United States
Environmental Protection
Agency
Office of Pollution
Prevention and Toxics
Washington. DC 20460
EPA 747-R-97-006
June. 1998
^y EPA R'sk Analysis to Support Standards for
Lead in Paint, Dust, and Soil
VOLUME I
Chapters 1 to 7
Appendix A
(,
^^e- X
^ -'
-------
EPA 747-R-97-006
June, 1998
RISK ANALYSIS TO SUPPORT STANDARDS
FOR LEAD IN PAINT, DUST, AND SOIL
VOLUME I
CHAPTERS 1 TO 7
APPENDIX A
Prepared
by
Battelle
505 King Avenue
Columbus, Ohio 43201
for
National Program Chemicals Division
Office of Pollution Prevention and Toxics
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
DISCLAIMER
Mention of trade names, products, or services does not
convey, and should not be interpreted as conveying, official EPA
approval, endorsement, or recommendation.
This report is copied on recycled paper.
-------
CONTRIBUTING ORGANIZATIONS
This study was funded and managed by the U.S. Environmental Protection Agency. The
risk analysis was conducted by Battelle Memorial Institute under contract to the Environmental
Protection Agency. Each organization's responsibilities are listed below.
Battelle Memorial Institute (Battelle)
Battelle was responsible for identifying and incorporating the results of relevant studies,
obtaining and managing relevant datasets, developing methodologies for risk assessment and risk
management, carrying out the risk analysis, and preparing the report. The Battelle Task Manager
was Ronald Menton. Robert Lordo, Nancy McMillan, and Nancy Niemuth were key contributors
to preparing the report. Brandon Wood was responsible for the statistical programming.
U.S. Environmental Protection Agency (EPA)
The Environmental Protection Agency was responsible for providing objectives of the
risk analysis, reviewing the developed methodology, contributing to the development of
conclusions, reviewing draft versions of the report, and managing the peer review and
publication of the report. The EPA Work Assignment Manager was Todd Holderman. The
Deputy Work Assignment Managers were Brad Schultz and Karen Lannon. The EPA Project
Officer was Sineta Wooten. Other EPA contributors include Barbara Leczynski, Janet Remmers,
John Schwemberger, Dave Topping, and the EPA/OPPT Risk Assessment Workgroup (chaired
by Lois Dicker).
HI
-------
TABLE OF CONTENTS
Page
EXECUTIVE SUMMARY ES-1
ES.1 BACKGROUND ES-1
ES.2 OBJECTIVES ES-1
ES.3 KEY ELEMENTS OF RISK ANALYSIS ES-2
ES.4 SUMMARY OF RESULTS ES-10
ES.4.1 RISK CHARACTERIZATION ES-10
ES.4.2 ANALYSIS OF EXAMPLE OPTIONS FOR RISK MANAGEMENT ES-12
ES.5 CONCLUSIONS ES-12
1.0 INTRODUCTION 1-1
1.1 Background 1-3
1.2 Statutory/Policy Constraints 1-4
1.3 Objectives , 1-6
1.4 Overview of Report 1-8
1.4.1 Organization of Report 1-8
1.4.2 General Approach 1-10
1.4.2.1 Approach for Risk Assessment 1-10
1.4.2.2 Approach for Risk Management 1-16
1.5 Peer Review 1-17
PART I - RISK ASSESSMENT
2.0 HAZARD IDENTIFICATION 2-1
2.1 Measures of Body-Lead Burden 2-3
2.2 Mechanisms of Lead Toxicity 2-5
2.2.1 Physiological Mechanisms 2-6
2.2.2 Neurotoxic Effects of Lead 2-6
2.2.3 Hematologic Effects of Lead 2-9
2.3 Health Effects of Lead Exposure 2-10
2.3.1 Neurological Effects of Lead 2-10
2.3.2 Other Effects of Lead 2-13
2.4 Representative Population 2-16
2.5 Selected Health Endpoints 2-18
2.5.1 Elevated Blood-Lead Concentration 2-18
2.5.2 IQ Point Deficits 2-19
2.6 Hazard Characterization 2-20
3.0 EXPOSURE ASSESSMENT 3-1
3.1 Sources and Pathways of Lead 3-4
3.2 Supporting Evidence in Lead Exposure Studies 3-9
3.2.1 Weight of Evidence on the Relationship Between '
Environmental-Lead Exposures and Increased Blood-Lead
Concentrations 3-9
3.2.2 More Detailed Description of the Most Useful Studies for this
Risk Assessment 3-13
iv
-------
TABLE OF CONTENTS
(Continued)
Page
3.2.2.1 Baltimore Repair and Maintenance (R&M) Study 3-24
3.2.2.2 Rochester Lead-in-Dust Study 3-26
3.2.2.3 Evaluation of the HUD Lead-Based Paint Hazard
Control Grant Program ("HUD Grantees") 3-28
3.2.2.4 Urban Soil Lead Abatement Demonstration Project (USLADP) . 3-31
3.2.2.5 Birmingham Urban Lead Uptake Study 3-33
3.2.2.6 Cincinnati Longitudinal Study 3-33
3.2.2.7 Brigham and Women's Hospital Longitudinal Study 3-34
3.3 Lead in Dust, Soil, and Paint in the Nation's Housing 3-35
3.3.1 The Distribution of Lead Levels in Household Dust, Soil, and Paint 3-35
3.3.1.1 HUD National Survey 3-36
3.3.1.2 The Baltimore Repair and Maintenance (R&M) Study 3-46
3.3.1.3 The Rochester Lead-in-Dust Study 3-49
3.3.1.4 Evaluation of the HUD Lead-Based Paint Hazard
Control Grant Program ("HUD Grantees") 3-54
3.3.1.5 Evaluating the Approach to Representing the Post-
1979 Housing Stock 3-61
3.3.2 Characterizing the Population of Children in the Nation's
Housing Stock 3-62
3.4 Distribution of Childhood Blood-Lead 3-62
3.4.1 Distribution of Blood-lead Concentration, as Measured by
NHANES III 3-63
3.4.2 Baltimore Repair and Maintenance (R&M) Study 3-69
3.4.3 Rochester Lead-in-Dust Study 3-70
3.4.4 Evaluation of the HUD Lead-Based Paint Hazard Control Grant
Program ("HUD Grantees") 3-71
3.5 Exposure Assessment Characterization 3-75
4.0 DOSE-RESPONSE ASSESSMENT 4-1
4.1 IEUBK Model 4-3
4.1.1 Description of the IEUBK Model 4-4
4.1.2 Inputs to the IEUBK Model 4-4
4.1.3 Estimating the Effect of Pica for Paint on Childhood Blood-Lead Levels . . 4-9
4.1.4 Estimating the National Distribution of Blood-Lead Using the
IEUBK Model 4-1iO
4.2 Empirical Model 4-11
4.2.1 Form of the Model 4-11
4.2.2 Variable Selection 4-12
4.2.3 Rochester Multimedia Model 4-14
4.2.4 Measurement Error Adjustment 4-15
4.2.5 Specification of the Empirical Model 4-16
4.2.6 Estimating the National Distribution of Blood-Lead Using the
Empirical Model 4-18
4.3 Utilizing Dust Lead Loadings 4-18
4.3.1 Wipe Versus Blue Nozzle (BN) Vacuum Conversions 4-19
4.3.2 Wipe Versus Baltimore Repair and Maintenance (BRM) Vacuum
Conversions 4-20
-------
TABLE OF CONTENTS
(Continued)
4.4 Health Outcomes 4-21
4.4.1 Decrements in IQ Scores 4-22
4.4.2 Increased Incidence of IQ Scores Less Than 70 4-25
4.5 Dose Response Characterization 4-29
5.0 RISK CHARACTERIZATION 5-1
5.1 Integrated Risk Analysis 5-4
5.1.1 Baseline Risk Characterization 5-7
5.1.2 Alternative Risk Characterization 5-11
5.2 Estimation of Risks Due to Background Exposure 5-14
5.3 Individual Risks 5-18
5.4 Risk Characterization Sensitivity and Uncertainty Analysis 5-23
5.4.1 Alternative Age Range of Children 5-25
5.4.2 Alternative Assumptions on Average IQ Score Decline Per Unit
Increase in Blood-Lead Concentration 5-28
5.4.3 Considering Potential Declines in Blood-Lead Concentration from
NHANES III Phase 2 Measures 5-29
5.4.4 Alternative Approach to Characterizing a Baseline Blood-Lead
Distribution from NHANES III Data 5-30
5.4.5 Uncertainty in Adjusting Dust-Lead Concentrations to Reflect the
Sample's Total Weight 5-31
5.4.6 Alternative Estimates for the Geometric Standard Deviation of
Blood-Lead Concentrations 5-34
5.4.7 Alternative Estimates for Daily Dietary Lead Intake Assumed in
Fitting the IEUBK Model 5-38
5.4.8 Alternative Assumptions on Paint Pica Tendencies in Children and
the Effect of Paint Pica on Blood-Lead Concentration 5-40
5.4.8.1 Empirical Model 5-41
5.4.8.2 IEUBK Model 5-42
5.4.9 Conclusions from Sensitivity Analysis 5-4,3
5.5 Risk Characterization Conclusions 5-44
PART II - RISK MANAGEMENT
6.0 ANALYSIS OF EXAMPLE OPTIONS FOR THE §403 STANDARDS 6-1
6.1 Intervention Activities 6-3
6.1.1 Interventions 6-5
6.1.2 Reductions in Environmental Lead Levels Following Interventions 6-7
6.1.3 Intervention Triggers 6-10
6.1.4 Reductions in Blood-Lead Levels Following Interventions 6-11
6.2 Methodology for Evaluating Risk Management Options 6-12
6.3 Results of the Evaluation of Example Risk Management Options 6-17
6.3.1 Evaluation of Example Dust Standards 6-19
6.3.2 Evaluation of Example Options for the Soil Standard 6-24
VI
-------
TABLE OF CONTENTS
(Continued)
Page
6.3.3 Evaluation of Example Options for the Trigger Levels of Paint
Intervention 6-29
6.3.4 Evaluation of the Effects of Varying Example Standard
Options for All Media 6-35
6.3.5 Risk Reduction Details for an Illustrative Set of Standards 6-40
6.4 Sensitivity and Uncertainty Analyses for Risk Management Analyses 6-43
6.4.1 Uncertainty in Converting Dust-Lead Loadings for Comparison
to Standards 6-43
6.4.2 Uncertainty in Converting Wipe Dust-Lead Loadings to Blue
Nozzle Dust-Lead Loadings for Determining Post-Intervention
Blood-Lead Distributions Using the Empirical Model 6-47
6.4.3 Alternative Assumptions on Post-Intervention Dust-Lead Loadings .... 6-48
6.4.4 Alternative Approach to Determining a Post-Intervention Blood-
Lead Concentration Distribution Using Directly-Measured Blood-
Lead Concentration Changes 6-50
6.4.5 Uncertainty in Assumptions Made in Determining Post-
Intervention Dust-Lead Concentrations 6-53
6.4.6 Alternative Estimates for the Geometric Standard Deviation of
Blood-Lead Concentrations 6-57
6.4.7 Alternative Estimates for Daily Dietary Lead Intake Assumed in
Fitting the IEUBK Model 6-59
6.4.8 Alternative Assumptions on Paint Pica Tendencies in Children
and the Effect of Paint Pica on Blood-Lead Concentration 6-60
6.4.8.1 Empirical Model 6-61
6.4.8.2 IEUBK Model 6-6,2
6.4.9 Standard Errors for Health Effect and Blood-Lead Concentration
Endpoints Due to Sampling Variability 6-64
6.5 Conclusion 6-67
7.0 REFERENCES 7-1
LIST OF APPENDICES
APPENDICES A THROUGH G
Appendix A Glossary A-1
Appendix B Health Effects Associated with Exposure to Lead and
Internal Lead Doses in Humans B-1
Appendix C1 Characterizing Baseline Environmental-Lead Levels in
the Nation's Housing Stock C1-1
Appendix C2 Method for Computing Confidence Intervals Associated with
Estimates in the Exposure Assessment and Risk Characterization .... C2-1
Appendix D1 Assumptions and Scientific Evidence to Account for
the Effect of Pica for Paint D1 -1
VII
-------
TABLE OF CONTENTS
(Continued)
Appendix D2 Results of Three Published Meta-Analyses on the
Relationship Between IQ Point Loss and Childhood
Blood-Lead Levels
Appendix E1 Methodology for Estimating Health Effects From
Blood-Lead Distribution
Appendix E2 Generating Distribution of Blood-Lead Concentrations
Based on Model-Predicted Geometric Mean and
Geometric Standard Deviation
Appendix F1 Methodology for Estimating Post-Intervention
Distribution of Children's Blood-Lead Concentrations
Resulting from Proposed §403 Rules
Appendix F2
Appendix G
Estimation of Primary Prevention Efficacy Using Model of
Bone-Lead Mobilization
Health Effects Associated with Exposure to Lead and Internal
Lead Doses in Humans
Page
D2-1
E1-1
E2-1
F1-1
F2-1
. G-1
LIST OF TABLES
l
Table ES-1. Interventions Defined for the §403 Risk Analysis, and the
Assumed Duration of Time During Which Lead Levels Are Reduced in
the Medium Targeted by the Intervention ES-6
Table ES-2. Information on the Two Statistical Models Used to Predict the Distribution
of Blood-Lead Concentration for Children Exposed to Specified
Environmental-Lead Levels ES-8
Table ES-3. Assumed Post-Intervention Lead Levels Affected By a Particular Intervention . . ES-9
Table ES-4. Estimated Baseline Number and Percentage of U.S. Children Aged 1-2 Years
Having Specific Health Effect and Blood-Lead Concentration Endpoints ES-11
Table ES-5. Examples of Environmental-Lead Levels Associated with a 5%
Likelihood that a Child Exposed to Such Levels Would Have a Blood-
Lead Concentration at or Above 10 A/g/dL ES-11
Table ES-6. Estimated Percentages of Occupied Housing Units in the 1997 U.S.
Housing Stock in Which Certain Interventions Are Triggered, for Three
Example Sets of Options for Dust and Soil Standards and Paint Triggers .... ES-13
Table ES-7a. Estimates of Health Effect and Blood-Lead Concentration Endpoints for
Children Aged 1 to 2 Years Under the IEUBK Model and Percent Declines
in These Estimates from Baseline, for Three Example Sets of Options for
Dust and Soil Standards and Paint Triggers ES-14
Table ES-7b. Estimates of Health Effect and Blood-Lead Concentration Endpoints for
Children Aged 1 to 2 Years Under the Empirical Model and Percent Decline
in These Estimates from Baseline, for Three Example Sets of Options for
Dust and Soil Standards and Paint Triggers ES-15
Table 2-1. Interpretation of Blood-Lead Concentration Categories and Follow-Up
Actions Recommended by CDC 2-19
viii
-------
TABLE OF CONTENTS
(Continued)
Page
Table 3-1. Childhood Lead Exposure Studies Conducted in Urban Communities That
Present Evidence of the Relationship Between Environmental-Lead Levels
and Blood-Lead Concentrations 3-11
Table 3-2. Childhood Lead Exposure Studies Conducted in Ore-Processing
Communities That Present Evidence of the Relationship Between
Environmental-Lead Levels and Blood-Lead Concentrations 3-12
Table 3-3a. Summary Information on Housing Surveyed in the Baltimore R&M Study
(pre-intervention phase), Rochester Lead-in-Dust Study, HUD Grantees
Program, and the HUD National Survey 3-15
Table 3-3b. Summary Information on Children Surveyed in the Baltimore R&M Study
(pre-intervention phase), Rochester Lead-in-Dust Study, HUD Grantees
Program, and the HUD National Survey 3-16
Table 3-3c. Information on Blood Sampling and Analysis in the Baltimore R&M Study
(pre-intervention phase), Rochester Lead-in-Dust Study, and the HUD
Grantees Program 3-18
Table 3-3d. Summary of Approaches for Soil Sampling and Analysis in the Baltimore
R&M Study (pre-intervention phase), Rochester Lead-in-Dust Study, HUD
Grantees Program, and the HUD National Survey 3-19
Table 3-3e. Summary of Approaches for Dust Sampling and Analysis in the Baltimore
R&M Study (pre-intervention phase), Rochester Lead-in-Dust Study, HUD
Grantees Program, and the HUD National Survey 3-20
Table 3-3f. Summary of Approaches for Paint Sampling and Analysis in the Baltimore
R&M Study (pre-intervention phase), Rochester Lead-in-Dust Study, HUD
Grantees Program, and the HUD National Survey 3-22
Table 3-4. Location of Grantees Participating in HUD Grantee Program Evaluation,
and Grantees' Criteria for Enrollment/Recruitment of Housing Units 3-30
Table 3-5. Estimated Total Number of Occupied Housing Units in the National Housing
Stock in 1997 According to Year-Built Category 3-37
Table 3-6. Summary of the Distribution of Lead Loadings in Floor-Dust Samples
Within Housing Units in the HUD National Survey, Weighted to Reflect
the Predicted 1997 Housing Stock 3-39
Table 3-7. Summary of the Distribution of Lead Concentrations in Floor-Dust
Samples Within Housing Units in the HUD National Survey, Weighted to
Reflect the Predicted 1997 Housing Stock 3-40
Table 3-8. Summary of the Distribution of Lead Loadings in Window Sill-Dust
Samples Within Housing Units in the HUD National Survey, Weighted to
Reflect the Predicted 1997 Housing Stock 3-40
Table 3-9. Summary of the Distribution of Lead Concentrations in Window Sill-Dust
Samples Within Housing Units in the HUD National Survey, Weighted to
Reflect the Predicted 1997 Housing Stock 3-41
Table 3-10. Summary of the Distribution of Soil-Lead Concentrations for Housing Units
in the HUD National Survey, Weighted to Reflect the Predicted 1997
Housing Stock 3-41
Table 3-11. Summary of XRF Paint Measurements Taken in the HUD National Survey,
Including the Percentage of Housing Units with Lead-Based Paint (LBP)
and Deteriorated LBP, by Component Category 3-42
ix
-------
TABLE OF CONTENTS
(Continued)
Page
Table 3-12. Summary of the Distribution of Observed Maximum XRF Lead Levels in
Paint for Housing Units in the HUD National Survey, Weighted to Reflect
the Predicted 1997 Housing Stock 3-43
Table 3-13. Predicted Numbers and Percentages of Units Having Lead-Based Paint in
the 1997 Occupied Housing Stock, Based on Information from the HUD
National Survey 3-43
Table 3-14. Imputed Environmental-Lead Measurements, by Age Category and
Presence of Lead-Based Paint, and Numbers of Units in the HUD
National Survey to Which Imputed Measurements Were Assigned in the
Risk Analyses 3-45
Table 3-15. Summary of Average Pre-lntervention Floor Dust-Lead Loading for
Occupied Housing Units in the Baltimore R&M Study 3-47
Table 3-16. Summary of Average Pre-lntervention Floor Dust-Lead Concentrations
for Occupied Housing Units in the Baltimore R&M Study 3-47
Table 3-17. Summary of Average Pre-lntervention Window Sill Dust-Lead Loading
for Occupied Housing Units in the Baltimore R&M Study 3-47
Table 3-18. Summary of Average Pre-lntervention Window Sill Dust-Lead
Concentrations for Occupied Housing Units in the Baltimore R&M Study 3-48
Table 3-19. Summary of Average Pre-lntervention Dripline Soil-Lead Concentrations
for Occupied Housing Units in the Baltimore R&M Study 3-48
Table 3-20. Summary of Observed Maximum XRF Paint-Lead Measurement at Pre-
lntervention for Occupied Housing Units Slated for R&M Intervention in
the Baltimore R&M Study 3-48
Table 3-21. Summary of Average Floor Dust-Lead Loading for Housing Units in the
Rochester Study 3-49
Table 3-22. Summary of Average Floor Dust-Lead Concentrations for Housing Units
in the Rochester Study 3-50
Table 3-23. Summary of Average Window Sill Dust-Lead Loading for Housing Units
in the Rochester Study 3-50
Table 3-24. Summary of Average Window Sill Dust-Lead Concentrations for Housing
Units in the Rochester Study 3-50
Table 3-25. Summary of Average Dripline Soil-Lead Concentrations for Housing Units
in the Rochester Study 3-51
Table 3-26. Summary of Average Soil-Lead Concentrations from Play Areas for
Housing Units in the Rochester Study 3-51
Table 3-27. Summary of XRF Paint Measurements Taken in the Rochester Lead-in-Dust
Study, Including the Percentage of Housing Units with Lead-Based Paint
(LBP) and Deteriorated LBP, by Component Category 3-52
Table 3-28. Summary of Observed Maximum XRF Paint-Lead Concentration for
Housing Units in the Rochester Study 3-53
Table 3-29. Summary of Area-Weighted Average Floor Dust-Lead Loadings (Pre-
lntervention, Using Wipe Collection Techniques) for Housing Units In the
HUD Grantee Program, According to Age of Unit and Grantee 3-56
Table 3-30. Summary of Area-Weighted Average Window Sill Dust-Lead Loadings
(Pre-lntervention, Using Wipe Collection Techniques) for Housing Units In
the HUD Grantee Program, According to Age of Unit and Grantee 3-57
-------
TABLE OF CONTENTS
(Continued)
Table 3-31. Summary of Play Area Soil-Lead Concentrations (Pre-lntervention) for
Housing Units In the HUD Grantee Program, According to Age of Unit
and Grantee 3-58
Table 3-32. Summary of Dripline Soil-Lead Concentrations (Pre-lntervention) for ,
Housing Units In the HUD Grantee Program, According to Age of Unit
and Grantee 3-59
Table 3-33. Summary of XRF Paint Measurements Taken in the HUD Grantee Program,
Including the Percentage of Housing Units with Lead-Based Paint (LBP)
and Deteriorated LBP, by Component Category 3-60
Table 3-34. Estimates of Geometric Mean Environmental-Lead Levels for HUD
National Survey Units Representing Post-1979 Housing in this Risk
Assessment and for Modern Urban Units in the Baltimore R&M Study 3-61
Table 3-35. Estimated Number of Children in the 1997 National Housing Stock, by
Age of Child and Year-Built Category 3-62
Table 3-36. Summary of Blood-Lead Concentration Data for Children Aged 1 -2 Years,
3-5 Years, and 1-5 Years, Based on NHANES III (Phase 2) Data 3-64
Table 3-37. Estimated Probabilities of Elevated Blood-Lead Concentrations in Children
Aged 1-2 Years, 3-5 Years, and 1-5 Years, Based on NHANES III
(Phase 2) Data 3-65
Table 3-38. Estimated Percentage of Children Aged 1 -2 Years (Within Selected
Subgroups) With Blood-Lead Concentrations At or Above 10 /jg/dL, and
the Geometric Mean and Geometric Standard Deviation of Blood-Lead
Concentration, Based on NHANES III (Phase 2) Data 3-66
Table 3-39. Estimated Geometric Mean Blood-Lead Concentration and Probabilities
of Elevated Blood-Lead Concentration in Children Aged 1 -2 Years, 3-5
Years, and 1-5 Years, by Age of Housing Unit, Based on NHANES III
(Phase 2) Data 3-67
Table 3-40. Estimated Percentage of Children With Blood-Lead Concentrations Exceeding
10 //g/dL, and the Geometric Mean and Geometric Standard Deviation of
Blood-Lead Concentration, for Children Aged 1 -2 Years According to Age
of Child's Residence and Either Family Income Level or Urban Status 3-68
Table 3-41. Summary Statistics on Blood-Lead Concentration Measured in the Initial
Round of Sampling in the Baltimore Repair and Maintenance Study 3-70
Table 3-42. Summary Statistics on Blood-Lead Concentration Measured in the
Rochester Lead-in-Dust Study 3-71
Table 3-43. Summary of Children's Blood-Lead Concentration in the HUD Grantees
Program, According to Blood Collection Method, Age of Child, and Grantee . . . 3-72
Table 3-44. Percentage of Children with Elevated Blood-Lead Concentration in the HUD
Grantees Program, According to Blood Collection Method, Age of Child,
and Grantee 3-73
Table 4-1. Summary of Default Parameter Values Used in the IEUBK Model
(Version 0.99D) 4-6
Table 4-2. Parameter Estimates and Associated Standard Errors for the Rochester
Multimedia Model 4-15
Table 4-3. Parameter Estimates and Associated Standard Errors for the Empirical
Model Used to Predict the National Distribution of Children's Blood-Lead
Concentration Based on Data from the HUD National Survey 4-17
XI
-------
TABLE OF CONTENTS
(Continued)
Table 4-4. Summary Information for Studies Included in the Schwartz (1994) Mela-
Analysis J 4-23
Table 4-5. Experts Who Participated in the Assessment of the Relationship Between
IQ Scores and Blood-Lead Levels by Wallsten and Whitfield 4-26
Table 4-6. Piecewise Linear Function for Estimating the Judged Increased Percentage
of Children Having IQ Scores Less Than 70 Due to Lead Exposure 4-26
Table 5-1. Estimated Baseline Number and Percentage of Children Aged 1-2 Years
Having Specific Health Effect and Blood-Lead Concentration Endpoints 5-9
Table 5-2. IEUBK and Empirical Model Predicted Health Effect and Blood-Lead
Concentration Endpoints for Children Aged 1 -2 Years 5-13
Table 5-3. IEUBK Model-Predicted Blood-Lead Concentrations for Children Aged 1-2
Years Under Different Soil-Lead Concentrations and at Dust-Lead
Concentrations Equal to Either 0 jjg/g or IEUBK Multiple Source Analysis
Default Values 5-17
Table 5-4. Percentage of Children Aged 1 -2 Years Having Specific Health Effect and
Blood-Lead Concentration Endpoints, Based on lEUBK-Predicted Blood-Lead
Concentrations Under Background Soil- and Dust-Lead Concentrations,
Compared with Estimates from the Baseline Risk Characterization 5-18
Table 5-5. Soil-Lead Concentrations at Which the Percentage of Children Aged 1-2
Years Having a Blood-Lead Concentration Above or Equal to 10 fjg/dL is
Estimated by the IEUBK Model at 1, 5, or 10%, Under Three Assumed
Dust-Lead Concentrations 5-20
Table 5-6. Floor Dust-Lead Loadings at Which the Percentage of Children Aged 1 -2
Years Having a Blood-Lead Concentration Above or Equal to 10 //g/dL is
Estimated by the Rochester Multimedia Model at 1, 5, or 10% Under Two
Assumed Soil-Lead Concentrations and Two Assumed Window Sill Dust-
Lead Loadings 5-21
Table 5-7. Window Sill Dust-Lead Loadings at Which the Percentage of Children Aged
1 -2 Years Having a Blood-Lead Concentration Above or Equal to 10 //g/dL is
Estimated by the Rochester Multimedia Model at 1, 5, or 10% Under Two
Assumed Soil-Lead Concentrations and Two Assumed Floor Dust-Lead
Loadings 5-23
Table 5-8. Components of the Risk Characterization Addressed By the Sensitivity
Analysis 5-24
Table 5-9a. Sensitivity Analysis for Estimated Baseline Number and Percentage of
Children Having Specific Health Effects and Blood-Lead Concentration
Endpoints for Two Age Groups of Children and Under Three Assumptions on
Average Decline in IQ Score per Unit Increase in Blood-Lead Concentration. . . . 5-26
Table 5-9b. Sensitivity Analysis for Estimated Baseline Average IQ Decrement for Two
Age Groups of Children and Under Three Assumptions on Average Decline
in IQ Score per Unit Increase in Blood-Lead Concentration 5-26
Table 5-10. Soil-Lead Concentration at Which the Percentage of Children Having a
Blood-Lead Concentration Above or Equal to 10 //g/dL is Estimated by
the IEUBK Model at 1, 5, or 10% Under Three Assumed Dust-Lead
Concentrations and Two Age Groups of Children 5-28
xii
-------
TABLE OF CONTENTS
(Continued)
Page
Table 5-11. Sensitivity Analysis for the Estimated Baseline Number and Percentage of
Children Aged 1 -2 Years Having Specific Health Effect and Blood-Lead
Concentration Endpoints, Assuming Various Percentage Declines in Blood-
Lead Concentration Since NHANES III Phase 2 5-30
Table 5-12. Sensitivity Analysis for Estimated Baseline Health Effect and Blood-Lead
Concentration Endpoints, for Children Aged 1 -2 Years, as Calculated Under
Two Approaches to Calculating the Baseline Distribution of Blood-Lead
Concentration Using NHANES III Data 5-32
Table 5-13. Sensitivity Analysis for Estimated Health Effect and Blood-Lead
Concentration Endpoints for Children Aged 1 -2 Years, as Calculated Using
the IEUBK Model and Under Four Approaches to Adjusting Dust-Lead
Concentrations for Low Tap Weight 5-34
Table 5-14. Sensitivity Analysis on the Model-Predicted, Pre-§403 Health Effect and
Blood-Lead Concentration Endpoints for Children Aged 1 -2 Years, Under
Three Alternative Values (1.4, 1.9, 2.1) for the Geometric Standard
Deviation (GSD) of the Blood-Lead Concentration Distribution and Under
the Value Used in the Risk Analysis (1.6) 5-36
Table 5-15. Sensitivity Analysis on IEUBK Model-Predicted Health Effect and Blood-Lead
Concentration Endpoints for Children Aged 1-2 Years, Under Three
Alternative Values (1.4, 1.9, 2.1) for the Geometric Standard Deviation
(GSD), Assuming a Background Soil-Lead Concentration of 20 //g/g and
One of Two Estimates of Background Dust-Lead Concentration 5-36
Table 5-16. Sensitivity Analysis on the Soil-Lead Concentrations at Which the
Percentage of Children Aged 1 -2 Years Having Blood-Lead Concentration at
Least 10/vg/dL is Estimated by the IEUBK Model at 1, 5, or 10%, Under
Three Assumed Dust-Lead Concentrations and for Alternative Assumptions
on the Geometric Standard Deviation (GSD) for the Blood-Lead Distribution. . . 5-37
Table 5-17. Sensitivity Analysis on the Floor Dust-Lead Loadings at Which the
Percentage of Children Aged 1 -2 Years Having Blood-Lead Concentration at
Least 10 //g/dL is Estimated by the Rochester Multimedia Model at 1, 5, or
10%, for Two Assumed Soil-Lead Concentrations and Two assumed Window
Sill Dust-Lead Loadings, and for Alternative Assumptions on the Geometric
Standard Deviation (GSD) for the Blood-Lead Distribution 5-37
Table 5-18. Sensitivity Analysis on the Window Sill Dust-Lead Loadings at Which the
Percentage of Children Aged 1 -2 Years Having Blood-Lead Concentration at
Least 10 //g/dL is Estimated by the Rochester Multimedia Model at 1, 5, or
10%, for Two Assumed Soil-Lead Concentrations and Two Assumed Floor
Dust-Lead Loadings, and for Alternative Assumptions on the Geometric
Standard Deviation (GSD) for the Blood-Lead Distribution 5-38
Table 5-19. Sensitivity Analysis on the IEUBK Model-Predicted, Pre-1403 Health Effect
and Blood-Lead Concentration Endpoints for Children Aged 1 -2 Years, Under
Two Alternative Values (1.29 fjg, 3.53 //g) for the Daily Lead Dietary Intake
Parameter and Under the Value Used in the Risk Analysis (5.78 /jg) 5-39
XIII
-------
TABLE OF CONTENTS
(Continued)
Page
Table 5-20. Sensitivity Analysis on the Soil-Lead Concentrations at Which the
Percentage of Children Aged 1 -2 Years Having Blood-Lead Concentration at
Least 10//g/dL is Estimated by the IEUBK Model at 1, 5, or 10%, for
Three Assumed Dust-Lead Concentrations and for Alternative Assumptions
on Daily Dietary Lead Intake 5-40
Table 5-21. Sensitivity Analysis on the Empirical Model-Predicted, Pre-1403 Health
Effect and Blood-Lead Concentration Endpoints for Children Aged 1 -2 Years,
Under Three Alternative Values (0%, 6%, 14%) for the Percentage of
Children with Paint Pica Tendencies., and Under the Value Used in the Risk
Analysis (9%) 5-41
Table 5-22. Sensitivity Analysis on the IEUBK Model-Predicted, Pre-1403 Health Effect
and Blood-Lead Concentration Endpoints for Children Aged 1-2 Years, Under
Three Alternative Sets of Assumptions on Paint Pica Effects, and Under the
Set of Assumptions Used in the Risk Analysis 5-43
Table 6-1. Interventions Defined for the §403 Risk Analysis Effort 6-6
Table 6-2. Expected Post-Intervention Lead Levels Associated With Performing §403
Interventions 6-8
Table 6-3. Intervention Triggers Defined for the §403 Risk Management Analyses 6-11
Table 6-4. Characterization of Impact of Various Example Options for Dust Standards:
Soil and Paint Standards Fixed (3,000 //g/g for Soil Removal, 5 ft2 Damaged
LBP for Paint Maintenance, 20 ft2 Damaged LBP for Paint Abatement) 6-18
Table 6-5. Characterization of Impact of Various Example Options for the Soil Standard:
Dust and Paint Standards fixed (100//g/fta for Floor Dust-Lead Loading, 500
//g/ft2 for Window Sill Dust-Lead Loading, 5 ft2 Damaged LBP for Paint
Maintenance, 20 ft2 Damaged LBP for Paint Abatement) 6-25
Table 6-6. Characterization of Impact of Various Options for Paint Intervention Triggers:
Example Dust and Soil Standards Fixed (100//g/ft2 for Dust-Lead Loading,
500//g/ft2 for Window Sill Dust-Lead Loading, 3,000//g/g for Soil Removal). . 6-30
Table 6-7. Characterization of Impact of Various Sets of Candidate Example Dust and
Soil, and Paint Intervention Triggers 6-36
Table 6-8. Estimated Distribution of Health Effect and Blood-Lead Concentration
Endpoints Prior to and After the Proposed §403 Rule for an Illustrative Set
of Standards 6-42
Table 6-9. Procedures for Which Alternative Assumptions Were Considered in
the Sensitivity Analysis Addressing Risk Management 6-44
Table 6-10. Number (and Percentage) of Units in the 1997 National Housing Stock
Projected to Exceed Various Combinations of Example Standards, As
Determined from Three Different Sets of Converted Dust-Lead Loadings 6-46
Table 6-11. Empirical Model-Predicted Post-§403 Health Effect and Blood-Lead
Concentration Endpoints for Children 1 -2 Years of Age, As Calculated Under
Three Assumptions on Post-Intervention Blue Nozzle Vacuum Dust-Lead
Loading 6-48
Table 6-12. Empirical Model-Predicted Post-§403 Percentages of Children Aged 1-2 Years
Experiencing Specific Health Effect and Blood-Lead Concentration Endpoints,
Under Various Assumptions on Post-Intervention Dust-Lead Loading 6-50
XIV
-------
TABLE OF CONTENTS
(Continued)
Table 6-13. Estimated Post-§403 Health and Blood-Lead Concentration Endpoints Based
on the Risk Assessment Approach and the Adjusted Blood-Lead Effects
Approach 6-54
Table 6-14. Geometric Mean Post-Intervention Floor Dust-Lead Concentration (//g/g),
and Percent Difference from Pre-lntervention Levels, for the Baltimore
R&M Study 6-56
Table 6-15. Geometric Mean Post-Intervention Floor Dust-Lead Concentration (//g/g), and
Percent Difference from Pre-lntervention Levels, for the Boston USLADP 6-56
Table 6-16. Sensitivity Analysis on the Estimated Post-§403 Health Effect and Blood-Lead
Concentration Endpoints for Children Aged 1 -2 Years, Under Three Alternative
Values (1.4, 1.9, 2.1) for the Geometric Standard Deviation (GSD) of the
Blood-Lead Concentration Distribution and Under the Value Used in the Risk
Analysis (1.6) 6-59
Table 6-17. Sensitivity Analysis on the IEUBK Model-Predicted Post-§403 Health Effect
and Blood-Lead Concentration Endpoints for Children Aged 1 -2 Years, Under
Two Alternative Values (1.29 fjg, 3.53 fjg) for the Daily Lead Dietary Intake
Parameter and Under the Value Used in the Risk Analysis (5.78^g) 6-60
Table 6-18. Sensitivity Analysis on the Empirical Model-Predicted Post-§403 Health Effect
and Blood-Lead Concentration Endpoints for Children Aged 1-2 Years, Under
Three Alternative Values (0%, 6%, 14%) for the Percentage of Children with
Paint Pica Tendencies, and Under the Value Used in the Risk Analysis (9%). . . 6-62
Table 6-19. Sensitivity Analysis on the IEUBK Model-Predicted Post-§403 Health Effect
and Blood-Lead Concentration Endpoints for Children Aged 1 -2 Years, Under
Three Alternative Sets of Assumptions on Paint Pica Effects, and Under the
Set of Assumptions Used in the Risk Analysis 6-65
Table 6-20. Estimates of Standard Errors Associated with Estimated Post-§403 Health
Effect and Blood-Lead Concentration Endpoints and with Number of Homes
Exceeding Standards, for Three Sets of Example Options for the §403
Standards 6-66
LIST OF FIGURES
Figure ES-1. Overview of the Risk Analysis Approach ES-3
Figure 1-1. Overview of the Risk Assessment and Risk Management Approach 1-11
Figure 2-1. Detailed Flowchart of the Approach to Hazard Identification 2-2
Figure 3-1. Detailed Flowchart of the Approach to Exposure Assessment 3-2
Figure 3-2. Pathways of Lead from the Environment to Humans, Main Organs of
Absorption and Retention, and Main Routes of Excretion 3-5
Figure 3-3. Set of Environmental Pathways Reported by Bornschein et al., 1986, Upon
Analysis of Data from the Cincinnati Longitudinal Study (Children Aged 18
Months) 3-6
Figure 3-4. Blood-Lead Concentration Versus Area-Weighted Arithmetic Average Floor
Dust-Lead Loading (Wipe Collection Method), for HUD Grantee and
Rochester Lead-in Dust Study Data 3.74
XV
-------
TABLE OF CONTENTS
(Continued)
Page
Figure 3-5. Blood-Lead Concentration Versus Area-Weighted Arithmetic Average
Window Sill Dust-Lead Loading, for the HUD Grantee, Rochester Lead-in-
Dust, and Baltimore R&M Studies 3-75"
Figure 4-1. Detailed Flowchart of the Approach to Dose-Response Assessment 4-2
Figure 4-2. IEUBK Model Predicted Geometric Mean Blood-Lead Concentration for
Children Aged 24 Months Plotted Separately Against Soil-Lead Concentration
and Dust-Lead Concentration for Fixed Default Values of the Remaining
Model Parameters 4-8
Figure 4-3. Estimated IQ Point Loss Due to Lead Exposure Plotted Against Concentration
of Lead in Soil and Dust, Utilizing IEUBK Model Predictions to Relate
Environmental Lead to Blood Lead 4-24
Figure 4-4. Judged Increase in Percentage of Children with IQ Below 70 Due to Lead
Exposure, Plotted Against Blood-Lead Concentration 4-27
Figure 4-5. Increase in Percentage of Children with IQ Below 70 Due to Lead Exposure
Plotted Against Concentration of Lead in Soil and Dust, Utilizing IEUBK
Model Predictions to Relate Environmental Lead to Blood Lead 4-28
Figure 4-6. Percentage of Children with Blood-Lead Concentration *10 and 20//g/dL
Due to Lead Exposure Plotted Against Geometric Mean Blood-Lead
Concentration, Assuming a GSD of 1.6 4-28
Figure 4-7. Percentage of Children with Blood-Lead Concentration ^10 and 20//g/dL
Due to Lead Exposure Plotted Against Concentration of Lead in Soil and
Dust Utilizing IEUBK Model Predictions to Relate Environmental Lead to
Blood Lead 4-29
Figure 5-1. Risk Characterization Overview 5-2
Figure 5-2. Summary of Risk Characterization Process 5-6
Figure 5-3. Baseline Distribution of Blood-Lead Concentrations Based on NHANES III,
Phase 2 (0.07 Percent of Children Had Blood-Lead Concentration Greater
than 32 j/g/dL) 5-8
Figure 5-4. Baseline Distribution of IQ Decrements Due to Elevated Blood-Lead
Concentration Based on NHANES III, Phase 2 (0.07 Percent of Children
Had in Excess of 8 IQ Points Lost) 5-10
Figure 5-5. Distribution of Blood-Lead Concentrations (//g/dL) for Children Aged 1-2
Years Based on NHANES III, and IEUBK and Empirical Model Predictions 5-12
Figure 5-6. Overview of Process for Estimating Risks Due to Background Lead Exposure. . 5-16
Figure 5-7. Percentage of Children's Blood-Lead Concentrations, as Predicted by the
IEUBK Model, That Will Exceed or Equal 10/yg/dL as a Function of Soil-
Lead Concentration for Three Dust-Lead Concentrations 5-20
Figure 5-8. Percentage of Children's Blood-Lead Concentrations, As Predicted By the
Rochester Multimedia Model, That Will Exceed or Equal 10//g/dL as a
Function of Floor Dust-Lead Loading for Two Soil-Lead Concentrations
and Two Window Sill Dust-Lead Loadings 5-22
Figure 6-1. Detailed Flowchart of the Approach to Risk Management 6-2
Figure 6-2. Calculation of Post-Intervention Floor Dust-Lead Concentration 6-10
Figure 6-3. Post-§403 Risk Management Process 6-13
Figure 6-4. Percentage of Homes Exceeding Example Candidate Dust Standards:
Soil Standard and Paint Intervention Trigger Held Fixed at Levels Given in
Table 6-4 6-21
XVI
-------
TABLE OF CONTENTS
(Continued)
Page
Figure 6-5a. Projected Health Endpoints Based on Various Example Options for Dust
Standards, Part 1; Soil Removal 3,000 //g/g, Paint Maintenance 5 ft2, Paint
Abatement 20 ft2. (Dashed reference line represents baseline risk.) 6-22
Figure 6-5b. Projected Health Endpoints Based on Various Example Options for Dust
Standards, Part 2; Soil Removal 3,000 //g/g, Paint Maintenance 5 ft2, Paint
Abatement 20 ft2. (Dashed reference line represents baseline risk.) 6-23
Figure 6-6. Percentage of the Nation's Homes Expected to Exceed Various Example
Candidate Soil Standards: Dust Standard and Paint Intervention Triggers
Held Fixed at Levels Given in Table 6-5 6-26
Figure 6-7a. Projected Health Endpoints Based on Various Example Options for the Soil
Standard, Part 1; Floor Dust 100 //g/ft2, Window Sill Dust 500 //g/ft2, Paint
Maintenance 5 ft2, Paint Abatement 20 ft2. (Dashed reference line
represents baseline risk.) 6-27
Figure 6-7b. Projected Health Endpoints Based on Various Options for the Soil Standard,
Part 2; Floor Dust 100//g/ft2, Window Sill Dust 500//g/ft2, Paint
Maintenance 5 ft2, Paint Abatement 20 ft2. (Dashed reference line
represents baseline risk.) 6-28
Figure 6-8. Percentage of Homes Exceeding Candidate Interior Paint Intervention Triggers:
Dust and Soil Example Standards Held Fixed at Levels Given in Table 6-6. . . . 6-31
Figure 6-9. Percentage of Homes Exceeding Candidate Exterior Paint Intervention
Triggers: Dust and Soil Example Standards Held Fixed at Levels Given in
Table 6-6 6-31
Figure 6-1 Oa. Projected Health Endpoints Based on Various Options for Paint Intervention
Triggers Part 1; Floor Dust 100 //g/ft2, Window Sill Dust 500 //g/ft2, Soil
Removal 3,000//g/g. (Dashed reference line represents baseline risk.) 6-33
Figure 6-1 Ob. Projected Health Endpoints Based on Various Options for Paint Intervention
Triggers, Part 2; Floor Dust 100//g/ft2, Window Sill Dust 500//g/ft2, Soil
Removal 3,000 //g/g. (Dashed reference line represents baseline risk.) 6-34
Figure 6-11 a. Projected Health and Blood-Lead Concentration Endpoints Based on Various
Example Sets of Options for Dust and Soil Standards, and Paint Intervention
Triggers, Part 1. (Dashed reference line represents baseline risk.) 6-38
Figure 6-11b. Projected Health Endpoints Based on Various Example Sets of Options for
Dust and Soil Standards, and Paint Intervention Triggers, Part 2. (Dashed
reference line represents baseline risk.) 6-39
Figure 6-12. Projected Post-Intervention Blood-Lead Concentration Distributions Based on
Empirical and IEUBK Models at Standards of Floor Dust-Lead - 100 //g/ft2;
Window Sill Dust-Lead - 500 //g/ft2; Soil-Lead Concentration - 2,000 //g/g;
Paint Maintenance - 5 ft2 Damaged LBP; and Paint Abatement - 20 ft2
Damaged LBP 6-41
Figure 6-13. Predicted Versus Observed Average Post-Intervention Floor Dust-Lead
Concentration (//g/g) (Boston USLADP Study Group Homes) 6-55
Figure 6-14. Predicted Versus Observed Average Post-Intervention Floor Dust-Lead
Concentration (//g/g) (Baltimore R&M Level III Homes) 6-55
Figure 6-15. Average Floor Dust-Lead Concentration Versus Average Fine Soil-Lead
Concentration (Baltimore USLADP Homes) 6-58
Figure 6-16. Predicted Versus Observed Average Post-Intervention Floor Dust-Lead
Concentration (//g/g) (Baltimore USLADP Treatment Group Homes) 6-58
xvii
-------
This page intentionally blank.
-------
EXECUTIVE SUMMARY
ES.1 BACKGROUND
Lead poisoning in children is recognized as a major health problem in the United States.
While there are many sources of lead in the human environment, lead-based paint hazards in
residential housing are considered the primary source of lead exposure for children. To help
develop a national strategy to eliminate lead-based paint hazards, the President of the United
States signed into law the Residential Lead-Based Paint Hazard Reduction Act of 1992 (42
U.S.C. 4851). This legislation included an amendment to the Toxic Substances Control Act
(Title IV: Lead Exposure Reduction), requiring the Administrator of the U.S. Environmental
Protection Agency (EPA) to enact a variety of activities to identify and reduce environmental
exposure to lead hazards. Specifically, §403 of TSCA (15 U.S.C. 2683) states:
"... the Administrator shall promulgate regulations which shall identify, for
purposes of this title and the Residential Lead-Based Paint Hazard Reduction Act
of 1992, lead-based paint hazards, lead-contaminated dust, and lead-contaminated
soil."
Under §403, the Agency is required to identify what constitutes a lead-based paint hazard (i.e.,
conditions that cause exposure to lead-contaminated dust, soil, or paint that would result in
adverse health effects to humans) and what constitutes lead contamination of dust and soil (i.e.,
the presence of lead levels which can pose a threat of adverse health effects). In particular, the
§403 rule to be established by the Agency will set standards for lead levels in dust and soil to
determine 1) whether a residential environment has lead-contaminated dust and soil, and 2)
whether a lead-based paint hazard is present in a residential environment.
ES.2 OBJECTIVES
This report presents the methods and findings of a risk analysis, which provides a
scientific foundation for the regulatory standards that the Agency will establish in response to
§403. This risk analysis consists of two parts. Part I (Chapters 2 through 5) constitutes the risk
assessment, or EPA's assessment of the health risks to young children from exposures to lead-
based paint hazards, lead-contaminated dust, and lead-contaminated soil in the nation's housing.
Part n (Chapter 6) constitutes an analysis of risk management options, which includes the
Agency's approach to estimating how these risks are reduced following promulgation of the §403
rule and illustrates use of this methodology for a broad range of example options for the §403
standards.
The objective of the risk assessment is to characterize baseline health risks to young
children from specific residential exposures to lead. The term baseline (or "pre-§403") refers to
conditions hi 1997, prior to promulgating any rule in response to §403. The objectives of risk
management are to develop and apply methodology to determine how risks are expected to be
reduced from baseline levels because of interventions conducted in response to the §403 rule (or
ES-1
-------
"post §403"), and to develop an approach to estimate numbers of children and housing units that
would be directly impacted by the rule. Information presented hi this risk analysis will ultimately
be used to consider various standards for rulemaking and as input to the Regulatory Impacts
Analysis (RIA) for the proposed rule, as well as any interim economic cost-benefit analyses.
ES.3 KEY ELEMENTS OF RISK ANALYSIS
Figure ES-1 provides an overview of the risk analysis conducted for the §403 regulation.
In the risk assessment, hazard identification, exposure assessment, and dose-response assessment
provide necessary information to risk characterization, where the baseline distribution of blood-
lead concentrations for U.S. children aged 12 to 35 months (cited as "aged 1-2 years" hi this
report) is determined, along with baseline health risks resulting from residential exposure to lead.
This overview focuses on the key issues addressed by this risk analysis in performing risk
characterization and risk management. These key issues include: examining the population of
interest; selecting the measurement and health endpoints; identifying data sources for this risk
analysis; defining intervention strategies for this analysis; and applying statistical models hi the
risk analysis.
Population of interest
While the health risk associated with lead exposure is significant for all humans, young
children are most sensitive in this regard, hi particular, the scientific literature indicates that the
following tend to be prevalent hi children aged 1-2 years:
a high level of hand-to-mouth activity, which increases the potential for ingesting
lead-contaminated dust, soil, and paint
a rapidly-developing central nervous system, making it highly susceptible to the
effects of lead
a peaking of the synaptic density of the frontal cortex of the brain; synaptic
development can be disrupted or delayed as a result of lead exposure
Also, the scientific literature best characterizes the relationship between childhood blood-lead
concentration and intelligence quotient (IQ) score within this age group (approximately a 0.25
drop hi IQ score is predicted for every 1.0 |ig/dL increase hi blood-lead concentration). These
neurotoxicological effects of lead exposure at this age may be irreversible. Therefore, as the
health effect and blood-lead concentration endpoints considered hi this risk analysis have high
sensitivity and are well characterized for this age group, and this age group is representative of
the population addressed by the statute, the population of interest for this risk analysis was U.S.
children aged 1-2 years.
ES-2
-------
Background
and
Objectives
(Chapter 1)
Hazard
Identification
(Chapter 2)
Exposure
Assessment
(Chapter 3)
Dose-Response
Assessment
(Chapter 4)
Risk
Characterization
(Chapter 5)
Risk
Management
(Chapter 6)
Conclusions on
Risk
Characterization
Conclusions on
Analysis of Example
Options for §403
Standards
Figure ES-1. Overview of the Risk Analysis Approach.
ES-3
-------
Selecting the measurement and health endpoints
To characterize the health risk associated with lead exposures to children aged 1-2 years,
this risk analysis considered the following: elevated blood-lead concentration and IP point
deficit. Blood-lead concentration is a measurement endpoint used hi many previous studies to
quantify the health effect associated with lead exposure, hi their guidelines for childhood lead
poisoning prevention, the Centers for Disease Control and Prevention (CDC) indicates that
procedures to prevent adverse health effects from lead exposure are triggered when a child's
blood-lead concentration exceeds specified thresholds. IQ point deficit is one of the many
adverse health effects resulting from lead exposure and is used in this report to represent the
neurotoxicological effects of lead. The risk characterization consisted of the following blood-
lead concentration and health effect endpoints:
The incidence of blood-lead concentration greater than or equal to 10 ug/dL1
The incidence of blood-lead concentration greater than or equal to 20 |ig/dL2
The incidence of IQ score less than 703 resulting from childhood lead exposure
The likelihood of an IQ score decline greater than or equal to 1 point due to
childhood lead exposure
The likelihood of an IQ score decline greater than or equal to 2 points due to
childhood lead exposure
The likelihood of an IQ score decline greater than or equal to 3 points due to
childhood lead exposure
Average IQ score decline in a child as a result of childhood lead exposure.
The latter four endpoints provide information on the distribution of IQ point deficit resulting
from lead-based paint hazards.
1 A blood-lead concentration of 10 iig/dL is the threshold at which CDC recommends frequent monitoring
of the child and community-wide lead poisoning prevention activities. Lead-related reductions in intelligence,
impaired hearing activity, and interference with Vitamin D metabolism have been documented in children with
blood-lead concentrations as low as 10 ug/dL.
2 A blood-lead concentration of 20 ug/dL is the threshold at which CDC recommends a complete medical
evaluation, an environmental assessment, and necessary environmental remediation for the child and his/her
environment Increased blood pressure, delayed reaction times, anemia, and kidney disease are among the adverse
health effects seen at this level.
3 An IQ score of 70 is two standard deviations below the population mean and is used as an indicator of
mental retardation.
ES-4
-------
Identifying data sources for this risk analysis
Data on the national distribution of blood-lead concentration in children aged 1-2 years
and on environmental-lead levels in the nation's housing stock were obtained to address the
above objectives. To the extent that they were available, data from recent, nationally
representative surveys were used in this risk analysis.
The national distribution of baseline blood-lead concentrations in children aged 1-2 years
was determined from data collected in Phase 2 of the third National Health and Nutrition
Examination Survey (NHANES HI), conducted from 1991-1994. Without disparaging the
national representativeness of the NHANES HI data, the use of NHANES HI data in this risk
analysis has the following limitations: blood-lead concentration data were sampled so that
potential effects on seasonality will be missed, and any further reduction in blood-lead
concentration that may have occurred between the middle of the survey and 1997 will not be
captured.
Data on numbers of housing units and on children within specified housing groups were
obtained from sources provided by the U.S. Bureau of the Census, such as the American Housing
Survey. In addition, information used to translate blood-lead concentration to the above health
effect endpoints was obtained from peer-reviewed articles in refereed journals.
Recognized as the leading source of data on environmental-lead levels in residential
environments, the National Survey of Lead-Based Paint in Housing, conducted from 1989-1990
by the U.S. Department of Housing and Urban Development (HUD), was the primary source of
data on baseline environmental-lead levels in dust and soil in the nation's housing stock. The
design and findings of the HUD National Survey have been peer reviewed and published in
several government reports. However, there are limitations associated with using the HUD
National Survey data: limited numbers of environmental samples were taken at each housing
unit, only 284 houses were sampled (which were all built prior to 1980), the study was conducted
over five years ago, and a dust collection device other than the wipe collection method being
adopted by the §403 rules was used. These limitations contribute to overall uncertainty in results
within the risk analysis.
Defining intervention strategies for this analysis
As part of the approach to determine how environmental-lead levels may change upon
implementing standards hi response to §403, the Agency identified a limited set of intervention
activities that were assumed to occur at housing units identified as exceeding one or more
standards. A brief description of each intervention activity is provided in Table ES-1, along with
when the activity is triggered at a specific residence and the assumed durations on the efficacy of
the intervention. This report does not attempt to provide detailed protocols on how each
intervention should be conducted. While the durations hi Table ES-1 were based on review of
available data hi the scientific literature, it should be understood that most published studies on
intervention effectiveness provide little or no information on efficacy beyond one or two years
after performance of the intervention activities.
ES-5
-------
Table ES-1. Interventions Defined for the §403 Risk Analysis, and the Assumed Duration
of Time During Which Lead Levels Are Reduced in the Medium Targeted by
the Intervention.
Intervention
Dust cleaning
Soil removal
Abatement of
exterior
lead-based paint
Maintenance of
exterior
lead-based paint
Abatement of
interior
lead-based paint
Maintenance of
interior
lead-based paint
When the Intervention is
Assumed to be Triggered1
After any interior paint
intervention, after soil removal, or
when dust-lead loadings on floors
or window sills exceed hazard
standards for dust
When average soil-lead
concentration for the entire yard
exceeds the hazard standard for
soil
When the level of lead in
deteriorated paint on exterior
surfaces exceeds the hazard
standard for paint, and the
amount (square footage) of such
paint warrants an abatement
When the level of lead in
deteriorated paint on exterior
surfaces exceeds the hazard
standard for paint, but the amount
(square footage) of such paint
does not warrant an abatement
When the level of lead in
deteriorated paint on interior
surfaces exceeds the hazard
standard for paint, and the
amount (square footage) of such
paint warrants an abatement
When the level of lead in
deteriorated paint on interior
surfaces exceeds the hazard
standard for paint, but the amount
(square footage) of such paint
does not warrant an abatement
Assumed Procedures Defining
the Intervention
Clean the unit using HEPA vacuums
and wet mopping
Soil from areas with elevated lead
concentrations is removed and
replaced with clean soil, or the areas
are permanently covered. A dust
cleaning intervention is assumed to
follow soil removal.
Deteriorated lead-based paint is
removed, and the affected surface
enclosed or encapsulated, if
necessary, using currently acceptable
practices and materials
Painted surfaces with deteriorated
lead-based paint are repaired by
feathering the edges of deteriorating
paint and repainting with new, lead-
free paint
Deteriorated lead-based paint is
removed, and the affected surface
enclosed or encapsulated, if
necessary, using currently acceptable
practices and materials. A dust
cleaning intervention is assumed to
follow this intervention.
Painted surfaces with deteriorated
lead-based paint are repaired by
feathering the edges of deteriorating
paint and repainting with new, lead-
free paint. A dust cleaning
intervention is assumed to follow this
intervention.
Assumed
Duration
Dust: 4 years
or permanent2
Soil:
Permanent
Dust:
Permanent
Paint:
20 years
Paint:
4 years
Paint:
20 years
Dust:
permanent2
Paint: 4 years
Dust: 4 years
1 The term "hazard standards" refers to standards to be established under §403 regulations against which to
compare a residential environment when evaluating the presence and magnitude of lead-based paint hazards.
These standards are expected to specify the condition and location of lead-based paint, and lead levels in
dust and soil.
2 Duration is assumed permanent if cleaning is accompanied by paint and soil abatements (20 years if
accompanied by paint abatement).
ES-6
-------
Applying statistical models in the risk analysis
In characterizing the risks posed by lead exposure to the nation's population of children
aged 1-2 years, the risk management analyses employed two statistical models: the Agency's
Integrated Exposure, Uptake, and Biokinetic (EEUBK) Model for Lead in Children (version
0.99D), and an empirical regression model developed for this risk analysis. While both models
use environmental-lead levels as measured in the HUD National Survey (as well as other factors)
to estimate a geometric mean blood-lead concentration for children aged 1-2 years in the U.S.
(not for individual children), they function and behave differently. The EEUBK model has been
studied extensively, has been utilized at a wide number of sites, and has undergone peer review
by EPA's Science Advisory Board. However, the EEUBK model was developed for applications
that differ somewhat from this study. While the empirical model was developed specifically for
this study, it is based on data collected in a single lead exposure study (The Rochester (NY)
Lead-in-Dust study, as documented in Lanphear et al., 1995), has not undergone formal peer
review, has not been applied elsewhere, and has not been studied hi depth. Table ES-2
summarize the characteristics of these two models.
The empirical model was developed to address some aspects of lead exposure that were
important to the risk management analyses but could not be directly addressed by the EEUBK
model. These include predicting blood-lead concentration based on dust-lead loadings rather
than concentrations, using data on lead loadings hi window sill dust as well as floor dust, and
representing the effect of pica tendency in the presence of deteriorated lead-based paint.
However, the empirical model has its own limitations, such as the utility of using data from a
single study to be nationally representative and the differences hi sampling methodology between
this study and the HUD National Survey. As no single model is optimal for application within
this risk analysis, use of these models allows two sets of results to be observed, with each set
having different advantages and limitations.
In the risk management methodology to evaluate example options for §403 standards, the
EEUBK and empirical models were each used to obtain two distributions of blood-lead
concentration: one resulting from exposure to environmental-lead levels existing prior to the
proposed rulemaking (pre-§403), and one resulting from lead exposures following interventions
that may occur at a residence in response to the proposed rule (post-§403). Then, for a given
model, the extent to which the pre-§403 and post-§403 distributions differed was characterized
by noting the percentage difference between key distribution parameters (the geometric mean and
geometric standard deviation) for the two distributions. These differences were then applied to
values of the parameters under the baseline distribution of blood-lead concentration (i.e., the pre-
§403 distribution determined from Phase 2 of NHANES ID). The result was an estimated
national distribution of blood-lead concentrations that represented conditions after implementing
the proposed §403 rule and that was directly comparable to the baseline distribution.
ES-7
-------
Table ES-2. Information on the Two Statistical Models Used to Predict the Distribution of
Blood-Lead Concentration for Children Exposed to Specified Environmental-
Lead Levels.
IEUBK model
Empirical model
Input parameters and
source of data inputs
Mass-weighted floor dust-lead
concentration (//g/g) for the unit (as
estimated from HUD National Survey
data, assuming Blue Nozzle vacuum
collection techniques)
Average soil-lead concentration (//g/g)
for the entire yard (as estimated from
HUD National Survey data)
IEUBK model default values were
used for all other model parameters:
lead in air, lead intake rates (from
diet, water, soil, dust), and
absorption
Area-weighted floor dust-lead loading
Oug/ft2) for the unit (as estimated
from HUD National Survey data,
assuming Blue Nozzle vacuum
collection techniques)
Area-weighted window sill dust-lead
loading (//g/ft2) for the unit (as
estimated from HUD National Survey
data, assuming Blue Nozzle vacuum
collection techniques)
Average soil-lead concentration
(//g/g) for the entire yard (as
estimated from HUD National Survey
data)
Categorical variable indicating the
extent to which deteriorated lead-
based paint was present and whether
the child puts paint chips in his/her
mouth
Endpoint predicted by
the model (under
conditions specified
by the input parameter
values)
Geometric mean blood-lead
concentration for children, over the
period of birth to seven years of age.
Model predictions at age 24 months
were used in risk management
analyses.
Geometric mean blood-lead
concentration for a group of children
aged 12-31 months (the age range
represented in the data used to develop
the model)
Source of data for
developing the model
Many scientific studies
Data from the Rochester Lead-in-Dust
Study (Rochester, NY; 84% of the
sampled units built prior to 1940,
approximately 40% of the sampled
children were African American)
Status of model
evaluation
The IEUBK model is peer reviewed and
is recommended as a risk assessment
tool to support OSWER Interim
Directive on Revised Soil Lead Guidance
for CERCLA Sites and RCRA Facilities
Has not undergone model evaluation
nor formal peer review except as part
of this document
Baseline values for the seven health effect and blood-lead concentration endpoints
defined earlier were calculated from the baseline distribution of blood-lead concentration.
Similarly, endpoint values under post-§403 conditions were calculated from the post-§403 blood-
lead concentration distribution. These two sets of endpoint values represent population-based
risks of lead exposure to children aged 1-2 years. It was of interest to determine how risks
declined from baseline to post-§403 conditions.
ES-8
-------
The IEUBK model and a third statistical model, the Rochester multimedia model, were
used to determine individual risks, or risks associated with children exposed to specified
environmental-lead levels. The Rochester multimedia model was developed as an intermediate
step in developing the empirical model from the Rochester Lead-in-Dust study data. This model
predicts blood-lead concentration as a function of wipe dust-lead loadings on floors and window
sills, dripline soil-lead concentration, and an indicator of paint/pica hazard. Soil-lead
concentration is considered at the dripline as most of the Rochester study units had soil-lead
concentration data only from the dripline. Individual risk characterization focused on the
probability that a child has a blood-lead concentration greater than or equal to 10 ^ig/dL. The
Rochester multimedia model was used to predict this probability as a function of dust-lead
loadings on floors and window sills, where dripline soil-lead concentration is assumed fixed at a
specified value. The IEUBK model was used to predict this probability as a function of average
soil-lead concentration for the yard, where dust-lead concentration is assumed fixed at a specified
value.
When the interventions in Table ES-1 are performed in a given housing unit, the
risk management analyses made assumptions on post-intervention lead levels for the media
affected by each intervention. These assumptions are documented in Table ES-3.
Table ES-3. Assumed Post-Intervention Lead Levels in Media Affected by a Particular
Intervention.
intervention
Dust cleaning
Soil removal
Abatement of exterior
lead-based paint
Maintenance of exterior
lead-based paint
Abatement of interior
lead-based paint
Maintenance of interior
lead-based paint
Assumed Post-Intervention Lead Levels tot Affected Medfa '
Floor dust-lead loading - the lower of 40 //g/ft2 and the pre-intervention level
Floor dust-lead cone, is determined by the methods documented in Section 6.1
Window sill dust-lead loading - the lower of 1 00 //g/ft2 and the pre-intervention
level
Soil-lead cone. = 1 50 //g/g in areas where soil removal is conducted
Floor dust-lead loading = the lower of 40 //g/ft2 and the pre-intervention level
Floor dust-lead cone, is determined by the methods documented in Section 6.1
Window sill dust-lead loading = the lower of 100 //g/ft1 and the pre-intervention
level
0 ft2 of deteriorated exterior lead-based paint
0 ft2 of deteriorated exterior lead-based paint
0 ft2 of deteriorated interior lead-based paint
Floor dust-lead loading = the lower of 40 //g/ft2 and the pre-intervention level
Floor dust-lead cone, is determined by the methods documented in Section 6.1
Window sill dust-lead loading - the lower of 100 //g/ft2 and the pre-intervention
level
0 ft* of deteriorated interior lead-based paint
Floor dust-lead cone, is determined by the methods documented in Section 6.1
ES-9
-------
ES.4 SUMMARY OF RESULTS
ES.4.1 RISK CHARACTERIZATION
The risk characterization was performed using the year 1997 as a point of reference. This
year represents a baseline point of reference for environmental-lead levels and health effect and
blood-lead concentration endpoints assuming any interventions or other activities conducted in
response to the §403 standards would not yet have been implemented. Using data from the
American Housing Surveys, this risk characterization predicts that the 1997 national housing
stock contains 99,272,000 occupied housing units containing nearly eight million children aged
1 to 2 years. Nearly six million of these children reside within the nearly 75 million housing
units built prior to 1980, which are of most importance to the §403 rule. As estimated by the
National Survey of Lead-Based Paint in Housing, approximately 83% of pre-1980 housing is
estimated to contain lead-based paint, and 18% is estimated to contain non-intact lead-based
paint, defined in the HUD National Survey as greater than 5 ft2 of peeling, chipping, or otherwise
deteriorated lead-based paint.
Baseline estimates of the blood-lead concentration and health effect endpoints associated
with children aged 1 to 2 years are provided in Table ES-4. These estimates are determined by
assuming that the distribution of blood-lead concentration in children aged 1-2 years is lognormal
with a geometric mean of 3.14 ug/dL and a geometric standard deviation of 2.09 ug/dL (as
estimated from the NHANES III Phase 2 data) and by assuming specified relationships between
blood-lead concentration and IQ decrement. According to this table, elevated blood-lead
concentrations continue to be present within children aged 1-2 years; approximately 458,000
children (5.75%) aged 1-2 years are estimated to have blood-lead concentrations that exceed 10
Hg/dL. Approximately 9,150 children are expected to have an IQ score below 70 as a result of
their exposure to lead.
Individual risks vs. population-based risks
The estimates in Table ES-4 represent risk to the entire population of children aged 1-2
years, given exposure to baseline levels of lead hi the nation's housing stock (i.e., population-
based risks). However, it is also of interest to characterize the probability of experiencing an
elevated blood-lead concentration for children exposed to specific levels of lead in dust and soil
(i.e., individual risks). Table ES-5 provides examples of estimated environmental-lead levels in
soil and dust associated with a 5% likelihood that children exposed to such levels will have a
blood-lead concentration at or above 10 ug/dL. This table indicates that the 5% likelihood is
unachievable when floor dust-lead concentrations are as high as 500 |ag/g. In addition, when
dripline soil-lead concentrations range from 100 to 400 ug/g and window sill dust-lead loadings
range from 200 to 500 ug/ft2, floor dust-lead loadings must be less than 10 ug/ft2 to achieve the
5% likelihood.
ES-10
-------
Table ES-4. Estimated Baseline Number and Percentage of U.S. Children Aged 1-2 Years
Having Specific Health Effect and Blood-Lead Concentration Endpoints.
Health Effect and Blood-Lead
Concentration Endpoints
Blood-lead concentration greater than or equal
to 20 //g/dL
Blood-lead concentration greater than or equal
to 1 0 //g/dL
IQ score less than 701
IQ score decrement of at least 1 '
IQ score decrement of at least 21
IQ score decrement of at least 31
Average IQ decrement in a child, resulting from
lead exposure
Baseline Estimates
Number of Children With
the Attribute
46,800
458,000
9,150
3,060,000
863,000
295,000
Percentage of ChBdren
With the Attribute
0.588%
5.75%
0.115%
38.5%
10.8%
3.70%
1.06
1 Resulting from lead exposure.
Table ES-5. Examples of Environmental-Lead Levels Associated with a 5% Likelihood that
a Child Exposed to Such Levels Would Have a Blood-Lead Concentration at or
Above 10//g/dL.
Model Used to
Predict Blood-
Lead Cone.
{Targeted
Medium)
IEUBK model
(soil)
Rochester
multimedia
model
(dust on floors
or window sills)
Environmental-Lead Levels Considered at
fixed Values in the Model
Floor dust-lead cone. = 1 00 //g/g
Floor dust-lead cone. = 200 //g/g
Floor dust-lead cone. = 500 fjg/g
Dripline soil-lead cone. = 100 //g/g,
Window sill dust-lead loading = 200 //g/ft2
Dripline soil-lead cone. = 1 00 fjg/g,
Window sill dust-lead loading = 500 //g/ft2
Dripline soil-lead cone. = 400 fjg/g,
Window sill dust-lead loading = 200 //g/ft2
Dripline soil-lead cone. = 400 //g/g.
Window sill dust-lead loading = 500 //g/ft2
Dripline soil-lead cone. = 1 00 fjglg,
Floor dust-lead loading = 25 //g/ft2
Dripline soil-lead cone. = 1 00 fjg/g,
Floor dust-lead loading = 1 00 //g/ft2
Dripline soil-lead cone. = 400 //g/g,
Floor dust-lead loading = 25 //g/ft2
Dripline soil-lead cone. = 400 //g/g,
Floor dust-lead loading = 1 00 //g/ft2
Environmental-Lead Levels in the Targeted
Medium Necessary to Control the
Likelihood of a Child with Blood-Lead
Concentration at or above 10 //g/dL to 5%
Soil-lead concentration = 370 //g/g
Soil-lead concentration = 240 //g/g
5% likelihood is unachievable at the given
floor dust-lead concentration
Floor dust-lead loading = 6.7 //g/ft2
Floor dust-lead loading = 2.0 //g/ft2
Floor dust-lead loading = 0.6 //g/ft2
Floor dust-lead loading = 0.2 //g/ft2
Window sill dust-lead loading = 74 //g/ft2
Window sill dust-lead loading = 26 //g/ft2
Window sill dust-lead loading = 1 2 //g/ft2
Window sill dust-lead loading = 4.2 //g/ft2
Note: Dust-lead loadings in this table are assumed to reflect wipe collection methods. Graphical portrayal of
the information presented in this table is found in Section 5.3.
ES-11
-------
ES.4.2 ANALYSIS OF EXAMPLE OPTIONS FOR RISK MANAGEMENT
To illustrate the methods developed in the risk management for evaluating various
options for the §403 standards, the methods were applied to different sets of example standards.
These examples are not meant to encompass all possible options for the §403 standards; the
Agency will consider other sets of candidate standards.
Table ES-6 contains estimated percentages of the housing stock for which the
interventions in Table ES-1 are expected to be triggered according to each of three sets of
example options for dust and soil standards, and paint triggers (the example options are specified
in the footnotes to this table). The number of pre-1980 housing units in which at least one
intervention is triggered ranges from 16.5 million units (22.2%) under example set A (higher
standards) to 46.2 million units (62.1%) under example set F (lower standards). Under example
set C, at least one intervention would be triggered hi approximately 28% of housing built prior to
1980 (20.7 million units) and hi approximately 22% of the entire national housing stock (21.6
million units).
For each set of example options hi Table ES-6, one or both of the dust-lead loading
standards (primarily the window sill standard) were exceeded hi more than half of the units hi
which at least one standard was exceeded. For those homes that did not exceed the example
option for the dust standard, about 1% had soil lead concentrations hi excess of the corresponding
soil standard.
Tables ES-7a and ES-7b provide estimates of the health effect and blood-lead
concentration endpoints for U.S. children aged 1-2 years, as calculated by the IEUBK model and
the empirical model, respectively, under the three sets of example options for dust and soil
standards and paint triggers considered hi Table ES- 6. For each endpoint and each example set,
estimates based on the IEUBK model are less than those based on the empirical model. While
the percentage difference from baseline differs considerably across the three example sets for
all endpoints (especially under the IEUBK model), the endpoints most sensitive to the levels of
the standards and triggers are the percentages of children with blood-lead concentrations at or
above 20 iig/dL or 10 fig/dL, and the percentages of children with IQ score decrements of greater
than 2 or 3.
ES.5 CONCLUSIONS
As estimated in this risk analysis, baseline distribution of childhood blood-lead
concentrations in the U.S. indicates mat elevated blood-lead concentrations remain prevalent.
According to this distribution, approximately 785,000 children aged 1-5 years (3.85%) are
estimated to have blood-lead concentrations at or above 10 ng/dL, the level of concern identified
by CDC. Among 1-2 year olds, the age group on which this risk analysis was focused,
approximately 458,000 children (5.75%) are estimated to have blood-lead concentrations at or
above 10 ug/dL, and 46,800 children (0.59%) are estimated to have blood-lead concentrations at
or above 20 ng/dL. Evidence from previous studies indicate that high percentages of these
children reside in certain environments, such as urban centers, older housing, or within low
ES-12
-------
Table ES-6. Estimated Percentages1 of Occupied Housing Units in the 1997 U.S. Housing
Stock in Which Certain Interventions Are Triggered for Three Example Sets of
Options for Dust and Soil Standards and Paint Triggers.
Intervention^)
Dust cleaning, triggered
by floor dust-lead
loading
Dust cleaning, triggered
by window sill dust-lead
loading
Soil removal
Exterior lead-based paint
abatement
Exterior lead-based paint
maintenance
Interior lead-based paint
abatement
Interior lead-based paint
maintenance
Dust cleaning, triggered
by floor OR window sill
dust-lead loading
Dust cleaning OR soil
removal
Any intervention
Example Set A*
% Units
Triggered
C)
10.3
0.215
3.03
3.84
0.453
2.80
10.3
10.6
17.5
%OfPre-
1980 Units
Triggered
(*)
12.6
0.287
4.05
5.12
0.605
3.73
12.6
12.9
22.2
Example Set C*
% Units
Triggered
4.04
12.5
2.49
5.77
3.49
2.43
2.92
13.9
14.6
21.8
% Of Pre-
1980 Units
Triggered
5.39
15.5
3.32
7.70
4.66
3.25
3.90
17.3
18.3
27.8
Example Set F4
% Units
Triggered
13.8
48.1
11.8
9.26
1.15
5.35
1.08
50.6
51.6
53.7
% of Pre-
1980 Units
Triggered
18.4
54.7
15.8
12.4
1.53
7.15
1.45
58.0
59.3
62.1
1 To assist in interpreting the percentages in this table, the 1997 occupied housing stock is estimated to
contain approximately 99,272,000 housing units, of which approximately 74,379,000 units are built prior to
1980.
2 Example dust and soil standards are 400 //g/ft2 for floors, 800 j/g/ft2 for window sills, and 5000 /yg/g for
soil. Dust standards assume wipe techniques. Paint intervention triggers are 10 ft2 of deteriorated lead-
based paint for paint maintenance, and 100 ft2 of deteriorated lead-based paint for paint abatement.
3 Example dust and soil standards are 100 yug/ft2 for floors, 500 //g/ft2 for window sills, and 2000 //g/g for
soil. Dust standards assume wipe techniques. Paint intervention triggers are 5 ft2 of deteriorated lead-based
paint for paint maintenance, and 20 ft2 of deteriorated lead-based paint for paint abatement.
4 Example dust and soil standards are 25 /yg/ft2 for floors, 25 /yg/ft2 for window sills, and 500 //g/g for soil.
Dust standards assume wipe techniques. Paint intervention triggers are 0 ft2 of deteriorated lead-based paint
for paint maintenance, and 5 ft2 of deteriorated lead-based paint for paint abatement.
(*) indicates that the estimate is essentially zero based on the available data.
ES-13
-------
Table ES-7a. Estimates of Health Effect and Blood-Lead Concentration Endpoints for
Children Aged 1 to 2 Years Under the IEUBK Model and Percent Declines in
These Estimates from Baseline for Three Example Sets of Options for Dust
and Soil Standards and Paint Triggers.
Measure
Percent of children with
blood-lead concentration
greater than or equal to
20//g/dL
Percent of children with
blood-lead concentration
greater than or equal to
10/yg/dL
Percent of children with IQ
score less than 70
resulting from lead exposure
Percent of children with IQ
score decrement of at
least 1 resulting from lead
exposure
Percent of children with IQ
score decrement of at
least 2 resulting from lead
exposure
Percent of children with IQ
score decrement of at
least 3 resulting from lead
exposure
Average IQ score decrement
per child resulting from lead
exposure
Geometric mean blood-lead
concentration (geometric
standard deviation)
Estimate for Children Aged 1-2 Years1
Example
Set A2
0.290%
3.92%
0.107%
34.5%
8.09%
2.37%
0.964
2.95
(2.00)
Example
SetC3
0.0539%
1.66%
0.0984%
28.3%
4.31%
0.858%
0.848
2.74
(1.84)
Example
SetF4
0.00198%
0.250%
0.0909%
15.1%
0.978%
0.0976%
0.666
2.25
(1.70)
Percent Decline in Estimate from
Baseline (Tabte ES-4)
Example
Set A2
50.7%
31.8%
7.0%
10.4%
25.1%
35.9%
9.1%
6.1%
Example
SetC*
90.8%
71.1%
14.4%
26.5%
60.1%
76.8%
20.0%
12.7%
Example
SetF*
99.7%
95.7%
21.0%
60.8%
90.9%
97.4%
37.2%
28.3%
To assist in interpreting the percentages in this table, the number of children aged 1 -2 years in the 1997
national housing stock is estimated as 7,961,000.
Example dust and soil standards are 400 //g/ft2 for floors, 800 //g/ft2 for window sills, and 5000 //g/g for
soil. Dust standards assume wipe techniques. Paint intervention triggers are 10 ft2 of deteriorated lead-
based paint for paint maintenance, and 100 ft2 of deteriorated lead-based paint for paint abatement.
Example dust and soil standards are 100 //g/ft2 for floors, 500 //g/ft2 for window sills, and 2000 //g/g for
soil. Dust standards assume wipe techniques. Paint intervention triggers are 5 ft2 of deteriorated lead-based
paint for paint maintenance, and 20 ft2 of deteriorated lead-based paint for paint abatement.
Example dust and soil standards are 25 //g/ft2 for floors, 25 //g/ft2 for window sills, and 500 //g/g for soil.
Dust standards assume wipe techniques. Paint intervention triggers are 0 ft2 of deteriorated lead-based
paint for paint maintenance, and 5 ft2 of deteriorated lead-based paint for paint abatement.
ES-14
-------
Table ES-7b. Estimates of Health Effect and Blood-Lead Concentration Endpoints for
Children Aged 1 to 2 Years Under the Empirical Model and Percent Decline
in These Estimates from Baseline for Three Example Sets of Options for Dust
and Soil Standards and Paint Triggers.
Measure
Percent of children with blood-
lead concentration greater than
or equal to 20 /yg/dL
Percent of children with blood-
lead concentration greater than
or equal to 10/yg/dL
Percent of children with IQ
score less than 70
resulting from lead exposure
Percent of children with IQ
score decrement of at least 1
resulting from lead exposure
Percent of children with IQ
score decrement of at least 2
resulting from lead exposure
Percent of children with IQ
score decrement of at least 3
resulting from lead exposure
Average IQ score decrement
per child resulting from lead
exposure
Geometric mean blood-lead
concentration (geometric
standard deviation)
Estimate for Children Aged 1-2 Years1
Example
Set A*
0.458%
5.03%
0.112%
37.1%
9.79%
3.16%
1.02
3.07
(2.05)
Example
SetC3
0.406%
4.70%
0.110%
36.3%
9.30%
2.93%
1.00
3.03
(2.04)
Example
SetF*
0.317%
4.09%
0.108%
34.7%
8.34%
2.49%
0.971
2.95
(2.01)
Percent Decline in Estimate from
Baseline (Table ES-4)
Example
SetA2
22.1%
12.5%
2.7%
3.6%
9.4%
14.6%
3.4%
2.3%
Example
SetC*
31.0%
18.3%
4.3%
5.7%
13.9%
20.8%
5.7%
3.5%
Example
SetF*
46.2%
28.9%
6.0%
9.8%
23.0%
32.6%
8.2%
5.9%
1 To assist in interpreting the percentages in this table, the number of children aged 1-2 years in the 1997
national housing stock is estimated as 7,961,000.
2 Example dust and soil standards are 400 /yg/ft2 for floors, 800 /yg/ft2 for window sills, and 5000 /yg/g for
soil. Dust standards assume wipe techniques. Paint intervention triggers are 10 ft2 of deteriorated lead-
based paint for paint maintenance, and 100 ft2 of deteriorated lead-based paint for paint abatement.
3 Example dust and soil standards are 100 //g/ft2 for floors, 500 /yg/ft2 for window sills, and 2000 /yg/g for
soil. Dust standards assume wipe techniques. Paint intervention triggers are 5 ft2 of deteriorated lead-based
paint for paint maintenance, and 20 ft2 of deteriorated lead-based paint for paint abatement.
4 Example dust and soil standards are 25 /yg/ft2 for floors, 25 /yg/ft2 for window sills, and 500 /yg/g for soil.
Dust standards assume wipe techniques. Paint intervention triggers are 0 ft2 of deteriorated lead-based
paint for paint maintenance, and 5 ft2 of deteriorated lead-based paint for paint abatement.
ES-15
-------
income households, where there is typically a higher likelihood of encountering lead-based paint
hazards.
Methodology was developed and applied for characterizing reductions in risk expected to
result after interventions are conducted in response to the proposed rule. While examples of the
risk reductions have been presented for selected sets of example options for §403 standards, they
have been included primarily to illustrate application of the risk management methodology and
are not necessarily meant to convey definitive patterns hi risk measures across different sets of
candidate standards. The Agency will apply the risk management methodology to evaluate
specific options for environmental-lead standards.
A major limitation hi the risk management methodology was the lack of nationally-
representative dust-lead loading data where samples were collected by wipe techniques. This
data gap existed for both baseline (pre-§403) and post-§403 conditions. While approaches were
used to help alleviate this data gap, such as conversions of dust-lead loading data from one
sample collection method to another and assumptions on post-intervention dust-lead loadings,
sensitivity analyses suggest that these approaches yielded estimates on numbers of affected
housing units and on risk characterization having considerable uncertainty.
Other aspects of the data modeling and analysis approaches, as well as assumptions made
in this process contribute to overall uncertainty in the results. Among these are the need to adjust
dust-lead concentrations hi the HUD National Survey to reflect underestimated sample weights,
high levels of variability in the HUD National Survey data, the age of the studies and surveys
providing data to this risk analysis, procedures to update the data to reflect 1997 conditions,
modeling the relationship between blood-lead concentration and health effect measures, the need
for data conversions, assumptions on lognormaliry hi the data distributions, and how statistical
models are applied in this setting. Despite the levels of uncertainty that they may generate, the
approaches and assumptions taken hi this analysis have a sound scientific basis.
This risk analysis indicates that while the presence of elevated blood-lead concentration
and health effects associated with lead exposure have declined over time, many U.S. children
continue to experience these health effects as a result of lead exposure. However, this risk
analysis has shown that it is possible to achieve substantial and important reductions in the
incidence of adverse health effects hi U.S. children when implementing standards to reduce lead
exposures under §403 rules.
ES-16
-------
1.0 INTRODUCTION
CHAPTER 1 SUMMARY
This introductory chapter provides the background, objectives, and
overview, which includes a description of the organization and general approach of
this report. §403 of the Toxic Substances Control Act, as created by Title X,
requires EPA to define standards for lead-based paint hazards, lead-contaminated
dust, and lead-contaminated soil. This report
documents the scientific basis for regulations called for
under §403. Hereinafter, regulations under $403 are
referred to by the term §403.
characterizes the health risks to young children from
exposures to lead
presents the methodology developed by EPA to estimate the
reduction in risk expected to result from promulgation of the
§403 standards
* applies the methodology to estimate the reductions in
childhood health risks and blood-lead concentrations
expected to result from example options for the §403
standards
estimates the percentages and numbers of children and
housing units affected by example options for the §403
standards.
This information is provided to help the risk managers evaluate and compare
various regulatory options for §403.
Title X of the Housing and Community Development Act, known as the Residential
Lead-Based Paint Hazard Reduction Act of 1992, contains legislation designed to evaluate and
reduce exposures to lead in paint, dust, and soil in the nation's housing. This act provides the
framework for developing a national strategy for reducing and preventing lead exposures to
children. Consistent implementation of this strategy by federal, state, local and private agencies
requires a uniform definition of lead hazards. Title X includes a provision that requires the U.S.
Environmental Protection Agency (EPA) to define standards for lead in paint, dust, and soil.
More specifically, §403 of the Toxic Substances Control Act (TSCA) is a part of Title IV, "Lead
Exposure Reduction," and was added to TSCA by Title X. Section 403 requires EPA to
''promulgate regulations which shall identify, for purposes of this title and the Residential Lead-
Based Paint Hazard Reduction Act of 1992, lead-based paint hazards, lead-contaminated dust,
and lead-contaminated soil."
The §403 regulations will set standards (condition and location of lead-based paint, levels
of lead in dust and soil) against which to compare a residential environment when evaluating the
1-1
-------
presence and magnitude of lead-based paint hazards. Federal, state, and local public health
agencies, as well as private property owners and other private sector interests, will use these
standards to determine in which homes actions should be taken to reduce or prevent the threat of
childhood lead poisoning. Blood-lead concentration is a commonly used indicator of exposure to
lead and of childhood lead poisoning. Following the conduct of actions taken in response to the
§403 standards, average blood-lead concentrations and collective health risks associated with
childhood lead poisoning will be reduced for children currently residing in the residence as well
as for those that may later live in the residence. Proper selection of the standards requires both an
understanding of the health risks associated with residential exposures to lead, the amount by
which these risks can be reduced through intervention strategies, and the numbers of homes and
children affected by the standards.
The purpose of this report is to document the scientific basis for the proposed §403
standards. First, the report summarizes EPA's assessment of the health risks to young children
from exposures to lead-based paint hazards, lead-contaminated dust, and lead-contaminated soil
in the nation's housing. Seven health effect and blood-lead concentration endpoints associated
with lead exposures are characterized for children aged 12 to 35 months (cited as aged 1-2 years
in this report). While health risks associated with lead exposures are significant for all young
children, health risks to children aged 1-2 years are utilized in this risk analysis because it was
found (Section 2.4) that the neurotoxicological effects of lead exposure may be best measured for
this age group and that the neurotoxicological effects of lead exposure at this age may be
irreversible. Second, the report documents the approach developed by EPA to estimate the
reductions in these risks following promulgation of the §403 standards, and applies this
methodology to evaluate example options for the §403 standards. The benefits of each example
option for §403 are expressed in terms of the reduction in health risks attained from actions taken
in response to promulgation of the §403 standards. Finally, the report provides estimates of the
numbers of homes and children that will be affected by various example standards.
Information presented in this risk analysis will ultimately be used to consider various
standards for rulemaking and as input to the Regulatory Impacts Analysis (RIA) for the proposed
rule, as well as any interim economic cost-benefit analyses. While the risk analysis provides
quantitative estimates of the impact of §403 hi terms of health and blood-lead concentration
endpoints and documents the scientific basis for these estimates, the RIA and other economic
analyses express the impact of the §403 standards in terms of costs: monetary costs of
implementing the regulation, monetary benefits associated with reductions hi health risks and
blood-lead concentrations for various options for the regulation, and the estimated economic
impacts of the regulations. The RIA also examines the likelihood of interventions actually taking
place. Finally, the RIA summarizes other regulatory actions designed to reduce risks from lead,
and presents environmental equity analyses for adults and children.
This report documents the critical decisions on risk-assessment-related tools and data that
are being relied upon in the RIA analyses and the actual rulemaking, which relate environmental
levels of lead to children's health effect and blood-lead concentration endpoints. This report
includes a description of the data used, an assessment of the strengths and weaknesses of that
data, and discussions of any additional uncertainties which result from using these particular data
1-2
-------
sets and tools to create estimates of risk reduction on a national basis. Section 1.1 provides
background on the §403 regulations. Statutory/policy constraints of the proposed rule are
discussed in Section 1.2. Objectives are presented in Section 1.3 and an overview of the report is
given in Section 1.4.
1.1 BACKGROUND
On October 29,1992, the Residential Lead-Based Paint Hazard Reduction Act of 1992
(42 U.S.C. 4851) was signed into law. Subtitle B of Title X amends TSCA, by adding Title IV,
"Lead Exposure Reduction." Title IV requires EPA to take certain actions to address lead-based
paint concerns, including establishing requirements for training and accreditation of contractors
conducting lead paint-related work. Section 403 of TSCA (15 U.S.C. 2683) states:
"... the Administrator shall promulgate regulations which shall identify, for
purposes of this title and the Residential Lead-Based Paint Hazard Reduction Act
of 1992, lead-based paint hazards, lead-contaminated dust, and lead-contaminated
soil."
This statute requires EPA to establish criteria for identifying lead-based paint hazards,
including lead-contaminated household dust and lead-contaminated residential soil. The statute
defines lead-based paint to be dried paint film with a lead content exceeding 1.0 mg of lead per
square cm of surface area (mg/cm2) or 0.5 percent (5,000 parts per million (ppm)) by weight1.
The §403 statute requires EPA to identify the condition and location of lead-based paint that
causes exposures to lead in paint, lead-contaminated dust and lead-contaminated soil that would
result in unacceptable health risks. The definitions of lead-based paint hazard, lead-contaminated
dust and lead-contaminated soil provided in Title X refer to human health effects' and human
health hazards. A glossary of terms defined in the statute and used in this risk analysis is
provided in Appendix A.
Congress concluded in Title X that exposure to deteriorated lead-based paint, lead in dust
and lead in soil comprises a serious health problem for American children. Congress further
stated that lead exposures to children, even at low levels, may result in intelligence quotient (IQ)
deficiencies, reading and learning disabilities, unpaired hearing, reduced attention span,
hyperactivity, and behavior problems. Actions taken to reduce childhood exposures that are
conducted in response to the proposed §403 standards are expected to reduce the incidence of
these adverse health effects hi young children. Title X states that the adverse health events
associated with childhood exposure to lead-based paint hazards can be reduced by abating
lead-based paint or by taking interim measures to prevent paint deterioration and limit children's
exposure to lead dust and paint chips.
The statute defines lead-based paint to be dried paint film with a lead content exceeding 1.0 mg/cm2.
However, other EPA and Federal programs (§1018, HUD Guidelines) have defined lead-based paint to be paint
with a lead content greater than or equal to 1.0 mg/cm3. To be consistent with other programs, lead-based paint is
defined as paint with lead content greater than or equal to 1.0 mg/cm2 in this risk analysis.
1-3
-------
1.2 STATUTORY/POLICY CONSTRAINTS
During the development of this risk analysis methodology, there were a number of
constraints that were imposed for statutory and programmatic reasons. Six of these were
especially significant.
First the §403 statute defines lead-based paint to be any painted surface that contains
more than 1.0 mg of lead per cm2 of surface area or 0.5% bv weight. Consequently, there are
likely to be residences with deteriorated paint that contributes to lead exposure but, according to
the above definition, do not contain "lead-based paint." Unless these residences contain dust or
soil lead at levels defined as hazardous under the §403 regulations, they would be outside the
scope of the Title X program.
Second, by statutory definition, intact lead-based paint is not considered a hazard unless it
is present on accessible, friction, or impact surfaces. Intact lead-based paint is lead-based paint
that is not deteriorating, chipping, or peeling. The nibbing and scraping of lead-based paint on
friction and impact surfaces may cause fine particles of lead to contaminate residential dust.
Accessible surfaces are those that are accessible for chewing or mouthing by young children.
Under §403 of TSCA, the Agency is required to identify any condition and location of lead-based
paint that would result in adverse human health effects. Condition refers to the extent to which
the painted surface is deteriorated. For deteriorated lead-based paint, there is an increased
potential for both the direct ingestion of paint chips containing lead and for the contamination of
the residential dust. This risk analysis uses the amount of deteriorated lead-based paint (in square
feet) to estimate the effects of lead-based paint on human health. Data linking lead-based paint
on friction or impact surfaces to lead in dust were limited. In addition, the effects of lead on
friction, impact, and accessible surfaces on childhood blood-lead concentrations are not well
quantified. Therefore, for the purpose of this report only, despite its eligibility, intact lead-based
paint is not evaluated as a potential hazard, even on friction, accessible or impact surfaces. The
agency has not excluded intact lead-based paint on these surfaces from the options for the
proposed §403 rule.
Third, there are two methods for measuring the amount of lead in household dust:
loading and concentration. Dust-lead loading measures the mass of lead collected per surface
area sampled and is usually expressed in terms of micrograms of lead collected per square foot
sampled (\ig Pb/ft2). Dust-lead concentration measures the mass of lead collected per mass of
dust collected and is usually stated hi terms of micrograms of lead collected per gram of dust
collected (ug Pb/g dust). Both are commonly used for evaluating exposures to lead hi dust.
Dust-lead loading measures the amount of lead present on the sampled surface, while dust-lead
concentration measures the amount of lead hi a given amount of dust. A high dust-lead loading
might represent a surface containing a large amount of dust at a low lead concentration or a
surface containing a small amount of dust at a high lead concentration. Both measures have been
used to predict blood-lead concentrations and there is currently no consensus on which measure
may be the better predictor. Ideally, EPA would use both loading and concentration data to
characterize hazards and to identify appropriate response actions. The Agency recognizes that
setting standards based on both measures might impede implementation of hazard evaluation on
1-4
-------
a large scale (i.e., in the nation's housing). For policy reasons, it was determined that dust-lead
levels would be characterized by a dust-lead loading2.
There are two limitations to this decision. First, residences with high dust-lead
concentrations and low dust-lead loadings will not be identified as exceeding the §403 dust-lead
loading standard. The second limitation involves the temporal nature of residential dust.
Samples for analysis of both dust-lead loading and dust-lead concentration are routinely collected
as grab samples during a scheduled visit to the residence. This means there is no control and no
uniformity over the length of time the surface has accumulated dust since the previous thorough
cleaning. Dust-lead concentration is assumed to be independent of this period of accumulation,
but dust-lead loading is highly dependent, as a long accumulation of dust with a low lead
concentration might give the same result as a short accumulation of dust with a high lead
concentration. (Although the §403 rulemaking will express the dust standard in terms of a dust-
lead loading, dust-lead concentration is employed as an intermediate step in the analysis when
using EPA's Integrated Exposure Uptake Biokinetic (EEUBK) model, as this model requires dust-
lead concentration.)
i
Fourth, there are two approaches for collecting samples of dust from a surface: wipe and
vacuum sampling. Although dust-lead loading can be measured using both wipe or vacuum
sampling, dust-lead concentration can be measured only via vacuum sampling. The interim
standards for dust lead in the Interim 403 Guidance Document (USEPA, 1995f) were defined in
terms of lead loading. That decision was based, in part, on the wider availability and familiarity
of wipe sampling compared to vacuum sampling. Currently, wipe sampling is the method most
risk assessors use and few are skilled in vacuum sampling. Therefore, the Agency made a policy
decision to define the §403 dust standards in terms of a dust sample collected via wipe sampling.
Fifth, the Agency had to determine for which surfaces it would propose standards for lead
in household dust. Dust can accumulate on multiple surfaces in the home: floors, furniture,
window sills, and window troughs. Elevated levels of lead in dust accumulated on the window
sill may not represent the same degree of health concern as do elevated levels of lead in floor
dust. To date, federal, state, and local public health agencies have primarily tested for the
presence of lead in dust on three horizontal surfaces: uncarpeted floors, ulterior window sills, and
window troughs. The Interim 403 Guidance Document (USEPA, 1995f) provided standards for
lead in dust on floors, window sills, and window troughs.
Technical analysis conducted to support the risk analyses indicated that dust on floors,
sills, and troughs are highly correlated. Several studies have observed that the correlation
between childhood blood-lead concentration and floor dust-lead loading is stronger than the
correlations with either dust-lead loading on sills or on troughs. Such a difference is not always
statistically significant nor is it always observed. Three recent studies (Lanphear, 1995; USEPA
1996a; USEPA, 1996b) report comparable correlations between blood-lead concentration and
This applies to interior dust exclusively. The soil standard will be expressed solely as a concentration
standard, i.e., ug Pb/g soil.
1-5
-------
dust-lead levels (loadings and concentrations) on floors, sills, and troughs. The conventional
wisdom, as reflected in the scientific literature, suggests that floor dust-lead levels are the dust
lead measure most relevant to childhood lead exposure due to the larger amount of tune children
spend in contact with floors compared to window sills and troughs. While the same dust can
settle on window sills and accumulate in window troughs, the exposure pathway (hand-to-mouth
behavior) is thought to primarily occur on the residential floor.
Collecting dust samples from both sills and troughs does not improve a risk assessor's
ability to characterize risk sufficiently to justify the sampling and analysis of both surfaces.
Because sills are easier to sample than troughs, the Agency has decided to estimate the risk and
risk reductions expected to result for various example options for the §403 dust standards for
floors and window sills but not for window troughs. However, the Agency recommends that
once dust-lead levels are identified to represent a hazard in a home, that appropriate action be
taken to reduce dust-lead levels on floors, window sills, window troughs and other surfaces in the
home.
Sixth, the regulations on lead levels in soil, as promulgated under §403. apply only to
"bare soil." At the time of the development of the risk analysis methodology, "bare soil" had not
yet been precisely defined. Currently, the application of the risk analysis methodology presented
in this document presumes that the standards for lead in soil apply to all residences that exceed
the standard (a soil-lead concentration), without consideration as to whether the soil is "bare."
Consequently, the numbers of homes identified in this report as exceeding example options for
the §403 soil standard will tend to be overestimated if the standard is defined in terms of bare soil
since some of the homes in the analyzed datasets may not contain bare soil. It should also be
noted that little or no data are available on the national prevalence of "bare soil," regardless of
how that term is ultimately defined in the §403 rulemaking.
1.3 OBJECTIVES
Although the standards defined by this rule will not require the conduct of any lead
exposure reduction activities, EPA recognizes that they will be used by federal, state, local, and
private entities hi their efforts to manage the hazards of lead in paint, dust, and soil. Therefore,
the objectives of this risk analysis are described below.
Risk Assessment Objectives
1. Document the scientific basis for the proposed §403 standards.
This report assesses the risks of childhood exposure to lead hi paint, dust and soil.
Each component of the risk assessment is documented in this report: hazard
identification, exposure assessment, and dose-response assessment. These individual
characterizations are integrated to assess the risks of lead exposures to children aged
1 -2 years and the reductions in these risks expected to take place as a result of the
§403 standards.
1-6
-------
2. Characterize the health risks to young children from specific residential exposures to
lead.
This document estimates risks to young children from specific residential sources of
lead. These sources are: (1) ulterior and exterior lead-based paint; (2) lead-
contaminated dust, which may contain lead derived from deteriorated interior paint,
tracked- or blown-in exterior soil, and other sources and (3) lead-contaminated soil,
which may contain lead from deteriorated paint, from past leaded-gasoline vehicle
emissions, or from other sources. >
This risk assessment focuses on risks to children aged 1-2 years. Other
populations also certainly face risks from lead'exposure, including children of
other ages, pregnant women, and the general adult population. Characterization
of risks and risk reduction for 1-2 year old children was chosen as being
representative of total risk and risk reduction. Also, the discussion in Section 2.4
provides evidence that this subgroup of children is among those most appropriate
for estimation of the endpoints considered in this risk assessment.
Risk Management Objectives
1. Develop methodology to estimate the reduction in risk expected to result from
promulgation of the §403 standards.
This report presents the approach developed by the Agency to characterize the
incremental risk reduction expected to result after interventions (actions taken to
reduce residential lead exposures) are conducted in response to the §403 standards.
Because the §403 rule does not mandate action be taken at any lead levels measured
in a residential environment, it was not possible to analyze the risk reductions
associated with specific interventions required by the regulation. Instead, the
Agency's approach is to characterize the risk reduction consequences that might occur
if broadly defined interventions are undertaken to reduce exposures to lead in dust,
soil, and paint. Intervention activities considered in this report are: cleaning of house
dust, maintenance of interior or exterior paint, encapsulation/abatement of ulterior or
exterior paint, and soil removal.
2. Apply the methodology to characterize the reduction in risk expected to result from
implementation of the §403 rule for a broad range of example standards.
This report implements the risk management methodology to explore the implications
of various example options for the §403 standards. The example standards examined
in this report are not meant to encompass all possible options for the §403 standards.
3. Estimate the numbers of children and housing units directly impacted by example
options for the §403 standards.
1-7
-------
This report estimates numbers of children and numbers of homes in the nation's
housing stock that would be affected by the rulemaking for a broad range of example
standards. The time frame of the risk analysis is 1997, with the assumption that all
actions resulting from the §403 rule occur within that time frame.
Note that the objectives of this report do not include the selection of the §403 standards.
Standard selection is a policy decision to be made by the Agency. The purpose of this report is to
provide relevant information on the scientific basis for setting the standards and the comparative
risk reductions that are expected to result for various example options for the §403 standards.
1.4 OVERVIEW OF REPORT
1.4.1 Organization of Report
This report serves to answer two questions:
What are the health risks to young children from exposures to lead in paint, dust, and
soil?
What are the expected reductions in the health risks as a result of actions conducted in
response to the proposed §403 standards?
To answer these two questions, the report is divided into two parts: Risk Assessment and
Risk Management. The first part, Risk Assessment, contains four chapters, and the second part,
Risk Management, includes one chapter. This section provides an overview of the organization
of the report.
