Lead Risk Assessment and Neurodevelopmental Toxicity

Lead and Compounds (Inorganic) — Risk Assessment Methodology

CASRN 7439-92-1 | RoseyFields Chemical-Specific Methodology Series

RoseyFields conducts lead risk assessments using the following methodology.

Lead is the leading example of a chemical for which the U.S. Environmental Protection Agency (EPA) has deliberately declined to derive the standard toxicity values — an oral Reference Dose (RfD), an inhalation Reference Concentration (RfC), or a quantitative cancer slope factor (CSF) or inhalation unit risk (IUR) — that anchor risk assessment for most chemicals. EPA’s Integrated Risk Information System (IRIS) file for lead states this directly: EPA’s RfD Work Group considered and rejected deriving an RfD in 1985, and reaffirmed that position in its most recent (2004) review. No inhalation RfC has ever been evaluated. No numerical cancer slope factor has been derived for either oral or inhalation exposure, notwithstanding a formal weight-of-evidence classification of lead as a probable human carcinogen. This is not an oversight or a data gap to be filled by RoseyFields; it reflects EPA’s considered judgment that the standard threshold-based, point-of-departure/uncertainty-factor (POD/UF) framework does not appropriately characterize lead’s risk. This methodology page explains why, and describes the alternative, blood-lead-level-based framework that RoseyFields applies instead when assessing lead exposure in litigation and regulatory matters.

1. Chemical Identity and Regulatory Background

Lead (Pb, CASRN 7439-92-1) is a naturally occurring, bluish-gray heavy metal historically valued for its malleability, corrosion resistance, and low melting point. Environmental and regulatory assessments generally address “lead and compounds (inorganic),” encompassing elemental lead and its inorganic salts (e.g., lead acetate, lead subacetate, lead oxide, lead chromate) as encountered in paint, plumbing, soil, dust, and industrial emissions; organolead compounds (e.g., tetraethyl lead) are addressed separately.

EPA’s IRIS file for lead was first posted in 1988, with the carcinogenicity section last substantively revised in 1988 and the noncancer (RfD/RfC) section last revised in 2004. Because IRIS has not derived quantitative toxicity values, the operative regulatory framework for lead is instead built around media-specific standards and blood-lead-level (BLL) benchmarks, developed and periodically updated across several agencies:

• Safe Drinking Water Act — EPA’s Lead and Copper Rule action level for lead in drinking water.
• Toxic Substances Control Act (TSCA), as amended by the Residential Lead-Based Paint Hazard Reduction Act (Title X) — standards for lead in paint, interior dust, and bare residential soil.
• Clean Air Act — National Ambient Air Quality Standard (NAAQS) for lead.
• CERCLA/Superfund — site-specific screening levels for lead in soil, developed using the modeling approach described in Section 4.
• CDC — the blood lead reference value (BLRV), a population-based benchmark (97.5th percentile of U.S. children’s BLLs) used to identify children with exposure higher than most of their peers. CDC set the BLRV at 5 µg/dL in 2012 and lowered it to 3.5 µg/dL in 2021 following the recommendation of the Lead Exposure Prevention Advisory Committee, reflecting the continued decline in background U.S. blood lead levels and the absence of an identified safe threshold.

RoseyFields tracks these values as they are updated, since (unlike an IRIS RfD) they are revised periodically as population blood lead distributions shift and as new epidemiological evidence accumulates on effects at progressively lower exposure levels.

RoseyFields — Chemical-Specific Risk Assessment Methodology: Lead

Exposure Sources and Pathways

Lead exposure is multi-source and highly setting-dependent, which is itself one of the reasons EPA has resisted a single generic toxicity value. Principal sources and pathways relevant to RoseyFields matters include:

• Deteriorating lead-based paint in pre-1978 housing, and the interior dust and exterior soil it generates — historically the dominant pathway for U.S. childhood exposure.
• Lead-contaminated residential and urban soil, from historical paint, leaded gasoline deposition, and industrial or mining activity.
• Drinking water contacting lead service lines, lead solder, or leaded brass fixtures, particularly where water is corrosive to plumbing.
• Occupational exposure in battery manufacturing and recycling, secondary lead smelting, radiator repair, firing-range work, and construction/renovation of older structures.
• Consumer products and imported goods, including some ceramics, cosmetics, traditional remedies, and toys.
• Ambient air near point sources such as smelters and, historically, from leaded gasoline combustion. Children under six are the population of greatest concern because of hand-to-mouth behavior that increases incidental ingestion of lead-contaminated dust and soil, and because of greater gastrointestinal absorption and neurodevelopmental susceptibility at this life stage. Pregnant women and their fetuses are a second population of concern because maternal skeletal lead, mobilized during pregnancy and lactation, can cross the placenta regardless of current external exposure.

