Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Jul 1.
Published in final edited form as: Neurotoxicol Teratol. 2023 May 10;98:107182. doi: 10.1016/j.ntt.2023.107182

Associations of prenatal exposure to per- and polyfluoroalkyl substances (PFAS) with measures of cognition in 7.5-month-old infants: an exploratory study

Elizabeth A Enright 1,*,†, Stephanie M Eick 2,*,†, Rachel Morello-Frosch 3,4, Andréa Aguiar 5,6, Megan L Woodbury 5,7, Jenna L N Sprowles 5,8, Sarah Dee Geiger 5,9, Jessica Trowbridge 4, Aileen Andrade 4, Sabrina Smith 10, June-Soo Park 4,10, Erin DeMicco 4, Amy M Padula 4, Tracey J Woodruff 4, Susan L Schantz 5,6
PMCID: PMC10762627  NIHMSID: NIHMS1952064  PMID: 37172619

Abstract

Background:

Prenatal exposure to per- and polyfluoroalkyl substances (PFAS) has been linked to a wide array of adverse maternal and child health outcomes. However, studies examining PFAS in relation to offspring cognition have been inconclusive.

Objective:

We examined whether prenatal exposure to a mixture of PFAS was related to cognition in 7.5-month-old infants.

Methods:

Our analytic sample included participants enrolled in the Chemicals in Our Bodies (CIOB) and Illinois Kids Development Study (IKIDS) cohorts (N = 163). Seven PFAS were measured in 2nd trimester maternal serum samples and were detected in >65% of participants. Infant cognition was measured with a visual recognition memory task using an infrared eye tracker when infants were 7.5 months old. This task included familiarization trials where each infant was shown two identical faces and test trials where each infant was shown the familiar face paired with a novel face. In familiarization, we assessed average run duration (time looking at familiarization stimuli before looking away) as a measure of information processing speed, in addition to time to familiarization (time to reach 20 seconds of looking at stimuli) and shift rate (the number of times infants looked between stimuli), both as measures of attention. In test trials, we assessed novelty preference (proportion of time looking to the novel face) to measure recognition memory. Linear regression was used to estimate associations of individual PFAS with cognitive outcomes, while Bayesian kernel machine regression (BKMR) was used to estimate mixture effects.

Results:

In adjusted single-PFAS linear regression models, an interquartile range increase in PFNA, PFOA, PFOS, PFHxS, PFDeA, and PFUdA was associated with an increase in shift rate, reflecting better visual attention. Using BKMR, increasing quartiles of the PFAS mixture was similarly associated with a modest increase in shift rate. There were no significant associations between PFAS exposure and time to reach familiarization (another measure of attention), average run duration (information processing speed), or novelty preference (visual recognition memory).

Conclusion:

In our study population, prenatal PFAS exposure was modestly associated with an increase in shift rate and was not strongly associated with any adverse cognitive outcomes in 7.5-month-old infants.

Keywords: Per- and polyfluoroalkyl substances, PFAS, infant cognition, visual recognition memory

1. Introduction

Per- and polyfluoroalkyl substances (PFAS) are a class of synthetic chemicals that are widely used in consumer products, including food packaging, clothing, and non-stick cookware (Sunderland et al., 2019). Commonly known as “forever chemicals,” exposures to PFAS have been linked to a wide array of negative long-term health consequences (Rappazzo, Coffman, & Hines, 2017). While exposures to specific PFAS (i.e., Perfluorooctanoic acid [PFOA] and Perfluorooctane Sulfonate [PFOS]) have decreased over the past decades (Dong et al., 2019; Hurley et al., 2013), representative studies of the United States (US) population have found that close to 100% of people continue to have detectable levels of multiple PFAS in their blood stream (Andrews & Naidenko, 2020; Calafat et al., 2007). This is particularly worrisome for pregnant people, as samples taken from placentas and umbilical cords after birth have consistently found detectable levels of PFAS, and studies find that fetuses are directly exposed in utero at levels associated with adverse pregnancy and developmental health effects (Kang et al., 2021; Morello-Frosch et al., 2016; Panagopoulos Abrahamsson et al., 2021). Given the widespread exposure to PFAS, there are concerns as to how prenatal exposures to these chemicals affect offspring. In animal studies, researchers have found that PFAS may alter neural cell differentiation and neuronal apoptosis (Johansson, Eriksson, & Viberg, 2009; Reistad, Fonnum, & Mariussen, 2013; Slotkin et al., 2008) and impact performance on cognitive and behavioral tasks (Jantzen, Annunziato, & Cooper, 2016; Johansson, Fredriksson, & Eriksson, 2008; Spulber et al., 2014). Hormones play a significant role in brain development (Porterfield & Hendrich, 1993), and PFAS may disturb the hormonal mediation of these neurodevelopmental processes (Coperchini et al., 2021). Yet, the connection between PFAS and cognition in humans remains unclear and understudied (For a review see Rappazzo, Coffman, & Hines, 2017). The National Academies of Sciences, Engineering, and Medicine (2022) recently put out a call for more research on the effects of PFAS on infants and children under 12 years of age as they believed that there was insufficient work to determine the effects of PFAS on neurodevelopment.

Considering the importance of cognitive development during infancy and early childhood for long-term well-being (Berk, 2010), it is essential to understand the implications of PFAS exposures in early life. Several human studies show that increased exposure to PFAS during early life is associated with delayed cognitive development. For example, prenatal exposure to PFAS has been associated with lower scores on the Bayley Scales of Infant Development among females at 6 months of age (Goudarzi et al., 2016), lower IQ scores at 8 years of age (Wang et al., 2015), and increased rates of attention deficit hyperactivity disorder (ADHD); (Lenters et al., 2019). Conversely, other studies assessing associations between child PFAS exposure levels and cognitive performance find that higher exposure is linked to better performance on cognitive tasks. For example, higher compared to lower levels of in utero and childhood PFOA among 6-12 year-olds were associated with an increase in full scale IQ and decrease in rates of ADHD (Stein, Savitz, & Bellinger, 2013). Similarly, among 8-year-old children, increasing prenatal and childhood PFOA and PFNA concentrations were associated with better working memory and perceptual reasoning tasks (Vuong et al., 2019). Additionally, other studies have been inconsistent with both positive and negative associations within the same cohort (Harris et al., 2018; Wang et al., 2015). Reviews examining the relationship between PFAS and cognitive outcomes have found the results to be inconclusive (Liew, Goudarzi, & Oulhote, 2018; Rappazzo, Coffman, & Hines, 2017; Roth & Wilks, 2014).

