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: Environ Res. 2020 Nov 18;194:110495. doi: 10.1016/j.envres.2020.110495

In utero pyrethroid pesticide exposure in relation to autism spectrum disorder (ASD) and other neurodevelopmental outcomes at 3 years in the MARBLES longitudinal cohort

Jacqueline M Barkoski 1, Claire Philippat 2, Daniel Tancredi 3, Rebecca J Schmidt 1,4, Sally Ozonoff 4, Dana Boyd Barr 5, William Elms 1, Deborah Bennett 1,*, Irva Hertz-Picciotto 1,3
PMCID: PMC7946720  NIHMSID: NIHMS1658966  PMID: 33220244

Abstract

Background:

We assessed the relationships between prenatal pyrethroid pesticide exposure and autism spectrum disorders (ASD) or non-typical development (non-TD) at 3 years.

Methods:

Participants were mother-child pairs (n=201) in the MARBLES (Markers of Autism Risk in Babies-Learning Early Signs) cohort. Because familial recurrence risk is high, MARBLES enrolls pregnant women with a family history of ASD. Children were clinically assessed at 3 years of age and classified into 3 outcome categories: ASD, typically developing (TD), or non-TD (and not meeting criteria for ASD). Repeated maternal second and third trimester urine samples were analyzed for pyrethroid metabolite 3-phenoxybenzoic acid (3-PBA). Multinomial logistic regression was used to obtain relative risk ratios (RRR) linking 3-PBA concentrations averaged across each trimester and over pregnancy with child’s outcome, either ASD or non-TD vs. TD. Models were adjusted for specific gravity, maternal pre-pregnancy BMI, prenatal vitamin use, birth year, home-ownership, and TCPy (3,5,6-trichloro-2-pyridinol) pregnancy concentrations.

Results:

The median specific gravity corrected 3-PBA concentration of all samples was 1.46 ng/ml. Greater second trimester 3-PBA concentrations were associated with modestly elevated relative risk ratios of ASD (RRR: 1.50 (95% CI 0.89 to 2.51), p = 0.12). There were no differences between non-TD and TD.

Conclusions:

This study found no evidence for differences in 3-PBA comparing non-TD with TD. A moderately elevated RRR was found comparing urinary 3-PBA concentrations for ASD versus TD; however, the confidence interval was wide and hence, these findings cannot be considered definitive.

Keywords: pyrethroid, autism, MARBLES, pesticide, neurodevelopment, pregnancy

1. Introduction

Autism spectrum disorder (ASD) is a severe neurodevelopmental disorder affecting one in 54 children in the United States (Maenner et al., 2016), and the prevalence has increased over the past 20 years. The etiology of ASD is not understood, but there are likely multiple contributing factors. Findings from studies on twins show genetic factors contribute considerably (Bailey et al., 1995; Folstein and Rutter, 1977), but environmental exposures also play a substantial role (Bai et al., 2019; Constantino et al., 2013; Gaugler et al., 2014; Hallmayer et al., 2011; Sandin et al., 2014). There is growing evidence to support the notion that environmental exposures such as air pollution (Yang et al., 2017), maternal metabolic conditions (Krakowiak et al., 2012; Li et al., 2016), and pesticides (Eskenazi et al., 2007; Roberts et al., 2007; Shelton et al., 2014a; von Ehrenstein et al., 2019) increase risk for ASD, and prenatal vitamin use may reduce risk (Guo et al., 2019; Levine et al., 2018; Schmidt et al., 2011; Schmidt et al., 2019; Schmidt et al., 2012; Surén et al., 2013).

Pyrethroids are a class of synthetic insecticides which act on voltage gated sodium channels in the nerve cell membrane (ASTDR, 2003). Pyrethroids cause persistent opening of the sodium ion channels allowing more sodium ions to pass and depolarizing the neuronal membrane (Shafer et al., 2005). In rodents pyrethroid exposures have been shown to affect neurotransmission, possibly altering brain development (Aziz et al., 2001; Malaviya et al., 1993). Pyrethroids have also been found to transcriptionally mimic autism in high-throughput screening using RALS-Seq methods (Simon et al., 2019). Animal studies demonstrated that although pyrethroids exhibit relatively low acute toxicity, there is potential for developmental neurotoxicity (Shafer and Hughes, 2010; Shafer et al., 2005; Soderlund, 2012).

Pyrethrum is a substance produced by the Chrysanthemum flower, and is toxic to insects. Pyrethroids are synthetic forms that were developed to be more stable in sunlight allowing greater persistence in the environment. With decreased photodegradation pyrethroid residue can persist in indoor environments for weeks after application (Starr et al., 2014). Pyrethroid pesticides are included in over 3,500 EPA registered products used in and around homes, on pets, in mosquito control and in agriculture (U.S. EPA, 2013). In 2001, the U.S. EPA banned the use of organophosphate pesticides (OP) in household products, which led to an increase of pyrethroid pesticides in this market (EPA, 2010). Pyrethroids are used, to a lesser extent, in agriculture particularly on crops such as cotton, vegetables, and nuts. Exposures occur through residential sources and residues on food (Lu et al., 2006a; Lu et al., 2006b).

Associations between organochlorine, organophosphate, and pyrethroid pesticides and ASD risk have been observed in large case-control studies (Roberts et al., 2007; Shelton et al., 2014b; von Ehrenstein et al., 2019). These studies have linked agricultural pesticide applications with maternal addresses during pregnancy assess pesticide exposures and studies utilizing biomarkers of pregnancy pesticide exposure and ASD have largely focused on OP’s. In one study pregnancy OP exposures were associated with pervasive developmental disorder (a type of autism spectrum disorder) at 24 months (Eskenazi et al., 2007). More recently, pregnancy OP biomarker concentrations were linked with ASD among girls in a high-risk population but not in the full population (Philippat et al., 2018). The research on pesticides and ASD is growing but studies on pyrethroid pesticides and ASD are limited.

