Essential Metals and Other Factors
Toxic elements such as lead are affected in their toxicokinetic or toxicological behavior
by interactions with a variety of biochemical factors, particularly nutrients.
10.8.4.1 Human Studies. In humans, the interactive behavior of lead and various nutritional
factors is expressed most significantly in young children, with such interactions occurring
against a backdrop of rather widespread deficiencies in a number of nutritional components.
Various surveys have indicated that iron, calcium, zinc, and vitamin deficiencies are wide-
spread among the pediatric population, particularly the poor. A number of reports have docu-
mented the association of lead absorption with suboptimal nutritional states for iron and cal-
cium, reduced intake being associated with increased lead absorption.
10.8.4.2 Animal Studies. Reports of lead-nutrient interactions in experimental animals have
generally described such relationships for a single nutrient, using relative absorption or
tissue retention in the animal to index the effect. Most of the recent data are for calcium,
Iron, phosphorus, and vitamin D. Many studies have established that diminished dietary calci-
um is associated with increased blood and soft-tissue lead content in such diverse species as
the rat, pig, horse, sheep, and domestic fowl. The increased body burden of lead arises from
both increased GI absorption and increased retention, indicating that the lead-calcium inter-
action operates at both the gut wall and within body compartments. Lead appears to traverse
10-67
-------
the gut via both passive and active transfer. It involves transport proteins normally operat-
ing for calcium transport, but is taken up at the site of phosphorus, not calcium, absorption.
Iron deficiency is associated with an increase of lead in tissues and increased toxicity,
effects that are expressed at the level of lead uptake by the gut wall, Iji vitro studies
indicate an interaction through receptor-binding competition at a common site, which probably
involves iron-binding proteins. Similarly, dietary phosphate deficiency enhances the extent
of lead retention and toxicity via increased uptake of lead at the gut wall, both lead and
phosphate being absorbed at the same site in the small intestine. Results of various studies
of the resorption of phosphate along with lead have not been able to identify conclusively a
mechanism for the elevation of tissue lead. Since calcium plus phosphate retards lead absorp-
tion to a greater degree than simply the sums of the interactions, an insoluble complex of all
these elements may be the basis of this retardation.
Unlike the inverse relationship existing for calcium, iron, and phosphate versus lead up-
take, vitamin D levels appear directly related to the rate of lead absorption from the GI
tract, since the vitamin stimulates the same region of the duodenum where lead is absorbed. A
number of other nutrient factors are known to have an interactive relationship with lead:
1. Increases in dietary Hpids increase the extent of lead absorption, with the extent
of the increase being highest with polyunsaturates and lowest with saturated fats,
e.g., tristearin.
2. The interactive relationship of lead and dietary protein is not clear cut, and
either suboptimal or excess protein intake will increase lead absorption.
3. Certain milk components, particularly lactose, greatly enhance lead absorption in
the nursing animal.
4. Zinc deficiency promotes lead absorption, as does reduced dietary copper.
Taken collectively, human and animal data dealing with the interaction of lead and nutri-
ents indicate that there are heterogeneous subsets of the human population. In terms of ped-
atric population risk for lead exposure, children having multiple nutrient deficiencies are 1n
the highest exposure risk category.
10.8.5 Interrelationships of Lead Exposure with Exposure Indicators and Tissue Lead Burdens
Three issues involving lead toxlcokinetics evolve toward a full connection between lead
exposure and its adverse effects: (1) the temporal characteristics of Internal Indices of
lead exposure; (2) the biological aspects of the relationship of lead in various media to
various indicators in internal exposure; and (3) the relationship of various Internal indica-
tors of exposure to target tissue lead burdens.
10-68
-------
10.8.5.1 Temporal Characteristics of Internal Indicators of Lead Exposure. The biological
half-life for newly absorbed lead in blood may be as short as weeks, or several months. Or, it
may be longer, depending on the mobile lead burden in the body. Compared to mineral tissues,
this medium reflects relatively recent exposure. If recent exposure is fairly representative
of exposure over a considerable period of time, e.g., exposure of lead workers, then blood
lead 1s more useful than for cases where exposure is intermittent or different across time, as
fn the case of lead exposure of children. Accessible mineralized tissue, such as shed teeth,
extend the time frame back to years of exposure, since teeth accumulate lead with age and as a
function of the extent of exposure. Such measurements are, however, retrospective in nature,
In that Identification of excessive exposure occurs after the fact and thus limits the possi-
bility of timely medical intervention, exposure abatement, or regulatory policy concerned with
ongoing control strategies.
Perhaps the most practical solution to the dilemma posed by both tooth and blood lead
analyses is jji situ measurement of lead in teeth or bone during the time when active accumu-
lation occurs, e.g., 2- to 3-year-old children. Available data using X-ray fluorescence anal-
ysis do suggest that such approaches are feasible and can be reconciled with such issues as
acceptable radiation hazard risk to subjects.
10.8.5.2 Biological Aspects of External Exposure/Internal Indicator Relationships. The
literature Indicates clearly that the relationship of lead 1n relevant media for human expo-
sure to blood lead is curvilinear when viewed over a relatively broad range of blood lead
values. This curvllinearity Implies that the unit change in blood lead per unit intake of
lead In some medium varies across this range of exposure, with comparatively smaller blood
lead changes occurring as Internal exposure increases.
Given our present knowledge, such a relationship cannot be taken to mean that body uptake
of lead 1s proportionately lower at higher exposure, because it may simply mean that blood
lead becomes an increasingly unreliable measure of target-tissue lead burden with increasing
exposure. While the basis of the curvilinear relationship remains to be identified, available
animal data suggest that it may be related to the increasing fraction of blood lead in plasma
as blood lead Increases above approximately 50-60 ug/dl.
10.8.5.3 Internal Indicator/Tissue Lead Relationships. In living human subjects, direct de-
termination of tissue lead burdens or how these relate to adverse effects in target tissues is
not possible. Some accessible Indicator (e.g., lead in a medium such as blood or a biochem-
ical surrogate of lead such as erythrocyte protoporphyHn), must be employed. While blood
lead still remains the only practical measure of excessive lead exposure and health risk, evi-
dence continues to accumulate that such an index has some limitations in either reflecting
tissue lead burdens or changes in such tissues with changes in exposure.
10-69
-------
At present, the measurement of plumburesis associated with challenge by a single dose of
a lead-chelating agent such as CaNa2EDTA is considered the best Indicator of the mobile,
potentially toxic fraction of body lead. Chelatable lead is logarithmically related to blood
lead, such that an incremental increase in blood lead is associated with an increasingly
larger increment of mobilizable lead. The problems associated with this logarithmic relation-
ship may be seen in studies of children and lead workers in whom moderate elevation 1n blood
lead levels can disguise levels of mobile body lead. In one recent multi-institution study of
210 children, for example, 12 percent of children with blood lead 30-39 ug/dl, and 38 percent
with levels of 40-49 pg/dl, had a positive EDTA-challenge response and required further eval-
uation or treatment. At blood lead levels such as these, the margin of protection against
severe intoxication is reduced. The biological basis of the logarithmic chelatable lead/
blood lead relationship rests, in large measure, with the existence of a sizeable bone lead
compartment that is mobile enough to undergo chelatlon removal and, hence, potentially mobile
enough to move into target tissues.
Studies of the relative mobility of chelatable lead over time indicate that, in former
lead workers, removal from exposure leads to a protracted washing out of lead (from bone re-
sorptlon of lead) to blood and tissues, with preservation of a bone burden amenable to
subsequent chelation. Studies with children are inconclusive, since the one investigation
directed to this end employed pediatric subjects who all underwent chelation therapy during
periods of severe lead poisoning. Animal studies demonstrate that changes in blood lead with
Increasing exposure do not agree with tissue uptake in a time-concordant fashion, nor does de-
crease 1n blood lead with reduced exposure signal a similar decrease 1n target tissue, parti-
cularly in the brain of the developing organism.
10.8.6 Metabolism of Lead AlkyIs
The lower alkyl lead components used as gasoline additives, tetraethyl lead (TEL) and
tetramethyl lead (TML), may themselves poise a toxic risk to humans. In particular, there 1s
among children a problem of sniffing leaded gasoline.
10.8.6.1 Absorption of Lead Alkyls In Humans and Animals. Human volunteers Inhaling labeled
TEL and TML show lung deposition rates for the lead alkyls of 37 and 51 percent, respectively,
values which are similar to those for partlculate inorganic lead. Significant portions of
these deposited amounts were eventually absorbed. Respiratory absorption of organolead bound
to partlculate matter has not been specifically studied as such.
While specific data for the GI absorption of lead alkyls in humans and animals are not
available, their close similarity to organotin compounds, which are quantitatively absorbed,
would argue for extensive GI absorption. In contrast to inorganic lead salts, the lower lead
10-70
-------
alkyls are extensively absorbed through the skin and animal data show lethal effects with per-
cutaneous uptake as the sole route of exposure.
10.8.6.2 Blotransformatlon and Tissue Distribution of Lead Alkyls. The lower lead alkyls TEL
and TML undergo monodealkylatlon 1n the liver of mammalian species via the P-450-dependent
mono-oxygenase enzyme system. Such transformation 1s very rapid. Further transformation
Involves conversion to the dlalkyl and Inorganic lead forms, the latter accounting for the
effects on heme biosynthesis and erythropoiesis observed in alkyl lead intoxication. Alkyl
lead is rapidly cleared from blood and shows a higher partitioning into plasma than Inorganic
lead, with tr1ethyl lead clearance being more rapid than that of the methyl analog.
Tissue distribution of alkyl lead in humans and animals primarily Involves the trialkyl
metabolites. Levels are highest in liver, followed by kidney, then brain. Of interest 1s the
fact that there are detectable amounts of trlalkyl lead from autopsy samples of human brain
even in the absence of occupational exposure. In humans, there appear to be two tissue com-
partments for triethyl lead, having half-times of 35 and 100 days.
10.8.6.3 Excretion of Lead Alkyls. With alkyl lead exposure, excretion of lead through the
renal tract is the main route of elimination. The chemical forms being excreted appear to be
species-dependent. In humans, trlalkyl lead in workers chronically exposed to alkyl lead is a
minor component of urine lead, approximately 9 percent.
10-71
-------
10.9 REFERENCES
Adamiak-Ziemba, J.; Bolanowska, W. (1970) Proba oceny ekspozycji zawodowej na czteroetylek
olowiu na podstawie jego stezen w powietrzu i stezen trojetylku olowiu w moczu [Occupa-
tional exposure to tetraethyl lead as estimated by tetraethyl lead concentration in air
and triethyl lead concentration in urine]. Med. Pr. 21: 172-179.
Alessio, L.; Bertazzi, P. A.; Monelli, 0. ; Toffoletto, F. (1976) Free erythrocyte protoporphyr-
in as an indicator of the biological effect of lead in adult males. III. Behavior of
free erythrocyte protoporphyrin in workers with past lead exposure. Int. Arch. Occup.
Environ. Health 38: 77-86.
Alexander, F. W.; Delves, H. T. (1981) Blood lead levels during pregnancy. Int. Arch. Occup.
Environ. Health 48: 35-39.
Alexander, F. W.; Delves, H. T.; Clayton, B. E. (1973) The uptake and excretion by children of
lead and other contaminants. In: Barth, D.; Berlin, A.; Engel, R.; Recht, P.; Smeets, J.,
eds. Environmental health aspects of lead: proceedings, international symposium; October
1972; Amsterdam, The Netherlands. Luxembourg: Commission of the European Communities; pp.
319-331.
Allcroft, R. (1950) Lead as a nutritional hazard to farm livestock. IV: Distribution of lead
in the tissues of bovines after ingestion of various lead compounds. J. Comp. Pathol. 60:
190-208.
Anders, E.; Bagnell, C. R., Jr.; Krigman, M. R.; Mushak, P. (1982) Influence of dietary pro-
tein composition on lead absorption in rats. Bull. Environ Contain. Toxicol. 28: 61-67.
Arai, F.; Yamamura, Y.; Yoshida, M. (1981) Excretion of triethyl lead, diethyl lead and inor-
ganic lead after injection of tetraethyl lead in rabbits. Sangyo Igaku 23: 496-504.
Araki, S. (1973) On the behaviour of "active deposit of lead (Teisinger)" in the Japanese free
from occupational exposure to lead. Ind. Health 11: 203-224.
Araki, S.; Ushio, K. (1982) Assessment of the body burden of chelatable lead: a model and its
application to lead workers. Br. J. Ind. Med. 39: 157-160.
Aungst, B. J.; Fung, H. (1981) Kinetic characterization of j_n vitro lead transport across the
rat small intestine. Toxicol. Appl. Pharmacol. 61: 39-47.
Aungst, B. J.; Dolce, J. A.; Fung, H. (1981) The effect of dose on the disposition of lead in
rats after intravenous and oral administration. Toxicol. Appl. Pharmacol. 61: 48-57.
Awad, L.; Huel, G.; Lazar, P.; Boudene, C. (1981) Facteurs de variation interindividuelle
de la plombemie [Factors of interindividual variations of blood lead levels]. Rev.
Epidemic!. Sante Publique 29: 113-124.
Azar, A.; Trochimowicz, H. J.; Maxfield, M. E. (1973) Review of lead studies in animals car-
ried out at Haskell Laboratory: two year feeding study and response to hemmorhage study.
In: Barth, D. ; Berlin, A.; Engel, R.; Recht, P.; Smeets, J., eds. Environmental health
aspects of lead: proceedings, international symposium; October 1972; Amsterdam, The
Netherlands. Luxembourg: Commission of the European Communities; pp. 199-210.
10-72
-------
Azar, A.; Snee, R. 0.; Habibi, K. (1975) An epidemiologies! approach to community air lead
exposure using personal air samplers. In: Griffin, T. B.; Knelson, J. H., eds. Lead.
Stuttgart, West Germany: Georg Thieme Publishers; pp. 254-290. (Coulston, F.; Korte, F.,
eds. Environmental quality and safety: supplement v. 2).
Bade, V.; Huber, F. (1970) Reaktionen von Organometallverbindungen. IV: Uber den Einfluss
der organischen Liganden und des Zentralatoms auf die Geschwindigkeit der Acidolyse von
Tetraorganoplumbanen und Tetraorganostannanen [Reactions of organometallic compounds. IV:
The influence of the organic ligand and the central atom on the speed of acidolysis of
organolead compounds and analogous tin compounds]. J. Organometal. Chem. 24: 387-397.
Baloh, R. W. (1974) Laboratory diagnosis of increased lead absorption. Arch. Environ. Health
28: 198-208.
Barltrop, D. (1969) Transfer of lead to the human foetus. In: Barltrop, D.; Burland, W. L.,
eds. Mineral metabolism in pediatrics. Philadelphia, PA: Davis Co.; pp. 135-151.
Barltrop, D. (1972) Children and environmental lead. In: Hepple, P., ed. Lead in the environ-
ment: proceedings of a conference; London, United Kingdom. London, United Kingdom:
Institute of Petroleum; pp. 52-60.
Barltrop, D. (1975) Assessment of the health hazard of various lead compounds. Atlanta, GA:
U.S. Department of Health, Education and Welfare, Center for Disease Control.
Barltrop, D. (1982) Nutritional and maturational factors modifying the absorption of inorganic
lead from the gastrointestinal tract. In: Hunt, V. R.; Smith, M. K.; Worth, D. Environ-
mental factors in human growth and development. Cold Spring Harbor, NY: Cold Spring
Harbor Laboratory, pp. 35-41. (Banbury report 11).
Barltrop, D.; Khoo, H. E. (1975) The influence of nutritional factors on lead absorption.
Postgrad. Med. J. 51: 795-800.
Barltrop, D.; Meek, F. (1975) Absorption of different lead compounds. Postgrad. Med. J. 51:
805-809.
Barltrop, D.; Meek, F. (1979) Effect of particle size on lead absorption from the gut. Arch.
Environ. Health 34: 280-285.
Barltrop, D.; Strehlow, C. D. (1978) The absorption of lead by children. In: Kirchgessner, M.,
ed. Trace element metabolism in man and animals - 3: proceedings of the 3rd international
symposium; July 1977; Freising, West Germany. Freising-Weihenstephan, West Germany:
Technische UniversitMt Miinchen; pp. 332-334.
Barltrop, D.; Strehlow, C. D.; Thornton, I.; Webb, J. S. (1974) Significance of high soil lead
concentrations for childhood lead burdens. Environ. Health Perspect. 7: 75-82.
Barnes, J. M.; Stoner, H. B. (1958) Toxic properties of some dialkyl and trialkyl tin salt.
Br. J. Ind. Med. 15: 15-22.
Barry, P. S. I. (1975) A comparison of concentrations of lead in human tissues. Br. J. Ind.
Med. 32: 119-139.
10-73
-------
Barry, P. S. I. (1978) Distribution and storage of lead in human tissues. In: Nriagu, J. 0.,
ed. The biogeochemistry of lead in the environment; part B. biological effects. Amster-
dam, The Netherlands: Elsevier/North Holland Biomedical Press; pp. 97-150. (Topics in
environmental health: v. IB).
Barry, P. S. I. (1981) Concentrations of lead in the tissues of children. Br. J. Ind. Med. 38:
61-71.
Barry, P. S. I. ; Mossman, D. B. (1970) Lead concentrations in human tissues. Br. J. Ind. Med.
27: 339-351.
Barton, J. C.; Conrad, M. E. (1981) Effect of phosphate on the absorption and retention of
lead in the rat. Am. J. Clin. Nutr. 34: 2192-2198.
Barton, J. C.; Conrad, M. E.; Harrison, L.; Nuby, S. (1978a) Effects of calcium on the absorp-
tion and retention of lead. J. Lab. Clin. Med. 91: 366-376.
Barton, J. C.; Conrad, M. E.; Nuby, S.; Harrison, L. (1978b) Effects of iron on the absorption
and retention of lead. J. Lab. Clin. Med. 92: 536-547.
Barton, J. C.; Conrad, M. E.; Harrison, L.; Nuby, S. (1980) Effects of vitamin D on the absorp-
tion and retention of lead. Am. J. Physiol. 238: G124-G130.
Batschelet, E.; Brand, L.; Steiner, A. (1979) On the kinetics of lead in the human body. J
Math. Biol. 8: 15-23.
Batuman, V.; Maesaka, J. K.; Haddad, B.; Tepper, E.; Landry, E.; Wedeen, R. P. (1981) The role
of lead in gout nephropathy. N. Engl. J. Med. 304: 520-523.
Batuman, V.; Landy, E.; Maesaka, J. K.; Wedeen, R. P. (1983) Contribution of lead to hypertens-
ion with renal impairment. N. Engl. J. Med. 309: 17-21.
Bell, R. R.; Spickett, J. T. (1981) The influence of milk in the diet on the toxicity of
orally ingested lead in rats. Food Cosmet. Toxicol. 19: 429-436.
Berg, L. R. ; Nordstrom, J. 0.; Ousterhout, L. E. (1980) The prevention of chick growth depres-
sion due to dietary lead by increased dietary calcium and phosphorus levels. Poult Sci
59: 1860-1863.
Bianco, A. ; Gibb, F. R. ; Morrow, P. E. (1974) Inhalation study of a submicron size lead-212
aerosol. In: Snyder, W. S., ed. Proceedings of the third international congress of the
International Radiation Protection Association; September 1973; Washington, DC. Oak
Ridge, TN: U.S. Atomic Energy Commission; pp. 1214-1219. Available from: NTIS, Sprlna-
field, VA; CONF-730907-P2. w
Bjo'rklund, H.; Lind, B.; Piscator, M.; Hoffer, B.; Olson, L. (1981) Lead, zinc, and copper
levels in intraocular brain tissue grafts, brain, and blood of lead-exposed rats
Toxicol. Appl. Pharmacol. 60: 424-430.
Blake, K. C. H. (1976) Absorption of 203Pb from gastrointestinal tract of man. Environ Res
11: 1-4.
Blake, K. C. H. (1980) Radioactive lead studies in the human [dissertation]. Capetown, South
Africa: University of Capetown.
10-74
-------
Boeckx, R. L. ; Posti, B.; Coodin, F. J. (1977) Gasoline sniffing and tetraethyllead poisoning
in children. Pediatrics 60: 140-145.
Bolanowska, W. (1968) Distribution and excretion of triethyllead in rats. Br. J. Ind. Med. 25:
203-208.
Bolanowska, W.; Garczyriski, H. (1968) Metabolizm czteroetylku ojowiu u krolikdw [Metabolism
of tetraethyl lead in rabbits]. Med. Prac. 19: 235-243.
Bolanowska, W. ; Piotrowski, J. ; Garczynski, H. (1967) Triethyllead in the biological material
in cases of acute tetraethyllead poisoning. Arch. Toxicol. 22: 278-282.
Booker, D. V.; Chamberlain, A. C.; Newton, D.; Stott, A. N. B. (1969) Uptake of radioactive
lead following inhalation and injection. Br. J. Radiol. 42: 457-466.
Boudene, C.; Malet, D.; Masse, R. (1977) Fate of 210Pb inhaled by rats. Toxicol. Appl.
Pharmacol. 41: 271-276.
Bruenger, F. W.; Stevens, W.; Stover, B. J. (1973) The association of 210Pb with constituents
of erythrocytes. Health Phys. 25: 37-42.
Brunekreef, B. D. (1984) The relationship between air lead and blood lead in children: a crit-
ical review. Sci. Total Environ. 38: 79-123.
Brunekreef, B.; Noy, D.; Biersteker, K.; Boleij, J. (1983) Blood lead levels of Dutch city
children and their relationship to lead in the environment. J. Air Pollut. Control Assoc.
33: 872-876.
Buchet, J.-P.; Roels, H.; Hubermont, G.; Lauwerys, R. (1978) Placenta! transfer of lead, mer-
cury, cadmium, and carbon monoxide in women: II. influence of some epidemiological fac-
tors on the frequency distributions of the biological indices in maternal and umbilical
cord blood. Environ. Res. 15: 494-503.
Bull, R. J.; Lutkenhoff, S. D.; McCarty, G. E.; Miller, R. G. (1979) Delays in the postnatal
increase of cerebral cytochrome concentrations in lead-exposed rats. Neuropharmacology
18: 83-92.
Bushnell, P. J.; DeLuca, H. F. (1981) Lactose facilitates the intestinal absorption of lead in
weanling rats. Science (Washington, DC) 211: 61-63.
Bushnell, P. J.; DeLuca, H. F. (1983) The effects of lactose on the absorption and retention
of dietary lead. J. Nutr. 113: 365-378.
Bushnell, P. J.; Levin, E. D. (1983) Effects of zinc deficiency on lead toxicity in rats.
Neurobehav. Toxicol. Teratol. 5: 283-288.
Butt, E. M.; Nusbaum, R. E.; Gilmore, T. C.; Didio, S. L.; Mariaro, Sister. (1964) Trace metal
levels in human serum and blood. In: Proceedings of the sixth annual air pollution med-
ical research conference; January 1963; San Francisco, CA. Arch. Environ. Health 8:
52-57.
Byington, K. H.; Yates, D. A.; Mullins, W. A. (1980) Binding of triethyllead chloride by hemo-
globin. Toxicol. Appl. Pharmacol. 52: 379-385.
10-75
-------
Castellino, N.; Aloj, S. (1964) Kinetics of the distribution and excretion of lead in the rat.
Br. J. Ind. Med. 21: 308-314.
Cavalleri, A.; Minoia, C.; Pozzoli, L. ; Baruffini, A. (1978) Determination of plasma lead
levels in normal subjects and in lead-exposed workers. Br. J. Ind. Med. 35: 21-26.
Cerklewski, F. L. (1979) Influence of dietary zinc on lead toxicity during gestation and
lactation in the female rat. J. Nutr. 109: 1703-1709.
Cerklewski, F. L. (1980) Reduction in neonatal lead exposure by supplemental dietary iron dur-
ing gestation and lactation in the rat. J. Nutr. 110: 1453-1457.
Cerklewski, F. L.; Forbes, R. M. (1976) Influence of dietary zinc on lead toxicity in the rat
J. Nutr. 106: 689-696.
Chamberlain, A. C. (1983) Effect of airborne lead on blood lead. Atmos. Environ. 17: 677-692.
Chamberlain, A. C. ; Heard, M. J. (1981) Lead tracers and lead balances. In: Lynam, 0. R.
Piantanida, L. G.; Cole, J. F., eds. Environmental lead: proceedings of the second inter-
national symposium on environmental lead research; December 1978; Cincinnati, OH. New
York, NY: Academic Press; pp. 175-198. (Coulston, F.; Korte, F., eds. Ecotoxicology and
environmental quality series).
Chamberlain, A. C.; Heard, M. J.; Little, P.; Newton, D.; Wells, A. C.; Wiffen, R. D. (1978)
Investigations into lead from motor vehicles. Harwell, United Kingdom: United Kingdom
Atomic Energy Authority; report no. AERE-R9198.
Chesney, R. W.; Rosen, J. F.; Hamstra, A. J.; Smith, C.; Mahaffey, K. ; DeLuca, H. F. (1981)
Absence of seasonal variation in serum concentrations of 1,25-dihydroxyvitamin D despite
a rise in 25-hydroxyvitamin D in summer. J. Clin. Endocrinol. Metab. 53: 139-142.
Chiesura, P. (1970) Escrezione urinaria di cataboliti del piombotetraetile nell'uomo [Urinary
excretion of tetraethyl lead catabolites in man]. Med. Lav. 61: 437-441.
Chisolm, J. J., Jr. (1981) Dose-effect relationships for lead in young children: evidence in
children for interactions among lead, zinc, and iron. In: Lynam, D. R.; Piantanida, L. Q-
Cole, J. F., eds. Environmental lead: proceedings of the second international symposium
on environmental lead research; December 1978; Cincinnati, OH. New York, NY: Academic
Press; pp. 1-7. (Coulston, F.; Korte, F., eds. Ecotoxicology and environmental quality
series).
Chisolm, J. J., Jr.; Barltrop, D. (1979) Recognition and management of children with increased
lead absorption. Arch. Dis. Child. 54: 249-262.
Chisolm, J. J., Jr.; Harrison, H. E. (1956) Quantitative urinary coproporphyrin excretion and
its relation to edathamil calcium disodium administration in children with acute lead
intoxication. J. Clin. Invest. 35: 1131-1138.
Chisolm, J. J. , Jr.; Mellits, E. D. ; Barrett, M. B. (1976) Interrelationships among blood lead
concentration, quantitative daily ALA-U and urinary lead output following calcium EDTA.
In: Nordberg, G. F.; ed. Proceedings of third meeting of the subcommittee on the toxi-
cology of metals under the Permanent Commission and International Association on Occupa-
tional Health; November 1974; Tokyo, Japan. Amsterdam, The Netherlands: Elsevier Pub-
lishing Co.; pp. 416-433.
10-76
-------
Clarkson, T. W. ; Kench, J. E. (1958) Uptake of lead by human erythrocytes in vitro. Biochem.
J. 69: 432-439.
Cohen, N. (1970) The retention and distribution of lead-210 in the adult baboon [disserta-
tion]. New York, NY: New York University. Available from: University Microfilms, Ann
Arbor, MI; publication no. 71-68.
Collins, M. F. ; Hrdina, P. D.; Whittle, E. ; Singhal, R. L. (1982) Lead in blood and brain
regions of rats chronically exposed to low doses of the metal. Toxicol. Appl. Pharmacol.
65: 314-322.
Conrad, M. E.; Barton, J. C. (1978) Factors affecting the absorption and excretion of lead in
the rat. Gastroenterology 74: 731-740.
Cramer, K.; Goyer, R. A.; Jagenburg, R.; Wilson, M. H. (1974) Renal ultrastructure, renal
function, and parameters of lead toxicity in workers with different periods of lead
exposure. Br. J. Ind. Med. 31: 113-127.
Cremer, J. E. (1959) Biochemical studies on the toxicity of tetraethyl lead and other organo-
lead compounds. Br. J. Ind. Med. 16: 191-199.
Cremer, J. E.; Callaway, S. (1961) Further studies on the toxicity of some tetra and trialkyl
lead compounds. Br. J. Ind. Med. 18: 277-282.
Dacre, J. C.; Ter Haar, G. L. (1977) Lead levels in tissues from rats fed soils containing
lead. Arch. Environ. Contam. Toxicol. 6: 111-119.
Danscher, G.; Hall, E.; Fredens, K.; Fjerdingstad, E.; Fjerdingstad, E. J. (1975) Heavy metals
in the amygdala of the rat: zinc, lead and copper. Brain Res. 94: 167-172.
David, 0. J.; Wintrob, H. L.; Arcoleo, C. G. (1982) Blood lead stability. Arch. Environ.
Health 37: 147-150.
Davis, R. K. ; Horton, A. W.; Larson, E. E.; Stemmer, K. L. (1963) Inhalation of tetramethyl-
lead and tetraethyllead: a comparison of the effects in rats and dogs. Arch. Environ.
Health 6: 473-479.
Day, J. P.; Hart, M.; Robinson, M. S. (1975) Lead in urban street dust. Nature (London) 253:
343-345.
Day, J. P.; Fergusson, J. E.; Chee, T. M. (1979) Solubility and potential toxicity of lead in
urban street dust. Bull. Environ. Contam. Toxicol. 23: 497-502.
Delves, H. T.; Clayton, B. E.; Carmichael, A.; Bubear, M.; Smith M. (1982) An appraisal of the
analytical significance of tooth-lead measurements as possible indices of environmental
exposure of children to lead. Ann. Clin. Biochem. 19: 329-337.
Delves, H. T.; Sherlock, J. C.; Quinn, M. J. (1984) Temporal stability of blood lead concen-
trations in adults exposed only to environmental lead. Hum. Toxicol. 3: 279-288.
Der, R.; Fahim, Z.; Hilderbrand, D.; Fahim, M. (1974) Combined effect of lead and low protein
diet on growth, sexual development, and metabolism in female rats. Res. Commun. Chem.
Pathol. Pharmacol. 9: 723-738.
10-77
-------
DeSilva, P. E. (1981) Determination of lead in plasma and studies on its relationship to lead
in erythrocytes. Br. J. Ind. Med. 38: 209-217.
Devoto, G.; Spinazzola, A. (1973) L'effet de la presence du plomb dans 1'air, dans Teau et
les denrees alimentaires sur son taux d'absorption chez differentes categories de per-
sonnes [The effects of the presence of lead in the air, in water and in foodstuffs on the
rate of absorption among different categories of people]. In: Barth, D.; Berlin, A.;
Engel, R.; Recht, P.; Smeets, J., eds. Environmental health aspects of lead: proceedings,
international symposium; October 1972; Amsterdam, The Netherlands. Luxembourg: Commission
of the European Communities; pp. 859-867.
Dobbins, A.; Johnson, D. R.; Nathan, P. (1978) Effect of exposure to lead on maturation of
intestinal iron absorption of rats. J. Toxicol. Environ. Health 4: 541-550.
Drill, S.; Konz, J.; Mahar, H.; Morse, M. (1979) The environmental lead problem: an assessment
of lead in drinking water from a multi-media perspective. Washington, DC: U.S. Environ-
mental Protection Agency; EPA report no. EPA-570/9-79-003. Available from: NTIS, Spring-
field, VA; PB-296556.
Duggan, M. J.; Williams, S. (1977) Lead-in-dust in city streets. Sci. Total Environ. 7: 91-97.
El-Gazzar, R. M.; Finelli, V. N.; Boiano, J.; Petering, H. G. (1978) Influence of dietary zinc
on lead toxicity in rats. Toxicol. Lett. 1: 227-234
Emerson, B. J. (1963) Chronic lead nephropathy: the diagnostic use of calcium EDTA and the
association with gout. Australas. Ann. Med. 12: 310-324. Forbes, G. B.; Reina, J. c.
(1972) Effect of age on gastrointestinal absorption (Fe, Sr, Pb) in the rat. J. Nutr
102: 647-652.
Everson, J,; Patterson, C. C. (1980) "Ultra-clean" isotope dilution/mass spectrometric anal-
yses for lead in human blood plasma indicate that most reported values are artificially
high, Clin. Chem. (Winston-Salem, NC) 26:1603-1607.
Fjerdingstad, E. J.; Danscher, G.; Fjerdingstad, E. (1974) Hippocampus: selective concen-
tration of lead in the normal rat brain Res. 80: 350-354.
Flanagan, P. R.; Hamilton, D. L.; Haist, J.; Valnerg, L. S. (1979) Inter-relationships between
iron and absorption in iron-deficient mice. Gastroenterology 77: 1074-1081.
Flanagan, P. R.; Chamberlain, M. J.; Valberg, L. S. (1982) The relationship between iron and
lead absorption in humans. Am. J. Clin. Nutr. 36: 823-829.
Forbes, G. B.; Reina, J. C. (1972) Effect of age on gastrointestinal ansorption (Fe, Sr, Pb)
in the rat. J. Nutr. 102: 647-652.
Fremlin, J. M.; Edmonds, M. I. (1980) The determination of lead in human teeth. Nucl. Instrum
Methods 173: 211-215.
Fukumoto, K.; Karai, I.; Horiguchi, S. (1983) Effect of lead on erythrocyte membranes. Br J
Ind. Med. 40: 220-223.
Gage, J. C.; Litchfield, M. H. (1968) The migration of lead from polymers in the rat gastro-
intestinal tract. Food Cosmet. Toxicol. 6: 329-338.
10-78
-------
Gage, J. C.; Litchfield, M. H. (1969) The migration of lead from paint films in the rat gas-
trointestinal tract. J. Oil Colour Chem. Assoc. 52: 236-243.
Garber, B. T.; Wei, E. (1974) Influence of dietary factors on the gastrointestinal absorption
of lead. Toxicol. Appl. Pharmacol. 27: 685-691.
Gershanik, J. J.; Brooks, G. G.; Little, J. A. (1974) Blood lead values in pregnant women and
their offspring. Am. J. Obstet. Gynecol. 119: 508-511.
Goldings, A. S.; Stewart, R. M. (1982) Organic lead encephalopathy: behavioral change and
movement disorder following gasoline inhalation. J. Clin. Psychol. 43: 70-72.
Goldstein, G. W.; Diamond, I. (1974) Metabolic basis of lead encephalopathy. In: Plum, F.,
ed. Brain dysfunction in metabolic disorders. Res. Publ. Assoc. Nerv. Ment. Dis. 53:
293-304.
Goldstein, G. W.; Asbury, A. K.; Diamond, I. (1974) Pathogenesis of lead encephalopathy: up-
take of lead and reaction of brain capillaries. Arch. Neurol. (Chicago) 31: 382-389.
Goyer, R. A. (1978) Calcium and lead interactions: some new insights. J. Lab. Clin. Med. 91:
363-365.
Goyer, R. A.; May, P.; Cates, M. M.; Krigman, M. R. (1970) Lead and protein content of iso-
lated intranuclear inclusion bodies from kidneys of lead-poisoned rats. Lab. Invest. 22:
245-251.
Grandjean, P. (1978) Regional distribution of lead in human brains. Toxicol. Lett. 2: 65-69.
Grandjean, P. ; Nielsen, T. (1979) Organolead compounds: environmental health aspects. Residue
Rev. 72: 98-148.
Grant, R. L. ; Calvery, H. 0.; Laug, E. P.; Morris, H. J. (1938) The influence of calcium and
phosphorus on the storage and toxicity of lead and arsenic. J. Pharmacol. Exp. Ther. 64:
446-457.
Grant, L. D.; Kimrnel, C. A.; West, G. L.; Martinez-Vargas, C. M.; Howard, J. L. (1980) Chronic
low-level lead toxicity in the rat: II. effects on postnatal physical and behavioral
development. Toxicol. Appl. Pharmacol. 56: 42-58.
Greenhalgh, J. R. ; James, A. C. ; Smith, H.; Hodgson, A. (1979) Absorption of lead ions from
the lung. In: National Radiological Protection Board annual research and development
report; no. NRPB/R&D 3. Harwell, United Kingdom: National Radiological Protection Board;
pp. 38-42.
Griffin, T. B. ; Coulston, F. ; Golberg, L. ; Wills, H. ; Russell, J. C.; Knelson, J. H. (1975)
Clinical studies on men continuously exposed to airborne particulate lead. In: Griffin,
T. B.; Knelson, J. H., eds. Lead. Stuttgart, West Germany: Georg Thieme Publishers; pp.
221-240. (Coulston, F. ; Korte, F. , eds. Environmental quality and safety: supplement v.
2).
Gross, S. B. (1981) Human oral and inhalation exposures to lead: summary of Kehoe balance ex-
periments. J. Toxicol. Environ. Health 8: 333-377.
10-79
-------
Gross, S. B.; Pfitzer, E. A. (1974) Influence of pathological change on lead in human tissues.
In: Hemphill, D. D., ed. Trace substances in environmental health - VIII: [proceedings of
University of Missouri's 8th annual conference on trace substances in environmental
health]; June; Columbia, MO. Columbia, MO: University of Missouri-Columbia; pp. 335-340.
Gross, S. B.; Pfitzer, E. A.; Yeager, D. W.; Kehoe, R. A. (1975) Lead in human tissues.
Toxicol. Appl. Pharmacol. 32: 638-651.
Hambidge, K. M. (1977) The role of zinc and other trace metals in pediatric nutrition and
health. Pediatr. Clin. North Am. 24: 95-106.
Hamilton, D. L. (1978) Interrelationships of lead and iron retention in iron-deficient mice.
Toxicol. Appl. Pharmacol. 46: 651-661.
Hammond, P. B. (1971) The effects of chelating agents on the tissue distribution and excretion
of lead. Toxicol. Appl. Pharmacol. 18: 296-310.
Hammond, P. B. (1973) The effects of D-penicillamine on the tissue distribution and excretion
of lead. Toxicol. Appl. Pharmacol. 26: 241-246.
Hammond, P. B.; O'Flaherty, E. J.; Gartside, P. S. (1981) The impact of air-lead on blood-
lead in man - a critique of the recent literature. Food Cosmet. Toxicol.19: 631-638.
Hammond, P. B.; O'Flaherty, E. J.; Gartside, P. S. (1982) Impact of air-lead on blood-lead in
man [letter]. Food Chem. Toxicol. 20: 493.
Hansen, J. P. B.; Ddssing, M.; Paulev, P.-E. (1981) Chelatable lead body burden (by calcium-
disodium EDTA) and blood lead concentration in man. J. Occup. Med. 23: 39-43.
Harris, P.; Holley, M. R. (1972) Lead levels in cord blood. Pediatrics 49:606-608.
Harrison, R. M. (1979) Toxic metals in street and household dusts. Sci. Total Environ. 11-
89-97.
Harrison, R. M.; Laxen, D. P. H. (1978) Sink processes for tetraalkyllead compounds in the
atmosphere. Environ. Sci. Technol. 12: 1384-1392.
Harrison, G. E.; Carr, T. E. F.; Sutton, A.; Humphreys, E. R. (1969) Effect of alginate on the
absorption of lead in man. Nature (London) 224: 1115-1116.
Hart, M. H.; Smith, J. L. (1981) Effect of vitamin D and low dietary calcium on lead uptake
and retention in rats. J. Nutr. Ill: 694-698.
Hayakawa, K. (1972) Microdetermination and dynamic aspects of in vivo alkyl lead compounds:
II. studies on the dynamic aspects of alkyl lead compound! in vivo. Jpn. J. Hyg. 26-
526-535.
Heard, M. J.; Chamberlain, A. C. (1982) Effect of minerals and food on uptake of lead from the
gastrointestinal tract in humans. Hum. Toxicol. 1: 411-415.
Heard, M. J.; Chamberlain, A. C. (1983) Uptake of lead by humans and effects of minerals and
food. Sci. Total Environ. 30: 245-253.
10-80
-------
Heard, M. J.; Wells, A. C. ; Newton, D.; Chamberlain, A. C. (1979) Human uptake and metabolism
of tetra ethyl and tetra methyl lead vapour labelled with 203Pb. In: International con-
ference: management and control of heavy metals in the environment; September; London,
United Kingdom. Edinburgh, United Kingdom: CEP Consultants, Ltd.; pp. 103-108.
Hofmann, W. (1982) Dose calculations for the respiratory tract from inhaled natural radio-
active nuclides as a function of age-II: basal cell dose distributions and associated
lung cancer risk. Health Phys. 43: 31-44.
Hofmann, W.; Steinhausler; Pohl, E. (1979) Dose calculations for the respiratory tract from
inhaled natural radioactive nuclides as a function of age-I: compartmental deposition,
retention and resulting dose. Health Phys. 37: 517-532.
Holtzman, R. B. (1978) Application of radiolead to metabolic studies. In: J. 0. Nriagu, ed.
The biogeochemistry of lead in the environment; part B. biological effects. Amsterdam,
The Netherlands: Elsevier/North-Holland Biomedical Press; pp. 37-96. (Topics in environ-
mental health: v. IB).
Horiguchi, S.; Utsunomiya, T. (1973) An estimate of the body burden of lead in the healthy
Japanese population: an attempt to assume absorption and excretion of lead in the healthy
Japanese population, part 2. Osaka City Med. J. 19: 1-5.
Horiuchi, K.; Horiguchi, S.; Suekane, M. (1959) Studies on the industrial lead poisoning. 1:
Absorption, transportation, deposition and excretion of lead. 6: The lead contents in
organ-tissues of the normal Japanese. Osaka City Med. J. 5: 41-70.
Hsu, F. S.; Krook, L.; Pond, W. G.; Duncan, J. R. (1975) Interactions of dietary calcium with
toxic levels of lead and zinc in pigs. J. Nutr. 105: 112-118.
Hunter, J. M. (1977) The summer disease: an integrative model of the seasonality aspects of
childhood lead poisoning. Soc. Sci. Med. 11: 691-703.
Hunter, J. M. (1978) The summer disease: some field evidence on seasonality in childhood lead
poisoning. Soc. Sci. Med. 12: 85-94.
Hursh, J. B.; Mercer, T. T. (1970) Measurement of *1!tPb loss rate from human lungs. J. Appl.
Physiol. 28: 268-274.
Hursh, J. B.; Suomela, J. (1968) Absorption of *i:iPb from the gastrointestinal tract of man.
Acta Radiol. 7: 108-120.
Indraprasit, S.; Alexander, G. V.; Gonick, H. C. (1974) Tissue composition of major and trace
elements in uremia and hypertension. J. Chronic Dis. 27: 135-161.
International Radiological Protection Commission, Task Group on Lung Dynamics. (1966) Deposit-
ion and retention models for internal dosimetry of the human respiratory tract. Health
Phys. 12: 173-207.
James, A. C. (1978) Lung deposition of sub-micron aerosols calculated as a function of age and
breathing rate. In: National Radiological Protection Board annual research and develop-
ment report. Harwell, United Kingdom: National Radiological Protection Board; pp. 71-75.
Johnson, N. E.; Tenuta, K. (1979) Diets and lead blood levels of children who practice pica.
Environ. Res. 18: 369-376.
10-81
-------
Jugo, S. (1980) Chelatable fraction of 203Pb in blood of young and adult rats. Environ. Res.
21: 336-342.
Jugo, S. ; Maljkovic, T. ; Kostial, K. (1975a) Influence of chelating agents on the gastro-
intestinal absorption of lead. Toxicol. Appl. Pharmacol. 34: 259-263.
Jugo, S. ; Maljkovic, T. ; Kostial, K. (1975b) The effect of chelating agents on lead excretion
in rats in relation to age. Environ. Res. 10: 271-279.
Kaplan, M. L.; Peresie, H. J.; Jeffcoat, M. K. (1980) The lead content of blood and deciduous
teeth in lead-exposed beagle pups. In: Needleman, H. L., ed. Low level lead exposure: the
clinical implications of current research. New York, NY: Raven Press; pp. 221-230.
Kehoe, R. A. (1961a) The metabolism of lead in man in health and disease: the normal metabo-
lism of lead. (The Harben lectures, 1960). J. R. Inst. Public Health Hyg. 24: 81-97.
Kehoe, R. A. (1961b) The metabolism of lead in man in health and disease: the metabolism of
lead under abnormal conditions. (The Harben lectures, 1960). J. R. Inst. Public Health
Hyg. 24: 129-143.
Kehoe, R. A. (1961c) The metabolism of lead in man in health and disease: present hygienic
problems relating to the absorption of lead. (The Harben lectures, 1960). J. R Inst
Public Health Hyg. 24: 177-203.
Kehoe, R. A.; Thamann, F. (1931) The behavior of lead in the animal organism: II. tetraethvl
lead. Am. J. Hyg. 13: 478-498.
Keller, C. A.; Doherty, R. A. (1980a) Distribution and excretion of lead in young and adult
female mice. Environ. Res. 21: 217-228.
Keller, C. A.; Doherty, R. A. (1980b) Bone lead mobilization in lactating mice and lead trans-
fer to suckling offspring. Toxicol. Appl. Pharmacol. 55: 220-228.
Keller, C. A.; Doherty, R. A. (1980c) Lead and calcium distributions in blood, plasma, and
milk of the lactating mouse. J. Lab. Clin. Med. 95: 81-89.
Kello, D. ; Kostial, K. (1973) The effect of milk diet on lead metabolism in rats. Environ
Res. 6: 355-360.
Kimmel, E. C.; Fish, R. H.; Casida, J. E. (1977) Bioorganotin chemistry: metabolism of organo-
tin compounds in microsomal monooxygenase systems and in mammals. J. Agric. Food Chem
25: 1-9.
Kimmel, C. A.; Grant, L. D.; Sloan, C. S.; Gladen, B. C. (1980) Chronic low-level lead tox-
icity in the rat. Toxicol. Appl. Pharmacol. 56: 28-41.
King, E.; Conchie, A.; Hiett, D.; Milligan, B. (1979) Industrial lead absorption. Ann. Occup
Hyg. 22: 213-239.
Klaassen, C. D.; Shoeman, D. W. (1974) Biliary excretion of lead in rats, rabbits, and dogs.
Toxicol. Appl. Pharmacol. 29: 434-446.
Klauder, D. S.; Petering, H. G. (1975) Protective value of dietary copper and iron against
some toxic effects of lead in rats. Environ. Health Perspect. 12: 77-80.
10-82
-------
Klauder, D. S.; Petering, H. G. (1977) Anemia of lead intoxication: a role for copper. J.
Nutr. 107: 1779-1785.
Klauder, D. S.; Murthy, L; Petering, H. G, (1973) Effect of dietary intake of lead acetate on
copper metabolism in male rats. In: Hemphill, D. 0., ed. Trace substances in environ-
mental health - VI: [proceedings of University of Missouri's 6th annual conference on
trace substances in environmental health]; June 1972; Columbia, MO. Columbia, MO: Uni-
versity of Missouri-Columbia; pp. 131-136.
Klein, A. W.; Koch, T. R. (1981) Lead accumulations in brain, blood, and liver after low
dosing of neonatal rats. Arch. Toxicol. 47: 257-262.
Klein, M.; Namer, R.; Harpur, E.; Corbin, R. (1970) Earthenware containers as a source of
fatal lead poisoning: case study and public-health considerations. N. Engl. 0. Med. 283:
669-672.
Kneip, T. J.; Mallon, R. P.; Harley, N. H. (1983) Biokinetic modelling for mammalian lead
metabolism. Neurotoxicology 4: 189-192.
Kochen, J. A.; Greener, Y. (1973) Levels of lead in blood and hematocrit: implications for the
evaluation of the newborn and anemic patient. Pediatr. Res. 7: 937-944.
Kostial, K.; Kello, D. (1979) Bioavailability of lead in rats fed "human" diets. Bull.
Environ. Contam. Toxicol. 21: 312-314.
Kostial, K.; Momcilovid, B. (1974) Transport of lead 203 and calcium 47 from mother to off-
spring. Arch. Environ. Health 29: 28-30.
Kostial, K.; Simonovic, J.; Pisonid, M. (1971) Lead absorption from the intestine in newborn
rats. Nature (London) 233: 564.
Kostial, K.; Kello, D.; Jugo, S.; Rabar, I.; Maljkovid, T. (1978) Influence of age on metal
metabolism and toxicity. Environ. Health Perspect. 25: 81-86.
Ku, Y.; Alvarez, G. H.; Mahaffey, K. R. (1978) Comparative effects of feeding lead acetate and
phospholipid-bound lead on blood and tissue lead concentrations in young and adult rats.
Bull. Environ. Contam. Toxicol. 20: 561-567.
Laug, E. P.; Kunze, F. M. (1948) The penetration of lead through the skin. J. Ind. Hyg.
Toxicol. 30: 256-259.
Lepow, M. L.; Bruckman, L.; Rubino, R. A.; Markowitz, S.; Gillette, M.; Kapish, J. (1974) Role
of airborne lead in increased body burden of lead in Hartford children. Environ. Health
Perspect. 7: 99-102.
Lloyd, R. D.; Mays, C. W.; Atherton, D. R.; Bruenger, F. W. (1975) 210Pb studies in beagles.
Health Phys. 28: 575-583.
Lorenzo, A. V.; Gewirtz, M.; Maher, C.; Davidowski, L. I. (1977) The equilibration of lead
between blood and milk of lactating rabbits. Life Sci. 21: 1679-1683.
Mackie, A. C.; Stephens, R.; Townsend, A.; Waldron, H. A. (1977) Tooth lead levels in
Birmingham children. Arch. Environ. Health 32: 178-185.
10-83
-------
Mahaffey, K. R.; Michaelson, I. A. (1980) The interaction between lead and nutrition. In:
Needleman, H. L., ed. Low level lead exposure: the clinical implications of current
research. New York, NY: Raven Press; pp. 159-200.
Mahaffey, K. R.; Goyer, R.; Haseman, J. K. (1973) Dose-response to lead ingestion in rats fed
low dietary calcium. J. Lab. Clin. Med. 82: 92-100.
Mahaffey, K. R.; Treloar, S.; Banks, T. A.; Peacock, B. J.; Parekh, L. E. (1976) Differences
in dietary intake of calcium, phosphorus and iron of children having normal and elevated
blood lead concentrations. J. Nutr. 106(7): xxx.
Mahaffey, K. R.; Annest, J. L.; Barbano, H. E.; Murphy, R. S. (1979) Preliminary analysis of
blood lead concentrations for children and adults: HANES II, 1976-1978. In: Hemphill, D
D. , ed. Trace substances in environmental health - XIII: [proceedings of University of
Missouri's 13th annual conference on trace substances in environmental health]; June-
Columbia, MO. Columbia, MO: University of Missouri-Columbia; pp. 37-51. ' '
Mahaffey Six, K.; Goyer, R. A. (1970) Experimental enhancement of lead toxicity by low dietarx/
calcium. J. Lab. Clin. Med. 76: 933-942. y
Mahaffey-Six, K.; Goyer, R. A. (1972) The influence of iron deficiency on tissue content and
toxicity of ingested lead in the rat. J. Lab. Clin. Med. 79: 128-136.
Manton, W. I.; Cook, J. D. (1979) Lead content of cerebrospinal fluid other tissue in amyo-
trophic lateral sclerosis (ALS). Neurology 29: 611-612.
Manton, W. I.; Cook, J. D. (1984) High accuracy (stable isotope dilution) measurements of
lead in serum and cerebrospinal fluid. Br. J. Ind. Med. 41: 313-319.
Manton, W. I.; Malloy, C. R. (1983) Distribution of lead in body fluids after inqestion nf
soft solder. Br. J. Ind. Med. 40: 51-57. T
Marcus, A. H. (1979) The body burden of lead: comparison of mathematical models for accumula
tion. Environ. Res. 19: 79-90. a
Marcus, S. M. (1982) Experience with D-penicillamine in treating lead poisoning Vet Hum
Toxicol. 24: 18-20. ' m'
Marcus, A. H. (1985a) Multicompartment kinetic models for lead: I. bone diffusion models for
long-term retention. Environ. Res. 36: 441-458.
Marcus, A. H. (1985b) Multicompartment kinetic models for lead: II. linear kinetics and vari-
able absorption in humans without excessive lead exposures. Environ. Res. 36: 459-472
Marcus, A. H. (1985c) Multicompartment kinetic model for lead: III. lead in blood plasma and
erythrocytes. Environ. Res. 36: 473-489.
Marcus, A. H. (1985d) Testing alternative nonlinear kinetic models in compartmental analysis
In: Eisenfeld, J.; Delisi, C., eds. Mathematics and computers in biomedical applications'
Amsterdam, The Netherlands: Elsevier; pp. 259-267.
Markowitz, M. E. ; Rosen, J. F. (1981) Zinc (Zn) and copper (Cu) metabolism in CaNa2 EDTA-
treated children with plumbism. Pediatr. Res. 15: 635.
10-84
-------
McLachlin, J. R.; Goyer, R. A.; Cherian, M. G. (1980) Formation of lead-induced inclusion
bodies in primary rat kidney epithelial cell cultures: effect of actinomycin D and
cycloheximide. Toxicol. Appl. Pharmacol. 56: 418-431.
Mehani, S. (1966) Lead retention by the lungs of lead-exposed workers. Ann. Occup. Hyg. 9:
165-171.
Momcilovic, B. (1978) The effect of maternal dose on lead retention in suckling rats. Arch.
Environ. Health 33: 115-117.
Momcilovic, B. ; Kostial, K. (1974) Kinetics of lead retention and distribution in suckling and
adult rats. Environ. Res. 8: 214-220.
Moore, M. R.; Meredith, P. A.; Campbell, B. C.; Watson, W. S. (1979) The gastrointestinal ab-
sorption of lead 203 chloride in man. In: Hemphill, D. D., ed. Trace substances in
environmental health- XIII: [proceedings of University of Missouri's 13th annual con-
ference on trace substances in environmental health]; June; Columbia, MO. Columbia, MO:
University of Missouri-Columbia; pp. 368-373.
Moore, M. R.; Meredith, P. A.; Watson, W. S.; Sumner, D. J.; Taylor, M. K.; Goldberg, A.
(1980) The percutaneous absorption of lead-203 in humans from cosmetic preparations
containing lead acetate, as assessed by whole-body counting and other techniques. Food
Cosmet. Toxicol. 18: 399-405.
Morgan, A.; Holmes, A. (1978) The fate of lead in petrol-engine exhaust particulates inhaled
by the rat. Environ. Res. 15: 44-56.
Morgan, A.; Holmes, A.; Evans, J. C. (1977) Retention, distribution, and excretion of lead by
the rat after intravenous injection. Br. J. Ind. Med. 34: 37-42.
Morrison, J. N.; Quarterman, J.; Humphries, W. R. (1977) The effect of dietary calcium and
phosphate on lead poisoning in lambs. J. Comp. Pathol. 87: 417-429.
Morrow, P. E.; Beiter, H.; Amato, F.; Gibb, F. R. (1980) Pulmonary retention of lead: an ex-
perimental study in man. Environ. Res. 21: 373-384.
Mortensen, R. A. (1942) The absorption of lead tetraethyl with radioactive lead as indicator.
J. Ind. Hyg. Toxicol. 24: 285-288.
Munro, I. C.; Willes, R. F.; Truelove, J. F. (1975) Absorption and tissue distribution of in-
organic lead in the developing infant monkey (Macaca irus). Toxicol. Appl. Pharmacol. 33:
128-129.
MykkMnen, H. M.; Wasserman, R. H. (1981) Gastrointestinal absorption of lead (203Pb) in
chicks: influence of lead, calcium, and age. J. Nutr. Ill: 1757-1765.
MykkSnen, H. M.; Dickerson, J. W. T.; Lancaster, M. C. (1979) Effect of age on the tissue
distribution of lead in the rat. Toxicol. Appl. Pharmacol. 51: 447-454.
National Academy of Sciences. National Research Council. (1974) Recommended dietary allow-
ances. 8th ed. Washington, DC: National Academy of Sciences.
National Academy of Sciences. National Research Council. (1976) Recommendations for the pre-
vention of lead poisoning in children. Washington, DC: National Academy of Sciences.
Available from: NTIS, Springfield, VA; PB-257645.
10-85
-------
National Academy of Sciences, Committee on Lead in the Human Environment. (1980) Lead in the
human environment. Washington, DC: National Academy of Sciences.
Needleman, H. L.; Shapiro, I. M. (1974) Dentine lead levels in asymptomatic Philadelphia
school children: subclinical exposure in high and low risk groups. Environ. Health
Perspect. 7: 27-31.
Nielsen, T.; Jensen, K. A.; Grandjean, P. (1978) Organic lead in normal human brains. Nature
(London) 274: 602-603.
Niklowitz, W. J.; Mandybur, T. I. (1975) Neurofibrillary changes following childhood lead
encephalopathy: case report. J. Neuropathol. Exp. Neurol. 34: 445-455.
Nozaki, K. (1966) Method for studies on inhaled particles in human respiratory system and
retention of lead fume. Ind. Health 4: 118-128.
Nygaard, S. P.; Ottosen, J. ; Hansen, J. C. (1977) Whole-blood lead concentration in Danes:
relation to age and environment. Dan. Med. Bull. 24: 49-51.
O'Flaherty, E. J.; Hammond, P. B.; Lerner, S. I. (1982) Dependence of apparent blood lead
half-life on the length of previous lead exposure in humans. Fundam. Appl. Toxicol ?
49-54.
Okazaki, H.; Aronson, S. M.; DiMaio, D. J.; Olvera, J. E. (1963) Acute lead encephalopathy of
childhood: histologic and chemical studies, with particular reference to angiopathic
aspects. In: Yahr, M. D., ed. Transactions of the American Neurological Association,
1963: v. 88. New York, NY: Springer Publishing Company, Inc.; pp. 248-250.
Ong, C. N. ; Lee, W. R. (1980a) Distribution of lead-203 in human peripheral blood in vitro
Br. J. Ind. Med. 37: 78-84. -'
Ong, C. N.; Lee, W. R. (1980b) High affinity of lead for fetal haemoglobin. Br. J. Ind. Med
37: 292-298.
Ong, C. N.; Lee, W. R. (1980c) Interaction of calcium and lead in human erythrocytes.Br J
Ind. Med. 37: 70-77.
Owen, G.; Lippman, G. (1977) Nutritional status of infants and young children: U.S.A. Pediatr
Clin. North Am. 24: 211-227.
Piomelli, S. ; Rosen, J. F.; Chisolm, J. J., Jr.; Graef, J. W. (1984) Management of childhood
lead poisoning. J. Pediatr. (St. Louis). 105: 523-532.
Pott, F.; Brockhaus, A. (1971) Vergleich der enteralen und pulmonalen Resorptionsquote von
Bleiverbindungen [Comparison of the enteral and pulmonary absorption rates of lead com-
pounds]. Zentralbl. Bakteriol. Parasitenkd. Infektionskrankh. Hyg. Abt. 1 Orig. Reihe B.
Pounds, J. G.; Marlar, R. J. ; Allen, J. R. (1978) Metabolism of lead-210 in juvenile and adult
rhesus monkeys Macaca mulatta. Bull. Environ. Contam. Toxicol. 19: 684-691.
Pounds, J. G.; Wright, R.; Kodell, R. L. (1982) Cellular metabolism of lead: a kinetic anal-
ysis in the isolated rat hepatocyte. Toxicol. Appl. Pharmacol. 66: 88-101.
Prerovska, I.; Teisinger, J. (1970) Excretion of lead and its biological activity several
years after termination of exposure. Br. J. Ind. Med. 27: 352-355.
10-86
-------
Prpic-Majic, D. ; Mueller, P. K.; Beritic, T.; Stanley, R.; Twiss, S. (1973) Delta-aminole-
vulinic acid dehydratase activity, lead blood levels, and the reticulocyte count. In:
Barth, D.; Berlin, A.; Engel, R.; Recht, P.; Smeets, J., eds. Environmental health as-
pects of lead: proceedings, international symposium; October 1972; Amsterdam, The Nether-
lands. Luxembourg: Commission of the European Communities; pp. 211-220.
Quarterman, J.; Morrison, J. N. (1975) The effects of dietary calcium and phosphorus on the
retention and excretion of lead in rats. Br. J. Nutr. 34: 351-362.
Quarterman, J.; Morrison, E. (1978) The effect of age on the absorption and excretion of lead.
Environ. Res. 17: 78-83.
Quarterman, J.; Morrison, J. N.; Carey, L. F. (1973) The influence of dietary calcium and
phosphate on lead metabolism. In: Hemphill, D. D., ed. Trace substances in environmental
health-VII: [proceedings of University of Missouri's 7th annual conference on trace
substances in environment health]; June; Columbia, MO. Columbia, MO: University of
Missouri-Columbia; pp. 347-352.
Quarterman, J.; Morrison, J. N.; Humphries, W. R. (1978a) The influence of high dietary cal-
cium and phosphate on lead uptake and release. Environ. Res. 17: 60-67.
Quarterman, J.; Morrison, E.; Morrison, J. N.; Humphries, W. R. (1978b) Dietary protein and
lead retention. Environ. Res. 17: 68-77.
Rabinowitz, M. B.; Wetherill, G. W.; Kopple, J. D. (1973) Lead metabolism in the normal human:
stable isotope studies. Science (Washington, DC) 182: 725-727.
Rabinowitz, M.; Wetherill, G. W.; Kopple, J. D. (1974) Studies of human lead metabolism by use
of stable isotope tracers. Environ. Health Perspect. 7: 145-153.
Rabinowitz, M.; Wetherill, G.; Kopple, J. (1976) Delayed appearance of tracer lead in facial
hair. Arch. Environ. Health 31: 220-223.
Rabinowitz, M. B.; Wetherill, G. W.; Kopple, J. 0. (1976) Kinetic analysis of lead metabolism
in healthy humans. J. Clin. Invest. 58: 260-270.
Rabinowitz, M. B.; Wetherill, G. W.; Kopple, J. D. (1977) Magnitude of lead intake from res-
piration by normal man. J. Lab. Clin. Med. 90: 238-248.
Rabinowitz, M. B.; Kopple, J. D.; Wetherill, G. W. (1980) Effect of food intake and fasting on
gastrointestinal lead absorption in humans. Am. J. Clin. Nutr. 33: 1784-1788.
Rabinowitz, M. B.; Needleman, H. L. (1982) Temporal trends in the lead concentrations of
umbilical cord blood. Science (Washington, DC) 216: 1429-1431.
Rabinowitz, M.; Leviton, A.; Needleman, H. (1984) Variability of blood lead concentrations
during infancy. Arch. Environ. Health 39: 74-77.
Rader, J. I.; Peeler, J. T.; Mahaffey, K. R. (1981a) Comparative toxicity and tissue dis-
tribution of lead acetate in weanling and adult rats. Environ. Health Perspect. 42:
187-195.
10-87
-------
Rader, J. I.; Celesk, E. M.; Peeler, J. T.; Mahaffey, K. R. (1981b) Comparative toxicity to
weanling and adult rats of lead acetate in water. In: Hemphill, D. D., ed. Trace sub-
stances in environmental health - XV: [proceedings of University of Missouri's 15th
annual conference on trace substances in environmental health]; June; Columbia, MO.
Columbia, MO: University of Missouri-Columbia; pp. 131-143.
Raghavan, S. R. V.; Gonick, H. C. (1977) Isolation of low-molecular-weight lead-binding
protein from human erythrocytes. Proc. Soc. Exp. Biol. Med. 155: 164-167.
Raghavan, S. R. V.; Culver, B. D.; Gonick, H. C. (1980) Erythrocyte lead-binding protein after
occupational exposure. I. Relationship to lead toxicity. Environ. Res. 22: 264-270.
Raghavan, S. R. V.; Culver, B. D.; Gonick, H. C. (1981) Erythrocyte lead-binding protein after
occupational exposure. II. Influence on lead inhibition of membrane Na , K adenosine-
triphosphatase. J. Toxicol. Environ. Health 7: 561-568.
Rastogi, S. C. ; Clausen, J. (1976) Absorption of lead through the skin. Toxicology 6-
371-376.
Rendall, R. E. G.; Baily, P.; Soskolne, C. L. (1975) The effect of particle size on absorption
of inhaled lead. Am. Ind. Hyg. Assoc. J. 36: 207-213.
Rosen, J. F.; Markowitz, M. E. (1980) D-penicillamine: its actions on lead transport in bone
organ culture. Pediatr. Res. 14: 330-335.
Rosen, J. F.; Zarate-Salvador, C.; Trinidad, E. E. (1974) Plasma lead levels in normal and
lead-intoxicated children. J. Pediatr. (St. Louis) 84: 45-48.
Rosen, J. F. ; Chesney, R. W.; Hamstra, A. J.; DeLuca, H. F.; Mahaffey, K. R. (1980) Reduction
in 1,25-dihydroxyvitamin D in children with increased lead absorption. N. Engl. J Med
302: 1128-1131.
Rosen, J. F.; Chesney, R. W.; Hamstra, A. J.; DeLuca, H. F.; Mahaffey, K. R. (1981) Reduction
in 1,25-dihydroxyvitamin D in children with increased lead absorption. In: Brown, S. S.-
Davis, D. S. , eds. Organ-directed toxicity: chemical indices and mechanisms. New York'
NY: Pergamon Press; pp. 91-95. '
Ryu, J. E.; Ziegler, E. E.; Fomon, S. J. (1978) Maternal lead exposure and blood lead concentr-
ation in infancy. J. Pediatr. (St. Louis) 93: 476-478.
Saenger, P.; Rosen, J. F.; Markowitz, M. E. (1982) Diagnostic significance of edetate disodium
calcium testing in children with increased lead absorption. Am. J. Dis. Child. 136-
312-315.
Sartor, F. A.; Rondia, D. (1981) Setting legislative norms for environmental lead exposure:
results of an epidemiological survey in the east of Belgium. Toxicol. Lett. 7: 251-257.
Scanlon, J. (1971) Umbilical cord blood lead concentration. Am. J. Dis. Child.121: 325-326.
Schroeder, H. A.; Tipton, I. H. (1968) The human body burden of lead. Arch. Environ. Healthl
7: 965978.
10-88
-------
Shapiro, I. M. ; Burke, A.; Mitchell, G.; Bloch, P. (1978) X-ray fluorescence analysis of lead
in teeth of urban children ui situ: correlation between the tooth lead level and the
concentration of blood lead and free erythroporphyrins. Environ. Res. 17: 46-52.
Shelling, D. H. (1932) Effect of dietary calcium and phosphorus on toxicity of lead in therat:
rationale of phosphate therapy. Proc. Soc. Exp. Biol. Med. 30: 248-254.
Sherlock, J.; Smart, G.; Forbes, G. I.; Moore, M. R. ; Patterson, W. J.; Richards, W. N. ;
Wilson, T. S. (1982) Assessment of lead intakes and dose-response for a population in Ayr
exposed to a plumbsolvent water supply. Hum. Toxicol. 1: 115-122.
Singh, N.; Donovan, C. M.; Hanshaw, J. B. (1978) Neonatal lead intoxication in a prenatally
exposed infant. J. Pediatr. (St. Louis) 93: 1019-1021.
Smith, C. M.; DeLuca, H. F.; Tanaka, Y.; Mahaffey, K. R. (1978) Stimulation of lead absorption
by vitamin D administration. J. Nutr. 108: 843-847.
Sobel, A. E.; Gawron, 0.; Kramer, B. (1938) Influence of vitamin D in experimentallead
poisoning. Proc. Soc. Exp. Biol. Med. 38: 433-435.
Sobel, A. E.; Yuska, H.; Peters, D. 0.; Kramer, B. (1940) The biochemical behavior of lead: I.
influence of calcium, phosphorus, and vitamin D on lead in blood and bone. J. Biol. Chem.