A shared limitation of these prior studies is that they have examined the effects of one PFAS at a time, failing to account for co-exposure to multiple PFAS, which may have additive or interactive health effects. Additionally, most of the prior work has focused on health effects of PFAS during middle childhood, neglecting infancy, a critical developmental period for early attention, memory, and information processing (Berk, 2010). Therefore, in the present study, we examined associations between a mixture of prenatal PFAS exposures and infant cognition, measured at 7.5 months. We assessed infant cognition using a visual recognition memory (VRM) task, which can be used with preverbal infants and has been shown to be sensitive to prenatal phthalate and psychosocial stress exposures (Dzwilewski et al., 2021; Merced-Nieves et al., 2021).

2. Materials and Methods

2.1. Study design and cohort

Participants in this study were a subset of those enrolled in the Chemicals in Our Bodies (CIOB) and Illinois Kids Development Study (IKIDS) prospective birth cohorts, which together form the ECHO.CA.IL cohort (Eick et al., 2021). Those included in the present study had information on maternal serum PFAS concentrations available and the infant had completed the VRM task at 7.5 months (N = 163 mother-infant pairs; N = 128 from IKIDS and N = 35 from CIOB). Recruitment and retention for the ECHO.CA.IL cohort has been described in detail elsewhere (Eick et al., 2021). Briefly, eligibility criteria for CIOB included being 1) at least 18 years of age, 2) fluent in English or Spanish, and 3) not pregnant with multiples, while pregnant people were eligible for IKIDS if they were 1) between 18 and 40 years of age, 2) able to speak English as their primary language, 3) not pregnant with multiples, 4) considered a low-risk pregnancy, and 5) resided within 30 minutes of the University of Illinois campus. Across both cohorts, information regarding maternal education, maternal age, maternal race/ethnicity, and marital status were obtained via self-reported interview questionnaire, while information on prepregnancy body mass index (BMI; kg/m2), parity, infant sex and gestational age at delivery were obtained via self-report and/or medical record abstraction. The CIOB and IKIDS study protocols were approved by the Institutional Review Boards at the University of California, San Francisco (10-00861) and Berkeley (2010-05-04), and the University of Illinois at Urbana-Champaign (09498), respectively.

2.2. Per- and polyfluoroalkyl substances (PFAS)

Serum samples were obtained from participants in the second trimester (range 12-28 weeks for CIOB and 16-20 weeks for IKIDS) and were stored at −80° C prior to analysis of PFAS. The Environmental Chemical Laboratory at the California Department of Toxic Substances Control (DTSC) quantified twelve PFAS using on-line solid-phase extraction coupled to liquid chromatography and tandem mass spectrometry (Eick et al., 2020), including: PFOA, PFOS, PFNA, perfluoro butane sulfonate (PFBS), perfluorohexanesulphonic acid (PFHxS), perfluoroheptanoic acid (PFHpA), perfluorodecanoic acid (PFDeA), perfluoroundecanoic acid (PFUdA), perfluorododecanoic acid (PFDoA), perfluorooctane sulfonamide (PFOSA), methyl-perfluorooctane sulfonamide acetic acid (Me- PFOSA-AcOH), and ethyl-perfluorooctane sulfonamide acetic acid (Et-PFOSA-AcOH). Values below the method detection limit (MDL), were assigned the machine-read value if available. When no machine-read values were available, the concentration was replaced with MDL/√2 (Homug & Reed, 1990). We restricted our analysis to those PFAS with ≥65% detection, which included PFOA, PFOS, PFNA, PFHxS, PFDeA, PFUdA, and Me-PFOSA-AcOH. All PFAS were natural log-transformed and standardized to the population interquartile range (IQR), which corresponds to an increase from the 25th to 75th percentile, for downstream analyses.

2.3. Visual Recognition Memory (VRM) Task

We assessed cognition when infants were 7- to 8-months of age using a VRM task modified from Rose and colleagues (1992) that has been described in detail elsewhere (Dzwilewski et al., 2020). Mothers and infants were invited to the lab to complete the VRM task. Each infant sat on their caregiver’s lap facing a large, high-definition TV screen approximately 1.5 m from the infant. The area encompassing the chair where the caregiver and infant were seated and the TV was surrounded by black curtains to minimize distractions. The EyeLink 1000 Plus system was used to present stimuli on the TV screen, and infants’ looking was tracked using an infrared eye tracker (SR Research Ltd., Mississauga, Ontario, Canada). During the task, caregivers were instructed to look down at their infant’s head and not at the TV screen, so that they did not influence their infant’s behavior. At the beginning of the assessment, researchers calibrated and validated infant gaze tracking (Dzwilewski et al., 2020).

The VRM task included black-and-white photographs of human faces presented side-by-side. There were 5 blocks, and each block consisted of 3 trials of the faces. The first trial of each block allowed infants to familiarize themselves with a face (i.e., the familiarization trial). Each infant was presented with two identical faces side-by-side until they accumulated 20 seconds of total looking time at the faces. Two test trials followed the familiarization trial. During the test trials, two faces were again shown, but one face was the previously familiarized face, and the second face was novel. In one test trial the novel face was shown on the left side of the TV screen, and in the other test trial the novel face was shown on the right side. During the test trials, each infant needed to accumulate at least 1 second of looking at the stimuli, and then the trial would continue for an additional 5 seconds regardless of their looking behaviors.

Data were processed using the SR Research DataViewer software (SR Research Ltd., Mississauga, Ontario, Canada) and pooled across cohorts. Infants were included in the present analyses if they completed all trials. In IKIDS, approximately half of the infants were familiarized to stimulus set 1 and the other half of infants were familiarized to stimulus set 2, such that the novel faces in set 1 were the familiar faces in set 2 and vice versa. All test trials were counterbalanced so that half of the infants saw the novel stimuli on the right side first and the other half of infants saw the novel stimuli on the left side first. In CIOB, infants saw set 2 as familiar. Therefore, VRM outcomes included in our primary analysis were restricted to IKIDS infants who saw set 2 as familiar.

