Invited Perspective: Identifying Childhood Lead Exposure Hotspots for Action

Lead poisoning and its long-term impacts are largely preventable, yet millions of children worldwide remain at risk for exposure to persistent lead hazards in the environment.¹ In the United States, federal and state regulatory policies aimed at controlling environmental sources of lead—including in residential paint, automotive gasoline, plumbing solder, food packaging, and some consumer products—have been instrumental in decreasing population exposures over time.² Despite the numerous successes in childhood lead poisoning prevention, lead exposure remains a “wicked problem,” denoting its complex nature and resistance to resolution.³

Recent estimates indicate that over 250,000 U.S. children 1–5 years of age have blood lead levels (BLLs) $\ge 5\mathrm{\mu }\mathrm{g}/\text{dL}$⁴ and 2.5% of U.S. children 1–5 years of age have BLLs $\ge 3.5\mathrm{\mu }\mathrm{g}/\text{dL}$, the recently updated Centers for Disease Control and Prevention (CDC) blood lead reference value in children.⁵ These estimates do not include older or younger children, occupationally exposed adults who can bring lead hazards home, or the $\text{approximately}\hspace{0.17em}\mathrm{500,000}$ women of childbearing age with BLLs $\ge 5\mathrm{\mu }\mathrm{g}/\text{dL}$ who could pass lead to the developing child during pregnancy and breastfeeding.⁶ Meanwhile, the evidence has mounted for adverse health effects at lower exposure levels, and it is now recognized that there is no safe BLL for children.⁷

Childhood lead exposure is associated with well-established risk factors at the population level, including race/ethnicity, low socioeconomic status, and housing age.^4,8–12 However, these risk factors are not uniformly or consistently distributed across the United States, and significant inequities in lead exposure sources and pathways exist.¹³ According to the Federal Lead Action Plan, identifying children and communities at increased risk of lead exposure is an important priority for local, state, and federal health agencies, as well as for health care providers and families with young children.¹⁴

Xue et al.¹⁵ present “a systematic approach to help identify high lead exposure locations” using state-based childhood blood lead surveillance data collected over an 11-year period (2006–2016) by the Michigan Department of Health and Human Services (MDHHS). The authors identified census tracts where children were identified with elevated BLLs (EBLLs) of $\ge 3.5$, $\ge 5$, or $\ge 10\mathrm{\mu }\mathrm{g}/\text{dL}$. “High %EBLL locations” were defined as “%EBLL exceedance rate” (the number of children tested in the census tract with EBLL divided by the total number of children tested in the census tract times 100) and as “population-adjusted %EBLL” (the number of children tested in the census tract with EBLL divided by the total population in the census tract times 100). Two statistical methods for identifying small geographic areas where children are at increased risk for lead exposure (“hotspots”) were used: those locations with %EBLLs in the highest 20th percentile and those locations that were clustered statistically. The authors then compared the hotspot locations identified by these two methods to a) each other, b) annual state BLL reports, and c) three existing lead exposure models/indices. Where blood lead surveillance data were available, they evaluated two statistical approaches to identify hotspots with multiple methods used to cross-validate the results. There was agreement between time periods, and high EBLL locations were relatively consistent between the EBLL threshold values across the most recent (2014–2016) period, with some different census tracts identified depending on the method used.

Of greater interest, perhaps, is the ability to reliably identify hotspots for public health action where blood lead surveillance data are lacking. This study provides some evidence that, in the absence of sufficient blood lead data, the three models/indices were moderately sensitive to identify high %EBLL locations, although some differences in rural and urban areas were observed. The data convergence analysis found 140 census tracts (or 5.8% of 2,401) that were not explained by housing–demographic factors. The U.S. Department of Housing and Urban Development’s Deteriorated Paint Index was developed to help identify and target homes with potential lead-based hazards for remediation and lead paint abatement efforts at the state and local (county, census tract) levels.¹⁶ The CDC/Agency for Toxic Substances and Disease Registry has developed a childhood Lead Exposure Risk Index (LERI) to help parents, health care providers, and the general public identify and map community risk for childhood lead exposure. LERI uses a nationally consistent approach with publicly available data across four domains—sociodemographic, housing, environmental, and geographic risk factors—weighted using nationally representative blood lead data to account for the relative importance of the risk factors.¹⁷