132: 239-265. Reprinted (1981) in Nutr. Rev. 39: 374-377.
Sorrel 1, M.; Rosen, J. F.; Roginsky, M. (1977) Interactions of lead, calcium, vitamin D, and
nutrition in lead-burdened children. Arch. Environ. Health 32: 160-164.
Steenhout, A.; Pourtois, M. (1981) Lead accumulation in teeth as a function of age with
different exposures. Br. J. Ind. Med. 38: 297-303.
Stephens, R. ; Waldron, H. A. (1975) The influence of milk and related dietary constituents on
lead metabolism. Food Cosmet. Toxicol. 13: 555-563.
Stevens, C. D.; Feldhake, C. J.; Kehoe, R. A. (1960) Isolation of triethyllead ion from liver
after inhalation of tetraethyllead. J. Pharmacol. Exp. Ther. 128: 90-94.
Stowe, H. D.; Goyer, R. A.; Krigman, M. M.; Wilson, M. ; Gates, M. (1973) Experimental oral
lead toxicity in young dogs: clinical and morphologic effects. Arch. Pathol. 95: 106-116.
Stuik, E. J. (1974) Biological response of male and female volunteers to inorganic lead. Int.
Arch. Arbeitsmed. 33: 83-97.
Thompson, J. A. (1971) Balance between intake and output of lead in normal individuals. Br.
J. Ind. Med. 28: 189-194.
Tola, S.; Hernberg, S.; Asp, S.; Nikkanen, J. (1973) Parameters indicative of absorption and
biological effect in new lead exposure: a prospective study. Br. J. Ind. Med. 30:
134-141.
10-89
-------
Trotter, M. ; Hixon, B. B. (1974) Sequential changes in weight, density, and percentage ash
weight of human skeletons from an early fetal period through old age. Anat. Rec. 179:
1-18.
U. S. Centers for Disease Control. (1978) Preventing lead poisoning in young children: a state-
ment by the Center for Disease Control. J. Pediatr. (St. Louis) 93: 709-720.
U. S. Environmental Protection Agency. (1977) Air quality criteria for lead. Research Triangle
Park, NC: Health Effects Research Lab, Criteria and Special Studies Office; EPA report
no. EPA-600/8-77-017. Available from: NTIS, Springfield, VA; PB-280411.
United Kingdom Central Directorate on Environmental Pollution. (1982) The Glasgow duplicate
diet study (1979/1980): a joint survey for the Department of the Environment and the
Ministry of Agriculture, Fisheries and Food. London, United Kingdom: Her Majesty's
Stationery Office; pollution report no. 11.
Victery, W.; Soifer, N. E. ; Weiss, J. S. ; Vander, A. J. (1981) Acute effects of lead on the
renal handling of zinc in dogs. Toxicol. Appl. Pharmacol. 61: 358-367.
Vitale, L. F.; Joselow, M. M.; Wedeen, R. P.; Pawlow, M. (1975) Blood leadan inadequate
measure of occupational exposure. J. Occup. Med. 17: 155-156.
Watson, W. S.; Hume, R.; Moore, M. R. (1980) Oral absorption of lead and iron. Lancet 2(8188)-
236-237.
Wedeen, R. P.; Maesaka, J. K.; Weiner, B.; Lipat, G. A.; Lyons, M. M.; Vitale, L. F.; Joselow
M. M. (1975) Occupational lead nephropathy. Am. J. Med. 59: 630-641.
Wedeen, R. P.; Mallik, D. K.; Batuman, V. (1979) Detection and treatment of occupational lead
nephropathy. Arch. Intern. Med. 139: 53-57.
Williams, M. K.; King, E.; Walford, J. (1969) An investigation of lead absorption in an elec-
tric accumulator factory with the use of personal samplers. Br. J. Ind. Med. 26: 202-216.
Willoughby, R. A.; Thirapatsakun, T.; McSherry, B. J. (1972) Influence of rations low 1n
calcium and phosphorus on blood and tissue lead concentrations in the horse. Am. J. Vet
Res. 33: 1165-1173.
Winneke, G.; Brockhaus, A.; Kramer, U.; Ewers, U.; Kujanek, G.; Lechner, H.; Janke, W. (1981)
Neuropsychological comparison of children with different tooth-lead levels: preliminary
report. In: International conference: heavy metals in the environment; Amsterdam, The
Netherlands. Edinburgh, United Kingdom: CEP Consultants, Ltd.; pp. 553-556.
World Health Organization, United Nations Environmental Programme. (1977) Lead. Geneva,
Switzerland: World Health Organization. (Environmental health criteria 3).
Yamamura, Y.; Takakura, J.; Hirayama, F.; Yamauchi, H.; Yoshida, M. (1975) Tetraethyl lead
poisoning caused by cleaning work in the aviation fuel tank. Jpn. J. Ind. Health 17-
223-235.
Yip, R. ; Norris, T. N.; Anderson, A. S. (1981) Iron status of children with elevated blood
lead concentrations. J. Pediatr. (St. Louis) 98: 922-925.
10-90
-------
Ziegler, E. E.; Edwards, B. B.; Jensen, R. L.; Mahaffey, K. R.; Fomon, S. 0. (1978) Absorption
and retention of lead by infants. Pediatr. Res. 12: 29-34.
Zielhuis, R. L.; del Castilho, P.; Herber, R. F. M.; Wibowo, A. A. E. (1978) Levels of lead
and other metals in human blood: suggestive relationships, determining factors. Environ.
Health Perspect. 25: 103-109.
10-91
-------
11. ASSESSMENT OF LEAD EXPOSURES AND ABSORPTION IN HUMAN POPULATIONS
11.1 INTRODUCTION
This chapter describes effects on internal body burdens of lead in human populations
resulting from exposure to lead in their environment. Particular attention is paid to changes
in indices of internal lead exposure that follow changes in external lead exposures. Blood
lead is the main index of internal lead exposure discussed here, although other indices, such
as levels of lead in teeth and bone, are also briefly discussed.
The following terms and definitions are used in this chapter. Sources of lead are those
components of the environment (e.g., gasoline combustion, smelters) from which significant
quantities of lead are released into various environmental media of exposure. Environmental
media are routes by which humans become exposed to lead (e.g., air, soil, food, water, dust).
External exposures are levels at which lead is present in any or all of the environmental
media. Internal exposures are amounts of lead present in various body tissues and fluids.
The present chapter is structured to achieve the following four main objectives:
(1) Elucidation of patterns of internal lead exposures in U.S. populations and
identification of important demographic covariates.
(2) Characterization of relationships between external and internal exposures to
lead by exposure medium (air, food, water or dust).
(3) Identification of specific sources of lead which result in increased internal
exposure levels.
(4) Estimation of the relative contributions of various sources of lead in the
environment to total internal 'exposure as indexed by blood lead level.
The existing scientific literature must be examined in light of the investigators' own
objectives and the quality of the scientific investigations performed. Although all studies
need to be evaluated in regard to their methodology, the more quantitative studies are evalu-
ated here in greater depth. A discussion of the main types of methodological points con-
sidered in such evaluations is presented in Section 11.2.
Patterns of internal exposure to lead in human populations are discussed in Section 11.3.
This begins with a brief examination of the historical record of internal lead exposure in
human populations. These data serve as a backdrop against which recent U.S. levels can be
contrasted and define the relative magnitude of external lead exposures in the past and
present. The contrast is structured as follows: historical data, recent data from popula-
tions thought to be isolated from urbanized cultures, and then U.S. populations showing
various degrees of urbanization and industrialization.
11-1
-------
The statistical treatment of distributions of blood lead levels in human populations is
the next topic discussed. As part of that discussion, the empirical characteristics of blood
lead distributions in well-defined homogeneous populations are denoted. Important issues
addressed include the proper choice of estimators of central tendency and dispersion, estima-
tors of percentile values and the potential influence of errors in measurement on statistical
estimation involving blood lead data.
Then recent patterns of internal exposure in U.S. and other populations showing change in
blood lead levels are discussed in detail. Estimates of internal lead exposure and identifi-
cation of demographic covariates are made. Studies examining the recent past for evidence of
change in internal exposure levels are presented. Next is an examination of extensive evi-
dence which points towards gasoline lead being an important determinant of changes in blood
lead level associated with exposures to airborne lead of populations in the United States and
elsewhere.
Section 11.4 focuses on general relationships between external exposures and levels of
internal exposure. The distribution of lead in man is diagramatically depicted by the compo-
nent model shown in Figure 11-1. Of particular importance for this document is the relation-
ship between lead in air and lead in blood. If lead in air were the only medium of exposure,
then the interpretation of a statistical relationship between lead in air and lead in blood
would be relatively simple. However, this is not the case. Lead is present in a number of
environmental media, as described in Chapter 7 and summarized in Figure 11-1. There are rela-
tionships between lead levels in air and lead concentrations in food, soil, dust, and water.
As shown in Chapters 6, 7, and 8, lead emitted into the atmosphere ultimately comes back to
contaminate the earth. However, only limited data are currently available that provide a
quantitative estimate of the magnitude of this secondary lead exposure. The implication is
that an analysis involving estimated lead levels in all environmental media may tend to under-
estimate the relationship between lead in blood and lead in air.
The discussion of relationships between external exposure and internal absorption com-
mences with air lead exposures. Both experimental and epidemiological studies are discussed.
Several studies are identified as being of greatest importance in determining the quantitative
relationship between lead in blood and lead in air. The form of the relationship between
blood lead and air lead is of particular interest and importance. After discussion of air
lead versus blood lead relationships, the chapter next discusses the relationship of blood
lead to atmospheric lead found in other environmental media. Section 11.5 describes studies
of specific lead exposure situations useful in identifying specific environmental sources of
lead that contribute to elevated body burdens of lead. The chapter concludes with a summary
of key information and conclusions derived from the scientific evidence reviewed.
11-2
-------
I CRUSTAL \ I PAINT 1
I WEATHERING] I J
SURFACE AND
GROUND WATER
INDUSTRIAL
EMISSIONS
FECES URINE
Figure 11-1. Pathways of lead from the environment to and within man.
11-3
-------
11.2 METHODOLOGICAL CONSIDERATIONS
11.2.1 Analytical Problems
Internal lead exposure levels in human populations have been estimated by analyses of a
variety of biological tissue matrices (e.g., blood, teeth, bone, and hair). Lead levels in
each of these matrices have particular biological meanings with regard to external exposure
status; these relationships are discussed in Chapter 10. The principal internal exposure
index discussed in this chapter is blood lead concentration. Blood lead concentrations are
most reflective of recent exposure to lead and bear a consistent relationship to levels of
lead in the external environment if the latter have been stable. Blood lead levels are vari-
ously reported as |jg/100 g, ug/100 ml, ng/dl, ppm, ppb, and nmol/1. The first four measures
are roughly equivalent, whereas ppb values are simply divisible by 1000 to be equivalent.
Actually there is a small, but not meaningful, difference in blood lead levels reported on a
per volume versus per weight difference. The difference results from the density of blood
being slightly greater than 1 g/ml. For the purposes of this chapter, data reported on a
weight or volume basis are considered equal. On the other hand, blood lead data reported on a
|jmol/l basis must be multiplied by 20.72 to get the equivalent pg/dl value. Data reported
originally as umol/1 in studies reviewed here are converted to ug/dl in this chapter.
As discussed in Chapter 9, the measurement of lead in blood has been accomplished via a
succession of analytical procedures over the years. The first reliable analytical methods
available were wet chemistry procedures, succeeded by increasingly automated instrumental
procedures. With these changes in technology there has been increasing recognition of the
importance of controlling for contamination in the sampling and analytical procedures. These
advances, as well as institution of external quality control programs, have resulted in
markedly improved analytical results. Data summarized in Chapter 9 show that a generalized
improvement in analytical results across many laboratories occurred during Federal Fiscal
Years 1977-1979. No further marked improvement was seen during Federal Fiscal Years
1979-1981.
Because of interest in being able to attribute specific proportions of blood lead as
coming from specific environmental sources, isotopic lead determinations in blood have become
an important analytic technique. As difficult as it is to determine blood lead levels accu-
rately, the achievement of accurate lead isotopic determinations is even more difficult.
Experience gained from the isotopic lead experiment (ILE) in Italy (reviewed in detail in
Section 11.3.6.2.1) has indicated that extremely aggressive quality control and contamination
control programs must be implemented to achieve acceptable results. With proper procedures,
meaningful differences on the order of a single nanogram are achievable.
11-4
-------
11.2.2 Statistical Approaches
Many studies have summarized the distribution of lead levels in humans. These studies
usually report measures of central tendency (means) and dispersion (variances). In this chap-
ter, the term "mean" refers to the arithmetic mean unless stated otherwise. This measure is
always an estimate of the average value, but it estimates the center of the distribution (50th
percentile) only for symmetric distributions. Many authors provide geometric means, which
estimate the center of the distribution if the distribution is lognormal. Geometric means are
influenced less by unusually large values than are arithmetic means. A complete discussion of
the lognormal distribution is given by Aitchison and Brown (1966), including formulas for con-
verting from arithmetic to geometric means.
Most studies also give sample variances or standard deviations in addition to the means.
If geometric means are given, then the corresponding measure of dispersion is the geometric
standard deviation. Aitchison and Brown (1966) give formulas for the geometric standard devi-
ation and, also, explain how to estimate percentiles and construct confidence intervals. All
of the measures of dispersion actually include three sources of variation: population varia-
tion, measurement variation, and variation due to sampling error. Values for these components
are needed in order to evaluate a study correctly. There are also sources of variation
related to the inclusion of predictive variables in the model, or their exclusion. Such vari-
ables include different lead uptakes attributable to exposure to lead in dust, soil, food,
water, paint in deteriorated housing, and other pathways. If included in the model, the
remaining sources of variation are due to unmeasured differences in intrinsic metabolism and
behavior. It has been the general goal in this chapter to include all attributable sources of
variation, thus reducing the estimates of variability to biological differences, uncertainties
in exposure, and measurement variations that cannot be further attributed. We recognize that
if only air lead exposure is controlled, then there will be additional variation in blood lead
response due to imperfectly controlled covariation of lead exposure from related pathways.
This additional variation can be dealt with in practice by use of a larger geometric standard
deviation.
A separate issue is the form of the distribution of blood lead values. Although the nor-
mal and lognormal distributions are commonly used, there are many other possible distribu-
tions. The form is important for two reasons: 1) it determines which is more appropriate,
the arithmetic or geometric mean, and 2) it determines estimates of the fraction of a popula-
tion exceeding given internal lead levels under various external exposures. Both of these
questions arise in the discussion of the distribution of human blood lead levels and are of
importance, ultimately, for deriving a rationale for standard-setting purposes.
11-5
-------
11.2.3 Confounding of Relevant Variables
Failure to include relevant variables is the most serious difficulty in evaluating stud-
ies on lead in human populations. This usually occurs when the blood lead response is wholly
attributed to some observed variable, e.g., the lead concentration in air, dust, or water.
Typical confounders for air lead include the following: (1) inhalation exposures not captured
by stationary air lead monitors, particularly those that occur from personal exposure to
leaded gasoline or its combustion products; (2) noninhalation exposures to air lead not cap-
tured by stationary monitors, e.g., ingestion of food products contaminated by lead fallout,
leaded dust, and soil; (3) ingestion of lead in water and food that is inadvertantly associ-
ated with air lead exposure. Socioeconomic factors may be important here also. See
Brunekreef (1984) and Snee (1982b,c) for additional comments.
Air lead concentrations are typically highest in urban centers where the concentration of
motor vehicles is greatest. (Communities with lead smelters are an exception). Suburban and
rural areas have much lower air lead concentrations. However, suburban and rural residents
may spend more time in motor vehicles due to longer trips to work, school, and shopping.
There is some reason to believe that higher lead concentrations may be found near and inside
automobiles (see Spengler et al., (1984), Section 11.3.6.2.1), thus offsetting the decreased
ambient air lead concentrations measured by stationary monitors in non-urban areas. Un-
fortunately, there is no way at this time to separate the response to average ambient air lead
levels from variations in personal lead exposure patterns.
Children are known to ingest quantities of dust and soil by normal hand-mouth contact. In
studies in which dust lead concentrations or hand lead quantities are measured, their contri-
bution is very large -- usually much larger than the lead intake by direct inhalation. In
smelter communities all of these variables -- ambient air lead, dust lead, soil lead, and lead
on children's hands -- are likely to be high. It may then be difficult to separate the contri-
butions of each of these components, and if any one is not measured, then its influence on
blood lead may be attributed to the other variables. This may cause little difficulty when in
fact there is a single source for all exposure pathways, but positive confounding may cause
difficulty in extrapolating the relationship to situations in which air and dust lead are less
strongly coupled. Similarly, the particle size distribution may change with distance from the
source (smelter, highway, etc.) and particle size is known to affect the fraction of lead ab-
sorbed by the lungs. However, air and dust lead concentrations also decrease with distance
from the source, thus leading to potential confounding of concentration and size effects.
This may be a factor in some smelter studies, e.g., the Silver Valley, Idaho, study discussed
later.
11-6
-------
Socioeconomic status (SES), sex, age, and race are also confounded with air lead. Lower
SES populations tend to be found in areas with high air lead concentration such as urban cen-
ters and smelter communities. There may also be systematic SES differences in use of lead-
soldered food and beverage cans and in exposure to food products with high lead content and in
personal and household cleanliness, as well. The latter is important because dust control can
substantially reduce blood lead burdens in children (Charney et al., 1983). Lower SES is also
associated with older housing stocks and increasing risk of encountering lead paint in poor
condition and lead pipes in water systems. Lower SES is also more likely to be associated
with inadequate dietary calcium, iron, and vitamins, all of which increase lead absorption and
the likely toxic effects of any given level of lead exposure. In addition, lower SES is also
more likely to imply reduced awareness of lead hazards and reduced resources for dealing with
such hazards. Other factors, such as the presence of pets in a household and the amount of
time spent playing outside, are not obviously related to SES.
Males have higher blood lead levels than females, at least beyond ages 10-11. The most
plausible explanations suggest differential exposure, with older boys and men typically spend-
ing more time in contact with motor vehicles, in jobs with potential lead exposure, and more
often outdoors. The risk factors have not been fully identified. Black children also often
have higher blood leads than do white children, even after adjusting for SES and other covar-
iates; the reason for this difference has also not been clarified, but may be related to posi-
tive confounding factors.
For modelling purposes, the appropriate geometric standard deviation removes a portion of
the total variation in blood lead due to differences in air lead exposure without removing the
variance due to these other factors. Controlling for race, urbanization, age, income, and
location may overcontrol in this case, since it may remove variance due to environmental expo-
sure factors that will remain after air lead is controlled to any given level. It may thus be
prudent and conservative to compensate for this overcontrol by increasing the geometric stan-
dard deviation when only air lead is used as a predictor variable.
All of the above factors make it difficult to analyze adequately such a highly confounded
environmental exposure variable as air lead. However, there appear to be enough studies in
which several of the possible confounding factors were also measured that it is possible to
obtain reasonable estimates of blood lead changes in response to differences in concentrations
of lead in air, dust, soil, water, and diet, seasonal variations, and personal risk factors
such as household quality, occupational exposure, and motor vehicle exposure. The remaining
sections of this chapter discuss studies from which such estimates are derived. Experimental
studies are much less subject to confounding, and where available, are generally preferred.
Unfortunately, experimental studies do not provide information about total environmental air
11-7
-------
lead exposure, which includes multiple exposure pathways and possible time lags of many years
due to passage of lead through the soil, the food chain, and water supplies. It is thus also
necessary to obtain information about total air lead exposure from observational studies. All
observational studies suffer confounding problems. This chapter focuses mainly on those ob-
servational studies in which a substantial number of the probable important confounding fac-
tors are either measured or are controlled by the design of the study. Less importance is
assigned to those studies in which too many important covariates have been omitted, or which
otherwise seem critically deficient.
11.3 LEAD IN HUMAN POPULATIONS
11.3.1 Introduction
This descriptive section presents information on dimensions of current internal exposures
to lead for United States populations. Several aspects of the current situation regarding
internal lead exposures are addressed. First, attention is focused on showing how current in-
dices of internal exposure compare with indices derived from historical samples. Also, the
question of how contemporaneous populations compare with one another with respect to internal
exposures is addressed. The primary data involved in this discussion are blood lead levels
from populations showing varying degrees of urbanization. Blood lead levels are lowest in
populations living remotely from urban influences and increase as one goes from rural to urban
areas, suggesting that higher blood lead levels are linked to urban lifestyles. Following
this discussion, data are presented on several large studies in the United States and a large
worldwide study. These data address two principal questions: 1) are there identifiable sub-
populations in the United States which exhibit higher than average blood lead levels, and
2) how do United States blood lead levels compare with other countries? This section next
presents studies which examine recent time trends in blood lead levels in the United States
and elsewhere, and then concludes with a discussion of evidence which points towards gasoline
lead being an important determinant of changes in blood lead levels associated with exposures
to airborne lead of populations in the United States and elsewhere.
11.3.2 Ancient and Remote Populations
One question of much interest in understanding environmental pollutants is the extent to
which current ambient exposures exceed background levels. Because lead is a naturally occur-
ring element it can be surmised that some level has been and will always be present in the
human body; the question of interest is what is the difference between body burdens of current
subgroups of the United States population and those "natural" levels. Information regarding
11-8
-------
this issue has been developed from studies of populations that lived in the past and popula-
tions that currently live in remote areas far from the influence of industrial and urban lead
exposures.
Man has used lead since antiquity for a variety of purposes. These uses have afforded
the opportunity for some segments of the human population to be exposed to lead and subse-
quently absorb it into the body. Because lead accumulates over a lifetime in bones and teeth
and because bones and teeth stay intact for extremely long times, it is possible to estimate
the extent to which populations in the past have been exposed to lead. Because of the prob-
lems of scarcity of samples and little knowledge of how representative the samples are of con-
ditions at the time, the data from these studies provide only rough estimates of the extent of
absorption. Further complicating the interpretation of these data are debates over proper
analytical procedures and the question of whether skeletons and teeth pick up or release lead
from or to the soil in which they are interred (Waldron et al., 1979; Waldron, 1981).
Waldron et al. (1979) have argued that any lead found in ancient bones probably is an
accurate reflection of exposure during life. They reported a small study which showed no cor-
relation between bone and soil lead concentrations. Later, however, Waldron (1981) reported a
study in which the postmortem bone lead levels appeared to be much too high to have been
developed during life. The bones were recovered from lead coffins. Electron microprobe
analysis on one bone from a lead coffin showed that the lead was concentrated on the surfaces
of the bone. This suggested that the lead in bones came from the lead coffin and led Waldron
(1981) to suggest that "in any further study of the lead content of bones from archaeological
sites, steps must be taken to assess environmental lead levels and if these are unusually
high, the results of the analyses should be viewed with suspicion." Barry and Connolly (1981)
express further concern over the use of paleontological remains as doubtful criteria for the
u» vivo assessment of lead exposure in past populations.
Despite these methodological difficulties, several studies provide data by which to esti-
mate internal exposure patterns among ancient populations, and some studies have included data
from both past and current populations for comparisons. Data from specific studies of bone
and teeth in ancient populations are summarized below in Section 11.3.2.1. In contrast to the
study of ancient populations using bone and teeth lead levels, several studies have looked at
the issue of lead contamination from the perspective of comparing blood lead levels in current
remote and urbanized populations. These studies using blood lead levels as an indicator found
mean blood concentrations in remote populations between 1 and 5 ug/dl (an order of magnitude
below current U.S. urban population means), as discussed in Section 11.3.2.2 below.
11-9
-------
11.3.2.1 Ancient Populations. Table 11-1 summarizes several studies that analyzed bones and
teeth to yield approximate estimates of lead absorption in the past. Some of these studies
also analyzed contemporary current samples so that a comparison between past and present could
be made. Studies summarized in Table 11-1 show an increase of lead levels in bone and teeth
from older to contemporary samples.
Samples from the Sudan (ancient Nubians) were collected from several different archaeo-
logical periods (Grandjean et al. , 1979). The oldest sample (3300-2900 B.C.) averaged 0.6
ug/g for bone and 0.9 ug/g for teeth. Data from the later time of 1650-1350 B.C. show a sub-
stantial increase in absorbed lead. Comparison of even the most recent ancient samples with a
current Danish sample showed a four- to eightfold increase over time.
The Shapiro et al. (1975) study compared the tooth lead content of ancient populations
with that of current remote populations and, also, with current urban populations. The
ancient Egyptian samples (1st and 2nd millenia) exhibited the lowest tooth lead levels, with
a mean of 9.7 ug/g. The more recent Peruvian Indian samples (12th century) had similar levels
(13.6 ug/g). The contemporary Alaskan Eskimo samples had a mean of 56.0 ug/9, while
Philadelphia samples had a mean of 188.3 ug/g. These data suggest an increasing pattern of
lead absorption from ancient populations to current remote and urban populations.
Data have also been obtained from ancient Peruvian and Pennsylvanian samples (Becker et
al., 1968). The Peruvian and Pennsylvanian samples for American Indian populations were from
approximately the same era (~1200-1400 A.D.). Little lead was used in these cultures as re-
flected by chemical analysis of bone lead content. The values were less than 5 ug/g for both
samples. In contrast, values obtained for modern samples from residents of Syracuse, New
York, ranged from 5 to 110 ug/g. Ericson et al. (1979) also analyzed bone speciments from
ancient Peruvians. Samples from 4500-3000 years ago to about 1400 years ago were reasonably
constant (<0.2 ug/g).
Fosse and Wesenberg (1981) reported a study of Norwegian teeth samples from several eras.
The older material from 1200-1800 A.D. was significantly lower in lead (1.22 to 1.81 ug/g)
than modern samples (3.73 to 4.12 ug/g).
Aufderheide et al. (1981) report a study of 16 skeletons from colonial America. Two
social groups, identified as plantation proprietors and laborers, had distinctly different
exposures to lead as shown by the analyses of the skeletal samples. The proprietor group
averaged 185 ug/g bone ash while the laborer group averaged 35 ug/g.
Changes in bone and tooth lead concentrations over time (as determined by the above or
other studies) have been evaluated by Angle and Mclntire (1982), as graphically depicted in
Figure 11-2. Lead concentrations in human bones apparently markedly increased among ancient
11-10
-------
TABLE 11-1. SUMMARY OF REPRESENTATIVE STUDIES OF PAST EXPOSURES TO LEAD
Population studied
Nubians1 vs. Modern Danes
Nubians
A- group
C-group
Pharonic
Merotic, X-group and Christians
Danes
Ancient Peruvians2
Ancient Pennsylvania!! Indians
Recent Syracuse, NY
Uvdal3
Modern Buskend County
Bryggen
Norway
Ancient Egyptian4
Peruvian Indian
Alaskan Eskimo
Philadelphian
Age of sample
3300 B.C. to 750 A.D. (5000 yrs. old)
3300 to 2900 B.C.
2000 to 1600 B.C.
1650 to 1350 B.C.
1 to 750 A.D.
Contemporary
500-600 yrs. old
500 yrs. old
Contemporary
Buried from before 1200 A.D. to 1804
Contemporary
Medieval Bergen
Contemporary
1st and 2nd millennia
12th century
Contemporary
Contemporary
Method of analysis
PASS, ASV
PASS, ASV
PASS, ASV
PASS, ASV
FASS, ASV
PASS, ASV
Arc emission spectroscopy
Arc emission spectrosocpy
Arc emission spectroscopy
AAS
AAS
AAS
AAS
ASV
ASV
ASV
ASV
Lead
M9/9
Bone
0.6t
i.ot
2.0t
1.2t
5.5t
<5tt
N.D.
5-110tt
levels,
dry weight
Tooth
0.9*
2.1*
5.0*
3.2*
25.7*
1.22**
4. 12**
1.81**
3.73**
9.7
13.6
56.0
188.3
!Grandjean et al. (1979).
2Becker et al. (1968).
3Fosse and Wesenberg (1981).
"Shapiro et al. (1975).
"Circunpulpal dentine.
(Temporal bone.
TtTibia/femur.
**Whole tooth, but values corrected for enamel and dentine.
-------
,
f-
A PERU / \
O EGYPT / \
0 NUBIA / \
DENMARK / \
A BRITAIN-ROMAN. / >
ANGLO SAXON / /^N T
u.s. / / V
O BRITAIN, contemporary ' f \
!
» ,
/ i
\
\
\
\
\
/ /j ii
/ ^ m
.._n A -^~ ^^ *-" A M A V
r r rn^Fn r r r i
200
150
-100
50
hi 0
O)
O
CD
5500 5000 4500 4000 3500 3000 2500 2000 1500 1000 500 PRESENT
BP
YEARS BEFORE PRESENT
Figure 11-2. Estimated lead concentrations in bones (//g/g) from 5500
years before present (BP) to the present time, from ancient Peru (Ericson
et al. 1979) and Egypt, Nubia, and Denmark (Grandjean et al. 1979).
Britain in the Roman and Anglo-Saxon (Waldron 1980) eras, contem-
porary British children (Barry 1981), and U.S. adults in the 1950s
(Schroeder and Tipton 1968).
Source: From Angle and Mclntire (1982).
Early (Britain-Roman, Anglo-Saxon) with soil
possibly contaminated with lead.
Early (Britain-Roman, Anglo-Saxon) with soil
believed not to be contaminated with lead.
Represents range of values.
11-12
-------
populations with the introduction of metallurgic processes and dramatic increases in produc-
tion and utilization of lead. For example, bone lead concentrations consistently below 3 ug/g
were found for premetallurgic societies in Peru, Egypt, Nubia, and Denmark, whereas concentra-
tions of lead in bones from England during the early Roman Empire era are reported to be
10-fold higher and to have reached 300 to 400 ug/g by the time of the Norman invasion. The
Danish bone lead levels also increased during medieval times and reached peak levels of about
40-50 ug/g in the eighteenth century. The data available for more recent contemporary popu-
lations in the twentieth century appear to be widely variable, ranging from 0.1 to 5.4 ug/g
reported for contemporary adults in Denmark to 7.5 to 195 ug/g reported for U.S. adults dying
in the 1950's. Overall, the available data (despite analytic errors in individual studies)
collectively suggest that contemporary Americans, especially urban populations, absorb mani-
fold higher levels of lead than did members of premetallurgic societies.
11.3.2.2 Remote Populations. Several studies have looked at the blood lead levels in current
remote populations (Piomelli et al., 1980; Poole et al., 1980). These studies are important
in defining baseline levels of internal lead exposures found in the world today.
Piomelli et al. (1980) studied blood lead levels of natives in a remote (far from indus-
trialized regions) section of Nepal. Portable air samplers were used to determine air lead
concentrations in the region. The lead content of the air samples proved to be less than the
detection limit, 0.004 ug/m3. A later study by Davidson et al. (1981) found an average air
lead concentration of 0.00086 ug/m3 in remote areas of Nepal, thus confirming the low air lead
levels reported by Piomelli et al. (1980).
Blood lead levels reported by Piomelli et al. (1980) for the Nepalese natives were low;
the geometric mean blood lead for this population was 3.4 ug/dl. Adult males had a geometric
mean of 3.8 ug/dl and adult females, 2.9 ug/dl. Children had a geometric mean blood lead of
3.5 ug/dl. Only 10 of 103 individuals tested had a blood lead level greater than 10 ug/dl.
The blood samples, which were collected on filter paper discs, were analyzed by a modification
of the Delves cup atomic absorption spectrophotometric method. Stringent quality control pro-
cedures were followed for both the blood and air samples. To put these Nepalese values in
perspective, Piomelli et al. (i960) reported analyses of blood samples collected and analyzed
by the same methods from Manhattan, New York. New York blood leads averaged about 15 ug/dl,
fivefold higher than the Nepalese values.
Poole et al. (1980) reported another study of a remote population, using contamination-
free micro-blood sampling and chemical analysis techniques. They reported acceptable preci-
sion at blood lead concentrations as low as 5 ug/dl, using spectrophotometry. One hundred
children were sampled from a remote area of Papua, New Guinea. Almost all of the children
came from families engaging in subsistence agriculture. The children ranged from 7 to 10
11-13
-------
years and included both sexes. Blood lead levels ranged from I to 13 ug/dl with a mean of
5.2. Although the data appear to be somewhat skewed to the right, they are in good agreement
with those of Piomelli for Nepalese subjects.
11.3.3 Levels of Lead and Demographic Covariates in U.S. and Other Populations
Several large surveys of blood lead levels give information on the major demographic co-
variates in U.S. populations (see also sections 7.3.2.2 and 7.3.2.3.) In addition to the
obvious covariates of age, sex, race, and urban-rural differences, there is a more subtle
effect of seasonality. Children show a strong midsummer peak (hence the characterization of
lead poisoning as "the summer disease" (Hunter, 1978)). This peak may be attributed to many
causes: 1) gasoline lead consumption and lead concentrations are higher in the summer; 2)
many people, especially children, spend more time outside during the summer; 3) more beverages
are consumed in the summer, increasing exposure from lead-soldered beverage cans; and 4) other
seasonal variations in diet, climate, and health status may affect blood lead levels. Thus,
seasonality has an effect on all of the demographic studies. The extent to which these demo-
graphic studies adjust for seasonality varies.
11.3.3.1 The NHANES II Study. The National Center for Health Statistics has provided the
best currently available picture of blood lead levels among United States residents as part of
the second National Health and Nutrition Examination Study (NHANES II) conducted from
February, 1976 to February, 1980 (Mahaffey et al., 1982; McDowell et al., 1981; Annest et al.,
1982; Annest and Mahaffey, 1984). These are the first national estimates of lead levels in
whole blood from a representative sample of the non-institutionalized U.S. civilian population
aged 6 months to 74 years.
From a total of 27,801 persons identified through a stratified, multi-stage probability
cluster sample of households throughout the United States, blood lead determinations were
scheduled for 16,563 persons including all children ages 6 months to 6 years, and one-half of
all persons ages 7-74. Sampling was scheduled in 64 sampling areas over the four-year period
according to a previously determined itinerary to maximize operational efficiency and response
of participants. Because of the constraints of cold weather, the examination trailers
traveled in the moderate climate areas during the winter, and the more northern areas during
the summer (McDowell et al., 1981).
All reported blood lead levels were based on samples collected by venipuncture. Blood
lead levels were determined by atomic absorption spectrophotometry using a modified Delves cup
micro-method. Specimens were analyzed in duplicate, with both determinations done independ-
ently in the same analytical run. Quality control was maintained by two systems, a bench
system and a blind insertion of samples. If the NHANES II replicates differed by more than
11-14
-------
7 |jg/dl, the analysis was repeated for the specimen (about 0.3 percent were reanalyzed). If
the average of the replicate values of either "bench" or "blind" control specimens fell out-
side previously established 95 percent confidence limits, the entire run was repeated. The
estimated coefficient of variation for the "bench" quality control ranged from 7 to 15 percent
(Mahaffey et al., 1979).
The reported blood lead levels were based on the average of the replicates. Blood lead
levels and related data were reported as population estimates; findings for each person were
inflated by the reciprocal of selection probabilities, adjusted to account for persons who
were not examined and poststratified by race, sex, and age. The final estimates closely ap-
proximate the U.S. Bureau of Census estimates for the civilian non-institutionalized popula-
tion of the United States as of March 1, 1978, aged 1/2-74 years.
Participation rates varied across age categories; the highest non-response rate (51
percent) was for the youngest age group, 6 months through 5 years. Among medically examined
persons, those with missing blood lead values were randomly distributed by race, sex, degree
of urbanization, and annual family income. These data are probably the best estimates now
available regarding the degree of lead absorption in the general United States population.
Forthofer (1983) has studied the potential effects of non-response bias in the NHANES II
survey and found no large biases in the health variables. This was based on the excellent
agreement of the NHANES II examined data, which had a 27 percent non-response rate, with the
National Health Interview Survey data, which had a 4 percent non-response rate.
The national estimates presented below are based on 9933 persons whose blood lead levels
ranged from 2.0 to 66.0 ug/dl. The median blood lead for the entire U.S. population is 13.0
ug/dl. It is readily apparent that blacks have a higher blood lead level than whites (medians
for blacks and whites were 15.0 and 13.0 ug/dl, respectively).
Tables 11-2 through 11-4 display the observed distribution of measured blood lead levels
by race, sex, and age. The possible influence of measurement error on the percent distribu-
tion estimates is discussed in Section 11.3.4. Estimates of mean blood lead levels differ
substantially with respect to race, age, and sex. Blacks have higher levels than whites, the
6-month to 5-year group is higher than the older age groups, and men are higher than women.
Overall, younger children show only a slight age effect, with 2- to 3-year-olds having slight-
ly higher blood lead levels than older children or adults (see Figure 11-3). In the 6-17 year
grouping there is a decreasing trend in lead levels with increasing age. Holding age con-
stant, there are significant race and sex differences; as age increases, the difference
between males and females in mean blood lead concentrations increases.
11-15
-------
TABLE 11-2. NHANES II BLOOD LEAD LEVELS OF PERSONS 6 MONTHS-74 YEARS, WITH WEIGHTED ARITHMETIC MEAN, STANDARD ERROR OF THE
MEAN, WEIGHTED GEOMETRIC MEAN, MEDIAN, AND PERCENT DISTRIBUTION, BY RACE AND AGE, UNITED STATES, 1976-80
Blood lead level pg/dl
Race and age
All racesd
All ages
6 months-5 years
6-17 years
18-74 years
White
Al 1 ages
6 months-5 years
6-17 years
18-74 years
Black
All ages
6 months-5 years
6-17 years
18-74 years
Estimated
population
in
thousands
203,554
16,852
44,964
141,728
174,528
13,641
37,530
123,357
23,853
2,584
6,529
14,740
Percent distribution0
Number b
exarai ned
9,933
2,372
1,720
5,841
8,369
1,876
1,424
5,069
1,332
419
263
650
Arith-
metic
mean
13.9
16.0
12.5
14.2
13.7
14.9
12.1
14.1
15.7
20.9
14.8
15.5
Standard
error of
the mean
0.24
0.42
0.30
0.25
0.24
0.43
0.30
0.25
0.48
0.61
0.53
0.54
Geometric
mean
12.8
14.9
11.7
13.1
12.6
14.0
11.3
12.9
14.6
19.6
14.0
14.4
Median
13.0
15.0
12.0
13.0
13.0
14.0
11.0
13.0
15.0
20.0
14.0
14.0
Less
than
10
22.1
12.2
27.6
21.2
23.3
14.5
30.4
21.9
13.3
2.5
12.8
14.7
10-19
62.9
63.3
64.8
62.3
62.8
67.5
63.4
62.3
63.7
45.4
70.9
62.9
20-29
13.0
20.5
7.1
14.3
12.2
16.1
5.8
13.7
20.0
39.9
15.6
19.6
30-39
1.6
3.6
0.5
1.8
1.5
1.8
0.4
1.8
2.3
10.2
0.7
2.0
40+
0.3
0.4
-
0.4
0.3
0.2
-
0.4
0.6
2.0
-
0.9
aAt the midpoint of the survey, March 1, 1978.
ntfith lead determinations from blood specimens drawn by venipuncture.
GNumbers may not add up to 100 percent due to rounding.
Includes data for races not shown separately.
-------
TABLE 11-3. NHANES II BLOOD LEAD LEVELS OF HALES 6 MONTHS-74 YEARS, WITH WEIGHTED ARITHMETIC MEAN, STANDARD ERROR OF THE
MEAN, WEIGHTED GEOMETRIC MEAN, MEDIAN, AND PERCENT DISTRIBUTION, BY RACE AND AGE, UNITED STATES, 1976-80
Blood lead level ug/dl
Race and age
All racesd
All ages
6 months-5 years
6-17 years
18-74 years
White
All ages
6 months-5 years
6-17 years
18-74 years
Black
All ages
6 ionths-5 years
6-17 years
18-74 years
Estimated
population
in
thousands
99,062
8,621
22,887
67,555
85,112
6,910
19.060
59,142
11,171
1,307
3,272
6,592
Number .
examined
4,945
1,247
902
2,796
4,153
969
753
2,431
664
231
129
304
Arith-
metic
mean
16.1
16.3
13.6
16.8
15.8
15.2
13.1
16.6
18.3
20.7
16.0
19.1
Standard
error of
the mean
0.26
0.46
0.32
0.28
0.27
0.46
0.33
0.29
0.52
0.74
0.62
0.70
Geometric
mean
15.0
15.1
12.8
15.8
14.7
14.2
12.4
15.6
17.3
19.3
15.3
18.1
Median
15.0
15.0
13.0
16.0
15.0
14.0
13.0
16.0
17.0
19.0
15.0
18.0
Less
than
10
10.4
11.0
19.1
7.6
11.3
13.0
21.4
8.1
4.0
2.7
8.0
2.3
Percent
10-19
65.4
63.5
70.1
64.1
66.0
67.6
69.5
64.8
59.6
48.8
69.9
56.4
distribution
20-29
20.8
21.2
10.2
24.2
19.6
17.3
8.4
23.3
31.0
35.1
21.1
34.9
30-39
2.8
4.0
0.7
3.4
2.6
2.0
0.7
3.3
4.1
11.1
1.0
4.5
40+
0.5
0.3
-
0.6
0.4
0.1
-
0.6
1.3
2.4
-
1.8
aAt the midpoint of the survey, March 1, 1978.
nrfith lead determinations from blood specimens drawn by venipuncture.
cNunbers nay not add to 100 percent due to rounding.
Includes data for races not shown separately.
-------
TABLE 11-4. NHANES II BLOOD LEAD LEVELS OF FEMALES 6 MONTHS-74 YEARS, WITH WEIGHTED ARITHMETIC MEAN,
STANDARD ERROR OF THE HE AN, WEIGHTED GEOMETRIC MEAN, MEDIAN, AND PERCENT DISTRIBUTION, BV RACE AND AGE, UNITED STATES, 1976-80
oc
Blood lead level
Race and age
All racesd
All ages
6 months -5 years
6-17 years
18-74 years
White
All ages
6 months-5 years
6-17 years
18-74 years
Black
All ages
6 months-5 years
6-17 years
18-74 years
Esti Bated
population
in a
thousands
104,492
8,241
22,077
74,173
89,417
6,732
18,470
64,215
12,682
1,277
3,256
8,148
Number .
examined
4,988
1,125
818
3,045
4,216
907
671
2,638
668
188
134
346
Arith-
metic
mean
11.9
15.8
11.4
11.8
11.7
14.7
11.0
11.7
13.4
21.0
13.6
12.7
Standard
error of
the mean
0.23
0.42
0.32
0.22
0.23
0.44
0.31
0.23
0.45
0.69
0.64
0.44
Geometric
mean
11.1
14.6
10.6
11.0
10.9
13.7
10.3
10.9
12.6
19.8
12.8
12.0
Median
11.0
15.0
11.0
11.0
11.0
14.0
11.0
11.0
13.0
20.0
13.0
12.0
Less
than
10
33.3
13.5
36.6
33.7
34.8
16.1
40.0
34.6
21.5
2.2
17.7
24.7
, M9/dl
Percent
10-19
60.5
63.2
59.3
60.6
59.6
67.3
56.9
59.9
67.3
41.6
71.9
68.1
distribution0
20-29
5.7
19.8
3.9
5.2
5.0
14.8
2.9
5.0
10.3
45.3
10.0
7.2
30-39
0.4
3.0
0.2
0.3
0.4
1.6
0.2
0.4
0.7
9.2
0.4
"
40+
0.2
0.5
-
0.2
0.2
0.2
-
0.2
0.1
1.7
-
"
aAt the Midpoint of the survey, March 1, 1978.
Trfith lead determinations fro* blood specimens drawn by venipuncture.
GNumbers may not add to 100 percent due to rounding.
Includes data for races not shown separately.
-------
5
O)
UJ
UJ
O
o
O
§
o
25
20
15
10
Black
White
AGE, yean
Figure 11-3. Geometric mean blood lead levels by race and age for
younger children in the NHANES II study. EPA calculations from
data furnished by the National Center for Health Statistics.
Source: Annest and Mahaffey (1984).
11-19
-------
For adults 18-74 years, males have greater blood lead levels than females for both whites
and blacks. There is a significant relationship between age and blood lead, but it differs
for whites and blacks. Whites have increasing blood lead levels until 35-44 years of age and
then decline, while blacks have increasing blood lead levels until 55-64.
This study showed a clear relationship between blood lead level and family income group.
For both blacks and whites, increasing family income is associated with lower blood lead
level. At the highest income level the difference between blacks and whites is the smallest,
although blacks still have significantly higher blood lead levels than whites. The racial
difference was greatest for the 6-month to 5-year age range.
The NHANES II blood lead data were also examined with respect to the degree of urbaniza-
tion at the place of residence. The three categories used were urban areas with population
greater than one million, urban areas with population less than one million, and rural areas.
Geometric mean blood lead levels increased with degree of urbanization for all race-age groups
except for blacks 18-74 years of age (see Table 11-5). Most importantly, urban black children
aged 6 months - 5 years appeared to have distinctly higher mean blood lead levels than any
other population subgroup.
11.3.3.2 The Childhood Blood Lead Screening Programs. In addition to the nationwide picture
presented by the NHANES II (Annest et al., 1982) study regarding important demographic corre-
lates of blood lead levels, Billick et al. (1979, 1982) provide large scale analyses of blood
lead values from childhood blood lead screening programs in specific cities that also address
this issue.
Billick et al. (1979) analyzed data from New York City blood lead screening programs from
1970 through 1976. The data include age in months, sex, race, residence expressed as health
district, screening information, and blood lead values expressed in intervals of 10 ug/dl.
Only the venous blood lead data (178,588 values), clearly identified as coming from the first
screening of a given child, were used. All blood lead determinations were done by the same
laboratory. The geometric means of the children's blood lead levels by age, race, and year of
collection are presented in Table 11-6. The annual means were calculated from the four quar-
terly means which were estimated by the method of Hasselblad et al. (1980).
The data obtained for New York are generally consistent with the nationwide results from
the NHANES II study. For example, all racial/ethnic groups show an increase in geometric mean
blood level with age for the first two years and a general decrease in the older age groups.
These age-related patterns are seen in Figure 11-4, which shows the trends for all years
(1970-1976) combined. Also, the childhood screening data described by Billick et al. (1979)
show higher geometric mean blood lead values for blacks than for Hispanics or for whites.
Table 11-6 presents these geometric means for the three racial/ethnic groups for seven years.
11-20
-------
TABLE 11-5. WEIGHTED GEOMETRIC MEAN BLOOD LEAD LEVELS
FROM NHANES II SURVEY BY DEGREE OF URBANIZATION OF PLACE OF
RESIDENCE IN THE U.S. BY AGE AND RACE, UNITED STATES 1976-80
(micrograms/deci1iter)
Race and age
All races
All ages
6 months-5
6-17 years
18-74 years
Whites
Al 1 ages
6 months-5
6-17 years
18-74 years
Blacks
All ages
6 months-5
6-17 years
Degree
Urban,
^1 million
years
- men:
women:
years
- men:
women:
years
18-74 years - men:
women:
14.
16.
13.
16.
12.
14.
15.
12.
16.
12.
14.
20.
14.
17.
11.
0
8
1
9
2
0
6
6
9
4
4
8
6
4
8
(2,395)a
(544)
(414)
(677)
(760)
(1,767)
(358)
(294)
(531)
(584)
(570)
(172)
(111)
(132)
(155)
12.
15.
11.
15.
11.
12.
14.
11.
15.
10.
14.
19.
13.
18.
12.
of urbanization
Urban,
<1 million
8
4
7
7
0
5
4
4
4
8
8
2
6
6
4
(3,869)
(944)
(638)
(1,050)
(1,237)
(3,144)
(699)
(510)
(889)
(1,046)
(612)
(205)
(113)
(134)
(160)
11.
13.
10.
15.
9.
11.
12.
10.
14.
9.
14.
16.
13.
18.
11.
Rural
9
0
7
1
8
8
7
5
8
8
4
5
0
3
3
(3
(1
(1
(3
(1
(1
,669)
(884)
(668)
,069)
,048)
,458)
(819)
(620)
,011)
,008)
(150)
(42)
(39)
(38)
(31)
Number with lead determinations from blood specimens drawn by venipuncture.
Source: Annest and Mahaffey, 1984; Annest et al., 1982.
11-21
-------
TABLE 11-6. ANNUAL GEOMETRIC MEAN BLOOD LEAD LEVELS FROM THE NEW YORK BLOOD LEAD SCREENING STUDIES
OF BILLICK ET AL. (1979). ANNUAL GEOMETRIC MEANS ARE CALCULATED FROM QUARTERLY
GEOMETRIC MEANS ESTIMATED BY THE METHOD OF HASSELBLAD ET AL. (1980)
(micrograms/deciliter)
Ethnic group
Black
Hispanic
White
Year
1970
1971
1972
1973
1974
1975
1976
1970
1971
1972
1973
1974
1975
1976
1970
1971
1972
1973
1974
1975
1976
1-12 mo
25.2
24.0
22.2
22.9
22.0
19.8
16.9
20.8
19.9
18.7
20.2
19.8
16.3
16.0
21.1
22.5
20.1
21.5
20.4
19.3
15.2
13-24 mo
28.9
29.3
26.0
26.6
25.5
22.4
20.0
23.8
22.6
20.5
21.8
21.5
18.7
17.4
25.2
22.7
21.6
21.8
21.7
17.9
18.2
25-36 mo
30.1
29.9
26.3
26.0
25.4
22.4
20.6
24.5
24.6
21.8
22.5
22.7
19.9
18.1
26.0
22.7
20.7
21.7
21.3
16.1
17.1
Age
37-48 mo
28.3
29.3
25.4
25.3
24.3
21.9
20.2
24.7
24.4
22.2
22.8
22.5
20.1
18.2
24.8
23.5
20.8
20.2
21.1
18.5
16.6
49-60 mo
27.8
28.2
24.7
24.4
23.4
21.2
19.5
23.8
23.9
21.8
22.0
21.9
19.8
18.0
26.0
21.6
21.0
21.3
20.6
16.8
16.2
61-72 mo
26.4
27.2
23.9
24.1
21.8
21.4
18.2
23.6
23.4
21.8
21.5
20.5
19.2
16.7
22.6
21.3
20.2
20.7
19.5
15.4
15.9
73- mo
25.9
26.5
23.3
23.3
21.9
18.9
18.4
23.0
23.5
21.0
21.7
20.2
17.2
17.2
21.3
19.5
17.3
18.4
17.3
15.9
8.8
All ages
27.5
27.7
24.5
24.6
23.4
21.1
19.1
23.4
23.1
21.1
21.8
21.3
18.7
17.4
23.8
21.9
20.2
20.8
20.2
17.1
15.1
-------
30
25
o 20
3
O
o
O
03
2
O IP
15
D BlHcks
O Whites
& Hispanics
123456
AGE ye,irs
Figure 11-4. Geometric mean blood lead values by race and age
for younger children in the New York City screening program
(1970-1976).
Source: Adapted from Hasselblad etal. 1980.
11-23
-------
Using the method of Hasselblad et al. (1980), the estimated geometric standard deviations were
1.41, 1.42, and 1.42 for blacks, Hispanics, and whites, respectively.
11.3.3.3 Levels of Lead and Demographic Covariates Worldwide. An international study conduc-
ted under the auspices of the United Nations Environment Program and the World Health Organi-
zation provides the first analytically comparable blood lead data set available to infer the
current similarities and differences in lead absorption from country to country (Friberg and
Vahter, 1983). Extensive attention was paid to quality control issues, with the resulting
blood lead determinations being very comparable from country to country. School teachers were
chosen as study subjects since they would be unlikely to have occupational exposures to lead
and also because they would have similarities in socioeconomic characteristics. A detailed
interview was administered to the subjects to obtain background data.
Figure 11-5, derived from data in the paper, displays the variability from country to
country. Unweighted geometric mean blood lead levels ranged from a low of 5.8 ug/dl in Japan
to 22.3 ug/dl in Mexico. Teachers in China, Israel, Japan, Sweden, and the United States all
had geometric mean blood leads below 8.0 yg/dl.
In general, males showed higher blood lead levels than females; on the average, male
teachers had blood lead levels 30 percent higher than females regardless of cigarette smoking
status. In most cases cigarette smokers had 10 percent higher blood lead levels than
nonsmokers.
11.3.4 Distributional Aspects of Population Blood Lead Levels
The importance of the form of the distribution of blood lead levels was briefly discussed
in Section 11.2.2. The distribution form determines which measure of central tendency (arith-
metic mean, geometric mean, median) is most appropriate. It is even more important in esti-
mating percentiles in the upper tail of the distribution, an issue of much importance in esti-
mating percentages (or absolute numbers) of individuals in specific population groups likely
to be experiencing various lead exposure levels.
Distribution fitting requires large numbers of samples taken from a relatively homo-
geneous population. A homogeneous population is one in which the distribution of values
remains constant when split into subpopulations. These subpopulations could be defined by
demographic factors such as race, age, sex, income, degree of urbanization, and degree of
exposure. Since these factors always have some effect, a relatively homogeneous population
will be defined as one with minimal effects from any factors that contribute to differences in
blood lead levels.
11-24
-------
24
2 5 22
20
§5
2*0
z
y|
EZ
p
Os
111 K
Ss
00
£3
QUI
u<
3
14
12
10
D
2
UJ
CD
4
O
£
Q
Z
CM
O
z
n
O
Z
UJ
K
(A
Z
0.
1
8
X
UJ
C
UJ
0.
^u
UI
(/>
UJ
B
(/)
r»
4
>
2
CO
O
STUDY LOCATION
Figure 1 1 -5. Unweighted geometric mean blood lead level for male and
female nonsmoking teachers (M9/dl) for several countries.
Source: Derived from Friberg and Vahter (1983).
11-25
-------
Several authors have suggested that the distribution of blood lead levels for any rela-
tively homogeneous population closely follows a lognormal distribution (Yankel et al., 1977;
Tepper and Levin, 1975; Azar et al. , 1975). Lognormality has been noted for other metals,
such as 90Sr, 144Ce, Pu, and Ti in various tissues of human populations (Cuddihy et al., 1979;
Schubert et al., 1967). Yankel et al. (1977), Tepper and Levin (1975), and Angle and Mclntire
(1979) all found their blood lead data to be lognormally distributed. Further analysis by EPA
of the Houston study of Johnson et al. (1974), the study of Azar et al. (1975), and the New
York children screening program reported by Billick et al. (1979) also demonstrated that a
lognormal distribution provided a good fit to the data.
The only nationwide survey of blood lead levels in the U.S. population is the NHANES II
survey (Annest et al., 1982). In order to obtain a relatively homogeneous subpopulation of
lower environmental exposure, the analysis was restricted to whites not living in an SMSA
(Standard Metropolitan Statistical Area), with a family income greater than $6,000 per year,
the poverty threshold for a family of four at the midpoint of study as determined by the U.S.
Bureau of Census. This subpopulation was split into four subgroups based on age and sex.
The summary statistics for these subgroups are in Table 11-7.
TABLE 11-7. SUMMARY OF UNWEIGHTED BLOOD LEAD LEVELS IN WHITES
NOT LIVING IN AN SMSA, WITH FAMILY INCOME GREATER THAN $6,000
Unweighted mean
Subgroup
Age 1/2 to 6
Age 6 to 18
Age 18+, men
Age 18+, women
Sample
size
752
573
922
927
Arith.
mean,
ug/dl
13.7
11.3
15.7
10.7
Geom.
mean,
ug/dl
12.9
10.6
14.7
10.0
Sample
median,
ug/dl
13.0
10.0
15.0
10.0
99th
percentile,
ug/dl
32.0
24.0
35.8
23.0
Arith.
std. dev. ,
ug/dl
5.03
4.34
5.95
4.14
Geom.
std. dev.
1.43
1.46
1.44
1.46
Each of these four subpopulations were fitted to five different distributions: normal,
lognormal, gamma, Weibull, and Wald (Inverse Gaussian) as shown in Table 11-8. Standard
chi-square goodness-of-fit tests were computed after collapsing the tails to obtain an
expected cell size of five. The goodness-of-fit test and likelihood functions indicate that
the lognormal distribution provides a better fit than the normal, gamma, or Weibull. A
histogram and the lognormal fit for each of the four subpopulations appear in Figure 11-6.
11-26
-------
TABLE 11-8. SUMMARY OF FITS TO NHANES II BLOOD LEAD LEVELS
OF WHITES NOT LIVING IN AN SMSA, WITH INCOME GREATER THAN $6,000,
FOR FIVE DIFFERENT TWO-PARAMETER DISTRIBUTIONS
Children <6 years
Normal
Lognormal
Gamma
Weibull
Wald
Normal
Lognormal
Gamma
Weibull
Wald
Chi-square
75.52
14.75
17.51
66.77
15.71
Chi-square
39.58
3.22
4.88
24.48
2.77
D.F.*
8
10
9
8
10
Children 6
D.F.*
6
8
7
6
8
p-value
0.0000
0.1416
0.0413
0.0000
0.1083
years S17
p-value
0.0000
0.9197
0.6745
0.0004
0.9480
log-
likelihood
-2280.32
-2210.50
-2216.51
-2271.57
-2211.83
log-
likelihood
-1653.92
-1607.70
-1609.33
-1641.35
-1609.64
deviation**
at 99th
percenti le
6.61
2.57
4.68
5.51
2.76
deviation**
at 99th
percent! le
2.58
-1.50
-0.64
1.72
-1.30
Men S18 years
Normal
Lognormal
Gamma
Weibull
Wald
Normal
Lognormal
Gamma
Weibull
Wald
Chi-square
156.98
12.22
34.26
132.91
14.42
Chi-square
66.31
7.70
11.28
56.70
10.26
D.F.*
10
13
12
11
13
Women £18
D.F.*
5
8
7
6
8
p-value
0.0000
0.5098
0.0006
0.0000
0.3450
years
p-value
0.0000
0.4632
0.1267
0.0000
0.2469
log-
likelihood
-2952.85
-2854.04
-2864.79
-2934.14
-2855.94
log-
likelihood
-2631.67
-2552.12
-2553.34
-2611.78
-2556.88
deviation**
at 99th
percentile
6.24
1.51
4.00
4.88
1.72
deviation**
at 99th
percentile
2.68
-1.18
0.90
1.73
-1.01
*D.F. = degrees of freedom.
**observed 99th sample percentile minus predicted 99th percentile.
11-27
-------
7.6
15.5
23.5
31.5
BLOOD LEAD LEVELS, pg/dl.
FOR 6-MONTH TO 6-YEAR-OLD-CHILDREN
BLOOD LEAD LEVELS, /ug/dl,
FOR 6-TO 17-YEAR OLD CHILDREN
u
ui
O
15.5
23.5
31.5
BLOOD LEAD LEVELS. M9/dl,
FOR MEN ^18 YEARS OLD
7.5
15.5
23.5 31.5
BLOOD LEAD LEVELS,
FOR WOMEN ^ 18 YEARS OLD
Figure 11-6. Histograms of blood lead levels with fitted lognormal
curves for the NHANES II study. All subgroups are white, non-SMSA
residents, with family incomes over $6000/year.
Source: (EPA calculations from data supplied by National Center
for Health Statistics.)
11-28
-------
The Wald distribution is quite similar to the lognormal distribution and appears to provide
almost as good a fit. Table 11-8 also indicates that the lognormal distribution estimates
the 99th percentile as well as any other distribution.
Based on the examination of the NHANES II data, as well as the results of the several
other studies discussed above, it appears that the lognormal distribution is the most appro-
priate for describing the distribution of blood lead levels in homogeneous populations with
relatively constant exposure levels. The lognormal distribution appears to fit well across
the entire range of the distribution, including the right tail.
The lognormal distribution describes both the mean and the variation of the populations
under study. It is obvious that even relatively homogeneous populations have considerable
variation among individuals. The estimation of this variation is important for determination
of the proportion of individuals above a given blood lead level. This variation is the result
of both analytic variation and population variation.
Analytic variation, which exists in any measurement of any kind, has an impact on the
bias and precision of statistical estimates. For this reason, it is important to estimate the
magnitude of variation. Analytic variation consists of both measurement variations (vari-
ation between measurements run at the same time) and variation created by analyzing samples at
different times (days). This kind of variation for blood lead determinations has been discus-
sed by Lucas (1981). The measurement variation alone does not follow a lognormal distribu-
tion, as was shown by Saltzman et al. (1983).
Values for the variation within groups (or mean square error) are available from several
studies discussed above, including the NHANES II Survey, the N.Y. Childhood Screening Study,
the Tepper-Leven Seven City Study, and the Azar et al. study. Variation, including analytic
variation, ranged from about 1.3 to 1.4 when expressed as a geometric standard deviation.
This value depends on the uniformness of the populations and the magnitude of the analytic
variation.
The NHANES II study provides excellent data for the study of this variation, since it has
excellent quality control and extensive information on demographic covariates. In order to
minimize the effects of location, income, sex, and age, an analysis of variance procedure was
used to estimate the variation for several age-race groups. The variables just mentioned were
used as main effects, and the resulting mean square errors of the logarithms are shown in
Table 11-9. The estimated geometric standard deviations have been adjusted for sex, age, in-
come, and place of residence. As a result, the values for geometric standard deviations tend
to be smaller than the unadjusted values for specific subgroups as reported by Annest and
Mahaffey (1984).
11-29
-------
TABLE 11-9. ESTIMATED MEAN SQUARE ERRORS RESULTING FROM
ANALYSIS OF VARIANCE ON VARIOUS SUBPOPULATIONS
OF THE NHANES II DATA USING UNWEIGHTED DATA
White,
Age Non-SMSA
0.5 to 6 0.0916
(1.35)*
6 to 18 0.0814
(1.33)
18+, men 0.1155
(1.40)
18+, women 0.1083
(1.39)
White, SMS A,
not central city
0.0839
(1.34)
0.0724
(1.31)
0.0979
(1.37)
0.0977
(1.37)
White,
central city
0.1074
(1.39)
0.0790
(1.33)
0.1127
(1.40)
0.0915
(1.35)
Black,
central city
0.0978
(1.37)
0.0691
(1.30)
0.1125
(1.40)
0.0824
(1.33)
Note: Mean square errors are based on the logarithm of the blood lead levels.
*Estimated geometric standard deviations are given in parentheses.
The analytic variation was estimated specifically for this study by Annest et al.
(1983b). The analytical variation was estimated as the sum of components estimated from the
high and low blind pool and from the replicate measurements in the study of Griffin et al.
(1975). The overall estimate of analytic variation for the NHANES II study was 0.02083
(estimated mean square error based on logarithms).
Analytic variation causes a certain amount of misclassification when estimates of the
percent of individuals above or below a given threshold are made. This is because the true
value of a person's blood lead could be below the threshold, but the contribution from analy-
tic variation may push the observed value over the threshold. The reverse is also possible.
These two types of misclassifications do not necessarily offset each other.
Annest et al. (1983b) estimated this misclassification rate for several subpopulations in
the NHANES II data using a threshold value of 30 ug/dl. In general, the percent truly greater
than this threshold was approximately 24 percent less than the prevalence of blood lead levels
equal to or greater than 30 ug/dl, estimated from the weighted NHANES II data. This is less
than the values predicted by Lucas (1981) which were based on some earlier studies.
The studies reviewed here provide estimates of geometric standard deviations for observed
blood lead distributions which consistently fall in the range of 1.3 to 1.4. The NHANES II
study, thought to provide the best available data set in terms of good quality control and
11-30
-------
other features such as sample size, yields estimates of geometric standard deviations for
various subgroups of young children (0.5 to 6 years old) in the range of 1.34 to 1.39 (uncor-
rected for analytic error). Variations in the site means of log(blood lead) were calculated
after controlling for race, income, and degree of urbanization. The remaining standard devi-
ation of 0.183 for site means indicates substantial variation in baseline exposure after
accounting for the major proxies for air lead. The geometric standard deviation attributable
to the non-air lead exposure sources can be estimated by adjusting the NHANES II blood lead
levels for the impact of gasoline lead by use of linear regression. Since gasoline lead
during 1976-1980 accounted for 85 to 90 percent of air lead, the effect at gasoline lead = 0
was reduced by an additional 15 percent to account for all air lead. The resulting geometric
standard deviation was 1.428. If this calculation is done only for children with blood lead
< 40 ug/dl (who are more likely to be helped by an air lead standard) then the geometric stan-
dard deviation is 1.419. Thus, a geometric standard deviation for the NHANES II population of
children without attribution of any source of lead exposure except gasoline lead and indus-
trial air lead emissions may be taken as approximately 1.42.
11.3.5 Time Trends in Blood Lead Levels Since 1970
In the past few years a number of reports have appeared that examined trends in blood
lead levels during the 1970's. In several of these reports some environmental exposure esti-
mates are available.
11.3.5.1 Time Trends in NHANES II Study Data. Blood lead data from NHANES II (see section
11.3.3.1 for full discussion of methodology) show a significant downward trend over time for
nationwide blood lead levels in the United States (Annest et al., 1983a). After accounting
for the effects of race, sex, age, region of country, season, income, and degree of urbaniza-
tion, a statistically significant negative association with date of sampling was found. Using
regression model-predicted blood lead levels, a 37 percent drop from 14.6 to 9.2 ug/dl from
the beginning to the end of the study was found. Overall nationwide mean blood lead levels
from these data presented in 28-day intervals from February, 1976 to February, 1980 are dis-
played in Figure 11-7. Similar decreases in average blood lead levels were noted for a number
of subgroups which compose the total sample (see Figure 11-8), with the declines ranging from
31 to 42 percent for various subgroups.
A variety of possible explanations for the nationwide decline in average blood lead
levels were examined. Analysis of quality control samples indicated that laboratory drift was
not the cause of the observed decline. Further statistical analyses ruled out the possibility
that the decline was entirely due to season, income, geographic region, or urban-rural differ-
ences. Annest et al. (1983a) suggested that although strong correlation does not prove cause
and effect, the most reasonable explanation for this trend appears to be reduction in the
11-31
-------
25
T5 20
5
_r
§
Q 15
CO
ro
Q
O
O
I
UI
10
WINTER 1976
(FEB.)
WINTER 1977
(FEB.)
WINTER 1978
(FEB.)
FALL 1978 WINTER 1979
(OCT.) (FEB.)
WINTER 1980
(FEB.)
I
I
I
I
10 15 20 25 30 35
CHRONOLOGICAL ORDER, 1 unit = 28 days
40
45
50
55
Figure 11 -7. Average blood lead levels of U.S. population aged 6 months74 years. United States,
February 1976February 1980, based on dates of examination of NHANES II examinees with
blood lead determinations.
Source: Annest et al. (1983a).
-------
Wj DU
Ul
0 40
s-
ca
5 20
O
1.
O
c
Z 0
OVERALL BLACK
WHITE MALE FEMALE 0.5-5 6-17 18-74
CL RACE BEX AGE IN YEARS
Figure 11 -8. Reduction in mean blood lead levels, according to race, sex,
and age. Data on sex and age are for whites.
Source: Annest et al. (1983a).
11-33
-------
amount of lead used in gasoline production over the same time period (as discussed in more
detail in Section 11.3.6.1).
