Abstract
Background:
Prenatal exposure to phthalates, organophosphate esters, and organophosphorous pesticides have been associated with neurodevelopmental deficits including language ability, however, few studies consider the effect of exposure mixtures and the potential longitudinal detriments over time.
Objective:
This study examines the influence of prenatal exposure to phthalates, organophosphate esters, and organophosphorous pesticides, on children’s language ability from toddlerhood to the preschool period.
Methods:
This study includes 299 mother-child dyads from Norway in the Norwegian Mother, Father and Child Cohort Study (MoBa). Prenatal exposure to chemicals were assessed at 17 weeks’ gestation, and child language skills were assessed at 18 months using the Ages and Stages Questionnaire communication subscale and at preschool age using the Child Development Inventory. We ran two structural equation models to examine the simultaneous influences of chemical exposures on parent-reported and teacher-reported child language ability.
Results:
Prenatal organophosphorous pesticides were negatively associated with preschool language ability through language ability at 18 months. Additionally, there was a negative association between low molecular weight phthalates and teacher-reported preschool language ability. There was no effect of prenatal organophosphate esters on child language ability at either 18 months or preschool age.
Conclusions:
This study adds to the literature on prenatal exposure to chemicals and neurodevelopment and highlights the importance of developmental pathways in early childhood.
Keywords: Organophosphorous pesticides, Phthalates, Organophosphate esters, Prenatal exposure mixtures, Language ability, MoBa
1. Introduction
As children’s language capacity increases (including the use of gestures, word comprehension, word production) throughout early childhood, so also does the child’s ability to interact with the outside world, playing an important role in social interactions, and acting as an early indicator of cognitive development[1–5]. The development of language is complex, and is influenced by genetics (e.g. [6]), socialization (e.g. [7]), and prenatal factors (e.g. [8]). The prenatal period is a particularly critical time for the development of the brain, due to the potential for prenatal experiences (i.e., chemicals) to cross the placenta and alter the “programming” of the fetus [9, 10]. While there is a large body of work examining the influence of prenatal chemicals on child neurodevelopment, there is less research focusing specifically on early language ability with considerable inconsistency in findings [8]. Moreover, the existing literature has tended to examine language development at a single point in time (e.g. [11–13]), without the consideration for language development trajectories, from gestures and early word production to expressive language and comprehension. Therefore, these studies are missing the potential developmental pathways or unique neurological mechanisms by which prenatal exposures to chemicals might influence children’s language development at different ages [14]. Therefore, the goal of this study is to examine the influence of prenatal exposure to multiple chemicals on child language ability from 18 months to preschool age (3–4 years old).
Humans are exposed to multiple chemicals simultaneously, and some non-persistent chemical exposures are higher among reproductive aged women compared to men in the same age range, likely due to increased use of personal care products and cosmetics among women [15, 16]. Pregnant women can also pass these chemicals to their developing fetus through placental transport mechanisms and their ability to cross the blood-brain barrier [10], rendering the prenatal period a critically important window to examine, particularly given the developing brain is sensitive to chemical impacts [16].
We focus on three chemical families with potential to impact language development: phthalates, organophosphorous pesticides (OPs), and organophosphate esters (OPEs) [8, 12, 17–20]. Multiple additional studies suggest deficits on a range of complementary neurodevelopmental endpoints [21–23]. Phthalates are rapidly metabolized compounds (half-lives about 12–24 hours) that are found in a range of consumer products, including shampoos and fragrances, but also vinyl flooring and housing materials [24] and have been shown to have endocrine altering effects and adversely influence hippocampal development [25]. OPs are insecticides that are used extensively in agriculture and exposure generally occurs through the consumption of conventionally grown fruits and vegetables. OP toxicity is due to irreversible inhibition of the enzyme acetylcholinesterase, which causes a build-up of acetylcholine in synapses [26]; however, adverse associations with neurodevelopmental outcomes have been observed at exposure concentrations too low to result in appreciable acetylcholinesterase inhibition, suggesting that alternate pathways to neurodevelopmental impairments may exist [27]. OPEs are a class of chemicals used as flame retardants and plasticizers that are found in everyday items like household furniture, personal care products, and nail polish, and can act as both a neurotoxicant [28] and an endocrine disrupter [29].
The existing literature examining phthalates, OPs, and OPEs in relation to early child language ability is sparse, inconsistent, and fails to account for real-world exposure mixtures and developmental trajectories. The lack of consistency in findings might be partially due to the lack of consideration of mixtures [30]. We address these limitations in a unique, nested study of the Norwegian Mother, Father and Child Cohort Study (MoBa), in which prenatal maternal urinary concentrations of phthalates, OPs, and OPEs were measured, and language development was assessed at 18 months and again in the preschool period. We characterize their influence on language development, while accounting for their intercorrelations and potentially age-dependent associations using structural equations models (SEM).
2. Materials and methods
2.1. Study Population
The Norwegian Mother, Father and Child Cohort Study (MoBa) is a population-based pregnancy cohort study conducted by the Norwegian Institute of Public Health [31]. Participants were recruited during their first ultrasound appointment at approximately 17 weeks’ gestation from all over Norway from 1999–2008 [32]. Approximately 41% of all pregnancies in the eligible period participated in MoBa. The cohort includes approximately 114,500 children, 95,200 mothers and 75,200 fathers. The current study is based on version 9 of the quality-assured data files released for research in preschool ADHD. This study was approved by the Regional Committee for Medical and Health Research Ethics in Norway, and the UNC IRB.
Our study population draws from nested sub-studies of MoBa that focused on ADHD [33–35], one that leveraged an onsite assessment for Preschool ADHD that incorporated a comprehensive set of neuropsychological assessments (33), and another that linked MoBa with the National Patient Registry (NPR) to identify children who had been clinically diagnosed with ADHD (34). Eligibility criteria for these studies included having a maternal urine sample collected during pregnancy and mothers being pregnant in 2003 or later (additional eligibility included elsewhere [33–35]).
For the current study, our analysis leveraged these already assembled subgroups (Supplemental Figure 1): (1) the Preschool ADHD sample: This population includes dyads for whom the parent reported the child had ADHD-like symptoms at the 36-month MoBa questionnaire, and the child was subsequently found to have clinically significant or subclinical symptoms of Preschool ADHD during an on-site assessment (n = 258); and children that were randomly selected to participate in the Preschool ADHD substudy who were free of clinical symptoms of ADHD during the onsite assessment (n = 79), for a total of 337 mother-child dyads. In total, 299 of these dyads had non-missing data across all required components for the longitudinal analysis. (2) NPR Registry ADHD cases: This population includes dyads in which the child was found to have a clinical diagnosis for ADHD registered. (3) MoBa Random Sample: Dyads that were randomly sampled from the eligible population irrespective of ADHD-like symptoms (n = 532). Together these groups yielded 1155 mother-child dyads, of whom 956 had complete exposure, covariate, and 18-month language data. Complete selection criteria for the underlying studies of ADHD and executive functions have been described previously [33–35].