2. Why the Standard POD/UF/RfD Framework Does Not Apply

For most chemicals, RoseyFields identifies a principal study and a point of departure (POD) — typically a No Observed-Adverse-Effect Level (NOAEL) or benchmark dose — then applies uncertainty factors to derive an RfD. EPA’s IRIS assessment explains, in its own words, why this approach was rejected for lead:

• No threshold: available evidence indicates some lead-associated effects — particularly changes in specific blood enzyme levels and children’s neurobehavioral development — occur at blood lead levels low enough to be essentially without an identified threshold, undermining the core assumption (a no-effect exposure level exists) on which the RfD methodology depends.
• Pre-existing body burden: lead bioaccumulates for decades in bone, and an individual’s current blood lead level reflects both ongoing intake and mobilization of this skeletal reservoir. A dose-response relationship derived from external dose alone (as a standard RfD requires) cannot account for this internal reservoir, which varies with age, health status, nutritional status, and, for women, pregnancy and lactation history.
• A single population-wide toxicity value would misrepresent risk for any individual whose body burden differs materially from the study population’s, which is the norm rather than the exception for lead.

Because no POD was selected and no RfD was derived, there is likewise no oral slope factor for translating a POD into a dose-response toxicity value, and Sections I.A.2 (Principal and Supporting Studies), I.A.3 (Uncertainty and Modifying Factors), and I.A.5 (Confidence in Risk Values) of the IRIS file are each marked “Not applicable.” RoseyFields therefore does not apply a POD, UF stack, or numerical RfD for lead in noncancer assessments, and flags as unsupportable any opposing analysis that purports to derive one from the IRIS file.

3. Pharmacokinetic (Biokinetic) Modeling in Place of Standard Dose Conversion

RoseyFields — Chemical-Specific Risk Assessment Methodology: Lead

In place of animal-to-human dose conversion and a static RfD, EPA and RoseyFields rely on physiologically based biokinetic models that predict blood lead level directly from estimated exposure, integrating absorption, distribution to blood and bone, and elimination over time. Two models anchor current practice:

Integrated Exposure Uptake Biokinetic (IEUBK) Model for Lead in Children

The IEUBK model, developed by EPA’s Superfund program, estimates a probability distribution of blood lead levels in children (typically ages 0–7) resulting from a specified combination of lead concentrations in soil, dust, water, air, diet, and paint. It was used in direct support of the TSCA lead-based paint/dust/soil hazard standards and remains the standard tool for site-specific childhood risk assessment at contaminated properties and Superfund sites. RoseyFields uses IEUBK outputs — predicted geometric mean BLL and the probability that a child’s BLL exceeds the current CDC BLRV of 3.5 µg/dL — as the operative, individualized substitute for a generic HQ calculation.

Adult Lead Model (ALM)

EPA’s Adult Lead Model addresses adult and, in particular, fetal risk from maternal lead exposure, predicting maternal blood lead level from estimated adult intake (chiefly occupational or site-related soil/dust exposure) and, from that, fetal blood lead level. It is used where the population of concern is pregnant workers or residents rather than young children.

Both models require site- or case-specific inputs (media concentrations, exposure frequency and duration, dietary and behavioral assumptions) rather than a single default exposure scenario, and RoseyFields calibrates these inputs to the facts of each matter — sampling data, exposure duration, and, where available, biomonitoring data — rather than relying on a generic default.

4. Uncertainty Factors: None Derived

EPA’s IRIS file lists Section I.A.3 (Uncertainty and Modifying Factors) as “Not applicable” because no RfD exists to which such factors could be applied. RoseyFields does not construct or apply an interspecies (UFA), intraspecies (UFH), subchronic-to-chronic (UFS), LOAEL-to-NOAEL (UFL), or database-uncertainty (UFD) factor for lead, and treats any purported UF-based lead RfD in an opposing report as inconsistent with the governing IRIS assessment.

The functional analogue to interspecies and intraspecies uncertainty factors for lead is instead built into the biokinetic models themselves and into the population-based derivation of the BLRV: the IEUBK and Adult Lead Models generate a distribution of predicted blood lead levels (not a single point estimate), explicitly incorporating inter-individual variability in intake, absorption, and biokinetics, while the BLRV is itself set at the 97.5th percentile of a current, nationally representative pediatric blood lead distribution — a population-variability benchmark rather than a study-derived safety margin.

