Associations of perinatal metals and metalloid exposures with early child behavioral development over time in the New Hampshire Birth Cohort Study

doi:10.21203/rs.3.rs-2128833/v1

Download PDF

Research Article

Associations of perinatal metals and metalloid exposures with early child behavioral development over time in the New Hampshire Birth Cohort Study

https://doi.org/10.21203/rs.3.rs-2128833/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 22 Mar, 2023

Read the published version in Exposure and Health →

You are reading this latest preprint version

Background

Research on the neurodevelopmental effects of metal/metalloids has focused mainly on outcomes assessed at one time point, even though brain development progresses over time. We investigated biomarkers of perinatal exposure to metal/loids and changes in child behavior over time.

Methods

We followed 268 participants from the prospective New Hampshire Birth Cohort Study between birth and age five years. We measured arsenic (As), copper (Cu), manganese (Mn), lead (Pb), selenium (Se) and zinc (Zn) in toenails from 6-week-old infants. The Behavioral Symptoms Index (BSI), externalizing, and internalizing symptoms were assessed using the Behavior Assessment System for Children, 2nd edition (BASC-2) at ages 3 and 5 years. Multivariable linear regression was used to estimate associations of metal/loids with behavior change, calculated as the difference in symptom raw scores between 3 and 5 years, in addition to the associations for symptom scores at 3 and 5 years separately. Sex-specific associations were also explored using stratified models and a sex-metal interaction term.

Results

Adjusted associations of metal/loids and change in behavior varied by exposure and outcome. Each 1 µg/g increase in toenail Cu was associated with improved behavior between 3 and 5 years [BSI: β= -3.88 (95%CI: -7.12, -0.64); Externalizing problems: β = -2.20 (95%CI: -4.07, -0.33)]. Increasing Zn was associated with increased externalizing behavior over time (β = 3.42 (95%CI: 0.60, 6.25). Sex-stratified analyses suggested more pronounced associations among boys compared to girls.

Conclusions

Perinatal exposure to metal/loids may alter behavioral development between ages 3 and 5 years. Findings support the need for more research on associations between metal/loids and neurodevelopment over longer time periods.

Environment

metals

neurodevelopment

children

prenatal

postnatal

Brain development is a dynamic and tightly choreographed process especially during prenatal and early postnatal periods (Dubois et al., 2014; Rice & Barone Jr., 2000). During these periods, developmental processes are influenced by genetics as well as positive and negative environmental inputs that can change the course of future brain development (Bellinger et al., 2016a; Debes et al., 2016; Rice & Barone Jr., 2000). Such environmental inputs include essential and neurotoxic metals and metalloids (hereafter referred to as metals) from anthropogenic and natural sources (Grandjean & Herz, 2015; Grandjean & Landrigan, 2014). While essential yet neurotoxic in excess metals such as manganese (Mn), copper (Cu), selenium (Se) and zinc (Zn) are required as part of a healthy diet, general populations can be over exposed to these ubiquitous metals and other toxic metals (e.g., arsenic (As) and lead (Pb)) by inhalation of contaminated air or dust, ingestion of contaminated drinking water or food, or less commonly, through dermal contact (Agency for Toxic Substances and Disease Registry (ATSDR), 2022; Agency for Toxic Substances and Disease Registry, 2012; ATSDR, 2004).

Although extensive research has assessed the relation of prenatal or early life exposure to metals with subsequent neurodevelopment, there is sparse research investigating the influence of environmental metals on change in behavioral function, requiring multiple time points of neurobehavioral assessment (Amorós et al., 2019a; Braun et al., 2017; Claus Henn et al., 2012; Coscia et al., 2010; Jedrychowski et al., 2009; Liu et al., 2018; Polanska et al., 2017; Tong et al., 1998; Wasserman et al., 2000). By investigating outcomes prospectively assessed at more than one time point, we can explore whether associations between prenatal and early life exposure to metals and neurobehavior are persistent over time, appear at later time points (delayed) or are transient and resolve over time (Bellinger et al., 2016a; Rice & Barone Jr., 2000).

Nearly all prior research associating prenatal or early-life metal exposure with neurodevelopment over multiple time points, has focused on cognitive outcomes (Claus Henn et al., 2012; Coscia et al., 2010; Jedrychowski et al., 2009; Liu et al., 2018; Polanska et al., 2017). Findings of this work suggest that associations with early-life metal exposures can vary by timing of outcome measurement. However, other neurodevelopmental domains are also important to quality of life; one in five children living in western societies experiences emotional and behavioral difficulties (Bayer et al., 2008) that relate to enhanced risk of future mental health disorders, substance use disorders, suicidal behavior and criminality (Babinski et al., 2021; Beesdo et al., 2009; Campbell, 1995). Many emotional and behavioral difficulties can be characterized as externalizing behaviors (i.e., negative behaviors projected outward, such as aggression) or internalizing behaviors (i.e., negative behaviors projected inward, such as anxiety and depression). Symptoms of emotional and behavioral difficulties in children can be reliably measured beginning around 2 years of age and in time, early symptoms may resolve or persist to become maladaptive behavioral disorders (e.g., ADHD, depression and anxiety)(Poulou, 2013; Wakschlag et al., 2010). Thus, examining trends in these symptoms, beginning at an early age, may help to inform timely interventions that reduce the risk of adverse long-term outcomes (Bayer et al., 2008; Luby et al., 2019).

We examined data from 268 mother-infant participants in the New Hampshire Birth Cohort Study to estimate associations of perinatal exposure to metals and change in behavior between ages 3 and 5 years. We measured metal concentrations in infant toenails collected at age 6 weeks and estimated associations with the change in composite measures from the Behavioral Assessment System for Children, 2nd edition (BASC-2) including the overall behavioral symptoms index, internalizing problems, and externalizing problems scores. We further explored sex-specific associations and associations at each time point of the outcome.

Study population

Participants for this analysis were part of the prospective New Hampshire Birth Cohort Study (NHBCS), initially designed to investigate associations between contaminants in private water systems and health outcomes of pregnant women and their offspring as described previously (Gilbert-Diamond et al. 2016). Briefly, participants have been recruited through selected prenatal care clinics in New Hampshire. Pregnant women were eligible for the study if they were: (1) aged 18–45 years; (2) pregnant with a singleton infant; (3) literate in English; (4) using a private drinking water system as their primary drinking water source at their residence; and (5) resided at the same address since their last menstrual period with intention to remain there throughout the pregnancy. At the time of this analysis, approximately 1,500 NHBCS participants had children born between 2010–2015 who were age eligible for 3- and 5-year behavioral assessments performed between 2014–2019.

Metals assessment in toenail clippings

From two to eight weeks postpartum, each participant mother was instructed to collect their infant’s nail clippings from all toes and place them in a pre-labeled collection vial (Davis et al., 2014). Infant toenails were collected at a median age of 6 weeks. This time point reflects late pregnancy to early neonatal exposure (Doherty et al. 2020). Information regarding sample preparation and measurement have been detailed elsewhere (Davis et al., 2014; Punshon et al., 2016). Briefly, toenail samples were digested via low-pressure microwave (CEM, Microwave-Assisted Reaction System) and analyzed for trace element concentrations using inductively coupled plasma mass spectrometry (ICP-MS) on an Agilent 7700x (Agilent Technologies, Santa Clara, CA).