2.2. Exposure biomarkers
Urine was collected from mothers at approximately 17 weeks’ gestation. These samples were shipped overnight at ambient temperature to the central MoBa biorepository in Oslo, Norway where they were aliquoted into 1mL cryotubes and frozen at −80°C. Prior quality control studies have been conducted to assess the impact of shipping temperature on phthalate and organophosphate pesticide biomarker measures, finding little impact [36]. In the current study, all analyses of environmental toxicant biomarkers were performed at the Norwegian Institute of Public Health.
Phthalate exposure biomarkers.
Maternal urine concentrations of 12 phthalate metabolites: monoethyl phthalate (MEP), a metabolite of diethyl phthalate, mono-iso-butyl phthalate (MiBP), a metabolite of di-iso-butyl phthalate; mono-n-butyl phthalate (MnBP), a metabolite of di-n-butyl phthalate; monobenzyl phthalate (MBzP), a metabolite of butyl benzyl phthalate (BBzP); mono-2-ethylhexyl phthalate (MEHP), mono- 2-ethyl-5-hydroxyhexyl phthalate (MEHHP), mono-2-ethyl-5-oxoy- hexyl phthalate (MEOHP), mono-2-ethyl-5-carboxypentyl phthalate (MECPP), and mono-2-methylcarboxyhexyl phthalate (MMCHP), metabolites of DEHP, and mono-4-methyl-7-hydroxyoctyl phthalate (OH-MiNP), mono-4-methyl-7-oxooctyl phthalate (oxo-MiNP), and mono-4-methyl-7-carboxyheptyl phthalate (cx-MiNP), metabolites of di-iso-nonyl phthalate (DiNP). Molar sums for DEHP and DiNP were computed using the 5 DEHP metabolites, and 4 DiNP metabolites respectively. On-line column switching liquid chromatography coupled with tandem mass spectrometry was used. All assays were conducted in batches with randomized cases and controls. The analytic approach and quality control procedures have been previously reported in more detail [34]. Each phthalate concentration was adjusted for batch and specific gravity and then natural log-transformed to be consistent with previous studies.
Organophosphorous pesticide exposure biomarkers.
Maternal urinary concentrations of six non-specific OP metabolites, dialkyl phosphates (DAPs) included three dimethyl phosphates [dimethyl phosphate (DMP), dimethyl thiophosphate (DMTP), and dimethyl dithiophosphate (DMDTP)] and three diethyl phosphates [diethyl phosphate (DEP), diethyl thiophosphate (DETP), and diethyl dithiophosphate (DEDTP)] were measured using ultra-performance liquid chromatography (UPLC) coupled with quadrupole-time-of-flight (QTOF) at the Norwegian Institute of Public Health [37]. Specific gravity-adjusted concentrations of dimethyl- and diethyl metabolites were summed by molar weight to calculate total dimethyl- (ΣDMP) and total diethyl phosphate (ΣDEP) concentrations and were subsequently log (natural) transformed. DEDTP was not included in ΣDEP because nearly 99% of values were below the Limit of Detection (LOD).
Organophosphate ester exposure biomarkers.
Four OPE metabolites, i.e., DPHP, di-n-butyl phosphate (DNBP), bis(2-butoxyethyl) hydrogen phosphate (BBOEP), and bis(1,3-dichloro-2-propyl) phosphate (BDCIPP), were measured in the urine samples using UPLC-QTOF by a modified method [38]. The modification was done in the sample preparation procedure and was adapted from an earlier published method by Cequier at al. (2016). The details of the modification are described elsewhere [39]. All assays were conducted in randomized batches, each of which included laboratory in-house as well as laboratory-blinded pooled urine quality control samples. BBOEP and BDCIPP were dichotomized at the LOD due to the lower detection frequency for these metabolites.
2.3. Measures of child language ability.
On the 18-month MoBa questionnaire, parents reported on the child’s communication ability using a shortened version of the Ages and Stages Questionnaire communication subscale [40]. Parents completed three items on a 3-point Likert scale (1 = Yes, often; 2 = Sometimes; 3 = Not yet) and the scale was reliable (σ = 0.60). Higher scores indicate better early communication ability.
Parents and teachers of children who returned for the on-site preschool ADHD assessment at approximately 3.5 years of age reported on child expressive language ability using the Child Development Inventory, expressive language subscale [41]. The CDI is a broad questionnaire that measures children’s health, development, and adjustment from ages 15 months to 6 years old. The expressive language subscale includes 50 items that assess expressive communication including simple gestures and complex language expressions and parents respond by marking yes or no to indicate if the statement describes the child’s behavior. The questionnaire has good sensitivity and specificity (>80%) of identifying delayed development in children. Higher scores indicate better language ability.
2.4. Covariates.
Covariate data were obtained through maternal self-report on MoBa questionnaires collected during pregnancy and in the postpartum period, as well as from the Medical Birth Registry of Norway (MRBN), which is a national health registry containing information about all births in Norway. The MBRN was used to collect maternal age, parity, birth year, and child sex at birth. Maternal self-report was used to obtain information on marital status, maternal education, maternal depressive symptoms, prenatal smoking and alcohol use, and maternal fish consumption during pregnancy.
2.5. Statistical Analysis
First, we conducted principal component analyses (PCA) for each chemical family (phthalates, pesticides, and organophosphate esters (OPEs) using varimax rotation and the factors were used in the subsequent analyses. Covariates were selected based on previous literature and then examined using a DAG (directed acyclic graph [42]) to determine the covariates to include in final models (supplemental Figure 2). Covariates were regressed out of the study variables (except for child sex), and the residuals were used in the main analyses. For the main analyses using the preschool sample (n=337), we used structural equation modeling (SEM) using Mplus 8 [43] to examine the associations of phthalates, OPs, and OPEs simultaneously, on child communication at 18 months and expressive language at preschool age. PCA coupled with SEM is one way to address chemical mixtures within a model. Specifically, PCA is a dimension reduction technique that can group the chemicals based on commonalities and remove concerns of multicollinearity. Within these SEMs, we modeled the principal components of each chemical family (phthalates, OPs, and OPEs) along with children’s language ability at 18 months and preschool age simultaneously. Directional pathways were included from the chemical family principal components to children’s language ability at 18 months and preschool age as well as a direct path from child language ability at 18 months to preschool age. Along with the modeled direct effects, we specified indirect effects using the MODEL INDIRECT command to test indirect effects between each chemical principal component and child language ability at preschool age through child language ability at 18 months. MPLUS calculates the indirect effect by multiplying the direct effects. We also included correlated pathways between the chemical families to address the potential for correlated exposures. In the case of a significant pathway between a chemical family and child language ability, estimates are interpreted as the effect of that chemical family on child language ability, while accounting for all other chemicals. This modeling technique will provide estimates about the joint effects of a chemical family on an outcome. Child expressive language was only assessed in the Preschool ADHD substudy and there was missing data on the outcomes at 18 months and preschool age; therefore, the analytic sample with child expressive language was smaller (n = 299). We ran two separate models, one that used parent-report of expressive language and another that used teacher-report of expressive language, to examine the potential for single reporter bias. Models of comparative fit were examined using the Chi-square goodness of fit index (p > 0.05), the comparative fit index (CFI) (≥ 0.95), SRMR (< 0.07), and the root mean-square error of approximation (RMSEA) (< 0.06). To examine potential indirect effects between the chemicals and expressive language through child communication at 18 months, we used bootstrapped estimates to examine the indirect pathway.