During the 5 familiarization trials, average run duration data was used as a measure of information processing speed, whereas time to familiarization and shift rate were used as measures of different aspects of attention. Run duration was the average time (in seconds) spent looking at the two stimuli before looking away, with shorter run duration times reflecting faster information processing (Dzwilewski et al., 2020; Rose et al., 1992). Time to familiarization was the average time (in seconds) to reach the familiarization criterion of 20 seconds looking at the stimuli, with shorter time to familiarization reflecting better visual attention (Dzwilewski et al., 2020; Rose et al., 1992). Shift rate was the average number of times infants look between stimuli during the familiarization trials, with higher shift rate scores reflecting better attention (Rose et al., 1992). During test trials, novelty preference was calculated to assess recognition memory. Novelty preference was the proportion of looking time that was spent focusing on the novel versus familiar stimulus, with higher scores reflecting better recognition memory (Dzwilewski et al., 2020; Rose et al., 1992). In all analyses, the four cognitive outcome measures were averaged across five blocks of trials for each infant.

2.4. Statistical Methods

We examined distributions of demographic characteristics in our overall analytic sample and across individual cohorts using means, standard deviations (SDs), frequencies, and counts. We similarly assessed distributions of PFAS using geometric means, geometric SDs, and selected percentiles. Generalized additive mixed models (GAMMs) were used to visually assess linearity between individual PFAS and VRM outcomes. Correlations between PFAS with ≥65% detection were estimated using Spearman correlation coefficients (ρ). Linear regression models were used to estimate unadjusted and adjusted associations between individual PFAS and VRM outcomes, respectively. VRM outcomes (shift rate, novelty preference, average run duration, time to reach familiarization) were treated as separate outcomes in individual models. We included an a priori indicator for cohort in all models (unadjusted and adjusted), as well as maternal age, parity, maternal education (an indicator of socioeconomic status), and infant age at assessment in adjusted models. Covariates were chosen via a Directed Acyclic Graph (DAG; Figure S1) and associations with exposures and outcomes in our study population. We did not adjust for smoking status or alcohol consumption, as few participants reported smoking (N = 1) or drinking alcohol (N = 6) during pregnancy.

To account for joint exposure to multiple PFAS, we implemented Bayesian kernel machine regression (BKMR) with component-wide variable selection (10,000 iterations; Bobb et al., 2015). BKMR estimates a nonparametric high-dimensional exposure-response function. Using BKMR, we assessed non-linearity by examining univariate exposure-response functions, which reflect the relationship between an individual PFAS and VRM outcome, holding the remaining PFAS constant at their median value. We then assessed interaction using bivariate exposure-response functions, which show evidence of interaction if the effect of one exposure differs across levels of another. The overall effect of the PFAS mixture was estimated by comparing the expected difference in VRM outcomes when exposures in the mixture were set at the 25th and 75th percentile, relative to when they were all fixed at their 50th percentile. As with our adjusted linear regression models, all BKMR models were adjusted for cohort, maternal age, parity, maternal education, and infant age at assessment.

As a secondary analysis, we assessed modification by infant sex using adjusted linear regression with an interaction term for PFAS*sex, as well as with models stratified by infant sex. We then conducted several sensitivity analyses to examine the robustness of our findings. First, we restricted our adjusted linear regression models to term births (gestational age at delivery ≥37 weeks gestation), as we hypothesized that gestational age at delivery was a mediator. To determine whether our results were robust against the influence of a single cohort, we estimated single PFAS associations using linear regression models stratified by cohort. Lastly, in models restricted to the IKIDS cohort, we assessed interaction by stimulus set using models stratified by set, and by including an interaction term for PFAS*set in adjusted linear regression models that were not restricted to set 2. P-values for interaction <0.1 were considered statistically significant.

3. Results

Of the 163 mother-infant pairs included in the present analysis, 128 were enrolled in IKIDS and 35 were enrolled in CIOB (Table 1). The average maternal age at delivery was 31 years (SD = 4.4) and was higher in CIOB compared to IKIDS (35 versus 30 years). Our analytic sample was well-educated, with more than half having a graduate degree (52.1%), and most participants self-identifying as White (73.6%). Among the infants included in this analysis, there were slightly more females than males (54.6% versus 45.4%). Relative to the overall study population, participants included in this analysis were more likely to have a college or graduate degree and self-identify as White (Eick et al., 2021). With respect to the VRM outcomes, the distribution of novelty preference and average run duration was similar across cohorts. Time to reach familiarization was higher in CIOB (65 versus 50 seconds), while shift rate was higher in IKIDS (0.41 looks between stimuli versus 0.31 looks between stimuli) (Table 1). We previously observed that sociodemographic characteristics are not strongly associated with any of our VRM outcomes (Eick et al., 2021).

Table 1.

Distribution of demographics and visual recognition outcome measures in ECHO.CA.IL analytic sample.

CIOB (N=35) IKIDS (N=128) Total (N=163)
Maternal Age at Delivery (years)
Mean (SD) 35 (4.7) 30 (3.8) 31 (4.4)
Pre-pregnancy Body Mass Index (kg/m2)
Mean (SD) 24 (3.5) 27 (6.3) 26 (5.9)
Missing 6 (17.1%) 0 (0%) 6 (3.7%)
Maternal Education
Less than College Degree 8 (22.9%) 18 (14.1%) 26 (16.0%)
College Degree 7 (20.0%) 45 (35.2%) 52 (31.9%)
Graduate Degree 20 (57.1%) 65 (50.8%) 85 (52.1%)
Maternal Race/Ethnicity
Asian/Pacific Islander 6 (17.1%) 7 (5.5%) 13 (8.0%)
Black <5 8 (6.3%) 9 (5.5%)
Latina 7 (20.0%) <5 10 (6.1%)
Other/Multi-Racial <5 8 (6.3%) 11 (6.7%)
White 18 (51.4%) 102 (79.7%) 120 (73.6%)
Infant Sex
Male 12 (34.3%) 62 (48.4%) 74 (45.4%)
Female 23 (65.7%) 66 (51.6%) 89 (54.6%)
Parity
1+ Births 19 (54.3%) 69 (53.9%) 88 (54.0%)
No Prior Births 13 (37.1%) 59 (46.1%) 72 (44.2%)
Missing 3 (8.6%) 0 (0%) 3 (1.8%)
Marital Status
Married/Living Together 32 (91.4%) 122 (95.3%) 154 (94.5%)
Single <5 6 (4.7%) 9 (5.5%)
Preterm Birth
No 28 (80.0%) 122 (95.3%) 150 (92.0%)
Yes 5 (14.3%) 6 (4.7%) 11 (6.7%)
Missing 2 (5.7%) 0 (0%) 2 (1.2%)
Gestational Age (weeks)
Mean (SD) 39 (1.9) 39 (1.5) 39 (1.6)
Missing 2 (5.7%) 0 (0%) 2 (1.2%)
Novelty Preference (%)
Mean (SD) 56 (6.2) 55 (6.4) 55 (6.3)
Time to Reach Familiarization (seconds)
Mean (SD) 65 (64) 50 (21) 53 (35)
Shift Rate (# looks between stimuli)
Mean (SD) 0.31 (0.1) 0.41 (0.18) 0.39 (0.17)
Average Run Duration (seconds)
Mean (SD) 4.8 (2.1) 4.5 (2.7) 4.6 (2.6)
Open in a new tab

Abbreviations: SD, standard deviation.