We selected six metals [arsenic (As), copper (Cu), manganese (Mn), lead (Pb), selenium (Se) and zinc (Zn)] a priori to include in our analyses based on previous evidence of associating with neurobehavioral outcomes and measuring reliably in toenails. Machine-reported concentrations below the limit of detection (LOD) were used (Whitcomb & Schisterman, 2008). Undetectable concentrations were imputed with the element specific LOD/sqrt(2) (See Table S1 for LOD information). Quality control procedures included digestion of fortified blanks, calibration verification, and inclusion of certified reference materials. Mean recovery of standard reference material [National Institute for Environmental Sciences (Japan) Certified Reference Material No. 13, hair] was 92% (± 7%) and the average coefficient of variation between duplicates was 5% (maximum CV = 7%). The Dartmouth Trace Elements Analysis Core (Hanover, NH) performed all chemical analyses.

Neurobehavioral assessments

Participant neurobehavior was assessed during early childhood (around 3 and 5 years). For each participant, at approximately 3 years of age, a parent or other caregiver was mailed the preschool Parent Rating Scales of the Behavior Assessment System for Children, 2nd edition (BASC-2). At approximately 5 years of age, caregivers completed these same rating scales during an in-person study visit or remotely.

The BASC-2 is a comprehensive assessment of child behavior and emotions (Kamphaus, 2015). The preschool assessment (valid from 2 through 5 years of age) is comprised of 134 questions, in which the parent indicates the frequency of a specific behavior using a Likert scale (Never, Sometimes, Often, Almost Always). Answers from these questions are then combined to derive 12 primary scales, either maladaptive (Aggression, Anxiety, Attention Problems, Atypicality, Depression, Hyperactivity, Somatization, Withdrawal), or adaptive (Activities of Daily Living, Adaptability, Functional Communication, Social Skills). Composite scores derived from the aforementioned scales that we used in primary analyses included: the Behavioral Symptoms Index (BSI), internalizing problems, and externalizing problems. Higher scores on the maladaptive composite scores indicate more problematic behavior. The BASC-2 includes three internal validity indices which help to identify assessments that were inappropriately completed, either due to likely inattention to question content (Response Pattern Index), inconsistent responses to similar questions (Consistency Index), or excessively negative responses (F-Index). Participants were excluded from the analyses if any of the BASC-2 internal validity indices were scored as extreme caution (n = 14 out of 282 (5%)).

Statistical analyses

We calculated univariate and bivariate descriptive statistics. Metal concentrations were natural log (ln) transformed to improve normality of residuals and reduce right skewed distributions. Bivariate correlations between transformed metals were examined using Pearson’s correlation coefficients. Regression diagnostics were examined for the computed outcomes (BASC score change for BSI, internalizing, and externalizing problems) to confirm normality and homoscedasticity of residuals. For each of the three BASC-2 measures, the difference in raw composite scores (BASC score at 5 years - BASC score at 3 years) was used as our primary outcome. Raw scores, instead of t-scores, were used in order to incorporate information on changes that occur over development and to optimize the variability of those changes over time. Regression diagnostics supported the use of untransformed outcome variables in our analyses.

Confounder selection

Covariates, collected from maternal questionnaires at their child’s time of birth and follow ups, were selected based on a priori evidence of potential relationships with both exposure and outcome and included child sex, the number of years between the two assessments (continuous), mother finished college (yes, no), maternal marital status (married, other), maternal smoking status during pregnancy (smoked, never smoked), maternal IQ (continuous), and maternal pregnancy 2010 Alternative Healthy Eating Index score (continuous). Maternal IQ was measured during the 5-yr in person assessments with the Kaufman Brief Intelligence Test, 2nd edition (KBIT2) (Kaufman & Kaufman, 2004). The Alternative Maternal Healthy Eating Index (AHEI-2010) is based on a food frequency questionnaire administered upon enrollment and is a measure of diet quality set by the Dietary Guidelines for Americans for a nutritionally adequate diet. The measure has 11 components for different food groups that are either recommended (e.g., fruits or vegetables) or not recommended (e.g., saturated fats or added sugars) in the daily diet and scores range from 0 to 100 with a higher score indicating a better-quality diet (Emond et al. 2018).

Multiple imputation

Maternal IQ was not assessed in all mothers (27% missing). We used multiple imputation to impute missing values for maternal IQ using chained equations with the MICE package in R. The variables included in the imputation were the covariates we adjusted for in main analyses. We assumed that data missingness was random and did not depend on unobserved data. We generated 40 imputed data sets and pooled across estimates and stand errors using Rubin’s Method (Rubin 2004).

Regression models

Because evidence exists for potential nonlinear associations as well as interactions between several of the metals in our models and neurobehavior (ATSDR, 2004; Brion et al., 2020; Hubbs-Tait et al., 2009; Peters et al., 2011; Vollet et al., 2016), we first explored nonlinearity and higher order interaction using Bayesian Kernel Machine Regression (BKMR) with multiple imputation (Bauer et al. 2020a). This method pools main effect estimates and 95% confidence intervals for each metal from the 40 datasets using Rubin’s method (Rubin 2004). BKMR analyses demonstrated little evidence for non-linearity and interaction between metals and the change in each BASC composite score between 3 and 5 years of age. Given that metals were not strongly correlated (rho = 0.05 to 0.57 with most less than 0.35), we were able to additionally confirm linear associations and no metal-metal interactions by using multivariable linear regression. We 1) compared model fit between models with an additional quadratic term for each metal to a nested model with only linear metal terms using a likelihood ratio test in the MICE R package (Buuren & Groothuis-Oudshoorn, 2011); and 2) statistically tested for interaction between ln-metals by including pairwise interactions between all metals in which statistical significance was indicated by a p-value of < 0.10 for the interaction term. In the final models, multivariable linear regression was used with simultaneous adjustment for all ln-metal measures but no ln-metal interaction or quadratic terms. From using 40 imputed datasets to impute maternal IQ for 27% of our sample, we obtained pooled estimates and 95% confidence intervals by combining information from the 40 mean and variance estimates for each metal using Rubin’s Method (Rubin, 2004).

Secondary analyses

We investigated if associations between metals and change in behavior over time were driven by a single time point by estimating associations at 3 or 5 years of age at assessment separately and comparing magnitudes of effect for each metal to results from primary analyses. Additionally, we explored whether associations between infant toenail metals and behavior differed between males and females given prior evidence of sex-specific metal effects on neurobehavior (Julia A. Bauer et al., 2020), by stratifying models by sex and including an interaction term for each metal (sex*metal) in regression models. Sex-metal interaction terms were considered statistically significant at p < 0.1.

Sensitivity analyses

We assessed the robustness of our findings in several ways. To assess the impact of baseline behavioral score, we scaled each outcome by the baseline score to compute a proportional change in score from 3 to 5 years [(BASC score at 5 years - BASC score at 3 years)/ (BASC score at 3 years)] and compared patterns of metal-behavior associations to the primary analyses. Because we used raw scores in our primary analysis, we repeated main analyses using BASC-2 composite t-scores to confirm the appropriateness of our use of raw scores. Additionally, findings were compared between models where adjustment for age at each assessment was used in lieu of time between assessments. Finally, we used one dataset with maternal IQ imputed and removed 16 influential points with a Cook’s distance < (4/n) (n = 252) and compared to the results from all participants (n = 268).