We ran additional sensitivity analyses of the association between prenatal exposure to chemicals and child communication at 18 months in the full sample (all children with 18-month language data) and restricted to the random sample of MoBa. The analysis restricted to the random sample was performed to examine whether associations in our primary analyses were driven by oversampling of children with preschool or childhood ADHD. Additionally, we used multigroup analyses to determine whether associations between toxicants and child communication differed by child sex. Lastly, we examined associations between the prenatal exposure to chemicals and expressive language at preschool age within the ADHD substudy independent of the 18-month assessment.
3. Results
Descriptive statistics of the sample are presented in Table 1 and distributions of the chemical metabolites in Table 2. Mothers of the Preschool sample (Table 1) were on average 30 years old (M = 30.18, SD = 4.12) and were mostly college educated or more (64.9%). Children (55.2% male) were born between 2004 and 2007 and were approximately 3 years old (M = 41.60 months, SD = 1.27, Range = 38.31– 44.81). The only difference in the Preschool sample compared to the full sample is that mothers in the Preschool sample were more likely be nulliparous (χ2 (1) = 9.61, p = 0.002).
Table 1.
Demographics of the sample.
| Preschool Sample N = 337 |
Random Sample N = 532 |
Full Sample N = 1155 |
||
|---|---|---|---|---|
| N (%) or Mean (SD) [Range] |
N (%) or Mean (SD) [Range] |
N (%) or Mean (SD) [Range] |
||
|
| ||||
| Maternal age at birth | 30.18 (4.12) [19–42] | 30.94 (4.20) [19–43] | 30.29 (4.45) [17–45] | |
| Child sex | Boy | 186 (55.2) | 266 (50.0) | 658 (57.0) |
| Girl | 151 (44.8) | 266 (50.0) | 496 (43.0) | |
| Birth year | 2003 | -- | -- | 32 (2.8) |
| 2004 | 46 (13.6) | 54 (10.2) | 195 (16.9) | |
| 2005 | 91 (27.0) | 125 (23.5) | 302 (26.1) | |
| 2006 | 114 (33.8) | 184 (34.6) | 340 (29.4) | |
| 2007 | 86 (25.5) | 169 (31.8) | 278 (24.1) | |
| 2008 | -- | -- | 8 (0.7) | |
| Prenatal smoking | No | 257 (76.3) | 452 (85.9) | 877 (78.3) |
| Yes | 78 (23.1) | 74 (14.1) | 243 (21.7) | |
| Missing | 2 | 6 | 35 | |
| Prenatal drinking | No | 272 (80.7) | 425 (86.7) | 903 (87.5) |
| Yes | 39 (11.6) | 65 (13.3) | 129 (12.5) | |
| Missing | 26 | 42 | 123 | |
| Maternal education | < College Completed | 111 (32.9) | 116 (22.1) | 381 (34.1) |
| College Completed | 141 (41.8) | 229 (43.6) | 440 (39.4) | |
| > College Completed | 78 (23.1) | 165 (31.4) | 267 (23.9) | |
| Other | 6 (1.8) | 15 (2.9) | 30 (2.7) | |
| Missing | 1 | 7 | 37 | |
| Parity | Nulliparous | 197 (58.5) | 261 (49.2) | 593 (51.5) |
| Multiparous | 139 (41.2) | 270 (50.8) | 558 (48.5) | |
| Missing | 1 | 2 | 4 | |
| ADHD group | Childhood (NPR) | NA | 2 (0.4) | 286 (24.7) |
| Preschool | 258 (76.8) | 7 (1.3) | 258 (22.4)a | |
| Chemical Metabolites b | ||||
| Phthalates | Log (μg/L) | Log (μg/L) | Log (μg/L) | |
| MEP | 4.74 (1.46) | 4.59 (1.47) | 4.72 (1.48) | |
| [1.37–8.93] | [0.84–8.82] | [0.84–9.26] | ||
| MiBP | 2.97 (.77) | 2.91 (.91) | 2.96 (.86) | |
| [1.01–5.19] | [0.52–6.33] | [0.52–6.33] | ||
| MnBP | 3.00 (.78) | 2.91 (.86) | 3.02 (.82) | |
| [1.10–5.94] | [0.69–11.16] | [0.69–11.16] | ||
| MBzP | 1.69 (.90) | 1.54 (.91) | 1.69 (.95) | |
| [−0.79–4.74] | [−0.57–4.63] | [−0.79–5.02] | ||
| Log (μmol/L) | Log (μmol/L) | Log (μmol/L) | ||
| DEHP | −1.22 (.73) | −1.31 (.69) | −1.26 (.68) | |
| [−2.47–3.11] | [−2.70–2.70] | [−2.70–3.11] | ||
| DiNP | −3.94 (.68) | −3.93 (.63) | −3.97 (.61) | |
| [−4.99– −0.04] | [−5.21–0.07] | [−5.21–0.07] | ||
| OPs | Log (nmol/L) | Log (nmol/L) | Log (nmol/L) | |
| ΣDMP | 4.04 (1.10) | 4.28 (1.20) | 4.11 (1.19) | |
| [1.15–7.22] | [−3.05–7.59] | [−3.05–7.59] | ||
| ΣDEP | 2.96 (.88) | 3.04 (.95) | 2.94 (.91) | |
| [0.71–5.49] | [−3.77–6.37] | [−3.77–6.37] | ||
| OPEs | Log (μg/L) | Log (μg/L) | Log (μg/L) | |
| DPHP | −0.63 (1.06) | −0.66 (1.04) | −0.58 (1.05) | |
| [−3.70–3.71] | [−6.23–2.31] | [−6.23–3.71] | ||
| DNBP | −1.26 (.76) | −1.29 (.84) | −1.30 (.79) | |
| [−3.09–3.34] | [−7.56–1.79] | [−7.56–3.34] | ||
| BBOEPc | 157 (46.6%) | 276 (51.9%) | 612 (53.0%) | |
| BDCIPP | 82 (24.3%) | 109 (20.5%) | 264 (22.9%) | |
Note:
10 children overlap in childhood and Preschool ADHD categories and are listed in the Preschool ADHD category.
Chemical metabolites are specific gravity corrected and natural log transformed.
BBOEP and BDCIPP are presented N and % greater than LOD. NPR=Norwegian Patient Registry.
Table 2.
Correlations of the study variables for the preschool subsample.
| LMWP | HMWP | OPs | OPEs | Comm | EL-P | EL-T | |
|---|---|---|---|---|---|---|---|
| LMWP | |||||||
| HMWP | −0.001 | ||||||
| OPs | 0.08 | 0.08 | |||||
| OPEs | 0.26** | 0.14* | 0.18** | ||||
| Communication 18 mos | 0.031 | 0.03 | −0.09 | −0.01 | |||
| Expressive Language-P Preschool age | −0.08 | 0.06 | −0.01 | −0.02 | 0.39** | ||
| Expressive Language-T Preschool age | −0.18** | 0.11 | −0.05 | −0.07 | 0.28** | 0.61** | |
| N | 336 | 336 | 336 | 336 | 312 | 334 | 318 |
| Mean | 0.002 | 0.06 | −0.02 | 0.008 | 22.64 | 45.68 | 48.08 |
| SD | 0.95 | 1.14 | 0.93 | 0.96 | 8.77 | 6.12 | 7.86 |
Note:
p < .05
p < .01.