Of the 7 PFAS that were detected in >65% of participants, the geometric mean was highest for PFOS (2.11 ng/mL) and PFOA (0.7 ng/mL) (Table 2), whereas PFUdA was detected at the lowest concentrations (0.04 ng/mL). Levels of PFNA, PFOA, PFHxS, and PFOS were slightly higher in IKIDS, whereas levels of PFDeA and PFUdA were slightly higher in CIOB (Table 2). PFAS were moderately correlated with one another, with PFNA and PFOA, and PFDeA and PFUdA having the strongest correlation (ρ = 0.73 for both; Figure S2). PFAS were not strongly correlated with VRM outcomes (Figure S2). We previously observed that PFAS levels were generally highest among those who self-identified as White or Multi-Racial, and had a college or graduate degree and lowest among those who self-identified as Black and had less than a high school education (Eick et al., 2021). The distribution of the remaining PFAS detected in <65% of participants in the overall cohort, as well as the distribution of PFAS levels across demographic characteristics, is provided elsewhere (Eick et al., 2021).

Table 2.

Distribution of second trimester serum levels of per- and polyfluoroalkyl substances (ng/mL) with ≥65% detection in the ECHO.CA.IL analytic sample (N=163).

Percentile
% Above MDL % Machine Readable Geometric Mean (Geometric SD) 5th 25th 50th 75th 95th
PFNA
Overall 98.77 99.39 0.28 (2.07) 0.09 0.19 0.27 0.42 0.7
CIOB 97.14 97.14 0.26 (2.78) 0.07 0.16 0.26 0.42 0.75
IKIDS 99.22 100 0.28 (1.88) 0.1 0.19 0.28 0.41 0.7
PFOA
Overall 98.77 99.39 0.7 (2.22) 0.2 0.43 0.72 1.19 2.06
CIOB 94.29 97.14 0.62 (3.14) 0.12 0.33 0.66 1.17 1.84
IKIDS 100 100 0.73 (1.96) 0.24 0.44 0.73 1.17 2.08
PFHxS
Overall 98.77 98.77 0.5 (2.87) 0.09 0.27 0.5 0.99 2.53
CIOB 97.14 97.14 0.28 (2.82) 0.04 0.17 0.36 0.56 1.05
IKIDS 99.22 99.22 0.58 (2.73) 0.11 0.31 0.57 1.11 2.99
PFOS
Overall 99.39 99.39 2.11 (2.33) 0.6 1.47 2.08 3.2 7.85
CIOB 97.14 97.14 1.43 (2.49) 0.43 1.15 1.75 2.43 3.34
IKIDS 100 100 2.34 (2.23) 0.65 1.6 2.2 3.88 8.17
Me-PFOSA-AcOH
Overall 79.75 93.25 0.04 (4.19) 0 0.01 0.05 0.1 0.22
CIOB 71.43 88.57 0.02 (3.58) 0 0.01 0.02 0.06 0.13
IKIDS 82.03 94.53 0.04 (4.23) 0.01 0.02 0.05 0.11 0.24
PFDeA
Overall 66.87 96.93 0.08 (2.48) 0.02 0.05 0.09 0.15 0.28
CIOB 74.29 88.57 0.11 (2.68) 0.03 0.06 0.11 0.16 0.3
IKIDS 64.84 99.22 0.08 (2.4) 0.02 0.04 0.08 0.14 0.27
PFUdA
Overall 68.71 91.41 0.04 (4.25) 0 0.02 0.05 0.11 0.27
CIOB 77.14 88.57 0.09 (3.1) 0.02 0.05 0.11 0.19 0.49
IKIDS 66.41 92.19 0.04 (4.33) 0 0.02 0.05 0.09 0.24
Open in a new tab

Abbreviations: MDL, method detection limit; SD, standard deviation; CIOB, chemicals in our bodies; IKIDS, Illinois kids development study.

3.1. Relationships between Individual PFAS and the VRM

Across unadjusted and adjusted single-pollutant models, an IQR increase in PFNA, PFOA, PFOS, PFHxS, PFDeA and PFUdA was associated with an increase in shift rate, potentially reflecting better attention (Figure 1; Table S1). The association was greatest in magnitude for PFUdA (β= 0.04, 95% confidence interval [CI]= 0.01, 0.07). When stratifying by cohort, the positive association between PFAS and shift rate persisted only among IKIDS participants, likely as a result of the larger sample size for IKIDS (Table S2). In the overall sample, an IQR increase in PFOS, PFHxS, and PFDeA was associated with a non-significant reduction in time to reach familiarization, reflecting better attention (Figure 1; Table S1). When novelty preference was the outcome of interest, an IQR increase in most PFAS was associated with a non-significant increase. No consistent patterns were observed between PFAS and time to reach familiarization in adjusted or unadjusted linear regression models (Figure 1; Table S1). When stratified by infant sex, the positive association between shift rate and PFNA, PFOA, PFOS, PFDeA and PFUdA persisted only among females, but were only statistically different for PFUdA (p-interaction=0.01) (Figure 2; Table S3). An IQR increase in PFNA, PFOA, and PFHxS was associated with a non-significant reduction in average run duration among males only (Figure 2; Table S3).

Figure 1.

Figure 1.

Open in a new tab

Adjusted linear regression coefficients and 95% confidence intervals for association between visual recognition memory outcomes with an interquartile range increase in second trimester PFAS (ng/mL) (N=160).

Note: PFAS were natural log transformed. Models are adjusted for cohort, maternal age, maternal education, parity, and infant age at assessment. The dotted lines indicate the null value of 0. Confidence intervals that do not cross the null are statistically significant at p-value<0.05.

Figure 2.

Figure 2.

Open in a new tab

Adjusted linear regression coefficients and 95% confidence intervals for association between visual recognition memory outcomes with an interquartile range increase in second trimester PFAS (ng/mL), stratified by infant sex (N=160).

Note: PFAS were natural log transformed. Models are adjusted for cohort, maternal age, maternal education, parity, and infant age at assessment. The dotted lines indicate the null value of 0. Confidence intervals that do not cross the null are statistically significant at p-value<0.05.

