Faculty Bibliography

BACKGROUND:A large body of evidence has documented the effects of ortho-phthalates, additives to soft plastic used in food packaging, on premature birth. Ortho-phthalates are increasingly replaced with alternatives including tere-phthalates and di(isononyl) cyclohexane-1,2-dicarboxylate (DINCH). A limited toxicological literature suggests that DINCH metabolites disrupt sex hormones and promote oxidant stress known to contribute to adverse birth outcomes, and epidemiologic studies have also been limited in sample size, inclusivity of different populations and measurements at different time points in pregnancy. METHODS:We leveraged a large American cohort comprised of mother-child dyads from twenty-two cohorts to assess associations of trimester-specific and pregnancy-wide DINCH metabolites with birth weight (BW) and gestational age (GA). FINDINGS/RESULTS:Average prenatal urinary DINCH was inversely associated with birth length, birthweight, and birthweight for gestational age z-scores, but not GA. A tenfold increase in the pregnancy-averaged DINCH metabolite cyclohexane-1,2-dicarboxylic acid-monocarboxy isooctyl ester (MCOCH) was associated with a decrease of 27.43 g (95%CI: -54.81, -0.04) in birthweight. Trimester-specific analyses revealed stronger associations for third trimester DINCH exposure with birth outcomes. INTERPRETATION/CONCLUSIONS:In a large American cohort representing a broad range of backgrounds, we identified an inverse association of DINCH with standardized measures of birth size, but not gestational age. Further studies should examine joint exposures to DINCH, tere- and orthophthalates, and repeated measures across pregnancy.

Limited studies have explored the modifying role of neighborhood factors in the association between prenatal air pollution exposure and child neurodevelopment. We investigated whether the relationship between prenatal ambient air pollution and autism-related traits differed by neighborhood-level opportunity and vulnerability. This study drew data from 8598 mother-child pairs from 35 sites participating in the Environmental influences on Child Health Outcomes (ECHO) Cohort. Ambient air pollutants, including fine particulate matter (PM_2.5), nitrogen dioxide (NO₂), and ozone (O₃), were estimated at residential addresses during pregnancy. The Social Vulnerability Index (SVI) and Child Opportunity Index (COI) at the census tract level were linked to residential addresses at birth. The associations of pregnancy average and trimester-specific air pollution with autism-related traits, as measured by the Social Responsiveness Scale (SRS), stratified by SVI and COI quintiles were evaluated using generalized estimating equation models with a p-value <0.10 for the interaction term as suggestive evidence of signficance. There was evidence of interaction between prenatal exposure to O₃ with SRS by SVI (interaction p = 0.07) and by COI (interaction p = 0.01). Prenatal exposure to O₃ during pregnancy was associated with higher SRS score in the lowest SVI (β = 0.05, 95% CI: -0.005, 0.10) and highest COI (β = 0.04, 95% CI:0.00, 0.08). Differential associations of trimester-specific O₃ and PM_2.5 by SVI and COI were also observed (interaction p < 0.10), although stratified associations within each SVI and COI category were not statistically significant. Our findings suggest that neighborhood level social factors may modify the associaitons between prenatal air pollution and autism-related traits and warrant further consideration of neighborhood factors in environmental epidemiological studies.

BACKGROUND:Perinatal exposure to phthalates, parabens, and phenols-common endocrine disrupting chemicals (EDCs)-has been associated with adverse health outcomes, but the underlying biological mechanisms remain unclear. This study aims to explore the associations between perinatal EDC exposure--individually and as a mixture--and the neonatal metabolome, which provides a readout of metabolic processes and the factors that impact them. METHODS:EDCs and the metabolome were characterized in umbilical cord blood plasma from 175 pregnant participants recruited through the Columbia Center for Children's Environmental Health. Nontargeted screening revealed 15 EDCs using gas chromatography coupled to high resolution mass spectrometry (HRMS) and 5835 metabolic features using liquid chromatography HRMS. Correlated features were grouped into 21 modules using weighted gene co-expression network analysis and into integrative networks using xMWAS. Linear regression and quantile g-computation were used to assess associations between EDC exposures and individual metabolic feature abundance and the modules. RESULTS:We observed 2208 metabolome-wide associations between individual metabolic features and EDCs modeled individually and one association with the EDC mixture. Eight EDCs were individually associated with 7 modules, and the EDC mixture was associated with 1 module. Pathway enrichment analyses revealed phenol and phthalate exposures were associated with fatty acid and energy metabolism. xMWAS networks demonstrated similar EDC-module and EDC-metabolic feature relationships. CONCLUSION/CONCLUSIONS:Perinatal EDC exposure, individually and as a mixture, is associated with individual metabolic features and metabolic networks measured in cord plasma. These results could offer insight into the biological processes underpinning relationships between perinatal EDC exposure and adverse health outcomes.

