Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Mar 1.
Published in final edited form as: Health Place. 2021 Feb 2;68:102517. doi: 10.1016/j.healthplace.2021.102517

Proximity to Sources of Airborne Lead is Associated with Reductions in Children’s Executive Function in the First Four Years of Life

Lisa M Gatzke-Kopp 1, Siri Warkentien 2, Michael Willoughby 3, Chris Fowler 4, David Folch 5, Clancy Blair 6
PMCID: PMC7965333  NIHMSID: NIHMS1665680  PMID: 33540187

Abstract

Although policies to remove lead from gasoline have resulted in a substantial reduction in airborne lead, multiple industries are known to generate lead that is released in the air. The present study examines the extent to which residential proximity to a documented source of airborne lead is associated with intellectual and executive function in children. Data were available for n = 849 children from the Family Life Project. Geolocation for children’s residences between birth and 36 months were referenced against the Environmental Protection Agency’s Risk Screening Environmental Indicators (RSEI) database, which estimates exposure for each ½ mile grid in the contiguous United States. Instrumental variable models were employed to estimate causal associations between exposure and cognitive outcomes measured at 36, 48, and 60 months, using census-documented density of manufacturing employment as the instrument. Models of continuous lead dosage indicated small negative effects for both child IQ and executive function (EF). These results indicate that RSEI estimates of airborne lead exposure are meaningfully associated with decrements in cognitive development.

Keywords: Toxic Release Inventory, air pollution, cognitive, development

Extensive research has documented the toxic effects of lead exposure on children’s intellectual development (Wu et al., 2018). In addition to the effects on intelligence, lead exposure has been associated with measures of self-regulation and attention, as well as externalizing behavior problems such as hyperactivity and aggression (Barg et al., 2018; Chiodo et al., 2007; Joo et al., 2018; Thomson et al., 1989). These deficits likely underlie the observed increase in risk that lead exposure poses for school suspension and juvenile detention (Aizer & Currie, 2017). Although lead exposure is frequently confounded with exposure to multiple neurotoxicants, research supports the unique contribution of lead (Boucher et al., 2014; Dorea, 2019; Frndak et al., 2019), reinforcing the importance of identifying all potential sources of exposure to better inform prevention and identification practices

Exposure to Airborne Lead

Policies intended to reduce childhood lead exposure, such as the eradication of leaded gasoline, have been demonstrated to be effective (Page, Cawse, & Baker, 1988), and are estimated to result in better cognitive and behavioral outcomes for children that translate into meaningful academic and occupational success long term (Nilsson, 2009). Unfortunately, despite actions taken to remove lead from commonly used goods, multiple sources of lead exposure persist in the environment. The most well-established are the continuing impact of sources of lead that predate protective policies. Specifically, housing built prior to 1978 typically contains lead-based paint that can be exposed when paint flakes from the wall, or during remodeling projects that generate dust, and older plumbing infrastructure that utilizes lead piping or lead soldering can contaminate public drinking water, particularly if the lead piping is permitted to corrode (Hanna-Attisha, LaChance, Sadler, & Schnepp, 2016). Policies pertaining to paint manufacturing and plumbing materials are designed to address ubiquitous household exposure, however, lead continues to be used in mining and manufacturing industries, which can serve as direct sources of exposure to workers (Rinsky et al., 2018), as well as a source of regional pollution of water and soil (Hai et al., 2018).

Less frequently addressed is the extent to which lead pollution is present in the air, reflecting an inescapable source of exposure for those in its proximity (Taylor, Isley, & Glover, 2019). Airborne sources likely represent a smaller contribution to blood lead levels relative to ingested sources. Indeed, the bioaccessibility of lead has been demonstrated to be much greater for ingestion than inhalation (Dartey et al., 2014), the latter of which is further influenced by particulate size (Chamberlain, 1985). However, multiple studies confirm a direct association between variance in lead levels and variance in airborne exposure induced through policy change (Hayes et al., 1994), naturally occurring seasonal variation (Havlena, Kanarek, & Coons, 2009), and variation in concentration (Hwang, Lin, Kin, & Wang, 2014), indicating that airborne lead directly impacts children’s physiological exposure.

High levels of airborne lead are a known byproduct of over 50 industries such as battery manufacturing, recycling, and smelting (Froines, Baron, Wegman, & O’Rourke, 1990). Pollution from these industries is, as expected, higher in countries without regulations (Gottesfeld & Pokhrel, 2011). Although multiple studies have examined the extent and nature of airborne lead exposure that these industries generate, much of this work has focused on workers who face high occupational exposure risk. Less research has examined the effect of the pollution generated on individuals residing in close proximity to these sources. It is particularly important to examine the potential risks that this pollution represents to children. Research indicates that children demonstrate a stronger association between airborne lead and blood lead than is observed in adults, suggesting that they may be more vulnerable to absorption through this route (Meng et al., 2013). Furthermore, lead is known to be taken up into bone structure (Griffin, Coulston, Wills, & Russell, 1975), which can ultimately be re-released to circulation later in life, particularly during pregnancy (Manton et al., 2003). Given the evidence that children show an association between regional concentrations of airborne lead and measured levels of blood lead, more research is needed to determine whether airborne lead exposure represents a risk to children’s cognitive development.

Variation in Exposure to Airborne Lead

The association between airborne lead pollution and industrial processes means that the amount of exposure varies across geographical locations as a function of proximity to the industrial sources, as well as across time, as industrial plants come and go from a given area (Browne, Husni, & Risk, 1999; Han, Guo, Zhang, Liao, & Nie, 2018; Moody & Grady, 2017). Researchers have capitalized on this variation to examine the impact of exposure for other types of air pollutants on children’s cognitive function. For instance, one group of researchers used residential addresses of infants born in an urban setting to estimate the magnitude of exposure to traffic-generated air pollution (Porta et al., 2016). A modest but significant effect was found, such that researchers observed a 1.4 point reduction in verbal IQ and a 1.2 point reduction in freedom from distractibility for each 10μg/m3 increase in nitrogen dioxide exposure, whereas measures of general particulate matter did not show a significant association with these cognitive measures. The magnitude of effects was strongest for early exposure (pregnancy through the preschool period) relative to the exposure estimated in the year before the IQ assessment at age 7, consistent with the vulnerability of the brain during early development (Porta et al., 2016).

Quantifying the impact of lead exposure is challenging in humans given the many correlated risk factors and the inability to employ scientific controls to determine causality. Exposure to lead is not random, but frequently correlated with a range of additional risk factors for adverse cognitive development including co-occurring toxic exposures, poverty, and parental characteristics. Some researchers have argued that these confounding factors have not been adequately accounted for in the existing findings, and as such, that the impact of low-level lead exposure has been exaggerated (Wilson & Wilson, 2016). In contrast, other researchers have argued that regression-based statistical approaches to controlling for confounding covariates leads to a systematic under-estimation of the true effects of lead exposure (Clay et al., 2019).