Associations were similar when restricted to term births, although CIs were less precise (Table S4). In analyses restricted to the IKIDS cohort, we observed evidence of effect modification between PFHxS and set when shift rate was the outcome, where the positive association only persisted among those who saw set 2, and between PFUdA and set when time to reach familiarization was the outcome, where the negative association persisted only among those who saw set 1 (Table S5).

3.2. Relationships between the PFAS mixture and the VRM

Using BKMR, the univariate exposure response functions showed that increasing PFOS, PFHxS, and PFDeA were associated with a slight decrease in time to reach familiarization, while Me-PFOSA-AcOH, PFOA, and PFUdA were associated with a slight increase. PFNA, PFOS, PFHxS and PFUdA were positively associated with shift rate (Figure 3). Me-PFOSA-AcOH was non-linearly associated with average run duration (Figure 3). When estimating the cumulative effect of the PFAS mixture, we observed a non-significant increase in shift rate and decrease in time to reach familiarization associated with increasing quantiles of the PFAS mixture (Figure 4). Me-PFOSA-AcOH was the strongest contributor to the model which included average run duration as the outcome (poster inclusion probability [PIP]=0.81), while PFUdA and PFOS contributed to most to the models which contained shift rate and time to reach familiarization, respectively (PIP=0.56, PIP=0.23, respectively). No PFAS were strong contributors to novelty preference. We observed no evidence of interactions between different PFAS with respect to novelty preference, time to reach familiarization, or shift rate (Figure S3, S4, and S5). When average run duration was the outcome of interest, we observed suggestive evidence of an interaction between PFUdA and the remaining PFAS, as well as Me-PFOSA-AcOH and the remaining PFAS (Figure S6).

Figure 3.

Figure 3.

Open in a new tab

Univariate exposure–response functions and 95% confidence intervals for the change in visual recognition memory outcomes, resulting from individual PFAS while fixing remaining exposures in the mixture at their 50th percentiles, estimated using BKMR (N=160).

Note: PFAS were natural log transformed. Models are adjusted for cohort, maternal age, maternal education, parity, and infant age at assessment.

Figure 4.

Figure 4.

Open in a new tab

Cumulative effect (estimates and 95% credible intervals) of the PFAS mixture on visual recognition memory outcomes, estimated using BKMR (N=160).

Note: PFAS were natural log transformed. Models are adjusted for cohort, maternal age, maternal education, parity, and infant age at assessment. The dotted lines indicate the null value of 0. Credible intervals that do not cross the null are statistically significant at p-value<0.05.

4. Discussion

In the present study, we examined the associations between prenatal exposure to a PFAS mixture and infant cognition at 7.5 months. Findings from this study suggest that prenatal exposure to PFAS has limited effects on infant cognition. Prenatal PFAS exposure was generally not significantly associated with time to reach familiarization, average run duration, or novelty preference which index attention, information processing speed, and recognition memory, respectively. Furthermore, there was only a modest association between prenatal PFAS exposure and shift rate suggesting that increased exposures were linked to slightly better attention during this task. Previously, we compared PFAS levels in our study population to those observed among reproductive women in NHANES, finding that levels in our study population were lower, thus we cannot rule out that higher levels of exposure would lead to adverse associations (Eick et al., 2021). Nonetheless, these findings are consistent with previous literature that found primarily null or inconsistent associations between prenatal PFAS exposure and offspring cognition (Rappazzo, Coffman, & Hines, 2017).

Several studies in humans have found that PFAS exposure is linked to attention as we found in the current study with shift rate, however the directionality of this effect has differed across studies. In humans, one study found that prenatal exposure to PFOSA, but not other PFAS, was linked to poorer selective attention at 5 years of age (Bach et al., 2022). Further, multiple studies have linked PFAS exposure to ADHD. A birth cohort study in Norway found that increasing levels of PFOS in breastmilk, reflecting early childhood exposure, was linked to higher rates of ADHD (Lenters et al., 2019). A birth cohort from Cincinnati, Ohio, USA found that prenatal exposure to both PFOS and PFNA were associated with hyperactive- impulsive type ADHD (Vuong et al., 2021), and a birth cohort study in Greenland found that increased prenatal PFOA levels were associated with hyperactivity (Høyer et al., 2015). Higher shift rates are generally viewed positively for two reasons (Rose et al., 2003). First, shift rates increase with age. Second, higher shift rates tend to be associated with higher novelty preference scores, indicative of better recognition memory. Based on these associations, researchers have argued that higher shift rates result from infants actively comparing stimuli. However, increased shift rates may also indicate less selective attention to a specific stimulus. For example, if an infant or child is constantly shifting their attention between many stimuli in the world around them., they may not spend enough time focusing on one individual stimulus. This possible explanation would align with the findings between PFAS exposure and less selective attention in childhood. More research is needed to explore the effects of PFAS and attention, as it is also possible that the impacts of PFAS on attention emerge later in childhood as attentional abilities continue to improve.

Additionally, our null findings for associations between PFAS exposure and information processing speed and visual recognition memory add to the growing body of work suggesting limited adverse cognitive effects linked to PFAS exposure. In particular, three different studies indicate that PFAS exposure is linked to limited adverse effects on memory. In a Boston-area birth cohort study, no associations were found between childhood PFAS exposure and visual recognition memory (Harris et al., 2018). In another study, PFOA and PFNA exposure during the prenatal and childhood periods was associated with improved working memory performance in childhood (Vunog et al., 2019). Finally, in a third study that differentiated between verbal and nonverbal working memory, the researchers found that increased prenatal PFAS exposure was linked to poorer nonverbal working memory but better verbal working memory (Skogheim et al., 2020). The results from the current study are in line with previous studies, most of which have failed to find associations between PFAS exposure and memory.

Many of the associations between PFAS and measures of cognition on the VRM task were non-significant. However, we found that an increase in PFNA, PFOA, PFOS, PFHxS, PFDeA, and PFUdA was associated with an increase in shift rate, reflecting better visual attention as well as a PFAS mixture was similarly associated with a modest increase in shift rate. These findings also are in line with some previous work suggesting that increased exposure to PFAS may be linked to better neurodevelopment. In a birth-cohort study, researchers found that higher prenatal exposure to PFOS was linked to a marginal positive association in general cognitive development at 4-5-years of age (Carrizosa et al., 2021). In another study, 7-year-old children who had higher levels PFNA exposure were less likely to have inattention/hyperactivity on the Strengths and Difficulties Questionnaire (Lien et al., 2016). Increased PFOA exposure has also been linked to decreased in inattention/Hyperactivity disorder at 6-12-years old (Stein et al., 2013).