11.3.5.2 Time Trends in the Childhood Lead Poisoning Screening Programs. Billick and col-
leagues have analyzed the results of blood lead screening programs conducted by the City of
New York (Billick et al., 1979; Billick, 1982). Most details regarding this data set were al-
ready described, but Table 11-10 summarizes relevant methodologic information for these analy-
ses and for analyses done on a similar data base from Chicago, Illinois. The discussion of
the New York data below is limited to an exposition of the time trend in blood lead levels
from 1970 to 1977.
Geometric mean blood lead levels decreased for all three racial groups and for almost all
age groups in the period 1970-76 (Table 11-6). Table 11-11 shows that the downward trend
covers the entire range of the frequency distribution of blood lead levels. The decline in
blood lead levels showed seasonal variability, but the decrease in time was consistent for
each season. The 1977 data were supplied to EPA by Dr. Billick.
In addition to this time trend observed in New York City, Billick (1982) examined similar
data from Chicago and Louisville. The Chicago data set was much more complete than the Louis-
ville one, and was much more methodologically consistent. Therefore, the Chicago data will
mainly be discussed here. The lead poisoning screening program in Chicago may be the longest
continuous program in the United States. Data used in this report covered the years 1967-
1980. Because the data set was so large, only a 1 in 30 sample of laboratory records was
coded for statistical analysis (similar to procedures used for New York described above).
The blood lead data for Chicago contains samples that may be repeats, confirmatory analy-
ses, or even samples collected during treatment, as well as initial screening samples. This
is a major difference from the New York City data, which had initial screening values only.
Chicago blood lead levels were all obtained on venous samples and were analyzed by one labora-
tory, the Division of Laboratories, Chicago Department of Health. Lead determinations were
done by atomic absorption. Racial composition was described in more detail than for New York,
but analysis showed there was no difference among the non-blacks, so they were pooled in the
final analysis.
Table 11-10 displays important characteristics of the Chicago and New York screening pro-
grams, including the number of observations involved in these studies. From tables in the ap-
pendices of the report (Billick, 1982), specific data on geometric mean blood lead values,
race, sex, and sampling data for both cities are available. Consistency of the data across
cities is depicted in Figure 11-9. The long-term trends are quite consistent, although the
seasonal peaks are somewhat less apparent. Although the data displayed are only for blacks
aged 25 to 36 months, very similar data are available for whites and other groups covered by
the study.
11-34
-------
1970 1971 1972 1973 1974 1976 1970 1977 1978 1979 1980
YEAR (Beginning Jan. 1)
Figure 11-9. Time dependence of blood lead levels for blacks, aged
25-36 months, in New York City and Chicago.
Source: Adapted from Billick (1982).
11-35
-------
TABLE 11-10. CHARACTERISTICS OF CHILDHOOD LEAD POISONING SCREENING DATA
New York
Chicago
Time period
Sampling technique
Analytic technique
Laboratory
Screening status
Race classification
and total number of
samples used in
analysis*
Raw data
Gasoline data
1970 - 1979
Venous
AAS
(Hasel method)
In house
Available/unknown
Unknown
White
Black
Hispanic
Other
TOTAL
69,658
5,922
51,210
41,364
4,398
172,552
Decade grouped
Tri-state (NY, NJ, CT)
1970 - 1979
SMSA 1974 - 1979
1967 - 1980 (QTR 2)
Venous
AAS
(Hasel method)
In house
Unavailable
Nonblack 6,459
Black 20,353
TOTAL 26,812
Ungrouped
SMSA
*New York data set only includes first screens while Chicago includes also
confirmatory and repeat samples.
TABLE 11-11. DISTRIBUTION OF BLOOD LEAD LEVELS FOR 13- TO 48-
MONTH-OLD BLACKS BY SEASON AND YEAR* FOR NEW YORK SCREENING DATA
January - March
Percent
Year <15Mg/dl 15 - 34ug/dl >34ug/dl
July - September
Percent
<15ug/dl 15 - 34ug/dl >34ug/dl
1970
1971
1972
1973
1974
1975
1976
1977
(insufficient sample size)
3.8
4.4
7.3
9.2
11.1**
21.1
28.4
69.5
76.1
80.3
73.8
77.5**
74.1
66.8
26.7
19.5
12.4
17.0
11.4**
4.8
4.8
3.4
1.3
4.3
2.7
8.2
7.3**
11.9
19.9
54.7
56.0
72.2
62.4
65.4
81.3**
75.8
72.9
42.0
42.7
23.4
34.9
26.4
11.4**
12.3
7.2
* data provided by I.H. Billick (1982).
**Percentages estimated using interpolation assuming a lognormal distribution.
11-36
-------
11.3.5.3 Newark. Cause et al. (1977) present data from Newark, New Jersey, that reinforce
the findings of Billick and coworkers. Gause et al. studied the levels of blood lead among 5-
and 6-year-old children tested by the Newark Board of Education during the academic years
1973-74, 1974-75, and 1975-76. All Newark schools participated in all years. Participation
rates were 34, 33, and 37 percent of the eligible children for the three years, respectively.
Blood samples collected by fingerstick onto filter paper were analyzed for lead by atomic
absorption spectrophotometry. The authors point out that fingerstick samples are more subject
to contamination than venous samples; and that because erythrocyte protoporphyrin confirmation
of blood lead values greater than 50 ng/dl was not done until 1974, data from earlier years
may contain somewhat higher proportions of false positives than later years.
Blood lead levels declined markedly during the 3-year study period. The percentage of
children with blood lead levels less than 30 ug/dl went from 42 percent for blacks in 1973-74
to 71 percent in 1975-76; similarly, the percentages went from 56 percent to 85 percent in
whites. The percentage of high risk children (>49 ug/dl) dropped from 9 to 1 percent in
blacks and from 6 to 1 percent in whites during the study period. Unfortunately, no com-
panion analysis was presented regarding concurrent trends in environmental exposures.
Foster et al. (1979), however, reported a study from Newark that examined the effective-
ness of the city's housing deleading program, using the current blood lead status of children
who had earlier been identified as having confirmed elevated blood lead levels; according to
the deleading program, these children's homes should have been treated to alleviate the lead
problem. After intensive examination, the investigators found that 31 of the 100 children
studied had lead-related symptoms at the time of Foster's study. Examination of the records
of the program regarding the deleading activity indicated a serious lack of compliance with
the program requirements. Given the results of Foster's study, it seems unlikely that the
observed trend was primarily caused by the deleading program.
11.3.5.4 Boston. Rabinowitz and Needleman (1982) studied umbilical cord blood lead levels
from 11,837 births between April, 1979 and April, 1981 in the Boston area. These represented
97 percent of the births occurring in a hospital serving a diverse population. Blood samples
were analyzed for lead by anodic stripping voltammetry after stringent quality control proce-
dures were used. External quality control checks were done by participation in the Blood Lead
Reference Program, conducted by the Centers for Disease Control. The average difference
between the investigators' results and the reference lab was 1.4 ug/dl.
The overall mean blood lead concentration was 6.56 ± 3.19 uQ/dl (standard deviation) with
a range from 0.0 to 37.0 (jg/dl. After regression of the individual values of blood lead
against the date of birth, a significant downward trend in blood levels was observed (~0.89
ug/dl/yr), representing a decrease of 14 percent per year (Figure 11-10). Figure 11-10 also
11-37
-------
12.0
s 10.0
2
8.0
O
o
O
*
c
ui
Q
O
O
6.0
4.0
Model Predicted
Actual Data
I I
I
4/79 7/79 10/79 1/80 4/80 7/80
MONTH AND YEAR OF COLLECTION
10180
1/81
4/81
Figure 11-10. Modeled umbilical cord blood lead levels by date of
sample collection for infants in Boston.
Source: Rabinowitz and Needleman (1982).
11-38
-------
illustrates the complicating aspect of seasonal trends in evaluating underlying secular
trends. The observed trend is similar to that noted in the NHANES II study described earlier.
Rabinowitz and Needleman (1982) list the following as possible causes of the decline: (1)
modification of the water supply to decrease the lead content; (2) reduction of the use of
lead in gasoline; (3) reduction in contamination of food by solder; and 4) changes in prenatal
practices, such as smoking or iron supplementation.
Rabinowitz and Needleman (1983) then sought to evaluate statistically possible reasons
for the observed two-year downward trend in umbilical cord blood lead levels. The authors
used pairwise product moment correlations for the monthly cord lead levels (about 500 per
month) and monthly amounts of gasoline lead in Massachusetts. A strong correlation was ob-
served: with the same month's data, the correlation coefficient was 0.716, which increased to
a peak correlation coefficient of 0.758 when a 1-month lag time was used. The authors indi-
cate that they did not observe similar trends in maternal tobacco smoking, education level,
and alcohol consumption. They did observe a positive (instead of negative) trend in tap water
lead concentrations. They conclude that gasoline lead exposure changes were probably the
cause of the observed trend in blood lead levels.
From the ongoing surveillance of consecutive births, Rabinowitz et al. (1984) also iden-
tified a cohort of 249 infants who were enrolled in an ongoing cohort study after meeting
certain eligibility standards. Indoor air was sampled for lead from the homes of children
when each child was 6, 18, and 24 months of age. Tapwater was collected after a 4-liter flush,
at 1 and 6 months of age. Seasonal biases in indoor/outdoor air lead ratios and the amounts
of time spent indoors may have been confounding variables which may have distorted upward the
underlying inhalation slope to the observed value near nine.
For each month there was generally available a mean air lead from 12 homes, water lead
from 23 homes, and blood leads for 500 births. The study period covered March, 1980 to April,
1981. The blood leads were then correlated with gasoline lead sales, indoor air, and tap-
water. A linear (although somewhat scattered) trend was found between lead in indoor air and
gasoline lead sales. Forty-eight percent of the variance in air lead could be accounted for
by the gasoline lead sales. Air lead and blood lead levels were highly correlated. The best
linear fit (r = 0.71) has a slope of 9 ug/dl/ug/m3 and an intercept of 4.9 ug/dl. No correla-
tion was observed between water and blood lead levels. Interestingly, a higher correlation
was found between gasoline lead sales and blood lead levels than between air lead and blood
lead.
Karalekas et al. (1983) report additional data from the Boston metropolitan area. Re-
sults of the lead screening program indicate that the percentage of screened children with
elevated blood lead levels declines over the period 1976-1981. Data on lead in water for this
11-39
-------
period are also presented. Water lead levels began to decline after the decline in blood lead
levels. This relationship in this data warrants further research.
11.3.5.5 Lead Studies in the United Kingdom. There has been a series of publications from
various workers in England who have been examining the question of whether or not time trends
in blood lead levels exist there as well as in the United States (Oxley, 1982; Elwood,
1983a,b; Quinn, 1983). These papers cover a variety of exposure situations and populations
studied. All of them obtained findings analogous to those described above for the United
States, in that there has been a general decline in blood lead levels over the decade of the
1970's; they differ, however, with regard to the magnitude of the decline, when the decline
began, and to what extent the decline may be attributable to a particular source of lead.
Oxley (1982) reported an analysis of blood lead levels found in blood samples drawn as a
part of preemployment medical examinations conducted by a major U.K.-based oil company during
1967-69 and 1978-80. Blood samples were collected by venipuncture and analyzed for lead by
two different methods. A comparative laboratory study also reported by Oxley suggested that
the data could be adjusted from one method to the other. Geometric mean blood lead levels de-
clined from 20.2 to 16.6 ug/dl.
Elwood (1983a) reported a time trend analysis of blood lead levels observed in adult
women studied over a 10-year period in eight surveys conducted in a variety of locations in
Wales. These were analyzed and examined for trends in blood lead levels. All women included
in this analysis came from surveys which were designed to generate representative samples of
adult women in residential areas. A high response rate (90 percent or more) was obtained in
each of the surveys. Venous blood samples were collected and analyzed for lead. A single
laboratory performed all of the analyses with an external reference laboratory performing
quality control checks in some of the surveys. Overall mean blood lead levels for the various
surveys fell more than 30 percent over the period 1972-1982. Two of the surveys were con-
ducted in the same area. Between 1974 and 1982, the mean blood lead concentration fell 37
percent. Surveys from mining areas showed that women there had higher blood lead levels than
in non-mining areas.
Elwood acknowledges that laboratory drift may be present in the data and also that the
surveys did not generate strictly comparable samples. Still, the observed decline was thought
to be real. No statistical analysis of the data is presented to examine the possible reasons
for the observed decline, but a number of possible environmental reasons were discussed. Re-
duced gasoline lead exposures as a reason were dismissed on the basis that while the lead con-
centration in gasoline had indeed declined, the overall use of petrol in England had in-
creased, therefore balancing the reduction. However, no data regarding traffic patterns or
gasoline usage in Wales were presented to verify this reasoning. A portion (amount unspeci-
11-40
-------
fied) of the reduction was attributed to a drop in dietary intake of lead due to the reduced
use of canned foods.
Elwood (1983b) also presents data from a more homogeneous setting. In 1969 a hematologic
survey of a random sample of 4070 women was conducted in one town in Wales. Detailed studies
were made of 121 of these women whose hemoglobin levels were below 10.5 g/100 ml. Samples of
their whole blood were deep frozen, and follow-up samples were obtained for some of the same
women in 1982. Follow-up and loss of original samples resulted in there being 26 women with
an available blood lead at both times and who were still living at the same address. The mean
fall in blood lead levels for these women was 23 percent, representing a fall of 3.5 ug/100
ml. Again Elwood does not attribute the decline to changes in gasoline lead or water supply,
but instead suggests that it may be due to changes in dietary intake although noting there are
no data on which to base a judgment.
King (1983), in commenting on the results of Elwood (1983a), noted that the blood lead
values before 1975 were probably falsely elevated due to matrix problems in the chemical ana-
lysis. This means the magnitude of the observed decline is probably less than that quoted by
Elwood (1983b). King (1983) further examined the question of the time trend by controlling
for region of Wales and reported that Elwood1s data showed a 50 percent increase in blood lead
levels from 1981 to 1982, a most unlikely outcome. Pirkle and Annest (1984) have also criti-
cized the Elwood (1983a) paper and concluded that various factors make reliable interpreta-
tions of Elwood1s data extremely difficult.
Quinn (1983) reports on the summarized findings of two large-scale survey effects in 1979
and 1981. Broad comparisons within the same authority showed an overall reduction approaching
10 percent (1 ug/100 ml). Quinn himself states, however, that these two survey efforts are
not strictly comparable in that the first round focused on representative population groups
while the second round focused on areas where lead may have presented a problem. No effort
was made to attribute the decline in blood lead levels to a particular source.
11.3.5.6 Other Studies. Okubo et al. (1983) examined a total of 1933 children from 5 to 18
years of age for blood lead using the Hessel method over the period 1975 to 1980 in an urban
area of Tokyo and in a nearby suburban area. The analysis of all blood lead was done by the
same laboratory. Over the time period of the study an apparent decrease in blood lead is
shown. A part of the difference in blood lead between urban and suburban groups is related to
the difference in average lead concentrations between the two areas. The difference of blood
lead between urban and suburban becomes greater when the comparison of blood lead between the
two areas is executed only among children who have lived in the same areas from their birth.
In an international study discussed in detail earlier, Friberg and Vahter (1983) compared
data on blood lead levels obtained in 1967 with data for 1981 (see Table 11-12). For areas of
11-41
-------
TABLE 11-12. COMPARISON OF MEDIAN BLOOD LEAD LEVELS (|jg/dl) IN SEVERAL COUNTRIES
FROM STUDIES OF GOLDWATER AND HOOVER (1967) AND FRIBERG AND VAHTER (1983)
Country
Japan
Israel
United States
Yugoslavia
Median blood lead
1967
21.0
15.0
18.0
15.0
Median blood lead
1981
6.0
8.2
7.5
9.2
% change
from 1967
71
45
58
39
the world where there were data collected by Goldwater and Hoover (1967) as well as the UN/WHO
study, there has been a substantial reduction in reported blood lead levels. A cautionary
note must be made, however, that the analytic and human sampling procedures are not the same
in the two studies. Therefore these data should be thought of as providing further but
limited evidence supporting a recent downward trend in blood lead levels worldwide.
11.3.6 Gasoline Lead as an Important Determinant of Trends in Blood Lead Levels
As noted in the preceding section, explanations have been sought for declining trends in
blood lead levels observed among population groups in the United States and certain other
countries since the early 1970s. Also noted was evidence presented by some investigators
which strongly suggests that gasoline lead usage is a major determinant of the reported down-
ward trends in blood lead levels. The present section examines additional, extensive evidence
which points towards gasoline lead being an important determinant of changes in blood lead
levels associated with exposures to airborne lead of populations in the United States and
elsewhere.
11.3.6.1 NHANES II Study Data. Blood lead data from the second National Health and Nutrition
Examination survey (NHANES II) were described earlier in Sections 11.3.3.1 and 11.3.5.1. One
striking feature of the NHANES II data was a dramatic decline in nationwide average blood lead
levels in the United States during the period (1976 to 1980) of the survey. In evaluating
possible reasons for the observed decrease in the NHANES II blood lead values, Annest et al.
(1983a) found highly significant associations between the declining blood lead concentrations
for the overall U.S. population and decreasing amounts of lead used in gasoline in the U.S.
during the same time period (see Figure 11-11). The associations persisted after adjusting
for race, age, sex, region of the country, season, income, and degree of urbanization (see
Table 11-13). Analogous strong associations (r = 0.95; p < 0.001) were also found for blood
lead levels for white children aged 6 months to 5 years in the NHANES II sample and gasoline
lead usage (Annest et al., 1983a).
11-42
-------
I
-&
GJ
O
O
£
UJ
a
O
-------
11-13. PEARSON CORRELATION COEFFICIENTS BETWEEN THE AVERAGE BLOOD LEAD LEVELS
FOR SIX-MONTH PERIODS AND THE TOTAL LEAD USED IN GASOLINE PRODUCTION
PER SIX MONTHS, ACCORDING TO RACE, SEX, AND AGEa
Overall (all races)
All
All
By
By
black6
whites
sex: Male
Female
age: 0.5-5 yr
6-17 yr
18-74 yr
Coefficients
January-June
and July-December
0.920
0.678
0.929
0.944
0.912
0.955
0.908
0.920
for 6-month periods
April -September .
and October-March
0.938
0.717
0.955
0.960
0.943
0.969
0.970
0.924
Averages
0.929
0.698
0.942
0.952
0.928
0.962
0.939
0.922
The lead values used to compute the averages were preadjusted by regression analysis to
account for the effects of income, degree of urbanization, region of the country, season,
and, when appropriate, race, sex, and age.
All correlation coefficients were statistically significant (p < 0.001) except those for
blacks (p < 0.05).
cAverages were based on six-month periods, except for the first and last time periods ,
which covered only February 1976 through June 1976 and January 1980 through February 1980,
respectively.
Averages were based on six-month periods, except for the last time period, which covered
only October 1979 through February 1980.
eBlacks could not be analyzed according to sex and age subgroups because of inadequate sample
sizes.
Questions have been raised by some commentors regarding whether or not (1) the NHANES II
survey design was adequate to allow for credible definition of time trends for nationwide
average blood lead concentrations, (2) the reported significant associations between NHANES II
blood lead data and U.S. gasoline usage are credible and reflect a causal relationship, and
(3) the entire decline in blood lead values is attributable to decreased gasoline lead usage
versus changes in other sources of lead exposure. These issues and alternative analyses con-
cerning the NHANES II blood lead/gasoline lead relationships were evaluated by an expert panel
(the NHANES II Time-Trend Analysis Review Group) convened by EPA.
11-44
-------
The NHANES II Time-Trend Analysis Review Group (1983) found the following: (1) strong
evidence that there was a substantial decline in the average level of blood lead in the U.S.
population during the NHANES II survey period; (2) after adjustment for relevant demographic
covariables, the magnitude of the change can be estimated for the total U.S. population and
for some major subgroups, provided careful attention is given to underlying model assumptions.
The Review Group also found a strong correlation between gasoline-lead usage and blood-lead
levels, and noted that in the absence of scientifically plausible alternative explanations,
the hypothesis that gasoline lead is an important causal factor for blood-lead levels must
receive serious consideration. Nevertheless, despite the strong association between the
decline in gasoline-lead usage and the decline in blood-lead levels, the survey results and
statistical analyses do not confirm the causal hypothesis. Rather, this finding is based on
the qualitatively consistent results of extensive analyses done in different but complementary
ways.
Further support for strong, likely causative, relationships between gasoline lead usage
and blood lead levels in the U.S. is provided by analyses carried out by Schwartz et al.
(1984). Those analyses not only evaluated NHANES II data, but, also, additional blood lead
data such as blood lead values from U.S. childhood lead-screening programs. Results obtained
were quite similar to those of Annest et al. (1983b), even after controlling for possible
alternative contributors to the blood lead decline, e.g., deleading of lead-painted housing
units or decreased food lead intake. Large numbers (thousands) of children were also esti-
mated by the analysis to have blood lead levels in excess of 30 pg/dl due in part to exposures
to lead emitted as a consequence of leaded gasoline usage in the United States.
Still further evidence for causative relationships between gasoline lead usage and
changes in human blood lead levels is provided by isotope studies of the type described next.
11.3.6.2 Isotope Studies. Two field investigations have attempted to derive estimates of the
amount of lead from gasoline that is absorbed by the blood of individuals. Both of these in-
vestigations used the fact that non-radioactive isotopes of lead are stable. The varying pro-
portions of the isotopes present in blood and environmental samples can indicate the source of
the lead. The Isotopic Lead Experiment (ILE) is an extensive study that attempted to use dif-
fering proportions of the isotopes in geologic formations to infer the proportion of lead in
gasoline that is absorbed by the body. The other study used existing natural shifts in iso-
topic proportions in an attempt to do the same thing.
11.3.6.2.1 Italy. The ILE is a large-scale community study in which the geologic source of
lead for antiknock compounds in gasoline was manipulated to change the isotopic composition of
the atmosphere (Garibaldi et al., 1975; Facchetti, 1979; Facchetti, 1985). Preliminary inves-
tigation of the environment of Northwest Italy, and the blood of residents there, indicated
11-45
-------
that the ratio of 206Pb/207Pb in blood was a constant, about 1.16, and the ratio in gasoline
was about 1.18. This preliminary study also suggested that it would be possible to substitute
for the currently used geologic sources of lead for antiknock production a geologically dis-
tinct source of lead from Australia that had an isotopic 206Pb/207Pb ratio of 1.04. It was
hypothesized that the resulting change in blood lead 206Pb/z07Pb ratios (from 1.16 to a lower
value) would indicate the proportion of lead in the blood of exposed human populations attri-
butable to lead in the air contributed by gasoline combustion in the study area.
Baseline sampling of both the environment and residents in the geographic areas of the
study was conducted in 1974-1975. The sampling included air, soil, plants, lead stock, gaso-
line supplies, etc. Human blood sampling was done on a variety of populations within the
area. Both environmental and human samples were analyzed for lead concentrations as well as
isotopic 206Pb/207Pb composition.
In August, 1975, the first switched (Australian lead-labeled) gasoline was introduced;
although it was originally intended to get a 100 percent substitution, practical and logisti-
cal problems resulted in only a 50 percent substitution being achieved by this time. By May,
1977, these problems were worked out and the substitution was practically complete. The sub-
stitution was maintained until the end of 1979, when a partial return to use of the original
sources of lead began. Therefore, the project had four phases: phase zero - background;
phase one - partial switch; phase two - total switch; and phase three - switchback.
Airborne lead measurements were collected in a number of sites to generate estimates of
the lead exposure that was experienced by residents of the area. Turin, the major city of the
region, was found to have a much greater level of atmospheric lead than the surrounding coun-
tryside. There also appeared to be fairly wide seasonal fluctuations.
The isotopic lead ratios obtained in the samples analyzed are displayed in Figure 11-12.
It can easily be seen that the airborne particulate lead rapidly changed its isotope ratio in
line with expectations. Changes in the isotope ratios of the blood samples appeared to lag
somewhat behind. Background blood lead ratios for adults were 1.1591 ± 0.0043 in rural areas
and 1.1627 ± 0.0022 in Turin in 1975. For Turin adults, a mean isotopic ratio of 1.1325 was
obtained in 1979, clearly less than background. Isotopic ratios for Turin schoolchildren,
obtained starting in 1977, tended to be somewhat lower than the ratios for Turin adults.
Preliminary analysis of the isotope ratios in air lead allowed for the estimation of the
fractional contribution of gasoline in the city of Turin, in small communities within 25 km of
Turin, and in small communities beyond 25 km (Facchetti and Geiss, 1982). At the time of
maximal use of Australian lead isotope in gasoline (1978-1979), about 87.3 percent of the air
11-46
-------
1.20
12
24
36
48
TIME, months
60 72
84
96
108
120
132
1.18
1.16
1.14
K
£
1.12
1.10
1.08
1.06
PHASED
>[« PHASE 1-
I 1
1 I I I
O GASOLINE
D BLOOD, ADULTS, TURIN
A BLOOD, ADULTS, >25 km
O BLOOD, ADULTS, <26 km
BLOOD. SCHOOL CHILDREN
BLOOD, TRAFFIC WARDENS
A AIRBORNE PARTICULATE, TURIN
AIRBORNE PARTICULATE, RURAL
-PHASE 2-
-PHASE 3-
I
I
1.04
1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984
YEAR
Figure 11-12. Change in 206Pb/207Pb ratios in gasoline, blood, and airborne particulate from
1974 to 1984.
Source: Facchetti (1985).
11-47
-------
lead in Turin and 58.7 percent of the air lead in the countryside was attributable to gaso-
line. The determination of lead isotope ratios was essentially independent of air lead con-
centrations. During that time, air lead averaged about 2.0 ug/m3 in Turin (from 0.88-4.54
ug/m3 depending on location of the sampling site), about 0.56 ug/m3 in the nearby communities
(0.30-0.67 ug/m3) and about 0.30 ug/m3 in more distant (> 25 km) locations. It is important
to note that the contribution calculations are for local lead in gasoline, not all lead from
gasoline. Large movements of air masses brought in air lead from other regions, especially
for the suburban and urban areas. In the absence of nearby lead industrial sources, this air
lead was at least substantially composed of non-Australian gasoline lead and would therefore
lead to an underestimate of the total contribution of gasoline lead to blood lead.
Blood lead concentrations and isotope ratios for 63 adult subjects were determined on two
or more occasions during phases 0-2 of the study. Their blood lead isotope ratios decreased
over time and the fraction of lead in their blood attributable to the Australian lead-labeled
gasoline could be estimated independently of blood lead concentration (see Appendix C for
estimation method). The mean fraction of blood lead attributable to the Australian lead-
labeled gasoline ranged from 21.4 ± 10.4 percent in Turin to 11.4 ±7.3 percent in the nearby
(< 25 km) countryside and 10.1 ±9.3 percent in the remote countryside. These likely represent
minimal estimates of fractions of blood lead derived from gasoline due to the following
reasons: (1) use of some non-Australian lead-labeled gasoline brought into the study area
from outside; (2) probable insufficient time to have achieved steady-state blood lead isotope
ratios by the time of the switchback; and (3) probable insufficient time to fully reflect de-
layed movement of the Australian lead from gasoline via environmental pathways in addition to
air.
These results can be combined with the actual blood lead concentrations to estimate the
fraction of gasoline uptake attributable or not attributable to direct inhalation. The
results are shown in Table 11-14 based upon the concept outlined in Facchetti and Geiss
(1982). From Section 11.4.1, we conclude that an assumed value of 6=1.6 is plausible for
predicting the amount of lead absorbed into blood at air lead concentrations less than 2.0
ug/m3. The predicted values for lead from gasoline in air (in the ILE) range from 0.28 to
2.79 ug/dl in blood due to direct inhalation. The total contribution to blood lead from
gasoline is much larger, from 3.21 to 4.66 ug/dl, suggesting that the non-inhalation con-
tribution of gasoline increases from 1.88 ug/dl in Turin to 2.33 ug/dl in the near region and
2.93 ug/dl in the more distant region. The non-inhalation sources include ingestion of dust
and soil lead, and lead in food and drinking water. Efforts are being made to quantify the
magnitude of these sources. The average direct inhalation of lead in the air from gasoline
11-48
-------
TABLE 11-14. ESTIMATED CONTRIBUTION OF LEADED GASOLINE TO BLOOD LEAD
BY INHALATION AND NON-INHALATION PATHWAYS
Location
Turin
<25 km
>25 km
Air Pb
fraction
from
??so(a)
linev '
0.873
0.587
0.587
Mean
air
Pb (b)
cone. ,v *
pg/m3
2.0
0.56
0.30
Blood Pb
fraction
from
0.214
0.114
0.101
Mean
blood
Pb , .
cone. ,
pg/dl
21.77
25.06
31.78
Blood
Pb
from
) gaso"(e)
' line/6'
pg/dl
4.66
2.86
3.21
Pb
from
gaso-
I1ne. (f)
in air,v J
pg/dl
2.79
0.53
0.28
Non-
inhaled
Pb from
gaso (g)
pg/dl
1.88
2.33
2.93
Estimated
fraction
gas-Pb
inhalaT
tion(R)
0.60
0.19
0.09
(a)
(b)
Fraction of air lead in Phase 2 attributable to lead in gasoline.
Mean air lead in Phase 2, pg/m3.
'c'Mean fraction of blood lead in Phase 2 attributable to lead in gasoline.
(d)
Mean blood lead concentration in Phase 2, pg/dl.
^Estimated blood lead from gasoline = (c) x (d)
' ^Estimated blood lead from gasoline inhalation = B x (a) x (b), B = 1.6.
^Estimated blood lead from gasoline, non-inhalation = (f)-(e)
^Fraction of blood lead uptake from gasoline attributable to direct inhalation = (f)/(e)
Data: Facchetti and Geiss (1982); Facchetti (1985).
is 9 to 19 percent of the total intake attributable to gasoline in the countryside and an
estimated 60 percent in the city of Turin. Note that in this sample, the blood lead con-
centrations were lowest in the city and highest in the more remote areas. This is not
obviously attributable to sex because the city sample was all male. Facchetti (1985) notes
that factors unaccounted for are presumably acting on the population of the ILE test area.
The lead concentration in tapwater in Turin is approximately 4 pg/1, while it ranges in the
country from 12 to 20 pg/1. Also, lead concentrations in Piedmont wines averaged 155 ± 67
pg/1. Daily wine consumption for rural drinkers ranges from 0.5 to 1 liter per day. Thus the
importance of wine consumption becomes evident. Other differences between city and county may
play a role. A more detailed statistical investigation is needed.
Spengler et al. (1984) have developed a modeling approach to try to explain these
results. Their hypothesized model suggests that in-vehicle lead exposure is important and may
explain part of the apparent anomaly of the blood lead levels in this study. That is,
Spengler et al. (1984) hypothesized that there is a large component of personal lead exposure
associated with gasoline use that is not captured by stationary ambient air lead monitors:
11-49
-------
personal exposure while riding in and working around motor vehicles using leaded gas. More
work on this problem is needed, particularly conduction of near- and in-vehicle studies.
Lead uptake may also be associated with occupation, sex, age, smoking, and drinking
habits. The linear exposure model used in Section 11.4 was also used here to estimate the
fraction of labeled blood lead from gasoline attributable to exposure via direct inhalation
and other pathways. EPA used the data in Facchetti and Geiss (1982) for the 35 subjects for
whom repeated measurements allowed estimation of the change in isotope ratios in the blood.
Their blood lead concentrations in Phase 2 were also determined, allowing for estimation of
the total gasoline contribution to blood lead. Possible covariates included sex, age,
cigarette smoking, drinking alcoholic beverages, occupation, residence location, and work
location. In order to obtain some crude comparisons with the inhalation exposure studies of
Section 11.4.1, EPA analyses assigned the air lead values listed in Table 11-15 to various
locations. Lower values for air lead in Turin would increase the estimated blood lead inhala-
tion slope above the estimated value of 1.70. Since the fraction of time subjects were
exposed to workplace air was not known, this was also estimated from the data as about 41
percent (i.e., 9.8 hours/ day). The results are shown in Figure 11-13 and Table 11-16. Of
all the available variables, only location, sex, and inhaled air lead from gasoline proved
statistically significant in predicting blood lead attributable to gasoline. The model
predictability is fairly good, with an R2 value of 0.654. It should be noted that a certain
amount of confounding of variables was unavoidable in this small set of preliminary data,
e.g., no female subjects in Turin or in occupations of traffic wardens, etc. There was a
systematic increase in estimated non-inhalation contributions from gasoline use for remote
areas, but the cause is unknown. The following interpretation for these results may be
offered: The air lead measurements used here represent community or ambient exposures. In
addition to the ambient air lead, there may have also been systematic differences in personal
exposure. Nevertheless, the estimated non-inhalation contribution of gasoline to blood lead
in the ILE study is significant (i.e., 1.8-3.4 ug/dl).
TABLE 11-15. ASSUMED AIR LEAD CONCENTRATIONS FOR MODEL
Residence or workplace code
Location
Air lead concentration
1-4
outside Turin
(a)
5
Turin residential
1.0 ug/m3(b)
6
Turin central
2.5 ug/m3(c)
(a) Use value for community air lead, 0.16 - 0.67 ug/m3.
(b) Intermediate between average traffic areas (1.71 ug/m3) and low traffic areas (0.88 ug/m3)
in Turin.
(c) Intermediate between average traffic areas (1.71 ug/m3) and heavy traffic areas (4.54
ug/m3) in Turin.
11-50
-------
Uj
3
O
isi
<
(3
uj 3
CD
t-
25km <25km
0.5
TURIN
1.0
1.5
2.0
AVERAGE AIR LEAD CONCENTRATION ATTRIBUTABLE TO GASOLINE
Figure 11-13. Estimated direct and indirect contributions of lead in
gasoline to blood lead in Italian men, based on EPA analysis of ILE data
(Table 11 -16).
TABLE 11-16. REGRESSION MODEL FOR BLOOD LEAD ATTRIBUTABLE TO GASOLINE
Variable
Coefficient ± standard error
Air lead from gas
Location
Turin
<25 ton
>25 km
Sex
1.70 4 1.04 pg/dl per
1.82 t 2.01 pg/dl
2.56 t 0.59 pg/dl
3.42 ± 0.85 pg/dl
2.03 ± 0.48
for women
11-51
-------
The preliminary linear analysis of the overall ILE data set (2161 observations) found
that total blood lead levels depended on other covariates for which there were plausible
mechanisms of lead exposure, including location, smoking, alcoholic beverages, age, and occu-
pation (Facchetti and Geiss, 1982). The difference between total blood lead uptake and blood
lead uptake attributable to gasoline lead has yet to be analyzed in detail, but these analyses
suggest that certain important differences may be found. Some reservations have been expres-
sed about the ILE study, both by the authors themselves and by Elwood (1983c). These include
unusual conditions of meteorology and traffic in Turin, and demographic characteristics of the
35 subjects measured repeatedly that may restrict the generalizabi1ity of the study.
Facchetti (1985) reports additional analysis which increases the number of blood leads from 35
to 63, alleviating this concern to some extent since the new results confirm the old. How-
ever, it is clear that changes in air lead attributable to gasoline were tracked by changes in
blood lead in Turin residents. The airborne particulate lead isotope ratio quickly achieved
new equilibrium levels as the gasoline isotope ratio was changed, and maintained that level
during the 2% years of Phase 2. The blood lead isotope ratios fell slowly during the change-
over period, and rose again afterwards as shown in Figure 11-12. Equilibrium was not clearly
achieved for blood lead isotope ratios, possibly due to large endogenous pools of old lead
stored in the skeleton and slowly mobilized over time. Even with such reservations, this
study provides a useful basis for relating blood lead and air lead derived from gasoline com-
bustion. Colombo and Fantechi (1983) have presented an analysis of the ILE study using a
dynamic model. The results of their analysis suggest that an appropriate estimate of the con-
tribution of locally consumed gasoline lead to blood lead is 26, 17, and 14 percent for the
subject groups of Turin, and near and far countryside, respectively. These values are similar
to but somewhat larger than those presented by Facchetti and Geiss (1982) and Facchetti
(1985).
11.3.6.2.2 United States. Manton (1977) conducted a long-term study of 10 subjects whose
blood lead isotopic composition was monitored for comparison with the isotopic composition of
the air they breathed. Manton had observed that the ratio of 2oepb/204pb in the a^r van-ed
with seasons in Dallas, Texas; therefore, the ratio of those isotopes should vary in the
blood. By comparing the observed variability, estimates could then be made of the amount of
lead in air that is absorbed by the blood.
Manton took monthly blood samples from all 10 subjects from April, 1974 until June, 1975.
The blood samples were analyzed for both total lead and isotopic composition. The recruited
volunteers included a mix of males and females, and persons highly and moderately exposed to
lead. However, none of the subjects was thought to be exposed to more than 1 ug/m3 of lead in
air. Lead in air samples was collected by hi-vol samplers primarily from one site in Dallas.
That site, however, had been shown earlier to vary in isotopic composition paralleling another
11-52
-------
site some 16 miles away. All analyses were carried out under clean conditions with care and
caution being exercised to avoid lead contamination.
The isotope ratio of 206Pb/204Pb increased linearly with time from about 18.45 to 19.35,
approximately a 6 percent increase. At least one of the two isotopic lead ratios increased
linearly in 4 of the 10 subjects. In one other, they increased, but erratically. In the
remainder of the subjects, the isotopic ratios followed smooth curves showing inflection
points. The curves obtained for the two subjects born in South Africa were 6 months out of
phase with the curves of the native-born Americans. The fact that the isotope ratios in 9 of
the 10 subjects varied regularly was thought to indicate that the non-airborne sources of lead
varied in isotopic composition very slowly.
The blood lead levels exhibited a variety of patterns, although none of the subjects
showed more than a 25 percent change from initial levels. This suggests a reasonably steady-
state external environment.
Manton carried his analyses further to estimate the percentage of lead in blood that
comes from air. He estimated that the percentage varied from 7 to 41 percent, assuming that
dietary sources of lead had a constant isotopic ratio while air varied. He calculated the
percent contribution according to the following equation.
, where
100+q a (11-1)
b = rate of change of an isotope ratio in blood,
a = rate of change of the same ratio in the air, and
q = constant defined as the number of atoms of the isotope in the
denominator of the airborne lead ratio mixed with 100 atoms of
the same isotope of lead from non-airborne sources.
The results are shown in Table 11-17. Slopes were obtained by least squares regression.
Percentages of airborne lead in blood varied between 7 ± 3 and 41 ± 3.
Stephens (1981) extended the analysis of data in Manton1s study (Table 11-18). He used
the observed air lead concentrations based on actual 24-hour air lead exposures in three
adults. He assumed values for breathing rate, lung deposition, and absorption into blood to
estimate the blood lead uptake attributable to 204Pb by the direct inhalation pathway. Sub-
jects 5, 6, and 9 absorbed far more air lead in fact than was calculated using the values in
Table 11-17. The total air lead contribution for those subjects was 8.4, 4.4, and 7.9 times,
respectively, larger than the direct inhalation. These estimates are sensitive to the assumed
parameter values.
11-53
-------
TABLE 11-17. RATE OF CHANGE OF 206pb/204pb AND 206pb/207pb
IN AIR AND BLOOD, AND PERCENTAGE OF AIRBORNE LEAD IN BLOOD OF SUBJECTS 1, 3, 5, 6, AND 9
Subject
(Air)
1
3
5
6
9*
Rate of change per day
206pb/204pb 206pb/207f>b
X 10" X 10"
17.60 ± 0.77 9.97 ± 0.42
. . . 0.70 ± 0.30
5.52 ± 0.55 ...
... 3.13 ± 0.34
6.53 ± 0.49 4.10 ± 0.25
3.25 2.01
Percentage of airborne lead in blood
From From
206pb/207pb 206Pb/207Pb
... ...
... 7 ± 3
31.4 ±3.4 ...
. . . 31.4 ± 3.7
37.1 ± 2.8 41.1 ± 3.0
18.5 20.0
Note: Errors quoted are one standard deviation
*From slope of tangent drawn to the minima of subject's blood curves. Errors
cannot realistically be assigned.
TABLE 11-18. CALCULATED BLOOD LEAD UPTAKE FROM AIR LEAD USING MANTON ISOTOPE STUDY
Blood uptake from air
Sub-
ject
5
6
9
Concen-
tration
0.22
1.09
0.45
ug/m3
M9/m
Mg/m
Expo-
sure*
15 m3/day
15 m3/day
15 m3/day
Deposi-
tion*
37%
37%
37%
Absorp-
tion*
50%
50%
50%
0.
3.
1.
Calcu-
lated
inhala-
tion
61 ug/d
0 ug/d
2 ug/d
Fraction of
lead uptake
from gasoline
Observed
5.1
13.2
9.9
ug/d
Mg/d
ug/d
by
direct
inhalation
0
0
0
.120
.229
.126
*assumed rather than measured exposure, deposition and absorption.
Source: Stephens, 1981, based on Manton, 1977; Table III.
In Manton (1985) the earlier isotope studies were greatly extended and the results
were reinterpreted. The recent study emphasized time changes in blood lead and in
206pb/207pb isotope ratios in three subjects in Dallas, Texas, from 1974 to 1983. Two of the
subjects .described earlier (Manton, 1977) were included here, a husband (subject 8) and his
first wife (subject 9). The more recent subject was the husband's second wife. The husband
11-54
-------
had grown up in South Africa and in England; thus he had deep bone pools of lead that reflec-
ted the Australian lead isotope ratio. As noted earlier, the husband's seasonal minima in
isotope ratio appeared to be the opposite of the two women with whom he shared a very similar
pattern of environmental exposures. Manton (1985) now attributes this to a large efflux of
lead from the skeletal pool. The husband's estimated dietary intake was 55 ug/day. If 10
percent of this is absorbed into blood (5.5 jjg/day), mean residence time of 40 days and volume
of distribution of 75 dl imply a dietary contribution to blood lead of about 3 ug/dl, much
less than his observed average of 17 ug/dl. There was little indication of large changes in
diet lead isotope ratio during this period, hence the changes in blood lead isotope ratio may
be attributed to changes in the air lead particulate isotope ratio, and to changes in isotope
ratio for endogenous sources. Manton attributes the large changes in isotope ratio in the
husband to changes in isotope ratio from lead resorbed from bone into the blood. His estimate
is that approximately 70 percent of the daily blood input is due to the endogenous skeletal
pool of this subject. The subject's wife also exhibited a variety of fluctuations in blood
lead level and isotope ratio due to childbirth and to short-term fluctuations in dietary lead.
The apparent effect of childbirth was to increase resorption of both skeletal calcium and
skeletal lead into blood. The contribution of airborne lead to blood lead isotope ratios thus
did not require correction for long-term secular changes in dietary lead isotope ratios. On
this basis the direct inhalation contribution was again calculated as about 20 to 60 percent
of the total uptake of atmospheric lead using p = 4.1. Manton's calculations are shown in
Table 11-19. The cumulative effects of long-term lead absorption on the mobilizable lead pool
in the skeleton have been ignored, but are apparently not negligible.
In summary, the direct inhalation pathway accounts for only a fraction of the total air
lead contribution to blood, the direct inhalation contribution being on the order of 12-23
percent of the total uptake of lead attributable to gasoline, using Stephen's assumptions, and
20-60 percent based on Manton1s analysis. This is consistent with estimates from the ILE
study, taking into account the much higher air lead levels in Turin.
11.3.6.3 Studies of Childhood Blood Lead Poisoning Control Programs. Billick et al. (1979)
presented several possible explanations for the observed decline (described in Section
11.3.5.2) in blood lead levels in New York City children as well as evidence supporting and
refuting each. The suggested contributing factors include the active educational and screen-
ing program of the New York City Bureau of Lead Poisoning Control, the decrease in the amount
of lead-based paint exposure as a result of rehabilitation or removal of older housing, and
changes in environmental lead exposure.
11-55
-------
TABLE 11-19. RESPIRED AND OTHER INPUTS OF AIRBORNE Pb TO BLOOD FOR SOME DALLAS RESIDENTS IN 1975"
1 »
1
tn
cr>
Subject
no.
3
6
8
9
Blood Pb,
ug/dl
8.4
12.6
5.5
17.4
Total Pb input,
ug/day
15
30
9.3
45
Percent
airborne Pb
in blood
31
39
>33
20
Total input
airborne Pb,
ug/day
4.5
11.7
9.0
Airborne Pb, 24- hr
concentration,
ug/-3
0.22
1.09
0.45C
0.45
. Respired
p ug/day
12 0.91
4.5 4.5
>4.1 1.9
7.7 1.9
Ai rborne
other,
ug/day
3.6
7.2
7.1
Pb
Respired
V
Respired & othe?
20
38
21
Data in first five col inns for subjects 3, 6, and 9 recalculated.
B is defined as the increment in blood Pb concentration (ug/dl) per unit increment in airborne Pb concentration
GFraction of airborne Pb in subject 8 calculated fro* Measured isotope ratios of air, blood, and diet on two
occasions in 1976. Figures quoted are ninina because skeletal input, which would have had an isotope
ratio less than that of the diet, has been ignored.
-------
Information was only available to partially evaluate the last source of lead exposure and
particularly only Jor ambient air lead levels. Air lead measurements were available during
the entire study period for only one station which was located on the west side of Manhattan
at a height of 56 m. Superposition of the air lead and blood lead levels indicated a simi-
larity in seasonal cycle and long-term decline. The authors cautioned against overinterpre-
tation because of the necessary assumptions in this analysis and because one air monitoring
site was used to be representative of the air lead exposure of New York City residents. With
this in mind, the investigators fitted a multiple regression model to the data to try to
define the important determinants of blood lead levels for this population. Age, ethnic
group, and air lead level were all found to be significant determinants of blood lead levels.
The authors further point out the possibility of a change in the nature of the population
being screened before and after 1973. They reran this regression analysis separately for
years both before and after 1973. The same results were still obtained, although the exact
coefficients varied.
Billick et al. (1980) extended their previous analysis of the data from the single moni-
toring site mentioned above. The investigators examined the possible relationship between
blood lead level and the amount of lead in gasoline used in the area. Figures 11-14 and 11-15
present illustrative trend lines in blood leads for blacks and Hispanics versus air lead and
gasoline lead, respectively. Gasoline lead was estimated by multiplying the sales of gasoline
by the estimated concentrations of lead in gasoline. Semiannual concentrations of lead for
the Mid-Atlantic Coast were interpolated to get quarterly values. Sales were computed using
figures for New York, New York plus New Jersey, New York plus Connecticut, or New York plus
New Jersey plus Connecticut: all gave similar results. The lead in gasoline trend line ap-
pears to fit the blood lead trend line better than the air lead trend, especially in the
summer of 1973.
Multiple regression analyses were calculated using six separate models. The best fitting
model had an R2 = 0.745. Gasoline lead content was included rather than air lead. The gaso-
line lead content coefficient was significant for all three racial groups. Partial correla-
tions with gasoline alone were not provided. The authors state a number of reasons for gaso-
line lead providing a better fit than air lead, including the fact that the single monitoring
site might not be representative.
Nathanson and Nudelman (1980) provide more detail regarding air lead levels in New York
City. In 1971, New York City began to regulate the lead content of gasoline sold. Lead in
gasoline was to be totally banned by 1974, but supply and distribution problems delayed the
effect of the ban. Ultimately, regulation of lead in gasoline was taken over by the U.S.
Environmental Protection Agency.
11-57
-------
O
I I I I I I
T
T
11'' 11'' I
i BLACK
_ _-. HISPANIC
AIR LEAD
I I I I I I I I I I I I
E
a.
_i
UJ
2.5 §
2.0
1.5
10
0.0
1970 1971 1972 1973 1974 1"75
OUARTERLY SAMSUNG DATE
1976
o
<
K
UJ
Figure 11-14. Geometric mean blood lead levels of New York City
children (aged 25-36 months) by ethnic group, and ambient air lead
concentration versus quarterly sampling period, 1970-1976.
Source: Billick et at. (1980).
11-58
-------
I I
I I
I I I I I I I I I I
BLACK
HISPANIC
. GASOLINE LEAD
B>
O)
O
r-
Q
6.0 2
5.0 g
O
4.0
3.0
Ol I I I I I I I I I I I I I t I I I I I I I I I I I I I IPO
1970 1971 1972 1973 1974 1975 1976
QUARTERLY SAMPLING DATE
Figure 11-15. Geometric mean blood lead levels of New York City
children (aged 25-36 months) by ethnic group, and estimated amount of
lead present in gasoline sold in New York, New Jersey, and Connecticut
versus quarterly sampling period, 1970-1976.
Source: Billick et al. (1980).
11-59
-------
New York City measured air lead levels during the periods June 1969 to September 1973 and
during 1978 at multiple sites. The earlier monitoring was done by 40 rooftop samplers using
cellulose filters analyzed by AAS. The latter sampling was done by 27 rooftop samplers using
glass fiber filters analyzed by X-ray fluorescence (XRF). There was excellent agreement
between the XRF and atomic absorption analyses for lead (r = 0.985). Furthermore, the XRF
analyses were checked against EPA AAS and again excellent agreement was found. The authors
did, however, point out that cellulose filters are not as efficient as glass fiber filters.
Therefore, the earlier results tend to be underestimates of air lead levels.
Quarterly citywide air lead averages generally declined during the years 1969-1978. The
maximum quarterly citywide average obtained was about 2.5 (jg/m3 for the third quarter of 1970.
The citywide trend corresponds to the results obtained from the single monitoring site used in
Billick et al.'s (1979) analysis. The citywide data suggest that the single monitoring site
in Manhattan is a responsible indicator of air lead level trends. The graph in Figure 11-16
reinforces this assertion by displaying the geometric mean blood lead levels for blacks and
Hispanics in the 25- to 36-month age groups and the quarterly citywide air lead levels for the
periods of interest. A good correspondence was noted.
As part of a detailed investigation of the relationship of blood lead levels and lead in
gasoline covering three cities, Billick (1982) extended the time trend analyses of New York
City blood lead data. Figure 11-17 presents the time trend line for geometric mean blood
leads for blacks aged 25-36 months extended to 1979. Similar results held for other ages.
The downward trend noted earlier was still continuing, although the slopes for both the blood
and gasoline lead seem to be somewhat shallower toward the most recent data. A similar
picture is presented by the percentage of children with blood lead levels greater than 30
(jg/dl. In the early 70's, about 60 percent of the screened children had these levels; by 1979
the percentage had dropped between 10 and 15 percent.
11.3.6.4 Frankfurt, West Germany. Sinn (1980; 1981) conducted a study specifically examining
the environmental and biological impact of the gasoline lead phasedown implemented in West
Germany on January 1, 1976. Frankfurt am Main provided a good setting for such a study
because of its physical character.
Air and dustfall lead levels at several sites in and about the city were determined be-
fore and after the phasedown was implemented. The mean air lead concentrations obtained
during the study are presented in Table 11-20. A substantial decrease in air lead levels was
noted for the low-level high traffic site (3.18 ug/m3 in 1975-76 to 0.68 ug/m3 in 1978-1979).
No change was noted for the background site while only minor changes were observed for the
other locations. Dustfall levels fell markedly (218 mg/cm2-day for 1972-1973 to 128
mg/cm2'day for 1977-1978). Traffic counts were essentially unchanged in the area during the
course of study.
11-60
-------
I I I I I I I I I I I I
BLACK
- HISPANIC
. AIR LEAD
i i i i I i i i I
J I I I I I l Inn
O
1970 1971 1972 1973 1974 1975 1976
QUARTERLY SAMPLING DATE
Figure 11-16. Geometric mean blood levels for blacks and Hispanics in
the 25-to-36-month age group and rooftop quarterly averages for
ambient citywide lead levels.
Source: Nathanson and Nudelman (1980).
11-61
-------
50
E
§ «
\
$?
25»
92
20
o
Ul
(9
10
i i n rn i
QEO. MEAN BLOOD Pb
GASOLINE LEAD
V- 'v
vv\A VW,
">>\/Vv.
"v'V-x
^V
TRISTATE x 4
* SMSA x 20
I I I I I
I I I
66 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81
YEAR
Figure 11-17. Time-dependence of blood lead and gas lead for blacks,
aged 25 to 36 months, in New York.
Source: Billick (1982).
11-62
-------
TABLE 11-20. MEAN AIR LEAD CONCENTRATIONS DURING THE VARIOUS BLOOD SAMPLING
PERIODS AT THE MEASUREMENT SITES DESCRIBED IN THE TEXT
1975-1976
1976-1977
1977-1978
1978-1979
Residential
low traffic
0.57
0.39
0.32
0.39
High traffic
(>20m)
0.59
0.38
0.31
0.31
High traffic
(3m)
3.18
1.04
0.66
0.68
Background
site
0.12
0.09
0.10
0.12
Source: Sinn (1980, 1981).
A number of population groups were included in the study of the blood lead levels; they
were selected for having either occupational or residential exposure to high density automo-
bile traffic. Blood samples were taken serially throughout the study (three phases in
December-January 1975-1976, December-January 1976-1977, and December-January 1977-1978).
Blood samples were collected by venipuncture and analyzed by three different laboratories.
All the labs used AAS although sample preparation procedures varied. A quality control
program across the laboratories was conducted. Due to differences in laboratory analyses,
attrition, and loss of sample, the number of subjects who could be examined throughout the
study was considerably reduced from the initial number recruited (124 out of 300).
Preliminary analyses indicated that the various categories of subjects had different
blood lead levels, and that males and females within the same category differed. A very com-
plicated series of analyses then ensued that made it difficult to draw conclusions because the
various years' results were displayed separately by each laboratory performing the chemical
analysis and by different groupings by sex and category. In Sinn's later report (1981), a
downward trend was shown to exist for males and females who were in all years of the study and
whose blood levels were analyzed by the same laboratory.
11.4 STUDIES RELATING EXTERNAL DOSE TO INTERNAL EXPOSURE
The purpose of this section is to assess the importance of environmental exposures in
determining the level of lead in human populations. Of prime interest are those studies that
yield quantitative estimates of the relationship between air lead exposures and blood lead
levels. Related to this question is the evaluation of which environmental sources of airborne
lead play a significant role in determining the overall impact of air lead exposures on blood
lead levels.
11-63
-------
A factor that complicates the analysis presented here is that lead does not remain sus-
pended in the atmosphere but rather falls to the ground, is incorporated into soil, dust, and
water, and enters the food chain over time (see Figure 11-1). Since man is exposed to lead
from all of these media, as will be demonstrated below, studies that relate air lead levels to
blood lead levels (especially experimental exposure studies) may underestimate the overall
impact of airborne lead on blood lead levels. In observational studies, on the other hand,
the effects of air lead will thus be confounded with lead exposures from other pathways. The
simultaneous presence of lead in multiple environmental media requires the use of multiple
variable analysis techniques or surrogate assessment of all other external exposures. Virtu-
ally no assessments of simultaneous exposures to all media have been done.
There are several key features that characterize good studies relating external exposure
to internal exposure of lead:
(1) The study population is well-defined.
(2) There is a good measure of the exposure of each individual.
(3) The response variable (blood lead) is measured with adequate quality control,
preferably with replicates.
(4) The statistical analysis model is biologically plausible and is consistent with
the data.
(5) The important covariates are either controlled for or measured.
Some studies of considerable importance do not address all of these factors adequately. Key
studies selected for discussion here are those which address enough of these factors suffi-
ciently well to establish meaningful relationships.
The choice of the statistical analysis model is important in determining these relation-
ships (for a more detailed discussion see Appendix 11B). The model used is especially criti-
cal in situations where lead is present in relatively low concentrations in one or more
environmental media. A large number of statistical models have been used to predict blood
lead from various environmental media. For simplicity, let PbB = blood lead, E. = environmen-
tal exposure from source j, and b. = the regression coefficient for source j. Using this
notation, the more common models can be written as follows:
Linear Model: PbB = b0 + bx Et + . . . + bg Eg + "error" (11-2)
Linear Model (log form): log(PbB) = Iog(b0 + bt Ej + . . . + b$ Eg) + "error" (11-3)
Log-log Model: log(PbB) = 1og(b0) + bi log(Et) + . . . + bg log(Es) + "error" (11-4)
Log Total Exposure Model: log(PbB) = b Iog(b0 + b^ Ex + ... + bc Ec) + "error" (11-5)
S 5>
11-64
-------
Power-Function Model: PbB = b0 + (bi Et + ... + bg Eg)C + "error" (11-6)
Cube-root Model: PbB = bQ + bt (E^^3 + "error" (11-7)
There is no question that the relationship between blood lead and environmental exposure
is nonlinear across the entire range of potential exposures, from very low to high levels. At
lower levels of exposure, however, the various models all provide adequate descriptions of the
observed data. The choice of a model must be based at least in part on the biological mecha-
nisms. At the very least, no model should be adopted which is inconsistent with biological
reality.
The compartment-type metabolic models described in Section 10.3.4 predict a linear
response to total lead intake. Compartment models are described by a system of coupled first-
order linear differential equations for the quantity of lead in various kinetically distinct
body pools, (see Appendix 11-A). These compartments or kinetic pools may or may not corres-
pond to distinct physiological systems. It is well known that if the kinetic rate coeffi-
cients and absorption coefficients in such a model are constant, then the equilibrium blood
lead in a steady-intake environment is
p.R _ (lead absorbed into blood, ug/d)_ (Pb mean residence time in blood, d) (11-8)
(PbB volume of distribution, dl)
The only allowable places for nonlinearity in intake are either in the absorption process, or
in the kinetics of lead distribution affecting the residence time. Nonlinearities affecting
distribution volume are less plausible. Some of the evidence relating to these mechanisms was
reviewed in Chapter 10. Chamberlain (1983) and U.S. EPA (1983) have concluded that after
several months of steady exposure to environmental lead, blood lead levels achieve a near-
equilibrium concentration that increases linearly with the ambient concentration no matter
what the exposure pathway (directly by air inhalation, or by ingestion of food, water, dust,
soil, or paint), provided the total exposure does not cause blood lead to exceed 30-40 |jg/dl.
However, when total lead exposure by any pathway becomes so great that blood lead levels
greatly exceed 60-80 ug/dl, then the blood lead concentrations increase much more slowly with
increasing exposure concentration than they did at lower levels.
On the other hand, the log-log and cube root models have slopes which approach infinity
as the exposure approaches zero. The curves are so highly nonlinear at low doses that the
models attribute nearly all of the increase of blood lead levels to the lowest exposures,
and attribute relatively little increase to any additional exposures. However, the data of
11-65
-------
Piomelli et al. (1980) on a population of Nepalese exposed to an air lead of 0.00086 pg/m3 had
a geometric mean blood lead level of 3.4 pg/dl. This is similar to the value predicted by the
log-log model of Goldsmith-Hexter.
The following sections give the models as presented by the original authors. In many
cases, EPA has fitted other models in order to show the sensitivity of analysis to the model
selected.
11.4.1 Air Studies
The studies emphasized in this section are those most relevant to answering the following
question: If there is moderate change in average ambient air lead concentrations due to
changes in environmental exposure (at or near existing EPA air lead standards), what changes
are expected in blood lead levels of individual adults and children in the population? Longi-
tudinal studies in which changes in blood lead can be measured in single individuals as re-
sponses to changes in air lead are discussed first. The cross-sectional relationship between
blood lead and air lead levels in an exposed population provides a useful but different kind
of information, since the population "snapshot" at some point in time does not directly mea-
sure changes in blood lead levels or responses to changes in air lead exposure. In this
chapter consideration is also restricted to those individuals without known excessive occupa-
tional or personal exposures (except, perhaps, for some children in the Kellogg/Silver Valley
study).
The previously published analyses of relevant studies have not agreed on a single form
for the relationship between air lead and blood lead. All of the experimental studies have at
least partial individual air lead exposure measures, as does the cross-sectional observational
study of Azar et al. (1975). The 1974 Kellogg/Silver Valley study (Yankel et al., 1977) has
also been analyzed using several models. Other population cross-sectional studies have been
analyzed by Snee (1981). The most convenient method for summarizing these diverse studies and
their several analyses is by use of the blood lead - air lead slope (p), where p measures the
change in blood lead that is expected for a unit change in air lead. If determined for indi-
vidual subjects in a study population, this slope is denoted p.. If the fitted equation is
linear, then p or p.. is the slope of the straight line relationship at any air lead level. If
the fitted relationship is nonlinear, then the slope of the relationship measures the expected
effect on blood lead of a small change in air lead at some given air lead value and thus will
be somewhat different at different air lead levels.
A basic assumption here is that the distribution of blood lead in human populations with
homogeneous exposure (same geometric mean blood lead) is lognormal; a second assumption is
that all such lognormal distributions have the same geometric standard deviation (g.s.d.) or
11-66
-------
coefficient of variation (c.v.) It is then possible to calculate the fraction of the popu-
lation in excess of any specific level of blood lead. Most subpopulations not occupationally
exposed to lead have geometric mean blood lead < 20 ug/dl, at which level the effects of a few
ug/dl change in blood lead can be well approximated by a linear function. On the other hand,
many important experimental studies involve subjects with much higher blood lead. The re-
sponse relationships derived from lead-exposed subjects (blood lead > 30 (jg/dl) usually show
much lower slopes b. when blood lead exceeds 40 ug/dl. These two uses of blood lead versus
J
intake models -- to predict the fraction of an exposed population at risk and to predict the
change in blood lead of subjects exceeding a criterion blood lead level when blood lead expo-
sure changes may require different blood lead slopes b.. These two uses are not neces-
J
sarily inconsistent, e.g., if there was a corresponding increase in biological variability of
response to high levels of intake offsetting the decreased slope. For this reason we sepa-
rately analyze the single-subject and population studies.
11.4.1.1 The Griffin et al. Study. The study of Griffin et al. (1975) has the largest number
of human subjects exposed to atmospheric particulate lead at near-ambient conditions, under
conditions of long-term controlled exposure. In two separate experiments conducted at the
Clinton Correctional Facility in 1971 and 1972, adult male prisoner volunteers were sequest-
ered in a prison hospital unit and exposed to approximately constant levels of lead oxide
(average 10.9 ug/m3 in the first study and 3.2 ug/m3 in the second). Volunteers were exposed
in an exposure chamber to an aerosol of submicron-sized particles of lead oxide, which was
prepared by burning tetra-ethyl lead in a propane flame. There was an approximate additional
10-15 percent exposure to ambient organic lead vapor. All volunteers were introduced into the
chamber 2 weeks before the initiation of the exposure; the lead exposures were scheduled to
last 16 weeks, although the volunteers could drop out whenever they wished. Twenty-four vol-
unteers, including 6 controls, participated in the 10.9 ug/m3 exposure study. Not all volun-
teers completed the exposure regimen. Blood lead levels were found to stabilize after appro-
ximately 12 weeks. Among 8 men exposed to 10.9 ug/m3 for at least 60 days, a stabilized mean
level of 34.5 ±5.1 ug/dl blood was obtained, as compared with an initial level of 19.4 ± 3.3
ug/dl. All but two of the 13 men exposed at 3.2 ug/m3 for at least 60 days showed increases
and an overall stabilized level of 25.6 ± 3.9 ug/dl was found, compared with an initial level
of 20.5 ±4.4 ug/dl. This represented an increase of about 25 percent above the base level.
The aerosols used in this experiment were somewhat less complex chemically, as well as
somewhat smaller, than those found in the ambient environment. The particle size obtained was
0.05-0.10 urn, which is smaller than true urban aerosol of 0.3 urn. Griffin et al. (1975), how-
ever, pointed out that good agreement was achieved on the basis of the comparison of their ob-
served blood lead levels with those predicted by Goldsmith and Hexter's (1967) equation; that
11-67
-------
is, Iog10 blood lead = 1.265 + 0.2433 Iog10 atmospheric air lead. The average diet content of
lead was measured and blood lead levels were observed at 1- or 2-week intervals for several
months. Eight subjects received the maximum 4-month exposure to 10.9 |jg/m3; nine subjects
were exposed for 1 - 3 months. Six subjects had the maximum 4-month exposure to 3.2 ug/m3,
and eight others had shorter exposures.
Compartmental models have been fitted to these data by 0'Flaherty et al. (1982) and by
EPA. The basis of these models is that the mass of lead in each of several distinct pools or
compartments within the body changes according to a system of coupled first-order linear dif-
ferential equations with constant fractional transfer rates (Batschelet et al., 1979; Rabino-
witz et al., 1976). Such a model predicts that when the lead intake changes from one constant
level to another, then the relationship between the mass of lead in each compartment and time
with constant intake has a single exponential term.
The subjects at 3.2 pg/m3 exhibited a smaller increase in blood lead, with correspond-
ingly less accurate estimates of the parameters. Several of the lead-exposed subjects failed
to show an increase.
EPA has reanalyzed these data using a two-compartment model for two reasons:
(1) Semi logarithmic plots of blood lead versus time for most subjects showed a two-
component exponential decrease of blood lead during the postexposure or washout
phase of the experiments. Rabinowitz et al. (1977) show that at least two pools
are necessary to model blood lead kinetics accurately. The first pool is tenta-
tively identified with blood and the most labile soft tissues. The second pool
probably includes soft tissues and labile bone pools.
(2) Kinetic models are needed to account for the subjects' lead burdens not being
in equilibrium at any phase of the experiments.
Previously published analyses have not used data for all 43 subjects, particularly for
the same six subjects (labeled 15-20 in both experiments) who served as controls both years.
These subjects establish a baseline for non-inhalation exposures to lead, e.g., in diet and
water, and allow an independent assessment of within-subject variability over time. EPA ana-
lyzed data for these subjects as well as others who received lead exposures of shorter dura-
tion.
The estimated blood lead inhalation slope, p, was calculated for each individual subject
according to the formula
_ (Change in intake, ug/day) x (mean residence time in blood, day) (11-9)
p (Change in air exposure, ug/nr*) x (Volume of distribution, dl)
11-68
-------
The changes in air exposure were 10.9-0.15 = 10.75 ug/m3 for 1970-71 and 3.2-0.15 = 3.05 pg/m3
in 1971-72. Paired sample t-tests of equal means were carried out for the six controls and
five subjects with exposure both years, and independent sample t-tests were carried out com-
paring the remaining 12 subjects the first year and nine different subjects the next year.
All standard error estimates include within-subject parameter estimation uncertainties as well
as between subject differences. The following are observations:
(1) Non-inhalation lead intake of the control subjects varied substantially during the
second experiment at 3.2 ug/m3, with clear indication of low intake during the 14-
day pre-exposure period (resulting in a net decrease of blood lead). There was an
increase in lead intake (resulting in either equilibrium or net increase of blood
lead) during the exposure period. Subjects 16 and 20 had substantial increases,
subjects 15 and 19 had moderate increases, and subject 18 had no increase in blood
lead during exposure. Subject 17 had a marked decline in blood lead, but the rate
of decrease was much faster in the pre-exposure period, suggesting an apparent in-
crease of intake during exposure periods even for this subject. These subjects had
not apparently achieved equilibrium in either blood or tissue compartments. Even
though these subjects were not exposed to air lead, the estimated difference between
blood lead intake before and during exposure of the other subjects was used to cal-
culate the apparent inhalation slope at that exposure. The pooled inhalation slope
estimated for all six controls (1.48 ± 0.82 s.e.) was significantly positive (Z =
1.76, one-tailed p <0.05), as shown in Table 11-21. No explanation for the in-
creased lead intake during the winter of 1971-72 can be advanced at this time, but
factors such as changes in diet or changes in resorption of bone lead are likely to
have had an equal effect on the lead-exposed subjects. No statistically significant
changes in the controls were found during the first experiment at 10.9 pg/m3.