All statistical analyses were conducted using R (version 3.6.0; R Development Core Team).

A total of 268 participants had available and valid data for BASC-2 outcomes assessed at 3 (mean ± SD: 3.3 ± 0.4) and 5 (mean ± SD: 5.2 ± 0.4) years of age and infant toenail metal measures. About half (49.3%) of participating children were male. Mothers of participants were mostly college educated (78%), married (91.8%), did not smoke during pregnancy (97%), and, on average, were 31.2 years of age upon enrollment (Table 1). Infant toenail Mn-Pb and Cu-Pb concentrations were moderately correlated (rho = 0.57 and 0.50, respectively) with the remainder ranging from rho = 0.05–0.41 (Figure S1). Maternal characteristics and metal concentrations did not differ by child sex (Table 1).

Adjusted associations between all infant toenail metals and the change from 3 to 5-year BASC-2 composite scores varied in direction depending on the element. Generally, each ln increase in toenail Cu was associated with reduced maladaptive behaviors between 3 and 5 years of age across the three BASC-2 composite scores [BSI: β= -3.88 (95%CI: -7.12, -0.64); Externalizing problems: β= -2.20 (95%CI: -4.07, -0.33); Internalizing problems: β= -1.24 (95%CI: -3.24, 0.76)]; (Fig. 1; Table S2). Exploratory sex-stratified analyses suggested that these negative associations with Cu were in boys, not girls with BSI boys: β = -6.39 (95%CI: -10.85, -1.93) vs. girls: β = 1.22 (95%CI: -3.83, 6.26); Externalizing problems boys: β= -3.62 (95%CI: -6.25, -1.00) vs. girls: β = 1.00 (95%CI: -1.77, 3.77); Internalizing problems boys: β= -2.30 (95%CI: -4.88, 0.29) vs. girls: β = 1.26 (95%CI: -2.10, 4.62). P-values for the sex*Cu term for each respective outcome were 0.03; 0.03; 0.11 (Fig. 1; Table S2). Associations of Pb and maladaptive behaviors from 3 to 5 years were generally null in models with all participants, but among boys, Pb was associated with increased changes in maladaptive behavior, while among girls Pb was associated with reduced maladaptive behavior, although sex-specific confidence intervals were imprecise: [BSI boys: β = 4.25 (95%CI: -0.53, 9.02) vs. girls: β= -3.86 (95%CI: -7.90, 0.17); Externalizing problems boys: β = 1.81 (95%CI: -1.00, 4.62) vs. girls: β= -2.66 (95%CI: -4.88, -0.43); Internalizing problems boys: β = 1.06 (95%CI: -1.71, 3.83) vs. girls: β= -2.86 (95%CI: -5.57, -0.16)]. P-values for the sex*Pb term for each respective outcome were 0.008; 0.007; 0.03. Zinc was associated with greater externalizing problems over time in models with all participants (β = 3.42, 95%CI: 0.60, 6.25), and especially among boys in sex-stratified models [β = 5.21 (95%CI: 0.24, 10.19) vs. girls β = 1.49 (95%CI: -1.85, 4.84)], however the p-value for sex*Zn interaction term was not statistically significant (p = 0.33).

We performed several secondary analyses. To investigate if associations between metals and change in behavior over time were driven by a single time point, we estimated associations with behaviors at 3 or 5 years of age separately (Figure S2). In models with all participants, the perinatal metal exposure biomarkers were not associated with BASC-2 measures at 3 years. Consistent with Cu’s association with longitudinal changes in behavior, at 5 years, it was associated with fewer externalizing behaviors: β= -1.98 (95%CI: -3.91, -0.05) vs. 3-year estimate: β = 0.92 (95%CI: -0.61, 2.45). Sex differences were also explored for these secondary analyses. The pattern of negative associations of Cu with longitudinal changes in behavior in boys and of negative associations of Pb with longitudinal changes in behavior in girls was apparent in analyses of 5-year, but not 3-year, assessments (Fig. 1; Figure S3; Figure S4). In general, the magnitude of observed associations was stronger when the change in behavior over time was the outcome as compared to outcomes assessed at a single time (3 or 5 years) (Figure S2).

Results from sensitivity analyses including 1) using proportional change for the outcome [(BASC score at 5 years - BASC score at 3 years)/ (BASC score at 3 years)] (Table S3 and Figure S5), 2) using BASC-2 t-scores for the outcomes (Figure S6) and 3) adjustment for age at each assessment in lieu of time between assessments (Figure S7) were all similar to results from the primary analyses. After removing influential points (n = 16) from our dataset, results were similar to our main analyses but associations between perinatal Cu or Zn and externalizing behavior was attenuated and no longer statistically significant (Table S4).

Table 1

Maternal and child characteristics of participants in the New Hampshire Birth Cohort Study who completed repeated BASC-2 assessments^a at 3- and 5-years of age and had available metal concentrations from infant toenails collected at age 6 weeks^b
	All participants (n = 268) N(%) or mean (SD)	Boys (n = 132) N(%) or mean (SD)	Girls (n = 136) N(%) or mean (SD)
Mother Graduated college Yes No	209 (78.0%) 59 (22.0%)	102 (77.3%) 30 (22.7%)	107 (78.7%) 29 (21.3%)
Married at Enrollment Yes No	246 (91.8%) 22 (8.2%)	121 (91.7%) 11 (8.3%)	125 (91.9%) 11 (8.1%)
Smoked during pregnancy Yes No	8 (3.0%) 260 (97.0%)	4 (3.0%) 128 (97.0%)	4 (2.9%) 132 (97.1%)
Maternal age at enrollment (years)	31.2 (4.1)	30.9 (4.2)	31.5 (4.0)
Maternal IQ	105.7 (13.2)	106.0 (13.3)	105.4 (12.9)
Maternal healthy eating index during pregnancy	54.9 (11.3)	54.8 (11.8)	55.1 (10.9)
Child sex male	49.3%	--	--
Child age at 3-year assessment (years)	3.3 (0.4)	3.4 (0.5)	3.3 (0.4)
Child age at 5-year assessment (years)	5.5 (0.4)	5.5 (0.4)	5.5 (0.4)
Age change between assessment (years)	2.2 (0.4)	2.2 (0.4)	2.2 (0.4)
Behavioral Symptoms Index Score at 3 years ^a	283.5 (32.3)	287.7 (30.3)	279.5 (33.7)
Behavioral Symptoms Index Score at 5 years ^a	286.4 (35.4)	289.5 (37.7)	283.4 (32.9)
Behavioral Symptoms Index Score change	2.9 (29.4)	1.8 (32.0)	3.9 (26.8)
Internalizing Problems Score at 3 years ^a	144.5 (18.7)	143.8 (17.8)	145.1 (19.5)
Internalizing Problems Score at 5 years ^a	148.6 (21.2)	147.4 (20.6)	149.8 (21.8)
Internalizing Problems Score change	4.1 (18.1)	3.6 (18.6)	4.7 (17.7)
Externalizing Problems Score at 3 years ^a	94.3 (13.7)	96.5 (12.8)	92.1 (14.3)
Externalizing Problems Score at 5 years ^a	99.3 (20.6)	102.5 (21.8)	96.2 (19.0)
Externalizing Problems Score change	5.0 (17.7)	6.0 (19.0)	4.1 (16.5)
Infant toenail metals (median (IQR))
Arsenic (µg/g)	0.05 (0.10)	0.05 (0.09)	0.05 (0.11)
Copper (µg/g)	4.6 (6.2)	4.6 (6.3)	4.6 (6.0)
Manganese (µg/g)	1.0 (1.8)	1.0 (1.8)	1.1 (1.7)
Lead (µg/g)	0.3 (0.6)	0.3 (0.6)	0.3 (0.6)
Selenium (µg/g)	1.2 (0.9)	1.1 (0.7)	1.2 (1.1)
Zinc (µg/g)	128.1 (102.7)	127.1 (104.8)	129.1 (92.8)
^a Raw scores from assessments used in analyses. ^b Descriptive statistics averaged across 40 multiple imputed datasets