It is possible that some of the associations between PFAS exposure and cognitive outcomes differ based on sex. We observed that the positive associations between PFNA, PFDeA, and PFUdA and shift rate only persisted among females. Other studies have also found sex differences based on exposure to PFAS and other chemicals (Spratlen et al., 2020). For example, higher exposure to PFAS has been linked to problems developing interpersonal skills in females, but not males, at four years of age (Niu et al., 2019). Higher exposure to PFAS has been linked to lower non-verbal working memory in pre-school age females specifically (Skogheim et al., 2020). Further, in 8-year-olds, PFAS exposure impacted full-scale IQ in females, but not males (Vuong et al., 2019).

Comparing the results of this study to other studies examining the effects of chemical and non-chemical exposures on VRM performance, it appears that PFAS are associated with fewer negative outcomes related to cognition in early infancy. In the same IKIDS cohort with comparable sample sizes, both prenatal phthalate exposure (Dzwilewski et al., 2021) and prenatal psychosocial stress exposure (Merced-Nieves et al., 2021) were linked to poorer outcomes on the VRM task. Specifically, prenatal phthalate exposure was associated with changes in both average run duration and novelty preference scores which suggested that exposure was linked to negative effects on information processing speed as well as recognition memory (Dzwilewski et al., 2021). Within a New York City based longitudinal birth cohort, the association between prenatal phthalate exposure and novelty preference, assessed via the Fagan Test of Infant Intelligence, persisted only among females (Ipapo et al., 2017). Another study assessing the impact of cocaine exposure in utero also found that cocaine-exposed infants had lower novelty preference scores relative to infants that were unexposed to cocaine (Singer et al., 2005). The work on prenatal exposure to stress found that all measures of stress were linked to at least one VRM outcome, and overall the findings suggested that stress was negatively associated with attention (Merced-Nieves et al., 2021). Given that the VRM is sensitive to other chemical and non-chemical exposures in the same cohort, this measure is a reliable way to index the effects of chemical and non-chemical exposures on early cognitive outcomes.

This study has a several strengths. First, we utilized the VRM task to assess cognition in early infancy, while most prior work has focused on childhood cognitive outcomes. While the VRM has not been validated in clinical samples, it is a measure adapted from experimental psychology studies and is a more objective measure, as opposed to other self-reported measures of neurodevelopment, as it uses an infrared eye tracker to precisely record infants’ looking behaviors. The dependent VRM measures variables have been relatively consistent across the CIOB and IKIDS cohorts as well as consistent with the original work of Rose and colleagues (1992). A second strength of this study was that we used a mixture approach to assess the cumulative effects of multiple PFAS on cognition in addition to assessing each PFAS alone. We also acknowledge our study’s limitations. First, the participants included in our analytic sample were largely white and had well-educated parents, which limits the generalizability of our results to other populations. Second, we had a modest sample size and lower PFAS levels relative to NHANES, both of which may have limited our statistical power, and this imprecision is reflected in our confidence intervals. However, as stated above similar sample sizes were adequate to detect associations with other chemical and non-chemical factors. Additionally, we did not adjust for multiple comparisons. However, we focused the interpretation of our results on identifying consistent patterns, as opposed to an overreliance on statistical significance of individual point estimates. Notably, adjustment for multiple comparisons is not always necessary in exploratory epidemiologic studies, as it may increase the possibility of a Type 2 error (Rothman, 1990). Lastly, we did not have information on possible sources of PFAS in infancy (e.g., breastmilk), which may impact neurodevelopment and should be explored in future work (Zheng et al., 2021).

4.1. Conclusions

Results from our study suggest that prenatal PFAS exposure is not strongly associated with information processing and memory in a VRM task. Unexpectedly, we observed that increasing PFAS exposures was associated with increased shift rates suggesting better attention, although associations were not always statistically significant. Our findings suggest that PFAS exposures are not adversely associated with cognition in early life.

Supplementary Material

Supplement

Highlights.

  • Infant cognition was assessed at 7.5 months using infrared eye tracking.

  • Prenatal PFAS exposure was modestly associated with better attention.

  • Prenatal PFAS exposure was not associated with any of the other cognitive measures assessed.

Acknowledgements:

We would like to thank the IKIDS and CIOB team for collecting the data, especially CIOB’s clinical research coordinators and IKIDS coordinators, graduate students, and postdocs. Additionally, we would like to thank the data analysis teams for helping to enter and compile the data. Thank you to Aileen Andrade, Cheryl Godwin de Medina, Cynthia Melgoza Canchola, Tali Felson, Harim Lee, Maribel Juarez, Lynn Harvey and Allison Landowski from the CIOB group and Ana Lucic, Shuk Han Ng, Mary Wakefield, Jeni Bushman, Kelsey Dzwilewski, Francheska Merced-Nieves, Mindy Howe, and Darcie Reckers from the IKIDS group. We also thank the study participants who participated in the CIOB and IKIDS studies. Lastly, we would like to thank the DTSC biomonitoring team for the laboratory analysis of PFAS in serum.

Funding Sources:

This work was supported by grants RD83543301 AND RD83543401 from the United States Environmental Protection Agency Children’s Environmental Health and Disease Prevention Research Center, P30 ES019776, P30 ES030284, P01 ES022841 P01 ES022848 and R01ES02705 from the National Institute of Environmental Health Sciences, and UG3OD023272, UH3OD023272, and 5U2COD023375-05 from the National Institutes of Health Environmental influences on Child Health Outcomes (ECHO) program. Stephanie Eick’s participation was partially supported by the JPB Environmental Health Fellowship.