The need for novel approaches to estimate the effect of airborne lead exposure on children’s development is critical. Airborne lead is a valid source of lead contributing to elevated blood lead levels in children but continues to be underexamined relative to other sources. Furthermore, the consequences of airborne lead exposure on children’s cognitive development has received only minimal attention in research studies. In this study, we estimate the effect of average airborne lead exposure, as indicated by the RSEI database from geocoded residential addresses at 2, 6, 15, 24, and 36 months, on children’s IQ at 36 months and Executive Function (EF) at 36, 48, and 60 months. Given that observational studies remain the primary way of understanding the effect of airborne toxins, methods that can account for measured and unmeasured confounding are needed. We use one such approach, instrumental variable models, with the instrument defined as the percentage of manufacturing employment at the Census tract level. We demonstrate that children living in tracts with higher density of manufacturing employment have higher levels of average airborne lead exposure during early childhood as indicated by the RSEI data. We hypothesize that children with higher levels of predicted average airborne lead exposure will have worse cognitive outcomes than children with lower levels.

Methods

Participants

Participants in the current study were drawn from the Family Life Project, a longitudinal study of semi-rural children followed prospectively since birth. Families (n = 1,292) were recruited from hospitals between September 1 2003 through August 31st 2004 in 3 counties in North Carolina and 3 counties in Pennsylvania, oversampling for poverty at both sites, and oversampling for African American families in North Carolina. More details about the recruitment strategy can be found in Vernon-Feagans & Cox, (2013). The analytic sample was restricted to children who had valid geolocation data at 2 months, at least one non-missing airborne lead exposure value across the five assessments that occurred during the first three years of children’s lives (the exposure period), and non-missing outcome data. Children who were enrolled in the study but were excluded from these analyses did not differ from the analytic sample with respect to child gender (52% vs. 51% male, p=.80), child race (43% vs. 42% Black, p=.78), or living with the biological dad in the household at birth (62% vs. 65%, p=.44). Children in the analytic sample were slightly less likely than those not in the sample to live in Pennsylvania (39% vs. 46%, p=.02) and to have parents with at least a high school diploma (74% vs. 82%, p=.01).

Procedures

Enrolled families received home visits when the child was 2, 6, 15, 24, 36, 48- and 60-months of age. At each home visit research assistants recorded the geographical coordinates of the participants’ residence. Data on child, family, and household characteristics is drawn from the home visit surveys that were administered between 2 and 36 months, and child outcomes were drawn from cognitive assessments that were administered to children at the 36-, 48-, and 60-month home visits. Participants provided written consent at each visit. All procedures were approved by the University of North Carolina IRB. Sample characteristics for all measures are reported in Table 1.

Table 1.

Sample characteristics

Variables N Mean SD Min Max
Outcomes
 Executive Function (3 years) 796 −0.53 0.54 −1.98 1.18
 Executive Function (4 years) 824 −0.11 0.51 −2.14 1.23
 Executive Function (5 years) 852 0.30 0.49 −1.98 1.30
 WPSSI (3 years) 859 94.20 16.58 45.00 142
Treatment Variable
 Airborne lead exposure (2–36 mos.) 1,036 4.68 5.35 0.01 21
Instrument
 % employed in manufacturing, tract-level 1,036 18.18 6.65 3.91 38.51
Covariates
Demographics
 Child sex (male) 1,036 0.51 0.50 0 1
 Child race (African American) 1,036 0.43 0.49 0 1
 Income-to-needs ratio (6–35 mos.) 1,036 1.94 1.52 0.04 15.50
 PC high school diploma 1,036 0.82 0.39 0 1
 PC 4-year college degree 1,036 0.15 0.35 0 1
 PC age 1,036 26.28 6.20 14.52 69.52
 Biological father in household (2 months) 1,036 0.65 0.48 0 1
 PC IQ estimate 1,036 92.34 13.17 57.00 138.00
 Biological mother/father history ADHD 1,036 0.05 0.22 0 1
 PC hostility 1,036 50.58 8.43 34.00 76.67
 PC depression 1,036 50.32 7.33 36.00 74.00
 Child low birth weight 1,036 0.08 0.27 0 1
 Average cotinine, log (6–24 months) 976 0.39 1.55 −2.59 4.48
Tract-level characteristics
 % living in poverty 1,036 0.17 0.10 0.03 0.56
Open in a new tab

Measures

Lead Exposure

Exposure to lead was extracted from the Environmental Protection Agency’s Risk Screening Environmental Indicators (RSEI) database. RSEI data incorporates information from the Toxics Release Inventory on the amount of toxic chemicals released, together with factors such as the chemical’s fate and transport through the environment, and each chemical’s relative toxicity to provide a toxicity-weighted concentration estimate for each ½ mile square grid across the United States. This dosage score accounts for the change in concentration over geographical distance from its source, as well as the estimated bio-absorption at a specific geographical point (Environmental Protection Agency, 2020). The resultant score is unitless and does not correspond to biologically assayed levels of toxins. RSEI scores are comparable only to one another, thus providing a value that classifies children’s exposure relative to one another to determine which children had the highest exposure. A RSEI toxicity and lead exposure value was assigned to each child for up to 5 time points based on the ½ mile square that corresponded to their recorded geographic location at the 2, 6, 15, 24, and 36-month home visits. Data were available for an average of M = 4.52, SD = 1.02 time points across the sample. Following precedent in previous studies (Chang, Fu, Li, Tam, & Wong, 2018), extreme outlier values were winsorized at the 95th percentile resulting in a winsorized M = 4.68, SD = 5.35, ranging from .01 to 21. Table 2 provides descriptive statistics of airborne lead exposure aggregated to the county level for the six original study counties (three in Pennsylvania and three in North Carolina), and Figure 1 illustrates the distribution of winsorized airborne lead exposure overall and at each time point.

Table 2.

Sample descriptive statistics of average airborne lead exposure (2–36 mos), by county

County N Mean SD Minimum Maximum
G3701630 117 0.65 0.74 0.01 4.12
G3701910 241 3.07 2.38 0.59 18.52
G3701950 263 12.06 4.83 2.11 21.00
G4200130 179 1.58 1.64 0.29 12.14
G4200210 187 1.93 1.38 0.70 14.45
G4200610 32 5.82 5.68 0.28 16.80
Open in a new tab

Estimated airborne lead exposure varies within counties (indicated by min-max) as well as across counties (indicated by mean).

Figure 1.

Figure 1.

Open in a new tab

Distribution of airborne lead at each assessment.

Box plots illustrate the sample distribution of airborne lead exposure at each assessment point. Exposure scores are generated by the RSEI index and are unitless.

Results indicate that although average airborne lead exposure varies across counties, high exposure levels are not restricted to certain counties (M = .65 to 12.06, minimums range from .01 to 2.11 and maximums range from 4.12 to 21). Exposure values across the 5 time points were highly correlated (ranging from .72 to .93, data not shown), with an alpha coefficient across time points of α = .91, indicating relatively stable within-person exposure. Given this stability, an average airborne lead exposure value was calculated for each child as the average of all available RSEI time points. Average lead exposure was defined as a continuous treatment in the current study, although sensitivity analyses were run on dichotomized versions with substantively similar results.