Findings for pyrethroid pesticide and child cognitive and behavioral outcomes are sparse. Pyrethroid biomarker concentrations from pregnancy urine samples have been associated with lower cognitive scores (Watkins et al., 2016), and increased behavioral difficulties (Viel et al., 2017) in early childhood. Detectable levels of pyrethroid biomarkers measured in prenatal urine were associated with a range of behavioral and executive functioning deficits (Furlong et al., 2017). Pyrethroids used on agricultural fields, in combination with other pesticides, near maternal residence was associated with lower IQs (Coker et al., 2017). In one study pregnancy air monitor samples of PBO (a synergist often in pyrethroid formulations) were associated with lower cognitive scores at 36 months (Horton et al., 2011). To our knowledge, prenatal pyrethroid metabolites have not yet been examined in relation to risk for ASD in the child.

Given the current state of knowledge on widespread exposures and growing use of pyrethroid pesticides and their neurotoxic properties, we investigated the effects of in utero pyrethroid pesticide exposure in relation to both children’s diagnosis of ASD or non-typical development (non-TD) at 3 years. The study population consisted of mother-child pairs from the MARBLES (Markers of Autism Risk in Babies – Learning Early Signs) Study, a longitudinal cohort of younger siblings of children with ASD in northern California.

2. Materials and methods

2.1. Study Population

The present study included 201 mother-child pairs enrolled in the MARBLES Study from 2007 – 2014 with pesticide metabolite concentrations from urine samples collected during pregnancy and child neurodevelopmental assessments. The MARBLES Study enrolls pregnant women from northern California whose expected child or children have an elevated risk for ASD. The vast majority of women had a previous child with ASD and a handful had a strong family history; these factors put the child they are expecting at higher risk to develop ASD Hertz-Picciotto et al., 2018). In this cohort, 23% of the children were diagnosed with ASD. Families were recruited from lists of children receiving services for ASD through the California Department of Developmental Services, from other studies at the University of California, Davis Medical Investigation of Neurodevelopmental Disorders (MIND) Institute, and from other referrals and self-referrals. Mothers were enrolled prior to or during pregnancy and were followed through pregnancy; infants were followed from birth to 3 years. Inclusion criteria for the full study were: (i) mother or father had one or more child(ren) with ASD and/or the gestating younger child had an older half-sibling or an equivalent or closer blood relative with ASD; (ii) mother was at least 18 years of age or older; (iii) mother was already pregnant or planning a pregnancy and biologically able to become pregnant; and (iv) mother lived within 2 hours of the Davis/Sacramento region at time of enrollment. The University of California, Davis (UCD) institutional review board approved the MARBLES Study and informed consent was obtained from each participant.

2.2. Child Neurodevelopmental Assessment

Participants are assessed at the MIND Institute by expert examiners at four ages, including 36 months, when a comprehensive diagnostic assessment is conducted. The Mullen Scales of Early Learning (MSEL), a norm-referenced developmental measure, is administered at each of these time points. Four subscales (visual reception, fine motor skills, receptive language, and expressive language) are combined into an overall composite score (Mullen, 1995). During the child’s 3-year visit, expert clinicians evaluate children on the Autism Diagnostic Observation Scale (ADOS), the gold-standard diagnostic tool (Lord et al., 2008). Three categories of children’s outcomes at 3 years are defined using an algorithmic approach previously developed by the Baby Siblings Research Consortium (Chawarska et al., 2014; Ozonoff et al., 2014), which takes into account scores on clinical assessments including the ADOS and MSEL. The first is ASD, defined by a child scoring at or above the ASD cutoff on the ADOS and meeting Diagnostic and Statistical Manual of Mental Disorders −5 (DSM-5) criteria for ASD. The second diagnostic category, non-typical development (non-TD), consists of those who have two or more MSEL subscale scores that are more than ≥ 1.5 standard deviation (SD) below the normative mean, or at least one MSEL subscale score that is more than 2 SD below the normative mean, and/or an elevated ADOS within 3 points of the ASD cutoff. The third category is typical development (TD), defined as a child who scores above 2.0 SD below the normative mean on all MSEL subscales, no more than one MSEL subscale that is between 1.5 and 2.0 SD below the normative mean; and an ADOS higher than 3 points below the ASD cutoff.

2.3. Exposure Data Collection and Analysis

During each trimester of pregnancy, participants were instructed to collect three first morning void (FMV) urine samples (one week apart) and one 24-hour urine sample. Samples were stored in home freezers and then collected during home visits and transported to UCD where they were then stored in −80° C freezers.

For analytical efficiency and cost-effectiveness, we pooled samples for participants with three or more samples in a trimester (Barkoski et al., 2018; Philippat et al., 2018; Shin et al., 2018). Gestational age was calculated using the date of the mothers’ last menstrual period (LMP) and, when available, pregnancy ultrasound information from medical records. The 1st trimester included gestational ages from LMP through 13 weeks, the 2nd trimester included gestational ages 14 – 27 weeks, and the 3rd trimester included gestational ages 28 weeks to birth. For each trimester the first sample was analyzed individually and the remaining samples were pooled. For participants with 1 – 2 samples in a trimester each sample was analyzed individually. Nine mothers had all of their spot and 24hr samples analyzed individually as part of a longitudinal evaluation of pesticide concentrations and their variability. Details of the collection methods and pooling strategy are also outlined in Barkoski et al. 2018 and Shin et al. 2018.

We had fewer samples from the 1st trimester because on average participants completed their first study visit early in the 2nd trimester (gestational week 18). Therefore, we only analyzed samples from the 2nd and 3rd trimesters. A total of 1,030 primary samples were collected from the 2nd and 3rd trimesters. After pooling samples we had 638 urine samples, representing 201 mother-child pairs, analyzed for pesticide concentrations.