Hypomineralized second primary molars (HSPM) affect approximately 6.8% of children. Molars affected by HSPM mineralize between 20 wk of gestation and 15 to 17 mo of age. Experimental studies suggest that exposure to endocrine-disrupting chemicals, including di-2-ethylhexyl phthalate (DEHP) and bisphenol A (BPA), may disrupt enamel formation, but evidence in humans remains limited. This study investigates the association between prenatal exposure to individual phthalates, BPA, and their mixtures and the odds of HSPM in children. It was embedded in the Generation R study, a population-based prospective birth cohort in the Netherlands. Maternal urinary concentrations of BPA and 11 phthalate metabolites were measured each trimester. HSPM was diagnosed at age 6 y based on European Academy of Paediatric Dentistry criteria. Multivariate logistic regression models and Bayesian weighted quantile sum regression were used to assess individual and mixture associations, respectively. Among 1,160 mother-child pairs, 10.2% of children were diagnosed with HSPM. In single-exposure models, higher prenatal concentrations of monobenzyl phthalate (mBzBP) were associated with lower odds of HSPM (odds ratio [OR] = 0.76; 95% confidence interval [CI]: 0.59 to 0.97). Second-trimester exposure to DEHP metabolites showed a nonsignificant trend toward increased HSPM odds (eg, mEOHP OR = 1.15; 95% confidence interval [CI]: 0.91 to 1.45). Sex-specific associations were observed: higher BPA and phthalate concentrations were associated with lower likelihood of HSPM in girls (eg, BPA OR = 0.70; 95% CI: 0.50 to 0.98) and higher odds in boys (p-interaction < 0.10). Mixture analyses showed a nonsignificant inverse association for overall prenatal exposure (OR = 0.84; 95% CI: 0.63 to 1.15), with mBzBP contributing most to the mixture. No overall association was observed for DEHP mixture, although sex-specific trends were consistent with single-exposure results. Prenatal exposure to phthalates and BPA may influence HSPM development. These findings highlight the need to consider early-life endocrine-disrupting chemical mixture exposure in HSPM etiology.

While maternal histories of adverse childhood experiences (ACEs) have been linked to altered immune signaling during pregnancy, the specific inflammatory pathways underlying these associations remain poorly characterized. We leveraged high-dimensional proteomics to examine associations between maternal ACEs and pregnancy-associated inflammatory signaling pathways defined by the Kyoto Encyclopedia of Genes and Genomes (KEGG) and tested whether these associations were moderated by neighborhood-level social vulnerability during pregnancy. Participants included pregnant individuals enrolled in the Environmental influences on Child Health Outcomes (ECHO) Cohort (N = 770). Circulating inflammatory proteins were assayed from maternal plasma using the Olink Target 96 Inflammation panel. Protein markers were mapped to Human Genome Organisation (HUGO) Gene Nomenclature Committee gene symbols and aggregated into KEGG pathway scores using curated definitions. Linear regression models evaluated associations between continuous maternal ACE scores and each pathway score. Following correction for multiple comparisons, higher maternal ACE exposure was associated with higher activity across 12 KEGG-annotated protein pathways, including AGE-RAGE, non-alcoholic fatty liver disease, chemokine, Th17 differentiation, HIF-1, cellular senescence, Rap1, NOD-like receptor signaling, transcriptional misregulation in cancer, and infection-related pathway signaling. Domain-specific analyses identified the most consistent associations for childhood household dysfunction. Neighborhood-level social vulnerability during pregnancy did not moderate ACE-inflammation associations but was independently associated with higher KSHV infection and NOD-like receptor signaling pathway scores. Together, these findings indicate that maternal ACEs and prenatal social vulnerability are independently associated with small, but directionally consistent, differences in overlapping inflammatory and growth-related protein signals during pregnancy.

BACKGROUND/UNASSIGNED:Pregnancy requires finely tuned immune changes that support implantation, placental development, maternal-fetal tolerance, and preparation for labor, yet the normative trajectories of circulating inflammatory proteins across gestation remain poorly defined. This cross-sectional study investigates how circulating inflammatory proteins vary with gestational age in pregnancy and examines the impacts of fundamental biological characteristics, such as gravidity and fetal sex. METHODS/UNASSIGNED:Data were drawn from 1154 pregnant individuals from six study sites of the National Institutes of Health Environmental influences on Child Health Outcomes (ECHO) Cohort. We used Olink high-throughput proteomic profiling to map cross-sectional associations between protein expression levels and gestational age at blood draw using linear, spline-based, and generalized additive modeling approaches. RESULTS/UNASSIGNED:Generalized additive models provided the best fit, revealing that immune changes across pregnancy were predominantly nonlinear. Sixty-one proteins showed significant associations with gestational age, with many exhibiting shared inflection points that aligned with major physiological transitions. A small subset of proteins also showed evidence of modification by fetal and maternal characteristics. CD244 displayed different gestational patterns by fetal sex, while CST5 and SIRT2 showed varied gestational associations by maternal gravidity. CONCLUSION/UNASSIGNED:The findings highlight pregnancy as a sequence of coordinated immune transitions rather than a simple linear shift and provide one of the most detailed characterizations to date of circulating inflammatory protein dynamics across human gestation. Establishing these normative trajectories offers a crucial reference for detecting early deviations that may signal risk for pregnancy complications and for identifying biomarkers in maternal and fetal health research.