Child Outcomes

WPPSI.

At 36-months children completed the Vocabulary and Block Design subscales of the Wechsler Preschool and Primary Scales of Intelligence (WPPSI; Wechsler, 2002) in order to provide an estimate of intellectual functioning (IQ) (Sattler, 2001). The sample mean was M = 94.2, SD = 16.58, and ranged from 45 to 142.

Executive Function.

Children’s executive function (EF) was measured at 36, 48 and 60 months using a battery of six tasks which encompass measures of inhibitory control, attention shifting, and working memory (Willoughby et al., 2012). All six tasks were scored using item response theory, with z-scores calculated and the mean of all z-scores used as the composite EF score for each child. Previous studies using these composite scores have demonstrated their acceptable psychometric properties (Willoughby et al., 2012).

Instrumental Variable

The instrumental variable (IV) should predict airborne lead exposure and be related to children’s cognitive outcomes only through that exposure. Because most individuals live near their employment (U.S. Department of Transportation, 2003), the density of manufacturing employment is related to the presence of nearby manufacturing plants which are known emitters of toxic chemicals (EPA, 2020). Density of manufacturing employment is drawn from Census data, which provides localized information on the demographic characteristics of residents, workforce and employer characteristics, and characteristics of the built environment. Data from the 2000 Census was used to calculate the percentage of individuals in each Census tract working in manufacturing, with the numerator defined as the total number of individuals working in manufacturing and the denominator defined as the total number of employed individuals. These data were pulled for each Census tract in North Carolina and Pennsylvania and merged with each child’s geolocation at 2 months. Across the Census tracts in the study sample, density of manufacturing employment ranged from 3.9% to 38.5% across 141 tracts.

Because children’s greatest source of lead exposure is likely to be in the home, we examined the association between Census tract-level %manufacturing and tract-level proportion of pre-1950s housing to determine whether this represented a potential confound of the instrumental variable. Surprisingly, manufacturing employment was strongly related to %pre-1950 housing in a negative direction (r = −.45), indicating that tracts with a greater proportion of manufacturing employment were less likely to contain older homes. The association between %pre-1950s housing and RSEI estimates of airborne lead was also negative (r = −.29). Given the direction of this association, % manufacturing was considered to be an appropriate instrumental variable. However, additional covariates were included in the model to account for other potential sources of influence on children’s cognitive development.

Covariates

Descriptive statistics for all variables are reported in Table 1. Covariates related to the child included sex and race. Children with low birth weight (< 2500g) (M = .08, SD = .27) were defined by maternal report at the 2-month home visit. Children’s environmental exposure to tobacco smoke was quantified by assaying cotinine, the primary metabolic byproduct of nicotine, from children’s salivary samples obtained at the 6-, 15-, and 24 -month home visits (see Gatzke-Kopp et al., 2018). Values were log transformed and averaged (M = .39, SD = 1.55).

Additional demographic variables included parental education and family composition as reported by parents at the initial home visit. Children’s exposure to poverty was measured using the income-to-needs ratio (INR) calculated by dividing total household income by the federal poverty threshold for a given family size at each assessment (6-, 15-, 24-, and 36-month) averaged across assessments (M = 1.94, SD = 1.52). Further, although evidence is somewhat mixed, the placement of manufacturing sites has been shown to be associated with neighborhood poverty and racial composition (see Shadbegian & Wolverton, 2010 for review). Neighborhood poverty in particular may be related to child cognitive outcomes through mechanisms other than lead exposure (e.g., sufficient access to nutrition or healthcare). We therefore include tract-level poverty as an additional place-based characteristic of children’s environments as a proxy for multiple community-level risk factors to further minimize this potential source of bias.

Models also condition on additional caregiver characteristics. Primary caregiver age at the time of the child’s birth (M = 26.3, SD = 6.20) was included, as was primary caregiver IQ (M = 92.3, SD = 13.17), which was assessed using an administration of the Vocabulary and Block Design subtests of the Wechsler Adult Intelligence Scale, 3rd edition (Weschler, 1997) at the 48-month home visit with standardized scores. Primary caregiver hostility and depression were assessed using items from the Brief Symptom Inventory (BSI; Derogatis, 1993) including a 6-item subscale assessing depression, and a 5-item subscale assessing hostility, both of which have demonstrated reliability and validity (Derogatis, 2000). Results from the 2-, 6-, 15-, and 24-month home visits were averaged (M = 50.6, SD = 8.43 for hostility; M = 50.3, SD = 7.33 for depression). Parental history of ADHD was measured by a single item that asked whether either the biological mother or father of the focal child had a childhood history of ADHD (M = .05, SD = .22). When the primary caregiver was not a biological parent of the target child, s/he answered the question with reference to the child’s biological parents.

Analysis Plan

The relationship between airborne lead exposure and child cognitive outcomes can be represented as:

Yi=β0+β1Li+β2Xi+μi (1)

where Yi represents the outcome variable (e.g., IQ at age 3, and EF at age 3, 4, or 5) for child i, Li is the average airborne lead exposure for child i; Xi is a vector of observable child, family, and neighborhood characteristics that are related to cognitive outcomes, and μi is the error term. Although this traditional regression model controls for multiple characteristics that are likely related to both the amount of lead a child is exposed to and their cognitive outcomes, there may continue to be unmeasured confounders that bias the effect estimate of lead on cognitive outcomes. For instance, if households choose to avoid areas with high airborne lead levels and these same families have advantages that are associated with higher IQ or executive function in their children, then the β1 estimate in equation (1) will be downwardly biased.

To deal with the endogeneity of lead exposure and unmeasured confounding, we use a two stage least squares (2SLS) instrumental variable (IV) approach to estimate equation (1) (Angrist and Krueger, 2001; Baiocci, Cheng, Small, 2014; Hernan and Robins, 2006). In the first stage of the 2SLS approach, instead of directly controlling for confounders to estimate the relationship between lead and child outcomes, the following equation estimates airborne lead exposure (the key predictor variable in our analysis) using an instrumental variable:

Li=α1Mi+α2Xi+εi (2)

where Mi (the instrument) is the percentage of individuals employed in manufacturing in the child’s Census tract, Xi is the same vector of observable child, family, and neighborhood characteristics as in (1), and εi are unmeasured factors for child i.

In the second stage, we use the predicted values of lead exposure (L^i) from (2) as the independent variable to predict child cognitive outcomes:

Yi=β1L^i+β2Xi+∈i (3)

where β1 provides the average causal effect of airborne lead exposure. The actual exposure has been replaced by predicted exposure, which is arguably unaffected by the common unmeasured causes that confound the relationship between airborne lead and child outcomes. Because the standard errors for the effect estimate must account for the uncertainty in estimation in the first and second stage, equations (2) and (3) are estimated simultaneously using Stata 16. Robust standard errors are estimated that account for the clustering of children within tracts (Baiocci, Cheng, Small, 2014).