(2) Among the controls, the estimated mean residence time in blood was slightly longer
for the first year than the second year, 41.8 ± 9.2 days versus 34.6 ± 6.5 days, but
a paired sample Z-test found that the mean difference for the controls (7.2 ± 11.2
days) was not significantly different from zero (see Table 11-22).
(3) Among the five subjects exposed to 10.9 ug/m3 the first year and 3.2 ug/m3 the
second year, the mean residence time in blood was almost identical (43.9 ± 9.4
versus 44.7 ±8.7 days).
11-69
-------
TABLE 11-21. GRIFFIN ET AL. (1975) EXPERIMENT INHALATION SLOPE ESTIMATES
Group
Controls
All exposed
At 3.2 ug/m3
1.48 ± 0.82 (n =
3.00 ± 0.76 (n =
6)*
14)
At 10.9 ug/m3
-0.20 ± 0.27 (n =
1.57 ± 0.26 (n =
6)
17)
Difference 1.52 ± 1.12 1.77 ± 0.37
(Exposed controls)
Pooled: (all subjects) 1.75 ± 0.35
(without subjects 1,6)** 1.78 ± 0.35
*n = number of subjects.
**Subjects 1 and 6 were "non-responders."
TABLE 11-22. GRIFFIN ET AL. (1975) EXPERIMENT MEAN RESIDENCE TIME IN BLOOD
Control
Exposed
3.2 ug/m3
experiment
34.6 ±6.5 days
40.8 ±4.4 days
10.9 ug/m3
experiment
41.8 ±9.2 days
40.6 ±3.6 days
(4) The average inhalation slope for all 17 subjects exposed to 10.9 ug/m3 is 1.77 ±
0.37 when the slope for the controls is subtracted. The corrected inhalation slope
for all 14 subjects exposed to 3.2 ug/m3 is 1.52 ± 1.12, or 1.90 ± 1.14 without
subjects 1 and 6 who were "non-responders." These are not significantly different.
The pooled slope estimate for all subjects is 1.75 ± 0.35. The pooled mean resi-
dence time for all subjects is 39.9 ±2.5 days.
Thus, in spite of the large estimation variability at the lower exposure level, the aver-
age inhalation slope estimate and blood lead half-life are not significantly different at the
two exposure levels. This suggests that blood lead response to small changes in air lead in-
halation is approximately linear at typical ambient levels.
11-70
-------
11.4.1.2 The Rabinowitz et al. Study. The use of stable lead isotopes avoids many of the
difficulties encountered in the analysis of whole blood lead levels in experimental studies.
Five adult male volunteers were housed in the metabolic research wards of the Sepulveda and
Wadsworth VA hospitals in Los Angeles for extended periods (Rabinowitz et al., 1974; 1976;
1977). For much of the time they were given low-lead diets with controlled lead content, sup-
plemented by tracer lead salts at different times.
Four subjects were initially observed in the ward for several weeks. Each subject was in
the semi-controlled ward about 14 hours per day and was allowed outside for 10 hours per day,
allowing the blood lead concentration to stabilize.
Subjects B, D, and E then spent 22-24 hours per day for 40, 25, and 50 days, respective-
ly, in a low-lead room with total particulate and vapor lead concentrations that were much
lower than in the metabolic wards or outside (see Table 11-23). The subjects were thereafter
exposed to Los Angeles air with much higher air lead concentrations than in the ward.
The calculated changes in lead intake upon entering and leaving the low-lead chamber are
shown in Table 11-24. These were based on the assumption that the change in total blood lead
was proportional to the change in daily lead intake. The change in calculated air lead
intakes (other than cigarettes) due to removal to the clean room were also calculated indepen-
dently by the lead balance and labeled tracer methods (Rabinowitz et al., 1976) and are con-
sistent with these direct estimates.
Rabinowitz and coworkers assumed that the amount of lead in compartments within the body
evolved as a coupled system of first-order linear differential equations with constant frac-
tional transfer rates. This compartmental model was fitted to the data. This method of
analysis is described in Appendix 11A.
Blood lead levels calculated from the three compartment model adequately predicted the
observed blood lead levels over periods of several hundred days. There was no evidence to
suggest homeostasis or other mechanisms of lead metabolism not included in the model. There
was some indication (Rabinowitz et al., 1976) that gut absorption may vary from time to time.
The calculated volumes of the pool with blood lead (Table 11-24) are much larger than the
body mass of blood (about 7 percent of body weight, estimated respectively as 4.9, 6.3, 6.3,
4.6, and 6.3 kg for subjects A-E). The blood lead compartment must include a substantial mass
of other tissue.
The mean residence time in blood in Table 11-24 includes both loss of lead from blood to
urine and transfer of a fraction of blood lead to other tissue pools. This parameter reflects
the speed with which blood lead concentrations approach a new quasi-equilibrium level. Many
years may be needed before approaching a genuine equilibrium level that includes lead that can
be mobilized from bones.
11-71
-------
TABLE 11-23. AIR LEAD CONCENTRATIONS* (ug/m3) FOR TWO SUBJECTS IN THE RABINOWITZ STUDIES
Environment Average Range
Subject A Outside (Sepulveda VA) 1.8 (1.2-2.4)
Inside (Sepulveda VA,
air-conditioned without
filter) 1.5 (1.0-2.7)
Inside (Wadsworth VA,
Open air room) 2.1 (1.8-2.6)
Subject B (Wadsworth VA)
Outside 2.0 (1.6-2.4)
In room (air conditioner
with filter, no purifier) 0.97 (0.4-2.1)
In room (with purifiers,
"clean air") 0.072 (0.062-0.087)
Open-air room 1.9 (1.8-1.9)
Organic vapor lead
Outside 0.10
"Clean air" 0.05
* 5-20 days exposure for each particulate lead filter.
One of the greatest difficulties in using these experiments is that the air lead expo-
sures of the subjects were not measured directly, either by personal monitors or by restric-
ting the subjects to the metabolic wards. The times when the subjects were allowed outside
the wards included possible exposures to ground floor and street level air, whereas the out-
side air lead monitor was mounted outside the third-floor window of the ward. The VA hospi-
tals are not far from major streets and the subjects' street level exposures could have been
much higher than those measured at about 10 m elevation (see Section 7.2.1.3). Some estimated
ratios between air concentrations at elevated and street level sites are given in Table 7-6.
A second complication is that the inside ward value of ug/m3 (Rabinowitz et al., 1977)
used for subject B may be appropriate for the Wadsworth VA hospital, but not for subject A in
the Sepulveda VA hospital (see Table 11-23). The changes in air lead values shown in Table
11-24 are thus nominal, and are likely to have systematic inaccuracies much larger than the
11-72
-------
TABLE 11-24. ESTIMATES OF INHALATION SLOPE, P, FOR RABINOWITZ STUDIES
Subject
A
B
C
D
E
Changes in
intake*,
ug/day
17 ± 5*
16 ± 3
15 ± 5*
9 ± 2
12 ± 2
Volume,**
kg
7.4 ± 0.6
10.0 ± 0.8
10.1 ± 1**
9.9 ± 1.2
11.3 ± 1.4
Residence!
time, days
34 ± 5
40 ± 5
37 ± 5
40 ± 5
27 ± 5
Changes in
air lead1',
ug/m3
2.5tt
2.0
2.2tt
2.0
2.0
Inhalation
slope, ug/d£
per ug/m3
2.98 ± 1.06
3.56 ± 0.93
2.67 ± 1.04
2.02 ± 0.60
1.59 ± 0.47
Maximujp
slope
4.38 ± 1.55
5.88 ± 1.54
4.16 ± 1.62
3.34 ± 0.99
2.63 ± 0.78
*From Rabinowitz et al. (1977), Table VI. Reduced intake by low-lead method for subjects
B, D, E, tracer method for A, balance method for C. Standard error for C is assumed by EPA
to be same as A.
**From Rabinowitz et al. (1976), Table II. EPA has assumed standard error with coefficient
of variation same as that for quantity of tracer absorbed in Table VI, except for subject C.
''"Estimates from Rabinowitz et al. (1976) Table II. Standard error estimate from combined
sample.
*+
"See text. For A and C, estimated from average exposure. For B, D, and E reduced by
0.2 ug/m3 for clean room exposure. Coefficient of variation assumed to be 10%.
"""Assumed density of blood 1.058 g/cm3.
Assuming outside air exposure is 2.1 ug/m3 rather than 4 ug/m3 for 10 hours.
nominal 10 percent coefficients of variation stated. The assumption is that for subjects B,
D, and E, the exposure to street level air for 10 hours per day was twice as large as the mea-
sured roof level air, i.e., 4 ug/m3; and the remaining 14 hours per day were at the ward level
of 0.97 ug/m3; thus the time-averaged level was [(10 x 4) + (14 x 0.97)]/24 = 2.23 ug/m3. The
average controlled exposures during the "clean room" part of the experiment were 23, 22, and
24 hours respectively for subjects B, D, E; thus averaged exposures were 0.19, 0.28, and 0.12
ug/m3, and reductions in exposure were about 2.0 ug/m3. This value is used to calculate the
slope. For subject A, the total intake due to respired air is the assumed indoor average of
1.5 ug/m3 for the Sepulveda VA hospital, combining indoor and outdoor levels [(10 x 4) + (14 x
1.5)]/24 = 2.54 ug/m3. For subject C the Wadsworth average applies. Other than uncertainties
in the air lead concentration, the inhalation slope estimates for Rabinowitz1s subjects have
less internal uncertainty than those calculated for subjects in Griffin's experiment.
11-73
-------
The inhalation slopes thus calculated are the lowest that can be reasonably derived from
this experiment, since the largest plausible air lead concentrations have been assumed. The
third-floor air monitor average of 2.1 (jg/m3 is a plausible minimum exposure, leading to the
higher plausible maximum inhalation slopes in the last column of Table 11-24. These are based
on the assumption that the time-averaged air lead exposure is smaller by [10(4-2.l)]/24 = 0.79
ug/m3 than assumed previously. It is also possible that some of this difference can be attri-
buted to dust ingestion while outside the metabolic ward.
11.4.1.3 The Chamberlain et al. Study. A series of investigations were carried out by
Chamberlain et al. (1975a,b; 1978) at the U.K. Atomic Energy Research Establishment in
Harwell, England. The studies included exposure of up to 10 volunteer subjects to inhaled,
ingested, and injected lead in various physical forms. The inhalation exposures included
laboratory inhalation of lead aerosols generated in a wind tunnel, or box, of various particle
sizes and chemical compositions (lead oxide and lead nitrate). Venous blood samples were
taken at several times after inhalation of 203Pb. Three subjects also breathed natural high-
way exhaust fumes at various locations for times up to about 4.5 hours.
The natural respiratory cycles in the experiments varied from 5.7 to 17.6 seconds (4 to
11 breaths per minute) and tidal volumes from 1.6 to 2.3 liters. Lung deposition of lead-
bearing particles depended strongly on particle size and composition, with natural exhaust
particles being more efficiently retained by the lung (30 - 50 percent) than were the chemical
compounds (20 - 40 percent).
The clearance of lead from the lungs was an extended process over time and depended on
particle size and composition, leaving only about 1 percent of the fine wind tunnel aerosols
in the lung after 100 hours, but about 10 percent of the carbonaceous exhaust aerosols. The
203Pb isotope reached a peak blood level about 30 hours after inhalation, the blood level then
representing about 60 percent of the initial lung burden.
A substantial fraction of the lead deposited in the lung appears to be unavailable to the
blood pool in the short term, possibly due to rapid transport to and retention in other tis-
sues including skeletal tissues. In long-term balance studies, some of this lead in the deep
tissue compartment would return to the blood compartment.
Lead kinetics were also studied by use of injected and ingested tracers, which suggested
that in the short term, the mean residence time of lead in blood could be calculated from a
one-pool model analysis.
Chamberlain et al. (1978) extrapolated these high-level, short-term exposures to longer
term ones. The following formula and data were used to calculate a blood-to-air level ratio
11-74
-------
[Tj ] [% Deposition] [% Absorption] [Daily ventilation]
p = -1 - (11-10)
[Blood volume] [0.693]
where:
Tj = biological half life
With an estimated value of T, = 18 days (mean residence time T, 70.693 = 26 days), with 50 per-
*i -j
cent for deposition in lung for ordinary urban dwellers, and 55 percent of the lung lead re-
tained in the blood lead compartment (all based on Chamberlain's experiments), with an assumed
ventilation of 20 nrVday over blood volume 5400 ml (Table 10-20 in Chamberlain et al . , 1978),
then
R = 26 day X 0.50 X 0.55 X 20 mVday =2.7 mVdl (11-11)
P 54 dl
This value of p could vary for the following reasons:
1. The absorption from lung to blood used here, 0.55, refers to short-term kinetics.
In the long term, little lead is lost through biliary or pancreatic secretions,
nails, hair, and sweat, so that most of the body lead is available to the blood
pool even if stored in the skeleton from which it may be resorbed. Chamberlain
suggests an empirical correction to 0.55 X 1.3 = 0.715 absorption.
2. The mean residence time, 26 days, is shorter than in Rabinowitz's subjects, and
the blood volume is less, 54 dl. It is possible that in the Rabinowitz study,
the mean times are longer and the blood pool size (100 dl) is larger than here
because Rabinowitz et al. included relatively fewer labile tissues such as kidney
and liver in the pool. Assuming 40 days mean residence time and 100 dl blood
volume the slope can be recalculated,
D _ 40 d X 0.50 X 0.55 X 20 m3/d _ , ? 3 , ., (11-12)
P -- IQO~dT -- ** m /dl-
3. The breathing rate could be much less, for inactive people.
11-75
-------
11.4.1.4 The Kehoe Study. Between 1950 and 1971, Kehoe exposed 12 subjects to various levels
of air lead under a wide variety of conditions. Four earlier subjects had received oral lead
during 1937-45. The inhalation experiments were carried out in an inhalation chamber at the
University of Cincinnati, in which the subjects spent varying daily time periods over extended
intervals. The duration was typically 112 days for each exposure level in the inhalation
studies, and at the end of this period it was assumed the blood lead concentration had reached
a near-equilibrium level. The experiments are described by Kehoe (1961a,b,c) and the data and
their analyses by Gross (1981) and Hammond et al. (1981). The studies most relevant to this
document are those in which only particles of lead sesquioxide aerosols in the submicron range
were used, so that there was at least one air lead exposure (other than control) for which the
time-averaged air lead concentration did not exceed 10 ug/m3. Only six subjects met these
criteria: LD (1960-63), JOS (1960-63), NK (1963-66), SS (1963-68), HR (1966-67), and DH
(1967-69). Subject DH had a rather high initial blood lead concentration (30 |jg/dl) that fell
during the course of the experiment to 28 pg/dl; apparently daily detention in the inhalation
chamber altered DH's normal pattern of lead exposure to one of lesser total exposure. The
Kehoe studies did not measure non-experimental airborne lead exposures, and did not measure
lead exposures during "off" periods. Subject HR received three exposure levels from 2.4-7.5
|jg/m3, subject NK seven exposure levels from 0.6-4.2 M9/m3, and subject SS 13 exposure levels
from 0.6-7.2 pg/m3- LD and JOS were each exposed to about 9, 19, 27, and 36 ug/m3 during
sequential periods of 109-113 days.
A great deal of data on lead content in blood, feces, urine and diet were obtained in
these studies and are exhibited graphically in Gross (1979) (see Figure 11-18). Apart from
the quasi-equilibrium blood lead values and balances reported in Gross (1979; 1981), there has
been little use of these data to study the uptake and distribution kinetics of lead in man.
EPA analyses used only the summary data in Gross (1981).
Data from Gross (1981) were fitted by least squares linear and quadratic regression
models. The quadratic models were not significantly better than the linear model except for
subjects LD and JOS, who were exposed to air levels above 10 \*q/m3. The linear terms predomi-
nate in all models for air lead concentrations below 10 ug/m3 and are reported in Table 11-25.
These data represent most of the available experimental evidence in the higher range of
ambient exposure levels, approximately 3-10 |jg/m3. Data for the four subjects with statis-
tically significant relationships are shown in Figure 11-19, along with the fitted regression
curve and its 95 percent confidence band.
11-76
-------
SUBJECT - SS
BALANCE
I
CO
Q.
o
<
a
co
Q.
o
5
<
Q
£
Q.
o
g
°
5
rtKfl^k***!^^
200 300 400 -500 600
900. 1000. 1100. 1200. 1300. 1400
TIME (days)
Figure 11-18. Data plots for individual subjects as a function of time for
Kehoe subjects, as presented by Gross (1979).
11-77
-------
TABLE 11-25. LINEAR SLOPE FOR BLOOD LEAD VERSUS AIR LEAD AT
LOW AIR LEAD EXPOSURE IN KEHOE'S SUBJECTS
Range
Linear Slopes
Subject
DHa
HRa
J0§b
NK<:
SSC
Linear Model
-0.
0.
0.
0.
2.
1.
34 ± 0.
70 ± 0.
67 ± 0.
64 ± 0.
60 ± 0.
31 ± 0.
28
46
07
11
32
20
p, m3/dl
, ± s. e.
Air*
Quadratic Model |jg/
0.
0.
1.
1.
1.
1.
14 ± 1.
20 ± 2.
01 ± 0.
29 ± 0.
55 ± 1.
16 ± 0.
25
14
19
06
28
78
5.
2.
9.
9.
0.
0.
6 -
4 -
4 -
3 -
6 -
6 -
'm
8.8
7.5
35.7
35.9
4.0
7.2
Blood,
ug/<
26 -
21 -
21 -
18 -
20 -
18 -
iSL
31
27
46
41
30
29
*Also, control = 0.
No statistically significant relationship between air and blood lead.
High exposures. Use linear slope from quadratic model.
Low exposures. Use linear slope from linear model.
11.4.1.5 The Azar et al. Study. Thirty adult male subjects were obtained from each of five
groups: 1) Philadelphia cab drivers; 2) OuPont employees in Starke, Florida; 3) DuPont em-
ployees in Barksdale, Wisconsin; 4) Los Angeles cab drivers; and 5) Los Angeles office workers
(Azar et al., 1975). Subjects carried air lead monitors in their automobiles and in their
breathing zones at home and work. Personal variables (age, smoking habits, water samples)
were obtained from all subjects, except for water samples from Philadelphia cab drivers.
Blood lead, ALAD urine lead, and other variables were measured. From two to eight blood sam-
ples were obtained from each subject during the air monitoring phase. Blood lead determina-
tions were done in duplicate. Table 11-26 presents the geometric means for air lead and blood
lead for the five groups. The geometric means were calculated by EPA from the raw data pre-
sented in the authors' report (Azar et al., 1975).
The Azar study has played an important role in setting standards because of the care used
in measuring air lead in the subjects' breathing zone. Blood lead levels change in response
to air lead levels, with typical time constants of 20-60 days. One must assume that the
subjects' lead exposures during preceding months had been reasonably similar to those during
the study period. Models have been proposed for these data by Azar et al. (1975), Snee (1981;
1982b), and Hammond et al. (1981) including certain nonlinear models.
Azar et al. (1975) used a log-log model for their analysis of the data. The model in-
cluded dummy variables, C1( C2, C3) C4, C5, which take on the value 1 for subjects in that
group and 0 otherwise (see Table 11-26 for the definitions of these dummy variables). The
fitted model using natural logarithms was
11-78
-------
30
I I /
0 1 234567
1 2 3
AIR LEAD, ualrn'
I I I I
SUBJECT LD /
0 6 10 15 20 26 30 35
0 5 10 15 20 26 30 35
AIR LEAD.
Figure 11-19. Blood level vs. air lead relationships for Kehoe inhalation
studies: linear relation for low exposures, quadratic for high exposures,
with 95% confidence bands.
11-79
-------
TABLE 11-26 GEOMETRIC MEAN AIR AND BLOOD LEAD LEVELS (ng/100 g)
FOR FIVE CITY-OCCUPATION GROUPS (DATA CALCULATED BY EPA)
.
Group
Cab drivers
Philadelphia, PA
Plant employees
Starke, FL
plant employees
Barksdale, WI
Cabdrivers
Los Angeles, CA
Office workers
Los Angeles, CA
Geometric mean
air lead,
ug/m3
2.59
0.59
0.61
6.02
2.97
GSD
1.16
2.04
2.39
1.18
1.29
1=^=^.1-=-=^..= ^-=r-= .',,. . . -=^ = LJ-J-±:
Geometric mean
blood lead,
ljg/100 g GSD
22.1 1.16
15.4 1.41
12.8 1.43
24.2 1.20
18.4 1.24
- r J : - ,:=
Sampl e
size
30
29
30
30
30
Code
Ci
C2
C3
C4
C5
Source: Azar et al. (1975).
log (blood Pb) = 2.951 Cj. + 2.818 C2 +
2.627 C3 + 2.910 C4 + 2.821 C5 + 0.153 log (air Pb) (11-13)
This model gave a residual sum of squares of 9.013, a mean square error of 0.063 (143 degrees
f freedom), and a multiple R2 of 0.502. The air lead coefficient had a standard error of
040 The fitted model is nonlinear on air lead, and so the slope depends on both air lead
d the intercept. Using an average intercept value of 1.226, the curve has a slope ranging
from 10.1 at an air lead level of 0.2 |jg/m3 to 0.40 at an air lead level of 9 pg/m3.
Snee (1982b) reanalyzed the same data and fitted the following power function model,
log (blood Pb) = log [12.1 (air Pb + 6.00 d + 1.46 C2
+ 0.44 C3 -»- 2.23 C4 + 6.26 C5)°-2669]
model gave a residual sum of squares of 9.101, a mean square error of 0.063 (142 degrees
freedom) and a multiple R2 of 0.497. Using an average constant value of 3.28, the slope
ranges
1 29 at an air lead of 0.2 to 0.51 at an air lead of 9.
i-f-->
11-80
-------
An important extension in the development of models for the data was the inclusion of
separate non-air contributions or background exposures for each separate group. The coeffi-
cients of the group variables, C., in the lead exposure model may be interpreted as measures
of total exposure of that group to non-air external sources (cigarettes, food, dust, water)
and to endogenous sources (lead stored in skeleton). Water and smoking variables were used to
estimate some external sources. (This required deleting another observation for a subject
with unusually high water lead.) The effect of endogenous lead was estimated using subject
age as a surrogate measure of cumulative exposure, since lead stored in the skeleton is known
to increase approximately linearly with age, for ages 20-60 (Gross et al., 1975; Barry, 1975;
Steenhout, 1982) in homogeneous populations.
In order to facilitate comparison with the constant p ratios calculated from the clinical
studies, EPA fitted a linear exposure model to the Azar data. The model was fitted on a loga-
rithmic scale to facilitate comparison of goodness of fit with other exposure models and to
produce an approximately normal pattern of regression residuals. Neither smoking nor water
lead provided significantly better fits to the log (blood lead) measurements after the effect
of age was removed.
Age and air lead may be confounded to some extent because the regression coefficient for
age may include the effects of prior air lead exposures on skeletal lead buildup. This would
have the effect of reducing the estimated apparent slope p.
Geometric mean regressions of blood lead on air lead were calculated by EPA for several
assumptions: (1) A linear model analogous to Snee's exposure model, assuming different non-
air contributions in blood lead for each of the five subgroups; (2) a linear model in which
age of the subject is also used as a surrogate measure of the cumulative body burden of lead
that provides an endogenous source of blood lead; (3) a linear model similar to (2), in which
the change of blood lead with age is different in different subgroups, but it is assumed that
the non-air contribution is the same in all five groups (as was assumed in the 1977 EPA Lead
Criteria Document); (4) a linear model in which both the non-air background and the change in
blood lead with age may differ by group; and (5) a nonlinear model similar to (4). None of
the fitted models are significantly different from each other using statistical tests of hypo-
theses about parameter subsets in nonlinear regression (Gallant, 1975).
11.4.1.6 Silver Vailey/KeHogg, Idaho Study. In 1970, EPA carried out a study of a lead
smelter in Kellogg, Idaho (Hammer et al., 1972; U.S. Environmental Protection Agency, 1972).
The study was part of a national effort to determine the effects of sulfur dioxide, total sus-
pended particulate and suspended sulfates, singly and in combination with other pollutants, on
human health. It focused on mixtures of the sulfur compounds and metals. Although it was
demonstrated that children had evidence of lead absorption, insufficient environmental data
were reported to allow further quantitative analyses.
11-81
-------
In 1974, following the hospitalization of two children from Kellogg with suspected acute
lead poisoning, the CDC joined the State of Idaho in a comprehensive study of children in the
Silver Valley area of Shoshone County, Idaho, near the Kellogg smelter (Yankel et al., 1977;
Landrigan et al., 1976).
The principal source of exposure was a smelter whose records showed that emissions aver-
aged 8.3 metric tons per month from 1955 to 1964 and 11.7 metric tons from 1965 to September,
1973. After a September, 1973 fire extensively damaged the smelter's main emission filtration
facility, emissions averaged 35.3 metric tons from October, 1973 to September, 1974 (Landrigan
et al. , 1976). The smelter operated during the fall and winter of 1973-74 with severely
limited air pollution control capacity. Beginning in 1971, ambient concentrations of lead in
the vicinity of the smelter were determined from particulate matter collected by hi-vol air
samples. Data indicated that monthly average levels measured in 1974 (Figure 11-20) were
three to four times the levels measured in 1971 (von Lindern and Yankel, 1976). Individual
exposures of study participants to lead in the air were estimated by interpolation from these
data. Air lead exposures ranged from 1.5 |jg/m3 to 30 ug/m3 monthly average (see Fi.gure 11-20).
Soil concentrations were as high as 24,000 ug/g and averaged 7000 |jg/g within one mile of the
smelter. House dusts were found to contain as much as 140,000 (jg/g and averaged 11,000 pg/g
in homes within one mile of the complex.
The study was initiated in May, 1974 and the blood samples were collected in August, 1974
from children 1-9 years old in a door-to-door survey (greater than 90 percent participation).
Social, family, and medical histories were conducted by interview. Paint, house, dust, yard
and garden soils, grass, and garden vegetable samples were collected. At that time, 385 of
the 919 children examined (41.9 percent) had blood lead levels in excess of 40 ug/dl, 41 chil-
dren (4.5 percent) had levels greater than 80 ug/dl. All but 2 of the 172 children living
within 1.6 km of the smelter had levels greater than or equal to 40 (jg/dl. Those two children
had moved into the area less than six months earlier and had blood lead levels greater than 35
|jg/dl. Both the mean blood lead concentration and the number of children classified as exhib-
iting excess absorption decreased with distance from the smelter (Table 11-27). Blood lead
levels were consistently higher in 2- to 3-year-old children than they were in other age
groups (Table 11-28). A significant negative relationship between blood lead level and hema-
tocrit value was found. Seven of the 41 children (17 percent) with blood lead levels greater
than 80 pg/dl were diagnosed as being anemic on the basis of hematocrit less than 33 percent,
whereas only 16 of 1006 children (1.6 percent) with blood lead levels less than 80 ug/dl were
so diagnosed. Although no overt disease was observed in children with higher lead intake,
differences were found in nerve conduction velocity. Details of this finding are discussed in
Chapter 12.
11-82
-------
"&
O
cc
K
Z
UJ
O
O
o
O
UJ
CC
z
LU
m
<
Q
UJ
>
cc
UJ
V)
CO
O
30
25
20
15
10
1971
1972
1973
1974
1975
Figure 11-20. Monthly ambient air lead concentrations in Kellogg,
Idaho, 1971 through 1975.
Source: von Lindern and Yankel (1976).
11-83
-------
TABLE 11-27. GEOMETRIC MEAN BLOOD LEAD LEVELS BY AREA COMPARED WITH
ESTIMATED AIR LEAD LEVELS FOR 1- TO 9-YEAR OLD CHILDREN
LIVING NEAR IDAHO SMELTER. (GEOMETRIC STANDARD DEVIATIONS,
SAMPLE SIZE, AND DISTANCES FROM SMELTER ARE ALSO GIVEN)3
Area
1
2
3
4
5
6
Geometric mean
blood lead,
ug/dl
65.9
47.7
33.8
32.2
27.5
21.2
GSD
1.30
1.32
1.25
1.29
1.30
1.29
Sample
size
170
192
174
156
188
90
% blood
lead
(>40 ug/dl)
98.9
72.6
21.4
17.8
8.8
1.1
Estimated
air lead,
(|jg/m3)
18.0
14.0
6.7
3.1
1.5
1.2
Distance from
smelter,
Km
0- 1.6
1.6- 4.0
4.0-10.0
10.0-24.0
24.0-32.0
about 75
EPA analysis of data from Yankel et al. (1977).
TABLE 11-28. GEOMETRIC MEAN BLOOD LEAD LEVELS BY AGE AND AREA FOR
SUBJECTS LIVING NEAR THE IDAHO SMELTER
(micrograms per deciliter)
Age group
Area
1
2
3
4
5
6
7
1
69*
50
33
31
27
21
28
2
72
51
36
35
35
25
30
3
75
55
36
34
29
22
28
4
75
46
35
31
29
23
32
5
68
49
35
31
29
20
30
6
66
50
35
35
28
22
26
7
63
47
31
30
25
20
37
8
60
42
32
32
27
22
30
9
57
40
32
30
24
17
20
Teenage
39
33
28
35
Adult
37
33
30
34
32
32
*Error in original publication (Yankel et al., 1977).
11-84
-------
Yankel et al. (1977) fitted the data to the following model.
-5
In (blood lead) = 3.1 + 0.041 air lead + (2.1 x 10 soil lead)
+ 0.087 dustiness - 0.018 age
+ 0.024 occupation
(11-15)
where air lead was in ug/m3; soil lead was in uQ/9; dustiness was 1, 2, or 3; age was in
years; and occupation (parental) was a Hollingshead index. The analysis included 879 sub-
jects, had a multiple R2 of 0.622, and a residual standard deviation of 0.269 (geometric
standard deviation of 1.31).
Walter et al. (1980) used a similar model to examine age specific differences of the re-
gression coefficients for the different variables. Those coefficients are summarized in Table
11-29. The variable that was most significant overall was air lead; its coefficient was ap-
proximately the same for all ages, corresponding to a change in blood lead of about 1 ug/dl
per unit increase of air lead (in ug/m3) at an air exposure of 1 ug/m3 and about 2.4 ug/dl per
unit increase in air at an air exposure of 22 ug/m3.
TABLE 11-29. AGE-SPECIFIC REGRESSION COEFFICIENTS FOR THE ANALYSIS OF
LOG (BLOOD LEAD) LEVELS IN THE IDAHO SMELTER STUDY
Age
1
2
3
4
5
6
7
8
9
* P
t P
Air
0.0467*
0.0405*
0.0472*
0.0366*
0.0388*
0.0361*
0.0413*
0.0407*
0.0402*
<0.01
<0.05
Dust
0.119t
o.ioet
o.iost
0.107t
0.052
0.070
0.053
0.051
O.OSlt
Occupation
0.0323
0.0095
0.0252
0.0348
0.0363t
0.0369t
0.0240
0.04221
0.0087
Pica
0.098
0.225*
0.077
0.117
0.048
0.039
0.106
0.010
0.108
Sex
0.055
0.002
0.000
0.032
-0.081
-0.092
-0.061
-o.ioet
-0.158*
Soil
(xlO4)
3.5
20. 6t
24.2*
32.1*
23.4*
38.4*
21. 3t
16.2
11.6
Intercept
3.017
3.567
3.220
3.176
3.270
3.240
3.329
3.076
3.477
N
98
94
115
104
130
120
113
105
104
The next most important variable that attained significance at a variety of ages was the
household dustiness level (coded as low = 0, medium = 1, or high = 2), showing a declining ef-
fect with age and being significant for ages 1-4 years. This suggested age-related hygiene
behavior and a picture of diminishing home orientation as the child develops. For ages 1-4
years, the coefficient indicates the child in a home with a "medium" dust level would have a
11-85
-------
blood lead level ~ 10 percent higher than a child in a home with a "low" dust level, other
factors being comparable.
The coefficients for soil lead - blood lead relationships exhibited a fairly regular pat-
tern, being highly significant (p <0.01) for ages 3-6 years, and significant (p <0.05) at ages
2-6 years. The maximum coefficient (at age 6) indicates a 4 percent increase in blood lead
per 1000 ug/g increase in soil lead.
Pica (coded absent = 0, present = 1) had a significant effect at age 2 years, but was in-
significant elsewhere; at age 2 years, an approximate 25 percent elevation in blood lead is
predicted in a child with pica, compared with an otherwise equivalent child without pica.
Parental occupation was significant at ages 5, 6, and 8 years; at the other ages, how-
ever, the sign of the coefficient was always positive, consistent with a greater lead burden
being introduced into the home by parents working in the smelter complex.
Finally, sex (coded male = 0; female = 1) had a significant negative coefficient for ages
8 and 9 years, indicating that boys would have lead levels 15 percent higher than girls at
this age, on the average. This phenomenon is enhanced by similar, but nonsignificant, nega-
tive coefficients for ages 5-7 years.
Snee (1982c) also reanalyzed the Idaho smelter data using a log-linear model. He used
dummy variables for age, work status of the father, educational level of the father, and
household dust level (cleanliness). The resulting model had a multiple R2 of 0.67 and a resi-
dual standard deviation of 0.250 (geometric standard deviation of 1.28). The model showed
that 2-year-olds had the highest blood lead levels. The blood lead inhalation slope was es-
sentially the same as that of Yankel et al. (1977) and Walter et al. (1980).
The above non-linear analyses of the Idaho smelter study are the only analyses which sug-
gest that the blood lead to air lead slope increases with increasing air lead, contrary to the
findings of decreasing slopes seen at high air lead exposures in other studies. An alterna-
tive to this would be to attempt to fit a linear model as described in Appendix 11-B. Expo-
sure coefficients were estimated for each of the factors shown in Table 11-30. The results
for the different covariates are similar to those of Snee (1982c) and Walter et al. (1980).
Because the previous analyses noted above indicated a nonlinear relationship, a similar
model with a quadratic air lead term added was also fitted. The coefficients for the other
factors remained about the same, and the improvement in the model was marginally significant
(p = 0.05). This model gave a slope of 1.16 at an air lead of 1 ug/m3, and 1.39 at an air
lead of 2 ug/m3. Both the linear and quadratic models, along with Snee's (1982b) model are
shown in Figure 11-21. The points represent mean blood lead levels adjusted for the factors
in Table 11-30 (except air lead) for each of the different exposure subpopulations.
11-86
-------
TABLE 11-30. ESTIMATED COEFFICIENTS* AND STANDARD ERRORS FOR THE IDAHO SMELTER STUDY
Factor
Intercept (ug/dl)
Air lead (ug/m3)
Soil lead (1000 ug/g)
Sex (male=l, female=0)
Pica (eaters=l, noneaters=0)
Education (graduate training=0)
At least high school
No high school
Cleanliness of home (clean=0)
Moderately clean
Dirty
Age (1 year old=0)
2 years old
3 years old
4 years old
5 years old
6 years old
7 years old
8 years old
9 years old
Work status (no exposure=0)
Lead or zinc worker
Coefficient
13.19
1.53
1.10
1.31
2.22
-
3.45
4.37
-
3.00
6.04
-
4.66
5.48
3.16
2.82
2.74
0.81
-0.19
-1.50
-
3.69
Asymptotic
standard error
1.90
0.064
0.14
0.59
0.90
1.44
1.51
0.65
1.06
1.48
1.32
1.32
1.25
1.24
1.23
1.28
1.21
0.61
Residual standard deviation = 0.2576 (geometric standard deviation = 1.29),
Multiple R2 = 0.662.
Number of observations = 860.
"Calculations made by EPA.
11-87
-------
Q
LU
b^
§
CD
Q
Ul
I
I
80
70
60
50
40
30
20
10
0
I I I I I | I I I I I I I II | I I I I I I !..f
LINEAR (EPA)
QUADRATIC (EPA)
LOG LINEAR ISNEE)
"I I I I I I I I I
10 15
AIR LEAD, pg/m3
20
25
Figure 11-21. Fitted equations to Kellogg Idaho/Silver Valley adjusted
blood lead data.
11-88
-------
Yankel et al. (1977), Walter et al. (1980), and Snee (1982c) make reference to a follow-
up study conducted in 1975. The second study was undertaken to determine the effectiveness of
control and remedial measures instituted after the 1974 study. Between August, 1974 and
August, 1975, the mean annual air lead levels decreased at all stations monitored. In order
of increasing distance from the smelter, the annual mean air lead levels for the one year
preceding each drawing were 18.0-10.3 ug/m3, 14.0-8.5 ug/m3, 6.7-4.9 ug/m3, and 3.1-2.5 ug/m3
at 10-24 km. Similar reductions were noted in house dust lead concentrations. In a separate
report, von Lindern and Yankel (1976) described reductions in blood lead levels of children
for whom determinations were made in both years. A number of factors complicate the interpre-
tation of the followup study, including the changes in time-varying concentrations of air lead
(Figure 11-20) from 1974 to 1975, and relocations of residence. The results demonstrated that
significant decreases in blood lead concentration resulted from exposure reductions.
11.4.1.7 Omaha. Nebraska Studies. Exposure from both a primary and secondary smelter in the
inner city area of Omaha, Nebraska, has been reported in a series of publications (Angle et
al., 1974; Angle and Mclntire, 1977, 1979; Mclntire and Angle, 1973). During 1970-1977, chil-
dren were studied from these areas: an urban school at a site immediately adjacent to a small
battery plant and downwind from two other lead emission sources; from schools in a mixed com-
mercial-residential area; and from schools in a suburban setting. Children's blood lead
levels by venipuncture were obtained by macro technique for 1970 and 1971, but Delves micro
assay was used for 1972 and later. The differences for the change in techniques were taken
into account in the presentation of the data. Air lead values were obtained by hi-vol sam-
plers and dustfall values were also monitored. Table 11-31 presents the authors' summary of
the entire data set, showing that as air lead values decrease and then increase, dustfall and
blood lead values follow. The authors used regression models, both log-linear and semilog, to
calculate (air lead)/(blood lead).
Specific reports present various aspects of the work. Black children in the two elemen-
tary schools closest to the battery plant had higher blood leads (34.1 ug/dl) than those in
elementary and junior high schools farther away (26.3 ug/dl). Best estimates of the air ex-
posures were 1.65 and 1.48 ug/m3, respectively (Mclntire and Angle, 1973). The latter study
compared three populations: urban versus suburban high school students, ages 14-18; urban
black children, ages 10-12, versus suburban whites, ages 10-12; and blacks ages 10-12 with
blood lead levels over 20 ug/dl versus schoolmates with blood lead levels below 20 ug/dl
(Angle et al., 1974). The urban versus suburban high school children did not differ signifi-
cantly, 22.3 ± 1.2 and 20.2 ± 7.0 ug/dl, respectively, with mean values of air lead concentra-
tions of 0.43 and 0.29 ug/m3. For 15 students who had environmental samples taken from their
homes, correlation coefficients between blood lead levels and soil and housedust lead levels
11-89
-------
TABLE 11-31. AIR, DUSTFALL AND BLOOD LEAD CONCENTRATIONS IN OMAHA, NE STUDY, 1970-19773
Group
All urban chi
1970-71
1972-73
1974-75
1976-77
'Air .
M9/m3 (N)D
Dustfall,
ug/m3 - mo (N)
Blood, ,
M9/dl (N)d
Idren, mixed commercial and residential site
1.48 ± 0.14(7;65)
0.43 ± 0.08(8;72)
0.10 ± 0.03(10;72)
0.52 ± 0.07(12;47)
10.6 ± 0.3(6)
6.0 ± 0.1(4)
8.8 (7)
31.4 ± 0.7(168)
23.3 ± 0.3(211)
20.4 ± 0.1(284)
22.8 ± 0.7(38)
Children at school in a commercial site
1970-71
1972-73
1974-75
1976-77
All suburban
1970-71
1972-73
1974-75
1976-77
1.69 ± 0.11(7;67)
0.63 ± 0.15(8;74)
0.10 ± 0.03(10;70)
0.60 ± 0.10(12;42)
children in a residential
0.79 ± 0.06(7;65)
0.29 ± 0.04(8;73)
0.12 ± 0.05(10;73)
25.9 ± 0.6(5)
14.3 ± 4.1(4)
33.9 (7)
site
4.6 ± 1.1(6)
2.9 ± 0.9(4)
34.6 ± 1.5(21)
21.9 ± 0.6(54)
19.2 ± 0.9(17)
22.8 ± 0.7(38)
19.6 ± 0.5(81)
14.4 ± 0.6(31)
18.2 ± 0.3(185)
aBlood lead 1970-71 is by the macro technique, corrected for an established
laboratory bias of 3 (jg/dl, macro-micro; all other values are by Delves micro
assay.
N = Number of months; number of 24-hour samples.
CN = Number of months.
N = Number of blood samples.
Source: Adapted from Angle and Mclntire, 1977.
were 0.31 and 0.29, respectively. Air, dust, and soil lead measurements at 37 sites were im-
puted to all children in the vicinity.
Suburban 10- to 12-year-olds had lower blood lead levels than their urban counterparts,
17.1 ± 0.7 versus 21.7 ±0.5 ug/dl (Angle et al., 1974). Air lead exposures were higher in
the urban than in the suburban population, although the average exposure remained less than 1
|jg/m3. Dustfall lead measurements, however, were very much higher; 32.96 mg/m2/month for ur-
ban 10- to-12-year-olds versus 3.02 mg/m2/month for suburban children.
Soil lead and house dust lead exposure levels were significantly higher for the urban
black high-lead group than for the urban low-lead group. A significant correlation (r = 0.49)
between blood lead and soil lead levels was found.
11-90
-------
Angle has reanalyzed the Omaha study (Angle et al. , 1984) using all of the data on chil-
dren from all years. There were 1075 samples from which blood lead (pg/dl), air (|jg/m3), soil
(ug/g), and house dust (H9/9) lead were available. The linear regression model, fitted in
logarithmic form, was
Pb-Blood = 15.67 + 1.92 Pb-Air + 0.00680 Pb-Soil + 0.00718 Pb-House Dust (11-16)
(±0.40) (±0.60) (±0.00097) (±0.00090)
(N = 1075, R2 = 0.20, S2 = 0.0901, GSD = 1.35)
Similar models fitted by age category produced much more variable results, possibly due to
small ranges of variation in air lead within certain age categories.
11.4.1.8 Roels et al. Studies. Roels et al. (1976, 1978, 1980) have conducted a series of
studies in the vicinity of a lead smelter in Belgium. Roels et al. (1980) report a follow-up
study in 1975 that included study populations from a rural-nonindustrialized area as well as
from the lead smelter area. The rural group consisted of 45 children (11-14 years). The
smelter area group consisted of 69 school children from three schools. These children were
divided into two groups; group A (aged 10-13) lived less than 1 km from the smelter and their
schools were very close to the smelter; group B consisted of school children living more than
1.5 km from the smelter and attending a school more distant from the smelter.
In 1974 the smelter emitted 270 kg of lead and the air lead levels were 1-2 orders of
magnitude greater than the current Belgian background concentration for air lead (0.23 ug/m3).
Soil and vegetation were also contaminated with lead; within 1 km the soil lead level was
12,250 ug/g. The concentration of lead in drinking water was less than 5 ng/1.
Environmental assessment included air, soil, and dust. Air monitoring for lead had been
continuous since September, 1973 at two sites, one for each of the two groups. In the rural
area, air monitoring was done at two sites for five days using membrane pumps. Lead was ana-
lyzed by flameless atomic absorption spectrophotometry. Dust and soil samples were collected
at the various school playgrounds, and were also analyzed by flameless atomic absorption. A
25 ml blood sample was collected from each child and immediately divided among three tubes.
One tube was analyzed for lead content by flameless atomic absorption with background correc-
tion. Another tube was analyzed for ALA-D activity while the third was analyzed for FEP. FEP
was determined by the Roels modification of the method of Sassa. ALA-D was assayed by the
European standard method.
11-91
-------
Air lead levels decreased from area A to area B. At both sites the airborne lead levels
declined over the two years of monitoring. The amount of lead produced at this smelter during
this time remained constant, about 100,000 metric tons/year. The median air lead level at the
closer site (A) dropped from 3.2 to 1.2 ug/m3, while at the far site (B) the median went from
1.6 to 0.5-0.8 ug/m3. The rural area exposure levels did not vary over the study period,
remaining rather constant at about 0.3 ug/m3.
Both smelter vicinity groups showed signs of increased lead absorption relative to the
rural population. Blood lead levels for group A were about three times those for the rural
population (26 versus 9 jjg/dl). The former blood lead levels were associated with about a 50
percent decrease in ALA-D activity and a 100 percent increase in FEP concentration. However,
FEP levels were not different for group B and rural area residents.
Later surveys of children (Roels et al., 1980) were conducted in 1976, 1977, and 1978;
the former two in autumn, the latter in spring. In total there were five surveys conducted
yearly from 1974-1978. A group of age-matched controls from a rural area was studied each
time except 1977. In 1976 and 1978 an urban group of children was also studied. The overall
age for the different groups ranged from 9 to 14 years (mean 11-12). The length of residence
varied from 0.5 to 14 years (mean 7-10 years). The subjects were always recruited from the
same five schools: one in the urban area, one in the rural area and three in the smelter area
(two <1 km and one, 2.5 km away). In all, 661 children (328 boys and 333 girls) were studied
over the years. Two hundred fourteen children came from less than 1 km from the smelter, 169
children from 1.5 to 2.5 km from the plant, 55 children lived in the urban area, and 223 chil-
dren lived in the rural area.
Air lead levels decreased from 1977 to 1978. However, the soil lead levels in the vicin-
ity of the smelter were still elevated (<1 km, soil lead = 2000-6000 ug/g). Dustfall lead in
the area of the near schools averaged 16.4-22.0 mg/m2-day at 500 m from the stack, 5.8-7.2
mg/m2-day at 700 m, about 2 mg/m2-day at 1000 m, and fluctuated around 0.5-1 mg/m2-day at 1.5
km and beyond. The particle size was predominantly 2 urn in diameter with a secondary peak
between 4 and 9 urn. The particle size declined with increasing distance from the smelter
(0.7-2.4 km).
The air lead and blood lead results for the five years are presented as Table 11-32. The
reported air leads are not calendar year averages. The table shows that blood lead levels
(electrothermal atomic absorption spectrophotometry) are lower in the girls than the boys.
Within 1 km of the smelter no consistent improvement in air lead levels was noted over the
years of the study. The mean blood leads for the children living at about 2.5 km from the
smelter never exceeded 20 ug/dl since 1975, although they were higher than for urban and rural
children.
11-92
-------
TABLE 11-32. MEAN AIRBORNE AND BLOOD LEAD LEVELS RECORDED DURING FIVE DISTINCT SURVEYS
(1974 to 1978) FOR STUDY POPULATIONS OF 11-YEAR-OLD CHILDREN LIVING LESS THAN 1 km
OR 2.5 km FROM A LEAD SMELTER, OR LIVING IN A RURAL OR URBAN AREA
Study
populations
1 Survey
(1974)
2 Survey
(1975)
3 Survey
(1976)
4 Survey
(1977)
5 Survey
(1978)
Setting
< 1 km
2.5 km
Rural
<1 km
2.5 km
Rural
<1 km
2.5 km
Urban
Rural
<1 km
2.5 km
< 1 km
2.5 km
Urban
Rural
Pb-Air,
pg/m3
4.06
1.00
0.29
2.94
0.74
0.31
3.67
0.80
0.45
0.30
3.42
0.49
2.68
0.54
0.56
0.37
Total
n
37
92
40
29
45
38
40
26
44
56
50
43
36
29
42
Blood
lead concentration,
Population
Mean ±
30.1 ±
9.4 ±
26.4 ±
13.6 ±
9.1 ±
24.6 ±
13.3 ±
10.4 ±
9.0 ±
28.9 ±
14.8 ±
27.8 ±
16.0 ±
12.7 ±
10.7 ±
SO
5.7
2.1
7.3
3.3
3.1
8.7
4.4
2.0
2.0
6.5
4.7
9.3
3.8
3.1
2.8
n
14
14
28
19
17
14
18
24
17
21
27
34
20
26
18
17
Boys
Mean ±
31.0 ±
21.1 ±
9.7 ±
27.4 ±
14.8 ±
8.2 ±
28.7 ±
15.6 ±
10.6 ±
9.2 ±
31.7 ±
15.7 ±
29.3 ±
16.6 ±
13.4 ±
11.9 ±
SD
5.5
3.4
1.6
6.5
3.6
2.1
8.0
2.9
2.0
2.3
9.5
4.8
9.8
3.5
2.3
3.0
pg/dl
n
23
64
21
12
31
20
16
9
23
29
16
23
10
11
25
Girls
Mean ±
29.6 ±
--
9.3 ±
25.4 ±
11.9 ±
9.5 ±
20.8 ±
9.8 ±
9.9 ±
8.7 ±
26.4 ±
13.0 +
26.5 ±
14.3 ±
11.5 ±
10.0 ±
SD
5.9
2.2
8.1
1.9
3.4
7.6
3.8
2.0
1.7
8.7
4.3
8.9
4.2
4.0
2.4
Source: Roels et al. (1980).
The researchers then investigated the importance of the various sources of lead in deter-
mining blood lead levels. Data were available from the 1976 survey on air, dust, and hand
lead levels. Boys had higher hand dust lead than girls. Unfortunately, the regression analy-
ses performed on these data were based on the group means of four groups.
EPA has reanalyzed the 1976 study using original data provided by Dr. Roels on the 148
children. The air lead, playground dust lead, and hand lead concentrations were all highly
correlated with each other. The hand lead measurements are used here with due regard for
their limitations, because day-to-day variations in hand lead for individual children are
believed to be very large. However, even though repeated measurements were not available,
this is among the most usable quantitative evidence on the role of ingested hand dust in
childhood lead absorption.
11-93
-------
Total lead content per hand is probably more directly related to ingested lead than is
the lead concentration in the hand dust. The linear regression model used above was fitted by
EPA using lead in air (ug/m3), lead in hand dust (ug/hand), lead in playground dust (ug/g),
and sex as covariates of blood lead. The lead variables were highly correlated, resulting in
a statistically significant regression but not statistically significant coefficients. Thus
the playground dust measurement was dropped and the following model obtained with almost as
small a residual sum of squares,
In(Pb-Blood) = ln(7.37 + 2.46 Pb-Air + 0.0195 Pb-Hand + 2.10 Male) (11-17)
(±.45)* (±.58)* (±.0062)* (±0.56)*
*Standard error of estimated regression coefficients.
The fitted model for the 148 observations gave an R2 of 0.654 and a mean square error (S2)
of 0.0836 (GSD = 1.335). The significance of the estimated coefficient establishes that
intake of lead-bearing dust from the hands of children does play a role in childhood lead ab-
sorption over and above the role that can be assigned to inhalation of air lead. Individual
habits of mouthing probably also affect lead absorption along this pathway. Note too that the
estimated inhalation slope, 2.46, is somewhat larger than most estimates for adults. However,
the effect of ingestion of hand dust appears to be almost as large as the effect of air lead
inhalation in children of this age (9-14 years). Roels et al. (1980), using group means,
concluded that the quantitative contribution of hand lead to children's blood lead levels was
far greater than that of air lead.
The high mutual correlations among air, hand, and dust lead suggest the use of their
principal components or principal factors as predictors. Only the first principal component
(which accounted for 91 percent of the total variance in lead exposure) proved a statistically
significant covariate of blood lead. In this form the model could be expressed as:
In(Pb-Blood) = ln(7.42 + 1.56Pb-Air + 0.0120Pb-Hand + 0.00212Pb-Dust + 2.29 Male) (11-18)
The estimated standard error on the inhalation slope is ±0.47. The difference between these
inhalation slope and hand lead coefficients is an example of the partial attribution of the
effects of measured lead exposure sources to those sources that are not measured.
11.4.1.9 Other Studies Relating Blood Lead Levels to Air Exposure.
The present chapter has thus far evaluated the effects of atmospheric lead on blood lead
in a disaggregate manner broken down according to exposure media, including direct inhalation
of atmospheric lead, ingestion of particulate lead that has fallen out as dust and surface
11-94
-------
soil, and air lead Ingested in consuming food and beverages (including lead absorbed from soil
and added during processing and preparation). Disaggregate analyses based on various pathways
for environmental lead of the type presented appear to provide a sensitive tool for predicting
blood lead burdens under changes of environmental exposure. However, some authors, e.g.,
Brunekreef (1984) make a strong argument for the use of air lead as the single exposure
criterion. Their argument is that exposure to air lead is usually of sufficient duration that
the contributions along other pathways have stabilized and are proportional to the air lead
concentration. In that case, the ratio between blood lead and air lead plus dust, food, and
other proportional increments must be much larger than for air lead by direct inhalation
alone.
The following studies provide information on the relationship of blood lead to air lead
exposures using aggregate analyses that include both direct and indirect air inputs. The
first group of studies are population studies which typically employed less accurate estimates
of individual exposures. The second group of studies represents industrial exposures at very
high air lead levels in which the response of blood lead appears to be substantially different
than at ambient air levels.
The Tepper and Levin (1975) study included both air and blood lead measurements. House-
wives were recruited from locations in the vicinity of air monitors. Table 11-33 presents the
geometric mean air lead and adjusted geometric mean blood lead values for this study. These
values were calculated by Hasselblad and Nelson (1975). Geometric mean air lead values ranged
from 0.17 to 3.39 ug/m3, and geometric mean blood lead values ranged from 12.7 to 20.1 M9/d1
Nordman (1975) reported a population study from Finland in which data from five urban and
two rural areas were compared. Air lead data were collected by stationary samplers. All
levels were comparatively low, particularly in the rural environment, where a concentration of
0.025 ug/m3 was seen. Urban-suburban levels ranged from 0.43 to 1.32 ug/m3.
A study was undertaken by Tsuchiya et al. (1975) in Tokyo using male policemen who
worked, but not necessarily lived, in the vicinity of air samplers. In this study, five zones
were established based on degree of urbanization, ranging from central city to suburban. Air
monitors were established at various police stations within each zone. Air sampling was con-
ducted from September, 1971 to September, 1972; blood and urine samples were obtained from
2283 policemen in August and September, 1971. Findings are presented in Table 11-34.
Goldsmith (1974) obtained data for elementary school (9- and 10-year-olds) and high
school students in 10 California communities. Lowest air lead exposures were 0.28 ug/m3 and
highest were 3.4 ug/m3. For boys in elementary school, blood lead levels ranged from 14.3 to
23.3 M9/dl; those for girls ranged from 13.8 to 20.4 ug/dl for the same range of air lead ex-
posures. The high school student population was made up of only males from some of the 10
towns. The air lead range was 0.77-2.75 ug/m3, and the blood lead range was 9.0-12.1 ug/dl.
11-95
-------
TABLE 11-33. GEOMETRIC MEAN AIR LEAD AND ADJUSTED BLOOD LEAD LEVELS FOR 11 COMMUNITIES
IN STUDY OF TEPPER AND LEVIN (1975) AS REPORTED BY HASSELBLAD AND NELSON (1975)
Geometric mean
air lead,
Community MS/1"3
Los Alamos, NM
Okeana, OH
Houston, TX
Port Washington, NY
Ardmore, PA
Lombard, IL
Washington, DC
Philadelphia, PA
Bridgeport, IL
Greenwich Village, NY
Pasadena, CA
0.17
0.32
0.85
1.13
1.15
1.18
1.19
1.67
1.76
2.08
3.39
Age and smoking
adjusted geometric
mean blood lead,
ug/dl
15.1
16.1
12.7
15.3
17.9
14.0
18.7
20.1
17.6
16.5
17.6
Sample
size
185
156
186
196
148
204
219
136
146
139
194
Multiple R2 = 0.240
Residual standard deviation = 0.262 (geometric standard deviation = 1.30)
TABLE 11-34. MEAN AIR AND BLOOD LEAD VALUES FOR FIVE ZONES IN TOKYO STUDY
Zones
1
2
3
4
5
Air lead
ug/m3
0.024
0.198
0.444
0.831
1.157
Blood lead,
ug/100 g
17.0
17.1
16.8
18.0
19.7
Source: Tsuchiya et al. 1975.
11-96
-------
The high school students with the highest blood lead levels did not come from the town with
the highest air lead value. However, a considerable lag time occurred between the collection
and analysis of the blood samples. In one of the communities the blood samples were refrig-
erated rather than frozen.
Another California study (Johnson et al., 1975, 1976) examined blood lead levels in rela-
tion to exposure to automotive lead in two communities, Los Angeles and Lancaster (a city in
the high desert). Los Angeles residents studied were individuals living in the vicinity of
heavily traveled freeways within the city. They included groups of males and females, aged 1
through 16, 17 through 34, and 34 and over. The persons selected from Lancaster represented
similar age and sex distributions. On two consecutive days, blood, urine, and fecal samples
were collected. Air samples were collected from one hi-vol sampler in Los Angeles, located
near a freeway, and two such samplers in Lancaster. The Los Angeles sampler collected for 7
days; the two in Lancaster operated for 14 days. Soil samples were collected in each area in
the vicinity of study subjects.
Lead in ambient air along the Los Angeles freeway averaged 6.3 ± 0.7 (jg/m3 and, in the
Lancaster area, the average was 0.6 ± 0.2 ug/m3. The mean soil lead in Los Angeles was 3633
ug/g, whereas that found in Lancaster was 66.9 (jg/g. Higher blood lead concentrations were
found in Los Angeles residents than in individuals living in the control area for all age
groups studied. Differences between Los Angeles and Lancaster groups were significant with
the sole exception of the older males. Snee (1981) has pointed out a disparity between blood
samples taken on consecutive days from the same child in the study. EPA reanalyses using
other criteria for outlier detection and removal obtained different inhalation slopes. This
calls into question the validity of using this study to quantify the air lead to blood lead
relationship.
Daines et al. (1972) studied black women living near a heavily traveled highway in New
Jersey. The subjects lived in houses on streets paralleling the highway at three distances:
3.7, 38.1, and 121.9 m. Air lead as well as blood lead levels were measured. Mean annual air
lead concentrations were 4.60, 2.41, and 2.24 (jg/m3, respectively, for the three distances.
The mean air lead concentration for the area closest to the highway was significantly differ-
ent from that in both the second and third, but the mean air lead concentration of the third
area was not significantly different from that of the second. The results of the blood lead
determinations paralleled those of the air lead. Mean blood lead levels of the three groups
of women, in order of increasing distance, were 23.1, 17.4, and 17.6 ng/dl respectively.
Again, the first group showed a significantly higher mean than the other two, but the second
and third groups' blood lead levels were similar to each other. Daines et al. (1972), in the
same publication, reported a second study in which the distances from the highway were 33.5
11-97
-------
and 457 m and in which the subjects were white upper middle class women. The air lead levels
were trivially different at these two distances, and the blood lead levels did not differ
either. Because the residents nearest the road were already 33 m from the highway, the dif-
ferences in air lead may have been insufficient to be reflected in the blood lead levels (see
Chapter 7).
A summary of linear relationships for other population studies has been extracted from
Snee (1981) and is shown in Table 11-35. The Fugas study is described later in Section
11.5.1.3. There is a large range of slope values (-0.1 to 3.1) with most studies in the range
of 1.0-2.0. Additional information on the more directly relevant studies is given in the
Summary Section 11.4.1.10.
TABLE 11-35. BLOOD LEAD-AIR LEAD SLOPES FOR SEVERAL POPULATION
STUDIES AS CALCULATED BY SNEE
Study
Tepper & Levin
(1975)
Johnson et al .
(1975)
Nordman (1975)
Tsuchiya et al.(1975)
Goldsmith (1974)
Fugas (1977)
Daines et al . (1972)
Johnson et al .
(1975)
Goldsmith (1974)
No.
subjects
1935
65
96
536
478
537
89
79
352
61
88a
37a
43
486
Sex
Female
Male
Female
Male
Female
Male
Male
Female
Male
Female
(spring)
Female (fall)
Male
(children)
Female
(children)
Male & female
(children)
Slope
1.1
0.8
0.8
1.2
0.6
3.1
-0.1
0.7
2.2
1.6
2.4
1.4
1.1
2.0
95% confidence
interval
±1.8
±0.7
±0.6
±1.0
±0.9
±2.2
±0.7
±0.7
±0.7
±1.7
±1.2
±0.6
±0.6
±1.3
Outlier results for four subjects deleted.
Source: Snee, 1981.
11-98
-------
A comprehensive review of studies of blood lead levels in children is presented by
Brunekreef (1984). Many of the studies did not include covariates by which air lead slopes
could be adjusted for dust or soil ingestion and other factors, leading to aggregate estimates
of air lead impacts (direct and indirect) on blood lead levels. The results of some of the
studies reviewed by Brunekreef are summarized in Table 11-36. Studies selected for Table
11-36 are those with identified air monitoring methods and reliable blood lead data. The
range of P values that Brunekreef (1984) reports is very large, and typical values of 3-5 are
larger than those adjusted slopes (1.52-2.46) derived by EPA in preceding sections.. If the
aggregate approach is accepted, then the blood lead versus total (both direct and indirect)
air lead slope for children may be approximately double the slope (~2.0) estimated for the
direct contribution due to inhaled air lead alone.
There is a great deal of information on blood lead responses to air lead exposures of
workers in lead-related occupations. Almost all such exposures are at air lead levels far in
excess of typical non-occupational exposures. The blood lead versus air lead slope p is very
much smaller at high blood and air levels. Analyses of certain occupational exposure studies
are shown in Table 11-37.
11.4.1.10 Summary of Blood Lead versus Inhaled Air Lead Relations. Any summary of the rela-
tionship of blood lead level and air lead exposure is complicated by the need for reconciling
the results of experimental and observational studies. Further, defining the form of the sta-
tistical relationship is problematical due to the lack of consistency in the range and accu-
racy of the air lead exposure measures in the various studies.
EPA has chosen to emphasize the results of studies that relate lead in air and lead in
blood under ambient conditions. At low air lead exposures there is no statistically signifi-
cant difference between curvilinear and linear blood lead inhalation relationships. Colombo
(1985) states that on the basis of experimental biological evidence, theory can provide a
steady-state relation of blood Pb to air Pb with a curved response and that the existing PbB
vs. PbA data are such that they can be fitted by several algebraically different PbA func-
tions, including a linear relationship. Colombo concludes, however, that the linear model is
preferred because it is consistent with other published models and it is much simpler in its
application. Therefore EPA has fitted linear relationships (Tables 11-38, 11-39, and 11-40)
to blood lead levels in the studies to be described next with the explicit understanding that
the fitted relationships are intended only to describe changes in blood lead due to modest
changes (of <3.0 ug/m3) in air lead among individuals whose blood lead levels do not exceed
30 pg/dl.
The blood lead inhalation slope estimates vary appreciably from one subject to another in
experimental and clinical studies, and from one study to another. The weighted slope and stan-
dard error estimates from the Griffin study in Table 11-21 (1.75 ± 0.35) were combined with
11-99
-------
TABLE 11-36. CHARACTERISTICS OF STUDIES ON THE RELATIONSHIP BETWEEN AIR LEAD AND BLOOD LEAD IN CHILDREN
O
O
Reference
Cavalleri et al. , 1981
Zielhuls et al. , 1979
Brunekreef et al , , 1981
Diemel et al . , 1981
Landrigan et al., 1975
Landrigan and Baker, 1981
Horse et al . , 1979
RoeU et al. , 1976, 1978,
1980
Yankel et al.. 1977
Walter et al., 1980
Snee, 1982c
Angle et al . , 1974
Angle and Hclntire, 1979
Billick et al. , 1979, 1980
Billick (1983)
Brunekreef et al . , 198}
Blood
Population sampling
3-6 n=110 venous
8-11 n=143
school populations, living
close to or at >4 .In fro*
a. lead shelter
1-7 n=690 (1977) venous
1-3 n=95 (1976)
volunteers (1976)
all children in area invited
(1977. 1978)
participation rate >50X
1-18 n=259 (exposed) venous
n=499 (control) 1972
n=140 (exposed) 1977
10-15 n=Z14 exposed 1974- venous
10-13 n=168 inter- 1978 puncture
mediate c cabined
10-13 n=223 rural
10-14 n=55 urban
1-9 n=1149 (1974) venous
n= 781 (1975) puncture
1-5 urban/suburban n=242 capillary
6-18 urban/ suburban/
industrial n=832 volunteers
0->6 n=178.533 venous
presented for screening
4-6 n=195 venous
nursery school populations,
living in city center or in
suburban area
Air
Quality control data sampling
yes; no interlabora- hi-vol (?)
tory comparison
yes; no information hi-vol
about participation
in inter laboratory
study
no hi-vol
yes; national and low volume
international inter-
laboratory program
no hi-vol
no hi-vol
yes, participation hi-vol
CDC blood lead
proficiency testing
program
yes; international low volume
quality control
progran
Unadjusted
slope
3.3
4.0
4.0
3.6
3.7
2.6
4.1-7.4
2.9-5.8
8.3-31.2
5.3
2.4-3.3
0.66
-2.63
2.10
15.8
24.5
18.5
Adjusted
slope Statistical model
group comparisons
3.6 group comparisons
nultiple regression,
single-log (1978)
group comparison
group comparisons
multiple regression
1-1.4 group comparisons/
multiple regression
1-2.5 single log
multiple regression,
log- log covariates
0.69 not included
5.2 multiple regression
2.9 with geometric group
means as dependent
variable
group comparisons and
8.5 multiple regression,
using log/log trans-
formations
Source: from Brunekreef (1984).
-------
TABLE 11-37. A SELECTION OF RECENT ANALYSES ON OCCUPATIONAL
8-HOUR EXPOSURES TO HIGH AIR LEAD LEVELS
Analysis
Ashf ord et al .
(1977)
King et al .
(1979)
Gartside
et al. (1982)
Bishop and
Hill (1983)
Study
Williams et al . , 1969
Globe Union
Delco-Remy
Factory 1, 1975
Factory 2a, 1975
Factory 3a, 1975
Delco-Remy,
1974-1976
Battery plants A
1975-1981 B
C
D
E
F
Air lead*,
ug/m3
50-300
35-1200
10-350
20-170
2-200
7-170
7-195
20-140
4-140
Blood lead,
(jg/dl
40-90
25-90
22-72
12-50
18-72
22-60
24-75
18-60
15-53
P
slope
0.19
0.10
0.032
0.07
0.0514
Nonlinear:
at 50:
0.081
0.045
0.048
0.022
0.045
0.101
*Assumed 8-hour exposure; divide by 3 for 24-hour equivalent.
those calculated similarly for the Rabinowitz study in Table 11-24 (2.14 ± 0.47) and the Kehoe
study in Table 11-25 (1.25 ± 0.35, setting subject DH = 0), yielding a pooled weighted slope
estimate of 1.64 ± 0.22 ug/dl per ug/m3. There are some advantages in using these experimen-
tal studies on adult males, but certain deficiencies need to be acknowledged. The Kehoe study
exposed subjects to a wide range of exposure levels while they were in the exposure chamber,
but did not control air lead exposures outside the chamber. The Griffin study provided rea-
sonable control of air lead exposure during the experiment, but difficulties in defining the
non-inhalation baseline for blood lead (especially in the important experiment at 3.2 ug/m3)
add much uncertainty to the estimate. The Rabinowitz study controlled well for diet and other
factors and since they used stable lead isotope tracers, they had no baseline problem. How-
ever, the actual air lead exposure of these subjects outside the metabolic ward was not well
determined.