Risk Assessment
Results of the hazard identification are provided in Chapter 2. The information provided
in Chapter 2 summarizes the existing knowledge, as documented in the literature, on the health
effects of lead exposures. Additional research on the toxicity and hazards of lead was not
conducted. For a more comprehensive assessment of lead toxicity, the reader is referred to the
evaluations conducted by EPA (USEPA, 1986) and the Agency for Toxic Substances and Disease
Registry (ATSDR) (ATSDR, 1993) and to other literature. Children aged 12 to 35 months (cited
as aged 1-2 years in this report) are selected as the reference group for assessing risks in Section
2.4. Health effects associated with deficits in IQ scores due to lead exposures and incidence of
blood-lead concentration exceeding two thresholds used by the Centers for Disease Control and
Prevention (CDC) are selected in Section 2.5 as the endpoints for characterizing the risks
associated with lead exposures hi this report.
Both environmental levels of lead and childhood blood-lead concentration are used hi
Chapter3 to assess lead exposures. Pathways and sources of environmental lead exposure are
summarized in Section 3.1. The rather extensive evidence on the relationship between childhood
blood-lead concentration and environmental-lead levels is summarized in Section 3.2. The
1-8
-------
distribution of lead in residential dust, soil, and paint in the nation's housing is estimated in
Section 3.3 based on data collected in the Department of Housing and Urban Development's
National Survey of Lead-Based Paint in Housing (HUD National Survey) (USEPA, 1995a;
USEPA, 1995g; and USEPA, 1995h). The distribution of blood-lead concentration for children
aged 1-2 years based on the data reported in Phase 2 of the third National Health and Nutrition
Examination Survey (NHANES ffl) (CDC, 1997) is presented in Section 3.4.
The methods used to characterize the dose-response relationships between environmental
lead exposures and the selected health effect and blood-lead concentration endpoints are
presented in Chapter 4. Two different types of models are presented in Chapter 4 for predicting
blood-lead concentrations from measures of environmental lead: EPA's IEUBK model and an
empirical model developed for this study. Because EPA's approach is to define the dust-lead
standard in terms of a wipe dust-lead loading and because dust samples in the HUD National
Survey were collected via vacuum sampling, Section 4.3 presents equations for converting a
dust-lead loading collected via a vacuum sampler to a wipe equivalent dust-lead loading for
purposes of this risk analysis. Finally, methods for computing the health effect and blood-lead
concentration endpoints evaluated in this report are detailed in Section 4.4.
Chapters integrates the characterizations from the hazard identification, exposure
assessment, and the dose-response assessment to characterize health risks to children aged 1-2
years from exposures to lead in paint, dust, and soil. Health risks are predicted for the nation's
children aged 1-2 years in 1997 before any actions (pre-§403) are taken hi response to the
proposed §403 standards (Section 5.1), for children exposed to background levels of lead
(Section 5.2), and for individual children exposed to specific levels of environmental lead
(Section 5.3). The risk characterization is provided hi Section 5.5.
Risk Management
Implications of example options for the §403 standards are explored in the second part of
the report: Risk Management. Chapter 6. Analysis of Example Options for the §403 Standards,
presents the methodology developed by the Agency to characterize the incremental risk
reductions expected to result from interventions conducted hi response to various §403 example
standards. The primary purpose of the risk management analyses is to document and illustrate
implementation of the approach developed for making relative comparisons among example
options for the §403 standards. To predict the distribution of environmental-lead levels that
would result from actions conducted in response to the §403 standards, the Agency identified a
limited set of intervention activities that are assumed to occur at housing units identified as
exceeding one or more of the standards (Section 6.1). The methodology utilized to predict health
effect and blood-lead concentration endpoints for children aged 1-2 years associated with
distributions of environmental-lead levels expected to take place following implementation of
the §403 standards is described in Section 6.2. Results of implementing these methods for
example options for the §403 standards are found in Section 6.3.
1-9
-------
1.4.2 General Approach
Figure 1-1 provides an overview of the risk analysis conducted for the §403 regulation.
Hazard identification, exposure assessment, and dose-response assessment provide necessary
information to characterize the risks associated with childhood lead exposures. The general
approach for these sections are summarized in Section 1.4.2.1. The general approach employed
in the Risk Management portion of the report is presented hi Section 1.4.2.2.
A scientific assessment of lead exposures on childhood health effects and blood-lead
concentrations is a very complex problem. The relationship between lead exposures and health
effects possesses both temporal and spatial sources of variation. The Agency is not aware of any
modeling tools or data sets that contain all of the information or variables required to estimate the
risk associated with lead exposures and the risk reductions expected to result from promulgation
of the rule. Therefore, it was necessary to link together multiple models and combine
information from multiple data sources in the risk analysis. The Agency acknowledges that there
is substantial uncertainty in the estimated risks associated with exposure to lead and in the
estimated risk reductions due the uncertainty hi the modeling tools, the assumptions underlying
Unking together various modeling tools, the variability in the measured data, and use of multiple
sources of data collected under different conditions using different techniques and procedures.
1.4.2.1 Approach for Risk Assessment
Hazard Identification
Figure 2-1 hi Chapter 2 presents the approach used for hazard identification. As with
most assessments of adverse health effects attributable to environmental exposures to lead, the
level of blood-lead concentration is used as the index of exposure. While the health risk
associated with lead exposure is significant for all young children, the population of interest
selected for this risk analysis was U.S. children aged 1-2 years because, as found in Section 2.4,
this subgroup of children is among those most appropriate for estimation of the endpoints
considered hi this risk assessment. Two types of endpoints are utilized to characterize the
adverse health effects hi this risk analysis: elevated blood-lead concentration and IQ point deficit.
The following seven health effect and blood-lead concentration endpoints are used to characterize
the risks associated with lead exposures in children aged 1-2 years:
The incidence of blood-lead concentration greater than or equal to 10 ug/dL
The incidence of blood-lead concentration greater than or equal to 20 ug/dL
The incidence of IQ score less than 70 due to childhood lead exposure
The likelihood of an IQ score decline greater than 1 point due to childhood lead
exposure
The likelihood of an IQ score decline greater than 2 points due to childhood lead
exposure
1-10
-------
Background
and
Objectives
(Chapter 1)
Hazard
Identification
(Chapter 2)
Exposure
Assessment
(Chapter 3)
Dose-Response
Assessment
(Chapter 4)
Risk
Characterization
(Chapter 5)
Risk Management
(Chapter 6)
Conclusions
on Risk
Characterization
Analysis of
Example Options for
§403 Standards
Figure 1-1. Overview of the Risk Assessment and Risk Management Approach.
1-11
-------
The likelihood of an IQ score decline greater than 3 points due to childhood lead
exposure
Average IQ score decline in a child as a result of childhood lead exposure.
Exposure Assessment
Figure 3-1 in Chapter 3 displays the approach utilized for exposure assessment. Both
environmental levels of lead and childhood blood-lead concentrations are employed to
characterize exposures to lead. Recognized as the leading source of data on environmental-lead
levels in residential environments, the National Survey of Lead-Based Paint in Housing,
conducted from 1989-1990 by the U.S. Department of Housing and Urban Development, was the
primary source of data on baseline environmental-lead levels in dust and soil in the nation's
housing stock. The design and findings of the HUD National Survey have been peer reviewed
and published in several government reports. However, there are limitations associated with
using the HUD National Survey data in this risk analysis, including limited numbers of
environmental samples taken at each housing unit, the sampling of only 284 houses (which were
all built prior to 1980), the age of the study, and use of a dust collection device other than the
wipe collection method being adopted by the §403 rules. These limitations contribute to overall
uncertainty in the analysis results.
The national distribution of baseline blood-lead concentrations in children aged 1-2 years
was determined from data collected hi phase 2 of the third National Health and Nutrition
Examination Survey (NHANES m), conducted from 1991-1994. While the national
representation of NHANES m results is widely accepted, some possible limitations in using
these data include ignoring any seasonality effects on blood-lead concentrations and any further
decline in concentrations that may have occurred since 1994.
Data from the Baltimore Repair and Maintenance Study and the Rochester Lead-in-Dust
Study are employed to supplement the national data on environmental-lead levels and childhood
blood-lead concentrations with data for inner-city homes and for older homes in an urban setting,
respectively.
Dose-Response Assessment
Figure 4-1 in Chapter 4 describes the approach taken for dose-response assessment. The
relationship between environmental-lead levels and the health effect and blood-lead
concentration endpoints are estimated in two steps because there is little scientific data for
estimating the relationship directly. First, blood-lead concentrations are estimated based on
environmental-lead levels, and then health effect and blood-lead concentration endpoints are
computed from the predicted blood-lead concentrations. The two-step dose-response relationship
was necessary because the majority of the scientific evidence on the relationship between lead
exposure and IQ point deficits are stated in terms of blood-lead concentrations.
1-12
-------
The Agency used two different tools for linking environmental- and blood-lead levels,
because no single tool is considered optimal for the risk analysis. The first tool is the Agency's
Integrated Exposure Uptake Biokinetic (IEUBK) model, a "biokinetic"-type model of lead
exposure. The EEUBK model employs exposure, uptake, and biokinetic information to predict a
distribution of blood-lead concentrations for children corresponding to a specific combination of
environmental-lead levels. Actually, the model predicts the center of this distribution, the
geometric mean. Because blood-lead concentrations tend to have a skewed distribution, the
geometric mean, rather than the arithmetic mean, is used to represent the center of the
distribution. A measure of variability and the predicted geometric mean are then used to estimate
the distribution of blood-lead concentrations associated with a specific combination of
environmental-lead levels. The variability estimate, referred to as the geometric standard
deviation (GSD), represents the variability in blood lead concentrations for a given set of
environmental exposures due to biological and behavioral variability in the exposed children.
The IEUBK model was initially developed in 1985 by EPA's Office of Air Quality
Planning and Standards (OAQPS) as a tool for setting air lead standards. The version used by the
Air program was peer reviewed and found acceptable by EPA's Clean Air Science Advisory
Committee of the Science Advisory Board (USEPA, 1990b). The IEUBK model, unfortunately,
has several limitations for use in the Agency's risk analysis. Specifically, the IEUBK model
1. does not incorporate dust-lead loading on residential surfaces;
2. does not incorporate dust lead exposure from residential window sills;
3. does not directly incorporate a child's tendency to pica behavior (an add-on
adjustment for pica, separate from the model's prediction of blood-lead concentration,
was necessary).
In addition, the IEUBK model's default parameters are meant to be adjusted for site-specific
applications. There are no parameters values available for use in a nationally representative
analysis such as this. A second tool, therefore, was developed that addressed each of these
limitations.
The second tool is an "empirical" model developed specifically for use hi this risk
analysis based on data collected in the Rochester Lead-in-Dust Study. The empirical model
relates environmental-lead levels observed at a residence to the blood-lead concentration
measured for a child living at the residence. For a given set of environmental-lead levels, the
model can be used to predict a geometric mean blood-lead concentration for children exposed to
the given lead levels. This result, along with a specified value of the GSD, is then applied to
estimate the distribution of predicted blood-lead concentrations.
The empirical regression model was developed using data from the Rochester Lead-in-
Dust Study, collected in the summer of 1993 to estimate the relationship between blood-lead
concentrations in young children and observed levels of lead hi environmental media (paint, dust
and soil) from their primary residences. The empirical model estimates the average log-
1-13
-------
transformed childhood blood-lead concentration associated with each studied home. The
variables used for prediction are soil-lead concentration, Blue Nozzle vacuum dust-lead loading
on floors (carpeted and uncarpeted), Blue Nozzle vacuum dust-lead loading on window sills, and
an indicator of paint/pica hazard. (The Blue Nozzle vacuum is the vacuum method employed in
the HUD National Survey). It is not intended as a general dose-response model, but rather as a
predictive model developed specifically for use in the risk analysis and specifically to predict
blood-lead concentrations from estimates of environmental lead as measured in the HUD
National Survey.
The choice of the Rochester Lead-in-Dust Study as the data upon which to develop the
empirical model was based on three factors:
1. all media, locations, and surfaces that are being considered for the §403 standards
were measured for lead in the Rochester study;
2. the Rochester study includes dust-lead loadings from wipe sampling and the §403
dust standard is expected to be based on dust-lead loading from wipe sampling; and,
3. the selection of homes and children in the Rochester study, although targeted, was
more random and more representative of a general population than is the case with
most recent lead exposure studies in non-smelter communities.
There are also limitations in the use of the empirical model in the Agency's risk analysis.
The empirical model is based on only one data set collected at a single city in the northeast.
More importantly, perhaps, it has not yet undergone formal peer review or model evaluation.
These limitations and those associated with the EEUBK model are why two tools were utilized for
linking environmental- and blood-lead levels.
All of the health and blood-lead concentration endpoints are computed using the
geometric mean and geometric standard deviation of predicted distributions of children's blood-
lead concentrations assuming that the distributions are approximately lognormal (a special
statistical distribution). The lognormal assumption is discussed in Sections 5.1.1 and 5.4.4. In
addition, computation of the IQ score declines greater than 1,2, or 3 points and average IQ score
decline are based on an average decrease of 0.257 IQ points per increase of one ug/dL in blood-
lead concentration (Schwartz, 1994). Estimating the incidence of IQ score less than 70 is based
on results in a paper by Wallsten and Whitfield (1986) on the relationship between reduced IQ
scores and blood-lead concentration.
Risk Characterization
Figure 5-1 in Chapter 5 shows the approach taken for risk characterization. Each
component of the risk assessment is integrated to characterize the risks to the nation's children
aged 1-2 years from exposures to lead in paint, dust, and soil. The hazard identification is used to
select the indicators of risk, the health effect and blood-lead concentration endpoints. The
exposure assessment is used to characterize the exposure of children, and the dose-response
1-14
-------
assessment is used to translate exposure into health effect and blood-lead concentration
endpoints.
The risk assessment characterizes risks associated with childhood lead exposure by
predicting incidence of the selected health effect and elevated blood-lead concentration endpoints
among 1-2 year old children for the year 1997. The 1997 baseline distribution of children's
blood-lead concentrations was calculated using data from Phase 2 of NHANES HI. Risks were
projected to 1997 because the rule is expected to be proposed in that year. Although Phase 2 of
NHANES ffl was conducted over the years 1991-1994, the Agency utilized the results of Phase 2
of NHANES ffl to represent the current blood-lead distribution. The Agency recognizes that
children's blood leads may be lower today (levels reported in previous NHANES surveys have
shown significant declines over time), but has no information upon which to project additional
declines from the 1991-1994 time frame to the present.
Using the IEUBK model and estimates of background soil-lead concentration, the
childhood health effect and blood-lead concentration endpoints were estimated at background
levels of lead in soil and dust. Selection of the background soil-lead concentration is described in
Section 5.2. The predicted health effect and blood-lead concentration endpoints at background
lead levels represent the risks that might exist if exposures to lead in paint, dust, and soil, except
soil at background concentration, could be eliminated. A comparison of the estimated endpoints
at background lead levels to the baseline estimates based on NHANES ffl provides an estimate of
the current risks to children due to exposure to lead in paint, soil, and dust.
The baseline risks based on Phase 2 of NHANES ffl are population-based risks; they
represent the risks posed by childhood lead exposure to our nation as a whole. The risk to
children exposed to specific levels of residential environmental lead were also computed using
the IEUBK and Rochester multimedia models. The Rochester multimedia model is a regression
model relating log-transformed childhood blood-lead concentration to dripline soil-lead
concentration, wipe dust-lead loading on floors (carpeted and uncarpeted), wipe dust-lead loading
on window sills, and an indicator of paint/pica hazard. The multimedia model was produced as
an intermediate step in the development of the empirical model. The empirical model was
developed specifically for estimating population-based risks. The multimedia model is a more
appropriate tool than the empirical model for estimating risks to children exposed to specific
environmental-lead levels. The IEUBK model was employed to predict the probability a child
will have a blood-lead concentration greater than or equal to 10 ug/dL for specific sets of soil-
lead and dust-lead concentrations. Three dust-lead concentrations were utilized (100,200, and
500 ppm) and soil-lead concentrations ranged from 25 to 2000 ppm. The Rochester multimedia
model was employed to predict the probability a child will have a blood-lead concentration
greater than or equal to 10 ug/dL for specific sets of soil-lead concentrations and floor and
window sill dust-lead loadings. Two soil-lead concentrations were utilized (100 and 400 ppm)
and dust-lead loadings ranged from 1 to 500 ng/ft2.
Sensitivity analyses were performed to gauge the robustness of the risk characterization
methodology. Factors evaluated include: 1) age group of interest, 2) relationship between blood-
lead concentration and IQ decrements, 3) assumptions on the national blood-lead distribution,
1-15
-------
4) adjustment made to correct for HUD National Survey dust sample analysis deficiencies,
5) variability of blood-lead concentrations, 6) daily dietary lead intake, and 7) assumptions on
contribution of paint pica tendencies to childhood blood-lead concentration.
1.4.2.2 Approach for Risk Management
The methodology developed for risk management is presented in the second part of the
risk analysis document, Risk Management. Figures 6-1 and 6-3 in Chapter 6 illustrate the
approach developed for analyzing example options for the §403 standards. The primary purpose
of the risk management analysis was to develop and apply methodology for analyzing example
options for the §403 standards. Under this methodology, both the IEUBK and empirical models
were used to obtain two distributions of blood-lead concentration: one resulting from exposure to
pre-§403 environmental-lead levels, and one resulting from exposure to lead levels expected to
result after intervention and other activities are conducted in response to the example option for
the §403 standards. Both models were applied to characterize the national distribution of blood-
lead concentrations of children aged 1-2 years in a pre-§403 environment using environmental
lead data collected in the HUD National Survey as inputs. The environmental-lead levels were
then adjusted to reflect the impact of interventions conducted in response to the §403 standards,
and a new blood-lead concentration distribution was estimated. The difference between these
distributions was assumed to be the decline in blood-lead concentrations attributable to
performing the interventions. Finally, the baseline distribution of blood-lead concentrations
(derived from Phase 2 of NHANES HI) and the just-estimated decline in the model-based
distribution of blood-lead concentration were used to derive a post-§403 blood-lead
concentration distribution.
Environmental-lead levels expected to result after interventions are conducted in response
to the §403 standards were determined as follows:
1. Observed levels of lead in paint, dust, and soil for HUD National Survey residential
units were compared to the example options for the §403 standards. Dust-lead
loadings were converted to wipe dust-lead loadings before comparison to the example
standards.
2. For those HUD National Survey residential units that had environmental-lead levels
above the example standards, interventions were triggered. If an intervention was
triggered, environmental-lead levels at the residential unit were set equal to assumed
post-intervention lead levels.
)
The Agency identified a limited number of intervention activities that were assumed to occur at
housing units identified as exceeding one or more of the example options for the §403 standards.
These intervention activities include both interim controls and more permanent abatement
measures. For each of the intervention activities, post-intervention environmental lead
conditions were assumed. Also, the expected duration of the reduction in environmental-lead
levels that result from each intervention was specified. In general, these intervention activities
are assumed to be medium-specific. For example, if a unit was identified as exceeding the
1-16
-------
standards for deteriorated exterior paint, the intervention activity would address only the exterior
paint. Exceptions are interventions dealing with either interior paint or activities involving soil
removal. These two interventions are assumed to be followed by cleaning of interior dust. This
is due to the expectation that these activities would create high levels of interior leaded dust that
would warrant special cleaning. This risk management analysis also assumes that proper
precautions are taken during the conduct of an exterior paint intervention to preclude the
requirement of a mandatory soil intervention following an exterior paint intervention.
The risk management methodology is utilized in Section 6.3 to analyze various example
options for the §403 standards. The example standards examined are not meant to encompass all
possible options for the §403 standards, and the Agency fully anticipates considering other sets of
candidate standards.
The major limitation with the approach developed for analyzing example options for the
§403 standards is the limited amount of data available for estimating pre- and post-§403
environmental-lead levels. This includes a lack of nationally-representative dust-lead loading
data (representing both pre- and post-§403 conditions) where samples were collected by wipe
techniques. This data limitation constitutes one of the major data gaps and limitations for the risk
management analyses. To help alleviate this limitation, sensitivity analyses were conducted to
assess the impact on the estimated endpoints of assumptions on methods for converting dust-lead
loadings, post-intervention environmental-lead levels, variability in childhood blood-lead
concentrations, daily dietary lead intake, and contribution of paint pica tendencies to childhood
blood-lead concentrations. In addition, an alternative approach for estimating the selected
endpoints that does not require specifying post-intervention environmental-lead levels was
examined in the sensitivity analysis for risk management.
1.5 PEER REVIEW
This report was reviewed independently by members of a peer review panel. The panel
consisted of a diverse group of six distinguished researchers who, together, had considerable
knowledge on all subject areas addressed in this report. The members of this panel and their
affiliations were:
Dr. Robert Bomshein, University of Cincinnati
Dr. Ruth Chen, Vanderbilt University Medical School
Mr. Victor Hasselblad, North Carolina State University
Dr. William Richards, Syracuse Research Institute
Dr. Charles Rohde, Johns Hopkins University
Mr. Joseph Schirmer, Wisconsin Division of Health, Bureau of Public Health
The charge given to this panel was to provide comments and responses to six general questions
and eight specific questions concerning the contents of this report. The six general questions
were as follows:
1-17
-------
1. Have we used the best available data?
2. Have we used this data appropriately?
3. Have we fairly characterized the variability, uncertainties, and limitations of the data
and our analysis?
4. Are there alternative approaches that would improve our ability to assess the relative
impacts of candidate options for paint, dust, and soil hazard standards?
5. The approach employs models that were primarily developed for use in site-specific
or localized assessments. Has the use and application of the IEUBK and empirical
model in this context been sufficiently explained and justified? Is our use of these
tools to estimate nationwide impacts technically sound?
6. Are there any critical differences in environmental lead-blood lead relationships found
in local communities that should be considered in interpreting our results at the
national level?
The eight specific questions (of which Question #4 had three parts) were as follows:
1. The HUD National Survey, conducted in 1989-90, measured lead levels in paint, dust
and soil in 284 privately owned houses. Does our use of this data constitute a
reasonable approach to estimating the national distribution of lead in paint, dust and
soil? Are any alternatives recommended? (Section 3.3)
2. Is the approach to evaluating the effects of pica for paint reasonable? Are there
alternative approaches that would be more appropriate? (Section 4.1.3, Appendix
Dl) '
3. The approach employs conversion factors to combine data from studies that used
different sample collection techniques. Is this appropriate? Is the method for
developing these conversion factor technically sound? (Section 4.3, Appendix X)
4a. IQ point deficits (Section 4.4) The approach characterizes IQ decrements in the
baseline blood-lead distribution, essentially implying that any blood-lead level above
zero results in IQ effects. Have we provided a sufficient technical justification for
this approach? Is this approach defensible and appropriate? (Section 4.4.1, Appendix
D2)
4b. IQ point deficits (Section 4.4) The characterization of IQ point loss in the population
includes the summation of fractional IQ points over the entire population of children.
Have we provided a sufficient technical justification for this approach? Is this
approach defensible and appropriate? (Section 4.4.1)
1-18
-------
4c. IQ point deficits (Section 4.4) One of the IQ-related endpoints is incidence of IQ less
than 70. Should consideration be given to what the IQ score was, or would have
been, prior to the decrement (i.e., should different consideration be given to cases
where a small, or even fractional, point decrement causes the <70 occurrence vs.
being <70 due to larger decrements)? If so, how might this be done?
5. Are the assumptions regarding duration and effectiveness of intervention activities
reasonable? (Section 6.1)
6. Removal and permanent cover (e.g., paving) are the only soil interventions
considered. Are there sufficient data on the effectiveness of other soil interventions
(e.g., vegetative cover) to quantify their ability to reduce exposure? (Section 6.1)
7. Are the combinations of standards used in Chapter 6 reasonably employed given the
potential interrelationships between levels of lead in different media? Are additional
data available on the interrelationship between lead levels in paint, dust, and soil prior
to and after abatement?
8. The approach for estimating health effect and blood-lead concentration endpoints
after interventions is based upon scaling projected declines in the distribution of
children's blood-lead concentrations to the distribution reported in Phase 2 of
NHANES HI. Under this approach, data collected in the HUD National Survey are
utilized to generate model-predicted distributions of blood-lead concentrations prior
to and after the rule making. The difference between the pre-section 403 and post-
section 403 model-predicted distributions is used to estimate the decline in the
distribution of children's blood-lead concentration. This decline is then
mathematically applied to the distribution reported in NHANES HI. Is this
adjustment scientifically defensible in general, and in the specific case where the
environmental datafrom the HUD survey-and the blood lead datafrom NHANES
ffl-were collected at different times (1989-90 vs. 1991-1994)? (Section 6.2)
The peer reviewers were also invited to provide general comments and suggestions concerning
the report.
In general, the peer reviewers concluded that the approaches and methods used in this
report were scientifically sound and that the data to which these methods were applied were the
most pertinent data available. However, the peer reviewers did provide useful suggestions for
revisions, as well as important issues to consider when interpreting results. The remainder of this
section discusses comments from the peer reviewers that were either important for interpreting
the study results or resulted in significant modifications to the report.
One reviewer suggested that additional confidence intervals be presented in the report.
Based on this comment, confidence intervals associated with the estimates of current risks
associated with childhood lead exposures (i.e., risks prior to implementing any proposed §403
rule) were calculated and presented hi Sections 3.4, 5.1, and 5.3. These risk estimates are labeled
1-19
-------
as "baseline" or "pre-§403" risks and are based on results reported in Phase 2 of the NHANES
ffl. However, confidence intervals were not calculated for other estimates within the report, such
as estimates of the reduced risks associated with lead exposures expected to exist after
implementing the proposed §403 standards. These risk estimates are labeled as "post-§403" risks
and are presented in Chapter 6, "Analysis of Example Options for the §403 Standards."
Confidence intervals for post-§403 risk estimates were not computed because it was not possible
to quantify important sources of variability required for their calculations, such as variability
associated with estimated post-intervention environmental-lead levels, conversions, and
intervention durations. Without including information on these sources of variability in the
equation calculation, the confidence intervals cannot be meaningfully interpreted.
A reviewer suggested that the Evaluation of the HUD Lead-Based Paint Hazard Control
Grant ("HUD Grantees") Program be included among the lead exposure studies whose data are
summarized within this document. This ongoing program provided among the most recent
information on lead levels in paint, dust, and soil within U.S. residences, as well as blood-lead
concentrations for children within these residences. Data collected prior to any interventions
conducted in this program were made available to this risk analysis, and were therefore added to
the data summaries in Sections 3.2, 3.3, and 3.4. However, when interpreting these data
summaries, one must consider that the housing units included in this program had high potential
for containing lead-based paint hazards or contained at least one child with an elevated blood-
lead concentration. This issue has also kept the HUD Grantees data from being used to develop
the empirical model used in this report to predict blood-lead concentration from environmental-
lead levels.
Comments were made to improve characterization of lead-based paint hazards from the
data available to the risk analysis. As a result, additional summary tables of lead levels in paint
were prepared for the lead-exposure studies whose data were summarized in Chapter 3. These
additional summaries included the percentage of surveyed units having lead-based paint on a
particular type of building component, the percentage having deteriorated lead-based paint on the
given component, and information on the distribution of maximum XRF measurement by
component type.
Upon suggestion from the peer reviewers, additional sensitivity analyses were performed
to investigate the impact of changes to the assumed values of various non-environmental
parameters in the IEUBK model, such as daily dietary lead intake and the geometric standard
deviation of children's blood-lead concentration under a common exposure scenario (Sections
5.4.6,5.4.7,6.4.6,6.4.7). Additional sensitivity analyses were conducted to assess assumptions
made when incorporating the effects of paint-chip pica on blood-lead concentration (Sections
5.4.8,6.4.8). Alternative assumptions on the percentage of children with paint-chip pica
tendencies, the percentage living in housing units with deteriorated lead-based paint who recently
ingested paint chips, and the blood-lead concentration of children who ingested paint chips either
recently or at some earlier time, were evaluated in the sensitivity analysis.
The risk analysis assumes a linear relationship between blood-lead concentration and
decline in IQ score. One peer review comment questioned whether there were ranges of blood-
1-20
-------
lead concentration where the relationship differs from what is assumed, thereby making it
nonlinear. Researchers have used both linear and log-linear models to predict decline in IQ score
as a function of blood-lead concentration. In his meta-analysis, Schwartz (1994) concluded that
the slope of the linear relationship becomes steeper at lower blood-lead concentrations,
suggesting a log-linear relationship. However, by assuming a linear relationship (i.e., the same
slope for all blood-lead concentrations), this risk analysis is more likely not to overestimate the
number of children with low blood-lead concentrations who benefit from intervention (if the
findings by Schwartz are true).
One peer reviewer suggested that the baseline risk estimates be adjusted to reflect
changes hi blood-lead concentration that may have occurred between when data used in this risk
analysis were collected and 1998. To address this issue, the sensitivity analysis includes an
investigation (Section 5.4.3) of how the baseline risk estimates are affected when blood-lead
concentrations are reduced by 10%, 20%, and 30% from values observed in Phase 2 of NHANES
HI (the survey whose data were used in this report to characterize blood-lead concentrations in
U.S. children). However, there was insufficient information to justify a single adjustment for the
entire nation. Any such adjustment would, therefore, have considerable variability and would
add considerably to the overall level of adjustment being made to the data in this risk analysis.
Other data adjustments recommended among the peer reviewers were not made for similar
reasons, such as adjusting for type of lead-based paint exposure (paint loading, surface type, paint
condition, substrate, and building component) and for regional differences (diet, architecture,
climate).
Peer reviewer comments were also useful in determining additional post-intervention
settings for dust-lead loadings and soil-lead concentrations for which risk estimates were
calculated. However, these additional results were added to Chapter 6 of the report only when
they provided additional information from what already existed hi this chapter.
EPA has established a public record for the peer review of this report under
administrative record AR-188, "Risk Analysis to Support Standards for Lead in Paint, Dust, and
Soil: Peer Review." The record is available in the TSCA Nonconfidential Information Center,
which is open from noon to 4 PM Eastern time Monday through Friday, except legal holidays.
The TSCA Nonconfidential Information Center is located hi Room NE-B607, Northeast Mall,
401 M Street SW, Washington, DC.
1-21
-------
This page intentionally blank
-------
PARTI
RISK ASSESSMENT
-------
This page intentionally blank.
-------
2.0 HAZARD IDENTIFICATION
CHAPTER 2 SUMMARY
This chapter presents information on the toxicity of lead, through a
discussion of how body-lead burden is measured, how lead works in the body, the
resulting adverse health effects, and populations at risk. Elevated blood-lead
concentration and selected Intelligence Quotient (IQ) measures are identified to
represent the adverse health effects resulting from lead exposure. The elevated
blood-lead concentration thresholds selected for this risk analysis are among those
established by CDC as levels of concern. In addition, a large body of evidence
suggests that IQ measures are impacted adversely in children exposed to lead.
These endpoints are used in this risk analysis to estimate the benefits of the
proposed §403 rule.
Blood-lead concentration is a commonly used measure of body lead burden.
An extensive body of research relates health effects of lead exposure to blood-lead
concentration. For example, lead-related reductions in intelligence, impaired hearing
acuity, and interference with vitamin D metabolism have been documented in
children at blood-lead concentrations as low as 10 to 15 ug/dL, with no apparent
threshold. At higher exposure levels, these effects become more pronounced and
other adverse health effects are observed in a broader range of body systems.
Increased blood pressure, delayed reaction times, anemia, and kidney disease may
become apparent at blood-lead concentrations between 20 and 40 ug/dL
Symptoms of very severe lead poisoning, such as kidney failure, abdominal pain,
nausea and vomiting, and pronounced mental retardation, can occur at blood-lead
levels as low as 60 ug/dL At even higher levels, convulsions, coma, and death
may result.
Figure 2-1 outlines the approach for the hazard assessment. The
conclusions from the hazard assessment are presented in Section 2.6.
The goal of the hazard identification is to answer the following questions:
1. How is lead exposure measured in the human body?
2. What measure of body lead burden should be used in this risk analysis?
3. How does lead work in the body?
4. What adverse health effects are linked to lead exposure?
5. What is the best population for measuring the adverse health effects of lead exposure
hi this risk analysis?
6. What health endpoints should be quantified in the risk analysis?
2-1
-------
Background
and
Objectives
JL.
HAZARD IDENTIFICATION
Measure Body-
Lead Burden
(Section 2.1)
-i
t
*-\
Identify the
Mechanisms
of Lead Toxicity
(Section 2.2)
B^'-^^tff^
Identify Health Effects
of Lead Exposure
(Section 2.3)
!-Sft&\f > »i ' "I"
Select a
Representative
Population
(Section 2.4)
Select Health Effect
Endpoints and Blood-Lead
Concentration Thresholds
(Section 2.5)
Exposure
Assessment
Dose-Response
Assessment
Risk
Characterization
Risk
Management
Conclusions on
Risk
Characterization
Conclusions on
Analysis of Example
Options for §403
Standards
Figure 2-1. Detailed Flowchart of the Approach to Hazard Identification.
2-2
-------
Questions 1,3, and 4 identify the hazardous effects of lead exposure. Questions 2, 5, and 6
address how the risk due to lead exposure is assessed herein. Methods for quantifying die health
endpoints identified in this chapter are described hi the dose response assessment (Chapter 4).
These methods are used hi the risk characterization (Chapter 5) and risk management analysis
(Chapter 6) to estimate the current and future risks of childhood lead exposure.
Figure 2-1 illustrates the relationship between the approach presented in this chapter and
the risk analysis approach. The information presented hi this chapter follows the flow of Figure
2-1 and the questions stated above. Namely, this chapter presents information on lead toxiciry,
through a discussion of how body-lead burden is measured (Section 2.1), how lead works in the
body (Section 2.2), and the resulting adverse health effects (Section 2.3). This chapter
documents several decisions that are relevant to the assessment of health risks due to lead
exposure. These include the selection of blood-lead concentration as the measure of body lead
burden (Section 2.1), selection of children aged 1-2 years as the representative population
(Section 2.4), and selection of specific elevated blood-lead concentration and health effect
endpoints (Section 2.5) that are used for the quantification of risk. The hazard characterization
(Section 2.6) summarizes the materials presented in this chapter and addresses the strengths and
weaknesses of the scientific evidence and decisions made, as they are relevant to this risk
assessment.
There is an extensive body of literature relating health effects of lead exposure to
measures of body-lead burden. This literature is summarized in several government reports,
including
Air Quality Criteria for Lead (USEPA, 1986)
The Nature and Extent of Lead Poisoning hi Children hi the United States: A Report
to Congress (ATSDR, 1988a)
Ah- Quality Criteria for Lead: Supplement to the 1986 Addendum (USEPA, 1990a)
Comprehensive and Workable Plan for the Abatement of Lead-Based Paint hi
Privately Owned Housing (HUD, 1990)
Preventing Lead Poisoning in Young Children (CDC, 1991)
Toxicological Profile for Lead (ATSDR, 1993)
These sources were used extensively hi the next sections, although the original sources are cited
for specific results whenever possible.
2.1 MEASURES OF BODY-LEAD BURDEN
For purposes of risk assessment, it would be ideal to precisely relate particular health
outcomes, such as learning deficits or decreased motor coordination, to environmental lead
levels. Unfortunately, most studies of lead hi the environment use measures of body-lead
2-3
-------
burden, such as blood-lead concentration, as biomarkers of lead exposure. Similarly, studies that
assess lead hazard interventions tend to use blood-lead concentration to measure intervention
effectiveness (USEPA, 1995b). There is extensive evidence that body-lead burden is associated
with lead levels in environmental media (USEPA, 1986; CDC, 1991). In addition, there is an
extensive body of literature relating health effects of lead exposure to measures of body-lead
burden (ATSDR, 1993; CDC, 1991).
The most common screening and diagnostic measure of body-lead burden is blood-lead
concentration. Blood-lead concentration has the advantage of being easily and inexpensively
measured. A disadvantage, however, is that it reflects a mixture of both recent and past
exposure. Because lead cycles between the blood and bone, a single blood lead measurement
cannot distinguish between low-level chronic exposure and high-level acute exposure (ATSDR,
1993). Both types of exposure could result in the same blood-lead concentration. Despite this
limitation, blood-lead concentration remains the one readily accessible measure that can
demonstrate hi a relative way the relationship of various effects to changes in lead exposure
(ATSDR, 1993).
Other measures of body-lead burden include lead hi bones, teeth, and hair. Of the other
measures, bone and tooth lead may be used to measure cumulative exposure to lead, while hair
lead is an indicator of more recent exposure. Bone-lead content may be measured by x-ray
fluorescence (XRF), although the reliability of this method has been questioned in the past
(Wedeen, 1988). While the reliability of the XRF method to measure bone-lead has improved in
recent years, it is still used primarily for research. Since teeth can store lead up to the time of
shedding or extraction, levels of lead in shed teeth have been used as an indicator of lead
exposure in some studies (Smith et al., 1983; Bergomi, et al., 1989; Pocock et al., 1989;
Needleman et al., 1990). Hair lead has been used as an indicator for intermediate exposure
(2 months) in children (Wilhelm et al., 1989). However, artificial hair treatments such as dyeing,
bleaching, or permanents, can invalidate metal analysis of hair in adults (Wilhelm et al., 1989).
In addition, external surface contamination problems make it difficult to differentiate between
externally and internally deposited lead (USEPA, 1986). Due to the disadvantages associated
with using bone, tooth, and hair lead as biomarkers of exposure, most researchers in the area of
lead exposure conclude that blood lead is the most efficient and useful way to assess body lead
burden.
Physiological changes that are known to implicate lead exposure may also be used as
biomarkers of exposure. For example, interference with heme synthesis following lead exposure
can lead to a reduction of hemoglobin concentration hi blood (Bernard and Becker, 1988) and an
increase in urinary coproporphyrin (USEPA, 1986). In addition, the concentration of erythrocyte
protoporphyrin (EP) rises above background at blood-lead levels of 25 to 30 ug/dL and there is
an association between blood-lead levels and EP (Hernberg et al., 1970; CDC, 1985). The level
of EP in blood is used as an indicator of past chronic exposure, since elevated EP reflects average
blood-lead levels for about 4 months following the exposure (Janin et al., 1985). In the case of
each of these physiological measures, other conditions may produce similar effects, leading to
false positive outcomes when these measures are used alone as biomarkers for body lead burden.
2-4
-------
Thus, generally, blood-lead levels are determined concurrently with these physiological
biomarkers.
2.2 MECHANISMS OF LEAD TOXICITY
Lead is a very dynamic element with a wide spectrum of effects in humans. Its effects are
seen at the subcellular level as well as at the level of general function that encompasses all
systems in the body. The subcellular mechanisms of action are included in this section, followed
by a discussion of the neurotoxic effects and the heme effects of lead poisoning. Wherever
possible, mechanisms included in the subcellular mechanisms section are related to the specific
effects of lead on the nervous system and the blood. There remain many gaps, however, in the
information needed to explain the varied mechanisms of lead in different organs.
Lead primarily enters the body through ingestion (eating and drinking) and inhalation
(breathing in air). It can also pass through the skin. Lead is absorbed, distributed throughout the
body, and removed from the body (excreted). The rate of lead absorption into the body depends
on the chemical and physical properties of the form of lead and the physiological characteristics
of the exposed person. For example, when inhaled, factors such as the lead particle size and
shape and the individual's ventilation rate influence how lead will be deposited in and absorbed
by the respiratory tract. Large particles, which may be encountered in an occupational setting,
tend to be deposited in the upper airways and may be indirectly absorbed by swallowing and
absorption in the stomach. Smaller particles tend to be deposited in the bronchial region of the
lung, and particles less than one micron, which is typical for urban air, reach the lower
respiratory tract where they can be directly absorbed across the thin walls of the alveolar sacs and
enter the blood. Ventilation rate is important because altering the inhalation rate may increase or
decrease the amount of the lead ultimately absorbed by the lung (Klaassen, 1993). The
respiratory deposition of airborne lead encountered in the general population ranges from 30-50
percent (Kehoe, 1961 a,b,c; Nozaki, 1966; Chamberlain, 1978; Morrow, 1980; Gross, 1981).
Several studies conducted in humans (Rabinowitz, 1977; Chamberlain, 1978; Morrow, 1980) and
animals (Pott and Brockhaus, 1971; Boudene, 1977; Kendall, 1975; Morgan and Holmes, 1978;
Greenhalgh, 1979) have indicated that lead deposited in the lower respiratory tract is completely
absorbed. Thus, the absorption rate is governed by the deposition rate and 30-50% of inhaled
lead is absorbed (USEPA, 1986). A respiratory deposition/absorption rate of 25-45% has been
estimated for children (USEPA, 1989a).
The amount of lead absorbed from the gastrointestinal tract of adults is 10-15% of the
amount ingested (Kehoe, 1961a,b,c; Hursh and Suomela, 1968; Harrison, 1969). In pregnant
women and children, the amount of lead absorbed via ingestion can increase to 50% (Alexander,
1973; Heard and Chamberlain, 1982; Rabinowitz and Needleman, 1982; USEPA, 1979). The
amount of lead absorbed by ingestion greatly increases during periods of iron or calcium
deficiency. Once absorbed, lead is distributed by the blood to the mineralizing tissues (bone and
teeth) and soft tissues (kidney, bone marrow, liver and brain). For adults, following exposure to
a single dose of lead, one-half of the lead from the original exposure remains hi the blood for
about 25 days after exposure, in soft tissues it remains for about 40 days, and in bone for more
than 25 years (Rabinowitz et al., 1976). Consequently, after a single exposure, a person's blood-
2-5
-------
lead concentration may begin to return to normal, but the total body burden (amount of lead in
the body) may still be elevated.
2.2.1 Physiological Mechanisms
The biological basis for many aspects of lead toxicity appears to relate to lead's ability to
bind (attach) to substances crucial to various physiological functions. For example, lead may
interfere with cell function by competing with essential minerals such as calcium and zinc for
binding sites on membranes and proteins. Lead binding to enzymatic proteins can inhibit the
activity of the enzyme and alter the processing of other chemicals (metabolism). Lead binding to
membranes or transport proteins can inhibit or alter ion transport across the membrane or within
a cell. The effects of lead are modulated by its distribution in the body, its affinity for various
binding sites, and differences in cellular composition and structure within tissues and organs. As
a result, there is no single well-defined mechanism that explains the toxicological activity of lead
in all tissues (USEPA, 1986).
Studies of the mechanism of lead toxicity at the cellular level implicate cell and
subcellular (organelle) membranes as a primary target for lead (USEPA, 1986). Lead-induced
alterations of ion transport, particularly calcium ions, are related to a number of the health effects
associated with lead exposure. Effects on ion-transport lead to inhibition of enzymes and or
signaling proteins and interferes with normal cellular processes. The overall impact of these
effects is to disturb the development and functioning of many organ systems, particularly the
central nervous system (USEPA, 1986).
The mitochondria appear to be particularly sensitive to lead (USEPA 1986). Lead causes
both structural changes and disturbances in mitochondria! function. Mitochondria exposed to
lead expand or swell and there is distortion and loss of the small folds of the inner membrane
called the cristae. The mitochondria! enzymes responsible for cellular respiration are largely
located within the cristae. Thus, lead uncouples energy metabolism and inhibits cellular
respiration (USEPA, 1986). Lead also alters the mitochondrial distribution of calcium (USEPA,
1986).
2.2.2 Neurotoxic Effects of Lead
The mechanisms for lead neurotoxicity are not well understood. Several mechanisms
have been proposed to explain why children are more sensitive than adults to the neurotoxic
effects of lead and how lead affects the nervous system on the molecular level.
For over a decade, the hippocampus has been considered to be the principal target of lead
in the brain because: 1) the hippocampus contains relatively high concentrations of zinc, and
zinc-dependent functions may be sensitive to lead, 2) the hippocampus contains a dense plexus
of cholinergic fibers that are affected by lead exposure, and 3) the hippocampus functions in
memory and learning (Petit, 1983). More recent investigations have shown that other areas of
the brain, particularly the mesolimbic system (Lasley and Lane, 1988; Moresco et al., 1988),
where low levels of lead have been found, cannot be excluded as a target site. Continuing
2-6
-------
research may help to determine which areas of the brain have an affinity for lead and are affected
by it.
A number of scientists working on the neurotoxicity of lead have proposed mechanisms
of how lead affects the nervous system. Among these scientists, Silbergeld (1992) and Bellinger
(1995) both discuss possible mechanisms for lead neurotoxicity in the context of
neurodevelopmental and neuropharmacological effects.
Neurodevelopmental Effects: During development, the central nervous system (the
brain and spinal cord) goes through a number of changes involving an overall growth in cell
numbers, with a resultant increase in the size of the organ. In addition, cells develop specialized
functions and there is a proliferation and outgrowth of the nerve cell projections that establish
connections between cells (Silbergeld, 1992). Many substances regulate these processes,
including growth factors, neurotransmitters functioning as trophic agents, and glycoprotein cell
adhesion molecules (Jacobson, 1990).
One of the potential mechanisms for lead's effect on the developing brain has been
investigated by Goldstein (1990), who suggests that the immature endothelial cells forming the
capillaries of the developing brain are more permeable to lead than are capillaries from mature
brains. As a result, lead in the blood may easily pass into the newly forming compartments of the
brain and affect many parts of this developing organ, hi comparison, the capillaries of adults are
developed and help to prevent the passage of lead (in its ionic form) across the blood-brain
barrier. It has been suggested that lead may affect the differentiation of capillary endothelial cells
in the fetal brain, similar to the way it affects developing neurons (Bressler and Goldstein, 1991).
This hypothesis provides a basis for the increased risk of pregnant women, infants, and young
children to the neurotoxic effects of lead.
Silbergeld (1991) found that exposure of fetal animals to lead affects both regional
growth and neuron-specific differentiation/synaptogenesis (development of synapses) in the
central nervous system. Of these, synaptogenesis appears to be the more sensitive (Regan, 1989;
Silbergeld, 1991). A synapse is a junction where the axon of one neuronal cell (or neuron)
terminates with the dendrite of another neuron. Nerve impulses move from one nerve cell to
another by traveling through the synapse. The normally functioning brain seems to exhibit a
deletion of synapses that are unused. Synapses which are frequently used are kept and
strengthened. Goldstein (1990,1992) suggests that lead may disrupt, or delay, this normal
synaptic developmental process, and that perhaps the resulting connections hi the brain are
"poorly chosen," leading to functional impairment. Although this hypothesis is speculative,
lead's ability to facilitate the unstimulated release or prevent the stimulated release of
neurotransmitters, which are important for the morphological organization of neurons, may be
related to how neurons are chosen to survive (Audesirk, 1985). This may result in a nervous
system that appears normal but in which cell to cell connections are not normal. These
abnormalities may be translated into neurobehavioral deficits which result in cognitive and
behavioral deficits.
2-7
-------
Neuropharmacological Effects of Lead: Lead may also act as a neuropharmacological
toxicant in the brain (Silbergeld, 1992; Bellinger, 1995). Silbergeld (1992) proposes that lead
interferes with the synaptic release of neurotransmitters from neurons and signal transduction.
Theoretically, these effects are reversible if lead is removed from the synapse. However,
exposure to lead for a long time may result in permanent alteration in cellular responsiveness at
pre- and post-synaptic levels. The pharmacologic effects of lead may include facilitated
transmitter release, modulation of ion conductance and, as a result, altered the
electrophysiological output of the neuron.
Disruption of ion transport at membranes may be the mechanism by which lead produces
its pharmacologic effects in the nervous system. Lead can substitute for calcium and zinc in ion
transport events at the synapse. While the exact biochemical mechanisms of lead toxicity are
unknown, at least some of its deleterious effects are attributed to interference with the functions
of sodium channels, calcium channels, calcium-binding modulators like calmodulin, messengers
like adenyl cyclase and protein kinase C ( Bessler and Goldstein, 1991). Lead may affect ion
channels by occupying zinc-binding sites and preventing ion movements (Alkondon, 1990).
At the neuron, mitochondrial release of calcium is quite sensitive to lead (Silbergeld and
Adler, 1978). Protein kinase C, which is very sensitive to lead, modulates receptor currents
affecting long-term potentiation and other forms of synaptic response that may underlie learning
and memory (Markovac and Goldstein, 1988). Dopamine-sensitive adenyl cyclate and
(Na+,K+)-ATPase, are also relatively sensitive to lead (Ewers and Erbe, 1980, Fox, 1991).
Neurotransmitter release or transmitter-gated ion channels are sensitive to higher concentrations
of lead (Kostial and Vouk, 1957; Silbergeld et al., 1974; Audesirk, 1985; Minnema et al., 1986;
Alkondon etal., 1990).
The differential ability to prevent lead entry into the neuron may be an important
protective mechanism to prevent the neurotoxic effects of lead. There has been speculation of a
lead-binding protein in humans (DuVal and Fowler, 1989) which may serve to concentrate and
transport lead to certain parts of the brain.
Peripheral Neuropathy: Lead induces degeneration of the protective Schwann cells in
the motor neurons of the peripheral nervous system. This causes segmental loss of the myelin
covering of the neuron and possible neuron degeneration (Fullerton, 1966). Dyck et al., (1980)
and Windebank et al. (1980) suggest that lead induces a breakdown in the blood-nerve barrier,
allowing lead and fluid to enter the endoneurium and disrupt the myelin membranes. The
degeneration of sciatic and tibial nerve roots is also possible. Sensory nerves are less sensitive to
lead than motor nerves. Peripheral neuropathy is usually present only after prolonged high
exposure to lead. Studies of occupationally exposed workers indicate that motor nerve
dysfunction can occur at blood-lead levels below 70 |ig/dL, possibly as low as 30 ug/dL (Araki
et al., 1980 and 1992; Rosen et al., 1983; Seppalainen, et al., 1983; Hirata and Kosaka, 1993;
Chia et al., 1996), when assessed clinically by the electrophysiologic measurement of nerve
conduction velocities. There is some evidence that these effects may be reversible (Araki et al.,
1980; Muijser et al., 1987).
2-8
-------
2.2.3 Hematologic Effects of Lead
Lead has adverse effects on heme synthesis and red blood cell formation. These effects
can result in anemia and decreased life span of red blood cells. Hemoglobin is a major
constituent of red blood cells. Hemoglobin consists of the protein globin and heme, which is a
metal complex consisting of an iron atom in the center of a porphyrin structure. The oxygenated
form of hemoglobin provides the red color to red blood cells. The effects of lead on heme
synthesis and hemoglobin production are described in detail in two reports, the Air Quality
Criteria for Lead (USEPA, 1986) and The Nature and Extent of Lead Poisoning in Children in
the United States: A Report to Congress (ATSDR, 1988a). A short summary follows.
Effect of Lead on Heme Synthesis: When an individual is exposed, lead quickly
reaches the blood, circulates in the body, and enters different tissues including the bone marrow,
where it can have an impact on various reactions involved in the formation of heme. The process
of heme biosynthesis starts with glycine and succinyl-coenzyme A, proceeds through the
formation of a molecule called protoporphyrin DC, and culminates with the insertion of iron into
the porphyrin ring to form heme. In addition to being a constituent of hemoglobin, heme is
found hi many hemoproteins, such as myoglobin, the P-450 component of the mixed-function
oxidase system, and the cytochromes of cellular energetics. Therefore, disturbing heme
biosynthesis by exposure to lead poses the potential for multiple-organ toxicity.
Lead's mechanism of action seems to be due to its effect on cellular mitochondria. Lead
enters the mitochondria of the cell where it impairs mitochondrial function and thus adversely
impacts the production of heme. In the mitochondria, lead increases the activity of the enzyme,
5-aininolevulenic acid synthetase (ALA-S), which increases the amount of 6-aminolevulinic acid
(ALA) formed. Lead, in the cytosol of the cell, decreases the activity of 6-aminolevulenic acid
dehydrase (ALA-D), an enzyme that catalyzes reaction of ALA in heme biosynthesis. The result
is an increase in the level of ALA (a potential neurotoxin) and a decrease hi the production of the
porphyrin needed for heme synthesis.
Ferrochelatase, an enzyme also found in the mitochondria, catalyzes the incorporation of
iron into protoporphyrin IX to form heme. Lead tends to inhibit ferrochelatase from
incorporating the iron into the protoporphyrin ring, thereby preventing the formation of heme.
Instead, there is an increase hi erythrocyte protoporphyrin hi the red blood cells. Erythrocyte
protoporphyrin (EP) can be measured hi blood as zinc protoporphyrin (ZPP) or free erythrocyte
protoporphyrin (FEP).
Effect of Lead on Hemoglobin Production and Red Cell Formation: As described
above, heme production is decreased by lead. Heme production mediates globin production
through a synchrony between the rates of globin and heme syntheses, hi the absence of heme, the
polyribosomes disaggregate and globin synthesis ceases. Accordingly, globin production is
decreased, resulting in decreased production of hemoglobin. These effects can lead to anemia
(reduction hi circulating red blood cell mass.) Lead exposure can lead to anemia in two ways. It
causes increased destruction of the red blood cells (hemolysis) and impairs red cell formation
resulting in hypochromatic (light colored) normocytic (normal size) cells.
2-9
-------
The molecular mechanism for the diminished red cell life span is thought to be due to
lead's inhibition of the active transport (Na+, K+)-ATPase enzyme system. If the active
transport is paralyzed, the cells accumulate sodium and water until a critical volume is reached
and men hemolysis (destruction of cells through rupture of the cell membrane) ensues. Also,
enzymes such as glucose-6-phosphate dehydrogenase (G6PD) may be affected by lead. G6PD
catalyzes the initial step in the pentose phosphate pathway of carbohydrate metabolism, through
which the reduced gluthathione reductase (GSH) is generated for maintaining the sulfhydryl
groups within the red blood cell and perhaps the red blood cell membrane.
2.3 HEALTH EFFECTS OF LEAD EXPOSURE
Lead is a powerful toxicant with no known beneficial purpose in the human body
(ATSDR, 1988a). The toxic effects of lead are seen primarily in the central nervous system, but
virtually all parts of the body can be damaged at high exposure levels. Specific health effects
from lead exposure, the blood-lead levels at which these effects have been observed, and the
scientific literature in which the effects were reported are summarized in Table B-l in
Appendix B. This table is reproduced from the Toxicological Profile for Lead (ATSDR, 1993).
2.3.1 Neurological Effects of Lead
The most severe neurological effect of lead is encephalopathy. Early symptoms of
encephalopathy include irritability, poor attention span, headache, muscular tremor, loss of
memory, and hallucinations. As encephalopathy increases more severe symptoms appear,
including delirium, convulsions, paralysis, coma, and death (Kumar et al., 1987). High-level
exposure to lead produces encephalopathy in children, starting with blood-lead levels of
approximately 80 to 100 ug/dL (Bradley and Baumgartner, 1958; Gant, 1938; Bradley et al.,
1956; NAS, 1972; Rummo et al., 1979; Smith et al., 1983; EPA, 1986).
The effect of lead on intelligence quotient (IQ) and other developmental indicators is
well-established for children with markedly elevated blood-lead concentrations. For example,
five point IQ decrements, fine motor dysfunction, and altered behavioral profiles were reported
among preschool children who ingested paint and plaster (pica) and whose blood-lead levels
were greater than 40 ug/dL (mean of 58 ug/dL), when compared with matched controls who did
not eat paint and plaster (de la Burde and Choate, 1972). At age 7 to 8, a three point IQ
decrement and impairment in learning and behavior were reported for these children, even
though blood-lead levels had declined (de la Burde and Choate, 1975). Blood-lead
concentrations for control children were not reported, but, given the date of the study, children in
the control population may have had what would now be considered elevated blood-lead levels.
A study that included children who had previously had encephalopathy indicated that these
children had increased incidence of hyperactivity and IQ decrements of approximately 16 points
resulting from lead exposure (Rummo et al., 1979). In the same study, asymptomatic children
with long-term exposure (mean blood-lead levels of 51-56 ug/dL) had IQ decrements of 5 points
on average, compared to control children (mean blood-lead levels of 21 ug/dL).
2-10
-------
Long-lasting impacts on intelligence, motor control, hearing, and neurobehavioral
development of children also have been documented at blood-lead levels that are not associated
with obvious symptoms and were once thought to be safe.
Results are available from four large-scale, longitudinal studies of lead exposed children
conducted in Boston, Cincinnati, Cleveland, and Port Pine, Australia. These studies indicate that
disturbances hi early neurobehavioral development occur at exposure levels that until recently
were considered safe, or even normal. In the Boston study, 4-8 point differences in performance
on the Bayley Mental Development Index (MDI) were reported at 6,12,18, and 24 months, after
adjusting for other covariates, when children with low blood-lead levels (prenatal mean of 1.8
|jg/dL) were compared to children with high blood-lead levels (prenatal mean of 14.6 |ig/dL)
(Bellinger et al., 1985a, 1985b, 1986a, 1986b, 1987a). These findings were confirmed in more
recent studies (Bellinger et al., 1989a, 1989b). Additional follow-up showed that deficits in
McCarthy General Cognitive Index scores at age 5 were significantly correlated with blood-lead
levels at age 24 months, although not with prenatal blood lead measures (Bellinger et al., 1991).
Similar results were reported in the Cincinnati study (Dietrich et al., 1986,1987a, 1987b). These
study results suggest that the effect of prenatal lead exposure on the MDI was mediated hi part
through its effects on bulb weight and gestational age, which were each significantly associated
with MDI scores (Dietrich et al., 1987a). Results reported for the Cleveland study were mixed,
but while the authors tended to conclude that there was not strong evidence of developmental
effects of lead (Emhart et al., 1985,1986,1987; Wolf et al., 1985; Ernhart and Green, 1990),
other reviewers suggest that such effects may be inferred from the reported results (EPA, 1986;
Davis and Svendsgaard, 1987; ATSDR, 1993). In the Port Pirie study, reduced MDI scores at 24
months were associated with postnatal blood-lead levels measured at age 6 months, but not with
prenatal exposure measured through cord and maternal blood-lead levels (Vimpani et al., 1985,
1989; Baghurst et al., 1987; Wigg et al., 1988). Results of a follow-up neurobehavioral
assessment conducted at age 3 to 4 years, using the McCarthy Scales of Children's Abilities,
indicated significant inverse correlations between postnatal blood-lead levels (geometric means
of 14 fig/dL at 6 months and approximately 21 ug/dL at 15 and 24 months) and ability test scores
(McMichael et al., 1988).
In addition to the effects on early neurobehavioral development, all four studies report
lower IQ scores at school-age for children who had earlier exhibited elevated blood-lead levels.
In Boston, slightly elevated blood-lead levels at age 24 months (mean of 6.5 fig/dL) were
associated with intellectual and academic performance deficits at age 10 years (Bellinger, 1992).
In Cincinnati, postnatal blood-lead levels measured through age 3 years were inversely associated
with IQ scores measured at age 5, although the effect was not statistically significant when
adjusted for covariates (Dietrich et al., 1993). hi Cleveland, a significant association was
reported between blood-lead concentration at age 2 (mean of 16.7 |ig/dL) and IQ measured at 5
years (Ernhart et al., 1989). In Port Pirie, statistically significant associations were reported
between IQ measured at age 7 and blood-lead levels from birth through age 7, with the strongest
associations for blood-lead levels measured at 15 months to 4 years (Baghurst et al., 1992).
Taken together, these studies provide strong evidence that low-level prenatal or early
postnatal exposure to lead results hi neurobehavioral developmental delays that persist through
2-11
-------
age 5. Strong relationships between blood-lead concentration in early childhood, age 15 months
to 4 years, and IQ scores were also reported, even when only slight elevations in blood-lead
levels were present.
Additional evidence of IQ point loss associated with lead exposure hi school-age children
is reported in cross-sectional studies throughout the world. For example, a study of Danish
children related tooth-lead concentration to performance on several psychometric tests (Hansen
et al., 1989). Children with elevated tooth-lead levels (above 18.7 ng/g) were matched by sex
and socioeconomic status with children with lower levels (below 5 ng/g). High lead children
scored lower on the Wechsler Intelligence Scales for Children (WISC) IQ test than children with
lower lead levels, although no difference in scores was observed for the Performance IQ and
several experimental tests. Impaired neuropsychological functioning due to lead exposure was
observed through differences in performance on the Bender Visual Motor Gestalt Test and on a
behavioral rating scale, hi addition, a study of school children in Edinburgh, Scotland (Fulton
et al., 1987) found that elevated blood-lead levels (mean of 11.5 |ig/dL) were associated with
lower scores on IQ tests and on mathematical and reading attainment tests, after adjusting for
covariates. No threshold in the relationship, below which lead does not have an effect on
intelligence and attainment, was observed even for blood-lead concentrations below 10 jxg/dL. A
study of Chinese children (Wang et al., 1989) also reported a significant dose-response
relationship between blood-lead concentration (above 10 ng/dL) and IQ scores, after adjusting
for covariates.
A significant effect of lead on IQ is not uniformly reported, however. Children randomly
selected from birth records hi Birmingham, United Kingdom, were assessed using a variety of
cognitive, performance, neuropsychological, and behavioral endpoints (Harvey et al., 1988). The
effect of lead (mean of 13.5 ug/dL) was not significant for most endpoints, and for none of the
three IQ measures, hi a study of 6 year old children hi London, both tooth lead and blood lead
were examined as predictors of intelligence (Smith et al., 1983; Pocock et al., 1989). Neither
measure of lead exposure was a significant predictor, once social factors were controlled. No
evidence of an association between blood-lead levels (mean of 12.75 ug/dL) and intelligence was
reported in another study of London children that included more middle class families
(Lansdown et al., 1986).
A possible explanation for these seemingly contradictory results is that the effect of lead
on IQ may be overshadowed by the effects of home and societal factors, such as birth order,
parental IQ and level of education, and socioeconomic status. For example, a study of 104
children under age 7 and of lower socioeconomic status indicated that MDI and IQ scores were
significantly associated with blood-lead levels ranging from 6 to 59 ^ig/dL, after controlling for
socioeconomic and other factors (Schroeder et al., 1985). In a five-year follow-up of 50 of these
children, IQ was inversely correlated with initial and concurrent blood-lead levels, but the effect
of lead was not significant when socioeconomic status and other covariates were included hi the
analysis (Schroeder and Hawk, 1987). However, hi a replication of the study among children of
uniformly low socioeconomic status, the effect of lead was evident at both the initial and five-
year follow-up (Hawk et al., 1986; Schroeder and Hawk, 1987). These results suggest that the
2-12
-------
effects of lead may be more easily detected in groups with similar home and societal
backgrounds.
Both current and long-term indicators of lead exposure have been studied to establish
which indicator was most strongly correlated with psychometric test scores (Bergomi et al.,
1989). Total and verbal IQ scores were negatively correlated with tooth-lead levels and
5-aminolevulenic acid dehydrase activity. Tooth-lead levels were also negatively correlated with
Toulouse Pieron test results, which evaluate figure identification ability, discrimination, and
attention. The most predictive measure of lead exposure was tooth lead, which is indicative of
cumulative lead exposure. Blood lead, which is indicative of a mix of current and past exposure,
and hair lead, which is indicative of short-term exposure, had little predictive value in this study.
A study of the long-term effects of low-level lead exposure found that children with
higher dentin lead levels were more likely to drop out of high school and have a reading
disability (Needleman et al., 1990). Higher lead levels were also associated with lower ranking
in high school class and increased absenteeism. Lower scores on vocabulary and grammatical-
reasoning tests were reported, along with poor hand-eye coordination, delayed reaction times,
and slowed finger tapping, compared to children with lower lead exposure. Earlier results
indicated that children with high dentin lead levels had deficits in IQ scores, speech and language
processing, attention, and classroom performance in first and second grades (Needleman et al.,
1979). IQ deficits continued through the fifth grade. In addition, children with higher lead levels
needed more special academic services and had a higher failure rate hi school (Bellinger et al.,
1986c).
A lead-related decrease in hearing acuity has been reported hi young children, with
hearing thresholds at 2000 Hz increasing with blood-lead levels in the range of 6 to 59 ug/dL
(ATSDR, 1993). Analysis of NHANES H data indicated that the probability of increased
hearing thresholds at 500,1000,2000, and 4000 Hz was associated with increased blood-lead
levels from below 4 ng/dL to over 50 ug/dL. In addition, this study reported increased
probability that a child was hyperactive and delayed developmental milestones (age at which
child first sat up, walked, and talked) associated with elevated blood lead (Schwartz and Otto,
1987).
2.3.2 Other Effects of Lead
Hematological Effects: The effects of lead on the blood's biochemical functions are
interrelated and have variable biological impact. Heme (the component of hemoglobin that binds
iron) is critical to the basic function of all cells due to its presence in the cytochromes involved in
energy production. As noted earlier, lead can disturb the formation of hemoglobin leading to
anemia at high exposure levels. The heme-mediated generation of an important hormonal
metabolite of vitamin D (1,25-dihydroxyvitamin D) may be disturbed by lead. This hormone
serves a number of functions hi humans, including the regulation of calcium metabolism. In
addition to the direct effects of lead on heme biosynthesis, there are potentially significant
indirect impacts on the central nervous system, caused by the accumulation of the potential
neurotoxicant, 6-aminolevulenic acid. Lead also inhibits coproporphyrin utilization and the
2-13
-------
conversion of zinc erythrocyte protoporphyrin (ZPP) into heme. The effects of lead on heme
biosynthesis are described in the Air Quality Criteria for Lead (USEPA, 1986).
Death: It is well known that severe lead poisoning can lead to encephalopathy and death.
There is some evidence, too, of higher death rates due to cerebrovascular disease among lead
workers (Malcolm and Barnett, 1982; Fanning, 1988; Michaels et al., 1991). In infants, high lead
levels have been suggested to cause Sudden Infant Death Syndrome (SUDS) (Drasch et al., 1988).
Hypertension: There may be a relationship between lead exposure and hypertension.
Increased heart rate and hypertension were observed in occupationally exposed workers after
only four weeks of exposure to high levels of lead (Marino, et al., 1989). Hypertension has also
been associated with lead exposure in the general population (Khera et al., 1980; Pirkle et al.,
1985; Harlan, 1988; Harlan et al., 1988), although the evidence is mixed (Pocock, et al., 1984,
1985,1988; Gartside, 1988; Coate and Fowles, 1989).
Gastrointestinal Effects: Colic is a consistent early symptom of lead poisoning. Colic is
characterized by the following symptoms: abdominal pain, constipation, cramps, nausea,
vomiting, anorexia, and weight loss. Although these symptoms typically occur at blood-lead
levels of 100 to 200 ng/dL, they have sometimes been noted hi workers whose blood-lead levels
were as low as 40 to 60 ng/dL (Table B-l).
Renal Effects: Both acute and chronic nephropathy (kidney disease) are known to be
caused by elevated lead exposure. The symptoms of acute nephropathy appear to be reversible.
The symptoms of chronic nephropathy, on the other hand, are irreversible. Acute nephropathy
has been reported hi children and lead workers, while chronic nephropathy is usually reported
only in lead workers. A summary of studies reporting acute or chronic nephropathy may be
found in ATSDR, 1993. Additional detail is reported in USEPA, 1986.
Vitamin D Metabolism: Lead may interfere with the conversion of vitamin D to its
hormonal form, 1,25-dihydroxyvitamin D. This effect is most apparent hi studies of children
with high lead exposure (Rosen et al., 1980; Mahaffey et al., 1982). No effect of lead on vitamin
D metabolism was observed in a study of children who received adequate amounts of calcium,
phosphorus, and vitamin D hi their diet and had low to moderate lead exposure. The average
lifetime blood-lead levels for these children ranged from 4.9 ng/dL to 23.6 [ig/dL (Koo et al.,
1991).
Thyroid: There is some evidence that lead may adversely affect thyroid function in
occupationally exposed workers (Tuppurainen et al., 1988). However, no effects of lead on
thyroid function have been reported in children (Siegel et al., 1989).
Growth: A number of epidemiological studies have reported an association between
blood-lead levels and growth in children (Nye, 1929; Johnson and Tenuta, 1979; Lauwers et al.,
2-14
-------
1986; Schwartz et al., 1986; Lyngbye et al., 1987; Angle and Kuntzelman, 1989). However, a
study of lead-poisoned subjects and nonexposed sibling controls failed to establish an association
between blood-lead levels and growth or the genetic predisposition for adult height (Sachs and
Moel, 1989). Moreover, a recent longitudinal study in Cleveland found no statistically
significant relationship between blood-lead levels and growth (height, weight, and head
circumference) from birth through age 4 years and 10 months (Greene and Emhart, 1991).
However, a separate analysis of 260 infants from this study found that growth rates, measured as
covariate-adjusted increases in stature from 3 and 15 months of age, were inversely correlated
with corresponding increases in blood-lead levels, although the observed relationship was
statistically significant only for infants exposed to higher prenatal blood-lead levels (maternal
blood-lead concentration >7.7 ug/dL) (Shukla et al., 1987,1989).
Development: Lead-related effects on children's development, such as reduced birth
weight, reduced gestational age, and neurobehavioral developmental deficits, have been reported.
The evidence for effects on birth weight and gestational age is mixed, with some studies
reporting reductions associated with lead exposure (Moore et al., 1982; McMichael et al., 1986),
while others report no differences (Needleman et al., 1984; Factor-Litvak et al., 1991; Green and
Ernhart, 1991). The evidence on neurobehavioral development is more consistent, with most
studies reporting an association between lead exposure and developmental deficits (Bellinger et
al., 1985a, 1985b, 1986a, 1986b, 1987a, 1989a, 1989b; Vimpani, et al., 1985,1989; Dietrich et
al., 1986,1987a, 1987b; Baghurst, et al., 1987; Wigg et al., 1988). A short summary of these
results is included hi Section 2.3.1. There is some evidence that early developmental deficits
related to prenatal lead exposure may not persist until age 4-5 years (Bellinger et al., 1991).
Immune System: The data on immunological effects of lead hi occupationally exposed
adults are inconsistent, but indicate that, while lead may have an effect on the cellular component
of the immune system, the humoral component is relatively unaffected (ATSDR, 1993). The
data on immunological effects of lead on children are very limited, but no effects have been
detected (Reigart and Graber, 1976; ATSDR, 1993).
Reproduction: High levels of lead have been shown to cause adverse effects on
reproduction hi both men and women. Women who are exposed to high levels of lead during
pregnancy have experienced an increased rate of miscarriages and stillbirths (Nordstrom, et al.,
1979; McMichael et al., 1986; Baghurst et al., 1987). In addition, women who were significantly
exposed during childhood may be at increased risk of spontaneous abortion and stillbirth and
their children more likely to experience learning disabilities (Hu et al., 1991). Effects of lead on
male reproductive functions, including reduced sperm production, have been reported hi studies
of occupationally exposed males (Lancrajan et al., 1975; Wildt et al., 1983; Chowdhury, et al.,
1986; Assennato et al., 1987). Reproductive effects of chronic low-level exposure are less
known. A recent prospective study found no effect on the rate of spontaneous abortions among
women who resided near a lead smelter compared to a control population of women who lived
25 miles away (Murphey et al., 1990).
2-15
-------
Genotoxic Effects: While the available evidence is contradictory, there is some evidence
to suggest that lead may have an effect on chromosomes. While increased frequencies of
chromosomal aberrations have been observed in occupationally-exposed workers, (Nordenson
et al., 1978; Huang et al., 1988), most studies report no such increase in workers (Schmid et al.,
1972; O'Riordan and Evans, 1974; Bauchinger et al., 1977; Maki-Paakkanen et al., 1981), or in
children (Bauchinger et al., 1977). Sister chromatid exchanges may (Grandjean et al., 1983;
Leal-Garza et al., 1986; Huang et al., 1988), or may not (Maki-Paakkanen et al., 1981; Dalpra
et al., 1983) be increased as a result of lead exposure. Concurrent exposure to other toxic
substances is a common problem in occupational exposure studies. Selection criteria employed
by Huang et al. were designed to minimize the effects of potential genotoxic factors other than
lead.
Cancer: Occupational exposure to lead has been associated with increased cancer risk.
Lead has been classified as a probable human carcinogen (Class B2) by EPA and a possible
human carcinogen (Group 2B) by the International Agency for Research on Cancer, based on
sufficient evidence of carcinogenicity in animals but inadequate evidence in humans (IARC,
1987; IRIS, 1993; EPA, 1989b). Increased risks of kidney cancer (Selevan et al., 1985;
Steenland et al., 1992; Cocco et al., 1997), lung cancer (Cooper et al., 1985; Gerhardsson et al.,
1986; Anttila et al., 1995; Lundstrom et al., 1997), glioma (Anttila et al., 1996), rectal cancer
(Fayerweather et al., 1997), and total malignant neoplasms (Cooper and Gaffey, 1975; Cooper,
1976,1981; Kang et al., 1980; Cooper et al., 1985; Anttila et al., 1995; Gerhardsson et al., 1995;
Lundstrom et al., 1997) have been observed in occupationally exposed workers. However, the
actual compounds of lead, routes of exposure, and levels of lead that may cause cancer in humans
are unknown. Furthermore, the potential for exposure to other carcinogens exists, particularly in
lead smelters.
2.4 REPRESENTATIVE POPULATION
As described in the previous section, adverse health effects of lead exposure have been
documented in people of all ages. Although occupational exposure to lead presents a serious
hazard and the subsequent health effects are well-documented, infants and young children are
more at risk from lead exposure than adults (USEPA, 1986; ATSDR, 1993). This intensified risk
is due to children's increased oral activity (e.g., hand-to-mouth behavior) and ability to absorb
lead, coupled with the susceptibility of their rapidly developing central nervous systems (Goyer,
1993; Bellinger, 1995).
The §403 regulations are intended to reduce the risk of childhood lead exposure through
the reduction of residential lead levels. Estimation of the benefits of reducing lead exposure in
children requires selection of an age group for characterizing the health risks of lead exposure.
The health benefits of reducing lead exposure are estimated for children aged 1-2 years (12-35
months) in this risk assessment. The selection of this age group was based on the most
appropriate age of child for the estimation of health effects, as described below.
The §403 standards are intended to protect all children, not just those aged 1-2 years.
However, it is assumed that the number of children outside this age range with elevated blood-
2-16
-------
lead concentrations, who did not have such elevations at this age, is relatively small, and so the
selection of children aged 1-2 years does not result in a gross underestimate of the health benefits
of reducing lead exposure. To assess the impact of using this age group for the characterization
of risk, an alternative age group of children aged 1-5 years is considered hi the sensitivity
analysis (Chapter 5).
Many lead exposure studies have been conducted on children aged 1-2 years, because
blood-lead concentrations tend to peak hi this period, hand-to-mouth activity is greatest, the
level of cognitive ability is sufficiently developed for testing, and children are more cooperative
for assessment, hi addition, several mechanisms for lead's effect on the developing brain were
identified in Section 2.2.2. These mechanisms provide a neurological basis for increased risk to
fetuses, infants, and young children exposed to lead. One mechanism, the disruption, or delay, of
synaptic development, suggests special concern for children aged 1-2 years. The synaptic density
of the frontal cortex of the brain peaks at age 2. Developmental disruptions at this critical time
could lead to permanent functional impairment in the brain. The effects of lead on the
developing brain may be estimated through IQ test scores later in life. Strong relationships
between blood-lead concentration hi early childhood and IQ scores have been reported in four
major longitudinal studies conducted hi Boston, Cincinnati, Cleveland, and Port Pine, Australia.
Lower IQ scores at school-age are reported for children who had earlier exhibited elevated blood-
lead levels.
In Boston, slightly elevated blood-lead levels at age 24 months (mean 6.5 //g/dL) were
associated with intellectual and academic performance deficits at age 10 years (Bellinger, 1992).
In fact, the correlation of IQ deficits at age 10 years was greatest with blood-lead concentration
measured at age 24 months. This is particularly significant, as IQ measures tend to be relatively
stable after age 10. Thus, hi addition to age 2 years being an important developmental period,
the long-term effects from lead exposure at this age appear to be particularly important.
hi Cincinnati, postnatal blood-lead levels measured through age 3 years were inversely
associated with IQ scores measured at age 5, although the effect was not statistically significant
when adjusted for covariates (Dietrich et al., 1993). hi Cleveland, a significant association was
reported between blood-lead concentration at age 2 (mean 16.7 //g/dL) and IQ measured at 5
years (Emhart et al., 1989). In Port Pirie, statistically significant associations were reported
between IQ measured at age 7 and blood-lead levels from birth through age 7, with the strongest
associations for blood-lead levels measured at 15 months to 4 years (Baghurst et al., 1992).
Three of these studies are included in the meta-analysis of Schwartz (1994), which is used
in this risk assessment to quantify IQ point decrements resulting from lead exposure. For the
longitudinal studies, Schwartz selected blood-lead concentration measures prior to age 3, because
basic cognitive abilities develop hi that period. Cross-sectional studies were also included in the
meta-analysis. For those studies, the blood-lead concentration and IQ scores were measured at
the same time. In a separate analysis, Schwartz concluded that the results reported by
longitudinal and cross-sectional studies were similar, so that estimates from the various study
designs could be combined.
2-17
-------
A second meta-analysis considered blood-lead concentrations from five longitudinal
studies, including the four described above (Pocock et al., 1994). Three measurements of blood-
lead concentration, at birth, at approximately age 2, and the mean of all post-natal measurements,
were related to IQ scores. This analysis determined that there was a strong relationship between
IQ scores and blood-lead concentration measured at age 2, but not with the other measurements.
Thus, the blood-lead concentration at age 2 may be used to predict the effect of lead on IQ scores
later in life.
2.5 SELECTED HEALTH ENDPOIMTS
The childhood lead poisoning problem encompasses a wide range of exposure levels,
with varying health effects at different levels of exposure. As described in Section 2.3, even low-
level exposure to lead can result in adverse health effects. At low levels, the health effects may
not be severe or obvious, but a large number of children are affected. As the exposure level
increases, the severity of the health effects increases, but the number of affected children
decreases.
Both individuals and society as a whole are damaged by adverse health effects associated
with lead exposure. In this section, several elevated blood-lead concentration and health effect
endpoints are identified. These endpoints are used in the risk characterization in Chapter 5, and
the risk management analysis hi Chapter 6, to estimate the numbers of children who may benefit
under the proposed §403 rule. Each endpoint may be used both to estimate the number of
children who will benefit under the proposed rule and also the economic benefit to society.
While the health effect endpoints were selected because they are indicative of health effects from
low exposures and the elevated blood-lead concentration endpoints are based on CDC's
guidelines for lead levels that cause effects, the ability to quantitatively measure health risks was
considered in selecting the health endpoints, as well. Economic benefits resulting from the rule
will be estimated in the §403 RIA.
2.5.1 Elevated Blood-Lead Concentration
Although an elevated blood-lead concentration is not a health effect in and of itself, the
relationship between blood-lead concentration and a range of adverse health effects is well-
established. In addition, CDC guidelines on childhood lead poisoning prevention traditionally
have been and currently are defined hi terms of blood-lead concentrations. Table 2-1
summarizes CDC's recommended actions for children with elevated blood-lead concentrations
(CDC, 1991). These actions include 1) more frequent rescreening, 2) parental education on
reducing lead exposure, 3) nutritional counseling, 4) environmental assessment and remediation,
5) medical evaluation, and 6) chelation therapy. The extent and expense of the recommended
interventions increases with blood-lead concentration. The classes defined hi Table 2-1 were
used to select the following two levels of elevated blood-lead concentration for which this risk
assessment estimates the number and percentage of children exceeding these levels:
2-18
-------
Table 2-1. Interpretation of Blood-Lead Concentration Categories and Follow-Up Actions
Recommended by CDC.
Class
I
IIA
IIB
III
IV
V
Blood-Lead
Concentration
U/g/dL)
s 9
10- 14
15- 19
20-44
45- 69
* 70
Recommended Action
A child in Class I is not considered to be lead-poisoned.
Many children (or a large proportion of children) with blood-lead levels in
this range should trigger communitywide childhood lead poisoning
prevention activities. Children in this range may need to be rescreened
more frequently.
A child in Class IIB should receive nutritional and educational
interventions and more frequent screening. If the blood-lead level
persists in this range, environmental investigation and intervention
should be done.
A child in Class III should receive environmental evaluation and
remediation and a medical examination. Such a child may need
pharmacologic treatment of lead poisoning.
A child in Class IV will need both medical and environmental
interventions, including chelation therapy.
A child with Class V lead poisoning is a medical emergency. Medical
and environmental management must begin immediately.
Source: Preventing Lead Poisoning in Young Children (CDC, 1991).
Incidence of blood-lead levels greater than or equal to 10 //g/dL: Adverse health
effects have been documented at blood-lead concentrations as low as 10 //g/dL (USEPA, 1986,
1990a; ATSDR, 1993). This level is the lowest blood-lead level that is considered elevated by
CDC. While extensive interventions are not always recommended, children with blood-lead
concentrations at or above 10 /ug/dL require more frequent rescreening at minimum, and may
require environmental or medical interventions. In addition, if many children in a community
have blood-lead concentrations above 10 ug/dL, community-wide intervention activities are
recommended (CDC, 1991).
Incidence of blood-lead levels greater than or equal to 20 //g/dL: Medical and
environmental interventions are recommended for all children with blood-lead concentrations at
or above 20 //g/dL.
2.5.2 IQ Point Deficits
In this section, IQ based health endpoints are identified to represent the
neurotoxicological effects of lead exposure. While tests that focus on a specific neurological
effect might be more sensitive to the effects of lead than IQ tests, the selection of a representative
effect is difficult. Differences in the level, tuning, and route of exposure for individuals may
result hi differing effects of lead. For example, early exposure to lead (before age 2) may affect
language skills, while later exposure is more likely to affect spatial-symbolic skills (Shaheen,
1984). In the absence of details of the exposure scenario, which are rarely available, exposure-
2-19
-------
related differences will be most apparent on tests, such as IQ tests, that measure performance
over a range of neurological functions (Bellinger, 1995). The relationship between blood-lead
concentration and IQ scores has been reported consistently in the literature and efforts have been
made to quantify this relationship by meta-analysis (Schwartz, 1993; Pocock et al., 1994;
Schwartz, 1994; Section 4.4; Appendix D). The following IQ-based health endpoints are used in
the risk assessment to represent the neurotoxicological effects of lead exposure:
IQ Points Lost: This effect .is used to represent the neurological loss due to low level
lead exposure. Lower IQ scores are associated with a lower level of educational attainment and
lower life-time earnings. The average IQ point loss in a child resulting from lead exposure is
estimated, along with incidences of IQ point loss £ 1, ^2, and ^3 points. These levels were
selected arbitrarily for presentation purposes. These small effects are not meaningful for
individual children, as the standard deviation associated with IQ tests is usually 5 points.
However, these effects may be estimated for the population of children and provide a useful
illustration for this risk analysis.
Increased Incidence of IQ scores less than 70: This effect measures the increased
likelihood of mental retardation resulting from lead exposure. An IQ of 70 is two standard
deviations below the population mean score of 100 and is used as an indicator of mental
retardation. Children who are mildly mentally retarded require special education classes hi
school. Children who are severely mentally retarded may require life-long institutional care.
2.6 HAZARD CHARACTERIZATION
Though lead causes a wide array of adverse health effects, particularly at high dose levels,
lead is most known for its adverse effects on the central nervous system. Young children are
most susceptible to adverse health effects associated with lead exposure due to their developing
central nervous systems and their increased ability to absorb lead. Long-lasting impacts on
intelligence, motor control, hearing, and neurobehavioral development of children have been
documented at levels of lead hi the body that are not associated with noticeable symptoms and
were once thought to be safe. There is no apparent threshold hi the level of lead associated with
some of these subtle neurological effects. At higher levels, lead affects the hematological,
gastrointestinal, renal, and reproductive systems. Severe cases of lead poisoning may result in
delirium, convulsions, paralysis, coma, and death. Although the evidence is less conclusive, lead
may also affect the immune system, thyroid function, growth and development in children, and
vitamin D metabolism. Lead has been associated with hypertension, chromosomal aberrations,
cancer, and increased risk of death due to cerebrovascular disease. The documented evidence on
the adverse biological responses to lead is one of the major strengths of this risk analysis.
Typically, hi studies assessing adverse health effects associated with lead exposure,
relationships between health effects and exposure are established using a measure of internal
rather than external exposure. A variety of direct measures (lead in blood, bones, teeth, and hair)
and indirect measures (hemoglobin and EP levels) of lead exposure were identified hi this
concentration is the most readily available and widely accepted measure of internal exposure.
2-20
-------
Thus blood-lead concentration was selected as the measure of body lead burden to quantify
environmental lead exposure.
To provide an endpoint that represents neurological effects and because the neurological
effects of lead are well-documented and generally accepted bv the scientific community, two
types of IQ-related health endpoints were identified: IP score decrements (average, z I. z2, and
^3 IQ points) and the increased incidence of IP scores less than 70 due to lead exposure. In
addition, incidences of elevated blood-lead concentrations above specified thresholds (10 and 20
were selected as surroates for the wide array of non-IO related health risks to both the
central nervous system and other organs. The relationship between elevated blood-lead levels
and adverse health effects is well-established. The blood-lead concentration thresholds selected
for this risk analysis were among those established by CDC as levels of concern and are generally
recognized by the scientific community. The neurotoxic and blood lead endpoints selected for
this risk analysis have been used to support previous regulatory decisions. It is likely that if
additional endpoints were included, the baseline risks to lead exposure would be larger and the
potential risk reduction might be larger.
A potential weakness of this risk analysis lies in the selection of the age group for which
health benefits are estimated. Children aged 1-2 years are targeted for estimation of health risks
in this risk analysis for two reasons: 1) increased vulnerability of 1-2 year olds due to their
rapidly developing central nervous system, and 2) both the normal hand-to-mouth activities of
this age group and the pica tendencies observed in some children may put children aged 1-2 years
most at risk to lead exposure. While older children may also experience adverse health effects, it
is assumed that few such children would not have been previously exposed to lead at age 1-2
years. If this is not the case, then the estimated health risks for children aged 1-2 years may
underestimate the risks for all young children. It is also assumed that children who experience an
acute increase in lead exposure while aged 1-2 .years suffer the same health consequences as
those whose exposure duration is longer. If this is not the case, then the health risks to this group
may be overestimated. To assess the impact on selecting children aged 1-2 years versus a larger
population, an alternative age group of children aged 1-5 years is considered in the sensitivity
analysis for the risk characterization (Chapter 5).
2-21
-------
This page intentionally blank.
-------
3.0 EXPOSURE ASSESSMENT
CHAPTER 3 SUMMARY
The goals of this exposure assessment are to document the important
sources of lead in the environment, to document the major pathways by which
children are exposed to lead, to characterize the current distribution of
environmental-lead levels in the nation's housing stock, and to characterize the
current distribution of average blood-lead concentration among the nation's
children. Information from the exposure assessment is used with the findings of
hazard identification (Chapter 2) and dose response assessment (Chapter 4) to
provide input to the risk characterization (Chapter 5) and analysis of example
options for risk management (Chapter 6).
Residential paint, dust, and soil are among those sources of lead which
contribute most significantly to overall lead exposure in humans. Several lead
exposure studies have concluded that the pathway of lead-contaminated soil and
dust to children's blood is an important means by which young children are exposed
to lead from lead-based paint hazards. Studies such as the Baltimore Repair and
Maintenance Study and the Rochester Lead-in-Dust Study conclude that elevated
lead levels in paint, dust, and soil continue to exist in residential environments,
particularly in older homes. Even at low to moderate levels, lead in residential dust
can affect children's blood-lead concentration.
The HUD National Survey of Lead-Based Paint in Housing was selected as
the data source for characterizing environmental-lead levels in the nation's housing
stock in this risk assessment. According to this survey, 83% of occupied units
built prior to 1980 are expected to contain lead-based paint, and 18% are expected
to contain more than five square feet of deteriorated lead-based paint. Dust- and
soil-lead concentrations and dust-lead loadings tend to increase with age of unit.
The baseline distribution of blood-lead concentration within children aged 1
to 2 years is derived in this risk assessment from data collected in Phase 2 of the
Third National Health and Nutrition Examination Survey (NHANES III). The
geometric mean blood-lead concentration for these children is 3.1 ug/dL, with a
95% confidence interval of 2.8-3.5 ug/dL. Approximately 6% of these children are
estimated to have blood-lead concentrations greater than or equal to 10 ug/dL.
Figure 3-1 outlines the approach for the exposure assessment. Conclusions
from the exposure assessment are presented in Section 3.5.
3-1
-------
Background
and
Objectives
Hazard
Identification
r
EXPOSURE
ASSESSMENT
*-
Identify Important Sources
and Pathways of Lead to Blood
(Section 3.1)
\
Identify Supporting Evidence of
Health Risks from Lead Exposure
(Section 3.2)
\
Identify Sources of Data on
Environmental-Lead and
Blood-Lead (Sections 3.3, 3.4)
i '
4l 4 4 4 4
(A A r A AmericanA f A HUD A A A BaltlmoreA f A RochetterA /" A HUD "A
WANES III W ) Housino ) ) National )( } MM }{ Vead-.n-Do.,) [ ) Grantee. ]
y Data y y^ ySurveyData/ ^ J Survey Data/ ^ y Study Datay \^ J Study Daiay ^ J Data y'
1 , ... ^
1 1
Characterize Baseline
Distribution of Blood-
Lead Concentration
in Children 1-2 Years
(Section 3.4.1)
1 »
D
Chil
S[
E
(Sec
1
L ^ ±
4
Characterize Provide Supporting
Environmental Information on Environmental
Lead Levels in Lead and Blood-Lead Levels
U.S. Housing (Sections 3.3.1,
(Section 3. 3.1) 3.4.2,3.4.3)
4
etermine # of
dren Exposed to
>ecific Sets of
nvironmental
Lead Levels
tions 3.3.2, 3.5)
J
i
r
-
Dose-
4--^ Response
Assessment
^
r
Risk
Characterization
Risk
Management
Conclusions
on Risk
Characterization
Conclusions on
Analysis of Example
Options for §403
Standards
Figure 3-1. Detailed Flowchart of the Approach to Exposure Assessment.
3-2
-------
This exposure assessment seeks to answer the following questions:
1. Do the predominant sources of residential lead exposure to children include lead-
based paint hazards, including lead-contaminated dust and lead-contaminated soil?
2. What is the distribution of environmental-lead levels in dust and soil in the nation's
housing stock?
3. Is there evidence of a relationship between lead-based paint exposures in residential
environments and children's blood-lead concentration?
4. What is the baseline distribution of blood-lead concentration in the representative
population (children aged 1-2 years)?
Information from the exposure assessment is used with the findings of hazard identification
(Chapter 2) and dose response assessment (Chapter 4) to provide input to the risk
characterization (Chapter 5) and risk management (Chapter 6).
Figure 3-1 presents the overall risk analysis approach, with the approach for the exposure
assessment detailed. This chapter is formatted similar to the outline in Figure 3-1. Section 3.1
provides documented sources and pathways of lead exposure hi the nation's residential
environment. A number of lead exposure studies have investigated the extent to which lead is
present in certain residential environments and how this lead exposure is reflected hi blood-lead
concentration in children. These studies are introduced and summarized hi Section 3.2. Section
3.3 characterizes lead exposure hi that portion of the national housing stock in which children
reside or can potentially reside (hereafter referred to simply as the "national housing stock").
Section 3.4 characterizes the distribution of childhood blood-lead concentrations in children aged
1-2 years. The characterizations in Sections 3.3 and 3.4 are provided for 1997, prior to when
regulations developed in response to §403 are expected to be proposed.
Answers to the above questions are summarized hi Section 3.5, along with key results to
be used in the risk characterization and any limitations associated with the data sources or
approaches used to obtain these results.
This chapter identifies the best sources of available data on housing stock characteristics,
population estimates, and environmental-lead levels hi housing units, hi order to make inferences
on residential lead exposure to children in the United States. The extent to which any exposure
assessment accurately portrays the exposure scenario of interest depends on the relevance and
representativeness of the data used hi the analyses and in the methods applied to these data to
meet the objectives of the exposure assessment. Therefore, an effort has been made in this
chapter to present the methods used, to identify assumptions and approximations made in the
analysis and when they were made, and to identify uncertainties and limitations hi the data.
Supporting information and detailed results to accompany the information in this chapter are
presented in Appendices Cl and C2.
3-3
-------
3.1 SOURCES AND PATHWAYS OF LEAD
Lead is a heavy, stable element occurring naturally in the earth's crust. Through natural
activity such as crustal weathering and human activity such as mining, this metal has been
distributed throughout the human environment. Lead's historic use as a raw material in various
manufactured and refined products has increased its introduction into the environment. As a
result, lead has been detected in water, soil, air, plants, animals, and humans. As lead does not
naturally biodegrade, its exposure potential tends to accumulate over time as more and more lead
is deposited in the environment.
Research has identified a variety of environmental sources and reservoirs of lead which
can contribute to overall lead exposure hi a child. Figure 3-2 illustrates the major sources and
reservoirs of lead, how lead is introduced into the human environment, and various pathways of
human exposure. According to this figure, both natural sources (e.g., crustal weathering) and
sources resulting from human activity (e.g., auto and industrial emissions, paint and industrial
dusts, solder, lead glazes) have contributed lead to various components of the human
environment. Lead in such media as inhaled air, dusts, food, or drinking water contributes to
human lead exposure via direct pathways between these reservoirs and man. As data supporting
the dangers of lead exposure have been identified, a combination of state and Federal action has
curtailed the impact of certain sources and reservoirs of lead hi the environment, resulting hi a
change in the predominance of historically significant sources.
hi the scientific literature (e.g., Bornschein et al., 1986), quantitative exposure models, or
pathways models, have been applied to data from environmental-lead studies to identify the most
significant pathways by which residential, childhood environmental-lead exposure occurs and to
provide quantitative estimates of the relative contributions of the numerous hypothesized
exposure paths. One such set of environmental pathways, as reported by Bornschein et al., 1986,
upon analysis of data from the Cincinnati Longitudinal Study (Section 3.2.2.5), is shown in
Figure 3-3.
The information that follows provides the current status of the sources of lead included hi
Figure 3-2 that have historically been recognized in the scientific literature as most associated
with elevated blood-lead concentrations hi children. Most of the information comes from
detailed investigations on sources of lead documented in USEPA (1986), CDC (1991), and
ATSDR(1993).
Airborne Lead
Historically, emissions from lead smelters, battery manufacturing plants, solid waste
incinerators, and automobiles have made major contributions to airborne lead levels. Fallout of
atmospheric lead contributes to lead levels hi soil, household dust, and street dust. Lead is
deposited on soil, plants, and animals, which thereby is incorporated into the food chain.
3-4
-------
SURFACE
AND GROUND
WATER
SOLDER,
LEAD
GLAZES
DRINKING
WATER
FECES
URINE
Figure 3-2. Pathways of Lead from the Environment to Humans, Main Organs of Absorption
and Retention, and Main Routes of Excretion.
(Sources: USEPA, 1986; USEPA, 1996a)
3-5
-------
Lead-Based
Paint
Soil
1
*-
1
id in
Lead in
Dust
^
Lead on
Children's
Hands
^
Lead in
Children's
Blood
t
\,
Figure 3-3. Set of Environmental Pathways Reported by Bornschein et al., 1986,
Upon Analysis of Data from the Cincinnati Longitudinal Study (Children
Aged 18 Months).
Until recently, leaded gasoline emissions was one of the primary sources of lead exposure
in the United States. Under Title H of the 1990 amendments to the Clean Air Act (42 U.S.C.
7545), EPA specified lead as a pollutant compound of concern and instituted a controlled phase-
out of leaded gasoline by December 31,1995 (§21 l(n) of Title II). As a result, there was a 73%
reduction in lead consumed in gasoline from 1975 to 1984 (USEPA, 1986) and a 64% reduction
in national lead emissions from 1985 to 1989 (ATSDR, 1993). This reduction has corresponded
to a similarly dramatic decrease in average lead concentration in children's blood (CDC, 1991;
Annest, 1983). The phase-out of leaded gasoline has contributed to airborne lead's becoming
only a minor lead-exposure pathway for children not exposed to specific point-emitting lead
sources (CDC, 1991).
Even in the absence of a point-emitting lead source, indoor air may be considered an
important indirect lead-exposure pathway when lead-based paint or lead-contaminated dust or
soil is disturbed during renovation and remodeling activities. Inadequate dust control or use of
paint stripping techniques that vaporize lead hi paint are ways that lead contaminates breathable
air during renovation and remodeling activities (USEPA, 1994b).
EPA has set a National Ambient Air Quality Standard of 1.5 ug of lead per cubic meter of
air (40 CFR 50.12). This standard is compared to the average of 15-16 air samples, each taken
for a 24-hour duration over a period of three months, to determine air quality compliance.
Drinking and Cooking Water
Detectable levels of lead are rare in surface and ground water that serve as sources of
drinking water hi this country. Typically, lead contamination of drinking water occurs after the
water leaves the treatment plant (CDC, 1991). By traveling within service lines and household
plumbing, drinking water can become contaminated upon encounter with lead pipes, connectors,
and solder. At a residence, water can also become contaminated by the lead or brass components
3-6
-------
of water fountains, coolers, faucets, and other fixtures. Through the authority of the 1986 Safe
Drinking Water Act and its amendments, EPA banned the use of lead materials and solders in
new plumbing and plumbing repairs, required that public water suppliers notify the public about
lead presence in drinking water, and encouraged local government measures to test and remediate
lead-contaminated drinking water in schools and day-care centers. As a result, drinking and
cooking water from municipal and other large drinking water distribution systems is generally
not a predominant source of lead exposure among lead-poisoned children (CDC, 1991).
Analysis of environmental-lead data from several studies, including the Baltimore R&M
Study and the Rochester Lead-in-Dust Study (Section 3.2), concluded that lead levels in drinking
water generally do not have a statistically significant effect on blood-lead concentrations. In both
of these studies, however, lead levels in water were low. However, due to the high absorption
rate of lead in water, lead in drinking water is still considered an important exposure source when
present (CDC, 1991).
As required by the Safe Drinking Water Act, the National Primary Drinking Water
Regulations (NPDWRs) for Lead and Copper (56 FR 26460, June 7,1991) set an action level for
lead in drinking water of 15 ppb and specified a maximum percentage of homes in a water
service area that could exceed this action level. Those systems that do not meet these standards
must inform the public, while taking measures to reduce lead levels and continue monitoring
procedures. The NPDWRs for Lead and Copper also set maximum contaminant level goals
(MCLGs). This rule set the MCLG for lead within drinking water at the tap to be 0 ppb (40 CFR
141,142).
Food
Many studies have shown that children's dietary intake of lead has receded over recent
years. For example, data from the U.S. Food and Drug Administration (FDA) indicate that
dietary lead intake in two-year-old children has declined from an approximate average of
30 ng/day in 1982 to 5 ng/day in the period 1986-1988 (CDC, 1991). U.S. FDA intervention and
outreach activities, along with reduced lead entering the food chain due to the phase-out of
leaded gasoline, have contributed to this decline. The phase-out of lead-soldered food cans
(1.4% of the U.S.-produced food and soft drink cans in 1989, compared to 47% of such cans
produced in 1980), along with public education on proper food storage and cooking techniques,
have made large contributions to reducing the amount of lead ingested with food (CDC, 1991).
Education is especially important in those areas of the country with traditions of using lead-
containing pottery in cooking and preparing folk remedies containing lead.
While production of lead-soldered food and soft drink cans have been virtually eliminated
hi the U.S., such cans may still be used by other countries who export food to the U.S. In
addition, lead can be introduced to food grown in lead-contaminated soil. Improper handling of
food in the home (e.g., storing food in containers such as lead-soldered cans and lead-glazed
pottery) can cause food to be a source of lead exposure. Thus, while lead exposures through food
ingestion have declined considerably hi recent years, these exposures can still occur if proper
precautions are not addressed.
3-7
-------
Lead-Based Paint
Lead-based paint (LBP) is currently considered the most significant high-dose source of
lead exposure in pre-school children (CDC, 1991). (Other sources such as lead plumbing and
historic reservoirs of lead deposited in soil and in house dust remain important for a significant
minority, especially in some non-urban areas.) From the turn of the century through the 1940's,
paint manufacturers used lead as a primary ingredient in many oil-based interior and exterior
house paints. Usage gradually decreased through the 1950s and 1960s, as largely lead-free latex
paints and exterior paint with lower lead concentrations were manufactured. In 1978, the
Consumer Product Safety Commission (CPSC) ruled that paint used for residence, toys,
furniture, and public areas must not contain more than 0.06% lead by weight. Nevertheless, the
presence of lead-based paint in the nation's housing stock remains high. An estimated 64 million
(or 83% of) privately-owned, occupied housing units built prior to 1980 contain some
components covered with lead-based paint (USEPA, 1995a), defined as containing at least
1.0 mg lead per square centimeter of painted surface. Approximately 12 million of these units
contain at least one child under the age of seven years. The estimated percentage of public
housing units in this category is even higher: 86% (USEPA, 1995a).
Human exposure to lead from lead-based paint is believed to be higher when the paint is
in a deteriorated state or is found on accessible, chewable, impact, or friction surfaces (USEPA,
1986; CDC, 1991). Thus, young children are especially susceptible to lead poisoning from lead-
based paint, as they may ingest lead-based paint chips or come into contact with dust or soil that
has been contaminated by deteriorated lead-based paint (see below). Both adults and children
can be exposed to hazardous levels of lead by ingesting paint-dust during hand-to-mouth
activities. The U.S. Department of Housing and Urban Development (HUD) has prepared
guidelines on controlling lead-based paint hazards, as improper control procedures can actually
increase the threat of lead-based paint exposure by dispersing fine lead dust particles in the air
and over accessible household surfaces (USHUD, 1995b; Farfel and Chisolm, 1990). The
potential for lead-based paint to contaminate a variety of environmental media within a
household makes lead-based paint the greatest source of public health concern regarding lead
exposure (CDC, 1991).
Contaminated Dust and Soil
While enforcement of national air quality standards continues to reduce the threat of lead
exposure via air from point sources, the fallout of atmospheric lead over time has resulted in a
continued exposure route through soil (USEPA, 1986). In addition, soil can become
contaminated by deteriorated lead-based paint or by the improper removal of lead-based paint
from a housing unit. The same soil, once tracked indoors, can become a component of
household dust causing yet another source of lead exposure. Children are exposed to lead from
soil or dust in then* homes during typical hand-to-mouth activities.