Specific gravity (SG) was measured from urine samples with a handheld refractometer (Atago Urine Specific Gravity Refractometer, PAL 10-S) at UCD. Distilled water was used to calibrate between each measurement. Then, urine samples were shipped overnight on dry ice to Emory University’s Rollins School of Public Health where they were analyzed for pesticide metabolite 3-phenoxybenzoic acid (3-PBA). Chemical analyses of 3-PBA were conducted according to previously established methods (Olsson et al., 2004). Briefly, 2 mL urine samples were spiked with an internal standard mixture consisting of isotopically labeled 3-PBA, and incubated with β-glucuronidase/sulfatase to liberate conjugated metabolites. The hydrolysates were extracted using mixed-mode solid-phase extraction cartridges and elutes were concentrated and analyzed using high-performance liquid chromatography/tandem mass spectrometry with both quantification and confirmation ions monitored (Barr et al., 2010). Samples were also analyzed for trans-2,2-(dichlorovinyl)-2,2-dimethylcyclopropane carboxylic acid (tDCCA), but over 35% of samples were <LOQ, and the correlation between tDCCA and 3-PBA was 0.88, and thus they were not included in the analysis.

2.4. Statistical Analyses

The goal of this project was to examine the relationship of quantified prenatal pyrethroid exposure with the child’s risk for either ASD or non-TD, as compared with TD. All statistical analyses were performed with SAS statistical software version 9.4 (SAS, 2002–2012). Pesticide concentrations below the limit of detection (LOD) were assigned a value of LOD/√2 where LOD for 3-PBA was 0.1 ng/mL. To account for urine dilution, metabolite concentrations were SG corrected using the formula described in Hauser et al. Pc = P x [(SGp – 1)/(SG – 1)] (Hauser et al., 2004), where Pc is the SG-corrected pesticide metabolite concentration (ng/mL), P is the measured metabolite concentration in ng/mL, SG is the specific gravity of the urine sample, and SGp (1.012) is the median specific gravity across all of the MARBLES Study samples. The SG-corrected 3-PBA concentrations were included in all analyses as the main exposure variable.

Because there are a varying number of urine samples per mother available for analysis, we calculated the average pesticide metabolite concentrations (Cave) for each trimester. We analyzed the first sample individually, and all remaining samples from the trimester as a pooled sample. In order for the average trimester concentration to have equal contribution from each sample, we multiplied the concentration of the pooled sample by the number of samples in the pool prior to adding the value to the sample analyzed individually. The average pesticide metabolite concentration for each participant was calculated using the following formula: Cave = (Cind + Cpooled x Npooled)/(Nind + Npooled), where Cind is the pesticide metabolite concentration in the individual sample, Cpooled is the pesticide metabolite concentration in the pooled sample, Npooled is the number of composites in the pooled sample, and Nind is the number of the individual sample (=1). Due to a skewed distribution, average SG-corrected 3-PBA concentrations were natural log transformed.

Participants had variable compliance with the urine sample protocol and on average, participants collected 5 (Standard Deviation = 2.6) urine samples during the second and third trimesters. We used analytical weights to account for the unbalanced number of urine samples per pregnancy and the correlations among samples that were averaged together from each pregnancy to form that woman’s average exposure, giving higher weights to more representative samples. These analytical weights reflected the “effective number” (Chih, 2011) of samples used to quantify a woman’s average exposure for a given measurement: weightij = ni/(1 + (ni – 1)ICCj), where ni is the number of individual urine samples that were pooled together for woman I (from a given trimester, if applicable) and ICCj is the within-woman intraclass correlation coefficient for compound j. Intraclass correlation coefficients for each compound were estimated using compound-specific between-mother and within-mother variance component estimated from maximum-likelihood analysis of pooled-sample and spot-sample observations, using a user-written program in SAS PROC NLMIXED, accounting for the variable number of samples collected for each pregnancy. The ICC for 3-PBA from repeated spot and pooled urine samples was 0.16 and this value was used for analytical weights in regression models.

Potential confounders were selected based on the literature regarding associations of covariates with pyrethroid exposure and with child neurodevelopment. We used a Directed Acyclic Graph (DAG) to represent assumed causal relationships between the exposure, outcome and covariates (Supplemental Material, Fig. S1). Bivariate associations between natural log-transformed 3-PBA SG-corrected concentrations and demographic factors were assessed using Kruskal–Wallis tests of association for categorical and with linear regression for continuous demographic variables. Covariates were retained in the final multivariable regression model based on whether addition of the covariate to the model changed the exposure beta coefficient by 10% or more. The final models were adjusted for maternal pre-pregnancy BMI (continuous), homeownership (owner, non-owner), and self-reported prenatal vitamin use during the first month of pregnancy (yes, no). Child’s birth year (continuous) also met criteria to be included. However, we also determined that there was a nonlinear association between year and natural log transformed 3-PBA concentrations based on Akaike information criterion (AIC) and standard errors. Hence, in all adjusted regression models year was centered at 2010 and linear and squared centered-year terms were included as covariates. Finally, we also adjusted for organophosphate (OP) exposure using natural log transformed TCPy-SG corrected (chlorpyrifos metabolite) average pregnancy concentrations because pyrethroid and OP pesticides have been used interchangeably, a large body of evidence on prenatal OP pesticides, much of it based on chlorpyrifos, has demonstrated their associations with impairments in child neurodevelopment (Eskenazi et al., 2007; Hertz-Picciotto et al., 2018; Rauh et al., 2006), and chlorpyrifos was the most commonly and widely used OP pesticide in the U.S. during the period of these MARBLES pregnancies. Other covariates that were considered but did not meet the confounding criteria were maximum parental education (less than college, college degree, graduate or professional degree), mother’s age at delivery (continuous years), child’s race/ethnicity (white, non-white, Hispanic), mother’s place of birth (US born, non-US born), parity (continuous), season of conception (warm, cold), and child’s sex (male, female).

Multinomial logistic regression models were used to test the association between natural log-transformed urinary prenatal 3-PBA concentrations, overall and within the 2nd and 3rd trimesters, and child diagnosis at 3 years. We obtained crude and adjusted relative risk ratio (RRR) estimates from multinomial logistic regression models for ASD and non-TD where TD children served as the reference in all models. Analytical weights, described above, were applied to multinomial logistic regression models, using weight statements in SAS PROC SURVEYLOGISTIC (for multinomial logistic regression models).