Among population studies, only the Azar study provides a slope estimate in which air lead
exposures are known for individuals. However, there was no control of dietary lead intake or
other factors that affect blood lead levels, and slope estimates assuming only air lead and
11-101
-------
TABLE 11-38. CROSS-SECTIONAL OBSERVATIONAL STUDY WITH MEASURED INDIVIDUAL AIR LEAD EXPOSURE
O
IVJ
Study
tear et al. (1975)
Study done in
1970-1971 In five
U.S. cities, total
sample size = 149.
Blood leads ranged
fron 8 to 40 pg/dl.
Air leads ranged
fro* 0.2 to 9.1
MS./"3
Analysis
Azar et al.
(1975)
Snee (1982b)
Hamond et al .
(1981)
EPA
EPA
EPA
EPA
EPA
EPA
In(PBB)
In(PBB)
= 0.153 In(PBA) *
Hodel
separate intercepts
= 0.2669 InCPBA + separate background
+ 1.0842
(PBB)"1'019 = 0.179 (PBA
In(PBB)
In(PBB)
In(PBB)
In(PBB)
In(PBB)
In(PBB)
-0.098
= ln(1.318 PBA +
= ln(2.902 PBA -
for each group)
= ln[1.342 PBA +
= ln[1.593 PBA +
slope)]
= ln[1.255 PBA +
age slope)]
* separate background
for each group
for each group)
for each group)
separate background for each group)
0.257 PBA2 + separate
separate background +
background
(age slope x age)]
common intercept + (age x separate age
separate background + (age x separate
= 0.25 ln[PBA + separate background +
age slope)]
(age x separate
R2
0.502
0.497
0.49
0.491
0.504
0.499
0.489
0.521
0.514
Hodel
d.f.*
6
7
8
6
7
7
7
11
12
Slope at an air lead of
(1
(0
(0
(0
(0
(0
1.0 jig/B3
2.57
.23, 3.91)
1.12
.29, 1.94)
1.08
1.32
.46, 2.17)
2.39
1.34
.32, 2.37)
1.59
.76. 2.42)
1.26
.46, 2.05}
about 1.0
(varies by
city)
2.0 tig/"3
1.43
(0.64. 2.30)
0.96
(0.25, 1.66)
(0
(0
(0
(0
1.07
1.32
.46, 2.17)
1.87
1.34
.32, 2.37)
1.59
.76, 2.42)
1.26
.46, 2.05)
about 1.0
(varies by
city)
Note: PBB stands for blood lead (ug/dl); PBA stands for air lead (pg/m3); slope means rate of change of blood lead per unit change in air lead at the
stated air lead value. The 95 percent confidence intervals for the slope are given in parentheses. These are approximate and should be used
with caution. The analyses labeled "EPA" are calculated fron the original authors' data.
*d.f. = degrees of freedom.
-------
TABLE 11-39. CROSS-SECTIONAL OBSERVATIONAL STUDIES ON CHILDREN WITH ESTIMATED AIR EXPOSURES
Study
Kellogg Idaho/Silver
Valley study con-
ducted in 1974 based
on about 880 chil-
dren. Air leads
ranged froa 0.5 to
22 ug/B3. Blood
leads ranged fro*
11 to 164
Kellogg Idaho/Silver
Valley study as above
Analysis
Yankel et al.
(1977)
Snee (1982c)
EPA
EPA
Walter et al.
(1980)
Snee (1982a)
In(PBB)
In(PflB)
In(PBB)
In(PBB)
In(PBB)
In(PBB)
Model
= 0.041 PBA + 2.1x10 soil + 0.087 dust
- 0.018 age + 0.024 parental occupation + 3.14
= 0.039 PBA * 0.065 In (soil) + tents for sex, parental
occupation, cleanliness, education, pica
= ln(1.53 PBA + 0.0011 soil + tents for sex, parental
occupation, cleanliness, 2 education, pica}
= 1n(1.13 PBA + 0.026 PBA + teras for soil, sex, parental
occupation, cleanliness, education, pica)
= separate slopes for air, dust, parental occupation, 0.
pica, sex, and soil by age
= 0.039 PBA + 0.055 In(soil) + tents for sex, parental
occupation, cleanliness, education, pica
0
0
0
0
Rz
.622
.666
.662
.656
56 to 0.70
0.
347
Model
d.f.*
6
25
18
19
7
25
Slope at an
1.0 ug/«J
1.16
(1.09, 1.23) (1
1.13
(1.06, 1.20) (1
1.53
(1.40, 1.66) (1
1.16
air
5.0
1
.27
1
.23
1
.40
1
lead of
pg/"3
.37
, 1-
.32
. I-
.53
, 1.
.39
46)
42)
66)
1.01 to 1.26 1.18 to 1.48
1.07
(0.89, 1.25) (1
1.
.01,
,25
, 1-
50)
restricted to 537 chil-
^
H*
1
f_»
O
CO
dren with air leads
below 10 ug/*3
Roels et al.
(1900)
Angle and Hclntire
(1979)
Reels et al.
(1980) based
on 8 groups
EPA analysis
on 148 subjects
Angle and
Kclntire (1979)
on 832 samples
ages 6-18
832 samples ages
6 to 18
Angle et al.
(1984) on 1074
saaples for ages
1-18
PBB = 0.
In(PBB)
In(PBB)
In(PBB)
In(PBB)
007 PBA * 11.50 log(Pb-Mand) - 4.27
= ln(2.46 PBA + 0.0195 (Pb-Hand) * 2.1 (Male) + 7.37)
= ln(8.1) + 0.03 In(PBA) + 0.10 In(Pb-Soil)
* 0.07 ln{Pb-House Oust)
= In (4.40 PBA + 0.00457 Pb-Soil
+ 0.00336 Pb-House Oust + 16.21)
= ln(1.92 PBA + 0.00680 Pb-Soil
* 0.00718 Pb-House Dust * 15.67)
0.
0.
0.
0.
0.
65
654
21
262
199
3
4
4
4
4
0.007
2.46
(1.31, 3.61)
0.6
4.40
(3.20, 5.60)
1.92
(0.74, 3.10)
0.
007
2.46
(1
0.
4.
(3.
1.
(0.
31,
14
40
20,
92
74,
3.61)
5.60)
3.10)
Note: PBB stands for blood lead (ug/dl); PBA stands for air lead (ug/«3); slope means rate of change of blood lead per unit change in air lead at the
stated air lead value. The 95 percent confidence intervals for the slope are given in parentheses. These are approximate and should be used
with caution. The analyses labeled "EPA" are calculated from the original authors' data.
*d.f.= degrees of freedon.
-------
TABLE 11-40. LONGITUDINAL EXPERIMENTAL STUDIES WITH MEASURED INDIVIDUAL AIR LEAD EXPOSURE
Experiment
Kehoe 1950-1971
1960-1969
Griffin et al.
1971-1972
Chamberlain et
al. 1973-1978
Rabinowitz
et al. 1973-1974
Analysis
Gross (1981)
Hammond et al.(1981)
Snee (1981)
EPA
Knelson et al.(1973)
Hammond et al.(1981)
Snee (1981)
EPA
Chamberlain et al.
(197B)
EPA
Snee (1981)
EPA
A
A
A
A
A
A
A
A
A
PBB
PBB
PBB
PBB
PBB
PBB
PBB
PBB
PBB
PBB
PBB
PBB
Model
Air lead,
ug/»3
= 0.57 A PBA
= p.A PBA, p. by subject from -0.6 to 2.94
= p?A PBA, pl by subject from 0.4 to 2.4
= pl PBA + background, Pi by subject from -.34 to 2.60
= 0.327 PBA *
= p A PBA. p =
= p. A PBA. p.
= pl A PBA. pl
and p = 1.7)
= p APBA, p =
= P APBA, p =
= p. APBA. p.
= pl APBA. pl
3.236 + (2.10 PBA
1.90 at 3.2 and p
by subject, p = 2
by subject, mean
at 10.9
1.2 calculated
2.7 calculated
by subject from 1.
by subject from 1.
+ 1.96) (In PBA + p.) by subject
= 1.54 at 10.9 '
.3 at 3.2 and p = 1.5 at 10.9
P = 1.52 at 3.2
7 to 3.9
59 to 3.56
0.6
0.6
0.6
0.6
0.15
0.15
0.2
to 36
to 36
to 36
to 9
, 3.2
, 10.9
to 2
Blood lead,
ug/dl
18
18
18
18
11
14
14
to
to
to
to
to
to
to
41
41
41
29
32
43
28
-------
location as covariables (1.32 ± 0.38) are not significantly different from the pooled experi-
mental studies.
Snee and Pfeifer (1983) have extensively analyzed the observational studies, tested the
equivalence of slope estimates using pooled within-study and between-study variance com-
ponents, and estimated the common slope. The result of five population studies on adult males
(Azar, Johnson, Nordman, Tsuchiya, Fugas) was an inhalation slope estimate ±95 percent confi-
dence limits of 1.4 ± 0.6. For six populations of adult females [Tepper-Levin, Johnson,
Nordman, Goldsmith, Daines (spring), Daines (fall)], the slope was 0.9 ± 0.4. For four popu-
lations of children [Johnson (male), Johnson (female), Yankel, Goldsmith], the slope estimate
was 1.3 ± 0.4. The between-study variance component was not significant for any group so de-
fined, and when these groups were pooled and combined with the Griffin subjects, the slope
estimate for all subjects was 1.2 ± 0.2.
The Azar slope estimate was not combined with the experimental estimates because of the
lack of control on non-inhalation exposures. Similarly, the other population studies in Table
11-35 were not pooled because of the uncertainty about both inhalation and non-inhalation lead
exposures. These studies, as a group, have lower slope estimates than the individual experi-
mental studies.
There are no experimental inhalation studies on adult females or on children. The inha-
lation slope for women should be roughly the same as that for men, assuming proportionally
smaller air intake and blood volume. The assumption of proportional size is less plausible
for children. Slope estimates for children from population studies have been used in which
some other important covariates of lead absorption were controlled or measured, e.g., age,
sex, and dust exposure in the environment or on the hands. Inhalation slopes were estimated
for the studies of Angle and Mclntire (1.92 ± 0.60), Roels (2.46 ± 0.58), and Yankel et al.
(1.53 ± 0.064). The standard error of the Yankel study is extremely low and a weighted pooled
slope estimate for children would reflect essentially that study alone. In this case the
small standard error estimate is attributable to the very large range of air lead exposures of
children in the Silver Valley (up to 22 ug/m3). The relationship is in fact not linear, but
increases more rapidly in the upper range of air lead exposures. The slope estimate at lower
air lead concentrations may not wholly reflect uncertainty about the shape of the curve at
higher concentrations. The median slope of the three studies is 1.92.
This estimate was not combined with the child population studies of Johnson or Goldsmith.
The Johnson study slope estimate used air lead measured at only two sites and is sensitive to
assumptions about data outliers (Snee, 1981), which adds a large non-statistical uncertainty
to the slope estimate. The Goldsmith slope estimate for children (2.0 ± 0.65) is close to
the estimate derived above, but was not used due to non-statistical uncertainties about blood
lead collection and storage.
11-105
-------
One can summarize the situation briefly:
(1) The experimental studies at lower air lead levels, 3.2 (jg/m3 or less, and lower
blood levels, typically 30 ug/dl or less, have linear blood lead inhalation rela-
tionships with slopes p. of 0-3.6 for most subjects. A typical value of 1.64 ±
0.22 may be assumed for adults.
(2) Population cross-sectional studies at lower air lead and blood lead levels are
approximately linear with slopes p of 0.8-2.0 for inhalation contributions.
(3) Cross-sectional studies in occupational exposures in which air lead levels are
higher (much above 10 (jg/m3) anc' blood lead levels are higher (above 40 ug/dl),
show a much more shallow linear blood lead inhalation relation. The slope p is
in the range 0.03-0.2.
(4) Cross-sectional and experimental studies at levels of air lead somewhat above
the higher ambient exposures (9-36 \jg/n\3) and blood leads of 30-40 (jg/dl can be
described either by a nonlinear relationship with decreasing slope or by a
linear relationship with intermediate slope, approximately p = 0.5. Several
biological mechanisms for these differences have been discussed (Hammond et al.,
1981; O'Flaherty et al., 1982; Chamberlain, 1983; Chamberlain and Heard, 1981).
Since no explanation for the decrease in steepness of the blood lead inhalation
response to higher air lead levels has been generally accepted at this time,
there is little basis on which to select an interpolation formula from low air
lead to high air lead exposures. The increased steepness of the inhalation
curve for the Silver Valley/ Kellogg study is inconsistent with the other
studies presented. It may be that smelter situations are unique and must be
analyzed differently, or it may be that the curvature is the result of impre-
cise exposure estimates.
(5) The blood lead inhalation slope for children is at least as steep as that for
adults, with a median estimate of 1.92 from three major studies (Yankel et al.,
1977; Roels et al., 1980; Angle and Mclntire, 1979).
(6) Slopes which include both direct (inhalation) and indirect (via soil, dust, etc.)
air lead contributions are necessarily higher than those estimates for inhaled
air lead alone. Studies using aggregate analyses (direct and indirect air
impacts) typically yield slope values in the range 3-5, about double the slope
due to inhaled air lead alone.
11.4.2 Dietary Lead Exposures Including Water
Another major pathway by which lead enters the body is by ingestion. As noted in Chap-
ters 6 and 7, the recycling of both natural and anthropogenic lead in the environment results
in a certain amount of lead being found in the food we eat and the water we drink. Both of
these environmental media provide external exposures to lead that ultimately increase internal
exposure levels in addition to internal lead elevations caused by direct inhalation of lead in
air. The Nutrition Foundation (1982) report presents a compilation of recent estimates of
11-106
-------
dietary intakes in the United States and Canada. The report gives information on relation-
ships between external lead exposures and blood lead levels. The mechanisms and absorption
rates for uptake of lead from food and water are described in Chapter 10. The purpose of the
present section is to establish (analogously to Section 11.4.1) the relationships between
external exposures to lead in food and drinking water and resulting internal lead exposures.
The establishment of these external and internal lead exposure relationships for the en-
vironmental media of food and water, however, is complicated by the inherent relationship be-
tween food and water. First, the largest component of food by weight is water. Second,
drinking water is used for food preparation and, as shown in Section 7.3.1.3, provides addi-
tional quantities of lead that are appropriately included as part of external lead exposures
ascribed to food. Third, the quantity of liquid consumed daily by people varies greatly and
substitutions are made among different sources of liquid: soft drinks, coffee, tea, etc., and
drinking water. Therefore, at best, any values of water lead intake used in drinking water
calculations are somewhat problematic.
A further troubling fact is the influence of lead in the construction of plumbing facil-
ities. Studies discussed in Section 7.3.2.1.3 have pointed out the substantial lead exposures
in drinking water that can result from the use of lead pipes in the delivery of water to the
tap. This problem is thought to occur only in limited geographic areas in the United States.
However, where the problem is present, substantial water lead exposures occur. In these areas
one cannot make a simplifying assumption that the lead concentration in the water component of
food is similar to that of drinking water; rather, one is adding a potentially major addi-
tional lead exposure to the equation.
Studies that have attempted to relate blood lead levels to ingested lead exposure have
used three approaches to estimate the external lead exposures involved: duplicate meals, fe-
cal lead determinations, and market basket surveys. In duplicate diet studies, estimated lead
exposures are assessed by having subjects put aside a duplicate of what they eat at each meal
for a limited period of time. These studies probably provide a good, but short term, estimate
of the ingestion intake. However, the procedures available to analyze lead in foods have his-
torically been subject to inaccuracies. Hence, the total validity of data from this approach
has not been established. Studies relying on the use of fecal lead determinations face two
major difficulties. First, this procedure involves the use of a mathematical estimate of the
overall absorption coefficient from the gut to estimate the external exposure. Until recent-
ly, these estimates have not been well documented and were assumed to be relatively constant.
Newer data discussed later show a much wider variability in the observed absorption coeffici-
ents than was thought to be true. These new observations cloud the utility of studies using
this method to establish external/internal exposure relationships. Secondly, it is difficult
to collect a representative sample.
11-107
-------
The last approach is the market basket approach. This approach uses the observed lead
concentrations for a variety of food items coupled with estimated dietary consumption of the
particular food items. Some studies use national estimates of typical consumption patterns
upon which to base the estimated exposures. Other studies actually record the daily dietary
intakes. This approach faces similar analytic problems to those found in the duplicate diet
approach. It also faces the problem of getting accurate estimates of dietary intakes. The
most current total diet study (Pennington, 1983) is described in Section 7.3.1.2.
Exposures to lead in the diet are thought to have decreased since the 1940's. Estimates
from that period were in the range of 400-500 ug/day for U.S. populations. Khandekar et al.
(1984) report a dietary intake of lead to be 245 ug/day. This was calculated from the lead
content in different food groups and the amount of each food group consumed by an average
resident of Bombay, India. Current estimates for U.S. populations are under 100 ug/day for
adults. Unfortunately, a good historical record regarding the time course of dietary expo-
sures is not available. In the years 1978-1982, efforts have been made by the American food
canning industry in cooperation with the FDA to reduce the lead contamination of canned food.
Data presented in Section 7.3.1.2.5 confirm the success of this effort. Seasonal variations
in blood lead might also be partially attributable to seasonal variations in the dietary in-
take of lead. The following evidence suggests that this does not happen. Table 11-41 is
taken from Human Nutrition Information Service (1983). The data suggest the following
pattern: (1) Consumption of canned vegetables and fruits is much lower in the spring and
summer, much higher in the fall and winter, which is the opposite of the pattern of blood lead
level variations and suggests that the attribution of seasonal changes to gasoline lead may be
an underestimate of its effects. (2) The pattern is similar for central city, suburban, and
nonmetropolitan households. (3) There is little seasonal variation for fruit and vegetable
juices and milk, and a slight increase of soft drink consumption in the summer. The magnitude
of such variations is too small to account for blood lead.
The specific studies available for review regarding dietary exposures will be organized
into three major divisions: lead ingestion from typical diets, lead ingestion from experimen-
tal dietary supplements, and inadvertent lead ingestion from lead plumbing.
11.4.2.1 Lead Ingestion from Typical Diets.
11.4.2.1.1 Ryu study on infants and toddlers. Ryu et al. (1983) reported a study of four
breast-fed infants and 25 formula-fed infants from 8-196 days of age. At 112 days of the
study, the formula-fed infants were separated into subgroups based upon how they were to re-
ceive their milk: homogenized whole cow milk obtaine'd in cartons from a local dairy, a com-
mercially available milk-based formula supplied in quart cans, and homogenized whole cow milk
11-108
-------
TABLE 11-41. HOUSEHOLD CONSUMPTION OF CANNED FOODS
(pounds per week)
Food
Canned fruits*
Central city
Suburban
Nonmetropolitan
Canned vegetables*
Central city
Suburban
Nonmetropolitan
Fresh fluid milk
Central city
Suburban
Nonmetropolitan
Processed milk
Central city
Suburban
Nonmetropolitan
Canned veg. juices*
Central city
Suburban
Nonmetropolitan
Canned fruit juices*
Central city
Suburban
Nonmetropolitan
Soft drinks (total)
Central city
Suburban
Nonmetropolitan
Spring
0.65
0.85
0.83
2.37
2.40
2.37
13.44
17.66
15.11
1.14
1.43
1.56
0.39
0.42
0.56
1.34
1.16
1.29
5.50
6.53
5.67
Summer
0.47
0.55
0.62
2.36
2.08
1.94
14.20
17.12
16.17
1.12
1.13
1.36
0.38
0.41
0.38
1.46
1.26
1.22
5.75
6.88
5.89
Fall
0.59
0.84
0.78
2.81
2.57
2.46
14.31
17.38
16.16
1.18
1.10
1.59
0.37
0.54
0.46
1.39
1.25
1.49
5.11
6.22
5.62
Winter
0.74
0.91
0.85
2.83
2.86
2.89
13.75
17.17
16.70
1.30
1.14
1.90
0.35
0.47
0.53
1.41
1.24
1.35
5.35
5.96
5.25
*Commercially canned.
supplied in quart cans and heat-treated in the same manner as the commercially available for-
mula. There were 10, 4, and 3 infants in each of these groups, respectively. In addition to
food concentrations, data were collected on air, dust, and water lead. Hemoglobin and FEP
were also measured.
11-109
-------
The trends in blood lead for the formula-fed infants are shown in Table 11-42. The re-
sults up to day 112 are averaged for all 25 infants. The estimated average intake was 17
ug/day for this time period. After day 112, the subgroup of seven infants fed either canned
formula or heat-treated cow's milk in cans (higher lead), had average estimated lead intake of
61 ug/day. This resulted in an increase of 7.2 ug/dl in the average blood lead level in
response to an increase of 45 ug/day in lead intake by day 196. However, since the blood lead
levels in this group had not reached equilibirum by this point, the slope calculated from this
data of 0.16 should be regarded as an underestimate.
TABLE 11-42. BLOOD LEAD LEVELS AND LEAD INTAKE VALUES FOR INFANTS IN THE STUDY OF RYU ET AL.
Age,
days
8
28
56
84
112
140
168
196
Blood
lead of
combined
group, ug/dl
Lower lead
6.2
7.0
7.2
8.9
5.8
5.1
5.4
6.1
Higher lead
9.3
12.1
14.4
Average lead
intake of
combined group, ug/day
17
17
17
17
17
Lower lead
16
16
16
Higher lead
61
61
61
Source: Ryu et al. (1983).
11.4.2.1.2 Rabinowitz infant study. As part of a longitudinal study of the sources of cur-
rent urban lead exposure, lead was measured in 100 breast milk samples and in 73 samples of
the infant formula used by non-nursing mothers (Rabinowitz et al., 1985a). Also, the blood
lead levels of the infants fed these diets were determined at birth and at six months of age.
Among the infants who were breast-fed, the lead content of their milks correlated very well
with their six-month blood lead levels (r = 0.42, p = 0.0003). The mean lead content of in-
fant formulas and breast milk were not significantly different, nor was the blood lead of
children fed one or the other. Lead levels in maternal milk correlated poorly with umbilical
cord blood lead (r = 0.18, p = 0.10). Since milk represents much of the diet of young infants
and because breast milk lead levels are stable, it is possible to relate blood lead and daily
dosage in this population.
11-110
-------
11.4.2.1.3 Rabinowitz adult study. This study on male adults was described in Section
11.4.1 and in Chapter 10, where ingestion experiments were analyzed in more detail (Rabinowitz
et al., 1980). As in other studies, the fraction of ingested stable isotope lead tracers ab-
sorbed into the blood was much lower when lead was consumed with meals (10.3 ±2.2 percent)
than between meals (35 ± 13 percent). Lead nitrate, lead sulfide, and lead cysteine as car-
riers made little difference. The much higher absorption of lead on an empty stomach implies
greater significance of lead ingestion from leaded paint and from dust and soil when consumed
between meals, as seems likely to be true for children.
11.4.2.1.4 Hubermont study. Hubermont et al. (1978) conducted a study of pregnant women liv-
ing in rural Belgium because their drinking water was suspected of being lead-contaminated.
This area was known to be relatively free of air pollution. Seventy pregnant women were re-
cruited and asked to complete a questionnaire. Information was obtained on lifetime residence
history, occupational history, smoking, and drinking habits. First flush tap water samples
were collected from each home with the water lead level determined by flameless atomic absorp-
tion spectrophotometry. Biological samples for lead determination were taken at delivery. A
venipuncture blood sample was collected from the mother, as was a fragment of the placenta; an
umbilical cord blood sample was used to estimate the newborn's blood lead status.
For the entire population, first-flush tap water samples ranged from 0.2 to 1228.5 ug/1.
The mean was 109.4, while the median was 23.2. The influence of water lead on the blood lead
of the mother and infants was examined by categorizing the subjects on the basis of the lead
level of the water sample, below or above 50 pg/1. Table 11-43 presents the results of this
study. A significant difference in blood lead levels of mothers and newborns was found for
the water lead categories. Placenta lead levels also differed significantly between water
lead groups. The fitted regression equation of blood lead level for mothers is given in
summary Table 11-51 in section 11.4.2.4.
11.4.2.1.5 Sherlock studies. Sherlock et al. (1982) reported a study from Ayr, Scotland,
which considered both dietary and drinking water lead exposures for mothers and children
living in the area. In December, 1980, water lead concentrations were determined from kettle
water from 114 dwellings in which the mother and child lived less than five years. The adult
women had venous blood samples taken in early 1981 as part of a European Economic Community
(EEC) survey on blood lead levels. A duplicate diet survey was conducted on a random sample
of these 114 women stratified by kettle water lead levels.
A study population of 11 mothers with infants less than 4 months of age agreed to parti-
cipate in the infant survey. A stratified sample of 31 of 47 adult volunteers was selected to
participate in the duplicate diet study.
Venous blood samples for adults were analyzed for lead immediately before the duplicate
diet study; in some instances additional samples were taken to give estimates of long-term
11-111
-------
TABLE 11-43. INFLUENCE OF LEVEL OF LEAD IN WATER ON BLOOD LEAD LEVEL IN BLOOD AND PLACENTA
Comparison
group
Age (years)
Pb-B mother
(ug/di)
Pb-B newborn
(Hg/dl)
Pb placenta
(M9/100 g)
Water Pb
(ug/i)
Water
level
Low**
High***
Low
High
Low
High
Low
High
Low
High
Mean
25.6
26.3
10.6
13.8
8.8
12.1
9.7
13.3
11.8
247.4
Median
24
25
9.9
13.1
8.5
11.9
8.2
12.0
6.3
176.8
Range
18-41
20-42
5.1-21.6
5.3-26.3
3.4-24.9
2.9-22.1
4.4-26.9
7.1-28
0.2-43.4
61.5-1228.5
Significance
NS*
<0.005
<0.001
<0.005
Source: Hubermont et al. (1978)
*NS means not significant.
**Water lead <50 ug/1.
***Water lead >50 ug/1.
exposure. Venous samples were taken from the infants immediately after the duplicate diet
week. Blood lead levels were determined by AAS with a graphite furnace under good quality
control. Two other laboratories analyzed each sample by different methods. The data reported
are based on the average value of the three methods.
Dietary intakes for adults and children were quite different; adults had higher intakes
than children. Almost one-third of the adults had intakes greater than 3 mg/week while only
20 percent of the infants had that level of intake. Maximum values were 11 mg/week for adults
and 6 mg/week for infants. The observed blood lead values in the dietary study had the dis-
tributions shown in Table 11-44.
Table 11-45 presents the crosstabulation of drinking water lead and blood lead level for
the 114 adult women in the study. A strong trend of increasing blood lead levels with increa-
sing drinking water lead levels is apparent. A curvilinear regression function fits the data
better than a linear one. A similar model including weekly dietary intake was fitted to the
data for adults and infants. These models are in summary Tables 11-49 and 11-52 in Section
11.4.2.4.
11-112
-------
TABLE 11-44. DISTRIBUTIONS OF OBSERVED BLOOD LEAD VALUES IN AYR
Groups
Adults
Infants
EEC directive
>20 ug/dl
55%
100%
50%
Blood lead values
>30 |jg/dl
16%
55%
10%
>35 ug/dl
2%
36%
2%
TABLE 11-45. BLOOD LEAD AND KETTLE WATER LEAD CONCENTRATIONS
FOR ADULT WOMEN LIVING IN AYR
Water lead, ug/1
Blood lead,
ug per 100 ml
<10
11-15
16-20
21-25
26-30
31-35
36-40
>40
Total
11-
<10 99
8 5
4 7
1 3
4
13 19
100-
299
3
12
9
2
2
28
300-
499
2
3
7
4
1
1
1
19
500-
999
3
5
4
2
1
4
19
1000-
1499
2
2
1
3
8
>1500
1
3
1
3
8
Total
13
17
22
25
12
10
4
11
114
The researchers also developed a linear model for the relationship between dietary intake
and drinking water lead. The equation indicates that, when the concentration of lead in water
was about 100 ug/1, approximately equal amounts of lead would be contributed to the total
week's intake from water and diet; as water lead concentrations increase from this value, the
principal contributor would be water.
A follow-up study on this same population was made from December, 1982 to March, 1983, as
reported by Sherlock et al. (1984). In April 1981, the pH of the water supply was increased
from pH 4.5-5.5 to about pH 8.5 by the addition of lime. The result was a decrease in the
median blood lead level from 21 to 13 ug/dl. The combined data set was used to give the re-
gression equation shown in Table 11-52 in Section 11.4.2.4.
11-113
-------
11.4.2.1.6 Central Directorate on Environmental Pollution study. The United Kingdom Central
Directorate on Environmental Pollution (1982) studied the relationship between blood lead
level and dietary and drinking water lead in infants. Subjects were first recruited by solic-
iting participation of all pregnant women attending two hospitals and residing within a single
water distribution system. Each woman gave a blood sample and a kettle water sample. The
women were then allocated to one of six potential study groups based on the concentration of
water lead.
At the start of the second phase (duplicate diet) a total of 155 women volunteered
(roughly 17-32 per water lead level category). During the course of the study, 24 mothers
withdrew; thus a final study population of 131 mothers was achieved.
When the children reached 13 weeks of age, duplicate diet for a week's duration was ob-
tained for each infant. Great care was exerted to allow collection of the most accurate
sample possible. Also, at this time a variety of water samples were collected for subsequent
lead analysis.
Blood samples were collected by venipuncture from mothers before birth, at delivery, and
about the time of the duplicate diet. A specimen was also collected by venipuncture from the
infant at the time of the duplicate diet. The blood samples were analyzed for lead by graph-
ite furnace AAS with deuterium background correction. Breast milk was analyzed analogously to
the blood sample after pretreatment for the different matrix. Water samples were analyzed by
flame atomic absorption; food samples were analyzed after ashing by flameless atomic absorp-
tion.
Both mothers and infants exhibited increased lead absorption by EEC (European Economic
Community) directive standards. The infants generally had higher blood leads than the
mothers. However, in neither population was there evidence of substantial lead absorption.
Water lead samples ranged from less than 50 to greater than 500 ug/1, which was expected
due to the sampling procedure used. First draw samples tended to be higher than the other
samples. The composite kettle samples and the random daytime samples taken during the dupli-
cate diet week were reasonably similar: 59 percent of the composite kettle samples contained
up to 150 ug/1, as did 66 percent of the random daytime samples.
Lead intakes from breast milk were lower than from duplicate diets. The lead intakes
estimated by duplicate diet analysis ranged from 0.04 to 3.4 mg/week; about 1/4 of the diets
had intakes less than 1.0 mg/week. The minimum intakes were truncated, as the limit of detec-
tion for lead was 10 ug/kg and the most common diets weighed 4 kg or more.
The central directorate data were reanalyzed by Lacey et al. (1985). Results from both
Lacey et al. (1985) and the United Kingdom Central Directorate on Environmental Pollution
(1982) are in Tables 11-49 to 11-52 in section 11.4.2.4. The authors used both linear and
cube root models to describe their data. Models relating blood lead levels of infants to
11-114
-------
dietary intake are in Table 11-49 in Section 11.4.2.4. Models relating blood lead levels for
both mothers and infants to first flush water lead levels and running water lead levels are in
Tables 11-51 and 11-52 in Section 11.4.2.4 respectively. In most cases, the nonlinear (cubic)
model provided the best fit. Figure 11-22 illustrates the fit for the two models showing in-
fant blood lead levels versus dietary lead intake.
11.4.2.1.7 Pocock study. Pocock et al. (1983) have recently reported an important study ex-
amining the relationship in middle-aged men of blood lead level and water lead levels. Men
aged 40-59 were randomly selected from the registers of general practices located in 24
British towns. Data were obtained between January, 1978 and June, 1980.
Blood lead levels were obtained on 95 percent of the 7378 men originally selected. The
levels were determined by microatomic absorption spectrophotometry. A strict internal and ex-
ternal quality control program was maintained on the blood lead determinations for the entire
study period. Tap water samples were obtained on a small subset of the population. About 40
men were chosen in each of the 24 towns to participate in the water study. First draw samples
were collected by the subjects themselves, while a grab daytime and flushed sample were col-
lected by study personnel. These samples were analyzed by several methods of AAS depending
on the concentration range of the samples.
Blood lead and water lead levels were available for a total of 910 men from 24 towns.
Table 11-46 displays the association between blood lead levels and water lead levels. Blood
lead levels nearly doubled from the lowest to highest water lead category.
The investigators analyzed their data further by examining the form of the relationship
between blood and water lead. This was done by categorizing the water lead levels into nine
intervals of first draw levels. The first group (<6 M9/D nad 473 men while the remaining
eight intervals had * 50 men each. Figure 11-23 presents the results of this analysis. The
authors state, "The impression is that mean blood lead increases linearly with first draw
water lead except for the last group with very high water concentrations." The regression
line shown in the figure is only for men with water lead levels less than 100 ug/1, and is
given in Table 11-51 in Section 11.4.2.4. A separate regression was done for the 49 men whose
water lead exposures were greater than 100 MS/I- The slope for the second line was only 23
percent of the first line.
Additional analyses were done examining the possible influence of water hardness on blood
lead levels. A strong negative relationship (r = 0.67) was found between blood lead level and
water hardness. There is a possibility that the relationship between blood lead and water
hardness was due to the relationship of water hardness and water lead. It was found that a
relationship with blood lead and water hardness still existed after controlling for water lead
level.
11-115
-------
E
8
t«»
1
Q
§
00
10
1.0
2.0
LEAD INTAKE, mg/wk
3.0
Figure 11 -22. Blood lead concentrations versus weekly lead
intake for bottle-fed infants. (Numbers are coincidental points.)
Source: United Kingdom Central Directorate on Environmental
Pollution (1982).
11-116
-------
TABLE 11-46. RELATIONSHIP OF BLOOD LEAD
AND WATER LEAD IN 910 MEN AGED 40-59 FROM 24 BRITISH TOWNS
First draw
water lead,
Mg/i
<50
50-99
100-299
£300
Total
Number of
men
789
69
40
12
910
Mean blood
lead
(M9/dl)
15.06
18.90
21.65
34.19
15.89
Standard
deviation
5.53
7.31
7.83
15.27
6.57
% with
blood lead
>35 pg/dl
0.7
4.3
7.5
41.7
1.9
Daytime
water lead,
H9/1
<50
50-99
100-299
£300
Total
845
36
23
5
909
15.31
19.62
24.78
39.78
15.85
5.64
7.89
9.68
15.87
6.44
0.7
8.3
17.4
60.0
1.8
Source: Pocock et al. (1983).
11-117
-------
1.25
1.2
1.0
§. 0.9
O*
O
O 0.8
0.7
o
50
100
320
FIRST DRAW WATER LEAD, ng/l
350
imiiii i -
61 52
473 60 51 50 65 49 49
Figure 11-23. Mean blood lead for men grouped by first draw water concentra-
tion.
Source: Pocock et al. (1983).
11-118
-------
The authors come to the following conclusion regarding the slope of the relationship be-
tween blood lead and water lead:
This study confirms that the relation is not linear at higher levels. Previous
research had suggested a power function relationshipfor example, blood lead in-
creases as the cube root of water lead. Our data, based on a large and more
representative sample of men, do not agree with such a curve, particularly at low
concentrations of water lead.
11.4.2.1.8 Thomas study. Thomas et al. (1981) studied blood lead levels among residents of a
hardwater area in the United Kingdom. They recruited a random sample of voters in an area
with 320 ppm calcium hardness. A tap water sample using first draw water was requested and
was returned by 70 percent of the selected voters. Sixty women in the dwellings with the
highest water blood level and 30 randomly selected women in dwellings in the lowest water lead
levels were selected for a blood lead determination; 84 women responded. Blood lead levels
were stratified by water lead levels and were compared to data gathered elsewhere from soft-
water areas. Substantial differences were noted, with the residents of the hardwater areas
having meaningfully lower blood lead levels. This is true even for residents in the hardwater
area with the lowest (<0.05 mg/1) water lead level.
11.4.2.1.9 Elwood study. Elwood et al. (1983) have investigated the potential of the degree
of water hardness to influence the relationship between lead concentrations in drinking water
and blood lead level. An experimental model was employed wherein two groups of women were
studied both before and after the water hardness of the drinking water for one group was
changed to 100 from 10 mg/1. Postconversion blood lead levels were obtained 6 months later.
Mean water lead levels fell slightly after the change in the area where the water was
hardened, whereas it increased slightly in the central area. Blood lead levels decreased in
the experimental areas while increasing in the central area. The decline in blood lead levels
was greater with increasing initial water lead levels.
11.4.2.2. Lead Ingestion from Experimental Dietary Supplements.
11.4.2.2.1 Kehoe study. Experimental studies have been used to study the relationship of
food lead and blood lead levels. Gross (1981) reanalyzed the results of Kehoe. Oral doses of
lead included 300, 1000, 2000, and 3000 ug/day. Each subject had a control period and an ex-
posure period. Some also had a post-exposure period. Blood samples were collected by veni-
puncture and analyzed by spectrographic and dithizone methods during the study years. The
ingestion doses were in addition to the regular ingestion of lead from the diet. The results
of the dose response analysis for blood lead concentrations are summarized in Table 11-47.
11-119
-------
TABLE 11-47. DOSE-RESPONSE ANALYSIS FOR BLOOD LEAD LEVELS IN THE KEHOE STUDY
AS ANALYZED BY GROSS (1981)
Difference from control1
Subject
SW
MR
EB
IF2
Added lead,
(jg/day
300
1000
2000
3000
Diet,
|jg/day
308
1072
1848
2981
Feces,
|jg/day
208
984
1547
2581
Urine,
pg/day
3
55
80
49
Blood,
Mg/dl
-1
17
33
19
JEach subject servced as his own control.
2Subject did not reach equilibrium.
Both subjects MR and EB had long exposure periods, during which time their blood lead
levels increased to equilibrium averages of 53 and 60 pg/dl, respectively. The exposure for
IF was terminated early before his blood lead had achieved equilibrium. No response in blood
lead was seen for subject SW whose supplement was 300 |jg/day.
11.4.2.2.2 Stuik study. Stuik (1974) administered lead acetate in two dose levels (20 and 30
ug/kg-day) to volunteers. The study was conducted in two phases. The first phase was con-
ducted for 21 days during February-March, 1973. Five males and five females aged 18-26 were
exposed to a daily dose of 20 ug Pb2+/kg. Five males served as controls. In the second
phase, five females received 20 ug Pb2+/kg and five males received 30 ug Pb2+/kg. Five
females served as controls. Pre-exposure values were established during the week preceding
the exposures in both phases. Blood lead levels were determined by Hessel's method.
The results of phase I for blood lead levels are presented in Figure 11-24. Blood lead
levels appeared to achieve an equilibrium after 17 days of exposure. Male blood lead levels
went from 20.6 to 40.9 (jg/g while females went from 12.7 to 30.4 ug/g. The males seemed to
respond more to the same body weight dose.
In phase II, males were exposed to a higher lead dose (30 ug/kg-day). Figure 11-25 dis-
plays these results. Male blood lead rose higher than in the first study (46.2 versus 40.9
ug/g); furthermore, there was no indication of a leveling off. Females also achieved a higher
blood lead level (41.3 versus 30.4 ug/dl), which the author could not explain. The pre-
exposure level, however, was higher for the second phase than the first phase (12.7 versus
17.3 pg/g).
11-120
-------
500
II I I
- CONTROL GROUP
EXPOSED MALE SUBJECTS. 20 ug'kg'day
> EXPOSED FEMALE SUBJECTS: 20 Mfl'kg'day
a
a
CD 300
£
100
/
/.
Pb EXPOSURE-
1 1
.
.
Ca EDTA
n
.MALE GROUP
^Ca EDTA ^
FtMALE GROUP .
I i
13 8 10 IS 17 22 29 31
DAYS
Figure 11-24. Average PbB levels, Exp. I.
Source: Stuik (1974).
38
46
800
1
300
100
II II II II
i CONTROL GROUP
_ _ EXPOSED MALE SUBJECTS: 30 -g kg 'day
. . _ EXPOSED FEMALE SUBJECTS: 20 ug/kg'day
s'*ls''~ ^
/"""" _.-^>*"*"
<'>
"2^^-^^ ^-^^
II II II II
20 47 11 14 18 21
DAYS
I I I
\ ~~
"*'""^_
v
Ca EDTA
1 i MALE GROUP i
26 27 34
Figure 11-25. Average PbB levels, Exp. II.
Source: Stuik (1974).
-------
11.4.2.2.3 Cools study. Cools et al. (1976) extended the research of Stuik (1974) by random-
ly assigning 21 male subjects to two groups. The experimental group was to receive a 30 (jg/kg
body weight dose of oral lead acetate for a period long enough to achieve a blood lead level
of 30.0 M9/9> when the lead dose would be adjusted downward to attempt to maintain the sub-
jects at a blood lead level of 40.0 ug/g. The other group received a placebo.
In the pre-exposure phase, blood lead levels were measured three times, while during ex-
posure they were measured once a week, except for the first three weeks when they were deter-
mined twice a week. Blood lead was measured by flame AAS according to the Westerlund modifi-
cation of Hessel's method.
Pre-exposure blood lead values for the 21 volunteers averaged 172 ppb. The effect of
ingestion of lead acetate on blood lead is displayed in Figure 11-26. After 7 days, mean
blood lead levels had increased from 17.2 to 26.2 ug/g. The time to reach a blood lead
level of 35.0 pg/g took 15 days on the average (range 7-40 days).
11.4.2.2.4 Schlegel study. Schlegel and Kufner (1979) report an experiment in which two sub-
jects received daily oral doses of 5 mg Pb2+ as an aqueous solution of lead nitrate for 6 and
13 weeks, respectively. Blood and urine samples were taken. Blood lead uptake (from 16-60
ug/dl in 6 weeks) and washout were rapid in subject HS, but less so in subject GK (from 12-29
ug/dl in 6 weeks). Time series data on other heme system indicators (FEP, ALA-D, ALA-U,
coproporphyrin III) were also reported.
11.4.2.2.5 Chamberlain study. This study (Chamberlain et al., 1978) was described in Section
11.4.1, and in Chapter 10. The ingestion studies on six subjects showed that the gut absorp-
tion of lead was much higher when lead was ingested between meals. There were also differ-
ences in absorption of lead chloride and lead sulfide.
11.4.2.3 Inadvertent Lead Ingestion from Lead Plumbing.
11.4.2.3.1 Early studies. Although the use of lead piping has been largely prohibited in
recent construction, occasional episodes of poisoning from this lead source still occur.
These cases most frequently involve isolated farms or houses in rural areas, but a surprising
urban episode was revealed in 1972 when Beattie et al. (1972a,b) showed the seriousness of the
situation in Glasgow, Scotland, which had very pure, but soft, drinking water as its source.
The researchers demonstrated a clear association between blood lead levels and inhibition of
the enzyme ALA-D in children living in houses with (1) lead water pipes and lead water tanks,
(2) no lead water tank but with more than 60 ft of lead piping, and (3) less than 60 ft of
lead piping. The mean lead content of the water as supplied by the reservoir was 17.9 ug/1;
those taken from the faucets of groups 1, 2, and 3 were 934, 239, and 108 ug/1, respectively.
11-122
-------
450
£
«n
£
i
EXPOSED (n
O CONTROLS
-------
Another English study (Crawford and Crawford, 1969) showed a clear difference between the
bone lead contents of the populations of Glasgow and London, the latter having a hard, nonsol-
vent water supply. In a study of 1200 blood donors in Belgium (De Graeve et al., 1975),
persons from homes with lead piping and supplied with corrosive water had significantly higher
blood lead levels.
11.4.2.3.2 Moore studies. Moore and colleagues have reported on several studies relating
blood lead levels to water lead levels. Moore (1977) studied the relationship between blood
lead level and drinking water lead in residents of a Glasgow tenement. The tenement was
supplied with water from a lead-lined water tank carried by lead piping. Water samples were
collected during the day. Comparative water samples were collected from houses with copper
pipes and from 15 lead-plumbed houses. Blood samples were taken wherever possible from all
inhabitants of these houses. The data indicated that if a house has lead-lined pipes, it is
almost impossible to reach the WHO standard for lead in water (100 ug/1). Linear regression
equations relating blood lead levels to first flush and running water lead levels are in
Tables 11-51 and 11-52 in Section 11.4.2.4.
Moore (1977) also reported the analysis of blood lead and water lead data collected over
a four-year period for different sectors of the Scottish population. The combined data showed
consistent increases in blood lead levels as a function of first draw water lead, but the
equation was nonlinear at the higher range. The water lead values were as high as 2000 ug/1.
The fitted regression equation for the 949 subjects is in Table 11-51 in Section 11.4.2.4.
Moore et al. (1981a,b) reported a study of the effectiveness of control measures for
plumbosolvent water supplies. In autumn and winter of 1977, they studied 236 mothers aged
17-37 in a postnatal ward of a hospital in Glasgow with no historical occupational expo-
sure. Blood lead and tap water samples from the home were analyzed for lead by AAS under a
quality control program.
A skewed distribution of blood lead levels was obtained with a median value of 16.6
ug/dl; 3 percent of the values exceeded 41 ug/dl. The geometric mean was 14.5 ug/dl. A cur-
vilinear relationship between blood lead level and water lead level was found. The log of the
maternal blood lead varied as the cube root of both first flush and running water lead concen-
trations. In Moore et al. (1979), further details regarding this relationship are provided.
Figure 11-27 presents the observed relationship between blood lead and water lead.
In April, 1978, a closed loop lime dosing system was installed. The pH of the water was
raised from 6.3 to 7.8. Before the treatment, more than 50 percent of random daytime water
samples exceeded 100 ug/1, the WHO standard. After the treatment was implemented, 80 percent
of random samples were less than 100 ug/1. It was found, however, that the higher pH was not
maintained throughout the distribution system. Therefore, in August, 1980, the pH was raised
11-124
-------
5
3.
O
O
OQ
235
25
24 26 25
24
NO IN
GROUP
Figure 11 -27. Cube root regression of blood lead on first flush water
lead. This shows mean ± S.D. of blood lead for pregnant women
grouped in 7 intervals of first flush water lead.
Source: Moore et al. (1979).
11-125
-------
to 9 at the source, thereby maintaining the tap water at 8. At this time, more than 95 percent
of random daytime samples were less than 100 ug/1.
In the autumn and winter of 1980, 475 mothers from the same hospital were studied. The
median blood lead was 6.6 ug/dl and the geometric mean was 8.1 ug/dl Comparison of the fre-
quency distributions of blood lead between these two blood samplings show a remarkable drop.
No other source of lead was thought to account for the observed change.
Sherlock et al. (1984) report that water treatment produced a sharp fall in water lead
concentrations and a decrease in the median blood lead concentrations from 21 to 13 ug/dl.
11.4.2.3.3 Thomas study. Thomas et al. (1979) studied women and children residing on two
adjacent housing estates. One estate was serviced by lead pipes for plumbing while the other
was serviced by copper pipe. In five of the homes in the lead pipe estate, the lead pipe had
been replaced with copper pipe. The source water is soft, acidic, and lead-free.
Water samples were collected from the cold tap in the kitchen in each house on three oc-
casions at two-week intervals. The following water samples were collected: daytime - first
water out of tap at time of visit; running - collected after tap ran moderately for 5 minutes
after the daytime sample; and first flush - first water out of tap in morning (collected by
residents). Lead was analyzed by a method (unspecified in report) that was reportedly under
quality control.
Blood samples were collected from adult females (2.5 ml venipuncture) who spent most of
the time in the home and from the youngest child (capillary sample). Blood samples were ana-
lyzed for lead by a quality-controlled unspecified method. Blood lead levels were higher in
the residents of the lead estate homes than in the residents of the copper estate homes.
Median levels for adult females were 39 and 14.5 ug/dl for the lead and copper estate homes
respectively. Likewise, children's blood lead levels were 37 and 16.6 M9/dl, respectively.
Water lead levels were substantially higher for the lead estate than for the copper estate.
This was true for all three water samples.
The researchers then monitored the effectiveness of replacing the lead pipe on reducing
both exposure to lead in drinking water and, ultimately, blood lead levels. This monitoring
was done by examining subsamples of adult females for up to 9 months after the change was
implemented. Water lead levels became indistinguishable from those found in the copper estate
homes. Blood lead levels declined about 30 percent after 3-4 months and 50 percent at 6
and 9 months. At 6 months the blood lead levels reached those of women living in the copper
estates. A small subgroup of copper estate females was also followed during this time. No
decline was noted among them. Therefore, it was very likely that the observed reduction in
blood lead levels among the other women was due to the changed piping.
11-126
-------
The researchers then analyzed the form of the relationship between blood lead levels and
water lead levels. They tried several different shapes for the regression line. Curvilinear
models provided better fits. Figure 11-28 depicts the scatter diagram of blood lead and water
lead. An EPA analysis of the data is in Table 11-51 in Section 11.4.2.4.
A later publication by Thomas (1980) extended his earlier analysis. This more extensive
analysis was limited to lead estate residents. Subjects who did not consume the first drawn
water from the tap had significantly lower blood lead levels than those who did (10.4 pg/dl
difference). No gradient was noted in blood lead levels with increasing water consumption.
Furthermore, no gradient in blood lead levels was noted with total beverage consumption (tea
ingestion frequency).
11.4.2.3.4 Worth study. In Boston, Massachusetts, an investigation was made of water distri-
bution via lead pipes. In addition to the data on lead in water, account was taken of socio-
economic and demographic factors as well as other sources of lead in the environment (Worth et
al., 1981). Participants, 771 persons from 383 households, were classified into age groups of
less than 6, 6-20, and greater than 20 years of age for analysis. A clear association
between water lead and blood lead was apparent (Table 11-48). For children under 6 years of
age, 34.6 percent of those consuming water with lead above the U.S. standard of 50 ug/1 had a
blood lead value greater than or equal to 35 ug/dl, whereas only 17.4 percent of those con-
suming water within the standard had blood lead values of greater than or equal to 35 ug/dl.
Worth et al. (1981) have published an extensive regression analysis of these data. Blood
lead levels were found to be significantly related to age, education of head of household, sex,
and water lead exposure. Of the two types of water samples taken, standing grab sample and
running grab sample, the former was shown to be more closely related to blood lead levels than
the latter. Regression equations are given in Tables 11-51 and 11-52 in Section 11.4.2.4.
11.4.2.4 Summary of Dietary Lead Exposures, Including Water. It is difficult to obtain accu-
rate dose-response relationships between blood lead levels and lead levels in food or water.
Dietary intake must be estimated by duplicate diets or fecal lead determinations. Water lead
levels can be determined with some accuracy, but the varying amounts of water consumed by dif-
ferent individuals add to the uncertainty of the estimated relationships.
Studies relating blood lead levels to dietary lead intake are compared in Table 11-49.
Two studies had subjects with relatively high dietary lead intakes. In the Sherlock et al.
(1982) study, 10 of 31 subjects had lead intake levels greater than 300 ug/day. In the United
Kingdom Central Directorate study (1982), 12 of 110 subjects had levels greater than 300
ug/day. These concentrations are high enough that the slope is clearly lower in this range
than it is in the 0-100 ug/day range. The estimates of slopes for the cube root models may
be overestimates in the low range (0-100 ug/day) for the reasons discussed in section 11.4.
11-127
-------
4.0
3.0
3.
Q
Q
O
O
2.0
1.0
MAXIMUM WATER LEAD
LEVELS ON 'COPPER ESTATE
MEDIAN WATER LEAD
LEVELS ON 'LEAD' ESTATE
3«
1.0 2.0
FIRST FLUSH WATER LEAD mg liter
3.0
Figure 11 28. Relation of blood lead (adult female) to first flush water
lead in combined estates. (Numbers are coincidental points: 9 = 9 or
more.) Curve a, present data; curve b, data of Moore at af. (1979).
11-128
-------
TABLE 11-48. BLOOD LEAD LEVELS OF 771 PERSONS IN RELATION
TO LEAD CONTENT OF DRINKING WATER, BOSTON, MA
Persons consuming water (standing grab samples)
Blood lead
levels, ug/dl
<35
>35
Total
<50
No.
622
61
683
ug Pb/1
Percent
91
9
100
£50
No.
68
20
88
IJ^Pb/l
Percent
77.3
22.7
100.0
Total
690
81
771
X2 = 14.35; df = 1.
Source: Worth et al. (1981).
Conversely, the linear equation is probably an underestimate. The slope from the Ryu study
was estimated directly from changes in infants and is the best estimate available. The esti-
mates for adults are more accurately estimated from the experimental studies.
The experimental studies are summarized in Table 11-50. Most of the dietary intake sup-
plements were so high that many of the subjects had blood lead concentrations much in excess
of 30 ug/dl for a considerable part of the experiment. Blood lead levels thus may not com-
pletely reflect lead exposure, due to the previously noted nonlinearity of blood lead response
at high exposures. The slope estimates for adult dietary intake are about 0.02 ug/dl increase
in blood lead per ug/day intake, but consideration of blood lead kinetics may increase this
value greatly. Such values are a bit lower than those estimated from the adult population
studies extrapolated to typical dietary intakes in Table 11-49, about 0.05 ug/dl per ug/day.
The value for infants is much larger.
The studies relating first flush and running water lead levels to blood lead levels are
in Tables 11-51 and 11-52, respectively. Many of the authors chose to fit cube root models to
their data, although polynomial and logarithmic models were also used. Unfortunately, the
form of the model greatly influences the estimated contributions to blood lead levels from
relatively low water lead concentrations. As indicated in section 11.4, the models producing
high estimated contributions are the cube root models and the logarithmic models. All others
are polynomial models, either linear, quadratic, or cubic. The slopes of these models tend to
be relatively constant at the origin.
11-129
-------
TABLE 11-49. STUDIES RELATING BLOOD LEAD LEVELS (ug/dl) TO DIETARY INTAKES (ug/day)
Study
Sherlock et al.
(1982) study of
31 adult women
in Ayr
Sherlock et al.
i- (1982) study of
7* infants in Ayr
>- combined with U.K.
^ Central Directorate
Study
U.K. Central
Directorate
(19B2) Study
of Infants in
Glasgow
Model
Analysis Model 82 D.F.
Sherlock et al. PbB = -1.4 + 3.6 ^~PbD~ 0.52 2
(1982)
Sherlock et al. PbB = 2.5 + 5.0 ^TED - 2
(1982)
U.K. Central PbB = 17.1 + 0.056(PbD) 0.39 2
Di rectorate or 3
on Environmental PbB = 3.9 * 4.6 /TH) 0.43 2
Pollution
(1982)
Estimated
blood
lead at
0 H20 Pb
-1.4
2.5
17.1
3.9
Predicted blood lead
contribution (ug/dl) for
a given dietary intake
(ug/d)
100 200 300
16.7 21.1 24.1
23.2 29.2 33.5
5.6 11.2 16.8
21.4 26.9 30.8
Slope from 100 to 200
ug/d, ug/dl per ug/d
0.034
0.060
O.OB6
0.053
Ryu et al. (1983)
study of infants
EPA
PbB = A + 0.16PbD
16.0
32.0 48.0
0.16
-------
TABLE 11-50. STUDIES INVOLVING BLOOD LEAD LEVELS (ug/dl) AND EXPERIMENTAL DIETARY INTAKES
i
i*
OJ
Study
Stuik (1974)
Study I
Study II
Cools et al.
(1976)
Schlegel and
Kufner (1979)
Gross (1979)
analysis of
Kehoe's
experiments
* Exposure
Subjects
5 adult male students
5 adult female students
5 adult male students
5 adult female students
5 adult male students
6 adult female students
11 adult males
10 adult males
1 adult male
1 adult male
1 adult male
1 adult male
1 adult male
1 adult male
(ug/d) = Exposure (ug/kg/day)
** Corrected for decrease of 2.2 ug/dl in
*** Assumed
**** Assumed
***** Removed
mean life 40d. This increases
limited absorption of lead.
Exposure
20 ug/kg/day -
20 ug/kg/day -
Controls
20 ug/kg/day
30 ug/kg/day
Controls
21 d.
21 d.
- 21 d.
30 ug/kg/day ~7 days
Control s
50 ug/kg/day - 6 wk.
70 ug/kg/day -13 wk.
300 ug/day
1000 ug/day
2000 ug/day
3000 ug/day
x 70 kg for males,
control males.
slope estimate for
55 kg for
short-tern
Fora of lead
Lead acetate
Lead acetate
Placebo
Lead acetate
Lead acetate
Placebo
Lead acetate
Placebo
Lead nitrate
Lead nitrate
Lead acetate
Lead acetate
Lead acetate
Lead acetate
females. Slope = (Final
studies. Stuik Study I
Blood lead
Initial
20.6
12.7
20.6
17.3
16.1
-17.0
17.2
16.5
12.4
- Initial Blood
would be 0.042,
Fi^al
40.9
30.4
18.4
41.3
46.2
-17.0
26.2
-19.0
64.0
30.4
-1
+17
+33
+19
Lead)/Exposure
Slope,*
per ug
0.017**,
0.018**,
0.022
0.014
0.027***
ug/dl
/d
***
***
Q.014
0.004****
[0]
0.017
0.016
0.006*****
(ug/d).
0.044 respectively for males, females.
from exposure before equilibrium.
-------
TABLE 11-51. STUDIES RELATING BLOOD LEAD LEVELS (ug/dl) TO FIRST-FLUSH WATER LEAD (ug/1)
11
t~*
1
1 »
CO
re
Study
Worth et al. (1981) study of 524
subjects in greater Boston. Water
leads (standing water) ranged from
<13 to 1108 MS/I- Blood leads
ranged from 6 to 71.
Moore et al. (1979) study of 232
mothers at delivery In Glasgow.
17* of the water leads were over
300 ug/1.
Hubermont et al. (1978) study of
70 pregnant women in rural Belgium.
Water leads ranged from 0.2 to
1228.5 ug/1. Blood leads ranged
from 5.1 to 26.3 ug/dl.
U.K. Central Directorate (1982)
study of 128 mothers in greater
Glasgow. Water leads ranged from
under <10 to 1060 ug/1. Blood
leads ranged from 2 to 39 ug/dl.
U.K. Central Directorate (1982)
study of 126 infants (as above).
Blood leads ranged from 1 to 51
(jg/dl .
Thomas et al. (1979) study of 115
adult Welsh feules. Water leads
ranged from <10 to 2800 ug/dl.
Blood leads ranged from 5 to 65
ug/dl .
Moore (1977) study of 75 residents
of a Glasgow tenement
Pococfc et al. (1983) study of 7735
en aged 40-59 in Great Britain.
Water leads restricted to <100 ug/1.
Moore (1984) study of 568
mothers in Scotland.
Analysis
Worth et al. (1981)
EPA
Moore et al. (1979)
Hubermont et al.
(1978)
U.K. Cen. Dir. (1982)
U.K. Cen. Oir. (1982)
Lacey et al. (1985)
EPA
U.K. Cen. Oir. (1982)
J.K. Cen. Oir. (1982)
Lacey et al. (198S)
EPA
EPA
Moore (1977)
Pocock et al. (1983)
Moore (1984)
Model
In (PbB) = 2.729 PbW - 4.699 (PbW)2 *
2.116 (PbW)3 + other terms for age.
sex, education, dust (PbW is in imj/1)
In (PbB) = In (.041 PbW - .000219
(PbW)2 » other terms for age, sex.
education, dust)
PbB = 5.81 + 2.73 {PbW)V3
PbB = 9.62 + 0.756 £n (PbW)
PbB = 13.2 + 1.8 (PbW)1/3
PbB = 18.0 + 0.009 PbW
PbB = 14.0 + 0.062 PbW
In (PbB) = In (14.2 + 0.033
PbW - 0.000031 PbW2)
PbB = 9.4 + 2.4 (rbV)1/3
PbB = 17.1 + 0.018 PbW
PbB = 14.0 + .062 PbW
In (PbB) = In (14.2 * 0.033
PbW - 0.000031 PbW2)
In (PbB) = [14.9 * 0.041 PbW - 0.000012
(PbW)2]
PbB = 15.7 * 0.015 PbW
PbB = 14.48 + 0.062 PbW
PbB = 5.5 + 2.63 (PbW)'/3
R2 '
0.18
0.18
0.44
0.14
0.11
0.05
0.10
0.17
0.12
0.15
0.61
0.34
0.59
Model
D.F.
14
11
2
2
2
2
2
3
2
2
2
3
3
2
2
2
Estimated
blood
lead at
0 H20 ft
20.5
21.1
5.8
8.4*
13.2
18.0
14.0
14.2
9.4
17.1
14.0
12.0
14.9
15.7
14.5
S.5
Predicted blood lead
contribution (+jg/dl ) for
a given water lead (ug/1)
5
0.3
0.2
4.7
2.4
3.1
0.0
0.3
0.2
4.1
0.1
0.3
0.2
0.2
0.1
0.3
4.S
10 25
0.6 1.4
0.4 1.0
5.9 8.0
3.0 3.7
3.9 5.3
0.1 0.2
0.6 1.6
0.3 0.8
5.2 7.0
0.2 0.4
0.6 1.6
0.5 1.2
0.4 1.0
0.2 0.4
0.6 1.-6
5.7 7.7
50
2.7
2.1
10.1
4.2
£.6
0.4
3.1
1.6
e.e
0.9
3.1
2.4
2.0
0.6
3-1
9.7
'Minimum water lead of 0.2 ug/dl used instead of 0.
-------
TABLE 11-52. STUDIES RELATING BLOOD LEAD LEVELS (ug/dl) TO RUNNING WATER LEAD (|jg/l)
Study
Analysis
Worth et al. (1981) study of 524 sub- EPA
jects in greater Boston. Water leads
ranged from <13 to 208 ug/dl. Blood
leads ranged fro* 6 to 71.
i
i *
t
t '
OJ
GJ
Worth et a). (1981) study restricted
to 390 subjects aged 20 or older.
Worth et al . (1981) study restricted
to 249 feules ages 20 to 50.
U.K. Central Directorate (1982)
study of 128 Mothers in greater
Glasgow. Water leads ranged froa
under 20 to 720 ug/1. Blood
leads ranged fro 1 to 39
^g/dl.
U.K. Central Directorate (1982)
study of 126 infants (as above).
Blood leads ranged fro* 1 to 51
ug/dl.
Moore (1977) study of 75 residents
of a Glasgow tenement.
Sherlock et al. (1982) study of 114
adult woven. Blood leads ranged
<5 to >61 pg/dl . Kettle water leads
ranged from <10 to >2570 (ig/1.
Sherlock et al. (1984) follow-up
study.
U.S. EPA (1980)
EPA
EPA
U.S. EPA (1980)
EPA
EPA
U.K. Cen.Dir. (1982)
U.K. Cen.Dir. (1982)
EPA
U.K. Cen.Dir. (1982)
U.K. Cen.Oir. (1982)
EPA
Moore (1977)
Sherlock et al.
(1982)
EPA
Sherlock et al.
(1984)
Model
In (PbB) = (0.0425 PbW + other terms for
age, sex, education, and dust)
PbB = 14.33 + 2.541 (PbW)1^
In (PbB) = In (18.6 + 0.071 PbW)
In (PbB) = In (0.073 PbW + other teras
for sex, education, and dust)
PbB = 13.38 + 2.487 (PbW)*/3
In (PbB) = In (17.6 + 0.067 PbW)
In (PbB) = (0.067 PbW + other terns
for education and dust)
PbB = 12.8 + 1.8 (PbW)l/3
PbB = 18.1 + 0.014 PbW
in (PbB) = In (13.4 + 0.071 PbW
-0.000104 PbW2)
PbB = 7.6 + 2.3 (PbW)'/3
PbB = 16.7 + 0.033 PbW
In (PbB) = In (12.3 + 0.068 PbW
-0.000056 PbW2)
PbB = 16.6 + 0.02 PbW
PbB =4.7+2.78 (PbW)'/3
In (PbB) = In (11.5 +_0.,033 PbW
-0.00001 PbW2)
PbB = 5.6+2.62 (PbW)'/3
Model
R2 D.F.
0.153
0.023
0.028
0.153
0.030
0.032
0.091
0.12
0.06
0.16
0.22
0.12
0.20
0.27
0.56
0.55
0.65
10
2
2
7
2
2
6
2
2
3
2
2
3
2
2
3
2
Estimated
blood
lead at
0 HjO Pb
21.
14.
18.
18.
13.
17.
17.
12.
IB.
13.
7.
16.
12.
16.
4.
11.
5.
3
3
6
8
4
6
6
8
1
4
6
7
3
6
7
5
6
Predicted blood lead
contribution (ug/dl) for
a given water lead (M3/D
5 10 25 50
0.2
4.4
0.4
0.4
4.3
0.3
0.3
3.1
0.1
0.4
3.9
0.2
0.3
0.1
4.8
0.2
4.5
0.4
5.4
0.7
0.7
5.4
0.7
0.7
3.9
0.4
0.7
5.0
0.3
0.7
0.2
6.0
O.J
5.6
1.1
7.4
1.8
1.8
7.3
1.7
1.7
5.3
0.4
0.7
6.7
0.8
1.7
0.5
8.1
0.8
7.7
2.1
9.4
3.6
3.7
9.2
3.4
3.4
6.6
0.7
3.3
8.5
1.6
3.3
1.0
10.2
1.6
9.7
-------
The problem of determining the most appropriate model(s) at low water lead levels
(0-25 ug/1) is extremely difficult. Most data sets estimate a relationship that is primarily
based on water lead levels of 50-2000 M9/1. and the problem becomes essentially a low-dose
extrapolation problem. The only study which estimates the relationship based primarily on
lower water lead levels (<100 |jg/1) is the Pocock et al. (1983) study. The data from this
study, as well as the authors themselves, suggest that in this lower range of water lead
levels, the relationship is linear. Furthermore, the contributions to blood lead levels esti-
mated from this study are quite consistent with the polynomial models from the other first-
flush water lead studies, such as Worth et al. (1981), United Kingdom Central Directorate on
Environmental Pollution (1982), and Thomas et al. (1979). For these reasons the Pocock et al.
(1983) slope of 0.06 is our best estimate for first-flush water lead studies. The slopes for
running water lead studies are about 1.5 to 2.0 times as large. The possibility does exist,
however, that the higher initial slopes from the cube-root and logarithmic models are correct.
11.4.3 Studies Relating Lead in Soil and Dust to Blood Lead
The relationship of exposure to lead contained in soil and house dust, and the amount of
lead absorbed by humans, particularly children, has been the subject of scientific investiga-
tion for some time (Duggan and Williams, 1977; Barltrop, 1975; Creason et al., 1975; Barltrop
et al., 1974; Roberts et al., 1974; Sayre et al., 1974; Ter Haar and Aronow, 1974; Fairey and
Gray, 1970). Duggan and Williams (1977) published an assessment of the risk of increased
blood lead resulting from the ingestion of lead in dust. Some of these studies have been con-
cerned with the effects of such exposures (Barltrop, 1975; Creason et al. , 1975; Barltrop et
al., 1974; Roberts et al., 1974; Fairey and Gray, 1970); others have concentrated on the means
by which the lead in soil and dust becomes available to the body (Sayre et al., 1974; Ter Haar
and Aronow, 1974; Brunekreef et al., 1983).
11.4.3.1 Omaha, Nebraska Studies. The Omaha studies were described in Section 11.4.1.7.
Soil samples were 2-inch cores halfway between the building and the lot line. Household dust
was collected from vacuum cleaner bags. The following analysis was provided courtesy of Dr.