BACKGROUND:Studies of prenatal exposure to non-persistent chemicals often assume biomarker measurements from a single urine sample can represent exposure pregnancy. The validity of this assumption depends on biomarker concentration variability across gestation, which is not well-characterized for many chemicals. OBJECTIVE:To assess the intra-individual variability of non-persistent chemical concentrations measured in 3 pregnancy urine samples. METHODS:We measured 154 analytes from 12 classes in 3 urine samples collected across gestation from 90 participants in 3 Environmental influences on Child Health Outcomes Cohort sites. For each cohort site, we estimated intraclass correlation coefficients (ICCs) of 58 analytes with >60% detection using log2-transformed concentrations of specific gravity (SG)-standardized values. We classified analyte ICCs by the site-specific median relative to the limits of detection (LODs) to evaluate the influence of exposure levels on reliability estimates. As secondary analyses, we calculated kappa statistics for 56 analytes with 20-60% detection to assess agreement of detection frequencies across the 3 timepoints. RESULTS:ICCs (range: -0.12 to 0.91) varied by cohort site, except for benzophenone-1 and propyl paraben (consistently >0.5) and bisphenol F (consistently <0.1). We observed site-specific trends for parabens, benzophenones, antimicrobials, phthalates, polycyclic aromatic hydrocarbons, and aromatic amines. Analytes with median:LOD ratios >50 were more stable throughout pregnancy (e.g., benzophenones, propyl paraben, antimicrobials, and naphthalenes), except for some phthalate metabolites, which may indicate high sporadic exposures. Analytes with median:LOD ratios <10 had lower ICCs, except for benzophenone-8, which may indicate low-level chronic exposure. Kappa statistics of the 56 analytes with 20-60% detection ranged from -0.22 to 0.53. SIGNIFICANCE/CONCLUSIONS:ICCs varied by cohort site. Lower ICCs were generally observed for analytes with lower concentrations. For the majority of non-persistent chemicals, a single pregnancy measure may better represent exposure near the time of sample collection rather than for the entire pregnancy. IMPACT/CONCLUSIONS:Understanding the intra-individual variability of urinary non-persistent chemical biomarkers across multiple pregnancy timepoints is important to inform study design, statistical analysis, and interpretation in future studies.

OBJECTIVE:To examine the association between prenatal exposure to a mixture of EDCs and attention-deficit/hyperactivity disorder (ADHD) symptoms in children. METHODS:Participating were mother-child pairs from the Environmental influences on Child Health Outcomes (ECHO) Cohort (n = 3,962, 18 Sites, 2001-2021). Maternal spot urine samples collected during pregnancy were sent to the Wadsworth Center Human Health Exposure Analysis Resource laboratory and analyzed using a single assay workflow. From those samples, 24 urinary metabolites of EDCs with detection frequency > 70%, were grouped as molar sum of high- and low-molecular-weight phthalates, polycyclic aromatic hydrocarbons, bisphenol S, organophosphate esters, and organophosphate and pyrethroid pesticides. Child ADHD symptoms were assessed using raw scores from the Child Behavior Checklist for preschool-aged (CBCL/1½-5) and school-aged (CBCL/6-18) Attention-Deficit/Hyperactivity Problems subscale. Scores ≥ the 85th and 95th percentiles were respectively classified as borderline and clinical thresholds. Quantile g-computation estimated the expected change in ADHD outcome associated with a one-quartile increase in all prenatal EDC mixture concentrations, using negative binomial models for raw scores and binomial models for percentile thresholds. RESULTS:Prenatal exposure to the EDC mixture was not associated with child ADHD symptoms in preschool-aged or school-aged children in adjusted models for either continuous raw scores or percentile thresholds. CONCLUSION/CONCLUSIONS:Overall our findings on prenatal EDC mixture exposure and ADHD symptoms were not robust to full adjustment. Future research that can delineate factors such as windows of vulnerability and trajectories of ADHD-related symptoms may help clarify mixed findings in the literature.

Aside from specific occupational exposures, ingestion is believed to be the primary route of exposure to PFAS, a class of forever chemicals known for their ability to repel oil and water, reduce friction, and resist high temperatures. The goal was to explore the geographic distribution of PFAS water contamination versus human exposure and identify dietary factors that may explain high PFAS levels among rural residents. We performed PFAS analyses among 160 rural participants, who provided blood samples and survey responses in our chemical exposure cross-sectional study. We mapped the geographic distribution of PFAS contamination based on water testing versus human samples results. We also used health surveys and food frequency questionnaire data to identify factors associated with high PFAS levels. The geographic mismatch between the distribution of PFAS water contamination and human exposure highlights the need to assess other potential sources, which identified potential dietary sources of these chemical hazards.