Integrated Science Assessment for Lead Appendix 3: Nervous System Effects External Review Draft March 2023


*>EPA

EPA/600/R-23/061

United States	.. , orm

Environmental Protection	iviarcnzuzj

Agency	www.epa.gov/isa

Integrated Science
Assessment for Lead

Appendix 3: Nervous System Effects
External Review Draft

March 2023

Health and Environmental Effects Assessment Division
Center for Public Health and Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency

-------
DISCLAIMER

This document is an external review draft for peer review purposes only. This information is
distributed solely for the purpose of predissemination peer review under applicable information quality
guidelines. It has not been formally disseminated by the Environmental Protection Agency. It does not
represent and should not be construed to represent any agency determination or policy. Mention of trade
names or commercial products does not constitute endorsement or recommendation for use.

External Review Draft

3-ii

DRAFT: Do not cite or quote

-------
DOCUMENT GUIDE

This Document Guide is intended to orient readers to the organization of the Lead (Pb) Integrated Science
Assessment (ISA) in its entirety and to the sub-section of the ISA at hand (indicated in bold). The ISA consists of
the Front Matter (list of authors, contributors, reviewers, and acronyms), Executive Summary, Integrated Synthesis,
and 12 appendices, which can all be found at https ://cfpub. epa. gov/ncea/isa/recordisplav .cfm?deid=3 57282.

Front Matter

Executive Summary

Integrative Synthesis

Appendix 1. Lead Source to Concentration

Appendix 2. Exposure, Toxicokinetics, and Biomarkers

Appendix 3. Nervous System Effects

Appendix 4. Cardiovascular Effects

Appendix 5. Renal Effects

Appendix 6. Immune System Effects

Appendix 7. Hematological Effects

Appendix 8. Reproductive and Developmental Effects

Appendix 9. Effects on Other Organ Systems and Mortality

Appendix 10. Cancer

Appendix 11. Effects of Lead in Terrestrial and Aquatic Ecosystems
Appendix 12. Process for Developing the Pb Integrated Science Assessment

External Review Draft

3-iii

DRAFT: Do not cite or quote

-------
CONTENTS

LIST OF TABLES	3-v

LIST OF FIGURES	3-vii

ACRONYMS AND ABBREVIATIONS	3-viii

APPENDIX 3 NERVOUS SYSTEM EFFECTS	3-1

3.1	Introduction 	3-2

3.2	Scope	3-2

3.3	Biological Plausibility	3-4

3.4	Overt Nervous System Toxicity	3-14

3.4.1	Epidemiologic Studies of Brain Structure and Function 	3-14

3.4.2	Experimental Animal Studies of Brain Structure and Function	3-18

3.4.3	Integrated Summary of Overt Nervous System Toxicity	3-25

3.5	Nervous System Effects Ascertained during Childhood, Adolescent, and Young Adult

Lifestages	3-25

3.5.1	Cognitive Function in Children 	3-26

3.5.2	Externalizing Behaviors: Attention, Impulsivity, and Hyperactivity in Children 	3-86

3.5.3	Externalizing Behaviors: Conduct Disorders, Aggression, and Criminal Behavior in
Children, Adolescents, and Young Adults	3-112

3.5.4	Internalizing Behaviors: Anxiety and Depression in Children	3-124

3.5.5	Motor Function in Children	3-137

3.5.6	Sensory Organ Function in Children	3-150

3.5.7	Social Cognition and Behavior in Children	3-159

3.6	Nervous System Effects Ascertained during Adult Lifestages	3-171

3.6.1	Cognitive Function in Adults	3-171

3.6.2	Psychopathological Effects in Adults	3-191

3.6.3	Sensory Organ Function in Adults	3-201

3.6.4	Neurodegenerative Diseases	3-209

3.7	Evidence Inventories - Data Tables to Summarize Study Details	3-225

3.8	References	3-493

External Review Draft

3-iv

DRAFT: Do not cite or quote

-------
LIST OF TABLES

Table 3-1
Table 3-2
Table 3-3
Table 3-4

Table 3-5

Table 3-6

Table 3-7

Table 3-8

Table 3-9

Table 3-10

Table 3-11

Table 3-12

Table 3-1E
Table 3-1T
Table 3-2E
Table 3-3E
Table 3-4E

Table 3-4T
Table 3-5E

Statistics associated with the international pooled analysis of data from seven

cohort studies.	3-28

Summary of evidence Indicating a causal relationship between Pb exposure and
cognitive effects in children.	3-83

Summary of evidence indicating a causal relationship of Pb exposure with
attention, impulsivity, and hyperactivity.	

Summary of evidence indicating a likely causal relationship between Pb
exposure and motor function in children.	

Evidence that is suggestive of, but not sufficient to infer, a causal relationship
between Pb exposure and sensory organ function in children.	

Evidence that is suggestive of, but not sufficient to infer, a causal relationship
between Pb exposure and social cognition and behavior in children.	

Animal toxicological studies of Pb exposure and cognitive function.

3-108

Summary of evidence for a likely to be causal association between Pb exposure

and conduct disorders, aggression, and criminal behavior in children and

adolescents.	3-122

Summary of evidence for a likely to be causal relationship between Pb exposure

and internalizing behaviors in children.	3-135

3-148

3-157

3-169

Summary of evidence for a likely causal relationship between Pb exposure and
cognitive effects in adults.	3-189

Summary of evidence for a likely to be causal relationship between Pb exposure

and psychopathological effects in adults.	3-200

Summary of the evidence that is suggestive of, but not sufficient to infer, a causal
relationship between sensory function in adults. 	3-207

Summary of evidence that is suggestive of, but not sufficient to infer, a causal
relationship between Pb exposure and neurodegenerative diseases in adults.	3-223

Epidemiologic studies of Pb exposure and overt nervous system toxicity. 	3-225

Animal toxicological studies of Pb exposure and brain function.	3-230

Epidemiologic studies of Pb exposure and full-scale intelligence quotient.	3-256

Epidemiologic studies of Pb exposure and infant development. 	3-273

Epidemiologic studies of Pb exposure and performance on neuropsychological

tests of cognitive function, i.e., learning, memory, and executive function.	3-280

3-290

Epidemiologic studies of Pb exposure, academic performance, and achievement. _3-317

External Review Draft

3-v

DRAFT: Do not cite or quote

-------
Table 3-6E

Epidemiologic studies of Pb exposure and cognitive effects: population or group
mean blood Pb levels >5 |jg/dl_.	3-324

Table

3-7E

Epidemiologic studies of Pb exposure and performance on neuropsychological
tests of attention, impulsivity, and hyperactivity, ADHD-related behaviors, and
clinical ADHD in children.

3-340

Table

3-7T

Animal toxicological studies of Pb exposure and externalizing and internalizing
behaviors.

3-358

Table

3-8E

Epidemiologic studies of Pb exposure and performance on neuropsychological
tests of attention, impulsivity, and hyperactivity, Attention Deficit Hyperactivity
Disorder-related behaviors, and clinical Attention Deficit Hyperactivity Disorder in
children; group or population mean blood Pb level >5 |jg/dl_, any study design.

3-366

Table

3-9E

Epidemiologic studies of Pb exposure and externalizing behaviors including
conduct disorders, aggression, and criminal behavior in children and
adolescents.

3-377

Table

3-1OE

Epidemiologic studies of Pb exposure and internalizing behaviors in children.

3-391

Table

3-11E

Epidemiologic studies of Pb exposure and motor function in children.

3-402

Table

3-11T

Animal toxicoloaical studies of Pb exposure and motor function.

3-418

Table

3-12E

Epidemiologic studies of Pb exposure and sensory organ function in children.

3-427

Table

3-13E

Epidemiologic studies of Pb exposure, social cognition, and behavior in children.

_ 3-436

Table

3-14E

Epidemiologic studies of exposure to Pb and coanitive function in adults.

3-446

Table

3-15E

Epidemiologic studies of Pb exposure and psychopathological effects in adults. _

_ 3-456

Table

3-16E

Epidemiologic studies of Pb exposure and sensorv oraan function in adults.

3-463

Table

3-16T

Animal toxicoloaical studies of Pb exposure and sensorv oraan function.

3-470

Table

3-17E

Epidemiologic studies of exposure to Pb and neurodegenerative disease in
adults.

3-472

Table 3-17T Animal toxicological studies of Pb exposure and neurodegeneration.	3-488

External Review Draft

3-vi

DRAFT: Do not cite or quote

-------
LIST OF FIGURES

Figure 3-1 Potential biological pathways for nervous system effects following developmental

exposure to Pb.	3-5

Figure 3-2 Potential biological pathways for nervous system effects following postweaning

exposure to Pb.	3-6

Figure 3-3 The relationship between blood Pb level at age 11 and brain outcomes in

adulthood.	3-16

Figure 3-4 Associations between blood Pb levels and full-scale intelligence quotient in

children.	3-31

Figure 3-5 Associations between biomarkers of Pb exposure and Bayley Score of Infant

Development Mental Development Index	3-38

Figure 3-6 Association of blood Pb level with reading and math scores among North

Carolina school children (third to eighth grades).	3-60

Figure 3-7 Relationship between concurrent blood Pb level and intelligence quotient among

Italian adolescents using a cubic spline fit.	3-62

Figure 3-8 Relationship between log-transformed blood Pb level and intelligence quotient

using an ordinary least squares fit.	3-63

Figure 3-9 Two distributions of intelligence test scores demonstrating the consequence in a

small shift in the mean score.	3-70

Figure 3-10 Scatter plots and regression lines of blood Pb level and 18-month Mental

Developmental Index among children in manganese (A) quintiles 1-4 and (B)

quintile 5.	3-74

Figure 3-11 Mean ± standard deviation behavior performance in the Go/No-Go task

according to quartiles of exposure for (A and B) cord blood Pb and (C) childhood
blood Pb level at age 11 years.	3-92

Figure 3-12 Associations of monthly airborne Pb exposure levels from birth to age 12 with
scores for anxiety and depression behaviors on the Behavior Assessment
System for Children.	3-127

Figure 3-13 Associations between biomarkers of Pb exposure and Bayley Score of Infant

Development Psychomotor Development Index	3-139

Figure 3-14 Differences in mean difference tooth Pb levels for autism spectrum disorder in
discordant twin pairs versus (A) non-autism spectrum disorder twin pairs or (B)
autism spectrum disorder concordant twin pairs.	3-160

Figure 3-15 Hazard rate ratios for Alzheimer's disease mortality by blood Pb level including

the lower 95% confidence interval.	3-219

External Review Draft

3-vii

DRAFT: Do not cite or quote

-------
ACRONYMS AND ABBREVIATIONS

AA	atomic absorption

AAS	atomic absorption spectrometry

Ap	amyloid beta

ABR	auditory brainstem response

AD	Alzheimer's disease

ADD	attention deficit disorder

ADHD	attention deficit/hyperactivity disorder

ADHD-RS	ADHD rating scale

ADOS	Autism Diagnostic Observation
Schedule

ADRA2A	alpha-2A-adrenergic receptor

ALAD	aminolevulinic acid dehydratase

ALS	Amyotrophic Lateral Sclerosis

ALSPAC	Avon Longitudinal Study of Parents

and Children

AOR	adjusted odds ratio

APP	amyloid precursor protein

AQCD	Air Quality Criteria Document

As	arsenic

avg	average

ASD	autism spectrum disorder

ASQ:I	Ages and Stages Questionnaire
Inventory

ASSQ	Autism Spectrum Screening
Questionnaire

ATP	adenosine triphosphate

BAARS	Barkley Adult ADHD-IV Rating Scale

BACE1	beta-secretase 1

BAEP	brain stem auditory evoked potential

BASC	Behavior Assessment System for
Children

BASC-2	Behavior Assessment System for
Children, second revision

BBB	blood-brain barrier

BDI	Beck Depression Inventory

BDNF	brain-derived neurotrophic factor

BKMR	Bayesian Kernel Machine Regression

BKT	BinetKamatTest

BLL	blood lead level

BMD	benchmark dose

BMDL	benchmark dose lower 95% confidence
limit

BMI	body mass index

BMS	Baltimore Memory Study

BNT	Boston Naming Test

BPA	bisphenol A

BPAQ	Buss-Perry Aggression Questionnaire

BrainAGE	Brain Age Gap Estimation

BRIEF	Behavior Rating Inventory of

Executive Functions
BRIEF-A	Behavior Rating Inventory of

Executive Functions for Adults

BRIEF-P	Behavior Rating Inventory of

Executive Functions for Preschool
Children

BRS	behavioral rating scale

BSI	Behavioral Symptoms Index

BSID	Bayley Scales of Infant and Toddler

Development

BSID-IIS	Bayley Scales of Infant and Toddler

Development - Spanish Version
Ca2+	calcium

CANTAB	Cambridge Neuropsychological Test

Automated Battery

CAR	Cortisol awakening response

CARES	Communities Actively Researching

Exposure Study

CARS	Childhood Autism Rating Scale

CAT	catalase

CBCL	Child Behavior Check List

CBLI	cumulative blood lead index

CCAAPS	Cincinnati Childhood Allergy and Air

Pollution Study

CCEI	Crown-Crisp Experiential Index

Cd	cadmium

CDIIT	Comprehensive Developmental

Inventory for Infants and Toddlers

CDK5	cyclin-dependent kinase 5

Ce	cesium

CEM	Coarsened Exact Matching

CERAD	Consortium to Establish a Registry for

Alzheimer's Disease
CHECK	Children's Health and Environmental

Chemicals in Korea
CHEER	Children's Health and Environmental

Research

CHMS	Child Health Monitoring System

CI	confidence interval

CIDI	Composite International Diagnostic

Interview

CKD	chronic kidney disease

CKiD	Chronic Kidney Disease in Children

CLS	Cincinnati Lead Study

CNS	central nervous system

External Review Draft

3-viii

DRAFT: Do not cite or quote

-------
C-P	central-to-peripheral

cpd	cycles per degree

CPR	Conditioned Position Responding

CPRS	Conners Parent Rating Scale

CPRS-R	Conners Parent Rating Scale-Revised

CPT	Continuous Performance Test

C-R	concentration-response

CREB	cyclic adenosine 3',4'-monophosphate
response element binding protein

CRISYS-R	Crisis in Family Systems-Revised

CRP	c-reactive protein

CRS	Conners' Rating Scale

CRS-R	Conners' Rating Scale-Revised

CRT	Combined Raven's Test

CSF	cerebrospinal fluid

C-TRF	Caregiver-Teacher Report Form

CTRS	Conners'Teacher Rating Scale

CTRS-R	Conners'Teacher Rating Scale-
Revised

C-V R2	cross validated R-square

CVA	cerebrovascular accident

CVD	cardiovascular disease

CVLT	California Verbal Learning Test

CVLT-C	California Verbal Learning Test-
Children's Version

d	day(s)

DAT1	dopamine transporter

DBD	Disruptive Behavior Disorder

DDE	dichlorodiphenyldichloroethylene

DI	deionized

DISCI	Disrupted-in-Schizophrenia-1

DMTS	Delayed Matching-to-Sample

DQ	development quotient

DRD2	Dopamine Receptor D2

DNAm	DNA methylation

DSC	Digit Symbol Coding

DSM	Diagnostic and Statistical Manual of

Mental Disorders

DSST	Digit Symbol Substitution Test

DTI	Diffusion Tensor Imaging

ECAT	elemental carbon attributable to traffic

ECDI	Early Child Development Inventory

EE	effect estimate

EEG	electroencephalogram

ELEMENT	Early Life Exposure in Mexico to
Environmental Toxicants

EMOCI	emotional regulation

EOG	end of grade

EPM	elevated plus maze

EPN	early postnatal

EPSC	excitatory postsynaptic currents

ERG	electroretinography

ERP	event-related potential

ETS	environmental tobacco smoke

F	female

F#	filial generation

FA	fractional anisotropy

FBB-ADHS	Fremdbeurteilungsbogen fur

Aufmerksamkeitsdefizit/Hyperaktivitat
storungen

Fe	iron

FFQ	Food Frequency Questionnaire

FI	fixed interval

FLEHS	Flemish Environment and Health Study

FR	fixed ratio

FSIQ	full-scale intelligence quotient

FST	forced swim test

GABA	gamma-aminobutyric acid

GCNT1	glucosaminyl (N-acetyl) transferase 1

GD	gestational day

GDS	Gesell Developmental Schedules

GFAAS	Graphite Furnace Atomic Absorption

Spectrometry

GMR	geometric mean ratio

GRIN	glutamate ionotropic receptor N-methyl
D-aspartate-type subunit

GSH	glutathione

GSI	Global Severity Index

GST	glutathione S-transferase

HCB	hexachlorobenzene

HDL	high-density lipoprotein

HFE	hemochromatosis

Hg	mercury

HHANES	Hispanic Health and Nutrition
Examination Survey

HI	hyperactivity and impulsivity

HNES	Home Nurture Environment Scale

HNRS	Heinz Nixdorf Recall Study

HOME	Health Outcomes and Measures of The

Environment Study

HPA	hypothalamic pituitary adrenal

hr	hour(s)

HR	hazard ratio

HR-ICP-MS	High Resolution Inductively Coupled
Plasma Mass Spectrometry

HRR	hazard rate ratio

HRT	hormone replacement therapy

ICD	International Classification of Diseases

External Review Draft

3-ix

DRAFT: Do not cite or quote

-------
ICP-DRC-MS Dynamic Reaction Cell for Inductively
Coupled Plasma Mass Spectrometry

ICP-MS	inductively coupled plasma mass

spectrometry

ICP-OES	inductively coupled plasma optical

emission spectroscopy
ICP-SFMS	inductively coupled plasma sector field

mass spectrometry
INMA	INfancia y Medio Ambiente

IQ	intelligence quotient

IQR	interquartile ratio

ISA	Integrated Science Assessment

ISAT	Illinois Standard Achievement Test

K6	Kessler Psychological Distress Scale

K-ABC	Kaufman Assessment Battery for

Children

K-ARS	Korean ADHD Rating Scale

K-CBCL	Korean Child Behavior Check List

KEDI	Korean Educational Development

Institute

KiTAP	Test of Attentional Performance for

Children

KNHANES	Korea National Health and Nutrition

Examination Surveys
K-SADS	Kiddie Schedule for Affective

Disorders and Schizophrenia

K-SADS-PL-K Kiddie Schedule for Affective

Disorders and Schizophrenia Present
and Lifetime - Korean Version

KXRF	K-Shell X-Ray Fluorescence

LASSO	least absolute shrinkage and selection

operator

In	natural log

LOD	limit of detection

LTP	long-term potentiation

LURF	Land Use Random Forest

M	male

Mat	maternal

MAT	Metropolitan Achievement Test

MCU	mitochondrial Ca2+ uniporter

MDAT	Malawi Development Assessment Tool

MDI	Mental Development Index

mDISC 1	mouse Disrupted-in-Schizophrenia-1

ME	maternal exposure

MEAP	Michigan Educational Assessment

Program

MeHg	methyl mercury

MHI-5	Mental Health Index 5-item

MIREC	Maternal-Infant Research on

Environmental Chemicals

MMSE	Mini Mental State Examination

Mn	manganese

mo	month(s)

MOCEH	Mothers'and Children's

Environmental Health

MRI	magnetic resonance imaging

MrOS	Osteoporotic Fractures in Men Study

MRS	magnetic resonance spectroscopy

MSCA	McCarthy Scales of Children's

Abilities

NaAc	sodium acetate

NAS	Normative Aging Study

NBAS	Neonatal Behavioral Assessment

Scales

NBNA	Neonatal Behavioral Neurological

Assessment

NCDS	Nunavik Child Development Study

NEI	National Emission Inventory

NHANES	National Health and Nutrition

Examination Survey
NHBCS	New Hampshire Birth Cohort Study

NHS	Nurses' Health Study

NMDAR	N-methyl-D-aspartate receptor

NPR	Norwegian Patient Registry

NR	not reported

NS	no stress

OD/CD	oppositional defiant and conduct
disorder

OFT	open-field test

OLS	ordinary least squares

OR	odds ratio

ORIEN	orientation/engagement

OTB	operant test battery

Pb	lead

PbO	lead oxide

PC	primary caregiver

PCBs	polychlorinated biphenyls

PCNA	proliferating cell nuclear antigen

PD	Parkinson's disease

PDI	Psychomotor Development Index

PECOS	Population, Exposure, Comparison,

Outcome, and Study

PEG	Parkinson's Environment and Genes

PERI	perinatal

PHDCN	Project on Human Development in

Chicago Neighborhoods

PIQ	Performance Intelligence Quotient

PIR	poverty-income ratio

PND	postnatal day

PPI	Psychopathic Personality Inventory

PR	prevalence ratio

External Review Draft

3-x

DRAFT: Do not cite or quote

-------
PROGRESS Programming Research in Obesity,	TUNEL
Growth, Environment and Social

Stressors	USV
PRP post-reinforcement pause

PTA	pure-tone average	VDR

pts	points	YEP

p-tau	phosphorylated tau	yiF

PTSD	post-traumatic stress disorder	yiQ

PW	postweaning	YMt

Q	quartile	WAIS

RNS	reactive nitrogen species	WASI

RO DI	reverse osmosis deionized

ROS	reactive oxygen species	wk

RR	relative risk	WHO

RSEI	Risk Screening Environmental	WIAT

Indicators	WISC
SCN suprachiasmatic nucleus

SCWT	Stroop Color-Word Test	WJTA

SD	standard deviation

SDQ	Strengths and Difficulties	WMC

Questionnaire	WMH

Se	selenium	WMS

SE	standard error	WPPSI

SES	socioeconomic status

SGA	small for gestational age	WRAMI

SGPD	System Genomics of Parkinson's

Disease	WRAT

SMBCS	Sheyang Mini Birth Cohort Study	XRF

SMS	Social Maturity Scale	yr

SOD	superoxide dismutase	YSR

Sp	specificity protein	^n

SPHERL	Study for Promotion of Health in

Recycling Lead
SPM	Standard Progressive Matrix

SQ	social quotient

SRP	self-report of personality

SRS	Social Responsiveness Scale

SWAN	Strengths and Weaknesses of ADHD

Symptoms and Normal Behavior Scale

T	tertile

TBD	to be determined

TBPS	Taiwan Birth Panel Study

TEACh	Test of Everyday Attention for
Children

TMT	Trail Making Test

TOKS	tin-ore kilns and smelters

TRD	Temporal-Response Differentiation

TRF	Teacher Report Form

TSCD	Tohoku Study of Child Development

TST	Tail Suspension Test

terminal deoxynucleotidyl transferase

dUTP nick end labeling

ultrasonic vocalizations

visual acuity

vitamin D receptor

visual evoked potential

variance inflation factor

Verbal Intelligence Quotient

visual-motor integration

Weschler Adult Intelligence Scale

Wechsler Abbreviated Scale of

Intelligence

week(s)

World Health Organization

Wechsler Individual Achievement Test

Wechsler Intelligence Scale for
Children

Woodcock-Johnson Test of
Achievement

working memory capacity
white matter hyperintensities
Weschler Memory Scale
Wechsler Preschool and Primary Scale
of Intelligence

Wide Range Assessment of Memory
and Learning

Wide Range Achievement Test

X-ray fluorescence

year(s)

youth self-report
zinc

External Review Draft

3-xi

DRAFT: Do not cite or quote

-------
APPENDIX 3 NERVOUS SYSTEM EFFECTS

Siiiniiiiirv of C'iiusiililv Dclei iniiiitlioiis lor Pb Kxposuiv iiiul \er\ous System KITecls

This appendix characterizes llie scientific e\ idence lluil supports cnusal11> Jclonninaluins lor
lead (Pit) exposure and ner\ mis s\ slem effects The l\ pes of studies e\alnalod w iill111 illis appondix arc
consislenl with the overall scope of llie ISA as dclailed 111 the Process Appendix (see Seclion 12 4) In
assessing llie overall e\ kIciicc. I lie strengths and limitations of indi\ id ua I sludies were e\ alnalod based
on scienlil"ic considerations delailed in "lahle 12-5 of the Process Appendin. (Seclion 12 <•> I) More
details on I lie causal framework used lo reach lliese conclusions are included in I lie Preamble lo I he ISA
(I S LPA. 2 (> 15) I'lie e\ idence presented throughout ill i s append i\ supporls I he following causahl\
conclusions

Outcome Group	Causality Determination

Nervous System Effects Ascertained during Childhood, Adolescent, and Young Adult Lifestages

Cognitive Function	Causal

Attention, Impulsivity and Hyperactivity	Causal

Conduct Disorders	Likely to be causal

Motor Function	Likely to be causal

Anxiety and Depression	Likely to be causal

„	..	Suggestive of, but not sufficient to infer, a causal

Sensory Function	, .r

1	relationship

Social Cognition and Behavior

Suggestive of, but not sufficient to infer, a causal
relationship

Nervous System Effects Ascertained during Adult Lifestages

Cognitive Effects	Likely to be causal

Psychopathological Effects	Likely to be causal

„ r- ..	Suggestive of, but not sufficient to infer, a causal

Sensory Function	, ,r

1	relationship

.. , ..	Suggestive of, but not sufficient to infer, a causal

Neurodegenerative Disease	, ,r , .

3	relationship

I'lie l\eculi\e Summar\. Inleyraled Synthesis. and all oilier appendices of llns Pit ISA can lie found al
111 ins: cl'nuh cna.jjox ncca isa recordisnla\ .cl'm'deid 3572S2

External Review Draft

3-1

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

3.1 Introduction

While Pb affects nearly every organ system, the nervous system appears to be one of the most
sensitive targets. The sections that follow provide an evaluation of the most policy-relevant scientific
evidence relating to the effects of lead (Pb) exposure on the nervous system. To maximize transparency
regarding the studies included in the appendix, the scope is defined in Section 3.2. Section 3.3, Biological
Plausibility, provides an overview of the biological pathways that potentially underlie the nervous system
effects discussed in subsequent section of Appendix. Section 3.4 summarizes overt nervous system
toxicity, including changes in brain structure and function. There is no causality determination in this
section; rather, data presented in the section may be referenced in the outcome-specific "Integrative
Summary and Causality Determination" discussions in later sections if they provide support for the
conclusions. Sections 3.5 and 3.6 describe the epidemiologic and experimental animal evidence that
pertains to specific endpoints or outcome groupings, which are organized by the lifestage at which they
are ascertained (i.e., childhood, adolescence, and young adult [Section 3.5] and adult [Section 3.6]
lifestages).

The strongest and most policy-relevant evidence within each section is discussed first. Within
Section 3.5, which focuses on exposures and outcomes ascertained during childhood lifestages, including
adolescence and early adulthood, the strongest evidence that is best substantiated at the lowest exposure
levels relates to Cognitive Effects (Section 3.5.1) and Attention, Impulsivity, and Hyperactivity
(Section 3.5.2) in children. Conduct Disorders are discussed in Section 3.5.3, followed by Anxiety and
Depression (Section 3.5.4), Motor Function (Section 3.5.5), Sensory Organ Function (Section 3.5.6), and
Social Cognition and Behavior (Section 3.5.7). The next section (Section 3.6) includes endpoints that are
ascertained during adult lifestages. The section begins with an assessment of the evidence pertaining to
Cognitive Effects in Adults (Section 3.6.1) followed by sections on Anxiety, Depression, and
Psychopathological Effects (Section 3.6.2), Sensory Function (Section 3.6.3), and Neurodegenerative
Diseases (Section 3.6.4). Within each section, the collective body of evidence is integrated within and
across scientific disciplines, and issues relevant for interpreting the scientific evidence as well as the
rationale for the causality determination are outlined for relevant endpoints or outcome groupings.

3.2 Scope

The scope of this appendix is defined by Population, Exposure, Comparison, Outcome, and Study
design (PECOS) statements. The PECOS statements define the objectives of the review and establish
study inclusion criteria, thereby facilitating identification of the most relevant literature to inform the
Lead Integrated Science Assessment (Pb ISA).1 In order to identify the most relevant literature, the body

1 The following types of publications are generally considered to fall outside the scope and are not included in the
ISA: review articles (which typically present summaries or interpretations of existing studies rather than bringing

External Review Draft

3-2

DRAFT: Do not cite or quote

-------
1	of evidence from the 2013 Pb ISA was considered in the development of the PECOS statements for this

2	appendix. Specifically, well-established areas of research; gaps in the literature; and inherent uncertainties

3	in specific populations, exposure metrics, comparison groups, and study designs identified in the 2013 Pb

4	ISA inform the scope of this appendix. The 2013 Pb ISA used different inclusion criteria than the current

5	ISA, and the studies referenced therein often do not meet the current PECOS criteria (e.g., due to higher

6	or unreported biomarker levels). Studies that were included in the 2013 Pb ISA, including many that do

7	not meet the current PECOS criteria, are discussed in this appendix to establish the state of the evidence

8	prior to this assessment. Except for supporting evidence used to demonstrate the biological plausibility of

9	Pb-associated nervous system effects, recent studies evaluated and subsequently discussed within this

10	appendix were only included if they satisfied all components of the following discipline-specific PECOS

11	statements:

12	Epidemiologic Studies:

13	Population: Any human population, including specific populations or lifestages that might be at

14	increased risk of a health effect;

15	Exposure: Exposure to Pb1 as indicated by biological measurements of Pb in the body, with a

16	specific focus on Pb in blood, bone, and teeth; validated environmental indicators of Pb

17	exposure,2 or intervention groups in randomized trials and quasi-experimental studies;

18	Comparison: Populations, population subgroups, or individuals with relatively higher versus

19	lower levels of the exposure metric (e.g., per unit or log unit increase in the exposure metric,

20	or categorical comparisons between different exposure metric quantiles);

21	Outcome: Nervous system effects including but not limited to cognitive function (e.g.,

22	intelligence quotient [IQ] decrement), externalizing and internalizing behaviors,

23	psychopathological effects, sensory organ function, motor function, and neurodegenerative

24	diseases; and

25	Study Design: Epidemiologic studies consisting of longitudinal and retrospective cohort studies,

26	case-control studies, cross-sectional studies with appropriate timing of exposure for the health

27	endpoint of interest, randomized trials and quasi-experimental studies examining

28	interventions to reduce exposures.

29	Experimental Studies:

30	Population: Laboratory nonhuman mammalian animal species (e.g., mouse, rat, guinea pig,

31	minipig, rabbit, cat, dog) of any lifestage (including preconception, in utero, lactation,

32	peripubertal, and adult stages);

forward new information in the form of original research or new analyses), Pb poisoning studies or clinical reports
(e.g., involving accidental exposures to very high amounts of Pb described in clinical reports that may be extremely
unlikely to be experienced under ambient air exposure conditions), and risk or benefits analyses (e.g., that apply
concentration-response functions or effect estimates to exposure estimates for differing cases).

1	Recent studies of occupational exposure to Pb were considered insofar as they addressed a topic area that was of
particular relevance to the National Ambient Air Quality Standards review (e.g., longitudinal studies designed to
examine recent versus historical Pb exposure).

2	Studies that estimate Pb exposure by measuring Pb concentrations in particulate matter with a nominal mean
aerodynamic diameter less than or equal to 10 |im3 (PMio) and particulate matter with a nominal mean aerodynamic
diameter less than or equal to 2.5 |im3 (PM25) ambient air samples are only considered for inclusion if they also
include a relevant biomarker of exposure. Given that size distribution data for Pb-PM are fairly limited, it is difficult
to assess the representativeness of these concentrations to population exposure (Section 2.5.3 (U.S. EPA. 2013aTI.
Moreover, data illustrating the relationships of Pb-PMio and Pb-PM2 5 with blood Pb levels (BLLs) are lacking.

External Review Draft

3-3

DRAFT: Do not cite or quote

-------
1 Exposure: Oral, inhalation, or intravenous routes administered to a whole animal (in vivo) that

2 results in a BLL of 30 (ig/dL or below;1'2

3 Comparators: A concurrent control group exposed to vehicle-only treatment or untreated

4 control;

5 Outcome: Nervous System effects; and

6 Study Design: Controlled exposure studies of animals in vivo.

3.3 Biological Plausibility

7 This section describes biological pathways that potentially underlie nervous system effects

8 resulting from exposure to Pb. Timing of exposure is important for the health effects for Pb. Exposures

9 during development can lead to improper formation and maturation of the nervous system and exposures

10 to the mature nervous system can lead to neurodegeneration. Figure 3-1 and Figure 3-2 graphically depict

11 these proposed pathways for health effects resulting from developmental exposure to Pb and later life

12 exposures, respectively. Proposed pathways are presented as a continuum of responses, connected by

13 arrows, which may ultimately lead to the apical nervous system health effects associated with exposures

14 to Pb at concentrations observed in epidemiologic studies. This discussion of "how" exposure to Pb may

15 lead to effects on the nervous system contributes to an understanding of the biological plausibility of

16 epidemiologic results evaluated throughout this appendix. Most of the studies cited in this subsection are

17 discussed in greater detail elsewhere in this appendix. The biological plausibility for Pb-induced effects

18 on the nervous system is supported by evidence from the 2013 Pb ISA and by recent evidence. Note that

19 the structure of the biological plausibility sections and the role of biological plausibility in contributing to

20 the weight-of-evidence analysis used in the current ISA are discussed in Section IS.7.2.

1 Pb mixture studies are included if they employ an experimental arm that involves exposure to Pb alone.

2 This level represents an order of magnitude above the upper end of the distribution of U.S. young children's BLL.
The 95th percentile of the 2011-2016 National Health and Nutrition Examination Survey distribution of BLL in
children (1-5 years; n = 2,321) is 2.66 (ig/dL (Eganetal.. 20211 and the proportion of individuals with BLL that
exceed this concentration varies depending on factors including (but not limited to) housing age, geographic region,
and a child's age, sex, and nutritional status.

External Review Draft

3-4

DRAFT: Do not cite or quote

-------
Note: The boxes above represent the effects for which there is experimental or epidemiologic evidence related to Pb exposure, and the arrows indicate a proposed relationship
between those effects. Solid arrows denote evidence of essentiality as provided, for example, by an inhibitor of the pathway used in an experimental study involving Pb exposure.
Dotted arrows denote a possible relationship between effects. Shading around multiple boxes is used to denote a grouping of these effects. Arrows may connect individual boxes,
groupings of boxes, and individual boxes within groupings of boxes. Progression of effects is generally depicted from left to right and color coded (white, exposure; green, initial effect;
blue, intermediate effect; orange, effect at the population level or a key clinical effect). Here, population-level effects generally reflect results of epidemiologic studies. When there are
gaps in the evidence, there are complementary gaps in the figure and the accompanying text below. The structure of the biological plausibility sections and the role of biological
plausibility in contributing to the weight-of-evidence analysis used in the 2022 Pb ISA are discussed in Section IS.7.2. Source: (Shadbeqian et al., 2019).

Figure 3-1 Potential biological pathways for nervous system effects following developmental exposure
to Pb.

External Review Draft

3-5

DRAFT: Do not cite or quote

-------
Altered Neurotransmitter
Signaling

Altered Calcium Signaling

Impaired Blood
Brain & Blood
Cerebrospinal
Fluid Barrier
Permeability

Note: The boxes above represent the effects for which there is experimental or epidemiologic evidence related to Pb exposure, and the arrows indicate a proposed relationship
between those effects. Solid arrows denote evidence of essentiality as provided, for example, by an inhibitor of the pathway used in an experimental study involving Pb exposure.
Dotted arrows denote a possible relationship between effects. Shading around multiple boxes is used to denote a grouping of these effects. Arrows may connect individual boxes,
groupings of boxes, and individual boxes within groupings of boxes. Progression of effects is generally depicted from left to right and color coded (white, exposure; green, initial effect;
blue, intermediate effect; orange, effect at the population level or a key clinical effect). Here, population-level effects generally reflect results of epidemiologic studies. When there are
gaps in the evidence, there are complementary gaps in the figure and the accompanying text below. The structure of the biological plausibility sections and the role of biological
plausibility in contributing to the weight-of-evidence analysis used in the 2022 Pb ISA are discussed in Section I.S.7.2. Source: (Shadbeaian et al.. 2019).

Figure 3-2 Potential biological pathways for nervous system effects following postweaning exposure to Pb.

External Review Draft

3-6

DRAFT: Do not cite or quote

-------
1

Plausible pathways connecting Pb exposure to apical events resulting from developmental and
later life exposures to Pb are proposed in Figure 3-1 and Figure 3-2, respectively. The proposed pathways
supported by the strongest evidence include the direct actions of Pb on cellular protein function and
subsequent initiation of oxidative stress-mediated pathways.

When Pb accumulates in the CNS, it can interfere with coordination of metal ions, which is
essential for the structure and function of many cellular proteins. Pb ions can compete with and displace
physiologically relevant ions (including Fe, Zn, Ca, and others) within proteins, leading to both altered
protein structure and function. As described in the 2013 Pb ISA, there is evidence that this ionic mimicry
and imbalance occurs in multiple organ systems, including the brain, and in proteins that perform diverse
functions including metabolism, inflammation, and oxidative stress responses. For example, Pb treatment
can disrupt Ca2+ signaling through interactions with calmodulin, voltage-gated Ca2+ channels, and various
adenosine triphosphate (ATP)ases (U.S. EPA. 2013a). There is also evidence that Pb can replace Zn ions
in Zn finger-binding motifs, which are present in several transcription regulating proteins (U.S. EPA.
2013a). Some research supports an interactive effect between Fe status and Pb exposure due to shared
metabolic and physiological profiles. Lifetime exposure to Pb in rats has been shown to affect Fe status
by increasing Fe content in the cortex and hippocampus of adult and aged animals and altering the
expression of divalent metal transporters (such as divalent metal transporter 1 and ferroportin) in the brain
(Zhu et al.. 2013). suggesting that Pb may interfere with Fe trafficking in the brain. Often the effect of Pb
can be reduced with exogenous supplementation of biologically relevant metals. Recent studies support
the protective role of supplementation of Ca2+ (Basha and Reddv. 2015; Gottipolu and Davuljigari. 2014).
Zn (Pcdroso et al.. 2017). Fe (Liu et al.. 2013c) or essential metal mixtures (Basha et al.. 2014) on
neurologic alterations from Pb. These data support the hypothesis that direct competition of Pb with
metals can cause neurologic effects.

The brain has the highest energy demand and metabolism of any organ. Because of this fact,
energy homeostasis is critical and energy imbalance can increase the brain's susceptibility to stressors and
cell death. Pb-induced alterations in energy production and metabolism have been measured in several
ways. As discussed in the 2013 Pb ISA, Pb exposure can alter many aspects of energy metabolism, with
animal models demonstrating effects following both developmental and adult exposures to Pb (discussed
in Section 3.4.2.1). In recent studies of developmental Pb exposure, Pb-induced impairments in energy
production throughout the body have been measured as reductions in the activity of glucose and glycogen
metabolizing enzymes (Baranowska-Bosiacka et al.. 2017) and alterations in the number and structure of
mitochondria (Ouvang et al.. 2019; Gassowska et al.. 2016a). Studies of Pb exposure in postweaning
animals showed similar reductions of metabolizing enzyme activity (Yun et al.. 2019; Verma et al.. 2005;
Yun and Hover. 2000; Sterling etal.. 1982). altered mitochondrial structure (Ouvang et al.. 2019;
Dabrowska et al.. 2015; Sun et al.. 2014). and ATPase activity (Thangaraian et al.. 2018). suggesting
alteration of energy metabolism may occur regardless of the timing of Pb exposure.

External Review Draft

3-7

DRAFT: Do not cite or quote

-------
1

Energy production involves the formation of reactive intermediate species including reactive
oxygen (ROS) and nitrogen species (RNS). Disruptions in the mitochondria and energy metabolism result
in increased levels of ROS and RNS. While ROS are a part of normal cellular functioning, uncontrolled
production or reduced elimination of ROS by antioxidant systems can result in oxidative stress and
cellular damage (for example, DNA damage, oxidization of cellular components). Evidence reviewed in
the 2013 Pb ISA suggests that Pb may exert toxicity by disrupting cellular metabolism, increasing
ROS/RNS concentrations, and depleting antioxidant capacity (U.S. EPA. 2013a'). Numerous recent
studies have reported dysregulation of oxidative stress concurrent with altered mitochondrial function
(Ahmad et al.. 2020; Karri et al.. 2018; Maiti et al.. 2017; Kumar and Muralidhara. 2014; Baranowska-
Bosiacka et al.. 2011). adding to the body of evidence. A study by Yang et al. (2014) found that Pb
downregulated the mitochondrial Ca2+ uniporter (MCU), resulting in increased ROS production in both
SH-SY5Y cells and in newborn rats. Yang and colleagues found that in vitro activation or overexpression
of MCU prevented Pb-induced oxidative stress whereas MCU inhibition or knockdown potentiated the
effects suggesting that alterations in mitochondrial function were responsible for Pb-induced ROS
production. In a similar manner, Pb has been shown to upregulate cyclophilin D, a protein that regulates
mitochondrial membrane potential, and in vitro knockdown or inhibition of cyclophilin D prevents the
Pb-induced loss of mitochondrial membrane potential (Ye et al.. 2020; Ye et al.. 2016a). Mitochondrial
function is also thought to be dependent on a dynamic balance between mitochondrial fission and fusion.
In a recent study, Pb reduced energy production and respiration while increasing mitochondrial ROS and
altering the expression of genes involved with mitochondrial dynamics both in vitro and in vivo
(Dabrowska et al.. 2015). In this study, knockdown of the transcription factor peroxisome proliferator-
activated receptor-y coactivator la, which protects the mitochondrial fusion and fission balance, increased
in vitro ROS production in response to Pb, further suggesting that altered mitochondrial activity results in
ROS production in response to Pb. Together, these data provide evidence that mitochondrial dysfunction
and altered energy metabolism is a source of oxidative stress.

Given their reactive nature, ROS and RNS can damage cellular proteins, lipids, and nucleic acids,
which can lead to functional and downstream signaling impairment. As discussed in the 2013 Pb ISA, Pb
exposure leads to elevated levels of ROS in neurons and other brain cells of exposed animals. Levels of
oxidative species have also been assessed indirectly by the presence of oxidative damage to DNA and
proteins as well as peroxidation of lipids. Studies assessed in the 2013 Pb ISA showed that Pb exposure
increased signs of oxidative damage in the brains of a variety of animal species (U.S. EPA. 2013a; Wu et
al.. 2008). Since publication of the 2013 Pb ISA, more recent studies have demonstrated Pb-induced
increases in ROS production and oxidative damage in the brain both during development (Hossain et al..
2016; Lu et al.. 2013) and postweaning (Singh et al.. 2019; Liu et al.. 2018a; Thangaraian et al.. 2018;
Singh et al.. 2017; Kumar and Muralidhara. 2014; Flora et al.. 2012). Proper regulation of oxidative stress
requires a balance between the presence of oxidative species (i.e., ROS and RNS) and levels of
antioxidant defense proteins (e.g., glutathione [GSH], catalase [CAT], and SOD). Along with increased
ROS/RNS production, depletion of antioxidant proteins or reductions in antioxidant enzyme activity
could contribute to an overall increase in oxidative stress. As discussed in the 2013 Pb ISA, animal

External Review Draft

3-8

DRAFT: Do not cite or quote

-------
1

studies and human panel studies have shown that BLL is associated with an increased ratio of oxidized to
unoxidized GSH (Mohammad et al.. 2008; Diouf et al.. 2006; Ercal et al.. 1996; Sandhir and Gill. 1995).
More recent studies showed similar impairment of antioxidant defenses in animal models of Pb exposure
during developmental (Lu et al.. 2013) and postweaning (Singh et al.. 2019; Thangaraian et al.. 2018;
Singh et al.. 2017; Flora et al.. 2012) Pb exposures. Furthermore, changes in antioxidant status and
oxidative stress can contribute to mitochondrial dysfunction, as described above. There is strong evidence
that Pb exposure across the lifespan disrupts multiple aspects of energy metabolism and oxidative stress
regulation.

Cellular damage caused by oxidative insults can trigger inflammation and vice versa; thus, it is
often difficult to disentangle which occurs first. Given the interrelated nature of these factors, they are
combined within the same gray box in the blood Pb diagrams (Figure 3-1 and Figure 3-2). Inflammation
is a hallmark of many neurological conditions and neurodegenerative diseases. Inflammation can be
triggered by the production of inflammatory mediators (e.g., cytokines) in response to cell or protein
damage. As discussed in the 2013 Pb ISA, Pb exposure results in signs of inflammation including
activation of inflammatory signaling pathways, inflammatory mediator production, and microglia cell
activation (U.S. EPA. 2013a). Several studies have observed increased inflammatory mediator levels and
activation of inflammatory signaling pathways (for example, tumor necrosis factor-alpha) in the brains of
animals exposed to Pb during development (Chibowska et al.. 2020; Hossain etal.. 2016; Ashok et al..
2015) and postweaning (Yang et al.. 2019; Liu et al.. 2018a). Proinflammatory markers interact with, and
in some cases infiltrate, the BBB, initiating neuroinflammation, as indicated by altered gene expression,
increased apoptosis, lipid and protein oxidation, and microglial activation(Saleh et al.. 2018; Shvachiv et
al.. 2018; Sobin et al.. 2013). Similarly, histologic and immunohistochemical signs of neuroinflammation
in the dentate gyrus have been reported in rats exposed to Pb continuously from 7 days postconception to
28 weeks of age, which corresponded to behavioral changes (Shvachiv et al.. 2018). In the same study, a
similar neuroinflammatory phenotype was observed in mice that were given an 8-week Pb abstinence
period between 12-week and 8-week Pb exposures (Shvachiv et al.. 2018). In sum, recent evidence
supports the plausibility of inflammation as an intermediate event in the development of neurological
health effects regardless of the timing of Pb exposure.

While a robust immune response can protect the brain from certain insults, prolonged
neuroinflammation is associated with several neurological and neurodegenerative diseases. AD,
characterized by the accumulation of A(3 and p-tau, has been associated with increased markers of
neuroinflammation. While neurodegenerative diseases are associated with old age, studies of
developmental exposures to Pb have shown that early life exposures are associated with Alzheimer's-like
pathology in adult animals. As discussed in the 2013 Pb ISA, Pb exposure in juvenile animals resulted in
the increased production of APP and higher levels of p-tau in offspring. Similarly, early life Pb exposure
of nonhuman primates led to Alzheimer's-like pathology later in adulthood (Wu et al.. 2008). Studies
published since the last ISA support and extend the findings that developmental exposures to Pb can lead
to increased levels of misfolded proteins (e.g., abnormal APP processing, A(3, tau protein) and

External Review Draft

3-9

DRAFT: Do not cite or quote

-------
1

Alzheimer's-like pathologies (e.g., p-tau accumulation) (Ashok et al.. 2015; Bihaqi and Zawia. 2013).
Evidence from exposures during development suggests that early life may represent a sensitive window
for insults associated with neurodegenerative disease (Liu et al.. 2014a). Some studies with exposure of
postweaning animals to Pb have shown increased inflammation associated with AD markers (Yang et al..
2019; Liu et al.. 2018a; Zhang et al.. 2012). In postweaning studies, treatment with molecules with anti-
inflammatory and antioxidative properties were able to prevent A(3 accumulation and reversed cognitive
and behavioral alterations in Pb-exposed mice (Liu et al.. 2020; Yang et al.. 2019; Liu et al.. 2018a').
Because of this, there is a solid line connecting the box containing inflammation, oxidative stress, and
altered energy metabolism to the accumulation of A(3 in Figure 3-2. Evidence of effects of Pb on other
neurodegenerative diseases are more limited. A recent study showed that exposure of postweaning rats to
Pb resulted in increased accumulation of a-synuclein, a protein associated with PD, in the hippocampus
that correlated with impaired learning and memory (Zhang et al.. 2012). Overall, new data support the
previous findings that Pb exposure can affect the development and progression of neurodegenerative
pathologies in developmentally and postweaning exposed animals.

Beyond Pb's ability to produce neuroinflammation and oxidative stress and increase expression
of disease-related proteins, excessive damage to cellular proteins or DNA can trigger cell death. While
cell death and neuronal population loss in adulthood contribute to brain pathology, the developing brain is
far more sensitive to disruption. Cell migration, differentiation, and pruning are all essential
neurodevelopmental processes that need to be carefully timed and orchestrated. Thus, increased or
aberrant cell loss results in improper nervous system development that could be responsible for the altered
mood, sensory, or cognitive functions observed in Pb-exposed children and animals. The 2013 Pb ISA
and Section 3.4.2.1 present several animal studies showing upregulation of apoptotic markers in various
regions of the brain following Pb treatment, at various lifestages, which was supported by similar findings
in in vitro experiments (U.S. EPA. 2013a). Recent studies also reported activation of pro-apoptotic
pathways in response to developmental Pb exposure (Ebrahimzadeh-Bideskan et al.. 2016; Hossain et al..
2016; Su et al.. 2016; Lu et al.. 2013). supporting and extending the experiments reviewed in the 2013 Pb
ISA. Another study showed histologic changes in the brain concomitant with increased markers of protein
and lipid damage (Saleh etal.. 2019). suggesting a relationship between cell death and structural changes
in the brain with oxidative damage. Developmental Pb exposure caused dysregulated myelination in the
brains of rats, which could be rescued with cotreatment with antioxidants (Nam et al.. 2020; Nam et al..
2019a). Myelination is an essential step in nervous system development, as myelin sheaths facilitate quick
and efficient electrical transmission along nerve cells to preserve nervous system function and
connectivity. Several studies have also demonstrated that treatment with compounds with antioxidant
capacity reduced the apoptotic signaling (Nam et al.. 2018b; Ebrahimzadeh-Bideskan et al.. 2016).
Together, these data provide the justification for a solid line from the gray box containing oxidative stress
and inflammation to the box containing cell injury/death in Figure 3-2.

Widespread cell loss in the mature nervous system can also lead to functional and structural
changes that can contribute to behavioral and cognitive changes. Cell death is also a common element in

External Review Draft

3-10

DRAFT: Do not cite or quote

-------
1

many neurodegenerative diseases. The animal studies showing upregulation of apoptotic markers in
various regions of the brain following Pb treatment discussed in the 2013 Pb ISA are strengthened by
similar findings in several new studies (Amcdu and Omotoso. 2020; Liu et al.. 2020; Abubakar et al..
2019; Singh et al.. 2019; Yang et al.. 2019; Liu et al.. 2018a; Thangaraian et al.. 2018; Maiti et al.. 2017;
Singh et al.. 2016; Flora et al.. 2012). In vitro exposure of neuronal cell lines to Pb resulted in reduced
cell viability and increased apoptosis (Ye et al.. 2020; Liu et al.. 2017; Neelima et al.. 2017; Meng et al..
2016; Su et al.. 2016; Ye et al.. 2016b; Ahmed et al.. 2013). Additional discussion of apoptotic markers
and brain structural changes following Pb exposure are discussed in Section 3.4.2.1. Like the
developmental exposure studies, demyelination was observed in the spinal cord following postweaning
exposure to Pb (da Silva et al.. 2020; Villa-Cedillo et al.. 2019). These data provide plausibility that adult
Pb exposures contribute to cognitive and behavioral changes. Some studies therapeutically targeted RNS
production in the mitochondria by treatment with fisetin, a polyphenolic compound with antioxidant
properties to ameliorate the activation of pro-apoptotic signaling (Yang et al.. 2019; Maiti et al.. 2017).
Their results suggest a role for oxidative stress in triggering the apoptotic cascade. In vitro treatment with
the antioxidant genistein also protected against cell death (Su et al.. 2016). Thangaraian et al. (2018)
showed that treatment with an anti-inflammatory and antioxidative compound, morin, was able to largely
restore proper brain architecture after Pb exposure. Together, these data provide the justification for a
solid line from the gray box containing oxidative stress and inflammation to the box containing cell
injury/death in Figure 3-2.

Inflammation and oxidative stress can also affect the integrity of the BBB, which provides a
selective barrier for entry from the circulation to the brain and spinal cord. As discussed in the 2013 Pb
ISA and 2006 Pb AQCD, Pb exposure in rodents was shown to increase permeability of the BBB and the
blood-CSF barrier. Interestingly, Pb alone does not have a large effect on BBB integrity but can prolong
BBB permeability in response to other stimuli (U.S. EPA. 2013a. 2006a). The effect of Pb on the BBB is
also selective in that the permeability of all solutes is not affected equally (U.S. EPA. 2013a. 2006a).
Disruption of the BBB could potentially promote increased Pb accumulation in the brain with prolonged
or repeated exposure. Two new studies assessed the integrity of the BBB following Pb exposure and
observed disruption of brain permeability with reduced levels of tight junction proteins and other
important capillary proteins (Wu et al.. 2020a; Song et al.. 2014). In adult rats, 8 weeks of Pb dosing
reduced expression of the tight junction proteins occludin and zonula occludens-1 at the BBB (Song et
al.. 2014). Pb has been implicated in alteration of the CSF barrier in rats (Zheng et al.. 1996). The CSF
can carry hormone signals important for brain development; thus, disruption of the cerebrospinal barrier
could affect proper hormone signaling for brain development. Indeed, Pb exposure was reported to
decrease transthyretin levels in the CSF, suggesting altered cerebrospinal barrier integrity (Zheng et al..
1996). There is likely interplay between Pb effects on endocrine and nervous system development. In
conclusion, there is a potential for Pb to affect the BBB and blood spinal cord barrier, which could alter
Pb availability and uptake into the nervous system.

External Review Draft

3-11

DRAFT: Do not cite or quote

-------
1

While most data suggest that Pb acts through a mode of action involving oxidative stress and
inflammation, additional signaling pathways are also affected by Pb exposure. Pb exposure can alter ion
balance, which has particular importance with regard to the effect on Ca2+ signaling. Calcium signaling is
vital for many fundamental neurological processes including membrane excitability, neurotransmitter
release, synaptogenesis, transmission, and other processes. Of particular relevance for this review,
neurotransmitter signaling is intimately connected with Ca2+ signaling. As discussed in the 2013 Pb ISA,
developmental exposure to Pb interferes with the evoked release of neurotransmitters by inhibiting Ca2+
transport through voltage-gated ion channels (Cooper and Manalis. 1984; Suszkiw et al.. 1984). Ca2+ is
also a ubiquitous second messenger, which can regulate many neuronal physiologic processes like gene
expression, membrane excitability, and dendrite development (Kawamoto et al.. 2012). Interestingly, in
the absence of stimulation, Pb has some Ca2+ mimetic activity that increases baseline neurotransmitter
release (Cooper and Manalis. 1984; Suszkiw et al.. 1984). In general, Pb exposure increased Ach levels,
increased dopaminergic signaling, and reduced NMDAR expression (U.S. EPA. 2013a). Animal models
suggest that Pb exposure during development leads to inhibition of acetylcholinesterase (AchE), thereby
increasing the levels of Ach and causing lasting neurodevelopmental changes that persist into adulthood
(Basha and Reddv. 2015). The authors found that addition of Ca2+ restored cholinergic signaling (Basha
and Reddv. 2015). These data help to justify the solid line from ionic mimicry to altered neurochemical
signaling in Figure 3-1. Similar studies in animals postweaning have shown similar decreases in AchE
activity (Galal et al.. 2019; Okesola et al.. 2019; Thangaraian et al.. 2018; Andrade et al.. 2017; Ferlemi et
al.. 2014; Phvu and Tangpong. 2013). Recent literature also supports altered dopaminergic signaling
following postweaning Pb exposure (Sobolewski et al.. 2020; Yousef et al.. 2019; Amos-Kroohs et al..
2016; Stansfield et al.. 2015; Basha et al.. 2014; Weston et al.. 2014; Corv-Slechta et al.. 2012; Graham et
al.. 2011). Ca2+ gradients are also responsible for generating action potentials. Alteration of intracellular
Ca2+ levels in neurons could cause deleterious effects on action potential generation and repolarization.
Recent evidence shows that hippocampal slices from 50-day old rats exposed to Pb both pre and
postnatally had enhanced pared pulse facilitation, suggesting Pb-induced dysregulation of Ca2+ signaling
(Zhang et al.. 2015b). This data support findings from a limited number of human MRI and MRS studies
that provide evidence of physical and physiological changes in the brain corresponding to increased blood
Pb that were discussed in the 2013 Pb ISA and 2006 Pb AQCD. Together there is evidence that Pb can
alter neurotransmitter release and signal potentiation, which in turn could contribute to the changes in
brain activity seen in various behavioral and cognitive diseases.

Beyond the actions of Pb discussed thus far, there is growing evidence of the effect of changes to
the epigenome in mediating toxicity. Epigenetic changes refer to alterations in the mechanisms that
regulate gene expression without altering DNA sequence. Epigenetic programming is a fundamental
developmental process, and the complex relationships between the genome, epigenome, and environment
can shape the health of present and future generations. Epigenetic alterations are often measured as
changes in histone and DNA methylation patterns as well as the levels of the enzymes (e.g.,
methyltransferases, acetylases, deacetylases) responsible for regulating epigenetic modification in situ.
Transcriptional regulators like microRNAs and long noncoding RNAs are also considered epigenetic

External Review Draft

3-12

DRAFT: Do not cite or quote

-------
1

modifiers. As discussed in the 2013 Pb ISA, developmental Pb treatment in mice and monkeys decreased
the activity of some DNA methyltransferases. Therapeutic treatment with a methyl donor improved Pb-
induced decrements in LTP and Morris water maze performance (Cao et al.. 2008). Gestational and
postnatal exposure to Pb in rats increased histone acetylation in the hippocampus, which corresponded to
a hyperactivity phenotype (Luo et al.. 2014). The authors suggested this was due to upregulation of
histone acetyltransferases, including p300. However, these changes occurred at BLLs in excess of
50 (ig/dL; thus, the relevance of these findings to ambient exposure in humans is questionable. Other
studies have reported changes in the expression of DNA methyltransferases (Schneider et al.. 2013) and
increased hypermethylation, especially in the hippocampus of female mice (Sanchez-Martin et al.. 2015).
The window of exposure, prenatal stress, and sex can all play a role in determining the epigenetic
modifications (Sobolewski et al.. 2018). While differences in epigenetic modifications and the effects of
Pb on epigenetic enzymes have been reported and linked to behavioral effects in animals, there remains
little evidence to connect epigenetic changes to alterations in specific pathways that have the potential to
cause neurobehavior effects. As a result, the arrow for epigenetic changes in is represented as a dotted
line when connecting to the box for neurodevelopmental disorders. Future research may elucidate a role
for epigenetic modification in the etiology of neurological diseases. Most of the Pb literature on
epigenetic changes has focused on heritable epigenetic changes during development; however, the effect
of postweaning exposure to Pb on epigenetic mechanisms is not well known. Very few studies have
evaluated epigenetic changes throughout the lifetime or with later life exposures. Individual studies have
found changes in the expression of a long noncoding RNA (Nan et al.. 2016) and a methyltransferase
(Schneider et al.. 2012). which might suggest that epigenetic modification could be affected during adult
exposures; however, there are too few studies to draw reliable conclusions. The conclusion of epigenetic
modification resulting in neurological effects is only plausible for developmental exposures with the
present available data.

In summary, Pb exposure can result in a range of neurocognitive and behavioral health effects
through a myriad of complex biological pathways. The pathways described here provide biological
plausibility for associations between Pb exposure and nervous system effects in in children and adults.
The developmental timing, sex, and presence of other stressors or enrichments alongside of Pb exposure
can affect the resulting health effects. The identified pathways share many common features, including Pb
interactions with cellular proteins, competing with and displacing other biologically relevant cations,
increased oxidative stress, and inflammation, which can have widespread effects on brain structure and
function. There is also evidence for disruptions of Ca2+ signaling, which can result in altered
neurotransmitter signaling and contribute to the development of neurological health effects. Epigenetic
modifications resulting from Pb developmental exposure have been reported but are still an area of active
investigation. The role of these epigenetic changes in the progression of neurological health effects with
later Pb exposures is unclear. Together the proposed pathways provide biological plausibility for
epidemiologic evidence of neurological effects and were used to inform causality determinations
throughout this appendix.

External Review Draft

3-13

DRAFT: Do not cite or quote

-------
1

3.4

Overt Nervous System Toxicity

The collective body of epidemiologic and experimental animal studies assessed in the 2013 Pb
ISA demonstrated the effects of Pb exposure on an array of nervous system outcomes. Overt nervous
system toxicity refers to a diverse group of endpoints that inform brain structure and function, including
brain histopathology changes, brain weight, electrophysiology, neuroinflammation, and neurotransmitter
analyses. The evidence, including uncertainties, is summarized in Section 3.4.3. Study details that
supplement the information provided in the text are in the evidence inventories (Table 3-1E and Table 3-
1T in Section 3.7). Previous Pb assessments reviewed epidemiologic studies that found associations of Pb
biomarkers with electrophysiologic or physical changes in the brains of adults assessed by imaging
technologies. Biological plausibility for the effects of Pb on overt nervous system toxicity was provided
by a small number of experimental animal study findings with dietary and lactational Pb exposure, with
some evidence at BLLs relevant to humans. Recent epidemiologic studies support and extend the
evidence pertaining to the association of lead exposure during childhood with brain structure and function
in adolescence or adulthood. A smaller set of cross-sectional studies also report associations between
childhood BLLs and overt nervous system outcomes. Multiple experimental animal studies report
changes in the brains of rats and mice following exposure to Pb. These include changes in histology,
neurotransmitter measures, brain weight, and electrophysiology measures, providing coherence for the
epidemiologic studies that show associations with decrements in cognition, neurodegeneration, or
increased behavioral problems. There is no causality determination for this section; rather, evidence in
this section may be referenced in the outcome-specific "Integrative Summary and Causality
Determination" discussions of Sections 3.5 and 3.6 if they provide biological plausibility or coherence for
the observations in the epidemiologic studies.

3.4.1 Epidemiologic Studies of Brain Structure and Function

Previous Pb assessments (U.S. EPA. 2013a. 2006b') reviewed a small body of epidemiologic
studies that found associations of Pb biomarkers with electrophysiologic and physical changes in the
brains of young adults as assessed by magnetic resonance imaging (MRI) or spectroscopy (MRS). The
implications of findings from most studies assessed in the 2006 Pb Air Quality Criteria Document
(AQCD) were limited by the small sample sizes (n = 12 to 45) and inadequate consideration of potential
confounding. However, analyses cohort of adults (ages 20-23 years) reviewed in the 2013 Pb ISA
included larger sample sizes and aimed to characterize potentially important lifestages of Pb exposures
(Yuan et al.. 2006). thus expanding the evidence pertaining to potential links between physiologic brain
changes and functional neurodevelopmental effects. Overall, the small number of studies in a limited
number of populations assessed in the 2013 Pb ISA showed physical and physiologic changes in areas of
the brain associated with neurodevelopmental function, providing biological plausibility for the

External Review Draft

3-14

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

associations observed between Pb biomarker levels and cognitive function decrements and behavioral
problems.

Several recent longitudinal studies add to the evidence characterizing the association of Pb
exposure during childhood with brain structure and function during adolescence or adulthood. A smaller
number of studies evaluated the cross-sectional association of childhood BLL with brain structure or
function in childhood. These studies are summarized below, and key information from the studies is
included in Section 3.7, Table 3-1E.

Reuben et al. (2020) conducted a study to examine the effect of childhood BLL, measured at age
11, on lower structural integrity of the brain at age 45. These investigators used data from the Dunedin
Study in New Zealand, which enrolled participants beginning in 1972 and 1973 and followed them
through April 2019. The mean early childhood BLL for the study participants was 10.99 (ig/dL. MRI was
used to assess multiple endpoints related to gray matter (cortical thickness, surface area, and hippocampal
volume), white matter (white matter hyperintensities, fractional anisotropy [FA]), and the gap between
chronological age and estimated brain age. In addition, cognitive function was estimated using the
Wechsler Adult Intelligence Scale (WAIS)-IV, self-reports, and informant reports (see Section 3.6.1). A
total of 564 of the original 1037 infants enrolled at birth were included in the analysis. Findings from the
study are depicted in Figure 3-3. In models adjusted for sex, maternal IQ, and socioeconomic status
(SES), associations were observed with cortical surface area, hippocampal volume, global FA, and the
gap between each study member's chronological age at imaging and their MRI-predicted age, but not with
all the MRI metrics assessed.

Two analyses of the Cincinnati Lead Study (CLS) have been conducted since the 2013 Pb ISA.
Confounders considered in these analyses included child characteristics, Home Observation for the
Measurement of Environment (HOME) score, maternal IQ, and SES (see Section 3.7, Table 3-1E for
study-specific confounders). Cecil (2011) examined the association of childhood BLL (childhood [3-
28 months] average) with volumetric MRI, MRS, diffusion tensor imaging (DTI), and functional MRI
outcomes ascertained between ages 19 and 24 years old. This study found that childhood BLL was
associated with decreased gray matter volume in several regions (i.e., medial and superior frontal gyri,
inferior parietal lobule and cerebellar hemispheres). Higher childhood BLL was also associated with
lower metabolite concentrations in several brain regions (white matter, left basal ganglia, left cerebellar
hemisphere, and vermis). DTI and functional MRI findings also suggested injury and compensatory
activity in specific brain regions. Overall, structural, organizational, and functional changes in the brain
regions responsible for regulating behavior were indicated by this study. In another study of participants
enrolled in the CLS, Beckwith et al. (2021) examined the relationships between childhood BLL (at
78 months), structural brain volume, and adult criminality. BLLs were associated with MRI-derived
decreases in white and gray matter volumes in the frontal parietal and temporal lobes. Decreased gray
matter volume in brain regions responsible for cognition and emotional regulation was also associated

External Review Draft

3-15

DRAFT: Do not cite or quote

-------
with criminal arrests, potentially supporting associations observed between Pb exposure and conduct
disorders that are described in Section 3.5.3.

[a] Cortical thickness (n = 508)
3.00-

I

£ 2.50-

• •
i ijiiiiliii;1: • •• :

mi ii I j-n» v

B

b = -0.004; 95% CI, -0.012 to 0.004; P = .39

10	20

Blood lead level at age 11 y, pg/dL

c] Hippocampal volume (n = 508)

h

hzzh

*	!' ¦ ii«* ¦

. i : i 15 ' • , • •

•	i • . •	• •

b = -0.10; 95% CI, -0.17 to -0.03; P = .006

10	20

Blood lead level at age 11 y, pg/dL

Fractional anisotropy (n = 503)

.11 " i - !• I	i

nr

s ! t

• I *

It

ill

+V1

i i i j! |' *. . • -

b = -0.12; 95% CI, -0.24 to -0.01; P = .04

10	20

Blood lead level at age 11 y, pg/dL

| B Cortical surface area (n = 508)

I

240-
220-



•
;

*
.

:

• •
• *

* . "



-r

i

1

i t
i •

rr

5 1 • . 1 • ' . • *

180-

f:

1





;1i^ » T



•



i

« 1

• 2

! • i I • • • •

• i • •

140-









b--

-1.19; 95% CI, -2.35 to -0.02; P = .05

0

10	20

Blood lead level at age 11 y, pg/dL

[p] WMH volume (n = 502)

i—~:

E 6-



b = 0.05; 95% CI, -0.02 to 0.13; P = .17

10	20

Blood lead level at age 11 y, pg/dL

|~F~ BrainAGE index (n = 512)

I

20-
10-

« !: ! •

i-li|fj|ii|j;L:	-

mriHimsii" -: :



b = 0.77; 95% CI, 0.02-1.51; P = .05

10	20

Blood lead level at age 11 y, pg/dL

BrainAGE = Brain Age Gap Estimation; Ci = confidence interval; MRI = magnetic resonance imaging; SES = socioeconomic status.
WMH = white matter hyperintensities; y = year.

Regression lines and their 95% CIs are plotted. The box plots show the distribution of BLLs and brain outcomes.

Beta coefficients shown are for an incremental increase of 5 pg/dL in childhood BLL. Source: Reuben et al. (2020).

Figure 3-3 The relationship between blood Pb level at age 11 and brain
outcomes in adulthood.

External Review Draft

3-16

DRAFT: Do not cite or quote

-------
1

Lamoureux-Tremblav et al. (2021) examined the association between pre- and postnatal
(measured concurrently to the MRI) BLLs and MRI findings in adolescents (mean age 18.3 years)
enrolled in a longitudinal study of Inuit from Northern Quebec exposed to Pb, mercury (Hg), and
polychlorinated biphenyls (PCBs). Functional MRI data were collected during fear conditioning and
extinction tasks with the aim of understanding emotional dysregulation that could lead to anxiety
disorders. These authors found higher differential activation in the right dorsolateral prefrontal cortex in
association with higher postnatal BLL. Differential effects in the high Pb exposure group were observed
during the fear extinction phase, and maintained discrimination between the safety and threatening signals
(CS+ > CS-). Activation of the dorsolateral prefrontal cortex has been associated with cognitive processes
for regulating the affective state. The mean cord blood Pb was 4.56 (ig/dL and the mean concurrent BLL
was 1.78 (ig/dL in this study. In an earlier study of this Inuit population, Ethier et al. (2012) measured
visual evoked potentials (VEPs) using electrodes on the scalp to provide a direct measure of brain
functions related to sensory function (i.e., visual contrast sensitivity and spatial vision) at age 5. The mean
cord BLL in this study was 4.6 (ig/dL. Amplitude and latency for standard VEP components were
measured (i.e., N75 [negative deflection at approximately 75 ms], P100 [positive deflection at
approximately 100 ms], and N150 [negative deflection at approximately 150 ms]) in this longitudinal
study. Multiple tests were conducted, and associations were reported with significance levels. Cord Pb
level was associated with adelay of the N150 component (e.g., [3=0.06 [95% CI: 0.01, 0.10]) and other
latency metrics, which may indicate a deficit in early visual processing in Pb-exposed children.
Confounders including child characteristics, maternal education, SES, drug and alcohol use, and other
metals were considered in the analyses of Inuit children (see Table 3-1E for study-specific confounders
considered). A subset of two separate studies was combined for this study, and participation rates based
on the original number of participants enrolled in the study were not reported.

In a cross-sectional analysis of children, Kim et al. (2018a) examined the interaction between
dopamine receptor D2 (DRD2) and BLL on the cortical thickness of 12 regions of the frontal lobe
ascertained via MRI. The D2 receptor is located in the prefrontal cortex of the brain and may contribute to
the pathology of attention deficit/hyperactivity disorder (ADHD). The authors relied on an age- and sex-
matched sample of children ages 6 to 17 years old with and without confirmed ADHD for the analysis.
This study found an interaction effect between a variant of DRD2 and BLL on reduced cortical thickness
of several regions in the frontal lobe in the ADHD group, but not in the healthy controls in regression
analyses adjusted for age, intracranial volume, and sex. A correlation between reduced cortical thickness
and poorer inattention score on the parent-reported ADHD rating scale was also reported, supporting a
link between Pb exposure, MRI findings, and functional decrements in attention. Results for Kim et al.
(2018a) are found in Section 3.7, Table 3-1E.

External Review Draft

3-17

DRAFT: Do not cite or quote

-------
1

3.4.1.1 Summary

Overall, multiple studies, including prospective studies following children through adolescence
and adulthood, found associations between BLL and physiologic changes in regions of the brain
responsible for cognition and behavior. The prospective studies considered important confounders
including the HOME score (CLS only), SES, drug and alcohol use, and maternal IQ. A smaller number of
studies indicated Pb-related changes in brain function or physiologic changes in children. Associations
between Pb exposure and a large number of metrics and brain regions were evaluated in the
epidemiologic studies, raising the likelihood of chance findings.

3.4.2 Experimental Animal Studies of Brain Structure and Function

Animal studies can offer insights into Pb-induced effects on brain structure and function through
investigations that cannot be conducted on human subjects. Experimental animal studies provide evidence
that Pb can cause alterations in brain development. The 2013 Pb ISA reviewed evidence that Pb exposure
produced increased levels of oxidative stress and inflammatory response markers (U.S. EPA. 2013a).
These effects were observed in many regions of the brain and were associated with changes in neuronal
and glial cell morphology, neurotransmitter levels, and brain electrophysiology. Additional discussion of
the biological pathways that potentially underlie these nervous system effects are discussed in Section 3.6.

Recent studies (see Figure 3-3 and Table 3-1T) support the results summarized in the 2013 Pb
ISA, showing increases in inflammatory responses and markers of oxidative stress in various regions of
the brain. Most studies evaluated oral dosing of Pb via drinking water or gavage with exposure durations
ranging from 14 days to 701 days depending on the study, with studies in both adult and developing
animals. The specific dosing regimen varied by study, but the present review focuses on studies that
resulted in a measured BLL <30 (ig/dL. In studies with multiple time points, the magnitude or severity of
effects generally increased with exposure duration. In addition, studies with developmental Pb exposure
identified pregnancy and early development as sensitive windows for Pb toxicity. Several studies have
examined brain architecture using histological methods following Pb exposure. In general, the magnitude
and severity of the effects increased with longer exposure durations, and some studies found Pb-induced
effects could be ameliorated by co-exposures with antioxidants (such as vitamins and specific lipids) or
by environmental conditions (e.g., rearing condition or enrichment).

3.4.2.1 Histopathology

The nervous system is made up of the central and peripheral nervous systems, which include a
diversity of cell types broadly grouped into neurons and glial cells. Neurons are the functional electrically
excitable cells in the brain. Glial cells can be further divided into microglia and macroglia (astrocytes,
oligodendrocytes, ependymal cells, Schwann cells, satellite cells, radial glia, and enteric glia), which are

External Review Draft

3-18

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

neuronal support cells with diverse functions including innate immunity, phagocytosis, myelination,
synaptic regulation, neuronal activity, and blood-brain barrier (BBB) integrity (Rca. 2015). In the central
nervous system (CNS), astrocytes, oligodendrocytes, and microglia are the major types of glial cells.
Normal brain development relies on coordination of all cell types across time. Pathology methods can be
used to study alterations in brain cell morphology, function, and composition.

Short-term studies (<30 days) in adults evaluating Pb exposure and brain morphology after oral
exposures in rats have observed abnormalities in treated animals and their offspring, including decreased
numbers of neurons or synapses (Nam et al.. 2018a; Saleh et al.. 2018; Gassowska et al.. 2016a; Han et
al.. 2014; Rahman et al.. 2012b). disorganized cells and lack of characteristic layering (Saleh et al.. 2019;
Saleh et al.. 2018; Zhou et al.. 2018). increased vacuolization (Saleh et al.. 2019). effects on dendritic
spines (Xiao et al.. 2020; Saleh etal.. 2018; Wang et al.. 2016; Du et al.. 2015; Rahman et al.. 2012b).
increased numbers of apoptotic cells (Saleh et al.. 2019; Meng et al.. 2016). and increased expression of
various proteins and biochemical parameters related to oxidative stress (Saleh et al.. 2018; Singh et al..
2017; Zhu et al.. 2013) in multiple brain regions including the cerebellum, cerebral cortex, and
hippocampus. In aggregate, this evidence suggests that Pb has the potential to disrupt the integrity of
single neurons and populations, which may contribute to overt toxicity. Pb exposure has also been shown
to interfere with the homeostasis of other essential metal ions, such as iron (Fe), in the brain (Zhu et al..
2013) and to inhibit various enzymes involved in energy production or glucose uptake (Zhao et al.. 2021)
and metabolism (reviewed in (ATSDR. 2020)). For all these studies, BLLs were below 30 (ig/dL, and
many were below 20 (ig/dL (see Evidence Inventory Table 3-1T). Further discussion of these mechanisms
is provided in Section 3.6.

Studies of long-term Pb exposure (>30 days) have also assessed brain morphology in adult mice
and rats following oral dosing and observed neuronal damage including irregular shape, vacuolization,
and cell degeneration in the cerebellum, hippocampus, and cerebral cortex (Liu et al.. 2022c; Saleh et al..
2019; Singh et al.. 2019; Saleh et al.. 2018; Sun et al.. 2014). Other histopathological lesions, including
karyopyknosis (pre-apoptotic chromatin condensation of cell nuclei) (Nan et al.. 2016) and swollen and
distorted mitochondria (Ouvang et al.. 2019; Gassowska et al.. 2016a; Sun et al.. 2014) were observed in
other studies. Several biochemical parameters related to oxidative stress were assessed including
apoptosis using immunohistochemical methods like terminal deoxynucleotidyl transferase dUTP nick end
labeling (TUNEL staining) (Singh et al.. 2019; Baranowska-Bosiacka et al.. 2017; Meng et al.. 2016; Nan
et al.. 2016; Su et al.. 2016; Baranowska-Bosiacka et al.. 2013) as well as caspase-3 (Wang et al.. 2021a)
and cell replication using proliferating cell nuclear antigen (PCNA) (Singh et al.. 2019). In a 12-week rat
dietary study, the authors reported neuronal damage and cognitive deficits accompanied by decreased
levels of synaptic proteins as well as decreased levels of receptors and proteins related to synaptic
plasticity regulation (N-methyl D-aspartate receptor [NMDAR], cyclic adenosine 3',4'-monophosphate
response element binding protein [CREB], brain-derived neurotrophic factor [BDNF]) (Liu et al.. 2022c).
BLLs in all of these studies were below 30 (ig/dL and many were below 20 (ig/dL (see Evidence
Inventory, Table 3-IT).

External Review Draft

3-19

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

There is a single inhalation study, adult mice were dosed for 6 weeks via whole body inhalation
of Pb oxide (PbO) nanoparticles, and hippocampal damage was observed including shrunken and
damaged neurons, as well as increased Pb content in the brain (Dumkova et al.. 2017). The study reported
BLLs of 13.99 (ig/dL. However, exposure to PbO nanoparticles via inhalation did not impact markers of
cell proliferation (PCNA) or cellular apoptosis (TUNEL) in the hippocampus (Dumkova et al.. 2017).
Despite the limited evidence available for inhaled Pb, continuity of the effects (i.e., neuronal damage) has
been reported across different routes of exposure.

Brain morphology has also been assessed in developing animals with a variety of exposure
contexts including pre-mating, during gestation, and during lactation. Pb is transferred across the placenta
and through lactation (Silbergeld. 1991; Bhattacharvva. 1983). Some studies included cross fostering
experiments to assess unique sensitivities at specific developmental windows, with some evidence
indicating that the postnatal period is particularly sensitive to Pb neurotoxicity (Barkur and Bairv. 2016).
Given the altricial nature of rodents, the postnatal period is roughly analogous to the third trimester of
human brain development. Similar to results in adults, studies of developmental Pb exposures observed
alterations in brain morphology including reduced numbers of neurons in the forebrain, hippocampus,
hypothalamus, and amygdala (Long et al.. 2022; Vigueras-Villasenor et al.. 2021; Wang et al.. 2021a;
Nam et al.. 2018a; Shvachiv et al.. 2018; Xiao et al.. 2014). damaged neurons (Long et al.. 2022; Wang et
al.. 2021a; Zhu et al.. 2013). reduced numbers of glial cells (Dominguez et al.. 2019; Sobin et al.. 2013).
changes in synapses (Sadeghi et al.. 2021; Wang et al.. 2021b; Gassowska et al.. 2016a; Gassowska et al..
2016b; Zhang et al.. 2015b; Xiao et al.. 2014). reduced numbers of mitochondria (Zhang et al.. 2015b).
swollen and shrunken mitochondria (Ouvang et al.. 2019; Gassowska et al.. 2016a; Baranowska-Bosiacka
et al.. 2013). altered levels of glycoconjugates (constituents of synaptic and neural membranes) (Sadeghi
et al.. 2021). and chromatin abnormalities (Ouvang et al.. 2019; Baranowska-Bosiacka et al.. 2013). BLLs
in all of these studies were below 30 (ig/dL and many were below 20 (ig/dL. In a rat developmental study
in which animals were dosed throughout pregnancy and lactation (gestational day [GD] 1 to postnatal day
[PND] 21) with BLLs of 6.86 (ig/dL, the authors reported pathological changes in synapses, including
swelling of nerve endings, thickened synaptic cleft structure, and abnormalities in synaptic vesicle density
(Gassowska et al.. 2016a). These changes in synapse morphology were accompanied by decreases in key
synaptic proteins, as well as BDNF, a key neurotrophic factor that supports the differentiation,
maturation, and survival of neurons in both development and adulthood (Gassowska et al.. 2016a).
Additionally, studies in adult humans suggested that decreases in BDNF are associated with
neurodegenerative diseases (Bathina and Das. 2015). These changes in synapses can result in synapse
dysfunction, which would contribute to altered neurotransmission. Adverse changes in neuronal dendrite
morphology were reported following developmental Pb exposures in multiple studies, including loss of
dendritic spines, reduced spine density, decreased spine length, and impaired spine maturity and
morphology at multiple developmental stages and brain regions (hippocampus, medial prefrontal cortex,
dentate gyrus) (Xiao et al.. 2020; Saleh et al.. 2018; Zhao et al.. 2018; Sepehri and Garni. 2016; Wang et
al.. 2016; Du et al.. 2015; Rahman et al.. 2012b). Dendritic spines are the morphological and structural

External Review Draft

3-20

DRAFT: Do not cite or quote

-------
1

basis for synaptic plasticity, learning, and memory (Frank et al.. 2018). providing biological plausibility
for altered learning and memory, as reported in Section 3.5.1.3.2.

The brain and CNS are separated from the blood by the BBB and the blood-cerebrospinal fluid
(CSF) barrier, which allows for selective transport of materials into the CNS. The BBB is formed by
several cell types including endothelial cells, astrocytes, pericytes, and microglia and plays an important
immunological role in protecting the brain from circulating pathogens and toxic substances. Two studies
assessed the integrity of the BBB following Pb exposure and observed disruption of brain permeability
with reduced levels of tight junction proteins and other important capillary proteins (Wu et al.. 2020a;
Song etal.. 2014). Increased permeability of the BBB may exacerbate neurotoxicity, as more toxicants
can penetrate the brain with repeated or continuous exposure. Additional research is needed to fully
elucidate the effects of Pb exposures on the BBB and the potential implications for CNS function and
disease.

Neuroinflammation is a complex response that involves microglia and astrocyte activation, as
well as other signaling proteins and cells (such as cytokines, reactive oxygen species, decreased
antioxidant activity). Inflammation is protective against pathogens but can result in neuronal injury or
neuronal loss in the CNS. Several studies in rodents evaluated and observed an effect of Pb exposure on
markers of neuroinflammation, including microglia and astrocyte activation, and the promotion of cellular
reactivity and inflammation (Wu et al.. 2020a; Saleh et al.. 2018; Shvachiv et al.. 2018). Numerous
studies also reported decreased numbers of microglia, which are glial cells that function primarily as
immune cells with macrophage activity, clearing cellular debris and dead neurons from nervous system
tissue (Dominguez et al.. 2019; Sobin et al.. 2013). Reduced numbers of microglia could indicate reduced
capacity for clearing cellular debris and responding to pathogens, which could contribute to functional
and morphological brain changes.

Experimental animal studies of rodents have also shown that Pb exposures affect measures of
brain metabolism, including reduced glycogen concentrations in various brain regions (such as the
forebrain, hippocampus, and cerebellum) (Baranowska-Bosiacka et al.. 2017) and reduced rates of
metabolism, which could indicate reduced glucose availability and poor metabolic cooperation between
neurons and astrocytes (Baranowska-Bosiacka et al.. 2017). A recent study directly measured decreased
hippocampal glucose metabolism following developmental Pb exposure (pre-mating through PND 10)
through reductions in glucose transporters (Zhao et al.. 2021). Glucose cannot be synthesized or stored in
neurons, hence glucose supply and transport are essential for neurophysiological processes with high
glucose demands (such as learning and memory). Because the brain has the highest energy demand and
metabolism of any organ, energy homeostasis is critical. In the above study, effects on glucose
metabolism persisted at PND 30 when the blood Pb concentration had returned to control levels (BLL
was 11.4 (ig/dL at PND 10 and 1.8 (ig/dL BLL at PND 30).

Taken together, animal studies provide strong evidence that Pb exposure impacts brain structure
and function. Altered brain morphology, increased brain inflammation, oxidative stress, and associated

External Review Draft

3-21

DRAFT: Do not cite or quote

-------
1

mitochondrial damage have all been consistently reported following Pb exposures. These effects were
observed across multiple brain regions, on different levels of brain organization, across a variety of
lifestages, and in both sexes.

3.4.2.2 Neurotransmitter Analysis

Neurotransmitters are molecules involved in the transmission of chemical signals between
neurons and target cells and are involved in controlling a wide variety of brain functions, including motor
function, learning, memory, metabolism, behavior, and hormone production. These neurochemical
systems have been implicated in the initiation and maintenance of some brain diseases and disorders, e.g.,
Parkinson's disease (PD), depression, aggression, and dementia (Monday et al.. 2018; Chichinadze et al..
2011; Haden and Scarpa. 2007; Webster. 2001). As described in the 2013 Pb ISA, exposures to Pb can
induce changes in brain neurochemistry and signaling that vary by brain region, neurotransmitter type,
and the sex of the animal. Pb can compete with calcium (Ca2+) for common binding sites and second
messenger system activation. When Pb activates a Ca2+-dependent system in the nervous system, it can
contribute to spurious neurotransmitter regulation and release because this system intimately relies on
Ca2+ signaling for its homeostasis. Pb-related alterations in neurotransmission are discussed in further
detail below.

A variety of neurotransmitters and their metabolites were evaluated in experimental animal
studies of rodents, across multiple brain regions (hypothalamus, cerebral cortex, nucleus accumbens,
frontal cortex, striatum, hippocampus, olfactory bulb, midbrain, cerebellum) and time points, including
serotonin and its metabolite 5-hydroxylindolacetic acid (Weston et al.. 2014; Mansouri et al.. 2013;
Graham et al.. 2011). norepinephrine and its metabolite methoxyhydroxyphenylglycol (Long et al.. 2022;
Basha et al.. 2014; Weston et al.. 2014; Biioor et al.. 2012; Graham et al.. 2011). dopamine and its
metabolites dihydroxyphenylacetic acid and homovanillic acid (Sobolewski et al.. 2020; Amos-Kroohs et
al.. 2016; Stansfield et al.. 2015; Basha et al.. 2014; Weston et al.. 2014; Corv-Slechta et al.. 2012;
Graham et al.. 2011). acetylcholine (Long et al.. 2022; Mansouri et al.. 2013). epinephrine (Basha et al..
2014). glutamate and its precursor glutamine (Long et al.. 2022). The direction of changes depended on
the brain tissue analyzed, time point, sex, and specific neurotransmitter assessed. However, multiple
studies found significant effects of Pb exposure on the dopamine system (Sobolewski et al.. 2020; Amos-
Kroohs et al.. 2016; Stansfield et al.. 2015; Basha et al.. 2014; Weston et al.. 2014; Corv-Slechta et al..
2012; Graham et al.. 2011). Some studies also reported Pb-induced changes in enzymes involved in
neurotransmitter turnover and cycling, including monoamine oxidase (Basha et al.. 2014) tyrosine
hydroxylase (Sobolewski et al.. 2020). glutamine synthase, and adenylate cyclase (Long et al.. 2022).

Pb exposure has demonstrated effects on several neurotransmitters, which are important signaling
molecules that control multiple brain functions. These effects were observed across multiple brain
regions, across a variety of lifestages, and in both sexes. Altered neurotransmitter signaling can contribute

External Review Draft

3-22

DRAFT: Do not cite or quote

-------
1

to multiple brain dysfunctions and disorders, providing biological plausibility for the health effects
discussed in subsequent sections.

3.4.2.3 Brain Weight

Organ weights are a frequently assessed in experimental animal studies as they can easily be
measured during animal necropsy. Importantly, brain weight is an indication of severe toxicity as the
body goes to great lengths to spare the brain at the expense of other bodily systems. Brain weights were
not reviewed in the 2013 Pb ISA. Several recent studies of rodents assessed brain weight following short
and long-term Pb exposures, and most of these studies reported nonsignificant findings (Vigucras-
Villasenor et al.. 2021; Mani et al.. 2020; Wu et al.. 2020a; Singh et al.. 2019; Rahman et al.. 2018; Saleh
et al.. 2018; Zhou et al.. 2018; Singh et al.. 2017; Barkur and Bairv. 2015a; Wang et al.. 2013; Rahman et
al.. 2012b). However, two studies did find significant decreases in brain weight (16-21% decrease): one
reported a decrease after a 90-day oral exposure to Pb in juvenile rats, which resulted in a BLL of
28.4 (ig/dL (Singh et al.. 2019) and the other reported an 18% decrease in cerebellum weight in treated
dams (27.7 (ig/dL BLL), as well as reduced fetal brain weight at parturition following gestational
exposure to Pb in drinking water (GD 1 to GD 20)(Saleh et al.. 2018). In addition to gross measures of
brain size and morphology (e.g., wet weight), studies using newer anatomical imaging methods have been
conducted since the 2013 Pb ISA. Three-dimensional imaging technologies, such as MRI and ultrasound
methods are being used for the analysis of neuroactivity and phenotypes in rodent toxicology studies
(Turnbull and Mori. 2007). Brain volume and MRI morphometry were assessed in a single study of
developing mice following dietary exposure to Pb (Abazvan et al.. 2014). The authors reported no
changes in lateral ventricle volume but did observe sex-specific changes in morphology including
enlarged lateral ventricles. As the technologies improve, imaging technologies offer promising results for
evaluating physiological functions like neural activity in whole intact animals. Study details are provided
in Table 3-1T.

3.4.2.4 Electrophysiology

The effects of Pb exposure on brain electrophysiology were not reviewed in the 2013 Pb ISA.
Several new studies in rodents have found that Pb exposure affects measures of brain electrophysiology,
including long-term potentiation (LTP) and evoked excitatory postsynaptic currents (EPSCs). LTP is the
process of signal transmission by which synaptic connections between neurons are activated and
strengthened and may be one of the mechanisms underlying learning and memory processes. Recording
of LTP is a recognized model for the study of memory (Lynch etal.. 1990).

Presynaptic plasticity can be assessed using paired-pulse stimulation, wherein two stimuli occur
in close succession. Hippocampal slices from rats were subjected to LTP induction and high-frequency

External Review Draft

3-23

DRAFT: Do not cite or quote

-------
1

tetanic stimulations, and the magnitudes of EPSCs were measured. In developmentally Pb-exposed rats,
the magnitudes of EPSCs were lower at PND 10, suggesting impaired hippocampal induction (Zhao ct al..
2018). Other studies have also assessed EPSCs with a longer exposure duration and different conditions
and found the ratios of EPSC responses between paired-pulse stimuli were significantly greater in
hippocampal slices from Pb-exposed rats (Zhang et al.. 2015b). These changes in EPSC responses were
accompanied by inhibition of synaptic vesicular release (Zhang et al.. 2015b). Depressed LTP following
Pb exposures was measured in (Zhou et al.. 2020a; Wang et al.. 2016; Liu et al.. 2012). These studies
additionally reported increased neuronal free Ca2+ concentration and inhibition of various signaling
proteins (Ca2+ calmodulin dependent protein kinase II and CREB), which were mediated by upregulation
of the ryanodine receptor. Ryanodine receptors are ion channels that are critical for maintaining
intracellular Ca2+ homeostasis. Changes in electrophysiological parameters affect neurotransmitter release
and cell signaling, which can affect brain function (described in Section 3.5.2.2).

In addition to these effects, (Zhu et al.. 2019a) reported alterations in cardiac sympathetic nerve
activity in rats while evaluating nerve discharge as a potential contributor to other health effects discussed
in the cardiovascular toxicity section (Appendix 4). The authors reported enhanced cervical sympathetic
nerve discharge 1 year after Pb exposure ended, suggesting that Pb-induced alterations to autonomic
nervous dysfunction can have lasting effects. This growing area of research recognizes the potential effect
of Pb on electrophysiology in the nervous system.

3.4.2.5 Circadian Rhythms

Two recent studies evaluated the effects of Pb exposure on circadian rhythms using rodent
models. The suprachiasmatic nucleus (SCN) of the hypothalamus is the primary regulator of circadian
physiological processes and is synchronized daily by signals of light. Vigueras-Villasenor et al. (2021)
subjected male rats to chronic Pb exposure from conception to euthanasia. In these adult rats, under a
standard 12:12-hour light-dark cycle, the authors observed daily delays in the nocturnal onset of
locomotor activity. With a 6-hour photoperiod delay, the activity rhythms of Pb-exposed rats entrained to
a new cycle faster than controls, and Pb treatment showed no significant effects when the photoperiod
was advanced by 6 hours. Histochemical analyses of the hypothalamus in light-pulsed Pb-treated animals
displayed decreases compared with controls in both photo-stimulated neurons (immunoreactivity to c-
Fos) and the neuronal population in the SCN. Hsu et al. (2021) assessed disturbances in rodent sleep
homeostasis by using electroencephalography and electromyography to score the sleep wake architecture
of sleep cycles and found that adult rats with chronic Pb exposure showed disturbances in sleep patterns
that were accompanied by altered clock gene expression and changes in the hypothalamus. These
alterations in behavior, sleep cycles, brain structure, neuronal function, and gene expression warrant
further investigations into the effects of Pb on the rhythm of vital circadian processes. The above studies
reinforce the importance of considering the time of day in studies measuring the effects of Pb.

External Review Draft

3-24

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

3.4.2.6

Summary

In conclusion, multiple studies measured a variety of nervous system endpoints in brains of rats
and mice following exposure to Pb including histology, neurotransmitter analysis, brain weight, and
electrophysiology measures. Histological analyses revealed reduced neuron counts, altered synapse
morphology, and increased apoptosis, as well as oxidative damage in several brain regions, including the
hippocampus, frontal cortex, and cerebellum. These regions were also found to have damaged
mitochondria, vacuolization, and morphological changes. Pb concentrations ranged from 4.7 (ig/dL to
28.4 (ig/dL in these studies, which were conducted in a variety of animal models, sexes, and lifestages.
These endpoints provide biological plausibility for effects on cognitive behavioral changes and diseases
described in subsequent sections.

3.4.3 Integrated Summary of Overt Nervous System Toxicity

Multiple studies measured nervous system endpoints in rats and mice following exposure to Pb
including histological changes in brain structure and morphology, neuroinflammation, neurotransmitter
analysis, brain weight, and electrophysiology measures. These findings that Pb exposures affect these
endpoints provide biological plausibility for Pb to elicit human cognitive behavioral changes and diseases
described in subsequent sections, and are generally coherent with the epidemiologic studies of overt
nervous system effects described in this section. Multiple epidemiologic studies, including prospective
studies following children through adolescence and adulthood and a smaller number of cross-sectional
studies of children, found associations between BLL and physiologic changes in regions of the brain
responsible for cognition and behavior. These prospective epidemiologic studies considered important
confounders; however, many Pb exposure metrics and brain regions were evaluated in the epidemiologic
studies, raising the possibility of chance findings.

As described in Section 3.1, no causality determinations will be developed for this section.
Instead, the evidence is considered supporting information that informs the health determinations in
Sections 3.5 and 3.6.

3.5 Nervous System Effects Ascertained during Childhood,
Adolescent, and Young Adult Lifestages

The collective body of epidemiologic and experimental animal studies assessed in the 2013 Pb
ISA demonstrated the effects of Pb exposure on an array of nervous system outcomes. Overall, the largest
body of evidence assessed in the 2013 Pb ISA, as well as the outcome that was best substantiated to occur
at the lowest Pb exposure levels, was related to cognitive effects in children. Multiple prospective studies
conducted in diverse populations consistently demonstrated associations of higher blood and tooth Pb

External Review Draft

3-25

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

levels with lower full-scale IQ (FSIQ), executive function, and academic performance and achievement.
Most studies examined representative populations and had moderate to high follow-up participation with
no indication of selective participation among children with higher BLLs and lower cognitive function.
Associations between BLL and cognitive function decrements were found with adjustment for several
potential confounding factors, most commonly SES, parental IQ, parental education, and parental
caregiving quality. In children aged 4-11 years, associations were found with prenatal, early childhood,
childhood average, and concurrent BLLs in populations with mean or group BLLs in the range of 2-
8 (ig/dL. Although examined less extensively than cognitive effects, a strong body of evidence also
indicated Pb-associated decrements in attention and increased hyperactivity.

3.5.1 Cognitive Function in Children

The evidence evaluated in the 2013 Pb ISA was sufficient to conclude that there is a "causal
relationship" between Pb exposure and decrements in cognitive function in children (U.S. EPA. 2013a).
Multiple prospective studies conducted in diverse populations consistently demonstrated associations of
higher blood and tooth Pb levels with lower FSIQ, executive function, and academic performance and
achievement. Most studies examined representative populations and had moderate to high follow-up
participation with no indication of selective participation among children with higher BLLs and lower
cognitive function. Associations between BLL and cognitive function decrements were found with
adjustment for several potential confounding factors, most commonly SES, parental IQ, parental
education, and parental caregiving quality. In children aged 4-11 years, associations were found with
prenatal, early childhood, childhood average, and concurrent BLLs in populations with mean or group
BLLs in the range of 2-8 (ig/dL. No critical lifestage or specific duration of Pb exposure within childhood
was uniquely associated with cognitive function decrements based on consideration of evidence from
epidemiologic and toxicological studies. Several epidemiologic studies found a supralinear concentration-
response (C-R) relationship (i.e., larger incremental effect at lower BLLs). A threshold for cognitive
function decrements was not discernable from the available evidence (i.e., examination of early childhood
blood Pb or concurrent [with peak <10 |ig/dL| blood Pb in the range of <1 to 10 (ig/dL). Evidence in
children was clearly supported by observations of Pb-induced impairments in learning and memory in
juvenile animals. Several studies in animals indicated learning impairments with prenatal, lactational,
post-lactational, and lifetime (with or without prenatal) Pb exposures that resulted in BLLs of 10-
25 (ig/dL. Biological plausibility for Pb-associated cognitive function decrements was supported by
observations of Pb-induced impairments in neurogenesis, synaptogenesis and synaptic pruning, LTP, and
neurotransmitter function in the hippocampus, prefrontal cortex, and nucleus accumbens.

The structure of the current assessment of cognitive effects in children is similar to that in the
2013 Pb ISA. Although the above measures of cognitive function are interrelated, the evidence for each of
these categories of outcomes (i.e., FSIQ, Bayley Scales of Infant Development [BSID],
neuropsychological tests of learning, memory, and executive function, and academic performance) was

External Review Draft

3-26

DRAFT: Do not cite or quote

-------
1

assessed separately, to the extent possible, in the order of strength of evidence. Studies assessing
cognitive function of school-age children using instruments that measure FSIQ are described in
Section 3.5.1.1, and studies assessing cognitive development in infants using the BSID and other
instruments are described in Section 3.5.1.2. Studies examining the associations of Pb exposure with
outcomes on neuropsychological tests of learning and memory and executive function in children as well
as analogous endpoints in animals are discussed in Sections 3.5.1.3 and 3.5.1.4, respectively. These
sections are followed by a discussion of studies that examine the association of Pb exposure with
academic achievement and performance in Section 3.5.1.5. The final sections discuss issues relevant for
the interpretation of the evidence base (Section 3.5.1.6) and the integrated summary and causality
determination (Section 3.5.1.7).

Because the conclusion from the 2013 Pb ISA was "causal," the PECOS statement for studies of
cognitive effects in children (see Section 3.2) was refined to emphasize recent studies that examined
lower BLLs more similar to those of current U.S. children (i.e., <5 (ig/dL). Details of these studies are
extracted into the evidence inventories (Section 3.7, Table 3-2E [FSIQ], Table 3-3E [Infant
Development], Table 3-4E [Learning, Memory, and Executive Function], and Table 3-5E [Academic
Achievement and Performance]). Studies with central tendency blood Pb concentrations that exceed
5 (ig/dL are extracted into Table 3-6E of Section 3.7. In addition to refining the PECOS statement to
focus on lower exposure levels, studies of younger children whose BLLs were less influenced by higher
past Pb exposures are considered particularly informative. Controls for important potential confounders
identified in the 2013 Pb ISA such as SES, parental education, quality of parental caregiving (often
measured as the HOME score), nutritional status, and birth weight were considered attributes of high-
quality studies (see section 4.3.13 of the 2013 Pb ISA (U.S. EPA. 2013a)). A summary of the recent
evidence, which is interpreted in the context of the entire body of evidence, is provided in the subsequent
sections. Overall, recent studies add to the evidence generally supporting the findings from the 2013 Pb
ISA.

3.5.1.1 Full-Scale IQ in Children

A large number of studies evaluated in the 2013 Pb ISA found a consistent pattern of associations
between higher BLL and lower FSIQ in children aged 4-17 years (see Figure 4-2 and Table 4-3 (U.S.
EPA. 2013a'). FSIQ has strong psychometric properties (i.e., reliability, consistency, validity), is among
the most rigorously standardized cognitive function measures, is relatively stable in school-age children,
and has been demonstrated to be predictive of educational achievement and life success. The strongest
evidence was provided by prospective studies with analyses of the association of blood or tooth Pb levels
measured in early childhood before FSIQ was assessed. These prospective studies typically considered
potential confounding by maternal IQ and education, SES, birth weight, smoking exposure, parental
caregiving quality, and in a few cases, other birth outcomes and nutritional factors. Associations were
found in diverse populations (e.g., Boston, MA; Cincinnati, OH; Rochester, NY; Cleveland, OH; Mexico
City, Mexico; Port Pirie, Australia; and Kosovo, formerly of Yugoslavia) in studies that examined

External Review Draft

3-27

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

children recruited from prenatal clinics, hospital maternity departments, or schools. Studies generally
reported high follow-up participation, which was supported by evidence that selection bias did not explain
the associations observed. The few studies reporting weak or null associations (i.e., Cleveland, Sydney
cohorts) were not stronger with respect to methodology or control for potential confounding and did not
weaken the far larger body of supporting evidence (U.S. EPA. 2013a).

The blood Pb-FSIQ association in children was further substantiated by an international pooled
analysis of seven prospective cohorts (Lanphcar et al.. 2019. 2005) as well as multiple meta-analyses that
combined results across various prospective and cross-sectional studies (Pocock et al.. 1994; Schwartz.
1994a; Needleman and Gatsonis. 1990). Schwartz (1994a) additionally demonstrated the robustness of
evidence to potential publication bias. The pooled analysis (Lanphcar et al.. 2019. 2005) examined several
BLL metrics and demonstrated that early childhood and concurrent childhood BLLs explained more
variation in FSIQ compared with the other blood Pb metrics, as indicated by the R-square values. The
coefficient for concurrent BLL had a smaller standard error (SE) than the coefficient for early childhood
BLL (Lanphear et al.. 2019). Across studies, no clear indication that Pb exposure during one critical
lifestage or time period within childhood was uniquely or more strongly associated with FSIQ (see
Section 3.5.1.6.3). Blood Pb-associated FSIQ decrements at ages 4-17 years were found with concurrent,
prenatal (maternal or cord), early childhood (e.g., age 2 or 4 years), multiple-year average, or lifetime
average BLLs. Associations were also found with tooth Pb levels.

Key statistics associated with the international pooled analyses of seven cohort studies are
presented in Table 3-1. The C-R function was nonlinear, with a larger incremental effect of Pb on IQ at
lower blood Pb concentrations (Lanphear et al.. 2019. 2005). The log-linear model coefficient (i.e., (3
coefficient) for concurrent BLL was -2.65 (95% confidence interval [CI]: -3.69,-1.61) per unit change
in natural log transformed BLL. The linear association observed for a subset of 103 children with peak
BLLs <7.5 (mean concurrent BLL = 3.2 (ig/dL) was -2.53 (95% CI: -4.48, -0.58). Linear coefficients for
higher BLLs and using a peak BLL cutoff point of 10 (ig/dL are included in Table 3-1, as are other key
statistics, including R-square values for various models.

Table 3-1 Statistics associated with the international pooled analysis of data
from seven cohort studies.

Main Finding from Analyses of the Pooled Dataset

Quantitative Result3

Log-linearb model coefficient for blood Pb metrics and IQ, adjusted for
site, HOME score, birth weight, maternal IQ, and maternal education

Early childhood: -2.21 (-3.38, -1.04)
Peak: -2.86 (-4.10, -1.61)

Lifetime average: -3.14 (-4.39, -1.88)
Concurrent: -2.65 (-3.69, —1.61 )c

IQ decrement over different concurrent blood Pb ranges based on the
log-linear model

2.4 to 30 [jg/dL: 6.7 IQ pts (4.1-9.3)
2.4 to 10 [jg/dL: 3.8 IQ pts (2.3-5.3)
10 to 20 [jg/dL: 1.8 IQ pts (1.1-2.6)

External Review Draft

3-28

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

Main Finding from Analyses of the Pooled Dataset	Quantitative Result3



20 to 30 |jg/dL: 1.1 IQ pts (0.7-1.5)

Linear coefficient,11 sample size (n) and concurrent BLL measurements
(mean, minimum, 5th and 95th percentiles, and maximum) for subset
with peak BLLs <7.5 |jg/dL

-2.53 (-4.48, -0.58 )
n = 118e

(3.3, 0.9, 1.1, 6.7, 7.4 pg/dL)

Linear coefficient,11 sample size (n) and concurrent blood Pb
measurements (mean, minimum, 5th and 95th percentiles, and
maximum) for subset with peak BLLs >7.5 |jg/dL

-0.15 (-0.19, -0.11)
n = 1215

(13.0, 0.1, 3.7, 34.2, 71.7)

Linear coefficient,11 sample size (n) and concurrent blood Pb d
measurements (mean, minimum, 5th and 95th percentiles and
maximum) for subset with peak blood Pb <10 |jg/dL

-0.77 (-1.65, 0.12)
n = 258

(4.4, 0.1, 1.4, 8.0, 9.8)

Linear coefficient,11 sample size (n) and concurrent blood Pb
measurements (mean, minimum, 5th and 95th percentiles, and
maximum) for subset with peak BLLs >10 pg/dL)

-0.13 (-0.22, -0.04)
n = 1075

(14.0, 0.1,4.4, 35.5, 71.7)

Blood Pb metric with the largest R2 for the relationship with IQ in the
log-linear models

Early childhood R2: 0.6433 = largest
Peak R2: 0.6401
Lifetime average R2: 0.6411
Concurrent R2: 0.6414

BLL = blood lead level; HOME = Health Outcomes and Measures of the Environment; IQ = intelligence quotient; Pb = lead; pts = points.
aResults reported by Lanphear et al. (2019) and/or Crump et al. (2013) and confirmed by Kirrane and Patel (2014).
b Coefficients are not standardized, i.e., coefficients indicate the decrement in full scale IQ per unit of natural log transformed
blood Pb. Standardized estimates (i.e., standardized to a 1 unit increase for the 10th—90th percentile interval of the biomarker
level and assumed to be linear within this interval) for this study are found in Evidence Inventory (Section 3.7, Table 3-2E).

°Slopes ranged from -2.36 to -2.94 in sensitivity analyses of concurrent blood Pb-IQ association, which omitted one cohort at a
time.

d Linear coefficients are standardized to a 1 pg/dL increase in blood Pb

eThe number of children from Boston cohort with peak BLLs <7.5 pg/dL was 28 after errors were corrected (Lanphear et al..

2019).

Several studies that conducted analyses stratified by BLL provide additional support for the
findings of Lanphear et al. (2005) and Lanphear et al. (2019). These studies comprise a compelling body
of evidence demonstrating a nonlinear C-R function for the association between BLL and intelligence.
This evidence is described in the 2013 Pb ISA (Section 4.3.12, Figure 4-15, and Table 4-16 of the U.S.
EPA (2013a)). In a recent analysis, Crump et al. (2013) examined the shape of the C-R function for the
pooled data using an alternative modeling strategy. Rather than model the natural log of BLL as was done
in the original analysis (Lanphear et al.. 2019. 2005). Crump et al. (2013) modeled the natural log (In) of
blood Pb + 1, which has the property of equaling zero when untransformed BLL equals zero. The authors
applied F-tests to nested models containing both In (BLL + 1) and non-transformed BLL and found that
the linear coefficient did not improve the prediction of the model, indicating that In (BLL +1) was a
better predictor across the full range of the data (e.g., 2.5-33.2, as 5th to 95th percentile concurrent
BLLs). In addition, Crump et al. (2013) considered confounding by additional covariates, which were
defined as site-specific in their final models. Despite the aforementioned differences in modeling
approach, the Crump et al. (2013) analysis corroborated the findings of the original analysis, providing
strong evidence in support of the nonlinear C-R function and the causal association between Pb exposure

External Review Draft

3-29

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

and cognitive effects in children. The coefficient for the Crump et al. (^O^log-linear association
between concurrent BLL and IQ was -3.32 (95% CI: -4.55, -2.08), somewhat larger than that reported
by Lanphear et al. (2019) (see Table 3-1).

Notably, the international pooled analysis (Lanphear et al.. 2019. 2005) included data from seven
longitudinal cohorts that were initiated before 1995. The median concurrent BLL was 9.7 (ig/dL (5th and
95th percentiles: 2.5-33.2 (ig/dL) with included studies reporting limits of detection of 1 (ig/dL. Several
other longitudinal and cross-sectional studies included in the 2013 Pb ISA, however, conducted analyses
of children with mean BLLs <5 (ig/dL, collectively providing strong evidence of an association between
Pb exposure and FSIQ at lower BLLs (Kim et al.. 2009; Chiodo et al.. 2007; Surkan et al.. 2007;
Bellinger and Needleman. 2003; Canfield et al.. 2003a) (see Figure 3-4).

Several recent longitudinal studies add to the evidence informing the relationship between BLL
and IQ in children. Heterogeneity in the magnitude and direction of the associations, which was
potentially explained by race/ethnicity, sex, and modeling choices such as adjustment for other metals or
chemicals was present. This heterogeneity did not weaken the larger body of supporting evidence.
Overall, recent studies generally corroborated previous epidemiologic observations of associations
between Pb exposure and IQ in children with relatively low blood Pb concentrations (<5 (ig/dL) (see
Figure 3-4 and Evidence Inventory Table 3-2E).

External Review Draft

3-30

DRAFT: Do not cite or quote

-------
Study

Prospective Studies

tTatsuta et al. 2020

tZhou et al 2020

Tohuku district. Japan Prenatal
Prenatal

Jiangsu Province, China Prenatal

tDesrochers-Couture et al. 2018 Quebec. Canada	Prenatal

tLee et al. 2021	8 hospitals. S Korea Prenatal

Bellinger et al 2003
tlglesias et al. 2011
tLee et al. 2021
tLee et al. 2021

Lanphear et al. 2005, 2019
Canfield et al. 2003
Kim et al. 2009

tTatsuta et al. 2020

tRuebner et al 2019	Multi-center U S

tlglesias et al. 2011	Northern Chile
tDesrochers-Couture et al. 2018 Quebec. Canada

Cross-sectional Studies

tDantzer et al. 2020	Cincinnati, OH

tMartin et al. 2021	East Liverpool. OH

tHong et al. 2015
tMenezes-Filho et al. 2018
tLucchini et al. 2012

5 regions. S Korea
Bahia. Brazil
Brescia. Italy

Mean

(pg/dL)

0.8

1.59

0.76
1.32

Boys

Girls

All

Boys

Girls

Boston area. MA
Northern Chile
8 hospitals, S Korea
8 hospitals, S Korea

International
Rochester. NY
4 Cities, Korea

Tohuku district, Japan

Early child; 2 yrs

Early child: 3-6 yrs	10.8	Pre-site closure

Early child: 4 yrs	1.41 (med)

Early child: 6 yrs	1.44(med)

FSIQ Adjusted for:

Age (yrs)

Hg, child Pb

Mn, Cd
Mn, Cd
Mn. Cd
3-4 Hg

Mn. Hg, Cd

10

10-13

5-15

5-15

Concurrent
Concurrent
Concurrent

Concurrent
Concurrent
Concurrent
Concurrent
Concurrent

Concurrent
Concurrent
Concurrent
Concurrent
Concurrent
Concurrent

3.2

3.3
1.7

1.2
3.5
0.7

0.57
1.13

1.8

1.64

1.71

4-10
5

6-10

<7.5 pg/dL
<10 pg/dL
Low Mn
High Mn
Boys
Girls

CKD patients
Post site closure 7-16
3-4

12

1-18

High Hair
Low Hair Mn

Hg. prenatal Pb *

12

7-9

8-11
7-12
11-14

l	I	I	I	I

-6.00	-4.00 -2.00	0.00	2.00

Beta values (95% CI) per 1 ug/dL increase in blood Pb

Note: Effect estimates are standardized to a 1 |jg/dL increase in blood Pb or a 10 |jg/g increase in bone Pb. If the Pb biomarker is log-transformed, effect estimates are standardized to
the specified unit increase for the 10th-90th percentile interval of the biomarker level. Effect estimates are assumed to be linear within the evaluated interval. Categorical effect
estimates are not standardized.

tStudies published since the 2013 Integrated Science Assessment for Lead.

Figure 3-4 Associations between blood Pb levels and full-scale intelligence quotient in children.

External Review Draft

3-31

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

Several recent prospective epidemiologic studies were also conducted that examined the
associations of prenatal or postnatal BLLs and effect measure modification by sex. Taylor et al. (2017)
used data from the Avon Longitudinal Study of Parents and Children (ALSPAC) to study the association
of maternal BLL and child IQ at ages 4 and 8 (i.e., Wechsler Preschool and Primary Scales of Intelligence
[WPPSI] and Wechsler Intelligence Scale for Children [WISC]-III, respectively). Little evidence of an
association with total IQ or verbal and performance components was observed. The change in raw score
on the WPPSI associated with maternal BLL was -0.32 (95% CI: -1.32. 0.68) per (ig/dL and the change
in score on the WISC-III was 0.26 (95% CI: -0.21, 0.73) per (ig/dL. Sex-stratified analyses, however,
indicated an inverse association in boys (-0.29 [95% CI: -1.02, 0.44]), but not in girls, among whom
positive associations were observed (0.73 [95% CI: 0.39, 1.33]). Note that these results are not included
in Figure 3-4 because they were reported as a change in the raw score rather than the normalized score.
Models were adjusted for covariates including maternal education and indicators of SES. The mean
prenatal and postnatal BLLs were 3.67 (ig/dL and 4.22 (ig/dL, respectively.

Tatsuta et al. (2020) examined the association of both cord BLL and postnatal BLL (at age 12)
with IQ (WISC-IV) among boys and girls enrolled in the Tohoku Study of Child Development, a
prospective birth cohort. In addition, the Boston Naming Test (BNT) was administered to assess language
abilities. This study found decrements in FSIQ score in association with postnatal BLL [[3 = -9.88 (95%
CI: -18.98, -0.78] among boys and a less precise association with IQ decrement that included the null
value among girls [[3 = -4.41 (95% CI: -15.94, 7.13]). Confounders considered in the analysis included
maternal IQ, parental SES (i.e., income) and Hg concentration in cord blood. Prenatal BLL was
associated with a relatively weak and imprecise decrease in FSIQ score among boys [[3 = -3.68 (95% CI:
-10.71, 3.35], The association of prenatal BLL with FSIQ was slightly positive but with the CI including
the null value among girls ([3 = 1.46 (95% CI: (-2.91, 5.83)]. Lower BNT scores (with cues) were
associated with both prenatal and postnatal BLL among boys. The associations of pre- and postnatal
BLLs with BNT were relatively weak or null in girls. The median postnatal BLL was 0.7 (ig/dL and the
median cord BLL was 0.8 (ig/dL in this study.

Desrochers-Couture etal. (2018) studied the association between cord, maternal, and childhood
(3-4 years old) BLLs with cognitive function (WPPSI-III at 3-4 years of age) among Inuit children
enrolled in the Maternal-Infant Research on Environmental Chemicals (MIREC) Study. Outcomes
included FSIQ, verbal IQ, performance IQ, and a general language composite. No associations were
observed between cord or childhood concurrent BLLs and FSIQ ([3 = -0.12 (95% CI: -0.25, 0.01)] and
[3 = 0.03 [95% CI: -0.14, 0.19]). The cord blood model adjusted for child age, sex, maternal education,
evaluation site, and cord blood Hg, while the postnatal model adjusted for child age, sex, evaluation site,
marital status, income, HOME score, Parenting Stress Index, and cord BLL. No pattern of Pb-associated
cognitive function decrements emerged with the verbal or performance components of IQ or with the
general language composite. An association was observed between cord BLL and performance IQ in boys
([3 = -5.69 [95% CI: -9.97, -1.41] but not in girls ([3 = 0.29 [95% CI: -3.79, 4.36]). The geometric mean

External Review Draft

3-32

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

concurrent blood Pb concentration was 0.70 (ig/dL, and the geometric mean cord blood Pb concentration
was 0.76 (ig/dL.

Zhou et al. (2020b) conducted a study to examine the association of cord blood concentrations of
trace elements, including Pb, manganese (Mn), and cadmium (Cd), with FSIQ, verbal IQ, and
performance IQ components among children from an agricultural region in China who were enrolled in a
prospective birth cohort. In models including each of these elements, cord BLL was not associated with
FSIQ ((3 = 0.67 [95% CI: -0.51, 1.85]). A similar lack of association was observed with performance and
verbal IQ and in sex-stratified analyses. Models were adjusted for covariates including maternal education
and family income. Each of the trace elements were included in the models but interactions between the
elements were not examined. The mean cord BLL among the children was 1.59 (ig/dL. Liu et al. (2015)
developed a predictive model to examine the association of Pb, Cd, and Hg in serum with FSIQ at age 5,
dropping variables based on the variance inflation factor (VIF >10). The final model for FSIQ (i.e.,
WPPSI) did not include cord serum Pb level; thus, no results pertaining to the association of Pb
concentration in serum with FSIQ were presented. Wang et al. (2022) investigated associations between
cord and concurrent venous blood concentrations of Pb, selenium (Se), As, Cu, Mn, and Cr and FSIQ
among children (6-8 years old) born in a hospital in Wujiang, Jiangsu Province. The geometric mean
concentrations of cord and venous blood Pb were 2.83 (ig/dL and 3.30 (ig/dL, respectively. Cord blood Pb
was weakly associated with reduced performance IQ ((3 = -0.11 [95% CI: -0.25, 0.03]) in boys, and
concurrent venous blood Pb was associated with reduced verbal IQ in girls ([3 = -0.49 [95% CI: -0.86,
-0.12]).

Lee et al. (2021) studied the association of Pb and other metals (i.e., Cd, Hg, and Mn) among
mother and infant pairs from eight hospitals in South Korea. In multivariable models including each of
the metals as well as covariates, imprecise negative associations of prenatal Pb exposure ([3 = -1.20 [95%
CI: -4.87, 2.01]), child BLL at age 4 ([3 = -1.83 [95% CI: -4.66, 1.01]), and child BLL at age 6
([3 = -2.61 [95% CI: -5.62, 0.40]) with FSIQ were observed. In another study of exposure to multiple
trace metals conducted in Wujiang, China, imprecise associations of maternal cord and early childhood
venous blood were observed with FSIQ (e.g., -4.77 [95% CI: -14.34, 4.79] comparing the upper quartile
of venous child BLL with the reference quartile) (Wang et al.. 2022). The geometric mean blood Pb
concentration was 2.30 (ig/dL (interquartile ratio [IQR]: 1.83-3.30 (.ig/dL). and the study included 113
children.

The recent body of evidence also includes a which evaluated the association of BLL in early
childhood (3-6 years) and BLL later in childhood (10-13 years) when IQ was also measured. The BLLs
that were measured later in childhood corresponded to the period when a major source of Pb exposure
was eliminated (i.e., following the closure of a Pb storage facility) (Iglesias et al.. 2011). The early
childhood mean BLL was 10.8 (ig/dL, and concurrent mean BLL was 3.5 (ig/dL. This study found an
FSIQ decrement associated with concurrent BLL ([3 = -0.94 [95% CI: (-1.77, -0.11)]) and a weaker, less
precise association with early childhood BLL ([3 = -0.14 [95% CI: -0.45, 0.16]). Verbal IQ was more

External Review Draft

3-33

DRAFT: Do not cite or quote

-------
1

strongly associated with concurrent BLL than performance IQ. These associations were adjusted for
important confounders including maternal IQ and education, HOME score, and SES; however,
participation was moderately low with approximately 43 percent of the children with early childhood
BLLs participating in the IQ assessment.

Braun et al. (2018) conducted a study to determine whether residential exposure interventions
would reduce BLL in children and further result in improvements in the IQ score assessed using the
WPPSI at 5 to 8 years of age. Eligible women were randomly assigned to either a Pb exposure reduction
or injury prevention group. The geometric mean BLLs for children from 1 to 8 years of age was
1.6 (ig/dL in the Pb exposure intervention group and 1.7 (ig/dL in the control group. Dust Pb loadings
were lower following the intervention, but no differences in BLL (i.e., risk of having blood Pb
concentration >2.5 or 5 (ig/dL) were observed among the children. The effect of the intervention on BLL
differed depending on race/ethnicity, however. Specifically, the relative risk (RR) of having an elevated
BLL (>2.5 (ig/dL) indicated a protective effect of the intervention among non-Hispanic black children
who received the Pb intervention (RR: 0.6 [95% CI, 0.4-1.0]), but not among non-Hispanic white
children who received the Pb intervention (RR: 1.0; 95% CI, 0.5-1.9; race/ethnicity x intervention p-
value = 0.06). No improvement in FSIQ was observed among children who received the Pb intervention,
nor did race/ethnicity modify the effect of the intervention on FSIQ.

Several cross-sectional analyses were also conducted. Dantzer et al. (2020) analyzed data drawn
from the Cincinnati Childhood Allergy and Air Pollution Study (CCAAPS), a longitudinal study that
followed children beginning at age 1 and included their caregivers. Children were assessed using the
WISC-IV at their age-12 study visit. BLL, toenail Pb concentration, and information on covariates were
also ascertained at the age-12 visit. A strong but imprecise association between BLL at age 12,
concurrently ascertained IQ -10.87 [95% CI: -16.89, -4.85]), and a relatively smaller association with
toenail Pb concentration (-1.70 [95% CI: -4.27, -0.86]) were observed after adjustment for caregiver IQ,
SES, BMI (sex was considered as a potential confounder). Toenail Pb concentration reflects blood Pb
concentration approximately months to a year before concurrent BLL due to the time it takes toenails to
grow. The concurrent BLL in this study was 0.57 (ig/dL.

Martin et al. (2021) found interactions between blood Pb and blood Mn level with IQ decrement
among children (n=57-62 depending on the analysis) enrolled in the Communities Actively Researching
Exposure Study (CARES) cohort in East Liverpool, Ohio. BLL was measured and FSIQ ascertained at
the first clinic visit, which occurred when the child was between 7 and 9 years of age. Stronger
associations between BLL and FSIQ were observed with increasing Mn concentrations in hair and
toenails. For example, the association of blood Pb with FSIQ ranged from 1.69 (95% CI: -3.04, 6.41)
when In hair Mn equaled 5 ng/g to -10.60 (95% CI: -17.17, -4.02) when In hair Mn equaled 7 ng/g. In
contrast, relatively imprecise associations between BLL and FSIQ at varying levels of blood Mn were
observed. The mean BLL among children in this study was 1.13 (ig/dL (range: 0.30-6.64). Havnes et al.
(2015) examined the association of BLL and FSIQ among the same cohort of children. The primary

External Review Draft

3-34

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

objective of this study was to examine the effect of Mn on child intelligence. Associations between BLL
and FSIQ were not reported, although a 1 (ig/dL increase in blood Pb was associated with lower
processing speed ((3 = -3.53 [95% CI: -6.95, -0.12]).

Ruebner et al. (2019) evaluated the association between BLLs and FSIQ among children with
chronic kidney disease (CKD). FSIQ was assessed using several instruments depending on the child's age
(i.e., Mullen Scales of Early Learning [age 12-29 months], WPPSI [30 months-5 years], and Wechsler
Abbreviated Scale of Intelligence [WASI; 6-18 years]). Concurrent BLL assessment was associated with
FSIQ decrement ([3 = -2.1 [95% CI: -3.9, -0.2]). Covariates considered as potential confounders
included race, poverty, maternal education, and factors related to CKD (i.e., CKD stage, duration,
glomerular versus non-glomerular diagnosis, hypertension, proteinuria, and anemia). The median BLL in
this study was 1.2 (ig/dL.

Hong et al. (2015) found an association between blood Pb concentration and lower FSIQ (-2.12
[95% CI: -3.79, -0.45] in a cross-sectional analysis of Korean school children from 8 to 11 years old.
This association persisted in models adjusted for paternal education and income, ADHD rating scale
score, Mn, and Hg (-1.95 [95% CI: -3.61, -0.29] per 10-fold increase). The mean BLL in this study was
1.80 (ig/dL.

Menezes-Filho etal. (2018) examined the association of concurrent BLL with intelligence
(WASI) among children from 7 to 12 years old. Mn in hair and toenails was also measured and the
interaction between metals evaluated. Child IQ was associated with BLL in this study (-2.78 [95% CI:
-4.66, -0.89]) in adjusted models. The mean BLL of children in this study was 1.64 (ig/dL. The effect of
BLL on child IQ was greater among children with higher toenail Mn concentrations.

Lucchini et al. (2012) conducted a cross-sectional analysis of children between the ages of 11 and
14 to examine the relationship between concurrent BLL and FSIQ as well as the potential interactions
with Mn and the aminolevulinic acid dehydratase (ALAD) genotype. A decrement in FSIQ score was
observed in association with BLL after adjustment for covariates including SES and maternal education
([3 = -2.24 [95% CI: -4.10, -0.37]). No interaction with Mn or ALAD was found. The mean BLL was
1.71 (ig/dL in this study.

3.5.1.1.1 Summary

A large number of studies evaluated in the 2013 Pb ISA found a consistent pattern of associations
between higher BLL and lower FSIQ in children aged 4-17 years (U.S. EPA. 2013a'). Multiple recent
longitudinal studies add to the evidence informing the relationship between BLL and IQ in children.
Heterogeneity in the magnitude and direction of the associations was present across studies. More
specifically, associations were observed in boys but not in girls in several studies (Tatsuta et al.. 2020;
Desrochers-Couture et al.. 2018; Taylor et al.. 2017). There was some indication that the heterogeneity

External Review Draft

3-35

DRAFT: Do not cite or quote

-------
1

across studies could be explained by modeling choices such as confounder adjustment for other metals.
For example, cross-sectional analyses found evidence that exposure to Mn may modify the association
between Pb exposure and IQ in some populations (Martin et al.. 2021; Menezes-Filho et al.. 2018).
However, studies that adjusted for multiple metals (e.g., Mn, Hg, Cd, and Pb) in regression models,
without examining the interaction between metals, found little evidence of an association between cord or
postnatal BLL and IQ (Zhou et al.. 2020b; Liu et al.. 2015). imprecise associations only in boys (Tatsuta
et al.. 2020; Desrochers-Couture et al.. 2018). or large IQ decrements after adjustment for Mn, Hg, and
ADHD rating score (Hong et al.. 2015). Overall, recent studies generally corroborated the epidemiologic
observations of associations between Pb exposure and IQ in children with relatively low blood Pb
concentrations (<5 (ig/dL) among some populations of children (see Figure 3-4 and Evidence Inventory
Table 3-2E). Consistent with findings from the 2013 Pb ISA, individual studies continue to report
associations of FSIQ with prenatal BLL (maternal and cord blood Pb) and postnatal BLLs measured at
various childhood lifestages. The heterogeneity in the observations across studies did not weaken the
larger body of evidence supporting the association of Pb exposure with cognitive effects in children at
BLLs <5 (ig/dL.

3.5.1.2 Infant Development

The BSID is an assessment instrument that was developed to identify children with
developmental delays. The early versions (e.g., (Bavlev. 1969)) have been expanded and refined, with
subsequent versions incorporating three domains of development (i.e., cognitive, language, and motor),
and parent-reported subtests that reflect social, emotional, and adaptive behaviors (Albers and Grieve.
2007). The current version of the BSID, the BSID-IV, retains the same number of domains but includes
fewer questions within each domain and requires less time to complete (Balasundaram and Avulakunta.
2021).

This section focuses on the Mental Development Index (MDI) and the cognitive and language
scales of later versions of the BSID. The MDI and cognitive/language scales are reliable indicators of the
current development and cognitive function of infants, integrating cognitive skills such as sensory and
perceptual acuities, discriminations, and response; acquisition of object constancy; memory learning and
problem-solving; vocalization and beginning of verbal communication; and basis of abstract thinking
(McCall et al.. 1972). However, the MDI test is not an intelligence test, and MDI scores, particularly
before ages 2-3 years, are not necessarily strongly correlated with later measurements of FSIQ in children
with normal development (U.S. EPA. 2013a).

In the review of the MDI evidence in the 2013 Pb ISA, emphasis was placed on results from
examinations at ages 2-3 years, which incorporate test items more similar to those in school-age IQ tests.
Most of the prospective studies reviewed in previous ISAs (U.S. EPA. 2013a. 2006b) found associations
of higher prenatal (cord and maternal BLL), earlier infancy, and concurrent BLL with lower MDI scores

External Review Draft

3-36

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

in children aged 2 to 3 years (see Table 4-4 of the 2013 Pb ISA). These blood Pb-associated decrements
in MDI were observed in populations with mean BLLs of 1.3 to 7.1 (ig/dL. Studies typically recruited
participants before or at birth without consideration of Pb exposure or maternal IQ and reported high to
moderate follow-up participation as well as nondifferential loss-to-follow-up. Most studies adjusted for
birth outcomes, maternal IQ, and education. Cord BLLs were associated with MDI, with additional
adjustment for SES and HOME score in the Boston cohort (Bellinger et al.. 1987) and for HOME score in
the Yugoslavia cohort ("Wasserman et al.. 1992). Some studies found a stronger association of MDI with
prenatal BLLs than child postnatal BLLs ((Hu et al.. 2006; Gomaa et al.. 2002; Bellinger et al.. 1987).

Among the studies assessed in the 2013 Pb ISA, several included children with mean BLLs less
than 5 (ig/dL (Hcnn et al.. 2012; Jedrvchowski et al.. 2009b; Hu et al.. 2006; Bellinger et al.. 1987).
Recent longitudinal epidemiologic studies of populations or including groups with maternal, cord, or
postnatal mean BLLs less than 5 (ig/dL add to the overall body of evidence (see Section 3.7, Table 3-3E).
These studies are presented in Figure 3-5.

External Review Draft

3-37

DRAFT: Do not cite or quote

-------
Study

Location

Blood Pb

Mean

Age at Outcome

Strata

Prospective Studies





(pg/dL)

(months)



Bellinger et al. 1987

Boston, MA

Prenatal (cord)

6.5

24

Pb6-7vs. *10 pg/dL (refmean: 14.6)*





Prenatal (cord)

1.8

24

Pb <3 vs. *10 pg/dL (refmean: 14.6) *

Jedrychowski et al. 2009

Krakow, Poland

Prenatal (cord)

1.23 (med)

24







Prenatal (cord)

1.23 (med)

36



Hu et al. 2006

Mexico City, Mexico

Prenatal (T1)

7.07

24







Prenatal (T3)

6.86

24







Prenatal (cord)

6.2

24



TKim et al. 2013

3 Cities, S Korea

Prenatal (early pregnancy)

1.4 (GM)

6







Prenatal (late pregnancy)

1.3 (GM)

6







Prenatal (early pregnancy)

1.4 (GM)

6

Cd <1.47 pg/L





Prenatal (late pregnancy)

1.3 (GM)

6

Cd >1.47 pg/L





Prenatal (early pregnancy)

1.4 (GM)

6

Cd<1.51 pg/L





Prenatal (late pregnancy)

1.3 (GM)

6

Cd >1.51 pg/L *

fValeri etal.2017

2 Districts, Bangladesh

Prenatal (cord)

1.8

20-40

Pabna





Prenatal (cord)

6

20-40

Sirajdikhan

Hu et al. 2006

Mexico City, Mexico

Concurrent

4.79

24



Claus Henn et al. 2012

Mexico City, Mexico

Concurrent (12 months)

5.1

12-36







Concurrent (24 months)

5

12-36







Concurrent (12 months)

5.1

12-36

Mn <2 pg/dL





Concurrent (24 months)

5

12-36

Mn <2 pg/dL

~

*
*

-4.00

	1	1	1	

-2.00	0.00	2.00

Beta values (95% CI) per 1 ugML increase in blood Pb

Note: Effect estimates are standardized to a 1 |jg/dL increase in blood Pb or a 10 |jg/g increase in bone Pb. If the Pb biomarker is log-transformed, effect estimates are standardized to
the specified unit increase for the 10th -90th percentile interval of the biomarker level. Effect estimates are assumed to be linear within the evaluated interval. Categorical effect
estimates are not standardized.

fStudies published since the 2013 Integrated Science Assessment for Lead.

Figure 3-5 Associations between biomarkers of Pb exposure and Bayley Score of Infant Development
Mental Development Index

External Review Draft

3-38

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

Several studies were conducted using data from Mexico City birth cohorts that enrolled low and
middle-income women seeking prenatal care at maternity hospitals belonging to the Mexican Institute of
Social Security (Y Ortiz et al.. 2017; Henn et al.. 2012; Hu et al.. 2006). Hu et al. (2006) and Sanchez et
al. (2011) were designed to elucidate the time window during pregnancy when the effect of Pb exposure
on neurodevelopment is most pronounced and are discussed in Section 3.5.1.6.3 and included in Table 3-
6E, which includes studies with central tendency BLLs >5 (ig/dL. Y Ortiz et al. (2017) examined the
modification of the Pb-neurodevelopment association by prenatal stress using the Crisis in Family
Systems-Revised (CRISYS-R) questionnaire, which assesses negative life events across several domains
(i.e., financial, legal, career, relationships, community and home violence, medical problems, other home
issues, discrimination or prejudice, and difficulty with authority). Using structural equation models, this
study found that 3rd trimester maternal BLL ((3 = -6.60 [95% CI: -13.49, 0.29] per unit of log-
transformed BLL) and the quadratic term for stress ((3 = -0.23 [95% CI: -0.45, -0.01] per unit of log-
transformed BLL) were associated with lower scores on the cognitive component of the BSID. A weak
more than multiplicative interaction between 3rd trimester maternal BLL and stress was also observed
([3 = 1.02 [95% CI: -0.78, 2.82]). Approximately 67% of the mother-infant pairs had complete
information for covariates, which included maternal education, IQ, and HOME score. Henn et al. (2012)
studied the interaction between postnatal blood Mn and Pb levels (age 12 and 24 months) and MDI score
at five different time points between 12 and 36 months of age among the Mexico City mother-infant pairs.
The coefficients for the association between BLL at 12 and 24 months with MDI score were -0.07 (95%
CI: -0.39, 0.25) and-0.08 (95% CI: -0.46, 0.30), respectively. Interactions between the highest quintile of
Mn and continuous BLL at 12 months were observed ([3 = -1.27 [95% CI: -2.18, -0.37]). The model was
adjusted for covariates including hemoglobin, maternal IQ, and maternal education.

Kim et al. (2013b. 2013c) studied the combined effect of prenatal exposure to Pb and Cd on
infant cognitive development at 6 months of age among participants in the Mothers' and Children's
Environmental Health (MOCEH) study, which enrolled infant-mother pairs from maternity clinics in
three Korean cities. Higher maternal BLL in late pregnancy was associated with lower MDI scores
([3 = -1.74 [95% CI: -3.37, -0.12]), while maternal BLL in early pregnancy was not (P = 0.02 [95% CI:
-1.20, 1.24] per (.ig/dL). This association was found after adjustment for Cd and other covariates
including maternal education and SES. A larger decrement in MDI was associated with late pregnancy
maternal BLL among those with Cd levels above the median (P = -3.20 [95% CI: -5.35, -1.06])
compared with the decrement observed among those with Cd levels below the median (P = -0.29 [95%
CI: -2.88, 2.30]). Further, an increase in MDI was observed in association with early pregnancy maternal
BLL among those with Cd levels below the median (P = 2.44 [95% CI: 0.04, 4.83]), indicating a pattern
of interaction between Pb and Cd exposure that may be dependent on the stage of pregnancy. In another
study of mother-infant pairs in Korea, Kim et al. (2018b) evaluated the associations between MDI and
various chemicals and metals, including Pb, in perinatal maternal whole blood and umbilical cord blood.
The median maternal and cord blood Pb concentrations were 2.7 (ig/dL and 1.2 (ig/dL, respectively.
Associations of blood Pb concentrations and MDI were assessed but not reported because they lacked
statistical significance.

External Review Draft

3-39

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

Valeri et al. (2017) examined the combined effect of cord blood concentrations of Pb, arsenic
(As), and Mn with cognitive and languages scores on the BSID. This study enrolled infant-mother pairs
from two birth cohorts in Bangladesh, which differed substantially regarding metal profiles and maternal
characteristics including maternal education. This study presents results from multiple regression modes
and also applied Bayesian Kernel Machine Regression (BKMR) in a prospective analysis that considered
covariates including maternal IQ, education, and HOME score. A weak association between increasing
cord Pb level and decreasing cognitive score was observed in the group with lower Mn and As
concentrations in cord blood ((3 = -0.01 [95% CI: -0.02, 0.00]) but not in the group with higher
concentrations of these metals ((3 = 0.01 [95% CI: -0.05, 0.07]).

Koshv et al. (2020) analyzed data from a birth cohort following children living in a slum in
Vellore, India. Blood Pb concentration at 15 and 24 months was averaged to determine the association
with raw cognition score on the BSID at age 2 ([3 = -0.2 [95% CI: -0.2, -0.03]). These results were
adjusted for covariates including SES, maternal IQ, and iron level. In another study, Shekhawat et al.
(2021) obtained cord blood Pb data and BSID-III scores at 6.5 months on average in a prospective cohort
study of mother-child pairs in western Rajasthan, India. The linear regression models showed no
significant associations of Pb levels and cognitive or language scores.

Paraiuli et al. (2015a) and Paraiuli et al. (2015b) assessed the association of cord BLLs with MDI
at 24 and 36 months of age, respectively, in a birth cohort of mother-child pairs recruited from a general
hospital in Bharatpur, Nepal. The median blood Pb concentration was 2.06 (ig/dL. Adjusting for in utero
Pb, As, and zinc (Zn) levels, HOME score, mother's age, parity, mother's education level, family income,
mother's body mass index (BMI) just before delivery, weight of the infant at birth and 24 months after
birth, gestational age, and infant age at the time of BSID-II assessment, no association was observed
between cord blood Pb and 24-month MDI ([3 = -4.21 [95% CI: -13.62, 5.20] per log-transformed BLL)
or 36-month MDI ([3 = 4.05 [95% CI: -3.21, 11.31] per log-transformed BLL).

Several recent studies assessed neurodevelopment using other validated instruments (Nozadi et
al.. 2021; Nvanza etal.. 2021; Zhou et al.. 2017; Vigeh et al.. 2014; Lin et al.. 2013). Zhou et al. (2017)
assessed 139 mother-child pairs from the Shanghai Stress Birth Cohort. Maternal whole blood and
maternal prenatal stress levels were assessed at 28-36 weeks of gestation, and the Gesell Developmental
Schedules (GDS) adapted for a Chinese population were administered to children at 24-36 months of age
in the study. This instrument measures development quotients (DQs) in five domains (gross motor, fine
motor, adaptive behavior, language, and social behavior) and has been validated for children 0-84 months
old. For this section on neurodevelopment, only the language domain is relevant. The Symptom
Checklist-90-Revised was used to produce a Global Severity Index (GSI) for evaluating overall maternal
emotional stress. After controlling for child sex, age, maternal age, gestational week, birth weight,
maternal education, and family monthly income, there was no association between prenatal maternal BLL
and child cognitive development. However, the authors observed interaction effects such that high
maternal stress appeared to exacerbate the effect of prenatal Pb exposure in several domains, including

External Review Draft

3-40

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

language ((3 = -33.82 [95% CI: -60.04, -7.59] per log-10 transformed unit of BLL), while low maternal
stress did not ((3 = -1.76 [95% CI: -13.03, 9.51] per log-10 transformed unit of BLL).

Vigeh et al. (2014) evaluated 174 children in Tehran, Iran up to 36 months postpartum in eight
developmental areas (social, self-help, gross motor, fine motor, expressive language, language
comprehension, letters, and numbers) using Harold Ireton's Early Child Development Inventory (ECDI).
Items for these areas were combined to generate a general development ECDI score, with higher scores
representing better development. This parent-reported measure is meant for use with children 15 months
to 6 years old and includes 60 age-discriminating items from the Minnesota Child Development
Inventory. To assess Pb exposure, three maternal whole blood samples and one umbilical cord blood
sample were collected from each mother-child pair in the first, second, and third trimesters and at
delivery, respectively. The authors observed increased odds (odds ratio [OR] = 1.74 [95% CI: 1.18, 2.57])
of a low ECDI score (<20% lower than expected for the children's age and sex) in the first trimester
(BLL=4.15 (ig/dL), adjusting for hematocrit, maternal education, BMI, family income, gestational age,
birth weight, and first born.

Lin et al. (2013) measured Pb and other metals (i.e., Mn, As, and Hg) in cord blood samples from
230 mother-infant pairs from the Taiwan Birth Panel Study (TBPS) and assessed development in
cognition, language, motor, social, and self-care skills among 2-year-old children with the Comprehensive
Developmental Inventory for Infants and Toddlers (CDIIT), which has been standardized for children 3 to
71 months old. The CDIIT uses DQs, and a score of 100 represents normal development. After adjusting
for maternal age, maternal education, fish intake >2 times/week during pregnancy, infant gender,
environmental tobacco smoke during pregnancy and after delivery, and HOME Inventory score, the linear
regression models showed that highly Pb-exposed (>75th percentile: 1.65 (ig/dL) children had lower
cognitive DQs ([3 = -5.35 [95% CI: -9.64, -1.06]) compared with those in the low-exposure (<75th
percentile) group. The authors also observed an interaction with Mn such that children who were highly
exposed to both Mn and Pb had larger deficits in cognitive ([3 = -8.19 [95% CI: -14.40, -1.98]) and
language ([3 = -6.81 [95% CI: -12.16, -1.46]) DQs compared with those with low exposure to just one or
both of these metals.

Nozadi et al. (2021) collected blood samples from pregnant mothers at the 36-week visit or at
time of delivery and administered the Ages and Stages Questionnaire Inventory (ASQ:I) at 10-13 months
of age to evaluate neurodevelopment. Trained staff scored children on five 65-70 item developmental
domains: communication, gross motor, fine motor, problem-solving, and personal-social. A 1 (ig/dL
increase in prenatal blood Pb was associated with small, imprecise decreases in problem-solving
([3 = -0.67 [95% CI: -1.54, 0.20]) scores.

Nvanza et al. (2021) collected dried blood spots from a finger prick to measure Pb (in addition to
Hg, Cd, and As) concentrations in pregnant mothers at 16-27 weeks of gestation from the Mining and
Health study in Northern Tanzania. The authors used the Malawi Developmental Assessment Tool
(MDAT) translated into Kiswahili to assess several functional domains, including social development, in

External Review Draft

3-41

DRAFT: Do not cite or quote

-------
1

children between 6 and 12 months old. MDAT has been validated for children 0-6 years old in rural sub-
Saharan Africa. Covariates in the Poisson regression model included maternal age, maternal education,
maternal and parental occupation, number of under-5-year-old siblings at home, family socioeconomic
wealth quintile, infant sex, infant age, birth weight, and height and weight at the time MDAT was
administered. Concentrations of Pb were low (median: 2.72 (.ig/dL). and the German Environmental
Survey for Children reference level of 3.5 (ig/dL was used to dichotomize Pb exposure groups into low
and high exposure groups. The authors did not observe significant associations between high Pb exposure
and language impairment. However, children highly exposed to both Hg (>0.08 (ig/dL) and Pb were more
likely to have global neurodevelopmental impairment (prevalence ratio [PR =1.4 [95% CI: 0.9, 2.1]).

3.5.1.2.1 Summary

Most of the prospective studies reviewed in previous ISAs (U.S. EPA. 2013a. 2006b') found
associations of higher prenatal (cord and maternal BLL), earlier infancy, and concurrent BLL with lower
MDI score in children aged 2 to 3 years (see Table 4-4 of the 2013 Pb ISA). These blood Pb-associated
decrements in MDI were observed in populations with mean BLLs of 1.3 to 7.1 (ig/dL. Studies typically
recruited participants before or at birth without consideration of Pb exposure or maternal IQ and reported
high to moderate follow-up participation and nondifferential loss-to-follow-up. Recent studies continue to
support associations between Pb exposure (i.e., maternal (Y Ortiz et al.. 2017; Vigeh et al.. 2014; Kim et
al.. 2013b. c), cord (Valcri et al.. 2017). and postnatal exposure (Lin et al.. 2013)) and poorer performance
on tests of neurodevelopment among mothers and infants with mean BLLs <5 (ig/dL (see Figure 3-5).
Although Zhou et al. (2017) found no association overall, this study reported decrements in several
domains of the GDS among infants of mothers reporting high maternal stress. Similarly, Y Ortiz et al.
(2017) found some evidence of interaction between Pb exposure and maternal stress..Several studies
found interactions between Pb, Mn, or Mn and As (Valeri et al.. 2017; Lin etal.. 2013; Henn et al.. 2012)
or Cd exposure (Kim et al.. 2013b. c). The direction of the interaction was not consistent across studies.
Overall, recent studies support findings from the 2013 Pb ISA and extend the evidence pertaining to
modification of the association between Pb exposure and infant neurodevelopment by maternal stress and
exposure to other metals.

3.5.1.3 Learning and Memory

The 2013 Pb ISA included many studies examining the associations of blood and tooth Pb levels
with neuropsychological tests of memory, learning, and executive function. These domains of cognitive
function are related to intelligence, and several were evaluated in the subtests of FSIQ. Further, indices of
memory, learning, and executive function are comparable to endpoints examined in experimental animal
studies.

External Review Draft

3-42

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

3.5.1.3.1

Epidemiologic Studies of Learning and Memory in Children

The studies evaluated in the 2006 Pb AQCD and the 2013 Pb ISA did not clearly indicate
associations between higher BLL and poorer performance on neuropsychological tests of memory or
learning (i.e., acquisition of new information) in children 4-17 years of age (see Table 4-5 (U.S. EPA.
2013a)'). The studies used various tests (e.g., spatial span total errors on the Cambridge
Neuropsychological Test Automated Battery [CANTAB], digit span or learning factor score on the
WISC, Kaufman Assessment Battery for Children [K-ABC], memory score on the McCarthy Scale of
Children's Abilities, California Verbal Learning Test [CVLT], and working memory on the Wide Range
Assessment of Memory and Learning [WRAML]) to assess learning and memory, which may account for
some of the heterogeneity observed in the findings. Notably, evidence for both memory and learning from
prospective analyses of several established cohorts (i.e., Rochester, Boston, and Cincinnati) was mixed
(Canfield et al.. 2004; Ris et al.. 2004; Stiles and Bellinger. 1993; Bellinger et al.. 1991; Dietrich et al..
1991). Cross-sectional studies included in the previous ISA, however, generally found associations
between higher concurrent BLLs and poorer learning and memory, including the large (n = 4,853) study
of children aged 5-16 years who participated in the National Health and Nutrition Examination Survey
(NHANES) III (Lanphear et al.. 2000). Associations of higher concurrent BLL and poorer memory in
children aged 5-16 years were also observed by Krieg et al. (2010) and Froehlich et al. (2007); however,
some studies reporting such associations had limited implications because they lacked consideration for
potential confounding (Counter et al.. 2008; Min et al.. 2007). Several studies included in the 2013 Pb
ISA were conducted in populations with mean BLLs <5 |ig/dL(Kricg et al.. 2010; Surkan et al.. 2007;
Lanphear et al.. 2000) and reported associations between increasing concurrent childhood blood Pb
concentration and lower performance on tests of learning and memory.

A small number of recent studies examined the association of Pb exposure with children's
performance on neuropsychological tests of learning and memory (see Section 3.7, Table 3-4E). Several
such studies examined the association between Pb exposure and performance on tests of learning and
memory in models that adjusted for several important confounders plus co-exposure to other metals or
chemicals. Yorifuii etal. (2011) evaluated the association of cord BLL with several components of IQ at
age 7 and age 14 in a Faroese birth cohort also exposed to methyl mercury (MeHg). IQ components
including attention and working memory, language, visuospatial reasoning, and memory were assessed
using the WISC-R and the children's version of the CVLT. The association of cord BLL with
neuropsychological tests of cognition was reported without adjustment for cord Hg, with adjustment for
cord Hg, and with a term for the interaction of cord blood Pb and cord Hg concentration. Poorer
performance on the digit span components of the WISC-R, which measure short-term memory, were most
consistently observed in association with cord BLL. The results for associations with performance on
some of the tests indicated that the interaction between Pb and methyl Hg (MeHg) may be less than
additive (i.e., the associations of cord blood with the neuropsychological test outcomes were most
discernable among children with hair Hg concentrations below 2.61 jxg/g and among those with the
lowest cord Hg concentrations (e.g., (3 = -0.27 (-0.42, -0.11) at age 14).

External Review Draft

3-43

DRAFT: Do not cite or quote

-------
1

Another recent study by Tatsuta et al. (2014) also examined exposure to multiple chemicals
including Pb, PCBs, and MeHg. The outcome in this study was performance on the K-ABC at 42 months
of age. No associations with sequential processing speed score (-2.14 [95% CI: -12.80, 8.53]) or mental
processing score (-3.32 [95% CI: -12.41, 5.77]) were observed after adjustment for variables including
other chemicals, maternal IQ, and family income (associations per unit of log [base not reported]
transformed BLL). Similarly, Oppenheimer et al. (2022) examined the association of cord BLL with
working memory assessed using the WRAML among children (13-17 years old) living near a superfund
site and thus exposed to multiple metals. Regression models were adjusted for prenatal concentrations of
dichlorodiphenyldichloroethylene (DDE), hexachlorobenzene (HCB), PCBs, Pb, and Mn as well as other
important confounders including HOME score and maternal IQ. The associations between verbal working
memory, symbolic working memory and working memory index differences were 0.12 (95% CI: -0.20,
0.45), 0.09 (95% CI: -0.25, 0.42), and 0.59 (95% CI: -0.97, 2.15) respectively. The interaction of Pb
exposure and sex was examined but no statistical evidence of the interaction was observed.

Summary

The studies evaluated in the 2006 Pb AQCD and the 2013 Pb ISA did not clearly indicate
associations between higher BLL and poorer performance on neuropsychological tests of memory or
learning (U.S. EPA. 2013a'). A small number of recent studies of children with mean BLLs <5 (ig/dL add
to the evidence informing the association of Pb exposure with performance on tests of memory and
learning; however, the results from these recent studies do not enhance the consistency of the evidence as
a whole. Some of the available studies consider co-exposure to other chemicals and metals as confounders
(Oppenheimer et al.. 2022; Tatsuta et al.. 2014) although there is evidence that such co-exposures may
interact with or modify the association between Pb and the outcome (Yorifuii etal.. 2011). The evidence
regarding the effect of Pb exposure on specific tests of learning and memory lacks consistency, overall.

3.5.1.3.2 Experimental Animal Studies of Learning and Memory

As described in the preceding sections, BLLs are consistently associated with decrements in FSIQ
in children but show variable associations with performance on tests of learning and memory. A
relationship between Pb exposure and cognitive function deficits is further supported by evidence for Pb-
induced impairments in memory and learning in animal models. Critical evidence for the association of
Pb with cognitive impairment comes from a series of studies describing the effects of lifetime Pb
exposure on nonhuman primates (Rice. 1992; Rice and Gilbert. 1990a; Rice. 1990; Rice and Karpinski.
1988). Cynomolgus monkeys (Macaca fascicularis) were dosed continuously from birth and tested
repeatedly throughout their lifetime. While these exposures yielded BLLs beyond values considered
relevant for the current assessment (>30 (.ig/dL). they provide key evidence of Pb-induced cognitive
impairments in a translationally relevant species.

External Review Draft

3-44

DRAFT: Do not cite or quote

-------
1

Learning and Memory - Morris Water Maze

In rodents, spatial learning and memory have been evaluated using several paradigms, including
the Morris water maze. Typically, the Morris water maze task is separated into two distinct phases.

During the training phase, spatial learning is assessed by measuring the time or distance required for a
rodent to swim to a submerged platform using visual cues beyond the maze (e.g., basic shapes). Slower
decreases in time to escape from the maze across training trials (i.e., escape latency) can be indicative of
impaired spatial learning. After the animals have learned the location of the hidden platform (confirmed
by steadily decreasing escape latencies across training trials), memory for the location of the platform is
assessed in the probe phase by removing the platform and measuring the time each animal spends in that
area of the maze. Decreased time spent or distance swam in the target zone can be indicative of a deficit
in spatial memory.

The 2013 Pb ISA (U.S. EPA. 2013a) reviewed the evidence suggesting that exposure to Pb
produced learning and memory impairments in laboratory rodents using the Morris water maze. Several
studies involved exposures to Pb of varying durations and across different developmental periods.
Significant impairments in both learning and memory were reported for developmental exposures
resulting in BLLs ranging from 23 to 70 (ig/dL. For example, Kuhlmann et al. (1997) compared the
effects of Pb exposure during various lifestages and reported impaired learning and long-term memory in
adult Long-Evans rats exposed during gestation and lactation (via maternal diet) or over a lifetime from
gestation through adulthood. Each of the exposure periods examined produced peak BLLs of 59 (ig/dL.
Exposure during adolescence only, which produced BLLs of 23 (ig/dL, did not affect memory. In contrast
with Kuhlmann, other studies reviewed in the previous ISA reported that postweaning Pb exposure
(8 weeks via drinking water) in Sprague Dawley rats resulted in significant deficits in both learning and
memory using the Morris water maze (Fan et al.. 2010; Fan et al.. 2009). Recent studies (seeevidence
inventory Table 3-4T) provide consistent evidence for Pb-induced impairments in learning and memory
following developmental exposures with lower BLLs than covered in the previous ISA (<30 (ig/dL).

Evidence reviewed in the previous ISA indicated that development (i.e., preconception, during
gestation, lactation) may be a critical window for Pb exposure to cause cognitive dysfunction later in life.
Several recent studies examined the effects of long-term Pb exposure that began during development and
continued into adulthood. In the study with the longest exposure duration that was relevant to this ISA,
Ouvang et al. (2019) developmentally exposed Sprague Dawley rats to Pb (0.05% Pb acetate in maternal
drinking water) beginning on GD 0. After weaning, animals were maintained on drinking water
containing (0.01% Pb acetate) until PND 679 (701 total days of exposure). This exposure resulted in a
final mean BLL of 22 (ig/dL. When assessed immediately following the end of exposure, exposed
animals displayed both impaired learning and memory in the Morris water maze task (31% fewer
crossings in the target zone during the probe trial compared with controls). Zhu et al. (2019b) exposed
rats to Pb (0.5 g/L Pb acetate in maternal drinking water) for 387 days beginning at conception, which
resulted in a final mean BLL of 29 (ig/dL. In the Morris water maze, exposed animals showed

External Review Draft

3-45

DRAFT: Do not cite or quote

-------
1

significantly increased escape latencies on the last day of training only, suggestive of slightly impaired
spatial learning. In the probe trial, exposed animals made 45% fewer crossings into the target zone than
controls, strongly suggestive of impaired spatial memory. In another long-term study, Zhou et al. (2020a)
developmentally exposed Sprague Dawley rats to Pb through maternal drinking water beginning at
conception and continuing through lactation. After weaning, animals were maintained on Pb in drinking
water (386 days). While this study examined several doses of Pb, only the lowest dose (0.5 g/L in water)
produced BLLs that were relevant to this ISA. Learning and memory were tested via the Morris water
maze during exposure at PND 21 (mean BLL 10 (ig/dL) and later immediately following the end of
exposure at PND 364 (mean BLL 15 (.ig/dL). At both time points, Pb-exposed animals took significantly
more time to escape the maze during training and spent less time in the target zone during the probe trial,
indicative of impaired learning and memory. The effect of Pb was slightly more pronounced at the earlier
timepoint (number of crossings in target zone was 36% lower than that of controls at PND 21, compared
with a 26% difference at PND 364), which may be due to improvement on the task with age.

In a study by Tartaglione et al. (2020). male and female Wistar rats were developmentally and
lactationally exposed to Pb beginning 4 weeks prior to conception (GD -28) to PND 23 (50 mg/L in
maternal drinking water), resulting in a final BLL of 26 (ig/dL, and displayed increased escape latencies
compared with controls. This effect was not sex-specific. Exposed animals showed mild memory deficits
in the form of increased latency to target zone and increased distance to target zone relative to controls
(no effect observed on crossings or time spent in target zone). While effects on memory were not sex-
specific, the authors reported that Pb significantly decreased path efficiency (ratio of the shortest possible
path length to the observed path length) in females only, which may indicate a sex-specific effect of Pb on
the processes that govern spatial integration.

Xiao et al. (2014) compared the effects of Pb on two separate developmental windows in Wistar
rats: one beginning prior to conception (2 mM Pb in maternal drinking water from GD -21 to PND 21,
57 days total) and the other beginning in adolescence (2 mM Pb in drinking water from PND 21 to 84,
63 days). The gestational exposure yielded a final BLL of 10 (ig/dL at PND 21, while the adolescent
exposure led to a final BLL of 4 (ig/dL. Animals from both exposure groups were tested in the Morris
Water Maze on PND 85. Exposed animals had significantly increased escape latencies and decreased
target zone time relative to control animals, indicating cognitive dysfunction. No difference was observed
between the exposure time frames, suggesting both may be similarly vulnerable to the cognitive effects of
Pb assessed by the Morris water maze. Similarly, Barkur and Bairv (2015b) employed a study design that
examined multiple different time frames of exposure: pregestational, gestational, and combined gestation
and lactation. All but the combined gestation and lactation group yielded BLLs that were relevant to this
ISA. Pb exposure during each period of development had a significant negative effect on memory
compared with the control groups (learning data were not reported). The gestation and lactation groups
exhibited similar magnitudes of effects, with the pregestational group showing the smallest difference
compared with the control. This study suggests that Pb affects memory following developmental

External Review Draft

3-46

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

exposure and that the periods of gestation and lactation may be more sensitive than the pregestational
period alone.

Wang et al. (2021a) exposed Sprague Dawley rats to 0.05 and 0.1% Pb in drinking water from the
beginning of gestation through the end of lactation. BLLs assessed on PND 21 resulted in BLLs of 24.9
and 30.4 (ig/dL for the two dose groups, respectively. On PND 21, Pb-exposed rats displayed
significantly increased escape latencies during acquisition, indicative of impaired learning. During the
probe trial, only animals in the highest Pb concentration had significantly fewer crossings in the target
zone.Betharia and Maher (2012) exposed Sprague Dawley rats starting at conception (10 |ig/mL in
maternal drinking water, GD 0 to PND 20) and assessed cognitive function via the Morris water maze at
two points: end of exposure (PND 21, BLLs of 0.98 (ig/dL) and later (PND 56, BLLs of 0.03 (ig/dL). This
study reported the lowest BLLs for animals tested in the Morris water maze paradigm. In contrast to many
of the recent studies reviewed here, when assessed immediately following the end of exposure, no effects
on learning or memory were observed. At the later time point, exposed females displayed significantly
impaired memory relative to untreated controls. This minor discrepancy could be due to the lower dose
used in the study. It is also possible that repeated experience with the paradigm across two sessions could
"unmask" a subtle effect on learning and memory produced by low-level exposure to Pb, though data on
cognitive function in animals with BLLs <1 (ig/dL remain limited.

In Anderson et al. (2012). rats were exposed starting prior to conception and then continuing
through lactation (GD -10 to PND 21) to a range of doses and assessed for learning after the end of
exposure. Only the lowest Pb concentration (250 ppm in drinking water) yielded BLLs relevant to this
ISA, with final levels of 19 (ig/dL in males and 18 (ig/dL in females. During the training phase, Pb
exposure had no effect on escape latency in either sex; however, Pb-exposed females displayed
significantly decreased path efficiency compared with untreated controls. While no effect on spatial
memory was observed during the probe trials, exposed females once again exhibited lower path efficiency
scores compared with untreated controls, suggesting that, in females particularly, Pb may influence
pathfinding processes. This study also determined that Pb partially blunted the positive effects of an
enriched environment on spatial learning, which may be relevant when considering how environmental
factors (e.g., SES) may interact with Pb exposure in humans.

In Zhao et al. (2018). rats were developmentally and lactationally exposed to multiple doses of Pb
from GD -14 to PND 10, resulting in final BLLs of 1 (ig/dL for 0.005% Pb and 1.5 (ig/dL for 0.01% Pb
on PND 30. The highest dose group (0.02% Pb in drinking water) yielded BLLs higher than relevant for
this ISA. At the two relevant doses, Pb exposure led to significant impairments in both learning and
memory. Additional recent studies provided evidence that developmental exposure to Pb resulted in
learning and memory deficits that persisted later into adolescence and adulthood (Xiao et al.. 2020; Li et
al.. 2016a; Zhang et al.. 2014; Rahman et al.. 2012b; Zhang et al.. 2012) In one discrepant study, Wang et
al. (2021b) exposed Sprague Dawley rats to Pb in drinking water (0.05-0.2% Pb) from 4 weeks prior to
conception to PND 21. Only the lowest exposure concentration (0.05%) resulted in a mean BLL relevant

External Review Draft

3-47

DRAFT: Do not cite or quote

-------
1

to this ISA (21.1 (ig/dL at PND 21). Learning and memory were also assessed via the Morris Water Maze
on PND 21; the authors reported no significant effects of Pb during acquisition or testing in the 0.05%
exposure group, though some effects on memory were seen at higher concentrations.