LMWP = low molecular weight phthalates, HWMP= high molecular weight phthalates, OPs = organophosphorous pesticides, OPEs=organophosphate esters, Comm = Communication, EL-P=expressive language-parent report, EL-T=expressive language-teacher report.
The principal components analysis was conducted using SPSS version 27 for each chemical family (phthalates, OPs, and OPEs) for data reduction purposes to model within-family chemical associations. Based on eigenvalues over 1.00, we identified 2 underlying factors for phthalates, 1 for OPs, and 1 for OPEs (Supplemental Table 1). The first phthalate factor had high loadings of MiBP, MnBP, and MBzP (0.79–0.86), and will henceforth be referred to as Low Molecular Weight Phthalates (LMWP). The second phthalate factor had high loadings of DEHP and DiNP (0.71 and 0.85), and will henceforth be referred to as High Molecular Weight Phthalates (HMWP). The OP factor had high loadings of both DMP and DEP (0.87 and 0.87). OPEs had high loadings of DPHP and DNBP (0.70 and 0.53). Correlations among the principal components and individual metabolites are presented in Figure 1. Metabolites within chemical families exhibited higher correlations than across families. Descriptive statistics and correlations for all study variables of the main analytic sample (the preschool sample) are presented in Table 2.
Figure 1.
Heatmap of correlations among all prenatal chemical exposures and composites for the full sample. Correlations are among natural log, controlled for specific gravity and batch. BBOEP and BDCIPP are dichotomized. OPEs= organophosphate esters. Outlined box reflects the PCA toxicant grouping correlations.
Our longitudinal model of prenatal exposure to chemicals associations with parent-reported expressive language development fit well (χ2 (7) = 7.01, p = 0.43, RMSEA = 0.002, CFI = 1.00, SRMR = 0.028; Figure 2 & Table 3). Prenatal exposure to higher levels of OPs were associated with lower levels of communication at 18 months (−0.15, 95% CI −0.27, −0.03). In other words, one standard deviation increase in prenatal OPs is associated with a decrease of 0.15 standard deviations in communication at 18 months. We did not find any other significant direct associations with chemicals on child communication at 18 months or parent-reported child expressive language at preschool age. As expected, communication at 18 months was positively associated with expressive language at preschool age (0.40, 95% CI 0.30, 0.50). Finally, we found an indirect effect from prenatal exposure to OPs to expressive language in preschool through communication at 18 months (β = −0.06, 95% CI: −0.11, −0.01).
Figure 2.
Associations of prenatal chemical exposures and parent-reported child language development to preschool age.
***p < .001, **p <.01, *p < .05. Preschool sample n=299. Standardized estimate and 95% confidence interval reported. Bolded lines are significant paths, dotted lines are non-significant paths. Major loadings of LMWP (low molecular weight phthalates) are MiBP, ,MnBP, and MBzP; HMWP (high molecular weight phthalates) are DEHP and DiNP; OP Pesticides are DMP and DEP, OPEs are DPHP and DNBP. Indirect effect of pesticides to expressive language through communication: β = −0.06 (−0.11, −0.01). Analysis controlled for birth year, maternal age, maternal education, parity, smoking during pregnancy, alcohol during pregnancy, maternal depression at 6 months, and fish intake.
Table 3.
Path estimates for the associations of prenatal chemical exposures and child language development to preschool age.
| Path | Estimate | CI |
|---|---|---|
|
| ||
| Parent-reported child language development model | ||
| LMWP → Communication (18mos) | 0.06 | −0.06, 0.18 |
| HMWP → Communication (18mos) | 0.04 | −0.08, 0.15 |
| OPs → Communication (18mos) | −0.15 | −0.27, −0.03 |
| OPEs → Communication (18mos) | −0.05 | −0.17, 0.07 |
| LMWP → Expressive Language (Preschool) | −0.07 | −0.18, 0.04 |
| HMWP → Expressive Language (Preschool) | 0.06 | −0.05, 0.16 |
| OPs → Expressive Language (Preschool) | 0.05 | −0.07, 0.15 |
| OPEs → Expressive Language (Preschool) | 0.04 | −0.07, 0.15 |
| Communication (18mos)→ Expressive Language (Preschool) | 0.40 | 0.30, 0.50 |
| OPs → Communication (18mos)→ Expressive Language (Preschool) | −0.06 | −0.11, −0.01 |
| Child sex →Communication (18mos) | 0.21 | −0.02, 0.44 |
| Child sex →Expressive Language (Preschool) | 0.16 | −0.05, 0.37 |
| Teacher-reported child language development model | ||
| LMWP → Communication (18mos) | 0.06 | −0.06, 0.18 |
| HMWP → Communication (18mos) | 0.04 | −0.07, 0.16 |
| OPs → Communication (18mos) | −0.15 | −0.27, −0.04 |
| OPEs → Communication (18mos) | −0.05 | −0.17, 0.08 |
| LMWP → Expressive Language (Preschool) | −0.13 | −0.24, −0.01 |
| HMWP → Expressive Language (Preschool) | 0.08 | −0.03, 0.19 |
| OPs → Expressive Language (Preschool) | −0.07 | −0.18, 0.05 |
| OPEs → Expressive Language (Preschool) | 0.001 | −0.12, 0.12 |
| Communication (18mos)→ Expressive Language (Preschool) | 0.25 | 0.14, 0.36 |
| OPs Communication (18mos)→ Expressive Language (Preschool) | −0.04 | −0.08, −0.004 |
| Child sex →Communication (18mos) | 0.11 | −0.01, 0.22 |
| Child sex →Expressive Language (Preschool) | 0.12 | 0.01, 0.23 |
Note: Preschool sample n=299 with all non-missing data. Standardized estimates and 95% confidence interval reported. LMWP = low molecular weight phthalates, HMWP = high molecular weight phthalates, OPs = organophosphorous pesticides, OPEs = organophosphate esters. Analysis controlled for birth year, maternal age, maternal education, parity, smoking during pregnancy, alcohol during pregnancy, maternal depression at 6 months, and fish intake.
Results from the longitudinal model of the prenatal exposure to chemicals associations with teacher-reported language development in preschool also fit well (χ2 (7) = 6.91, p = 0.44, RMSEA = 0.00, CFI = 1.00, SRMR = 0.028) (Figure 3 & Table 3). Again, we found a negative direct effect of OPs on parent-reported communication at 18 months (−0.15, 95% CI −0.27, −0.04), and an indirect effect from OPs to teacher-rated preschool expressive language through parent-reported communication at 18 months (β = −0.04, 95% CI −0.08, −0.004). In other words, one standard deviation increase in prenatal OPs is associated with a decrease of 0.15 standard deviations in communication at 18 months. Additionally, prenatal exposure to higher levels of LMWP was directly associated with lower expressive language at the preschool age (β = −0.13, 95% CI −0.24, −0.01), an association that was not mediated through communication at 18 months. In other words, one standard deviation increase in prenatal LMWPs is associated with a 0.13 standard deviation decrease in expressive language at the preschool age.
Figure 3.