3. Results

MARBLES study population characteristics for 201 mother-child pairs with prenatal 3-PBA concentrations, from 638 urine samples, and a diagnosis at 36 months of age are reported in Table 1. The number of urine samples collected in a trimester ranged from 1 – 6 samples and the majority of mothers (71%) collected 2 samples in a trimester. The majority of children were boys (60.2%) and slightly more than half of the participants were white (55.2%). Births included in this study ranged from years 2007 – 2014 with the most children born in 2009. Year-to-year fluctuations are to be expected in this type of study, as the pool of potential participants in small, and the study requires a significant time commitment. In the present study population, mothers’ mean age at delivery was 34 years and the majority of parents were college educated (68.8%) or owned a home (62.1%).

Table 1:

Characteristics of mother-child pairs from the Markers of Autism Risks in Babies – Learning - Early Signs (MARBLES) cohort with pesticide biomarker concentrations from urine samples collected during pregnancy and child neurodevelopmental assessments (n=201, 2007 – 2014.

Participant Characteristics N % ASD Non-TD TD
Child’s sex
 Boys 121 60.2 33 (73.3%) 34 (63.0%) 54 (52.9%)
 Girls 80 39.8 12 (26.7%) 20 (37.0%) 48 (47.1%)
Birth year
 2007 –2008 48 23.9 4 (8.9%) 11(20.4%) 33 (32.4%)
 2009 48 23.9 11 (24.4%) 13 (24.1%) 24 (23.5%)
 2010 – 2011 50 24.9 13 (31.1%) 12 (22.2%) 24 (23.5)
 2012 – 2014 55 27.4 16 (35.6) 18 (33.3%) 21 (20.5%)
Child’s race/ethnicity
 White 111 55.2 21 (46.7%) 27 (50.0%) 63 (61.8%)
 Non-white (Black or multiracial) 31 15.4 8 (17.8%) 11 (20.4%) 12 (11.8%)
 Hispanic 59 29.4 16 (35.6%) 16 (29.6%) 27 (26.5%)
Gestational age at delivery (weeks)
 ≤ 37 weeks 20 10.0 2 (4.4%) 4 (7.4%) 14 (13.7%)
 > 37 weeks 181 90.0 43 (95.6%) 50 (93.6%) 88 (86.3%)
Parity
 1* 83 43.0 14 (33.3%) 22 (42.3%) 47 (47.5%)
 > 1 110 57.0 28 (66.7%) 30 (57.7%) 52 (52.5%)
Season of birth (months)
 Cold 100 49.8 20 (44.4%) 24 (44.4%) 56 (54.9%)
 Warm 101 50.3 25 (55.6%) 30 (55.6%) 46 (45.1%)
Maternal pre-pregnancy BMI
 Normal/ underweight 106 53.0 20 (45.5%) 27 (50.0%) 59 (57.8%)
 Overweight 49 24.5 13 (29.6%) 13 (24.1%) 23 (22.6%)
 Obese 45 22.5 11 (25.0%) 14 (25.9%) 20 (19.6%)
Mom’s age at delivery
 < 35 years 105 52.2 19 (42.2%) 29 (53.7%) 57 (55.9%)
 ≥ 35 years 96 47.8 26 (57.8%) 25 (46.3%) 45 (44.1%)
Prenatal vitamin use
 Yes 109 55.9 15 (35.7%) 30 (57.7%) 64 (63.4%)
 No 86 44.1 27 (64.3%) 22 (42.3%) 37 (36.6%)
Parental education
 Less than college 62 31.2 13 (30.2%) 20 (37.0%) 29 (28.4%)
 Bachelor degree 89 44.7 21 (48.8%) 21 (38.9%) 47 (46.1%)
 Graduate or professional degree 48 24.1 9 (20.9%) 13 (24.1%) 26 (25.5%)
Homeowner
 Yes 121 62.1 23 (54.8%) 31 (59.6%) 67 (66.3%)
 No 74 37.9 19 (45.2%) 21 (40.4%) 34 (33.7%)
Open in a new tab
*

Two children had parity equal to zero, and were high risk due to the parents having siblings with ASD. ASD=autism spectrum disorder, non-TD= non-typical development, TD= typical development

Participants 3-PBA urinary concentrations were detectable in 97% of samples with the median of average pregnancy concentrations 2.06 ng/mL (Table 2). We compared 3-PBA concentrations in MARBLES participants, whose pregnancies were in years 2007–2014, to concentrations from a sample of women of reproductive age from NHANES 2009 – 2010. When compared to the NHANES sample, the MARBLES participants had the highest concentrations and greatest proportion above the LOD. We found that pregnancies with male births had higher 3-PBA concentrations.

Table 2.

Distribution of 3-PBA concentrations (ng/mL) measured from urine samples collected during pregnancy from 201 subjects from the MARBLES cohort, 2007 – 2014, and comparison to NHANES pesticide metabolite concentrations (ng/mL) from adult women 2009 – 2010.

Selected Percentiles
N 25th 50th 75th 95th
MARBLES Study                
Pregnancy1                
    by sample ng/mL - SG 6382 0.84 1.46 2.73 10.53
    by participant ng/mL - SG 2013 1.25 2.06 3.50 9.96
2nd trimester                
    by sample ng/mL - SG 2702 0.81 1.45 2.54 10.53
    by participant ng/mL - SG 1343 1.00 1.81 3.04 9.05
3rd trimester                
    by sample ng/mL - SG 3682 0.85 1.46 2.86 10.29
    by participant ng/mL - SG 1903 1.13 1.93 3.55 13.22
NHANES
Adult women (2009 – 2010) ng/mL 1392 NR 0.4 1.06 6.5
Open in a new tab

CI= Confidence Interval, SG=specific gravity, 3-PBA=3 Phenoxybenzoic acid, NR=not reported,

1

Samples from the 2nd and 3rd trimesters

2

Total samples,

3

Average 3-PBA concentration among women included in the present study

Of the 201 mother-child pairs with pregnancy 3-PBA concentrations and the 3-year clinical diagnosis, seven had missing information on covariates, leaving a total of 194 participants with complete information on exposure, 3-year clinical outcome, and covariates. Results of the clinical assessment were 41 children with ASD, 52 in the non-TD group, and 101 classified as TD.