The IV approach provides unbiased effect estimates (β1) only when a valid instrument is identified that (i) is correlated with children’s exposure, (ii) does not share common causes with children’s cognitive outcomes, and (iii) has no effect on child outcomes except through the potential effect on exposure. We demonstrate the correlation (i) in the first-stage results in the Results section. Although (ii) and (iii) are not directly testable, to strengthen the likelihood that the instrument is independent of unmeasured confounders, we include a broad set of measured covariates that may be associated with living in a high density manufacturing tract and are associated with child outcomes through mechanisms other than lead exposure. We assume manufacturing density is not related to children’s cognitive outcomes except through airborne lead exposure, a plausible assumption given that the type of work performed by children’s neighbors is unlikely to affect children’s early development due to limited contact outside of family or caregivers.

Results are first presented for the first stage of the IV model to demonstrate the strength of the instrument. Second-stage results provide the effect estimates of airborne lead exposure, with separate analyses conducted for each outcome. Results are shown as standardized coefficients to ease comparisons across outcomes on varying scales (i.e., IQ and executive function) and outcomes with unfamiliar scales (executive function). Standardized coefficients also facilitate the interpretation of changes in RSEI lead values. As noted previously, RSEI values are unitless (i.e., they are not directly comparable to biologically assayed levels of lead in their current form). All variables were standardized to have mean 0 and variance 1.

Results

Table 3 presents correlations for the key outcomes, instrumental variable (density of manufacturing employment), and all covariates for the analytic sample. First-stage results are presented in Table 4. Across outcome models, the first-stage results are identical and vary only slightly based on small differences in the sample size for different outcomes. Results in Table 4 are shown for the largest sample size—IQ measured at 36 months (N = 849) – and indicate that the instrument (density of manufacturing employment) is statistically significant (p < 0.001). For each percentage increase in tract employment in manufacturing, there is an increase of 0.4 in the average airborne lead exposure. The range between the lowest and highest manufacturing employment density tracts is over 34, translating to changes of over 15 in airborne lead, or more than 3 times the SD. In addition to a significant association between density of manufacturing employment and airborne lead exposure, the strength of the instrument is supported by the partial R2 (0.24) which measures the correlation between lead exposure and manufacturing employment density after partialling out the effects of all other covariates (Bound, Jaeger, Baker, 1995). Furthermore, the F-statistic in this model is approximately 39, well above the suggested threshold of 10 for a single instrument, again indicating a sufficiently strong instrument (Stock, Wright, Yogo, 2002).

Table 3.

Sample correlations

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21)
(1) EF (3 years) 1.00
(2) EF (4 years) 0.34 1.00
(3) EF (5 years) 0.31 0.59 1.00
(4) WPSSI (3 years) 0.41 0.54 0.47 1.00
(5) Lead exp. (2–36 mos.) −0.16 −0.18 −0.15 −0.18 1.00
(6) % manufacturing, tract −0.24 −0.26 −0.22 −0.27 0.58 1.00
(7) Child sex (male) −0.09 −0.14 −0.12 −0.12 −0.02 −0.04 1.00
(8) Child race (Black) −0.31 −0.35 −0.31 −0.38 0.38 0.44 −0.02 1.00
(9) INR (6–36 mos.) 0.24 0.28 0.27 0.39 −0.15 −0.23 0.05 −0.36 1.00
(10) PC high school diploma 0.11 0.20 0.15 0.21 −0.09 −0.13 0.00 −0.12 0.28 1.00
(11) PC 4-year college degree 0.21 0.26 0.25 0.36 −0.06 −0.18 0.02 −0.27 0.53 0.20 1.00
(12) PC age 0.13 0.14 0.20 0.21 −0.07 −0.18 0.03 −0.22 0.34 0.32 0.36 1.00
(13) Bio dad in HH (2 mos) 0.22 0.22 0.23 0.27 −0.19 −0.24 0.05 −0.48 0.39 0.31 0.29 0.38 1.00
(14) PC IQ estimate 0.28 0.36 0.35 0.40 −0.17 −0.26 −0.02 −0.42 0.39 0.27 0.39 0.22 0.32 1.00
(15) Bio mom/dad ADHD 0.02 −0.04 −0.02 −0.03 −0.11 −0.05 −0.01 −0.14 −0.03 −0.15 −0.02 −0.06 0.00 0.02 1.00
(16) PC hostility 0.05 0.03 0.00 0.01 0.01 −0.08 −0.03 −0.05 −0.07 −0.03 −0.08 −0.01 −0.04 0.09 0.07 1.00
(17) PC depression −0.05 −0.11 −0.12 −0.16 0.06 0.01 −0.03 0.12 −0.20 −0.08 −0.18 −0.09 −0.17 −0.06 0.07 0.68 1.00
(18) Child low birth weight −0.11 −0.12 −0.14 −0.08 −0.02 0.00 −0.05 0.07 −0.09 −0.06 −0.07 −0.02 −0.07 −0.11 0.01 0.01 0.06 1.00
(19) Avg. Log cotinine, 6–24m −0.20 −0.22 −0.21 −0.29 0.07 0.12 −0.02 0.13 −0.45 −0.30 −0.41 −0.34 −0.26 −0.30 0.10 0.15 0.21 0.07 1.00
(20) % living in poverty, tract −0.15 −0.21 −0.20 −0.25 0.32 0.40 0.03 0.40 −0.33 −0.18 −0.22 −0.20 −0.29 −0.24 −0.01 0.04 0.10 0.04 0.26 1.00
Open in a new tab

Table 4.

First stage results

Airborne Lead Exposure (2–36 mos.)
Coefficient SE
Child sex (male) −0.01 0.34
Child race (Black) 1.28 * 0.65
Income-to-needs ratio (6–36 mos.) 0.05 0.10
PC high school diploma −0.76 0.50
PC 4-year college degree 1.28 ** 0.41
PC age 0.02 0.03
Biological dad in household (2 mos.) −0.03 0.54
PC IQ estimate 0.00 0.01
Biological mother/father history ADHD −1.81 ** 0.56
PC hostility 0.02 0.02
PC depression 0.03 0.03
Child low birth weight −0.68 0.59
Average cotinine, logged (6–24 mos.) 0.02 0.12
% poverty, tract-level 4.95 4.51
% working in manufacturing, tract-level 0.41 *** 0.0664
Constant −6.87 *** 1.9104
Observations 849
Partial R2 0.24
Robust F-statistic 38.8
Open in a new tab
***

p<.001;

**

p<.01; p<.05

Second-stage results provide the estimated effect of airborne lead exposure on child cognitive outcomes (Table 5). Results indicate that airborne lead exposure has a negative effect on executive function that remains throughout early childhood. A 1-SD increase in average airborne lead exposure from birth through age 3 leads to a 0.21 SD decrease in EF at age 3 (a decrease of 0.12 on the original EF scale). The negative effects of airborne lead persist in subsequent years, and by age 4 and 5, an increase of 1 SD in airborne lead exposure leads to a decline of 0.23 SD and 0.18 SD in EF, respectively. These correspond to declines of 0.12 and 0.09 on original EF scales. Results also suggest that airborne lead negatively affects child cognition, with a 1SD increase in airborne lead associated with a 0.15 SD decline in IQ measured at age 3, which corresponds to 2.5 points on the original scale.