Angle. The model is also described in Section 11.4.1.8, and provided the coefficients and
standard errors shown in Table 11-53.
11.4.3.2 Stark Study. Stark et al. (1982) used a large-scale lead screening program in New
Haven, Connecticut, during 1974-77 as a means of identifying study subjects. The screening
program had blood lead levels on 8289 children, ages 1-72 months, that represented about 80
percent of the total city population in that age group. From this initial population, a much
smaller subset of children was identified for a detailed environmental exposure study. Using
the classifying criteria of residential stability and repeatable blood lead levels (multiple
11-134
-------
TABLE 11-53. COEFFICIENTS AND STANDARD ERRORS FOR OMAHA STUDY MODEL
Asymptotic
Factor Coefficient Standard Error
Intercept (ug/dl) 15.67 0.398
Air lead (ug/m3) 1.92 0.600
Soil lead (mg/g) 6.80 0.966
House dust (mg/g) 7.18 0.900
Multiple R2 = 0.198
Sample size = 1075
Residual standard deviation = 0.300 (geometric standard deviation = 1.35)
measurements fell into one of three previously defined blood lead concentration categories), a
potential study population of 784 was identified. Change of residence following identifica-
tion and refusal to let sanitarians make inspections resulted in 407 children being dropped;
the final study population contained 377 children.
With the exception of dietary lead intake, each child's potential total external lead
exposure was assessed. Information was obtained on lead in air, house dust, interior and
exterior paint, and soil near and far from the home. A two percent sample of homes with
children having elevated lead levels had tap water lead levels assessed. No water lead levels
above the public health service standard of 50 ug/1 were found. Socioeconomic variables were
also obtained.
For all children in the study, micro blood samples were taken and analyzed for lead by
AAS with Delves cup attachment. Blood lead values were found to follow a lognormal distri-
bution. Study results were presented using geometric means and geometric standard deviation.
Among the various environmental measurements a number of significant correlation coefficients
were observed. However, air lead levels were independent of most of the other environmental
variables. Environmental levels of lead did not directly follow socioeconomic status. Most
of the children, however, were in the lower socioeconomic groups.
Multiple regression analyses were performed by Stark et al. (1982) and by EPA*, using all
926 blood lead measurements. Stark and coworkers derived a log-log model with R2 = 0.11, and
no significant effects of race or age were found. EPA fitted a linear exposure model in loga-
rithmic form with results shown in Table 11-54. Significant differences among age groups were
*NOTE: The term EPA analyses refers to calculations done at EPA. A brief discussion of the
methods used is contained in Appendix 11-B; more detailed information is available at EPA
upon request.
11-135
-------
TABLE 11-54. MULTIPLE REGRESSION MODELS FOR BLOOD LEAD
OF CHILDREN IN NEW HAVEN, CONNECTICUT, SEPTEMBER 1974 - FEBRUARY 1977
Regression Coefficients and Standard Errors
Covariate
Summer - winter
Dust, ug/g
Housekeeping quality
Soil near house, ug/g
Soil at curb, ug/g
Paint, child's bedroom
Paint outside house
Paint quality
Race = Black
Residual standard deviations
Multiple R2
Sample size (blood samples)
Ages
0-1 yr
6.33 ± 2.11*
0.00402 ± 0.00170*
4.38 ± 2.02*
0.00223 ± 0.00091*
0.00230 ± 0.00190
0.0189 ± 0.0162
-0.0023 ± 0.0138
0.89 ± 1.71
2.16 ± 2.05
0.1299
0.289
153
Ages
2-3 yr
3.28 ± 1.30*
0.00182 ± 0.00066*
1.75 ± 1.17
-0.00016 ± 0.00042
0.00203 ± 0.00082*
0.0312 ± 0.0066*
0.0200 ± 0.0069*
3.38 ± 0.96*
0.07 ± 1.09
0.0646
0.300
334
Ages
4-7 yr
2.43 ± 1.38*
0.00022 ± 0.00077
-1.61 ± 1.12
0.00060 ± 0.00041
0.00073 ± 0.00079
0.0110 ± 0.0064*
0.0172 ± 0.0067*
4.14 ± 1.15*
5.81 ± 1.00*
0.1052
0.143
439
"Significant positive coefficient, one-tailed p <0.05.
noted, with considerably improved predictability (R2 = 0.29, 0.30, 0.14 for ages 0-1, 2-3, and
4-7). Sex was not a significant variable, but Race = Black was significant at ages 4-7. Air
lead did not significantly improve the fit of the model when other covariates were available,
particularly dust, soil, paint, and housekeeping quality. However, the range of air lead
levels was small (0.7-1.3 ug/m3) and some of the inhalation effect may have been confounded
with dust and soil ingestion. Seasonal variations were important at all ages.
EPA analyses of data from children in New Haven (Stark et al., 1982) found substantial
evidence for dust and soil lead contributions to blood lead, as well as evidence for increased
blood lead due to decreased household cleanliness. These factors are somewhat correlated with
each other, but the separate roles of increased concentration and cleanliness could be distin-
guished. Overall dust, soil, and paint lead levels were not presented in the published
papers, but data presented by year of housing construction indicate that meaningful lead expo-
sures were present. Geometric mean dust lead levels varied from 239 ppm for houses built in
11-136
-------
1960-1969 to 756 ppm for those built in 1910-1919. Soil lead levels varied from 131 ppm to
1273 ppm for 1970-1977 and 1920-1929, respectively.
11.4.3.3 The Silver Valley/Kellogg Idaho Study. The Silver Valley/Kellogg Idaho study was
discussed in section 11.4.1.6. Yankel et al. (1977) showed that lead in both soil and dust
was independently related to blood lead levels. In their opinion, 1000 ug/g soil lead ex-
posure was cause for concern. Walter et al. (1980) showed that children aged 3 through 6
showed the strongest relationship between soil lead and blood lead, but 2-year-olds and 7-year-
olds also had a significant relationship (Table 11-29). The slope of 1.1 for soil lead (1000
ug/g) to blood lead ((jg/dl) represents an average relationship for all ages.
The Silver Valley-Kellogg Idaho study also gave some information on house dust lead, al-
though this data was less complete than the other information. Regression coefficients for
these data are in Tables 11-29 and 11-30. In spite of the correlation of these predictors,
significant regression coefficients could be estimated separately for these effects.
11.4.3.4 Blood Lead Levels of Dutch City Children. Brunekreef et al. (1983) reported on a
very extensive study on blood lead and environmental variables in native Dutch children 4-6
years old. Three hundred seventy-one children participated in the blood lead survey and 195
children in the environmental study as well. The environmental evaluation was carried out in
April-June 1981 in the cities of Rotterdam, the Hague, and Zoeterraeer. Blood was sampled by
venipuncture. The environmental variables included:
In the home of each child:
lead in drinking water (one first-draw sample)
lead deposition indoors, using 2 greased deposition plates per home and an averaging
time of 4 weeks
lead in floor dust, using a special vacuum cleaner to take 2 duplicate samples 4 weeks
apart
lead in 0-5 cm top soil in gardens, if present
In living area:
lead deposition outdoors on 5-10 spots per area with an averaging time of 4 weeks
lead in street dust using the vacuum cleaner method, taking 30-40 duplicate samples
per area on 2 occasions 4 weeks apart
In the classroom/school:
lead in drinking water (one running sample)
lead deposition indoors, applying 2 plates in 2 classrooms per school with an averag-
ing time of 4 weeks
lead in floor dust, taking 2 duplicate samples in 2 different classrooms per school,
4 weeks apart
lead in playground dust, using the vacuum cleaner method to take 4 duplicate samples
on two occasions 4 weeks apart
lead in 0-5 cm top soil in playground
lead on dominant hand of child, after playing outdoors for at least 30 minutes in
school playground on a dry day.
11-137
-------
Resulting blood lead levels and environmental lead measurements are shown in Tables 11-55
to 11-58.
Multiple regression analyses were done by Brunekreef et al. in logarithmic rather than
linear form. The equation is as follows.
In PbB = 1.882 + 0.163C In (lead deposition outdoors) - 0.003 (year of construc-
tion - 1900) + 0.135b (hand dirtiness) - 0.380d (milk consumption) + 0.116b
(presence of pets) + 0.1063 (mouthing behavior) - 0.069 (number of rooms)
ap <0.01. bp <0.005. Cp <0.001. dp <0.0001. (11-19)
Multiple regression analysis for combined inner city and suburban populations give the
following: n = 193, R2 = 0.519, F-total = 28.5.
Lead deposition outdoors was an important factor, but only in the combined sample, so
confounding cannot be ruled out. This appears, however, to be the single most important
environmental source, particularly in conjunction with hand dirtiness and with mouthing
behavior. Further analyses of these data are proposed. The difference of about 2 ug/dl
between city and suburban children (adjusted for all other covariates) can hardly be attri-
buted to direct inhalation of ambient air lead which differs slightly from city to suburb
(0.12-0.13 ug/m3), and hence must be attributed to other pathways. The large coefficient for
milk reflects the known importance of calcium in lead metabolism and is also related to mouth-
ing behaviors, including pica. The presence of pets probably increases the exposure to dirt.
This study thus corroborates the importance of various non-inhalation pathways for lead in
children, particularly the dust-hand-mouth pathway.
Dr. Brunekreef has (personal communication, February 8, 1984) fitted his data on Dutch
children to a linear model in logarithmic form, as the Environmental Protection Agency has
done elsewhere in the present document. The regression coefficients are all statistically
significant, and variables are as in his 1983 paper. The logarithmic linear model had vari-
ance s2 = 0.06272 and R2 = 0.521; it thus provided an (insignificantly) better description of
the data than the original log-log model.
11.4.3.5 Charney Study. Charney et al. (1980) conducted a case control study of children
ages 1.5-6 with highly elevated and non-elevated blood lead levels. Cases and controls were
initially identified from the lead screening programs of two Rochester, New York, health
facilities. Cases were defined as children who had at least two blood lead determinations
between 40 and 70 ug/dl and FEP values greater than 59 ug/dl during a 4-month period. Con-
trols were children who had blood lead levels equal to or less than 29 ug/dl and FEP equal to
11-138
-------
TABLE 11-55. AIR LEAD LEVELS IN THE ROTTERDAM AREA (BRUNEKREEF ET AL., 1983)
Geometric mean air lead level in pg/m3
yg/
-Ji
Sampling location January-March, 1981 April-June, 1981
Rotterdam (center) 0.27 0.22
Maassluis (upwind suburb) 0.14 0.10
TABLE 11-56. BLOOD LEAD LEVELS IN pg/100 ml FOR CHILDREN WHO
PARTICIPATED IN BLOOD SURVEY AND ENVIRONMENTAL SURVEY
City
Rotterdam (center)
Rotterdam (suburb)
The Hague
Zoetermeer
Number
54
72
16
53
Geometric
mean
13.1
8.2
11.5
7.9
Percenti le
Range
7-31
5-15
7-21
4-15
50
13
8
11
8
90
19
11
19
11
98
23
14
21
14
Difference between city and suburb significant (t-test on arithmetic means; p <0.001).
TABLE 11-57. SCHOOL VARIABLES (ARITHMETIC MEANS) FOR MEASURED LEAD CONCENTRATIONS
City
Rotterdam3
Rotterdam
Zoetermeer
In
drinking
water,
pg/1
6
1
1
Deposition
indoors,
|jg/m2/d
11.74
4.29
4.59
On
floors,
pg/m2
100
29
40
On
schoolyard,
pg/m2
1120
364
337
In sandy
playground,
mg/kg
6
5
6
alnner city.
Suburb.
11-139
-------
TABLE 11-58. RESULTS OF LEAD MEASUREMENTS REPORTED BY BRUNEKREEF ET AL. (1983)
City
Lead deposition
Rotterdam.
Rotterdam
The Hague .
Zoetermeer
Lead on streets
Rotterdam.
Rotterdam
The Hague .
Zoetermeer
Lead in garden
Rotterdam^
Rotterdam
The Hague .
Zoetermeer
Lead deposition
Rotterdam.
Rotterdam
The Hague b
Zoetermeer
Lead on floors
Rotterdam.
Rotterdam
The Hague ,
Zoetermeer
Concentration
2
outdoors (arithmetic mean, ug/m
643
220
369
125
2
(geometric mean, ug/m )
532
318
428
126
soil (geometric mean, mg/kg)
336
43
278
21
2
Range
/d)
394-957
144-315
317-439
73-278
168-2304
113-1155
81-1339
46-497
6-184
35-527
3-75
indoors (geometric mean, ug/m /d)
2.86 0.10-20.86
0.99
4.32
1.51
2
(geometric mean, ug/m )
81
30
58
32
Lead in drinking water (geometric mean, ug/1)
Rotterdam. 20
Rotterdam
The Hague .
Zoetermeer
2
21
1
0.10-8.40
1.95-27.05
0.48-4.40
5-740
1-410
22-166
3-201
1-126
1-50
1-85
1-4
n
9
6
5
10
37
36
10
21
1
56
6
33
48
67
13
49
43
62
11
50
46
60
16
53
t-test
p <0.001
p <0.001
p <0.001
p <0.001
p <0.005
p <0.005
p <0.001
p <0.001
p <0.001
p <0.001
p <0.001
p <0.001
p <0.001
p <0.001
p <0.001
p <0.001
p <0.025
p <0.025
p <0.001
p <0.001
p <0.001
p <0.001
Lead on hands (geometric mean, ug/hand)
Rotterdam.
Rotterdam .
Zoetermeer
alnner city.
Suburb.
12
5
4
1-96
1-21
1-18
44
65
37
p <0.001
p <0.001
p <0.001
11-140
-------
or less than 59 ug/dl. High-level children were selected first and low-level children were
group-matched based on age, area of residence, and social class of the family. Home visits
were made to gain permission as well as to gather questionnaire and environmental data. Lead
analyses of the various environmental samples were done at several different laboratories. No
specification was provided regarding the analytical procedures followed.
The matching procedure worked well for age, and mother's educational level and employment
status. There were more blacks in the high lead group as well as more Medicaid support.
These factors were then controlled in the analysis; no differences were noted between the high
and low blood lead groups regarding residence on high traffic density streets (>10,000 vehi-
cles/ day) or census tract of residence.
The two groups differed regarding mean house dust lead levels (1265 ug/sample for high
and 123 pg/sample for low). Median values also differed, 149 versus 55 ug/sample. One-third
of the children in the low blood lead group had house dust lead samples with more lead than
those found in any middle class home previously investigated.
There were considerably greater quantities of lead on the hands of the high blood lead
group compared with the low lead group (mean values were 49 and 21 pg/sample, respectively).
Hand and house dust lead levels were correlated (r = 0.25) but the relationship was not
linear. At the low end of the house dust lead values, hand dust was always low but the con-
verse was not true: not every child exposed to high house dust lead had high hand dust
levels.
In addition to hand and house dust lead, other factors differentiated the high and low
blood lead groups. Although both groups had access to peeling paint in their homes (^2/3),
paint lead concentrations exceeding 1 percent were found more frequently in the high as oppo-
sed to the low group. Pica (as defined in Chapter Seven) was more prevalent in the high lead
group as opposed to the low lead group.
Since the data suggested a muHifactorial contribution of lead, a multiple regression
analysis was undertaken. The results suggest that hand lead level, house dust lead level,
lead in outside soil, and history of pica are very important in explaining the observed vari-
ance in blood lead levels.
11.4.3.6 Charleston Studies. In one of the earliest investigations regarding soil lead expo-
sures, Fairey and Gray (1970) conducted a retrospective study of lead poisoning cases in
Charleston, South Carolina. Two-inch core soil samples were collected from 170 randomly
selected sites in the city and were compared with soil samples taken from homes where 37 cases
of lead poisoning had occurred. The soil lead values obtained ranged from 1 to 12,000 ug/g,
with 75 percent of the samples containing less than 500 ug/g. A significant relationship
between soil lead levels and lead poisoning cases was established; 500 pg/g was used as the
11-141
-------
outpoint in the chi-square contingency analysis. Fairey and Gray were the first to examine
this complex problem and, although their data support the soil lead hypothesis, the relation-
ship between soil lead and blood lead levels could not be quantified. Furthermore, because no
other source of lead was measured, any positive association could have been confounded by
additional sources of lead, such as paint or air.
A later study by Galke et al. (1975), in Charleston, used a house-to-house survey to re-
cruit 194 black preschool children. Soil, paint, and air lead exposures, as measured by traf-
fic density, were established for each child. When the population was divided into two groups
based on the median soil lead value (585 ug/g), a 5 ug/dl difference in blood lead levels was
obtained. Soil lead exposure for this population ranged from 9 to 7890 ug/g. Vehicle traffic
patterns were defined by area of recruitment as being high or low. A multiple regression
analysis of the data showed that vehicle traffic patterns, lead level in exterior siding
paint, and lead in soil were all independently and significantly related to blood lead levels.
Using the model described in Appendix 11B, the following coefficients and standard errors were
obtained as shown in Table 11-59.
TABLE 11-59. COEFFICIENTS AND STANDARD ERRORS FROM MODEL OF CHARLESTON STUDY
Factor
Intercept ()jg/dl)
Pica (1 = eater, 0 = otherwise)
Traffic pattern (1 = high, 0 = low)
Siding paint (mg/cm2)
Door paint (mg/cm2)
Soil lead (mg/g)
Multiple R2 = 0.386
Residual standard deviation = 0.2148 (geometric
Coefficient
25.92
7.23
7.11
0.33
0.18
1.46
standard deviation = 1.24)
Asymptotic
standard error
1.61
1.60
1.48
0.11
0.12
0.59
11.4.3.7 Barltrop Studies. Barltrop et al. (1974) described two studies in England-investi-
gating the soil lead to blood lead relationship. In the first study, children aged 2 and 3
and their mothers from two towns chosen for their soil lead content had their blood lead
levels determined from a capillary sample. Hair samples were also collected and analyzed for
lead. Lead content of the suspended particulate matter and soil was measured. Soil samples
for each home were a composite of several 2-inch core samples taken from the yard of each
home. Chemical analysis of the lead content of soil in the two towns showed a 2- to 3-fold
difference, with the values in the control town about 200-300 (jg/g compared with about 700-
1000 ug/g in the exposed town. A difference was also noted in the mean air lead content of
11-142
-------
the two towns, 0.60 compared with 0.29 ug/m3. Although this difference existed, both
air lead values were thought low enough not to affect the blood level values differentially.
Mean surface soil lead concentrations for the two communities were statistically different,
the means for the high and low community being 909 and 398 H9/9. respectively. Despite this
difference, no statistically significant differences in maternal blood lead levels or chil-
dren's blood or hair lead levels were noted. Further statistical analysis of the data, using
correlational analysis on either raw or log-transformed blood lead data, likewise failed to
show a significant relationship of soil lead with either blood lead or hair lead.
The second study was reported in both preliminary and final form (Barltrop et al. , 1974;
Barltrop, 1975). In the more detailed report (Barltrop, 1975), children's homes were clas-
sified by their soil lead content into three groups: less than 1,000; 1,000 - 10,000; and
greater than 10,000 pg/g. As shown in Table 11-60, children's mean blood lead levels increased
correspondingly from 20.7 to 29.0 M9/dl. Mean soil lead levels for the low and high soil
exposure groups were 420 and 13,969 ug/g, respectively. Mothers' blood levels, however, did
not reflect this trend; nor were the children's fecal lead levels different across the soil
exposure areas.
TABLE 11-60. MEAN BLOOD AND SOIL LEAD CONCENTRATIONS IN ENGLISH STUDY
Category
of soil lead
d-jg/g)
<1000
1000-10000
> 10000
Sample
size
29
43
10
Children's
blood lead
(ng/dl)
20.7
23.8
29.0
Soil lead
((jg/g)
420
3390
13969
Source: Barltrop, 1975.
An analysis of the data in Table 11-60 gives the following model
blood lead (ug/dl) = 0.64 soil lead (1000 |jg/9) + 20-98 (11-20)
No confidence intervals were calculated since the calculations were based on means.
11.4.3.8 The British Columbia Studies. Neri et al. (1978) studied blood lead levels in
children living in Trail, British Columbia. Capillary blood samples were collected and
analyzed for lead by anodic stripping voltammetry. Duplicate samples were analyzed and the
11-143
-------
results were discarded whenever the values differed by more than 8 ug/dl. This procedure
probably helped control to some degree the commonly encountered positive bias in blood lead
levels observed when capillary samples are used. An episode of poisoning of horses earlier
had been traced to ingestion of lead. Environmental monitoring at that time did not suggest
that a human health risk, existed. However, it was later thought wise to conduct a study of
lead absorption in the area.
Trail had been the site of a smelter since the turn of the century. The smelter had
undergone numerous changes for reasons of both health and productivity. At the time of the
blood lead study, the smelter was emitting 300 pounds of lead daily, with ambient air lead
levels at about 2 ug/m3 in 1975. Nelson, BC was chosen as the control city. The cities are
reasonably close (-30 miles distant), similar in population, and served by the same water
basin. The average air lead level in Nelson during the study was 0.5 ug/m3.
Initial planning called for the sampling of 200 children in each of three age groups (1"3
years, 1st grade and 9th grade) from each of the two sites. A strike at the smelter at the
onset of the study caused parts of the Trail population to move. Hence, the recruited sample
deviated from the planned one. School children were sampled in May, 1975 at their schools
while the 1- to 3-year olds were sampled in September, 1975 at a clinic or home. This delayed
sampling was intentional to allow those children to be exposed to the soil and dust for the
entire summer. Blood and hair samples were collected from each child.
The children in the younger age groups living in Trail had higher blood lead levels than
those living in Nelson. An examination of the frequency distributions of the blood lead
levels showed that the entire frequency of the distribution shifted between the residents of
the two cities. Interestingly, there was no difference in the ninth grade children.
Table 11-61 displays the results of the soil lead levels along with the blood lead levels
obtained in the earlier study. Blood lead levels were higher for 1- to 3-year olds and first
graders in the two nearest-to-smelter categories than in the far-from-smelter category.
Again, no difference was noted for the ninth graders.
An EPA analysis of the Neri et al. (1978) data gives the following models for children I"
to 3-years old
Blood lead (ug/dl) = 0.0076 soil lead (ug/g) + 15.43, and (11-21)
Blood lead (ug/dl) = C.C046 soil lead (ug/g) + 16.37 (11-22)
for children in grade one. No confidence intervals were calculated since the analysis was
based on means.
11-144
-------
TABLE 11-61. LEAD CONCENTRATION OF SURFACE SOIL AND CHILDREN'S
BLOOD BY RESIDENTIAL AREA OF TRAIL, BRITISH COLUMBIA
Residential
area(s)
1 and 2
5
9
3, 4, and 8
6 and 7
Total
Mean
soil lead
concentration, ug/g,
± standard error
(and no. of samples)
225 ± 39 (26)
777 ± 239 (12)
570 ± 143 (11)
1674 ± 183 (53)
1800 ± 212 (51)
1320 ± 212 (153)
Blood
M9/dl
error
1- to 3-
year olds
17.2 ± 1.1 (27)
19.7 ± 1.5 (11)
20.7 ± 1.6 (19)
27.7 ± 1.8 (14)
30.2 ± 3.0 (16)
22.4 ± 1.0 (87)
lead concentration,
, mean ± standard
(and no. of children)
Grade one
children
18.0 ±1.9 (18)
18.7 ±2.3 (12)
19.7 ± 1.0 (16)
23.8 ± 1.3 (31)
25.6 ±1.5 (26)
21.9 ± 0.7 (103)
Source: Schmitt et al., 1979.
11.4.3.9 The Baltimore Charney Study: A Controlled Trial of Household Dust Lead Reduction.
Charney et al. (1983) selected children from the Lead Poisoning Clinic of the John F. Kennedy
Institute in Baltimore. The children were all 15-72 months old at the time of enrollment and
had at least two venous blood lead levels between 30 and 49 |jg/dl and FEP < 655 ug/dl. The
children were also required to have had the same place of residence for at least the preceding
six months. Their houses had to have been deleaded in accordance with standard procedures
used by the Baltimore City Health Department. Experimental control subjects were recruited on
the basis of attendance at routine periodic blood lead monitoring. Alternative identification
numbers were used for allocation to experimental and control groups. Home visits were made
for children in the experimental group and a 930 cm2 area of the floor or windowsill was wiped
with an alcohol-treated cloth towel and the dust lead content analyzed. A "dust control team"
then visited each home twice monthly and wet-mopped all surfaces with >100 ug Pb per 930 cm2.
The child's caretaker was advised to wet-mop these surfaces and other "hot spots" more fre-
quently, to wash the child's hands before meals and at bedtime, and to restrict access to high-
lead areas.
Both the 14 experimental subjects receiving the above treatment and the 35 control sub-
jects started the study with about the same moderately elevated blood lead levels, 38.6 ±
5.2 ug/dl at the start of the experiment. These levels had remained almost stationary for six
months before the experiment, increasing only 1 ug/dl on average. After a year of dust con-
trol , the experimental subjects had reduced their PbB levels by 6.9 ug/dl, whereas the control
11-145
-------
subjects had reduced their PbB levels insignificantly (0.7 (jg/dl). Five of the control sub-
jects actually had increased PbB by 6-12 ug/dl, and one by 20 |jg/dl. None of the dust-
controlled subjects had any PbB increase, and most showed a decrease of at least 6 ug/dl.
Four experimental subjects had PbB < 30 |jg/dl by the end of. the experiment.
Dust lead levels in experimentally cleaned homes returned to nearly the previous high
values within two weeks. There was no significant relation between reduction of leaded dust,
initial level of leaded dust, and the reduction in a child's blood lead level. This lack of
apparent correlation may have been due to failure to control or monitor hand washing, finger-
sucking and mouthing behavior, access to "hot spots," and time spent in the home. Further-
more, attempts at dust control may have been more successful in some of the control homes than
in others, resulting in blood lead reduction in at least some individual cases. Since advice
on dust control was offered to caretakers of lead-burdened children visiting the Clinic, it
may be presumed that some measure of dust control would have occurred in any event. Dust lead
values in the experimental homes were high compared to homes in other areas (13/14 had sites
>100 pg/930 cm2). While many potentially important factors were not completely controlled
during the trials, the importance of dust ingestion is evident. This study also points out
the difficulties in quantifying the dust-hand-mouth pathway using familiar measures of house-
hold dust lead and concentration. Since the reduction in blood lead levels cannot be plausi-
bly attributed to factors other than household dust control (e.g., relocation of residence or
change in diet), the experimental evidence for the importance of household dust in elevation
of blood lead levels in U.S. urban children is very strong.
11.4.3.10 Gallacher Study. A report from England (Gallacher et al. , 1984) provides addi-
tional informative data on the importance of dust to blood lead levels. They were interested
in the effect of pica on blood lead levels. Mothers and children aged 1-3 years were recruit-
ed from 4 areas of Wales chosen for presumed lead exposure: 1) roadside dwellings; 2) cul
de sac dwellings; 3) an old mining area; and 4) a control area. Comprehensive environmental
sampling accompanied the study of blood lead levels. Indoor air samples, soil from play
areas, pavement dust, house dust, and tap water samples were collected and analyzed for lead
content. Capillary blood samples were collected from the children, while venous samples were
collected from the mothers. Blood samples were analyzed for lead by atomic absorption spec-
trophotometry. The accuracy of the capillary sampling was checked; the authors concluded that
contamination was not a problem but that the values of the capillary samples were 37 percent
higher than venous samples. They attributed the difference as "probably owing to haemoconcen-
tration of capillary blood."
Results from the environmental sampling indicated that for many of the environmental
media, lead exposures were reasonably constant over a several-month period. The authors state
11-146
-------
that, "Coefficient of variation..., based on duplicate pairs and after logarithmic transfor-
mation, was 9 percent for pavement dwellings (22 dwellings) and 10 percent for housedust
(25 dwellings). The coefficient of variation of child hand lead using the 'wet wipe1 tech-
nique was 19 percent (based on 17 children)." The coefficient of variation of the blood lead
sample of venous blood was around 7 percent.
In both children and mothers, the mining area differed the most from the control area.
The excess of lead in the blood of children was 30 percent for the mining area; in mothers the
excess was about 50 percent.
Pica as determined by questionnaire showed no consistent association with any area or all
areas combined. On the other hand, the analysis of the wet wipe study provided interesting
results. Within the roadside dwellings, the cul de sacs, and the control areas, mean lead
levels of wet wipe samples were remarkably similar for mothers' hands, children's hands, and
kitchen surfaces. But the mining area had a 40 percent excess for mothers' hands, 45 percent
for children's hands, and 35 percent for kitchen surfaces, compared to the control area.
However, the only difference which was statistically significant was for the children.
Correlation analysis was performed between blood lead concentrations and hand lead con-
centrations. In the mining area, which was the most contaminated area, the correlation co-
efficient was 0.38, which was statistically significantly different from zero. In the non-
contaminated areas, a statistically significant relationship was found between blood lead and
kitchen surface. No statistically significant relations were seen for the mothers. Thus
these data give additional support to the notion of normal hand-to-mouth activity being a
pathway by which lead in dust can get into the blood of children.
11.4.3.11 Other Studies of Soil and Dusts. Rabinowitz et al. (1985c) report in a study dis-
cussed in Section 11.3.5.4 that lead levels in indoor dust and outdoor soil were strongly pre-
dictive of blood lead levels. Their sample consisted of Boston urban and suburban infants
followed from birth to 2 years of age whose mothers had a mean age of 29 years and 15 years
mean schooling.
Lepow et al. (1975) studied the lead content of air, house dust, and dirt, as well as the
lead content of dirt on hands, food and water, to determine the cause of chronically elevated
blood lead levels in 10 children 2 to 6 years old in Hartford, Connecticut. Lead-based paints
had been eliminated as a significant source of lead for these children. Ambient air lead con-
centrations varied from 1.7 to 7.0 ug/m3. The mean lead concentration in dirt was 1,200 ug/g
and in dust, 11,000 ug/g. The mean concentration of lead in dirt on children's hands was
2,400 ug/g- The mean weight of samples of dirt from hands was 11 mg', which represented only a
small fraction of the total dirt on hands. Observation of the mouthing behavior in these
young children led to the conclusion that the hands-in-mouth exposure route was the principal
cause of excessive lead accumulation.
11-147
-------
Several studies have investigated the mechanism by which lead from soil and dust gets
into the body (Sayre et al., 1974; Ter Haar and Aronow, 1974). Sayre et al. (1974) in
Rochester, New York, demonstrated the feasibility of house dust as a source of lead for chil-
dren. Two groups of houses, one inner city and the other suburban, were chosen for the study.
Lead-free sanitary paper towels were used to collect dust samples from house surfaces and the
hands of children (Vostal et al., 1974). The medians for the hand and household samples were
used as the outpoints in the chi-square contingency analysis. A statistically significant
difference between the urban and suburban homes for dust levels was noted, as was a relation-
ship between household dust levels and hand dust levels (Lepow et al., 1975).
Ter Haar and Aronow (1974) investigated lead absorption in children that can be at-
tributed to ingestion of dust and dirt. They reasoned that because the proportion of the
naturally occurring isotope of 210Pb varies for paint chips, airborne particulates, fallout
dust, house dust, yard dirt, and street dirt, it would be possible to identify the sources of
ingested lead. They collected 24-hour excreta from eight hospitalized children on the first
day of hospitalization. These children, 1 to 3 years old, were suspected of having elevated
body burdens of lead, and one criterion for the suspicion was a history of pica. Ten children
of the same age level, who lived in good housing in Detroit and the suburbs, were selected as
controls and 24-hour excreta were collected from them. The excreta were dried and stable lead
as well as 210Pb content determined. For seven hospitalized children, the stable lead mean
value was 22.43 ug/g dry excreta, and the eighth child had a value of 1640 ug/g. The con-
trols' mean for stable lead was 4.1 ug/g dry excreta. However, the respective means for 21°Pb
expressed as pCi/g dry matter were 0.044 and 0.040. The authors concluded that because there
is no significant difference between these means for 210Pb, the hypothesis that young children
with pica eat dust is not supported. The authors further concluded that children with
evidence of high lead intake did not have dust and air suspended particulate as the sources of
their lead. It is clear that air suspended particulate did not account for the lead levels in
the hospitalized children. However, the 210Pb concentrations in dust and feces were similar
for all children, making it difficult to estimate the dust contribution.
Heyworth et al. (1981) studied a population of children exposed to lead in mine tailings.
These tailings were used in foundations and playgrounds, and had a lead content ranging from
10,000 to 15,000 ug/g- In December, 1979, venous blood samples and hair were collected from
181 of 346 children attending two schools in Western Australia. One of the schools was a pri-
mary school; the other was a combined primary and secondary school. Parents completed ques-
tionnaires covering background information as well as information regarding the children's
exposure to the tailings. Blood lead levels were determined by the AAS method of Farrely and
Pybos. Good quality control measures were undertaken for the study, especially for the blood
11-148
-------
lead levels. Blood lead levels were higher in boys versus girls (mean values were 14.0 and
10.4 ug/dl, respectively). This difference was statistically significant. Five percent of
the children (n = 9) had blood lead levels greater than 25 Mg/dl; five of these children had
blood lead levels greater than 30 ug/dl. Blood lead levels decreased significantly with age
and were slightly lower in children living on properties on which tailings were used. However,
they were higher for children attending the school that used the tailings in the playground.
Landrigan et al. (1982) studied the impact on soil and dust lead levels on removal of
leaded paint from the Mystic River Bridge in Masschusetts. Environmental studies in 1977 in-
dicated that surface soil directly beneath the bridge had a lead content ranging from 1300 to
1800 ug/g. Analysis of concomitant trace elements showed that the lead came from the bridge.
A concurrent survey of children living in Chelsea (vicinity of bridge) found that 49 percent
of 109 children had blood lead levels greater than or equal to 30 ug/dl. Of children living
more distant from the bridge, 37 percent had that level of blood lead.
These findings prompted the Massachusetts Port Authority to undertake a program to delead
the bridge. Paint on parts of the bridge that extended over neighborhoods was removed by
abrasive blasting and replaced by zinc primer. Some care was undertaken to minimize both the
occupational as well as environmental exposures to lead as a result of the blasting process.
Concurrently with the actual deleading work, a program of air monitoring was established
to check on the environmental lead exposures being created. In June, 1980, four air samples
taken at a point 27 m from the bridge had a mean lead content of 5.32 ug/m3. As a result of
these findings air pollution controls were tightened; mean air lead concentrations 12 meters
from the bridge in July were 1.43 ug/m3.
Samples of the top 1 cm of soil were obtained in July, 1980 from within 30, 30-80, and
100 m from the bridge. Comparison samples from outside the area were also obtained. Samples
taken directly under the bridge had a mean lead content of 8127 ug/g. Within 30 m of the
bridge, the mean content was 3272 ug/g, dropping to 457 ug/g at 30 to 80 m. At 100 m the soil
lead level dropped to 197 ug/g. Comparison samples ranged from 83 to 165 ug/g depending on
location.
Fingerstick blood samples were obtained on 123 children 1-5 years of age living within
0.3 km of the bridge in Charlestown. Four children (3.3 percent) had blood lead levels
greater than 30 ug/dl, with a maximum of 35 ug/dl. All four children lived within two blocks
of the bridge. Two of the four had lead paint in their homes but it was intact. None of the
76 children living more than two blocks from the bridge had blood leads greater than or equal
to 30 ug/dl, a statistically significant difference.
ShelTshear's (1973) case report from New Zealand ascribes a medically diagnosed case of
lead poisoning to high soil lead content in the child's home environment. Shellshear et al.
11-149
-------
(1975) followed up his case report of increased lead absorption resulting from exposure to
lead contaminated soil with a study carried out in Christchurch, New Zealand. Two related
activities comprised the study. First, from May, 1973 to November, 1973, a random study of
pediatric admissions to a local hospital was made. Blood samples were taken and analyzed for
lead. Homes were visited and soil samples were collected and analyzed for lead. Lead anal-
yses for both soil and blood were conducted by AAS. Second, a soil survey of the area was
undertaken. Whenever a soil lead value greater than 300 ug/g was found and a child aged 1-5
was present, the child was referred for blood testing.
The two methods of subject recruitment yielded a total of 170 subjects. Eight (4.7 per-
cent) of the children had blood lead equal to or greater than 40 |jg/dl , and three of them had
a blood lead equal to or greater than 80 pg/dl. No correlation with age was noted. The mean
blood lead of the pediatric admissions was 17.5 pg/dl with an extremely large range (4-170
|jg/dl). The mean blood lead for soil survey children was 19.5 pg/dl.
Christchurch was divided into two sections based on the date of development of the area.
The inner area had developed earlier and a higher level of lead was used there in the house
paints. The frequency distribution of soil lead levels showed that the inner zone samples had
much higher soil lead levels than the outer zone. Furthermore, analysis of the soil lead
levels by type of exterior surface of the residential unit showed that painted exteriors had
higher soil lead values than brick, stone, or concrete block exteriors.
Analysis of the relationship between soil lead and blood lead was restricted to children
from the sampled hospital who had lived at their current address for at least one year. Table
11-62 presents the analysis of these results. Although the results were not statistically
significant, they are suggestive of an association.
TABLE 11-62. ANALYSIS OF RELATIONSHIP BETWEEN SOIL LEAD AND BLOOD LEAD IN CHILDREN
Area of city
Inner zone
Outer zone
Soil
Mean
1950
150
lead (uq/q)
Range
30-11000
30-1100
n
21
47
Blood
Mean
25.4
18.3
lead (pq/dl)
Range
4-170
5-84
Source: Shell shear (1973).
Analysis of the possible effect of pica on blood lead levels showed the mean blood lead
for children with pica to be 32 Mg/d1 while those without pica had a mean of 16.8 ug/dl. The
pica blood lead mean was statistically significantly higher than the non-pica mean.
11-150
-------
Mielke et al. (1984) reports elevated blood lead and FEP levels among Hmong children
living in Minneapolis, Minnesota. The lead sources for these children included soil lead,
house paint, and leaded gasoline from vehicle traffic. Fifty percent of children with lead
poisoning (FEP > 50 pg/dl, blood lead > 30 ug/dl) inhabited homes which had soil lead levels
of 500 to 1000 (jg/g.
Wedeen et al. (1978) reported a case of lead nephropathy in a black female who exhibited
geophagia. The patient, who had undergone chelation therapy, eventually reported that she had
a habit of eating soil from her garden in East Orange, New Jersey. During spring and summer,
she continuously kept soil from her garden in her mouth while gardening. She even put a sup-
ply away for winter. The soil was analyzed for lead and was found to contain almost 700 ug/g.
The authors estimated that the patient consumed 100-500 mg of lead each year. One month after
initial hospitalization her blood lead level was 70 |jg/dl.
11.4.3.12 Summary of Soil and Dust Lead . Studies relating soil lead to blood lead levels
are difficult to compare. The relationship obviously depends on depth of soil lead, age of
the children, sampling method, cleanliness of the home, mouthing activities of the children,
and possibly many other factors. Brunekreef et al. (1983) studied a population of urban and
rural children in the Netherlands. The analyses are described in detail in Section 11.4.3.4.
Blood lead levels increased with increasing outside dustfall, with increased lead on chil-
dren's hands, and with pets in the household, and decreased with increasing number of rooms
(due to dilution or confounded SES effects). Dust lead and its related transport factors sub-
stantially increased blood lead. Table 11-63 gives some estimated slopes taken from several
different studies. The range of these values is quite large, ranging from 0.6 to 6.8. This
range is similar to the range of 1.0 to 10.0 reported by Ouggan (1980, 1983). Two studies
providing good data for slope estimates are the Stark et al. (1982) study and the Angle and
Mclntire (1982) study. These two studies gave slope estimates of 2.2 and 6.8 ug/dl per 1000
M9/g, respectively.
The relationship of house dust lead to blood lead is even more difficult to obtain.
Table 11-64 contains some values for three studies that give data permitting such caculations.
The median value of 1.8 pg/dl per 1000 ug/g for children 2-3 years old in the Stark study may
also represent a reasonable value for use here.
11.4.4 Paint Lead Exposures
A major source of environmental lead exposure for some in the general population comes
from lead contained in both interior and exterior paint on dwellings. The amount of lead
present, as well as its accessibility, depends upon the age of the residence (because older
11-151
-------
TABLE 11-63. ESTIMATES OF THE CONTRIBUTION OF SOIL LEAD TO BLOOD LEAD
Study
Angle and Mclntire
(1982) study of
children in
Omaha, NE
Stark et al.
(1982) study
of children in
New Haven, CT
Range of soil
lead values
(M9/9)
16-4792
30 - 7000
(age 0-1)
30 - 7600
(age 2-3)
Depth of
sample
2"
V
Estimated
slope (X103)
6.8
2.2
2.0
Sample
size
1075
153
334
R2
0.198
0.289
0.300
Yankel et al.
(1977) study
of children
in Kellogg, ID
Galke et al.
(1975)
study of
children in
Charleston, SC
Barltrop et
al. (1975)
study of
children in
England
50 - 24,600
9 - 7890
3/4"
2"
420 - 13,969
(group means)
2"
1.1
1.5
0.6
860
194
82
0.662
0.386
NA*
Neri et al.
(1978) study
of children
in British
Columbia
225-1800 NA
(group means,
age 1-3)
225-1800 NA
(group means,
age 2-3)
7-6 87 NA
4-6 103 NA
*NA means Not Available.
11-152
-------
TABLE 11-64. ESTIMATES OF THE CONTRIBUTION OF HOUSEDUST TO BLOOD LEAD IN CHILDREN
Range of dust
Study lead values (|jg/g)
Angle and Mclntire
(1979) study in
Omaha, NE
Stark et al. (1982)
study in New Haven,
CT
Yankel et al. (1977)
study in Kellogg,
ID
18-5571
70-7600
40-7600
9-4900
50-35,600
Age range
in years
1-18
6-18
o-i
2-3
4-7
0-4
5-9
Estimated Sample
slope (X103) size
7.18
3.36
4.02
1.82
0.02
0.19
0.20
1074
832
153
334
439
185
246
R2
0.198
0.262
0.289
0.300
0.143
0.721
0.623
buildings contain paint manufactured before lead content was regulated) and the physical con-
dition of the paint. It is generally accepted by the public and by health professionals that
lead-based paint is one major source of overtly symptomatic pediatric lead poisoning in the
United States (Lin-Fu, 1973).
The level and distribution of lead paint in a dwelling is a complex function of history,
geography, economics, and the decorating habits of its residents. Lead pigments were the
first pigments produced on a large commercial scale when the paint industry began its growth
in the early 1900's. In the 1930's lead pigments were gradually replaced with zinc and other
opacifiers. By the 1940's, titanium dioxide became available and is now the most commonly
used pigment for residential coatings. There was no regulation of the use of lead in house
paints until 1955, when the paint industry adopted a voluntary standard that limited the lead
content in interior paint to no more than 1 percent by weight of the nonvolatile solids. At
about the same time, local jurisdictions began adopting codes and regulations that prohibited
the sale and use of interior paints containing more than 1 percent lead (Berger, 1973a,b).
In spite of the change in paint technology and local regulations governing its use, in-
terior paint with significant amounts of lead was still available in the 1970's. Studies by
Berger (1973b) and by the U.S. Consumer Product Safety Commission (1974) showed a continuing
decrease in the number of interior paints with lead levels greater than 1 percent. By 1974,
only 2 percent of the interior paints sampled were found to have greater than 1 percent lead
in the dried film (U.S. Consumer Product Safety Commission, 1974).
The level of lead in paint in a residence that should be considered hazardous remains in
question. Not only is the total amount of lead in paint important, but also the accessibility
11-153
-------
of the painted surface to a child, as well as the frequency of ingestion, must be considered.
Attempts to set an acceptable lead level, i_n situ, have been unsuccessful, and preventive con-
trol measures of lead paint hazards have been concerned with lead levels in currently manufac-
tured paint. In one of its reviews, the NAS concluded: "Since control of the lead paint
hazard is difficult to accomplish once multiple layers have been applied in homes over two to
three decades, and since control is more easily regulated at the time of manufacture, we re-
commend that the lead content of paints be set and enforced at time of manufacture" (National
Academy of Sciences, 1976).
Legal control of lead paint hazards is being attempted by local communities through
health or housing codes and regulations. At the Fedjral level, the Department of Housing and
Urban Development has issued regulations for lead hazard abatement in housing units assisted
or supported by its programs. Generally, the lead level considered hazardous ranges from 0.5
to 2.5 nig/cm2., but the level of lead content selected appears to depend more on the sensiti-
vity of field measurement (using X-ray fluorescent lead detectors) than on direct biological
dose-response relationships. Regulations also require lead hazard abatement when the paint is
loose, flaking, peeling, or broken, or in some cases when it is on surfaces within reach of a
child's mouth.
Some studies have been carried out to determine the distribution of lead levels in paint
in residences. A survey of lead levels in 2370 randomly selected dwellings in Pittsburgh pro-
vides some indication of the lead levels to be found (Shier and Hall, 1977). Figure 11-29
shows the distribution curves for the highest lead level found in dwellings for three age
groupings. The curves bear out the statement often made that paint with high levels of lead
is most frequently found in pre-1940 residences. One cannot assume, however, that hTglTTead
paint is absent in dwellings built after 1940. In the case of the houses surveyed in
Pittsburgh, about 20 percent of the residences built after 1960 have at least one surface with
more than 1.5 mg/cm2.
The distribution of lead within an individual dwelling varies considerably. Lead paint
isjnost frequently found on doors and windows where lead levels greater than 1.5 mg/cm2 were
found on 2 percent of the surfaces surveyed, whereas only about 1 percent of the walls had
lead levels greater than 1.5 mg/cm2 (Shier and Hall, 1977).
In a review of the literature, Lin-Fu (1973) found general acceptance that the presence of
lead in paint is necessary but not sufficient evidence of a hazard. Accessibility in terms
of peeling, flaking, or loose paint also provide evidence for the presence of a hazard. Of the
total samples surveyed, about 14 percent of the residences had accessible paint with a lead
content greater than 1.5 mg/cm2. As discussed in Section 7.3.2.1.2, one must note that lead
oxides of painted surfaces contribute to the lead level of house dust.
11-154
-------
LEAD LEVEL (X). mg/cni
Figure 11 -29. Cumulative distribution of lead levels in dwelling
units.
Source: Shier and Hall (1977).
11-155
-------
It is not possible to extrapolate the results of the Pittsburgh survey nationally.
However, additional data from a pilot study of 115 residences in Washington, DC, showed
similar results (Hall 1974).
An attempt was made in the Pittsburgh study to obtain information about the correlation
between the quantity and condition of lead paint in buildings, and the blood lead of children
who resided there (Urban, 1976). Blood lead analyses and socioeconomic data for 456 children
were obtained, along with the information about lead levels in the dwelling. Figure 11-30 is
a plot of the blood lead levels versus the fraction of surfaces within a dwelling with lead
levels of at least 2 mg/cm2. Analysis of the data shows a low correlation between the blood
lead levels of the children and fraction of surfaces with lead levels above 2 mg/cm2, but
there is a stronger correlation between the blood lead levels and the condition of the painted
surfaces in the dwellings in which children reside. This latter correlation appeared to be
independent of the lead levels in the dwellings.
Yaffe et al. (1983) report data that suggests that soil lead possibly derived from
exterior paint was an important source for a selected group of children. They used a stable
lead isotope ratio technique.
Hammond et al. (1981, 1982) conducted a study of Cincinnati children with the dual pur-
pose of determining whether inner city children with elevated blood lead levels have elevated
fecal lead and whether fecal lead correlates with lead-base paint hazard in the home or traf-
fic density as compared with blood lead. Subjects with high blood lead levels were primarily
recruited. Some comparison children with low blood lead levels were also identified. The
three comparison children had to be residentially stable so that their low blood lead levels
were reflective of the lead intake of their current environment. The subjects from the inner
city were usually from families in extremely depressed socio-economic circumstances. Stool
samples were collected on a daily basis for up to 3 weeks, then analyzed for lead. Fecal lead
levels were expressed both as mg/kg-day and as mg/m2-day.
An environmental assessment was made at the home of each child. Paint lead exposure was
rated on a three-point scale (high, medium, and low) based on paint lead level and integrity
of the painted wall. Air lead exposure was assessed by the point scale (high, medium, and
low) based on traffic density, because there are no major point sources of lead in the
Cincinnati area.
Blood samples were collected on an irregular basis but were taken sufficiently often to
have at least one sample from a child from every house studied. The blood samples were
analyzed for lead by two laboratories that had different histories of performance in the CDC
proficiency testing program. All blood lead levels used in the statistical analysis were ad-
justed to a common base. Because of the variable number of fecal and blood lead levels, the
data were analyzed using a nested analysis of variance.
11-156
-------
(/)
QC
O
1
3 30
LU
>
LU
o25
<
LU
_J
Q 20
O
O
_J
m 15
i 1 1 1 1 1 1 1 T"
SURFACES IN BAD CONDITION, i.e., PEELING,
__ CHALKING. OR POOR SUBSTRATE
ALL SURFACES
A
^ Q ._..
. « T "^ n £
=-r « s
0
nf 1 I I I I I I | |
«
1
__
T
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
FRACTIONS OF SURFACES WITH LEAD >2 mg/cm2
Figure 11-30. Correlation of children's blood lead levels with
fractions of surfaces within a dwelling having lead
concentrations > 2 mg Pb/cm2.
Source: Urban (1976).
1.0
11-157
-------
The homes of the children were found to be distributed across the paint and traffic lead
exposure categories. Both fecal lead levels and blood levels were positively associated with
interior paint lead hazard. A marginal association between fecal lead levels and exterior
paint hazard was also obtained. Neither fecal lead nor blood lead was found to be associated
with traffic density; the definition of the high traffic density category, however, began at a
low level of traffic flow (7500 cars/day).
Examination of fecal and blood lead levels by sex and race showed that black males had
the highest fecal lead excretion rates followed by white males and black females. White
females were only represented by two subjects, both of whom had high fecal lead excretion.
Blood lead levels were more influenced by race than by sex. The results suggested that
children in high and medium paint hazard homes (high = at least 1 surface with >0.5 percent
Pb, peeling or loose) were probably ingesting paint in some form. This could not be con-
firmed, however, by finding physical evidence in the stools.
Long-term stool collection in a subset of 13 children allowed a more detailed examination
of the pattern of fecal lead excretion. Two patterns of elevated fecal lead excretion were
noted. The first was a persistent elevation compared with controls; the second was markedly
elevated occasional spikes against a normal background.
One family moved from a high-hazard home to a low one during the course of the study.
This allowed a detailed examination of the speed of deleading of fecal and blood lead level.
The fecal levels decreased faster than the blood lead levels. The blood leads were still
elevated at the end of the collection.
Gilbert et al. (1979) studied a population of Hispanic youngsters in Springfield,
Massachusetts, in a case control study designed to compare the presence of sources of lead in
homes of lead-poisoned children and appropriately matched controls. Cases were defined as
children having two consecutive blood lead levels greater than 50 pg/dl. Controls were chil-
dren with blood lead levels less than or equal to 30 ug/dl who had no previous history of lead
intoxication and were not siblings of children with blood lead levels greater than 30 ug/dl-
Study participants had to be residentially stable for at least 9 months and not have moved
into their current home from a lead contaminated one. All blood lead levels were analyzed by
Delves cup method of AAS. Cases and controls were matched by age (±3 months), sex, and neigh-
borhood area. The study population consisted of 30 lead intoxication cases and 30 control
subjects.
Home visits were undertaken to gather interview information and conduct home inspection.
Painted surfaces were assessed for integrity of the surface and lead content. Lead content
was measured by X-ray fluorometry. A surface was scored as positive if the lead content
exceeded 1.2 mg/cm2. Drinking water lead was assessed for each of the cases and was found to
11-158
-------
contain less than 50 M9/1. thought by the authors to be sufficiently low so as not to consti-
tute a hazard. Tap water samples were not collected in the homes of the controls. Soil
samples were collected from three sites in the yard and analyzed for lead by X-ray fluoro-
metry.
Cases and controls were compared on environmental lead exposures and interview data using
McNemar's test for paired samples. The odds ratio was calculated as an estimator of the rela-
tive risk on all comparisons. Statistically significant differences between cases and con-
trols were noted for lead in paint and the presence of loose paint. Large odds ratios (>10)
were obtained, suggesting a very strong association of blood lead level and paint lead expo-
sure. There appeared to be little influence of age or sex on the odds ratios.
Significant differences between cases and controls were obtained for both intact and
loose paint by individual surfaces within specific living areas of the home. Surfaces acces-
sible to children were significantly associated with lead poisoning status while inaccessible
surfaces generally were not. Interestingly, the odds ratios tended to be larger for the in-
tact surface analysis than for the loose paint one.
Median paint lead levels in the homes of cases were substantially higher than those in
the homes of controls. The median paint lead for exterior surfaces in cases was about
16-20 ug/cm2 and about 10 MQ/cm2 for interior surfaces. Control subjects lived in houses in
which the paint lead generally was less than 1.2 ug/cm2 except for some exterior surfaces.
Soil lead was significantly associated with lead poisoning; the median soil lead level for
homes of cases was 1430 M9/9i while the median soil lead level for control homes was 440 pg/g.
Rabinowitz et al. (1985b) report that refinishing activity in homes with high paint lead
was associated with elevations of blood lead averaging 69 percent. Blood lead levels of 249
infants were measured semiannually from birth to two years of age. Also, home paint was
sampled and any recent home refinishing was recorded. Mean blood lead correlated signifi-
cantly with the amount of lead in the indoor paint.
Two other studies have attempted to relate blood lead levels and paint lead as determined
by X-ray fluorescence. Reece et al. (1972) studied 81 children from two lower socioeconomic
communities in Cincinnati. Blood leads were analyzed by the dithizone method. There was con-
siderable lead in the home environment, but it was not reflected in the children's blood lead.
Analytical procedures used to test the hypothesis were not described; neither were the raw
data presented.
Galke et al. (1975), in their study of inner-city black children, measured the paint lead,
both interior and exterior, as well as soil and traffic exposure. In a multiple regression
analysis, exterior siding paint lead was found to be significantly related to blood lead
levels.
11-159
-------
Evidence indicates that a source of exposure in childhood lead poisoning is peeling lead
paint and broken lead-impregnated plaster found in poorly maintained houses. There are also
reports of exposure cases that cannot be equated with the presence of lead paint. Further,
the analysis of paint in homes of children with lead poisoning has not consistently revealed a
hazardous lead content (Lin-Fu, 1973). For example, one paper reported 5466 samples of paint
obtained from the home environment of lead poisoning cases in Philadelphia between 1964 and
1968. Among these samples of paint, 67 percent yielded positive findings, i.e., paint with
more than 1 percent lead (Tyler, 1970).
Data published or made available by the Centers for Disease Control also show that a sig-
nificant number of children with undue lead absorption occupy buildings that were inspected
for lead-based paint hazards, but in which no hazard could be demonstrated (U.S. Centers for
Disease Control, 1977a; Hopkins and Houk, 1976). Table 11-65 summarizes the data obtained
from the HEW-funded lead-based paint poisoning control projects for Fiscal Years 1981, 1979,
1978, 1975, and 1974. These data show that in Fiscal Years 1974, 1975, and 1978, in 40-50
percent of confirmed cases of elevated blood lead levels, a possible source of lead paint
hazard was not located. In fiscal year 1981, the U.S. Centers for Disease Control (1982a,b),
screened 535,730 children and found 21,897 with lead toxicity. Of these, 15,472 dwellings
were inspected and 10,666 or approximately 67 percent were found to have leaded paint. The
implications of these findings are not clear. The findings are presented in order to place in
proper perspective both the concept of total lead exposure and the concept that lead paint is
one source of lead that contributes to the total body load. The background contribution of
lead from other sources is still not known, even for those children for whom a potential lead
paint hazard has been identified; nor is it known what proportion of lead came from which
source.
11-160
-------
TABLE 11-65. RESULTS OF SCREENING AND HOUSING INSPECTION IN CHILDHOOD LEAD
POISONING CONTROL PROJECT BY FISCAL YEAR
Results
Children screened
Children with elevated
lead exposure
Dwellings inspected
Dwellings with
lead hazard
1981
535,730
21,897
15,472
10,666
1979
464,751
32,537
17,911
12,461
Fiscal year
1978
397,963
25,801
36,138
18,536
1975
440,650
28,597a
30,227
17,609
1974
371,955
16,228a
23,096
13,742
Confirmed blood lead level >40 \ig/dl.
Source: U.S. Centers for Disease Control (1977a, 1979, 1980, 1982a,b);
Hopkins and Houk, 1976.
11.5 SPECIFIC SOURCE STUDIES
The studies reviewed in this section all provide important information regarding specific
environmental sources of airborne lead that play a role in population blood lead levels.
These studies also illustrate several interesting approaches to this subject.
11.5.1 Primary Smelter Populations
Some studies of nonindustry-employed populations living in the vicinity of industrial
sources of lead pollution were triggered because evidence of severe health impairment had been
found. Subsequently, extremely high exposures and high blood lead concentrations were found.
The following studies document the excessive lead exposure that developed, as well as some of
the relationships between environmental exposure and human response.
11.5.1.1 El Paso. Texas. In 1972, the Centers for Disease Control studied the relationships
between blood lead levels and environmental factors in the vicinity of a primary smelter lo-
cated in El Paso, Texas emitting lead, copper, and zinc. The smelter had been in operation
since the late 1800's (Landrigan et al., 1975; U.S. Centers for Disease Control, 1973). Daily
hi-vol samples collected on 86 days between February and June, 1972, averaged 6.6 ug/m3.
These air lead levels fell off rapidly with distance, reaching background values approximately
5 km from the smelter. Levels were higher downwind, however. High concentrations of lead in
soil and house dusts were found, with the highest levels occurring near the smelter. The geo-
metric means of 82 soil and 106 dust samples from the sector closest to the smelter were 1791
11-161
-------
and 4022 ug/g, respectively. Geometric means of both soil and dust lead levels near the
smelter were significantly higher than those in study sectors 2 or 3 km farther away.
Sixty-nine percent of children 1 to 4 years old living near the smelter had blood lead
levels greater than 40 ug/dl, and 14 percent had blood lead levels that exceeded 60 ug/d1
Concentrations in older individuals were lower; nevertheless, 45 percent of the children 5 to
9 years old, 31 percent of the individuals 10 to 19 years old, and 16 percent of the indivi-
duals above 19 had blood lead levels exceeding 40 ug/dl. The data presented preclude calcula-
tions of means and standard deviations.
Data for people aged 1-19 years of age living near the smelter showed a relationship
between blood lead levels and concentrations of lead in soil and dust. For individuals with
blood lead levels greater than 40 ug/dl, the geometric mean concentration of lead in soil at
their homes was 2587 pg/g, whereas for those with a blood lead concentration less than 40
ug/dl, home soils had a geometric mean of 1419 ng/g. For house dust, the respective geometric
means were 6447 and 2067 |jg/g. Length of residence was important only in the sector nearest
the smelter.
Additional sources of lead were also investigated. A relationship was found between
blood lead concentrations and lead release from pottery, but the number of individuals exposed
to lead-glazed pottery was very small. No relationships were found between blood lead levels
and hours spent out-of-doors each day, school attendance, or employment of a parent at the
smelter. The reported prevalence of pica also was minimal.
Data on dietary intake of lead were not obtained because there was no food available from
sources near the smelter since the climate and proximity to the smelter prevented any farming
in the area. It was unlikely that the dietary lead intakes of the children from near the
smelter or farther away were significantly different. It was concluded that the primary
factor associated with elevated blood lead levels in the children was ingestion or inhalation
of dust containing lead.
Morse et al. (1979) conducted a follow-up investigation of the El Paso smelter to deter-
mine whether the environmental controls instituted following the 1972 study had reduced the
lead problem described. In November, 1977, all children 1 to 18 years old living within
1.6 km of the smelter on the U.S. side of the border were surveyed. Questionnaires were ad-
ministered to the parents of each participant to gather background data.
Venous blood samples were drawn and analyzed for lead by modified Delves cup spectropho-
tometry. House dust and surface soil samples, as well as sample pottery items, were taken
from each participant's residence. Dust and soil samples were analyzed for lead by AAS.
Pottery lead determinations were made by the extraction technique of Klein. Paint, food, and
water specimens were not collected because the earlier investigations of the problem had
demonstrated these media contributed little to the lead problem in El Paso.
11-162
-------
Fifty-five of 67 families with children (82 percent) agreed to participate in the study.
There were 142 children examined in these homes. The homes were then divided into two groups.
Three children lived in homes within 0.8 km of the smelter. Their mean blood lead level in
1977 was 17.7 |jg/dl. By contrast, the mean blood lead level of 160 children who lived within
0.8 km of the smelter in 1972 had been 41.4 M9/dl. In 1977, 137 children lived in homes lo-
cated 0.8-1.6 km from the smelter. Their mean blood lead level was 20.2 (jg/dl. The mean
blood level of 96 children who lived in that same area in 1972 had been 31.2 (jg/dl.
Environmental samples showed a similar improvement. Dust lead fell from 22,191 to 1,479
Mg/g while soil lead fell from 1,791 to 427 ug/g closest to the smelter. The mean air lead
concentration at 0.4 km from the smelter decreased from 10.0 to 5.5 ug/m3 and at 4.0 km from
2.1 to 1.7 ug/m3. Pottery was not found to be a problem.
11.5.1.2 CDC-EPA Study. Baker et al. (1977b), in 1975, surveyed 1774 children 1-5 years old,
most of whom lived within 4 miles of lead, copper, or zinc smelters located in various parts
of the United States. Blood lead levels were modestly elevated near 2 of the 11 copper and 2
of the 5 zinc smelters. Although blood lead levels in children were not elevated in the
vicinity of three lead smelters, their FEP levels were somewhat higher than those found in
controls. Increased levels of lead and cadmium in hair samples were found near lead and zinc
smelters; this was considered evidence of external exposure. No environmental determinations
were made for this study..
11.5.1.3 Meza Valley, Yugoslavia. A series of Yugoslavian studies investigated exposures to
lead from a mine and a smelter in the Meza Valley over a period of years (Fugas et al., 1973;
Graovac-Leposavic et al. 1973; Milic et al., 1973; Djuric et al., 1971, 1972). In 1967,
24-hour lead concentrations measured four on different days varied from 13 to 84 |jg/m3 in the
village nearest the smelter, and concentrations of up to 60 ug/m3 were found as far as 5 km
from the source. Mean particle size in 1968 was less than 0.8 urn. Analysis of some common
foodstuffs showed concentrations that were 10-100 times higher than corresponding foodstuffs
from the least exposed area (Mezica) (Djuric et al., 1971). After January, 1969, when partial
control of emissions was established at the smelter, weighted average weekly exposure was cal-
culated to be 27 |jg/m3 in the village near the smelter. In contrast to this, the city of
Zagreb (Fugas et al., 1973), which has no large stationary source of lead, had an average
weekly air lead level of 1.1 ug/m3.
In 1968, the average concentration of ALA in urine samples from 912 inhabitants of 6 vil-
lages varied'by village from 9.8-13 mg/1. A control group had a mean ALA of 5.2 mg/1. Data
on lead in blood and the age and sex distribution of the villagers were not given (Djuric et
al-, 1971).
11-163
-------
Of the 912 examined, 559 had an ALA level greater than 10 mg/1 of urine. In 1969, a more
extensive study of 286 individuals with ALA greater than 10 mg/1 was undertaken (Graovac-
Leposavic et al. 1973). ALA-U increased significantly from the previous year. When the pub-
lished data were examined closely, there appeared to be some discrepancies in interpretation.
The exposure from dust and from food might have been affected by the control devices, but no
data were collected to establish this. In one village, Zerjua, ALA-U dropped from 21.7 to 9.4
mg/1 in children 2-7 years of age. Corresponding ALA-U values for 8- to 15-year-olds and for
adult men and women were reduced from 18.7 to 12.1, from 23.9 to 9.9, and from 18.5 to 9.0
mg/1, respectively. Because lead concentrations in air (Fugas et al., 1973), even after 1969,
indicated an average exposure of 25 ug/m3, it is possible that some other explanation should
be sought. The author indicated in the report that the decrease in ALA-U showed "the depen-
dence on meteorologic, topographic, and technological factors" (Graovac-Leposavic et al.,
1973).
Fugas (1977) in a later report estimated the time-weighted average exposure of several
populations studied during the course of this project. Stationary samplers as well as per-
sonal monitors were used to estimate the exposure to airborne lead for various parts of the
day. These values were then coupled with estimated proportions of time at which these expo-
sures held. In Table 11-66, the estimated time-weighted air lead values as well as the
observed mean blood lead levels for these studied populations are presented. An increase in
blood lead values occurs with increasing air lead exposure.
TABLE 11-66. MEAN BLOOD LEAD LEVELS IN SELECTED YUGOSLAVIAN
POPULATIONS, BY ESTIMATED WEEKLY TIME-WEIGHTED AIR LEAD EXPOSURE
Population
Rural I
Rural II
Rural III
Postmen
Customs officers
Street car drivers
Traffic policemen
N
49
47
45
44
75
43
24
Time- weighted
air lead, (ug/m3)
0.079
0.094
0.146
1.6
1.8
2.1
3.0
Blood lead level,
Mean
7.9
11.4
10.5
18.3
10.4
24.3
12.2
(pg/dl)
SO
4.4
4.8
4.0
9.3
3.3
10.5
5.1
Source: Fugas, 1977.
11-164
-------
11.5.1.4 Kosovo Province. Yugoslavia. Residents living in the vicinity of the Kosovo smelter
were found to have elevated blood lead levels (Popovac et al., 1982). In this area of
Yugoslavia, five air monitoring stations had been measuring air lead levels since 1973. Mean
air lead varied from 7.8 to 21.7 Mg/m3 in 1973; by 1980 the air lead averages ranged from 21.3
to 29.2 pg/m3. In 1978 a pilot study suggested that there was a significant incidence of ele-
vated blood lead levels in children of the area. Two major surveys were then undertaken.
In August, 1978, letters were sent to randomly selected families from the business commu-
nity, hospitals or lead-related industries in the area. All family members were asked to come
to a hospital for primary screening by erythrocyte protoporphyrin. A central population of
comparable socioeconomic and dietary background was collected from a town without lead emis-
sions. Blood levels were determined primarily for persons with EP greater than 8 Mg/g Hgb.
EP was measured by a hematofluorimeter, while blood ]ead was determined by the method of
r . . ^,,4-inn with firaohite furnace and background correction.
Fernandez using atomic absorption witn graym^ »
Mean EP values were higher in the 1978 survey for exposed residents compared to controls
in the average age group. EP values seemed to decline with age. Similar differences were
noted for blood lead levels. The observed mean blood leads, ranging from 27.6 in the greater
than 15 year age group to 50.9 pg/dl in the 5 to 10 year group, suggest substantial lead ex-
posure of these residents. In the control group the highest blood lead level was 19 ug/dl.
In December 1980 a second survey was conducted to obtain a more representative sample of
' aro, Letters were sent again, and 379 persons responded. EP levels
persons residing in the area. Letter* - *,...,,
-man uprsus 1978, although the differences were not statistically
were hiaher in all ages in 198U versus
significant The air lead levels increased from 14.3 ^ in 1978 to 23.8 ug/m3 in 1980.