Lead-contaminated soil and dust are thought to be the major pathway by which young
children are exposed to lead from lead-based paint hazards (USEPA, 1986). Exterior house paint
can flake off or leach into the soil around the outside of a home, contaminating children's play
3-8
-------
areas. Indoors, normal wear of lead-based paint (especially around windows and doors) and
contaminated soil tracked into the house can contaminate interior dust. Lead-based paint takes
the form of paint chips or tiny, exfoliated flakes when contaminating interior dust; the
contribution of these two forms to contaminated dust has not been studied in great detail. When
lead takes the form of small particles, as it typically does when found within household dust (Que
Hee et al., 1985), it is more easily absorbed into the body (Mahaffey, 1977).
A number of studies have assessed the effect of dust- and soil-lead levels on childhood
blood-lead concentrations. A few studies have concluded that the effect of residential lead-based
paint on blood-lead levels occurs via the pathway of dust- and soil-lead to blood. For example,
the pathways diagram in Figure 3-3 indicated that a significant lead pathway from exterior dust to
interior dust to hands to blood was identified through analysis of data from the Cincinnati
Longitudinal Study (Section 3.2.2.5), with lead in paint and soil contributing to lead in exterior
dust (Bornschein et al., 1986). Analysis of data from the Brigham and Women's Hospital
Longitudinal Study (Section 3.2.2.6) concluded that the pathway from soil to window sill-dust to
floor-dust to blood was statistically significant (Menton et al., 1995). It is likely that exposure of
young children to lead in dust and soil is primarily due to their propensity to mouth fingers, toys,
and other nonfood items that contain contaminated dust. Pathways analyses of data from such
studies as the Cincinnati Longitudinal Study (Bornschein et al., 1986) found a significant
pathway of lead from hand dust to blood, suggesting that hand-to-mouth activities are an
important contributor to childhood blood-lead concentrations.
3.2 SUPPORTING EVIDENCE IN LEAD EXPOSURE STUDIES
To determine the extent to which lead-based paint hazards are associated with elevated
blood-lead concentrations in children residing in the nation's housing stock, this exposure
assessment has documented evidence as reported within a vast library of government reports,
published articles, and proceedings. Section 3.2.1 identifies recent human characterization and
intervention studies that address the relationship between childhood blood-lead concentration and
environmental-lead levels in the nation's housing, along with their general findings. A selected
number of these studies provide the most useful information for this risk assessment; these
studies are presented in greater detail in Section 3.2.2.
3.2.1 Weight of Evidence on the Relationship between Environmental-Lead
Exposures and Increased Blood-Lead Concentrations
Extensive evidence of the relationship between childhood blood-lead concentrations and
environmental-lead levels is offered in the scientific literature. Evidence from two types of
studies is available. Human characterization studies investigate the association between elevated
blood-lead concentrations and elevated levels of lead in a child's residential environment.
Intervention studies investigate the impact on children's blood-lead concentrations of reducing
childhood lead exposure via a range of intervention strategies. Human characterization studies
have demonstrated that elevated blood-lead concentrations are associated with elevated lead
levels in the dust, paint, and soil of the surrounding environment. Intervention studies can
contribute to conclusions about causation. If children receiving an intervention strategy that
3-9
-------
targets a particular lead exposure source (e.g., paint, dust, or soil) exhibit greater reductions in
blood-lead concentrations than those reported for a suitable control population, then the targeted
source may be at least partially responsible for the prior exposure.
A review of intervention studies (USEPA, 1995b) concluded that reductions in blood-
lead concentrations have occurred following interventions aimed at lead in paint, dust, and soil.
While such studies suggest causation, their results are not necessarily indicative of the magnitude
of the association between the levels of lead in targeted environmental media and blood-lead
concentrations. This is because intervention studies typically examine children already exposed
to environmental lead. Exposed children retain a store of lead in their tissues that routinely
mobilizes into the blood (Gulson et al., 1995). This mobilization may be heightened following
an intervention (Schroeder and Tipton, 1968; Rabinowitz, 1991) as the change in exposure
caused by the intervention disrupts the body's equilibrium. Blood-lead concentrations following
the intervention, therefore, represent a combination of the now reduced environmental lead
exposure and the increasingly (at least temporarily) mobilized lead stores.
During the past 25 years, studies have been conducted to investigate the sources
responsible for lead exposure in children. These studies include investigations of the sources and
extent of lead exposure in both urban and ore-processing communities. The studies listed in
Tables 3-1 and 3-2 provide evidence regarding associations between childhood blood-lead
concentrations and environmental-lead levels in urban and ore-processing communities,
respectively. Many of these studies are limited, small, or not relevant to the current exposure
situation. However, the results of these studies are qualitatively similar in that the association1
between environmental-lead levels and blood-lead concentration is consistently positive and,
when considered without the confounding from additional variables, usually found to be
statistically significant given the collected data. When confounding variables (e.g., age, race or
ethnicity, socioeconomic status, housing condition) are included hi the analysis of data from
these studies, the estimated strength of the relationship between blood-lead and environmental-
lead levels is modified, and, sometimes, is no longer significant given the collected data. It is
difficult to combine results from multiple studies into one representative, quantitative measure of
the relationship between blood-lead concentration and environmental-lead levels, due to the
qualitative dissimilarities among the studies (e.g., differences in sampling and analysis methods,
sampling locations, target populations, and types of communities).
Early childhood lead exposure studies emphasized exposure to lead in paint, leaded
gasoline emissions, and emissions from industrial sources. These studies, therefore, measured
lead levels in these media and sought to relate them directly to resident children's blood-lead
concentrations. Due to the assessment by many researchers in childhood lead exposure that
ingestion of dust and soil via hand-to-mouth behavior represents the principal mechanism of lead
exposure in young children today (CDC, 1991), more recent studies have focused principally on
lead exposure from residential soil and dust. As indicated in Figure 3-2, residential soil and dust
are assumed to have been contaminated by these same original sources: lead-based paint,
industrial emissions or tailings, and leaded gasoline emissions.
3-10
-------
Table 3-1. Childhood Lead Exposure Studies Conducted in Urban Communities That
Present Evidence of the Relationship Between Environmental-Lead Levels and
Blood-Lead Concentrations.
Study/Community
Baltimore (MD) Lead-
Based Paint Abatement
and Repair and
Maintenance Study
Rochester (NY) Lead-in-
Dust Study
Evaluation of the HUD
Lead-Based Paint Hazard
Control Grant Program
Baltimore (MD) Urban
Soil Lead Abatement
Demonstration Project
(USLADP)
Boston (MA) USLADP
Cincinnati (OH) USLADP
Birmingham (UK) Urban
Lead Uptake Study
Cincinnati (OH)
Longitudinal Study
Brigham and Women's
Hospital Longitudinal
Study (Boston, MA)
New Haven, CT
Omaha, NE
Study
Duration
1992-1997
1993
1994-
present
1988-1991
1989-1991
1989-1991
1984-1985
1980-1987
1980-1983
1977
1970-1977
Study Type
Intervention
(Abatement Efficacy)
Human
Characterization
Intervention (cost and
effectiveness)
Intervention (Soil
Abatement Effioapvl
Human
Characterization
Human
Characterization
Human
Characterization
Human
Characterization
Human
Characterization
Referenced)
Farfel and Lim, 1 995
USEPA, 1996b
USHUD, 1995a;
Lanphear et al., 1995
Lanphear et al., 1996a
Lanphear et al., 1996b
Emond et al., 1997
NCLSH and UCDEH, 1997
NCLSH and UCDEH, 1994
USEPA, 1996a;
Weitzman et al., 1993;
Aschengrau et al., 1994
Davies et al., 1990;
Thornton et al., 1990;
Davies et al., 1987
Bornschein et al., 1985a; Que
Hee et al., 1985; Bornschein
et al., 1985b; Bornschein
et al., 1986
Bellinger et al., 1986b;
Rabinowitz et al., 1985a;
Rabinowitz et al., 1985b;
Rabinowitz et al., 1984a;
Rabinowitz et al., 1984b;
Rabinowitz et al., 1982
Stark et al., 1982;
Stark et al., 1978
Angle and Mclntire, 1979;
Angle et al., 1974;
Angle et al., 1984
3-11
-------
Table 3-2. Childhood Lead Exposure Studies Conducted in Ore-Processing Communities
That Present Evidence of the Relationship Between Environmental-Lead Levels
and Blood-Lead Concentrations.
Study/Community
Palmerton (PA) Lead Exposure
Study
Bingham Creek (UT)
Environmental and Human Health
Lead and Arsenic Study
Leadville/Lake County (CO)
Environmental Health Study
Granite City (IL) Educational
Intervention Study
Butte-Silver Bow (MT)
Environmental Health Study
Clear Creek/Central City (CO)
Mine Waste Exposure Study
Midvale (UT) Community Lead
Study
Child Lead Exposure Study
(Leeds, AL)
Philadelphia (PA) Neighborhood
Lead Study
Leadville (CO) Metals Exposure
Study
Silver Creek Mine Tailings
Exposure Study (Park City, UT)
Telluride, ID
Kellogg (ID) Revisited
Helena Valley (MT) Child Lead
Study
El Paso, TX
Study
Duration
1994
1993
1991
1991
1990
1990
1989
1989
1989
1988
1987
1986
1983
1983
1971-1973
Study
Type
Human
Characterization
Human
Characterization
Human
Characterization
Intervention
Human
Characterization
Human
Characterization
Human
Characterization
Human
Characterization
Human
Characterization
Human
Characterization
Human
Characterization
Human
Characterization
Human
Characterization
Human
Characterization
Human
Characterization
Reference(s)
Bornschein, 1996a
Bornschein, 1996b
Bornschein, 1997
Kimbrough et al., 1994
Butte-Silver Bow Dept. of
Health, et al., 1991
ATSDR, 1992
Bornschein et al., 1990
ATSDR, 1991 a
ATSDR, 1991b
Colorado Dept. Of Health,
et al., 1990
ATSDR, 1988b
Bornschein et al., 1989
Panhandle District Health
Dept. et al., 1986
Lewis and Clark County
Health Dept. et al., 1986
Landrigan et al., 1975
3-12
-------
Due to the reduction in lead sources such as gasoline emissions over time, the most recent
lead exposure studies provide a more accurate picture of the relationship between child blood-
lead concentrations and lead-based paint hazards. Additionally, while deteriorated lead-based
paint is a common lead source in older, ore-processing communities, studies in these
communities provide less evidence than general urban studies as to the association between
elevated blood-lead concentrations and lead-based paint hazards, due to the presence of hazards
from industrial sources. Consequently, this risk assessment has relied primarily on information
from recent studies conducted in urban areas in the absence of specific point emission sources.
While this approach may underestimate the exposure of some high-risk subpopulations heavily
exposed to area sources such as leaded gasoline depositions near highways, bridge and structure
painting and refinishing, and low-level non-ferrous metal processing operations (e.g., battery
recycling and radiator shops), it is consistent with the intent of Title X to reduce hazards
associated with lead-based paint.
3.2.2 More Detailed Description of the Most Useful Studies for
This Risk Assessment
Upon review of the design, analysis approach, and conclusions of the ten studies listed in
Table 3-1, nine studies were identified as most relevant to address the questions on childhood
lead exposure presented at the beginning of this exposure assessment chapter. These studies
were
the Baltimore Lead-Based Paint Repair and Maintenance (R&M) study (pre-
intervention phase);
the Rochester Lead-in-Dust Study;
Evaluation of the HUD Lead-Based Paint Control Grant Program (HUD Grantees);
the three studies constituting the Urban Soil Lead Abatement Demonstration Project
(USLADP);
the Birmingham Urban Lead Uptake study;
the Cincinnati Longitudinal study, and
the Brigham and Women's Hospital Longitudinal study.
Three of these nine studies, the Baltimore R&M Study (Section 3.2.2.1), the Rochester Lead-in-
Dust Study (Section 3.2.2.2), and the HUD Grantees Program (Section 3.2.2.3), provide the most
useful and available data on the relationship between environmental-lead levels and childhood
blood-lead concentration, while the HUD National Survey of Lead-Based Paint in Housing
(Section 3.3) was the primary source of environmental-lead data used in this risk analysis.
3-13
-------
Summaries of housing units and children sampled in these four studies and the approach to blood
sampling are presented in Table 3-3a through 3-3c, while summaries of the approach to
collecting environmental-lead data in these studies are presented in Tables 3-3d through 3-3f.
(No blood-lead data were collected in the HUD National Survey.) Environmental-lead data from
these studies are summarized in Section 3.3, while children's blood-lead concentrations are
summarized in Section 3.4. These four studies were selected to provide data for this risk analysis
for the following reasons:
The studies had available data for lead in paint, dust, and soil.
The Baltimore R&M study, the Rochester study, and the HUD Grantees program also
had data available on lead in children's blood.
These studies were conducted recently.
These studies were not conducted in locations with a specific point source of lead.
These studies were conducted in the United States (source control may be different in
other countries).
The remaining six studies in the above list provided useful information to this exposure
assessment on the relationship between environmental-lead levels and children's blood-lead
concentration. However, data from these studies were not used in this risk analysis for the
following reasons:
As the three USLADP studies were longitudinal intervention studies, their designs
and implementation were not appropriate for assessing general residential lead
exposure (Section 3.2.2.4). In addition, dust samples were collected in two of these
studies using a Sirchee-Spittler vacuum method. No method was available for
converting lead loadings in these dust samples to corresponding loadings based on a
wipe dust collection technique, which was necessary for such data to be used in this
risk assessment.
The Birmingham Urban Lead Uptake study was conducted outside of the U.S. and
over ten years ago, thus considered less representative of current childhood lead
exposure in the U.S. as compared to more recent studies.
The Cincinnati Longitudinal Study and the Brigham and Women's Hospital
Longitudinal Study were conducted too long ago for their data to be considered in
this risk assessment.
3-14
-------
Table 3-3a. Summary Information on Housing Surveyed in the Baltimore R&M Study (pre-intervention phase), Rochester
Lead-in-Dust Study, HUD Grantees Program, and the HUD National Survey.
Information on
Housing
Eligibility
# of housing
units surveyed
Year surveyed
housing was
built
Dates of
environmental
and blood
sampling
Baltimore R&M Study
(pre-intervention phase)
Structurally sound houses in Baltimore, MD, with
at least one eligible child. Houses slated for R&M
interventions must have lead-based paint on at
least one surface or be built prior to 1 941 .
Previously-abated houses built prior to 1941 were
abated from 5/88 to 2/91. Modern urban units
were built after 1979 and located within a single
urban neighborhood.
At enrollment Later
Occupied vacant dropped
from study
R&M units: 56 39 20
Previously-
abated units: 16 0 0
Modern urban
units: 16 0 0
R&M units and previously-abated units were built
prior to 1941.
Modern urban units were built after 1979.
R&M units: 3/93 to 11/94
All other units: 1/93 to 7/93
Rochester Lead-in-Dust Study
Houses in Rochester, NY, with
eligible children born at
hospitals and clinics that
provided necessary information
for enrollment.
205
Pre-1940: 84%
1940-1969: 11%
1970-1979: <1%
Post-1979: 5%
8/93 to 11/93
HUD Grantees Program
(pre-intervention phase,
as of 9/97)
Privately-owned, low- and
middle-income houses likely to
contain lead-based paint
hazards and on which
interventions could be
performed in this program.
Housing eligibility differed
among the 1 4 participating
grantees (see Table 3-4).
4,999 housing units enrolled
Pre-1940: 89.6%
1940-1959: 9.2%
1960-1977: 0.9%
Post-1977: 0.3%
(522 housing units
had no age specified)
2/94 to 9/97
(environmental sampling)
5/94 to 8/97
(blood sampling)
HUD National Survey
(privately-owned units
only)
Occupied permanent
housing in the 48
conterminous states built
prior to 1 980 with the
potential for containing
children
284
Pre-1940: 27%
1940-1959: 31%
1960-1979: 42%
11/89 to 3/90
(environmental sampling
only)
-------
Table 3-3b. Summary Information on Children Surveyed in the Baltimore R&M Study (pre-intervention phase), Rochester
Lead-in-Dust Study, HUD Grantees Program, and the HUD National Survey.
Information on
Children
Group being
considered in this
table
Number of
children in the
above group
Age breakdown in
the above group
Racial breakdown
of the above
group
" Baltimore R&M Study
I;: (pre-intervention phase)
Children living in units at the time of
enrollment and prior to any interventions
performed in the study, and having blood
sample data at pre-intervention
115 children in 87 units (16 modern urban
units, 1 5 previously-abated units, and 56
R&M units; from 1 to 4 children sampled
per unit).
Note: Excluded from the above group are
48 children whose first blood sample in the
study was taken prior to moving into a
vacant study unit in which R&M
interventions were complete.
Aae at blood draw
0-1 2 months: 11%
13-24 months: 30%
25-36 months: 29%
37-72 months: 30%
African-American: 100%
Rochester Lead-in-Dust Study
Children contributing blood
samples in the surveyed units
205 (one per housing unit)
12-1 8 months: 44%
18-24 months: 28%
24-30 months: 28%
African-American: 42%
White: 42%
Puerto Rican/Hispanic: 8%
Other: 8%
HUD Grantees Program
(pre-intervention phase,
as of 9/97)
Children contributing blood
samples in the surveyed units
prior to intervention
1,306 children in 830 housing
units
(from 1 to 5 children per
housing unit)
< 1 year: 5%
1-2 years: 38%
3-4 years: 36%
> 4 years: 21%
(66 children had no
age specified)
African-American: 44%
White: 26%
Hispanic: 15%
Asian: 10%
Native American: 1 %
Other: 4%
(75 children had no
race specified)
% HUD National Survey
(privately-owned units only)
Children less than seven years
of age who are the youngest
residents in a surveyed
housing unit (no blood
sampling was done in this
study)
90 (i.e., 90 housing units had
at least one child less than
seven years of age)
0-11 months: 11%
12-23 months: 17%
24-35 months: 19%
36-47 months: 16%
48-59 months: 1 7%
60-71 months: 10%
72-83 months: 1 1 %
White, non-Hispanic: 67%
Hispanic: 1 8%
African-American: 1 1 %
Other: 2%
Asian/Pacific: 1 %
No information: 1 %
0>
-------
Table 3-3b. Summary Information on Children Surveyed in the Baltimore R&M Study (pre-intervention phase), Rochester
Lead-in-Dust Study, HUD Grantees Program, and the HUD National Survey. (Continued)
Information on
Children
Baltimore R&M Study
(pro-intervention phase)
Rochester Lead-in-Dust Study
HUD Grantees Program
(ore-intervention phase,
as of 9/97)
HUD National Survey
(privately-owned units only)
Frequency of
paint pica activity
in the above
group
Pica never occurs: 87%
Pica < 1 day per month: 3.5%
Pica < 1 day per week: 3.5%
Pica > 1 day per week: 3.5%
Daily pica: 2%
Pica never observed: 1 %
Pica never occurs; 90%
Pica rarely occurs: 6%
Pica sometimes occurs: < 1 %
Pica often occurs: 2%
Pica always occurs: < 1 %
No information collected
No information collected
Frequency of soil
pica activity in the
above group
Pica never occurs: 82.5%
Pica < 1 day per month: 5%
Pica < 1 day per week: 4%
Pica > 1 day per week: 3.5%
Daily pica: 3%
Pica never observed/other: 2%
Pica never occurs; 52%
Pica rarely occurs: 22%
Pica sometimes occurs: 22%
Pica often occurs: 4%
Pica always occurs: 1 %
No information collected
No information collected
-------
Table 3-3c. Information on Blood Sampling and Analysis in the Baltimore R&M Study (pre-intervention phase), Rochester
Lead-in-Dust Study, and the HUD Grantees Program.
Information on ' * :
Blood Sampling
and Analysis
Eligibility
(see above for numbers of children
sampled and demographic breakdowns)
Method of blood sampling
Chemical analysis method
Baltimore R&M Study
(pre-intervention phase)
6-60 months of age (at enrollment)
no disability
spent at least 75% of time at the
unit
no definite and immediate plans to
move from the unit at the time of
enrollment
Note: No restriction was placed on the
minimum amount of time that the child
has lived in the given housing unit.
Venipuncture
GFAA/ASV
Rochester Lead-in-Dust Study
12-31 months of age
resided in same house since six
months of age
spent at least 20 hours per week at
primary residence
exclusions were made if
confounding factors could affect
blood-lead cone.
Venipuncture
GFAA
HUD Grantees Program
(pre-intervention phase,
as of 9/97)
Eligibility is dependent on the
participating grantee
Venipuncture: 877 children (67.2%)
Fingerstick: 429 children (32.8%)
GFAAS/ASV
CO
00
Note: No blood samples were collected in the HUD National Survey.
-------
Table 3-3d. Summary of Approaches for Soil Sampling and Analysis in the Baltimore R&M Study (pre-intervention phase),
Rochester Lead-in-Dust Study, HUD Grantees Program, and the HUD National Survey.
Exterior SoU
Sampling and
Analysis
tt composite samples
from dripline
tt composite samples
from entryway
tt composite samples
from remote areas
# composite samples
from play areas
tt core samples per
composite
depth of core samples
sampling
approach/method
laboratory sample
preparation
laboratory analysis
method
Baltimore R&M Study
(pre-Biterverrtk>n phase)
1 per housing unit in 28 units
-
--
--
3
0.5 in.
Core samples taken from
randomly-determined areas using
a 6" stainless steel recovery
probe and collected into a
polystyrene liner
Samples were dried, sieved, and
homogenized. Samples were
digested using SW 846-3015
and SW 846-3051.
GFAA(SW 846-7421)
Rochester Lead-in-Dust Study
1 per housing unit in 1 86 units
-
-
1 per housing unit in 87 units
12 (dripline samples)
8-10 (play area samples)
0.5 in.
3 dripline core samples were
taken at each side of the unit
and composited. Samples were
composited in polyethylene
bags.
Samples were mixed and sieved
into fine (250 /jm) and coarse (2
mm) fractions. Each fraction
was digested using SW 846-
3050 and analyzed separately.
FAA (method 239.1)
HUD Grantees Program
(pre-intervention phase, as of
9/97)
1 per housing unit in 557 units
-
-
1 per housing unit in 330 units
5-10
0.5 - 1.0 in.
5-10 dripline samples were
taken from all sides of the
building (21 from foundation and
2' from each other). 5-10
samples from play areas were
collected along x-shaped grids
(each sample at least 1 ' from
each other).
No information
EPA (method SW-846)
HUD National Survey
(privately-owned units only)
1 per housing unit in 249 units
1 per housing unit in 260 units
1 per housing unit in 253 units
1 to 3 per housing unit in 6
units (total of 1 1 samples)
3
10 cm
Dripline core samples were
taken from a common side of
the unit. Remote samples were
taken halfway between the unit
and its property boundary.
SW-846 digestion protocol used
ICP-AES
w
<0
-------
Table 3-3e. Summary of Approaches for Dust Sampling and Analysis in the Baltimore R&M Study (pre-intervention phase),
Rochester Lead-in-Dust Study, HUD Grantees Program, and the HUD National Survey.
' pf :' Dust Sampling « | ;i ;; '; ;;?: :;:
and Analysis ::;:
FLOORS
Rooms sampled
# samples collected
per room
Sample compositing
n samples analyzed
from floors
tt samples analyzed
from floors labeled as
uncarpeted
# samples analyzed
from floors labeled as
carpeted
Baltimore R&M Study |
(pre-intervention phase)
All interior rooms
2 (from randomly-determined
areas along the perimeter of
the room)
3 composites were formed per
unit:
Samples from all first-story
rooms with windows
Samples from all second-
story rooms with windows
Samples from all rooms
with no windows
490 composite samples in at
least 1 22 housing units
(includes composites
containing dust from both
carpeted and uncarpeted
floors)
352 samples in 1 22 housing
units (composite dust samples
from uncarpeted floors only)
53 composite samples in 34
housing units
(composite dust samples from
carpeted floors only)
Rochester Lead-in-Dust Study
Child's bedroom, kitchen, play
area, living room, entryway
3 from 1 ft2 areas, one per
sampling method, taken side-
by-side in the midpoint of the
room or where the child plays
most frequently
No compositing done
817 samples in 205 housing
units
405 samples in 205 housing
units
412 samples in 205 housing
units
HUD Grantees Program ;
(pre-intervention phase, as of
- :s ,, £ ;'; 9/97)- ,
Entryway, children's principal
play room, kitchen, up to two
children's bedrooms (one or
more additional rooms were
occasionally sampled)
From 1 to 5 dust sample results
were reported per room (98%
of rooms had one sample
result).
No compositing done.
12,260 samples in 2,846
housing units
9,044 samples in 2,797
housing units
3,216 samples in 1,396
housing units
: HUD National Survey
(privately-owned units only)
One wet room, one dry room
(both selected randomly from all
such rooms), and entryway. (See
Glossary in Appendix A for
definitions of wet and dry rooms)
Note: Common areas in
multifamily units were also
sampled, but their data were not
used in this risk analysis.
1 (each from a 1 ft2 area)
No compositing done
838 samples taken from 282
housing units (includes samples
from floor surfaces with no
recorded surface type)
335 samples taken from 214
housing units
470 samples taken from 241
housing units
fO
O
-------
Table 3-3e. Summary of Approaches for Dust Sampling and Analysis in the Baltimore R&M Study (pre-intervention phase),
Rochester Lead-in-Dust Study, HUD Grantees Program, and the HUD National Survey. (Continued)
Dust Sampling
and Analysis
WINDOW
SILLS
Rooms sampled
n samples collected
per room
Sample compositing
approach
# samples analyzed
Sample collection method(s)
laboratory sample preparation
laboratory analysis method
Baltimore R&M Study
(pro-intervention phase]
All interior rooms with
windows
Equal to the number of
windows in the room available
for sampling
All window sill dust samples
were composited into a single
sample
268 samples in 1 35 housing
units
BRM vacuum sampler
Digested using SW 846-3015
and SW 846-3051.
ICP-AESISW 846-60 10)
(GFAA (SW846-7421) was
used if levels were below the
ICP limit of quantitation)
Rochester lead-in-Dust Study
Child's bedroom, play area,
living room
3 (one per sampling method
on a common window sill)
No compositing done
363 samples in 205 housing
units
BRM vacuum sampler
DVM vacuum sampler
Wipe sampling
Digested using SW846-3051
FAA (method 239.1)
(GFAA (method 239.2) was
used if levels were below FAA
detection limits)
HUD Grantees Program :?«
(pre-intervention phase, as of
=< « s? 9/97) ::i:: H,
Kitchen, bedrooms, principal
play room, up to two additional
bedrooms (one or more
additional rooms were
occasionally sampled)
1 or 2 dust sample results were
reported per room (99% of
rooms had one sample result).
No compositing done
5,526 samples in 2,702
housing units
Wipe sampling on floors and
window sills
DVM vacuum sampler on some
carpeted floors
Digested using SW846
Flame AA or ICP
HUD National Survey £
:: (privately-owned units only)
One wet room and one dry room
(selected randomly from all such
rooms)
Note: Common areas in
multifamily units were also
sampled, but their data were not
used in this risk analysis.
1
No compositing done
392 samples in 245 housing units
Blue Nozzle vacuum sampler
SW-846 digestion protocol used
GFAA
(A)
to
-------
Table 3-3f. Summary of Approaches for Paint Sampling and Analysis in the Baltimore R&M Study (pre-intervention phase),
Rochester Lead-in-Dust Study, HUD Grantees Program, and the HUD National Survey.
Pabit Sampling
and Analysis
Baltimore R&M Study
(pro-intervention phase)
Rochester Lead-in-Dust Study
HUD Grantees Program
(pro-intervention phase,
as of 9/97)
HUD National Survey
(privately-owned units only)
OJ
ro
Interior rooms
sampled
Interior
components
sampled
Exterior
components
sampled
tt of sampled
components per
housing unit
No specified rooms or components
were identified for paint sampling
in the study protocol. Paint
sampling was done as a screening
procedure to determine the
presence of lead-based paint in
only those units slated for R&M
intervention. Sampled
components were not selected
randomly or by any other sampling
protocol.
Kitchen, child's bedroom, play
area, entryway
Components sampled at a
minimum: window sill, window
sash, window well, trim, door,
door jamb, painted floor
Components sampled at a
minimum: door, door jamb, siding,
masonry
No more than 15 samples were to
be taken per unit.
Extensive sampling of painted
components in all interior rooms,
on the building's exterior, and on
various painted exterior surfaces
associated with the unit, was
done to determine the presence
and location of lead-based paint.
Within a room, components with
different painting histories were to
be tested separately.
One wet room and one dry room.
Also one common area room in
multifamily units. One room of
each type was selected randomly
from all such rooms.
Painted components were grouped
into four strata:
Walls, ceilings, and floors
Metal substrates
Nonmetal substrates
Other components
In each sampled room, one
component from each stratum
was sampled. Then, an additional
component anywhere in the unit
was sampled.
Painted components on a single
exterior wall were grouped into
four strata:
Wall
Metal substrates
Nonmetal substrates
Other components
On this wall, one component from
each stratum was sampled (if
available). Then, an additional
component anywhere on the wall
was sampled.
From 1 to 34 (average of 17 per
unit)
-------
Table 3-3f. Summary of Approaches for Paint Sampling and Analysis in the Baltimore R&M Study (pre-intervention phase),
Rochester Lead-in-Dust Study, HUD Grantees Program, and the HUD National Survey. (Continued)
Paint Sampling
and Analysis :
Baltimore R&M Study
(pre-intervention phase)
Rochester Lead-in-Dust Study
HUD Grantees Program
. (pre-intervention phase,
* as of 9/97}
HUD National Survey J?
(privately-owned units only)
in situ
measurement
device used
XRF
Microlead I XRF
XRF (device type may vary among
the grantees)
MAP-3 XRF
laboratory
measurement
device used
AAS or ICP (used only when in
situ XRF could not be used)
Not specified (up to 10 paint chip
samples could be taken from each
unit for laboratory analyses when
XRF results fall between 0.4 and
1.5 mg/cm2)
OJ
ro
CO
Method to rating
paint condition
Paint condition was specified at
the housing unit level for older
housing only at enrollment:
none/little peeling paint (65 of 125
units) vs. extensive peeling paint
(12 of 125 units)
Three categories:
Good (0-5% deteriorated)
Fair (5-15% deteriorated)
Poor (> 15% deteriorated)
Only sampled components were
rated.
Three categories:
Good: Paint intact does not
chalk
Fair: Largely intact with
cracks and chipping
Poor: Peeling, chalking,
blistering, flaking
Four categories:
0% deteriorated
Less than 10% deteriorated
10-25% deteriorated
> 25% deteriorated
Total square feet of deteriorated
paint was recorded for some (not
all) sampled components
containing lead-based paint.
-------
The remaining subsections provide details on key objectives and conclusions on the
effects of childhood lead exposure along with an overview of the sampling designs, for the above
nine studies.
3.2.2.1 Baltimore Repair and Maintenance (R&M) Study
The objectives of the Lead-Based Paint Abatement and Repair and Maintenance Study
(USEPA, 1996b), cited as the "Baltimore R&M Study" in this document, were to characterize the
efficacy of comprehensive lead-based paint abatement for up to six years after the abatements
and to characterize the efficacy and costs of three levels (low, medium and high) of less costly
Repair and Maintenance interventions. Environmental-lead and blood-lead data measured prior
to performing interventions in this study were used in this risk analysis (USEPA, 1996b). These
data were provided by Kennedy Krieger Institute, who was responsible for the overall design and
conduct of the study.
In 1992, three groups of housing units were recruited for this study. In the first group, 16
dwellings were chosen from 90 occupied, low-income housing units that were built prior to 1941
and were abated between May, 1988, and April, 1992, as part of the Baltimore City and Kennedy
Krieger Institute Pilot Abatement Projects. The second group, slated to receive R&M
interventions in this study, consisted of 95 vacant or occupied, low-income dwellings in
Baltimore City built prior to 1941. Twenty of these housing units were later removed from the
study. Finally, 16 occupied, modern urban dwellings believed to be free of lead-based paint were
chosen as control units. These units were chosen from clusters of urban houses built after 1979.
At enrollment, all occupied units had to include at least one eligible child aged 6 to 60 months
who spent most of his/her time at the unit. All vacant units were to become occupied following
R&M interventions. All children in this study were African-American.
Prior to any intervention in this study, blood-lead concentrations were measured for 115
children that lived in 87 of the housing units at the time of enrollment, hi addition, blood-lead
concentrations were measured on 48 children before they moved into one of 39 housing units that
were vacant at the time of enrollment and that had R&M interventions performed in this study
prior to being occupied. These blood-lead concentration data and environmental-lead data from
samples collected prior to any R&M intervention performed in this study were used in this risk
assessment.
The BRM vacuum method, consisting of a modified HVS3 cyclone collector, was the
primary dust sampling method used in the R&M Study (USEPA, 1995c). Within each housing
unit, rooms were divided into three groups: first-story rooms with windows, second-story rooms
with windows, and all rooms with no windows. Within each group of rooms, a composite
sample of floor dust from multiple areas along the perimeter of each room was collected, hi
addition, the following four composite dust samples were collected in each unit: dust from first-
story window sills, dust from first-story window wells, dust from second-story window sills, and
dust from second-story window wells.
3-24
-------
Lead levels in paint were measured through in situ x-ray fluorescence (XRF)
measurement in only those units slated for R&M intervention, in order to determine whether
lead-based paint existed in these units. Only, those components suspected of being covered with
lead-contaminated paint were measured, and no specified protocol was followed to take these
measurements. As a result, paint-lead measurements in this study do not represent a random
sampling of painted surfaces in a housing unit and should be not used to make generalizations on
lead levels in paint within these types of housing.
At 28 units, three soil samples were taken from (Vz inch) soil cores collected at the
foundation (dripline) and composited. The number of units with soil samples was small due to
the lack of available dripline soil to sample. Two-hour stagnation drinking water samples were
also collected from units occupied prior to interventions. Dust, soil, and water samples were
analyzed for lead using inductively coupled plasma-atomic emission spectrometry or graphite
furnace atomic absorption spectroscopy. A structured questionnaire collected information on
study children and the households.
The primary conclusions made from summary and analysis of the pre-intervention data in
this study were as follows:
Pre-intervention dust-lead loadings in units slated for R&M interventions in this study
were higher than those in previously-abated units by approximately one to two orders
of magnitude. Furthermore, dust-lead concentrations in previously-abated units were
higher than those in modern urban units by approximately two to three orders of
magnitude. Differences of approximately one order of magnitude were observed in
pre-intervention dust-lead concentration between units slated for R&M interventions
and previously-abated units, and between previously-abated units and modern urban
units.
Dust-lead levels in previously-abated units were moderately elevated, despite the
abatement efforts on these units that preceded this study by two to four years. This
can be partially due to the location of these units in older neighborhoods, or to
residual contamination from abatement.
Blood-lead concentrations were low for children living in modern urban units
compared to units slated for R&M intervention and previously-abated units. The
continued presence of elevated blood-lead concentrations in previously-abated units
implies that the abatement effort performed in these units prior to this study did not
necessarily reduce blood-lead concentrations to acceptable levels in these units.
Significant linear correlation was observed between blood-lead concentrations and
environmental-lead levels when considering all children regardless of housing group.
Lead levels in drinking water were very low and considered not to be a lead hazard to
children in this study.
3-25
-------
Details on pre-intervention environmental-lead levels and blood-lead concentrations in this study
are presented in Sections 3.3.1.2 and 3.4.2, respectively.
3.2.2.2 Rochester Lead-in-Dust Study
The Rochester study (USHUD, 1995a and Lanphear et al., 1995), conducted in 1993, was
a cross-sectional design study whose primary objective was to obtain information on the
association between lead levels in house dust and blood-lead concentrations of resident children.
Children between the ages of 12 and 31 months and living hi the city of Rochester, NY, were
eligible for this study, provided:
they or their environment had not undergone recent interventions that were likely to
alter blood or dust lead (e.g., major renovation, recent ingestion of prescribed iron
products, or any medical or environmental intervention for an elevated blood-lead
level),
they did not spend more than 20 hours per week away from home, and
they did not live with an adult exposed to lead from an occupational or recreational
activity (Lanphear et al., 1995).
Random sampling techniques were used to recruit children bom from March 1,1991, to
September 30,1992, at either Rochester General Hospital, Strong Memorial Hospital, or St.
Mary's Hospital. Data for 205 families and children were included in the analysis. Succinct
descriptions of the quality control procedures employed for all laboratory samples during the
Rochester Lead-in-Dust Study are given in Lanphear et al., 1995. The Rochester study dataset is
publicly available and can be obtained from the National Center for Lead-Safe Housing.
During visits to the home of each study participant, an environmental health team
obtained a venipuncture blood sample from the eligible child, completed a behavioral
questionnaire for the household regarding lead exposure, collected environmental samples
(interior dust, exterior soil, water), and took in situ measurements of lead hi paint. The dust
samples were collected from floors, window sills, and window wells within rooms hi which the
child was frequently present. This risk analysis considered only dust-lead loading results from
samples collected using wipe techniques ("Little Ones" baby wipes). However, because a
secondary objective of the Rochester study was to evaluate various dust sampling methods
relative to predicting children's blood lead levels, dust samples were also collected using the
University of Cincinnati Dust Vacuum Method (DVM) and the BRM vacuum (USEPA, 1995c).
Lead concentrations of dust samples collected using the BRM vacuum are also summarized.
Side-by-side dust samples were collected at specific locations, with each sample corresponding
to a particular collection method and the wipe sample being the first to be collected. Dust
samples were analyzed using either flame or graphite furnace atomic absorption spectroscopy.
Soil samples, taken at the play area and dripline, were analyzed using flame atomic absorption
spectroscopy. Further details on sample collection are available in Lanphear et al., 1995.
3-26
-------
Children enrolled in the Rochester study were not specifically recruited because of
elevated blood-lead concentrations. However, a disproportionate percentage of these children
exhibited two risk factors associated with elevated blood-lead concentrations: residing in older
housing (at least 84% of the homes were built prior to 1940) and belonging to low-income
families (55% of households had incomes below $15,500).
Based on analysis of the public dataset, the geometric mean blood-lead concentration for
the 205 children in the Rochester study was 6.38 ug/dL, with a geometric standard deviation of
1.85. Twenty-three percent of the children had blood-lead concentrations above 10 ug/dL, 8%
above 15 |ig/dL, and 3% above 20 u£/dL. Further summaries of blood-lead concentrations in
this study are presented in Section 3.4.3.
A statistical approach using linear regression techniques (Neter and Wasserman, 1974)
was used to determine those environmental variables and questionnaire variables most important
to predicting blood-lead concentration in children (USHUD, 1995a and Lanphear et al., 1995).
In addition to wipe dust-lead loading, the following factors were significantly associated with
increased blood-lead concentrations among children: African-American race, children engaging
in soil pica, single parent household, and high ferritin levels. Adjusting for these factors, wipe
dust-lead loading accounted for 10.1% of the variation in blood-lead concentrations (USHUD,
1995a and Lanphear et al., 1995).
The Rochester study also investigated the relationship between soil-lead concentration
and children's blood-lead concentration. One composite soil sample was obtained from a
maximum of 12 core samples (3 per side of house) taken two feet away from the foundation, and
a second composite sample was obtained from 8-10 samples taken where the child frequently
played. A coring device was used to take samples at a depth of l/i inches only where bare soil
was present. When asked "How often does [the study child] put dirt or sand in his/her mouth,"
27% of the home interview respondents indicated that the study child for which they were
responding "sometimes," "often," or "always" did. The remaining 53% indicated the study child
"never" or "rarely" did (USHUD 1995a). Soil-lead concentration was a significant (positive)
predictor of blood-lead concentration, even when adjusting for dust-lead loading (USHUD,
1995a).
The Rochester study concluded that lead-contaminated dust significantly contributes to
children's blood-lead concentrations, even when those concentrations are in the low to moderate
range (Lanphear et al., 1996b). This relationship differs according to the dust sampling method
and the type of surface sampled. At the relatively low levels of dust-lead loadings and
concentrations in this study, dust-lead loadings were found to be a better predictor of blood-lead
concentration than were dust-lead concentrations. Of the three dust collection methods
considered, dust-lead loadings from samples collected using either wipe or BRM-vacuum
methods were more highly correlated with blood-lead concentrations than were loadings from
DVM dust samples (Lanphear et al., 1995). As data were collected in late summer and autumn,
the seasonal effect on these relationships could not be measured.
3-27
-------
Summaries of environmental-lead levels observed in the Rochester study are presented in
Section 3.3.1.3. Data from the Rochester study were employed to develop the empirical model
used in this risk assessment (Section 4.2).
3.2.2.3 Evaluation of the HUD Lead-Based Paint Hazard Control
Grant Program ("HUD Grantees")
Since 1994. grantees participating in the HUD Lead-Based Paint Hazard Control Grant
Program have conducted interventions in privately-owned low- and middle-income housing to
control lead-based paint hazards. The grantees are primarily affiliated with states and local
governments. In this program, HUD has supplied an additional grant to the National Center for
Lead-Safe Housing (NCLSH) to evaluate the cost and efficacy of the interventions being
conducted. In this evaluation, fourteen grantees are collecting data on environmental, biological,
demographic, housing, cost, and hazard-control aspects of the interventions they are performing.
NCLSH is conducting this evaluation with the Department of Environmental Health at the
University of Cincinnati (UCDEH).
Among the data being collected in this evaluation are the following:
lead loadings in dust samples using wipe collection techniques, determined prior to
and following any environmental intervention in a housing unit. Carpeted or
uncarpeted floors, window sills, and window wells were sampled. Rooms sampled
included entryways, children's principal play room (or living room), kitchen, and up
to two children's bedrooms. The DVM sampler was occasionally used on carpets, but
these data were not considered in this exposure assessment.
blood-lead concentration for children between the ages of six months and six years,
determined prior to and following any environmental intervention in a housing unit.
While the program recommended venipuncture collection techniques, some grantees
are using fingerstick methods. Blood samples were analyzed by graphite furnace
atomic absorption spectrophotometry (GFAAS) or by anodic stripping voltammetry
(ASV).
lead levels on painted surfaces prior to intervention to determine the presence and
location of lead-based paint. Portable XRF measurement techniques were used, but
laboratory testing of paint chips was also employed when XRF measurements were
indeterminant.
soil-lead concentration prior to and following any environmental intervention, where
composite soil samples were collected from the dripline (foundation) and from
children's play areas. As soil sampling was optional in this program, the availability
of soil-lead concentration data is limited.
demographic information on the household and on the resident children, such as
income level, age of house, age of child, and mouthing behavior.
3-28
-------
Only pre-intervention data collected from February, 1994, to September, 1997, were available for
this exposure assessment. These data provide some of the most recent information on
environmental-lead measurements in housing units with high potential for containing lead-based
paint hazards, and the relationship of these measurements with children's blood-lead
concentration.
The grantees followed specified sampling protocols and used standard data collection
forms developed specifically for this evaluation (NCLSH and UCDEH, 1994). However, as it
was HUD's desire to emphasize local control of the individual programs, each grantee was given
some freedom in developing their approach to recruitment and enrollment. Some grantees
targeted high-risk neighborhoods, while others enrolled only homes with a lead-poisoned child,
while still others considered unsolicited applications. The locations at which data were collected,
along with the enrollment criteria, are summarized in Table 3-4.
Preliminary conclusions made on the pre-intervention data from the HUD Grantees
evaluation program are as follows:
Blood-lead concentrations and environmental-lead levels tend to vary widely across
grantee locations, primarily due to how housing units were targeted for enrollment by
each grantee, and methods used to obtain and analyze the samples.
Compared to the national housing stock as a whole, housing units enrolled in the
evaluation program are more likely to contain lead-based paint hazards (e.g., older or
low-income housing, or the neighborhood has a history of lead-based paint hazards)
or to contain children with elevated blood-lead concentrations. As a result, blood-
lead concentrations and environmental-lead levels tended to be high for most housing
units. However, when interpreting results of any analyses of data from this program,
one should be aware of regional or strategy selection biases that may be present.
Results of interim comparisons of environmental-lead levels and blood-lead concentrations
between pre-intervention and post-intervention periods are found in NCLSH and UCDEH, 1997.
Summaries of pre-intervention environmental-lead measurements from the HUD
Grantees evaluation program are presented in Section 3.3.1.4, while pre-intervention blood-lead
concentrations and their observed relationships with selected environmental-lead parameters are
summarized hi Section 3.4.3. Note that these data are considered preliminary, as only data
collected through September, 1997, were available to this exposure assessment.
3-29
-------
Table 3-4. Location of Grantees Participating in HUD Grantee Program Evaluation, and
Grantees' Criteria for Enrollment/Recruitment of Housing Units.
Location of
Grantee
Alameda
County2
Baltimore
Boston
California2
Chicago
Cleveland2
Massachusetts
Milwaukee2
Minnesota2
New Jersey3
New York City3
Rhode Island2
Vermont2
Wisconsin2
Enrollment Plan
Targeting 4 high-risk cities (Alameda, Berkeley, Emeryville, Oakland); many
units contain a lead-poisoned child
Targeting 3 neighborhoods, 2 of which have histories of lead-poisoning;
predominantly rowhouses
Enrolling only units which have received an order to abate based on the
identification of a lead-poisoned child
Targeting older homes in low-income neighborhoods
Targeting 5 neighborhoods; units are selected based on reports of a lead-
poisoned child and after a special compliance hearing is held
Using two criteria independently: one targets units with a lead-poisoned
child, and the other targets homes in a single neighborhood
Primarily enrolling units under existing orders to abate because of the
presence, at some time, of a lead-poisoned child (Brockton, Chelsea,
Lawrence, and Worchester)
Targeting several of the lowest income neighborhoods in the city; units are
selected from referrals of families with a lead-poisoned child
Minneapolis/St. Paul: targeting units with a lead-poisoned child
Duluth: targeting units with deteriorated housing conditions
Selecting units in conjunction with concurrent comprehensive housing
renovation/rehabilitation
Targeting neighborhoods with the highest percentages of lead poisonings;
one of two programs is specifically targeting families with newborn babies
living in deteriorated housing
Enrolling only units that meet Section 8 Housing Quality Standards, and the
owner cannot own more than 1 2 units
Considering referrals of families with lead-poisoned children, non-profit
housing developers who learn of the program when applying for federal
HOME funds, and unsolicited applications
Each of the 1 2 sub-grantees within the state (not counting Milwaukee) use
own criteria (no information given)
TOTAL NUMBER OF HOUSING UNITS ENROLLED1
» Units
Enrolled1
334
649
158
186
185
264
327
477
282
119
387
383
954
294
4999
1 Through September, 1997. Environmental-lead and/or blood-lead data were not available for some units.
2 Grantee collected soil samples as well as samples of other environmental media.
3 Grantee did not collect blood samples.
3-30
-------
3.2.2.4 Urban Soil Lead Abatement Demonstration Project (USLADP)
The USLADP, authorized in 1986 under the Superfund Amendments and Reauthorization
Act, was conducted in 1988-1991 to determine whether reducing lead levels in soil accessible to
children decreases their blood-lead concentration (USEPA, 1996a). While other observational
studies of childhood lead exposure such as the Rochester study have shown that differences in
soil lead exposure are associated with differences in blood-lead concentration, this project
specifically addressed whether controlled reductions in external soil lead exposure were
associated with reductions in blood-lead concentrations. The USLADP consisted of three studies
conducted in Baltimore, MD, Boston, MA, and Cincinnati, OH. This project considered soil
abatements in urban areas and focused on inner-city children. The USLADP Integrated Report
(USEPA, 1996a) is the source for study details reported here.
In Baltimore, data were analyzed for 185 children aged 6 to 72 months. These children
resided in either the study area (expectation of moderate risk of lead poisoning) or a control area.
The Boston study included 149 children aged 6 to 48 months, considered to be at risk for lead
exposure and residing in one of the study areas (history of high incidence of lead poisoning).
Only children with blood-lead concentrations ranging from 7 to 24 ug/dL were included in the
Boston study. In Cincinnati, families with children under five years of age and residing in one of
the study areas (selected as having similar socioeconomic and housing type characteristics) were
enrolled in the study. Data for 206 children were analyzed from the Cincinnati study.
Within each city, a series of neighborhoods were considered in the study from which the
participating households were selected. Selected units within certain neighborhoods were to
have interventions performed, while units in other neighborhoods were selected as control units.
For purposes of data summary and analysis, study units were grouped according to intervention
strategy. Environmental media sampled included dust, soil, drinking water, and paint.
Household interviews were also conducted to obtain information on such factors as household
behavior and socioeconomic status.
The following two main conclusions were drawn from the USLADP (USEPA, 1996a):
1. "When soil is a significant source of lead in the child's environment, under certain
conditions, the abatement of that soil will result in a reduction in exposure that will
cause a reduction in childhood blood lead concentrations. "
2. "Although these conditions for a reduction in blood are not fully understood, it is
likely that five factors are important in determining the magnitude of any possible
reduction: (1) the past history of exposure of the child to lead, as reflected in thepre-
abatement blood lead; (2) the initial soil lead concentration and the magnitude of the
reduction in soil lead concentrations; (3) the initial interior house dust lead loading
and the magnitude of reduction in house dust lead loading; (4) the magnitude of other
sources of lead exposure, relative to soil; and (5) the strength of the exposure
pathway between soil and the child relative to other lead exposure pathways in the
child's environment."
3-31
-------
The five factors specified in the second conclusion contributed to differences among the studies
on the impact of soil abatement. The Baltimore and Cincinnati studies concluded that the impact
of soil abatement on blood-lead concentrations was limited, hi part due to low pre-intervention
soil-lead concentrations and the extent to which soil was contributing to lead in house dust. Soil-
lead concentrations were higher in the Boston study, where soil abatement was associated with
declines in blood-lead concentration. However, researchers in the Boston study concluded that
these declines are generally modest, and as a result, may not warrant the resources required to
conduct a soil abatement when only low levels of lead exposure are present (Weitzman et al.,
1993).
The entire soil region surrounding the residence was partitioned into distinct areas (e.g.,
front, back), and samples were taken from each partition. At each core sample, the top 2" and
bottom 2" of the sample core were retained. A single core sample was taken when less than two
meters in either direction were available for sampling. Larger areas had core samples taken at the
foundation and at the boundary.
Dust samples were collected by vacuum methods in all three cities. In Baltimore, the
Sirchee-Spittler vacuum sampler was used to collect dust samples from a 4' x 4' sample area
demarcated with tape. A minimum of three areas were sampled: the main entrance to the
household and two areas often frequented by the child when playing. In Boston, the same
sampler and sampling sites were used, but a plastic 25 cm x 25 cm frame was used instead of
tape. In Cincinnati, the DVM sampler was used with a plastic 25 cm x 25 cm frame. Dust was
sampled from a floor area adjacent to the main entrance from a floormat placed by sample
collection personnel. In addition, a composite of dust samples from floors was collected from at
least three areas including the child's bedroom and a high traffic area in the main living area, and
a composite of dust samples was collected from at least three window well and sill areas
including from within the child's bedroom and the main living area. Dustfall and exterior surface
dust were also measured.
The design of the USLADP studies allowed EPA to evaluate some of the effects of soil
lead abatement. The design and implementation of the study was appropriate for a longitudinal
intervention study, and no other uses of the study data were anticipated. As a result, data from
these studies were not used for general exposure assessment in this risk analysis. Other reasons
for not using these data in the risk analysis are as follows:
1. The housing units and children sampled in the Baltimore study were not intended to
represent a cross-section of Baltimore children, nor a cross-section of housing hi
Baltimore. They were chosen because they had a large number of lower-income pre-
school children and were believed to have high yard soil lead concentrations.
2. The Boston study was designed to include only children whose blood lead was not
too high (< 24 ug/dL), not too low (at least 7 ng/dL), and who lived hi housing with
high yard soil lead (in general, at least 1000 ug/g).
3-32
-------
3. The Cincinnati study was designed to control for lead paint as a factor that may
confound direct or indirect soil lead exposure. The control measure was to include
only completely rehabilitated housing, where almost all of the lead-based paint had
been stripped off of walls and trim.
4. hi the Boston and Baltimore studies, dust samples were collected using the Sirchee-
Spittler vacuum method (USEPA, 1995c). No convenient method has been
established for converting Sirchee-Spittler dust-lead loadings to wipe dust-lead
loadings necessary for the risk analysis.
5. The Cincinnati study collected soil-lead concentrations at the neighborhood level. It
is not clear how to relate neighborhood soil-lead measurement concentrations to a
specific child's exposure.
6. Age of housing unit is not reported for the majority of units in the USLADP.
3.2.2.5 Birmingham Urban Lead Uptake Study
This study was conducted hi Birmingham, England, from 1984-1985, and consisted of
183 randomly-selected children, aged 24 months (+ 2 months), born in and still residing hi urban
Birmingham. A stratified subset of 106 children were selected for the study, of which 97
completed the study. The objective of the study was to simultaneously examine lead uptake via
all identified environmental pathways for young children hi an urban environment.
Soil samples were collected using a stainless-steel trowel surface scrape (0-5 cm). One
composite soil sample was obtained from 25 core samples. A specially adapted vacuum was
used to collect dust samples from the child's main play area, the child's bedroom and under the
doormat. All exposed floor space was sampled. Samples were also taken from the bag of the
vacuum cleaner most often used by the household.
The main conclusion from this study was that childhood blood-lead concentration was
found to be significantly associated with a combination of dust-lead loading, the rate of touching
objects, water-lead concentration, and smoking habits of the parents. Only an estimated 3% of a
child's average total uptake of lead per day was attributed to breathable air; the remainder was
attributed to dust, food, and water ingestion.
Because this study was conducted outside of the United States over ten years ago, it was
considered less representative of current childhood lead exposure in the United States than more
recent studies. Therefore, the data from this study were not used in this risk analysis.
3.2.2.6 Cincinnati Longitudinal Study
Objectives of the Cincinnati Longitudinal study, conducted from 1980-1987, were to
provide a complete picture of a child's lead exposure history and to investigate the factors
responsible for excessive lead exposure. Approximately 250 expectant mothers residing within a
3-33
-------
prespecified set of census tracts in the Cincinnati (OH) area were enrolled for this study. These
census tracts were identified as having a long history of producing children with elevated blood
lead levels. The mothers were patients at one of three prenatal clinics. Once these mothers
delivered, blood-lead concentrations were measured in the children from birth through 5 years of
age.
Soil samples were collected by surface scrapings. Surface scrapings were collected from
the child's play area outside, if one existed. Interior dust samples were collected from areas
where the child frequents using a personal sized vacuum within a 484 cm2 plastic frame. A
maximum of five sites were sampled within the home. Each sample entailed three sweeps of the
vacuum within the frame. Exterior dust samples were collected via scraping exterior surfaces
with a stainless steel spatula. Paint-lead levels from a maximum of 15 surfaces were measured
using XRF techniques. Dust collection from children's hands was performed via repeated
wiping with multiple pre-moistened wipes.
This study observed high levels of lead contamination in the residential environments,
with most contamination occurring hi areas immediately outside of the unit and within the
entranceways. Statistical analyses indicated that the pathway from exterior dust to interior dust
to hands to blood was of most significance in this study (Figure 3-3).
Due to the age of this study, data were not used hi this risk analysis.
3.2.2.7 Briqham and Women's Hospital Longitudinal Study
The objective of this early study was to examine the relationship between children's
blood-lead levels and various environmental factors from late pregnancy to two years of age.
Children were selected from births occurring between April 1979 and April 1981 at Brigham and
Women's Hospital hi Boston, MA. Births were categorized into the highest, lowest, and middle
deciles of umbilical cord blood lead. The 249 infants selected were nearly equally drawn from
three distinct categories of cord blood levels. All families resided hi an urban environment
within a 12 mile radius of hospital, spoke English as their primary language, and the infants had
no serious illness. These families were predominantly white and middle- to upper-middle class.
In addition to umbilical cord blood, blood samples were collected at 6,12,18, and 24 months of
age.
At 18 and 24 months, soil samples were collected at a distance of three meters from any
road or structure. Dust samples were collected at 1,6,18, and 24 months using wipe techniques
from a living room surface (floor or furniture top) and from a window sill. Samples were
collected from within a plastic frame having a 930 cm2 opening (a 465 cm2 opening for window
sills). Lead levels in paint were measured by a PGT model XE-3 XRF instrument. Air samples
were collected from personal air monitors, and drinking water samples were collected from the
kitchen tap after a 4-liter flush.
Mean blood-lead concentrations at 24 months was 6.8 ug/dL. At 24 months, blood-lead
concentration was found to be significantly associated with soil-lead concentration, dust-lead
3-34
-------
loading, the presence of deteriorated paint, and the occurrence of recent refinishing activities at
the residence. Water-lead and airborne-lead levels were not significant factors. These findings
agreed with earlier studies which considered children with higher blood-lead concentrations. In
addition, blood-lead concentrations were found to be approximately 44% higher within
specimens collected in summer months, indicating a possible seasonality factor associated with
blood-lead concentration.
Due to the age of the study, data from this study were not used in this risk analysis.
3.3 LEAD IN DUST. SOIL. AND PAINT IN THE NATION'S HOUSING
This section provides information on the distribution of environmental-lead levels in the
nation's housing stock, with a focus on lead in residential dust, soil, and paint. This risk analysis
uses data from the HUD National Survey of Lead-Based Paint hi Housing to characterize the
distribution of environmental-lead levels in the nation's occupied housing stock in 1997. These
environmental-lead data are summarized in Section 3.3.1. To supplement the national
environmental-lead data with data for certain categories of housing units, such as inner-city
homes and older homes in an urbanized setting, environmental-lead data from the Baltimore
R&M Study (Section 3.2.2.1), the Rochester Lead-in-Dust Study (Section 3.2.2.2), and the HUD
Grantees Program (Section 3.2.2.3) are also summarized in this section. Section 3.3.1 also
includes estimated numbers of occupied housing units in the 1997 national housing stock.
To provide a link between childhood and residential environmental lead exposures,
Section 3.3.2 presents estimated numbers of children of specific age groups in 1997 residing
within housing units of specific ages.
3.3.1 The Distribution of Lead Levels in Household Dust, Soil, and Paint
In this section, environmental-lead levels in residences are summarized for four studies.
The first study presented, the HUD National Survey, is the primary source of data on
environmental-lead levels in the nation's occupied housing stock. While this study was designed
to be a nationally-representative study of environmental-lead in the nation's housing built prior to
1980, it is used here to characterize the nation's housing in 1997, prior to §403 interventions.
The other three studies, the Baltimore R&M Study, the Rochester study, and the HUD Grantees
program, provide supporting information on environmental-lead levels for specific housing
groups or exposure conditions.
As discussed in Section 3.2.2, the four studies presented in this section were selected to
provide data for this risk analysis for the following reasons:
The studies had available data for lead in paint, dust, and soil.
The Baltimore R&M Study, the Rochester study, and the HUD Grantees program
also had data available on lead in children's blood.
3-35
-------
These studies were conducted recently.
These studies were not conducted in locations with a specific point source of lead.
These studies were conducted in the United States (source control may be different
in other countries).
3.3.1.1 HUD National Survey
For this risk analysis, the primary source of information on environmental-lead levels in
the national housing stock was the National Survey of Lead-Based Paint in Housing (USEPA,
1995a; USEPA, 1995g; and USEPA 1995h). This survey was sponsored by the U.S. Department
of HUD, in response to a mandate in the 1987 amendments to the Lead-Based Paint Poisoning
Prevention Act to obtain "an estimate of the amount, characteristics and regional distribution of
housing in the United States that contains lead-based paint hazards at differing levels of
contamination." Conducted in 1989-1990, the privately-owned unit portion of the survey (cited
as the "HUD National Survey" in this document) measured lead levels hi paint, dust, and soil
within 284 privately-owned, occupied housing units. The units were selected via a statistically-
based sampling design to represent the national housing stock built prior to 1980. Units built in
1980 or later were not included in the survey, as they were assumed to be free of lead-based paint
as a result of the Consumer Product Safety Commission's 1978 regulation on the maximum
allowable lead in paint used for residences, toys, furniture, and public areas. CPSC's maximum
lead level is below the thresholds usually used to define lead-based paint (1.0 mg/cm2, or 0.5%
lead by weight).
The design of the HUD National Survey stipulated that housing units be distributed
across three age categories (pre-1940,1940-1959,1960-1979) based on proportions indicated in
the 1987 American Housing Survey and that multi-family units be oversampled (USEPA, 1995a;
USEPA, 1995g). To take into account the oversampling of multi-family units and the
overrepresentation of certain demographic groups in the final sample, the HUD National Survey
assigned sampling weights to each surveyed housing unit. The sum of all 284 sampling weights
equaled the number of pre-1980 privately-owned, occupied units in the national housing stock at
the time of the survey. Sampling weights in the HUD National Survey were determined
according to four demographic variables associated with the units:
Age category of unit
Number of units in the building
Census region
Presence of a child under age 7 years.
3-36
-------
The method to assigning sampling weights ensured that inferences based on the 284 privately-
owned homes sampled in the HUD National Survey would be representative of the pre-1980
national housing stock.
In order to use the environmental-lead levels from the HUD National Survey to
characterize environmental-lead levels in the 1997 national housing stock, it was necessary to
revise the sampling weights of the HUD National Survey units to represent the 1997 occupied
housing stock, both publicly-owned and privately-owned. (While environmental data for only
the 284 privately-owned units in the HUD National Survey were used in this risk analysis, the
revised 1997 sampling weights for these units represent both privately-owned and publicly-
owned units in the national housing stock.) Using data from the U.S. Bureau of the Census, the
method for revising the sampling weights is documented in Section 1.1.2 of Appendix Cl; Table
Cl-7 of Appendix Cl lists the revised weights for each unit. The revised weights, therefore,
indicate the number of units in the 1997 national housing stock that are represented by the given
HUD National Survey unit, and therefore, represented by its environmental-lead levels. The
estimated numbers of units in the 1997 national housing stock are presented in Table 3-5, within
four age categories.
Table 3-5. Estimated Total Number of Occupied Housing Units in the National Housing
Stock in 1997 According to Year-Built Category-
Year In Which the Unit
Was Built
Pre-1940
1940-1959
1960-1979
Post-1979
Number of
National Survey
Units
77
87
120
281
Estimated Total:
Estimated Numbers of Units in the 1997
National Housing Stock
19,676,000
19,718,000
34,985,000
24,893,000
99,272,000
1 Units built from 1960-1979 and containing no lead-based paint were placed in this
category as well as in the 1960-1979 category.
The HUD National Survey did not consider units built after 1979, as all such units were
assumed to be free of lead-based paint. In characterizing the 1997 national housing stock from
the HUD National Survey, post-1979 housing was represented by the 28 units built between 1960
and 1979 and containing no lead-based paint (i.e., the predicted maximum amount of lead in
paint within the unit was less than 1.0 mg/cm2). Therefore, the revised sampling weights for
these 28 units are the sum of two parts: one part representing 1960-1979 units, and the other
representing post-1979 units. This approach assumes that environmental lead levels in post-1979
homes are similar to environmental lead levels in homes built between 1960 and 1979 which do
not contain lead-based paint. See Section 1.1.3 of Appendix Cl and Section 3.3.1.5 on the
rationale for selecting these 28 units to represent the post-1979 housing stock and on the method
for obtaining the portion of the sampling weight representing post-1979 units.
3-37
-------
In the HUD National Survey, lead loadings (ug of lead per square-feet of area sampled)
and lead concentrations (ug of lead per gram of sample) were measured from dust samples
collected on floors, window sills, and window wells. Dust samples were collected using the Blue
Nozzle vacuum method. Lead concentrations in the soil at each unit were measured by collecting
soil samples along the foundation, the entryway to the unit, and from remote areas in the yard,
using a soil corer with plunger. Lead levels hi paint (milligrams of lead per square-centimeter of
painted surface) were measured using in situ XRF techniques in selected rooms as well as on the
exterior of the unit. Detailed protocols for sample collection are available in USEPA, 1995g.
hi the HUD National Survey, the dust-lead concentration equaled the amount of lead hi
the entire dust sample, divided by the tap weight. "Tap weight" is the portion of a dust sample
that was tapped out of the sample collection filter. Note that the tap weight could be less than the
actual weight of the collected sample. Therefore, the dust-lead concentration measurements used
hi this risk analysis were adjusted for the effect of underestimated sample weights. Details on the
method used to adjust the tap weights is available hi USEPA, 1996c. Lead concentrations for
dust samples with a tap weight of less than 0.7 mg were omitted from risk analyses.
hi this risk analysis, data from the 284 privately-owned units hi the HUD National Survey
were used to characterize environmental-lead levels in the nation's occupied housing. Table
Cl-7 of Appendix Cl contains the following summary of environmental-lead levels for each of
these units:
two weighted arithmetic averages of dust-lead loading: one for floors and one for
window sills (where each sample's results were "area-weighted," or weighted
according to area of sample location)
two weighted arithmetic averages of dust-lead concentration: one for floors and one
for window sills (where each sample's results were "mass-weighted," or weighted
according to mass of sample)
the weighted arithmetic average soil-lead concentration (where remote sample results
were weighted twice that of the entryway and dripline results)
the maximum observed amount of lead hi paint, determined for both the ulterior and
the exterior, as measured by in situ XRF techniques.
Note that the last bullet indicates the maximum observed (or measured) paint-lead concentration
hi a unit To identify whether a unit was suspected of containing any lead-based paint, even in
unsampled areas, statistical modeling was performed hi the HUD National Survey to obtain a
predicted maximum XRF measurement for each unit. If the predicted maximum XRF
measurement for a unit was greater than or equal to 1.0 mg/cm2, the unit was considered to
contain lead-based paint (USEPA, 1995a). hi this risk analysis, a unit's predicted maximum XRF
measurement was used only to identify the presence of lead-based paint within the unit
3-38
-------
Using the environmental-lead measurements and the updated 1997 sampling weights for
the HUD National Survey units from Appendix Cl, Tables 3-6 and 3-7 summarize the estimated
lead loadings and concentrations, respectively, in floor-dust samples across units in the 1997
housing stock. Tables 3-8 and 3-9 summarize lead loadings and concentrations, respectively, in
window sill-dust samples. Table 3-10 summarizes lead concentrations in soil. To summarize the
results of XRF paint testing across the surveyed units, Table 3-11 presents information on the
extent of XRF sampling, the presence of lead-based paint, the presence of deteriorated lead-based
paint, and the distribution of maximum XRF measurements in a unit, for specific categories of
interior and exterior painted components. Table 3-12 presents summaries of each unit's
maximum XRF measurement for paint within three age group categories. The percentages of
units in the 1997 housing stock having lead-based paint, as well as the percentages having
damaged lead-based paint, are estimated in Table 3-13. The statistics calculated in these tables
are summaries of the observed data (using 1997 estimated sampling weights) and do not make
any distribution assumptions. Variability in these data may result in unexpected (and likely
insignificant) trends across age categories.
Table 3-6. Summary of the Distribution of Lead Loadings in Floor-Dust Samples Within
Housing Units in the HUD National Survey, Weighted to Reflect the Predicted
1997 Housing Stock.
Surveyed Units,
According to the Year
Unit Was Built
All Units Built Before
1940
All 1940-1959 Units
All 1960-1979 Units
1960-1979 units with
no LBP2
Floor Dust-Lead Loadings U/g/ft1}1
Geometric
Mean
22.6
8.74
4.14
3.14
Geometric
Standard
Deviation
3.63
3.34
2.45
2.06
Sth
Percentile
2.83
1.25
1.20
1.21
25th
Percentile
8.47
4.20
2.28
1.76
Median
17.2
8.32
4.04
2.84
75th
Percentile
46.2
22.5
7.63
5.66
95th
Percentile
197
72.0
21.2
12.2
1 The statistics presented in this table were calculated on area-weighted (i.e., weighted for area of sample
location) arithmetic mean dust-lead loadings from floors for the 284 privately-owned, occupied National
Survey units (see Appendix CD. These loadings are converted to represent loadings from dust samples
obtained from wipe collection techniques. In the summaries, each unit is weighted by its 1997 weight,
which is presented in Appendix C1.
2 Units with no LBP have a predicted maximum XRF value (interior and exterior) less than 1.0 mg/cmz. These
units represent post-1979 units in this risk analysis.
3-39
-------
Table 3-7. Summary of the Distribution of Lead Concentrations in Floor-Dust Samples
Within Housing Units in the HUD National Survey, Weighted to Reflect the
Predicted 1997 Housing Stock.
Surveyed Units,
According to the Year
Unit Was Built
All Units Built Before
1940
All 1940- 1959 Units
All 1960-1979 Units
1960-1979 units with
no LBP2
Floor Dust-Lead Concentrations (pg/g)1
Geometric
Mean
505
201
121
91.9
Geometric
Standard
Deviation
4.00
2.64
3.01
2.16
5th
Percentile
86.6
32.2
24.5
21.5
25th
Percentile
246
101
72.1
53.5
Median
406
218
137
86.5
75th
Percentile
813
330
223
165
95th
Percentile
2260
1240
647
429
1 The statistics presented in this table were calculated on mass-weighted (i.e., weighted for mass of sample)
arithmetic mean dust-lead concentrations from floors for the 284 privately-owned, occupied National
Survey units (see Appendix CD. These concentrations were adjusted to reflect the weight of the entire
dust sample, not just the tap weight (USEPA, 1996c). In the summaries, each unit is weighted by its
1997 weight, which is presented in Appendix C1.
1 Units with no LBP have a predicted maximum XRF value (interior and exterior) less than 1.0 mg/cm2.
These units represent post-1979 units in this risk analysis.
Table 3-8. Summary of the Distribution of Lead Loadings in Window Sill-Dust Samples
Within Housing Units in the HUD National Survey, Weighted to Reflect the
Predicted 1997 Housing Stock.
Surveyed Units,
According to the Year
Unit Was Built
All Units Built Before
1940
All 1940-1 959 Units
All 1960- 1979 Units
1 960-1 979 units with
no LBP2
Window Sill Dust-Lead Loadings U'g/ft1)1
Geometric
Mean
168
22.0
16.2
8.17
Geometric
Standard
Deviation
16.7
10.7
14.6
9.94
5th
Percentile
0.797
0.659
0.250
0.122
25th
Percentile
8.32
6.77
2.82
2.58
Median
96.4
27.0
18.1
8.11
75th
Percentile
808
177
217
57.8
95th
Percentile
6190
1290
575
127
1 The statistics presented in this table were calculated on area-weighted (i.e., weighted for area of sample
location) arithmetic mean dust-lead loadings from window sills for the 284 privately-owned, occupied
National Survey units (see Appendix C1). These loadings are converted to represent loadings from dust
samples obtained from wipe collection techniques. In the summaries, each unit is weighted by its 1997
weight, which is presented in Appendix C1.
2 Units with no LBP have a predicted maximum XRF value (interior and exterior) less than 1.0 mg/cm2.
These units represent post-1979 units in this risk analysis.
3-40
-------
Table 3-9. Summary of the Distribution of Lead Concentrations in Window Sill-Dust
Samples Within Housing Units in the HUD National Survey, Weighted to
Reflect the Predicted 1997 Housing Stock.
Surveyed Unit*,
According to the Year
Unit Was Built
All Units Built Before
1940
All 1940-1 959 Units
All 1960-1 979 Units
1960-1979 units with
no LBP2
Window Sill Dust-Lead Concentrations U/g/g)1
Geometric
Mean
1710
471
377
239
Geometric
Standard
Deviation
5.24
4.08
4.91
3.26
5th
Percentile
72.1
48.1
34.4
26.0
25th
Percentile
500
244
148
124
Median
1690
510
516
267
75th
Percentile
6680
1330
1480
492
95th
Percentile
10200
4470
1570
1140
1 The statistics presented in this table were calculated on mass-weighted (i.e., weighted for mass of sample)
arithmetic mean dust-lead concentrations from window sills for the 284 privately-owned, occupied
National Survey units (see Appendix CD. These concentrations were adjusted to reflect the weight of the
entire dust sample, not just the tap weight (USEPA, 1996c). In the summaries, each unit is weighted by
its 1997 weight, which is presented in Appendix C1.
2 Units with no LBP have a predicted maximum XRF value (interior and exterior) less than 1.0 mg/cm2
These units represent post-1979 units in this risk analysis.
Table 3-10. Summary of the Distribution of Soil-Lead Concentrations for Housing
Units in the HUD National Survey, Weighted to Reflect the Predicted
1997 Housing Stock.
Surveyed units.
According to the Year
Unit Was Built
All Units Built Before
1940
All 1940-1 959 Units
All 1960-1979 Units
1960-1979 units with
no LBP2
Soil-Lead Concentrations fy/g/g}1
Geometric
Mean
463
92.6
32.8
22.4
Geometric
Standard
Deviation
3.09
3.15
2.56
2.31
5th
Percentile
35.0
22.0
6.13
5.58
25th
Percentile
138
47.6
20.4
13.6
Median
394
81.4
31.5
21.2
75th
Percentile
841
171
62.5
45.0
95th
Percentile
2000
485
183
82.5
1 The statistics presented in this table were calculated on weighted arithmetic mean soil-lead concentrations
for the 284 privately-owned, occupied National Survey units (see Appendix C1). Within each unit's
average, remote sample results were weighted twice that of the entryway and dripline results. In the
summaries, each unit was weighted by its 1997 weight, which is presented in Appendix C1.
2 Units with no LBP have a predicted maximum XRF value (interior and exterior) less than 1.0 mg/cm2.
These units represent post-1979 units in this risk analysis.
3-41
-------
Table 3-11. Summary of XRF Paint Measurements Taken in the HUD National Survey,
Including the Percentage of Housing Units with Lead-Based Paint (LBP) and
Deteriorated LBP, by Component Category.1
Component Category
# Units
with
XRF
Data
Percent of Unit* With a
Given # of
Measurements Reported
1
2
>2
Percent of
Units with
LBP2
Percerrtiles of the Distribution of
Maximum XRF Measurement in a
Unit (mg/cmz)
25th
Percentile
50th
Percentile
75th
Percentite
Percent of
Units with
Deteriorated
LBP3
Interior Components
Cabinets
Ceiling
Door components
(trim, systems)
Floors
Stairs (trim)
Trim (baseboard,
molding)
Walls
Window sills
Other window
components (trim,
systems)
99
247
219
11
14
173
256
176
169
93.9
24.7
15.1
90.9
92.9
45.1
0.8
54.0
21.3
6.1
75.3
27.9
9.1
7.1
38.7
2.3
44.9
35.5
0.0
0.0
57.1
0.0
0.0
16.2
96.9
1.1
43.2
13.1
21.1
26.9
0.0
35.7
23.7
21.9
29.5
33.7
0.1
0.3
0.5
0.3
0.3
0.5
0.3
0.5
0.5
0.5
0.6
0.6
0.4
0.6
0.6
0.6
0.6
0.6
0.6
0.7
1.0
0.6
1.8
0.9
0.8
1.2
1.3
2.2
3.7
3.2
0.0
7.7
3.5
4.4
6.3
5.5
Exterior Components
Door components
(trim, systems)
Porch/stair
components
(includes columns,
rails)
Trim (soffits, fascia)
Walls/siding
Window sills
Other window
components
(trim, systems)
153
89
153
146
111
132
33.3
59.6
99.4
100.0
98.2
99.2
50.3
29.2
0.7
0.0
1.8
0.8
16.3
11.2
0.0
0.0
0.0
0.0
40.5
33.7
30.7
33.6
45.9
42.4
0.5
0.4
0.4
0.3
0.4
0.4
0.6
0.6
0.6
0.6
0.6
0.6
2.4
1.6
1.5
1.6
3.7
2.9
9.0
11.1
10.3
6.4
17.3
16.1
1 This table is a summary of observed XRF measurements within the 284 privately-owned housing units in the
HUD National Survey. Summaries are unweighted (i.e., do not reflect sampling weights associated with the
units).
2 Percentage of units with XRF data whose maximum XRF measurement is at least 1.0 mg/cm2
3 Percentage of units having both XRF data and data on the extent of paint deterioration, whose maximum
XRF measurement is at least 1.0 mg/cm2, and at least one of these measurements is from a component
containing some deteriorated paint.
3-42
-------
Table 3-12. Summary of the Distribution of Observed Maximum XRF Lead Levels in Paint
for Housing Units in the HUD National Survey, Weighted to Reflect the
Predicted 1997 Housing Stock.
Year Unit
Was Built2
#
Nations]
Survey
Units'
Obsorved Maximum XRF Paint-Lead Levels (rng/cm*)'
Geometric
Mem
Geometric
Standard
Deviation
5th
Percerrtfle
26th
PercentUo
Median
75th
PercentBe
95th
Percentlte
Interior of Unit
Before 1 940
1940-1959
1960-1979
72
83
116
1.86
1.02
0.712
3.78
2.42
1.79
0.300
0.400
0.300
0.600
0.600
0.500
1.45
0.800
0.600
6.10
1.70
0.900
11.5
7.30
2.50
Exterior of Unit
Before 1940
1940-1959
1960-1979
60
76
103
3.14
1.45
0.719
3.75
3.05
2.43
0.300
0.200
0.00
0.700
0.600
0.500
4.20
1.40
0.600
7.70
2.60
0.900
29.0
13.0
5.10
1 The statistics presented in this table were calculated on observed maximum XRF paint-lead level for National
Survey units across both interior and exterior painted surfaces (see Appendix CD. Each unit's observed
maximum XRF paint-lead level was weighted by the 1997 weight for the unit, which is presented in
Appendix C1.
2 No units built after 1979 were included in the HUD National Survey. In this risk analysis, these units are
assumed to be free of LBP.
3 Number of privately-owned units in the HUD National Survey in which an observed maximum XRF paint-
lead level was available (for either the interior or exterior).
Table 3-13. Predicted Numbers and Percentages of Units Having Lead-Based Paint in the
1997 Occupied Housing Stock, Based on Information from the HUD National
Survey.1
Year Unit Was Built
Before 1940
1940-1959
1960-1979
After 1979
All Housing
Number (%) of Units with
Lead-Based Paint
17,248,000(87.7%)
18,047,000(91.5%)
26,452,000 (75.6%)
0(0%)
61,747,000(62.2%)
Number (%) of Units with More Than
5 ft2 of Deteriorated Lead-Based Paint
7,755,000 (39.4%)
3,065,000(15.5%)
2,651,000(7.6%)
0 (0%)
13,470,000(13.6%)
1 A unit in the HUD National Survey is labeled as containing LBP if its predicted maximum XRF value in either
the interior or the exterior is greater than or equal to 1.0 mg/cm2. Results are weighted using the 1997
weights presented in Appendix C1.
3-43
-------
As the §403 dust-lead standards will be defined as lead loadings from dust collected using
wipe techniques (Section 1.1), it was necessary to express dust-lead loadings from the HUD
National Survey database as loadings based on wipe dust collection, even though a Blue Nozzle
vacuum collection technique was used in this survey. As a result, in this section and throughout
this document, dust-lead loading data based on Blue Nozzle vacuum techniques were converted
to wipe-equivalent dust-lead loadings prior to summarizing these data (e.g., Table C1-7 of
Appendix Cl, Table 3-6, Table 3-8). Methods used to perform these conversions, specially
developed for this risk analysis, are presented in Section 4.3 of Chapter 4.
For some HUD National Survey units, measurements were not reported for either dust-
lead loading, dust-lead concentration, or soil-lead concentration. In these situations, it was
necessary for modeling purposes to represent these units (and their associated sampling weights)
with some type of measurement. As discussed hi Section 1.3 of Appendix Cl, the value assigned
to a unit having a missing value for a particular data parameter equaled the average value across
units in the same category of year built and lead-based paint status (i.e., presence or absence of a
maximum XRF value in the ulterior or exterior at or above 1.0 mg/cm2). Table 3-14 presents
these average environmental-lead levels and numbers of National Survey housing units with a
missing value. The data summaries hi Table Cl-7 of Appendix Cl and in Tables 3-6 and 3-7
were calculated after replacing missing values with these imputed values.
Tables 3-6 through 3-9 indicate that the geometric means and medians for dust-lead
loadings and dust-lead concentrations decrease with the age of the unit. This finding is
consistent with the hypothesis that the potential for dust contamination by lead is higher in older
units, due to their propensity to contain greater amounts of lead-based paint and to be located in
older neighborhoods with lead-contaminated soil. Window sill dust-lead loadings and
concentrations in units built prior to 1940 were considerably higher than those of the other units.
These tables also indicate that lead loadings and concentrations tend to be higher on window sills
than floors, especially in older units. The same trends were observed in soil-lead concentration
(Table 3-10), whose geometric mean and median decreased with the age of the unit, and whose
levels were considerably higher hi pre-1940 units than in the other units.
The components tested for lead-based paint in the HUD National Survey were selected
based on a predetermined sample design. Therefore, the data summarized in Table 3-11 and 3-12
represent both lead-contaminated and lead-free painted surfaces. The primary components
sampled included ulterior doors, windows, walls, and ceilings, and exterior trim, doors, windows,
and siding. In general, XRF measurements were low among the tested surfaces (Table 3-11),
with those components containing lead-based paint in more than 25% of units limited to exterior
components and ulterior door, window, and stair components. In general, less than 10% of units
had deteriorated lead-based paint present on a particular component. One exception was for
exterior window surfaces, where from 16% to 17% of units had deteriorated lead-based paint.
The relationship between lead levels in paint and age of unit is strongest for the median and
upper percentiles (Table 3-12), indicating that while low paint-lead measurements are observed
hi all housing regardless of age, higher paint-lead measurements are more prevalent in older
units.
3-44
-------
Table 3-14. Imputed Environmental-Lead Measurements, by Age Category and Presence
of Lead-Based Paint1, and Numbers of Units in the HUD National Survey to
Which Imputed Measurements Were Assigned in the Risk Analyses.
Environmental-Lead Measurement
Floor Dust-Lead Loading fc/g/ft2)
(wipe)3
Window Sill Dust-Lead Loading (pg/ft2)
(wipe)3
Floor Dust-Lead Loading (//g/ft2)
(Blue Nozzle vacuum)4
Window Sill Dust-Lead Loading (//g/ft2)
(Blue Nozzle vacuum)4
Floor Dust-Lead Concentration d/g/g)
(Blue Nozzle Vacuum)5
Dripline Soil-Lead Concentration U/g/g)3-4-8
Remote Soil-Lead Concentration U/g/g)3-5
Imputed Measurement2
(Number of HUD National Survey units in which imputed
measurements were assigned)
Pre-1940 Units
LBP
Present
46.2
(1)
2300.
(6)
17.9
(1)
207.
(6)
-
1126.
(10)
555.
(13)
LBP Not
Present
-
-
-
-
-
453.
(3)
105.
(5)
1940-1959 Units
LBP
Present
--
309.
(7)
-
34.5
(7)
-
373.
(4)
253.
(8)
LBP Not
Present
-
17.4
(1)
-
3.73
(1)
-
45.8
(1)
32.8
(1)
1960-1979 Units
LBP
Present
7.67
(2)
217.
(21)
4.25
(2)
28.3
(21)
740.
(2)
84.1
(3)
42.7
(4)
LBP Not
Present
-
81.5
(4)
-
12.2
(4)
-
-
-
Post-1979
Units (LBP
Not Present)
-
83.
(4)
-
12.3
(4)
-
-
--
1 Units with lead-based paint have a predicted maximum XRF measurement (in the interior or the exterior)
greater than or equal to 1.0 mg/cm2.
2 For a given measurement type, imputed measurements were the average measurement across housing units
within the given category of year-built and presence of lead-based paint. Numbers in parentheses indicate
the number of units having no data for the given measurement type (i.e., the imputed measurement in the
given cell was used to represent the measurement for the number of units in parentheses). Cells containing
dashes had no housing units in the given category needing to have measurements imputed for the given
measurement type. The numbers of housing units entering into each imputed value can be determined from
Table 3-5 (second column) minus the numbers in parentheses in this table.
3 This measurement was used to determine whether units exceeded example standards (soil interventions
were triggered by the average of dripline and remote soil-lead concentrations).
4 This measurement was used as input to the empirical model (Section 4.2) to obtain a distribution of
predicted blood-lead concentrations from environmental-lead measurements.
5 This measurement was used as input to the IEUBK model (Section 4.1) to obtain a distribution of predicted
blood-lead concentrations from environmental-lead measurements (dripline and remote soil-lead
concentrations were averaged prior to input to the IEUBK model).
This risk analysis predicts that approximately 62% of the 1997 occupied housing stock
contain lead-based paint (Table 3-13), based on information from the HUD National Survey and
under the assumption that no units built after 1979 contain lead-based paint. This percentage is
less than 83%, the percentage of pre-1980 occupied housing predicted to contain lead-based paint
according to the HUD National Survey (USEPA, 1995a). The estimate of 62% is relative to all
occupied housing, even units built after 1979. The percentages of units with lead-based paint
3-45
-------
within the three pre-1980 year-built categories match those reported in the National Survey report
(USEPA, 1995a). Table 3-13 also indicates that approximately 14% of units are predicted to
contain more than five square feet of deteriorated lead-based paint, with over half of these units
built prior to 1940. Where only housing built prior to 1980 is considered, this percentage
increases to 18% (Tables 3-13 and 3-5).
3.3.1.2 The Baltimore Repair and Maintenance (R&M) Study
In the Baltimore R&M Study (Section 3.2.2.1), the BRM vacuum sampler was used to
collect dust samples. The BRM dust-lead loadings were converted to wipe equivalent dust-lead
loadings using the conversion equations presented in Section 4.3. Tables 3-15 and 3-16
summarize pre-intervention lead loadings and concentrations, respectively, in floor-dust samples
across study units that were occupied prior to intervention. Tables 3-17 and 3-18 summarize lead
loadings and concentrations, respectively, hi window sill-dust samples. Table 3-19 summarizes
lead concentrations hi soil samples taken at the dripline. Table 3-20 presents summaries of the
observed maximum XRF paint-lead measurement within study units slated for R&M
interventions, for the interior only and the exterior only, as well as for the entire unit. Recall that
XRF measurements were not made hi the previously abated and modern urban homes.
Tables 3-15 through 3-18 indicate that geometric mean dust-lead levels are highest for
units slated for R&M intervention, while modern urban units have geometric mean levels that are
as much as an order of magnitude lower than the other two housing groups. However, units
slated for R&M interventions should not be considered representative of occupied inner city
homes. As many of these units were in poor condition prior to the interventions, they represent a
worst case setting for residential environmental-lead levels. Dust-lead loadings for previously-
abated units and units slated for R&M interventions are considerably higher than those reported
for pre-1940 housing in the HUD National Survey. Units slated for R&M interventions have
very high dust-lead concentrations and window sill dust-lead loadings, due to the deteriorated
condition of most of these units. Modern urban units have dust-lead levels that are slightly
higher than the HUD National Survey units built from 1960-1979 and containing no LBP.
Soil-lead concentrations summarized hi Table 3-19 are based on small numbers of units,
due to the lack of available dripline soil to sample at many of the study units. Geometric mean
soil-lead concentrations presented hi Table 3-19 are high compared to those hi the HUD National
Survey.
The paint-lead measurements summarized in Table 3-20 are extremely high, as the data
represent only units slated for R&M interventions. These units were likely to contain large
amounts of lead-based paint, and paint-lead measurements were taken primarily from
components suspected of containing LBP to identify and prioritize surfaces requiring LBP
intervention. Thus, the data summarized hi Table 3-20 reflect a LBP-contaminated environment
and are not typical of all painted surfaces in occupied housing.
3-46
-------
Table 3-15. Summary of Average Pre-lntervention Floor Dust-Lead Loading for Occupied
Housing Units in the Baltimore R&M Study.
Unit Category
All Study Units
Previously Abated Units
Units Slated for R&M
Intervention
Modern Urban Units
# Units
90
16
58
16
Floor Dust-Lead Loading (//g/ft2)1
Geometric
Mean
40.9
45.6
58.6
10.0
Geometric
Standard
Deviation
2.3
1.6
1.7
1.6
Minimum
4.5
23.1
22.0
4.5
Maximum
266
125
266
17.4
1 The statistics presented in this table were calculated on area-weighted arithmetic mean dust-lead loadings
from floors for each unit. These loadings have been converted to represent loadings from dust samples
obtained from wipe collection techniques (see Section 4.3).
Table 3-16. Summary of Average Pre-lntervention Floor Dust-Lead Concentrations for
Occupied Housing Units in the Baltimore R&M Study.
Unit Category
All Study Units
Previously Abated Units
Units Slated for R&M
Intervention
Modern Urban Units
# Units
90
16
58
16
Floor Dust-Lead Concentration (//g/g)1
Geometric
Mean
1300
1210
2440
145
Geometric
Standard
Deviation
4.1
2.5
2.6
2.2
Minimum
48.9
332
426
48.9
Maximum
60300
7360
60300
704
1 The statistics presented in this table were calculated on mass-weighted arithmetic mean dust-lead
concentrations from floors for each unit, where dust was sampled using the BRM vacuum method.
Table 3-17. Summary of Average Pre-lntervention Window Sill Dust-Lead Loading for
Occupied Housing Units in the Baltimore R&M Study.
Unit Category
All Study Units
Previously Abated Units
Units Slated for R&M
Intervention
Modern Urban Units
# Units
90
16
58
16
Window Sill Dust-Lead Loading {pg/ft*)1
Geometric
Mean
326
158
675
48.2
Geometric
Standard
Deviation
3.2
2.2
1.6
1.6
Minimum
23.0
46.6
203
23.0
Maximum
1880
833
1880
86.7
1 The statistics presented in this table were calculated on area-weighted arithmetic mean dust-lead loadings
from window sills for each unit. These loadings have been converted to represent loadings from dust
samples obtained from wipe collection techniques (see Section 4.3).
3-47
-------
Table 3-18. Summary of Average Pre-lntervention Window Sill Dust-Lead Concentrations
for Occupied Housing Units in the Baltimore R&M Study.
Unit Category
All Study Units
Previously Abated Units
Units Slated for R&M Intervention
Modern Urban Units
#
Units
90
16
58
16
Window SHI Dust-Lead Concentration (//g/g)1
Geometric
Mean
5600
1880
20100
161
Geometric
Standard
Deviation
8.5
4.6
2.4
2.7
Minimum
7.2
255
28010
7.2
Maximum
141000
31500
141000
447
1 The statistics presented in this table were calculated on mass-weighted arithmetic mean dust-lead
concentrations from window sills for each unit, where dust was sampled using the BRM vacuum method.
Table 3-19. Summary of Average Pre-lntervention Dripline Soil-Lead Concentrations for
Occupied Housing Units in the Baltimore R&M Study.
Unit Category
All Study Units
Previously Abated Units
Units Slated for R&M
Intervention
Modern Urban Units
# Units
28
2
16
10
Soil-Lead Concentration (//g/g)
Geometric
Mean
445
219
1260
61.1
Geometric
Standard Deviation
5.1
1.6
2.0
1.7
Minimum
28.9
1570
336
28.9
Maximum
3540
3060
354
153.7
Table 3-20. Summary of Observed Maximum XRF Paint-Lead Measurement at Pre-
lntervention for Occupied Housing Units Slated for R&M Intervention in the
Baltimore R&M Study.1
Location Within a Unit
Entire Unit
Exterior Only
Interior Only
# Units
36
35
36
Observed Maximum XRF Paint-Lead Measurement (mg/cm2)
Geometric Mean
38.4
24.8
28.2
Geometric Standard
Deviation
1.7
2.6
1.8
Minimum
9.3
0.6
7.4
Maximum
98.1
86.3
98.1
1 XRF data were not available for previously-abated units and modern urban units in the study.
3-48
-------
3.3.1.3 The Rochester Lead-in-Dust Study
Like the previous studies in this section, the Rochester Lead-in-Dust Study database
contains a variable for the year in which the surveyed housing unit was built. As a result, this
section presents summaries of environmental-lead levels across all surveyed units, as well as for
each of the four age categories considered hi the HUD National Survey. However, Rochester
study representatives later informed the risk analysis team that the specified year was
occasionally determined from public tax assessor records and may not always reflect the actual
year of construction. For units experiencing a major renovation effort, the specified year may be
the year of the latest such effort, and not the original construction year. Therefore, conclusions
on environmental-lead levels for specific age categories should be made cautiously, especially for
units categorized as "after 1979," as the actual year of construction for some units may be earlier
than that specified.
Across study units, Tables 3-21 and 3-22 summarize lead loadings and concentrations,
respectively, in floor-dust samples. Tables 3-23 and 3-24 summarize lead loadings and
concentrations, respectively, in window sill-dust samples. In these four tables, dust-lead loadings
were summarized for only those samples collected using wipe techniques, while dust-lead
concentrations were summarized for only those samples collected using BRM vacuum
techniques. Table 3-25 summarizes lead concentrations hi soil samples taken at the dripline
where bare soil was present, while Table 3-26 summarizes lead concentrations in soil samples
taken at the child's principal play area hi the yard, where bare soil was present. Table 3-27
presents information on the extent of XRF sampling, the presence of lead-based paint, the
presence of deteriorated lead-based paint, and the distribution of maximum XRF measurements
hi a unit, for specific categories of ulterior and exterior painted components. Table 3-28 presents
summaries of the observed maximum XRF paint-lead measurements within the study units, for
the ulterior only and the exterior only, as well as for the entire unit.
Table 3-21. Summary of Average Floor Dust-Lead Loading for Housing Units in the
Rochester Study.
Year Category1
All Units
Before 1 940
1940-1959
1960-1979
After 1979
# Units
205
172
19
4
10
Floor Dust-Lead Loading (pg/ft1)*
Geometric
Mean
17.7
19.8
8.4
7.8
15.0
Geometric Standard
Deviation
3.2
3.2
2.6
2.4
3.3
Minimum
1.2
1.7
1.2
2.1
3.5
Maximum
8660
8660
26.9
13.2
250
1 Reflects age of housing unit as specified in public tax assessor records.
2 The statistics presented in this table were calculated on area-weighted arithmetic mean dust-lead loadings
from floors for each unit. Results included in the summaries are only for dust samples collected using wipe
techniques.
3-49
-------
Table 3-22. Summary of Average Floor Dust-Lead Concentrations for Housing Units in the
Rochester Study.
Year Category1
All Units
Before 1 940
1940-1959
1960-1979
After 1979
# Units
204
172
18
4
10
Floor Dust-Lead Concentration U/g/g)2
Geometric
Mean
351
396
209
60.8
226
Geometric Standard
Deviation
3.7
3.6
4.6
2.7
3.0
Minimum
8.3
8.3
16.5
16.9
57.0
Maximum
40700
40700
7900
164
1120
1 Reflects age of housing unit as specified in public tax assessor records.
2 The statistics presented in this table were calculated on mass-weighted arithmetic mean dust-lead
concentrations from floors for each unit. Results included in the summaries are only for dust samples
collected using BRM vacuum techniques.
Table 3-23. Summary of Average Window Sill Dust-Lead Loading for Housing Units in the
Rochester Study.
Year Category1
All Units
Before 1940
1940-1959
1960-1979
After 1979
# Units
196
164
18
4
10
Window Sill Dust-Lead Loading (pg/ft*)2
Geometric
Mean
196
234
72.0
52.3
113
Geometric
Standard Deviation
4.0
3.7
6.2
1.4
1.9
Minimum
2.8
2.8
2.8
36.2
26.9
Maximum
14900
14900
439
70.7
320
1 Reflects age of housing unit as specified in public tax assessor records.
2 The statistics presented in this table were calculated on area-weighted arithmetic mean dust-lead loadings
from window sills for each unit. Results included in the summaries are only for dust samples collected using
wipe techniques.
Table 3-24. Summary of Average Window Sill Dust-Lead Concentrations for Housing
Units in the Rochester Study.
Year Category1
All Units
Before 1940
1940-1959
1960-1979
After 1979
# Units
199
166
19
4
10
Window Sill Dust-Lead Concentration U/g/g)2
Geometric
Mean
2700
3860
497
473
674
Geometric Standard
Deviation
8.4
7.3
9.9
2.9
8.6
Minimum
3.1
15.8
5.3
160
3.1
Maximum
368000
368000
15000
1900
8630
1 Reflects age of housing unit as specified in public tax assessor records.
2 The statistics presented in this table were calculated on mass-weighted arithmetic mean dust-lead
concentrations from window sills for each unit. Results included in the summaries are only for dust samples
collected using BRM vacuum techniques.
3-50
-------
Table 3-25. Summary of Average Dripline Soil-Lead Concentrations for Housing Units in
the Rochester Study.
Year Category1
All Units
Before 1940
1940-1959
1960-1979
After 1979
# Units
186
158
14
4
10
Dripline Soil-Lead Concentration ifjglg}
Geometric
Mean
731
938
291
66.4
135
Geometric Standard
Deviation
3.7
3.2
3.3
1.8
3.1
Minimum
12.3
12.3
29.7
29.0
26.0
Maximum
21000
21000
1790
111
876
1 Reflects age of housing unit as specified in public tax assessor records.
Note: Data in this table reflect bare soil areas along the unit's dripline.
Table 3-26. Summary of Average Soil-Lead Concentrations from Play Areas for Housing
Units in the Rochester Study.
Year Category1
All Units
Before 1940
1940-1959
1960-1979
After 1979
# Units
87
79
6
1
1
Play Area Soil-Lead Concentration ipglg]
Geometric
Mean
267
278
185
138
215
Geometric Standard
Deviation
2.8
2.8
3.1
Minimum
28.0
28.0
55.4
138
215
Maximum
7300
7300
767
138
215
1 Reflects age of housing unit as specified in public tax assessor records.
Note: Data in this table reflect bare soil areas in the child's principal play area in the yard.
3-51
-------
Table 3-27. Summary of XRF Paint Measurements Taken in the Rochester Lead-in-Dust
Study, Including the Percentage of Housing Units with Lead-Based Paint
(LBP) and Deteriorated LBP, by Component Category.
Component Category1
Baseboards/trim
Door jambs
Door surfaces
Floors
Walls
Window sashes
Window sills
Window wells
# Units
with
XRF
Data
Percent of Units With a
Given # of
Measurements Reported
1
2
>2
Percent of
Units with
LBP'
Percentfles of the Distribution of
Maximum XRF Measurement in a
Unit (mg/cm2)
25th
Percenttte
50th
Percentfle
Interior Components
20
196
164
81
14
165
195
124
60.0
11.2
50.6
76.5
85.7
30.9
21.5
40.3
30.0
88.8
49.3
23.5
7.1
69.1
74.4
58.9
10.0
0.0
0.0
0.0
7.1
0.0
4.1
0.8
40.0
48.5
19.5
19.8
28.6
70.9
66.7
91.9
0.5
0.5
0.5
0.5
0.5
0.7
0.7
8.3
0.8
0.9
0.5
0.5
0.5
2.1
1.7
17.1
75th
Percerrtlle
Percent of
Units with
Deteriorated
LBP*
2.4
5.4
0.7
0.6
3.5
6.4
4.4
25.5
30.0
13.3
6.7
16.0
28.6
41.8
35.4
74.2
Exterior Components
Door jambs
Door surfaces
Porch floors
Trim
Walls/siding
200
193
112
39
95
100.0
100.0
92.9
79.5
93.7
0.0
0.0
6.3
7.7
6.3
0.0
0.0
0.9
12.8
0.0
63.5
22.3
58.9
84.6
83.2
0.5
0.5
0.5
1.9
2.5
6.5
0.5
2.3
6.9
10.7
26.2
0.8
8.0
17.1
24.6
22.0
7.8
47.3
79.5
22.1
1 Percentage of units with XRF data whose maximum XRF measurement is at least 1.0 mg/cm2.
2 Percentage of units having both XRF data, whose maximum XRF measurement is at least 1.0 mg/cm2, and
at least one of these measurements is from a component containing some paint in "average" or "poor"
condition.
3-52
-------
Table 3-28. Summary of Observed Maximum XRF Paint-Lead Concentration for Housing
Units in the Rochester Study.
Year Category1
#
Units
%of
Units
with LBP2
Observed Maximum XRF Paint-Lead Levels {mg/cm*)
Geometric
Mean
Geometric Standard
Deviation
Minimum
Maximum
Entire Unit (interior and exterior)
All Units
Before 1940
1940-1959
1960-1979
After 1 979
205
172
19
4
10
89%
95%
68%
50%
40%
12.8
16.6
5.50
1.04
1.93
3.88
3.05
4.97
1.89
5.92
0.50
0.50
0.50
0.57
0.50
59.8
59.8
37.6
1.93
39.4
Interior of Unit
All Units
Before 1940
1940-1959
1960-1979
After 1979
205
172
19
4
10
83%
91%
63%
0%
20%
7.58
9.90
2.86
0.61
1.32
4.38
3.67
4.71
1.12
5.72
0.50
0.50
0.50
0.57
0.50
57.6
57.6
32.9
0.73
39.4
Exterior of Unit
All Units
Before 1 940
1940-1959
1960-1979
After 1 979
204
171
19
4
10
79%
84%
63%
50%
40%
8.14
10.3
3.47
1.00
1.63
4.91
4.40
5.22
1.97
4.95
0.50
0.50
0.50
0.50
0.50
59.8
59.8
37.6
1.93
37.4
1 Reflects age of housing unit as specified in public tax assessor records.
2 A unit is assumed to contain LBP if its maximum XRF value is greater than or equal to 1.0 mg/cm3
Note from Table 3-21 that at least 84% of the study units were built prior to 1940.
Therefore, while the Rochester study considers units in an urban environment and does not
attempt to target a particular lead exposure environment in recruiting the units, most of the units
are older units. In addition, most units contain families with low income levels.
As was also seen in the HUD National Survey, dust-lead loadings and concentrations are
highest among the units in the "before 1940" category (Tables 3-21 through 3-24). For units not
in the "after 1979" category, the geometric mean floor dust-lead levels were often lower in the
Rochester study than in the HUD National Survey. While the ten units in the "after 1979"
category had higher geometric mean dust-lead levels than for the 1940-1959 and 1960-1979
categories, some of these units may have been built prior to 1980 and renovated after 1979.
Geometric mean soil-lead concentrations were higher than those observed in the HUD
National Survey. Less than half of the units had soil samples taken from play areas (Table 3-26),
where geometric mean concentrations were generally lower than at the dripline for older units.