In our study of U.S. of children from a general population sample of New Hampshire residents, we estimated associations between metals measured in infant toenails, a measure of perinatal metal exposure, and change in maladaptive behaviors from 3- to 5-years of age using repeated measures of children’s behavior. Overall, Cu was associated with improved behavior over time while Zn was associated with worse behavior over time, particularly externalizing behaviors. In secondary analyses both associations were more prominent among boys as compared to girls suggesting that boys may be more sensitive to metal-neurobehavior effects over time for this outcome and age range. Similarly, among boys, but not girls, Pb was associated with worse behavior from 3 to 5 years of age.

Problem behaviors assessed at 3 to 5 years of age are relatively stable in general populations as compared to earlier time points or outcomes (Bartels et al., 2004), consequently investigating this time frame may reduce variability from factors other than perinatal metal exposure. Characterizing maladaptive behavioral trajectories in early childhood (e.g., increasing internalizing or externalizing symptoms over time) could help to identify early indicators of risk for later psychopathology such as depression, anxiety or ADHD (Luby et al., 2019), potentially allowing for earlier intervention, mental health screening, or other preventative or ameliorative efforts (Wakschlag et al., 2010, 2019).

In our study infant toenails were used to estimate perinatal exposure to metals. Little research exists using infant toenails as biomarkers of metal exposures (Davis et al., 2014; Rodrigues et al., 2015). Toenails are composed of keratinous material rich in sulfhydrl groups, which bind metals, allowing toenails to be good potential biomatrices for metals exposure. They are easy to collect and store, and may reflect long term metals exposure due to their slow growth rate (Goullé et al., 2009; Gutiérrez-González et al., 2019; He, 2011). As for validity, evidence exists on the correlation between environmental/biological media and toenail concentrations of As, Mn, Se and Zn (Gutiérrez-González et al., 2019; Rodrigues et al., 2015; Signes-Pastor et al., 2021). Some evidence exists for toenail Pb to correlate with environmental dust in Pakistani urban and industrial areas (Mohmand et al., 2015) and in Zambian children living near mining industry (Ndilila et al., 2014). Toenail Pb has also been correlated with whole blood Pb in children exposed to higher Pb than the general country population in Ohio and Vietnam (Dantzer et al., 2020; Salcedo-Bellido et al., 2021; Sanders et al., 2014). However, toenail Pb was more weakly associated with cognitive outcomes as compared to blood Pb in Ohio children (Dantzer et al., 2020). Among the limited number of studies using toenail Cu as a biomarker of Cu exposure (Gutiérrez-González et al., 2019), toenail Cu has been correlated with indoor dust in populations living near to Cu-related industry (Berasaluce et al., 2019).

Comparisons from existing literature can be made to our study participant’s toenail metal concentrations. One study from Bangladesh with known water Mn and As contamination reported higher 1-month-old infant toenail metal concentrations than our U.S. rural general population cohort (median As: Rodrigues et al.: 0.5–0.7 µg/g vs. present study: 0.05 µg/g; Mn: Rodrigues et al.: 3.3–6.6 µg/g vs. present study: 1.0 µg/g). Another study of school-aged children (6–13 years) living in rural Canada with elevated Mn concentrations in their drinking water had twice the toenail Mn concentrations as our study (Ntihabose et al., 2018).

We observed a protective association of perinatal Cu with change in maladaptive behavioral symptoms from 3 to 5 years of age, which is consistent with the properties of Cu as an essential nutrient that supports brain development and possibly promotes neuroprotection (Bica et al., 2014; Hung et al., 2012; Kaler, 2011). However, findings remain mixed from previous studies in which Cu has been associated with enhanced (Jedynak et al. 2021; Liu et al. 2018), null (Polanska et al. 2017) or poor (Amorós et al. 2019a) neurobehavioral skills in children. Further, sparse research exists for estimating associations between metal mixtures, including Cu and neurodevelopmental trajectories. However, a recent Mexican study reported comparable findings to ours wherein higher second trimester maternal blood Cu was associated with improved cognition between ages 6 and 24 months, assessed with repeated measures of the Bayley Scales of Infant Development, 3rd edition (BSID-III) (Liu et al., 2018). They observed an interaction between Cu and Pb, where the positive slope of the trajectory associated with Cu was lowest at high levels of Pb as compared to low levels of Pb. Our study did not observe interactions between metals. These differences may be due to use of different biomarkers of exposure (our study: infant toenails vs. second trimester blood (Liu et al.)) as well as assessment of different outcomes at different ages (maladaptive behavior from 3–5 years vs. cognition changes from 6–24 months). A prospective study of Spanish children found that first trimester serum Cu was associated with reduced cognition at 12 months and verbal ability at 5 years, although beta estimates were small in magnitude and cognitive assessments were different for the two time points (Bayley Scales of Infant Development (BSID) vs. McCarthy Scales for Children) (Amorós et al., 2019). A Polish prospective birth cohort reported null associations between maternal plasma Cu measured at multiple perinatal time points and infant cognition assessed by the BSID at 12 or 24 months of age (Polanska et al., 2017). Collectively, the aforementioned research supports the need for more studies to investigate the protective and harmful effects of Cu over time and identify which Cu biomarkers may be most informative for children’s neurodevelopment.