Cowarina the 1980 blood lead results with the 1978 control group shows that the 1980
levels were hi'gher in each age group. °^**r than 15 yearS had hl'Qher """ bl00d lead
levels than the females (3*3 versus,32 4 pg/dl) ^^ ^ ^ ^^ ^ ^
1.5.1.5 S!^^^^m^^^^^^, The exposed population
lead smelter and children from . a nursery school and 80 primary school children
consisted of 85 children '*** ^ ^^ ^ ^ ^ ^ ^.^
aged 8 to 11. The control pop^at ^ ^^ ^ & ^ ^ ^ ^^
school children aged 8-11. inC area had fl much higher lifetime exposure.
the older children living in the s ^ ^^ sgmpies ^ ^ stHpping voltammetry by
Blood lead analysis was p ^ ^ ^^ ^ iQ_m ^g/^ Reporled reproduci-
Morrell 's method. Precision was ^^^ reanalyzed by AAS using graphite furnace
bility was also good. All samp ^ ^ ^^ ^ ^^ ^^ obtained by the
and background correction by tne ^^ (average difference 1.4 jig/dl; correla-
second method were quite similar to tnose
t-ion coefficient, 0.962).
11-165
-------
Air was sampled for lead for 1 month at three sampling sites. The sites were located at
150 m, 300 m, and 4 km from the wall of the lead smelter. The average air lead levels were
2.32, 3.43, and 0.56 ng/m3, respectively.
A striking difference in blood lead levels of the exposed and control populations was ob-
served; levels in the exposed population were almost twice that in the control population.
There was no significant difference between nursery school and primary school children. The
geometric mean for nursery school children was 15.9 and 8.2 for exposed and control, respecti-
vely. For primary school it was 16.1 and 7.0 ug/dl. In the exposed area, 23 percent of the
subjects had blood lead levels between 21 and 30 ug/dl and 3 percent greater than 31 ug/dl
No control children had blood lead levels greater than 20 ug/dl. The air leads were between
2-3 (jg/m3 in the exposed and 0.56 ug/m3 in the control cases.
11.5.1.6 Hartwell Study. Hartwell et al. (1983) report a study of 4 primary smelters: two
lead and two zinc. Study subjects were recruited in accordance with a statistical sampling
plan based on diffusion modeling. Subjects were recruited to represent a variety of aqes:
1-5 years, 6-18 years, 20-40 years, and, in two sites, >60 years. Environmental samples
covering the important environmental sources of lead were obtained, as were blood samples
Unfortunately, air sampling was only conducted for about 1 month in each of the study areas.
Dust, water, and soil samples were also collected and analyzed for lead. Table 11-67 sum-
marizes the descriptive results of this study in terms of blood lead levels. Table 11-68
presents the Spearman correlation coefficient obtained.
11.5.2 Battery Plants
Studies of the effects of storage battery plants have been reported from France and Italy
(Dequidt et al., 1971; De Rosa and Gobbato, 1970). The French study found that children from
an industrialized area containing such a plant excreted more ALA than those living in a diffe-
rent area (Dequidt et al., 1971). Increased urinary excretion of lead and coproporphyrins was
found in children living up to 100 m from a battery plant in Italy (De Rosa and Gobbato,
1970). Neither study gave data on plant emissions or lead in air.
11.5.3 Secondary Smelters
Zielhuis et al. (1979) studied children living in the vicinity of the Arnhem secondary
lead smelter. In 1976 they recruited children to serve as subjects and controls. The chil-
dren chosen were 2 and 3 years old. Parents were asked to complete a questionnaire for back-
ground information. Two-mi venous samples were collected from 17 children living less than 1
km, from 54 children living 1-2 km, and from 37 children living greater than 2 km from the
smelter (control group). Blood samples were analyzed for lead by graphite furnace AAS and for
11-166
-------
TABLE 11-67. LEVELS OF LEAD RECORDED IN HARTWELL ET AL. (1983) STUDY
Smelter
Bartlesville
Palmerton
Ajo
Anaconda
Distance from
smelter
3.5-24.0
1.3-3.7
0.8-4.3
0.8-1.5
11.0-26.0
5.4-14.5
3.3-9.9
0.3-2.8
3.4-68.0
1.0-6.4
0.5-2.3
0.5-1.3
10.0-26.0
3.5-21.0
2.0-11.0
2.0-3.5
Air
131
203
299
309
361
563
128
278
94
108
191
256
141
176
91
255
Dust
241
409
386
441
263
201
198
438
74.2
60.0
64.7
116
235
164
210
398
Water
6.04
4.56
6.81
7.63
8.7
6.0
2.8
1.8
6.9
11.5
13.3
3.1
3.10
3.52
3.02
3.83
Soil
34.8
243
829
821
532
117
326
331
57.8
64.5
76.5
94.8
75
115
294
424
PbB
Ages 1-5
10.5
24.7
39.6
18.8
10.3
11.3
12.6
15.9
9.9
10.6
10.5
9.2
21.0
17.3
18.9
21.5
Ages 6-10
12.4
12.9
21.8
20.3
12.4
10.2
11.2
10.3
7.8
7.7
6.9
6.9
19.0
11.9
14.3
17.9
TABLE 11-68. SPEARMAN CORRELATIONS OF LEAD IN AIR, WATER, DUST, SOIL, AND PAINT
WITH LEAD LEVELS IN BLOOD: BY SITE AND AGE GROUPS, 1978-1979
Aae (yr)
Bartlesville
Palmerton
Air
Water
Dust
Soil
Paint
Air
Water
Dust
Soil
Paint
1-5
Blood
0.40*
0.05
0.20
0.33*
-0.06
-0.12
-0.06
0.06
0.16
-0.02
6-18
Blood
0.22*
0.14
0.10
0.13
0.06
0.02
0.11
-0.07
0.20
0.06
20-40
Blood
0.27*
0.07
0.21
0.07
-0.12
-0.01
-0.05
0.23*
Over 60
Blood
0.19
0.23
0.00
-0.06
*Significantly different from zero at 0.05 level.
11-167
-------
FEP by the method of Piomelli. Air measurements for lead were made in autumn, 1976. Samples
were established about 2 km northeast and about 0.4 km north of the plant. Air lead levels
ranged from 0.8 to 21.6 (jg/m3 northeast and from 0.5 to 2.5 (jg/m3 north of the plant.
Blood leads were statistically significantly higher closer to the smelter. For all chil-
dren the mean blood lead level was 19.7 ug/dl for the less than 1 km and 11.8 |jg/dl for the
controls (>2 km). Similarly, FEP levels were higher for the closer (41.9 ug/100 ml erythro-
cytes) children as opposed to the control (32.5 ug/100 ml RBC). Higher blood levels were
associated with lower socioeconomic status.
Further investigation of this smelter was undertaken by Brunekreef et al. (1981) and
Diemel et al. (1981). In May, 1978, venipuncture blood samples were collected from 95 one- to
three-year-old children living within 1 km of the smelter. Blood leads were determined by
graphite AAS.
Before the blood sampling, an environmental sampling program was conducted. The samples
collected are listed in Table 11-69. Questionnaires were administered to collect background
and further exposure information. A subset of 39 children was closely observed for 1 or 2
days for mouthing behavior. Table 11-69 also presents the overall results of the environmen-
tal sampling. As can be readily seen, there is a low exposure to airborne lead (geometric
mean) 0.41 ug/m3 with a range of 0.28-0.52 |jg/m3). Soil exposure was moderate, although high.
Interior dust was high in lead (geometric mean of 967 ug/g with a maximum of 4741 ug/g). In a
few homes, high paint lead levels were found. Diemel et al. (1981) extended the analysis of
the environmental samples. They found that indoor pollution was lower than outside. In
Arnhem, it was found that lead is carried into the homes in particulate form by sticking to
shoes. Most of the lead originated from soil from gardens and street dust.
Simple correlation coefficients were calculated to investigate the relationship between
log blood lead and the independent variables. Significantly, correlations were found with
quantity of house dust, quantity of deposited lead indoors, observational score of dustiness,
age of child, and the average number of times an object is put in the mouth. Multiple regre-
ssion analyses were calculated on four separate subpopulations. Among children living in
houses with gardens, the combination of soil lead level and educational level of the parents
explained 23 percent of the variations of blood lead. In children without gardens, the amount
of deposited lead indoors explained 26 percent of the variance. The authors found that an in-
crease in soil lead level from 100-600 ug/g resulted in an increase in blood lead of 6.3
ug/dl.
In a Dallas, Texas, study of two secondary lead smelters, the average blood lead level of
exposed children was found to be 30 ug/dl versus an average of 22 ug/dl in control children
(Johanson and Luby, 1972). For the two study populations, the air and soil lead levels were
3.5 and 1.5 ug/m3 and 727 and 255 ug/9, respectively.
11-168
-------
TABLE 11-69. ENVIRONMENTAL PARAMETERS AND METHODS: ARNHEM LEAD STUDY, 1978d
Parameter
Method
Geometric mean
Range
I
I»
uo
1. Lead in ambient air
(ug/m3)
2. Lead in dustfall
(ug/m3-day)
3. Lead in soil
(pg/g)
4. Lead in street dust
(ug/g)
5. Lead in indoor air
(ug/m3)
6. Lead in dustfall
indoors (ug/m3*day)
7. Lead in floor dust
(M9/g)
8. Easily available
lead indoors
9. Lead in tapwater
10. Dustiness of homes
High-volume samples; 24-hr measurements
at 6 sites, continuously for 2 months
Standard deposit gauges; 7-day measurements
at 22 sites, semicontinuously for 3 months
Sampling in gardens of study populations;
analysis of layers from 0 to 5 cm and
5 to 20 cm
Samples at 31 sites, analysis of fraction
<0.3mm
Low-volume samples; 1-month measurements
in homes of study population, continuously
for 2 months
Greased glass plates, of 30 x 40 cm; 1-month
measurements in homes of study population,
continuously for 3 months
Vacuum cleaner with special filter
holder; 5 samples, collected on 3 different
occasions; with intervals of approximately
1 month, in homes of study populations
Wet tissues, 1 sample in homes of study
population
Proportional samples, during 1 week in
homes of study population
Visual estimation, on a simple scale ranging
from 1 (clean) to 3 (dusty); 6 observations
in homes of study population
0.41
467
240
690
0.26
7.34
fine 957
course 282
0.28-0.52
108-2210
21-1126
77-2667
0.13-0.74
1.36-42.35
463-4741
117-5250
85% of samples <20 ug Pb/tissue
5.0
(arithmetic) mean
<0.5-90.0
All lead analyses were performed by atomic absorption spectrophotometry, except part of the tapwater analysis,
which was performed by anodic stripping voltametry. Lead in tapwater analyzed by the National Institute of
Drinking Water Supply in Leidscherdam. Soil and street dust analyzed by the Laboratory of Soil and Plant
Research in Oosterbeek. (Zielhuis, et. al., 1979; Diemel, et. al., 1981)
-------
In Toronto, Canada, the effects of two secondary lead smelters on the blood and hair lead
levels of nearby residents have been extensively studied (Ontario Ministry of the Environment,
1975; Roberts et al., 1974). In a preliminary report, Roberts et al. (1974) stated that blood
and hair lead levels were higher in children living near the two smelters than in children
living in an urban control area. Biologic and environmental lead levels were reported to de-
crease with increasing distance from the base of the smelter stacks.
A later and more detailed report identified a high rate of lead fallout around the two
secondary smelters (Ontario Ministry of the Environment, 1975). Two groups of children living
within 300 m of each of the smelters had geometric mean blood lead levels of 27 and 28 ug/dl ,
respectively; the geometric mean for 1231 controls was 17 (jg/dl. Twenty-eight percent of the
sample children tested near one smelter during the summer and 13 percent of the sample chil-
dren tested near the second smelter during the winter had blood lead levels greater than 40
ug/dl. Only 1 percent of the controls had blood lead levels greater than 40 ug/dl . For chil-
dren, blood lead concentrations increased with proximity to both smelters, but this trend did
not hold for adults, generally. The report concluded that soil lead levels were the main de-
terminant of blood lead levels; this conclusion was disputed by Horn (1976).
Blood lead levels in 293 Finnish individuals, aged 15-80, were significantly correlated
with proximity to a secondary lead smelter (Nordman et al., 1973). The geometric mean blood
lead concentration for 121 males was 18.1 ug/dl > for *72 females, it was 14.3 (jg/dl. In 59
subjects who spent their entire day at home, a positive correlation was found between blood
lead and distance from the smelter up to 5 km. Only one of these 59 individuals had a blood
lead greater than 40 pg/dl , and none exceeded 50 ug/dl .
11.5.4 Secondary Exposure of Children
Excessive intake and absorption of lead on the part of children can result when parents
who work in a dusty environment with a high lead content bring dust home on their clothes,
shoes, or even their automobiles. Once they are home, their children are exposed to the dust.
Landrigan et al. (1976) reported that the 174 children of smelter workers who lived with-
in 24 km of the smelter had significantly higher blood lead levels, a mean of 55.1 ug/dl , than
the 511 children of persons in other occupations living in the same areas whose mean
blood lead levels were 43.7 ug/dl. Analyses by EPA of the data collected in Idaho showed that
employment of the father at a lead smelter, at a zinc smelter, or in a lead mine resulted in
higher blood lead levels in the children living in the same house as opposed to those children
whose fathers were employed in different locations (Table 11-70). The effect associated with
parental employment appears to be much more prominent in the most contaminated study areas
nearest to the smelter. This may be the effect of an intervening socioeconomic variable: the
11-170
-------
TABLE 11-70. GEOMETRIC MEAN BLOOD LEAD LEVELS FOR CHILDREN BASED ON REPORTED OCCUPATION
OF FATHER, HISTORY OF PICA, AND DISTANCE OF RESIDENCE FROM SMELTER
(micrograms per deciliter)
Area
1
2
3
4
5
6
Distance
from
smelter, km
1.6
1.6 to 4.0
4.0 to 10.0
10.0 to 24.0
24.0 to 32.0
75
Lead
smelter
worker
No
Pica Pica
78.7 74.2
50.2 52.2
33.5 33.3
30.3
24.5
-
Lead/zinc mine
worker
Pica
75.3
46.9
36.7
38.0
31.8
-
No
Pica
63.9
46.9
33.5
32.5
27.4
-
Zinc smelter
worker
No
Pica Pica
69.7 59.1
62.7 50.3
36.0 29.6
40.9 36.9
-
-
Other
occupations
No
Pica Pica
70.8
37.2
33.3
-
28.0
17.3
59.9
46.3
32.6
39.4
26.4
21.4
Source: Landrigan et al. 1976.
lowest paid workers, employed in the highest exposure areas within the industry, might be ex-
pected to live in the most undesirable locations, closest to the smelter.
Landrigan et al. (1976) also reported a positive history of pica for 192 of the 919 chil-
dren studied in Idaho. This history was obtained by physician and nurse interviews of
parents. Pica was most common among 2-year-old children and only 13 percent of those with
pica were above age 6. Higher blood lead levels were observed in children with pica than in
those without pica. Table 11-70 shows the mean blood lead levels in children as they were af-
fected by pica, occupation of the father, and distance of residence from the smelter. Among
the populations living nearest to the smelter, environmental exposure appears to be sufficient
at times to more than overshadow the effects of pica, but this finding may also be caused by
inadequacies inherent in collecting data on pica. These data indicate that in a heavily con-
taminated area, blood lead levels in children may be significantly increased by the inten-
tional ingestion of nonfood materials having a high lead content.
Data on the parents' occupation are, however, more reliable. It must be remembered also
that the study areas were not homogeneous socioeconomically. In addition, the specific type
of work an individual does in an industry is probably much more important than simply being
employed in a particular industry. The presence in the home of an industrial employee exposed
occupationally to lead may produce increases in the blood lead levels ranging from 10 to 30
percent.
11-171
-------
The importance of the infiltration of lead dusts onto clothing, particularly the under-
garments, of lead workers and their subsequent transportation has been demonstrated in a num-
ber of studies on the effects of smelters (Martin et al. , 1975). It was noted in the United
Kingdom that elevated blood lead levels were found in the wives and children of workers even
though they resided some considerable distance from the facility. It was most prominent in
the workers themselves, who had elevated blood lead levels. Quantities of lead dust were
found in workers' cars and homes. It apparently is not sufficient for a factory merely to
provide outer protective clothing and shower facilities for lead workers. In another study in
Bristol, 650-1400 ug/g of lead was found in the undergarments of workers as compared with 3-13
|jg/g in undergarments of control subjects. Lead dust will remain on the clothing even after
laundering: up to 500 mg of lead has been found to remain on an overall garment after washing
(Lead Development Association, 1973).
Baker et al. (1977a) found blood lead levels greater than 30 ug/dl in 38 of 91 children
whose fathers were employed at a secondary lead smelter in Memphis, TN. House dust, the only
source of lead in the homes of these children, contained a mean of 2687 ug/g compared with 404
ug/g in the homes of a group of matched controls. Mean blood lead levels in the workers'
children were significantly higher than those for controls and were closely correlated with
the lead content of household dust. In homes with lead in dust less than 1000 ug/g, 18 chil-
dren had a mean blood lead level of 21.8 ± 7.8 ug/dl, whereas in homes where lead in dust was
greater than 7000 ug/g, 6 children had mean blood lead levels of 78.3 ± 34.0 ug/dl. See Sec-
tion 7.3.2.1.6 for a further discussion of household dust.
Other studies have documented increased lead absorption in children of families where at
least one member was occupationally exposed to lead (Fischbein et al., 1980a). The occupa-
tional exposures involved battery operations (Morton et al., 1982; U.S. Centers for Disease
Control, 1977b; Dolcourt et al., 1978, 1981; Watson et al. , 1978; Fergusson et al., 1981) as
well as other occupations (Snee, 1982b; Rice et al., 1978).
In late summer of 1976, a battery plant in southern Vermont provided the setting for the
first documented instance of increased lead absorption in children of employees in the battery
industry. The data were first reported by the U.S. Centers for Disease Control (1977b) and
more completely by Watson et al. (1978). Reports of plant workers exposed to high levels of
lead stimulated a study of plant employees and their children in August and September, 1975.
In the plant, lead oxide powder is used to coat plates in the construction of batteries.
Before the study, the work setting of all 230 employees of the plant had been examined and 62
workers (22 percent) were identified as being at risk for high lead exposure. All of the
high-risk workers interviewed reported changing clothes before leaving work and 90 percent of
them reported showering. However, 87 percent of them stated that their work clothes were
washed at home.
11-172
-------
Of the high-risk employees, 24 had children between the ages of 1 and 6 years. A case-
control study was conducted in the households of 22 of these employees. Twenty-seven children
were identified. The households were matched with neighborhood controls, including 32 control
children. None of the control family members worked in a lead industry. Capillary blood
specimens were collected from all children and the 22 battery plant employees had venous spec-
imens taken. All blood samples were analyzed for lead by AAS. Interviewers obtained back-
ground data, including an assessment of potential lead exposures.
About 56 percent of the employees' children had blood leads greater than 30 ug/dl com-
pared with about 13 percent of the control children. Mean blood lead levels were signifi-
cantly different, 31.8 ug/dl and 21.4 ug/dl, respectively. Blood lead levels in children were
significantly correlated with employee blood lead levels.
House dust lead levels were measured in all children's homes. Mean values were 2239.1
M9/Q and 718.2 ug/g for employee and control homes, respectively; this was a statistically
significant difference. Examination of the correlation coefficient between soil lead and
blood lead levels in the two sets of homes showed a marginally significant coefficient in the
employee households but no correlation in the control homes. Tap water and paint lead levels
did not account for the observed difference in blood leads between children of workers and
neighborhood controls. It is significant that these findings were obtained despite the chang-
ing of clothes at the plant.
Morton et al. (1982) conducted their study of children of battery plant workers and con-
trols during February-March, 1978. Children were included in the study if one parent had at
least 1 year of occupational exposure, if they had lived at the same residence for at least 6
months and if they were from 12-83 months of age. Children for the control group had to have
' . , atinnal PXDosure to lead for 5 years, and had to have lived at the same ad-
no parental occupational exposure >.» j >
dress at least 6 months. ,
Th- t four children were control-matched to the exposed group by neighborhoods and age
(+1 year"/" ^matching was thought necessary for sex because in this age group blood lead
~ Th seiection of the control population attempted to adjust for
levels are unaffected by sex. me
both socioeconomic status as well as exposure to automotive lead.
pillary blood specimens were collected concurrently for each matched pair. Blood lead
p . J . ..0 rnr iab using a modified Delves cup AAS procedure. Blood lead
levpl*; wpre measured by tne i»uv, !«" =
levels for the employees for the previous year were obtained from company records. Question-
... * A ,t thP same time as the blood sampling to obtain background informa-
naires were administered at tne bai"<=
complete the interview to try to get a more accurate picture
tion. The homemaker was asked to comH.e
of the hygiene practices followed by the employees.
11-173
-------
Children's blood lead levels differed significantly between the exposed and control
groups. Fifty-three percent of the employees' children had blood lead levels greater than 30
jjg/dl, while no child in the control population had a value greater than 30 ug/dl. The mean
blood lead for the children of the employees was 49.2 ug/dl with a standard deviation of 8.3
(jg/dl. These data represent the population average for yearly individual average levels. The
employees had an average greater than 60 |jg/dl. Still, this is lower than the industry
average. Of the eight children with blood levels greater than 40 ug/dl, seven had fathers
with blood lead greater than 50 ug/dl. Yet there was not a significant correlation between
children's blood lead level and father's blood lead level.
Investigations were made into the possibility that other lead exposures could account for
the observed difference in blood lead levels between children of employees and control chil-
dren. In 11 of the 33 pairs finally included in the study, potential lead exposures other
than fathers' occupations were found in the employee child of the matched pair. These in-
cluded a variety of lead sources such as automobile body painting, casting of lead, and
playing with spent shell casings. The control and exposed populations were again compared
after removing these 11 pairs from consideration. There was still a statistically significant
difference in blood lead level between the two groups of children.
An examination of personal hygiene practices of the workers showed that within high-ex-
posure category jobs, greater compliance with recommended lead containment practices resulted
in lower mean blood lead levels in children. Mean blood leads were 17.3, 36.0, and 41.9 ug/dl
for good, moderately good, and poor compliance groups, respectively. In fact, there was only
a small difference between the good hygiene group within the high-exposure category and the
mean of the control group (17.3 ug/dl versus 15.9 ug/dl). Insufficient sample sizes were
available to evaluate the effect of compliance on medium and low lead exposures for fathers.
Dolcourt et al. (1978) investigated lead absorption in children of workers in a plant
that manufactures lead-acid storage batteries. The plant became known to these researchers as
a result of finding an elevated blood lead level in a 20-month-old child during routine
screening. Although the child was asymptomatic, his mother proved not to be. Two siblings
were also found to have elevated blood lead levels. The mother was employed by the plant; her
work involved much hard labor and brought her into continual contact with powdery lead oxide.
No uniforms or garment covers were provided by the company. As a result of these findings,
screening was offered to all children of plant employees.
During February to May, 1977, 92 percent of 63 eligible children appeared for screening.
Age ranged from 10 months to 15 years. About equal numbers of girls and boys underwent
screening. Fingerstick blood samples were collected on filter paper and were analyzed for
lead by AAS. Children with blood lead levels equal to or greater than 40 ug/dl were referred
11-174
-------
for more detailed medical evaluation including an analysis of a venous blood specimen for
lead. Dust samples were collected from carpeting in each home and analyzed for lead by gra-
phite furnace AAS. Home tap water was analyzed for lead by AAS, and house paint was analyzed
for lead by XRF.
Of the 58 children who had the initial fingerstick blood lead elevation, 69 percent had
blood lead levels equal to or greater than 30 pg/dl. Ten children from six families had blood
lead levels equal to or greater than 40 ug/dl, and blood lead levels were found to vary
markedly with age. The 0- to 3-year-old category exhibited the highest mean (48.6 ug/dl) with
the 3- to 6-year-olds the next highest (38.2 ug/dl). Lowest mean values were found in the
equal to or greater than 10-year-old group (26.7 ug/dl).
More detailed investigation of the six families with the highest blood lead levels in
their children revealed the following: five of the six lived in rural communities, with no
pre-existing source of lead from water supply, house paint, industrial emissions, or heavy
automobile traffic. However, dust samples from the carpets exhibited excessively high lead
concentrations. These ranged from 1700 to 84,050 ug/g.
Fergusson et al. (1981) sampled three population groups: general population, employees
of a battery plant, and children of battery plant employees, using hair lead levels as indices
of lead. Hair lead levels ranged from 1.2 to 110.9 ug/g in the 203 samples from the general
population. The distribution of hair lead levels was nearly lognormal. Employees of the bat-
tery factory had the highest hair lead levels (median ~250 ug/g), while family members (median
~40 ug/g) had a lesser degree of contamination and the general population (median ~5 ug/g)
still less.
Analysis of variance results indicated a highly significant difference between mean lead
levels of the general survey and family members of the employees, and a significant difference
between the mean lead levels in the hair of the employees and their families. No significant
differences were found comparing mean hair lead levels among family members in terms of age
and sex. The analyses of the house dust suggested that the mechanism of exposure of family
members is via the lead in dust that is carried home. Mean dust lead level among the homes of
factory employees was 5580 ug/g while the dust inside of houses along a busy road was only
1620 ug/g. Both of these concentrations are for particles less than 0.1 mm.
Dolcourt et al. (1981) reported two interesting cases of familial exposure to lead caused
by recycling of automobile storage batteries. The first case was of a 22-member, four-
generation family living in a three-bedroom house in rural eastern North Carolina. The great-
grandfather of the index case worked at a battery recycling plant. He had two truckloads of
spent casings delivered to the home to serve as fuel for the wood stove; the casings were
burned over a 3-month period.
11-175
-------
The index case presented with classic signs of acute lead encephalopathy, the most severe
and potentially fatal form of acute lead poisoning. The blood lead level was found to be 220
pg/dl. Three months after initial diagnosis and after chelation therapy, she continued to
have seizures and was profoundly mentally retarded. Dust samples were obtained by vacuum
cleaner and analyzed for lead by flameless AAS. Dust from a sofa near the wood stove con-
tained 13,283 ug/g lead, while the kitchen floor dust had 41,283 ug/g. There was no paint
lead. All other members of the family had elevated blood lead levels ranging from 27-256
ug/dl.
The other case involved a truck driver working in a low-exposure area of a battery
recycling operation in rural western North Carolina. He was operating an illegal battery re-
cycling operation in his home by melting down reclaimed lead on the kitchen stove. No family
member was symptomatic for lead symptoms but blood lead levels ranged from 24 to 72 ug/dl
Soil samples taken from the driveway, which was paved with fragments of the discarded battery
casings, contained 12-13 percent lead by weight.
In addition to families being exposed as a result of employment at battery plants stu-
dies have been reported recently for smelter worker families (Rice et al., 1978; Snee 1982c)
Rice et al. studied lead contamination in the homes of secondary lead smelters. Homes of em-
ployees of secondary smelters in two separate geographic areas of the country were examined to
determine whether those homes had a greater degree of lead contamination than homes of workers
in the same area not exposed to lead. Both sets of homes (area I and II) were examined at the
same time of the year.
Thirty-three homes of secondary smelter employees were studied; 19 homes in the same or
similar neighborhoods were studied as controls. Homes studied were in good condition and were
one- or two-family dwellings. Blood lead levels were not obtained for children in these
homes. In the homes of controls, a detailed occupational history was obtained for each
employed person. Homes where one or more residents were employed in a lead-contaminated
environment were excluded from the analysis.
House dust samples were collected by Vostal's method and were analyzed for lead by AAS
In one of the areas, samples of settled dust were collected from the homes of employees and
controls. Dust was collected over the doorways. In homes where the settled dust was collec-
ted, zinc protoporphyrin (ZPP) determinations were made in family members of the lead workers
and in the controls.
In both areas, the wipe samples were statistically significantly higher in the homes of
employees compared to controls (geometric mean 79.3 ± 61.8 ug/g versus 28.8 ± 7.4 ug/g Area I;
112.0 ± 2.8 ug/g versus 9.7 ± 3.9 ug Area II). No significant differences were found between
workers' homes or controls between Area I and Area II. Settled dust lead was significantly
11-176
-------
higher in the homes of employees compared to controls (3300 versus 1200 ug/g). Lead contents
of participate matter collected at the curb and of paint chips collected in the home were not
significantly different between employee homes and controls. Zinc protoporphyrin determina-
tions were done on 15 children, 6 years or younger. ZPP levels were higher in employee chil-
dren than in control children. Mean levels were 61.4 ug/ml and 37.6 ug/ml, respectively.
It should be noted again that the wipe samples were not different between employee homes
in the two areas. Interviews with employees indicated that work practices were quite similar
in the two areas. Most workers showered and changed before going home. Work clothes were
washed by the company. Obviously, much closer attention needs to be paid to other potential
sources of lead introduction into the home (e.g., automobile surfaces).
From Mexico (Molina-Ballesteros et al., 1983) comes a report of yet another occupation
which can contribute to the lead burden of children whose parents work in settings contami-
nated by lead. One hundred and fifty-three children belonging to pottery-making families with
home workshops were studied, as well as 80 randomly selected children serving as controls.
Venipuncture blood samples were collected and analyzed by atomic absorption spectrophotometry.
Mean blood lead levels were 15 ug/dl higher for children whose parents had the home pottery
workshops than for control children. The mean blood lead level in the exposed children was
39.5 M9/d"l, which indicates a high degree of lead absorption in these children.
11.5.5 Miscellaneous Studies
11.5.5.1 Studies Using Indirect Measures of Air Exposure.
11.5.5.1.1 Studies in the United States. A 1973 Houston study examined the blood lead levels
of parking garage attendants, traffic policemen, and adult females living near freeways
(Johnson et al., 1974). A control group for each of the three exposed populations was selec-
ted by matching for age, education, and race. Unfortunately, the matching was not altogether
successful; traffic policemen had less education than their controls, and the garage employees
were younger than their controls. Females were matched adequately, however. It should be
noted that the mean blood lead values for traffic policemen and parking garage attendants, two
groups regularly exposed to higher concentrations of automotive exhausts, were significantly
higher than the means for their relevant control groups. Statistically significant differ-
ences in mean values were not found, however, between women living near a freeway, and control
women living at greater distances from the freeway.
A study of the effects of lower-level urban traffic densities on blood lead levels was
undertaken in Dallas, Texas, in 1976 (Johnson et al., 1978). The study consisted of two
phases. One phase measured air lead values for selected traffic densities and conditions,
ranging from equal to or less than 1,000 to about 37,000 cars/day. The second phase consisted
11-177
-------
of an epideim'ological study of traffic density and blood lead levels among residents. Figure
11-31 shows the relationship between arithmetic means of air lead and traffic density. As can
be seen from the graph, a reasonable fit was obtained.
^
"01
Z
LU
u
o
o
Q
2.0
I I I I I
Y = 0.6598 + 0.0263 X
X = TRAFFIC COUNT/1000
0 4.000 8.00012.00016,00020,00024.00028,00032.00036,00038,000
TRAFFIC VOLUME, cars/day
Figure 11-31. Arithmetic mean of air lead levels by traffic
volume. Dallas, 1976.
Source: Johnson et al. (1978).
In addition, for all distances measured (1.5-30.5 m from the road), air lead concentra-
tions declined rapidly with distance from the street. At 15 m, concentrations were about 55
percent of the street concentrations. In air lead collections from 1.5 to 30.5 m from the
street, approximately 50 percent of the airborne lead was in the respirable range (<1 pm), and
the proportions in each size class remained approximately the same as the distance from the
street increased.
Soil lead concentrations were higher in areas with greater traffic density, ranging from
73.6 pg/g at less than 1,000 cars per day to a mean of 105.9 at greater than 19,500 cars per
day. The maximum soil level obtained was 730 ug/g. Dustfall samples for 28 days from nine
locations showed no relationship to traffic densities, but outdoor levels were at least 10
times the indoor concentration in nearby residences.
11-178
-------
In the second phase, three groups of subjects, 1 to 6 years old, 18 to 49 years old, and
50 years and older, were selected in each of four study areas. Traffic densities selected
were less than 1,000, 8,000-14,000, 14,000-20,000, and 20,000-25,000 cars/day. The study
groups averaged about 35 subjects, although the number varied from 21 to 50. The smallest
groups were from the highest traffic density area. No relationship between traffic
density and blood lead levels in any of the age groups was found (Figure 11-32). Blood lead
levels were significantly higher in children, 12-18 ug/dl, than in adults, 9-14 pg/dl.
Caprio et al. (1974) compared blood "lead levels and proximity to major traffic arteries
in a study reported in 1971 that included 5226 children in Newark, New Jersey. Over 57 per-
cent of the children living within 30.5 m of roadways had blood lead levels greater than 40
ug/dl. For those living between 30.5 and 61 m from the roadways, more than 27 percent had
such levels, and at distances greater than 61 m, 31 percent exceeded 40 ug/dl. The effect of
automobile traffic was seen only in the group that lived within 30.5 m of the road.
No other sources of lead were considered in this study. However, data from other studies
on mobile sources indicate that it is unlikely that the blood lead levels observed in this
study resulted entirely from automotive exhaust emissions.
In 1964, Thomas et al. (1967) investigated blood lead levels in 50 adults who had lived
for at least 3 years within 76 m of a freeway (Los Angeles) and those of 50 others who had
lived for a similar period near the ocean or at least 1.6 km from a freeway. Mean blood lead
levels for those near the freeway were 22.7 ±5.6 for men and 16.7 ± 7.0 ug/dl for women.
These concentrations were higher than for control subjects living near the ocean: 16.0 ±8.4
ug/dl for men and 9.9 ± 4.9 ug/dl for women. The higher values, however, were similar to
those of other Los Angeles populations. Measured mean air concentrations of lead in Los
Angeles for October, 1964, were as follows: 12.25 ± 2.70 ug/m3 at a location 9 m from the San
Bernardino freeway; 13.25 ± 1.90 ug/m3 at a fourth-floor location 91.5 m from the freeway; and
4.60 ± 1.92 ug/1"3 1-6 km from the nearest freeway. The investigators concluded that the dif-
ferences observed were consistent with coastal inland atmospheric and blood lead gradients in
the Los Angeles basin and that the effect of residential proximity to a freeway (7.6-76 m) was
not demonstrated.
Ter Haar and Chadzynski report a study of blood lead levels of children living near three
heavily travelled streets in Detroit (Ter Haar, 1981; Ter Haar and Chadzynski, 1979). Blood
lead levels were not found to be related to distance from the road but were related to condi-
tions of housing and age of the child after multiple regression analyses.
11.5.5.1-2 British studies. In a Birmingham, England, study, mean blood lead levels in 41
males and 58 females living within 800 m of a highway interchange were 14.41 and 10.93 ug/dl,
respectively, just before the opening of the interchange in May, 1972 (Waldron, 1975). From
11-179
-------
25
Z
O
<
c
t-
LU
O
o
o
o
8
CO
20
15
10
MALES<9
«o
MALES>49
FEMALES 19-49
rs
FEMALES >49
I
I
< 1,000 1,00013.500 13.500 19.500-
19,500 38,000
TRAFFIC DENSITY, cars/day
Figure 11-32. Blood lead concentration and traffic density by
sex and age, Dallas, 1976.
Source: Johnson et al. (1978).
11-180
-------
October, 1972, to February, 1973, the respective values for the same individuals were 18.95
and 14.93 ug/dl. In October, 1973, they were 23.73 and 19.21 pg/dl. The investigators noted
difficulties in the blood collection method during the baseline period and changed from capil-
lary to venous blood collection for the remaining two sets of samples. To interpret the
significance of the change in blood collection method, some individuals gave both capillary
and venous blood at the second collection. The means for both capillary and venous bloods
were calculated for the 18 males and 23 females who gave both types of blood samples (Barry,
1975). The venous blood mean values for both these males and females were lower by 0.8 and
0.7 ug/dl , respectively. If these differences were applied to the means of the third series,
the mean for males would be reduced to 24.8 (jg/dl and that for the females to 18.7 ug/dl.
These adjusted means still show an increase over the means obtained for the first series.
Comparing only the means for venous bloods, namely series two and three, again shows an in-
crease for both groups. The increase in blood lead values was larger than expected following
the model of Knelson et al. (1973), because air lead values near the road were approximately I
ug/m3. The investigators concluded that either the lead aerosol of very small particles
behaved more like a gas so that considerably more than 37 percent of inhaled material was
absorbed, or that ingestion of lead-contaminated dust might be responsible.
Studies of taxicab drivers have employed different variables to represent the drivers'
lead exposure (Flindt et al., 1976; Jones et al., 1972): one variable was night versus day-
shift drivers (Jones et al., 1972); the other was mileage driven (Flindt et al., 1976). No
difference was observed, in either case.
The studies reviewed show that automobiles produce sufficient emissions to increase air
and nearby soil concentrations of lead as well, and to increase blood lead concentrations in
children and adults. The problem is of greater importance when houses are located within
100 ft (30 m) of the roadway.
11.5.5.2 Miscellaneous Sources of Lead. The habit of cigarette smoking is a source of lead
exposure. Shaper et al. (1982) report that blood lead concentration is higher for smokers
than nonsmokers and that cigarette smoking makes a significant independent contribution to
blood lead concentration in middle-aged men in British towns. A direct increase in lead in-
take from cigarettes is thought to be responsible. Hopper and Mathews (1983) comment that
current smoking has a significant effect on blood lead level, with an average increase of 5.8
percent in blood lead levels for every 10 cigarettes smoked per day. They also report that
past smoking history had no measurable effect on blood lead levels. Hasselblad and Nelson
(1975) report an average increase in women's blood lead levels of 1.3 ug/dl for smokers com-
pared to nonsmokers in the study of Tepper and Levin (1975).
11-181
-------
Although no studies are available, it is conceivable that destruction of lead-containing
plastics (to recover copper), which has caused cattle poisoning, also could become a source of
lead exposure for humans. Waste disposal is a more general problem because lead-containing
materials may be incinerated and may thus contribute to increased air lead levels. This
source of lead has not been studied in detail. Tyrer (1977) cautions of the lead hazard in
the recycling of waste.
The consumption of illicitly distilled liquor has been shown to produce clinical cases of
lead poisoning. Domestic and imported earthenware (De Rosa et al., 1980) with improperly
fired glazes have also been related to clinical lead poisoning. This source becomes important
when foods or beverages high in acid are stored in earthenware containers, because the acid
releases lead from the walls of the containers.
Particular cosmetics, popular among some Oriental and Indian ethnic groups, contain high
percentages of lead that sometimes are absorbed by users in quantities sufficient to be toxic.
Ali et al. (1978) and Attenburrow et al. (1980) discuss the practice of surma and lead poison-
ing. In addition to lead-containing cosmetics causing lead poisoning, folk remedies have also
been linked to lead poisoning (U. S. Centers for Disease Control, 1983a,b). Two Mexican folk
remedies, Azarcon and greta, have been implicated as causing lead poisoning in children (U.S.
Centers for Disease Control, 1983a). These products have a high lead content (70-90 percent)
and are primarily lead tetroxide and lead oxide for Azarcon and greta, respectively. There
have been a minimum of 15 reported cases of lead poisoning associated with these products. A
survey of Mexican-Hispanics living in Los Angeles estimated that 7.1-21.1 percent of Mexican-
Hispanic households had at some time used these products.
A folk medicine used by Hmong refugees from Northern Laos has also been implicated in
lead poisoning of children (U.S. Centers for Disease Control, 1983b). The product, "pay-loo-
ah," has a variable composition and texture, making control more difficult. Other sources of
lead are presented in Table 11-71.
TABLE 11-71. SOURCES OF LEAD
Source References
Gasoline sniffing Kaufman and Wiese (1978)
Coodin and Boeckx (1978)
Hansen and Sharp (1978)
Colored gift wrapping Bertagnolli and Katz (1979)
Gunshot wound Dillman et al. (1979)
Drinking glass decorations Anonymous (1979)
Electric kettles Wigle and Charlebois (1978)
Hair dye Searle and Harnden (1979)
Snuff use Filippini and Simmler (1980)
Firing ranges Fischbein et al. (1979, 1980b)
Glazed pottery Acra et al. (1981)
===== 11-182 -= ~=
-------
11.6 SUMMARY AND CONCLUSIONS
Using the bones and teeth of ancient populations, studies show that levels of internal
exposures of lead today are substantially elevated over past levels. Studies of current
populations living in remote areas far from urbanized cultures show blood lead levels in the
range of 1-5 |jg/dl. In contrast to the blood lead levels found in remote populations, data
from current U.S. populations have geometric means ranging from <10 to 20 ug/dl depending on
age, race, sex, and degree of urbanization. These higher current exposure levels appear to be
associated with industrialization and widespread commercial use of lead, e.g., in gasoline
combustion.
Age appears to be one of the single most important demographic covariates of blood lead
levels. Blood lead levels in children up to six years of age are generally higher than those
in non-occupationally exposed adults. Children aged two to three years tend to have the high-
est levels, as shown in Figure 11-33. Blood lead levels in non-occupationally exposed adults
may increase slightly with age due to skeletal lead accumulation.
Sex has a differential impact on blood lead levels depending on age. No significant dif-
ferences exist between males and females less than seven years of age. Males above the age of
seven generally have higher blood lead levels than females.
Race also plays a role, in that blacks generally have higher blood lead levels than
either whites or Hispanics and urban black children (aged 6 months-5 years) have markedly
higher blood lead concentrations than any other racial or age group. Possible genetic factors
associated with race have yet to be fully untangled from differential exposure levels and
other factors as important determinants of blood lead levels.
Blood lead levels also generally increase with degree of urbanization. Data from NHANES
II show blood lead levels in the United States, averaged over 1976-1980, increasing from a
geometric mean of 11.9 ug/dl in rural populations to 12.8 ug/dl in urban populations of less
than one million, and increasing again to 14.0 ug/dl in urban populations of one million or
more.
Blood lead levels, examined on a population basis, have similarly skewed distributions.
Blood lead levels, from a population thought to be homogeneous in terms of demographic and
lead exposure characteristics, approximately follow a lognormal distribution. The geometric
standard deviations, an estimation of dispersion, for four different studies are shown in
Table 11-72. The values, including analytic error, are about 1.4 for children and possibly
somewhat smaller for adults. This allows an estimation of the upper tail of the blood lead
distribution, the group at higher risk. A somewhat larger geometric standard deviation of
1.42 may be derived from the NHANES II study when only gasoline and industrial air lead
emission exposures are assumed to be controllable sources of variation.
11-183
-------
40
35 U-
30
01
cT
<
25
CD
20
15
IDAHO STUDY
NEW YORK SCREENING - BLACKS
NEW YORK SCREENING - WHITES
NEW YORK SCREENING - HISPANICS
NHANES II STUDY - BLACKS
NHANES II STUDY - WHITES
\
\
o*
d
j j
5
AGE, yr
10
Figure 11-33. Geometric mean blood lead levels by race and age for younger children in the
NHANES II study, and the Kellogg/Silver Valley and New York Childhood Screening Studies.
11-184
-------
TABLE 11-72. SUMMARY OF BLOOD LEAD POOLED GEOMETRIC STANDARD DEVIATIONS
AND ESTIMATED ANALYTIC ERRORS
Pooled geometric standard deviations
Study
NHANES II
N.Y. Childhood
Inner city
black children
1.37a
1.41
Inner city
white children
1.39a
1.42
Adult
females
1.36b
Adult
males
1.40b
Estimated
analytic
error
0.021
_c
Screening Study
Tepper-Leven
Azar et al.
1.30
1.29
0.056°
0.042d
Note: To calculate an estimated person-to-person GSD, compute Exp [((In(GSD))2 -
1/2-1
Analytic Error)
aA geometric standard deviation of 1.42 may be derived when only gasoline and industrial
air lead emission exposures are assumed to be controllable sources of variability.
Pooled across areas of differing urbanization
cNot known, assumed to be similar to NHANES II
dTaken from Lucas (1981).
Recent U.S. blood lead levels show a downward temporal trend occurring consistently
across race, age, and geographic location. The downward pattern commenced in the early part
of the 1970's and has continued into 1980. The downward trend has occurred from a shift in
the entire distribution and not through a truncation in the high blood lead levels. This con-
sistency suggests a general causative factor, and attempts have been made to identify the
causative element. Reduction in lead emitted from the combustion of leaded gasoline is a prime
candidate.
Studies of data from blood lead screening programs (i.e., New York City) suggest that the
downward trend in blood lead levels noted earlier is due to the reduction in air lead levels,
which has been attributed to the reduction of lead in gasoline. The NHANES II analysis found
a highly significant association between the declining blood lead concentrations for the over-
all U.S. population and decreasing amounts of lead used in gasoline in the United States
during the same time period. Two studies used isotope ratios of lead to estimate the relative
proportion of lead in the blood coming from airborne lead. From one study, by Manton, it can
be estimated that between 7 and 41 percent of the blood lead in study subjects in Dallas
11-185
-------
resulted from airborne lead. Additionally, these data provide a means of estimating the in-
direct contribution of air lead to blood lead. By one estimate, only 10 - 20 percent of the
total airborne contribution in Dallas is from direct inhalation.
From the ILE data in Facchetti and Geiss (1982) and Facchetti (1985), as shown in Table
11-73, the direct inhalation of air lead may account for 60 percent of the total adult blood
lead uptake from leaded gasoline in a large urban center, but inhalation is a much less impor-
tant pathway in suburban parts of the region (19 percent of the total gasoline lead contribu-
tion) and in the rural parts of the region (9 percent of the total gasoline lead contribu-
tion). EPA analyses of the preliminary results from the ILE study separated the inhalation
and non-inhalation contributions of leaded gasoline to blood lead into the following three
parts: (1) an increase of about 1.7 |jg/dl in blood lead per (jg/m3 of air lead, attributable
to direct inhalation of the combustion products of leaded gasoline; (2) a sex difference of
about 2 ug/dl attributable to lower exposure of women to indirect (non-inhalation) pathways
for gasoline lead; and (3) a non-inhalation background attributable to indirect gasoline lead
pathways, such as ingestion of dust and food, increasing from about 2 ug/dl in Turin to 3
|jg/dl in remote rural areas. The non-inhalation background represents only two to three years
of environmental accumulation at the new experimental lead isotope ratio. It is not clear how
to numerically extrapolate these estimates to U.S. subpopulations; but it is evident that even
in rural and suburban parts of a metropolitan area, the indirect (non-inhalation) pathways for
exposure to leaded gasoline make a significant contribution to blood lead. This can be seen in
Table 11-73. It should also be noted that the blood lead isotope ratio responded fairly
rapidly when the lead isotope ratio returned to its pre-experimental value, but it is not yet
possible to estimate the long-term change in blood lead attributable to persistent exposures
to accumulated environmental lead.
The strongest kind of scientific evidence about causal relationships is based on an ex-
periment in which all possible extraneous factors are controlled. The evidence derived from
the Isotopic Lead Experiment (ILE) comes very close to this ideal. The experimental inter
vention consisted of replacing the normal 206Pb/207Pb isotope ratio by a very different ratio.
There is no plausible mechanism by which other concurrent lead exposure variables (food, water
and beverages, paint, and industrial emissions) could have also changed their isotope ratios.
Hence the very large changes in isotope ratios in blood were responding to the change in
gasoline. There was no need to carry out detailed aerometric and ecological modeling to track
the leaded gasoline isotopes through the various environmental pathways. In fact, our analy-
ses (Section 11.3.6.2.1) show that consideration of inhalation of community air lead alone
will substantially under estimate the total effect of gasoline lead, at least in the 35 sub-
jects whose blood leads were tracked in the ILE Preliminary Study. This may be partially
explained by the differences in the lead concentration measured by stationary monitors
11-186
-------
TABLE 11-73. ESTIMATED CONTRIBUTION OF LEADED GASOLINE TO BLOOD LEAD
BY INHALATION AND NON-INHALATION PATHWAYS
Location
Turin
<25 km
>25 km
Air lead
fraction
from
gasoline3
0.873
0.587
0.587
Blood
lead
fraction
from
i b
gasol me
0.214
0.114
0.101
Blood
lead
from
gasol ipe
in air
(M9/dl)
2.79
0.53
0.28
Blood lead
not inhaled
from .gaso-
line0
(ug/dl)
1.88
2.33
2.93
Estimated
fraction
gas- lead
inhalation6
0.60
0.19
0.09
^Fraction of air lead in Phase 2 attributable to lead in gasoline.
Mean fraction of blood lead in Phase 2 attributable to lead in gasoline.
^Estimated blood lead from gasoline inhalation = p x a x b, p = 1.6.
Estimated blood lead from gasoline, non-inhalation = f-e.
fraction of blood lead uptake from gasoline attributable to direct inhalation = f/e.
Source: Facchetti and Geiss (1982), pp. 52-56; Facchetti (1985).
compared to those that would be measured by personal monitors, expecially if higher exposures
occur in certain microenvironments. Diet lead is also an explanation for the large excess of
gasoline lead isotope ratio in blood beyond that expected from inhalation of ambient air lead,
both from gasoline lead entering the food chain and added by food processing and preparation.
The subjects in the ILE study cannot be said to represent some defined population, and it is
not clear how the results can be extended to U.S. populations. Turin's unusual meteorology,
high lead levels, and "reversed" urban-rural gradient of the subjects in the ILE study indi-
cate the need for future research. But in spite of the variable gasoline lead exposures of
the subjects, there is strong evidence that changes in gasoline lead produce large changes in
blood lead.
Because the main purpose of this chapter is to examine relationships of lead in air and
lead in blood under ambient conditions, the results of studies most appropriate to this area
have been emphasized. A summary of the most appropriate studies appears in Table 11-74. At
air lead exposures of 3.2 ug/m3 or less, there is no statistically significant difference be-
tween curvilinear and linear blood lead inhalation relationships. At air lead exposures of 10
pg/rn3 or more, either nonlinear or linear relationships can be fitted. Thus, a reasonably
consistent picture emerges in which the blood lead to air lead relationship by direct inhala-
tion was approximately linear in the range of normal ambient exposures of 0.1-2.0 ug/m3 (as
discussed in Chapter 7). Differences among individuals in a given study (and among several
11-187
-------
TABLE 11-74. SUMMARY OF BLOOD INHALATION SLOPES, (B)
pg/dl per (jg/m3
Population
Children
Children
(P)
Study Slope, Model sensitivity
Study type N (jg/dl per ug/m3 of slope*
Angle and Population 1074 1.92 (1.40 - 4 40)a>b'c
Mclntire, 1979
Omaha, NE
Roels et al . Population 148 2.46 (1.55 - 2 46)a>b
(1980)
Belgium
Children
Adult males
Adult males
Yankel et al. Population
(1977); Walter
et al. (1980)
Idaho
Azar et al.
(1975). Five
groups
Griffin et al.
(1975), NY
prisoners
Population
Experiment
879
149
43
1.52
1.32
1.75
(1.07 - 1.52)a'b>c
(1.08 - 2.39)b'C
(1.52 - 3.38)d
Adult males
Adult males
Gross
(1979)
Rabinowitz et
al. (1973,1976,
1977)
Experiment 6 1.25
Experiment 5 2.14
l
(1.25 -
(2.14 -
1.55)b
3.51)e
^Selected from among the most plausible statistically equivalent models.
models, slope at 1.0 (jg/m3.
Sensitive to choice of other correlated predictors such as dust and soil
bSensitive to linear versus nonlinear at low air lead.
Sensitive to age as a covariate.
^Sensitive to baseline changes in controls.
Sensitive to assumed air lead exposure.
For nonlinear
lead.
11-188
-------
studies) are large, so that pooled estimates of the blood lead inhalation slope depend upon
the weight given to various studies. Several studies were selected for analysis, based upon
factors described earlier. EPA analyses* of experimental and clinical studies (Griffin et
al., 1975; Rabinowitz et al., 1974, 1976, 1977; Kehoe 1961a,b,c; Gross, 1981; Hammond et al.,
1981) suggest that blood lead in adults increases by 1.64 ± 0.22 ug/dl from direct inhalation
of each additional M9/n>3 of air lead. EPA analysis of Azar's population study (Azar et al.
1975) yields a slope of 1.32 ± 0.38 for adult males. EPA analyses of population studies
(Yankel et al., 1977; Roels et al., 1980; Angle and Mclntire, 1979) suggest that, for chil-
dren, the median blood lead increase is 1.97 ug/dl per ug/m3 for inhaled air lead.
These slope estimates are based on the assumption that an equilibrium level of blood lead
is achieved within a few months after exposure begins. This is only approximately true, since
lead stored in the skeleton may return to blood after some years. Chamberlain et al. (1978)
suggest that long-term inhalation slopes should be about 30 percent larger than these estima-
tes. Inhalation slopes quoted here are associated with a half-life of blood lead in adults of
about 30 days. 0'Flaherty et al. (1982) suggest that the blood lead half-life may increase
slightly with duration of exposure, but this has not been confirmed (Kang et al., 1983).
One possible approach would be to regard all inhalation slope studies as equally infor-
mative and to calculate an average slope using reciprocal squared standard error estimates as
weights. This approach has been rejected for two reasons. First, the standard error estima-
tes characterize only the internal precision of an estimated slope, not its representativeness
(i.e., bias) or predictive validity. Secondly, experimental and clinical studies obtain more
information from a single individual than do population studies. Thus, it may not be appro-
priate to combine the two types of studies.
Estimates of the inhalation slope for children are only available from population
studies. The importance of dust ingestion as a non-inhalation pathway for children is estab-
lished by many studies. A pooled slope estimate, 1.97 ± 0.39, has been derived for air lead
inhalation based on those studies (Angle and Mclntire, 1979; Roels et al., 1980; Yankel et
al., 1977) from which the air inhalation and dust ingestion contributions can both be esti-
mated. Aggregate analyses of data from these and several other studies typically yield slope
estimates in the range of 3-5 for the combined impact of both direct (inhaled) and indirect
(via dust, etc.) contributions of air lead to blood lead in children.
*Note: The term EPA analyses refers to calculations done at EPA. A brief discussion of the
methods used is contained in Appendix 11-B; more detailed information is available at EPA
upon request.
11-189
-------
While direct inhalation of air lead is stressed, this is not the only air lead contribu-
tion that needs to be considered. Smelter studies allow partial assessment of the air lead
contributions to soil, dust, and finger lead. Conceptual models allow preliminary estimation
of the propagation of lead through the total food chain as shown in Chapter 7. Useful mathe-
matical models to quantify the propagation of lead through the food chain need to be devel-
oped. The direct inhalation relationship does provide useful information on changes in blood
lead as responses to changes in air lead on a time scale of several months. The indirect
pathways through dust and soil and through the food chain may thus delay the total blood lead
response to changes in air lead, perhaps by one or more years. The Italian ILE study facili-
tates partial assessment of this delayed response from leaded gasoline as a source.
Dietary absorption of lead varies greatly from one person to another and depends on the
physical and chemical form of the carrier, on nutritional status, and on whether lead is in-
gested with food or between meals. These distinctions are particularly important for consump-
tion by children of leaded paint, dust, and soil. Typical values of 10 percent absorption of
ingested lead into blood have been assumed for adults and 25 to 50 percent for children.
It is difficult to obtain accurate dose-response relationships between blood lead levels
and lead levels in food or water. Dietary intake must be estimated by duplicate diets or
fecal lead determinations. Water lead levels can be determined with some accuracy, but the
varying amounts of water consumed by different individuals add to the uncertainty of the esti-
mated relationships.
Quantitative analyses relating blood lead levels and dietary lead exposures have been re-
ported. Studies on infants provide estimates that are in close agreement. Only one indi-
vidual study is available for adults (Sherlock et al. 1982); another estimate from a number of
pooled studies is also available. These two estimates are in good agreement. Most of the
subjects in the Sherlock et al. (1982) and United Kingdom Central Directorate on Environmental
Pollution (1982) studies received quite high dietary lead levels (>300 ug/day). The fitted
cube root equations give high slopes at lower dietary lead levels. On the other hand, the
linear slope of the United Kingdom Central Directorate on Environmental Pollution (1982) study
is probably an underestimate of the slope at lower dietary lead levels. For these reasons,
the Ryu et al. (1983) study is the most believable, although it only applies to infants and
also probably underestimates to some extent the value of the slope. Estimates for adults
should be taken from the experimental studies or calculated from assumed absorption and half-
life values. Most of the dietary intake supplements were so high that many of the subjects
had blood lead concentrations much in excess of 30 ug/dl for a considerable part of the ex-
periment. Blood lead levels thus may not completely reflect lead exposure, due to the
previously noted nonlinearity of blood lead response at high exposures. The slope estimates
11-190
-------
for adult dietary intake are about 0.02 ug/dl increase in blood lead per ng/day intake, but
consideration of blood lead kinetics may increase this value to about 0.04. Such values are a
bit lower than slopes of about 0.05 ug/dl per ug/day estimated from the population studies ex-
trapolated to typical dietary intakes. The value for infants is larger.
The relation between blood lead and water lead is not clearly defined and is often de-
scribed as nonlinear. Water lead intake varies greatly from one person to another. It has
been assumed that children can absorb 25-50 percent of lead in water. Many authors chose to
fit cube root models to their data, although polynomial and logarithmic models were also used.
Unfortunately, the form of the model greatly influences the estimated contributions to blood
leads from relatively low water lead concentration.
Although there is close agreement in the quantitative analyses of the relationship bet-
ween blood lead level and dietary lead, there is a larger degree of variability in results of
the various water lead studies. The relationship is curvilinear, but its exact form is yet to
be determined. At typical levels for U.S. populations, the relationship appears linear. The
only study that determines the relationship based on lower water lead values (<100 ng/1) is
the Pocock et al. (1983) study. The data from this study, as well as the authors themselves,
suggest that in this lower range of water lead levels, the relationship is linear. Further-
more, the estimated contributions to blood lead levels from this study are quite consistent
with the polynomial models from other studies. For these reasons, the Pocock et al. (1983)
slope of 0.06 is considered to represent the best estimate. The possibility still exists,
however, that the higher estimates of the other studies may be correct in certain situations,
especially at higher water lead levels (>100 ug/1).
Studies relating soil lead to blood lead levels are difficult to compare. The relation-
ship obviously depends on depth of soil lead, age of the children, sampling method, cleanli-
ness of the home, mouthing activities of the children, and possibly many other factors. Var-
ious soil sampling methods and sampling depths have been used over time, and as such they may
not be directly comparable and may produce a dilution effect of the major lead concentration
contribution from dust which is located primarily in the top 2 cm of the soil. Increases in
soil dust lead significantly increase blood lead in children. From several studies (Yankel et
al., 1977; Angle and Mclntire, 1979) EPA estimates an increase of 0.6-6.8 ug/dl in blood lead
for each increase of 1000 ug/g in soil lead concentration. Values of about 2.0 ug/dl per
1,000 |jg/g soil lead from the Stark et al. (1982) study may represent a reasonable median
estimate. The relationship of housedust lead to blood lead is difficult to obtain. House-
hold dust also increases blood lead, as children from the cleanest homes in the Silver Valley/
Kellogg Study had 6 ug/dl less lead in blood, on average, than those from the households with
the most dust.
11-191
-------
A number of specific environmental sources of airborne lead have been identified as
having a direct influence on blood lead levels. Primary lead smelters, secondary lead
smelters, and battery plants emit lead directly into the air and ultimately increase soil and
dust lead concentrations in their vicinity. Adults, and especially children, have been shown
to exhibit elevated blood lead levels when living close to these sources. Blood lead levels
in these residents have been shown to be related to air, as well as to soil or dust exposures.
The habit of cigarette smoking is a source of lead exposure. Other sources include the
following: lead based cosmetics, lead-based folk remedies, and glazed pottery.
11-192
-------
11.7 REFERENCES
Acra, A.; Dajanl, R.; Raffoul, Z.; Karahagoplan, Y. (1981) Lead-glazed pottery: a potential
health hazard in the Middle East. Lancet 1(8217): 433-434.
AHchlson, J.; Brown, J. A. C. (1966) The lognormal distribution: with special reference to
Its uses 1n economics. London, United Kingdom: Cambridge University Press. (University of
Cambridge, Department of Applied Economics Monographs: v. 5).
All, A. R.; Smales, 0. R. C.; Aslam, M. (1978) Surma and lead poisoning. Br. Med. J. 2(6142):
915-916.
Angle, C. R.; Mclntire, M. S. (1977) Is busing good for your blood lead? Presented at: 2nd an-
nual French-American congress on clinical and analytical toxicology; August; St. Adele,
Canada. Washington, DC: National Institute of Environmental Health Sciences.
Angle, C. R.; Mclntire, M. S. (1979) Environmental lead and children: the Omaha study. J.
Toxlcol. Environ. Health 5: 855-870.
Angle, C. R.; Mclntire, M. S. (1982) Children, the barometer of environmental lead. Adv.
Pedlatr. 27: 3-31.
Angle, C. R. ; Mclntire, M. S. ; Colucci, A. V. (1974) Lead in air, dustfall, soil, housedust,
milk and water: correlation with blood lead of urban and suburban school children. In:
Hemphill, D. D., ed. Trace substances in environmental health - VIII: [proceedings of
University of Missouri's 8th annual conference on trace substances in environmental
health]; June; Columbia, MO. Columbia, MO: University of Missouri-Columbia; pp. 23-29.
Angle, C. R.; Marcus, A.; Cheng, I.-H.; Mclntire, M. S. (1984) Omaha childhood blood lead and
environmental lead: a linear total exposure model. Environ. Res. 35: 160-170.
Annest, J. L. (1983) Trends in the blood lead levels of the U.S. population: the second
National Health and Nutrition Examination Survey (NHANES II) 1976-1980. In: Rutter, M.;
Russell Jones, R., eds. Lead versus health: sources and effects of low level lead ex-
posure. New York, NY: John Wiley and Sons; pp. 33-58.
Annest, J. L.; Mahaffey, K. (J.984) Blood lead levels for persons ages 6 months-74 years:
United States, 1976-80. Hyattsville, MD: U.S. Department of Health and Human Services,
National Center for Health Statistics; DHHS pub. no. (PHS) 84-1683. (Vital and Health
Statistics: series 11, no. 233). Available from: GPO, Washington, DC; S/N 017-022-
00846-7.
Annest, J. L.; Mahaffey, K. R.; Cox, D. H.; Roberts, J. (1982) Blood lead levels for persons 6
months-74 years of age: United States, 1976-80. Hyattsville, MD: U.S. Department of
Health and Human Services; DHHS pub no. (PHS) 82-1250. (Advance data from vital and
health statistics of the National Center for Health Statistics: no. 79).
Annest J. L.; Pirkle, J. L.; Makuc, D.; Neese, J. W.; Bayse, D. D.; Kovar, M. G. (1983a)
Chronological trend in blood lead levels between 1976 and 1980. N. Engl. J. Med. 308:
1373-1377.
11-193
-------
Annest, J. L.; Casady, R. J.; White, A. A. (1983b) The NHANES II study: analytic error and Its
effects on national estimates of blood lead levels: United States, 1976-80. Available for
inspection at: U.S. Environmental Protection Agency, Environmental Criteria Assessment
Office, Research Triangle Park, NC.
Anonymous. (1979) Lead and cadmium may leach from drinking glass decorations. J. Am Med
Assoc. 241: 544.
Ashford, N. A.; Gecht, R. D., Hattis, D. B.; Katz, J. I. (1977) The effects of OSHA medical
removal protection on labor costs of selected lead industries. Cambridge, MA: Massac-
husetts Institute of Technology, Center for Policy Alternatives; CPA report no
CPA-77/11. Available from: NTIS, Springfield, VA; PB-278653.
Attenburrow, A. A.; Campbell, S.; Logan, R. W.; Goel, K. M. (1980) Surma and blood lead levels
in Asian children in Glasgow. Lancet 1(8163): 323.
Aufderheide, A. C.; Neiman, F. D.; Wittmers, L. E. , Jr.; Rapp, G. (1981) Lead in bone. II:
Skeletal-lead content as an indicator of lifetime lead ingestion and the social cor-
relates in an archaeological population. Am. J. Phys. Anthro. 55: 285-291.
Azar, A.; Snee, R. D.; Habibi, K. (1975) An epidemiologic approach to community air lead ex-
posure using personal air samplers. In: Griffin, T. B.; Knelson, J. H,, eds. Lead. Stut-
tgart, West Germany: Georg Thieme Publishers; pp. 254-290. (Coulston, F.; Korte, F., eds.
Environmental quality and safety: supplement v. 2).
Baker, E. L., Jr.; Folland, D. S.; Taylor, T. A.; Frank, M.; Peterson, W.; Lovejoy, G.; Cox,
D.; Housworth, J.; Landrigan, P. J. (1977a) Lead poisoning in children of lead workers:
house contamination with industrial dust. N. Engl. J- Med. 296: 260-261.
Baker, E. L., Jr.; Hayes, C. G.; Landrigan, P. J.; Handke, J. L.; Leger, R. T.; Housworth, W.
J.; Harrington, J. M. (1977b) A nationwide survey of heavy metal absorption in children
living near primary copper, lead, and zinc smelters. Am. J. Epidemiol. 106: 261-273.
Barltrop, D. (1975) Significance of lead-contaminated soils and dusts for human populations.
Arh. High. Rada Toksikol. Suppl. 26: 81-96.
Barltrop, D.; Strehlow, C. D.; Thornton, I.; Webb, J. S. (1974) Significance of high soil lead
concentrations for childhood lead burdens. Environ. Health Perspect. 7: 75-82.
Barltrop, D.; Strehlow, C. D.; Thornton, I.; Webb. J. S. (1975) Absorption of lead from dust
and soil. Postgrad. Med. J. 51: 801-804.
Barry, P. S. I. (1975) Lead levels in blood. Nature (London) 258: 775.
Barry, P. S. I.; Connolly, R. (1981) Lead concentrations in mediaeval bones. Int. Arch. Occup.
Environ. Health 48: 173-177.
Batschelet, E. ; Brand, L. ; Steiner, A. (1979) On the kinetics of lead in the human body. J.
Math. Biol. 8: 15-23.
Beattie, A. D. ; Dagg, J. H.; Goldberg, A.; Wang, I.; Ronald, J. (1972a) Lead poisoning in
rural Scotland. Br. Med. J. 2(5812): 488-491.
11-194
-------
Beattle, A. D. ; Moore, M. R.; Devenay, W. T.; Miller, A. R.; Goldberg, A. (1972b) Environ-
mental lead pollution In an urban soft-water area. Br. Med. J. 2(5812): 491-493.
Becker, R. 0.; Spadaro, J. A.; Berg, E. W. (1968) The trace elements of human bone. J. Bone
Jt. Surg. 50A: 326-334.
Berger, H. W. (1973a) The NBS lead paint poisoning project: housing and other aspects.
Washington, DC: U.S. Department of Commerce, National Bureau of Standards; NBS technical
note 759.
Berger, H. W. (1973b) Final report, phase I lead paint survey sampling plan and preliminary
screening. Washington, DC: U.S. Department of Commerce, National Bureau of Standards; NBS
report no. 10958.
Bernard, S. R. (1977) Dos1metr1c data and metabolic model for lead. Health Phys. 32: 44-46.
BertagnolH, J. F.; Katz, S. A. (1979) Colored gift wrapping papers as a potential source of
toxic metals. Int. J. Environ. Anal. Chem. 6: 321-325.