References

  1. Andrews DQ, & Naidenko OV (2020). Population-wide exposure to per-and polyfluoroalkyl substances from drinking water in the United States. Environmental Science & Technology Letters, 7(12), 931–936. [Google Scholar]
  2. Bach CC, Liew Z, Matthiesen NB, Henriksen TB, Bech BH, Nøhr EA, … & Olsen J (2022). In utero exposure to perfluoroalkyl and polyfluoroalkyl substances and attention and executive function in the offspring: A study in the Danish National Birth Cohort. Environmental Research, 212, 113262. [DOI] [PubMed] [Google Scholar]
  3. Berk LE (2010). Exploring lifespan development. [Google Scholar]
  4. Bobb JF, Valeri L, Claus Henn B, et al. , 2015. Bayesian kernel machine regression for estimating the health effects of multi-pollutant mixtures. Biostatistics. 16(3), 493–508. doi: 10.1093/biostatistics/kxu058 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Calafat AM, Wong LY, Kuklenyik Z, Reidy JA, & Needham LL (2007). Polyfluoroalkyl chemicals in the US population: data from the National Health and Nutrition Examination Survey (NHANES) 2003–2004 and comparisons with NHANES 1999–2000. Environmental health perspectives, 115(11), 1596–1602. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Carrizosa C, Murcia M, Ballesteros V, Costa O, Manzano-Salgado CB, Ibarluzea J, … & Lopez-Espinosa MJ (2021). Prenatal perfluoroalkyl substance exposure and neuropsychological development throughout childhood: The INMA Project. Journal of Hazardous Materials, 416, 125185. [DOI] [PubMed] [Google Scholar]
  7. Coperchini F, Croce L, Ricci G, Magri F, Rotondi M, Imbriani M, & Chiovato L (2021). Thyroid disrupting effects of old and new generation PFAS. Frontiers in Endocrinology, 11, 612320. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dong Z, Wang H, Yu YY, Li YB, Naidu R, & Liu Y (2019). Using 2003–2014 US NHANES data to determine the associations between per-and polyfluoroalkyl substances and cholesterol: Trend and implications. Ecotoxicology and environmental safety, 173, 461–468. [DOI] [PubMed] [Google Scholar]
  9. Dzwilewski KL, Merced-Nieves FM, Aguiar A, Korrick SA, & Schantz SL (2020). Characterization of performance on an automated visual recognition memory task in 7.5-month-old infants. Neurotoxicology and teratology, 81, 106904. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Dzwilewski KL, Woodbury ML, Aguiar A, Shoaff J, Merced-Nieves F, Korrick SA, & Schantz SL (2021). Associations of prenatal exposure to phthalates with measures of cognition in 7.5-month-old infants. Neurotoxicology, 84, 84–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Eick SM, Enright EA, Geiger SD, Dzwilewski KL, DeMicco E, Smith S, … & Schantz SL (2021). Associations of maternal stress, prenatal exposure to per-and polyfluoroalkyl substances (PFAS), and demographic risk factors with birth outcomes and offspring neurodevelopment: an overview of the ECHO. CA. IL prospective birth cohorts. International journal of environmental research and public health, 18(2), 742. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Eick SM, Hom Thepaksorn EK, Izano MA, Cushing LJ, Wang Y, Smith SC, … & Morello-Frosch R (2020). Associations between prenatal maternal exposure to per-and polyfluoroalkyl substances (PFAS) and polybrominated diphenyl ethers (PBDEs) and birth outcomes among pregnant women in San Francisco. Environmental Health, 19(1), 1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Goudarzi H, Nakajima S, Ikeno T, Sasaki S, Kobayashi S, Miyashita C, … & Kishi R (2016). Prenatal exposure to perfluorinated chemicals and neurodevelopment in early infancy: The Hokkaido Study. Science of the Total Environment, 541, 1002–1010. [DOI] [PubMed] [Google Scholar]
  14. Harris MH, Oken E, Rifas-Shiman SL, Calafat AM, Ye X, Bellinger DC, … & Sagiv SK (2018). Prenatal and childhood exposure to per-and polyfluoroalkyl substances (PFASs) and child cognition. Environment international, 115, 358–369. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hornung RW, & Reed LD (1990). Estimation of average concentration in the presence of nondetectable values. Applied occupational and environmental hygiene, 5(1), 46–51. [Google Scholar]
  16. Høyer BB, Ramlau-Hansen CH, Obel C, Pedersen HS, Hernik A, Ogniev V, … & Toft G. (2015). Pregnancy serum concentrations of perfluorinated alkyl substances and offspring behaviour and motor development at age 5–9 years–a prospective study. Environmental health, 14(1), 1–11 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Hurley S, Goldberg D, Wang M, Park JS, Petreas M, Bernstein L, … & Reynolds P (2018). Time trends in per-and polyfluoroalkyl substances (PFASs) in California women: declining serum levels, 2011–2015. Environmental science & technology, 52(1), 277–287. [DOI] [PubMed] [Google Scholar]
  18. Ipapo KN, Factor-Litvak P, Whyatt RM, Calafat AM, Diaz D, Perera F, … & Herbstman JB (2017). Maternal prenatal urinary phthalate metabolite concentrations and visual recognition memory among infants at 27 weeks. Environmental research, 155, 7–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Jantzen CE, Annunziato KM, & Cooper KR (2016). Behavioral, morphometric, and gene expression effects in adult zebrafish (Danio rerio) embryonically exposed to PFOA, PFOS, and PFNA. Aquatic toxicology, 180, 123–130. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Johansson N, Eriksson P, & Viberg H (2009). Neonatal exposure to PFOS and PFOA in mice results in changes in proteins which are important for neuronal growth and synaptogenesis in the developing brain. Toxicological Sciences, 108(2), 412–418. [DOI] [PubMed] [Google Scholar]
  21. Johansson N, Fredriksson A, & Eriksson P (2008). Neonatal exposure to perfluorooctane sulfonate (PFOS) and perfluorooctanoic acid (PFOA) causes neurobehavioural defects in adult mice. Neurotoxicology, 29(1), 160–169. [DOI] [PubMed] [Google Scholar]
  22. Kang H, Kim HS, Yoon YS, Lee J, Kho Y, Lee J, … & Kim YA (2021). Placental Transfer and Composition of Perfluoroalkyl Substances (PFASs): A Korean Birth Panel of Parent-Infant Triads. Toxics, 9(7), 168. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Lenters V, Iszatt N, Forns J, Čechová E, Kočan A, Legler J, … & Eggesbø M (2019). Early-life exposure to persistent organic pollutants (OCPs, PBDEs, PCBs, PFASs) and attention-deficit/hyperactivity disorder: A multi-pollutant analysis of a Norwegian birth cohort. Environment International, 125, 33–42. [DOI] [PubMed] [Google Scholar]
  24. Lien GW, Huang CC, Shiu JS, Chen MH, Hsieh WS, Guo YL, & Chen PC (2016). Perfluoroalkyl substances in cord blood and attention deficit/hyperactivity disorder symptoms in seven-year-old children. Chemosphere, 156, 118–127. [DOI] [PubMed] [Google Scholar]