3-53
-------
The paint testing protocol for the Rochester study indicated that 10 to 15 surfaces were
tested in each housing unit and that the selected surfaces either contained paint in the greatest
state of deterioration or were most accessible to children. As indicated in Table 3-27, the tested
surfaces were primarily door and window components, as well as painted floors, trim, and
exterior siding. Typically, no more than two measurements were taken of each type of
component within each unit. As seen in Table 3-27, the majority of units had some lead-based
paint on exterior surfaces and interior window components. Interior door components and trim
also had high incidences of lead-based paint. Window wells and exterior trim contained
deteriorated lead-based paint in at least three-fourths of the units, while at least 40% of the units
had deteriorated lead-based paint on window sashes and exterior porch floors.
Table 3-28 indicates that units built prior to 1940 had the highest geometric mean paint-
lead measurements and the highest percentage of units with lead-based paint in the study. While
at least 40% of the units within each year category had XRF lead measurements at or above
1.0 mg/cm2 (i.e., the criterion for determining the presence of lead-based paint), caution must be
taken when interpreting this result. One should consider the following when making any
conclusion from the XRF paint-lead measurement summary in Table 3-28:
As described above, the year category may not represent the actual construction year
for some housing units, especially units with recent specified years.
The Microlead I XRF instrument was used in this study without substrate correction.
A study of in situ XRF instrument performance (USEPA, 1995J) found that
measurements by this instrument tend to be positively biased, especially when not
corrected for type of substrate.
3.3.1.4 Evaluation of the HUD Lead-Based Paint Hazard Control Grant Program ("HUD
Grantees")
The Department of Environmental Health at the University of Cincinnati (UCDEH)
provided pre-intervention data collected in the HUD Grantees evaluation program from February,
1994, to September, 1997. These data are considered preliminary, as data continue to be
collected in this program.
An area-weighted arithmetic average dust-lead loading (under wipe collection techniques)
was calculated for floors and window sills within each housing unit. Only dust-lead loadings
from targeted rooms were considered (entryways, play rooms, kitchens, and up to two
bedrooms). Tables 3-29 and 3-30 summarize these area-weighted averages across housing units,
for floors and window sills, respectively. Results are summarized across all housing units, for
units in each of four age categories, and by grantee. Note that observed differences across
grantees are not simply due to geographical variation, but also to other differences among the
grantees, such as differences hi the recruitment process and in dust sampling and analysis
procedures.
3-54
-------
For both Tables 3-29 and 3-30, the geometric mean dust-lead loading in pre-1940 housing
units was higher than that for units built from 1940-1959 and from 1960-1977. (As seen in Table
3-3a, nearly 90% of housing units in the program were built prior to 1940, so considerably more
data are available to characterize lead levels in this category of units). For pre-1978 units,
geometric means were generally higher than in the Rochester study and the HUD National
Survey, but lower than in the Baltimore R&M study. Primarily the result of the different
grantees, the geometric standard deviations associated with the data for a given age category of
housing units are higher than observed in the Baltimore R&M and the Rochester Lead-in-Dust
studies.
In Tables 3-29 and 3-30, it is uncertain why average dust-lead loadings are so high in the
few units built after 1977. One likely reason is cross-contamination from other lead-
contaminated housing units in the same neighborhood. In addition, these units were likely
targeted for lead abatement activities, which contributed to the magnitude of their dust-lead
loadings. Three of the four post-1977 units were recruited by the Rhode Island grantee, and the
other by the Minnesota grantee.
Tables 3-31 and 3-32 summarize soil-lead concentrations measured in the HUD Grantees
program at play areas and at the dripline, respectively. Results are presented hi the same manner
as hi Tables 3-29 and 3-30. Note that since soil sampling was considered optional in the HUD
Grantees program, soil-lead concentration data were available for less than 15% of the enrolled
housing units, and for only eight of the 14 grantees. Only one unit with soil data was specified as
being built after 1959, although some units had no age information specified. While each
housing unit represented in Tables 3-31 and 3-32 had a single soil-lead concentration
measurement for the specified yard location, the analyzed soil sample was a composite of 5-10
subsamples.
Play area soil-lead concentrations had a geometric mean of 415 ug/g across 330 units
(Table 3-31), while dripline soil-lead concentrations had a geometric mean of 1180 ug/g across
557 units (Table 3-32). These geometric means were somewhat higher than those reported in the
HUD National Survey, the Baltimore R&M study, and the Rochester study. While the geometric
means for play area soil were similar between pre-1940 units and units built from 1940-1959, the
geometric mean dripline soil-lead concentration for pre-1940 units was over twice that of units
built from 1940-1959 (Table 3-32).
Table 3-33 presents information on the extent of XRF sampling, presence of lead-based
paint, presence of deteriorated lead-based paint, and distribution of maximum XRF
measurements in a unit, for the components most commonly tested. A variety of components
were tested, with ulterior component testing occurring for more housing units than exterior
component testing, and with greater than two measurements taken within each component
classification in most housing units. The XRF measurements were higher than those observed in
the HUD National Survey and the Rochester study, with over 70% of units having lead-based
3-55
-------
paint on all component types except interior cabinets, ceilings, floors, and stairs. In a majority of
housing units, the lead-based paint on these components was in a deteriorated state. For specific
door and window components, as well as exterior porch and stair components, the maximum
XRF measurement exceeded 19 mg/cm2 in at least 25% of the housing units. Thus, lead-based
paint was quite prevalent, and lead in the paint was at high levels, among the HUD Grantee units.
Table 3-29. Summary of Area-Weighted Average Floor Dust-Lead Loadings (Pre-
Intervention, Using Wipe Collection Techniques) for Housing Units In the
HUD Grantee Program, According to Age of Unit and Grantee.1
AH Units
Units Built Prior to 1940
Unit* Built 1940-1959
Units Built 1960-1977
Units BuiH After 1977
Alameda County Grantee
Baltimore Grantee
Boston Grantee
California Grantee
Cleveland Grantee
Massachusetts Grantee
Minnesota Grantee
New Jersey Grantee
Rhode Island Grantee
Wisconsin Grantee
Milwaukee Grantee
Chicago Grantee
New York City Grantee
Vermont Grantee
#of
Units
with
Data
2826
2260
211
35
4
177
405
99
86
105
221
208
57
199
221
294
119
277
358
Area-Weighted Average Floor Dust-Lead Loadings (//g/ft2)
Geometric
Mean
55.4
52.3
31.9
15.5
372
24.6
164
68.9
22.1
91.2
70.0
38.2
54.8
59.5
26.9
31.3
28.8
49.4
86.5
Geometric
Standard
Deviation
5.3
5.1
5.2
2.1
5.6
4.0
4.1
4.2
4.6
5.6
5.5
3.2
5.4
4.6
5.8
4.2
3.5
6.4
4.7
Minimum
0.088
0.088
2.12
4.03
32.5
3.54
17.7
5.00
3.54
3.54
1.64
14.1
3.54
5.22
1.77
2.06
3.54
0.088
7.07
25th
Percentile
17.7
17.7
8.43
14.1
120
7.52
50.1
21.2
7.01
25.4
19.2
17.7
14.1
17.8
6.49
11.0
10.5
18.8
22.5
Median
40.5
38.1
22.3
14.1
580
18.0
166
57.5
13.1
84.0
64.5
22.8
29.0
53.6
16.5
26.5
28.9
34.3
51.6
75th
Percentile
160
145
75.8
17.2
1160
65.8
456
170
47.9
280
210
61.5
247
157
78.1
73.8
62.8
95.3
195
Maximum
26400
19600
26400
247
1750
1230
7470
2490
2220
10900
16600
5060
4250
2260
2780
5810
26400
19600
15600
1 Summaries include data collected through September, 1997. Only data for wipe dust samples collected in
entry ways, play rooms, kitchens, and two targeted bedrooms were considered. Eleven sample results
labeled as statistical outliers were not included in calculating area-weighted averages at the housing unit
level.
3-56
-------
Table 3-30. Summary of Area-Weighted Average Window Sill Dust-Lead Loadings (Pre-
Intervention, Using Wipe Collection Techniques) for Housing Units In the
HUD Grantee Program, According to Age of Unit and Grantee.1
All Units
Units Built Prior to 1940
Units Built 1940-1969
Units Built 1960-1977
Units Built After 1977
Alameda County Grantee
Baltimore Qrantee
Boston Grantee
California Grantee
Cleveland Grantee
Massachusetts Grantee
Minnesota Grantee
New Jersey Grantee
Rhode Island Grantee
Wisconsin Grantee
Milwaukee Grantee
Chicago Grantee
New York City Grantee
Vermont Grantee
#of
Units
with
Data
2669
2220
201
35
2
177
404
87
77
97
198
189
57
189
218
273
117
263
323
Area-Weighted Average Window Sill Dust-Lead Loadings (i/g/ft2)
Geometric
Mean
475
505
260
61.0
2380
150
1910
474
333
757
407
396
142
740
341
546
456
278
306
Geometric
Standard
Deviation
6.7
6.7
6.1
2.2
4.5
4.7
5.1
6.5
4.9
5.0
6.0
5.6
5.2
6.4
6.2
6.6
6.0
5.7
7.5
Minimum
0.32
0.32
2.79
22.9
816
7.41
36.3
12.0
19.8
45.8
2.60
23.5
10.3
8.75
3.54
19.1
11.3
0.32
10.6
25th
Percentile
115
126
63.6
33.7
816
42.1
597
138
93.4
261
133
88.3
39.4
201
93.7
127
127
97.6
63.6
Median
403
428
208
45.6
2380
131
2130
419
268
677
368
339
84.4
655
294
413
475
184
214
75th
Percentile
1730
1810
1000
105
6940
440
5850
1650
1040
1610
1260
1230
425
2120
1250
2110
1200
708
1050
Maximum
1 32000
1 32000
29400
289
6940
19100
1 30000
47200
10200
54400
76100
100000
8130
1 32000
73900
53600
36500
57100
98100
1 Summaries include data collected through September, 1997. Only data for wipe dust samples collected in
entry ways, play rooms, kitchens, and two targeted bedrooms were considered. Two sample results labeled
as statistical outliers were not included in calculating area-weighted averages at the housing unit level.
3-57
-------
Table 3-31. Summary of Play Area Soil-Lead Concentrations (Pre-lntervention) for
Housing Units In the HUD Grantee Program, According to Age of Unit and
Grantee.1
All Units
Units Built Prior to 1940
Units Built 1940-1959
Units Built 1960-1 977
Units Built After 1977
Alameda County Grantee
California Grantee
Cleveland Grantee
Minnesota Grantee
Rhode island Grantee
Wisconsin Grantee
Milwaukee Grantee
Vermont Grantee
All Other Grantees
#of
Units
with
Data
330
190
12
0
1
69
8
99
44
41
38
11
20
0
Play Area Soil-Lead Concentrations (pg/g)
Geometric
Mean
415
357
223
481
483
271
633
469
457
132
690
151
Geometric
Standard
Deviation
4.5
4.9
5.5
2.7
3.5
2.6
2.4
4.8
16.3
1.8
5.1
Minimum
0.005
0.005
5.00
481
41.0
22.0
43.0
54.0
5.00
0.005
407
3.80
25th
Percentile
259
210
80.6
.
481
241
148
390
284
370
100
442
49.5
Median
480
456
402
481
501
462
589
464
621
300
538
140
75th
Percentile
870
800
723
481
1010
585
1010
687
1060
572
1300
547
Maximum
12000
12000
2630
481
4170
830
12000
4100
5210
2100
2050
3730
1 Summaries include data collected through September, 1997. One play area soil-lead concentration was
reported for each housing unit represented in this table.
3-58
-------
Table 3-32. Summary of Dripline Soil-Lead Concentrations (Pre-lntervention) for Housing
Units In the HUD Grantee Program, According to Age of Unit and Grantee.1
AB Unit*
Units Built Prior to 1940
Units Built 1940-1959
Units Built 1960-1977
Units BuUt After 1977
Alameda County Grantee
California Grantee
Cleveland Grantee
Minnesota Grantee
Rhode Island Grantee
Wisconsin Grantee
Milwaukee Grantee
Vermont Grantee
All Other Grantees
#of
Unite
with
Data
557
266
17
0
1
97
8
99
44
60
66
12
171
0
Dripline Soft-lead Concentrations (ftglg)
Geometric
Mean
1180
1030
478
.
330
776
331
2380
593
1390
564
1970
1540
Geometric
Standard
Deviation
3.7
3.9
3.2
2.7
1.9
2.3
2.7
3.3
7.3
2.5
3.3
Minimum
0.07
0.07
66.0
.
330
30.0
94.0
420
45.0
66.0
0.007
327
25.0
25th
Percentile
557
534
174
330
395
269
1350
280
638
400
1160
692
Median
1250
1080
530
330
710
360
2140
559
1500
859
1690
1500
75th
Percentile
2580
2150
925
330
1390
456
4520
1150
2780
1500
3490
3380
Maximum
52700
50600
5390
330
21100
780
16400
8120
26200
5730
10300
52700
1 Summaries include data collected through September, 1997. One dripline soil-lead concentration was
reported for each housing unit represented in this table.
3-59
-------
Table 3-33. Summary of XRF Paint Measurements Taken in the HUD Grantee Program,
Including the Percentage of Housing Units with Lead-Based Paint (LBP) and
Deteriorated LBP, by Component Category.1
Component Category
# Unit*
with
XRF
Data
Percent of Units With a
Given* of
Measurements Reported
1
2
>2
Percent of
Units with
LBP2
Percentiles of the Distribution of
Maximum XRF Measurement in a
Unit (mg/cm2)
25th
Percentile
50th
Percentile
75th
Percentile
Percent of
Units with
Deteriorated
LBP3
Interior Components
Cabinets
Ceiling
Doors/transoms -
friction surfaces
(jamb, threshold)
Doors/transoms --
other surfaces (door
surface, casing,
transom sash)
Floor
Stairs
Trim (baseboards,
chair rails, crown
molding)
Walls
Window sills
Window wells
Other window
components
(casing/apron,
jamb/track, sash)
2139
2841
2953
3348
2195
439
3152
3371
3085
1958
3208
22.3
7.50
12.7
1.76
29.5
42.6
4.79
2.46
5.80
15.7
2.37
28.6
5.17
9.11
2.78
15.8
27.3
6.00
2.52
6.13
12.3
3.58
49.0
87.3
78.2
95.5
54.7
30.1
89.2
95.0
88.1
72.0
94.0
39.6
55.0
80.9
84.5
35.6
55.1
75.1
77.7
79.2
94.5
91.9
0.100
0.300
1.70
2.20
0.100
0.200
1.00
1.20
1.40
6.10
5.30
0.500
1.60
9.90
10.0
0.300
1.70
6.60
7.60
6.40
10.0
10.0
3.60
9.90
16.4
19.3
2.40
10.0
14.2
15.1
12.8
24.3
22.0
19.2
36.6
63.4
65.0
28.5
40.3
47.1
52.3
64.0
88.9
79.3
Exterior Components
Ceiling
Doors/transoms -
friction surfaces
(jamb, threshold)
Door surfaces
Porch/stair
components
Trim (including
gutters, downspouts,
soffits, fascias)
Walls/siding
Window wells
Other window
components (sashes,
casing)
640
1308
1602
1282
1021
1417
631
980
74.5
23.2
13.0
18.1
46.7
36.6
53.1
40.9
18.1
25.5
17.4
14.9
31.1
33.0
28.5
28.6
7.34
51.3
69.5
67.0
22.1
30.5
18.4
30.5
76.7
89.8
87.7
70.7
77.6
75.9
89.4
95.3
1.65
6.35
4.60
0.500
1.70
1.10
3.10
6.15
9.90
12.4
10.0
7.10
9.60
7.80
9.90
10.0
18.5
25.6
22.4
15.9
14.3
14.9
15.2
21.3
61.7
81.3
78.3
63.8
66.9
63.8
76.4
83.3
1 Summaries include data collected through September, 1997.
2 Percentage of units with XRF data whose maximum XRF measurement is at least 1.0 mg/cm2.
3 Percentage of units having both XRF data, whose maximum XRF measurement is at least 1.0 mg/cm2, and
at least one of these measurements is from a component containing some paint in "fair" or "poor" condition.
3-60
-------
3.3.1.5 Evaluating the Approach to Representing the Post-1979 Housing Stock
Recall from Section 3.3.1.1 that the HUD National Survey did not include housing units
built after 1979. As a result, this risk analysis used the 28 surveyed housing units built from
1960 to 1979 and not containing lead-based paint to represent the national housing stock built
after 1979 (these units also were included among those housing units representing the national
housing stock built from 1960-1979). Under this approach, results in Tables 3-6 through 3-10
and Table 3-13 represent a summary of environmental-lead levels for the entire 1997 occupied
national housing stock. In order to evaluate the assumption that using HUD National Survey
units constructed between 1960 and 1979 with no lead-based paint accurately represents post-
1979 housing units, Table 3-34 presents geometric means of environmental lead levels for
modern urban units in the Baltimore R&M study, which were built after 1979, along with
environmental lead levels for HUD National Survey units with no lead-based paint and
constructed between 1960 and 1979. Geometric means presented in this table suggest that the
environmental-lead levels used to represent the post-1979 housing stock in this risk analysis are
similar to or lower than estimates for post-1979 housing from the Baltimore R&M study. When
interpreting the results in Table 3-34, one must consider the different dust sampling methods
used in the two studies, along with the different conversion factors used to obtain wipe-
equivalent dust-lead loadings.
Table 3-34. Estimates of Geometric Mean Environmental-Lead Levels for HUD National
Survey Units Representing Post-1979 Housing in this Risk Assessment and
for Modern Urban Units in the Baltimore R&M Study.
Study
(Subset of housing
units considered)
HUD National Survey
(1960-1979 units with
no lead-based paint)3
Baltimore R&M Study
(Modern urban units)
#of
Housing
Units
28
164
Lead Concentrations1 U/g/g)
Boor Dust
91.9
145
Window Sill
Dust
239
161
Oripline
Soil
27.4
61.1
Lead Loadings fc/g/ft")
assuming wipe technique*2
Floor Dust
3.14
10.0
Window Sill
Dust
8.17
48.2
1 Dust-lead concentrations in this table are geometric means (across housing units) of the mass-weighted
mean concentrations for each housing unit. Concentrations for HUD National Survey units reflect Blue
Nozzle vacuum methods, while concentrations for the Baltimore R&M Study reflect BRM vacuum methods.
Soil-lead concentrations are geometric means (across housing units) of mean dripline soil-lead concentrations
for each housing unit.
2 Dust-lead loadings in the HUD National Survey and the Baltimore R&M Study were converted to wipe-
equivalent dust-lead loadings according to type of surface and dust collection method. Table entries are
geometric means (across housing units) of the area-weighted mean dust-lead loadings for each housing unit.
3 Environmental-lead levels in these units also represent levels in post-1979 housing in this risk assessment.
4 Only 10 housing units had data for dripline soil-lead concentration.
3-61
-------
3.3.2 Characterizing the Population of Children in the Nation's Housing Stock
Tables 3-6 through 3-10 in Section 3.3.1.1 characterize measured lead levels in dust and
soil in the 1997 occupied housing stock, prior to implementing interventions that would occur
under the proposed §403 rule. These summaries were based on data from the HUD National
Survey with sampling weights revised to represent the 1997 national occupied housing stock. To
characterize the extent to which these environmental-lead levels provide exposures to children
and to characterize the benefits associated with performing interventions under §403 rules, it was
necessary to estimate numbers of children of specific age groups who reside within the 1997
national housing stock depicted in Table 3-5 of Section 3.3.1.1.
Methods used to obtain numbers of children in the national housing stock are presented in
Section 1.2 of Appendix Cl. These methods used estimates developed by the U.S. Bureau of the
Census of the 1997 birth rate, the average number of children per 1,000 people, and the average
number of residents per housing unit. While this risk analysis focused on characterizing the
blood-lead concentrations and associated health effects for children aged 1-2 years (i.e., 12-35
months), the sensitivity analysis within the risk characterization (Section 5.4) also considered
children aged 1-5 years (i.e., 12-71 months). Therefore, the methods in Appendix Cl were
applied to both age groups.
Table 3-35 provides the estimated number of children residing in the 1997 housing stock
according to age of housing unit and age of child. Numbers of children associated with the
1997 sampling weights for each HUD National Survey unit are displayed in Table C1-7 of
Appendix Cl.
Table 3-35. Estimated Number of Children in the 1997 National Housing Stock, by Age of
Child and Year-Built Category-
Years in Which Housing
Units Were Built
Prior to 1 940
1940-1959
1960-1979
After 1979
All Housing1
Age Of Child Within These Housing Units
1-2 Years
1,578,000
1,581,000
2,805,000
1,996,000
7,961,000
1-5 Years
4,043,000
4,051,000
7,188,000
5,115,000
20,397,000
1 Values in this row may differ from sum of previous rows due to rounding.
3.4 DISTRIBUTION OF CHILDHOOD BLOOD-LEAD
As described in Section 2.4, the population of interest in this risk analysis is children aged
1-2 years (i.e., 12-35 months). To characterize the national distribution of blood-lead
concentration for these children, this risk analysis uses data from Phase 2 of the Third National
Health and Nutrition Examination Survey (NHANES IE). Information on NHANES III and a
3-62
-------
summary of the blood-lead concentrations from Phase 2 of this survey are presented in Section
3.4.1. These data are used in Section 5.1 to establish a baseline distribution of blood-lead
concentration for this risk analysis to measure health effects in children aged 1-2 years. In
Section 5.1, three factors were used to establish this baseline distribution: the geometric mean
blood-lead concentration for this age group as presented in Section 3.4.1, the corresponding
geometric standard deviation of these data, and an assumption that blood-lead concentrations for
this group of children follow a lognormal distribution.
Additional information on blood-lead concentrations to support the NHANES III data is
provided in Sections 3.4.2 through 3.4.4 through summary statistics from the Baltimore R&M
study, the Rochester Lead-in-Dust study, and the HUD Grantees program. While blood-lead
concentrations in these latter three studies are not representative of lead exposure on a national
scale, they provide supporting information on the prevalence of elevated blood-lead
concentrations in children living in inner-city locations, in children living in primarily older
housing in an urbanized setting, or hi children living in housing units likely to contain lead-based
paint.
3.4.1 Distribution of Blood-lead Concentration, as Measured by NHANES III
The National Health and Nutrition Examination Surveys, conducted by the CDC's
National Center for Health Statistics (NCHS), trace the health and nutritional status of the
noninstitutionalized, civilian U. S. population. The surveys consist of adult, youth, and family
questionnaires, followed by standardized physical examinations.
NHANES ffl, conducted from 1988 to 1994, was the seventh in a series of national
examination studies conducted by NCHS since 1960. The target population for NHANES ffl
included the civilian noninstitutionalized population 2 months of age and older. The primary
objectives of NHANES ffl were the following (CDC, 1992):
"To produce national population health parameters; to estimate the national prevalence
of selected diseases and disease risk factors; to investigate secular trends in selected
diseases and risk factors; to contribute to the understanding of disease etiology; and to
investigate the natural history of selected diseases."
Approximately 40,000 persons were sampled in NHANES ffl, including approximately 3,000
children aged 1-2 years. Phase 1 of NHANES ffl was conducted from 1988-1991, while Phase 2
was conducted from 1991-1994. Phase 2 provided the most recently-collected data on blood-
lead concentrations available to this risk analysis. These data included 987 children aged 1-2
years, and 2,392 children aged 1-5 years. Results from Phase 2 presented in this section were
calculated from datasets obtained from CDC's National Center for Health Statistics.
Study participants in NHANES ffl were subjected to a physical examination conducted by
a physician, a dentist, and health technicians. For participants aged 12 months and older, these
examinations included taking a blood sample via venipuncture. This sample was analyzed for
lead content by graphite furnace atomic absorption spectrophotometry.
3-63
-------
To provide for a nationally representative sample, a complex survey design was employed
in NHANES III (CDC, 1992; CDC, 1994). Although estimates of national population health and
nutrition parameters were the primary objectives of the survey, suitably precise estimates for
certain age and race groups were obtained through oversampling. As part of the survey design,
each subject was assigned a sampling weight indicating the total number in the U.S. population
represented by the given subject in the survey. This weight was determined from 1993 Current
Population Survey (CPS) data on demographic groups associated with the subject. As a result,
the NHANES IE provides national and subpopulation estimates of the distribution of childhood
blood-lead concentrations.
Table 3-36 contains descriptive summaries of the blood-lead concentration data collected
in Phase 2 of NHANES III for three age groups: 1-2 years, 3-5 years, and 1-5 years. As seen in
this table, the geometric mean blood-lead concentration for children aged 1-2 years is 3.14
Hg/dL, with a geometric standard deviation of 2.09. Lower geometric means, and slightly lower
geometric standard deviations, were observed in the other two age groups, supporting the
hypothesis that blood-lead concentration tends to peak at some age between 1 and 2 years. The
geometric mean of 3.14 |ig/dL is a decline from 4.03 |ig/dL observed for children aged 1-2 years
in Phase 1 of NHANES III. The geometric mean and geometric standard deviation blood-lead
concentration for children aged 1 -2 years form the basis for the baseline risk characterization in
Section 5.1.1.
Table 3-36. Summary of Blood-Lead Concentration Data for Children Aged 1-2 Years, 3-5
Years, and 1-5 Years, Based on NHANES III (Phase 2) Data.
Age
Range
(years)
1-2
3-5
1-5
# Children
with
Blood-Lead
Cone.
Reported
987
1,405
2,392
Sum of
NHANES
Sample
Weights1
8,262,537
11,916,489
20,179,026
Blood-Lead Concentration (/t/g/dL)2
Minimum
0.70
0.70
0.70
Maximum
32.6
56.6
56.6
Geometric
Mean
3.14
2.51
2.74
Geometric
Standard
Deviation
2.09
2.05
2.08
95% Confidence
Interval on the
Geometric Mean
(2.82, 3.49)
(2.31, 2.73)
(2.52, 2.99)
1 Weights assigned at the time of the physical examination. Included in this sum are the weights for all
children who were examined, including those who did not have a blood-lead concentration reported.
2 Summaries are calculated using sample weights to reflect the entire nation in this age group.
Table 3-36 also presents approximate 95% confidence intervals on the geometric mean
blood-lead concentration for the three age groups. These intervals indicate the level of variability
that is associated with the geometric mean estimates. This variability was estimated using
statistical software that takes into account the correlation in the data resulting from the complex
survey design. As seen hi Table 3-36, the confidence interval for the 1-2 year age group does not
overlap the confidence interval for the 3-5 year age group, providing some evidence that the two
geometric means differ from a statistical perspective. Methods for calculating these confidence
intervals are documented in Appendix C2.
3-64
-------
For these same three age groups, Table 3-37 displays estimates of the probability that
U.S. children have blood-lead concentrations at or above a given threshold: 10, 15, 20, and 25
ug/dL. These probabilities were estimated by determining the numbers of children in Phase 2 of
NHANES III whose blood-lead concentrations were at or above the given threshold, with the
results for each child weighted by his/her sampling weight. This approach differs from that taken
in the risk analysis to estimate these probabilities (Chapter 5). The estimated probabilities in
Table 3-37 are accompanied by approximate 95% confidence intervals, calculated using the
methods documented in Section 2.0 of Appendix C2.
Table 3-37. Estimated Probabilities of Elevated Blood-Lead Concentrations in Children
Aged 1-2 Years, 3-5 Years, and 1-5 Years, Based on NHANES III (Phase 2)
Data.
Age
Range
(year*)
1-2
3-5
1-5
# Children
with Blood-
Lasd Cone.
Reported
987
1,405
2,392
Sum of NHANES
Sample Weights1
8,262,537
11,916,489
20,179,026
Percentages with Elevated Blood-Lead Concentration2
(96% Confidence Interval)
ilOA/fl/dL
5.88%
(3.75, 9.22)
3.45%
(2.20, 5.41)
4.41%
(2.93, 6.65)
=>15j/fl/dL
1.96%
(0.980, 3.90)
0.943%
(0.662, 1.34)
1.34%
(0.804, 2.25)
s20^a/dt
0.431%
(0.191,0.977)
0.315%
(0.153,0.648)
0.361%
(0.173,0.756)
*25//g/dL
0.148%
(0.021, 1.04)
0.285%
(0.125,0.648)
0.231%
(0.080, 0.667)
1 Weights assigned at the time of the physical examination. Included in this sum are the weights for all
children who were examined, including those who did not have a blood-lead concentration reported.
2 The methods for estimating these percentages differ from those used in the risk characterization
(Section 5.1.1).
According to the methods used in Table 3-37, approximately 5.88% of the nation's
children aged 1-2 years have blood-lead concentrations of 10 ug/dL or above, the current action
level established by the CDC. The estimated probability of children aged 1-2 years having
blood-lead concentrations of 20 ^ig/dL or above is only 0.431%.
For children aged 1-2 years, Table 3-38 presents geometric mean blood-lead
concentration and percentage of children with a blood-lead concentration at or above 10 ug/dL
for selected subgroups of the U.S. population (family income level, urban status, and selected
race groups). Also included in Table 3-38 are approximate 95% confidence intervals associated
with these estimates, calculated using the procedures in Appendix C2. These results illustrate
that socioeconomic status is an important factor hi the incidence rate of elevated lead exposure.
Across all children, low-income families and children in urban centers have the highest
percentages of children with blood-lead concentration at 10 ug/dL or above, with income level
having more of an effect on this percentage. Urban centers are usually associated with high
environmental-lead exposure due to the high density of older buildings containing lead-based
paint and remaining fallout of leaded gasoline emissions from urban traffic. While Table 3-38
also presents results by race group, sample sizes are typically too small to warrant adequate
comparison, as indicated by the large widths of the 95% confidence intervals. However, there is
clearly a trend of high blood-lead concentration among certain race groups, such as non-Hispanic
African-Americans.
3-65
-------
Table 3-38. Estimated Percentage of Children Aged 1-2 Years (Within Selected
Subgroups) With Blood-Lead Concentrations At or Above 10 //g/dL, and
the Geometric Mean and Geometric Standard Deviation of Blood-Lead
Concentration, Based on NHANES III (Phase 2) Data.
FAMILY INCOME LEVEL1
Low
Mid
High
% :> 10 //g/dL
(95% conf. int.)
Geometric Mean
(95% conf. int.)
Sample size
% 2 10 //g/dL
(95% conf. int.)
Geometric Mean
(95% conf. int.)
Sample size
% *10//g/dL
(95% conf. int.)
Geometric Mean
(95% conf. int.)
Sample size
URBAN STATUS2
Population
2 1 ,000,000
Population
< 1 ,000,000
% ;> 10 //g/dL
(95% conf. int.)
Geometric Mean
(95% conf. int.)
Sample size
% i 10 //g/dL
(95% conf. int.)
Geometric Mean
(95% conf. int.)
Sample size
All Children
Aged 1-2
Years
Selected Race Groups of Children 1 -2 Years
Non-Hispanic
White
Non-Hispanic
African
American
Mexican
American
Other
10.1%
(7.4, 13.7)
4.27
(3.86, 4.73)
501
3.7%
(1.8, 7.8)
2.79
(2.49, 3.13)
271
3.1%
(1.1, 8.9)
2.46
(2.19, 2.76)
215
10.8%
(6.2, 18.9)
3.74
(3.03, 4.60)
75
4.0%
(1.5, 10.7)
2.71
(2.31, 3.17)
114
0.6%
(0.1, 5.0)
2.28
(2.00, 2.60)
113
14.2%
(8.8, 22.8)
5.38
(4.68, 6.19)
183
1.6%
(0.2, 11.7)
3.81
(3.12, 4.67)
75
11.4%
(3.4, 38.2)
4.21
(3.06, 5.78)
35
6.7%
(3.8, 11.7)
3.29
(2.84, 3.81)
543
5.0%
(2.0, 12.4)
2.98
(2.48, 3.57)
444
3.3%
(1.2, 9.4)
2.79
(2.29, 3.40)
111
5.0%
(1.6, 15.0)
2.68
(2.20, 3.26)
191
12.3%
(5.4, 27.8)
4.92
(3.99, 6.07)
168
7.9%
(3.8, 16.3)
4.56
(3.55, 5.86)
125
5.3%
(2.3, 12.1)
3.69
(3.05, 4.45)
212
8.3%
(4.5, 15.4)
2.98
(2.50, 3.56)
71
3.7%
(0.7, 18.0)
2.43
(1.68, 3.49)
47
6.2%
(1.3, 29.4)
4.56
(3.37, 6.17)
31
0% observed
in these data
1.92
(1.25, 2.93)
11
13.0%
(3.0, 56.1)
2.92
(1.62, 5.25)
20
5.8%
(2.9, 11.7)
3.29
(2.74, 3.95)
218
5.3%
(0.9, 32.1)
3.21
(2.67, 3.86)
112
10.3%
(3.5, 30.3)
3.25
(2.31,4.58)
46
0% observed
in these data
3.47
(2.35, 5.12)
16
1 Income level was defined by poverty income ratio (PIR) categorized as low (0
-------
As age of housing unit has historically been recognized as an important factor associated
with the presence and magnitude of lead-based paint hazards, the blood-lead concentration data
collected in Phase 2 of NHANES III were also summarized according to age of housing unit.
Table 3-39 presents the geometric mean blood-lead concentration (and geometric standard
deviation) and the percentages of children at or above specified concentration thresholds for each
category of age of house considered in NHANES III. The geometric mean blood-lead
concentration and the percentage of children at or above 10 ug/dL are presented according to
combinations of age of house and either family income level or urban status in Table 3-40.
Approximate 95% confidence intervals are provided in both tables.
Table 3-39. Estimated Geometric Mean Blood-Lead Concentration and Probabilities of
Elevated Blood-Lead Concentration in Children Aged 1-2 Years, 3-5 Years,
and 1-5 Years, by Age of Housing Unit, Based on NHANES III (Phase 2) Data.
Year
Housing
Unit Was
Built
» Children1
Geometric Mean
Blood-Lead Cone.
0/g/dL)
(95% conf. int.)
Percentages with Elevated Blood-Lead Concentration
(95% confidence interval)
2lO//g/dL
i15//g/dL
2:20 //g/dL
225/yg/dL
Children Aged 1 -2 Years
Pro- 1946
1946-
1973
Post-1973
153
361
315
4.46
(3.77, 5.27)
3.27
(2.94, 3.64)
2.37
(2.12, 2.65)
12.2%
(6.26, 23.9)
6.54%
(3.59, 11.9)
2.17%
(0.93, 5.05)
6.01%
(2.54, 14.2)
1.05%
(0.438, 2.50)
0.255%
(0.069, 0.937)
0.139%
(0.018, 1.07)
0.323%
(0.045, 2.31)
0% observed
in these data
0% observed
in these data
0% observed
in these data
0% observed
in these data
Children Aged 3-5 Years
Pre-1946
1946-
1973
Post-1973
215
528
429
3.40
(2.69, 4.29)
2.55
(2.30, 2.83)
1.83
(1.68, 2.00)
6.16%
(3.27, 11.6)
3.43%
(1.72, 6.83)
1.23%
(0.32, 4.72)
1.89%
(0.947, 3.75)
0.432%
(0.164, 1.14)
0.551%
(0.120, 2.52)
0.465%
(0.156, 1.39)
0.171%
(0.039, 0.752)
0.141%
(0.020, 0.985)
0.303%
(0.079, 1.16)
0.171%
(0.039, 0.752)
0.141%
(0.020, 0.985)
Children Aged 1-5 Years
Pre-1946
1946-
1973
Post-1973
368
889
744
3.79
(3.12, 4.60)
2.81
(2.58, 3.06)
2.04
(1.86, 2.24)
8.60%
(5.22, 14.2)
4.64%
(2.88, 7.46)
1.62%
(0.60, 4.40)
3.54%
(1.97, 6.37)
0.670%
(0.349, 1.29)
0.427%
(0.134, 1.36)
0.334%
(0.103, 1.08)
0.230%
(0.130, 0.406)
0.082%
(0.012, 0.579)
0.181%
(0.046, 0.716)
0.105%
(0.024, 0.465)
0.082%
(0.012, 0.579)
1 Number of children with blood-lead concentration reported and for which the age category of his/her housing
unit was known.
3-67
-------
Table 3-40. Estimated Percentage of Children With Blood-Lead Concentrations Exceeding
10//g/dL, and the Geometric Mean and Geometric Standard Deviation of
Blood-Lead Concentration, for Children Aged 1 -2 Years According to Age of
Child's Residence and Either Family Income Level or Urban Status.
FAMILY INCOME LEVEL1
Low
Mid
High
% i10//g/dL
(95% conf. int.)
Geometric Mean
U/g/dL)
(95% conf. int.)
Sample size
% *10;/g/dL
(95% conf. int.)
Geometric Mean
(//g/dL)
(95% conf. int.)
Sample size
% :> 10 //g/dL
(95% conf. int.)
Geometric Mean
U/g/dL)
(95% conf. int.)
Sample size
URBAN STATUS2
Population
21,000,000
Population
< 1,000,000
% i10//g/dL
(95% conf. int.)
Geometric Mean
Ot/g/dL)
(95% conf. int.)
Sample size
% *10/yg/dL
(95% conf. int.)
Geometric Mean
(//g/dL)
(95% conf. int.)
Sample size
Year Housing Unit Was Built
Pre-1946
1946-1973
Post-1973
19.7%
(11.5, 33.9)
6.18
(5.20, 7.33)
73
10.2%
(2.32, 44.6)
3.73
(2.76, 5.04)
39
6.52%
(1.22, 35.0)
3.78
(2.84, 5.02)
41
9.01%
(5.52, 14.7)
4.11
(3.49, 4.83)
201
3.73%
(1.42, 9.82)
2.82
(2.37, 3.36)
97
5.76%
(1.49, 22.2)
2.65
(2.07, 3.41)
63
7.81%
(3.65, 16.7)
3.60
(3.14, 4.13)
126
0.59%
(0.11, 3.26)
2.33
(2.01, 2.71)
99
0% observed in
these data
1.87
(1.68, 2.08)
90
16.5%
(8.89, 30.7)
4.87
(3.70, 6.43)
85
7.79%
(1.61,37.8)
4.06
(3.31,4.99)
68
6.67%
(3.60, 12.4)
3.47
(3.05, 3.94)
224
6.38%
(1.97, 20.7)
3.05
(2.55, 3.64)
137
1.22%
(0.53, 2.78)
2.44
(2.20, 2.70)
158
3.16%
(0.92, 10.9)
2.30
(1.84, 2.88)
157
1 Income level was defined by poverty income ratio (PIR) categorized as low (0
-------
The overall conclusion made by Phase 2 of NHANES m is that blood-lead concentrations
in U.S. children continued to decline in the early 1990s. However, the percentage of children
with elevated blood-lead concentration remains disproportionately high in certain subpopulations
(e.g., low-income households, urban area households) that have a greater likelihood of
encountering lead hazards.
3.4.2 Baltimore Repair and Maintenance (R&M) Study
Childhood blood-lead concentrations collected prior to any interventions performed in the
Baltimore R&M Study (Section 3.2.2.1) provide evidence of elevated blood-lead concentrations
for children hi inner-city, high-exposure environments. Units slated for R&M interventions were
documented to contain lead-based paint and elevated lead levels in household dust. Modern
urban units, assumed to be free of lead-based paint, acted as negative controls. Previously-abated
units were abated for lead-based paint previous to this study, and therefore reflect a post-
abatement environment.
Table 3-41 summarizes blood-lead concentrations measured in the initial round of
sampling (i.e., at study enrollment, prior to any interventions that may have occurred in the study,
or when a child moves into a vacant unit following R&M intervention). In the initial sampling of
blood among 93 children aged 1-2 years in this study (collected from 1993-1994), the overall
geometric mean blood-lead concentration is 9.94 ug/dL, which is over three times the value of
3.1 jig/dL obtained from Phase 2 of NHANES in (Table 3-36). In particular, geometric mean
blood-lead concentrations for children aged 1-2 years in previously-abated units and units slated
for R&M intervention (11.9 jig/dL and 10.6 ug/dL, respectively) are high, while for 1-2 year old
children residing among the modem urban units, where potential for lead exposure was reduced,
the geometric mean was 2.82 ug/dL. Thus, it is possible that previously-abated units continue to
have lead exposures that affect children's blood lead concentration.
The percentages of children with blood-lead concentrations at or above 10,15,20, or 25
lig/dL are high for all but the modern urban units (Table 3-41). When measured at enrollment or
prior to interventions performed hi the study, blood-lead concentrations were at or above 10
ug/dL for approximately 47% of 1-2 year olds compared to the Phase 2 NHANES HI estimate of
5.88% (Table 3-37). Again, this large difference is likely due to the increased lead exposure
associated with these children compared to the national population. Of the different groups of
study units in the study, the highest percentage of children with blood-lead concentrations above
10 ug/dL occurred for previously-abated units (67%). Recall from Section 3.3.1.2 that
environmental-lead levels were high in these units as well, suggesting that not all environmental-
lead exposures were removed as a result of the abatements performed previous to this study. In
contrast, no children aged 1-2 years who resided in modern urban units had elevated blood-lead
concentrations.
3-69
-------
Table 3-41. Summary Statistics on Blood-Lead Concentration Measured in the Initial
Round of Sampling in the Baltimore Repair and Maintenance Study.
Age
Range of
Children
(years)
Number
of
Children
Blood-Lead Concentration (pg/dL)
Minimum
Maximum
Geometric
Mean
Geometric
Standard
Deviation
Percentages with Elevated
Blood-Lead Concentration (%)
i10
j/g/dL
;>15
fjgldL
*20
//g/dL
225
//9/dL
All Initial Round Blood-Lead Concentrations1
All
1-2
163
93
0.9
0.9
65.5
65.5
10.4
9.94
2.12
2.29
58.3
53.8
35.0
33.3
14.7
16.1
10.4
12.9
Children Living in Study Units at the Time of Enrollment (Pre-lntervention)
All2
1-2
115
68
0.9
0.9
65.5
65.5
8.78
8.38
2.14
2.34
47.0
44.1
26.1
26.5
9.6
11.8
5.2
7.4
Children in Previously-Abated Units
All3
1-2
23
12
3.65
3.65
28.8
24.2
12.7
11.9
1.60
1.71
73.9
66.7
43.5
50.0
13.0
8.3
4.3
0.0
Children in Units Slated for Repair and Maintenance (Pre-lntervention)
All2
1-2
69
41
1.75
1.75
65.5
65.5
10.2
10.6
1.87
1.99
49.3
51.2
27.5
26.8
10.1
14.6
5.8
9.8
Children in Modem Urban (control) Units
All4
1-2
19
14
0.9
0.9
10.2
5.8
3.04
2.82
1.74
1.67
5.3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1 Includes children moving into vacant study units following R&M interventions (blood sampling performed
prior to moving into these units).
2 Children aged 6-57 months.
3 Children aged 1 0-57 months.
4
Children aged 1 6-43 months.
3.4.3 Rochester Lead-in-Dust Study
Blood-lead concentration data were collected in 1993 in the Rochester study (Section
3.2.2.2) for 205 children aged 12-30 months of age. While units having recent major renovations
or the potential for lead contamination from exterior sources were not considered in this study,
no attempt was made to include units based on environmental-lead levels or the presence of lead-
based paint. Therefore, this study characterizes lead exposure conditions in a particular urban
setting, but not the inner-city setting portrayed in the Baltimore R&M Study.
Table 3-42 summarizes the blood-lead concentrations from the Rochester study (as stored
hi the study's public use database). The geometric mean blood-lead concentration was 6.38
ug/dL (geometric standard deviation, 1.85), twice the geometric mean reported hi Phase 2 of
NHANES HI for children aged 1-2 years. Twenty-three percent of the children had blood-lead
concentrations at or above 10 ug/dL, 8% at or above 15 ng/dL, and 3% at or above 20 ug/dL.
Therefore, blood-lead concentrations in the Rochester study are generally higher than those for
Phase 2 of NHANES HI, but lower than those for the Baltimore R&M study.
3-70
-------
Table 3-42. Summary Statistics on Blood-Lead Concentration Measured in the Rochester
Lead-in-Dust Study.
Number
of
Children
205
Blood-Lead Concentration fy/g/dL)
Minimum
1.4
Maximum
31.7
Geometric
Mean
6.38
Geometric
Standard
Deviation
1.85
Percentages with Elevated Blood-Lead
Concentration
*10//g/dL
23.4
*15//g/dL
7.8
*20jfg/dL
2.9
*28 fjgldL
1.5
3.4.4 Evaluation of the HUD Lead-Based Paint Hazard Control
Grant Program ("HUD Grantees")
Blood-lead concentrations of children residing in households participating in the
evaluation phase of the HUD Grantees program (Section 3.2.2.3) were measured, along with
environmental-lead levels in various media. The population of children targeted for participation
in the program differed among the fourteen grantee recipients, due to the different enrollment
criteria among the grantees (see Table 3-4). These criteria included targeting high-risk
neighborhoods, enrolling only homes with a lead-poisoned child, and considering unsolicited
applications. Pre-intervention data collected through September 1997 are presented hi this
exposure assessment; these data provide some of the most recent information on the relationship
between children's blood-lead concentration and environmental-lead levels.
Across all grantees, blood-lead concentration data were collected for 471 children aged
1-2 years and for 657 children aged 3-5 years. Either venipuncture or fingerstick blood collection
methods were used. Table 3-43 summarizes blood-lead concentrations by blood collection type,
by age, and by grantee. The geometric mean blood-lead concentration via venipuncture
collection method is 9.2 ug/dL for children aged 1-2 years, and 7.6 ug/dL for children aged 3-5
years. These geometric mean values are from two to three tunes higher than those obtained for
children at the same age group from Phase 2 of NHANES ni (Table 3-36), reflecting the
procedure of selecting higher-risk children for the HUD Grantees program. The differing
enrollment criteria across grantees also contributed to considerable differences hi the geometric
mean blood-lead concentration among the grantees. Under venipuncture, the geometric means
for individual grantees reporting more than one blood-lead result ranged from 4.0 ug/dL
(California, which only targeted older units) to 18.9 ug/dL (Cleveland, which targeted units with
lead-poisoned children). The geometric mean blood-lead concentration via fingerstick method is
9.5 ug/dL for children aged 1-2 years and 8.9 ug/dL for children aged 3-5 years. When data were
available for more than one child under fingerstick collection methods, the geometric means
ranged from 5.5 ug/dL (Wisconsin) to 13.2 ug/dL (Milwaukee).
3-71
-------
Table 3-43. Summary of Children's Blood-Lead Concentration in the HUD Grantees
Program, According to Blood Collection Method, Age of Child, and Grantee.
Age Range
1-2 Years
3-5 Years
1-5 Years
Grantee
Alameda
County
Baltimore
Boston
California
Cleveland
Massachusetts
Minnesota
New Jersey
Rhode Island
Wisconsin
Chicago
New York City
Vermont
Age Range
1-2 Years
3-5 Years
1-5 Years
Grantee
Baltimore
Cleveland
Massachusetts
Minnesota
Rhode Island
Wisconsin
Milwaukee
Vermont
Number of
Children
Blood-Lead Concentration (i/g/dL)
Geometric
Mean
Geometric
Standard
Deviation
Minimum
25th
Percentile
Median
75th
Percentile
Maximum
Blood Collection Method = Venlpuncture
313
442
755
9.2
7.6
8.2
2.3
2.2
2.2
0.7
0.7
0.7
5.0
4.0
4.9
10.0
8.0
9.0
17.0
15.0
16.0
53.0
48.0
53.0
(Children Aged 1-2 Years only)
26
23
19
19
46
45
75
1
13
8
15
15
8
4.8
8.0
9.9
4.0
18.9
8.7
10.7
3.0
8.2
7.6
13.5
4.6
12.4
2.3
1.9
2.0
2.0
1.7
1.9
2.4
.
2.0
1.7
1.7
1.6
1.6
1.4
2.0
3.0
1.4
4.0
3.0
0.7
3.0
2.0
4.0
4.0
2.0
6.0
3.0
6.0
6.0
2.6
14.0
5.0
6.0
3.0
6.0
4.9
11.0
4.0
8.5
4.8
7.0
14.0
3.5
18.0
9.0
11.0
3.0
9.0
7.5
14.0
5.0
14.4
6.6
10.0
19.0
7.2
28.0
15.0
22.0
3.0
14.0
11.5
19.0
5.0
17.0
24.8
26.0
24.0
15.0
53.0
40.0
43.0
3.0
21.0
16.0
28.0
12.0
22.0
Blood Collection Method = Fingerstick
158
215
373
9.5
8.9
9.1
2.0
2.0
2.0
2.0
2.0
2.0
6.0
5.0
5.0
9.0
9.0
9.0
16.0
15.0
15.0
62.0
49.0
62.0
(Children Aged 1 -2 Years only)
1
1
3
1
9
36
83
24
9.0
13.0
7.1
33.0
8.2
5.5
13.2
7.0
.
.
1.7
.
1.5
1.4
2.0
1.6
9.0
13.0
4.0
33.0
5.0
3.5
2.0
3.5
9.0
13.0
4.0
33.0
7.0
4.0
9.0
5.0
9.0
13.0
9.0
33.0
7.0
5.0
15.0
6.5
9.0
13.0
10.0
33.0
11.0
7.0
20.0
11.0
9.0
13.0
10.0
33.0
15.0
14.0
62.0
16.0
3-72
-------
The percentages of children with blood-lead concentrations at or above 10, 15, 20 or 25
ug/dL are summarized in Table 3-44. Fifty two percent of children aged 1-2 years had blood-
lead concentrations (venipuncture) at or above 10 ug/dL, compared to the estimates of 5.88% for
Phase 2 of NHANES III (Table 3-37), 53.8% for the Baltimore R&M study (Table 3-41), and
23.4% for the Rochester Lead-in-Dust study (Table 3-42). For individual grantees, the
percentage of children aged 1-2 years with blood-lead concentrations (venipuncture) at or above
10 ng/dL varies from 16% (California) to 91% (Cleveland). Percentages under the fmgerstick
method are similar to that under venipuncture, but less data were available to estimate them.
Table 3-44. Percentage of Children with Elevated Blood-Lead Concentration in the HUD
Grantees Program, According to Blood Collection Method, Age of Child, and
Grantee.
Age Range
1-2 Years
3-5 Years
1-5 Years
Grantee
Alameda
County
Baltimore
Boston
California
Cleveland
Massachusetts
Minnesota
New Jersey
Rhode Island
Wisconsin
Chicago
New York City
Vermont
Age Range
1-2 Years
3-5 Years
1-5 Years
Grantee
Baltimore
Cleveland
Massachusetts
Minnesota
Rhode Island
Wisconsin
Milwaukee
Vermont
Number of
Children
Percentage of Children with Elevated Blood-Lead Concentration {%)
* 10 //g/dL
* 15/yg/dL
* 20 //g/dL
2 25/^g/dL
Blood Collection Method = Venipuncture
313
442
755
52
41
45
35
26
30
19
10
14
12
5
8
(Children Aged 1-2 Years only)
26
23
19
19
46
45
75
1
13
8
15
15
8
23
39
53
16
91
44
61
0
38
38
80
7
63
12
22
47
5
70
29
44
0
23
13
47
0
50
4
9
11
0
48
9
31
0
8
0
20
0
13
0
4
0
0
37
7
21
0
0
0
7
0
0
Blood Collection Method = Fingerstlck
158
215
373
45
46
45
30
26
28
14
14
14
11
7
9
(Children Aged 1-2 Years only)
1
1
3
1
9
36
83
24
0
100
33
100
33
3
69
29
0
0
0
100
11
0
51
13
0
0
0
100
0
0
25
0
0
0
0
100
0
0
19
0
3-73
-------
Figures 3-4 and 3-5 illustrate the nature of the linear relationship observed in the HUD
Grantees program between (log-transformed) children's blood-lead concentration and the
household's (log-transformed) area-weighted arithmetic average dust-lead loading for floors and
window sills, respectively. Also included in these figures are the linear relationship associated
with the data from Rochester Lead-in-Dust study and, in Figure 3-5, the Baltimore R&M study.
The regression lines span the ranges of the observed dust-lead loadings. Note that the slopes of
these lines are relatively parallel across the grantees and the two other studies. This indicates that
the relationships between blood-lead concentration and dust-lead loading are relatively consistent
across grantees. In particular, blood-lead concentration and dust-lead loading data from the
Rochester study (used in developing the empirical model in Chapter 4) have similar relationships
to what is observed hi the HUD Grantees program, which considers data from a much larger
geographical area and under various exposure conditions.
100-
10
1-
0.1 1.0 10.0 100.0 1000.0 10000.0
Carpeted and Uncarpeted Floor Wipe Area-Weighted Arithmetic Average Dust-Lead Loading (ug/ft2)
Atameda County
Cleveland
Wteooraki
QQQ
L L L Chicago
Q |% ft
w O 17
III rM. .j lanrl
T I r nnooo raw
M-M-M NawYbifcClty
Figure 3-4. Blood-Lead Concentration Versus Area-Weighted Arithmetic Average
Floor Dust-Lead Loading (Wipe Collection Method), for HUD Grantee
and Rochester Lead-in Dust Study Data.
3-74
-------
100-
10-
1
10 100 1000 10000
Window Sill Area-Weighted Arithmetic Average Dust-Lead Loading (ug/ft2)
100000
A A A Atanwd* County
e-E-E Ctovrimd
I I I
www
LJ L) U Bwornofw
K K K MW«*M
000 MttmorvR&M
f* f* f* B-.*--
w O O DOMDn
Q G G £TO*
tfcb Chicago
POD
i I r Rhocto Wind
M-M-M NiwKMcCty
Figure 3-5. Blood-Lead Concentration Versus Area-Weighted Arithmetic
Average Window Sill Dust-Lead Loading, for the HUD Grantee,
Rochester Lead-in-Dust, and Baltimore R&M Studies.
3.5 EXPOSURE ASSESSMENT CHARACTERIZATION
Conclusions across multiple studies have indicated that lead-based paint hazards.
including lead-contaminated soil and lead-contaminated dust, are primary contributors to overall
lead exposure in young children. Lead-based paint, especially when in a deteriorated state or
when found on accessible, chewable, impact, or friction surfaces, is a significant high-dose
source of lead exposure in pre-school children, hi turn, several lead exposure studies have
concluded that the pathway of lead-contaminated soil and dust to children's blood is an important
means by which young children are exposed to lead from lead-based paint hazards.
For this risk analysis, the exposure media associated with lead-based paint hazards within
the nation's housing are segmented into three categories: 1) deteriorated lead-based paint; 2)
yard soil contaminated from exterior sources such as deteriorated lead-based paint on the exterior
of the residence; and 3) household dust contaminated from sources such as deteriorated lead-
based paint on ulterior surfaces and lead-containing soil tracked in from the exterior of the
residence.
3-75
-------
This risk analysis selected the HUD National Survey of Lead-Based Paint in Housing as
the primary data source for characterizing the distribution of environmental-lead levels in dust
and soil in the nation's housing stock. The HUD National Survey is considered the most
complete and representative characterization of lead levels in dust and soil in the nation's
housing built prior to 1980. As part of the survey, each surveyed unit was assigned a sampling
weight corresponding to the number of units hi the national housing stock whose environmental
conditions were represented by the given unit. Data from the American Housing Survey
(sponsored by the Bureau of the Census and the U.S. Department of HUD) were used to revise
these sampling weights to reflect the 1997 national housing stock.
Using HUD National Survey data with revised sampling weights, this risk analysis has
estimated that the 1997 national housing stock contains over 99 million occupied housing units.
of which 62% are expected to contain lead-based paint and 14% to contain more than five square
feet of deteriorated lead-based paint. When considering only occupied units built prior to 1980.
83% are expected to contain lead-based paint and 18% are expected to contain more than five
square feet of deteriorated lead-based paint. Geometric mean environmental-lead levels in paint,
dust, and soil tend to increase with age of unit. This finding is consistent with several studies
that provide evidence of a link between the presence of lead-based paint (especially in a
deteriorated state) and age of housing unit.
The design and findings of the HUD National Survey have been peer reviewed and
published in several government reports. However, among the limitations associated with using
data from this survey (with revised sampling weights) in this risk analysis are the following:
The survey did not collect blood samples from children within the sampled units,
thereby preventing this data source from being used to characterize the relationship
between lead-based paint hazards and blood-lead concentration.
The survey did not take more than three dust samples from within a housing unit;
these samples were not necessarily taken from areas experiencing high levels of
activity among resident children.
Considerable measurement error may be present in the measured dust-lead
concentrations due to small amounts of dust collected in some samples. (Adjusting
dust-lead concentrations for small dust mass required a laboratory study to be
conducted as part of this risk assessment as discussed hi USEPA, 1996c.)
Measurement error and its impact on the relationship between dust-lead and blood-
lead concentration, as characterized by data from the Rochester study, is investigated
in Emond et al., 1997.
Dust samples were collected using the Blue Nozzle vacuum method, which differs
from the wipe sampling method assumed hi the §403 rulemaking when determining
whether dust-lead loadings in a housing unit exceed pre-specified standards. (This
situation required a dust-lead loading conversion procedure as discussed in Section
4.3 of Chapter 4.)
3-76
-------
The HUD National Survey did not attempt to sample lead levels in paint on all
painted surfaces in a surveyed housing unit, creating considerable uncertainty in how
housing units are characterized as containing lead-based paint.
The survey did not sample housing built after 1979. (To represent such housing, this
risk assessment had to employ inference procedures.)
It is uncertain how environmental-lead levels in housing have changed since the
survey was performed in 1989-1990.
The need to update the sampling weights to represent the 1997 national housing stock
introduced additional uncertainty to these weights. While the updating procedure was
based on data from the nationally-representative American Housing Survey, these
data were collected in 1993 and required additional procedures to be developed in
this risk assessment to update the results to 1997.
The sample size of 284 housing units may be considered too small to adequately
represent the nation's housing stock.
The field sampling occurred in winter months, which can influence the values of
environmental-lead and blood-lead measurements.
A collection of human characterization studies and intervention studies were identified in
which significant positive relationships were consistently observed between lead-based paint
exposures in residential environments and children's blood-lead concentration. Three recent
studies, the pre-intervention phases of the Baltimore (MD) Repair & Maintenance Study
(conducted in a variety of housing types in an inner-city setting) and the Evaluation of the HUD
Lead-Based Paint Hazard Control Grant program, as well as the Rochester (NY) Lead-in-Dust
study (conducted in an urban/suburban setting), provided extensive information on
environmental-lead levels and blood-lead concentrations in children exposed to such levels in
their respective cities. Data from these three studies were summarized as part of this exposure
assessment, to supplement the information on baseline distributions of environmental-lead levels
in housing and blood-lead concentrations in young children. Conclusions from these studies
indicate that elevated lead levels in paint, dust, and soil continue to exist in residential
environments, particularly in older housing units. Even at low to moderate lead levels, lead-
contaminated dust can affect children's blood-lead concentration.
Despite the general agreement across studies that blood-lead concentration is positively
associated with environmental-lead levels, it is difficult to combine these findings into a single,
quantitative measure of association. One reason is the presence of qualitative dissimilarities
among the studies. These dissimilarities include differences in study design, sampling and
analysis protocols, study objectives, target populations, and study locations. Another reason is
that each study collects different types of supporting data, some of which may reduce the
influence of some environmental-lead variables on blood-lead concentration when they are
represented in statistical modeling procedures.
3-77
-------
The extent to which a particular child is exposed to lead-based paint hazards is
determined not only by environmental factors, but also by the child's activity patterns and by
household characteristics. Examples include the number of hours the child is inside/outside the
residence, the amount of time the child is away from the residence (including any exposures
he/she may encounter there), the presence of pets, the occupation of adults in the household,
cleaning habits of the residence, the presence of air conditioning, and the frequency that windows
are opened. Therefore, lead exposures pose a health hazard to a child only to the extent that the
child encounters the exposures and they become bioavailable to the child.
The baseline distribution of blood-lead concentration in the target population (children
aged 1-2 yeare) was derived in this risk analysis from data collected in Phase 2 of NHANES PL
These data, the latest available on the national distribution of blood-lead concentration
(representing the period 1991-1994), verify the hypothesis that blood-lead concentration tends to
peak in children at ages 1-2 years. Phase 2 of NHANES HI estimates that the geometric mean
blood-lead concentration for this age group is 3.14 [ig/dL: approximately 5.88% of these children
have blood-lead concentrations at 10 ug/dL or higher. While these statistics have declined from
earlier years, they continue to be unacceptably high, especially for children in urban centers or
within low income households due to their higher likelihood of encountering lead-based paint
hazards. While the national representation of NHANES m results is widely accepted, some
possible limitations in using these data include ignoring any seasonally effects on blood-lead
concentration and any further decline in concentrations that may have occurred since 1994.
Information on average family sizes (as measured in the American Housing Survey) and
numbers of children per person in the U.S. population (as projected by the U.S. Census Bureau)
was used to determine numbers of children exposed to each environmental-lead condition as
represented by housing units within the HUD National Survey. Possible limitations associated
with using this information include assumptions that these numbers remain consistent across the
entire national housing stock.
3-78
-------
4.0 DOSE-RESPONSE ASSESSMENT
CHAPTER 4 SUMMARY
Chapter 4 presents the approach for characterizing the relationship between
environmental lead exposure and the resulting adverse health effects. The
relationship is established in two stages. First, the relationship between
environmental lead levels and blood-lead concentration is characterized. Two
different models, the IEUBK and empirical models, are used to characterize this
relationship. Then the relationship between blood-lead concentration and specific
elevated blood-lead concentration and health effect endpoints is established. This
two-stage relationship is applied in this risk analysis (Chapter 5), using
environmental data from the HUD National Survey, to estimate the number of
children who will benefit from the §403 rule.
This chapter describes the two models that are used to relate
environmental- lead levels to blood-lead concentration and establishes the
relationship between blood lead and the specific elevated blood-lead concentration
and health effect endpoints. Methods for converting environmental lead levels
measured by different sampling methods are also presented.
Figure 4-1 outlines the approach for the dose-response assessment. The
conclusions from the dose-response assessment are presented in Section 4.5.
This chapter seeks to answer the following questions:
1. What is the dose-response relationship between environmental-lead exposure and
the blood-lead concentration and health effect endpoints evaluated in this risk
analysis?
la. What is the dose-response relationship between environmental-lead exposure and
childhood blood-lead concentration?
Ib. What is the dose-response relationship between childhood blood-lead
concentration and health effects?
2. Can lead loadings in dust samples collected using a vacuum sampler be converted to
wipe-equivalent dust-lead loadings?
4-1
-------
Background
and
Objectives
Hazard
Identification
r
1.
Exposure
Assessment
DOSE-RESPONSE
ASSESSMENT
Present
IEUBK Model
(Section 4.1)
Present
Empirical Model
(Section 4.2)
Apply Appropriate
Inputs to IEUBK Model
(Sections 4.1.2, 4.1.3)
Convert Post-
Intervention Wipe
Dust-Lead Loadings
(Section 4.3)
w
Apply Appropriate
Inputs to Empirical
Model
(Section 4.2.2)
Estimate National
Distribution of Blood
Lead Concentration
Using the IEUBK Model
(Section 4.1.4)
Estimate National
Distribution of Blood
Lead Concentration Using
the Empirical Model
(Section 4.2.6)
Convert Blood-Lead
Distributions to Health
Effect Endpoints
(Section 4.4)
Risk
Characterization
Risk
Management
Conclusions
on Risk
Characterization
Conclusions on
Analysis of Example
Options for $403
Standards
Figure 4-1. Detailed Flowchart of the Approach to Dose-Response Assessment.
4-2
-------
Answering question 1, the primary question addressed in this chapter, is problematic, as only
limited data exist for relating health outcomes directly to environmental-lead levels. The link
between lead exposure and health effects is usually studied hi terms of a measure of body-lead
burden, such as blood-lead concentration, rather than environmental-lead levels. Therefore, the
answer to question 1 is obtained by addressing questions la and Ib. The relationship between
environmental-lead levels and health outcomes is computed in a two stage process. This two
stage dose-response relationship is used to characterize the risk due to lead exposure under
present environmental conditions (Chapter 5) and to estimate the risk under environmental
conditions predicted to occur for various examples of options for the §403 standards (Chapter 6).
The specific health effect and blood-lead concentration endpoints utilized in this risk analysis
were identified hi Chapter 2. Question 2 is necessary as §403 dust-lead standards are expected to
be defined hi terms of a wipe dust-lead loading, and dust samples hi the HUD National Survey
(the primary source of data on environmental-lead levels in the nation's housing stock used in
mis risk analysis) were collected via vacuum sampling.
Figure 4-1 illustrates the relationship between material presented in this chapter and other
key elements of the risk analysis. Two models are utilized to relate environmental-lead levels to
blood-lead concentrations: the Integrated Exposure, Uptake, and Biokinetic (IEUBK) Model for
Lead hi Children (USEPA, 1994a, 1995d) and an empirical model that was developed
specifically for this risk analysis. The application of the IEUBK model in this risk analysis is
described hi Section 4.1. The development of the empirical model and its application hi this risk
analysis are described hi Section 4.2. Briefly, each model is applied to characterize the national
distribution of blood-lead concentrations of children aged 1-2 years both prior to and following
implementation of §403 rulemakmg ("pre-§403" and "post-§403"). Data collected in the HUD
National Survey serve as inputs to both models for the estimation of the pre-§403 distribution.
Estimation of post-§403 environmental lead distributions is discussed hi Chapter 6. Section 4.3
presents conversion equations developed to relate dust-lead loadings under different dust
sampling methods (e.g., Blue Nozzle vacuum) to wipe dust-lead loadings. These conversions are
used to compare environmental levels to standards and to prepare the post-intervention data for
input to the models. The national distributions of blood-lead concentrations predicted by each
model are used as input to the second stage models relating blood-lead concentrations to health
outcomes. Section 4.4 presents the approach for relating blood-lead concentration to the elevated
blood-lead concentration and health effects endpoints identified in Chapter 2. The dose-response
characterization (Section 4.5) provides summary answers to the above questions and addresses
the strengths and weaknesses of the scientific evidence and decisions made, as they are relevant
to this risk analysis.
4.1 IEUBK MODEL
This section describes how EPA's Integrated Exposure, Uptake, and Biokinetic (IEUBK)
Model for Lead in Children (USEPA, 1994a, 1995d) is used hi this risk analysis to model the
dose-response relationship between environmental-lead levels in the nation's housing stock and
blood-lead concentration hi children aged 1-2 years.
4-3
-------
The precursor to the biokinetic part of the IEUBK model was developed in 1985 by
EPA's Office of Air Quality Planning and Standards (OAQPS) as a tool for setting air lead
standards. The version used by the Air program was peer reviewed and found acceptable by
EPA's Clean Air Science Advisory Committee of the Science Advisory Board (USEPA, 1990b).
The IEUBK model has been recommended as a risk assessment tool to support the
implementation of the July 14,1994 Office of Solid Waste and Emergency Response (OSWER)
Interim Directive on Revised Soil Lead Guidance for CERCLA Sites and RCRA Facilities. The
most current version, Version 0.99D, of the IEUBK model is used in this risk analysis.
4.1.1 Description of the IEUBK Model
The IEUBK model employs exposure, uptake, and biokinetic information to predict a
distribution of blood-lead levels in children corresponding to a specific combination of
environmental-lead levels. The predicted distribution may be used to predict the probability of
elevated blood-lead levels in children exposed to similar environmental-lead levels. The model
addresses three components of environmental risk assessment: 1) multimedia nature of exposures
to lead, 2) the differential bioavailability of various sources of lead, 3) the pharmacokinetics of
internal distribution of lead to bone, blood, and other tissues, and 4) inter-individual variability in
blood-lead levels.
Specifically, the model uses lead concentrations measured in dust, soil, air, water, diet,
and other ingested media to estimate a longitudinal exposure pattern from birth to seven years of
age (USEPA, 1995d). The model then estimates a distribution of blood-lead levels for a
population of children receiving similar exposures. The center of this distribution, the geometric
mean, is predicted by the model. A constant empirical estimate is used by the model to represent
the variability about the geometric mean. In statistical terminology, this variation is referred to as
the geometric standard deviation (GSD). The GSD characterizes the inter-individual and
biological variability in blood-lead levels of children exposed to similar environmental-lead
levels. The EEUBK model is not intended to predict the blood-lead level of an individual child
and cannot substitute for a medical evaluation of an individual child.
It is beyond the scope of this document to describe the IEUBK model in detail. Very
briefly, the model has three distinct functional components that work together in series: exposure,
uptake, and biokinetic components. Each model component is a set of complex equations and
parameters. The Technical Support Document (USEPA, 1995d) provides the scientific basis of
the parameters and equations used in the model, while the Guidance Manual (USEPA, 1994a)
includes a detailed description of the exposure pathways, absorption mechanism, and biokinetic
compartments and associated compartmented transfers of lead.
4.1.2 Inputs to the IEUBK Model
This section describes the inputs to the IEUBK model used in this risk analysis. Three
sets of parameters are used in the IEUBK model equations. (1) Exposure parameters are used to
estimate the amount of environmental lead that is taken into the body, through breathing or
ingestion. (2) Uptake parameters estimate the amount of lead that is absorbed from
-------
environmental sources. (3) Biokinetic parameters characterize the transfer of lead between
compartments of the body (for example, between blood and bone) and the elimination of lead
from the body. The IEUBK model allows the user to input values for most exposure and uptake
parameters. The biokinetic parameter values, however, are not accessible.
For this risk analysis, soil- and dust-lead concentrations from the HUD National Survey
(Section 3.3.1.1) are used as inputs to the IEUBK model to predict national distribution of blood-
lead concentrations that represents baseline (pre-§403) conditions, while adjusted concentrations
are used to predict a blood-lead concentration distribution for post-§403 conditions. IEUBK
model default values are applied for all other parameters. The default parameter values for the
BEUBK model and the calculation of input values based on the HUD National Survey are
described in this section.
IEUBK Model Default Parameters
When exposure and uptake parameter values are not specified, the IEUBK model
program provides default values. Table 4-1 presents the default values for the exposure and
uptake parameters. The default parameter values are based on various studies and are considered
the best available estimates for urban residents with no unusual lead exposure (USEPA, 1994a,
1995d). For example, the default air lead concentration is 0.1 |ig/m3, which is approximately the
average 1990 urban air lead concentration (USEPA, 1991). Thus, blood-lead concentrations
estimated using the default parameter values for exposure other than dust and soil represent the
'background' blood-lead levels that cannot be avoided (USEPA, 1994a). The use of default
parameter values is documented in detail in the BEUBK Guidance Manual (USEPA, 1994a).
While the Guidance Manual encourages the use of site-specific estimates, the default parameter
values are appropriate for assessment of national risk. In addition, site-specific estimates for the
default parameter values utilized in this risk analysis were not available.
Data from many different scientific studies of lead biokinetics, contact rates of children
with environmental media, and data on the presence and behavior of environmental lead were
utilized in developing the IEUBK model default parameter values. Details on these data sources
and the derivation of the default parameter values are provided in the Technical Support
Document (USEPA, 1995d). In brief, default values fall into five general categories: exposure
rates, exposure concentrations, uptake of ingested lead, biokinetic parameters, and variability in
blood-lead levels. Key (to this risk analysis) default parameter values in each category are
described briefly below.
Exposure rates: The age-weighted dust and soil ingestion rates used as defaults in the
model (85-135 mg/day) represent central tendency values within the range of values seen in
different studies. The default proportion (45%) of total dust and soil ingested that is derived
from soil is based primarily on a study of Dutch children in day care centers (USEPA, 1994a),
contrasting dust plus soil ingestion on days with good weather with dust ingestion on days with
rainy weather (presumably little outdoor activity on those days).
4-5
-------
Table 4-1,
Parameter
Setting*
Parameter
Setting*
Parameter
Setting*
Parameter
Setting*
Parameter
Setting
Summary of Default Parameter Values Used in the IEUBK Model (Version
0.99D).
Air Parameters
Vary air concentration by year?
No
Outdoor air lead concentration
0.10//g/m3
Indoor air lead concentration
30% of outdoor value
* All air parameters use default values
Diet Intake Parameters
Lead intake in diet, by age of child
0-1 yrs
5.53 //g/day
1-2 yrs
5.78
fig/day
2-3 yrs
6.49 //g/day
3-4 yrs
6.24 //g/day
4-5 yrs
6.01
//g/day
5-6 yrs
6.34 //g/day
6-7 yrs
7. 00 //g/day
* All diet intake parameters use default values
Water Intake Parameters
Lead Cone, in
Water
4//g/L
Drinking water consumption, by age of child
0-1 yrs
0.20
L/day
1-2 yrs
0.50
L/day
2-3 yrs
0.52
L/day
3-4 yrs
0.53
L/day
4-5 yrs
0.55
L/day
5-6 yrs
0.58
L/day
6-7 yrs
0.59 L/day
* All water intake parameters use default values
Soil and Dust Intake Parameters
Soil/Dust
Ingestion
Weighting
Factor
45% soil;
55% dust
Total soil + dust intake, by age of child
0-1 yrs
0.085
g/day
1-2 yrs
0.135
g/day
2-3 yrs
0.135
g/day
3-4 yrs
0.135
g/day
4-5 yrs
0.1 g/day
5-6 yrs
0.09
g/day
6-7 yrs
0.085 g/day
* Soil and dust lead concentrations are input. All other parameters use default values.
Absorption Method Parameters
Half Saturation
Level
100 //g/day
Total Absorption
Soil
30%
Dust
30%
Water
50%
Diet
50%
Alt.
0%
Fraction of Total Assumed Passive
Absorption
Soil
0.20
Dust
0.20
Water
0.20
Diet
0.20
Alt.
0.20
Blood Lead Parameter
Parameter
Setting
Geometric Standard Deviation (GSD)
1.6
Exposure concentrations: The default dust- and soil-lead concentrations are not used in
this risk analysis. Default diet values (5.53-7.00 //g/day) are based on data from the Food and
Drug Administration. No other data were available. Model default water values were considered
adequate for communities without a particular water-lead problem. The default air lead
concentration (0.1 ug/m3) is approximately the average 1990 urban air lead concentration.
4-6
-------
Uptake of ingested lead: Lead bioavailability varies across the chemical forms in
which lead can exist. Many factors complicate the estimation of bioavailability, including
nutritional status and timing of meals relative to lead intake (lead uptake generally increases as
dietary levels of calcium, iron, phosphate, vitamin D, fats, etc. decrease), age, and magnitude of
exposure. The default media-specific bioavailabilities in the EEUBK model are central tendency
estimates.
Biokinetic parameters: The data on which these parameter values are based originate
from a variety of separate investigations, including as much clinical data as were available
(USEPA, 1995d). The biokinetic parameters cannot be changed by the user.
Variability in blood leads (GSD): A variety of factors may cause children exposed to
similar environmental-lead concentrations to have varying blood-lead concentrations. These
include differences hi children's tendency to ingest soil or dust, hygiene habits, the potential for
soil or dust to be deposited on food, and biological factors that may affect the absorption and
processing of lead. The complexity of these factors suggests that the overall variability
encompassed by the GSD cannot be determined by aggregating the variability in each of these
factors into an overall GSD estimate. Instead, an empirical estimate of the variability in blood-
lead concentrations, a GSD of 1.6, was estimated from residential community blood-lead studies
(USEPA, 1995d). This estimate is applied for predictions of the national distribution of blood-
lead concentrations utilizing both the IEUBK and empirical models (Section 4.2).
Figure 4-2 illustrates the relationship between blood-lead concentration predicted by the
IEUBK model and specified soil- or dust-lead concentration, for children aged 24 months. The
solid line illustrates the relationship for a fixed dust-lead concentration of 200 ppm and varying
soil-lead concentrations. For the dashed line, the soil-lead concentration was fixed at 100 ppm
and dust-lead concentration varied. These fixed values are similar to the geometric mean dust-
lead concentration (192 ppm) and soil-lead concentration (78 ppm), reported hi the HUD
National Survey. From the dashed line in Figure 4-2, the predicted blood-lead concentration is 3
ug/dL for a dust-lead concentration of 100 ppm and a soil-lead concentration of 100 ppm.
Similarly, from the solid line, the predicted blood-lead concentration for 200 ppm soil- and dust-
lead concentrations is 4.5 ug/dL. It is important to recognize that each point on the predicted
curve represents a geometric mean blood-lead level for children exposed to similar
environmental-lead levels. The blood-lead levels for individual children will vary.
Utilizing the HUD National Survey Data
The IEUBK model is used in this risk analysis to predict a national distribution of
children's blood-lead concentrations. A nationally representative sample of environmental-lead
levels hi housing is required to provide inputs to the IEUBK model for this purpose. The HUD
National Survey is a recent nationally representative study that assessed environmental-lead
levels in paint, dust and soil hi residential housing.
4-7
-------
30J
c
o
I
20-
m 10-
o
i
1000 2000 3000
Lead Concentration in Soil or Floor Dust (ppm)
Fixed Soll=100 ppm; Variable Dus*
Fixed Duit=200 ppm; Variable Soil
4000
5000
Figure 4-2. IEUBK Model Predicted Geometric Mean Blood-Lead Concentration for Children
Aged 24 Months Plotted Separately Against Soil-Lead Concentration and
Dust-Lead Concentration for Fixed Default Values of the Remaining Model
Parameters
In the HUD National Survey (Section 3.3.1.1), one floor-dust sample was collected from
each of three locations (wet, dry, and entry rooms) using a Blue Nozzle vacuum sampler. The
mass weighted dust-lead concentration of these three samples is input to the IEUBK model, to
represent the dust-lead concentration to which a child is exposed. Three soil samples were
collected, one each from dripline, entryway, and remote locations. A factor weighted soil-lead
concentration (0.25 * dripline measurement + 0.25 * entryway measurement + 0.5 * remote
measurement) is used to represent the average soil-lead concentration hi the yard. This weighting
scheme avoids double counting the concentration near the house (i.e., the dripline and entryway
samples) and does not estimate the amount of time children spend in specific areas of the yard.
There are some limitations inherent in using the HUD National Survey data to provide
inputs to the IEUBK model. For example, the IEUBK model uses the specified lead
concentrations hi conjunction with soil and dust ingestion rates and bioavailability factors to
determine the dose of lead absorbed by the body. This dose is then used to predict the geometric
mean blood-lead concentration for children exposed to the specified lead concentrations. An
important assumption is that the dust- and soil-lead concentrations input to the IEUBK model are
representative of the actual lead concentration to which a child is exposed. Thus, risk assessors
typically use children's activity patterns to guide the selection of dust and soil samples. It is not
4-8
-------
possible, however, to determine whether children are actually exposed to soil- and dust-lead
levels represented by the samples collected in each HUD National Survey home. Therefore, it is
uncertain whether the mass weighted dust-lead concentrations and the factor weighted soil-lead
concentrations are typical of childhood lead exposures. A number of factors affect the actual
exposure scenario for an individual child, including the number of hours the child spends playing
inside and outside the residence, the amount of time that the child spends away from home, the
presence of pets that spend tune inside and outside, the frequency and thoroughness of house
cleaning, air conditioning, parental occupation, and a host of similar factors. Such factors can
differ among neighborhoods, communities, and time periods (Stark et al., 1982; Bornschein
etal., 1985b)
4.1.3 Estimating the Effect of Pica for Paint on Childhood Blood-Lead Levels
The exposure pathway from lead-based paint to childhood blood-lead concentration can
be both direct and indirect. Indirect exposure takes place when deteriorated lead-based paint
contaminates residential dust or soil, which is then ingested by the child. Direct exposure takes
place through the ingestion of paint chips. While the IEUBK model estimates the geometric
mean blood-lead concentration for children receiving indirect exposure to lead-based paint
through the soil- and dust-lead concentrations used as model inputs, it does not include a direct
mechanism for estimating the contribution of paint chip ingestion to childhood blood lead. This
section describes how this risk analysis accounts for the effect of pica for paint on the
geometric mean blood-lead concentrations predicated by the IEUBK model. Note that this
approach was developed specifically for this risk analysis and is not a component of the EEUBK
model.
As described in Section 4.1.2, environmental conditions observed in the HUD National
Survey are used as input to the IEUBK model. For each home hi the HUD National Survey, the
IEUBK model is used to predict the geometric mean blood-lead concentration of children
exposed to those environmental conditions. The distribution of blood-lead levels in the
population of children aged 1-2 years is then characterized by allowing each home hi the HUD
National Survey to represent a proportion of the total number of children aged 1-2 years in the
country. For homes without damaged lead-based paint, the predicted geometric mean blood-lead
concentration and the assumed geometric standard deviation of 1.6 are used to model the
distribution of blood-lead levels in children represented by each home.
For homes with damaged lead-based paint (defined as greater than 0 ft2 of ulterior or
exterior deteriorated lead-based paint), adjustments are made to the IEUBK model predictions to
account for the effect of pica for paint on children's blood-lead concentrations. In this adjustment
procedure, the children represented by each of these homes are assigned into three groups: 1)
children who have recently ingested paint chips (0.03%), 2) children who ingested paint chips at
some time (8.97%), and 3) children who do not ingest paint chips (91%). The distribution of
blood-lead levels for children in the three groups is estimated as follows:
1. Children who have recently ingested paint chips (0.03%) - Blood-lead concentration
is assigned the value 63 fig/dL with no variation.
4-9
-------
2. Children who ingested paint chips at some time (8.97%) - Geometric mean blood-
lead concentration is 3.0 ug/dL greater than the geometric mean blood-lead
concentration predicted by the IEUBK model. The adjusted geometric mean blood-
lead concentration and the assumed geometric standard deviation of 1.6 ug/dL are
used to model the distribution of blood-lead levels for these children.
3. Children who do not ingest paint chips (91.0%) - The IEUBK model predicted
geometric mean blood-lead concentration and the assumed geometric standard
deviation of 1.6 ng/dL are used to model the distribution of blood-lead levels for
these children.
The scientific evidence and assumptions used to select percentages of children assigned
to each group and the adjustments to blood-lead concentrations for children who have ingested
paint chips are described in Appendix Dl.
4.1.4 Estimating the National Distribution of Blood-Lead Using the IEUBK Model
For prediction of the pre-§403 national distribution of children's blood-lead
concentrations, estimates of soil- and dust-lead concentrations observed hi the HUD National
Survey are used as inputs to the IEUBK model, as described in Section 4.1.2. If the input values
for the IEUBK model were missing for a home in the HUD National Survey, then an imputed
value is used in the risk analysis. The imputed values are summarized in Table 3-14 of Section
3.3.1.1, with more details provided in Appendix C1. The IEUBK model is then used to predict
the geometric mean blood-lead concentration associated with each home in the HUD National
Survey.
To predict a post-§403 national distribution of children's blood-lead concentrations, the
following method was used to prepare soil- and dust-lead concentrations hi the HUD National
Survey data for input into the IEUBK model:
1. Observed levels of lead in environmental variables in the HUD National Survey were
compared to candidate §403 standards. Blue-nozzle vacuum floor and window sill
dust-lead loadings were converted to wipe dust-lead loadings before comparison to
the §403 standards. Although sill dust-lead levels are not provided as input to the
EEUBK model, they are used to determine which homes require an intervention.
2. §403 interventions were triggered hi HUD National Survey residential units that had
levels of lead hi environmental variables that were above the candidate standard. If
an intervention was triggered, assumed post-intervention lead levels in
environmental variables were substituted for observed levels. Post intervention dust-
lead concentrations for use hi the prediction are determined from methods
documented hi Section 4.3.
The geometric mean blood-lead concentrations predicted by the IEUBK model, an assumed
geometric standard deviation of 1.6, and population weights adjusted to the 1997 population of
4-10
-------
children (aged 1-2 years), are used to predict the pre- and post-§403 national distributions of
blood-lead concentrations.
Although the IEUBK model simulates a longitudinal exposure pattern from birth to seven
years of age, in order to simplify calculations for this risk analysis, a specific age was selected at
which blood-lead concentrations are estimated. The representative population for the risk
analysis is children aged 1-2 years. IEUBK model-predicted blood-lead concentrations were
examined for each month over the 12-35 month period. Predicted blood-lead concentrations at
age 24 months were found to be approximately equal to the mean predicted values over the entire
two-year period. Thus, IEUBK model predictions at age 24 months are utilized in the risk
analysis to characterize blood-lead concentrations of children aged 1-2 years.
4.2 EMPIRICAL MODEL
This section describes the development and application of an empirical model in this risk
analysis. The empirical model was developed using data from the Rochester Lead-in-Dust Study
to estimate the relationship between blood-lead levels in young children and observed levels of
lead in environmental media (paint, dust and soil) from then* primary residences. The purpose of
this model is to serve as a basis for predicting a national distribution of children's blood-lead
concentrations as a function of environmental lead-levels observed in the HUD National Survey.
Variables were selected for the model from among those that were measured in both studies, or
could be constructed in both studies using the available data. The mathematical form of the
model, variables included in the model, and parameter estimates based on the Rochester study
are presented in Sections 4.2.1 through 4.2.3. The model was then adjusted to account for
systematic differences and differences in error structure between the Rochester study variables
and the analogous HUD National Survey variables (Section 4.2.4). The final form of the
empirical model is presented in Section 4.2.5. The application of the empirical model to predict
the national distribution of blood-lead concentrations is described in Section 4.2.6.
The choice and construction of variables, the mathematical form of the empirical model,
assessment of goodness of fit and influential points, and the treatment of measurement error in
predictor variables are described in detail in Appendix G. The empirical model has not yet
undergone formal peer review or model evaluation, and is based on data from only one source
(the Rochester Lead-in-Dust Study). It is not intended as a general dose-response model, but
rather as a predictive model developed specifically for use hi this risk analysis and specifically to
predict a national distribution of blood-lead concentrations from estimates of environmental lead
as measured hi the HUD National Survey.
4.2.1 Form of the Model
The empirical model is log-linear in nature, expressing natural-log transformed blood-
lead concentration as a linear combination of natural-log transformed exposure variables and
select covariates. A typical multimedia exposure log-linear model for blood-lead concentrations
might appear as follows:
4-11
-------
ln(PbBj) = p0 + p, ln(Dust() + P2 In(Soilj) + p3 ln(Paint;) + y ' Covariatef + e,.
where PbBj is the observed blood-lead concentration of the Ith child, Dustj, Soilj, and Paintj are
the environmental lead levels in the home of the i* child, Covariatej represents one or more
variables with strong predictive value, and e-t (the residual error) is assumed to follow a normal
distribution with mean zero and variance
When translated back into the original scale of observed blood-lead concentrations, the
log-linear model yields a multiplicative relationship between environmental-lead levels and
blood-lead concentration;
PbBj = exp(P0) Dustf1 Soilj Pz Paintj p3 Covariatej Y exp(e;)
Thus, for example, the effect of dust lead on blood lead is dependent on the combined
effects of all of the other variables included hi the model. Furthermore, the difference between
predicted blood-lead concentrations for children exposed to dust-lead loadings of 5 and 50 |ig/ft2
is the same as that between children exposed to dust-lead loadings of 500 and 5000 ug/ft2 if the
values of the other variables are constant. Although the multiplicative interpretation of the log-
linear model is not considered biologically or physically plausible, for low to moderately exposed
children, the log-linear model often fits the data better than statistical models with a more
plausible, biological/physical basis (Rust et al, 1996; Jiang and Succop, 1996).
4.2.2 Variable Selection
The criteria used for the selection of predictor variables in the empirical model
emphasized use of measures of environmental lead and other factors observed hi both the
Rochester Lead-in-Dust Study and the HUD National Survey. Variables whose translation
between the two studies was straightforward, whose statistical relationship with blood-lead
concentration in the Rochester study was significant, and whose values in the HUD National
Survey covered a wide range, were used in the empirical model.
The predictor variables selected for the final model are described below. Each variable is
first defined as in the Rochester study for model development purposes. Next, the definition of
an analogous variable based on the HUD National Survey data is presented. This latter definition
was employed when applying the empirical model to the environmental data from the HUD
National Survey to estimate a national distribution of children's blood-lead concentrations.
Floor Dust-Lead Loading
The empirical model was developed using the natural logarithm of the area-weighted
arithmetic average (wipe) dust-lead loading from carpeted and uncarpeted floors in the Rochester
study. For this risk analysis, the natural logarithm of the area-weighted arithmetic average floor
dust-lead loading from 3 sample locations (wet, dry and entry rooms) in HUD National Survey
4-12
-------
homes (as measured using Blue Nozzle vacuum techniques) is used as the measure of lead in
floor dust.
Window Sill Dust-Lead Loading
The empirical model was developed using the natural logarithm of the area-weighted
arithmetic average (wipe) dust-lead loading from window sills hi the Rochester study. For this
risk analysis, the natural logarithm of the area-weighted arithmetic average (Blue Nozzle
Vacuum) dust-lead loading from window sills hi 2 sample locations (wet and dry rooms) in HUD
National Survey homes is used as the measure of lead in window sill dust.
Soil-Lead Concentration
The empirical model was developed using the natural logarithm of the dripline soil-lead
concentration (fine soil fraction) hi the Rochester study. Dripline soil samples hi the Rochester
study were thoroughly homogenized and sieved into coarse and fine fractions using a 2 mm mesh
sieve followed by a 250 urn mesh sieve. These two soil fractions were chemically analyzed
separately, and results from the fine soil fraction were selected for statistical analysis. For this
risk analysis, the natural logarithm of the weighted average concentration of samples collected
from dripline, entryway, and remote locations (with weights of 25%, 25%, and 50%,
respectively) from HUD National Survey homes is used as the measure of soil-lead
concentration.
Extent of Paint/Pica Hazard
The empirical model was developed using a paint/pica variable that took into account the
presence and condition of lead-based paint (LBP) hi the home and the tendency of the child to
ingest paint chips. The following question hi the Rochester study questionnaire was designed to
measure mouthing behavior or pica tendencies in resident children:
How often does the child put paint chips hi his/her mouth?
The possible responses to this question were: 0 = Never, 1 = Rarely, 2 = Sometimes,
3 = Often, and 4 = Always. For the empirical model, a categorical variable (paint/pica) was
constructed that was nonzero when the home contained some damaged or deteriorated ulterior
lead-based paint (determined by whether any paint had a condition of fair or poor) and the
response to the above pica question was 1 or greater (i.e., at least rarely). This variable was
defined to have values of 0,1, and 2, which were defined as follows:
0 No LBP present (maximum XRF reading < 1 mg/cm2) or condition of paint is rated
as Good or child does not exhibit pica;
1 LBP present (maximum XRF reading z 1 mg/cm2) and paint condition is Fair or
Poor and child exhibits pica rarely;
4-13
-------
2 LBP present (maximum XRF reading ^ 1 mg/cm2) and paint condition is Fair or
Poor and child exhibits pica at least sometimes.
For the Rochester study, condition of the paint was characterized as Good when less than 5% of
the surface was deteriorated, Fair when 5% to 15% of the surface was deteriorated, and Poor
when more than 15% of the surface was deteriorated.
For this risk analysis, the value of the paint/pica variable for each home in the HUD
National Survey is determined by a combination of the presence of deteriorated lead based paint
and an assumed 9% of children aged 1-2 years who exhibit pica for paint (Appendix Dl). For
homes with no deteriorated lead-based paint, the value of the paint-pica variable is set to zero.
For homes that were found to contain deteriorated lead-based paint, it was assumed that 9% of
children living in a similar environmental would ingest paint chips at some time. For those
children, the value of the paint/pica variable is set at 1.5, which is the average response to the
paint pica question in Rochester among children who exhibited pica for paint. The remaining
91% of the children living in homes with deteriorated lead-based paint are assumed to exhibit no
pica for paint. Thus, the paint/pica variable is set equal to zero for 91% of children living in
homes with deteriorated lead-based paint.
The development of the empirical model for this risk analysis is complicated by the fact
that the sampling methodology used to measure lead exposures in the HUD National Survey is
different from that used in the Rochester Lead-in-Dust Study. Specifically, two of the lead
exposure measurements from the HUD National Survey are blue nozzle vacuum floor dust-lead
loading and blue nozzle vacuum window sill dust-lead loading, compared to floor dust-lead
loading and window still wipe dust-lead loading hi Rochester. Thus, these variables have
different interpretations hi the two studies.
In addition, the soil variable from the HUD National Survey is the weighted average of
samples collected from dripline, entryway and remote locations (with weights of 25%, 25%, and
50%, respectively), whereas the soil variable from the Rochester study is based on a composite
sample from the dripline area only. Also, the paint/pica variable from the HUD National Survey
data was based on the measures of paint on both ulterior and exterior surfaces, whereas the
variable from the Rochester study was based on measures of paint on only interior surfaces.
Lead-based paint on deteriorated exterior surfaces was not considered hi the estimation of the
paint/pica model parameter based on Rochester data, because nearly every home surveyed in the
Rochester study had deteriorated lead-based paint on exterior surfaces. The differences in
paint/pica variable construction between the Rochester study and HUD National Survey are
considered minor in comparison to the differences in the dust-lead loading and soil variables.
4.2.3 Rochester Multimedia Model
As a first step in developing the empirical model, a multi-media predictive model was
developed using data from the Rochester Lead-in-Dust Study which explained children's blood-
lead concentration as a function of dust-lead loadings from floors and window sills, drip-line
soil-lead concentration and the paint/pica variable. The Rochester multimedia model was log-
4-14
-------
linear in nature, and specific details on model development are found in Appendix G. (The
model is referred to as "the Multi-Media Predictive model based on Rochester data" in
Appendix G.) Table 4-2 provides parameter estimates and associated standard errors for the
multimedia predictive model.
Table 4-2. Parameter Estimates and Associated Standard Errors for the Rochester
Multimedia Model
Parameter
Po
p,
P2
33
04
R2
°2Er,or
Variable Description
Intercept
log(PbF):Area-WeightedArithmeticMean(Wipe)Dust-Lead
LoadingfromAnyFloor(CarpetedorUncarpeted)
log(PbW):Area-weightedArithmeticMean(Wipe)Dust-
LeadLoadingfromWindowSills
log(PbS):DriplineSoil-LeadConcentration(finesoilfraction)
PbP:lndicatoroflnteriorPaint/PicaHazard
Coefficientof Determination
Error
Estimate
(Standard Error)
0.418(0.240)
0.066(0.040)
0.087(0.036)
0.114(0.035)
0.248(0.100)
21.67%
0.316
The Rochester multimedia model is used in the risk characterization (Chapter 5) to
determine the probability that a child exposed to specific levels of lead in paint, dust and soil will
have a blood-lead concentration at or above 10
4.2.4 Measurement Error Adjustment
The fact that the Rochester multimedia model lead exposure variables for paint, dust and
soil are subject to measurement error raises concerns about the need to account for this
measurement error in the model building process. The term "measurement error" is used to
describe uncertainty in the predictor variables attributable to sampling, spatial, laboratory and/or
temporal variability. The presence of measurement error in predictor variables, if not accounted
for in the statistical models, could result in biased predictions (Fuller, 1987). In addition,
because different sampling methods were used in the Rochester study and the HUD National
Survey, adjustments for those different sampling methods may be needed when applying the
empirical model to the HUD National Survey data.
The first question to be asked when addressing measurement error is: Is an adjustment
for measurement error necessary? The appropriateness of an adjustment for measurement error
depends on the use of the statistical model. One primary differentiation in model use concerns
whether the model is being used to characterize the relationship between observed blood-lead
levels in children and "true" lead exposures, or whether the model is being used to predict blood-
lead levels based on some source of measured levels of environmental lead. The primary use of
4-15
-------
the empirical model in the §403 mlemaking is for the latter case (prediction). Therefore, a
classic errors-in-variables adjustment was not considered necessary (Carroll et al., 1995).
However, to predict the national distribution of childhood blood-lead concentrations
(prior to and following implementation of §403 rules), the empirical model must be combined
with environmental data observed in a nationally representative sample (the HUD National
Survey). An empirical model unadjusted for the effects of differences in measurement error in
the lead exposure predictor variables would be appropriate for prediction of the national
distribution of blood-lead concentrations, if the following four assumptions were acceptable:
1. The sampling scheme for environmental lead implemented in the Rochester study (or
other studies used for model building) is similar to the sampling scheme
implemented in the HUD National Survey.
2. The sampling collection devices and instruments used to measure lead have similar
properties with respect to measurement error between the Rochester study and the
HUD National Survey.
3. The distribution of observed environmental lead levels is similar between the
Rochester study and the HUD National Survey.
4. The characteristics of the true exposure relationship hi the Rochester study is the
same as in the U.S. as a whole.
Investigation of the data from the Rochester study and the HUD National Survey
suggested that the first three assumptions were unacceptable. Therefore, an adjustment for the
differences in measurement error between predictor variables used in the model building process
and input variables from the HUD National Survey used in the prediction process is appropriate.
Although this can be considered an adjustment for "measurement error," the resulting model
should not be interpreted as the "true" relationship between blood-lead and environmental lead
exposure (measured without error). Rather, this adjustment accounts for the differences in
variability of the measured data in the two studies to facilitate a better prediction of the national
distribution of childhood blood-lead concentrations using the data from the HUD National
Survey.
If the fourth assumption is not acceptable, it is questionable whether the Rochester study
is an appropriate source of data for informed decisions concerning lead exposures nationwide.
There is no evidence to suggest that the fourth assumption is unacceptable.
4.2.5 Specification of the Empirical Model
When using the empirical model to predict a national distribution of children's blood-lead
concentrations, differences hi dust and soil variables between the Rochester study and the HUD
National Survey are accounted for by first establishing a relationship between blood-lead and
environmental variables, as measured by methods used in the Rochester study (the Rochester
4-16
-------
Multimedia Model), and then adjusting this relationship to use environmental variables as
measured in the HUD National Survey. The adjustment takes into account both systematic
differences and differences in error structures between the two sets of data and involves fitting a
classic errors in variables model (Carroll et al., 1995) as an intermediate step. The method
provides an empirical model of the relationship between blood-lead concentration and floor
and window sill dust-lead loadings and other covariates as observed in the HUD National
Survey.
In addition, the intercept of the empirical model was adjusted so that the geometric mean
of the predicted national distribution of children's blood-lead concentrations matches that
observed in Phase 2 of NHANES HI.
The final mathematical form of the empirical model is:
ln(PbB) =
t ln(PbFBN)
ln(PbWBN)
ln(PbS)
PbP
where PbB represents the blood-lead concentration, PbFBN and PbWBN correspond to average
dust-lead loadings from interior floors and window sills respectively (assuming the Blue Nozzle
vacuum technique), PbS represents average soil-lead concentration for the yard, PbP represents
paint/pica hazard, and e represents the residual error left unexplained by the model. These
predictor variables were introduced in Section 4.2.1 . Table 4-3 provides parameter estimates and
associated standard errors for the model parameters. The standard errors provided in Table 4-3
were estimated using a bootstrap algorithm. The empirical model is not intended to be used to
estimate the effect of a single medium on blood-lead levels. The model should only be used to
predict a distribution of blood-lead levels when environmental lead levels for all media are
known or estimated. Individual parameter estimates in Table 4-3 should not be interpreted in
isolation. Specific details on the development of the empirical model are found in Appendix G.
Table 4-3. Parameter Estimates and Associated Standard Errors for the Empirical Model
Used to Predict the National Distribution of Children's Blood-Lead
Concentration Based on Data from the HUD National Survey
Variable' ^Li-V'jM,
Intercept
FloorDust-LeadLoading(BlueNozzleVacuum)
WindowSiliOust-LeadLoading(BlueNozzle
Vacuum)
Soil-LeadConcentration(YardAverage)
Paint/Pica
Error
Parameter
3o
P,
P2
B3
fc
t^Error
Estimate
(Standard Error)
0.651(0.154)
0.032(0.044)
0.050(0.031)
0.094(0.043)
0.256(0.098)
0.313
4-17
-------
4.2.6 Estimating the National Distribution of Blood-Lead Using the Empirical Model
The empirical model is used to predict a national distribution of children's blood-lead
concentrations both before and after interventions resulting from the §403 standards.
Environmental conditions observed in the HUD National Survey are used as input to the
empirical model for predicting blood-lead levels in children 1-2 years old. A population of
children aged 1-2 years is both the representative population for this risk analysis and similar to
the age group that was recruited in the Rochester Lead-in-Dust Study (thus the empirical model is
representative of children in this age group).
The empirical model is used to estimate the average log-transformed childhood blood-
lead concentration associated with each home in the HUD National Survey. Input variables were
constructed from observed levels of lead in each residential unit, as described in Section 4.2.2,
for prediction of the current national distribution of children's blood-lead concentrations. If the
input values for the empirical model were missing for a home, then an imputed value is used in
the risk analysis. The imputed values are summarized in Table 3-14 of Section 3.3.1.1, with
more details provided in Appendix Cl.
To predict a post-§403 national distribution of children's blood-lead concentrations, the
following method was used to prepare soil- and dust-lead concentrations in the HUD National
Survey Data for input into the empirical model:
1. Observed levels of lead in environmental variables in the HUD National Survey were
compared to proposed §403 standards. Blue-nozzle vacuum floor and window sill
dust-lead loadings were converted to wipe dust-lead loadings before comparison to
the §403 standards.
2. §403 interventions were triggered in HUD National Survey residential units that had
levels of lead in environmental variables that were above the proposed standard. If
an intervention was triggered, assumed post-intervention lead levels in
environmental variables were substituted for observed levels.
The geometric mean blood-lead concentrations predicted by the empirical model, an assumed
geometric standard deviation of 1.6, and population weights adjusted to the 1997 population of
children (aged 1-2 years) are used to predict the pre- and post-§403 national distributions of
blood-lead concentrations.
4.3 UTILIZING DUST LEAD LOADINGS
The HUD National Survey is the only national survey of environmental lead levels and
therefore was used for prediction of a national distribution of blood-lead concentrations. Dust
lead measurements in the HUD National Survey were collected by the Blue Nozzle (BN) vacuum
method. However, §403 standards for dust will be expressed as a measured lead loading
collected by a dust wipe sample. As a result, the following conversions are necessary in the risk
analysis methodology:
4-18
-------
Converting Blue Nozzle dust-lead loadings observed in the HUD National Survey to
wipe-equivalent dust-lead loadings to determine the extent to which homes in the
United States are impacted by example options for the §403 dust-lead standard.
Converting post-intervention wipe dust-lead loadings to Blue Nozzle-equivalent
dust-lead loadings for input into the empirical model.
In addition, conversion factors were used to convert dust-lead loadings under the BRM
dust sampling method employed hi the Baltimore R&M Study to wipe equivalents for production
of prevalence tables in Chapter 3.