Bmick, I. H. (1977) [Presentation to the U.S. Environmental Protection Agency, Scientific
Advisory Board Subcommittee on Scientific Criteria for Environmental Lead, October 7].
Copy supplied to U.S. Environmental Protection Agency, Environmental Criteria and Assess-
ment Office, Research Triangle Park, NC.
Blllick, I. H. (1982) Prediction of pedlatrlc blood lead levels from gasoline consumption
[submitted to docket for public hearing on lead phasedown proposed rulemaking, April 15.]
Available from: U.S. Environmental Protection Agency, Central Docket Section Washington,
DC; docket no. A-81-36; document no. IVA.4.
Blllick, I. H. (1983) Sources of lead 1n the environment. In: Rutter, M.; Russell Jones, R.,
eds. Lead versus health: sources and effects of low level lead exposure. New York, NY:
John Wiley and Sons, Ltd.; pp. 59-77.
BUHck, I. H. ; Curran, A. S.; Shier, D. R. (1979) Analysis of pediatHc blood lead levels in
New York City for 1970-1976. Environ. Health Perspect. 31: 183-190.
BUHck, I. H.; Curran, A. S.; Shier, D. R. (1980) Relation of pedlatrlc blood lead levels to
lead in gasoline. Environ. Health Perspect. 34: 213-217.
BilHck, I. H.; Shier, D. R.; Splegelman, C. H. (1982) Sensitivity of trends in geometric mean
blood levels to random measurement errors. Sci. Total Environ. 24: 233-248.
Bishop, L.; Hill, W. J. (1983) A study of the relationship between blood lead levels and occu-
pational air lead levels. Am. Stat. 37: 471-475.
Brunekreef, B. D. (1984) The relationship between air lead and blood lead in children: a
critical review. Sci. Total Environ. 38: 79-123.
Brunekreef, B.; Veenstra, S. J.; Biersteker, K.; Bolelj, J. S. M. (1981) The Arnhem lead
study: 1. lead uptake by 1- to 3-year-old children living 1n the vicinity of a secondary
lead smelter 1n Arnhem, The Netherlands. Environ. Res. 25: 441-448.
11-195
-------
Brunekreef, B.; Noy, 0.; Biersteker, K.; Boleij, J. (1983) Blood lead levels of Dutch city
children and their relationship to lead in the environment. J. A1r Pollut. Control Assoc
33: 872-876.
Caprio, R. J.; Margulls, H. L.; Joselow, M. M. (1974) Lead absorption 1n children and Its
relationship to urban traffic densities. Arch. Environ. Health 28: 195-197.
Cavalleri, A.; Baruffini, A.; Minoia, C.; Bianco, L. (1981) Biological response of children to
low levels of inorganic lead. Environ. Res. 25: 415-423.
Chamberlain, A. C. (1983) Effect of airborne lead on blood lead. Atmos. Environ. 17: 677-692.
Chamberlain, A. C. ; Heard, M. J. (1981) Lead tracers and lead balances. In: Lynam, D. R.
Piantanida, L. G.; Cole, J. F., eds. Environmental lead: proceedings of the second inter-
national symposium on environmental lead research; December 1978; Cincinnati, OH New
York, NY: Academic Press; pp. 175-198. (Coulston, F.; Korte, F., eds. Ecotoxicology and
environmental quality series).
Chamberlain, A. C.; Clough, W. S.; Heard, M. J.; Newton, D.; Stott, A. N. B.; Wells, A. C
(1975a) Uptake of inhaled lead from motor exhaust. Postgrad. Med. J. 51: 790-794.
Chamberlain, A. C. ; Clough, W. S.; Heard, M. J-; Newton, D.; Stott, A. N. B.; Wells, A. C.
(1975b) Uptake of lead by inhalation of motor exhaust. Proc. R. Soc. London Ser. B 192-
77-110.
Chamberlain, A. C.; Heard, M. J.; Little, P.; Newton, D.; Wells, A. C.; Wlffen, R. D. (1978)
Investigations into lead from motor vehicles. Harwell, United Kingdom: United Kingdom
Atomic Energy Authority; report no. AERE-R9198.
Charney, E.; Sayre, J. ; Coulter, M. (1980) Increased lead absorption in Inner city children:
where does the lead come from? Pediatrics 65: 226-231.
Charney, E. ; Kessler, B. ; Farfel, M. ; Jackson, D. (1983) Childhood lead poisoning: a con-
trolled trial of the effect of dust-control measures on blood lead levels. N. Engl. J.
Med. 309: 1089-1093.
Colombo, A. (1985) The underdefined nature of the blood lead-air lead relationship from bio-
logical and statistical grounds. Atmos. Environ. 19: 1485-1493.
Colombo, A.; Fantechi, R. (1983) Isotoplc lead experiment: a dynamic analysis of the Isotopic
lead experiment results. Luxembourg: Commission of the European Communities.
Coodln, F. J. ; Boeckx, R. (1978) Lead poisoning from sniffing gasoline [letter]. N. Engl. J
Med. 298: 347.
Cools, A.; Salle, H. J. A.; Verberk, M. M.; Zlelhuis, R. L. (1976) Biochemical response of
male v'olunteers ingesting inorganic lead for 49 days. Int. Arch. Occup. Environ. Health
38: 129-139.
Crawford, M. D.; Crawford, T. (1969) Lead content of bones in a soft and a hard water area.
Lancet 1(7597): 699-701.
11-196
-------
Creason, J. P.; Hlnners, T. A.; Bumgarner, J. E.; Pinkerton, C. (1975) Trace elements 1n hair,
as related to exposure 1n metropolitan New York. CUn. Chem. (W1nston-Salem, NC) 21:
603-612.
Cuddlhy, R. G.; McClellan, R. 0.; Griffith, W. C. (1979) Variability 1n target organ deposi-
tion among Individuals exposed to toxic substances. Toxlcol. Appl. Pharmacol. 49:
179-187.
Dalnes, R. H.; Smith, 0. W.; Felldano, A.; Trout, J. R. (1972) A1r levels of lead inside and
outside of homes. Ind. Med. Surg. 41: 26-28.
Davidson, C. I.; Grimm, T. C. Nasta, M. A. (1981) Airborne lead and other elements derived
from local fires 1n the Himalayas. Science (Washington, DC) 214: 1344-1346.
De Graeve, J.; Rondla, D.; Jamln, P. (1975) Concentration du plomb dans 1'eau potable et
plombemle d'une population adulte [Lead concentration 1n drinking water and plumbism 1n
an adult population]. In: Recent advances 1n the assessment of the health effects of
environmental pollution: proceedings, International symposium, vol. 2; June 1974; Paris,
France. Luxembourg: Commission of the European Communities; pp. 523-535.
De Rosa, E.; Gobbato, F. (1970) Ep1dem1a d1 saturnlsmo non professional per Inquinamento da
effluentl 1ndustr1al1 [Non-professional saturnism: a pollution epidemic from Industrial
emission]. Ig. Mod. 63: 472-484.
De Rosa, E.; Rossi, A.; Toffolo, D.; Br1ghent1, F. ; Rosa, A.; Caroldl, S. (1980) The ceramics
Industry and lead poisoning: long-term testing. Scand. J. Work Environ. Health 6:
312-315.
Dequldt, J.; Vaast, D.; Lespagnol, A. (1971) Rlsques d1Impregnation saturnine au volslnage
d'uslnes de traUement du plomb [Saturnism hazards In lead processing plant areas].
Pollut. Atmos. 13: 289-292.
Dlemel, J. A. L.; Brunekreef, B.; Bolelj, J. S. M.; Blersteker, K.; Veenstra, S. J. (1981) The
Arnhem lead study: II. Indoor pollution, and Indoor/outdoor relationships. Environ. Res.
25: 449-456.
DUlman, R. 0.; Crumb, C. K.; Lldsky, M. J. (1979) Lead poisoning from a gunshot wound: report
of a case and review of the literature. Am. J. Med. 66: 509-514.
Djurid, D.; Kerln, Z.; Graovac-Leposavld, L. ; Novak, L.; Kop, M. (1971) Environmental con-
tamination by lead from a mine and smelter. Arch. Environ. Health 23: 275-279.
DjuHd, D.; Graovac-Leposavld, L.; M1l1d, S.; Senlcar, L. (1972) Lead contamination of Meza
Valley: lead mobilization with Ca-Na2-EDTA from the body of inhabitants. Prac. Lek.
(Praque) 24: 49-50.
Dolcourt, J. L.; HamHck, H. J.; O'Tuama, L. A.; Wooten, J.; Barker, E. L., Jr. (1978) In-
creased lead burden 1n children of battery workers: asymptomatic exposure resulting from
contaminated work clothing. Pediatrics 62: 563-566.
Dolcourt, J. L.; Finch, C. ; Coleman, G. D. ; KHmas, A. J.; M1lar, C. R. (1981) Hazard of lead
exposure 1n the home from recycled automobile storage batteries. Pediatrics 68: 225-230.
Duggan, M. J- (1980) Lead in urban dust: an assessment. Water Air Soil Pollut. 14: 309-321.
11-197
-------
Duggan, M. J. (1983) Contribution of lead 1n dust to children's blood lead. Environ Health
Perspect. 50: 371-381.
Duggan, M. J.; Williams, S. (1977) Lead-in-dust 1n city streets. Sc1. Total Environ. 7: 91-97.
Elwood, P. C. (1983a) For debate ... changes 1n blood lead concentrations 1n women 1n Wales
1972-82. Br. Med. J. 286: 1553-1555.
Elwood, P. C. (1983b) Blood lead and petrol lead [letter]. Br. Med. J. 286: 1515.
Elwood, P. C. (1983c) Turin 1sotop1c lead experiment. Lancet 1(8329): 869.
Elwood, P. C.; Phillips, K. M.; Lowe, N. ; Phillips, J. K.; Toothlll, C. (1983) Hardness of
domestic water and blood lead levels. Hum. Toxicol. 2: 645-648.
EHcson, J. E.; Shlrahata, H.; Patterson, C. C. (1979) Skeletal concentrations of lead in
ancient Peruvians. N. Engl. J. Med. 300: 946-951.
Facchettl, S. (1979) Isotope study of lead 1n petrol. In: International conference: management
and control of heavy metals 1n the environment; September; London, United Klnadom
Edinburgh, United Kingdom: CEP Consultants, Ltd.; pp. 95-102. y
Facchettl, S. (1985) Isotoplc lead experiment - an update. Presented at: Lead environmental
health: the current Issues; May; Durham, NC. Durham, NC: Duke University Medical Center.
Facchettl, S.; Geiss, F. (1982) Isotoplc lead experiment: status report. Luxembourg: Com-
mission of the European Communities; Publication no. EUR 8352 EN.
Falrey, F. S.; Gray, J. W., III. (1970) Soil lead and pediatrlc lead poisoning in Charleston
S.C. J. S. C. Med. Assoc. 66: 79-82.
Fergusson, J. E.; Hibbard, K. A.; Ting, R. L. H. (1981) Lead In human hair: general survey -
battery factory employees and their families. Environ. Pollut. Ser. B 2: 235-248.
F1lipp1n1, L.; Slmmler, F. (1980) Ble1-Intox1kat1on durch Schnupftabak [Lead Intoxication due
to taking snuff]. Dtsch. Med. Wochenschr. 105: 1504-1506.
Fischbeln, A.; Rice, C.; Sarkozl, L.; Kon, S. H.; Petrocci, M.; Selikoff, I. J. (1979) Ex-
posure to lead 1n firing ranges. J. Am. Med. Assoc. 241: 1141-1144.
Fischbeln, A.; Cohn, J.; Ackerman, G. (1980a) Asbestos, lead, and the family: household risks.
J. Fam. Pract. 10: 989-992.
Fischbeln, A.; Nicholson, W. J.; Weisman, I. (1980b) Comparative lead emissions from con-
ventional and jacketed ammunition. Am. Ind. Hyg. Assoc. J. 41: 525-527.
FUndt, M. L. H.; King, E.; Walsh, 0. B. (1976) Blood lead and erythrocyte 6-am1nolevulinic
acid dehydratase levels in Manchester taxi drivers. Br. J. Ind. Med. 33: 79-84.
Forthofer, R. N. (1983) Investigation of nonresponse bias in NHANES II. Am. J. Epidemiol. 117:
507-515.
Fosse, G.; Wesenberg, G. B. R. (1981) Lead, cadmium, zinc and copper 1n deciduous teeth of
Norwegian children in the pre-1ndustrial age. Int. J. Environ. Stud. 16: 163-170.
11-198
-------
Foster, J. D.; Louria, D. B; Stinson, L. (1979) Influence of documented lead poisoning on en-
vironmental modification programs 1n Newark, New Jersey. Arch. Environ. Health 34:
368-371.
Frlberg, L.; Vahter, M. (1983) Assessment of exposure to lead and cadmium through biological
monitoring: results of a UNEP/WHO global study. Environ. Res. 30: 95-128.
Fugas, M. (1977) Biological significance of some metals as air pollutants. Part I: lead.
Research Triangle Park, NC: U.S. Environmental Protection Agency, Health Effects Research
Laboratory; EPA report no. EPA-600/1-77-041. Available from: NTIS, Springfield, VA;
PB-274055.
Fugas, M.; Wilder, B.; Paukovid, R.; Hrsak, J.; Stelner-Skreb, D. (1973) Concentration levels
and particle size distribution of lead in the air of an urban and an Industrial area as a
basis for the calculation of population exposure. In: Barth, D.; Berlin, A.; Engel, R.;
Recht, P.; Smeets, J., eds. Environmental health aspects of lead: proceedings, inter-
national symposium; October 1972; Amsterdam, The Netherlands. Luxembourg: Commission of
the European Communities, pp. 961-968.
Galke, W. A.; Hammer, D. I.; Kell, J. E.; Lawrence, S. W. (1975) Environmental determinants of
lead burdens 1n children. In: Hutchlnson, T. C.; Epstein, S.; Page, Al. L.; Van Loon, J.;
Davey, T., eds. International conference on heavy metals in the environment: symposium
proceedings, vol. 3; October; Toronto, ON, Canada. Toronto; ON, Canada: Institute for
Environmental Studies; pp. 53-74. See also: Washington, DC: U.S Environmental Protection
Agency; EPA report no. EPA-600/J-78-022 1975. Available from: NTIS, Springfield, VA;
PB-283567.
Gallacher, J. E. J.; Elwood, P. C.; Phillips, K. M.; Davles, B. E.; Jones, D. T. (1984)
Relation between pica and blood lead in areas of differing lead exposure. Arch. Dis.
Child. 59: 40-44.
Gallant, A. R. (1975) Testing a subset of the parameters of a nonlinear regression model. J.
Am. Stat. Assoc. 70: 927-932.
Garibaldi, P.; Facchetti, S.; Quagllardl, A.; Vanlni, G.; Gaddo, P. P.; De Bortoli, M.;
Gaglione, P. (1975) Petrols addltivated with Isotoplcally differentiated lead: proposal
of an experiment to estimate the Incidence of traffic on the environment pollution by
lead - first experimental results. In: Recent advances In the assessment of the health
effects of environmental pollution: proceedings, international symposium, vol. 3; June
1974; Paris, France. Luxembourg: Commission of the European Communities; pp. 1287-1299.
Gartside, P. S.; Buncher, C. R.; Lerner, S. (1982) Relationship of air lead and blood lead for
workers at an automobile battery factory. Int. Arch. Occup. Environ. Health 50: 1-10.
Cause, D.; Chase, W.; Foster, J.; Louria, 0. B. (1977) Reduction In lead levels among children
in Newark. J. Med. Soc. N. J. 74: 958-960.
Gilbert, C.; Tuthlll, R. W.; Calabrese, E. J.; Peters, H. A. (1979) A comparison of lead
hazards 1n the housing environment of lead poisoned children versus nonpoisoned controls.
J. Environ. Sci. Health Part #A14: 145-168.
Goldsmith, J- R- (1974) Food chain and health Implications of airborne lead. Sacramento, CA:
State of California, A1r Resources Board; report no. ARB-R-102-74-36. Available from:
NTIS, Springfield, VA; PB-248745.
11-199
-------
Goldsmith, J. R.; Hexter, A. C. (1967) Respiratory exposure to lead: ep1dem1olog1cal and ex-
perimental dose-response relationships. Science (Washington, DC) 158: 132-134.
Goldwater, L. J.; Hoover, A. W. (1967) An international study of "normal" levels of lead 1n
blood and urine. Arch. Environ. Health. 15: 60-63.
Grandjean, P.; Nielsen, 0. V.; Shapiro, I. M. (1979) Lead retention in ancient Nubian and
contemporary populations. J. Environ. Pathol. Toxlcol. 2: 781-787.
Graovac-Leposavlc, L.; Djuric, D.; Valjarevlc, V.; Senlcar, H.; Senlcar, L.; M1l1c", S.; Dellc",
V. (1973) Environmental lead contamination of Meza Valley - study on lead exposure of
population. In: Barth, D.; Berlin, A.; Engel, R.; Recht, P.; Smeets, J., eds. Environ-
mental health aspects of lead: proceedings, International symposium; October 1972;
Amsterdam, The Netherlands. Luxembourg: Commission of the European Communities; pp.
685-705.
Griffin, T. B.; Coulston, F.; Golberg, L.; Wills, H.; Russell, J. C.; Knelson, J. H. (1975)
Clinical studies on men continuously exposed to airborne particulate lead. In: Griffin,
T. B.; Knelson, J. H., eds. Lead. Stuttgart, West Germany: Georg Thieme Publishers; pp.
221-240. (Coulston, F.; Korte, F., eds. Environmental quality and safety, supplement v.
2).
Gross, S. B. (1979) Oral and Inhalation lead exposures 1n human subjects (Kehoe balance ex-
periments). New York, NY: Lead Industries Association.
Gross, S. B. (1981) Human oral and Inhalation exposures to lead: summary of Kehoe balance ex-
periments. J. Toxlcol. Environ. Health 8: 333-377.
Gross, S. B.; Pfitzer, E. A.; Yeager, D. W.; Kehoe, R. A. (1975) Lead 1n human tissues.
Toxicol. Appl. Pharmacol. 32: 638-651.
Hall, W. (1974) Survey plans and data collection and analysis methodologies: results of a
pre-survey for the magnitude and extent of the lead based paint hazard 1n housing.
Washington, DC: U.S. Department of Commerce, National Bureau of Standards; report no.
NBSIR 74-426. Available from: NTIS, Springfield, VA; COM-74-11074.
Hammer, D. I.; Finklea, J. F.; Hendricks, R. H.; Hlnners, T. A.; Rlggan, W. B.; Shy, C. M.
(1972) Trace metals 1n human hair as a simple ep1demiolog1c monitor of environmental ex-
posure. In: HemphUl, D. D., ed. Trace substances 1n environmental health - V: [pro-
ceedings of University of Missouri's 5th annual conference on trace substances in en-
vironmental health]; June 1971; Columbia, MO. Columbia, MO: University of Missouri-
Columbia; pp. 25-38.
Hammond, P. B.; O'Flaherty, E. J.; Gartslde, P. S. (1981) The impact of air-lead on blood-lead
1n man - a critique of the recent literature. Food Cosmet. Toxicol. 19: 631-638.
Hammond, P. B.; O'Flaherty, E. J.; Gartslde, P. S. (1982) Impact of air-lead on blood-lead 1n
man [letter]. Food Chem. Toxlcol. 20: 493.
Hansen, K. S.; Sharp, F. R. (1978) Gasoline sniffing, lead poisoning, and myoclonus. J. Am.
Med. Assoc. 240: 1375-1376.
11-200
-------
Hartwell, T. D.; Handy, R. W.; Harris, B. S.; Williams, S. R.; Gehlbach, S. H. (1983) Heavy
metal exposure in populations living around zinc and copper smelters. Arch. Environ.
Health 38: 284-295.
Hasselblad, V. ; Nelson, W. (1975) Additional analyses of the seven city lead study. In:
Griffin, T. B.; Knelson, J. H., eds. Lead. Stuttgart, West Germany: Georg Thieme Pub-
lishers; pp. 147-151. (Coulston, F. ; Korte, F. , eds. Environmental quality and safety:
supplement v. 2).
Hasselblad, V.; Stead, A. G. ; Galke, W. (1980) Analysis of coarsely grouped data from the log-
normal distribution. J. Am. Stat. Assoc. 75: 771-778.
Heyworth, F.; Splckett, J.; Dick, M.; Margetts, B.; Armstrong, B. (1981) Tailings from a lead
mine and lead levels in schoolchildren: a preliminary report. Med. J. Aust. 2: 232-234.
Hopkins, D. R.; Houk, V. N. (1976) Federally-assisted screening projects for childhood lead
poisoning control: the first three years (July 1972-June 1975). Am. J. Public Health 66:
485-486.
Hopper, J. L; Mathews, J. D. (1983) Extensions to multivariate normal models for pedigree
analysis: II. modeling the effect of shared environment in the analysis of variation in
blood lead levels. Am. J. Epidemiol. 117: 344-355.
Horn, J. D. (1976) Significant overestimation of blood lead levels in practice: a consequence
of using the micro method of blood lead analysis. ON, Canada: University of Western
Ontario (Prepublication draft).
Hubermont, G.; Buchet, J.-P.; Roels, H.; Lauwerys, R. (1978) Placental transfer of lead, mer-
cury and cadmium in women living in a rural area: importance of drinking water in lead
exposure. Int. Arch. Occup. Environ. Health 41: 117-124.
Human Nutrition Information Service. (1983) Food consumption: households in the United States,
seasons and year 1977-78; nationwide food consumption survey 1977-78. Washington, DC:
U.S. Department of Agriculture; NFCS 1977-78 report no. H-6. Available from: GPO,
Washington, DC; S/N 001-000-04335-8.
Hunter, J. M. (1978) The summer disease: some field evidence on seasonal1ty in childhood lead
poisoning. Soc. Sc1. Med. 12: 85-94.
Jacquez, J. A. (1972) Compartmental analysis 1n biology and medicine: kinetics of distribution
of tracer-labeled materials. New York, NY: Elsevier Publishing Co.
Johanson, W. C., Jr.; Luby, J. P. (1972) A report on a study to determine the blood lead
levels in Dallas children. Dallas, TX: Dallas Health Department.
Johnson, D. E.; Tlllery, J. B.; Hosenfeld, J. M.; Register, J. W. (1974) Development of
analytic techniques to measure human exposure to fuel additives. Research Triangle Park,
NC: U.S. Environmental Protection Agency, National Environmental Research Center; EPA
report no. EPA-650/1-74-003. Available from: NTIS, Springfield, VA; PB-232124.
Johnson, D. E.; Tillery, J. B.; Prevost, R. J. (1975) Levels of platinum, palladium, and lead
1n populations of southern California. Environ. Health Perspect. 12: 27-33.
11-201
-------
Johnson, D. E.; Prevost, R. J.; Tillery, J. B.; Camann, D. E.; Hosenfeld, J. M. (1976) Base-
line levels of platinum and palladium In human tissue. Research Triangle Park, NC: U.S.
Environmental Protection Agency, Health Effects Research Laboratory; EPA report no.
EPA-600/1-76-019. Available from: NTIS, Springfield, VA; PB-251885.
Johnson, D. E.; Prevost, R. J. ; Tlllery, J. B.; Thomas, R. E. (1978) The distribution of
cadmium and other metals 1n human tissue. Research Triangle Park, NC: U.S. Environmental
Protection Agency; EPA report no. EPA-600/1-78-035. Available from: NTIS, Springfield,
VA; PB-285200.
Jones, R. D.; Commins, B. T.; Cernik, A. A. (1972) Blood lead and carboxyhaemoglobin levels 1n
London taxi drivers. Lancet 2(7772): 302-303.
Kang, H. K. ; Infante, P. F.; Carra, J. S. (1983) Determination of blood-lead elimination
patterns of primary lead smelter workers. J. Toxicol. Environ. Health 11: 199-210.
Karalekas, P. C., Jr.; Ryan, C. R.; Taylor, F. B. (1983) Control of lead, copper, and Iron
pipe corrosion in Boston. J. Am. Water Works Assoc. 75: 92-95.
Kaufman, A.; W1ese, W. (1978) Gasoline sniffing leading to increased lead absorption in
children. CUn. Pediatr. (Philadelphia) 17: 475-477.
Kehoe, R. A. (1961a) The metabolism of lead in man in health and disease: the normal
metabolism of lead. (The Harben lectures, 1960). J. R. Inst. Public Health Hyg. 24:
81-97.
Kehoe, R. A. (1961b) The metabolism of lead in man in health and disease: the metabolism of
lead under abnormal conditions. (The Harben lectures, 1960). J. R. Inst. Public Health
Hyg. 24: 129-143.
Kehoe, R. A. (1961c) The metabolism of lead 1n man 1n health and disease: present hygienic
problems relating to the absorption of lead. (The Harben lectures, 1960). J. R. Inst.
Public Health Hyg. 24: 177-203.
Khandekar, R. N.; Mishra, U. C.; Vohra, K. G. (1984) Environmental lead exposure of an urban
Indian population. Sci. Total Environ. 40: 269-278.
King, E. (1983) Changes in blood lead concentrations in women in Wales 1972-82. Br. Med. J.
286: 2059-2060.
King, E.; Conchie, A.; H1ett, D.; Milllgan, B. (1979) Industrial lead absorption. Ann. Occup.
Hyg. 22: 213-239.
Knelson, J. H.; Johnson, R. J.; Coulston, F.; Golberg, L.; Griffin, T. (1973) Kinetics of
respiratory lead uptake in humans. In: Barth, D.; Berlin, A.; Engel, R.; Recht, P.;
Smeets, J., eds. Environmental health aspects of lead: proceedings, international sym-
posium; October 1972; Amsterdam, The Netherlands. Luxembourg; Commission of the European
Communities, pp. 391-401.
Lacey, R. F.; Moore, M. R.; Richards, W. N. (1985) Lead in water, Infant diet and blood: the
Glasgow duplicate diet study. Sci. Total Environ. 41: 235-257.
11-202
-------
Landrigan, P. J.; Baker, E. L. (1981) Exposure of children to heavy metals from smelters:
epidemiology and toxic consequences. Environ. Res. 25: 204-224.
Landrigan, P. J. ; Gehlbach, S. H.; Rosenblum, B. F. ; Shoults, J. M.; Candelaria, R. M. ;
Barthel, W. F.; Liddle, J. A.; Smrek, A. L.; Staehling, N. W.; Sanders, J. F. (1975)
Epidemic lead absorption near an ore smelter: the role of participate lead. N. Engl. J.
Med. 292: 123-129.
Landrigan, P. J.; Baker, E. L., Jr.; Feldman, R. G.; Cox, D. H.; Eden, K. V.; Orenstein, W.
A.; Mather, J. A.; Yankel, A. J.; von Lindern, I. H. (1976) Increased lead absorption
with anemia and slowed nerve conduction 1n children near a lead smelter. J. Pedlatr. (St.
Louis) 89: 904-910.
Landrigan, P. J.; Baker, E. L., Jr.; Himmelstein, J. S.; Stein, G. F.; Wedding, J. P.; Straub,
W. E. (1982) Exposure to lead from the Mystic River bridge: the dilemma of deleading. N.
Engl. J. Med. 306: 673-676.
Lead Development Association. (1973) Medical aspects of lead absorption 1n Industrial pro-
cesses: proceedings of a meeting; July 1972; London, United Kingdom. London, United
Kingdom: Lead Development Association.
Lepow, M. L.; Bruckman, L.; Gillette, M.; Markowitz, S. ; Roblno, R. ; Kapish, J. (1975) In-
vestigations Into sources of lead in the environment of urban children. Environ. Res. 10:
415-426.
L1n-Fu, J. S. (1973) Vulnerability of children to lead exposure and toxlcity: parts one and
two. N. Engl. J. Med. 289: 1229-1233; 1289-1293.
Lucas, J. M. (1981) Effect of analytical variability on measurements of population blood lead
levels. Am. Ind. Hyg. Assoc. J. 42: 88-96.
Mahaffey, K. R.; Annest, J. L.; Barbano, H. E.; Murphy, R. S. (1979) Preliminary analysis of
blood lead concentrations for children and adults: HANES II, 1976-1978. In: Hemphlll, D.
D., ed. Trace substances in environmental health - XIII: [proceedings of University of
Missouri's 13th annual conference on trace substances in environmental health]; June;
Columbia, MO. Columbia, MO: University of Missouri-Columbia; pp. 37-51.
Mahaffey, K. R.; Annest, J. L.; Roberts, J.; Murphy, R. S. (1982) National estimates of blood
lead levels: United States, 1976-1980: association with selected demographic and socio-
economic factors. N. Engl. J. Med. 307: 573-579.
Manton, W. I. (1977) Sources of lead in blood: identification by stable isotopes. Arch.
Environ. Health 32: 149-159.
Manton, W. I. (1985) Total contribution of airborne lead to blood lead. Br. J. Ind. Med. 42:
168-172.
Manton, W. I.; Cook, J. D. (1984) High accuracy (stable Isotope dilution) measurements of lead
in serum and cerebrosplnal fluid. Br. J. Ind. Med. 41: 313-319.
Marcus, A. H. (1985) Multicompartment kinetic model for lead: III. lead in blood plasma and
erythrocytes. Environ. Res. 36: 473-489.
11-203
-------
Marcus, A. H. (1985) Testing alternative nonlinear kinetic models In corapartmental analysis
In: Elsenfeld, J.; Del1s1, C., eds. Mathematics and computers in biomedlcal aoolications'
Amsterdam, The Netherlands: Elsevler; pp. 259-267. H wuons.
Martin, A. E.; Fairweather, F. A.; Buxton, R. St. J.; Roots, L. M. (1975) Recent epldemlo-
loglcal studies on environmental lead of Industrial origin. In: Recent advances 1n the
assessment of the health effects of environmental pollution; v. 2. proceedings Inter-
national symposium; June 1974; Paris, France. Luxembourg: Commission of the
Communities; pp. 1113-1122.
McDowell, A.; Engel, A.; Massey, J. T.; Maurer, K. (1981) Plan and operation of the second
National Health and Nutrition Examination Survey, 1976-80. Washington, DC- U S
Department of Health and Human Services, National Center for Health Statistics- DHH<;
publication no. (PHS) 81-1317. (Vital and health statistics series 1, programs and
collection procedures: no. 15).
Mclntire, M. S.; Angle, C. R. (1973) Air lead/blood lead 1n G-6-PD deficient black school
children. In: Barth, D.; Berlin, A.; Engel, R. ; Recht, P.; Smeets, J., eds. Environmental
health aspects of lead: proceedings, International symposium; October 1972; Amsterdam
The Netherlands. Luxembourg: Commission of the European Communities, Centre for Tnfnr-mal
tion and Documentation; pp. 749-761. inrorma
Mlelke, H. W.; Anderson, J. C. ; Berry, K. J.; Mielke, P. W.; Chaney, R. L. ; Leech, M (1983)
Lead concentrations in inner-city soils as a factor in the child lead orobl'em Am i
Public Health 73: 1366-1369.
Mielke, H.; Blake, B.; Burroughs, S.; Hassinger, N. (1984) Urban lead levels in Minneapolis*
the case of the Hmong children. Environ. Res. 34: 64-76.
Milic, S.; Stankovic, M.; Delic, V.; Djordjevlc, V. (1973) Biochemical parameters in evalua-
tion of environmental lead exposure. In: Barth, D.; Berlin, A.; Engel, R.; Recht P.-
Smeets, J., eds. Environmental health aspects of lead: proceedings, International'sym-
posium; October 1972; Amsterdam, The Netherlands. Luxembourg: Commission of the European
Communities, pp. 561-568.
Molina-Ballesteros, 6.; ZufUga-Charles, M. A.; Ortega, A. C.; Solfs-Camara R., P.; Solfs-
Cdmara, V. P. (1983) Lead concentrations 1n the blood of children from pottery-making
families exposed to lead salts 1n a Mexican village. Bull. Panam. Health Organ. 17-
35-41.
Moore, M. R. (1977) Lead in drinking water in soft water areashealth hazards. Sc1 Total
Environ. 7: 109-115.
Moore, M. R.; Goldberg, A.; Meredith, P. A.; Lees, R.; Low, R. A.; Pocock, S. J. (1979) The
contribution of drinking water lead to maternal blood lead concentrations. CUn. Ch1m
Acta 95: 129-133.
Moore, M. R.; Goldberg, A.; Fyfe, W. M.; Low, R. A.; Richards, W. N. (1981a) Lead in water 1n
Glasgow - a story of success. Scott. Med. J. 26: 354-355.
Moore, M. R.; Goldberg, A.; Fyfe, W. M.; Richards, W. N. (1981b) Maternal lead levels after
alterations to water supply [letter]. Lancet 2(8239): 203-204.
Morse, D. L.; Watson, W. N.; Housworth, J.; Witherell, L. E.; Landrigan, P. J. (1979) Exposure
of children to lead in drinking water. Am. J. Public Health 69: 711-712.
11-204
-------
Morton, D. E. ; Saah, A. J. ; Silberg, S. L. ; Owens, W. L.; Roberts, M. A.; Saah, M. D. (1982)
Lead absorption 1n children of employees in a lead-related industry. Am. J. Epidemiol
115: 549-555.
Nathanson, B. ; Nudelman, H. (1980) Ambient lead concentrations in New York City and their
health implications. Bull. N. Y. Acad. Med. 56: 866-875.
National Academy of Sciences. National Research Council. (1976) Recommendations for the pre-
vention of lead poisoning 1n children. Washington, DC: National Academy of Sciences.
Available from: NTIS, Springfield, VA; PB-257645.
Neri, L. C.; Johansen, H. L.; Schmltt, N.; Pagan, R. T.; Hewitt, D. (1978) Blood lead levels
in children in two British Columbia communities. In: Hemphill, D. D. , ed. Trace sub-
stances in environmental health - XII: [proceedings of University of Missouri's 12th
annual conference on trace substances in environmental health]; June; Columbia, MO.
Columbia, MO: University of Missouri-Columbia; pp. 403-410.
NHANES II Time Trend Analysis Review Group. (1983) Report of the NHANES II Time Trend Analysis
Review Group. [From meeting; March 1983; Research Triangle Park, NC. Available for in-
spection at: U.S. Environmental Protection Agency, Environmental Criteria and Assessment
Office, Research Triangle Park, NC.
Nordman, C. H. (1975) Environmental lead exposure in Finland: a study on selected population
groups [dissertation]. Helsinki, Finland: University of Helsinki.
Nordman, C. H.; Hernberg, S. ; Nlkkanen, J.; Ryhanen, A. (1973) Blood lead levels and erythr-
ocyte 6-aminolevullnic acid dehydratase activity In people living around a secondary lead
smelter. Work Environ. Health 10: 19-25.
Nutrition Foundation, Inc. (1982) Assessment of the safety of lead and lead salts 1n food: a
report of the Nutrition Foundation's Expert Advisory Committee. Washington, DC: The
Nutrition Foundation.
O'Flaherty, E. J.; Hammond, P. B.; Lerner, S. I. (1982) Dependence of apparent blood lead
half-life on the length of previous lead exposure in humans. Fundam. Appl. Toxicol. 2:
49-54.
Okubo, T. ; Omae, K.; Sakurai, H.; Sugita, M.; Tsuchiya, K. (1983) The relationship of airborne
lead to blood lead in school children. J. UOEH 5(suppl.): 137-144.
Ontario Ministry of the Environment. (1975) Report of the lead data analysis task force.
Toronto, ON, Canada: Ontario Ministry of the Environment.
Oxley, G. R. (1982) Blood lead concentrations: apparent reduction over approximately one
decade. Int. Arch. Occup. Environ. Health 49: 341-343.
Pennlngton, J. A. T. (1983) Revision of the total diet study food list and diets. J. Am. Diet.
Assoc. 82: 166-173.
PiomelH, S.; Corash, L.; Corash, M. B.; Seaman, C.; Mushak, P.; Glover, B.; Padgett, R.
(1980) Blood lead concentrations in a remote Himalayan population. Science (Washington,
DC) 210: 1135-1137.
11-205
-------
Pirkle, J. L.; Annest, J. L. (1984) Blood lead levels [letter]. N. Enql J Mod
1125-1126. ' ' °-
Pocock, S. J.; Shaper, A. G. ; Walker, M.; Wale, C. J.; Clayton, B.; Delves T Lacev R F
Packham, R. F.; Powell, P. (1983) Effects of tap water lead, water hardness, alcohol' and
cigarettes on blood lead concentrations. J. Ep1dem1ol. Comm. Health 37: 1-7. '
Poole, C.; Smythe, L. E.; Alpers, M. (1980) Blood lead levels 1n Papua New Guinea
living 1n a remote area. Sc1. Total Environ. 15: 17-24.
Popovac, D.; Grazlano, J.; Seaman, C.; Kaul, B.; Colakovic, B.; Popovac R Osmani T
Haxhlu, M.; Begraca, M.; Bozovic, Z.; M1k1c, M. (1982) Elevated blood'lead In a
tlon near a lead smelter 1n Kosovo, Yugoslavia. Arch. Environ. Health 37: 19-23
Quinn, M. J. (1983) Factors affecting blood lead concentrations 1n the UK. In: International
conference: heavy metals 1n the environment; September; Heidelberg, West Germanv
burgh, United Kingdom: CEP Consultants, Ltd.; pp. 294-297. y"
Rablnowitz, M. B. ; Needleman, H. L. (1982) Temporal trends 1n the lead concentrat-innc
umbilical cord blood. Science (Washington, DC) 216: 1429-1431. '«ions
Rabinowltz, M.; Needleman, H. L. (1983) Petrol lead sales and umbilical cord blood lead
in Boston, Massachusetts [letter]. Lancet 1(8314/5): 63.
Rabinowltz, M.; Needleman, H. (1984) Variability of blood lead concentrations durina
Arch. Environ. Health 39: 74-77. 9
Rabinowltz, M. B.; Wetherill, G. W.; Kopple, J. D. (1973) Lead metabolism in the normal
stable isotope studies. Science (Washington, DC) 182: 725-727.
Rablnowitz, M.; Wetherill, G. W.; Kopple, J. D. (1974) Studies of human lead metabolism bv use
of stable isotope tracers. Environ. Health Perspect. 7: 145-153.
Rabinowltz, M. B.; Wetherill, G. W.; Kopple, J. D. (1976) Kinetic analysis of lead metabolism
in healthy humans. J. CUn. Invest. 58: 260-270.
Rabinowltz, M. B.; Wetherlll, G. W.; Kopple, J. D. (1977) Magnitude of lead intake from
respiration by normal man. J. Lab. din. Med. 90: 238-248.
Rabinowltz, M. B.; Kopple, J. D.; Wetherill, G. w. (1980) Effect of food Intake and fastlnq on
gastrointestinal lead absorption in humans. Am. J. CUn. Nutr. 33: 1784-1788.
Rablnowitz, M.; Needleman, H.; Burley, M.; Finch, H.; Rees, J. (1984) Lead 1n umbilical blood
indoor air, tap water, and gasoline in Boston. Arch. Environ. Health 39: 299-301. '
Rabinowltz, M. B.; Leviton, A.; Needleman, H. L. (1985a) Lead in milk and infant blood a
dose-response model. Arch. Environ. Health: in press. '
Rablnowitz, M.; Leviton, A.; Bellinger, D. (1985b) Home refinishlng, lead paint, and infant
blood lead levels. Am. J. Public Health 75: 403-404.
Rabinowltz, M.; Leviton, A.; Needleman, H.; Bellinger, D.; Waternaux, C. (1985c) Environmental
correlates of infant blood lead levels. In: Bornschein, R. L.; Rablnowitz, M. B., eds
The second international conference on prospective studies of lead; April 1984-
Cincinnati, OH. Environ. Res. 38: 96-107.
11-206
-------
Reece, R. M.; Reed, A. J.; Clark, C. S.; Angoff, R.; Casey, K. R.; Challop, R. S.; McCabe, E.
A. (1972) Elevated blood lead levels and the jjn situ analysis of wall paint by X-ray
fluorescence. Am. J. D1s. Child. 124: 500-502.
R1ce, C.; Fischbein, A.; L1l1s, R.; Sarkozl, L.; Kon, S.; Sellkoff, I. J. (1978) Lead con-
tamination in the homes of employees of secondary lead smelters. Environ. Res 15-
375-380.
Roberts, T. M.; Hutchlnson, T. C.; Paciga, J.; Chattopadhyay, A.; Jervls, R. E.; VanLoon, J.;
Parkinson, D. K. (1974) Lead contamination around secondary smelters: estimation of
dispersal and accumulation by humans. Science (Washington, DC) 186: 1120-1123.
Roels, H.; Buchet, J.-P.; Lauwerys, R.; Hubermont, G.; Bruaux, P.; Claeys-Thoreau, F. ;
Lafontalne, A.; Van Overschelde, J. (1976) Impact of air pollution by lead on the heme
blosynthetlc pathway in school-age children. Arch. Environ. Health 31: 310-316.
Roels, H. A.; Buchet, J.-P.; Lauwerys, R.; Bruaux, P.; Claeys-Thoreau, F.; Lafontalne, A.;
van Overschelde, J.; Verduyn, G. (1978) Lead and cadmium absorption among children near a
nonferrous metal plant: a follow-up study of a test case. Environ. Res. 15: 290-308.
Roels, H. A.; Buchet, J.-P.; Lauwerys, R. R.; Bruaux, P.; Claeys-Thoreau, F.; Lafontaine, A.;
Verduyn, G. (1980) Exposure to lead by the oral and the pulmonary routes of children
living in the vicinity of a primary lead smelter. Environ. Res. 22: 81-94.
Ryu, J. E.; Zlegler, E. E.; Nelson, S. E.; Fomon, S. J. (1983) Dietary Intake of lead and
blood lead concentration 1n early Infancy. Am. J. D1s. Child. 137: 886-891.
Saltzman, B. E.; Yeager, D. W.; Melners, B. G. (1983) ReproducibilUy and quality control 1n
the analysis of biological samples for lead and mercury. Am. Ind. Hyg. Assoc. J. 44-
263-267.
Sayre, J. W.; Charney, E.; Vostal, J.; Pless, I. B. (1974) House and hand dust as a potential
source of childhood lead exposure. Am. J. D1s. Child. 127: 167-170.
Schlegel, H.; Kufner, G. (1979) Long-term observation of biochemical effects of lead in human
experiments. J. CUn. Chem. Clin. Blochem. 17: 225-233.
Schmltt, N.; PhlHon, J. J.; Larsen, A. A.; Harnadek, M.; Lynch, A. J. (1979) Surface soil as
a potential source of lead exposure for young children. Can. Med. Assoc. J. 121-
1474-1478.
Schubert, J. ; Brodsky, A.; Tyler, S. (1967) The log-normal function as a stochastic model of
the distribution of stront1um-90 and other fission products in humans. Health Phys. 13-
1187-1204.
Schwartz, J. D.; Janney, A.; Pitcher, H. (1984) The relationship between gasoline lead and
blood lead. Available from: Joel D. Schwartz, U.S. Environmental Protection Agency
Office of Policy Analysis, Washington, DC.
Searle, C. E.; Harnden, D. G. (1979) Lead 1n hair-dye preparations [letter]. Lancet 2(8151)-
1070.
Shaper, A. G.; Pocock, S. J.; Walker, M. ; Wale, C. J.; Clayton, B.; Delves, H. T.; Hinks L
(1982) Effects of alcohol and smoking on blood lead 1n middle-aged British men Br Med'
J. 284: 299-302.
11-207
-------
Shapiro, I. M.; Mitchell, G.; Davidson, I.; Katz, S. H. (1975) The lead content of teeth-
evidence establishing new minimal levels of exposure in a living preindustrializert human
population. Arch. Environ. Health 30: 483-486. uman
Shellshear, I. D. (1973) Lead poisoning in childhood: a case report with environmental
cations. N.Z. Med. J. 78: 251-254.
Shellshear, I. D.; Jordan, L. D. ; Hogan, D. J.; Shannon, F. T. (1975) Environmental lead ex-
posure in ChHstchurch children: soil lead a potential hazard. N. Z. Med. J. 81: 382-386
Sherlock, J.; Walters, B. (1983) Dietary intake of heavy metals and its estimation Chpm TnH
(London) (4 July): 508-518. ' ' ina>
Sherlock, J. ; Smart, G. ; Forbes, G. I.; Moore, M. R.; Patterson, W. J.; Richards W N
Wilson, T. S. (1982) Assessment of lead intakes and dose-response for a population in Avr
exposed to a plumbsolvent water supply. Hum. Toxicol. 1: 115-122. y
Sherlock, J. C.; Ashby, D.; Delves, H. I.; Forbes, G. I.; Moore, M. R.; Patterson W 1
Pocock, S. J.; Quinn, M. J.; Richards, W. N.; Wilson, T. S. (1984) Reduction in 'exposure
to lead from drinking water and its effect on blood lead concentrations. Hum To*irni 7.
383-392. ' OX1C01- J-
Shier, D. R.; Hall, W. G. (1977) Analysis of housing data collected in a lead-based oaint
survey in Pittsburgh, Pennsylvania: part I. Washington, DC: National Bureau of Standee-
Report no. NBSIR 77-1250. ^anaaras,
Sinn, W. (1980) Ober den Zusammenhang von Luftbleikonzentration und Bleigehalt des Blutes von
Anwohnern und BerufstStigen im Kerngebiet einer Grossstadt (Blutbleistudie Frankfurt)- I
Versuchsanlage und DifferenzprUfung [Relationship between lead concentration in the air
and blood levels of people living and working in the centre of a city (Frankfurt blood
lead study): I. experimental method and examination of differences]. Int. Arch Occun
Environ. Health 47: 93-118. ' p-
Sinn, W. (1981) Relationship between lead concentration in the air and blood lead levels of
people living and working in the center of a city (Frankfurt blood lead study): II. cor-
relations and conclusions. Int. Arch. Occup. Environ. Health 48: 1-23.
Snee, R. D. (1981) Evaluation of studies of the relationship between blood lead and air lead
Int. Arch. Occup. Environ. Health 48: 219-242.
Snee, R. D. (1982a) Development of an air quality standard for lead from community studies
Environ. Sci. Technol. 16: 241-246.
Snee, R. D. (1982b) Models for the relationship between blood lead and air lead. Int. Arch
Occup. Environ. Health 50: 303-319.
Snee, R. D. (1982c) Silver Valley lead study: further analysis of the relationship between
blood lead and air lead. J. Air Pollut. Control Assoc. 32: 170-175.
Snee, R. D.; Pfeifer, C. G. (1983) [Letter to Dr. David E. Weil]. January 31. Available for
inspection at: U.S. Environmental Protection Agency, Environmental Criteria and Assess-
ment Office, Research Triangle Park, NC; docket no. ECAO-CD-81-2, WS-6-22.
11-208
-------
Spengler, J. D.; Billick, I.; Ryan, P. B. (1984) Modeling population exposures to airborne
lead. In: Berglund, B.; Undvall, T.; Sundell, J., eds. Indoor air: v. 4, chemical char-
acterization and personal exposure; August; Stockholm, Sweden. Stockholm, Sweden: Swedish
Council for Building Research; pp. 87-94.
Stark, A. D. ; Quah, R. F.; Meigs, J. W.; DeLoulse, E. R. (1982) The relationship of environ-
mental lead to blood-lead levels in children. Environ. Res. 27: 372-383.
Steenhout, A. (1982) Kinetics of lead storage in teeth and bones: an epidemiologic approach.
Arch. Environ. Health 37: 224-231.
Stephens, R. (1981) Human exposure to lead from motor vehicle emissions. Int. J. Environ.
Stud. 17: 73-83.
Stuik, E. J. (1974) Biological response of male and female volunteers to inorganic lead. Int.
Arch. Arbeltsmed. 33: 83-97.
Tepper, L. B.; Levin, L. S. (1975) A survey of air and population lead levels 1n selected
American communities. In: Griffin, T. B.; Knelson, J. H., eds. Lead. Stuttgart, West
Germany: Georg Thieme Publishers; pp. 152-196. (Coulston, F.; Korte, F., eds. Environ-
mental quality and safety: supplement v. 2).
Ter Haar, G. (1981) Lead debate [letter]. Environ. Sc1. Technol. 15: 722.
Ter Haar, G.; Aronow, R. (1974) New information on lead in dirt and dust as related to the
childhood lead problem. Environ. Health Perspect. 7: 83-89.
Ter Haar, G.; Chadzynski, L. (1979) An Investigation of elevated blood lead levels in Detroit
children. Arch. Environ. Health 34: 145-150.
Thomas, H. F. (1980) Domestic water usage and blood lead levels. Public Health 94: 294-295.
Thomas, H. V.; Milmore, B. K.; Heidbreder, G. A.; Kogan, B. A. (1967) Blood lead of persons
living near freeways. Arch. Environ. Health 15: 695-702.
Thomas, H. F.; Elwood, P. C.; Welsby, E.; St. Leger, A. S. (1979) Relationship of blood lead
in women and children to domestic water lead. Nature (London) 282: 712-713.
Thomas, H. F.; Elwood, P. C.; Toothlll, C.; Morton, M. (1981) Blood and water lead 1n a hard
water area. Lancet 8228: 1047-1048.
Tsuchiya, K.; Sugita, M.; Seki, Y.; Kobayashi, Y.; Horl, M.; Park, Ch. B. (1975) Study of lead
concentrations in atmosphere and population in Japan. In: Griffin, T. B.; Knelson, J. H.,
eds. Lead. Stuttgart, West Germany: Georg Thieme Publishers; pp. 95-145. (Coulston, F.;
Korte, F., eds. Environmental quality and safety: supplement v. 2).
Tyler, R. L. (1970) Philadelphia combats "silent epidemic" in the "ghetto": lead poisoning
control. J. Environ. Health 33: 64-71.
Tyrer, F. H. (1977) A cautionary tale of a lead hazard 1n recycling of waste. J. Soc. Occup.
Med. 27: 26-30.
U S Centers for Disease Control. (1973) Human lead absorption - Texas. Morb. Mortal. Wkly.
Rep. 22: 405-407.
11-209
-------
U. S. Centers for Disease Control. (1977a) Surveillance of childhood lead polsonlna -United
States. Morb. Mortal. Wkly. Rep. 26: 49. *
U. S. Centers for Disease Control. (1977b) Increased lead absorption in children of lead
workers - Vermont. Morb. Mortal. Wkly. Rep. 26: 61-62.
U. S. Centers for Disease Control. (1979) Surveillance of childhood lead polsonlna -United
States. Morbidity and Mortality Weekly Report 28: 117-118.
U. S. Centers for Disease Control. (1980) Surveillance of childhood lead polsonlna -United
States. Morb. Mortal. Wkly. Rep. 29: 170-171. u
U. S. Centers for Disease Control. (1982a) Surveillance of childhood lead polsonlna -United
States. Morb. Mortal. Wkly. Rep. 31: 118-119. y
U. S. Centers for Disease Control. (1982b) Blood-lead levels in U. S. population
Mortal. Wkly. Rep. 31: 132-134. "
U. S. Centers for Disease Control. (1983a) Lead poisoning from Mexican folk
-California. Morb. Mortal. Wkly. Rep. 32: 554-555.
U. S. Centers for Disease Control. (1983b) Folk remedy-associated lead polsonlna in Hmonn
children - Minnesota. Morb. Mortal. Wkly. Rep. 32: 555-556. y
U. S. Consumer Product Safety Commission. (1974) A report to Congress in compliance with the
lead based paint poisoning prevention act, as amended (P.L. 93-151). Washington DC- U
S. Consumer Product Safety Commission. Available from: NTIS, Springfield, VA; PB-245225!
U. S. Environmental Protection Agency. (1972) Helena Valley, Montana, area environmental pol-
lution study. Research Triangle Park, NC: Office of Air Programs; Office of Air Proarams
publication no. AP-91. Available from: NTIS, Springfield, VA; PB-207126.
U. S. Environmental Protection Agency. (1980) Ambient water quality criteria for lead.
Washington, DC: U. S. Environmental Protection Agency; EPA report no. EPA-440/5-80-057.
Available from: NTIS, Springfield, VA; PB81-117681.
U. S. Environmental Protection Agency. (1983) A1r quality criteria for lead: v. I-IV [external
review draft no. 1]. Research Triangle Park, NC: Environmental Criteria and Assessment
Office; EPA report no. EPA-600/8-83-028A. Available from: NTIS, Springfield VA-
PB84-144591.
United Kingdom Central Directorate on Environmental Pollution. (1982) The Glasgow duplicate
diet study (1979/1980): a joint survey for the Department of the Environment and the
Ministry of Agriculture, Fisheries and Food. London, United Kingdom: Her Majesty's Sta-
tionery Office; pollution report no. 11.
Urban, W. D. (1976) Statistical analysis of blood lead levels of children surveyed in Pitts-
burgh, Pennsylvania: analytical methodology and summary results. Washington, DC: U.S.
Department of Commerce, National Bureau of Standards; report no. NBSIR 76-1024. Available
from: NTIS, Springfield, VA; PB-255876.
von Llndern, I.; Yankel, A. J. (1976) Presentation to the Shoshone heavy metals project com-
mittee (Seattle, WA; September 1975). In: Shoshone lead health project: work summary.
Boise, ID: Idaho Department of Health & Welfare; pp. 73-89.
11-210
-------
Vostal, J. J.; Taves, E.; Sayre, J. W.; Charney, E. (1974) Lead analysis of house dust: a
method for the detection of another source of lead exposure In Inner city children.
Environ. Health Perspect. 7: 91-97.
Waldron, H. A. (1975) Lead levels 1n blood of residents near the M6-A38(M) Interchange,
Birmingham. Nature (London) 253: 345-346.
Waldron, H. A. (1981) Postmortem absorption of lead by the skeleton. Am. J. Phys. Anthro. 55:
395-398.
Waldron, H. A.; Khera, A.; Walker, G.; Wlbberley, G. ; Green, C. J. S. (1979) Lead concentra-
tions 1n bones and soil. J. Archaeol. Sc1. 6: 292-298.
Walter, S. D.; Yankel, A. J.; von Lindern, I. H. (1980) Age-specific risk factors for lead
absorption 1n children. Arch. Environ. Health 35: 53-58.
Watson, W. N.; Wltherell, L. E.; Glguere, G. C. (1978) Increased lead absorption 1n children
of workers In a lead storage battery plant. J. Occup. Med. 20: 759-761.
Wedeen, R. P.; MalUk, D. K.; Batuman, V.; Bogden, J. D. (1978) Geophaglc lead nephropathy:
case report. Environ. Res. 17: 409-415.
W1gle, D. T.; Charlebols, E. J. (1978) Electric kettles as a source of human lead exposure.
Arch. Environ. Health 33: 72-78.
Williams, M. K.; King, E.; Walford, J. (1969) An Investigation of lead absorption 1n an
electric accumulator factory with the use of personal samplers. Br. J. Ind. Med. 26:
202-216.
Worth, D.; Matranga, A.; Lleberman, M.; DeVos, E.; Karelekas, P.; Ryan, C.; Craun, G. (1981)
Lead 1n drinking water: the contribution of household tap water to blood lead levels. In:
Lynam, D. R.; P1antan1da, L. G.; Cole, J. F., eds. Environmental lead: proceedings of the
second International symposium on environmental lead research; December 1978; Cincinnati,
OH. New York, NY: Academic Press; pp. 199-225. (Coulston, F.; Korte, F., eds. Ecotoxl-
cology and environmental quality series).
Yaffe, Y.; Flessel, C. P.; Wesolowskl, J. J.; del Rosarlo, A.; Gu1rgu1s, G. N.; Matias, V.;
Gramllch, J. W.; Kelly, W. R.; DeGarmo, T. E.; Coleman, G. C. (1983) Identification of
lead sources 1n California children using the stable Isotope ratio technique. Arch.
Environ. Health 38: 237-245.
Yankel, A. J.; von Lindern, I. H.; Walter, S. D. (1977) The Silver Valley lead study: the
relationship between childhood blood lead levels and environmental exposure. J. A1r
Pollut. Control Assoc. 27: 763-767.
Z1elhu1s, R. L.; del Castllho, P.; Herber, R. F. M.; Wibowo, A. A. E.; Salle", H. J. A. (1979)
Concentrations of lead and other metals 1n blood of two and three year-old children
living near a secondary smelter. Int. Arch. Occup. Environ. Health 42: 231-239.
11-211
-------
APPENDIX 11A
COMPARTMENTAL ANALYSIS
Many authors have noted that under conditions of constant lead exposure, blood lead con-
centrations change from one level to another apparent equilibrium level over a period of
several months. A mathematical model is helpful in estimating the new apparent equilibrium
level even when the duration of the experiment is not sufficiently long for this equilibrium
level to have been achieved. The model assumes that lead in the body is held in some number
of homogeneous and well -mixed pools or compartments. The compartments have similar kinetic
properties and may or may not correspond to identifiable organ systems. In a linear kinetic
model it is assumed that the rate of change of the mass of lead in compartment i at time t,
denoted X^ (t), is a linear function of the mass of lead in each compartment. Denote the frac-
tional rate of transfer of lead into compartment i from compartment j by K. . (fraction per
day), and let I..(t) be the total external lead input into compartment i at time t in units
such as ug/day. The elimination rate from compartment i is denoted K~.. The compartmental
model is
for each of the n compartments. If the inputs are all constant, then each X^t) is the sum of
(at most) n exponential functions of time (see for example, Jacquez, 1972).
For the one-compartment model
Ij - KQ1 Xt(t) (11-24)
with an initial lead burden X:(0) at time 0,
Xi(t) = Xj(0) exp(-KQlt) + [(l!/K01) (l-exp(-K01t)] (11-25)
The mass of lead at equilibrium is Ii/KQi ug. We may think of this pool as "blood lead". If
the pool has volume Vj then the equilibrium concentration is li/Kg! Vx ug/dl. Intake from
several pathways will have the form
Ix = At (Pb-AIr) + A2 (Pb-D1et)+ ' ' ' (11-26)
11A-1
-------
so that the long-term concentration is
VO Pb-Air + ' ' ' (11-27)
The inhalation coefficient is p = A^K^V!. The blood lead half-life is 0.693/KQl.
Models with two or more compartments will still have equilibrium concentrations in blood
and other compartments that are proportional to the total lead intake, and thus increase
linearly with increasing concentrations in air, dust, and diet. The relationship between the
exponential parameters and the fractional transfer coefficients will be much more complicated,
however.
Models with two or three pools have been fitted by Rabinowitz et al. (1976, 1977) and by
Batschelet et al. (1979). The pools are tentatively identified as mainly blood, soft tissue
and bone. But as noted in Section 11.4.1.1, the "blood" pool is much larger than the volume
of blood itself, and so it is convenient to think of this as the effective volume of distri-
bution for pool 1. A five-pool model has been proposed by Bernard (1977), whose pools are
mainly blood, liver, kidney, soft bones and hard bone.
The major conclusion of this Appendix is that linear kinetic mechanisms imply linear
relationships between blood lead and lead concentrations in environmental media. An extended
discussion of nonlinear kinetic mechanisms is given in Chapter 10, based on analyses in
Marcus (1985). One important mechanism involves an apparent limitation on the amount of
lead that can be absorbed by the red blood cells. However, at blood lead levels <30 ug/dl
this limitation does not greatly affect the linearity of the relationship between blood lead
and lead exposure.
11A-2
-------
APPENDIX 11B
FITTING CURVES TO BLOOD LEAD DATA
The relationship between blood lead and the concentrations of lead in various environ-
mental media is a principal concern of this chapter. It is generally accepted that the geo-
metric mean blood lead is some function, f, of the concentration of air lead and of lead in
diet, dust, soil, and other media. It has been observed that blood lead levels have a highly
skewed distribution even for populations with relatively homogeneous exposure, and that the
variability in blood lead is roughly proportional to the geometric mean blood lead or to the
arithmetic mean (constant coefficient of variation). Thus, instead of the usual model in
which random variations are normally distributed, a model is assumed here in which the random
deviations are multiplicative and lognormally distributed with geometric mean 1 and geometric
standard deviation (GSD) e°. The model is written
Pb-Blood = f (Pb-Air, etc.) eoz (11-28)
where z is a random variable with mean 0 and standard deviation 1. It has a Gaussian or
normal distribution. The model is fitted to data in logarithmic form
In(Pb-Blood) = In (f) (11-29)
even when f is assumed to be a linear function, e.g.,
f = p Pb-Air + pQ + pi Pb-Dust + ... (11-30)
The nonlinear function, fitted by most authors (e.g., Snee, 1982b), is a power function with
shape parameter \,
f = (p Pb-Air + p + Pi Pb-Dust + ...)A (11-31)
These functions can all be fitted to data using nonlinear regression techniques. Even when
the nonlinear shape parameter A. has a small statistical uncertainty or standard error as-
sociated with it, a highly variable data set may not clearly distinguish the linear function
(A = 1) from a nonlinear function (\ $ 1). In particular, for the Azar data set, the residual
sum of squares is shown as a function of the shape parameter \, in Figure 11B-1. When only a
11B-1
-------
9.3
9.2 r-
ffi
oc
<
a
w
u.
O
g.o
(A 8.9
Q
55
fla
O.O
8.7
8.6 J-
I T
MINIMUM SIGNIFICANT
DIFFERENCE FOR 1 DF
A =0.26
MINIMUM SIGNIFICANT
DIFFERENCE FOR 5 DF
SSE FOR In (Pb-Blood) = A In 3 Pb-Air + Z/3. C.)
5 5
SSE FOR In (Pb-Blood) = A In I/} Pb Air + Z/J C. +Z/?'. C. Age)
1 J J 1 J J
0.1 0.2 0.3 0.4 0.5 0-6 0.7 0.8 0.9 1.0
POWER EXPONENT. A
Figure 11 B-1. Residual sum of squares for nonlinear regression models for Azar
data (N = 149).
11B-2
-------
separate intercept (background) is assumed for each subpopulation, the best choice is A =
0.26; but when age is also used as a covariate for each subpopulation, then the linear model
is better. However, the approximate size of the difference in residual sum of squares
required to decide at the 5 percent significance level that a nonlinear model is better (or
worse) than a linear model is larger than the observed difference in sum of squares for any
A>0.2 (Gallant, 1975). Therefore, a linear model is used unless evidence of nonlinearity is
very strong, as with some of Kehoe's studies and the Silver Valley/Kellogg study. Non-
linearity is detectable only when blood lead is high (much above 35 or 40 ug/dl), and intake
is high, e.g., air lead much above 10 ug/m3. Additional research is needed on the relation-
ship between lead levels and lead intake from all environmental pathways.
11B-3
-------
APPENDIX 11C
ESTIMATION OF GASOLINE LEAD CONTRIBUTIONS TO ADULT
BLOOD LEAD BURDENS BASED ON ILE STUDY RESULTS
As discussed in Chapter 11 (pp. 11-118 to 11-123) the results of the Isotopic Lead Ex-
periment (ILE) carried out in Northern Italy provide one basis by which to estimate contribu-
tions of lead in gasoline to blood lead burdens of populations exposed in the ILE study area.
Figures 11C-1 to 5 of this appendix, reprinted from Facchetti and Geiss (1982), illustrate
changes in isotopic 206Pb/207Pb ratios for 35 adult subjects, for whom repeated measurements
were obtained over time during the ILE study. The percent of total blood lead in those sub-
jects contributed by Australian lead-labeled gasoline (petrol) used in automotive vehicles in
the ILE study area was estimated by the approach reprinted below verbatim from Appendix 17 of
Facchetti and Geiss (1982):
The main purpose of the ILE project was the determination of the contribution of petrol
lead to total lead in blood. A rough value for the fraction of petrol lead in blood can be
derived from the following equations:
R! X + f (1-X) = R1
R2 X + f (1-X) = R"
each of them referring to a given time at which equilibrium conditions hold.
R' and R" represent the blood lead isotopic ratios measured at each of the two times; if
R1 and R« represent the local petrol lead isotopic ratios measured at the same times, X is the
fraction of local petrol lead in blood due to petrols affected by the change in the lead
isotopic ratio, irrespective of its pathway to the blood i.e., by inhalation and ingestion
(e.g., from petrol lead fallout). The term (1-X) represents the fraction of the sum of all
other external sources of lead in the blood (any «other» petrol lead included), factor f
being the unknown isotopic ratio of the mixture of these sources. It is assumed that X and f
remained constant over the period of the experiment, which implies a reasonable constancy of
both the lead contributing sources in the test areas and the living habits which, in practice,
might not be entirely the case.
Data from individuals sampled at the initial and final equilibrium phases of the ILE
study together with petrol lead isotopic ratios measured at the same times, would ideally
provide a means to estimate X for Turin and countryside adults. However, for practical
reasons, calculations were based on the initial and final data of the subjects whose first
11C-1
-------
sampling was done not later than 1975 and the final one during phase 2. Their complete
follow-up data are shown in Table 27. For RI and R2 the values measured in the phases 0 and 2
of ILE were used (Rj = 1.186, RZ = 1.060). Hence, as averages of the individual X and f
results, we obtain:
Turin
countryside
<25 km
countryside
>25 km
K! = 0.237 ± 0.054
fi = 1.1560 ± 0.0033
X2 - 0.125 ± 0.071
f2 = 1.1542 i 0.0036
X3 - 0.110 ± 0.058
f3 = 1.1576 ± 0.0019
i.e 24%
i.e. 12%
i.e 11%
1.16 -
Figure 11C-1. Individual values of blood Pb-206/Pb-207 ratio
for subjects follow-up in Turin (12 subjects).
Source: Facchetti and Geiss (1982).
11C-2
-------
I I I I I I I I I I I I
£
8
£
1.16
1.15
1.14
1.13
1.12
-PHASE 0»
I I I
PHASE 1 »+
I I I I I
PHASE 2 H-
I I I I
74
75
76
77
78
79
80
Figure 11C-2. Individual values of blood zoepb/zoJRb ratio
for subjects follow-up in Castagnetto (4 subjects).
T
I I I
1.16
1.15
£
S
a 1.14
a.
1.13
1.12
DRUENTO
- FIANO
-PHASE 0.
I I T
-PHASE 1.
J I
-4*
I J_
-PHASE 2.
I I
I I
74
75
76
77
78
79
Figure 11C-3. Individual values of blood 2o6pb/207Pb ratio
for subjects follow-up in Druento and Fiano (6 subjects).
Source: Facchetti and Geiss (1982).
11C-3
-------
1.16
1.15
1.14
1.13
l.i:
I I 1 I I I I I I I I I I
\ v '
\ V V '-.'
- MOLE
SANTENA
.PHASE 2.
.*-. PHASE 0-^L, PHASE 1 »»h
I I l'~ I I I L I I I 1 I
74
75
76
77
78
79 80
Figure 11C-4. Individual values of blood 2°«Pb/207Pb ratio
lor subjects follow-up in Nole and Santena (9 subjects).
8
1.15
1.14
1.13
1.12
I
-PHASE 1
PHASE 0
L I l' I I I 'I I I I I 1
PHASE 2.
H*
74
75
76
77
78
79
80
Figure 11C-5. Individual values of blood 2ocpb/207p|, ratio
for subjects follow-up in Viu (4 subjects).
Source: Facchetti and Geiss (1982).
11C-4
t>U.S GOVERNMENT PRINTINfl OFFICE: i 9 8 6- 6 k 6- 1 1 8 / I* 0 6 » 3
------- |