  25. Liew Z, Goudarzi H, & Oulhote Y (2018). Developmental exposures to perfluoroalkyl substances (PFASs): an update of associated health outcomes. Current environmental health reports, 5(1), 1–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Merced- Nieves FM, Dzwilewski KL, Aguiar A, Lin J, & Schantz SL (2021). Associations of prenatal maternal stress with measures of cognition in 7.5- month- old infants. Developmental psychobiology, 63(5), 960–972. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Morello-Frosch R, Cushing LJ, Jesdale BM, Schwartz JM, Guo W, Guo T, … & Woodruff TJ (2016). Environmental chemicals in an urban population of pregnant women and their newborns from San Francisco. Environmental science & technology, 50(22), 12464–12472. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. National Academies of Sciences, Engineering, and Medicine. (2022). Guidance on PFAS Exposure, Testing, and Clinical Follow-Up. [PubMed]
  29. Niu J, Liang H, Tian Y, Yuan W, Xiao H, Hu H, … & Miao M (2019). Prenatal plasma concentrations of Perfluoroalkyl and polyfluoroalkyl substances and neuropsychological development in children at four years of age. Environmental Health, 18(1), 1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Panagopoulos Abrahamsson D, Wang A, Jiang T, Wang M, Siddharth A, Morello-Frosch R, … & Woodruff TJ (2021). A comprehensive non-targeted analysis study of the prenatal exposome. Environmental science & technology, 55(15), 10542–10557. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Porterfield SP, & Hendrich CE (1993). The role of thyroid hormones in prenatal and neonatal neurological development—current perspectives. Endocrine reviews, 14(1), 94–106. [DOI] [PubMed] [Google Scholar]
  32. Rappazzo KM, Coffman E, & Hines EP (2017). Exposure to perfluorinated alkyl substances and health outcomes in children: a systematic review of the epidemiologic literature. International journal of environmental research and public health, 14(1), 691. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Reistad T, Fonnum F, & Mariussen E (2013). Perfluoroalkylated compounds induce cell death and formation of reactive oxygen species in cultured cerebellar granule cells. Toxicology letters, 218(1), 56–60. [DOI] [PubMed] [Google Scholar]
  34. Rose SA, Feldman JF, & Jankowski JJ (2003). The building blocks of cognition. The Journal of pediatrics, 143(4), 54–61. [DOI] [PubMed] [Google Scholar]
  35. Rose SA, Feldman JF, & Wallace IF (1992). Infant information processing in relation to six-year cognitive outcomes. Child Development, 63(5), 1126–1141. 10.1111/j.1467-8624.1992.tb01684.x [DOI] [PubMed] [Google Scholar]
  36. Roth N, & Wilks MF (2014). Neurodevelopmental and neurobehavioural effects of polybrominated and perfluorinated chemicals: a systematic review of the epidemiological literature using a quality assessment scheme. Toxicology letters, 230(2), 271–281. [DOI] [PubMed] [Google Scholar]
  37. Rothman KJ. (1990). No adjustments are needed for multiple comparisons. Epidemiology. 1(1):43–6. [PubMed] [Google Scholar]
  38. Singer LT, Eisengart LJ, Minnes S, Noland J, Jey A, Lane C, & Min MO (2005). Prenatal cocaine exposure and infant cognition. Infant Behavior and Development, 28(4), 431–444. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Skogheim TS, Villanger GD, Weyde KVF, Engel SM, Suren P, Øie MG, … & Aase H (2020). Prenatal exposure to perfluoroalkyl substances and associations with symptoms of attention-deficit/hyperactivity disorder and cognitive functions in preschool children. International journal of hygiene and environmental health, 223(1), 80–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Slotkin TA, MacKillop EA, Melnick RL, Thayer KA, & Seidler FJ (2008). Developmental neurotoxicity of perfluorinated chemicals modeled in vitro. Environmental health perspectives, 116(6), 716–722. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Spratlen MJ, Perera FP, Lederman SA, Rauh VA, Robinson M, Kannan K, … & Herbstman J (2020). The association between prenatal exposure to perfluoroalkyl substances and childhood neurodevelopment. Environmental Pollution, 263, 114444. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Spulber S, Kilian P, Wan Ibrahim WN, Onishchenko N, Ulhaq M, Norrgren L, … & Ceccatelli S (2014). PFOS induces behavioral alterations, including spontaneous hyperactivity that is corrected by dexamfetamine in zebrafish larvae. PloS one, 9(4), e94227. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Stein CR, Savitz DA, & Bellinger DC (2013). Perfluorooctanoate (PFOA) and neuropsychological outcomes in children. Epidemiology (Cambridge, Mass.), 24(4), 590. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Sunderland EM, Hu XC, Dassuncao C, Tokranov AK, Wagner CC, & Allen JG (2019). A review of the pathways of human exposure to poly-and perfluoroalkyl substances (PFASs) and present understanding of health effects. Journal of exposure science & environmental epidemiology, 29(2), 131–147. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Taibl KR, Schantz S, Aung M, Padula A, Geiger S, Smith S, … & Eick SM (2022). Associations of per-and polyfluoroalkyl substances (PFAS) and their mixture with oxidative stress biomarkers during pregnancy. Environment International, 107541. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Vuong AM, Webster GM, Yolton K, Calafat AM, Muckle G, Lanphear BP, & Chen A (2021). Prenatal exposure to per-and polyfluoroalkyl substances (PFAS) and neurobehavior in US children through 8 years of age: the HOME study. Environmental research, 195, 110825. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Vuong AM, Yolton K, Xie C, Dietrich KN, Braun JM, Webster GM, … & Chen A (2019). Prenatal and childhood exposure to poly-and perfluoroalkyl substances (PFAS) and cognitive development in children at age 8 years. Environmental research, 172, 242–248. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Wang Y, Rogan WJ, Chen HY, Chen PC, Su PH, Chen HY, & Wang SL (2015). Prenatal exposure to perfluroalkyl substances and children’s IQ: The Taiwan maternal and infant cohort study. International journal of hygiene and environmental health, 218(7), 639–644. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Zheng G, Schreder E, Dempsey JC, Uding N, Chu V, Andres G, Sathyanarayana S, Salamova A. Per- and Polyfluoroalkyl Substances (PFAS) in Breast Milk: Concerning Trends for Current-Use PFAS. (2021) Environ Sci Technol. 55(11):7510–7520. doi: 10.1021/acs.est.0c06978. Epub 2021 May 13. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Supplement
NIHMS1952064-supplement-Supplement.docx (736.5KB, docx)

RESOURCES