Teacher model with associations of prenatal chemical exposures and child language development to preschool age.
***p < .001, **p <.01, *p < .05. Preschool sample n=299. Standardized estimate and 95% confidence interval reported. Bolded lines are significant paths, dotted lines are non-significant paths. Major loadings of LMWP (low molecular weight phthalates) are MiBP, ,MnBP, and MBzP; HMWP (high molecular weight phthalates) are DEHP and DiNP; OP Pesticides are DMP and DEP, OPEs are DPHP and DNBP. Indirect effect of pesticides to expressive language through communication: β = −0.04 (−0.08, −0.004). Analysis controlled for birth year, maternal age, maternal education, parity, smoking during pregnancy, alcohol during pregnancy, maternal depression at 6 months, and fish intake.
In sensitivity analyses, we examined the influence of prenatal exposure to chemicals on the 18 month and preschool language assessments separately to compare to prior literature. When examining the association between the chemical exposures and child communication at 18 months among all children with 18-month data (n = 956), and separately within the random sample (n = 471), we found that the magnitude and direction of associations were similar to the longitudinal model, although the statistical significance sometimes changed (Supplemental Table 2). This suggests that oversampling ADHD cases or limiting the study to only children with longitudinal communication information at 18 months and the preschool period, did not induce selection bias, or meaningfully alter interpretation. As expected, statistical significance and precision in estimates was often highest in the subset with the largest sample size. We further explored multigroup analyses to examine whether there were differences in toxicant associations for boys and girls (Supplemental Table 3) and found little evidence of sex-specific associations. Specifically, the magnitude and direction of toxicant associations were very similar for boys and girls. Finally, we examined the associations between the prenatal exposure to chemicals and child expressive language at preschool age as reported by the parent and teacher, without accounting for children’s earlier language ability (Supplemental Table 4). We found that the parent model looked similar to the longitudinal model; specifically, there were no substantial direct associations between phthalates, OPs and OPEs and preschool expressive language. The teacher model also exhibited patterns similar to the SEM longitudinal model, although the latter associations were estimated with greater precision. Both models suggest modest negative associations between OPs and low molecular weight phthalates and expressive language. It is important to note that these models for expressive language do not account for earlier child communication ability.
4. Discussion
We examined associations between prenatal exposure to multiple chemicals and longitudinal measures of early childhood language development from toddlerhood to preschool age. We found that prenatal exposure to OPs was negatively associated with preschool age expressive language through deficits in early communication ability measured at 18 months. We also found that low molecular weight phthalates (i.e., MiBP, MnBP, and MBzP) were negatively associated with preschool age expressive language, although these associations were not mediated by effects on communication at 18-months. Our results therefore suggest the presence of multiple age-specific mechanistic pathways through which prenatal exposure to chemicals influence child language ability.
Our finding that prenatal OP exposure was negatively associated with early communication skills at 18 months, which in turn, was associated with later child expressive language skills in preschool, is in accordance with a more extensive neurodevelopmental literature examining the impact of OPs on early child cognitive development [44–46]. The Ages and Stages communication subscale included on the MoBa 18-month questionnaire is intended to capture deficits in early communication (production and comprehension), and as such, may be identifying children’s early inability to gesture and form basic sentences. This finding is supported by a study of an effect of OPs on early language ability in toddlerhood [11], and supports the null findings from Donauer and colleagues [47] and Cartier and colleagues [48], given we did not find a direct association between prenatal OPs and language ability in the preschool age. While these two studies and our study did not find a direct effect of prenatal exposure to OPs on preschool language ability, the current study could suggest that these previous null effects might be due to earlier neurological deficits. In the case of the current study, these earlier neurological effects are child communication ability at 18 months. This is supported by research examining the effects of OPs on broader cognitive functioning in infancy and toddlerhood—that includes early language correlates—(e.g., 49, 22) as evinced by the indirect pathway from prenatal OPs to preschool language ability through child language ability at 18 months.
In contrast, we found that phthalates were not associated with parent reported early communication ability at 18 months, or parent-reported expressive language in the preschool period, which is in line with a previous study that found no association between phthalates mixtures and parent-reported children’s language ability at 3 years old [19]. However, other studies have reported negative associations between phthalates and parent-reported communication skills in early childhood (approximately 3 years old) [12, 20]. These two previous studies more specifically found that LMWP (e.g., MEP, MBzP) were associated with lower communication scores using the Communicative Development Inventories and utterances count, although these studies did not account for correlated phthalate exposures or co-exposure to other potential neurotoxicants like OPs and OPEs. Although we did not find associations with parent-reported expressive language, we did find an association of LMWP with teacher reported expressive language ability in the preschool period, potentially highlighting differences in sensitivity across reporters. Teachers generally have better training to more competently assess a child’s communication ability [51], particularly as they observe multiple children at any given time. As approximately half of the children included were first born, it is possible their parents are less able to identify problems in child language development, since as new parents they may be observing child development for the first time. However, we found that multiparous mothers rated their children lower on their expressive abilities at preschool age as compared to nulliparous mothers (t (224.72) = −2.74, p = 0.007). Examining the teacher model in comparison to the parent model can potentially help researchers consider any potential reporter bias that might be evident when the parent is reporting on both communication skills at 18 months and at the preschool age. The associations we found in the teacher model may be more indicative of the hypothetical true effect of chemicals on language development and support previous literature finding associations between phthalates and cognitive development [50], as compared to a potential parent bias in reporting on their child.
Finally, we found that prenatal exposure to organophosphate esters (OPEs) were not associated with child language ability at any age. Only one other study has examined the relationship between OPEs and language ability, finding no association except in sex stratified models where BDCIPP had a positive effect on male’s language ability [18]. While BDCIPP was including in the PCA analysis for OPEs, the highest loadings were DPHP and DNBP, so our study may not be ideally suited to identify any unique effects of BDCIPP on language ability. Additionally, our sample had lower exposure to BDCIPP as compared to Doherty et al., and thus used a dichotomous variable of below or above the limit of detection, possibly hindering our ability to examine the potential association of BDCIPP. There is only a small literature examining the effects of OPEs on children’s neurodevelopment, suggesting deleterious effects on children’s early cognitive abilities from two to seven years old [23, 52]. Our study adds to this limited literature by suggesting that any associations with cognition may not manifest in deficits in language specifically.
There are several limitations to this study. First, while participants in this study were drawn from the well-characterized MoBa cohort which enrolled almost half of all pregnant women in Norway during the eligible period, enrolled participants tended to disproportionately represent older, married women with fewer high-risk behaviors [48]. Additionally, this specific subset of the MoBa cohort oversampled children with symptoms of ADHD, which might limit the generalizability of the findings. Research suggests links between ADHD and language problems, and a recent meta-analysis— including children ages 3–18— highlighted the cross-sectional associations between ADHD and language problems [54]. However, norms from other samples using the CDI suggest that this current sample has a relatively normative language development [49.29 versus our sample: 45.60 and 48.08] (e.g. [55]). Additionally, comparing the magnitude of associations between all children with 18-month data, and only those selected into the random sample (Supplemental Table 2), we found that OPs had the same magnitude of association, suggesting that oversampling on ADHD symptoms was not driving any associations between toxicants and children’s language ability.