One recent study investigated the effects of Pb exposure on the cognitive function of adolescent
rodents. Liu et al. (2022c) exposed 4-week-old Sprague Dawley rats to 0.2% Pb for 12 weeks, which
yielded a mean BLL of 17.3 (ig/dL. The authors reported no effect of Pb on learning during the
acquisition phase; however, during the probe trial, Pb-exposed animals exhibited significantly fewer
crossings relative to untreated controls, suggestive of memory impairment. These recent studies provide
broadly consistent evidence that Pb produces learning and memory impairments, with developmental
periods potentially representing a more sensitive window for exposure.

Learning and Memory - Novel Objection Recognition

Another commonly applied measure of long-term memory in animal models is the novel object
recognition task. Following habituation to an empty arena, animals are placed in the arena with two
identical mundane objects and allowed to explore freely. During the testing phase (-24 hours after
training), animals are returned to the arena with one object from the first day and one novel object and
allowed to explore. The time spent examining each object is recorded. Because rodents tend to explore
unfamiliar objects, these durations can be used to calculate a recognition index, which serves as a measure
of memory. Decreased recognition indices (less time spent with the novel object) suggest impaired
memory. The previous ISA did not incorporate novel object recognition data, but one recent study used
the paradigm to assess long-term memory following Pb exposure relevant to the current assessment.

Tartaglione etal. (2020) observed that long-term exposure to Pb via maternal drinking water (GD
-28 to PND 23), which yielded relatively high BLLs of -26 (ig/dL, caused a significant decrease in novel
object recognition index in females only when tested at PND 60-72. The result of this single study is
generally consistent with the pattern of memory impairment observed following developmental Pb
exposure, though evidence from the novel object recognition paradigm remains limited. Sex, exposure
timing, and behavioral history may also influence effects on long-term memory, yet the contribution of
each of these factors remains unclear.

Learning and Memory - Y Maze

Another measure of spatial memory in rodents is the Y maze, which relies on the natural
inclination of rodents to explore new areas rather than revisit previously explored areas. After the animal
in placed in the Y-shaped maze, spontaneous alterations (i.e., entries into an arm different than the most
recently visited arm) and total arm entries are recorded. Re-entries into the most recently visited arm from
the center of the maze (decreased spontaneous alteration) may indicate dysfunction in working spatial
memory. The previous ISA reviewed evidence from only one study that utilized the Y maze: (Niu et al..

External Review Draft

3-48

DRAFT: Do not cite or quote

-------
1

2009) reported that Wistar rats exposed to Pb from lactation up to 12 weeks of age displayed learning
impairments starting at 8 weeks of age. These exposures resulted in BLLs of 17 (ig/dL, which are relevant
to the current assessment.

Three recent studies utilized the Y maze to assess spatial memory following Pb exposures that
produced comparable BLLs (Table 3-4T), and the results were inconsistent. Xiao et al. (2020) reported
that female Sprague Dawley rats with long-term developmental exposure to Pb (125 ppm in drinking
water from GD -7 to PND 68) displayed a significant decrease in spontaneous alterations compared with
control females (70 versus 55%), which suggests a deficit in spatial working memory. In contrast,
Tartaglionc etal. (2020) did not observe any changes in spontaneous alterations following a shorter
exposure in male and female Wistar rats (50 mg/L in drinking water from GD -28 to PND 23).
Tartaglionc etal. (2020) did report a significant decrease in arm entries made by exposed rats, which may
indicate a Pb-induced alteration in exploratory behavior rather than an effect on memory. No sex effects
were observed in this study. Similarly, Abazvan et al. (2014) conducted a dietary exposure to Pb from
conception to adulthood (approximately 6 months), which yielded BLLs of 26 (ig/dL in males and
35 (ig/dL in females (not relevant to this ISA). The authors reported no significant effect of Pb on
alterations in the Y maze in either sex. While these recent studies were focused on assessing memory
rather than learning in the Y maze, the effects of Pb on Y maze performance and the influence of the
developmental window remain unclear. Further investigation may be needed.

Learning and Memory - Fear Conditioning

Another measure of learning and memory, fear conditioning, is a task in which animals are
trained to associate a particular conditioned stimulus (e.g., auditory tone) with an aversive unconditioned
stimulus (e.g., mild foot shock). After repeated pairings of the conditioned and unconditioned stimuli
(acquisition), animals are exposed to the conditioned stimulus and the conditioned response (e.g.,
freezing, defined as lack of non-respiratory movement) is recorded. Decreases in freezing behavior may
indicate memory deficits, as the animal is no longer associating the tone with the aversive stimulus.
Several variations on this procedure may be employed to interrogate different brain regions and
processes, such as "trace" fear conditioning, wherein an interval occurs between the tone and the aversive
stimulus. Though fear conditioning data were not incorporated in the previous ISA, four recent studies
examined the effects of Pb exposures that produced BLLs relevant to the current assessment on
associative memory.

To assess the influence of exposure window on Pb-induced cognitive impairment.Anderson et al.
(2016) exposed male and female Long-Evans rats to Pb (150, 375, and 750 ppm in chow) during three
separate exposure windows (perinatal [GD -10 to PND 21], early postnatal [PND 0 to PND 21], and
long-term postnatal [PND 0 to PND 55]), all of which resulted in BLLs <10 (ig/dL (summarized in
Table 3-4T). The authors used a "trace" fear conditioning paradigm with memory testing at 1, 2, and
10 days after conditioning. Anderson et al. (2016) reported significant effects of Pb that differed by sex,

External Review Draft

3-49

DRAFT: Do not cite or quote

-------
1

exposure window and dose. In females, learning impairments were observed only in the highest dose
group of the perinatally exposed animals. Some memory deficits were noted in the early postnatal
exposure group but only at lower doses. Interestingly, following long-term postnatal exposure, females
only displayed memory problems at the lowest doses of Pb. This result is not easily explained by
variation in BLL (i.e., the lowest dose group did not have higher BLLs than the other dose groups). In
males, minor learning deficits were noted in the early postnatal and long-term exposure groups. Memory
impairment was noted in perinatally exposed males at the lowest and highest doses only. The results of
this experiment suggest that both sex and exposure window influence the effects of Pb on learning and
memory, and that these effects may not follow the traditional dose-response relationship reported using
other paradigms.

In a subsequent study by the same group, Verma and Schneider (2017) compared the effects of
Pb on associative memory in two different rat strains using a design similar to the previous study
(Anderson et al.. 2016) to examine the influence of exposure window. There were no significant
differences in BLL between Sprague Dawley and Long-Evans rats (Table 3-4T). The authors reported no
Pb effects on acquisition across sex, indicating no effect on learning within the fear conditioning
paradigm. In Long-Evans females, animals exposed during the early postnatal period showed a marked
decrease in percent time freezing during the memory tests (day 1, control: 90% time freezing versus
treated: 68% time freezing). Consistent with the previous study from this group, the effect was more
pronounced after the initial acquisition trials (day 10, control: 70% time spent freezing versus treated:
29% time freezing). Conversely, in Long-Evans males, there was no effect in the postnatal exposure
group, yet significant impairments were detected in the perinatal exposure group starting on the 2nd day
after acquisition (day 2, control: 68% freezing versus treated: 48% freezing). Once again, the effect was
more pronounced later in the experiment (day 10, control: 75% freezing versus treated: 39% freezing).
Interestingly, no significant effect of Pb on learning or memory was observed in Sprague Dawley rats of
either sex with BLLs of approximately 5 (ig/dL.

Wang et al. (2016) exposed male Sprague Dawley rats to 100 ppm Pb in drinking water from
PND 24 to 56 and then assessed memory using a context-dependent fear conditioning paradigm in which
the rats were returned to the same test chambers without atone or shock 24 hours after acquisition, and
freezing was recorded. In this version of the test, environmental context serves as a cue that the animals
associate with the aversive stimulus. The authors reported a dramatic decrease in % time spent freezing in
treated animals during the memory test 24 hours later (64% time spent freezing in controls compared with
only 8% time spent freezing in treated animals). This long-term adolescent exposure, which produced
BLLs of 13 (ig/dL, resulted in significant memory dysfunction. The divergent results between Verma and
Schneider (2017) and Wang et al. (2016) may be explained by variations in the paradigms used and the
duration, timing of the exposures, and resulting BLLs.

Abazvan et al. (2014) conducted dietary exposure to Pb from conception to adulthood
(approximately 6 months) in mutant (double transgenic) Disrupted-in-Schizophrenia-1 (DISCI; a genetic

External Review Draft

3-50

DRAFT: Do not cite or quote

-------
1

risk factor for schizophrenia) mice, which yielded BLLs of 26 (ig/dL in males and 35 (ig/dL in females
(outside PECOS). Single transgenic mice that possessed the mouse DISCI (mDISCl) transgene but did
not express mDISCl served as controls. Using a contextual fear conditioning paradigm, these authors
reported no effect of Pb on fear extinction in mutant or control mice. These studies suggest developmental
Pb may adversely affect learning and memory within the fear conditioning paradigm, but these effects
may be sensitive to factors such as sex, strain or genetics, dose, and timing of exposure.

Learning and Memory-Avoidance

Another measure of learning and memory in animal models is the avoidance paradigm, which is a
fear-aggravated test that relies on animals learning to avoid environments where they experienced an
aversive stimulus. In the passive, "step-through" variation of the test, animals are placed in an arena with
at least two compartments, separated by gates that allow passage between compartments. During training,
animals will receive an aversive stimulus (foot shock) in the darkened chamber. The animals are later
placed in an illuminated chamber and the time that elapses before the animals enter the dark chamber is
recorded (entry latency). Shorter entry latencies are associated with impaired memory. The previous ISA
did not incorporate any passive avoidance studies. Four recent studies examined passive avoidance
behavior following Pb exposure.

Barkur and Bairv (2015b) compared the effects of Pb (0.2% in maternal drinking water) on
associative learning in male Wistar rats across several different developmental exposure periods and
durations, all but the longest of which produced BLLs <30 (ig/dL. All exposed animals, except for the
pregestation group (GD -30 to GD 1), displayed decreased entry latencies relative to controls, indicative
of impaired memory. These effects persisted out to 48 hours after the initial exploration trial. Similarly,
Barkur etal. (2011) observed that male Wistar rats exposed via maternal drinking water (0.2%) from
GD 0 to PND 21 had significantly shorter entry latencies when assessed at PND 25 and again at
PND 120. It should be noted that BLLs were >30 (ig/dL when measured on PND 25 but the levels
decreased to -0.5 (ig/dL by PND 120. Following long-term exposure to Pb via drinking water (50 ppm,
GD 0 to PND 45), Biioor et al. (2012) reported that male and female offspring displayed significantly
shorter entry latencies than their untreated counterparts.

One study utilized a "step-down" version of the test, wherein animals are placed on a platform
above a grid that delivers a mild electric shock. Over the course of training, animals should learn to
associate the grid with the aversive stimulus and avoid stepping down off the platform. During testing,
both latency (time elapsed before stepping down) and errors (number of times the animal stepped down
onto the grid) are recorded to assess memory. Following long-term developmental exposure to Pb (0.4%
in maternal drinking water from GD 0 to PND 21, BLLs of-14 (.ig/dL). Kunming mice displayed
significantly decreased step-down latency and increased errors relative to controls (Zhang et al.. 2014).
suggestive of both learning and memory impairment. These studies suggest that exposure to Pb during
development results in negative effects on associative learning and memory that may persist into

External Review Draft

3-51

DRAFT: Do not cite or quote

-------
1

adulthood and that these effects are influenced by the developmental window during which exposure
occurs.

Learning and Memory with Stress

The paradigm of combined Pb and stress exposure experienced by a laboratory animal has been
examined by the Cory-Slechta laboratory with a focus on the common pathway of an altered
hypothalamic pituitary adrenal (HPA) axis and brain neurotransmitter levels. Effects on learning varied,
depending on the timing of stress, Pb exposure concentration, and sex of the animal. Pb-stress interactions
were found with dietary Pb exposures that resulted in BLLs relevant to this ISA. The evidence
additionally indicated that associations of Pb exposure and stress with learning deficits (multiple
schedules of repeated learning and performance in females) may be related to aberrations in
corticosterone and dopamine. Several recent studies with Pb exposures relevant to the current ISA
included stress components in their experimental designs, providing further evidence that supports an
interaction between stress experience and the effects of Pb on cognitive function.

The Cory-Slechta laboratory expanded upon their previous research by investigating the
interaction between Pb and prenatal stress in males, with additional comparisons between maternal-only
and lifelong exposure (Cory-Slechta et al.. 2012). In contrast to previous reports on females, prenatal
stress with Pb exposure was reported to enhance learning accuracy with a repeated learning and
performance schedule. The authors postulated that this effect may be due to increases in the response rate,
which have been observed in both Pb and stress independently. Thus, this seemingly positive result may
reflect an increase in response rate or impulsivity.

In (Anderson et al.. 2012). researchers examined the influence of differential rearing conditions
(enriched or barren) on Morris water maze performance in Sprague Dawley rats exposed to Pb. The
authors reported a significant positive effect of the environment on learning (decreased escape latencies)
in males regardless of Pb exposure, though this effect was only present during the first two acquisition
trials. The same trend was observed in female rats, though Pb was shown to dull the advantage provided
by enrichment, with complete negation of the advantage observed at the high dose in females. While no
consistent effect was noted on time spent in the target quadrant, Pb significantly impaired the path
efficiency relative to controls in both sexes, and this effect was ameliorated by an enriched environmental
status. While these studies on the influence of Pb and stress on learning in rodents produced variable
results, they provide evidence supporting the interaction between Pb and stress and suggest that these
effects are further influenced by sex, age, and timing of exposure.

Summary

Several recent studies in laboratory animal models with exposures resulting in mean BLLs
<30 (ig/dL add to the substantial body of evidence indicating that Pb exposures can impair learning and

External Review Draft

3-52

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

memory. Compared with studies in the 2013 Pb ISA, more recent studies demonstrate that the
relationship between Pb exposures and learning and memory impairments is present at lower BLLs.
Recent studies provided evidence that early-life exposures were associated with learning and memory
impairments later in adulthood, indicating that development is a critical window for the effects of Pb on
cognitive function. Additionally, new evidence suggests that longer durations of Pb exposure (especially
those encompassing developmental windows) produced greater learning and memory impairments. The
few studies reporting weak or null effects were not stronger with respect to the design or methodology
and did not weaken the much larger body of supporting evidence.

3.5.1.4 Executive Function in Children

3.5.1.4.1 Epidemiologic Studies

Epidemiologic evidence presented in the 2006 Pb AQCD and the 2013 Pb ISA indicated a
consistent pattern of associations between higher childhood blood or tooth Pb levels and poorer
performance on tests of executive function in children and young adults (see Table 4-8 (U.S. EPA.
2013a')'). Associations were found with indices of executive function such as strategic planning, organized
search, flexibility of thought and action to a change in situation, and control of impulses assessed by
various tests including the Intra-Extra Dimensional Set Shift, Wisconsin Card Sorting Test, and Stroop
Color-Word test (SCWT). The strongest evidence was provided by prospective analyses. These analyses
included several birth cohorts in Boston and Rochester and examined BLLs that preceded the outcome
assessment, with adjustment for several potential confounding factors (Canfield et al.. 2004; Canfield et
al.. 2003b; Bellinger et al.. 1994a; Stiles and Bellinger. 1993). Moderate to high follow-up participation
that was not biased to those with higher BLLs and lower cognitive function was an additional strength of
the studies. A small number of cross-sectional studies also found concurrent blood Pb-associated
decrements in executive function, including an analysis of the Rochester cohort (Froehlich et al.. 2007).
and some studies were limited due to their lack of consideration of potential confounding (Nelson and
Espy. 2009; Vega-Dienstmaier et al.. 2006). A cross-sectional analysis by Cho et al. (2010) with a
concurrent child mean BLL of 1.9 (ig/dL did not find an association with performance on SCWT.

A small number of recent studies expanded the evidence base pertaining to the effect of Pb on
executive functions in children (study details can be found in Section 3.7, Table 3-4E). Most of these
recent studies assessed executive function using parent-teacher ratings on the Behavior Rating Inventory
of Executive Function (BRIEF) (Gioia et al.. 2002). This instrument comprises three scales including a
Behavioral Regulation Index, which has several components (i.e., emotional control, shift, and inhibit).
Other scales of the BRIEF are the metacognition index and the global executive composite. Higher
BRIEF scores indicate executive function-related behavioral dysfunction.

External Review Draft

3-53

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

Fruh et al. (2019) studied mother-child pairs participating in Project Viva, a longitudinal birth
cohort in eastern Massachusetts. Maternal blood Pb concentration in erythrocytes was measured during
the second trimester of pregnancy and parents rated their child's behavior using the BRIEF in mid-
childhood (median 7.7 years). The associations (i.e., (3 coefficients) with the parent and teacher-rated
BRIEF Behavioral Regulation Index were imprecise, i.e., 1.15 (95% CI: -0.22, 2.52) and 0.77 (95% CI:
-0.57, 2.10) per 1 (ig/dL increase in maternal erythrocyte Pb, respectively. In another analysis of these
data, Fruh et al. (2021) aimed to determine the association of joint exposure to Pb, Mn, Se, and MeHg
with scores on the BRIEF and Strengths and Difficulties Questionnaire (SDQ) using BKMR and quantile
g-computation. Individual beta coefficients for each metal from multiple regression models generally
agreed with the original results. Specifically, maternal Pb concentration in erythrocytes (2nd trimester)
was associated with worse parental ratings on the BRIEF global executive composite ((3 = 1.11 [95% CI:
(-0.12, 2.34] per unit increase in maternal erythrocyte Pb). Notably, the mixture was also associated with
poorer parent ratings on the BRIEF in BKMR models.

Sex-specific findings were observed in a study by Merced-Nieves et al. (2022). The researchers
examined the association of prenatal BLL with behavioral tasks on the operant test battery (OTB) (i.e.,
Conditioned Position Responding [CPR], Temporal-Response Differentiation [TRD], Delayed Matching-
to-Sample [DMTS], and Incremental Repeated Acquisition), which assess executive functions, at age 6-
7 years. Maternal blood Pb in late pregnancy was not associated with greater response latencies in the
CPR ((3 = 0.00 [95% CI: -0.08, 0.08]) and DMTS ([3 = 0.08 [95% CI: -0.04, 0.20]) tasks, although a
small increase average latency to initiate a response in the TRD task ([3 = 0.14 [95% CI: -0.00, 0.29]).
The association of blood Pb concentrations with two operant tasks were modified by child sex, indicating
Pb-associated changes in the CPR task were more pronounced in girls, and Pb-associated changes in the
TRD task were more pronounced in boys. The mean BLLs during the first trimester, second trimester, and
delivery, in umbilical cord blood, and postnatal were 3.7, 3.9, 4.3, 3.4, and 2.4 (ig/dL, respectively.

Ruebner et al. (2019) evaluated the association between BLLs and executive function among
children with CKD. In addition to adjusting for important potential confounders including SES and
maternal education, the author adjusted for clinical variables in their models (i.e., CKD stage, duration,
glomerular versus non-glomerular diagnosis, hypertension, proteinuria, and anemia). Executive
functioning was assessed with the Delis-Kaplan Executive Function System Tower Subset
(subjects >6 years) and rated by parents using BRIEF for Preschool Children (BRIEF-P; 2-5 years) and
the standard BRIEF (6-18 years) or self-reported by adults (18 years and older) using BRIEF for Adults
(BRIEF-A). Associations between BLL and behavioral symptoms on BRIEF did not persist in models
that controlled for potential confounders including race, poverty, maternal education, and clinical factors
related to CKD (quantitative results not reported). The median BLL in this study was 1.2 (ig/dL.

External Review Draft

3-54

DRAFT: Do not cite or quote

-------
1

Summary

Strong evidence of associations between Pb exposure and indices of executive function was
described in the 2013 Pb ISA. Studies included prospective analyses of several birth cohorts with
moderate to high follow-up rates in Boston and Rochester that examined the BLLs that preceded the
outcome assessment and adjusted for several potential confounding factors (Canfield et al.. 2004;

Canfield et al.. 2003b; Bellinger etal.. 1994a; Stiles and Bellinger. 1993). Recent studies that assessed
executive functions using parent or teacher behavioral ratings on BRIEF are mixed; however, findings
from these studies do not diminish the evidence from the earlier well-conducted studies that relied on
neuropsychological testing.

3.5.1.4.2 Toxicological Studies of Executive Function

The epidemiologic evidence reviewed above indicated associations between higher childhood
BLLs and poorer performance on tests of executive functions in children and young adults. Pb was
associated with impaired strategic planning, organized search, flexibility of thought and action to a
change in situation, and control of impulses (described in Section 3.5.1.4.1). In rodents, reversal learning
is one of the main frameworks used to measure cognitive flexibility, an important component of executive
function. Reversal learning tasks assess the ability of animals to actively suppress reward-related response
and disengage from ongoing behavior when the conditions governing the response are altered. For
example, if rodents are trained to press the leftmost of two levers to receive a reward, then the
experimenters cease to reward presses of the left lever and begin rewarding right-lever presses, the
animals must learn that the conditions have changed and adjust their behavior accordingly. Perseverance
(i.e., continuing to press the left lever after the rules have changed) represents impaired cognitive
flexibility and execution dysfunction. This basic paradigm can be expanded to parse specific components
of executive function.

Two recent studies from the same group assessed executive function using an attention-set
shifting task paradigm following relevant Pb exposures during development. Neuwirth et al. (2019c)
compared performance in Long-Evans rats exposed to Pb during different windows of development.
Briefly, rats were trained to dig for treats by relying on environmental cues that indicated which of two
bowls contained a buried treat. Trained rats were run through a series of discrimination trials including
interdimensional and extradimensional shifts to test cognitive flexibility. An interdimensional shift occurs
when the relevant cue changes but remains within the same dimension (e.g., the baited bowl is still
indicated by a scent, but the correct scent has changed from lavender to peppermint). More complicated
extradimensional shifts require animals to recognize that the relevant cue has changed dimensions (e.g.,
the baited bowl is no longer indicated by scent but by the texture of the media).

Male rats exposed to Pb via lactation (150 ppm in maternal chow) during the early postnatal
period (PND 0 to 22), which yielded BLLs of ~6 (ig/dL, displayed substantial learning deficits in the form

External Review Draft

3-55

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

of increased Trials-to-criterion for the olfactory (the relevant dimension) discrimination component.
Indeed, males exposed during the early postnatal period were unable to complete discrimination training
and progress to the next task in the same manner as the control males, indicative of a substantial learning
impairment as the result of Pb exposure. No effect was observed in female rats following exposure in the
postnatal period, despite similar BLLs of ~5 (ig/dL. Even though male rats perinatally exposed (GD -14
to PND 22, BLLs of ~6 (ig/dL) successfully completed discrimination training, they struggled during
testing and displayed significant increases in Trials-to-criterion across simple and complex discrimination
tasks yet solved extradimensional shift tasks in a fashion comparable to control males. In contrast to
males, perinatally exposed females performed poorly in extradimensional shift tasks and showed
improved performance relative to controls in discrimination training and reversal (Ncuwirth et al.. 2019c).

In a subsequent study by the same group (Ncuw irth et al.. 2019b). the authors reported that
perinatal exposure (GD 0 to PND 22), which resulted in BLLs -10 (ig/dL on PND 22, also produced sex-
specific learning deficits in discrimination training and impaired reversal learning. These studies provide
evidence that exposure to Pb during different developmental windows may produce differential patterns
of executive dysfunction and that these changes may be sex-specific. While the previous ISA did not
include any toxicological evidence that explicitly addressed executive function, the findings of (Ncuw irth
et al.. 2019c) and (Neuwirth et al.. 2019b) are consistent with the evidence reviewed in the previous ISA
that indicated Pb exposure contributed to cognitive dysfunction and that the effects of Pb on cognition
were often sex-specific.

Summary

The previous ISA did not incorporate any evidence of the relationship between Pb exposure and
executive function in animal models. Two recent studies from the same group provided evidence that Pb
exposure broadly impairs measures of executive function in a reversal learning paradigm. These effects
were sex-specific, with greater effects reported in males. While these reports are consistent with one
another, evidence for the association between Pb exposure and impaired executive function in animal
models with BLLs <30 (ig/dL remains limited.

3.5.1.5 Academic Performance and Achievement in Children

Poorer academic performance and achievement is linked with lower FSIQ and may have
important implications for success later in life (U.S. EPA. 2013a). The 2006 Pb AQCD and the 2013 Pb
ISA described associations of higher blood and tooth Pb levels in children aged 5-18 years with poorer
performance on tests of math, reading, and spelling skills, lower probability of high school completion,
lower class rank, and lower teacher ratings of academic functioning. Notably, associations were reported
in prospective studies examining performance on academic achievement tests (Chandramouli et al.. 2009;
Min et al.. 2009; Miranda et al.. 2009) and an additional analysis of adolescents participating in NHANES

External Review Draft

3-56

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

(Lanphear et al.. 2000). Several prospective studies (Min et al.. 2009; Miranda et al.. 2009) and cross-
sectional studies (Krieg et al.. 2010; Chiodo et al.. 2007; Surkan et al.. 2007; Lanphear et al.. 2000) were
conducted in populations with population or group mean BLLs <5 (ig/dL. In addition, prospective studies
in Boston and New Zealand found associations of tooth Pb levels measured at an earlier age (e.g., ages 6-
8 years) with school performance ascertained at age 18 from school records (Fergusson et al.. 1997;
Needleman et al.. 1990). suggesting the effect of early exposure on Pb may be persistent. The strengths of
the Fergusson et al. (1997) analysis included a low probability of selection bias, coherence with results
indicating associations between higher tooth Pb levels and lower teacher ratings of math, reading, and
writing abilities at ages 12-13 years (Fergusson et al.. 1993). and consideration of important covariates
including SES, parental education, and HOME score. The Needleman et al. (1990) study was relatively
small with no adjustment for parental caregiving quality.

Recent prospective studies of groups or populations with mean BLLs <5 (ig/dL add to the
evidence supporting an effect of Pb exposure on academic achievement and performance (study details
can be found in Section 3.7, Table 3-5E). Prospective studies have been conducted in Detroit, Chicago,
North Carolina, and a 57-county region in New York State (all counties outside New York City).

Zhang et al. (2013) studied children enrolled in Detroit public schools to determine the
association between childhood BLL measured before age 6 and performance on standardized tests for
math, reading, and science in grades 3,5, and 8. Compared with students with lower BLLs (defined as
levels <1 (ig/dL), students with higher BLLs (defined as levels between 1 and 5 (ig/dL) had increased risk
of scores that were classified as less than proficient (OR = 1.42 [95% CI: 1.24, 1.63] for math, OR =1.33
[95% CI: 1.10, 1.62] for science, and OR= 1.45 [95% CI: 1.27, 1.67] for reading. Logistic regression
models were adjusted for covariates including SES (i.e., free and reduced school lunch participation) and
maternal education.

Evens et al. (2015) conducted a study in children enrolled in the Chicago public school system.
The association between childhood BLLs measured before 72 months and failure on standardized tests for
math and English in grade 3 was examined. This study found that a 1 (ig/dL increase childhood BLL was
associated with an increased risk of failure on the reading and math tests (RR = 1.06 [95% CI: 1.05, 1.07]
and RR = 1.06 [95% CI: 1.05, 1.07], respectively) after adjustment for covariates including child
characteristics, preterm birth, maternal education, and SES (i.e., participation in the free and reduced
lunch program). The associations of BLL with reading failure in white, black and hispanic children were
1.14 (95%CI: 1.08, 1.20), 1.05 (95% CI: 1.04, 1.06) and 1.08 (95%CI: 1.05, 1.11), respectively. The
associations of BLL with math failure were in white, black and hispanic children were 1.11 (95% CI:
1.05, 1.18), 1.05 (95% CI: 1.04, 1.06) and 1.09 (95% CI: 1.06, 1.12), respectively. The mean BLL was
4.81 (ig/dL in this study. Blackowicz et al. (2016) extended this analysis through their examination of the
association of BLL and failure on standardized tests for math or reading among Hispanic children
enrolled in the Chicago school system. An association between 1 ug/dL change in BLL and failures in

External Review Draft

3-57

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

reading (RR= 1.07 [95% CI: 1.05, 1.10]) and math (RR= 1.09 [95% CI: 1.06, 1.12]) were observed. The
mean BLL was 4.16 (ig/dL in Hispanic children in this study.

In a statewide study of North Carolina school children (Shadbcgian et al.. 2019). children with
higher BLLs had, on average, lower scores in both math and reading (averaged over grades 3 and 8) than
children with lower BLLs. Compared with children with BLL <1 (ig/dL, the authors reported a decrease
in the test-score percentile of 0.95 (0.66, 1.24) for math and 1.41 (1.12, 1.70) for reading in children with
a BLL of 5 (ig/dL. Shadbegian et al. (2019) included interaction terms between BLL and the grade of
testing that further indicated that the deficit in the test score persisted from grade 3 to grade 8.

SkerfVing et al. (2015) studied the association of childhood BLL (age 7-12 years) with school
performance in the ninth grade at age 16 among Swedish school children. School performance was based
on a 1-5 point passing grade scale or a 4-level merit system in which 0, 10, 15, or 20 points were
assigned for each increasing level of performance. This study found a 0.11-point decrease (95% CI:
-0.18, -0.05) per 1 (ig/dL increase in BLL for school performance using the grading scale and a -10.90
(95% CI: -15.49, -6.31) point decrease using the merit scale, among school children with BLLs
<5 (ig/dL. The models were adjusted for covariates including parent's income, education, and father's IQ
score on the military conscription exam. The association with IQ ([3 = -0.20 [95% CI: -0.39, -0.02])
among those evaluated for military conscription at age 18 was also examined (see Section 3.6.1).

3.5.1.5.1 Summary

Associations of higher blood and tooth Pb levels in children aged 5-18 years with poorer
performance on tests of math, reading, and spelling skills, lower probability of high school completion
and lower-class rank, and lower teacher ratings of academic functioning were observed in previous
assessments (U.S. EPA. 2013a). Recent studies in populations of children (age 6-16 years) with BLLs
<5 (ig/dL support and extend these observations of poorer academic performance in association with
increasing Pb exposure.

3.5.1.6 Relevant Issues for Interpreting the Evidence Base

3.5.1.6.1 Concentration-Response Function

With each previous assessment (U.S. EPA. 2013a. 2006b). the epidemiologic and toxicological
study findings have shown that progressively lower BLLs or Pb exposures are associated with cognitive
deficits in children. The 2006 AQCD found that cognitive effects in children were associated with BLLs
of 10 (ig/dL and lower, while the evidence assessed in the 2013 Pb ISA found that an association between
BLLs and cognitive effects in children was substantiated to occur in populations of young children with

External Review Draft

3-58

DRAFT: Do not cite or quote

-------
1

mean BLLs between 2 and 8 (ig/dL. The conclusions of the 2013 Pb ISA were based on studies that
examined early childhood BLLs (i.e., age <3 years), considered peak BLLs in their analysis (i.e., peak
<10 (ig/dL), or examined concurrent BLLs in young children (i.e., age 4 years). The lower bound of this
mean BLL range was derived from Miranda et al. (2009). who examined the association between early
childhood BLL and academic performance among school-aged (grade 4) children. Although some
individual recent studies found associations of Pb exposure with cognitive effects in children with mean
BLLs <2 (ig/dL (e.g., (Martin et al.. 2021; Dantzer et al.. 2020; Desrochers-Couture et al.. 2018; Hong et
al.. 2015)). the studies generally involved somewhat older children with lengthier exposure histories, or
employed modeling strategies designed to answer relatively narrow research questions (e.g., the effect of
joint exposure to Pb and other metals or the effect of concurrent Pb exposure independent from prenatal
exposure). Consequently, the studies did not provide evidence that would change the conclusion of the
2013 Pb ISA that cognitive effects in children are best substantiated in young children with mean BLLs
between 2 and 8 (ig/dL. Studies that might extend the evidence related to exposure-response relationships
(i.e., recent studies that reflect the lower early childhood Pb exposures, which are now more common in
the U.S.]) are generally lacking. Overall, the recently available studies were not designed, and may not
have the sensitivity (Cooper et al.. 2016). to detect the effect or hazard at very low BLLs, nor do they
provide evidence to support the existence of an exposure threshold across the range of BLLs that were
examined. The finding of higher mean IQ with decreasing blood Pb concentration observed across
epidemiologic studies, however, indicates that the absolute magnitude of the effect of Pb exposure on
cognitive function is smaller with decreasing BLL.

Despite limitations, several recent studies describe the cognitive effects over the range of Pb
exposure examined. Shadbegian et al. (2019) extended the analysis conducted by Miranda et al. (2009)
also using data from the statewide study of North Carolina school children while focusing on lower BLLs
(<10 (ig/dL) and characterized the persistence of Pb effects across grades. Among children with BLLs of

5 (ig/dL or lower, a decrease in the test score percentile of 0.95 was found for math and a decrease of 1.41
was found for reading when comparing children with a BLL of 5 to those with a BLL <1 (ig/dL. Figure 3-

6 depicts the association of blood Pb levels with math and reading test score performance (average
percentile decrement), with 95% CIs, among children in the third and eighth grades.

External Review Draft

3-59

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

123456789 10

123456789 10





c

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

Pb exposure on cognitive function and the supralinear concentration relationship between Pb exposure
and FSIQ. The beta coefficients from the log-linear models, which indicate larger incremental effects of
Pb at lower blood Pb concentration, for (Lanphear et al.. 2019. 2005) and Crump et al. (2013) were
comparable (i.e., (3 = -2.65 [95% CI: -3:69, -1:61] per unit of natural log transformed BLL and (3 = -3.32
[95% CI: -4.55, -2.08] per unit of natural log transformed BLL+1, respectively).

Attenuation of C-R relationships at higher exposure or dose levels has been reported in the
occupational literature. Reasons proposed to explain the attenuation include greater exposure
measurement error and saturation of biological mechanisms at higher levels as well depletion of the pool
of susceptible individuals at higher exposure levels (Stavncr et al.. 2003). Possible explanations specific
to nonlinear relationships observed in studies of Pb exposure in children include a lower incremental
effect of Pb due to covarying risk factors such as low SES, poor caregiving environment, and higher
exposure to other environmental factors (Schwartz. 1994a). differential activity of mechanisms at
different exposure levels, and confounding by omitted or mis-specified variables (U.S. EPA. 2013a).
Review of the evidence did not reveal a consistent set of covarying risk factors to explain the differences
in blood Pb IQ C-R relationship across high and low Pb exposure groups observed in epidemiologic
studies. Recent studies in populations with mean concentrations of 2 (ig/dL or lower indicated that some
of the observed heterogeneity at lower BLLs may be explained by the underlying distribution in at-risk
factors, including other metals. In addition, some recent studies with similarly low blood Pb
concentrations reported effect modification by sex. These studies are discussed in more detail in
Section 3.5.1.6.2.

A limited number of recent studies examine the shape of the C-R function for the relationship
between Pb exposure and cognitive effects in children. Lucchini et al. (2012) conducted a cross-sectional
analysis of children between the ages of 11 and 14 to examine the relationship between concurrent BLL
and FSIQ. The mean BLL was 1.71 (ig/dL in this study. The relationship between BLL and IQ using a
restricted cubic spline fit is plotted in Figure 3-7. As shown in the plot, the decrement in IQ is not
constant over the range in BLLs (0.44-10.2 (.ig/dL). The study was conducted in an area where ferroalloy
plants had operated and the extent to which the children in the study were exposed to higher Pb levels
during early childhood was not clear from this cross-sectional analysis.

External Review Draft

3-61

DRAFT: Do not cite or quote

-------
2

A	e

RbtupfflLJ

d

10

IQ = intelligence quotient.

Source: Lucchini et al. (2012).

Figure 3-7 Relationship between concurrent blood Pb level and intelligence
quotient among Italian adolescents using a cubic spline fit.

Lucchini et al. (2012) also plotted relationship between the log-transformed BLL (ordinary least
squares fit) and FSIQ (Figure 3-8). A decrement in FSIQ score was observed in association with In
concurrent BLL after adjustment for covariates including SES and maternal education (-2.24 [95% CI:
-4.10, -0.37)]. Consistent with evidence reviewed in the 2013 Pb ISA, the log transformation of BLL
implies a larger incremental decrement in IQ at lower BLLs. Lucchini et al. (2012) calculated the
benchmark dose (BMD) for blood Pb, which is the dose that results in a specific IQ loss (i.e., a loss of one
IQ point), and its lower 95% confidence limit (BMDL) using this C-R function. The BMDL calculated
from these data is 0.11 (.ig/dL. As noted previously the older children in this study may have had higher
past Pb exposure that was not reflected in their concurrent BLL.

External Review Draft

3-62

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

130 -

# ft #

* # •

• *
~ *

» *4» » *	*

* * # •

120	• ~

*

* *	t •

110-

2

D

»- 100

# ~ ** * *
• • « • *

*

*

* « *

*

*

* * *
* ~ *

•



***** *



* *
** * * * *



*-*	* * *



~ ** *
*

' -—^	^

*	* * « • *	I ~*

• • * •
• • *
* ******
• ~ * *

• ~
• * * * *

so	: i **.*

* * ~

80

* *

* *

i

0.5	12	5	10

Pb (pgltiL)

IQ = intelligence quotient.

Source: Lucchini et al. (2012).

Figure 3-8 Relationship between log-transformed blood Pb level and
intelligence quotient using an ordinary least squares fit.

Experimental animal studies support the findings in epidemiologic studies indicating the effect of
Pb exposure on cognition at low exposures. Clear support from animal toxicological studies that
demonstrated decrements in learning, memory, and executive function with dietary exposures resulting m
relevant BLLs was assessed in the 2013 Pb ISA. Recent experimental animal studies further support
impairments in cognitive function at BLLs <20 ng/dL. It is well-documented (and reviewed in the
previous ISA) that Pb exposures resulting in BLLs > 20 p.g/dL consistently produced deficits in cognitive
function. Recent evidence (reviewed in the current ISA; see Section 3.5.1.3.2) suggests that BLLs
resulting from lower-level exposures (5 - 10 ug/dL) also lead to cognitive function deficits in animal
models. For example, Zhou et al. (2020a); Zhao et al. (2018); Xiao et al. (2014); Bethana and Maher
(2012); Corv-Slechta et al. (2012) all reported significant cognitive deficits following exposures yielding
BLLs <10 |ig/dL. Betharia and Maher (2012) reported the lowest BLLs for animals tested in the Morris

External Review Draft

3-6,3

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

water maze (0.98 j^ig/dL at PND 21 and later 0.03 (ig/dL at PND 56) which produced mild effects on
memory in females at the later time point only.

The reader should be aware that BLLs reported in many of these studies were measured later
during experimentation and do not necessarily reflect peak Pb burden or Pb burden during the most
sensitive window of brain development for young animals and children; BLLs in toxicological studies
should be interpreted in the context of both exposure time course and blood collection. Discussion of
BLLs and cognitive function is further complicated by the exposure window. The evidence generally
supports the notion that Pb exposures during brain development led to changes in cognition that persist
after cessation of exposure and BLLs have decreased. For example, Barkur and Bairv (2015b) compared
multiple different windows of exposure and reported the greatest decreases in learning and memory
following gestational and lactational windows, consistent with the altricial nature of rodent brain
development.

While nonlinear C-R relationships including U- or inverted U-shaped curves for various
endpoints, including those related to cognitive impairment, were demonstrated in the toxicological
literature discussed in the previous ISA, these toxicological findings are distinct from epidemiologic
findings of supralinear relationships in that some U- or inverted U-shaped relationships do not indicate
Pb-induced impairments at higher exposure concentrations (U.S. EPA. 2013a). Recent animal studies do
not provide evidence for an inverted dose relationship, rather increased Pb doses (and resulting BLLs)
generally resulted in greater cognitive impairment. This may be related to the refined PECOS used in the
current ISA, which did not incorporate studies reporting BLLs higher than 30 (ig/dL, thus narrowing the
range of doses integrated. Thus, recent evidence generally supports dose-dependent effects of Pb on
cognitive function at relevant BLLs.

In summary, recent studies support and extend the evidence pertaining to the effect of Pb
exposure on cognitive function in children at low BLLs. These effects are best substantiated to occur in
study populations with mean BLLs between 2 and 8 (ig/dL. Association between Pb exposure in
populations of children below 2 (ig/dL are reported, extending the evidence described in the 2013 Pb ISA;
however, heterogeneity at lower exposure levels (i.e., not all studies report positive associations) has been
observed. Recent experimental studies of rodents continue to support impairments in cognitive function at
BLLs <30 (ig/dL. Compelling evidence for a larger decrement in cognitive function per unit increase in
blood Pb among children with lower mean blood Pb concentrations, compared with children with higher
mean blood Pb concentrations, across a broad range of BLLs (e.g., 5th percentile of 2.5 ug/dL up to 95th
percentile of 33 ug/dL) was supported by a reanalysis of a pooled international dataset Crump, 2013,
3838553}. Recent studies with an adequate range of Pb exposure measured during relevant time periods
that would be required to evaluate exposure-response relationships are generally lacking. Considering the
collective body of studies, no evidence of a threshold for cognitive effects in children across the range of
BLLs examined in epidemiologic studies was reported.

External Review Draft

3-64

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

3.5.1.6.2 Confounding

The 2013 Pb ISA described multiple factors that influence cognitive function and behavior in
children including parental IQ and education, SES of the family, quality of the caregiving environment
(i.e., HOME score), and other environmental exposures (U.S. EPA. 2013a; Wasserman and Factor-Litvak.
2001). These other risk factors often are correlated with blood, tooth, and bone Pb levels, and thus, are
considered as potential confounding factors in epidemiologic analyses. The collective epidemiologic
evidence consistently demonstrates associations of higher blood and tooth Pb levels with cognitive
function decrements and poorer behavior in children. These associations were observed in diverse
populations in the U.S., Mexico, Europe, Asia, and Australia. Associations have been observed across
studies that used different methods to control for confounding and adjusted for different potential
confounding factors, commonly maternal IQ and education, SES, and HOME score. Several studies have
found associations with additional adjustments for smoking exposure, birth outcomes, and nutritional
factors. Multiple recent studies adjusted for exposure to other metals or environmental chemicals e.g.,
Zhou et al. (2020b') and Liu et al. (2015); however, there remains uncertainty regarding the
appropriateness of the adjustment for other metals as confounders in studies that did not examine the
potential for interactions (see Section 3.5.1.6.5 for evidence related to interactions between Pb and other
metals).

As noted in the 2013 Pb ISA, no single method to control for potential confounding is without
limitation, and there is potential for residual confounding by unmeasured factors. However, the
consistency of findings among different populations and study methods with consideration of several well
characterized potential confounding factors as described above increases confidence that the associations
observed between Pb biomarker levels and neurodevelopmental effects in children represented a
relationship with Pb exposure. Recent studies expanded the evidence reporting associations between Pb
exposure and nervous system effects in children after consideration of covariates including sex, maternal
stress and race or ethnicity as potential effect modifiers as opposed to confounders (see Section 3.5.1.6.5).
Biological plausibility was derived from extensive evidence provided by animal toxicological studies that
are experimental in design and thus, not vulnerable to confounding. These experimental animal studies
demonstrate the effect of Pb on cognition and behavior as well as changes in neurogenesis, synaptic
pruning, and neurotransmitter function in the hippocampus, prefrontal cortex, and nucleus accumbens of
the brain (U.S. EPA. 2013a). Recent experimental animal studies support the evidence described in the
2013 Pb ISA, provide additional evidence for Pb-induced impairments in learning and memory (short and
long-term) assessed by several methods not discussed in the 2013 Pb ISA, and extend the limited
evidence related to Pb-induced impairment of executive functions. These experimental animal studies
provide strong support that the effects observed in epidemiologic studies cannot be explained by
confounding.

External Review Draft

3-65

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

3.5.1.6.3 Lifestages

Epidemiologic studies reviewed in the 2013 Pb ISA consistently showed that BLLs measured
during various lifestages and time periods (i.e., prenatal, early childhood, childhood average, and
concurrent with the outcome) were associated with cognitive function decrements in children (U.S. EPA.
2013a). Epidemiologic studies consistently pointed to inverse associations between FSIQ in school-aged
children and BLLs measured at various lifestages and time periods (Table 4-14 U.S. EPA (2013a)). In an
analysis of data from seven prospective studies Lanphear et al. (2019) found that increases in early
childhood (age 6-24 months on average), peak, concurrent, and lifetime average BLLs were associated
with decreases in FSIQ in children at ages 4-10 years. The investigators reported that the best predictor of
IQ decrement, as indicated by the model R2 value, was early childhood blood Pb concentration
(R2 = 0.6433), although the R2 value for the concurrent metric (0.6414) was nearly identical (Lanphear et
al.. 2019; Crump et al.. 2013). These results illustrated the challenge of distinguishing a critical time
period when exposures are highly temporally correlated. Epidemiologic studies that aimed to improve the
characterization of important lifestages and time periods of Pb exposure by examining children in whom
BLLs were not strongly correlated with exposure over time indicated FSIQ decrements in association
with higher concurrent BLLs but did not conclusively demonstrate stronger findings for early versus
concurrent BLLs (Table 4-15 of U.S. EPA (2013a)). Considering the collective body of epidemiologic
evidence reviewed in the 2013 Pb ISA, there was no clear indication of a single critical lifestage or
duration of Pb exposure that is uniquely associated with the risk of neurodevelopmental effects in
children. These observations in the epidemiologic literature were supported by experimental animal
evidence. Consistent with findings from the 2013 Pb ISA, more recent studies continue to report
associations with prenatal BLLs (maternal and cord blood Pb) and postnatal BLLs measured at various
childhood lifestages despite some heterogeneity in the magnitude and direction of the associations at
BLLs <5 (ig/dL.

Maternal Pb exposure presents an exposure risk during gestation and early infancy, when
important neurodevelopmental processes are known to occur. Substantial fetal Pb exposure may occur
from mobilization of maternal skeletal Pb stores (Gulson et al.. 2003; Hu and Hernandez-Avila. 2002) and
its transfer across the placenta (Section 3.2.2.4 of U.S. EPA (2013a)). Among studies that examined BLLs
at multiple time periods, some found a larger decrement in MDI per unit increase in prenatal blood Pb
than concurrent blood Pb ((Hu et al.. 2006; Gomaa et al.. 2002). Table 4-14 of U.S. EPA (2013a)).

Prenatal and early postnatal (age 6 months) BLLs were also associated with cognitive function in studies
that included school-aged children (ages 5-17 years) (Table 4-14 of U.S. EPA (2013a)). Sanchez et al.
(2011) extended the analysis of Hu et al. (2006). which was designed to elucidate the time window during
pregnancy that the effect of Pb exposure on neurodevelopment is most pronounced among participants in
a birth cohort study in Mexico City. These authors compared methods to model exposure and found that
the MDI score at age 2 was sensitive to the choice of method. A decrease in MDI score of 2.74 (95% CI -
5.78 to 0.29) per natural log increase in BLL during the first trimester was observed, using a window-

External Review Draft

3-66

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

specific regression, while the corresponding decrease was larger [(3=—4.13 (95% CI, -7.54, -0.72) using a
multiple informant model.

As described above, however, most of these studies also found cognitive function decrements in
association with postnatal BLLs, and the results did not identify an individual critical postnatal time
period of blood Pb measurement associated with cognitive function decrements. Maternal pregnancy-cord
BLL correlations of 0.53-0.81, depending on the stage of pregnancy, were reported by Schell et al.
(2003). Depending on the magnitude of child exposure, the contribution of maternal blood Pb to child
BLLs appears to diminish rapidly over a period of a few months following birth, after which child BLLs
may be influenced mainly by postnatal Pb exposures (Section 3.4.1 of U.S. EPA (2013a)).

Recent studies observed associations between Pb exposure during prenatal and childhood
lifestages (i.e., maternal (Y Ortiz et al.. 2017; Vigeh et al.. 2014; Kim et al.. 2013b. c), cord (Valcri et al..
2017). and postnatal exposure (Lin et al.. 2013)) and poorer performance on tests of neurodevelopment
among mothers and infants with mean BLLs <5 (ig/dL. The is some evidence indicating that there
heterogeneity in the magnitude and direction of the observed associations in recent studies may be
explained, in part, by co-exposure to other metals or maternal stress

Experimental animal studies demonstrated that prenatal or early postnatal or lifetime Pb exposure
alters brain development via changes in synaptic architecture and neuronal outgrowth, leading to
impairments in memory and learning (Sections 4.3.10.4, 4.3.10.10, and 4.3.2.3 of U.S. EPA (2013a)).
Gestational or infancy Pb exposures are not necessary to induce cognitive function decrements in juvenile
animals, however. Studies of monkeys have found that Pb exposures during lifestages and time periods
extending from infancy through the juvenile or adult periods resulted in impaired cognitive function
(Rice. 1992; Rice and Gilbert. 1990a; Rice. 1990; Rice and Karpinski. 1988). These findings are
consistent with studies of individuals aged 3 to 30 years, which showed that brain development
ascertained using MRI continues throughout adolescence, indicating the potential for alterations to
neurodevelopment later in childhood (Gerber et al.. 2009; Lenroot and Giedd. 2006).

Additional recent animal studies support the notion that various exposure periods (i.e.,
preconception, gestation, lactation) may represent a critical periods during which Pb exposure can cause
cognitive impairment later in life. In rodents, developmental exposure to Pb was consistently associated
with persistent cognitive effects observed both early (Tartaglionc et al.. 2020; Zhao et al.. 2018; Barkur
and Bairv. 2015b; Anderson et al.. 2012) and later in life (Liu et al.. 2022c; Xiao et al.. 2014; Betharia
and Maher. 2012). Few studies were designed to compare exposures across multiple different
developmental windows (Barkur and Bairv. 2015b; Xiao et al.. 2014). These studies reported similar
magnitudes of effects between developmental windows, suggesting that individual periods of
development may be similarly sensitive to Pb. Generally, longer exposures that spanned multiple
developmental periods (e.g., preconception through lactation) produced not only the highest BLLs but the
largest effects on cognition (Zhou et al.. 2020a; Zhu et al.. 2019b).

External Review Draft

3-67

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

Unlike other organ systems, the unidirectional nature of CNS development limits the capability of
the developing brain to compensate for cell loss, and environmentally induced cell death can result in a
permanent reduction in cell numbers (Baver. 1989). Hence, when normal development is altered, the early
effects have the potential to persist into adult life even in the absence of concurrent exposure, magnifying
the potential public health effects. A limited number of studies examined the persistence of the effects of
Pb on cognitive function. A recent study by Shadbcgian et al. (2019) indicated that poorer performance on
tests of reading and math associated with earlier childhood Pb exposure persisted from grade 3 to grade 8.
Some epidemiologic evidence reviewed in the 2013 Pb ISA indicated associations of earlier childhood
blood or tooth Pb levels in adolescents or adults with decreased cognitive function (Mazumdar et al..
2011; Ris et al.. 2004; Stiles and Bellinger. 1993). Recent studies support and extend this evidence.
Specifically, childhood Pb exposure was observed to have long-term cognitive consequences in young-
(18-19 years) (Skcrfving et al.. 2015) or mid-adulthood (38 or 45 years of age) (Reuben et al.. 2020;
Reuben et al.. 2017). These epidemiologic studies did not examine adult BLLs, thus the relative influence
of adult Pb exposure was not ascertained. The persistence of effects of early exposures, however, is
supported by findings of impaired learning in adult monkeys exposed to Pb only during infancy (Rice.
1992; Rice and Gilbert. 1990a; Rice. 1990). Additional recent studies in rodents provided support for the
persistence of effects of early exposures of Pb (Xiao et al.. 2020; Li et al.. 2016a; Xiao et al.. 2014; Zhang
et al.. 2014; Rahman et al.. 2012b; Zhang et al.. 2012; Kuhlmann etal.. 1997).

There is some evidence that the effects of early Pb exposure on cognitive function are not fixed.
Results indicated higher cognitive function in children at ages 1-8 years who had declines in BLL over
durations of 6 months to 5 years compared with children with smaller declines, no change, or increases in
BLLs in some studies (Hornung et al.. 2009; Chen et al.. 2005; Liu et al.. 2002; Ruff etal.. 1993;

Bellinger et al.. 1990). This evidence pertains to populations with declines from higher BLLs at baseline
(20-55 (ig/dL) or larger declines over time (i.e., 8, 14 (ig/dL) than those expected for most of the current
population of U.S. children. No recent studies that provided additional information on this topic were
identified.

To conclude, the collective body of epidemiologic evidence reviewed in the 2013 Pb ISA did not
provide strong evidence to identify an individual critical lifestage or timing of Pb exposure with regard to
neurodevelopmental effects in children (U.S. EPA. 2013a). Recent studies support this conclusion.
Evidence indicates that prenatal BLLs are associated with mental development in very young children
aged <2 years. Several studies indicated that increases in postnatal (earlier childhood, lifetime average,
concurrent) BLLs were associated with larger cognitive function decrements in children aged 4-10 years
than were similarly sized increases in prenatal BLLs. These results suggest that per unit increase,
postnatal Pb exposures that are reflected in concurrent or cumulative BLLs or tooth Pb levels may have a
larger magnitude of effect on cognitive function decrements as children age (U.S. EPA. 2013a). The
identification of critical lifestages and time periods of Pb exposure is complicated by the fact that BLLs in
older children, although affected by recent exposure, are also influenced by Pb stored in their bone and
maternal Pb stores. Thus, associations of neurodevelopmental effects with concurrent BLL in children

External Review Draft

3-68

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

may reflect the effects of past and recent Pb exposures. Nonetheless, the epidemiologic evidence for
associations of neurodevelopmental effects with multiple lifestages or time periods of Pb exposure,
including more recent exposures, is supported by evidence in monkeys that Pb exposures in infancy,
lifetime exposure starting from birth, or lifetime exposure starting during the juvenile period induce
impairments in cognitive function when assessed between the ages of 6 and 10 years.

3.5.1.6.4 Public Health Significance

The 2006 Pb AQCD and the 2013 Pb ISA (U.S. EPA. 2013a. 2006b) concluded that
neurodevelopmental effects in children were among the effects best substantiated as occurring at the
lowest BLLs. Evidence from several cohorts of children indicated that there is a supralinear C-R
relationship between blood Pb and FSIQ (i.e., larger incremental effect of blood Pb on FSIQ at lower
levels) and no threshold was identified for Pb-associated neurodevelopmental effects in the range of BLLs
examined. The evidence reviewed in the current assessment supports these conclusions and continues to
clearly indicate that neurodevelopmental effects in children are among the greatest public health concern
associated with Pb exposure.

Cognitive function in children has been assessed using a variety of tests, including FSIQ, BSID,
academic performance, and academic achievement. As noted in the 2013 Pb ISA (U.S. EPA. 2013a).

FSIQ has strong psychometric properties (i.e., reliability, consistency, validity), is among the most
rigorously standardized cognitive function measures, is relatively stable in school-age children, and has
been predictive of educational achievement and life success. In children aged 6 months to 3 years, the
BSID is commonly used to assess mental development; however, the BSID MDI is not an intelligence test
and MDI scores are not necessarily strongly correlated with later measurements of FSIQ in children with
normal development. Lower FSIQ is also linked to poorer academic performance and achievement, both
of which have important implications for success later in life including reduced earning potential and
productivity (Lin et al.. 2016; U.S. EPA. 2013a; Salkever. 1995; Schwartz. 1994b). Analyses of end-of-
grade tests from North Carolina indicated that early childhood BLL is associated with reduced
performance on the tests, the cumulative effect of Pb and low SES is more pronounced at the lower end of
the test score distribution (Miranda et al.. 2009). and the effects of Pb exposure persisted from grade 3 to
grade 8 (Shadbcgian et al.. 2019). Tests of academic achievement generally measure a child's
understanding of a given curriculum that is developed and implemented through the school system; thus,
because exams are typically specific to each state, data cannot be directly compared across states.

The World Health Organization (WHO) definition of "health" is "the state of complete physical,
mental, and social well-being and not merely the absence of disease or infirmity" (WHO. 1948). By this
definition, decrements in health status that are not severe enough to result in the assignment of a clinical
diagnosis might reflect a decrement in the well-being of an individual. Further, deficits in subtle indices
of health or well-being may not be observable except in aggregate, at the population level; therefore, a
critical distinction between population and individual risk is essential for interpreting the public health

External Review Draft

3-69

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

significance of study findings. This concept of population risk is relevant to the interpretation of findings
regarding IQ in the assessment of their public health significance. Specifically, Weiss (1988) discussed
the hypothetical effects of a small shift in the population distribution of IQ score. As shown in Figure 3-9,
these authors anticipate that even a small shift in the population mean IQ may be significant from a public
health perspective because such a shift, given a normal distribution, could yield a larger proportion of
individuals functioning in the low range of the IQ distribution, which is associated with increased risk of
educational, vocational, and social failure (Section 4.3.13), as well as reduce the proportion of individuals
with high IQ scores. Although the change in population mean IQ score may be small relative to the
standard error for the IQ measurement, a study that is large enough will have adequate statistical power to
detect small changes at the population level. Bias may be introduced if the measurement error of the
outcome is highly correlated with the exposure, but there is no evidence to suggest that individuals with
higher BLLs test systematically lower than their true IQ.

IQ = intelligence quotient.

Note: Two distributions of intelligence test scores. (Left): Based on a mean of 100 (the standardized average, with SD of 15).
(Right): Demonstrating a 5% reduction model, based on a mean score of 95. This is a conceptual model that assumes that the
incremental C-R between Pb exposure and IQ is similar across the full range of IQ and is not based on actual data. The figure
shows that the effect of a small shift in population mean IQ score may result in a larger proportion of individuals with IQ scores
below 70 and a smaller proportion with IQ scores above 130.

Source: Reproduced with permission of Elsevier; from Weiss (1988).

Figure 3-9 Two distributions of intelligence test scores demonstrating the
consequence in a small shift in the mean score.

3.5.1.6.5 Potentially At-Risk Populations

The 2013 Pb ISA described physiologic factors that influence the internal distribution of Pb (U.S.
EPA. 2013a). Blood and bone Pb measurements are influenced to varying degrees by biokinetic processes
including absorption, distribution, metabolism, and excretion. These processes are affected by age,
genetics, diet, and co-exposures to other metals and chemicals, which are summarized in the executive
and integrative summaries (https://cfpub.epa.gov/ncea/isa/recordisplav.cfm?deid=357282). In addition to

External Review Draft

3-70

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

these physiological factors, several population characteristics that explain differential Pb exposure have
been identified. These factors included age, sex, race and ethnicity, proximity to Pb sources, and
residential sources and are also discussed in the executive and integrative summaries
(https://cfpub.epa.gov/ncea/isa/recordisplav.cfm?deid=357282.) The factors potentially related to
increased risk of Pb-induced cognitive effects (i.e., factors identified in epidemiologic studies that
conduct stratified analyses and compare the magnitude of the observed association across stratum) are
discussed below.

Age

This section on Potentially At-Risk Populations emphasizes stratified results described in some
epidemiologic studies, as opposed to the large body of longitudinal studies following mothers and infants
throughout childhood that comprises the most compelling body of evidence in support of conclusions
regarding childhood as an "at-risk" factor. As noted in previous sections of the document (i.e.,

Section 3.5.1.1 and 3.5.1.6.1), recent evidence supports the finding from the 2013 Pb ISA that cognitive
effects in young children is the outcome that best substantiated to occur at the lowest exposure levels.
Strong evidence indicates increased risk of Pb-induced neurocognitive effects during several childhood
lifestages throughout gestation, childhood, and into adolescence (see Section 3.5.1.6.3). Moreover, the
integrated synthesis (Section 7.4.2.2) of this document concludes that, "In consideration of the evidence
base (e.g., stratified and longitudinal analyses) and integrating across disciplines of toxicokinetics,
exposure, and health, there is adequate evidence to conclude that children are an at-risk population."

Sex

Multiple epidemiologic studies included in the 2013 Pb ISA examined Pb-related effects on
cognition separately in males and females. Studies on cognition from the CLS cohort and a study in
Poland reported larger magnitude Pb-associated cognitive effects in males (Jedrvchowski et al.. 2009a;
Ris et al.. 2004; Dietrich et al.. 1987). whereas studies from Australia indicated that females were at
increased risk of Pb-associated cognitive effects (Tong et al.. 2000; Baghurst et al.. 1992; McMichael et
al.. 1992). While toxicological evidence supporting sex-specific effects of Pb on cognitive function was
summarized in the previous ISA (Virgolini et al.. 2008; Yang et al.. 2003; Mcgivern et al.. 1991). the
number of studies considering sex as a factor was limited. Recent evidence provides more support for the
sex-biased effects. One study reported a male-specific effect (Anderson et al.. 2016) and several studies
demonstrated female-specific effects (Tartaglione et al.. 2020; Verma and Schneider. 2017; Anderson et
al.. 2012; Betharia and Maher. 2012). The evidence supports a conclusion that there are sex-related
differences in the effects of Pb on cognitive function, yet it remains difficult to parse the exact nature and
direction of sex-specific effects given the variation in outcomes examined, exposure timing and the
considerable number of studies that only reported data from one sex at a time.

External Review Draft

3-71

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

Several recent epidemiologic studies examined sex-stratified associations of Pb exposure with
cognitive effects (Tatsuta et al.. 2020; Zhou et al.. 2020b; Desrochers-Couture et al.. 2018; Taylor etal..
2017). Tatsuta et al. (2020) found a decrement in FSIQ score in association with postnatal BLL
((3 = -9.880 [95% CI: -2.905, 5.831]) among boys, with a smaller less precise association among girls
((3 = -4.406 [95% CI: -15.94, 7.129]). Prenatal BLL was associated with a smaller and less precise
decrease in FSIQ score among boys ([3 = --3.683 [95% CI: -10.714, 3.349]) but not among girls
([3 = 1.463 [95% CI: -2.905, 5.831]). A lower BNT score (with cues) was associated with both pre- and
postnatal BLL among boys in this study. Desrochers-Couture etal. (2018) studied the association
between cord, maternal and childhood (3-4 years old) BLLs with cognitive function (WPPSI-III at age 3-
4 years) among Inuit children in Quebec. Although no associations were observed between cord or
childhood concurrent BLLs and FSIQ, an association was observed between cord BLL and performance
IQ in boys ([3 = -3.28 [95% CI: -5.31,-1.18] per doubling) that was not present in girls ([3 = 0.16 [95%
CI: -1.76, 2.06] per doubling). Although associations were imprecise, Taylor etal. (2017) found an
association between increased maternal BLL and IQ decrements in boys but not in girls enrolled in the
ALSPAC study (e.g., -0.29 [95% CI: -1.02, 0.44] versus 0.73 [95% CI: 0.39, 1.33], respectively on the
WISC). Using models that adjusted for co-exposure to metals (Mn and Cd), Zhou et al. (2020b) found no
association between cord BLL and FSIQ in boys or girls. Further, using models that adjusted for other
chemicals (i.e., DDE, HCB, PCBs, and Mn), Oppenheimer et al. (2022) found no statistical evidence of
an interaction between prenatal Pb exposure and sex.

Maternal Stress

Toxicological studies assessed in the 2013 Pb ISA demonstrated that early life exposure to Pb and
maternal stress can result in dysfunction of the HPA axis (U.S. EPA. 2013a). Recent toxicological
evidence provides further support for this interaction between maternal stress and Pb exposure during
development. Anderson et al. (2012) demonstrated that exposure to Pb blunted the positive effects of
environmental enrichment on learning in the Morris water maze paradigm. Interestingly, Corv-Slechta et
al. (2012) reported that contrary to the effect previously reported in females, maternal stress improved the
performance of Pb-exposed male offspring in a repeated performance and learning paradigm compared
with Pb-exposed males without maternal stress. While this study supports the interaction between
maternal stress and Pb exposure, it remains unclear whether stress would positively influence other facets
of cognitive function.

Recent epidemiologic studies examined maternal stress as a modifier of the association between
Pb exposure and neurodevelopment. Y Ortiz et al. (2017) used the CRISYS-R questionnaire, which
assesses negative life events across several domains (i.e., financial, legal, career, relationships,
community and home violence, medical problems, other home issues, discrimination or prejudice, and
difficulty with authority) to examine this effect. Third trimester maternal BLL was associated with the
cognitive component of the BSID in this study and a weak interaction (i.e., lower cognitive scores as BLL
and stress increase) between log-transformed maternal blood Pb and stress was observed ([3 = 1.02 [95%

External Review Draft

3-72

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

CI: -0.78, 2.82]). In another study, Zhou et al. (2017) assessed mother-child pairs from the Shanghai
Stress Birth Cohort. Maternal whole blood and maternal prenatal stress levels were assessed at 28-
36 weeks of gestation, and the GDS adapted for a Chinese population were administered to children 24-
36 months old in the study. No association between prenatal maternal BLL and child cognitive
development was observed; however, an interaction effect was observed such that high maternal stress
appeared to exacerbate the effect of prenatal Pb exposure in several domains, including language
((3 = -33.82 [95% CI: -60.04, -7.59] per log-10 transformed unit of BLL), while low maternal stress did
not ((3 = -1.76 [95% CI: -13.03, 9.51] per log-10 transformed unit of BLL, p-interaction = 0.02).

Other Metal Exposure (Cd, Mn, Hg, As)

A limited number of studies included in the 2013 Pb ISA examined the modification of
association between Pb exposure and cognitive function by other metals (U.S. EPA. 2013a'). Larger Pb-
associated decrements in IQ (Kim et al.. 2009) and neurodevelopment (Henn et al.. 2012) were observed
in children with higher Mn levels. Henn et al. (2012) also observed an interaction between the highest
quintile of Mn and BLL at 12 months (Figure 3-10).

Several recent epidemiologic studies of the association of Pb exposure with FSIQ examined
interactions between Pb exposure and other metals or modification of the Pb-FSIQ association by other
metals. For example, some cross-sectional analyses found evidence that coexposure to Mn may heighten
the effect of Pb in some populations (Martin et al.. 2021; Menezes-Filho et al.. 2018). while another study
found no interaction between Pb exposure and Mn (or ALAD) (Lucchini et al.. 2012). Several studies of
the association between Pb exposure and infant development also point to possible interactions with other
metals. Lin etal. (2013) observed an interaction with Mn such that children who were highly exposed to
both Mn and Pb had larger neurodevelopmental deficits compared with those with low exposure to just
one or both these metals. Kim et al. (2013b. 2013c) observed a larger decrement in MDI in association
with late pregnancy maternal BLL among those with Cd levels above the median ([3 = -3.20 [95% CI:
-5.35, -1.06]) compared with the decrement among those with Cd levels below the median ([3 = -0.29
[95% CI: -2.88, 2.30]). In contrast to the findings of Henn et al. (2012). Valeri et al. (2017) observed an
association between increasing cord Pb level and neurodevelopmental decrements in children with lower
cord blood Mn and As ([3 = -0.01 [95% CI: -0.02, 0.00]), but not in the group with higher concentrations
of these metals (or metalloids) ([3 = 0.01 [95% CI: -0.05, 0.07]). Nvanza et al. (2021) also observed
modification of maternal blood Pb and global neurodevelopmental status by blood Hg concentrations.

External Review Draft

3-73

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

o
E

125

105

35

65

45

a

Mn quintiles 1-4

« «

~

• 0

125

105

85

65

45

~i—
10

~i—
15

-1—

20

Mn quintile 5

-1—

10

-1—

15

-1—

20

BPb (iigAIL!

BPb (|ig/dL)

BPb = blood Pb; MDI = Mental Developmental Index; Mn = manganese.
Source: Henn et al. (2012).

Figure 3-10 Scatter plots and regression lines of blood Pb level and 18-month
Mental Developmental Index among children in manganese (A)
quintiles 1-4 and (B) quintile 5.

Only one recent animal study incorporated combined exposure to Pb and Mn at relevant levels
(<30 (ig/dL Pb). Betharia and Maher (2012) reported that both Pb and Mn individually impaired memory
in the Morris water maze, but the effects of the mixture were not significantly different from those of the
control. Interestingly, during the learning (acquisition) phase only, the mixture enhanced performance,
suggesting a possible antagonistic effect of these two metals on the development of spatial learning
processes. Given the lack of evidence available on combined metal exposures, the possible interaction
between multiple metals at relevant levels in animals remains unclear.

Many recent toxicological studies provided evidence for the interaction of Pb and other metals
(Mn, Cd, Ar, Hg, Fe) but were not PECOS-relevant (e.g., in vitro studies, high levels, non-mammalian
models) and are summarized in the biological plausibility Section 3.3

Socioeconomic Status

SES has been examined as an effect modifier in multiple studies of Pb-induced cognitive effects
(U.S. EPA. 2013a. 2006b). Larger blood Pb-associated decreases in cognitive function were found with
lower SES in several studies (Ris et al.. 2004; Tong et al.. 2000; Bellinger et al.. 1990). In contrast, a
meta-analysis of eight studies found a smaller decrement in FSIQ for studies in disadvantaged
populations than for studies in advantaged populations (Schwartz. 1994a). While the results indicate that
BLL is associated with FSIQ deficits in both higher and lower sociodemographic groups, they do not

External Review Draft

3-74

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

clearly indicate whether groups with different SES differ in Pb-related changes for cognitive function
(Murphy et al.. 2013).

No recent epidemiologic studies examined SES as a modifier of the association between Pb
exposure and cognitive effects in children.

Race/Ethnicity

The evidence reviewed in the 2013 Pb ISA pertaining to the modification of the effect of Pb
exposure on cognitive function in children by race or ethnicity was limited to one study (U.S. EPA.
2013a). Miranda et al. (2007) presented data indicating that the association between early childhood
exposure to Pb and declines in reading and mathematics scores was similar between black and white
children. In a recent study, Braun et al. (2018) examined the effects of residential exposure interventions
on dust Pb loadings, BLL, and neurodevelopmental outcomes in children (4 to 8 years old). Although no
intervention effect on BLL was found, overall, the geometric mean childhood BLLs for children 1 to
8 years old was lower in non-Hispanic black children (See Appendix 2:

https://cfpub.epa.gov/ncea/isa/recordisplav.cfm?deid=357282). No intervention effect on other
neurodevelopmental outcomes (e.g., FSIQ, BSID, BRIEF) were observed.

Pre-existing Disease

Studies that examined the effect of Pb exposure on cognitive function in children across strata
defined by pre-existing disease status were not reviewed in previous assessments (U.S. EPA. 2013a.
2006b).

A recent study examined the association of Pb exposure with IQ and executive functioning using
BRIEF among children with CKD (Ruebner et al.. 2019). Concurrent BLL assessment was associated
with FSIQ decrement in adjusted models ((3 = -2.1 [95% CI: -3.9, -0.2] per 1 (ig/dL increase in BLL).
Associations between BLL and behavioral symptoms indicating executive function problems did not
persist in models that controlled for potential confounders including race, poverty, maternal education,
and clinical factors related to CKD.

Nutritional Factors

The 2006 Pb AQCD included studies that indicated individuals with Fe deficiency and
malnourishment had greater inverse associations between Pb and cognition (U.S. EPA. 2006b);
nutritional factors were not examined as effect modifiers of the association between Pb exposure and
cognitive effects in children in more recent studies reviewed in the 2013 Pb ISA (U.S. EPA. 2013a).

External Review Draft

3-75

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

No recent epidemiologic studies were available to inform this topic. Recent toxicological studies
(that were PECOS-relevant) investigated the influence of different dietary factors on the effects of Pb. Liu
et al. (2022c) exposed rats to 0.2% Pb in drinking water in combination with either standard rodent chow
or a high-fat diet and then assessed cognitive function using the Morris water maze paradigm. High-fat
diet increased the BLL of rats compared with Pb-exposed rats maintained on standard chow. High-fat diet
enhanced the effect of Pb on learning during the acquisition phase compared with the Pb-control diet
group. During the probe trial, Pb-exposed animals (both diets) had significantly fewer crossings into the
target quadrant compared with untreated animals. Interestingly, a similar memory impairment was
observed in the non-Pb + high-fat diet group, suggesting a role for high-fat diets in cognitive impairment
independent of Pb exposure. The contribution of Pb versus high-fat diet remains unclear based on this one
study.

Al-Oahtani et al. (2022) supplemented Pb exposure in mice with green tea extract and reported
that green tea ameliorated the negative effects of Pb exposure on both learning and memory assessed in
an active avoidance paradigm. Additionally, Long et al. (2022) reported that probiotic supplementation
(Limosilactobacillus fermentum) in Pb-exposed rats partially mitigated the cognitive deficits observed in
an active avoidance paradigm. These studies support a role for dietary factors in the neurotoxicity of Pb
but the diversity of nutritional factors investigated and the small number of studies make it difficult to
determine their importance.

Genetics

Polymorphisms in certain genes have been implicated in the absorption, retention, and
toxicokinetics of Pb in humans (U.S. EPA. 2013a. 2006b). Studies assessed in the 2013 Pb ISA indicated
that the presence of ALAD variants was associated with an increase in Pb-related cognitive effects in
adults, but there was limited information for children. In studies of children, inverse associations with
poorer rule learning and reversal, spatial span, and planning were exacerbated among those lacking the
DRD4 gene (Froehlich et al.. 2007). Two additional studies found no evidence that the
methylenetetrahydrofolate reductase 677T allele or variants of the DRD2 or dopamine transporter
(DAT1) genes modified the effect of Pb on neurodevelopment (Kordas et al.. 2011; Pilsner et al.. 2010).

Several recent studies add to the limited body of evidence in children. Bah et al. (2022) found that
the effect of low Pb exposure on children's IQ was less among those with the ALAD1 genotype. Roonev
et al. (2018) found interaction effects between variants of glutamate ionotropic receptor NMDA-type
subunits 2A and 2B (GRIN2A and GRIN2B) and Pb exposure on performance on tests of learning,
memory, and executive function, which were more pronounced in boys. Kordas etal. (2011) found that
children with the DRD2 TT genotype (variant) scored higher than children with CC genotype (wild type)
on the Bayley MDI and McCarthy memory scale. However, the variants did not modify the relationship
between BLLs and MDI or McCarthy memory scale scores. Bozack et al. (2021) found that prenatal Pb
exposure was associated with DNA methylation in regions annotated to genes involved in

External Review Draft

3-76

DRAFT: Do not cite or quote

-------
1	neurodevelopment. Overall, the evidence pertaining to interactions between genes and Pb exposure in

2	children remains limited.

Other Factors

3	No studies that examined maternal smoking as a modifier of the association between Pb exposure

4	and cognitive effects were included in previous assessments (U.S. EPA. 2013a. 2006b). Recent studies

5	did not examine maternal smoking as a modifier of the association between Pb exposure and cognitive

6	effects in children.

7	BMI was not examined as an effect modifier of the association between Pb exposure and

8	cognitive effects in children in studies reviewed in the 2013 Pb ISA (U.S. EPA. 2013a). No recent studies

9	have examined this factor as an effect modifier.

10	Maternal self-esteem modified the association between BLL and infant development in Surkan et

11	al. (2008). which was assessed in the 2013 Pb ISA. No recent epidemiologic studies were available to

12	inform this topic.

13	Cognitive reserve was not examined as an effect modifier of the association between Pb exposure

14	and cognitive effects in children in studies reviewed in the 2013 Pb ISA (U.S. EPA. 2013a). No recent

15	epidemiologic studies were available to inform this topic.

3.5.1.7 Integrated Summary and Causality Determination: Cognitive Effects in
Children

16	The evidence from epidemiologic and experimental evidence that supports the causality

17	determination for cognitive effects in children is outlined in Table 3-2. Overall, recent evidence supports

18	the conclusion from the 2013 Pb ISA that there is a causal relationship between Pb exposure and

19	cognitive effects in children.

20	Studies evaluated in the 2013 Pb ISA found a consistent pattern of associations between higher

21	BLLs and lower FSIQ in children aged 4-17 years (see Figure 4-2 and Table 4-3 (U.S. EPA. 2013a)). The

22	strongest evidence was provided by prospective studies with analyses of the association of blood or tooth

23	Pb levels measured in early childhood before FSIQ was assessed. These prospective studies typically

24	considered potential confounding by maternal IQ and education, SES, birth weight, smoking exposure,

25	parental caregiving quality, and in a few cases, other birth outcomes and nutritional factors. Associations

26	were found in diverse populations (e.g., Boston, MA; Cincinnati, OH; Rochester, NY; Cleveland, OH;

27	Mexico City, Mexico; Port Pirie, Australia; and Kosovo, Yugoslavia) in studies that examined children

28	recruited from prenatal clinics, hospital maternity departments, or schools. Studies generally reported

29	high follow-up participation supported by evidence that selection bias did not explain the associations

30	observed.

External Review Draft

3-77

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

Multiple recent longitudinal studies of children with mean BLLs <5 (ig/dL add to the evidence
informing the relationship between BLL and IQ in children. Heterogeneity in the magnitude and direction
of the associations was present across these studies, however. More specifically, associations were
observed in boys but not in girls in several studies (Tatsuta et al.. 2020; Desrochers-Couture et al.. 2018;
Taylor et al.. 2017). There was also some indication that the heterogeneity across studies could be
explained by modeling choices such as confounder adjustment for other metals. For example, cross-
sectional analyses found evidence that exposure to Mn may modify the association between Pb exposure
and IQ in some populations Martin et al. (2021); (Mcnczcs-Filho et al.. 2018). However, studies that
adjusted for multiple metals (e.g., Mn, Hg, Cd, and Pb) in regression models, without examining the
interaction between metals, found little evidence of an association between cord or postnatal BLL and IQ
(Zhou et al.. 2020b; Liu et al.. 2015). imprecise associations only in boys (Tatsuta et al.. 2020;
Desrochers-Couture et al.. 2018). or large IQ decrements after adjustment for Mn, Hg, and ADHD rating
score Hong et al. (2015). Overall, recent studies generally corroborated the epidemiologic observations of
associations between Pb exposure and IQ in children with relatively low blood Pb concentrations
(<5 (ig/dL) among some groups of children (see Section 3.5.1.1). Consistent with findings from the 2013
Pb ISA, studies continue to report associations with prenatal BLL (maternal and cord blood Pb) and
postnatal BLLs measured at various childhood lifestages despite the aforementioned heterogeneity at
BLLs <5 (ig/dL. Overall, the heterogeneity did not weaken the larger body of supporting evidence.

In the review of the MDI evidence in the 2013 Pb ISA, emphasis was placed on results from
examinations at ages 2-3 years, which incorporate test items more similar to those in school-age IQ tests.
Among these studies, several included children with mean BLLs less than 5 (ig/dL (Henn et al.. 2012;
Jedrvchowski et al.. 2009b; Hu et al.. 2006; Bellinger et al.. 1987). Most of the prospective studies
reviewed in previous ISAs (U.S. EPA. 2013a. 2006b) found associations of higher prenatal (cord and
maternal BLL), earlier infancy, and concurrent BLL with lower MDI scores in children aged 2 to 3 years
(Figure 3-10). These blood Pb-associated decrements in MDI were observed in populations with mean
BLLs of 1.3 to 7.1 (ig/dL. Studies typically recruited participants before or at birth without consideration
of Pb exposure or maternal IQ and reported high to moderate follow-up participation as well as
nondifferential loss-to-follow-up. Most studies adjusted for birth outcomes, maternal IQ, and education.
Cord BLLs were associated with MDI, with additional adjustment for SES and HOME score in the
Boston cohort (Bellinger et al.. 1987) and for HOME score in the Yugoslavia cohort (Wasserman et al..
1992). Some studies found a stronger association of MDI with prenatal than child postnatal BLLs ((Hu et
al.. 2006; Gomaa et al.. 2002; Bellinger et al.. 1987).

Recent studies continue to support associations between Pb exposure measured during prenatal or
childhood lifestages and poorer performance on tests of neurodevelopment, among mothers and infants
with mean BLLs <5 (ig/dL (i.e., maternal (Y Ortiz et al.. 2017; Vigeh et al.. 2014; Kim et al.. 2013b. c),
cord (Valeri et al.. 2017). and postnatal (Lin et al.. 2013) BLLs). Although Zhou et al. (2017) found no
association overall, this study reported decrements on several domains of the GDS among infants of
mothers reporting high maternal stress. Similarly, Y Ortiz et al. (2017) found some evidence of

External Review Draft

3-78

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

interactions between Pb exposure and maternal stress..Several studies found interactions between Pb and
Mn or Mn and As (Valcri et al.. 2017; Lin et al.. 2013; Henn et al.. 2012) or Cd exposure (Kim ct al..
2013b); Kim et al. (2013c). Overall, recent studies support findings from the previous reviews and extend
the evidence pertaining to modification of the association between Pb exposure and infant
neurodevelopment by maternal stress and exposure to other metals. The MDI and other tests that measure
neurodevelopment in infants and toddlers are not intelligence tests. Notably, MDI scores, particularly
before ages 2-3 years, are not necessarily strongly correlated with later measurements of FSIQ in children
with normal development and thus, are not weighted heavily in the consideration of causality (U.S. EPA.
2013a).

Experimental animal studies evaluated in the 2013 Pb ISA demonstrated that prenatal and early
postnatal or lifetime Pb exposure alters brain development via changes in synaptic architecture and
neuronal outgrowth, leading to impairments in memory and learning (Sections 4.3.10.4, 4.3.10.10, and
4.3.2.3 of U.S. EPA (2013a)). A small number of recent experimental animal studies were designed to
compare exposures across multiple different developmental windows (Barkur and Bairv. 2015b; Xiao et
al.. 2014); these studies reported similar magnitudes of effects between developmental windows,
suggesting that individual periods of development may be similarly sensitive to Pb. Generally, longer
exposures that spanned multiple developmental periods (e.g., preconception through lactation) produced
not only the highest BLLs but the largest effects on cognition (Zhou et al.. 2020a; Zhu et al.. 2019b).
Overall, these studies provide strong support for observations in epidemiologic studies that Pb exposure
during the prenatal, childhood, and adolescent lifestages is associated with cognitive effects. Recent
animal studies also provide evidence to support the observation that development (i.e., preconception,
gestation, lactation) may represent a critical window for Pb exposure to cause cognitive impairment later
in life (see Section 3.6.1). In rodents, developmental exposure to Pb was consistently associated with
persistent cognitive effects observed both early (Tartaglionc et al.. 2020; Zhao et al.. 2018; Barkur and
Bairv. 2015b; Anderson et al.. 2012) and later in life (Liu et al.. 2022c; Xiao et al.. 2014; Bctharia and
Maher. 2012).

Learning, memory, and executive function are domains of cognitive function that are related to
intelligence, and several are evaluated in the subtests of FSIQ. Additionally, indices of memory, learning,
and executive function are comparable to endpoints examined in experimental animal studies. The studies
evaluated in the 2006 Pb AQCD and 2013 Pb ISA did not clearly indicate associations between higher
BLL and poorer performance on neuropsychological tests of memory or learning (U.S. EPA. 2013a). The
ascertainment of the outcomes varied across studies, potentially explaining the heterogeneity of the
epidemiologic observations. Notably, evidence for both memory and learning decrements from
prospective analyses of several established cohorts (i.e., Rochester, Boston, and Cincinnati) was mixed
(Canfield et al.. 2004; Ris et al.. 2004; Stiles and Bellinger. 1993; Bellinger et al.. 1991; Dietrich et al..
1991). Cross-sectional studies included in the previous ISA, however, generally found associations
between higher concurrent BLLs and poorer learning and memory. Several recent studies of children with
mean BLLs <5 (ig/dL add to the evidence informing the association of Pb exposure with performance on

External Review Draft

3-79

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

tests of memory and learning; however, these recent studies do not enhance the consistency of the
evidence as a whole. Some of the available studies consider co-exposure to other chemicals and metals as
confounders (Oppcnhcimcr et al.. 2022; Tatsuta et al.. 2014) despite evidence that such co-exposures may
interact with or modify the association between Pb and the outcomes (Yorifuji et al.. 2011). Several
recent studies of rodents with exposure resulting in mean BLLs <30 (ig/dL add to the evidence indicating
coherence between the epidemiologic and toxicological findings pertaining to learning and memory
observed in the 2013 Pb ISA.

Strong evidence of associations between Pb exposure and indices of executive function was
described in the 2013 Pb ISA. Studies included prospective analyses of several birth cohorts with
moderate to high follow-up rates in Boston and Rochester that examined BLLs before the outcome
assessment and adjusted for several potential confounding factors (Canfield et al.. 2004; Canfield et al..
2003b; Bellinger etal.. 1994a; Stiles and Bellinger. 1993). Recent studies relying on parent or teacher
behavioral ratings on BRIEF did not generally report associations. The previous ISA did not incorporate
any evidence of the relationship between Pb exposure and executive function in animal models. Recent
studies from a single laboratory provided evidence that Pb exposure broadly impairs measures of
executive function in a reversal learning paradigm. These effects were sex-specific, with greater effects
reported in males. While these reports are consistent with one another, evidence for the association
between Pb exposure and impaired executive function in animal models with BLLs <30 (ig/dL remains
limited.

As described in Sections 3.5.1.1 and 3.5.1.2 and summarized above, heterogeneity in the
epidemiologic results for FSIQ and infant development at BLLs <5 (ig/dL may be explained in part by
sex, exposure to other metals, or maternal stress. Experimental animal studies offer some support for the
observations regarding sex and maternal stress. The limited evidence evaluated in the 2013 Pb ISA
(Virgolini et al.. 2008; Yang et al.. 2003; Mcgivern et al.. 1991). combined with recent evidence, provides
more consistent support for the sex-biased effects in both male (Anderson et al.. 2016) and female
(Tartaglionc et al.. 2020; Verma and Schneider. 2017; Anderson et al.. 2012; Betharia and Maher. 2012)
animals. The exact nature and direction of sex-specific effects given the variation in outcomes examined
remains unclear, however. Toxicological studies assessed in the 2013 Pb ISA demonstrated the potential
for Pb and maternal stress to result in dysfunction of the HPA axis (U.S. EPA. 2013a). Recent
toxicological evidence provides further support for the interaction between maternal stress and Pb
exposure during the exposure period (Anderson et al.. 2012; Corv-Slechta et al.. 2012) but it remains
unclear whether stress would positively influence some facets of cognitive function. Given the lack of
evidence available on combined metal exposures, the possible interaction between multiple metals at
relevant levels in animals remains unclear.

Poorer academic performance and achievement is linked with lower FSIQ and may have
important implications for success later in life (U.S. EPA. 2013a). Associations of higher blood and tooth
Pb levels in children aged 5-18 years with poorer performance on tests of math, reading, and spelling

External Review Draft

3-80

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

skills, lower probability of high school completion and lower-class rank, and lower teacher ratings of
academic functioning were observed in previous assessments (U.S. EPA. 2013a). Recent studies in
populations of children (age 6-16 years) enrolled in school districts including North Carolina, Detroit,
and Chicago with BLLs <5 (ig/dL support and extend these observations of poorer academic performance
in association with increasing Pb exposure in populations with mean BLLs <5 (ig/dL.

Recent studies support and extend the evidence pertaining to the effect of Pb exposure on
cognitive function in children at low BLLs. Compelling evidence for a larger decrement in cognitive
function per unit increase in blood Pb among children with lower mean blood Pb concentrations,
compared with children with higher mean blood Pb concentrations, was supported by a reanalysis of a
pooled international dataset (Crump et al.. 2013). Crump et al. (2013) also supported the finding of
Lanphear et al. (2019) that the C-R function is adequately modeled as linear at BLLs <10 (ig/dL. Recent
studies with an adequate range of Pb exposure measured during relevant time periods that would be
required to evaluate exposure-response relationships were generally lacking, however. Considering the
collective body of studies, no evidence of a threshold for cognitive effects in children across the range of
BLLs examined in epidemiologic studies was reported.

The total body of evidence evaluated in this and previous assessments is sufficient to
conclude that there is a causal relationship between Pb exposure and decrements in cognitive
function in children. This conclusion reflects the consistency of the results from epidemiologic studies
of FSIQ, Bayley MDI, and academic performance and achievement, as well as the coherence of evidence
across epidemiologic and toxicological studies of learning and memory. Biological plausibility is
provided by studies that describe pathways involving the interaction of Pb with cellular proteins, in some
cases competing with and displacing other biologically relevant cations. This interaction leads to
increased oxidative stress and the presence of inflammation, which can have widespread effects on brain
structure and function, as well as disruptions of Ca2+ signaling. These disruptions can result in altered
brain signaling and contribute to the development of neurological health effects. Consistent evidence of
associations between Pb exposure and Bayley MDI provides additional support for this conclusion.

Recent studies support the conclusion of the 2013 Pb ISA that Pb-associated cognitive effects in children
occur in populations with mean BLLs between 2 and 8 (ig/dL. As noted in the 2013 Pb ISA, this
conclusion was based on studies that examined early childhood BLLs (i.e., age <3 years), considered peak
BLLs in their analysis (i.e., peak <10 (.ig/dL). or examined concurrent BLLs in young children (i.e., age
4 years). Some recent studies found associations of Pb exposure with cognitive effects in children with
mean BLLs <2 (ig/dL; however, the studies with mean BLLs <2 (ig/dL lack the aforementioned attributes
(i.e., early childhood BLLs, consideration of peak BLLs, or examination of concurrent BLLs in young
children) and exhibit heterogeneity in both the magnitude and direction of the associations at the lowest
blood Pb concentrations. The observed heterogeneity may be explained in part by the underlying
distribution and complex relationship between covariates in the populations studied, including sex,
maternal stress, and co-exposures to other metals and neurotoxic chemicals, at relatively low BLLs (<5
(ig/dL). Epidemiologic and toxicological studies continue to strongly support the finding that exposure

External Review Draft

3-81

DRAFT: Do not cite or quote

-------
1	during multiple lifestages (prenatal through adolescence and early adulthood) is associated with cognitive

2	effects in children. Recent studies extend the evidence indicating that early life exposures are associated

3	with cognitive effects in adulthood. No evidence of a threshold for cognitive effects in children across the

4	range of BLLs examined in epidemiologic studies was reported.

External Review Draft

3-82

DRAFT: Do not cite or quote

-------
Table 3-2 Summary of evidence Indicating a causal relationship between Pb exposure and cognitive effects
in children.

Rationale for Causality	Key Evident	References'3	Levels Associated

Determination	wjth Effectsc

Consistent associations from multiple,
prospective epidemiologic studies with
relevant BLLs

Evidence from prospective studies for decrements
in FSIQ in association with prenatal, earlier
childhood, peak, concurrent, lifetime average BLLs
and tooth Pb levels in children ages 4-17 yr in
multiple U.S. locations, Mexico, Europe, Australia.

U.S. EPA (2013a)

Section 4.3.2.1, Table 4-3

Blood Pb (various
time periods and
lifestages): Means 3-
16 [jg/dL

Recent prospective studies observe associations of Section 3.5.4.1
Pb with FSIQ; however, heterogeneity in the
magnitude and direction of the associations is
present.

Blood Pb (various
time periods and
lifestages) <5 [jg/dL
(<2 [jg/dL in some
studies)

Some recent epidemiologic studies indicate
potential effect modification or interactions of Pb
with sex, other metals, and maternal stress
potentially explaining heterogeneity in the observed
associations at BLLs <5 [jg/dL.

Section 3.5.1.6.5

Evidence from prospective studies for lower scores U.S. EPA (2013a)

on tests of executive function and academic

performance in association with earlier childhood or

lifetime average BLLs or tooth Pb levels in children

ages 5-20 yr in multiple U.S. locations, U.K., New

Zealand.

Recent evidence generally relies on outcome
ascertainment based on the BRIEF is inconsistent.

The direction and magnitude of associations were
not consistent for learning and memory.

Recent evidence does not enhance the
consistency.

Section 3.5.1.4

U.S. EPA (2013a)
Section 3.5.1.3

Blood Pb (various
time periods and
lifestages) <5 [jg/dL

External Review Draft

3-83

DRAFT: Do not cite or quote

-------
Rationale for Causality	Key Evidence13	References'3	Levels Associated

Determination3	wjth EffectsC

Supporting evidence from cross-sectional studies of U.S. EPA (2013a)
children ages 3-16 yr, but most did not consider
potential confounding by parental caregiving
quality. Includes large NHANES III analysis.

Blood Pb (various
time periods and
lifestages) <5 [jg/dL

Several studies indicate supralinear C-R	U.S. EPA (2013a)

relationship, with larger decrements in cognitive Section 3 5 16 1
function per unit increase in blood Pb at lower BLLs
in children ages 5-10 yr
Reanalysis of international pooled analysis
substantiates this finding.

Crump et al. (2013)

Epidemiologic evidence helps rule out Several epidemiologic studies found associations U.S. EPA (2013a)
chance, bias, and confounding with with adjustment for SES, maternal IQ and	Section 3 5 16 2

reasonable confidence	education, HOME score. Several adjust for birth

weight, smoking. A few, nutritional factors.

Experimental animal studies with
relevant exposures provide coherence
and help rule out chance, bias, and
confounding with reasonable
confidence

Impaired learning and associative ability in juvenile
and adult animals as indicated by performance in
tasks of visual discrimination, water maze, y maze,
and operant conditioning with schedules of
reinforcement with relevant dietary Pb exposure.

Recent studies of executive function in rodents add
to the evidence; however, evidence for impaired
executive function in animal models with BLLs
<30 [jg/dL remains limited.

U.S. EPA (2013a)
Section 4.3.2.3

Section 3.5.1.4.2

Blood Pb (after
prenatal/ lactation,
lactation only,
prenatal/lifetime Pb
exposure): 10-
25 [jg/dL

Experimental animal studies with	Recent studies in rodents suggest that factors such Section 3.5.1.6.5

relevant exposures provide coherence as sex and maternal stress may influence the

for epidemiologic observations of effect effects of Pb on cognitive function.

modification by sex or interactions of Pb

with other metals or maternal stress

External Review Draft

3-84

DRAFT: Do not cite or quote

-------
Rationale for Causality
Determination3

Key Evidence13

References'3

Pb Biomarker
Levels Associated
with Effects0

Biological plausibility demonstrated Pathways involving oxidative stress, inflammation U.S. EPA (2013a)

and Ca2+ signaling result in impaired neuron	Section 3 6

development, synaptic changes, LTP, and
neurotransmitter changes.

Recent studies support and extend findings related Section 3.3
to overt nervous system effects.

BLL = blood lead level; BRIEF = Behavior Rating Inventory of Executive Functions; Ca2+ = calcium ion; C-R = concentration-response; FSIQ = full-scale intelligence quotient;

HOME = Health Outcomes and Measures of the Environment; IQ = intelligence quotient; LTP = long-term potentiation; NHANES = National Health and Nutrition Examination Survey;
Pb = lead; SES = socioeconomic status; yr = year(s).

aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Table I and Table II of the Preamble to the ISAs (U.S. EPA. 2015).
bDescribes the key evidence and references, supporting or contradicting, contributing most heavily to causality determination and, where applicable, to uncertainties or
inconsistencies. References to earlier sections indicate where the full body of evidence is described.

°Describes the Pb biomarker levels at which the evidence is substantiated.

External Review Draft

3-85

DRAFT: Do not cite or quote

-------
3.5.2 Externalizing Behaviors: Attention, Impulsivity, and Hyperactivity in
Children

The evidence evaluated in the 2013 Pb ISA was sufficient to conclude that there is a "causal
relationship" between Pb exposure and effects on attention, impulsivity, and hyperactivity in children.
Several prospective studies demonstrated associations of blood or tooth Pb levels measured years before
outcomes with attention decrements and hyperactivity in children (7-20 years) as assessed using objective
neuropsychological tests and rated by parents and teachers. Most of the prospective studies examined
representative populations with no indication of participation that was conditional on BLLs and behavior.
The results from the prospective studies were generally adjusted for potential confounding by SES as well
as parental education and caregiving quality, with some studies also considering parental cognitive
function, birth outcomes, substance abuse, and nutritional factors. With respect to the timing of exposure
discerned from prospective studies, blood Pb-associated attention decrements and hyperactivity were
found in populations with prenatal (maternal or cord) or postnatal (i.e., 3-60-month average, age 6 years,
or lifetime average through age 11-13 years) mean BLLs of 7 to 14 (ig/dL and in groups with BLLs
>10 (ig/dL at 30 months of age. Biological plausibility for these observations in children was provided by
experimental animal studies that demonstrated increases in impulsivity or impaired response inhibition
with relevant postweaning and lifetime Pb exposures that resulted in BLLs of 11 to 30 (ig/dL.
Demonstrated Pb-induced impairments in neurogenesis, synaptic pruning, and dopamine transmission in
the prefrontal cerebral cortex, cerebellum, and hippocampus also supported the biological plausibility of
the associations observed in the epidemiologic studies. Although coherence across and within lines of
evidence was demonstrated, the small number of studies of diagnosed ADHD were limited by their cross-
sectional or case-control design, inconsistent adjustment for SES and parental education, and lack of
consideration for potential confounding by parental caregiving quality.

There are three major domains of externalizing behavior disorders: (1) ADHD, (2)
undersocialized aggressive conduct disorder, and (3) socialized aggressive conduct disorder (as reviewed
in (Whitcomb and Merrell. 2012)). Although these domains are interrelated, to the extent possible, this
section (3.5.2) will maintain a similar structure as the 2013 Pb ISA by focusing on the ADHD domain,
which encompasses characteristics including but not limited to short attention span, distractibility,
impulsivity, and hyperactivity. Within the ADHD domain of externalizing behaviors, most epidemiologic
studies of Pb exposure focus on attention, impulsivity, and hyperactivity. Some epidemiologic studies
examined composite indices of multiple behaviors, and a few studies have examined physician-diagnosed
ADHD. Domain-specific neuropsychological assessments of attention, impulsivity, and hyperactivity
with strong psychometric properties and rigorous validation were emphasized in the 2013 Pb ISA and
provide the strongest evidence for the causality determination. Studies that evaluated the association of Pb
exposure with externalizing behaviors assessed using teacher and parent ratings contributed to the overall
evidence, but the limitations of these studies were noted. Specifically, these studies were subject to

External Review Draft

3-86

DRAFT: Do not cite or quote

-------
greater measurement errors compared with studies of outcomes assessed through neuropsychological
testing.

Control for confounding is considered an attribute of a well-conducted, high-quality study.
Greater weight is given to studies that consider important potential confounders in their design or
statistical analyses. As noted in the 2013 Pb ISA (U.S. EPA. 2013a). associations between Pb biomarker
levels and externalizing behaviors may be confounded by parental SES, education, and IQ; nutritional
status; and the quality and stability of the caregiving environment (often evaluated using the HOME score
(Totsika and Svlva. 2004)). The research available for evaluation in the 2013 Pb ISA did not establish a
direct relationship between parental psychopathology and child Pb exposure or one between parental
psychopathology and poorer parental caregiving quality. Thus, parental psychopathology itself was not
considered to be a potential confounder of associations between child Pb and externalizing behaviors.
Although parental psychopathology was hypothesized to modify the association between Pb exposure and
externalizing behavior, no studies available for evaluation in the 2013 Pb ISA evaluated parental
psychopathology as an effect modifier. To the extent that parental psychopathology could affect child Pb
exposure indirectly through parental caregiving quality, however, it was noted that confounder control
would be achieved in studies that included an adjustment for the HOME score or similar metrics.

Greater emphasis is also placed on prospective studies with repeated assessments of BLLs and
studies of children with BLLs that are less influenced by higher past Pb exposures (i.e., younger children).
Studies assessing effects in populations with BLLs that are most relevant to current U.S. children (e.g.,
<5 (ig/dL) are also emphasized, e.g., (Cho et al.. 2010; Nicolescu et ak. 2010; Chandramouli et ak. 2009;
Nigg et ak. 2008; Chen et ak. 2007; Chiodo et ak. 2007). In the current ISA, when considering the causal
relationship of Pb exposure with attention, impulsivity, and hyperactivity, PECOS statements (see
Section 3.2) were refined to focus on the most informative studies. Longitudinal epidemiologic studies
with mean (or central tendency) BLLs <5 (ig/dL are highlighted in the text as are the most reliable
biomarkers of Pb exposure (i.e., blood, bone, teeth, or nails). Consideration of potential confounding and
modification of the observed associations by the aforementioned factors was evaluated when considering
the overall quality of the study. Measures of central tendency for Pb biomarker levels used in each study,
along with other study-specific details, including study population characteristics and select effect
estimates, are highlighted in evidence inventory Table 3-7E (Epidemiologic Studies) and Table 3-7T
(Toxicological Studies). In addition, studies with central tendency blood Pb concentrations that exceed
5 |ig/dk are extracted into Table 3-8E of Section 3.7 (Evidence Inventories). An overview of the recent
evidence is provided below. Overall, recent studies generally support findings from the 2013 Pb ISA.

External Review Draft

3-87

DRAFT: Do not cite or quote

-------
3.5.2.1

Attention in Children

3.5.2.1.1 Epidemiologic Studies of Attention in Children

Attention is the ability to maintain a consistent focus on an activity or relevant stimuli and can be
assessed by examining sustained attention, concentration, or distractibility. The preponderance of
evidence pertaining to the externalizing behaviors included in the 2013 Pb ISA evaluated the association
of Pb exposure with measures of attention (U.S. EPA. 2013a). Most prospective studies found
associations of blood or tooth Pb levels with decrements in neuropsychological tests of attention as well
as parent and teacher ratings of attention. One strength of the prospective studies is that they characterized
the sequence of Pb exposure (i.e., prenatal blood Pb, postnatal blood Pb before the outcome, concurrent,
lifetime average blood Pb, and tooth Pb), establishing the temporal relationship between exposure and
outcome. In addition, the studies reported moderate to high follow-up participation that was not
conditional on blood or tooth Pb levels and controlled for important confounders (i.e., parental education,
IQ, and caregiving quality; SES). Overall, these studies showed a pattern of lower attention with higher
blood or tooth Pb level (see Figure 4-9 and Table 4-11 of the 2013 Pb ISA (U.S. EPA. 2013a')'). Mean
BLLs were generally within the range of 7-14 (ig/dL for most of the prospective studies, and cross-
sectional studies generally supported findings from the longitudinal analyses (see Section 4.3.3.1 of the
2013 Pb ISA (U.S. EPA. 2013a)).

A small number of recent longitudinal epidemiologic studies of children with relatively low BLLs
(i.e., <5 (ig/dL) add to the evidence for associations between Pb exposure and decrements in
neuropsychological tests of attention or parent and teacher ratings of behaviors that indicate attention
problems (see Section 3.5.2.4).

Neugebauer et al. (2015) conducted an analysis of the Duisburg birth cohort data to examine the
association of maternal BLLs at 32 weeks gestation with performance on neuropsychological tests of
attention (Test of Attentional Performance for Children [KiTAP]) and parent-rated ADHD behaviors on
the German Symptom Checklist for ADHD (Fremdbeurteilungsbogen fur

Aufmerksamkeitsdefizit/Hyperaktivitatstorungen [FBB-ADHS]) in childhood. Maternal blood Pb was
most strongly associated with specific KiTAP subtests, i.e., number of omissions (geometric mean ratio
[GMR] = 1.15 [95% CI: 1.00, 1.33]) and reduced performance speed (GMR= 1.14 [95% CI: 0.98, 1.33]).
Maternal blood Pb concentration was positively associated with the inattention component of the FBB-
ADHS indicating that inattention increases with increasing BLL(GMR = 1.05 [95% CI: 0.99, 1.12]).

In a study of a subset (n = 27) of Inuit children (Boucher et al.. 2012b) (see Section 3.5.2.5) that
used a modified Posner paradigm to assess the association between prenatal and concurrent childhood Pb
exposure with visuospatial attention, vigilance, and impulsivity, Ethieretal. (2015) found that concurrent
ln-transformed blood Pb was associated with some tests of attention including longer reaction times
([3 = 0.52 [95% CI: -0.10, 1.14] per SD increase in ln-transformed Pb) in amodel adjusted forage, sex,

External Review Draft

3-88

DRAFT: Do not cite or quote

-------
and current PCB exposure. In another study, Tatsuta et al. (2014) adjusted for PCBs and MeHg (in
addition to maternal IQ and family income) and found no associations with sequential processing score
(-2.14 [95% CI: -12.80, 8.53] per unit of log transformed BLL [base not specified]) or mental processing
score (-3.32 [95% CI: -12.4, 5.77] per unit of log transformed BLL [base not specified])). Yorifuji et al.
(2011) found that cord blood Pb was associated with some neuropsychological tests of attention and
working memory on the WISC-R (i.e., digit span) and that the interaction between cord blood Pb and cord
Hg level may be less than additive. For example, the association of cord Pb with performance on the digit
span forward at age 7 was [3 = -0.11 [95% CI: -0.29, -0.07] per log-transformed unit of BLL without
accounting for the interaction between cord Pb and cord Hg concentration. This association was more
pronounced when Hg exposure was lower. Specifically, the unstandardized associations of log-
transformed cord BLL with the neuropsychological test outcomes were most discernable among children
with hair Hg concentrations below 2.61 jxg/g, which was the lowest cord Hg concentration. For example,
a lower digit span forward score on the WISC-R ([3 = -1.70 [95% CI: -3.12, -0.28]) at age 7 and a lower
digit span backward score on the WISC-R ([3 = -2.73 [95% CI: -4.32, -1.14]) at age 14 were observed
among children with the lowest Hg exposure.

Ruebner et al. (2019) evaluated the association between BLLs and attention and hyperactivity
among children with CKD. Attention was assessed using either Conners' Kiddie Continuous Performance
(K-CPT; 4-5 years) or Conners' CPT II (>6 years), which produce scores for omission and commission
errors, correct detection rate, response variability, reaction time, and summary measures for sustained
attention and inhibitory control. A 1.8 T-score point increase (i.e., worse performance) in CPT variability
(95% CI: 0.2, 3.5), which indicates problems with sustained attention and attention regulation, was
associated with childhood blood Pb (on average, ~2 years before outcome ascertainment) in this study.
Covariates considered as potential confounders included race, poverty, maternal education, and factors
related to CKD. The median BLL in this study was 1.2 (ig/dL.

A small number of studies examined the gene-environment interaction between Pb exposure and
genotypes associated with attention decrements. Roonev et al. (2018) studied children in Lisbon, Portugal
to determine the association of baseline BLL (8-12 years old) and variants of GRIN2A and GRIN2B,
which regulate neurodevelopmental processes, with performance on neuropsychological tests, including
tests of attention, during the 7-year follow-up period. A pattern of association indicating poorer
performance on tests of attention with increasing baseline Pb exposure was not observed. Choi et al.
(2020) enrolled children (5-18 years old) with ADHD and healthy controls without ADHD to evaluate
interactions between Pb exposure and noradrenergic pathway-related genotypes (i.e., [DAT1], dopamine
receptor D4 [DRD4], and alpha-2A-adrenergic receptor [ADRA2A]). ADHD was assessed using the
ADHD rating scale (ADHD-RS) and neuropsychological tests of attention (i.e., CPT and SCWT) were
also administered. BLLs were associated with omission errors ([3 = 3.75 [95% CI; 0.09, 7.40]) in models
adjusted for IQ, age, and sex, which were found to partly mediate the effect of Pb on ADHD symptoms in
a path analysis model. An interaction effect was detected between the ADRA2A Dral genotype and Pb
levels on omission errors ([3 = 5.07 [95% CI: 0.20, 9.93]). Multiple comparisons were made in this

External Review Draft

3-89

DRAFT: Do not cite or quote

-------
analysis (e.g., associations with additional CPT components including commission errors, response time,
and response time variability, ADHD-RS components and SCWT components were not observed),
increasing the likelihood of chance findings.

Summary

Prospective studies in the 2013 Pb ISA showed strong support for an association between pre-
and postnatal Pb exposure (range: 7-14 (ig/dL) and decreased scores on neuropsychological tests and
parent/teacher ratings of attention. Cross-sectional studies from the 2013 Pb ISA corroborated these
observations. A small number of recent studies reported associations of maternal and cord BLLs <5 (ig/dL
with some measures of inattention; however, the results for multiple subtests were reported, potentially
increasing the likelihood of chance findings. In addition, there is uncertainty regarding the patterns of
exposure associated with maternal and cord BLLs. Recent studies add to the limited evidence regarding
co-exposure to Hg and gene-environment interactions (see Section 3.5.2.6.3).

3.5.2.1.2 Toxicological Studies of Attention

In support of the associations described in the preceding sections for BLL with attention
decrements in children, studies have found Pb-induced decreases in attention in animals, although results
have not been consistent across studies. Although tests in animals often measure aspects of both attention
and impulsivity, behaviors measured with signal detection tests with distraction can be inferred as
predominately assessing sustained attention. In this test, animals earn food rewards by responding to a
target stimulus and not responding to a distracting light. Poorer sustained attention and greater
distractibility are indicated by lack of response to the target and increased response to the distracter light,
respectively. The 2006 Pb AQCD (U.S. EPA. 2006a') reported inconsistent effects of Pb exposure in
animals on performance in this test. For example, postweaning Pb exposure that produced BLLs of 16
and 28 (ig/dL induced small decreases in attention in adult rats, as indicated by small increases in
omission and commission errors but only during sessions with long intervals between stimuli (Brockel
and Corv-Slechta. 1999). Lifetime Pb exposure from birth (mean peak BLLs of 15 and 25 (ig/dL for the
50 and 100 (ig/kg/day groups, respectively) was found to induce distractibility in monkeys at age 9-
10 years, as indicated by increased responses to irrelevant cues, i.e., distracting stimuli, in a spatial
discrimination reversal task. Repeated reversal testing revealed that these deficits likely were not due to
sensory or motor impairment (Gilbert and Rice. 1987).

In animals, Pb-induced decrements in attention have been inferred from tests designed to assess
impulsivity but that have elicited behaviors that suggest deficits in attention. For example, a study
reported that impaired performance on auditory threshold tasks in Pb-exposed monkeys was likely due to
lack of attention (Laughlin et al.. 2009). Rhesus monkeys were exposed to Pb acetate from gestation
(drinking water of mothers, 3 months prior to mating) to birth or postnatally from birth to age 5.5 months

External Review Draft

3-90

DRAFT: Do not cite or quote

-------
at weaning, resulting in bone Pb levels of 7 and 13 jxg/g for prenatal and postnatal groups at 11 years of
age, respectively, and average BLLs of 35 and 46 (ig/dL, respectively, during Pb exposure. Animals were
tested at age 13 years when BLLs had returned to baseline levels. The inability of some of the monkeys to
engage or focus attention on the task at hand yielded fewer available measurements in Pb-exposed
animals versus controls. These observations were made in monkeys with higher peak BLLs than those
relevant to this ISA. No recent studies have evaluated attention in animals following exposure that
resulted in BLLs relevant to the current ISA.

3.5.2.2 Impulsivity in Children

3.5.2.2.1 Epidemiologic Studies of Impulsivity in Children

Measures specific to impulsivity were examined in relatively few epidemiologic studies of
children compared with measures of attention, and most studies including evaluations of impulsivity that
were included in the 2013 Pb ISA were cross-sectional in design (U.S. EPA. 2013a). The available
evidence indicated Pb-associated poorer performance on tests of response inhibition. Response inhibition
is a measure of impulsivity and has been assessed in children via stop signal tasks, which measure the
execution of action in response to stimuli and the inhibition of that action when given a stop signal.
Associations of blood and tooth Pb with parent and teacher ratings of impulsivity were also reported.
These studies generally adjusted for potential confounding by SES, sex, parental education, and smoking;
however, parental IQ or caregiving quality was not examined in most studies. The relatively small body
of epidemiologic evidence (see Figure 4-9 and Table 4-11 (U.S. EPA. 2013a)') was coherent with results
from experimental animal studies (see Section 4.3.3.1 (U.S. EPA. 2013a)').

Analyses of Inuit children have been conducted since the 2013 Pb ISA, examining Pb exposures
and impulsivity. Boucher et al. (2012a) examined response inhibition deficits assessed with the Go/No-
Go task and event-related potentials (ERPs) derived from electroencephalogram (EEG) recordings during
task performance among Inuit school children residing in Arctic Quebec. Cord and concurrent blood Pb
concentrations were associated with increased impulsivity after adjustment for covariates including child
age, sex, SES, maternal nonverbal reasoning abilities, and Hg (Figure 3-11). In addition, Ethier et al.
(2015) studied a subset of this population (n = 27) using a modified Posner paradigm to assess the
association between pre- and concurrent childhood Pb exposure with visuospatial attention, vigilance, and
impulsivity. The study found that cord Pb was associated with greater impulsivity ((3 = 0.42 [95% CI:
0.08, 0.76] per SD increase in ln-transformed Pb).

External Review Draft

3-91

DRAFT: Do not cite or quote

-------
100

¦= 90

70

a

[

T M



1







5 70

60

50

a



90

t

[





£

80

cc





[





Z

§5 70

o

c

L

a> 60
o

C_3



50

0.8-2.5 2.5-3.7 3.7-5.6 5.6-20.9
Cord blood Pb ([ig/dL)

0.8-2.5 2.5-3.7 3.7-5.6 5.6-20.9
Cord blood Pb (iig/dL)

Source: Boucher et al. (2012a)

Figure 3-11

0.4-1.3 1.3-2.0 2.0-2.9 2.9-12.8
11-year blood Pb (|ig/dL)

Mean ± standard deviation behavior performance in the Go/No-Go
task according to quartiles of exposure for (A and B) cord blood
Pb and (C) childhood blood Pb level at age 11 years.

Summary

Studies of impulsivity in children in the 2013 Pb ISA were limited by their quantity and lack of
temporality but generally indicated associations of Pb exposure with worse scores on tests of response
inhibition and on parent and teacher ratings of impulsivity. These studies also often lacked confounder
control for parental IQ or caregiving quality, which are key potential confounders. Recent analyses of
Inuit children add support for the relationship between Pb exposure and impulsivity with additional
consideration of potential confounders including maternal nonverbal reasoning abilities (Boucher et al..
2012a).

3.5.2.2.2 Toxicological Studies of Impulsivity

The associations described between higher BLL and greater impulsivity in children are supported
by findings in animals for Pb-induced increases in perseveration and impaired ability to inhibit
inappropriate responses. In animals, these effects are supported by studies reviewed in the 1986 and 2006
Pb AQCDs (US EPA, 2006b, 1986b) and studies incorporated into the 2013 Pb ISA. Animal studies
provide more consistent evidence for the effects of Pb exposure on impulsivity than on sustained
attention. As mentioned earlier, behaviors displayed by animals in a variety of tests can be identified as
reflecting impulsivity. These include tests of differential reinforcement of low rates of responding, fixed
interval (FI) schedule performance, FI with extinction, or fixed ratio (FR)/waiting-for-reward. Greater
impulsivity is indicated by premature responses, decreased pause time between two scheduled events, and
increased perseveration.

External Review Draft

3-92

DRAFT: Do not cite or quote

-------
Behaviors observed in tests of operant conditioning with FI reinforcement schedules have also
been used to indicate impaired learning in animals (Section 3.5.3.2), and the interactions observed
between Pb exposure and maternal or offspring stress may also apply to effects on impulsivity. Maternal
exposure to 150 ppm Pb with and without stress co-exposure was found to increase overall FI rate and
decrease Post-reinforcement Pause (PRP) in rats. Lifetime (from gestation) Pb exposure resulting in BLLs
of 11-16 (ig/dL increased the overall FI rate without stress co-exposure and decreased PRP with stress co-
exposure (Rossi-George et al.. 2011). suggesting that stress may interact with Pb exposure to affect
attention. Discrimination reversal learning has been shown to be affected by Pb exposure. In these tasks,
an animal is trained to choose between two alternative responses and is then required to reverse the
association. Perseveration or lack of inhibition of the original response can be interpreted to involve
impulsivity. Spatial and non-spatial discrimination reversal was significantly affected in monkeys after Pb
exposure during infancy, after infancy, or continuously from birth, and was exacerbated with distracting
stimuli (Rice. 1990; Rice and Gilbert. 1990b; Gilbert and Rice. 1987). These monkeys had BLLs in the
range of 15-36 (ig/dL, which includes values relevant to this ISA. Hilson and Strupp (1997) found Pb
exposure (Pb acetate in drinking water at GD 1-PND 28, yielding a BLL of 26|ig/dL in the lower dose
group) in rats slowed reversal learning in an olfactory discrimination task. However, analysis of the
response patterns showed that Pb exposure shortened the perseverative responding phase of reversal
learning and lengthened the post-perseverative phase of chance responding, indicating impairments in
associative ability, not response inhibition. Thus, it is more likely that Pb negatively affected associative
learning rather than impulsivity in this study (Hilson and Strupp. 1997). which is inconsistent with the FI
data in monkeys. Due to the small number of studies following relevant Pb exposures in rodents, it
remains unclear whether Pb affects impulsivity in FI.

The effects of Pb exposure on impulsivity also have been demonstrated in a study reporting that
Pb-exposed animals wait a shorter period of time for reward in FR/waiting for reward testing. In this test,
animals can obtain food by pressing a lever a fixed number of times (FR component). Free food is then
delivered at increasingly longer time intervals, so long as the animal inhibits additional lever presses.
Animals can reset the schedule to return to the FR component at any time. Brocket and Corv-Slechta
(1998) exposed male Long-Evans rats to 0, 50, or 150 ppm Pb acetate in drinking water from weaning,
which produced respective BLLs of <5, 11, and 29 (ig/dL after 3 months of exposure. After 40 days of
exposure, the 150 ppm Pb-exposed rats responded more quickly in the FR component and reset the
schedule (thus shortening the waiting period) more often than did the 50 ppm Pb-exposed rats and
controls. In the waiting component, average wait time was significantly lower in both Pb exposure groups
compared with controls. The rats exposed to 150 ppm Pb also had higher response rates and earned more
reinforcers per session but had a higher response to reinforcement-ratio than did the 50 ppm Pb group and
controls, which indicated less efficient responses.

Weston et al. (2014) used the delayed discounting paradigm following developmental exposure
with or without prenatal restraint stress. The delayed discounting protocol offered animals the choice
between a large reward after a long delay or a small reward after a short delay. Pb increased long-delay

External Review Draft

3-93

DRAFT: Do not cite or quote

-------
responding, slowed acquisition of delayed discounting performance, and increased failures almost
exclusively in males. Consistent with (Hilson and Strupp. 1997). these results more likely represent
impaired learning or cognitive flexibility rather than simply increased impulsivity.

In summary, several studies in animals indicate that Pb exposure of rodents and nonhuman
primates from birth or after weaning changes behavior in ways consistent with increased impulsivity,
primarily as indicated by impaired response inhibition. It is also important to note that many of the
measures of impulsivity discussed in this section are sensitive to disruption by impairments in learning
and executive function, which is consistent with several of the studies summarized in Section 3.5.3.2.
Some observations of Pb-induced impulsivity in animals were made with BLLs considered relevant for
this ISA. The observations for Pb-induced increases in impulsivity in animals provide support for
associations found in children of higher blood and tooth Pb levels with lower response inhibition and
higher ratings of impulsivity.

Transgenerational Effects of Pb on Impulsivity

The paradigm of combined Pb and stress exposure experienced by a laboratory animal has been
examined with a focus on the common pathway of altered HPA axis and brain neurotransmitter levels.
Studies investigating the interactions between Pb and prenatal stress on learning and memory are
reviewed in Section 3.5.1.3.2. These findings were expanded on in a recent study that investigated the
transgenerational effects of combined Pb and prenatal stress in mice (Sobolewski et al.. 2020). The
authors reported that Pb exposure in the gestating female (F0 generation) resulted in sex-specific effects
in the third filial (F3) generation (no direct exposure to Pb), with F3 females displaying significantly
elevated response rates in an FI schedule of reward compared with control lineages, suggesting an
impulsive behavioral phenotype (Sobolewski et al.. 2020). This and other transgenerational effects were
accompanied by Pb-induced alterations in neurotransmitters, BDNF expression, and DNA methylation.
The authors postulated that lineage effects may be mediated through some combination of maternal
responses to pregnancy, maternal behavior, or epigenetic modifications (Sobolewski et al.. 2020). While
these findings were limited to a single study, they support the possibility that exposure to Pb may
influence the behavior of subsequent generations.

3.5.2.3 Hyperactivity in Children

3.5.2.3.1 Epidemiologic Studies of Hyperactivity in Children

Studies reviewed in the 2006 Pb AQCD (U.S. EPA. 2006b) indicated associations between higher
concurrent BLLs or tooth Pb levels and higher parent or teacher ratings of hyperactivity in children aged
6-11 years in the U.S., Asia, and New Zealand (Rabinowitz et al.. 1992; Silvaetal.. 1988; Gittleman and
Eskenazi. 1983; Needleman et al.. 1979; David et al.. 1976). The case-control or cross-sectional design of

External Review Draft

3-94

DRAFT: Do not cite or quote

-------
studies limited understanding of the temporal sequence between Pb exposure and hyperactivity. A
prospective study (Chandramouli et al.. 2009) included in the 2013 Pb ISA also found associations
between BLL and hyperactivity as rated by teachers and parents. Overall, studies indicated associations in
children 3-12 years old with mean concurrent BLLs of 3.7-12 (ig/dL. Studies of recent hyperactivity
symptoms as rated by parents and teachers are discussed in Section 3.5.2.4.

3.5.2.3.2 Toxicological Studies of Hyperactivity

The 2006 Pb AQCD (U.S. EPA. 2006c') reviewed the evidence that developmental exposure to Pb
could affect locomotor activity in laboratory animals. Findings summarized in this document included
four studies showing increased activity with developmental Pb exposure and three studies showing no
change in activity. The 2013 Pb ISA (U.S. EPA. 2013a) only described one new study in this category,
which showed a decrease in activity in mice after maternal Pb exposure (Leasure et al.. 2008). Effects of
developmental Pb exposure on rodent locomotor activity are commonly assessed using an open-field test.
The activities examined vary across studies (e.g., distance traveled, counts of square crossings). Because
there are myriad potential explanations for changes in rodent activity, it can be difficult to draw
conclusions from "simple" tests like open-field. The results from such tests are best interpreted alongside
additional behavioral assays, which, together, may better model the complexity of human behavior.
Conclusions for an effect of developmental Pb exposure on locomotor activity were not reached in earlier
U.S. EPA Pb reviews due to mixed results in these tests. Studies described in the 2013 Pb ISA (U.S. EPA.
2013a) and 2006 Pb AQCD (U.S. EPA. 2006c) as observing the effects of maternal Pb exposure, which
resulted in mean BLLs no higher than 30 (ig/dL, are Munoz et al. (1989). Rodrigues et al. (1996). Moreira
et al. (2001). Trombini et al. (2001). De Marco et al. (2005). and Leasure et al. (2008).

Recent studies (see Table 3-1 IT) observed the activity of early postnatal rodents after
developmental exposures to Pb with varying durations. Tartaglione et al. (2020) exposed rats to Pb from
pregestation to offspring weaning and tested offspring. They observed a decrease in neonatal spontaneous
activity on PND 10 and no change in spontaneous activity on PND 4, 7, and 12. Another group reported
that two groups of CD 1 mice exposed to either a high or low dose of Pb through lactation exhibited
hyperactivity (PND 7, 11, 15, 19) compared with controls in open-field testing Duan etal. (2017).
Interestingly, when locomotor data from early postnatal studies are pooled, nine out of nine sets of
lactationally exposed animals (BLLs: 9.6-28.9 (ig/dL) tested between PND 14 and 23 were hyperactive
(Duan et al.. 2017; De Marco et al.. 2005; Moreira et al.. 2001; Rodrigues et al.. 1996). These sets
consisted of both male and female rodents, except for one set of only males in Moreira et al. (2001).

Evidence inventory (Section 3.7) Table 3-1 IT also includes recent studies that monitored the
effects of developmental exposure to Pb on immature rodents postweaning. Basha and Reddv (2015)
found decreased locomotor activity in male rats with gestational exposure to Pb when tested on both
PND 21 and PND 28. Betharia and Maher (2012) studied open-field behavior in Sprague Dawley rats
with gestational and lactational exposure to Pb. There were no differences in total square crossings for

External Review Draft

3-95

DRAFT: Do not cite or quote

-------
both males and females tested at both PND 24 and PND 59. Flores-Montova and Sobin (2015) saw no
effects on open-field tasks (PND 28) in two groups each of male and female C57BL/6 mice after
postnatal exposure (PND 0-28) to low levels of Pb acetate. Developmentally Pb-exposed rats from the
Tartaglionc etal. (2020) study described in the previous paragraph exhibited no change compared with
control in open-field activity when tested on PND 30. Neuwirth et al. (2019a) observed no effect on
locomotor activity in open-field tests (PND 36-45) for two groups of Long-Evans rats with different
levels of in utero and lactational exposure to Pb. Zou et al. (2015) reported that exposure to Pb acetate in
drinking water for 3 weeks (PND 37-58) increased spontaneous locomotor activity in juvenile ICR mice
tested on PND 58. These recent studies do not indicate effects on the activity of rodents when tested in
adolescence after Pb exposure from mothers, and they do not support the findings of hyperactivity in the
similarly exposed and tested mice described by Trombini et al. (2001).

Faulk et al. (2014). Basha et al. (2014). and Wang et al. (2016) evaluated activity in adult rodents
after developmental Pb exposure. Faulk et al. (2014) measured activity and horizontal movements along
with ambulatory activity by adult offspring of dams (Agouti mouse) exposed to Pb for weeks from
pregestation until weaning. Mean BLLs for offspring were not reported; however, maternal BLLs tested
at weaning were below the limit of detection in the control group and 4.1, 25.1, and 32.1 (ig/dL in the
three respective exposure groups of 2.1, 16, and 32 ppm. Overall horizontal activity was different across
Pb exposures in females but not in males. At 9 months, female offspring exposed to 2.1 ppm Pb had
higher average horizontal activity compared with controls. There was a sex-specific difference in
ambulatory measurement (subset of total horizontal activity), with only exposed females showing
significant differences from controls. Ambulatory activity was lower in females at the 32-ppm exposure
level at 3 months versus control offspring. Males did not exhibit significant differences at any time point
or exposure level. Although there was suggestive evidence of differences in the life-course patterns of
vertical activity by exposure among females, neither sex showed statistically significant differences
between Pb-exposed and control offspring. Testing adult rats at 4, 12, and 18 months of age, Basha et al.
(2014) found consistent decreases in locomotor activity associated with lactational Pb exposure. While
mean BLLs at these testing periods were shown to be below the 30 (ig/dL limit for PECOS relevance, it is
notable that the mean BLL for this group of animals was determined to be 49.5 (ig/dL at PND 45. Wang
et al. (2016) measured distance traveled in the open field by Sprague Dawley rats aged 116-122 days
after adolescent Pb exposure in drinking water from PND 24 to 56. They observed no effect of this
exposure.

Evidence of Pb exposure-induced intergenerational effects on rodent behavior was also reviewed
in the 2006 Pb AQCD (U.S. EPA. 2006c). Trombini et al. (2001) observed increased open-field
ambulation in F2 generation rats derived from female offspring of Pb-treated pregnant mothers. Recently,
Sobolewski et al. (2020) exposed F0 mice to Pb during pregnancy and lactation, bred offspring with
unexposed mates for two generations (F1 and F2), and then evaluated behavior in the F3 generation. F3
females demonstrated a small increase in locomotor activity, regardless of lineage.

External Review Draft

3-96

DRAFT: Do not cite or quote

-------
Overall, there are still mixed indications on locomotor activity from Pb exposure studies with
BLLs <30 |ig/dL. which may be due to differential dosing, timing of exposures, and activity
measurements. However, in a rare set of four individual studies wherein these critical factors were
analogous, Pb exposure during lactation induced hyperactivity in rodents when tested within a PND 14 to
23 window (Duan et al.. 2017; De Marco et al.. 2005; Moreira et al.. 2001; Rodrigues etal.. 1996). Pb-
induced hyperactivity in rodents provides some support for hyperactivity observed in children but may be
more appropriately interpreted in the context of additional behavioral assays.

3.5.2.4 Parent and Teacher Ratings of ADHD-related Behavior

In addition to finding associations with attention, impulsivity, and hyperactivity, epidemiologic
studies also found associations between higher concurrent BLLs and higher parent and teacher ratings of
ADHD-related behaviors, calculated as a composite of the various behaviors evaluated in the diagnosis of
ADHD (see Section 4.3.3.1 of the 2013 Pb ISA (U.S. EPA. 2013a)). Most of these studies were limited
due to their cross-sectional design and lack of validation of ADHD ratings with clinical diagnosis.
Although diagnostic guidelines for ADHD exist, the exact criteria or specific behaviors required can vary.
Thus, within studies, there were variations among subjects in the types of behaviors they displayed that
led to a diagnosis of ADHD. Further the available studies considered age, sex, and SES or parental
education but generally not both as potential confounders, and none of the studies considered parental
caregiving quality.

Recent longitudinal studies add to the body of evidence examining the association between
prenatal and childhood Pb exposure and parent/teacher-rated ADHD symptoms in populations with
relatively low blood Pb concentrations (<5 (ig/dL). This group of studies includes some that found
associations with hyperactivity using the SDQ, which is a screening questionnaire that includes five
domains. Sioen et al. (2013) analyzed data from the Flemish Environment and Health Study (FLEHS I,
2002-2006), a birth cohort comprising mother-infant pairs to examine the association between cord blood
Pb and ADHD-related behaviors for 281 infants whose parents returned the SDQ (26.4%). A positive
association of cord blood Pb concentration with hyperactivity score >7 was observed (OR: 2.94 [95% CI:
1.17, 7.38 per log ug/dL increase in BLL]). In another study using this assessment instrument, Fruh et al.
(2019) analyzed data from mother-child pairs participating in Project Viva, a longitudinal birth cohort in
eastern Massachusetts. Maternal blood Pb concentration in erythrocytes was measured during the second
trimester of pregnancy and parents rated their child's behavior using the SDQ in mid-childhood (median
7.7 years). The associations (i.e., (3 coefficients) with the parent and teacher-rated hyperactivity
component of the SDQ were 0.10 (95% CI: -0.21, 0.41) and 0.20 (95% CI: -0.24, 0.64), respectively.
While behavior assessments and maternal blood Pb measurements were available for fewer than half of
Project Viva participants, important confounders including HOME score, maternal IQ, and parental
education were considered in this study.

External Review Draft

3-97

DRAFT: Do not cite or quote

-------
Several of these longitudinal epidemiologic studies used the Behavior Assessment System for
Children (BASC) to assess both the behaviors and emotions of children (Reynolds and Kamphaus. 2015).
BASC-2 includes individual subscales for attention and hyperactivity as well as an overall behavioral
skills index (BSI) composite score. Specific rating scales and forms related to attention, hyperactivity, and
impulsivity are emphasized in this section (e.g., clinical scales such as "attention problems" or
"hyperactivity" on the teacher or parent rating scale forms).

Horton et al. (2018) analyzed data from the Early Life Exposure in Mexico to Environmental
Toxicants (ELEMENT) Project birth cohort in Mexico City to determine the association of weekly tooth
Pb concentration (prenatal through 1 year postnatal) with BASC-2 scores assessed between 8 and
11 years old. Distributed lag models were used to identify specific time windows of increased risk due to
Pb exposure. Tooth Pb concentration estimated to correspond with the 8 to 11 months postnatal period
was associated with parent-rated behavioral symptoms overall ((3 = 0.22 units [95% CI: 0.06, 0.38] per
natural log-transformed unit increase in dentine Pb concentration), and hyperactivity ((3 = 0.19 units [95%
CI = 0.02, 0.37] per natural log-transformed unit increase in dentine Pb concentration) after adjustment
for gestational age and maternal education. Approximately 12% of the original cohort was enrolled in this
study; participants differed with respect to several characteristics including child birth weight and
maternal IQ. Rasnick et al. (2021) conducted a study that estimated monthly air Pb exposure. The authors
also aimed to identify sensitive time windows of exposure; however, they attempted to distinguish
exposure to Pb in air by controlling for concurrent BLL (age 12 years) in their analysis of the Cincinnati
Study of Allergy and Air Pollution study data. Air Pb exposure was estimated using validated land use
regression models and behavioral outcomes, including attention and hyperactivity, were assessed using
BASC-2 administered at age 12. Distributed lag models to predict outcome responses based on current
and past (i.e., lagged) predicted air Pb exposures did not identify associations during any of the lifestages
examined. Models were adjusted for community deprivation, residential greenspace, and elemental
carbon attributable to traffic (ECAT), in addition to concurrent BLL.

In addition to examining attention using Conners' (Section 3.5.2.1.1), Ruebner et al. (2019)
evaluated the association of BLLs with parent-rated attention and hyperactivity symptoms on BASC-2.
This study was unique in that it enrolled children with CKD. Associations with parent ratings did not
persist in models that controlled for potential confounders including race, poverty, maternal education,
and clinical factors related to CKD. The median BLL in this study was 1.2 (ig/dL.

Several other instruments, including FBB-ADHS, the Child Behavior Checklist (CBCL), the
Disruptive Behavior Disorder (DBD) rating scale, the Barkley Adult ADHD-IV Rating Scale (BAARS),
Conners' Rating Scale (CRS), the Strengths and Weaknesses of ADHD Symptoms and Normal Behavior
Scale (SWAN), and DuPaul's ADHD rating scale were used to assess total ADHD in recent prospective
or case-control studies.

Neugebauer et al. (2015) conducted an analysis of the Duisburg birth cohort data to determine the
association of maternal BLL at 32 weeks gestation with parent-rated ADHD behaviors in childhood

External Review Draft

3-98

DRAFT: Do not cite or quote

-------
(average age 9.5 years old) assessed using FBB-ADHS. Maternal blood Pb was associated with overall
ADHD symptoms ((3 = 1.06 [95% CI: 1.01, 1.12]), with the strongest association observed for the
impulsivity component ((3 = 1.13 [95% CI: 1.06, 1.22]). These associations were observed after
adjustment for confounders including parental education, but not SES.

Liu et al. (2014b) examined the association of early childhood blood Pb concentration at 3, 4, or
5 years old (mean: 6.8 (ig/dL) with parent and teacher ratings of ADHD behaviors among Chinese school
children at age 6 using CBCL and the Caregiver-Teacher Report Form (C-TRF). The outcome was
modeled as a continuous and also as a dichotomous variable (i.e., clinically significant behavior problems
when T-score >60). The associations (i.e., (3) between increased blood Pb concentrations (per (ig/dL) and
ADHD behavior problems were 0.001 (95% CI: -0.002, 0.002) for problems reported on CBCL and 0.07
(-0.18 to 0.32) for behavior problems reported on C-TRF. The associations (i.e., OR) with clinically
significant ADHD behavior reported by parents on CBCL were 1.08 (95% CI: 0.99, 1.18) among children
overall, 1.04 (95% CI: 0.94, 1.16) among boys, and 1.15 (95% CI: 0.98, 1.35) among girls. The
participation rate was 81% in this study. Models were adjusted for confounders including parental
caregiving quality but not SES. Another prospective study evaluated the association of Pb exposure with
caregiver ratings on CBCL. In this study of adolescents, Winter and Sampson (2017) examined the
relationship between average BLLs in childhood (6 years old or younger) with impulsivity between 16 to
18 years old. These authors found a 0.06 SD (95% CI: 0.01, 0.12) increase in impulsivity score, after
adjustment for caregiver education and SES. Participants were originally enrolled in the mid-1990s and a
random sample of those that continued to participate in 1999 and 2002 was randomly selected for this
study, with 67% of those selected agreeing to participate.

Choi et al. (2016) investigated the association of childhood BLLs (geometric mean
BLL = 1.56 (ig/dL) with parent-rated ADHD symptoms later in childhood assessed using DuPaul's
ADHD rating scale. Approximately 72% (n = 2,159) of 2,967 eligible participants provided blood Pb
measurements and ADHD assessments, and 2052 were free of ADHD symptoms at baseline. A positive
association between childhood blood Pb and the development of ADHD symptoms at the 2-year follow-
up visit was observed in this study (RR: 1.55 [95% CI: 1.00, 2.40] >2.17 versus <2.17 (ig/dL) after
adjustment for residential area, household income, parental marital status, family history of psychiatric
disorders, preterm birth, and birth weight. A stronger association was observed among children with
higher BLLs and who resided in a single parent home (RR: 3.57 [95% CI 1.60, 7.98]).

Another longitudinal analysis examined the association of both cord BLL (mean: 4.7 (ig/dL) and
childhood (mean 2.7 (ig/dL) BLL with ADHD symptoms among Inuit children in Quebec (Boucher et al..
2012b). In this study, teachers completed the DBD rating scale to indicate Diagnostic and Statistical
Manual of Mental Disorders (DSM)-IV symptoms of ADHD inattentive type and ADHD hyperactive-
impulsive type. Associations between child concurrent log-transformed BLL and hyperactive/impulsive-
type ADHD symptoms assessed using DBD were observed (OR = 4.01 [95% CI: 1.06, 5.23] tertile 2
versus tertile 1; OR = 5.52 [95% CI: 1.38, 22.12] tertile 3 versus tertile 1). Inattentive-type ADHD

External Review Draft

3-99

DRAFT: Do not cite or quote

-------
symptoms on the DBD were not associated with Pb exposure. (Desrochcrs-Couture et al.. 2019) extended
this study by conducting a mediation analysis to estimate the direct and indirect associations of childhood
BLLs with adolescent externalizing behaviors, including ADHD symptoms assessed using BAARS. The
study found an association between childhood BLL and child hyperactivity and impulsivity, assessed by
teachers on CBCL ((3 = 0.45 [95% CI: 0.13, 0.78]). Neither a direct ((3 = 0.09 [95% CI: -0.11, 0.28]) nor
an indirect ([3 = -0.02 [95% CI: -0.06, 0.03]) association with adolescent ADHD symptomology assessed
using BAARS was observed. A wide array of covariates was considered as potential confounders
including both maternal education and SES.

Hong et al. (2015) found an association between blood Pb concentration and higher parent and
teacher-rated ADHD-RS symptoms ([3 = 1.04 [95% CI: 0.18, 1.90] and [3 = 1.90 [95% CI: 0.74, 3.05],
respectively) in a cross-sectional analysis of Korean school children from 8 to 11 years old after
adjustment for demographic factors (age, sex, residential region, paternal education level, and SES). This
association remained positive but was attenuated in models additionally adjusted for FSIQ, Mn, and Hg
([3 = 0.68 [95% CI: -0.20, 1.56] and [3 = 1.49 [95% CI: 0.32, 2.67], parent- and teacher-rated symptoms,
respectively). The mean BLL in this study was 1.80 (ig/dL. Associations indicating an increase in
commission errors on CPT were also observed.

Nigg et al. (2016) conducted a case-control study of children from Michigan (mean
BLL = 0.74 (ig/dL (cases) and 0.94 (ig/dL controls). In this study, ADHD composite indices were derived
for (1) inattention/disorganization and (2) composite hyperactivity-impulsivity using relevant scales of the
DuPaul, Conners', and SWAN scales. This study found an interaction between the hemochromatosis
(HFE) C282Y genotype, which is involved in iron metabolism, and BLL in predicting parent and teacher
reports of hyperactivity-impulsivity but not inattention. For example, the association between z scores of
BLL and hyperactivity was significantly stronger among those with the HFE C282Y mutation ([3 = 0.74,
[95% CI: 0.52, 0.96]) compared with those with the wild type genotype ([3 = 0.28 [95% CI: 0.15, 0.41]).
This study also found an interaction between z scores of logio-transformed BLL and sex (association
larger in boys) in predicting parent and teacher-rated hyperactivity and impulsivity but not attention.

3.5.2.4.1 Summary

Cross-sectional studies in the 2013 Pb ISA found associations between higher concurrent BLL
and higher parent and teacher ratings of ADHD-related behaviors, calculated as a composite of the
various behaviors that are evaluated in the diagnosis of ADHD (U.S. EPA. 2013a'). The evidence from
prospective studies was limited to Chandramouli et al. (2009). which found associations between BLL
and hyperactivity as rated by teachers and parents. Parent and teacher ratings generally lacked validation
and the available studies considered SES or parental education but generally not both as potential
confounders. None of the studies considered parental caregiving quality. Recent longitudinal studies that
established the temporality between the exposure and the outcome add to the body of evidence examining
the association between prenatal and childhood Pb exposure and parent/teacher-rated ADHD symptoms in

External Review Draft

3-100

DRAFT: Do not cite or quote

-------
populations with relatively low blood Pb concentrations (<6 (ig/dL). Across studies, associations were
observed with tooth Pb concentration, childhood BLLs, and maternal or cord (2-5 (ig/dL) BLLs. Studies
of caregiver-reported ADHD symptoms generally report associations with composite indices (Choi et al..
2016; Hong et al.. 2015; Neugebauer et al.. 2015; Liu et al.. 2014b; U.S. EPA. 2013a). and there is some
evidence indicating that the associations with impulsivity and hyperactivity symptoms (Dcsrochcrs-
Couture et al.. 2019; Fruh et al.. 2019; Horton et al.. 2018; Winter and Sampson. 2017; Nigg et al.. 2016;
Neugebauer et al.. 2015; Sioen et al.. 2013; Boucher et al.. 2012b) are stronger than the associations with
inattention symptoms. The majority of recent studies were prospective and generally reported moderate or
high participation rates. Some studies addressed the validity of caregiver assessed outcomes by evaluating
internal consistency (Rasnick et al.. 2021; Desrochers-Couture et al.. 2019). and Nigg et al. (2016)
addressed reliability and validity concerns by using structural equation modeling to create latent factors
for inattention and hyperactivity-impulsivity for each informant. Confounder adjustment remains
somewhat inconsistent across studies, although Liu et al. (2014b) and Fruh et al. (2019) adjusted for the
quality of parental caregiving, Choi et al. (2016) adjusted for family history of psychiatric disorders, and
several considered both SES and parental education (Desrochers-Couture et al.. 2019; Ruebner et al..
2019; Horton et al.. 2018; Winter and Sampson. 2017; Boucher et al.. 2012b). There is uncertainty
regarding the patterns of exposure that are associated with maternal and cord BLLs and BLLs in older
children because they may be influenced by higher past exposure.

3.5.2.5 Clinically Diagnosed ADHD

In the 2013 Pb ISA, results from a small body of cross-sectional studies indicated associations
between concurrent BLL and the prevalence of ADHD symptom ratings (Section 3.5.2.4) and clinically
diagnosed ADHD in children aged 4-17 years. The temporal relationship between Pb exposure and
ADHD was not established in these studies, and concurrent blood Pb concentrations in older children may
reflect higher past exposures. Additionally, outcomes in the ADHD symptom rating studies lacked
validation, and confounding was inconsistently addressed across studies. Because of these limitations, the
evidence specifically for these total ADHD index ratings and clinically diagnosed ADHD were
emphasized less in the 2013 Pb ISA (U.S. EPA. 2013a) than evidence for individual behaviors when
drawing conclusions about the effects of Pb exposure on attention, impulsivity, and hyperactivity.

Recent studies add to the evidence and address some of the uncertainties pertaining to the studies
included in the 2013 Pb ISA. Notably, Ji et al. (2018) analyzed data from the Boston Birth Cohort (1479
mother-infant pairs) to examine the association of early childhood Pb exposure (i.e., earliest (< age 4)
blood Pb concentration recorded during routine screening) with the development of ADHD later in
childhood. ADHD was assessed using electronic medical records (International Classification of Diseases
[ICD]-9 codes: 314.0, 314.00, 314.01, 314.1, 314.2,314.8, and 314.9, or ICD-10 codes: F90.0, F90.1,
F90.2, F90.8, and F90.9). Several important potential confounders (i.e., parental education, SES but not
quality of parental caregiving) were controlled for in the analysis, and child sex, maternal high-density

External Review Draft

3-101

DRAFT: Do not cite or quote

-------
lipoprotein (HDL), and maternal stress were considered as potential effect modifiers. Ji et al. (2018)
analyzed the association modeling blood Pb concentration as continuous and as categorical variables.
When blood Pb was analyzed using three categories, the OR comparing children with BLLs between 2
and 4 (ig/dL to children with BLLs <2 (ig/dL was 1.08 (95% CI: 0.81-1.44). The OR comparing children
with BLLs between 5 and 10 (ig/dL to children with BLLs <2 (ig/dL was 1.73 (95% CI: 1.09-2.73).

When blood Pb was modeled as a continuous variable, the OR was 1.12 [95% CI: 1.00, 1.25) per ug/dL
increase in BLL. Sex-stratified analyses comparing children with BLLs between 5 and 10 (ig/dL to
children with BLLs <5 (ig/dL indicated no association among girls (OR = 0.68 [95% CI: 0.27, 1.69]) and
a strong association among boys (OR = 2.49 [95% CI: 1.46-4.26]). Joint analyses indicated a 10-fold
increase in the magnitude of the association between childhood blood Pb concentration and ADHD
diagnosis among those with multiple risk factors (i.e., male sex, inadequate maternal HDL, higher
maternal stress).

Several additional recent studies also extend the evidence. Park et al. (2016) conducted a hospital
based case-control study in Busan, South Korea comparing the odds of higher blood Pb concentration
among diagnosed ADHD cases, which were confirmed using the Korean version of the Kiddie Schedule
for Affective Disorders and Schizophrenia Present and Lifetime (K-SADS-PL-K), to the odds of higher
blood concentration among controls that were frequency matched by age and sex and adjusted for other
potential confounders. Blood Pb was measured when the cases and controls were recruited into the study.
Higher blood Pb concentration was associated with increased risk of ADHD (OR: 1.60 [95 % CI: 1.04-
2.45] per unit increase in log BLL); however, blood Pb concentrations were not associated with ADHD-
RS score or CPT profiles among the ADHD cases. In a smaller case-control study that examined the
association of childhood Pb exposure with diagnosed ADHD among children living near a former smelter
in Omaha, Nebraska, Kim et al. (2013a) found a positive association (OR: 2.52 [95% CI: 1.07, 5.92] per
unit increase in natural log-transformed BLL).

In a recent cross-sectional analysis of NHANES (2003-2004) data, Geieretal. (2018) examined
the association of concurrent blood Pb concentration with self-reported doctor diagnosed attention deficit
disorder (ADD) among children and adolescents 10-19 years old. This study observed a positive
association between concurrent blood Pb concentration and ADD after adjusting for age, race, sex, and
SES (OR: 1.29 [95% CI: 1.03, 1.55]). In a previous analysis Braun et al. (2006) found an association in
children aged 4-15 years participating in NHANES (1999-2002). ADHD ascertained by the parent report
of ADHD diagnosis is subject to reporting bias; however, the examination of multiple risk factors and
outcomes in NHANES reduces the likelihood of biased participation and reporting of ADHD by parents
of children specifically with higher Pb exposure.

3.5.2.5.1 Summary

The 2013 Pb ISA assessed a small body of cross-sectional studies that examined the associations
between concurrent BLLs and the prevalence of clinically diagnosed ADHD. The temporal relationship

External Review Draft

3-102

DRAFT: Do not cite or quote

-------
between Pb exposure and ADHD was not established in these studies and it was noted that concurrent
blood Pb concentration in older children potentially reflects higher past exposures. Due to these
limitations, the evidence pertaining to total ADHD index ratings (see Section 3.5.2.4) and clinically
diagnosed ADHD was emphasized less than evidence for individual behaviors in drawing conclusions
about the effects of Pb exposure on attention, impulsivity, and hyperactivity in the 2013 Pb ISA (U.S.
EPA. 2013a). Further, the available studies did not consistently adjust for SES, parental education, and
quality of parental caregiving. A small number of recent studies add to the evidence showing consistent
associations between Pb exposure and diagnosed ADHD. One recent epidemiologic study (Ji et al.. 2018)
addressed several of the uncertainties identified in the literature included in the 2013 Pb ISA. Specifically,
this study employed a prospective design, and adjusted for parental education and SES (although not
quality of parental caregiving). Notably, ADHD was ascertained using ICD codes recorded on electronic
records and ADHD type was not distinguished in this study.

3.5.2.6 Relevant Issues for Interpreting the Evidence Base
3.5.2.6.1 Lifestages

Environmental exposures during critical lifestages spanning from childhood into adolescence can
affect key physiological systems that orchestrate brain development and plasticity (see Section 3.4.1.6.4
of U.S. EPA (2013a)). Epidemiologic studies examined in the 2013 Pb ISA consistently showed that
BLLs measured during various lifestages and time periods, including the prenatal period, early childhood,
later childhood, and averaged over multiple years, are associated with attention decrements, impulsivity,
and hyperactivity in children. These observations of Pb-associated elevated risk are well supported by
findings in animals that prenatal and early postnatal or lifetime Pb exposures alter brain development via
changes in synaptic architecture (Section 4.3.10.4 of U.S. EPA (2013a')') and neuronal outgrowth
(Section 4.3.10.10 of U.S. EPA (2013a)). potentially leading to increases in impulsivity (Section 4.3.3.1
of U.S. EPA (2013a')'). Potential mechanisms of lifestage-specific sensitivities are further reviewed in
Section 3.3. Recent studies support this conclusion from the 2013 Pb ISA.

A limited number of epidemiologic studies employed methods designed to further elucidate
critical lifestages for Pb exposure but did not change the overall conclusion in the 2013 Pb ISA. Horton et
al. (2018) used distributed lag models to identify specific time windows of increased Pb-associated
externalizing behaviors. This study found that tooth Pb concentration corresponding with the 8 to
11 months postnatal period was associated with parent-rated behavioral symptoms overall (0.22 units
[95% CI: 0.06, 0.38] per natural log-transformed unit increase in dentine Pb concentration), and
hyperactivity ((3 = 0.19 units [95% CI = 0.02, 0.37] per natural log-transformed unit increase in dentine Pb
concentration) after adjustment for gestational age and maternal education. In another study, Rasnick et
al. (2021) also aimed to distinguish critical windows of exposure to Pb, focusing on Pb concentration in
air by controlling for concurrent BLL (age 12 years). Air Pb exposure was estimated using validated land

External Review Draft

3-103

DRAFT: Do not cite or quote

-------
use regression models and behavioral outcomes, including attention and hyperactivity, were assessed
using BASC-2 administered at age 12. Distributed lag models to predict outcome responses based on
current and past (i.e., lagged) predicted air Pb exposures did not identify associations during any of the
lifestages examined.

3.5.2.6.2	Public Health Significance

The strongest evidence indicating a causal relationship between Pb exposure and attention,
impulsivity, and hyperactivity assessed in the 2013 Pb ISA was derived from studies that relied on
neuropsychological testing (U.S. EPA. 2013a). Domain-specific neuropsychological assessments of
attention, impulsivity, and hyperactivity have strong psychometric properties and rigorous validation;
however, deficits on these neuropsychological tests do not directly correspond to a diagnosis of ADHD
nor do they necessarily predict long-term consequences that might be associated with some types of
ADHD. Studies that evaluated the association of Pb exposure with behavioral symptoms of ADHD
assessed using teacher and parent ratings contributed to the overall evidence in the 2013 Pb ISA, but the
limitations of these studies were noted. The bulk of the recent evidence comprises prospective studies of
parent or teacher ratings of ADHD behavioral symptoms. Although the recent studies addressed some
uncertainties in the previous ISA related to the temporal association of the exposure with the outcome and
controlled for potential confounding, the majority of the studies lacked validation and were subject to
greater measurement error compared with studies of outcomes assessed through neuropsychological
testing. Studies of diagnosed ADHD are also subject to limitations. Although diagnostic guidelines for
ADHD exist, the exact criteria or specific behaviors required for diagnosis may vary across studies. The
recent study by Ji et al. (2018) addressed several of the uncertainties regarding the association of Pb
exposure with clinical ADHD. This study was prospective in design, assessed early childhood BLL
(<4 years old), and adjusted for parental education and SES (although not quality of parental caregiving);
however, ADHD was ascertained using ICD codes recorded on electronic records and ADHD type was
not distinguished.

3.5.2.6.3	Potentially At-Risk Populations

Sex

Studies examining sex as an at-risk factor for attention, hyperactivity and impulsivity outcomes
were not assessed in the 2013 Pb ISA. A recent study by Nigg et al. (2016) found an interaction between
BLL and sex in predicting parent and teacher-rated hyperactivity and impulsivity but not attention. The
association was larger in boys in this study.

External Review Draft

3-104

DRAFT: Do not cite or quote

-------
Maternal Smoking

Maternal smoking during pregnancy was examined in a study of children's concurrent BLLs and
the prevalence of ADHD among children aged 8-15 years. An interaction was observed between
children's current BLLs and prenatal tobacco smoke exposure; those children with high Pb levels and
prenatal tobacco smoke exposure had the highest odds of ADHD (Froehlich et al.. 2009). Recent studies
have not examined maternal smoking as an at-risk factor.

Co-exposure to Other Metals or Chemicals

Studies examining other metals as an at-risk factor for attention, hyperactivity and impulsivity
outcomes were not assessed in the 2013 Pb ISA. Some recent studies adjusted for other metals or
chemicals (e.g., PCBs) (Ethier et al.. 2015; Tatsuta et al.. 2014). and effect modifications were observed
in other studies (Yorifuii et al.. 2011). For example, Yorifuii et al. (2011) found a less-than-additive
interaction between cord Pb and Hg concentrations. Specifically, a lower digit span forward score on the
WISC-R ((3 = -1.70 [95% CI: -3.12, -0.28] per log-transformed BLL) at age 7 and a lower digit span
backward score on the WISC-R ((3 = -2.73 [95% CI: -4.32, -1.14] per log-transformed BLL) at age 14
were observed among children with the lowest Hg exposure.

Gene-Environment Interactions

Studies examining gene-environment interactions in the context of attention, hyperactivity and
impulsivity outcomes were not assessed in the 2013 Pb ISA. Interactions between child BLL and genes
that regulate neurodevelopmental processes were observed in studies of attention (Choi et al.. 2020;
Roonev et al.. 2018). Genes that were implicated included variants of GRIN2A and GRIN2B and
genotypes involved in the regulation of noradrenergic pathways. In addition, Nigg et al. (2016) found an
interaction between the HFE C282Y genotype and BLL in predicting parent and teacher reports of
hyperactivity-impulsivity but not inattention. Specifically, the association between z scores of BLL and
hyperactivity was significantly stronger among those with the HFE C282Y mutation ([3 = 0.74 [95% CI:
0.52, 0.96]) compared with those with the wild type genotype ([3 = 0.28 [95% CI: 0.15, 0.41]).

3.5.2.7 Integrated Summary and Causality Determination: Attention, Impulsivity, and
Hyperactivity

Attention, hyperactivity, and impulsivity are included within the ADHD domain of externalizing
behaviors. Although not studied as extensively as cognitive function, several epidemiologic studies have
examined the relationship between Pb exposure in children and attention, impulsivity, and hyperactivity
in children and young adults.

External Review Draft

3-105

DRAFT: Do not cite or quote

-------
The majority of these studies examined attention, and some also examined impulsivity or
hyperactivity. Thus, the focus of the evaluation is on the evidence related to attention, but the evaluation
also draws on coherence with evidence for impulsivity and hyperactivity, including evidence in animals
and that suggesting potential modes of action. The collective epidemiologic evidence base for attention in
children comprises many prospective and cross-sectional studies, which were also reviewed in the 2006
Pb AQCD and the 2013 Pb ISA and some recently published studies. Most of these studies reported
associations between childhood blood or tooth Pb levels and attention decrements, impulsivity, and
hyperactivity (Table 3-7E). A small number of recent longitudinal studies contributed to this evidence.
Not all results were uniform with regard to precision and the magnitude of the association, but results
mostly showed a pattern of attention decrements, impulsivity, and hyperactivity with higher blood or
tooth Pb levels.

Whether prospective, cross-sectional, or longitudinal, most studies relied on population-based
recruitment from prenatal clinics, hospitals at birth, or schools and reported moderate to high
participation. Several of the studies reviewed in the previous ISA demonstrated increased loss-to-follow-
up in certain groups (e.g., lower SES or HOME scores), which has the potential to introduce selection
bias and reduce the generalizability of findings. A strong indication that participation in the study was
biased to those with higher BLLs and greater deficits in attention, hyperactivity, or impulsivity was not
observed. Recent studies incorporated adjustments for these and other covariates. Repeated testing in
children was common but the consistent pattern of association observed across the ages, BLL, and
behavioral outcomes examined increases confidence that the evidence is not unduly biased by the
increased probability of finding associations by chance alone. Coherence with animal studies, which are
less vulnerable to confounding, further supports the pattern of associations described in the preceding
sections.

The strongest epidemiologic evidence indicating an association of Pb exposure with inattention
and hyperactivity is described in the 2013 Pb ISA U.S. EPA (2013a). Prospective studies showed strong
support for an association between Pb exposure (range: 7-14 (ig/dL) and decreased scores on
neuropsychological tests and parent/teacher ratings of attention and hyperactivity. Cross-sectional studies
from the 2013 Pb ISA generally corroborated these observations. Studies of impulsivity in children in the
2013 Pb ISA were limited by their quantity and lack of temporality but generally indicated associations of
Pb exposure with worse scores on tests of response inhibition and on parent/teacher ratings of impulsivity
in cross-sectional analyses. A small number of recent prospective studies with mean maternal and cord
BLLs <5 (ig/dL report associations with some measures of inattention (Ethier et al.. 2015; Neugebaueret
al.. 2015). In addition, recent analyses of Inuit children add support for the relationship between child and
cord BLL and impulsivity (Boucher et al.. 2012a'). There is uncertainty regarding the patterns of exposure
associated with maternal and cord BLLs.

Most of the aforementioned studies of parent and teacher ratings of ADHD-related behaviors in
the 2013 Pb ISA were largely cross-sectional in design. The evidence from prospective studies was

External Review Draft

3-106

DRAFT: Do not cite or quote

-------
limited to Chandramouli et al. (2009). which found associations between BLLs and hyperactivity as rated
by teachers and parents. Parent and teacher ratings generally lacked validation and the available studies
considered SES or parental education but generally not both as potential confounders, and none of the
studies considered parental caregiving quality. The bulk of the recent evidence comprises prospective
studies that established the temporality of the association between Pb exposure and parent or teacher
ratings of ADHD symptoms and clinical ADHD. Across studies, associations were observed with tooth
Pb concentration, childhood (<6 (.ig/dL). and maternal or cord (2-5 (ig/dL) BLLs. Studies of caregiver-
reported ADHD symptoms generally reported associations with composite indices, and there is some
evidence that the associations with impulsivity and hyperactivity symptoms are stronger than the
associations with inattention symptoms. The majority of the recent studies were prospective and generally
reported moderate or high participation rates. Some studies addressed the validity of caregiver assessed
outcomes by evaluating internal consistency (Rasnick et al.. 2021; Desrochers-Couture et al.. 2019). and
Nigg et al. (2016) addressed reliability/validity concerns by using structural equation modeling to create
latent factors for inattention and hyperactivity-impulsivity for each informant. Confounder adjustment has
become more consistent across recent studies. In addition to the studies relying on parent and teacher
behavior ratings, a small number of recent studies add to the evidence showing consistent associations
between Pb exposure and diagnosed ADHD. One recent epidemiologic study (Ji et al.. 2018) addressed
several of the uncertainties identified in the literature included in the 2013 Pb ISA. Specifically, this study
employed a prospective design, assessed early childhood BLL, and adjusted for parental education and
SES (although not quality of parental caregiving). Notably, in this study, ADHD was ascertained using
ICD codes recorded on electronic records and ADHD type was not distinguished. Uncertainty remains
regarding the patterns of exposure associated with maternal and cord BLLs and BLLs in older children
because they may be influenced by higher past exposures.

In summary, the total body of evidence evaluated in this and previous assessments is
sufficient to conclude that there is a causal relationship between Pb exposure and attention
decrements, impulsivity, and hyperactivity. This conclusion reflects the consistency of the results from
epidemiologic studies of externalizing behaviors in children and young adults, incorporating various
objective neuropsychological tests and reporting from teachers and parents. The conclusion also
incorporates the coherence of evidence across epidemiologic and toxicological studies of externalizing
behaviors and biological plausibility provided by studies that outline pathways by which Pb may interfere
with the proper development, connectivity, and function of systems underlying externalizing behaviors.
These findings are consistent with the conclusion in the 2013 Pb ISA.

External Review Draft

3-107

DRAFT: Do not cite or quote

-------
Table 3-3 Summary of evidence indicating a causal relationship of Pb exposure with attention, impulsivity,
and hyperactivity.

Rationale for

Causality
Determination3

Key Evidence13

References'3

Pb Biomarker Levels Associated with Effects0

Row heading if applicable

Consistent
associations from
multiple prospective
epidemiologic
studies with relevant
BLLs

Evidence from prospective
studies for attention decrements
and hyperactivity in association
with prenatal (maternal or cord),
early childhood, and lifetime
umber blood Pb and tooth Pb
levels in children ages 7-17 yr
and young adults 19-20 yr in the
United States, United Kingdom
Australia, New Zealand.

Evidence from prospective
studies of parent or teacher-
rated ADHD composite
symptom indices derived from
widely used, structured
instruments.

Burns et al. (1999)

Ris etal. (2004)

Ferausson et al. (1993)

Bellinger et al. (1994a)

Chandramouli et al. (2009)

Leviton et al. (1993)

Section 4.3.3.1 U.S. EPA (2013a)

Neuqebauer et al. (2015)

Ethieret al. (2015)

Choi, 2016, 3351960; Neugebauer, 2015, 2920111;
Liu, 2014, 5438421

Blood Pb:

Means 2 to 8.3 [jg/dL (prenatal maternal or cord),

8.3 [jg/dL age 6 yr, 13.4 age 3-60 month, 14 [jg/dL

(lifetime avg to age 11-13 yr)

Group with age 30 mo >10 [jg/dL

Tooth Pb (ages 6-8 yr): Means: 3.3, 6.2 |jg/g

Childhood BLL <6; maternal and cord BLL 2-
5 [jg/dL.