Table 5.

Coefficients of airborne lead exposure from IV 2SLS model

Lead exposure, (fitted value)
Coefficient SE
Outcomes
Executive Function (3 years) −0.213** −0.008
Executive Function (4 years) −0.227** −0.007
Executive Function (5 years) −0.178** −0.006
WPSSI (3 years) −0.152* −0.212
Open in a new tab

NOTE: Coefficients are standardized beta coefficients. Standard errors are clustered at the tract level.

Discussion

Airborne lead has received relatively less attention as a source of children’s lead exposure since the removal of lead from gasoline. However, airborne lead pollution continues to occur as a byproduct of many mining and manufacturing industries. The current study examined the association between children’s exposure to airborne lead across the first three years of life, and measures of their cognitive development between the ages of 3 and 5. This association was modeled using an instrumental variable approach designed to account for unobserved confounding in the data to enable the estimate of causal associations. The instrumental variable models detected a significant effect for children’s IQ measured at age 3, as well as their EF measured at ages 3, 4, and 5. While airborne lead may represent a relatively lower risk factor compared to ingested sources, results nevertheless indicate an impact of airborne exposure on cognitive development.

Sources of Lead in the Environment

Research examining the relative contributions of environmental lead exposure to children’s blood lead levels regularly identify household dust and water contamination as the largest contributing sources of exposure (Etchevers et al., 2015), although these studies rarely consider airborne lead in analyses. One study of children in a region of Australia with a high concentration of airborne lead found that exposure was associated with higher blood lead levels relative to children in low-pollution regions, but that elevations remained below the 5μg/dL cutoff (Rossi et al., 2012). It is possible that airborne exposure, in isolation, is not a potent risk factor for elevated blood lead. However, the fact that airborne lead does contribute to blood lead levels suggests that it should be considered in the context of the multiple potential sources children may encounter. For instance, approximately 39% of the participants in the current study were drawn from counties in Pennsylvania, which has the highest percentage of children under the age of 6 with blood lead levels greater than 5 μg/dL, largely due to the prevalence of older housing (Shah, Oleske, Gomez, Davidow, & Bogden, 2017). Children exposed to decaying lead paint or corroding plumbing may be placed at additional risk through the additive contributions of airborne pollution. As such, policies establishing maximum exposure thresholds in drinking water or environmental release should consider the potential for cumulative exposure to exceed recommended thresholds. The current findings indicate that airborne pollution is associated with decrements in executive function and IQ and should not be discounted. Although the models employed in the current study controlled for a wide range of demographic factors to demonstrate an independent effect of airborne lead on cognitive development, the burden of airborne exposure is not, in reality, distributed equitably. Airborne exposure is concentrated in certain geographic regions, which are often characterized by multiple risk factors making children in these areas especially vulnerable.

Contextualizing the size of estimated effects in this study is challenging given that correlations between airborne exposure and blood lead levels are unknown in our sample, despite previously documented associations between airborne lead and blood lead levels in young children (Meng et al., 2013). In their meta-analysis, Lanphear and colleagues (2005) found a decrease of 4 IQ points for an increase in blood lead levels from 4 to 35 ug/dL which represented a shift from the 5th to 95th percentile for early childhood blood lead levels in their data. Our estimated negative effect of 2.5 IQ points for one standard deviation increase in airborne lead translates to nearly 8 points when moving from the 5th to 95th percentile in airborne lead exposure, suggesting that very large increases in airborne lead pose a potential risk for cognitive development.

Lead exposure (inhaled or ingested) and executive function outcomes have received less attention, and studies that have examined executive function have tended to focus on middle childhood and adolescence. Although a generally negative relationship between low levels of blood lead and executive function has been found, results are less consistent for toddler and preschool ages (younger than 5 years old) (Albert & Liu, 2020). Researchers have noted that this is in part due to difficulty in measuring these domains in younger children. This study represents a contribution in examining airborne lead exposure on valid and reliable assessments of early childhood (ages 3, 4, and 5) executive function. Within our sample, we find that moving from the 5th to 95th percentile in airborne lead exposure translates to 0.35 decrease in executive function at age 4 (on its original scale). This change is nearly equal to the average difference in executive function within this sample between ages 3 and 4 (mean EF at 3 years = −.53, mean EF at 4 years = −.11), a period of rapid executive function skill development (Zelazo, Blair, and Willoughby, 2016).

Potential Confounding Variables

Given that random assignment to different toxic exposures is impossible and unethical, observational studies must carefully consider all sources of confounding. The IV approach used in this study is intended to account for the unobserved confounding that limits traditional regression or propensity score estimates of the relationship between lead exposure and child outcomes. Prior research has similarly used IV approaches to estimate the causal effects of lead exposures on child and adolescent outcomes, including cognitive ability (Clay et al., 2019) and risky or delinquent behavior (Reyes, 2015; Aizer & Currie, 2019). Our analyses attempted to account for a host of correlated risk factors that may confound the association between lead exposure and cognitive development. These included individual and family demographic factors such as poverty, as well as parental characteristics that could confer genetic (e.g. IQ, history of ADHD) or experiential (e.g. symptoms of psychopathology) risk for children’s development. These also included quantifying children’s exposure to environmental tobacco smoke, a potent source of environmental lead exposure in children (Apostolou et al., 2012), as well as community-level risk factors (i.e. poverty composition at the Census tract) that maybe related to both manufacturing employment and children’s cognitive development. The instrument in combination with the rich set of included measured covariates minimizes these threats to selection bias. However, there is continued need for caution in considering the instrument’s complete independence of outcomes conditional on the observed covariates. In addition, although one strength of the IV approach is its ability to produce causal estimates in the presence of omitted variables, effect estimates from IV methods are often susceptible to increased variance and estimates with wide confidence intervals particularly with a limited sample size.

Limitations and Conclusions

It is important to note that RSEI scores of airborne lead exposure are estimated based on the Toxic Release Inventory; reporting required for all commercial enterprises to track their use and disposal of toxic substances. As such, this is not a direct measure of lead sampled from the air and may therefore underestimate true airborne exposure. One of the most frequently identified sources of lead detected in air samples is the resuspension of lead from the soil (e.g. dust) (Pingitore, Clague, Amaya, Maciejewska, & Reynoso, 2009), and actual airborne lead concentrations may also be affected by water and soil contamination. Given that lead emitted into the air itself settles to the ground and contributes to soil lead levels (Sheets et al., 2001), it is possible that airborne pollution creates an accumulating cycle of exposure as lead transitions repeatedly between soil and air.