Methods for detecting hotspots are valuable in practice, and have been used for understanding the spatial distribution of, for example, radionuclides,¹⁸ traffic crashes,¹⁹ small protein molecules,²⁰ and COVID-19 deaths.²¹ A risk-based approach to hotspot identification can be used to focus limited resources on localized areas of environmental contamination to guide remediation efforts.²² Wartenberg²³ first suggested that a geographic information system (GIS) approach could be used by lead screening programs to improve targeting of high-risk populations for lead exposure. A literature review on the use of GIS for childhood lead exposure identified at least 23 published studies that suggested numerous risk factors.²⁴

The role of “place” is increasingly recognized as having a critical role in population health and health equity.²⁵ However, the power of GIS for statistical modeling to predict lead exposure risk has yet to be fully realized owing to multiple limitations of model inputs and approaches. Such analyses depend on reliable assessment of both temporal and spatial components of exposure. One limitation is the inability to geocode all addresses for positional accuracy. Miranda et al.²⁶ overcame this limitation by supplementing address information with tax assessor data at the individual parcel level in North Carolina; however, similar data may not be available or consistent beyond the local level. Additional limitations include: 1) a lack of national datasets for environmental lead levels (e.g., soil, water, leaded pipes, air), 2) complex exposure data distributions are often modeled as dichotomous variables, and 3) reliance on other variables with substantial measurement error increases model uncertainties. There may also be limited ability to capture contextual constructs that change over time, such as housing and neighborhood gentrification.

In addition, there are several limitations to reliance solely on lead risk assessment questionnaires, at the individual level, and on historical blood lead surveillance data, at the population level, to define risk. The CDC/American Academy of Pediatrics (AAP) lead risk assessment questionnaire has never been demonstrated to reasonably predict risk of EBLLs, and in a systematic review, it performed little better than chance at predicting lead poisoning risk among children.^27,28 Blood lead screening requirements differ widely across states and local health jurisdictions, with some states (and Federal Medicaid for enrolled children) requiring all young children to receive a blood lead test, but most use various targeted screening criteria to determine which children should be tested.²⁹ Because the MDHHS uses a targeted screening approach, despite efforts to “ground truth” the data with various sensitivity analyses, the children tested may not be representative of the entire state. Local, state, and national childhood blood lead surveillance data are imperfect owing to the reliance on health care providers to identify appropriate children for testing (without sufficient tools to determine a child’s individual risk) and incomplete reporting of test results by clinicians and clinical laboratories to public health departments.³⁰

Childhood lead exposure remains a critical public health challenge and environmental justice issue that disproportionately impacts economically disadvantaged communities of color. The U.S. Environmental Protection Agency has restated its commitment to reduce community lead exposure by addressing multimedia exposure pathways with “all our applicable statutory authorities, across all our relevant programs, and in coordination with our federal partners and stakeholders.”³¹ Eliminating childhood lead exposure is an achievable goal through the continued use of evidence-based practices at the individual and population levels.^32,33 Primary prevention strategies that control or eliminate sources of lead from the environment—before children are exposed—remain the most effective way to prevent the harmful effects of lead exposure. Ongoing “lead-free” city initiatives (e.g., in Flint, Michigan; Cleveland, Ohio; Rochester, New York) have demonstrated proof of concept. Our failure to this point to completely address lead hazards in the environment represents a lost opportunity to protect and promote healthy child development. Although there is still a lot of work to do, concerted use of evidence-based practices, primary prevention strategies, and approaches like those by Xue et al.¹⁵ have every chance of turning the tide. We need to capitalize on the political will and current momentum at the federal, state, and local levels to finally eliminate sources of childhood lead exposure in the environment.

Refers to https://doi.org.1289/10.1289/EHP9705