Ratings for impulsivity and
hyperactivity more strongly
associated with Pb exposure

Prospective analysis found
associations with impulsivity in
Inuit children

(Desrochers-Couture etal., 2019; Fruh et al.. 2019; Childhood BLL <6; maternal and cord BLL 2-
Horton et al.. 2018: Winter and Sampson. 2017: Nigg 5 [jg/dL.
et al.. 2016: Neugebauer et al.. 2015: Sioen et al..

2013; Boucher etal.. 2012b)
Boucher et al. (2012a)

Mean 4.7 (cord), 2.7 (concurrent, average age
11.3 yr)

Prospective analysis find
association with attention
decrements in children with CKD

Ruebner et al. (2019)

Child blood Pb
age 4-18 yr

-2 yr before outcome assessment at

External Review Draft

3-108

DRAFT: Do not cite or quote

-------
Rationale for







Causality

Key Evidence13

References'3

Pb Biomarker Levels Associated with Effects0

Determination3





Limited evidence

Co-exposure to Hg modified the

Yorifuii etal. (2011)



evaluates the

risk of Pb-associated effects on





potential modification

attention (less than additive





of Pb associations to

effect observed)





other metals or







genes

Interactions between BLL and

Niqa etal. (2016)





genes that regulate

Roonev et al. (2018)





neurodevelopmental processes

Choi et al. (2020)





observed.





Association observed in

Ji etal. (2018)

Mean: 2.2 [jg/dL (<4 yr of age)



prospective study of clinical







ADHD diagnosed before age 6,







with adjustment for parental







education and SES.







No association found with

Wasserman et al. (2001)

Blood Pb: Mean 7.2 [jg/dL for lifetime (to age 4-5 yr)



ratings of attention problems in



avg



children ages 4-5 yr in whom







ratings may be measured less







reliably.





Supporting evidence

Associations of concurrent BLL

Section 4.3.3.1 U.S. EPA (2013a)

Concurrent (ages 5-7.5 yr) blood Pb: Means 5.0-

from cross-sectional

with attention decrements,



5.4 [jg/dL

studies

impulsivity, and hyperactivity in







children ages 5-7.5 yr. Some







populations had high prenatal







drug or alcohol exposure.





External Review Draft

3-109

DRAFT: Do not cite or quote

-------
Rationale for

Causality
Determination3

Key Evidence13

References'3

Pb Biomarker Levels Associated with Effects0

Epidemiologic
studies help rule out
chance, bias, and
confounding with
reasonable
confidence

Most prospective and some
cross-sectional studies found
associations with adjustment for
SES, maternal education, and
parental caregiving quality
(HOME score). Some also
considered parental IQ,
smoking, birth outcomes. A few
considered substance abuse,
nutritional factors, and family
history of psychiatric disorders.

Studies had population-based
recruitment with moderate to
high follow-up participation not
conditional on blood or tooth Pb
level.

Section 4.3.3.1 U.S. EPA (2013a)

HOME score: Liu et al. (2014b) and
Fruh etal. (2019)

Family history of psychiatric disorders:

Choi etal. (2016)

SES and parental education: Desrochers-Couture
et al. (2019): Ruebner et al. (2019): Horton et al.
(2018): Winter and Sampson (2017): Boucher et al.
(2012b)

Consistent evidence
in animals with
relevant exposures

Several studies report increased Duan et al. (2017): De Marco et al. (2005): Moreira

open-field activity in rodents et al. (2001): Rodriques et al. (1996)

following developmental Pb

exposure, consistent with

hyperactivity.

Blood Pb: 19-28 [jg/dL in mice with lactational
exposure (tested PND 15-19); 10-29 [jg/dL in rats
with lactational exposure (tested PND 14-23)

Evidence from
lifetime Pb exposure
in nonhuman
primates suggests
that Pb produces
attention decrements

Lifetime Pb exposure in
nonhuman primates was
reported to increase
distractibility in a spatial
discrimination task. Evidence
from lifetime Pb exposure in
nonhuman primates supports
the findings in humans.

Gilbert and Rice (1987)

Blood Pb: 15-25 pg/dL

Evidence from lifetime Pb
exposure in nonhuman primates
suggests that Pb increased
perseveration and errors of
commission in a spatial
discrimination reversal task

Rice (1990): Rice and Gilbert (1990b): Gilbert and
Rice (1987)

Blood Pb; 15-36 pg/dL

External Review Draft

3-110

DRAFT: Do not cite or quote

-------
Rationale for







Causality

Key Evidence13

References'3

Pb Biomarker Levels Associated with Effects0

Determination3





Blood Pb; 15-

Evidence from lifetime Pb

Rice (1990): Rice and Gilbert (1990b): Gilbert and

Blood Pb; 15 pg/dL

36 |jg/dL

exposure in nonhuman primates

Rice (1987)





suggests that Pb increased







perseveration and errors of







commission in a spatial







discrimination reversal task





Evidence describes

Both in vitro and in vivo

Section 3.6



biologically plausible

evidence suggests that Pb





pathways

exposure may influence brain







development,







neurotransmission, connectivity,







neuronal integrity, all of which







may underlie the observed







alterations in externalizing







behaviors.





ADHD = attention deficit/hyperactivity disorder; avg = average; BLL = blood lead level; CKD = chronic kidney disease; HOME = Health Outcomes and Measures of The Environment
Study; mo = month(s); Pb = lead; PND = postnatal day; SES = socioeconomic status; yr = year(s).

aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Table I and Table II of the Preamble to the ISAs (U.S. EPA. 2015).
bDescribes the key evidence and references, supporting or contradicting, contributing most heavily to causality determination and, where applicable, to uncertainties or
inconsistencies. References to earlier sections indicate where the full body of evidence is described.

°Describes the Pb biomarker levels at which the evidence is substantiated.

External Review Draft

3-111

DRAFT: Do not cite or quote

-------
3.5.3 Externalizing Behaviors: Conduct Disorders, Aggression, and Criminal
Behavior in Children, Adolescents, and Young Adults

There are two domains of conduct disorders that are considered in the ISA: undersocialized
aggressive conduct disorder, and socialized aggressive conduct disorder ("Whitcomb and Merrell. 2012).
As discussed in the 2013 Pb ISA (U.S. EPA. 2013a). these domains are combined in this assessment
because they cannot be disentangled based on the available epidemiologic literature. This section also
considers evidence for criminal offenses, which are associated with conduct disorders (U.S. EPA. 2013a).
Although not described explicitly as a part of either domain of conduct disorders, evidence for criminal
offenses is reviewed with conduct disorders because conduct disorders can be predictors of subsequent
delinquency and criminality (Soderstrom et al.. 2004; Babinski etal.. 1999; Paier. 1998).

The evidence reviewed in the 2013 Pb ISA is sufficient to conclude that a "causal relationship is
likely to exist" between Pb exposure and conduct disorders in children and young adults (U.S. EPA.
2013a). Prospective studies consistently indicated that earlier childhood (e.g., age 30 months 6 years) or
lifetime average (to age 11-13 years) BLLs or tooth (from ages 6-8 years) Pb levels were associated with
criminal offenses in young adults aged 19-24 years, and with higher parent and teacher ratings of
behaviors related to conduct disorders in children ages 7-17 years (see Table 4-12 of (U.S. EPA. 2013a)
and (U.S. EPA. 2006b)). Pb-associated increases in conduct disorders were found in populations with
mean BLLs of 7-14 (ig/dL. These associations were found without indication of strong selection bias and
with adjustment for SES, parental education and IQ, parental caregiving quality, family functioning,
smoking, and substance abuse. Supporting evidence was provided by cross-sectional studies of children
participating in NHANES, e.g., (Braun et al.. 2008). and a meta-analysis of prospective and cross-
sectional studies (Marcus et al.. 2010). In addition, there was coherence across related measures of
conduct problems in epidemiologic studies. Evidence for Pb-induced aggression in animals was mixed,
however, with increases in aggression found in some studies of adult animals with gestational plus
lifetime Pb exposure but not juvenile animals. The strongest evidence for the 2013 causality conclusion
was provided by prospective epidemiologic studies, with support from cross-sectional studies of criminal
offenses and ratings of behaviors related to conduct disorders. Associations with lower BLLs that were
not influenced by higher earlier Pb exposures as in older children and adults were not well characterized,
however.

Studies published since 2013 from both cohort and cross-sectional studies add to this evidence
base, which continues to support a "likely to be causal" relationship, as described in Table 3-4 and below.
The central tendency Pb levels, study-specific details, and selected effect estimates are highlighted in
Table 3-9E.

External Review Draft

3-112

DRAFT: Do not cite or quote

-------
3.5.3.1

Epidemiologic Studies of Conduct Disorders in Children and Adolescents

The 2013 Pb ISA describes cohort and cross-sectional studies demonstrating associations of Pb
exposure with behaviors related to conduct disorders, including criminal offenses (U.S. EPA. 2013a).
Collectively, the evidence from prospective cohort studies indicated associations of aggressive, antisocial,
delinquent, and criminal behavior with biomarkers of Pb exposure. Cross-sectional studies also provided
evidence on these associations, though there is more uncertainty in data from this study design due to
limitations in assessing temporality.

Recent studies have evaluated associations between Pb and conduct disorders, aggressive
behavior, and other related measures of externalizing behavior. Most of these studies were prospective
cohort studies (Tlotlcng et al.. 2022; Desrochers-Couture et al.. 2019; Reuben et al.. 2019; Beckwith et
al.. 2018; Nkomo et al.. 2018; Nkomo et al.. 2017; Liu et al.. 2014b; Sioen et al.. 2013; Boucher et al..
2012b; Tatsuta et al.. 2012). Several utilized self-report tools (e.g., Youth Self-Report [YSR], Buss-Perry
Aggression Questionnaire [BPAQ], Psychopathic Personality Inventory [PPI], Antisocial Behavior
Interview) to assess aggression, violence, or other socio-behavioral problems among adolescents and
young adults aged 14-24 years. These studies reported central tendency BLLs at ages 6.5-13 years
ranging from 2.3 to 8 (ig/dL or mean bone Pb of 8.7 (ig/dL (Tlotleng et al.. 2022; Desrochers-Couture et
al.. 2019; Beckwith et al.. 2018; Nkomo et al.. 2018; Nkomo et al.. 2017). In analyses adjusted for most
key confounders, associations were observed for: physical violence ((3: 0.05; 95% CI: 0.04, 0.05) (Nkomo
et al.. 2017); direct aggression ((3 [95% CI] comparing those with BLLs >10 (ig/dL to those with BLLs
<5 ng/dL: 0.43 [0.08, 0.78]) (Nkomo etal.,2018): anger aggression ([3 = 0.25 [95% CI: 0.04, 0.37])
(Tlotleng et al.. 2022); and PPI (overall [3 = 0.22 [95% CI: 0.06, 0.38]; female [3 = 0.16 [95% CI: -0.05,
0.37]; male [3 = 0.22 [95% CI: -0.02, 0.47]) (Beckwith et al.. 2018)(Table 3-9E). Although the PPI serves
as a measure of psychopathic personality traits, psychopathy more generally includes behavioral factors
such as aggression and criminal conduct, in addition to personality traits. As discussed in the 2013 Pb ISA
(U.S. EPA. 2013a). an analysis of this same cohort reported associations between BLLs and criminal and
violent criminal arrests at ages 19-24 (Wright et al.. 2008). (Beckwith et al.. 2018) also noted that BLLs
were associated with volumetric reductions in gray matter in the frontal lobe and white matter in several
brain regions. Considered together, these studies provide support for an association between childhood Pb
exposure and psychopathy in adolescents and young adults that may stem from changes in brain
morphology. In addition to these studies examining total effects, one prospective study of Pb and self-
reported behavioral outcomes conducted mediation analyses and reported an association between Pb and
adolescent externalizing behavior mediated through child externalizing behavior ([3: 0.18, 95% CI: 0,
0.36) (Desrochers-Couture et al.. 2019). There was also evidence of a small but imprecise direct effect,
though there was likely limited power to detect a direct effect given the small sample size and correlation
between child and adolescent externalizing behavior. The observed association between BLLs and
adolescent externalizing behavior is at least partially mediated through child externalizing behavior in this
study population.

External Review Draft

3-113

DRAFT: Do not cite or quote

-------
Other prospective cohort studies used observer assessments (e.g., parent or teacher ratings) to
assess conduct disorders, aggression, and related behaviors among children with central tendency blood
or cord blood Pb ranging from 0.4 to 14.3 (ig/dL (Fruh et al.. 2019; Ruebner et al.. 2019; Liu et al..
2014b; Sioen et al.. 2013; Boucher et al.. 2012b; Tatsuta et al.. 2012). Some of these studies focused on
the prenatal period as a potentially sensitive period of exposure: three of these studies evaluated Pb levels
in cord blood (Sioen et al.. 2013; Boucher et al.. 2012b; Tatsuta et al.. 2012) and one evaluated second
trimester maternal BLLs (Fruh et al.. 2019). All but two (Ruebner et al.. 2019; Boucher et al.. 2012b) of
these studies evaluated the outcome in children with mean age <8 years. These analyses reported
generally null associations (Table 3-9E).

There were also some cross-sectional evaluations of blood Pb and behavioral problems (e.g.,
aggression, oppositional, externalizing, antisocial) covering children with central tendency BLLs ranging
from 0.7 to 11.08 (ig/dL (Liu et al.. 2022b; Desrochers-Couture et al.. 2019; Reuben et al.. 2019; Barg et
al.. 2018; Rodrigucs et al.. 2018; Boucher et al.. 2012b; Naicker et al.. 2012; Nigg et al.. 2010). These
analyses utilized a mix of self-report and observer assessment tools to evaluate the outcomes of interest in
children aged 6-13 years. Positive associations were reported in most studies, including for child
externalizing behavior ((3: 0.23; 95% CI: 0.08, 0.38) and child oppositional defiant and conduct disorder
(OD/CD) ((3: 0.37; 95% CI: 0.06, 0.69) (Desrochers-Couture et al.. 2019); "attacking people" (boys only,
see table for unstandardized estimate; (Naicker et al.. 2012)); teacher-reported aggressive and rule-
breaking behavior (referred to in the paper as "externalizing behavior") (log-transformed concurrent Pb (3:
0.14; 95% CI: 0.01, 0.26) (Boucher et al.. 2012b); parent-reported externalizing composite ((3 for SD
increase in symptoms scores per SD increase in loglO transformed BLL = 0.21 [95% CI: 0.05, 0.37]) and
oppositional behavior ((3 for SD increase in symptoms scores per SD increase in log 10 transformed BLL
= 0.09 [95% CI: -0.09, 0.27]) (Nigg etal.,2010); antisocial behavior ([3 = 0.02 [95% CI: 0.00, 0.04])
(Reuben et al.. 2019); and antisocial/aggressive behavior factor (Parent-reported [3 = 0.20 [95% CI: 0.05,
0.34]; Child-reported [3 = 0.20 [95% CI: 0.04, 0.35]) (Liu et al.. 2022b). Both of the null studies evaluated
the outcome in groups of children that included individuals aged <8 years (Barg et al.. 2018; Rodrigues et
al.. 2018); it is possible that behavioral ratings are less reliable at younger ages or that these outcomes
manifest at later ages (Blair. 2001). More research is needed to disentangle these issues. Overall, cross-
sectional studies were less of a consideration in drawing conclusions on the effects of Pb, given their
inherent limitations with regard to temporality.

Among recent studies, there were also evaluations of the association between Pb and suspensions,
arrests, juvenile delinquency, and crime (including violent crime) (Wright et al.. 2021; Becklev et al..
2018; Boutwell et al.. 2017; Amato et al.. 2013). The strongest evidence comes from two prospective
studies. Using data from the CLS on multiple measures of blood Pb (from prenatal to age 6 years;
mean = 14.4 (ig/dL) and arrests from ages 18-33, (Wright et al.. 2021) observed numerous positive
associations, including for adult arrests (RR =1.01 [95% CI: 1.00, 1.03]), lifetime arrests (RR = 1.02
[95% CI: 1.00, 1.03]), arrests for violent crime (RR = 1.02 [95% CI: 0.99, 1.04]), and arrests for drug
crime (RR= 1.03 [95% CI: 1.01, 1.06]) (Wright et al.. 2021). Additionally, (Amato et al.. 2013) reported

External Review Draft

3-114

DRAFT: Do not cite or quote

-------
that Pb exposure during the first 3 years of life (based on a BLL >10 (ig/dL and <20 (ig/dL) increased the
odds of school suspensions in fourth grade, compared with those without Pb exposure during the first
three years of life (based on a BLL <5 (ig/dL) (OR: 2.66; 95% CI: 2.12, 3.32) (Amato et al.. 2013).

Supporting evidence comes from several prospective studies with some limitations that affect
interpretation and confidence as well as one ecologic study. Criminal offending, comprising both criminal
conviction and self-report offending, was evaluated in a prospective cohort study based in Dunedin, New
Zealand in which the mean 11-year-old BLL was 11.01 (ig/dL (Becklev et al.. 2018). In sex-adjusted
analyses of convictions, the authors reported that increased childhood BLL was associated with increased
odds of at least one nonviolent criminal conviction for ages 15-38 years (OR: 1.05; 95% CI: 1.00, 1.10).
However, sex-adjusted analyses of other criminal conviction endpoints (e.g., any criminal conviction,
recidivistic conviction, one-time conviction, violent offense) were inconclusive (see Table 3-9E). While
this study had extensive follow-up (27 years) and both subjective and objective measures of the outcome,
the limited adjustment for confounders is a key concern. Analyses were only adjusted for sex without
consideration of other potential covariates, leaving the possibility of residual confounding. In an ecologic
study of 106 census tracts in St. Louis, Missouri, United States, (Boutwell et al.. 2017) reported that a 1%
increase in the proportion of elevated blood Pb tests (>5 mg/dL) among children within a census tract was
associated with increased risk ratios for firearm crimes (RR: 1.03; 95% CI: 1.03, 1.04), assault crimes
(RR: 1.03; 95% CI: 1.02, 1.03), robbery crimes (RR: 1.03; 95% CI: 1.02, 1.04), and homicides (RR: 1.03;
95% CI: 1.01, 1.04). The association with rape was inconclusive (RR: 1.01; 95% CI: 0.99, 1.03). While
ecologic studies can be useful for hypothesis generation and understanding patterns among groups, the
lack of control for individual-level confounding factors in such studies leaves concern for risk of bias.

3.5.3.1.1 Summary

Overall, recently published epidemiologic studies support the findings from the previous ISA.
The strongest evidence published since 2013 comes from prospective cohort studies of 1) self-reported
conduct and aggression-related outcomes (Tlotlcng et al.. 2022; Desrochers-Couture et al.. 2019;
Beckwith et al.. 2018; Nkomo et al.. 2018; Nkomo et al.. 2017). and 2) external measures of delinquency
(e.g., criminal arrests, school suspensions) (Wright et al.. 2021; Amato et al.. 2013). These studies
evaluated outcomes among individuals aged 7-33 years in relation to earlier (or cumulative) Pb levels.
BLLs were <10 (ig/dL in the studies of self-reported conduct and aggression-related outcomes and higher
in studies of external measures of delinquency (e.g., (Wright et al.. 2021); mean 14.4 (.ig/dL). These
studies controlled for most relevant confounders, and the prospective study design inherently ensured
appropriate temporality between the exposure and outcome. Additional supporting evidence comes from
cross-sectional studies using either self-report or observer-reported outcome measures among individuals
aged 6-13 years with concurrent BLLs ranging from 0.7-11.08 (ig/dL (Liu et al.. 2022b; Desrochers-
Couture et al.. 2019; Reuben et al.. 2019; Boucher et al.. 2012b; Naicker et al.. 2012; Nigg et al.. 2010).

External Review Draft	3-115	DRAFT: Do not cite or quote

-------
3.5.3.2	Toxicological Studies of Aggression

There are no recent PECOS-relevant studies examining the relationship between Pb exposure and
aggression. Available toxicological studies of aggression were described in the 2006 Pb AQCD (U.S.
EPA. 2006c) and 2013 Pb ISA (U.S. EPA. 2013a). The evidence supported effects of Pb exposure on
changes in social behavior of rodents and nonhuman primates. In animals, the social behavior most
comparable to conduct disorders in children is aggression; however, the effects of Pb on aggression in
animals were inconsistent. In animals, aggression was assessed as threats, attacks, bites, chases, and
offensive posture in encounters with other animals. Pb exposure was found to have no effect on
aggression in some studies as well as to decrease and increase aggression in others. Pb exposure generally
was not found to affect aggression in juvenile animals; however, increased aggression was found in adult
animals with high concentrations of gestational plus postnatal dietary Pb exposure. Recent PECOS-
relevant studies have not further examined the effects of Pb on aggression. Additional reported effects on
social behaviors described in the 2006 Pb AQCD (U.S. EPA. 2006c) and 2013 Pb ISA (U.S. EPA. 2013a)
included Pb-induced increases in social and sexual investigation, as indicated by sniffing, grooming,
following, mounting, and lordosis behavior. Despite the limited new evidence, observations for Pb-
induced changes in aggression in animals provide support for associations of altered aggression outcomes
in children. Furthermore, many of the more general overt nervous system toxicology studies discussed in
Sections 3.4.2 and 3.3 assessed a variety of endpoints, including brain structural changes and
neurotransmitter analysis, that can contribute to understandings of the mechanistic underpinning of
observed behavioral changes providing additional biological plausibility.

3.5.3.3	Relevant Issues for Interpreting the Evidence Base

3.5.3.3.1 Concentration-Response Function

The evidence base for this outcome is more limited compared with that for cognitive deficits, and
the shape of the C-R function cannot be determined from available studies. However, it is important to
highlight that in studies reviewed for the 2013 Pb ISA, effects were observed at central tendency BLLs of
5-10 (ig/dL (Nigg et al.. 2008; Wright et al.. 2008; Chiodo et al.. 2007; Wasserman et al.. 2001) and
<5 (ig/dL (Braun et al.. 2008). Among studies published since 2013, effects on conduct disorder,
aggression, and crime were observed at central tendency BLLs of 5-10 (ig/dL (Tlotleng et al.. 2022;
Beckwith et al.. 2018; Nkomo et al.. 2018; Nkomo et al.. 2017; Naicker et al.. 2012) as well as at central
tendency BLLs <5 (ig/dL (Liu et al.. 2022b; Desrochers-Couture et al.. 2019; Boucher et al.. 2012b; Nigg
et al.. 2010). However, it should be noted that there is less confidence in these studies of BLLs <5 (ig/dL
as they were all cross-sectional analyses (Liu et al.. 2022b; Desrochers-Couture et al.. 2019; Boucher et
al.. 2012b; Nigg et al.. 2010). Further work is needed to better understand whether the potential effects of
Pb on this outcome persist at BLLs <10 (ig/dL.

External Review Draft

3-116

DRAFT: Do not cite or quote

-------
3.5.3.3.2 Potentially At-Risk Populations

Sex

The 2013 Pb ISA identified one study that evaluated the role of sex as an at-risk factor. Wright et
al. (2008) examined early life BLLs and criminal arrests in adulthood and reported that risks attributable
to Pb exposure were greater among males than females (Wright et al.. 2008).

Several new studies evaluated the role of sex as an at-risk factor through sex-stratified analyses of
Pb exposure and conduct disorders. The results were generally inconclusive regarding sex as an at-risk
factor. In a prospective study of blood Pb concentrations at 3-5 years and teacher-rated behavioral
problems at age 6 years, sex-stratified results were similar to non-stratified results, with null associations
for conduct disorder and aggression-related outcomes (Liu et al.. 2014b). In a prospective study of Pb
exposure and self-reported aggressive behavioral characteristics, associations for some outcomes (e.g.,
"attacks people") were observed in boys (but not observed or reported for girls); the authors suggested
this may be due to lower BLLs in girls compared with boys (Naickcr et al.. 2012).

In a cross-sectional study of first grade children (mean 6.7 years) and teacher-rated behavioral
problems, sex-stratified results were generally null and similar to the non-stratified results. However,
some analyses indicated stronger associations among females (e.g., Behavioral Regulation Index (PR
[95% CI]: girls = 1.03 [1.00, 1.05]; boys = 0.99 [0.97, 1.01]), though the sample size was limited (n = 83
for girls) (Barg et al.. 2018). The authors suggested these results could be explained by teacher
expectations and perceptions of girls compared with boys, with effects on girls being more noticeable due
to gender norms and expectations rather than greater susceptibility to Pb exposure (Barg et al.. 2018).

Finally, in a study of BLLs measured at age 6.5 years and PPI between ages 19 and 24, sex-
stratified models indicated stronger associations in males, though associations were also present in
females (Bcckwith et al.. 2018). Sex-stratified analyses indicated that Pb-associated gray matter volume
loss was only present in females, while Pb-related white matter loss was more widespread in males,
including an overlap in frontal white matter loss associated with both PPI scores and BLLs (Bcckw ith et
al.. 2018).

Pre-existing Conditions

One study evaluated the association between Pb (median BLL 1.2 (ig/dL) and aggression/conduct
problems among children with CKD, a population at elevated risk of neurocognitive dysfunction (Gerson
et al.. 2006; Gipson et al.. 2004). No associations were observed (Ruebner et al.. 2019).

External Review Draft

3-117

DRAFT: Do not cite or quote

-------
3.5.3.3.3 Confounding

The 2013 Pb ISA described multiple factors that influence conduct disorder and related outcomes
including sex, race, SES, parental education, parental IQ, and quality of the caregiving environment (i.e.,
HOME score) (U.S. EPA. 2013a). These risk factors are often correlated with blood, tooth, and bone Pb
levels, and thus, are considered as potential confounding factors in epidemiologic analyses. As noted in
the 2013 Pb ISA, no single method to control for potential confounding is without limitation, and there is
potential for residual confounding by unmeasured factors. However, consistency of results across studies
utilizing different approaches to control for confounding can increase confidence across the body of
evidence.

Recent studies demonstrate associations between Pb exposure and conduct disorder after
controlling for different combinations of the aforementioned key covariates as well as additional relevant
covariates. However, it should be noted that in the current evidence base, the vast majority of studies that
identified associations did not specifically adjust for the HOME score (Liu et al.. 2022b; Tlotlcng et al..
2022; Desrochers-Couture et al.. 2019; Reuben et al.. 2019; Barg et al.. 2018; Becklev et al.. 2018;

Nkomo et al.. 2018; Rodrigues et al.. 2018; AbuShadv et al.. 2017; Boutwell et al.. 2017; Nkomo et al..
2017; Liu et al.. 2014b; Amato et al.. 2013; Sioen et al.. 2013; Boucher et al.. 2012b; Naicker et al.. 2012;
Tatsuta et al.. 2012; Nigg et al.. 2010). Yet, most of these studies did adjust for other potentially related
covariates such as social adversity, house crowding, family violence, and SES, which mitigates some of
the concern about residual confounding due to exclusion of the HOME score. Additionally, as highlighted
in the previous ISA, a meta-analysis by Marcus et al. indicated that the lack of adjustment for variables
such as SES or HOME score does not warrant limiting inferences from a particular study (U.S. EPA.
2013a; Marcus et al.. 2010).

When there is uncertainty in epidemiologic evidence due to potential confounding, it is often
helpful to consider associated toxicological data. Aggressive behavior in rodents is mediated by several
brain regions, including the hypothalamus, prefrontal cortex, dorsal raphe nucleus, nucleus accumbens,
and olfactory system (Takahashi and Miczek. 2014) along with other neurochemical systems including
neurotransmitters, neuropeptides, and neuromodulators (i.e., serotonin, dopamine, vasopressin, oxytocin,
testosterone, estrogen, corticotrophin releasing factor, opioids, neuronal nitric oxidate synthase, and
monoamine oxidase A)(Takahashi and Miczek. 2014). Pb-induced changes on many of these
neurochemical endpoints has been reported and are described in Section 3.3, which lends some limited
biological plausibility from the animal evidence without influence of potential confounding factors. While
no new studies on Pb-induced aggressive behavior in mammals were identified with BLLs of relevance to
this ISA, the previous experimental animal studies support the evidence described in the 2013 Pb ISA.

External Review Draft

3-118

DRAFT: Do not cite or quote

-------
3.5.3.3.4 Lifestages

Environmental exposures during critical lifestages spanning from childhood into adolescence can
affect key physiological systems that orchestrate brain development and plasticity (see Section 3.5.1.6.4).
Epidemiologic evidence assessed in the 2013 Pb ISA indicated associations of earlier childhood blood or
tooth Pb levels with behaviors related to conduct disorders in adolescents or adults (Fergusson et al..
2008; Wright et al.. 2008); however, these epidemiologic studies did not examine adult BLLs, thus the
relative influence of adult Pb exposure cannot be ascertained.

Recent studies observed associations of Pb exposure assessed via blood, cord blood, or bone
between delivery and age 13 years with outcomes evaluated among children, adolescents, and young
adults aged 7-33 years (Liu et al.. 2022b; Tlotleng et al.. 2022; Wright et al.. 2021; Desrochers-Couture et
al.. 2019; Reuben et al.. 2019; Beckwith et al.. 2018; Nkomo et al.. 2018; Nkomo et al.. 2017; Amato et
al.. 2013; Naicker et al.. 2012). Evidence published since 2013 is weaker for exposures that occur during
the prenatal period (Fruh et al.. 2019; Sioen et al.. 2013; Tatsuta et al.. 2012) and for most studies
assessing outcomes prior to the age of 8 years (Fruh et al.. 2019; Liu et al.. 2014b; Sioen et al.. 2013;
Tatsuta et al.. 2012). It is possible that outcome assessment tools that measure conduct disorder and
related aggressive traits are less reliable in this age group, aggressive patterns have not yet stabilized, or
the particular type of aggression associated with Pb exposure does not manifest until later years (Blair.
2001). Overall, Pb exposure during lifestages spanning childhood and into adolescence may confer risk
for conduct disorders and related outcomes.

3.5.3.3.5 Public Health Significance

The global prevalence of conduct disorders in 2019 was estimated to be 40.1 million (95% CI: 29
million, 52 million), with the highest burden experienced by individuals 0-14 years of age (GBP 2019
Mental Disorders Collaborators. 2022). Early life conduct disorders and other "antisocial behaviors" are
an important public health issue due to their persistence within an individual (Lvnam et al.. 2009). their
costs (both social and economic) to society (Sumner etal.. 2015; Mccollister et al.. 2010). and their
association with risk-taking behaviors, comorbid mental health conditions, and premature mortality
(Reves. 2015; Maughan et al.. 2014; Glenn et al.. 2013). For example, in one recent study based in New
Zealand, children with conduct problems accounted for 9.0% of the population but 53.3% of convictions,
15.7% of emergency department visits, 20.5% of prescription fills, 13.1% of injury claims, and 24.7 % of
welfare benefit months (Rivenbark et al.. 2018).

3.5.3.4 Integrated Summary and Causality Determination: Conduct Disorders

The 2013 Pb ISA concluded that the relationship between Pb exposure and conduct disorders was
"likely to be causal" (U.S. EPA. 2013a). This causality determination was primarily based on

External Review Draft

3-119

DRAFT: Do not cite or quote

-------
epidemiologic evidence. In particular, prospective cohort studies provided key evidence of the association
between blood or tooth Pb levels and 1) parent or teacher ratings of delinquent, aggressive, and antisocial
behavior (Chandramouli et al.. 2009; Dietrich et al.. 2001; Burns et al.. 1999). and 2) criminal offenses
(Fcrgusson et al.. 2008; Wright et al.. 2008) in children and adolescents across diverse locations.
Supporting evidence was provided by cross-sectional studies (Braun et al.. 2008; Chiodo et al.. 2007).

Recent epidemiologic studies support the findings from the previous ISA. The strongest evidence
published since the 2013 Lead ISA comes from prospective cohort studies of 1) self-reported conduct and
aggression-related outcomes (Tlotlcng et al.. 2022; Desrochers-Couture et al.. 2019; Beckwith et al..
2018; Nkomo et al.. 2018; Nkomo et al.. 2017). and 2) external measures of delinquency (e.g., criminal
arrests, school suspensions) (Wright et al.. 2021; Amato et al.. 2013). BLLs were <10 (ig/dL in studies of
self-reported conduct and aggression-related outcomes and higher in studies of external measures of
delinquency (e.g., (Wright et al.. 2021); mean 14.4 (.ig/dL). The studied controlled for most relevant
confounders, and the study design inherently ensured appropriate temporality between the exposure and
outcome. Additional supporting evidence comes from cross-sectional studies using either self-report or
observer-reported outcome measures among individuals aged 6-13 years with concurrent BLLs ranging
from 0.7-11.08 (ig/dL (Liu et al.. 2022b; Desrochers-Couture et al.. 2019; Reuben et al.. 2019; Boucher et
al.. 2012b; Naickeretal.. 2012; Nigg et al.. 2010). Although the evidence generally suggests positive
associations, null results may be explained by age at outcome or exposure. For example, many studies
with null associations evaluated the outcome in groups of children that included individuals <8 years of
age. It is possible that behavioral ratings are less reliable among this younger age group and/or abnormal
behaviors do not manifest until later in childhood. It should also be noted that both studies focusing
exclusively on newborn exposure (i.e., measurement of Pb in cord blood) were null, which potentially
indicates that the prenatal period may not be a relevant sensitive period of exposure for this outcome.

Despite the growing epidemiologic evidence, the central uncertainty present in the 2013 Pb ISA
database remains: there is limited and inconsistent evidence from animal toxicological studies. Available
toxicological studies of aggression were described in the 2006 Pb AQCD (U.S. EPA. 2006c) and 2013 Pb
ISA (U.S. EPA. 2013a). No new PECOS-relevant studies examining the relationship between Pb
exposure and aggression have been reported. Despite the lack of new PECOS-relevant studies, Pb-
induced changes on many neurochemical endpoints that contribute to aggressive behaviors have been
reported and are described in Section 3.3, which lends biological plausibility from the animal evidence.

In summary, despite limitations that remain in the animal toxicology database, recently published
prospective and cross-sectional epidemiological studies build on the evidence base from the 2013 Pb ISA.
Biological plausibility for these associations is supported by evidence linking early life Pb exposure to
later life volumetric reductions in gray matter in the frontal lobe and white matter in several brain regions
(Beckwith et al.. 2018). Given consistent positive associations observed across various populations and
based on multiple outcome assessment approaches at relevant Pb exposure levels, there is sufficient

External Review Draft

3-120

DRAFT: Do not cite or quote

-------
evidence to conclude that there is likely to be a causal relationship between Pb exposure and conduct
disorders, aggression, and criminal behavior.

External Review Draft

3-121

DRAFT: Do not cite or quote

-------
Table 3-4 Summary of evidence for a likely to be causal association between Pb exposure and conduct
disorders, aggression, and criminal behavior in children and adolescents.

Rationale for Causality
Determination3

Key Evidence13

References'3

Pb Biomarker Levels Associated
with Effects0

Consistent results from
epidemiologic studies
with relevant blood or
bone Pb levels,
adequate control of
relevant confounders

Evidence from prospective studies
(demonstration of a temporal sequence)
using self-report measures of aggressive or
related externalizing behavior among
individuals ages 14-24 yr in relation to
earlier average blood Pb or bone Pb

Tlotlena et al. (2022): Beckwith et al. (2018):
Nkomo etal. (2018): Nkomo et al. (2017)

Wright et al. (2021): Amato etal. (2013)

Blood Pb: age 6.5-13 mean = 5.6-
8 [jg/dL

Bone Pb: age 9 mean = 8.7 |jg/g

Evidence from prospective studies
(demonstration of a temporal sequence) of
arrests (ages 18-33 yr) and suspensions
(ages 9-10 yr) in relation to earlier average
blood Pb

Supporting evidence from cross-sectional
studies using both self-report and observer-
reported measures of aggressive or
externalizing behavior among individuals
ages 6-13 yr

Blood Pb: prenatal to age 6
mean = >10 [jg/dL

Liu et al. (2022b): Desrochers-Couture et al.
(2019): (Reuben et al.. 2019): Boucher et al.
(2012b): Naicker et al. (2012): Nigg etal. (2010)

Desrochers-Couture etal. (2019)

Blood Pb (concurrent): age 6-13
mean = 0.7-11.08 pg/dL

Supporting evidence from a mediation
analysis from a prospective cohort study
indicating indirect association of BLL on
adolescent externalizing behavior via child
externalizing behavior

Blood Pb: age 11 geometric
mean = 2.3 [jg/dL

Beckwith etal. (2018)

Supporting evidence from a prospective
study demonstrating association between
early life BLL and volumetric reductions in
gray matter in the frontal lobe and white
matter in several brain regions (mean age
26.8 yr)

Blood Pb: age 6.5 mean = 8.0 [jg/dL

External Review Draft

3-122

DRAFT: Do not cite or quote

-------
Rationale for Causality
Determination3

Key Evidence13

References'3

Pb Biomarker Levels Associated
with Effects0

Most studies had sufficient adjustment for
relevant confounders. While most did not
adjust for HOME score specifically, they did
consider other related variables, such as
income, parental IQ, parental education,
SES, and/or neighborhood safety

Evidence strongest for outcomes assessed
among children >9 years

Experimental animal Supporting evidence from animals exposed U.S. EPA (2013a)

studies with relevant prenatally and postnatally

exposures provide

coherence and help rule

out chance, bias, and

confounding with

reasonable confidence

Biological plausibility Changes in key brain regions and	U.S. EPA (2013a)

demonstrated	neurochemical systems implicated in

behavioral changes.

BLL = blood lead level; HOME = Health Outcomes and Measures of the Environment; IQ = intelligence quotient; SES = socioeconomic status; yr = year(s).

aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Table I and Table II of the Preamble to the ISAs (U.S. EPA. 2015).

bDescribes the key evidence and references, supporting or contradicting, contributing most heavily to causality determination and, where applicable, to uncertainties or
inconsistencies. References to earlier sections indicate where the full body of evidence is described.

°Describes the Pb biomarker levels at which the evidence is substantiated.

External Review Draft

3-123

DRAFT: Do not cite or quote

-------
3.5.4

Internalizing Behaviors: Anxiety and Depression in Children

The evidence evaluated in the 2013 Pb ISA was sufficient to conclude that a "causal relationship
was likely to exist" between Pb exposure and internalizing behaviors in children (U.S. EPA. 2013a).
Prospective studies in a few populations found associations of higher lifetime average blood (mean:
-14 (ig/dL) or childhood tooth (from ages 6-8 years) Pb levels with higher parent and teacher ratings of
internalizing behaviors such as withdrawn behavior and symptoms of depression and anxiety in children
aged 8-13 years. There was no strong indication of biased reporting of behaviors for children with higher
BLLs. The few cross-sectional associations in populations with mean concurrent BLLs of ~5 (ig/dL were
inconsistent. Pb-associated increases in internalizing behaviors were found with adjustment for maternal
education and SES-related variables. Consideration for potential confounding by parental caregiving
quality was inconsistent. Despite some uncertainty in the epidemiologic evidence, the biological
plausibility for the effects of Pb on internalizing behaviors was provided by a small number of
experimental animal study findings with dietary lactational Pb exposure, with some evidence at BLLs
relevant to humans. Additional toxicological evidence demonstrating Pb-induced changes in the HPA axis
and dopaminergic and gamma-aminobutyric acid (GABA) systems provided additional support. Overall,
the strongest evidence was from prospective studies in a few populations of children and the coherence
with evidence from a small number of experimental animal studies with relevant Pb exposures. Some
uncertainty related to potential confounding by parental caregiving quality remained.

Measures of central tendency for Pb biomarker levels used in each study, along with other study-
specific details, including study population characteristics and select effect estimates, are highlighted in
Table 3-10E (Epidemiologic Studies) and Table 3-7T (Toxicological Studies). An overview of the recent
evidence is provided below. Overall, recent studies generally support findings from the 2013 Pb ISA.

3.5.4.1 Epidemiologic Studies of Internalizing Behaviors in Children

Several epidemiologic studies evaluated in the 2013 Pb ISA linked biomarkers of Pb exposure in
children with internalizing behaviors characterized by directing feelings and emotions inward, i.e.,
withdrawn behavior, symptoms of depression, fearfulness, and anxiety. These studies did not clearly
indicate that Pb exposure affected a particular domain of internalizing behaviors, i.e., withdrawn
behavior, somatic symptoms, anxiety, and depression. However, a consistent pattern of associations with
BLLs was observed across ages and across multiple internalizing behaviors. The strongest evidence was
provided by prospective studies conducted across multiple locations, i.e., Boston, Port Pirie, Australia,
and Yugoslavia ("Wasserman et al.. 2001; Burns et al.. 1999; Wasserman et al.. 1998; Bellinger et al..
1994b). Collectively, these studies found associations between internalizing behaviors in children (ages
3-13 years) and Pb levels based on cord blood, concurrent blood (age 3 years), lifetime average blood,
and teeth. Moderate to high follow-up rates in most studies increased confidence that selection bias did

External Review Draft

3-124

DRAFT: Do not cite or quote

-------
not explain the pattern of associations observed in the studies. Factors, which were well documented to be
correlated with both Pb exposure and internalizing behaviors, including SES, parental caregiving quality
(speculated to mediate the potential correlation between parental psychopathology and Pb exposure), and
parental education, were considered as potential confounders across most studies. Although internalizing
behaviors are likely to have a strong familial component, the available evidence did not support parental
psychopathology as a direct confounder of the child Pb-internalizing behavior association. Studies that
included both teacher and parent ratings were emphasized. The most common instrument used to assess
internalizing behaviors was CBCL. Summary scores for internalizing behaviors, associated syndromes,
and DSM-IV scales (e.g., anxiety and depression) can be derived using CBCL.

Recent studies also analyze the association of Pb exposure with internalizing behaviors assessed
using CBCL. Using community survey data from the Project on Human Development in Chicago
Neighborhoods (PHDCN), Winter and Sampson (2017) examined the relationship between average BLL
in childhood (6 years old or younger) and anxiety or depression in adolescence (mean age 17 years old).
These authors found a 0.09 SD (0.03, 0.16) increase in anxiety or depression score, after adjustment for
covariates including caregiver education and SES. Participants were originally enrolled in the mid-1990s
and a random sample of those continuing to participate in 1999 and 2002 was randomly selected for this
study, with 67% of those selected agreeing to participate. Liu et al. (2014b) examined the association of
early childhood blood Pb concentration (3, 4, or 5 years old) with both parent and teacher ratings of
internalizing behavior at age 6 using CBCL and C-TRF, respectively. The outcomes were modeled as
both continuous and dichotomous variables (i.e., clinically significant behavior problems with T-score
>60) and adjusted for potential confounders including parent's educational level, father's occupation, and
child IQ. The emotional reactivity syndrome component of the teacher-rated internalizing problem scale
and the DSM-IV oriented anxiety were associated with child BLL when scores were modeled as
continuous terms ((3 = 0.32 [95% CI: 0.06, 0.59] and (3 = 0.25 [95% CI: 0.02, 0.50], respectively). The
ORs were 1.10 (95% CI: 1.03, 1.18) for the association of child BLL with clinically significant teacher-
reported internalizing behavior and 1.10 (95% CI: 1.01, 1.19) for clinically significant anxiety problems.
The participation rate was 81% in this study. The mean BLL of the children in this study was 6.4 (ig/dL
and the study had a high participation rate and included both teacher and parent ratings of internalizing
behavior.

Joo et al. (2018) analyzed data from the MOCEH study, a Korean prospective birth cohort of
mother-child pairs that were followed for 5 years. Maternal (early and late pregnancy), cord, and multiple
postnatal blood Pb concentrations were measured, and internalizing behaviors were assessed by the parent
using the Korean-CBCL at age five. The interaction between Pb exposure and child sex was evaluated
with further model adjustment for covariates including maternal educational level, and SES. Late
pregnancy and cord BLLs were associated with increasing internalizing behavior ratings in boys ([3 = 2.55
[95% CI: 0.22, 4.88] and [3 = 2.44 [95% CI: -0.74, 5.63], respectively), while postnatal (ages 2 and 5)
BLL was associated with increasing internalizing behavior ratings in girls ([3 = 2.94 [95% CI: 0.36, 5.52]

External Review Draft

3-125

DRAFT: Do not cite or quote

-------
and (3 = 5.65 [95% CI: 0.5, 10.8]). A total of 579 women of the 1751 originally enrolled in the cohort
provided data for this study.

Recent studies also examined the association of Pb exposure with internalizing behaviors using
SDQ. SDQ includes five scales (i.e., peer relationship problems, hyperactivity, emotional problems,
conduct problems and prosocial behavior) with the results for emotional problems discussed in this
section. Fruh et al. (2019) studied mother-child pairs participating in Project Viva, a longitudinal birth
cohort in eastern Massachusetts. Maternal blood Pb concentration in erythrocytes was measured during
the second trimester of pregnancy and parents rated their child's behavior using the SDQ (see also
Sections 3.5.1, 3.5.2, 3.5.3) in mid-childhood (median 7.7 years). The associations (i.e., (3 coefficients)
with the parent- and teacher-rated emotional components of the SDQ were 0.30 (95% CI: 0.05, 0.55) and
0.07 (95% CI; -0.22, 0.35), respectively. A stronger association with the emotional component of the
SDQ for girls compared with boys was reported by parents ((3 = 0.52 [0.18, 0.86] for girls versus [3 = 0.17
[95% CI: -0.17, 0.50] for boys). Note that the higher scores on the emotional problem scale indicate
worse performance. Behavior assessments and maternal blood Pb measurements were available for fewer
than half of study participants; however, important confounders including HOME score, maternal IQ, and
parental education were considered in this study. Sioen et al. (2013) analyzed data from a birth cohort
(FLEHS I, 2002-2006) comprising mother-infant pairs born in the Netherlands. This study examined the
association of cord blood for 281 infants whose parents returned the SDQ (26.4% response rate). No
association of cord blood Pb concentration with emotional symptom score >5 was observed (OR: 0.90
[95% CI: 0.52, 1.55] per doubling of BLL on log-scale).

Rokoff et al. (2022) used Conners' Parent and Teacher Ratings Scales (CPRS and CTRS) at age
8 years and the BASC-2 self-report of personality (SRP) at age 15 years to assess internalizing behaviors
among children enrolled in a birth cohort study in New Bedford, MA. This study examined the
association of cord blood Pb with internalizing behaviors and also considered exposure to
organochlorines (hexachlorobenzene, p,p'-dichlorodiphenyl dichloroethylene, polychlorinated biphenyls)
and Mn, which were also measured in cord blood. BKMR analysis indicated linear associations and no
interactions between cord Pb, Mn, and organochlorines. Cord blood Pb was positively associated with
BASC anxiety score at age 15 ([3 = 1.78 [95% CI: 0.58, 2.99] BASC-2 SRP anxiety score increase per
doubling Pb) but not with Conners' anxious-shy score at age 8 years. Additionally, a positive association
of cord blood Pb with depression score at age 15 was observed ([3 = 0.79 [95% CI: -0.39, 1.97]). The
Connor's psychosomatic score was positively associated with cord Pb, and this association was stronger
in boys ([3 = 2.08 [95% CI: 0.07, 4.10]) than in girls ([3 = 0.48 [95% CI: -1.00, 1.97]). A total of 528 of
the original 788 (67%) mother-infant pairs participated in the 15-year follow-up. The models were
adjusted for SES, maternal age, smoking, seafood, alcohol intake during pregnancy, maternal IQ, quality
of parental caregiving, and child characteristics (sex, race/ethnicity, age at assessment).

Several additional studies used the BASC-2 to assess associations with Pb exposure. Rasnick et
al. (2021) designed a study to identify sensitive time windows of exposure to Pb in air. These authors

External Review Draft

3-126

DRAFT: Do not cite or quote

-------
controlled for concurrent BLL (age 12 years) in their analysis of the Cincinnati Study of Allergy and Air
Pollution study data. Air Pb exposure was estimated using validated land use regression models, and
behavioral outcomes, including depression and anxiety, were assessed using the BASC-2 administered at
age 12. Models were adjusted for community deprivation, residential greenspace, and ECAT, in addition
to concurrent BLL. Distributed lag models that predicted outcome responses based on current and past
(i.e., lagged) predicted air Pb exposures identified a sensitive window in late childhood for anxiety but not
depression (Figure 3-12). The sensitive time window is indicated by months when the estimated 95% CI
did not include the null value.

BASC-2 = Behavior Assessment System for Children; edf =effective degrees of freedom.

The solid lines show the predicted change in score and the gray shading indicates the 95% CIs.

Source: Rasnick et al. (2021).

Figure 3-12 Associations of monthly airborne Pb exposure levels from birth to
age 12 with scores for anxiety and depression behaviors on the
Behavior Assessment System for Children.

Ruebner et al. (2019) evaluated the association between BLLs and attention among children with
CKD. Internalizing behavior symptoms were assessed using the parental rating scales of the BASC-2,
which includes a composite score for internalizing problems. Associations between BLL and behavioral
symptoms on BASC-2 did not persist in models that were controlled for potential confounders including
race, poverty, maternal education, and clinical factors related to CKD. The median BLL in this study was
1.2 (ig/dL.

Two additional prospective studies examined the association Pb concentration in teeth and
toenails with internalizing behavior on the BASC; these studies relied on a low proportion of the original
cohort, however. Horton et al. (2018) analyzed data from the ELEMENT Project birth cohort in Mexico
City to determine the association of weekly tooth Pb concentration (prenatal through 1 year postnatal)
with BASC-2 scores assessed between 8 and 11 years old. Approximately 12% of the original cohort was
enrolled in this study. Participants differed with respect to child birth weight and maternal IQ. A 0.4-unit

External Review Draft

3-127

DRAFT: Do not cite or quote

-------
increase in anxiety score was associated with a log-transformed unit increase in tooth Pb concentration at
12 months, while no consistent pattern of association was observed with increased internalizing behavior
symptoms overall. Dohertv et al. (2020) followed children enrolled in the New Hampshire Birth Cohort
Study (NHBCS) to examine the association of toenail Pb concentration with parent-rated internalizing
behaviors on the BASC-2. Data were available for approximately 300 of the 2000 women enrolled in the
study. No consistent pattern of association between pre- or postnatal toenail Pb concentration was
observed with internalizing behaviors after adjustment for confounders, including parent education and
parent perception of the parent-child relationship.

3.5.4.1.1 Summary

The 2013 Pb ISA included several prospective studies with moderate to high participation rates
that controlled for potential confounders including SES, parental education, and quality of parental
caregiving. These studies found associations of higher lifetime average blood (mean: -14 (ig/dL) or
childhood tooth (from ages 6-8 years) Pb levels with higher parent and teacher ratings of internalizing
behavior on the CBCL in children aged 8-13 years. Several recent longitudinal epidemiologic studies
with high to moderate participation rates, which relied on an expanded array of instruments to assess
internalizing behaviors (i.e., CBCL, SDQ, CPRS, CTRS, and BASC-2), reported associations with blood
Pb concentration (childhood average, prenatal, and postnatal BLLs <7 (ig/dL). Several studies in children
evaluated sex (Rokoff et al.. 2022; Fruh et al.. 2019; Joo et al.. 2018) as an effect modifier. The majority
of analyses controlled for important potential confounders including the quality of parental caregiving
(Rokoff et al.. 2022; Fruh et al.. 2019) maternal education and SES (Rokoff et al.. 2022; Fruh et al.. 2019;
Winter and Sampson. 2017; Liu et al.. 2014b). No association with internalizing behaviors was observed
for the blood Pb of children with CKD or in prospective studies of Pb concentration in blood (Sioen et al..
2013). teeth (Horton et al.. 2018). or toenails (Dohertv et al.. 2020). which reported relatively low
participation rates. The limited number of studies that aimed to distinguish types of internalizing
behaviors indicated associations with the anxiety component (Rokoff et al.. 2022; Rasnick et al.. 2021).

3.5.4.2 Toxicological Studies of Anxiety and Depression

Evidence in the 2013 Pb ISA consistently supported increases in emotionality in Pb-treated
animals. Postnatal exposure to Pb in female Long-Evans rats, resulting in mean BLLs between 13 and
31 (ig/dL, increased disruption, and frustration in response to errors and reward omission in
discrimination task trials (Beaudin et al.. 2007; Stangle et al.. 2007). Pb-exposed female Rhesus macaques
displayed increased negative responses to repeated tactile stimuli (i.e., tactile defensiveness) during
adolescence (mean BLLs of 31 (ig/dL) (Moore et al.. 2008). Furthermore, decreased exploratory
behaviors in the open-field test were also reported in male Wistar rats following Pb exposure from
gestation through weaning (Souza Lisboa et al.. 2005). Additional evidence for increased anxiety,

External Review Draft

3-128

DRAFT: Do not cite or quote

-------
evaluated via the elevated plus maze, was found in one study of postnatally exposed rats (mean BLLs
35 (ig/dL at \vcaning)(Fox et al.. 2010); however, no significant effects were found in another study of
postnatal Pb cxposurc(Molina et al.. 2011). Inconsistent evidence for depression-like behaviors measured
in the forced swim test was also reported (Souza Lisboa et al.. 2005; Stewart et al.. 1996).

Tests of anxiety-like behavior (i.e., emotionality) in rodents are often designed to exploit the
approach-avoidance conflict. Rodents must balance their motivation to explore novel environments (to
gather food and resources) with the need to evade predators and other threats. The open-field test (OFT)
allows for observation of rodent behavior within a bare, brightly lit, open area. Decreases in measures of
exploration (e.g., rearing, sniffing) may indicate a shift towards an anxiety-like phenotvpe.Basha et al.
(2014) found that postnatal Pb exposure in male rats decreased rearing and sniffing in the OFT apparatus
between PND 45 and 18 months, well after exposure was terminated. Grooming was also decreased at
PND 45, 4 months, and 12 months, which may indicate an altered response to stress in comparison to
controls. The same study also utilized the hole board test as another method to evaluate rodents' interest
in exploration of a novel environment. Animals displayed anxiety-like behavior (i.e., decreases in head
dip count and head dip duration) between PND 45 and 18 months. A follow-up study evaluated male
Wistar rats using a prenatal Pb exposure paradigm that resulted in BLLs of 11 (ig/dL at PND 21 and
found decreased exploratory behaviors in both the OFT and hole board test between PND 21 and
4 months (Basha and Reddv. 2015). Decreases in head dipping behavior were also reportedby Flores-
Montova and Sobin (2015). who evaluated male and female C57BL/6 mice following exposure to Pb
from PND 0 to PND 28 that resulted in low BLLs (mean between 3 and 12 (ig/dL). Further analysis of
individual BLLs and head dipping behavior suggested a negative association (i.e., head dipping behaviors
decreased as BLLs increased).

Enhanced thigmotaxis (i.e., tendency to remain close to the walls of the arena) within the OFT is
also associated with an anxiety-like phenotype. Betharia and Maher (2012) reported that low dose Pb
treatment had no significant effects on the latency of rodents to enter the center of the arena at PND 24 or
PND 59. The Sprague Dawley rats used in this study were exposed to Pb through their mothers from
gestation until PND 20 and had a mean BLL of 9 (ig/dL at PND 2, which decreased to <1 (ig/dL when
behavior was assessed. Another recent study found that adolescent Pb exposure (between PND 24 and
PND 56) in male Sprague Dawley rats, resulting in mean BLLs of 13 (ig/dL, significantly decreased the
time spent exploring the center of the arena compared with controls shortly after exposure was terminated
(Wang et al.. 2016). However, Shvachiv et al. (2018) found no significant effect of developmental Pb
exposure on adult Wistar rats using the same measure, despite employing a longer exposure paradigm that
resulted in higher BLLs than Wang et al. (2016). Interestingly, Abazvan et al. (2014) reported OFT
findings suggestive of an anxiolytic effect of Pb exposure (i.e., increased central activity and increased
rearing) in male transgenic mice that were heterozygous for mDISC 1 (associated with increased risk for
psychiatric disorders including schizophrenia) but phenotypically normal.

External Review Draft

3-129

DRAFT: Do not cite or quote

-------
Six recent studies have evaluated the potential anxiogenic effects of Pb using the elevated plus
maze (EPM). The EPM is comprised of four arms—two closed and two open (i.e., with or without
walls)—and anxious behavior is indicated by an increase in the preference for the closed arms (or
inversely, decreased preference for the open arms). Despite not finding conclusive results indicative of
increased anxiety in the OFT, Shvachiv et al. (2018) found that Pb-treated Wistar rats (male and female
adults exposed consistently or intermittently since gestation) spent significantly less time in the open arms
of the EPM. This finding was corroborated in a subsequent study by the same research group, which
investigated lifetime Pb exposure in Wistar rats and found that the percent of time animals spent in the
open arms significantly decreased at 12, 20, and 28 weeks of age (Shvachiv et al.. 2020). Interestingly,
the greatest decrease in open arm presence was observed at 20 weeks. Abazvan et al. (2014) also
demonstrated an anxiety-like phenotype using the EPM in 6-month-old male and female mice following
lifetime exposure to Pb.Tartaglionc et al. (2020) found that exposure to Pb from gestation to weaning
significantly decreased entries into the open arms, decreased head dipping behavior and stretch-attend
postures in female Wistar rats at PND 60 (mean BLLs of 25 (.ig/dL): however, only the decreases in
stretch-attend postures were observed in males. One study, Neuwirth et al. (2019a). found no significant
behavioral differences in the EPM in adolescent Long-Evans rats following gestational and
developmental exposure in either dosing group (peak BLLs 3-11 (ig/dL for lower dose group and 9-
18 (ig/dL for higher dose group).

Sobolewski et al. (2020) investigated the potential for transgenerational effects of Pb on this
endpoint by exposing female C57BL/6J mice (F0) prior to mating and during gestation, resulting in
offspring (Fl) with BLLs of 10-15 (ig/dL at PND 6-7. The developmentally exposed F1 generation was
paired with unexposed mice at PND 60 to produce the F2 generation, and the process was repeated to
produce the F3 generation which had no direct Pb exposure. F3 females spent significantly more time in
the open arms of the EPM. This effect could be further traced to descendants of the Fl sire line instead of
the Fl dam line. No significant effects were detected in F3 males.

The influence of Pb exposure on rodent behavior in the forced swim test (FST) and tail
suspension test (TST) has also been evaluated in recent studies. These tests are classically considered
models of emotional despair, with animals exhibiting both escape-directed behaviors and periods of
immobility (e.g., floating or hanging). Originally used to screen for antidepressant drugs, decreases in
immobility in the FST or TST following chemical exposure are interpreted as an antidepressant effect;
however, it was recently suggested that immobility is instead an adaptive response to the acute stress of
the FST or TST, and decreased immobility may be reflective of a maladaptive coping strategy or,
potentially, an anxiety-like phenotype (Anvan and Amir. 2018; Molendijk and de Kloet. 2015). Corv-
Slechtaetal. (2013) reported that C57BL/6 mice which had been exposed to Pb from gestation to
adulthood had significantly decreased immobile bouts in the FST compared with control animals. In
another recent study, postnatal exposure to Pb in male and female CD 1 mice significantly increased their
time spent resisting in the TST (Duan et al.. 2017). These recent results indicate that, at least under some
experimental testing paradigms (producing mean BLLs as low as roughly 6 (.ig/dL). Pb exposure results in

External Review Draft

3-130

DRAFT: Do not cite or quote

-------
what has classically been considered an antidepressant effect but may be more aptly attributed to an
altered response to stress.

3.5.4.2.1 Summary

Studies in the previous ISA consistently supported increases in emotionality in rodents and
nonhuman primates following developmental Pb exposure that produced mean BLLs as low as 13 (ig/dL.
Recent studies largely support and expand on this conclusion. Consistent decreases in rodent exploratory
behaviors in the OFT and hole board test (e.g., rearing, sniffing, head dipping) were found in Pb-exposed
rodents with peak BLLs from 3 to greater than 30 (ig/dL, lower than previously demonstrated. An
anxiety-like phenotype was also demonstrated in the EPM by multiple studies, with only one study
reporting null effects. Sobolewski et al. (2020) also demonstrated potential sex-specific transgenerational
effects of Pb exposure on this endpoint. Inconsistent effects of Pb on thigmotactic behavior were reported
by a few studies, which was not an endpoint discussed in the previous ISA. Two studies demonstrated
decreased immobility in classical tests of depression-like behavior, suggestive of an antidepressant effect,
but the relevance of these tests to human depression is unclear. While limited studies reported null results,
they were not stronger with respect to design or methodology anddid not significantly weaken the larger
body of evidence.

3.5.4.3 Relevant Issues for Interpreting the Evidence Base

3.5.4.3.1	Concentration-Response Function

Bayesian Kernel Machine Regression (BKMR) and five-chemical linear regression models were
used to examine covariate adjusted associations between Pb exposure and CPRS Anxious-Shy T-score at
age 8 and BASC-second revision Anxiety T-score at age 15 Rokoff et al. (2022). BKMR analysis
indicated linear associations between Pb exposure and these outcomes, and no interactions between cord
Pb, Mn, and organochlorines.

3.5.4.3.2	Potentially At-Risk Populations

The 2013 Pb ISA did not describe populations of children potentially at higher risk of Pb-
associated internalizing behaviors. Recent epidemiologic studies presented sex-stratified results or
examined interactions between Pb exposure and other chemicals.

External Review Draft

3-131

DRAFT: Do not cite or quote

-------
Sex

Fruh et al. (2019) studied mother-child pairs participating in Project Viva, a longitudinal birth
cohort in eastern Massachusetts. This study found a stronger association of maternal BLL with the
emotional component of the SDQ measured in mid-childhood for girls compared with boys ((3 = 0.52
[0.18, 0.86] for girls v. (3 = 0.17 [95% CI: -0.17, 0.50] for boys). Note that higher scores on the emotional
problem scale indicate worse performance. In another study, Joo et al. (2018) found that late pregnancy
and cord BLL was associated with increasing internalizing behavior ratings on the CBCL in boys
([3 = 2.55 [95% CI: 0.22, 4.88] and [3 = 2.44 [95% CI: -0.74, 5.63], respectively), while postnatal (age 2
and 5) BLL was associated with increasing internalizing behavior ratings on the CBCL in girls ([3 = 2.94
[95% CI: 0.36, 5.52] and [3 = 5.65 [95% CI: 0.5, 10.8]). In a study that used Conners' rating scale to
ascertain internalizing behaviors, Rokoff et al. (2022) found the psychosomatic score was positively
associated with cord Pb and this association was stronger in boys than in girls ([3 = 2.08 [95% CI: 0.07,
4.10] versus [3 = 0.48 [95% CI: -1.00, 1.97]). Of the experimental animal studies that evaluated both
sexes, a small number identified behavioral changes in Pb-exposed females while detecting minimal or no
changes in their male counterparts on the EPM (Sobolewski et al.. 2020; Tartaglione et al.. 2020).

Overall, no consistent pattern was observed across the limited number of epidemiologic and toxicologic
studies that presented sex-stratified results. Each study used a different instrument to ascertain the
outcomes.

Other Metals

A recent study examined the interaction effect between prenatal Pb exposure and other metals on
internalizing behavior scores on the BASC and the CPRS. BKMR analysis indicated no interactions
between cord blood Pb, Mn, and organochlorines that would indicate a deviation from additivity in a
study by Rokoff et al. (2022).

3.5.4.3.3 Lifestages

Epidemiologic studies consistently show that BLLs measured during various lifestages and time
periods, including the prenatal period, early childhood, and later childhood, and averaged over multiple
years, are associated with increases in internalizing behaviors. The identification of critical lifestages and
time periods of Pb exposure is complicated further by the fact that BLLs in older children, although
affected by recent exposure, are also influenced by Pb stored in bone due to rapid growth-related bone
turnover in children relative to adults. Thus, associations of neurodevelopmental effects with concurrent
BLL in children may reflect the effects of past and recent Pb exposures. Recent prospective studies add to
the evidence from the strongest studies in the 2013 Pb ISA that found associations with childhood
average blood and tooth Pb levels in children. These recent studies found associations between
internalizing behaviors and early childhood, maternal, and cord BLLs. Toxicological studies also provide

External Review Draft

3-132

DRAFT: Do not cite or quote

-------
support that the sensitive exposure window is not limited to a single phase of development. Rather,
effects of Pb exposure on anxiety or depression-like behavior in animals have been found following
gestational and postnatal exposure, exposure starting in adolescence, and lifetime exposure.

3.5.4.4 Integrated Summary of Internalizing Behaviors in Children

The 2013 Pb ISA concluded that a causal relationship was likely to exist between Pb exposure in
children and internalizing behaviors based on the available evidence (U.S. EPA. 2013a). Prospective
studies demonstrated associations between higher average blood (roughly 14 (ig/dL) or tooth Pb levels
and higher parent and teacher ratings of internalizing behaviors, including withdrawn behavior and
symptoms of depression, fearfulness, and anxiety in children (aged 8-13). These associations were
present after adjustment for SES, birth outcomes, and parental education, but some uncertainty regarding
potential confounding by parental caregiving quality remained. Results from cross-sectional studies
evaluating lower concurrent BLLs (5 (ig/dL) were inconsistent. Increased emotionality in rodents and
monkeys was demonstrated at BLLs as low as 13 (ig/dL after exposure to Pb during development, and
biological plausibility was supported by findings of alterations in the HPA axis and dopaminergic and
GABAergic systems.

Several recent longitudinal epidemiologic studies with high to moderate participation rates relied
on an expanded array of instruments to assess internalizing behaviors (i.e., CBCL, SDQ, PRS, CTRS, and
BASC-2) compared with the studies in the 2013 Pb ISA. These studies observed associations with blood
Pb exposure (early childhood and prenatal BLLs <7 (ig/dL). A limited number of studies evaluated child
sex (Rokoff et al.. 2022; Fruh et al.. 2019; Joo et al.. 2018) as an effect modifier but were not consistent
with regard to sex-specific effects. The majority of analyses controlled for important potential
confounders including the quality of parental caregiving (Rokoff et al.. 2022; Fruh et al.. 2019). maternal
education, and SES (Rokoff et al.. 2022; Fruh et al.. 2019; Winter and Sampson. 2017; Liu et al.. 2014b);
however, each potential confounder was not uniformly considered across studies. No association between
blood Pb and internalizing behaviors was observed among children with CKD or in prospective studies of
Pb concentration in blood (Sioen et al.. 2013). teeth (Horton et al.. 2018) or toenails (Dohertv et al..
2020). which reported relatively low participation rates. The limited number of studies that aimed to
distinguish types of internalizing behaviors indicated associations with the anxiety component (Rokoff et
al.. 2022; Rasnick et al.. 2021). Recent studies that found associations with prenatal or cord BLLs add to
the evidence. Uncertainty remains, however, regarding the exposure patterns associated with prenatal
BLLs as well as BLLs in older children and adults.

Recent experimental animal studies provide coherence with the previous findings that moderate
to high peak BLLs (12 to >30 (ig/dL) increase anxiety-like behaviors on the EPM, hole board test, and
OFT following Pb exposure during a single developmental window (including prenatal (Basha and
Reddv. 2015). postnatal (Basha etal.. 2014). or adolescent periods(Wang et al.. 2016)) or throughout

External Review Draft

3-133

DRAFT: Do not cite or quote

-------
development and beyond (Shvachiv et al.. 2020; Tartaglione et al.. 2020; Shvachiv et al.. 2018; Abazvan
et al.. 2014). Overall, experimental animal studies provide more extensive support for anxiety-like
behaviors than for depression-like behaviors. However, two recent studies reported that Pb exposure
decreased immobility in classical tests of emotional despair following postnatal or lifetime Pb
exposure(Duan et al.. 2017; Corv-Slechta et al.. 2013). In addition to the well demonstrated effects at
moderate to high BLLs, two recent studies found altered behaviors in a nose poke task and FST following
Pb exposures resulting in low BLLs (3.2-10 (.ig/dL); moreover, one study was able to demonstrate
exposure-response relationships (i.e., higher BLLs were associated with greater behavioral
changcsHFlorcs-Montova and Sobin. 2015).

In summary, recent prospective epidemiologic studies add to previous evidence of a positive
association with higher average blood (prenatal, early childhood, lifetime) or childhood tooth Pb levels
with multiple measures of internalizing behaviors in children (aged 4-17). While multiple confounding
factors were accounted for in these associations, the studies did not uniformly adjust for parental
caregiving quality. Additional uncertainty regarding the exposure patterns associated with the prenatal
BLLs as well as BLLs in older children and adults remains. Limited cross-sectional studies reviewed in
the previous ISA produced inconsistent findings in populations with BLLs <5 (ig/dL; these
inconsistencies have not been addressed in recent studies. Recent toxicological studies strengthen the
overall evidence base, providing further support for anxiety-like behaviors following developmental and
cumulative exposures that result in BLLs that are relevant to humans. Overall, the evidence is sufficient
to conclude that there is likely to be a causal relationship between Pb exposure and internalizing
behaviors in children.

External Review Draft

3-134

DRAFT: Do not cite or quote

-------
Table 3-5 Summary of evidence for a likely to be causal relationship between Pb exposure and internalizing
behaviors in children.

Rationale for Causality
Determination3

Key Evidence13

References'3

Pb Biomarker Levels Associated with
Effects0

Epidemiologic Studies

Consistent results from
prospective epidemiologic
studies with relevant
exposures

Evidence from prospective studies for
higher ratings of internalizing behaviors in
children ages 8-13 yr in Boston and Port
Pirie cohorts in association with tooth or
lifetime average BLLs.

Section 4.3.4.1, (U.S. EPA. 2013a)
Burns et al. (1999)

Bellinger et al. (1994b)

Blood Pb lifetime (to age 11-13 yr)
average mean: -14 [jg/dL

Tooth Pb (age 6 yr) mean: 3.4 |jg/g

Evidence from prospective studies for
higher rating of internalizing behaviors in
children 6-17 yr (cohorts in eastern MA,
Chicago, Cincinnati, and China) in
association with early childhood and
prenatal BLLs.

Winter and Sampson (2017)
Liu et al. (2014b)

Fruh et al. (2019)

Rokoff et al. (2022)

Early childhood <7 [jg/dL (median/mean)

Maternal and cord blood Pb, <2 [jg/dL
(median)

Associations also found in children aged
4-5 yr in former Yugoslavia in
association with lifetime average BLL

Wasserman etal. (2001)

Blood Pb lifetime (to age 4-5 yr) average
mean: 7.2 [jg/dL

Prospective studies had population-
based recruitment with moderate follow-
up participation. Participation not
conditional on tooth/BLLs and behavior

Inconsistent results in cross-sectional
studies with mean BLLs < 5

Section 4.3.4.1, (U.S. EPA. 2013a)

External Review Draft

3-135

DRAFT: Do not cite or quote

-------
Rationale for Causality
Determination3

Key Evidence13

References'3

Pb Biomarker Levels Associated with
Effects0

Uncertainty regarding
potential confounding

Epidemiologic associations found with
adjustment for SES, birth outcomes,
parental education. Studies did not
uniformly adjust for parental caregiving
quality.

Section 3.7, Table 3-10E



Uncertainty regarding the
exposure patterns associated
with observed BLLs.

Uncertainty in regarding past exposure
among women in studies reporting
associations with maternal or cord BLL

Fruh etal. (2019)
Rokoffetal. (2022)



Toxicological Studies

And, supporting animal
evidence with relevant
exposures from multiple
studies

Gestational, lactational, and adolescent
exposures increasing anxiety-like
behaviors and altered stress coping
response.

Corv-Slechta etal. (2013)
Flores-Montova and Sobin (2015)
Shvachiv et al. (2020)

Peak BLLs: 3-27 pg/dL

BLL = blood lead level; Pb = lead; yr = year(s); SES = socioeconomic status.

aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Table I and Table II of the Preamble to the ISAs (U.S. EPA. 2015).

bDescribes the key evidence and references, supporting or contradicting, contributing most heavily to causality determination and, where applicable, to uncertainties or
inconsistencies. References to earlier sections indicate where the full body of evidence is described.

°Describes the Pb biomarker levels at which the evidence is substantiated.

External Review Draft

3-136

DRAFT: Do not cite or quote

-------
3.5.5

Motor Function in Children

The evidence assessed in the 2013 Pb ISA is sufficient to conclude that a "causal relationship is
likely to exist" between Pb exposure and decrements in motor function in children. Evidence from
prospective studies of Cincinnati and Yugoslavia birth cohorts indicated associations of decrements in
fine and gross motor function with higher neonatal, concurrent, and lifetime average BLLs in children
aged 4.5-6 years and with higher earlier childhood (ages 0-5 years on average, age 78 months) BLLs in
children aged 15-17 years (Bhattacharva et al.. 2006; Ris et al.. 2004; Bhattacharva et al.. 1995; Dietrich
et al.. 1993). The means for these blood Pb metrics ranged from 4.8 to 12 (ig/dL. These associations were
found with adjustment for several potential confounding factors, including SES, parental caregiving
quality, and child health with no indication of substantial selection bias. Evidence from cross-sectional
studies was less consistent, however (see Section 4.3.8 of (U.S. EPA. 2013a)'). The biological plausibility
for associations observed in children was supported by a study that found poorer balance in male mice
with relevant gestational to early postnatal (PND 10) Pb exposures. Overall, the strongest evidence was
from a small number of prospective cohort studies of children with limited support from studies in mice
with relevant exposures.

Measures of central tendency for Pb biomarker levels used in each study, along with other study-
specific details, including study population characteristics and select effect estimates, are highlighted in
Table 3-1 IE (Epidemiologic Studies) and Table 3-1 IT (Toxicological Studies). An overview of the recent
evidence is provided below. Overall, recent epidemiologic studies support findings from the 2013 Pb ISA
and a limited number of recent experimental animal studies provide coherence for their observations
demonstrating effects at relevant exposure concentrations.

3.5.5.1 Epidemiologic Studies of Motor Function

Several epidemiologic studies examined the association between Pb exposure and decrements in
motor function in children. The findings generally support an association between Pb exposure and
decrements in motor function; however, they varied by the specific measure of motor function as well as
the timing of exposure measurement. Most studies were cohort studies and assessed motor function using
a comprehensive motor score, such as the Psychomotor Development Index (PDI) score, from a version
of the BSID (Jiang et al.. 2022; Kao et al.. 2021; Rygiel et al.. 2021; Shekhawat et al.. 2021; Kim et al..
2018b; Y Ortiz et al.. 2017; Paraiuli et al.. 2015b; Paraiuli et al.. 2015a; Liu et al.. 2014c; Kim et al..
2013c; Henn et al.. 2012). A few studies used a motor score from the Chinese version of the GDS (Liu et
al.. 2022a; Zhou et al.. 2017). The remaining studies assessed specific tasks, such as balance, manual
dexterity, coordination, and fine motor speed (Taylor et al.. 2018; Boucher et al.. 2016; Taylor et al..
2015).

External Review Draft

3-137

DRAFT: Do not cite or quote

-------
Studies using the Bayley scales to measure motor function in infants and toddlers (i.e., through
age 3) generally found associations between some Pb exposure metrics and decreased motor score. Kim et
al. (2013c). Kim et al. (2018b). Y Ortiz et al. (2017). Liu et al. (2014c). Rygiel et al. (2021). and
Shekhawat et al. (2021) observed a decrease in motor score using maternal or cord BLLs, as well as other
blood Pb metrics, in several birth cohorts in multiple countries. Associations between BLLs and PDI are
presented in Figure 3-13.

External Review Draft

3-138

DRAFT: Do not cite or quote

-------
Study

Prospective Studies

fKim etal. 2013

fKim etal.2018

tLiu etal.2014
TRygiel etal.2021

Location

3 Cities, S Korea

4 Cities, S Korea

Guangdong, China
Mexico City', Mexico

Claus Henn et al. 2012 Mexico City, Mexico

Blood Pb

Prenatal (late pregnancy)
Prenatal (early pregnancy)
Prenatal (late pregnancy)
Prenatal (early pregnancy)
Prenatal (late pregnancy)
Prenatal (late pregnancy)
Prenatal (late pregnancy)
Prenatal (late pregnancy)
Prenatal (cord)

Prenatal (T1)

Prenatal (T2)

Prenatal (T3)

Prenatal (T1)

Prenatal (T2)

Prenatal (T3)

Child (12 months)

Child (24 months)

Mean

(ng/dL)

Age at Outcome

(months)

Strata

1.4 (GM)

6

1.3 (GM)

6

1.4 (GM)

6

1.3 (GM)

6

1.4 (GM)

6

1.3 (GM)

6

2.7 (median)

13-24

NR

13-24

NR

13-24

5.63 (ref: 1.35)

36

5.27 (GM)

12

4.74 (GM)

12

4.98 (GM)

12

5.27 (GM)

24

4.74 (GM)

24

4.98 (GM)

24

5.1

12-36

5

12-36

Cd<1.47Mg/L
Cd >1.47 (jg/L
Cd <1.51 pg/L
Cd >1.51 pg/L

Boys	*

Girls	"

Pb £3.92 vs. *1.89 pg/dL

	1	1	1	1	

-15.00	-10.00	-5.00	0.00

Beta values (95% CI) per 1 ug/dL increase in blood Pb

Figure 3-13 Associations between biomarkers of Pb exposure and Bayley Score of Infant Development
Psychomotor Development Index

Note: Effect estimates are standardized to a 1 |jg/dL increase in blood Pb or a 10 |jg/g increase in bone Pb. If the Pb biomarker is log-transformed, effect estimates are standardized to
the specified unit increase for the 10th -90th percentile interval of the biomarker level. Effect estimates are assumed to be linear within the evaluated interval. Categorical effect
estimates are not standardized. Associations that could not be standardized are not included on the plot.
fStudies published since the 2013 Integrated Science Assessment for Lead.

External Review Draft

3-139

DRAFT: Do not cite or quote

-------
Kim et al. (2013c) found that PDI score at 6 months of age decreased with increasing BLLs
measured in the third trimester (median = 39th week) ((3 = -1.38 [95% CI: -3.31, 0.55] per 1 (ig/dL
increase in BLL) in the Korean MOCEH study. Another Korean study using the CHECK cohort (Kim et
al.. 2018b) also observed decrements in PDI among 13-24 month old infants in association with perinatal
maternal BLLs ((3 per 1-jxg/dL blood Pb = -15.45 [95% CI: 30.12, -0.79]). Sex-stratified results were
slightly negative but not significant. In China, Liu et al. (2014c) observed an association between
increasing prenatal (umbilical cord blood) Pb levels and worse PDI score at 36 months of age. Compared
with low prenatal Pb (<1.89 (.ig/dL). children exposed to high prenatal Pb (>3.92 (ig/dL) were more likely
to have a lower PDI score ([3 = -1.30 [95% CI: -1.57, -1.03]), after adjusting for potential confounders.

Several studies in Mexico also examined the association of BLL with PDI assessed in infants.
Rvgiel et al. (2021) found a small negative association between prenatal (trimester-specific) BLLs and
PDI scores at 12 months in the ELEMENT Project study ([3 per 1 (ig/dL increase in 1st trimester
Pb = -0.24 [95% CI: -0.95, 0.48]; [3 per 1 (ig/dL increase in 2nd trimester Pb = -0.38 [95% CI: -1.10,
0.35]; [3 per 1 (ig/dL increase in 3rd trimester Pb = -0.33 [95% CI: -1.06, 0.40]). At 24 months, the
negative association persisted but with a smaller magnitude of effect. Rvgiel et al. (2021) also examined
whether DNA methylation mediated the association and found that DNA methylation of cgl8515027
located within glucosaminyl (N-acetyl) transferase 1 (GCNT1) had a suppressive (positive indirect) effect
on the inverse relationship between second trimester BLLs (ln-transformed) and PDI scores at 12 months
([3indirect = 1.25 (95% CI: -0.11, 3.32]). In the Programming Research in Obesity, Growth, Environment
and Social Stressors (PROGRESS) birth cohort in Mexico, Y Ortiz et al. (2017) found a negative
association between motor score at 24 months of age and log-transformed BLLs measured during the
third trimester ([3 = -11.01 [95% CI: -17.55, -4.48]), but not for BLLs measured during the second
trimester ([3 = 1.97 [95% CI: -2.46, 6.40]). In a study using childhood BLLs, Henn et al. (2012) found a
small negative association with repeated measures of PDI scores in another cohort of children in Mexico.
From adjusted mixed-effects models with repeated measures of PDI scores at 12, 18, 24, 30, and
36 months, there was a negative association between 12-month blood Pb and PDI scores ([3 per 1-fxg/dL
blood Pb = -0.27 [95% CI: -0.56, 0.02]). Similarly, from adjusted mixed-effects models with repeated
measures of PDI scores at 24, 30, and 36 months, there was a negative association between 24-month
blood Pb and PDI scores ([3 per 1-fxg/dL blood Pb = -0.18 [95% CI: -0.53, 0.17]).

Several studies are not pictured in Figure 3-13. Shekhawat et al. (2021) found that children with
cord blood Pb concentrations of 5-10 (ig/dL had reduced gross motor skills on the BSID at an average
age of 6.5 months ([3 = -0.29 [95% CI: -5.00, 0.11]) for each 1 (ig/dL increase in cord BLL. Additionally,
in a birth cohort of mother-child pairs recruited from Bharatpur General Hospital in Nepal, Paraiuli et al.
(2015a) and Paraiuli et al. (2015b) assessed the association of cord BLLs with PDI at 24 months old and
36 months of age, respectively. Negative but non-significant associations were observed between log-
transformed cord BLLs and 24-month PDI ([3 = -4.83 [95% CI: -16.53, 6.86]) nor 36-month PDI ([3 =
-2.56 [95% CI: -9.71, 4.59]). Notably, two additional studies that used biomarkers other than blood did
not find associations, i.e., Jiang et al. (2022) measured Pb in meconium (at birth) and in hair and

External Review Draft

3-140

DRAFT: Do not cite or quote

-------
fingernails (at 3 years of age) in Taiwan and did not find an association with any motor score (total, fine
motor, or gross motor) at 3 years of age. Another study in Taiwan (Kao et al.. 2021) that used hair and
fingernail biomarkers of Pb concentrations similarly did not report significant associations with motor
development among infants less than 3 years old.

Several other instruments were used to assess motor function in infants and toddlers. Paraiuli et
al. (2013) measured Pb, As, and Zn levels in cord blood and used the third edition of the Brazelton
Neonatal Behavioral Assessment Scales (NBAS III) to assess neurodevelopment in one-day-old newborns
in Chitwan, Nepal. The NBAS III contains 27 behavioral and 18 reflex items and is used for infants up to
2 months old. The multivariate model was adjusted for parity, family income, mother's age, education,
BMI, birth weight, gestational age, and age in hours at NBAS assessment. The NBAS motor cluster score
was inversely associated with the log-transformed cord BLLs ((3 = -2.15 [95% CI: -4.27, -0.03]). Liu et
al. (2014d) used the Neonatal Behavioral Neurological Assessment (NBNA), which is based on the
NBAS and has five clusters of behavior: passive tone, active tone, primary reflexes, and general
assessment. Newborns in this study were assessed at 3 days old, and the NBNA has been validated among
Chinese newborns between 2 and 28 days old. Associations between maternal BLL in the first trimester
and the NBNA scores were observed ((3 = -4.86 [95% CI: -8.83, -0.89] per unit of log-transformed Pb).
Less precise associations of second trimester, third trimester, and cord BLLs with decreased motor
function were also observed. Among toddlers (2-3 years old) Zhou et al. (2017) and Liu et al. (2022a)
both used the Chinese version of the GDS to calculate a motor score. For every log 10 (|ig/dL) increase in
maternal blood Pb (measured at 28-36 weeks of gestation), Zhou et al. (2017) observed a positive
association for gross motor development ([3 = 3.31 [95% CI: -6.11, 12.73] per log-10 transformed unit of
BLL) as well as fine motor development ([3 = 0.49 [95% CI: -11.27, 12.24] per log-10 transformed unit
of BLL); however, the effect estimates were extremely imprecise. On the other hand, for each In (|ig/L)
increase in maternal Pb, Liu et al. (2022a) observed a negative association for gross motor development
([3 = -2.32 [95% CI: -3.61, -1.03] per ln-transformed unit of BLL). Furthermore, Nvanza et al. (2021)
did not find associations between high Pb exposure and fine or gross motor impairment assessed by the
MDAT.

Several additional studies were conducted using assessment instruments that measure children's
(7 years or older) ability to perform certain tasks. In the ALSPAC, Taylor et al. (2015) conducted a heel-
to-toe test in children at age 7 years, beam walking test (to measure dynamic balance) at age 10 years, and
balancing test with eyes closed (to measure static balance) also at age 10 years. Pb levels measured in
maternal blood (<18 weeks of gestation) and Pb levels measured in child blood (30 months old) were not
associated with any measure of motor function in this study (Taylor et al.. 2015). In another analysis of
ALSPAC data, Taylor et al. (2018) examined the association between first trimester BLLs and different
measures of coordination. Compared with prenatal blood Pb <5 (ig/dL, children exposed to higher levels
(>5 (.ig/dL) of prenatal Pb were more likely to fail the tests of manual dexterity (threading lace, peg board
using preferred hand, and peg board using non-preferred hand). When comparing the highest blood Pb
quartile to the lowest blood Pb quartile, the only association remaining was for failing the peg board using

External Review Draft

3-141

DRAFT: Do not cite or quote

-------
the preferred hand (OR for quartile 4 versus quartile 1 = 1.23 [95% CI: 0.92, 1.66]). Prenatal Pb exposure
was not associated with tests of balance and the results were inconsistent for ball skills (inverse
association for >5 (ig/dL versus <5 (ig/dL; positive association for quartile 4 versus quartile 1). In the
Nunavik Child Development Study in Canada, Boucher et al. (2016) measured manual dexterity, fine
motor speed, and visuomotor integration in children (ages 8.5-13.3 years). BLLs (log-transformed)
measured at birth (cord blood) and at age 11 years were negatively associated with manual dexterity ((3
for cord blood Pb = -0.08 [p > 0.10]; (3 for child blood Pb = -0.17 [95% CI: -0.34, 0.00]) and fine motor
speed ([3 for cord blood Pb = -0.19 [95% CI: -0.33, -0.05]; [3 for child blood Pb = -0.21 [95% CI: -0.37,
-0.05]). For visuomotor integration, there was no association with cord blood Pb ([3 for cord blood
Pb = -0.01 [p > 0.10]) and a positive association with child blood Pb ([3 for child blood Pb = 0.10
[p > 0.10]). The magnitude of effect was greater for child BLLs. Nozadi et al. (2021) collected blood
samples from pregnant mothers at the 36-week visit or at the time of delivery and administered the ASQ:I
at 10-13 months of age to evaluate communication, gross motor, fine motor, problem-solving, and
personal-social development. A 1-jxg/dL increase in prenatal blood Pb was associated with a decrease in
fine motor ([3 = -0.63 [95% CI: -1.19, -0.08]) scores. Palaniappan et al. (2011) observed decrements of
WRAVMA scores in association with 1-fxg/dL increase in concurrent BLLs (Drawing: [3 = -0.29 [95%
CI: -0.51, -0.07]; Matching: [3 = -0.14 [95% CI: -0.31, 0.02]; Pegboard: [3 = -0.19 [95% CI: -0.38,
0.01]; Composite: [3 = -0.26 [95% CI: -0.45, -0.07]).

3.5.5.1.1 Summary

Evidence from prospective studies of Cincinnati and Yugoslavia birth cohorts indicated
associations of decrements in fine and gross motor function with higher neonatal, concurrent, and lifetime
average BLLs in young children with higher earlier childhood BLLs. Several recent birth cohort studies
observed lower scores on the Bayley PDI in association with maternal Pb exposure (no clear pattern by
trimester of pregnancy), cord BLL, and postnatal concurrent blood Pb (Rygiel et al.. 2021; Y Ortiz et al..
2017; Liu et al.. 2014c; Kim et al.. 2013c; Henn et al.. 2012). Pb-associated decrements in motor function
were observed in neonates (Liu et al.. 2014d; Paraiuli et al.. 2013) and in some but not all studies of
toddlers that assessed motor function using GDS (Liu et al.. 2022a; Zhou et al.. 2017) or children's
(greater than 7 years old) abilities to perform certain tasks indicative of gross motor function (i.e.,
balance) (Taylor et al.. 2015). although associations with fine motor function were observed (Taylor et al..
2018; Boucher et al.. 2016).

3.5.5.2 Toxicological Studies of Motor Function

As described above and in previous reviews (U.S. EPA. 2013a. 2006c'). epidemiologic studies
provide evidence of associations between Pb exposures and fine and gross motor decrements, mainly in
children. Evaluating performance in neurobehavioral toxicological studies with Pb exposure in rodents

External Review Draft

3-142

DRAFT: Do not cite or quote

-------
can substantiate observed Pb exposure effects on motor function seen in humans. In past assessments,
evidence in animal toxicological studies has been limited due to a lack of investigations with relevant Pb
exposures. The purpose of this section is to update the collection of evidence available concerning Pb
exposure-induced effects on motor function in animal models. Studies that examined various indices of
locomotor activity are evaluated above in the Toxicological Studies of Hyperactivity section
(Section 3.5.2.3.2).

Previous ISAs (U.S. EPA. 2013a. 2006c) highlighted rotarod and air righting reflex experiments
with rodents to discuss the effects of Pb on development of motor coordination and balance. Typical
rotarod tests compare the latency to fall for subjects placed on a rotating rod. Falling off more quickly
indicates decreased coordination and/or balance, wo rotarod studies discussed in previous EPA reviews
described effects of developmental Pb exposure on rotarod performance that resulted in relevant BLLs
less than 30 (ig/dL. Interestingly, Moreira et al. (2001) saw no effect of Pb exposure, from the beginning
of gestation through lactation, on Wistar rat rotarod performance at PND 70 with PND 23 mean BLLs of
21 (ig/dL. In contrast, Leasure et al. (2008) observed substandard performance in pregestational through
lactation Pb-exposed male, but not female, mice with peak BLLs of less than 10 (ig/dL. Since Leasure et
al. (2008). no other PECOS-relevant studies have assessed rotarod performance in rodents exposed to Pb
throughout the entire developmental period. Two recent rotarod studies with mice, by Flores-Montova
and Sobin (2015) and Zou et al. (2015). showed no decrements in performance in rotarod tests after
postnatal-only exposure to Pb in drinking water for PND 0-28 and 37-58 for respective studies.

While the outcomes of these two latest rotarod studies were mostly negative, additional
investigations evaluating the effects of developmental Pb exposure on coordination and balance in
neonatal rats using surface righting reflex, negative geotaxis reflex, and ascending wire mesh tests yielded
mixed results. Surface righting reflex tests are run by placing pups in a supine position and then recording
the time it takes to flip onto their feet. Slower times to flip indicate postural imbalances. For negative
geotaxis reflex, or slant-board tests, pups are placed on a slanted board and the time it takes for the pup to
face upward is recorded. Slower times to turn upward indicate that the vestibular response to gravity cues
or motor coordination required for turning are underdeveloped. Success in ascending wire mesh tests also
requires coordination, as the animals are required to climb to the top of a mesh out of a water bath in a
predetermined period of time. In a study comparing the developmental effects on male Wistar rats with
pregestational, gestational, or lactational Pb exposure, pups exposed during gestation achieved negative
geotaxis significantly faster than unexposed counterparts when tested on PND 8, 10, and 12 (Rao Barkur
and Bairv. 2016). In contrast, in the same study, Rao Barkur and Bairv (2016) observed no difference in
negative geotaxis times between control, pregestation alone, and lactation alone Pb-exposed pups. No
effects on surface righting reflex on PND 3 through 5 were observed for pups belonging to the previously
mentioned exposure groups. The day of achievement in ascending wire mesh tests (PND 14-18) was
delayed for animals in both gestation and lactation Pb-exposed groups but not for those in the
pregestational group (Rao Barkur and Bairv. 2016). Betharia and Maher (2012) exposed pregnant
Sprague Dawley rats to Pb (II) acetate trihydrate via drinking water from the beginning of gestation

External Review Draft

3-143

DRAFT: Do not cite or quote

-------
through lactation and until weaning. Development of the surface righting reflex of control and exposed
offspring was tested from PND 1 to 10. Slower righting times were observed for Pb-exposed offspring on
PND 1; however, from PND 2 through 10, there were no differences between Pb-exposed and control
groups. Basha and Reddv (2015) observed a significant increase in righting time in righting reflex tests
done on PND 6 and 7 and an increase in latency to turn in negative geotaxis tests for male Wistar rats
tested on PND 8, 9, and 10 after in utero exposure to Pb. Tartaglionc et al. (2020) saw no decrements in
righting reflex time or negative geotaxis achievement on PND 4, 7, 10, and 12 from Pb exposure in the
offspring of dams exposed to Pb from 1 month pre-mating to offspring weaning.

Additional motor function experiments with early postnatal weaning in Wistar rats were carried
out in studies with developmental Pb exposures. Tartaglione et al. (2020) recorded on PND 4, 7, 10, and
12 the duration of neonatal motor patterns of rat pups from dams exposed to Pb before mating until
offspring weaning. On PND 10, Pb-exposed pups spent less time in locomotion compared with controls,
in favor of head rising and wall climbing movements, demonstrating a stereotyped/preservative profile.
Basha and Reddv (2015) observed a prenatal Pb-induced strength deficit when rats were subject to
forelimb hang tests on PND 13, 14, 15, and 16 but not on day 12. This indicated Pb-induced
underdevelopment of fine motor ability. Rao Barkur and Bairv (2016) tested rats on PND 6, 8, 10, and 12
for Pb-induced effects on swimming development. They observed no difference in swimming body angle
or limb movements for ISA-relevant pregestation, gestation, or lactation-exposed groups compared with
control. These novel studies warrant further investigation into the effects of Pb exposure at different
concentrations and stages of development on neonatal movement patterns and forelimb hang tests.

3.5.5.2.1 Summary

The evidence supporting the link between developmental Pb exposure and deficits in motor
function in animal models has expanded on account of recent studies utilizing Pb-exposed rodents with
mean BLLs <30 (ig/dL. These new studies illustrate the effects of Pb exposure on both gross and fine
motor development in novel paradigms. In addition to the effect on rotarod performance (Leasure et al..
2008) described in the previous ISA, developmental Pb-induced decrements in righting reflex, negative
geotaxis reflex (Basha and Reddv. 2015). ascending wire mesh (Rao Barkur and Bairv. 2016). and
forelimb hang tests (Basha and Reddv. 2015) were observed. Interestingly, gestational Pb exposure was
present among each type of study that yielded decrements in these measurements of motor function;
therefore, it may be a more sensitive window compared with lactation or postnatal exposures. In terms of
design or methodology, studies that found weak or null relationships were not stronger and did not
weaken the overall body of corroborating data. Key aspects such as exposure levels and timing, ages of
animals at testing, and slant-board angles were variable between the few relevant studies. Altogether, the
results from these recent studies support the conclusions from the previous ISA. However, due to the
limited number of reproduced experiments, these recent studies do not enhance the consistency of the
evidence. In addition to the fine motor, motor reflex, and coordination, and balance studies described in

External Review Draft

3-144

DRAFT: Do not cite or quote

-------
this section, effects of developmental Pb exposure on locomotor activity are evaluated separately in the
Toxicological Studies of Hyperactivity section (Section 3.5.2.3.2) above. Briefly, due to heterogeneity in
study design, the evidence for effects of developmental Pb on locomotor activity is mixed; however, a set
of four independent studies with analogous conditions showed hyperactivity in rodents when tested within
a PND 14 to 23 window after lactational Pb exposure (Duan et al.. 2017; De Marco et al.. 2005; Moreira
et al.. 2001; Rodrigucs et al.. 1996).

3.5.5.3 Relevant Issues for Interpreting the Evidence Base

3.5.5.3.1 Potentially At-Risk Populations

Sex

A limited number of toxicological studies have reported sex differences in Pb-related effects on
motor function. Among studies in the 2013 Pb ISA, sex-specific differences in mice were observed for
gross motor skills, with balance and coordination most affected among males at the lowest Pb exposures
(Leasurc et al.. 2008).

Recent epidemiologic studies that evaluated sex as a potential modifier of the association between
Pb exposure and motor function add to the evidence (Liu et al.. 2022a; Y Ortiz et al.. 2017). Y Ortiz et al.
(2017) found that the observed association between maternal blood Pb during the third trimester and
lower PDI scores was not different between boys and girls. Liu et al. (2022a) found that the association of
maternal blood Pb exposure with gross motor development quotient on the GDS was modified by sex
(-3.43 [95% CI: -6.16, -0.69] in boys and -1.18 [95% CI: -2.81, 0.44] in girls per ln-transformed unit).

Maternal Self-esteem

Maternal self-esteem has been shown to modify associations between BLLs and health effects in
children. In one study, high maternal self-esteem appeared to attenuate the negative effects of the child's
increased BLLs on PDI scores (Surkan et al.. 2008). In this study, larger decreases in PDI scores were
associated with increased BLLs among children whose mothers were in the lower quartiles of self-esteem
(Surkan et al.. 2008). Maternal self-esteem was not evaluated as an effect modifier in recent studies of Pb
exposure and motor function among children.

External Review Draft

3-145

DRAFT: Do not cite or quote

-------
Maternal Stress

In a recent epidemiologic study, Y Ortiz et al. (2017) found that the observed association of
maternal blood Pb during the third trimester with lower PDI scores differed depending on maternal stress.
Contrary to expectations, higher PDI scores were observed with higher maternal stress.

3.5.5.3.2 Lifestages

Multiple lifestages during childhood are implicated in the effects of Pb exposure on motor
function in children. Analyses of children enrolled in the Cincinnati cohort at age 6 years indicated
associations of concurrent, lifetime average, and neonatal Pb exposure with poorer upper limb dexterity
and fine motor composite score. Studies conducted in the Cincinnati cohort found that prenatal or
neonatal BLLs were not consistently associated with motor function decrements at ages 4-10 years
(Bhattacharva et al.. 1995; Dietrich et al.. 1993). Several recent birth cohort studies support findings from
the 2013 Pb ISA with observations of lower scores on the Bayley PDI in association with maternal Pb
exposure (no clear pattern by trimester of pregnancy), cord BLL, and postnatal concurrent blood Pb
(Rvgiel et al.. 2021; Y Ortiz et al.. 2017; Liu et al.. 2014c; Kim et al.. 2013c; Henn et al.. 2012). Animal
toxicological studies mentioned above and in previous ISAs indicate the potential for delays in gross
motor development with gestational and/or early postnatal Pb exposure (Rao Barkur and Bairv. 2016;
Basha and Reddv. 2015; Leasure et al.. 2008) and for fine motor decrements with gestational Pb exposure
(Basha and Reddv. 2015). Apart from the study by Leasure et al. (2008). which tested balance in adults,
these studies measured and found diminished motor development in early postnatal rodents (Rao Barkur
and Bairv. 2016; Basha and Reddv. 2015).

3.5.5.4 Integrated Summary and Causality Determination: Motor Function in Children

The evidence assessed in the 2013 Pb ISA is sufficient to conclude that a "causal relationship is
likely to exist" between Pb exposure and decrements in motor function in children. Key evidence came
from prospective analyses of the CLS and Yugoslavia cohorts demonstrating associations of BLLs with
poorer motor function with consideration of potential confounders including SES, parental caregiving
quality and education, smoking birth outcomes, sex, and child health. Among children that participated in
the Cincinnati cohort, higher earlier childhood BLLs (age 0-5 year average [median: 11.7 |ig/dL| or age
78 month) were associated with poorer fine (i.e., grooved pegboard and finger tapping) (Ris et al.. 2004)
and gross motor function (i.e., postural balance) (Bhattacharva et al.. 2006) assessed in adolescence (ages
12, 15-17 years). In addition, assessments of children enrolled in the Cincinnati cohort at age 6 years
indicated associations of concurrent (mean: 10.1 (.ig/dL). lifetime average (mean: 12.3 (.ig/dL). and
neonatal (mean: 4.8 (ig/dL) but not prenatal maternal (mean: 8.4 (ig/dL) BLLs with poorer upper limb
dexterity, fine motor composite score (Dietrich et al.. 1993). and poorer postural balance (Bhattacharva et

External Review Draft

3-146

DRAFT: Do not cite or quote

-------
al.. 1995). Wasserman et al. (2000) also examined the association of Pb exposure with motor function. In
this prospective analysis of the Yugoslavian cohort, an association of lifetime average BLL (exact levels
not reported) with decrements in fine but not gross motor function at age 4.5 years was observed
(Wasserman et al.. 2000). Evidence from cross-sectional studies for associations between motor function
and concurrent BLL was mixed in populations with mean BLLs of 2-5 (ig/dL (Min et al.. 2007; Surkan et
al.. 2007; Despres et al.. 2005). Recent epidemiologic and toxicologic studies generally support findings
from the 2013 Pb ISA. The key evidence, as it relates to the causal framework, is summarized in Table 3-
6.

Several recent birth cohort studies report lower scores on the Bayley PDI in association with
maternal Pb exposure (no clear pattern by trimester of pregnancy), cord BLL, and postnatal concurrent
blood Pb (Rvgiel et al.. 2021; Y Ortiz et al.. 2017; Liu et al.. 2014c; Kim et al.. 2013c; Henn et al.. 2012).
Pb-associated decrements in motor function were also observed in neonates (Liu et al.. 2014d; Paraiuli et
al.. 2013). A limited number of studies of children greater than 7 years old were conducted. Taylor et al.
(2015) did not report associations with certain tasks indicative of gross motor function (i.e., balance),
although associations with decreased fine motor function were observed (Taylor et al.. 2018; Boucher et
al.. 2016).

Recent toxicological studies provide limited biological plausibility by showing effects on motor
function in rodent models from developmental Pb exposure resulting in BLLs <30 (ig/dL within one order
of magnitude of recent concentrations observed in humans. Epidemiologic evidence of developmental Pb-
induced impairment of balance and coordination is supported by observations of poorer rotarod
performance in male mice exposed to Pb during gestation (Leasure et al.. 2008). In addition, evidence
from epidemiologic studies indicating Pb-induced delayed gross motor development in children is
reinforced by toxicological studies that display slower times to achievement by postnatal rats
gestationally exposed to Pb in surface righting reflex (gestational Pb), negative geotaxis reflex
(gestational Pb) (Basha and Reddv. 2015). and ascending wire mesh tests (gestational Pb; lactational Pb)
(Rao Barkur and Bairv. 2016). Epidemiologic studies revealing Pb-induced decrements in children's fine
motor skills are supported by the observed grip strength deficits for gestational Pb-exposed early postnatal
rats in forelimb hang tests (Basha and Reddv. 2015). Additional studies on Pb-induced changes on several
neurochemical endpoints that factor into impaired motor function have been reported and are described in
Section 3.6. Overall, the evidence is sufficient to conclude that there is likely to be a causal
relationship between Pb exposure and motor function in children.

External Review Draft

3-147

DRAFT: Do not cite or quote

-------
Table 3-6 Summary of evidence indicating a likely causal relationship between Pb exposure and motor
function in children.

Rationale for Causality
Determination3

Key Evidence13

References'3

Pb Biomarker Levels Associated with
Effects0

Row heading if applicable

Consistent findings from a few
prospective epidemiologic studies
with relevant BLLs

Evidence from prospective studies for fine and gross Ris et al. (2004)	Blood Pb means

motor function decrements in children ages 4.5-17 yr Dietrich etal. (1993) Cincinnati: neonatal (10 day) 4.8 pg/dL,

rna,al' BhaeMmaetaU1Saa"n=urren' (age Syr) 11.6 Mg/dL, lifetime (to

earlier childhood, concurrent, lifetime avg BLLs.	/onnm age 15-17 yr) avg 12.3 pg/dL, age 0-5 yr

High follow-up participation, no selective attrition in Wasserman et al. (2000) gvg 11 y Mg/dL
Cincinnati cohort, higher loss-to-follow-up in	Former Yugoslavia' NR

Yugoslavia cohort with lower maternal IQ, HOME. Section 4.3.7, (U.S. EPA.

Both studies adjusted for maternal IQ, parental 2013a)
education, SES, HOME score

Studies used various, widely used tests to assess
outcomes.

Mixed evidence for lower (concurrent) BLLs from
cross-sectional studies that considered several
potential confounding factors.

Section 4.3.7, (U.S. EPA.
2013a)

Consistent findings from prospective
studies of infants and toddlers

Lower scores on the Bayley PDI in association with
maternal Pb exposure (no clear pattern by trimester
of pregnancy), cord BLL and postnatal concurrent
blood Pb

Kim et al. (2013c)
Y Ortiz et al. (2017)
Liu et al. (2014c)
Rvaiel et al. (2021)
Henn et al. (2012)

Limited evidence in neonates

Pb-associated decrements in motor function
reflexes) observed

e.g., Paraiuli et al. (2013)
Liu et al. (2014d)

Limited experimental animal evidence Deficient gross motor coordination and balance in
at relevant exposures	rodents with developmental Pb exposure (less time

on rotarod, slower righting and negative geotaxis
reflexes, delayed day of achievement for ascending
wire mesh test)

Leasure et al. (2008) Blood Pb:-10 pg/dL in mice after
Basha and Reddv (2015) progestational through lactation exposure,
„ „ , , „ . 5-11 ug/dL in rats after gestational
(20°16)	exposure, 27 pg/dL in rats after lactational

*	'	exposure

Fine motor (grip strength) deficits in early postnatal
rats with gestational Pb exposure

Basha and Reddv (2015) Blood Pb: 11.2 pg/dL after gestational
exposure

External Review Draft

3-148

DRAFT: Do not cite or quote

-------
Rationale for Causality
Determination3

Key Evidence13

References'3

Pb Biomarker Levels Associated with
Effects0

Limited experimental animal evidence Poorer balance (fell off rotarod more quickly) in adult Leasure et al. (2008)
at relevant exposures provide	male but not female mice with pregestational through

coherence for epidemiologic	lactation dietary Pb exposure

observations of effect modification by
sex

Biological plausibility demonstrated Pathways involving oxidative stress, inflammation andU.S. EPA (2013a)

Ca2+ signaling result in impaired neuron development, section 3 6
synaptic changes, and neurotransmitter changes.

Recent studies support and extend findings related to Section 3.3
overt nervous system effects

avg = average; BLL = blood lead level; Ca2+ = calcium ion; IQ = intelligence quotient; NR = not reported; Pb = lead; PDI = Psychomotor Developmental Index; SES = socioeconomic
status; yr = year(s).

"Based on aspects considered in judgments of causality and weight of evidence in causal framework in Table I and Table II of the Preamble to the ISAs (U.S. EPA. 2015).

'Describes the key evidence and references, supporting or contradicting, contributing most heavily to causality determination and, where applicable, to uncertainties or inconsistencies.
References to earlier sections indicate where the full body of evidence is described.

°Describes the Pb biomarker levels at which the evidence is substantiated.

Blood Pb: -10 [jg/dL in mice after
pregestational through lactation exposure

External Review Draft

3-149

DRAFT: Do not cite or quote

-------
3.5.6

Sensory Organ Function in Children

The 2013 Pb ISA included separate causality conclusions for auditory and visual function. This
ISA combines these categories and makes one causality determination for Sensory Organ Function (see
Section 3.5.6.4)

3.5.6.1 Auditory Function in Children

The evidence assessed in the 2013 Pb ISA was sufficient to conclude that "a causal relationship is
likely to exist" between Pb exposure and decrements in auditory function in children (U.S. EPA. 2013a).
Evidence from a prospective study (Dietrich et al.. 1992) and small number of cross-sectional studies of
U.S. children, including NHANES and Hispanic Health and Nutrition Examination Survey (HHANES)
analyses (Schwartz and Otto. 1991. 1987) indicated associations of higher BLLs with increases in hearing
thresholds as well as decreases in auditory processing or auditory evoked potentials, with adjustment for
potential confounding by SES in most studies and by child health and nutritional factors in some studies.
The high participation rates in a prospective birth cohort study (Dietrich et al.. 1992) reduced the
likelihood of biased participation by children with higher BLLs. Across studies, associations were found
with BLLs measured at various time periods, including prenatal maternal, neonatal (10 days, mean
4.8 (ig/dL), lifetime average (to age 5 years), and concurrent (ages 4-19 years) BLLs (median 8 (ig/dL).
Evidence for Pb-associated increases in hearing thresholds or latencies of auditory evoked potentials was
also found in adult monkeys with lifetime dietary Pb exposure. However, these effects in adult animals
were demonstrated at higher peak or concurrent BLLs (i.e., 33-150 (ig/dL) than those relevant to this
ISA; thus, the biological plausibility for epidemiologic observations was unclear.

In the current ISA, several recent cross-sectional studies support the conclusion in the 2013 Pb
ISA regarding the association of Pb exposure with hearing loss; however, results were inconsistent for
other audiometric parameters. Recent toxicological studies provide additional evidence for hearing loss
and auditory processing deficits in rodents at relevant BLLs. Measures of central tendency for Pb
biomarker levels used in each study, along with other study-specific details, including study population
characteristics and select effect estimates, are highlighted in Table 3-12E (Epidemiologic Studies) and
Table 3-16T (Toxicological Studies). An overview of the recent evidence is provided below.

3.5.6.1.1 Epidemiologic Studies of Auditory Function

Several recent epidemiologic studies examined the association between Pb exposure and
decrements in auditory function in children. The findings generally support a positive association between
Pb exposure and hearing loss. For other audiometric parameters, however, the results were inconsistent.

External Review Draft

3-150

DRAFT: Do not cite or quote

-------
Most studies of auditory function were cross-sectional. In a meta-analysis of studies from Iran, Korea,
China, and the United States, Yin et al. (2021) observed a positive association between Pb exposure and
hearing loss indicated by pure-tone average (PTA) >25 dB in children and adolescents (3-19 years)
(combined OR per unit increase in Pb = 1.53 [95% CI: 1.24, 1.87]). The pooled OR was based on only
two studies (Xu et al.. 2020; Choi and Park. 2017). Xu et al. (2020) conducted a case-control analysis of
preschool-aged children (3 to 7 years of age) who resided in an area contaminated with Pb and Cd and in
an uncontaminated reference area. This study found associations of exposures with hearing loss,
potentially affected by epigenetic changes. The OR for Pb-associated hearing loss in both ears was 1.40
(95% CI: 1.06, 1.84 per unit change log-transformed BLL) after adjustment for characteristics of the
child, parental education, SES, and noise exposure. Choi and Park (2017) measured speech- and high-
frequency hearing loss in adolescents (12-19 years) and adults (20-87 years) in the Korea National
Health and Nutrition Examination Surveys (KNHANES). Hearing loss was defined as PTA >15 dB in
adolescents. For each doubling of blood Pb, there was a positive association with speech-frequency
hearing loss (OR =1.2 [95% CI: 0.48, 3.05]) and high-frequency hearing loss (>25 dB) (OR = 1.26 [95%
CI: 0.73, 2.16]) among adolescents.

In addition to the aforementioned studies, among adolescent NHANES participants (ages 12 to
19 years), Shargorodskv et al. (2011) found a positive association between blood Pb and hearing loss.
Hearing loss was defined as low or high-frequency PTA >15 dB in either ear. Compared with study
participants with low BLLs (<1 (ig/dL), those with the highest level (>2 (ig/dL) were more likely to have
any hearing loss (OR = 1.95 [95% CI: 1.24-3.07]), particularly high-frequency hearing loss (OR = 2.22
[95% CI: 1.39-3.56]). The direction of effect for low-frequency hearing loss was the same but at a
smaller magnitude (OR =1.13 [95% CI: 0.61-2.07]). Among younger children (3-7 years with median
BLL <5 (ig/dL) in China, a positive association between blood Pb and hearing loss was also observed
(Liu et al.. 2018c). For each (ig/dL increase in blood Pb, the odds of any hearing loss increased by 1.24
times (OR = 1.24 [95% CI: 1.03, 1.49]). This association was not as evident for high-frequency hearing
loss (OR= 1.08 [95% CI: 0.84, 1.38]) and low-frequency hearing loss (OR= 1.02 [95% CI: 0.87, 1.19]).

Auditory function in children was also measured according to the auditory brainstem response
(ABR) (Silver et al.. 2016; Alvarenga et al.. 2015; Pawlas et al.. 2015). In an unadjusted descriptive
analysis of children (4-13 years) in Poland, BLLs were positively correlated with brainstem auditory
evoked potentials (BAEP) and pure-tone audiometry and negatively correlated with acoustic otoemission
(Pawlas et al.. 2015). In multivariable analyses stratified by polymorphisms in the ALAD and vitamin D
receptor (VDR) genes, the associations for BAEP per (ig/dL increase in blood Pb were generally null
(Pawlas et al.. 2015). Silver et al. (2016) measured ABR in newborns (average 2 days old) in China.
Compared with a low (<2 (ig/dL) BLL measured during late pregnancy, infants exposed to medium (2-
3.8 (ig/dL) and high (>3.8 (ig/dL) Pb levels were more likely to have a higher ABR central-to-peripheral
(C-P ratio) (Silver et al.. 2016). When using Pb levels measured in cord blood and during mid-pregnancy,
however, the association for ABR C-P ratio moved toward the null (Silver et al.. 2016). Although

External Review Draft

3-151

DRAFT: Do not cite or quote

-------
quantitative results were not provided for a study of children (18 months-14 years) in Brazil, Alvarenga
et al. (2015) observed no association between cumulative BLLs and BAEP.

Summary

A prospective study in the 2013 Pb ISA (Dietrich et al.. 1992) found an association of Pb
exposure with decreased auditory processing. In addition, cross-sectional studies found increased hearing
thresholds in children aged 4-19 years that participated in NHANES and HHANES in association with
higher concurrent BLLs. Recent cross-sectional and case-control studies of young children and
adolescents generally support a positive association between Pb exposure and hearing loss (Xu et al..
2020; Choi and Park. 2017; Shargorodskv et al.. 2011). whereas the results were inconsistent for ABR
(Silver etal.. 2016; Alvarenga et al.. 2015; Pawlas et al.. 2015).

3.5.6.1.2 Toxicological Studies of Auditory Function

Toxicological evidence for effects on auditory function in the 2013 Pb ISA was limited to one
study (U.S. EPA. 2013a'). This study evaluated auditory thresholds using a behavioral task in 13-year-old
monkeys (Macaca mulatto) who had previously been exposed to Pb either gestationally or postnatally
(Laughlin et al.. 2009). Potentially due to limitations noted within the study, small but nonsignificant
increases in the auditory threshold were reported in Pb-exposed animals compared with controls. Stronger
associations between Pb exposure, auditory threshold shifts, and latency in BAEP were reviewed in the
2006 Pb AQCD (U.S. EPA. 2006a). Importantly, the associations demonstrated in the 2013 Pb ISA and
2006 Pb AQCD occurred at higher BLLs (>30 (ig/dL) that would not be considered PECOS-relevant for
this ISA.

Changes in auditory thresholds using BAEP have been further assessed in three recent rodent
studies (Table 3-16T). Jamesdaniel et al. (2018) exposed male C57B1/6 mice from PND 33 to PND 61 to
Pb and subsequently detected 8-12-dB upward shifts in hearing thresholds (indicative of hearing loss)
between 4 and 32 kHz. In contrast, another recent study using similarly aged male CBA/CaJ mice and a
longer exposure paradigm (11 weeks) found no significant effect of Pb on hearing thresholds at 8, 16, and
32 kHz (Carlson et al.. 2018). The final study, which exposed male and female Sprague Dawley rats
postnatally to Pb did not detect significant differences in hearing thresholds between 4 and 28 kHz at
PND 60 (Zhu et al.. 2016). Animals in the two studies that did not detect an effect had lower BLLs than
those in the study that did (3-8 (ig/dL versus 29 (ig/dL). However, due to the small number of studies, the
existence of an exposure threshold for this effect remains uncertain.

Recent studies have also investigated the effect of Pb exposures on auditory processing, which
was not discussed in previous IS As. Zhu et al. (2016) exposed rat pups to Pb through their dams' drinking
water until weaning, when they began drinking Pb-free water. BLLs of the pups were roughly 8 (ig/dL
during exposure and had returned to baseline levels by PND 40. At PND 60, the Pb-exposed rats were

External Review Draft

3-152

DRAFT: Do not cite or quote

-------
found to have a decreased ability to discriminate between target and nontarget sound bursts. Additionally,
these rats were found to have a reduced spike rate-following ability and decreased cortical response
synchronization, indicative of a deficit in auditory cortical temporal processing. The same research group
published a follow-up study using a similar exposure paradigm to investigate another aspect of auditory
processing (i.e., sound localizationXLiu et al.. 2019). In a sound-azimuth discrimination task, Pb-exposed
animals took significantly longer to reach target accuracy and had significantly greater deviations (i.e.,
difference between the location of the desired response versus the location of incorrect response)
compared with control animals. These behavioral impairments were accompanied by a degraded sound-
azimuth selectivity in the primary auditory cortex neurons.

Summary

Earlier experimental animal studies have found decreased auditory function in adult monkeys and
rodents after lifetime exposure to Pb in animals with peak BLLs greater than 30 (ig/dL, but the persistence
of these effects at lower BLLs and in juvenile animals was uncertain. Three recent studies evaluated
auditory thresholds using BAEP in rodents exposed to Pb starting in the postnatal or juvenile period.
Jamesdaniel et al. (2018) found 8-12-dB upward shifts in hearing thresholds between 4 and 32 kHz in
young adult mice (peak BLLs of 29 (.ig/dL). Studies evaluating lower mean BLLs from 3 to 8 (ig/dL did
not report differences in BAEP thresholds. However, mice with mean peak BLLs of 8 (ig/dL had
significant deficits in auditory processing, including decreased sound discrimination and sound
localization ability paired with dysfunction in the auditory cortical neurons (Liu et al.. 2019; Zhu et al..
2016).

3.5.6.2 Visual Function

The evidence reviewed in the 2013 Pb ISA was inadequate to determine whether a causal
relationship exists between Pb exposure and visual function in children (U.S. EPA. 2013a). A study in
children and a few studies in animals showed Pb-associated increases in supernormal electroretinograms;
however, the biological plausibility of the observations was unclear. Overall, the available epidemiologic
and toxicological evidence was of insufficient quantity, quality, and consistency to support a causality
conclusion.

3.5.6.2.1 Epidemiologic Studies of Visual Function

Only a few epidemiologic studies examined the association between Pb exposure and decrements
in visual function in children (Silver et al.. 2016; Fillion et al.. 2013). Since the measures of visual
function differed between studies, it is difficult to draw any conclusions about Pb exposure and visual
function in children. Silver et al. (2016) measured grating visual acuity (VA) in 6-week-old infants in

External Review Draft

3-153

DRAFT: Do not cite or quote

-------
China. Compared with low (<2 (ig/dL) BLLs measured during late pregnancy, infants exposed to medium
(2-3.8 (ig/dL) and high (>3.8 (ig/dL) Pb levels were more likely to have lower grating VA (Silver et al..
2016). When using Pb levels measured in cord blood and during mid-pregnancy, however, the association
for grating VA was attenuated and moved closer toward the null (Silver etal.. 2016). In Brazil, Fillion et
al. (2013) measured contrast sensitivity (cycles per degree [cpd]) and acquired color vision loss (color
confusion index, CCI) in study volunteers that included adolescents (age range: 15-66 years). Based on
the entire study population, blood Pb exposure was negatively associated with the intermediate spatial
frequency of contrast sensitivity (12 cycles/degree); however, results varied by spatial frequency (Fillion
et al.. 2013). For CCI, there was a small positive association with blood Pb (Fillion et al.. 2013).

Summary

Overall, the available epidemiologic and toxicological evidence assessed in the 2013 Pb ISA was
of insufficient quantity, quality, and consistency to support a causality conclusion. A limited number of
recent epidemiologic studies are available for consideration; however, measures of visual function
differed between studies limiting observations regarding the consistency of the evidence overall.

3.5.6.2.2 Toxicological Studies of Visual Function

The evidence base pertaining to effects on visual function in the 2013 Pb ISA was largely
supported by seminal literature reviewed previously in the 1986 and 2006 Pb AQCDs showing reduced
VA, retinal alterations, and changes in CNS visual processing areas and subcortical neurons involved in
vision (U.S. EPA. 2013a. 2006a. 1986). Electroretinography (ERG), which measures the bioelectrical
response of the retina to a light stimulus, is used to detect abnormalities in retinal functioning. Fox et al.
(2008) found that Pb exposure in female Long-Evans rats (gestation through PND 10, measured at
PND 90) induced supernormal ERGs (i.e., increases in the response amplitude) at low and moderate
exposure levels (BLLs of 12 and 24 (ig/dL) and subnormal ERGs (i.e., decreases in the response
amplitude) in the high exposure group (BLL of 46 (ig/dL). Earlier studies have also found Pb-related
aberrations in ERGs, but the direction of this effect is inconsistent (i.e., both subnormal and supernormal
responses have been detected) (Fox et al.. 1997; Lilienthal et al.. 1988). As discussed in Giddabasappa et
al. (2011). the effect direction may be related to both the lifestage during exposure (gestational versus
postnatal) and the Pb dose. This study also demonstrated that low to moderate gestational Pb exposure
(BLLs: 10 and 27 (ig/dL) increased and prolonged retinal progenitor cell proliferation, resulting in
selectively increased rod photoreceptor and bipolar cell neurogenesis in C57BL/6 mice at PND 60
(Giddabasappa et al.. 2011). Adult monkeys (Macaca fascicularis) with lifetime Pb exposure, producing
BLLs from 50-115 (ig/dL, had temporal vision dysfunction but no change in spatial function (Rice.
1998). In contrast to these effects, Laughlin et al. (2008) found that Pb exposure in Rhesus monkeys
(exposed from PND 8-26 weeks; BLLs of 35-40 (ig/dL) did not significantly affect the development of

External Review Draft

3-154

DRAFT: Do not cite or quote

-------
photopic spatial acuity assessed using a modified Teller preferential looking paradigm. Recent PECOS-
relevant studies have not further examined the effects of Pb on visual function.

3.5.6.3	Relevant Issues for Interpreting the Evidence Base

3.5.6.3.1 Potentially At-Risk Populations

Genes

Pawlas et al. (2015) conducted multivariable analyses stratified by polymorphisms in the ALAD
and VDR genes and found that the associations for BAEP and pure-tone audiometry per |ig/L increase in
blood Pb were generally null (Pawlas et al.. 2015).

3.5.6.4	Integrated Summary and Causality Determination: Sensory Organ Function

The 2013 Pb ISA presented two causality determinations related to sensory function in children:
auditory function and visual function (U.S. EPA. 2013a). The evidence was sufficient to conclude that a
causal relationship was likely to exist between Pb exposure and auditory function decrements in children,
based on consistent findings from epidemiologic studies and studies in nonhuman primates. For visual
function, the evidence was inadequate to determine if a causal relationship exists. In this ISA, recent
studies inform a single causality determination for sensory organ function because there are relatively few
studies within this outcome grouping.

Auditory processing decrements were previously demonstrated in a prospective study by Dietrich
et al. (1992). In 5-year-old children, elevated BLLs during infancy (mean BLLs of 4.8 (ig/dL at 10 days
old) were associated with poorer performance on a test for auditory processing disorders after adjusting
for confounding factors including SES, HOME score, a variety of birth outcomes, maternal alcohol
consumption, maternal smoking, and overall child health. Recently, experimental animal studies
demonstrated that postnatal Pb exposure resulting in mean peak BLLs of 8 (ig/dL also caused significant
deficits in auditory processing, including decreased sound discrimination and sound localization ability
paired with dysfunction in the auditory cortical neurons (Liu et al.. 2019; Zhu et al.. 2016).

Multiple large cross-sectional NHANES and HHANES studies have shown that higher BLLs
(children aged 4-19; BLLs 8 (ig/dL) are associated with increased hearing thresholds (Schwartz and Otto.
1991. 1987). These associations remained after adjustment for age, sex, race, family income, parental
education, and nutritional factors. Recent cross-sectional and case-control studies continued to
demonstrate associations with BLLs and hearing loss in young children (aged 3-7, BLLs ~3 to 6 (ig/dL)
and adolescents (aged 12-19, BLLs ~1 to 8 (.ig/dL). particularly at higher frequencies (Xu et al.. 2020; Liu

External Review Draft

3-155

DRAFT: Do not cite or quote

-------
et al.. 2018c; Choi and Park. 2017; Shargorodskv et al.. 201IV Furthermore, hearing threshold increases
were previously demonstrated in adult nonhuman primates after developmental or lifetime Pb exposure,
although BLLs in these studies were greater than 30 (ig/dL (Laughlin et al.. 2009; Rice. 1997). Recent
experimental animal studies have not further evaluated hearing thresholds in nonhuman primates and
instead have focused on BAEPs in rodents. Jamesdaniel et al. (2018) found 8-12-dB upward shifts in
auditory thresholds between 4 and 32 kHz in young adult mice exposed during adolescence (peak BLLs
29 (ig/dL). Similar studies did not detect differences in BAEPs in rodents with lower peak BLLs (3 to
8 (ig/dL). Likewise, a few recent epidemiologic studies also evaluated BAEP with inconsistent results.

Although Pb-induced alterations in subcortical visual neurons, visual processing areas, and retinal
development have been demonstrated, supporting the biological plausibility of Pb-associated effects on
vision (U.S. EPA. 2013a). evidence relating to visual function in epidemiological and toxicological
studies remains limited and inconsistent. Silver etal. (2016) found that decreased VA in infants was
associated with maternal BLLs higher than 2 (ig/dL in late pregnancy, but this association was weaker
with BLLs in both mid-pregnancy and cord blood. Studies in nonhuman primates failed to detect changes
in VA at BLLs above 35 (ig/dL, although one reported decrements in temporal acuity as a result of Pb
exposure (Laughlin et al.. 2008; Rice. 1998). Another recent study found associations with blood Pb and
decrements in contrast sensitivity and color vision, an endpoint that has not been previously studied, in a
study population which included adolescents (15-66 years old)(Fillion et al.. 2013). Studies in both
humans and animals have found significant but inconsistent changes in ERGs (Fox et al.. 2008;
Rothenberg et al.. 2002; Fox etal.. 1997). though it is unclear if these findings translate to functional
visual changes.

In conclusion, recent studies of auditory function build upon the evidence base established in the
2013 Pb ISA, particularly with respect to auditory processing and hearing loss at low BLLs.
Epidemiologic studies evaluating BAEP are not entirely consistent and toxicological studies conducted in
young animals remain limited. Based on previous and recent evidence relating to auditory processing and
function, the evidence is suggestive of, but not sufficient to infer, a causal relationship between Pb
exposure and sensory function in children.

External Review Draft

3-156

DRAFT: Do not cite or quote

-------
Table 3-7 Evidence that is suggestive of, but not sufficient to infer, a causal relationship between Pb
exposure and sensory organ function in children.

Rationale for Causality
Determination3

Key Evidence13

References'3

Pb Biomarker Levels Associated with
Effects0

Auditory Function

Consistent findings from a few
epidemiologic studies with
relevant BLLs

Prospective study found associations of
prenatal (maternal), neonatal, yearly age
1 to 5 yr, lifetime avg BLLs with poorer
auditory processing in children at age
5 yr in Cincinnati.

Dietrich etal. (1992)

Blood Pb means: neonatal (10 day)
4.8 [jg/dL, yearly age 1 to 5 yr 10.6-
17.2 [jg/dL, lifetime (to age 5 yr) avg NR

Cross-sectional and case-control studies
for increased hearing thresholds in
children ages 3-19 yr, including analyses
of NHANES, HHANES and KNHANES in
association with higher concurrent BLLs.

Section 4.3.6.1, (U.S. EPA. 2013a)

Xu et al. (2020)

Liu etal. (2018c)

Sharqorodskv et al. (2011)

Choi and Park (2017)

Blood Pb median:

HHANES: 8 pg/dL; NHANES: NR

Means 3.63-5.69 pg/dL (3-7 yr)

NHANES (2005-2008): med ~1 pg/dL
(12-19 yr)

KNHANES: GM: 1.26 pg/dL (15.6 yr)

Epidemiologic evidence helps Prospective study adjusted for SES,

to rule out chance, bias and
confounding with reasonable
confidence

HOME score, birth outcomes, obstetrical
complications, maternal smoking. Several
other factors considered.

Cross-sectional and case-control studies
considered potential confounding by age,
sex, race, income, parental education,

nutritional factors.

Dietrich etal. (1992)

Xu et al. (2020)

Liu etal. (2018c)
Sharqorodskv et al. (2011)
Choi and Park (2017)

External Review Draft

3-157

DRAFT: Do not cite or quote

-------
Rationale for Causality
Determination3

Key Evidence13

References'3

Pb Biomarker Levels Associated with
Effects0

Uncertainty due to lack of
animal evidence in juveniles
and limited evidence at
relevant exposure levels

Hearing loss in adult monkeys and
decreased BAEP in young adult rodents
at higher exposure levels.

Rice (1997)

Lauqhlin et al. (2009)
Jamesdaniel et al. (2018)

Peak BLLs >29 |jg/dL



Decrements in sound discrimination and
localization in young adult rodents.

Zhu et al. (2016)
Liu et al. (2019)

Peak BLLs 8.2 pg/dL

Visual Function

Limited evidence from
epidemiologic studies

Associations with some tests of grating
VA and contrast sensitivity observed.

Silver etal. (2016)
Fillion etal. (2013)



Uncertainty due to limited
animal evidence in juveniles
and at relevant exposures

Higher than relevant postnatal Pb
exposure did not cause changes in VA in
infant nonhuman primates in infants but
did decrease temporal acuity in adults.

Lauahlin et al. (2008)
Rice (1998)

BLLs >35 pg/dL

Biological plausibility
demonstrated

Pb-induced alterations in ERGs,
subcortical visual neurons, visual
processing areas, and retinal
development demonstrated.

(U.S. EPA. 2013a)



avg = average; BAEP = brainstem auditory evoked potentials; BLL = blood lead level; ERG = electroretinography; GM = geometric mean; HHANES = Hispanic Health and Nutrition
Examination Survey; HOME = Health Outcomes and Measures of the Environment; KNHANES = Korea National Health and Nutrition Examination Survey; NHANES = National
Health and Nutrition Examination Survey; NR = not reported; Pb = lead; SES = socioeconomic status; VA = visual acuity; yr = year(s).

aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Table I and Table II of the Preamble to the ISAs (U.S. EPA. 2015). Note that

the change from "likely to be causal" for auditory effects in children in the 2013 Lead ISA, to "suggestive of, but not sufficient to infer, a causal relationship" for sensory organ function

in children reflects minor changes to the causal framework, rather than a weakening of the evidence base pertaining to auditory effects in children.

bDescribes the key evidence and references, supporting or contradicting, contributing most heavily to causality determination and, where applicable, to uncertainties or

inconsistencies. References to earlier sections indicate where the full body of evidence is described.

°Describes the Pb biomarker levels at which the evidence is substantiated.

External Review Draft

3-158

DRAFT: Do not cite or quote

-------
3.5.7

Social Cognition and Behavior in Children

In addition to neurodevelopmental disorders covered in previous sections—including ADHD,
intellectual and developmental disabilities, and motor disorders—there is an emerging body of research
on autism spectrum disorder (ASD) and other conditions related to social cognition and behavior. The
2013 Pb ISA (U.S. EPA. 2013a) did not evaluate any epidemiologic studies examining associations
between Pb exposure and autism. ASD is generally characterized by restricted interests and behaviors,
including stereotyped patterns of behavior and sensory sensitivities. To meet the DSM criteria for ASD, a
child must have persistent deficits in social communication and demonstrate repetitive behaviors (APA.
2013). Social cognition, which is often impaired among individuals with ASD, involves the ability to
interpret and respond to social cues, communication, and interaction. These traits (or behaviors) can be
measured on a continuum in the general population with scores exhibiting a fairly normal distribution,
with scores at the extreme impaired end indicating a higher risk for ASD (Constantino. 2011). Deficits in
social cognition have been associated with lifelong educational, vocational, adaptive functioning, and
mental health challenges among individuals with and without a clinically diagnosed disorder. Autism
diagnosis (e.g., via the ICD code), the CBCL, Social Responsiveness Scale (SRS), BASC-2, BSID-II and
III, CDIIT, ASQ:I, GDS, Social Maturity Scale (SMS), MDAT, and ECDI have been used in studies
examining the association of Pb exposure with social cognition and behavior.

3.5.7.1 Epidemiologic Studies of Social Cognition and Behavior

There have been a number of recent studies of ASD and deficits in social cognition and related
behaviors. Many of these recent studies did not control for potential confounders and/or did not include
robust statistical methods to estimate C-R relationships between Pb exposure and outcome, and are not
considered further in this section (Filon et al.. 2020; Oin et al.. 2018; Skalnv et al.. 2017; Macedoni-
Luksic et al.. 2015; Alabdali et al.. 2014; Yassa. 2014; De Palma et al.. 2012; Blaurock-Busch et al..
2011; Tianetal.. 2011). Instead, the ensuing discussion focuses on a number of autism and social
cognition studies that include more comprehensive control for potential confounders. The relevant studies
provide some evidence of a positive association between Pb exposure and ASD, along with generally
consistent supporting evidence of an association with decrements in social cognition. Measures of central
tendency for BLLs used in each study, along with other study-specific details, including study population
characteristics and select effect estimates, are highlighted in Table 3-13E of Section 3.7. An overview of
the recent evidence is provided below.

Two recent studies used robust modeling approaches to assess the C-R relationship between
exposure to Pb and ASD (Arora et al.. 2017; Kim et al.. 2016). While each study examined different
biomarkers of exposure and exposure windows, both indicated associations between Pb exposure and
ASD. Arora et al. (2017) conducted a difference-in-differences analysis of a small case-control study of

External Review Draft

3-159

DRAFT: Do not cite or quote

-------
8- to 12-year-old twins with discordant or concordant ASD status. Pb was measured in shed deciduous
teeth using a method that provided temporal estimates of tooth Pb levels ranging from 20 weeks before
birth to 30 weeks after birth. To estimate the relationship between tooth Pb and ASD across this exposure
window, the authors used distributed lag models to estimate the smoothed mean differences in tooth Pb
levels in discordant pairs minus the mean differences in concordant twins at each time point. In this case,
concordant twins served as the control group to account for natural variations in Pb exposure within a
dyad. One analysis used concordant twins without ASD and the other used concordant twins with ASD as
the control groups. In both cases, the difference in tooth Pb levels between discordant twins was greater
than the difference in concordant twins across the entire exposure window, though there appeared to be
bimodal peaks in tooth Pb differences from about 10 to 15 weeks before birth and 10 to 20 weeks after
birth (see Figure 3-14).

Time since birth (weeks)	Time since birth (weeks)

ASD = autism spectrum disorder.

Black line represents the difference in mean differences in tooth Pb levels between discordant ASD twins and: A) control twins; or B)
concordant ASD twins. Gray bands are unadjusted 95% CIs, while blue bands are adjusted for intra-twin correlations. Values above
zero represent increased levels in ASD cases compared with the non-ASD sibling after taking into account average difference in
control twins.

Source: Arora et al. (2017).

Figure 3-14 Differences in mean difference tooth Pb levels for autism

spectrum disorder in discordant twin pairs versus (A) non-autism
spectrum disorder twin pairs or (B) autism spectrum disorder
concordant twin pairs.

In a large cohort study of children in South Korea, Kim et al. (2016) analyzed blood Pb in relation
to autistic behaviors measured by parental response to the Autism Spectrum Screening Questionnaire
(ASSQ) and the SRS at ages 11-12 years old. BLLs at study enrollment (7-8 years old) were associated
with higher scores on the ASSQ (number of autistic behaviors) and SRS (severity across domains of

External Review Draft

3-160

DRAFT: Do not cite or quote

-------
social awareness, cognition, communication, motivation, and mannerisms). There were null associations
with blood Pb measured at 9-10 years and attenuated, but still positive, associations with concurrent
BLLs (11-12 years old). Nonparametric generalized additive models indicated an approximately linear
relationship between BLLs at enrollment and scores on the SRS. In addition to continuous models, the
authors also dichotomized ASSQ scores and reported a 45% increase (95% CI: 10%, 93%) in the odds of
a positive screen for autism (ASSQ score >17) per 1 (ig/dL increase in BLLs at enrollment. Notably,
symptoms of ASD manifest as early as infancy and although BLLs in this study were measured prior to
assessment of autistic behaviors, the relevant exposure window likely preceded exposure measurement.

In contrast to the results from Kim et al. (2016) and Arora et al. (2017). recent case-control
studies did not observe increased adjusted mean childhood BLLs in ASD cases compared with controls
(Rahbar et al.. 2021; Rahbar et al.. 2015). In addition to matching cases and controls on age and sex, the
authors estimated mean differences using a linear model controlling for a variety of demographic and SES
factors, including maternal age. Similarly, a recent case-control study reported a null association between
tertiles of maternal BLLs and ASD in children, though there was some evidence of a nonlinear
association in a cubic spline model (Skogheim et al.. 2021). The study populations for these analyses
included children and mothers with lower (<2 (ig/dL; (Skogheim et al.. 2021; Rahbar et al.. 2015)) and
higher (>7 (ig/dL; (Rahbar et al.. 2021)) mean or median BLLs.

One additional large retrospective study in Northeast China (Dong et al.. 2022) compared current
BLLs in children with moderate/severe versus mild autism, as determined by CARS scores. Mean BLLs
for the mild and moderate/severe groups were 2.58 (SD: 1.08) (ig/dL and 3.25 (SD: 1.89) (ig/dL,
respectively. Adjusting for age, place of residence, type of caregivers, parental education level, and
gastrointestinal problem, autism severity was positively associated with BLL ((3 = 0.03 [95% CI: 0.01,
0.05]).

Providing supporting evidence to the autism studies described above, several cohort studies
investigated the associations between Pb exposure and social cognition in children without autism. Most
of these studies reported inverse relationships of increasing prenatal Pb exposure with corresponding
small decrements in social cognition measures. Three studies additionally investigated effect modification
by various other factors.

Rvgiel et al. (2021) assessed the relationship between maternal blood Pb and infant behavioral
development at 12 to 24 months of age in a small analysis of three birth cohorts from the ELEMENT
study in Mexico City. The authors used the behavioral rating scale (BRS) of the BSID-II to examine
attention, social engagement, orientation, motivation, and emotional response, giving rise to two social
cognition outcomes: orientation/engagement (ORIEN) and emotional regulation (EMOCI). The authors
reported small but consistent decreases in 24-month EMOCI and ORIEN percentiles with increasing BLL
at each trimester. Associations were greatest for the second trimester, with 1.13% (95% CI: -2.63%,
0.37%) and 0.98% (95% CI: -2.83%, 0.88%) decreases in 24-month EMOCI and ORIEN percentiles,
respectively, for 1 (ig/dL increase in blood Pb. In an examination of the mediation of trimester-specific Pb

External Review Draft

3-161

DRAFT: Do not cite or quote

-------
exposure by DNA methylation at several previously identified CpG sites, the authors observed both
enhancing and suppressive effects of DNA methylation on the association between blood Pb and
neurocognitive outcomes, depending on the gene locus, with the majority playing a suppressive role.
Shekhawat et al. (2021) similarly reported null but slightly inverse associations of cord blood Pb and
social-emotional scores from the BSID-III in a study of mother-child pairs in western Rajasthan, India.
Neurocognitive assessments were conducted at the average age of 6.5 months. Additionally, Nozadi et al.
(2021) reported nonsignificant decreases in communication ((3 = -0.15 [95% CI: -0.58, 0.28]) and
personal-social ((3 = -0.11 [95% CI: -0.72, 0.50]) domain scores on the ASQ:I.

Some cohort studies examined interactions between Pb levels and the levels of other trace
elements (Nvanza et al.. 2021; Dohertv et al.. 2020; Lin et al.. 2013). Lin et al. (2013) measured maternal
blood Pb and assessed child development (including social and self-care skills) in the TBPS with the
CDIIT, as described in Section 3.5.1.2. The authors observed that highly Pb-exposed (>75th percentile:
1.65 (ig/dL) children had lower social DQs ([3 = -5.89 [95% CI: -10.81, -0.97]) compared with those
with low prenatal Pb exposure. In addition, the authors reported nonsignificant decreases in social or self-
help DQs among those with higher Pb and Mn concentrations in an interaction analysis. Nvanza etal.
(2021) measured Pb, Hg, Cd, and As concentrations using dried blood spots from pregnant mothers at 16-
27 weeks of gestation in Northern Tanzania. Adjusting for maternal age, maternal education, maternal
and parental occupation, number of under-five siblings at home, family socioeconomic wealth quintile,
infant sex, infant age, birth weight, and height and weight at neurocognitive testing, the authors did not
observe an association between high Pb exposure (>3.5 (ig/dL) and social impairment on the MDAT,
which is described in Section 3.5.1.2. However, the interaction analysis with maternal blood Hg levels
showed that children highly exposed to both Hg (>0.08 (ig/dL) and Pb were more likely to have global
neurodevelopmental impairment (PR= 1.40 [95% CI: 0.90, 2.10]). Dohertv et al. (2020) measured
concentrations of Pb and other metals (As, Cu, Mn, Se, and Zn) in maternal prenatal and postnatal
toenails and infant toenails at 6 weeks of life from mother-infant pairs in the New Hampshire Birth
Cohort. The three exposure assessments estimated exposures that occurred during periconception and
early pregnancy, mid-pregnancy, and late pregnancy and early neonatal life, respectively. The authors did
not observe significant associations between Pb exposure at the three timepoints and total SRS-2 scores
or the adaptive skills composite on the BASC-2 (see Section 3.7, Table 3-13E). Pb concentrations did not
appear to interact with other metals on the total SRS-2 score or the BASC-2 adaptive skills composite,
and sex-stratified analyses revealed inconsistent associations among girls.

One additional study assessed effect modification by maternal psychosocial measures. Zhou et al.
(2017) investigated the interactions of maternal BLL in whole blood and maternal prenatal stress levels
with child development (including adaptive behavior and social domains) using the GDS. Among those
with high maternal stress levels (GSI: P75-P100), adaptive behavior DQs decreased by 17.93 points
(95% CI: -35.83, -0.03) per loglO-transformed (ig/dL increase in maternal BLL. Social behavior DQs
also saw substantial decrements ([3 = -41.00 [95% CI: -63.11, -18.89] per log-10 transformed unit of
BLL).

External Review Draft

3-162

DRAFT: Do not cite or quote

-------
One cross-sectional study (Rucbner et al.. 2019) evaluated the association between concurrent
BLL and neurocognitive outcomes including adaptive skills among children with CKD, using parent
ratings on the BASC-2. More details about the study are covered in Section 3.5.2.1.1. Higher BLL was
associated with worse adaptive skills composite scores ((3 = -3.1) in univariable analyses; however, this
association did not remain after adjusting for key sociodemographic and clinical confounders.

One additional prospective study (Vigeh et al.. 2014) measured domains of social and self-help
skills but presented only associations with composite neurodevelopment test scores (described in
Section 3.5.1.2), which impedes parsing of specific social cognition effects of Pb exposure. In addition,
Kim et al. (2018b) evaluated concentrations of Pb in maternal serum, cord blood, urine, and breast milk in
association with neurodevelopmental and behavioral outcomes, including social quotient (SQ) measures
from the SMS among 13-24-month-old children. However, they reported only statistically significant
results in the paper, precluding quantitative results for blood Pb and SQ.

Recent epidemiologic studies utilized a wide range of outcome measures, including diagnostic
tests of autism (e.g., ICD code, DSM classification, ASSQ, and the Autism Diagnostic Observation
Schedule [ADOS]), behavior rating systems (e.g., CBCL, SRS-2, BASC-2, and SMS), and
neurodevelopmental assessments with social behavioral subtests (e.g., BSID-II and III, CDIIT, ASQ:I,
GDS, MDAT, and ECDI). Psychometric tests of social cognition often add valuable dimensional
information regarding the severity and type of social deficit among children with autistic traits, and the
wide variety of tests used in the evaluated studies provided insight into diverse aspects of problems with
social cognition, including communication, adaptive and self-help skills, social engagement, and
emotional behavior. One limitation, however, is that this variety complicates a straightforward
interpretation of results due to the lack of consistency of measures. Many behavioral tests provide
outcomes that overlap with domains discussed in other sections such as externalizing behavior
(Section 3.5.3) and internalizing behavior (Section 3.5.4), which can limit parsing of effects. Vigeh et al.
(2014) examined social and adaptive skills but reported quantitative results using only the global
neurodevelopmental composite score from the ECDI. Other studies (Nvanzaetal.. 2021; Rygiel et al..
2021) used rating subscale measures (i.e., EMOCI and ORIEN from the BRS; social development score
from the MDAT) that are not widely used in the literature, making it difficult to compare results across
studies.

3.5.7.1.1 Summary

Two recent high-quality studies of Pb exposure and ASD reported positive associations between
increased Pb exposure and higher risk of ASD diagnosis or symptomatology (Arora et al.. 2017; Kim et
al.. 2016). One retrospective study also observed a positive association between greater autism severity
and current BLL (Dong et al.. 2022). However, some case-control studies did not find evidence of a
positive association (Rahbar et al.. 2021; Skogheim et al.. 2021; Rahbar et al.. 2015). Although most
autism studies except one (Kim et al.. 2016) were case-control studies, two of the case-control studies

External Review Draft

3-163

DRAFT: Do not cite or quote

-------
accounted for temporality of exposure and outcome by analyzing prenatal maternal blood (Skoghcim et
al.. 2021) or using tooth Pb measurement methods that allow ascertainment of perinatal Pb exposure
levels (Arora et al.. 2017). Several cohort studies observed null or slight impairments of social
dimensions scores. Recent studies of social cognition in children without ASD used a wide variety of
psychosocial and neurodevelopmental instruments, such as BASC-2, BSID-II and III, CDIIT, ASQ:I,
GDS, SRS, SMS, MDAT, and ECDI, to obtain scores of social, emotional, and adaptive abilities. These
studies were mostly prospective in design and accounted for some key potential confounders, including
maternal age, parental education, SES, and caregiving.

3.5.7.2 Toxicological Studies of Social Cognition and Behavior

The previous ISA incorporated evidence of the effects of Pb exposure on social cognition and
behavior. Donald et al. (1986) reported sex-specific effects of Pb exposure on social investigatory
behavior in mice, wherein males and females exposed to Pb displayed enhanced social interaction but at
different times after exposure. In a subsequent publication, Donald et al. (1987) reported that Pb exposure
increased non-social behavior in males while females displayed decreased non-social behavior. The
previous evidence suggests that Pb may influence social behavior in rodents in a sex-specific manner, but
the direction of the effect was not clear.

There is limited recent toxicological evidence available on the effects of Pb exposure on social
cognition and behavior. A single study by Tartaglione et al. (2020) examined homing test and ultrasonic
vocalizations (USV). USV are calls emitted by pups when separated from their mother and siblings and
are markers of early emotional and communication development. Pups prenatally and lactationally

exposed to Pb exhibited reduced numbers of calls at PND 4 and 12, with no significant differences at
PND 7 and 10 from control animals. The same study also performed a homing test, which assesses
discriminative performance and maternal preference behavior by separating the pup from the dam and
recording the time taken to return to the nest from a maze. The time spent is a measure of both olfactory
discrimination and social preference. The authors reported no difference in homing test performance
between control and Pb-exposed pups at PND 12 (Tartaglione et al.. 2020). In summary, there is limited
evidence from the toxicological literature examining potential relationships between developmental Pb
exposure and social behavior, which represents an area of uncertainty.

3.5.7.3 Relevant Issues for Interpreting the Evidence Base

3.5.7.3.1 Concentration-Response Function

Evaluation of the shape of the C-R function in recent studies of social cognition is limited,
making it challenging to draw conclusions. Across studies, associations between Pb exposure and social

External Review Draft

3-164

DRAFT: Do not cite or quote

-------
cognition and behavior were observed at median or geometric mean maternal and cord BLLs ranging
from 3.3 to 5.5 (ig/dL, and BLLs measured in children ranging from 1.6 to 3.9 (ig/dL (Table 3-8). Kim et
al. (2016) used penalized regression splines to examine the C-R relationship between BLLs at 7-8 years
old and SRS scores at 11-12 years old. The C-R relationship was approximately linear across the range of
the BLL distribution, though there is more confidence in the shape of the C-R relationship (i.e., more
narrow confidence limits) closer to the mean, where there is a higher density of observations. Spline
models for most of the SRS subscales are also approximately linear, except for social cognition, which
has a sublinear relationship with BLLs (i.e., a smaller slope below the mean).

3.5.7.3.2 Potentially At-Risk Populations

Maternal Stress

Stratifying by maternal stress, Zhou et al. (2017) found that social behavior ((3 = -41.00, 95% CI:
-63.11, -18.89 per log-10 transformed unit of BLL) and adaptive behavior ((3 = -17.93, 95% CI: -35.83,
-0.03 per log-10 transformed unit of BLL) in toddlers were adversely associated with higher Pb levels
among children of mothers with low prenatal stress.

Co-exposure to Other Metals or Chemicals

Lin et al. (2013) observed slight impairments to social and self-help DQs among those with high
concentrations of both Pb (>1.65 (ig/dL) and Mn (>5.93 (ig/dL).

Gene-Environment Interactions

Rvgiel et al. (2021) found both enhancing and suppressing effects of DNA methylation at several
CpG sites in mediation analyses. Methylation of cg23280166 within CCSER1, a gene which has been
associated with ADHD, suppressed the association between second trimester Pb levels and ORIEN and
EMOCI scores at 24 months old, while methylation at cgl8515027 (GCNT1), positively mediated the
association between first and second trimester BLLs and 24-month EMOCI scores. Likewise, DNA
methylation of cg23280166 (VPS11) also positively mediated the relationship between third trimester
BLLs and 24-month EMOCI scores.

Pre-existing Conditions

One recent study evaluated adaptive behavior among children with CKD. After adjusting for
sociodemographic and CKD-related variables, they did not observe a significant association (Rucbncr et
al.. 2019).

Sex

Dohertv et al. (2020) observed inconsistent associations between Pb and SRS-2 total and
BASC-2 adaptive skills composite scores in sex-stratified analyses. Female infant toenail Pb

External Review Draft

3-165

DRAFT: Do not cite or quote

-------
concentration was positively associated with adaptive skills ((3 = 0.26 [95% CI: 0.07, 0.45] per log-2
transformed unit of BLL) while maternal prenatal toenail Pb was negatively associated ((3 = -0.19 [95%
CI: -0.34, -0.04] per log-2 transformed unit of BLL). There were additionally inconsistent interactions
with other metals (i.e., As and Se) among female but not male infants. Sample size limited statistical
precision in sex-stratified analyses, which can help explain the inconsistencies.

3.5.7.3.3 Confounding

Several sociodemographic characteristics were considered for confounder control in recent
epidemiologic studies. Child age at outcome measurement was included in all but three studies
(Shckhawat et al.. 2021; Kim et al.. 2018b: Vigeh et al.. 2014). Child sex was included in all studies but
two (Nozadi et al.. 2021: Vigeh et al.. 2014). Parental education appears to be associated with BLLs and
measures of social cognition and autism status. All studies except Rygiel et al. (2021) adjusted for or
considered parental education. Rygiel et al. (2021) was the only study to include maternal IQ as a
potential confounder. Many studies also included SES among their covariates (Nvanza et al.. 2021:

Rygiel etal.. 2021: Ruebner et al.. 2019: Zhou et al.. 2017: Vigeh et al.. 2014). HOME score was
included in only one study (Lin et al.. 2013).

Various birth factors are relevant for discussion. Maternal age was consistently included as a
potential confounder in most studies. Maternal age is known to contribute to autism risk and is correlated
with Pb exposure; hence, lack of inclusion in models may introduce bias. Additionally, autism, like many
developmental disorders, is more prevalent as delivery diverges in both directions from 40 weeks of
gestation. As such, gestational age and birth weight were included in most studies. Breastfeeding, parity,
maternal smoking and alcohol intake, and food consumption during pregnancy were also included in
multiple studies.

Genetics also plays a large role in the association between Pb exposure and social cognition
abilities. Arora et al. (2017) used a case-control design with twin pairs, which allowed for matching on
genetic factors to some extent. Rahbar et al. (2021) evaluated interaction effects of glutathione S-
transferase (GST) genes (GSTP1, GSTM1, and GSTT1), which have been linked to detoxification of
environmental pollutants and to autism status. However, they did not find evidence of a significant
interaction. Rygiel et al. (2021) examined mediation by DNA methylation at various CpG sites linked to
prenatal Pb levels.

Co-exposures and mixtures with other trace metals were considered in several studies. Nozadi et
al. (2021) found positive correlations of BLLs with Mn and Cd. The authors used an algorithm to identify
the control variables for each metal they analyzed, including all co-occurring metals and demographics;
however, none met the inclusion criteria of being significantly associated with both the exposure and
outcome, and the final model did not include any covariates. Additionally, Lin et al. (2013) observed that
Mn and As were positively correlated with Pb. Nvanza et al. (2021) further found positive correlations

External Review Draft

3-166

DRAFT: Do not cite or quote

-------
between Cd and Pb levels. Kim et al. (2016) adjusted for Hg and was the only study to adjust for a metal
in its final model. Finally, two studies examined potential modifying effects of Mn (Lin et al.. 2013) and
Hg (Nvanza et al.. 2021).

3.5.7.3.4 Lifestages

No epidemiologic studies examining the relationship between Pb exposure and social cognition
and behavior in children were included in the 2013 Pb ISA (U.S. EPA. 2013a). Recent studies
demonstrated that BLLs measured during various lifestages and time periods (i.e., prenatal, early
childhood, later childhood, and concurrent with outcome assessment) are associated with ASD and
decrements in social cognition. Due to differences in study designs and the variety of psychometric tests
used to assess aspects of social cognition, it is difficult to compare the magnitude of associations across
studies to characterize important lifestages and time periods of Pb exposure. There is some examination
of different exposure measurement windows within studies. In the case-control study of twins described
previously, Arora et al. (2017) used laser ablation inductively coupled plasma mass spectrometry (ICP-
MS) to estimate pre- and postnatal Pb exposure from shed deciduous teeth. Differences in tooth Pb levels
were consistently higher in discordant ASD twins across the exposure period (20 weeks prenatal to
30 weeks postnatal) compared with concordant and control twins, with bimodal peaks around 10 to
15 weeks before birth and 10 to 20 weeks after birth (see Figure 3-14). This is consistent with results
from a birth cohort study that reported negative associations between maternal BLLs and social cognition
in infants (Rvgiel et al.. 2021). The observed associations were strongest in magnitude with maternal
BLLs measured in the second trimester compared with BLLs in the first and third trimesters. Although
the limited number of studies that evaluate different exposure windows makes it difficult to draw firm
conclusions on critical lifestages, the nature of ASD as a developmental disorder suggests that prenatal
and early infant exposures may be of particular importance.

3.5.7.4 Integrated Summary and Causality Determination: Social Cognition and
Behavior

The 2013 Pb ISA (U.S. EPA. 2013a) did not include a causality determination for social
cognition and behavior in children. There were no epidemiologic studies on social cognition and behavior
in children in the previous ISA, and only a few toxicological studies that examined social behavior in
mice. The number of studies examining autism and social cognition in relation to Pb exposure has
increased substantially since the 2013 Pb ISA (U.S. EPA. 2013a). highlighted by recent epidemiologic
studies that provide some evidence that Pb exposure is associated with increased ASD incidence and
symptomology, as well as decrements in social, emotional, and adaptive abilities. Recent toxicological
evidence, along with studies reviewed in the 2013 Pb ISA, provide some evidence of Pb-induced changes
in social behavior in mice, but the direction of the observed changes was inconsistent.

External Review Draft

3-167

DRAFT: Do not cite or quote

-------
A recent novel epidemiologic analysis of twins provides strong evidence of an association
between Pb exposure and ASD. Arora et al. (2017) examined tooth Pb levels with respect to ASD status
among discordant and concordant twin pairs and observed higher Pb levels in the affected twin among
discordant monozygotic and dizygotic pairs. In contrast, concordant twins demonstrated similar levels of
exposure. This study also provided some insight into a sensitive time window in which the association
between tooth Pb levels and autistic status was highest between 10-15 weeks before birth and 10-
20 weeks after birth. Additional support was provided by a prospective cohort study, which reported that
the number and severity of autistic behaviors in young children was positively associated with low BLLs
(geometric mean: 1.58-1.64 (ig/dL) at several points prior to outcome assessment (Kim et al.. 2016).
There is some uncertainty about the relevance of the exposure window in this study given that the earliest
Pb measurements occurred at 7-8 years old, which is close to the outcome assessment age (11-12 years
old) and later than autistic behaviors typically manifest. Additionally, covariates in this study did not
include maternal age, which is an important potential confounder for developmental disorders like autism;
therefore, lack of adjustment for this variable weakens the conclusions that can be drawn from the
analysis. Dong et al. (2022) provides support for the positive association between autism severity and
BLLs among children 2 to 13 years old at low levels of current blood Pb (mild group mean: 2.58 (ig/dL;
moderate/severe group mean: 3.25 (.ig/dL): however, the study's retrospective design and the wide range
of the ages of assessed children introduce uncertainty regarding potential reverse causality.

Several prospective studies among children without autism provide some additional support for
associations between Pb exposure and measures of social impairment in children (Nozadi et al.. 2021;
Nvanza et al.. 2021: Rygiel et al.. 2021: Shekhawat et al.. 2021: Zhou et al.. 2017: Lin et al.. 2013).
Median or geometric mean maternal and cord BLLs in these studies ranged from 3.3 to 5.5 (ig/dL, and
BLLs measured in children ranged from 2.7 to 3.9 (ig/dL. These studies had moderate to good follow-up
participation rates, and follow-up durations ranged from 6.5 months to 3 years. Furthermore, they
demonstrated good confounder control, adjusting for maternal age and some measure of SES or parental
education. Notably, the use of non-specific composite test scores (Vigeh et al.. 2014) and lesser-used
subscales (Nvanza et al.. 2021: Rygiel et al.. 2021) limits the specificity and generalizability of some
studies. Additionally, results from recent studies were not entirely consistent, as some analyses did not
observe associations (Rahbar et al.. 2021: Skogheim et al.. 2021: Dohertv et al.. 2020: Ruebner et al..
2019; Rahbar et al.. 2015). These included mostly case-control studies, one prospective cohort study, and
one cross-sectional study. Although all three case-control studies adjusted for maternal age and various
relevant covariates among matched pairs, Ruebner et al. (2019) did not.

Two toxicological studies in the 2013 Pb ISA reported a potential sex-based effect modification
of the effect of Pb exposure on social behavior (Donald et al.. 1987. 1986). Female and male mice
exhibited social interaction and non-social behavior at different timings and in different directions. One
recent study observed that rats exposed to Pb made fewer ultrasonic vocalizations than did control rats at
PND 4 and 12 but not at PND 7 and 10 (Tartaglione et al.. 2020). The authors additionally did not

External Review Draft

3-168

DRAFT: Do not cite or quote

-------
observe differences between exposed and control rats on the homing test, which evaluates olfactory
discrimination and social preference.

In summary, a number of recent prospective epidemiologic studies provide evidence of a positive
association of increases in blood, bone, and tooth Pb levels with reduced social cognition and increased
autistic behaviors in children. The epidemiologic evidence, however, is not entirely consistent.
Additionally, with the exception of a novel case-control study that provides strong support for a positive
association between dentine Pb levels and autism risk, much of the epidemiologic evidence on ASD is
limited by the potential for unmeasured confounding by maternal age or uncertainty regarding the Pb
exposure window. Furthermore, the wide range of measures used in the evaluated studies simultaneously
adds dimensionality and complicates interpretation of the results. Only one recent experimental animal
study on Pb exposure and social cognition was available. This study, combined with the toxicological
evidence reviewed in the previous ISA suggests that Pb exposure may influence social cognition and
communication, though the direction of these effects is inconsistent. In summary, the body of evidence
is suggestive of, but not sufficient to infer, a causal relationship between Pb exposure and social
cognition and behavior in children.

Table 3-8 Evidence that is suggestive of, but not sufficient to infer, a causal
relationship between Pb exposure and social cognition and
behavior in children.

Rationale for

Causality
Determination3

Key Evidence13

References'3

Pb Biomarker Levels
Associated with Effects0

Consistent evidence
from a few high-
quality epidemiologic
studies with relevant
blood, bone, and
tooth Pb levels

Greater difference in tooth Pb
levels among twins discordant
for ASD status than among
concordant twins.

Arora etal. (2017)

Lower scores on test of social
cognition in a prospective study
in South Korea in association
with earlier childhood and
concurrent mean BLLs.

Kim etal. (2016)

Deciduous tooth Pb NR
(early and postnatal Pb
levels)

Child blood Pb GM:
7-8 y: 1.64 pg/dL
9-1 Oy: 1.58 pg/dL
11—12 y: 1.58 pg/dL

Evidence from multiple
prospective cohort studies for
small decrements in scores on
tests of social cognition among
children without autism ages
6.5 mo-3 yr at low levels of
exposure.

Shekhawat etal. (2021)

Rvaiel etal. (2021)

Cord blood Pb GM:
4.14 pg/dL

Mat. blood Pb GM (SD):
1st tri.: 5.27 (1.93) pg/dL
2nd tri.: 4.74 (1.96) pg/dL
3rd tri.: 4.98 (1.93) pg/dL

Infant blood GM (SD):
12 mo: 3.92 (1.80) pg/dL

External Review Draft	3-169	DRAFT: Do not cite or quote

-------
Rationale for

Causality
Determination3

Key Evidence13

References'3

Pb Biomarker Levels
Associated with Effects0

24 mo: 3.49 (1.93) pg/dL

Zhou etal. (2017)

Mat. blood Pb GM (95%
CI): 3.30 (3.05, 3.57) pg/dL

Section 3.5.7.1

Epidemiologic

Prospective studies had Section 3.5.7.1

studies help rule out

population-based recruitment

chance, bias, and

with moderate to good follow-up

confounding with

participation not conditional on

reasonable

bone Pb/BLLs and social

confidence

cognition scores.