This section presents the equations for the above conversions. The conversion equations
are presented for dust samples collected from floors and window sills, since §403 rules will
include standards for those housing components. These equations were established using data
from environmental field studies where dust samples were collected by both sampling techniques
from adjoining ("side-by-side") sample areas. Detailed information concerning the development
of all the conversion equations discussed in this section, including a discussion of the studies and
sample sizes used to estimate the equations is available in USEPA, 1997.
The use of a measurement error adjustment for the conversion from Blue Nozzle vacuum
to wipe-equivalent dust-lead loadings is described in USEPA, 1997, as well. This adjustment
was required because the data employed to develop the conversion equations possessed different
distributional characteristics than the data to which the conversion equations were applied to
(HUD National Survey). Similar adjustments were not employed for the other conversions. For
the wipe to Blue Nozzle vacuum dust-lead loading conversion, there was not enough information
to determine whether an adjustment was appropriate, since the conversion is applied simply to
the assumed post-intervention wipe lead loadings. In the case of the BRM to wipe conversions,
the data sets were similar and no adjustment was needed.
4.3.1 Wipe Versus Blue Nozzle (BN) Vacuum Conversions
Three studies reported side-by-side wipe and BN vacuum dust-lead measurements:
1. CAPS Pilot Study (USEPA, 1995i)
2. National Center for Lead-Safe Housing (NCLSH)/Westat Study (Westat, 1995)
3. Baltimore Repair and Maintenance (R&M) Pilot Study (Battelle, 1992)
To obtain conversion factors from one collection method to another, log-linear regression
models were fitted to dust-lead loading data for each study separately. Weighted averages of the
parameter estimates from each model were used to obtain the following equations (written in the
scale of the original data) for conversions between wipe and BN vacuum sampling methods.
Confidence intervals and prediction intervals are provided in USEPA, 1997.
4-19
-------
1. Equations used to predict a wipe dust-lead loading from a BN vacuum dust-lead
loading:
Uncarpeted Floors:
Homes built prior to 1940: Wipeload = 5.66 (BNload)0809
Homes built 1940-1969: Wipeload = 4.78 (BN10J0800
Homes built 1960-1979: Wipeload = 4.03 (BNload)°707
Window Sills:
All homes: Wipeload = 2.95 (BNload)U8
Note that differences among the three categories of houses determined by age
prompted different conversion equations for dust-lead loading on uncarpeted floors,
but different equations were not necessary for window sills. As an example of using
these equations, a BN dust-lead loading of 100 ug/ft2 on an uncarpeted floor in a
house built prior to 1940 would be converted to a wipe dust-lead loading of 235
Hg/ft2. In developing these equations, the observed BN lead loadings ranged from
1.0 to 2,164 ng/ft2 on floors, and 1.4 to 8,964 jig/ft2 on window sills. Extrapolation
is necessary for BN loadings outside this range.
2. Equations used to predict a BN vacuum dust-lead loading from a wipe dust-lead
loading:
Uncarpeted Floors:
All units: BNload = 0.185 (Wipeload)° 931
Window Sills:
All units: BNload = 0.955 (Wipeload)°583
Thus, for example, a wipe lead loading of 100 ug/ft2 on an uncarpeted floor would be
converted to a BN lead loading of 13.5 jig/ft2. In developing these equations, the
observed wipe lead loadings ranged from 7.6 to 6,755 ng/ft2 on floors, and from 3.0
to 425,000 ng/ft2 on window sills.
4.3.2 Wipe Versus Baltimore Repair and Maintenance (BRM) Vacuum Conversions
In characterizing dust-lead loadings in the Baltimore R&M Study (Section 3.2.2.1) for
this risk analysis, it was necessary to express the loadings relative to wipe collection techniques,
rather than BRM vacuum techniques. Therefore, it was necessary to convert BRM dust-lead
loadings to wipe-equivalent loadings to obtain the summary statistics provided hi Chapter 3.
Four studies reported side-by-side wipe and BRM vacuum dust-lead measurements:
4-20
-------
1. R&M Mini Study (Farfel, 1994)
2. Rochester Lead-in-Dust Study (USHUD, 1995a)
3. NCLSH 5-Method Comparison Study (Westat, 1995)
4. Milwaukee Low-Cost Intervention Study (USEPA, 1997)
These studies are described in USEPA, 1997.
An analogous approach to that presented in Section 4.3.1 was used to develop the
following equations for predicting a wipe dust-lead loading from a BRM vacuum dust-lead
loading:
Uncarpeted Floors: Wipe = 8.34 BRM °371
Carpeted Floors: Wipe = 3.01 BRM °227
Window Sills: Wipe = 14.8 BRM °453
For instance, a BRM dust-lead loading of 100 ug/ft2 on an uncarpeted floor would be converted
to a wipe dust-lead loading of 46.0 ug/ft2. In developing these equations, the observed BRM lead
loadings ranged from 0.1 to 74,100 ug/ft2 on uncarpeted floors, from 1.4 to 141,000 ug/ft2 on
carpeted floors, and from 0.3 to 4,170,000 (ig/ft2 on window sills.
Note that the floor dust-lead samples in the Baltimore R&M Study were collected as
composite samples (i.e., sub-samples from different locations combined into a single sample),
which eliminates the ability to distinguish uncarpeted floor samples from carpeted floor samples.
However, it was possible to determine the number of uncarpeted and carpeted subsamples within
each composite sample. Therefore, floor dust-lead loadings from the Baltimore R&M Study
were converted to wipe-equivalent loadings as follows:
Wipe = p 8.34 BRM0371 + (1-p) 3.01 BRM0227,
where p represents the proportion of the composite sample obtained from uncarpeted floors, and
BRM represents the dust-lead loading under BRM vacuum sampling techniques. For example, a
BRM dust-lead loading of 100 ug/ft2 in a composited floor-dust sample consisting of 3
uncarpeted and 2 carpeted subsamples would be converted to a floor wipe dust-lead loading of
31.1 ug/ft2.
4.4 HEALTH OUTCOMES
This section presents the approach for determining the incidence of adverse health
outcomes resulting from lead exposure in young children and characterizes the relationship of
certain elevated blood-lead concentration thresholds and other health effect endpoints to blood-
lead concentrations. These relationships are applied to predicted geometric mean blood-lead
concentrations, as predicted by the IEUBK and empirical models, to relate environmental lead
exposure to health effects. The risk characterization in Chapter 5 and risk management analysis
in Chapter 6 apply these relationships to the predicted national distribution of blood-lead
concentrations to estimate incidences of elevated blood lead concentrations and health effects hi
children aged 1-2 years. For example, the relationship between blood-lead concentration and IQ
4-21
-------
scores is used to estimate the average IQ point loss due to lead exposure and the percentage of
children with IQ point decrements greater than or equal to one, two, or three IQ points.
4.4.1 Decrements in IQ Scores
The IQ point loss health effect represents the neurological loss that a child experiences
due to low level lead exposure. The relationship between blood-lead concentration and IQ point
decrements has received considerable study, as described in Section 2.3. Multiple attempts have
been made to quantify the effect using meta-analysis to combine the varying estimates reported in
the scientific literature (Schwartz, 1993; Pocock et al., 1994; Schwartz, 1994). Meta-analysis is
an often used statistical technique that is used to combine the results of statistical summaries and
inferences across multiple studies. Several estimates of the relationship between blood-lead
concentration and IQ scores were identified, one of which is applied within this risk analysis.
Two additional estimates are used hi the sensitivity analysis (Section 5.4.2) to determine the
extent to which the uncertainty in estimating this parameter affects the characterization of risk.
The approach to developing the estimate used in the risk analysis is described below.
Schwartz (1994) conducted a random effects meta-analysis to quantify the relationship
between blood-lead concentrations and IQ scores. The results from seven studies were employed
to characterize the decrease in IQ score associated with increased blood-lead concentration. The
three longitudinal and four cross-sectional studies included in the meta-analysis are summarized
in Table 4-4. Each study estimated the effect that blood-lead concentration has on full-scale IQ
score in primary school age children. Estimates of the IQ score decrease from three of the
studies were not statistically different from zero. Additional details are provided in Appendix
D2, Tables D2-1 and D2-2. A summary of the Schwartz (1994) article and a comparison of the
results to those reported in similar papers (Schwartz, 1993; Pocock, et al, 1994) are also
presented in Appendix D2.
The seven studies used linear or log-linear regression models to model the relationship
between IQ scores and childhood blood-lead levels, along with other potentially important
covariates. A log-linear regression model is a regression model fitted to the logarithm of the
independent variables; in this application the independent variable is blood-lead concentration
and the dependent variable is IQ score. Three of the studies included in the meta-analysis
employed log-linear models, while the four remaining studies employed linear models. Schwartz
conducted a threshold analysis, to determine whether there is a level, below which a relationship
between blood-lead concentration and IQ score is not apparent. On the contrary, Schwartz
concluded that the slope appears to be steeper at lower blood-lead concentrations. This
conclusion is consistent with the log-linear form of the regression model. Despite this, the linear
relationship is assumed in this risk analysis. The assumption of a linear model reduces the
likelihood of overestimating the number of children with low blood-lead concentrations at risk,
or who may benefit from actions taken in response to the §403 standards.
4-22
-------
Table 4-4. Summary Information for Studies Included in the Schwartz (1994) Meta-
Analysis.
Study
Hawk.etal
(1986)
Hatzakis.etal
(1987)
Fulton, etal
(1987)
Yule.etal
(1981)
Bellinger,etal
(1992)
Dietrich, etal
(1993)
Baghurst,
etal(1992)
Number
of
Children
75
509
501
166
147
231
494
Blood-Lead Concentration
(pg/dL)
Range
6.2-47.4
7.4-63.9
3.3-34.0
7.0-33.0
0-253
na
<12.2-
>28.2
Mean (SD)
20.9(9.7)
23.7(9.2)
11.52
13.5(4.1)
6.5(4.9)
15.2(11.3)
20(na)
Estimated
Effect1
on IQ Score
(SE)
2.55(1.5)
2.66(0.7)
2.56(0.9)
5.6(3.2)
5.8(2.1)
1.3(0.9)
3.33(1.5)
Other Study Information
Cross-sectionalstudyof
childrenage3-7inLenoirand
NewHanovercounties,NC;
Linearregressionmodel
Cross-sectionalstudyof
primaryschoolagechildrenin
aleadsmeltercommunity
( La vrion, Greece);
Linearregressionmodel
Cross-sectionalstudyof
primaryschoolagechildrenin
Edinburgh, Scotland;
Log-linearregressionmodel
Cross-sectionalstudyof
primaryschoolagechildrenin
London, England;
Log-linearregressionmodel
LongitudinalstudyinBoston,
MA;Bloodleadatage2;IQ
measuredatschoolage;
Linearregressionmodel
Longitudinalstudyin
Cincinnati.OH; Integrated
bloodleaduptoage3; 1 Q
measuredatschoolage;
Linearregressionmodel
LongitudinalstudyinPort
Pirie,Australia;lntegrated
bloodleaduptoage3;IQ
measuredatschoolage;
Log-linearregressionmodel
1 EffectrepresentsaveragedeclinesinlQpointsassociatedwithanincreaseinblood-leadconcentrationfrom
10/yg/dLto20/;g/dL.
2 GeometricMeanwasreportedforthisstudy.
3 Theexactrangewasnotreported.The90thpercentilewas12.5/yg/dLandallchildrenwerebelow25
fjg/dL.
4-23
-------
Based on the modeled relationships reported for each study, Schwartz concluded that a
doubling of blood-lead concentration from 10 ug/dL to 20 ug/dL results in a loss of 2.57 IQ
(SE = 0.41) points, on average. Therefore, a 1 ug/dL increase in blood-lead concentration results
in a loss of 0.257 IQ points, on average. (For an individual child, a greater or lesser IQ point loss
may be observed.) This relationship is most applicable for blood-lead concentrations between 10
and 20 |ig/dL, because of the modeling assumptions made in the studies that used log-linear
models. However, the relationship is applied over a much broader range of blood-lead
concentrations in the risk analysis. Similar effects were observed in the studies that employed
linear and log-linear regression models, as shown in Table 4-4. In addition, the blood-lead
concentrations ranged from <6.2 ug/dL to 63.9 ug/dL in the studies that employed linear
regression models.
The relationship between environmental-lead levels and IQ point loss is presented hi
Figure 4-3, utilizing predicted geometric mean blood-lead concentrations from the IEUBK
model. For each curve, the soil- or dust-lead concentrations were varied over a range of values,
while all other IEUBK model parameters were held fixed as described in Section 4.1. Then the
predicted geometric mean blood-lead concentration from the IEUBK model was used to estimate
the average IQ point loss using the relationship established above. For example, approximately
2.3 IQ points are expected to be lost as a result of exposure to a soil or floor dust-lead
concentration of 1,000 ppm.
(A
C *
3
Q.
2 3
1000 2000 3000 4000
Lead Concentration In Soil or Oust (ppm)
Fixed Sons 100 ppm: Vorlobl. Dud
Fixed Dutt=200 ppm: Variable Soil
5000
Figure 4-3. Estimated IQ Point Loss Due to Lead Exposure Plotted Against
Concentration of Lead in Soil and Dust, Utilizing IEUBK Model
Predictions to Relate Environmental Lead to Blood Lead
4-24
-------
The relationship between IQ score decrement and blood-lead concentration is used in this
risk analysis to estimate the average IQ point decrement for children exposed to various
environmental-lead levels. Of interest is the extent to which average IQ score decrement is
reduced upon promulgation of the §403 rule. Both the IEUBK and the empirical models are used
to estimate the distribution of blood-lead concentrations of children exposed to a given set of
environmental conditions observed in the HUD National Survey homes. The predicted blood-
lead concentrations are multiplied by 0.257 to estimate the corresponding IQ point loss due to
lead exposure for children exposed to these conditions. Then, the average IQ decrement and
percentages of children with decrements of £ 1, ^2, and ^3 IQ points due to lead exposure are
calculated.
4.4.2 Increased Incidence of IQ Scores Less Than 70
The increased incidence of IQ scores less than 70 resulting from lead exposure represents
an increased likelihood of mental retardation resulting from lead exposure. An IQ of 70 is two
standard deviations below the population mean IQ of 100 and can be used as an indicator of
mental retardation. Children who are mildly mentally retarded require special education classes
in school. Children who are severely mentally retarded may require life-long institutional care.
There are limited data available to estimate the increased likelihood of mental retardation
resulting from lead exposure. Because of the lack of data, Wallsten and Whitfield (1986) used
judgmental probability encoding methods to assess health risks due to lead exposure, particularly
in the area of lower IQ scores. As part of this assessment, the increased percentage of children
having IQ scores less than 70 was estimated for populations of children with elevated blood-lead
levels. Judgmental probability encoding methods rely on expert judgement to estimate the effect
of interest and are not applied when sufficient data are available to make the estimate.
In the Wallsten and Whitfield study, care was taken to select experts whose opinions
spanned the range of respected opinion. The six experts who participated in the assessment of
the relationship between IQ scores and blood-lead levels are listed in Table 4-5. These experts
were asked to consider a hypothetical experiment in which a large number of children were
randomly assigned at birth to either a control group, or one of six lead-exposure groups. Lead
exposure was to remain fixed until the children reached age seven, at which time the Wechsler
Intelligence Scale for Children - Revised (WISC-R) IQ test would be administered. Blood-lead
levels were to be measured at age three. The lead exposure levels were such that at age three,
members of each of the lead-exposure groups had blood-lead levels of 5,15,25,35,45, and 55
ug/dL. The experts were asked to estimate the mean and standard deviation of IQ scores in the
control group. The experts also estimated the expected mean IQ differences between the control
group and each exposure group. Each expert assumed that the IQ standard deviation hi exposure
groups was the same as that of the control group. This information was used to estimate the
increased percentage, due to lead exposure, of children having IQ scores less than 70.
4-25
-------
Table 4-5. Experts Who Participated in the Assessment of the Relationship Between
IQ Scores and Blood-Lead Levels by Wallsten and Whitfield.
Expert
KimDietrich
ClaireErnhart
HerbertNeedleman
MichaelRutter
GerhardWinneke
WilliamYule
Affiliation
Universityof Cincinnati
ClevelandMetropolitanGeneralHospital
Universityof Pittsburgh
Instituteof Psych iatry,London,UK
Universityof Dusseldorf,Dusseldorf,WestGermany
Instituteof Psychiatry , London, U K
If the expert thought it necessary, separate judgements were made according to
socioeconomic status (SES). For this purpose, low SES was defined as children living in
households with incomes at, or below, the fifteenth percentile; and high SES was defined as
children living in households with incomes above the fifteenth percentile. Five of the six experts
chose to make separate judgements based on socioeconomic status.
At blood-lead levels ranging from 2.5 to 27.5 ug/dL, the distribution of increased
percentage of children having IQ scores less than 70 was reported by Wallsten and Whitfield, for
each expert and SES (low and high). These distributions were combined by calculating the
weighted average of the low SES median (15% of weight) and high SES median (85%). The
increased percentage of children having IQ scores less than 70, due to lead exposure, was
estimated from the weighted average of the medians as a piecewise linear function of blood-lead
concentration. This function is reported in Table 4-6 and illustrated in Figure 4-4, over a range
of blood-lead concentrations. For example, 0.6% = (-0.281 + 0.0432 x 20) of children with
blood-lead concentrations of 20 ug/dL are expected to have IQ scores less than 70 due to lead
exposure, above and beyond those whose IQ would naturally fall below that level. The
relationship is extrapolated to include blood-lead concentrations below 2.5 and above
27.5 ug/dL.
Table 4-6. Piecewise Linear Function for Estimating the Judged Increased Percentage of
Children Having IQ Scores Less Than 70 Due to Lead Exposure.
Range of Blood-Lead (PbB) Levels
U/g/dU
0
-------
2.0
o
1
2 1.5
C
" '-0
"o
_g
c
£ O.S
0.0
10 20 30
Blood-Lead Concentration (yug/dL)
40
Figure 4-4. Judged Increase in Percentage of Children with IQ Below 70 Due to Lead
Exposure, Plotted Against Blood-Lead Concentration
The relationship between environmental-lead levels and the increased percentage of
children having IQ scores less than 70 is presented in Figure 4-5, utilizing EEUBK model
predicted geometric mean blood-lead concentrations. For each curve, the soil- or dust-lead levels
were varied over a range of values, while all other parameters were held fixed. The predicted
geometric mean blood-lead concentration from the IEUBK model was used to estimate the
increased percentage of children with IQ scores less than 70 due to lead exposure. For example,
an additional 0.2% of children exposed to soil- or floor dust-lead concentrations of 1,000 ppm
would be expected to have IQ scores less than 70 as a result of the exposure.
Figure 4-6 illustrates the relationships between geometric mean blood-lead concentration
and the predicted percentage of children with a blood-lead concentration greater than or equal to
10 and 20 ug/dL, over a range of geometric mean blood-lead levels. This relationship was
computed assuming a geometric standard deviation of 1.6 ug/dL and that blood-lead
concentrations have a log-normal distribution. The same assumptions are applied hi the risk
characterization (Chapter 5). The relationships between lead concentrations in soil and dust and
the incidence of blood-lead levels greater than or equal to 10 and 20 ug/dL are illustrated in
Figure 4-7, utilizing geometric mean blood-lead concentrations as predicted by the EEUBK
model. For each curve, the soil- or dust-lead concentrations were varied over a range of values,
while all other model parameters were held fixed. The BEUBK model predicted geometric mean
blood-lead concentration and the geometric standard deviation of 1.6 were used to calculate the
percentage of children exposed to these environmental conditions who would have blood-lead
concentrations greater than or equal to 10 and 20 ug/dL. As can be seen hi Figure 4-7,
approximately 40% of children exposed to soil or dust-lead concentrations of 1,000 ppm are
4-27
-------
2.0
_o
0>
CO
£
2
o
"5
I
c
1.0
0.5
o
8
o
u
0.0
1000 2000 3000 4000
Lead Concentration In Soil or Dust (ppm)
Fixed So)l= 100 ppm; Vorfablt Quit
Flxtd Dust= 200 ppm; Variabl* Soil
5000
Figure 4-5. Increase in Percentage of Children with IQ Below 70 Due to Lead Exposure
Plotted Against Concentration of Lead in Soil and Dust, Utilizing IEUBK Model
Predictions to Relate Environmental Lead to Blood Lead.
o
"6
100
90
80
70
60
SO
I -
e
£ 30
20
10
10 20 30
Geometric Mean Blood Lead Concentration (/Jg/dL)
PbB>10pg/dL - - PbB > 20/ig/dL
40
Figure 4-6. Percentage of Children with Blood-Lead Concentration *10 and 20 //g/dL Due
to Lead Exposure Plotted Against Geometric Mean Blood-Lead Concentration,
Assuming a GSD of 1.6.
4-28
-------
100
90
80
i "'
i »
u
-
30
20
10
0
1000 2000 3000 4000
Lead Concentration In Soil or Dust (ppm)
Flxtd Soll=100 ppm: Variable Quit
Fixed Du«f«200 ppm; Variabl* Soil
5000
Figure 4-7. Percentage of Children with Blood-Lead Concentration 210 and 20 //g/dL Due
to Lead Exposure Plotted Against Concentration of Lead in Soil and Dust,
Utilizing IEUBK Model Predictions to Relate Environmental Lead to Blood Lead.
expected to have blood-lead concentrations of at least 10 u.g/dL, whereas fewer than 5% of these
children are likely to have blood-lead concentrations exceeding 20 ng/dL.
4.5 DOSE RESPONSE CHARACTERIZATION
This chapter summarized the approach taken to establish the relationship between
exposures to lead hi dust, soil, and paint and childhood blood-lead concentration and health
effect endpoints. This relationship is used hi the risk characterization (Chapter 5) to describe the
risk to children aged 1-2 years under present environmental conditions. The relationship is also
used hi the risk management analysis (Chapter 6) to estimate the risk to children aged 1-2 years
under candidate §403 standards. Establishing this relationship is very problematic, because only
limited data exist relating specific health outcomes directly to environmental-lead levels.
Most environmental lead studies relate measures of residential lead exposure to measures of body
lead burden, rather than directly to health effects. In addition, most studies of health effects of
lead exposure relate specific health outcomes to measures of body lead burden, rather than
directly to environmental-lead levels. Therefore, it is necessary to establish the relationship
1-lead levels and health outcomes for this risk analysis in two steps. First
models. Then, incidence of elevated blood-lead concentrations and health effect risks are
estimated from those blood-lead concentrations.
4-29
-------
Relationship Between Environmental Lead and Blood Lead
Two models are applied to relate environmental-lead levels to blood-lead concentrations.
the IEUBK model and the empirical model. The IEUBK model takes user inputs for exposure
and uptake through a biokinetic process of distributing the lead to key tissues to predict a
distribution of blood-lead concentrations in children exposed to a specific combination of
environmental conditions. The precursor to the biokinetic part of the IEUBK model was
developed in 1985 as a tool for setting air lead standards and that version of the model was peer
reviewed and found acceptable by EPA's Science Advisory Board. The most current version of
the IEUBK model is used in this risk analysis. Although the IEUBK model was developed for
point source applications, it is being used in this risk analysis to predict a national distribution of
blood-lead concentrations.
The empirical model is a log-linear regression model, developed using data from the
Rochester Lead-in-Dust Study to estimate the relationship between blood-lead concentrations hi
young children and observed lead levels in then" primary residence. This model was developed
specifically for this risk analysis and has not yet undergone peer review.
While the two models provide useful alternative views of the relationship between
environmental and blood lead, neither model is optimal for this risk analysis. For example, the
IEUBK model utilizes dust-lead concentrations, while the §403 standards for dust will be defined
hi terms of dust-lead loadings. Furthermore, the IEUBK model does not include a direct
mechanism for the contribution of lead-based paint to childhood blood-lead levels (i.e., pica for
paint). Thus estimated blood-lead concentrations are adjusted hi homes with damaged lead-
based paint to reflect this exposure pathway. This adjustment for paint pica has not undergone
peer review. Although the empirical model was developed specifically for this risk analysis,
different sampling methods were used hi the Rochester study, upon which the empirical model is
based, and the HUD National Survey, which is used for predicting the national distribution of
blood-lead concentrations. Despite these shortcomings, the use of two different modeling
approaches provides a more robust analysis for the risk analysis than either approach alone.
Relationship Between Blood Lead and Health Endpoints
Both the IEUBK model and the empirical model are used to predict the national
distribution of blood-lead concentrations of young children. Incidence of elevated blood-lead
concentration endpoints are calculated from the predicted national distributions, with no further
modeling steps required.
While the existence of a relationship between decreased IQ scores and increased blood-
lead concentrations is generally accepted hi the scientific community, the quantification of the
relationship is more problematic. Several estimates of the relationship between IO scores and
blood-lead concentrations were considered. The most central and most widely accepted of these
is utilized hi the risk assessment to calculate the average IO decrement and the numbers of
children with IO decrement of ^ 1. ^2. and ^3 points due to lead exposure. Two additional
4-30
-------
estimates are utilized in the sensitivity analysis (Section 5.4.2) to determine the extent to which
the uncertainty in this parameter affects the risk characterization.
The relationship between blood-lead concentrations and the remaining health endpoint,
the increased incidence of IQ scores below 70 due to lead exposure, was estimated using a
piecewise linear function based on the distributions of IQ scores estimated by 6 experts who
participated in a 1985 study that utilized judgmental probability encoding methods. Because of
the lack of data to estimate this relationship, the use of expert judgement was unavoidable.
Impact on Risk Characterization
For the risk characterization, levels of lead in dust, soil, and paint from the HUD National
Survey are provided as inputs to each of the models described above for the prediction of the
national distribution of blood-lead concentrations for children aged 1-2 years. Although the
HUD National Survey is the most comprehensive survey of residential-lead levels available, the
application of these modeling tools to the HUD National Survey data is a limitation.
An important assumption in risk assessment is that the soil- and dust-lead concentrations
represent a child's actual lead exposure. For site-specific risk assessments, children's activity
patterns are used to guide the selection of sampling locations. It is uncertain whether the HUD
National Survey data represent typical childhood lead exposure levels.
The empirical model was developed from environmental lead measures from a single
study in one city and is being applied to nationally representative data. The empirical model
predictor variables are similar, by design, to those available in the HUD National Survey.
However, differences in sampling protocols exist between the studies, resulting in important
differences in variables used to develop the empirical model and the HUD National Survey
variables used for prediction. Most important is that dust samples were collected using different
techniques in the two studies. In addition, dripline soil samples were used in developing the
empirical model, while both dripline and remote area soil samples from the HUD National
Survey are used for prediction. Information on ulterior paint condition and pica tendency was
used hi developing the empirical model, but both interior and exterior paint condition are used
for prediction. Also, because no information on pica tendency was available for HUD National
Survey homes, the proportion of children who exhibit pica for paint was estimated and this
proportion was applied in all homes with damaged lead-based paint. To address these
differences, the empirical model includes an adjustment for measurement error that takes into
account both systematic differences and differences in error structures between the Rochester and
HUD National Survey studies.
Impact on Risk Management Analysis
The IEUBK and empirical models are used in the risk management analysis to predict
changes in blood-lead concentrations associated with reductions in environmental lead. In
addition to the concerns described above, use of these tools hi a post-§403 environment required
developing relationships between different dust-lead measurements. These relationships are used
4-31
-------
only in the risk management analysis. Use of these relationships to convert lead levels from one
sampling method to another is a weak link in the risk management analysis.
The following conversions are applied in the analysis of example options for risk
management:
\_i from pre-intervention blue nozzle (BN) vacuum lead loadings to pre-intervention
wipe lead loadings, to determine whether an intervention is required:
2t from post-intervention wipe lead loadings to BN vacuum lead loadings, for input
to the empirical model:
A great deal of uncertainty is associated with these conversions. (1) There was very little
data upon which to base the conversions. The scarcity of data results in highly variable
parameter estimates and makes it likely that individual data points may influence the analysis.
The method for developing the conversion equations was designed to minimize the effect of
influential observations and to account for differing variability across studies. (2) The range of
data used to develop the conversions does not span the range of the HUD National Survey data.
Thus, extrapolation is required to convert the lower portion of HUD National Survey BN lead
loadings to wipe lead loadings. Fortunately, the affected homes are not expected to exceed any
realistic set of standards. (3) There is considerable variability inherent in wipe lead loading
measurements. The sensitivity analysis in Section 6.4 considers the effect of the uncertainty
associated with these standards on the evaluation of risk management options.
4-32
-------
5.0 RISK CHARACTERIZATION
CHAPTER 5 SUMMARY
This chapter characterizes risks associated with current residential lead
exposures (pre-%403) for children aged 1-2 years. Risks are quantified by
predicting incidences of selected health effect and blood-lead concentration
endpoints. Baseline risk is computed using NHANES III, Phase 2 survey data.
Alternative risk estimates are computed as a function of environmental-lead levels
using HUD National Survey data. The tools developed and presented in Chapter 4
are used to produce, from the observational data, the risk estimates presented in
this chapter.
Estimates of risk to 1-2 year olds due to background lead exposure levels are
also presented in this chapter. These estimates help to quantify the maximum
possible reduction in adverse health effects which could be achieved by any type of
$403 rule. These are useful for comparison with the other risk estimates of this
chapter and Chapter 6.
Because individual children are exposed to specific residential environments
having specific (possibly known) levels of environmental lead, risk to children
exposed to specific environmental-lead levels are presented. These estimates are
based on the IEUBK model and the Rochester multimedia model presented in
Chapter 4.
Sensitivity and uncertainty analyses are performed in this chapter to gauge the
robustness of the risk characterization methodology to minor changes in
assumptions. The risk characterization is sensitive to the assumed relationship
between blood-lead concentration and IQ score decrements. Risk estimates based
on the HUD National Survey data are also sensitive to this relationship, to the
adjustment made to tap weights in the HUD National Survey dust samples, and to
assumptions on the geometric standard deviation associated with blood-lead
concentrations under specific exposure scenarios.
Figure 5-1 outlines the approach for the risk characterization. Risk
characterization conclusions, which include conclusions from hazard identification,
exposure assessment, and dose response assessment, are presented in Section 5.5.
This chapter answers four questions:
1. How can the material presented in the Hazard Identification, Exposure Assessment,
and Dose-Response Assessment chapters be integrated to characterize health risks to
young children due to residential exposures to lead in paint, dust, and soil?
2. What are the health risks to children exposed to specific levels of environmental
lead?
5-1
-------
Background
and
Objectives
Hazard
Identification
Exposure
Assessment
Dose-Response
Assessment
t
. RISK r , <\v^>; . , ;... ;.-- "-..-...,.:
CHARACTERIZATION KC ^s--S*
-'
Conduct the
Integrated
Risk Analysis
(Section 5.1)
' I
^f
Conduct a
Baseline Risk
Characterization
Using
NHANES III
(Section 5.1.1)
*^j
~y
Perform en
Alternative Risk
Characterization
(Section 5.1.2)
^ I -.-
Estimate « , .- ^ -
Health Effects T
Distributions " **
Prior to 5403 """" *
(Section 5.1) -
:" .y?1: StA^''*;-'-"'!'-, ' - - -
Estimate '
Risks Due to
Background '
Exposure .
(Section 5.2)
-t
r
s
Determine r
Individual «
Risks
(Section 5.3) ^
- Conduct Risk
Characterization
Sensitivity/
Uncertainty
Analysis
(Section 5.4)
". i .,.. - i
Determine
IEUBK Model-
Based Blood-Lead
Distributions at
Background
Exposures
(Section 5.2)
Estimate
Background
Heelth Effects -
(Section 5.2)
T
Risk
Characterization
Conclusions
(Section 5.5)
...._
Use IEUBK
and Rochester
Multimedia Model
to Estimate Blood-
Lead Distributions
at Specific
Exposure Levels
(Section 5.3)
Estimate
Health Effects
at Specific
Exposure Levels
(Section 5.3)
Present Blood-
Lead and Health
Effect Distribu-
tions Under
Alternative
: Assumptions
(Section 5.4)
Present
Conclusions on
Sensitivity/
Uncertainty
Analysis
(Section 5.4)
-I
_J
1
r
Risk
Management
1
r
Conclusions on
Analysis of Example
Options for 5403
Standards
...
Figure 5-1. Risk Characterization Overview.
5-2
-------
3. How does uncertainty in each step of the risk characterization affect estimates of
health risks to young children?
4. What are the overall conclusions of the risk assessment?
To answer the first question, Section 5.1 presents estimated prevalences of selected
adverse health effect and blood-lead concentration endpoints among the nation's 1-2 year olds,
projected to the year 1997. Where possible, 95% confidence intervals are presented to
characterize the uncertainty associated with the estimates. Each component of the risk
assessment is integrated to produce this characterization. The Hazard Identification component
is used to select the indicators of risk, the health effect and blood-lead concentration endpoints.
The Exposure Assessment component is used to characterize the exposure of children, and the
Dose-Response Assessment component is used to translate exposure into health effect and blood-
lead concentration endpoints.
To provide the reader a basis of comparison for current risk estimates presented in
Section 5.1, risks due to childhood lead exposure at background lead levels are estimated in
Section 5.2. Risks are estimated as the percentage of the nation's 1-2 year olds expected to
experience selected health effect and blood-lead concentration endpoints. In this context,
background levels of environmental lead are defined as levels which might have existed if
humans had not introduced lead into the environment through leaded gasoline consumption,
lead-based paint usage, and various other activities.
Section 5.3 presents a risk characterization for children exposed to specific levels of
residential environmental lead. The tools used to construct the individual risk estimates are the
appropriate blood-lead concentration/environmental-lead exposure relationships presented in
Chapter 4. Because the empirical model was developed specifically to characterize population
risk based on HUD National Survey data, the Rochester multimedia model and the IEUBK
model were used to characterize individual risks rather than the empirical model. Where
possible, 95% confidence intervals are presented to characterize the uncertainty associated with
the estimates. This information serves to answer question 2 above.
Sensitivity analyses in Section 5.4 are performed to gauge the uncertainty associated with
the estimated adverse health effect and blood-lead concentration endpoints due to methodological
assumptions. These analyses focus on potential weaknesses identified in the risk characterization
process. The sensitivity analyses were designed to produce a range of estimates for an unknown
parameter within which the true value of the parameter may reasonably be expected to fall.
Figure 5-1 provides an overview and roadmap to this chapter. The overall risk
assessment conclusions are discussed in Section 5.5.
5-3
-------
5.1 INTEGRATED RISK ANALYSIS
This risk analysis characterizes risks associated with childhood lead exposure by
predicting the incidence (both number and percentage) of selected health effect and blood-lead
concentration endpoints among 1-2 year old children. In this section, risks are projected for
1997, prior to EPA proposing §403 standards, using pre-1997 data adjusted, when possible, to
reflect current conditions. Risks due to childhood lead exposure characterized in this section are
population-based risks. The particular risk endpoints chosen for characterization of risk were
first presented in Section 2.4. These endpoints are
Incidence of blood-lead concentration (PbB) greater than or equal to 20 (ig/dL
Incidence of blood-lead concentration greater than or equal to 10 |ig/dL
Incidence of IQ score less than 70 resulting from lead exposure
Incidence of IQ score decrement greater than or equal to 1 resulting from lead
exposure
Incidence of IQ score decrement greater than or equal to 2 resulting from lead
exposure
Incidence of IQ score decrement greater than or equal to 3 resulting from lead
exposure
Average IQ decrement in a child, resulting from lead exposure.
As discussed in Section 2.4, blood-lead concentration endpoints are not health effects, but serve
as surrogates for a number of lead-associated health effects.
Because this risk analysis attempts to estimate population-based risks for the U.S.
population of 1-2 year olds, nationally representative data are required. However, nationally
representative data for which the selected endpoints can be estimated directly are not available.
There are, however, two studies measuring different aspects of lead exposure whose results are
nationally representative. These studies, the HUD National Survey and NHANES ffl, are
discussed in Sections 3.3 and 3.4, respectively, of the Exposure Assessment chapter. The former
study measures environmental lead exposure from residential dust, soil, and paint. The latter
study provides information on children's blood-lead concentrations. While the HUD National
Survey was conducted in 1989-1990, and Phase 2 of NHANES HI was conducted from 1991 to
1994, data from both studies are being used in this risk analysis to characterize children's lead
exposure in 1997, adjusted, when possible, to reflect current conditions. For example, this risk
analysis has updated the sampling weights associated with the HUD National Survey units to
make the model-predicted risk estimates more representative of conditions in 1997. (See Section
3.3 for an overview and Appendix Cl for details.) It would be ideal to use data collected in 1997
to characterize 1997 health risks to young children, but no such data were available.
5-4
-------
When blood-lead concentration data are directly available (i.e., NHANES HI), incidences
of the health effect and blood-lead concentration endpoints are calculated from the distribution of
blood-lead concentrations characterized by these data, using the methods hi Section 4.4.
However, this risk analysis also uses statistical models to characterize the blood-lead
concentration distribution as a function of environmental-lead levels from the HUD National
Survey. Once blood-lead concentration is predicted from environmental exposure using either
the IEUBK model (Section 4.1) or the empirical model (Section 4.2), incidences of the selected
endpoints are calculated from the blood-lead concentration distribution, using methods hi
Section 4.4.
The risk estimates computed directly from blood-lead concentration data collected in
NHANES m, Phase 2, are preferred to those based on model predicted blood-lead concentrations
using environmental-lead exposures in the HUD National Survey, hi this section, risk estimates
are calculated under both approaches. The model-predicted estimates of risk are of interest
because:
The §403 standards for lead in paint, dust, and soil will directly impact levels of lead
hi dust, soil, and paint. An assessment of options for standards requires quantifying
the reduction hi adverse health effect and blood-lead concentration endpoints
associated with reduced environmental-lead exposures expected to result from the
proposed rule. The approach used in Chapter 6 is based on the two-step process for
computing health effect and blood-lead concentration endpoints presented hi
Chapter 4.
Presenting an approach for characterizing current risks due to childhood lead
exposure, which uses both steps of the two-step process, illustrates how each
component of the risk assessment feeds into the risk characterization. Sensitivity
analyses on the integrated risk analysis produced by this approach reflect how the
uncertainty associated with each component of the risk assessment affects the overall
risk characterization.
Subsection 5.1.1 presents the baseline risk characterization, obtained using NHANES ffl,
Phase 2 data. The alternative characterization of current risks, based on the IEUBK or empirical
model applied to the HUD National Survey data, is presented hi subsection 5.1.2. Figure 5-2
summarizes both approaches. Approximate 95% confidence intervals, which reflect sampling
variability from NHANES m and uncertainty hi the blood-lead concentration to IQ score
decrement relationship, are presented for most of the health effect and blood-lead concentration
endpoints hi the baseline risk characterization. Because of the many steps to the alternative risk
characterization (e.g., blood-lead concentration had to predicted for each HUD National Survey
unit), confidence intervals are not presented for the alternative risk characterization.
5-5
-------
Integrated Risk Characterization Process
Baseline Risk
Characterization
Environmental-Lead Exposure
Dust Soil Paint
Prediction of childhood blood-
lead concentrations from
environmental-lead exposure
(internal dose)
NHANES Ill-weighted geometric
mean and geometric standard
deviation of 1-2 year old
blood-lead concentrations
National blood-lead
concentration distribution for
1-2 year old children
Assumed linear relationship
between blood-lead
concentration and
IQ score decrements,
and relationship between
blood-lead concentration
and probability of IQ<70
Prediction of selected
health effect and blood-lead
concentration endpoints from
blood-lead concentration
geometric mean and
geometric standard deviation
Alternative Risk
Characterization
HUD National Survey Data
(with tap weight adjustment)
IEUBK or empirical model
Weighted geometric mean
and geometric standard
deviation of predicted
blood-lead concentration
Assumed linear relationship
between blood-lead
concentration and
IQ score decrements,
and relationship between
blood-lead concentration
and probability of IQ<70
Risks due to childhood
lead exposure
Incidence of selected
health effect and blood-lead
concentration endpoints
Risks due to childhood
lead exposure
I I Steps in the process
Data sets or results of applying steps
Figure 5-2. Summary of Risk Characterization Process.
5-6
-------
5.1.1 Baseline Risk Characterization
The baseline risk characterization is based on the distribution of blood-lead
concentrations for children aged 1-2 years determined in Phase 2 of NHANES ffl. This survey
provides a national distribution of blood-lead concentration for the period 1991-1994.
In NHANES HI, each surveyed participant was assigned a sampling weight, where the
sum of the sampling weights equaled the total U.S. population. For children aged 1-2 years, the
sum of these weights (6,789,000) was less than the total number predicted in 1997 by this risk
analysis (7,961,000; see Table 3-35 of Chapter 3). Thus, numbers of children experiencing
particular health effect and blood-lead concentration endpoints are estimated by multiplying the
estimated percentage of children experiencing the given endpoint by the predicted number of
children hi the age group (i.e., 7,961,000 1-2 year olds).
The bar graph hi the top part of Figure 5-3 summarizes blood-lead concentrations for
children aged 1-2 years as measured hi NHANES ffl, Phase 2. Each bar indicates the percentage
of children having a given blood-lead concentration, where the concentration for each surveyed
child is weighted by his/her sampling weight. Using information portrayed in this bar graph, the
estimated geometric mean blood-lead concentration for children aged 1-2 years was 3.14 ng/dL.
and the estimated geometric standard deviation was 2.1.
The distribution of blood-lead concentrations used hi the baseline risk characterization
was assumed to be lognormal with the same geometric mean and geometric standard deviation as
that observed hi NHANES HI, Phase 2. This distribution is plotted as the smooth curve in the
top graph of Figure 5-3; the lognormal distribution closely resembles the bar graph for the
NHANES HI data. The close agreement between the lognormal distribution and the bar graph
was used to validate the lognormal assumption for blood-lead concentrations.
The bottom graph hi Figure 5-3 presents the cumulative distribution function (cdf) of the
blood-lead concentration distribution, which is used to determine the estimated percentage of
children having a blood-lead concentration below a specified value. Two curves are presented:
the cdf of the observed NHANES HI data (jagged curve) and the cdf of the lognormal distribution
used in the baseline risk characterization (smooth curve). The close agreement between the two
curves indicates that the lognormal assumption is appropriate. Using the procedures documented
in Appendix El, the smooth curve is used to determine the baseline estimates of the two blood-
lead concentration endpoints: the percentage of children with a blood-lead concentration at or
above 10 ug/dL, and the percentage at or above 20 ug/dL. These estimated percentages, along
with the associated numbers of children and approximate 95% confidence intervals for these
percentages, are provided hi Table 5-1. These confidence intervals were computed based on the
lognormality assumption, using standard error estimates calculated to account for the complex
survey design employed hi NHANES HI. This methodology is presented in Appendix C2,
Section 2.0. These estimates and confidence intervals are slightly different than those presented
in CDC, 1997, and Section 3.4 because they are based on the lognormality assumption.
5-7
-------
0 1 2 3 4 6 a 7 I 910t112«141S1617««202122a3242S282rata>a03ia2
Blood-Lead Concentration (pg/dL)
wo
94%
3
b
l
o
s
o i 2 3 4 a e r a aiotii2i3Ui8iai7iaia202i2223242saa2raa293oai32
Blood-Lead Concentration (pg/dL)
Figure 5-3. Baseline Distribution of Blood-Lead Concentrations Based on NHANES III,
Phase 2 (0.07 Percent of Children Had Blood-Lead Concentration Greater
than 32 /ig/dL).
5-8
-------
Table 5-1. Estimated Baseline Number and Percentage of Children Aged 1-2 Years
Having Specific Health Effect and Blood-Lead Concentration Endpoints.
Health Effect and Blood-Lead
Concentration Endpoints
PbB i 20 /yg/dL1
PbB* lOj/g/dL1
IQ score less than 702
IQ score decrement * 13
IQ score decrement 2 23
IQ score decrement i 33
Average IQ decrement
Estimated Baseline
Number of Children
46,800
458,000
9,130
3,060,000
863,000
294,000
Baseline Percentage
of Children
Estimate
0.588%
5.75%
0.115%
38.5%
10.8%
3.70%
Estimate
1 .06 points
95% Confidence
Interval
(0.256, 1.35)
(3.73, 8.84)
4
(24.6, 60.2)
(4.86, 24.1)
(1.28, 10.7)
95% Confidence
Interval
(0.703, 1.41)
1 Determined from Figure 5-3.
2 Determined from methods in Section 4.4.2
3 Determined from Figure 5-4.
4 A confidence interval could not be estimated for this endpoint.
Given the lognormal distribution of blood-lead concentration portrayed in Figure 5-3 and
using the procedures documented in Section 4.4 and Appendix El, a baseline distribution of
children's IQ point decrements due to residential lead exposure was calculated. This distribution
is plotted in the top graph of Figure 5-4; the cdf of the distribution appears in the bottom graph.
From this distribution, baseline estimates of the numbers and percentages of children
experiencing a specified IQ point decrement resulting from lead exposure were calculated and are
presented in Table 5-1, along with approximate 95% confidence intervals for these percentages.
The confidence intervals reflect sampling variability from NHANES in and uncertainty in the
relationship between blood-lead concentration and IQ decrement, and are calculated using
methods described in Appendix C2, Section 3.0. Table 5-1 also contains baseline estimates for
numbers and percentages of children having IQ score less than 70, as well as average IQ
decrement per child, that are expected to result from lead exposure.
Since Phase 2 of NHANES III was conducted, there have been many ongoing local, state,
and federal initiatives to reduce childhood blood-lead concentrations. These actions contribute to
differences between the actual distribution of childhood blood-lead concentrations in 1997 and
that estimated above. Specifically, if the government strategies already in place are effective, the
estimated distribution may assign higher probabilities to elevated blood-lead concentrations than
does the actual distribution for 1997.
The sensitivity analysis (Section 5.4) addresses the impact on the baseline risk
characterization of various assumptions on adjusting the NHANES data to reflect 1997
conditions. Such adjustments were not made in the baseline risk characterization presented in
5-9
-------
100 -f
80
60
40
20
100H
80
60
I
40
20
345
Number of IQ Points Lost
345
Number of IQ Points Lost
Figure 5-4. Baseline Distribution of IQ Decrements Due to Elevated Blood-Lead
Concentration Based on NHANES III, Phase 2 (0.07 Percent of Children Had
in Excess of 8 IQ Points Lost).
5-10
-------
this section, primarily due to the lack of available information to support making one type of
adjustment over another, and due to questions on whether a single adjustment can be made at the
national level. Nevertheless, data from Phase 2 of NHANES III are considered the best available
data for estimating the distribution of blood-lead concentrations in 1997.
5.1.2 Alternative Risk Characterization
Using the environmental-lead data from the HUD National Survey (Section 3.3.1) as
input, the IEUBK model (Section 4.1) and empirical model (Section 4.2) were each used to
predict an alternative national distribution of blood-lead concentrations for 1-2 year old children
in 1997 (pre-§403). Using methods presented in Appendix El, these blood-lead concentration
distributions were then used to estimate health effect and blood-lead concentration endpoints.
This section details this alternative risk characterization and presents comparisons among the
three approaches to risk characterization: NHANES HI (Section 5.1.1); HUD National
Survey/IEUBK model; and HUD National Survey/empirical model. The empirical model was
developed with the property that the geometric mean blood-lead concentration predicted by this
model (based on the HUD National Survey data) would match that obtained from NHANES ffl.
As detailed in Sections 4.1 and 4.2, respectively, the alternative risk characterization
applies the IEUBK and empirical models to environmental-lead levels associated with each
housing unit in the HUD National Survey. Specifically, a geometric mean blood-lead
concentration for children aged 1-2 years is predicted for each unit in the National Survey, and a
geometric standard deviation (GSD) of 1.6 is assumed for each unit. Data from various studies
indicate that the inherent variability in blood-lead concentration among children exposed to
similar environmental-lead levels corresponds to a GSD of 1.6, the default GSD recommended in
the IEUBK guidance manual (USEPA, 1994a). The predicted geometric mean associated with
the national distribution of children's blood-lead concentrations is calculated by taking a
weighted geometric mean of the IEUBK or empirical model-predicted blood-lead concentrations
associated with each HUD National Survey unit, with each unit weighted by its sampling weight
adjusted for 1997 population totals. The predicted national GSD is calculated as a function of
variability in blood-lead concentration at a given exposure level and variability associated with
different exposure levels. (Details are in Appendix E2.) The alternative national distribution of
children's blood-lead concentration is assumed to have a lognormal distribution with this
geometric mean and geometric standard deviation.
The predicted 1997 distributions of 1-2 year old children's blood-lead concentrations
obtained by applying the IEUBK and the empirical models to environmental-lead data from the
HUD National Survey are graphically displayed in Figure 5-5 along with the distribution reported
in Section 5.1.1 based on NHANES in data. The bottom graph in Figure 5-5 presents the
national distribution of children's blood-lead concentration as a cdf. The empirical model-
predicted blood-lead distribution has a geometric mean value that is identical to NHANES HI (by
design), but has much shorter tails than the distribution based on NHANES ffl, due to lower
variability associated with the distribution based on the empirical model.
5-11
-------
50-
25
20-
10
Btood-Uad Otitrtbutlen
NHANES IN Prt-lntwvcnlten
HUO/IEUWC Pi*-lnt«mntton
HUD/EMPIRICAL Pr*-lnl*rv«n»on
0 I
t 10 11 12 19 14 19 16 17 It It 20 21 22 29 24 29 2* 27 21 21 JO 91 92
Blood-Lead Concentration
100-
80-
g ^
a.
1
| 40H
o
20
ttood-Uod OlMrtbullon
NHANES III *r«-lnten«nttofi
HUO/IEIMK Pr*-lnt«rvwi«ton
1O 11 IX 11 14 19 1C 17 II It 20 21 22 21 24 23 2t 17 U 2* JO 11 12
Blood-Lead Concentration (/zg/dL)
Figure 5-5. Distribution of Blood-Lead Concentrations (pg/dL) for Children
Aged 1-2 Years Based on NHANES III, and IEUBK and Empirical
Model Predictions.
5-12
-------
The geometric mean blood-lead concentration based on NHANES III data is 3.14 |ag/dL
(GSD=2.1), while the predicted geometric means of 3.9 ug/dL (GSD=2.3) and 3.14 ug/dL
(GSD=1.7) are obtained from the IEUBK model and the empirical model, respectively. The
lEUBK-predicted geometric mean and GSD are close to those reported in Phase 1 of NHANES
III for 1-2 year olds (4.1 ^ig/dL, with GSD = 2.1). The time periods of data collection in the
HUD National Survey (1989-1990) and NHANES IE, Phase 1 (1988-1991) overlap more closely
than do the HUD National Survey and NHANES ffl, Phase 2 (1991-1994).
The IEUBK and empirical model-predicted national blood-lead concentration
distributions were calculated to estimate the incidence of health effect and blood-lead
concentration endpoints using the methods presented hi Appendix El. Table 5-2 presents the
estimated health effect and blood-lead concentration endpoints. The estimated percentages of
children 1-2 years old having blood-lead concentration at or above 10 ng/dL are 5.75% as
estimated by the baseline risk characterization, 12.4% as predicted by the IEUBK model, and
1.5% as predicted bv the empirical model. The empirical model predicts a lower estimate of the
percentage of children affected by health effect and blood-lead concentration endpoints than the
baseline risk characterization due to the smaller estimated geometric standard deviation (1.7 vs.
2.1). Conversely, the IEUBK model predicts larger estimates of the percentage of children
affected compared to the baseline estimate because it predicts a higher geometric mean and
geometric standard deviation than does the baseline risk characterization. These findings reflect
differences in the use and development of these models.
Table 5-2. IEUBK and Empirical Model Predicted Health Effect and Blood-Lead
Concentration Endpoints for Children Aged 1-2 Years.
Blood-Lead Concentration
and Health Effect Endpoints
PbB ^ 20 /yg/dL
PbBi 10//g/dL
IQ score less than 70 resulting from lead exposure
IQ decrement * 1 resulting from lead exposure
IQ decrement z 2 resulting from lead exposure
IQ decrement 2 3 resulting from lead exposure
Average IQ decrement resulting from lead exposure
Percentage of Children Aged 1-2 Years
Empirical Model
Prediction
0.0278%
1.54%
0.0997%
34.5%
4.53%
0.718%
0.932 points
IEUBK Model
Prediction
2.24%
12.4%
0.146%
50.4%
19.9%
8.95%
1 .40 points
Baseline Risk
Characterization
0.588%
5.75%
0.115%
38.5%
10.8%
3.70%
1 .06 points
While comparisons can be made between health effect and blood-lead concentration
endpoints estimated from the NHANES data and HUD National Survey data using the IEUBK
and empirical models, such comparisons should not be used to evaluate the quality of either
model. Too many other factors (e.g., survey times are not the same, HUD National Survey may
not represent actual lead levels in U.S. housing, no post-1979 housing was actually sampled in
the HUD National Survey) vary between the two methods of estimating health effect and blood-
lead concentration endpoints. This makes it impossible to affirm that observed differences are
due to model deficiencies.
5-13
-------
The IEUBK model was developed and calibrated for children at certain large area lead
sites identified in the Superfund program. In general, these are children in housing for which
lead in soil contributes significantly to lead in house dust, and this lead is accessible and
bioavailable. It is not clear that the default parameters, which were selected as appropriate for
Superfund sites (USEPA, 1994a; Section 4.1), are applicable to all U.S. children. Conditions at
general residential sites have not been investigated to the same level of detail.
The empirical model was derived from a single urban lead exposure study (the Rochester
Lead-in-Dust Study) with parameters adjusted for measurement error and "calibrated" for use
with HUD National Survey data (see Section 4.2). The relationship between blood-lead
concentration and lead in soil and dust is not as strong based on the Rochester data as that
predicted by the IEUBK model. This contributes to the smaller geometric standard deviation
observed when applying the empirical model to the HUD National Survey data, compared to
applying the IEUBK model. Empirical studies rarely find relationships as song as that predicted
by the IEUBK model. This is due in part to underlying assumptions of the IEUBK model and in
part to factors in the real world which attenuate the observed relationships between blood-lead
concentrations and environmental-lead levels. Factors that contribute to the attenuation include
variations in children's biokinetics of lead, behavioral activities, hand washing practices and
nonresidential lead exposures, house cleaning practices, and extent of groundcover.
Comparing estimates of health effect and blood-lead concentration endpoints between the
IEUBK and empirical models is also restricted because the two models require different inputs
from the HUD National Survey data. Specifically, the IEUBK model requires only the floor
dust-lead concentration, while the empirical model uses the floor and window sill dust-lead
loadings. The dust-lead concentrations measured in the HUD National Survey required a tap
weight adjustment before use hi the risk analysis (USEPA, 1996c).
5.2 ESTIMATION OF RISKS DUE TO BACKGROUND EXPOSURE
This risk analysis addresses adverse health effects due to exposure to lead-based paint
hazards (i.e., the condition, location, and amount of lead-based paint that causes exposure to lead
in paint, lead-contaminated dust and lead-contaminated soil that would result in adverse health
risks). Therefore, it is desirable to quantify the potential reduction in health effect and blood-lead
concentration endpoints that would occur by completely eliminating lead exposures associated
with lead-based paint hazards. However, it is difficult to quantify the contribution of sources
other than lead-based paint (for example, smelters or battery plants) to lead in dust and soil.
Therefore, to quantify the portion of adverse health effects which could be prevented by reducing
childhood lead exposure from all sources, this section estimates health effect and blood-lead
concentration endpoints at background levels of lead in dust and soil. Because complete
elimination of lead-based paint hazards will not necessarily reduce levels of lead in dust and soil
to background levels, this analysis puts an upper bound on the potential reductions in adverse
health effects resulting from promulgation of the §403 rule.
The approach is based on using the IEUBK model to predict the average blood-lead
concentration for children aged 1-2 years that are exposed to background levels of lead. The
5-14
-------
specified levels are based upon the best available scientific evidence, but actual levels
representative of an environment free of lead hazards are unknown. No contribution due to pica
for paint is considered hi this analysis as it is assumed there is no lead-based paint. Figure 5-6
presents an overview of the process for estimating background risks.
To address the uncertainty regarding soil-lead and dust-lead concentrations corresponding
to lead levels representative of an environment free of lead hazards, four possible soil-lead
concentrations and two different dust-lead concentrations were considered. The various soil-lead
concentrations used corresponded to no soil lead (0 |xg/g), background soil-lead concentration
(16 |ig/g rounded up to 20 |ig/g) in non-urban environments without lead-based paint, and two
larger soil-lead concentrations for comparative purposes (50 and 100 fig/g). The background
soil-lead concentration was obtained from Shacklette et al. (1984).
The two dust-lead concentrations considered corresponded to no dust lead (0 |ig/g) and
the default dust-lead concentration assumed by the IEUBK Multiple Source Analysis (USEPA
1994a). The default value for dust-lead concentration in the Multiple Source Analysis assumes
that both soil and air contribute to dust-lead concentrations. The default dust-lead concentration
is computed as
dust-lead concentration = Q.7Q*(soil-Iead concentration) +
100 \ig/g/\ig/m3*(air-lead concentration)
where the air-lead concentration is set equal to 0.10 (ig/m3 (the approximate average 1990 urban
air-lead concentration (USEPA, 1991)). All other default values defined by the IEUBK model
were used hi these analyses (Section 4.1). The LEUBK Guidance Manual (USEPA, 1994a)
contains a complete discussion of the Multiple Source Analysis and other default values used for
this analysis.
The results of this analysis are presented in Table 5-3. The second and third columns of
the table show the predicted geometric mean blood-lead concentration for both definitions of
background dust-lead concentration. The geometric mean blood-lead concentration estimated at
a soil- and dust-lead concentration of 0 (ig/g (i.e., blood-lead concentration resulting from other
sources, such as food and water) suggests an upper bound on adverse health effect reductions due
to promulgation of the §403 rule (i.e., complete elimination of lead-based paint hazards,
including lead-based paint in soil and dust, and removing soil as a source of exposure, is not
anticipated to reduce the national geometric mean blood-lead concentration below this point).
Estimating the geometric mean blood-lead concentration at soil-lead concentrations of 50 and
100 (ig/g and dust-lead concentrations set according to the IEUBK Multiple Source Analysis
provides additional understanding of the impact of sources of lead other than lead-based paint on
potential benefits of the §403 rule.
5-15
-------
Background Exposure Risk
Characterization Process
Environmental-Lead Exposure
Dust Soil Paint
Background Environmental-Lead Levels, e.g.
Soil: 20 ng/g
Dust: 14 ng/g
Paint: No LBP
Prediction of childhood blood-
lead concentration from
environmental-lead exposure
IEUBK model (no pica)
For the above example, the geometric
mean blood-lead concentration is
computed to be 1.86 ng/dL.
National blood-lead
concentration distribution for
1-2 year old children
Employs computed geometric mean blood-lead
concentration for 1-2 year olds, geometric
standard deviation of 1.6, and assumption of
lognormal distribution.
Prediction of adverse health
effects and blood-lead concentration
endpoints from blood-lead
concentration geometric mean and
geometric standard deviation
Assumed linear relationship between blood-lead
concentration and IQ score decrements
Incidence of adverse
health effects and blood-lead
concentration endpoints
Risks due to background childhood lead exposure
I I Steps in the process
j [ Data sets or results of applying steps
Figure 5-6. Overview of Process for Estimating Risks Due to Background Lead Exposure.
5-16
-------
Table 5-3. IEUBK Model-Predicted Blood-Lead Concentrations for Children Aged 1-2
Years Under Different Soil-Lead Concentrations and at Dust-Lead
Concentrations Equal to Either 0 //g/g or IEUBK Multiple Source Analysis
Default Values.
Background
Soil-Lead
Concentration
Ufg/g)
0
20
50
100
IEUBK Geometric Mean Predicted Blood-Lead
Concentration (//g/dL)
Dust Lead Concentration =
Multiple Source Analysis Default1
1.61
1.86
(41 %)2
2.23
2.83
Dust Lead
Concentration = 0/yg/g
1.52
1.66
(47%)*
1.86
2.20
Weighted Geometric
Mean from
NHANES III Study
U/g/dU
3.14
1 Dust-lead concentration = 0.7 * (soil-lead concentration) + 10/yg/g.
2 Total potential for reduction in geometric mean blood-lead concentration for 1-2 year olds is 41%.
3 Total potential for reduction in geometric mean blood-lead concentration for 1-2 year olds is 47%.
The weighted geometric mean from Phase 2 of NHANES III, 3.14 ug/dL, is shown in the
last column of Table 5-3. The difference between 3.14 ug/dL and that estimated at
background lead exposures may be used to estimate the total potential for reduction in
adverse health effects due to childhood lead exposure.
Using a soil-lead concentration of 20 ug/g as an example, when dust-lead concentration is
assumed to be 0 ug/g. the geometric mean blood-lead concentration could be reduced up to 47%
(47% = 100 x (3.14 -1.66 / 3.14Y). Under conditions reflected bv the Multiple Source Analysis
results, it could be reduced up to 41%.
Results in Table 5-3 for the soil-lead concentration of 20 ug/g were used to estimate
health effect and blood-lead concentration endpoints due to background lead exposure. Health
effect and blood-lead concentration endpoints were computed using methodologies detailed in
Chapter 4 and Appendix El and consistent with Section 5.1, assuming that
1. Blood-lead concentrations are lognormally distributed,
2. The geometric standard deviation of blood-lead concentrations for children with
comparable lead exposures is 1.6,
3. The geometric mean blood-lead concentration is either 1.86 ug/dL (Multiple Source
Analysis dust-lead concentration of 24 ug/g) or 1.66 ug/dL (dust-lead concentration
of 0 ug/g).
Table 5-4 presents health effect and blood-lead concentration endpoints calculated based on these
assumptions and using data from Phase 2 of NHANES III.
5-17
-------
Table 5-4. Percentage of Children Aged 1-2 Years Having Specific Health Effect and
Blood-Lead Concentration Endpoints, Based on lEUBK-Predicted Blood-Lead
Concentrations Under Background Soil- and Dust-Lead Concentrations,
Compared with Estimates from the Baseline Risk Characterization.
Health Effect and
Blood-Lead Concentration
Endpoints
PbB 2 20 /yg/dL
PbB 2 10//g/dL
IQ score less than 70
IQ score decrement 2 1
IQ score decrement * 2
IQ score decrement 2 3
Average IQ decrement
Percentage of Children Aged 1-2 Years
Soil-Lead Cone. = 20 //g/g
Dust Lead Cone. = 24 //g/g1
2.17x 10'6%
0.0173%
0.0877%
5.82%
0.116%
0.00466%
0.534 points
Soil-Lead Cone. = 20 //g/g
Dust Lead Cone. = 0 //g/g
5.93 x 10-0%
0.00665%
0.0868%
3.50%
0.0506%
0.00166%
0.476 points
Baseline Risk
Characterization
(Section 5.1.1)
0.588%
5.75%
0.115%
38.5%
10.8%
3.70%
1 .06 points
1 Multiple Source Analysis default.
5.3 INDIVIDUAL RISKS
Risk estimates in Section 5.1 represent risk to the nation's population of 1-2 year old
children (12-35 months) exposed to residential environmental-lead levels existing within the
1997 housing stock. These population-based risks characterize hazards posed by childhood lead
exposure to the nation as a whole. The risk to children exposed to specific levels of residential
environmental lead are better characterized by estimating the impact of the specific
environmental exposure levels if those levels are known or can be estimated. The exposure
levels of a child may be estimated, for example, based on the results of a risk assessment
conducted in that child's residence.
The individual risks presented in this chapter refer to the risks estimated for (a population
of) children exposed to specified levels of environmental lead. These risks are different from the
population-based risks of Section 5.1 in that those risks are estimated based on the distribution of
environmental-lead levels occurring in the U.S. housing stock. Individual risks do not represent
the risks of a specific child, as such risks are best determined by medical professionals who have
access to specific information on the characteristics of the child as well as his/her environment.
The estimates of individual risks take the form of, for example, the percentage of children
exposed to an average floor dust-lead loading of 200 fig/ft2, window sill dust-lead loading of 700
Hg/ft2, and soil-lead concentration of 2,000 ppm who experience a blood-lead concentration
greater than or equal to 10 jig/dL.
5-18
-------
There are two tools available for estimating individual risks based on residential
environmental-lead levels: the IEUBK model (Section 4.1) and the Rochester multimedia model
(Section 4.2). The Rochester multimedia model differs from the empirical model in three ways:
it has not been adjusted for HUD National Survey specific measurement error, it was not
calibrated to produce a geometric mean blood-lead concentration of 3.14 ng/dL (the geometric
mean in Phase 2 of NHANES ffl) when applied to the HUD National Survey data, and it predicts
blood-lead concentrations based on wipe dust-lead loadings rather than Blue Nozzle vacuum
dust-lead loadings. These differences make the Rochester multimedia model more appropriate
than the empirical model for use in the individual risk analysis.
The IEUBK model is used to estimate the risks of childhood blood-lead concentration
greater than or equal to 10 ug/dL for various soil-lead concentrations. The Rochester multimedia
model takes as input dripline soil-lead concentration, and therefore, was not considered
appropriate for evaluating the risks at specified yard average soil-lead concentrations. The
Rochester multimedia model is employed to estimate the risks of childhood blood-lead
concentration greater than or equal to 10 ug/dL for various combinations of floor and window
sill dust-lead loading. The IEUBK model employs dust-lead concentration as its input measure
of dust lead, and therefore, was not considered appropriate for evaluating the risks at specified
dust-lead loadings.
Figure 5-7 presents estimates of the percentage of children's blood-lead concentrations
that will exceed (or equal) 10 ug/dL as a function of soil-lead concentration, as predicted by the
IEUBK model, assuming dust-lead concentrations of 100,200, and 500 ug/g. These dust-lead
concentrations were chosen as representative of typical dust-lead concentrations in homes built in
the period 1960-1979,1940-1959, and pre-1940, respectively (see Table 3-7).
Figure 5-7 suggests that the percentage of children aged 1-2 years having blood-lead
concentrations greater than or equal to 10 ug/dL is larger than 5% for any soil-lead concentration,
if the children are exposed to a dust-lead concentration of 500 ug/g (top-most curve). For
children exposed to a dust-lead concentration between 100 and 200 ug/g, a soil-lead
concentration between 250 and 350 ug/g would provide an approximately 5% chance of having a
blood-lead concentration greater than or equal to 10 ug/dL. Table 5-5 presents the soil-lead
concentrations predicted to maintain the percentage of children having blood-lead concentrations
above (or equal to) 10 ug/dL at 1%, 5%, and 10% for the dust-lead concentrations considered in
Figure 5-7.
As shown in Table 5-5, for all dust-lead concentrations considered, the soil-lead
concentrations estimated to maintain risks at 1% and 5% are less than 500 ppm; even a soil-lead
concentration of 0 ppm would not achieve these levels of protection if children are exposed to
dust-lead concentrations of 500 ng/g. Based on Table 5-5, if dust-lead concentrations are equal
to 100 ng/g, a soil-lead concentration of approximately 400 ug/g maintains the percentage of
blood-lead concentrations greater than or equal to 10 ug/dL at 5%.
5-19
-------
/M
2
0.
i
o
S
5
100-
90
80
70
60
so}
40
30
20
10
0 250 500 750 1000 1250 1500 1750 2000
Soil-Lead Concentration (pg/g)
Dust-Leod Concentration (\igJg): 100 200 500
Figure 5-7. Percentage of Children's Blood-Lead Concentrations, as Predicted by the
IEUBK Model, That Will Exceed or Equal 10//g/dL as a Function of Soil-Lead
Concentration for Three Dust-Lead Concentrations.
Table 5-5. Soil-Lead Concentrations at Which the Percentage of Children Aged 1-2 Years
Having a Blood-Lead Concentration Above or Equal to 10 //g/dL is Estimated
by the IEUBK Model at 1, 5, or 10%, Under Three Assumed Dust-Lead
Concentrations.
Floor Dust-Lead
Concentration 0/g/g)
100
200
500
Soil-Lead Concentration (//g/g)
1%
155
35
Not achievable
5%
365
245
Not achievable
10%
515
395
25
Figure 5-8 graphs the estimated percentage of children having blood-lead concentrations
above or equal to 10 ug/dL as a function of floor dust-lead loadings. The percentages are plotted
for soil-lead concentrations of 100 and 400 ug/g and window sill dust-lead loadings of 200 and
500 ng/ft2, with separate plots for the two specified window sill dust-lead loadings. Soil-lead
concentrations of 100 and 400 ug/g were chosen based on the previous analysis of soil-lead
concentrations. Window sill dust-lead loadings were chosen because they are representative of
homes built in the period pre-1940 (see Table 3-8) and interim guidance levels (USEPA, 1995h).
5-20
-------
Figure 5-8 indicates that at the specified soil-lead concentrations and window sill dust-
lead loadings, floor dust-lead loading must be less than 10 ug/ft2 to control the percentage of
children predicted to have blood-lead concentrations greater than or equal to 10 ug/dL at 5%.
When soil-lead concentration is 100 ug/g and window sill dust-lead loadings is 200 ug/ft2, a risk
of 10% can be achieved by floor dust-lead loadings around 90 ug/ft2.
Tables 5-6 and 5-7 present the floor and window sill dust-lead loadings, respectively, that
are predicted to maintain the percentage of children having blood-lead concentrations above or
equal to 10 ug/dL at 1%, 5%, and 10% for specified levels of soil-lead concentration and dust-
lead loading on window sills or floors, respectively. Approximate 95% upper confidence
bounds, which account for the variability of parameter estimates from the Rochester multimedia
model, are also provided in Tables 5-6 and 5-7. Confidence bounds are included in Tables 5-6
and 5-7 but not Table 5-5 because of fundamental differences in the IEUBK and Rochester
multimedia models. Specifically, the Rochester model was estimated empirically, and thus,
uncertainty in the model can be quantified. (This captures variability in the relationship between
environmental lead and blood-lead concentration in the Rochester study and assumes that this
relationship is representative of the entire nation). The IEUBK model, however, is mechanistic,
and no measure of uncertainty in prediction of mean blood-lead concentration associated with
input environmental-lead levels is available. The methodology used to compute the upper
confidence bounds is provided in Appendix C2, Section 5.0.
Floor dust-lead loadings that control the percentage (risk) of children having blood-lead
concentrations above (or equal to) 10 ug/dL at 1% are less than 1 ug/ft2 for all soil-lead
concentrations and window sill dust-lead loadings considered. Floor dust-lead loadings that
maintain risk at the 10% level are less than 100 ug/ft2. Similar results are observed for window
sill dust-lead loadings in Table 5-7. Window sill and floor dust-lead loadings that control risk at
5% range from 4.2 to 74 ug/ft2 and from 0.2 to 6.7 ug/ft2, respectively.
Table 5-6. Floor Dust-Lead Loadings at Which the Percentage of Children Aged 1-2 Years
Having a Blood-Lead Concentration Above or Equal to 10 //g/dL is Estimated
by the Rochester Multimedia Model at 1, 5, or 10% Under Two Assumed Soil-
Lead Concentrations and Two Assumed Window Sill Dust-Lead Loadings.
Soil-Lead
Concentration
(P0/9>
100
400
Window Sill
Dust-Lead
Loading
fo/g/ff)
200
500
200
500
Floor Dust-Lead Loading (//g/ft*)
1%
Estimate
0.05
0.02
Not
achievable
95% Upper
Confidence
Bound
1.32
0.42
0.13
0.04
5%
Estimate
6.7
2.0
0.61
0.18
95% Upper
Confidence
Bound
170
54
17
5.3
10%
Estimate
89
27
8.1
2.4
95% Upper
Confidence
Bound
2200
710
220
71
5-21
-------
T?
30 f
25
15
10
30
25
? 20
All
2
£ 15
10
5
100 200 300 400
Floor Dust-Lead Loading (ug/fP)
Soil-Lead Concentration (ug/g): 100 400
a) Window Sill Dust-Lead Loading = 200 ug/ft2
500
100
200
300
400
500
Floor Dust-Lead Loading (ug/fP)
Soil-Lead Concentration (ug/g): 100 400
b) Window Sill Dust-Lead Loading = 500 ug/ft2
Figure 5-8. Percentage of Children's Blood-Lead Concentrations, As Predicted By the
Rochester Multimedia Model, That Will Exceed or Equal 10 jig/dL as a
Function of Floor Dust-Lead Loading for Two Soil-Lead Concentrations and
Two Window Sill Dust-Lead Loadings.
5-22
-------
Table 5-7. Window Sill Dust-Lead Loadings at Which the Percentage of Children Aged
1-2 Years Having a Blood-Lead Concentration Above or Equal to 10 /yg/dL
is Estimated by the Rochester Multimedia Model at 1, 5, or 10% Under Two
Assumed Soil-Lead Concentrations and Two Assumed Floor Dust-Lead
Loadings.
Soil-Lead
Concentration
100
400
Floor Dust-
Lead Loading
(ftgm*}
25
100
25
100
Window Sill Dust-Lead Loading (pg/ft2)
1%
Estimate
1.9
0.65
0.30
0.11
95% Upper
Confidence
Bound
25
9.5
4.5
1.7
5%
Estimate
74
26
12
4.2
95% Upper
Confidence
Bound
990
380
180
67
10%
Estimate
520
180
85
30
95% Upper
Confidence
Bound
7000
2700
1300
480
5.4 RISK CHARACTERIZATION SENSITIVITY AND UNCERTAINTY ANALYSIS
Results presented in this risk characterization are dependent on a number of factors,
including the assumptions and data analysis approaches taken, the outcomes of supporting data
analyses, and the availability of sufficient data. Sensitivity analyses address the extent to which
variations in key assumptions and approaches affect the outcome. These variations are
associated with overall uncertainty. Thus, sensitivity analysis within the risk characterization
evaluates how sensitive the results and conclusions of the characterization are to the uncertainty
present in the analysis.
Table 5-8 summarizes the components of the risk characterization that were addressed by
the sensitivity analysis and the alternative approach considered for each component. This section
presents the findings of the sensitivity analysis. Justification for each alternative approach
considered in the analysis is provided, and reasons for not including certain factors of the risk
assessment in the sensitivity analysis are discussed.
Note that the sensitivity analysis does not consider other options for obtaining estimated
numbers of housing units in the 1997 housing stock or numbers of children residing in the
housing stock (presented in Chapter 3). In preliminary analyses, it was observed that regardless
of the method used to obtain an estimated number of units (or children) within the four categories
determined by housing age, the percentage of the total housing stock (or the total population of
children) within each group remained relatively constant. Therefore, it was not deemed
necessary to consider alternative methods for determining numbers of housing units or children.
5-23
-------
Table 5-8. Components of the Risk Characterization Addressed By the Sensitivity
Analysis.
Risk Characterization Component
Approach Taken in
the
Risk Analysis
Altemative(s) Considered in the
Sensitivity Analysis
Determine
:hildren
appropriate age group of
Age group = 1-2
years (i.e., 12 to 35
months) for both
population and
individual risk
characterization
Baseline (Population) Risk Characterization:
Age group = 1-5 years
(i.e., 12 to 71 months)
Individual Risk Characterization (IEUBK
Model):
Age group = 0-5 years
(i.e., 0 to 71 months)
Section 5.4.1
Determine an average IQ point loss
associated with every 1 /yg/dL increase in
blood-lead concentration in children
Average IQ point loss
= 0.257.
Alt. #1: Average IQ point loss = 0.185
Alt. 92: Average IQ point loss = 0.323
Section 5.4.2
Determine the baseline (pro-§403) blood-
ead concentration in children aged 1-2
years in U.S. housing
Use data collected in
Phase 2 of NHANES
III, with no
adjustments to the
reported blood-lead
concentrations
Consider across-the-board declines of 10%,
20%, and 30% in blood-lead concentration
from those concentrations reported in Phase
2 of NHANES III
Section 5.4.3
Assume a lognormal
distribution (See
Section 5.1.1)
Use the observed distribution data reported
in the NHANES III without any assumption of
lognormality.
Section 5.4.4
When using modeling techniques to predict
pre-intervention values of the health effect
and blood-lead concentration endpoints,
determine an appropriate value for the
geometric standard deviation (GSD) of the
blood-lead concentrations associated with
a given environmental-lead exposure
scenario
Assume a GSD of 1.6
Alt. *1; Assume a GSD of 1.4
Alt. #2: Assume a GSD of 1.9
Alt. »3: Assume a GSD of 2.1
Section 5.4.6
When using the IEUBK model to predict
pre-intervention values of the health effect
and blood-lead concentration endpoints,
determine an appropriate value for daily
dietary lead intake for a child aged 1-2
years (an input parameter to the IEUBK
model)
Assume daily dietary
lead intake is 5.78 tig
(the model's default
value for children
aged 1-2 years)
Alt. #1: Daily dietary lead intake = 1.29//g
Alt. #2: Daily dietary lead intake = 3.53 //g
Section 5.4.7
When using modeling techniques to predict
pre-intervention values of the health effect
and blood-lead concentration endpoints,
adjust model-based results to reflect the
effects of paint pica tendencies on blood-
lead concentration
Make assumptions on
the prevalence of
paint pica and the
effects of paint pica
on blood-lead
concentration that are
documented in
Section 4.1.3 and
Appendix D1
Alt. #1: Make no adjustment for paint pica
effects
Alt. #2: Assume a lower prevalence of paint
pica and lower effects of paint pica on
blood-lead concentration than that used in
the risk analysis
Alt. #3: Assume a higher prevalence of
paint pica and higher effects of paint pica on
blood-lead concentration than that used in
the risk analysis
Section 5.4.8
5-24
-------
Table 5-8. Components of the Risk Characterization That Were Addressed By the
Sensitivity Analysis (Continued)
Risk Characterization Component
Approach Taken in
the
Risk Analysis
Alternative(s) Considered in the
Sensitivity Analysis
When using the IEUBK model to predict a
pre-intervention geometric mean blood-lead
concentration, adjust dust-lead
concentrations in the HUD National Survey
to reflect the dust sample's total weight,
not just the tap weight
For a given dust
sample, use a
regression model
(USEPA, 1996c) to
predict the ratio of
the sample's total
weight to its tap
weight, then divide
the sample's reported
lead concentration by
this ratio.
Alt.
-------
Table 5-9a. Sensitivity Analysis for Estimated Baseline Number and Percentage of
Children Having Specific Health Effects and Blood-Lead Concentration
Endpoints for Two Age Groups of Children and Under Three Assumptions on
Average Decline in IQ Score per Unit Increase in Blood-Lead Concentration.
Health Effect and Blood-Lead
Concentration Endpoints1
PbB 2 20 fjg/dL
PbB 2 10/yg/dL
IQ score less than 70
IQ score decrement 2 1
IQ score decrement 2 2
IQ score decrement 2 3
0.185 decline/
1//g/dL increase
0.257 decline/
1/yg/dL increase
0.323 decline/
1/yg/dL increase
0.185 decline/
1/yg/dL increase
0.257 decline/
1/yg/dL increase
0.323 decline/
1/;g/dL increase
0.185 decline/
1/yg/dL increase
0.257 decline/
1/yg/dL increase
0.323 decline/
1//g/dL increase
Children Aged 1-2 Years
Having the Given Health
Effect
Number
(millions)
0.0468
0.458
0.00913
1.83
3.06
4.04
0.368
0.863
1.41
0.101
0.294
0.557
Percentage {%)
(95% CD
0.588
(0.256, 1 .35)
5.75
(3.73, 8.84)
0.115
23.0
(10.2, 51.7)
38.5
(24.6, 60.2)
50.7
(38.0, 67.8)
4.62
(1.23, 17.4)
10.8
(4.86, 24.1)
17.8
(10.2, 31.0)
1.27
(0.238, 6.81)
3.70
(1.28, 10.7)
7.00
(3.27, 15.0)
Children Aged 1-5 Years Having
the Given Health Effect
Number
(millions)
0.0673
0.785
0.0216
3.61
6.45
8.85
6.20
1.57
2.71
0.154
0.485
0.973
Percentage (%)
(95% CD
0.330
(0.153, 0.714)
3.85
(2.55, 5.79)
0.106
17.7
(7.21,43.3)
31.6
(19.1, 52.2)
43.4
(31.3, 60.2)
3.04
(0.737, 12.5)
7.69
(3.25, 18.2)
13.3
(7.27, 24.2)
0.756
(0.130, 4.40)
2.38
(0.781, 7.27)
4.77
(2.13, 10.7)
1 For IQ score decrement, this column also includes the assumption on average IQ score decline per 1 //g/dL increase in
blood-lead concentration.
Shaded cells correspond to results that were presented in Table 5-1.
Table 5-9b. Sensitivity Analysis for Estimated Baseline Average IQ Decrement for Two
Age Groups of Children and Under Three Assumptions on Average Decline in
IQ Score per Unit Increase in Blood-Lead Concentration.
Assumption on Average IQ Score Decline per 1 .0
//g/dL increase in Btood-Lead Concentration
0.185
0.257
0.323
Average IQ Decrement (95% Confidence Interval)
Children Aged 1-2 Years
0.761 (0.418, 1.10)
1.06(0.703, 1.41)
1.33 (0.961, 1.70)
Children Aged 1-5 Years
0.663 (0.368, 0.958)
0.921 (0.619, 1.22)
1.16(0.848, 1.47)
Shaded cell corresponds to results that were presented in Table 5-1.
5-26
-------
Effect on risk analysis: The percentages for the 1-5 year age group are approximately
15%-35% lower than those for the 1-2 year age group. For example, Table 5-9a indicates that the
expected percentage of children aged 1-2 years having blood-lead concentration of at least 10
Hg/dL is approximately 6%, compared to approximately 4% for children aged 1-5 years.
However, these differences are generally not statistically significant at a 5% error rate. Table
5-9b presented estimated average IQ score loss; similar declines are observed here. Observed
declines are likely the result of lower blood-lead concentrations introduced to the distribution by
increasing the representation of older children.
Characterizing Individual-Based Risks
Individual risks were also calculated for the age group 0-5 years. However, there are
restrictions associated with applying a broader age range in characterizing individual risks. In
particular, the Rochester multimedia model (Section 4.2), used to estimate individual risks as a
function of dust-lead loadings on floors and window sills, was developed from data hi the
Rochester Lead-in-Dust Study, which included only children aged 12-31 months. Therefore, the
Rochester multimedia model could not be used to obtain risk estimates for children aged 0-5
years.
The IEUBK model (Section 4.1) was used to estimate individual risks as a function of
soil-lead concentration, at specified values of dust-lead concentration. As the IEUBK model
produces longitudinal blood-lead concentration estimates for children aged 0-7 years under
specific exposure scenarios, it was possible to consider the broader age range of 0-5 years in this
sensitivity analysis.
As described in Section 4.1.4, when using the IEUBK model to characterize risks to
children in a specified age range, a representative age within this range (in months) was selected
to estimate blood-lead concentration. The IEUBK model was used to obtain predicted blood-
lead concentrations for each month hi the range 0-71 months. The predicted blood-lead
concentration at 49 months was approximately equal to the geometric mean of blood-lead
concentrations over the entire age range for a number of lead exposure scenarios. Therefore, in
this sensitivity analysis, IEUBK model predictions at age 49 months were used to characterize
blood-lead concentrations of children aged 0-5 years.
Table 5-10 presents soil-lead concentrations predicted by the IEUBK model to control the
percentage of 0-5 year old children estimated to have blood-lead concentration greater than or
equal to 10 ug/dL at 1%, 5%, and 10% under three dust-lead concentrations: 100,200, and 500
Hg/g. Comparable results first presented hi Table 5-5 for children aged 1-2 years are also
provided in Table 5-10.
Effect on risk analysis: As expected, when considering the broader age range of 0-5
y^ars, the maximum soil-lead concentration necessary to control the percentage of children with
elevated blood-lead concentration at 1 or 5% is somewhat higher than that associated with the
1-2 year olds (under a fixed floor dust-lead concentration). However, the increased soil-lead
concentration is still within the same working range observed for 1-2 year olds.
5-27
-------
Table 5-10. Soil-Lead Concentration at Which the Percentage of Children Having a Blood-
Lead Concentration Above or Equal to 10 /sg/dL is Estimated by the IEUBK
Model at 1, 5, or 10% Under Three Assumed Dust-Lead Concentrations and
Two Age Groups of Children.
Floor Dust-Lead
Concentration
U/g/g)
100
200
500
Soil-Lead Concentration (pg/g)
Children Aged 1-2 Years
1%
155
35
5%
365
245
Not achievable
10%
515
395
25
Children Aged 0-5 Years
1%
230
110
5%
480
360
Not achievable
10%
655
535
165
Shaded cells correspond to results that were presented in Table 5-5.
5.4.2 Alternative Assumptions on Average IQ Score Decline Per Unit Increase in Blood-
Lead Concentration
As discussed in Chapter 4, results of the meta-analysis documented in Schwartz (1994)
indicate that an average IQ point loss of 0.257 is predicted for every 1.0 ug/dL increase in blood-
lead concentration. This relationship was used in the risk characterization to characterize health
effects associated with elevated blood-lead concentration. In the sensitivity analysis, two
alternative average IQ point loss estimates were considered: 0.185 and 0.323. The lower value
of 0.185 was selected based on a meta-analysis that combined the findings of prospective studies
that relate blood-lead concentration for children approximately two years of age to IQ scores at
age 5 to 10 years, as reported in Pocock et al. (1994). The higher value of 0.323 corresponds to
an examination in Schwartz (1994) on the existence of a threshold in the relationship between IQ
score and blood-lead concentration. For four studies where the mean blood-lead concentration
was 15 ug/dL or lower, the estimated average IQ point loss was reported to be 0.323. The
estimates of 0.185 and 0.323 result in a lower and higher estimate, respectively, of the adverse
health effect endpoints. The sensitivity analysis did not consider alternative methods for
estimating the probability of observing IQ scores less than 70.
In Tables 5-9a and 5-9b of the previous subsection, the estimated percentages of children
with IQ score decrements greater than 1, 2, or 3 under assumptions of an IQ score decline of
0.185, 0.257, and 0.323 points for every 1.0 ug/dL increase in blood-lead concentration are
presented, along with approximate 95% confidence intervals associated with these percentages.
These percentages reflect baseline (pre-§403) conditions and are presented for children aged 1-2
years and 1-5 years.
Effect on risk analysis: The magnitude of the decline in IQ score associated with a 1
ug/dL increase in blood-lead concentration has a considerable impact on the likelihood that a
child will experience an IQ score decrement greater than 1,2, or 3, with its effect increasing as
the decrement of interest increases. As seen in Table 5-9a, the estimated percentage of children
5-28
-------
aged 1-2 years with an IQ score decrement of at least one, as calculated using the low estimate of
IQ point loss (0.185), more than doubles when the high estimate (0.323) is used instead (from
23% to 51%). A more than fivefold increase in the percentage of children aged 1-2 years with an
IQ score decrement of at least three is observed between the low and high estimates of IQ point
loss (from 1.3% to 7.0%). In Table 5-9b, average IQ decrement increases from 0.76 to 1.33 for
children aged 1-2 years, with a similar increase observed for children aged 1 -5 years. However,
as the 95% confidence intervals overlap across the three assumptions on IQ score decline, the
observed differences in the estimates across these assumptions are not statistically significant at a
5% error rate.
5.4.3 Considering Potential Declines in Blood-Lead Concentration from NHANES III
Phase 2 Measures
Blood-lead concentrations in the U.S. population have consistently declined in recent
years. Therefore, it is likely that blood-lead concentrations have continued to decline since 1994,
the last year of Phase 2 of NHANES HI for which blood-lead concentration data were utilized in
this risk analysis. This portion of the sensitivity analysis investigated how baseline estimates of
the blood-lead concentration and health effect endpoints for children aged 1-2 years (Section
5.1.1) may change under different assumptions on the decline in geometric mean blood-lead
concentration since 1994.
Between the two phases of NHANES ffl, the measured geometric mean blood-lead
concentration for children aged 1-2 years declined 22.5%, from 4.05 |ig/dL hi Phase 1 (1988-
1991) to 3.14 ug/dL hi Phase 2 (1991-1994). Therefore, this sensitivity analysis calculated
baseline estimates of the blood-lead concentration and health effect endpoints for children aged
1-2 years, where each blood-lead concentration measurement in Phase 2 of NHANES m was
reduced by the same amount: 10%, 20%, or 30%. Table 5-11 presents these estimates, along
with the estimates reported in the risk analysis (where no reduction was assumed).
Effect on risk analysis: According to Table 5-11, a 10% across-the-board decline hi
blood-lead concentration reduced the estimated number of children whose blood-lead
concentration was at or above 20 ug/dL from 46,800 to 30,900, a decline of 34%, while the
estimated number at or above 10 ng/dL was reduced by 26%. Similar percentage declines were
also observed for numbers of children with IQ score decrements of 2 or 3 as a result of lead
exposure. A 30% across-the-board decline reduced the estimated number of children at or above
10 ug/dL by 66%, and the estimated number of children at or above 20 |ig/dL by 77%.
The results hi Table 5-11 are based on the assumption that the blood-lead concentrations
for each child in the population are reduced by the same percentage. In reality, some subgroups
achieve less of a decline, with others achieving a greater decline. However, considering different
percentage declines for different subgroups would be very difficult, and the resulting estimates of
the health effect and blood-lead concentration endpoints would likely differ only slightly from
that observed in Table 5-11.
5-29
-------
Table 5-11. Sensitivity Analysis for the Estimated Baseline Number and Percentage of
Children Aged 1-2 Years Having Specific Health Effect and Blood-Lead
Concentration Endpoints, Assuming Various Percentage Declines in Blood-
Lead Concentration Since NHANES III Phase 2.
Health Effect and Blood-Lead
Concentration Endpoints
PbB :> 20 /;g/dL
PbB * 10//g/dL
IQ score less than 70
IQ score decrement 2 1
IQ score decrement * 2
IQ score decrement * 3
Average IQ score decrement
Geometric Mean (jjg/dL)
Numbers (%) of Children Aged 1-2 Years
Risk analysis
estimate (Table 5-1)
46,800 (0.588%)
458,000 (5.75%)
9,130
(0.115%)
3,060,000 (38.5%)
863,000 (10.8%)
294,000 (3.70%)
1.06
3.14
Percentage Decline in Blood-Lead Concentration
Since NHANES III Phase 2
10%
30,900
(0.388%)
340,000
(4.27%)
8,610
(0.108%)
2,640,000
(33.2%)
669,000
(8.40%)
213,000
(2.68%)
0.951
2.82
20%
18,900
(0.238%)
239,000
(3.00%)
8,160
(0.102%)
2,190,000
(27.6%)
493,000
(6.19%)
146,000
(1.83%)
0.845
2.51
30%
10,600
(0.133%)
1 56,000
(1.96%)
7,760
(0.098%)
1 ,740,000
(21.8%)
340,000
(4.27%)
91,900
(1.15%)
0.740
2.20
5.4.4 Alternative Approach to Characterizing a Baseline Blood-Lead Distribution from
NHANES III Data
As discussed in Section 5.1.1, the baseline distribution of blood-lead concentration in
children aged 1-2 years was assumed to be lognormal, with geometric mean (3.14 ug/dL) and
geometric standard deviation (2.1 ug/dL) calculated from NHANES III Phase 2 data. Health
effect and blood-lead concentration endpoints were then calculated from this distribution.
Although a plot of the fitted lognormal distribution is in close agreement with the NHANES III
data (see Figure 5-3), some deviation from the lognormal distribution was evident, especially in
the upper tail. Therefore, an alternative approach to characterizing the baseline distribution
using the NHANES III data considered an empirical distribution (i.e., the distribution of the
observed data, represented by the bar chart in Figure 5-3), with no lognormal assumption. This
alternative approach was applied in the sensitivity analysis to evaluate the effect of the lognormal
assumption on the results of the baseline risk characterization.
The NHANES III, Phase 2 database included blood-lead concentrations for 987 children
aged 1-2 years at the time of their survey interview. At the time of the physical examinations,
each child in the survey was assigned a sampling weight corresponding to the number of children
in the country being represented by the child. In this risk assessment, these weights were scaled
to represent the 1997 population (see Section 5.1.1). This combination of blood-lead
concentration and sample weight for each surveyed child provided an empirical distribution of
5-30
-------
blood-lead concentration for children aged 1-2 years. The percentage of children with a blood-
lead concentration greater than or equal to 10 ^ig/dL was estimated from this distribution by
summing the sample weights for children with blood-lead concentrations greater than or equal to
10 ng/dL, then dividing by the total of all sampling weights. This same approach as used to
summarize the NHANES HI data in Section 3.4.1. The percentage of children experiencing
certain IQ decrements as a result of lead exposure was obtained by transforming the IQ
decrement to a specific blood-lead concentration threshold, assuming a 0.257 IQ score decrement
for every 1.0 (ig/dL of blood-lead concentration, then estimating the percentage of children at or
above this threshold. The percentage of children with an IQ score less than 70 due to lead
exposure was calculated by determining the probability of this occurrence for each surveyed
child based on his/her blood-lead concentration (using the methods in Section 4.4.2), then
multiplying by the child's sample weight, summing the results across children, and dividing by
the total of all sampling weights.
Effect on risk analysis: The values of the health effect and blood-lead concentration
endpoints under the empirical distribution of blood lead concentration, as well as under the
baseline distribution used hi the risk assessment, are provided in Table 5-12. The estimated
health effect and blood-lead concentration endpoints are very similar for the two distributions.
The empirical distribution method estimates a higher percentage of children with blood-lead
concentration greater than or equal to 10 ug/dL (5.88% versus 5.75% for the risk assessment
baseline distribution), but a smaller percentage exceeding 20 ug/dL (0.43% versus 0.59%).
Estimated average 10 score loss is virtually identical for the two distributions, with a larger
difference observed for an IQ decrement of 3 or higher. The empirical distribution yields higher
estimates for the endpoints associated with the most severe effects compared to the lognormal
distribution due to its greater emphasis of data in the upper tail of the distribution.
5.4.5 Uncertainty in Adjusting Dust-Lead Concentrations to Reflect the Sample's Total
Weight
Section 5.1.2 presented a model-based approach to characterizing a pre-intervention
distribution of blood-lead concentrations. Under this method, environmental-lead levels
measured in the HUD National Survey are used as input to a model to predict the geometric
mean blood-lead concentration hi a particular population exposed to such levels. Model inputs
for the IEUBK model (Section 4.1) include dust-lead concentrations from the HUD National
Survey (Section 3.3.1). As the HUD National Survey reported lead concentration within a dust
sample as the amount of lead in the entire sample divided by the weight of only that portion of
the sample that could be tapped out of the vacuum collection cassette (i.e., the "tap weight"), it
was necessary to adjust these reported concentrations to reflect the sample's total weight prior to
using the dust-lead concentrations in the risk assessment. Otherwise, the reported lead
concentration might overestimate the true concentration in the sample. The methods developed
in this risk assessment for making the adjustments to dust-lead concentration are documented in
USEPA, 1996c. In this component of the sensitivity analysis, geometric mean blood-lead
concentrations (as determined by fitting the IEUBK model to the dust-lead concentrations) and
the resulting health effects were recalculated using three alternative sets of adjusted dust-lead
concentrations that were either lower or higher than those used hi the risk assessment.
5-31
-------
Table 5-12. Sensitivity Analysis for Estimated Baseline Health Effect and Blood-Lead
Concentration Endpoints, for Children Aged 1 -2 Years, as Calculated Under
Two Approaches to Calculating the Baseline Distribution of Blood-Lead
Concentration Using NHANES III Data.
Health Effect end Blood-Lead
Concentration Endpoints
PbB * 20 //g/dL
PbB* 10|/g/dL
IQ score less than 70
IQ score decrement z 1
IQ score decrement * 2
IQ score decrement 2 3
Average IQ decrement
Children Aged 1-2 Years
Risk Analysis Estimates
(Table 5-1 )(%)
0.588
5.75
0.115
38.5
10.8
3.70
1 .06 points
Estimates Based on Alternative
(Empirical) Approach (%)
0.431
5.88
0.115
38.3
10.5
4.47
1 .06 points
The adjustment of dust-lead concentrations in National Survey units to correct for tap
weight bias involved dividing each dust sample's reported concentration by the predicted ratio of
the total weight of the sample to its tap weight (USEPA, 1996c). A sample's predicted ratio was
a function of its tap weight (mg) and was obtained from the following formula:
Predicted Ratio = 30.71 - 19.54*X + 3.36*X2 - 0.061*W
where X equals the minimum of 2.903 and the sample's tap weight, and W equals the maximum
of (tap weight - 2.903) and zero. If the tap weight was greater than or equal to 23.44 mg, the
predicted ratio was taken to be one. Dust-lead concentrations were omitted from the risk analysis
for samples having a tap weight of less than 0.7 mg.
The above formula was determined from a nonlinear regression analysis performed as
part of a laboratory study (USEPA, 1996c). Thus, the coefficients 30.71, -19.54, 3.36, and -
0.061 in the formula were estimates based on the data collected in the laboratory study.
Therefore, assuming normality, lower and upper one-sided 90% confidence bounds on the
predicted ratio were determined in the following manner:
Lower 90% confidence bound on ratio = Predicted Ratio - (1.3*SEPredictedRatio)
Upper 90% confidence bound on ratio = Predicted Ratio + (1 .3*SEPredictedRatio)
where SEp^^,^ is the standard error of the predicted ratio, which was a function of the
standard errors of the estimated coefficients and the co variances between pairs of coefficients
that were obtained in the regression analysis. The lower and upper confidence bounds on the
5-32
-------
predicted ratio presented are for the mean predicted ratio, and, as such, do not account for house
to house variability. Accounting for house to house variability would result in a smaller, lower
bound and a larger upper bound. The above estimates of the upper and lower bounds are useful,
however, to portray the impact of the assumptions for tap weight adjustments on the risk
estimates.
The upper and lower confidence bounds were used to obtain low and high estimates,
respectively, for the adjusted dust-lead concentrations in the HUD National Survey:
Low estimate of adjusted dust-lead cone. = unadjusted dust-lead cone.
upper 90% conf. bound on ratio
High estimate of adjusted dust-lead cone. = unadjusted dust-lead cone.
lower 90% conf. bound on ratio
Therefore, the low and high estimates represent two alternative sets of adjusted dust-lead
concentrations for the HUD National Survey dust samples. In addition, as the predicted ratios
are never less than one, the adjusted concentrations can never be larger than their respective
unadjusted values. Thus, the unadjusted concentrations serve as upper limits on the adjusted
concentrations. This sensitivity analysis considered three alternative sets of dust-lead
concentrations when predicting geometric mean blood-lead concentrations under the IEUBK
model: the set of low adjusted estimates (Alternative set #1), the set of high adjusted estimates
(Alternative set #2), and the set of unadjusted concentrations (Alternative set #3)'.
For each alternative set of dust-lead concentrations, the mass-weighted arithmetic mean
dust-lead concentrations for both floors and window sills were calculated for each National
Survey unit (i.e., each sample concentration was weighted by the sample's tap weight). The
IEUBK model was then applied to these data for each unit to obtain a predicted baseline
geometric mean blood-lead concentration for the nation.
From this geometric mean, health effect and blood-lead concentration endpoints were
calculated. The findings were compared across the three alternative sets of concentrations to
evaluate the impact of the adjustment method. Table 5-13 presents estimated pre-intervention
health effect and blood-lead concentration endpoints (obtained using the IEUBK model on the
HUD National Survey data) as reported hi the risk characterization (and presented in Table 5-2)
and under the three alternative sets of adjusted dust-lead concentrations.
1 Dust-lead concentrations with tap weights less than 0.7 mg were omitted from all alternative sets.
5-33
-------
Table 5-13. Sensitivity Analysis for Estimated Health Effect and Blood-Lead
Concentration Endpoints for Children Aged 1 -2 Years, as Calculated Using
the IEUBK Model and Under Four Approaches to Adjusting Dust-Lead
Concentrations for Low Tap Weight.
Blood-Lead Concentration and
Health Effect Endpoints
Geometric mean blood-lead
concentration U/g/dL)
(geometric standard deviation in parentheses)
PbB* 20/yg/dL(%)
PbB 2 10/yg/dL{%)
IQ score less than 70 (%)
IQ decrement 2 1 (%)
IQ decrement 2 2 (%)
IQ decrement 2 3 (%)
Average IQ decrement (# points)
Risk Analysis
Estimates
(Table 5-2)
3.92 (2.26)
2.24
12.4
0.146
50.4
19.9
8.95
1.40
Estimates
Under
Alternative
#1'
3.85 (2.25)
2.13
1"2.0
0.144
49.5
19.3
8.62
1.38
Estimates
Under
Alternative
#22
3.96 (2.22)
2.10
12.2
0.144
50.9
19.8
8.73
1.40
Estimates Under
Alternative #3
(no adjustment)
4.45 (2.32)
3.74
16.9
0.169
56.3
25.4
12.7
1.63
1 Low estimates for the tap-weight adjusted dust-lead concentration (unadjusted concentration divided by the upper 90%
confidence bound for the ratio of total dust weight to tap weight).
2 High estimates for the tap-weight adjusted dust-lead concentration (unadjusted concentration divided by the lower 90%
confidence bound for the ratio of total dust weight to tap weight).
Effect on risk analysis: Table 5-13 indicates that using the upper or lower 90%
confidence bound on the ratio of the tap weight to the entire sample's weight to adjust HUD
National Survey dust-lead concentration values (i.e., Alternatives #1 and #2) has little impact on
risk estimates. Therefore, risk estimates are probably not sensitive to the exact method chosen to
do the tap weight adjustment. However, health effect and blood-lead concentration endpoints for
the no-adjustment alternative (#3) suggest estimates of health risks due to childhood lead
exposure are more severe if no tap weight adjustment is done. Based on the dust sample
collection and analysis protocol used in the HUD National Survey, it is clear that some
adjustment needs to be made to the dust-lead concentration values to remove potential bias
associated with these values.
5.4.6 Alternative Estimates for the Geometric Standard Deviation of Blood-Lead
Concentrations
In using the IEUBK and empirical models to characterize the distribution of blood-lead
concentrations in children aged 1-2 years exposed to a specified set of environmental-lead levels,
this risk analysis assumes that the geometric standard deviation (GSD) of this distribution is 1.6.
This component of the sensitivity analysis considers alternatives to this GSD value.