Adjusted RRRs for the pregnancy average 3-PBA concentrations and for trimester-specific averages are reported in Table 3, with non-adjusted model results in the Supplemental Material (Table S1). The reported relative risk ratios represent relative changes in the outcome for a 1-unit change in the corresponding natural log-transformed prenatal 3-PBA concentration, equivalent to an increase in concentration of approximately 2.7-fold. For natural-log average pregnancy 3-PBA concentrations we observed an elevated relative risk ratio for ASD compared to TD (RRR: 1.34 (95% CI 0.89 to 2.03)) and a lower relative risk ratio for non-TD compared to TD (RRR: 0.78 (95% CI 0.48 to 1.28)). The RRR’s for 2nd trimester 3-PBA concentrations in relation to ASD (RRR: 1.50 (95% CI 0.89 to 2.51) p=0.12) and non-TD development (RRR: 1.17 (95% CI 0.69 to 1.98) p=0.56) were greater than estimates for 3rd trimester exposures or the combined pregnancy exposures.

Table 3:

Adjusted multinomial logistic regression analysis assessing natural log-transformed prenatal 3-PBA concentrations overall and within the 2nd and 3rd trimester, in relation to children's 3-year diagnosis (n=194).

ASD1
 Non-Typical1
Timing N RRR 95% CI p-value RRR 95% CI p-value
Pregnancy average (2nd and 3rd trimesters) 194 1.34 (0.89 to 2.03) 0.16 0.78 (0.48 to 1.28) 0.33
2nd trimester 133 1.50 (0.89 to 2.51) 0.12 1.17 (0.69 to 1.98) 0.56
3rd trimester 183 1.31 (0.85 to 2.02) 0.23 0.77 (0.49 to 1.20) 0.25
Open in a new tab
1

Typically developing children served as the reference, models adjusted for pre-pregnancy BMI, birth year, birth year squared, homeownership, prenatal vitamin use, average TCPy, RRR=Relative Risk Ratio, CI=Confidence Interval, ASD=autism spectrum disorder

Note: The reported relative risk ratios pertain to a 1-unit change in the corresponding natural log-transformed prenatal 3-PBA concentration, equivalent to an increased concentration of approximately 2.7-fold.

4. Discussion

In the present study, greater in utero 3-PBA concentrations from 2nd trimester urine samples were associated with a modest increase (50%) in risk for ASD at 3-years, relative to their risk for TD. Previously published studies analyzing pyrethroids in association with ASD only considered exposures from agricultural pesticide use. In a record-linkage study of births in the California central valley from 1996–1998 and agricultural pesticide applications near the residence at birth, Roberts et al. found an association of ASD with prenatal exposures to bifenthrin, a pyrethroid (Roberts et al., 2007). Bifenthrin is the most stable pyrethroid, with an exceptionally long half-life outdoors, making it one of the most potent pyrethroid pesticides on the market (Cao et al., 2014). However, bifenthrin does not metabolize to 3-PBA and we did not assess a bifenthrin specific metabolite. In a similar study, von Ehrenstein et al. analyzed a wide array of pesticides in California births in the San Joaquin Valley from 1998–2010; nearby prenatal agricultural applications of permethrin, but not bifenthrin, were associated with a slight increase in ASD risk (OR=1.10, 95% CI=(1.01, 1.20)) (von Ehrenstein et al., 2019). Both of these studies relied on the database from California Pesticide Use Reports, (PUR) which undergoes extensive quality control (California Department of Pesticide Regulation, 2020), and have been validated by a high correlation with airborne concentrations at distances similar to the buffers used by these studies (Wofford et al 2014). Although these studies did not use biomarkers of pyrethroid exposure, the PUR data capture exposures over time, and in agricultural communities, which may represent the predominant pesticide exposures. Our study catchment area included some participants from agricultural communities, but also encompassed several metropolitan areas.

There are few published studies investigating prenatal pyrethroid exposure in relation to child cognitive development, and thus far the findings are inconsistent. In the ELEMENT study from Mexico City, Mexico, researchers reported a marginal association between prenatal 3-PBA concentrations and Bayley Scales of Infant Development (BSID) Mental Development Index (MDI) scores at 24 months but not at 36 months. As compared with the lowest exposure category, participants in the medium and high 3-PBA categories had reduced MDI scores at 24 months (Watkins et al., 2016).

A study from California’s Salinas Valley investigated residential proximity to agricultural pesticide use and child intelligence quotient (IQ), as evaluated on the WISC, at 7 years. In this study Gunier et al. observed lower IQ scores for children whose mothers were living within 1 km of agricultural pyrethroid applications during pregnancy. Children had significantly lower full scale IQ (β: −2 (95% CI −3.7 to −0.3)), perceptual reasoning (β: −2.1 (95% CI −4.0 to −0.2)), and verbal comprehension (β: −1.8 (95% CI −3.4 to −0.3)) scores (Gunier et al., 2017). Agricultural applications would be a major source of pesticide exposure in the Salinas Valley because participants lived in an agricultural community.

Pyrethroid pesticide exposures have been studied in relation to child behaviors and cognition in the PELAGIE study, a prospective cohort study from France. Prenatal pyrethroid exposures, as measured in a single pregnancy urine sample, were not significantly associated with child cognitive development at 6 years (Viel et al., 2015). Prenatal cis-3-(2,2-dichlorovinyl)-2,2-dimethylcyclopropant carboxylic acid (DCCA) concentrations were associated with internalizing behavioral difficulties at 6 years (Viel et al., 2017). In this French population trans-DCCA, which is a metabolite of permethrin and cypermethrin, was commonly detected. However, the majority of urine samples were below LOD for 3-PBA indicating this population of French mothers are exposed to different types of pyrethroids and possibly have different exposure routes compared to the US population. Detection limits for 3-PBA were lower in the PELAGIE cohort with LOD = 0.008 ng/mL compared to 0.1 ng/mL in the present study. Prenatal concentrations of 3-PBA were categorized due to the large number of samples below the LOD in the PELAGIE Study, and were not associated with childhood behaviors, despite higher proportions detected and significant associations between childhood 3-BPA and increased odds of behavioral disorders for the externalizing and reverse-scored prosocial behavior subscales. A variety of behavioral and executive functioning deficits were also found to be related to whether pyrethroid metabolites were detected prenatally in a study based in New York (Furlong et al., 2017).