All studies controlled for Table 3-13E



maternal age, education and/or



SES. Some controlled for



HOME score, maternal IQ, and



exposures to other pollutants.

Limited supporting
evidence from case-
control and cross-
sectional studies

Null findings from case-control
studies conducting adjusted
mean comparisons of ASD
cases and typically developing
controls with lower and higher
mean or median BLLs,
adjusting for maternal age,
various demographic and
lifestyle factors and dietary
consumption.

Skoaheim et al. (2021)

Rahbaretal. (2015)

Mat. blood Pb GM
cases: 0.83 pg/dL
controls: 0.88 pg/dL

Child blood Pb GM (SD)
cases: 2.25 (2.23) pg/dL
controls: 2.73 (1.85) pg/dL

Rahbaretal. (2021)

Child Blood Pb GM
cases: 7.11 pg/dL
controls: 8.48 pg/dL

Greater BLLs among children
2-13 years old with
moderate/severe vs. mild
autism in a retrospective study.

Dona et al. (2022)

Child blood Pb mean (SD)
Mild: 2.58 (1.08) pg/dL
Moderate/severe: 3.25
(1.89) pg/dL

Null finding from cross-sectional
study. Lacked control for
maternal age at delivery.

Ruebner et al. (2019)

Child blood Pb med:
1.2 pg/dL

Limited experimental
animal evidence at
relevant exposures

Mixed evidence of enhanced or
reduced social interaction
behavior among Pb-exposed
mice. Some suggestion of sex-
specific effect modification.

Donald et al. (1986)

Donald et al. (1987)

External Review Draft

3-170

DRAFT: Do not cite or quote

-------
Rationale for

Causality	Key Evidence*	References*	Associated wfth Efflits^

Determination3	Associated witn tnects

Reduced ultrasonic	Tartaalione etal. (2020) Med blood Pb after

vocalizations among Pb-	exposure during

exposed rats. No evidence of a	pregnancy and lactation:

difference on homing tests of	0.26 |jg/ml_ PND 23
olfactory discrimination and
social preference.

ASD = autism spectrum disorder; BLL = blood lead level; CI = confidence interval; GM = geometric mean; HOME = Health
Outcomes and Measures of the Environment; IQ = intelligence quotient; Mat = maternal; med = median; mo = month(s); NR = not
reported; Pb = lead; PND = postnatal day; SD = standard deviation; SES = socioeconomic status; tri = trimester; yr = year(s).
aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Table I and Table II of the
Preamble to the ISAs (U.S. EPA, 2015).

bDescribes the key evidence and references, supporting or contradicting, contributing most heavily to causality determination and,
where applicable, to uncertainties or inconsistencies. References to earlier sections indicate where the full body of evidence is
described.

°Describes the Pb biomarker levels at which the evidence is substantiated.

3.6 Nervous System Effects Ascertained during Adult Lifestages

The strongest evidence of Pb-associated nervous system effects in adults without occupational
exposure pertained to cumulative exposure and cognitive effects (U.S. EPA. 2013b). Prospective studies
indicated associations of higher baseline tibia (means 19, 20 jxg/g) or patella (mean 25 jxg/g) Pb levels
with declines in cognitive function in adults (age >50 years) over 2- to 4-year periods. Pb-associated
cognitive function decrements were found with adjustment for potential confounding factors such as age,
education, SES, current alcohol use, and current smoking. Supporting evidence was provided by cross-
sectional studies, which found stronger associations with bone Pb level than concurrent BLL. Cross-
sectional studies also considered more potential confounding factors, including dietary factors, physical
activity, medication use, and comorbid conditions. The multiple exposures and health outcomes examined
in many studies reduced the likelihood of biased participation specifically by adults with higher Pb
exposure and lower cognitive function. Uncertainties remained due to residual confounding by age and
lack of information on the patterns of exposure associated with the BLLs observed in the epidemiologic
studies.

3.6.1 Cognitive Function in Adults

The evidence reviewed in the 2013 Pb ISA was sufficient to conclude that "a causal relationship
is likely to exist" between long-term cumulative Pb exposure and cognitive function decrements in adults
(U.S. EPA. 2013a). Prospective studies of the Normative Aging Study (NAS) and Baltimore Memory
Study (BMS) cohorts indicated associations of higher baseline tibia (means 19, 20 jxg/g) or patella (mean
25 jxg/g) Pb levels with declines in cognitive function in adults (age >50 years) over 2- to 4-year periods

External Review Draft

3-171

DRAFT: Do not cite or quote

-------
among adults without occupational exposure (see Table 4-10 (U.S. EPA. 2013a)'). While the specific
covariates differed between studies, these bone Pb-associated cognitive function decrements were found
with adjustment for potential confounding factors such as age, education, SES, current alcohol use, and
current smoking. Supporting evidence was provided by cross-sectional analyses of the NAS, BMS, and
the Nurses' Health Study (NHS), which found stronger associations with bone Pb level than concurrent
BLL. Cross-sectional analyses also considered more potential confounding factors, including dietary
factors, physical activity, medication use, and comorbid conditions. The multiple exposures and health
outcomes examined in many studies reduced the likelihood of biased participation specifically by adults
with higher Pb exposure and lower cognitive function. The specific timing, frequency, duration, and
magnitude of Pb exposures contributing to the associations observed with bone Pb levels were not
discernable from the evidence. Further, there was potential for residual confounding by age. The effects
of recent Pb exposures on cognitive function decrements in adults were indicated in Pb-exposed workers
by associations found with BLLs, although these studies did not consider potential confounding by other
workplace exposures. The biological plausibility for the effects of Pb exposure on cognitive function
decrements in adults was provided by findings that relevant lifetime Pb exposures from gestation, birth, or
after weaning induce learning impairments in adult animals and by evidence for the effects of Pb altering
neurotransmitter function in the hippocampus, prefrontal cortex, and nucleus accumbens.

Recent epidemiologic studies provide consistent evidence that higher bone Pb levels or past
childhood BLLs, capturing cumulative or past exposures, are associated with decrements in cognitive
function (one or multiple domains) during young-, mid- or older-adulthood periods (Table 3-14E). Across
populations, higher Pb levels were associated with decrements in FSIQ, global cognitive function,
executive function, visuospatial and visuomotor skills, language, and memory. Much of this evidence was
provided by extended analyses (about 15 years of follow-up data) of the NAS and NHS cohorts
considered in the 2013 Pb ISA, and additional recent evidence from prospective cohort studies from
Sweden and New Zealand that explored the effects of early childhood Pb exposure (7-12 years) on IQ
and various cognitive domains during young adulthood (18-19 years). Findings from these recent
prospective cohort studies, even after adjustments of various sociodemographic factors and maternal and
childhood IQ, emphasize the important role of early childhood Pb exposure and persistent effects on adult
cognition. Overall, the longitudinal design with longer follow-up periods, multiple and repeatedly
measured cognitive outcomes, and multiple risk factors and confounders accounted for in the studies
reduce the bias and strengthen the study findings related to the effects of Pb exposure on adult IQ and
cognitive function. The specific frequency, duration, and magnitude of Pb exposures that influence the
Pb-cognitive function associations are yet to be fully understood. In addition, some uncertainties remain.
Sex and age differences in bone kinetics and turnover may contribute to differences in biomarker Pb
levels and add uncertainty in modeling. Further, variability within and between bone Pb biomarkers
associated with specific cognitive domains also adds to uncertainty in the observed associations. Recent
evidence from animal studies supports the notion that postnatal exposure to Pb (either during adolescence
or continuing into adulthood) negatively affects learning and memory in rodents. Additionally, adult
rodents exposed during early developmental periods displayed impairments in tests of learning and

External Review Draft

3-172

DRAFT: Do not cite or quote

-------
memory conducted in adulthood (reviewed in Section 3.4). This suggests that early life Pb exposure
contributes to cognitive dysfunction that persists into adulthood, which is new evidence in this review.
Additionally, animals exposed to Pb during adulthood display similar cognitive impairment, though there
is still uncertainty regarding the influence of age on Pb exposures during adulthood. These studies add to
the current evidence base suggesting a potential role of both early and later life Pb exposures and
biological plausibility for the effects of Pb exposure on cognitive function decrements in adults.

A summary of the recent evidence, which is interpreted in the context of the entire body of
evidence, is provided in the subsequent sections. Measures of central tendency for Pb biomarker levels
used in each study, along with other study-specific details, including study population characteristics and
select effect estimates, are highlighted in Section 3.7, Table 3-14E (Epidemiology) and Table 3-4T
(Toxicology).

3.6.1.1 Epidemiologic Studies of Cognitive Function in Adults

The 2013 Pb ISA (U.S. EPA. 2013a) concluded that despite some variability across domains in
cognitive function, the evidence of an association between increased bone Pb levels and decreased
cognition in adults was generally consistent. Domains evaluated included executive function, visuospatial
skills, and learning and memory. The evidence for this conclusion was obtained from racially and
socioeconomically diverse (BMS studies) and predominately white (NAS and NHS studies) cohort study
populations. Several recent epidemiologic studies that examined cognitive performance in adults without
occupational Pb add to the body of evidence; however, uncertainties remain regarding residual
confounding by age and the timing, duration, frequency, and level of Pb exposure that contributed to
associations observed with cognitive function.

Recent studies include several longitudinal and cross-sectional analyses of Pb exposure, or multi-
metal exposures including Pb, that showed declines in cognitive function in adults. Many cross-sectional
studies evaluated associations between concurrent BLLs and various domains of cognitive function, while
longitudinal studies evaluated the adult bone Pb levels or childhood BLLs to assess the relationship
between long-term Pb exposure and cognitive decline. An overview of the recent evidence is provided
below.

Key evidence for Pb-associated cognitive decline among adult populations was provided by
recent longitudinal studies (Reuben et al.. 2020); Farooqui et al. (2017); (Reuben et al.. 2017; SkerfVing et
al.. 2015; Power et al.. 2014). Two of these studies included the extended analysis of the previously
explored Veterans Affairs NAS and NHS cohorts Farooqui et al. (2017); (Power et al.. 2014). In a recent
study including a subset of the NAS cohort with healthy men aged 51-98 years, Farooqui et al. (2017)
examined the associations between long-term Pb exposure quantified using bone biomarkers (mean
patella Pb: 30.6 jxg/g and tibia Pb: 21.6 jxg/g) and longitudinal changes in cognition (repeatedly measured
up to five visits over the 15 years follow-up period) using linear mixed effect and Cox proportional

External Review Draft

3-173

DRAFT: Do not cite or quote

-------
hazard models adjusted for age at the first cognitive test, past education level, baseline smoking status,
and alcohol intake. The study found that higher patella bone Pb concentration (IQR: 21 jxg/g) was
associated with a 0.062 point lower baseline Mini Mental State Examination (MMSE) score (95% CI: -
0.012, 0.003), 0.008 units/year MMSE decline (95% CI: -0.015, 0) over 15 years, and an increased risk
of having an MMSE score below 25 (threshold considered to represent cognitively not normal or at risk
for dementia) (hazard ratio [HR] = 1.10 (95% CI: 0.99, 1.21)). Similar but weaker and less precise
associations were observed when tibia Pb and MMSE outcomes were assessed. The study also used
"global cognition" as a separate proxy for worsening cognitive impairment, and combined seven test
scores assessed in NES2, CERAD, and WAIS-R. Weaker associations were observed between both
patella and tibia Pb and global cognition (both baseline and longitudinal change). When separate
cognitive domains were assessed, patella Pb was associated with faster longitudinal decline in language
and memory domains, whereas similar but weaker associations were observed with tibia Pb. Another
longitudinal study examined a subset of the NHS cohort with healthy women aged 45-74 years (Power et
al.. 2014). The authors examined the associations between past Pb exposures using bone and blood
biomarkers (mean patella Pb: 12.6 jxg/g and tibia Pb: 10.5 jxg/g; mean blood Pb: 2.9 (ig/dL) and cognitive
decline (repeatedly measured using a telephone battery of cognitive tests assessing learning, memory,
executive function, and attention during 2-4 waves over the 13-year follow-up period) using linear mixed
effect models adjusted for alcohol consumption, smoking status, education, husband's education,
menopausal status and hormone therapy use, physical activity, ibuprofen use, aspirin use, vitamin E
supplementation, percentage of residential census tract of white race or ethnicity, and median income of
residential census track. A weak and imprecise association was observed for an excess annual decline in
the overall cognitive test Z-score per SD increase in tibia bone Pb concentration (-0.002 standard units;
95% CI: -0.005, 0.000). When individual cognitive tests were considered, a decline on the East Boston
Memory Test as well as immediate (a measure of episodic memory) and category fluency (a measure of
executive function and memory) was observed in relation to increased tibia Pb concentration (Power et
al.. 2014). There was little evidence for associations between patella Pb or blood Pb and the decline in
overall cognition, verbal memory, or individual cognitive tests. Associations between either tibia or
patella Pb concentration and cognitive function decline observed in these longitudinal studies are
supported by (Weuve et al.. 2013). a cross-sectional study that assessed bone Pb and cognitive function
(assessed using telephone cognitive assessment battery) among participants from an existing case-control
study. The Pb concentration observed in this study is similar to that reported in the NHS cohort. Separate
analyses were performed for the group of participants with PD and for all participants including both PD
and control groups. Analysis of the PD group showed that higher tibia Pb was significantly associated
with worse overall performance (as shown by the global cognitive score) and worse performance on the
majority of telephone cognitive tests (in the model adjusted for age at cognitive assessment, sex, race,
education, smoking history). Patella Pb concentration, however, was not consistently associated with
cognitive performance. In the model with both PD and control groups, interactions were observed for the
association between tibia Pb and global cognition by case-control status. Among the participants, a
10 jxg/g increase in tibia Pb corresponded to a decrease in the global cognitive score by 0.12 standard

External Review Draft

3-174

DRAFT: Do not cite or quote

-------
units (95% CI: -0.22, -0.01), but the association among controls was weak (0.06 standard units, 95% CI:
-0.09, 0.20).

Evidence of the cognitive effects of cumulative Pb exposure observed in the NAS and NHS
studies is strengthened and extended further by the findings from cohort studies that examined the
associations of childhood Pb exposure and continued long-term exposures with cognitive impairments in
young adults (18-19 years) (Skerfving et al.. 2015) or in mid-adulthood (38 or 45 years of age) (Reuben
et al.. 2020; Reuben et al.. 2017). SkerfVing et al. (2015) included samples of 7-12 year-old school
children in southern Sweden and followed them over time to examine the association between childhood
BLL (age 7-12 years old, mean blood Pb: 3.4 (ig/dL) and cognitive performance (IQ) assessed for
military conscription at 18-19 years of age using generalized linear models. The study found an IQ loss of
0.127 (95% CI: -0.209, -0.045) points per (ig/dL increase in childhood BLL for all participants and an IQ
loss of 0.204 (95% CI: -0.392, -0.016) points per (ig/dL increase in childhood BLL among those with
childhood BLLs <50 |ig/L. even in multivariable models adjusted for parent's income, education, and
father's IQ. (Reuben et al.. 2020; Reuben et al.. 2017) examined a New Zealand birth cohort with
participants born in 1972-1973 (a time when Pb exposure in New Zealand cities were higher than
international standards) who were part of the Dunedin Multidiscipl itiary Health and Development Study.
Infants were followed from birth through adulthood. Blood collection at 11 years of age provided blood
biomarker data for Pb (mean blood Pb: 10.99 (.ig/dL). Cognitive performance was assessed using
objective tests of cognitive performance such as the WISC-R during childhood at ages 7 and 9 years.
Cognitive performance was also assessed using the WAIS-IV when participants were 38 years old
(Reuben et al.. 2017) and again when they were 45 years old (Reuben et al.. 2020). (Reuben et al.. 2020).
in addition to the objective tests, also included subjective reports of everyday cognitive functioning
(memory or attention problems) at age 45 years as provided by study participants and their nominated
informants. The studies examined the association between childhood blood Pb and adult cognitive
outcomes or cognitive decline (change in IQ score between childhood and mid-adulthood), using OLS
multiple regression models. (Reuben et al.. 2017) found that each 1 (.ig/dL higher level of blood Pb in
childhood was associated with a 0.39-point lower score in adult FSIQ (95% CI: -0.67, -0.12), and a 0.32-
point decline (95% CI: -0.50, -0.15) after adjusting for sex, childhood IQ, maternal IQ, and childhood
SES. Similarly, with additional years of data (Reuben et al.. 2020) continued to show significant
associations between childhood BLL and IQ at 45 years of age. Each 1 (ig/dL higher childhood BLL was
associated with a -0.41 (95% CI: -0.68, -0.15) point decline in full-scale IQ from baseline. When using a
residualized change model to adjust for autocorrelation between baseline and follow-up IQ, the decline in
IQ was similar (-0.39 [95% CI: -0.58, -0.21]). The study also found that the relationship between
childhood blood Pb and adult IQ persisted and remained significant even after adjustment for brain
structure measures. Results pertaining to brain structure are discussed in Section 3.4.1.

Several recent cross-sectional studies also examined the associations of Pb exposure and
decrements in cognitive function. The majority used concurrent blood biomarkers (Sasaki and Carpenter.
2022; Xiao et al.. 2021; Przvbvla et al.. 2017; Souza-Talarico et al.. 2017; Khalil et al.. 2014; van

External Review Draft	3-175	DRAFT: Do not cite or quote

-------
Wiingaarden et al.. 201IV A few studies used other biomarkers such as toenails (Me ram at et al.. 2017).
bone ("Weuve et al.. 2013). or urine (Sasaki and Carpenter. 2022). As described further below, results
were not entirely consistent across studies.

Three of the cross-sectional studies examining the blood Pb-cognitive function association used
data from various NHANES cycles including participants aged 60-84 years (Sasaki and Carpenter. 2022;
Przvbvla et al.. 2017; van Wiingaarden et al.. 2011). van Wiingaarden et al. (2011) examined the
associations of blood Pb (mean blood Pb: 2.46 (ig/dL) with self-reported confusion and memory problems
using data from NHANES (1999-2008). Data from the 1999-2002 NHANES cycle were used to estimate
the association between BLL and performance on the Digit Symbol Substitution Test (DSST) scores.
After adjustment for age, sex, education level, ethnicity, poverty-income ratio (PIR), self-reported health
status, and comorbid conditions, no association of BLLs with self-reported confusion or memory
problems or DSST performance was observed (van Wiingaarden et al.. 2011). Two other studies,

Przvbvla et al. (2017) and Sasaki and Carpenter (2022). included the NHANES participants from 1999-
2002 and 2011-2014 cycles, respectively. These studies explored the cross-sectional associations of blood
and urine biomarkers of multiple metals and metalloids (separately and jointly) on cognitive function.
(Przvbvla et al.. 2017) used a path analysis approach to model multiple exposures of 14 chemicals
simultaneously while adjusting for multiple comparisons. The study found that the association of BLL
(Geometric mean: 2.17 (ig/dL) with lower cognition scores was attenuated when the model controlled for
smoking status. Specifically, a 1-SD increase in BLL was weakly associated with a slightly lower Digit
Symbol Coding (DSC) test score from WAIS-III (f3 = -0.10, 95% CI: -0.20, -0.00) after controlling for
co-exposure and sociodemographic covariates. The study also performed stratified analysis by sex and
age (above and below median age) and found a greater magnitude of associations for female and higher
age categories (>10%), despite a lack of statistical evidence of an interaction. (Sasaki and Carpenter.
2022) used two stage linear regression models. First, they performed single metal analyses separately for
each of seven metals or metalloids followed by a second analysis including multiple metals or metalloids
from the stage 1 analysis to examine the associations with immediate, delayed, and working memory
quantified using CERAD and DSST. When single metals were assessed, increased blood Pb concentration
was associated with decrements in performance on all three cognitive tests after adjusting for
sociodemographic. behavior, and clinical characteristics (immediate recall: (3 = -0.58, 95% CI: -0.91,
-0.24; delayed recall: (3 = -0.19, 95% CI: -0.35, -0.02; Digital Symbol Substitution: (3 = -1.08, 95% CI:
-2.12, -0.05). Multi-metal analysis stratified by age group (60-70 and >70 years old) suggested greater
declines in immediate recall among participants over the age of 70. Khalil et al. (2014) examined the
association between concurrent blood Pb concentration (mean blood Pb: 2.25 (ig/dL) and cognitive
function among a subset of non-institutionalized community dwelling non-Hispanic Caucasian men
65 years and older who participated in the Osteoporotic Fractures in Men Study (MrOS) cohort study.
Cognitive function was assessed using the Modified MMSE (3MS) and the Trail Making Test Part B.
Higher scores on the 3MS and faster time on the Trail Making Test Part B both represent better
performance. Multivariable analysis found no association between blood Pb concentration and cognitive
function (Khalil et al.. 2014).

External Review Draft

3-176

DRAFT: Do not cite or quote

-------
Souza-Talarico et al. (2017) examined the association between blood Pb (and interactions
between blood Pb and Cd) and working memory capacity (WMC) in a population of 125 older adults
aged 50-82 years, in the metropolitan area of Sao Paulo, Brazil. The study also explored the mediating
role of antioxidant capacity (using various oxidative stress biomarkers) in the heavy metals-memory
associations. Using regression models accounting for age, sex, income, and hemoglobin, the study did not
find an association between blood Pb (mean 2.1 (ig/dL) and WMC ((3 = 0.106, 95% CI:-0.208, 0.417);
however, an interaction between blood Pb and blood Cd level was observed as well as a significant
inverse association between the blood Cd x blood Pb interaction term and WMC was observed (Souza-
Talarico et al.. 2017). The Monte Carlo Method for Assessing Mediation test for mediation revealed that
the association between the blood Cd x blood Pb interaction term and WMC was significantly mediated
by total antioxidant capacity.

(Xiao et al.. 2021) examined the association between multiple metals (22 metals including Pb;
mean blood Pb: 5.15 (.ig/dL) and cognitive function measured using MMSE in participants aged >60 years
from Guangxi, Southern China. The study used least absolute shrinkage and selection operator (LASSO)
penalized regression to identify main metals associated with cognitive function. Twelve metals (including
Pb) selected from LASSO were then explored in a multi-metal generalized linear regression model
adjusted for age, gender, education attainment, annual income, BMI, smoking, alcohol drinking,
insomnia, and physical activity. No association was observed for blood Pb and cognitive function after
adjustment for other metals.

Notably, a limitation of cross-sectional studies of concurrent BLLs is that the relative contribution
of the recent versus past Pb exposure is not well characterized. A recent prospective study was designed
to address the uncertainties related to the exposure patterns associated with BLLs observed in studies of
adults (Yu et al.. 2021). Yu et al. (2021) examined the association of BLLs and neurocognitive
performance among newly hired employees at battery manufacturing and Pb recycling plants with no
previous occupational Pb exposure, a subset of participants in the Study for Promotion of Health in
Recycling Lead (SPHERL) cohort study. Baseline blood Pb concentration was measured, and the
participants were followed annually over a 2-year period to measure blood Pb biomarkers and assess if
higher recent occupational exposure to Pb was associated with neurocognitive dysfunction. The
participants completed the DSST and SCWT at baseline and annual follow-up visits. The geometric mean
blood Pb at baseline and first and second follow-up visits were 3.97 (ig/dL, 13.4 (ig/dL, and 12.8 (ig/dL,
respectively, showing an almost three-fold increase in blood Pb over the 2 years of occupational
exposure. The study used a linear mixed model to examine the changes in DSST and SCWT
corresponding to changes in blood Pb separately for the 1- and 2-year visits. Despite the three-fold
increase in blood Pb concentration, the study found no association between blood Pb and cognitive
function. The change in latency time and error rate based on the DSST test showed an increase from
baseline to follow-up, with an increase in the follow-up-to-baseline blood Pb concentration ratio, but the
association was weak and imprecise in the fully adjusted models (change in latency: 0.55%, 95% CI:
-0.33, 1.42; error rate: OR: 1.01, 95% CI: 1.00, 1.03)

External Review Draft

3-177

DRAFT: Do not cite or quote

-------
3.6.1.1.1 Summary

Longitudinal cohort studies evaluated in the 2013 Pb ISA indicated that there was generally
consistent evidence of an association between increased bone Pb levels and decreased cognition in adults
despite some variability in the associations observed between Pb biomarkers and domains in cognitive
function (U.S. EPA. 2013a). Recent prospective cohort studies add to the body of evidence informing the
relationship between Pb exposure and cognitive performance in adults without occupational Pb exposure.
More specifically, recent cohort studies indicated that higher adult bone Pb levels, which indicate
cumulative Pb exposure (tibia mean range: 10.5, 21.6 jxg/g, patella mean range: 12.6. 30.6 jxg/g) or
childhood BLLs (mean range: 3.4 ug/dL, 10.99 ug/dL at 7 12 years of age), were associated with
decrements in cognitive function or IQ during young-, mid-, or older-adulthood periods (Table 3-14E).
There was some variability within and between bone Pb biomarkers in the associations with various
domains of cognitive function tested within studies; however, higher Pb levels were associated with
decrements in FSIQ, global cognitive function, executive function, visuospatial and visuomotor skills,
language, and memory. Extended analyses of the NAS and NHS cohorts with 13 to 15 years of follow-up
add to the evidence base (Farooqui et al.. 2017; Power etal.. 2014). These studies found associations of
cumulative Pb exposure with decrements in cognitive function in adults after adjustment for potential
confounding by combinations of factors including demographic, socioeconomic, behavioral, clinical, and
neighborhood-level factors. In addition, findings from recent prospective cohort studies in Sweden and
New Zealand that explored the effects of early childhood Pb exposure (7-12 years) on IQ and cognitive
domains during young adulthood (18-19 years) SkerfVing et al. (2015) and mid-adulthood (38-45 years)
(Reuben et al„ 2020; Reuben et al.. 2017) found that higher childhood BLLs were associated with
declines in IQ ascertained in adulthood after adjustment for demographic and socioeconomic factors,
maternal IQ. and childhood IQ scores. These findings provide new insight into the persistence of Pb-
associated cognitive function decrements. A single prospective cohort study designed to provide insight
on the effect of recent Pb exposure on cognitive function found no association in young workers despite
the increase in Pb exposure from a geometric mean of 4 (ig/dL at baseline to 13 (ig/dL over the 2-year
study period (Yu et al.. 2021 (.Overall, the longitudinal design with longer follow-up periods, multiple and
repeatedly measured cognitive outcomes, and multiple risk factors and confounders accounted for in
epidemiologic studies investigating long-term cumulative exposure and early childhood exposure reduce
uncertainties and strengthen the overall evidence related to the association of Pb exposure with cognitive
function in adulthood. The specific frequency, duration, and magnitude of Pb exposures that influence the
Pb-cognitive function associations are yet to be fully understood. In addition, some uncertainties remain.
Sex (male versus female, premenopause versus postmenopause) and age (young versus mid-aged versus
old-aged adults) differences in bone kinetics and turnover, as well as disease comorbidity, particularly at
middle- and older-adulthood lifestages may potentially lead to differences in bone Pb and blood Pb loads
and add complexity when modeling the associations since inclusion of only age or sex in the model may
not fully account for these differences. Further, variability within and between bone Pb biomarkers
associated with specific cognitive domains also adds to uncertainty in the observed association.

External Review Draft

3-178

DRAFT: Do not cite or quote

-------
3.6.1.2

Toxicological Studies of Cognitive Function in Adults

3.6.1.2.1 Learning and Memory - Morris Water Maze

This section specifically reviews studies that exposed animals to Pb during either adulthood or
late adolescence. Studies that exposed animals during development (i.e., pregestation, gestation, lactation)
are reviewed in Section 3.5.3.2. Animals exposed to Pb via drinking water in adulthood displayed
impaired learning and memory. Using the Morris water maze, Mansouri et al. (2012) found that short-
term Pb exposure (50 mg/L Pb in drinking water, PND 70 to 100), which produced mean BLLs of
8 (ig/dL, significantly impaired both learning and memory, though the magnitude of the effect on memory
was smaller compared with the effect observed in other studies of developmental exposure
(Section 3.5.3.2). Also using the Morris water maze, Mansouri et al. (2013) reported that long-term Pb
exposure (50 ppm in drinking water, PND 60 to 240), which produced peak BLLs of 11-19 (ig/dL,
significantly impaired learning and memory performance in both sexes. Cognitive impairment was also
observed in studies that utilized daily administration of Pb via gavage. Singh et al. (2019) reported
significantly increased escape latencies and path lengths (i.e., distance traveled to reach the platform,
another measure of learning) in exposed rats following long-term Pb exposure via gavage (2.5 mg/kg,
PND 90 to 180), which produced peak BLLs of 28 (ig/dL. No probe phase was conducted in this study.
Additionally, Su et al. (2016) found that male rats gavaged with Pb solutions daily (200 ppm, PND 20 to
76) displayed significant impairments in both the learning and memory components of the Morris water
maze.

In two recent studies, Zou et al. (2015) and Han et al. (2014) reported significant learning and
memory deficits following short-term exposure of juvenile animals. Zou et al. (2015) exposed mice from
PND 35 to 56 with mean BLLs of 22 (ig/dL, while Han et al. (2014) exposed rats from PND 21 to 42 and
reported mean BLLs of 15 (ig/dL. Other studies examined the effects of long-term Pb exposure on
juvenile rodents and found similar effects. For example, (An et al.. 2014) exposed groups of juvenile rats
to multiple doses of Pb for 56 days (PND 28 to 84). All examined doses (100, 200, and 300 ppm in
drinking water) produced BLLs relevant to this ISA. At the time of Morris water maze assessment
(PND 84), the mean BLL ranged from 11 to 23 (ig/dL. All exposed animals displayed impaired memory
during the probe trial relative to controls. Only animals in the two highest dose groups were reported to
show learning deficits during training, suggesting that, with juvenile exposures, Pb may have a greater
effect on memory than learning processes. Another study that exposed juvenile mice to Pb in drinking
water (0.2%) for 90 days (PND 28 to 112) assessed Morris water maze performance in the same animals
at multiple time points during and immediately following exposure (Wu et al.. 2020b). This long-term
exposure produced relatively high mean BLLs of 28 (ig/dL, and all exposed animals showed signs of
impaired learning and memory in the maze. Interestingly, both measures of cognition improved in the
exposed animals over time, which may reflect either increasing familiarity with the task or clearance of
Pb over time. In contrast to all other studies in young and juvenile animals, Li et al. (2013) reported that

External Review Draft

3-179

DRAFT: Do not cite or quote

-------
rats given Pb in drinking water for 84 days (from PND 28 to 112) with peak BLLs of 16 (ig/dL showed no
indication of learning or memory impairment in the Morris water maze. Despite this one discrepant study,
recent evidence supports the notion that postnatal exposure to Pb (either during adolescence or continuing
into adulthood) negatively affects learning and memory in rodents, which contrasts with several of the
key studies reviewed in the previous ISA.

3.6.1.2.2 Summary

Four recent studies of rodents with exposure resulting in mean BLLs <30 (ig/dL add to the
evidence informing the association of both short- and long-term Pb exposure during adulthood with
measures of learning and memory in rodents. While these studies are consistent with one another,
toxicological evidence for the effects of Pb on cognitive function in adults remains limited. Additionally,
a few recent studies in juvenile rodents also provide some support for the association between postnatal
Pb exposure either during adolescence or continuing into adulthood and cognitive impairment,
specifically learning and memory.

3.6.1.3 Relevant Issues for Interpreting the Evidence Base

3.6.1.3.1 Concentration-Response Function

The 2013 Pb ISA reviewed a small number of studies that examined the shape of the C-R
relationship between blood or bone Pb levels and cognitive function. Studies using BMS and NAS
cohorts assessed nonlinearity using quadratic terms, penalized splines, or visual inspection of bivariate
plots. Prospective analyses of the NAS cohorts provided some evidence of nonlinearity (Wang et al..
2007; Weisskopf et al.. 2007) Figures 4-7 and 4-8 from 2013 Pb ISA). Weisskopf et al. (2007) found that
a 20 jxg/g difference in patella Pb level was associated with a 0.07-ms increase in response latency (95%
CI: 0.04, 0.12; larger values mean slower reaction times in the pattern comparison test) among all men
and a 0.15-ms increase among men with patella Pb level <60 jxg/g. These results suggest that Pb-
associated latency worsens with increasing Pb up to 60 jxg/g and levels off at higher values. Similarly,
Wang et al. (2007) found that among NAS men with an HFE gene variant, there was a larger decline in
MMSE score (a global examination of cognitive function with low scores indicating poor cognitive
performance) per unit increase in tibia Pb level at higher tibia Pb levels.

In the current review, the shape of the C-R function was not assessed in studies that examined the
associations of Pb biomarkers with cognitive function in adults. The majority of studies selected
analytical models that assumed linear associations in the Pb-cognitive function associations. A few
studies in the recent review examining the influence of early childhood Pb exposure on cognitive
impairments at the young- (18-19 years) (Skerfving et al.. 2015) or mid-adulthood periods (38 or 45 years

External Review Draft

3-180

DRAFT: Do not cite or quote

-------
of age) Reuben et al. (2017) and Reuben et al. (2020) performed restricted analysis for a subset of the
population exposed to higher or lower Pb levels and provided insight into possible nonlinear relationships
or threshold effects. SkerfVing etal. (2015) examined the influence of early childhood Pb exposure on
long-term cognitive impairments at young adulthood (18-19 years) using generalized linear models. The
study found an IQ loss of 0.127 (95% CI: -0.209, -0.045) points per (ig/dL increase in childhood BLLs for
all participants, and a slightly larger IQ loss (i.e., 0.204 [95% CI: -0.392, -0.016] point per (ig/dL increase
in childhood BLL) for the populations with childhood BLLs <50 |ig/L. (Reuben et al.. 2020; Reuben et
al.. 2017) examined the association of childhood blood Pb (11 years) with cognitive performance during
mid-adulthood and cognitive decline (change in IQ score between childhood and mid-adulthood) for the
overall sample as well as separately for participants above or below the historic level of concern (i.e.,
>10 (ig/dL). Reuben et al. (2017) found a 1,97-IQ-point reduction in adulthood (95% CI: -3.34, -0.59)
for the overall sample, a 4.25-IQ-point reduction for individuals above the level of concern, and a 2.73-
IQ-point reduction for individuals below the level of concern for each 5-(ig/dL increase in the childhood
blood Pb. Similarly, IQ decline from childhood to adulthood suggested a mean decline of a 1.61 IQ points
(95% CI: -2.48, -0.74) in adulthood for the overall sample, a mean decline of 1.68 IQ points for
participants above the level of concern, and a mean increase of 1.22 IQ points for participants below the
level of concern for each 5-(ig/dL increase in the childhood blood Pb. Similar results were observed by
(Reuben et al.. 2020) in the same population studied in (Reuben et al.. 2017) and followed till 45 years of
age. Overall, these childhood exposure studies suggested persistence and continued cognitive effects of
childhood Pb exposure through mid-adulthood, and the strength of associations were higher in magnitude
for the participants with childhood exposure above the historic level of concern (>10 (ig/dL).

The limited recent toxicological evidence generally supports the dose-dependent effects of Pb on
cognitive function at relevant BLLs in adult animals. Only one study examined juvenile animals exposed
to multiple concentrations of Pb and reported greater decrements in learning and memory at higher doses
(An et al.. 2014).

3.6.1.3.2 Potentially At-Risk Populations
Age and Sex:

In the 2013 Pb ISA, an analysis using the NAS cohort reported an interaction between Pb and age
(Wright et al.. 2003). The study reported that the inverse association between age and cognitive function
was greater among those with high blood or patella Pb levels. Specifically, in the highest quartile of
patella Pb, each year increase in age led to a four-fold steeper decline in the MMSE score relative to the
effect of age in the lowest quartile of patella Pb. Effect estimates were in the same direction for tibia Pb,
but the interaction was not statistically significant.

External Review Draft

3-181

DRAFT: Do not cite or quote

-------
Two recent epidemiologic studies using NHANES data from the 1999-2002 and 2011-2014
cycles explored the cross-sectional associations of blood and urine biomarkers of multiple metals and
metalloids (separately and jointly) with cognitive function and provided some insights into potential
effect modifications of Pb-associated decrements in cognitive function (Sasaki and Carpenter. 2022;
Przvbvla et al.. 2017). Stratified analysis by sex and/or age groups (above and below median age groups)
performed in these studies suggested a greater magnitude of Pb-cognitive function associations (beta
estimates>10%) for females and for older age categories , however, the statistical test for the interactions
suggested no significant difference between the sex and age categories.

Toxicological studies investigating potential sex differences in Pb-induced cognitive impairment
are limited. A study by Mansouri et al. (2013) reported that Pb produced similar decrements in learning
and memory in both male and female animals. Given the lack of toxicological evidence available, the
possible influence of sex on Pb-induced cognitive impairment in adult animals remains unclear.

Pre-existing conditions:

One study evaluated the association of bone Pb (mean ranges for various age groups: tibia Pb:
4.4-9.2 jxg/g; patella Pb: 5.9-15.2 jxg/g) with cognitive function among individuals with PD and controls
participating in a case-control study ("Weuve et al.. 2013). The patella Pb and tibia Pb concentrations
reported in this study for all study participants increased with increasing age. The highest Pb
concentrations were found in study participants in the 75-81 years old category, and the lowest
concentrations were found in participants in the 54-65 years old category. When the data were analyzed
separately for participants with PD, higher tibia Pb concentration was significantly associated with lower
scores on all of the telephone cognitive tests (adjusted difference in scores per 10 jxg/g increase in bone
Pb: Telephone Interview for Cognitive status (TICS) test: -0.20 [-0.4, -0.00]; digit span forward: -0.23
[-0.43, -0.03]; digit span backward: -0.19 [-0.37, -0.00]) and global cognitive score (adjusted
difference in scores per 10 jxg/g increase in bone Pb: -0.13 [-0.25, -0.01]). When the overall (cases and
control) data were analyzed, significant interactions were observed for the association between tibia Pb
and global cognition by case-control status. Participants with PD showed worse scores compared with
controls (1 SD increase in tibia Pb led to worsening of the global cognitive score by 0.12 units [95% CI:
-0.22, -0.01] among cases; controls: 0.06 [95% CI: -0.09, 0.20]).

Genetics:

Studies investigating the association between Pb levels and cognitive function in 2013 Pb ISA
extensively evaluated the effect modification by ALAD and HFE gene variants. The evidence was
provided by an NHANES analysis (Krieg et al.. 2009) as well as multiple analyses from the NAS cohort
examining different tests of cognitive function (Raian et al.. 2008; Weuve et al.. 2006). In the study using
a cohort from NHANES III, associations with concurrent BLLs were more pronounced in groups with CC
and CG ALAD genotypes (i.e., ALAD2 carriers) for several indices of cognitive function (Krieg et al..

External Review Draft

3-182

DRAFT: Do not cite or quote

-------
2009). In the NAS cohort of men, Weuve et al. (2006) found that higher concurrent BLL but not bone Pb
level was associated with a larger decrease in a test of general cognitive function among ALAD2 carriers.
Another NAS study examined the function of specific cognitive domains (e.g., vocabulary, memory,
visuospatial skills) and found variable evidence for effect modification by ALAD genotype across tests
(Raian et al.. 2008). For example, among ALAD2 carriers, concurrent BLL was associated with a more
pronounced decrease in vocabulary score but less pronounced decrease in a memory index and no
difference in the associations with other cognitive tests. For tibia and patella Pb levels, ALAD genotype
was found to modify associations with different tests, for example, executive function and perceptual
speed. It is not clear why the direction of effect modification would vary among different cognitive
domains. The limited number of populations examined, and the different cognitive tests performed in
each study, make it difficult to conclusively summarize findings for effect modification by ALAD
variants. However, in the limited available body of evidence, blood and bone Pb levels were generally
associated with lower cognitive function in ALAD2 carriers.

Longitudinal analysis of the NAS cohort also indicated that HFE gene variants modified the
blood Pb-cognition association (Wang et al.. 2007). Wang et al. (2007) found an IQR higher tibia Pb level
(15 jxg/g) was associated with a 0.22 point steeper annual decline (95% CI: -0.39, -0.05) in the MMSE,
which assesses cognitive impairment in a number of domains, among the men with at least one HFE
variant allele (H63D or C282Y variant). The association was found to be nonlinear, with larger Pb-
associated declines observed at higher tibia Pb levels. Tibia Pb level was not associated with a decline in
MMSE score in men with the HFE wildtype genotype. Moreover, the deleterious association between
tibia Pb and cognitive decline appeared progressively worse in participants with increasingly more copies
of HFE variant alleles (p-trend = 0.008). These findings suggest that HFE polymorphisms greatly enhance
susceptibility to Pb-related cognitive impairment in a pattern consistent with allelic dose.

None of the studies in the current review examined the effect modification by genetic variants.
Other Metal Exposure:

Various studies that examined other metals either evaluated the relationship of each metal
separately with the outcomes of interest or included multiple metals jointly in the model (Sasaki and
Carpenter. 2022; Xiao et al.. 2021; Przvbvla et al.. 2017) and did not explore the interactions between the
metals during co-exposure. Only one study in this review explored the interaction between metals and
their associations with cognitive function. In a study of adults 50 to 82 years old in Sao Paulo, Brazil,
Souza-Talarico et al. (2017) examined the associations of heavy metals (Cd and Pb) in blood and WMC
separately as well as together in a model for metal interactions. The study found no significant association
between blood Pb and WMC in the model including Pb only, but significant interactions were observed
between blood Cd and blood Pb and the inverse association with WMC ((3 = -0.38, p < 0.001).

External Review Draft

3-183

DRAFT: Do not cite or quote

-------
3.6.1.3.3 Lifestages

The identification of critical lifestages and time periods of Pb exposure is complicated by the fact
that the majority of adult cognitive studies used concurrent adult BLLs. Although possibly affected by
recent exposure, BLLs are also influenced by Pb stored in bone. Thus, associations in adult studies using
concurrent BLL may reflect the effects of past and recent Pb exposures on cognitive outcomes. Some
cohort studies in the 2013 Pb ISA and the current review using bone Pb suggested the effects of
cumulative long-term Pb exposure on cognitive impairment during adulthood. However, it is still difficult
to specifically identify exposures at particular lifestages (prenatal, infancy, early and late childhood, early
adulthood, etc.) that could have led to the long-term cognitive impairment observed in these studies. Few
recent prospective studies evaluating early childhood exposures at 7-12 years of age and long-term
cognitive impairment and decline at young- (18-19 years) (Skerfving ct al.. 2015) and mid-adulthood (38
or 45 years of age) (Reuben et al.. 2020; Reuben et al.. 2017) provided an insight into critical lifestages
(i.e., early childhood and persistence of the cognitive effects through adulthood). SkerfVing et al. (2015)
examined the association of childhood BLL in children from southern Sweden (age 7-12 years old, mean
blood Pb: 3.4 (ig/dL) with cognitive performance (IQ) at the age of 18-19 years. They found an IQ loss of
0.127 (-0.209, -0.045) points per (ig/dL increase in childhood BLL for all participants and an IQ loss of
0.204 (-0.392, -0.016) points per (ig/dL increase in childhood BLL among those with childhood BLLs
<50 |ig/L even in multivariable models adjusted for parent's income, education, and father's IQ. Reuben
et al. (2017)and Reuben et al. (2020) followed a New Zealand birth cohort to examine the association of
childhood Pb level (age 11 years, mean blood Pb: 10.99 (ig/dL) with cognitive performance and decline at
38 years (Reuben et al.. 2017) and 45 years (Reuben et al.. 2020). Reuben et al. (2017) found that each 5-
(ig/dL higher level of blood Pb in childhood was associated with a 1.97-point decrease in IQ score (95%
CI: -3.34, -0.59) and a 1.61-point decline (95% CI: -2.48, -0.74) in adult FSIQ after adjusting for sex,
childhood IQ, maternal IQ, and childhood SES. With additional years of data. Reuben et al. (2020) also
showed a significant association between childhood BLL and IQ at 45 years of age. Each 5-(.ig/dL higher
level of blood Pb in childhood was associated with a 2.07-point decrease in the fiill-IQ score (95% CI:
-3.39, -0.74), and a 1.97-point decline (95% CI: -2.92, -1.03) after adjusting for covariates. These study
findings suggest early childhood is a critical lifestage that can influence childhood cognition and
continued effects through adulthood.

Overall, recent rodent studies of learning and memory evaluating Pb exposure at various
lifestages suggest cognitive impairment. However, the magnitude of the effect at different lifestages has
been shown to differ. Adult animals may be less sensitive than juvenile animals, and juvenile animals
may be less sensitive than animals exposed during development (reviewed in Section 3.5.2). This general
pattern is consistent with evidence describing critical windows for brain development (Section 3.7).
Additionally, critical evidence for the association of Pb with cognitive impairment across lifestages comes
from a series of studies describing the effects of lifetime Pb exposure on nonhuman primates (Rice. 1992;
Rice and Gilbert. 1990a; Rice. 1990; Rice and Karpinski. 1988). Cynomolgus monkeys (Macaca
fascicularis) were dosed continuously from birth and tested repeatedly throughout their lifetime. While

External Review Draft

3-184

DRAFT: Do not cite or quote

-------
these exposures yielded BLLs beyond values considered relevant for the current assessment (>30 (.ig/dL).
they provide key evidence of Pb-induced cognitive impairments that persisted into adulthood in a
translationally relevant species. However, given the limited number of studies conducted in juvenile and
adult animals, and the lack of studies examining the same endpoint across multiple age groups, the precise
role of exposures at various stages on the cognitive effects in adult animals remains unclear.

3.6.1.4 Integrated Summary and Causality Determination: Cognitive Function in
Adults

The 2013 Pb ISA (U.S. EPA. 2013a) concluded that the available evidence was sufficient to
conclude "a causal relationship is likely to exist" between long-term cumulative Pb exposure and
cognitive function decrements in adults. This causality determination was based on a small body of
prospective studies that indicated strong associations of higher baseline tibia (means 19, 20 jxg/g) or
patella (mean 25 jxg/g) Pb levels with declines in cognitive function in adults (age >50 years) over 2- to 4-
year periods among adults without occupational exposure (i.e., NAS and BMS cohorts). Supporting
evidence was provided by cross-sectional analyses of similar cohorts (i.e., NAS, BMS, and NHS), which
revealed stronger associations with cognitive impairment for bone Pb level than for concurrent BLL. The
specific timing, frequency, duration, and magnitude of Pb exposures that contributed to the associations
observed with bone Pb levels were not discernable from the evidence. Further, there was potential for
residual confounding by age. The biological plausibility for the effects of Pb exposure on cognitive
function decrements in adults was provided by findings that relevant lifetime Pb exposures from
gestation, birth, or after weaning induce learning impairments in adult animals and by evidence for the
effects of Pb altering neurotransmitter function in the hippocampus, prefrontal cortex, and nucleus
accumbens.

Results from recent epidemiologic and animal studies add to the evidence base reviewed in the
2013 Pb ISA. Recent epidemiologic studies consistently report that higher bone Pb levels or past
childhood BLLs, capturing cumulative or past exposures, were associated with poor cognitive
performance or decrements in cognitive function during young-, mid-, or older-adulthood periods
(Table 3-14E). However, it is important to acknowledge that there was some variability in the magnitude
and direction of associations observed between different Pb biomarkers (tibia versus patella versus blood
Pb biomarkers) and various cognitive function tests and domains included in the studies. Across
populations, higher Pb levels were associated with decrements in FSIQ, global cognitive function,
executive function, visuospatial and visuomotor skills, language, and memory. Much of this evidence on
adult cognitive outcomes was obtained from analyses of the NAS and NHS cohorts, including recent
analyses that extended follow-up periods beyond the analyses evaluated in the 2013 Pb ISA. Recent
evidence also comes from early childhood exposure cohort studies conducted in Sweden and New
Zealand. These studies strengthen findings that childhood Pb exposures are associated with decrements in
IQ and cognitive function during young- and mid-adulthood. Longitudinal study designs with longer

External Review Draft

3-185

DRAFT: Do not cite or quote

-------
follow-up periods, multiple and repeatedly measured cognitive outcomes, and multiple risk factors and
confounders accounted for in the studies reduce the bias and strengthen the study findings related to Pb
exposure and adult cognitive function. Further, significant findings from new studies specifically
investigating the influence of early childhood Pb exposure on adult IQ and cognitive outcomes, even after
adjustments for various confounders and childhood IQ, provide important insights on the role of early
childhood Pb exposures, not only the effects on childhood IQ and cognitive functions but also the
continued cognitive effects during adulthood. The specific timing, frequency, duration, and magnitude of
Pb exposures that contribute to adverse cognitive outcome during adulthood are yet to be fully
understood.

Strong evidence for cognitive function declines associated with cumulative or past Pb exposures
was provided by prospective cohort studies that demonstrated increased bone Pb levels (tibia mean: 10.5,
21.6 jxg/g, patella mean: 12.6, 30.6 jxg/g) measured at baseline were associated with cognitive decline
over the follow-up period of 13-15 years (Farooqui etal.. 2017; Power et al.. 2014). Findings from these
studies suggest that long-term Pb exposure may contribute to ongoing declines in cognitive function in
adults. These associations remained significant even after adjustment for potential confounding by
combinations of factors including demographic, socioeconomic, behavioral, clinical, and neighborhood
level factors. Potential associations observed in the longitudinal studies were also supported by a cross-
sectional study that suggested increased bone Pb level (tibia mean range: 4.4-9.2 jxg/g) was associated
with cognitive function outcomes among cases of PD ("Weuve et al.. 2013). Additional support for the
effects of cumulative or past Pb exposure is provided by an analysis of past blood Pb exposures during
childhood (either low or high Pb exposure scenarios; blood Pb mean: 3.4 (ig/dL at 7-12 years,
10.99 (ig/dL at 11 years) and studies that followed study participants through young or mid-adulthood
(Reuben et al.. 2020: Reuben et al.. 2017): SkerfVing et al. (2015). These studies indicated that higher
childhood BLL was associated with declines in IQ at 18 to 19 years old and at 38 years or 45 years old.

Findings from cross-sectional studies that assessed the relationships of concurrent blood
biomarkers and cognitive function outcomes were more mixed. An important point to note is that the
mean BLLs reported in these recent cross-sectional studies are at the low end (2.1 (ig/dL to 5.1 (ig/dL).
Two studies including NHANES data suggested significant inverse association between concurrent BLLs
and cognitive outcomes (Sasaki and Carpenter. 2022: Przvbvla et al.. 2017). while others suggested null
associations (Xiao et al.. 2021: Souza-Talarico et al.. 2017: Khalil et al.. 2014: van Wiingaarden et al..
2011). The NHANES studies demonstrating significant associations considered multiple metals in their
analytical models including Pb, used advanced model approaches to handle multiple exposures and issues
around multiple comparison and multi-collinearity, and adjusted for sociodemographic, behavioral, and
clinical characteristics. These approaches reduced the bias and uncertainty in the study findings.

However, one caveat to the interpretation of the study findings from the cross-sectional studies including
concurrent blood Pb biomarkers is that, given bone Pb is a major contributor of BLLs, the relative
contributions of recent and past Pb exposures on the observed association with cognitive function is
unclear. A recent study using a prospective design addressed some of the concern around the health

External Review Draft

3-186

DRAFT: Do not cite or quote

-------
effects of recent Pb exposure (Yu et al.. 2021). The study included a group of individuals with no prior
occupational exposure and recently hired young workers at battery manufacturing and Pb recycling
plants. The association between neurocognitive performance and blood Pb was examined prior to and up
to 2 years after the first occupational exposure (Geo mean baseline: 3.97 (ig/dL; 13.4 (ig/dL, and
12.8 (ig/dL at the first and second follow-up visits). The study did not observe significant associations of
changes in neurocognitive function in the workers with an over three-fold increase in blood Pb
concentration over the 2-year follow -up period. This may be because the time from the first occupational
exposure was too short for the increased Pb levels to have neurocognitive effects.

Despite the large body of recent epidemiologic evidence, particularly from longitudinal studies,
some uncertainties remain. Sex (male versus female, premenopause versus postmenopause for female)
and age (young versus mid-aged versus old-aged adults) differences in bone kinetics and turnover, as well
as disease comorbidity particularly at middle- and older-adulthood lifestages may potentially lead to
differences in tibia, patella, and blood Pb. This adds complexity to modeling potential associations
between Pb biomarkers and cognitive effects in adulthood, since the inclusion of age and sex in
regression models may not fully account for these differences. Potential sex and age differences in bone
kinetics and turnover could have contributed to differences in the magnitude of associations observed for
different bone Pb biomarkers and cognitive function in the NHS and NAS cohorts. Specifically, the role
of specific bone biomarkers (i.e., tibia or patella Pb) as indicators of Pb exposure for specific age and sex
groups in relation to individual cognitive domains is yet to be fully understood. Relatedly, an important
uncertainty in the association between Pb exposure and cognitive function in adults is the observed
heterogeneity of associations across cognitive domains and specific bone Pb biomarkers (i.e., within and
between bone Pb biomarker variability associated with specific cognitive domains). For instance, the
findings from the NHS cohort with shorter follow-up Weuve et al. (2009) suggested that higher tibia Pb
in women was inversely associated with the overall cognitive score. The associations with the majority of
the domain-specific cognitive scores were also negative (except for letter fluency, which was positive) but
the estimates were imprecise. The recent extended analysis of the NHS cohortPower et al. (2014). on the
other hand, suggested that higher tibia Pb in women was inversely associated with individual cognitive
scores representing executive function and memory domains, and was unexpectedly positively associated
with immediate verbal memory domain. Associations with other cognitive domains or the overall
cognitive score were imprecise. Similarly, analyses from the NAS men cohort with shorter versus
extended follow-up periods indicated heterogeneous associations with global and domain scores as well.
Results from the analysis with shorter follow-up(Weisskopf et al.. 2007) suggested stronger inverse
association between higher patella Pb and declines in the visuospatial and visuomotor domains over time,
but weaker and imprecise associations were observed for other domains. In contrast, results from the
extended analvsisFarooqui etal. (2017) observed that higher patella Pb was associated with faster
longitudinal decline in MMSE (a measure of global cognition) and declines in the language and memory
domains, whereas associations with other cognitive scores or domains were imprecise. It is important to
realize that the approach for grouping the cognitive test scores to form specific cognitive domains or
overall score was slightly different between the studies, making it difficult to directly compare the

External Review Draft

3-187

DRAFT: Do not cite or quote

-------
magnitude and direction of the observed associations. There is limited evidence at present to elucidate the
heterogeneity in the observed associations of bone Pb biomarkers with cognitive function domains.

Four recent studies of rodents with exposure resulting in mean BLLs <30 (ig/dL add to the
evidence informing the association of both short- and long-term Pb exposure during adulthood with
measures of learning and memory in rodents. While these studies are consistent with one another,
toxicological evidence for effects of Pb on cognitive function in adults remains limited., A few recent
studies in juvenile rodents also provide some support for the association between postnatal Pb exposure
either during adolescence or continuing into adulthood and cognitive impairment, specifically learning
and memory. Older studies in nonhuman primates demonstrated that early life exposure to Pb may
produce cognitive impairment in adulthood. Hence, these findings add to the current evidence base
suggesting potential roles of both early and later life Pb exposures to produce cognitive function
decrements in adults. Additionally, animal and in vitro studies lend biological plausibility to the
association between adult Pb exposure and adult cognitive impairment, showing that Pb has negative
effects on neuronal function and integrity, neurotransmission, and synaptic plasticity in regions of the
brain associated with learning and memory (Section 3.6).

In summary, recent prospective cohort studies add to the previous body of evidence informing the
inverse association between bone or BLL and cognitive performance and decline in adults without
occupational Pb exposure. These studies provided better insights into the role of both early and later life
as critical lifestages for exposure-outcome relationships. The prospective design with longer follow-up
periods, multiple and repeatedly measured cognitive outcomes, and multiple risk factors and confounders
accounted for in these recent epidemiologic studies reduce uncertainty particularly on the timing or
lifestage of exposure and strengthen the overall evidence for an association between Pb exposure and
decreased cognitive function in adulthood. However, some uncertainties remain. Sex and age differences
in bone kinetics and turnover may contribute to differences in biomarker Pb levels and add uncertainty in
modeling. Variability within and between bone Pb biomarker associated with specific cognitive domains
also adds to uncertainty in the observed association. Recent evidence from animal studies supports the
biological plausibility for the effects of Pb exposure on cognitive function in adults and the notion that
postnatal exposure to Pb (either during adolescence or continuing into adulthood) may also negatively
affect learning and memory. Overall, the collective evidence is sufficient to conclude that there is
likely to be a causal relationship between Pb exposures and cognitive effects in adults.

External Review Draft

3-188

DRAFT: Do not cite or quote

-------
Table 3-9 Summary of evidence for a likely causal relationship between Pb exposure and cognitive effects in
adults.

Pb Biomarker

Rationale for Causality Determination3	Key Evidence13	References'3 Levels Associated

with Effects0

Consistent findings from prospective epidemiologic
studies with relevant adult bone Pb levels or early
childhood BLLs suggesting significant association
of long-term and early childhood Pb exposure with
cognitive impairment and decline

Prospective analyses in NAS cohort of white men and NHS
cohort of white women found cognitive function decrements
over the 13 to 15 yr follow-up in association with patella or
tibia Pb levels.

Prospective analyses of childhood Pb exposure and long-term
cognitive impairments suggested persistent effects on
cognition during adulthood by showing lower IQ at young- and
mid-adulthood periods, and significant decline in IQ between
childhood and adulthood periods due to exposure to higher
childhood Pb levels.

(Power et al.. 2014)

(Farooaui et al..
2017)

(Skerfvinq et al..
2015)

(Reuben et al..
2017)

(Reuben et al..
2020)

Mean patella Pb:
12.6 |jg/g, 30.6 |jg/g
mean tibia Pb:
10.5 |jg/g, 21.6 pg/g

Mean childhood (7-
12 yr) blood Pb:
3.4 [jg/dL,
10.99 [jg/dL

Models adjusted for various confounding factors including
baseline individual-level, socioeconomic, demographics,
behavioral, and clinical factors, as well as various
neighborhood level variables. The early childhood Pb
exposure studies also adjusted for parental education, HOME
scores, parent IQ and childhood IQ.

External Review Draft

3-189

DRAFT: Do not cite or quote

-------
Rationale for Causality Determination3

Key Evidence13

Pb Biomarker
References'3 Levels Associated
with Effects0

Supporting cross-sectional epidemiologic studies

Cross-sectional analysis of bone Pb and cognitive function
among cases ad controls of PD found lower cognitive
performance score with increased tibia Pb among cases.

Cross-sectional analysis of concurrent blood Pb (along with
multi-metals) and cognitive function using NHANES data and
advanced modeling approach suggested significant inverse
association between concurrent BLL and cognitive outcomes.

Models adjusted for socioeconomic and demographic factors,
education, health status, comorbidities, and co-exposure to
other metals.

Weuve, 2013,
6718818

(Przvbvla et al..
2017)

(Sasaki and
Carpenter. 2022)

Mean tibia Pb: 4.4-
9.2 |jg/g (for age
groups)

Mean patella Pb:
5.9-15.2 |jg/g (for
age groups)

Geometric mean
blood Pb: 2.17 pg/dL

Mean blood Pb:
1.9 pg/dL

Consistent evidence in animals with relevant
exposures

Recent evidence from animal studies supports the notion that (Mansouri et al.

postnatal exposure to Pb (either during adolescence or
continuing into adulthood) negatively affects learning
and memory in rodents.

2012)

(Mansouri et al..

2013)

(Singh et al.. 2019)
(Su et al.. 2016)

Mean BLL: 8 pg/dL

Peak BLL: 11-
19 pg/dL

Peak BLL: 28 pg/dL
Mean BLL: 8.4 pg/dL

Some uncertainty remains

Sex and age differences in bone kinetics and turnover may
contribute differences in biomarker Pb levels and adds
uncertainty in modeling. In addition, within and between bone
Pb biomarker variability associated with specific cognitive
domains also adds to uncertainty in the observed association.

BLL = blood lead level; NAS = Normative Aging Study; NHANES = National Health and Nutrition Examination Survey; Pb = lead; PD = Parkinson's disease.

aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Table I and Table II of the Preamble to the ISAs (U.S. EPA. 2015).

bDescribes the key evidence and references, supporting or contradicting, contributing most heavily to causality determination and, where applicable, to uncertainties or
inconsistencies. References to earlier sections indicate where the full body of evidence is described.

°Describes the Pb biomarker levels at which the evidence is substantiated.

External Review Draft

3-190

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

3.6.2

Psychopathological Effects in Adults

The evidence assessed in the 2013 Pb ISA was sufficient to conclude that "a causal relationship is
likely to exist" between Pb exposure and psychopathological effects in adults. Cross-sectional studies in a
small number of distinct U.S. populations demonstrated associations of higher concurrent blood or tibia
Pb levels with self-reported symptoms of depression and anxiety in adults (Bouchard et al.. 2009; Raian
et al.. 2007; U.S. EPA. 2006b). The examination of multiple exposures and outcomes in the available
studies does not provide a strong indication of biased reporting of psychopathological effects specifically
by adults with higher Pb exposures. In adults, Pb-associated increases in depression and anxiety were
found with adjustments for age, SES, and in the NAS, daily alcohol intake. The biological plausibility for
epidemiologic evidence was provided by observations of depression-like behavior in animals with dietary
lactational Pb exposure, with some evidence at relevant BLLs. In addition, Pb-induced changes in the
HPA axis and dopaminergic and GABAergic systems were demonstrated in animals. Overall, the
strongest evidence was from epidemiologic studies of adults without occupational Pb exposure, with
additional support from a few experimental animal studies; however, uncertainties related to residual
confounding of bone Pb associations by age in epidemiologic studies remained. Overall, recent studies
add to the evidence generally supporting the findings from the 2013 Pb ISA.

3.6.2.1 Epidemiologic Studies of Psychopathological Effects in Adults

A limited number of epidemiologic studies evaluated in the 2006 Pb AQCD (U.S. EPA. 2006c')
and 2013 Pb ISA (U.S. EPA. 2013a) examined the relationship between blood or bone Pb levels and
psychopathological effects in adults. All of these studies were cross-sectional and most examined
occupationally exposed populations. In addition to the occupational studies, which provided consistent
evidence of positive associations between BLLs (mean levels >15 (ig/dL) and the prevalence of self-
reported symptoms of depression, anxiety, and tension, a prospective analysis of the NAS cohort (Raian
et al.. 2007) and a cross-sectional analysis of NHANES participants (Bouchard et al.. 2009) reported
positive associations between much lower concentrations of concurrent blood (NHANES; mean ~
6 (ig/dL) or bone (NAS) Pb levels and symptoms of depression and anxiety or prevalent major depressive
disorder, respectively.

Recent evidence includes a number of prospective cohort studies and cross-sectional analyses of
Pb exposure and general psychopathological effects or internalizing symptoms, as well as a case-control
study of schizophrenia. In general, recent prospective studies provide evidence of an association between
blood or bone Pb levels and psychopathological effects in adults. Results from cross-sectional studies are
inconsistent. Additionally, with blood or bone Pb levels, it is difficult to characterize the specific timing,
duration, frequency, and level of Pb exposure that contributed to associations observed with cognitive
function. This uncertainty may apply particularly to assessments of BLLs, which in nonoccupationally

External Review Draft

3-191

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

exposed adults reflect both current exposures and cumulative Pb stores in bone that are mobilized during
bone remodeling. Measures of central tendency for blood and bone Pb levels used in each study, along
with other study-specific details, including study population characteristics and select effect estimates, are
highlighted in Table 3-15E. An overview of the recent evidence is provided below.

A few recent prospective cohort studies report generally consistent evidence of associations
between blood or bone Pb levels and psychopathological effects in adults, although the results for specific
endpoints are not entirely consistent. In a recent prospective cohort study examining a subset of the NAS
cohort, Peters et al. (2011) used structural equation models to examine the interrelations of childhood and
adult SES, bone Pb levels, pessimism, and depression in older adults. After controlling for childhood and
adult SES through a combination of latent variables—including parental and participant education,
occupation, and home ownership—a 10 jxg/g increase in tibia Pb levels measured prior to psychological
measurement was independently associated with a 0.3-unit increase (95% CI: 0.0, 0.6) in pessimism level
on the Life Orientation Test. An independent association between bone Pb levels and depression was not
observed when controlling for pessimism (quantitative results not reported), but pessimism was strongly
associated with increased odds of depression (OR = 1.04 [95% CI: 1.02, 1.05] per 1 unit increase in
pessimism level), indicating a potential mediating effect of pessimism on the relationship between bone
Pb levels and depression. Other recent prospective cohort studies examined the relationship between
childhood BLLs and a wider range of psychopathological effects (Reuben et al.. 2019; McFarlane et al..
2013). In an analysis of the Dunedin cohort in New Zealand, Reuben et al. (2019) reported that each 1-
(ig/dL increase in BLLs at age 11 was associated with a 0.27-point increase (95% CI: 0.02, 0.51) in
standardized general psychopathology scores (mean [SD]: 100 [15]) in early adulthood. In more specific
analyses of psychopathology components, the association with general psychopathology appeared to be
driven by positive associations with symptoms of internalizing behavior and thought disorder. These
results are somewhat consistent with a follow-up analysis of the Port Pirie cohort, a birth cohort from a
South Australian Pb-smelting town (McFarlane et al.. 2013). In this study, BLLs averaged over the first
7 years of life were not associated with depressive symptoms measured during follow-up at ages 25
through 29. However, the authors did report positive associations between BLLs and some internalizing
behaviors in women (e.g., social phobia, specific phobia, and anxiety problems). There were generally
null associations for the same outcomes in men. A notable limitation of this study is that attrition in the
original cohort led to a small sample size that was even further reduced by sex-stratified models, resulting
in limited power to detect an association. Additionally, due to high community exposure to Pb, the
participants in the Reuben et al. (2019) and McFarlane et al. (2013) studies had a high mean childhood
BLL (11.08 and 17.2 (ig/dL, respectively). The mean BLLs in these studies are not directly comparable to
other studies in this section, which used concurrent BLLs in adult populations that likely had higher past
exposures.

Associations between BLLs and mood states were inconsistently observed in cross-sectional
studies of pregnant women in China (Li et al.. 2017) and Japan (Ishitsuka et al.. 2020). In a small study of
pregnant women in Shanghai, Li et al. (2017) observed nonlinear associations between BLLs and

External Review Draft

3-192

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

depression, anxiety, and psychological distress in nonparametric models. Based on visual inspection of
the spline curves, the authors ran separate piecewise linear regression models for each outcome with a
knot at 2.57 (ig/dL. In the piecewise models, BLLs were associated with increased depression, anxiety,
and psychological stress at levels below 2.57 (ig/dL, whereas smaller inverse associations were observed
above the knot. In contrast, a much larger cross-sectional study of pregnant women across Japan noted
that BLLs measured during middle or late pregnancy were not associated with increased odds of Kessler
Psychological Distress Scale (K6) scores greater than or equal to 5 or 13 (Ishitsuka et al.. 2020). The
authors used different cut points to account for potential differences in the optimal sensitivity-specificity
tradeoff for assessing depression in their study population. The contrasting results in these studies are not
readily explained by variations in BLLs, as Li et al. (2017) observed positive associations at low BLLs in
linear spline models, and the population analyzed by Ishitsuka et al. (2020) had a geometric mean BLL
<1 (ig/dL.

Other recent cross-sectional studies assessed the relationship between blood or bone Pb levels and
mood states measured by validated questionnaires or self-reported physician's diagnosis in a range of
study populations, including a population-based analyses ofNHANES (Berk et al.. 2014) and KNHANES
participants (Nguyen et al.. 2022). older women participating in subcohorts of the NHS (Eum et al..
2012). and older adults from two communities selected using cluster-based sampling of communities in
Luan, China (Fan et al.. 2020). Similar to previously discussed studies, results from these analyses were
inconsistent. Fan et al. (2020) reported monotonic increases in odds of depression associated with
increasing blood Pb exposure quartiles. In this study, older adults in the highest quartile of exposure
(BLLs <3.06 (ig/dL) had just over twice the odds of depression as those with BLLs <2.03 (ig/dL
(OR = 2.03 [95% CI: 1.23, 3.35]). In contrast, Eum et al. (2012). Berk et al. (2014). and Nguyen et al.
(2022) reported null associations between bone or blood Pb levels and anxiety or depression. However, in
a subgroup analysis restricting the study population to pre- and postmenopausal women taking hormone
replacement therapy (HRT), Eum et al. (2012) reported that increasing tibia Pb tertiles were associated
with monotonically increasing odds of phobic anxiety and lower scores on the Mental Health Index 5-
item (MHI-5; indicating worse depressive symptoms). The authors restricted the analysis by HRT status
to account for potential exposure measurement error in the non-HRT population, resulting from higher
variability in bone turnover. Notably, patella Pb was not associated with depressive symptoms in this
population and was inversely associated with phobic anxiety. Since tibia Pb has a longer half-life than
patella Pb, the results could be indicative of a long-term exposure window contributing to changes in
anxiety and depression. However, the subgroup analyses also had small sample sizes and thus a lack of
precision and higher probability of chance findings.

While most recent studies of psychopathological effects in adults examined internalizing
behaviors or general psychopathological effects, a small case-control study in China evaluated serum
heavy metal levels in association with the risk of schizophrenia (Ma etal.. 2019). The authors reported
that increases in serum Pb levels were associated with a large, but imprecise increase in the odds of
schizophrenia (OR = 3.15 [95% CI: 1.24, 7.99] per ng/mL increase in BLL). Although the cases and

External Review Draft

3-193

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

controls were matched on age and sex, it is unclear what, if any, other potential confounders were
included in the adjusted models. This finding was ostensibly consistent with a pooled analysis of two
small cohorts evaluated in the 2013 Pb ISA that observed an association between higher S-ALAD levels
and increased odds of schizophrenia spectrum disorder in adolescents and adults (Opler et al.. 2008).
However, this pooled analysis had a number of limitations that precluded any conclusions regarding a
relationship between Pb exposure and schizophrenia, including the lack of direct measurements of Pb
biomarker levels and limited consideration for potential confounding. As noted previously, Reuben et al.
(2019) reported a positive association between childhood BLLs and symptoms of thought disorder in
early adulthood. However, the metric for thought disorders included factor loadings for obsessive
compulsive disorder and mania in addition to schizophrenia, making it difficult to distinguish an
independent relationship between BLLs and schizophrenia.

3.6.2.1.1 Summary

A limited number of cross-sectional studies evaluated in the 2013 Pb ISA (U.S. EPA. 2013a)
provided consistent evidence of positive associations between blood and bone Pb levels and the
prevalence of self-reported symptoms of depression, anxiety, and tension. Recent prospective analyses
provide additional support for a positive association between bone and BLLs and psychopathological
effects in older adults, although results from cross-sectional studies are inconsistent.

3.6.2.2 Toxicological Studies of Psychopathological Effects in Adults

No studies in the 2013 Pb ISA evaluated the effect of adult-only exposure on anxiety and
depression-like behaviors in animals. Nevertheless, developmental studies consistently supported an
effect of Pb exposure on these endpoints in adult animals (U.S. EPA. 2013a). Of particular importance,
the increased reactivity to errors and reward omission reported by Beaudin et al. (2007) and Stangle et al.
(2007) extended to adulthood, well after postnatal exposure was terminated. Furthermore, Pb exposure
during adolescence reduced immobility on the FST in adult rats (Stewart et al.. 1996).

A few recent studies have investigated adult-only exposure and anxiety-like outcomes using the
OFT and EPM. Similar to developmental exposures, adult male mice displayed anxiety-like behavior (i.e.,
increased time spent in closed arms) in the EPM following 6 weeks of Pb exposure given via oral gavage
(mean BLLs 7.1 (ig/dL) (Al-Qahtani et al.. 2022). Singh et al. (2019) also found that oral gavage of Pb for
90 days decreased the amount of time rodents spent in the open arms of the EPM compared with control
animals. Another study of long-term adult exposure (126 days) to Pb in Wistar rats found increased
rearing and grooming, but not sniffing, in males, and no significant effects in females in the OFT
(Mansouri et al.. 2013). The same research group found that Pb did not affect these endpoints (i.e.,
rearing, sniffing) following shorter-term exposure (30 days) in adult animals (Mansouri et al.. 2012).

External Review Draft

3-194

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

Studies that employed developmental exposure paradigms are discussed in more detail in
Section 3.5.4.2. However, they provide consistent evidence to support the notion that the effects of
developmental exposures may persist into adulthood. In one study of gestational exposure, there were
significant decreases in exploratory behaviors in the OFT and hole board test in Wistar rats at 4 months
old (BLLs peaked at 12 (ig/dL and decreased to 6 (ig/dL by 4 months) (Basha and Reddv. 2015). A
similar study utilizing postnatally exposed rats found that some measures of decreased exploratory
behavior in the OFT and hole board test persisted until 18 months, when the study was terminated (Basha
ct al.. 2014). The mean BLL at PND 45 in this study was high (50 (ig/dL) but had decreased to 11 (ig/dL
by 18 months. Three additional studies investigated lifetime Pb exposure (gestation through 6-
12 months), and all found significant treatment effects on EPM or FST behavior during adulthood
(Shvachiv et al.. 2020. 2018; Abazvan et al.. 2014; Corv-Slechta et al.. 2013).

Toxicological studies also provide biological plausibility to support a connection between
exposure to Pb and schizophrenia. As discussed in the 2013 Pb ISA, antagonists of NMDAR's glycine
site have been shown to exacerbate schizophrenia symptoms in affected individuals and induce a
schizophrenic phenotype in unaffected subjects (Covle and Tsai. 2004). Previous studies have shown that
Pb is a potent allosteric inhibitor at NMDARs (Hashcmzadch-Gargari and Guilarte. 1999; Guilarte. 1997).
A recent study found that developmental Pb exposure (BLLs of 22 (ig/dL at PND 50) reduced the number
of parvalbumin-positive GABAergic interneurons in the median prefrontal cortex and hippocampus and
induced subcortical dopaminergic hyperactivity, consistent with studies of schizophrenic patients
(Stansfield et al.. 2015; Volman etal.. 2011). Pb exposure may also affect DISCI, a gene-protein pair
associated with increased susceptibility to schizophrenia and other mental disorders. You et al. (2012)
found that expression of the DISCI protein was increased in the hippocampus of rats exposed to Pb
during gestation and lactation. Abazvan et al. (2014) utilized a transgenic mouse model expressing mutant
DISCI (mDISCl) to evaluate a potential gene-environment interaction using lifetime Pb exposure. Pb-
exposed mDISCl mice exhibited behavioral and structural abnormalities consistent with schizophrenia,
which were not found in unexposed mDISCl mice or Pb exposure regular mice (i.e., heterozygous for
mDISCl but phenotypically normal).

3.6.2.2.1 Summary

Toxicological studies providing support for adult psychopathological effects in the previous ISA
used developmental exposure paradigms. Recent developmental exposure studies were consistent with the
previous evidence and were predominantly focused on anxiety-like behaviors. Multiple studies
demonstrated the persistence of these effects into adulthood (up to 1.5 years), in some cases long after
termination of Pb exposure .A few recent studies focused on adult-only exposures found some
associations with anxiety-like behavior after 42-126 days of exposure but not following a 30-day
exposure; additional studies are needed to strengthen this line of evidence. Two recent studies also

External Review Draft

3-195

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

provided further biological plausibility support for an NMDAR-mediated association between Pb
exposure and schizophrenia.

3.6.2.3 Relevant Issues for Interpreting the Evidence Base

3.6.2.3.1	Concentration-Response Function

One recent study used a nonparametric model to assess the C-R relationship between BLLs and
depressive symptoms in pregnant women (Li et al.. 2017). The authors observed nonlinear associations
between BLLs and depression, anxiety, and psychological distress scores and used the nonparametric
models to determine knots for a piecewise linear regression model. The subsequent models indicated
positive associations between BLLs and depression, anxiety, and psychological stress at levels below
2.57 (ig/dL, whereas smaller inverse associations were observed for BLLs above 2.57 (ig/dL. Other cross-
sectional studies reported inconsistent evidence of associations despite evaluating populations with low
mean BLLs. However, in studies analyzing BLLs in adult populations with higher past exposures, it is a
challenge to ascertain the level, timing, frequency, and duration of Pb exposure that contributed to
observed associations.

3.6.2.3.2	Potentially At-Risk Populations

A few of the recent epidemiologic studies detailed in this section evaluated populations that are
potentially at-risk for Pb-related health effects. The conclusions that can be drawn from these analyses are
limited. A prospective analysis of young adults reported sex-specific associations between childhood
BLLs and internalizing symptoms in early adulthood (McFarlanc et al.. 2013). The observed associations,
which were only present in stratified models including women, were extremely imprecise due to a small
sample size that was even further reduced by stratification. The small sample size in this study reduced
the statistical power to detect an association and the likelihood that an observed result reflects a true
effect, making it difficult to draw firm conclusions on these sex-specific comparisons. Additionally, two
cross-sectional studies examining depressive symptoms in pregnant women observed inconsistent
evidence of an association with BLLs (Ishitsuka et al.. 2020; Li et al.. 2017).

3.6.2.3.3	Confounding

The studies evaluated in this section controlled for a range of potential confounding variables that
may be associated with both Pb exposure and psychopathological effects, including age, sex, SES factors,
and marital status (see Table 3-15E). One potential confounder that was not considered is the use of
antidepressants. The use of medication to treat depression could be a particularly important confounder in

External Review Draft

3-196

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

studies that evaluate depressive symptoms. A lack of adjustment for antidepressant use in those studies is
likely to bias effect estimates towards the null.

3.6.2.3.4 Lifestages

Toxicological studies provide consistent evidence that developmental and lifetime exposures to
Pb can lead to increases in anxiety-like behaviors. Comparatively fewer studies have investigated the
effects of adult-only exposure, but some effects have been demonstrated. Epidemiologic studies provide
some supporting evidence for the importance of developmental and cumulative exposures. In a recent
prospective cohort study, Peters et al. (2011) reported that increased depressive symptoms in older adults
were associated with tibia Pb levels, a measure of cumulative exposure. Other recent prospective analyses
observed associations between childhood BLLs and increased internalizing symptoms in young adults
(Reuben et al.. 2019; McFarlane et al.. 2013). Additionally, (Reuben et al.. 2019) also observed positive
associations between childhood BLLs and internalizing symptoms at the time of BLL testing, which is
coherent with toxicological evidence that suggests the persistence of developmental effects into
adulthood. Given the uncertainties regarding potentially higher historical exposures in adults, the cross-
sectional epidemiologic studies evaluated in this section are less suited to address the importance of
concurrent exposures.

3.6.2.4 Integrated Summary and Causality Determination: Psychopathological Effects
in Adults

The 2013 Pb ISA (U.S. EPA. 2013a) concluded that the available evidence was sufficient to
conclude that "a causal relationship is likely to exist" between Pb exposure and psychopathological
effects in adults. This causality determination was based on a small body of epidemiologic evidence that
demonstrated consistent positive associations between concurrent blood or bone Pb levels and self-
reported symptoms of depression, anxiety, and panic disorder in large studies of adults (i.e., NHANES,
NAS). The epidemiologic evidence was supported by coherence in animal toxicological studies that
demonstrated depression-like behavior and emotionality in rodents exposed to dietary lactational Pb with
or without additional post-lactational exposure. Epidemiologic associations were observed in study
populations of young (20-39 years old) and older (44-98 years old) adults. Because of the cross-sectional
design of the epidemiologic studies, there was uncertainty regarding the temporal sequence between Pb
exposure and psychopathological symptoms in adults. This uncertainty is somewhat reduced with results
for tibia Pb since it is an indicator of cumulative Pb exposure. Nonetheless, because these studies included
adults with likely higher past Pb exposures, uncertainties exist regarding the Pb exposure level, timing,
frequency, and duration contributing to the associations observed with blood or bone Pb levels. An
uncertainty in the toxicological evidence base was the limited number of studies that administered
exposures resulting in BLLs that are not relevant to humans. Recent epidemiologic and toxicological

External Review Draft

3-197

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

evidence continues to link Pb exposure to psychopathological effects in adults, though some uncertainties
still remain. The key evidence, as it relates to the causality determination, is presented in Table 3-10 and
Table 3-11 and summarized below.

Recent evidence from prospective epidemiologic studies provides further support for positive
associations between Pb exposures and pathological effects, including increased internalizing symptoms.
Specifically, a study of older men in the NAS cohort provides evidence of an (indirect) association
between Pb exposure and depression in adults (Peters etal.. 2011). and an analysis of a cohort in New
Zealand similarly reported that increased childhood BLLs were associated with increased internalizing
symptoms in young adults (Reuben et al.. 2019). Together, these studies address an uncertainty from the
previous ISA regarding the temporality of the exposure and outcome. Another recent prospective study of
young adults from the Port Pirie cohort reported null associations between childhood BLLs and
depression in adults but noted several sex-specific associations between BLLs and other internalizing
symptoms in young women (McFarlane et al.. 2013). The small analytic sample of the stratified models
used in this analysis reduces the likelihood of detecting a true effect and decreases confidence that
observed associations represent a true effect. Notably, supporting evidence from recent cross-sectional
epidemiologic studies conducted in diverse populations is largely inconsistent. However, these studies, as
well as the recent prospective analyses described previously, do not adjust for potential confounding by
antidepressant medication. Given the existing evidence for associations between Pb exposure and
psychopathological effects, the use of antidepressants could be associated with both higher BLLs and
reduced symptoms of depression. Therefore, a lack of adjustment for antidepressant use in studies that
evaluate self-reported depressive symptoms could bias effect estimates towards the null.

The epidemiologic evidence is supported by coherence with results from an expanded number of
toxicological studies conducted at BLLs relevant to humans. Recent toxicological studies examine
multiple exposure windows and provide strong support for Pb-induced anxiety-like behaviors following
developmental and cumulative exposures. Multiple studies demonstrate the persistence of these effects
into adulthood (up to 1.5 years), in some cases long after termination of Pb exposure. The evidence for
effects resulting from adult-only exposures is more limited, though there is some evidence for an increase
in anxiety-like behavior following 42-126 days of exposure but not following a 30-day exposure. The
2013 Pb ISA also highlighted Pb-induced changes in the dopaminergic and GABAergic systems and the
HPA axis, which underlie biological plausibility for the changes in mood and emotional state that have
been observed in epidemiologic and toxicological studies (U.S. EPA. 2013a). Recent studies continue to
demonstrate changes in corticosterone and glucocorticoid receptors (i.e., HPA axis changes (Corv-Slcchta
et al.. 2012)) and the dopaminergic system (Section 3.4.2.2).

In addition to studies of depression, anxiety, and mood-related disorders, some recent studies
examined the relationship between Pb exposure and schizophrenia. A recent case-control study in China
reported a positive, but imprecise association between serum Pb levels and schizophrenia prevalence in
adults (Ma et al.. 2019). Because serum Pb was measured after schizophrenia was diagnosed, the results

External Review Draft

3-198

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

do not establish temporality between exposure and outcome. Additionally, the reported analytic
methodology does clarify the confounding variables considered outside of the age- and sex-based
matching of cases and controls. Recent toxicological studies provide biological plausibility to support a
connection between exposure to Pb and schizophrenia. Consistent with evidence from the 2013 Pb ISA,
two recent studies support Pb-induced pathophysiological features in rodents consistent with
schizophrenia, likely through inhibition of NMDAR activity. Although the toxicological evidence
presents a biologically plausible pathway through which exposure to Pb could lead to schizophrenia, the
limited quantity and quality of the epidemiologic evidence precludes meaningful consideration of this
endpoint in the causality determination for Pb exposure and psychopathological effects.

In summary, a limited number of recent prospective epidemiologic studies add to previous
evidence of a positive association between bone or BLLs and psychopathological effects in older adults
and addresses prior uncertainties regarding the temporality of exposure and outcome. Recent
toxicological studies strengthen the overall evidence base, providing further support for anxiety-like
behaviors following developmental and cumulative exposures that result in BLLs that are relevant to
humans. Recent cross-sectional epidemiologic evidence is less consistent than results from studies
evaluated in the 2013 Pb ISA (U.S. EPA. 2013a'). and there is potential for unmeasured confounding by
antidepressant use. Overall, the collective evidence is sufficient to conclude that there is likely to be a
causal relationship between Pb exposure and psychopathological effects in adults.

External Review Draft

3-199

DRAFT: Do not cite or quote

-------
Table 3-10 Summary of evidence for a likely to be causal relationship between Pb exposure and
psychopathological effects in adults.

Rationale for Causality
Determination3

Pb Biomarker Levels
Key Evidence13	References'3 Associated with

Effects0

Limited epidemiologic evidence from Prospective analyses reported positive associations between tibia Pb levels
high-quality prospective cohort studies and depressive symptoms in older men, mediated by pessimism; and
with relevant bone Pb levels	childhood BLLs and psychopathology scores in early adulthood.

Some supporting evidence from a prospective analysis reported imprecise
positive associations between childhood BLLs and prevalence of social
phobia, specific phobia, PTSD, anxiety problems, somatic problems, and
antisocial personality problems in young adult women.

Consistent evidence in animals with Increased anxiety-like behaviors in adulthood (up to 18 mo) following	Section 3.5.4.2 Peak blood Pb after

relevant exposures	developmental or lifetime Pb exposure provides coherence with epidemiologic	lifetime exposure: 7-

evidence.	24 [jg/dL

Peak blood Pb after
developmental
exposure: 12-50 [jg/dL

Section 3.6.2.1 Mean/median range
across studies:

Bone: 10.3-12.5 |jg/dL
Blood: 0.58-3.97 pg/dL

Uncertainty regarding potential	Most studies included adjustment for age, sex, SES factors, and marital

confounding	status, Table 3-15E but did not consider use of antidepressants

Peters et al. Mean: 20.6 pg/g
120111	Mean: 11.08 pg/dL

Reuben et al.

(2019)

McFarlane et al. Mean: 17.2 pg/dL
(2013)

Inconsistent supporting evidence from A limited body of cross-sectional studies provides inconsistent evidence of
cross-sectional epidemiologic studies associations between generally lower blood and bone Pb levels and
with relevant blood and bone Pb levels depression and anxiety.

BLL = blood lead level; mo = months; Pb = lead; PTSD = post-traumatic stress disorder; SES = socioeconomic status.

aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Table I and Table II of the Preamble to the ISAs (U.S. EPA. 2015).
bDescribes the key evidence and references, supporting or contradicting, contributing most heavily to causality determination and, where applicable, to uncertainties or
inconsistencies. References to earlier sections indicate where the full body of evidence is described.

°Describes the Pb biomarker levels at which the evidence is substantiated.

External Review Draft

3-200

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

3.6.3

Sensory Organ Function in Adults

The 2013 Pb ISA included separate causality conclusions for auditory and visual function. This
ISA combines these categories and makes one causality determination for Sensory Organ Function (see
Section 3.6.3.5). Recent studies are summarized in Table 3-16E and Table 3-16T. An overview of the
recent evidence is provided below.

3.6.3.1 Auditory Function

The evidence assessed in the 2013 Pb ISA was "suggestive of a causal relationship" between Pb
exposure and auditory function decrements in adults (U.S. EPA. 2013a). The strongest evidence was
provided by the analysis of NAS men, which revealed associations between higher tibia Pb levels and a
higher rate of elevations in hearing threshold over 20 years (Park et al.. 2010). Findings demonstrating
decreased auditory evoked potentials in animals provided biological plausibility for the observations in
this epidemiologic study, but uncertainties related to effects on auditory function in adult animals with
relevant Pb exposures remained.

3.6.3.1.1 Epidemiologic Studies of Auditory Function

Several recent epidemiologic studies examined the association between Pb exposure and
decrements in auditory function in adults (Tu et al.. 2021; Yin et al.. 2021; Wang et al.. 2020; Kang et al..
2018; Choi and Park. 2017; Shiue. 2013; Choi et al.. 2012). The findings generally supported an
association between Pb exposure and hearing loss in adults. These studies are described below.

Most studies were cross-sectional, using data from NHANES (Tu et al.. 2021; Shiue. 2013; Choi
et al.. 2012) and KNHANES (Kang et al.. 2018; Choi and Park. 2017). Choi et al. (2012) and Tu et al.
(2021) measured blood Pb in adult NHANES participants (20-69 years) in 1999-2004 and 2011-2012,
respectively. Choi et al. (2012) observed an increased likelihood of hearing loss per doubling of blood Pb
(OR = 1.09 [95% CI: 0.95, 1.26]) as well as an increased percent change in hearing threshold (%
change = 5.41 [95% CI: 2.12, 8.81]). Similar findings were noted in quintile analyses, with higher blood
Pb quintiles having a greater magnitude of effect when compared with the lowest quintile (0.20-
0.80 (ig/dL) (Choi et al.. 2012). Tu et al. (2021) measured hearing loss at speech frequency and at high
frequency. In quartile analyses, the magnitude of effect increased with each blood Pb quartile for each
type of hearing loss (Tu et al.. 2021). Compared with the lowest quartile (<0.07 (.ig/dL). the highest blood
Pb quartile (>0.16 (ig/dL) was positively associated with speech-frequency hearing loss (OR = 1.46 [95%
CI: 0.81, 2.64]) and high-frequency hearing loss (OR= 1.98 [95% CI: 1.27, 3.10]). In sex-stratified
analyses, the direction of effect remained but the magnitude of effect was greater among males. In age-

External Review Draft

3-201

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

stratified analyses, a positive association remained for ages <35 years and 35-52 years for both types of
hearing loss; however, ORs were generally imprecise for the younger age group (<35 years). On the
contrary, there appeared to be an inverse association between BLLs and hearing loss in the oldest age
group (>52 years) (Tu et al.. 2021). Also using NHANES data, Shiue (2013) measured Pb exposure in
urine and did not find an adverse association with self-reported hearing among older adults (>50 years).
For self-reported ear ringing, urinary Pb had a slightly positive association (Shiue. 2013). In KNHANES,
Choi and Park (2017) measured speech- and high-frequency hearing loss in adolescents (12-19 years) and
adults (20-87 years). Hearing loss was defined as pure-tone average >25 dB in adults. For each doubling
of blood Pb, there was an increased likelihood of speech-frequency hearing loss (OR =1.15 [95% CI:
0.94, 1.41]) and high-frequency hearing loss (OR = 1.30 [95% CI: 1.08, 1.57]). In quartile analyses, the
magnitude of effect increased with each blood Pb quartile when compared with the lowest quartile (Choi
and Park. 2017). In another analysis conducted in the KNHANES population, Kang et al. (2018) observed
an association between BLLs and hearing impairment in adults (20-87 years). In quartile analyses, the
magnitude of effect for high-level frequency hearing impairment was greatest in the highest blood Pb
quartile compared with the lowest blood Pb quartile in males (OR = 1.63 [95% CI: 1.16, 2.29]) as well as
in females (OR = 1.50 [1.03-2.20]). For low-frequency hearing impairment, the associations were less
consistent by quartile and more attenuated (Kang et al.. 2018).

In a meta-analysis of studies from Iran, Korea, China, and the United States, Yin et al. (2021)
observed consistent positive associations between Pb exposure and any hearing loss (combined OR per
unit increase in Pb = 1.42 [95% CI: 1.22, 1.67]), low-frequency hearing loss (combined OR = 1.31 [95%
CI: 1.17, 1.47]), and high-frequency hearing loss (combined OR = 1.96 [95% CI: 1.48, 2.60]). When
stratified by age group, the association persisted in adults (>20 years; combined OR per unit increase in
Pb = 1.34 [95% CI: 1.18, 1.52]) (Yin et al.. 2021). Despite these results, there are still some
inconsistencies in the recent literature. In a case-control study of adults in China who participated in a
survey of hearing loss, Wang et al. (2020) did not observe an association with blood Pb before and after
adjusting for workplace noise exposure.

Summary

The strongest evidence described in the 2013 Pb ISA was provided by the analysis of NAS men
for associations of higher tibia Pb level with a higher rate of elevations in hearing threshold over 20 years
(Park et al.. 2010). Several recent cross-sectional analyses of NHANES and KNHANES generally support
an association of Pb exposure (i.e., concurrent BLLs) with hearing loss; however, recent studies are not
entirely consistent.

External Review Draft

3-202

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

3.6.3.1.2 Toxicological Studies of Auditory Function

Recent animal studies on the auditory effects of Pb exposures have investigated exposures
beginning during development (postnatal or adolescent), discussed in Section 3.5.6.1.2. In particular,
Carlson et al. (2018) did not detect any significant changes in BAEP in 4-month-old adult mice with very
low BLLs (3 (ig/dL). The strongest evidence for adult effects of Pb exposure was presented in the
previous ISA (U.S. EPA. 2013a). Lifetime Pb exposure was found to increase hearing thresholds and
latencies in BAEP in adult monkeys (aged 8-13 years) (Rice. 1997; Lilienthal and Winneke. 1996).
Moreover, Laughlin et al. (2009) detected small nonsignificant shifts in auditory threshold in 13-year-old
Rhesus monkeys following gestational or postnatal Pb exposure. However, these effects were
demonstrated at higher BLLs than are relevant to this ISA (33-150 (ig/dL).

3.6.3.2 Visual Function

The evidence pertaining to visual function assessed in the 2013 Pb ISA was limited. A case-
control study found higher Pb in retinal tissue from macular degeneration cases but lacked rigorous
statistical analysis and examination of potential confounding. Studies in adult animals showed differential
effects on ERGs, depending on the timing and concentration of exposure. Because the available
epidemiologic and toxicological evidence was of insufficient quantity, quality, and consistency, the 2013
Pb ISA concluded the "evidence is inadequate to determine that a causal relationship exists between Pb
exposure and visual function decrements in adults."

3.6.3.2.1 Epidemiologic Studies of Visual Function

Only a few epidemiologic studies examined the association between Pb exposure and decrements
in visual function in adults (Paulsen et al.. 2018; Fillion et al.. 2013; Shiue. 2013). Fillion et al. (2013)
measured contrast sensitivity (cpd) and acquired color vision loss (CCI) in adolescents and adults (15-
66 years) in Brazil. Blood Pb exposure was negatively associated with the intermediate spatial frequency
of contrast sensitivity (12 cycles/degree); however, results varied by spatial frequency. For CCI, there was
a slightly positive association with blood Pb, but the effect estimate was imprecise (Fillion et al.. 2013).
In another study of contrast sensitivity, Paulsen et al. (2018) used the Pelli-Robson letter sensitivity chart
and did not observe an association with blood Pb in a U.S.-based cohort (HR for blood Pb >2.06 |ig/L
versus. <2.06 |ig/L = 0.91 [95% CI: 0.69, 1.18]). Visual function has also been measured using self-
reported eyesight. Using NHANES data, Shiue (2013) measured Pb exposure in urine and did not find an
association with self-reported visual impairment among older adults (>50 years).

External Review Draft

3-203

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

Summary

The epidemiologic evidence pertaining to the association of Pb exposure with visual function in
adults remains limited. A small number of recent studies examining contrast sensitivity or acquired color
vision loss found inconsistent results for associations with blood or urine Pb level (Paulsen et al.. 2018;
Fillion et al.. 2013; Shiue. 2013).

3.6.3.2.2 Toxicological Studies of Visual Function

No recent PECOS-relevant studies have evaluated the effects of Pb exposure on visual function.
Section 3.5.6.2.2 summarizes the literature discussed in the 2013 Pb ISA (U.S. EPA. 2013a).

3.6.3.3	Olfactory Function

The 2013 Pb ISA did not assess any evidence on the relationship between Pb exposure and
olfactory function in adults (U.S. EPA. 2013a).

3.6.3.3.1 Epidemiologic Studies of Olfactory Function

In the Heinz Nixdorf Recall Study (HNRS) in Germany, male participants were recruited in
2000-2003 and followed up in 2011-2014 (Casiens et al.. 2018). Casiens et al. (2018) examined the
effect of Pb exposure on odor identification using the Sniffin' sticks odor identification test of 12 odors.
Participants were classified as normosmic (identified >9 odors), hyposmic (identified 7-9 odors), and
functionally anosmic (identified <7 odors). Compared with the lowest BLL at baseline (<5 (.ig/dL). the
highest BLL (>9 (ig/dL) was associated with impaired odor identification (proportional OR = 1.96 [95%
CI: 0.94, 4.11]). Similar results were observed using BLLs measured at follow-up (proportional OR =1.57
[95% CI: 0.47, 5.19]).

3.6.3.4	Relevant Issues for Interpreting the Evidence Base

3.6.3.4.1 Potentially At-Risk Populations

Sex

Tu et al. (2021) measured hearing loss at speech frequency and at high frequency. In quartile
analyses, the magnitude of effect increased with each blood Pb quartile for each type of hearing loss (Tu
et al.. 2021). In sex-stratified analyses, the direction of effect remained but the magnitude of effect was

External Review Draft

3-204

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

greater among males. Kang et al. (2018) also observed an association between BLLs and hearing
impairment in adults. In quartile analyses, the magnitude of effect for high-level frequency hearing
impairment was greatest in the highest blood Pb quartile compared with the lowest blood Pb quartile, with
similar associations observed in males (OR= 1.63 [95% CI: 1.16,2.29]) and females (OR= 1.50 [1.03-
2.20]).

3.6.3.5 Integrated Summary and Causality Determination: Sensory Organ Function in
Adults

In the 2013 Pb ISA, causality determinations were separately determined for auditory and visual
function in adults, while no studies of olfactory function were evaluated (U.S. EPA. 2013a). For auditory
function in adults, the evidence was suggestive of, but not sufficient to infer, a causal relationship, based
primarily on findings from a few epidemiologic studies. Similar to the conclusion for children, the
evidence relating to visual function in adults was inadequate to determine if a causal relationship exists.
In the current ISA, the evidence for auditory, visual, and olfactory function are evaluated together, with a
single causality determination for sensory organ function.

The strongest evidence described in the 2013 Pb ISA was provided by the analysis of NAS men
for associations between higher tibia Pb levels and a higher rate of elevations in hearing threshold over
20 years (Park et al.. 2010). Several recent cross-sectional analyses of NHANES and KNHANES
generally support an association of Pb exposure (i.e., concurrent BLLs) with hearing loss (Tu et al.. 2021;
Kang etal.. 2018); Choi and Park (2017); (Choi et al.. 2012); however, recent studies are not entirely
consistent (Wang et al.. 2020). Hearing loss and altered responses on BAEPs in adult nonhuman primates
and rodents following lifetime or developmental Pb exposure have been demonstrated at BLLs as low as
29 (ig/dL (Jamesdaniel et al.. 2018; Laughlin et al.. 2009; Rice. 1997). The few studies that investigated
BLLs from 3-8 (ig/dL did not report altered BAEP in rodents, though effects on auditory processing may
occur at these lower exposure levels (Liu et al.. 2019; Carlson et al.. 2018; Zhu etal.. 2016).

The epidemiologic and experimental animal evidence pertaining to the association of Pb exposure
with visual function in adults remains limited. Toxicological studies have demonstrated biological
plausibility for Pb-induced effects on vision, including dysfunction of subcortical visual neurons, visual
processing areas, and retinal development (U.S. EPA. 2013a). A small number of recent studies
examining contrast sensitivity or acquired color vision loss found inconsistent results for associations
with blood or urine Pb level (Paulsen et al.. 2018; Fillion et al.. 2013; Shiue. 2013). Deficits in visual
temporal acuity have been demonstrated in adult nonhuman primates, although peak exposure levels are
higher than considered relevant for this assessment (Rice. 1998). Altered responses to ERGs have been
detected at a wide range of BLLs, but the direction of this effect (i.e., supernormal or subnormal
responses) is overall inconsistent (Fox et al.. 2008; Rothenberg et al.. 2002; Fox etal.. 1997).

External Review Draft

3-205

DRAFT: Do not cite or quote

-------
1	Olfactory function was not discussed in the 2013 Pb ISA. Recently, baseline BLL was associated

2	with reduced odor identification in a prospective analysis of the German HNRS (Casiens et al.. 2018).

3	In conclusion, prospective and cross-sectional analyses of hearing loss in adults is generally

4	consistent, but few experimental animal studies have investigated this endpoint at relevant BLLs.

5	Additional uncertainties remain for effects of Pb on visual and olfactory function. Based on studies of

6	auditory function, the evidence is suggestive of, but not sufficient to infer, a causal relationship

7	between Pb exposure and sensory function decrements in adults.

External Review Draft

3-206

DRAFT: Do not cite or quote

-------
Table 3-11 Summary of the evidence that is suggestive of, but not sufficient to infer, a causal relationship
between sensory function in adults.

Rationale for Causality
Determination3

Key Evidence13

References'3

Pb Biomarker Levels Associated with
Effects0

Auditory Function

Generally consistent
associations observed in
multiple epidemiologic
studies.

Prospective study found association
between tibia Pb level and higher rate of
increase in hearing threshold over 23 yr in
males enrolled in the NAS.

Park etal. (2010)

Tibia Pb mean: 22.5 |jg/g, measured near
end of follow-up



Cross-sectional analyses of NHANES and
KNHANES find Pb-associated effects on
hearing loss

Shiue (2013)

Choi et al. (2012) (Tu et al.. 2021) Kanq

etal. (2018)

Choi and Park (2017)



Uncertainty at relevant
exposure levels in
experimental animal studies

Hearing loss and altered responses on
BAEPs in adult nonhuman primates and
rodents demonstrated

Rice (1997)

Lauqhlin et al. (2009)
Jamesdaniel et al. (2018).

Lifetime or developmental Pb exposure
>29 [jg/dL



No altered BAEP in rodents, though
effects on auditory processing may occur

Carlson et al. (2018)
Zhu etal. (2016)
Liu etal. (2019).

3-8 [jg/dL

Visual Function

Inconsistent results across
limited epidemiologic studies

Associations of blood or urine Pb level
with contrast sensitivity or acquired color
vision loss were inconsistent

Shiue (2013)

Paulsen et al. (2018) Fillion et al. (2013)



Limited evidence from
experimental animal studies

Deficits in visual temporal acuity
demonstrated in adult nonhuman
primates

Rice (1998)

>30 [jg/dL

External Review Draft

3-207

DRAFT: Do not cite or quote

-------
Rationale for Causality
Determination3

Key Evidence13

References'3

Pb Biomarker Levels Associated with
Effects0

Biological plausibility
demonstrated

Dysfunction of subcortical visual
neurons, visual processing areas, and
retinal development

(U.S. EPA. 2013a)



Olfactory Function

Single study indicates
association

Impaired odor identification associated
with BLL in a single study

Casiens et al. (2018)

>9 vs. <5 |jg/dL

BAEP = brainstem auditory evoked potentials; BLL = blood lead level; KNHANES = Korean National Health and Nutrition Examination Survey NAS = Normative Aging Study;
NHANES = National Health and Nutrition Examination Survey; Pb = lead; yr = year(s).

aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Table I and Table II of the Preamble to the ISAs (U.S. EPA. 2015).

bDescribes the key evidence and references, supporting or contradicting, contributing most heavily to causality determination and, where applicable, to uncertainties or
inconsistencies. References to earlier sections indicate where the full body of evidence is described.

°Describes the Pb biomarker levels at which the evidence is substantiated.

External Review Draft

3-208

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

3.6.4

Neurodegenerative Diseases

The 2013 Pb ISA concluded that the evidence was "inadequate to determine that a causal
relationship exists" between Pb exposure and neurodegenerative diseases (U.S. EPA. 2013a). Evidence
was inconclusive for Amyotrophic Lateral Sclerosis (ALS) (see Section 4.3.9.2 of (U.S. EPA. 2013a))
and Alzheimer's disease (AD; see Section 4.3.9.1 of (U.S. EPA. 2013a)); however, a few case-control
studies each found higher BLLs in adults with essential tremor and higher bone Pb levels in adults with
PD (Weisskopf et al.. 2010; U.S. EPA. 2006b).The evidence was considered inconclusive overall due to a
limited number of studies, the potential for reverse causation (specifically for case-control studies in
which the reduced physical activity among cases could result in greater bone turnover and greater release
of Pb from bones into blood as compared with controls), and limited consideration for potential
confounding factors.

Recent epidemiologic studies indicated potential relationships between Pb exposure and some
neurodegenerative disease endpoints among non-occupational cohorts. The strongest evidence in the
current review includes well-designed studies of ALS and PD outcomes. Findings for AD, tremor, and
motor function were inconclusive. Measures of central tendency for Pb biomarker levels used in each
study, along with other study-specific details, including study population characteristics and select effect
estimates, are highlighted in Section 3.7, Table 3-17E. A large number of toxicological studies adds to the
evidence suggesting that developmental exposure to Pb increases the expression of pathophysiological
markers of AD, including amyloid beta (A(3) peptides, tau, and phosphorylated tau (p-tau), at lower BLLs
than were investigated in the previous ISA (<10 (ig/dL). Toxicological studies investigating potential
associations between Pb exposure and PD, ALS, and essential tremor remain limited in this review.
However, toxicological evidence for PD and ALS are supported by some studies in the 2013 review that
provided pathophysiological evidence for Pb-induced decreases in dopaminergic cell activity in the
substantia nigra, which can contribute to PD development, and Pb exposure affecting neurophysiologic
changes associated with ALS.

3.6.4.1 Epidemiologic Studies of Neurodegenerative Diseases

3.6.4.1.1 Alzheimer's Disease

MMSE is a widely used screening tool for AD and other types of dementia. Lower scores on
MMSE were consistently associated with higher bone Pb levels, which indicated long-term or cumulative
exposure to Pb, in the NAS studies assessed in the 2013 Pb ISA (Wang et al.. 2007; Weisskopf et al..
2004; Wright et al.. 2003). However, across studies, lower MMSE scores were not consistently associated
with BLL in adults (Weuve et al.. 2006; Nordberg et al.. 2000). There is uncertainty regarding the extent

External Review Draft

3-209

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

to which BLLs reflect past versus recent exposure in adults. Evidence regarding the association of Pb
exposure with clinical diagnosis of AD was limited to studies which did not find associations with higher
occupational exposure to Pb (Graves et al.. 1991) or higher Pb concentration in the brains (Haraguchi et
al.. 2001) in AD cases compared with unaffected controls. Overall, the latter studies considering the AD
endpoints had sufficient limitations (e.g., case-control design that may be subject to reverse causation,
leading to increased Pb levels among AD cases than controls, and limited consideration for potential
confounding). Consequently, the evidence is inconclusive regarding the effect of Pb exposure on AD.

Recent studies add to the evidence base, including an analysis of the NAS cohort that examined
the association of bone Pb biomarkers with cognitive impairment, including MMSE score, Farooqui et al.
(2017). and two studies that examined the association of blood Pb biomarkers with the clinical endpoints
of AD risk or AD mortality (Horton et al.. 2019; Yang et al.. 2018) in non-occupational cohorts (Table 3-
16T). Among the older male participants in the NAS higher patella, Pb concentration (IQR: 21 jxg/g) was
associated with increased risk (HR: 1.10, 95% CI: 0.99, 1.21) of having an MMSE score below 25
(threshold that represent cognitively not normal or at risk for dementia), while less support was observed
for an association with tibia Pb concentration (HR: 1.03, 95% CI: 0.88, 1.22) (Farooqui et al.. 2017).
Studies that specifically assessed clinically diagnosed AD or mortality did not provide strong evidence of
an association. A case-control study by Yang et al. (2018) included participants from clinical settings in
Taiwan and used standard case-control as well as propensity score-matched approaches to assess the
relationship between the heavy metals (Pb, Cd, Se, Hg) and AD risk. Findings from the multivariable
analysis showed the association between BLL and AD risk in tertiles, either in the full population tertile 2
(OR 1.00, 95% CI 0.56-1.79) and tertile 3 (OR 0.87, 95% CI 0.49-1.55) or propensity score-matched
population (tertile 2: OR 1.16, 95% CI 0.55-2.47; and tertile 3: OR 1.12, 95% CI 0.53-2.39), was
imprecise. A cohort study by Horton et al. (2019) used national data from five NHANES cycles (1999-
2008) and followed a large cohort of 8,080 participants from 1999 till December 2014 for AD-related
mortality to examine the longitudinal association between blood Pb and AD mortality. Results from Cox
proportional hazard models adjusted for various confounders and competing risks for AD mortality (death
due to cancer, cardiovascular disease (CVD), cerebrovascular accident [CVA], nephritis, and respiratory
disease) indicated that BLLs of 1.5 and 5 (ig/dL had 1.2 (95% CI = 0.70, 2.1) and 1.4 (95% CI = 0.54,
3.8) times the rate of AD mortality compared with those with a BLL of 0.3 (ig/dL, respectively. The
associations observed for various BLL categories with respect to the reference category of BLL 0.3 (ig/dL
were in a positive direction with increased AD risk for increasing BLL categories; however, the
associations were imprecise. The imprecise effect estimates are likely due to the small number of AD
mortality cases (n = 81), which resulted from AD mortality being determined from the listing of the
immediate cause of death rather than the underlying cause of death. This means the study may be
underpowered, potentially resulting in an unstable effect estimate.

External Review Draft

3-210

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

3.6.4.1.2 Amyotrophic Lateral Sclerosis

Case-control and cohort studies examining the association of BLL and ALS risk or ALS survival
that were included in the 2006 AQCD for Pb or the 2013 Pb ISA produced inconsistent results (Fang et
al.. 2010; Kamel et al.. 2008; Kamel et al.. 2002; Vinceti et al.. 1997). Recent case-control and cohort
studies that assessed biomarkers of Pb before disease development addressed uncertainties related to
temporality and reverse causality identified in previous reviews, thus expanding the support for an
association of BLL with ALS risk and survival.

Strong evidence for the association of Pb and ALS is provided by a recent prospective cohort
study that used data from the National Registry of Veterans in the United States with ALS cases
ascertained between April 2003 and September 2007 (with blood samples collected from January to
September 2007) and followed through the date of death or July 2013 (April 2003-Sep 2007) Fang et al.
(2017). The study was novel in that it assessed ALS mortality and survival and its association with blood
Pb and also bone turnover (formation and resorption) biomarkers. The association of ALS survival time
with blood Pb indicated that increased blood Pb was significantly associated with the increased mortality
and thus shorter survival after ALS diagnosis (HR: 1.23 [95% CI: 1.02, 1.49]) in the model mutually
adjusted for bone resorption and formation and other confounding variables. The observation of the
association between BLL and ALS after adjustment for biomarkers of bone turnover reduced
uncertainties related to reverse causality.

Additional studies indicating associations of Pb concentration in CSF and Pb in air also provide
some support for an association between Pb exposure and ALS. A case-control study in Italy used CSF
biomarkers for heavy metals including Pb Vinceti et al. (2017). The odds of ALS were greater in the
highest tertile of CSF Pb concentration than in the lowest tertile; however, the effect estimate was
imprecise (OR: 1.39, 95% CI 0.48-4.25). Another case-control study of a large nationally representative
U.S. sample (cases: 26,199 and controls: 78,597) used the U.S. healthcare claims database from the
Symphony Health's Integrated Dataverse Andrew et al. (2022). Participants with the first ALS diagnosis
after 6 months of enrollment in the database were included (diagnosis years 2013-2019). Controls were
matched based on age and sex, selected from the Symphony Health network, and also required to have a
minimum of 6 months enrollment in the database. The study did not use the Pb biomarkers but rather used
the airborne contaminants level data for 268 contaminants (including Pb) obtained from the U.S. EPA
National Emission Inventory (NEI) for 2008 to estimate the past exposure prior to the ALS onset (2013-
2019) at the participants' location of residence. The study used a three-phase approach to assess ALS risk
nationwide: discovery, validation, and confirmation. First, in the discovery phase, the study identified
major contaminants (out of 268 contaminants) that were significantly associated with the ALS risk.
Second, in the validation phase, the study evaluated various combinations of contaminants associated
with the ALS risk. In the final confirmatory phase, the study used cohorts only from NH, VT, and OH,
incorporating their detailed residential history information to capture changes in exposure due to
residential move prior to ALS diagnosis. The discovery phase identified 49 airborne contaminants

External Review Draft

3-211

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

(including Pb) associated with ALS risk. The relationship between these 49 contaminants and the risk of
ALS was further analyzed in the validation cohort, and airborne Pb and five PCBs were identified
associated with an increased risk for ALS (Pb: OR: 1.02, 95% CI: 1.01-1.03). The confirmatory analysis
using NH/VT and OH based cohorts with detailed residential history to calculate 5-, 10-, and 15-year past
exposure prior to diagnosis suggested significant increased risk for ALS associated with 10-year Pb
exposure history when the >75th percentile group was compared with the <5 0th percentile group
(NH/VT: OR: 2.03, 95% CI: 1.46-2.80, and OH: OR: 1.60, 95% CI: 1.28-1.98) in a multivariable model.
Despite the strong design and larger sample size of the study, the inference regarding the Pb-ALS risk for
this study should be interpreted with caution given the uncertainty regarding the relationship between the
estimated concentration of Pb in the air and Pb concentration in biomarkers as well as the influence of
potential unmeasured confounders.

3.6.4.1.3	Parkinson's Disease

A limited number of case-control studies assessed in the 2013 Pb ISA found positive associations
between bone Pb concentration and PD. A recent study by Paul et al. (2021) examined participants from
two large and independent population-based case-control studies (total n > 2,600)—the System Genomics
of Parkinson's Disease (SGPD), a consortium of three studies from across Australia and New Zealand;
and the Parkinson's Environment and Genes (PEG) study, a population-based study from three
agricultural counties of Central California—to explore the association of cumulative Pb exposure on the
PD risk. The study used novel epigenetic biomarkers of cumulative Pb exposure (i.e., DNA methylation
[DNAm] Pb data in the patella and tibia developed in the NAS cohort). The study analyzed the
relationship between DNAm Pb and PD separately for two cohorts and meta-analyzed the results. The
findings from the multivariable adjusted model suggested that PD risk was strongly associated with the
DNAm biomarker for tibia Pb levels in both cohorts (SGPD cohort: OR: 2.06, 95% CI: 1.66-2.56; PEG
cohort: OR: 1.60, 95% CI: 1.20, 2.15; meta-analyzed results [meta-OR: 1.89, 95% CI: 1.59-2.24]). The
results for DNAm in the patella for the two cohorts were inconsistent.

3.6.4.1.4	Tremor

A limited number of studies examined the association between BLLs and tremor (the most
common indicator of neurological diseases) in the previous review. The studies were potentially
influenced by reverse causation because inactivity due to disease condition and subsequent bone
resorption can lead to increased BLLs. A recent study used a cohort of men in the Veterans Affairs NAS
and assessed the longitudinal relationship of tremor score with bone Pb in the tibia (n = 670, mean:
21.23 jxg/g) and patella (n = 672, mean: 27.98 jxg/g), in addition to BLL (n = 807, mean: 5.01 (ig/dL). Ji et
al. (2015) found that over 8 years of follow-up, neither blood Pb nor bone Pb was associated with the
tremor score. However, among men younger than the median age (68.9 years), the tremor score increased

External Review Draft

3-212

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

as the quintile of blood Pb increased (p = 0.03), with men in the highest quintile scoring 0.35 (95% CI:
0.03, 0.67) points higher on the tremor scale than those in the lowest quintile. This pattern was not
evident for bone Pb. The tremor score in the study was based on assessments of drawing capability and
not the clinical diagnosis.

3.6.4.1.5 Motor Function

Several epidemiologic studies examined the association between Pb exposure and decrements in
motor function in adults (Casiens et al.. 2018; Khalil et al.. 2014; Grashow et al.. 2013; Ji et al.. 2013;
Shiue. 2013; Min et al.. 2012). Motor function was assessed using measures of balance, walking speed,
coordination, and strength. Inconsistencies in the results made it difficult to draw conclusions about the
association between Pb exposure and motor function in adults.

Most studies were cross-sectional in design. Results from studies that measured balance were
inconsistent. Among older adults (>50 years) in NHANES 2003-2004, Shiue (2013) observed an inverse
association between urinary Pb and balance disorders defined as self-reported dizziness, difficulty with
balance, or difficulty with falling in the past 12 months. For every log unit increase in urinary Pb (unit not
specified), the likelihood of having a balance disorder decreased (OR = 0.56 (95% CI: 0.38, 0.84]) (Shiue.
2013). Among adults (>40 years) who participated in the NHANES Balance Component, Min et al.
(2012) found that higher levels of blood Pb were generally associated with an increased likelihood of
failing a balance test. Balance dysfunction was evaluated using the Romberg Test of Standing Balance on
Firm and Compliant Support Surfaces, which measured a participant's ability to maintain balance under
various test conditions. Compared with the lowest quintile of blood Pb (<1.2 (.ig/dL). the likelihood of any
balance dysfunction (failing any balance test) increased in the fourth quintile (2.3-3.2 (ig/dL; OR = 5.23
[95% CI: 0.59, 46.43]) and fifth quintile (3.3-48 ^ig/dL; OR = 33.33 [95% CI: 1.94, 573.16]) (Min et al..
2012V

Among older adults (50-85 years) in NHANES, Ji et al. (2013) assessed the relationship between
blood Pb and walking speed. Walking speed was measured by timing a participant's walk for 20 feet at
their usual walking pace. Compared with the lowest quintile of blood Pb (0.2 to <1.2 (ig/dL), walking
speed decreased with increasing quintiles of blood Pb in women (p-trend = 0.005). In the highest quintile
of blood Pb (3.0 to <53.0 (ig/dL), the estimated mean walking speed in women was 0.11 feet/second
slower ([3 = -0.11 [95% CI: -0.19, -0.04]). On the contrary, BLL did not appear to be associated with
walking speed in men (Ji etal.. 2013). In MrOS, Khalil et al. (2014) examined the association between
blood Pb and walking speed and strength among older non-Hispanic Caucasian men (>65 years). In this
cross-sectional analysis, blood Pb did not appear to be associated with grip strength, walking speed, or
narrow-walk pace. Leg extension power (watts) measured with the Nottingham power rig had a negative
association for every log-|ig/dL increase in blood Pb ([3 = -0.03 [95% CI: -1.97, 2.03]). In addition, blood
Pb had a negative association with the participants' ability to stand from a chair without using their arms

External Review Draft

3-213

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

(OR per log-(ig/dL increase in blood Pb = 0.97 [95% CI: 0.88, 1.07]) (Khali 1 et al.. 2014). The reported
associations were very imprecise.

A few cohort studies measured fine motor abilities and hand-eye coordination in adults. In the
HNRS in Germany, Casiens et al. (2018) examined the effect of Pb exposure on fine motor abilities
among male participants who were recruited in 2000-2003 and followed up in 2011-2014. Fine motor
abilities were measured at follow-up (at ages 55-86 years) and included four tasks (tapping, aiming, line
tracing, and steadiness) carried out separately with each hand. Compared with the lowest BLL at baseline
(<5 (ig/dL), the highest BLL (>9 (ig/dL) was positively associated with tapping hits (OR = 1.35 [95% CI:
0.49, 3.70]) and steadiness errors (OR = 1.36 [95% CI: 0.50, 3.66]) but negatively associated with aiming
errors (OR = 0.56 [95% CI: 0.22, 1.42]) and line tracing errors (OR = 0.93 [95% CI: 0.32, 2.74]). The
magnitude of each association increased when using BLLs measured at follow-up, except for tapping hits,
which changed to a negative association (OR = 0.98 [95% CI: 0.82, 1.16]). In general, the ORs were
imprecise (Casiens et al.. 2018). In the Department of Veterans Affairs NAS, Grashow et al. (2013)
examined the association between bone Pb and a coordination task which involved inserting metal pegs
into a grooved pegboard. Bone Pb was measured at the patella and the midtibial shaft and was positively
associated with the grooved pegboard completion time. In other words, the pegboard test took longer to
complete for every 10 jj.g/g increase in patella Pb ([3 = 1.97 [95% CI: 0.55, 3.38]) and 10 jj.g/g increase in
tibia bone Pb ([3 = 3.11 [95% CI: 1.16, 5.06]) (Grashow et al.. 2013).

3.6.4.1.6 Summary

In summary, recent epidemiological studies suggested potential relationships between Pb
exposure and some neurodegenerative disease endpoints among non-occupational cohorts. Similar to the
conclusion of the 2013 Pb ISA, the direction and strength of the association is stronger for some
endpoints than for others. In the 2013 Pb ISA, evidence was inconclusive for ALS and AD; however, a
few case-control studies indicated relationships between higher BLLs in adults and essential tremor, and
between higher bone Pb levels in adults and PD. In the current review, stronger evidence from better
designed case-control and cohort studies is available for ALS and PD outcomes, whereas findings for AD,
tremor, and motor function are inconclusive.

Specifically for AD, two studies in the current review examined the association of blood Pb
biomarkers with clinical diagnosis of AD risk or AD mortality in non-occupational cohorts. This
represents a change from the previous ISA, in which the evaluated studies examined cognitive
impairment (e.g., as indicated by MMSE scores, which are used to screen for dementia and AD) rather
than the clinical diagnosis of AD. A recent case-control study that examined the relationship of BLLs and
AD risk reported an imprecise positive association between Pb exposure and AD risk (Yang et al.. 2018).
The inability to establish temporal relationships with this case-control study, given the inclusion of
prevalent cases of AD, leads to uncertainty because of potential reverse causality. Another cohort study
that investigated the relationships between BLLs and AD mortality addressed the temporality issue with

External Review Draft

3-214

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

the case-control study design. The study observed positive relationships between blood Pb and AD
mortality risk, but the association was imprecise (Horton et al.. 2019). For tremor outcomes, no
significant association was observed in one of the cohort studies that considered both blood and bone Pb
biomarkers as well as the tremor score for the overall participants (Ji et al.. 2015). However, the results
suggested an association of blood Pb and tremor score, particularly in younger men, when analysis was
performed by age categories. Studies reviewed for Pb exposure and motor function in adults also yielded
inconsistent findings. Hence, the relationships of Pb biomarkers with AD risk, tremor, or motor function
are inconclusive.

Studies for ALS in this review included one cohort study that examined the relationship between
BLLs and ALS survival among U.S. veterans (Fang et al.. 2017) and a large case-control study of
participants from a healthcare claims dataset examining associations between past airborne Pb exposures
and ALS risk (Andrew et al.. 2022). The findings from the cohort study by (Fang et al.. 2017) suggested
that increased levels of baseline blood Pb were associated with shorter ALS survival even after mutually
accounting for bone resorption and formation. This finding is in agreement with the results of a study
included in the previous review that suggested positive associations between Pb in blood or bone with
ALS risk (Fang et al.. 2010; Kamel et al.. 2002). Fang et al. (2017) further added to the evidence by
demonstrating that the association of bll with ALS survival persisted after adjustment for a biomarker of
bone turnover. Increased bone turnover in ALS patients could cause higher BLLs, potentially explaining
the associations between bll and ALS risk and survival (i.e., reverse causality). Another study
investigating ALS risk using a case-control design with a large sample from a healthcare database and
well characterized exposure and outcomes examined whether previous airborne Pb exposures were related
to ALS development. The study found significant positive associations between Pb exposure and
increased ALS risk (Andrew et al.. 2022). specifically for residential Pb exposure over the past 10 years.
The use of estimated airborne exposure without corresponding measurement of Pb exposure biomarkers is
a potential limitation of this study. For the PD outcome, only one case-control study was available in this
review. The study found a significant association between DNAm tibia Pb levels, used as an epigenetic
biomarker of bone Pb, and PD risk (Paul et al.. 2021).

3.6.4.2 Toxicological Studies of Neurodegenerative Diseases

Although the evidence was inconclusive overall, a few toxicological studies in the 2013 Pb ISA
suggested that Pb exposure in early life could influence AD-like pathologies (U.S. EPA. 2013a). AD
exhibits several neuropathologic hallmarks, such as senile plaques and neurofibrillary tangles (comprised
of A(3 and hyperphosphorylated tau aggregates, respectively), as well as synaptic loss and neuronal death.
Developmental exposure to Pb in rodents, resulting in BLLs >40 (ig/dL, increased A(3 peptides,
hyperphosphorylated tau, and other related endpoints (Li et al.. 2010; Basha et al.. 2005). Importantly,
these effects were not found following adult-only exposure. Wu et al. (2008) also demonstrated that 23-
year-old monkeys (Macaca fascicularis) given Pb from birth to PND 400 had elevated A(3 and amyloid

External Review Draft

3-215

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

plaques in their frontal cortex compared with unexposed age-matched controls. BLLs in these animals
ranged from 19 to 26 (ig/dL at PND 400 but had returned to baseline by adulthood. One previous study
performed in transgenic superoxide dismutase 1 (SOD1) mice (a model of ALS) found that adolescent
exposure to Pb reduced astrocyte reactivity and extended the survival time but had no significant effects
on the onset of disease in this model (Barbcito et al.. 2010). Tavakoli-Nezhad et al. (2001). reviewed in
the 2006 AQCD, demonstrated Pb-induced decreases in dopaminergic cell activity in the substantia nigra,
which is associated with PD. No recent PECOS-relevant studies have investigated the effects of Pb
exposure on ALS-relevant endpoint or endpoints related to essential tremor.

The potential relationship between Pb exposure and AD has been further explored in recent
literature (Table 3-17T). Of the two major A(3 isoforms (A(340 and A(342), the less predominant isoform,
A(342, is typically considered more prone to aggregation (Xiao et al.. 2015). A low A(342/A(340 ratio in
plasma and CSF has also been shown to indicate an increased risk for AD (Graff-Radford et al.. 2007).
(Zhou et al.. 2018) found increased expression of A(342 in the cerebral cortex and hippocampus of
Sprague Dawley rats following Pb exposure during adolescence. Gestational and lactational Pb exposure
resulting in lower BLLs, between 4-10 (ig/dL, also significantly increased A(340 in the cerebral cortex of
Kunming mice (Li et al.. 2016d). Utilizing a transgenic mouse model (Tg-SwDI), Gu et al. (2012)
reported that adolescent Pb exposure significantly increased both A(340 and A(342 in the cerebral cortex,
hippocampus, and CSF; however, the ratio of A(342/A(340 was not significantly different. Pb-treated
animals in this study also had increased amyloid plaque formation, which was co-localized with brain Pb
deposits. Increases in the expression of amyloid precursor protein (APP) and beta-secretase 1 (BACE1;
the enzyme that cleaves APP into A(3) have been demonstrated in some recent studies (Wu et al.. 2020b;
Zhou et al.. 2018; Sun et al.. 2014). but not all (Gu et al.. 2012).

Changes in the expression of both total tau (t-tau) and p-tau are considered biomarkers of AD in
humans. One recent study demonstrated that developmental Pb exposure (GD 0-PND 21; resulting in
BLLs of 7 (ig/dL) increased t-tau and p-tau in the cerebral cortex and cerebellum but not the hippocampus
of juvenile Wistar rats (Gassowska et al.. 2016b). These changes coincided with some evidence of
enhanced activity of two tau kinases (glycogen synthase kinase-3(3 and cyclin-dependent kinase 5
[CDK5]) in relevant brain regions. Another study found that postnatal Pb exposure in Wistar rats caused
significant changes in t-tau, p-tau, and related phosphatases in the hippocampus, but these effects were
transient and inconsistent at the time points tested (PND 21 and PND 30) (Rahman et al.. 2012b). Wu et
al. (2020b) also found no significant increases in hippocampal p-tau in Pb-exposed C57B1/6 mice at
4 months old; however, increases in hippocampal p-tau became apparent at 13 months old and persisted
until 16 months old. In the prefrontal cortex, p-tau was significantly elevated above age-matched control
values at 4 month and 13 months, but not 16 months. Importantly, BLLs in these rodents at 4 months
(when exposure was terminated) were nearly 60 (ig/dL and did not decrease to PECOS-relevant values
until 16 months (28 (ig/dL). One additional study Zhang et al. (2012) reported that 8 weeks of Pb
exposure increased p-tau in the hippocampus at BLLs as low as 10 (ig/dL. Notably, this study also

External Review Draft

3-216

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

reported increased alpha-synuclein in the hippocampus. Alpha-synuclein is a major constituent of Lewy
bodies, which are a neuropathologic hallmark of PD (Babaetal.. 1998).

In the 1980s, Rice (1990) established a cohort of monkeys (Macaca fascicularis) exposed to Pb in
the first 400 days of life (resulting in BLLs between 19-26 (ig/dL) and terminated at 23 years old. At the
time of termination, BLLs had returned to control levels. Using tissues from these animals, Wu et al.
(2008) found increases in the protein expression of A(3 and APP, as well as increases in the gene
expression of APP and specificity protein 1 (Spl, a transcriptional regulator of APP and tau), reported in
the previous ISA. Recently, Bihaqi and Zawia (2013) extended these findings by analyzing the cerebral
cortex tissue for changes in tau-related endpoints. Compared with age-matched controls, Pb-exposed
animals had significantly increased expression of t-tau, p-tau, and CDK5 (a tau kinase). These findings
were further supported by neuropathological changes (i.e., increases in p-tau immunoreactivity and
deposits). This study also found that mRNA levels of tau, CDK5, Spl, and Sp3 were significantly
increased.

Recent studies have also measured endpoints outside of those related to A(3 and tau.

Dysregulation of lipid pathways has been implicated in AD and neurodegenerative disorders (Di Paolo
and Kim. 2011). Zhou et al. (2018) found that Pb exposure decreased total and free cholesterol levels via
dysregulation of cholesterol metabolism in the cerebral cortex and hippocampus. Feng et al. (2019) found
that lifetime exposure to Pb significantly decreased neuronal density in the cerebral cortex at PNW 70.
This change was accompanied by a decrease in overall brain volume. In addition, several studies that
reported on AD-related neuromolecular changes also reported significant impairment of learning and
memory assessed via the Morris water maze (Wu et al.. 2020b; Li et al.. 2016d; Gu et al.. 2012; Rahman
et al.. 2012b). However, these results cannot be definitively attributed to AD-related neurobehavioral
changes, due to the well-known effects of Pb on cognitive function, which are unrelated to AD.

Deficiencies in balance, walking speed, coordination, and strength can also arise from insults to
the motor system in adulthood. Recent toxicological studies exposed mature rodents to Pb and
investigated the effects on motor function. Typical rotarod tests compare the latency to fall for subjects
placed on a rotating rod. Falling off more quickly indicates decreased coordination or balance. Locomotor
activity tests (e.g., measurements of distance traveled, counts of square crossings) can detect gross motor
problems as well; however other influences on behavior may factor into differences in the amount of
movement. Fine motor forelimb grip strength can be determined in rodents by pulling subjects holding
onto a measurement-taking tension bar. Mansouri et al. (2012) subjected Wistar rats to Pb acetate for
30 days and observed hyperactivity in open-field tests in males but not females on the final day
(PND 100). They also observed no effect on rotarod performance in both males and females on PND 100.
However, in a subsequent study, Mansouri et al. (2013) found that long-term exposure to Pb acetate in
drinking water (155-159 days starting at PND 55-60) resulted in substandard rotarod performance
(7.5 months old) for male Wistar rats, whereas the exposure had no effect on performance of female rats.
Similarly, in a long-term exposure study by Singh et al. (2019). male Wistar rats exposed daily to Pb

External Review Draft

3-217

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

acetate by oral gavage from 3 months to 6 months of age performed worse in rotarod and grip strength
tests compared with their saline-treated counterparts. Singh et al. (2019) also found decreased activity in
6-month-old Pb-treated rats. Al-Oahtani et al. (2022) observed a decrease in locomotor activity in 15-
week-old male mice after a 6-week Pb treatment period.

Summary

In summary, recent studies have significantly expanded the toxicological literature base
established in the last Pb ISA for AD. Significant increases in A(340 and A(342 following Pb exposure
were consistently detected in the cerebral cortex, hippocampus, and CSF in multiple studies at BLLs of 4-
30 (ig/dL. Pb-induced amyloid plaque formation was also reported in a transgenic mouse model of AD
(Tg-SwDI) (Gu et al.. 2012). Aged cynomolgus monkeys (23 years old), exposed to Pb during infancy,
had both amyloid plaques and tau deposits in their cerebral cortex; however, these findings are limited
somewhat by the small sample size (Bihaqi and Zawia. 2013; Wu et al.. 2008). In rodents, mean BLLs
<10 (ig/dL were shown to increase the expression and phosphorylation of tau in multiple brain regions,
but this effect was not entirely consistent between studies. Overall, recent studies have primarily focused
on exposure paradigms beginning during development, but one study demonstrated effects from an
exposure beginning in early adulthood (8 weeks) (Gu et al.. 2012). One study reported that the PD-related
protein, alpha-synuclein, was increased in the hippocampus in Pb-treated male rats (Zhang et al.. 2012).
Recent studies have not expanded on previous findings on ALS-related endpoints or contributed evidence
related to essential tremor.

3.6.4.3 Relevant Issues for Interpreting the Evidence Base

3.6.4.3.1 Concentration-Response Function

The shape of the C-R function was not examined in the studies of the association of Pb
biomarkers with neurodegenerative diseases in adults in the past review. The majority of studies in the
current review also did not explore the shape of the C-R function in the Pb-neurodegenerative disease
associations. Horton et al. (2019) explored the relationship between BLL and AD mortality using Cox
regression models that incorporated design effect or competing risks. The study found an increase in the
hazard rate ratio (HRR) by 14%-30% with each unit increase in BLL (see Figure 3-15 below). Given the
small number of AD mortality events in the study population, the effect estimates were imprecise and had
larger CIs. The authors also performed categorical analysis to explore blood Pb-AD mortality association.
Results from Cox proportional hazard models adjusted for various confounders and competing risks for
AD mortality (death due to cancer, CVD, CVA, nephritis, and respiratory disease) indicated that
participants in the 1.5 and 5 (ig/dL BLL categories had 1.2 (95% CI = 0.70, 2.1) and 1.4 (95% CI = 0.54,
3.8) times the rate of AD mortality compared with those with a blood Pb reference of 0.3 (ig/dL,

External Review Draft

3-218

DRAFT: Do not cite or quote

-------
1	respectively. The associations observed for various BLL categories with respect to the reference category

2	of BLL 0.3 (ig/dL were imprecise.

2.2

2.1
2

1.9
1.8
1.7
1.6
1.5
1.4

1.3

1.2
1.1

1

0.9
0.8

0.7
0.6
0.5

	Design e fleet

™ L€I-Design effect

	- Null line

	Competing risk

™- — — LCI-Competing risk

0.5

1.5

2 2.5 3
BLL (ng/dL)

3.5

4.5

BLL=blood lead level; HRR = hazard rate ratio; LCMower confidence interval.

Source: Reproduced with permission from Horton et al. (2019).

Figure 3-15 Hazard rate ratios for Alzheimer's disease mortality by blood Pb
level including the lower 95% confidence interval.

3.6.4.3.2 Potentially At-Risk Populations

Genetics

3	Several studies in the 2013 Pb review evaluating the association between Pb and MMSE (a

4	marker for AD) and effect modification by genetic variants provided support for effect modification of

5	the association by the ALAD genotype (details on Potentially At-Risk Populations in the Cognitive

6	Function Section: 3.6.1.3.2). A study by (Fang et al.. 2010) examining the association of BLL and ALS

7	risk and effect modification by the ALAD genotype suggested a significant Pb-ALS association among

8	ALAD1-1 carriers but a weaker and imprecise association among ALAD2 carriers. Tests to identify an

9	interaction between Pb and the ALAD genotype in the Pb-ALS association suggested no significant
10 difference in association between ALAD 1-1 versus ALAD2 carriers, however (p = 0.32).

External Review Draft

3-219

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

In the current Pb review, (Ji et al.. 2015) considered both bone and blood Pb biomarkers among
the NAS cohort and performed stratified analysis by ALAD gene (ALAD-2 carriers or non-carriers).

They found no effect modification by the ALAD genotype for the association between Pb biomarkers and
elevated tremor.

Age and Sex

A few studies in the current review explored the effect modification of the Pb-neurodegenerative
disease associations by age or sex. (Ji et al.. 2015) examined the associations of bone and blood Pb
biomarkers with tremor among the NAS cohort. The results suggested that among younger cohorts (i.e.,
below the median age of 68.9 years), the tremor score increased significantly with increasing quintile of
blood Pb (p = 0.03), and those in the highest quintile scored 0.35 (95% CI: 0.03, 0.67) points higher than
those in the lowest quintile. This pattern was not apparent when bone Pb biomarkers were used. Similarly.
(Paul et al.. 2021) performed stratified analysis of the DNAm estimated tibia and patella Pb
concentrations and PD risk by sex and found a significant association when tibia Pb concentration was
used. The magnitude of risk was higher for men in the SGPD cohort (OR and 95% CI: men: 2.48 [1.86,
3.34]; women: 1.67 [1.21, 2.33]), and the risk was higher for women in the PEG cohort (OR and 95% CI:
men: 1.49 [1.02, 2.20]; women: 1.81 [1.17, 2.85]).

3.6.4.4 Integrated Summary and Causality Determination: Neurodegenerative
Diseases

The 2013 Pb ISA (U.S. EPA. 2013a) concluded that the available evidence was "inadequate to
determine that a causal relationship exists between Pb exposure and neurodegenerative diseases in
adults." This conclusion was based on a limited number of studies that examined the association of blood
Pb or bone Pb levels with essential tremor, PD, ALS, and AD. These studies were not sufficient to reach a
conclusion regarding the presence or absence of an effect due largely to the potential for reverse causation
(i.e., reduced physical activity among cases resulting in greater bone turnover and higher BLLs), and
limited consideration for potential confounding factors. Limited studies in monkeys and rodents found
that developmental Pb exposure induced pathologies that underlie AD, and rodent studies suggested
neurophysiologic characteristics and changes related to ALS and PD. Recent epidemiologic studies
expanded the evidence bases indicating associations between Pb exposure and some neurodegenerative
diseases among non-occupational cohorts. The strongest evidence in the current review includes well-
designed case-control and cohort studies of ALS and PD, whereas findings from studies of AD, tremor,
and motor function remain inconclusive. Findings from recent toxicological studies add to the evidence
suggesting that developmental exposure to Pb increases the expression of proteins related to AD,
including A|3, tau and p-tau at lower BLLs than the values investigated in the previous ISA (<10 (.ig/dL).
Alterations in neuropathologic hallmarks of AD in older monkeys were also demonstrated following
developmental Pb exposure.

External Review Draft

3-220

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

Studies for ALS in this review included one well-designed cohort study that examined the
relationship of the blood Pb biomarker and ALS survival after ALS diagnosis among U.S. veterans (Fang
et al.. 2017) and one large case-control study that examined associations between past airborne Pb
exposures and ALS risk in participants from a healthcare claims dataset (Andrew et al.. 2022). The
findings from the cohort study by (Fang et al.. 2017) suggested that increased levels of past blood Pb
prior to mortality follow-up were associated with shorter ALS survival even after mutually accounting for
bone resorption and bone formation, thus reducing the uncertainty due to reverse causality. (Andrew et
al.. 2022) found significant positive associations between airborne Pb exposure and increased ALS risk,
specifically for residential Pb exposure over the past 10 years. For PD outcomes, only one case-control
study investigated the relationship between the risk of PD and epigenetic biomarkers by quantifying
DNAm tibia and patella Pb concentrations as cumulative Pb exposure. The authors found a significant
association between DNAm tibia Pb levels and PD risk (Paul et al.. 2021).

Toxicological studies investigating potential associations between Pb exposure and ALS or PD
remain limited. In one study reviewed in the previous ISA, Pb exposure was found to induce
neurophysiologic changes in a rodent model of ALS. Neurophysiologic characteristics of PD, such as
decreased activity of dopaminergic neurons in the substantia nigra and increased expression of
hippocampal alpha-synuclein, have also been demonstrated following Pb exposure.

The bulk of the epidemiologic evidence in the 2013 Pb ISA drawn upon to evaluate the
association of Pb exposure AD focused on cognitive impairment identified using dementia screening
instruments such as the MMSE. Recent studies that examined the association of blood Pb biomarkers
with clinical endpoints of AD risk (Yang et al.. 2018) or AD mortality (Horton et al.. 2019) in non-
occupational cohorts add to the evidence. The association between blood Pb exposure and AD risk
observed in Yang et al. (2018) was imprecise. Uncertainties related to potential reverse causality and
timing of exposure were not addressed in this study. (Horton et al.. 2019) addressed concerns raised for
the case-control study design used in Yang et al. (2018) but also observed imprecise relationships
between blood Pb and AD mortality risk, (Horton et al.. 2019). For tremor outcomes, no significant
association was observed between blood Pb and the tremor score when all study participants were
analyzed; however, an increase in the tremor score was observed for increasing BLL among younger
participants (Ji et al.. 2015). Studies reviewed for Pb exposure and motor function in adults also provided
inconsistent findings. Overall, the relationships of Pb biomarkers with AD risk, tremor, or motor function
are inconclusive. In contrast to the inconclusive epidemiologic study findings on AD, a large number of
recent toxicological studies add to the evidence which suggests that developmental exposure to Pb
increases the expression of proteins related to AD, including A(3, tau, and p-tau, at lower BLLs than the
values investigated in the previous ISA (<10 (ig/dL). In older monkeys (23 years), the neuropathologic
hallmarks of AD (i.e., amyloid plaques and tau deposits) were also demonstrated following
developmental Pb exposure.

External Review Draft

3-221

DRAFT: Do not cite or quote

-------
1

2

3

4

5

6

7

8

9

10

11

12

13

14

In summary, recent epidemiologic studies of varying quality strengthen the evidence for the
association of Pb exposure with ALS and PD, and reduce some uncertainties related to the potential for
reverse causality and the lack of temporality for these clinical endpoints. However, findings from
epidemiologic studies of AD, tremor, and motor function are inconclusive and uncertainties remain
regarding reverse causality, temporality issues, and exposure, and outcome assessments. The number of
epidemiological studies exploring these various endpoints are still relatively limited and vary in quality.
In contrast to the epidemiologic evidence, multiple recent toxicological studies add to the evidence
indicating that developmental exposure to Pb increases the expression of proteins related to AD, including
A(3, tau, and p-tau at lower BLLs than the values investigated in the previous ISA (<10 (.ig/dL).

Alterations in neuropathologic hallmarks of AD in older monkeys were also demonstrated following
developmental Pb exposure. However, toxicological studies investigating potential associations between
Pb exposure and ALS or PD remain limited. Overall, the evidence from epidemiologic and
experimental animal studies is suggestive of, but not sufficient to infer, a causal relationship between
Pb exposure and neurodegenerative diseases.

External Review Draft

3-222

DRAFT: Do not cite or quote

-------
Table 3-12 Summary of evidence that is suggestive of, but not sufficient to infer, a causal relationship
between Pb exposure and neurodegenerative diseases in adults.

Rationale for Causality
Determination3

Key Evidence13

References'3

Pb Biomarker
Levels Associated
with Effects0

Row heading if applicable

At least one high-quality
prospective cohort or case-control
study finds associations with ALS
and PD.

A prospective analysis of U.S. veterans found that higher baseline BLL was
associated with increased mortality / shorter survival after ALS diagnosis. The
association persisted after controlling for confounders including a biomarker of bone
turnover (formation and reabsorption), thus addressing the issue of reverse
causality.

Fang et al.
(2017)

Support from recent case-control studies using novel exposure metrics that found
associations between higher estimated air Pb exposure and ALS risk and between a
DNAm biomarker of tibia Pb and PD.

(Paul et al..
2021)

Limited number of studies address Well-designed case-control studies and prospective studies of ALS and PD
uncertainty due to temporality and assessed exposure prior to disease development and accounted for increased bone
reverse causation.	turnover resulting from the disease state.

External Review Draft

3-223

DRAFT: Do not cite or quote

-------
Rationale for Causality
Determination3

Key Evidence13

Pb Biomarker
References'3 Levels Associated
with Effects0

Coherence for AD provided by Amyloid plaques and/or tau deposits in transgenic rodents and aged monkeys
consistent evidence in animals following Pb exposure,
with relevant exposures.

Wu et al. (2008)

Bihaqi and	PeakBLLs:19-

Zawia (2013) 30 |jg/dL

Peak BLLs: 4-
58 [jg/dL

Evidence describes biologically
plausible pathways.

Evidence suggests that exposure to Pb results in neuronal cell death associated with Section 3.7
oxidative stress, neuroinflammation and altered energy metabolism, all of which may
underlie general neurodegenerative processes.

AD = Alzheimer's disease; ALS = amyotrophic lateral sclerosis; BLL = blood lead level; DNAm = DNA methylation; Pb = lead; PD =Parkinson's disease.

aBased on aspects considered in judgments of causality and weight of evidence in causal framework in Table I and Table II of the Preamble to the ISAs (U.S. EPA. 2015).

bDescribes the key evidence and references, supporting or contradicting, contributing most heavily to causality determination and, where applicable, to uncertainties or

inconsistencies. References to earlier sections indicate where the full body of evidence is described.

°Describes the Pb biomarker levels at which the evidence is substantiated.

1

External Review Draft

3-224

DRAFT: Do not cite or quote

-------
3.7

Evidence Inventories - Data Tables to Summarize Study Details

Table 3-1E Epidemiologic studies of Pb exposure and overt nervous system toxicity.

Reference and Study
Design

Study Population Exposure Assessment Outcome

Confounders

Effect Estimates and
95% Clsa

Yuan etal. (2006)
Cincinnati, OH

CLS
n: 24

Blood

Age at measurement:
BLL from 3 to 78
months averaged

Mean: 14.18
Range: 4.77-31.06
IJg/dL

MRI (subject asked to generate
verbs to activate language with
bilateral finger tapping)

birth weight and
marijuana use;
consideration of IQ,
sex, SES, gestational
age

Increasing BLL
associated with
decreased brain
activation in the left
frontal gyrus and left
middle temporal gyrus,
regions (semantic
language function)

External Review Draft

3-225

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population Exposure Assessment Outcome

Confounders

Effect Estimates and
95% Clsa

tReuben et al. (2020)

Dunedin
New Zealand
1972-2019
Cohort

Dunedin Study
n: 512

Whole blood Pb (pg/dL)
was measured via
GFAAS

Age at measurement:
11 yr

Mean (SD): 10.99
(4.63) [jg/dL
Max: 31 [jg/dL

Cortical thickness, cortical
surface area, hippocampal
volume, WMH volumFe,
BrainAGE index

High resolution images
showing cortical thickness,
cortical surface area, bilateral
hippocampal volume, WMH,
and FA were produced using
T1-weighted, fluid-attenuated
inversion recovery and
diffusion-weighted sequences
with a Siemens Skyra 3T
scanner with 64-channel head
and neck coil. BrainAGE index
was calculated as a composite
measure of all measured
indices. Outcomes were
assessed at 45 yr of age.

Sex, maternal IQ,
childhood

Betas

BrainAGE Index: 0.03
(0.00, 0.06)

Hippocampal Volume:
0.00 (-0.01, 0.00))

Cortical Surface Area:
-0.05 (-0.09, 0.00)

Cortical Thickness
(mm):

0.00 (0.00, 0.00)
WMH:

0.00 (0.00, 0.01)

Age at outcome:

External Review Draft

3-226

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population Exposure Assessment Outcome

Confounders

Effect Estimates and
95% Clsa

Cecil (2011)

Cincinnati, Ohio, Cincinnati
Children's Hospital Medical
Center
United States

Enrollment(mothers): 1979-
1984. Follow-up: Birth to 24 y
Cohort

CLS
n: 159

Whole blood Pb
measured using anodic
stripping voltammetry

Age at measurement:
3-78 months old

Pb prenatally and at
intervals to age 17 yr;
imaging 19-24 yr

Mean (SD) mcg/dL:
prenatal 8.3 (3.7), 3-
12 mo 10.6 (5.1).
(Reported in Dietrich et
al. 1993)

MRI brain assessments of 4
types: volumetric (morphology),
spectroscopy (chemical
concentrations), diffusivity
(organization), and functionality
(activation related to tasks).

Brain MRI/fMRI measures of
four types. (1) Volume of gray
matter. (2) Spectroscopy -
metabolites linked to neuronal
function and myelin
architecture: N-acetyl
aspartate, creatine and
phosphocreatine,
phosphocholines and
glycerolphosphocholine , myo-
inositol, glutamate and
glutamine. (3) Diffusivity in
white matter regions reflecting
axonal and myelin effects: FA;
mean, axial and radial
diffusivity. (4) Functionality:
activation related to task
performance.

Age at outcome:

19-24 years

Varied by outcome;
included age at imaging
and birth weight.

Reported a negative
association between
childhood BLL and gray
matter volume in
several regions: medial
and superior frontal
gyri, inferior parietal
lobule, cerebellar
hemispheres

Reported an
association between
higher childhood BLL
and lower metabolite
concentrations in
several regions: white
matter, left basal
ganglia, left cerebellar
hemisphere, vermis

External Review Draft

3-227

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population Exposure Assessment Outcome

Confounders

Effect Estimates and
95% Clsa

tBeckwith et al. (2021)	CLS

n: 123

Cincinnati, OH
United States

in utero to up to 33 yr of age
Cohort

MRI brain volumetrics of white

Pb was measured in
whole blood at 10 d, on
a quarterly basis up to
60 mo, and monthly at
66, 72, and 78 mo.
Samples were collected
mainly by venipuncture,
but occasionally by heel
or finger stick. Pb
concentrations were
quantified using anodic
stripping voltammetry

Age at measurement:
0-78 months

and gray matter, focusing on
regions involved in cognitive
and emotional function.

Age at time of imaging,
birth weight, total
intracranial volume.

BLLs were associated
with MRI-derived
decreases in white and
gray matter volumes in
the frontal, parietal, and
temporal lobes.

Decreased gray matter
volume in brain regions
responsible for
cognition and emotional
regulation associated
with criminal arrests

MRI scans (Voxel based
morphometry) were used to
examine spatial differences in
regional gray and white matter
volumes in adulthood (mean
age 26.8 yr) associated with
childhood blood Pb
concentrations at 78 mo.

Age at outcome:

18-33 years

10 days, on a quarterly
basis up to 60 mo, and
monthly at 66, 72, and
78 mo

Mean (SD) blood Pb at
78 mo: 7.82 pg/dL (4.2)
Max: 24.75 pg/dL

External Review Draft

3-228

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population Exposure Assessment Outcome

Confounders

Effect Estimates and
95% Clsa

tLamoureux-Tremblav et al.
(2021)

Nunavik, Northern Quebec

Canada

Cohort

NCDS
n: 71

Blood

Cord and concurrent
blood; GFAAS with
Zeeman background
correction

Age at measurement:
16-22

Cord blood: median
3.73 [jg/dL, mean
4.56 [jg/dL. Adolescent
blood: median
1.52 [jg/dL, mean
1.78 [jg/dL
Max: 17.81 pg/dL

Activation of the human neural
fear circuitry

fMRI analysis of brain
activation in response to a
validated fear conditioning and
extinction stimulus test

Age at outcome:

16-22 years

Sex, age, SES, and

alcohol/drug

consumption.

Higher differential
activation in the right
dorsolateral prefrontal
cortex in association
with higher postnatal
BLL.

tEthieret al. (2012)

Nunavik, Quebec
Canada
11 yr
Cohort

Prospective 11 yr Blood, Maternal Blood Neurological
study of Inuit
children from
Nunavik
n: 149

Concurrent venous
blood; GFAAS with
Zeeman background
correction (Perkin Elmer
model ZL 4100).

Age at measurement:
Pre-natal and 11 years

At birth mean:

4.6 pg/dL, SD: 3.1; At
11 yr mean: 2.6 pg/dL;
SD: 2.3

Achromatic pattern-reversal
VEPs with different visual
contrast levels were
administered, using generated
vertical sinusoidal gratings with
a spatial frequency of 2.5
cycles per degree. Children
viewed stimuli binocularly and
were instructed to fixate on a
small red dot. Pattern-reversal
VEPs were recorded from the
scalp over the visual cortex at
Oz derivation according to the
International 10-20 system

Age at outcome:

10-13 years

Analysis of variance
models controlled for
current Se, cord Se,
and gender.

Betas

N150 Latency
95% Contrast Level:
0.056 (0.099, 0.014)

*Note- 95% CIs were
converted from author
reported p-values

External Review Draft

3-229

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population Exposure Assessment Outcome

Confounders

Effect Estimates and
95% Clsa

tKimetal. (2018a)

Seoul
Korea

Enrollment 2010-2015
Case-Control

n: 150

Blood

Cortical Thickness

Blood Pb was measured Cortical thickness of brain
using atomic absorption regions was ascertained via

Age, intracranial
volume, gender.

spectrometer graphite
furnace

Age at measurement:
6-17 years

Mean (SD) - Cases: 1.3
(0.6) [jg/dL; Controls:
1.5 (0.7) [jg/dL

whole-brain structural MRI.

Age at outcome:
6-17 years

An interaction between
DRD2 and BLL on the
cortical thickness of the
frontal lobe in the
ADHD group, and a
brain-behavior
correlation between
cortical thickness and
the ADHD-RS
inattention score was
observed.

BLL = blood lead level; BrainAGE = Brain Age Gap Estimation; CI = confidence interval; CLS = Cincinnati lead study; d = day(s); FA = fractional anisotropy;
fMRI = functional magnetic resonance imaging; GFAAS = graphite furnace atomic absorption spectrometry; IQ = intelligence quotient; mo = month(s);
MRI = magnetic resonance imaging; Pb = lead; SD = standard deviation; Se = selenium; SES = socioeconomic status; VEP = visual evoked potential;
WMH = white matter hyperintensities; yr = year(s).

aEffect estimates are standardized to a 1 [jg/dL increase in BLL or a 10 |jg/g increase in bone Pb level, unless otherwise noted. For studies that report results
corresponding to a change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the
biomarker and standardized accordingly.

bResults are not standardized (e.g., BLL distribution data needed to calculate the standardized estimate was not reported or categorical data was analyzed).
fStudies published since the 2013 Integrated Science Assessment for Lead.

Table 3-1T Animal toxicological studies of Pb exposure and brain function.

Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported ((jg/dL)

Endpoints Examined

Graham et al. (2011)

Rat (Sprague Dawley)

PND 4 to PND28,

Oral, gavage

PND 29:

PND 11, 19, 29: Neurotransmitter



Control (vehicle), M/F,

every other day









n = 4-8





0.289 [jg/dL for Control





1 mg/kg, M/F, n = 4-8





3.27 |jg/dL for 1 mg/kg





10 mg/kg, M/F, n = 4-8





12.6 |jg/dL for 10 mg/kg



External Review Draft

3-230

DRAFT: Do not cite or quote

-------
Study	Species (Stock/Strain), Timing o, Expose BLL As Reported lMg/dL)	Endp.in
-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported (pg/dL)

Endpoints Examined

Corv-Slechta et al. (2012) Rat (Long-Evans)

Control (tap water), M,
n = 12

50 ppm Pb, M, n = 12

GD-60 to 10 mo Oral, drinking PND 5-6:
water
Oral,
lactation
In utero

10-11 mo: Neurotransmitter

<5 [jg/dL for Control
12.5 [jg/dL for 50 ppm
2.5 mo:

<5 [jg/dL for Control
6.43 [jg/dL for 50 ppm
10 mo:

<5 [jg/dL for Control
8.98 [jg/dL for 50 ppm

Weston et al. (2014)

Rat (Long-Evans) GD -60 to PND 21

Oral,

PND 5-6-Males:

PND 60: Neurotransmitter



Control (tap water), M/F,

lactation







n = 18-22 (9-11/9-11)

In utero

0.76 [jg/dL for Control





50 ppm, M/F, n = 18-22



15.7 [jg/dL for 50 ppm





(9-11/9-11)













PND 5-6 Females:









0.82 [jg/dL for Control









14.7 pg/dL for 50 ppm



External Review Draft

3-232

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported (pg/dL)

Endpoints Examined

Han et al. (2014)

Rat (Wistar)

Control (tap water), M,
n = 8

2 mM - postweaning
(PW), M, n = 8

2 mM - ME, M, n = 8

PW group: PND 21 Oral, drinking PND 21
to PND 42	water

Oral,

ME group: GD -21 lactation
to PND 20	In utero

PND 68: Histopathology

7.36 [jg/L (0.74 pg/dL) for
Control

NR for 2 mM - PW

146.6 pg/L (14.7 pg/dL) for
2 mM-ME

PND 63:

9.22 pg/L (0.92 pg/dL) for
Control

147.9 pg/L (14.8 pg/dL) for
2 mM - PW

46.13 pg/L (4.6 pg/dL) for
2 mM-ME

Barkurand Bairv (2015a) Rat (Wistar)

Pregestation

Control (untreated), M/F, exposure - GD -30

n = 9

to GD 0
Lactation only

0.2% solution -
Pregestation, M/F, n = 9 exposure - PND 0
to PND 22

0.2% solution -
Lactation, M/F, n = 9

0.2% solution -
Gestation, M/F, n = 9

Gestational
exposure - GD 0 to
GD 20

Gestation and
Lactation exposure

In utero PND 22:

0.19 pg/dL for Control

3.04 pg/dL for 0.2%
Pregestation

5.26 pg/dL for 0.2% Gestation

26.8	pg/dL for 0.2% Lactation

31.9	pg/dL for 0.2% Gestation
and Lactation

PND 30: Brain Weight

0.2% solution -
Gestation and Lactation, - GD 0 to PND 22
M/F, n = 9

External Review Draft

3-233

DRAFT: Do not cite or quote

-------
Study	Species (Stock/Strain), TMngo, Exposure BLL As Reported (|jg/dL)	Endpolnts Examined

Bashaetal. (2014)

Rat (Not Specified)

PND 1 to PND21

Oral,

PND 45:

PND 45, 4 mo, 12 mo, 18 mo:



Control (deionized



lactation



Neurotransmitter Analysis



water), M, n = 6





0.42 [jg/dL for Control





0.2% solution, M, n = 6





49.5 [jg/dL for 0.2% solution











4 mo:











0.56 [jg/dL for Control











14.4 [jg/dL for 0.2% solution











12 mo:











0.46 [jg/dL for Control











6.96 [jg/dL for 0.2% solution











18 mo:











0.12 [jg/dL for Control











11.2 [jg/dL for 0.2% solution



Rahman et al. (2012b)

Rat (Wistar)

PND 1 to PND 30

Oral, drinking

PND 21:

PND 21, 30: Brain Weight,



Control (tap water), M/F,



water



Histopathology



n = 4-10



Oral,

1.4 |jg/dL for Control









lactation







0.2% solution, M/F,





12.1 [jg/dL for 0.2% solution





n = 4-10

















PND 30:



1.2 [jg/dL for Control
12.8 [jg/dL for 0.2% solution

External Review Draft

3-234

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported (pg/dL)

Endpoints Examined

Nam et al. (2019b)

Rat (Sprague Dawley) GD 0 to 22

Control (not specified),

M/F, n = 12

0.2 % Solution, M/F,
n = 12

In utero PND21:

0.64 [jg/dL for Control
17.30 [jg/dL for 0.2% solution

PND 21: Histopathology

Saleh et al. (2018)

Rat (Sprague Dawley) GD 1 to GD 20

Control (deionized
water), M/F, n = 8 litters

160 ppm, M/F, n = 8
litters

In utero Maternal Blood Pb GD 20:
5.1 [jg/dL for Control
27.7 [jg/dL for 160 ppm

GD 20: Brain Weight, Histopathology

Li et al. (2016b)

Mice (Kunming)

Control

0.1% Pb
0.5 % Pb
1.0% Pb

GD 0 to PND 21 Oral, drinking 10 pg/dl_ for 0 ppm, 40
water

PND 21: Brain Histopathology

Mena et al. (2016)

Rat (Sprague Dawley) PND 0 to PND 21

Control (deionized
water), M/F, n = 7

300 ppm, M/F, n = 7

Oral,
lactation

PND 35:

7.61 |jg/L (0.76 pg/dL) for
Control

84.3 pg/L (8.43 pg/dL) for
300 ppm

NR: Histopathology

External Review Draft

3-235

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported (pg/dL)

Endpoints Examined

Amos-Kroohs et al.
(2016)

Rat (Sprague Dawley)

Control (sodium
acetate), M/F, n = 16
(8/8) per time point

1 mg/kg Pb, M/F, n = 16
(8/8) per time point

10 mg/kg Pb, M/F,
n = 16 (8/8) per time
point

PND 4 to PND 28 Oral, gavage PND 29:

1.27 [jg/dL for Control
2.76 pg/dL for 1 mg/kg
9.07 pg/dL for 10 mg/kg

PND 29: Neurotransmitter

Rahman et al. (2018)

Rat (Wistar)

Control (tap water), M/F,
n = 20

0.2% solution, M/F,
n = 20

PND 1 to PND 21

Oral, drinking PND 21:
water

PND 21, 30: Brain Weight

Oral,
lactation

2.2	[jg/dL for Control

12.4 [jg/dL for 0.2% solution
PND 30:

3.3	[jg/dL for Control

22.7 [jg/dL for 0.2% solution

Baranowska-Bosiacka et Rat (Wistar)	GD 0 to PND 21

al. (2017)	Control (distilled water),

M/F, n = 8

0.1% solution, M/F,
n = 8

Oral,
lactation
In utero

PND 28:

0.05 [jg/dL for Control
6.90 [jg/dL for 0.1% solution

PND 28: Histopathology

Baranowska-Bosiacka et Rat (Wistar)
al. (2013)	Control (distilled water),

M/F, n = 36 (17/19)

0.1% solution, M/F,
n = 36 (18/18)

GD 0 to PND 21

Oral,
lactation
In utero

PND 28:

0.93 [jg/dL for Control
6.86 [jg/dL for 0.1% solution

PND 28: Histopathology

External Review Draft

3-236

DRAFT: Do not cite or quote

-------
Study	Species (Stock/Strain), TMngo, Exposure BLL As Reported (|jg/dL)	Endpolnts Examined

Wanqetal. (2013)

Rat (Sprague Dawley)

GDOto PND 1,

Oral, drinking

PND 72:

PND 72: Brain Weight



Control (untreated), M/F,

PND 1 to PND 21,

water







n = 6

PND 21 to 42

Oral,

34.99 |jg/L (3.5 pg/dL) for









lactation

Control





0.2% Pb (w/v), M/F,



In utero







n = 6 - Gestational





35.78 pg/L (3.58 pg/dL) for 0.2





Exposure





% solution Gestational





0.2% Pb (w/v), M/F,





65.97 pg/L (6.60 pg/dL) for





n = 6 - Lactational





0.2% solution Lactational





Exposure

















110.67 pg/L (11.07 pg/dL) for





0.2% Pb (w/v), M/F,





0.2% solution Ablactational





n = 6 - Ablactational











Exposure









Wanaetal. (2016)

Rat (Sprague Dawley)

PND 24 to PND 56

Oral, drinking

PND 56:

PND 60-66: LTP, Neuronal Morphology



Control (tap water), M,



water







n = 7





11 pg/L (1.1 pg/dL) for Control





100 ppm, M, n = 9





133 pg/L (13.3 pg/dL) for











100 ppm



Li etal. (2016c)

Mice (Kunming)

GDOto PND 21

Oral,

PND 21:

PND 21: Histopathology



Control (deionized



lactation







water), M/F, n = 10



In utero

8.27 pg/L (0.827 pg/dL) for











Control





0.1% (1000 ppm), M/F,











n = 10





41.05 pg/L (4.11 pg/dL) for











0.1% solution





0.2% (2000 ppm), M/F,











n = 10





82.93 pg/L (8.29 pg/dL) for











0.2% solution





0.5% (5000 ppm), M/F,











n = 10





105.33 pg/L (10.53 pg/dL) for











0.5% solution



External Review Draft

3-237

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported ((jg/dL)

Endpoints Examined

Sobolewski etal. (2018)

Mice (C57BL/6)

GD -60 to PND 21

Oral,

PND 6-7:

PND 60: Epigenetics



Control (deionized



lactation







water), M/F, n = 6-12



In utero

0.37 pg/dL for Control





100 ppm, M/F, n = 6-12





10.2 pg/dL for 100 ppm



Barkurand Bairv (2016)

Rat (Wistar)

Control (tap water with
acetic acid), M/F, n = 8

0.2% solution,
pregestation only (PG),
M/F, n = 8

GD -30 to PND 21

Oral,
lactation
In utero

PND 22:

0.5 pg/dL for Control

9.4 pg/dL for 0.2% solution,
pregestation only (PG)

PND 30: Histopathology

0.2% solution, gestation
only, M/F, n = 8

0.2% solution, lactation,
M/F, n = 8

0.2% solution, gestation
and lactation, M/F, n = 8

16.6 [jg/dL for 0.2% solution,
gestation only

30.1 [jg/dL for 0.2% solution,
lactation

33.4 [jg/dL for 0.2% solution,
gestation and lactation

Shvachiv et al. (2018)

Rat (Wistar)

Control (tap water), M/F,
n = 8

0.2% (p/v) solution
(distilled water), M/F,
n = 9 - Intermittent
exposure

0.2% (p/v) solution, M/F,
n = 9 - Continuous
exposure

Intermittent
Exposure: GD 7 to
PND 84, PND 140
to PND 196

Continuous
Exposure: GD 7 to
PND 196

Oral, drinking PND 196:
water

Oral,
lactation
In utero

<0.1 [jg/dL for Control

18.8 [jg/dL for 0.2%
(Intermittent)

24.4 pg/dL for 0.2%
(Continuous)

PND 189: Brain Histopathology

Stansfield et al. (2015) Rat (Lonq-Evans) GD 0 to PND 50

Oral, diet

PND 50:

PND 50: Neurotransmitter Analysis,

Control (chow), M/F,

Oral,



Brain Histopathology

n = 4-7

lactation

0.6 pg/dL for Control





In utero





1500 ppm, M/F, n = 4-7



22.2 pg/dL for 1500 ppm



External Review Draft

3-238

DRAFT: Do not cite or quote

-------
Study Species (Stock/Strain),	Timing o, Expose BLL As Reported lMg/dL)	Endp.in
-------
Zhao etal. (2018)

Rat (Sprague Dawley) GD -14 to PND 10 Oral,
Control (tap water), M,	lactation

n = 8	In utero

0.005% solution, M,
n = 8

0.01% solution, M, n = 8
0.02% solution, M, n = 8

External Review Draft

3-240

PND 0:

1.9 |jg/dL for Control
17.9 [jg/dL for 0.005% solution

23.2	[jg/dL for 0.01 % solution
48.8 [jg/dL for 0.02% solution
PND 3:

I.9	|jg/dL for Control

6.7 [jg/dL for 0.005% solution

II.5	|jg/dL for 0.01% solution
23.1 [jg/dL for 0.02% solution
PND 7:

1.3 |jg/dL for Control

8.1	[jg/dL for 0.005% solution

12.3	[jg/dL for 0.01 % solution
18.7 [jg/dL for 0.02% solution
PND 10:

1.2	|jg/dL for Control

5.6 [jg/dL for 0.005% solution
7.0 [jg/dL for 0.01 % solution
12.3 [jg/dL for 0.02% solution
PND 14:

0.7 [jg/dL for Control

PND 30: Electrophysiology,
Histopathology

DRAFT: Do not cite or quote

-------
Study	Species (Stock/Strain), TMngo, Exposure BLL As Reported (|jg/dL)	Endpolnts Examined

4.0	[jg/dL for 0.005% solution

5.5 [jg/dL for 0.01% solution

8.9 [jg/dL for 0.02% solution
PND 21:

1.1	|jg/dL for Control

2.5 [jg/dL for 0.005% solution
2.5 [jg/dL for 0.01% solution
2.98 [jg/dL for 0.02% solution
PND 30:

1.5 [jg/dL for Control
1.0 [jg/dL for 0.005% solution
1.5 [jg/dL for 0.01% solution
1.5 [jg/dL for 0.02% solution

Dominquez etal. (2019) Mice (C57BL/6)	PND 0 to PND 28 Oral,

Control (tap water), M/F,	lactation

n = 10 (7/3)

30 ppm, M/F, n = 10
(6/4)

330 ppm, M/F, n = 10
(4/6)

PND 28 - Females:
0.02 [jg/dL for Control
3.03 [jg/dL for 30 ppm
12.79 [jg/dL for 330 ppm
PND 28- Males:
0.03 [jg/dL for Control
3.68 [jg/dL for 30 ppm
15.42 ug/dL for 330 ppm

PND 28: Histopathology

External Review Draft

3-241

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported (pg/dL)

Endpoints Examined

Du et al. (2015)

Rat (Sprague Dawley)

Control (distilled water),
M/F, n = 8

250 ppm, M/F, n = 8

PNDOto PND90

Oral, drinking PND 30:
water

Oral,	13.9 pg/L (1.4 pg/dL) for

lactation	Control

205.6 pg/L (20.6 pg/dL) for
250 ppm

PND 60:

15.0 pg/L (1.5 pg/dL) for
Control

321.9 pg/L (32.2 pg/dL) for
250 ppm

PND 90:

11.8 pg/L (1.2 pg/dL) for
Control

PND 30, 60, 90: Histopathology

379.2 pg/L (37.9 pg/dL) for 250
ppm

Mansouri et al. (2013)

Rat (Wistar)

Control (tap water or
water + NaAc), M/F,
n = 16 (8/8)

50 ppm, M/F, n = 16
(8/8)

PND 55 to
PND 181

Oral, drinking PND 178-181 - Females:
water

NR for Control
10.6 pg/dL for 50 ppm
PND 178-181 - Males:
NR for Control
18.9 pg/dL for 50 ppm

PND 161-179: Neurotransmitter

External Review Draft

3-242

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported (pg/dL)

Endpoints Examined

Zhou etal. (2018)

Rat (Sprague Dawley) PND 24 to PND 52 Oral, drinking PND 52:

Control (distilled water),	water

13.3 |jg/L (1.3 [jg/dL) for
Control

M, n = 10

0.5% solution, M, n = 10
1.0% solution, M, n = 10
2.0% solution, M, n = 10

148.9 [jg/L (14.9 pg/dL) for
0.5% solution

231.3 pg/L (23.1 pg/dL) for
1.0% solution

PND 24, 31, 38, 45, 52: Brain Weight,
Brain Histopathology

293.4 ug/L (29.3 pg/dL) for
2.0% solution

Dumkova et al. (2017)

Mice (ICR)

Control, F, n = 10

106/cm3 PbO
nanoparticles, F, n = 10

NR (24 g) - 6 wk

continuous

exposure

Inhalation After 6 wk treatment:

11 ng/g (1.16 pg/dL) for
Control

132 ng/g (13.99 pg/dL) for
106/cm3

After 6 wk treatment: Histopathology

External Review Draft

3-243

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported (pg/dL)

Endpoints Examined

Xiao etal. (2014)

Rat (Wistar)	Pre-weaning: GD

Control (tap water), M/F, -21 to PND 21
n = 10 (5/5)

Postweaning:
Pre-weaning: 2 mM PND 21 to PND 84
solution, M/F, n = 10
(5/5)

Postweaning: 2 mM
solution, M/F, n = 10
(5/5)

Oral, drinking PND 21 - Pre-weaning:
water

10.09 |jg/L(1 ijg/dL) for
Control

Oral,
lactation
In utero

PND 84 and PND 91: Histopathology

103.8 [jg/L (10.4 [jg/dL) for 2
mM solution

PND 21 - Postweaning:

Not Reported

PND 91 - Pre-weaning:

10.32 |jg/L(1 [jg/dL) for
Control

39.27 |jg/L (3.9 [jg/dL) for 2
mM solution

PND 91 - Postweaning:

10.32 |jg/L(1 [jg/dL) for
Control

105.45 [jg/L (10.5 [jg/dL) for 2
mM solution

Sobin et al. (2013)

Mice (C57BL/6)	PND 1 to PND 28

Control (tap water), M/F,
n = 30

30 ppm, M/F, n = 30
230 ppm, M/F, n = 30
330 ppm, M/F, n = 30

Oral,
lactation

PND 28:

0.22 [jg/dL for Control
4.12 pg/dL for 30 ppm
10.31 |jg/dLfor230 ppm
13.84 [jg/dL for 330 ppm

PND 28: Histopathology

External Review Draft

3-244

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported (pg/dL)

Endpoints Examined

Sobolewski et al. (2020) Mice (C57BL/6)

F0:

Control (deionized
water), F, n = 10

100 ppm, F, n = 10
F1:

see Figure 1, n = 12

F1: GD -60 to PND Oral,
23-27	lactation

In utero

F1 PND 6-7:

0 [jg/dL for Control

12.5 pg/dL for 100 ppm (F0
dosing)

F3 PND 6-7:

0 ng/dL for Control

0 pg/dL for 100 ppm (F0
dosing)

PND 60-120 (variable by endpoint):
Neurotransmitter, Epigenetics

F2:

see Figure 1, n = 12

F3:

see Figure 1, n = 8-10

Ouvanq et al. (2019)

Rat (Sprague Dawley) GD 0 to PND 679 Oral, drinking	wk 97:

Control (tap water), M/F,	water

n = 6-10	Oral,	0 mg/L (0 pg/dL) for Control

lactation

0.05/0.01% solution,	In utero	0.216 mg/L (21.6 pg/dL) for

M/F, n = 6-10	0.05/0.01% solution

PND 679: Histopathology

Saleh et al. (2019)

Rat (Sprague Dawley) NR (190-220g) - Oral, drinking After 20 d treatment:
Control (deionized	20 d of treatment water

water), F, n = 8	5.4 [jg/dL for Control

After 20 d treatment: Brain Weight
Histopathology

160 ppm, F, n = 8

23.8 [jg/dL for 160 ppm

External Review Draft

3-245

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported (pg/dL)

Endpoints Examined

Singh et al. (2019)

Rat (Wistar)	3 mo to 6 mo

Control (distilled water),

M, n = 5

2.5 mg/kg, M, n = 5

Oral, gavage

6 mo:

5.76 [jg/dL for Control
28.4 pg/dL for 2.5 mg/kg

6 mo: Brain Weight Brain
Histopathology

Xiao et al. (2020)

Rat (Sprague Dawley) GD -7 to PND 68

Control (tap water), F,
n = 10

125 ppm, F, n = 10

Oral, drinking
water
Oral,
lactation
In utero

PND 68:

24.23 ng/mL (2.4 pg/dL) for
Control

205 ng/mL (20.5 pg/dL) for
125 ppm

PND 22, 68: Histopathology

Sun et al. (2014)

Rat (Sprague Dawley)	NR (230-260 g) -

Control (tap water), NR,	3 mo of treatment
n = 20

580 ppm, NR, n = 20

Oral, drinking
water

After 3 mo treatment:	After 3 mo treatment: Histopathology

3.0 pg/L (0.3 pg/dL) for Control

56.8 pg/L (5.7 pg/dL) for
580 ppm

Su et al. (2016)

Rat (Sprague Dawley) PND 20 to PND 76

Control (deionized water
with 0.9% saline), M,
n = 4

200 ppm, M, n = 4

Oral, gavage PND 76:

PND 76: Histopathology

7.99 pg/L (0.8 pg/dL) for
Control

84.17 pg/L (8.4 pg/dL) for
200 ppm

Song et al. (2014)

Rat (Sprague Dawley) PND 20-22 to
Control (tap water), M, PND 76-78
n = 9

100 pg/mL, M, n = 9
200 pg/mL, M, n = 9
300 pg/mL, M, n = 9

Oral, drinking
water

PND 76-78:

0.73 pg/dL for Control
4.7 pg/dL for 100 pg/mL
10.1 pg/dL for 200 pg/mL
12.3 pg/dL for 300 pg/mL

PND 76-78: Histopathology

External Review Draft

3-246

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain),
n, Sex

SS?™ EKr BLL As Reported (Mg/dL)

Endpoints Examined

Zhou et al. (2020a)

Rat (Sprague Dawley) GD 1 to PND 364

Control (distilled water),

M, n =5-11

0.5 g/L solution, M,
n = 5-11

2.0 g/L solution, M,
n = 5-

Oral, drinking PND 21:
water

0 mg/L (0 |jg/dL) for Control

0.1 mg/L (10 pg/dL) for 0.5 g/L
solution

0.36 mg/L (36 pg/dL) for
2.0 g/L solution

PND 364:

0 mg/L (0 |jg/dL) for Control

0.15 mg/L (15 pg/dL) for
0.5 g/L solution

0.51 mg/L (51 pg/dL) for
2.0 g/L solution

PND 21, 364: Histopathology,
Electrophysiology

Liu et al. (2019)

Rat (Sprague Dawley) PND 1 to PND 21

Control (tap water), F,
n = 12

58 mg/L, F, n = 11

Oral,
lactation

PND 9:

0 pg/dL for Control
7.9 pg/dL for 58 mg/L
PND 21:

0 pg/dL for Control,
8.2 pg/dL for 58 mg/L
PND 40:

0 pg/dL for Control
0 pg/dL for 58 mg/L

PND 93: Histopathology

External Review Draft

3-247

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported (pg/dL)

Endpoints Examined

Nan et al. (2016)

Mice (C57BL/6)	PND 21 to PND 56

Control (sterile water),

NR, n = 30

9.6 mmol/L, NR, n = 30

Oral, drinking
water

PND 7:

0 |jg/L (0 pg/dL) for Control

106.3 [jg/L (10.6 pg/dL) for 9.6
mmol/L

PND 14:

0 |jg/L (0 pg/dL) for Control

293.2 |jg/L (29.3 pg/dL) for 9.6
mmol/L

PND 35:

0 |jg/L (0 pg/dL) for Control

959.6 |jg/L (96 pg/dL) for 9.6
mmol/L

PND 56: Histopathology

Singh et al. (2017)

Rat (Wistar)	NR (160-200 g) -

Control (distilled water),	14 days of

M, n = 3-6	treatment

7.5 mg/kg, M, n = 3-6

Oral, gavage 12 hr after last treatment:
5.54 pg/dL for Control
30.28 pg/dL for 7.5 mg/kg

12 hr after last treatment: Brain Weight,
Histopathology

Biiooret al. (2012)

Rat (Wistar)

Control (deionized
water), M/F, n = 10

50 ppm, M/F, n = 10

GD Oto PND 45

Oral, drinking
water
Oral,
lactation
In utero

PND 45:

4.06 pg/dL for Control
10.65 pg/dL for 50 ppm

PND 45: Neurotransmitter

External Review Draft

3-248

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported (pg/dL)

Endpoints Examined

Wanqetal. (2021a)

Rat (Sprague Dawley) GD 0 to PND 21

Control (deionized
water), M, n = 3

0.05% solution, M, n = 3

0.1% solution, M, n = 3

Oral,
lactation
In utero

PND 21:

23.1 |jg/L (2.31 pg/dL) for
Control

248 |jg/L (24.8 pg/dL) for
0.05% solution

PND 21: Histopathology

302 pg/L (30.2 pg/dL) for 0.1%
solution

361 pg/L (36.1 pg/dL) for 0.2%
solution

Liu et al. (2022c)

Rat (Sprague Dawley)

PND 35 to PND

Oral, drinking

PND 119:

PND 119: Histopathology



Control (tap water), M,

119

water







n = 10





10.9 pg/L (1.09 pg/dL) for











Control





0.2% solution, M, n = 10

















176 pg/L (17.6 pg/dL) for 0.2%











solution



Hsu et al. (2021)

Rat (Sprague Dawley)

PND 42 to PND 77

Oral, drinking

PND 84:

PND 78 to PND 84: Electrophysiology



Control (deionized



water







water), M, n = 6





0.9 pg/L (0.09 pg/dL) for











Control





250 ppm, M, n = 6

















15.3 pg/L (1.53 pg/dL) for











250 ppm



Sadeahi etal. (2021)

Rat (Wistar)

GD 0 to PND 50

Oral, drinking

PND 50:

PND 50: Histopathology



Control (untreated), M,



water







n = 5



Oral,

0.58 pg/dL for Control









lactation







1500 ppm, M, n = 5



In utero

3.4 pg/dL for 1500 ppm



Viqueras-Villasenor et al.

Rat (Wistar)

GD 0 to PND 21

Oral,

PND 110:

PND 90 to PND 110: Histopathology

(2021)

Control (tap water), M,



lactation







n = 20



In utero

2.04 pg/dL for Control





320 ppm, M, n = 20





26.3 pg/dL for 320 ppm



External Review Draft

3-249

DRAFT: Do not cite or quote

-------
Study	Species (Stock/Strain), TMngo, Exposure BLL As Reported (|jg/dL)	Endpolnts Examined

Lonq et al. (2022)

Rat (Sprague Dawley) 6 wk to 18 wk

Oral, drinking

00

NR: Histopathology, Neurotransmitter



Control (untreated), M,

water





n = 12



2.14 [jg/L (0.214 (jg/dL) for









Control





200 mg/L solution, M,









n = 12



32.48 |jg/L (3.25 pg/dL) for









200 mg/L solution



Abazvan et al. (2014)

Mice (CAMKII-tTA; GD 0 to PND 180

Oral, diet

6 mo - Females:

PND 180: Brain Volume, Brain MRI,



heterozygous or

Oral,



Morphometric Measurements in several



homozygous for

lactation, in

0.6 pg/dL for Control (het)

regions



mDISCI)

utero





Control (het), M/F,



0.8 pg/dL for Control (mutant)





n = 5-10











34.9 pg/dL for 1500 ppm (het)





Control (mutant), M/F,









n = 5-10



33.3 pg/dL for ppm (mutant)





1500 ppm(het), M/F,



6 mo - Males:





n = 5-10













1.1 pg/dL for Control (het)





1500 ppm (mutant), M/F,









n = 5-10



1.1 pg/dL for Control (mutant)









26.1 pg/dL for 1500 ppm (het)









25.0 pg/dL for 1500 ppm









(mutant)



External Review Draft

3-250

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported (pg/dL)

Endpoints Examined

Zhu etal. (2013)

Rat (Sprague Dawley) GD 0 to PND 490

Control (untreated), M,
n = 11-13

510 mg/L, M, n = 11-13

Oral, drinking	PND 21

water	g (jg/L for control

Oral,

lactation, in	0.27 mg/L (27 pg/dL) for

utero	510 mg/L

PND 287
0 |jg/L for control

0.24 mg/L (24 pg/dL) for
510 mg/L

PND 490
0 |jg/L for control

PND 21 to PND 490: Histopathology

0.25 mg/L (25 pg/dL) for
510 mg/L

Nam etal. (2018a)

Rat (Sprague Dawley) GD 0 to PND 21

Control (distilled water),

M/F, n = 12

0.2% solution M/F,
n = 12

Oral, drinking PND 21

1.28 pg/dL for control

PND 21: Histopathology

water
Oral

lactation, in
utero

12.67 pg/dL for 0.2% solution

Gassowska et al. (2016a) Rat (Wistar)

Control (drinking water),
M/F, n = 4-8

0.1% solution M/F,
n = 4-8

GD 0 to PND 28 Oral, drinking PND 28

water	0.93 pg/dL for control

Oral

lactation, in 6 86 Mg/dL for QAo/o
utero

PND 28: Histopathology

External Review Draft

3-251

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported (pg/dL)

Endpoints Examined

Gassowska et al. (2016b) Rat (Wistar)	GD 0 to PND 28

Control (drinking water),

M/F, n = 4-8

0.1% solution, M/F,
n = 4-8

Oral, drinking PND 28

0.93 [jg/dL for control

PND 28: Histopathology

water
Oral

lactation, in
utero

6.86 [jg/dL for 0.1%solution

Sepehri and Ganii (2016) Rat (Wistar)

Control, M, n = 8

0.05% solution, M, n = i

GD 5 to PND 25

Oral, drinking

water

Oral

lactation, in
utero

PND 25

0.78 [jg/dL for control
28.3 [jg/dL for 0.05%

PND 25: Histopathology

Zhu et al. (2019a)

Rat (Sprague Dawley) GD-10 to 12 mo

Oral, drinking

12 mo

12 mo: Electrophysiology



Control (distilled water),

water

0 [jg/dL for control





M, n = 10

Oral









lactation, in
utero

0.27 mg/L (27 pg/dL) for





0.5 g/L, M, n = 10

0.5 g/L



Zhana et al. (2015b)

Rat (Long-Evans)

GD-10 to PND 50 Oral, diet

PND 50

PND 50: Histopathology,



Control (0 ppm), M/F,

Oral

0.8 pg/dL for control

Electrophysiology



n = 10

lactation, in
utero

21.1 pg/dL for 1500 ppm





1500 ppm, M/F, n = 10







Wang, 2021,

Rat (Sprague Dawley)

GD -28 to PND 21 Oral,

PND 21:



10296633@@author-

Control (deionized

lactation





year

water), M/F, n = 12

In utero

23.9 pg/L (2.39 pg/dL) for







Control





0.05% solution, M/F,









n = 10



206 pg/L (20.6 pg/dL) for
0.05% solution



External Review Draft

3-252

DRAFT: Do not cite or quote

-------
Study	Species (Stock/Strain), TMngo, Exposure BLL As Reported (|jg/dL)	Endpolnts Examined

Wu et al. (2020a)

Mouse (C57BL/6J)

Control (ultra-pure
water), F, n = 8

200 mg/L, F, n = 8

Mouse (APP/PS1)

Control (ultra-pure
water), F, n = 8

200 mg/L, F, n = 8

GD -7 to 7 mo

Oral, diet
Oral

lactation, in
utero

PND 207-210

Mouse (C57BL/6J)

19.71 [igIL (1.97 pg/dL) for
control

84.53 [igIL (8.45 ng/dL) for
200 mg/L

Mouse (APP/PS1)

19.96 |jg/L(1.99 pg/dL) for
control

205.49 pg/L(20.54 pg/dL) for
200 mg/L

7 mo: Brain Weight, Histopathology

Manietal. (2020)

Rat (Wistar)

8 mo to NR

Oral, gavage

NR

NR: Histopathology, Brain Weight



Control





2.3 pg/dL for Control





10 mg/kg





8.5 pg/dL for 10 mg/kg





50 mg/kg





16.4 pg/dL for 50 mg/kg





100 mg/kg





16.3 pg/d for 100 mg/kg



External Review Draft

3-253

DRAFT: Do not cite or quote

-------
Study	Species (Stock/Strain), TMngo, Exposure BLL As Reported (|jg/dL)	Endpolnts Examined

Shvachiv et al. (2020)

Rat (Wistar)	GD 7 to 3 mo

Control (tap water), M/F,
n = 12	GD 7 to 5 mo

0.2% solution (3 mo), GD 7 to 7 mo
M/F, n = 12

0.2% solution (5 mo),
M/F, n = 12

0.2% solution (7 mo),
M/F, n = 12

Oral, drinking 3 mo:
water

°ra|.	<1 for Control

lactation

in utero 24.0 Mg/dL for 0.2% solution
5 mo:

<1 for Control

24.8	[jg/dL for 0.2% solution
7 mo:

<1 for Control

26.9	[jg/dL for 0.2% solution

External Review Draft

3-254

DRAFT: Do not cite or quote

-------
Study	Species (Stock/Strain), TMngo, Exposure BLL As Reported (|jg/dL)	Endpolnts Examined

PND 10	PND 30: Histopathology,

0.6 pg/dL for control	Electrophysiology

11.4 pg/dL for 109 ppm
PND 21

0.85 pg/dL for control
3.5 pg/dL for 109 ppm

PND 30

0.98 pg/dL for control
1.8 pg/dL for 109 ppm

APP = amyloid precursor protein; BLL = blood lead level; Dl = deionized; F = female; F0 = gestating female; GD = gestational day; LTP = long-term potentiation; M = male;
ME = maternal exposure; MRI = magnetic resonance imaging; mo = month(s); NaAc = sodium acetate; NR = not reported; Pb = lead; PG = pregestation; PND = postnatal day;
PW = postweaning; wk = week(s); yr = year(s).

Zhao et al. (2021)	Rat (Sprague Dawley) GD-14 to PND 10 Oral, drinking

Control, M, n = 6	water

109 ppm, M, n = 6

Oral,
lactation
In utero

External Review Draft

3-255

DRAFT: Do not cite or quote

-------
Table 3-2E Epidemiologic studies of Pb exposure and full-scale intelligence quotient.

Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

Lanphear et al. (20051
and Lanphear et al.
(20191

International pooled
analysis: Prospective
cohorts from Boston,
Cincinnati, Cleveland,
Mexico City, Port Pirie,
Rochester, and
Yugoslavia.

Followed from birth
(1979-1995) up to age
10 yr

n = 1,333 children

Blood

Median (5th-95th)

Early childhood (6-
24 mo):

12.7 (3.5-34.5)

Peak: 18 (6.0-47.0)

Lifetime avg (through
outcome measurement at
4.8-10 yr):

11.9 (3.6-34.5)

Concurrent: 9.7 (2.5-
33.2)

FSIQ: WISC-III, WISC-R, HOME score, birth weight, Early Childhood:

WPPSI, WISC-S	maternal IQ and

(depending on the cohort) education. Also
Ages 4 8-10 vr	considered potential

confounding by child sex,
birth order, marital status,
maternal age, prenatal
smoking status and
alcohol use.

-0.137 (-0.209, -0.064)

Lifetime avg:

-0.206 (-0.285, -0.126)

Concurrent:

-0.187 (-0.26, -0.114)

Peak:

-0.126 (-0.182, -0.071)

Lanphear et al. (20051
and Lanphear et al.
(20191, subset of with
peak BLLs <7.5 [jg/dL

n = 103 children

Same

Concurrent
Mean: 3.2

Same

Same

-2.53 (-4.48, -0.58

Canfield et al. (2003a1	n — 101

Rochester, NY	Children recruited from

Prospective cohort	dust control studV

Born 1994-1995 followed
from age 6 mo to 5 yr

Blood	FSIQ

Stanford-Binet

Concurrent, children with Age 5 yr
peak <10

Mean: 3.3

Child sex, Fe status, birth -1.8 (-3.0, -0.60)

weight, maternal race,

education, IQ, income,

and prenatal smoking

status, HOME score.

External Review Draft

3-256

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

Bellinger and Needleman n - 48 children
(20031	Recruited at birth

Boston area, MA.

Prospective

Followed from birth
(1979-1981) to age 10 yr.

Blood

Early childhood (age 2 yr)
Mean (SD)

Peak <10: 3.8 (range: 1-
9.3)

Detection limit NR

FSIQ: WISC-R
Age 10 yr

HOME score (age 10 and
5), child stress events,
race, maternal IQ, age,
marital status, SES, sex,
birth order, # residence
changes before age 5 yr.
Also considered potential
confounding by family
stress, maternal age,
psychiatric factors, child
serum ferritin levels

-1.6 (-2.9, -0.2)

Surkan et al. (2007)

n = 389

Blood

WISC-III

Caregiver IQ, child age, 1.0 (reference)

Boston, MA and

Children recruited from



Age 6-10 yr

SES, race, birth weight. _g 12 (-3 3 3 1)

Farmington, ME

trial of amalgam dental
fillings.

Concurrent

Also considered site, sex, „„ „

—6 0 (—11 — 1 4^
birth order, caregiver ' v ' '



Groupl: 1-2



education and marital

Cross-sectional



Group 2: 3-4



status, parenting stress,

Sep 1997-Mar 2005



Group 3: 5-10

Mean (SD):
2.2 (1.6)



and maternal utilization of
prenatal and annual
health care (not parental
caregiving quality.)

Chiodo et al. (20071
Detroit, Ml area

Cross-sectional

495 children (born 1989-
1991) age 7 yr,

Blood

Concurrent
Mean (SD): 5.0 (3.0)

WPPSI
Age 7 yr

Maternal

psychopathology, IQ,
prenatal smoking,
prenatal marijuana, SES,
HOME score, caretaker
education and marital
status, # children in
home, child sex. Also
considered child age,
maternal age, custody,
cocaine use, prenatal
alcohol use.

-0.19 (-0.30, -0.08)
Note: standardized
regression coefficient.
95% CI estimated using
the reported p-value of
0.01

External Review Draft

3-257

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

Kim et.a.1. (20091	261 children

Seoul, Seongnam, Ulsan,

and Yeoncheon, South School recruitment

Korea

Cross-sectional

Blood

Concurrent

Mean (SD): 1.7 (0.80)

Age: 8-11 yr

KEDI-WISC
Ages 8-11 yr

Blood Mn <1.4 [jg/dL
-2.4 (-6.0, 1.1)

Children born 1996-1999

Maternal age, education
and prenatal smoking
status paternal
education, yearly income,	Ha

smoking exposure status "3.2 (-6.1, -0.24)
after birth, child age, sex,
and birth weight (not
parental caregiving quality
or IQ)

tBraunetal. (2018)

Cincinnati, OH
United States
Mar2003-Jan2006

Followed for 8 yr
Cohort

HOME study
n: 355 (Intervention
group: 174, Control
group: 181)

Clinical trial of pregnant
women, mean gestation
of 16 wk and residence in
a house built in or before
1978

Intervention to reduce Pb
exposure

Blood	WPPSI

Maternal and child blood; Age at outcome:
ICP-MS	5-8 yr

Dust Pb loadings floor,
interior windowsill and
window at 20 wk
gestation, child age 1 and
2 yr; GFAAS.

Age at Measurement:

16, 26 wk of gestation,
delivery (maternal);

1,2,3,4,5, 8 yr (child)

Baseline GM (Intervention
and control groups):
maternal: 0.7 and
0.7 [jg/dL , floor dust Pb:

1.5 and 1.9 pg/sq ft,
windowsill dust Pb: 28
and 33 pg/sq ft, window
trough dust Pb: 574 and
510 |jg/sq ft.

NA

Mean FSIQ score
difference15: 0.5 (-3.3,
24.2), comparing the
treatment to the injury
prevention control group.

External Review Draft

3-258

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

t Taylor et al. (20171
UK

Cohort

April 1, 1991-Dec 31,
1992 (followed until age
4-8 yr)

ALSPAC
n: 4285

Mother-infant pairs.

Blood	FSIQ, VIQ, PIQ (WPPSI

or WISC-III).

Maternal and child venous
blood; ICP-MS	Age at outcome:

4-8 yr

Age at Measurement:

Prenatal (mean
gestational age 11 wk)
and postnatal (30 mo)

Prenatal: 3.67 [jg/dL;

Child BLL: 4.22 pg/dL.

Family adversity index,
housing tenure,
household crowding,
smoking in the first
trimester, alcohol
consumption in the first

WISC-Boysb: -0.29
(-1.02, 0.44)
WISC-Girlsb: 0.73 (0.13,
1.33)

WPPSI-Girlsb: -0.65
(-2.065, 0.765)

trimester, maternal age at WPSSI-Boysb: -0.54

index birth, parity,	(-2.015, 0.935)

maternal education,

length of time the mother

lived in Avon, child sex,

child age at testing,

weighted life events

score, and hemoglobin

level.

tTatsuta et al. (20201

Tohoku district, coastal
area Japan

2002-2006 (enrollment)
through 2015-2018 (12 yr
followup)

Cohort

TSCD coastal cohorts
n: 289 mother-child pairs
(singleton births); 148
boys and 141 girls

Blood

Cord and child venous
blood; ICP-MS

Age at Measurement:
Delivery (cord), 12 yr
(child)

FSIQ (Japanese version
WISC-IV), age equivalent
ranking and scores for
verbal comprehension,
perceptual reasoning,
working memory, and
processing speed
composites; BNT (cues
and no cues)

Median: Cord - 0.8 pg/dL, Age at outcome:
12-yr = 0.7 pg/dL	12 yr

95th: Cord: 1.4 pg/dL, 12-
yr: 1.1 pg/dL

Birth weight, drinking or
smoking during
pregnancy, the Raven's
score (parent assessment
for child at 18 mo of age),
passive smoking status at
12 yr old, family income,
WISC/BNT tester, and
cord blood total Hg

Cord-Boys: (3=—3.683 i
Cord-Girls: (3=10.714,
3.349)

Child-Boys: (3=1.463
(-2.905, 5.831)

Child-Boys: (3=-9.88
(-18.977, -0.782)

Child-Girls: -4.406
(-15.94, 7.129)

External Review Draft

3-259

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tDesrochers-Couture et
al. (70181

10 study sites
Canada

2008-2011

Followed 3-4 yr from birth
Cohort

MIREC Study	Blood

n: 609

Maternal, cord, and
Birth cohort: Mother-infant postnatal child (venous)
pairs recruited during 1st ICP-MS
trimester	Age at Measurement:

Maternal (6-13 wk, 32-
34 wk), birth (cord); and
3-4 yr (postnatal child)

GM: 1st trimester:

0.62 [jg/dL
0.59 [jg/dL
0.76 [jg/dL
0.70 [jg/dL
Max: 1st trimester:

3rd trimester:
cord blood:
child blood:

4.14 [jg/dL
3.93 [jg/dL
3.52 [jg/dL
5.49 [jg/dL

FSIQ, VIQ, PIQ, General
Language composite

WPPSI-3rd Edition, short
version. The age-
standardized WPPSI-111
Canadian norms
were used to calculate the
scores.

Age at outcome:

Between 2 yr 6 mo and
3 yr 11 mo

3rd trimester:
cord blood:
child blood:

Cord blood model: child
age, child sex, maternal
education, evaluation site
and cord blood Hg (log—2
scale).

Child blood model: child
age, child sex, evaluation
site, marital status,
familial income, HOME
total score, Parenting
Stress Index, and cord
blood Pb (log—2 scale).

Cord: (3=-0.123 (-0.251,
0.005)

Child: (3=0.027 (-0.135,
0.188)

Cord-Boys: (3=—5.686
(-9.968, -1.405)
Cord-Girls: (3=0.287
(-3.787, 4.361)

tZhou et al. (2020b)

Jiangsu Province
China

June 2009-Jan 2010 to
June 2016-July 2017

Cohort

Sheyang Mini Birth
Cohort Study
n: 296

Blood, Urine

FSIQ, VIQ, PIQ (Chinese Sex, maternal age,

Cord blood tested for Mn,
Cd and Pb using GFAAS.
Birth cohort- mother-infant Postnatal urine samples
pairs from an agricultural urine also tested for the
region.	elements.

Age at Measurement:
Cord; postnatal Urine NR

GM: cord blood:
15.88 |jg/L, urine:
1.43 |jg/L,

75th: cord blood:
21.83 |jg/L, urine:
2.27 |jg/L,

Max: cord blood:

1168.20 |jg/L, urine:
62.47 |jg/L,

version WISC-R).

Age at outcome:
6-7 yr (school-aged
children)

maternal education,
family annual income,
family inhabitation area,
and passive smoking;
multiple effects were also
assessed by entering all
other metals in the model.
Sex-stratified analysis
conducted.

Cord-Girls: 0.615 (-0.909,
2.138)

Cord-Boys: 0.835
(-1.164, 2.833)

Cord-All: 0.67 (-0.514,
1.854)

External Review Draft

3-260

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tLiu et al. (20151

Chongqing, China

March 4, 2003-June 19,
2003 (enrollment)
Followed 5 yr

Cohort

Birth cohort with mother-
infant pairs
n: 149

Mothers gave birth in 4
hospitals in Tongliang
county.

Blood

VIQ, PIQ, and FSIQ
(Shanghai version
WPPSI)

Maternal and cord serum
(Hg, Cd, Pb); Pb
determined with AAS.

Ratio of maternal to cord
serum levels (i.e.,
placental transport ratio of equivalent percentile
metal) estimated.	ranks were estimated.

Raw scores were
converted to composite
scores and age

Age at Measurement:
Delivery

Age at outcome:
5 yr

Maternal age, educational
level, vitamins, placental
transport ratios, maternal
exposure to ETS. Pb, Hg
and Cd were considered
in multivariable models
(covariates retained
based on evaluation of
VIFs).

Final predictive model did
not include associations
with Pb

Mean: 3.45 [jg/dL.

External Review Draft

3-261

DRAFT: Do not cite or quote

-------
fWang_eiaL12022} n: 148

Wujiang
China

Birth cohort established
from 2009-2010; follow-
up from 2016-2017.

Cohort

Blood

Blood Pb measured via
AAS.

Age at Measurement:
Mean, SD (mo): 89.90,
3.77

Cord blood:

GM = 28.26 |jg/L,
Median = 27.56 |jg/L;
Venous blood:
GM = 22.99 |jg/L,
Median = 23.80 |jg/L
75th: Cord blood:
38.42 |jg/L; Venous
blood: 33.00 |jg/L
Max: Cord blood:
249.00 |jg/L; Venous
blood: 71.40 pg/L.

Cognitive Effects in
children - FSIQ

Child health personnel
conducted the WISC-CR
for FSIQ, PIQ, and VIQ.
Scores were age
converted and
standardized.

Age at outcome:

General linear models
adjusted for child sex,
maternal age at delivery,
age of children, maternal
education level, paternal
education level, monthly
household income, parity,
inhabitation area, passive
smoking.

"Table 3, Beta (95% CI):

VIQ, total population, cord
blood:

Q2 vs. Q1: -0.296
(-7.005, 6.413)

Q3 vs. Q1: 4.468 (-2.840,
11.776)

Q4 vs. Q1: -1.275
(-8.231, 5.682)

VIQ, boys, cord blood:

Q2 vs. Q1: -2.752
(-12.594, 7.091)

Q3 vs. Q1: 5.681 (-6.288,
17.649)

Q4 vs. Q1: 2.119 (-8.913,
13.150)

VIQ, girls, cord blood:

Q2 vs. Q1: 5.679 (-5.242,
16.600)

Q3 vs. Q1: 6.510 (-4.501,
17.521)

Q4 vs. Q1: -1.331
(-11.933, 9.271)

VIQ, total population,
venous blood:

Q2 vs. Q1: 4.942 (-3.957,
13.841)

Q3 vs. Q1: -0.536
(-9.560, 8.487)

External Review Draft

3-262

DRAFT: Do not cite or quote

-------
External Review Draft

Q4 vs. Q1: -3.304
(-12.117, 5.509)

VIQ, boys, venous blood:

Q2 vs. Q1: 5.592 (-7.193,
18.378)

Q3 vs. Q1: 8.858 (-4.271,
21.988)

Q4 vs. Q1: 7.143 (-5.649,
19.935)

VIQ, girls, venous blood:

Q2 vs. Q1: 3.179
(-10.810, 17.169)

Q3 vs. Q1: -13.548
(-27.506, 0.411)

Q4 vs. Q1: -14.964
(-28.412, -1.517), p-
value = 0.036

PIQ, total population, cord
blood:

Q2 vs. Q1: -5.584
(-14.011, 2.842)

Q3 vs. Q1: -0.441
(-9.620, 8.738)

Q4 vs. Q1: -9.365
(-18.103, -0.628), p-
value = 0.038

PIQ, boys, cord blood:

Q2 vs. Q1: -13.080
(-24.907, -1.254), p-

DRAFT: Do not cite or quote

-------
External Review Draft

value = 0.035

Q3 vs. Q1: -9.686
(-24.067, 4.695)

Q4 vs. Q1: -7.592
(-20.848, 5.663)

PIQ, girls, cord blood:

Q2 vs. Q1: -0.002
(-14.192, 14.188)

Q3 vs. Q1: 4.023
(-10.284, 18.331)

Q4 vs. Q1: -13.293
(-27.069, 0.483)

PIQ, total population,
venous blood:

Q2 vs. Q1: -7.293
(-17.605, 3.020)

Q3 vs. Q1: -8.176
(-18.633, 2.281)

Q4 vs. Q1: -4.507
(-14.720, 5.706)

PIQ, boys, venous blood:

Q2 vs. Q1: -8.218
(-22.276, 5.841)

Q3 vs. Q1: -6.701
(-21.138, 7.737)

Q4 vs. Q1: -4.294
(-18.360, 9.772)

PIQ, girls, venous blood:

DRAFT: Do not cite or quote

-------
External Review Draft

Q2 vs. Q1: -10.417
(-29.104, 8.270)

Q3 vs. Q1: -16.397
(-35.043, 2.248)

Q4 vs. Q1: -6.994
(-24.957, 10.968)

FSIQ, total population,
cord blood:

Q2 vs. Q1: -3.369
(-10.711, 3,974)

Q3 vs. Q1: 2.396 (-5.603,
10.394)

Q4 vs. Q1: -6.087
(-13.700, 1.527)

FSIQ, boys, cord blood:

Q2 vs. Q1: -8.599
(-19.015, 1.818)

Q3 vs. Q1: -1.434
(-14.100, 11.232)

Q4 vs. Q1: -2.552
(-14.227, 9.123)

FSIQ, girls, cord blood:

Q2 vs. Q1: 2.986 (-9.126,
15.098)

Q3 vs. Q1: 5.743 (-6.469,
17.955)

Q4 vs. Q1: -8.635
(-20.394, 3.123)

FSIQ, total population,

DRAFT: Do not cite or quote

-------
RefereDCeesignnd	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

venous blood:

Q2 vs. Q1: -0.950
(-10.606, 8.706)

Q3 vs. Q1: -4.930
(-14.722, 4.861)

Q4 vs. Q1: -4.773
(-14.336, 4.790)

FSIQ, boys, venous
blood:

Q2 vs. Q1: -1.367
(-14.675, 11.941)

Q3 vs. Q1: 1.366
(-12.300, 15.033)

Q4 vs. Q1: 1.929
(-11.386, 15.243)

FSIQ, girls, venous blood:

Q2 vs. Q1: -3.771
(-20.071, 12.529)

Q3 vs. Q1: -17.326
(-33.590, -1.062), p-
value = 0.044

Q4 vs. Q1: -13.625
(-29.293, 2.0"3)"

External Review Draft

3-266

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tlglesias et al. (20111

Antofagasta,

Northern Chile
1998-2005

Cohort

Sepulveda study
n: 192

Children lived or attended
school in area
contaminated by Pb
mineral concentrate
stored in open sites at
railroad terminal (closed
in 1998).

Blood

Child's venous blood;
AAS.

Age at Measurement:
In 1998 at 0-7 yr old;
2005 at 7-16 yr old

FSIQ, VIQ, PIQ (Chilean
version of WISC-R).

Age at outcome:

7-16 yr

in

Mean and Median: blood
Pb 1998: 10.8 and
10 [jg/dL; blood Pb 2005:
3.5 and 3.2 [jg/dL
75th: blood Pb 1998:
14 [jg/dL; blood Pb 2005:
4.3 [jg/dL

Max: Blood Pb 1998:
33 [jg/dL; blood Pb 2005:
14 [jg/dL.

Sex, birth weight, birth
order, # of siblings, milk
type during the first six
months of life, history of
anemia, SES (household
income, home ownership,
and school type: public or
private), parental
education, maternal
smoking during
pregnancy, maternal IQ,
children's stimulation at
home, HOME score.

Concurrent: -0.94 (-1.77,
-0.11)

Early childhood: -0.14
(-0.445, 0.165)

tRuebner et al. (20191

46 centers
U.S.

Cohort

3 enrollment periods,
2005-2009, 2011-2014,
2016-2020

Followed up to 9 yr

CKiD Cohort study
n: 412

Children with mild to
moderate CKD

Blood

Child venous blood; ICP-
MS. The BLL
measurement closest to
the time of neurocognitive
testing was used for
analysis (concurrent).

Age at measurement:
NR; 2, 4, or 6 years after
study entry

Median: 1.2 (ig/dL
75th: 1.8 (ig/dL
Max: 5.1 (ig/dL

FSIQ

Mullen Scales of Early
Learning (age 12-29 mo),
WPPSI (30 mo-5 yr), and
WAS I (6-18 yr).

The last available test
results were to evaluate
long-term effects. Mean
time between BLL and
neurocognitive testing
was 2.3 yr.

Age at outcome:

1-16 yr

Age, sex, race, poverty,
and maternal education

Concurrent: (3=-2.1
(-3.95, -0.25)

External Review Draft

3-267

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tLee et al. (20211

Seoul, Gyeonggi, and
Incheon provinces
South Korea

2008-2017 (Recruitment
2008-2010; follow-up
from 2012-2017)

Cohort

Environment and
Development of Children
n: 502

Blood

Whole blood Pb from
mothers during their
second trimester of
pregnancy and children at
ages 4y and 6y were
analyzed by atomic
absorption
spectrophotometry.

Age at Measurement:
Maternal mean age
(SD) = 31.3 (3.5) yr.
Children at 4yo and 6yo.

Prenatal GM (SD) = 1.32
(1.32) [jg/dL; children at
4yo = 1.43 (1.38) pg/dL;
children at 6yo = 1.43
(1.35) pg/dL
75th:

Prenatal = 1.56 pg/dL;
children at

4yo = 1.72 pg/dL; children
at 6yo = 1.70 pg/dL
95th:

Prenatal = 2.11 pg/dL;
children at

4yo = 2.47 pg/dL; children
at 6yo = 2.28 pg/dL.

Cognitive Effects (FSIQ)

Intelligence quotients of
children assessed using
'KEDI-WISC.

Age at outcome:

Multivariate models
adjusted by maternal
education level, exposure
to ETS during the
pregnancy, maternal age,
and maternal IQ.

Table-3 - Estimated
coefficients and 95% CI of
associations between
single metals and
children's IQ at 6-years
old (standardized)

Prenatal period: -1.202
(-4.87, 2.467)

At age 4: -1.829 (-4.664,
1.006)

At age 6: -2.614 (-5.623,
0.396)

tDantzer et al. (20201

Greater Cincinnati,
Metro area
U.S.

Ohio

CCAAPS
n: 344

Cohort recruited at birth
Oct 2001 —Jul 2003

Cross-sectional analysis
of data collected at age
12

Blood, nails

Postnatal child venous
blood, toenail; ICP-MS.
Mean blood Pb:
0.57 pg/dL; toenail Pb:
0.66 pg/g; information
also available by gender
and race

Age: 12 yr.

FSIQ (WISC-IV)
Age at outcome:
12 yr

Caregiver IQ, community
deprivation index, and
BMI. Sex considered as a
potential confound.

Concurrent (blood):
B=-10.871 (-16.893,
-4.848)

B=-1 .70 (-4.27, -0.862)

External Review Draft

3-268

DRAFT: Do not cite or quote

-------
tMartin et al. (20211

East Liverpool, Ohio
United States
2013-2014
Cross-Sectional

CARES
n: 66

BLLs from the children
were analyzed by ICP-
MS.

Age at Measurement:
Mean (SD) = 8.4 (0.9) yr

GM (SD) = 1.13
(1.96) [jg/dL
Max: 6.64 [jg/dL.

Cognitive Effects

Cognitive performance
was assessed using the
WISC-IV.

Age at outcome:

Regression models were
adjusted for sex, income,
and In (serum cotinine).

"Table 3 - Interaction
effects between Ln Blood
Pb-Ln Hair Mn, 13 (95%
CI)

Blood Pb (per 1 In [jg/dL
difference)

FSIQ

At In hair Mn = 5 ng/g:
1.686 (-3.039, 6.412)

At In hair Mn = 6.25 ng/g:
-4.447 (-8.333, -0.561)

At In hair Mn = 7 ng/g:
-8.133 (-13.4, -2.867)

At In hair Mn = 7.5 ng/g:
-10.596 (-17.172, -4.02)

Perceptual Reasoning

At In hair Mn = 5 ng/g:
2.392 (-4.055, 8.839)

At In hair Mn = 6.25 ng/g:
-4.612 (-9.918, 0.694)

At In hair Mn = 7 ng/g:
-8.808 (-15.996, -1.62)

At In hair Mn = 7.5 ng/g:
-11.616 (-20.592,
-2.639)

Processing Speed

At In hair Mn = 5 ng/g:
-0.62 (-4.263, 3.024)

At In hair Mn = 6.25 ng/g:
-2.565 (-5.565, 0.435)

External Review Draft

3-269

DRAFT: Do not cite or quote

-------
RefereDCeesignnd	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

At In hair Mn = 7 ng/g:
-3.725 (-7.788, 0.337)

At In hair Mn = 7.5 ng/g:
-4.502 (-9.576, 0.573)

Verbal Comprehension

At In hair Mn = 5 ng/g:
1.208 (-3.451, 5.867)

At In hair Mn = 6.25 ng/g:
-4.141 (-7.976, -0.306)

At In hair Mn = 7 ng/g:
-7.349 (-12.549, -2.149)

At In hair Mn = 7.5 ng/g:
-9.49 (-15.976, -3.004)

Working Memory

At In hair Mn = 5 ng/g:
2.376 (-2.404, 7.157)

At In hair Mn = 6.25 ng/g:
-2.133 (-6.067, 1.8)

At In hair Mn = 7 ng/g:
-4.831 (-10.161, 0.498)

At In hair Mn = 7.5 ng/g:
-6.635 (-13.286, 0.0"6)"

External Review Draft

3-270

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tHavnes et al. (20151

Marietta or Cambridge,
Ohio, and surrounding
communities
U.S.

Oct 2008-March 2013
Cross-sectional

CARES
n: 404

Participants resided in
study area throughout
their life; not moving for at
least 1 yr

Blood

Child venous blood: ICP-
MS

Age at Measurement:
7-9 yr

GM: 0.82 pg/dL
Max: NR.

FSIQ (WISC-IV), 4
domains of intellectual
functioning (reasoning,
processing speed,
working memory and
verbal comprehension)

Age at outcome:
7-9 yr

Sex, parent's IQ, parent Processing speed: (3 = -
education, parent	3.53 [-6.95, -0.12)

confidence T-score, Mn or

Pb, community residence Association with FSIQ NR
(FSIQ models only); other
sets of variables added
depending on domain.

(Note: main effect is Mn)

tHong et al. (20151

5 administrative regions
South Korea
NR

Cross-sectional

n: 1001

General population of
children

Blood

Venous blood; GFAAS
Age at Measurement:
8-11 yrold

Median: 1.81 pg/dL
75th: 2.25 pg/dL,
95th: 3.01 pg/dL
Max: 6.16 pg/dL.

KEDI-WISC

Age at outcome:
8-11 yrold

Age, sex, residential
region, paternal education
level, and yearly income
Iog10-transformed blood
Hg, Mn, urine
concentrations of cotinine,
phthalate metabolites

B=-1.948 (-3.608,
-0.288), adjusted for
other metals

B=-2.113 (-3.73,
-0.496), adjusted for
ADHD and CPT
B=-2.118 (-3.792,
-0.445), adjusted for
socio-demographic
factors

tMenezes-Filho et al.
(20181

Salvador, Bahia
Brazil

Cross-sectional
Study years: NR

School-based cohort
n: 225

Children from 4
elementary schools in
industrial town.

Blood

Child venous blood Pb,
hair and toenails tested
for Mn; GFAAS

Age at Measurement:
7-12 yr

Mean: 1.64 pg/dL,
Median: 1.15 pg/dL, only
about 2% of children
above the Centers for
Disease Control and
Prevention ref value of
5 pg/dL

75th: 2.1 pg/dL
Max: 15.6 pg/dL.

IQ estimated using
vocabulary and matrix
reasoning (WASI).

Age at outcome:
7-12 yr

Age, Maternal IQ

Effect modification by Mn
assessed. Both Pb and
Mn were log-transformed
to include in the model.
Interaction between Pb
and Mn assessed.

Put results for Model B
which is adjusted for Mn,
age and maternal IQ

External Review Draft

3-271

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tLucchini et al. (2012)

Junior high school-age

Blood

IQ tested using WISC-III

Sex, age at testing,

B=-2.248 (-4.111,

children from 20 local



(verbal IQ and

parental education, SES,

-0.385)

Valcamonica and Garda

public schools

Children's venous blood;

performance IQ

family size, parity order,



Lake areas in Province of

n: 299

GFAAS

assessed).

BMI



Brescia











Italy



Age at Measurement:

Age at outcome:





Cross-sectional



11-14 yr

11-14 yr









1.71 pg/dL, Median: 1.50











75th: 2.10 pg/dL











Max: 10.2 pg/dL.







AAS = atomic absorption spectroscopy; ALSPAC = Avon Longitudinal Study of Parents and Children; avg = average; BLL = blood lead level; BNT = Boston Naming Test;

CARES = Communities Actively Researching Exposure Study; CCAAPS = Cincinnati Childhood Allergy and Air Pollution Study; Cd = cadmium; CI = confidence interval;
CKD = chronic kidney disease; CKiD = Chronic Kidney Disease in Children Study; ETS = environmental tobacco smoke; FSIQ = full-scale intelligence quotient; GFAAS = graphite
furnace atomic absorption spectrometry; GM = geometric mean; Hg = mercury; HOME = Health Outcomes and Measures of the Environment; ICP-MS = inductively coupled plasma
mass spectrometry; IQ = intelligence quotient; KEDI = Korean Educational Development Institute; MIREC = Maternal-Infant Research on Environmental Chemical; Mn = manganese;
mo = month(s); NA = not available; NR = not reported; Pb = lead; PC = primary caregiver; PIQ = performance IQ; Q = quartile; SD = standard deviation; SES = socioeconomic status;
VIF = variance inflation factor; VIQ = verbal IQ; WASI = Wechsler Abbreviated Scale of Intelligence; WISC = Weschler Intelligence Scale for Children; wk = week(s);

WPPSI = Wechsler Preschool and Primary Scale of Intelligence yr = year(s).

aEffect estimates are standardized to a 1 |jg/dL increase in BLL or a 10 |jg/g increase in bone Pb level, unless otherwise noted. For studies that report results corresponding to a
change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the biomarker and standardized accordingly.

bResults are not standardized (e.g., BLL distribution data needed to calculate the standardized estimate was not reported or categorical data was analyzed).

fStudies published since the 2013 Integrated Science Assessment for Lead.

External Review Draft

3-272

DRAFT: Do not cite or quote

-------
Table 3-3E Epidemiologic studies of Pb exposure and infant development.

Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

Bellinger et al. (1987)

Boston, MA
U.S.

Apr. 1979 - Apr. 1981
(enrollment)

Followed through 2 yr
Cohort

Birth cohort, n = 182
infants

Recruitment from births
at Brigham and Women's
Hospital

Blood

Cord blood; anodic stripping
voltammetry (ASV)

Age at measurement:
Delivery

Mean (SD): 6.6 (3.2) (ig/dL
Low: <3 [jg/dL
Medium: 6-7 [jg/dL
High: >10 |jg/dL

MDI assessed using
BSID-II

Age-standardized
scores (mean: 100,
SD: 16)

Age at outcome: 2 yr

Maternal age, race, IQ,
education, years of
smoking, and alcohol
drinks/wk in 3rd
trimester, SES, HOME
score, child sex, birth
weight, gestational age,
birth order.

Beta:

Cord blood:

Low vs. high: -3.8 (-6.3,
-1.3)

Medium vs. high: -4.8
(-7.3, -2.3)

Concurrent blood
reported not to be
associated with MDI,
quantitative data not
reported.

Jedrvchowski et al.
(2009b)

Krakow, Poland

2001-2004 (enrollment)
Followed through 3 yr

Cohort

Birth cohort, n = 381-415 Blood
children

Recruited pregnant
mothers from prenatal
clinics in Krakow inner
city in 1st and 2nd
trimesters.

Cord blood; ICP-MS

Age at measurement:
Delivery

GM (95% CI): 1.29 (1.24,
1.34) (ig/dL
Median: 1.23 [jg/dL

MDI assessed using
BSID-II (Polish
version)

Standardized scores

Age at outcome:
12, 24, 36 mo

Maternal education and
prenatal smoking, child
sex and birth order

Beta

Age 2 yr: -1.8 (-3.4,
-0.14)

Age 3yr: -1.6 (-2.9,
-0.21)

Henn etal. (2012)

Mexico City
Mexico

1997-2000 (enrollment)

Followed through 24
months

N: 455

Blood

MDI assessed using Sex, gestational age,

Women recruited during Child venous blood; ICP-MS
pregnancy or at delivery

Age at measurement:
12, 24 mo

Mean (SD):

12 months: 5.1 (2.6) pg/dL

BSID-II (Spanish
version)

Age at Outcome:

12, 18, 24, 30, 36
mo

hemoglobin, maternal IQ,
maternal education, and
visit

Beta

12-months:
0.25)

24-months:
0.30)

-0.07 (-0.39,

-0.08 (-0.46,

12 mo Pb * Mn <2 pg/dL:
-0.31 (-1.25, 0.62)

24 mo Pb * Mn <2 pg/dL:
-1.27 (-2.18, -0.37)

External Review Draft

3-273

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

Cohort



24 months: 5.0 (2.9) pg/dL







Hu et al. (2006)

Mexico City
Mexico

1997-2000 (enrollment)

Followed through 24
months

Cohort

N: 83 (cord)-146 (24- Blood
months child blood)

Women recruited during
pregnancy or at delivery

MDI assessed using Maternal age and IQ,

Maternal blood, cord blood,
and child venous blood; ICP-
MS

Age at measurement: T1, T2,
T3 (maternal), delivery (cord),
12, 24 mo (child)

Mean (SD):

Maternal Tl: 7.07 (5.10)
Hg/dL; T3: 6.86 (4.23) ng/dL
Cord: 6.20 (3.88) pg/dL

Child 24 mo: 4.79 (3.71)
pg/dL

BSID-II (Spanish
version)

Age at Outcome:
24 months

child sex, current weight,
height-for-age Z score,
and concurrent blood Pb
(in models examining
prenatal blood Pb)

Maternal T1: -0.76
(-1.50, -0.03)

Maternal T3: -0.43
(-1.10, 0.27)

Cord: -0.06 (-0.87, 0.74)

Child 24 mo: -0.23
(-0.92, 0.45)

t Y Ortiz et al. (2017)

Mexico City
Mexico

Jul 2007-Feb 2011
Followed through 24 mo

Cohort

PROGRESS birth cohort
n: 536

Women <20 wk of
gestation and planning to
reside in Mexico City for
the next 3 yr.

Blood

Maternal blood; ICP-MS.

Age at measurement:
T2, T3

Mean:

T2: 3.7 pg/dL
T3: 3.9 pg/dL.

Cognitive and

language

development

development

assessed using

BSID-III.

Standardized scores
(mean: 100, SD:
15). Cognitive,
language and motor
scores were jointly
considered for
standardizing.

Age at outcome:
24 mo

Infant sex, birth weight,
gestational age, maternal
age, maternal IQ (WAIS
Spanish version), HOME
score.

Beta

Cognitive
Development:

T2: 0.76 (-3.35, 4.87)bc
T3: -6.60 (-13.49,
0.29)bc

Stress2: -0.23 (-0.45,
-0.01 )bc

T3*Stress: 1.02 (-0.78,

2.82)bc

Language

Development:

T2: 0.97 (-3.18, 5.12)bc
T3: -6.00 (-12.94,
0.94)bc

External Review Draft

3-274

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tKim et al. (2013c) and MOCEH study
Kimetal. (2013b)

Seoul,
Ulsan
Korea

Cheonan and

2006-2010

Followed through 6 mo
Cohort

n: 884

Mothers recruited before
20th wk of pregnancy
between and were in
locations (Seoul,
Cheonan and Ulsan).

Blood

Maternal venous blood;
GFAAS with Zeeman
background correction,
measured for Pb amd Cd

Age at measurement:

Early (<20 wk) and late
pregnancy (med = 39 wk)

Early pregnancy: 1.4 (GM),
2.1 (90th), 9.8 (max) pg/dL

Late pregnancy: 1.3 (GM); 2.1
(90th), 4.3 (max) pg/dL

MDI assessed using
BSID-II (Korean
version)

Age-standardized
scores (mean: 100,
SD: 15).

Age at outcome:
6 mo

Birth weight, infant sex,
maternal age and
education, family income,
breastfeeding status,
residential area.

Beta
Early:

Overall: 0.02 (-1.20,
1.24)

Cd <1.47 pg/L: 2.44
(0.04, 4.83)

Cd >1.47 pg/L: -0.87
(-2.52, 0.78)

Late:

Overall: -1.74 (-3.37,
-0.12)

Cd <1.51 pg/L: -0.29
(-2.88, 2.30)

Cd >1.51 pg/L: -3.20
(-5.35, -1.06)

tKim et al. (2018b)

4 cities: Seoul, Anyang,
Ansan and Jeju
Korea

2011-2012 (enrollment)
Followed through 24 mo

Cohort

CHECK cohort
n: 140

birth cohort- pregnant
women recruited from 4
cities in Korea before
delivery.

Blood

Maternal and cord blood;
method NR

Age at measurement:
Delivery

Median (IQR):

Maternal: 2.7 (3.5, 5.7) pg/dL
Cord: 1.2 (0.8, 1.7) pg/dL

MDI assessed using
BSID-II (Korean
version)

Age at outcome:
13-24 mo

BPA, and phthalates,
maternal age
(continuous), birth
delivery mode
(categorical), monthly
household income
(categorical), child's sex,
and BDI (continuous) of
the mother, gestational
age (continuous),
primiparous (categorical),
and

pre-pregnancy BMI
(categorical).

Associations of blood Pb
concentrations and MDI
were assessed but not
reported because they
lacked statistical
significance.

tValeri etal. (2017)

Birth cohort

Blood

Cognitive and

Child sex, age attesting,

Beta



n: 825 (Pabna: 409,



language

maternal age and

Pabna

Pabna and Sirajdikhan

Sirajdikhan: 416)

Cord blood; ICP-MS,

development using

education, maternal IQ,

Cognitive: 0.012 (-0.05,
0.074)

districts



measured for Pb, As, and Mn

BSID-III (Bengali

HOME score, ETS,

Bangladesh

Mother-infant pairs,



version, adapted for

protein intake.

2010-2013 (enrollment)



Age at measurement:

rural Bangaldesh)





External Review Draft

3-275

DRAFT: Do not cite or quote

-------
Referenc^and Study study Population	Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

Followed through 20-
40 mo

Cohort

pregnant women enrolled Delivery
in the first trimester.

GM: Pabna: 1.8 [jg/dL,
Sirajdikhan: 6.0 [jg/dL.
75th: Pabna: 2.4, Sirajdikhan:
9.7

Max: Pabna: 79.1 [jg/dL,
Sirajdikhan: 36.0 [jg/dL.

Two primary
outcomes derived by
summing across raw
scores of cognitive
and language
development. Z-
scores were
calculated.

Language: -0.014
(-0.076, 0.048)

Sirajdikhan

Cognitive: -0.011
(-0.024, 0.001)

Language: -0.004
(-0.026, 0.017)

Age at outcome:
20-40 mo

tKoshv et al. (2020)

Old Town, Salavanpet
and neighboring areas in
Vellore, South India

Mar 2010-Feb 2012
(enrollment)

Followed through 5 yr

Cohort

Etiology, Risk Factors
and Interactions of
Enteric Infections and
Malnutrition and the
Consequences for Child
Health and Development
(MAL-ED) Network
n: 228 (followed 2 yr) and
212 children (followed
5 yr)

Birth cohort of mother-
infant pairs in eight
adjacent urban slum
dwelling areas

Blood

Child venous blood; GFAAS.
Mean BLL derived by
averaging BLLs at 15 and
24 mo for analysis at 2 yr, and
15, 24 and 36 mo for analysis
at 5 yr.

Age at measurement:
7, 15, 24 and 36 mo

Mean: 15 mo: 0.5 pmol/L,
24 mo: 0.6 pmol/L, 36 mo:
0.6 pmol/L

Cognitive and
language
development
assessed using
BSID-III (culturally
adapted and
translated)

Raw scores of
cognition and
expressive and
receptive language
domains.

Age at outcome:
24 mo

Child sex, maternal
intelligence raw scores,
SES, mean body Fe
levels.

Beta

Cognitive: -0.2 (-0.2,
-0.03)

Expressive language:
-0.2 (-0.3, -0.1)

Receptive language:
-0.04 (-0.1, 0.02)

tShekhawat et al. (2021) n: 117

Western Rajasthan
India

2018-2019 (enrollment)

Followed through 6.5 mo
(average)

Mother-child pairs in third
trimester or at delivery

Blood

Cord blood; ICP-OES

Age at measurement:
Delivery

GM = 4.14 [jg/dL;
mean = 4.77 ± 3.3 [jg/dL;

Cognitive and
language
development
assessed using
BSID-III

Age at outcome:
6.5 mo (average)

Maternal age, gravida,
gestational age, maternal
education, child sex and
weight, preterm birth,
maternal food intake
during pregnancy,
smoking, alcohol
consumption, maternal
residential and

13 (95 % CI)
(A) Umbilical cord Pb
level <5 [jg/dL (n = 70)
Composite cognitive:
0.19 (-0.03, 0.34)
Composite language:
0.21 (-0.23, 0.42)
Subscale receptive
language: 0.11 (-0.6,
1.66)

Subscale expressive

External Review Draft

3-276

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

Cohort

median = 4.23 [jg/dL
75th: 5.1 |jg/dL.

occupational history,
delivery type.

language: 0.22 (-0.03,
1.87)

(B) Umbilical cord Pb
level = 5.0-10.5 ug/dL
(n = 47)

Composite cognitive:
-0.13 (-0.77, 0.28)
Composite language:
-0.05 (-0.7, 0.47)
Subscale receptive
language: -0.04 (-3.5,
2.5)

Subscale expressive
language: 0.04 (-3.8,
2.9)

tParaiuli et al. (2015a) Birth cohort from
Bharatpur General
Hospital
n: 100

Chitwan, Bharatpur

District

Nepal

Sep-Oct 2008
(enrollment)

Followed through 24 mo

Resided in area for at
least 2 yr delivered at
term (i.e., >37 wk).

Blood

Cord blood; ICP-MS,
measured for Pb, As and Zn

Age at measurement:
Delivery

Median: 2.06 [jg/dL
Max: 22.08 pg/dL.

MDI assessed using Maternal age and

BSID-II

Age at outcome:
24 mo

education, BMI,
gestational age, family
income, parity, birth
weight, weight at 24 mo,
child age assessment,
As, Zn, HOME score
(smoking and alcohol
consumption not included
given low prevalence).

Beta

-4.21 (-13.62, 5.20)c

Cohort

tParaiuli et al. (2015b)

Chitwan, Bharatpur

district

Nepal

Sep-Oct 2008
(enrollment)

Followed through 36 mo
Cohort

Birth cohort from
Bharatpur General
Hospital
n: 100

Resided in area for at
least 2 yr delivered at
term (i.e., >37 wk).

Blood

Cord blood; ICP-MS,
measured for Pb, As and Zn

Age at measurement:
Delivery

Median: 2.06 pg/dL
Max: 22.08 pg/dL.

MDI assessed using Maternal age and

BSID-II

Age at outcome:
36 mo

education, BMI,
gestational age, family
income, parity, birth
weight, weight at 24 mo,
child age at assessment,
As, Zn, HOME score
(smoking and alcohol
consumption not included
given low given low
prevalence).

Beta

4.05 (-3.21, 11.31)

External Review Draft

3-277

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

t Zhou etal. (2017)

Shanghai
China

2010-2012 (enrollment)
Followed through 24-
36 mo

Cohort

Shanghai Stress Birth
Cohort Study
n: 139

Women enrolled in
prenatal clinics of
maternity hospitals
during mid-to-late
pregnancy.

Blood

Maternal blood; AAS

Age at measurement:
28-36 wk of gestation

GM (95% CI): 3.30 (3.05,
3.57) [jg/dL.

Language
development
assessed using
GDS (Chinese
version)

Age at outcome:
24-36 mo

Maternal age at
enrollment, economic
status, maternal
education, gestational
week, child sex, birth
weight and age.

Beta per log—10
transformed BLL

Language development
Overall: -6.76 (-17.29,
3.77)d

Low stress: -1.76
(-13.03, 9.51 )d
High stress: -33.82
(-60.04, -7.59)d

tViqeh etal. (2014)

Tehran
Iran

October 2006 - March
2011

Followed through 36 mo
Cohort

Birth cohort
n: 174

Mother-infant pairs
recruited in first trimester
(8-12 wk).

Blood

Maternal blood, cord blood;
ICP-MS

Age at measurement:
3 trimesters during pregnancy
and delivery

Mean:

Maternal T1: 4.15 pg/dL, T2:
3.44 pg/dL, T3: 3.78 pg/dL

Cord: 2.86 pg/dL
Max:

Maternal T1: 20.5 pg/dL, T2:
7.5 pg/dL, T3: 8.0 pg/dL
Cord: 6.9 pg/dL

Mental development
composite assessed
using the ECDI by
Harold Ireton
(language
comprehension,
expressive
language, gross
motor, self-help,
social interaction).

Cutoff point scores
for development
delay were score
<20% of that
expected for
children's age.

Age at outcome:
36 mo

Maternal educational,
BMI, family income,
gestational age, birth
weight, birth order (first
born).

OR

Total ECDI: 1.74 (1.18,
2.5).

tLinetal. (2013)
Taipei, Taiwan

April 2004-Jan 2005
(enrollment)

Birth cohort
n: 230

Mother-infant pairs from
medical center, local
hospital, and obstetric
clinics.

Blood

Maternal blood, cord blood;
ICP-MS, measured for Pb,
Mn, As, and Hg.

Cognitive and
language
development
assessed using
CDIIT

Maternal age, education,
infant gender, ETS
during pregnancy and
after delivery, fish intake,
and HOME score.

Beta

Cognitive

High vs. Low Pb: -5.35
(-9.642, -1.058)b
High MnlowPb: -4.15
(-9.618, 1.318)b

External Review Draft

3-278

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

Followed through 2 yr

Panel Study

Pb categories:
Low: <16.45 |jg/L
High: >16.45 pg/L
Mn categories:
Low: <59.59 |jg/L
High: >59.59 pg/L

Age at measurement:
delivery

Mean: 13 |jg/L, GM:
10.61 |jg/L
75th: 16.45 |jg/L
Max: 43.22 pg/L

Age at outcome:
2 yr

Low Mn*high Pb: -4.79
(-10.298, 0.718)b

High Mn*high Pb: -8.19
(-14.403, -1.977)b

Language

High vs. Low Pb: -2.53
(-6.234, 1.174)b
High Mn*low Pb: -1.56
(-6.264, 3.144)b
Low Mn*high Pb: 0.22
(-4.523, 4.963)b

HighMn*highPb: -6.81
(-12.161,-1.459)b

tNozadi etal. (2021)

Navajo Nation
United States

February 2013-June
2018 (enrollment)
Followed through 10-
13 mo

Cohort

Navajo Birth Cohort

Study

n: 327

Blood

Maternal blood, child blood;
ICP-DRC-MS.

Age at measurement:

Delivery or 36-wk visit
(maternal); 10, 13 mo (child)

GM = 0.410 pg/dL;
median = 0.37 pg/dL
75th: 0.51 pg/dL
95th: 1.20 pg/dL

Problem-solving
scores assessed
using ASQ:I.

Age-adjusted
scores.

Age at outcome: 10-
13 mo

Urine strontium and
arsenic.

Beta

Problem-Solving: -0.67
(-1.54, 0.20)

tNvanza et al. (2021)

Northern Tanzania
Tanzania

2015-2017 (enrollment)
Followed through 6-
12 mo

Mining and Health
Prospective Longitudinal
Study in Northern
Tanzania
n: 439

Birth cohort of mother-
child pairs recruited in
2nd trimester

Maternal dried blood spots;
ICP-MS, measured for Pb,
Hg, and Cd

Age at measurement:

T2

Median: 2.72 pg/dL

Language and
global

neurodevelopment
assessed using
MDAT. Scores in
each domain
classified as normal
(>90th percentile on
all items in that
domain or <90th

Maternal age and
education, maternal and
paternal occupation,
number siblings under
5 yr at home, and family
SES, infant sex, age,
birth weight, height, and
weight as a proxy for
nutritional status.
(Covariates with p < 0.20

Prevalence ratio

Language Development:
1.0 (1.0, 1.0)

Global

neurodevelopmental
status: 1.0 (0.9, 1.0)

External Review Draft

3-279

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

Cohort

75th: 4.25 pg/dL
Max: 14.5 pg/dL

percentile on one or
two items in the
domain) or impaired
(<90th percentile on
more than two items
in a domain).

retained in the final
models.)

Hg>0.08 pg/dL * Pb> 3.5
pg/dL: 1.4 (0.9, 2.1)

Age at outcome:
6-12 mo

AAS = atomic absorption spectroscopy; As = arsenic; ASQ:I = Ages and Stages Questionnaires: Inventory; BDI = Beck Depression Inventory; BLL = blood lead level; BMI = body
mass index; BPA = bisphenol A; BSID = Bayley Scales of Infant and Toddler Development; CARES = Communities Actively Researching Exposure Study; CCAAPS = Cincinnati
Childhood Allergy and Air Pollution Study; Cd = cadmium; CDIIT = Comprehensive Developmental Inventory for Infants and Toddlers; CHECK = Children's Health and Environmental
Chemicals in Korea; CI = confidence interval; CKD = chronic kidney disease; CKiD = Chronic Kidney Disease in Children Study; ECDI = Early Child Development Inventory;
ELEMENT = Early Life Exposures in Mexico to Environmental Toxicants; ETS = environmental tobacco smoke; FSIQ = full-scale IQ; GDS = Gesell Developmental Schedules;
GFAAS = graphite furnace atomic absorption spectrometry; GM = geometric mean; Hg = mercury; HOME = Health Outcomes and Measures of the Environment; ICP-DRC-
MS = dynamic reaction cell for inductively coupled plasma mass spectrometry; ICP-MS = inductively coupled plasma mass spectrometry; ICP-OES = inductively coupled plasma
optical emission spectrometry; IQ = intelligence quotient; MDAT = Malawi Development Assessment Tool; MDI = Mental Developmental Index; Mn = manganese; mo = month(s);
MOCEH = Mothers' and Children's Environmental Health; NBAS = Neonatal Behavioral Assessment Scale; NR = not reported; OR = odds ratio; Pb = lead; PDI = Psychomotor
Developmental Index; PROGRESS = Programming Research in Obesity, Growth, Environment and Social Stressors; SD = standard deviation; SES = socioeconomic status; T1 =
first trimester of pregnancy; T2 = second trimester of pregnancy; T3 = third trimester of pregnancy; WISC = Weschler Intelligence Scale for Children; wk = week(s); yr = year(s).

aEffect estimates are standardized to a 1 pg/dL increase in BLL or a 10 pg/g increase in bone Pb level, unless otherwise noted. For studies that report results corresponding to a

change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the biomarker and standardized accordingly.

bThe CI was calculated from a p-value and the true CI may be wider or narrower than calculated.

°Results are unstandardized because the log base used for exposure transformation was unspecified in the study.

dResults are unstandardized because the Pb level distribution data was not available.

fStudies published since the 2013 Integrated Science Assessment for Lead.

Table 3-4E Epidemiologic studies of Pb exposure and performance on neuropsychological tests of cognitive
function, i.e., learning, memory, and executive function.

Reference and Study

Design Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

Lanphear et al. (2000) U.S. NHANES

Blood

Digit span

Child sex, race/ethnicity,

-0.05 (-0.09, -0.01)





WISC-R

poverty index ratio,









reference adult education,



External Review Draft

3-280

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

United States

1988-1994
Cross-sectional

n = 4,853 children ages

6-16 yr (born 1972-1988) Concurrent

GM (SD): 1.9 (7.0)

Large U.S. representative 63.5% <2.5

study of multiple risk Detection limit = 0.5
factors and outcomes

Interval analyzed: 1-

Age at outcome: 6-16 yr

Linear regression

serum ferritin and cotinine
levels. Did not consider
potential confounding by
parental cognitive function
or caregiving quality.

Krieq et al. (2010)

United States

1991-1994
Cross-sectional

U.S. NHANES

n = 773 children ages 12-
16 yr (born 1972-1982)

Large U.S. representative
study of multiple risk
factors and outcomes

Concurrent
GM (SD): 1.9 (7.0)
63.5% <2.5
Detection limit = 0.5
Interval analyzed: 1-5

Digit span
WISC-R

Age at outcome: 6-16 yr
Log-linear regression

Child sex, caregiver
education, family income,
race/ethnicity, test
language. Did not
consider potential
confounding by parental
cognitive function or
caregiving quality.

-0.34 (-0.59, -0.08)

Surkan et al. (2007)

n = 389 children

Blood

General memory index,

Caregiver IQ, child age,

-0.69 (-4.4, 3.0)







WRAML

SES, race, birth weight.

-6.7 (-12, -1.2)

Boston, Massachusetts
and Farmington, Maine

6-10 yr

Concurrent
Group 1: 1-2

Age at outcome: 6-10 yr

Also considered potential
confounding by site, sex,
birth order, caregiver

United States

Recruitment from trial of

Group 2: 3-4



education and marital





amalgam fillings

Group 3: 5-10



status, parenting stress,



Cross-sectional





and maternal utilization of





Mean (SD): 2.2 (1.6)



prenatal and annual
health care but not
parental caregiving
quality.



tYorifuii et al. (2011)

Faroese island
Denmark

1986-1987 (enrollment)
Followed through 7-14 yr

Cohort

Birth cohort

n: At age 7: 896, At age
14: 808

Birth cohort of mother-
infant pairs

Blood, hair

Cord blood;
electrothermal AAS.
Age at measurement:
At birth

GM of cord blood Pb:
1.57 [jg/dL
75th: 2.2 pg/dL

Verbal and visuospatial
reasoning, language,
learning, and memory
assessed using WISC-R
similarities, WISC-R block
designs, BNT, and CVLT-
C.

Age at outcome:

7, 14 yr

Age, sex, maternal
Raven's score, paternal
employment and
education, maternal
education, daycare at age
7, medical risk, and
maternal alcohol use and
smoking during
pregnancy

Beta

WISC-R at 7 years old
with cord mercury
Block Design: -0.011
(-0.083, 0.062)

Similarities: -0.122
(-0.38, 0.135)

External Review Draft

3-281

DRAFT: Do not cite or quote

-------
Reference and Study

Design	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

Digit Span Forward: -0.1
(-0.183, -0.016)

WISC-R at 7 years old
Block Design: -0.004
(-0.013, 0.006)

Similarities: 0.019
(-0.051, 0.088)

Digit Span Forward:
-0.028 (-0.049, -0.006)

CVLT-C at 14 years old
plus interaction with cord
mercury

Recognition: -0.053
(-0.136, 0.031)

Long-term Recall: -0.1
(-0.256, 0.057)

Short-term Recall: -0.009
(-0.178, 0.16)

Learning: -0.438 (-0.965,
0.089)

CVLT-C at 14 years old
Recognition: -0.001
(-0.022, 0.021)

Long-term Recall: 0.041
(0, 0.082)

Short-term Recall: 0.013
(-0.031, 0.058)

Learning: 0.037 (-0.103,
0.177)

External Review Draft

3-282

DRAFT: Do not cite or quote

-------
Reference and Study

Design	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

Boston Naming Test at 14
years old plus interaction
with cord mercury
With Cues: 0.002 (-0.337,
0.342)

No Cues: -0.095 (-0.473,
0.283)

Boston Naming Test at 14
years old

With Cues: 0.033 (-0.056,
0.122)

No Cues: 0.003 (-0.096,
0.102)

WISC-R at 14 years old
plus interaction with cord
mercury

Block Design: 0.241
(-0.575, 1.057)

Similarities: -0.147
(-0.383, 0.089)

Digit Span Backward:
-0.16 (-0.253, -0.067)

Digit Span Forward:
-0.107 (-0.199, -0.016)

Digit Span: -0.267
(-0.423, -0.111)

WISC-R at 14 years old
Block Design: -0.023
(-0.238, 0.191)

Similarities: -0.007
(-0.069, 0.055)	

External Review Draft

3-283

DRAFT: Do not cite or quote

-------
Reference and Study

Design	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

Digit Span Backward:
-0.035 (-0.06, -0.009)

Digit Span Forward:
-0.024 (-0.048, 0)

Digit Span: -0.059 (-0.1,
-0.017)

CVLT-C at 7 years old
plus interaction with cord
mercury

Recognition: -0.094
(-0.2, 0.011)

Long-term Recall: 0.037
(-0.141, 0.215)

Short-term Recall: -0.068
(-0.22, 0.084)

Learning: -0.501 (-0.981,
-0.02)

CVLT-C at 7 years old
Recognition: -0.003
(-0.032, 0.026)

Long-term Recall: 0.033
(-0.014, 0.081)

Short-term Recall: 0.043
(0.003, 0.084)

Learning: 0.073 (-0.057,
0.202)

Boston Naming Test at 7
years old plus interaction
with cord mercury	

External Review Draft

3-284

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

With Cues: -0.138
(-0.469, 0.193)

No Cues: -0.046 (-0.376,
0.284)

Boston Naming Test at 7
years old

With Cues: 0.042 (-0.045,
0.129)

tTatsuta et al. (2014)

Sendai, Tohoku region
Japan

Study years NR
Followed through 42 mo

Cohort

TSCD birth cohort
n: 387

Mother-infant pairs urban
areas of the Tohoku
district

Blood

Cord blood; ICP-MS.

Age at measurement:
Delivery

Median: 1.0 pg/dL
Max: 1.8 pg/dL

Intelligence and
achievement (K-ABC)

Age at outcome:
42 months

Child sex, birth order,
alcohol and smoking
habits, duration of
breastfeeding, annual
family income at 42 mo,
and maternal IQ (Raven
SPM)

No Cues: 0.039 (-0.049,
0.127)

Beta

Adjusted Model
Mental Processing Score:
-3.319 (-12.41, 5.774)

Sequential Processing
Score: -2.136 (-12.80,
8.531)

tOppenheimer et al.
(2022)

New Bedford Harbor,

Massachusetts

United States

1993-1998 (enrollment)

Followed through 2008-

2014

Cohort

New Bedford Cohort
n: 373

Blood

Cord blood; isotope
dilution ICP-MS

Age at measurement:
Delivery

Mean (SD): 1.4
(0.9) pg/dL
Max: 9.4 pg/dL

Cognitive Effects

Cognitive effects were
assessed using four
subtests of the Delis-
Kaplan Executive
Function System. These
included Trail Making:
Number-Letter Switching
condition, Verbal Fluency:
Category Switching
condition, Design
Fluency: Filled Dots and
Empty Dots Switching
condition, and Color-Word
Interference:

Multiple linear regression
models adjusted for child
race, sex, age at exam,
year of birth, HOME
score, maternal marital
status at child's birth,
maternal IQ, maternal
seafood consumption
during pregnancy,
maternal smoking during
pregnancy, maternal and
paternal education, and
annual household income
at child's birth, and study
examiner.

Beta

WRAML

Verbal Working Memory:
0.12 (-0.20, 0.45)

Symbolic Working
Memory: 0.09 (-0.246,
0.42)

Working Memory Index
Differences: 0.59 (-0.97,
2.15)

External Review Draft

3-285

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

Inhibition/Switching
condition.

Age at Outcome:
14-18 yr

Cho et al. (2010)

Seoul (metropolitan),
Seongnam (suburban),
Ulsan and Incheon
(industrial), and
Yeoncheon (rural)
South Korea

2009

Cross-sectional

n = 639 children (8-11 yr) Blood

Color-Word score

School-based recruitment Child blood; GFAAS with SCWT
Zeeman background

correction	Age at outcome:

Age at measurement: 8-11 F
8-11 yr

Mean (SD): 1.9 (0.67)
pg/dL

10th—90th: 1.2-2.8 pg/dL

Age, sex, paternal
education, maternal IQ,
child IQ, birth weight,
urinary cotinine,
residential area. Did not
consider potential
confounding by parental
caregiving quality.

Beta

0 (-0.02, 0.02)

tFruhetal. (2019)

Eastern Massachusetts
U.S.

1999-2002 (enrollment)
Followed through 7 yr

Cohort

Project Viva
n: 1006

Birth cohort of mother-
child pairs

Blood

Maternal venous
erythrocyte blood
specimens; ICP-MS

Age at measurement:
2nd to 3 rd trimester of
pregnancy (median: 27.9
weeks)

Med (IQR): 1.1 (0.06)
pg/dL

Executive Function
also Section 3.5.1)

see

Parent teacher ratings on
BRIEF

Age at outcome:

7 yr

Scores standardized for
child age and sex;
Additional adjustment for
maternal 2nd trimester Hg
and Mn levels, nulliparity,
smoking during
pregnancy, IQ, and
education; Paternal
education; HOME
composite score and
household income; and
child race/ethnicity.

Beta

BRIEF Parent-Reported

Behavioral Regulation
Index

All: 1.15 (-0.217, 2.517)

Girls: 1.717 (0.025, 3.408)

Boys: 0.85 (-1.058,

2.758)

Metacognition Index
All: 0.95 (-0.25, 2.15)

Girls: 1.483 i
3.075)

-0.108,

Boys: 0.6 (-1.05, 2.25)

External Review Draft

3-286

DRAFT: Do not cite or quote

-------
Reference and Study

Design	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

General Executive
Composite

All: 1.217 (-0.1, 2.533)

Girls: 1.95 (0.1, 3.8)

Boys: 0.783 (-0.958,
2.525)

BRIEF Teacher-
Reported

Behavioral Regulation
Index

All: 0.767 (-0.567, 2.1)

Girls: 0.933 (-1.142,
3.008)

Boys: 0.75 (-0.9, 2.4)

Metacognition Index
All: 0.683 (-0.717, 2.083)

Girls: 0.9 (-1.308, 3.108)

Boys: 0.683 (-1.142,
2.508)

General Executive

Composite

All: 0.7 (-0.65, 2.05)

Girls: 0.883 (-1.258,
3.025)

Boys: 0.683 (-1.008,
2.375)

External Review Draft

3-287

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tFruh etal. (2021)

Eastern Massachusetts,
U.S.

1999-2002 (enrollment),
Followed through 6-11 yr

Cohort

Project Viva
n: 2128

Birth cohort of mother-
child pairs

Blood

Maternal venous
erythrocyte blood
specimens; ICP-MS

Age at measurement:
2nd to 3 rd trimester of
pregnancy (median: 27.9
weeks)

Med (IQR): 1.1 (0.06)
pg/dL

Global Executive
Composite (GEC); Total
Difficulties score

BRIEF; SDQ

Age at Outcome:
6-11 yr

Scores on BRIEF
standardized forage, sex
and model adjusted for
child race/ethnicity,
maternal parity, maternal
smoking status, maternal
IQ, maternal education,
paternal; education,
hemoglobin, HOME
score, household income,
and fish consumption.
SDQ model adjusted for
age, sex, and above
covariates. Pb, Mn, Se
and MeHg included
together in the models.

Beta

SDQ Total Difficulties:
0.617 (-0.058, 1.292)

BRIEF GEC:
1.11 (-0.12, 2.34)

tRuebner et al. (2019)

46 centers
U.S.

Study Years: NR
Followed through: 1-16 yr

Cohort

CKiD Cohort study
n: 412

Children with mild to
moderate CKD

Blood

ICP-

Child venous blood
MS. The BLL
measurement closest to
the time of neurocognitive
testing was used for
analysis (concurrent).

Age at measurement:
NR; 2, 4, or 6 years after
study entry

Median: 1.2 pg/dL
75th: 1.8 Mg/dL
Max: 5.1 [jg/dL

Executive function (see Age, sex, race, poverty,
also Section 3.5.1 [FSIQ], maternal education.
Section 3.5.2 [attention
and hyperactivity])

Age-specific assessments
administered at visit 3, 5,

7, or 9. Last available
results used (mean time
between BLL and
outcome

assessment = 2.3 yr).

Delis-Kaplan Executive
Function System Tower
Subset (>6 yr), BRIEF-P
(2-5 yr), BRIEF (6-18),

BRIEF-A (>18 yr)

Age at outcome:

1 to >18 yr

Adjusted BRIEF results
were not reported
because they were not
statistically significant.

tMerced-Nieves et al.
(2022)

PROGRESS Cohort
n = 549

Blood

Various measures from
Condition Position
Responding (CPR),

Child's age at testing, Betas for BLL at T3
maternal education (
-------
Reference and Study

Design	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

Mexico City
Mexico

2007-2011 (enrollment)
Followed through 6-7 yr
Cohort

Birth cohort

Maternal, cord, and child
blood; Agilent 8800 ICP
Triple Quad

Age at measurement:
Maternal: T2, T3, delivery
Cord: delivery
Postnatal: 4-6 yr

Mean (SD)

Maternal T2: 2.7 (2.7)
pg/dL

Maternal T3: 3.9 (2.8)
pg/dL

Maternal at delivery: 4.3
(3.2) pg/dL

Cord: 3.4 (2.6) pg/dL

Child: 2.4 (2.6) pg/dL

Temporal Response
Differentiation (TRD),
Delayed Matching-to-
Sample (DMTS), and
Incremental Repeated
Acquisition (IRA) from the
OTB

Age at Outcome:

6-7 yr

>high school), and SES.
Modification by sex
examined.

Observing response
latency:

0.001s (-0.08, 0.08s)
TRD

Average latency:
0.14s (-0.001, 0.29s)

DMTS

Average observing
response latency:
0.08s (-0.04, 0.20s)

IRA

Effective response rate:
-0.01s (-0.03, -0.002s)

AAS = atomic absorption spectrometry; BLL = blood lead level; BNT = Boston Naming Test; BRIEF = Behavior Rating Inventory of Executive Functions;

CANTAB = Cambridge Neuropsychological Test Automated Battery; CI = confidence interval; CKD = chronic kidney disease; CKiD = Chronic Kidney Disease in
Children Study; CVLT-C = California Verbal Learning Test-Children's version; ETS = environmental tobacco smoke; FSIQ = full-scale IQ; GFAAS = graphite
furnace atomic absorption spectrometry; GM = geometric mean; HOME = Health Outcomes and Measures of the Environment; ICP-MS = inductively coupled
plasma mass spectrometry; K-ABC = Kaufman Assessment Battery for Children; MANAs = Metals, Arsenic and Nutrition in Adolescents Study; MeHg = methyl
mercury; NHANES = National Health and Nutrition Examination Survey; NR = not reported; OTB = Operant Test Battery; Pb = lead; SCWT = Stroop Color-Word
test; SD = standard deviation; SDQ = Strengths and Difficulties Questionnaire; SES = socioeconomic status; SPM = Standard Progressive Matrices; T1 = first
trimester of pregnancy; T2 = second trimester of pregnancy; T3 = third trimester of pregnancy; TSCD = Tohoku Study of Child Development; WISC = Weschler
Intelligence Scale for Children; WRAML = Wide Range Assessment of Memory and Learning; WRAT = Wide Range Achievement Test; yr = year(s).

aEffect estimates are standardized to a 1 pg/dL increase in BLL or a 10 pg/g increase in bone Pb level, unless otherwise noted. For studies that report results
corresponding to a change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the biomarker
and standardized accordingly.

fStudies published since the 2013 Integrated Science Assessment for Lead.

External Review Draft

3-289

DRAFT: Do not cite or quote

-------
Table 3-4T

Animal toxicological studies of Pb exposure and cognitive function.



Study

Species (Stock/Strain), n, Timing of
Sex Exposure

Exposure Details

BLL as Reported (pg/dL)

Endpoints
Examined

Corv-Slechta et al. (2012) Rat (Lonq-Evans) GD-60to10mo

Control (tap water), M,
n = 12

Oral, drinking water
Oral, lactation
In utero

PND 5-6:

<5 [jg/dL for Control
12.5 [jg/dL for 50 ppm

2-3 mo to 10 mo:
Operant Behavior



50 ppm, M, n = 12



2.5 mo:

<5 [jg/dL for Control
6.43 [jg/dL for 50 ppm
10 mo:

<5 [jg/dL for Control
8.98 [jg/dL for 50 ppm



Zou et al. (2015)

Mice (ICR) ~5 wk to 8 wk

Control (distilled water), M,
n = 10

250 mg/L solution, M,
n = 10

Oral, drinking water

8 wk:

1.8 pg/dL for Control
21.7 pg/dL for 250 mg/L

8 wk: Morris
water maze

External Review Draft

3-290

DRAFT: Do not cite or quote

-------
stlldy	Species (Stock/Strain),	TWngo,	Exposllre Detellls	BLL as Reported (|jg/dL)

Corv-Slechtaetal. (2013) Mice (C57BL/6) GD-60 to 12 mo

Oral, drinking water

PND 75 - Females:

7-12 mo:

Control (distilled deionized

Oral, lactation



Operant Behavior

water) - NS, M/F, n = 10-

In utero

-------
stlldy	Species (Stock/Strain),	TWngo,	Exposllre Detellls	BLL as Reported (|jg/dL)

Weston etal. (2014)

Rat (Long-Evans) GD -60 to PND 21

Control (tap water), M/F,
n = 22 (11/11)

50 ppm, M/F, n =22 (11/11)

Oral, lactation
In utero

PND 5-6-Males:
0.76 pg/dL for Control
15.7 pg/dL for 50 ppm
PND 5-6 - Females:
0.82 pg/dL for Control
14.7 pg/dL for 50 ppm

>PND 60:

Operant Behavior

Betharia and Maher

Rat (Sprague Dawley) GD 0 to PND 20

Oral, lactation

PND 2:

PND 21-25, 56-

(2012)

PND 21-25:

In utero



60: Morris water







1.77 ng/g (0.188 pg/dL) for

maze



Control (RO Dl water), M/F,



Control





n = 11-13













85.17 ng/g (9.02 pg/dL) for





10 |jg/mL , M/F, n = 11-13



10 pg/mL





PND 56-60:



PND 25:





Control (RO Dl water), M/F,



0.83 ng/g (0.088 pg/dL) for





n = 9-11



Control





10 pg/mL, M/F, n = 9-11



9.21 ng/g (0.98 pg/dL) for









10 pg/mL









PND 60:









0.23 ng/g (0.024 pg/dL) for









Control









0.30 ng/g (0.032 pg/dL) for









10 pg/mL



External Review Draft

3-292

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain), n,
Sex

Timing of
Exposure

Exposure Details

BLL as Reported (pg/dL)

Endpoints
Examined

Han et al. (2014)

Rat (Wistar)

Control (tap water), M, n =
2 mM - PW, M, n = 8
2 mM-ME, M, n = 8

PW group: PND 21
8 to PND 42

ME group: GD -21
to PND 20

Oral, drinking water
Oral, lactation
In utero

PND 21:

7.36 |jg/L (0.74 pg/dL) for Control
NR for 2 mM - PW

146.6 pg/L (14.7 pg/dL) for 2 mM

-	ME

PND 63:

9.22 pg/L (0.92 pg/dL) for Control

147.9 pg/L (14.8 pg/dL) for 2 mM

-	PW

46.13 pg/L (4.6 pg/dL) for 2 mM -
ME

PND 63 to
PND 68: Morris
water maze

Flores-Montova et al.
(2015)

Mice (C57BL/6)

Control (Sodium treated
water), M/F, n = 10 (8/2)

30 ppm, M/F, n = 10 (5/5)

330 ppm, M/F, n = 13 (7/6)

GD Oto PND 28

Oral, drinking water

PND 28 - Females:
0.02 pg/dL for Control
2.63 pg/dL for 30 ppm
12.92 pg/dL for 330 ppm
PND 28- Males:
0.31 pg/dL for Control
3.10 pg/dL for 30 ppm
15.21 pg/dL for 330 ppm

PND 28: Novel
Odor Recognition

Rahman et al. (2012a)

Rat (Wistar)

Control, M/F, n = 6

0.2% solution (0.002 g/mL),
M/F, n = 10

PND 1 to PND 21 Oral, drinking water

PND 21:

1.35 pg/dL for Control
12.40 pg/dL for 0.2% solution

PND 21: Morris
water maze

External Review Draft

3-293

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain), n,
Sex

Timing of
Exposure

Exposure Details

BLL as Reported (pg/dL)

Endpoints
Examined

Rahman et al. (2012b)

Rat (Wistar)

Control (tap water), M/F,
n = 6

0.2% solution, M/F, n = 10

PND 1 to PND 30

Oral, drinking water
Oral, lactation

PND 21:

1.4 pg/dL for Control
12.1 [jg/dL for 0.2% solution
PND 30:

1.2 [jg/dL for Control
12.8 [jg/dL for 0.2% solution

PND 21, 30:
Morris water
maze

Mansouri et al. (2012)

Rat (Wistar)

Control (distilled water),
M/F, n = 16 (8/8)

50 mg/L, M/F, n = 16 (8/8)

PND 70 to PND 100

Oral, drinking water

PND 100-Males:
2.05 [jg/dL for Control
8.8 [jg/dL for 50 mg/L
PND 100 - Females:
2.17 [jg/dL for Control
6.8 [jg/dL for 50 mg/L

PND 100: Morris
water maze,
Novel Object
Recognition

External Review Draft

3-294

DRAFT: Do not cite or quote

-------
Anderson et al. (2016)

Rat(Long-Evans)

Control (untreated), M/F ,
n = 16 (8/8)

150 ppm, M/F
per duration

375 ppm, M/F
per duration

750 ppm, M/F
per duration

Perinatal exposure
group: GD -10 to
PND 21

Early postnatal
exposure group:
PND 0 to PND 21

Long-term postnatal
exposure group:
PND 0 to PND 55

, n = 16 (8/8)
, n = 16 (8/8)
, n = 16 (8/8)

Oral, diet
Oral, lactation
In utero

External Review Draft

3-295

PND 65 - Perinatal exposure
females:

0 [jg/dL for Control
1.36 [jg/dL for 150 ppm
2.13 [jg/dL for 375 ppm

2.08	[jg/dL for 750 ppm

PND 65 - Early postnatal
exposure females:

0 [jg/dL for Control

2.11 |jg/dL for 150 ppm

2.0 [jg/dL for 375 ppm

3.09	[jg/dL for 750 ppm

PND 65 - Long-term exposure
females:

0 [jg/dL for Control

4.5 [jg/dL for 150 ppm

5.75 [jg/dL for 375 ppm

9.58 [jg/dL for 750 ppm

PND 65 - Perinatal exposure
males:

0 [jg/dL for Control
1.25 |jg/dL for 150 ppm
2.42 [jg/dL for 375 ppm
2.47 [jg/dL for 750 ppm

PND 55, 56, 57,
and 65: Trace
Fear Conditioning

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain), n, Timing of
Sex	Exposure

Exposure Details	BLL as Reported (pg/dL)

PND 65 - Early postnatal
exposure males:

0 [jg/dL for Control

1.64 [jg/dL for 150 ppm

1.95 [jg/dL for 375 ppm

2.83 [jg/dL for 750 ppm

PND 65 - Long-term exposure
males:

0 [jg/dL for Control
2.01 [jg/dL for 150 ppm
8.0 [jg/dL for 375 ppm
7.46 [jg/dL for 750 ppm

Menq et al. (2016)	Rat (Sprague Dawley) PND 0 to PND 21 Oral, lactation

Control (deionized water),

M/F, n = not specified

300 ppm Pb, M/F, n = not
specified

7.61 |jg/L (0.76 pg/dL) for Control PND 35-40:

Morris water
84.3 pg/L (8.43 pg/dl_) for	maze

300 ppm

External Review Draft

3-296

DRAFT: Do not cite or quote

-------
stlldy	Species (Stock/Strain),	TWngo,	Exposllre Detellls	BLL as Reported (|jg/dL)

Li etal. (2016a)

Mice (Kunming)

GD 1 to PND 21 Oral, drinking water

PND 21:

PND 21, 22, 23,



Control (untreated), M/F,





24, 25, 26: Morris



n = 10



9.8 |jg/L (0.98 |jg/dL) for Control

water maze



0.1% solution (1000 ppm),



42.5 |jg/L (4.25 pg/dL) for 1,000





M/F, n = 10



ppm





0.2% solution (2000 ppm),



85.3 pg/L (8.53 pg/dL) for 2000





M/F, n = 10



ppm





0.5% solution (5000 ppm),



106.4 pg/L (10.64 pg/dL)for





M/F, n = 10



5000 ppm



Li etal. (2016d)

Mice (Kunming)

GD 0 to PND 21 Oral, lactation

PND 21:

PND 21: Morris



Control (distilled water),

In utero



water maze



M/F, n = 10



10.62 pg/L (1.1 pg/dL) for Control





0.1% solution (mass



40.71 pg/L (4.1 pg/dL) for 0.1%





fraction), M/F. n = 10



solution





0.2% solution (mass



81.77 pg/L (8.2 pg/dL) for 0.2%





fraction), M/F, n = 10



solution





0.5% solution (mass



103.36 pg/L (10.3 pg/dL) for





fraction), M/F, n = 10



0.5% solution



Menq et al. (2016)

Rat (Sprague Dawley)

PND 1 to PND 21 Oral, lactation

PND 35:

NR: Morris water



Control (deionized water),





maze



M/F, n = 7



5.6 pg/L (0.56 pg/dL) for Control





300 ppm, M/F, n = 7



84.84 pg/L (8.48 pg/dL) for









300 ppm



External Review Draft

3-297

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain),
Sex

n,

Timing of
Exposure

Exposure Details

BLL as Reported (pg/dL)

Endpoints
Examined

Wanqetal. (2013)

Rat (Sprague Dawley)

Control (untreated), M/F,
n = 6

0.2% solution (w/v), M/F,
n = 6 - Gestational
Exposure

0.2% solution (w/v), M/F,
n = 6 - Lactational
Exposure

0.2% solution (w/v), M/F,
n = 6 - Ablactational
Exposure

GD Oto PND 1,
PND 1 to PND 21,
PND 21 to 42

Oral, drinking water
Oral, lactation
In utero

PND 72:	PND 65 to

PND 69: Morris
34.99 |jg/L (3.5 pg/dL) for Control water maze

35.78 |jg/L (3.58 pg/dL) for 0.2 %
solution Gestational

65.97 pg/L (6.60 pg/dL) for 0.2%
solution Lactational

110.67 pg/L (11.07 pg/dL) for
0.2% solution Ablactational

Wanqetal. (2016)

Rat (Sprague Dawley)

Control (tap water), M, n = 7

100 ppm, M, n = 9

PND 24 to PND 56 Oral, drinking water

PND 56:

11 pg/L (1.1 pg/dL) for Control

133 pg/L (13.3 pg/dL) for
100 ppm

PND 60-66:
Trace Fear
Conditioning

Zhang et al. (2014)	Mice (Kunming)	GD Oto PND 21 Oral, lactation	PND 36:	PND 29, 30:

Control (distilled water),	In utero	Passive

M/F, n = 12	18.5 pg/L (1.9 pg/dL) for Control Avoidance Test,

PND 31-35:

0.4% solution, M/F, n = 13	136.7 pg/L (13.7 pg/dL) for 0.4% Morris water

Solution	maze

External Review Draft

3-298

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain), n,
Sex

Timing of
Exposure

Exposure Details

BLL as Reported (pg/dL)

Endpoints
Examined

Barkurand Bairv (2015b) Rat (Wistar)

Control (untreated), M,
n = 6

0.2% solution - Gestational,
M, n = 6

0.2% solution - Lactational
, M, n = 6

0.2% solution -
Gestation + Lactation , M,
n = 6

0.2% solution -
Pregestational, M, n = 6

GD -30 to PND 21 Oral, lactation
In utero

PND 22:

0.18 [jg/dL for Control

3.02 [jg/dL for 0.2% solution -
Pregestation

5.30 [jg/dL for 0.2% solution -
Gestational

26.65 [jg/dL for 0.2% solution -
Lactational

32.0 [jg/dL for 0.2% solution -
Gestation + Lactation

PND 30 to
PND 36: Morris
water maze,
PND 26, 27, 28:
Passive

Avoidance Test

Barkuretal. (2011)

Rat (Wistar)	GD 0 to PND 21

Control (tap water), M, n = 9

0.2% solution (w/v), M, n =9

Oral, lactation
In utero

PND 120:

0.24 [jg/dL for control
0.47 [jg/dL for 0.2% solution

PND 120:
Passive

Avoidance Test

External Review Draft

3-299

DRAFT: Do not cite or quote

-------
Verma and Schneider Rat (Long-Evans)
(2017)	Control, M/F, n = 32 (16/16)

PERI: GD -14 to
PND 21

150 ppm chow (PERI), M/F,
n = 32 (16/16)

150 ppm chow (EPN), M/F,
n = 32 (16/16)

EPN: PND 0 to
PND 21

External Review Draft

Oral, lactation	PND 14 - PERI Males:	PND 56, 57, 65:

In utero	Trace Fear

-------
Study

Species (Stock/Strain), n,
Sex

Timing of
Exposure

Exposure Details

BLL as Reported (pg/dL)

Endpoints
Examined









-------
Verma and Schneider Rat (Sprague Dawley)

(2017)	Control, M/F, n = 36 (18/18)

PERI: GD -14 to
PND 21

150 ppm Chow (PERI),
M/F, n = 36 (18/18)

150 ppm Chow (EPN), M/F,
n = 36 (18/18)

EPN: PND 0 to
PND 21

External Review Draft

Oral, lactation	PND 14 - PERI Males:	PND 56, 57, 65:

In utero	Trace Fear

-------
Study

Species (Stock/Strain), n,
Sex

Timing of
Exposure

Exposure Details

BLL as Reported (pg/dL)

Endpoints
Examined









-------
Anderson et al. (2012) Rat (Long-Evans)	GD-10toPND21 Oral, lactation

Control (untreated chow),

M/F, n =28(11/17)

250 ppm Chow, M/F, n = 15
(10/5)

750 ppm Chow, M/F, n = 25
(13/12)

1500 ppm Chow, M/F,
n = 23 (12/11)

External Review Draft

3-304

PND 1 - Males:

0 [jg/dL for Control
18.9 [jg/dL for 250 ppm
52.5 [jg/dL for 750 ppm

52.5	[jg/dL for 1500 ppm
PND 1 - Females:

0 [jg/dL for Control
21.9 [jg/dL for 250 ppm
47.2 [jg/dL for 750 ppm
56.7 [jg/dL for 1500 ppm
PND 7-Males:

0 [jg/dL for Control
8.5 [jg/dL for 250 ppm
29.1 [jg/dL for 750 ppm
35.7 [jg/dL for 1500 ppm
PND 7 - Females:
0 [jg/dL for Control
14.7 [jg/dL for 250 ppm
26.9 [jg/dL for 750 ppm

37.6	[jg/dL for 1500 ppm
PND 14- Males:

0 [jg/dL for Control

PND 55: Morris
water maze

DRAFT: Do not cite or quote

-------
stlldy	Species (Stock/Strain),	Tlnjngof	Exposllre Detellls	BLL as Reported (|jg/dL)

10.5	[jg/dL for 250 ppm

18.6	[jg/dL for 750 ppm
24.8 [jg/dL for 1500 ppm
PND 14 - Females:

0 [jg/dL for Control
11.8 |jg/dL for 250 ppm
20.2 [jg/dL for 750 ppm
26.4 [jg/dL for 1500 ppm

PND 21 - Males:

0 [jg/dL for Control

18.6	[jg/dL for 250 ppm

28.8	[jg/dL for 750 ppm

28.7	[jg/dL for 1500 ppm
PND 21 - Females:

0 [jg/dL for Control

17.9	[jg/dL for 250 ppm
27.4 [jg/dL for 750 ppm

29.8	[jg/dL for 1500 ppm

External Review Draft

3-305

DRAFT: Do not cite or quote

-------
Zhao et al. (2018)	Rat (Sprague Dawley) GD -14 to PND 10 Oral, lactation

Control (tap water), M, n = 8	In utero

0.005% solution , M, n = 8

0.01% solution, M, n = 8

0.02% solution, M, n = 8

External Review Draft

3-306

PND 0:

1.9 |jg/dL for Control
17.9 [jg/dL for 0.005% solution

23.2	[jg/dL for 0.01 % solution
48.8 [jg/dL for 0.02% solution
PND 3:

I.9	|jg/dL for Control

6.7 [jg/dL for 0.005% solution

II.5	|jg/dL for 0.01% solution
23.1 [jg/dL for 0.02% solution
PND 7:

1.3 |jg/dL for Control

8.1	[jg/dL for 0.005% solution

12.3	[jg/dL for 0.01 % solution
18.7 [jg/dL for 0.02% solution
PND 10:

1.2	|jg/dL for Control

5.6 [jg/dL for 0.005% solution
7.0 [jg/dL for 0.01 % solution
12.3 [jg/dL for 0.02% solution
PND 14:

0.7 [jg/dL for Control

PND 30: Morris
water maze

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain), n,
Sex

Timing of
Exposure

Exposure Details

BLL as Reported (pg/dL)

Endpoints
Examined

4.0	[jg/dL for 0.005% solution
5.5 [jg/dL for 0.01% solution
8.9 [jg/dL for 0.02% solution
PND 21:

1.1	|jg/dL for Control

2.5 [jg/dL for 0.005% solution
2.5 [jg/dL for 0.01% solution
2.98 [jg/dL for 0.02% solution
PND 30:

1.5 pg/dL for Control
1.0 [jg/dL for 0.005% solution
1.5 pg/dL for 0.01% solution
1.5 pg/dL for 0.02% solution

Neuwirth et al. (2019b)

Rat(Long-Evans)

Control (tap water), M/F,
n = 12 (6/6)

363.83 |jM solution, M/F,
n = 12 (6/6)

GD Oto PND 22

Oral, lactation
In utero

PND 22:

NR for Control

5.3-15 [jg/dL for 364 |jM solution
PND 56-90:

ND for Control,

ND for 364 pM

PND 56-90:
Attention Set
Shifting Test

External Review Draft

3-307

DRAFT: Do not cite or quote

-------
stlldy	Species (Stock/Strain),	TWngo,	Exposllre Detellls	BLL as Reported (|jg/dL)

Neuwirth et al. (2019c)

Rat(Long-Evans)

PERI: GD -14 to Oral, lactation

PND 14 - Females:

NR: Attention Set



Control, M/F, n = 12 (6/6)

PND 22 In utero



Shifting Test







-------
Study

Species (Stock/Strain), n,
Sex

Timing of
Exposure

Exposure Details

BLL as Reported (pg/dL)

Endpoints
Examined

Tartaqlione etal. (2020)

Rat (Wistar)

Control (tap water), M/F
n = 16 (9/7)

50 mg/L, M/F, n = 16 (9/7)

GD -28 to PND 23

Oral, lactation
In utero

PND 23:

0.007 pg/mL (0.7 pg/dL) for
Control

0.255 pg/mL (25.5 pg/dL) for
50 mg/L

PND 35: Y Maze
- Spontaneous
Alternation,
PND 63-65:
Novel Object
Recognition,
PND 68-72:
Morris water
maze

Xiao etal. (2014)

Rat (Wistar)

Control (tap water), M/F,
n = 10 (5/5)

Pre-weaning: 2 mM

Pre-weaning: GD
-21 to PND 21

Postweaning:
PND 21 to PND 84

Oral, drinking water
Oral, lactation
In utero

PND 21 - Pre-weaning:
10.09 pg/L (1 pg/dL) for Control
103.8 pg/L (10.4 pg/dL) for 2 mM

PND 85 to 90:
Morris water
maze

solution, M/F, n = 10 (5/5)

Postweaning: 2 mM
solution, M/F, n = 10 (5/5)

solution

PND 21 - Postweaning:

Not Reported

PND 91 - Pre-weaning:

10.32 |jg/L(1 |jg/dL) for Control

39.27 |jg/L (3.9 pg/dL) for 2 mM
solution

PND 91 - Postweaning:

10.32 |jg/L(1 pg/dL) for Control

105.45 pg/L (10.5 pg/dL) for 2
mM solution

External Review Draft

3-309

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain), n,
Sex

Timing of
Exposure

Exposure Details

BLL as Reported (pg/dL)

Endpoints
Examined

Sobolewski et al. (2020) Mice (C57BL/6)

F0:

F1: GD -60 to PND Oral, lactation
23-27	In utero

Control (distilled Dl water),
F, n = 10

100 ppm, F, n = 10
F1:

see Figure 1, n = 12
F2:

see Figure 1, n = 12
F3:

see Figure 1, n = 8-10

F1 PND 6-7:

0 [jg/dL for Control

12.5 pg/dL for 100 ppm (F0
dosing)

F3 PND 6-7:

0 ng/dL for Control

0 [jg/dL for 100 ppm (F0 dosing)

PND 60-120
(variable by
endpoint): Fl
Training

Ouvanq et al. (2019)

Rat (Sprague Dawley)

Control (tap water), M/F,
n = 6-10

0.05/0.01% solution, M/F,
n = 6-10

GD Oto PND 679

Oral, drinking water
Oral, lactation
In utero

wk 97:

0 mg/L (0 pg/dL) for Control

0.216 mg/L (21.6 pg/dL) for
0.05/0.01% solution

PND 674 to
PND 679: Morris
water maze

Singh et al. (2019)

Rat (Wistar)	3 mo to 6 mo

Control (distilled water), M,
n = 5

2.5 mg/kg, M, n = 5

Oral, gavage

6 mo:

5.76 pg/dL for Control
28.4 pg/dL for 2.5 mg/kg

6 mo: Morris
water maze

Xiao et al. (2020)

Rat (Sprague Dawley)

Control (tap water), F,
n = 10

125 ppm, F, n = 10

GD -7 to PND 68

Oral, drinking water
Oral, lactation
In utero

PND 68:

24.23 ng/mL (2.4 pg/dL) for
Control

205 ng/mL (20.5 pg/dL) for
125 ppm

PND 56 -61:
Morris water
maze, PND 55: Y
Maze -
Spontaneous
Alternation

External Review Draft

3-310

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain), n,
Sex

Timing of
Exposure

Exposure Details

BLL as Reported (pg/dL)

Endpoints
Examined

Su et al. (2016)

Rat (Sprague Dawley)

Control (deionized water
with 0.9% saline), M, n = 15

200 ppm, M, n = 16

PND 20 to PND 76 Oral, gavage

PND 76:

7.99 |jg/L (0.8 pg/dL) for Control

84.17 |jg/L (8.4 pg/dL) for
200 ppm

PND 76: Morris
water maze

External Review Draft

3-311

DRAFT: Do not cite or quote

-------
An et al. (2014)	Rat (Sprague Dawley) 4 wk to 12 wk

Control (deionized water
with NaAc), M, n = 12

100 ppm, M, n = 12

200 ppm, M, n = 12

300 ppm, M, n = 12

External Review Draft

Oral, drinking water

5	wk:	12-wk: Morris

water maze

0.96 [jg/dL for Control
7.07 |jg/dL for 100 ppm
11.54 |jg/dLfor200 ppm
14.76 |jg/dL for 300 ppm

6	wk:

0.96 [jg/dL for Control
8.13 |jg/dL for 100 ppm
12.92 |jg/dL for 200 ppm
16.65 [jg/dL for 300 ppm

7	wk:

0.96 [jg/dL for Control
9.68 |jg/dL for 100 ppm
13.37 |jg/dLfor200 ppm
19.48 |jg/dL for 300 ppm

8	wk:

0.96 [jg/dL for Control
9.64 |jg/dL for 100 ppm
17.07 |jg/dLfor200 ppm
22.02 [jg/dL for 300 ppm

9	wk:

0.96 [jg/dL for Control

3-312

DRAFT: Do not cite or quote

-------
stlldy	Species (Stock/Strain),	Tlnjngof	Exposllre Detellls	BLL as Reported (|jg/dL)

12.12 [jg/dL for 100 ppm
20.7 [jg/dL for 200 ppm
22.28 [jg/dL for 300 ppm

10	wk:

0.96 [jg/dL for Control
11.48 [jg/dL for 100 ppm
17.75 |jg/dLfor200 ppm
24.69 [jg/dL for 300 ppm

11	wk:

0.96 [jg/dL for Control

11.51	[jg/dL for 100 ppm

17.52	|jg/dLfor200 ppm
22.18 |jg/dL for 300 ppm

12	wk:

0.96 [jg/dL for Control
11.41 [jg/dL for 100 ppm
17.23 |jg/dLfor200 ppm
22.57 [jg/dL for 300 ppm

External Review Draft

3-313

DRAFT: Do not cite or quote

-------
stlldy	Species (Stock/Strain),	TWngo,	Exposllre Detellls	BLL as Reported (|jg/dL)

Li et al. (2013)	Rat(Wistar)	4wkto16wk	Oral, drinking water	4 mo:	4 mo: Morris

Control (tap water), M/F,	water maze



n = 16 (8/8)

500 ppm, M/F, n = 16 (8/8)



29.99 |jg/L (3 |jg/dL) for Control

159.54 |jg/L (16 |jg/dL) for
500 ppm



Zhu et al. (2019b)

Rat (Sprague Dawley) GD 0 to 12 mo

Oral, drinking water

12 mo:

NR: Morris water



Control (deionized water),

Oral, lactation



maze



M/F, n = 32

In utero

-------
Study

Species (Stock/Strain), n,
Sex

Timing of
Exposure

Exposure Details

BLL as Reported (pg/dL)

Endpoints
Examined

Zhang et al. (2012)

Rat (Sprague Dawley)

Control (deionized water),

NR (40-60 g)

Oral, drinking water

+8 wk from start of exposure:

M, n = 10



49.9 ng/mL (5 pg/dL) for Control

100 ppm, M,

n = 10

100.9 ng/mL (10.1 pg/dL) for





100 ppm

200 ppm, M,

n = 10







128.6 ng/mL (12.9 pg/dL) for

300 ppm, M,

n = 10

200 ppm

147.7 ng/mL (14.8 pg/dL) for
300 ppm

+8 wk from start
of exposure:
Morris water
maze

Hong et al. (2021)

Rat (Sprague Dawley)

Control (tap water), M/F,
n = 50

GD Oto PND 21

Oral, lactation
In utero

0.009 mg/L for Control,
0.291 mg/L for 1 g/L Pb -
PND 21

PND 21-27:
Morris water
maze

1 g/L Pb solution, M/F,
n = 50

Biiooret al. (2012)

Rat (Wistar)

Control (deionized water),
M/F, n = 10

50 ppm, M/F, n = 10

GD 0 to PND 45

Oral, drinking water
Oral, lactation
In utero

PND 45:

4.06 pg/dL for Control
10.65 pg/dL for 50 ppm

PND 45: Passive
Avoidance Test

Wang et al. (2021a)

Rat (Sprague Dawley)

Control (deionized water),
M, n = 8

0.05% solution, M, n = 8
0.1% solution, M, n = 8

GD Oto PND 21

Oral, lactation
In utero

PND 21:

23.1 pg/L (2.31 pg/dL) for Control

248 pg/L (24.8 pg/dL ) for 0.05%
solution

302 pg/L (30.2 pg/dL) for 0.1%
solution

PND 21: Morris
water maze

361 pg/L (36.1 pg/dL) for 0.2%
solution

External Review Draft

3-315

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain), n,
Sex

Timing of
Exposure

Exposure Details

BLL as Reported (pg/dL)

Endpoints
Examined

Liu et al. (2022c)

Rat (Sprague Dawley)

Control (tap water), M,
n = 10

0.2% solution, M, n = 10

PND 35 to PND 119 Oral, drinking water

PND 119:

10.9 |jg/L (1.09 pg/dL) for Control

176 pg/L (17.6 pg/dL) for 0.2%
solution

PND 119: Morris
water maze

Wang et al. (2021b)

Rat (Sprague Dawley)

Control (deionized water),
M/F, n = 12

0.05% solution, M/F, n = 10

GD -28 to PND 21

Oral, lactation
In utero

PND 21:

23.9 pg/L (2.39 pg/dL) for Control

206 pg/L (20.6 pg/dL) for 0.05%
solution

PND 21: Morris
water maze

Al-Qahtani et al. (2022)

Mice (Albino)

Control (distilled water), M,
n = 10

0.2 mg/kg, M, n = 10

8-9 wk to 14-15 wk Oral, gavage

14-15 wk:

1.2 pg/100 mL (1.2 pg/dL) for
Control

7.1 pg/100 mL (7.1 pg/dL) for
0.2 mg/kg

NR: Active
Avoidance Test

Long et al. (2022)

Rat (Sprague Dawley)

Control (untreated), M,
n = 12

200 mg/L solution, M,
n = 12

6 wk to 18 wk

Oral, drinking water

18 wk:

2.14 pg/L (0.214 pg/dL) for
Control

32.48 pg/L (3.25 pg/dL) for
200 mg/L solution

NR: Morris water
maze, NR: Active
Avoidance Test

BLL = blood lead level; CI = confidence interval; EPN = early postnatal; F = female; F1 = first filial generation; Fl
M = male; ME = maternal exposure; mo = month(s); NaAc = sodium acetate; NR = not reported; NS = no stress;
stress; PW = postweaning; RO Dl = reverse osmosis deionized; SD = standard deviation; wk = week.

= fixed interval; GD = gestational day; LOD = limit of detection;
Pb = lead; PERI = perinatal; PND = postnatal day; PS = prenatal

External Review Draft

3-316

DRAFT: Do not cite or quote

-------
Table 3-5E Epidemiologic studies of Pb exposure, academic performance, and achievement.

Reference and Study

Design	Study Population Exposure Assessment Outcome

Chandramouli et al.
(2009)

Avon
U.K

Jul. - Dec. 1992 (birth)
Followed 8 yr

10% random subsample
of Avon Longitudinal
Study of Parents and
Children (ALSPAC)
n =488

School children

Blood

Earlier childhood venous
blood; AAS using micro
sampling flame
atomisation

Age at measurement:
30 mo

Academic achievement

Standardized
Achievement Test

Age at outcome:

7 yr

Confounders

Maternal education and
smoking, home
ownership, home
facilities score, family
adversity index, paternal
SES, parenting attitudes
at 6 mo, child sex. Also
considered child IQ

Effect Estimates and
95% Clsa

Per doubling BLLb
-0.3 (-0.5, -0.1)

Cohort

Mean (SD): NR
Group 1: 0-<2 |jg/dL

Group 2: 2-<5 [jg/dL

Group 3: 5-<10 [jg/dL

Group 4: >10 pg/dL

Miranda et al. (2009)

School children, n=

Blood

Academic achievement

Sex, age of blood Pb

Score vs. blood Pb



57,568

Surveillance database



measurement, race,

category 1 [jg/dL

North Carolina





4th grade EOG test score

enrollment in
free/reduced lunch

2 [jg/dL:

U.S.

Screened for Pb at age

Age at measurement: 9-

for reading (2001-2005)

program, parental

-0.30 (-0.58, -0.01)



9-36 mo in 100 NC

36 mo



education, charter

3 [jg/dL:

1995 through 1999

counties





school.

-0.46 (-0.73, -0.19)

(screening)









4 [jg/dL:











-0.52 (-0.79, -0.24)

Cohort









5 [jg/dL:











-0.80 (-1.08, -0.51)

External Review Draft

3-317

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment Outcome

Confounders

Effect Estimates and
95% Clsa

Min etal. (2009)

Cleveland, OH
1994-1996 (birth)
Followed to age 11 yr
Cohort

Birth cohort, n = 267
86% African-American
with high prevalence of
prenatal drug and alcohol
exposure

Blood

Earlier childhood (age
4 yr)

Mean (SD): 7.0 (4.1)

Interval analyzed: 3.0
(10th percentile)-10

WJTA (math and reading
scores)

Age at outcome:

11 yr

Sex, caregiver education,
family income,
race/ethnicity, test
language.

Math score: -2.5 (-4.6,
-0.38)

Reading score: -2.9
(-4.4, -1.4)

Lanphearetal. (2000) U.S. NHANES

United States

1988-1994

Cross-sectional

n = 4,853 children ages
6-16 yr (born 1972-
1988)

Large U.S.

representative study of
multiple risk factors and
outcomes

Blood

Concurrent
GM (SD): 1.9 (7.0)
63.5% <2.5
Detection limit = 0.5
Interval analyzed: 1-5

WRAT (arithmetic and
reading scores)

Age at outcome: 6-16 yr

Linear regression

Child sex, race/ethnicity, -0.05 (-0.09, -0.01)

poverty index ratio,

reference adult

education, serum ferritin

and cotinine levels. Did

not consider potential

confounding by parental

cognitive function or

caregiving quality.

Krieq etal. (2010)

United States

1991-1994

Cross-sectional

U.S. NHANES
n = 773 children ages
12-16 yr (born 1972-
1982)

Large U.S.

representative study of
multiple risk factors and
outcomes

Blood

Concurrent
GM (SD): 1.9 (7.0)
63.5% <2.5
Detection limit = 0.5
Interval analyzed: 1-5

WRAT (arithmetic and
reading scores)

Age at outcome: 6-16 yr

Log-linear regression

Child sex, caregiver
education, family income,
race/ethnicity, test
language. Did not
consider potential
confounding by parental
cognitive function or
caregiving quality.

-0.34 (-0.59, -0.08)

External Review Draft

3-318

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment Outcome

Confounders

Effect Estimates and
95% Clsa

Surkan et al. (2007)

Boston, Massachusetts
and Farmington, Maine

United States
Cross-sectional

n = 389 children
6-10 yr

Recruitment from trial of
amalgam fillings

Blood

Concurrent
Group 1: 1-2
Group 2: 3-4
Group 3: 5-10
Mean (SD): 2.2 (1.6)

WIAT (reading and math
composites)

Age at outcome: 6-10 yr

Caregiver IQ, child age,
SES, race, birth weight.
Also considered potential
confounding by site, sex,
birth order, caregiver
education and marital
status, parenting stress,
and maternal utilization
of prenatal and annual
health care but not
parental caregiving
quality.

-0.69 (-4.4, 3.0)
-6.7 (-12, -1.2)

Chiodoet al. (2007)
Detroit, Michigan
United States

Cross-sectional

n = 495 children (born
1989-1991)age 7 yr

Blood

Concurrent
Mean (SD): 5.0 (3.0)

Test of Early Reading
Ability—2

MAT (math and reading
scores)

Age 7 yr

Maternal

psychopathology, IQ,
prenatal smoking,
prenatal marijuana, SES,
HOME score, caretaker
education and marital
status, # children in
home, child sex. Also
considered child age,
maternal age, custody,
cocaine use, prenatal
alcohol use.

-0.19 (-0.30, —0.08)c

Ferqusson et al. (1997) n = 881 children

Christchurch
New Zealand

1977 (birth)
Followed to age 18

Cohort

Tooth Pb (age 6-8 yr) Percent leaving school Maternal age,

Mean (SD): 6.2 (3.7) pg/g

Christchurch Health and
Development Study birth
cohort

without school certificate
Ages 16-18 yr

0-2 pg/g: 15.6

punitiveness, standard of
living, breastfeeding
duration, parental
conflict, grade, residence
on busy roads. Also
considered potential
confounding by sex,
ethnicity, maternal
education, family size,
HOME, SES, ethnicity,
parental change, birth
order, single parent.

16.7
18.1

3-5 pg/g:
6-8 pg/g:
9-11 pg/g: 19.7
12+ pg/g: 24.1
p < 0.05

External Review Draft

3-319

DRAFT: Do not cite or quote

-------
Reference and Study

Design	Study Population Exposure Assessment Outcome

Confounders

Effect Estimates and
95% Clsa

Needleman et al. (1990) n = 132 children (1st/2nd Tooth Pb (1st/2nd grade) Failure to graduate high Maternal age at birth,

Failure to graduate: 7.4

Chelsea and Somerville,
MA

United States

1975-1978 (enrollment)
followed to age 18 yr

Cohort

grade) in Massachusetts
schools

distribution
<10 ppm: 50%
10-19.9 ppm: 22.7%
>20 ppm: 27.3%

school

Highest grade achieved

Logistic regression

education, and IQ, family (1.4, 41) d

size, SES, sex, age at
testing, birth order,
alcohol use, mother and
child left hospital
together. Did not
examine potential
confounding by parental
caregiving quality.

OR >20 ppm vs.
<10 ppm

Highest grade achieved:
-0.03 (-0.05, -0.01)

per natural log increase
in tooth Pb

tZhanq et al. (2013)

Detroit, Michigan
United States

1990-2008 (born)

Followed through grade
3-8 (2008-2010)

Cohort

Students in public
schools in Detroit
n: 21281 (8831-3rd
grade, 7708- 5th grade,
4742- 8th grade)

At least 1 of the 3 tests
(math, science and
reading) taken and BLL
before 6 yr of age

Blood

Venous blood collected
for surveillance by Detroit
Department of Health
and Wellness Promotion
Age at measurement:
Before 6 yr of age (mean
age: 3.1)

Max: Highest BLL before
age 6 yr: 7.12 [jg/dL

Academic achievement
(math, science and
reading) in grade 3, 5,
and 8

Educational attainment in
math, science and
reading on MEAP.

Age at outcome:
3, 5, and 8 grades

Grade level, gender,
race, language, maternal
education, SES (i.e.,
school lunch status).

ORs of Scoring "Less
Than Proficient" on
MEAP Tests (Ref=<1
Ijg/dL)

1-5 /jg/dL

Mathematics: 1.42 (1.24,
1.63)

Science: 1.33 (1.10,
1.62)

Reading: 1.45 (1.27,
1.67)

6-10 iig/dL

Mathematics: 2.00 (1.74,
2.30)

Science: 2.22 (1.82,
2.72)

Reading: 2.21 (1.92,
2.55)

>10 iig/dL

Mathematics: 2.40 (2.07,

2.77)

Science: 2.26 (1.84,

2.78)

Reading: 2.69 (2.31,
3.12)

External Review Draft

3-320

DRAFT: Do not cite or quote

-------
Reference and Study

Design	Study Population Exposure Assessment Outcome

Confounders

Effect Estimates and
95% Clsa

tEvensetal. (2015)

Chicago metropolitan
area, 6 counties
U.S.

1994-1998 (born)
Followed 9-10 yr, 2003
2006
Cohort

Math Failure
1 jjg/dL increase

All Children:
1.06 (1.05, 1.07)
NH White:
1.11 (1.05, 1.18)
NH Black:
1.05 (1.04, 1.06)
Hispanic:
1.09 (1.06, 1.12)

Chicago public school

children

n: 46796

Blood

BLLs obtained from
Chicago Blood Pb
Surveillance program;
ICP-MS orAAS.

Age at measurement:
<72 mo (mean age:
45 mo)

Mean: 4.81 pg/dL

Academic achievement

3rd grade ISAT scores in
Reading and Math; 4
score categories, i.e.,
failure, below standard,
meets standard and
exceeds standard.

Age at outcome:

9-10 yr

Sex, mother's education,
low-income, very low
birth weight/preterm,
child's age at time of
BLL, ISAT vs. Iowa, race
(Interaction with race
ethnicity explored).

RR

Reading Failure
1 jjg/dL increase

All Children:
1.06 (1.05, 1.07)
NH White:
1.14 (1.08, 1.20)
NH Black:
1.05 (1.04, 1.06)
Hispanic:
1.08 (1.05, 1.11)

f Blackowicz et al.
(2016)

Chicago
U.S.

1994-1998 (birth)

Followed through 2003-
2006 (3rd grade)

Cohort

School children
n: 12319

Chicago Public Schools.

Blood

Chicago Blood Pb
Registry provided data
on BLL measured
between birth and 2006
Age at measurement:
between birth and 2006
(most recent was used in
analysis)

4.16 [jg/dL

School performance

3rd grade performance
based on ISAT scores

Age at outcome:
3rd grade

Child sex, maternal
education, low-income,
preterm birth, small for
gestational age, child's
age at time of BLL, ISAT
vs. Iowa, and Hispanic
subgroup (Mexican-
American vs. other
Hispanic and Puerto
Rican vs. Other
Hispanic);

Beta

Reading scores: -0.11
(-0.134, -0.086)

Math scores: -0.096
(-0.12, -0.072)

RR

Reading failure: 1.07
(1.05, 1.10)

Math failure: 1.09 (1.06,
1.12)

External Review Draft

3-321

DRAFT: Do not cite or quote

-------
Reference and Study

Design	Study Population Exposure Assessment Outcome

Confounders

Effect Estimates and
95% Clsa

tShadbeqian et al.

(2019)

North Carolina Statewide
U.S.

1990-2004 (birth)
Followed 6 yr (3-8
grade)

Cohort

NC Pb Poisoning
Prevention Program
Cohort

n: 560,624 (54% of the
Pb surveillance registry)

Living in NC between
2000-2012 with BLL
<10 |jg/dL at 0-5 yr

Blood

Child blood (BLL <10,
BLL <5, and a matched
group via CEM with BLL
<5 ijg/dL)

Age at measurement:
0-5 yr

Full sample (BLL
<10 |jg/dL) mean: 3.66,
Full sample (BLL
<5 |jg/dL) mean: 2.89,
CEM Matched sample
(BLL <5 |jg/dL) mean:
2.40

Academic achievement

Percentile scores on
standardized EOG tests
for math and reading

Age at outcome:

Grade 3 and Grade 8

Child's sex,
race/ethnicity, SES,
Medicaid enrollment,
birth month, and age
upon entry to grade 3,
mother's age, marital
status, parental alcohol
and tobacco use, highest
educational achievement
at the time of the child's
birth, vector representing
school, grade and year
combination

CEM to balance
distributions between
groups

Beta

Decrease in Test-Score
Percentile in Children
with 5-6 ijg/dL vs. BLL <
1 jjg/dL

Math: 0.95 (0.66, 1.24)
Reading: 1.41 (1.12,
1.70)

2- and 3-way interactions
for BLL*grade,
covariates*grade,
BLL*grade*covariates.

External Review Draft

3-322

DRAFT: Do not cite or quote

-------
Reference and Study

Design	Study Population Exposure Assessment Outcome

Confounders

Effect Estimates and
95% Clsa

tSkerfvinq et al. (2015)

Landskrona and

Trelleborg

Sweden

1978-2007 (enrollment
during primary school)

Followed through age 16

Cohort

Primary school children
n: 3176

Blood

Child venous blood;
flame or electrothermal
atomization AAS

Age at measurement:
7-12 yr

34 |jg/L; Median: 30
75th: 44
90th: 60
Max: 162

School performance (see
also Section 3.6.1 [adult
cognitive function])

School performance after
nine-year compulsory
schooling. 4-5 categories
from not passing to
passing with merit (based
on ranking)

Age at outcome:

16 yr

Child and parent country
of birth, parental
education, total family
income, father's IQ.

Beta

Merits

Children (BLL < 5 pg/dL):
-10.9 (-15.486, -6.314)

All Children: -6.36
(-9.986, -2.734)

Grades

Children (BLL < 5 pg/dL):
-0.112 (-0.177, -0.047)

All Children: -0.155
(-0.21, -0.1)

Note: CIs estimated from
p-values.

AAS = atomic absorption spectroscopy; BLL = blood lead level; CEM = coarsened exact matching; CI = confidence interval; EOG = end of grade; GM = geometric mean;

HOME = Health Outcomes and Measures of the Environment; ICP-MS = inductively coupled plasma mass spectrometry; ISAT = Illinois Standard Achievement Test; IQ = intelligence
quotient; MAT = Metropolitan Achievement Test; MEAP = Michigan Educational Assessment Program; Ml = Michigan; NHANES = National Health and Nutrition Examination Survey;
NR = not reported; Pb = lead; SD = standard deviation; SES = socioeconomic status; WIAT = Wechsler Individual Achievement Test; WJTA = Woodcock-Johnson Test of
Achievement; WRAT = Wide Range Achievement Test; yr = year(s).

aEffect estimates are standardized to a 1 |jg/dL increase in BLL or a 10 |jg/g increase in bone Pb level, unless otherwise noted. For studies that report results corresponding to a
change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the biomarker and standardized accordingly.

bResults are not standardized (e.g., BLL distribution data needed to calculate the standardized estimate was not reported or categorical data was analyzed).

The CI was calculated from a p-value and the true CI may be wider or narrower than calculated.

fStudies published since the 2013 Integrated Science Assessment for Lead.

External Review Draft

3-323

DRAFT: Do not cite or quote

-------
Table 3-6E Epidemiologic studies of Pb exposure and cognitive effects: population or group mean blood Pb
levels >5 (jg/dL.

Reference^nd Study study Population Exposure Assessment

Al-Saleh et al. (2020)

Saudi Arabia
2011-2013 (enrollment)
Followed through 2011-
2013 and 2017-2018
Cohort

Lactating mother-infant
pairs

n: 82 (36 males and 46
female children).

Blood, Hair, Urine, Breast
Milk

Maternal blood, spot
urine, breast milk, and
hair, child spot urine and
hair; AAS with
electrothermal atomizer

Age at measurement:
Maternal measurements
made during lactation
Infants at 3-12 mo
(lactation) and children at
5-8 yr old

Lactation:

GM: maternal urine
:5.881 |jg/L, hair
:1.717 |jg/g, blood GM:
2.346 [jg/dL, breastmilk:
46.483 |jg/L; Infant urine:
4.946 |jg/L, hair:
2.894 ijg/g;

Outcome

Neurodevelopmental
performance and visual-
motor integration (Test of
Nonverbal Intelligence
2nd edition and Beery
VMI 3rd edition,
respectively.)

Age at outcome:
5-8 yr old

Confounders

Child's age and sex,
maternal age, BMI,
parity, lifestyle,
educational level, SES,
residential

characteristics, urinary
cotinine levels (an index
of exposure to
secondhand smoke).

Effect Estimates and
95% CIs

Beta (95% Cl)a

BVMI: 0.012 (-1.989,
2.014)

TONI: -0.044 (-1.809,
1.721)

Early childhood:

GM: urine: 2.563 |jg/L,
hair: 0.850 |jg/g max:
urine: 20.826 |jg/L, hair:
4.470 |jg/g

External Review Draft

3-324

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

Barq et al. (2018)

Montevideo
Uruguay
Study years NR
Cross-sectional

n: 206

Children living in areas
considered high risk for
metal exposure.

Blood

Child venous blood;
flame AAS or GFAAS
Age at measurement:
5-8 yr

4.2 [jg/dL

Executive function

BRIEF (teacher rating)

Age at outcome:

5-8 yr

Child IQ, iron status, and
BMI, blood Pb testing
method, household
possessions, maternal
education, current parent
smoking.

PR (95% Cl)b

BRIEF- Global Executive
Composite

Children with BLL > 5 vs
<5 [jg/dL: 1.02 (0.96,
1.09)

Boys: 1.00 (0.98, 1.01)
Girls: 1.01 (0.99, 1.04)

Cai et al. (2021)

Guangxi
China

Study years NR
Cross-sectional

School children
n: 255

Participants living near a
Pb and zinc mine (-500
m distance between
school and mine).

Blood

Child venous blood (Pb
intoxication and non-Pb
intoxication groups);
GFAAS

Age at measurement:
7-12 yr

Median Pb level: Rice
samples: 0.10 mg/kg,
Blood: 84.8 |jg/L
75th: Blood: 115.4 pg/L
Max: Rice: 0.53 mg/kg

Perception and
reasoning

Raven's SPM

Age at outcome:
7-12 yr

Age, gender, physical
condition, lifestyle habits,
educational attainment
and smoking habit of
parents, family
environment and
economy.

Beta (95% Cl)b

IQ, RSPM: -0.58 (-1.031,
-0.129)

Jeong et al. (2015)

Multi-center
South Korea
Cross-sectional

May 2006-Dec 2010

MOCEH study
n: 194

Birth cohort- mother-
infant pairs followed
through 60 mo of age.
Cross-sectional analysis
conducted.

Blood

Child venous blood;
GFAAS

Age at measurement:
60 mo

GM: 13.01 pg/L
Max: 35.05 pg/L

FSIQ, VIQ, PIQ (Korean
WPPSI-R)

Age at outcome:

60 mo

Sex, parental education,
family income,
breastfeeding status,
CRP level, mother's BLL
during pregnancy

Note: Mediation analysis
to examine the
relationship of BLL, iron
deficiency and IQ.

Beta (95% Cl)b

Verbal IQ and In-BLL
(pg/L): -9.587 (-16.829,
-2.344)

External Review Draft

3-325

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

Kao etal. (2021)

Taipei
Taiwan
2011-2014
Cross-Sectional

recruited from Taipei
MacKay Memorial
Hospital

n:139 children less than
3 years of age

Hair, fingernails

Child hair, fingernails;
ICP-MS

Age at Measurement:
Mean (SD) 2.8 (0.4)
years (children under 3y)

Mean (SD): hair 2.9 (4.8)

pg/g,

nails 0.8 (5.1) |jg/g

BSID-III cognitive and
language development
scores

Age at outcome: 2.8 ±
0.4 yrs

General linear models
adjusted for sex,
gestational age at birth,
age of the house (years),
leafy-vegetable intake
(servings/week), and the
area of surface roads
within 100 m of the
residence

Regression results were
not reported because
they were not
statistically significant.

Kordas et al. (2011)

Mexico City, Mexico
Jan 1994-June 1995
Followed for 48 mo
Cohort

Birth cohort

n: 24 mo = 220,
48 mo = 186

Mother-infant pairs from
3 hospitals serving low-
and middle-income
women

Blood

Maternal, cord blood, and
child blood; GFAAS

Age at measurement:
Delivery (maternal, cord),
24 and 48 mo (child)

Mean: Maternal BLL at
delivery: 8.6 [jg/dL, Cord
blood: 6.6, BLL at 24 mo:
8.1, BLL at 48 mo: 8.1

Neurodevelopment using
BSID-II (MDI), MSCA
(general cognitive index
and memory scale).

Age at outcome:
24 (BSID) and 48 mo
(MSCA)

OLS linear regression

Birth weight, gestational
age, child sex; maternal
age, years of schooling,
IQ, smoking status,
marital status crowding in
the house, type of floor in
the house.

(Stratified analysis by
child development 48 mo
and gene polymorphism
also conducted.)

Beta (95% Cl)b

McCarthy Scales of
Children's Abilities, 48
months
GCI

Concurrent BLL: -0.6
(-0.992, -0.208)

Cord BLL: -0.2 i
0.388)

-0.788,

Memory Score

Concurrent BLL: -0.3
(-0.496, -0.104)

Cord BLL: 0.1 (-0.096,
0.296)

Bayley Scales of Infant
Development II, 24
months
MDI

Concurrent BLL: -0.1
(-0.492, 0.292)

Cord BLL: -0.7 (-1.288,
-0.112)

External Review Draft

3-326

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

Kuanq et al. (2020)

Nanjing

China

2012

Cross-sectional

Public primary school
children

n: 742

Excluded students with
congenital mental
retardation (third-degree
relatives included) and
diseases.

Blood

Child venous blood;
MS

ICP-

Age at measurement:
7-11 yr

Mean: 30.4 |jg/L; Median:
26.1 |jg/L

School performance

Standardized scores on
Chinese, Math and
English added for total
scores.

Age at outcome:
7-11 yr

Age and sex (tests
administered on the
same day).

Beta (no p-value, Cis,
or SE reported)13

Total: -0.168

Chinese: -0.042

Math: -0.039

English: -0.087

Lee et al. (2017)

Korea

Enrolled 2006-2015,
followed to age 5 (2015)
Cohort

Mothers' and Children's
Environmental Health
(MOCEH)
n: 251

Maternal Blood

GFAAS with Zeeman
background correction

Age at Measurement:
At birth (cord blood)

GM 0.957 [jg/dL
Max: 3.17 [jg/dL

Cognitive Development

The mental
developmental index
(MDI) of the Korean
BSID-II (K-BSID-II) was
administered to infants
who were 6, 12, 24, and
36 months-old. The
Korean language version
of the Wechsler
Preschool and Primary
Scale of Intelligence a€"
Revised (K-WPPSI-R)
was administered to
children at 60 months.

Partial correlation
analysis adjusted for
maternal education, sex
of child, and family
income.

Strongest correlations
between scores
measured at closest
time. Scores more stable
in those at extreme ends
of cognitive development.

External Review Draft

3-327

DRAFT: Do not cite or quote

-------
RefereDCesignnd	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

Beta for log-transformed
BLL (95% CI)b (ref: <8
pg/dL)

FSIQ

8-10 [jg/dL: -1.28
(-4.01, 1.46)

>10 [jg/dL: -1.45 (-3.50,
0.67)

Chinese score

8-10 [jg/dL: -3.20

(-5.78, -0.63)

>10 [jg/dL: -4.02 (-7.11,

-0.93)

Math score

8-10 [jg/dL: -5.25

(-8.14, -2.36)

>10 [jg/dL: -5.27 (-8.73,

-1.81)

English score

8-10 [jg/dL: -4.33 (-7.32,

-1.34)

>10 [jg/dL: -5.18 (-8.76,
-1.59)

Liu etal. (2018b)

PROGRESS study

Blood

Cognitive development

SES, mother's NR



n: 665



using BSID-III. BSID

hemoglobin during the

Mexico City, Mexico



Maternal blood; joint

scores were centered

second trimester of

Followed for 24 mo

Mother-infant pairs.

exposure to Mn, Pb, Co,

and scaled and

pregnancy, mother's

Cohort



Cr, Cs, Cu, As, Cd, and

presented as z-scores

educational level, child





Sb

normalized to expected
mean of 100 and SD of

gender, mother's WASI
IQ, and Fenton's birth





Age at measurement:

15.

weight z-scores.





Prenatal exposure (2nd









trimester)

Age at outcome:







NR

6, 12, 18, and 24 mo



External Review Draft	3-328	DRAFT: Do not cite or quote

Liu etal. (2013a)

Jiangsu province
China

2004-2005 (enrollment)
followed 5 yr

Cohort

China Jintan Child
Cohort Study
n: 1341 children (603
girls and 738 boys)

Community based cohort
of preschool children.

Blood

Early child blood;
GFAAS.

Age at measurement:
3, 4 or 5 yr

mean: 6.43 [jg/dL

FSIQ, VIQ, PIQ (Chinese
version ofWPPSI-R).

See also Table 3 (School
performance was
assessed by
standardized tests on 3
major subjects: Chinese,
English and Math.)

Age at outcome:

6 yr (IQ), 8-10 yr (school

performance)

Child age at blood Pb
test, child gender,
residence as defined as
school location, blood
iron level, parent
education, parent
occupation, and father's
smoking.

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

Marques et al. (2014)

State of Rondonia,

Western Amazon

Brazil

Cohort

2007-2012

Population-based cohort
n: 96 (TOKS=51 and
Itapua = 45)

Low SES populations
living in rural and urban
areas including children
living in the vicinity of
TOKS (i.e., multiple
metal exposure).

Breastmilk

Breastmilk; GFAAS
Age at measurement:
6 mo of breastfeeding
(from Marques 2013c)

TOKS: 10.04 pg/L
(mean), 8.2 pg/L
(median); 29.4 pg/L
(max)

Itapua: 3.89 pg/L (mean)
2.5 pg/L (median),
16.2 pg/L (max)

Neurodevelopment
(milestones including age
of walking and talking,
Bayley MDI and PDI);
milestones assessed
based on mothers'
recollection at the time of
visit.

Age at outcome:

6 and 24 mo (Bayley
MDI, PDI)

Birth weight, income,
maternal education,
breastfeeding status.

Beta (95% Cl)b

MDI 6 M -0.293 (-0.50,

0.08)

MDI 24 M -0.234 (-0.60,
0.13)

PDI 6 M -0.062 (-0.28,
0.16)

PDI 24 M -0.129 (-0.34,
0.08)

Age of walking -0.219
(-0.43, 0.002)

Age of talking -0.066
(-0.28, -0.16)

External Review Draft

3-329

DRAFT: Do not cite or quote

-------
RefereDCesignnd	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

Maqzamen et al. (2015)

Wisconsin (Milwaukee or
Racine)

United States
Enrollment: born during
1996-2000 (Follow-up of
blood lead level: before
child's third birthday;
Follow-up for WKCE
scores: 4th grade)

Cohort

Wisconsin Children's
Lead Levels and
Educational Outcomes
Project (CLLEO)
n: 1076

Blood

Blood lead records used
to categorize children as
not exposed (<5 mg/dL)
or exposed (>10 mg/dL
and <20 mg/dL)

Age at Measurement:
18-36 mo

43% of sample defined
as exposed

Academic achievement:
Wisconsin Knowledge
and Concepts Exam
(WKCE) math and
reading scores

WKCE reading and math
scores obtained from the
Wisconsin Department of
Public Instruction with
parental consent.

Child gender, race,
parental < HS education,
free lunch program,
English language learner,
and child health rating by
parents (excellent vs
other); interactions with
Pb tested for each
covariate

Beta for entire
distribution of math
scores'3

OLS = -8.94 (-14.84,
-3.05)

Beta for math scores in
quantilesb

10th percentile = -17.00
(-32.13, -3.27)
50th percentile = -8.00
(-15.24, -0.36)
90th percentile = -4.50
(-10.55, 4.50)

Beta for entire
distribution of reading
scores'3

OLS = -13.66 (-19.94, -
7.37)

Beta for reading scores
in quantilesb

10th percentile = -18.00
(-48.72, -3.32)
50th percentile = -14.50
(-20.72, -5.61)
90th percentile = -7.50 (
-15.58, 2.07)

External Review Draft

3-330

DRAFT: Do not cite or quote

-------
Rawat et al. (2022)	n: 43	Blood

India	Blood Pb was measured

Not reported	via LeadCare II testing

Cross-Sectional	analyser

Age at Measurement:

4-12 years

GM (SD) 19.93 (9.22)
ug/dL

Max: 37.4 ug/dL

% change in results from
Draw-A-Person test for
Group B compared to
Group A:

Line characteristic: Thick
and sharp = 41%; Soft =
-32%

IQ level, performance on
Draw-A-Person Test

The Draw-A-Person test
and the IQ test were
administered in the study
setting

Age at Outcome:
4-12 years

There were no
adjustments for
confounders, as simple
statistics were employed.

IQ - Mean (SD) scoreb

Group A (<10 ug/dL Pb,
n=9): 122.33 (4.03)

Group B (>10 ug/dL Pb,
n=34): 96.03 (12.76) p-
value for difference =
0.006

Detailing: With = 32%;
Without = -9%

Shading: With = -21%;
Without = 24%

Distortion: With = 50%;
Without = -17%

Colours: Warm = 41%;
Cool = -27%

Group A (<10 ug/dL,
n=9) vs Group B (>10
ug/dL, n=34) drawings
had the following
characteristics:

Thick and sharp lines =
33% vs 47%; Soft lines=
78% vs 53%%

External Review Draft

3-331

DRAFT: Do not cite or quote

-------
RefereDCesignnd	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

With detailing = 22% vs
29%%; Without detailing
= 78% vs 71%%

With shading = 33% vs
26%; Without shading =
67% vs 82%

With distortion = 33% vs
50%; Without distortion=
89% vs 74%

Warm colors= 67% vs
94%; Cool colors= 89%
vs 65%

Rodriaues et al. (2016)

Sirajdikhan and Pabna

districts

Bangladesh

2008-2011 (enrollment)
Cross-sectional

Birth cohort in
Bangladesh

n: 525 (Sirajdikhan: 239;
Pabna: 286)

Pregnant women
(gestational age <16 wk).

Blood

Child concurrent whole
blood tested using the Pb
Care II

Water samples from tube
well tested for As and Mn
during first trimester of
pregnancy and follow-up
visits at age of 1 mo,
12 mo and 20-40 mo.

Age at measurement:
20-40 mo

Median: Sirajdikhan:
7.6 [jg/dL; Pabna: 
-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

Roy etal. (2011)

Chennai, India (4
representative industrial
and traffic zones)
2005-2006
Cross-sectional

School Children (3-7 yr)
n: 725

3 schools from each of
the 4 zones randomly
selected (12 schools
total); children lower and
upper kindergarten and
first grades.

Blood

Postnatal venous blood;
PbCare Analyzer

Age at measurement:
3-7 yr of age

Overall mean:
11.5 |jg/dL; mean by
genotypes: Taq A1/A1:
11.66 [jg/dL; Taq
A1/A2+A2/A2:
11.42 [jg/dL

Max: Overall: 40.5 [jg/dL

IQ using BKT (mental
age divided by
chronological age and
multiplied by 100)

Tamil-translated Binet-
Kamat Scales of
Intelligence

Age at outcome:
3-7 yr of age

Age + age 2, sex,
midarm circumference,
average monthly family
income, and family size,
parental education.
Note: stratified analysis
by genotypes (3
categories) conducted.

Beta (95% Cl)b
IQ (BKT, Tamil-
translated): -4.22 (-7.10,
-1.36)

tRvaiel etal. (2021)

Mexico City
Mexico
1997-2005
Cohort

ELEMENT project
n: 85

Mother-child pairs
recruited at the Mexican
Social Security Institute

Blood

Maternal and child
venous blood; ICP-MS,
GFAAS

Age at measurement:
Tl, T2, T3 (maternal);
12, 24 mo (child)

Maternal blood GM (SD):
T1: 5.27 (1.93) pg/dL
T2: 4.74 (1.96) pg/dL
T3: 4.98 (1.93) pg/dL

MDI assessed using
BSID-II (Spanish version)

Age at outcome: 12-
24 mo

Maternal IQ (WAIS),
maternal age, infant
weight, length, SES,
infant age and sex,
current infant BLL.

Beta (95% CI) for 12-
month MDIb

Tl: 0.31 (0.00,0.62)
T2: 0.11 (-0.63,0.86)
T3: 0.41 (-0.34, 1.17)

Beta (95% CI) for 24-
month MDIb
Tl: -0.16 (-0.99,0.66)
T2: -0.23 (-1.05, 0.59)
T3: 0.28 (-0.50, 1.06)

Infant blood GM (SD):
12 mo: 3.92 (1.80) pg/dL
24 mo: 3.49 (1.93) pg/dL

tSanchez et al. (2011) ELEMENT study
n: 169

Mother-child pairs

Blood

Maternal blood; ICP-MS

MDI assessed using
BSID-II (Spanish
version)

Maternal age, IQ,
duration of
breastfeeding, sex,

Beta (95% Cl)b
T1: -5.42 (-10.2, -0.64)
T2: 0.88 (-5.34, 7.09)

External Review Draft

3-333

DRAFT: Do not cite or quote

-------
Mexico City
Mexico

1997-1999 (enrollment)

Followed through 24
months

Cohort

recruited during
pregnancy or before
conception.

Age at measurement:

each trimester of
pregnancy

Mean (SD):

1st trimester (n = 139):
13.7 (3.4) [jg/dL
2nd trimester (n = 159):
24.5 (2.8) [jg/dL

3rd trimester (n = 147):
35.2 (1.9) [jg/dL
Max:

1st trimester: 20.4 [jg/dL
2nd trimester: 33.7 [jg/dL

3rd trimester: 39.0 [jg/dL.

Scores standardized for
mother's age, mother's
IQ, duration of
breastfeeding, sex, and
weight and height z-
score at 24 months

Age at outcome:

24 mo

weight, and height Z-
score at 24 mo.

T3: 1.22 (-3.65, 6.08)

Saxena et al. (2022)

Araihazar
Bangladesh
2012-2016
Cross-Sectional

Metals, Arsenic, &
Nutrition in Adolescents
study (MANAs)
n: 572

Blood

Whole blood Pb
quantified using ICP-MS.
Age at Measurement:
Mean (SD) = 14.6 (0.7)
years

Mean = 98.7 A|jg/L;
Median = 91.29 A|jg/L

Cognitive Effects

The Cambridge
Neuropsychological Test
Automated Battery
(CANTAB) was
administered to the
adolescents to assess
aspects of executive
function.

Linear regression models
adjusted for BMI, head
circumference, child's
years of education,
maternal intelligence
(WASI), paternal years of
education, wall type, sex,
and other blood metals -
arsenic, cadmium,
manganese, and
selenium.

Beta (95% Cl)c

Delayed Match to
sample: -3.67 (-6.59, -
0.75)

Planning: -0.05 (-0.42,
0.32)

Rapid visual processing:
-0.01 (-0.03, 0.01)
Reaction time: 0.03 (-
0.01, 0.07)

Spatial recognition
memory: 1.9 (-0.88,
4.68)

Spatial span: -0.16 (-
0.43,0.11)

Spatial working
memory: 1.09 (-2.54,
4.72)

External Review Draft

3-334

DRAFT: Do not cite or quote

-------
Soetrisno and Delqado-
Saborit (2020)

West Java (Depok,
Bogor and Bekasi)
Sukatani village (control)
Indonesia
Cross-sectional

School children living in
urban locations near e-
waste facility; control site
n: 44 (22 from Bogor and
22 from Sukatani)

Children selected from
schools per teachers/
principal

recommendation.

Hair, soil, water

Hair samples from
children in Bogor and
Sukatani village. BLLs
from 36 children in Bogor
area (2010).

Age at measurement:
6-9 yr

Soil Pb mean: Depok-
Bekasi: 3653 mg/kg;
Sukatani: 93.2 mg/kg;
Water Pb: all 10 samples
below LOD; Hair Pb:
Depok-Bekasi:
0.155 mg/g; Sukatani:
0.0729 mg/kg
Max: Soil Pb: Depok-
Bekasi: 7662 mg/kg;
Sukatani: 115 mg/kg;

Hair Pb: Depok-Bekasi:
0.841 mg/g; Sukatani:
0.255 mg/kg

Academic achievement
(see also Section 3.5.1.4,
executive function)

Performance on reading,
math, writing expression
and oral language, arts,
science, social sciences,
and sports collected from
the school official alumni
report. TMT B.

Age at outcome:

6-9 yr

Age, parental education,
environmental tobacco
smoke at home, and
residential traffic
exposure.

Beta (95% Cl)d

Change in TMT-B
(seconds) per mg/g unit
of hair Pb: 54 (-3.8, 114)

Sun etal. (2015)

Jiangsu Province
China
Nov 2011
Cross-sectional

Chinese National Health Blood, Urine
Research Program
n: 446

IQ (CRT). Primary score Father's education,

Participants recruited
from three primary
schools located in the
three towns.

Child venous blood
samples; ICP-MS
method. Morning urine
samples collected was
tested for heavy metals.
Age at measurement:
9-13 yr

GM BLL: 33.13 pg/L;
Arithmetic mean BLL:
36.99 pg/L
75th: 43.39 pg/L
90th: 56.85 pg/L
Max: 101 pg/L

was converted to
standard IQ scores.

Age at outcome:
9-13 yr

mother's education, BMI,
annual family income,
gender, age.

Beta (95% Cl)b

-6.61 (-13.15, -0.07)

External Review Draft

3-335

DRAFT: Do not cite or quote

-------
Tassiopoulos et al.
(2017)

22 PHACS clinical
research sites in the
United States, including
Puerto Rico
USA

Enrollment began in
2007; BLL data available
from 1998-2014,
developmental data
available from
1996-2010,
developmental data
available from
1996-2010
Cohort

Surveillance Monitoring
of ART Toxicities
(SMARTT)
n: 546 children with a
Bayley-111 at one year of
age who had a BPb
between 9 months of age
and up to 3 months after
the Bayley-lll; 634
children with a Bayley
Screen at 3 years of age
and a BPb between 9
months of age and up to
3 months after the Bayley
Screen

Blood

Blood lead obtained
between the ages of 1
and 3 years as part of
standard of care or local
guidelines are abstracted
from the medical chart
when available
Age at Measurement: 1
yr (n=546) and 3 yrs
(n=634)

Cognitive Effects

Cognition and language
neurodevelopment using
BSID-III.

Age at outcome: 1 year

At 3 years of age,
developmental function
was assessed with the
Bayley Screening Test
(Bayley Screen), 19 which
includes a subset of
items from the Bayley-lll
with the domains of
cognition, receptive
communication,
expressive

communication, and fine
and gross motor
development.

Sex; race; ethnicity;
maternal IQ (evaluated
with the Wechsler
Abbreviated Scale of
Intelligence20); maternal
education, primary
language, living
arrangement, and living
situation; household
income; geographic
region; prenatal tobacco
exposure; postnatal
tobacco exposure within
the home; age at the
developmental
evaluation; and help from
others caring for the
child.