This study was limited in its geographical coverage to the initial six counties in North Carolina and Pennsylvania where study participants were born and any counties to which children moved during the study’s defined period of exposure between birth and 36 months. As a function of the original study design (see Vernon-Feagans & Cox, 2013), counties were selected for their lack of urban centers. It is possible, therefore, that results would not replicate in densely populated regions. For instance, assumptions that individuals reside in general proximity to their workplace may not hold in regions with available mass transit systems. In addition, more urban regions may be more likely to show a positive association between pre-1950s housing and manufacturing employment, as investments in new housing developments may be driven toward suburban regions and away from manufacturing sites. The results of the current study demonstrate the value of the RSEI dataset for examining regional effects of toxicity exposure, but it remains important to consider potentially moderating factors of geographical locations in efforts to replicate these findings in other regions of the United States.

The harmful effects of elevated blood lead levels for children are well-known, but airborne lead has received far less attention as a potential source. This study provides evidence of the negative consequences of airborne lead exposure on children’s executive function, which has implications for later developmental outcomes including academic achievement (Ribner et al., 2017). Although some evidence suggests that having nutritional deficiencies, particularly of metals such as iron or zinc, may put certain children at greater risk of absorbing environmental lead (Ravenscroft et al., 2018; Shah-Kulkarni et al., 2016), there is no evidence that nutritional supplementation beyond the correction of a deficiency can reduce blood lead levels (Rosado et al., 2006). Because there are few promising treatments for elevated lead once it is absorbed, regardless of the source, awareness of exposure risk and its consequences is critical. Exposure prevention remains a public health imperative (Kordas, 2017).

Figure 2. IV results of airborne lead exposure on child cognitive outcomes.

Figure 2.

Open in a new tab

Effect estimates (standardized beta coefficients) and confidence intervals indicate a significant negative effect of airborne lead exposure across all cognitive measures (blue).

  • Children’s residential location affects their degree of exposure to airborne lead

  • Instrumental variable models provide an estimate of causal effects of exposure

  • Regional variation in airborne lead is associated with lower cognitive function

Funding:

This research was supported by a grant from National Institutes of Health Office of the Director UG3OD023332-01; UH3OD023332-01 as well as the previous grants from the National Institute of Child Health and Human Development 1PO1HD39667 and 2PO1HD039667 Co-funding was provided by the National Institute of Drug Abuse, NIH Office of Minority Health, National Center on Minority Health and Health Disparities, and the Office of Behavioral and Social Sciences Research. We would like to express our gratitude to all of the families, children, and teachers who participated in this research and to the Family Life Project (FLP) research assistants for their hard work and dedication to the FLP. This study is part of the Family Life Project (https://flp.fpg.unc.edu/).

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Competing Financial Interests: The authors have no financial conflicts to declare.