We found that Zn was associated with an increase in maladaptive behaviors between 3 and 5 years, especially for externalizing behavior problems, although after removing influential points this association was weakened. While Zn is known for its beneficial effects on neurodevelopment (Adamo & Oteiza, 2010; Boscarino et al., 2021; Georgieff, 2007), recent prospective pediatric studies have observed harmful associations between increased Zn and children’s neurodevelopment (Horton et al., 2018; Polanska et al., 2017; Yang et al., 2013). A birth cohort study in Mexico City found that increased tooth Zn, reflecting exposures from birth to 11 months of age, was associated with more internalizing symptoms, but not with externalizing symptoms or BSI among 8-year-olds assessed with the BASC-2 (Horton et al., 2018). In contrast, our results were null for associations of perinatal metal exposures, including Zn, with internalizing symptoms. This may suggest that in the New Hampshire child population at 3–5 years, internalizing symptoms may not be sensitive to perinatal metals exposure. Or it may be outcome related, that metal associated changes in internalizing symptoms are more easily ascertained at older ages (Horton et al., participants’ mean age was 8 years). In terms of exposure, differences between our study and Horton et. al, include biomarkers and timing of exposure (infant 6-week toenails representing perinatal exposure in our study vs. deciduous teeth reflecting postnatal exposure for the first year of life in Horton et al., 2018). While these exposure time periods overlap around the neonatal period, they otherwise reflect different exposure time frames and potential routes of exposure; our perinatal metal toenail biomarker will additionally reflect maternal exposure, upregulation of nutrient absorption and placental transport of nutrient metals, while the tooth metal biomarker reflects postnatal exposures from birth to 11 months (Horton et al., 2018) and thereby incorporates postnatal infant hand-to-mouth behavior, introduction of solid foods and more infant formula use (Cohen Hubal et al., 2000; Ljung et al., 2011; Zota et al., 2016). Not only do metal exposures vary over the course of late pregnancy through the first of life, but so do neurodevelopmental processes. As such, susceptible time points of neurodevelopmental insult vary between these exposure time periods and could explain differences in our study findings and Horton et al., 2018 (Rice et al., 2000).

In exploratory sex-stratified analyses, associations of metals with longitudinal changes in behavior were typically more pronounced among boys as compared to girls. Among boys, perinatal Cu was associated with reduced problem behaviors over time, while Zn and Pb were associated with increased problem behaviors over time. It has been hypothesized that sex differences in metal-neurodevelopmental associations may be consequent to sex-specific anatomical and functional differences in brain development (Llop et al., 2013). For example, brain regions during development are sexually dimorphic and estrogens, which play a role in neuroprotection, are denser in some areas of developing female brains than in males (Rao & Kölsch, 2003; Scallet & Meredith, 2002; Vahter et al., 2002). In addition, male and female brains may rely on different neural pathways and information processing strategies to complete the same task (Cahill, 2006; Gillies et al., 2014). The few studies investigating sex as an effect modifier in the association between Cu and neurodevelopment reported more pronounced harmful associations between Cu and neurobehavior among boys(Amorós et al. 2019a; Kicinski et al. 2015; Polanska et al. 2017; Zhou et al. 2015) but there is one report of stronger Cu-associated decrements in visuospatial abilities in girls compared to boys (Rechtman et al. 2020). These findings generally support our observation that behavioral changes in boys were more sensitive to Cu than in girls (albeit a protective rather than harmful effect). A prospective study in Spain assessed prenatal Cu at multiple time points of cognitive development and reported an increase of 10 µg/L of 1st trimester maternal serum Cu was associated with poorer cognition on the BSID mental developmental index at 1 year among boys compared to improved performance in girls. Likewise at 5 years, Cu-associated decrements in cognition (using the verbal scale of the McCarthy Scales of Children’s Abilities) were displayed in boys as compared to improved performance in girls, however sex differences were not statistically significant (Amorós et al. 2019a). Mixed findings have also been reported in multiple studies for sex differences in the association between Pb and children’s neurobehavior (Bauer et al. 2020b). For example, one study using three time points (12, 24 and 36 month) of the BSID reported stronger inverse associations between cord blood Pb and 36-month mental development index (MDI) among boys while null results were found in girls (Jedrychowski et al. 2009).

There are limitations for this study. The sample size was relatively small and therefore had limited statistical power to detect associations in sex-stratified analyses. For this reason, we presented these analyses as exploratory. We used one exposure metric for metals collected at one time point. However, unmeasured exposures at other critical developmental time periods may be more or less potent for child maladaptive behavior (Bauer et al. 2020b). This may be important especially when considering sex-specific associations. For example, a prospective Korean study reported sex- and exposure time-specific associations between blood Pb and total, internalizing and externalizing problems in 2 to 5-year-old children. Boys were more susceptible to prenatal Pb exposure, while girls were more susceptible to postnatal exposure (Joo et al. 2018). Using a repeat outcomes design allowed for exploration of changes in behavior over time, an outcome with potentially important functional implications but for which prior research is limited. While estimating metal-behavioral associations in early childhood is important in terms of understanding later mental health etiology, it is also possible that environmental metals are differently associated with behavioral symptoms at later time points of development (Luby et al., 2019; Wakschlag et al., 2010, 2019). For example, a US study of 6–8-year-old children also reported negative associations between Pb (measured in maternal erythrocytes) and behavioral difficulties, however Pb was more strongly associated with behavioral difficulties among girls as compared to boys (Fruh et al. 2019). Having access to only two time points for the outcome also hindered employment of new trajectory research methods (Liu et al. 2018). Other research studies examining associations of environmental metals on an outcome with more than two time points have used linear mixed models (Claus Henn et al., 2012; Coscia et al., 2010; Liu et al., 2018) or generalized estimating equations (Jedrychowski et al., 2009; Wasserman et al., 2000) to increase statistical power and allow for an overall estimate of the exposure effect across time and estimates at individual time points. Given that our data were only from two time points, using the change in BASC scores allowed us to investigate this trajectory question using both linear models as well as to initially explore metal mixture associations using BKMR with change in score as the outcome.

By using more than one time point for the outcome, our study contributes to understanding the potential role of metal exposures on neurodevelopmental trajectories, an area of sparse research (Braun et al. 2017). Neurodevelopmental trajectories reflect the progressive advancement of cognitive and behavioral skills that occurs over time as a child matures. Exposures during critical windows of development may change a trajectory’s course toward a more maladaptive or adaptive phenotype (Bellinger et al., 2016). Alterations in developmental trajectories, in turn, can be manifested as persistent deficits, developmental delays, or transient deficits (Rice & Barone Jr., 2000). Our findings support the potential for changes in behavior over time in young children to be informative for chemical exposure effects. In this context, we were able to identify metals that may improve, and others that may impair, behavioral progression in young children. If replicable, such findings could help guide interventions. In addition, given the potential for maladaptive behavioral trajectories in early childhood (e.g., our study’s observation of metal-associated increasing externalizing symptoms over time) to be an early indicator of risk for later psychopathology such as depression, anxiety or ADHD (Luby et al., 2019), assessment of behavioral changes over time would allow for earlier intervention, mental health screening, or other preventative or ameliorative efforts (Wakschlag et al., 2010, 2019).

We used repeated measures of maladaptive behavior in a cohort of typically developing children to investigate whether perinatal exposure to metals may change behavior between the ages of 3 and 5 years. Overall, Cu was associated with reduced maladaptive behaviors over time while Zn was associated with increased maladaptive behaviors over time. In secondary analyses both associations were more prominent among boys as compared to girls suggesting that boys were more sensitive to metal-neurobehavior effects over this time period, supporting the call for exploration of sex differences in these associations to protect children’s health. Our findings support the need for future research to investigate environmental influences on the progression of childhood neurobehavioral development.

Funding

This work was supported in part by grants from the National Institute of Environmental Health (P01ES022832, P20ES018175, P42ES007373), the National Institute of General Medical Sciences (P20GM104416), and the U.S. Environmental Protection Agency (RD-83544201) and the National Cancer Institute T32CA134286.

Competing interests

The authors state they have no conflicts of interests.

Data Availability

Use of the data may be possible under certain conditions by contacting the New Hampshire Birth Cohort Study Principal Investigator: Margaret Karagas, PhD, ([email protected]).