5. Reference Dose / Reference Concentration: Not Derived

EPA has not established, and RoseyFields does not apply, a numerical oral RfD or inhalation RfC for lead. This has been EPA’s consistent position since 1985 and was reaffirmed in the 2004 IRIS update, which states that current knowledge of lead pharmacokinetics indicates that risk values derived by standard procedures would not accurately reflect potential risk because of the difficulty in accounting for pre-existing body burden. No inhalation RfC has been evaluated at all.

In place of an RfD/RfC, RoseyFields’ operative noncancer benchmarks for lead are:

• CDC’s blood lead reference value (3.5 µg/dL, effective 2021) for children ages 1–5, used as the outcome benchmark for IEUBK-modeled exposures.

RoseyFields — Chemical-Specific Risk Assessment Methodology: Lead

• Media-specific regulatory standards — the drinking water action level, the TSCA lead-in-paint/dust/soil standards, and the lead NAAQS — used as exposure-side screening levels.
• Site-specific IEUBK or Adult Lead Model output, compared against the BLRV or an agreed litigation/regulatory benchmark, used for individualized or property-specific risk characterization.

6. Carcinogenicity Classification

EPA classified lead and lead compounds as Group B2, a probable human carcinogen, under the 1986 EPA cancer guidelines, based on sufficient evidence of carcinogenicity in animals and inadequate evidence in humans.

Animal Evidence

More than ten investigators have demonstrated the carcinogenic potential of soluble lead salts (chiefly acetates and phosphates) administered orally or by injection to rats and mice, with bilateral renal carcinoma as the most characteristic response.

• Azar et al. (1973) fed lead acetate at 10–2,000 ppm to rats for two years; renal tumor incidence rose from 5/50 at 500 ppm to 16/20 at 2,000 ppm in males, with 7/20 females affected at 2,000 ppm, while no tumors occurred at 10–100 ppm or in controls.
• Kasprzak, Hoover, and Poirier (1985) fed 1% lead subacetate (8,500 ppm Pb) to male Sprague-Dawley rats for 79 weeks; 44.8% of surviving rats developed renal tumors (adenomas and adenocarcinomas), with none in controls.
• Koller, Kerkvliet, and Exon (1986) administered lead acetate (2,600 ppm) in drinking water to male Sprague-Dawley rats for 76 weeks; 13 of 16 treated rats (81%) developed renal tubular carcinoma, versus none of 10 controls.
• Van Esch and Kroes (1969) fed basic lead acetate to Swiss mice for two years; 6 of 25 males at 0.1% developed renal tumors, versus none in controls (a lower incidence at 1.0% was attributed to early mortality). Hamsters given lead subacetate showed no significant renal tumor response in the same study.

Supporting mechanistic evidence includes cell transformation of Syrian hamster embryo cells by lead acetate (DiPaolo, Nelson, and Casto, 1978), enhancement of viral transformation and adenovirus induction by lead acetate and lead oxide (Casto, Meyers, and DiPaolo, 1979), and sister chromatid exchange associated with occupational lead exposure (Grandjean, Wulf, and Niebuhr, 1983).

Human Evidence

EPA characterized the human epidemiological evidence as inadequate. Dingwall-Fordyce and Lane (1963) and Nelson, Kiremidjian-Schumacher, and Stotzky (1982) found no association between occupational lead exposure and cancer mortality. Selevan et al. (1985), studying primary lead smelter workers, found a slight overall decrease in cancer mortality (SMR = 95) alongside nonsignificant excesses in respiratory (SMR = 111) and kidney cancer (SMR = 204). Cooper and Gaffey (1975), updated by Cooper (1985), found statistically significant excesses in total cancer mortality, stomach cancer, and lung cancer among battery-plant workers, with similar but nonsignificant excesses among smelter workers. Across these studies, exposures to lead were not quantified, co-exposures to other metals (arsenic, cadmium, zinc) were not controlled, smoking was not addressed, no consistent tumor site emerged, and no dose-response relationship was demonstrated — collectively precluding a conclusion of human carcinogenicity from this body of evidence.