References

  1. Aizer A, & Currie J (2017). Lead and juvenile delinquency: new evidence from linked birth, school and juvenile detention records. Working Paper 23392, National Bereau of Economic Research. http://www.nber.org/papers/w23392. Retrieved on July 13th 2020. [Google Scholar]
  2. America’s Health Rankings analysis of U.S. Census Bureau, American Community Survey, United Health Foundation, AmericasHealthRankings.org, Accessed July 2020.
  3. Angrist, Joshua D, and Krueger Alan B.. (2001). Instrumental Variables and the Search for Identification: From Supply and Demand to Natural Experiments. Journal of Economic Perspectives 15(4): 69–85. [Google Scholar]
  4. Apostolou A, Garcia-Esquinas E, Fadrowski JJ, McLain P, Weaver VM, & Navas-Acien A (2012). Secondhand tobacco smoke: A source of lead exposure in US children and adolescents. American Journal of Public Health, 102, 714–722. doi: 10.2105/AJPH.2011.300161 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Barg G, Daleiro M, Queirolo EI, Ravenscroft J, Manay N, Peregalli F, & Kordas K (2018). Association of low lead levels with behavioral problems and executive function deficits in schoolers from Montevideo, Uruguay. International Journal of Environmental Research and Public Health, 15, 2735. Doi: 10.3390/ijerph15122735. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Boucher O, Muckle G, Jacobson JL, Carter RC, Kaplan-Estrin M,… Jacobson SW (2014). Domain-specific effects of prenatal exposure to PCBs, mercury, and lead on infant cognition: Results from the environmental contaminants and child development study in Nunavik. Environmental Health Perspectives, 122, 310–316. doi: 10.1289/ehp.1206323. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Browne DR, Husni A, & Risk MJ (1999). Airborne lead and particulate levels in Semarang, Indonesia and potential health impacts. Science of the Total Environment, 227, 145–154. Doi: 10.1016/s0048-9697(99)00022-4. [DOI] [PubMed] [Google Scholar]
  8. Chamberlain AC (1985). Prediction of response of blood lead to airborne and dietary lead from volunteer experiments with lead isotopes. Proceedings of the Royal Society of London Biological Sciences, 224 (1235), 149–182. Doi: 10.1098/rspb.1985.0027. [DOI] [PubMed] [Google Scholar]
  9. Chang XS, Fu K, Li T, Tam L, & Wong G (2018). Corporate environmental liabilities and capital structure. Available at SSRN 3200991. [Google Scholar]
  10. Chiodo LM, Covington C, Sokol RJ, Hannigan JH, Jannise J, Ager J, Greenwald M, & Delaney-Black V (2007). Blood lead levels and specific attention effects in young children. Neurotoxicology & Teratology, 29, 538–546. Doi: 10.1016/j.ntt.2007.04.001. [DOI] [PubMed] [Google Scholar]
  11. Clay K, Portnykh M, & Severnini E (2019). The legacy lead depositions in soils and its impact on cognitive function in preschool-aged children in the United States. Economics and Human Biology, 33, 181–192. [DOI] [PubMed] [Google Scholar]
  12. Dartey E, Berlinger B, Thomassen Y, Ellingsen DG, Odland JO, Nartey VK, Yehoah FA, & Weinbruch S (2014). Bioaccessibility of lead in airborne particulates from car battery repair work. Environmental Science and Process Impacts, 16, 2782–2788. Doi: 10.1039/c4em00455h. [DOI] [PubMed] [Google Scholar]
  13. Dorea JG (2019). Environmental exposure to low-level (Pb) co-occurring with other neurotoxicants in early life and neurodevelopment of children. Environmental Research, 177, 108641. Doi: 10.1016/j.envres.2019.108641. Environmental Protection Agency (2020, 09/17). Understanding RSEI Results. https://www.epa.gov/rsei/understanding-rsei-results [DOI] [PubMed] [Google Scholar]
  14. Etchevers A, Le Tertre A, Lucas J-P, Bretin P, Oulhote Y, Le Bot B, & Glorennec P (2015). Environmental determinants of different blood lead levels in children : A quantile analysis from a nationwide survey. Environment International, 74, 152–159. Doi: 10.1016/j.envint.2014.10.007. [DOI] [PubMed] [Google Scholar]
  15. Frndak S, Barg G, Canfield RL, Quierolo EI, Manay N, Kordas K (2019). Latent subgroups of cognitive performance in lead- and manganese-exposed Uruguayan children: examining behavioral signatures. Neurotoxicology, 73, 188–198. Doi: 10.1016/j.neuro.2019.04.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Froines JR, Baron S, Wegman DH, & O’Rourke S (1990). Characterizatino of the airborne concentrations of lead in U.S. industry. American Journal of Industrial Medicine, 18, 1–17. Doi: 10.1002/ajim.4700180102. [DOI] [PubMed] [Google Scholar]
  17. Gatzke-Kopp LM, Willoughby MT, Warkentien SM, O’Connor T, Granger DA, & Blair C (2018). Magnitude and chronicity of environmental smoke exposure across infancy and early childhood in a sample of low-income children. Nicotine & Tobacco Research. 21, 1665–1672. doi: 10.1093/ntr/nty228 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Gottesfeld P, & Pokhrel AK (2011). Review: Lead exposure in battery manufacturing and recycling in developing countries and amoung children in nearby communities. Journal of Occupational Environmental Hygeine, 8, 520–532. Doi: 10.1080/15459624.2011.601710. [DOI] [PubMed] [Google Scholar]
  19. Griffin TB, Coulston F, Wills H, & Russell JC (1975). Biologic effects of airborne particulate lead on continuously exposed rats and rhesus monkeys. Environmental Quality and Safety, 2, 202–220. [PubMed] [Google Scholar]
  20. Hai DN, Tung LV, Van DK, Bihh TT, Phyong HL, … Khue PM (2018). Lead environmental pollution and childhood lead poisoning at Ban Thi Commune, Bac Kan Province, Vietnam. Biomedical Research International, 18, 5156812. Doi: 10.1155/2018/5156812. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Han Z, Guo X, Zhang B, Liao J, & Nie L (2018). Blood lead levels of children in urban and suburban areas in China (1997 – 2015): Temporal and spatial variations and influencing factors. Science of the Total Environment, 625, 1659–1666. Doi: [DOI] [PubMed] [Google Scholar]
  22. Hanna-Attisha M, LaChance J, Sadler RC, & Schnepp AC (2016). Elevated blood lead levels in children associated with the Flint drinking water crisis: a spatial analysis of risk and public health response. American Journal of Public Health, 106, 283–290. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Harris MH, Gold DR, Rifas-Shiman SL, Melly SJ, Zanobetti A, Coull BA, Schwartz JD, Gryparis A, Kloog I, Koutrakis P, Bellinger DC, Belfort B, Webster TF, White RG, Sagiv SK, & Oken E (2016). Prenatal and childhood traffic-related air pollution exposure and childhood executive function and behavior. Neurotoxicology & Teratology, 57 60–70. Doi: 10.1016/j.ntt.2016.06.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Havlena J, Kanare MS, & Coons M (2009). Factors associated with the seasonality of blood lead levels among preschool Wisconsin children. WMJ, 108, 151–155. [PubMed] [Google Scholar]
  25. Hayes EB, Orbach HG, Fernandez AM, Lyne S, Matte TD, & McElvaine MD (1994). Long-term trends in blood lead levels among children in Chicago: Relationship to air lead levels. Pediatrics, 93, 195–200. [PubMed] [Google Scholar]
  26. Hernán Miguel A., and Robins James M.. (2006). Instruments for Causal Inference: An Epidemiologist’s Dream? Epidemiology, 17(4): 360–372. [DOI] [PubMed] [Google Scholar]
  27. Hwang Y-H, Lin Y-S, Lin C-Y, & Wang I-J (2014). Incense burning at home and the blood lead level of preschoolers in Taiwan. Environmental Science & Pollution Research International, 21, 13480–13487. Doi: 10.1007/s11356-014-3273-1. [DOI] [PubMed] [Google Scholar]
  28. Jedrychowski W, Perera FP, Jankowski J, Mrozek-Budzyn D, Mroz E, … Lisowska-Miszcyk I (2009). Very low prenatal exposure to lead and mental development of children in infancy and early childhood: Krakow prospective cohort study. Neuroepidemiology, 32, 270–278. Doi: 10.1159/000203075. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Joo H, Choi J,H, Burm E, Park H, Hong Y-C, … Ha M (2018). Gender difference in the effects of lead exposure at different time windows on neurobehavioral development in 5-year-old children. Science of the Total Environment, 615, 1086–1092. Doi: 10.1016/j.scitotenv.2017.10.007. [DOI] [PubMed] [Google Scholar]
  30. Kordas K (2017). The “lead diet”: Can dietary approaches prevent or treat lead exposure? Journal of Pediatrics, 185, 224–231.e1. doi: 10.1016/j.jpeds.2017.01.069. [DOI] [PubMed] [Google Scholar]