OR (95% Cl)b (for BLLs
>=5 vs <5 ug/dL)

Cognitive delay 1.64
(0.95, 2.90)

Receptive

communication 0.83
(0.47,1.43)

Expressive
communication 0.91
(0.52,1.58)

External Review Draft

3-336

DRAFT: Do not cite or quote

-------
Tung et al. (2022)

Providence, Rhode
Island

United States
Mother-newborn
assessed for exposure
and outcome within 2
hours of delivery and 24
hours of delivery,
respectively.
Cross-Sectional

Rhode Island Health
Study (RICHS)
n: 192

Placental Blood Pb

Placental Pb
concentrations quantified
using ICP-MS.

Age at Measurement:
24 hours

Mean = 4.49 ng/g among
those with detectable Pb

BSID

Newborns' neurologic
integrity, behavioral
function, and signs of
stress assessed by NICU
Network Neurobehavioral
Scale (NNNS). Latent
Profile Analyses used to
place children in
subgroups with discrete
profiles.

Multinomial regression
models adjusted for
infant gender, maternal
age, maternal race,
maternal BMI, education
status, and smoking
status during pregnancy.

OR (95% Cl)b for
neurobehavioral profile
membership associated
with detectable Pb
(>LOD, dichotomized) vs.
Profile 2 membership

Profile 1: 0.95 (0.38,

2.35)

Profile 3: 0.97 (0.42,

2.25)

Profile 4: 0.91 (0.38,

2.20)

Profile 5: 3.42 (0.88,
13.32), p<0.1

Profile characteristics: 5
= Highest arousal,
excitability and
hypertonicity with lowest
quality of movement and
regulation (most
extreme). Other profiles:
1 = High attention and
quality of movement; 2
[referent] = Average with
lowest lethargy; 3 =
Average, required more
handling; 4 = More signs
of lethargy, hypotonicity,
nonoptimal reflexes, low
attention and arousal.

*Unstandardized due to
BLL distribution
information

External Review Draft

3-337

DRAFT: Do not cite or quote

-------
Wan etal. (2021)
China

Cross-Sectional

n: 333

Blood

Blood samples were
collected from a previous
study and analyzed for
Pb. Authors did not
report analytical method.
Age at Measurement:
Children aged 9-11
years; exposure group
mean (SE) = 9.93 (0.85)
years; control group (SE)
= 9.62 (0.73)

Median for exposure
group = 7.163 [jg/dL;
median for control group
= 3.703 [jg/dL

FSIQ

Intelligence was tested
using the Combined
Raven's Test in China
(CRT-C2).

Multivariable linear
regression adjusted for
children's age and
gender, father's and
mother's age, education
levels and occupations,
passive smoking of the
children, and annual
family incomes.

Beta (95% Cl)a
-1.2 (-1.7, -0.60)

Wang etal. (2012)

Taizhou region (Luqiao
city and Lanxi city),
Zhejiang Province (for
exposure site) and
Chun'an, Zhejiang
province (reference site)
China

June 2010
Cross-sectional

School-based study
n: 329 (Luqiao: 108,
Lanxi: 151, Chun'an: 70)

Schools located neare-
waste recycling center
and tinfoil manufacturing
area (Luqiao and Lanxi
cities). Comparison
group schools in area
dominated by agriculture
(Chun'an).

Blood, urine
Child's venous blood,
urine; ICP-MS.

Age at measurement:
11-12 yr

GM: Luqiao: 6.97 [jg/dL,
Lanxi: 8.11 |jg/dL,
Chun'an: 2.78 pg/dL (42—
53% had BLL>10 pg/dL
in Luqiao and Lanxi and
no one had BLL>10 in
Chun'an)

Max: Luqiao:

57.24 pg/dL, Lanxi:
59.98 pg/dL, Chun'an:
7.59 pg/dL

IQ (CRT) calculated from
raw score.

Age at outcome:

11-12 yr

Child's sex, birth weight,
BMI, gestation at delivery
and the mother's age at
delivery, years of
education, yearly income,
tobacco exposure during
pregnancy and alcohol
exposure during
pregnancy.

Beta (95% Cl)b

IQ (CRT)

Female: -0.097 (-0.178,
-0.016)

Male: -0.096 (-0.175,
-0.016)

External Review Draft

3-338

DRAFT: Do not cite or quote

-------
AAS = atomic absorption spectrometry; ADHD = attention deficit/hyperactivity disorder; As = arsenic; BASC = Behavior Assessment System for Children; BKT = Binet Kamat Test Of
Intelligence; BLL = blood lead level; BMI = body mass index; BRIEF = Behavior Rating Inventory of Executive Functions; BSID = Bayley Scales of Infant and Toddler Development;
CANTAB = Cambridge Neuropsychological Test Automated Battery; CBLI = cumulative blood lead index; CI = confidence interval; Co = cobalt; Cr = chromium; CRS = Conners'
Rating Scales; CRT = Combined Raven's Test; Cs = cesium; Cu = copper; ELEMENT = Early Life Exposures in Mexico to Environmental Toxicants; ETS = environmental tobacco
smoke; Fe = iron; FSIQ = full-scale intelligence quotient; GFAAS = graphite furnace atomic absorption spectrometry; GM = geometric mean; Hg = mercury; HNES = Home Nurture
Environment Scale; HOME = Health Outcomes and Measures of the Environment; ICP-MS = inductively coupled plasma mass spectrometry; ICP-MS-DRC = inductively coupled
plasma mass spectrometry; KXRF = K-Shell X-Ray fluorescence; LOD = limit of detection; MDAT = Malawi Developmental Assessment Tool; MDI = Mental Developmental Index;
Mn = manganese; mo = month(s); MOCEH = Mothers' and Children's Environmental Health; MSCA = McCarthy Scales of Children's Abilities; NR = not reported; OLS = ordinary
least squares; Pb = lead; PDI = psychomotor developmental index; PIQ = performance intelligence quotient; PROGRESS = Programming Research in Obesity, Growth, Environment
and Social Stressors; Sb = antimony; SES = socioeconomic status; SPM = Standard Progressive Matrices; TMT = Trail Making Test; TOKS = tin, ores, kiln, smelters; VIQ = verbal
intelligence quotient; VMI = visual-motor integration WAIS = Wechsler Adult Intelligence Scale-Revised; WASI = Wechsler Abbreviated Scale of Intelligence; WISC = Weschler
Intelligence Scale for Children; WPPSI = Wechsler Preschool and Primary Scales of Intelligence; WRAT = Wide Range Achievement Test.

aEffect estimates are standardized to a 1 |jg/dL increase in BLL or a 10 |jg/g increase in bone Pb level, unless otherwise noted. For studies that report results corresponding to a
change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the biomarker and standardized
accordingly.

bEffect estimates are not standardized because data pertaining to the BLL distribution and/or base for the log-transformation were not reported.

°Per natural log increased in centered BLLs (i.e., BLL/median)

dResults are unstandardized due to the biomarker (hair)

fStudies published since the 2013 Integrated Science Assessment for Lead.

External Review Draft

3-339

DRAFT: Do not cite or quote

-------
Table 3-7E Epidemiologic studies of Pb exposure and performance on neuropsychological tests of attention,
impulsivity, and hyperactivity, ADHD-related behaviors, and clinical ADHD in children.

RefereDCeesfgnn	studV Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
Clsa

tNeuqebauer et al.
(2015)

Duisburg
Germany

2000-2002
(enrollment)
Followed through
2009-2011

Cohort

Duisburg birth
cohort study
n: 114

Pregnant women
and their offspring

Blood

Maternal venous blood; AAS

Age at measurement:
32 wk gestation (prenatal)

Mean (SD): 2.216 (1.083)
pg/dL

Med: 2.0 pg/dL
95th: 4.2 pg/dL
Max: 6.3 pg/dL

Attentional performance
using KiTAP with 5 subtests:
alertness, distractibility,
Go/No-go, divided attention,
flexibility; ADHD-associated
behavior using FBB-ADHS

Age at outcome:

Mean: 8.5 yr (KiTAP); 9.5 yr

(FBB-ADHS)

maternal
diseases, parental
lifestyle, childbirth
outcomes, HOME
Score

gMR:

KITAP

Inattention (omissions):
1.15 (1.00, 1.33)

Attention (performance
speed): 1.14 (0.98, 1.33)

FBB-ADHS

Overall ADHD: 1.061 (1.009,

1.115)

Impulsivity

1.133 (1.055, 1.216)

Hyperactivity

1.047 (0.992, 1.106)

Inattention

1.054 (0.989, 1.123)

External Review Draft

3-340

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
Clsa

tEthieret al. (2015)

Arctic Quebec

(Puvirnituq)

Canada

1993-1998
(enrollment)
Followed through
2009-2020

Cohort

NCDS
n: 27

Subsample of
school-aged
children who
participated in the
Cord Blood
Monitoring Program

Blood

Cord blood and concurrent
venous blood; GFAAS

Age at measurement:

Delivery (cord); 8.6-12.6 yr old
(concurrent)

Mean (SD): 5.4 (4.1) pg/dL
Max: 17.8 pg/dL

Selective spatial attention

Visuo-spatial attention-shift
task (adapted from Posner
paradigm)

Age at outcome:
8.6-12.6 yr

Sex, age at testing
time, SES,
breastfeeding
duration, maternal
alcohol, marijuana,
cigarettes use

(Each model used a
different set of
confounders)

Beta perSD increase in In-
transformed Pb:

Cord Blood

Reaction time: 0.02b
Omission Error: -0.02b
False Alarm: 0.42 (0.08,
0.76)c

Accuracy: -0.27b
Validity Effect: -0.05b

Concurrent Blood

Reaction time: 0.52 (-0.10,
1.14)c

Omission Error: -0.10b
False Alarm: -0.16b
Accuracy: -0.17b
Validity Effect: -0.13b

tTatsuta et al. (2014) TSCD birth cohort
n: 387

Sendai, Tohoku region
Japan

Study years NR

Followed through
42 mo

Cohort

Mother-infant pairs
urban areas of the
Tohoku district

Blood

Cord blood; ICP-MS.

Age at measurement:
Delivery

Median: 1.0 pg/dL
Max: 1.8 pg/dL

Sequential processing and Child sex, birth order, Betas: K-ABC

mental processing scores
(K-ABC)

Age at outcome:
42 mo

alcohol and smoking
habits, duration of
breastfeeding, annual
family income at
42 mo, and maternal
IQ (Raven SPM)

Sequential Processing:
-2.136 (-12.80, 8.531 )d
Mental Processing: -3.319
(-12.41, 5.774)d

External Review Draft

3-341

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
Clsa

tYorifuii et al. (2011) Faroese birth cohort Blood

Attention/working memory Child age, sex,

Faroese island
Denmark

1986-1987
(enrollment)

Followed through 7-
14 yrs

Cohort

n: 896 (7 yrs), 808
(14 yrs)

Mother-infant pairs

Cord blood; electrothermal
AAS with Zeeman background
correction.

Age at measurement:

Delivery

GM: 1.57 [jg/dL
75th: 2.2 pg/dL

assessed using WISC-R
digit span

Age at outcome:

7, 14 yr

maternal IQ (RPM),
paternal employment
and education,
maternal

education, daycare at
age 7, medical risk,
and maternal alcohol
use and smoking
during pregnancy

Beta per log-transformed
Pb:

7 yr

Digit span forward: -0.11
(-0.29, 0.07)d
<2.61 ug/g Hg: -1.70
(-3.12, -0.28)

14 yr

Digit span: -0.21 (-0.53,
0.11)d

Digit span forward: -0.04
(-0.23, 0.14)d
Digit span backward: -0.17
(-0.37, 0.04)d

<2.61 ug/g Hg: -2.73
(-4.32, -1.14)

tRuebner et al. (2019) CKiD Cohort study Blood

46 centers
United States

Study Years: NR
Followed through 1-
16 yr

Cohort

n: 412

Children ages 1-
16 yr at recruitment
with mild to
moderate CKD

Child venous blood; ICP-MS.
The BLL measurement closest
to the time of neurocognitive
testing was used for analysis
(concurrent).

Age at measurement:

NR; 2, 4, or 6 years after study

entry

Median: 1.2 pg/dL
75th: 1.8 pg/dL
Max: 5.1 pg/dL

Attention, hyperactivity,
response inhibition

and

Child age, sex, race,
poverty, and maternal
education

Age-specific neurocognitive
assessments (K-CPT, CPT
III, BASC-2) administered 3,
5, 7 or 9 years after study
entry. The last available test
results were used to
evaluate long-term effects.
Mean time between BLL
and neurocognitive testing
was 2.3 yr.

Age at outcome:

1 to >18 yr; median: 15.4 yr

Beta: K-CPT/CPT
Attention: 1.8 (0.15, 3.45)

Adjusted BRIEF and
BASC-2 results were not
reported because they were
not statistically significant.

External Review Draft

3-342

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
Clsa

tRoonev et al. (2018) Casa Pia Clinical Blood

Lisbon
Portugal

1997-2005
(enrollment age 8-12
yr)

Followed 7 yr age 15-
19 yr

Cohort

Trial of Dental
Amalgams in
Children
n: 330

Children aged 8-12
yr at baseline in the
Casa Pia school
system

Child venous blood; flameless
AAS

Age at measurement:
8-12 yr old (baseline)

Mean (SD):

Boys: 5.26 (2.73) pg/dL
Girls: 4.42 (2.19) pg/dL
Max: 15.0 [jg/dL

Neuropsychological tests of
attention

Stroop word, Stroop color,
Stroop color/word, WISC-III
Digitspan, WAIS-III, WMS-
III, Trail Making A, Adult
Trail Making A

Age at outcome:

15-19 yr (annual
assessment for 7 yr)

Age at baseline, race,
and nonverbal IQ
(home environment,
parent's SES,
medical histories
similar across
subjects

Median beta: Boys

Stroop word: -0.118
(-0.257, 0.021)

Stroop color: -0.114
(-0.246, 0.018)

Stroop color/word: -0.117
(-0.232, -0.001)

WAIS-III digitspan: -0.049
(-0.112, 0.015)

WMS-III spatialspan: -0.012
(-0.077, 0.054)

Adult Trailmaking A: -0.02
(-0.148, 0.108)

Median beta: Girls

Stroop word: -0.01 (-0.182,

0.162)

Stroop color: 0.033 (-0.142,
0.208)

Stroop color/word: -0.019
(-0.165, 0.126)

WAIS-III digitspan: -0.06
(-0.139, 0.02)

WMS-III spatialspan: -0.019
(-0.103, 0.066)

Adult Trailmaking A: -0.165
(-0.35, 0.021)

External Review Draft

3-343

DRAFT: Do not cite or quote

-------
tChoi et al. (2020)

Seoul
Korea

Aug. 2010-Feb. 2015
Case-control

Blood

Child venous blood; GFAAS
with Zeeman background
correction

Age at measurement:
5-18 yr

Mean: 1.4 (cases) vs. 1.3
(controls) [jg/dL

ADHD status
diagnosed with K-
SADS-PL

(neuropsychological
testing by board
certified

child/adolescent
psychiatrist)

n = 355 (259
ADHD, 96 controls)

5-18 yr old patients
at a child and
adolescent
psychiatry
outpatient clinic of
Seoul National
University Hospital

External Review Draft

Inattention and	Age, sex, IQ

hyperactivity/impulsvity
assed using ADHD-RS IV
(parent rating)

Attention and executive
function assessed using
computerized SCWT and
CPT

Age at outcome:
5-18 yr

Beta direct effects: ADHD-
RS

Total ADHD severity: 2.254
(-0.278, 4.785)

Inattention: 1.053 (-0.387,
2.493)

Hyperactivity/lmpulsivity:
1.259 (-0.042, 2.560)

Beta direct effects: Conners'
CPT

Inattention (errors of
omission): 3.748 (0.091,
7.404)

Impulsivity (errors of
commission): -0.925
(-4.412, 2.562)

Response Time: 2.515
(0.013, 5.017)

Response Time Variability:
2.647 (-0.846, 6.140)

Beta direct effects: Stroop

Stroop word: -1.143
(-3.316, 1.031)

Stroop color: -0.729
(-2.832, 1.375)

Stroop color/word: 0.491
(-1.876, 2.857)

Stroop color/word
interference: 1.618 (-0.963,
4.199)

Beta interaction: Stroop

DAT1 x Pb on Inattention
(errors omission): 10.613
(-0.237, 21.463)

3-344

DRAFT: Do not cite or quote

-------
RefereDCesignnd	Study Population Exposure Assessment

Outcome

Confounders Effect Est,™f?s and 95%

DAT1 x Pb on Response
Time Variability: -0.198
(-10.527, 10.132)

DRD4 x Pb on Inattention
(errors omission): -0.911
(-7.380, 5.558)

DRD4 x Pb on Response
Time Variability: -4.065
(-10.166, 2.036)

ADRA2A Mspl * Pb on
Inattention (errors omission):
2.870 (-2.340, 8.079)

ADRA2A Mspl x Pb on
Response Time Variability:
-1.588 (-6.526, 3.350)
ADRA2A Dral * Pb on
Inattention (errors omission):
5.066 (0.197, 9.934)

ADRA2A Dral x Pb on
Response Time Variability:
3.392 (-1.233, 8.017)

External Review Draft

3-345

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
Clsa

tBoucher et al.
(2012a)

Nunavik region,

Montreal

Canada

1993-1998
(enrollment)

Followed through Sep.
2005-Apr. 2007

Cohort

Cord Blood
Monitoring Program
(CBMP)

One child from the
Environmental
Contaminants and
Child Development
Study (1996-2000)

n: 196

School children
without known
neurodevelopmenta
I disorder or
medication for
attention problems

Blood

Cord and child blood; GFAAS
with Zeeman background
correction (cord), ICP-MS
(child).

Age at measurement:

Delivery (cord), 9-13 yr (child)

Cord: 4.8 [jg/dL (mean),
3.7 [jg/dL (med); 20.9 pg/dL
(max)

Concurrent child: 2.2 pg/dL
(mean), 2.0 pg/dL (med),
12.8 pg/dL (max)

Impairment in response
inhibition (Go/No-Go, ERPs
measured by EEG)

Electro-oculogram was
recorded from bipolar
miniature electrodes placed
vertically above and below
the right eye.

Age at outcome:

9-13 yr

Child age, sex, status
as adoptee; transport
by plane from remote
to larger village for
assessment; time of
assessment;
maternal age at
delivery; SES;
maternal nonverbal
reasoning abilities;
breastfeeding
duration; maternal
smoking, marijuana
use, binge drinking
during pregnancy;
docosahexaenoic
acid concentrations in
cord and child plasma
samples; Hg, PCBs

Beta per log-transformed
Pb: Cord blood

Mean Reaction Time (RT),
correct go trials: -0.05b
Mean RT, incorrect no-go
trials: -0.10b
Percent correct go trials:
-0.21 (-0.36, —0.06)c
Percent correct no-go trials:
-0.17 (-0.29, —0.05)c

Concurrent blood

Mean RT, correct go trials:
0.03b

Mean RT, incorrect no-go
trials: 0.03b

Percent correct go trials:
-0.12b

Percent correct no-go trials:
-0.16 (-0.27, —0.05)c

Rabinowitz et al.
(1992)

Taiwan

Study period NR
Cross-sectional.

N: 493

Mix of children
residing in urban or
rural environments
or near a smelter

Children grades 1-
3 recruited from
schools

Tooth

Child deciduous tooth;
method NR

Age at measurement: grades
1-3

Mean (SD): 4.6 (3.5) pg/g

Hyperactivity Syndrome

Boston Teacher
Questionnaire (BTQ)

Age at Outcome: Grades 1-
3

Sex, # adults at
home.

Also considered
grade, child longest
hospital stay parental
education, SES, birth
outcomes,
handedness,
language at home,
and prenatal maternal
medicine, alcohol,
and smoking.

OR vs. <2.3 pg/g as
reference

2.3-7 pg/g: 1.9 (0.53, 6.5)
>7 pg/g: 2.8 (0.68, 12)

External Review Draft

3-346

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
Clsa

Chandramouli et al.
(2009)

Avon
U.K

10% random
subsample of Avon
Longitudinal Study
of Parents and
Children (ALSPAC)

n =488

Jul.-Dec. 1992 (birth) schoolchildren
Followed through 8 yr

Cohort

Blood

Earlier childhood venous
blood; AAS using micro
sampling flame atomisation

Age at measurement:
30 mo

Mean (SD): NR
Group 1: 0-<2 |jg/dL

Group 2: 2-<5 [jg/dL

Group 3: 5-<10 [jg/dL

Group 4: >10 pg/dL

Parent and teacher rated
hyperactivity and attention

SDQ (7 yr), Development
and Well-Being Assessment
(DAWBA) (8 yr), Test of
Everyday Attention for
Children (TEACh) (8 yr)

Age at outcome:

7-8 yr

Maternal education
and smoking, home
ownership, home
facilities score, family
adversity index,
paternal SES,
parenting attitudes at
6 mo, child sex. Also
considered child IQ.

OR for increased score:
TEACh

Group 1: reference
Group 2: 1.03(0.66, 1.61)
Group 3: 0.99 (0.62, 1.57)
Group 4: 1.14(0.54, 2.40)

SDQ hyperactivity
Group 1: reference
Group 2: 0.84 (0.47, 1.52)
Group 3: 1.25 (0.67, 2.33)
Group 4: 2.82 (1.08, 7.35)

tSioen et al. (2013) Flemish Health and Blood

Flanders
Belgium

Oct. 2002 - Dec. 2003
(enrollment)

Followed through June
2011

Cohort

Environment Study
(FLEHS 1)
n: 270

Birth cohort of
Flemish children
living in either rural
or urban areas

Cord blood, HR-ICP-MS

Age at measurement:
Delivery

median = 14.3 pg/L
75th: 25.3 pg/L

Hyperactivity

SDQ with 5 domains:
emotional, conduct,
hyperactivity, peer and
social problems

Age at outcome:

7-8 yr

Maternal and paternal
BMI, maternal age,
weight increase of
mother during
pregnancy, smoking
during pregnancy,
smoking behavior of
maternal

grandmother before
birth of mother,
parental education,
current parental
smoking, child sex,
serious infections of
child since birth (also
tested interaction by
sex)

OR per doubling of log-
transformed Pb:

Hyperactivity: 2.940 (1.172,
7.380)d

Total difficulties: 2.167
(0.741, 6.334)d

External Review Draft

3-347

DRAFT: Do not cite or quote

-------
RefereDCesignnd	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
Clsa

tFruhetal. (2019)

Eastern

Massachusetts

U.S.

1999-2002
(enrollment)

Followed through age
7 yr

Project Viva
n: 1006

Birth cohort of
mother-child pairs

Blood

Maternal venous blood; ICP-
MS

Age at measurement:

T2

Median: 1.1 pg/dL

Parent teacher ratings of
hyperactivity using SDQ

Standardized for child age
and sex

Age at outcome:

7 yr

Maternal 2nd
trimester Hg and Mn
levels, nulliparity,
smoking during
pregnancy, IQ, and
education; paternal
education; HOME
composite score and
household income;
and child
race/ethnicity

Beta per In-transformed Pb
for hyperactivity:

SDQ-parent: 0.10 (-0.21,
0.41)

SDQ-teacher: 0.20 (-0.24,
0.64)

Cohort

tHorton et al. (2018) ELEMENT Project Teeth

Mexico City
Mexico

born 1994-2006 and
followed through age
6-16
Cohort

n: 133

healthy, low to
moderate income
mother (18-39 yr
old)-child pairs

tooth Pb (prenatal, postnatal
metrics derived); laser ablation
ICP-MS

Age at measurement:
tooth Pb concentration
corresponded to prenatal and 8-11 yr old
300 days after birth

Externalizing behavior
(attention and hyperactivity)

BASC-2: BSI, hyperactivity
and attention symptoms

Age at outcome:

Maternal age at
delivery, maternal
education, smoking,
maternal IQ

Beta per In-transformed Pb:
Attention: 0.19 (0.02, 0.37)e
BSI (composite): 0.22 (0.06,
0.38)e

Figure 1c

External Review Draft

3-348

DRAFT: Do not cite or quote

-------
RefereDCesignnd	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
Clsa

tRasnick et al. (2021) CCAAPS
n: 263

Cincinnati, OH

Born: Oct 2001 —Jul
2003

Exposure: 2001-2005
Cohort

Air

Attention problems using
BASC-2

LURF, air sampling at 24 sites

(C-V R2 = 0.89), predicted air Age at outcome: 12 yr

concentration at child's

residence.

Children residing>1,500 m or
<400 m from major highway
eligible.

Maternal education,
community-level
deprivation, blood Pb
concentrations,
greenspace, and
traffic related air
pollution.

Beta per 1 ng/m3 increase in
monthly air Pb exposure:

Attention: 0.8 (0.1, 1.5)

Median: 0.51 ng/m3 (range 0-
10.8 ng/m3)

tLiuetal. (2014b)

Jintan, Jiangsu

province

China

Sep. 1, 2004 - Apr.
30, 2005 (age 3-5 yrs)
Followed to age 6 yrs

Cohort

China Jintan Child Blood
Cohort Study
n: 1025 children

Chinese preschool
children

Venous child blood; GFAAS
Age at measurement:
3-5 yr old

Mean (SD): 6.4 (2.6) pg/dL
median = 6.0 (ig/dL
75th: 7.5 (ig/dL
90th: 9.4 (ig/dL
Max: 32 (ig/dL

Attention and ADHD
problems

CBCL (Chinese version);
Caregiver-Teacher Report
Form; normalized T scores

Age at outcome:

6 yr

Age at BLL test, sex, Beta:

preschool residence,
father's educational
level, mother's
educational level,
father's occupation,
parents' marital
status, single child
status, and child IQ

CBCL

Attention

0.002)

ADHD: 0.

0.386)

C-TRF

Attention

0.002)

ADHD: 0.
0.322)

0.001 (-0.002,
136 (-0.115,

0.001 (-0.002,
073 (-0.177,

OR:

C-TRF

ADHD all: 1.08 (0.99, 1.18)
Boys: 1.04 (0.94, 1.16)

Girls: 1.15 (0.98, 1.35)

External Review Draft

3-349

DRAFT: Do not cite or quote

-------
RefereDCesignnd	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
Clsa

tWinter and Sampson PHDCN

Blood

(2017)

Chicago, Illinois
U.S.

1995-1997 (birth)

Followed through 2013
(age 17 yrs)

Cohort

n: 254

Avg BLL before age 6;
Children and	methods NR

caregivers living in Age at measurement:
Chicago	6 yr old or younger

Mean: 6.4 [jg/dL

Impulsivity score

CBCL PC questionnaire

Age at outcome:

Mean: 17 yr old

Age at CBCL
assessment, sex,
race/ethnicity,
primary caregiver
immigrant
generational status,
marital status,
education level,
Temporary
Assistance for Needy
Families receipt, and
the proportion of
residential

neighborhood that is
non-Hispanic Black,
Hispanic, below the
poverty line, and
tested for Pb
exposure

Beta: CBCL

Impulsivity: 0.06 (0.005,
0.115)

tChoi etal. (2016)

10 Cities
South Korea

2006-2010
(enrollment at 1st -2nd
grade)

Followed through age
7-9 yrs

Cohort

CHEER
n: 2195

Elementary school
children

Blood

Child venous blood; AAS with
Zeeman background
correction

Age at measurement:

7-9 yrs

GM: 1.56 [jg/dL

ADHD symptomology

DuPaul's ADHD rating scale
per DSM-IV

Age at outcome:

After age 7-9 yr

Age, sex, residential
area, monthly
household income,
parental marital
status, family history
of psychiatric
disorders (anxiety
disorder, ADHD,
autism and
schizophrenia),
preterm birth and
birth weight

RR (BLL >2.17 vs. <2.17
|jg/dL) for ADHD symptoms:

1.552 (1.002, 2.403)

Single parent home and BLL
>2.17 pg/dLvs. 2-parent
home and BLL <2.17 [jg/dL:
3.567 (1.595, 7.980)

External Review Draft

3-350

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
Clsa

tBoucher et al.
(2012b)

Nunavik, Arctic

Quebec

Canada

1993-2000

(enrollment)

Sep. 2005-Feb. 2010

(follow-up)

Cohort

Cord Blood
Monitoring Program
and Environmental
Contaminants and
Child Development
Study
n: 279

Inuit Children

Blood

Cord and child venous blood;
AAS (cord), ICP-MS (child)

Age at measurement:
delivery (cord), 11.3 yr (child)

Mean (SD): 4.7 (3.3) pg/dL
(cord); 2.7 (2.2) pg/dL (child)

Median: 3.7 pg/dL (cord); 2.1
pg/dL (child)

Max: 20.9 pg/dL (cord); 12.8
pg/dL (child)

ADHD symptomology
assessed using the TRF
from CBCL and the DBD
rating scale

Age at outcome:
11.3 yr (average)

Child age and sex,
SES, age of the
biological mother at
birth, maternal
tobacco use during
pregnancy, and birth
weight, Hg

Cord Blood:

Attention problems Beta
(95% CI) per log-
transformed Pb: 0.05 (-0.10,
0.19)d

OR (95% CI)

ADHD inattentive type

1st fertile referent

2nd fertile 2.77 (1.00, 7.65)d

3rd fertile 2.87 (1.04, 7.94)d

ADHD hyperactive-impulsive
type

1st fertile referent

2nd fertile 0.95 (0.30, 3.00)d

3rd fertile 2.92 (1.07, 8.04)d

Child Blood:

Attention problems Beta
(95% CI) per log-
transformed Pb: 0.08 (-0.05,
0.21 )d

OR (95% CI)

ADHD inattentive type

1st fertile referent

2nd fertile 1.06 (0.42, 2.66)d

3rd fertile 1.01 (0.38, 2.64)d

ADHD hyperactive-impulsive

type

1st fertile referent

2nd fertile 4.01 (1.06, 15.23)d

3rd fertile 5.52(1.38, 22.12)d

External Review Draft

3-351

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
Clsa

tDesrochers-Couture
etal. (2019)

Nunavik, Northern

Quebec

Canada

Nov. 1993-Mar. 2002
(enrollment)
Sep. 2005-Feb. 2010
(1st follow-up)

Jan. 2013-Feb. 2016
(2nd follow-up)

Cohort

NCDS-childhood
n: 212

Inuit children from
14 coastal villages
in Nunavik,

Quebec, subsample
from the Cord Blood
Monitoring Program
and NIH-infancy
study

Blood

Cord and child venous blood;
GFAAS (cord), ICP-MS (child)

Age at measurement:

Delivery (cord), 11.4, 18.5 yr
(child)

GM (GSD): 3.80 (1.84) pg/dL
(cord); 2.34 (1.86) pg/dL
(child); 1.63 (2.00) pg/dL
(adolescent)

Median:3.73 pg/dL (cord); 2.07
pg/dL (child); 1.52 pg/dL
(adolescent)

Max: 17.80 pg/dL (cord);
12.83 pg/dL (child); 18.13
pg/dL (adolescent)

Teacher-rated ADHD
symptomology

Teacher assessed DBD and
TRF, Achenbach's YSR,
BAARS

Age at outcome:

11.4, 18.5 yr (average)

Child age, sex, SES,
maternal age at
delivery, maternal
tobacco smoking
during pregnancy,
and birth weight

Beta (95% CI):

Child Blood:

Child externalizing behavior:

0.23 (0.08, 0.38)

Child ADHD: 0.45 (0.13,
0.78)

Direct effect: 0.09 (-0.11,
0.28)

Indirect effect: -0.02 (-0.06,
0.03)

Adolescent externalizing
behavior mediated through
child externalizing behavior
((3: 0.09, 95% CI: 0, 0.17)

tHonq etal. (2015)

5 administrative
regions
South Korea

Study years NR
Case-control

n: 1001

General population
of children in 3rd to
4th grades

Blood

Child venous blood; GFAAS
with Zeeman background
correction

Age at measurement:
8-11 yr

Median: 1.81 pg/dL
75th: 2.25 pg/dL,

95th: 3.01 pg/dL
Max: 6.16 pg/dL

ADHD symptomology

Teacher/parent ratings
ADHD symptoms (ADHD-
RS); CPT

Age at outcome:

8-11 yr

Age, gender,
residential region,
paternal education
level, and yearly
income Iog10-
transformed blood
Hg, Mn, urine
concentrations of
cotinine, phthalate
metabolites full-scale
IQ

Beta (95% CI) Child blood:

ADHD-RS, parent-rated
1.04 (0.18, 1.90)
ADHD-RS, teacher-rated
1.90 (0.74, 3.05)

Additionally adjusted for
FSIQ, Mn, and Hg:

ADHD-RS, parent-rated
0.68 (-0.20, 1.56)

ADHD-RS, teacher-rated
1.49 (0.32, 2.67)

External Review Draft

3-352

DRAFT: Do not cite or quote

-------
RefereDCesignnd	Study Population Exposure Assessment

Outcome

Confounders Effect Est,™f?s and 95%

tNiqq etal. (2016)

Michigan
United States

Study years NR

Case-control

n: 386 children, 6-
17 yrold from 267
families (148
singletons, 119
sibling pairs)

Non-ADHD: 147
ADHD: 122

Blood

Child venous blood; ICP-MS
Age at measurement: 6-17 yr

Non-ADHD: mean (SD) = 0.74
(0.35) [jg/dL

ADHD: mean (SD) = 0.94
(0.52) [jg/dL

Composite parent and
teacher ratings of ADHD
symptoms using 3 scales:

ADHD-RS: inattention and
hyperactivity-impulsivity
symptom scores

CRS-R: cognitive
(inattention) and
hyperactivity problems
subscales

Gross annual income,
HFE mutations, race,
parenting behavior,
OD/CD, Fe
hemoglobin level, sex

SWAN: inattention and
hyperactivity symptom
scores

Age at outcome: 6-17 yr

Betas of hyperactivity-
impulsivity scores perz-
score increase in Pb
modified by HFE C282Y
mutation

Parent ratings:

Mutation: 0.74 (0.52, 0.96)e
Wild-type: 0.28 (0.15,
0.41)e

Male: 0.31 (0.14, 0.48)e
Female: 0.09 (-0.16, 0.34)e

Teacher ratings:

Mutation: 0.47 (0.22, 0.72)e
Wild-type: 0.29 (-0.04,
0.12)e

Male: 0.19 (0.07, 0.31 )e
Female: 0.11 (-0.03, 0.25)e

External Review Draft

3-353

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
Clsa

tJoo etal. (2018)

Seoul, Ulsan,
Cheonan
South Korea

2006-2011
(enrollment)

Followed through 5 yr
Cohort

MOCEH

n: 575 mother-child
pairs

Pregnant women at
12-20 wk of
pregnancy in
prenatal clinics and
public health
centers

Blood

Maternal venous, cord, and
child venous blood; AAS

Age at measurement: 20 wk
(maternal); delivery (maternal
and cord); 2, 3, and 5 yr (child)

GM: Maternal 1.28 [jg/dL
(early), 1.24 (late) 0.9 (cord);
Child 1.55 (age 2), 1.43 (age
3), 1.29 (age 5)

Attention and aggressive
behavior combined

K-CBCL: Externalizing
behavior (attention and
aggressive behavior
combined);

Age at outcome:

5 yr

Maternal age at
childbirth, parity,
maternal educational
level, household
income, residential
area, and
breastfeeding

Beta (95% CI):

Externalizing behavior at
5 yr

Maternal-early pregnancy
Male: -0.72 (-3.12, 1.69)
Female: -0.45 (-2.16, 1.26)
Maternal-late pregnancy
Male: 2.99 (0.55, 5.43)
Female: 0.24 (-2.18, 2.66)

Cord blood

Male: 3.09 (-0.08, 6.26)
Female: -0.16 (-3.33, 3.01)

Child blood-2 yr

Male: 0.55 (-1.52, 2.62)

Female: 3.50 (0.97, 6.03)

Child blood-3 yr

Male: 1.13 (-1.42, 3.68)

Female: 2.05 (-1.35, 5.45)

Child blood-5 yr
(concurrent)

Male: 1.42 (-2.12, 4.95)
Female: 4.53 (-0.81, 9.86)

External Review Draft

3-354

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
Clsa

tJi et al. (2018)

Boston,

Massachusetts
U.S.

1998-2013
(enrollment)

Followed through 2016

Cohort

Boston Birth Cohort Blood
n: 299 ADHD
cases, 1180
neurotypical
controls

Mother-infant pairs

Child blood; method NR,
obtained from electronic
medical records

BLLs

Age at measurement:
<4 yr; the earlier BLL was
selected where multiple BLLs
were recorded

Mean (SD): 2.2 (1.6) pg/dL

Diagnosed ADHD

Physician-diagnosed ADHD
from electronic medical
records

Age at outcome:

Median: 6 yr

Maternal age at
delivery, maternal
race/ethnicity,
maternal education,
smoking during
pregnancy,
intrauterine infection,
parity, child's sex,
mode of delivery,
preterm birth, and
birth weight

OR (95% CI)

Continuous BLL: 1.118
(1.003, 1.247)

Categorical BLL:

2-4 vs. <2 (ig/dL: 1.08
(0.81, 1.44)e
5-10 vs. <2 pg/dL: 1.73
(1.09, 2.73)e

Sex-stratified:

Girls 5-10 vs. <5 (ig/dL:
0.68 (0.27, 1.69)e
Boys 5-10 vs. <5 (ig/dL:
2.49 (1.46, 4.26)e
Joint Effects of sex and
BLL category:

Girls*5-10 (ig/dL: 0.69
(0.28, 1.7 l)e

Boys*<5 (ig/dL: 3.02 (2.24,
4.06)e

Boys*5-10 pg/dL: 7.48
(4.29, 13.02)e

tParketal. (2016)
Busan

South Korea
Apr.-Sep. 2013
Case-control

n: 114 cases
(diagnosed ADHD),
114 controls

Recruitment from
child psychiatric
and pediatric clinics
from four university
hospitals

Blood

Child venous blood; GFAAS
with Zeeman background
correction

Age at measurement:

6-12 yr

GM (GSD): 1.90 (0.86) pg/dL
(cases); 1.59 (0.68) pg/dL
(controls)

Diagnosed ADHD

Diagnosed ADHD
(confirmed by [K-SADS-PL-
K]); CPT and parent-rated
ADHD symptoms among
ADHD cases

Age at outcome:

6-12 yr

Age, sex-matched OR (95% CI) per log-

controls; gestational
age, birth weight,
SES, parental
education, and
parents' smoking
behavior

transformed Pb:
ADHD total
1.60 (1.04-2.45)d

External Review Draft

3-355

DRAFT: Do not cite or quote

-------
RefereDCesignnd Study P°P"'ation

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
Clsa

tKimetal. (2013a)

Omaha, Nebraska
U.S.

Aug 2007-Dec 2009
Case-control

n: 71 cases
(diagnosed ADHD);
58 controls

Children living near
a former refinery

Blood

Child venous blood; ICP-MS

Age at measurement:

5-12 yr

GM: 1.29 [jg/dL (cases); 1.33
[jg/dL (controls); 1.65 [jg/dL
(inside Pb investigation area);
1.01 |jg/dL (outside Pb
investigation area)

ADHD

Physician-diagnosed
according to DSM-IV

Age at outcome:
5-12 yr

Matched on age, sex, OR (95% CI) per In
race and adjusted for transformed Pb
maternal smoking,

SES, and
environmental
tobacco exposure

ADHD Overall 2.52 (1.07-
5.92)e

tGeieretal. (2018)

NHANES

Blood

ADD

Sex, age, SE

ES, race OR (95% CI):

Representative sample
U.S.

n: 2109
Children

Child venous blood: ICP-MS

Self-reported doctor
diagnosed ADD



ADD 1.292 (1.025-1.545)

2003-2004



Age at measurement:
10-19 yr

Age at outcome:
10-19 yr





Cross-sectional



Mean (SD): 1.16 (1.27) pg/dL







External Review Draft

3-356

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
Clsa

Braun et al. (2006)

Representative sample
U.S.

1999-2002
Cross-sectional

NHANES
n: 4704

Children 4-15 yr

Blood

Venous blood: GFAAS
Age at measurement:
4-15 yr old

Parent-reported ADHD with Age, sex, race,

prescription stimulant use
Age at outcome: 4-15 yr old

Quintiles:
ND-0.7 [jg/dL

0.8-1.0 [jg/dL

1.1-1.3 Mg/dL

1.4-2.0 [jg/dL

679
795
857
745

prenatal ETS
exposure, postnatal
ETS exposure, BLLs,
preschool or
childcare attendance,
health insurance
coverage, and ferritin
levels

AOR (95% CI) Child blood:

2nd quintile (0.8-1.0): 1.1
(0.4-3.4)e

3rd quintile (1.1-1.3): 2.1
(0.7-6.8)e

4th quintile (1.4-2.0): 2.7
(0.9-8.4)e

5th quintile (>2.0): 4.1 (1.2-
14.0)e

>2.0 [jg/dL: 995

AAS = atomic absorption spectrometry; ADHD = attention deficit/hyperactivity disorder; ADHD-RS = ADHD rating scale; ADRA2A = adrenoceptor alpha 2A; AOR = adjusted odds
ratio; BAARS = Barkley Adult ADHD-IV Rating Scale; BASC = Behavior Assessment System for Children; BLL = blood lead level; BMI = body mass index; BRIEF = Behavior Rating
Inventory of Executive Functions; BSI = Behavioral Symptoms Index; CARES = Communities Actively Researching Exposure Study; CBCL = Child Behavior Check List;
Cd = cadmium; CHEER = Children's Health and Environment Research; CI = confidence interval; CKiD = Chronic Kidney Disease in Children Study; CPT = Continuous Performance
Test; CRS-R = Conners' Rating Scale-Revised; C-TRF = Caregiver-Teacher Report Form; DAT1 = dopamine transporter; DBD = Disruptive Behavior Disorders; DRD2 = dopamine
receptor D2; DSM = Diagnostic and Statistical Manual of Mental Disorders; EEG = electroencephalogram; ELEMENT = Early Life Exposures in Mexico to Environmental Toxicants;
ERP = event-related potentials; FBB-ADHS = Fremdbeurteilungsbogen fur Aufmerksamkeitsdefizit/HyperaktivitatstSrungen; Fe = iron; FLEHS = Flemish Health and Environment
Study; GFAAS = graphite furnace atomic absorption spectrometry; GM = geometric mean; GMR = geometric mean ratio; HFE = hemochromatosis; Hg = mercury; HOME = Health
Outcomes and Measures of the Environment; ICP-MS = inductively coupled plasma mass spectrometry; K-ABC = Kaufman Assessment Battery For Children; K-CPT = Conners'
Kiddie Continuous Performance; KiTAP = Test of Attentional Performance for Children; K-SADS-PL-K = Kiddie Schedule for Affective Disorders and Schizophrenia Present and
Lifetime - Korean Version; LURF = Land Use Random Forest; Mn = manganese; mo = month(s); MOCEH = Mothers' and Children's Environmental Health; NCDS = Nunavik Child
Development Study; NR = not reported; OD/CD = oppositional defiant and conduct disorder; OR = odds ratio; Pb = lead; PCBs = polychlorinated biphenyls; PHDCN = Project on
Human Development in Chicago Neighborhoods; RR = relative risk; RT = ; SCWT = Stroop Color-Word Test; SD = standard deviation; SDQ = Strengths and Difficulties
Questionnaire; SE = standard error; SES = socioeconomic status; SPM = Standard Progressive Matrices; SRS = Social Responsiveness Scale; SWAN = Strengths and Weaknesses
of ADHD Symptoms and Normal Behavior Scale; T1 = first trimester of pregnancy; T2 = second trimester of pregnancy; T3 = third trimester of pregnancy; TEACh = Test of Everyday
Attention for Children; TSCD = Tohoku Study of Child Development; WAIS = Wechsler Adult Intelligence Scale; wk= week; WMS = Weschler Memory Scale; yr = year(s).

aEffect estimates are standardized to a 1 |jg/dL increase in BLL or a 10 |jg/g increase in bone Pb level, unless otherwise noted. For studies that report results corresponding to a
change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the biomarker and standardized accordingly.

bResults are unstandardized because they did not have an associated SE, CI, or p-value reported in the study.

The CI was calculated from a p-value and the true CI may be wider or narrower than calculated.

dResults are unstandardized because the log base used for exposure transformation was unspecified in the study.

eResults are unstandardized because the Pb level distribution data was not available.

fStudies published since the 2013 Integrated Science Assessment for Lead.

External Review Draft

3-357

DRAFT: Do not cite or quote

-------
Table 3-7T Animal toxicological studies of Pb exposure and externalizing and internalizing behaviors.

Study

Species (Stock/Strain), n, Sex

Timing of Exposure

Exposure
Details

BLL as Reported
(Hg/dL)

Endpoints Examined

Externalizing Behavior

Tartaqlione etal. (2020)

Rat (Wistar)

Control (tap water), M/F n = 16
(9/7)

50 mg/L, M/F, n = 16 (9/7)

GD 28 to PND 23

Oral,
lactation
In utero

PND 23:

0.007 |jg/ml_ (0.7 pg/dL)
for Control

0.255 pg/mL

(25.5 pg/dL) for 50 mg/L

PND 4, 7, 10, 12: Ultrasonic
Vocalizations

Internalizing Behavior

Corv-Slechta et al.
(2013)

Mice (C57BL/6)

Control (distilled deionized
water) - NS, M/F, n = 8-16

Control (distilled deionized
water) - prenatal stress (PS),
M/F, n = 8-16

100 ppm (NS), M/F, n = 8-16

GD -60 to 12 mo

Oral,

drinking

water

Oral,

lactation

In utero

PND 75 - Females:

-------
Study	Species (Stock/Strain), n, Sex Timing of Exposure	BLL funMn'1'*0^	Endpoints Examined

-------
Study

Species (Stock/Strain), n, Sex

Timing of Exposure

Exposure
Details

BLL as Reported
(Hg/dL)

Endpoints Examined



Control (RO Dl water), M/F,
n =10-11

10 pg/mL, M/F, n = 10-11





9.21 ng/g (0.98 pg/dL)
for 10 pg/mL

PND 60:

0.23 ng/g (0.024 pg/dL)
for Control

0.30 ng/g (0.032 pg/dL)
for 10 pg/mL



Faulk etal. (2014)

Mice (Agouti)

Control (distilled water), M/F,
n = 30

2.1 ppm, M/F, n = 28
16 ppm, M/F, n = 33
32 ppm, M/F, n = 29

GD -14 to PND21

Oral,
lactation
In utero

PND 21 (Maternal BLL):
-------
Study	Species (Stock/Strain), n, Sex Timing of Exposure	BLL funMn'1'*0^	Endpoints Examined

solution
18 mo:

0.12 [jg/dL for Control

11.2 [jg/dL for 0.2%
solution

Mansouri et al. (2012)

Rat (Wistar) PND 70 to PND 100
Control (distilled water), M/F,
n = 16 (8/8)

50 mg/L, M/F, n = 16 (8/8)

Oral,

drinking

water

PND 100-Males: PND 100: OFT

2.05 [jg/dL for Control

8.8 [jg/dL for 50 mg/L

PND 100 - Females:

2.17 [jg/dL for Control

6.8 [jg/dL for 50 mg/L

Duan et al. (2017)

Mice (CD1) PND 1 to PND 21

Oral,

PND 21: PND 7, 11, 15, 19: TST, OFT



Control (distilled water), M/F,

lactation





n = 5



16.2 |jg/L (1.6 [jg/dL) for







Control



27 ppm, M/F, n = 5











191.8 [jg/L (19.2 pg/dL)



109 ppm, M/F, n = 5



for 27 ppm

283.4 |jg/L (28.3 |jg/dL)
for 109 ppm

PND 35:

14.3 |jg/L (1.4 [jg/dL) for
Control

283.4 |jg/L (28.3 |jg/dL)
for 27 ppm

External Review Draft

3-361

DRAFT: Do not cite or quote

-------
Study	Species (Stock/Strain), n, Sex Timing of Exposure	BLL funMn'1'*0^	Endpoints Examined









376.9 [jg/L (37.7 pg/dL)











for 109 ppm



Wanaetal. (2016)

Rat (Sprague Dawley)

PND 24 to PND 56

Oral,

PND 56:

PND 60-66: OFT



Control (tap water), M, n = 7



drinking











water

11 pg/L (1.1 pg/dL) for





100 ppm, M, n = 9





Control











133 pg/L (13.3 pg/dL) for











100 ppm



Shvachiv et al. (2018)

Rat (Wistar)

Intermittent Exposure:

Oral,

PND 196:

PND 189: OFT, EPM



Control (tap water), M/F, n = 8

GD 7 to PND 84,

drinking









PND 140 to PND 196

water

<0.1 pg/dL for Control





0.2% (p/v) solution (distilled



Oral,







water), M/F, n = 9 - Intermittent

Continuous Exposure:

lactation

18.8 pg/dL for 0.2%





exposure

GD 7 to PND 196

In utero

(Intermittent)





0.2% (p/v) solution, M/F, n = 9





24.4 pg/dL for 0.2%





- Continuous exposure





(Continuous)



Basha and Reddv (2015)

Rat (Wistar)

GD 6 to GD 21

In utero

PND 21:

PND 21, PND 28, 4 mo: OFT,



Control (deionized water), M,







Hole Board Test



n = 8





0.21 pg/dL for Control





0.2 % solution, M, n = 8





11.2 pg/dL for 0.2%











solution











PND 28:











0.33 pg/dL for Control











12.3 pg/dL for 0.2%











solution











4 mo:











0.19 pg/dL for Control



External Review Draft

3-362

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain), n, Sex Timing of Exposure

Exposure
Details

BLL as Reported
(Hg/dL)

Endpoints Examined

5.9 |jg/dL for 0.2%
solution

Stansfield etal. (2015)

Rat (Long-Evans) GD 0 to PND 50

Control (chow), M/F, n = 11-23

1500 ppm, M/F, n = 11-23

Oral, diet
Oral,
lactation
In utero

PND 50:

0.6 [jg/dL for Control
22.2 |jg/dL for 1500 ppm

PND 50: Locomotor Activity

Flores-Montova and

Mice (C57BL/6) PND 0 to PND 28

Oral,

>PND 28 Males:

>PND 28 Hole Board Test, OFT

Sobin (2015)

Control (distilled water), M/F,

drinking







n = 19 (8/11)

water

0.2 [jg/dL for Control







Oral,







30 ppm, M/F, n = 26 (16/10)

lactation

3.93 [jg/dL for 30 ppm





230 ppm, M/F, n = 16 (12/4)



9.39 [jg/dL for 230 ppm









>PND 28 Females:









0.19 [jg/dL for Control









3.19 [jg/dL for 30 ppm









12.14 |jg/dLfor230 ppm



Neuwirth et al. (2019a)

Rat (Long-Evans) GD 0 to PND 22

Control (tap water), M/F,
n = NR

150 ppm, M/F, n = NR
1000 ppm, M/F, n = NR

Oral,
lactation
In utero

PND 22:

-------
Study	Species (Stock/Strain), n, Sex Timing of Exposure	BLL funMn'1'*0^	Endpoints Examined

-------
Study

Species (Stock/Strain), n, Sex Timing of Exposure

Exposure
Details

BLL as Reported
(Hg/dL)

Endpoints Examined



F3:









see Figure 1, n = 8-10







Sinqh et al. (2019)

Rat (Wistar) 3 mo to 6 mo

Control (distilled water), M,
n = 5

2.5 mg/kg, M, n = 5

Oral, gavage

6 mo:

5.76 [jg/dL for Control
28.4 |jg/dL for 2.5 mg/kg

6 mo: EPM, Locomotor Activity

Al-Qahtani et al. (2022) Mice (Albino)	8-9 wk to 14-15 wk Oral, gavage 14-15 wk:	NR: EPM, Locomotor Activity

Control (distilled water), M,

n = 10	1.2|jg/100mL

(1.2 |jg/dL) for Control

0.2 mg/kg, M, n = 10

7.1 |jg/100 mL

(7.1 |jg/dL) for 0.2 mg/kg

BLL = blood lead level; EPM = elevated plus maze; F = female; FST = forced swim test; GD = gestational day; LOD = limit of detection; M = male; MRI = magnetic resonance
imaging; mo = month(s); NR = not reported; NS = no stress; OFT = open-field test; Pb = lead; PG = pregestation; PND = postnatal day; PS = prenatal stress; TST = tail suspension
test; wk = week(s); yr = year(s).

External Review Draft

3-365

DRAFT: Do not cite or quote

-------
Table 3-8E Epidemiologic studies of Pb exposure and performance on neuropsychological tests of attention,
impulsivity, and hyperactivity, Attention Deficit Hyperactivity Disorder-related behaviors, and
clinical Attention Deficit Hyperactivity Disorder in children; group or population mean blood Pb
level >5 [jg/cIL, any study design.

Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders Effect Estimates and 95% Clsa

tArbuckle et al. (2016a)

representative population

Canada

2007-2009

Cross-sectional

CHMS	Blood

n: 1080

Child venous blood;
Representative sample of analytic method NR
children	Age at measurement:

6-11 yrold

GM: 0.90
95th: 1.96 pg/dL

ADHD symptoms

SDQ, parent-reported
ADD/ADHD

Age at outcome:
6-11 yrold

Age, sex,	ORb

neonatal unit,	Parent-Reported Outcomes

sltkmg child	ADD/ADHD: 2.08 (1.01, 4.25)
aqe

(Supplemental	Any Learning Disability: 1.41 (0.73,

Table 1)	Z70)

Psychotropic Medicine Taken: 2.91
(1.47, 5.79)

SDQ

Total Difficulties, Prenatal Smoking:
10.57 (2.81, 39.69)

Total Difficulties, No Prenatal
Smoking: 1.98 (1.41, 2.79)

Emotional Symptoms: 1.25 (0.60,
2.59)

Hyperactivity/lnattention: 2.75
(1.46, 5.16)

External Review Draft

3-366

DRAFT: Do not cite or quote

-------
RefereDCesignnd	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tArbuckle etal. (2016b)

representative sample
Canada
2007-2009
Cross-sectional

CHMS	Blood

n: 2097

Child venous blood
Representative sample of Age at measurement:
children	6-19 yr old

ADHD symptomology

SDQ, reported ADD or
ADHD

Age at outcome:
6-19 yr old

Smoking, sex, ORb

income	Parent or Self-Reported Outcomes,

Ages 6-19

ADD/ADHD: 2.39 (1.32, 4.32)

Learning Disability (Low Income):
0.81 (0.37, 1.81)

Learning Disability (High Income):
2.78 (1.40, 5.51)

Medicine Taken (Fasting Sample):
0.83 (0.34, 2.02)

Medicine Taken (Non-Fasting):
4.20 (1.92, 9.17)

SDQ, Ages 6-17

Total Difficulties: 2.16 (1.33, 3.51)

Emotional Symptoms: 1.08 (0.68,
1.71)

Hyperactivity/lnattention: 2.33
(1.59, 3.43)

External Review Draft

3-367

DRAFT: Do not cite or quote

-------
RefereDCesignnd	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tBarq et al. (2018)

Montevideo

Uruguay

Cross-sectional

n: 206

Children living in areas
considered high risk for
metal exposure

Blood

Child venous blood;
flame AAS or GFAAS
Age at measurement:
5-8-year-old

4.2 [jg/dL

teacher-rated ADHD and
hyperactive behavior

CRS-R: hyperactive,
oppositional, cognitive,
and ADHD-like behaviors
(teacher ratings)

Age at outcome:
5-8-year-old

Child IQ, iron
status, and
BMI, blood Pb
testing method,
household
possessions,
maternal
education,
current parent
smoking

PRs

Cognitive Problems/Inattention
Total population (>5 vs 5 |jg/dL):
1.02 (0.967, 1.076)

Girls: 1.01 (0.995, 1.025)

Boys: 1.01 (0.99, 1.03)

Hyperactivity

Total population (>5 vs 5 |jg/dL):
1.01 (0.947, 1.077)

Girls: 1.02 (1, 1.04)

Boys: 0.99 (0.97, 1.01)

ADHD Index

Total population (>5 vs 5 |jg/dl_):
1.01 (0.952, 1.072)

Girls: 1.01 (0.99, 1.03)

Boys: 1.00 (0.98, 1.02)

External Review Draft

3-368

DRAFT: Do not cite or quote

-------
RefereDCesignnd	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tChan et al.
(2015)

10 locations

U.S.

Cohort

National Institute of Child
Health and Human
Development, Study of
Early Child Care and
Youth Development
n: 266

School children

Teeth (Shed molars)

ICP-OES
Mean: 0.46 |jg/g

Disruptive behavior and
ADHD subscales

TBD completed by 3rd
grade teachers; scores
for (1) Total Disruptive
Behavior; (2) subscale
scores for ADHD,
hyperactivity/impulsivity,
inattention, and OD

Race, sex,

paternal

education,

maternal

education,

marital status,

and family SES

Change in behavior score per
|jg/g of Pb concentration in
teeth:0

DBD: -0.05
ADHD: -0.03
Impulsive: -0.06
Inattention: 0.00
Defiance: -0.09

Age at outcome:
teeth collected at 8-11 yr
old (body burden)

tForns et al. (2014)

Catalonia

Spain

Cohort

INMA
n: 385

Children of mothers
enrolled in the
population-based cohort
as part of the INMA
(Environment and
Childhood) Project

Urine

Maternal urine; ICP-MS,
values below LOD were
imputed

Age at measurement:
T1, T3

Median: 3.44, 1st; 3.63
3rd

75th: 4.64 1st, 4.84

ADHD symptoms

ADHD-DSM-IV criteria
and MSCA

Age at outcome:
4 yr old

Age, maternal
social class,
and maternal
mental health

Change in neuropsychological
outcomes per ng/mL increase in
mother's urinary Pb
concentration:

T1

GCI MSCA: 1.46 (-2.76, 5.69)
EF MSCA: 0.34 (-3.95, 4.63)

T3

GCI MSCA:
EF MSCA: -

-1.27
0.74 (-

-5.71, 3.17)
¦5.24, 3.75)

IRR:

T1

Inattention: 0.92 (0.57, 1.46)
Hyperactivity: 1.04 (0.65, 1.65)

T3

Inattention: 0.71 (0.43, 1.18)
Hyperactivity: 1.04 (0.64, 1.70)

External Review Draft

3-369

DRAFT: Do not cite or quote

-------
RefereDCesignnd	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tGu etal. (2018)

Wuhan
China

Case-control

Hospital based case-
control, recruitment:
n: 389 cases; 392
controls

Children 6-18 yr old

Blood

Venous sample, whole
blood, AAS
Median: 5.685 [jg/dL

ADHD, subtypes
inattention, hyperactivity
and impulsivity and
combined

Cases: ADHD DSM-IV
(subtypes defined by
inattention, hyperactivity
and impulsivity [HI] and
combined type [C]);
WISC and Parent
Symptom Questionnaire

Age at outcome:
6-18 yr old

ORb

BLL<56.85|jg/L, GG: Reference

Age, sex
(cases and
controls
compared on
IQ, maternal
alcohol,
smoking,
parental
relationship,
breastfeeding) BLL>56.85|jg/L, GA/AA: 1.871
(1.014, 3.451)

BLL>56.85|jg/L, GG: 1.865 (1.132,
3.075)

BLL<56.85|jg/L, GA/AA: 1.255
(0.806, 1.954)

tGump etal. (2017)

Upstate New York
U.S.

Cross-sectional

Environmental
Exposures and Child
Health Outcomes
n: 203

children residing in low-
to middle-income
communities

Blood

venous blood;

Age at measurement:
9-11 yrold

Externalizing behavior:
attention, impulsivity,
hyperactivity

DBD for ADHD
inattentive type and
ADHD hyperactive-
impulsive type; ASQ:I
questionnaire for ASD
(parent-rated); acute
vagal response for stress
(heart rate variability)

Age at outcome:
9-11 yrold

Sex, race, age,
and SES and
Hg

Beta (95%)b

ADHD-Inattention (Score) 0.01
(-0.14, 0.15)

ADHD-Hyperactivity (Score) 0.16
(0.02, 0.30)

Oppositional Defiant Disorder
(Score) 0.16 (0.02, 0.31)

External Review Draft

3-370

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders Effect Estimates and 95% Clsa

tHuanq et al. (2016)

Mexico City
Mexico
1997-2001
Cross-sectional

ELEMENT
n: 578

Mother-child pairs

Blood

Venous blood; ICP-MS
Age at measurement:
6-13 yr old

Mean: 3.4 [jg/dL

ADHD symptomology

CRS-R, CRS-R DSM-IV

Age at outcome:
6-13 yr old

Maternal
marital status,
age,

educational

years,
smoking during
pregnancy,
child's age,
sex, birth
weight.

Adjusted associations between a 1-
[jg/dL increase in blood lead

Cognitive Problem/Inattention -
0.03 (-0.3, 0.2)

Hyperactivity 1.2 (0.3, 2.0)

ADHD Index 0.02 (-0.2, 0.3)
CGI Restless-Impulsive 1.2 (0.3,
2.0)

CRS-R DSM-IV
Inattentive 0 (-0.3, 0.3)
Hyperactive-Impulsive 1.1 (0.2, 2.0)
Total 0.03 (-0.2, 0.3)

tJoo et al. (2017)

Cheonan
South Korea
2008-2010
Case-Control

n: 214 cases (=19 on the
K-ARS or ADHD
diagnosis); 214 control
(49 elementary schools)

Elementary school
children

Blood

Venous blood, AA
spectrophotometry
GM: 1.65 (cases) [jg/dL;
1.49 [jg/dL (controls)

ADHD symptomology
K-ARS

Age at outcome:
6 to 10 yr old

Maternal
education,
family history
of ADHD,
parental
marital status,
and teenage
mother

ORb

All ADHD: 1.28 (0.89, 1.83)

Inattention: 1.63 (1.03, 2.58)

Hyperactivity/impulsivity: 1.04
(0.53, 2.07)

tKicinski et al. (2015)

n: 606

Blood

Sustained attention,
short-term memory,

R gender, age,
smoking,

Effect estimates between BLL and
neurobehavioral outcomes not

Flanders

Third year secondary

venous blood, ICP-MS

manual motor speed

passive

reported due lack of statistical

Belgium

school students in two

Age at measurement:



smoking,

significance

2008 and 2011

industrial areas in

13.6-17 yr old

CPT, NES

household



Cross-sectional

Flanders, Belgium

Mean: 13.8 [jg/dL
95th: 28.1 pg/dL

Age at outcome:
13.6-17 yr old

income per
capita, the
highest
occupational
category of
either parent,
and the

education level
of the mother



External Review Draft

3-371

DRAFT: Do not cite or quote

-------
RefereDCesignnd	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tLinetal. (2019)

n: 164

Blood, Bone

ADHD symptoms and
comorbidities

Children's age;
sex, passive

Xinhua

Children who visited a

Child venous blood; AAS



smoking (the

China

lead specialty clinic in

Tibia bone; XRF

Vanderbilt-ADHD

frequency of

Aug. 2014 - Aug. 2015

Xinhua Hospital from

Age at measurement:

Diagnostic-Parent-Rating

smoking by

Cross-sectional

August 2014-August

3-15 yr

Scale

parents and



2015



other





GM:

Age at outcome:

household





Blood:

3-15 yr

members in







the presence





Low: 4.3 pg/dL



of children),





High: 19.6 pg/dL



parity,





Bone:



maternal







education





Low: 0.3 pg/g



levels and





High: 12.8 pg/g



family yearly
income

ORb

Inattention

BLL <10 [jg/dL: Reference
BLL >10 [jg/dL: 3.3 (0.9, 12.4)

Hyperactivity/impulsivity
BLL <10 [jg/dL: Reference
BLL >10 [jg/dL: 2.0 (0.5, 7.5)

Oppositional defiant disorder
BLL <10 [jg/dL: Reference
BLL >10 [jg/dL: 2.7 (0.8, 8.9)

tLiuetal. (2014e)

Guiyu
China

2009 -2011
Cross-sectional

n: 240

Native 3-7 year old
kindergarten children
who have resided in
Guiyu for more than 2
years after birth

Blood

Child venous blood;
GFAAS

Age at measurement:
3-7 yr

Median: 7.33 [jg/dL
75th: 9.13 pg/dL

ADHD symptomology

ADHD (H,I,C) perDSM-
IV; CPRS-R, CTRS-R,
Rutter Child Behavior
Questionnaire (antisocial
behavior, neurotic)

Age at outcome:
3-7 yr

Age and
gender,
residential site,
time, heavy
metal exposure

Measures of association for blood
lead and attention outcomes not
reported. Study only reports
correlation analyses.

External Review Draft

3-372

DRAFT: Do not cite or quote

-------
RefereDCesignnd	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tLucchini et al. (2012)

Valcamonica and Garda
Lake areas in Province of
Brescia
Italy

Cross-sectional

Junior high school-age
children from 20 local
public schools
n: 299

Blood

Child venous blood;
GFAAS

Age at measurement:
11-14 yr

1.71 pg/dL, Median: 1.50
75th: 2.10 pg/dL
Max: 10.2 pg/dL

Conners-Wells'
Adolescent Self-Report
Scale Long Form

10 subscales: family
problems, emotional
problems, conduct
problems, cognitive
problems/inattention,
anger control problems,
hyperactivity, ADHD
index, DSM-IV
(disattention), DSM-IV
(hyperactivity/impulsivity)
, and DSM-IV (Total)

Age at outcome:

11-14 yr

Sex, age at
testing,
parental
education,

3, family
parity
order, BMI

size,

Betas

Performance IQ: -1.991 (-3.918,
-0.064)

Verbal IQ: -1.863 (-3.79, 0.064)

Total IQ (Table 4): -2.237 (-4.101,
-1.372)

Total IQ (Table 5): -2.248 (-4.111,
-0.385)

tMunoz et al. (2020)

Arica
Chile

2009-2015
Cross-sectional

n: 2656

Children enrolled in a
heavy metal intervention
program

Blood

Child venous blood; AAS
Age at measurement:
3-17 yr

Median: 1.0 pg/dL
75th: 2.0 pg/dL

Parent-reported attention
deficit and hyperactivity
recorded in medical
records

Age at outcome:

3-17 yr

Age, sex,
parents' report
of children
exposure to
secondhand
tobacco
smoke,
housing
material quality

ORb

BLL >5 pg/dL: 2.33 (1.32, 4.12)

tRodriques et al. (2018) n: 225

Blood

Salvador, Bahia
Brazil

Cross-sectional

Children living near alloy GFAAS
plant	Age at measurement:

7-12 yr

1.2 pg/dL
Max: 15.6 pg/dL

Child behavior

CBCL: 8 domains
including attention

Age at outcome:
7-12 yr

Sex, age,

height-for-age

Z-score,

maternal

schooling,

socioeconomic

classification,

and community

violence index,

as well as

maternal IQ

Change in Behavior (Total raw
score) per iog-fjg/dL decrease in
BLL:

-1.08 (-11.5, 9.3)

Change in Behavior (Total score T)
per log-ijg/dL decrease in BLL:
-0.74 (-5.3, 3.8)

External Review Draft

3-373

DRAFT: Do not cite or quote

-------
RefereDCesignnd	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tSkoqheim et al. (2021)

Norway
2002-2009

Case-control

Norwegian Mother,
Father and Child Cohort
Study (MoBa)
n: 397 ASD cases, 1034
controls

Children

Blood

Maternal whole blood;
ICP-SFMS atwk 17 of
gestation

Age at Measurement:
Prenatal, Week 17 og
gestation

GM (95% CI) (cases):
0.835 (0.797, 0.875)
pg/dL

GM (95% CI) (controls):
0.882 (0.860, 0.905)
pg/dL

ADHD

Diagnosis of ADHD
(NPR)

Age at outcome: 3 or less

Child sex, birth
weight, birth
year, and SGA,
maternal age
at delivery,
education,
parity, pre-
pregnancy
BMI, kg/m2),
self-reported
smoking and
alcohol intake
during
pregnancy,
FFQ-based
estimates of
seafood intake
(g/day), and
dietary iodine
intake (pg/day)

ORb

ADHD
Q1 (Reference):
Q2: 1.15 (0.87
Q3: 0.84 (0.63
Q4: 1.09 (0.82

1.52)
1.12)
1.45)

tSobin et al. (2015)
U.S.

Cross-sectional

n: 421

Elementary school
children

Blood

Attention

2 samples 60 days apart Age at outcome:
averaged; ICP-MS or Pb 5.1-11.8 yr old
Care I

Age at measurement:

5.1—11.8 yr old

Mean: 2.7 [jg/dL (males);

2.4 [jg/dL (females)

Sex, age and
mother's level
of education

Beta

Motor dexterity non-dominant hand
time (s): 1.93 (-1.343, 5.203)

Working memory misses: 0.11
(0.051, 0.169)

Working memory false alarms
errors: 0.05 (-0.009, 0.109)

Visual attention 5-choice movement
time (ms): 26.07 (11.331, 40.809)

External Review Draft

3-374

DRAFT: Do not cite or quote

-------
RefereDCesignnd	Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tSoetrisno and Delqado-
Saborit (2020)

West Java (Depok,

Bogor and Bekasi)
Sukatani village (control)
Indonesia
Cross-sectional

School children living in
urban locations near e-
waste facility; control site
n: 44 (22 from Bogor and
22 from Sukatani)

Children selected from
schools per teachers/
principal

recommendation

Hair, soil, water

hair samples from
children in Bogor and
Sukatani village. BLLs
from 36 children in Bogor
area (2010).

Age at measurement:
6-9 yr

Soil Pb mean: Depok-
Bekasi: 3653 mg/kg;
Sukatani: 93.2 mg/kg;
Water Pb: all 10 samples
below LOD; Hair Pb:
Depok-Bekasi:
0.155 mg/g; Sukatani:
0.0729 mg/kg
Max: Soil Pb: Depok-
Bekasi: 7662 mg/kg;
Sukatani: 115 mg/kg;

Hair Pb: Depok-Bekasi:
0.841 mg/g; Sukatani:
0.255 mg/kg

Visual attention TMT A
Age at outcome:

6-9 yr

Age, parental
education,
environmental
tobacco smoke
at home, and
residential
traffic exposure

Change in TMT-A (seconds) per
mg/g unit of hair Pb
2.5 (-55, 60)

External Review Draft

3-375

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tZhanq et al. (2015a)

n: 243

Blood

ADHD symptomology

Age, sex,

OR









father's work in

ADHD: 2.4 (1.1, 5.2)

Guangdong

Preschool children

Child venous blood,

Parent rating per DSM-IV

e-waste,

China

residing near e-recycling

GFAAS

ADHD criteria

Serum ferritin,



Jan. 2012-May 2012

plant

Median: 7.9 [jg/dL



E-waste



Cross-sectional



95th: 16.9 pg/dL

Age at outcome:

workshops









3-7 yr

around the











house



ADD = attention deficit disorder; ADHD = attention deficit/hyperactivity disorder; ASD = autism spectrum disorder; ASQ:I = Ages and Stages Questionnaire
Inventory; BLL = blood lead level; BMI = body mass index; BRIEF = Behavior Rating Inventory of Executive Functions; CBCL = Child Behavior Check List;

CHMS = Child Health Monitoring System; CI = confidence interval; CPRS-R = Conners Parent Rating Scale-reformed; CPT = Continuous Performance Test;
CTRS = Conners' Teacher Rating Scale; DBD = Disruptive Behavior Disorders; DSM = Diagnostic and Statistical Manual of Mental Disorders; ELEMENT = Early
Life Exposures in Mexico to Environmental Toxicants; FFQ = Food Frequency Questionnaire; GFAAS = graphite furnace atomic absorption spectrometry;
GM = geometric mean; Hg = mercury; ICP-MS = inductively coupled plasma mass spectrometry; ICP-OES = inductively coupled plasma optical emission
spectrometry; ICP-SFMS = inductively coupled plasma sector field mass spectrometry; INMA = Infancia y Medio Ambiente (Environment and Childhood);
IQ = intelligence quotient; K-ARS = Korean ADHD Rating Scale; LOD = limit of detection; mo = month(s); NPR = Norwegian Patient Registry; NR = not reported;
OFT = open-field test; OD = oppositional defiant; Pb = lead; SDQ = Strengths and Difficulties Questionnaire; SES = socioeconomic status; SGA = small for
gestational age; TBD = to be determined; TMT A = Trail Making Test: attention; XRF = x-ray fluorescence; yr = year(s).

aEffect estimates are standardized to a 1 [jg/dL increase in BLL or a 10 |jg/g increase in bone Pb level, unless otherwise noted. For studies that report results
corresponding to a change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the
biomarker and standardized accordingly.

bEffect estimate is unstandardized due to insufficient blood lead distribution information or insufficient information regarding log transformation.

cResult did not report confidence interval nor p-value

fStudies published since the 2013 Integrated Science Assessment for Lead.

External Review Draft

3-376

DRAFT: Do not cite or quote

-------
Table 3-9E Epidemiologic studies of Pb exposure and externalizing behaviors including conduct disorders,
aggression, and criminal behavior in children and adolescents.

Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tTatsuta et al.
(2012)

Tohoku district
Japan

Study years NR
Followed through
30 mo

TSCD birth cohort
n: 306

Mother/child pairs in
Japan

Blood

Cord blood, HR-ICP-MS

Age at measurement:
delivery

Median = 1.0 [jg/dL
95th: 1.7 [jg/dL

Externalizing
behavior composite
(oppositional,
aggressive)

CBCL

Age at Outcome:
2.5 yr

Child age, birth weight, sex,
maternal age at pregnancy, delivery
type, birth order, drinking/smoking
habits in pregnancy, duration of
breastfeeding, maternal IQ,
Evaluation of Environmental
Stimulation score

Externalizing behavior
beta = -0.032b (not
significant)

Cohort

tSioen et al. (2013) Flemish Health and Blood

Flanders
Belgium

Oct. 2002 - Dec.
2003 (enrollment)
Followed through
June 2011

Cohort

Environment Study
(FLEHS 1)
n: 270

Cord blood, HR-ICP-MS

Age at measurement:
delivery

Birth cohort of
Flemish children living
in either rural or urban Median = 14 3 Mg/L

areas

75th: 25.3 pg/L

Conduct problems

SDQ with 5
domains:

emotional, conduct,
hyperactivity, peer
and social
problems

Age at outcome:
7 - 8 yr

Maternal and paternal BMI, maternal
age, weight increase of mother
during pregnancy, smoking during
pregnancy, smoking behavior of
maternal grandmother before birth of
mother, parental education, current
parental smoking, child sex, serious
infections of child since birth (also
tested interaction by sex)

OR per doubling of log-
transformed Pb:

Conduct problems:
1.182 (0.319, 4.385)°

External Review Draft

3-377

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tLiuetal. (2014b)

Jintan, Jiangsu

province

China

Sep. 1, 2004 - Apr.
30, 2005 (age 3-5
yrs)

Followed through
age 6 yr

Cohort

China Jintan Child
Cohort Study
n: 1025 children

Chinese preschool
children

Blood

Child venous blood; GFAAS
Age at measurement:

3-5 yr

Mean (SD): 6.4 (2.6) pg/dL
median = 6.0 (ig/dL
75th: 7.5 (ig/dL
90th: 9.4 (ig/dL
Max: 32 pg/dL

Aggressive
behavior and
oppositional defiant
problems

CBCL (Chinese
version); Caregiver-
Teacher Report
Form; normalized T
scores

Age at outcome:
6 yr

Age at BLL test, sex, preschool
residence, father's educational level,
mother's educational level, father's
occupation, parents' marital status,
single child status, and child IQ

Parent:

Aggressive (3 (95% CI):
-0.018 (-0.264, 0.229)

Oppositional (3 (95%
CI): -0.03 (-0.28,

0.22)

Teacher:

Aggressive (3 (95% CI):
0.001 (-0.001, 0.003)
Oppositional (3 (95%
CI): 0.223 (-0.038,
0.484)

Aggressive OR (95%
CI): overall 1.07 (0.98,
1.17); boys 1.03 (0.93,

1.14);	girls 1.21 (0.99,
1.47)

Oppositional OR (95%
CI): overall 1.06 (0.98,

1.15);	boys 1.02 (0.92,
1.13); girls 1.12 (0.97,
1.29)

tNkomo et al.
(2017)

Soweto/Johannesb
urg

South Africa

Apr. 23 - Jun. 8,
1990 (enrollment)
Followed 15-16 yr

Cohort

Birth to Twenty Plus
(BT20+)
n: 1322

684 females, 87.2%
Black African; 10.4%
mixed ancestry urban
residents; White and
Indian participants
excluded due to low
numbers

Blood

Child venous blood; GFAAS
with Zeeman background
correction

Age at measurement:

13 yr

Mean (SD) = 5.76 (2.42) pg/dL
median = 5.62 pg/dL
75th: 7.08 pg/dL
Max: 28 pg/dL

Violent behavior

YSR - violent
behavior

Age at outcome:
15-16 yr

Child sex, ethnicity, maternal
education, public/private hospital,
SES (unclear covariate adjustment
reporting)

physical violence (3
(95% CI): 0.05 (0.04,
0.05)

External Review Draft

3-378

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tNkomo et al.
(2018)

Soweto/Johannesb
urg

South Africa

Apr. 23 - Jun. 8,
1990 (enrollment)
Followed 14-15 yr

Cohort

Birth to Twenty Plus
(BT20+)
n: 1086

Black African and
mixed ancestry urban
residents; White and
Indian participants
excluded due to low
numbers

Blood

Child venous blood; GFAAS
with Zeeman background
correction

Age at measurement:

13 yr

mean (SD) = 5.6 (2.3) pg/dL
GM = 5.1 pg/dL
median = 5.4 pg/dL

Aggressive
behavior

YSR

Age at outcome:
14-15 yr

Child sex, maternal age, maternal
education at birth, marital status,
public/private hospital, SES

Direct aggression
(BLLs >10 pg/dL vs.
<5 pg/dL): 0.43 (0.08,
0.78)d

tBoucher et al.
(2012b)

Nunavik, Arctic

Quebec

Canada

1992-2000
(enrollment)

2005-2010 (follow-
up)

Cohort

Cord Blood
Monitoring Program
and Environmental
Contaminants and
Child Development
Study
n: 279

Inuit Children

Blood

Cord blood; AAS

Child venous blood; ICP-MS

Age at measurement:
Avg: delivery (cord) 11.3 yr
(child)

Mean: 4.7 (cord);
Max: 20.9 (cord);

2.7 (child)
12.8 (child)

Externalizing
behavior and
OD/CD problems

CBCLand DBD
rating scale

Age at outcome:
Avg: 11.3 yr old

Child age and sex, SES, age of the
biological mother at birth, maternal
tobacco use during pregnancy, and
birth weight, Hg

Externalizing behavior

Cord: 0.09 (-0.05,
0.23 )c

Child: 0.14 (0.01,
0.26)c

OR

OD/CD

2nd vs. 1st fertile: 1.90
(0.88, 4.11 )c
3rd vs. 1st fertile: 1.53
(0.67, 3.49)c

External Review Draft

3-379

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tBeckwith et al.
(2018)

Cincinnati, OH
U.S.

1979-1984
(enrollment)

Followed through
19-24 yr

Cohort

CLS
n: 250

Young adults from
birth cohort

Recruited pregnant
women in 1st or 2nd
trimester from inner
city neighborhoods
with historically
elevated incidence of
childhood lead
poisoning

Blood

Child venous blood; ASV (see
(Roda et al.. 1988))

Age at measurement:

78 mo

Mean: 7.99 [jg/dL
Max: 24.75 pg/dL

PPI score and gray
and white matter
volume in cingulate
and ventromedial
prefrontal cortex

PPI; high resolution
anatomical MRI
with Voxel Based
Morphometry to
calculate brain
volume changes

Age at outcome:
19-24 yr old

Sex, race, age at time of imaging,
gestational age at birth, weight at
birth, maternal IQ, participant IQ,
HOME score, adult marijuana
usage, maternal prenatal alcohol
use, maternal prenatal cigarette use,
maternal narcotic use, and maternal
prenatal marijuana use

Beta
PPI

Overall: 0.22 (0.06,
0.38)e

Female: 0.16 (-0.05,
0.37)e

Male: 0.22 (-0.02,
0.47)e

tDesrochers-
Couture et al.
(2019)

Nunavik, Northern

Quebec

Canada

Nov. 1993-Mar.
2002 (enrollment)

Sep. 2005-Feb.
2010 (1st follow-up)

Jan. 2013-Feb.
2016 (2nd follow-up)

Cohort

NCDS-childhood
n: 212

Inuit children from 14
coastal villages in
Nunavik, Quebec,
subsample from the
Cord Blood
Monitoring Program
and NIH-infancy study

Blood

Cord and venous child blood;
GFAAS (cord), ICP-MS (child)

Age at measurement:

Cord: delivery; Avg child: 11.4
and 18.5 yrs

GM (GSD): 3.80 (1.84) pg/dL
(cord); 2.34 (1.86) pg/dL
(child); 1.63 (2.00) pg/dL
(adolescent)

Median:3.73 pg/dL (cord); 2.07
pg/dL (child); 1.52 pg/dL
(adolescent)

Max: 17.80 pg/dL (cord);
12.83 pg/dL (child); 18.13
pg/dL (adolescent)

Externalizing
behavior; behavior
problems;
substance use

Behaviors -

externalizing

(CBCL),

hyperactivity-

impulsivity (DBD,

BAARS),

oppositional

defiant/conduct

disorder (DBD,

DISC); substance

use

Age at outcome:
childhood (11 yr);
adolescence (18 yr)

Child age, sex, SES, age of
biological mother at delivery,
maternal tobacco smoking during
pregnancy, birth weight, blood Hg,
house crowding, education of
primary caregiver

Beta (95% CI):

Child Blood:

Child externalizing

behavior:

0.23 (0.08, 0.38)

Direct effect on
adolescent
externalizing: 0.34 (-
0.38, 1.06)

Indirect effect on
adolescent
externalizing: 0.18 (0,
0.36)

Child OD/CD: 0.37
(0.06, 0.69)

Direct effect on
adolescent CD: 0.01 (-
0.10,0.13)

Indirect effect o
adolescent CD (0.01, -
0.01,0.03)

External Review Draft

3-380

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tTlotlenq et al.
(2022)

Johannesburg
South Africa

April-June 1990
(birth), sub-cohort
established at age
9 yr, followed
through 23-24 yr

Cohort

Young adults, n = 100 Bone

Sub-cohort (Bone
Health Cohort) of
singleton children
(born April-June
1990) from BT20
Cohort enrolling
women in 2nd and
3rd trimester residing
in Soweto-
Johannesburg

Child tibia; KXRF

Age at measurement: NR

Mean (SD), min, med (IQR),
max: 8.7 (5.3), 0, 9 (5-12.5),

21 pg/g

Males (n = 53): 8.1 (4.4), 0, 8
(5-11), 18 pg/g
Females (n = 47): 9.4 (6.1), 0,
10 (4-14), 21 pg/g

Aggression scores
(anger, physical,
verbal, hostility)

BPAQ

Age at outcome:
23-24 yr

Age, sex, exposure to family
violence, attitude toward
neighborhood, exposure to crime
and violence in the neighborhood

Level of schooling, alcohol and drug
abuse, presence of both parents at
home, home environment, and SES
(maternal education, housing type,
participant's education/occupation)
also considered.

Beta per 1 pg/g
increase in bone Pb

Anger aggression: 0.25
(0.04, 0.37)

Physical aggression:
0.093 (-0.01, 0.27)

Verbal aggression:
0.093 (-0.05, 0.23)
Hostility: 0.03 (-0.19,
0.26)

tReuben et al.
(2019)

Dunedin
New Zealand

Apr. 1, 1972-Mar.
31, 1973
(enrollment)
Followed through
Dec. 2012

Cohort

Dunedin
Multidisciplinary
Health and
Development Study

N: 579

Birth cohort of
nationally
representative
(majority white)
children with high
rates of participation
and follow-up

Blood

Child venous blood; GFAAS

Age at measurement: 11 yr

Mean: 11.08 pg/dL
(94% above 5 pg/dL)

Antisocial behavior
in children

Rutter Child Scale
(averaged
parent/teacher
ratings)

Age at outcome:
11 yr

Sex, childhood SES, maternal IQ,
and family history of mental illness.

Beta

Antisocial behavior:
0.02 (0.00, 0.04)

External Review Draft

3-381

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

Chandramouli et al.
(2009)

Avon
U.K

Jul. - Dec. 1992
(enrollment)

Followed through
8 yr

Cohort

10% random
subsample of Avon
Longitudinal Study of
Parents and Children
(ALSPAC)

n = 488
Birth cohort

Blood

Child venous blood; AAS with
micro sampling flame
atomisation

Age at measurement: 30 mo

Mean (SD): NR
Group 1: 0-<2 |jg/dL
Group 2: 2-<5 [jg/dL
Group 3: 5-<10 [jg/dL
Group 4: >10 pg/dL

Antisocial activities

Parent/teacher
ratings on

Antisocial Behavior
Interview

Age at outcome:
8 yr

Maternal education and smoking,
home ownership, home facilities
score, family adversity index,
paternal SES, parenting attitudes at
6 mo, child sex. Also considered
child IQ.

ORs for increased
score

Group 1 (0-<2 pg/dL):
ref

Group 2 (2-<5 pg/dL):
0.93 (0.47, 1.83)

Group 1 (5-<10 pg/dL):
1.44 (0.73, 2.84)

Group 1 (>10 pg/dL):
2.90 (1.05, 8.03)

Wright et al. (2008)

Cincinnati, OH
United States

1979-1984
(enrollment)
Followed through
19-24 yr

Cohort

CLS
n: 250

Young adults from
birth cohort

Recruited pregnant
women in 1st or 2nd
trimester from inner
city neighborhoods
with historically
elevated incidence of
childhood lead
poisoning

Blood

Child blood; ASV

Age at measurement: 6 yr

Median (5th—95th):
6 yr: 6.8(3.4-18) pg/dL
0-6 yr avg: 12 (6.0-26) pg/dL

Criminal arrests
County records

Age at outcome:
19-24 yr

Maternal IQ and education, sex,

Also considered potential
confounding by maternal prenatal
smoking, marijuana use, narcotic
use, and prior arrests, HOME score,
birth weight, # children in the home,
public assistance in childhood.

RRs (yes/no)

Age 6 blood Pb: 1.05
(1.01, 1.09)

Age 0-6 yr avg blood
Pb:

1.01 (0.98, 1.05)

External Review Draft

3-382

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tFruh et al. (2019) Project Viva
n: 1006

Eastern

Massachusetts Birth cohort of
U s	mother-child pairs

1999-2002
(enrollment)

Followed through
age 7 yr

Cohort

Blood

Maternal venous blood; ICP-
MS

Age at measurement:

T2

Median: 1.1 pg/dL

Parent teacher
ratings of conduct
problems using
SDQ

Standardized for
child age and sex

Age at outcome:
7 yr old

Maternal 2nd trimester Hg and Mn
levels, nulliparity, smoking during
pregnancy, IQ, and education;
paternal education; HOME
composite score and household
income; and child race/ethnicity

Parent ratings:

Overall (3 (95% CI):
0.10 (-0.10, 0.30)

Boys: 0.07 (-0.18,
0.32)

Girls: 0.13 (-0.13,
0.40)

Teacher ratings:

Overall (3 (95% CI):
0.18 (-0.08, 0.44)

Boys: 0.18 (-0.17,
0.53)

Girls: 0.17 (-0.13,
0.46)

tRuebner et al.
(2019)

46 centers
U.S.

Study Years: NR
Follow-up: NR

Cohort

CKiD Cohort study
n: 412

Children ages 1-16 yr
at recruitment with
mild to moderate CKD

Blood

Child venous blood; ICP-MS.
The BLL measurement closest
to the time of neurocognitive
testing was used for analysis
(concurrent).

Age at measurement:

NR; 2, 4, or 6 years after study

entry

Median: 1.2 [jg/dL
75th: 1.8 [jg/dL
Max: 5.1 [jg/dL

Externalizing
behaviors,
composite index on
the BASC-2 (see
also 3.5.1 and
3.5.2)

The last available
test results were
used to evaluate
long-term effects.
Mean time between
BLL and
neurocognitive
testing was 2.3 yr.
Age at outcome:
1-16 yr

Age, sex, race, poverty, and
maternal education

Adjusted BASC-2
results were not
reported because they
were not statistically
significant.

External Review Draft

3-383

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tNaicker et al.
(2012)

Johannesburg
South Africa

Apr. - Jun. 1990
(enrollment)
Followed through
20 yr

Birth to Twenty cohort Blood
(Bt20)

n: 1041 (487 boys,

554 girls)

Singleton children
representative of
South Africa
population

Child venous blood; GFAAS
Age at measurement:

13 yr

median = 5.4 [jg/dL;
GM = 5.2 [jg/dL
Max: 28.1 pg/dL

Rule-breaking
behavior,
aggressive
behavior

YSR (adapted from
CBCL for use in
adolescents)

Age at outcome:
13 yr

SES, maternal education,
demographic factors

Attacking people -
boys (3 (95% CI): 0.54
(0.09, 0.98)d

Cohort

tRodriques et al.
(2018)

Simoes Filho,
Salvador, Bahia
Brazil

Study years NR
Cross-sectional

Simoes Filho,
n: 225

Brazil Blood

Children aged 7-
12 yr, attending public
in town near ferro-Mn
alloy plant

Child venous blood; GFAAS
Age at measurement:
7-12 yr

median = 1.2 [jg/dL
Max: 15.6 [jg/dL

Behavioral
problems/disruptive
behavior
(externalizing
behavior,
aggressive
behavior, rule-
breaking behavior)

CBCL

Age at outcome:
7-12 yr

Gender, age
violence score

community
maternal IQ

Attacking people -
Adjusted total T-score
(3 (95% CI): -0.74
(-5.3, 3.8)d

External Review Draft

3-384

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tBarq et al. (2018) Montevideo sample
n: 206

Montevideo

Uruguay	Children in urban

area

Study years NR
Cross-sectional

Blood

Child venous blood (fasting);
AAS with flame or graphite
furnace ionization
Age at measurement:
6-8 yr (mean = 6.75 yr)

mean = 4.2 [jg/dL

Behavior problems Child IQ, iron status, BMI, household BRIEF:

(e.g., oppositional)

CRS-R; BRIEF

Age at outcome:
6-8 yr

(mean = 6.75 yr)

possessions, maternal education,
current parent smoking (also looked
at sex and Pb evaluation method in
sensitivity analyses)

Behavioral Regulation
Index (PR [95% CI]):
overall = 1.01 (1.00,

1.03);	girls = 1.03
(1.00, 1.05);
boys = 0.99 (0.97,
1.01)

CTRS-R:

Oppositional (PR [95%
CI]): overall = 1.00
(0.98, 1.02);
girls = 1.01 (0.99,

1.04);	boys = 0.99
(0.96, 1.02)

tLiu et al. (2022b) Healthy Brains and
Behavior

Philadelphia
County, PA;
Suburbs of
Philadelphia, PA
United States

Study years NR
Cross-sectional

n: 131

Blood

Child blood; HR-ICP-MS.
Age at Measurement:
11-12 yr

Mean = 2.2 Apg/dL; Median
1.10 Apg/dL
75th: 1.8 Apg/dL
Max: 35.4 Apg/dL

Parent-report and
child-report of
externalizing
behavior
(composite)

Scores derived
from factor
analyses of 14
validated measures
of antisocial/
aggressive
behavior (RPQ,
CBCL, YSR, APSD,
CODDS, AQ from
BPAQ)

OLS regression adjusted for sex and
race.

Beta for externalizing
behavior:

Parent-reported: 0.20
(0.05, 0.34)

Child-reported: 0.20
(0.04, 0.35)

Age at outcome:
11-12 yr

External Review Draft

3-385

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tNiqq etal. (2010) n = 326

Study location and Recruitment by

year NR

Case-control

community
advertisements,
mailings, outreach to
clinics

Blood

Externalizing

composite score

(oppositional and

x „	conduct symptoms)

Age at Measurement: 6-17 yr Qn pgrent K_SADS

Child venous blood; ICP-MS

Mean (SE) = 0.73 (0.04) [jg/dL Oppositional

behavior on parent
and teacher CRS

Age at Outcome:
6-17 yr

Household income, maternal
smoking, child age, sex, blood
hemoglobin, child FSIQ (WISC-IV)

Beta for SD increase in
scores per SD
increase in log-10 Pb
Parent ratings:

K-SADS externalizing
composite: 0.21 (0.05,
0.37)

CRS oppositional
behavior: 0.09 (-0.09,
0.27)

Teacher ratings:
CRS oppositional
behavior: 0.11 (-0.01,
0.23)

tAmato et al.
(2013)

Milwaukee, Wl
United States

Study years NR

Followed 7-10 yr
(blood Pb before
age 3, outcome
assessment at 4th
grade)

Wisconsin Childhood
Pb Poisoning
Prevention Project /
Milwaukee Public
School

n: 1076 unexposed;
2687 exposed; 3763
total

Exposed individuals
were more likely to be
black or Hispanic, and
be on assisted lunch
programs

Blood

Maximum child blood Pb,
methods varied by providers
Age at measurement:

<3 yr

Mean NR; reported exposed
(BLL 10-20 |jg/dL) vs.
unexposed (<5 |jg/dL)
Max: 20 [jg/dL

School
suspensions

Unduplicated
suspension count

Age at outcome:
10 yr (4th grade)

Gender, race/ethnicity, income
(free/reduced lunch)

Suspensions

OR for exposed
(10-20 |jg/dL) vs.
unexposed (<5 |jg/dL):
2.66 (2.12, 3.32)

Cohort

External Review Draft

3-386

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tBoutwell et al.
(2017)

106 census tracts in Blood
St. Louis City, MO
n: 59,645 children; NR

St. Louis City, MO 15,734 violent crimes Age at measurement:

United States

Study years: NR
16-year period

Other - ecological
study

St Louis residents

<72 mo age

NR

Violent crime
(crimes with
firearm, assault
crimes, robbery
crimes, homicides,
rape)

Police

department/uniform
crime report -
violent crime
(crimes with
firearm, assault
crimes, robbery
crimes, homicides,
rape)

Concentrated disadvantage; mean
age of housing; proportion occupied
by renters; domestic assaults

RR for 1% increase in
proportion of elevated
blood tests in the
census tract.

Firearm crimes: 1.03
(1.025, 1.035)

Assault: 1.03 (1.025,
1.035)

Robbery: 1.03 (1.02,
1.04)

Homicides: 1.03
(1.015, 1.045)

Rape: 1.01 (0.99, 1.03)

External Review Draft

3-387

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tBecklev et al.
(2018)

Apr. 1, 1972-Mar.
31, 1973
(enrollment)
Followed through
38 yr

Cohort

Dunedin
Multidisciplinary
Health and
Development Study

N: 553

Birth cohort of
nationally
representative
(majority white)
children with high
rates of participation
and follow-up

Blood

Child venous blood; GFAAS
Age at measurement:

11 yr

mean = 11.01 pg/dL
Max: 31 pg/dl_

Criminal offending Sex

(criminal conviction,

recidivism,

conviction for

violent offense,

self-reported

criminal offending)

Official conviction
records from
central police
computer; self-
reported offending
interview

Age at outcome:

38 yr

OR (ref: no conviction)
Any criminal
conviction: 1.042 (1,
1.086)

One-time: 1.046 (0.99,
1.104)

Recidivistic: 1.039
(0.986, 1.095)
Nonviolent: 1.051
(1.003, 1.101)

Violent offense: 1.025
(0.962, 1.092)

Beta (self-report
offending)

15 yr: 0.1 (0.015,
0.185)

18 yr: 0.06
0.14)

21 yr: 0.01
0.085)

26 yr: 0.06
0.135)

32 yr: 0.04
0.12)

38 yr: 0.02

-0.02,

-0.065,

-0.015,

-0.04,

-0.06, 0.1)

External Review Draft

3-388

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

tWriqht et al.
(2021)

Cincinnati, OH
United States

1979-1984
(enrollment)
Followed through
2013

Cohort

CLS
n: 254

Young adults from
birth cohort

Recruited pregnant
women in 1st or 2nd
trimester from inner
city neighborhoods
with historically
elevated incidence of
childhood lead
poisoning

Blood

Maternal blood (n= 219) and
child blood; ASV

Age at measurement: prenatal,
<60 mo (avg child), 60-78 mo
(avg late child), 78 mo (late
child)

Mean (SD):

Average child (0-60 mo): 14.4
(6.6) pg/dL

Total number of
arrests for each
subject (2003-2013
and lifetime),
violent crimes, drug
crimes, and
property crimes

Hamilton County
public records

Age at outcome:
18-24 yr, 27-33 yr

Birth weight (grams), maternal age
at delivery, Appearance, Pulse,
Grimace, Activity, and Respiration
scores taken at 1 min, self-reported
maternal drug use during pregnancy
that includes reports of alcohol,
marijuana and tobacco use,
maternal IQ measured by the WAIS-
R, and HOME Inventory scores
across the first 3 yr

IRR

Arrests 2003-2013 for
6-year blood Pb,
controlling for prior
arrests 1998-2003:
1.008 (0.995, 1.021)

6 year blood Pb
Lifetime Arrests:
1.016(1.002, 1.03)
Property Arrests: 1
(0.977, 1.023)

Drug Arrests: 1.032
(1.005, 1.06)

Violent Arrests: 1.016
(0.992, 1.039)

Adult Arrests: 1.014
(1, 1.027)

EEs for other blood Pb
sources are available
but not listed for the
sake of space.

External Review Draft

3-389

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% Clsa

BASC-2 = Behavior Assessment System for Children; BLL = blood lead level; BMI = body mass index; BRIEF = Behavior Rating Inventory of Executive Functions; CBCL = Child
Behavior Check List; CI = confidence interval; CKD = chronic kidney disease; CKiD = Chronic Kidney Disease in Children Study; CLS = Cincinnati Lead Study; CTRS-R = Conners'
Teacher Rating Scale-Revised; DBD = Disruptive Behavior Disorder; DISC = Disrupted-in-Schizophrenia; EES = Evaluation of Environmental Stimulation; FLEHS = Flemish
Environment and Health Study; GFAAS = graphite furnace atomic absorption spectrometry; Hg = mercury; HOME = Health Outcomes and Measures of The Environment Study; ICP-
MS = Inductively Coupled Plasma Mass Spectrometry; IQ = intelligence quotient; IQR = interquartile range; K-SADS = Kiddie Schedule for Affective Disorders and Schizophrenia;
Mn = manganese; mo = month(s); NCDS = Nunavik Child Development Study; NR = not reported; Pb = lead; PPI = Psychopathic Personality Inventory; RR = relative risk;
SDQ = Strengths and Difficulties Questionnaire; SE = standard error; SES = socioeconomic status; T2 = second trimester of pregnancy; TSCD = Tohoku Study of Child
Development; WAIS-R = Weschler Adult Intelligence Scale-Revised; yr = year(s); YSR = Youth Self-Report.

aEffect estimates are standardized to a 1 |jg/dL increase in BLL or a 10 |jg/g increase in bone Pb level, unless otherwise noted. For studies that report results corresponding to a

change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the biomarker and standardized accordingly.

bResults are unstandardized because they did not have an associated SE, CI, or p-value reported in the study.

°Results are unstandardized because the log base used for exposure transformation was unspecified in the study.

dResults are unstandardized because the Pb level distribution data was not available.

eThe CI was calculated from a p-value and the true CI may be wider or narrower than calculated.

fStudies published since the 2013 Integrated Science Assessment for Lead.

External Review Draft

3-390

DRAFT: Do not cite or quote

-------
Table 3-10E Epidemiologic studies of Pb exposure and internalizing behaviors in children.

Reference and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates
and 95% Clsa

Wasserman et al. (2001)

Pristina
Yugoslavia

1984-1985 (enrollment)
Followed through 1999

Cohort

N: 191

Blood

Recruitment Child blood; method NR
from prenatal

clinics	Age at outcome:

Delivery to 4-5 yr

Lifetime (to age 4-5 yr) avg
blood

Mean (SD) of log—10 Pb:
0.86 (0.12) [jg/dL, Mean:
-7.2 [jg/dL

Internalizing behavior scores
and subscores (i.e.,
anxious/depressed, somatic
complaints, and withdrawn)
assessed using maternal
ratings of CBCL

Age at Outcome: 4-5 yr

Sex, ethnicity, age, maternal
education and smoking
history, HOME score, birth
weight

Betas for log—10
change in outcome:

Internalizing
composite: 0.152
(0.023, 0.281)
Anxious/depressed:
0.041 (-0.089, 0.17)
Somatic complaints:
0.107 (-0.062,
0.276)

Withdrawn: 0.066
(-0.073, 0.205)

External Review Draft

3-391

DRAFT: Do not cite or quote

-------
Refe,enDeeSfgn„dS,Udy Population Exposure Assessment

Outcome

Confounders

Effect Estimates
and 95% Clsa

Burns et al. (1999)

Port Pirie
Australia

May 1979-May 1982
(enrollment)

Followed to age 11-13
years

Cohort

Female:

Internalizing
composite: 2.1 (0.0,
4.2)b

Anxious/depressed:
1.3 (0.1, 2.5)b
Somatic complaints:
0.3 (-0.4, 0.9)b

Withdrawn: 0.6 (0.0,
1.1)b

Port Pirie
Cohort Study
(PPCS)

N: 322

Recruited
90% of live
births in a
lead smelting
community

Blood

Lifetime avg (to age 11-13
yr) blood

GM (95% CI) [jg/dL

Boys: 14.3 (13.5,
Girls: 13.9 (13.2,

15.1)
14.6)

Internalizing behavior scores
and subscores (i.e.
anxious/depressed, somatic
complaints, and withdrawn)
assessed using maternal
ratings of CBCL

Age at Outcome: 11 -13 yr

Maternal age, prenatal
smoking status, IQ,
concurrent psychopathology,
and education, birth weight,
type of feeding, length of
breastfeeding, paternal
education and occupation,
birth order, family functioning,
parental smoking, marital
status, HOME score, child IQ

Beta
Male:

Internalizing
composite: 0.8 (-0.9,
2.4)b

Anxious/depressed:
0.8 (-0.2, 1.8)b
Somatic complaints:
-0.1 (-0.7, 0.4)b
Withdrawn: 0.1
(-0.4, 0.7)b

External Review Draft

3-392

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates
and 95% Clsa

Bellinger et al. (1994b)

Boston, MA
US

1979-1980 (birth) followed
to age 8 yrs

Cohort

N:1782

Recruitment
at birth
hospital

Blood, Tooth

Cord blood and shed
deciduous teeth. Dentin
taken from zone representing
cumulative postnatal
deposition; ASV
Age at Measurement: 6 yr

Tooth mean (SD): 3.4 (2.4)
pg/dL

Range: 0.1-28.9
10th—90th percentiles:
1.2-6.3

95th percentile: 7.4

Cord blood mean (SD): 6.8
(3.1) |jg/dL

Interval analyzed: 0.1-35.1
95th percentile: 12.2

Internalizing behavior T
scores and subscores (i.e.
anxious/depressed, somatic
complaints, and withdrawn)
assessed using teacher
ratings of CBCL

Age at Outcome: 8 years old

Prepregnant weight, race,
Cesarean section, maternal
marital status, prenatal care,
paternal education, colic, child
current medication use,
sibship size, sex, birth weight.
Also considered potential
confounding by public
assistance, prenatal smoking,
maternal education but not
parental caregiving quality.

Betas for In-
transformed change
in internalizing T
score:

Cord blood: -0.07
(-0.23, 0.10)

Tooth: 0.43 (0.09,
0.77)

tWinter and Sampson
(2017)

Chicago, Illinois
U.S.

born 1995-1997 to 2013,
followed through 17 yr old
Cohort

PHDCN
n: 254

Children and
caregivers
living in
Chicago

Blood

Child venous and capillary
blood; methods NR
Age at measurement:
before 6 yr

Avg BLL before 6 yr
Mean: 6.4 [jg/dL

Internalizing on the CBCL
(i.e., anxiety and
depression); see also
Section 3.5.2 (impulsivity)
PC questionnaire)

Age at outcome:

Mean: 17 yr old

Age, sex, race/ethnicity; PC's
immigrant generational status,
marital status, education,
Temporary Assistance for
Needy Families receipt;
proportion residential
neighborhood that is non-
Hispanic Black, Hispanic, and
below the poverty line;
proportion of the child's
residential neighborhood
tested for Pb exposure

Beta

Anxiety/depression:
0.09 (0.03, 0.16)

External Review Draft

3-393

DRAFT: Do not cite or quote

-------
tLiuetal. (2014b)

Jintan, Jiangsu province
China

Sep. 1, 2004-Apr. 30,
2005 (age 3-5 yrs)
Followed to age 6 yrs

Cohort

China Jintan
Child Cohort
Study
n: 1025
children

Chinese

preschool

children

Blood

Child venous blood; GFAAS

Age at measurement:
3-5 yr old

Mean (SD): 6.4 (2.6) pg/dL

median = 6.0 pg/dL
75th: 7.5 pg/dL
90th: 9.4 pg/dL
Max: 32 pg/dL

External Review Draft

Internalizing problems
composite and subscores
(emotionally reactive,
anxious/depressed, somatic
complaints, withdrawn, and
sleep)

CBCL (Chinese version);
Caregiver-Teacher Report
Form; normalized T scores

Age at outcome:

6 yr

Age at BLL test, sex,
preschool residence, father's
educational level, mother's
educational level, father's
occupation, parents' marital
status, single child status, and
child IQ

Internalizing
problems

Parent beta: -0.029
(-0.280, 0.222)
Teacher beta: 0.223
(-0.037, 0.484)
Teacher OR: 1.10
(1.03, 1.18)

Emotionally Reactive
Parent beta: -0.117
(-0.365, 0.131)
Teacher beta: 0.322
(0.058, 0.587)
Teacher OR: 1.10
(1.02, 1.19)

Anxiety/Depression:
Parent beta: 0.101
(-0.151, 0.354)
Teacher beta: 0.001
(-0.001, 0.003)
Teacher OR: 1.12
(1.03, 1.23)

Somatic Complaints
Parent beta: -0.171
(-0.436, 0.094)
Teacher beta: 0.001
(-0.003, 0.001)
Teacher OR: 1.01
(0.90, 1.13)

Withdrawn
Parent beta: 0.096
(-0.158, 0.349)
Teacher beta: 0.001
(-0.001, 0.003)
Teacher OR: 1.02
(0.93, 1.12)

3-394

DRAFT: Do not cite or quote

-------
Refe,enDeeSfgn„dS,Udy Population Exposure Assessment

Outcome

Confounders

Effect Estimates
and 95% Clsa

Clinically significant
anxiety

Parent beta: 0.044
(-0.212, 0.299)
Teacher beta: 0.253
(0.016, 0.500)
Teacher OR: 1.12
(1.03, 1.23)

External Review Draft

3-395

DRAFT: Do not cite or quote

-------
tJoo etal. (2018)

Seoul, Ulsan, Cheonan
South Korea

2006-2011 (enrollment)
Followed through 5 yr

Cohort

MOCEH
n: 575
mother-child
pairs

pregnant
women

Blood

Maternal venous blood, cord
blood, and child blood; AAS

Age at measurement: 20 wk
gestation (maternal); delivery
(cord); 2,3 and 5 yr (child)

GM:

Maternal 1.28 [jg/dL (early),
1.24 (late) 0.9 (cord)

Child 1.55 (age 2), 1.43 (age
3), 1.29 (age 5)

Internalizing behavior

K-CBCL (emotional
reactivity, anxious/
depressed, somatic
complaints, and
withdrawn/depressed states)
See also Section 3.5.2

Age at outcome:

5 yr old

Maternal age at childbirth,
parity, maternal educational
level, household income,
residential area, and
breastfeeding

Beta (95% CI):
Internalizing at 5 yr

Maternal-early
pregnancy

Male: -0.16 (-2.54,
2.23)

Female: -0.13
(-1.86, 1.60)

Maternal-late
pregnancy
Male: 2.55 (0.22,
4.88)

Female: -0.18
(-2.66, 2.31)

Cord blood

Male: 2.44 (-0.74,
5.63)

Female: -1.00
(-4.30, 2.29)

Child blood-2 yr

Male: -0.03 (-2.07,
2.00)

Female: 2.94 (0.36,
5.52)

Child blood-3 yr

Male: 0.25 (-2.33,
2.82)

Female: 2.76 (-0.73,
6.26)

Child blood-5 yr
(concurrent)

Male: 1.23 (-2.10,
4.56)

Female: 5.65 (0.50,
10.80)

External Review Draft

3-396

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates
and 95% Clsa

tFruhetal. (2019)

Eastern Massachusetts
U.S.

1999-2002 (enrollment)
Followed through age 7 yr

Cohort

Project Viva Blood
n: 1006

Maternal venous blood; ICP-
Birth cohort of MS
mother-child Age at measurement:
pairs	T2

Median: 1.1 pg/dL

Parent teacher ratings of
emotional problems

SDQ

Standardized for child age
and sex

Age at outcome:

7 yr old

Maternal 2nd trimester Hg
and Mn levels, nulliparity,
smoking during pregnancy,
IQ, and education; paternal
education; HOME composite
score and household income;
and child race/ethnicity

Parent ratings:

Overall (3 (95% CI):
0.30 (0.05, 0.55)
Boys: 0.17 (-0.17,
0.50)

Girls: 0.52 (0.18,
0.86)

Teacher ratings:

Overall (3 (95% CI):
0.07 (-0.22, 0.35)
Boys: 0.02 (-0.33,
0.37)

Girls: 0.12 (-0.31,
0.54)

tSioen etal. (2013)

Flanders
Belgium

Oct. 2002 - Dec. 2003
(enrollment)

Followed through June 2011
Cohort

Flemish
Health and
Environment
Study
(FLEHS 1)
n: 270

Birth cohort of
Flemish
children living
in either rural
or urban
areas

Blood

Cord blood, HR-ICP-MS

Age at measurement:
delivery

median = 14.3 |jg/L
75th: 25.3 pg/L

Emotional problems

SDQ with 5 domains:
emotional, conduct,
hyperactivity, peer and social
problems

Age at outcome:

7 - 8 yr

Maternal BMI, age at
pregnancy, weight increase
during pregnancy, smoking,
paternal BMI, if parents
smoke, smoking behavior
maternal grandmother before
the birth of the mother,
parental education, child sex,
serious child infections

OR per doubling of
log-transformed Pb:
Emotional problems:
0.900 (0.524, 1.547)b

External Review Draft

3-397

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates
and 95% Clsa

tRokoff et al. (2022)

New Bedford, MA
Born: 1993-1998

Cohort

Children
residing near
Superfund
site

n: 468 of 788
mother-infant
pairs.

Blood

Cord blood; ICP-MS

Child blood; medical records,
method NR

Age at measurement:
delivery

Mean (SD) cord BLL

CPRS: 1.37 [jg/dL (0.94)

BASC-2: 1.37 pg/dL (0.95)

Mean (SD) peak postnatal
BLL

CPRS: 6.68 pg/dL(3.95)
BASC-2: 6.58 pg/dL (3.87)

Internalizing Behavior

Anxiety, Depression,
Somatization, and
Internalizing Problems on
BASC-2 SRP

Anxious-Shy and
Psychosomatic on CPRS

Anxious-Shy on CTRS

Age(s) at outcome: 8 years
(CPRS and CTRS) and 15-
years (BASC-2)

Maternal age, marital status,
parity, parental education),
household income, maternal
smoking, alcohol consumption
during pregnancy, pre-
pregnancy weight, height, and
gestational weight gain, BMI,
prenatal social disadvantage
index, HOME score, maternal
IQ

No interactions between
chemicals; linear regression
models adjusted for Mn and
organochlorines,

BASC-2 SRP

Anxiety: 1.78 (0.58,
2.99)

Depression: 0.79 (-
0.39, 1.97)

CPRS

Psychosomatic

Boys: 2.08 (0.07,
4.10)

Girls: 0.48 (-1.00,
1.97)

C-R functions
presented

tRasnick et al. (2021)

Cincinnati, OH

Born: Oct2001-Jul 2003
Exposure: 2001-2005

Cohort

CCAAPS
n: 263

Air

LURF, air sampling at 24
sites (C-V R2 = 0.89),
predicted air concentration at
child's residence.

Children residing>1,500 m or
<400 m from major highway
eligible.

Median: 0.51 ng/m3 (range 0-
10.8 ng/m3)

Internalizing and
Externalizing Behavior

BASC-2; internalizing
behaviors (anxiety,
depression, somatization),
externalizing behaviors
(aggression, conduct
problems, and hyperactivity),
behavioral symptoms index
(attention problems,
atypicality, and withdrawal)

Age at outcome: 12 yr

Maternal education,
community-level deprivation,
blood Pb concentrations,
greenspace, and traffic
related air pollution.

B (anxiety score)=3.1
(95% CI: 0.4, 5.7)
per ng/m3

Note: no association
with depression,
somatization,
conduct problems,
hyperactivity,
withdrawal behaviors

External Review Draft

3-398

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates
and 95% Clsa

tRuebner et al. (2019)

46 centers
U.S.

Study Years: NR
Follow-up: 1-16 yr

Cohort

CKiD Cohort
study
n: 412

Children with
mild to
moderate
CKD

Blood

Child venous blood; ICP-MS.
The BLL measurement
closest to the time of
neurocognitive testing was
used for analysis
(concurrent).

Age at measurement:
NR; 2, 4, or 6 yr after study
entry

Median: 1.2 pg/dL
75th: 1.8 Mg/dL
Max: 5.1 [jg/dL

Internalizing behaviors,
composite index on the
BASC-2 (see also 3.5.1
3.5.2)

Age, sex, race, poverty, and
maternal education

and

The last available test results
were used to evaluate long-
term effects. Mean time
between BLL and
neurocognitive testing was
2.3 yr.

Age at outcome:

1-16 yr

Adjusted BASC-2
results were not
reported because
they were not
statistically
significant.

tHorton et al. (2018)

Mexico City
Mexico

born 1994-2006 and
followed through age 6-16
Cohort

ELEMENT
Project
n: 133

healthy, low
to moderate
income
mother (18-
39 yr old)-
child pairs

Tooth

Tooth Pb (prenatal, postnatal
metrics derived); laser
ablation ICP-MS
Age at measurement:
tooth Pb concentration
corresponded to prenatal and
300 days after birth

Figure 1c

Internalizing behavior on the
BASC-2. See Section 3.5.2
(attention and hyperactivity)
and BSI

Age at outcome:

8-11 yrold

Maternal age at delivery,
maternal education, smoking,
maternal IQ

Beta: BASC-2

Internalizing (10
months): NR
Anxiety (12 months):
0.4 (95% CI NR)C

Internalizing
composite result was
not reported because
it was not statistically
significant.

External Review Draft

3-399

DRAFT: Do not cite or quote

-------
Refe,enDeeSfgn„dS,Udy Population Exposure Assessment

Outcome

Confounders

Effect Estimates
and 95% Clsa

tDohertv et al. (2020)

New Hampshire
U.S.

2009 to 2014-2019
Cohort

NHBCS
n: 371

(SRS-2); 318
(BASC-2)

Mother-child
pairs

Toenails

Maternal and infant toenails;
Median (maternal prenatal):
0.14 |jg/g (SRS-2), 0.13 pg/g
(BASC-2); Median (maternal
postnatal): 0.10 pg/g (SRS),
0.11 pg/g (BASC-2); Median
(infant): 0.35 pg/g (SRS-2),
0.37 pg/g

Internalizing Behaviors on
the BASC-2; see also
Section 3.5.2.2

Age at outcome:

3 yr

Maternal age, maternal BMI,
parental education, maternal
smoking, marital status,
parity, child age at last
breastfeeding, Healthy Eating
Index score, year of birth, sex,
and age of the child at testing

Exposure was log2
transformed.

Betas per 1 pg/g
increase in toenail
Pb concentration.

Total

Maternal prenatal:
-0.14 (-0.28, 0.00)d
Maternal postnatal:
0.06 (-0.05, 0.18)d
Child: 0.01 (-0.14,
0.16)d

Males

Maternal prenatal:
-0.17 (-0.36, 0.01 )d
Maternal postnatal:
0.31 (0.15, 0.47)d
Child: 0.09 (-0.13,
0.31 )d

Females

Maternal prenatal:
-0.16 (-0.33, 0.01 )d
Maternal postnatal:
-0.04 (-0.20, 0.13)d

Child: -0.15 (-0.36,
0.06)d

External Review Draft

3-400

DRAFT: Do not cite or quote

-------
Refe,enDeeSfgn„dS,Udy Population Exposure Assessment

Outcome

Confounders

Effect Estimates
and 95% Clsa

BASC = Behavioral Assessment System for Children; BLL = blood lead level; BMI = Body Mass Index; BRIEF = Behavior Rating Inventory of Executive Functions; CBCL = Child
Behavior Check List; CCAAPS = Cincinnati Childhood Allergy and Air Pollution Study; CI = confidence interval; CKD = chronic kidney disease; CKiD = Chronic Kidney Disease in
Children Study; CPRS = Conners' Parent Rating Scale; C-TRF = Caregiver-Teacher Report Form; CTRS = Conners' Teacher Rating Scale; C-V R2 = cross validated R-square;
DSM = Diagnostic and Statistical Manual of Mental Disorders; ELEMENT = Early Life Exposures in Mexico to Environmental Toxicants; FSIQ = full-scale intelligence quotient;
GFAAS = graphite furnace atomic absorption spectrometry; ICP-MS = inductively coupled plasma mass spectrometry; HOME = Home Observation for Measurement of the
Environment; IQ = intelligence quotient; K-CBCL = Korean Child Behavior Check List; LURF = Land Use Random Forest; MOCEH = Mothers' and Children's Environmental Health;
NHBCS = New Hampshire Birth Cohort Study; NR = not reported; OR = odds ratio; Pb =lead; PC = primary caregiver; PHDCN = Project on Human Development in Chicago
Neighborhoods; SDQ = Strengths and Difficulties Questionnaire; SRP = Self-Report of Personality; SRS = Social Responsiveness Scale; yr = year(s); T2 = second trimester of
pregnancy.

aEffect estimates are standardized to a 1 |jg/dL increase in BLL or a 10 |jg/g increase in bone Pb level, unless otherwise noted. For studies that report results corresponding to a
change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the biomarker and standardized accordingly.

bResults are unstandardized because the log base used for exposure transformation was unspecified in the study.

°Results are unstandardized because they did not have an associated SE, CI, or p-value reported in the study.

dResults are unstandardized because the biomarker used for Pb exposure measurement is toenails.

fStudies published since the 2013 Integrated Science Assessment for Lead.

External Review Draft

3-401

DRAFT: Do not cite or quote

-------
Table 3-11E Epidemiologic studies of Pb exposure and motor function in children.

Reference
and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

Ris et al.
(20041

Cincinnati, OH
US

1979-1985
(enrollment)

Cohort

Cincinnati Lead Blood
Study (CLS)

N: 195

Birth cohort
recruited
prenatally from
obstetrical
clinics

Prenatal maternal blood: NR

Average Childhood blood (mean of 20 quarterly
concentrations obtained from 3-60 months): NR

78 month blood lead: NR

Visuoconstructio Maternal IQ, SES,
n (Block Design total average HOME
Subtest, ROCF-
Accuracy) and
Fine-Motor
(Grooved
Pegboard Test,

FingerTapping
Test) factors

Age at outcome:
15-17 yr

scores,

and adolescent

marijuana

consumption

Beta

Visuoconstruction
Prenatal: -0.157 (-0.277,
-0.037)b

Average: 0.028 (-0.052, 0.108)b
78-month: 0.014 (-0.088, 0.116)b

Fine-motor

Prenatal: -0.017 (-0.056, 0.022)b

Average: -0.016 (-0.041, 0.009)b

78-month: -0.046 (-0.077,
—0.015)b

Bhattacharva
et al. (19951

Cincinnati, OH
US

1979-1984
(enrollment)

Cohort

Cincinnati Lead Blood
Program Project

N: 202

Pregnant
mothers living in
older houses in
poor condition
and with
chipping lead-
based paint and
lead laden dust

GM (SD) (min-max) ug/dL:

Prenatal maternal blood: 8.0 (1.58) (2-22)

Average Childhood blood (mean of 20 quarterly
concentrations obtained from birth to 5 years):
11.9 (1.5) (4-28)

Postural
balance,
including sway
area (SA) and
sway length (SL)

Age at outcome:
5 yr

Age, height, weight,
birth length, birth
weight, Hgb, TIBC,
Minimum middle-ear
pressure, smoking
during pregnancy,
HOME score at 36
months, foot area,
sports participation,
race, known
occurrences of
bilateral otitis media

Betas
Eyes open

SA: 0.059 (0.024, 0.093)
SL: 0.145 (0.088, 0.201)
Eyes closed
SA: 0.043 (0.01, 0.076)
SL: 0.121 (0.069, 0.173)
Eyes open, foam
SA: 0.046 (-0.175, 0.266)
SL: 0.113 (0.065, 0.16)
Eyes closed, foam
SA: 0.055 (0.018, 0.091)
SL: 0.86 (0.277, 1.443)

External Review Draft

3-402

DRAFT: Do not cite or quote

-------
Dietrich et al. |\|- 245
(19931

Cincinnati,
US

OH

1979-1984
(enrollment)

Cohort

Pregnant
mothers living in
older houses in
poor condition
and with
chipping lead-
based paint and
lead laden dust

Blood

Maternal and child venous blood

Age at measurement: T1 (maternal), delivery, 1

2, 3, 4, 5, 6 yr (child)

Mean (SD) (min-max) ug/dL:

T1: 8.4 (3.8) (1-27)

Neonatal: 4.8 (3.1) (1-22)

1	year: 10.5 (4.9) (3-35)

2	years: 17.1 (8.3) (6-49)

3	years: 16.2 (7.6) (4-50)

4	years: 14.0 (7.1) (4-45)

5	years: 11.9 (6.4) (3-38)

6	years: 10.1 (5.6) (2-33)

External Review Draft

3-403

Beta

Bilateral coordination
Prenatal: -0.04 (-0.197, 0.117)b
Neonatal: -0.15 (-0.326, 0.026)b
Average (3-60 months): -0.11
(-0.188, -0.032)b
Concurrent: -0.18 (-0.258,
—0.102)b

Visual-motor control
Prenatal: 0.06 (-0.097, 0.217)b
Neonatal: -0.1 (-0.296, 0.096)b
Average (3-60 months): -0.05
(-0.148, 0.048)b
Concurrent: -0.12 (-0.218,
-0.022)b

Upper-limb speed and dexterity

Prenatal: -0.2 (-0.435, 0.035)b

Neonatal: -0.45 (-0.724,
—0.176)b

Average (3-60 months): -0.19
(-0.327, -0.053)b
Concurrent: -0.31 (-0.447,
—0.173)b

Fine motor composite
Prenatal: -0.14 (-0.552, 0.272)b
Neonatal: -0.49 (-0.96, -0.02)b
Average (3-60 months): -0.28
(-0.515, -0.045)b

Concurrent: -0.46 (-0.715,
-0.205)b

DRAFT: Do not cite or quote

-------
Reference
and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

Wasserman et
al. (2000)

K. Mitrovica
and Pristina

Kosovo,
Yugoslavia

1985-1986
(enrollment)

Followed 54
months

Cohort

Yugoslavia

Prospective

Study

N: 283 children

Pregnant
women
recruited from
K. Mitrovica
(lead smelter,
refinery, and
battery factory)
and Pristina
(town 40 km
south)

Blood

Prenatal maternal, delivery, and subsequent 6-
month interval venous blood: NR

Fine motor and
gross motor
composites
assessed using
BOTMP

Visual motor
integration
assessed using
Beery Test of
VMI

Age at outcome:
54 months

Maternal age,
parental education,
number of siblings,
living arrangement,
HOME score at 3
years, maternal
intelligence at 2 years
(RSPM), birthweight,
BMI at 54 months,
child sex,
opportunities to
practice motor skill,
incomplete
lateralization

Beta for log—10 transformed Pb

Fine motor composite: -0.17

(-1.503, 1.163)c

Gross motor composite: 0.03

(-1.538, 1.598)c

VMI: -0.24 (-0.632, 0.152)c

tKim et al.

MOCEH study

(2013c) and

n: 884

Kim et al.



(2013b)

Mothers



recruited before

Seoul,

20th wk of

Cheonan and

pregnancy

Ulsan

between and

Korea

were in

2006-2010

locations

Followed 6 mo

(Seoul,

Cohort

Cheonan and



Ulsan)

Blood

Maternal blood samples measured for Pb, Cd in
early (<20 wk) pregnancy and late
(med = 39 wk) pregnancy

Age at measurement:

Early and late pregnancy

Early pregnancy: 1.4 (GM), 2.1 (90th), 9.8
(max) [jg/dL

Late pregnancy: 1.3 (GM); 2.1 (90th), 4.3
(max) [jg/dL

GM also available separately by 3 sites

PDI assessed
using BSID-II
(Korean version)

Age at outcome:
6 mo

Birth weight, infant
sex, maternal age
and education, family
income, breastfeeding
status, residential
area.

Beta

Early: 0.28 (
Late: -1.38 i

¦1.19, 1.75)
-3.31, 0.55)

Early:

Cd <1.47 |jg/L: 2.70 (0, 5.39)

Cd >1.47 |jg/L: -1.17 (-3.27,
0.94)

Late:

Cd <1.51 |jg/L: 0.18 (-2.70, 3.07)

Cd >1.51 |jg/L: -2.86 (-5.55,
-0.16)

External Review Draft

3-404

DRAFT: Do not cite or quote

-------
Reference
and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tKim et al.
(2018b)

4 cities: Seoul,
Anyang, Ansan
and Jeju
Korea

2011-2012
(enrollment)
Followed
through 24 mo

Cohort

CHECK cohort
n: 140

birth cohort-

pregnant

women

recruited from 4
cities in Korea
before delivery

Blood

Maternal blood; method NR
Age at measurement: delivery
Median (IQR):

Maternal: 2.7 (3.5, 5.7) pg/dL
Cord: 1.2 (0.8, 1.7) pg/dL

PDI assessed
using BSID-II
(Korean version)

Age at outcome:
13-24 mo

BPA, and phthalates,
maternal age
(continuous), birth
delivery mode
(categorical), monthly
household income
(categorical), child's
sex, and BDI
(continuous) of the
mother, gestational
age (continuous),
primiparous
(categorical), and
pre-pregnancy BMI
(categorical)

Beta (maternal blood)

Overall: -15.45 (-30.12, -0.79)
Boys: -18.32 (-45.35, 8.71)
Girls: -7.48 (-42.10, 27.15)

tY Ortiz et al. PROGRESS
(2017)

Mexico City
Mexico

Jul 2007-Feb
2011
Followed
through 24 mo

birth cohort
n: 536

Women <20 wk
of gestation and
planning to
reside in Mexico
City for the next
3 yr.

Blood

Maternal blood analyzed using ICP-MS.
Age at measurement:

T2, T3

Mean:

T2: 3.7 pg/dL
T3: 3.9 pg/dL

Motor

development
assessed using
BSID-III.

Standardized
scores (mean:
100, SD: 15).

Age at outcome:
24 mo

Infant sex, birth
weight, gestational
age, maternal age,
maternal IQ (WAIS
Spanish version),
HOME score.

Beta for log-transformed Pb
Motor Development:
T2: 1.97 (-2.46, 6.40)bc
T3: -11.01 (-17.55, -4.48)bc

Cohort

External Review Draft

3-405

DRAFT: Do not cite or quote

-------
Reference
and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tLiu et al.

Birth cohort

(2014c)

from 3 medical



centers

Pearl River

n: 362 mother-

Delta Region,

infant pairs (141

Guangdong

high Pb group

China

with cord BLL



>3.92 pg/dL and

Jan 2009-Jan

102 low Pb

2010

group

(enrollment)

<1.89 pg/dL)

Followed for



3 yr



Cohort



tRvaiel et al.

ELEMENT

Blood

Cord blood and child blood analyzed using
GFAAS. 2 exposure groups created based on
cord BLL below 25th percentile (low) and above
the 75th percentile (high). Age at measurement:
At birth (cord), 6, 12, 24 and 36 mo (postnatal
child)

High and low Pb groups: cord BLLs: 5.63 and
1.35 [jg/dL; 6 mo BLL: 4.03 and 2.85 pg/dL;
12 mo: 4.87 and 3.79 pg/dL; 24 mo: 4.39 and
3.31 pg/dL; 36 mo: 3.94 and 3.28 pg/dL

PDI assessed
using BSID-II
(Chinese
version)

Age at outcome:
36 mo

Birth weight, sex,
maternal education,
IQ (WISC-R),
hemoglobin level,
smoking, age,
parental occupations,
household annual
income, HNES total
score

Beta comparing PDI score at 36
mo in high exposed (Cord BLL
>3.92 pg/dL) vs. low exposed
(Cord BLL <1.89 pg/dL):

-1.302 (-1.572, -1.031)

(2021)

Mexico City
Mexico

1997-2005
Cohort

project
n: 85

Mother-child
pairs recruited
at the Mexican
Social Security
Institute

Blood

Maternal and child venous blood; ICP-MS,
GFAAS

Age at measurement:

Tl, T2, T3 (maternal); 12, 24 mo (child)

Maternal blood GM (SD):

PDI assessed
using BSID-II
(Spanish
version)

Age at outcome:
12-24 mo

Maternal IQ (WAIS),
maternal age, infant
weight, length, SES,
infant age and sex,
current infant BLL

T1
T2
T3

5.27 (1.93) pg/dL
4.74 (1.96) pg/dL
4.98 (1.93) pg/dL

A large number of results were
obtained from the mediation
analysis. In summary, T1, T2,
and T3 BLLs were associated
with nonsignificant decreases in
12-month PDI. This association
persisted for 24-month PDI at
less magnitude for T2 Pb.

Beta for 12-month PDI

Tl
T2
T3

Infant blood GM (SD):
12 mo: 3.92 (1.80) pg/dL
24 mo: 3.49 (1.93) pg/dL

-0.24 (-0.95, 0.48)
-0.38 (-1.10, 0.35)
-0.33 (-1.06, 0.40)

Beta for mediation by GCNT1
cgl8515027 methylation of In-
transformed T2 BLLs and 12-
month PDI:

Indirect: 1.25 (-0.11, 3.32)

External Review Draft

3-406

DRAFT: Do not cite or quote

-------
Reference
and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tShekhawat et n: 117
al. (2021)

Western

Rajasthan

India

2018-2019
(enrollment)

Follow-up at
6.5 mo
(average)

Cohort

Mother-child
pairs in third
trimester or at
delivery

Blood

Cord blood; ICP-OES

GM = 4.14 [jg/dL; mean = 4.77 ± 3.3 pg/dL;
median = 4.23 pg/dL
75th: 5.1 pg/dL

Motor

development
assessed using
BSID-III

Age at outcome:
6.5 mo

Maternal age,
gravida, gestational
age, maternal
education, child sex
and weight, preterm
birth, maternal food
intake during
pregnancy, smoking,
alcohol consumption,
maternal residential
and occupational
history, delivery type

(3 (95 % CI)

Umbilical cord Pb level
<5 pg/dL (n = 70)

Composite motor:
-0.048 (-0.28, 0.19)
Subscale fine motor:
-0.10 (-1.80, 0.68)

Subscale gross motor:

0.14 (-0.84, 0.94)

Umbilical cord Pb level 5.0-
10.5 pg/dL (n = 47)

Composite motor:

0.01 (-1.19, 0.23)

Subscale fine motor:
0.03 (-3.34, 4.1)

Subscale gross motor:

-0.29 (-5.00, 0.11)

Henn et al.
(2012)

N: 455

pregnancy or at
delivery

Women

Mexico City recruited during
Mexico

1997-2000
(enrollment)

Followed for
24 months

Blood

Child venous blood, ICP-MS

Age at measurement: 12, 24 months

12 months mean (SD): 5.1 (2.6) pg/dL
24 months mean (SD): 5.0 (2.9) pg/dL

PDI assessed
using BSID-II
(Spanish
version)

Age at Outcome:
12, 18, 24, 30,
36 months

Sex, gestational age,
hemoglobin, maternal
IQ, maternal
education, and visit

Beta

12-month BLL:
0.02)

24-month BLL:
0.17)

-0.27 (-0.56,

-0.18 (-0.53,

Cohort

External Review Draft

3-407

DRAFT: Do not cite or quote

-------
Reference

Study
Population









and Study
Design

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tParaiuli et al.

Pregnant

Blood

PDI assessed

Maternal age and

Beta

(2015a)

women visiting



using BSID-II

education, BMI,

-4.83 (-16.53, 6.86)



the Bharatpur

Cord blood; ICP-MS, measured for Pb, As and



gestational age,

Chitwan,

General

Zn

Age at outcome:

family income, parity,



Bharatpur

Hospital



24 mo

birth weight, weight at



District

n: 100

Age at measurement:



24 mo, child age



Nepal

Birth cohort:

At birth



assessment, As, Zn,
HOME score



Sep-Oct 2008

women were

Median: 2.06 [jg/dL



(smoking and alcohol



Followed

selected if living

Max: 22.08 pg/dL



consumption not



through 24 mo

in the study





included given low



area for at least





prevalence)



Cohort

2 yr and were at
term pregnancy
(>37 wk of
gestation) from









tParaiuli et al.

Birth cohort

Blood

PDI assessed

Maternal age and

Beta

(2015b)

from Bharatpur



using BSID-II

education, BMI,

-2.56 (-9.71, 4.59)



General

Cord blood; ICP-MS, measured for Pb, As and



gestational age,

Chitwan,

Hospital

Zn

Age at outcome:

family income, parity,



Bharatpur

n: 100



36 mo

birth weight, weight at



district



Age at measurement:



24 mo, child age at



Nepal

Resided in area
for at least 2 yr

At birth



assessment, As, Zn,
HOME score



Sep-Oct 2008

delivered at

Median: 2.06 pg/dL



(smoking and alcohol



Followed

term (i.e.,

Max: 22.08 pg/dL



consumption not



through 36 mo

>37 wk)





included given low









given low prevalence)



Cohort











External Review Draft

3-408

DRAFT: Do not cite or quote

-------
Reference
and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tJianq et al.
(2022)

Taipei
Taiwan

August 2008-

December

2009

(enrollment)

Follow-up 3
years

Cohort

N: 53 children
Meconium
(n=36)

Hair (n=52)

Fingernail
(n=43)

Longitudinal
birth cohort
study at a
medical center
hospital in
northern Taiwan

Meconium, Hair, Fingernail

All metals analyzed using ICP-MS.

Child meconium collected at birth
Child hair and fingernails collected at age 1
month

Median (min, max):

Meconium: 25.6 (2.00, 8815) ng/g
Hair: 3.61 (0.31, 25.1) pg/g

Fingernail: 0.84 (0.06, 24.3) pg/g

Motor

development
assessed using
BSID-III.

Raw total motor
scores were
standardized to
expected mean
of 100 and SD of
15. Raw fine
motor and gross
motor scores
were

standardized to
expected mean
of 10 and SD of
3.

Age at Outcome:
3 years

Maternal age and
education, newborn
birth head
circumference and
sex, and As and Cd
levels

Beta for log—10 transformed Pb
and log—10 transformed motor
development score

Meconium

Motor: -0.00001 (-0.021, 0.021 )e
Fine motor: 0.009 (-0.048,
0.065)e

Gross motor: -0.014 (-0.064,
0.036)e

Hair

Motor: 0.020 (-0.009, 0.049)e
Fine motor: 0.046 (-0.015,
0.107)e

Gross motor: 0.006 (-0.043,
0.054)e

Fingernails

Motor: -0.003 (-0.025, 0.019)e
Fine motor: -0.001 (-0.053,
0.052)e

Gross motor: -0.004 (-0.046,
0.037)e

External Review Draft

3-409

DRAFT: Do not cite or quote

-------
Reference
and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tZhou et al.

Shanghai

(2017)

Stress Birth



Cohort Study

Shanghai

n: 139

China



2010-2012

Women

(enrollment)

enrolled in

Followed for

prenatal clinics

24-36 mo after

of maternity

birth

hospitals during

Cohort

mid-to-late



pregnancy.

tLiu et al.

Guangxi Birth

(2022a)

Cohort Study



N: 703 children

Blood

Maternal blood Pb measured using AAS.
Age at measurement:

28-36 wk of gestation

GM: 3.30 pg/dL

Gross motor and
fine motor
development
assessed using
GDS (Chinese
version)

Age at outcome:
24-36 mo

Maternal age at
enrollment, economic
status, maternal
education, gestational
week, child sex, birth
weight and age

Beta per log—10 transformed BLL
Gross motor development: 3.31
(-6.11, 12.73)c

Fine motor development: 0.49
(-11.27, 12.24)c

Guangxi region
China

July-

September
2015

(enrollment)

Followed until
July-

September
2018 (3 years)

Pregnant
women
recruited from
eight maternity
and child
healthcare
hospitals in six
cities of
Guangxi

Blood, urine

Prenatal maternal serum (first, second, and third
trimesters)

Infant urine

Age at measurement: NR

Maternal serum med (25th, 75th): 0.78 (0.54,
1.24) pg/L

Infant urine med (25th, 75th): 0.22 (0.14, 0.37)
pg/L

Gross motor
development
using GDS
(Chinese
version)

Age at outcome:
2.57 (SD: 0.14)
yrs

Maternal age, pre-
pregnancy BMI,
children's age,
children's gender,
blood sampling time,
delivery mode,
delivery gestational
week, birth head
circumference.

Beta per In-transformed pg/L
increase in Pb

Overall: -2.321 (-3.614, -1.029)c
Male: -3.426 (-6.162, -0.691 )c
Female: -1.182 (-2.805, 0.442)c

Cohort

External Review Draft

3-410

DRAFT: Do not cite or quote

-------
Reference
and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tTavlor et al.
(2015)

Avon
UK

Apr. 1, 1991 -
Dec. 31, 1992
(expected
delivery date)
Followed 10
years

Cohort

Subsample of
ALSPAC Study

N: 582 child
blood
N:4285
prenatal
maternal blood

Pregnant
women in
former Avon
Health
Authority

Blood

Maternal blood collected in early pregnancy
(med: 11 weeks of gestation); ICP-MS

Child venous blood

Age at measurement: 30 months

Mean (SD)

Prenatal: 3.67 (1.47) ng/dL
Child: 4.22 (3.12) jig/dL

Balance (heel-
to-toe test) from
the Movement
Assessment
Battery for
Children
(Movement
ABC)

Age at outcome:
7 years

Static and
dynamic balance
tests based on
BOTMP

Age at outcome:
10 years

Sex, passive smoking
at 77 or 103 months
old (weekdays and
weekends), and
concurrent Ca and Fe
intakes

OR for >5 [jg/dL vs. <5 [jg/dL Pb
Prenatal Pb:

Heel-to-toe test: 1.01 (0.95, 1.01)

Dynamic balance: 1.02 (0.95,
1.09)

Static balance: 0.98 (0.92, 1.06)
Child Pb:

Heel-to-toe test: 0.98 (0.92, 1.05)
Dynamic balance: 1.01 (0.93,
1.09)

Static balance: 1.03 (0.94, 1.12)

External Review Draft

3-411

DRAFT: Do not cite or quote

-------
Reference
and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tTavlor et al.
(2018)

Avon
UK

Apr. 1, 1991 -
Dec. 31, 1992
(expected
delivery date)
Followed 10
years

Cohort

Subsample of
ALSPAC Study

N: 1558

Pregnant
women in
former Avon
Health Authority

Blood

Maternal blood; ICP-MS
Age at measurement:

Early pregnancy (med: wk 11 of gestation)
Mean (SD)

Prenatal: 3.66 (1.55) ng/dL

Balance (heal-
to-toe test), ball
skills (beanbag
toss), and
manual dexterity
(threading lace
and placing
pegs)

Movement

Assessment

Battery for

Children

(Movement

ABC)

Age at outcome:
7 years

Sex, maternal
education, smoking in
pregnancy, alcohol in
pregnancy, maternal
age and parity

OR for >5 [jg/dL vs. <5 [jg/dL Pb
Balance: 0.99 (0.74, 1.33)

Ball skills: 0.88 (0.58, 1.32)
Threading lace: 1.12 (0.83, 1.50)
Peg board - preferred hand: 1.19
(0.88, 1.60)

Peg board - non-preferred hand:
1.14 (0.85, 1.54)

OR for highest quartile (NR) vs.
lowest quartile (<5 (ig/dL) of Pb
Balance: 0.98 (0.73, 1.31)

Ball skills: 1.07 (0.71, 1.63)
Threading lace: 1.01 (0.75, 1.35)

Peg board - preferred hand: 1.23
(0.92, 1.66)

Peg board - non-preferred hand:
0.99 (0.73, 1.32)

External Review Draft

3-412

DRAFT: Do not cite or quote

-------
Reference
and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tBoucher et al.
(2016)

Nunavik,
Quebec

Canada

October 2005
- February
2010 (outcome
assessment)

Cohort

Nunavik Child Blood
Development
Study

N: 265 school
children

Cord blood and child concurrent venous blood,
ICP-MS

Age at measurement: birth (cord), 11.3 years
(child)

Phone

recruitment of
children with

umbilical cord Cord mean, median (SD): 4.7, 3.7 (3.4) pg/dL
blood samples
obtained under

the Arctic Cord Child blood mean, median, SD: 2.7, 2.0 (2.1)

Blood	MQ/dL

Monitoring

Program

Fine motor
performance on
Santa Ana Form
Board (manual
dexterity), Finger
Tapping (fine
motor speed),
and Stanford-
Binet Copying
(visuo-motor
integration)

Age at outcome:
11.3 years (SD:
0.8)

Form Board: child age
and sex, social
environment,
maternal age, parity,
marital status,
smoking during
pregnancy; Finger
Tapping: child age
and sex, social
environment;
Stanford-Binet: child
age and sex, social
environment, marital
status

Others considered:
adoption status,
primary caregiver's
years of education,
Peabody Picture
Vocabulary Test,
RPM, parity, mother
fluency in
English/French,
assimilation to
Western culture,
alcohol and illicit drug
use during
pregnancy, other
contaminants, nutrient
biomarkers

Beta for log-transformed Pb
Cord blood

Manual dexterity: -0.08df

Fine motor speed: -0.19 (-0.33,

-0.05)bf

Visuo-motor integration: -0.01df
Child blood

Manual dexterity: -0.17 (-0.34,

0)b,f

Fine motor speed: -0.21 (-0.37,
-0.05)bf

Visuo-motor integration: 0.1df

External Review Draft

3-413

DRAFT: Do not cite or quote

-------
Reference
and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tParaiuli et al.

n: 79

(2013)





women living in

Chitwan Valley

the study area

Nepal

(i.e., Chitwan)

Sep-Oct 2008

for at least 2 yr,

Cross-

at term

sectional

pregnancy



when the



mothers visited



the hospital



(more than



37 wk of



gestation), age



of 18-40 yr,



singleton birth,



and no report of



diabetes,



hypertension, or



preeclampsia

tLiu et al.

Birth cohort

(2014d)

n: 415 mother-



child pair (219

Shenzhen,

high Pb group

Guangdong

>4.89 pg/dL at

China

first trimester

Jan 2009-Jan

and 196 low Pb

2010

group

Followed for

<1.96 pg/dL)

26-30 wks



Cohort

Pregnant



women



recruited during



the early



pregnancy (10-



14 wk)

Blood

Cord blood Pb concentrations determined using
ICP-MS.

Age at measurement:
delivery

Neurodevelopm Maternal age, parity,

mean: 31.7 |jg/L;
75th: 35.1 pg/L
Max: 220.8 pg/L

median: 20.6 pg/L

ent assessed
using Brazelton
NBAS III

Age at outcome:
1 day old

mother's education
level; annual family
income, mother's
BMI, birth weight,
gestational age, age
of baby at NBAS
assessment

13 (95 % CI) change in score per
1 ug/L increase in blood Pb

Habituation: 1.44 (-1.19, 4.07)

Orientation: -0.12 (-8.34, 8.10)

Motor system: -2.15 (-4.27,
-0.03)

State organization: 2.15 (-1.58,
5.88)

State regulation: -0.75 (-3.86,
2.36)

Autonomic Stability: 0.71 (-0.48,
1.90)

Abnormal reflex: 1.07 (-1.32,
3.46)

Blood

Maternal, cord blood analyzed using GFAAS.
Maternal BLL classified as low or high

Age at measurement:

First, second and third trimester and at delivery

Low and High BLL groups: First trimester:
1.22 pg/dL and 6.49 pg/dL; second trimester:
1.01 pg/dL and 5.63 pg/dL; third trimester:
1.19 pg/dL and 6.31 pg/dL; and delivery:
1.26 pg/dL and 6.65 pg/dL

Neurodevelopm
ent assessed
using NBNA

Age at outcome:
3 days

Infant sex, maternal
hemoglobin, IQ,
tobacco use and
parents' occupation,
education, yearly
household income.

Beta for change in NBNA score
per log-transformed Pb

T1
T2
T3

-4.86 (-8.831, —0.889)f
-3.98 (-8.180, 0.220)f
-3.65 (-6.609, 1.309)f

Cord: -3.39 (-7.531, 0.751 )f

External Review Draft

3-414

DRAFT: Do not cite or quote

-------
Reference
and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tNozadi et al.
(2021)

Navajo Nation
United States

Enrollment
February 2013
- June 2018
Follow-up at
age 10-13 mos

Cohort

Navajo Birth
Cohort Study
(NBCS)

n: 327

Blood

Maternal blood Pb from the 36-week visit or at
the time of delivery was processed using ICP-
DRC-MS.

Age at Measurement:

Mean (SD) maternal age at birth = 27.4 (5.87)
years. Children assessed at 10 and 13 months.

GM = 0.410 Apg/dL; median
75th: 0.51 Apg/dL
95th: 1.20 Apg/dL

0.37 Apg/dL

Neurodevelopm
ent assessed
using Ages and
Stages

Questionnaire

Inventory

(ASQ:I)

Age at outcome:
10, 13 months

Multivariable linear
regression for fine
motor adjusted for
blood cadmium, urine
cesium, urine arsenic,
and mother's
education; gross
motor adjusted for
urine strontium

Beta

Fine motor: -0.63 (-1.19, -0.08)
Gross motor: 0.14 (-0.47, 0.75)

tKao et al.
(2021)

Taipei

Taiwan

2011-2014

Cross-
Sectional

recruited from
Taipei MacKay
Memorial
Hospital

n:139 children
less than 3
years of age

Hair, fingernails

Pb concentrations in hair and fingernails were
measured using ICP-MS

Age at Measurement:

Mean (SD) 2.8 (0.4) years (children under 3y)

GM (SD): hair 2.9 (4.8) i1/4g/g, nails 0.8 (5.1)
i1/4g/g

Motor

development
assessed using
BSID-III

Age at outcome:
2.8 ± 0.4 yrs

Sex, gestational age
at birth, age of the
house (years), leafy-
vegetable intake
(servings/week), and
the area of surface
roads within 100 m of
the residence

Regression results were not
reported because they were not
statistically significant.

External Review Draft

3-415

DRAFT: Do not cite or quote

-------
Reference
and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tNvanza et al.
(2021)

Northern
Tanzania
Tanzania

2015-2017
(enrollment)
Followed for
12 mo
Cohort

Mining and
Health
Prospective
Longitudinal
Study in
Northern
Tanzania
n: 439

Birth cohort of
mother-child
pairs recruited
in 2nd trimester

Maternal dried blood spots; ICP-MS, measured
for Pb, Hg, and Cd

Age at measurement:

T2

Median: 2.72 [jg/dL
75th: 4.25 pg/dL
Max: 14.5 pg/dL

Gross motor and
fine motor
development
assessed using
MDAT.

Scores in each
domain
classified as
normal (>90th
percentile on all
items in that
domain or <90th
percentile on
one or two items
in the domain) or
impaired (<90th
percentile on
more than two
items in a
domain).

Maternal age and
education, maternal
and paternal
occupation, number
siblings under 5 yr at
home, and family
SES, infant sex, age,
birth weight, height
and weight as a proxy
for nutritional status.
(Covariates with
p < 0.20 retained in
the final models.)

Prevalence ratio

Gross motor development: 1.0

(0.9, 1.0)

Fine motor development: 1.0
(0.9, 1.0)

Age at outcome:
between 6 and
12 mo

External Review Draft

3-416

DRAFT: Do not cite or quote

-------
Reference
and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% Clsa

tPalaniappan n = 755 school Blood
et al. (2011) children (age 3-

7 y)	Child venous blood; LeadCare Analyzer

Chennai

India	Children	Age at measurement: 3-7 yr

attending public
2003-3006 schools in	Mean (SD): 11.5 (5.3) [jg/dL

Chennai
Cross-	(kindergarten -

sectional	1st grade)

Visual-motor

(drawing),

visual-spatial

(matching), fine

motor

(pegboard)

subtests and

composite

assessed using

WRAVMA

Age at outcome:
3-7 yr

Standardized
scores (mean:
100, SD: 15)

Gender, age,
hemoglobin level,
average monthly
income of the family
(categorical) and
parent education
(categorical)

Beta

Drawing: -0.29 (-0.51, -0.07)
Matching: -0.14 (-0.31, 0.02)
Pegboard: -0.19 (-0.38, 0.01)
Composite: -0.26 (-0.45, -0.07)

AAS = atomic absorption spectroscopy; As = arsenic; BASC = Behavior Assessment System for Children; BDI = Beck Depression Inventory; BLL = blood lead level; BOTMP =
Bruininks-Oseretsky Test of Motor Proficiency; BPA = bisphenol A; BRIEF = Behavior Rating Inventory of Executive Functions; BSID = Bayley Scales of Infant and Toddler
Development; CBCL = Child Behavior Check List; CCAAPS = Cincinnati Childhood Allergy and Air Pollution Study; CHECK = Health and Environmental Chemicals in Korea;
CI = confidence interval; CKD = chronic kidney disease; CKiD = Chronic Kidney Disease in Children; CPRS = Conners Parent Rating Scale; C-TRF = Caregiver-Teacher Report
Form; CTRS = Conners Teacher Rating Scale; DSM = Diagnostic and Statistical Manual of Mental Disorders; ELEMENT = Early Life Exposure in Mexico to Environmental
Toxicants; FSIQ = full-scale intelligence quotient; GDS = Gesell Development Schedules; GFAAS = graphite furnace atomic absorption spectrometry; GM = geometric mean;

HOME = Health Outcomes and Measures of The Environment Study; IQ = intelligence quotient; MOCEH = Mothers' and Children's Environmental Health; NBAS = Neonatal
Behavioral Assessment Scales; NBNA = Neonatal Behavioral Neurological Assessment; NHBCS = New Hampshire Birth Cohort Study; NR = not reported; OR = odds ratio;
Pb = lead; PDI = Psychomotor Development Index; PHDCN = Project on Human Development in Chicago Neighborhoods; SDQ = Strengths and Difficulties Questionnaire;
SES = socioeconomic status; SRP = self-report of personality; SRS = Social Responsiveness Scale; T1 = first trimester of pregnancy; T2 = second trimester of pregnancy; T3 = third
trimester of pregnancy; WAIS = Weschler Adult Intelligence Scale; WISC-R = Wechsler Intelligence Scale for Children; WRAVMA = Wide Range Assessment of Visual-Motor
Abilities; Zn = zinc.

aEffect estimates are standardized to a 1 |jg/dL increase in BLL or a 10 |jg/g increase in bone Pb level, unless otherwise noted. For studies that report results corresponding to a
change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the biomarker and standardized
accordingly.

The CI was calculated from a p-value and the true CI may be wider or narrower than calculated.

°Results are unstandardized because the Pb level distribution data was not available.

dResults are unstandardized because they did not have an associated SE, CI, or p-value reported in the study.
eResults are unstandardized because the biomarker used for Pb exposure measurement is not blood, tooth, or bone.

'Results are unstandardized because the log base used for exposure transformation was unspecified in the study.
fStudies published since the 2013 Integrated Science Assessment for Lead.

External Review Draft

3-417

DRAFT: Do not cite or quote

-------
Table 3-11T Animal toxicological studies of Pb exposure and motor function.

Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported ((jg/dL)

Endpoints Examined

Moreira et al. (2001)

Rat (Wistar)

GD 0 to PND 22

Oral,

PND 23:

PND 23: OFT



Control (tap water), M,
n = 7-9



lactation
In utero

< 0.01 [jg/dL for Control

PND 70: OFT, Rotarod Test



500 ppm, M, n = 7





21.24 [jg/dL for 500 ppm
PND 70:

<	0.01 [jg/dL for Control

<	0.01 [jg/dL for 500 ppm



Leasure et al. (2008)

Mice (C57BL/6))

Control (tap water), M/F,
n = 12-18 (6-9/6-9))

27 ppm, M/F, n = 12-18
(6-9/6-9)

GD -14 to PND 10 Oral,

lactation
In utero

PND 0-10:

<	1 [jg/dL for Control

<	10 [jg/dL for 27 ppm
PND 30:

<	1 [jg/dL for Control

<	1 |jg/dL for 27 ppm

1 yr: Rotarod Test, Locomotor
Activity

External Review Draft

3-418

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported ((jg/dL)

Endpoints Examined

tFlores-Montova and Sobin

Mice (C57BL/6)

PND 0 to PND 28

Oral,

PND 28 - Males:

PND 28: OFT, Rotarod Test

(2015)

Control (distilled water),



drinking







M/F, n = 19 (8/11)



water

0.2 [jg/dL for Control









Oral,







30 ppm, M/F, n = 26



lactation

3.93 [jg/dL for 30 ppm





(16/10)

















9.39 [jg/dL for 230 ppm





230 ppm, M/F, n = 16











(12/4)





PND 28 - Females:











0.19 [jg/dL for Control











3.19 [jg/dL for 30 ppm











12.14 |jg/dLfor230 ppm



tZou etal. (2015)

Mice (ICR)

Control (distilled water),
M, n = 10

250 mg/L solution, M,
n = 10

-5 wk to 8 wk

Oral,

drinking

water

8 wk:

1.8 |jg/dL for Control
21.7 |jg/dL for 250 mg/L

8 wk: Rotarod Test,
Locomotor Activity

tRao Barkur and Bairv (2016)

Rat (Wistar)

Control (tap water), M,
n = 12

0.2% solution, PG, M,
n = 12

0.2% solution, G, M,
n = 12

0.2% solution, L, M,
n = 12

PG: GD -30 to
GD 0

G: GD 1 to GD 21

L: PND 1 to
PND 21

Oral,
lactation
In utero

PND 22:

0.19 [jg/dL for Control
3.03 [jg/dL for PG
5.51 [jg/dL for G
26.86 [jg/dL for L

PND 3, 4, 5: Surface Righting
Reflex, PND 6, 8, 10, 12:
Swimming Performance,
PND 8, 10, 12: Negative
Geotaxis, PND 14-18:
Ascending Wire Mesh,

External Review Draft

3-419

DRAFT: Do not cite or quote

-------
Study

Species ^Stock/Strain), Timing o, Exposure BLL As Reported (|jg/dL)	Endpoints Examined

tBetharia and Maher(2012) Rat (Sprague Dawley) GD 0 to PND 20 Oral,

PND 24:	lactation

In utero

Control (RO Dl water),

M/F, n = 11-13

10 |jg/mL, M/F, n = 11-
13

PND 59:

Control (RO Dl water),

M/F, n = 10-11

10	|jg/mL, M/F, n = 10-

11

PND 2:	PND 1-10: Surface Righting

Reflex, PND 24, 59: OFT

1.77 ng/g (0.188 pg/dL) for
Control

85.17 ng/g (9.02 pg/dL) for
10 |jg/mL

PND 25:

0.83 ng/g (0.088 |jg/dL) for
Control

9.21 ng/g (0.98 pg/dL) for
10 |jg/mL

PND 60:

0.23 ng/g (0.024 pg/dL) for
Control

0.30 ng/g (0.032 pg/dL) for
10 pg/mL

External Review Draft

3-420

DRAFT: Do not cite or quote

-------
Study	Species (Stock/Strain), Timing o, Exposure BLL As Reported (|jg/dL)	Endpoin.s Examined

tBasha and Reddv(2015)

Rat (Wistar)

GD 6 to GD 21

In utero

PND 21:

PND 4-7: Surface Righting



Control (deionized







Reflex, PND 8-10: Negative



water), M, n = 8





0.21 [jg/dL for Control

Geotaxis, PND 12 -16:











Forelimb Hang, PND 21,



0.2 % solution, M, n = 8





11.2 [jg/dL for 0.2% solution

PND 28, 4 mo: Locomotor











Activity









PND 28:











0.33 [jg/dL for Control











12.3 [jg/dL for 0.2% solution











4 mo:











0.19 [jg/dL for Control











5.9 [jg/dL for 0.2% solution



tTartaqlione et al. (2020)

Rat (Wistar)

GD -28 to PND 23

Oral,

PND 23:

PND 4, 7, 10, 12: Neonatal



Control (tap water), M/F



lactation



Spontaneous Movement,



n = 16 (9/7)



In utero

0.007 |jg/ml_ (0.7 [jg/dL) for

PND 4, 7, 10, 12: Surface









Control

Righting Reflex, PND 4, 7, 10,



50 mg/L, M/F, n = 16







12: Negative Geotaxis,



(9/7)





0.255 |jg/mL (25.5 pg/dL) for

PND 30: OFT









50 mg/L



tFaulketal. (2014)

Mice (Agouti)

GD -14 to PND 21

Oral,

PND 21 (Maternal BLL):

PND 90, 180, and 270:



Control (distilled water),



lactation



Locomotor Activity



M/F, n = 30



In utero

-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported ((jg/dL)

Endpoints Examined

tBashaetal. (2014)

Rat (Not Specified)

PND 1 to PND 21

Oral,

PND 45:

PND 45, 4 mo, 12 mo, 18 mo:



Control (deionized



lactation



OFT, Locomotor Activity



water), M, n = 6





0.42 [jg/dL for Control





0.2% solution, M, n = 6





49.5 [jg/dL for 0.2% solution











4 mo:











0.56 [jg/dL for Control











14.4 [jg/dL for 0.2% solution











12 mo:











0.46 [jg/dL for Control











6.96 [jg/dL for 0.2% solution











18 mo:











0.12 [jg/dL for Control











11.2 [jg/dL for 0.2% solution



tMansouri et al. (2012)

Rat (Wistar)

Control (distilled water),
M/F, n = 16 (8/8)

50 mg/L, M/F, n = 16
(8/8)

PND 70 to
PND 100

Oral,

drinking

water

PND 100 - Males:
2.05 [jg/dL for Control
8.8 [jg/dL for 50 mg/L
PND 100 - Females:

PND 100: OFT, Rotarod Test

2.17 [jg/dL for Control
6.8 [jg/dL for 50 mg/L

External Review Draft

3-422

DRAFT: Do not cite or quote

-------
Study	Species (Stock/Strain), Timing o, Exposure BLL As Reported (|jg/dL)	Endpoin.s Examined

tDuan etal. (2017)	Mice(CD1)	PND 1 to PND 21 Oral,	PND21:	PND 7, 11, 15, 19: OFT

Control (distilled water),	lactation

M/F, n = 5	16.2 |jg/L (1.6 |jg/dL) for Control

27 ppm, M/F, n = 5	191.8 pg/L (19.2 pg/dL) for

27 ppm

109 ppm, M/F, n = 5

283.4 pg/L (28.3 pg/dL) for
109 ppm

PND 35:

14.3 pg/L (1.4 pg/dL) for Control

283.4 pg/L (28.3 pg/dL) for
27 ppm

376.9 pg/L (37.7 pg/dL) for
109 ppm

tWanqetal. (2016)

Rat (Sprague Dawley)

PND 24 to PND 56

Oral,

PND 56: PND 60-66: OFT



Control (tap water), M,



drinking





n = 7



water

11 pg/L (1.1 pg/dL) for Control



100 ppm, M, n = 9





133 pg/L (13.3 pg/dL) for 100 ppm

tShvachiv et al. (2018)

Rat (Wistar)

Intermittent

Oral,

PND 196: PND 189: OFT



Control (tap water), M/F,

Exposure: GD 7 to

drinking





n = 8

PND 84, PND 140

water

<0.1 pg/dL for Control





to PND 196

Oral,





0.2% (p/v) solution



lactation

18.8 pg/dL for 0.2% (Intermittent)



(distilled water), M/F,

Continuous

In utero





n = 9 - Intermittent

Exposure: GD 7 to



24.4 pg/dL for 0.2% (Continuous)



exposure

PND 196







0.2% (p/v) solution, M/F,









n = 9 - Continuous









exposure







External Review Draft

3-423

DRAFT: Do not cite or quote

-------
Study	Species (Stock/Strain), Timing o, Exposure BLL As Reported (|jg/dL)	Endpoin.s Examined

tStansfield etal. (2015)

Rat(Long-Evans)

Control (chow), M/F,
n = 11-23

1500 ppm, M/F, n = 11-
23

GD Oto PND 50

Oral, diet
Oral,
lactation
In utero

PND 50: PND 50: Locomotor Activity
0.6 [jg/dL for Control
22.2 |jg/dL for 1500 ppm

tNeuwirth et al. (2019a)

Rat (Long-Evans)

GD Oto PND 22

Oral,

PND 22: PND 36-45: OFT



Control (tap water), M/F,



lactation





n =48 (30/18)



In utero

-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported (pg/dL)

Endpoints Examined

tSobolewski et al. (2020)

Mice (C57BL/6)

FO:

Control (distilled Dl
water), F, n = 10

100 ppm, F, n = 10

F1:

see Figure 1, n = 12
F2:

see Figure 1, n = 12
F3:

see Figure 1, n = 8-10

F1: GD -60 to PND Oral,
23-27	lactation

In utero

F1 PND 6-7:

0 [jg/dL for Control

12.5 pg/dL for 100 ppm (F0
dosing)

F3 PND 6-7:

0 ng/dL for Control

0 [jg/dL for 100 ppm (F0 dosing)

PND 60-120 (variable by
endpoint): Locomotor Activity

tSinqh et al. (2019)

Rat (Wistar) 3 mo to 6 mo

Control (distilled water),

M, n = 5

2.5 mg/kg, M, n = 5

Oral,
gavage

6 mo:

5.76 [jg/dL for Control
28.4 pg/dL for 2.5 mg/kg

6 mo: Locomotor Activity,
Rotarod Test

tViqueras-Villasenor et al.