5-34
-------
Three alternative GSD values were considered: 1.4, 1.9, and 2.1. The value 1.4 was the
GSD for the distribution of blood-lead concentrations for the national population of children
aged 1-2 years in 1979 and 1980, as estimated by NHANES H. The value 2.1 was the GSD as
estimated in both phases of NHANES ffl. The value 1.9 falls approximately halfway between
the NHANES HI estimate of 2.1 and the value of 1.6 used in the risk analysis.
In the risk analysis, a GSD of 1.6 is assumed to reflect only "inter-individual" variability,
or variability among children exposed to the same environmental-lead levels. Sources of "inter-
individual" variability include behavioral differences, the extent of accessibility to and contact
with available lead, measurement variability, biological diversity, and differences in food
consumption. In contrast, GSD values calculated from blood-lead concentrations measured in
the NHANES (e.g., the lowest and highest values specified in the previous paragraph) reflect
variability due to these sources, as well as "inter-neighborhood" variability, or variability
associated with exposure to different environmental-lead levels. Therefore, the range of
alternative values is highly likely to contain the "true" value for the GSD which reflects only
variability among children exposed to the same environmental-lead levels.
As the results in Table 5-2 and Tables 5-4 through 5-7 are dependent on the value of the
"inter-individual" GSD, the numbers in these tables were recalculated under the three alternative
GSD values. Section 6.4.6 presents results of applying these alternative values to estimate
endpoint values following promulgation of the proposed §403 rules, under a specific set of
example options for the §403 standards.
Effect on risk analysis: Under the three alternative GSD assumptions, Table 5-14
presents model-predicted, pre-§403 health effect and blood-lead concentration endpoints, while
Table 5-15 presents these estimates under dust-lead and soil-lead concentrations that represent
background conditions. According to these tables, the choice of GSD value impacts results
under the empirical model more than the IEUBK model. For example, the pre-§403 probability
of a child having a blood-lead concentration at or above 10 ug/dL, as estimated by the empirical
model, ranges from 0.316% to 7.02% as the GSD varies from 1.4 to 2.1, respectively (Table
5-14). In contrast, this probability ranges from 10.4% to 17.3% under the IEUBK model. When
considering the IQ parameters, those representing the greatest health effects were most sensitive
to the GSD value.
The effect of alternative GSD values on estimates of individual risks (Section 5.3) is
investigated in Tables 5-16 through 5-18. These tables have the same formats as Tables 5-5
through 5-7 and provide estimates of the maximum environmental-lead level in a specific
medium that would keep the percentage of children with blood-lead concentrations at or above
10 ug/dL to within a specified threshold, assuming fixed lead levels in other media. These
maximum lead levels decline as the GSD increases; levels are generally very low and/or
unachievable for GSDs of 1.9 and 2.1. The estimates differ substantially between GSD values of
1.6 and 1.4, indicating that the assumed value of the GSD has a large impact on the individual
risk estimates.
5-35
-------
Table 5-14. Sensitivity Analysis on the Model-Predicted, Pre-§403 Health Effect and
Blood-Lead Concentration Endpoints for Children Aged 1 -2 Years, Under
Three Alternative Values (1.4, 1.9, 2.1) for the Geometric Standard
Deviation (GSD) of the Blood-Lead Concentration Distribution and Under the
Value Used in the Risk Analysis (1.6).
Health Effect and Blood-
Lead Concentration
Endpoints
PbB * 20//g/dL(%)
PbB i 10/yg/dL(%)
IQ < 70 (%)
IQ decrement * 1 (%)
IQ decrement a 2 (%}
IQ decrement z 3 (%)
Average IQ decrement
(# points)
Pre-5403 Predictions: IEUBK Model
GSD = 1.4
1.41
10.4
0.135
50.4
17.8
7.09
1.33
GSD = 1.6
2.24
12.4
0.146
50.4
19.9
8.95
1.40
GSD = 1.9
3.86
15.5
0.167
50.3
22.9
11.8
1.54
GSD = 2.1
5.06
17.3
0.184
50.3
24.5
13.6
1.65
Pre-1403 Predictions: Empirical Model
GSD = 1.4
0.0006
0.316
0.0950
30.6
1.62
0.0984
0.883
GSD =1.6
0.0278
1.54
0.0997
34.5
4.53
0.718
0.932
GSD =1.9
0.372
4.70
0.111
37.8
9.48
2.89
1.03
GSD =2.1
0.922
7.02
0.120
39.2
12.4
4.73
1.10
Note: Results in shaded cells were presented in Table 5-2.
Table 5-15. Sensitivity Analysis on IEUBK Model-Predicted Health Effect and Blood-Lead
Concentration Endpoints for Children Aged 1 -2 Years, Under Three
Alternative Values (1.4, 1.9, 2.1) for the Geometric Standard Deviation
(GSD), Assuming a Background Soil-Lead Concentration of 20 //g/g and One
of Two Estimates of Background Dust-Lead Concentration.
Health Effect and Blood-
Lead Concentration
Endpoints
PbB a 20/yg/dL(%)
PbB 2 10j/g/dL(%)
IQ < 70 (%)
IQ decrement 2 1 (%)
IQ decrement * 2(%)
IQ decrement * 3 {%)
Average IQ decrement
(# points)
Dust-Lead Concentration = Multiple Source
Analysis Default
GSD = 1.4
8.39x1 0'13
2.88x1 0"B
0.0871
1.41
0.001 1
2.40x10*
0.506
GSD =1.6
2.17x10-*
0.0173
0.0877
5.82
0.116
0.0047
0.534
GSD = 1.9
0.0108
0.439
0.0902
12.5
1.29
0.211
0.587
GSD = 2.1
0.0684
1.17
0.0932
16.0
2.69
0.665
0.629
Dust-Lead Concentration = 0 //g/g
GSD =1.4
6.96x1 0'14
4.72x1 0"*
0.0863
0.568
0.0002
3.38x1 0'7
0.451
GSD = 1.6
5.93x1 0"8
0.0067
0.0868
3.50
0.0506
0.0017
0.476
GSD = 1.9
0.0053
0.257
0.0886
9.22
0.804
0.119
0.524
GSD = 2.1
0.0397
0.775
0.0909
12.5
1.87
0.428
0.562
Note: Results in shaded cells were presented in Table 5-4. When background dust-lead concentration is taken to be the
multiple source analysis default, the endpoints are estimated assuming a geometric mean of 1.86 (jg/dL, and when
background dust-lead concentration is taken to be zero, the endpoints are estimated assuming a geometric mean of
1.66 fjg/dl (see Table 5-3).
5-36
-------
Table 5-16. Sensitivity Analysis on the Soil-Lead Concentrations at Which the Percentage
of Children Aged 1 -2 Years Having Blood-Lead Concentration at Least 10
//g/dL is Estimated by the IEUBK Model at 1, 5, or 10%, Under Three
Assumed Dust-Lead Concentrations and for Alternative Assumptions on the
Geometric Standard Deviation (GSD) for the Blood-Lead Distribution.
Fkior Dust-
Lead Cone.
U/g/g)
100
200
500
Soil-Lead Concentration (//g/g)
%chlldren with blood-lead cone.
2 10 //g/dL = 1%
QSD =
1.4
360
235
na
QSD=
1.6
155
35
na
QSD =
1.9
na
na
na
QSD =
2.1
na
na
na
%children with blood-lead cone.
2 10 //g/dL = 5%
QSD =
1.4
565
440
75
QSD =
1.6
365
245
na
GSD-
1.9
180
55
na
QSD =
2.1
95
na
na
%chlldren with blood-lead cone.
* 10 //g/dL = 10%
QSD =
1.4
700
575
210
QSD =
1.6
515
395
25
QSD=.
1.9
325
205
na
QSD =
2.1
240
115
na
Note: Results in shaded cells were presented in Table 5-5. 'na" = not achievable.
Table 5-17. Sensitivity Analysis on the Floor Dust-Lead Loadings at Which the
Percentage of Children Aged 1 -2 Years Having Blood-Lead Concentration at
Least 10 //g/dL is Estimated by the Rochester Multimedia Model at 1, 5, or
10%, for Two Assumed Soil-Lead Concentrations and Two assumed
Window Sill Dust-Lead Loadings, and for Alternative Assumptions on the
Geometric Standard Deviation (GSD) for the Blood-Lead Distribution.
Soil-
Lead
Cone.
(pg/g)
100
400
Window
Sill
Duct-
Lead
Loading
0/g/ft2)
200
500
200
500
Floor Dust-Lead Loading (pg/ft*)
%ch8dren with blood-lead cone.
* 10 //g/dL - 1%
QSD=
1.4
5.8
1.7
0.53
0.16
QSD=
1.6
0.05
0.02
na
na
QSD =
1.9
na
na
na
na
QSD =
2.1
na
na
na
na
%ch8dren with blood-lead cone.
i 10 //g/dL = 5%
QSD =
1.4
190
56
17
5.1
QSD =
1.6
6.7
2.0
0.61
0.18
GSD =
1.9
0.09
0.03
0.01
na
GSD=
2.1
0.01
na
na
na
%children with blood-lead cone.
4 10 //g/dL = 10%
QSD =
1.4
1200
360
110
32
QSD =
1.6
89
27
8.1
2.4
GSD =
1.9
3.2
0.95
0.29
0.09
QSD =
2.1
0.45
0.14
0.04
0.01
Note: Results in shaded cells were presented in Table 5-6. 'na' = not achievable.
5-37
-------
Table 5-18. Sensitivity Analysis on the Window Sill Dust-Lead Loadings at Which the
Percentage of Children Aged 1 -2 Years Having Blood-Lead Concentration at
Least 10 i/g/dL is Estimated by the Rochester Multimedia Model at 1, 5, or
10%, for Two Assumed Soil-Lead Concentrations and Two Assumed Floor
Dust-Lead Loadings, and for Alternative Assumptions on the Geometric
Standard Deviation (GSD) for the Blood-Lead Distribution.
Soil-
Lead
Cone.
U/g/a)
100
400
Floor
Dust-
Lead
Loading
fc/g/ft*)
25
100
25
100
Window Sffi Dust-Lead Loading bug/ft1)
%ch9dren with blood-lead cone.
* 10ffg/dL = 1%
QSD =
1.4
66
23
11
3.7
GSD«
1.6
1.9
0.65
0.30
0.11
QSD =
1.9
0.02
0.01
na
na
GSD =
2.1
na
na
na
na
%chiklren with blood-toad cone.
i 10//g/dL = 5%
GSD~
1.4
920
320
150
52
GSD-
1.6
74
26
12
4.2
GSD =
1.9
2.9
1.0
0.46
0.16
GSD-
2.1
0.43
0.15
0.07
0.02
%chlldren with blood-lead cone.
* 10//g/dL= 10%
GSD-
1,4
3700
1300
610
210
GSD-
1.6
520
180
85
30
GSD-
1.9
42
15
6.8
2.4
GSD-
2.1
9.5
3.3
1.6
0.54
Note: Results in shaded cells were presented in Table 5-7. "na" = not achievable.
5.4.7 Alternative Estimates for Daily Dietary Lead Intake Assumed in Fitting the
IEUBK Model
When the IEUBK model was applied in this risk analysis to obtain a geometric mean
blood-lead concentration as a function of soil-lead concentration and floor dust-lead
concentration, default values were used for parameters that represent lead exposures associated
with sources other than soil and dust. In this component of the sensitivity analysis, alternative
values for one of these parameters, daily dietary lead intake, are considered. Insufficient
information was available to consider alternative values for the other parameters.
The LEUBK model's default value for daily dietary lead intake hi children aged 1-2 years
is 5.78 ug (Table 4-1). This value was estimated from data collected in the Food and Drug
Administration's Market Basket Survey from 1986 (fourth quarter) to 1988 (third quarter), as
well as from other sources. This value was used when fitting the IEUBK model to data within
the risk analysis, hi this sensitivity analysis, two alternative values were considered: 1.29 ug and
3.53 ug. The former value was reported for children aged two years in a more recent Market
Basket Survey, and the latter is the average of the former value and the value of 5.78 jig used in
the risk analysis. As lead concentrations in food are assumed to decline over tune, no alternative
value larger than 5.78 ug was considered.
As the results in Tables 5-2 through 5-5 are dependent on the value of the daily diet
intake parameter hi the IEUBK model, these tables were recalculated under the two alternative
values for this parameter.
Effect on risk analysis: Under the two alternative values for the daily diet intake
parameter (as well as the default value of 5.78 ug used in the risk analysis), Table 5-19 presents
EEUBK model-predicted, pre-§403 health effect and blood-lead concentration endpoints, as well
as at dust-lead and soil-lead concentrations representing background conditions. When the daily
5-38
-------
Table 5-19. Sensitivity Analysis on the IEUBK Model-Predicted, Pre-§403 Health Effect
and Blood-Lead Concentration Endpoints for Children Aged 1-2 Years, Under
Two Alternative Values (1.29//g, 3.53/yg) for the Daily Lead Dietary Intake
Parameter and Under the Value Used in the Risk Analysis (5.78 fjg).
Health Effect and Blood-
Lead Concentration
Endpoints
% with PbB 2 20 fjgldL
% with PbB 2 10//g/dL
% with IQ < 70
% with IQ decrement 2 1
% with IQ decrement 2 2
% with IQ decrement 2 3
Average IQ decrement
(# points)
Geometric mean blood-
lead cone. U/g/dL)
Pre-§403
Predictions
1.29
jrg
2.12
9.78
0.136
38.5
15.2
7.24
1.18
2.95
3.53
W
2.13
11.0
0.140
44.5
17.4
7.98
1.29
3.45
5.78
PQ
2.24
1-2.4
0.146
50.4
19.9
8.95
1.40
3.92
Background Predictions at Soil-Lead Concentration = 20 //g/g
Dust-Lead Concentration =
Multiple Source Analysis Default
1.29
PS
1.21x10-*
5.96x106
0.0841
0.220
7.68X10"4
1.07x10'6
0.293
1.02
3.53
m
1.08x10-°
1.87x10'3
0.0858
1.72
0.0166
4.25x10"
0.413
1.44
5.78
pa
2.17x10*
0.0173
0.0877
5.82
0.116
0.0047
0.534
1.86
Dust-Lead Concentration = 0 j/g/f
1.29
Pa
4.48x10-10
4.46x10-*
0.0833
0.0420
7.40x10'6
6.87x1 0'7
0.232
0.81
3.53
pa
1.65x10-'
4.47x10"
0.0850
0.749
0.0047
9.19x10-B
0.356
1.24
5.78
Pa
5.93x10*
0.0067
0.0868
3.50
0.05O6
0.0017
0.476
1.66
Note: Results in shaded cells were presented in Tables 5-2 through 5-4.
dietary lead intake is equal to 3.53 U£, the geometric mean blood-lead concentrations are from 12
to 25 percent lower than those presented in the risk analysis. The decline is even greater, from 25
to 51 percent, when the daily dietary lead intake is equal to 1.29 ug. Only slight differences are
observed in the IEUBK model-predicted pre-§403 endpoints for the three values of daily dietary
lead intake. However, the assumed value of the daily dietary lead intake has a larger impact on
the estimated endpoints for background environmental-lead exposures. Estimated endpoints at
background lead exposures are more sensitive to the assumed value of daily dietary lead intake
because dietary lead comprises a larger portion of the total lead exposure at background
environmental-lead levels.
The effect of alternative values for daily dietary lead intake on IEUBK model-predicted
individual risks (Section 5.3) is investigated in Table 5-20. This table, a companion of Table 5-
5, provides estimates of the maximum soil-lead concentration necessary to keep the percentage of
children with blood-lead concentrations at or above 10 ug/dL to within specified percentages,
assuming a fixed value for floor dust-lead concentration. According to this table, the maximum
soil-lead concentration does not differ appreciably when daily dietary lead intake is varied.
Therefore, results of this sensitivity analysis conclude that when soil-lead and dust-lead
concentrations are above background levels, the value of daily dietary lead intake has a small
effect on the IEUBK model-predicted health effect and blood-lead concentration endpoints.
5-39
-------
Table 5-20. Sensitivity Analysis on the Soil-Lead Concentrations at Which the
Percentage of Children Aged 1 -2 Years Having Blood-Lead Concentration at
Least 10 //g/dL is Estimated by the IEUBK Model at 1, 5, or 10%, for Three
Assumed Dust-Lead Concentrations and for Alternative Assumptions on
Daily Dietary Lead Intake.
Floor Dust-Lead
Concentration
U/g/g)
100
200
500
Soil-Lead Concentration (pg/g)
% children with Mood-lead
concentration * 10 //g/dL = 1%
1.29 pg
280
155
na
3.53 pg
220
95
na
5.78 pg
155
35
na
% children with Wood-lead
concentration i lOpg/dL = 5%
1.29pg
490
365
na
3.53 pg
425
305
na
6.78 pg
365
245
na
% children with Mood-lead
concentration 2 TOpg/dL « 10%
1.29 ps
635
515
150
3.53 pg
575
455
85
5.78 pg
516
395
25
Note: Results in shaded cells were presented in Table 5-5. "na" = not achievable.
5.4.8 Alternative Assumptions on Paint Pica Tendencies in Children and the Effect of
Paint Pica on Blood-Lead Concentration
Section 4.1.3 and Appendix Dl present the method used in this risk analysis for obtaining
a model-predicted geometric mean blood-lead concentration for those children who have
ingested paint chips. The set of assumptions used by this method differs according to which
model is being used to predict the geometric mean blood-lead concentration:
Assumptions under the empirical model:
9% of children aged 1-2 years have paint pica tendencies
a value of 1.5 is used for the model's paint pica parameter when predicting the
geometric mean blood-lead concentration for children having paint pica tendencies
and living in a housing unit with damaged lead-based paint, and a value of zero is
used for all other situations
Assumptions under the IEUBK model;
9% of children aged 1 -2 years have paint pica tendencies
0.03% of children aged 1-2 years living in housing units containing damaged lead-
based paint have recently ingested paint chips.
children aged 1 -2 years who recently ingested paint chips have a blood-lead
concentration of 63 ug/dL.
children aged 1 -2 years who ingested paint chips at some time, but not recently, have
a 3 |ig/dL increase in their geometric mean blood-lead concentration from children
who do not ingest paint chips.
5-40
-------
This component of the sensitivity analysis considers how alternatives to these assumptions
impact model-predicted, pre-§403 health effect and blood-lead concentration endpoints. As the
estimates of individual risks presented in Section 5.3 assume no deteriorated lead-based paint is
present, there is no effect of paint pica on these estimates.
As the sets of assumptions for handling paint pica differ between the empirical and
IEUBK models, each model is addressed separately within the following two subsections.
5.4.8.1 Empirical Model
When addressing the approach to handling paint pica under the empirical model, the
sensitivity analysis considers three alternatives to the assumed percentage of children aged 1-2
years that have paint pica tendencies: 0%, 6%, and 14%. The assumption of 0% is equivalent to
making no adjustment for paint pica, while the assumptions of 6% and 14% correspond to the
lower and upper limits of an approximate 95% confidence interval on the percentage of children
with paint pica tendencies hi the Rochester Lead-in-Dust study.
Effect on risk analysis: Table 5-21 presents the empirical model-predicted, pre-§403
health effect and blood-lead concentration endpoints, under the three alternative assumptions on
the percentage of children with paint pica tendencies, as well as under the 9% assumption used in
the risk analysis. As seen in this table, the differences among the pica percentage assumptions
are relatively minor. The largest observed difference was a 14% difference in the probability of
observing a child with blood-lead concentration of at least 20 |ig/dL between the risk analysis
estimate (0.0278%) and the percentage that is estimated when no paint pica tendencies are
assumed (0.0239%). All other differences from the risk analysis estimates in Table 5-21 were
less than 10%.
Table 5-21. Sensitivity Analysis on the Empirical Model-Predicted, Pre-§403 Health
Effect and Blood-Lead Concentration Endpoints for Children Aged 1-2 Years,
Under Three Alternative Values (0%, 6%, 14%) for the Percentage of
Children with Paint Pica Tendencies, and Under the Value Used in the Risk
Analysis (9%).
Health Effect ami BJood-Laad
Concentration Endpoints
PbB * 20 fjg/dl (%)
PbBi 10^a/dL(%)
IQ < 70 (%)
IQ decrement * 1 (%)
IQ decrement * 2 (%)
IQ decrement * 3 (%)
Average IQ decrement {# points)
0% of children
w/paint pica
0.0239
1.43
0.0993
34.0
4.30
0.657
0.924
6%ofchBdren
w/paint pica
0.0265
1.50
0.0996
34.3
4.45
0.697
0.929
9% of children
w/paint pica
0,0278
1.S4
0,0997
34.6
4.53
0.718
0.932
14% of children
w/paint pica
0.0302
1.60
0.100
34.7
4.66
0.752
0.936
Note: Results in shaded cells were presented in Table 5-2.
5-41
-------
5.4.8.2 IEUBK Model
The approach to handling effects of paint pica on geometric mean blood-lead
concentrations estimated from the IEUBK model is more complex than that for the empirical
model, due to the greater number of assumptions going into the approach. In the sensitivity
analysis, three sets of alternative assumptions were considered:
Alternative set #1:
Alternative set #2:
Alternative set #3:
Assumes 0% of children have paint pica tendencies. (This is equivalent to
making no adjustment for paint pica, so no alternatives need be specified
for the other assumptions.)
Assumes that paint pica tendencies have less of an impact than that
assumed in the risk analysis:
6% of children aged 1 -2 years have paint pica tendencies (the lower
bound of a 95% confidence interval on the percentage as estimated
from the Rochester Lead-in-Dust study)
0.01 % of children aged 1 -2 years living in housing units containing
damaged lead-based paint have recently ingested paint chips.
children aged 1 -2 years who recently ingested paint chips have a
blood-lead concentration of 55 ug/dL (a low estimate based on
information from McElvaine et al., 1992).
children aged 1-2 years who ingested paint chips at some time, but not
recently, have a 15% increase hi their geometric mean blood-lead
concentration from children who do not ingest paint chips (the lower
bound of a 95% confidence interval on the percentage in the Rochester
Lead-in-Dust study).
Assumes that paint pica tendencies have more of an impact than that
assumed in the risk analysis:
14% of children aged 1 -2 years have paint pica tendencies (the upper
bound of a 95% confidence interval on the percentage as estimated
from the Rochester Lead-in-Dust study)
0.10% of children aged 1 -2 years living in housing units containing
damaged lead-based paint have recently ingested paint chips.
children aged 1-2 years who recently ingested paint chips have a
blood-lead concentration of 63 ug/dL.
children aged 1-2 years who ingested paint chips at some time, but not
recently, have a 100% increase in their geometric mean blood-lead
concentration from children who do not ingest paint chips (the upper
bound of a 95% confidence interval on the percentage hi the Rochester
Lead-in-Dust study).
5-42
-------
Effect on risk analysis: Table 5-22 presents the IEUBK model-predicted, pre-§403
health effect and blood-lead concentration endpoints, under the three alternative sets of
assumptions, as well as under the set of assumptions used in the risk analysis. Results differed
only slightly between the approach used in the risk analysis to adjust for paint pica and when no
adjustment is made (alternative set #1). The differences in results were larger between the
approach used in the risk analysis and when a larger impact of paint pica is assumed (alternative
set #3), but remained relatively minor.
Table 5-22. Sensitivity Analysis on the IEUBK Model-Predicted, Pre-§403 Health Effect
and Blood-Lead Concentration Endpoints for Children Aged 1 -2 Years, Under
Three Alternative Sets of Assumptions on Paint Pica Effects, and Under the
Set of Assumptions Used in the Risk Analysis.
Health Effect and Blood-Lead
Concentration Endpoints
PbB :> 20 //g/dL (%)
PbB^ 10//g/dL(%)
IQ < 70 (%)
IQ decrement :> 1 (%)
IQ decrement * 2 (%)
IQ decrement z 3 (%)
Average IQ decrement
(# points)
Pica Assumptions
in the Risk
Analysis
2.24
12.4
0.146
50.4
19.9
8.95
1.40
Pica Alternative
Set#1
(no adjustment)
2.16
12.2
0.144
50.1
19.6
8.75
1.39
Pica Alternative
Set #2
(tow adjustment)
2.18
12.3
0.145
50.2
19.7
8.81
1.39
Pica Alternative
Set #3
(high adjustment)
2.60
13.3
0.151
50.9
20.9
9.73
1.44
Note: Results in shaded cells were presented in Table 5-2.
5.4.9 Conclusions from Sensitivity Analysis
Several analyses were conducted to assess the sensitivity of the characterization of current
risks to the uncertainty in the underlying assumptions and methods utilized in the risk
assessment.
Estimating baseline health effect and blood-lead concentration endpoints for a broader
age group.
Estimating baseline health effect and blood-lead concentration endpoints using three
different assumptions on the decline in IQ score associated with a unit increase in
blood-lead concentration.
Estimating baseline health effect and blood-lead concentration endpoints when blood-
lead concentrations are assumed to have declined since 1994.
5-43
-------
Estimating baseline health effect and blood-lead concentration endpoints using an
empirical approach to characterizing the distribution of blood-lead concentrations
from NHANES m data.
Estimating baseline health effect and blood-lead concentration endpoints using the
IEUBK model, under three different approaches to adjusting HUD National Survey
dust-lead concentrations to reflect the sample's total weight, under alternative
estimates for the geometric standard deviation of the blood-lead concentration
distribution, under alternative estimates for a child's daily dietary lead intake, and
under different assumptions on the prevalence of paint pica tendencies and their effect
on blood-lead concentration.
Estimating baseline health effect and blood-lead concentration endpoints using the
empirical model, under alternative estimates for the geometric standard deviation of
the blood-lead concentration distribution and under different assumptions on the
prevalence of paint pica tendencies.
i
It may be concluded from this sensitivity analysis that the risk characterization is sensitive to
uncertainty in the relationship between declines in IQ score and increases in blood-lead
concentration (Section 5.4.2X The alternative risk characterization (Section 5.1.2^ is also
sensitive to whether or not the HUD National Survey dust-lead concentration measurements are
adjusted to account for tap weights and on assumptions on the geometric standard deviation
associated with blood-lead concentration at a given lead exposure level. Failing to make some
tap weight adjustment leads to significantly increased risk estimates. Risk estimates were also
increased when the geometric standard deviation is increased, or if any decline in blood-lead
concentrations since 1994, when NHANES HI Phase 2 was completed, was ignored.
The assumption of an 0.257 decrease in IQ score for an increase of one |ig/dL in blood-
lead concentration has considerable impact on the estimates of numbers of children with
specified IQ decrements and average decline in IQ due to lead exposures. However, even if the
decline is less severe (0.185 vs. 0.257), approximately 1.45 million children 1-2 years old, and
2.95 million children 1-5 years old suffer IQ decrements greater or equal to than 2 points due to
exposures to lead-based paint hazards, lead-contaminated dust, and lead-contaminated soil.
5.5 RISK CHARACTERIZATION CONCLUSIONS
There is essentially no question in the scientific community as to whether or not lead is a
hazard. Lead is a neurotoxin causing reductions in IQ scores as well as other neurological
problems, including, in extreme cases, illness and death.
One to two year old children are particularly sensitive to lead for two reasons. Their
rapidly developing central nervous systems are more readily damaged than the central nervous
systems of older age groups. Also, the frequent hand-mouth activity exhibited by 1-2 year olds
increases the quantity of environmental lead ingested. Thus, the most sensitive age group is also
the most exposed.
5-44
-------
In this risk assessment, estimates of internal lead dose are a necessary intermediary step
between environmental exposure to and adverse health effects of lead. Blood-lead concentration
is the most commonly used measure of internal lead dose or lead body burden and the measure
for which the most data are available to make the connection between environmental lead
exposure and adverse health effects. In fact, many general adverse health effects resulting from
exposure to lead are quantified in this risk assessment by the incidences of children with blood-
lead concentrations above certain values.
Quantification of the health hazard due to childhood lead exposure is a difficult problem.
There are many diverse adverse health effects that have been associated with lead exposure. In
this risk assessment, risks are quantified by estimating incidences of seven health effect and
blood-lead concentration endpoints among 1-2 year old children hi 1997. The selected health
effect and blood-lead concentration endpoints are
Incidence of blood-lead concentration greater than or equal to 10 ug/dL
Incidence of blood-lead concentration greater than or equal to 20 ug/dL
Incidence of IQ decrement greater than or equal to 1 resulting from lead exposure
Incidence of IQ decrement greater than or equal to 2 resulting from lead exposure
Incidence of IQ decrement greater than or equal to 3 resulting from lead exposure
Average IQ decrement resulting from lead exposure
Incidence of IQ less than 70 resulting from lead exposure.
The blood-lead concentration endpoints, while not health effects, serve as surrogates for a
number of other adverse health effects.
There are several sources of lead hi the environment, including lead-based paint, lead
plumbing, lead solder, lead glazes, industrial emissions, and auto emissions (before lead was
banned as a gasoline additive). These sources have resulted hi exposure to lead in multiple
media (e.g., dust, soil, paint, food, water) to which children are exposed.
The HUD National Survey provides the most comprehensive, nationally representative
data on lead in residential dust, soil, and paint. According to this survey,
8.9% of the nation's housing contain floor dust-lead loadings greater than 100 fig/ft2,
12.8% have soil-lead concentrations greater than 400 ug/g, and
13.6% contain more than 5 ft2 of damaged lead-based paint
Data from Phase 2 of NHANES ffl (conducted from 1991-1994) provide the most
comprehensive and current nationally representative data on children's blood-lead
concentrations. These data indicate that many children continue to have elevated blood-lead
concentrations. Specifically, 5.75% of the nation's 1-2 year old children are estimated to have
blood-lead concentrations greater than or equal to 10 ug/dL. Estimates for certain subgroups
(e.g., low-income and inner-city children) are more alarming. NHANES ffl and HUD National
Survey data are consistent in suggesting that some subgroups are disproportionately exposed to
5-45
-------
lead, i.e., both environmental- and blood-lead measurements are above average for some
subpopulations, including low-income and inner-city children.
Nationally representative data are not available for relating the health effect and blood-
lead concentration endpoints directly to environmental lead levels. Therefore, an alternative risk
characterization was performed to predict blood lead distribution as a function of environmental
exposure measures collected in the HUD National Survey, using both the EEUBK and empirical
models. The alternative risk characterization based on the EEUBK model and the HUD National
Survey data produced higher risk estimates than did the baseline risk characterization, while the
alternative risk characterization based on the empirical model produced lower risk estimates. All
methods of characterizing risks indicate substantial numbers of children suffer adverse health
effects. For example, the baseline risk characterization indicates that 39% of children aged 1-2
years (3.060.000) will experience at least a 1 point IP decrement as a result of lead exposure.
while 3.7% (294.000) will experience a 3 point decrement.
Using NHANES m data, the IEUBK model, and estimated national background soil-lead
concentration, estimates of the maximum possible reduction in childhood blood-lead
concentrations and adverse health effects were produced. These estimates suggest that if the
contributions of lead-based paint and all other anthropogenic sources contributing to lead in dust
and soil could be removed from total lead exposure, the geometric mean 1-2 year old blood-lead
concentration would be reduced by approximately 40-45%. This reduction is considered an
estimate of the maximum reduction that promulgation of the §403 rule could achieve. It is
expected that this maximum cannot be achieved, as levels of lead in all dust and soil will not be
reduced to background levels.
In Section 5.3, risks to children exposed to specific levels of environmental lead were
characterized, with particular attention paid to determining levels of lead in dust and soil which
would independently be protective of children. The EEUBK and Rochester multimedia models
were both used for this characterization. If floor dust-lead concentration is equal to 100 ng/g. the
EEUBK model predicts that a soil-lead concentration of 370 |ig/g will control the probability of a
1-2 year old child having a blood-lead concentration greater than or equal to 10 [ig/dL (risk) to be
5%.
Using the Rochester multimedia model, floor (window sill) dust-lead loadings which
control the percentage of children having blood-lead concentration at or above 10 ug/dL to 5%
can be predicted when soil-lead concentration and window sill (floor) dust-lead loading are equal
to 400 [ig/g and 200 ^ig/ft2 (25 \ig/ft2\ respectively. This analysis resulted in a dust-lead loading
of 0.6 lie/ft2 for floors (12 ug/ft2 for window sillsV
Sensitivity analyses were performed to gauge the robustness of the methodology
employed for characterizing population-based risk. These included
Estimating baseline health effect and blood-lead concentration endpoints for a broader
age group.
5-46
-------
Estimating baseline health effect and blood-lead concentration endpoints using three
different assumptions on the decline in IQ score associated with a unit increase in
blood-lead concentration.
Estimating baseline health effect and blood-lead concentration endpoints when blood-
lead concentrations are assumed to have declined since 1994.
Estimating baseline health effect and blood-lead concentration endpoints using an
empirical approach to characterizing the distribution of blood-lead concentrations
from NHANES m data.
Estimating baseline health effect and blood-lead concentration endpoints using the
IEUBK model, under three different approaches to adjusting HUD National Survey
dust-lead concentrations to reflect the sample's total weight, under alternative
estimates for the geometric standard deviation of the blood-lead concentration
distribution, under alternative estimates for a child's daily dietary lead intake, and
under different assumptions on the prevalence of paint pica tendencies and their effect
on blood-lead concentration.
Estimating baseline health effect and blood-lead concentration endpoints using the
empirical model, under alternative estimates for the geometric standard deviation of
the blood-lead concentration distribution and under different assumptions on the
prevalence of paint pica tendencies.
The risk characterization is most sensitive to assumptions about the relationship between blood-
lead concentration and IQ score decrements. The alternative risk characterization (Section 5.1.2)
was significantly impacted by the decision to perform a tap weight adjustment to the HUD
National Survey dust samples, but much less impacted by the particular adjustment made when a
range of reasonable adjustments were considered. The alternative risk characterization was also
significantly impacted by assumptions on the geometric standard deviation associated with
blood-lead concentrations for a given lead exposure scenario, i.e., by the GSD used to represent
inter-individual variability in blood-lead concentrations.
Sensitivity analyses were also performed to gauge the robustness of the individual risk
methodology and the methodology to predict risks at background exposure levels. In particular,
the impact on individual risks of considering children 0-5 rather than 1-2 years of age was
examined and determined not to significantly impact conclusions. The impact of varying the
GSD assumed to characterize inter-individual variability on both individual risks and risks
predicted at background exposure levels was also considered and determined to be very
significant. A sensitivity analysis varying the dietary lead intake parameter of the IEUBK model
confirmed that while individual risk conclusions were not sensitive to this parameter, estimates
of risk at background levels of lead exposure were. This conclusion is highly intuitive as dietary
lead plays a much larger role at background environmental-lead exposures.
5-47
-------
In summary all methods employed in this risk assessment indicate significant risk to
children from exposure to lead and significant evidence that those risks are related to levels of
lead in paint, dust and soil in the residential environment.
5-48
-------
PART II
RISK MANAGEMENT
-------
This page intentionally blank
-------
6.0 ANALYSIS OF EXAMPLE OPTIONS FOR THE §403 STANDARDS
CHAPTER 6 SUMMARY
This chapter presents the methodology developed by EPA to characterize
reductions in childhood health effect and blood-lead concentration endpoints
expected to result after interventions are conducted in response to the proposed
$403 rule. This chapter also applies the risk management methodology to estimate
the risk reductions for a broad range of example options for the §403 standards.
For each example standard, projected health effect and blood-lead concentration
endpoints associated with predicted residential lead exposures in the post-%403
environment are compared to baseline estimates computed in Chapter 5. Post-
§403 risk is estimated separately using the IEUBK model and an empirical model
applied to environmental-lead levels observed in the HUD National Survey. Post-
%403 environmental-lead levels are adjusted for the assumed effects of
interventions initiated by the proposed §403 rule, under various example options for
standards for lead in dust, soil, and paint.
Results presented in this chapter are dependent on a number of
assumptions. A sensitivity analysis was performed to characterize this dependence.
Alternative assumptions and procedures were considered for characterizing post-
intervention blood-lead distributions. However, the largest differences in results,
especially those representing the most extreme health effects, tended to appear
when making alternative assumptions on post-intervention environmental-lead
levels.
Figure 6-1 presents the approach for risk management analysis.
Conclusions for risk management are presented in Section 6.5.
This chapter has two primary objectives:
1. Present the methodology used by EPA for evaluating options for the §403 standards.
2. Illustrate the application of this methodology for a broad range of example
standards.
The methodology and its application using example §403 standards represent risk
management analysis. The role of risk management in the overall risk analysis is outlined in
Figure 6.1 and Figure 6.3 hi Section 6.2 illustrates the approach for evaluating example options
for the §403 standards.
6-1
-------
Background
and Objectives
(Chapter 1)
Hazard
Identification
(Chapter 2)
Exposure
Assessment
(Chapter 3)
Dose-Response
Assessment
(Chapter 4)
U2S&V.-5-'
Risk
Characterization
(Chapter 5)
RISK MANAGEMENT
Present Interventions
and Their Effects on
Environmental-Lead
Levels
(Section 6.1)
Develop Methodology
for Evaluating Risk
Management
(Section 6.2)
Implement Sensitivity/
Uncertainty Analysis
for Risk Management
(Section 6.4)
Predict Post-
Intervention Blood-Lead
Distributions for
Various Examples
(Section 6.3)
Determine Blood-
Lead and Health
Effect Distributions
Under Alternative
Assumptions
(Section 6.4)
Estimate Health
Effects
Distributions for
Various Examples
(Section 6.3)
Present Conclusions
on Sensitivity/
Uncertainty Analysis
(Section 6.4)
Conclusions
on Risk
Characterization
I
Conclusions on
Analysis of
Example Options
for §403 Standards
(Section 6.5)
Figure 6-1. Detailed Flowchart of the Approach to Risk Management.
6-2
-------
The risk management methodology is used to predict childhood health effect and blood-
lead concentration endpoints that are expected to result after activities are conducted in response
to the proposed §403 standards. Estimating the selected endpoints after the §403 rulemaking is
promulgated (post-§403) required developing an answer to each of the following questions:
1. What will home owners do in response to the proposed rulemaking?
2. How much will environmental-lead levels change due to the activities of home
owners?
3. How many homes will be affected by the rulemaking?
4. How much will blood-lead levels change due to changes in the distribution of
environmental-lead levels?
5. How much will health effect endpoints change due to changes in the distribution of
blood-lead concentrations for children aged 1-2 years.
For the purposes of the risk management analyses, a set of six interventions were defined
and utilized for modeling actions homeowners would take in response to the §403 standards.
The definition of each of these intervention strategies includes then* efficacy in terms of the
expected reductions hi environmental-lead levels, the expected duration of their effectiveness,
and the circumstances (i.e., environmental-lead levels) under which they will be performed
(Section 6.1). The defined interventions are an essential component of the methodology
developed by EPA to evaluate various options for risk management. The risk management
methodology is presented in Section 6.2. This methodology is utilized to estimate the number of
homes affected by the rulemaking, and to predict post-§403 blood-lead concentration and health
effect endpoints for children aged 1-2 years. Application of the risk management methodology is
illustrated for a broad range of example §403 standards in Section 6.3. A sensitivity analysis on
the effects of the uncertainty present in the key assumptions, parameters, data sources, and
analysis tools is presented in Section 6.4. The sensitivity analysis examines the impact on the
predicted blood-lead concentration and health-effects distributions of changes to the key pieces
of the risk management methodology. Finally, conclusions derived from the analysis of the
various example options for the §403 standards are stated in Section 6.5.
6.1 INTERVENTION ACTIVITIES
Once defined, the proposed §403 rule will prompt intervention activities targeting
residential lead hazards. These interventions will be conducted on behalf of children already
exposed to the targeted lead hazards, as well as children who would otherwise be exposed if the
hazards are not abated or controlled. For the purposes of the risk management analyses, a lead
hazard intervention is defined as any non-medical activity that seeks to prevent a child from
being exposed to the lead hi his or her surrounding environment. An intervention, therefore, may
range from the education of parents regarding the dangers of a young child's hand-to-mouth
activity, to the abatement of lead-based paint.
6-3
-------
An intervention conducted on behalf of children already exposed to the targeted hazard is
termed secondary prevention (e.g., paint abatement hi the home of a child who has an elevated
blood-lead concentration). A primary prevention intervention prevents exposure before it occurs
(e.g., paint abatement in a home before a new family with children moves hi). The distinction
between primary and secondary prevention efforts is one of the population targeted rather than
the activity conducted. In fact, a given intervention can have both primary and secondary
prevention benefits.
One objective of §403 is to prompt primary prevention interventions targeting lead
hazards hi residential soil, dust, and paint. (Secondary prevention will, of course, also take
place.) As the risk analysis needs to model the expected benefits following promulgation of
§403, measures of the effectiveness of these lead hazard interventions are required.
Unfortunately, there is no information currently available hi the scientific literature regarding the
efficacy (as measured by either avoided health outcomes or by prevented changes hi children's
blood-lead concentrations) of primary prevention interventions targeting paint, dust, or soil.
There are limited data on the effectiveness of secondary prevention interventions (USEPA,
1995b, 1998).
Research suggests that primary prevention interventions will produce greater efficacy
than secondary prevention interventions (Gulson et al., 1995). Bone-lead stores accumulated by
exposed children continue to mobilize into the blood following an intervention and may mask the
intervention's full effectiveness. The reported effectiveness of interventions studied hi the
literature, therefore, has shortcomings as an estimate of the efficacy stemming from the
interventions when performed as primary prevention. Thus, these blood lead declines from a
secondary prevention situation are not used to assess the health effect and blood-lead
concentration endpoints stemming from the §403 rule. However, a method which uses the
changes in blood-lead concentrations from secondary prevention settings with a modeled effect
of the bone lead stores is examined hi the sensitivity analysis (Section 6.4.4).
Data on changes hi environmental-lead levels for interventions targeting paint, dust, and
soil were used to estimate environmental-lead levels following interventions conducted as a
result of §403. It is important to note, however, that only some of the interventions considered
viable hi the regulatory and scientific communities have been studied and documented hi the
literature. Where published data are available, the reported post-intervention environmental-lead
levels may then be translated into blood-lead concentrations representing the benefit of primary
prevention interventions. The translation is accomplished using both the IEUBK model and the
empirical model. Where little or no environmental effectiveness information is available to
characterize a particular intervention, EPA has used its current understanding of the intervention
to develop an estimated effectiveness.
Fully characterizing an intervention requires addressing four questions:
1. What ''triggers" the intervention?
2. What procedures are conducted during the intervention?
6-4
-------
3. How effective is the intervention at reducing environmental-lead levels?
4. What is the duration for which the environmental-lead levels will remain reduced?
The interventions and their associated procedures utilized in the risk management
analyses are discussed in Section 6.1.1. The effectiveness of these methods in reducing
environmental-lead levels and blood-lead concentrations is documented in Section 6.1.2 and
6.1.4, respectively. Section 6.1.3 discusses the circumstances under which each of the defined
interventions is triggered. The methods used to predict childhood health effect and blood-lead
concentration endpoints after performing the defined interventions in response to the proposed
§403 rule are presented in Section 6.2.
The characteristic of greatest importance associated with an intervention is its ultimate
effect on children's blood-lead levels. In this analysis, this characteristic is estimated using two
blood-lead concentration prediction models, the IEUBK model (Section 4.1), and an empirical
model (Section 4.2). The impacts of the interventions, as triggered by different example options
for the standards, are illustrated in Section 6.3.
6.1.1 Interventions
For the purposes of evaluating various example options for risk management, a total of
six interventions were defined for lead in paint, dust, and soil. The six interventions are dust
cleaning, exterior LBP maintenance, exterior LBP encapsulation/abatement, interior LBP
maintenance, ulterior LBP encapsulation/abatement, and soil removal. For interior paint and
exterior paint, two intervention approaches were defined. These two approaches are intended to
reflect the viable range in scope achieved by paint interventions. For residential soil, there are
several options for reducing exposure to elevated soil-lead levels - soil removal, soil till, or sod,
mulch, or pavement application. As EPA has data only on the effectiveness of soil removal (no
data on soil cover efficacy are available in the scientific literature), this approach is the only soil
intervention evaluated in the analysis of example options for risk management. For residential
dust, a dust-cleaning method was included to follow interior LBP interventions and soil removal.
The dust-cleaning method is also applied in homes where interior dust-lead loadings are high and
no residential sources of lead (LBP or soil lead) are identified. Table 6-1 presents these six
interventions by defining the procedures conducted and the expected duration of the
intervention's benefits.
The procedures defined in Table 6-1 for each of the interventions are consistent with
intervention practices currently recommended by EPA (and mandated in some communities) in
§402 and by HUD in its "Guidelines for the Evaluation and Control of Lead-Based Paint Hazards
in Housing," (HUD, 1995b). For example, paint removal must be conducted using appropriate
precautions (e.g., avoid soil or dust contamination), and LBP encapsulation must utilize materials
approved as encapsulants (i.e., remain effective for 20 years). The procedures exclude
interventions previously utilized but now considered hazardous, such as open-flame burning or
abrasive sanding of lead-based paint.
6-5
-------
Table 6-1. Interventions Defined for the §403 Risk Analysis Effort.
Intervention
Procedures Defining the Intervention
Expected Duration1
Dust Cleaning
Cleaning the unit using HEPA vacuums and wet
mopping.
4 years or
permanent2
Exterior
LBP
Maintenance
Painted surfaces with deteriorated LBP are repaired by
feathering the edges of deteriorating paint and
repainting with new, lead-free paint (less than 0.06%
lead by weight). Measures are taken to preclude soil
contamination during intervention.
4 years for paint
Encapsulation/
Abatement
Deteriorated LBP is removed, and the affected surface
encapsulated or enclosed, if necessary, using
currently acceptable practices and materials.
Measures are taken to preclude soil contamination
during intervention.
20 years for paint
Maintenance
Interior
LBP
Painted surfaces with deteriorated LBP are repaired by
feathering the edges of deteriorating paint and
repainting with new, lead-free paint. Window sills are
covered with permanent barrier. A Dust Cleaning of
the affected area follows the intervention.
4 years for paint,
4 years for dust
Encapsulation/
Abatement
Deteriorated LBP is removed, and the affected surface
encapsulated or enclosed, if necessary, using
currently acceptable practices and materials. A Dust
Cleaning of the housing unit follows the intervention.
20 years for paint,
permanent2 for dust
Soil Removal
Soil from areas with elevated lead concentrations are
removed and replaced with clean soil, or the areas are
permanently covered. A Dust Cleaning of the housing
unit follows the intervention.
Permanent
1 Duration is defined as the length of time before the lead levels in the targeted medium or conditions of the medium require
further intervention.
2 If the cleaning is accompanied by paint and soil abatements, the duration of reduced dust-lead levels is permanent (20
years if accompanied by paint abatement).
The specified durations of the interventions reflect the length of time before the targeted
media returns to levels or conditions requiring further interventions. For example, the duration
of a paint intervention represents the estimated period of tune before formerly intact or repaired
surfaces deteriorate. The expected duration of an intervention is assumed applicable only to
residential environments consistent with the circumstances under which that intervention would
be triggered (Section 6.1.3). When defining the duration of interior lead-based paint abatements,
moreover, the duration of reduced ulterior residential dust-lead levels is also defined. Since paint
interventions target only deteriorated lead-based paint, it would be unrealistic to assume that
dust-lead levels remain low permanently. The once intact lead-based paint could, over time,
deteriorate and produce elevated lead levels in residential house dust. Some deterioration over
time, if not due to normal abrasion (e.g., opening and closing windows), can be avoided if the
intact paint is properly maintained. Moreover, one can reasonably expect such maintenance from
most homeowners.
6-6
-------
Unfortunately, there were only limited data available for estimating the duration of the
methods defined in Table 6-1. Though numerous intervention studies are documented in the
literature, none traced effectiveness for more than a couple of years (most less than one year).
The effectiveness durations, and their underlying motivations, outlined in the draft HUD
Regulatory Impact Analysis (RIA) (ICF, 1995, pages 3-21 through 3-22) provided a reasonable
starting point. The HUD RIA utilized 4 and 8 years as the duration of reduced dust-lead levels
following interim paint controls and paint abatements, respectively. These durations were based
on estimates of the annual rate of increased dust-lead loading (ng/ft2 per year) stemming from
residential recontamination reported in studies of LBP interventions conducted in Baltimore and
Cincinnati (page 3-22). A standard of 100 ug/ft2 was considered hi deriving the recontamination
rates (e.g., floor dust lead was estimated to reaccumulate to levels exceeding 100 ug/ft2 by four
years following repair of the deteriorated lead-based paint). Though a different standard would
imply a different duration, a constant duration of four years was assumed following paint
maintenance in the risk analysis. The duration for reduced dust-lead loading following paint
abatement was assumed to be consistent with the duration assumed for the intervention itself.
Lead levels would only re-elevate when their source reappeared. The HUD RIA assumed a paint
duration of 20 years following lead-based paint abatement. This degree of efficacy is consistent
with HUD's definition of a LBP abatement practice: requiring the abatement to be effective for
at least 20 years hi order to be called an abatement. Since "paint repair should provide
approximately five years of protection against significant amounts of deteriorated LBP" (page
3-22), the HUD RIA assumed a paint duration of 5 years following lead-based paint
maintenance. A more conservative effectiveness duration of 4 years (Table 6-1) is assumed in
the risk management analyses for both the paint itself and the surrounding dust. Finally, the
assumption that soil removal intervention has a permanent effectiveness (Table 6-1) was made
since the soil exhibiting elevated lead concentrations has been either removed or permanently
covered.
6.1.2 Reductions in Environmental Lead Levels Following Interventions
The effectiveness of the interventions outlined hi Table 6-1 is defined in terms of how
environmental-lead levels are reduced following conduct of the intervention. Table 6-2 presents
the assumed post-intervention environmental-lead levels for each of the interventions described
in Table 6-1. For each intervention, the post-intervention lead levels are defined for those media
expected to be affected by the intervention. For example, ulterior paint abatement can be
expected to prompt reductions hi ulterior dust-lead loadings as well as in interior paint-lead
loadings. Where relevant, additional details are provided regarding the effectiveness of the
interventions.
The interventions outlined hi Tables 6-1 and 6-2 are intended to include state-of-the-art
practices. As a result, defining the effectiveness of these interventions as measured by reduced
environmental-lead levels is difficult. Though numerous intervention studies are documented in
the literature, many utilized methods that today would be considered inappropriate. The
available information on intervention effectiveness often is of little relevance. Where possible,
however, the available data were utilized.
6-7
-------
Table 6-2. Expected Post-Intervention Lead Levels Associated With Performing §403
Interventions.
Intervention
Post-intervention Lead Ltvai
Commwn* on Performing
th» Intervention
Dust Cleaning1
:loors: Wipe dust-lead loading equals minimum of 40
//g/ft1 and pre-intervention level
Dust-lead concentration is determined by the
approach outlined in Figure 6-2
Window Sills: Wipe dust-lead loading equals minimum
of 100//g/ft1 and pre-intervention level
it is assumed that this intervention
would occur as the sole
intervention only if dust-lead levels
were above the standard, and if no
sources of lead exposure remain in
the housing unit.
Exterior
LBP
Maintenance
0 square feet of deteriorated exterior LBP
Deteriorated LBP is eliminated as a
potential exposure source for the
duration specified in Table 6-1.
Encapsulation/
Abatement
0 square feet of deteriorated exterior LBP
Deteriorated LBP is eliminated as a
potential exposure source for the
duration specified in Table 6-1.
Maintenance
0 square feet of deteriorated interior LBP
Floor and window sill dust-lead loading unchanged from
pre-intervention levels
Floor dust-lead concentration is determined by the
approach outlined in Figure 6-2
Deteriorated LBP is eliminated as a
potential exposure source for the
duration specified in Table 6-1.
Interior
LBP
Encapsulation/
Abatement
0 square feet of deteriorated interior LBP
Floors: Wipe dust-lead loading equals minimum of 40
//g/ft1 and pre-intervention level
Dust-lead concentration is determined by the
approach outlined in Figure 6-2
Window Sills: Wipe dust-lead loading equals minimum
of 100 //g/ft1 and pre-intervention level
Deteriorated LBP is eliminated as a
potential exposure source for the
duration specified in Table 6-1.
Soil Removal
Soil-lead concentration equals 150 ppm in areas where
soil removal is conducted
Floors: Wipe dust-lead loading equals minimum of 40
//g/ft* and pre-intervention level
Dust-lead concentration is determined by the
approach outlined in Figure 6-2
Window Sills: Wipe dust-lead loading equals minimum
of 100 //g/ft* and pre-intervention level
Residential dust is not
recontaminated by the intervention
1 Triggered by window sill as well as floor dust-lead loadings exceeding their respective standards.
Encapsulation/abatement of interior paint is assumed to reduce residential floor and
window sill dust-lead loadings to 40 and 100 ng/ft2, respectively, while effectively eliminating
(for the duration outlined in Table 6-1) the hazard from deteriorated lead-based paint. The same
degree of effectiveness with regard to residential dust was assumed for soil removal and dust
cleaning. Dust-lead loadings were not reduced following maintenance of interior paint because
the dust cleaning accompanying interior paint maintenance was assumed to be conducted in the
affected area only. These values were selected after considering the efficacy reported for housing
units in the Denver Comprehensive Abatement Performance (CAP) Study and in the Baltimore
Experimental Paint Abatement Study. The geometric mean floor vacuum dust-lead loading
measured in abated units studied by the Denver CAP Study was 29.0 fig/ft2 approximately two
years following extensive paint abatements; the geometric mean window sill vacuum dust-lead
loading was 91.6 ^ig/ft2 among the same units (page 34 of USEPA, 1995e). Similarly, the
Baltimore Experimental Paint Abatement Study reported a geometric mean floor wipe dust-lead
6-8
-------
loading of 40.9 ng/ft2 among 13 housing units 18-42 months following complete paint
abatements; a geometric mean of 103 jig/ft2 was reported for the unit's window sill wipe dust-
lead loadings at the same time (page 62 of USEPA, 1995c).
The flowchart presented in Figure 6-2 illustrates the approach for determining post-
intervention floor dust-lead concentrations. For instance, the upper most branch in the top half of
Figure 6-2 (Branch 1) presents the method for determining post-intervention dust-lead
concentration if a soil intervention and either paint abatement or maintenance are conducted:
post-intervention dust-lead concentration on floors is set equal to 20 percent of the pre-
intervention dust-lead concentration. The upper-most branch in the lower half of Figure 6-2
(Branch 4) presents the method for determining post-intervention dust-lead concentration if a soil
intervention is not conducted and either paint abatement or maintenance is conducted: post-
intervention dust-lead concentration on floors is set equal to 20 percent of the pre-intervention
dust-lead concentration. The assumption of an 80 percent reduction in dust-lead concentration
following conduct of a paint intervention was developed by examining the results of paint
interventions conducted in the Study Group homes in the Boston phase of the Urban Soil Lead
Abatement Performance Study (USLADP), and in the R&M Level HI homes in the Baltimore
R&M Study.
The remaining branches portrayed in Figure 6-2 are applicable to determining post-
intervention dust-lead concentration if no paint abatement or maintenance is conducted.
Branches 2 and 3 are employed if a soil intervention is conducted. Branches 5, 6 and 7 are
applied if no soil intervention is conducted. Branches 2 through 6 are based on the assumption
that approximately 80 percent of interior floor dust mass comes from the surrounding soil, and
the additional assumption that the lead in soil is uniformly distributed across particle sizes that
migrate and become interior dust. The assumption that 80 percent of the mass of interior floor
dust stems from soil was examined using the data from the Baltimore phase of the USLADP.
The assumption that the lead hi soil is uniformly distributed across particle sizes that can migrate
and become interior dust was made by necessity. EPA is not aware of any available data to
evaluate this assumption.
Reducing amounts of deteriorated LBP to zero square feet following the four paint
interventions is consistent with the procedures defined for the interventions and their assumed
durations. These interventions are defined as utilizing practices that ensure the surfaces with
deteriorated paint remain intact following the intervention for the specified duration. Recall that
the durations were defined to recognize the potential for paint, intact at the time of the
intervention, becoming deteriorated over time. Thus, the potential hazard posed by deteriorated
paint is assumed to be completely eliminated by each of the interventions (both ulterior and
exterior) for the durations specified in Table 6-1.
The post-intervention soil-lead concentration assumed following soil removal is primarily
dependent upon the concentration of lead in the backfill soil used to replace the contaminated
soil being removed. Usually soil with a lead concentration at background or minimal level is
utilized for backfill. Given that the recontamination of the new soil by remaining lead-
contaminated soil at the residence may occur, EPA chose to assume a post-intervention soil-lead
concentration of 150 ppm.
6-9
-------
Paint Intervention
Triggered?
Post-Intervention Floor
Dust-Lead Concentration
Soil
Removal
Triggered?
YES
YES
PbDpoa=0.2*PbDpn
Deteriorated Lead-
Based Paint Present?
YES
PbDpoa = mm(PbDpn, 0.8*PbSp
-------
elevated levels. If either dust, soil, or paint exhibit levels in excess of those specified by the §403
standards, appropriate interventions that target the problematic media are assumed.
Table 6-3 summarizes the circumstances under which each of the defined interventions in
Table 6-1 is conducted. At a given residence, the circumstances outlined in Table 6-3 could
trigger multiple interventions. The choice of an encapsulation/abatement approach versus a
maintenance approach to paint intervention is based on the extent to which deteriorated lead-
based paint is present. Although in practice the choice between paint maintenance and
encapsulation/abatement is based on a number of factors, including surface area and location of
deteriorated LBP (USHUD, 1995b), for the purposes of the risk management analyses only
surface area of deteriorated LBP is being considered. As noted earlier, dust cleaning is prompted
as a clean-up activity following an interior paint intervention or soil removal, or when elevated
dust-lead levels are observed despite the absence of residential sources of lead exposure (e.g.,
soil or paint). Non-residential lead sources are assumed absent.
Table 6-3. Intervention Triggers Defined for the §403 Risk Management Analyses.
Intervention
Dust Cleaning
Exterior
LBP
Interior
LBP
Maintenance
Encapsulation/
Abatement
Maintenance
Encapsulation/
Abatement
Soil
Removal
Circumstances Prompting Conduct of the Intervention
Follows any interior paint intervention or soil removal, and when dust-lead
loadings are elevated despite absence of residential sources of lead
exposure (e.g., no deteriorated LBP or elevated soil lead).
When deteriorated exterior LBP is present, but not extensive (e.g., confined
to a limited area).
When deteriorated exterior LBP is present and extensive (e.g., not confined
to a limited area).
When deteriorated interior LBP is present, but not extensive (e.g., confined
to one area of the housing unit).
When deteriorated interior LBP is present and extensive (e.g., greater than
one area of the housing unit).
When residential soil-lead concentrations exceed the soil standard. It is
assumed this degree of intervention would only be warranted in specific
areas of the yard (e.g., dripline, entry way).
6.1.4 Reductions in Blood-Lead Levels Following Interventions
For each home in the HUD National Survey, the post-intervention environmental-lead
levels presented in Table 6-2 were employed to predict the blood-lead concentrations of resident
children. If an intervention was triggered by one or more of the example standards, then the
environmental-lead levels in the post-intervention time frame were set equal to those displayed in
Table 6-2. The IEUBK and empirical models (discussed hi Sections 4.1 and 4.2) were employed
to predict a geometric mean blood-lead concentration for children aged 1-2 years exposed to
environmental conditions represented by the residence following the intervention activity. In this
manner, the risk management analyses estimated the impact of various example options for the
§403 standards on environmental-lead levels in the nation's housing and blood-lead
concentrations in children aged 1-2 years.
6-11
-------
6.2 METHODOLOGY FOR EVALUATING RISK MANAGEMENT OPTIONS
This section describes the methods developed to predict distributions of blood-lead
concentrations and values of specific health effect endpoints following promulgation of the
proposed §403 rule. This methodology is applied in Section 6.3 to characterize risk reductions
associated with various example sets of standards. These example standards are associated with
interventions designed to reduce environmental-lead levels, which in turn are expected to reduce
lead exposure and consequently reduce blood-lead concentrations. The consequent distribution
of environmental-lead levels expected to result from the example standards, together with the
BEUBK or empirical models, were used to estimate the post-§403 distribution of blood-lead
concentrations. The health effect and blood-lead concentration endpoints were then computed
from the post-§403 distribution of blood-lead concentrations.
The methodology presented in this section is illustrated in Figure 6-3. In this figure,
boxes with rectangular corners represent datasets or tables of results. Boxes with rounded
corners represent steps in the process that transform the inputted data either through a
predictive model (e.g., the IEUBK and empirical models are used to predict blood-lead
concentrations from environmental-lead levels) or computation.
The four numbered steps in the process are illustrated by the four boxes with rounded
corners in Figure 6-3. These steps form the basis of the methodology and are now presented in
more detail.
Step 1: Predict Post-S403 Environmental-Lead Levels. Data from the HUD National
Survey were used to predict post-§403 environmental-lead levels. As described in Chapter 3, the
HUD National Survey included 284 homes for which data are given on lead levels in dust, soil,
and paint. Each of the homes in the survey represented a specified number of U.S. homes having
similar environmental-lead levels. These data were used to infer how many U.S. homes would
have lead levels above specified examples for the §403 standards for dust, soil, and paint.
In the risk management analyses, example dust standards are defined in terms of wipe
dust-lead loadings. However, dust samples in the HUD National Survey were collected with the
Blue Nozzle vacuum method. Therefore, to infer which HUD National Survey houses had floor
or window sill dust-lead loadings above the wipe standard, the Blue Nozzle lead loadings were
converted to "equivalent" wipe dust-lead loadings and compared to the example standards. This
conversion process is summarized in Section 4.3 and described hi more detail hi USEPA, 1998.
Other conversions necessary for predicting blood-lead levels from environmental-lead levels are
discussed in Step 2.
Table 6-2 in Section 6.1.3 presented the assumed effect of a §403 intervention on
environmental-lead levels, and Table 6-3 displayed the circumstances assumed for triggering an
intervention. To illustrate how the procedures hi the tables were applied hi the risk management
analyses, the following text discusses their application for a particular example set of standards to
house number 0411207 in the HUD National Survey. The standards and intervention trigger
levels assumed for this example are:
6-12
-------
HUD National Survey Data on Environmental Levels
(Pre-§403 Environmental Levels)
Dust
Soil
Paint
§403
Intervention
Paint
Post- §403
Environmental
Levels
lEUBK/Empirical
PbB Prediction Model
Modeled
Pre-5403
Blood-Lead
Distribution
Modeled
Post- §403
Blood-Lead
Distribution
NHANES
III
Calculate Change In Modeled Blood-
Lead Distribution (Post Minus Pro)
and
Derive Post-§ 403 Blood-Lead
Distribution Based on Modeled Change
and NHANES III
Risk Summarization
(Estimation of Health Effects)
Impact of Standards
IK Houses Affected
Health Endpolnts
PbB> 20
PbB> 10
IQ <"70
Average
IQ decrement *1 .2,3
Steps in the process
Data sets or results of applying steps
Figure 6-3. Post-§403 Risk Management Process.
6-13
-------
floor dust-lead loading: 200 ug/fta
window sill dust-lead loading: 500 ug/ft2
soil-lead concentration: 3,000 (ig/g
paint (maintenance): 10-40 ft2 of damaged LBP
paint (abatement): 40 ft2 of damaged LBP.
House number 0411207 had the following environmental-lead levels measured in the National
Survey, representing baseline (pre-intervention) levels:
floor dust-lead loading (wipe-equivalent): 235.5 |ig/ft2
floor dust-lead concentration: 1,812 ng/g
window sill dust-lead loading (wipe-equivalent): 14.6 ng/ft2
soil-lead concentration: 805 |ig/g
maximum XRF: 0.4 mg/cm2
damaged LBP: 0 ft2.
Because there is no damaged LBP at this house, no paint intervention is triggered regardless of
the example intervention trigger level. There is no change, therefore, in the paint-lead loading or
in the amount of deteriorated LBP. The pre-intervention soil-lead concentration was 805 |ag/g
which is below the example soil removal standard of 3,000 ng/g. So this remains unchanged.
The floor dust-lead loading of 235.5 (ig/ft2 exceeds the example standard of 200 |ig/ft2, triggering
a dust cleaning. The floor dust-lead loading is, therefore, reduced to 40 ug/ft2. The window sill
dust-lead loading was unchanged, as it was below the assumed post-intervention window sill
dust-lead loading of 100 ng/ft2. As no soil or paint interventions were triggered, only a dust
cleaning was triggered. Branch 6 of Figure 6-2 indicates that the post-intervention floor dust-
lead concentration is equal to the minimum of the pre-intervention concentration (1,812 iig/g)
and 80% of the post-intervention soil-lead concentration (0.8 x 805 = 644 |ig/g). Thus, the floor
dust-lead concentration is reduced to 644 ug/g. Therefore, the post-§403 environmental-lead
levels assumed for house 0411207 based on the example standards listed above are:
floor dust-lead loading: 40 ug/ft2
floor dust-lead concentration: 644 ug/g
window sill dust-lead loading: 14.6 ug/ft2
soil-lead concentration: 805 ng/g
maximum XRF: 0.4 mg/cm2
damaged LBP: 0 ft2.
Step 2: Use EBUBK/empirical models to predict blood-lead concentrations. The second
step in characterizing post-§403 health risks was to use the residential environmental-lead levels
and assumptions on pica tendencies to predict blood-lead concentrations hi children. As
discussed in Chapter 4, two different models were used to predict blood-lead levels from
environmental-lead levels. It was also necessary for reasons described hi Step 3 (below), to
characterize the pre-§403 distribution of children's blood-lead concentrations using the same two
models used to predict post-§403 blood-lead levels.
6-14
-------
This section consists, therefore, of three parts: a description of the two models used to
predict blood-lead levels from environmental-lead levels, the prediction of pre-§403 blood-lead
levels, and the prediction of post-§403 blood-lead levels.
Description of Models:
The first model utilized is EPA's IEUBK model introduced in Section 4.1. This model
was used to predict blood-lead levels for children aged 1-2 years. The IEUBK model uses dust-
lead concentrations and soil-lead concentrations to predict a geometric mean blood-lead
concentration. The IEUBK model predicted blood-lead concentration was then adjusted to
account for the contribution of damaged LBP at the house, and a tendency for paint pica. Details
of this pica adjustment were presented in Section 4.1.3.
A second model, referred to as the empirical model, was also used to predict blood-lead
levels from environmental-lead levels. This model was developed using the data in the
Rochester Lead-in-Dust Study. It requires as inputs dust-lead loadings from floors and window
sills collected by the Blue Nozzle vacuum method, as well as soil-lead concentrations.
Information on the amount of damaged LBP is directly incorporated in the empirical model to
estimate the contribution of pica for paint or childhood blood-lead concentration. Details of the
model are provided in Section 4.2 and Appendix G. Thus, the two models used to predict blood-
lead concentrations depend on different sets of inputs and emphasize the inputs differently. It is
therefore expected that there are differences in the geometric mean blood-lead concentrations
predicted by the two models.
For each home in the HUD National Survey, both the IEUBK and the empirical model
were used to predict a geometric mean blood-lead concentration for children aged 1-2 years who
would be exposed to the given environmental conditions. However, not all of these children will
have the same blood-lead levels. Appendix E2 describes the approach taken to characterize the
variability in blood-lead levels about the estimated geometric mean associated with each house
and how this information was used to determine a distribution of children's blood-lead
concentrations.
Predicting Pre-§403 Blood-lead Levels:
The environmental-lead levels collected in each home in the HUD National Survey were
used as input to the blood-lead concentration prediction models to obtain a baseline distribution
of blood-lead concentrations more directly comparable to the distribution predicted from the
post-§403 distribution of environmental-lead levels. As described above, the IEUBK model
requires as input floor dust-lead concentrations and soil-lead concentrations. Blue Nozzle dust-
lead concentrations were used as input to the model. The empirical model requires floor and
window sill dust-lead loadings measured by the Blue Nozzle vacuum method, and soil-lead
concentrations. Both models require an indication of the presence of damaged LBP. Because
these model-required inputs were all measured in the HUD National Survey, no conversions were
necessary to estimate the pre-§403 blood-lead concentrations.
6-15
-------
Predicting Post-§403 Blood-lead Levels:
The predicted post-§403 environmental-lead levels determined in Step 1 were used as
input to the blood-lead concentration prediction models to estimate the post-§403 distribution of
blood-lead levels. However, in the case of an intervention requiring dust cleaning, the projected
post-intervention floor dust-lead loading of 40 ^ig/ft2 is based on the wipe method (see Table
6-2). For input to the IEUBK model, the post-intervention dust-lead concentration was
determined according to the approach in Figure 6-2. For input into the empirical model, 40
ug/ft2 was converted to a Blue Nozzle equivalent dust-lead loading of 5.8 ug/ft2 using the method
described in Section 4.3. Similarly, the projected post-intervention window sill dust-lead loading
of 100 ug/ft2 based on the wipe method was converted to a Blue Nozzle equivalent dust-lead
loading of 14 ug/ft2.
Step 3: Adjust predicted post-§403 blood-lead concentrations using (NHANES HP
baseline information. Step 2 described the process used to estimate the pre- and post-§403
distribution of blood-lead concentrations predicted to derive from environmental-lead levels.
Step 3 determines the change in blood-lead concentrations resulting from the intervention (post-
§403 minus pre-§403), and applies this change to the distribution of blood-lead concentrations
inferred from Phase 2 of NHANES m. This step is necessary because the NHANES m data are
regarded as the basis for the most reliable baseline characterization of children's blood-lead
concentration available. The IEUBK and empirical models applied in Step 2 to environmental-
lead levels derived from the HUD Survey, however, are the best tools available for estimating the
change in blood-lead concentration associated with an intervention. Thus, there are three inputs
to this step of the process:
1. A model-predicted, pre-§403 distribution of blood-lead concentration based on
unadjusted HUD National Survey environmental-lead levels
2. A model-predicted, post-§403 distribution of blood-lead concentration based on
projected post-§403 environmental-lead levels
3. A baseline distribution of blood-lead concentration based on NHANES in data.
In this step, the difference between pre-§403 modeled blood-lead concentration and post-§403
modeled blood-lead concentration is applied to the baseline distribution of blood-lead
concentration inferred from NHANES HI. The details of this step are described in Steps (1)
through (4) of Appendix Fl. The result is an estimate of the post-§403 geometric mean and the
geometric standard deviation of blood-lead levels in the nation. These estimates are used to
predict health risks to children in the next step.
In reality, of course, a myriad of other factors in addition to residential environmental-
lead levels will influence the post-§403 national distribution of blood-lead levels. The baseline
distribution of blood-lead concentration was first presented in Section 5.1.1. For reasons stated
in Section 5.1.1, the baseline distribution, which is based on data collected from 1991 to 1994
was not projected to 1997. Because of this, both baseline risks due to lead exposure and
predicted post-§403 risks due to lead exposure may be overestimated for 1997.
6-16
-------
Step 3 was necessary to ensure that the predicted distribution of blood-lead
concentrations (as derived from the HUD National Survey environmental-lead levels) would
yield appropriate results when high example standards are considered. In particular, if the
selected example standards were set sufficiently high as to trigger no interventions, then the
predicted distribution should agree with the distribution predicted by NHANES ffl (which is
regarded as the best characterization available). When the model predictions are used directly,
no such agreement is evident. Appendix Fl describes the details of the approach that was
developed to correct this inconsistency.
Step 4: Predict health effects and blood lead endpoints for children 1 -2 years old. The
last step hi the process is the summarization of health effect and blood-lead concentration
endpoints associated with the baseline and the predicted post-§403 distributions of blood-lead
concentrations. This step estimates the proportion of children with blood-lead levels at or above
specified thresholds, the proportion of children anticipated to experience IQ decrements of
specified amounts due to elevated blood lead concentrations, the proportion of children with IQ
levels below 70 due to elevated blood-lead concentrations, and the average IQ point decrement in
children due to elevated blood-lead concentrations. Each of these endpoints is estimated from
the geometric mean and geometric standard deviation of children's blood-lead concentrations,
assuming a lognormal distribution. The mathematical approach used to make these inferences is
described in Appendix El.
6.3 RESULTS OF THE EVALUATION OF EXAMPLE RISK MANAGEMENT OPTIONS
This section applies the methods presented in Section 6.2 to evaluate the health effect and
blood-lead concentration endpoints associated with various example options for the §403
standards. Several different example sets of standards were selected for illustrative purposes.
The example standards examined are not meant to encompass all possible options for the §403
rule, and the Agency fully anticipates considering other sets of candidate standards. To simplify
the presentation, the effect of changing the levels hi the example standards is examined
separately for dust, soil, and paint. To accomplish this, it was necessary to hold the example
standards for soil and paint fixed at a specified level while the levels were varied for dust.
To illustrate the approach taken hi this section, Table 6-4 presents estimated health effect
and blood-lead concentration endpoints and percentages of housing units exceeding the standards
under six sets of example options (A-F) for floor and window sill dust-lead loading standards.
In this example, the soil removal is triggered if soil-lead concentration exceeds 3,000 u.g/g, paint
maintenance is prompted at 5 ft2 of damaged LBP, and paint abatement at 20 ft2 of damaged
LBP. For this illustration, the example standards range from 25 to 400 ug/ft2 (hi reverse order)
for floor dust-lead loading, and from 25 to 800 ug/ft2 for window sills. Each column (A-F) hi
Table 6-4 is devoted to a specific pair of example standards for floor and window sill dust-lead
loading. For each example set of standards, the rows in the top part of Table 6-4 indicate the
percentage of homes that would exceed each of the example floor and window sill dust-lead
loading standards, the percentage of homes that would exceed either the floor or window sill
dust-lead loading example standards, and the percentage of homes that would exceed any one of
the example standards for dust, soil, or paint specified in this table.
6-17
-------
Table 6-4. Characterization of Impact of Various Example Options for Dust Standards:
Soil and Paint Standards Fixed (3,000 //g/g for Soil Removal, 5 ft2 Damaged
LBP for Paint Maintenance, 20 ft2 Damaged LBP for Paint Abatement).
Example Options for Dust-Lead Loading Standard (pg/ft*)
EXAMPLE OPTION CODE
Floor Dust Standard
Window SHI Dust Standard
Percentage of Homes
Exceeding Floor Dust
Standard
Percentage of Homes
Exceeding Window SHI Dust
Standard
Percentage of Homes
Exceeding Any Dust
Standard
Percentage of Homes
Exceeding Any Dust, Soil, or
Paint Standard
A
400
800
0.00
10.3
10.3
19.5
B
200
600
0.694
12.5
13.0
21.0
C
100
500
4.04
12.5
13.9
21.6
D
100
200
4.04
24.3
25.5
30.9
E
50
100
8.28
32.5
34.4
37.9
F
25
25
13.8
48.1
50.6
52.6
Baseline
Predicted Health Effect and Blood-Lead Concentration Endpoints (Based on Empirical Model)
PbBi20pg/dL(%)
PbBi10pg/dL(%)
IQ<70<%)
IQ decrements 1 (%)
IQ decrement;^ (%)
10 decrements 3 (%)
Avg. IQ decrement
0.442
4.93
0.111
36.9
9.65
3.09
1.02
Predicted Health Effect and
PbB*20//g/dL{%>
PbB*10pg/dL(%)
IQ<70<%)
IQ decrements 1 {%}
IQ decrement:^ (%)
IQ decremented (%)
Avg. IQ decrement
0.168
2.97
0.104
32.4
6.61
1.71
0.920
0.430
4.85
0.111
36.7
9.53
3.04
1.01
0.427
4.84
0.111
36.6
9.50
3.02
1.01
0.410
4.72
0.110
36.3
9.33
2.94
1.01
0.396
4.63
0.110
36.1
9.19
2.88
1.00
0.395
4.62
0.110
36.1
9.17
2.87
1.00
0.588
5.75
0.115
38.5
10.8
3.70
1.06
Blood-Lead Concentration Endpoints (Based on IEUBK Model)
0.153
2.82
0.103
31.9
6.35
1.61
0.910
0.0991
2.28
0.101
30.5
5.44
1.24
0.885
0.101
2.23
0.100
29.4
5.27
1.22
0.868
0.0867
2.01
0.0994
28.1
4.84
1.09
0.847
0.0749
1.74
0.0978
25.4
4.21
0.942
0.803
0.588
5.75
0.115
38.5
10.8
3.70
1.06
As column A of Table 6-4 indicates, virtually no houses in the nation would be expected
to exceed a floor dust-lead loading standard of 400 ng/ft2,10.3 percent of the nation's homes
would be expected to have a window sill dust-lead loading exceeding 800 ng/ft2, and 10.3
percent of the nation's homes would be expected to exceed at least one of these two standards.
Notice that the percentage of homes exceeding any example dust standard is less than or equal to
the sum of these two percentages. Consider example option C, with the example floor dust
standard set at 100 ^ig/ft2 and the example window sill dust standard set at 500 ng/ft2. The
percentage of homes estimated to exceed the example floor standard of 100 (ig/ft2 is 4.04, and the
percentage of homes estimated to exceed the example window sill option is 12.5. The
percentage of homes estimated to exceed at least one of these example standards of 500 ng/ft2 is
13.9. This means that 12.5 + 4.04 - 13.9 = 2.64 percent of homes are estimated to exceed both
example dust standards.
6-18
-------
Continuing down the rows of column A of Table 6-4, 19.5 percent of the nation's homes
would be expected to exceed any of the example standards and paint intervention trigger levels
considered in this column:
floor dust-lead loading: 400 ug/ft2
window sill dust-lead loading: 800 ug/ft2
soil-lead concentration: 3,000 ug/g
paint (maintenance): Greater than 5 ft2 of damaged LBP
paint (abatement): Greater than 20 ft2 of damaged LBP.
As mentioned above, a total of 10.3 percent of the homes were projected to exceed at least one of
the example dust-lead loading standards. This means that 19.5 - 10.3 = 9.2 percent of the
nation's homes are projected to exceed a soil-lead concentration of 3,000 fig/'g or exceed 5 ft2 of
damaged LBP, but not exceed either of the two dust-lead loading standards.
The middle section of Table 6-4 presents estimates, based on the empirical model, of the
projected health effect and blood-lead concentration endpoints after implementation of the
various example options for the §403 standards. For example, if the proposed §403 rule consists
of the standards in column A, 0.442 percent of the nation's children (aged 1-2 years) would be
projected, using the empirical model, to have blood-lead concentration at or above 20 ug/dL, and
approximately 4.93 percent of children would be projected to have blood-lead concentration at or
above 10 ug/dL following promugation of §403. The percentage of children expected to have an
IQ score below 70 due to elevated blood-lead concentration is 0.11 percent. The next three rows
provide estimates of the percentage of children expected to have IQ decrements greater than or
equal to 1,2, and 3 IQ points. For the example option hi column A, the estimates under the
empirical model are 37 percent, 9.7 percent, and 3.1 percent, respectively. The next row gives
the estimated average IQ decrement associated with elevated blood-lead concentration.
Interventions triggered by the first set of standards would be projected to result in (an arithmetic)
average IQ decrement of 1.02.
The bottom section of Table 6-4 presents the same information as the middle section, but
with the projected health effects determined using the IEUBK model to predict blood-lead
concentrations instead of the empirical model. For convenience in making comparisons, the
baseline percentages displayed in Table 5-1 (estimated using blood-lead concentration data from
NHANES HI) are provided in the last column. Additional discussion of the results in Table 6-4
is provided in the next subsection.
6.3.1 Evaluation of Example Dust Standards
To examine the impact of various example options for the dust-lead loading standards on
childhood health effect and blood-lead concentration endpoints, Table 6-4 considers six example
combinations of floor and window sill dust-lead loadings standards. The last row of the top
section of Table 6-4 indicates that the number of homes that would be affected by any of the
selected example options ranges from about 20 percent to 53 percent, with the percentage
increasing as the example standards decrease.
6-19
-------
Examining the associated health effect and blood-lead concentration endpoints in the
middle and bottom sections of Table 6-4 reveals that the greatest reduction in health effects is
achieved between example options B and C for the EEUBK model and example options C and D
for the empirical model, with smaller reductions in health effects achieved between successive
lowering of the example standards. The reductions in health effect and blood-lead concentration
endpoints dimmish and the number of houses impacted increases as the example options decline.
This is most clearly evident for the percentage of children with blood-lead concentration at or
above 20 |ig/dL or 10 ug/dL, and the percentage of children that will have an IQ decrement
greater than 2 or 3. For example, based on the empirical model, 4.9 percent of the nation's
children would be anticipated to have blood-lead concentration at or above 10 |ig/dL under the
example option A (floor 400 ng/ft2; window sill 800 ug/ft2) while the IEUBK model predicts 3.0
percent. These percentages can be compared with the baseline estimate of 5.75 percent. Under
example option C (floor 100 |ig/ft2; window sill 500 ng/ft2), the projected number of children
with blood-lead levels at or above 10 ug/dL decreases to 4.8 percent (empirical model) and 2.3
percent (TEUBK model). For the lowest example dust standards considered in this analysis
(option F: floor 25 |ig/ft2; window sill 25 ug/ft2), the estimated percentage of children with
blood-lead concentration at or above 10 ug/dL is reduced to 4.6 percent (empirical model) and
1.7 percent (IEUBK model). However, significantly more homes would be affected by this pair
of example standards.
Figure 6-4 shows graphically the percentages of homes that would exceed the example
standards given in Table 6-4. For each of the six example sets of standards, Figure 6-4 plots the
percentage of homes that would exceed the floor dust standard, the percentage that would exceed
the example window sill dust standard, the percentage that would exceed any example dust
standard, and the percentage that would exceed any example standard with soil standards and
paint intervention triggers held fixed at the levels indicated hi Table 6-4.
According to Figure 6-4, the difference between the percentage of homes exceeding
example window sill dust standards and the percentage exceeding example floor dust standards is
smallest when the example standards are set at 100 and 500 ug/ft2 (example option C) and largest
for the lowest standards (example options E and F), while the difference between the percentage
of homes exceeding example window sill dust standards and the percentage exceeding either
example dust standard is small for all six example sets. The latter implies that most homes
exceeding the floor dust standard once defined also exceed the window sill dust standard, i.e.,
very few homes that exceed the example option for a floor dust standard did not exceed the
corresponding example option for the window sill standard (at least for the pairs of example
standards considered hi this analysis).
Figures 6-5a and 6-5b contain a total of seven graphs, one for each of the seven health
effect and blood-lead concentration endpoints presented hi Table 6-4. Each graph hi Figures
6-5a and 6-5b illustrates how values for a particular endpoint (as specified along the graph's
vertical axis) are affected by each example option for the dust standards. In addition, each graph
illustrates how the percentage of homes exceeding at least one example paint intervention
6-20
-------
E
o
o
o
TJ
O
O
o»
s
o
X
E
o
o
0>
o>
2
"c
o
a.
100-
90-
an.
O\J
70
60
50
40
30
20
10
0
oors:
Indov
% Exceeding Floor Dust Standard
«-o--o % Exceeding Window Sill Dust Standard
e-°-e % Exceeding Either Dust Standard
a-B-a % Exceeding Any Standard
e- ° ^
^^ f *
^ -^*
/^_ -
-^i5^ . ~~*
ABC
400 200 100
H Sills: 800 500 500
//^
- ~J*^'
^° ^^ '
^0*
__ <
__ ^"^
D E F
1 00 50 2J
200 1 00 21
Candidate §403 Standards for Dust
Figure 6-4. Percentage of Homes Exceeding Example Candidate Dust Standards:
Soil Standard and Paint Intervention Trigger Held Fixed at Levels Given
in Table 6-4.
trigger, or dust or soil standard represented in Table 6-4 (as specified along the graph's horizontal
axis) is affected by changes in the example dust standards. Each graph contains two curves: a
solid curve illustrating predictions based on the empirical model, and a dashed curve representing
predictions based on the IEUBK model. The solid curve is higher than the dashed curve in all
graphs, indicating that the empirical model predicts higher values for the endpoints than the
IEUBK model. Six letters are plotted on each curve, with each letter corresponding to one of the
six example options. The example set of standards associated with a particular letter is specified
at the top of Table 6-4 and in the horizontal axis label in Figure 6-4. The dashed horizontal
reference lines in Figure 6-4 indicates the baseline risk as determined from NHANES ffl.
The graphs in Figures 6-5a and 6-5b show the impact of example options for the dust
standards on the health effect and blood-lead concentration endpoints, and the number of homes
impacted by the example standards. Note the generally consistent shape of each of the curves in
these figures. A sharp decline in the curve indicates a large change in the health effect or
blood-lead concentration endpoint. accompanied bv a relatively small increase in the
number of homes requiring an intervention. A less steep decline indicates a large increment
6-21
-------
0.7
0.6
0.5
£ 0.4
o
CM
M
3 0.3
O_
0.2
0.0
CD °'12
M
10
0 10
0.08
g
° 0.06
V
o
0.04
0.02
0.00
A*
^ D g p
*
EMPIRICAL
IEUBK
10 20 30 40 50 60 70 80
Percentage of Homes Exceeding Any Standard
7W " D E ~ " F "
M«,.,=
EMPIRICAL
IEUBK
6
4
s
0
m
£
2
0
1.1
1.0
0.9
0.8
|0.7
b 0.6
a>
0
0 0.5
f 0.4
0.3
0.2
0.1
0.0
* 0 E F
A
***lj
EMPIRICAL
IEUBK
10 20 30 40 50 60 70 80
Percentage of Homes Exceeding Any Standard
^B n F F
E F
Uod.l:
EMPIRICAL
IEUBK
10 20 30 40 50 60 70
Percentage of Homes Exceeding Any Standard
BO
10 20 30 40 50 60 70
Percentage of Homes Exceeding Any Standard
80
Figure 6-5a. Projected Health Endpoints Based on Various Example Options for Dust Standards, Part 1; Soil Removal
3,000 /ig/g. Paint Maintenance 5 ft2. Paint Abatement 20 ft2. (Dashed reference line represents baseline
risk.)
-------
40
35
30
25
1 201
I 15
10
EMPIRICAL
IEUBK
-D E-
F
10 20 30 40 50 60 70
Percentage of Homes Exceeding Any Standard
80
12
10
8
I 6
g 4
-D-
10 20 30 40 50 60 70
Percentage of Homes Exceeding Any Standard
80
IS)
w
4.0
3.5
3.0
S" 2.5
« 2.0
5 1.5
g
1.0
0.5
0.0
EMPIRICAL
IEUBK
10 20 30 40 50 60 70
Percentage of Homes Exceeding Any Standard
80
Figure 6-5b. Projected Health Endpoints Based on Various Example Options for Dust Standards, Part 2; Soil Removal
3,000 i/glg. Paint Maintenance 5 ft2. Paint Abatement 20 ft2. (Dashed reference line represents baseline
risk.)
-------
in the number of homes requiring an intervention, accompanied by a small reduction in the
endpoint. In each case, the steepest drop occurs between example options A (floor 400
lie/ft2; window sill 800 ng/ft^ and C (floor 100 ng/ft2; window sill 500 ng/ft*). and then
gradually levels off as lower example standards are considered, despite the associated
impact on greater and greater numbers of homes. This pattern is consistent between the
empirical and IEUBK models, and across endpoints, with some endpoints reflecting the pattern
more drastically than others. However, the estimated reduction hi health effect and blood-lead
concentration endpoints associated with decreases in example standards is generally smaller
under the empirical model compared to the IEUBK model.
hi Figures 6-5a and 6-5b, the projected health effect and blood-lead concentration
endpoints as a result of implementing §403 with the various example standards can be compared
to the baseline (current estimated) values represented by the horizontal reference lines. For
example, each example dust standard results hi a substantial improvement relative to the baseline
for the percentage of children with blood-lead concentration at or above 20 u.g/dL or 10 |ag/dL,
and the percentage of children anticipated to have an IQ decrement greater than 2 or 3 resulting
from elevated blood-lead concentration, hi contrast, there is little reduction from baseline hi the
percentage of children predicted to have IQ below 70, hi the percentage of children expected to
have IQ decrement greater than 1, and hi the average IQ decrement, over the range of example
standards considered hi this analysis.
6.3.2 Evaluation of Example Options for the Soil Standard
Table 6-5 presents results for a range of example options (150 to 5,000 |ig/g) for the §403
soil standard, hi this table, the example floor dust-lead loading standard is set at 100 ng/ft2, the
example window sill dust-lead loading standard at 500 ng/ft2, paint maintenance at 5 ft2 of
damaged LBP, and paint abatement at 20 ft2 of damaged LBP. For each of these example options
(A-H), the top section indicates the percentage of homes that would exceed the example soil
standard and the percentage of homes that would exceed at least one of the example standards for
dust or soil, or the paint intervention triggers specified hi Table 6-5. The remaining rows are
analogous to those displayed in Table 6-4. Values of health effect and blood-lead concentration
endpoints are presented first for the empirical model, and then for the IEUBK model.
Table 6-5 predicts that the number of houses that would exceed at least one of the given
example standards ranges from 22 percent to 32 percent. Over the range of example soil
standards, the projected post-§403 percentage of children with blood-lead concentration at or
above 20 ug/dL ranges from 0.43 percent to 0.31 percent based on the empirical model, and from
0.12 to less than 0.001 percent based on the IEUBK model. The projected percentages of
children having blood-lead concentration at or above 10 ug/dL range from 4.9 to 4.0 percent
based on the empirical model, and from 2.5 to 0.2 percent based on the DEUBK model. Thus,
while the IEUBK model projects lower incidence of elevated blood-lead concentrations for each
example standard, both models project substantial reductions hi this incidence over the range of
example standards.
6-24
-------
Table 6-5. Characterization of Impact of Various Example Options for the Soil Standard:
Dust and Paint Standards fixed (100//g/ft2 for Floor Dust-Lead Loading, 500
//g/ft2 for Window Sill Dust-Lead Loading, 5 ft2 Damaged LBP for Paint
Maintenance, 20 ft2 Damaged LBP for Paint Abatement).
Example Options for Soil Lead Concentration Standard (pg/g)
EXAMPLE OPTION CODE
Soil Standard
Percentage of Homes
Exceeding Soil Standard
Percentage of Homes
Exceeding Any Dust
Soil, or Paint Standard
A
6000
0.215
21.5
B
3000
0.746
21.6
C
2000
2.49
21.8
D
1500
3.27
21.8
E
1000
5.82
22.3
F
500
11.8
25.3
Q
300
16.9
27.8
H
150
23.9
31.8
i
8
to
CO
Predicted Health Effect and Blood-Lead Concentration Endpolnts (Based on Empirical Model)
PbBi20//g/dL (%)
PbBilO/ifl/dL (%)
10X70 <%)
IQ decrement* 1 (%)
IQ decrement 2 2 <%)
IQ decrements (%1
Avg. IQ decrement
Predicted
PbB*20//g/dLf%)
PbB*10jUfl/dL[%>
IQ<70<%)
IQ decrements 1 (%)
IQ decrements 2 {%}
IQ decremented (%)
Avg. IQ decrement
0.433
4.87
0.111
36.7
9.56
3.05
1.01
0.427
4.84
0.111
36.6
9.50
3.02
1.01
0.406
4.70
0.110
36.3
9.30
2.93
1.00
Health Effect and Blood-Lead
0.119
2.49
0.102
31.1
5.80
1.38
0.895
0.0991
2.28
0.101
30.5
5.44
1.24
0.885
0.0539
1.66
0.0984
28.3
4.31
0.858
0.848
0.397
4.65
0.110
36.2
9.22
2.89
1.00
0.375
4.51
0.110
35.9
9.00
2.79
0.994
0.340
4.26
0.109
35.2
8.62
2.61
0.981
0.318
4.11
0.108
34.8
8.38
2.50
0.972
0.305
4.01
0.108
34.5
8.22
2.44
0.967
0.588
5.75
0.115
38.5
10.8
3.70
1.06
Concentration Endpoint* {Based on IEUBK Model)
0.0408
1.44
0.0975
27.3
3.88
0.725
0.834
0.0207
1.02
0.0957
25.1
2.99
0.479
0.802
0.00399
0.430
0.0928
20.2
1.58
0.174
0.741
0.00171
0.275
0.0918
18.0
1.12
0.103
0.715
0.000862
0.188
0.0911
16.1
0.839
0.0663
0.692
0.588
5.75
0.115
38.5
10.8
3.70
1.06
The percentage of children projected to have IQ scores below 70 is insensitive to changes
in the example soil standards based on the empirical model (staying at about 0.11 percent), and
ranges from 0.10 percent to 0.09 percent based on the IEUBK model.
Figure 6-6 shows the percentage of homes that would exceed each of the eight example
options for the soil standards. The bottom curve indicates the percentage of homes exceeding the
soil standard, and the top curve indicates the percentage exceeding any example standard, with
the example dust standard and paint intervention triggers held fixed at the levels indicated in the
caption of Table 6-5. Relatively small percentages of homes are predicted to exceed example
soil standards of 1,500 ug/g or greater.
The seven graphs in Figures 6-7a and 6-7b illustrate how values for a particular health
effect or blood-lead concentration endpoint (as specified along the graph's vertical axis) are
affected by the various example options for the soil standard. Each graph also illustrates how the
percentage of homes exceeding at least one example paint, dust, or soil standard represented in
6-25
-------
o
o
o
55
o
2
o
J
o»
c
TJ
u
X
UJ
n
100
90
80
70
60
50
40
30
20
10
0
% Exceeding Soil Standard
% Exceeding Any Standard
o-
.--*
Soil Rem.: 5000
B
3000
C
2000
0
1500
E
1000
F
500
G
300
H
150
Candidate §403 Standards for Soil Removal
Figure 6-6. Percentage of the Nation's Homes Expected to Exceed Various
Example Candidate Soil Standards: Dust Standard and Paint
Intervention Triggers Held Fixed at Levels Given in Table 6-5.
Table 6-5 (as specified along the graph's horizontal axis) is affected by changes in the example
soil standard. Each graph contains two curves: a solid curve illustrating predictions based on the
empirical model, and a dashed curve representing predictions based on the IEUBK model. As
seen in Figures 6-5a and 6-5b, the empirical model predicts higher values for the endpoints than
the IEUBK model. Each example option is represented in the plots by its letter code (A through
H) specified at the top of Table 6-5 and in the horizontal axis label in Figure 6-6.
Results hi Figures 6-7a and 6-7b indicate that the IEUBK model predicts significant
reduction in percentage of children experiencing IQ decrements greater than or equal to 1,2, and
3, and blood-lead concentrations across the range of example options A (5,000 ug/g) to F (500
ug/g). The empirical model predicts only small differences. This is due to differences between
the soil-lead to blood-lead relationship embodied by the IEUBK model as compared to the
empirical model. Specifically, at soil-lead concentrations greater than 1,500 (ig/g the IEUBK
model generally predicts much higher blood-lead concentrations than the empirical model.
Conversely, at the assumed post-intervention soil-lead concentration (150 ug/g) the IEUBK
model generally predicts lower blood-lead concentrations. Thus, as the option for soil standard
6-26
-------
IS)
-J
0.7
0.6
0.5
£ 0.4
0
< °'
0.
0.
0.
0.
1
0
9
8
7
6
5
4
3
2
1
0
6
H^
^^^^*H
A
B
6
E
Modil:
p
EMPIRICAL -G LJ
IEUBK > - rl
1 1 1 1 1 1 1 1
10 20 30 40 50 60 70 80
Percentage of Homes Exceeding Any Standard
*^O-H
|
1
FQ
^^~ "H
Uod.l:
EMPIRICAL
IEUBK
10 20 30 40 50 60 70
Percentage of Homes Exceeding Any Standard
80
10 20 30 40 50 60 70
Percentage of Homes Exceeding Any Standard
80
Figure 6-7a. Projected Health Endpoints Based on Various Example Options for the Soil Standard, Part 1; Floor Dust
100/ig/ft2, Window Sill Dust 500//g/ft2, Paint Maintenance 5 ft2. Paint Abatement 20 ft2. (Dashed reference
line represents baseline risk.)
-------
at
N>
oo
40
35
30
*?
~ 25
| 20
1 15
g
10
5
0
4.0
3.5
3.0
f ».
| 2.0-
o
S 1.5
g
1.0
0.5
0.0
£
^»^-G H
£
E
''f
-H
Mod.l:
EMPIRICAL
IEUBK
10 20 30 40 50 60 70 80
Percentage of Homes Exceeding Any Standard
ft
%=LH
A
B
B
Modi!: E
EMPIRICAL ''C f^ .
IEUBK G-H
12
10
g 8
cs
J^l
I '
0
*
0 4
0-
A
^^G
ft
6
E
"G H
EMPIRICAL " "
IEUBK
10 20 30 40 50 60 70 80
Percentage of Homes Exceeding Any Standard
10 20 30 40 50 60 70
Percentage of Homes Exceeding Any Standard
80
Rgure 6-7b. Projected Health Endpoints Based on Various Options for the Soil Standard, Part 2; Floor Dust 100 /ig/ft2.
Window Sill Dust 500 //g/ft2. Paint Maintenance 5 ft2. Paint Abatement 20 ft2. (Dashed reference line
represents baseline risk.)
-------
decreases (A-F) and soil-lead concentrations are assumed to be reduced to post-intervention
concentrations at more homes, the impact predicted by the IEUBK model is much greater than
that predicted by the empirical model. Results presented for percentage of children with IQ less
than 70 due to lead exposure and blood-lead concentrations greater than or equal to 20 ng/dL are
less sensitive to the model differences discussed above.
Figures 6-7a and 6-7b also indicate that there is very little change in the percentage of
homes exceeding any standard between example options A through E. This is indicated by the
small horizontal displacement between these letters. Conversely, for example options E through
H larger increases in the percentage of homes exceeding any standard are indicated.
In Figures 6-7a and 6-7b, the projected post-§403 health effect and blood-lead
concentration endpoints under the various example options can be compared to the baseline
(current estimated) endpoints using the horizontal reference lines. Each example option for the
soil standard results in a large decrease relative to the baseline for the percentage of children with
blood-lead concentration at or above 20 |ig/dL or 10 ^ig/dL, and the percentage of children
anticipated to have an IQ decrement greater than 2 or 3 resulting from elevated blood-lead
concentration. In contrast, there is little reduction from baseline in the percentage of children
predicted to have IQ below 70 or in the percentage of children expected to have IQ decrement
greater than 1 over the range of example standards considered.
There is a clear benefit projected for even the largest example soil-lead concentration
standard, and there is additional benefit predicted for the lower example standards. There are
gains to be made in health effect and blood-lead concentration endpoints for example soil
standards as low as 500 (ig/g.
6.3.3 Evaluation of Example Options for the Trigger Levels of Paint Intervention
Table 6-6 presents results for a range of paint intervention trigger levels. Example
options considered for requiring paint maintenance range from 0 to 10 ft2 of damaged LBP, and
example options for requiring paint abatement range from 5 to 100 ft2 of damaged LBP. In Table
6-6, the example floor dust-lead loading standard is set at 100 ng/ft2, the example window sill
dust-lead loading standard at 500 |ig/ft2, and the example soil standard at 3,000 ng/g. For each of
these options (A-E), the top section of Table 6-6 indicates the percentages of homes that would
exceed the trigger levels for interior and exterior paint maintenance, the percentages that would
exceed the trigger levels for interior and exterior paint abatement, and the percentage of homes
that would exceed at least one of the example standards for dust or soil, or the trigger levels for
paint specified in Table 6-6. The remaining rows are analogous to those in Tables 6-4 and 6-5.
Estimated values of the health effect and blood-lead concentration endpoints are first presented
for the empirical model and then for the IEUBK model.
6-29
-------
Table 6-6. Characterization of Impact of Various Options for Paint Intervention Triggers:
Example Dust and Soil Standards Fixed (100/yg/ft2 for Dust-Lead Loading, 500
//g/ft2 for Window Sill Dust-Lead Loading, 3,000 //g/g for Soil Removal).
Example Options for Point Standard (ft1 damaged LBP)
EXAMPLE OPTION CODE
Paint Maintenance Trigger
(Interior or Exterior)
Paint Abatement Trigger
(interior or Exterior)
_ in r .IT i
IQ decrement 2 2 {%)
IQ decrement;^ {%)
Avg. EQ decrement
0.437
4.90
0.111
36.8
9.59
3.07
1.01
0.428
4.85
0.111
36.7
9.52
3.03
1.01
0.426
4.83
0.111
36.6
9.50
3.02
1.01
0.425
4.82
0.111
36.6
9.49
3.02
1.01
0.423
4.81
0.111
36.6
9.46
3.00
1.01
0.340
4.26
0,109
35.2
8.62
2.61
0.981
Predicted Health Effect and Blood-Lead Concentration Endpoints (Based on IEUBK Modef>
PbBiZO <%)
PbBi10<%)
IQ<70(%)
IQ decrements 1 {%}
IQ decrement:^ {%}
IQ decremented (%)
Avg. fQ decrement
0.162
2.92
0.103
32.2
6.52
1.67
0.917
0.0991
2.28
0.101
30.5
5.44
1.24
0.885
0.0973
2.26
0.101
30.4
5.40
1.23
0.883
0.0966
2.24
0.101
30.3
5.37
1.22
0.882
0.0947
2.22
0.101
30.2
5.32
1.21
0.880
0.588
5.75
0.115
38.5
10.8
3,70
1.06
Figures 6-8 and 6-9 display the percentage of homes that would exceed the different
trigger levels of paint intervention for interior and exterior paint, respectively. In each of these
figures, the difference between the percentage of homes exceeding either intervention trigger
level and the percent exceeding the paint abatement trigger represents the percentage of homes
that have enough damaged LBP to exceed the paint maintenance trigger but not the abatement
trigger.
6-30
-------
g so-
_o
S «>
2
o
o
o> 30-
S
o
X
« 20-
o
o 10-
1
o
O. 0-
Matnten
Abatem
«-^e % Exceeding Interior Paint Abatement Trigger
->»% Exceeding Either Interior Paint Trigger
e-o-e % Exceeding Any Standard
P-
/
/ .--*
A
once: 10
ent: 100
0 e
~~~~
-o-
HI . *
B C
5 2
40 20
e E
^Z^ "
D E
1 0
10 5
Candidate §403 Standards for Paint
Figure 6-8. Percentage of Homes Exceeding Candidate Interior Paint Intervention Triggers:
Dust and Soil Example Standards Held Fixed at Levels Given in Table 6-6.