  31. Lee W, Yoon J-H, Roh J, Lee S, Seok H, Lee J-H, P. K. Jung, & Won J-U (2016). The association between low blood lead levels and the prevalence of prehypertension among nonhypertensive adults in Korea. American Journal of Human Biology, 28, 729–735. Doi: 10.1002/ajhb.22857. [DOI] [PubMed] [Google Scholar]
  32. Manton WI, Angle CR, Stanek KL, Kuntzelman D, Reese YR, & Kuehnemann TJ (2003). Release of lead from bone in pregnancy and lactation. Environmental Research, 92, 139–151. Doi: 10.1016/S0013-9351(03)00020-3. [DOI] [PubMed] [Google Scholar]
  33. Meng Q, Richmond-Bryant J, Davis JA, Cohen J, Svendsgaard D, Brown JS, Tuttle L, Hubbard H, Rice J, Vinikoor-Imler L, Sacks JD, Kirrane E, Kotchmar D, Hines E, & Ross M (2013). Contribution of particle-size fractionated airborne lead to blood lead during the National Health and Nutrition Examination Survey, 1999–2008. Environmental Science & Technology, 48, 1263–1270. Doi: 10.1021/es4039825. [DOI] [PubMed] [Google Scholar]
  34. Moody H, & Grady SC (2017). Lead emissions and population vulnerability in the Detroit (Michigan, USA) metropolitan area, 2006–2013: A spatial and temporal analysis. International Journal of Environmental Research & Public Health, 14, 1445. Doi: 10.3390/ijerph14121445. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Nilsson JP (2009). The long-term effects of early childhood lead exposure: Evidence from the phase-out of leaded gasoline. Working Paper, Institution for Labor Market Policy Evaluation. [Google Scholar]
  36. Obeng-Gyasi E (2019). Lead exposure and cardiovascular disease among young and middle-aged adults. Medical Science, 7, 103. Doi: 10.3390/medsci7110103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Page RA, Cawse PA, & Baker SJ (1988). The effect of reducing petrol lead on airborne lead in Wales, U.K. Science of the Total Environment, 68, 71–77. Doi: 10.1016/00489697(88)90362-2. [DOI] [PubMed] [Google Scholar]
  38. Pingitore NE, Clague JW, Amaya MA, Maciejewska B, & REynoso JJ (2009). Urban airborne lead: X-ray absorption spectroscopy establishes soil as dominant source. PLoS One, 4:e5019. Doi: 10.1371/journal.pone.0005019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Porta D, Narduzzi S, Badaloni C, Bucci S, Cesaroni G, Volelli V, Davoli M, Sunyer J, Zirro E, Schwartz J, & Forastiere F (2016). Air pollution and cognitive development at age 7 in a prospective Italian birth cohort. Epidemiology, 27, 228–236. Doi: 10.1097/EDE.0000000000000405. [DOI] [PubMed] [Google Scholar]
  40. Ravenscroft J, Roy A, Queirolo EI Manay N, Martinez G, Peregalli F, & Kordas K (2018). Drinking water lead, iron and zinc concentrations as predictors of blood lead levels and urinary lead excretion in school children from Montevideo, Uruguay. Chemosphere, 212, 694–704. Doi: 10.1016/j.chemosphere.2018.07.154. [DOI] [PubMed] [Google Scholar]
  41. Ribner AD, Willoughby MT, Blair CB, The Family Life Project Key Investigators Executive function buffers the association between early math and later academic skills. Front Psychol. 2017. 10.3389/fpsyg.2017.00869. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Rico JA, Kordas K, Lopez P, Rosado JL, Vargas GG, Ronquillo D, & Stoltzfus RJ (2006). Efficacy of iron and/or zinc supplementation on cognitive performance of lead-exposed Mexican Schoolchildren: A randomized, placebo-controlled trial. Pediatrics, 117, e518–527. Doi: 10.1542/peds.2005-1172. [DOI] [PubMed] [Google Scholar]
  43. Rinsky JL, Higgins S, Angelon-Gaetz K, Hogan D, Lauffer P, …Moore Z (2018). Occupational and take-home lead exposure among lead oxide manufacturing employees, North Carolina, 2016. Public Health Report, 133, 700–706. Doi: 10.1177/0033354918795442. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Rosado JL, Lopez P, Kordas K, Garcia-Vargas G, Ronquillo D, Alatorre J, & Stoltzfus RJ (2006). Iron and/or since supplementation did not reduce blood lead concentrations in children in a randomized, placebo-controlled trial. Journal of Nutrition, 136, 2378–2383. Doi: 10.1093/jn/136.9.2378. [DOI] [PubMed] [Google Scholar]
  45. Rossi E, McLaughlin V, Joseph J, Bulsara M, Coleman K, Douglas C, & Robertson A (2012). Community blood lead survey with emphasis on preschool children following lead dust pollution in Esperance, Western Australia. Australian & New Zealand Journal of Public Health, 36, 171–175. Doi: 10.1111/j.1753-6045.2011.00814x. [DOI] [PubMed] [Google Scholar]
  46. Sattler JM (2001). Assessment of children: Cognitive applications. (4th ed.). San Diego, CA: [Google Scholar]
  47. Satter.Shadbegian R, & Wolverton A (2010). Location decisions of US polluting plants: Theory, empirical evidence, and consequences (No. 2168-2018-8121). Retrieved from https://ageconsearch.umn.edu/record/280885/.
  48. Shah KK, Oleske JM, Gomez HF, Davidow AL, & Bogden JD (2017). Blood lead concentrations of children in the United States: A comparison of states using two very large databases. Journal of Pediatrics, 185, 218–223. Doi: 10.1016/j.jpeds.2017.01.059. [DOI] [PubMed] [Google Scholar]
  49. Shah-Kulkarni S, Ha M, Kim B-M, Kim E, Hong Y-C, …Ha E-H (2016). Neurodevelopment in early childhood affected by prenatal lead exposure and iron intake. Medicine, 95, e2508. Doi: 10.1097/MD0000000000002508. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Sheets RW, Kyger JR, Biagioni RN, Probst S, Boyer K, & Barke K (2001). Relationship between soil lead and airborne lead concentrations at Springfield, Missouri, USA. Science of the Total Environment, 271, 79–85. Doi: 10.1016/s0048-9697(00)00832-9. [DOI] [PubMed] [Google Scholar]
  51. Taylor MP, Isley CF, & Glover J (2019). Prevalence of childhood lead poisoning and respiratory disease associated with lead smelter emissions. Environment International, 127, 340–352. Doi: 10.1016/j.envint.2019.01.062. [DOI] [PubMed] [Google Scholar]
  52. Thomson GO, Raab GM, Hepburn WS, Hunter R, Fulton M, & Laxen DP (1989). Blood-lead levels and children’s behavior-results from the Edinburgh Lead Study. Journal of Child Psychology and Psychiatry, 30, 515–528. Doi: 10.1111/j.1469-7610.1989.tb00265.x. [DOI] [PubMed] [Google Scholar]
  53. US Department of Transportation, Bureau of Transportation Statistics. (2003). From Home to Work, the Average Commute is 26.4 Minutes, 3(4). Retrieved from https://www.nrc.gov/docs/ML1006/ML100621425.pdf. [Google Scholar]
  54. Environmental Protection Agency US. (2020). Toxics Release Inventory (TRI) 2018 National Analysis. Retrieved from https://www.epa.gov/sites/production/files/2020-02/documents/tri_national_analysis_complete_report.pdf.
  55. Vernon-Feagans L, & Cox M (2013). The family life project: An epidemiological and developmental study of young children living in poor rural communities. Monographs of the Society for Research in Child Development, 78, 1–150. Doi: 10.1111/mono.12047. [DOI] [PubMed] [Google Scholar]
  56. Wechsler D (1955). Wechsler Adult Intelligence Scale (WAIS). Journal of Consulting Psychology, 19, 319–320. Doi: 10.1037/h0039221. [DOI] [Google Scholar]
  57. Wechsler D (2002). WPPSI-III Administration and Scoring Manual. San Antonio, TX: Psychological Corporation. [Google Scholar]
  58. Willoughby MT, Wirth RJ, Blair CB, & The Family Life Project I. (2012). Executive function in early childhood: longitudinal measurement invariance and developmental change. Psychological Assessment, 24, 418–431. Doi: 10.1037/a0025779. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Wilson IH, & Wilson SB (2016). Confounding and causation in the epidemiology of lead. International Journal of Environmental Health Research, 26, 5–6, 467–482. Doi: 10.1080/09603123.2016.1161179. [DOI] [PubMed] [Google Scholar]
  60. Wu Y, Sun J, Wang M, Yu G, Yu L, & Wang C (2018). The relationship of children’s intelligence quotients and blood lead and zinc levels: A meta-analysis and system review. Biological Trace Element Research, 182, 185–195. Doi: 10.1007/s12011-017-1093 [DOI] [PubMed] [Google Scholar]
  61. Zelazo PD, Blair CB, and Willoughby MT (2016). Executive Function: Implications for Education (NCER 2017–2000) Washington, DC: National Center for Education Research, Institute of Education Sciences, U.S. Department of Education. This report is available on the Institute website at http://ies.ed.gov [Google Scholar]

RESOURCES