Rat (Wistar) GD 0 to PND 21

Oral,

PND 110:

PND 90 to PND 110:

(2021)

Control (tap water), M,

lactation



Locomotor Activity



n = 8

In utero

2.04 [jg/dL for Control





320 ppm, M, n = 8



26.3 [jg/dL for 320 ppm



External Review Draft

3-425

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported (pg/dL)

Endpoints Examined

tAI-Qahtani et al. (2022)

Mice (Albino)

Control (distilled water),
M, n = 10

0.2 mg/kg, M, n = 10

8-9 wkto 14-15 wk

Oral,
gavage

14-15 wk:

1.2 |jg/100 mL (1.2 pg/dL) for
Control

7.1 |jg/100 mL (7.1 pg/dL) for
0.2 mg/kg

NR: Locomotor Activity

BLL = blood lead level; F# = filial generation; F = female; GD = gestational day; LOD = limit of detection; M = male; MRI = magnetic resonance imaging; mo = month(s);

NaAc = sodium acetate; NR = not reported; OFT = open-field test Pb = lead; PG = pregestation; PND = postnatal day; wk = week(s); yr = year(s).

aEffect estimates are standardized to a 1 |jg/dL increase in BLL or a 10 |jg/g increase in bone Pb level, unless otherwise noted. For studies that report results corresponding to a
change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the biomarker and standardized accordingly.

fStudies published since the 2013 Integrated Science Assessment for Lead.

External Review Draft

3-426

DRAFT: Do not cite or quote

-------
Table 3-12E Epidemiologic studies of Pb exposure and sensory organ function in children.

Reference and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates
and 95% CIs

Dietrich et al. (1992)

Cincinnati, U.S.
Cross-sectional

The Cincinnati
lead study cohort

n: 259

Blood

Age at measurement: prenatal-5
years

Mean (SD) [jg/dL:

Prenatal 8.2 (3.8)

Neonatal 4.8 (3.3)

Central auditory
processing abilities and
cognitive developmental
status

Age at outcome: 5 years

Measures of fetal distress
and growth, perinatal
complications, postnatal
indices of health and
nutritional status,
sociodemographic
characteristics, and
psychosocial features of the
home environment

Betab

Filtered Word
Score (total
number of words
correctly identified
in both ears)

Prenatal: -0.12
Neonatal: -0.26
Mean lifetime
through 5 years:
-0.07

Schwartz and Otto (1991)
HHANES, U.S.
Cross-sectional

Hispanic Health
and Nutrition
Examination
Survey
n: 3545

Blood

Age at measurement: 6-19 years

Median (25th' 75th) pg/dL:

Mexican Americans 8 (6, 11)
Cuban American 8 (6, 10)

Puerto Ricans 8 (6, 11)

Elevated hearing
threshold

Audiometric evaluations
were performed for all
subjects Beltone model
200-C audiometers were
used in the survey;
Hearing threshold was
defined as the lowest
intensity of a pure tone
that was just audible to
the subject.

Age at outcome: 6-19
years

NR

An increase15 in
BLL from 7
~microg/dl to 18
pg/dl was
associated with an
approximately 2-dB
loss of hearing at
all frequencies

External Review Draft

3-427

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates
and 95% CIs

Schwartz and Otto (1987)

NHANES II
U.S.

Cross-sectional

NHANES II	Blood

n: 4519	Age at measurement: 4-19 years

Range of Pb: 6 to 47 [jg/dL

Hearing thresholds

Standard Beltone Model
200C audiometers were
used and ca liberated
weekly with B&K Model
2203 sound level meters
in accordance with 1969
ANSI specifications.
Tests were conducted at
500, 1000, 2000, and
4000Hz on each ear.

Age at outcome: 4-19
years

Race, lead, ear discharge,
cold in last 2-week, other
ear condition, chronic ear
discharge, income, dietary
calcium, sex, current cold,
ringing in ear(s), earache,
previous running ear,
diagnosed hearing
impairment, degree of
urbanization, head of
household education level

The risk of
elevated hearing
thresholds at 500,
1000, 2000 and
4000 Hz increased
with increasing
PbB for both ears

tYin etal. (2021)

n: 234-7596 in
studies

Blood	Hearing loss

Age at measurement: 3-87 years

All studies included in the
meta-analysis controlled for
age and sex. Adjustment for
other potential confounders
varies by studies, but
includes monthly income,
education levels, smoking
status, BMI, ethnicity, work
duration, ototoxic
medication, blood lead,
occupational noise, loud
noise, and firearm noise,
and hypertension and
diabetes

OR (95% Cl)b
1.53 (1.24,1.87)

tChoi and Park (2017)

Korea

2010-2012

Cross-sectional

KNHANES
n: 5187 adults
and 853
adolescents

Blood

Hearing loss (>15dB) at
speech frequency;
Hearing loss (>15dB) at
high frequency

Age, age squared, sex,
education, BMI, current
cigarette smoking

OR (95% Cl)b

Hearing Loss (>15
dB) High-frequency
PTA

Pb Quartile 2
(0.978-1.260):
0.89 (0.39, 2.03)

External Review Draft

3-428

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates
and 95% CIs

Graphite furnace atomic
absorption spectrometry

Age at Measurement:
adolescents 12-19 years
(mean±SE 15.6±0.10)

Geometric mean (95% CI) (age-
adjusted): 1.26 [jg/dL (1.22, 1.30)

Pure-tone air conduction
hearing thresholds were
obtained for each ear at
frequencies of 0.5, 1, 2, 3,
4, and 6 kHz over an
intensity range of -10
to 110 dB -10 to 110 dB.

Age at outcome: 12-19
years

Pb Quartile 3
(1.261-1.557):

1.88 (0.83, 4.25)

Pb Quartile 4
(1.562-5.904):

1.38 (0.63, 3.02)
Per doubling of Pb:
1.26 (0.73, 2.16)
Hearing Loss (>15
dB) Speech-
frequency PTA

Pb Quartile 2
(0.978-1.260):
1.17 (0.41, 3.32)

Pb Quartile 3
(1.261-1.557):

1.08 (0.38, 3.08)
Pb Quartile 4
(1.562-5.904):
1.24 (0.34, 4.49)
Per doubling of Pb:
1.2 (0.48, 3.05)

External Review Draft

3-429

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates
and 95% CIs

tXu et al. (2020)

China

October-December 2014
Cross-sectional

n: 116

Blood	DNA methylation and

hearing loss

Graphite furnace atomic

absorption spectrometry (GFAAS, Age at outcome: 3-7

Jena Zeenit 650, Germany)	years

Age at measurement: 3-7 years

Median ± SEM (P25, P75):

Exposed group

5.29 ± 0.29 (3.61, 7.40)

Refence group

3.63 ± 0.24 (2.98, 4.77)

Both continuous variables
for child age, gender,
weight, height and BMI, and
categorical variables for
presence of family member
smoking, residence distance
to the road, residence
nearby noise, residence
renovation noise within a
year, often listening music
with earphones within a
year, often watching
television programs in loud
noise, and often play (i.e.,
toys or music, etc.) in loud
noise

Beta (95% Cl)b

Q1 0.139 (0.007,
2.968

Q2 0.051 (0.003,
0.977)

Q3 0.16 (0.016,
1.58)

Q4 2.765 (1.795,
15.237)

OR (95% CI)

Hearing loss in
both ears 1.40
(1.06, 1.84)

Left ear 1.46 (1.12,
1.91)

tSharqorodskv et al. (2011)
NHANES, U.S.

2005-2008
Cross-sectional

NHANES
n: 2535

Blood

Inductively coupled plasma mass
spectrometry

Age at Measurement: 12-19
years

Weighted Mean (95% CI):
Age 12-13: 1.00 pg/dL
(0.92-1.09 pg/dL)

Age 14-15: 0.93 pg/dL
(0.87-0.99 pg/dL)

Age 16-17: 0.85 pg/dL
(0.79-0.91 pg/dL)

Age 18-19: 0.93 pg/dL
(0.84-1.03 pg/dL)

Any Hearing Loss (>15
dB),

High-Frequency Hearing
Loss, Low-Frequency
Hearing Loss

Age at outcome: 12-19
years

Age, sex, race-ethnicity,
PIR, history of 3 or more ear
infections, loud noise
exposure, and smoking

OR (95% Cl)b (<1
|jg/dL reference)

Any >15 dB

1-1.99 pg/dL 0.99

(0.67-1.46)

>2 pg/dL 1.95

(1.24-3.07)

High-Frequency

1-1.99 pg/dL 1.20

(0.80-1.80)

>2 pg/dL 2.22

(1.39-3.56)

Low-Frequency

1-1.99 pg/dL 1.24

(0.82-1.86)

>2 pg/dL 1.13

(0.61-2.07)

External Review Draft

3-430

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates
and 95% CIs

tLiuetal. (2018c)

Guiyu (e-waste recycling
area) & Haojing (exposure
control, no e-waste
processing), China

2014

Cross-sectional

n: 234 (146
exposed; 88
reference)

Blood

Graphite furnace atomic

absorption spectrometry (GFAAS,
Jena Zeenit 650, Germany)

Age at Measurement:

3-7 years

Mediant SE: 4.94± 0.20 [jg/dL in
exposed; 3.85± 1.81 pg/dL in
reference

Hearing loss, Low
frequency hearing loss,
High frequency hearing
loss

Age at outcome: 3-7
years

Child age, gender, weight,
height, BMI, parent
education level, family
member smoking, family
monthly income, residence
distance to the road,
residence nearby noise,
residence renovation noise
within a year, often listening
to music with earphones
within a year, often watching
television programs in loud
noise, and often play (i.e.,
toys or music, etc.) in loud
noise

OR (95% Cl)b

Hearing loss total
1.24 (1.029, 1.486)
Low frequency
1.02 (0.869, 1.190)
High frequency
1.08 (0.839, 1.379)

External Review Draft

3-431

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates
and 95% CIs

tPawlas etal. (2015)

Upper Silesia, Poland
1996-2001 and 2008-2010
Cross-sectional

Two cohorts
merged

n: 483

Blood

Graphite furnace atomic

absorption spectrometry

Age at Measurement:
4-13 years

Median: 4.50 [jg/dL

Pure-tone audiometry
(PTA),

Brainstem auditory
evoked potentials
(BAEP),

Acoustic otoemission

Age at Outcome: 4-13
years

Cohort, mother's
education (dichotomized
into 'secondary school or
higher', or 'less',
corresponding to primary
and apprenticeship) and
smoking during pregnancy,
and the child's sex, birth
weight, apgar score, history
of mumps, age, and
pressure in middle ear on
both sides

Beta (95% Cl)b

ALAD Msp/
ALAD1-1 0.3
(0.15, 0.45)
ALAD*2 0.42
(-0.03, 0.87)
ALAD Rsa1
TT+TC 0.3 (0.1,
0.5)

CC 0.2 (-0.05,
0.45)

VDR Bsml
bb 0.03 (-0.22,
0.28)

Bb+BB 0.4 (0.25,

0.55)

VDR taq1

TT 0.04 (-0.21,

0.29)

Tt+tt 0.4 (0.2, 0.6)
VDR fokl
FF+Ff 0.4 (0.25,
0.55)

ff-0.1 (-0.6, 0.4)

External Review Draft

3-432

DRAFT: Do not cite or quote

-------
tSilver et al. (2016)

Sanhe County, Hebei
Province, China

November 2009- November
2011

Cohort

n: 391 (ARB:

auditory
brainstem
response), 1148
(VA: visual
acuity)

Maternal Blood

Atomic absorption spectrometry
(A AS)

Age at Measurement:

Pregnant woman 18 years or
older

Mean (SD) gestational age at
mid-pregnancy visit

ABR subset 15.7 (2.2) weeks
VA subset 15.5 (1.9) weeks
Mean (SD) gestational age at
late-pregnancy
ABR subset 38.8 (1.3) weeks

VA subset 39.3 (1.3) weeks
Mean (SD) gestational age at
birth

ABR subset 39.2 (1.1) weeks
VA subset 39.7 (1.1) weeks

ABR Pb median
2.9 [jg/dL at mid-pregnancy
3.0 [jg/dL at late-pregnancy
<2.0 [jg/dL at birth (cord blood)
GM (SD)

2.4 (2.5) [jg/dL at mid-pregnancy
2.7 (2.3) [jg/dL at late-pregnancy
<2.0 [jg/dL at birth (cord blood);
VA median

2.9 [jg/dL at mid-pregnancy, 3.3
[jg/dL at late-pregnancy, 2.1
[jg/dL at birth (cord blood)

GM (SD)

2.4 (2.6) [jg/dL at mid-pregnancy
2.9 (2.2) [jg/dL at late-pregnancy
<2.0 [jg/dL at birth (cord blood)

External Review Draft

3-433

Auditory brainstem
response (ABR);

Grating visual acuity (VA)

Age at Outcome:

ABR mean 2 days old
VA mean 6 weeks old

Sex, age attesting, cord
blood iron status,
gestational age, birth
weight, head circumference

Beta (95% Cl)b

ARB C-P ratio
Mid pregnancy
lead High (>3.8
[jg/dL) 0.02 (-0.01
-0.05)

Mid pregnancy
lead Med. (2-3.8
[jg/dL) 0.02 (-0.01
-0.05)

Late-pregnancy
lead High (>3.8
[jg/dL) 0.05 (0.02,
0.08)

Late-pregnancy
lead Med. (2-3.8
[jg/dL) 0.03 (0.01,
0.06)

Cord lead High
(>3.2 (jg/dL) 0
(-0.02, 0.03)

Cord lead Med. (2-
3.2 (jg/dL) 0
(-0.03, 0.03)

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates
and 95% CIs

tAlvarenqa etal. (2015)
Brazil

Followed 35 5 months

Contemporary cross-
sectional cohort

n: 130 children
(80 males & 50
females)

Blood

Atomic absorption
spectrometry with graphite
furnace

Age at Measurement:

18 months -14 years (Mean: 6
years 8 months! 2 years 3
months)

Mean: 12.2 pg/dL; SD=5.7 pg/dL
Median: 10.2 [jg/dL

Auditory brainstem
response

Age at outcome: 18
months -14 years

Age, gender, cumulative
blood lead levels, and

date of the audiological
assessment

Beta (95% Cl)b

Wave III, in relation
to wave I
Constant 4.00
(3.97, 4.04)

Wave I RE 0.58
(0.44, 0.72)

Male RE 0.09
(0.05, 0.13)
Constant 4.03
(3.99, 4.06)

Wave I LE 0.61
(0.45, 0.77)

Male LE 0.07
(0.03, 0.11)

Wave V, in relation
to wave III
Constant 5.77
(5.74, 5.80)

Wave I RE 0.81
(0.68, 0.94)

Male RE 0.073
(0.03, 0.11)
Constant
5.78 (5.75, 5.81)
Wave I LE 0.85
(0.73, 0.97)

Male LE 0.08
(0.05, 0.12)

External Review Draft

3-434

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study
Population

Exposure Assessment

Outcome

Confounders

Effect Estimates
and 95% CIs

tFillion et al. (2013)

Lower Tapajos River Basin,
State of Para, Brazil

May to July 2006

Cross-sectional

n: 228

Blood

Inductively coupled plasma mass
spectrometry (ICP-MS, Perkin
Elmer DRC II)

Age at Measurement:
15-66 years (median=33.0
years)

Mean=12.8±8.4 pg/dL;
Median=10.5 [jg/dL

Contrast sensitivity

Age, sex, current smoking Beta (95% Cl)b

(cycles per degree, cpd); (yes vs. no), current drinking Spatial frequency

Acquired color vision loss
(color confusion index,
CCI)

Age at outcome: 15-66
years

(yes vs. no)

with %EPA
1.5 cpd -1.32
(-4.30; 1.65)

3 cpd 2.06 (-2.87;
6.99)

6 cpd 0.60 (-6.04;
7.25)

12 cpd -13.33
(-23.28; -3.49)
18 cpd -2.43
(-6.64; 1.79)
CCI 0.16 (-0.03;
0.33)

BLL = blood lead level; CI = confidence interval; OR = odds ratio; Pb = lead; PTA = pure-tone average.

aEffect estimates are standardized to a 1 |jg/dL increase in BLL or a 10 |jg/g increase in bone Pb level, unless otherwise noted. For studies that report results corresponding to a
change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the biomarker and standardized accordingly.

bEffect estimates are not standardized because data pertaining to the BLL distribution and/or base for the log-transformation were not reported.

fStudies published since the 2013 Integrated Science Assessment for Lead.

External Review Draft

3-435

DRAFT: Do not cite or quote

-------
Table 3-13E Epidemiologic studies of Pb exposure, social cognition, and behavior in children.

Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% CIs

tKimetal. (2016)

South Korea
2005-2006
(enrollment); 2009-
2010 (follow-up)
Cohort

CHEER study Blood
n: 2,437

Child blood; GFAAS
Children recruited Age at measurement:
from 33 elementary 7_8 yr 0|di g_10 yr old, and 11-
12 yr old

schools across 10
Korean cities

GM (pg/dL):

7-8 y: 1.64; 9-10 y: 1.58; 11-12
y: 1.58

75th (pg/dL):

7-8 y: 2.36; 9-10 y: 2.08; 11-12
y: 2.05
95th (pg/dL):

7-8 y: 3.47; 9-10 y: 3.05; 11-12
y: 3.05

Autistic behaviors

Parent responses to
ASSQ and SRS

Age at outcome:
11-12 yr

Child sex, fetal and
environmental tobacco
smoke, parental
education levels,
family income, low
birth weight,
breastfeeding,
gestational age, fish
intake, and blood Hg
level

Change in SRS Scores*
Exposure at 7-8 yrs
1.37 (0.75, 1.98)
Exposure at 9-10 yrs
0.56 (-0.33, 1.44)
Exposure at 11-12 yrs
0.39 (-0.47, 1.25)

Change in ASSQ Scores*
Exposure at 7-8 yrs
0.09 (0.03, 0.14)
Exposure at 9-10 yrs
-0.02 (-0.09, 0.05)
Exposure at 11-12 yrs
0.03 (-0.04, 0.10)

*Higher score indicates more
autistic behaviors

OR Autism (ASSQ > 17)
Exposure at 7-8 yrs
1.45 (1.10, 1.93)
Exposure at 9-10 yrs
0.86 (0.60, 1.23)
Exposure at 11-12 yrs
0.97 (0.70, 1.35)

External Review Draft

3-436

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% CIs

tArora et al. (2017) Roots of Autism Tooth

Sweden

2011-2016
Cohort

and ADHD Twin
Study in Sweden
n: 32 twin pairs
and 12 individual
twins

Monozygotic and
dizygotic twins
discordant for
ASD; discordance
defined as >2
points differences
on the Autism-Tics,
ADHD and other
Comorbidities
subscale

Shed deciduous teeth, validated
by maternal, cord, and serial
child blood Pb; laser ablation
ICP-MS

Age at measurement: estimating
various timepoints of prenatal
and postnatal Pb exposure

Mean NR

ASD diagnosis

ADOS-2, SRS-2
among discordant twins
for ASD (ICD10 [Autism
or Asperger's]; DSM-5
[ASD])

Age at outcome:

8-12 yr

Genetic factors

Child sex, zygosity,
gestational age, the
average birth weight of
the twin pairs, and the
SD of the birth weight
in the twin pairs.

OR of log-transformed Pb for
ASD case vs. non=ASD twin
control: 1.5 (0.9, 2.5)bd

More quantitative results
depicted graphically (see
Figure 3-2)

tSkoqheim et al.
(2021)

Nationwide
Norway

2002-2009
(enrollment)

Case-control

Norwegian Mother,
Father and Child
Cohort Study
(MoBa)

n: 397 ASD cases,
1034 controls

Children from a
birth cohort

Blood

ASD diagnosis

Maternal whole blood; ICP-SFMS NPR
Age at measurement:
wk 17 of gestation

Exposure Quartiles:

Q1
Q2
Q3
Q4

0.16-0.65 [jg/dL
0.65-0.86 [jg/dL
0.86-1.12 [jg/dL
1.12-8.24 [jg/dL

Age at outcome: NR

Birth year and child
sex-matched controls

Child sex, birth weight,
birth year, and SGA,
maternal age at
delivery, education,
parity, pre-pregnancy
BMI, kg/m2), self-
reported smoking and
alcohol intake during
pregnancy, FFQ-
based estimates of
seafood intake (g/day),
and dietary iodine
intake (pg/day)

OR for In-transformed Pb

Q1:

Ref.





Q2:

0.80

(0.57,

1.12)'

Q3:

0.79

(0.56,

1.12)'

Q4:

0.81

(0.57,

1.15)'

External Review Draft

3-437

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% CIs

tRahbar et al. (2015) Jamaican Autism
Study

Blood

Kingston
Jamaica

December 2009-
March 2012

Case-control

n: 100 cases; 100 Child venous blood; ICP-MS

controls

Children 2-8 yrat
enrollment

Age at measurement:

2-8 yr

GM (SD) (cases): 2.25 (2.23)
pg/dL

GM (SD) (controls): 2.73 (1.85)
pg/dL

ASD diagnosis

DSM-IV-TR criteria,
ADOS

Age at outcome:
2-8 yr

Age, sex-matched
controls

maternal age, parental
education levels,
parish at child's birth,
SES (i.e., car
ownership by the
family), consumption
of shellfish (lobsters,
crabs), and Teflon use
(pots, pans, and
dishes) for cooking

GMD for In-transformed Pb
(ASD Cases vs. Controls):

-0.17 (-0.86, 0.52)bc

tRahbar et al. (2021) n: 30 cases; 30

Karachi
Pakistan

Study years NR

Case-control

controls

children at clinics
affiliated with Aga
Khan University

Blood

Child venous blood; ICP-MS
Age at measurement:

2-8 yr

GM (cases): 7.11 [jg/dL;
GM (controls): 8.48 [jg/dL

ASD diagnosis

DSM-IV-TR criteria,
ADOS

Age at outcome:
2-12 yr

Age, sex-matched
controls

GMD for In-transformed Pb
(ASD Cases vs. Controls):

maternal age, parental -1.37 Mg/dL (-3.28, 0.54)bc

education level, and

SES (i.e., car

ownership by the

family) and dietary

consumptions

dummy variables that
represented the
matched pairs

External Review Draft

3-438

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% CIs

tDonq et al. (2022)

Northeast China

October 2017-
January 2020

Case-Control

n: 512 children with Blood
ASD

Children diagnosed
with ASD at First
Hospital of Jilin
University

Child serum

Age at measurement: 2-13 yr
Mean (SD)

Mild Autism: 2.58 (1.08) pg/dL

Moderate/severe: 2.58 (1.08)
pg/dL

ASD severity

Severity of autism
symptoms determined
by CARS

Age at outcome 2-13 yr

Age, place of
residence, caregivers,
parental education
level, gastrointestinal
problems.

Also considered sex,
siblings, parental age
at pregnancy,
household income,
family history of mental
illness, vitamin intake
during pregnancy,
eating problems,
sleeping problems,
gastrointestinal
problems, ADHD
comorbidity

Beta

0.03 (0.01, 0.05)c

tRvaiel et al. (2021) ELEMENT project Blood

Orientation/engagement Maternal IQ (WAIS),

Mexico City
Mexico

1997-2005
Cohort

n: 85

Mother-child pairs
recruited at the
Mexican Social
Security Institute

Maternal and child venous blood;
ICP-MS, GFAAS

Age at measurement:

Tl, T2, T3 (maternal); 12, 24 mo

(child)

Maternal blood GM (SD):
T1: 5.27 (1.93) pg/dL
T2: 4.74 (1.96) pg/dL
T3: 4.98 (1.93) pg/dL

and emotional
regulation

ORIEN and EMOCI
scores from BRS of
BSID-IIS

Age at outcome:
24 mo

12-

maternal age, infant
weight, length, SES,
infant age and sex,
current infant BLL

A large number of results were
obtained from the mediation
analysis. In summary, T2 BLLs
were consistently inversely
associated with 24-month
EMOCI and ORIEN scores

Beta

24-month EMOCI at T2:
-1.13% (-2.63, 0.37)

24-month ORIEN at T2:
-0.98% (-2.83, 0.88)

Infant blood GM (SD):
12 mo: 3.92 (1.80) pg/dL
24 mo: 3.49 (1.93) pg/dL

External Review Draft

3-439

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders Effect Estimates and 95% CIs

tShekhawat et al.
(2021)

Western Rajasthan
India

2018-2019
(enrollment)
Follow-up at 6.5 mo
(average)

Cohort

n: 117

Mother-child pairs
in third trimester or
at delivery

Blood

Cord blood; ICP-OES

GM = 4.14 [jg/dL;
mean = 4.77 ± 3.3 [jg/dL;
median = 4.23 [jg/dL

75th: 5.1 [jg/dL

Social-emotional
development score
using BSID-III

Maternal age, gravida,
gestational age,
maternal education,
child sex and weight,

13 (95 % CI) for socio-emotional
development scores

Pb < 5 [jg/dL: 0.19 (-0.46,

0.46)

Age a, outcome: 6.S mo	05

(-0.60, 0.86)
pregnancy, smoking, v ' '

alcohol consumption,

maternal residential

and occupational

history, delivery type

tNozadi et al. (2021)

Navajo Nation
United States

February 2013-June
2018 (enrollment)

Followed through 10-
13 mo

Cohort

Navajo Birth
Cohort Study
(NBCS)

n: 327

Children of
mothers (age 14-
45 yr) living across
Navajo Nation with
community
exposure to metal
mixtures from
abandoned
uranium mines

Blood

Maternal blood,
DRC-MS.

child blood; ICP-

Age at measurement:

Delivery or 36-wk visit (maternal);
10, 13 mo (child)

GM = 0.410 [jg/dL;
median = 0.37 [jg/dL
75th: 0.51 pg/dL
95th: 1.20 pg/dL

Communication and
personal-social domain
scores using the ASQ:I.
Age-adjusted scores.

Age at outcome:
13 mo

10-

Age.

Also considered
maternal age, marital
status, maternal
occupation and
education, household
income,

concentrations of
various metals in
urine, blood, and
serum

Beta (95% CI)

Communication:
0.28)

Personal-Social:
0.50)

-0.15 (-0.58,

-0.11 (-0.72,

External Review Draft

3-440

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% CIs

tLinetal. (2013)
Taipei, Taiwan

April 2004-Jan 2005
(enrollment)

Followed through 2 yr

Panel Study

TBPS
n: 230

Singleton full-term
children of non-
smoking mothers
without
occupational
exposure attending
medical center,
hospital, and
clinics in Taipei

Blood

Maternal blood, cord blood; ICP-
MS, measured for Pb, Mn, As,
and Hg.

Pb categories:

Low: <16.45 |jg/L
High: >16.45 pg/L
Mn categories:

Low: <59.59 pg/L
High: >59.59 pg/L

Age at measurement:
delivery

Mean: 13 pg/L, GM: 10.61 pg/L
75th: 16.45 pg/L
Max: 43.22 pg/L

Social and self-help
ability DQs

CDIIT

Age at outcome:
2 yr

Maternal age,
maternal education,
child sex,

environmental tobacco
smoke during
pregnancy and after
delivery, fish intake,
and HOME Inventory
score

Beta
Social

High vs. Low Pb: -5.89
(-10.81, -0.97)c
High Mn x low Pb: 2.83
(-3.442, 9.102)c
Low Mn x high Pb: -2.9
(-9.231, 3.431 )c

High Mn x high Pb:
(-14.144, 0.124)c

-7.01

Self-help

High vs. Low Pb: -1.26
(-5.905, 3.385)c
High Mn x low Pb: 0.49
(-5.429, 6.409)c
Low Mn x high Pb: 0.35
(-5.608, 6.308)c

High Mn x high Pb: -2.38
(-9.103, 4.343)c

tNvanza et al. (2021)

Northern Tanzania
Tanzania

2015-2017
(enrollment)
Followed through
12 mo

Cohort

Mining and Health
Prospective
Longitudinal Study
in Northern
Tanzania

n: 439

Birth cohort of
mother-child pairs
recruited in 2nd
trimester

Maternal dried blood spots; ICP-
MS, measured for Pb, Hg, and
Cd

Age at measurement:
second trimester

Median: 2.72 pg/dL
75th: 4.25 pg/dL
Max: 14.5 pg/dL

Social development
domain using MDAT.
Scores classified as
normal (>90th
percentile on all items in
the domain or <90th
percentile on one or two
items in the domain) or
impaired (<90th
percentile on more than
two items in the
domain).

Age at outcome:
6-12 mo

Maternal age and
education, maternal
and paternal
occupation, number
siblings under 5 yr at
home, and family SES,
infant sex, age, birth
weight, height and
weight as a proxy for
nutritional status
(covariates with
p < 0.20 retained in
the final models)

Prevalence ratio:

Social status development:
1.01 (1.00, 1.02)

External Review Draft

3-441

DRAFT: Do not cite or quote

-------
tDohertv et al.
(2020)

New Hampshire
U.S.

2009 to 2014-2019
Cohort

NHBCS

n: 371 (SRS-2);
318 (BASC-2)

Mother-child pairs

Toenails

Maternal and infant toenails;
Median (maternal prenatal):
0.14 |jg/g (SRS), 0.13 pg/g
(BASC); Median (maternal
postnatal): 0.10 pg/g (SRS),
0.11 pg/g (BASC); Median
(infant): 0.35 pg/g (SRS),
0.37 pg/g

External Review Draft

Composite score
(Social Awareness,
Social Cognition, Social
Communication, Social
Motivation, and
Restricted Interests and
Repetitive Behavior) on
SRS-2.

Adaptive skills
composite on BASC-2;
see also Section 3.5.2.2

Age at outcome:

3 yr old

Maternal age,
maternal BMI, parental
education, maternal
smoking, marital
status, parity, child age
at last breastfeeding,
Healthy Eating Index
score, year of birth,
sex, and age of the
child at testing

Beta per log2-transformed pg/g
increase in toenail Pb

Total SRS-2

Maternal prenatal: -0.08

(-0.20, 0.04)e

Maternal postnatal: 0.03

(-0.08, 0.13)e

Child: -0.06 (-0.19, 0.06)e

Males

Maternal prenatal: -0.06
(-0.23, 0.11 )e

Maternal postnatal: -0.01

(-0.14, 0.13)e

Child: -0.08 (-0.25, 0.10)e

Females

Maternal prenatal: -0.04
(-0.19, 0.11 )e
Maternal postnatal: 0.07
(-0.08, 0.21 )e
Child: -0.05 (-0.21, 0.11)e

Total Adaptive Skills
Maternal prenatal: -0.06
(-0.19, 0.07)e
Maternal postnatal: 0.08
(-0.03, 0.19)e
Child: 0.08 (-0.06, 0.22)e

Males

Maternal prenatal: -0.01
(-0.19, 0.18)e

3-442

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% CIs

Maternal postnatal: 0.08

(-0.07, 0.24)e

Child: 0.10 (-0.13, 0.32)e

tZhou etal. (2017)

Shanghai
China

2010-2012

Followed through 24-
36 mo

Cohort

Females

Maternal prenatal: -0.19
(-0.34, —0.04)e

Maternal postnatal: 0.07

(-0.08, 0.23)e

Child: 0.26 (0.07, 0.45)e

Shanghai Stress
Birth Cohort study

n: 139

Mother-infant pairs
in prenatal clinics
of maternity
hospitals during
mid-to-late
pregnancy

Blood

Maternal whole blood

Age at measurement: 28-36 wk
of gestation

GM (95% CI): 3.30 (3.05, 3.57)
pg/dL

Adaptive and social
behavior domain DQs
from GDS

Age at outcome: 24-
36 mo

Maternal age at
enrollment, SES,
maternal education,
gestational week, child
sex, birth weight and
age

Beta per log—10 transformed
BLL

Adaptive:

Overall: 3.60 (-3.64, 10.83)b
Low stress: 7.57 (-0.12,
15.27)b

High stress: -17.93 (-35.83,
—0.03)b

Social:

Overall: -6.45 (-15.55, 2.65)b
Low stress: -0.07 (-9.57,
9.44)b

High stress: -41.00 (-63.11,
-18.89)b

External Review Draft

3-443

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% CIs

tRuebner et al.
(2019)

46 centers
U.S.

Study Years: NR
Follow-up: 1-16 yr

Cross-sectional

CKiD Cohort study Blood
n: 412

Children (age 1-
16 yr) with mild to
moderate CKD

Child venous blood; ICP-MS. The
BLL measurement closest to the
time of neurocognitive testing
was used for analysis
(concurrent).

Age at measurement:

NR; 2, 4, or 6 years after study

entry

Median: 1.2 pg/dL
75th: 1.8 Mg/dL
Max: 5.1 [jg/dL

Adaptive skills,
composite index on the
BASC-2 (see also 3.5.1
and 3.5.2)

The last available test
results were used to
evaluate long-term
effects. Mean time
between BLL and
neurocognitive testing
was 2.3 yr.

Age at outcome:
3, 5, or 7 yr after study
entry

Child age, sex, race,
poverty, and maternal
education

Adjusted BASC-2 results
were not reported because
they were not statistically
significant.

tViqeh et al. (2014) Birth cohort

Tehran
Iran

October 2006 -
March 2011
Followed through
36 mo

Cohort

n: 174

Mother-infant pairs
recruited in first
trimester (8-
12 wk).

Blood

Maternal blood,
MS

cord blood; ICP-

Age at measurement:

3 trimesters during pregnancy
and delivery

Mean: 1st trimester: 4.15 [jg/dL,
2nd trimester: 3.44, 3rd trimester:
3.78, umbilical cord: 2.86
Max: 1st trimester: 20.5 [jg/dL,
2nd trimester: 7.5, 3rd trimester:
8.0, umbilical cord: 6.9

Mental development
assessed using the
ECDI by Harold Ireton
(language
comprehension,
expressive language,
gross motor, self-help,
social interaction).
Cutoff point scores for
development delay was
score <20% of that
expected for children's
age.

Age at outcome:

36 mo

Maternal educational,
BMI, family income,
gestational age, birth
weight, birth order (first
born)

OR

Total ECDI: 1.74 (1.18, 2.5)

External Review Draft

3-444

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% CIs

tKimetal. (2018b)

4 cities: Seoul,
Anyang, Ansan and
Jeju
Korea

Pregnancy (2011 —
2012) through 24 mo
of age

Cohort

CHECK cohort
n: 140

birth cohort-
pregnant women
recruited from 4
cities in Korea
before delivery,

Blood

Prenatal maternal blood collected
during hospital visit: 2.7 [jg/dL

Cord blood: 1.2 [jg/dL

Adaptive behaviors
assessed using SMS

Association was
examined using multiple
linear regression
analysis.

Age at outcome:
13-24 mo

BPA, and phthalates, Associations of blood Pb

maternal age
(continuous), birth
delivery mode
(categorical), monthly
household income
(categorical), child's
sex, and BDI
(continuous) of the
mother, gestational
age (continuous),
primiparous
(categorical), and
pre-pregnancy BMI
(categorical)

concentrations and SQ were
assessed but not reported
because they lacked statistical
significance.

AAS = atomic absorption spectroscopy; ADHD = attention deficit/hyperactivity disorder; ADOS = Autism Diagnostic Observation Schedule; ASD = autism spectrum disorder;

ASQ = Ages and Stages Questionnaire Inventory; ASSQ = Autism Spectrum Screening Questionnaire; BASC = Behavior Assessment System for Children; BDI = Beck Depression

Inventory; BLL = blood lead level; BMI = body mass index; BPA = bisphenol A; BRS = behavioral rating scale; BSID = Bayley Scales of Infant and Toddler Development;

CARS = Childhood Autism Rating Scale; CDIIT = Comprehensive Developmental Inventory for Infants and Toddlers; CHECK = Children's Health and Environmental Chemicals in

Korea; CHEER = Children's Health and Environment Research; CKiD = Chronic Kidney Disease in Children; DQ = development quotient; DSM = Diagnostic and Statistical Manual of

Mental Disorders; GM = geometric mean; ECDI = Early Child Development Inventory; ELEMENT = Early Life Exposures in Mexico to Environmental Toxicants; GDS = Gesell

Developmental Schedules; GFAAS = graphite furnace atomic absorption spectrometry; HOME = Health Outcomes and Measures of the Environment; ICP-MS = inductively coupled

plasma mass spectrometry; ISAT = Illinois Standard Achievement Test; MAT = Metropolitan Achievement Test; MEAP = Michigan Educational Assessment Program;

MDAT = Malawi Development Assessment Tool; Mn = manganese; mo = month(s); NHANES = National Health and Nutrition Examination Survey; NHBCS = New Hampshire Birth

Cohort Study; NHNPR = Norwegian Patient Registry; NR = not reported; OR = odds ratio; Pb = lead; SD = standard of deviation; SES = socioeconomic status; SGA = small for

gestational age; SMS = Social Maturity Scale; SRS = Social Responsiveness Scale; SQ = social quotient; TBPS = Taiwan Birth Panel Study; wk = week(s); yr = year(s).

aEffect estimates are standardized to a 1 |jg/dL increase in BLL or a 10 |jg/g increase in bone Pb level, unless otherwise noted. For studies that report results corresponding to a
change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the biomarkerand standardized
accordingly.

bResults are unstandardized because the Pb level distribution data was not available.

The CI was calculated from a p-value and the true CI may be wider or narrower than calculated.

dResults are unstandardized because the log base used for exposure transformation was unspecified in the study.

eResults are unstandardized because the biomarker used for Pb exposure measurement is not blood, tooth, or bone.

fStudies published since the 2013 Integrated Science Assessment for Lead.

External Review Draft

3-445

DRAFT: Do not cite or quote

-------
Table 3-14E Epidemiologic studies of exposure to Pb and cognitive function in adults.

Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% CIs

tPower et al. (2014)
Boston, MA, U.S

1993-2008
Cohort

Participants selected from
cohort study (Nurse's Health
Study) and part of case-
control ancillary study

n: 584

Bone, Blood

Bone Pb: K-XRF at the
midtibial shaft and the
patella, blood Pb
concentrations; GFAAS with
Zeeman background
correction in year 1993-
2004

Age at measurement:
registered nurses aged 45-
74 yr

Tibia Pb cone:

10.5 ± 9.7 |jg/g, Patella Pb

cone: 12.6 ±11.7 |jg/g.

Blood Pb cone:

2.9 ± 1.9 [jg/dL

Cognitive decline

Cognitive decline
assessed using a
telephone battery
of cognitive tests
during 2-4 waves
over the period of
follow-up, 1995-
2008. All 9
cognitive scores
were Z-transformed
with high score
representing better
performance.

Alcohol
consumption,
smoking status,
education,
husband's
education,
menopausal
status/hormone
therapy use,
physical activity,
ibuprofen use,
aspirin use, vitamin
E supplementation,
the % of residential
census tract of
white race/ ethnicity,
and median income
of residential
census track.

Beta (95% Cl)a

Tibia

Verbal Memory
-0.002 (-0.006, 0.003)
Overall Cognition
-0.002 (-0.005, 0)
Patella

Verbal Memory
-0.001 (-0.005, 0.002)
Overall Cognition
-0.001 (-0.004, 0.001)
Blood

Verbal Memory
0.003 (-0.021, 0.027)
Overall Cognition
-0.007 (-0.023, 0.009)

External Review Draft

3-446

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% CIs

tFarooqui et al. (2017)

Boston, MA, U.S.

1993-2007

Cohort

Participants selected from
cohort study (Veterans Affairs
NAS

n: 741 subjects in MMSE and
715 in Global cognition

Bone

Patella (trabecular bone)
and tibia (cortical bone)
bone Pb was measured
using KXRF spectroscopy in
1993

Age at measurement:
healthy men aged 51-98 yr

Patella Pb cone:

30.6 ± 19.44 |jg/g, and tibia

Pb cone: 21.6 ± 13.33 |jg/g

Changes in
cognition

Cognition was
assessed using the
MMSE, NES2,
CERAD and WAIS-
R during 3-5 visits
over the period of
15 yr of follow-up.

Age at first cognitive
test, past education
level, baseline
smoking status and
alcohol intake.

Beta (95% Cl)a Pb and MMSE
overtime

Tibia

IQR change in Pb -0.051
(-0.137, 0.035)

IQR change in Pb*time -0.007 (-

0.018, 0.004)

Patella

IQR change in Pb -0.061
(-0.12, -0.002)

IQR change in Pb*time -0.008
(-0.015, 0)

HR (95% Cl)b

Patella 1.095 (0.993, 1.207)
Tibia 1.033 (0.875, 1.22)

Beta (95% Cl)b Pb and Global
Cognition overtime

Patella -0.119 (-0.247, 0.009)
Tibia -0.137 (-0.318, 0.043)

External Review Draft

3-447

DRAFT: Do not cite or quote

-------
tWeuve etal. (2013)

Boston, MA, United
States

2003-2007

Cross-sectional

PD cases confirmed by
movement disorder specialists
using the U.K. Brain Bank
criteria

n: 151 subjects (101 cases
and 50 controls)

Bone

Bone Pb measured using
KXRF spectrometric
estimates of Pb
concentrations in Tibia and
Patella bones.

Age at measurement: cases
and controls (spouses, in-
laws, or friends of the cases)
aged 54-81 yr

Patella Pb cone by age at
cognitive interview
categories:

54-64.9 yr: 5.9 ± 10.3 |jg/g
65-69.9 yr: 9.2 ± 7.8 |jg/g
70-74.9 yr: 7.7 ± 10.5 pg/g
75-80.9 yr: 15.2 ± 10.2 pg/g

Tibia Pb cone by age at
cognitive interview
categories:

54-64.9

yr:

4.4

+

11.1 pg/g

65-69.9

yr:

8.8

+

10.5 pg/g

70-74.9

yr:

6.8

+

8.8 pg/g

75-80.9

yr:

9.2

+

11.5 pg/g

External Review Draft

3-448

Cognition function

Cognitive function
assessed using a
telephone cognitive
assessment battery
of 9 tests based on
a validated
telephone battery
for assessing age-
related cognitive
decline. Added test
of cognitive
domains that
typically decline in
PD. All 9 cognitive
scores were z-
transformed with
high score
representing better
performance.

Age at cognitive
assessment, sex,
race, education,
smoking history

Adjusted difference (95% Cl)b

Patella

Telephone interview for
cognitive assessment (TICS)
-0.08 (-0.32 to 0.15)

Delayed 10-word recall
0.05 (-0.18 to 0.28)

Delayed 10-word recognition
0.01 (-0.22 to 0.24)

Animal naming
-0.11 (-0.32 to 0.10)
"F" naming
-0.07 (-0.30 to 0.17)

Digit span forward
-0.02 (-0.27 to 0.22)

Digit span backward
0.05 (-0.17 to 0.27)

Oral trails B minus A
0.03 (-0.23 to 0.28)

Global score
-0.01 (-0.14 to 0.13)

Tibia

Telephone interview for
cognitive status (TICS)
-0.20 (-0.40 to -0.00)

Delayed 10-word recall
-0.04 (-0.23 to 0.16)

Delayed 10-word recognition
-0.01 (-0.21 to 0.20)

Animal naming
-0.11 (-0.29 to 0.07)
"F" naming
-0.19 (-0.39 to 0.01)

Digit span forward
-0.23 (-0.43 to -0.03)

Digit span backward
-0.19 (-0.37 to -0.00)

Oral trails B minus A
-0.06 (-0.29 to 0.17)

Global score

-0.13 (-0.25 to -0.01)

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% CIs

tSkerfvinq et al. (2015) n: 927

Landskrona and
Trelleborg, Southern
Sweden

1978-2007 followed for
4-12 yr

Cohort

Blood

Between 1978 and 1994, B-
Pb levels were determined
using flame or

electrothermal atomization
atomic absorption
spectrometry; between 1995
and 2007, B-Pb levels were
determined using inductively
coupled plasma mass
spectrometry

Age at measurement: 7-12
years

Mean: 34 |jg/L

IQ assessed for
military
conscription

IQ (measured
logical, verbal,
spatial abilities, and
technical
understanding)
assessed as a part
of military
conscription
examinations.

Age at outcome:
18-19 years

Age at blood
sampling, sex,
parents' education,
family economy,
and country of birth
of child and parents

Beta (SE)a
IQ

All subjects

-0.127 (-0.209, -0.045)
Blood Pb <50 |jg/L
-0.204 (-0.392, -0.016)

tReuben et al. (2017) Dunedin Multidisciplinary
Health and Development
Dunedin, New Zealand Study

1972/73-2012

Cohort

n: 565

Blood

Graphite fumance atomic

absorption

spectrophotometry

Age at measurement: 11 yr

Mean (SD):

10.99 ±4.63 [jg/dL

Full -scale IQ
(other domains
such as verbal
comprehension,
perpetual

reasoning, working
memory,

processing speed)

Cognitive function
assessed using
Wechsler Adult
Intelligence Scale -
IV (WAIS-IV) at the
age of 38 yr.

Childhood IQ scores
(age 7 and 9 yr),
their mothers' IQ
score, and their
socioeconomic
background

Change in IQa (95% CI)
Adjusted by sex
-0.394 (-0.669, -0.119)
Fully adjusted
-0.322 (-0.496, -0.148)
Change in perceptual
reasoning3 (95% C)
-0.414 (-0.627, -0.201)
Change in working memory3
(95% CI)

-0.252 (-0.476, -0.028)
Change in socioeconomic
status3 (95% CI)
-0.358 (-0.635, -0.081)

External Review Draft

3-449

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% CIs

tReuben et al. (2020) Dunedin Multidisciplinary
Health and Development
Dunedin, New Zealand Study

1972/73-2019

Cohort

n: 564

Blood

Furnace atomic absorption
spectrophotometry

Age at measurement: 11 yr

Mean (SD):

10.99 ±4.63 [jg/dL

Full-scale IQ and

self-reported

information)

Cognitive
performance
assessed
objectively using
Wechsler Adult
Intelligence Scale -
IV (WAIS-IV) and
subjectively via
informant and self-
reports at the age
of 45 yr.

Childhood IQ scores
(age 7 and 9 yr),
their mothers' IQ
score, and their
socioeconomic
background

Change in IQa(95% CI)
-0.414 (95% CI: -0.679,
-0.149)

Residualized Change in IQa
(95% CI)

-0.394 (-0.583, -0.205)

tKhalil et al. (2014)

Multicity (6 clinical
sites), U.S

May 2007 to Nov 2008
Cross-sectional

Population-based cohort study Blood
(MrOS)

n: 445	Venous blood samples

tested for Pb levels using
AAS

Age at measurement: non-
Hispanic Caucasian men
aged >65 yr

Mean (SD)

2.25 ± 1.20 [jg/dL

Cognitive function
assessed using 3
MS

Age at outcome:

>65 yr

Age, education,
smoking, alcohol
consumption and
BMI

Beta (95% Cl)b

Cognitive Function in Adults

3MS

-0.01 (-1.10,1.07)

Cognitive Function in Adults
Trail Making B
2.72 (-7.65, 13.09)

External Review Draft

3-450

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% CIs

tSouza-Talarico et al.
(2017)

Sao Paulo City,

Brazil

Cross-sectional

n: 125 (104 women and 21
men)

Blood

Venous blood samples
tested for heavy metals (Cd
and Pb) levels using ICP-MS

Age at measurement:
Healthy older adults
between 50 and 82 yr
(M = 65.9)

Mean (SD)

2.1 ± 0.970 [jg/dL

MMSE and
Informant
Questionnaire on
Cognitive Decline
used to rule out
cognitive and
functional
impairments.

Age at outcome: 50
and 82 yr
(M = 65.9)

Age, sex, income,
education,
hemoglobin,
hematocrit

Betab

WMC Pb 0.106 (AR: 0.057)

BCd x BPb interaction-term and
WMC - 0.378(p< 0.001)

Table 3 adjusted EE for Pb and
WMC, with and without
controlling for Oxygen Radical
Absorbance Capacity total:
standardized

tvan Wiinaaarden et al.
(2011)

Nationwide, U.S.

1999-2008

Cross-sectional

NHANES 1999-2008 (for self-
reported confusion and
memory problems) and
NHANES 1999-2002 (for
DSST)

n: 9526 participants (7277
from NHANES 1999-2008
and 2299 participants from
NHANES 1999-2002)

Blood

Venous blood samples
tested for Pb concentration
using AAS with Zeeman
background correction.

Age at measurement: >60 yr

Blood Pb cone: 2.46 [jg/dL
(range 0.18-54.00 pg/dL)

Cognitive function

Cognitive function
assessed by self-
reported responses
on limitation in
cognitive
functioning, and
DSST (a subset of
the WAIS-III) for
subset of
participants.

Age, sex, ethnicity,
education level,
PIR, self-reported
general health
status

OR (95% Cl)b
1.01 (0.65, 1.56)

Age at outcome:

>60 yr

External Review Draft

3-451

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% CIs

tPrzvbyla etal. (2017)
Nationwide, U.S.

1999-2002

Cross-sectional

NHANES cycles 1999-2000
and 2001 -2002;

n: 498

Blood

Blood samples tested for
chemicals (Pb, Cd and
PCBs) concentrations; Pb
and Cd measured using
ICP-MS.

Age at measurement: 60-
84 yr

Mean 2.17 |jg/dL (95% CI:
2.07, 2.27)

Cognitive function Race/ethnicity, age,

education level,
Cognitive function PIR, sex and
assessed using the smoking status
DSC Module ofthe
WAIS-III.

Age at outcome:
60-84 yr

Betas per natural log increase in
BLL

Cognitive Functioning

All Participants: -0.10 (-0.20,

-0.01)

Females:

-0.12 (-0.26, 0.01)

Males:

-0.09 (-0.24, 0.06)

Age 60-69:

-0.13 (-0.28, 0.01)

Age 70-74:

-0.08 (-0.2, 0.04)

External Review Draft

3-452

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% CIs

tSasaki and Carpenter NHANES cycles 2011-12 and Blood and Urine

(2022)

Nationwide, U.S.

2011-2014

Cross-sectional

2013-14 and tested for
different sets of chemicals for
different subgroups
n: 3042

Venous blood samples and
urine samples tested for
seven metals and metalloids
(including Pb) using ICP-MS

Age at measurement: 60-
80 yr

Blood mean Pb: 19.0 |jg/L
Urine mean Pb: 0.72 [jg/dL

Cognitive function	Age, sex, ethnicity,

education level,

Immediate and	depression,

delayed memory	diabetes, alcohol

assessed using the	consumption, and

CERAD, and	smoking
working memory
assessed using the
DSST.

Age at outcome:
60-80 yr

Beta (95% Cl)b

Blood

CERAD Immediate recall:
-0.58 (-0.91, -0.24)

CERAD Delayed recall:
-0.19 (-0.35, -0.02)

Digit symbol substitution:
-1.08 (-2.12, -0.05)

CERAD immediate recall as a
function of age
60s Years Old Group:
-0.37 (-0.87, 0.13)

>70 Years Old Group:
-0.85 (-1.44. -0.27)

Urine

CERAD Immediate recall:
-0.26 (-0.58, 0.06)

CERAD Delayed recall:
-0.03 (-0.19, 0.13)

Digit symbol substitution:
-1.03 (-2.01, -0.06)

tXiao et al. (2021)

Guangxi, southern
China

Aug 2016-July 2018

Cross-sectional

n: 2879

Blood

Venous blood samples
tested for 22 metals
(including Pb) using ICP-MS.

Age at measurement: >60 yr
Blood Pb: Median: 51.5 |jg/L

Cognitive function Age, gender,

Cognitive function
assessed using the
MMSE.

Age at outcome:
>60 yr

education
attainment, annual
income, BMI,
smoking, alcohol
drinking, insomnia,
and physical activity

Beta (95% Cl)a
Cognitive function
Single-pollutant model
-0.018 (-0.06, 0.023)
Multi-pollutant model
-0.019 (-0.063, 0.025)

External Review Draft

3-453

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95% CIs

tMeramat et al. (2017) Neuroprotective Model for
Healthy Longevity among

Malaysia

May 2013 to January
2014

Cross-sectional

Malaysia Older adult
n: 317

Nail

Toenails (clipped from all
toes) assessed for trace
elements (Al, Ca, Cd, Co,
Fe, Pb, Zn, Se, Cu and Cr)
using ICP-MS.

Age at measurement: >60 yr

Pb cone: Cognitive impaired
group (n = 197): 0.55
± 0.03 |jg/g; and Normal
cognitive group (n = 120):
0.35 ± 0.013 |jg/g

Cognitive
impairment
assessed using
Montreal Cognitive
Assessment - a
Malay version

Age at outcome:
>60 yr

Age, sex, years of
education and
smoking habits

OR (95% Cl)a

Cognitive impairment 2.471

(1.535-3.980)

tYu et al. (2021)
Nationwide, U.S.

Jan 2015-Sep 2017

Cohort

SPHERL longitudinal study Blood

n: 260 (260: DSST cohort and Venous blood samples

Cognitive function

Cognitive function

168: SCWT cohort) with
baseline and annual follow-up
blood Pb measurements and
neurocognitive function
assessments.

tested for Pb concentration changes assessed

using ICP-MS.

Age at measurement: mean
age 29.4 years

Blood Pb cone: DSST
cohort: Geo mean: 3.97 (5-
95th percentage interval (PI)
0.90-14.3) [jg/dL at
baseline, 13.4 (PI 3.70-30.3)
[jg/dL and 12.8 (PI 2.80-
29.2) [jg/dL at the first and
second follow-up visits,
respectively.

using the DSST
and ST at baseline
and annual follow-
up visits.

Age at outcome:
mean age 29.4
years

Age, sex, ethnicity,
change in age,
baseline BMI,
changes in body
weight, education,
baseline blood Pb,
baseline
neurocognitive
function test,
baseline values and
changes in smoking
status, total/HDL
ratio, cholesterol
and alcohol
consumption

OR (95% Cl)a
DSST

1.012 (0.997, 1.028)

External Review Draft

3-454

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population

Exposure Assessment

Outcome

Confounders Effect Estimates and 95% CIs

3MS = Modified Mini Mental State Examination; BrainAGE = Brain Age Gap Estimation; CERAD = Consortium to Establish a Registry for Alzheimer's Disease; DSC = Digital Symbol
Coding; DSST = Digit Symbol Substitution Test; EE = effect estimate(s); KXRF = K-Shell X-Ray fluorescence; MMSE = mini mental status exam; MrOS = Osteoporotic Fractures in Men
Study; NAS = Normative Aging Study; NES2 = Neurobehavioral Evaluation System 2; PD = Parkinson's disease; WAIS-III = Wechsler Adult Intelligence Scale, Third Edition; WAIS-
R = Wechsler Adult Intelligence Scale-Revised; WMC = working memory capacity; SPHERL = Study for Promotion of Health in Recycling Lead; SCWT = Stroop Color-Word Test.

aEffect estimates are standardized to a 1 |jg/dL increase in BLL or a 10 |jg/g increase in bone Pb level, unless otherwise noted. For studies that report results corresponding to
a change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the biomarker and standardized
accordingly.

bResult not standardized because data pertaining to the BLL distribution and/or base for the log-transformation were not reported
fStudies published since the 2013 Integrated Science Assessment for Lead.

1

External Review Draft

3-455

DRAFT: Do not cite or quote

-------
Table 3-15E Epidemiologic studies of Pb exposure and psychopathological effects in adults.

Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
CIs

Raian et al. (2007)
Boston, MA, U.S.
1991-2002
Cohort

Veterans Affairs
NAS
n: 1,075

Closed cohort of
male volunteers with
no chronic medical
conditions at entry.
97% white

Bone

Bone Pb measured in the mid-
tibia shaft and patella using
KXRF

Age at measurement: 21-80
years

Mean: -67.5 yr old

Mean (SD):

Tibia: 22.1 (13.8) pg/g

Patella: 31.4 (19.6) pg/g

Depression and anxiety

Depressive and anxiety
symptoms were measured using
the BRIEF Symptom Inventory
(depression and anxiety were
determined to be present for
participants that scored 1 SD
above the mean for a normal
population). Participant followed
up was 3 years.

Age, alcohol
consumption,
education, time
between

assessments, and
cumulative smoking

Anxiety OR (95% cl)a
Tibia: 1.13 (0.99, 1.29)
Patella: 1.09 (0.99, 1.19)

Depression OR (95% Cl)a
Tibia: 1.11 (0.98, 1.38)
Patella: 1.05 (0.96, 1.16)

Bouchard et al.
(2009)

U.S.

1999-2004
Cross-sectional

NHANES
n: 1,987

Blood

Blood Pb measured in venous
whole blood samples using
ICP-MS

Age at measurement:
20-39 yr old

Geo. mean: 1.24 [jg/dL
20th %ile: 0.7 pg/dL
40th %ile: 1.0 pg/dL
60th %ile: 1.4 pg/dL
80th %ile: 2.1 pg/dL

Depression

WHO CIDI was administered.
Major depressive disorder
diagnosed according to DSM-IV

criteria.

Age at outcome: 20-39 yr

Age, sex,
race/ethnicity,
education, and PIR

Major Depressive Disorder

OR (95% Clf

Q1
Q2
Q3
Q4
Q5

Ref.

1.39 (0.71, 2.72)
1.28 (0.69, 2.38)
1.41 (0.76, 2.6)
2.32 (1.13, 4.75)

External Review Draft

3-456

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
CIs

tPeters et al. (2011)

Boston, MA
U.S

1991-1997(Bone Pb

measurements);

1993-2003

(Psychological

measurements)

Cohort

Veterans Affairs

NAS

n: 412

Closed cohort of
male volunteers with
no chronic medical
conditions at entry.
97% white

Bone

Bone Pb measured in the mid-
tibia shaft using KXRF
Age at measurement:

Mean: -65.3 yr old

Mean: 20.6 |jg/g

Pessimism and Depression

A subscale of the Life
Orientation Test was used to
assess pessimistic attitudes.
Depressive symptoms were
measured using the BRIEF
Symptom Inventory (depression
was determined to be present
for participants that scored 1 SD
above the mean for a normal
population).

Age at outcome: Mean -68.3 yr

Age, health
behaviors,
childhood
adult SES

Difference in Pessimism
Level on the Life Orientation
Test (95% CI)

0.21 (0.00, 0.43)

tReuben et al.
(2019)

Dunedin, New
Zealand

Enrollment: 1972-73;
Follow-up through
2012

Cohort

Dunedin
Multidisciplinary
Health and
Development Study

Cohort of children
3 yr old at
enrollment followed
through 32 yr of age.
Study population
was nationally
representative
(majority white) and
had high rates of
participation and
follow-up.

Blood

Blood Pb measured in venous
blood samples using GFAAS
Age at measurement: 11 yr

Mean: 11.08 pg/dL
(94% above 5 |jg/dL)

General Psychopathology,
Externalizing Symptoms,
Internalizing Symptoms, and
Thought Disorder Symptoms in
Adults

Psychopathology symptoms
were assessed using the
Diagnostic Interview Schedule.
Factor loadings from each of 11
disorders were used to create
hierarchical measures for
psychopathology and each of its
constituent psychiatric spectra

Sex, childhood SES,
maternal IQ, and
family history of
mental illness.

Change in symptom scores
(95%CI) (standardized to a
mean [SD] of 100 [15])a

General Psychopathology
0.27 (0.02, 0.51)

Externalizing Symptoms
0.15 (-0.10, 0.40)
Internalizing Symptoms
0.28 (0.04, 0.53)

Thought Disorder
0.26 (0.01, 0.51)

Age at outcome:
and 38 yr

18, 21, 26, 32,

External Review Draft

3-457

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
CIs

tMcFarlane et al.
(2013)

Port Pirie, Australia

1979-1982
(enrollment); 2008-
2009 (follow-up)

Cohort

Port Pirie cohort
Study
n: 210

Mother-singleton
infant pairs enrolled
in Pb-smelting town
from 1979-1982.
Assessed
periodically from
birth to 7 yr, again
from 11 to 13 yr, and
for this study, at 25
to 29 yr

Blood

Blood Pb measured in capillary
blood samples using GFAAS
Age at measurement:
6, 15, and 24 mo; 3-7 yr

Mean: 17.2 [jg/dL
(birth to 7-yr average)

Drug and alcohol abuse,

DSM-IV Disorders (Alcohol
abuse, drug abuse, social
phobia, specific phobia, PTSD,
alcohol dependence, panic
attack, major depressive
disorder) and adult self-report
DMV-IV oriented subscale
(anxiety, somatic problems,
depressive problems,
hyperactivity, inattention,
antisocial personality problems,
avoidant personality problems)

Age at outcome:

25 to 29 years

HOME, maternal
education, paternal
occupation,
mothers' age at
birth, breastfeeding,
and single parent
family status

OR (95% Cl)a

Social Phobia

Women: 1.05 (0.93, 1.188)

Men: 0.96 (0.80, 1.15)

Specific Phobia
Women: 1.13 (0.99, 1.29)
Men: 1.02 (0.71, 1.47)

Major Depressive Disorder
Women: 0.89 (0.77, 1.03)
Men: 0.89 (0.68, 1.16)

External Review Draft

3-458

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
CIs

tLi etal. (2017)

Shanghai (inner and
outer districts)

China

2010

Cross-sectional

n: 1,701

Stratified cluster
sampling of
pregnant women
(gestational wk 28-
36)

Blood

Blood Pb measured in venous
blood samples using GFAAS
Geo mean: 3.97 [jg/dL
Max: 14.84 pg/dL

Maternal stress

Life Event Stress Scale for
Pregnant Women, Symptom
Checklist-9-Revised (GSI
[measure of psychological
distress], anxiety and
depression scores)

Age at outcome:

13-42 yr old (wk 28-36)

Maternal age at
enrollment,
ethnicity, maternal
education, and
family monthly
income, years

residing in Shanghai Maternai stress

Change in maternal stress,
anxiety, and depression
scores per 10-fold increase in
BLLs (results from piecewise
linear models)3

<2.57 [jg/dL: 0.22 (0.05, 0.4)
>2.57 [jg/dL: -0.07 (-0.16,
0.01)

Depression

<2.57 [jg/dL: 0.34 (0.12,
0.56)

>2.57 [jg/dL: -0.09 (-0.19,
0.02)

Anxiety

<2.57 [jg/dL: 0.25 (0.04,
0.46)

>2.57 [jg/dL: -0.08 (-0.18,
0.02)

tlshitsuka et al.
(2020)

Japan

2011-2014

Cross-sectional

Japan Environment Blood
and Children's Study
n: 17,267

Pregnant women
recruited out of 15
regional centers
across Japan

Blood Pb measured in whole
blood samples using ICP-MS
Age at measurement:
31 yr (mean)

Geo. mean: 0.58 [jg/dL
Max: 6.75 [jg/dL

Maternal Depression

K6. Depression measured as
scores >5 or 13 (two cutoff
points for sensitivity).

Age at outcome:

Mean age: 31 yr

Age, parity, marital
status, education,
employment status,
household income,
and smoking and
alcohol status

OR (95% Cl)b

K6 >13

1.00 (0.76, 1.32)
K6 >5

0.98 (0.88, 1.09)

External Review Draft

3-459

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
CIs

tBerketal. (2014)

NHANES
n: 15,140

Blood

Depression

Age, sex, poverty,
family income,

Depression OR (95% CI)

U.S.



Blood Pb measured in venous

Depression measured as >9 on

ethnicity, and

Q4 vs. Q1*:



General population,

whole blood samples using

the nine-item depression

country of birth

0.98 (0.78, 1.25)

2005-2010

>18 yr old

ICP-MS

module of the Patient Health



Age at measurement:

Questionnaire









>18 yr old
Mean: NR





*Quartile levels NR

Cross-sectional



Age at outcome:
>18 yr old





tNauven et al. (2022) KNHANES
n: 16,371

South Korea

2009-2013 and
2016-2017

Cross-Sectional

General population;
mean age: 42.6 yr
old (SD: 18.12)

Blood

Blood Pb was measured in
venous whole blood using
GFAAS

Age at measurement (mean):
42.6 yr old (SD: 18.12)

Geo. Mean:

1.84	[jg/dL (w/o depression)

1.85	[jg/dL (w/ depression)

Depression

Self-reported physician's
diagnosis or treatment for
depression

Age at outcome (mean): 42.6 yr
old (SD: 18.12)

Sex, urbanicity,
household income,
physical activity,
occupation, BMI,
alcohol
consumption,
education level, and
smoking status

OR (95% Cl)a
1.02 (0.90, 1.16)

External Review Draft

3-460

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
CIs

tEumetal. (2012)

Boston, MA
United States

Subsample 1: Bone
Pb Measure 1993-
1995;

Subsample 2: Bone
Pb Measure 2001-
2004.

Psychological
Questionnaires:
1988, 19992,1996,
2000, 2004

Cohort

Nurses' Health
Study
n: 617

Women from two
subsample studies
of the NHS cohort

Bone

Midtibial shaft and patella bone
Pb measured using KXRF
Age at measurement:

Mean: 60.9 yr

Mean:

Tibia: 10.3 |jg/g;

Patella: 12.5 |jg/g

Tibia Tertiles:

T1: <7.0 |jg/g
T2: 7.0-11.5 pg/g
T3: >11.5 pg/g

Patella Tertiles:
T1
T2
T3

<8.5 pg/g
8.5-14.5 pg/g
>14.5 pg/g

Phobic anxiety and depressive
symptoms

Depression symptoms
measured using MHI-5; Anxiety
symptoms measured using
phobic anxiety scale of the
Crown-Crisp Experiential Index
(CCEI)

Age at outcome:

Mean:

MHI-5: 59.4 yr
CCEI: 59.2 yr

Substudy group,
age at bone Pb
measure, age at
MHI-5 or CCEI
measurement,
education,
husband's
education, alcohol
consumption, pack-
years of smoking,
and employment
status at MHI-5 or
CCEI assessment

OR (95% Cl)b (Tertile 3 vs
Tertile 1)

CCEI >4
All women

Tibia: 1.10 (0.73, 1.64)
Patella: 0.75 (0.49, 1.15)

Women on HRT
Tibia: 2.79 (1.02, 7.59)
Patella: 0.23 (0.07, 0.69)

MHI-5 Point Difference
(lower scores indicate worse
symptoms)

All women (T3 vsT1)

Tibia: -1.06 (-3.05, 0.94)

Patella: -7.78 (-11.73,
-3.83)

Women on HRT (T3 vsT1)
Tibia: 0.61 (-1.55, 2.78)
Patella: 0.51 (-3.91, 4.94)

External Review Draft

3-461

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and 95%
CIs

tFan et al. (2020)

Cohort Study of
Elderly Health and

Blood

Depressive symptoms

Age, gender, region,
marital status,

OR (95% Cl)b

Luan city, Anhui

Environmental

Blood Pb measured in venous

Chinese revision of the geriatric

monthly income,

Depression

Q1: Ref.

province, China

Controllable Factors

whole blood samples using

depression scale

education level,



n: 994

ICP-MS



alcohol intake,

2016



Age at measurement:

Age at outcome:

smoking, and BMI

Q2: 1.28 (0.79, 2.08)

Older adults (>60 yr

>60 yr old

>60 yr old



Q3: 1.36 (0.84, 2.22)

Cross-sectional

old) selected using
cluster sampling

Quartiles





Q4: 2.03 (1.23, 3.35)



from two









communities in

Q1: <2.03 [jg/dL









Luan, China

Q2: 2.03-2.68 pg/dL
Q3: 2.68-3.06 pg/dL
Q4: >3.06 pg/dL







tMaetal. (2019)

n: 190 (95 cases, 95
controls)

Blood

Schizophrenia

Marital status
(others not

OR (95% Cl)b per 1 ng/mL
increase

Hebei Province



Serum Pb measured in venous

Physician-diagnosed

specified).

3.15 (1.24, 7.99)

China

First-episode drug-

blood samples using ICP-MS

schizophrenia using ICD-10

Population matched



naive patients ages

Age at measurement:

criteria

on age and sex



2018-2019

18 to 60 yr old were
recruited from a

18-60 yr old

Age at outcome:





Case-control

psychiatric hospital.

Median: 0.61 ng/mL (serum)

18-60 yr old





Age and sex-
matched controls
without known
psychiatric problems

75th: 0.79 ng/mL (serum)







recruited from an
affiliated hospital

AAS = atomic absorption spectrometry; BLL = blood lead level; BMI = body mass index; CCEI = Crown-Crisp Experiential Index; CI = confidence interval; CIDI = Composite
International Diagnostic Interview; GFAAS = graphite furnace atomic absorption spectrometry; K6 = Kessler Psychological Distress Scale; KXRF = K-Shell X-Ray fluorescence;
MHI-5 = Mental Health Index 5-item; NAS = Normative Aging Study; NR = not reported; Pb = lead; PIR = poverty-income ratio; PTSD = post-traumatic stress disorder; Q = quartile;
SD = standard deviation; SES = socioeconomic status; WHO = World Health Organization; wk = week(s); yr = year(s).

aEffect estimates are standardized to a 1 |jg/dL increase in BLL or a 10 |jg/g increase in bone Pb level, unless otherwise noted. For studies that report results corresponding to a
change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the biomarker and standardized accordingly.
bResult not standardized because data pertaining to the BLL distribution and/or base for the log-transformation were not reported
fStudies published since the 2013 Integrated Science Assessment for Lead.

External Review Draft

3-462

DRAFT: Do not cite or quote

-------
Table 3-16E Epidemiologic studies of Pb exposure and sensory organ function in adults.

Reference and Study
Design

Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

Park etal. (2010)

Eastern Massachusetts
U.S.

Enrollment and outcome
assessment: 1962-1996;
bone Pb measurements:
1991-1996

Cohort

NAS
n: 448

Bone

Bone Pb levels measured
in the midtibial shaft and
patella with a KXRF
instrument.

Age at measurement:
Mean (SD) at bone Pb
measurement = 64.9

(7.3)	yr; mean (SD) at first
audiometric test = 42.5

(8.4)	yr

Mean (SD) in tibia = 22.5
(14.2) |jg/g; mean (SD) in
patella = 32.5 (20.4) pg/g

Sensory Organ Function

Pure-tone averages
assessed by audiologists
with the modified Hughson-
Westlake procedure. Air
conduction hearing
thresholds measured for
each ear by audiologists
using either a Beltone 15C
or a Grason-Stadler 1701
audiometer.

Cross-sectional
analyses and logistic
regression analyses
adjusted for age,
race, education, BMI,
pack-years of
cigarettes, diabetes,
hypertension,
occupational noise,
and noise notch.

Hearing loss OR (95%
Cl)b

Tibia 1.19 (0.92, 1.53)

Patella 1.48 (1.14, 1.91)

EE in Hearing thresholds
(dB HL) with one
interquartile range
Increment in bone lead
measure

Tibia PTA 0.83 (-0.18,
1.83)

Patella PTA 1.58 0.62,
2.55

tShiue (2013)
U.S.

2003-2004
Cross-sectional

NHANES

n: 712 (vision); 732
(hearing); 669
(balance)

NHANES age 50 and
above

Urine

Urinary Pb was detected
by mass spectrometry
Age at measurement:
50 yr

Not Reported

Vision: excellent, good, and
fair eyesight (self-reported)
were classified as good;
poor and very poor were
classified as poor

Hearing: good and little
trouble hearing (self-
reported) were classified as
good; lots of trouble and
deaf were classified as poor

Balance: "During the past
12 mo, have you had
dizziness, difficulty with
balance, or difficulty with

Age, sex, ethnicity,
urine creatinine,
survey weighting

OR (95% Cl)b

Vision 1.15 (0.67-1.97)

Hearing 0.97 (0.63-1.51)
Balance 0 68 (0.51-0.91)

External Review Draft

3-463

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

failing?"

Ear ringing: "ears ringing,
roaring, or buzzing in the last
year"

Age at outcome:

50 yr

tKanq et al. (2018)
Korea
2010-2013
Cross-sectional

KNHANES
n: 6409

Representative sample
of the entire Korean
population. Study
participants were at
least 20 yr old and
underwent pure-tone
audiometry and blood
Pb test.

Blood

Blood Pb was measured
using GFAAS and
classified into quartiles by
sex

Age at measurement:
20-87 yr (mean ± SE:
47.1 ± 0.3 yr)

Weighted mean ± SE
(Men): Q1 = 1.56
± 0.01 |jg/dL;
Q2 = 2.22 ±0.01 [jg/dL
Q3 = 2.82 ±0.01 [jg/dL
Q4 = 4.22 ±0.08 pg/dL
Weighted mean ± SE
(Women):

Q1 = 1.12 ± 0.01 pg/dL
Q2 = 1.61 ± 0.01 pg/dL
Q3 = 2.11 ± 0.01 pg/dL
Q4 = 3.03 ±0.03 pg/dL

Low-frequency hearing
impairment;

High-frequency hearing
impairment

Pure-tone audiometry was
performed on both ears at
0.5, 1, 2, 3, 4, and 6 kHz. A
binaural pure-tone average
threshold was used and two
binaural averages were
computed, one across 0.5, 1,
and 2 kHz and the other
across 3, 4, and 6 kHz to
determine the low- and high-
frequency thresholds.
Hearing impairment was
then determined according
to whether an average
threshold exceeded 25 dB in
the respective frequency
band.

Age at outcome:

20-87 yr (mean ± SE:
47.1 ± 0.3 yr)

Age, BMI, education,
smoking, alcohol
consumption,
exercise, diabetes
mellitus,

hypertension, noise
exposure

OR (95% Cl)b

Hearing Loss - Low
Frequency
Females
Q2
Q3
Q4

1.271 (0.726, 2.224)
1.308 (0.784, 2.183)
0.932 (0.541, 1.605)

Males

Q4
Q3
Q2

1.026 (0.813, 1.295)
1.028 (0.661, 1.598)
1.17 (0.772, 1.773)

Hearing Loss - High

Frequency

Females

Q2: 0.947 (0.608, 1.475)

Q3: 1.013 (0.698, 1.471)
Q4: 1.502 (1.027, 2.196)

Males

Q2: 1.368 (1.006, 1.86)
Q3: 1.402 (1.005, 1.955)

External Review Draft

3-464

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

Q4:

1.629 (1.161, 2.286)

tChoi and Park (2017)

Korea National Health and
Nutrition Examination Survey
(KNHANES), Korea

2010-2012

Cross-sectional

KNHANES
n: 5187 adults

Blood

Measured using Graphite
furnace atomic absorption
spectrometry

Age at Measurement: 20-
87 yrs

90th: Adults: Geometric
mean (age-adjusted) 2.12
[jg/dL (95% CI: 2.08, 2.15)
Adolescents: Geometric
mean (age-adjusted) 1.26
[jg/dL (95% CI: 1.22, 1.30)

Hearing loss (>25dB) at
speech frequency;

Hearing loss (>25dB) at high
frequency

Pure-tone air conduction
hearing thresholds were
obtained for each ear at
frequencies of 0.5, 1, 2, 3, 4,
and 6 kHz over an intensity
range of-10 to 110 dB
-10 to 110 dB.

Age at outcome: 20-87 yrs

Adjusted for age;
age squared; sex;
education; BMI;
current cigarette
smoking; current
diagnosis of
hypertension and
diabetes; and
occupational,
recreational, and
firearm noise
exposures

OR (95% Cl)b

Hearing Loss (>25 dB)
High-frequency PTA

Pb Quartile 2 (1.594-
2.146):

1.13 (0.83, 1.53)
Pb Quartile 3 (2.48-
2.822):

1.35 (1, 1.81)

Pb Quartile 4 (2.823-

26.507):

1.7 (1.25, 2.31)

Per doubling of Pb:

1.3 (1.08, 1.57)

Speech-Frequency PTA

Pb Quartile 2:

0.94 (0.65, 1.35)

Pb Quartile 3:

1.29 (0.92, 1.78)

Pb Quartile 4:

1.25 (0.87, 1.79)

Per doubling of Pb:

1.15 (0.94, 1.41)

tWana et al. (2020)

Zhejiang Province
(Hangzhou, Jiangshan,
Tonglu, Jiaxing, Anji,
Jinyun), China

n: 2016

Blood

Measured by graphite
furnace atomic absorption
spectrometry

Hearing loss

The devices utilized in this
research were an
audiometer (AT235,
Interacoustics AS, Assens,
Denmark) and standard
headphones (TDH-39,

Income, education,
hypertension,
diabetes,

hyperlipidemia, otitis
media, migraine,
anemia, smoking,
alcohol consumption,
daily fruit and

OR (95% Cl)b Q1 Ref
Q2 1.135 (0.806, 1.599)
Q3 1.038 (0.731, 1.475)
Q4 1.016 (0.7, 1.475)

External Review Draft

3-465

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

2016 to 2018
Case-control

Age at Measurement:
21-89 years

Logarithmic-transformed
levels of Pb

Case group (1.58±0.17
[jg/cSL) and control group
(1.57±0.16 fjg/dL)

Telephonies Corporation,
Farmingdale, USA)

Age at outcome: 21-89
years

vegetable intake,
and workplace noise
exposure

tChoi etal. (2012)

NHANES, U.S

1999-2004

Cross-sectional

NHANES
n: 3698

Blood

Simultaneous
multielement atomic
absorption spectrometer
(SIMAA 6000;

PerkinElmer, Norwalk, CT)
with Zeeman background
correction

Age at Measurement:
20-69 years

Age-adjusted geometric
mean (95% CI) = 1.54
[jg/dL (1.49, 1.60)

Hearing threshold;

Hearing loss

Pure-tone air conduction
hearing thresholds were
obtained for both ears at
frequencies of 0.5-8 kHz
over an intensity range of-
10 to 120 dB.

Age at outcome: 20-69
years

Age and age2, sex,
race/ethnicity [non-
Hispanic white
(reference), Mexican
American, non-
Hispanic black,
other], education [<
high school
(reference), high
school, > high
school], BMI
(continuous),
ototoxic medication
use (yes/no),
cigarette smoking
[never smoker
(reference), < 20
pack-years, > 20
pack-years],
hypertension
(yes/no), type 2
diabetes (yes/no),
and either blood lead
or blood cadmium
(for the

corresponding
cadmium or lead
model), occupational
noise exposure
(0*NET score,

OR (95% Cl)b
Hearing Loss:

Quintile 2 (0.90-1.30
Mg/dL)

1.08	(0.55, 2.12)

Quintile 3 (1.4-1.8 pg/dL)
1.1 (0.58, 2.05)

Quintile 4 (1.90-2.70
MQ/dL)

1.21 (0.67, 2.22)

Quintile 5 (2.80-54
MQ/dL)

1.36 (0.75, 2.48)
Per doubling of Pb

1.09	(0.95, 1.26)

External Review Draft

3-466

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

continuous),
nonoccupational
firearm noise
(yes/no) and any
recreational noise
(yes/no)

tYin etal. (2021)

Iran, Korea, China, United

States

Other

n: 234-7596 in :
studies

Blood

Age at measurement:
3-87 years

Hearing loss

All studies included
in the meta-analysis
controlled for age
and sex. Adjustment
for other potential
confounders varies
by studies, but
includes monthly
income, education
levels, smoking
status, BMI,
ethnicity, work
duration, ototoxic
medication, blood
lead, occupational
noise, loud noise,
and firearm noise,
and hypertension
and diabetes

OR (95% Cl)b
1.34 (1.18, 1.52)

tTu etal. (2021)

NHANES, U.S.

2011-2012
Cross-sectional

NHANES
n: 1503

Blood

Measured by plasma
mass spectrometry

Age at measurement:
20-69 years

Median=1.07 |jg/l
95th: 1.62 pg/l

Speech-frequency hearing
loss;

High-frequency hearing loss

For each ear, 0 5, 1, 2, 3, 4
and 6 kHz frequencies
were used for assessing
pure-tone air conduction
hearing thresholds
over a -10 to 110 dB
intensity ranges. The
average of four

Age, sex, education,
marital status, BMI,
smoking, noise
exposure,
hypertension
and diabetes

OR (95% Cl)b

HFHL 1-98 (1-27, 3-10)
SFHL 1.46 (0.81, 2.64)

External Review Draft

3-467

DRAFT: Do not cite or quote

-------
RefereDCesignnd	Study Population Exposure Assessment	Outcome	Confounders Effect 9E|0^es and

audiometric frequencies
(0 5, 1, 2 and 4 kHz) was
used to identify
speech-frequency hearing
loss (SFHL), while the
average of three audiometric
frequencies (3, 4 and 6 kHz)
was used to identify
high-frequency hearing loss
(HFHL). SFHL or HFHL > 25
dB in either ear was sued to
define hearing loss, based
on the WHO definition for
this condition

Age at outcome: 20-69
years

tPaulsen et al. (2018)

Beaver Dam Offspring Study
Beaver Dam, Wisconsin,

U.S.

Baseline data collection
June 8, 2005, through
August 4, 2008 with two
follow-up examinations
occurred at 5-year intervals:
one was conducted between
July 12, 2010, and March 21,
2013, and the other between
July 1, 2015, and November
13, 2017

BOSS
n: 1983

Blood

Measured by Inductively

coupled plasma mass
spectrometry

Age at Measurement:
21-84 years

Central tendency BLL: NR

Contrast sensitivity
impairment

Age, alcohol
consumption,
smoking, AMD,
cataract, plaque site,
VA impairment, and
sex

HR (95% Cl)b
0.91 (0.696, 1.

19)

Cohort

tFillion et al. (2013)

n: 228

Blood

Contrast sensitivity (cycles
per degree, cpd);

Age, sex, current
smoking (yes vs. no)

Beta (95% Cl)b
Spatial frequency with
%EPA

External Review Draft

3-468

DRAFT: Do not cite or quote

-------
Reference and Study
Design

Study Population Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

Lower Tapajos River Basin,

State of Para

Brazil

May to July 2006
Cross-sectional

Measured by Inductively
coupled plasma mass
spectrometry (ICP-MS)

Age at Measurement:
15-66 years
(median=33.0 years)

Mean=12.8±8.4 pg/dL;
Median=10.5 [jg/dL

Acquired color vision loss
(color confusion index, CCI)

Age at outcome: 15-66
years

current drinking (yes

vs. no)

1.5 cpd -1.32 (-4.30;
1.65)

3 cpd 2.06 (-2.87; 6.99)
6 cpd 0.60 (-6.04; 7.25)
12 cpd -13.33 (-23.28;
-3.49)

18 cpd -2.43 (-6.64;
1.79)

CCI 0.16 (-0.03; 0.33)

BLL = blood lead level; CI = confidence interval; CIDI = Composite International Diagnostic Interview; KXRF = K-shell X-Ray fluorescence; NAS = Normative Aging Study; OR = odds
ratio; Pb = lead; PTA = pure tone average; Q = quartile; RR= relative risk; SD = standard deviation; SE = standard error; WHO = World Health Organization; yr = year(s).
aEffect estimates are standardized to a 1 |jg/dL increase in BLL or a 10 |jg/g increase in bone Pb level, unless otherwise noted. For studies that report results corresponding to a
change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the biomarker and standardized accordingly.
bResult not standardized because data pertaining to the BLL distribution and/or base for the log-transformation were not reported.
fStudies published since the 2013 Integrated Science Assessment for Lead

External Review Draft

3-469

DRAFT: Do not cite or quote

-------
Table 3-16T Animal toxicological studies of Pb exposure and sensory organ function.

Study

Species (Stock/Strain),
n, Sex

Timing of
Exposure

Exposure
Details

BLL As Reported
(Hg/dL)

Endpoints Examined

tJamesdaniel et al. (2018)

Mice (C57BL/6)

Control (tap water), M,
n =6

2 mM, M, n = 6

PND 33 to
PND 61

Oral,

drinking

water

PND 61:

10 [jg/L (1 [jg/dL) for
Control

293 |jg/L (29.3 pg/dL) for
2 mM

PND 61: Auditory threshold (via BAEP)

tCarlson et al. (2018)

Mice (CBA/CaJ)

Control (deionized water),
M, n = 16

0.03 mM, M, n = 8

5 wk to
16 wk

Oral,

drinking

water

16 wk:

-------
Study

Species (Stock/Strain), Timing of Exposure BLL As Reported
n, Sex	Exposure Details	(|jg/dL)

Endpoints Examined

tLiuetal. (2019)

Rat (Sprague Dawley)

Control (tap water), F,
n = 12

58 mg/L, F, n = 11

PND 1 to
PND 21

Oral,

drinking

water

PND 9:

0 [jg/dL for Control
7.9 [jg/dL for 58 mg/L
PND 21:

0 [jg/dL for Control,
8.2 [jg/dL for 58 mg/L
PND 40:

0 [jg/dL for Control
0 [jg/dL for 58 mg/L

PND 93: Sound-Azimuth Discrimination
Training

BAEP = brainstem auditory evoked potentials; BLL = blood lead level; F = female; KNHANES = Korea National Health and Nutrition Examination Surveys; M = male; Pb = lead;
PND = postnatal day; wk = week(s).

fStudies published since the 2013 Integrated Science Assessment for Lead.

External Review Draft

3-471

DRAFT: Do not cite or quote

-------
Table 3-17E Epidemiologic studies of exposure to Pb and neurodegenerative disease in adults.

Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

Wang et al. NAS
(200Z)	n: 358

Normative
Aging Study
(NAS)

US

Bone Pb

measurement

(1991-1999),

Mini-Mental

State

Examination
(MMSE) twice
(1993-1998 and
1995-2000)

Tibia and patella	Cognitive decline

cognitive assessment battery
Measured by K-X-ray fluorescence was the MMSE, a global

examination of cognitive
function that assesses
orientation, immediate and
Median: 19 and 23 |jg/g for tibia and short-term recall, verbal and
patella	written skills, and attention

and ability to follow
commands

Age at measurement 21-81 years

Adjusted for age, years of Change in MMSE score

education, nonsmoker,
former smoker, pack-
years, nondrinker,
alcohol consumption,
English as first language,
computer experience,
and diabetes

Age at outcome: 21-81 years

per IQR (15 pg/g)
increase in tibia Pb by
class of HFE genotype3

Wild-type -0.02 (-0.10 to
0.07)

One HFE variant allele -
0.14 (-0.33 to 0.04)
Two HFE variant alleles -
0.63 (-1.04 to -0.21)

Cross-sectional

Weisskopf et al. NAS
(2004)	n: 466

Normative
Aging Study,

US

1991 and 2002
Cross-sectional

Tibia, patella, and blood

Bone Pb measured by ABIOMED
KXRF instrument and blood Pb
measured by Zeeman background-
corrected fiameless atomic
absorption (graphite furnace)

Age at measurement 21-81 years

Median pg/g (interquartile range)

Patella 27 (19, 40) pg/g
Tibia 21 (15, 29) pg/g
Blood 5 (3, 7) pg/dL

Cognitive decline

cognitive assessment battery
was the MMSE, a global
examination of cognitive
function that assesses
orientation, immediate and
short-term recall, verbal and
written skills, and attention
and ability to follow
commands

Age at first MMSE test,
alcohol intake, and days
between the two MMSE
tests as continuous
variables, as well as
education (<12 years, 12
years, 13-15 years, >16
years), smoking status
(never, former, current),
computer experience
(yes/no), and English as
a first language (yes/no)

Difference in change in
MMSE score per IQR
increase in Pba
-0 25 (-0 45, -0 05)

External Review Draft

3-472

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

(Wright et al.
2003)

Normative
Aging Study,
US

1991-1997
Cross-sectional

NAS
n: 1033

Tibia, patella, and blood

Bone Pb measured by ABIOMED
KXRF instrument and blood Pb
measured by Zeeman background-
corrected fiameless atomic
absorption (graphite furnace)

Age at measurement 21-81 years
Mean (SD)

Patella 29 5 (21.2) |jg/g
Tibia 22 4 (15.3) |jg/g
Blood 4.5 (2.5) pg/dL

Cognitive decline	Age, alcohol intake,

cognitive assessment battery education history

was the MMSE, a global

examination of cognitive

function that assesses

orientation, immediate and

short-term recall, verbal and

written skills, and attention

and ability to follow

commands

and OR (95% Cl)a MMSE<24

Tibia 1.02 (1.00,1.04)
Patella 1.02(1.00,1.03)

External Review Draft

3-473

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

(Weuve et al.. NAS
2006)	n: 1171

Normative
Aging Study,

US

1991 and 2002
Cross-sectional

Tibia, patella, and blood

Measured by graphite furnace

atomic absorption with Zeeman
background correction

Age at measurement 21-81 years

Median (and first and third quartiles)
of tibia and patella were 19(13, 28)
and 27 (18, 39) pg/g

Blood 5 2 (<1-28) pg/dl

ALAD genotype
modifications on cognition

cognitive assessment battery
was the MMSE, a global
examination of cognitive
function that assesses
orientation, immediate and
short-term recall, verbal and
written skills, and attention
and ability to follow
commands

Age at cognitive
assessment and age-
squared, years of
education (<8, 9-11, 12,
13-15, 16, >17 years),
computer experience (an
additional measure of
socioeconomic status),
and length of time
between the lead and
cognitive assessments,
were smoking status
(current v past or never),
alcohol consumption
(none, 0 1-4 9 g/day,
5 0-9 9 g/day, >10

Mean difference in MMSE
score per IQR increase in
Pb (95% Clf
Tibia

Among ALAD-2 carriers
-0 16 (-0 58 to 0.27)
Among ALAD wildtypes
-0 05 (-0.21 to 0.12)
Patella

Among ALAD-2 carriers
-0 26 (-0 64 to 0.12)
Among ALAD wildtypes
-0 07 (-0 23 to 0.09)
Blood

Among ALAD-2 carriers
0 26 (-0 54 to 0.01)

g/day, or missing), calorie Among ALAD wildtypes

adjusted calcium intake -0.04 (-0.16 to 0.07)

(in tertiles), regular

energy expenditure on

leisure time physical

activity (in tertiles), and

diabetes (physician

diagnosed or fasting

blood glucose >126

mg/dl)

(Nordberq et al.. Kungsholmen project
2000)	n: 762

Stockholm

1994-1996

Cross-sectional

Blood

Measured using Graphite furnace
atomic absorption spectrometry

Age at measurement: 75+ (mean
age of 88.4 years)

Mental Performance
MMSE

Age and BMI

No association was
reported (quantitative
estimate NR)

Mean (SD) 3.7 (2.3) (jg/dl

External Review Draft

3-474

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

tFarooqui et al. Participants selected Bone

(2017)

Boston, MA,
United States

1993-2007

Cohort

from cohort study
(Veterans Affairs NAS);
healthy men aged 51-
98 yr

n: 741 subjects in MMSE
and 715 in Global
cognition

Patella (trabecular bone) and tibia
(cortical bone) bone Pb was
measured using KXRF
spectroscopy
in 1993

Patella mean (SD)

30.6 ± 19.44 |jg/g, and tibia mean

(SD) 21.6 ± 13.33 |jg/g

Changes in cognition

Cognition was assessed
using the MMSE, NES2,
CERAD and WAIS-R during
3-5 visits over the period of
15 yr of follow-up.

Age at first cognitive test, HR (95% Cl)b

past education level,
baseline smoking status
and alcohol intake.

Cognition
MMSE <25
Tibia

1.05 (0.82, 1.35)

Patella

1.21 (0.99, 1.49)

Beta (95% Cl)b

Global Cognition
(Summary score of NES2,
CERAD and WAIS-R)

Tibia

-0.206 (-0.453, 0.089)
Patella

-0.25 (-0.518, 0.019)

Cognition

MMSE

Tibia

-0.077 (-0.206, 0.052)
Patella

-0.128 (-0.251, -0.0004)

External Review Draft

3-475

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

tYanq et al.
(2018)

Multicity
(Taichung city,
Changhua,and
Nantou County);
Taiwan

Feb 2015-
2016

¦Oct

Case-control

Participants were
recruited from the China
Medical University
Hospital. Cases recruited
the Department of
Neurology, and controls
from the

' Department of Family
Medicine who were
receiving general health
check-up; aged >50 yr.

n: Full sample: 434 (170
AD and 264 controls);
Propensity-score-
matched sample: 84 AD
and 84 controls.

Blood assessed for heavy metals AD risk

Blood samples tested for heavy
metals (Pb, Cd, Se, Hg). Blood Pb
measured through ICP-MS

Blood Pb cone: Full samples: AD:
2.50 ± 1.35 [jg/dL, Controls:
2.36 ± 1.02 [jg/dL

Propensity matched samples: AD:
2.58 ± 1.35 [jg/dL, Controls:
2.50 ±1.18 [jg/dL

Physician AD diagnosis
based on definition of the
Diagnostic and Statistical
Manual Fourth Edition
Criteria; MMSE test of
cognitive function

Age, gender, education,
exercise habits,
hypertension, diabetes,
cardiovascular diseases,
depression, anxiety

OR (95% Cl)b

Full population
Total: 1.05 (0.86-1.28)
Tertile 2 vs 1: 1.00 (0.56-
1.79)

Tertile 3 vs 1: 0.87 (0.49-
1.55)

Propensity score-matched
population

Total: 1.06 (0.83-1.35)

Tertile 2 vs 1: 1.16 (0.55-
2.47)

Tertile 3 vs 1: 1.12 (0.53-
2.39)

tHorton et al.
(2019)

Nationwide,
United States

1999-2008
followed till
2014

Cohort

Participants selected Blood assessed for Pb
from the five NHANES

cycles and included 1999 Blood samples collected during the

AD mortality

to 2008 who were
followed till 2014 for
death; aged >60 yr.

n: 8,080 subjects

NHANES mobile examination
center visit assessed for Pb using
ICP-DRC-MS.

The identification of AD
mortality is based on the
immediate cause of death in
the National Death Index
record. Cause of death was
Blood Pb cone: Geo mean and 95% coded according to the
CI: 2.1 (2.02, 2.11) pg/dL	ICD-10, revision 10; G30

was used to indicate AD.

Age, sex, poverty status,
race/ethnicity, and
smoking status, and
competing risks for AD
mortality.

HRR (95% Cl)b
0.3 [jg/dL:
ref

0.5 [jg/dL:

1.1	(0.89, 1.3)

1	[jg/dL:

1.2	(0.77, 1.8)
1.5 [jg/dL:

1.2	(0.7, 2.1)

2	[jg/dL:

1.3	(0.66. 3)

3	[jg/dL:

1.3	(0.6, 3.0)
5 [jg/dL:

1.4	(0.54, 3.8)

External Review Draft

3-476

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

(Vinceti et al.
1997)

n: 15 cases and 36
controls

Blood

ALS measured by ALS
severity scale



Age at measurement:



Santa Mafia

(mean ±SD)



Nuova Hospital

Patients 65.9±14.0 years



in Reggio

Controls 64.4±12.9



Emilia, northern





Italy

Mean (SD)





Controls 108.3±44.4 |jg/l



December 31st,

Patients 127.1±67.8 pg/l



1994





Case-control





(Kameletal., n: 109 cases and 256

Blood and bone

ALS

2002) controls



A board-certified neurologist



Blood lead was measured using

(T. L. M. or J. M. S.)

New England

graphite furnace atomic absorption

evaluated potential cases.

and U.S.

spectroscopy. Bone lead was

Diagnosis of ALS was based



measured using in vivo K-XRF

on criteria published by the

1993-1996



World Federation of



Age at measurement: 30-80 yrs

Neurology.

Case-control





Patients and controls Correlation coefficient
matched on year of birth ALSSS (p-value)b
and gender, confounders j0ta| -0 440 (0 101)
NR

Cases and controls
matched on age, sex,
and region

OR (95% Cl)b
Blood 1.9 (1.4, 2.6)
Tibia 2.3 (0.4, 14.5)
Patella 3.6 (0.6, 20.6)

Mean (SE)

Blood [jg/dl
Cases 5.2 (0.4)
Controls 3.4 (0.4)
Patella |jg/g
Cases 20.5(2.1)
Controls 16.7 (2.0)
Tibia |jg/g
Cases 14.9 (1.6)
Controls 11.1 (1.6)

External Review Draft

3-477

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

Kamel et al.
(2008)

New England

and

U.S.

Enrollment:
1993 - 1996;
follow-up
through
December 31,
2003

Cohort

n: 110

Bone and blood

Bone Pb measured in the tibia and
patella using K X-ray fluorescence;
blood Pb measured using atomic
absorption spectrometry

Age at Measurement:

Median (range) = 60 (30-79) years

Blood Pb median = 4 [jg/dL; patella
Pb median = 15 |jg/g; tibia Pb mean
= 13 |jg/g

Max: Blood Pb max = 14 [jg/dL;
patella Pb max = 107 |jg/g; tibia Pb
max = 61 |jg/g

Neurodegenerative Disease -
ALS

Amyotrophic lateral sclerosis
(ALS) was diagnosed by
board-certified neurologists
and based on the World
Federation of Neurology El
Escorial criteria; related
symptoms were documented
from interviews. Cause of
death was identified by the
National Death Index (NDI).

Cox proportional hazard
analyses adjusted for
age, sex, and ever
smoked, except for sex-
stratified models, which
included age and ever
smoked.

HR (95% Cl)a
Diagnosis to death 0.9
(0.8 to 1.0)

Symptoms to death 0.9
(0.8 to 1.0)

(Fang et al..
2010)

U.S.

2003-2007
Case-control

n: 184 cases and 194
controls

Blood	ALS	Age

Neurologists with expertise in
Pb measured by inductively coupled ALS reviewed medical
plasma mass spectrometry	records to determine motor

neuron disease diagnosis in
accordance with the original
El Escorial Criteria, including
ALS (International
Classification of Diseases,

Ninth Revision (ICD-9) code
335 20), progressive
muscular atrophy (ICD-9
code 335 21), progressive
bulbar palsy (ICD-9 code
335 22), pseudobulbar palsy
(ICD-9 code 335 23), primary
lateral sclerosis (ICD-9 code
335 24), and other motor
neuron diseases (ICD-9 code
335 29).	

OR (95% Cl)b
Overall 1.9 (1.3, 2.7)

Age at measurement: mean (SD)
Cases 63 3(34—83)

Controls 63 4 (34-84)

Pb Mean (SD)

3 4 jjg/dL (2.5)

External Review Draft

3-478

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

tFanq et al.
(2017)

Nationwide,
United States

2007-2013

Cohort

Veterans with ALS in the
U.S. National Registry of
Veterans, and other
veterans with ALS not
treated within the
Veterans Affairs
healthcare system, with
ALS from April 2003 to
September 2007 and
followed till the date of
death or July 25, 2013;
non-Hispanic
Caucasian men aged
34-83 yr

n: 145 U.S. Veterans
with ALS, who were
male, diagnosed with
ALS by neurologist.

Blood assessed for Pb

Whole blood collected during Jan-
Sep 2007 assessed for Pb using
ICP-MS.

Biomarkers for bone formation
measured in plasma. Bone
formation was measured using
procollagen type I N-terminal
propeptide and bone resorption
measured using C-terminal collagen
crosslinks.

Blood Pb cone: 2.35 ± 1.28 [jg/dL

ALS survival	Age at diagnosis,

diagnostic certainty, site
Trained neurologist with of onset, diagnostic delay
expertise in ALS assigned and revised ALS
diagnoses using an algorithm Functional Rating Scale
based on the revised El Score
Escorial Criteria

HR (95% Cl)a
1.234 (1.021, 1.49)

tVinceti et al. Cases were ALS patients Cerebral spinal fluid assessed for ALS

(2017)

Emilia-

Romagna,

Italy

May 1998—April
2011

and controls were
selected from hospital-
admission of no ALS;
mean age cases: 52 yr

n: 76 (38 ALS cases and
38 controls)

heavy metals

CSF evaluated for heavy metals
(Pb, Cd, Hg) using ICP-MS

Median Pb cone: Cases: 155 ng/L,
Controls: 132 ng/L

Age, sex, and total
selenium

Probable ALS diagnosis
using the revised El Escorial
Criteria.

In the highest fertile of
exposure, OR (95% Cl)b
1.39 (0.48, 4.25)

Case-control

External Review Draft

3-479

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

tAndrew et al.
(2022)

Nationwide,
United States

2013-2019
Case-control

Participants are from the
healthcare claims
dataset from Symphony
Health with ALS
diagnosis after 6 mo
enrollment in the

268 Airborne contaminants

Airborne exposure to Pb and other
contaminants assessed from U.S.
EPA's NEI database for 2008 to
estimate exposure prior to ALS
database prior to the first onset. Qata was usec| t0 estimate

ALS ICD code. Controls residential exposure at the zip3

are individuals similar to |ocations of the ALS patients and

ALS cases based on controls

age, sex, and length of

database history with min

of 6 mo in database;

cases and controls age

18-80 yr (63% were 55-

75 yr)

n: Cases: 26,199 and
controls: 78,597

ALS

ALS based on the healthcare
claims

Family income, race,
age, and sex

OR (95% Cl)b
Discovery and
Validation Cohorts

1.39 (95% CI 0.48-4.25)

New

Hampshire/Vermont

[Q1+Q2: <1.37 tons (Ref)
Q3: 1.37-26.1
Q4: >26.1]

5-year Exposure History
Q3: 1.79 (1.32, 2.43)
Q4: 1.11 (0.8, 1.55)

10-year Exposure History
Q3: 2.42 (1.76, 3.33)
Q4: 2.03 (1.46, 2.8)

15-year Exposure History
Q3: 1.83 (1.34, 2.52)
Q4: 1.73 (1.26, 2.38)

Ohio

[Q1+Q2: <14.7 tons (Ref)
Q3: 14.7-50.8
Q4: >50.8]

5-year Exposure History
Q3: 0.48 (0.37, 0.61)
Q4: 0.39 (0.3, 0.51)

10-year Exposure History
Q3: 1.05 (0.83, 1.33)
Q4: 1.6 (1.28, 1.98)

15-year Exposure History
Q3: 0.94 (0.74, 1.18)
Q4: 1.07 (0.86, 1.34)

External Review Draft

3-480

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

tPaul et al.
(2021)

Australia and
New Zealand,
and Central
California

Case-control

Participants for this study
comes from two publicly
available PD studies:
SGPD consortium of
three studies across
Australia and New
Zealand with cases and
controls. PEG a
population-based study
from three agricultural
counties of Central
California with cases and
controls.

Epigenetic biomarkers for	PD

cumulative Pb exposure (tibia and
patella), i.e., DNAm Pb

Epigenetic biosensors identified
with site-by-sire analysis and
combined via machine learning
algorithm on KXRF in vivo
measures of bone Pb. To determine
Pb biomarker level in two cohorts,
the regression coefficients were
extracted from NAS and applied to
the DNAm beta matrices.

Age (DNAm Age in
SGPD), sex, ancestry
(PEG only), smoker
(PEG only), blood cell
composition, and mean
Meth By Sample

OR (95% Cl)b
Tibia SPGD
1.54 (1.22, 1.95)
Patella SPGD
0.70 (0.53, 0.93)
Tibia PEG
1.52 (1.25, 1.86)

n: SGPD cohort: 959
cases and 930 controls;
PEG cohort: 569 cases
and 238 controls

DNAm tibia Pb: SGPD cohort:
cases: 3.41 ± 0.4, controls:
3.48 ± 0.4; PEG cohort: cases:
3.06 ± 0.4, controls: 3.03 ± 0.3

tJi et al. (2015) Participants selected are Blood, Bone assessed for Pb

Boston, MA,
United States

NAS

Cohort

subgroup of participants
from cohort study
(Veterans Affairs NAS);
healthy men aged 50-
98 yr.

n: 807

Blood samples tested for Pb
concentration using Zeeman
background-corrected flameless
atomic absorption graphite furnace.
Bone Pb concentration measured
with KXRF at both the tibia and the
patella starting in 1991.

Blood Pb concentration:

5.01 ± 2.72 [jg/dL

Tibia Pb cone: 21.23 ± 13.29 |jg/g
(); patella Pb cone:

27.98 ± 18.38 pg/g)

Tremor

Tremor score was created
based on an approach using
hand-drawing samples that
were derived from figure
copying testing performed as
part of larger cognitive test
battery (CERAD, MMSE, and
VMI) assessed over a mean
follow-up of 8.0 ± 3.2 yr after
bone Pb measurement.

ALAD genotype was
determined by amplifications
of 0.5 pl_ of whole blood
using two sets of primers
specific for a portion of the
ALAD gene.

Age, age squared,
alcohol consumption,
smoking status,
education level

OR (95% Cl)b
Blood

Quintile 2 vs 1
0.09 (-0.10, 0.29)
Quintile 3 vs 1
0.06 (-0.16, 0.28)
Quintile 4 vs 1
0.07 (-0.13, 0.27)
Quintile 5 vs 1
0.07 (-0.16, 0.30)
Patella

Quintile 2 vs 1
-0.07 (-0.30, 0.15)
Quintile 3 vs 1
-0.14 (-0.37, 0.09)
Quintile 4 vs 1
-0.22 (-0.45, 0.01)

External Review Draft

3-481

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

Quintile 5 vs 1
-0.02 (-0.27, 0.22)
Tibia

Quintile 2 vs 1
0.03 (-0.20, 0.26)
Quintile 3 vs 1
0.03 (-0.20, 0.26)
Quintile 4 vs 1
0.13 (-0.11, 0.36)
Quintile 5 vs 1
-0.07 (-0.32, 0.17)

tKhalil et al.
(2014)

Pittsburgh, PA
United States

2007-2009

Cross-sectional

MrOS	Blood

n: 445

Blood Pb measured using AAS
Non-Hispanic Caucasian Age at measurement:
men (community dwelling Mean = 79.5 ± 5 yr
non-institutionalized) at
least 65 yr of age	Mean = 2.25 [jg/dL;

enrolled in MrOS at the SD = 1.20 [jg/dL; Median = 2 [jg/dL
University of Pittsburgh Max: 10 |jg/dL
clinic. Eligibility criteria
included the ability to
walk unaided and without
bilateral hip
replacements.

Grip strength (kg);
Leg extension power (watts);
Walking speed (m/s);
Narrow-walk pace (m/s);
Use arms to stand up
(yes/no)

Grip strength was measured
on a Jamar dynamometer;
Leg extension power was
measured with the
Nottingham power rig;

Walking speed was assessed
on a standard 6-m walking
course;

Narrow-walk pace (an indirect
measure of dynamic balance)
was assess while keeping
each foot within a 20-
centimeter wide lane on the
6-m walking course;

Stand from a chair without
using the arms was
measured as yes/no.

Age at outcome:

65 yr

Age, education, smoking,
alcohol consumption,

BMI

Beta (95% Cl)a

Leg extension power

-0.03 (-1.97, 2.03)

Ability to stand from a
chair without using their
arms 0.97 (0.88, 1.07)

External Review Draft

3-482

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

tShiue (2013)

United States

2003-2004

NHANES

n: 712 (vision); 732
(hearing); 669 (balance)

NHANES age 50 and
above

Cross-sectional

Urine

Urinary Pb was detected by mass

spectrometry

Age at measurement:

50 yr

Not Reported

Vision;

Hearing;

Balance;

Ear ringing

Vision: excellent, good, and
fair eyesight (self-reported)
were classified as good; poor
and very poor were classified
as poor

Hearing: good and little
trouble hearing (self-reported)
were classified as good; lots
of trouble and deaf were
classified as poor
Balance: "During the past
12 mo, have you had
dizziness, difficulty with
balance, or difficulty with
failing?"

Ear ringing: "ears ringing,
roaring, or buzzing in the last
year"

Age at outcome:

50 yr

Age, sex, ethnicity,
creatinine, survey
weighting

urine OR (95% Cl)b

Vision 1.15 (0.67-1.97)
Hearing 0.97 (0.63-1.51)
Balance 0 68 (0.51-0.91)

External Review Draft

3-483

DRAFT: Do not cite or quote

-------
tCasiens et al.
(2018)

HNRS
n: 1188

Ruhr area, a Men from the Heinz

German	Nixdorf Recall Study.

industrial region Recruitment details not

with a high provided.

volume of steel

production

Germany

Baseline

recruitment

2000-2003;

Follow-up

2011-2014

Cohort

Blood

Blood Pb was measured in aliquots
of whole blood archived at baseline
and at follow-up using ICP-MS
Age at measurement:

Median = 58 yr at baseline (range
45-75 yr) and 68 yr (range 55-
86) yr at follow-up

Median = 3.29 (IQR 2.55-
4.32) [jg/dL at baseline; 2.59 (IQR
1.99—3.39) [jg/dL at follow-up
Max: 67.73 [jg/dL at baseline;
39.68 [jg/dL at follow-up

Odor identification;

Tapping hits;

Aiming errors;

Line tracing errors;

Steadiness errors

Odor identification: Sniffin
sticks odor identification test
of 12 odors, participants
classified as normosmic if >9
odors identified, hyposmic if
7-9 odors identified, and
functionally anosmic if <7
odors identified
Tapping hits: tapping a stylus
within 32 s as fast as
possible; hits <10th percentile
were considered as
substantially impaired manual
dexterity

Aiming errors: 20 small plates
with a diameter of 5 mm
standing in a line (distance 4
mm) had to be touched with a
stylus as fast as possible;
errors >90th percentiles were
considered as substantially
impaired manual dexterity
Line tracing errors: drawing a
stylus through a curvy course
of a groove without touching
side walls or bottom; errors
>90th percentiles were
considered as substantially
impaired manual dexterity
Steadiness errors: maintain a
precise arm-hand position by
holding a stylus for 32 s in a
5.8 mm hole without touching
sides or bottom; errors >90th
percentiles were considered
as substantially impaired
manual dexterity

Occupational
qualification, age,
smoking status, alcohol
consumption, total test
time

OR (95% Cl)b
Group 1: <5 [jg/dL (Ref)
Group 2: 5 - <9 [jg/dL
Group 3: >9 [jg/dL

Odor identification

baseline

G2: 0.91 (0.65, 1.28)
G3: 1.96 (0.94, 4.11)
follow-up

G2: 1.04 (0.55, 1.94)
G3: 1.57 (0.47, 5.19)

Motor Performance
Series

Steadiness Errors
baseline

G2: 0.99 (0.62, 1.59)
G3: 1.36 (0.5, 3.66)
follow-up

G2: 1.16 (0.5, 2.69)
G3: 1.75 (0.41, 7.58)

Line tracing errors
baseline

G2: 1.09 (0.68, 1.76)
G3: 0.93 (0.32, 2.74)

follow-up

G2: 1.01 (0.41, 2.48)
G3: 0.59 (0.08, 4.11)

Aiming errors
baseline

G2: 1.07 (0.75, 1.53)
G3: 0.56 (0.22, 1.42)

follow-up

G2: 1.35 (0.73, 2.51)

External Review Draft

3-484

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

Age at outcome:
55-86 yr

G3: 0.42 (0.09, 2.08)

Tapping Hits
baseline

G2: 0.87 (0.53, 1.44)
G3: 1.35 (0.49, 3.7)

follow-up

G2: 2.63 (1.26, 5.49)
G3: 0.8 (0.14, 4.59)

tJi et al. (2013) NHANES

Blood

Walking speed (ft/sec)

United States

1999-2002

Cross-sectional

Blood Pb was measured using AAS Time to walk 20 ft (at usual

Age at measurement:
50-85 yr (Median = 61.2 yr)

n: 3,593 (1,798 women;

1,795 men)

NHANES data from the
1999-2000 and 2001-
2002 surveys,	Mean ± SD:

participants were 50 yr of Women = 2.17 ± 0.04 [jg/dL;
age	Men = 3.18 ± 0.08 [jg/dL ("there's a

slight discrepancy in the SDs in
Table 1 vs. text on p. 712)

Median: Women = 1.72 [jg/dL;
Men = 2.41 |jg/ dL

walking pace)

Age at outcome:

50-85 yr (median = 61.2 yr)

Model 4 (fully adjusted):
age, education, ethnicity,
height, waist
circumference, alcohol,
smoking, physical
activity, arthritis,
diabetes, heart condition,
hypertension,
homocysteine, C-reactive
protein

Beta (95% CI)

Walking Speed-Men 4.4
to <54.0

-0.029 (-0.155, 0.097)
Walking Speed-Women
3.0 to <53.0

-0.114 (-0.191, -0.038)

Walking Speed-Men 3.1
to < equal to 4.3

0.082 (-0.012, 0.176)

Walking Speed-Women

2.2	to <2.9

-0.104 (-0.187, -0.021)

Walking Speed-Men 2.4
to <3.0

-0.17 (-0.26, -0.08)
Walking Speed-Women
1.7 to <2.1

-0.024 (-0.118, -0.063)

Walking Speed-Men 1.8
to <2.3

0.057 (-0.051, 0.165)

Walking Speed-Women

1.3	to <1.6

-0.027 (-0.055, 0)

External Review Draft

3-485

DRAFT: Do not cite or quote

-------
tMin et al.
(2012)

United States

1999-2004

Cross-sectional

NHANES
n: 5574

Adults who participated
in the NHANES Balance
Component and had
blood Pb and Cd
measurements and data
for all covariate variables

Blood

Blood Pb was measured using a
multielement AAS with Zeeman
background correction.

Age at measurement:

40 yr

Weighted mean (participant without
balance dysfunction): 2.09 [jg/dL
(95% CI: 2.01, 2.18); Weighted
mean (participants with balance
dysfunction): 2.39 ug/dL (95% CI:
2.29, 2.49)

Max: 48 pg/dl_

Balance dysfunction

Balance dysfunction was
evaluated by the Romberg
Test of Standing Balance on
Firm and Compliant Support
Surfaces, which measured
the participant's ability to
maintain balance under four
test conditions: Test 1)
maintain balance while
standing (with feet together
and arms folded across the
waist, holding each elbow
with the opposite hand) for 15
sec.; Test 2) maintain
balance while standing for 15
sec with eyes closed so that
only vestibular and
proprioceptive (i.e., leg
muscle position sense)
information is available; Test
3) maintain balance while
standing for 30 sec on a
foam-padded surface, which
reduces proprioceptive input
but does not affect visual or
vestibular input; Test 4)
maintain balance while
standing for 30 sec on a
foam-padded surface with
eyes closed, so that input is
available from the vestibular
system only. Each condition
was scored on a pass or fail
basis. The time to failure (i.e.,
loss of balance) was also
recorded for test condition 4,
with those who passed the
test assigned the maximum
value of 30 sec.

Age, sex, race/ethnicity,
education, pack-years of
smoking, alcohol
consumption, histories of
stroke and diabetes,
intakes of Ca2+ and iron

OR (95% CI)

Balance Dysfunction
(Quintile 5 [3.3^8 |jg/dL])

33.334 (1.939, 573.157)
Balance Dysfunction
(Quintile 4 [2.3-
3.2 pg/dL])

5.234 (0.59, 46.429)
Balance Dysfunction
(Quintile 3 [1.8-
2.2 pg/dL])

0.665 (0.05, 8.783)

Balance Dysfunction
(Quintile 2 [1.3-
1.7 pg/dL])

3.707 (0.544, 25.282)

Age at outcome:
40 yr

External Review Draft

3-486

DRAFT: Do not cite or quote

-------
Reference and
Study Design

Study Population

Exposure Assessment

Outcome

Confounders

Effect Estimates and
95% CIs

tGrashow et al.
(2013)

Greater Boston
area, MA,

United States

Grooved
pegboard May
2005-

December 2009
Neuroskill July

2004 and
November 2007

Cohort

Normative Aging Study
n: 362 for grooved
pegboard test; 328 for
the Neuroskill test

Bone

Grooved pegboard
(completion time, seconds);
Neuroskill (Signature score,
%); Neuroskill (Im pattern
score, %)

Bone Pb was measured at the
patella and the midtibial shaft using
an ABIOMED KXRF instrument.

Elderly, majority	"Tibia and patella bone Pb

Caucasian men originally concentrations reflect cumulative Pb Grooved pegboard test:
recruited from the greater exposure over different time	Subjects were asked to insert

Boston, Massachusetts windows: patella Pb reflects	the metal pegs into each of

area in the 1960s	exposure over the last decade, the 25 holes in sequence as

while tibia Pb half-life is on the orderquickly as possible with their

Age, smoking, education,
computer experience,
income

of decades"

Age at measurement:

NR

Mean patella Pb = 25.0 mg/g bone

(SD = 20.7); Mean tibia

Pb = 19.2 mg/g bone (SD = 14.6)

dominant hand without
practice trials;

Neuroskill tests (signature
score, %): Subjects were
asked to provide five samples
of their signature in
succession, written in their
natural manner;

Neuroskill tests (Im pattern
score, %): Subjects were
asked to provide five samples
of a series of cursive Ims (Im
pattern) using the
instrumented pen

Beta (95% Cl)a

Neuroskill-lm pattern
score

Patella 0.45 (0.178,

0.723)

Tibia 0.847 (0.163, 1.53)

Neuroskill-Signature

Score

Patella 0.08 (-0.47, 0.63)

Tibia -0.293 (-1.063,
0.477)

Grooved pegboard-
dominant hand
completion time

Patella 1.965 (0.553,
3.378)

Tibia 3.107 (1.157, 5.057)

Age at outcome:
Mean age = 69.1 yr
(SD = 7.2)

External Review Draft

3-487

DRAFT: Do not cite or quote

-------
Stu^y "d^stgn Study Population	Exposure Assessment	Outcome	Confounders	Effect Estimates and

AAS = atomic absorption spectroscopy; BLL = blood lead level; Cd = cadmium; CERAD = Consortium to Establish a Registry for Alzheimer's Disease; CI = confidence
interval; CSF = cerebrospinal fluid; EE = effect estimate(s); HNRS = Heinz Nixdorf Recall Study; HRR = hazard rate ratio; KXRF = K-Shell X-Ray fluorescence;

MMSE = Mini Mental State Examination; mo = month(s); MrOS = Osteoporotic Fractures in Men Study; NAS = Normative Aging Study; NEI = National Emission
Inventory; NHANES = National Health and Nutrition Examination Survey; OR = odds ratio; Pb = lead; SD = standard deviation; sec = second(s); VMI = visual-motor
integration; yr = year(s).

aEffect estimates are standardized to a 1 |jg/dL increase in BLL or a 10 |jg/g increase in bone Pb level, unless otherwise noted. For studies that report results
corresponding to a change in log-transformed Pb biomarkers, effect estimates are assumed to be linear within the 10th to 90th percentile interval of the
biomarker and standardized accordingly.

bResult not standardized because data pertaining to the BLL distribution and/or base for the log-transformation were not reported.
fStudies published since the 2013 Integrated Science Assessment for Lead.

Table 3-17T Animal toxicological studies of Pb exposure and neurodegeneration.

Study

Species (Stock/Strain), n,	Timing of	Exposure

Sex	Exposure	Details

BLL as Reported (pg/dL)

Endpoints Examined

Zhou etal. (2018)

Rat (Sprague Dawley)

Control (distilled water), M,
n = 10

0.5% solution, M, n = 10
1.0% solution, M, n = 10
2.0% solution, M, n = 10

PND 24 to PND 52 Oral, drinking
water

PND 52:

13.3 |jg/L (1.3 pg/dL) for Control

148.9 |jg/L (14.9 pg/dL) for 0.5%
solution

231.3	pg/L (23.1 pg/dL) for 1.0%
solution

293.4	pg/L (29.3 pg/dL) for 2.0%
solution

PND 24, 31, 38, 45,
52: Amyloid protein
expression, Brain
Cholesterol,
Expression of BACE1
and APP

External Review Draft

3-488

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain), n,
Sex

Timing of
Exposure

Exposure
Details

BLL as Reported (pg/dL)

Endpoints Examined

Li etal. (2016d)

Mice (Kunming)

Control (distilled water), M/F,
n = 10

0.1% solution (mass fraction),
M/F. n = 10

0.2% solution (mass fraction),
M/F, n = 10

0.5% solution (mass fraction),
M/F, n = 10

GD to PND21

Oral, lactation
In utero

PND 21:

10.62 |jg/L (1.1 pg/dL) for Control

40.71 pg/L (4.1 pg/dL) for 0.1%
solution

81.77 pg/L (8.2 pg/dL) for 0.2%
solution

103.36 pg/L (10.3 pg/dL) for 0.5%
solution

PND 21: Amyloid
protein expression

Gu etal. (2012)

Mice (Tg-SwDI)

Control (Na-acetate water),
NR, n = 4-7

50 mg/kg, NR, n = 4-7

4-8 wk to 10-14 wk

Oral, gavage

10-14 wk:

1.83 pg/dL for Control
29.5 pg/dL for 50 mg/kg

10-14 wk: Beta-
amyloid and APP
expression

Wu et al. (2020b)

Mice (C57BL/6)

Control (distilled deionized
water), M, n = 7-10

0.2% solution, M, n = 7-10

4 wk to 4 mo

Oral, drinking
water

16 mo:

66.4 pg/L (6.6 pg/dL) for Control

278.9 pg/L (27.9 pg/dL) for 0.2%
solution

16 mo: Expression of
BACE1 and APP,
Phosphorylated tau
expression

Sun etal. (2014)

Rat (Sprague Dawley)

Control (tap water), NR,
n =20

580 ppm, NR, n = 20

NR (230-260 g) -
3 mo of treatment

Oral, drinking
water

After 3 mo treatment:

3.0 pg/L (0.3 pg/dL) for Control

56.8 pg/L (5.7 pg/dL) for 580 ppm

After 3 mo treatment:
Immunohistochemistry
of APP

Gassowska et al. (2016b)

Rat (Wistar)

Control (tap water), M/F, n = 8
0.1% solution, M/F, n = 8

GD Oto PND 21

Oral, lactation
In utero

PND 28:

0.93 pg/dL for Control
6.86 pg/dL for 0.1% solution

PND 28: Tau protein
expression and
phosphorylation

External Review Draft

3-489

DRAFT: Do not cite or quote

-------
Study	Species (Stock/Strain), n, Tlmmgof	Exposure	BLL as Reported 	Endpoints Examined

Rahman et al. (2012b)

Rat (Wistar) PND 1 to PND 30

Control (tap water), M/F,
n =6-10

0.2% solution, M/F, n = 6-12

Oral, drinking
water

Oral, lactation

PND 21:

1.4 |jg/dL for Control
12.1 [jg/dL for 0.2% solution
PND 30:

1.2 [jg/dL for Control
12.8 [jg/dL for 0.2% solution

PND 21, 30: Tau
protein expression
and phosphorylation

Zhana et al. (2012)

Rat (Sprague Dawley) NR (40-60 g)

Oral, drinking

+8 wk from start of exposure

+8 wk from start of



Control (deionized water), M,

water



exposure:



n = 10



49.9 ng/mL (5 |jg/dL) for Control

Phosphorylated tau









expression, Alpha-



100 ppm, M, n = 10



100.9 ng/mL (10.1 pg/dL) for 100 ppm

Synuclein expression



200 ppm, M, n = 10



128.6 ng/mL (12.9 pg/dL) for 200 ppm





300 ppm, M, n = 10



147.7 ng/mL (14.8 pg/dL) for 300 ppm



Bihaqi and Zawia (2013)

Monkey (Macaca fascicularis) PND 0 to PND 400
Control, F, n = 4

1.5 mg/kg/day, F, n = 5

Oral, infant
formula
Oral, gelatin
capsules

PND 400:

3-6 [jg/dL for Control
19-26 [jg/dL for 1.5 mg/kg/day

23 yr:

Tau protein
expression and
phosphorylation, Tau
phosphorylation

External Review Draft

3-490

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain), n,	Timing of	Exposure

Sex	Exposure	Details

BLL as Reported (pg/dL)

Endpoints Examined

Feng et al. (2019)

Rat (Sprague Dawley)

Control (deionized water),
M/F, n = 8

0.8 g/L (maternal) and 0.3 g/L
(pup), M/F, n = 8

1.5 g/L (maternal) and 0.9 g/L
(pup), M/F, n = 8

GD -10 to
PND 490

Oral, drinking
water

Oral, lactation
In utero

PND 21:

0 mg/L (0 pg/dL) for Control
0.29 mg/L (29 pg/dL) for 0.8 g/L
0.69 mg/L (69 pg/dL) for 1.5 g/L
PND 287:

0 mg/L (0 pg/dL) for Control
0.29 mg/L (29 pg/dL) for 0.8 g/L
0.61 mg/L (61 pg/dL) for 1.5 g/L
PND 490:

0 mg/L (0 pg/dL) for Control
0.31 mg/L (31 pg/dL) for 0.8 g/L
0.58 mg/L (58 pg/dL) for 1.5 g/L

PND 21, 287, 490:
Neuronal Density,
Brain Volume

Mansouri et al. (2012)

Rat (Wistar)	PND 70 to PND

Control (distilled water), M/F, 100
n = 16 (8/8)

50 mg/L, M/F, n = 16 (8/8)

Oral, drinking
water

PND 100-Males:
2.05 pg/dL for Control
8.8 pg/dL for 50 mg/L

PND 100: Open Field
Test, Rotarod Test

PND 100 - Females:
2.17 pg/dL for Control
6.8 pg/dL for 50 mg/L

External Review Draft

3-491

DRAFT: Do not cite or quote

-------
Study

Species (Stock/Strain), n,	Timing of	Exposure

Sex	Exposure	Details

BLL as Reported (pg/dL)

Endpoints Examined

Mansouri et al. (2013)

Rat (Wistar)

Control (tap water or
water+NaAc), M/F, n = 16
(8/8)

50 ppm, M/F, n = 16 (8/8)

PND 55 to PND
181

Oral, drinking
water

PND 178-181 - Females:

NR for Control

10.6 [jg/dL for 50 ppm

PND 155-159:
Rotarod Test

PND 178-181 - Males:
NR for Control
18.9 pg/dL for 50 ppm

Singh et al. (2019)	Rat (Wistar)	3 mo to 6 mo

Control (distilled water), M, n
= 5

2.5 mg/kg, M, n = 5

Oral, gavage 6 mo:

5.76 [jg/dL for Control
28.4 pg/dL for 2.5 mg/kg

6 mo: Locomotor
Activity, Rotarod Test

Al-Qahtani et al. (2022) Mice (Albino)	8-9 wk to 14-15 wk Oral, gavage 14-15wk:	NR: Locomotor

Control (distilled water), M, n	Activity

= 10	1.2 [jg/100 mL (1.2 pg/dL) for Control

0.2 mg/kg, M, n = 10

7.1 pg/100 mL (7.1 pg/dL) for 0.2
mg/kg

APP = amyloid precursor protein; BACE1 = beta-secretase 1; F = female; GD = gestational day; M = male; mo = month(s); NR = not reported; Pb = lead; PND = postnatal day; wk =
week(s); yr = year(s).

External Review Draft

3-492

DRAFT: Do not cite or quote

-------
3.8 References

Abazvan. B: Dziedzic. J: Hua. K: Abazvan. S: Yang. C: Mori. S: Pletnikov. MY: Guilarte. TR. (2014). Chronic
exposure of mutant DISCI mice to lead produces sex-dependent abnormalities consistent with
schizophrenia and related mental disorders: A gene-environment interaction study. Schizophr Bull 40: 575-
584. http://dx.doi.org/10.1093/schbul/sbt071.

Abubakar. K: Mailafiva. MM: Danmaigoro. B: Chiroma. SM: Rahim. E: Zakari. M. (2019). Curcumin attenuates
lead-induced cerebellar toxicity in rats via chelating activity and inhibition of oxidative stress.

Biomolecules 9: 453. http://dx.doi.org/10.3390/biom9090453.

AbuShadv. MM: Fathv. HA: Fathv. GA: abd el Fatah. S: Ali. A: Abbas. MA. (2017). Blood lead levels in a group of
children: The potential risk factors and health problems. JPediatr(Rio J) 93: 619-624.
http://dx.doi.Org/10.1016/i.iped.2016.12.006.

Ahmad. F: Hague. S: Ravinavagam. V: Ahmad. A: Kamli. MR: Barreto. GE: Ghulam Md Ashraf. GE. (2020).
Developmental lead (Pb)-induced deficits in redox and bioenergetic status of cerebellar synapses are
ameliorated by ascorbate supplementation. Toxicology 440: 152492.
http://dx.doi.Org/10.1016/i.tox.2020.152492.

Ahmed. MB: Ahmed. MI: Meki. AR: Abdraboh. N. (2013). Neurotoxic effect of lead on rats: Relationship to
apoptosis. International Journal of Health Sciences 7: 192-199. http://dx.doi.org/10.12816/0006Q42.

Al-Oahtani. A: Aiarem. J: Okla. MK: Rubnawaz. S: Alamri. SA: Al-Oahtani. WH: Al-Himaidi. AR: Elgawad. HA:
Akhtar. N: Maodaa. SN: Abdel-Maksoud. MA. (2022). Protective effects of green tea supplementation
against lead-induced neurotoxicity in mice. Molecules 27: 993.
http://dx.doi.org/10.3390/molecules27030993.

Al-Saleh. I: Moncari. L: Jomaa. A: Elkhatib. R: Al-Rouai. R: Eltabache. C: Al-Raiudi. T: Alnuwavsir. H: Nester. M:
Aldhalaan. H. (2020). Effects of early and recent mercury and lead exposure on the neurodevelopment of
children with elevated mercury and/or developmental delays during lactation: A follow-up study. Int J Hyg
Environ Health 230: 113629. http://dx.doi.Org/10.1016/i.iiheh.2020.113629.

Alabdali. A: Al-Avadhi. L: El-Ansarv. A. (2014). A key role for an impaired detoxification mechanism in the
etiology and severity of autism spectrum disorders. Behavioral and Brain Functions 10: 14.
http://dx.doi.org/10.1186/1744-9081-10-14.

Albers. CA: Grieve. AJ. (2007). Test review: Bayley, N. (2006). Bayley Scales of Infant and Toddler Development-
Third Edition [Review]. J Psychoeduc Assess 25: 180-198. http://dx.doi.org/10.1177/0734282906297199.

Alvarenga. KF: Morata. TC: Lopes. AC: Feniman. MR: Corteletti. LC. (2015). Brainstem auditory evoked potentials
in children with lead exposure. Braz J Otorhinolaryngol 81:37-43.
http://dx.doi.Org/10.1016/i.biorl.2013.12.001.

Amato. MS: Magzamen. S: Imm. P: Havlena. JA: Anderson. HA: Kanarek. MS: Moore. CF. (2013). Early lead
exposure (<3 years old) prospectively predicts fourth grade school suspension in Milwaukee, Wisconsin
(USA). Environ Res 126: 60-65. http://dx.doi.Org/10.1016/i.envres.2013.07.008.

Amedu. NO: Omotoso. GO. (2020). Lead acetate-induced neurodegenerative changes in the dorsolateral prefrontal
cortex of mice: The role of Vitexin. Environ Anal Health Toxicol 35: e2020001.
http://dx.doi.org/10.5620/eaht.e2020Q01.

External Review Draft

3-493

DRAFT: Do not cite or quote

-------
Amos-Kroohs. RM: Graham. PL: Grace. C'E: Braun. AA: Schaefer. TL: Skelton. MR: Vorhees. CV: Williams. MT.
(2016). Developmental stress and lead (Pb): Effects of maternal separation and/or Pb on corticosterone,
monoamines, and blood Pb in rats. Neurotoxicology 54: 22-33.
http://dx.doi.Org/10.1016/i.neuro.2016.02.011.

An. J: Cai. T: Che. H: Yu. T: Cao. Z: Liu. X: Zhao. F: Jing. J: Shen. X: Liu. M: Du. K: Chen. J: Luo. W. (2014).

The changes of miRNA expression in rat hippocampus following chronic lead exposure. Toxicol Lett 229:
158-166. http://dx.doi.Org/10.1016/i.toxlet.2014.06.002.

Anderson. DW: Mettil. W: Schneider. JS. (2016). Effects of low level lead exposure on associative learning and
memory in the rat: Influences of sex and developmental timing of exposure. Toxicol Lett 246: 57-64.
http://dx.doi.Org/10.1016/i.toxlet.2016.01.011.

Anderson. DW: Pothakos. K: Schneider. JS. (2012). Sex and rearing condition modify the effects of perinatal lead
exposure on learning and memory. Neurotoxicology 33: 985-995.
http://dx.doi.Org/10.1016/i.neuro.2012.04.016.

Andrade. V: Mateus. ML: Batoreu. MC: Aschner. M: Dos Santos. AM. (2017). Toxic mechanisms underlying motor
activity changes induced by a mixture of lead, arsenic and manganese. EC Pharmacol Toxicol 3: 31-42.

Andrew. A: Zhou. J: Gui. J: Harrison. A: Shi. X: Li. M: Guetti. B: Nathan. R: Tischbein. M: Pioro. E: Stommel. E:
Bradley. W. (2022). Airborne lead and poly chlorinated biphenyls (PCBs) are associated with amyotrophic
lateral sclerosis (ALS) risk in the U.S. Sci Total Environ 819: 153096.
http://dx.doi.Org/10.1016/i.scitotenv.2022.153096.

Anvan. J: Amir. S. (2018). Too depressed to swim or too afraid to stop? A reinterpretation of the forced swim test as
a measure of anxiety-like behavior [Comment]. Neuropsychopharmacology 43: 931-933.
http://dx.doi.org/10.1038/npp.2017.260.

APA (American Psychiatric Association). (2013). Diagnostic and statistical manual of mental disorders (5th ed.).
Arlington, VA. http://dx.doi.org/10.1176/appi.books.9780890425596.

Arbuckle. TE: Davis. K: Bovlan. K: Fisher. M: Fu. J. (2016a). Bisphenol A, phthalates and lead and learning and
behavioral problems in Canadian children 6-11 years of age: CHMS 2007-2009. Neurotoxicology 54: 89-
98. http://dx.doi.Org/10.1016/i.neuro.2016.03.014.

Arbuckle. TE: Davis. K: Bovlan. K: Fisher. M: Fu. J. (2016b). Processed data for CHMS 2007-2009: Bisphenol A,
phthalates and lead and learning and behavioral problems in Canadian children 6-19 years of age. Data in
Brief 8: 784-802. http://dx.doi.Org/10.1016/i.dib.2016.06.017.

Arora. M: Reichenberg. A: Willfors. C: Austin. C: Gennings. C: Berggren. S: Lichtenstein. P: Anckarsater. H:

Tammimies. K: Bolte. S. (2017). Fetal and postnatal metal dysregulation in autism. Nat Commun 8: 15493.
http://dx.doi.org/10.1038/ncommsl5493.

Ashok. A: Rai. NK: Tripathi. S: Bandvopadhvav. S. (2015). Exposure to As-, Cd-, and Pb-mixture induces A(3,
amyloidogenic APP processing and cognitive impairments via oxidative stress-dependent
neuroinflammation in young rats. Toxicol Sci 143: 64-80. http://dx.doi.org/10.1093/toxsci/kfu208.

ATSDR (Agency for Toxic Substances and Disease Registry). (2020). Toxicological profile for lead [ATSDR Tox
Profile]. Atlanta, GA: U.S. Department of Health and Human Services, Agency for Toxic Substances and
Disease Registry, https://www.atsdr.cdc.gov/toxprofiles/tp 13.pdf.

Baba. M: Nakaio. S: Tu. PH: Tomita. T: Nakava. K: Lee. VM: Troianowski. JO: Iwatsubo. T. (1998). Aggregation
of alpha-synuclein in Lewy bodies of sporadic Parkinson's disease and dementia with Lewy bodies. Am J
Pathol 152: 879-884.

Babinski. LM: Hartsough. CS: Lambert. NM. (1999). Childhood conduct problems, hyperactivity-impulsivity, and
inattention as predictors of adult criminal activity. J Child Psychol Psychiatry 40: 347-355.
http://dx.doi.org/10.1111/1469-7610.00452.

Baghurst. PA: McMichael. AJ: Wigg. NR: Vimpani. GV: Robertson. EF: Roberts. RJ: Tong. SL. (1992).

Environmental exposure to lead and children's intelligence at the age of seven years: The Port Pirie cohort
study. N Engl J Med 327: 1279-1284. http://dx.doi.org/10.1056/NEJM1992102932718Q5.

External Review Draft

3-494

DRAFT: Do not cite or quote

-------
Bah. HAF: Dos Anios. ALS: Gomes-Junior. EA: Bandeira. MJ: de Carvalho. C'F: Dos Santos. NR: Martinez. VQ:
Adorno. EV: Menezes-Filho. JA. (2022). Delta-aminolevulinic acid dehydratase, low blood lead levels,
social factors, and intellectual function in an Afro-Brazilian children community. Biol Trace Elem Res 200:
447-457. http://dx.doi.org/10.1007/sl2011-021-Q2656-8.

Balasundaram. P: Avulakunta. ID. (20211. Bayley scales of infant and toddler development [StatPearls]. Treasure
Island, FL: StatPearls Publishing. Retrieved from https://www.ncbi.nlm.nih.gov/books/NBK567715/

Baranowska-Bosiacka. I: Falkowska. A: Gutowska. I: Gassowska. M: Kolasa-Wolosiuk. A: Tarnowski. M:

Chibowska. K: Goschorska. M: Lubkowska. A: Chlubek. D. (2017). Glycogen metabolism in brain and
neurons - astrocytes metabolic cooperation can be altered by pre- and neonatal lead (Pb) exposure.
Toxicology 390: 146-158. http://dx.doi.Org/10.1016/i.tox.2017.09.007.

Baranowska-Bosiacka. I: Gutowska. I: Marchetti. C: Rutkowska. M: Marchlewicz. M: Kolasa. A: Prokopowicz. A:
Wiernicki. I: Piotrowska. K: Baskiewicz. M: Safranow. K: Wiszniewska. B: Chlubek. D. (2011). Altered
energy status of primary cerebellar granule neuronal cultures from rats exposed to lead in the pre- and
neonatal period. Toxicology 280: 24-32. http://dx.doi.Org/10.1016/i.tox.2010.ll.004.

Baranowska-Bosiacka. I: Struzvnska. L: Gutowska. I: Machalinska. A: Kolasa. A: Klos. P: Czapski. GA:

Kurzawski. M: Prokopowicz. A: Marchlewicz. M: Safranow. K: Machalinski. B: Wiszniewska. B:

Chlubek. D. (2013). Perinatal exposure to lead induces morphological, ultrastructural and molecular
alterations in the hippocampus. Toxicology 303: 187-200. http://dx.doi.Org/10.1016/i.tox.2012.10.027.

Barbeito. AG: Martinez-Palma. L: Vargas. MR: Pehar. M: Mafiav. N: Beckman. JS: Barbeito. L: Cassina. P. (2010).
Lead exposure stimulates VEGF expression in the spinal cord and extends survival in a mouse model of
ALS. Neurobiol Dis 37: 574-580. http://dx.doi.Org/10.1016/i.nbd.2009.ll.007.

Barg. G: Daleiro. M: Oueirolo. EI: Ravenscroft. J: Mafiav. N: Peregalli. F: Kordas. K. (2018). Association of low
lead levels with behavioral problems and executive function deficits in schoolers from Montevideo,
Uruguay. Int J Environ Res Public Health 15: 2735. http://dx.doi.org/10.3390/iierphl5122735.

Barkur. RR: Bairv. LK. (2015a). Assessment of oxidative stress in hippocampus, cerebellum and frontal cortex in rat
pups exposed to lead (Pb) during specific periods of initial brain development. Biol Trace Elem Res 164:
212-218. http://dx.doi.org/10.1007/sl2011-014-Q221-3.

Barkur. RR: Bairv. LK. (2015b). Evaluation of passive avoidance learning and spatial memory in rats exposed to

low levels of lead during specific periods of early brain development. Int J Occup Med Environ Health 28:
533-544. http://dx.doi.org/10.13075/iiomeh. 1896.00283.

Barkur. RR: Bairv. LK. (2016). Histological study on hippocampus, amygdala and cerebellum following low lead
exposure during prenatal and postnatal brain development in rats. Toxicol Ind Health 32: 1052-1063.
http://dx.doi.org/10.1177/0748233714545624.

Barkur. RR: Rao. MS: Bairv. LK. (2011). Low lead exposure during foetal and early postnatal life impairs passive
avoidance learning in adulthood in rats. Arh Hig Rada Toksikol 62: 147-153.
http://dx.doi.org/10.2478/10004-1254-62-2011-207Q.

Basha. CD: Reddv. RG. (2015). Long-term changes in brain cholinergic system and behavior in rats following
gestational exposure to lead: Protective effect of calcium supplement. Interdiscip Toxicol 8: 159-168.
http://dx.doi.org/10.1515/intox-2015-0025.

Basha. DC: Reddv. NS: Rani. MU: Reddv. GR. (2014). Age related changes in aminergic system and behavior
following lead exposure: Protection with essential metal supplements. Neurosci Res 78: 81-89.
http://dx.doi.Org/10.1016/i.neures.2013.09.007.

Basha. MR: Murali. M: Siddiqi. HK: Ghosal. K: Siddiqi. OK: Lashuel. HA: Ge. YW: Lahiri. DK: Zawia. NH.

(2005). Lead (Pb) exposure and its effect on APP proteolysis and A beta aggregation. FASEB J 19: 2083-
2084. http://dx.doi.org/10.1096/fi.Q5-4375fie.

Bathina. S: Das. UN. (2015). Brain-derived neurotrophic factor and its clinical implications. Archives of Medical
Science 11: 1164-1178. http://dx.doi.org/10.5114/aoms.2015.56342.

Baver. SA. (1989). Cellular aspects of brain development [Review]. Neurotoxicology 10: 307-320.

External Review Draft

3-495

DRAFT: Do not cite or quote

-------
Bavlev. N. (19691. Bayley scales of infant development: Manual. New York, NY: The Psychological Corporation.

Beaudin. SA: Stangle. DE: Smith. PR: Levitskv. DA: Strupp. BJ. (2007). Succimer chelation normalizes reactivity
to reward omission and errors in lead-exposed rats. Neurotoxicol Teratol 29: 188-202.
http://dx.doi.Org/10.1016/i.ntt.2006.ll.004.

Becklev. AL: Caspi. A: Broadbent. J: Harrington. H: Houts. RM: Poulton. R: Ramrakha. S: Reuben. A: Moffitt. TE.
(2018). Association of childhood blood lead levels with criminal offending. JAMA Pediatr 172: 166-173.
http://dx.doi.org/10.1001/iamapediatrics.2017.40Q5.

Beckwith. TJ: Dietrich. KN: Wright. JP: Altave. M: Cecil. KM. (2018). Reduced regional volumes associated with
total psychopathy scores in an adult population with childhood lead exposure. Neurotoxicology 67: 1-26.
http://dx.doi.Org/10.1016/i.neuro.2018.04.004.

Beckwith. TJ: Dietrich. KN: Wright. JP: Altave. M: Cecil. KM. (2021). Criminal arrests associated with reduced

regional brain volumes in an adult population with documented childhood lead exposure. Environ Res 201:
111559. http://dx.doi.Org/10.1016/i.envres.2021.111559.

Bellinger. D: Hu. H: Titlebaum. L: Needleman. HL. (1994a). Attentional correlates of dentin and bone lead levels in
adolescents. Arch Environ Health 49: 98-105. http://dx.doi.org/10.1080/00Q39896.1994.9937461.

Bellinger. D: Leviton. A: Allred. E: Rabinowitz. M. (1994b). Pre- and postnatal lead exposure and behavior
problems in school-aged children. Environ Res 66: 12-30. http://dx.doi.org/10.1006/enrs.1994.1041.

Bellinger. D: Leviton. A: Sloman. J. (1990). Antecedents and correlates of improved cognitive performance in
children exposed inutero to low levels of lead. Environ Health Perspect 89: 5-11.
http://dx.doi.org/10.2307/3430890.

Bellinger. D: Leviton. A: Waternaux. C: Needleman. H: Rabinowitz. M. (1987). Longitudinal analyses of prenatal
and postnatal lead exposure and early cognitive development. N Engl J Med 316: 1037-1043.
http://dx.doi.org/10.1056/NEJM198704233161701.

Bellinger. D: Needleman. HL. (2003). Intellectual impairment and blood lead levels [Letter], N Engl J Med 349:
500-502. http://dx.doi.org/10.1056/NEJM20030731349Q515.

Bellinger. D: Sloman. J: Leviton. A: Rabinowitz. M: Needleman. HL: Waternaux. C. (1991). Low-level lead
exposure and children's cognitive function in the preschool years. Pediatrics 87: 219-227.
http://dx.doi.Org/10.1542/peds.87.2.219.

Berk. M: Williams. LJ: Andreazza. AC: Pasco. JA: Dodd. S: Jacka. FN: Movlan. S: Reiner. EJ: Magalhaes. PVS.
(2014). Pop, heavy metal and the blues: Secondary analysis of persistent organic pollutants (POP), heavy
metals and depressive symptoms in the NHANES National Epidemiological Survey. BMJ Open 4:
e005142. http://dx.doi.org/10.1136/bmiopen-2014-005142.

Betharia. S: Maher. TJ. (2012). Neurobehavioral effects of lead and manganese individually and in combination in
developmentally exposed rats. Neurotoxicology 33: 1117-1127.
http://dx.doi.Org/10.1016/i.neuro.2012.06.002.

Bhattacharva. A: Shukla. R: Dietrich. K: Bornschein. R: Berger. O. (1995). Effect of early lead exposure on
children's postural balance. Dev Med Child Neurol 37: 861-878. http://dx.doi.org/10. Ill 1/i. 1469-
8749.1995.tb 11939.x.

Bhattacharva. A: Shukla. R: Dietrich. KN: Bornschein. RL. (2006). Effect of early lead exposure on the maturation
of children's postural balance: A longitudinal study. Neurotoxicol Teratol 28: 376-385.
http://dx.doi.Org/10.1016/i.ntt.2006.02.003.

Bhattacharvva. MH. (1983). Bioavailability of orally administered cadmium and lead to the mother, fetus, and
neonate during pregnancy and lactation: An overview. Sci Total Environ 28: 327-342.
http://dx.doi.org/10.1016/S0048-9697(83)80030-8.

Bihaai. SW: Zawia. NH. (2013). Enhanced taupathy and AD-like pathology in aged primate brains decades after

infantile exposure to lead (Pb). Neurotoxicology 39: 95-101. http://dx.doi.Org/10.1016/i.neuro.2013.07.010.

External Review Draft

3-496

DRAFT: Do not cite or quote

-------
Biioor. AR: Sudha. S: Venkatesh. T. (20121. Neurochemical and neurobehavioral effects of low lead exposure on
the developing brain. Indian J Clin Biochem 27: 147-151. http://dx.doi.org/10.1007/sl2291-012-019Q-2.

Blackowicz. MJ: Hrvhorczuk. DO: Rankin. KM: Lewis. DA: Haider. D: Lanphear. BP: Evens. A. (2016). The

impact of low-level lead toxicity on school performance among Hispanic subgroups in the Chicago Public
Schools. Int J Environ Res Public Health 13: 774. http://dx.doi.org/10.3390/iierphl3080774.

Blair. RJR. (20011. Neurocognitive models of aggression, the antisocial personality disorders, and psychopathy. J
Neurol Neurosurg Psychiatry 71: 727-731. http://dx.doi.Org/10.1136/innp.71.6.727.

Blaurock-Busch. E: Amin. OR: Rabah. T. (20111. Heavy metals and trace elements in hair and urine of a sample of
arab children with autistic spectrum disorder. Maedica 6: 247-257.

Bouchard. MF: Bellinger. DC: Weuve. J: Matthews-Bellinger. J: Gilman. SE: Wright. RO: Schwartz. J: Weisskopf.
MG. (2009). Blood lead levels and major depressive disorder, panic disorder, and generalized anxiety
disorder in US young adults. Arch Gen Psychiatry 66: 1313-1319.
http://dx.doi.org/10.1001/archgenpsvchiatrv.20Q9.164.

Boucher. O: Burdea MJ: Muckle. G: Saint-Amour. D: Avotte. P: Dewaillv. E: Nelson. CA: Jacobson. SW:

Jacobson. JL. (2012a). Response inhibition and error monitoring during a visual go/no-go task in Inuit
children exposed to lead, polychlorinated biphenyls, and methylmercury. Environ Health Perspect 120:
608-615. http://dx.doi.org/10.1289/ehp. 1103828.

Boucher. O: Jacobson. SW: Plusauellec. P: Dewaillv. E: Avotte. P: Forget-Dubois. N: Jacobson. JL: Muckle. G.

(2012b). Prenatal methylmercury, postnatal lead exposure, and evidence of attention deficit/hyperactivity
disorder among Inuit children in Arctic Quebec. Environ Health Perspect 120: 1456-1461.
http://dx.doi.org/10.1289/ehp.1204976.

Boucher. O: Muckle. G: Avotte. P: Dewaillv. E: Jacobson. SW: Jacobson. JL. (2016). Altered fine motor function at
school age in Inuit children exposed to PCBs, methylmercury, and lead. Environ Int 95: 144-151.
http://dx.doi.Org/10.1016/i.envint.2016.08.010.

Boutwell. BB: Nelson. EJ: Oian. Z: Vaughn. MG: Wright. JP: Beaver. KM: Barnes. JC: Petkovsek. M: Lewis. R:
Schootman. M: Rosenfeld. R. (2017). Aggregate-level lead exposure, gun violence, homicide, and rape.
PLoS ONE 12: e0187953. http://dx.doi.org/10.1371/iournal.pone.0187953.

Bozack. AK: Rifas-Shiman. SL: Coull. BA: Baccarelli. AA: Wright. RO: Amarasiriwardena. C: Gold. PR: Oken. E:
Hivert. MF: Cardenas. A. (2021). Prenatal metal exposure, cord blood DNA methylation and persistence in
childhood: An epigenome-wide association study of 12 metals. Clinical Epigenetics 13: 208.
http://dx.doi.org/10.1186/sl3148-021-01198-z.

Braun. JM: Froehlich. TE: Daniels. JL: Dietrich. KN: Hornung. R: Auinger. P: Lanphear. BP. (2008). Association of
environmental toxicants and conduct disorder in U.S. children: NHANES 2001-2004. Environ Health
Perspect 116: 956-962. http://dx.doi.org/10.1289/ehp. 11177.

Braun. JM: Hornung. R: Chea A: Dietrich. KN: Jacobs. DE: Jones. R: Khourv. JC: Liddv-Hicks. S: Morgan. S:
Vanderbeek. SB: Xu. Y: Yolton. K: Lanphear. BP. (2018). Effect of residential lead-hazard interventions
on childhood blood lead concentrations and neurobehavioral outcomes: A randomized clinical trial. JAMA
Pediatr 172: 934-942. http://dx.doi.org/10.1001/iamapediatrics.2018.2382.

Braun. JM: Kahn. RS: Froehlich. T: Auinger. P: Lanphear. BP. (2006). Exposures to environmental toxicants and
attention deficit hyperactivity disorder in U.S. children. Environ Health Perspect 114: 1904-1909.
http://dx.doi.org/10.1289/ehp.9478.

Brockel. BJ: Corv-Slechta. DA. (1998). Lead, attention, and impulsive behavior: Changes in a fixed-ratio waiting-
for-reward paradigm. Pharmacol Biochem Behav 60: 545-552. http://dx.doi.org/10.1016/S0Q91-
3057(98)00023-9.

Brockel. BJ: Corv-Slechta. DA. (1999). The effects of postweaning low-level Pb exposure on sustained attention: A
study of target densities, stimulus presentation rate, and stimulus predictability. Neurotoxicology 20: 921-
933.

External Review Draft

3-497

DRAFT: Do not cite or quote

-------
Burns. JM: Baghurst. PA: Sawyer. MG: McMichael. AJ: Tong. SL. (1999). Lifetime low-level exposure to

environmental lead and children's emotional and behavioral development at ages 11-13 years: The Port
Pirie cohort study. Am J Epidemiol 149: 740-749. http://dx.doi.org/10.1093/oxfordiournals.aie.a009883.

Cai. OL: Peng. DJ: Zhao. L: Chen. JW: Li. Y: Luo. HL: Ou. SY: Huang. ML: Jiang. YM. (2021). Impact of lead

exposure on thyroid status and iq performance among school-age children living nearby a lead-zinc mine in
China. Neurotoxicology 82: 177-185. http://dx.doi.Org/10.1016/i.neuro.2020.10.010.

Canfield. RL: Gendle. MH: Corv-Slechta. DA. (2004). Impaired neuropsychological functioning in lead-exposed
children. Dev Neuropsychol 26: 513-540. http://dx.doi.org/10.1207/sl5326942dn2601 8.

Canfield. RL: Henderson. CR. Jr: Corv-Slechta. DA: Cox. C: Jusko. TA: Lanphear. BP. (2003a). Intellectual

impairment in children with blood lead concentrations below 10 |ig per deciliter. N Engl J Med 348: 1517-
1526. http://dx.doi.org/10.1056/NEJMoa022848.

Canfield. RL: Kreher. DA: Cornwell. C: Henderson. CR. Jr. (2003b). Low-level lead exposure, executive
functioning, and learning in early childhood. Child Neuropsychol 9: 35-53.
http://dx.doi.Org/10.1076/chin.9.l.35.14496.

Cao. XJ: Huang. SH: Wang. M: Chen. JT: Ruan. DY. (2008). S-adenosyl-L-methionine improves impaired

hippocampal long-term potentiation and water maze performance induced by developmental lead exposure
in rats. Eur J Pharmacol 595: 30-34. http://dx.doi.Org/10.1016/i.eiphar.2008.07.061.

Carlson. K: Schacht. J: Neitzel. RL. (2018). Assessing ototoxicity due to chronic lead and cadmium intake with and
without noise exposure in the mature mouse. J Toxicol Environ Health A 81: 1041-1057.
http://dx.doi.org/10.1080/15287394.2018.152132Q.

Casiens. S: Pesch. B: van Thriel. C: Zschiesche. W: Behrens. T: Weiss. T: Pallapies. D: Arendt. M: Dragano. N:
Moebus. S: Jockel. KH: Briining. T. (2018). Associations between blood lead, olfaction and fine-motor
skills in elderly men: Results from the Heinz Nixdorf Recall Study. Neurotoxicology 68: 66-72.
http://dx.doi.Org/10.1016/i.neuro.2018.06.013.

Cecil. KM. (2011). Effects of early low-level lead exposure on human brain structure, organization and functions. J
Dev Orig Health Dis 2: 17-24. http://dx.doi.org/10.1017/S204017441000Q486.

Chan. TJH: Gutierrez. C: Qgunseitan. OA. (2015). Metallic burden of deciduous teeth and childhood behavioral
deficits. Int J Environ Res Public Health 12: 6771-6787. http://dx.doi.org/10.3390/iierphl20606771.

Chandramouli. K: Steer. CD: Ellis. M: Emond. AM. (2009). Effects of early childhood lead exposure on academic
performance and behaviour of school age children. Arch Dis Child 94: 844-848.
http://dx.doi.org/10.1136/adc.2008.149955.

Chen. A: Cai. B: Dietrich. KN: Radcliffe. J: Rogan. WJ. (2007). Lead exposure, IQ, and behavior in urban 5- to 7-
year-olds: Does lead affect behavior only by lowering IQ? Pediatrics 119: e650-e658.
http://dx.doi.org/10.1542/peds.2006-1973.

Chen. A: Dietrich. KN: Ware. JH: Radcliffe. J: Rogan. WJ. (2005). IQ and blood lead from 2 to 7 years of age: Are
the effects in older children the residual of high blood lead concentrations in 2-year-olds? Environ Health
Perspect 113: 597-601. http://dx.doi.org/10.1289/ehp.7625.

Chibowska. K: Korbecki. J: Gutowska. I: Metrvka. E: Tarnowski. M: Goschorska. M: Barczak. K: Chlubek. D:

Baranowska-Bosiacka. I. (2020). Pre- and neonatal exposure to lead (Pb) induces neuroinflammation in the
forebrain cortex, hippocampus and cerebellum of rat pups. International Journal of Molecular Sciences 21:
1083. http://dx.doi.org/10.3390/iims21031083.

Chichinadze. K: Chichinadze. N: Lazarashvili. A. (2011). Hormonal and neurochemical mechanisms of aggression
and a new classification of aggressive behavior. Aggression and Violent Behavior 16: 461-471.
http://dx.doi.Org/10.1016/i.avb.2011.03.002.

Chiodo. LM: Covington. C: Sokol. RJ: Hannigan. JH: Jannise. J: Ager. J: Greenwald. M: Delanev-Black. V. (2007).
Blood lead levels and specific attention effects in young children. Neurotoxicol Teratol 29: 538-546.
http://dx.doi.Org/10.1016/i.ntt.2007.04.001.

External Review Draft

3-498

DRAFT: Do not cite or quote

-------
Cho. SC: Kim. BN: Hong. YC: Shin. MS: Yoo. H.T: Kim. JW: Bhang. SY: Cho. TH: Kim. HW (2010). Effect of
environmental exposure to lead and tobacco smoke on inattentive and hyperactive symptoms and
neurocognitive performance in children. J Child Psychol Psychiatry 51: 1050-1057.
http://dx.doi.org/10.1111/i. 1469-7610.2010.02250.X.

Choi. JW: Jung. AH: Nam. S: Kim. K: Kim. J: Kim. S: Kim. BN: Kim. JW. (2020). Interaction between lead and

noradrenergic genotypes affects neurocognitive functions in attention-deficit/hyperactivity disorder: A case
control study. BMC Psychiatry 20: 407. http://dx.doi.org/10.1186/sl2888-020-02799-3.

Choi. WJ: Kwon. HJ: Lim. MH: Lim. JA: Ha. M. (2016). Blood lead, parental marital status and the risk of

attention-deficit/hyperactivity disorder in elementary school children: A longitudinal study. Psychiatry Res
236: 42-46. http://dx.doi.Org/10.1016/i.psvchres.2016.01.002.

Choi. YH: Hu. H: Mukheriee. B: Miller. J: Park. SK. (2012). Environmental cadmium and lead exposures and

hearing loss in U.S. adults: The National Health and Nutrition Examination Survey, 1999 to 2004. Environ
Health Perspect 120: 1544-1550. http://dx.doi.org/10.1289/ehp. 1104863.

Choi. YH: Park. SK. (2017). Environmental exposures to lead, mercury, and cadmium and hearing loss in adults and
adolescents: KNHANES 2010-2012. Environ Health Perspect 125: 067003.
http://dx.doi.org/10.1289/EHP565.

Constantino. JN. (2011). The quantitative nature of autistic social impairment [Review]. PediatrRes 69: 55R-62R.
http://dx.doi.org/10.12Q3/PDR.0b013e318212ec6e.

Cooper. GP: Manalis. RS. (1984). Interactions of lead and cadmium on acetylcholine release at the frog
neuromuscular junction. Toxicol Appl Pharmacol 74: 411-416. http://dx.doi.org/10.1016/0Q41-
008X(84)90294-1.

Cooper. GS: Lunn. RM: Agerstrand. M: Glenn. BS: Kraft. AD: Luke. AM: Ratcliffe. JM. (2016). Study sensitivity:
Evaluating the ability to detect effects in systematic reviews of chemical exposures. Environ Int 92-93:
605-610. http://dx.doi.Org/10.1016/i.envint.2016.03.017.

Corv-Slechta. DA: Merchant-Borna. K: Allen. JL: Liu. S: Weston. D: Conrad. K. (2013). Variations in the nature of
behavioral experience can differentially alter the consequences of developmental exposures to lead,
prenatal stress, and the combination. Toxicol Sci 131: 194-205. http://dx.doi.org/10.1093/toxsci/kfs260.

Corv-Slechta. DA: Virgolini. MB: Liu. S: Weston. D. (2012). Enhanced stimulus sequence-dependent repeated

learning in male offspring after prenatal stress alone or in conjunction with lead exposure. Neurotoxicology
33: 1188-1202. http://dx.doi.Org/10.1016/i.neuro.2012.06.013.

Counter. SA: Buchanan. LH: Ortega. F. (2008). Zinc protoporphyrin levels, blood lead levels and neurocognitive
deficits in Andean children with chronic lead exposure. ClinBiochem 41: 41-47.
http://dx.doi.org/10.1016/i.clinbiochem.2007.10.002.

Covle. JT: Tsai. G. (2004). The NMD A receptor glycine modulatory site: A therapeutic target for improving

cognition and reducing negative symptoms in schizophrenia [Review]. Psychopharmacology 174: 32-38.
http://dx.doi.org/10.1007/s00213-003-17Q9-2.

Crump. KS: Van Landingham. C: Bowers. TS: Cahov. D: Chandalia. JK. (2013). A statistical reevaluation of the
data used in the Lanphear et al. (2005) pooled-analysis that related low levels of blood lead to intellectual
deficits in children [Review]. Crit Rev Toxicol 43: 785-799.
http://dx.doi.org/10.3109/10408444.2Q13.832726.

da Silva. DRF: Bittencourt. LO: Aragao. WAB: Nascimento. PC: Leao. LKR: Oliveira. ACA: Crespo-Lopez. ME:
Lima. RR. (2020). Long-term exposure to lead reduces antioxidant capacity and triggers motor neurons
degeneration and demyelination in spinal cord of adult rats. Ecotoxicol Environ Saf 194: 110358.
http://dx.doi.Org/10.1016/i.ecoenv.2020.110358.

Dabrowska. A: Luis Venero. J: Iwasawa. R: Hankir. MK: Rahman. S: Boobis. A: Haiii. N. (2015). PGC-la controls
mitochondrial biogenesis and dynamics in lead-induced neurotoxicity. Aging 7: 629-647.
http://dx.doi.org/10.18632/aging.10079Q.