50 i
% Exceeding Exterior Paint Abatement Trigger
-o--o % Exceeding Either Exterior Paint Trigger
Exceeding Any Standard
Maintenance:
Abatement:
Candidate §403 Standards for Paint
Figure 6-9. Percentage of Homes Exceeding Candidate Exterior Paint Intervention
Triggers: Dust and Soil Example Standards Held Fixed at Levels Given in
Table 6-6.
6-31
-------
These figures show that relatively few homes exceeded even the lowest example option
considered for the paint intervention triggers. Therefore, there is very little change hi the
percentage of homes that would exceed any of the combinations of trigger levels for paint from
the lowest to the highest square footages of deteriorated LBP. However, this analysis is based on
the limited data available on deteriorated lead-based paint hi the HUD National Survey.
The seven graphs hi Figures 6-10a and 6-1 Ob illustrate how values for a particular health
effect or blood-lead concentration endpoint (as specified along the graph's vertical axis) are
affected by the various example options for the pair of paint trigger levels. Each graph also
illustrates how the percentage of homes exceeding at least one paint intervention trigger level or
dust or soil example standard represented in Table 6-6 (as specified along the graph's horizontal
axis) is affected by changes hi the paint intervention trigger levels. Each graph contains two
curves: a solid curve illustrating predictions based on the empirical model, and a dashed curve
representing predictions based on the IEUBK model. As was seen hi previous figures, the
empirical model predicts higher values for the endpoints than does the IEUBK model. Each
example option for the pair of paint intervention trigger levels is represented hi the plots by its
letter code (A through E) specified at the top of Table 6-6 and hi the horizontal axis label in
Figures' 6-8 and 6-9 (a given set of paint intervention triggers is assumed to hold for both the
interior and exterior).
With the exception of example option A, the graphs hi Figures 6-10a and 6-1 Ob show
very little variation across the various paint intervention trigger levels. The percentage of homes
exceeding the example options for §403 standards ranged only from 20 percent at the highest
paint intervention trigger considered (option A) to 23 percent at the lowest paint intervention
trigger (option E). The values of health effect and blood-lead concentration endpoints, although
less than their baseline pre-§403 respective values, are very similar for paint intervention trigger
options B through E. For instance, the percentage of children with a blood-lead concentration at
or above 20 |ig/dL ranges from 0.10 percent at option B to 0.99 percent at option E, based on the
IEUBK predicted blood-lead concentrations. Example option A for the paint intervention trigger
(maintenance 10 ft2; abatement 100 ft2) provided noticeably higher health effect and blood-lead
concentration endpoints than the next example option (option B: maintenance 5 ft2; abatement
40 ft2). The limited ranges for the predicted health effect and blood-lead concentration endpoints
and for the percentages of homes exceeding the intervention triggers, particular among options B
through E might be due to the following:
1. It is very difficult to study the effects of an individual environmental medium on
childhood blood-lead concentration. As discussed hi Section 3.1, deteriorated or
damaged LBP is a source of lead contamination for both soil and household dust.
Thus, most of the homes that exceed a paint intervention trigger also exceeded either
the dust or soil example standards, and therefore, much of the post-§403 benefits
expected to result from the paint intervention triggers overlap with risk reductions
expected to result from the dust and soil example standards.
6-32
-------
0.7
0.6
0.5
£ 0.4
s
*
2 0.3
Q.
0.2
0.0
0)
<> °'12
u
0.10
0.08
g
° 0.06
V
0.04
0.02
0.00
Akti
*V.
A
'aa
Modal: ^*
EMPIRICAL
IEUBX
10 20 30 40 50 60 70 80
Percentage of Homes Exceeding Any Standard
OK
*&
Modd:
EMPIRICAL
IEUBK
6
5
4
^^
o
« 3
2
Q-
2
1
0
1.1
1.0
0.9
0.8
1 °'7
i °-6
o 0.5
| 0.4
0.3
0.2
0.1
0.0
*1
A
'£
UcxJ.I:
EMPIRICAL
IEUBK
10 20 30 40 50 60 70 80
Percentage of Homes Exceeding Any Standard
tm
AA
Mod.l:
EMPIRICAL
IEUBK
10 20 30 40 50 60 70
Percentage of Homes Exceeding Any Standard
BO
10 20 30 40 50 60 70
Percentage of Homes Exceeding Any Standard
80
Figure 6-10a. Projected Health Endpoints Based on Various Options for Paint Intervention Triggers Part 1; Floor Dust
100 //g/ft2. Window Sill Dust 500 //g/ft2. Soil Removal 3,000 //g/g. (Dashed reference line represents
baseline risk.)
-------
40
35
30
g
25
M
°c
| 20
o
u
S 15
S
10
5
0
d>
TJ 4.0
U
3.5
3.0
^ 2.5
"E
| 2.0
u
S 1.5
2
1.0
0.5
0.0
/B
/L
«
Modtl:
EMPIRICAL
IEUBK
10 20 30 40 50 60 70 80
Percentage of Homes Exceeding Any Standard
*e
A
',
to
Modtl:
EMPIRICAL
IEUBK
12
10
S 8
04
*
i 6
E
5
1
0 4
2
0-
&i
A
*,
*^^h*
^2
Modil:
EMPIRICAL
IEUBK
10 20 30 40 50 60 70 80
Percentage of Homes Exceeding Any Standard
10 20 30 40 50 60 70
Percentage of Homes Exceeding Any Standard
80
Figure 6-1 Ob. Projected Health Endpoints Based on Various Options for Paint Intervention Triggers, Part 2; Floor Dust
100 jig/ft2. Window Sill Dust 500 //g/ft2, Soil Removal 3,000 //g/g. (Dashed reference line represents
baseline risk.)
-------
2. Not all surfaces were examined for the presence of damaged or deteriorated LBP in
the HUD National Survey. In general, only two interior rooms and one exterior
surface were examined.
3. The tools available for assessing the impact of damaged lead-based paint are limited.
Both the empirical and IEUBK models for predicting blood-lead concentrations
based on environmental-lead levels are limited hi then* usage of paint-lead
measurements. IEUBK model-predicted blood-lead concentrations are adjusted for
the contribution of paint ingested due to pica using the procedures, developed only
for the purposes of this analysis, presented in Section 4.1.3. The empirical model
was developed from data collected in the Rochester Lead-in-Dust Study, which does
not express the amount of damaged LBP in the same manner as the HUD National
Survey. Pica for paint also plays a role hi this model, and the estimate of the
prevalence of pica for paint used hi this risk analysis may be somewhat inaccurate.
6.3.4 Evaluation of the Effects of Varying Example Standard Options for All Media
Analyses summarized in Tables 6-4 through 6-6 permit an assessment of the impact on
the nation's housing and health effects of children for various example standard or trigger
options for an individual environmental medium. A range of example standards for one
environmental medium is considered while the example standards for the other media are held
fixed at a specified level. However, those results do not show the effect of varying the example
standards simultaneously for dust, soil, and paint. Table 6-7 presents results when the example
standards for all media are varied over the ranges previously considered in this chapter. Table
6-7 is structured similarly to Tables 6-4 through 6-6. Each column represents a unique
combination of example standards displayed at the very top hi the shaded rows. For example,
column A in Table 6-7 represents an option in which the candidate example standards are 400
Hg/ft2 for floor dust-lead loading, 800 u.g/ft2 for window sill dust-lead loading, 5,000 |ig/g for
soil-lead concentration, and with 10 ft2 of damaged lead-based paint prompting paint
maintenance, and 100 ft2 prompting paint abatement. Below the example options for standards
are presented the estimated percentage of homes that exceed one or more of the example
standards. The first row in this section provides the estimated percentage of homes that would
exceed the example floor dust standard. Analogous information is provided hi the next seven
rows for the window sill dust standard, the soil standard, and the interior and exterior paint
maintenance triggers, and ulterior and exterior paint abatement triggers. Finally, the last row of
this section provides the estimated percentage of the nation's homes that would exceed one or
more of the example standards or intervention triggers.
The third and fourth sections of Table 6-7 provide the estimated post-§403 health effect
and blood-lead concentration endpoints, based on the empirical model and the IEUBK model,
respectively. The rows hi these two sections are analogous to those in Tables 6-4 through 6-6.
6-35
-------
Table 6-7. Characterization of Impact of Various Sets of Candidate Example
Dust and Soil, and Paint Intervention Triggers.
Exempt* Option* for Standards/Trigger*
EXAMPLE OPTION CODE
Roor Dust-Lead Loading
(TO/*1)
Window Sill Dim-Lead Loading
(//g/ft»)
Soil-Lead Concentration (//g/g)
Paint Maintenance (interior and
exterior) (ft1 damaged LBP)
Paint Abatement (Interior and
exterior) (ft* damaged LBP)
A
400
800
5000
10
100
B
200
500
3000
10
40
C
100
500
2000
5
20
D
100
200
1500
2
10
E
50
100
1000
1
10
F
25
25
500
0
5
Current
Interim
Guidance
I
100
500
5000
2
10
Baseline
Percentage of Home* Exceeding Example St*nd*rd*/TrJflgeri
Floor Dust
Window Sill Dtwt
Soil
Interior Point Maintenance
Exterior Paint Maintenance
Interior Paint Abatement
Exterior Paint Abatement
Percentage of Homes
Exceeding Any Standard
0.00
10.3
0.215
2.80
3.84
0.453
3.03
17.5
0.694
12.5
0.746
2.27
2.41
0.980
4.46
19.5
4.04
12.5
2.49
2.92
3.49
2.43
5.77
21.8
4.04
24.3
3.27
2.22
3.09
3.25
6.87
31.2
8.28
32.5
5.82
2.75
3.22
3.25
6.87
38.4
13.8
48.1
11.8
1.08
1.15
5.35
9.26
53.7
4.04
12.5
0.215
2.22
3.09
3.25
6.87
22.0
Predicted Health Effect and Blood-Lead Concentration Ertdpoint* (Based on Empirical Modal}
PbBi20 <%)
PbBi10(%)
IQ<70<%)
IQ decrements; 1 (%)
IQ decrement;^ (%)
IQ decrement;^ (%)
Avg. IQ decrement
0.458
5.03
0.112
37.1
9.79
3.16
1.02
0.439
4.91
0.111
36.8
9.62
3.08
1.02
0.406
4.70
0.110
36.3
9.30
2.93
1.00
0.381
4.53
0.110
35.9
9.04
2.81
0.995
0.350
4.33
0.109
35.4
8.71
2.66
0.984
0.317
4.09
0.108
34.7
8.34
2.49
0.971
0.431
4.86
0.111
36.7
9.54
3.04
1.01
O.S88
5.75
0.115
38.6
10.8
3.70
1,06
Predicted Health Effect and Blood-Lead Concentration Endpolnt* (Based on IEUBK Modefl
PbBiZO <%)
Pbfte 10 :<%).
IQ<70<%)
IQ decrements 1 {%>
IQ decrement:^ (%)
IQ decrements (%)
Avg. IQ decrement
0.290
3.92
0.107
34.5
8.09
2.37
0.964
0.235
3.51
0.106
33.5
7.45
2.08
0.943
0.0539
1.66
0.0984
28.3
4.31
0.858
0.848
0.0409
1.39
0.0971
26.2
3.71
0.702
0.816
0.0164
0.841
0.0945
22.5
2.52
0.392
0.764
0.00198
0.250
0.0909
15.1
0.978
0.0976
0.666
0.117
2.47
0.102
31.0
5.76
1.37
0.894
0.588
5.75
0.115
38.5
10.8.
3.70
1,06
6-36
-------
A total of seven example options for the standards are assessed in Table 6-7.
Environmental-lead levels are highest for example option A (floor: 400 |ig/ft2; window sill: 800
ug/ft2; soil: 5,000 ug/g; paint maintenance: 10 ft2 damaged LBP; paint abatement: 100 ft2
damaged LBP) and are lowest for example option F (floor: 25 (ig/ft2; window sill: 25 ng/ft2;
soil: 500 ug/g; paint maintenance: 0 ft2 damaged LBP; paint abatement: 5 ft2 damaged LBP). In
addition, example option I corresponds to the interim standards presented in the interim rule
(floor: 100 (ig/ft2; window sill: 500 (ig/ft2; soil: 5,000 |ig/g; paint maintenance: 2 ft2 damaged
LBP; paint abatement: 10 ft2 damaged LBP). For comparison purposes, the baseline values for
the health and blood-lead concentration endpoints are displayed in the last column of the table.
The last row of the second section indicates that the percentage of homes affected by the
various example sets of standards ranges from 17.5 percent to 53.7 percent. This is a wider range
than was observed for any of the individual environmental medium. This is because the example
options considered in these tables represent a broader range of example standards than what was
considered in the analyses illustrating the effect of varying the standard for a single medium.
Over this range of example standards, the percentage of children expected to have blood-
lead concentration at or above 20 |ig/dL ranged from 0.46 to 0.32 percent based on the empirical
model and from 0.29 to 0.002 percent based on the IEUBK model. The percentage of children
with blood-lead concentration at or above 10 ng/dL ranged from 5.0 to 4.1 percent based on the
empirical model and from 3.9 to 0.3 percent based on the IEUBK model. The percentage of
children expected to have an IQ below 70 as a result of lead exposure ranged from 0.112 to 0.108
percent based on the empirical model, and from 0.107 to 0.091 percent based on the IEUBK
model.
The seven graphs in Figures 6-1 la and 6-1 Ib illustrate how values for a particular health
effect or blood-lead concentration endpoint (as specified along the graph's vertical axis) are
affected by the example options in Table 6-7. Each graph also illustrates how the percentage of
homes exceeding at least one example standard (as specified along the graph's horizontal axis)
changes among the different sets of example standards. Each graph contains two curves: a solid
curve illustrating predictions based on the empirical model, and a dashed curve representing
predictions based on the IEUBK model. As was seen in previous figures, the empirical model
predicts higher values for the endpoints than does the IEUBK model. Each example set of
standards is represented by its letter code (A through F) specified at the top of Table 6-7.
In Figures 6-1 la and 6-1 Ib, the incremental reduction in the estimated health effect or
blood-lead concentration endpoint for each unit change in the number of homes affected is
represented by the slope of the line connecting any two plotted points. For each graph, the slope
is steepest between example options A and C. This property was also present in the graphs
(Figures 5a, 5b, 7a, 7b, lOa, and lOb) illustrating the effects of changes in example standards for
the individual environmental medium.
6-37
-------
<)
00
o
0.7
0.6
0.5
0.4
2 0.3
Q.
0.2
0.1
0.0
£
o
r*
v
O
0.12
0.10
0.08
0.06
0.04
0.02
0.00
D..
10 20 30 40 SO 60 70
Percentage of Homes Exceeding Any Standard
80
-06-
EMPIRICAL
IEUBK
10 20 30 40 SO 60 70
Percentage of Homes Exceeding Any Standard
80
CD
.O
O.
EMPIRICAL
IEUBK
10 20 30 40 SO 60 70
Percentage of Homes Exceeding Any Standard
80
1.1
1.0
0.9
0.8
I
| °'6
o
g 0.5
d»
5 0.4
0.3
0.2
0.1
0.0
EMPIRICAL
IEUBK
10 20 30 40 50 60 70
Percentage of Homes Exceeding Any Standard
80
Figure 6-11 a. Projected Health and Blood-Lead Concentration Endpoints Based on Various Example Sets of Options for
Dust and Soil Standards, and Paint Intervention Triggers, Part 1. (Dashed reference line represents baseline
risk.)
-------
9>
u
(O
40
35
30
~ 25
n
| 20
u
5 15
g
10
5'
0
EMPIRICAL
ItUBK
10 20 30 40 50 60 70
Percentage of Homes Exceeding Any Standard
80
4.0
3.5
3.0
2.0
u
5 1.5
O
1.0
0.5
0.0
- EMPIRICAL
---- IEUBK
10 20 30 40 50 60 70
Percentage of Homes Exceeding Any Standard
80
12
10
8
2 4
I
«
C
EMPIRICAL
IEU8K
10 20 30 40 50 60 70
Percentage ot Homes Exceeding Any Standard
80
Figure 6-11b. Projected Health Endpoints Based on Various Example Sets of Options for Dust and Soil Standards, and Paint
Intervention Triggers, Part 2. (Dashed reference line represents baseline risk.)
-------
There is, again, a generally consistent shape to each of the curves in Figures 6-1 la and
6-1 Ib. In each case, the steepest drop occurs between example options A and C. This pattern is
consistent between the empirical and IEUBK models; however, incremental changes predicted by
the empirical model are generally less than those predicted by the IEUBK model. While example
option C is estimated to affect about the same number of homes as the current interim guidance
(21.8 compared to 22.0 percent), the estimated health effect and blood-lead concentration
endpoints for the interim standards are generally higher. However, the actual difference hi the
endpoints between the two sets of example standards may be inconsequential relative to the
uncertainty in the estimated endpoints.
As also observed when considering each medium individually, an option that establishes
even a relatively high example standard for all environmental media results hi a substantial
improvement relative to the baseline for the percentage of children at or above 20 ug/dL or
10 |ig/dL, and the percentage of children anticipated to have an IQ decrement greater than 2 or 3
resulting from elevated blood-lead concentration. However, even varying the example set of
standards encompassing all environmental media results hi little change in the percentage of
children predicted to have an IQ below 70 due to elevated blood-lead concentration or hi the
percentage of children expected to have an IQ decrement greater than 1 due to elevated blood-
lead concentration.
6.3.5 Risk Reduction Details for an Illustrative Set of Standards
This section provides a more detailed characterization of projected health effect and
blood-lead concentration endpoints associated with a particular illustrative set of dust and soil
standards, and paint intervention triggers. The illustrative standards considered are 100 ug/ft2for
floor dust-lead loading. 500 [ig/ft2 for window sill dust lead loading. 2.000 |ig/g for soil-lead
concentration removal. 5 ft2 damaged LBP for paint maintenance, and 20 ft2 damaged LBP for
paint abatement (i.e., option C of Table 6-7).
Under these illustrative standards, Figure 6-12 displays the projected post-§403
distribution of blood-lead concentrations hi children aged 1-2 years based on the empirical model
and the IEUBK model in both histogram and cumulative distribution function (cdf) format. The
pre-§403 (baseline) distribution is also presented in Figure 6-12. The histogram indicates the
general shape of the distribution of blood-lead concentrations, while the cdf provides the
probability that a child has a blood-lead concentration below any specified value. The cdf
enables the reader to easily estimate the percentage of children having blood-lead concentrations
within any particular interval of concentrations.
Qualitatively, the distribution associated with the lEUBK-predicted, post-§403 blood-lead
concentrations appears to the left of the corresponding distribution based on the empirical model,
which does not appear to be substantially different than the baseline (pre-§403) distribution. The
IEUBK model-predicted distribution of blood-lead concentrations indicates that the reduction in
the number of children with elevated blood-lead concentration under the illustrative set of
standards is more substantial than that based on the empirical model.
6-40
-------
30 H
25
20
o 15H
10-
0-
Blood-L«od Distribution
NHANES III Pr«-lnl«rv«nllon
HUD/IEUBK Post-Intervention
HUD/EMPIRICAL Poit-lnt.rvintlon
-iiiiiiiiir
12 13 14 15 16 17 18 19 20 21
-iiiiiiiiir
23 24 ' 25 26 27 28 29 30 31 32
Blood-Lead Concentration (/ug/dL)
100-
80-
60-
40
20-
c
o
e
a.
-------
Table 6-8 compares the baseline distribution of blood-lead concentrations and health
effect endpoints to the post-§403 distribution based on the empirical and IEUBK models for the
illustrative set of standards considered in this section: 100 ug/ft2 for floor dust-lead loading, 500
ug/ft2 for window sill dust-lead loading, 2,000 ug/g for soil-lead concentration, 5 ft2 damaged
LBP for paint maintenance, and 20 ft2 damaged LBP for paint abatement. The top half of Table
6-8 characterizes the distribution of children's blood-lead concentrations. Estimated numbers
and percentages of children with blood-lead concentration in various intervals are provided. The
bottom half of Table 6-8 estimates various health endpoints under the baseline and post-§403
projections for this example set of standards.
Table 6-8. Estimated Distribution of Health Effect and Blood-Lead Concentration
Endpoints Prior to and After the Proposed §403 Rule for an Illustrative Set of
Standards.1
PbB U/g/dL)
Total
0 <: PbB <1
1 i PbB <3
3s PbB <5
5* PbB <10
10 * PbB < 15
15 * PbB <20
20 * PbB < 25
PbB 225
Pre-§403
ft Children2
7.960
477
3,310
2,080
1,640
325
85.9
27.9
18.9
Percent
100
5.99
41.6
26.1
20.6
4.08
1.08
0.350
0.238
Post- §403
(Empirical Model)
# Children2
7.960
475
3,460
2,110
1,550
275
66.7
20.1
12.2
Percent
100
5.96
43.4
26.5
19.5
3.46
0.838
0.252
0.154
Post-1403
(IEUBK Model)
ft Children2
7.960
385
4,060
2,230
1,150
112
16.3
3.19
1.10
Percent
100
4.83
51.0
28.0
14.5
1.41
0.205
0.0401
0.0138
Inferred Health Effects
IQ < 70
IQ decrement 2 1
IQ decrement z 2
IQ decrement * 3
Average IQ
decrement
Houses Affected
9.13
3,060
863
294
0.115
38.5
10.8
3.70
1.06
# Houses
0
Percent
0
8.79
2,890
741
233
0.110
36.3
9.30
2.93
1.00
# Houses
21,600
Percent
21.8
7.84
2,250
343
68.3
0.0984
28.3
4.31
0.858
0.848
# Houses
21,600
Percent
21.8
1 100 yt/g/ft2 for floor dust lead loading, 500 /yg/ft2 for window sill dust-lead loading, 2,000 //g/g for soil-lead
concentration, 5 ft2 damaged LBP for paint repair, and 20 ft2 damaged LBP for paint abatement.
2 Numbers of children aged 1-2 years in thousands.
6-42
-------
6.4 SENSITIVITY AND UNCERTAINTY ANALYSES FOR RISK MANAGEMENT
ANALYSES
There are numerous procedures and assumptions discussed and presented that contribute
to the final results in this chapter. Sensitivity analyses address the extent to which variations in
key assumptions and approaches affect the estimated outcomes, thereby contributing to overall
uncertainty hi the results. As it was not feasible to consider variations in all aspects of the
analysis, the sensitivity analysis considered approaches and assumptions which had the potential
for producing the largest expected deviation. The alternative approaches considered hi the
sensitivity analysis and the comparison of their findings with the final results had to be
manageable within the context of the sensitivity analysis. Table 6-9 summarizes seven factors
addressed by the sensitivity analysis for risk management analyses where alternative
approaches) were considered; these alternative approaches are included in Table 6-9. Sections
6.4.1 through 6.4.8 present the sensitivity analyses under each of these factors.
An eighth factor considered hi the sensitivity analysis was the method for determining
post-intervention dust-lead concentrations (Section 6.1.3). However, instead of presenting
results under one or more alternative assumptions (as was done with the seven factors in Table
6-9), graphs and tables were prepared that illustrate how results calculated under this method
compare to those from published studies. Section 6.4.5 presents these findings.
There is also uncertainty hi the estimated post-§403 health effect and blood-lead
concentration endpoints due to the variability in the data used to obtain these estimates. Standard
errors associated with post-intervention estimates of the health effect and blood-lead
concentration endpoints are presented hi Section 6.4.9 for three sets of example options for the
standards.
6.4.1 Uncertainty in Converting Dust-Lead Loadings for Comparison to Standards
Because the §403 dust-lead loading standards will be defined in terms of lead loadings for
dust samples collected with wipe collection techniques, and because dust samples in the HUD
National Survey were collected using a Blue Nozzle vacuum, it was necessary to convert the
HUD National Survey dust-lead loadings (for both floors and window sills) to wipe dust-lead
loadings in the risk management analysis. Different formulas were used (Section 4.3; Table 6-9)
to predict a wipe dust-lead loading from a Blue Nozzle vacuum dust-lead loading, depending on
the age of the house and whether a floor or window sill was sampled. These formulas assume
that the expected value of the log-transformed wipe dust-lead loading (log(Wipe)) given a Blue
Nozzle vacuum dust-lead loading of "Vac," takes the form
a + p*log(Vac).
6-43
-------
Table 6-9. Procedures for Which Alternative Assumptions Were Considered in the
Sensitivity Analysis Addressing Risk Management.
Procedure
Approach Taken in the
Risk Management Analyses
Aftemattve(s) Considered in the
Sensitivity Analysis
Convert Blue Nozzle vacuum
dust-lead loadings reported in
the National Survey to wipe
dust-lead loadings, so that the
area-weighted geometric
mean for a housing unit can
be compared to example dust-
lead loading standards
As indicated in Section 4.3, convert
each Blue Nozzle vacuum dust-lead
loading ("Vac") to a wipe dust-lead
loading ("Wipe") using the following
formulas:
Floors:
Pre-1940: Wipe = 5.66(Vac)0809
1940-1959: Wipe = 4.78(Vac)080°
1960-1979: Wipe = 4.03(Vac)0707
Window Sills:
Wipe = 2.95'(Vac)118
Alt. #1 (low estimate): Assign the lower 90%
confidence bound on the estimated wipe dust-
lead loading obtained from the adjacent
formulas to each sample result.
Alt. #2 (high estimate): Assign the upper 90%
confidence bound on the estimated wipe dust-
lead loading obtained from the adjacent
formulas to each sample result.
(Section 6.4.1)
Convert the specified post-
intervention wipe dust-lead
loadings of 40 //g/ft2 for
floors and 100 //g/ft2 for
window sills to Blue Nozzle
dust-lead loadings for input to
the empirical model
As indicated in Section 4.3, convert the
wipe dust-lead loading to a Blue Nozzle
vacuum dust-lead loading ("BN") as
follows:
Floors:
BN = 0.185*(40)0931
Window Sills:
BN = 0.955*(100)0683
5.7//g/ft2
14.0pg/ft2
Alt. #1 (low estimate): Assign the lower 90%
confidence bound on the estimated Blue
Nozzle vacuum dust-lead loading obtained from
the adjacent formulas.
Alt. #2 (high estimate): Assign the upper 90%
confidence bound on the estimated Blue
Nozzle vacuum dust-lead loading obtained from
the adjacent formulas.
(Section 6.4.2)
Determine a post-§403 blood-
lead concentration distribution
under the empirical model as
a function of post-intervention
dust-lead loadings
Consider post-intervention dust-lead
loadings of 40 ^g/ft2 for floors and 100
pg/ft2 for window sills
Consider the following alternative post-
intervention dust-lead loadings:
- 20 fjg/H* for floors and 50 //g/ft* for
window sills
-- 100//g/ft2 for floors and 250//g/ft2 for
window sills
(Section 6.4.3)
Determine a method for
characterizing the post-§403
distribution of blood-lead
concentration, and comparing
health effects between pre-
and post-§403.
Apply the methods in Section 6.3 to
obtain pre- and post-intervention
distributions.
Rather than predicting post-§403 blood-lead
concentration as a function of environmental-
lead levels, use the average efficacy observed
in abatement studies with an adjustment for
bone-lead stores.
(Section 6.4.4)
When predicting the post-
intervention values of the
blood-lead distribution and
health effect endpoints,
determine an appropriate
value for the geometric
standard deviation (GSD) of
the blood-lead concentrations
associated with a given
environmental-lead exposure
scenario
Assume a GSD of 1.6
Alt, n: Assume a GSD of 1.4
Alt. #2: Assume a GSD of 1.9
Alt. #3: Assume a GSD of 2.1
Section 6.4.6
When using the IEUBK model
to predict post-intervention
values of the blood-lead
distribution and health effect
endpoints, determine an
appropriate value for daily
dietary lead intake for a child
aged 1-2 years (an input
parameter to the IEUBK
model)
Assume daily dietary lead intake is
5.78 (jg (the IEUBK model's default
value for children aged 1 -2 years)
Daily dietary lead intake = 1.29 |/g
Daily dietary lead intake = 3.53 fjg
Section 6.4.7
6-44
-------
Table 6-9. Procedures for Which Alternative Assumptions Were Considered in the
Sensitivity Analysis Addressing Risk Management. (Continued)
Procedure
Approach Taken in the
Risk Management Analyses
Alternatives] Considered in the
Sensitivity Analysis
When using modeling
techniques to predict post-
intervention values of the
blood-lead distribution and
health effect endpoints,
adjust model-based results to
reflect the effects of paint
pica tendencies on blood-lead
concentration
Make assumptions on the prevalence of
paint pica and the effects of paint pica
on blood-lead concentration that are
documented in Section 4.1.3 and
Appendix D1
Alt.
Make no adjustment for paint pica
effects
Alt. #2: Assume a lower prevalence of paint
pica and lower effects of paint pica on blood-
lead concentration than that used in the risk
analysis
Alt. #3: Assume a higher prevalence of paint
pica and higher effects of paint pica on blood-
lead concentration than that used in the risk
analysis
Qo/*tinn R A ft
where values of a and P are provided in Table 6-9. Assuming lognormality, upper and lower
one-sided 90% confidence bounds on the expected value of log(Wipe) are
(predicted value of log(Wipe)) ±1.3* SE(a + p * log(Vac))
where SE(a + P * log (Vac)) is the standard error of the expected value of log(Wipe). Upper and
lower 90% confidence bounds on the untransformed expected wipe dust-lead loadings are
obtained by exponentiating the bounds for the expected log-transformed loading.
The confidence bounds were used to define two alternative sets of converted dust-lead
loadings in the sensitivity analysis:
Alternative set #1:
Wipe dust-lead loading equals the lower 90% confidence bound on
the expected wipe dust-lead loading obtained from the formulas in
Table 6-9.
Alternative set #2:
Wipe dust-lead loading equals the upper 90% confidence bound on
the expected wipe dust-lead loading obtained from the formulas in
Table 6-9.
Note that alternative set #1 is a low estimate of the converted loading value, while alternative set
#2 is a high estimate. Under both sets, area-weighted arithmetic mean dust-lead loadings for
both floors and window sills were calculated for each HUD National Survey unit. The means
were used to determine whether candidate dust-lead loading standards were exceeded for a given
unit. In this part of the sensitivity analysis, numbers and percentages of units exceeding various
combinations of example environmental-lead standards were calculated under each set of
converted dust-lead loadings.
6-45
-------
Table 6-10 considers numbers of units exceeding an example floor dust-lead loading
standard of 100 ^ig/ft2, exceeding an example window sill dust-lead loading standard of 500
Hg/ft2, either of these two example standards, or any of the example standards for dust, soil, or
paint. These numbers were calculated for the wipe-equivalent dust-lead loadings used in the risk
management analyses, Alternative set #1, or Alternative set #2.
Table 6-10. Number (and Percentage) of Units in the 1997 National Housing Stock
Projected to Exceed Various Combinations of Example Standards, As
Determined from Three Different Sets of Converted Dust-Lead Loadings.
Example Standards, or
Combination of Standards
Floor-dust standard of 1 00
//g/ft2
Window sill-dust standard of
500 //g/ft2
Floor- or window sill- dust
standard
At least one dust or soil
standard, or paint intervention
trigger3
Number (%) of Units Exceeding the Example Standard(s)
Approach Used in
Risk Management
Analyses1
4,010,000
(4.04%)
1 2,400,000
(12.5%)
1 3,800,000
(13.9%)
21,600,000
(21.8%)
Using low Alternative
Estimates for
Converted Dust-Lead
Loading2
2,320,000
(2.34%)
9,760,000
(9.83%)
11,600,000
(11.7%)
20,300,000
(20.5%)
Using High Alternative
Estimates for Converted
Dust-Lead Loading2
5,750,000
(5.80%)
12,900,000
(13.0%)
15,800,000
(16.0%)
23,500,000
(23.6%)
1 See Section 4.3 on the methods for performing conversions from Blue Nozzle vacuum to wipe dust-lead loadings.
1 Low and high estimates correspond to the lower 90% confidence bound and upper 90% confidence bound, respectively,
for the estimates considered in the second column of this table.
3 Example soil standard and paint intervention triggers are as follows: soil-lead concentration of 2,000 //g/g for soil
removal, 5 ft1 of deteriorated lead-based paint for paint maintenance, and 20 ft1 of deteriorated lead-based paint for paint
abatement.
Effect on risk analysis: The largest variation between the two alternative sets of dust-
lead loadings occurred when considering only the example floor-dust standard. Under
Alternative set #2 (high converted values), 5.75 million units exceed the example floor-dust
standard of 100 ng/ft2, compared to four million units under the set of converted values used in
the risk management analyses, and 2.32 million units under Alternative set #1 (the low converted
values). This finding implies that the risk management analysis may be underestimating the
numbers of homes exceeding example standards by as much as 50%. However, a dust-cleaning
intervention is triggered if either the floor or window sill dust-lead loading standard is exceeded.
The impact of the uncertainty in the dust-lead loading conversion equation was smaller for the
number of homes in which either the example floor dust standard or window sill dust standard
was exceeded. The number of units triggering an intervention by exceeding either example dust
standard ranged from a low estimate of 11.6 million to a high estimate of 15.8 million, which is a
16% decrease or increase, respectively, from the estimate of 13.8 million units calculated in the
risk management analysis.
6-46
-------
6.4.2 Uncertainty in Converting Wipe Dust-Lead Loadings to Blue Nozzle Dust-Lead
Loadings for Determining Post-Intervention Blood-Lead Distributions Using the
Empirical Model
As described in Section 4.2, the empirical model is a multi-media regression model
developed especially for this risk analysis to predict the geometric mean blood-lead concentration
of children 1-2 years old as a function of environmental-lead levels at a child's primary
residence. Because data from the HUD National Survey are utilized to predict children's blood-
lead concentrations, the dust-lead loadings for floors and window sills inputted to the empirical
model are assumed to represent dust samples collected using the Blue Nozzle vacuum method
(i.e., the method used in the HUD National Survey). However, the dust-lead loading on floors
and window sills following dust-cleaning interventions were specified in terms of a wipe dust-
lead loading (Table 6-2). Thus, a means of converting post-intervention dust-lead loadings from
wipe to Blue Nozzle vacuum loadings was necessary.
Two formulas were used (Table 6-9) to predict a Blue Nozzle vacuum dust-lead loading
as a function of a wipe dust-lead loading, depending on whether a floor or window sill was
sampled. These formulas indicate that the expected value of the log-transformed Blue Nozzle
dust-lead loading (log(BN)) given a wipe dust-lead loading of "Wipe" takes the form
a + p*log(Wipe)
where estimates of a and P are provided in Table 6-9. Therefore, assuming lognormality, upper
and lower one-sided 90% confidence bounds on the expected value of log(BN) are
(predicted value of log(BN)) ±1.3* SE(
-------
Thus, Alternative #1 represents low estimates of the converted loadings, while Alternative #2
represents high estimates. Table 6-11 presents the resulting health effects under each of these
two alternatives, as well as under the converted loadings employed in Section 6.3.
Table 6-11. Empirical Model-Predicted Post-§403 Health Effect and Blood-Lead
Concentration Endpoints for Children 1-2 Years of Age, As Calculated
Under Three Assumptions on Post-Intervention Blue Nozzle Vacuum Dust-
Lead Loading1
Hearth Effect and Blood-Lead
Concentration Endpoints
PbB*20 (%)
PbBilO (%)
IQ<70 t%>
IQ decrement si (%)
1Q decrements 2 (%>
IQ decrement^ (%>
Avg. IQ decrement
Poet-Intervention Blue Nozzle Dutt-Lead Loading
Values Used in the Risk
Management Analyses (5.7
//g/ft* for floors, 14.0
pg/ft* for window sills)
0.406
4.70
0.110
36.3
9.30
2.93
1.00
Alternative #1
(4.5 //gm* for floors. 12.4
//g/ft* for window sills)
0.400
4.67
0.110
36.2
9.24
2.90
1.00
Alternative 92
<7.6//g/ft» for floors, 15.8
//g/ft* for window sOs>
0.412
4.74
0.111
36.4
9.36
2.96
1.01
1 Health effects are calculated assuming the following:
Example dust-lead loading standards of 100 //g/ft2 for floors and 500 //g/ft1 for window sills
Example soil-lead concentration standard of 2,000 //g/g
Paint maintenance is performed if more than 5 ft1, but less than 20 ft* of deteriorated lead-based paint exists.
Paint abatement is performed if more than 20 ft1 of deteriorated lead-based paint exists.
Blue Nozzle dust-lead loadings for floors and window sills equal to the minimum of the average pro-intervention
Blue Nozzle loading and the loading specified in the column heading.
Soil-lead concentrations equal to 150 fjg/g after soil removal intervention
0 ft1 of deteriorated lead-based paint after all paint interventions
Effect on risk analysis: For each alternative, deviation from the results for the risk
management analyses was negligible.
6.4.3 Alternative Assumptions on Post-Intervention Dust-Lead Loadings
Assumed post-intervention environmental-lead levels used hi the risk analysis were
provided in Table 6-2. The sensitivity analysis considered alternatives to the assumed post-
intervention wipe dust-lead loading following dust cleaning, ulterior paint abatement, and soil
removal, in order to observe how the health effect and blood-lead concentration estimates under
the empirical model were affected by assumptions on post-intervention dust-lead loadings. Two
sets of alternative post-intervention wipe dust-lead loadings for floors and window sills were
considered:
20 ug/ft2 for floors and 50 ug/ft2 for window sills, and
100 ug/ft2 for floors and 250 ug/ft2 for window sills.
6-48
-------
(The loadings used in the risk management analyses were 40 ng/ft2 for floors and 100 ng/ft2 for
window sills.) The sensitivity analysis did not address alternative soil-lead concentration values
following soil removal (150 ug/g), or amounts of deteriorated lead-based paint following paint
interventions (0 ft2).
Note that assumptions on post-intervention dust-lead loadings affect estimates of the
distribution of post-§403 blood-lead concentration and the health effect endpoints only when
these estimates are determined by the empirical model (Section 4.2). The IEUBK model (Section
4.1) uses post-intervention dust-lead concentration as input, and the methods used to determine
post-intervention dust-lead concentrations are not affected by assumptions on post-intervention
dust-lead loadings (Section 6.1.3). Therefore, health effect and blood-lead concentration
endpoints are estimated only under the empirical model here.
Table 6-12 summarizes the post-intervention estimates of childhood health effect and
blood-lead concentration endpoints (based on the empirical model) under the alternative post-
intervention dust-lead loadings. Results in Table 6-12 were calculated assuming the following
example dust and soil standards and paint intervention triggers:
Dust-lead loadings (under wipe sampling techniques) of 100 ug/ft2 for floors and
500 |ig/ft2 for window sills
Soil-lead concentration of 2,000 ug/g
Paint maintenance is performed if more than 5 ft2, but less than 20 ft2 of deteriorated
lead-based paint exists
Paint abatement is performed if more than 20 ft2 of deteriorated lead-based paint
exists.
Effect on risk analysis: Table 6-12 indicates that the health effect and blood-lead
concentration endpoints most affected by changes in the observed post-intervention dust-lead
loadings are those indicating the most extreme effects (e.g., IQ decrement of at least 3, blood-
lead concentration of at least 20 ug/dL). The percentage of children with blood-lead
concentration at or above 20 ug/dL differs from the estimate reported in the risk analysis by
approximately 4 to 6 percent under the two alternative post-intervention dust-lead loadings, while
an approximate 3 percent difference is observed for the percent of children with blood-lead
concentrations at or above 10 ug/dL. Virtually no difference in the estimated percentage of
children with IQ less than 70 or in average IQ decrement in a child as a result of lead exposure is
observed between the two alternatives.
6-49
-------
Table 6-12. Empirical Model-Predicted Post-§403 Percentages of Children Aged 1-2
Years Experiencing Specific Health Effect and Blood-Lead Concentration
Endpoints, Under Various Assumptions on Post-Intervention Dust-Lead
Loading.
Health Effect and Blood-Lead
Concentration Endpoints
PbB*20 {%)
PbB*10 <%)
IQ<70(%>
IQ decrements 1 (%}
IQ decrement 2 2 (%)
IQ decrements 3 (%)
Avg. IQ decrement
0 ft* Deteriorated Lead-Based Paint after all Paint Intervention*
Soil-Lead Concentration after Soil Removal Intervention = 150pg/g
Dust-Lead Loading*:
Floors = 20 //g/ff
Window Sills - 50/ig/ft*
0.388
4.59
0.110
36.0
9.12
2.85
0.998
Dust-Lead Loading1:
ROOTS => 40//g/fta
Window Sills * 1 00 //g/ft*
0.406
4.70
0.1 10
36.3
9.30
2,93
1.00
Dust-Lead Loading1:
Floors a 100j/g/ft*
Window SO* * 250//fl/ftl
0.429
4.85
0.111
36.7
9.53
3.04
1.01
1 After dust cleaning, soil removal, or interior paint abatement this analysis assumes the following example
dust and soil standards and paint intervention triggers:
Dust-lead loadings (under wipe techniques) of 100 jug/ft2 for floors and 500 //g/ft1 for window sills
Soil-lead concentration of 2,000 fjg/g
Paint maintenance is performed if more than 5 ft2, but less than 20 ft2 of deteriorated lead-based paint
exists
Paint abatement is performed if more than 20 ft2 of deteriorated lead-based paint exists.
Shaded cells correspond to results for example option C in Table 6-7.
6.4.4 Alternative Approach to Determining a Post-Intervention Blood-Lead Concentration
Distribution Using Directly-Measured Blood-Lead Concentration Changes
An alternative to the approach presented in Section 6.2 to characterizing a post-
intervention blood-lead concentration distribution was performed utilizing published results on
the effectiveness of lead hazard intervention strategies among children exposed to residential
lead hazards. This approach is desirable since blood-lead concentrations are a more direct
measure of intervention effectiveness than are environmental-lead levels. The scientific
literature reports the results of a range of non-medical intervention strategies conducted to reduce
the lead exposure of children residing at the targeted residences (USEPA, 1995b). The strategies
included lead-based paint abatement, interior dust abatement via routine cleaning procedures,
elevated soil-lead abatement, and intensive educational efforts (USEPA, 1995b). The
effectiveness of these strategies as measured by declines in children's blood-lead concentrations
may be used to estimate the post-intervention blood-lead concentration distribution. As such,
this approach represents a somewhat independent (of many of the procedures and data used in
risk management) estimation of a post-intervention distribution.
6-50
-------
As summarized in USEPA, 1995b, the intervention strategies reported 18-34% declines
in the blood-lead concentrations of exposed children six to twelve months following the conduct
of the intervention. Lead-based paint abatement (of all deteriorated LBP), biweekly dust
abatement (of areas with elevated dust lead), soil abatement (removal and replacement of top 6"),
and intensive education (visit by semi-professional outreach worker) reported comparable
declines of approximately 25% one year following conduct of the intervention (USEPA, 1995b).
Each of these four intervention studies reported significantly greater declines among the study
population than among a suitable control populationno control population was studied for the
educational intervention associated with the 34% declineproviding reassurance that the
interventions themselves were responsible for much of the reported declines. For the purpose of
this sensitivity analysis, therefore, the average decline in children's blood-lead concentration
resulting from an intervention was taken to be 25%'.
This degree of effectiveness may not be suitable for estimating the post-intervention
blood-lead distribution since the reported declines were for children already exposed (i.e., already
exhibiting elevated blood-lead concentrations due to exposure to the targeted lead source). By
contrast, the promulgation of §403 will prompt preventive interventions (primary prevention)
conducted prior to any lead exposure to resident children. Measures of secondary prevention
effectiveness may not be representative of primary intervention effectiveness because lead
present in blood is a combination of current environmental exposure and internal reservoirs of
lead stored in bone and soft tissue (Gulson et al., 1995; Smith et al., 1996; Rabinowitz et al.,
1976; Manton, 1985). The reported declines in exposed children's blood-lead concentrations,
therefore, may underestimate the primary prevention effectiveness of an intervention (Gulson
etal., 1995).
A methodology was developed to estimate the impact of body lead burdens on measures
of secondary intervention effectiveness to adjust the reported secondary prevention effectiveness
(see Appendix F2). For a two-year-old exposed child, it is estimated that a secondary
intervention prompting a 25% decline in blood-lead concentration at one year following
intervention would actually prompt 33% declines were the intervention primary in character
(Table F2-1 of Appendix F2). Based on this result, a 33% efficacy will be utilized for the
purposes of this portion of the sensitivity analysis. As a comparison, the IEUBK model indicates
a 41% primary prevention efficacy when lead-based paint hazards are eliminated and dust- and
soil-lead levels are lowered to background levels (Section 5.2).
It is worth noting that the scientific literature also includes two recent journal articles
regarding the percentage of lead hi blood that may be attributed to body lead stores (Gulson et al.,
1995; Smith et al., 1996). Such results, of course, have relevance to this aspect of the sensitivity
analysis. Both articles indicate that between 40-70% of lead in an adult's blood may be
attributed to mobilized bone-lead stores. The fact that these studies examined adults is critical
1 In all four studies, the control population did exhibit some decline which may be attributed to increased
awareness of environmental lead and its hazards. As similar awareness may be expected to accompany §403
prompted interventions, it was not deemed necessary to adjust the reported study population declines by the declines
associated with the control populations.
6-51
-------
because the percentage of blood lead attributable to bone-lead stores varies considerably with age
(Rabinowitz, 1991). Higher percentages are associated with older individuals (Rabinowitz,
1991). Thus, the population of 1-2 year olds considered in this risk analysis may have lower
percentages of their blood lead attributable to mobilized bone lead. Greater primary prevention
efficacy is reported for, say, 7 year old children than for 2 year old children (Table F2-1 in
Appendix F2). If the methodology used in this alternative approach is used to make inferences
on adults, it too suggests that 40-70% of blood lead is attributable to mobilized bone-lead stores.
This alternate approach to estimating a post-intervention national distribution of blood-
lead concentrations for 1997 children aged 1-2 years was implemented based on the estimated
33% decline in blood-lead concentration following an intervention. This alternative estimate of
primary prevention effectiveness, which adjusts the blood-lead changes for body-lead stores and
hereafter is denoted the 'adjusted blood-lead effects model', was then compared to post-
intervention distributions based on the EEUBK model and the empirical model.
The methodology for this comparison is summarized as follows:
1. Environmental-lead levels for each HUD National Survey unit were used as input to
the IEUBK and empirical models to predict the geometric mean blood-lead
concentration for children aged 1-2 years old exposed to environmental-lead levels
similar to that hi the National Survey unit. The contribution of pica was estimated
using the methodology documented hi Section 4.1.3.
2. For each unit in the HUD National Survey, lead levels hi paint, dust, and soil were
compared to the following example dust and soil standards and paint intervention
triggers (example option C in Table 6-7):
100 jig/ft2 for floor dust-lead loading and SOO ug/ft2 for window sill dust-lead
loading,
2,000 ug/g for soil-lead concentration,
Paint maintenance is performed if more than 5 ft2, but less than 20 ft2 of
deteriorated lead-based paint exists,
Paint abatement is performed if more than 20 ft2 of deteriorated lead-based
paint exists.
3. For each HUD National Survey unit, if an intervention was triggered, then the post-
intervention geometric mean blood-lead concentration was set equal to 67% of the
geometric mean computed in (1). If an intervention was not triggered, then the post-
intervention geometric mean blood-lead concentration equaled the geometric mean
calculated in (1).
6-52
-------
4. The geometric mean blood-lead concentration calculated in (3) and an assumed
geometric standard deviation of 1.6 were used to generate a distribution of blood-lead
concentrations for each unit in the HUD National Survey. The distributions were then
combined over all of the HUD National Survey units to yield estimated post-
intervention blood-lead concentrations under the IEUBK model or the empirical
model (Appendix E2).
Table 6-13 summarizes the health effect and blood-lead concentration endpoint values as
estimated in the baseline risk characterization (Section 5.1.1), in the risk management analysis
(Section 6.3), and under the adjusted blood-lead effects model. The table also includes the
geometric mean and geometric standard deviation of the blood-lead distributions.
Effect on risk analysis: According to Table 6-13, the post-intervention geometric mean
blood-lead concentrations under the adjusted blood-lead effects model were estimated to be 2.89
and 2.88 |ig/dL for the IEUBK and empirical models, respectively. The IEUBK model-predicted
geometric mean reported in the risk management analysis is slightly lower (2.74 |ig/dL), while
that predicted using the empirical model is slightly higher (3.03 ug/dL). Under the IEUBK
model, the estimated percentages of children with blood-lead concentration at or above 10 or 20
ug/dL are greater using the adjusted blood-lead effects approach than those predicted in the risk
management analysis. This results from the differences in the geometric standard deviations of
blood-lead concentrations between the two approaches (1.97 and 1.84). Under the empirical
model, percentages of children with blood-lead concentrations at or above than 10 or 20 ug/dL
are less using the adjusted blood-lead effects approach than those predicted in the risk
management analysis. This results from the differences hi the geometric mean blood-lead
concentrations between the two approaches.
6.4.5 Uncertainty in Assumptions Made in Determining Post-Intervention Dust-Lead
Concentrations
As the IEUBK model requires dust-lead levels to be input as concentrations for predicting
the geometric mean blood-lead concentration associated with a given exposure scenario (Section
4.1), it was necessary to develop a method for determining (ulterior) floor dust-lead
concentrations that result from interventions performed under §403 rules. This method was
presented hi Section 6.1.3. In this section, uncertainty associated with key assumptions made in
this method is characterized.
To determine post-intervention floor dust-lead concentrations, the following two
assumptions were made:
1. an 80% reduction in floor dust-lead concentration results whenever a paint
intervention is conducted (regardless of any other type of intervention that may be
conducted)
2. the amount of floor-dust lead that is attributable to soil is equal to 80% of the amount
of lead in the soil.
6-53
-------
Table 6-13. Estimated Post-§403 Health and Blood-Lead Concentration Endpoints Based
on the Risk Assessment Approach and the Adjusted Blood-Lead Effects
Approach.
Health Effect and
Blood-Lead
Concentration
Endpoints
PbB * 20 (%)
PbB* 10(%)
IQ < 70 (%)
IQ decrement * 1 (%)
IQ decrement * 2 (%)
IQ decrement * 3 (%)
Avg. IQ decrement
Geom. Mean PbB
(GSD)
Baseline
(Section
5.1,1)
0.588
5.75
0.115
38.5
10.8
3.70
1.06
3.14
(2.09)
Post- §403 Estimates
Under the Adjusted Blood Lead
Effects Model
iEUBK Model
0.213
3.33
0.105
33.0
7.16
1.96
0.934
2.89
(1.97)
Empirical
Model
0.302
3.89
0.107
33.4
7.94
2.37
0.949
2.88
(2.03)
Post-5403
Estimates Under the
Risk Management Analysis
IEUBK Model
0.0539
1.66
0.0984
28.3
4.31
0.858
0.848
2.74
(1.84)
Empirical
Model
0.406
4.70
0.110
36.3
9.30
2.93
1.00
3.03
(2.04)
Example dust and soil standards were set at: 100/jg/ft1 for floor dust-lead loading, 500pg/ft2 for window sill dust-lead
loading, and 2,000 //g/g for soil-lead concentration. Paint maintenance is performed if more than 5 ft1, but less than 20
ft2, of deteriorated lead-based paint exists. Paint abatement is performed if more than 20 ft1 of deteriorated lead-based
paint exists.
GSD = geometric standard deviation.
To investigate the uncertainty associated with Assumption #1, post-intervention floor
dust-lead concentrations measured in two studies were compared to those preicted by the
algorithm in Section 6.1.3. The two studies were the Boston phase of the Urban Soil Lead
Abatement Demonstration Project (USLADP; Section 3.2.2.4) and the Baltimore Repair and
Maintenance (R&M) study (Section 3.2.2.1). These studies were selected because pre- and post-
intervention floor dust-lead concentrations were measured and because they assessed the efficacy
of paint interventions (among other interventions). The algorithm presented in Section 6.1.3 was
used to predict the post-intervention dust-lead concentration (i.e., an 80% reduction from pre-
intervention levels) for "study group" units in the Boston USLADP and "R&M Level III" units in
the Baltimore R&M study. Figures 6-13 and 6-14 plot the predicted versus observed average
post-intervention floor dust-lead concentrations in these units for the Boston USLADP and
Baltimore R&M study, respectively. The solid line in both plots indicates equality. In both
plots, the line of equality appears to be a good fit to the data points, indicating that the 80%
reduction in dust-lead concentration from pre-intervention conditions is a good estimate of the
post-intervention dust-lead concentration. However, there is considerable variability between the
data points and this line, indicating that while the assumption is good when considering an
average across all units, it may not be appropriate in certain units.
6-54
-------
10000
c
0
- 1000
5
o
o.
c
o
o
c
o
o
1
o
100
10
10 100 1000 10000
Observed Floor Oust Lead Concentration Post Intervention
Figure 6-13. Predicted Versus Observed Average Post-Intervention Floor Dust-Lead
Concentration (//g/g) (Boston USLADP Study Group Homes).
100000
c
I 10000
_c
I
2 1000
M
o
100
10
10 100 1000 10000 100000
Observed Floor Dust Lead Concentration Post Intervention
Figure 6-14. Predicted Versus Observed Average Post-Intervention Floor Dust-Lead
Concentration (/ig/g) (Baltimore R&M Level III Homes).
6-55
-------
To investigate the extent to which floor dust-lead concentration declines following
interventions, Tables 6-14 and 6-15 present geometric mean concentrations at specific times
following intervention and how these geometric means have declined from pre-intervention
values. Table 6-14 show results for Baltimore R&M study units according to housing
type/group. This table shows that 80% declines are typical for R&M IE study units (which had
the most intensive intervention strategies) throughout the months following intervention. Similar
results are seen in the "study group" of units in Table 6-15, which shows results for the Boston
USLADP.
Table 6-14. Geometric Mean Post-Intervention Floor Dust-Lead Concentration (//g/g), and
Percent Difference from Pre-lntervention Levels, for the Baltimore R&M
Study.
% Months
Post-
Intervention
Pre-
intervention
06
12
18
24
30
Modem Urban
Unit*
Qeom.
Mean
85.8
92.1
55.1
72.1
45.0
65.1
%Dtff.
from
Pre-lnt.
7.3%
-3.5%
-16.0%
-47.6%
-24.1%
Previously Abated
Units
Qeom.
Mean
736.4
876.5
715.1
731.2
523.9
531.5
%DrH.
from
Pre-tnt.
19.0%
-2.9%
-0.7%
-28.9%
-27.8%
R&M 1 Units
Qeom.
Mean
1,413
846.7
769.8
490.9
716.9
%Drff.
from
Pre-lnt
-40.1%
-45.5%
-65.3%
-49.3%
R&M H Units
Qeom.
Mean
1,930
621.9
684.0
484.3
332.4
%DJff.
from Pre-
lnt.
-
-67.8%
-64.6%
-74.9%
-82.8%
R&M HI Units
Qeom.
Mean
3,970
931.1
578.6
718.2
547.3
442.7
%Diff.
from
Pre-lnt.
-76.5%
-85.4%
-81.9%
-86.2%
-88.8%
Table 6-15. Geometric Mean Post-Intervention Floor Dust-Lead Concentration (//g/g), and
Percent Difference from Pre-lntervention Levels, for the Boston USLADP.
Study Phase
1
Recontamination
No. 1
Recontamination
No. 2
# Months
Post-
intervention
Pre-
lntervention
6
11
Study Group
Qeom.
Mean
6,623
3,108
1,294
% Diff.
from Pre-
lnt.
-53.1%
-80.5%
Control Group A
Geom.
Mean
4,202
1,458
1,300
%Dtff.
from Pre-
lnt.
-65.3%
-69.1%
Control Group 8
Geom.
Mean
5,178
1,493
1,886
%Diff.
from Pre-
lnt.
-71.2%
-63.6%
6-56
-------
Pre-intervention data from the Baltimore phase of the USLADP were used to investigate
Assumption #2. Figure 6-15 plots (pre-intervention) floor dust-lead concentration versus (pre-
intervention) fine soil-lead concentration for units in this phase. The solid line in Figure 6-15
represents a lower bound on dust-lead concentration when assuming that the soil contributes 80%
of the mass of dust. Only 12% of the units have data which fall below this line, which is within
range of what can be expected under Assumption #2 given the measurement errors in soil-lead
and dust-lead concentrations. Figure 6-15 also contains lines that represent soil contributions of
20%, 40% and 60% of the total mass of floor dust.
Post-intervention dust-lead concentrations measured in the Baltimore USLADP were
compared to those predicted by the algorithm in Section 6.1.3. Because paint interventions were
not conducted in the Baltimore USLADP, this comparison provides an assessment of
assumption 2. Figure 6-16 plots predicted post-intervention floor dust-lead concentration versus
measured concentration, with the solid line representing equality. This plot does not indicate a
particular bias in the prediction procedure for these units, supporting the approach taken for units
with no paint interventions. However, large differences between the observed and predicted
post-intervention concentrations are present for certain units.
6.4.6 Alternative Estimates for the Geometric Standard Deviation
of Blood-Lead Concentrations
The sensitivity of pre-§403 model-based estimates of the health effect and blood-lead
concentration endpoints to various assumptions on the GSD for childhood blood-lead
concentrations was presented hi Section 5.4.6. Three alternative GSD values were considered:
1.4,1.9, and 2.1. In this section, post-§403 estimates of the health effect and blood-lead
concentration endpoints are estimated (under a single set of example options for standards, using
both the IEUBK and empirical models) under these same alternative GSD values. See Section
5.4.6 for additional details on how the alternative GSD values were selected and on interpreting
the GSD hi this risk analysis.
Effect on risk analysis: For the three alternative GSD values, as well as for the GSD of
1.6 used in the risk analysis, Table 6-16 presents the estimated post-§403 health effect and blood-
lead concentration endpoints for the example standards specified in the footnote to the table. As
was seen in Table 5-14, post-§403 risks increase as the assumed GSD increases (i.e., larger
percentage of children with blood-lead concentrations greater than or equal to 10 or 20 ug/dL).
The IEUBK model is considerably more sensitive than the empirical model to the GSD value.
For example, the probability of a child having a blood-lead concentration at or above 10 ug/dL
increases by 41% under the IEUBK model (from 1.46% to 2.07%) when the GSD increases from
1.4 to 2.1, compared to only a 7% increase under the empirical model (from 4.56% to 4.88%).
The probability of a child having a blood-lead concentration at or above 20 ug/dL more than
doubles under the ffiUBK model (from 0.0404% to 0.0865%), while only a 16% increase is
observed under the empirical model (from 0.378% to 0.440%). Higher sensitivity to the GSD
value was also observed for the IEUBK model versus the empirical model for the IQ parameters.
6-57
-------
_o
I
"c
o
o
3
o
o
.3
n
O
100000i
10000
1000
100
10
o o
10 100 1000
Soil-Lead Concentration Fine Sieve (ppm)
10000
ooo Treatment
60%
Control
40% -
80%
20%
Figure 6-15. Average Floor Dust-Lead Concentration Versus Average Fine Soil-Lead
Concentration (Baltimore USLADP Homes).
g 100000
c 10000
O-
c
| 1000
o
I
1
o
I
100
10
10 100 1000 10000
Observed Floor Dust Lead Concentration Post Intervention
100000
o o o UP
No LBP
R«(«r«nc« Un»
* * PbD < .8 x PbS
***lBP7(Ult«lng)
Figure 6-16. Predicted Versus Observed Average Post-Intervention Floor Dust-Lead
Concentration (//g/g) (Baltimore USLADP Treatment Group Homes).
6-58
-------
Table 6-16. Sensitivity Analysis on the Estimated Post-§403 Health Effect and Blood-
Lead Concentration Endpoints for Children Aged 1 -2 Years, Under Three
Alternative Values (1.4, 1.9, 2.1) for the Geometric Standard Deviation
(GSD) of the Blood-Lead Concentration Distribution and Under the Value
Used in the Risk Analysis (1.6).1
Health Effect and Blood-
LMd Concentration
Endpoints
PbB * 20 (%)
PbBk 10(%)
IQ < 70 (%)
IQ decrement 2 1 (%)
IQ decrement * 2 (%)
IQ decrement *3 (%)
Average IQ decrement
(ft points)
Predictions Using the IEUBK Model
GSD «
1.4
0.0404
1.46
0.0977
27.8
3.94
0.731
0.841
GSD »
1.6
0,0539
1.66
0*0384
28.3
4.3t
o.ase
0.848
GSD**
1.9
0.0742
1.93
0.0994
28.8
4.77
1.03
0.857
QSO =
2,1
0.0865
2.07
0.0999
29.1
5.01
1.12
0.862
Predictions U*ing the Empirical ModeP
GSD -
1.4
0.378
4.56
0.110
36.2
9.11
2.82
1.0O
GSD =
t.e
0.406
4.70
0.110
36*3
9.30
2,93
1.00
GSD »
1.9
0.430
4.83
0.111
36.4
9.47
3.02
1.01
GSD *
2.1
0.440
4.88
0.111
36.5
9.53
3.06
1.01
1 The specified GSD represents variability associated with blood-lead concentrations in children aged 1-2 years who are
exposed to the same set of environmental-lead levels. Health effects are calculated assuming the following:
Example dust-lead loading standards of 100 j/g/ft2 for floors and 500 //g/ft2 for window sills
Example soil-lead concentration standard of 2000 fjglg
Paint maintenance is performed if more than 5 ft2, but less than 20 ft2 of deteriorated lead-based paint exists
Paint abatement is performed if more than 20 ft2 of deteriorated lead-based paint exists
Shaded cells correspond to results presented in Table 6-7 (under example options "C"). Only IQ decrement and occurrences
of IQ < 70 that result from exposure to lead-based paint hazards are considered in calculating health effect end points.
6.4.7 Alternative Estimates for Daily Dietary Lead Intake Assumed in
Fitting the IEUBK Model
Section 5.4.7 considered how alternative values for daily dietary lead intake in children
aged 1-2 years affected IEUBK model-based, pre-§403 estimates of the health effect and blood-
lead concentration endpoints. The alternative values were 1.29 ug and 3.53 ng, compared to the
value of 5.78 ug considered in the risk analysis, hi this section, post-§403 health effect and
blood-lead concentration endpoints are estimated (using the IEUBK model, under a single set of
example options for standards) under these same alternative assumptions on daily dietary lead
intake. See Section 5.4.7 for details on how the alternative values were selected.
Effect on risk analysis: Under the two alternative daily diet intake values (as well as the
default value used in the risk analysis), Table 6-17 presents the IEUBK model-predicted post-
§403 health effect and blood-lead concentration endpoints for the example standards provided in
the footnote to the table. The probability of a child having a blood-lead concentration at or above
20 ng/dL is reduced by 34% (from 0.0539% to 0.0355%) when daily dietary lead intake
decreases from 5.78 ug to 1.29 ng, while the probability of a child having a blood-lead
concentration at or above 10 ug/dL is reduced by 20% (from 1.66% to 1.32%).
6-59
-------
Table 6-17. Sensitivity Analysis on the IEUBK Model-Predicted Post-§403 Health Effect
and Blood-Lead Concentration Endpoints for Children Aged 1-2 Years, Under
Two Alternative Values (1.29 //g, 3.53 //g) for the Daily Lead Dietary Intake
Parameter and Under the Value Used in the Risk Analysis (5.78 //g).1
Health Effect and Blood-Lead
Concentration Endpoints
PbB 2 20 (%)
PbB 2 10(%)
IQ < 70 (%)
IQ decrement 2 1 (%)
IQ decrement 2 2 (%)
IQ decrement 23 (%)
Average IQ decrement (# points)
Geometric mean blood-lead
cone. (i/g/dL)
IEUBK Model-Predicted Post-1403 Estimates
Lead intake: 1 .29 //g/day
0.0355
1.32
0.0970
26.4
3.62
0.658
0.821
2.68
Lead intake: 3.53 pg
0.0497
1.43
0.0967
24.6
3.67
0.744
0.791
2.53
Lead intake: 5.78 fjg
0.0539
1,66
0.0984
28.3
4.31
0.858
0.848
2.74
Hearth
effects are calculated assuming the following:
Example dust-lead loading standards of 100 //g/ft2 for floors and 500 A/g/ft2 for window sills
Example soil-lead concentration standard of 2000 //g/g
Paint maintenance is performed if more than 5 ft2, but less than 20 ft2 of deteriorated lead-based paint exists
Paint abatement is performed if more than 20 ft2 of deteriorated lead-based paint exists
Shaded cells correspond to results presented in Table 6-7 (under example options "C"). Only IQ decrement and occurrences
of IQ < 70 that result from exposure to lead-based paint hazards are considered in calculating health effect endpoints.
In general, the impact of varying the daily dietary lead intake on the estimated endpoints
is minimal. For example, the geometric mean post-§403 blood-lead concentration for daily
dietary lead intakes of 1.29 and 5.78 fig were 2.74 and 2.68 (ig/dL, respectively. The post-§403
geometric mean is computed by multiplying the pre-§403 geometric mean (determined by
NHANES HI) by the ratio of the model-predicted geometric means (see appendix Fl and Step 3
in Section 6.2). The ratio (post-§403 geometric mean divided by pre-§403 geometric mean) is
determined by fitting the LEUBK model to pre- and post-§403 environmental-lead data. Because
changing the daily dietary lead intake has a similar effect on the IEUBK model-predicted pre-
and post-§403 geometric means, the ratio of the IEUBK model-predicted geometric means is
robust to variations in the daily dietary lead intake.
6.4.8 Alternative Assumptions on Paint Pica Tendencies in Children and the
Effect of Paint Pica on Blood-Lead Concentration
Section 5.4.8 considered alternative assumptions on the method for obtaining a model-
predicted geometric mean blood-lead concentration for children with a history of ingesting paint
chips. This section considers how these alternative assumptions affect estimated post-§403
health effect and blood-lead concentration endpoints. Results of this sensitivity analysis are
presented separately for each model.
6-60
-------
6.4.8.1 Empirical Model
When applying the empirical model to characterize the distribution of blood-lead
concentration in children aged 1-2 years, it is assumed that 9% of children residing in housing
units with deteriorated lead-based paint ingest paint chips in some manner. The sensitivity
analysis considers three alternatives to this assumed percentage: 0%, 6%, and 14%. The
assumption of 0% is equivalent to making no adjustment for paint pica, while the assumptions of
6% and 14% correspond to the lower and upper limits of an approximate 95% confidence
interval on the percentage of children with paint pica tendencies hi the Rochester Lead-in-Dust
study.
Effect on risk analysis: Table 6-18 presents the post-§403 health effect and blood-lead
concentration endpoints, as estimated by the empirical model, under the three alternative
assumptions on the percentage of children with paint pica tendencies hi units with deteriorated
lead-based paint (the assumed set of example options for the standards is provided hi a footnote
to the table). Values under the 9% assumption used in the risk analysis are also included in this
table for comparison purposes.
Results hi Table 6-18 indicate that as the assumed pica percentage increases, the
estimated endpoints decrease. The reason for this trend will be explained in terms of the
estimated geometric means given in the last row. The post-§403 geometric mean is computed by
multiplying the pre-§403 geometric mean (determined by NHANES HI) by the ratio of the
model-predicted geometric means (see Appendix Fl and Step 3 in Section 6.2). A total of 55
housing units in the HUD National Survey contained deteriorated lead-based paint. Upon
conducting paint interventions under the example standards considered in Table 6-18, only 9
housing units continued to contain deteriorated lead-based paint. Therefore, increasing the
percentage of children hi such housing who have paint pica tendencies increases the pre-§403
model-predicted geometric mean more than the post-§403 model-predicted geometric mean.
Therefore, increasing the percentage of children with paint pica decreases the ratio of the model-
predicted geometric means and consequently reduced the post-§403 geometric mean.
The change hi the estimated endpoints (based on the empirical model) is generally small.
The percentage of children with blood-lead concentration greater than or equal to 20 ug/dL
increased by 6.9% (from 0.406% to 0.434%) when the 9% assumption was decreased to 0%; the
percentage increase in other endpoints is even less. When the assumption is increased from 9%
to 14%, a 3.7% decline in the percentage of children with blood-lead concentration greater than
or equal to 20 ug/dL (from 0.406% to 0.391%) is observed.
6-61
-------
Table 6-18. Sensitivity Analysis on the Empirical Model-Predicted Post-§403 Health
Effect and Blood-Lead Concentration Endpoints for Children Aged 1 -2 Years,
Under Three Alternative Values (0%, 6%, 14%) for the Percentage of
Children with Paint Pica Tendencies, and Under the Value Used in the Risk
Analysis (9%).1
Health Effect and Blood-Lead
Concentration Endpoints
PbB 2 20 (%)
PbB 2 10 (%)
IQ < 70 (%)
IQ decrement i 1 (%)
IQ decrement 2 2 (%)
IQ decrement 2 3 (%)
Average IQ decrement (# points)
Geometric mean blood-lead concentration U/g/dL)
0%
0.434
4.87
0.111
36.7
9.55
3.05
1.01
3.048
6%
0.415
4.76
0.111
36.4
9.38
2.97
1.01
3.038
9%
0.406
4.70
0.110
36.3
9.30
2.93
1.00
3.034
14%
0.391
4.61
0.110
36.2
9.17
2.86
1.00
3.026
1 Health effects are calculated assuming the following:
Example dust-lead loading standards of 100 pg/ft2 for floors and 500 jug/ft2 for window sills
Example soil-lead concentration standard of 2000 fjg/g
Paint maintenance is performed if more than 5 ft2, but less than 20 ft2 of deteriorated lead-based
paint exists
Paint abatement is performed if more than 20 ft2 of deteriorated lead-based paint exists
Shaded cells correspond to results presented in Table 6-7 (under example options "C"). Only IQ decrement
and occurrences of IQ < 70 that result from exposure to lead-based paint hazards are considered in calculating
health effect endpoints.
6.4.8.2 IEUBK Model
The approach to accounting for the effects of paint pica on geometric mean blood-lead
concentrations estimated from the IEUBK model is more complex than that for the empirical
model, due to the greater number of assumptions going into the approach. Assumptions in the
risk analysis are as follows:
9% of children aged 1-2 years have paint pica tendencies
0.03% of children aged 1-2 years living in housing units containing damaged lead-
based paint have recently ingested paint chips.
children aged 1-2 years who recently ingested paint chips have a blood-lead
concentration of 63 (ig/dL.
children aged 1-2 years who ingested paint chips at some time, but not recently, have
a 3 ug/dL increase in their geometric mean blood-lead concentration from children
who do not ingest paint chips.
6-62
-------
In the sensitivity analysis, three sets of alternative assumptions were considered:
Alternative set #1: Assumes 0% of children have paint pica tendencies. (This is equivalent to
making no adjustment for paint pica.)
Alternative set #2: Assumes that pica tendencies have a lower impact than that observed in the
risk analysis:
6% of children aged 1-2 years have paint pica tendencies (the lower
bound of a 95% confidence interval on the percentage in the Rochester
Lead-in-Dust study).
0.01 % of children aged 1 -2 years living in housing units containing
damaged lead-based paint have recently ingested paint chips.
children aged 1-2 years who recently ingested paint chips have a blood-
lead concentration of 55 ug/dL (a low estimate based on information
from McElvaine et al., 1992).
children aged 1-2 years who ingested paint chips at some time, but not
recently, have a 15% increase hi their geometric mean blood-lead
concentration from children who do not ingest paint chips (the lower
bound of a 95% confidence interval on the percentage increase as
estimated from the Rochester Lead-in-Dust study).
Alternative set #3: Assumes that pica tendencies have a larger impact than that observed in the
risk analysis:
14% of children aged 1-2 years have paint pica tendencies (the upper
bound of a 95% confidence interval on the percentage hi the Rochester
Lead-in-Dust study).
0.10% of children aged 1 -2 years living in housing units containing
damaged lead-based paint have recently ingested paint chips.
children aged 1-2 years who recently ingested paint chips have a blood-
lead concentration of 63 (ig/dL.
children aged 1-2 years who ingested paint chips at some time, but not
recently, have a 100% increase in their geometric mean blood-lead
concentration from children who do not ingest paint chips (the upper
bound of a 95% confidence interval on the percentage as estimated from
the Rochester Lead-in-Dust study).
Effect on risk analysis: Table 6-19 presents estimated post-§403 endpoints, as
estimated by the IEUBK model under the three alternative sets of assumptions, as well as under
the set of assumptions used in the risk analysis. As seen hi Table 6-18, the estimated endpoints
decrease as the prevalence of paint pica and the effect of paint pica on blood-lead concentration
increases. The reason for the decreasing trend is similar to that explained in the previous
subsection.
6-63
-------
According to Table 6-19, the set of pica assumptions considered in the risk analysis yields
estimated endpoints closer to those under the low-end alternative sets (sets #1 and #2) than under
the high-end alternative set #3. The percent increase in the estimated endpoints between the risk
analysis assumptions and alternative set #1 is no higher than 9%, while percent declines between
the risk analysis assumptions and alternative set #3 are as high as 35% (e.g., the percentage of
children with blood-lead concentrations at or above 20 ng/dL declines from 0.0539% to
0.0348%). Therefore, if assumptions on the prevalence and health effects of pica are actually
greater than those considered in the risk analysis, the post-§403 estimated endpoints may be less
than those estimated in the risk analyses.
6.4.9 Standard Errors for Health Effect and Blood-Lead Concentration Endpoints Due to
Sampling Variability
The health effect and blood-lead concentration endpoints presented in Tables 6-4 to 6-7
are based on models for predicting blood-lead concentration from environmental lead measured
in the HUD National Survey, conversions between various types of measured data, assumed
relationship between IQ point loss and blood-lead concentration, and assumptions on the post-
intervention environmental-lead levels. Earlier subsections investigated the sensitivity of the risk
analysis to assumptions on conversions, relationship between IQ point loss and blood-lead
concentration, and post-intervention environmental-lead levels by modifying the assumptions and
recalculating the health effect and blood-lead concentration endpoints. In this section,
uncertainty in the estimated post-§403 health effect and blood-lead concentration endpoints as a
result of sampling variability in the HUD National Survey and NHANES HI is characterized.
As described in Section 3.3, the HUD National Survey collected samples from 284
homes. The environmental-lead levels hi these homes are used to represent a sample of the
environmental-lead levels in the nation's housing. If a different set of 284 homes was sampled,
then the estimated health effect and blood-lead concentration endpoints would be different. For
three sets of example options for the §403 standards, statistical analyses were conducted to
characterize the variability in the estimated post-§403 health effect and blood-lead concentration
endpoints due to the sampling variability of the 284 homes. For each set of example standards,
standard errors were computed for each of the estimated health effect and blood-lead
concentration endpoints based on a Monte Carlo (bootstrap) analysis (Efron and Tibshirani,
1993). The standard errors were derived by recomputing the endpoints for each of 1,000
different samples of size 284 drawn with replacement from the 284 homes. For each of the 1,000
samples generated, a sample was taken with replacement from the 987 children aged 1-2 years hi
the NHANES m, Phase 2 data. Then, for each of these 1,000 sets of samples, the same
procedures used hi the risk management analyses (Section 6.2) were applied to compute each of
the endpoints.
6-64
-------
Table 6-19. Sensitivity Analysis on the IEUBK Model-Predicted Post-§403 Health Effect
and Blood-Lead Concentration Endpoints for Children Aged 1-2 Years, Under
Three Alternative Sets of Assumptions on Paint Pica Effects, and Under the
Set of Assumptions Used in the Risk Analysis.1
Health Effect and Blood-Lead
Concentration Endpoints
PbB * 20 (%)
PbBi 10(%)
IQ < 70 (%)
IQ decrement * 1 (%)
IQ decrement * 2 (%)
IQ decrement 2 3 (%)
Average IQ decrement (#
points)
Geometric mean blood-lead
concentration U/g/dL)
Ptea Assumptions
in the Risk
Analysis
0,0538
t.66
0.0984
284
4,31
0.858
0-848
2.74
Pica Alternative
S«t#1
{no adjustment)
0.0586
1.74
0.0988
28.6
4.45
0.904
0.853
2.755
Pic* Attamative
Set #2
(tow adjustment)
0.0568
1.71
0.0986
28.5
4.40
0.887
0.852
2.752
Pica Alternative
Set #3
flifcb adjustment)
0.0348
1.34
0.0972
27.1
3.69
0.663
0.830
2.715
1 Health effects are calculated assuming the following:
Example dust-lead loading standards of 100 j/g/ft2 for floors and 500 //g/ft2 for window sills
Example soil-lead concentration standard of 2000 //g/g
Paint maintenance is performed if more than 5 ft2, but less than 20 ft2 of deteriorated lead-based
paint exists
Paint abatement is performed if more than 20 ft2 of deteriorated lead-based paint exists
Shaded cells correspond to results presented in Table 6-7 (under example options "C"). Only IQ decrement
and occurrences of IQ < 70 that result from exposure to lead-based paint hazards are considered in calculating
health effect endpoints.
Table 6-20 displays the standard errors for the estimated health effect and blood-lead
concentration endpoints under each of the three sets of example standards, along with estimates
of the standard errors of these estimates. Approximate 95% confidence intervals for the
estimated endpoints can be computed by adding and subtracting two times the standard error to
the respective endpoint. For instance, under the first set of standards presented in Table 6-20, the
lower bound of the 95% percent confidence interval for the percentage of homes exceeding any
of the standards is 17.5 - (2 * 2.1) = 13.3%, while the upper bound is 17.5 + (2 * 2.1) = 21.7%.
hi general, the standard errors displayed in Table 6-20 are quite small. This suggests that
other sources are likely to have a larger impact on overall uncertainty than the sampling
variability hi the HUD National Survey and in NHANES III. Other sources include uncertainty
associated with the conversion equations, assumptions on post-intervention environmental-lead
levels, the ability of the models (IEUBK and empirical) to predict blood-lead concentration from
environmental levels, the relationship between IQ point loss and blood-lead concentration, the
6-65
-------
Table 6-20. Estimates of Standard Errors Associated with Estimated Post-§403 Health
Effect and Blood-Lead Concentration Endpoints and with Number of Homes
Exceeding Standards, for Three Sets of Example Options for the §403
Standards.
Example Options for Standard*
Floor Dust-Lead Loading
Window Sill Dust-Lead
Loading
Soil-Lead Concentration
Paint Maintenance Trigger
Paint Abatement Trigger
STANDARD/TARGET
Floor Dust
Window Sill Dust
Soil Removal
Interior Paint Maintenance
Exterior Paint Maintenance
Interior Paint Abatement
Exterior Paint Abatement
Exceeding Any Standard
400
800
5000
10
100
Estimate
Standard
Error
100
500
200O
5
20
Estimate
Standard
Error
25
25
500
0
&
Estimate
Standard
Error
Percentage of Homes Exceeding Example Standards*
0.00
10.3
0.215
2.80
3.84
0.453
3.03
17.5
0.00
1.7
0.273
0.92
1.11
0.382
0.97
2.1
4.04
12.5
2.49
2.92
3.49
2.43
5.77
21.8
1.13
1.9
0.87
0.94
1.05
0.87
1.34
2.3
13.8
48.1
11.8
1.08
1.15
5.35
9.26
53.7
1.7
2.7
1.8
0.59
0.59
1.26
1.61
2.7
Predicted Health Effect and Blood-Lead Concentration Endpointe (Based on
Empirical Model)**
PbBi20 (%)
PbBi10(%)
10 < 701%)
IQ decrements 1 (%)
1Q decrement^ (%)
IQ decrement;^ (%}
Avg. }Q decrement
0.458
5.03
0.112
37.1
9.79
3.16
1.02
0.094
0.53
0.002
1.3
0.78
0.39
0.03
0.406
4.70
0.110
36.3
9.30
2.93
1.00
0.088
0.52
0.002
1.3
0.78
0.38
0.03
0.317
4.09
0.108
34.7
8.34
2.49
0.971
0.074
0.49
0.002
1.3
0.75
0.35
0.027
Predicted Health Effect and Blood-Lead Concentration Endpoints {Based on
IEUBK Model)**
PbB z 20 (%)
Pb8s10(%)
IQ<70{%)
IQ decrement* 1 (%)
id decrement i 2 (%)
IQ decrement^ <%)
Avg. JQ decrement
0.290
3.92
0.107
34.5
8.09
2.37
0.964
0.081
0.56
0.002
1.5
0.87
0.40
0.030
0.0539
1.66
0.0984
28.3
4.31
0.858
0.848
0.0429
0.59
0.0024
2.3
1.12
0.367
0.038
0.00198
0.250
0.0909
15.1
0.978
0.0976
0.666
0.00354
0.167
0.001 1
2.6
0.469
0.0801
0.031
1 Standard error estimates for percentage of homes affected by standards are based on 2,000 bootstrap replicates.
' Standard error estimates for health effects are based on 1,000 bootstrap replicates.
6-66
-------
assumption of lognormality in blood-lead concentration, and uncertainty in sample mean and
sample standard deviation (on a log scale) associated with blood-lead concentration. In addition,
these standard errors were computed assuming simple random sampling and do not account for
the complex survey design employed in the HUD National Survey.
6.5 CONCLUSION
The primary purpose of the risk management analyses is to develop and apply
methodology for analyzing example options for the §403 standards. To that end, various example
options for the §403 standards for lead in paint, dust, and soil were evaluated in this chapter. The
example options were assessed by predicting the incremental risk reductions expected to result
after interventions are conducted in response to the proposed §403 rule.
Estimating the impact of the proposed §403 rule on health effect and blood-lead
concentration endpoints for children aged 1-2 years is a very complicated and challenging
problem. A series of technical analyses were conducted to address this problem. The following
four points summarize the analyses conducted.
First, estimating the impact of the proposed §403 rule required estimating the
distribution of environmental-lead levels expected to result from promulgation of the
§403 standards. This was accomplished by assuming that homeowners would take
actions in response to the various example standards.. Predicting the actual responses
of homeowners to the proposed rule is a difficult problem. For the purposes of the
risk management analyses, a set of six interventions were defined and utilized for the
analyses of various example options for the §403 standards: one dust intervention,
one soil intervention, two exterior paint interventions, and two interior paint
interventions. The effectiveness and duration of effectiveness of each of the six
interventions is defined in terms of environmental-lead levels. To the extent possible,
the assumed efficacies and durations are based on data in the scientific literature.
Second, determining the impact of the various example options for the proposed §403
standards required estimating the numbers and percentages of homes affected by each
example option for the §403 standards for lead in paint, dust, and soil. The HUD
National Survey is the most complete and extensive set of data on lead levels in paint,
dust, and soil hi the nation's housing. However, this study was conducted over six
years ago, collected measures of dust lead that required extensive conversions, was
limited to homes built prior to 1979, and involved only 284 homes. A detailed and
involved methodology was developed to update the numbers of homes to 1997,
convert the dust lead measures, and estimate environmental-lead levels in homes built
post-1979.
Third, estimating the impact of the proposed §403 rule required estimating the
distribution of blood-lead concentrations for children aged 1-2 expected to result from
example options for the §403 standards. Even if the distribution of environmental-
lead levels for example options for the proposed §403 standards could be determined,
6-67
-------
estimating the distribution of blood-lead concentrations associated with the post-§403
environmental-lead levels is a very complicated problem. There are many factors
other than the measured amount of lead in the child's home that contribute to a
child's blood-lead concentration: nutrition, activity patterns, and lead exposures at
day cares, schools, and at play areas outside of the home. More factors are listed in
Section 4.1.2. Predicting a blood-lead concentration distribution associated with a
specific set of environmental exposures is a difficult problem. Predicting the national
distribution of blood-lead concentrations across a wide range of environmental
exposures for children aged 1-2 years is an order of magnitude more complex.
Two different types of models were used to predict blood-lead concentrations: EPA's
IEUBK model and an empirical model developed for this study. The IEUBK model
has been studied extensively, has been utilized at a wide number of sites, and has
undergone peer review by EPA's Science Advisory Board. However, the application
of the IEUBK model in this study differs from those it was developed for. The
empirical model was developed specifically for this study based on the data collected
in a single study (Lanphear et al, 1995). It has not undergone peer review, has not
been applied elsewhere, and has not been studied in much depth. The two models
function and behave very differently, and that is why two different models were used.
A detailed and involved methodology was developed to predict the distribution of
blood-lead concentrations for children aged 1-2 years associated with distributions of
environmental-lead levels expected to result for various example options for the
proposed §403 standards. It is essential that we recognize that the predicted post-
§403 blood-lead distributions may not be very accurate, are not very robust, and
should not be used as indicators of what will happen in the future following
promulgation of the S403 standards for lead in paint, dust, and soil. On the other
hand, the predicted post-§403 blood-lead distributions are useful for making relative
comparisons among example options for the §403 standards.
Fourth, characterizing health benefits associated with the reduction of lead-based
paint hazards under various example options for the proposed §403 standards
required estimation of health effects from blood-lead concentrations. Seven health
effect and blood-lead concentration endpoints were used to characterize the health
benefits associated with various example options for the proposed §403 standards for
lead in paint dust and soil. The prediction of health effects related to IQ scores from
blood-lead distributions is based on the best available data and tools. Nevertheless, as
stated above for blood-lead concentration, predicted post-§403 health effect and
blood-lead concentration endpoints are meant to be used only for making relative
comparisons between example options for the §403 standards.
Tables, developed from these four analyses, that predict the health effect and blood-lead
concentration endpoints for children aged 1-2 years in the year 1997 following proposal of the
§403 rule for example standards are presented in this chapter. The primary conclusion from
these analyses is that health benefits tend to be most sensitive, and numbers of affected housing
6-68
-------
units least sensitive to changes in the example standards on dust-lead loadings and soil-lead
concentration when these example standards are at the upper end of the ranges considered. At
the lower end of the ranges of the example dust and soil standards, health benefits are less
sensitive, while the numbers of affected housing units are highly sensitive.
For example, consider again the plots displayed in Figure 6-1 la for the various example
standards. The percentage of children aged 1-2 years predicted to have a blood-lead concentration
greater than or equal to 10 jag/dL, based on the IEUBK model predictions, following
promulgation of the example option labeled as point A (3.9%) is twice as large as that for the
example option labeled as point C (1.7%). However, the percentages of homes affected by the
two example options (17.5% and 21.8%) are similar. On the other hand, the percentage of
children aged 1-2 years predicted to have a blood-lead concentration greater than or equal to 10
ug/dL following promulgation of the example option E (0.84%) is very similar to that predicted
for the example option F (0.25%) even though the percentages of homes affected by the two
example options are substantially different (38.4% and 53.7%).
A secondary conclusion of the analyses is that there are relatively small differences in the
selected endpoints and percentages of homes affected among the example options considered for
the paint intervention trigger levels. However, this conclusion must be interpreted with caution,
as the available data on deteriorated lead-based paint hi the nation's housing stock were
considered very limited, and the models are limited in their ability to handle paint as a predictor
variable. Because there is not sufficient data and information to perform a quantitative analysis
of example options for the paint intervention triggers, it may be best to only qualitatively
evaluate these options.
When comparing values of the health effect and blood-lead concentration endpoints
between baseline (pre-§403) and post-§403 (under the example standards studied) conditions, the
largest differences occurred for the percentages of children with blood-lead concentration at or
above 10 or 20 ug/dL and the percentages of children with IO decrement of greater than or equal
to 2 or 3. Smaller declines from baseline were observed in the percentage of children with IQ
score less than 70, the percentage of children with IQ decrement of greater than 1, and average IQ
decrement in a child. Across all endpoints, larger differences from baseline were observed under
the IEUBK model than under the empirical model.
The major limitation associated with how example options for environmental-lead
standards were investigated hi this chapter is the limited amount of data available for estimating
pre- and post-§403 environmental-lead levels. This includes a lack of nationally-representative
dust-lead loading data (representing both pre- and post-§403 conditions) where samples were
collected by wipe techniques. This data limitation constitutes one of the major data gaps and
limitations for the risk management analyses. To help alleviate this limitation, sensitivity
analyses were conducted to examine the impact of changes in post-intervention environmental-
lead levels on risk reductions and on determining wipe-equivalent dust-lead loadings for
comparisons to example standards and for determining a post-intervention blood-lead
concentration distribution. Two conclusions from the sensitivity analysis on dust-lead loading
data were as follows:
6-69
-------
Estimated numbers of housing units in which dust cleanings are triggered based on
pre-intervention dust-lead loadings may be biased by as many as one million units in
either direction due to necessary conversions of these loadings to wipe equivalents.
Deviating the assumptions on post-intervention (wipe) dust-lead loadings (40 |ig/ft2
for floors and 100 ug/ft2 for window sills) most notably affected estimated endpoints
representing extreme effects (i.e., high blood-lead levels, large IQ decrements).
One component of the sensitivity analysis examined an alternative approach to estimating
health effect and blood-lead concentration endpoints which does not require specifying post-
intervention environmental-lead levels (Section 6.4.5). This approach narrowed the extent to
which the IEUBK and empirical models differed hi then: estimates of post-intervention health
effect and blood-lead concentration endpoints.
The analyses of various example options for the §403 standards clearly indicates that the
risks to children's health associated with exposures to lead hi paint, dust, and soil can be reduced.
The standards established by the proposed §403 rule (once defined) will help reduce the health
risks to our nation's children. Depending on the methodology utilized, an illustrative example
for the §403 standards (floor dust-lead loading of 100 fig/ft2, window sill dust-lead loading of
500 ug/ft2, soil-lead concentration of 2000 ug/g, paint maintenance warranted at 5 ft2 deteriorated
LBP, and paint abatement at 20 ft2 deteriorated LBP) indicates that the percentage of children
aged 1-2 years with a blood-lead concentration at or above 10 ug/dL ranged from 1.83 to 4.85 %
compared to the current baseline estimate of 5.75%. This corresponds to approximately 70 to
300 thousand fewer children aged 1-2 years old with a blood-lead concentration at or above 10
Hg/dL. Reductions in other health measures would also be achieved, hi addition, reductions in
health measures would also be achieved for children of other age groups.
6-70
-------
7.0 REFERENCES
Adebonojo, P.O. (1974) "Hematologic Status of Urban Black Children in Philadelphia:
Emphasis on the Frequency of Anemia and Elevated Blood Lead Levels." Cli Pediatr. 13:874-
Alessio, L., Bertazzi, P.A., Monelli, O., et al. (1976) "Free Erythrocyte PTotoporphyrin as an
Indicator of the Biological Effect of Lead in Adult Males: n. Comparison Between Free
Erythrocyte Protoporphyrin and Other Indicators of Effect." International Archives of
Occupational and Environmental Health. 37:89-105.
Alexander, F.W., et al. (1973) "The Uptake and Excretion by Children of Lead and Other
Contaminants." In: Barth, D., et al., eds. Environmental Health Aspects of Lead; Proceedings,
International Symposium, Oct. 1972. Amsterdam, The Netherlands. Luxembourg: Commission
of the European Communists. 319-331.
Alkondon, M., et al. (1990) "Selective Blockade of NMDA-activated Channel Currents May Be
Implicated hi Learning Deficits Caused by Lead." FEES Lett. 261:124-130.
Alomran, A.H. and Shleamoon, M.N. (1988) "The Influence of Chronic Lead Exposure on
Lymphocyte Proliferative Response and Immunoglobulin Levels in Storage Battery Workers."
Journal of Biological Science Research. 19:575-585.
Alvares, A.P., Kapelner, S., Sassa, S., et al. (1975) "Drug Metabolism hi Normal Children, Lead-
Poisoned Children, and Normal Adults." Clin Pharmacol Ther. 17:179-183.
Amitai, Y., Brown, M.J., Graef, J.W., Cosgrove, E. (1991) "Residential Deleading: Effects on
the Blood Lead Levels of Lead-Poisoned Children." Pediatrics. 88(5):893-897.
Angle, C.R. and Mclntire, M.S., Colucci, A.V. (1974) "Lead hi Air, Dustfall, Soil, House Dust,
Milk and Water: Correlation with Blood Lead of Urban and Suburban School Children." In:
Trace Substances in Environmental Health - Vm, D. D. Hemphill, Ed., 23-29.
Angle, C.R. and Mclntire, M.S. (1978) "Low Level Lead and Inhibition of Erythrocyte
Pyrimidine Nucleotidase." Environmental Research. 17:296-302.
Angle, C.R., Mclntire, M.S. (1979) "Environmental Lead and Children: The Omaha Study."
Journal ofToxicological and Environmental Health. 5:855-870.
Angle, C.R., Mclntire, M.S., Swanson, M.S., et al. (1982) "Erythrocyte Nucleotides in Children-
Increased Blood Lead and Cytidine Triphosphate." Pediatr Res. 16:331-334.
Angle, C.R., Marcus, A., Cheng, I-H., et al. (1984) "Omaha Childhood Blood Lead and
Environmental Lead: A Linear Total Exposure Model." Environmental Research. 35:160-170.
7-1
-------
Angle, C.R. and Kuntzelman, D.R. (1989) "Increased Erythrocyte Protoporphyrins and Blood
Lead A Pilot Study of Childhood Growth Patterns." J Toxicol Environ Health. 26:149-156.
Annest, J.L. (1983) "Trends in the Blood-Lead Levels of the U.S. Population: The Second
National Health and Nutrition Examination Survey (NHANES JJ) 1976-1980." In: Lead Versus
Health: Sources and Effects of Low Level Lead Exposure, Rutter, M., Russell Jones, R., eds.
New York: John Wiley and Sons, 33-58.
Anttila, A., Heikkila, P., Pukkala, E., Nykyri, E., Kauppinen, T., Hernberg, S., et al., (1995)
"Excess Lung Cancer Among Workers Exposed to Lead" Scandinavian Journal of Work
Environment and Health. 21:460-469.
Anttila, A., Heikkila, P., Nykyri, E., Kauppinen, T., Pukkala, E., Hernberg, S., and Hemminki, K.
(1996) "Risk of Nervous System Cancer Among Workers Exposed to Lead." Journal of
Occupational and Environmental Medicine. 38:131-136.
Araki, S., Honma, T., Yanagihara, S., et al. (1980) "Recovery of Slowed Nerve Conduction
Velocity in Lead-Exposed Workers." 46:151-157.
Araki, S., Murata, K., Yokoyama, K., and Uchida, E. (1992) "Auditory Event-related Potential
(P300) in Relation to Peripheral Nerve Conduction in Workers Exposed to Lead, Zinc, and
Copper: Effects of Lead on Cognitive Function and Central Nervous System." American
Journal of Industrial Medicine. 21:539-547.
Arnvig, E., Grandjean, P., Beckmann, J. (1980) "Neurotoxic Effects of Heavy Lead Exposure
Determined with Psychological Tests." Toxicol Lett. 5:399-404.
Aschengrau, A., Beiser, A., Bellinger, D., Copenhafer, D., Weitzman, M. (1994) "The Impact of
Soil Lead Abatement on Urban Children's Blood Lead Levels: Phase n Results from the Boston
Lead-in-Soil Demonstration Project" Environmental Research. 67:125-148.
Assennato, G., Baser, M., Molinini, R., et al. (1987) "Sperm Count Suppression Without
Endocrine Dysfunction in Lead-Exposed Men." Archives of Environmental Health. 42:124-127.
ATSDR (1988a) "The Nature and Extent of Lead Poisoning in Children hi the United States: A
Report to Congress." U.S. Department of Health and Human Services, Public Health Service,
Agency for Toxic Substances and Disease Registry.
ATSDR (1988b) "The Silver Creek Mine Tailings Exposure Study, Park City, Utah." Final
Report of the Agency for Toxic Substances and Disease Registry, Public Health Service, U.S.
Department of Health and Human Services, June 1988.
ATSDR (199 la) "Child Lead Exposure Study, Leeds, Alabama." Final Report of the Agency for
Toxic Substances and Disease Registry, Public Health Service, U.S. Department of Health and
Human Services, September 1991.
7-2
-------
ATSDR (1991b) "Philadelphia Neighborhood Lead Study." Final Report of the Agency for
Toxic Substances and Disease Registry, Public Health Service, U.S. Department of Health and
Human Services, September 1991.
ATSDR (1992) "Clear Creek/Central City Mine Waste Exposure Study, Part I: Smuggler
Mountain Site." Final Report of the Agency for Toxic Substances and Disease Registry, Public
Health Service, U.S. Department of Health and Human Services, September 1992.
ATSDR (1993) "Toxicological Profile for Lead." Final Report of the Agency for Toxic
Substances and Disease Registry, Public Health Service, U.S. Department of Health and Human
Services, April 1993.
Audesirk, G. (1985) "Effects of Lead Exposure on the Physiology of Neurons." Prog.
Neurobiol 24:199-231.
Awad El Karim, M.A., Hamed, A.S., Elhaimi, Y.A.A., et al. (1986) "Effects of Exposure to Lead
among Lead-Acid Battery Factory Workers in Sudan." Archives of Environmental Health.
41:261-265.
Baghurst, P.A., Robertson, E.F., McMichael, A.J., et al. (1987) "The Port Pirie Cohort Study:
Lead Effects on Pregnancy Outcome and Early Childhood Development." Neurotocicology.
8:395-401.
Baghurst, P.A., McMichael, A.J., Wigg, N.R., Vimpani, G., Robertson, E.F., Roberts, R.J., Tong,
S. (1992) "Environmental Exposure to Lead and Children's Intelligence at the Age of Seven
Years: The Port Pirie Cohort Study." New England Journal of Medicine. 327:1279-1284.
Baker, EX. Jr., Landrigan, P.J., Barbour, A.G., et al. (1979) "Occupational Lead Poisoning in the
United States: Clinical and Biochemical Findings Related to Blood Lead Levels." British
Journal of Industrial Medicine. 36:314-322.
Baker, E.L., Feldman, R.G., White, R.F., et al. (1983) "The Role of Occupational Lead Exposure
in the Genesis of Psychiatric and Behavioral Disturbances." Acta Phsychiatr Scand Suppl.
67:38-48.
Baloh, R.W., Spivey, G.H., Brown, C.P., et al. (1979) "Subclinical Effects of Chronic Increased
Lead Absorption A Prospective Study: n. Results of Baseline Neurologic Testing." Journal
of Occupational Medicine. 21:490-496.
Barry, P.S.I, and Mossman, D.B. (1970) "Lead Concentration hi Human Tissues." British
Journal of Industrial Medicine. 27:339-351.
Barry, P.S.I. (1975) "A Comparison of Concentrations of Lead in Human Tissues." British
Journal of Industrial Medicine. 32:119-139.
7-3
-------
Barry, P.S.I. (1981) "Concentrations of Lead in the Tissues of Children." British Journal of
Industrial Medicine. 38:61-71.
Batschelet, E., Brand, L., Steiner, A. (1979) "On the Kinetics of Lead in the Human Body."
J. Math Biology. 8:15-23.
Batuman, V., Maesaka, J.K., Haddad, B., et al. (1981) "The Role of Lead in Gout Nephropathy."
New England Journal of Medicine. 304:520-523.
Batuman, V., Landy, E., Maesaka, J.K., et al. (1983) "Contribution of Lead to Hypertension with
Renal Impairment." New England Journal of Medicine. 309:17-21.
Bauchinger, M., Dresp, J., Schmid, E., et al. (1977) "Chromosome Analyses of Children After
Ecological Lead Exposure." Mut. Res. 56:75-79.
Bellinger, D.C. and Needleman, H.L. (1983) "Lead and the Relationship Between Maternal and
Child Intelligence." Journal of Pediatrics. 102:523-527.
Bellinger, D.C., Needleman, H.L., Leviton, A., et al. (1984) "Early Sensory-Motor Development
and Prenatal Exposure to Lead." Neurobehav Toxicol Teratol. 6:387-402.
Bellinger, B.C., Leviton, A., Waternaux, C., et al. (1985a) "A Longitudinal Study of the
Developmental Toxicity of Low-Level Lead Exposure in the Prenatal and Early Postnatal
Periods." hi: Lekkas TD, ed. International Conference on Heavy Metals hi the Environment,
Athens, Greece, September, Vol 1. Edinburgh, United Kingdom: CEP Consultants, Ltd, 32-34.
Bellinger, D.C., Leviton, A., Waternaux, C., et al. (1985b) "Methodological Issues in Modeling
the Relationship Between Low-Level Lead Exposure and Infant Development: Examples from
the Boston Lead Study." Environmental Research. 38:119-129.
Bellinger, D.C., Leviton, A., Needleman, H.L., et al. (1986a) "Low-Level Lead Exposure and
Infant Development in the First Year." Neurobehav Toxicol Teratol. 8:151-161.
Bellinger, D.C., Leviton, A., Rabinowitz, M., et al. (1986b) "Correlates of Low-Level Lead
Exposure in Urban Children at Two Years of Age." Pediatrics. 77:826-833.
Bellinger, D., Needleman, H.L., Bromfield, R., Mikntz, M. (1986c) "A Follow-up Study of the
Academic Attainment and Classroom Behavior of Children with Elevated Dentine Lead Levels."
Bio Trace Elem Res. 6:207-23.
Bellinger, D.C., Leviton, A., Waternaux, C., et al. (1987a) "longitudinal Analyses of Prenatal
and Postnatal Lead Exposure and Early Cognitive Development" New England Journal of
Medicine. 316:1037-1043.
7-4
-------
Bellinger, D., Sloman, J., Leviton, A., et al. (1987b) "Low Level Lead Exposure and Child
Development: Assessment at Age 5 of a Cohort Followed from Birth." In: Lindberg, S.E.,
Hutchinson, T.C., eds. International Conference on Heavy Metals in the Environment. New
Orleans, LA, September, Vol. 1. Edinburgh, UK: CEP Consultants, Ltd., 49-53.
Bellinger, D., Leviton, A., Waternaux, C., et al. (1989a) "Low-Level Lead Exposure and Early
Development in Socioeconomically Advantaged Urban Infants." In: Smith M, Grant L.D. Sors,
A., eds. Lead Exposure and Child Development: An International Assessment. Lancaster, UK:
Kluwer Academic Publishers.
Bellinger, D.C., Leviton, A., Watemaux, C., et al. (1989b) "Low-level Lead Exposure, Social
Class, and Infant Development." Neurotoxicol Teratol. 10:497-504.
Bellinger, D., Sloman, J., Leviton, A., Rabinowitz, M., Needleman H.L., Watemaux, C. (1991)
"Low-level Lead Exposure and Children's Cognitive Function hi the Preschool Years."