No Quantitative Cancer Slope Factor or Unit Risk

EPA’s Carcinogen Assessment Group declined to derive a numerical oral slope factor or inhalation unit risk for lead, for the same pharmacokinetic reasons underlying the RfD decision: age, health, nutritional state, body burden, and exposure duration all influence lead absorption, retention, and excretion in ways that standard low-dose

RoseyFields — Chemical-Specific Risk Assessment Methodology: Lead

extrapolation cannot capture. RoseyFields therefore does not calculate an excess lifetime cancer risk for lead using a slope factor, and treats lead’s carcinogenicity qualitatively — as a B2 weight-of-evidence classification supported by consistent, high-incidence animal renal tumor findings — rather than quantitatively.

7. Application in RoseyFields Risk Assessments

Because lead has no RfD, RfC, or CSF, RoseyFields’ practical workflow departs from its standard Hazard Quotient/Hazard Index and slope-factor-based approach:

• Exposure and dose reconstruction: RoseyFields compiles site- or case-specific data on lead concentrations in the relevant media (soil, interior dust, drinking water, air, occupational surfaces) together with exposure duration, frequency, and, where available, biomonitoring (blood lead) data for the individuals or population at issue.
• Biokinetic modeling in place of HQ/HI: rather than dividing an estimated dose by an RfD, RoseyFields inputs the reconstructed exposure into the IEUBK model (children) or Adult Lead Model (adults, including pregnant workers/residents) to generate a predicted blood lead level or distribution, and compares that prediction against the current CDC BLRV (3.5 µg/dL) or another benchmark appropriate to the matter (e.g., an applicable regulatory or consent-decree threshold).
• Comparison to media-specific standards: where individualized biokinetic modeling is not warranted or feasible, RoseyFields compares measured or estimated media concentrations directly to the governing regulatory standard (drinking water action level, TSCA paint/dust/soil standards, or NAAQS) as a screening-level indicator of exposure significance.
• Cancer risk characterization: RoseyFields presents lead’s carcinogenicity qualitatively, citing the B2 classification and the specific supporting animal studies, rather than computing an excess lifetime cancer risk figure, and explains to counsel and testifying experts why no such calculation is defensible on the current IRIS record.
• Litigation and regulatory support: this framework is applied in toxic tort and property-damage matters (e.g., lead paint, contaminated soil, or drinking water claims), regulatory compliance and enforcement matters under TSCA, the Safe Drinking Water Act, and CERCLA, and site remediation decision-making, where

RoseyFields’ role is typically to reconstruct exposure, generate a defensible predicted blood-lead-level or exceedance analysis, and explain the qualitative basis for lead’s carcinogenicity classification to the trier of factor regulator.

8. Key Uncertainties and Sensitive Populations

• No identified threshold: because effects on children’s neurodevelopment and specific blood enzymes have not been shown to have a threshold, any exposure reduction is presumptively risk-reducing, and there is no “safe” exposure level analogous to an RfD against which to declare a case-specific exposure acceptable.
• Body burden and multi-source exposure: current blood lead level reflects both recent intake and mobilization of decades-old skeletal lead, complicating attribution of a given individual’s BLL to any single source or defendant, and requiring careful, source-specific exposure and pathway analysis in multi defendant matters.
• Children under six: heightened vulnerability from hand-to-mouth behavior, higher gastrointestinal absorption efficiency than adults, and a developing central nervous system particularly sensitive to lead’s neurotoxic effects.
• Pregnant women and fetuses: maternal skeletal lead mobilized during pregnancy and lactation can elevate fetal and infant exposure independent of the mother’s current external exposure, which the Adult Lead Model is specifically designed to address.
• Evolving benchmarks: unlike a static IRIS RfD, the CDC blood lead reference value has been revised twice (5 µg/dL in 2012; 3.5 µg/dL in 2021) as population blood lead levels have declined and as evidence of effects at progressively lower levels has accumulated; RoseyFields confirms the benchmark in effect at the

RoseyFields — Chemical-Specific Risk Assessment Methodology: Lead relevant exposure period and, separately, at the time of assessment, since these may differ and both may be relevant depending on the claim.

• Co-exposures and confounding in epidemiological evidence: the human occupational studies underlying the B2 classification did not control for smoking or co-exposure to other metals, a limitation RoseyFields flags when the qualitative human evidence is at issue.
• Screening-level literature reviews: EPA’s most recent (2002) screening-level reviews of both the noncancer and cancer literature for lead identified, respectively, new studies of potential relevance and no critical new cancer studies; RoseyFields monitors subsequent peer-reviewed literature and agency guidance (CDC, ATSDR, and any future IRIS reassessment) for developments that could affect this methodology.

Bibliography