External Review Draft

3-499

DRAFT: Do not cite or quote

-------
Dantzer. J: Ryan. P: Yolton. K: Parsons. PJ: Palmer. CD: Cecil. K: Unrine. JM. (2020). A comparison of blood and
toenails as biomarkers of children's exposure to lead and their correlation with cognitive function. Sci Total
Environ 700: 134519. http://dx.doi.Org/10.1016/i.scitotenv.2019.134519.

David. OJ: Hoffman. SP: Sverd. J: Clark. J: Voeller. K. (1976). Lead and hyperactivity. Behavorial response to
chelation: A pilot study. Am J Psychiatry 133: 1155-1158. http://dx.doi.org/10.1176/aip.133.10.1155.

De Marco. M: Halpern. R: Barros. HMT. (2005). Early behavioral effects of lead perinatal exposure in rat pups.
Toxicology 211: 49-58. http://dx.doi.Org/10.1016/i.tox.2005.02.007.

De Palma. G: Catalan! S: Franco. A: Brighenti. M: Apostoli. P. (2012). Lack of correlation between metallic
elements analyzed in hair by ICP-MS and autism. J Autism Dev Disord 42: 342-353.
http://dx.doi.org/10.1007/sl0803-011-1245-6.

Despres. C: Beuter. A: Richer. F: Poitras. K: Veilleux. A: Avotte. P: Dewaillv. E: Saint-Amour. D: Muckle. G.

(2005).	Neuromotor functions in Inuit preschool children exposed to Pb, PCBs, and Hg. Neurotoxicol
Teratol 27: 245-257. http://dx.doi.Org/10.1016/i.ntt.2004.12.001.

Desrochers-Couture. M: Courtemanche. Y: Forget-Dubois. N: Belanger. RE: Boucher. O: Avotte. P: Cordier. S:
Jacobson. JL: Jacobson. SW: Muckle. G. (2019). Association between early lead exposure and
externalizing behaviors in adolescence: A developmental cascade. Environ Res 178: 108679.
http://dx.doi.Org/10.1016/i.envres.2019.108679.

Desrochers-Couture. M: Oulhote. Y: Arbuckle. TE: Fraser. WD: Seguin. JR: Quelle! E: Forget-Dubois. N: Avotte.
P: Boivin. M: Lanphear. BP: Muckle. G. (2018). Prenatal, concurrent, and sex-specific associations
between blood lead concentrations and IQ in preschool Canadian children. Environ Int 121: 1235-1242.
http://dx.doi.Org/10.1016/i.envint.2018.10.043.

Pi Paolo. G: Kim. TW. (2011). Linking lipids to Alzheimer's disease: Cholesterol and beyond [Review]. Nat Rev
Neurosci 12: 284-296. http://dx.doi.org/10.1038/nrn3012.

Dietrich. KN: Berger. OG: Succop. PA. (1993). Lead exposure and the motor developmental status of urban six-
year-old children in the Cincinnati Prospective Study. Pediatrics 91: 301-307.
http://dx.doi.Org/10.1542/peds.91.2.301.

Dietrich. KN: Douglas. RM: Succop. PA: Berger. OG: Bornschein. RL. (2001). Early exposure to lead and juvenile
delinquency. Neurotoxicol Teratol 23: 511-518. http://dx.doi.org/10.1016/S0892-0362(01)00184-2.

Dietrich. KN: Krafft. KM: Shukla. R: Bornschein. RL: Succop. PA. (1987). The neurobehavioral effects of early
lead exposure. In SR Schroeder (Ed.), Toxic substances and mental retardation: Neurobehavioral
toxicology and teratology (pp. 71-95). Washington, DC: American Association on Mental Deficiency.

Dietrich. KN: Succop. PA: Berger. OG: Hammond. PB: Bornscheia RL. (1991). Lead exposure and the cognitive
development of urban preschool children: The Cincinnati lead study cohort at age 4 years. Neurotoxicol
Teratol 13: 203-211. http://dx.doi.org/10.1016/0892-0362(91)90012-L.

Dietrich. KN: Succop. PA: Berger. OG: Keith. RW. (1992). Lead exposure and the central auditory processing
abilities and cognitive development of urban children: The Cincinnati lead study cohort at age 5 years.
Neurotoxicol Teratol 14: 51-56. http://dx.doi.org/10.1016/0892-0362(92)90028-9.

Diouf. A: Garcon. G: Diop. Y: Ndiave. B: Thiaw. C: Fall. M: Kane-Barry. O: Ba. D: Haguenoer. JM: Shirali. P.

(2006).	Environmental lead exposure and its relationship to traffic density among Senegalese children: A
cross-sectional study. Hum Exp Toxicol 25: 637-644. http://dx.doi.org/10.1177/0960327106Q74591.

Dohertv. BT: Romano. ME: Gui. J: Punshon. T: Jackson. BP: Karagas. MR: Korrick. SA. (2020). Periconceptional
and prenatal exposure to metal mixtures in relation to behavioral development at 3 years of age.
Environmental Epidemiology 4: e0106. http://dx.doi.org/10.1097/EE9.00000000000001Q6.

Dominguez. S: Flores-Montova. MG: Sobin. C. (2019). Early chronic exposure to low-level lead alters total
hippocampal microglia in pre-adolescent mice. Toxicol Lett 302: 75-82.
http://dx.doi.Org/10.1016/i.toxlet.2018.10.016.

External Review Draft

3-500

DRAFT: Do not cite or quote

-------
Donald. JM: Cutler. MG: Moore. MR. (19861. Effects of lead in the laboratory mouse: 1. Influence of pregnancy
upon absorption, retention, and tissue distribution of radiolabeled lead. Environ Res 41: 420-431.
http://dx.doi.org/10.1016/S0013-9351(86180136-0.

Donald. JM: Cutler. MG: Moore. MR. (19871. Effects of lead in the laboratory mouse. Development and social
behaviour after lifelong exposure to 12 (iM lead in drinking fluid. Neuropharmacology 26: 391-399.
http://dx.doi.org/10.1016/0028-3908(87190194-8.

Dong. HY: Feng. JY: Li. HH: Yue. XJ: Jia. FY. (2022). Non-parental caregivers, low maternal education,

gastrointestinal problems and high blood lead level: Predictors related to the severity of autism spectrum
disorder in Northeast China. BMC Pediatr 22: 11. http://dx.doi.org/10.1186/sl2887-021-03086-0.

Du. Y: Ge. MM: Xue. W: Yang. OO: Wang. S: Xu. Y: Wang. HL. (2015). Chronic lead exposure and mixed factors
of genderxagexbrain regions interactions on dendrite growth, spine maturity and NDR kinase. PLoS ONE
10: e0138112. http://dx.doi.org/10.1371/iournal.pone.0138112.

Duaa Y: Peng. L: Shi. H: Jiang. Y. (2017). The effects of lead on GAB Aergic interneurons in rodents. Toxicol Ind
Health 33: 867-875. http://dx.doi.org/10.1177/0748233717732902.

Dumkova. J: Smutna. T: Vrlikova. L: Le Coustumer. P: Vecera. Z: Docekal. B: Mikuska. P: Capka. L: Fictum. P:
Hampl. A: Buchtova. M. (2017). Sub-chronic inhalation of lead oxide nanoparticles revealed their broad
distribution and tissue-specific subcellular localization in target organs. Part Fibre Toxicol 14:55.
http://dx.doi.org/10.1186/sl2989-017-0236-v.

Ebrahimzadeh-Bideskan. AR: Hami. J: Alipour. F: Haghir. H: Fazel. AR: Sadeghi. A. (2016). Protective effects of
ascorbic acid and garlic extract against lead-induced apoptosis in developing rat hippocampus. Metab Brain
Dis 31: 1123-1132. http://dx.doi.org/10.1007/sl 1011-016-9837-7.

Egan. KB: Cornwell. CR: Courtney. JG: Ettinger. AS. (2021). Blood lead levels in U.S. children ages 1-11 years,
1976-2016. Environ Health Perspect 129: 37003. http://dx.doi.org/10.1289/EHP7932.

Ercal. N: Treeratphan. P: Hammond. TC: Matthews. RH: Grannemann. NH: Spitz. DR. (1996). In vivo indices of

oxidative stress in lead-exposed C57BL/6 mice are reduced by treatment with meso-2,3-dimercaptosuccinic
acid orN-acetylcysteine. Free Radic Biol Med 21: 157-161. http://dx.doi.org/10.1016/Q891-
5849(96)00020-2.

Ethier. AA: Muckle. G: Bastien. C: Dewaillv. E: Avotte. P: Arfken. C: Jacobson. SW: Jacobson. JL: Saint-Amour.
D. (2012). Effects of environmental contaminant exposure on visual brain development: A prospective
electrophysiological study in school-aged children. Neurotoxicology 33: 1075-1085.
http://dx.doi.Org/10.1016/i.neuro.2012.05.010.

Ethier. AA: Muckle. G: Jacobson. SW: Avotte. P: Jacobson. JL: Saint-Amour. D. (2015). Assessing new dimensions
of attentional functions in children prenatally exposed to environmental contaminants using an adapted
Posner paradigm. Neurotoxicol Teratol 51: 27-34. http://dx.doi.Org/10.1016/i.ntt.2015.07.005.

Eum. KD: Korrick. SA: Weuve. J: Okereke. O: Kubzanskv. LP: Hu. H: Weisskopf. MG. (2012). Relation of

cumulative low-level lead exposure to depressive and phobic anxiety symptom scores in middle-age and
elderly women. Environ Health Perspect 120: 817-823. http://dx.doi.org/10.1289/ehp. 1104395.

Evens. A: Hrvhorczuk. D: Lanphear. BP: Rankin. KM: Lewis. DA: Forst. L: Rosenberg. D. (2015). The impact of
low-level lead toxicity on school performance among children in the Chicago Public Schools: A
population-based retrospective cohort study. Environ Health 14: 21. http://dx.doi.org/10.1186/sl2940-Q15-
0008-9.

Faa G: Feng. C: Li. Y: Wang. CH: Yan. J: Li. W: Feng. JG: Shi. XL: Bi. YY. (2009). Selection of nutrients for
prevention or amelioration of lead-induced learning and memory impairment in rats. Ann Occup Hyg 53:
341-351. http://dx.doi.org/10.1093/annhvg/mepO 19.

Faa G: Feng. C: Wu. F: Ye. W: Lin. F: Wang. C: Yan. J: Zhu. G: Xiao. Y: Bi. Y. (2010). Methionine choline

reverses lead-induced cognitive and N-methyl-d-aspartate receptor subunit 1 deficits. Toxicology 272: 23-
31. http://dx.doi.Org/10.1016/i.tox.2010.03.018.

External Review Draft

3-501

DRAFT: Do not cite or quote

-------
Fan. Y: Sheng. J: Liang. C: Yang. L: Liu. K: Wang. O: Zhang. D: Ma. Y: Li. X: Xie. S: Cao. H: Wang. S: Tao. F.
(2020). Association of blood lead levels with the risk of depressive symptoms in the elderly Chinese
population: Baseline data of a cohort study. Biol Trace Elem Res 194: 76-83.
http://dx.doi.org/10.1007/sl2011-019-Q1755-x.

Fang. F: Kwee. LC: Allen. KD: Umbach. DM: Ye. W: Watson. M: Keller. J: Oddone. EZ: Sandler. DP: Schmidt. S:
Kamel. F. (2010). Association between blood lead and the risk of amyotrophic lateral sclerosis. Am J
Epidemiol 171: 1126-1133. http://dx.doi.org/10.1093/aie/kwa063.

Fang. F: Peters. TL: Beard. JD: Umbach. DM: Keller. J: Mariosa. D: Allen. KD: Ye. W: Sandler. DP: Schmidt. S:

Kamel. F. (2017). Blood lead, bone turnover, and survival in amyotrophic lateral sclerosis. Am J Epidemiol
186: 1057-1064. http://dx.doi.org/10.1093/aie/kwxl76.

Farooaui. Z: Bakulski. KM: Power. MC: Weisskopf. MG: Sparrow. D: Spiro. A. Ill: Vokonas. PS: Nie. LH: Hu. H:
Park. SK. (2017). Associations of cumulative Pb exposure and longitudinal changes in Mini-Mental Status
Exam scores, global cognition and domains of cognition: The VA Normative Aging Study. Environ Res
152: 102-108. http://dx.doi.Org/10.1016/i.envres.2016.10.007.

Faulk. C: Barks. A: Sanchez. BN: Zhang. ZZ: Anderson. OS: Peterson. KE: Dolinov. DC. (2014). Perinatal lead
(Pb) exposure results in sex-specific effects on food intake, fat, weight, and insulin response across the
murine life-course. PLoS ONE 9: el04273. http://dx.doi.org/10.1371/iournal.pone.0104273.

Feng. C: Liu. S: Zhou. F: Gao. Y: Li. Y: Du. G: Chen. Y: Jiao. H: Feng. J: Zhang. Y: Bo. D: Li. Z: Fan. G. (2019).
Oxidative stress in the neurodegenerative brain following lifetime exposure to lead in rats: Changes in
lifespan profiles. Toxicology 411: 101-109. http://dx.doi.Org/10.1016/i.tox.2018.ll.003.

Fergussoa DM: Boden. JM: Horwood. LJ. (2008). Dentine lead levels in childhood and criminal behaviour in late
adolescence and early adulthood. J Epidemiol Community Health 62: 1045-1050.
http://dx.doi.org/10.1136/iech.2007.072827.

Fergussoa DM: Horwood. LJ: Lvnskev. MT. (1993). Early dentine lead levels and subsequent cognitive and
behavioural development. J Child Psychol Psychiatry 34: 215-227. http://dx.doi.org/10. Ill 1/i. 1469-
7610.1993.tb00980.x.

Fergussoa DM: Horwood. LJ: Lvnskev. MT. (1997). Early dentine lead levels and educational outcomes at 18
years. J Child Psychol Psychiatry 38: 471-478. http://dx.doi.org/10.Ill 1/i. 1469-7610.1997.tb01532.x.

Ferlemi. AY: Avgoustatos. D: Kokkosis. AG: Protonotarios. V: Constantinou. C: Margaritv. M. (2014). Lead-
induced effects on learning/memory and fear/anxiety are correlated with disturbances in specific
cholinesterase isoform activity and redox imbalance in adult brain. Physiol Behav 131: 115-122.
http://dx.doi.Org/10.1016/i.phvsbeh.2014.04.033.

Fillion. M: Lemire. M: Philibert. A: Frenette. B: Weiler. HA: Davee Guimaraes. JR: Larribe. F: Barbosa. F. Jr:

Mergler. D. (2013). Toxic risks and nutritional benefits of traditional diet on near visual contrast sensitivity
and color vision in the Brazilian Amazon. Neurotoxicology 37: 173-181.
http://dx.doi.Org/10.1016/i.neuro.2013.04.010.

Filon. J: Ustvmowicz-Farbiszewska. J: Kraiewska-Kulak. E. (2020). Analysis of lead, arsenic and calcium content in
the hair of children with autism spectrum disorder. BMC Public Health 20: 383.
http://dx.doi.org/10.1186/sl2889-020-08496-w.

Flora. SJ: Gautam. P: Kushwaha. P. (2012). Lead and ethanol co-exposure lead to blood oxidative stress and

subsequent neuronal apoptosis in rats. Alcohol Alcohol 47: 92-101. http://dx.doi.org/10.1093/alcalc/agrl52.

Flores-Montova. MG: Alvarez. JM: Sobin. C. (2015). Olfactory recognition memory is disrupted in young mice
with chronic low-level lead exposure. Toxicol Lett 236: 69-74.
http://dx.doi.Org/10.1016/i.toxlet.2015.04.013.

Flores-Montova. MG: Sobin. C. (2015). Early chronic lead exposure reduces exploratory activity in young
C57BL/6J mice. J Appl Toxicol 35: 759-765. http://dx.doi.org/10.1002/iat.3064.

External Review Draft

3-502

DRAFT: Do not cite or quote

-------
Forns. J: Fort. M: Casas. M: Caceres. A: Guxens. M: Gascoa M: Garcia-Esteban. R: Julvez. J: Grimalt. JO: Sunver.
J. (2014). Exposure to metals during pregnancy and neuropsychological development at the age of 4 years.
Neurotoxicology 40: 16-22. http://dx.doi.Org/10.1016/i.neuro.2013.10.006.

Fox. DA: Campbell. ML: Blocker. YS. (19971. Functional alterations and apoptotic cell death in the retina following
developmental or adult lead exposure. Neurotoxicology 18: 645-664.

Fox. DA: Kala. SV: Hamilton. WR: Johnson. JE: O'Callaghan. JP. (2008). Low-level human equivalent gestational
lead exposure produces supernormal scotopic electroretinograms, increased retinal neurogenesis, and
decreased retinal dopamine utilization in rats. Environ Health Perspect 116: 618-625.
http://dx.doi.org/10.1289/ehp. 11268.

Fox. DA: Qpanashuk. L: Zharkovskv. A: Weiss. B. (2010). Gene-chemical interactions in the developing

mammalian nervous system: Effects on proliferation, neurogenesis and differentiation. Neurotoxicology
31: 589-597. http://dx.doi.Org/10.1016/i.neuro.2010.03.007.

Frank. AC: Huang. S: Zhou. M: Gdalvahu. A: Kastellakis. G: Silva. TK: Lu. E: Wen. X: Poirazi. P: Trachtenberg.
JT: Silva. AJ. (2018). Hotspots of dendritic spine turnover facilitate clustered spine addition and learning
and memory. Nat Commun 9: 422. http://dx.doi.org/10.1038/s41467-017-02751-2.

Froehlich. TE: Lanphear. BP: Auinger. P: Hornung. R: Epstein. JN: Braun. J: Kahn. RS. (2009). Association of
tobacco and lead exposures with attention-deficit/hyperactivity disorder. Pediatrics 124: E1054-E1063.
http://dx.doi.org/10.1542/peds.2009-0738.

Froehlich. TE: Lanphear. BP: Dietrich. KN: Corv-Slechta. DA: Wang. N: Kahn. RS. (2007). Interactive effects of a
DRD4 polymorphism, lead and sex on executive functions in children. Biol Psychiatry 62: 243-249.
http://dx.doi.Org/10.1016/i.biopsvch.2006.09.039.

Fruh. V: Rifas-Shiman. SL: Amarasiriwardena. C: Cardenas. A: Bellinger. DC: Wise. LA: White. RF: Wright. RO:
Okea E: Henn. BC. (2019). Prenatal lead exposure and childhood executive function and behavioral
difficulties in project viva. Neurotoxicology 75: 105-115. http://dx.doi.Org/10.1016/i.neuro.2019.09.006.

Fruh. V: Rifas-Shiman. SL: Coull. BA: Devick. KL: Amarasiriwardena. C: Cardenas. A: Bellinger. DC: Wise. LA:
White. RF: Wright. RO: Oken. E: Henn. BC. (2021). Prenatal exposure to a mixture of elements and
neurobehavioral outcomes in mid-childhood: Results from Project Viva. Environ Res 201: 111540.
http://dx.doi.org/10.1016/i.envres.2021.111540.

Galal. MK: Elleithv. EMM: Abdrabou. MI: Yasin. NAE: Shaheen. YM. (2019). Modulation of caspase-3 gene
expression and protective effects of garlic and spirulina against CNS neurotoxicity induced by lead
exposure in male rats. Neurotoxicology 72: 15-28. http://dx.doi.Org/10.1016/i.neuro.2019.01.006.

Gassowska. M: Baranowska-Bosiacka. I: Moczvdlowska. J: Frontczak-Baniewicz. M: Gewartowska. M:

Struzvnska. L: Gutowska. I: Chlubek. D: Adamczvk. A. (2016a). Perinatal exposure to lead (Pb) induces
ultrastructural and molecular alterations in synapses of rat offspring. Toxicology 373: 13-29.
http://dx.doi.Org/10.1016/i.tox.2016.10.014.

Gassowska. M: Baranowska-Bosiacka. I: Moczvdlowska. J: Tarnowski. M: Pilutin. A: Gutowska. I: Struzvnska. L:
Chlubek. D: Adamczvk. A. (2016b). Perinatal exposure to lead (Pb) promotes Tau phosphorylation in the
rat brain in a GSK-3(3 and CDK5 dependent manner: Relevance to neurological disorders. Toxicology 347-
349: 17-28. http://dx.doi.Org/10.1016/i.tox.2016.03.002.

GBP 2019 Mental Disorders Collaborators. (2022). Global, regional, and national burden of 12 mental disorders in
204 countries and territories, 1990-2019: A systematic analysis for the Global Burden of Disease Study
2019. Lancet Psychiatry 9: 137-150. http://dx.doi.org/10.1016/S2215-0366(21)00395-3.

Geier. DA: Kern. JK: Geier. MR. (2018). A cross-sectional study of the relationship between blood lead levels and
reported attention deficit disorder: An assessment of the economic impact on the United States. Metab
BrainDis 33: 201-208. http://dx.doi.org/10.1007/sll011-017-Q146-6.

Gerber. AJ: Peterson. BS: Giedd. JN: Lalonde. FM: Celano. MJ: White. SL: Wallace. GL: Lee. NR: Lenroot. RK.
(2009). Anatomical brain magnetic resonance imaging of typically developing children and adolescents
[Review]. J Am Acad Child Adolesc Psychiatry 48: 465-470.
http://dx.doi.org/10.1097/CHI.0b013e31819f2715.

External Review Draft

3-503

DRAFT: Do not cite or quote

-------
Gersoa AC: Butler. R: Moxev-Mims. M: Wentz. A: Shinnar. S: Lande. MB: Mendlev. SR: Waradv. BA: Furth. SL:
Hooper. SR. (2006). Neurocognitive outcomes in children with chronic kidney disease: Current findings
and contemporary endeavors [Review]. Ment Retard Dev Disabil Res Rev 12: 208-215.
http://dx.doi.org/10.1002/mrdd.20116.

Giddabasappa. A: Hamilton. WR: Chanev. S: Xiao. W: Johnson. JE: Mukheriee. S: Fox. DA. (2011). Low-level
gestational lead exposure increases retinal progenitor cell proliferation and rod photoreceptor and bipolar
cell neurogenesis in mice. Environ Health Perspect 119: 71-77. http://dx.doi.org/10.1289/ehp. 1002524.

Gilbert. SG: Rice. DC. (1987). Low-level lifetime lead exposure produces behavioral toxicity (spatial discrimination
reversal) in adult monkeys. Toxicol Appl Pharmacol 91: 484-490. http://dx.doi.org/10.1016/0Q41-
008X(87)90070-6.

Gioia. GA: Isauith. PK: Retzlaff. PD: Espy. KA. (2002). Confirmatory factor analysis of the Behavior Rating
Inventory of Executive Function (BRIEF) in a clinical sample. Child Neuropsychol 8: 249-257.
http://dx.doi.org/10.1076/chin.8.4.249.13513.

Gipson. PS: Wetherington. CE: Duquette. PJ: Hooper. SR. (2004). The nervous system and chronic kidney disease
in children [Review]. Pediatr Nephrol 19: 832-839. http://dx.doi.org/10.1007/s00467-004-1532-v.

Gittleman. R: Eskenazi. B. (1983). Lead and hyperactivity revisited: An investigation of non-disadvantaged

children. Arch Gen Psychiatry 40: 827-833. http://dx.doi.org/10.1001/archpsvc.1983.017900700170Q2.

Glenn. AL: Johnson. AK: Raine. A. (2013). Antisocial personality disorder: a current review [Review]. Curr
Psychiatry Rep 15: 427. http://dx.doi.org/10.1007/sll920-013-0427-7.

Gomaa. A: Hu. H: Bellinger. D: Schwartz. J: Tsaih. SW: Gonzalez-Cossio. T: Schnaas. L: Peterson. K: Aro. A:
Hernandez-Avila. M. (2002). Maternal bone lead as an independent risk factor for fetal neurotoxicity: A
prospective study. Pediatrics 110: 110-118. http://dx.doi.org/10.1542/peds. 110.1.110.

Gottipolu. RR: Davuliigari. CB. (2014). Perinatal exposure to lead: Reduction in alterations of brain mitochondrial
antioxidant system with calcium supplement. Biol Trace Elem Res 162: 270-277.
http://dx.doi.org/10.1007/sl2011-014-0112-7.

Graff-Radford. NR: Crook. JE: Lucas. J: Boeve. BF: Knopman. PS: Ivnik. RJ: Smith. GE: Younkin. LH: Petersen.
RC: Younkia SG. (2007). Association of low plasma A beta 42/A beta 40 ratios with increased imminent
risk for mild cognitive impairment and Alzheimer disease. Arch Neurol 64: 354-362.
http://dx.doi.Org/10.1001/archneur.64.3.354.

Graham. PL: Grace. CE: Braun. AA: Schaefer. TL: Skelton. MR: Tang. PH: Vorhees. CV: Williams. MT. (2011).
Effects of developmental stress and lead (Pb) on corticosterone after chronic and acute stress, brain
monoamines, and blood Pb levels in rats. Int J Pev Neurosci 29: 45-55.
http://dx.doi.Org/10.1016/i.iidevneu.2010.09.008.

Grashow. R: Spiro. A: Taylor. KM: Newton. K: Shrairman. R: Landau. A: Sparrow. P: Hu. H: Weisskopf. M.
(2013). Cumulative lead exposure in community-dwelling adults and fine motor function: Comparing
standard and novel tasks in the VA Normative Aging Study. Neurotoxicology 35: 154-161.
http://dx.doi.Org/10.1016/i.neuro.2013.01.005.

Graves. AB: VanPuiin. CM: Chandra. V: Fratiglioni. L: Hevman. A: Jorm. AF: Kokmen. E: Kondo. K: Mortimer.
JA: Rocca. WA: Shalat. SL: Soininen. H: Hofman. A. (1991). Occupational exposures to solvents and lead
as risk factors for Alzheimer's disease: A collaborative re-analysis of case-control studies. Int J Epidemiol
2: S58-S61. http://dx.doi.org/10.1093/iie/20.Supplement 2.S58.

Gu. H: Robison. G: Hong. L: Barrea. R: Wei. X: Farlow. MR: Pushkar. YN: Pu. Y: Zheng. W. ei. (2012). Increased
(3-amyloid deposition in Tg-SWPI transgenic mouse brain following in vivo lead exposure. Toxicol Lett
213: 211-219. http://dx.doi.Org/10.1016/i.toxlet.2012.07.002.

Gu. X: Yuan. FF: Huang. X: Hou. YW: Wang. M: Lin. J: Wu. J. (2018). Association of PIK3CG gene
polymorphisms with attention-deficit/hyperactivity disorder: A case-control study. Prog
Neuropsychopharmacol Biol Psychiatry 81: 169-177. http://dx.doi.Org/10.1016/i.pnpbp.2017.10.020.

External Review Praft

3-504

PRAFT: Po not cite or quote

-------
Guilarte. TR. (19971. Pb2+ inhibits NMDA receptor function at high and low affinity sites: Developmental and
regional brain expression. Neurotoxicology 18: 43-51.

Gulson. BL: Mizon. KJ: Korsch. MJ: Palmer. JM: Donnelly. JB. (2003). Mobilization of lead from human bone

tissue during pregnancy and lactation - A summary of long-term research. Sci Total Environ 303: 79-104.
http://dx.doi.org/10.1016/80048-9697(02100355-8.

Gump. BB: Dvkas. MJ: MacKenzie. JA: Dumas. AK: Hruska. B: Ewart. CK: Parsons. PJ: Palmer. CD: Bendinskas.
K. (2017). Background lead and mercury exposures: Psychological and behavioral problems in children.
Environ Res 158: 576-582. http://dx.doi.Org/10.1016/i.envres.2017.06.033.

Haden. SC: Scarpa. A. (2007). The noradrenergic system and its involvement in aggressive behaviors. Aggression
and Violent Behavior 12: 1-15. http://dx.doi.Org/10.1016/i.avb.2006.01.012.

Han. XJ: Xiao. YM: Ai. BM: Hu. XX: Wei. O: Hu. OS. (2014). Effects of organic selenium on lead-induced

impairments of spatial learning and memory as well as synaptic structural plasticity in rats. Biol Pharm Bull
37: 466-474. http://dx.doi.org/10.1248/bpb.bl3-00892.

Haraguchi. T: Ishizu. H: Takehisa. Y: Kawai. K: Yokota. O: Terada. S: Tsuchiva. K: Ikeda. K: Morita. K: Horike. T:
Kira. S: Kuroda. S. (2001). Lead content of brain tissue in diffuse neurofibrillary tangles with calcification
(DNTC): The possibility of lead neurotoxicity. Neuroreport 12: 3887-3890.
http://dx.doi.org/10.1097/00001756-200112210-000Q6.

Hashemzadeh-Gargari. H: Guilarte. TR. (1999). Divalent cations modulate N-methyl-D-aspartate receptor function
at the glycine site. J Pharmacol Exp Ther 290: 1356-1362.

Havnes. EN: Sucharew. H: Kuhnell. P: Alden. J: Barnas. M: Wright. RO: Parsons. PJ: Aldous. KM: Praamsma. ML:
Beidler. C: Dietrich. KN. (2015). Manganese exposure and neurocognitive outcomes in rural school-age
children: The communities actively researching exposure study (Ohio, USA). Environ Health Perspect 123:
1066-1071. http://dx.doi.org/10.1289/ehp. 1408993.

Henn. BC: Schnaas. L: Ettinger. AS: Schwartz. J: Lamadrid-Figueroa. H: Hernandez-Avila. M: Amarasiriwardena.
C: Hu. H: Bellinger. DC: Wright. RO: Tellez-Roio. MM. (2012). Associations of early childhood
manganese and lead coexposure with neurodevelopment. Environ Health Perspect 120: 126-131.
http://dx.doi.org/10.1289/ehp.100330Q.

Hilson. JA: Strupp. BJ. (1997). Analyses of response patterns clarify lead effects in olfactory reversal and

extradimensional shift tasks: Assessment of inhibitory control, associative ability, and memory. Behav
Neurosci 111: 532-542. http://dx.doi.org/10.1037/0735-7044.111.3.532.

Hong. SB: Im. MH: Kim. JW: Park. F,J: Shin. MS: Kim. BN: Yoo. HJ: Cho. IH: Bhang. SY: Hong. YC: Cho. SC.
(2015). Environmental lead exposure and attention-deficit/hyperactivity disorder symptom domains in a
community sample of South Korean school-age children. Environ Health Perspect 123: 271-276.
http://dx.doi.org/10.1289/ehp.130742Q.

Hong. T: Li. SM: Jia. B: Huang. Y: Shu. K: Yuan. KW: Chen. L: Li. LX: Liu. L: Liu. ZY. (2021). DNA methylation
changes in the hippocampus of learning and memory disorder offspring rats of lead exposure during
pregnant and lactation period. Ann Palliat Med 10: 1059-1069. http://dx.doi.org/10.21037/apm-19-421.

Hornung. RW: Lanphear. BP: Dietrich. KN. (2009). Age of greatest susceptibility to childhood lead exposure: A

new statistical approach. Environ Health Perspect 117: 1309-1312. http://dx.doi.org/10.1289/ehp.080Q426.

Hortoa CJ: Weng. HY: Wells. EM. (2019). Association between blood lead level and subsequent Alzheimer's
disease mortality. Environmental Epidemiology 3: e045.
http://dx.doi.org/10.1097/EE9.0000000000000Q45.

Hortoa MK: Hsu. L: Henn. BC: Margolis. A: Austin. C: Svensson. K: Schnaas. L: Gennings. C: Hu. H: Wright. R:
Tellez Roio. MM: Arora. M. (2018). Dentine biomarkers of prenatal and early childhood exposure to
manganese, zinc and lead and childhood behavior. Environ Int 121: 148-158.
http://dx.doi.Org/10.1016/i.envint.2018.08.045.

External Review Draft

3-505

DRAFT: Do not cite or quote

-------
Hossain. S: Bhowmick. S: Jahan. S: Rozario. L: Sarkar. M: Islam. S: Basunia. MA: Rahman. A: Choudhurv. BK:
Shahialal. H. (2016). Maternal lead exposure decreases the levels of brain development and cognition-
related proteins with concomitant upsurges of oxidative stress, inflammatory response and apoptosis in the
offspring rats. Neurotoxicology 56: 150-158. http://dx.doi.Org/10.1016/i.neuro.2016.07.013.

Hsu. CY: Chuang. YC: Chang. FC: Chuang. HY: Chiou. TTY: Lee. CT. (2021). Disrupted sleep homeostasis and
altered expressions of clock genes in rats with chronic lead exposure. Toxics 9: 217.
http://dx.doi.org/10.3390/toxics909Q217.

Hu. H: Hernandez-Avila. M. (2002). Invited commentary: Lead, bones, women, and pregnancy - The poison within?
[Comment]. Am J Epidemiol 156: 1088-1091. http://dx.doi.org/10.1093/aie/kwfl64.

Hu. H: Tellez-Roio. MM: Bellinger. D: Smith. D: Ettinger. AS: Lamadrid-Figueroa. H: Schwartz. J: Schnaas. L:
Mercado-Garcia. A: Hernandez-Avila. M. (2006). Fetal lead exposure at each stage of pregnancy as a
predictor of infant mental development. Environ Health Perspect 114: 1730-1735.
http://dx.doi.org/10.1289/ehp.9067.

Huang. SY: Hu. H: Sanchez. BN: Peterson. KE: Ettinger. AS: Lamadrid-Figueroa. H: Schnaas. L: Mercado-Garcia.
A: Wright. RO: Basu. N: Cantonwine. DE: Hernandez-Avila. M: Tellez-Roio. MM. (2016). Childhood
blood lead levels and symptoms of attention deficit hyperactivity disorder (ADHD): A cross-sectional study
of Mexican children. Environ Health Perspect 124: 868-874. http://dx.doi.org/10.1289/ehp.1510067.

Iglesias. V: Steenland. K: Maisonet. M: Pino. P. (2011). Exposure to lead from a storage site associated with
intellectual impairment in Chilean children living nearby. Int J Occup Environ Health 17: 314-321.
http://dx.doi.org/10.1179/107735211799041841.

Ishitsuka. K: Yamamoto-Hanada. K: Yang. L: Mezawa. H: Konishi. M: Saito-Abe. M: Sasaki. H: Nishizato. M:

Sato. M: Koeda. T: Ohva. Y. (2020). Association between blood lead exposure and mental health in
pregnant women: Results from the Japan environment and children's study. Neurotoxicology 79: 191-199.
http://dx.doi.Org/10.1016/i.neuro.2020.06.003.

Jamesdaniel. S: Rosati. R: Westrick. J: Ruden. DM. (2018). Chronic lead exposure induces cochlear oxidative stress
and potentiates noise-induced hearing loss. Toxicol Lett 292: 175-180.
http://dx.doi.Org/10.1016/i.toxlet.2018.05.004.

Jedrvchowski. W: Perera. F: Jankowski. J: Mrozek-Budzvn. D: Mroz. E: Flak. E: Edwards. S: Skarupa. A:

Lisowska-Miszczvk. I. (2009a). Gender specific differences in neurodevelopmental effects of prenatal
exposure to very low-lead levels: The prospective cohort study in three-year olds. Early Hum Dev 85: 503-
510. http://dx.doi.Org/10.1016/i.earlhumdev.2009.04.006.

Jedrvchowski. W: Perera. FP: Jankowski. J: Mrozek-Budzva D: Mroz. E: Flak. E: Edwards. S: Skarupa. A:

Lisowska-Miszczvk. I. (2009b). Very low prenatal exposure to lead and mental development of children in
infancy and early childhood: Krakow prospective cohort study. Neuroepidemiology 32: 270-278.
http://dx.doi.org/10.1159/000203Q75.

Jeong. KS: Park. H: Ha. E: Hong. YC: Ha. M: Park. H: Kim. BN: Lee. SJ: Lee. KY: Kim. .TH: Kim. Y (2015).

Evidence that cognitive deficit in children is associated not only with iron deficiency, but also with blood
lead concentration: A preliminary study. J Trace Elem Med Biol 29: 336-341.
http://dx.doi.Org/10.1016/i.itemb.2014.08.014.

Ji. JS: Elbaz. A: Weisskopf. MG. (2013). Association between blood lead and walking speed in the National Health
and Nutrition Examination Survey (NHANES 1999-2002). Environ Health Perspect 121: 711-716.
http://dx.doi.org/10.1289/ehp.1205918.

Ji. JS: Power. MC: Sparrow. D: Spiro. A. Ill: Hu. H: Louis. ED: Weisskopf. MG. (2015). Lead exposure and tremor
among older men: The VA normative aging study. Environ Health Perspect 123: 445-450.
http://dx.doi.org/10.1289/ehp.1408535.

Ji. Y: Hong. X: Wang. G: Chatteriee. N: Rilev. AW: Lee. LC: Surkan. PJ: Bartell. TR: Zuckerman. B: Wang. X.

(2018). A prospective birth cohort study on early childhood lead levels and attention deficit hyperactivity
disorder: New insight on sex differences. J Pediatr 199: 124-13 l.el28.
http://dx.doi.Org/10.1016/i.ipeds.2018.03.076.

External Review Draft

3-506

DRAFT: Do not cite or quote

-------
Jiang. C'B: Kao. CS: Chien. LC: Chen. YJ: Liao. KW. (2022). Associations among prenatal and postnatal arsenic,
lead, and cadmium exposures and motor development in 3-year-old children: A longitudinal birth cohort
study in Taiwan. Environ Sci Pollut Res Int 29: 43191-43200. http://dx.doi.org/10.1007/sl 1356-021-
18321-5.

Joo. H: Choi. JH: Burm. E: Park. H: Hong. YC: Kim. Y: Ha. F,H: Kim. Y: Kim. BN: Ha. M. (2018). Gender

difference in the effects of lead exposure at different time windows on neurobehavioral development in 5-
year-old children. Sci Total Environ 615: 1086-1092. http://dx.doi.Org/10.1016/i.scitotenv.2017.10.007.

Joo. H: Lim. MH: Ha. M: Kwon. HJ: Yoo. SJ: Choi. KH: Paik. KC. (2017). Secondhand smoke exposure and low
blood lead levels in association with attention-deficit hyperactivity disorder and its symptom domain in
children: A community-based case-control study. Nicotine Tob Res 19: 94-101.
http://dx.doi.org/10.1093/ntr/ntwl52.

Kamel. F: Umbach. DM: Munsat. TL: Shefner. JM: Hu. H: Sandler. DP. (2002). Lead exposure and amyotrophic
lateral sclerosis. Epidemiology 13: 311-319.

Kamel. F: Umbach. DM: Stallone. L: Richards. M: Hu. H: Sandler. DP. (2008). Association of lead exposure with
survival in amyotrophic lateral sclerosis. Environ Health Perspect 116: 943-947.
http://dx.doi.org/10.1289/ehp. 11193.

Kang. GH: Uhm. JY: Choi. YG: Kang. EK: Kim. SY: Choo. WO: Chang. SS. (2018). Environmental exposure of
heavy metal (lead and cadmium) and hearing loss: Data from the Korea National Health and Nutrition
Examination Survey (KNHANES 2010-2013). Ann Occup Environ Med 30: 22.
http://dx.doi.org/10.1186/s40557-018-0237-9.

Kao. CS: Wang. YL: Chuang. TW: Jiang. CB: Hsi. HC: Liao. KW: Chien. LC. (2021). Effects of soil lead exposure
and land use characteristics on neurodevelopment among children under 3 years of age in northern Taiwan.
Environ Pollut 286: 117288. http://dx.doi.Org/10.1016/i.envpol.2021.117288.

Karri. V: Ramos. D: Martinez. JB: Odena. A: Oliveira. E: Coort. SL: Evelo. CT: Marimaa ECM: Schuhmacher. M:
Kumar. V. (2018). Differential protein expression of hippocampal cells associated with heavy metals (Pb,
As, and MeHg) neurotoxicity: Deepening into the molecular mechanism of neurodegenerative diseases. J
Proteomics 187: 106-125. http://dx.doi.org/10.1016/i.jprot.2018.06.020.

Kawamoto. EM: Vivar. C: Camandola. S. (2012). Physiology and pathology of calcium signaling in the brain. Front
Pharmacol 3: 61. http://dx.doi.org/10.3389/fphar.2012.00061.

Khalil. N: Faulkner. KA: Greenspan. SL: Caulev. JA. (2014). Associations between bone mineral density, grip
strength, and lead body burden in older men. J Am Geriatr Soc 62: 141-146.
http://dx.doi.org/10. Ill 1/jgs. 12603.

Kicinski. M: Vriiens. J: Vermier. G: Hond. ED: Schoeters. G: Nelen. V: Bruckers. L: Sioea I: Baevens. W: Van

Larebeke. N: Viaene. MK: Nawrot. TS. (2015). Neurobehavioral function and low-level metal exposure in
adolescents. Int J Hyg Environ Health 218: 139-146. http://dx.doi.Org/10.1016/i.iiheh.2014.09.002.

Kim. JI: Kim. JW: Lee. JM: Yun. HJ: Sohn. CH: Shin. MS: Kim. B: Chae. J: Roh. J: Kim. BN. (2018a). Interaction
between DRD2 and lead exposure on the cortical thickness of the frontal lobe in youth with attention-
deficit/hyperactivity disorder. Prog Neuropsychopharmacol Biol Psychiatry 82: 169-176.
http://dx.doi.Org/10.1016/i.pnpbp.2017.ll.018.

Kim. KN: Kwon. HJ: Hong. YC. (2016). Low-level lead exposure and autistic behaviors in school-age children.
Neurotoxicology 53: 193-200. http://dx.doi.Org/10.1016/i.neuro.2016.02.004.

Kim. S: Arora. M: Fernandez. C: Landero. J: Caruso. J: Chen. A. (2013a). Lead, mercury, and cadmium exposure
and attention deficit hyperactivity disorder in children. Environ Res 126: 105-110.
http://dx.doi.Org/10.1016/i.envres.2013.08.008.

Kim. S: Eom. S: Kim. HJ: Lee. JJ: Choi. G: Choi. S: Kim. S: Kim. SY: Cho. G: Kim. YD: Suh. E: Kim. SK: Kim. S:
Kim. GH: Moon. HB: Park. J: Kim. S: Choi. K: Eun. SH. (2018b). Association between maternal exposure
to major phthalates, heavy metals, and persistent organic pollutants, and the neurodevelopmental
performances of their children at 1 to 2 years of age-CHECK cohort study. Sci Total Environ 624: 377-
384. http://dx.doi.Org/10.1016/i.scitotenv.2017.12.058.

External Review Draft

3-507

DRAFT: Do not cite or quote

-------
Kim. Y: Ha. EH: Park. H: Ha. M: Kim. Y: Hong. YC: Kim. EJ: Kim. BN. (2013b). Erratum to: "Prenatal lead and
cadmium co-exposure and infant neurodevelopment at 6 months of age: The Mothers and Children's
Environmental Health (MOCEH) study" [NeuroToxicology 35 (2013) 15-22] [Erratum], Neurotoxicology
37: 248-249. http://dx.doi.Org/10.1016/i.neuro.2013.04.009.

Kim. Y: Ha. EH: Park. H: Ha. M: Kim. Y: Hong. YC: Kim. EJ: Kim. BN. (2013c). Prenatal lead and cadmium co-
exposure and infant neurodevelopment at 6 months of age: The Mothers and Children's Environmental
Health (MOCEH) study. Neurotoxicology 35: 15-22. http://dx.doi.Org/10.1016/i.neuro.2012.ll.006.

Kim. Y: Kim. BN: Hong. YC: Shin. MS: Yoo. HJ: Kim. JW: Bhang. SY: Cho. SC. (2009). Co-exposure to

environmental lead and manganese affects the intelligence of school-aged children. Neurotoxicology 30:
564-571. http://dx.doi.Org/10.1016/i.neuro.2009.03.012.

Kirrane. EF: Patel. MM. (2014). Identification and consideration of errors in Lanphear et al. (2005), "Low-level
environmental lead exposure and children's intellectual function: An international pooled analysis".
Available online at https://www.regulations.gov/document/EPA-HO-ORD-2011-0051-005Q (accessed

Kordas. K: Ettinger. AS: Bellinger. DC: Schnaas. L: Tellez Roio. MM: Hernandez-Avila. M: Hu. H: Wright. RO.
(2011). A dopamine receptor (DRD2) but not dopamine transporter (DAT1) gene polymorphism is
associated with neurocognitive development of Mexican preschool children with lead exposure. J Pediatr
159: 638-643. http://dx.doi.Org/10.1016/i.ipeds.2011.03.043.

Koshv. B: Srinivasan. M: Zachariah. SM: Karthikevan. AS: Roshan. R: Bose. A: Mohan. VR: John. S: Ramanuiam.
K: Mulivil. J: Kang. G. (2020). Body iron and lead status in early childhood and its effects on development
and cognition: A longitudinal study from urban Vellore. Public Health Nutr 23: 1896-1906.
http://dx.doi.org/10.1017/S13689800190Q4622.

Krieg. EF. Jr: Butler. MA: Chang. MH: Liu. TB: Yesupriva. A: Lindegren. ML: Dowling. N. (2009). Lead and
cognitive function in ALAD genotypes in the Third National Health and Nutrition Examination Survey.
Neurotoxicol Teratol 31: 364-371. http://dx.doi.Org/10.1016/i.ntt.2009.08.003.

Krieg. EF. Jr: Butler. MA: MH. C: Liu. TB: Yesupriva. A: Dowling. N: Lindegren. ML. (2010). Lead and cognitive
function in VDR genotypes in the third National Health and Nutrition Examination Survey. Neurotoxicol
Teratol 32: 262-272. http://dx.doi.Org/10.1016/i.ntt.2009.12.004.

Kuang. W: Chea Z: Shi. K: Sun. H: Li. H: Huang. L: Bi. J. (2020). Adverse health effects of lead exposure on
physical growth, erythrocyte parameters and school performances for school-aged children in eastern
China. Environ Int 145: 106130. http://dx.doi.Org/10.1016/i.envint.2020.106130.

Kuhlmann. AC: Mcglothan. JL: Guilarte. TR. (1997). Developmental lead exposure causes spatial learning deficits
in adult rats. Neurosci Lett 233: 101-104. http://dx.doi.org/10.1016/S0304-3940(97)00633-2.

Kumar. VL: Muralidhara. (2014). Ameliorative effects of ferulic Acid against lead acetate-induced oxidative stress,
mitochondrial dysfunctions and toxicity in prepubertal rat brain. Neurochem Res39:2501-2515.
http://dx.doi.org/10.1007/sllQ64-014-1451-7.

Lamoureux-Tremblav. V: Chauret. M: Muckle. G: Maheu. F: Suffren. S: Jacobson. SW: Jacobson. JL: Avotte. P:
Lepore. F: Saint-Amour. D. (2021). Altered functional activations of prefrontal brain areas during
emotional processing of fear in Inuit adolescents exposed to environmental contaminants. Neurotoxicol
Teratol 85: 106973. http://dx.doi.Org/10.1016/i.ntt.2021.106973.

Lanphear. BP: Dietrich. K: Auinger. P: Cox. C. (2000). Cognitive deficits associated with blood lead concentrations
<10 microg/dL in US children and adolescents. Public Health Rep 115: 521-529.
http://dx.doi.Org/10.1093/phr/115.6.521.

Lanphear. BP: Hornung. R: Khourv. J: Yolton. K: Baghurst. P: Bellinger. DC: Canfield. RL: Dietrich. KN:

Bornschein. R: Greene. T: Rothenberg. SJ: Needleman. HL: Schnaas. L: Wasserman. G: Graziano. J:
Roberts. R. (2005). Low-level environmental lead exposure and children's intellectual function: An
international pooled analysis. Environ Health Perspect 113: 894-899. http://dx.doi.org/10.1289/ehp.7688.

External Review Draft

3-508

DRAFT: Do not cite or quote

-------
Lanphear. BP: Hornung. R: Khourv. J: Yolton. K: Baghurst. P: Bellinger. DC: Canfield. RL: Dietrich. KN:

Bornschein. R: Greene. T: Rothenberg. SJ: Needleman. HL: Schnaas. L: Wasserman. G: Graziano. J:
Roberts. R. (2019). Erratum: "Low-level environmental lead exposure and children's intellectual function:
An international pooled analysis" [Erratum], Environ Health Perspect 127: 099001.
http://dx.doi.org/10.1289/EHP5685.

Laughlin. NK: Luck. ML: Laskv. RE. (20081. Postnatal lead effects on the development of visual spatial acuity in
rhesus monkeys (Macaca mulatta). Dev Psychobiol 50: 608-614. http://dx.doi.org/10.1002/dev.20315.

Laughlin. NK: Luck. ML: Laskv. RE. (20091. Early lead exposure effects on an auditory threshold task in the rhesus
monkey (Macaca mulatta). Dev Psychobiol 51: 289-300. http://dx.doi.org/10.1002/dev.20364.

Leasure. JL: Giddabasappa. A: Chanev. S: Johnson. JE. Jr: Pothakos. K: Lau. YS: Fox. DA. (2008). Low-level
human equivalent gestational lead exposure produces sex-specific motor and coordination abnormalities
and late-onset obesity in year-old mice. Environ Health Perspect 116: 355-361.
http://dx.doi.org/10.1289/ehp.10862.

Lee. H: Park. H: Ha. E: Hong. YC: Ha. M: Park. H: Kim. BN: Lee. SJ: Lee. KY: Kim. JH: .Teong. KS: Kim. Y
(2017). Stability of cognitive development during the first five years of life in relation to heavy metal
concentrations in umbilical cord blood: Mothers' and Children's Environmental Health (MOCEH) birth
cohort study. Sci Total Environ 609: 153-159. http://dx.doi.Org/10.1016/i.scitotenv.2017.07.074.

Lee. KS: Kim. KN: Ahn. YD: Choi. YJ: Cho. J: Jang. Y: Lim. YH: Kim. JT: Shin. CH: Lee. YA: Kim. BN: Hong.
YC. (2021). Prenatal and postnatal exposures to four metals mixture and IQ in 6-year-old children: A
prospective cohort study in South Korea. Environ Int 157: 106798.
http://dx.doi.Org/10.1016/i.envint.2021.106798.

Lenroot. RK: Giedd. JN. (2006). Brain development in children and adolescents: Insights from anatomical magnetic
resonance imaging [Review]. Neurosci Biobehav Rev 30: 718-729.
http://dx.doi.Org/10.1016/i.neubiorev.2006.06.001.

Leviton. A: Bellinger. D: Allred. EN: Rabinowitz. M: Needleman. H: Schoenbaum. S. (1993). Pre- and postnatal
low-level lead exposure and children's dysfunction in school. Environ Res 60: 30-43.
http://dx.doi.org/10.1006/enrs.1993.10Q3.

Li. HW: Deng. JG: Du. ZC: Yan. MS: Long. ZX: Pham Thi. PT: Yang. KD. (2013). Protective effects of mangiferin
in subchronic developmental lead-exposed rats. Biol Trace Elem Res 152: 233-242.
http://dx.doi.org/10.1007/sl2011-013-961Q-2.

Li. N: Li. X: Li. L: Zhang. P: Qiao. M: Zhao. O: Song. L: Yu. Z. (2016a). Original research: The expression of

MMP2 and MMP9 in the hippocampus and cerebral cortex of newborn mice under maternal lead exposure.
Exp Biol Med 241: 1811-1818. http://dx.doi.org/10.1177/1535370216647808.

Li. N: Qiao. M: Zhang. P: Li. X: Li. L. i: Yu. Z. (2016b). The effects of early life lead exposure on the expression of
glycogen synthase kinase-3(3 and insulin-like growth factor 1 receptor in the hippocampus of mouse pups.
Biol Trace Elem Res 169: 114-120. http://dx.doi.org/10.1007/sl2011-015-Q382-8.

Li. N: Qiao. M: Zhao. O: Zhang. P: Song. L: Li. L: Cui. C. (2016c). Effects of maternal lead exposure on RGMa and
RGMb expression in the hippocampus and cerebral cortex of mouse pups. Brain Res Bull 127: 38-46.
http://dx.doi.Org/10.1016/i.brainresbull.2016.08.010.

Li. N: Yang. G: Wang. Y: Qiao. M: Zhang. P: Shao. J: Yang. G. (2016d). Decreased IDE and IGF2 expression but
increased A(U0 in the cerebral cortex of mouse pups by early life lead exposure. Brain Res Bull 121: 84-
90. http://dx.doi.Org/10.1016/i.brainresbull.2016.01.004.

Li. N: Yu. ZL: Wang. L: Zheng. YT: Jia. JX: Wang. O: Zhu. MJ: Liu. XL: Xia. X: Li. WJ. (2010). Increased tau

phosphorylation and beta amyloid in the hipocampus of mouse pups by early life lead exposure. Acta Biol
Hung 61: 123-134. http://dx.doi.Org/10.1556/ABiol.61.2010.2.l.

Li. S: Xu. J: Liu. Z: Yan. CH. (2017). The non-linear association between low-level lead exposure and maternal
stress among pregnant women. Neurotoxicology 59: 191-196.
http://dx.doi.Org/10.1016/i.neuro.2016.07.005.

External Review Draft

3-509

DRAFT: Do not cite or quote

-------
Lilienthal. H: Lenaerts. C: Winneke. G: Hennekes. R. (1988). Alteration of the visual evoked potential and the
electroretinogram in lead-treated monkeys. Neurotoxicol Teratol 10: 417-422.
http://dx.doi.org/10.1016/0892-0362(88190002-5.

Lilienthal. H: Winneke. G. (19961. Lead effects on the brain stem auditory evoked potential in monkeys during and
after the treatment phase. Neurotoxicol Teratol 18: 17-32. http://dx.doi.org/10.1016/0892-0362(95102010-
1.

Lin. CC: Chen. YC: Su. FC: Lin. CM: Lia.o. HF: Hwang. YH: Hsieh. WS: Jeng. SF: Su. YN: Chen. PC. (2013). In
utero exposure to environmental lead and manganese and neurodevelopment at 2 years of age. Environ Res
123: 52-57. http://dx.doi.Org/10.1016/i.envres.2013.03.003.

Lin. D: Lutter. R: Ruhm. C. (2016). Cognitive Performance and Labor Market Outcomes. National Bureau of
Economic Research, http://dx.doi.org/10.3386/w22470.

Lin. Y: Huang. L: Xu. J: Specht. AJ: Yan. C: Geng. H: Shen. X: Nie. LH: Hu. H. (2019). Blood lead, bone lead and
child attention-deficit-hyperactivity-disorder-like behavior. Sci Total Environ 659: 161-167.
http://dx.doi.Org/10.1016/i.scitotenv.2018.12.219.

Listos. J: Baranowska-Bosiacka. I: Talarek. S: Listos. P: Orzelska. J: Fidecka. S: Gutowska. I: Kolasa. A: Rvbicka.
M: Chlubek. D. (2013). The effect of perinatal lead exposure on dopamine receptor D2 expression in
morphine dependent rats. Toxicology 310: 73-83. http://dx.doi.Org/10.1016/i.tox.2013.05.007.

Liu. C: Huang. L: Huang. S: Wei. L: Cao. D: Zan. G: Tan. Y: Wang. S: Yang. M: Tian. L: Tang. W: He. C: Shen. C:
Luo. B: Zhu. M: Liang. T: Pang. B: Li. M: Mo. Z: Yang. X. (2022a). Association of both prenatal and early
childhood multiple metals exposure with neurodevelopment in infant: A prospective cohort study. Environ
Res 205: 112450. http://dx.doi.Org/10.1016/i.envres.2021.112450.

Liu. C: Yang. W: Ma. J: Yang. H: Feng. Z: Sun. J: Cheng. C: Jiang. H. (2018a). Dihydromyricetin inhibits lead-
induced cognitive impairments and inflammation by the adenosine 5'-monophosphate-activated protein
kinase pathway in mice. J Agric Food Chem 66: 7975-7982. http://dx.doi.org/10.1021/acs.iafc.8b02433.

Liu. CM: Tian. ZK: Zhang. YJ: Ming. OL: Ma. JO: Ji. LP. (2020). Effects of Gastrodin against Lead-Induced Brain
Injury in Mice Associated with the Wnt/Nrf2 Pathway. Nutrients 12: 1-13.
http://dx.doi.org/10.3390/nul2061805.

Liu. F: Xue. Z: Li. N: Huang. H: Ying. Y: Li. J: Wang. L: Li. W. (2014a). Effects of lead exposure on the expression
of amyloid (3 and phosphorylated tau proteins in the C57BL/6 mouse hippocampus at different life stages. J
Trace Elem Med Biol 28: 227-232. http://dx.doi.Org/10.1016/i.itemb.2014.01.002.

Liu. J: Li. L: Wang. Y: Yan. C: Liu. X. (2013a). Impact of low blood lead concentrations on IQ and school

performance in Chinese children. PLoS ONE 8: e65230. http://dx.doi.org/10.1371/iournal.pone.0065230.

Liu. J: Liu. X: Wang. W: McCaulev. L: Pinto-Martin. J: Wang. Y: Li. L: Yan. C: Rogan. WJ. (2014b). Blood lead
concentrations and children's behavioral and emotional problems: A cohort study. JAMA Pediatr 168: 737-
745. http://dx.doi.org/10.1001/iamapediatrics.2Q14.332.

Liu. J: Portnov. J: Raine. A: Gladieux. M: McGarrv. P: Chea A. (2022b). Blood lead levels mediate the relationship
between social adversity and child externalizing behavior. Environ Res 204: 112396.
http://dx.doi.Org/10.1016/i.envres.2021.112396.

Liu. JA: Chen. YJ: Gao. DG: Jing. J: Hu. OS. (2014c). Prenatal and postnatal lead exposure and cognitive

development of infants followed over the first three years of life: A prospective birth study in the Pearl
River Delta region, China. Neurotoxicology 44: 326-334. http://dx.doi.Org/10.1016/i.neuro.2014.07.001.

Liu. JA: Gao. DJ: Chen. YM: Jing. J: Hu. OS: Chen. YJ. (2014d). Lead exposure at each stage of pregnancy and
neurobehavioral development of neonates. Neurotoxicology 44: 1-7.
http://dx.doi.Org/10.1016/i.neuro.2014.03.003.

Liu. MC: Liu. XO: Wang. W: Shen. XF: Che. HL: Guo. YY: Zhao. MG: Chen. JY: Luo. WJ. (2012). Involvement
of microglia activation in the lead induced long-term potentiation impairment. PLoS ONE 7: e43924.
http://dx.doi.org/10.1371/iournal.pone.0043924.

External Review Draft

3-510

DRAFT: Do not cite or quote

-------
Liu. MC: Xu. Y: Chen. YM; Li. J: Zhao. F: Zheng. G: Jing. JF: Ke. T: Chen. JY: Luo. WJ. (2013b). The effect of
sodium selenite on lead induced cognitive dysfunction. Neurotoxicology 36: 82-88.
http://dx.doi.Org/10.1016/i.neuro.2013.03.008.

Liu. R: Bai. L: Liu. M: Wang. R: Wu. Y: Li. O: Ba. Y: Zhang. H: Zhou. G: Yu. F: Huang. H. (2022c). Combined
exposure of lead and high-fat diet enhanced cognitive decline via interacting with CREB-BDNF signaling
in male rats. Environ Pollut 304: 119200. http://dx.doi.Org/10.1016/i.envpol.2022.119200.

Liu. SH: Bobb. JF: Claus Henn. B: Gennings. C: Schnaas. L: Tellez-Roio. M: Bellinger. D: Arora. M: Wright. RO:
Coull. BA. (2018b). Bayesian varying coefficient kernel machine regression to assess neurodevelopmental
trajectories associated with exposure to complex mixtures. Stat Med 37: 4680-4694.
http://dx.doi.org/10.1002/sim.7947.

Liu. W: Huo. X: Liu. D: Zeng. X: Zhang. Y: Xu. X. (2014e). S 100(3 in heavy metal-related child attention-deficit
hyperactivity disorder in an informal e-waste recycling area. Neurotoxicology 45: 185-191.
http://dx.doi.Org/10.1016/i.neuro.2014.10.013.

Liu. X: Dietrich. KN: Radcliffe. J: Ragan. NB: Rhoads. GG: Rogan. WJ. (2002). Do children with falling blood lead
levels have improved cognition? Pediatrics 110: 787-791. http://dx.doi.org/10.1542/peds. 110.4.787.

Liu. X: Wei. F: Cheng. Y: Zhang. Y: Jia. G: Zhou. J: Zhu. M: Shan. Y: Sun. X: Yu. L: Merzenich. MM: Lurie. DI:
Zheng. O: Zhou. X. (2019). Auditory training reverses lead (Pb)-toxicity-induced changes in sound-
azimuth selectivity of cortical neurons. Cereb Cortex 29: 3294-3304.
http://dx.doi.org/10.1093/cercor/bhvl99.

Liu. X: Zheng. G: Wu. Y: Shen. X: Jing. J: Yu. T: Song. H: Chen. J: Luo. W. (2013c). Lead exposure results in
hearing loss and disruption of the cochlear blood-labyrinth barrier and the protective role of iron
supplement. Neurotoxicology 39: 173-181. http://dx.doi.Org/10.1016/i.neuro.2013.10.002.

Liu. XO: Huang. R: Zhou. X: Cai. TJ: Chen. JJ: Shi. XW: Deng. HB: Luo. WJ. (2017). Presence of nano-sized

chitosan-layered silicate composites protects against toxicity induced by lead ions. Carbohydr Polym 158:
1-10. http://dx.doi.Org/10.1016/i.carbpol.2016.ll.084.

Liu. Y: Huo. X: Xu. L: Wei. X: Wu. W: Wu. X: Xu. X. (2018c). Hearing loss in children with e-waste lead and
cadmium exposure. Sci Total Environ 624: 621-627. http://dx.doi.Org/10.1016/i.scitotenv.2017.12.091.

Liu. YF: Chen. O: Wei. XP: Chen. L: Zhang. X: Chen. K: Chen. J: Li. TY. (2015). Relationship between perinatal
antioxidant vitamin and heavy metal levels and the growth and cognitive development of children at 5
years of age. Asia Pac J Clin Nutr 24: 650-658. http://dx.doi.Org/10.6133/apicn.2015.24.4.25.

Long. X: Wu. H: Zhou. Y: Wan. Y: Kan. X: Gong. J: Zhao. X. (2022). Preventive effect of Limosilactobacillus
fermentum SCHY34 on lead acetate-induced neurological damage in SDrats. Front Nutr 9: 852012.
http://dx.doi.org/10.3389/fnut.2022.852012.

Lu. X: Jin. C: Yang. J: Liu. O: Wu. S: Li. D: Guan. Y: Cai. Y. (2013). Prenatal and lactational lead exposure

enhanced oxidative stress and altered apoptosis status in offspring rats' hippocampus. Biol Trace Elem Res
151: 75-84. http://dx.doi.org/10.1007/sl2011-Q12-9531-5.

Lucchini. RG: Zoni. S: Guazzetti. S: Bontempi. E: Micheletti. S: Broberg. K: Parrinello. G: Smith. DR. (2012).

Inverse association of intellectual function with very low blood lead but not with manganese exposure in
Italian adolescents. Environ Res 118: 65-71. http://dx.doi.Org/10.1016/i.envres.2012.08.003.

Luo. M: Xu. Y: Cai. R: Tang. Y: Ge. MM: Liu. ZH: Xu. L: Hu. F: Ruan. DY: Wang. HL. (2014). Epigenetic histone
modification regulates developmental lead exposure induced hyperactivity in rats. Toxicol Lett 225: 78-85.
http://dx.doi.Org/10.1016/i.toxlet.2013.ll.025.

Lvnam. PR: Charnigo. R: Moffitt. TE: Raine. A: Loeber. R: Stouthamer-Loeber. M. (2009). The stability of
psychopathy across adolescence. Dev Psychopathol 21: 1133-1153.
http://dx.doi.org/10.1017/S0954579409990Q83.

Lynch. G: Kessler. M: Arai. A: Larson. J. (1990). The nature and causes of hippocampal long-term potentiation
[Review], Prog Brain Res 83: 233-250. http://dx.doi.org/10.1016/s0079-6123(08)61253-4.

External Review Draft

3-511

DRAFT: Do not cite or quote

-------
Ma. J: Yan. L: Guo. T: Yang. S: Guo. C: Liu. Y: Xie. O: Wang. J. (2019). Association of typical toxic heavy metals
with schizophrenia. Int J Environ Res Public Health 16: 4200. http://dx.doi.org/10.3390/iierphl621420Q.

Macedoni-Luksic. M: Gosar. D: Biorklund. G: Orazem. J: Kodric. J: Lesnik-Musek. P: Zupancic. M: France-Stiglic.
A: Sesek-Briski. A: Neubauer. D: Osredkar. J. (2015). Levels of metals in the blood and specific
porphyrins in the urine in children with Autism Spectrum Disorders. Biol Trace Elem Res 163: 2-10.
http://dx.doi.org/10.1007/sl2011-014-Q121-6.

Magzamen. S: Amato. MS: Imm. P: Havlena. JA: Coons. MJ: Anderson. HA: Kanarek. MS: Moore. CF. (2015).

Quantile regression in environmental health: Early life lead exposure and end-of-grade exams. Environ Res
137: 108-119. http://dx.doi.Org/10.1016/i.envres.2014.12.004.

Maiti. AK: Saha. NC: More. SS: Panigrahi. AK: Paul. G. (2017). Neuroprotective efficacy of mitochondrial
antioxidant MitoQ in suppressing peroxynitrite-mediated mitochondrial dysfunction inflicted by lead
toxicity in the rat brain. NeurotoxRes 31: 358-372. http://dx.doi.org/10.1007/sl2640-Q16-9692-7.

Mani. MS: Joshi. MB: Shettv. RR: Dsouza. VL: Swathi. M: Kabekkodu. SP: Dsouza. HS. (2020). Lead exposure
induces metabolic reprogramming in rat models. Toxicol Lett 335: 11-27.
http://dx.doi.Org/10.1016/i.toxlet.2020.09.010.

Mansouri. MT: Naghizadeh. B: Lopez-Larrubia. P: Cauli. O. (2012). Gender-dependent behavioural impairment and
brain metabolites in young adult rats after short term exposure to lead acetate. Toxicol Lett 210: 15-23.
http://dx.doi.Org/10.1016/i.toxlet.2012.01.012.

Mansouri. MT: Naghizadeh. B: Lopez-Larrubia. P: Cauli. O. (2013). Behavioral deficits induced by lead exposure
are accompanied by serotonergic and cholinergic alterations in the prefrontal cortex. Neurochem Int 62:
232-239. http://dx.doi.Org/10.1016/i.neuint.2012.12.009.

Marcus. DK: Fulton. JJ: Clarke. EJ. (2010). Lead and conduct problems: A meta-analysis. J Clin Child Adolesc
Psychol 39: 234-241. http://dx.doi.org/10.1080/153744110Q3591455.

Marques. RC: Bernardi. JV: Dorea. JG: de Fatima Ramos Moreira. M: Malm. O. (2014). Perinatal multiple exposure
to neurotoxic (lead, methylmercury, ethylmercury, and aluminum) substances and neurodevelopment at six
and 24 months of age. Environ Pollut 187: 130-135. http://dx.doi.Org/10.1016/i.envpol.2014.01.004.

Martin. KV: Sucharew. H: Dietrich. KN: Parsons. PJ: Palmer. CD: Wright. R: Amarasiriwardena. C: Smith. PR:
Havnes. EN. (2021). Co-exposure to manganese and lead and pediatric neurocognition in East Liverpool,
Ohio. Environ Res 202: 111644. http://dx.doi.org/10.1016/i.envres.2021.111644.

Maughan. B: Stafford. M: Shah. I: Kuh. D. (2014). Adolescent conduct problems and premature mortality: follow-
up to age 65 years in a national birth cohort. Psychol Med 44: 1077-1086.
http://dx.doi.org/10.1017/S00332917130014Q2.

Mazumdar. M: Bellinger. DC: Gregas. M: Abanilla. K: Bacic. J: Needleman. HL. (2011). Low-level environmental
lead exposure in childhood and adult intellectual function: A follow-up study. Environ Health 10: 24.
http://dx.doi.org/10.1186/1476-069X-10-24.

McCall. RB: Hogartv. PS: Hurlburt. N. (1972). Transitions in infant sensorimotor development and the prediction of
childhood IQ [Review]. Am Psychol 27: 728-748. http://dx.doi.org/10.1037/h0033148.

Mccollister. KE: French. MT: Fang. H. (2010). The cost of crime to society: new crime-specific estimates for policy
and program evaluation. Drug Alcohol Depend 108: 98-109.
http://dx.doi.Org/10.1016/i.drugalcdep.2009.12.002.

McFarlane. AC: Searle. AK: Van Hooff. M: Baghurst. PA: Sawyer. MG: Galletlv. C: Sim. MR: Clark. LS. (2013).
Prospective associations between childhood low-level lead exposure and adult mental health problems: the
PortPirie cohort study. Neurotoxicology 39: 11-17. http://dx.doi.Org/10.1016/i.neuro.2013.08.003.

Mcgivern. RF: Sokol. RZ: Bermaa NG. (1991). Prenatal lead exposure in the rat during the third week of gestation:
Long-term behavioral, physiological and anatomical effects associated with reproduction. Toxicol Appl
Pharmacol 110: 206-215. http://dx.doi.org/10.1016/S0041-008X(05)80003-l.

External Review Draft

3-512

DRAFT: Do not cite or quote

-------
McMichael. AJ: Baghurst. PA: Vimpani. GV: Robertson. EF: Wigg. NR: Tong. SL. (1992). Sociodemographic

factors modifying the effect of environmental lead on neuropsychological development in early childhood.
Neurotoxicol Teratol 14: 321-327. http://dx.doi.org/10.1016/0892-0362(92190038-0.

Menezes-Filho. JA: Carvalho. CF: Rodrigues. JLG: Arauio. CFS: Dos Santos. NR: Lima. CS: Bandeira. MJ:

Marques. BLS: Anios. ALS: Bah. HAF: Abreu. N: Philibert. A: Mergler. D. (2018). Environmental co-
exposure to lead and manganese and intellectual deficit in school-aged children. Int J Environ Res Public
Health 15: 2418. http://dx.doi.org/10.3390/iierphl5112418.

Meng. H: Wang. L: He. JH: Wang. ZF. (2016). The Protective Effect of Gangliosides on Lead (Pb)-Induced
Neurotoxicity Is Mediated by Autophagic Pathways. Int J Environ Res Public Health 13: 365.
http://dx.doi.org/10.3390/iierphl3040365.

Meramat. A: Raiab. NF: Shahar. S: Sharif. RA. (2017). DNA damage, copper and lead associates with cognitive

function among older adults. J Nutr Health Aging 21: 539-545. http://dx.doi.org/10.1007/sl2603-016-Q759-
1.

Merced-Nieves. FM: Chelonis. J: Pantic. I: Schnass. L: Tellez-Roio. MM: Braun. JM: Paule. MG: Wright. RJ:

Wright. RO: Curtin. P. (2022). Sexually dimorphic associations between prenatal blood lead exposure and
performance on a behavioral testing battery in children. Neurotoxicol Teratol 90: 107075.
http://dx.doi.Org/10.1016/i.ntt.2022.107075.

Min. JY: Min. KB: Cho. SI: Kim. R: Sakong. J: Paek. D. (2007). Neurobehavioral function in children with low
blood lead concentrations. Neurotoxicology 28: 421-425. http://dx.doi.Org/10.1016/i.neuro.2006.03.007.

Min. KB: Lee. KJ: Park. JB: Min. JY. (2012). Lead and cadmium levels and balance and vestibular dysfunction

among adult participants in the National Health and Nutrition Examination Survey (NHANES) 1999-2004.
Environ Health Perspect 120: 413-417. http://dx.doi.org/10.1289/ehp. 1103643.

Min. MO: Singer. LT: Kirchner. HL: Minnes. S: Short. E: Hussain. Z: Nelson. S. (2009). Cognitive development
and low-level lead exposure in poly-drug exposed children. Neurotoxicol Teratol 31: 225-231.
http://dx.doi.Org/10.1016/i.ntt.2009.03.002.

Miranda. ML: Kim. D: Galeano. M: Paul. CJ: Hull. AP: Morgan. SP. (2007). The relationship between early

childhood blood lead levels and performance on end-of-grade tests. Environ Health Perspect 115: 1242-
1247. http://dx.doi.org/10.1289/ehp.9994.

Miranda. ML: Kim. D: Reiter. J: Galeano. MAO: Maxson. P. (2009). Environmental contributors to the achievement
gap. Neurotoxicology 30: 1019-1024. http://dx.doi.Org/10.1016/i.neuro.2009.07.012.

Mohammad. IK: Mahdi. AA: Raviraia. A: Naimul. I: Iqbal. A: Thuppil. V. (2008). Oxidative stress in painters

exposed to low lead levels. Arh Hig Rada Toksikol 59: 161-169. http://dx.doi.org/10.2478/100Q4-1254-59-
2008-1883.

Molendiik. ML: de Kloet. ER. (2015). Immobility in the forced swim test is adaptive and does not reflect

depression. Psychoneuroendocrinology 62: 389-391. http://dx.doi.Org/10.1016/i.psvneuen.2015.08.028.

Molina. RM: Phattanarudee. S: Kim. J: Thompson. K: Wessling-Resnick. M: Maher. TJ: Brain. JD. (2011).

Ingestion of Mn and Pb by rats during and after pregnancy alters iron metabolism and behavior in
offspring. Neurotoxicology 32: 413-422. http://dx.doi.Org/10.1016/i.neuro.2011.03.010.

Monday. HR: Younts. TJ: Castillo. PE. (2018). Long-term plasticity of neurotransmitter release: Emerging

mechanisms and contributions to brain function and disease [Review]. Annu Rev Neurosci 41: 299-322.
http://dx.doi.org/10.1146/annurev-neuro-080317-Q62155.

Moore. CF: Gaiewski. LL: Laughlin. NK: Luck. ML: Larson. JA: Schneider. ML. (2008). Developmental lead

exposure induces tactile defensiveness in rhesus monkeys (Macaca mulatta). Environ Health Perspect 116:
1322-1326. http://dx.doi.org/10.1289/ehp. 11203.

Moreira. EG: Vassilieff. I: Vassilieff. VS. (2001). Developmental lead exposure: Behavioral alterations in the short
and long term. Neurotoxicol Teratol 23: 489-495. http://dx.doi.org/10.1016/S0892-0362(01)00159-3.

External Review Draft

3-513

DRAFT: Do not cite or quote

-------
Munoz. C: Garbe. K: Lilienthal. H: Winneke. G. (1989). Neuronal depletion of the amygdala resembles the learning
deficits induced by low level lead exposure in rats. Neurotoxicol Teratol 11: 257-264.
http://dx.doi.org/10.1016/0892-0362(89190068-8.

Mufioz. MP: Rubilar. P: Valdes. M: Mufioz-Ouezada. MT: Gomez. A: Saavedra. M: Iglesias. V. (2020). Attention
deficit hyperactivity disorder and its association with heavy metals in children from northern Chile. Int J
Hyg Environ Health 226: 113483. http://dx.doi.Org/10.1016/i.iiheh.2020.113483.

Murphy. PL: Patel. M: Kirrane. E: Vinikoor-Imler. L. (2013). Comments on: Chari, R.; Burke, T.A.; White, R.H.;
Fox, M.A. Integrating susceptibility into environmental policy: An analysis of the National Ambient Air
Quality Standard for lead. Int. J. Environ. Res. Public Health 2012, 9, 1077-1096 [Comment]. Int J Environ
Res Public Health 10: 712-716. http://dx.doi.org/10.3390/iierphlQ020712.

Naicker. N: Richter. L: Mathee. A: Becker. P: Norris. SA. (2012). Environmental lead exposure and socio-

behavioural adjustment in the early teens: The birth to twenty cohort. Sci Total Environ 414: 120-125.
http://dx.doi.org/10.1016/i.scitotenv.2011.11.013.

Nam. S: Seo. J: Nahm. SS: Chang. BJ. (2019a). Effects of ascorbic acid on osteopontin expression and axonal
myelination in the developing cerebellum of lead-exposed rat pups. Int J Environ Res Public Health 16:
983. http://dx.doi.org/10.3390/iierphl6060983.

Nam. SM: Ahn. SC: Go. TH: Seo. JS: Nahm. SS: Chang. BJ: Lee. JH. (2018a). Ascorbic acid ameliorates

gestational lead exposure-induced developmental alteration in GAD67 and c-kit expression in the rat
cerebellar cortex. Biol Trace Elem Res 182: 278-286. http://dx.doi.org/10.1007/sl2011-Q17-1086-z.

Nam. SM: Chang. BJ: Kim. JH: Nahm. SS: Lee. JH. (2018b). Ascorbic acid ameliorates lead-induced apoptosis in
the cerebellar cortex of developing rats. Brain Res 1686: 10-18.
http://dx.doi.Org/10.1016/i.brainres.2018.02.014.

Nam. SM: Cho. IS: Seo. JS: Go. TH: Kim. JH: Nahm. SS: Chang. BJ: Lee. JH. (2019b). Ascorbic acid attenuates
lead-induced alterations in the synapses in the developing rat cerebellum. Biol Trace Elem Res 187: 142-
150. http://dx.doi.org/10.1007/sl2011-Q18-1354-6.

Nam. SM: Choi. SH: Cho. HJ: Seo. JS: Choi. M: Nahm. SS: Chang. BJ: Nah. SY. (2020). Ginseng gintonin

attenuates lead-induced rat cerebellar impairments during gestation and lactation. Biomolecules 10: 385.
http://dx.doi.org/10.3390/biomlQ030385.

Nan. A: Zhou. X: Chen. L: Liu. M: Zhang. N: Zhang. L: Luo. Y: Liu. Z: Dai. L: Jiang. Y. (2016). A transcribed
ultraconserved noncoding RNA, Uc. 173, is a key molecule for the inhibition of lead-induced neuronal
apoptosis. Onct7: 112-124. http://dx.doi.org/10.18632/oncotarget.6590.

Needleman. HL: Gatsonis. CA. (1990). Low-level lead exposure and the IQ of children: A meta-analysis of modern
studies. JAMA 263: 673-678. http://dx.doi.org/10 1001 /iama. 1990.03440050067035.

Needleman. HL: Gunnoe. C: Leviton. A: Reed. R: Peresie. H: Maher. C: Barrett. P. (1979). Deficits in psychologic
and classroom performance of children with elevated dentine lead levels. N Engl J Med 300: 689-695.
http://dx.doi.org/10.1056/NEJM1979032930013Ql.

Needleman. HL: Schell. A: Bellinger. D: Leviton. A: Allred. EN. (1990). The long-term effects of exposure to low
doses of lead in childhood. An 11-year follow-up report. N Engl J Med 322: 83-88.
http://dx.doi.org/10.1056/NEJM199001113220203.

Neelima. A: Raianna. A: Bhanuprakash. RG: Chettv. CS: Suresh. C. (2017). Deleterious effects of combination of
lead and (3-amyloid peptides in inducing apoptosis and altering cell cycle in human neuroblastoma cells.
Interdiscip Toxicol 10: 93-98. http://dx.doi.org/10.1515/intox-2017-0Q15.

Nelson. MM: Espy. KA. (2009). Low-level lead exposure and contingency-based responding in preschoolers: An
exploratory study. Dev Neuropsychol 34: 494-506. http://dx.doi.org/10.1080/875656409Q2964565.

Neugebauer. J: Wittsiepe. J: Kasper-Sonnenberg. M: Schoneck. N: Scholmerich. A: Wilhelm. M. (2015). The

influence of low level pre- and perinatal exposure to PCDD/Fs, PCBs, and lead on attention performance
and attention-related behavior among German school-aged children: Results from the Duisburg Birth
Cohort Study. Int J Hyg Environ Health 218: 153-162. http://dx.doi.Org/10.1016/i.iiheh.2014.09.005.

External Review Draft

3-514

DRAFT: Do not cite or quote

-------
Neuwirth. LS: Emenike. BU: Barrera. ED: Hameed. N: Rubi. S: Dacius. TF: Skeen. JC: Bonitto. JR: Khairi. E:

Igbal. A: Ahmed. I: Jose. TJ: Lynch. K: Khan. M: Alvira. AL: Mathew. N: Kaur. S: Masood. S: Tranguilee.
B: Thiruverkadu. V. (2019a). Assessing the anxiolytic properties of taurine-derived compounds in rats
following developmental lead exposure: A neurodevelopmental and behavioral pharmacological pilot
study. In J Hu; F Piao; SW Schaffer; A Elldrissi; JY Wu (Eds.), Taurine 11 (pp. 801-819). Singapore,
Singapore: Springer, http://dx.doi.org/10.1007/978-981-13-8023-5 69.

Neuwirth. LS: Kim. Y: Barrerra. ED: Jo. C: Chrisphonte. JM: Hameed. N: Rubi. S: Dacius. TF. Jr: Skeen. JC:
Bonitto. JR: Khairi. E: Iqbal. A: Ahmed. I: Masood. S: Tranguilee. B: Thiruverkadu. V. (2019b). Early
neurodevelopmental exposure to low lead levels induces fronto-executive dysfunctions that are recovered
by taurine co-treatment in the rat attention set-shift test: Implications for taurine as a
psychopharmacotherapy against neurotoxicants. In J Hu; F Piao; SW Schaffer; A El Idrissi; JY Wu (Eds.),
Taurine 11 (pp. 821-846). Singapore, Singapore: Springer. http://dx.doi.org/10.1007/978-981-13-8Q23-
5 70.

Neuwirth. LS: Masood. S: Andersoa DW: Schneider. JS. (2019c). The attention set-shifting test is sensitive for
revealing sex-based impairments in executive functions following developmental lead exposure in rats.
Behav Brain Res 366: 126-134. http://dx.doi.Org/10.1016/i.bbr.2019.03.022.

Nguyen. HP: Oh. H: Hoang. NHM: Jo. WH: Kim. MS. (2022). Environmental science and pollution research role of
heavy metal concentrations and vitamin intake from food in depression: A national cross-sectional study
(2009-2017). Environ Sci Pollut Res Int 29: 4574-4586. http://dx.doi.org/10.1007/sll356-021-15986-w.

Nicolescu. R: Petcu. C: Cordeanu. A: Fabritius. K: Schlumpf. M: Krebs. R: Kramer. U: Winneke. G. (2010).

Environmental exposure to lead, but not other neurotoxic metals, relates to core elements of ADHD in
Romanian children: Performance and questionnaire data. Environ Res 110: 476-483.
http://dx.doi.Org/10.1016/i.envres.2010.04.002.

Nigg. J: Knottnerus. G: Martel. M: Nikolas. M: Cavanagh. K: Karmaus. W: Rapplev. M. (2008). Low blood lead
levels associated with clinically diagnosed attention-deficit/hyperactivity disorder and mediated by weak
cognitive control. Biol Psychiatry 63: 325-331. http://dx.doi.Org/10.1016/i.biopsvch.2007.07.013.

Nigg. JT: Elmore. AL: Nataraian. N: Friderici. KH: Nikolas. MA. (2016). Variation in an iron metabolism gene

moderates the association between blood lead levels and attention-deficit/hyperactivity disorder in children.
Psychol Sci 27: 257-269. http://dx.doi.org/10.1177/0956797615618365.

Nigg. JT: Nikolas. M: Knottnerus. GM: Cavanagh. K: Friderici. K. (2010). Confirmation and extension of

association of blood lead with attention-deficit/hyperactivity disorder (ADHD) and ADHD symptom
domains at population-typical exposure levels. J Child Psychol Psychiatry 51: 58-65.
http://dx.doi.org/10.1111/i. 1469-7610.2009.02135.X.

Niu. RY: Sun. ZL: Cheng. ZT: Li. ZG: Wang. JD. (2009). Decreased learning ability and low hippocampus
glutamate in offspring rats exposed to fluoride and lead. Environ Toxicol Pharmacol 28: 254-258.
http://dx.doi.Org/10.1016/i.etap.2009.04.012.

Nkomo. P: Mathee. A: Naicker. N: Galpin. J: Richter. LM: Norris. SA. (2017). The association between elevated
blood lead levels and violent behavior during late adolescence: The South African Birth to Twenty Plus
cohort. Environ Int 109: 136-145. http://dx.doi.Org/10.1016/i.envint.2017.09.004.

Nkomo. P: Naicker. N: Mathee. A: Galpin. J: Richter. LM: Norris. SA. (2018). The association between

environmental lead exposure with aggressive behavior, and dimensionality of direct and indirect aggression
during mid-adolescence: Birth to Twenty Plus cohort. Sci Total Environ 612: 472-479.
http://dx.doi.Org/10.1016/i.scitotenv.2017.08.138.

Nordberg. M: Winblad. B: Fratiglioni. L: Basun. H. (2000). Lead concentrations in elderly urban people related to
blood pressure and mental performance: Results from a population-based study. Am J Ind Med 38: 290-
294. http://dx.doi.org/10.1002/1097-0274(200009)38:3<290::AID-AJIM7>3.0.CQ:2-T.

Nozadi. SS: Li. L: Luo. L: MacKenzie. D: Erdei. E: Du. R: Roman. CW: Hoover. J: O'Donald. E: Burnette. C:

Lewis. J. (2021). Prenatal metal exposures and infants' developmental outcomes in a Navajo population. Int
J Environ Res Public Health 19: 425. http://dx.doi.org/10.3390/iierphl9010425.

External Review Draft

3-515

DRAFT: Do not cite or quote

-------
Nvanza. EC: Bernier. FP: Martin. JW: Manvama. M: Hatfield. J: Dewev. D. (2021). Effects of prenatal exposure
and co-exposure to metallic or metalloid elements on early infant neurodevelopmental outcomes in areas
with small-scale gold mining activities in Northern Tanzania. Environ Int 149: 106104.
http://dx.doi.org/10.1016/i.envint.2020.106104.

Okesola. MA: Aiibove. BO: Ovinlove. BE: Oio. OA. (2019). Neuromodulator effects of ethyl acetate fraction of
Zingiber officinale Roscoe extract in rats with lead-induced oxidative stress. J Integr Med 17: 125-131.
http://dx.doi.Org/10.1016/i.ioim.2019.01.002.

Qpler. MGA: Buka. SL: Groeger. J: McKeague. I: Wei. C: Factor-Litvak. P: Bresnahan. M: Graziano. J: Goldstein.
JM: Seidman. LJ: Brown. AS: Susser. ES. (2008). Prenatal exposure to lead, 5-aminolevulinic acid, and
schizophrenia: Further evidence. Environ Health Perspect 116: 1586-1590.
http://dx.doi.org/10.1289/ehp.10464.

Qppenheimer. AY: Bellinger. DC: Coull. BA: Weisskopf. MG: Korrick. SA. (2022). Prenatal exposure to chemical
mixtures and working memory among adolescents. Environ Res 205: 112436.
http://dx.doi.Org/10.1016/i.envres.2021.112436.

Ouvang. L: Zhang. W: Du. G: Liu. H: Xie. J: Gu. J: Zhang. S: Zhou. F: Shao. L: Feng. C: Fan. G. (2019). Lead
exposure-induced cognitive impairment through RyR-modulating intracellular calcium signaling in aged
rats. Toxicology 419: 55-64. http://dx.doi.Org/10.1016/i.tox.2019.03.005.

Paier. KA. (1998). What happens to "bad" girls? A review of the adult outcomes of antisocial adolescent girls
[Review]. Am J Psychiatry 155: 862-870. http://dx.doi.org/10.1176/aip. 155.7.862.

Palaniappan. K: Roy. A: Balakrishnan. K: Gopalakrishnan. L: Mukheriee. B: Hu. H: Bellinger. DC. (2011). Lead
exposure and visual-motor abilities in children from Chennai, India. Neurotoxicology 32: 465-470.
http://dx.doi.Org/10.1016/i.neuro.2011.03.011.

Paraiuli. RP: Fuiiwara. T: Umezaki. M: Watanabe. C. (2013). Association of cord blood levels of lead, arsenic, and
zinc with neurodevelopmental indicators in newborns: A birth cohort study in Chitwan Valley, Nepal.
EnvironRes 121: 45-51. http://dx.doi.Org/10.1016/i.envres.2012.10.010.

Paraiuli. RP: Fuiiwara. T: Umezaki. M: Watanabe. C. (2015a). Home environment and cord blood levels of lead,
arsenic, and zinc on neurodevelopment of 24 months children living in Chitwan Valley, Nepal. J Trace
Elem Med Biol 29: 315-320. http://dx.doi.Org/10.1016/i.itemb.2014.08.006.

Paraiuli. RP: Umezaki. M: Fuiiwara. T: Watanabe. C. (2015b). Association of cord blood levels of lead, arsenic, and
zinc and home environment with children neurodevelopment at 36 months living in Chitwan Valley, Nepal.
PLoS ONE 10: e0120992. http://dx.doi.org/10.1371/iournal.pone.0120992.

Park. JH: Sep. JH: Hong. YS: Kim. YM: Kang. JW: Yoo. JH: Chueh. HW: Lee. JH: Kwak. M.T: Kim. J: Woo. HP:
Kim. DW: Bang. YR: Choe. BM. (2016). Blood lead concentrations and attention deficit hyperactivity
disorder in Korean children: A hospital-based case control study. BMC Pediatr 16: 156.
http://dx.doi.org/10.1186/sl2887-016-0696-5.

Park. SK: Elmarsafawv. S: Mukheriee. B: Spiro. A. Ill: Vokonas. PS: Nie. H: Weisskopf. MG: Schwartz. J: Hu. H.

(2010). Cumulative lead exposure and age-related hearing loss: The VA Normative Aging Study. Hear Res
269: 48-55. http://dx.doi.Org/10.1016/i.heares.2010.07.004.

Paul. KC: Horvath. S: Del Rosario. I: Bronstein. JM: Ritz. B. (2021). DNA methylation biomarker for cumulative
lead exposure is associated with Parkinson's disease. Clinical Epigenetics 13: 59.
http://dx.doi.org/10.1186/sl3148-021-01051-3.

Paulsen. AJ: Schubert. CR: Johnson. LJ: Chen. Y: Dalton. PS: Klein. BEK: Klein. R: Pinto. A: Cruickshanks. KJ.
(2018). Association of cadmium and lead exposure with the incidence of contrast sensitivity impairment
among middle-aged adults. JAMA Ophthalmol 136: 1342-1350.
http://dx.doi.org/10.1001/iamaophthalmol.2018.3931.

Pawlas. N: Broberg. K: Olewinska. E: Kozlowska. A: Skerfving. S: Pawlas. K. (2015). Genetic modification of

ALAP and VPR on lead-induced impairment of hearing in children. Environ Toxicol Pharmacol 39: 1091-
1098. http://dx.doi.Org/10.1016/i.etap.2015.03.008.

External Review Praft

3-516

PRAFT: Po not cite or quote

-------
Pedroso. TF: Oliveira. CS: Fonseca. MM: Oliveira. VA: Pereira. ME. (2017). Effects of zinc and n-acetylcysteine in
damage caused by lead exposure in young rats. Biol Trace Elem Res 180: 275-284.
http://dx.doi.org/10.1007/sl2011-017-10Q9-z.

Peters. JL: Kubzanskv. LP: Ikeda. A: Spiro. A. Ill: Wright. RO: Weisskopf. MG: Kim. D: Sparrow. D: Nie. LH:
Hu. H: Schwartz. J. (2011). Childhood and adult socioeconomic position, cumulative lead levels, and
pessimism in later life: The VA Normative Aging Study. Am J Epidemiol 174: 1345-1353.
http://dx.doi.org/10.1093/aie/kwr269.

Phvu. MP: Tangpong. J. (2013). Protective effect of Thunbergia laurifolia (Linn.) on lead induced

acetylcholinesterase dysfunction and cognitive impairment in mice. BioMed Res Int 2013: 186098.
http://dx.doi.org/10.1155/2013/186098.

Pilsner. JR: Hu. H: Wright. RO: Kordas. K: Ettinger. AS: Sanchez. BN: Cantonwine. D: Lazarus. AL: Cantoral. A:
Schnaas. L: Tellez-Roio. MM: Hernandez-Avila. M. (2010). Maternal MTHFR genotype and haplotype
predict deficits in early cognitive development in a lead-exposed birth cohort in Mexico City. Am J Clin
Nutr 92: 226-234. http://dx.doi.org/10.3945/aicn.20Q9.28839.

Pocock. SJ: Smith. M: Baghurst. P. (1994). Environmental lead and children's intelligence: A systematic review of
the epidemiological evidence [Review]. Br Med J 309: 1189-1197.
http://dx.doi.org/10.1136/bmi.309.6963.1189.

Power. MC: Korrick. S: Tchetgen Tchetgen. EJ: Nie. LH: Grodstein. F: Hu. H: Weuve. J: Schwartz. J: Weisskopf.

MG. (2014). Lead exposure and rate of change in cognitive function in older women. Environ Res 129: 69-
75. http://dx.doi.Org/10.1016/i.envres.2013.12.010.

Przvbvla. J: Houseman. EA: Smit. E: Kile. ML. (2017). A path analysis of multiple neurotoxic chemicals and
cognitive functioning in older US adults (NHANES 1999-2002). Environ Health 16: 19.
http://dx.doi.org/10.1186/sl2940-017-0227-3.

Oin. YY: Jian. B: Wu. C: Jiang. CZ: Kang. Y: Zhou. JX: Yang. F: Liang. Y. (2018). A comparison of blood metal
levels in autism spectrum disorder and unaffected children in Shenzhen of China and factors involved in
bioaccumulation of metals. Environ Sci Pollut Res Int 25: 17950-17956. http://dx.doi.org/10.1007/sl 1356-
018-1957-7.

Rabinowitz. MB: Wang. JD: Soong. WT. (1992). Children's classroom behavior and lead in Taiwan. Bull Environ
Contam Toxicol 48: 282-288. http://dx.doi.org/10.1007/BF0Q194385.

Rahbar. MH: Ibrahim. SH: Azam. SI: Hessabi. M: Karim. F: Kim. S: Zhang. J: Ali. NG: Loveland. KA. (2021).

Concentrations of lead, mercury, arsenic, cadmium, manganese, and aluminum in the blood of Pakistani
children with and without autism spectrum disorder and their associated factors. Int J Environ Res Public
Health 18: 8625. http://dx.doi.org/10.3390/iierphl8168625.

Rahbar. MH: Samms-Vaughan. M: Dickerson. AS: Loveland. KA: Ardiomand-Hessabi. M: Bressler. J:

Shakespeare-Pellington. S: Grove. ML: Pearson. DA: Boerwinkle. E. (2015). Blood lead concentrations in
Jamaican children with and without Autism Spectrum Disorder. Int J Environ Res Public Health 12: 83-
105. http://dx.doi.org/10.3390/iierphl20100Q83.

Rahman. A: Al-Awadi. AA: Khan. KM. (2018). Lead affects vitamin D metabolism in rats. Nutrients 10: 264.
http://dx.doi.org/10.3390/nulQ030264.

Rahman. A: Khan. K: Al-Khaledi. G: Khan. I: Attur. S. (2012a). Early postnatal lead exposure induces tau
phosphorylation in the brain of young rats. Acta Biol Hung 63: 411 -425.
http://dx.doi.Org/10.1556/ABiol.63.2012.4.l.

Rahman. A: Khan. KM: Al-Khaledi. G: Khan. I: Al-Shemarv. T. (2012b). Over activation of hippocampal

serine/threonine protein phosphatases PP1 and PP2A is involved in lead-induced deficits in learning and
memory in young rats. Neurotoxicology 33: 370-383. http://dx.doi.Org/10.1016/i.neuro.2012.02.014.

Raian. P: Kelsev. KT: Schwartz. JD: Bellinger. DC: Weuve. J: Sparrow. D: Spiro. A. Ill: Smith. TJ: Nie. H: Hu. H:
Wright. RO. (2007). Lead burden and psychiatric symptoms and the modifying influence of the delta-
aminolevulinic acid dehydratase (ALAD) polymorphism: The VA Normative Aging Study. Am J
Epidemiol 166: 1400-1408. http://dx.doi.org/10.1093/aie/kwm220.

External Review Draft

3-517

DRAFT: Do not cite or quote

-------
Raian. P; Kelsev. KT: Schwartz. JD: Bellinger. DC: Weuve. J: Spiro. A. Ill; Sparrow. D: Smith. TJ: Nie. H:

Weisskopf. MG: Hu. H; Wright. RO. (2008). Interaction of the delta-aminolevulinic acid dehydratase
polymorphism and lead burden on cognitive function: The VA Normative Aging Study. J Occup Environ
Med 50: 1053-1061. http://dx.doi.org/10.1097/JQM.0b013e3181792463.

Rao Barkur. R: Bairv. LK. (20161. Comparison of the developmental milestones and preweaning neurobehavioral
parameters in rat pups exposed to lead (Pb) during gestation, lactation and pregestation period. Drug Chem
Toxicol 39: 248-255. http://dx.doi.org/10.3109/01480545.2Q15.1082136.

Rasnick. E: Ryan. PH: Bailer. AJ: Fisher. T: Parsons. PJ: Yolton. K: Newman. NC: Lanphear. BP; Brokamp. C.

(2021). Identifying sensitive windows of airborne lead exposure associated with behavioral outcomes at
age 12. Environmental Epidemiology 5: el44. http://dx.doi.org/10.1097/EE9.000000000000Q144.

Rawat. PS: Singh. S: Mahdi. AA; Mehrotra. S. (2022). Environmental lead exposure and its correlation with
intelligence quotient level in children. J Trace Elem Med Biol 72: 126981.
http://dx.doi.Org/10.1016/i.itemb.2022.126981.

Rea. P. (2015). Overview of the nervous system. In P Rea (Ed.), Essential clinically applied anatomy of the
peripheral nervous system in the limbs (pp. 1-40). London, United Kingdom: Academic Press.
http://dx.doi.org/10.1016/B978-0-12-803062-2.00Q01-2.

Reuben. A; Caspi. A; Belskv. DW: Broadbent. J: Harringtoa H; Sugden. K; Houts. RM; Ramrakha. S: Poulton. R;

Moffitt. TE. (2017). Association of childhood blood lead levels with cognitive function and socioeconomic
status at age 38 years and with IQ change and socioeconomic mobility between childhood and adulthood.
JAMA 317: 1244-1251. http://dx.doi.org/10.1001/iama.2017.1712.

Reuben. A; Elliott. ML; Abraham. WC: Broadbent. J: Houts. RM; Ireland. D: Knodt. AR: Poulton. R: Ramrakha. S:
Hariri. AR; Caspi. A; Moffitt. TE. (2020). Association of childhood lead exposure with MRI measurements
of structural brain integrity in midlife. JAMA 324: 1970-1979. http://dx.doi.org/10.100l/iama.2020.19998.

Reuben. A; Schaefer. JD; Moffitt. TE; Broadbent. J: Harrington. H; Houts. RM; Ramrakha. S: Poulton. R; Caspi. A.
(2019). Association of childhood lead exposure with adult personality traits and lifelong mental health.
JAMA Psychiatry 76: 418-425. http://dx.doi.org/10.1001/iamapsvchiatrv.2Q18.4192.

Reves. JW. (2015). Lead exposure and behavior: Effects on antisocial and risky behavior among children and
adolescents. Economic Inquiry 53: 1580-1605. http://dx.doi.org/10. Ill 1/ecin. 12202.

Reynolds. CR: Kamphaus. RW. (2015). BASC-3: Behavior assessment system for children (3rd ed.). Bloomington,
MN: NCS Pearson.

Rice. DC. (1990). Lead-induced behavioral impairment on a spatial discrimination reversal task in monkeys exposed
during different periods of development. Toxicol Appl Pharmacol 106: 327-333.
http://dx.doi.org/10.1016/0041-008X(90)90251-Q.

Rice. DC. (1992). Lead exposure during different developmental periods produces different effects on FI
performance in monkeys tested as juveniles and adults. Neurotoxicology 13: 757-770.

Rice. DC. (1997). Effects of lifetime lead exposure in monkeys on detection of pure tones. Fundam Appl Toxicol
36: 112-118. http://dx.doi.org/10.10Q6/faat.1996.2268.

Rice. DC. (1998). Effects of lifetime lead exposure on spatial and temporal visual function in monkeys.
Neurotoxicology 19: 893-902.

Rice. DC: Gilbert. SG. (1990a). Lack of sensitive period for lead-induced behavioral impairment on a spatial
delayed alternation task in monkeys. Toxicol Appl Pharmacol 103: 364-373.
http://dx.doi.org/10.1016/0041-008X(90)90236-N.

Rice. DC: Gilbert. SG. (1990b). Sensitive periods for lead-induced behavioral impairment (nonspatial discrimination
reversal) in monkeys. Toxicol Appl Pharmacol 102: 101-109. http://dx.doi.org/10.1016/0Q41-
008X(90)90087-B.

Rice. DC: Karpinski. KF. (1988). Lifetime low-level lead exposure produces deficits in delayed alternation in adult
monkeys. Neurotoxicol Teratol 10: 207-214. http://dx.doi.org/10.1016/0892-0362(88)90019-0.

External Review Draft

3-518

DRAFT: Do not cite or quote

-------
Ris. MP: Dietrich. KN: Succop. PA: Berger. QG: Bornschein. RL. (2004). Early exposure to lead and
neuropsychological outcome in adolescence. J Int Neuropsychol Soc 10: 261-270.
http://dx.doi.org/10.1017/S13556177041Q2154.

Rivenbark. JG: Odgers. CL: Caspi. A: Harrington. H: Hogan. S: Houts. RM: Poulton. R: Moffitt. TE. (2018). The
high societal costs of childhood conduct problems: evidence from administrative records up to age 38 in a
longitudinal birth cohort. J Child Psychol Psychiatry 59: 703-710. http://dx.doi.org/10. Ill 1/icpp. 12850.

Roda. SM: Greenland. RD: Bornschein. RL: Hammond. PB. (1988). Anodic stripping voltammetry procedure
modified for improved accuracy of blood lead analysis. Clin Chem 34: 563-567.
http://dx.doi.Org/10.1093/clinchem/34.3.563.

Rodrigues. ALS: Rocha. JBT: Mello. CF: Souza. DO. (1996). Effect of perinatal lead exposure on rat behaviour in
open-field and two-way avoidance tasks. Basic Clin Pharmacol Toxicol 79: 150-156.
http://dx.doi.org/10.1111/i. 1600-0773.1996.tb00259.x.

Rodrigues. EG: Bellinger. DC: Valeri. L: Hasan. MPS. I: Ouamruzzaman. O: Golam. M: Kile. ML: Christian! DC:
Wright. RO: Mazumdar. M. (2016). Neurodevelopmental outcomes among 2- to 3-year-old children in
Bangladesh with elevated blood lead and exposure to arsenic and manganese in drinking water. Environ
Health 15: 44. http://dx.doi.org/10.1186/sl2940-016-0127-v.

Rodrigues. JLG: Arauio. CFS: Dos Santos. NR: Bandeira. MJ: Anios. ALS: Carvalho. CF: Lima. CS: Abreu. JNS:
Mergler. D: Menezes-Filho. JA. (2018). Airborne manganese exposure and neurobehavior in school-aged
children living near a ferro-manganese alloy plant. Environ Res 167: 66-77.
http://dx.doi.Org/10.1016/i.envres.2018.07.007.

Rokoff. LB: Shoaff. JR: Coull. BA: Enlow. MB: Bellinger. DC: Korrick. SA. (2022). Prenatal exposure to a mixture
of organochlorines and metals and internalizing symptoms in childhood and adolescence. Environ Res 208:
112701. http://dx.doi.Org/10.1016/i.envres.2022.112701.

Roonev. JPK: Woods. NF: Martin. MP: Woods. JS. (2018). Genetic polymorphisms of GRIN2A and GRIN2B
modify the neurobehavioral effects of low-level lead exposure in children. Environ Res 165: 1-10.
http://dx.doi.Org/10.1016/i.envres.2018.04.001.

Rossi-George. A: Virgolini. MB: Weston. D: Thiruchelvam. M: Corv-Slechta. DA. (2011). Interactions of lifetime
lead exposure and stress: Behavioral, neurochemical and HPA axis effects. Neurotoxicology 32: 83-99.
http://dx.doi.Org/10.1016/i.neuro.2010.09.004.

Rothenberg. SJ: Schnaas. L: Salgado-Valladares. M: Casanueva. E: Geller. AM: Hudnell. HK: Fox. DA. (2002).
Increased ERG a- and b-wave amplitudes in 7- to 10-year-old children resulting from prenatal lead
exposure. Invest Ophthalmol Vis Sci 43: 2036-2044.

Roy. A: Hu. H: Bellinger. DC: Mukheriee. B: Modali. R: Nasaruddin. K: Schwartz. J: Wright. RO: Ettinger. AS:
Palaniapan. K: Balakrishnan. K. (2011). Hemoglobin, lead exposure, and intelligence quotient: Effect
modification by the DRD2 Taq IA polymorphism. Environ Health Perspect 119: 144-149.
http://dx.doi.org/10.1289/ehp.0901878.

Ruebner. RL: Hooper. SR: Parrish. C: Furth. SL: Fadrowski. JJ. (2019). Environmental lead exposure is associated
with neurocognitive dysfunction in children with chronic kidney disease. Pediatr Nephrol 34: 2371-2379.
http://dx.doi.org/10.1007/s00467-019-043Q6-7.

Ruff. HA: Biiur. PE: Markowitz. M: Ma. YC: Rosen. JF. (1993). Declining blood lead levels and cognitive changes
in moderately lead-poisoned children. JAMA 269: 1641-1646.
http://dx.doi.org/10 1001/iama. 1993.03500130055032.

Rygiel. CA: Dolinov. DC: Bakulski. KM: Aung. MT: Perng. W: Jones. TR: Solano-Gonzalez. M: Hu. H: Tellez-
Roio. MM: Schnaas. L: Marcela. E: Peterson. KE: Goodrich. JM. (2021). DNA methylation at birth
potentially mediates the association between prenatal lead (Pb) exposure and infant neurodevelopmental
outcomes. Environ Epigenet 7: dvab005. http://dx.doi.org/10.1093/eep/dvab005.

Sadeghi. A: Khordad. E: Ebrahimi. V: Raoofi. A: Alipour. F: Ebrahimzadeh-Bideskan. A. (2021). Neuroprotective
effects of vitamin C and garlic on glycoconjugates changes of cerebellar cortex in lead-exposed rat
offspring. J Chem Neuroanat 114: 101948. http://dx.doi.Org/10.1016/i.ichemneu.2021.101948.

External Review Draft

3-519

DRAFT: Do not cite or quote

-------
Saleh. HA: Abd El-Aziz. GS: Mustafa. HN: El-Fark. M: Mai. A: Aburas. M: Deifalla. AH. (2019). Thymoquinone
ameliorates oxidative damage and histopathological changes of developing brain neurotoxicity. J
Histotechnol 42: 116-127. http://dx.doi.org/10.1080/01478885.2Q19.1619654.

Saleh. HA: Abdel El-Aziz. GS: Mustafa. HN: Saleh. AHA: Mai. AO: Deifalla. AHS: Aburas. M. (2018). Protective
effect of garlic extract against maternal and foetal cerebellar damage induced by lead administration during
pregnancy in rats. Folia Morphol (Warsz) 77: 1-15. http://dx.doi.org/10.5603/FM.a2017.0Q63.

Salkever. PS. (1995). Updated estimates of earnings benefits from reduced exposure of children to environmental
lead. Environ Res 70: 1-6. http://dx.doi.org/10.1006/enrs.1995.1038.

Sanchez-Martin. FJ: Lindauist. DM: Landero-Figueroa. J: Zhang. X: Chen. J: Cecil. KM: Medvedovic. M: Puga. A.
(2015). Sex- and tissue-specific methylome changes in brains of mice perinatally exposed to lead.
Neurotoxicology 46: 92-100. http://dx.doi.Org/10.1016/i.neuro.2014.12.004.

Sanchez. BN: Hu. H: Litman. HJ: Tellez-Roio. MM. (2011). Statistical methods to study timing of vulnerability
with sparsely sampled data on environmental toxicants. Environ Health Perspect 119: 409-415.
http://dx.doi.org/10.1289/ehp. 1002453.

Sandhir. R: Gill. KD. (1995). Effect of lead on lipid peroxidation in liver of rats. Biol Trace Elem Res 48: 91-97.
http://dx.doi.org/10.1007/BF02789081.

Sasaki. N: Carpenter. DO. (2022). Associations between metal exposures and cognitive function in American older
adults. Int J Environ Res Public Health 19: 2327. http://dx.doi.org/10.3390/iierphl9042327.

Saxena. R: Gamble. M: Wasserman. GA: Liu. X: Parvez. F: Navas-Acien. A: Islam. T: Factor-Litvak. P: Uddin.

MN: Kioumourtzoglou. MA: Gibson. EA: Shahriar. H: Slavkovich. V: Ilievski. V: Loiacono. N: Balac. O:
Graziano. JH. (2022). Mixed metals exposure and cognitive function in Bangladeshi adolescents.
Ecotoxicol Environ Saf 232: 113229. http://dx.doi.Org/10.1016/i.ecoenv.2022.113229.

Schell. LM: Denham. M: Stark. AD: Gomez. M: Ravenscroft. J: Parsons. PJ: Avdermir. A: Samelson. R. (2003).

Maternal blood lead concentration, diet during pregnancy, and anthropometry predict neonatal blood lead
in a socioeconomically disadvantaged population. Environ Health Perspect 111: 195-200.
http://dx.doi.org/10.1289/ehp.5592.

Schneider. JS: Kidd. SK: Anderson. DW. (2013). Influence of developmental lead exposure on expression of DNA
methyltransferases and methyl cytosine-binding proteins in hippocampus. Toxicol Lett 217: 75-81.
http://dx.doi.Org/10.1016/i.toxlet.2012.12.004.

Schneider. JS: Mettil. W: Anderson. DW. (2012). Differential effect of postnatal lead exposure on gene expression
in the hippocampus and frontal cortex. J Mol Neurosci 47: 76-88. http://dx.doi.org/10.1007/sl2031-011-
9686-0.

Schwartz. J. (1994a). Low-level lead exposure and children's IQ: A meta-analysis and search for a threshold.
Environ Res 65: 42-55. http://dx.doi.org/10.1006/enrs.1994.102Q.

Schwartz. J. (1994b). Societal benefits of reducing lead exposure. Environmental Progress 66: 105-124.
http://dx.doi.org/10.1006/enrs. 1994.1048.

Schwartz. J: Otto. D. (1987). Blood lead, hearing thresholds, and neurobehavioral development in children and
youth. Arch Environ Health 42: 153-160. http://dx.doi.org/10.1080/00Q39896.1987.9935814.

Schwartz. J: Otto. D. (1991). Lead and minor hearing impairment. Arch Environ Health 46: 300-305.
http://dx.doi.org/10.1080/00Q39896.1991.9934391.

Sepehri. H: Ganii. F. (2016). The protective role of ascorbic acid on hippocampal CA1 pyramidal neurons in a rat
model of maternal lead exposure. J Chem Neuroanat 74: 5-10.
http://dx.doi.Org/10.1016/i.ichemneu.2016.01.005.

Shadbegian. R: Guignet. D: Klemick. H: Bui. L. (2019). Early childhood lead exposure and the persistence of
educational consequences into adolescence. Environ Res 178: 108643.
http://dx.doi.Org/10.1016/i.envres.2019.108643.

External Review Draft

3-520

DRAFT: Do not cite or quote

-------
Shargorodskv. J: Curhan. SG: Henderson. E: Eavev. R: Curhan. GC. (2011). Heavy metals exposure and hearing
lossinUS adolescents. Arch Otolaryngol Head Neck Surg 137: 1183-1189.
http://dx.doi.org/10.1001/archoto.2011.202.

Shekhawat. PS: Janu. VC: Singh. P: Sharma. P: Singh. K. (2021). Association of newborn blood lead concentration
with neurodevelopment outcome in early infancy. J Trace Elem Med Biol 68: 126853.
http://dx.doi.Org/10.1016/i.itemb.2021.126853.

Shiue. I. (2013). Urinary environmental chemical concentrations and vitamin D are associated with vision, hearing,
and balance disorders in the elderly. Environ Int 53: 41-46. http://dx.doi.Org/10.1016/i.envint.2012.12.006.

Shvachiv. L: Geraldes. V: Amaro-Leal. A: Rocha. I. (2018). Intermittent low-level lead exposure provokes anxiety,
hypertension, autonomic dysfunction and neuroinflammation. Neurotoxicology 69: 307-319.
http://dx.doi.Org/10.1016/i.neuro.2018.08.001.

Shvachiv. L: Geraldes. V: Amaro-Leal. A: Rocha. I. (2020). Persistent effects on cardiorespiratory and nervous

systems induced by long-term lead exposure: Results from a longitudinal study. Neurotox Res 37: 857-870.
http://dx.doi.org/10.1007/sl2640-020-0Q162-8.

Silbergeld. EK. (1991). Lead in bone: Implications for toxicology during pregnancy and lactation [Review]. Environ
Health Perspect 91: 63-70. http://dx.doi.org/10.2307/3430984.

Silva. PA: Hughes. P: Williams. S: Faed. JM. (1988). Blood lead, intelligence, reading attainment, and behaviour in
eleven year old children in Dunedin, New Zealand. J Child Psychol Psychiatry 29: 43-52.
http://dx.doi.org/10.1111/i. 1469-7610.1988.tb00687.x.

Silver. MK: Li. XO: Liu. YH: Li. M: Mai. XO: Kaciroti. N: Kilenv. P: Tardif. T: Meeker. JD: Lozoff. B. (2016).
Low-level prenatal lead exposure and infant sensory function. Environ Health 15: 65.
http://dx.doi.org/10.1186/sl2940-016-0148-6.

Singh. PK: Nath. R: Ahmad. MK: Rawat. A: Babu. S: Dixit. RK. (2016). Attenuation of lead neurotoxicity by
supplementation of polyunsaturated fatty acid in Wistar rats. Nutr Neurosci 19: 396-405.
http://dx.doi.org/10.1179/1476830515Y.0000000Q28.

Singh. PK: Singh. MK: Yadav. RS: Dixit. RK: Mehrotra. A: Nath. R. (2017). Attenuation of lead-induced
neurotoxicity by omega-3 fatty acid in rats. Ann Neurosci 24: 221-232.
http://dx.doi.org/10.1159/0004818Q8.

Singh. PK: Singh. MK: Yadav. RS: Nath. R: Mehrotra. A: Rawat. A: Dixit. RK. (2019). Omega-3 fatty acid
attenuates oxidative stress in cerebral cortex, cerebellum, and hippocampus tissue and improves
neurobehavioral activity in chronic lead-induced neurotoxicity. Nutr Neurosci 22: 83-97.
http://dx.doi.org/10.1080/1028415X.2Q17.1354542.

Sioen. I: Den Hond. E: Nelen. V: Van de Mieroop. E: Croes. K: Van Larebeke. N: Nawrot. TS: Schoeters. G.
(2013). Prenatal exposure to environmental contaminants and behavioural problems at age 7-8 years.
Environ Int 59: 225-231. http://dx.doi.Org/10.1016/i.envint.2013.06.014.

Skalnv. AV: Simashkova. NV: Klvushnik. TP: Grabeklis. AR: Biorklund. G: Skalnava. MG: Nikonorov. AA:
Tinkov. AA. (2017). Hair toxic and essential trace elements in children with autism spectrum disorder.
Metab Brain Dis 32: 195-202. http://dx.doi.org/10.1007/sll011-Q16-9899-6.

Skerfving. S: Lofmark. L: Lundh. T: Mikoczv. Z: Stromberg. U. (2015). Late effects of low blood lead

concentrations in children on school performance and cognitive functions. Neurotoxicology 49: 114-120.
http://dx.doi.Org/10.1016/i.neuro.2015.05.009.

Skogheim. TS: Wevde. KVF: Engel. SM: Aase. H: Suren. P: 0ie. MG: Biele. G: Reichborn-Kiennerud. T:

Caspersen. IH: Hornig. M: Haug. LS: Villanger. GD. (2021). Metal and essential element concentrations
during pregnancy and associations with autism spectrum disorder and attention-deficit/hyperactivity
disorder in children. Environ Int 152: 106468. http://dx.doi.Org/10.1016/i.envint.2021.106468.

External Review Draft

3-521

DRAFT: Do not cite or quote

-------
Sobin. C: Flores-Montova. MG: Gutierrez. M: Parisi. N: Schaub. T. (2015). 5-Aminolevulinic acid dehydratase

single nucleotide polymorphism 2 (ALAD2) and peptide transporter 2*2 haplotype (hPEPT2*2) differently
influence neurobehavior in low-level lead exposed children. Neurotoxicol Teratol 47: 137-145.
http://dx.doi.Org/10.1016/i.ntt.2014.12.001.

Sobin. C: Montova. MGF: Parisi. N: Schaub. T: Cervantes. M: Armiios. RX. (2013). Microglial disruption in young
mice with early chronic lead exposure. Toxicol Lett 220: 44-52.
http://dx.doi.Org/10.1016/i.toxlet.2013.04.003.

Sobolewski. M: Abston. K: Conrad. K: Marvin. E: Harvey. K: Susiario. M: Corv-Slechta. DA. (2020). Lineage- and
sex-dependent behavioral and biochemical transgenerational consequences of developmental exposure to
lead, prenatal stress, and combined lead and prenatal stress in mice. Environ Health Perspect 128: 27001.
http://dx.doi.org/10.1289/EHP4977.

Sobolewski. M: Varma. G: Adams. B: Anderson. DW: Schneider. JS: Corv-Slechta. DA. (2018). Developmental
lead exposure and prenatal stress result in sex-specific reprograming of adult stress physiology and
epigenetic profiles in brain. Toxicol Sci 163: 478-489. http://dx.doi.org/10.1093/toxsci/kfV046.

Soderstrom. H: Siodin. AK: Carlstedt. A: Forsman. A. (2004). Adult psychopathic personality with childhood-onset
hyperactivity and conduct disorder: A central problem constellation in forensic psychiatry. Psychiatry Res
121: 271-280. http://dx.doi.org/10.1016/S0165-1781(03)00270-1.

Soetrisno. FN: Delgado-Saborit. JM. (2020). Chronic exposure to heavy metals from informal e-waste recycling
plants and children's attention, executive function and academic performance. Sci Total Environ 717:
137099. http://dx.doi.Org/10.1016/i.scitotenv.2020.137099.

Song. H: Zheng. G: Shea XF: Liu. XO: Luo. WJ: Chen. JY. (2014). Reduction of brain barrier tight junctional
proteins by lead exposure: Role of activation of nonreceptor tyrosine kinase Src via chaperon GRP78.
Toxicol Sci 138: 393-402. http://dx.doi.org/10.1093/toxsci/kfu007.

Souza-Talarico. JN: Marcourakis. T: Barbosa. F. Jr: Barros. SBM: Rivelli. DP: Pompeia. S: Caramelli. P:

Plusauellec. P: Lupien. SJ: Catucci. RF: Alves. AR: Suchecki. D. (2017). Association between heavy metal
exposure and poor working memory and possible mediation effect of antioxidant defenses during aging. Sci
Total Environ 575: 750-757. http://dx.doi.Org/10.1016/i.scitotenv.2016.09.121.

SouzaLisboa. SFD: Goncalves. G: Komatsu. F: Oueiroz. CAS: Almeida. AA: Moreira. EG. (2005). Developmental
lead exposure induces depressive-like behavior in female rats. Drug Chem Toxicol 28: 67-77.
http://dx.doi.org/10.1081/DCT-39696.

Stangle. DE: Smith. PR: Beaudin. SA: Strawderman. MS: Levitskv. DA: Strupp. BJ. (2007). Succimer chelation
improves learning, attention, and arousal regulation in lead-exposed rats but produces lasting cognitive
impairment in the absence of lead exposure. Environ Health Perspect 115: 201-209.
http://dx.doi.org/10.1289/ehp.9263.

Stansfield. KH: Ruby. KN: Spares. BP: McGlothan. JL: Liu. X: Guilarte. TR. (2015). Early-life lead exposure

recapitulates the selective loss of parvalbumin-positive GABAergic interneurons and subcortical dopamine
system hyperactivity present in schizophrenia. Transl Psychiatry 5: e522.
http://dx.doi.org/10.1038/tp.2Q14.147.

Stavner. L: Steenland. K: Posemeci. M: Hertz-Picciotto. I. (2003). Attenuation of exposure-response curves in
occupational cohort studies at high exposure levels. Scand J Work Environ Health 29: 317-324.
http://dx.doi.org/10.5271/siweh.737.

Sterling. G: O'Neill. K: McCaffertv. M: O'Neill. J. (1982). Effect of chronic lead ingestion by rats on glucose
metabolism and acetylcholine synthesis in cerebral cortex slices. J Neurochem 39: 592-596.
http://dx.doi.org/10.1111/i. 1471-4159.1982.tb03989.x.

Stewart. PW: Blaine. C: Cohen. M: Burright. RG: Ponovick. PJ. (1996). Acute and longer term effects of meso-2,3
dimercaptosuccinic acid (PMSA) on the behavior of lead-exposed and control mice. Physiol Behav 59:
849-855. http://dx.doi.org/10.1016/0031-9384(95)02185-X.

Stiles. KM: Bellinger. PC. (1993). Neuropsychological correlates of low-level lead exposure in school-age children:
A prospective study. Neurotoxicol Teratol 15: 27-35. http://dx.doi.org/10.1016/0892-0362(93)90042-M.

External Review Praft

3-522

PRAFT: Po not cite or quote

-------
Su. P: Zhang. J: Wang. S: Aschner. M: Cao. Z: Zhao. F: Wang. D: Chen. J: Luo. W. (2016). Genistein alleviates
lead-induced neurotoxicity in vitro and in vivo: Involvement of multiple signaling pathways.
Neurotoxicology 53: 153-164. http://dx.doi.Org/10.1016/i.neuro.2015.12.019.

Sumner. SA: Mercy. JA: Dahlberg. LL: Hillis. SD: Klevens. J: Hourv. D. (2015). Violence in the United States:
Status, Challenges, and Opportunities. JAMA 314: 478-488. http://dx.doi.org/10.1001/iama.2Q15.8371.

Sun. H: Chen. W: Wang. D: Jin. Y: Chen. X: Xu. Y: Huang. L. (2015). Inverse association between intelligence
quotient and urinary retinol binding protein in Chinese school-age children with low blood lead levels:
Results from a cross-sectional investigation. Chemosphere 128: 155-160.
http://dx.doi.Org/10.1016/i.chemosphere.2015.01.036.

Sun. L: Zhou. XL: Yi. HP: Jiang. SJ: Yuan. H. (2014). Lead-induced morphological changes and amyloid precursor
protein accumulation in adult rat hippocampus. Biotech Histochem 89: 513-517.
http://dx.doi.org/10.3109/10520295.2014.9Q4926.

Surkan. PJ: Schnaas. L: Wright. RJ: Tellez-Roio. MM: Lamadrid-Figueroa. H: Hu. H: Hernandez-Avila. M:

Bellinger. DC: Schwartz. J: Perroni. E: Wright. RO. (2008). Maternal self-esteem, exposure to lead, and
child neurodevelopment. Neurotoxicology 29: 278-285. http://dx.doi.org/10.1016/i.neuro.2007.11.006.

Surkan. PJ: Zhang. A: Trachtenberg. F: Daniel. DB: McKinlav. S: Bellinger. DC. (2007). Neuropsychological
function in children with blood lead levels <10 (ig/dL. Neurotoxicology 28: 1170-1177.
http://dx.doi.Org/10.1016/i.neuro.2007.07.007.

Suszkiw. J: Toth. G: Murawskv. M: Cooper. GP. (1984). Effects of Pb2+ and Cd2+ on acetylcholine release and
Ca2+ movements in synaptosomes and subcellular fractions from rat brain and Torpedo electric organ.
Brain Res 323: 31-46. http://dx.doi.org/10.1016/0006-8993(84)90262-2.

Takahashi. A: Miczek. KA. (2014). Neurogenetics of aggressive behavior: Studies in rodents. In K Miczek; A
Meyer-Lindenberg (Eds.), Neuroscience of aggression (pp. 3-44). Berlin, Germany: Springer.
http://dx.doi.org/10.1007/7854 2013 263.

Tartaglione. AM: Serafini. MM: Raggi. A: Iacoponi. F: Zianni. E: Scalfari. A: Minghetti. L: Ricceri. L: Cubadda. F:
Calamandrei. G: Viviani. B. (2020). Sex-dependent effects of developmental lead exposure in Wistar rats:
Evidence from behavioral and molecular correlates. International Journal of Molecular Sciences 21: 2664.
http://dx.doi.org/10.3390/iims21082664.

Tassiopoulos. K: Huo. Y: Braun. J: Williams. PL: Smith. R: Aschengrau. A: Nichols. S: Hazra. R: Mever. WA. Ill:
Knapp. K: Devgoo. NS: Seage. GR. III. (2017). Blood lead levels and neurodevelopmental function in
perinatally HIV-exposed, uninfected children in a U.S.-based longitudinal cohort study. AIDS Res Hum
Retroviruses 33: 919-928. http://dx.doi.org/10.1089/aid.2016.0265.

Tatsuta. N: Nakai. K: Kasanuma. Y: Iwai-Shimada. M: Sakamoto. M: Murata. K: Satoh. H. (2020). Prenatal and
postnatal lead exposures and intellectual development among 12-year-old Japanese children. Environ Res
189: 109844. http://dx.doi.org/10.1016/i.envres.2020.109844.

Tatsuta. N: Nakai. K: Murata. K: Suzuki. K: Iwai-Shimada. M: Kurokawa. N: Hosokawa. T: Satoh. H. (2014).

Impacts of prenatal exposures to polychlorinated biphenyls, methylmercury, and lead on intellectual ability
of 42-month-old children in Japan. Environ Res 133: 321-326.
http://dx.doi.Org/10.1016/i.envres.2014.05.024.

Tatsuta. N: Nakai. K: Murata. K: Suzuki. K: Iwai-Shimada. M: Yaginuma-Sakurai. K: Kurokawa. N: Nakamura. T:
Hosokawa. T: Satoh. H. (2012). Prenatal exposures to environmental chemicals and birth order as risk
factors for child behavior problems. Environ Res 114: 47-52.
http://dx.doi.Org/10.1016/i.envres.2012.02.001.

Tavakoli-Nezhad. M: Barron. AJ: Pitts. DK. (2001). Postnatal inorganic lead exposure decreases the number of
spontaneously active midbrain dopamine neurons in the rat. Neurotoxicology 22: 259-269.
http://dx.doi.org/10.1016/S0161-813X(01)00010-9.

Taylor. CM: Emond. AM: Lingam. R: Golding. J. (2018). Prenatal lead, cadmium and mercury exposure and

associations with motor skills at age 7 years in a UK observational birth cohort. Environ Int 117: 40-47.
http://dx.doi.Org/10.1016/i.envint.2018.04.032.

External Review Draft

3-523

DRAFT: Do not cite or quote

-------
Taylor. CM: Humphriss. R: Hall. A: Golding. J: Emond. AM. (2015). Balance ability in 7- and 10-year-old children:
Associations with prenatal lead and cadmium exposure and with blood lead levels in childhood in a
prospective birth cohort study. BMJ Open 5: e009635. http://dx.doi.org/10.1136/bmiopen-2015-009635.

Taylor. CM: Kordas. K: Golding. J: Emond. AM. (2017). Effects of low-level prenatal lead exposure on child IQ at
4 and 8 years in a UK birth cohort study. Neurotoxicology 62: 162-169.
http://dx.doi.Org/10.1016/i.neuro.2017.07.003.

Thangaraian. S: Vedagiri. A: Somasundaram. S: Sakthimanogaran. R: Murugesan. M. (2018). Neuroprotective

effect of morin on lead acetate- induced apoptosis by preventing cytochrome c translocation via regulation
of Bax/Bcl-2 ratio. Neurotoxicol Teratol 66: 35-45. http://dx.doi.Org/10.1016/i.ntt.2018.01.006.

Tian. Y: Green. PG: Stamova. B: Hertz-Picciotto. I: Pessah. IN: Hansen. R: Yang. X: Gregg. JP: Ashwood. P:

Jickling. G: Van de Water. J: Sharp. FR. (2011). Correlations of gene expression with blood lead levels in
children with autism compared to typically developing controls. Neurotox Res 19: 1-13.
http://dx.doi.org/10.1007/sl2640-0Q9-9126-x.

Tlotleng. N: Naicker. N: Mathee. A: Todd. AC: Nkomo. P: Norris. SA. (2022). Association between bone lead
concentration and aggression in youth from a sub-cohort of the birth to twenty cohort. Int J Environ Res
Public Health 19: 2200. http://dx.doi.org/10.3390/iierphl904220Q.

Tong. S: McMichael. AJ: Baghurst. PA. (2000). Interactions between environmental lead exposure and
sociodemographic factors on cognitive development. Arch Environ Health 55: 330-335.
http://dx.doi.org/10.1080/00039890009604Q25.

Totsika. V: Svlva. K. (2004). The home observation for measurement of the environment revisited. Child and
Adolescent Mental Health 9: 25-35. http://dx.doi.Org/10.1046/i.1475-357X.2003.00073.x.

Trombini. TV: Pedroso. CG: Ponce. D: Almeida. AA: Godinho. AF. (2001). Developmental lead exposure in rats: Is
a behavioral sequel extended at F2 generation? Pharmacol BiochemBehav 68: 743-751.
http://dx.doi.org/10.1016/S0091-3057(01)00473-7.

Tu. YO: Fan. GR: Wu. N: Wu. H: Xiao. HJ. (2021). Association of plasma lead, cadmium and selenium levels with
hearing loss in adults: National Health and Nutrition Examination Survey (NHANES) 2011-2012. Br J Nutr
128: 1100-1107. http://dx.doi.org/10.1017/S00071145210Q4335.

Tung. PW: Burt. A: Karagas. M: Jackson. BP: Punshon. T: Lester. B: Marsit. CJ. (2022). Association between

placental toxic metal exposure and NICU Network Neurobehavioral Scales (NNNS) profiles in the Rhode
Island Child Health Study (RICHS). Environ Res 204: 111939.
http://dx.doi.Org/10.1016/i.envres.2021.111939.

Turnbull. DH: Mori. S. (2007). MRI in mouse developmental biology [Review]. NMR Biomed 20: 265-274.
http://dx.doi.org/10.1002/nbm. 1146.

U.S. EPA (U.S. Environmental Protection Agency). (1986). Air quality criteria for lead [EPA Report]. (EPA/600/8-
83/028aF-dF). Research Triangle Park, NC. http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=32647.

U.S. EPA (U.S. Environmental Protection Agency). (2006a). Air quality criteria for lead [EPA Report].
(EPA/600/R-05/144aF-bF). Research Triangle Park, NC.
https://cfpub.epa.gov/ncea/risk/recordisplay.cfm?deid=158823.

U.S. EPA (U.S. Environmental Protection Agency). (2006b). Air quality criteria for lead (Final report, 2006):
Volume I of II [EPA Report], (EPA/600/R-05/144aF). Washington, DC.
http ://cfpub .epa. gov/ncea/CFM/recordisplav ,cfm?deid= 158823.

U.S. EPA (U.S. Environmental Protection Agency). (2006c). Air quality criteria for ozone and related
photochemical oxidants. (EPA/600/R-05/004aF-cF). Research Triangle Park, NC.
http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=149923.

U.S. EPA (U.S. Environmental Protection Agency). (2013a). Integrated science assessment for lead [EPA Report].
(EPA/600/R-10/075F). Washington, DC. https://nepis.epa. gov/Exe/ZvPURL,cgi?Dockev=P 100K82L.txt.

External Review Draft

3-524

DRAFT: Do not cite or quote

-------
U.S. EPA (U.S. Environmental Protection Agency). (2013b). Integrated science assessment for ozone and related
photochemical oxidants [EPA Report]. (EPA/600/R-10/076F). Research Triangle Park, NC: U.S.
Environmental Protection Agency, Office of Research and Development, National Center for
Environmental Assessment-RTP Division. http://cfpub.epa.gov/ncea/isa/recordisplav.cfm?deid=247492.

U.S. EPA (U.S. Environmental Protection Agency). (2015). Preamble to the Integrated Science Assessments [EPA
Report]. (EPA/600/R-15/067). Research Triangle Park, NC: U.S. Environmental Protection Agency, Office
of Research and Development, National Center for Environmental Assessment, RTP Division.
https ://cfpub .epa. gov/ncea/isa/recordisplav. cfm?deid=310244.

Valeri. L: Mazumdar. MM: Bobb. JF: Henn. BC: Rodrigues. E: Sharif. OlA: Kile. ML: Ouamruzzaman. O: Afroz.
S: Golam. M: Amarasiriwardena. C: Bellinger. DC: Christiani. DC: Coull. BA: Wright. RO. (2017). The
joint effect of prenatal exposure to metal mixtures on neurodevelopmental outcomes at 20-40 months of
age: Evidence from rural Bangladesh. Environ Health Perspect 125: 067015.
http://dx.doi.org/10.1289/EHP614.

van Wiingaarden. E: Winters. PC: Corv-Slechta. DA. (2011). Blood lead levels in relation to cognitive function in
older U.S. adults. Neurotoxicology 32: 110-115. http://dx.doi.org/10.1016/i.neuro.2010.11.002.

Vega-Dienstmaier. JM: Salinas-Pielago. JE: del Rosario Gutierrez-Campos. M: Mandamiento-Avauipa. RD: del
Carmen Yara-Hokama. M: Ponce-Canchihuaman. J: Castro-Morales. J. (2006). Lead levels and cognitive
abilities in Peruvian children. Braz J Psychiatry 28: 33-39. http://dx.doi.org/10.1590/S1516-
44462006000100008.

Verma. M: Schneider. JS. (2017). Strain specific effects of low level lead exposure on associative learning and
memory in rats. Neurotoxicology 62: 186-191. http://dx.doi.Org/10.1016/i.neuro.2017.07.006.

Verma. SK: Dua. R: Gill. KD. (2005). Impaired energy metabolism after co-exposure to lead and ethanol. Basic Clin
Pharmacol Toxicol 96: 475-479. http://dx.doi.org/10.1111/i. 1742-7843.2005.pto 96611.x.

Vigeh. M: Yokovama. K: Matsukawa. T: Shinohara. A: Ohtani. K. (2014). Low level prenatal blood lead adversely
affects early childhood mental development. J Child Neurol 29: 1305-1311.
http://dx.doi.org/10.1177/0883073813516999.

Vigueras-Villasefior. RM: Chavez-Saldafia. MP: Landero-Huerta. DA: Montes. S: Rios. C: Roias. P: Molina-

Obregoa HA: Duran. P: Roias-Castafieda. JC. (2021). Chronic lead exposure alters photic entrainment of
locomotor activity rhythm and neuronal photoactivation in the suprachiasmatic nucleus of the adult rat. J
Chem Neuroanat 117: 101991. http://dx.doi.Org/10.1016/i.ichemneu.2021.101991.

Villa-Cedillo. SA: Nava-Hernandez. MP: Soto-Dominguez. A: Hernandez-Ibarra. JA: Perez-Truiillo. JJ: Saucedo-
Cardenas. O. (2019). Neurodegeneration, demyelination, and astrogliosis in rat spinal cord by chronic lead
treatment. Cell Biol Int 43: 706-714. http://dx.doi.org/10.1002/cbin. 11147.

Vinceti. M: Filippini. T: Mandrioli. J: Violi. F: Bargellini. A: Weuve. J: Fini. N: Grill. P: Michalke. B. (2017). Lead,
cadmium and mercury in cerebrospinal fluid and risk of amyotrophic lateral sclerosis: A case-control study.
J Trace Elem Med Biol 43: 121-125. http://dx.doi.Org/10.1016/i.itemb.2016.12.012.

Vinceti. M: Guidetti. D: Bergomi. M: Caselgrandi. E: Vivoli. R: Olmi. M: Rinaldi. L: Rovesti. S: Solime. F. (1997).
Lead, cadmium, and selenium in the blood of patients with sporadic amyotrophic lateral sclerosis. Ital J
Neurol Sci 18: 87-92. http://dx.doi.org/10.1007/BFQ1999568.

Virgolini. MB: Rossi-George. A: Weston. D: Corv-Slechta. DA. (2008). Influence of low level maternal Pb

exposure and prenatal stress on offspring stress challenge responsivity. Neurotoxicology 29: 928-939.
http://dx.doi.Org/10.1016/i.neuro.2008.09.010.

Volman. V: Behrens. MM: Seinowski. TJ. (2011). Downregulation of parvalbumin at cortical GABA synapses
reduces network gamma oscillatory activity. J Neurosci 31: 18137-18148.
http://dx.doi.org/10.1523/JNEUROSCI.3041-ll.2011.

Wan. C: Pan. S: Lin. L: Li. J: Dong. G: Jones. KC: Liu. F: Li. D: Liu. J: Yu. Z: Zhang. G: Ma. H (2021). DNA
methylation biomarkers of IQ reduction are associated with long-term lead exposure in school aged
children in Southern China. Environ Sci Technol 55: 412-422. http://dx.doi.org/10.1021/acs.est.0c01696.

External Review Draft

3-525

DRAFT: Do not cite or quote

-------
Wang. DH: Xu. H: Zheng. YH: Gu. PS: Zhu. YJ: Ren. Y: Wang. SC.': Yang. L: Xu. LW. (2020). Environmental
exposure to lead and cadmium and hearing loss in Chinese adults: A case-control study. PLoS ONE 15:
e0233165. http://dx.doi.org/10.1371/iournal.pone.Q233165.

Wang. FT: Hu. H: Schwartz. J: Weuve. J: Spiro. AS. Ill: Sparrow. D: Nie. HL: Silverman. EK: Weiss. ST: Wright.
RO. (2007). Modifying effects of the HFE polymorphisms on the association between lead burden and
cognitive decline. Environ Health Perspect 115: 1210-1215. http://dx.doi.org/10.1289/ehp.9855.

Wang. R: Wu. Z: Bai. L: Liu. R: Ba. Y: Zhang. H: Cheng. X: Zhou. G: Huang. H. (2021a). Resveratrol improved
hippocampal neurogenesis following lead exposure in rats through activation of SIRT1 signaling. Environ
Toxicol 36: 1664-1673. http://dx.doi.org/10.1002/tox.23162.

Wang. R: Wu. Z: Liu. M: Wu. Y: Li. O: Ba. Y: Zhang. H: Cheng. X: Zhou. G: Huang. H. (2021b). Resveratrol
reverses hippocampal synaptic markers injury and SIRT1 inhibition against developmental Pb exposure.
Brain Res 1767: 147567. http://dx.doi.Org/10.1016/i.brainres.2021.147567.

Wang. T: Guan. RL: Liu. MC: Shen. XF: Chen. JY: Zhao. MG: Luo. WJ. (2016). Lead exposure impairs

hippocampus related learning and memory by altering synaptic plasticity and morphology during juvenile
period. Mol Neurobiol 53: 3740-3752. http://dx.doi.org/10.1007/sl2035-Q15-9312-l.

Wang. X: Miller. G: Ding. G: Lou. X: Cai. D: Chen. Z: Meng. J: Tang. J: Chu. C: Mo. Z: Han. J. (2012). Health risk
assessment of lead for children in tinfoil manufacturing and e-waste recycling areas of Zhejiang Province,
China. Sci Total Environ 426: 106-112. http://dx.doi.Org/10.1016/i.scitotenv.2012.04.002.

Wang. XM: Liu. WJ: Zhang. R: Zhou. YK. (2013). Effects of exposure to low-level lead on spatial learning and

memory and the expression of mGluRl, NMDA receptor in different developmental stages of rats. Toxicol
Ind Health 29: 686-696. http://dx.doi.org/10.1177/0748233712436641.

Wang. Y: Wang. Y: Yan. C. (2022). Gender differences in trace element exposures with cognitive abilities of

school-aged children: A cohort study in Wujiang City, China. Environ Sci Pollut Res Int 29: 64807-64821.
http://dx.doi.org/10.1007/sl 1356-022-20353-4.

Wasserman. G: Graziano. JH: Factor-Litvak. P: Popovac. D: Morina. N: Musabegovic. A: Vrenezi. N: Capuni-
Paracka. S: Lekic. V: Preteni-Rediepi. E: Hadzialievic. S: Slavkovich. V: Kline. J: Shrout. P: Stein. Z.
(1992). Independent effects of lead exposure and iron deficiency anemia on developmental outcome at age
2 years. J Pediatr 121: 695-703. http://dx.doi.org/10.1016/S0022-3476(05)81895-5.

Wasserman. GA: Factor-Litvak. P. (2001). Methodology, inference and causation: Environmental lead exposure and
childhood intelligence [Comment]. Arch Clin Neuropsychol 16: 343-352. http://dx.doi.org/10.1016/S0887-
6177(00)00085-8.

Wasserman. GA: Liu. X: Pine. PS: Graziano. JH. (2001). Contribution of maternal smoking during pregnancy and
lead exposure to early child behavior problems. Neurotoxicol Teratol 23: 13-21.
http://dx.doi.org/10.1016/S0892-0362(00)00116-1.

Wasserman. GA: Musabegovic. A: Liu. X: Kline. J: Factor-Litvak. P: Graziano. JH. (2000). Lead exposure and
motor functioning in 4 1/2-year-old children: The Yugoslavia prospective study. J Pediatr 137: 555-561.
http://dx.doi.org/10.1067/mpd.2000.109111.

Wasserman. GA: Staghezza-Jaramillo. B: Shrout. P: Popovac. D: Graziano. J. (1998). The effect of lead exposure on
behavior problems in preschool children. Am J Public Health 88: 481-486.
http://dx.doi.Org/10.2105/AJPH.88.3.481.

Webster. R. (2001). Neurotransmitters, drugs and brain function. New York, NY: Wiley, https://www.wilev.com/en-
us/Neuro transmitters %2C+Drugs+and+Brain+Function-p-9780471978190.

Weiss. B. (1988). Neurobehavioral toxicity as a basis for risk assessment [Review]. Trends Pharmacol Sci 9: 59-62.
http://dx.doi.org/10.1016/0165-6147(88)90118-6.

Weisskopf. MG: Proctor. SP: Wright. RO: Schwartz. J: Spiro. A. Ill: Sparrow. D: Nie. HL: Hu. H. (2007).

Cumulative lead exposure and cognitive performance among elderly men. Epidemiology 18: 59-66.
http://dx.doi.org/10.1097/01.ede.000Q248237.35363.29.

External Review Draft

3-526

DRAFT: Do not cite or quote

-------
Weisskopf. MG: Weuve. J: Nie. H: Saint-Hilaire. MH: Sudarskv. L: Simon. DK: Hersh. B: Schwartz. J: Wright.
RQ: Hu. H. (2010). Association of cumulative lead exposure with Parkinson's Disease. Environ Health
Perspect 118: 1609-1613. http://dx.doi.org/10.1289/ehp.1002339.

Weisskopf. MG: Wright. RO: Schwartz. J: Spiro. A. Ill: Sparrow. D: Aro. A: Hu. H. (2004). Cumulative lead

exposure and prospective change in cognition among elderly men: The VA Normative Aging Study. Am J
Epidemiol 160: 1184-1193. http://dx.doi.org/10.1093/aie/kwh333.

Weston. HI: Weston. DP: Allen. JL: Corv-Slechta. DA. (2014). Sex-dependent impacts of low-level lead exposure
and prenatal stress on impulsive choice behavior and associated biochemical and neurochemical
manifestations. Neurotoxicology 44: 169-183. http://dx.doi.Org/10.1016/i.neuro.2014.06.013.

Weuve. J: Kelsev. KT: Schwartz. J: Bellinger. D: Wright. RO: Raian. P: Spiro. A. Ill: Sparrow. D: Aro. A: Hu. H.
(2006). Delta-aminolevulinic acid dehydratase polymorphism and the relation between low level lead
exposure and the Mini-Mental Status Examination in older men: The Normative Aging Study. Occup
Environ Med 63: 746-753. http://dx.doi.org/10.1136/oem.2006.027417.

Weuve. J: Korrick. SA: Weisskopf. MA: Ryan. LM: Schwartz. J: Nie. HL: Grodstein. F: Hu. H. (2009). Cumulative
exposure to lead in relation to cognitive function in older women. Environ Health Perspect 117: 574-580.
http://dx.doi.org/10.1289/ehp. 11846.

Weuve. J: Press. DZ: Grodstein. F: Wright. RO: Hu. H: Weisskopf. MG. (2013). Cumulative exposure to lead and
cognition in persons with Parkinson's disease. Mov Disord 28: 176-182.
http://dx.doi.org/10.1002/mds.25247.

Whitcomb. S: Merrell. KW. (2012). Behavioral, social, and emotional assessment of children and adolescents (4th
ed.). New York, NY: Routledge. http://dx.doi.org/10.4324/9780203818596.

WHO (World Health Organization). (1948). Preamble to the Constitution of the World Health Organization as

adopted by the International Health Conference. New York, 19-22 June, 1946; signed on 22 July 1946 by
the representatives of 61 States (Official Records of the World Health Organization, no. 2, p. 100) and
entered into force on 7 April 1948. In Constitution of the World Health Organization (pp. 2). Geneva,
Switzerland, http://whalibdoc.who.int/hist/official records/constitution.pdf.

Winter. AS: Sampson. RJ. (2017). From lead exposure in early childhood to adolescent health: A Chicago birth
cohort. Am J Public Health 107: 1496-1501. http://dx.doi.org/10.2105/AJPH.2017.303903.

Wright. JP: Dietrich. KN: Ris. MP: Hornung. RW: Wessel. SD: Lanphear. BP: Ho. M: Rae. MN. (2008).

Association of prenatal and childhood blood lead concentrations with criminal arrests in early adulthood.
PLoS Med 5: elOl. http://dx.doi.org/10.1371/iournal.pmed.005Q101.

Wright. JP: Lanphear. BP: Dietrich. KN: Bolger. M: Tullv. L: Cecil. KM: Sacarellos. C. (2021). Developmental lead
exposure and adult criminal behavior: A 30-year prospective birth cohort study. Neurotoxicol Teratol 85:
106960. http://dx.doi.Org/10.1016/i.ntt.2021.106960.

Wright. RO: Tsaih. SW: Schwartz. J: Spiro. A. Ill: McDonald. K: Weiss. ST: Hu. H. (2003). Lead exposure
biomarkers and mini-mental status exam scores in older men. Epidemiology 14: 713-718.
http://dx.doi.org/10.1097/01.EDE.0000Q81988.85964.db.

Wu. J: Basha. MR: Brock. B: Cox. DP: Cardozo-Pelaez. F: Mcpherson. CA: Harry. J: Rice. DC: Malonev. B: Chen.
D: Lahiri. DK: Zawia. NH. (2008). Alzheimer's disease (AD)-like pathology in aged monkeys after
infantile exposure to environmental metal lead (pb): Evidence for a developmental origin and
environmental link for AD. J Neurosci 28: 3-9. http://dx.doi.org/10.1523/ineurosci.4405-07.20Q8.

Wu. S: Liu. H: Zhao. H: Wang. X: Chen. J: Xia. D: Xiao. C: Cheng. J: Zhao. Z: He. Y. (2020a). Environmental lead
exposure aggravates the progression of Alzheimer's disease in mice by targeting on blood brain barrier.
Toxicol Lett 319: 138-147. http://dx.doi.Org/10.1016/i.toxlet.2019.ll.009.

Wu. Z: Bai. L: Tu. R: Zhang. L: Ba. Y: Zhang. H: Li. X: Cheng. X: Li. W: Huang. H. (2020b). Disruption of

synaptic expression pattern and age-related DNA oxidation in a neuronal model of lead-induced toxicity.
Environ Toxicol Pharmacol 76: 103350. http://dx.doi.Org/10.1016/i.etap.2020.103350.

External Review Draft

3-527

DRAFT: Do not cite or quote

-------
Xiao. J: Wang. T: Xu. Y: Gu. X: Li. D: Niu. K: Wang. T: Zhao. J: Zhou. R: Wang. HL. (2020). Long-term probiotic
intervention mitigates memory dysfunction through a novel H3K27me3-based mechanism in lead-exposed
rats. Transl Psychiatry 10: 25. http://dx.doi.org/10.1038/s41398-020-Q719-8.

Xiao. L: Zan. G: Oin. J: Wei. X: Lu. G: Li. X: Zhang. H: Zou. Y: Yang. L: He. M: Zhang. Z: Yang. X (2021).

Combined exposure to multiple metals and cognitive function in older adults. Ecotoxicol Environ Saf 222:
112465. http://dx.doi.Org/10.1016/i.ecoenv.2021.112465.

Xiao. Y: Fu. H: Han. X: Hu. X: Gu. H: Chen. Y: Wei. O: Hu. O. (2014). Role of synaptic structural plasticity in

impairments of spatial learning and memory induced by developmental lead exposure in Wistar rats. PLoS
ONE 9: el 15556. http://dx.doi.org/10.1371/iournal.pone.0115556.

Xiao. Y: Ma. B: McElhenv. D: Parthasarathv. S: Long. F: Hoshi. M: Nussinov. R: Ishii. Y. (2015). A(3(l-42) fibril
structure illuminates self-recognition and replication of amyloid in Alzheimer's disease. Nat Struct Mol
Biol 22: 499-505. http://dx.doi.org/10.1038/nsmb.2991.

Xu. L: Huo. X: Liu. Y: Zhang. YL: Oin. OL: Xu. XJ. (2020). Hearing loss risk and DNA methylation signatures in
preschool children following lead and cadmium exposure from an electronic waste recycling area.
Chemosphere 246: 125829. http://dx.doi.Org/10.1016/i.chemosphere.2020.125829.

Y Ortiz. MT: Tellez-Roio. MM: Treio-Valdivia. B: Schnaas. L: Osorio-Valencia. E: Coull. B: Bellinger. D: Wright.
RJ: Wright. RO. (2017). Maternal stress modifies the effect of exposure to lead during pregnancy and 24-
month old children's neurodevelopment. Environ Int 98: 191-197.
http://dx.doi.org/10.1016/i.envint.2016.11.005.

Yang. W: Tian. ZK: Yang. HX: Feng. ZJ: Sun. JM: Jiang. H: Cheng. C: Ming. OL: Liu. CM. (2019). Fisetin

improves lead-induced neuroinflammation, apoptosis and synaptic dysfunction in mice associated with the
AMPK/SIRT1 and autophagy pathway. Food Chem Toxicol 134: 110824.
http://dx.doi.Org/10.1016/i.fct.2019.110824.

Yang. X: Wang. B: Zeng. H: Cai. C: Hu. O: Cai. S: Xu. L: Meng. X: Zou. F. (2014). Role of the mitochondrial
Ca2+ uniporter in Pb2+-induced oxidative stress in human neuroblastoma cells. Brain Res 1575: 12-21.
http://dx.doi.Org/10.1016/i.brainres.2014.05.032.

Yang. Y: Ma. Y: Ni. L: Zhao. S: Li. L: Zhang. J: Fan. M: Liang. C: Cao. J: Xu. L. (2003). Lead exposure through
gestation-only caused long-term learning/memory deficits in young adult offspring. Exp Neurol 184: 489-
495. http://dx.doi.org/10.1016/S0014-4886(03)00272-3.

Yang. YW: Liou. SH: Hsueh. YM: Lvu. WS: Liu. CS: Liu. HJ: Chung. MC: Hung. PH: Chung. CJ. (2018). Risk of
Alzheimer's disease with metal concentrations in whole blood and urine: A case-control study using
propensity score matching. Toxicol Appl Pharmacol 356: 8-14.
http://dx.doi.Org/10.1016/i.taap.2018.07.015.

Yassa. HA. (2014). Autism: A form of lead and mercury toxicity. Environ Toxicol Pharmacol 38: 1016-1024.
http://dx.doi.Org/10.1016/i.etap.2014.10.005.

Ye. F: Li. X: Li. F: Li. J: Chang. W: Yuan. J: Chen. J. (2016a). Cyclosporin A protects against lead neurotoxicity
through inhibiting mitochondrial permeability transition pore opening in nerve cells. Neurotoxicology 57:
203-213. http://dx.doi.Org/10.1016/i.neuro.2016.10.004.

Ye. F: Li. X: Li. L: Yuan. J: Chen. J. (2016b). t-BHQ provides protection against lead neurotoxicity via Nrf2/HO-l
pathway. Oxid Med Cell Longev 2016: 2075915. http://dx.doi.org/10.1155/2016/2075915.

Ye. F: Li. XY: Liu. YW: Jiang. AL: Li. XT: Yang. LY: Chang. W: Yuan. J: Chen. J. (2020). CypD deficiency

confers neuroprotection against mitochondrial abnormality caused by lead in SH-SY5Y cell. Toxicol Lett
323: 25-34. http://dx.doi.Org/10.1016/i.toxlet.2019.12.025.

Yia JZ: E. M: Chao. H. (2021). Population-based study of environmental lead exposure and hearing loss: A
systematic review and meta-analysis [Review]. Public Health 197: 63-67.
http://dx.doi.Org/10.1016/i.puhe.2021.06.009.

External Review Draft

3-528

DRAFT: Do not cite or quote

-------
Yorifuii. T: Debes. F: Weihe. P: Grandiean. P. (2011). Prenatal exposure to lead and cognitive deficit in 7- and 14-
year-old children in the presence of concomitant exposure to similar molar concentration of
methylmercury. Neurotoxicol Teratol 33: 205-211. http://dx.doi.Org/10.1016/i.ntt.2010.09.004.

You. YY: Sun. LG: Peng. B: Li. Y: Ben. SB: Gao. S. (2012). Increased hippocampal Disrupted-In-Schizophrenia 1
expression in mice exposed prenatally to lead. Neural Regen Res 7: 1939-1945.
http://dx.doi.Org/10.3969/i.issn.1673-5374.2012.25.003.

Yousef. AO: Fahad. AA: Abdel Moneim. AE: Metwallv. DM: El-Khadragy. MF: Kassab. RB. (2019). The
neuroprotective role of coenzyme Q10 against lead acetate-induced neurotoxicity is mediated by
antioxidant, anti-inflammatory and anti-apoptotic activities. Int J Environ Res Public Health 16: 2895.
http://dx.doi.org/10.3390/iierphl6162895.

Yu. YL: Thiis. L: Saenen. N: Melgareio. JD: Wei. D: Yang. W: Yu. C: Roels. HA: Nawrot. T: Maestre. GE:

Staessen. J: Zhang. ZY. (2021). Two-year neurocognitive responses to first occupational lead exposure.
Scand J Work Environ Health 47: 233-243. http://dx.doi.org/10.5271/siweh.3940.

Yuan. W: Holland. SK: Cecil. KM: Dietrich. KN: Wessel. SD: Altave. M: Hornung. RW: Ris. MP: Egelhoff. JC:
Lanphear. BP. (2006). The impact of early childhood lead exposure on brain organization: A functional
magnetic resonance imaging study of language function. Pediatrics 118: 971-977.
http://dx.doi.org/10.1542/peds.2006-0467.

Yun. S: Wu. Y: Niu. R: Feng. C: Wang. J. (2019). Effects of lead exposure on brain glucose metabolism and insulin
signaling pathway in the hippocampus of rats. Toxicol Lett 310: 23-30.
http://dx.doi.Org/10.1016/i.toxlet.2019.04.011.

Yun. SW: Hover. S. (2000). Effects of low-level lead on glycolytic enzymes and pyruvate dehydrogenase of rat
brain in vitro: Relevance to sporadic Alzheimer's disease? Journal of Neural Transmission 107: 355-368.
http://dx.doi.org/10.1007/s00702005003Q.

Zhang. J: Cai. T: Zhao. F: Yao. T: Chen. Y: Liu. X: Luo. W: Chen. J. (2012). The role of a-synuclein and tau

hyperphosphorylation-mediated autophagy and apoptosis in lead-induced learning and memory injury. Int J
Biol Sci 8: 935-944. http://dx.doi.org/10.7150/iibs.4499.

Zhang. J: Yan. C: Wang. S: Hou. Y: Xue. G: Zhang. L. (2014). Chrysophanol attenuates lead exposure-induced
injury to hippocampal neurons in neonatal mice. Neural Regen Res 9: 924-930.
http://dx.doi.org/10.4103/1673-5374.133141.

Zhang. NH: Baker. HW: Tufts. M: Raymond. RE: Salihu. H: Elliott. MR. (2013). Early childhood lead exposure and
academic achievement: Evidence from Detroit public schools, 2008-2010. Am J Public Health 103: E72-
E77. http://dx.doi.org/10.2105/AJPH.2012.301164.

Zhang. RB: Huo. X: Ho. GY: Chen. XJ: Wang. HW: Wang. TY: Ma. L. (2015a). Attention-deficit/hyperactivity

symptoms in preschool children from an e-waste recycling town: Assessment by the parent report derived
from DSM-IV. BMC Pediatr 15: 51. http://dx.doi.org/10.1186/sl2887-015-0368-x.

Zhang. XL: Guariglia. SR: McGlothan. JL: Stansfield. KH: Stanton. PK: Guilarte. TR. (2015b). Presynaptic
mechanisms of lead neurotoxicity: Effects on vesicular release, vesicle clustering and mitochondria
number. PLoS ONE 10: e0127461. http://dx.doi.org/10.1371/iournal.pone.0127461.

Zhao. ZH: Du. KJ: Wang. T: Wang. JY: Cao. ZP: Chen. XM: Song. H: Zheng. G: Shen. XF. (2021). Maternal lead
exposure impairs offspring learning and memory via decreased GLUT4 membrane translocation. Front Cell
Dev Biol 9: 648261. http://dx.doi.org/10.3389/fcell.2021.648261.

Zhao. ZH: Zheng. G: Wang. T: Du. KJ: Han. X: Luo. WJ: Shen. XF: Chen. JY. (2018). Low-level gestational lead
exposure alters dendritic spine plasticity in the hippocampus and reduces learning and memory in rats. Sci
Rep 8: 3533. http://dx.doi.org/10.1038/s41598-018-21521-8.

Zheng. W: Shen. H: Blaner. WS: Zhao. O: Ren. X: Graziano. JH. (1996). Chronic lead exposure alters transthyretin
concentration in rat cerebrospinal fluid: The role of the choroid plexus. Toxicol Appl Pharmacol 139: 445-
450. http://dx.doi.org/10.1006/taap.1996.0186.

External Review Draft

3-529

DRAFT: Do not cite or quote

-------
Zhou. CC: Gao. ZY: Wang. J: Wu. MO: Hu. S: Chen. F: Liu. JX: Pan. H: Yan. CH. (2018). Lead exposure induces
Alzheimers's disease (AD)-like pathology and disturbes cholesterol metabolism in the young rat brain.
Toxicol Lett 296: 173-183. http://dx.doi.Org/10.1016/i.toxlet.2018.06.1065.

Zhou. F: Du. G: Xie. J: Gu. J: Jia. O: Faa Y: Yu. H: Zha. Z: Wang. K: Ouvang. L: Shao. L: Feng. C: Fan. G.

(2020a). RyRs mediate lead-induced neurodegenerative disorders through calcium signaling pathways. Sci
Total Environ 701: 134901. http://dx.doi.Org/10.1016/i.scitotenv.2019.134901.

Zhou. L: Xu. J: Zhang. J: Yan. C: Lin. Y: Jia. Y: Hu. W. (2017). Prenatal maternal stress in relation to the effects of
prenatal lead exposure on toddler cognitive development. Neurotoxicology 59: 71-78.
http://dx.doi.Org/10.1016/i.neuro.2017.01.008.

Zhou. T: Guo. J: Zhang. J: Xiao. H: Oi. X: Wu. C: Chang. X: Zhang. Y: Liu. O: Zhou. Z. (2020b). Sex-specific

differences in cognitive abilities associated with childhood cadmium and manganese exposures in school-
age children: A prospective cohort study. Biol Trace Elem Res 193: 89-99.
http://dx.doi.org/10.1007/sl2011-019-017Q3-9.

Zhu. G: Dai. B: Chen. Z: He. L: Guo. J: Dan. Y: Liang. S: Li. G. (2019a). Effects of chronic lead exposure on the
sympathoexcitatory response associated with the P2X7 receptor in rat superior cervical ganglia. Auton
Neurosci 219: 33-41. http://dx.doi.Org/10.1016/i.autneu.2019.03.005.

Zhu. G: Fan. G: Feng. C: Li. Y: Chen. Y: Zhou. F: Du. G: Jiao. H: Liu. Z: Xiao. X: Lia F: Yan. J. (2013). The
effect of lead exposure on brain iron homeostasis and the expression of DMT1/FP1 in the brain in
developing and aged rats. Toxicol Lett 216: 108-123. http://dx.doi.org/10.1016/i.toxlet.2012.11.024.

Zhu. G: Peng. T: Peng. C: Li. H. (2019b). Chronic lead exposure decreases the expression of Huntingtin-associated
protein 1 (HAP1) through Repressor element-1 silencing transcription (REST). Toxicol Lett 306: 1-10.
http://dx.doi.Org/10.1016/i.toxlet.2019.02.003.

Zhu. X: Liu. X: Wei. F: Wang. F: Merzenich. MM: Schreiner. CE: Sun. X: Zhou. X. (2016). Perceptual training
restores impaired cortical temporal processing due to lead exposure. Cereb Cortex 26: 334-345.
http://dx.doi.org/10.1093/cercor/bhu258.

Zou. Y: Feng. W: Wang. W: Chen. Y: Zhou. Z: Li. O: Zhao. T: Mao. G: Wu. X: Yang. L. (2015). Protective effect
of porcine cerebral hydrolysate peptides on learning and memory deficits and oxidative stress in lead-
exposed mice. Biol Trace Elem Res 168: 429-440. http://dx.doi.org/10.1007/sl2011-015-0329-Q.

External Review Draft

3-530

DRAFT: Do not cite or quote

-------