Another limitation of our study is that the preschool-aged assessment was limited to those children who were selected to participate in an onsite clinical assessment, and thus the sample available for longitudinal analyses was small. Additionally, the 18-month language assessment included a reduced number of items, although it showed good reliability in the sample. This reduced measure might provide a limited range of language ability at 18 months. Finally, our study had only a single assessment of maternal chemical exposures during pregnancy, and this might not be indicative of exposure levels throughout pregnancy due to short half-lives of the exposures examined. As there might be high levels of intra-individual variability, the use of a single sample could increase the risk of misclassification [56–57]. Studies with multiple measures of prenatal exposure would better characterize exposure over the entire period of pregnancy.
While there are limitations to this analysis, this paper also has several strengths. This is the first study to examine the association between the prenatal exposure to multiple chemicals and child language development at multiple points in time simultaneously. The inclusion of multiple developmental periods allows us to examine age-specific patterns of language development, which may suggest different mechanistic pathways. This study also considers the effects of multiple exposures simultaneously, while accounting for their intercorrelations. The majority of studies that have examined prenatal exposure to these chemicals in relation to early language ability have utilized single pollutant models; however, exposure mixture methods can help better account for the real-world experience of expectant mothers. There are a multitude of analytic methods to examine mixtures and the specific statistical approach to estimating the health or developmental impacts of exposure mixtures should be driven by the research question [58]. In this study, we sought to quantify the joint effect of total exposure to classes of compounds on the development of language, while accounting for total exposure to other chemical classes, so we combined PCA with SEM. SEM allows for the longitudinal modeling to address our developmental question. In comparison to PCA with SEM, alternative methods like WQS can pinpoint specific chemicals as bad actors within a group of related exposures [30], while Bayesian Kernel Machine Regression examines individual and joint effects of exposures [59]. Each method is important in examining different questions related to chemical mixture effects on outcomes (e.g., PCA: what is the overall effect of these chemicals on an outcome; WQS: what is the bad actor?) [59]. The results of the current study can act as both informational, given the importance of examining joint effects, but also as a screening approach to highlight groups of compounds with possible harmful impacts that future studies may disentangle to investigate bad actors.
The emergence of language in early childhood foreshadows downstream development of school readiness and other competencies. Understanding the influences on language development can not only pinpoint mechanisms on which to intervene to reduce language deficits, but also help us consider language itself as an anchor to alter children’s developmental trajectories. We found that prenatal exposure to OPs and phthalates are associated with age-specific deficits in language development as rated by parents and teachers at 18 months and the preschool period. Much remains unknown with respect to the role of chemical exposures on early childhood language development. Future work should focus on broad language development constructs and differentiate between language production and language comprehension, which we were unable to do in the present study. The etiology of language development is complex, and this work further highlights the importance of the prenatal environment as a mechanism of influence that are associated with deficits in early language acquisition and ability, which could signal increased behavioral problems and academic difficulties in later childhood that extend into adolescence.
Supplementary Material
Highlights.
We assessed prenatal chemical mixtures on early child language ability.
Early language ability mediated prenatal pesticides and low preschool language.
Prenatal phthalates were associated with lower preschool-age language ability.
This study advances the literature by considering cascading, and age-specific effects.
Acknowledgements
The Norwegian Mother, Father and Child Cohort Study is supported by the Norwegian Ministry of Health and Care Services and the Ministry of Education and Research. We are grateful to all the participating families in Norway who take part in this on-going cohort study.
Funding:
This study was funded in part by the National Institute of Health (NIH) and National Institute of Health Science (NIEHS) R01ES021777, P30ES010126, and F32 ES031832-01 (A.M. Ramos) and by the Intramural Research Program of the NIH/NIEHS. The Norwegian Mother, Father, and Child Cohort study (MoBa) is supported by the Norwegian Ministry of Health and Care Services (HOD) and the Ministry of Education and Research, NIH/NIEHS (nu. NO1-ES-75558), NIH/National Institute of Neurological Disorders (NINDS) (nu. 1 U01 NS 047537-01 and nu. 2 U01 NS 047537-06A1). The Preschool ADHD study, a sub-study within MoBa, was funded by grants and funds from the Norwegian Ministry of Health, The Norwegian Health Directorate, The South Eastern Health Region, and the G&PJ Sorensen Fund for Scientific Research, The Norwegian Resource Center for ADHD, Tourette’s Syndrome, and Narcolepsy.
Footnotes
Declaration of interests
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.
Ethics:
The current study is based on version 9 of the quality-assured data files released for research in preschool ADHD. The establishment of MoBa and initial data collection was based on a license from the Norwegian Data Protection Agency and approval from The Regional Committees for Medical and Health Research Ethics. The MoBa cohort is currently regulated by the Norwegian Health Registry. The current study was approved by the Regional Committee for Medical Research Ethics in Norway and reviewed and determined to be exempt from further review by the Office of Human Research Ethics at the University of North Carolina at Chapel Hill.
Credit Author Statement
Amanda M. Ramos: conceptualization, methodology, formal analysis, writing, visualization
Amy H. Herring: methodology, writing, supervision
Gro D. Villanger: writing, review, data acquisition,
Cathrine Thomsen: writing, review, data acquisition
Amrit K. Sakhi: writing, review, data acquisition
Enrique Cequier: review, data acquisition
Heidi Aase: conceptualization, data acquisition, writing, funding acquisition, supervision
Stephanie M. Engel: conceptualization, resources, formal analysis, writing, visualization, data acquisition, funding acquisition, supervision
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.Abreu PMJ, Gathercole SE, Martin R: Disentangling the relationship between working memory and language: The roles of short-term storage and cognitive control. Learning and Individual Differences 2011, 21:569–574. [Google Scholar]