Ethics Approval and Consent Statement

All participants provided written, informed consent and consent to publish on their results upon enrollment. All protocols were approved by the Dartmouth College Institutional Review Board (Dartmouth Committee for the Protection of Human Subjects Approval Reference #20844).

Adamo AM, Oteiza PI (2010) Zinc deficiency and neurodevelopment: the case of neurons. Biofactors 36(2):117–124. https://doi.org/10.1002/biof.91
Amorós R, Murcia M, González L, Soler-Blasco R, Rebagliato M, Iñiguez C, Carrasco P, Vioque J, Broberg K, Levi M, Lopez-Espinosa MJ, Ballester F, Llop S (2019a) Maternal copper status and neuropsychological development in infants and preschool children. Int J Hyg Environ Health 222(3):503–512. https://doi.org/10.1016/J.IJHEH.2019.01.007
Amorós R, Murcia M, González L, Soler-Blasco R, Rebagliato M, Iñiguez C, Carrasco P, Vioque J, Broberg K, Levi M, Lopez-Espinosa MJ, Ballester F, Llop S (2019b) Maternal copper status and neuropsychological development in infants and preschool children. Int J Hyg Environ Health 222(3):503–512. https://doi.org/10.1016/J.IJHEH.2019.01.007
ATSDR (2004) Interaction Profile: Lead, Manganese, Zinc, and Copper. https://www.atsdr.cdc.gov/interactionprofiles/ip06.html
Babinski LM, Hartsough CS, Lambert NM (2021) Childhood Conduct Problems, Hyperactivity-impulsivity, and Inattention as Predictors of Adult Criminal Activity. J Child Psychol Psychiat 40(3):347–355. https://doi.org/10.1111/1469-7610.00452
Bartels M, Hudziak JJ, van den Oord EJCG, Rietveld MJH, van Beijsterveldt CEM, Boomsma DI (2004) Genetic and environmental mechanisms underlying stability and change in problem behaviors at ages 3, 7, 10, and 12. Dev Psychol 40(5):852–867. https://doi.org/10.1037/0012-1649.40.5.852
Bauer JA, Devick KL, Bobb JF, Coull BA, Bellinger D, Benedetti C, Cagna G, Fedrighi C, Guazzetti S, Oppini M, Placidi D, Webster TF, White RF, Yang Q, Zoni S, Wright RO, Smith DR, Lucchini RG, Henn BC (2020) Associations of a Metal Mixture Measured in Multiple Biomarkers with IQ: Evidence from Italian Adolescents Living near Ferroalloy Industry. Environ Health Perspect 128(9). https://doi.org/https://doi.org/10.1289/EHP6803
Bauer JA, Fruh V, Howe CG, White RF, Henn C, B (2020) Associations of Metals and Neurodevelopment: a Review of Recent Evidence on Susceptibility Factors. Curr Epidemiol Rep 1–26. https://doi.org/10.1007/s40471-020-00249-y
Bayer J, Hiscock H, Ukoumunne O, Price A, Wake M (2008) Early childhood aetiology of mental health problems: a longitudinal population-based study. J Child Psychol Psychiatry Allied Discip 49(11):1166–1174. https://doi.org/10.1111/J.1469-7610.2008.01943.X
Beesdo K, Knappe S, Pine DS (2009) Anxiety and Anxiety Disorders in Children and Adolescents: Developmental Issues and Implications for DSM-V. Psychiatr Clin North Am 32(3):483. https://doi.org/10.1016/J.PSC.2009.06.002
Bellinger DC, Matthews-Bellinger JA, Kordas K (2016a) A developmental perspective on early-life exposure to neurotoxicants. Environ Int 94:103–112. https://doi.org/10.1016/J.ENVINT.2016.05.014
Bellinger DC, Matthews-Bellinger JA, Kordas K (2016b) A developmental perspective on early-life exposure to neurotoxicants. Environment International, vol 94. Elsevier Ltd, pp 103–112. https://doi.org/10.1016/j.envint.2016.05.014
Bica L, Liddell JR, Donnelly PS, Duncan C, Caragounis A, Volitakis I, Paterson BM, Cappai R, Grubman A, Camakaris J, Crouch PJ, White AR (2014) Neuroprotective Copper Bis(thiosemicarbazonato) Complexes Promote Neurite Elongation. PLoS ONE 9(2):e90070. https://doi.org/10.1371/journal.pone.0090070
Boscarino G, Gasparini C, Conti MG, di Chiara M, Terrin G (2021) Zinc levels in neonatal life influence long-term neurodevelopment. J Perinatol 2021 41:5(5):1196–1197. https://doi.org/10.1038/s41372-021-01012-x. 41
Braun JM, Yolton K, Stacy SL, Erar B, Papandonatos GD, Bellinger DC, Lanphear BP, Chen A (2017) Prenatal environmental chemical exposures and longitudinal patterns of child neurobehavior. Neurotoxicology 62:192–199. https://doi.org/10.1016/j.neuro.2017.07.027
Brion LP, Heyne R, Lair CS (2020) Role of zinc in neonatal growth and brain growth: review and scoping review. Pediatr Res 2020 89:7(7):1627–1640. https://doi.org/10.1038/s41390-020-01181-z. 89
van Buuren S, Groothuis-Oudshoorn K (2011) mice: Multivariate Imputation by Chained Equations in R. J Stat Softw 45(3). https://doi.org/10.18637/jss.v045.i03
Cahill L (2006) Why sex matters for neuroscience. Nat Rev Neurosci 7(6):477–484. https://doi.org/10.1038/nrn1909
Campbell SB (1995) Behavior Problems in Preschool Children: A Review of Recent Research. J Child Psychol Psychiatry 36(1):113–149. https://doi.org/10.1111/J.1469-7610.1995.TB01657.X
Claus Henn B, Schnaas L, Ettinger AS, Schwartz J, Lamadrid-figueroa H (2012) Associations of Early Childhood Manganese and Lead Coexposure with Neurodevelopment. Environ Health Perspect 126(1):126–131
Coscia JM, Ris MD, Succop PA, Dietrich KN(2010) Cognitive Development of Lead Exposed Children from Ages 6 to 15 Years: An Application of Growth Curve Analysis. Http://Dx.Doi.Orgi>/10.1076/Chin.9.1.10.14498, 9(1), 10–21. https://doi.org/10.1076/CHIN.9.1.10.14498
Davis MA, Li Z, Gilbert-Diamond D, Mackenzie TA, Cottingham KL, Jackson BP, Lee JS, Baker ER, Marsit CJ, Karagas MR (2014) Infant toenails as a biomarker of in utero arsenic exposure. J Exposure Sci Environ Epidemiol 24(5):467–473. https://doi.org/10.1038/jes.2014.38
Debes F, Weihe P, Grandjean P (2016) Cognitive deficits at age 22 years associated with prenatal exposure to methylmercury. Cortex 74:358–369. https://doi.org/10.1016/j.cortex.2015.05.017
Doherty 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. Environ Epidemiol 4(4):e0106. https://doi.org/10.1097/EE9.0000000000000106
Dubois J, Dehaene-Lambertz G, Kulikova S, Poupon C, Hüppi PS, Hertz-Pannier L (2014) The early development of brain white matter: A review of imaging studies in fetuses, newborns and infants. Neuroscience 276:48–71. https://doi.org/10.1016/J.NEUROSCIENCE.2013.12.044