Pediatrics. 87(2):219-227.
Bellinger, D., Stiles, K.M., Needleman, H. (1992) "Low-Level Lead Exposure, Intelligence, and
Academic Achievement: A Long-term Follow-up Study." Pediatrics. 90(6):855-861.
Bellinger, D.C. (1995) "Interpreting the Literature on Lead and Child Development: The
Neglected Role of the 'Experimental System'." Neuratoxicology and Teratology. 17(3):201-212.
Bergomi, M., Borella, P., Fantuzzi, G., et al. (1989) "Relationship Between Lead Exposure
Indicators and Neuropsychological Performance hi Children." Dev Med Child Neurol 31:181-
190.
Bernard, B.P., Becker, C.E. (1988) "Environmental Lead Exposure and the Kidney." Clin
Toxicol 26:1-34.
Bert, J.L., van Dusen, L.J., Grace, J.R. (1989) "A Generalized Model for the Prediction of Lead
Body Burdens." Environmental Research. 48:17-127.
Belts, P.R., Astley, R., Raine, D.N. (1973) "Lead Intoxication in Children in Birmingham."
British Medical Journal 1:402-406.
Biagini, G., Caudarella, R., Vangelista, A. (1977) "Renal Morphological and Functional
Modification in Chronic Lead Poisoning." In: Brown, S.S., ed. Clinical Chemistry and Chemical
Toxicology of Metals. Elsevier/North-Holland Biomedical Press, 123-126.
Bomschein, R.L., Hammond, P.B., Dietrich, K.N., Succop, P.A., Krafft, K.M., Clark, C.S.,
Pearson, D., Que Hee, S. (1985a) "The Cincinnati Prospective Study of Low-Level Lead
Exposure and Its Effects on Child Development: Protocol and Status Report." Environmental
Research 38:4-18.
7-5
-------
Bornschein, R.L., Succop, P.A., Dietrich, R.N., Clark, C.S., Que Hee, S., Hammond, P.B.
(1985b) "The Influence of Social and Environmental Factors on Dust Lead, Hand Lead, and
Blood Lead Levels in Young Children." Environmental Research 38:108-118.
Bornschein, R.L., Succop, P.A., Krafft, K.M., Clark, C.S., Peace, B., Hammond, P.B. (1986).
"Exterior Surface Dust Lead, Interior House Dust Lead and Childhood Lead Exposure in an
Urban Environment." In: Trace Substances in Environmental Health n, 1986, A Symposium,
D.D. Hemphill, ed. Columbia, MO: University of Missouri.
Bornschein, R.L., Grote, J., Mitchell, T., et al. (1989) "Effects of Prenatal Lead Exposure on
Infant Size at Birth." In: Smith, M., Grand, L.D., Sors, A., eds. Lead Exposure and Child
Development: An International Assessment. Lancaster, UK: Kluwer Academic Publishers.
Bornschein, R., Clark, S., Pan, W., Succop, P. (1990) "Midvale Community Lead Study." Final
Report of the University of Cincinnati Medical Center, July 1990.
Bornschein, R. (1996a) "Palmerton Lead Exposure Study." Final report of the University of
Cincinnati.
Bornschein, R. (1996b) "Bingham Creek Environmental and Human Health Lead and Arsenic
Study." Final report of the University of Cincinnati.
Bornschein, R. (1997) "Leadville/Lake County Environmental Health Study." Final report of
the University of Cincinnati.
Boudene, C., Malet, D. and Musse, R. (1977) "Fate of210 Pb Inhaled by Rats." Toxicol. Appl
Pharmacol 41:271-276.
Bradley, I.E., Powell, A.E., Niermann, W., et al. (1956) "The Incidence of Abnormal Blood
Levels of Lead in a Metropolitan Pediatric Clinic: With Observation on the Value of
Coproporphyrinuria as a Screening Test." Journal of Pediatrics. 49:1-6.
Bradley, J.E. and Baumgartner, R.J. (1958) "Subsequent Mental Development of Children with
Lead Encephalopathy, as Related to Type of Treatment." Journal of Pediatrics. 53:311-315.
Braunstein, G.D., Dahlgren, J., Loriaux, D.L. (1978) "Hypogonadism in Chronically Lead-
poisoned Men." Infertility. 1:33-51.
Bressler, J.P. and Goldstein, G.W. (1991) "Mechanisms of Lead Neurotoxicity." Biochem.
Pharmacol. 41:479-484.
7-6
-------
Brody, D. J., Pirkle, J.L., Kramer, R. A., Flegal, K.M., Matte, T. D., Gunter, E.W., Paschal, D.C.
(1994) "Blood Lead Levels in the US Population: Phase I of the Third National Health and
Nutritional Examination Survey (NHANES HI, 1988 to 1991)," Journal of the American Medical
Association. 272(4):277-283.
Buc, H.A. and Kaplan, J.C. (1978) "Red-cell Pyrimidine S'-nucleotidase and Lead Poisoning."
Clin ChimActa. 87:49-55.
Buchet, J.P., Roels, H., Bernard, A., et al. (1980) "Assessment of Renal Function of Workers
Exposed to Inorganic Lead, Cadmium, or Mercur Vabor. Journal of Occupational Medicine.
22:741-750.
Bureau of the Census, HUD (1995). American Housing Survey for the United States in 1993.
Current Housing Reports, No. H150/93. Bureau of the Census, U.S. Department of Commerce;
and the Office of Policy Development and Research, U.S. Department of Housing and Urban
Development, February, 1995.
Bureau of the Census, HUD (1996). Housing Completions, December 1995/January 1996.
C22/96-1, March 1996.
Butte-Silver Bow Department of Health, and Department of Environmental Health, University of
Cincinnati. (1991) "The Butte-Silver Bow Environmental Health Lead Study." Draft Final
Report, 10 June 1991.
Campara, P., D'Andrea, F., Micciolo, R., et al. (1984) "Psychological Performance of Workers
With Blood-lead Concentration Below the Current Threshold Limit Value." International
Archives of Occupational Environmental Health. 53:233-246.
Carroll, R.J., Ruppert, D. and Stefanski, L.A. (1995) Measurement Error in Nonlinear Models.
New York, Chapman & Hall.
Centers for Disease Control (1985) "Preventing Lead Poisoning in Young Children." Altanta,
GA: U.S. Department of Health and Human Services, Public Health Service, Centers for Disease
Control. Publication No. 99-2230,7-19.
Centers for Disease Control (1991) "Preventing Lead Poisoning hi Young Children: A Statement
by the Centers for Disease Control." Public Health Service, Centers for Disease Control and
Prevention, U.S. Department of Health and Human Services, October 1991.
Centers for Disease Control (1992) "Sample Design: Third National Health and Nutrition
Examination Survey. Series 2: Data Evaluation and Methods Research." National Center for
Health Statistics, Centers for Disease Control and Prevention, U.S. Department of Health and
Human Services, DHHS Publication No. (PHS) 92-1887.
7-7
-------
Centers for Disease Control (1994) "Plan and Operation of the Third National Health and
Nutrition Examination Survey, 1988-94." National Center for Health Statistics, Centers for
Disease Control and Prevention, U.S. Department of Health and Human Services, DHHS
Publication No. (PHS) 94-1308.
Centers for Disease Control (1997) "Update: Blood Lead Levels - United States, 1991-1994."
U.S. Department of Health and Human Services, Morbidity and Mortality Weekly Report
46(7):141-146.
Chamberlain, A.C., et. al. (1978) "Investigations into Lead from Motor Vehicles." Hartwell,
United Kingdom: United Kingdom Atomic Energy Authority; Report No. AERE-R9198.
Charney, E., Kessler, B., Farfel, M., Jackson, D. (1983) "Childhood Lead Poisoning: A
Controlled Trial of the Effect of Dust-Control Measures on Blood Lead Levels." New England
Journal of Medicine. 309:1089-1093.
Chia, S-E, Chia, K-S, Chia, H-P, Ong, C-N, and Jeyaratnam, J. (1996) "Three Year Follow-up of
Serial Nerve Conduction among Lead-exposed Workers." Scandinavian Journal of Work
Environment and Health. 22:374-380.
Chisolm, J.J., Jr. and Harrison, H.E. (1956) "The Exposure of Children to Lead." Pediatrics.
18:943-958.
Chisolm, J.J., Jr. (1962) "Aminoaciduria as a Manifestation of Renal Tubular Injury in Lead
Intoxication and a Comparison with Patterns of Aminoaciduria Seen in Other Diseases." Journal
of Pediatrics. 60:1-17.
Chisolm, J.J., Jr. (1965) "Chronic Lead Intoxication in Children." Dev Med Child Neurol. 7:529-
536.
Chisolm, J.J., Thomas, D.J., Hammill, T.G. (1985) "Erythrocyte Porphobilinogen Synthase
Activity as an Indicator of Lead Exposure to Children." Clin Chem. 31:601-605.
Chowdhury, A.R., Chinoy, N.J., Gautam, A.K., et al. (1986) "Effect of Lead on Human Semen."
Adv Contracept Deliv Syst. 2:208-211.
Christoffersson, J.O., Ahlgren, L., Schutz, A., Skerfving, S., Mattsson, S. (1986) "Decrease of
Skeletal Lead Levels in Man after End of Occupational Exposure." Archives of Environmental
Health. 41:5:312-318.
Coate, D. and Fowles, R. (1989) "Is There Statistical Evidence for a Blood Lead-Blood Pressure
Relationship?" Journal of Economics. 8:173-184.
Cocco, P., Fu, H., Boffetta, P., Carta, P., Flore, C., Flore, V., et al. (1997) "Mortality of Italian
Lead Smelter Workers." Scandinavian Journal of Work Environment and Health. 23(l):15-23.
7-8
-------
Cohen, N., Jaakkola, T., Wrenn, M.E. (1973) "Lead-210 Concentrations in the Bone, Blood, and
Excreta of a Former Uranium Miner." Health Physics. 24:601-609.
Colorado Department of Health, University of Colorado at Denver, and ATSDR (1990)
"Leadville Metals Exposure Study." Final Report by the Colorado Department of Health,
Division of Disease Control and Environmental Epidemiology; University of Colorado at
Denver, Center for Environmental Sciences; and the Agency for Toxic Substances and Disease
Registry, Public Health Service, U.S. Department of Health and Human Services, April, 1990.
Cooney, G.H., Bell, A., McBride, W., et al. (1989) "Low-Level Exposures to Lead: The Sydney
Lead Study." Dev MedCMdNeuroL 31:640-649.
Cooney, G., Bell, A., Stavron, C. (1991) "Low Level Exposures to Lead and Neurobehavioural
Development: The Sydney Study at Seven Years." Heavy Metals in the Environment.
Edinburgh: CEP Consultants, 1991:16-9.
Cooper, W.C. (1981) "Mortality in Employees of Lead Production Facilities and lead Battery
Plants, 1971-1975." In Lynam, D.R., et al. eds. Environmental Lead: Proceedings of the
Second International Symposium on Environmental Lead Research, December 1975, Cincinnati,
OH. New York, NY: Academic Press, 111-143.
Cooper, C., Wong, O., and Kheifets, L. (1985) "Mortality Among Employees of Lead Battery
Plants and Lead-Producing Plants, 1947-1980." Scandinavian Journal of Work Environment and
Health. 11:331-345.
Cooper, W.C. and Gaffey, W.R. (1975) "Mortality of Lead Workers." Journal of Occupational
Medicine. 17:100-107.
Cooper, W.C. (1976) "Cancer Mortality Patterns in the Lead Industry." Ann, NYAcadSci.
271:250-259.
Cooper, W.C., Wong., O., Kheifets., L. (1985) "Mortality Among Employees of Lead Battery
Plants and Lead Producing Plants, 1947-1980." Scandinavian Journal of Work Environment and
Health.ll:331-345.
Cooper, W.C. (1988) "Deaths from Chronic Renal Disease in U.S. Battery and Lead Production
Workers." Environmental Health Perspectives. 78:61-63.
Copley, C.G. (1983) "The Effect of Lead Hazard Source Abatement and Clinic Appointment
Compliance on the Mean Decrease of Blood Lead and Zinc Protoporphyrin Levels. Mimeo."
City of St. Louis, Department of Health and Hospitals, Division of Health, Office of the Health
Commissioner, St. Louis, MO.
7-9
-------
Cramer, K., Goyer, R.A., Jagenburg, R., et al. (1974) "Renal Ultrastructure, Renal Function, and
Parameters of Lead Toxicity in Workers with Different Periods of Lead Exposure." British
Journal of Industrial Medicine. 31:113-127.
Cristy, M., Leggett, R.W., Dunning, D.E., Jr., Eckerman, K.F. (1986) "Relative Age-Specific
Radiation Dose Commitment Factors for Major Radionuclides Released from Nuclear Fuel
Facilities." NUREG/CR-4628 ORNL/TM-9890.
Cullen, M.R., Kayne, R.D., Robins, J.M. (1984) "Endocrine and Reproductive Dysfunction in
Men Associated with Occupational Inorganic Lead Intoxication." Archives of Environmental
Health. 39:431-440.
Dalpra, L., Tibiletti, M.G., Nocera, G., et al. (1983) "SCE Analysis in Children Exposed to Lead
Emission from a Smelting Plant." MutRes. 120:249-256.
Davies, D.J., Watt, J.M., Thornton, I. (1987) "Lead Levels in Birmingham Dusts and Soils." The
Science of the Total Environment. 87(2-3):177-185.
Davies, D.J.A, Thornton, I., Watt, J.M., Culbard, E.B., Harvey, P.O., Delves, H.T., Sherlock,
J.C., Smart, G.A., Thomas, J.F.A., Quinn, M.J. (1990) "Lead Intake and Blood Lead hi Two-
Year-Old U.K. Urban Children." The Science of the Total Environment. 90:13-29.
Davis, J.M. and Svendsgaard, D.J. (1987) "Lead and Child Development." Nature. 329:297-300.
Day, J.C. (1993) Population Projections of the United States, by Age, Sex, Race, and Hispanic
Origin: 1993 to 2050. U.S. Bureau of the Census, Current Population Reports P25-1104.
de la Burde, B. and Choate, M.S., Jr. (1972) "Does Asymptomatic Lead Exposure hi Children
Have Latent Sequelae?" Journal of Pediatrics. 81:1088-1091.
de la Burde, B. and Reames, B. (1973) "Prevention of Pica, The Major Cause of Lead Poisoning
in Children." American Journal of Public Health. 63(8):737-743.
de la Burde, B. and Choate, M.S., Jr. (1975) "Early Asymptomatic Lead Exposure and
Development at School Age." Journal of Pediatrics. 87:638-642.
deKort, W.L.A.M., Verschoor, M.A., Wibowo, A.A.E., et al. (1987) "Occupational Exposure to
Lead and Blood Pressure: A Study of 105 Workers." American Journal of Industrial Medicine.
11:145-156.
Dersimonian, R. and Laird, N. (1986) "Meta-Analysis hi Clinical Trials." Controlled Clinical
Trials. 7:177-188.
7-10
-------
Dietrich, K.N., Krafft, K.M., Bier, M, et al. (1986) "Early Effects of Fetal Lead Exposure:
Neurobehavioral Findings at 6 Months." InternationalJownal of Biosocial and Medical
Record. 8:151-168.
Dietrich, K.N., Krafft, K.M., Bornschein, R.L., et al. (1987a) "Low-Level Fetal Lead Exposure
Effect on Neurobehavioral Development in Early Infancy." Pediatrics. 80:721-730.
Dietrich, K.N., Krafft, K.M., Shukla, R., et al. (1987b) "The Neurobehavioral Effects of Early
Lead Exposure." Monogr Am Assoc Ment Defic. 8:71-95.
Dietrich, K.N., Berger, O.G., Succop, P.A., Hammond, P.B., Bornschein, I. (1993) "The
Developmental Consequences of Low To Moderate Prenatal and Postnatal Lead Exposure:
Intellectual Attainment in the Cincinnati Lead Study Cohort Following School Entry."
Neurotoxicology and Teratology. 15:37-44.
Drasch, G.A., Kretschmer, E., Lochner, C. (1988) "Lead and Sudden Infant Death:
Investigations on Blood Samples of SID Babies." Eur JPediatr. 147:79-84.
Du Val, G.E. and Fowler, B.A. (1989) "Preliminary Purification and Characterization Studies of
a Low Molecular Weight High Affinity Cytosolic Lead Binding Protein in Rat Brain. Biochem.
Biophys. Res. Commun. 159:177-184.
Dyck, P.J., et al. (1980) "Blood Brain Barrier in Rat and Cellular Mechanisms of Lead-induced
Segmental Demyelination " J. Neuropathol. Exp. Neural. 39:700-709.
Efron, Bradley and Tibshirani, Robert J. (1993) "An Introduction to the Bootstrap." Chapman &
Hall, New York.
Elwood., P.C., Davey-Smith, G., Oldham, P.D., et al. (1988) "Two Welsh Surveys of Blood Lead
and Blood Pressure." Environmental Health Perspectives. 78:119-121.
Emond, M.J., Lanphear, B.P., et al. (1997). "Measurement Error and Its Impact on the Estimated
Relationship between Dust Lead and Children's Blood Lead." Environmental Research. 72:82-
92.
Erenberg, G., Rinsler, S.S., Fish, B.G. (1974) "Lead Neuropathy and Sickle Cell Disease."
Pediatrics. 54:438-441.
Ernhart, C.B., Landa, B., Schell, N.B. (1981) "Subclinical Levels of Lead and Developmental
Deficit A Multivariate Follow-up Reassessment." Pediatrics. 67:911-919.
Ernhart, C.B., Wolf, A.W., Kennard, M.J., et al. (1985) "Intrauterine Lead Exposure and the
Status of the Neonate." In: Lekkas TD, ed. International Conference on Heavy Metals in the
Environment, Athens, Greece, September. Vol. 1. Edinburgh, United Kingdom: CEP
Consultants, Ltd. 35-37.
7-11
-------
Ernhart, C.B., Wolf, A.W., Kennard, M.J., et al. (1986) "Intrauterine Exposure to Low Levels of
Lead: The Status of the Neonate." Archives of Environmental Health. 41:287-291.
Ernhart, C.B., Morrow-Tlucak, M., Marler, M.R., et al. (1987) "Low Level Lead Exposure in the
Prenatal and Early Preschool Periods: Early Preschool Development." Neurotoxicol Teratol .
9:259-270.
Emhart, C.B., Morrow-Tlucak, M., Worf, A.W., Super, D., Drotar, D. (1989) "Low Level Lead
Exposure in the Prenatal and Early Preschool Periods; Intelligence Prior to School Entry."
Neurotoxicology and Teratology. 11:161 -70.
Ernhart, C.B., and Greene, T. (1990) "Low-Level Lead Exposure in Prenatal and Early Preschool
Periods: Language Development." Archives of Environmental Health. 45:342-354.
Ewers, U. and Erbe, R. (1980) "Effects of Lead, Cadmium, and Mercury on Brain Adenylate
Cyclase." Toxiclogy. 16:227-237.
Ewers, U., Stiller-Winkler, R., Idel, H. (1982) "Serum Immunoglobulin, Complement C3, and
Salivary IgA Level in Lead Workers." Environmental Research. 29:351-357.
Factor-Litvak, P., Graziano, J.H., Kline, J.K., et al. (1991) "A Prospective Study of Birthweight
and Length of Gestation in Population Surrounding a Lead Smelter in Kosovo, Yugoslavia." Int
JEpidemiol 20:772-728.
Fanning, D. (1988) "A Mortality Study of Lead Workers, 1926-1985." Archives of
Environmental Health. 43:247-251.
Farfel, M.R. and Chisolm, J.J., Jr. (1990) "Health and Environmental Outcomes of Traditional
and Modified Practices for Abatement of Residential Lead-Based Paint." American Journal of
Public Health 80(10):1240-1245.
Farfel, M.R., Las, P.S.J., Rhode, C.A., Lim, B.S., and Bannon, D. (1994) "Comparison of Wipe
and Cyclone Methods for the Determination of Lead in Residential Dusts," Applied Occupational
and Environmental Hygiene, 9:1006-1012.
Farfel, M.R. and Lim, B.S. (1995) "The Lead Paint Abatement and Repair and Maintenance
Study in Baltimore." In: Lead in Paint, Soil, and Dust: Health Risks, Exposure Studies, Control
Measures, Measurement Methods, and Quality Assurance, ASTM STP 1226, Michael E. Beard
and S. D. Allen Iske, Eds. Philadelphia: American Society for Testing and Materials, 107-118.
Fayerweather, W.E., Karns, M.E., Nuwayhid, I.A., and Nelson, T.J. (1997) "Case-Control Study
of Cancer Risk in Tetraethyl Lead Manufacturing." American Journal of Industrial Medicine.
31:28-35.
7-12
-------
Fox, D.A. (1991) "Visual and Auditory System Alterations Following Developmental or Adult
Lead Exposure: A Critical Review." In Human Lead Exposure (Needleman,H.L., ed) 105-123,
CRC Press, Boca Raton, Florida.
Fuller, W.A. (1987) Measurement Error Models. New York, John Wiley.
Fullerton, P.M. (1966) "Chronic Peripheral Neuropathy Produced by Lead Poisoning in Guinea
Pigs." J. Neuropathol. Exp. Neurol. 25:214-236.
Fulton, M, Raab, G., Thomson, G., Laxen, D., Hunter, R., Hepburn, W. (1987) "Influence of
Blood Lead on the Ability and Attainment of Children in Edinburgh." Lancet. 1:1221-1226.
Gant, V.A. (1938) "Lead Poisoning." Industial Medicine. 7:679-699.
Gartside, P.S. (1988) "The Relationship of Blood Lead Levels and Blood Pressure hi NHANES:
n. Additional Getzcalculations." Environmental Health Perspectives. 78:31-34.
Gerhardsson, L., Lundstrom, N.G., Nordberg, G., and Wall, S. (1986) "Mortality and Lead
Exposure: A Retrospective Cohort Study of Swedish Smelter Workers." British Journal of
Industrial Medicine. 43:707-712.
Gerhardsson, L., Hagmar, L., Rylander, L., Skerfving, S. (1995) "Mortality and Cancer Incidence
Among Secondary Lead Smelter Workers." Occup Environ Med. 52:667-672.
Glickman, L., Valciukas, J.A., Lilis, R., et al. (1984) "Occupational Lead Exposure: Effects on
Saccadic Eye Movements." International Archives of Occupational Environmental Health.
54:115-125.
Goldstein, G. (1990) "Lead Poisoning and Brain Cell Function." Environ. Health Perspect.
89:91-94.
Goldstein, G. (1992) "Developmental Neurobiology of Lead Toxicity." In Human Lead
Exposure (Needleman, H. ed.) 125-135.
Goyer, R.A. (1993) "Lead Toxicity: Current Concerns." Environmental Health Perspectives.
100:177-187.
Grandjean, P. and Lintrup, J. (1978) "Erythrocyte-Zn-protoporphyrin as an Indicator of Lead
Exposure." ScandJClin Lab Invest. 38:669-675.
Grandjean, P. (1979) "Occupational Lead Exposure hi Denmark: Screening with the
Haematofluorometer." British Journal of Industrial Medicine. 36:52-58.
Grandjean, P., Wulf, H.C., Neibuhr, E. (1983) "Sister Chromatid Exchange in Response to
Variations in Occupational Lead Exposure." Environmental Research. 32:199-204.
7-13
-------
Grandjean, P., Hollnagel, H., Hedegaard, L., et al. (1989) "Blood Lead-Blood Pressure Relation:
Alcohol Intake and Hemoglobin as Confounders." Am JEpidemiol. 129:732-739.
Greene, T. and Emhart, C.B. (1991) "Prenatal and Preschool Age Lead Exposure: Relationship
with Size." Neurotoxicology and Teratology. 13:417-427.
Greenhalgh, J.K., et al. (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. 38-42.
Gross, S. (1981) "Human Oral and Inhalation Exposures to Lead: Summary of Kehoe Balance
Experiments." J. Toxicol Environ. Health. 8:333-377.
Gulson, B.L., Mahaffey, K.R., Mizon, K.J., Korsch, M.J., Cameron, M.A., Vimpani, G. (1995)
"Contribution of Tissue Lead to Blood Lead in Adult Female Subjects Based on Stable Lead
Isotope Methods." Journal of Lab and Clinical Medicine. 125:703.
Haas, T., Wieck, A.G., Schaller, K.H., et al. (1972) "The Usual Lead Load in New-born Infants
and Their Mothers." Zentralblattjur Bakteriologie [B]. 155:341-349. (German)
Haenninen, H., Hemberg, S., Mantere, P., et al. (1978) "Psychological Performance of Subjects
with Low Exposure to Lead " Journal of Occupational Medicine. 20:683-689.
Haenninen, H., Mantere, P., Hernberg, S., et al. (1979) "Subjective Symptoms in Low-Level
Exposure to Lead." Neurotoxicology. 1:333-347.
Hammond, P.B., Bomschein, R.L., Succop, P. (1985) "Dose-effect and Dose-response
Relationships of Blood Lead to Erythrocytic Protoporphyrin hi Young Children." In:
Bomschein, R.L., Rabinowitz, M.B., eds. The Second International Conference on Prospective
Studies of Lead, Cincinnati, OH: April, 1984. Environmental Research. 38:187-196.
Hansen, O.N., Trillingsgaard, A., Beese, L, et al. (1989) "A Neuropsychological Study of
Children with Elevated Dentine Lead Level: Assessment of the Effect of Lead hi Different
Socio-economic Groups." Neurotoxicol Teratol. 11:205-213.
Harlan, W.R. (1988) "The Relationship of Blood Lead Levels to Blood Pressure in the US
Population. Environmental Health Perspectives. 78:9-13.
Harlan, W.R., Landis, J.R., Schmouder, R.L., et al. (1988) "Blood Lead and Blood Pressure:
Relationship in the Adolescent and Adult US Population." Journal of the American Medical
Association. 253:530-534.
Harrison, G.E., et al. (1969) "Effect of Alginate on the Absorption of Lead in Man." Nature
(London). 224:1115-1116.
7-14
-------
Harvey, P.O., Hamlin, M.W., Kumar, R., et al. (1984) "Blood Lead, Behavior and Intelligence
Test Performance in Preschool Children." Set Total Environ. 40:45-60.
Harvey, P.O., Hamlin, M.W., Kumar, R., Morgan, G., Spurgeon, A., Delves, H.T. (1988)
"Relationship Between Blood Lead, Behaviour, Psychometric and Neuropsychological Test
Performance in Young Children." British Journal ofDevelomental Psychology. 29:43-52.
Hatzakis, A., Kolevi, A., Katsouyanni, K., et al. (1987) "Psychometric Intelligence and
Attentional Performance Deficits in Lead Exposed Children." International Conference on
Heavy Metals in the Environment. Edinburgh, Scotland; CEP Consultants. 16:57-67.
Hatzakis, A., Kokkeni, A., Maranelias, C., Katsouyanni, K., Salaminios, F., Kalandidi, A., et al.
(1989) "Psychometric Intelligence Deficits hi Lead-exposed Children." In: Smith MA., Grant,
I.D., Sora Al, eds. Lead Exposure and Child Development. London: Kluwer, 1989:211-23.
Hawk, B.A., Schroeder, S.R., Robinson, G., et al. (1986) "Relation of Lead and Social Factors to
IQ of Low-SES Children: A Partial Replication. " American Journal of Mental Deficiency
91:178-183.
Heard, M.J. and 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. and Chamberlain, A.C. (1984) "Uptake of Pb by Human Skeleton and Comparative
Metabolism of Pb and Alkaline Earth Elements." Health Physics. 47:857-865.
Hernberg, S., Nikkanen, J., Mellin, G., et al. (1970) "6-Aminolevulinic Acid Dehydrase as a
Measure of Lead Exposure." Archives of Environmental Health. 21:140-145.
Hirata, M. and Kosaka, H. (1993) "Effects of Lead Exposure on Neurophysiological Parameters."
Environmental Research 63:60-69.
Hogstedt, C., Hane, M., Agrell, A., et al. (1983) "Neuropsychological Test Results and
Symptoms Among Workers with Well-defined Long-term Exposure to Lead." British Journal of
Industrial Medicine. 40:99-105.
Holness, D.L. and Nethercott, J.R. (1988) "Acute Lead Intoxication in a Group of Demolition
Workers." Applied Industrial Hygiene. 3:338-341.
Hu, H., Pepper, L., Goldman, R. (1991) "Effect of Repeated Occupational Exposure to Lead,
Cessation of Exposure, and Chelation on Levels of Lead in Bone." American Journal of
Industrial Medicine. 20-723-735.
Huang, X.P., Feng, Z.Y., Zhai, W.L., et al. (1988) "Chromosomal Aberrations and Sister
Chromatid Exchanges in Workers Exposed to Lead." Biomed Environ Sci. 1:382-387.
7-15
-------
Hursh, J.B. and Suomela, J. (1968) "Absorption of212 Pb from the Gastrointestinal Tract of
Man." ActaRad. 7:108-120.
Hyrhorczuk, D., Rabinowitz, M., Hessel, S., Hoffman, D., Hogan, M., Mallin, K., Finch, H.,
Chris, P., Berman, E. (1985) "Elimination Kinetics of Blood Lead in Workers with Chronic Lead
Intoxication." American Journal of Industrial Medicine. 8:33-42.
ICF (1995) "Regulatory Impact Analysis of the Proposed Rule on Lead-Based Paint:
Requirements for Notification, Evaluation and Reduction of Lead-Based Paint Hazards in
Federally-Owned Residential Property and Housing Receiving Federal Assistance." Draft report
for the Office of Lead-Based Paint Abatement and Poisoning Prevention, U.S. Department of
Housing and Urban Development.
Integrated Risk Information System (IRIS) on Lead (1993), Environmental Criteria and
Assessment Office, Office of Health and Environmental Assessment, Office of Research and
Development, U.S. Environmental Protection Agency. Cincinnati, OH.
International Agency for Research on Cancer (IARC), Overall Evaluations of Carcinogenicity:
An Updating of IARC Monographs, Volumes 1 to 42. Lyon: IARC Monographs on the
Evaluation of Carcinogenic Risk to Humans, Supplement 7.
Jacobson, M. (1990) Developmental Neurobiology, 4th ed. Plenum, New York.
Janin, Y., Couinaud, C., Stone, A., et al. (1985) "The 'Lead-Induced Colic' Syndrome in Lead
Intoxication." SurgAnn. 17:287-307.
Jiang, Q. and Succop, P. A., "A Study of the Specification of the Log-Log and Log-Additive
Models for the Relationship Between Blood-Lead and Environmental Lead." (1996). Journal of
Agricultural, Biological, and Environmental Statistics. 1(4): 1-9.
Johnson, N.E. and Tenuta, K. (1979) "Diets and Lead Blood Levels of Children Who Practice
Pica." Environmental Research. 18:869-376.
Kang, H.K., Infante, P.F., Carra, J.S. (1980) "Occupational Lead Exposure and Cancer (Letter)."
Science. 207:935-936.
Kehoe, R.A. (196 la) "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.
7-16
-------
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.
Khera, A.K., Wibberley, D.G., Edwards, K.W., et al. (1980) "Cadmium and Lead Levels in
Blood and Urine in a Series of Cardiovascular and Normotensive Patients." International
Journal of Environmental Studies. 14:309-312.
Kimbrough, R.D., LeVois, M., Webb, D.R. (1994) "Management of Children with Slightly
Elevated Blood Lead Levels." Pediatrics. 93(2): 188-191.
Kirkby, H. and Gyntelberg, F. (1985) "Blood Pressure and Other Cardiovascular Risk Factors of
Long-term Exposure to Lead." Scandinavian Journal of Work Environment and Health. 11:15-
19.
Klaassen, C.D., et al. (1993) "Cassartt and Doull's Toxicology: The Basic Science of Poisons."
4th Edition, McGraw-Hill Publishers, New York.
Kleinbaum, D.G., Kupper, L.L., and Morgenstern, H. (1982) "Epidemiologic Research:
Principle and Methods. Van Norstrand Reinhold Co., New York.
Koo, W.W.R., Succop, P.A., Bornschein, R.L., et al. (1991) "Serum Vitamin D Metabolites and
Bone Mineralization in Young Children with Chronic Low to Moderate Lead Exposure."
Pediatrics. 87:680-687.
Kostial, K. and Vouk, V.B. (1957) Lead Ions and Synaptic Transmission in the Superior Cervical
Ganglion of the Cat. Br. J. PharmacoL 12:219-226.
Kotok, D. (1972) "Development of Children with Elevated Blood Levels: A Controlled Study."
Journal of Pediatrics. 80:57-61.
Kotok, D., Kotok, R., Heriot, T. (1977) "Cognitive Evaluation of Children with Elevated Blood
Lead Levels." Am JDis Child. 131:791-793.
Kuhnert, P.M., Erhard, P., Kuhnert, B.R. (1977) "Lead and 6-ammoluvulinic Acid Dehydratase
in RBC's of Urban Mothers and Fetuses." Environmental Research. 14:73-80.
Kumar, S., Jain, S., Aggarwal, C.S., et al. (1987) "Encephalopathy Due to Inorganic Lead
Exposure in an Adult" Jpn JMed 26:253-254.
Lancranjan, I., Popescu, H.I., Gavanescu, O., et al. (1975) "Reproductive Ability of Workmen
Occupationally Exposed to Lead." Archives of Environmental Health. 30:396-401.
7-17
-------
Landis, J.R. and Flegal, K.M. (1988) "A Generalized Mantel-Haenszel Analysis of the
Regression of Blood Pressure on Blood Lead Using NHANES n Data " Environmental Health
Perspectives. 78:35-41.
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 Particulate Lead." New England Journal of
Medicine. 292(3):123-129.
Landrigan, P.J., Baker, EX., Jr., Feldman, R.G., et al. (1976) "Increased Lead Absorption with
Anemia and Slowed Nerve Conduction in Children Near a Lead Smelter." Journal of Pediatrics.
89:904-910.
Lanphear, B.P., Emond, M., Jacobs, D.E., Weitzman, M., Tanner, M., Winter, N.L., Yakir, B.,
Eberly, S. (1995) "A Side-by-Side Comparison of Dust Collection Methods for Sampling Lead-
Contaminated House Dust." Environmental Research 68(2): 114-123.
Lanphear, B. P., M. Weitzman, and S. Eberly (1996a) "Racial Differences hi Urban Children's
Environmental Exposures to Lead," American Journal of Public Health, Vol. 86, No. 10, pp.
1460-1463.
Lanphear, B. P., M. Weitzman, N. L. Winter, S. Eberly, B. Yakir, M. Tanner, M. Emond, and T.
D. Matte (1996b) "Lead-Contaminated House Dust and Urban Children's Blood Lead Levels,"
American Journal of Public Health, Vol. 86, No. 10, pp. 1416-1421.
Lansdown, R., Yule, W., Urbanowicz, M.A., et al. (1986) "The Relationship Between Blood
Lead Concentrations, Intelligence, Attainment and Behavior in a School Population: The Second
London Study." International Archives of Occupational Environmental Health. 57:225-235.
Lasley, S.M. and Lane, J.D. (1988) "Diminished Regulation of Mesolimbic Dopaminergic
Activity in Rat after Chronic Inorganic Lead Exposure." Toxicol. Appl. Pharmacol. 95:474-483.
Lauwers, M.C., Hauspie, R.C., Susanne, C., et al. (1986) "Comparison of Biometric Data of
Children with High and Low Levels of Lead in the Blood." Am JPhys Anthropol. 69:107-116.
Lauwerys, R., Buchet, J-P, Roels, H.A., et al. (1974) "Relationship Between Urinary 6-
aminolevulinic Acid Excretion and the Inhibition of Red Cell 6-aminolevulinate Dehydratase by
Lead." Clin Toxicol. 7:383-388.
Lauwerys, R., Buchet, J-P, Roels, H.A., et al. (1978) "Placental Transfer of Lead, Mercury,
Cadmium, and Carbon Monoxide hi Women: I. Comparison of the Frequency Distributions of
the Biological Indices in Maternal and Umbilical Cord Blood." Environmental Research.
15:278-289.
7-18
-------
Leal-Garza, C., Moates, De Oca R, Cerda-Flores, R.M., et al. (1986) "Frequency of Sister-
Chromatid Exchanges (SCE) in Lead Exposed Workers." Arch Invest Med. 17:267-276.
Leggett, R.W., Eckerman, K.F., and Williams, L.R. (1982) "Strontium-90 in Bone: A Case
Study in Age-Dependent Dosimetric Modeling." Health Physics. 43(3):307-322.
Lewis and Clark County Health Department, Montana Department of Health and Environmental
Sciences, Centers for Disease Control, U.S. Department of Health and Human Services, and U.S.
EPA (1986) "East Helena, Montana: Child Lead Study, Summer 1983." Final Report, July
1986.
Lilis, R., Gavrilescu, N., Nestorescu, B., et al. (1968) "Nephropathy in Chronic Lead Poisoning."
British Journal of Industrial Medicine. 25:196-202.
Lilis, R., Eisinger, J., Blumberg, W., et al. (1978) "Hemoglobin, Serum Iron, and Zinc
Protoporphyrin in Lead-exposed Workers." Environmental Health Perspectives. 25:97-102.
Lundstrfim, N.G., Nordberg, G., Englyst, V., Gerhardsson, L., Hagmar, L., Jin, T., et al. (1997)
"Cumulative Lead Exposure - Relationship to Mortality and Lung Cancer Morbidity in a Cohort
of Smelter Workers." Scandinavian Journal of Work Environment and Health. 23(1):24-30.
Lyngbye, T., Hansen, O.N., Grandjean, P. (1987) "The Influence of Environmental Factors on
Physical Growth in School Age: A Study of Low Level Lead Exposure." In: Lindberg S.E.,
Hutchinson, T.C., eds. International Conference on Heavy Metals in the Environment, Vol. 2,
New Orleans, LA, September. Edinburgh, UK: CEP Consultants, Ltd. 210-212.
Mahaffey, K.R. (1977) "Quantities of Lead Producing Health Effects in Humans: Sources and
Bioavailability." Environmental Health Perspectives. 19:285-295.
Mahaffey, K.R., Rosen, J.F., Chesney, R.W., et al. (1982) "Association Between Age, Blood
Lead Concentration, and Serum 1,25-dihydroxycholecalciferol Levels in Children." Am J din
Nutr. 35:1327-1331.
Maki-Paakkanen, J., Sorsa, M., Vainio, H. (1981) "Chromosome Aberrations and Sister
Chromatid Exchanges in Lead-Exposed Workers." Hereditas. 94:269-275.
Malcolm, D. and Barnett, H.A.R. (1982) "A Mortality Study of Lead Workers: 1925-76."
British Journal of Industrial Medicine. 39:404-410.
Mantere, P., Haenninen, H., Hemberg, S. (1982) "Subclinical Neurotoxic Lead Effects: Two-
Year Follow-up Studies with Psychological Test Methods." Neurobehav Toxicol Teratol. 4:725-
727.
Manton, W.I. (1985) "Total Contribution of Airborne Lead to Blood Lead." British Journal of
Industrial Medicine. 42:168-72.
7-19
-------
Maranelli, G. and Apostoli, P. (1987) "Assessment of Renal Function in Lead Poisoned
Workers." Occup Environ Chem Hazards. 344-348.
Marcus, A.H. (1985) "Multicompartment Kinetics Models for Lead. I. Bone Diffusion Models
for Long-Term Retention." Environmental Research. 36:441-458.
Marino, P.E., Franzblau, A., Lilis, R., et al. (1989) "Acute Lead Poisoning in Construction
Workers: The Failure of Current Protective Standards." Archives of Environmental Health.
44:140-145.
Markovac, J. and Goldstein, G.W. (1988) "Picomolar Concentration of Lead Stimulate Brain
Protein Kinase C." Nature. 334:71-73.
Markowitz, M.E., Bijur, P.E., Ruff, H.A., Rosen, J.F. (1993) "Effects of Calcium Disodium
Versenate (Canasta) Chelation in Moderate Childhood Lead Poisoning." Pediatrics.
92(2):265-271.
Matte, T.D., Figueroa, J.P., Burr, G., et al. (1989) "Lead Exposure Among Lead-acid Battery
Workers in Jamaica " American Journal of Industrial Medicine. 16:167-177.
McBride, W.G., Black, B.P., English, B.J. (1982) "Blood Lead Levels and Behavior of 400
Preschool Children." MedJAust. 10:2(l):26-29.
McElvaine, M.D., DeUngria, E.G., Matte, T.D., Copley, C.G., and Binder, S. (1992) "Prevalence
of Radiographic Evidence of Paint Chip Digestion Among Children with Moderate to Severe
Lead Poisoning, St. Louis, Missouri, 1989 Through 1990." Pediatrics. 89(4):740-742.
McMichael, A.J., Vimpani, G.V., Robertson, E.F., et al. (1986) "The Port Pirie Cohort Study:
Maternal Blood Lead and Pregnancy Outcome." JEpidemiol Community. 40:18-25.
McMichael, A.J., Baghurst, P.A., Wigg, N.R., et al. (1988) "Port Pirie Cohort Study:
Environmental Exposure to Lead and Children's Abilities at the Age of Four Years." New
England Journal of Medicine. 319:468-476.
Menton, R.G., Burgoon, D.A., Marcus, A.H. (1995) "Pathways of Lead Contamination for the
Brugham and Women's Hospital Longitudinal Lead Study." In: Lead hi Paint, Soil, and Dust:
Health Risks, Exposure Studies, Control Measures, Measurement Methods, and Quality
Assurance, ASTM STP 1226, Michael E. Beard and S. D. Allen Iske, Eds. Philadelphia:
American Society for Testing and Materials, 92-106.
Meredith, P.A., Moore, M.R., Campbell, B.C., et al. (1978) "6-aminolaevulinic Acid Metabolism
in Normal and Lead-exposed Humans." Toxicology. 9:1-9.
Michaels, D., Zoloth, S.R., Stem, F.B. (1991) "Does Low-level Lead Exposure Increase Risk of
Death?: A Mortality Study of Newspaper Printers." Int J Epidemiol. 20:978-983.
7-20
-------
Milbum, H., Mitran, E., Crockford, G.W. (1976) "An Investigation of Lead Workers for
Subclinical Effects of Lead Using Three Performance Tests." Ann Occup Hyg. 19:239-249.
Minnema, D.J., et al. (1986) "Effect of in Vitro Inorganic Lead on Dopamine Release from
Supervised Rat Striatal Synaptosomes. Toxicol Appl. Parmacol. 84:400-411.
Moore, M.R., Goldberg, A., Pocock, S.J., et al. (1982) "Some Studies of Maternal and Infant
Lead Exposure in Glasgow." Scott MedJ. 27:113-122.
Moresco, R.M., et al. (1988) "Lead Neurotoxicity: A Role for Dupamine Receptors."
Toxicology, 53:315-322.
Morgan, A. and Holmes, A. (1978) "The Fate of Lead in Petrol-Engine Exhaust Particulates
Inhaled by the Rat." Environ. Res. 15:44-56.
Morrow, P.E., et al. (1980) "Pulmonary Retention of Lead: An Experimental Study in Man."
Environ. Res. 21:373-384.
Muijser, H., Hoogendijk, E.M., Hoosma, J., et al. (1987) "Lead Exposure During Demolition of a
Steel Structure Coated with Lead-based Paints, n. Reversible Changes in the Conduction
Velocity of the Motor Nerves in Transiently Exposed Workers." Scandinavian Journal of Work
Environment and Health. 13:56-61.
Murphy, M.J., Graziano, J.H., Popovac, D., et al. (1990) "Past Pregnancy Outcomes Among
Women Living in the Vicinity of a Lead Smelter in Kosovo, Yugoslavia " American Journal of
Public Health. &Q33-35.
NCLSH and UCDEH (1994) "Evaluation of the HUD Lead-Based Paint Hazard Control Grant
Program in Private Housing: Overall Design and Data Collection Forms and Procedures."
Report prepared by the National Center for Lead-Safe Housing and the University of Cincinnati
Department of Environmental Health, OMB No. 2539-0004,14 February 1994.
NCLSH and UCDEH (1997) "Evaluation of the HUD Lead-Based Paint Hazard Control Grant
Program: Fourth Interim Report, Progress as of September 15,1996." Report prepared by the
National Center for Lead-Safe Housing and the University of Cincinnati Department of
Environmental Health for the U.S. Department of Housing and Urban Development, February
1997.
National Academy of Sciences. (1972) "Lead: Airborne Lead in Perspective: Biologic Effects of
Atmospheric Pollutants." Washington, DC: National Academy of Sciences. 71-177,281-313.
Needleman, H.L., Gunnoe, C., Leviton, A., et al. (1979) "Deficits in Psychologic and Classroom
Performance of Children with Elevated Dentine Lead Levels." New England Journal of
Medicine. 300:689-695.
7-21
-------
Needleman, H.L., Rabinowitz, M., Leviton, A., et al. (1984) "The Relationship Between Prenatal
Exposure to Lead and Congenital Anomalies." Journal of the American Medical Association.
251:2956-2959.
Needleman, H.L., Geiger, S.K., Frank, R. (1985) "Lead and IQ Scores: A Reanalysis (letter)."
Science. 227:701-704.
Needleman, H.L., Schell, A., Bellinger, D., Leviton, A., Allred, E.N. (1990) "The Long-Term
Effects of Exposure to Low Doses of Lead in Childhood: An 11-Year Follow-up Report." The
New England Journal of Medicine. 322(2):83-88.
Neri, L.C., Hewitt, D., Orser, B. (1988) "Blood Lead and Blood Pressure: Analysis of Cross-
sectional and Longitudinal Data from Canada." Environmental Health Perspectives. 78:123-
126.
Neter, J. and Wasserman, W. (1974) Applied Linear Statistical Models. Homewood, IL: Richard
D. Irwin, Inc.
Nilsson, U., Attewell, R., Christoffersson, J.O., Schutz, A., Ahlgren, S., SkerfVing, S., Mattson,
S. (1991) "Kinetics of Lead in Bone and Blood After End of Occupational Exposure."
Pharmacology and Toxicology. 69:477-484.
Nordberg, G.R., Mahaffey, K.R., and Fowler, B.A. (1991) "Introduction and Summary."
International Workshop on Lead in Bone: Implications for Dosimetry and Toxicology.
Environmental Health Perspectives. 91:3-7.
Nordenson, I., Beckman, G., Beckman, L., et al. (1978) "Occupational and Environmental Risks
In and Around a Smelter in Northern Sweden: IV. Chromosomal Aberrations hi Workers
Exposed to Lead." Hereditas. 88:263-267.
Nordstrom, S., Beckman, L., Nordenson, I. (1979) "Occupational and Environmental Risks in
and Around a Smelter in Northern Sweden: V. Spontaneous Abortion Among Female
Employees and Decreased Birth Weight hi Then- Offspring." Hereditas. 90:291-296.
Nozaki, K. (1966) "Method for Studies on Inhaled Particles hi Human Respiratory System and
Retention of Lead Fumes." Ind. Health. 4:118-128.
Nye, L.J.J. (1929) "An Investigation of the Extraordinary Incidence of Chronic Nephritis hi
Young People in Queensland." MedJAust. 2:145-159.
Ong, C.N., Endo, G., Chia, K.S., et al. (1987) "Evaluation of Renal Function hi Workers with
Low Blood Lead Levels." In: Fao V., Emmett, E.A., Maroni, M., et al., eds. Occupational and
Environmental Chemical Hazards. Chichester: Ellis Horwood Limited, 327-333.
7-22
-------
O'Riordan, MX. and Evans, H.J. (1974) "Absence of Significant Chromosome Damage in Males
Occupationally Exposed to Lead." Nature. 247:50-53.
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 " JClin Invest. 56:1164-1169.
Paglia, D.E., Valentine, W.N., Fink, K. (1977) "Lead Poisoning: Further Observations on
Erythrocyte Pyrimidine-nucleotidase Deficiency and Intracellular Accumulation of Pyrimidine
Nucleotides." JClin Invest. 60:1362-1366.
Pagliuca, A., Mufti, G.J., Baldwin, D., et al. (1990) "Lead Poisoning: Clinical, Biochemical, and
Hematological Aspects of a Recent Outbreak." JClin Path. 43:277-281.
Panhandle District Health Department, Idaho Department of Health and Welfare, Centers for
Disease Control, and U.S. EPA. (1986) "Kellogg Revisited -1983: Childhood Blood Lead and
Environmental Status Report." Final Report of the U.S. Public Health Service, July 1986.
Parkinson, D.K., Ryan, C., Bonnet, J., et al. (1986) "A Psychiatric Epidemiologic Study of
Occupational Lead Exposure." Am JEpidemiol. 123:261-269.
Parkinson, D.K., Hodgson, M.J., Bromet, E.J., et al. (1987) "Occupational Lead Exposure and
Blood Pressure." British Journal of Industrial Medicine. 44:744-748.
Pasternak, G., Becker, C.E., Lash, A., et al. (1989) "Cross-sectional Neurotoxicology Study of
Lead-exposed Cohort." Clin Toxicol. 27:37-51.
Petit, T.L., et al. (1983) "Early Lead Exposure and the Hippocampus: A Review and Recent
Advances." Neurotoxicology. 4:79-94.
Piomelli, S., Seaman, C., Zullow, D., et al. (1982) "Threshold for Lead Damage to Heme
Synthesis in Urban Childrea " Proc NatlAcadSci. 7:3335-3339.
Pirkle, J.L., Schwartz, J., Landis, J.R., et al. (1985) "The Relationship Between Blood Lead
Levels and Blood Pressure and its Cardiovascular Risk Implications." Am JEpidemiol. 121:246-
258.
Pirkle, J.L., Brody, D.J., Gunter, E.W., Kramer, R.A., Paschal, D.C., Flegal, K.M., Matte, T.D.
(1994) "The Decline in Blood Lead Levels in the United States: The National Health and
Nutrition Examination Surveys (NHANES)." Journal of the American Medical Association.
272(4):284-291.
Pocock, S.J., Shaper, A.G., Ashby, D., et al. (1984) "Blood Lead Concentration, Blood Pressure
and Renal Function. British Medical Journal 289-872-874.
7-23
-------
Pocock, S.J., Shaper, A.G., Ashby, D., et al. (1985) "Blood Lead and Blood Pressure in Middle-
aged Men." In: Lekkas, T.D., ed. International Conference on Heavy Metals in the Environment,
Vol. 1, Athens, Greece, Sepember. Edinburgh, United Kingdom: CEP Consultants, Ltd., SOS-
SOS.
Pocock, S.J., Shaper, A.G., Ashby, D., et al. (1988) "The Relationship Between Blood Lead,
Blood Pressure, Stroke, and Heart Attacks in Middle-aged British Men." Environmental Health
Perspectives. 78:23-30.
Pocock, S.J., Ashby, D., Smith, M.A. (1989) "Lead Exposure and Children's Intellectual
Performance: The Institute of Child Health/Southampton Study." In: Smith, M., Grant, L.D.,
Sors, A., eds. Lead Exposure and Child Development: An International Assessment. Lancaster,
UK: Kluwer Academic Publishers.
Pocock, S.J., Smith, M., Baghurst, P. (1994) "Environmental Lead and Children's Intelligence:
A Systematic Review of the Epidemiological Evidence." BMJ. 309:1189-1197.
Pollock, C.A. and Ibels, L.S. (1986) "Lead Intoxication in Paint Removal Workers on the Sidney
Harbour Bridge." MedJAust. 145:635-639.
Pott, F. and Brockhaus, A. (1971) "Vergleich Der Enteralen Und Pulmonalen Resorptionsquote
von Bleiverbindungen (Comparison of the Enteral and Pulmonary Absorption Rates of Lead
Compounds)." Zentralbl Bakteriol. Parasitendic Infectionsbrankh. Hyg. Abt 1 Orig. Reihe B.
Pueschel, S.M., Kopito, L., Schwachman, H. (1972) "Children with an Increased Lead Burden:
A Screening and Follow-up Study." Journal of the American Medical Association. 222:462-466.
Que Hee, S.S., Peace, B., Clark, S., Boyle, J.R., Bomschein, R.L., Hammond, P.B. (1985)
"Evolution of Efficient Methods to Sample Lead Sources, Such as House Dust and Hand Dust, in
the Homes of Children." Environmental Research 38:77-95.
Rabinowitz, M.B., Wetherill, G.W., Kopple, J.D. (1973) "Lead Metabolism in the Normal
Human: Stable Isotope Studies." Science. 182:725-727.
Rabinowitz, M.B., Wetherill, G.W., Kopple, J.D. (1976) "Kinetic Analysis of Lead Metabolism
in Healthy Humans." Journal of Clinical Investigation. 58:260-270.
Rabinowitz, M.B. et. al. (1977) "Magnitude of Lead Intake from Respiration by Normal Man."
J. Lab. Clin. Med. 90:238-249.
Rabinowitz, M.B., and Needleman, H.L. (1982) "Temporal Trends in the Lead Concentrations of
Umbilical Cord Blood." Science. 216:1429-1430.
Rabinowitz, M.B., and Needleman, H.L. (1984a) "Environmental, Demographic, and Medical
Factors Related to Cord Blood Lead Levels." Biological Trace Element Research. 6:57-67.
7-24
-------
Rabinowitz, M.B., Needleman, H.L., Leviton, A. (1984b) "Variability of Blood Lead During
Normal Infancy." Archives of Environmental Health. 39:74-77
Rabinowitz, M, Leviton, A., Needleman, H., Bellinger, D., Waternaux, C. (1985a).
"Environmental Correlates of Infant Blood Lead Levels in Boston." Environmental Research.
38:96-107.
Rabinowitz, M., Leviton, A., Bellinger, D. (1985b) "Home Refinishing, Lead Paint, and Infant
Blood Lead Levels." American Journal of Public Health. 75(4):403-404.
Rabinowitz, M.B., Leviton, A., Needleman, H. (1986) "Occurrence of Elevated Protoporphyrin
Levels in Relation to Lead Burden in Infants." Environmental Research. 39:253-257.
Rabinowitz, M.B. (1991) "Toxicokinetics of Bone Lead." Environmental Health Perspectives.
91:33-37.
Regan, C.M. (1989) "Lead-impaired Neurodevelopment: Mechanisms and Threshold Values in
the Rodent." Neurotoxicol. Teratol. 11:533-537.
Reigart, J.R. and Graber, C.D. (1976) "Evaluation of the Humoral Immine Response of Children
with Low Level Lead Exposure." Bull Environ Contam Toxicol. 16:112-117.
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.
Rodamilans, M., Osaba, M.J., To-Figueras, J., et al. (1988) "Lead Toxicity on Endocrine
Testicular Function in an Occupationally Exposed Population." Hum Toxicol. 7:125-128.
Roels, H.A., Lauwerys, R.R., Buchet, J-P, et al. (1975) "Response of Free Erythrocyte Porphyrin
and Urinary-6-aminolevulinic Acid hi Men and Women Moderately Exposed to Lead " bit Arch
Arbeitsmed 34:97-108.
Roels, H.A., Buchet, J-P, Lauwerys, R., et al. (1976) "Impact of Air Pollution by Lead on the
Heme Biosynthetic Pathway in School-age Children." Archives of Environmental Health.
31:310-316.
Roels, H.A., Balis-Jacques, M.N., Buchet, J-P, et al. (1979) "The Influence of Sex and Chelation
Therapy on Erythrocyte Protoporphyrin and Urinary-6-aminolevulinic Acid in Lead-exposed
Workers." Journal of Occupational Medicine. 21:527-539.
Roels, H.A. and Lauwerys, R. (1987) "Evaluation of Dose-effect and Dose-response
Relationships for Lead Exposure in Different Belgian Population Groups (Fetus, Child, Adult
Men and Women)." Trace Elements in Medicine. 4:80-87.
7-25
-------
Rosen, J.F., Zarate-Salvador, C., Trinidad, E.E. (1974) "Plasma Lead Levels in Normal and
Lead-intoxicated Children." Journal of Pediatrics. 84:45-48.
Rosen, J.F., Chesney, R.W., Hamstra, A.J., et al. (1980) "Reduction in 1,25-dmydroxyvitamine
D in Children with Increased Lead Absorption.'Wew England Journal of Medicine. 302:1128-
1131.
Rosen, L, Wildt, K., Gullberg, B., et al. (1983) "Neurophysiological Effects of Lead Exposure."
Scandinavian Journal of Work Environment and Health 9:431-441.
Rothenberg, S.J., Schnaas, L., Cansino-Ortiz, S., et al. (1989) "Neurobehavioral Deficits After
Low Level Lead Exposure in Neonates: The Mexico City Pilot Study." Neurotoxicol Teratol.
11:85-93.
Rummo, J.H., Routh, D.K., Rummo, N.J., et al. (1979) "Behavioral and Neurological Effects of
Symptomatic and Asymptomatic Lead Exposure in Children." Archives of Environmental
Health. 34:120-125.
Rust, S.W., Burgoon, D.A., Lanphear, B.P., and Eberly, S. (1996) "Log-additive versus log-
linear analysis of lead-contaminated house dust and children's blood-lead levels." Environmental
Research (hi press).
Rust, S.W., Kumar, P., Burgoon, D.A., Schultz, B. (1997) "Influence of Bone-lead Stores on the
Observed Effectiveness of Lead Hazard Intervention." (hi review).
Ryan, C.M., Morrow., L., Parkinson, D., et al. (1987) "Low Level Lead Exposure and
Neuropsychological Functioning hi Blue Collar Males." hit JNeurosci. 36:29-39.
Sachs, H.K. and Moel, D.I. (1989) "Height and Weight Following Lead Poisoning in
Childhood." American Journal of Diseases and Children. 143:820-822.
Saenger, P., Markowitz, M.E., Rosen, J.F. (1984) "Depressed Excretion of 6p-hydroxycortisol in
Lead-toxic Children." J Clin Endocrinol Metab. 58:363-367.
Schmid, E., Bauchinger, M., Pietruck, S., et al. (1972) "Cytogenic Action of Lead hi Human
Peripheral Lymphocytes In Vitro and/n Vivo" MutRes. 16:401-406. (German)
Schneitzer, L., Osborn, H.H., Bierman, A., et al. (1990) "Lead Poisoning hi Adults from
Renovation of a Older Home." AnnEmergMed. 19:415-420.
Schroeder; H.A. and Tipton, I.H. (1968) "The Human Body Burden of Lead." Archives of
Environmental Health. 17:965-977.
Schroeder, S.R., Hawk, B., Otto, D., Mushak, P., Hicks, R.E. (1985) "Separating the Effects of
Lead and Social Factors on IQ." Environmental Research. 91:178-183.
7-26
-------
Schroeder, S.R. and Hawk, B. (1987) "Psycho-social Factors, Lead Exposure and IQ." Monogr
Am Assoc Ment Defic. 8:97-137.
Schutz, A., Skerfving, S., Ranstam, J., Christoffersson, J. (1987) "Kinetics of Lead in Blood after
the End of Occupational Exposure." Scand.J. Work Environ. Health. 13:221-231.
Schwartz, J., Angle, C., Pitcher, H. (1986) "Relationship Between Childhood Blood Lead Levels
and Stature." Pediatrics. 77:281-288.
Schwartz, J. and Otto, D.A. (1987) "Blood Lead, Hearing Thresholds, and Neurobehavioral
Development in Children and Youth." Archives of Environmental Health. 42:153-160.
Schwartz, J. (1988) "The Relationship Between Blood Lead and Blood Pressure in the NHANES
H Survey." Environmental Health Perspectives. 78:15-22.
Schwartz, J., Landrigan, P.J., Feldman, R.G., et al. (1988) "Threshold Effect in Lead-induced
Peripheral Neuropathy." Journal of Pediatrics. 112:12-17.
Schwartz, J., Landrigan, P.J., Baker, EX., Jr. (1990) "Lead-induced Anemia: Dose-response
Relationships and Evidence for a Threshold." American Journal of Public Health. 80:165-168.
Schwartz, J. (1993) "Beyond LOEL's, p Values, and Vote Counting: Methods for Looking at the
Shapes and Strengths of Association." Neurotoxicology and Teratology. 14(2-3):237-246.
Schwartz, J. (1994) "Low-Level Lead Exposure and Children's IQ: A Meta-Analysis and Search
for a Threshold." Environmental Research. 65:42-55.
Secchi, G.C., Erba, L., Cambiaghi, G. (1974) "6-aminolevulinic Acid Dehydratase Activity of
Erythrocytes and Liver Tissue hi Man: Relationship to Lead Exposure." Archives of
Environmental Health. 28:130-132.
Selander, S. and Cramer, K. (1970) "Interrelationships Between Lead in Blood, Lead in Urine,
and ALA hi Urine During Lead Work." British Journal of Industrial Medicine. 27:28-39.
Selevan, S.G., Landrigan, P.J., Stern, F.B., et al. (1985) "Mortality of Lead Smelter Workers."
Am. J. Epidemiol. 122:673-683.
Seppalainen, A.M., Hernberg, S., Vesanto, R., et al. (1983) "Early Neurotoxic Effects of
Occupational Lead Exposure: A Prospective Study." Neurotoxicology. 4:181-192.
Seto, D.S.Y. and Freeman, J.M. (1964) "Lead Neuropathy in Childhood." Am JDis Child.
107:337-342.
7-27
-------
Shacklette, H.T., and Boerngen, J.G., (1984) "Element Concentrations in Soils and Other
Surficial Materials of the Conterminous United States," U.S. Geological Survey Professional
Paper 1270. U.S. Government Printing Office, Washington, D.C.
Shaheen, S. (1984) "Neuromaturation and Behavior Development: The Case of Childhood Lead
Poisoning." Dev. Psych. 20:542-550.
ShukhL, R. Bornschein, R.L., Dietrich, K.N., et al. (1987) "Effects of Fetal and Early Postnatal
Lead Exposure on Child's Growth in Stature The Cincinnati Lead Study." In: Lindberg, S.E.
Hutchinson, T.C., eds. International Conference on Heavy Metals in the Environment, Vol. 1.
New Orleans, LA, September. Edinburgh, UK: CEP Consultants, Ltd., 210-212.
Shukla, R., Bornschein, R.L., Dietrich, K.N., et al. (1989) "Fetal and Infant Lead Exposure:
Effects on Growth in Stature." Pediatrics. 84:604-612.
Siegel, M., Forsyth, B., Siegel, L., et al. (1989) "The Effect of Lead on Thryoid Function in
Children." Environmental Research. 49:190-196.
Silbergeld, E.K., et al. (1974) "Lead: Evidence for a Prejunctional Effect on Neuro Muscular
Function." Nature. 247:49-50.
Silbergeld, E.K. and Adler, H.S. (1978) "Subcellular Mechanisms of Lead Neurotoxicity." Brain
Res. 148:451-467.
Silbergeld, E.K. (1991) "Aneurological Perspective on Lead Toxicity." In Human Lead
Exposure (Needleman, H.L. ed.) 89-103. CRC Press, Boca Raton, Florida.
Silbergeld, E.K. (1992) "Mechanisms of Lead Neurotoxiciry, or Looking Beyond the Lampost."
FASEB Journal. 6:3201-3206.
Silva, P.A., Hughes, P., Williams, S., Faed, J.M. (1988) "Blood Lead, Intelligence, Reading
Attainment, and Behaviour in Eleven Year Old Children in Dunedin, New Zealand." Journal
of Child Psychology and Psychiatry. 29:43-52.
Silver, W. and Rodriguez-Torres, R. (1968) "Electrocardiographic Studies in Children with Lead
Poisoning." Pediatrics. 41:1124-1127.
Smith, M., Delves, T., Lansdown, R., et al. (1983) "The Effects of Lead Exposure on Urban
Children: The Institute of Child Health/Southampton Study." Dev Med Child Neurol. 25 (suppl
47).
Smith, F.L., H, Rathmell, T.K., Marcil, G.E. (1938) "The Early Diagnosis of Acute and Latent
Plumbism." Am JClinPathol. 8:471-508.
7-28
-------
Smith, D.R., Osterloh, J.D., Flegal, A.R. (1996) "Use of Endogenous Stable Lead Isotopes to
Determine Release of Lead from the Skeleton." Environmental Health Perspectives. 104(1):60-
66.
Spivey, G.H., Baloh, R.W., Brown, C.P., et al. (1980) "Subclinical Effects of Chronic Increased
Lead AbsorptionA Prospective Study: ffl. Neurologic Findings at Follow-up Examintion."
Journal of Occupational Medicine. 22:607-612.
Staes, C., Matte, T., Copley, G., Flanders, D., Binder, S. (1994) "Retrospective Study of the
Impact of Lead-based Paint Hazard Remediation on Children's Blood Lead Levels hi St. Louis,
Missouri." American Journal of Epidemiology. 139(10):1016-1026.
Staessen, J., Yeoman, W.B., Fletcher, A.E., et al. (1990) "Blood Lead Concentration, Renal
Function, and Blood Pressure in London Civil Servants." British Journal of Industrial Medicine.
47:442-447.
Staessen, J., Sartor, F., Roels, H., et al. (1991) "The Association Between Blood Pressure,
Calcium and Other Divalent Cations: A Population Study." Journal of Human Hypertension.
5:485-494.
Stark, A.D., Meigs, J.W., Fitch, R.A., DeLouise, E.R. (1978) "Family'Operational Cofactors in
the Epidemiology of Childhood Lead Poisoning." Archives of Environmental Health. 33:222-
226.
Stark, A.D., Quah, R.F., Meigs, J.W., DeLouise, E.R. (1982) "The Relationship of
Environmental Lead to Blood Lead Levels hi Children." Environmental Research. 27:372-383.
Steenland, K., Selevan, S., Landrigan, P. (1992) "The Mortality of Lead Smelter Workers: An
Update." Am J Public Health. 82:1641-1644.
Stollery, B.T., Banks, H.A., Broadbent, D.E., et al. (1989) "Cognitive Functioning in Lead
Workers." British Journal of Industrial Medicine. 46:698-707.
Stuik, EJ. (1974) "Biological Response of Male and Female Volunteers to Inorganic Lead." Int
Arch Arbeitsmed. 33:83-97.
Thornton I, Davies D.J., Watt J.M., Quinn M.J. (1990) "Lead Exposure in Young Children From
Dust and Soil in the United Kingdom." Environmental Health Perspectives. 89:55-60.
Tola, S., Hernberg, S., Asp, S., et al. (1973) "Parameters Indicative of Absorption and Biological
Effect in New Lead Exposure: A Prospective Study." British Journal of Industrial Medicine.
30:134-141.
7-29
-------
Triebig, G., Weltle, D., Vanentin, H. (1984) "Investigations on Neurotoxicity of Chemical
Substances at the Workplace: V. Determination of the Motor and Sensory Nerve Conduction
Velocity in Persons Occupationally Exposed to Lead." International Archives of Occupational
Environmental Health. 53:189-204.
Tuppurainen, M, Wagar, G., Kurppa, K. (1988) "Thyroid Function as Assessed by Routine
Laboratory Tests of Workers with Long-term Lead Exposure." Scandinavian Journal of Work
Environment and Health. 14:175-180.
U.S. Department of Housing and Urban Development, Office of Policy Development and
Research. (1990) "Comprehensive and Workable Plan for the Abatement of Lead-Based Paint in
Privately Owned Housing," Report to Congress, December 1990.
U.S. Department of Housing and Urban Development (1995a) "The Relation of Lead-
Contaminated House Dust and Blood Lead Levels Among Urban Children." Volumes I and n.
Final Report to the U.S. Department of Housing and Urban Development Grant from The
University of Rochester School of Medicine, Rochester, New York, and The National Center for
Lead-Safe Housing, Columbia Maryland.
U.S. Department of Housing and Urban Development (1995b) "Guidelines for the Evaluation
and Control of Lead-Based Paint Hazards in Housing." Office of Lead-Based Paint Abatement
and Poisoning Prevention.
U.S. Environmental Protection Agency (1979) "The Environmental Lead Problem: An
Assessment of Lead in Drinking Water From a Multi-Media Perspective." EPA 570/9-79-003.
U.S. Environmental Protection Agency (1986) "Air Quality Criteria for Lead." Research
Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research and
Development, Office of Health and Environmental Assessment, Environmental Criteria and
Assessment Office. EPA 600/8-83-028F.
U.S. Environmental Protection Agency (1989a) "Review of the National Ambient Air Quality
Standards for Lead: Exposure Analysis Methodology and Validation," Office of Air Quality
Planning and Standards. EPA-450/2-89-011.
U.S. Environmental Protection Agency (1989b) "Evaluation of the Potential Carcinogenicity of
Lead and Lead Compounds: In Support of Reportable Quantity Adjustments Pursuant to
CERCLA Section 102." Office of Health and Environmental Assessment. EPA/600/8-89/045 A
(External Review Draft).
U.S. Environmental Protection Agency (1990a) "Air Quality Criteria for Lead: Supplement to
the 1986 Addendum." Office of Research and Development, Washington, D.C. EPA/600/8-
89/049F.
7-30
-------
U.S. Environmental Protection Agency (1990b). "National Air Quality and Emissions Trends
Report, 1988." Office of Air Quality Planning and Standards, U.S. Environmental Protection
Agency. EPA 450/4-90-002.
U.S. Environmental Protection Agency (1991). "National Air Quality and Emissions Trends
Report 1989."Office of Air Quality Planning and Standards, U.S. Environmental Protection
Agency. EPA 450/4-91-003.
U.S. Environmental Protection Agency (1994a) "Guidance Manual for the Integrated Exposure
Uptake Biokinetic Model for Lead in Children." Washington, DC. EPA/540/R-93/081, PB93-
963510.
U.S. Environmental Protection Agency (1994b) "Reducing Lead Hazards When Remodeling
Your Home." Office of Pollution Prevention and Toxics, U.S. Environmental Protection
Agency. EPA 747-R-94-002, April 1994.
U.S. Environmental Protection Agency (1995a) "Report on the National Survey of Lead-Based
Paint in Housing: Base Report." Office of Pollution Prevention and Toxics, U.S. Environmental
Protection Agency, EPA 747-R95-003, April 1995.
U.S. Environmental Protection Agency (1995b) "Review of Studies Addressing Lead Abatement
Effectiveness." Office of Pollution Prevention and Toxics, U.S. Environmental Protection
Agency. EPA 747-R-95-006, July 1995.
U.S. Environmental Protection Agency (1995c) "Sampling House Dust for Lead: Basic
Concepts and Literature Review." Office of Pollution Prevention and Toxics, U.S.
Environmental Protection Agency. EPA 747-R-95-007, September 1995.
U.S. Environmental Protection Agency (1995d) "Technical Support Document: Parameters and
Equations Used in the Integrated Exposure Uptake Biokinetic Model for Lead hi Children (v
0.99D)." EPA 540/R-94/040.
U.S. Environmental Protection Agency (1995e) "Comprehensive Abatement Performance Study,
Volume I: Summary Report." Office of Pollution Prevention and Toxics, U.S. Environmental
Protection Agency. EPA 230-R-94-013a.
U.S. Environmental Protection Agency (1995f) "Guidance on Identification of Lead-Based Paint
Hazards." Notice. Federal Register, pp 47248-47257, September 11,1995.
U.S. Environmental Protection Agency (1995g) "Report on the National Survey of Lead-Based
Paint hi Housing: Appendix I: Design and Methodology." Office of Pollution Prevention and
Toxics, U.S. Environmental Protection Agency, EPA 747-R95-004, April 1995.
7-31
-------
U.S. Environmental Protection Agency (1995h) "Report on the National Survey of Lead-Based
Paint in Housing: Appendix II: Analysis." Office of Pollution Prevention and Toxics, U.S.
Environmental Protection Agency, EPA 747-R95-005, April 1995.
U.S. Environmental Protection Agency (1995i) "Comprehensive Abatement Performance Pilot
Study." Volumes 1 and 2. EPA 747-R-93-007.
U.S. Environmental Protection Agency (1995J) "A Field Test of Lead-Based Paint Testing
Technologies: Summary Report." EPA 747-R-95-002a, May 1995.
U.S. Environmental Protection Agency (1996a) "Urban Soil Lead Abatement Demonstration
Project, Volume I: Integrated Report." Office of Research and Development, U.S.
Environmental Protection Agency. EPA 600-P-93-001aF, April 1996.
U.S. Environmental Protection Agency (1996b) "Lead-Based Paint Abatement and Repair and
Maintenance Study in Baltimore: Pre-Intervention Findings," EPA 747-R-95-012, August 1996.
U.S. Environmental Protection Agency (1996c) "Adjustments to the HUD National Survey Dust
Data for Section 403 Analyses," EPA 747-R-96-011, December 1996.
U.S. Environmental Protection Agency (1997) "Conversion Equations for Use in Section 403
Rulemaking." Office of Pollution Prevention and Toxics, U.S. Environmental Protection
Agency. EPA 747-R-96-0012.
U.S. Environmental Protection Agency (1998) "Review of Studies Addressing Lead Abatement
Effectiveness: Updated Edition." Office of Pollution Prevention and Toxics, U.S.
Environmental Protection Agency. EPA 747-B-98-001.
Valciukas, J.A., Lilis, R., Eisinger, J., et al. (1978) "Behavioral Indicators of Lead Neurotoxicity:
Results of a Clinical Field Survey." International Archives of Occupational Environmental
Health. 41:217-236.
Vershoor, M., Wibowo, A., Herber, R., et al. (1987) "Influence of Occupational Low-level Lead
Exposure on Renal Parameters." American Journal of Industrial Medicine. 12:341-351.
Vimpani, G.V., Wigg, N.R., Robertson, E.F., et al. (1985) "The Port Pirie Cohort Study: Blood
Lead Concentration and Childhood Developmental Assessment". Presented at: Lead
Environmental Health: Current Issues, May, Duke University, Durham, NC.
Vimpani, G.V., Baghurst, P.A., Wigg, N.R., et al. (1989) "The Port Pirie Cohort Study
Cumulative Lead Exposure and Neurodevelopmental Status at Age 2 Years: Do HOME Scores
and Maternal IQ Reduce Apparent Effects of Lead on Bayley Mental Scores?" In: Smith, M.,
Grant., L.D., Sors, A., eds. Lead Exposure and Child Development: An International
Assessment. Lancaster, UK: Kluwer Academic Press.
7-32
-------
Wada, O., Yano, Y, Ono, T., et al. (1973) "The Diagnosis of Different Degrees of Lead
Absorption in Special References to Choice and Evaluation of Various Parameters Indicative of
an Increased Lead Absorption." Industrial Health. 11:55-67.
Wallsten, T.S. and Whitfield, R.G. (1986) "Assessing the Risks to Young Children of Three
Effects Associated with Elevated Blood-lead Levels." Report by Argonne National Laboratory.
Report No. ANL/AA-32. Sponsored by the U.S. EPA Office of Air Quality Planning and
Standards.
Wang, L., Xu S.E., Zhang, G.D, Wang, W.Y. (1989) "Study of Lead Absorption and its Effect on
Children's Development." Biomedical and Environmental Sciences. 2:325-330.
Ward, N.I., Watson, R., Bryce-Smith, D. (1987) "Placental Element Levels in Relation to Fetal
Development for Obstetrically Normal Births: A Study of 37 Elements: Evidence for the Effects
of Cadmium, Lead, and Zinc on Fetal Growth and for Smoking as a Source of Cadmium." Int J
BiosocRes. 9:63-81.
Wedeen, R.P., Mallik, O.K., Batuman, V. (1979) "Detection and Treatment of Occupational
Lead Nephropathy." Archives of Internal Medicine. 139:53-57.
Wedeen, R.P. (1988) "Bone Lead, Hypertension, and Lead Nephropathy." Environmental Health
Perspectives. 78:57-60.
Weiss, S.T., Munoz, A., Stein, A., et al. (1986) "The Relationship of Blood Lead to Blood
Pressure in a Longitudinal Study of Working Men. " Am JEpidemiol 123:800-808.
Weiss, S.T., Munoz, A., Stem, A., et al. (1988) "The Relationship of Blood Lead to Systolic
Blood Pressure hi a Longitudinal Study of Policemen." Environmental Health Perspectives.
78:53-56.
Weitzman, M., Aschengrau, A., Bellinger, D., Jones, R., Hamlin, J.S., Beiser, A. (1993) "Lead-
Contaminated Soil Abatement and Urban Children's Blood Lead Levels." Journal of the
American Medical Association. 269(13):1647-1654.
Westat (1995) "NCLSH Westat Blue Nozzle Study." Prepared by Westat, Inc. for the National
Center for Lead-Safe Housing, revised January, 1995.
Wibberley, D.G., Khera, A.K., Edwards, J.H., et al. (1977) "Lead Levels in Human Placentae
from Normal and Malformed Births." JMed Genet. 14:339-345.
Wigg, N.R., Vimpani, G.V., McMichael, A.J., et al. (1988) "Port Pirie Cohort Study: Childhood
Blood Lead and Neuropsychological Development at Age Two Years." JEpidemiol Community
Health. 42:213-219.
7-33
-------
Wildt, K., Eliasson, R., Berlin, M. (1983) "Effects of Occupational Exposure to Lead on Sperm
and Semen." In: Clarkson, T.W., Nordberg, G.F., Sager, P.R., eds. Reproductive and
Developmental Toxicity of Metals. Proceedings of a Joint Meeting, Rochester, NY, May 1982.
New York, NY: Plenum Press. 279-300.
Wilhelm, M., Lombeck, I., Hafher, D., et al. (1989) "Hair Lead Levels in Young Children from
the F.R.G." JTrace Elements and Electrolytes in Health and Disease, 3:165-170.
Williamson, A.M. and Teo, R.K.C. (1986) "Neurobehavioral Effects of Occupational Exposure
to Lead " British Journal of Industrial Medicine. 43:374-380.
Windebank, A.J., et al. (1980) "The Endoneurial Content of Lead Related to the Onset and
Severity of Segmental Demyelination." J. Neuropathol. Exp. Neural. 38:692-699.
Winneke, G., Beginn, U. Ewert, T., et al. (1984) "Understanding of Subvclinical lead Effects on
the Nervous System of Children with Known Prenatal Exposure in Nordenham" Schriftenr.
Ver. Wasser Boden Lufthyg. 59:215-30. (German)
Winneke, G., Beginn, U., Ewert, T., et al. (1985a) "Comparing the Effects of Perinatal and Later
Childhood Lead Exposure on Neurophysiological Outcome." Environmental Research. 38:155-
167.
Winneke, G., Brockhaus, A., Collet, W., et al. (1985b) "Predictive Value of Different Markers of
Lead-exposure for Neuropsychological Performance." In: Lekkas, T.D., ed. International
Conference on Heavy Metals in the Environment, Athens, Greece, September, Vol. 1. Edinburgh,
United Kingdom: CEP Consultants, Ltd., 44-47.
Winneke, G., Brockhaus, A., Ewers, V., Kramer, V., and Neuf, M. (1990) "Results from the
European Multicenter Study on Lead Neurotoxicity in Childen: Implications for Risk
Assessment." Neurotoxicol Teratol 12:553-559.
Wolf, A.W., Ernhart, C.B., White, C.S. (1985) "Intrauterine Lead Exposure and Early
Development." hi: Lekkas T.D., ed. International Conference: Heavy Metals in the
Environment, Athens, Greece, September, Vol. 2. Edinburgh, United Kingdom: CEP
Consultants, Ltd. 153-155.
Wrenn, M.E., Cohen, M., Rosen, J.C., Eisenbud, M., Blanchard, R.L. (1972) "In-vivo
Measurements of Lead-210 in Man. Assessment of Radioactive Contamination in Man." IAEA.
129-146.
Yule, W., Lansdown, R., Millar, I., Urbanowiez, M. (1981) "The Relationship Between Blood
Lead Concentration, Intelligence, and Attainment in a School Population: a Pilot Study." Dev
Med Child Neuro. 23:567-576.
Zimmennan-Tansella, C., Campara, P., D'Andrea, F., et al. (1983) "Psychological and Physical
Complaints of Subjects with Low Exposure to Lead." Hum Toxicol. 2:651-623.
7-34
-------
APPENDIX A
Glossary
A-1
-------
APPENDIX A
GLOSSARY
Abatement: Any set of measures designed to permanently eliminate lead-based paint hazards
in accordance with standards established by Federal agencies.
* Accessible or Chewable Surface: An interior or exterior surface painted with lead-based paint
that is accessible for a young child to mouth or chew.
Arithmetic Mean: The sum of a set of measurements divided by the number of measurements.
Background Lead Exposure: Exposure to environmental lead that is not the result of human
activity such as lead-based paint or industrial sources.
Baseline: Conditions prior to implementing interventions in response to §403 rules. Baseline
risk characterization is performed in this risk analysis using blood-lead concentration data from
Phase 2 of NHANES in and by assumptions on the relationship between blood-lead
concentration and IQ score decrement.
Biokinetics: Processes affecting the movement of molecules from one internal body
compartment to another, including elimination from the body.
Blood-Lead Concentration: Blood-lead concentration measures the mass of lead collected per
volume of whole blood collected and is usually expressed in terms of micrograms of lead
collected per deciliter of blood collected (ug Pb/dL blood).
Blue Nozzle Sampler: Refers to the vacuum sampler used to collect dust samples hi the HUD
National Survey and the Baltimore R&M Pilot study. The sampling flow rate is cited as 16 liters
per minute. The sampler consists of a rotary vane pump connected to the same filter and
sampling cassette used hi the DVM sampler.
Body-Lead Burden: The level of lead carried hi a body.
BRM Sampler: Refers to the vacuum sampler developed and utilized to collect dust hi EPA's
Baltimore Repair and Maintenance Study. It is a modified version of the HVS3 sampler,
employing a portable handheld vacuum and other modifications to make it easy to use and to
access small areas.
Confidence Interval: An interval that contains the true value of a parameter with a certain
degree of confidence.
* As defined in Section 1001 of Tide X and Section 401 of TSCA Title IV amendment.
A-2
-------
Conversion Factors: Use of regression models in this risk analysis to convert observed lead
measurements from one format to another, typically to correct for differences in dust collection
method. For example, a conversion factor was used to express the Blue Nozzle vacuum dust-
lead loadings reported in the HUD National Survey as wipe-equivalent dust-lead loadings in
order to determine which housing units in the HUD National Survey exceeded example standards
for wipe dust-lead loadings.
Cumulative Distribution Function (CDF): For any number x, the CDF F(x) of a random
variable X is the probability that the observed value of X will be at most x.
Deteriorated Paint: Any interior or exterior paint that is peeling, chipping, chalking or
cracking or any paint located on an interior or exterior surface or fixture that is damaged or
deteriorated.
Dripline Soil Sample: Any soil sample collected from the drip line area about the residence.
This is usually approximately 1-3 feet from the side (e.g. foundation) of the house, under the
eaves.
Drv Room: (see Wet Room).
Dust Abatement: Removing settled dust from a housing unit using HEP A vacuums and wet
mopping.
Dust Cleaning: Intervention where settled dust that is likely to be lead-contaminated is removed
from residential surfaces using HEP A vacuums and wet mopping.
Dust-Lead Loading: Dust-lead loading measures the mass of lead collected per surface area
sampled and is usually expressed in terms of micrograms of lead collected per square foot
sampled (ug Pb/ft2).
Dust-Lead Concentration: Dust-lead concentration measures the mass of lead collected per
mass of dust collected and is usually stated in terms of micrograms of lead collected per gram of
dust collected (ug Pb/g dust).
DVM Sampler: A device used to collect dust samples using a vacuum (personal air sampler)
operating at a rate of two to three liters of air per minute. It was designed to collect only dust that
would most likely stick to a child's hand, not total lead on a surface. Thus, it tends to have low
collection efficiency for particles larger than 250 microns.
* As define in Section 1001 of Title X and Section 401 of TSCA Title IV amendment.
A-3
-------
Efficacy: Refers to the effectiveness of a method of abatement and is defined as the generalized
evaluation of several key factors including the usability of a method, its hazard abatement
effectiveness, and the amount of hazardous dust lead generated by a method, measured by air and
post-cleanup wipe samples.
Empirical Model: A statistical regression model developed for this risk analysis from data
collected in the Rochester lead-in-Dust study. The resulting model which predicts geometric
mean blood-lead concentration for children aged 12-30 months as a function of environmental
lead levels (dust-lead loading, soil-lead concentration, extent of deteriorated lead-based paint
hazard) is used to predict a national distribution of children's blood-lead levels for this age
group.
Encapsulation: A method of "abatement" that involves the coating and sealing of surfaces with
durable coatings formulated to be elastic, long-lasting (e.g., at least 20 years), and resistant to
cracking, peeling, algae, and fungus.
Enclosure: The resurfacing or covering of surfaces by sealing or caulking them with
mechanically affixed, durable materials so as to prevent or control chalking, flaking, lead-
containing substances from being part of house dust or accessible to children.
Entrvwav Soil: Any soil sample collected immediately adjacent to the entryway of the
residence.
EPI Study: A targeted epidemiology study which measures both children's blood-lead
concentrations and environmental lead levels as well as other factors (e.g., behavioral,
demographic) influencing a child's blood-lead level.
Epidemiology: In broad terms epidemiology is concerned with the distribution of disease, and it
is now customary to include within its orbit the study of chronic disease as well as communicable
diseases which give rise to epidemics of the classical sort.
Expected Value: The average value of a statistic if it were calculated from an infinite number of
equal-sized samples from a given population.
Exposure Route: The manner by which a chemical or pollutant enters an organism after contact
(e.g., by ingestion, inhalation).
Exposure Pathway: The physical course a chemical or pollutant takes from its source to the
organism exposed.
Exposure: Contact between a chemical, physical, or biological agent (e.g., lead) with the outer
boundary of an organism (e.g., a child's skin). Exposure is quantified as the concentration of the
agent in the medium in contact integrated over the time duration of that contact.
A-4
-------
Friction Surface: An interior or exterior surface that is subject to abrasion or friction,
including certain window, floor, and stair surfaces.
Geometric Mean: The n* root of the product of n values. Also, the exponentiation of the
"arithmetic mean" of a set of n natural log-transformed values.
Geometric Standard Deviation (GSD): The exponentiation of the "standard deviation" of a set
of n natural log-transformed values.
HEPA: A High Efficiency Paniculate Accumulator vacuum used in dust cleaning, fitted with a
filter capable of filtering out particles of 0.3 microns or greater from a body of air at 99.97
percent efficiency or greater.
Histogram: A bar graph associating frequencies or relative frequencies with data intervals. The
values of the variable are by convention represented on the horizontal scale, and the
vertical scale represents the frequency or relative frequency of data values in each standard
grouping of possible values for the variable. It illustrates the general shape of the observed data
distribution.
Human Exposure Studies: Studies which investigate the association between elevated blood-
lead concentration and elevated levels of lead in a child's residential environment. Examples of
human exposure studies are the Rochester Lead-in-Dust study and the Brigham and Women's
Hospital Longitudinal study.
HVS3 Vacuum Sampler: Vacuum method originally developed to measure pesticides in house
dust and is now recognized as an ASTM standard for collecting floor dust samples to be analyzed
for lead content.
IEUBK Model: EPA's Integrated Exposure Uptake Biokinetic Model for Lead, designed to
model exposure from lead in air, water, soil, dust, diet, and paint and other sources using
pharmacokinetic modeling methods to predict blood-lead concentrations in children 6 months to
7 years of age.
*Impact Surface: An interior or exterior surface that is subject to damage by repeated impacts,
for example, certain parts of door frames.
Individual Risks: Hazards posed for children exposed to specified levels of lead in certain
media within the residential environment.
As define in Section 1001 of Title X and Section 401 of TSCA Title IV amendment
A-5
-------
Intelligence Quotient (IO): A score used to express the apparent relative intelligence of a
person determined by dividing his/her mental age as reported on a standardized test by his/her
chronological age and multiplying by 100. This risk analysis used IQ score decrement as a
means of measuring the neurological effects of lead.
Intercept: See Slope.
*
Interim Controls: A set of measures designed to temporarily reduce human exposure or likely
human exposure to lead-based paint hazards, including specialized cleaning, repairs,
maintenance, painting, temporary containment; ongoing monitoring of lead-based paint hazards
or potential hazards, and the establishment and operation of management and resident education
programs.
Intervention: A procedure implemented to reduce or eliminate a lead-based paint hazard within
a specific medium within a residence, when some type of mechanism is triggered for that
medium (e.g., dust-lead standard is exceeded). Interventions considered in this risk analysis
include dust cleaning, soil removal, paint maintenance, and paint abatement.
Intervention Studies: Studies which investigate the impact on children's blood-lead
concentration of reducing childhood lead exposure via a range of intervention strategies.
Intervention studies can contribute to conclusions about whether specific lead exposures are the
cause behind elevated blood-lead concentration. Examples of intervention studies are the
Baltimore R&M study and the Urban Soil-Lead Abatement Demonstration Project.
*Lead-Based Paint Hazard: Any condition that causes exposure to lead from lead-
contaminated dust, lead-contaminated soil, lead-contaminated paint that is deteriorated or present
in accessible surfaces, friction surfaces, or impact surfaces that would result in adverse human
health effects as established by EPA.
*Lead-Based Paint (LBP): Dried paint film that has a lead content exceeding 1.0 mg/cm2 or 0.5
percent (5,000 parts per million (ppm)) by weight.
*Lead-Contaminated Soil: Bare soil on residential real property that contains lead at or in
excess of the levels determined to be hazardous to human health by EPA.
Lead-Contaminated Dust: Surface dust in residential dwellings that contains an area or mass
concentration of lead in excess of levels determined by EPA to pose a threat of adverse health
effects in pregnant women or young children.
Log-Linear Regression Model: A regression model in which the natural logarithm of the
independent (predictor) variables is taken before fitting the model.
As define in Section 1001 of Title X and Section 401 of TSCA Title IV amendment
A-6
-------
Lognormal Distribution: A nonnegative random variable X is said to have a lognormal
distribution if the natural logarithm of X has a normal (Gaussian) distribution.
, Minimum and Range: The largest and smallest observations in a data distribution
are called maximum and minimum respectively. The difference between the maximum value and
the minimum value is defined as the range.
Measurement Error: Error in an observed measurement attributable to sampling, laboratory,
spatial and/or temporal variability.
Measurement Error Model: A regression model which attempts to account for measurement
error in the observed predictor variables.
Meta-Analysis: The statistical integration of the results of independent studies.
Microgram (ug): A microgram is 1/1,000,000 of a gram or 1/1,000 of a milligram.
Monte Carlo Analysis: An estimation method where approximations are obtained by repeated
random sampling or simulation.
Negative Predictive Value (NFV1: Probability of a resident child having a blood-lead
concentration below some specified threshold value, given that observed lead levels in a
specified medium within the dwelling is below the standard for that medium.
90% Confidence Bound on a Statistic: The upper and lower limits of a 90% confidence
interval.
Paint Maintenance: Intervention where all surfaces with deteriorated lead-based paint are
repaired by feathering the edges of deteriorating paint and repainting with new, lead-free paint.
Paint Abatement: Intervention where all surfaces with deteriorated lead-based paint are
encapsulated, enclosed, or removed using currently acceptable practices and materials.
Parameter: A characteristic of a population, such as the population mean or variance.
Percentile: A particular value in a set or distribution of numbers for which a specified
percentage of the numbers are less than the given value. For instance, the 5th percentile of a set
of blood-lead concentrations is the blood-lead concentration value such that 5% of the numbers
are less man the value and 95% are greater than it. The 50th percentile is also known as the
median.
Performance Characteristics Analysis: An analysis used to characterize the performance of
options for the §403 standards based on the data from IEUBK model or Empirical model.
A-7
-------
Perimeter Soil Sample: Any soil sample collected from the perimeter or remote areas of the
residence's yard. (Note: in the Rochester Lead-in-Dust study, this terminology referred to
samples collected adjacent to the foundation).
Pharmacokingtics; The study of the time course of absorption, distribution, metabolism, and
excretion of a foreign substance (e.g., a drug or pollutant) in an organism's body.
Pica: An abnormal tendency to mouth or attempt to consume non-food objects, such as paint
chips.
Piecewise Linear Function: The domain of a function divided into finite pieces such that in
each piece the function is linear.
Play-yard Soil Sample: Any soil sample collected in areas where the child usually played. In
the HUD National Survey, this was frequently a local playground. In other studies, this refers to
an exterior site at the residence.
Population: A population of items is defined to be any set of items for which one wants to study
and make inferences. Associated with each item in a population are one or more numbers or
attributes of interest, which are called variables.
Population Risks: Hazards posed by childhood lead exposure to our nation as a whole.
Positive Predictive Value flPPV): Probability of a resident child having a blood-lead
concentration above some specified threshold value given that observed lead levels in a specified
medium within the dwelling is above the standard for that medium.
Primary Prevention Intervention: A primary prevention intervention prevents human
exposure before it occurs (e.g. paint abatement occurs in the home before a new family with
children moves in).
Probability Samples: Samples selected from a statistical population such that each sample has a
known probability of being selected.
Probability: Given an experiment with an associated sample space, the objective of probability
is to assign to each event a number, which will provide a measure of the likelihood that A will
occur when the experiment is performed.
Random Samples: Samples selected from a statistical population such that each sample has an
equal probability of being selected.
A-8
-------
Reduction: Measures designed to reduce or eliminate human exposure to lead-based paint
hazards through methods including interim controls and abatement.
Regression Model: A statistical representation of the relationship between a dependent variable
such as blood-lead concentration to one or more independent variables such as environmental
lead exposures. For example, a regression model could indicate that blood-lead concentration is
an additive function of environmental lead levels.
Removal and Replacement: A method of abatement that entails removing substrates such as
windows, doors, trim, or soil that have lead-contaminated surfaces and installing new (and
presumably lead-free) or deleaded components.
Residual Error: The difference between the modeled predicted value of a random variable
under specified conditions and the observed value of that variable under the same conditions.
Risk: The probability of deleterious health or environmental effects.
Risk Assessment: Within the context of this risk analysis report, risk assessment is that portion
of the risk analysis consisting of hazard identification (Chapter 2), exposure assessment (Chapter
3), dose-response assessment (Chapter 4), and risk characterization (Chapter 5). Within the
context of identifying lead-based paint hazards in a residence, risk assessment is an on-site
investigation to determine and report the existence, nature, severity, and location of lead-based
paint hazards within a specific residential dwelling.
Rochester Multimedia Model: A regression model obtained in the process of developing the
"empirical model" (using data from the Rochester Lead-in-Dust study) which expresses blood-
lead concentration for children aged 12-31 months as a function of environmental-lead levels
(dust-lead loading, soil-lead concentration, extent of deteriorated lead-based paint hazard). This
model differs from the empirical model in that it does not take into account measurement error in
the predictor variables and assumes dust-lead loadings are based on wipe collection techniques.
This model was used to characterize individual risks in this risk analysis.
Sample: A small part of something designed to show the nature or quality of the whole.
Exposure-related measurements are usually samples of environmental or ambient media,
exposures of a small subset of a population for a short time, or biological samples, all for the
purpose of inferring the nature and quality of parameters important to evaluating exposure.
* As define in Section 1001 of Title X and Section 401 of TSCA Title IV amendment.
A-9
-------
Sampling Weights: In a complex survey design, a sampling weight is assigned to a sampling
unit to denote the total number of units in the population that is represented by that sampling
unit. Sampling weights are necessary to make results of the survey representative of the
population. For example, the sampling weight assigned to one of the 284 households in the HUD
National Survey represents the number of homes that house represents nationally.
Secondary Prevention Intervention: A secondary prevention intervention reduces or
eliminates human exposure on behalf of humans already exposed to the targeted hazard (e.g.
paint abatement occurs in the home of a child who has an elevated blood-lead concentration).
Sensitivity Analysis: An investigation to determine the extent to which variations in key
assumptions and approaches affect the results and conclusions of the analysis.
Sensitivity and Specificity: Sensitivity is the probability of a dwelling being above the media
standards (e.g., soil lead, dust lead, etc.) given that there is a resident child with blood
concentration above some specified threshold value. Specificity is the probability of a dwelling
being below the media standards given that there is a resident child with blood concentration
below some specified threshold value.
Sirchee-Spittler Sampler: Vacuum method used to collect dust samples in the Boston and
Baltimore phases of EPA*s Urban Soil-Lead Abatement Demonstration Project (USLADP). It is
a hand-held, battery-powered vacuum unit designed to collect forensic evidence.
Slope: If the regression model is a simple regression model such that y=a+f3x+e, then (3 is called
the slope, and a is called the intercept. The slope is interpreted as the amount by which y
changes when x is increased by one unit.
Soil Removal: Intervention where soil from areas with elevated lead concentrations are removed
and replaced with clean soil, or the areas are permanently covered.
Soil-Lead Concentration: A measure of the mass of lead collected per mass of soil collected
and is usually stated in terms of micrograms of lead collected per gram of soil collected (ug Pb/g
soil). These units are also sometimes referred to as parts per million (ppm).
Standard Error: The standard deviation of errors around a fitted regression model.
Standard Deviation: A measure of the dispersion of a set of values that is the square root of the
"arithmetic mean" of the squares of the deviation of each value from the "arithmetic mean" of
the values.
Subpopulation: A subset of the population of interest that is used for analysis. Usually the
subpopulation is taken to be representative of the entire population.
A-10
-------
Tails: The portion of a distribution containing extreme values are called the tails of the
distribution.
Tap Weight: The weight of the dust that was tapped out of the blue nozzle vacuum cassette and
analyzed for lead. Note that a dust sample's tap weight is lower than its actual weight, as some
dust may remain in the cassette.
Target Housing: Any housing constructed prior to 1978, except for housing of the elderly or
persons with disabilities (unless any child who is less than 6 years of age resides or is expected to
reside in such housing for the elderly or persons with disabilities), or any 0-bedroom dwelling.
Threshold: The value above which something is true or will take place and below which it is
not or will not.
True Negative Rate: Alternative terminology for specificity.
True Positive Rate: Alternative terminology for sensitivity.
Uptake: The process by which a substance is absorbed into the body.
Vacuum Sample: Collecting dust over a specified area by vacuuming the area. The contents of
the vacuum bag or filter cassette are then analyzed for the amount of dust and the amount of lead.
Results from vacuum sampling can be expressed as "dust-lead loadings" or "dust-lead
concentrations".
Variability: A measure used to describe how data vary about the center of the distribution. It
also tells the spread of the data.
Wet Room: An interior room hi a house which is either a kitchen, bathroom, laundry, or utility
room is classified as a 'wet room', otherwise the room may be classified as a 'dry' room.
Terminology used in the HUD National Survey.
Window Sill: The portion of the horizontal window ledge that protrudes into the ulterior of the
room, adjacent to the window sash when the window is closed.
Window Trough: The portion of the horizontal window sill that receives the window sash when
the window is closed, often located between the storm window and the interior window sash.
This is also sometimes referred to as a window well.
As define in Section 1001 of Title X and Section 401 of TSCA Title IV amendment
A-11
-------
Wipe Sample: Dust that is collected over a specified area by wiping the area with a moist cloth.
The cloth and the dust on the cloth are then analyzed for the amount of lead. Results from wipe
sampling are in the form of "dust-lead loadings." Section 403 standards for lead in dust will
likely be specified in terms of wipe dust-lead loadings.
XRF: "X-ray fluorescence" is a principle used by instruments to determine the lead
concentration in substances, usually in milligrams of lead per square centimeter of surface area
(mg/cm2).
A-12
-------
50272-101
REPORT DOCUMENTATION
PAGE
1. REPORT NO.
EPA 747-R-97-006
3. Recipient's Accession No.
4. Title and Subtitle
Risk Analysis to Support Standards for Lead in Paint, Dust, and Soil
5. Report Date
June 1998
7. Author(s) Menton, RG, Lordo, RA, McMillan, N, Niemuth, NA, Burgoon, DA, Kinateder, JG,
Strauss, WJ and Wood, BJ
8. Performing Organization
Rept. No.
9. Performing Organization Name and Address
Battelle Memorial Institute
505 King Avenue
Columbus, Ohio 43201-2693
10. Project/Task/Work Unit No.
G003470-06
11. Contract(C) or Grant(G) No.
(C) 68-D5-0008
(G)
12. Sponsoring Organization Name and Address
U.S. Environmental Protection Agency
Office of Pollution Prevention and Toxics
401 M Street, S.W.
Washington, D.C. 20460
13. Type of Report & Period
Covered
Technical Report
14.
15. Supplementary Notes
Support staff involved in the risk analysis and the production of this report included John Menkedick, Ying-Uang Chou, Greg Stark,
James Ma, Agnes Kovacs, and Steve Naber.
16. Abstract (Limit 200 words)
Title X of the Housing and Community Development Act, known as the Residential Lead-Based Paint Hazard Reduction Act of 1992,
contains legislation designed to evaluate and reduce exposures to lead in paint, dust, and soil in the nation's housing. As part of Title X,
the Toxic Substances Control Act (TSCA) was amended to include Title IV, 'Lead Exposure Reduction*. Section 403 of TSCA requires
EPA to define standards for lead in paint, dust, and soil. Federal, state, and local public health agencies, as well as private property
owners and other private sector interests, will use these standards to determine in which homes actions should be taken to reduce or
prevent the threat of childhood lead poisoning.
This report documents the scientific basis for the proposed §403 standards. First, the report summarizes EPA's assessment of the health
risks to young children from exposures to lead-based paint hazards, lead-contaminated dust, and lead-contaminated soil. Second, the
report documents the approach developed by EPA to estimate the reduction in hearth risks expected to take place following promulgation
of the $403 standards. Finally, the report provides estimates of the numbers of homes and children that will be affected by various
example options for the standards.
17. Document Analysis
a. Descriptors
Blood-lead concentration, childhood health risks, dust lead concentration, lead hazards, lead based paint, risk assessment, risk
characterization, risk management
b. Identifiers/Open-Ended Terms
abatement, dose-response assessment, exposure assessment, hazard identification, HUD National Survey, interventions,
NHANES III, Section 403, sensitivity and uncertainty analysis. Title X
c. COSATI Field/Group
18. Availability Statement
Release Unlimited
19. Security Class (This Report)
Unclassified
20. Security Class (This Page)
Unclassified
21. No. of Pages
530
22. Price
(SeeANSI-239.18)
OPTIONAL FORM 272 (4-77)
(Formerly NTIS-35)
Department of Commerce
------- |