Urinary pesticide concentrations from the MARBLES study population were higher than have previously been reported and the majority of samples in the MARBLES population were above LOD. This allowed us to investigate associations with a continuous term for 3-PBA exposures. Higher concentrations detected in MARBLES samples likely reflects growing use of pyrethroids since the 2001 EPA restriction of chlorpyrifos.

The use of pooled samples in the present study reduced exposure misclassification that can occur when participants are assigned to an exposure category based on urinary concentrations from a single spot sample (Barkoski et al., 2018). We reported an ICC = 0.16 for 3-PBA from multiple pregnancy urine samples indicating greater within- than between-person variability showing that a single urine sample from pregnancy may not capture long-term exposure for pyrethroids due to their quick metabolism. Participants in the present study could be exposed from residential and dietary sources with the latter contributing more variability in exposure. Since the restriction of chlorpyrifos, pyrethroids are commonly used in residential pesticide products making home sources a primary exposure route (Atwood, 2017; Lu et al., 2006b; U.S. EPA, 2013).

Nevertheless, there are limitations with urinary pesticide metabolites. Even though we collected multiple urine samples from each trimester (1 – 6 samples), we still may have failed to capture exposures due to the rapid metabolism (about 7 hours) of these compounds (Ratelle et al., 2015). The literature on prenatal pyrethroid exposures and child neurodevelopment is inconsistent, which in part could be due to exposure misclassification in previous studies, which used a single spot urine sample to assess exposure. Published studies reporting associations for prenatal pesticide exposure and ASD have used agricultural pesticide applications as a proxy for pesticide exposure. The present study used a biomarker for pesticide exposures, which captures all exposure routes. However due to the rapid metabolism of pesticides, a urinary biomarker may fail to capture exposure over long periods. Pyrethroid metabolite 3-PBA is a valid biomarker for 6 common pyrethroid pesticides in humans but it may also reflect exposures to the metabolite in the environment. A potentially relevant finding was the detection of 3-PBA in house dust samples collected with floor wipes from homes in northern California (Trunnelle et al., 2014), indicating that metabolites from home applications may linger. Overall, we expect that the resulting errors in our use of urinary metabolites at various intervals throughout pregnancy would tend to be non-differential, particularly since the collection of urine samples occurred before the child was born and years before the diagnoses were made. In this circumstance, the bias in results would on average be towards the null and could increase the uncertainty, reflected in the confidence intervals, but bias away from the null is also possible.

Given the overall small sample size, the number of children affected by ASD or non-TD outcomes was only a few dozen of each. Precision is somewhat low, as reflected in the width of the confidence intervals. A larger sample would increase statistical power to detect smaller effects that would still have public health relevance and tighter confidence intervals.

In comparison with prior studies of pyrethroids, especially those examining ASD, we considered and adjusted for a wider array of potential risk and protective factors previously associated with ASD or neurodevelopmental outcomes in other studies. Nevertheless, bias from unmeasured, unknown, and/or omitted confounders cannot be ruled out. Another potential limitation was the lack of 1st trimester or postnatal urinary 3-PBA concentrations. Shelton et al. observed associations between pyrethroid exposures during both the 1st and 3rd trimester and increased risk for ASD (Shelton et al., 2014b); von Ehrenstein found associations with pyrethroid exposures in the first year of life (von Ehrenstein et al., 2019). We had fewer participants with urine samples from the 1st trimester therefore the present study focused on exposures from the latter two trimesters of pregnancy. Unique developmental processes in the brain occur in the early first trimester, beginning with the demethylation of both the maternal and paternal DNA, and when structures such as the neural tube form, migrational pathways are established, and even preconceptional exposures may influence the central nervous system. Nevertheless, later gestational and postnatal exposures can also have long-lasting effects on brain development.

Finally, the MARBLES Study population differs from the general population because of the design: the familial/sibling recurrence in this study was 23%, and despite a shared environment with their older affected sibling, many of the MARBLES children likely have an increased genetic susceptibility for ASD. Our results therefore may not be generalizable to populations with lower risk. In a population with a high genetic load, it is difficult to predict whether environmental contributions to ASD will be amplified (e.g., as a result of gene-by-environment interactions) or may be secondary to, or dwarfed by, strong genetic susceptibility that would confer higher background risk for ASD.

Major strengths of this study include exposure measurements from multiple biological specimens during pregnancy, blinding of laboratory personnel to the child outcomes, the high quality clinical data stemming from administration of child developmental assessments by highly trained clinicians with research reliability, and collection and adjustment for a number of potential risk/protective covariates, including sociodemographic information, that reduced confounding and potential selection bias. Additionally, the majority of urine samples were above the LOD. This allowed us to assess linear relationships between continuous measures for the exposure and outcomes. Maternal urine samples were pooled within a trimester to reduce analytical costs and to improve exposure assessment. We have previously reported that pooled samples can reduce exposure misclassification in this population (Barkoski et al., 2018).

This may be the first study to assess a biomarker of pyrethroid exposure, based on up to 6 urine samples from both the 2nd and 3rd trimesters during pregnancy, in relation to clinical ASD diagnosis at 3-years. Larger studies in general population cohorts with repeated urinary measures starting earlier in pregnancy and that consider gene-environment interactions should be conducted to clarify the potential for adverse neurodevelopmental effects of pyrethroid pesticides.

5. Conclusions

Exposures to pyrethroid pesticides in utero were evaluated as 3-PBA concentrations from multiple samples collected during the 2nd and 3rd trimesters in a cohort at high risk for developing ASD, the MARBLES Study, due to a sibling with ASD. During the 2nd trimester, a higher concentration of urinary 3-PBA was associated with a modest increased risk for ASD at 3-years, but these results are not definitive, in part due to the relatively small number of cases. Our findings do indicate exposures to pyrethroid pesticides are common among pregnant women in northern California, Given the growing use of these pesticides, determining whether pregnancy exposures pose a risk to the developing brain, and if so, at what levels, is increasingly critical. In the interests of healthy development for future children, studies of pesticides with accurate exposure ascertainment and larger sample sizes are needed.