Emond JA, Karagas MR, Baker ER, Gilbert-Diamond D (2018) Better Diet Quality during Pregnancy Is Associated with a Reduced Likelihood of an Infant Born Small for Gestational Age: An Analysis of the Prospective New Hampshire Birth Cohort Study. J Nutr 148(1):22–30. https://doi.org/10.1093/JN/NXX005
Fruh V, Rifas-Shiman SL, Amarasiriwardena C, Cardenas A, Bellinger DC, Wise LA, White RF, Wright RO, Oken E, Henn C, B (2019) Prenatal lead exposure and childhood executive function and behavioral difficulties in project viva. Neurotoxicology 75:105–115. https://doi.org/10.1016/j.neuro.2019.09.006
Georgieff MK (2007) Nutrition and the developing brain: nutrient priorities and measurement. Am J Clin Nutr 85(2):614S–620. https://doi.org/10.1093/AJCN/85.2.614S. S
Gilbert-Diamond D, Emond JA, Baker ER, Korrick SA, Karagas MR (2016) Relation between in Utero arsenic exposure and birth outcomes in a cohort of mothers and their newborns from New Hampshire. Environ Health Perspect 124(8):1299–1307. https://doi.org/10.1289/ehp.1510065
Gillies GE, Virdee K, McArthur S, Dalley JW (2014) Sex-dependent diversity in ventral tegmental dopaminergic neurons and developmental programing: A molecular, cellular and behavioral analysis. Neuroscience 282:69–85. https://doi.org/10.1016/j.neuroscience.2014.05.033
Goullé JP, Saussereau E, Mahieu L, Bouige D, Groenwont S, Guerbet M, Lacroix C (2009) Application of inductively coupled plasma mass spectrometry multielement analysis in fingernail and toenail as a biomarker of metal exposure. J Anal Toxicol 33(2):92–98. https://doi.org/10.1093/JAT/33.2.92
Grandjean P, Herz KT(2015) Trace elements as paradigms of developmental neurotoxicants: Lead, methylmercury and arsenic. In Journal of Trace Elements in Medicine and Biology (Vol. 31, pp. 130–134). Elsevier GmbH. https://doi.org/10.1016/j.jtemb.2014.07.023
Grandjean P, Landrigan PJ (2014) Neurobehavioural effects of developmental toxicity. Lancet Neurol 13(3):330–338. https://doi.org/10.1016/S1474-4422(13)70278-3
Gutiérrez-González E, García-Esquinas E, de Larrea-Baz NF, Salcedo-Bellido I, Navas-Acien A, Lope V, Gómez-Ariza JL, Pastor R, Pollán M, Pérez-Gómez B (2019) Toenails as biomarker of exposure to essential trace metals: A review. Environ Res 179:108787. https://doi.org/10.1016/J.ENVRES.2019.108787
He K (2011) Trace elements in nails as biomarkers in clinical research. Eur J Clin Invest 41(1):98–102. https://doi.org/10.1111/j.1365-2362.2010.02373.x
Horton MK, Hsu L, Henn BC, Margolis A, Austin C, Svensson K, Schnaas L, Gennings C, Hu H, Wright R, Rojo M, M. T., Arora M (2018) Dentine biomarkers of prenatal and early childhood exposure to manganese, zinc and lead and childhood behavior. Environ Int 121(Pt 1):148–158. https://doi.org/10.1016/j.envint.2018.08.045
Hubbs-Tait L, Mulugeta A, Bogale A, Kennedy TS, Baker ER, Stoecker BJ (2009) Main and interaction effects of iron, zinc, lead, and parenting on children’s cognitive outcomes. Dev Neuropsychol 34(2):175–195. https://doi.org/10.1080/87565640802646759
Hung LW, Villemagne VL, Cheng L, Sherratt NA, Ayton S, White AR, Crouch PJ, Lim S, Leong SL, Wilkins S, George J, Roberts BR, Pham CLL, Liu X, Chiu FCK, Shackleford DM, Powell AK, Masters CL, Bush AI, Barnham KJ (2012) The hypoxia imaging agent CuII(atsm) is neuroprotective and improves motor and cognitive functions in multiple animal models of Parkinson’s disease. J Exp Med 209(4):837–854. https://doi.org/10.1084/jem.20112285
Jedrychowski W, Perera F, Jankowski J, Mrozek-Budzyn D, Mroz E, Flak E, Edwards S, Skarupa A, Lisowska-Miszczyk I (2009) 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(8):503–510. https://doi.org/10.1016/J.EARLHUMDEV.2009.04.006
Jedynak P, Maitre L, Guxens M, Gützkow KB, Julvez J, López-Vicente M, Sunyer J, Casas M, Chatzi L, Gražulevičienė R, Kampouri M, McEachan R, Mon-Williams M, Tamayo I, Thomsen C, Urquiza J, Vafeiadi M, Wright J, Basagaña X, Philippat C (2021) Prenatal exposure to a wide range of environmental chemicals and child behaviour between 3 and 7 years of age – An exposome-based approach in 5 European cohorts. Sci Total Environ 763:144115. https://doi.org/10.1016/J.SCITOTENV.2020.144115
Joo H, Choi JH, Burm E, Park H, Hong YC, Kim Y, Ha EH, 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. https://doi.org/10.1016/J.SCITOTENV.2017.10.007
Kaler SG (2011) ATP7A-related copper transport diseases-emerging concepts and future trends. Nat Reviews Neurol 7(1):15–29. https://doi.org/10.1038/nrneurol.2010.180
Kamphaus RW (2015) Behavior Assessment System for Children, Second Edition (BASC-2). The Encyclopedia of Clinical Psychology. John Wiley & Sons, Inc, pp 1–6. https://doi.org/10.1002/9781118625392.wbecp447
Kicinski M, Vrijens J, Vermier G, Hond E, den, Schoeters G, Nelen V, Bruckers L, Sioen I, Baeyens W, van Larebeke N, Viaene MK, Nawrot TS (2015) Neurobehavioral function and low-level metal exposure in adolescents. Int J Hyg Environ Health 218(1):139–146. https://doi.org/10.1016/j.ijheh.2014.09.002
Levin-Schwartz Y, Henn C, Gennings B, Coull C, Placidi BA, Horton D, Smith MK, Lucchini DR, Wright RO (2021) Integrated measures of lead and manganese exposure improve estimation of their joint effects on cognition in Italian school-age children. Environ Int 146:106312. https://doi.org/10.1016/J.ENVINT.2020.106312
Liu SH, Bobb JF, Henn C, Gennings B, Schnaas C, Tellez-Rojo L, Bellinger M, Arora D, Wright M, Coull BA (2018) Bayesian varying coefficient kernel machine regression to assess neurodevelopmental trajectories associated with exposure to complex mixtures. Stat Med. https://doi.org/10.1002/sim.7947
Llop S, Lopez-Espinosa MJ, Rebagliato M, Ballester F (2013) Gender differences in the neurotoxicity of metals in children. Toxicology 311(1–2):3–12. https://doi.org/10.1016/j.tox.2013.04.015