- 2.Gibson JL, Newbury DF, Durkin K, Pickles A, Conti-Ramsden G, Toseeb U: Pathways from the early language and communication environment to literacy outcomes at the end of primary school; the roles of language development and social development. Oxford Review of Education 2021, 47:260–283. [Google Scholar]
- 3.Gooch D, Thompson P, Nash HM, Snowling MJ, Hulme C: The development of executive function and language skills in the early school years. Journal of Child Psychology and Psychiatry 2016, 57:180–187. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Hentges RF, Devereux C, Graham SA, Madigan S: Child Language Difficulties and Internalizing and Externalizing Symptoms: A Meta-Analysis. Child Development 2021, 92:691–715. [DOI] [PubMed] [Google Scholar]
- 5.Luyster RJ, Seery A, Talbott MR, Tager-Flusberg H: Identifying Early-Risk Markers and Developmental Trajectories for Language Impairment in Neurodevelopmental Disorders. Developmental Disabilities Research Reviews 2011, 17:151–159. [DOI] [PubMed] [Google Scholar]
- 6.Kovas Y, Hayiou-Thomas ME, Oliver B, Dale PS, Bishop DVM, Plomin R: Genetic Influences in Different Aspects of Language Development: The Etiology of Language Skills in 4.5-Year-Old Twins. Child Development 2005, 76:632–651. [DOI] [PubMed] [Google Scholar]
- 7.Rogers CR, Nulty KL, Betancourt MA, DeThorne LS: Causal effects on child language development: A review of studies in communication sciences and disorders. Journal of Communication Disorders 2015, 57:3–15. [DOI] [PubMed] [Google Scholar]
- 8.Dzwilewski KL, Schantz SL: Prenatal Chemical Exposures and Child Language Development. Journal of Communication Disorders 2015, 57:41–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Barker DJ: In utero programming of chronic disease. Clinical Science 1998, 95(2):115–128. [PubMed] [Google Scholar]
- 10.Grandjean P, Landrigan PJ: Neurobehavioural effects of developmental toxicity. The Lancet Neurology 2014, 13:330–338. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Andersen HR, Dalsager L, Jensen IK, Timmermann CAG, Olesen TS, Trecca F, Nielsen F, Schoeters G, Kyhl HB, Grandjean P et al. : Prenatal exposure to pyrethroid and organophosphate insecticides and language development at age 20–36 months among children in the Odense Child Cohort. International Journal of Hygiene and Environmental Health 2021, 235:113755. [DOI] [PubMed] [Google Scholar]
- 12.Bornehag CG, Lindh C, Reichenberg A, Wikström S, Unenge Hallerback M, Evans SF, Sathyanarayana S, Barrett ES, Nguyen RHN, Bush NR et al. : Association of Prenatal Phthalate Exposure With Language Development in Early Childhood. JAMA Pediatrics 2018, 172:1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Bornehag CG, Reichenber A, Swan SH: Language Development of Young Children Is Not Linked to Phthalate Exposure-Reply. JAMA Pediatr 2019, 173(5):499. [DOI] [PubMed] [Google Scholar]
- 14.Bellinger DC, Matthews-Bellinger JA, Kordas K: A developmental perspective on early-life exposure to neurotoxicants. Environment International 2016, 94:103–112. [DOI] [PubMed] [Google Scholar]
- 15.Silva MJ, Barr DB, Reidy JA, Malek NA, Hodge CC, Caudill SP, Brock JW, Needham LL, Calafat AM: Urinary levels of seven phthalate metabolites in the U.S. population from the National Health and Nutrition Examination Survey (NHANES) 1999–2000. Environmental Health Perspectives 2004, 112:331–338. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Woodruff TJ, Zota AR, Schwartz JM: Environmental Chemicals in Pregnant Women in the United States: NHANES 2003–2004. Environmental Health Perspectives 2011, 119:878–885. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Doherty BT, Hammel SC, Daniels JL, Stapleton HM, Hoffman K: Organophosphate Esters: Are These Flame Retardants and Plasticizers Affecting Children’s Health? Current Environmental Health Reports 2019, 6(4):201–213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Doherty BT, Hoffman K, Keil AP, Engel SM, Stapleton HM, Goldman BD, Olshan AF, Daniels JL: Prenatal exposure to organophosphate esters and cognitive development in young children in the Pregnancy, Infection, and Nutrition Study. Environmental Research 2019, 169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Loftus CT, Bush NR, Day DB, Ni Y, Tylavsky FA, Karr CJ, Kannan K, Barrett ES, Szpiro AA, Sathyanarayana S et al. : Exposure to prenatal phthalate mixtures and neurodevelopment in the Conditions Affecting Neurocognitive Development and Learning in Early childhood (CANDLE) study. Environment International 2021, 150:106409. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Olesen TS, Bleses D, Andersen HR, Grandjean P, Frederiksen H, Trecca F, Bilenberg N, Kyhl HB, Dalsager L, Jensen IK et al. : Prenatal phthalate exposure and language development in toddlers from the Odense Child Cohort. Neurotoxicology and Teratology 2018, 65:34–41. [DOI] [PubMed] [Google Scholar]
- 21.Radke EG, Braun JM, Nachman RM, Cooper GS: Phthalate exposure and neurodevelopment: A systematic review and meta-analysis of human epidemiological evidence. Environment International 2020, 137:105408. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Sapbamrer R, Hongsibsong S: Effects of prenatal and postnatal exposure to organophosphate pesticides on child neurodevelopment in different age groups: A systematic review. Environmental Science and Pollution Research 2019, 26:18267–18290. [DOI] [PubMed] [Google Scholar]
- 23.Vuong AM, Yolton K, Cecil KM, Braun JM, Lanphear BP, Chen A: Flame Retardants and Neurodevelopment: An Updated Review of Epidemiological Literature. Current Epidemiology Reports 2020, 7:220–236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Wormuth M, Scheringer M, Vollenweider M, Hungerbühler K: What Are the Sources of Exposure to Eight Frequently Used Phthalic Acid Esters in Europeans? Risk Analysis 2006, 26:803–824. [DOI] [PubMed] [Google Scholar]
- 25.Bellinger DC: Prenatal Exposures to Environmental Chemicals and Children’s Neurodevelopment: An Update. Safety and Health at Work 2013, 4:1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Gilden RC, Huffling K, Sattler B: Pesticides and Health Risks. Journal of Obstetric, Gynecologic & Neonatal Nursing 2010, 39:103–110. [DOI] [PubMed] [Google Scholar]
- 27.Ray DE, Richards PG: The potential for toxic effects of chronic, low-dose exposure to organophosphates. Toxicology Letters 2001, 120:343–351. [DOI] [PubMed] [Google Scholar]
- 28.Dishaw LV, Powers CM, Ryde IT, Roberts SC, Seidler FJ, Slotkin TA, Stapleton HM: Is the PentaBDE replacement, tris (1,3-dichloro-2-propyl) phosphate (TDCPP), a developmental neurotoxicant? Studies in PC12 cells. Toxicology and Applied Pharmacology 2011, 256:281–289. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Preston EV, McClean MD, Claus Henn B, Stapleton HM, Braverman LE, Pearce EN, Makey CM, Webster TF: Associations between urinary diphenyl phosphate and thyroid function. Environment International 2017, 101:158–164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Lazarevic N, Barnett AG, Sly PD, Knibbs LD: Statistical Methodology in Studies of Prenatal Exposure to Mixtures of Endocrine-Disrupting Chemicals: A Review of Existing Approaches and New Alternatives. Environmental Health Perspectives 2019, 127:026001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Magnus P, Irgens LM, Haug K, Nystad W, Skjaerven R, Stoltenberg C, Group MS: Cohort profile: The Norwegian Mother and Child Cohort Study (MoBa. International Journal of Epidemiology 2006, 35:1146–1150. [DOI] [PubMed] [Google Scholar]