Supplementary Material

1

  • First study of pregnancy pyrethroid pesticide biomarker and risk of ASD at 3-years.

  • Higher concentration of urinary 3-PBA associated with modest increased risk for ASD.

  • Pesticide exposure characterization difficult due to temporal variability and multiple sources.

Acknowledgements

We would like to thank the MARBLES Study participants for making this research possible.

Funding

The project described was supported by NIEHS [grant numbers R01ES020392, R24ES028533, P30ES023513, P01ES11269, and P30ES019776], U.S EPA STAR [grant number 83543201], NICHD [grant number U54HD079125], the Autism Science Foundation, and the University of California, Davis MIND Institute.

The University of California, Davis institutional review board approved the MARBLES Study and informed consent was obtained from each participant.

Footnotes

Declaration of interests

None

X The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

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.

References

  1. ASTDR, 2003. Toxicological Profile for Pyrethrins and Pyrethroids. Agency for Toxic Substances and Disease Registry. [PubMed] [Google Scholar]
  2. Atwood D, Paisley-Jones C,, Pesticides Industry Sales and Usage 2008 – 2012. US Environmental Protection Agency, 2017. [Google Scholar]
  3. Aziz MH, et al. , 2001. Neurodevelopmental consequences of gestational exposure (GD14-GD20) to low dose deltamethrin in rats. Neurosci Lett. 300, 161–5. [DOI] [PubMed] [Google Scholar]
  4. Bai D, et al. , 2019. Association of genetic and environmental factors with autism in a 5-country cohort. JAMA Psychiatry. 76, 1035–1043. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bailey A, et al. , 1995. Autism as a strongly genetic disorder: evidence from a British twin study. Psychol Med 25, 63–77. [DOI] [PubMed] [Google Scholar]
  6. Barkoski J, et al. , 2018. Variability of urinary pesticide metabolite concentrations during pregnancy in the MARBLES Study. Environ Res 165, 400–409. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Barr DB, et al. , 2010. Urinary concentrations of metabolites of pyrethroid insecticides in the general U.S. population: National Health and Nutrition Examination Survey 1999–2002. Environ Health Perspect 118, 742–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cao Z, et al. , 2014. Nanomolar bifenthrin alters synchronous Ca2+ oscillations and cortical neuron development independent of sodium channel activity. Mol Pharmacol 85, 630–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chawarska K, et al. , 2014. 18-month predictors of later outcomes in younger siblings of children with autism spectrum disorder: a baby siblings research consortium study. J Am Acad Child Adolesc Psychiatry. 53, 1317–1327 e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Chih C-P, 2011. The design effects of cluster sampling on the estimation of mean lengths and total mortality of reef fish. Fisheries Research. 109, 295–302. [Google Scholar]
  11. Coker E, et al. , 2017. Association between Pesticide Profiles Used on Agricultural Fields near Maternal Residences during Pregnancy and IQ at Age 7 Years. Int J Environ Res Public Health. 14, 506. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Constantino JN, et al. , 2013. Autism recurrence in half siblings: strong support for genetic mechanisms of transmission in ASD. Mol Psychiatry. 18, 137–8. [DOI] [PubMed] [Google Scholar]
  13. EPA US, Revaluation: Pyrethrins and Pyrethroids. Pesticides: Regulating Pesticides, Vol. 2013, 2010. [Google Scholar]
  14. Eskenazi B, et al. , 2007. Organophosphate pesticide exposure and neurodevelopment in young Mexican-American children. Environ Health Perspect 115, 792–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Folstein S, Rutter M, 1977. Infantile autism: a genetic study of 21 twin pairs. J Child Psychol Psychiatry. 18, 297–321. [DOI] [PubMed] [Google Scholar]
  16. Furlong MA, et al. , 2017. Prenatal exposure to pyrethroid pesticides and childhood behavior and executive functioning. Neurotoxicology. 62, 231–238. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Gaugler T, et al. , 2014. Most genetic risk for autism resides with common variation. Nat Genet 46, 881–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Gunier RB, et al. , 2017. Prenatal Residential Proximity to Agricultural Pesticide Use and IQ in 7-Year-Old Children. Environ Health Perspect 125, 057002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Guo B-Q, et al. , 2019. Maternal multivitamin supplementation is associated with a reduced risk of autism spectrum disorder in children: a systematic review and meta-analysis. Nutrition Research. 65, 4–16. [DOI] [PubMed] [Google Scholar]
  20. Hallmayer J, et al. , 2011. Genetic heritability and shared environmental factors among twin pairs with autism. Arch Gen Psychiatry. 68, 1095–1102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hauser R, et al. , 2004. Temporal variability of urinary phthalate metabolite levels in men of reproductive age. Environ Health Perspect 112, 1734–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hertz-Picciotto I, et al. , 2018. A prospective study of environmental exposures and early biomarkers in autism spectrum disorder: Design, protocols and preliminary data from the MARBLES study. Environmental Health Perspectives. 126, 117004–117004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hertz-Picciotto I, Sass JB, Engel S, Bennett DH, Bradman A, Eskenazi B, Lanphear B, Whyatt R. Organophosphate exposures during pregnancy and child neurodevelopment: Recommendations for essential policy reforms. PLos Medicine October 24, 2018. 10.1371/journal.pmed.1002671 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Horton MK, et al. , 2011. Impact of prenatal exposure to piperonyl butoxide and permethrin on 36-month neurodevelopment. Pediatrics. 127, e699–706. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Krakowiak P, et al. , 2012. Maternal metabolic conditions and risk for autism and other neurodevelopmental disorders. Pediatrics. 129, e1121–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Levine SZ, et al. , 2018. Association of maternal use of folic acid and multivitamin supplements in the periods before and during pregnancy with the risk of autism spectrum disorder in offspring. JAMA Psychiatry. 75, 176–184. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Li M, et al. , 2016. The association of maternal obesity and diabetes with autism and other developmental disabilities. Pediatrics. 137, e20152206. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Lord C, et al. , 2008. ADOS: Autism diagnostic observation schedule. Hogrefe; Boston. [Google Scholar]