Luby J, Allen N, Estabrook R, Pine DS, Rogers C, Krogh-Jespersen S, Norton ES, Wakschlag L (2019) Mapping infant neurodevelopmental precursors of mental disorders: How synthetic cohorts & computational approaches can be used to enhance prediction of early childhood psychopathology. Behav Res Ther 123:103484. https://doi.org/10.1016/J.BRAT.2019.103484
Ntihabose R, Surette C, Foucher D, Clarisse O, Bouchard MF (2018) Assessment of saliva, hair and toenails as biomarkers of low level exposure to manganese from drinking water in children. Neurotoxicology 64:126–133. https://doi.org/10.1016/J.NEURO.2017.08.011
Peters C, Muñoz B, Sepúlveda FJ, Urrutia J, Quiroz M, Luza S, de Ferrari Gv, Aguayo LG, Opazo C (2011) Biphasic effects of copper on neurotransmission in rat hippocampal neurons. J Neurochem 119(1):78–88. https://doi.org/10.1111/j.1471-4159.2011.07417.x
Polanska K, Hanke W, Krol A, Gromadzinska J, Kuras R, Janasik B, Wasowicz W, Mirabella F, Chiarotti F, Calamandrei G (2017) Micronutrients during pregnancy and child psychomotor development: Opposite effects of Zinc and Selenium. Environ Res 158:583–589. https://doi.org/10.1016/J.ENVRES.2017.06.037
Poulou MS(2013) Emotional and Behavioural Difficulties in Preschool. Journal of Child and Family Studies 2013 24:2, 24(2), 225–236. https://doi.org/10.1007/S10826-013-9828-9
Punshon T, Li Z, Marsit CJ, Jackson BP, Baker ER, Karagas MR (2016) Placental Metal Concentrations in Relation to Maternal and Infant Toenails in a U.S. Cohort. Environ Sci Technol 50(3):1587–1594. https://doi.org/10.1021/acs.est.5b05316
Rechtman E, Curtin P, Papazaharias DM, Renzetti S, Cagna G, Peli M, Levin-Schwartz Y, Placidi D, Smith DR, Lucchini RG, Wright RO, Horton MK(2020) Sex-specific associations between co-exposure to multiple metals and visuospatial learning in early adolescence. Translational Psychiatry 2020 10:1, 10(1), 1–10. https://doi.org/10.1038/s41398-020-01041-8
Rice D, Barone S Jr (2000) Evidence from Humans and Animal Models Development of the Brain in Utero. Environ Health Perspect 108(January):511–533. https://doi.org/10.1289/ehp.00108s3511. Critical Periods of Vulnerability for the Developing Nervous System: Evidence from Humans and Animal Models Critical Periods of Vulnerabilityfor the Developing Nervous System
Rodrigues EG, Kile M, Dobson C, Amarasiriwardena C, Quamruzzaman Q, Rahman M, Golam M, Christiani DC (2015) Maternal–infant biomarkers of prenatal exposure to arsenic and manganese. J Exposure Sci Environ Epidemiol 2015 25:6(6):639–648. https://doi.org/10.1038/jes.2015.45. 25
Rubin DB(2004) Multiple Imputation for Nonresponse in Surveys. In Wiley-Interscience
S Y, N H, K H (2010) Growth rate of human fingernails and toenails in healthy American young adults. J Eur Acad Dermatology Venereology: JEADV 24(4):420–423. https://doi.org/10.1111/J.1468-3083.2009.03426.X
Tong S, Baghurst PA, Sawyer MG, Burns J, McMichael AJ (1998) Declining Blood Lead Levels and Changes in Cognitive Function During Childhood: The Port Pirie Cohort Study. JAMA 280(22):1915–1919. https://doi.org/10.1001/JAMA.280.22.1915
Vahter M, Åkesson A, Lidén C, Ceccatelli S, Berglund M (2007) Gender differences in the disposition and toxicity of metals. Environ Res 104(1):85–95. https://doi.org/10.1016/J.ENVRES.2006.08.003
Vollet K, Haynes EN, Dietrich KN (2016) Manganese Exposure and Cognition Across the Lifespan: Contemporary Review and Argument for Biphasic Dose-Response Health Effects. Curr Environ Health Rep 3(4):392–404. https://doi.org/10.1007/s40572-016-0108-x
Wakschlag LS, Roberts MY, Flynn RM, Smith JD, Krogh-Jespersen S, Kaat AJ, Gray L, Walkup J, Marino BS, Norton ES, Davis MM (2019) Future Directions for Early Childhood Prevention of Mental Disorders: A Roadmap to Mental Health, Earlier. J Clin Child Adolesc Psychology: Official J Soc Clin Child Adolesc Psychol Am Psychol Association Div 53(3):539. https://doi.org/10.1080/15374416.2018.1561296. 48
Wakschlag LS, Tolan PH, Leventhal BL (2010) “Ain’t Misbehavin”: Towards a Developmentally-Specified Nosology for Preschool Disruptive Behavior. J Child Psychol Psychiatry Allied Discip 51(1):3. https://doi.org/10.1111/J.1469-7610.2009.02184.X
Wasserman GA, Liu X, Popovac D, Factor-Litvak P, Kline J, Waternaux C, LoIacono N, Graziano JH (2000) The Yugoslavia Prospective Lead Study: contributions of prenatal and postnatal lead exposure to early intelligence. Neurotoxicol Teratol 22(6):811–818. https://doi.org/10.1016/S0892-0362(00)00106-9
White R, Campbell R, Echeverria D, Knox S, Janulewicz P (2009) Assessment of neuropsychological trajectories in longitudinal population-based studies of children. J Epidemiol Commun Health 63(Suppl 1):i15–26. https://doi.org/10.1136/jech.2007.071530. Suppl 1
Yang X, Yu XD, Fu HH, Li LL, Ren TH (2013) Different levels of prenatal zinc and selenium had different effects on neonatal neurobehavioral development. Neurotoxicology 37:35–39. https://doi.org/10.1016/J.NEURO.2013.04.001
Zhou G, Ji X, Cui N, Cao S, Liu C, Liu J, Zhou G, Ji X, Cui N, Cao S, Liu C, Liu J (2015) Association between Serum Copper Status and Working Memory in Schoolchildren. Nutrients 7(9):7185–7196. https://doi.org/10.3390/nu7095331

SuppMBASCtrajectory9.23.22.docx

Download PDF

Journal Publication

published 22 Mar, 2023

Read the published version in Exposure and Health →

Reviewers agreed at journal
09 Oct, 2022
Reviewers invited by journal
04 Oct, 2022
Editor invited by journal
04 Oct, 2022
Editor assigned by journal
04 Oct, 2022
First submitted to journal
03 Oct, 2022

You are reading this latest preprint version

Associations of perinatal metals and metalloid exposures with early child behavioral development over time in the New Hampshire Birth Cohort Study

Status:

Journal Publication

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

Introduction

Methods

Study population

Metals assessment in toenail clippings

Neurobehavioral assessments

Statistical analyses

Confounder selection

Multiple imputation

Regression models

Secondary analyses

Sensitivity analyses

Results

Discussion

Conclusions

Declarations

References

Supplementary Files

Status:

Journal Publication

Version 1