- 32.Magnus P, Birke C, Vejrup K, Haugan A, Alsaker E, Daltveit AK, Handal M, Haugen M, Høiseth G, Knudsen GP et al. : Cohort Profile Update: The Norwegian Mother and Child Cohort Study (MoBa. International Journal of Epidemiology 2016, 45:382–388. [DOI] [PubMed] [Google Scholar]
- 33.Choi G, Villanger GD, Drover SSM, Sakhi AK, Thomsen C, Nethery RC, Zeiner P, Knudsen GP, Reichborn-Kjennerud T, Øvergaard KR et al. : Prenatal phthalate exposures and executive function in preschool children. Environment International 2021, 149:106403. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Engel SM, Villanger GD, Nethery RC, Thomsen C, Sakhi AK, Drover SSM, Hoppin JA, Zeiner P, Knudsen GP, Reichborn-Kjennerud T et al. : Prenatal Phthalates, Maternal Thyroid Function, and Risk of Attention-Deficit Hyperactivity Disorder in the Norwegian Mother and Child Cohort. Environmental Health Perspectives 2018, 126:057004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Kamai EM, Villanger GD, Nethery RC, Thomsen C, Sakhi AK, Drover SSM, Hoppin JA, Knudsen GP, Reichborn-Kjennerud T, Zeiner P et al. : Gestational Phthalate Exposure and Preschool Attention Deficit Hyperactivity Disorder in Norway. Environmental Epidemiology 2021, 5:161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Hoppin JA, Ulmer R, Calafat AM, et al. Impact of urine preservation methods and duration of storage on measured levels of environmental contaminants. J Expo Sci Environ Epidemiol. 2006;16(1):39–48. doi: 10.1038/sj.jea.7500435 [DOI] [PubMed] [Google Scholar]
- 37.Cequier E, Sakhi AK, Haug LS, Thomsen C: Development of an ion-pair liquid chromatography-high resolution mass spectrometry method for determination of organophosphate pesticide metabolites in large-scale biomonitoring studies. Journal of Chromatography 2016:32–41. [DOI] [PubMed] [Google Scholar]
- 38.Cequier E, Marcé RM, Becher G, Thomsen C: A high-throughput method for determination of metabolites of organophosphate flame retardants in urine by ultra performance liquid chromatography-high resolution mass spectrometry. Analytica Chimica Acta 2014, 845:98–104. [DOI] [PubMed] [Google Scholar]
- 39.Choi G, Keil AP, Villanger GD, Richardson DB, Daniels JL, Hoffman K, Sakhi AK, Thomsen C, Herring AH, Drover SSM et al. : Pregnancy exposure to common-detect organophosphate esters and phthalates and maternal thyroid function. The Science of the Total Environment 2021, 782:146709. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Squires J, Potter L, Bricker D: The ages and stages user’s guide. In.: Paul H. Brookes Publishing Co.; 1999. [Google Scholar]
- 41.Ireton H, Glascoe FP: Assessing children’s development using parents’ reports. The Child Development Inventory. Clin Pediatr (Phila) 1995, 34(5):248–255. [DOI] [PubMed] [Google Scholar]
- 42.Textor J, van der Zander B, Gilthorpe MS, Liśkiewicz M, Ellison GTH. Robust causal inference using directed acyclic graphs: the R package ‘dagitty.’ Int J Epidemiol. Published online January 15, 2017:dyw341. doi: 10.1093/ije/dyw341 [DOI] [PubMed] [Google Scholar]
- 43.Muthen LK, Muthen BO: Mplus User’s Guide, 8th edn: Muthen & Muthen; 1998–2017. [Google Scholar]
- 44.Eskenazi B, Marks AR, Bradman A, Harley K, Barr DB, Johnson C, Morga N, Jewell NP: Organophosphate Pesticide Exposure and Neurodevelopment in Young Mexican-American Children. Environmental Health Perspectives 2007, 115:792–798. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.González-Alzaga B, Lacasaña M, Aguilar-Garduño C, Rodríguez-Barranco M, Ballester F, Rebagliato M, Hernández AF: A systematic review of neurodevelopmental effects of prenatal and postnatal organophosphate pesticide exposure. Toxicology Letters 2014, 230:104–121. [DOI] [PubMed] [Google Scholar]
- 46.Liu J, Schelar E: Pesticide Exposure and Child Neurodevelopment. Workplace Health & Safety 2012, 60:235–243. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Donauer S, Altaye M, Xu Y, Sucharew H, Succop P, Calafat AM, Khoury JC, Lanphear B, Yolton K: An Observational Study to Evaluate Associations Between Low-Level Gestational Exposure to Organophosphate Pesticides and Cognition During Early Childhood. American Journal of Epidemiology 2016, 184:410–418. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Cartier C, Warembourg C, Le Maner-Idrissi G, Lacroix A, Rouget F, Monfort C, Limon G, Durand G, Saint-Amour D, Cordier S et al. : Organophosphate Insecticide Metabolites in Prenatal and Childhood Urine Samples and Intelligence Scores at 6 Years of Age: Results from the Mother-Child PELAGIE Cohort (France). Environ Health Perspect 2016, 124(5):674–680. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Eskenazi B, Marks AR, Bradman A, et al. Organophosphate Pesticide Exposure and Neurodevelopment in Young Mexican-American Children. Environ Health Perspect. 2007;115(5):792–798. doi: 10.1289/ehp.9828 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Zhang Q, Chen XZ, Huang X, Wang M, Wu J. The association between prenatal exposure to phthalates and cognition and neurobehavior of children-evidence from birth cohorts. Neurotoxicology. 2019;73:199–212. doi: 10.1016/j.neuro.2019.04.007 [DOI] [PubMed] [Google Scholar]
- 51.Hauerwas LB, Stone CA: Are parents of school-age children with specific language impairments accurate estimators of their child’s language skills? Child Language Teaching and Therapy 2000, 16:73–86. [Google Scholar]
- 52.Liu W, Luo D, Xia W, et al. Prenatal exposure to halogenated, aryl, and alkyl organophosphate esters and child neurodevelopment at two years of age. J Hazard Mater. 2021;408:124856. doi: 10.1016/j.jhazmat.2020.124856 [DOI] [PubMed] [Google Scholar]
- 53.Nilsen RM, Vollset SE, Gjessing HK, Skjærven R, Melve KK, Schreuder P, Alsaker ER, Haug K, Daltveit AK, Magnus P: Self-selection and bias in a large prospective pregnancy cohort in Norway. Paediatric and Perinatal Epidemiology 2009, 23:597–608. [DOI] [PubMed] [Google Scholar]
- 54.Korrel H, Mueller KL, Silk T, Anderson V, Sciberras E: Research Review: Language problems in children with Attention-Deficit Hyperactivity Disorder – a systematic meta-analytic review. Journal of Child Psychology and Psychiatry 2017, 58:640–654. [DOI] [PubMed] [Google Scholar]
- 55.Byrne JM, Backman Je, Bawden HN: Minnesota Child Development Inventory: A normativey study. Canadian Psychology 1995, 36(2):115–130. [Google Scholar]
- 56.Johns LE, Cooper GS, Galizia A, Meeker JD: Exposure assessment issues in epidemiology studies of phthalates. Environment International 2015, 85:27–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57.Spaan S, Pronk A, Koch HM, et al. Reliability of concentrations of organophosphate pesticide metabolites in serial urine specimens from pregnancy in the Generation R Study. J Expo Sci Environ Epidemiol. 2015;25(3):286–294. doi: 10.1038/jes.2014.81 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58.Braun Joseph M, Gennings Chris, Hauser Russ, Webster Thomas F. What Can Epidemiological Studies Tell Us about the Impact of Chemical Mixtures on Human Health? Environ Health Perspect. 2016;124(1):A6–A9. doi: 10.1289/ehp.1510569 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59.Hamra GB, Buckley JP. Environmental Exposure Mixtures: Questions and Methods toAddress Them. Curr Epidemiol Rep. 2018;5(2):160–165. doi: 10.1007/s40471-018-0145-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.