  29. Lu C, et al. , 2006a. A longitudinal approach to assessing urban and suburban children’s exposure to pyrethroid pesticides. Environ Health Perspect 114, 1419–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Lu C, et al. , 2006b. Organic diets significantly lower children’s dietary exposure to organophosphorus pesticides. Environ Health Perspect 114, 260–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Maenner MJ, Shaw KA, Baio J, EdS, Washington A, Patrick M, et al. 2020. Prevalence of autism spectrum disorder among children aged 8 years - autism and developmental disabilities monitoring network, 11 sites, united states, 2016. MMWR Surveill Summ 69:1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Malaviya M, et al. , 1993. Perinatal effects of two pyrethroid insecticides on brain neurotransmitter function in the neonatal rat. Vet Hum Toxicol 35, 119–22. [PubMed] [Google Scholar]
  33. Mullen EM, 1995. Mullen Scales of Early Learning. Circle Pines, Minnesota: American Guidance Serviecs, Inc. [Google Scholar]
  34. Olsson AO, et al. , 2004. A liquid chromatography--tandem mass spectrometry multiresidue method for quantification of specific metabolites of organophosphorus pesticides, synthetic pyrethroids, selected herbicides, and deet in human urine. Anal Chem 76, 2453–61. [DOI] [PubMed] [Google Scholar]
  35. Ozonoff S, et al. , 2014. The broader autism phenotype in infancy: when does it emerge? Journal of the American Academy of Child and Adolescent Psychiatry. 53, 398–407 e2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Philippat C, et al. , 2018. Prenatal exposure to organophosphate pesticides and risk of autism spectrum disorders and other non-typical development at 3 years in a high-risk cohort. Int J Hyg Environ Health. 221, 548–555. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Ratelle M, et al. , 2015. Toxicokinetics of permethrin biomarkers of exposure in orally exposed volunteers. Toxicol Lett 232, 369–75. [DOI] [PubMed] [Google Scholar]
  38. Rauh VA, et al. , 2006. Impact of prenatal chlorpyrifos exposure on neurodevelopment in the first 3 years of life among inner-city children. Pediatrics. 118, e1845–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Roberts EM, et al. , 2007. Maternal residence near agricultural pesticide applications and autism spectrum disorders among children in the California Central Valley. Environ Health Perspect 115, 1482–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Sandin S, et al. , 2014. The familial risk of autism. JAMA. 311, 1770–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. SAS II, 2002–2012. SAS Software Version 9.4. Cary, NC. [Google Scholar]
  42. Schmidt RJ, et al. , 2011. Prenatal Vitamins, One-carbon Metabolism Gene Variants, and Risk for Autism. Epidemiology. 22, 476–485. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Schmidt RJ, et al. , 2019. Association of Maternal Prenatal Vitamin Use With Risk for Autism Spectrum Disorder Recurrence in Young Siblings. JAMA Psychiatry. 76, 391–398. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Schmidt RJ, et al. , 2012. Maternal periconceptional folic acid intake and risk of autism spectrum disorders and developmental delay in the CHARGE (CHildhood Autism Risks from Genetics and Environment) case-control study. Am J Clin Nutr 96, 80–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Shafer TJ, Hughes MF, 2010. Accumulation of pyrethroid compounds in primary cultures from rat cortex. Toxicol In Vitro. 24, 2053–7. [DOI] [PubMed] [Google Scholar]
  46. Shafer TJ, et al. , 2005. Developmental neurotoxicity of pyrethroid insecticides: critical review and future research needs. Environ Health Perspect 113, 123–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Shelton JF, et al. , 2014a. Neurodevelopmental disorders and prenatal residential proximity to agricultural pesticides: The charge study. Environ Health Perspect 122, 1103–1109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Shelton JF, et al. , 2014b. Neurodevelopmental Disorders and Prenatal Residential Proximity to Agricultural Pesticides: The CHARGE Study. Environ Health Perspect [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Shin HM, et al. , 2018. Prenatal exposure to phthalates and autism spectrum disorder in the MARBLES study. Environ Health. 17, 85. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Simon JM, et al. , 2019. High-throughput screening and classification of chemicals and their effects on neuronal gene expression using RASL-seq. Sci Rep 9, 4529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Soderlund DM, 2012. Molecular mechanisms of pyrethroid insecticide neurotoxicity: recent advances. Arch Toxicol 86, 165–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Starr JM, et al. , 2014. A test house study of pesticides and pesticide degradation products following an indoor application. Indoor Air. 24, 390–402. [DOI] [PubMed] [Google Scholar]
  53. Surén P, et al. , 2013. Association between maternal use of folic acid supplements and risk of autism spectrum disorders in children. JAMA 309, 507–577. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Trunnelle KJ, et al. , 2014. Urinary pyrethroid and chlorpyrifos metabolite concentrations in northern California families and their relationship to indoor residential insecticide levels, part of SUPERB. Environ Sci Technol [DOI] [PubMed] [Google Scholar]
  55. U.S. EPA, E. P. A., Pesticides: Regulating Pesticides. Vol. 2015, 2013. [Google Scholar]
  56. Viel JF, et al. , 2017. Behavioural disorders in 6-year-old children and pyrethroid insecticide exposure: the PELAGIE mother-child cohort. Occup Environ Med 74, 275–281. [DOI] [PubMed] [Google Scholar]
  57. Viel JF, et al. , 2015. Pyrethroid insecticide exposure and cognitive developmental disabilities in children: The PELAGIE mother-child cohort. Environ Int 82, 69–75. [DOI] [PubMed] [Google Scholar]
  58. von Ehrenstein OS, et al. , 2019. Prenatal and infant exposure to ambient pesticides and autism spectrum disorder in children: population based case-control study. BMJ. 364, l962. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Watkins DJ, et al. , 2016. Urinary 3-phenoxybenzoic acid (3-PBA) levels among pregnant women in Mexico City: Distribution and relationships with child neurodevelopment. Environ Res 147, 307–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Yang C, et al. , 2017. The association between air pollutants and autism spectrum disorders. Environ Sci Pollut Res Int 24, 15949–15958. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1658966-supplement-1.docx (121.2KB, docx)

RESOURCES