Heat stress has been implicated as a risk factor for cardiovascular disease (CVD). Wet bulb globe temperature (WBGT) is considered a more physiologically relevant measure of heat stress than temperature alone, accounting for direct sunlight, relative humidity, solar radiation, and wind speed. Pregnant individuals may be particularly vulnerable to heat stress due to physiological changes in thermoregulation. Pregnancy may serve as a sensitive period influencing future health, but it is unknown if prenatal heat stress impacts postpartum cardiovascular health. We examined associations between prenatal, residential WBGT exposure and postpartum Pathobiologic Determinants of Atherosclerosis in Youth (PDAY) risk scores, which predict future atherosclerosis. Among 193 participants from the Maternal and Developmental Risks from Environmental and Social Stressors pregnancy cohort, annual postpartum PDAY scores were calculated across 6 years postpartum, using age, non-high-density lipoprotein cholesterol, high-density lipoprotein cholesterol, hypertension, obesity, and hyperglycemia from in-person questionnaires and biomarker samples. We used hierarchical generalized additive models (GAMs) to examine associations between pregnancy average and trimester-specific WBGT and postpartum PDAY scores. Effect modification by prenatal psychosocial stressors, urban heat island index (UHII), and climate vulnerability index (CVI) was assessed through stratification. Distributed lag non-linear models (DLNM) with mixed effects assessed weekly associations. Models were adjusted for maternal age at consent, pre-pregnancy body mass index, household income, maternal education, race/ethnicity, and conception season. Model fit indices indicated a nonlinear relationship between average prenatal WBGT and postpartum PDAY scores, with a positive association emerging at temperatures above 19 °C, though confidence intervals were wide. We were unable to observe evidence of associations with trimester-specific WBGT, sensitive windows of exposure, or effect modification by psychosocial stressors, UHII, or CVI. Our data suggested an association at pregnancy average WBGT greater than 19 °C with postpartum cardiometabolic health risk scores, however sparsity at temperature extremes limited our conclusions. The relatively limited exposure variability in our study region and small sample size may have limited our ability to observe more robust associations.

We evaluated whether maternal HSP27 and HSP70 responded to heat stress during pregnancy and whether HSPs were associated with preterm birth in a secondary analysis of a prospective cohort with repeated biomarkers (N = 227 from MADRES). Serum HSPs were measured during early and late pregnancy using ELISA. Daily heat stress in the 30 days before HSP measurement was computed as Daily Maximum Heat Index (DMHI) and Wet Bulb Globe Temperature (WBGT). Distributed lag models identified sensitive windows of heat adaptation; logistic regression models assessed HSPs' effects on preterm birth (<37 gestational weeks, n = 18 [7.9%]). Models were adjusted for sociodemographics, study design, and heat-exposure predictors. HSP27 and HSP70 concentrations decreased over pregnancy. Higher acute DMHI exposure was associated with higher HSP27 and HSP70 during early pregnancy (1-15 days lags; strongest effect per 1-SD increase: 0.6% [95% CI: 0.001-1.2%] for HSP27 at 9 days; 8.3% [2.6-14.5%] for HSP70 at 1 day) and higher HSP27 during late pregnancy. Mothers with larger HSP70 increases over pregnancy were associated with higher preterm birth risks (OR = 5.2, [1.5, 18.6]) versus those with stable/decreasing levels. HSPs may indicate acute heat stress during pregnancy, with increased HSP70 associated with preterm birth risks.

Poor pregnancy sleep quality has been linked to adverse pregnancy and birth outcomes, particularly in late pregnancy when sleep symptoms worsen. We aimed to identify potential important predictors of late-pregnancy sleep quality from social and environmental stressors and pregnancy health conditions, using a machine-learning approach. In the MADRES cohort, 687 mothers reported late-pregnancy sleep quality (Jenkins Sleep Scale [JSS]), with 26.3% reporting poor sleep quality (JSS≥12). Gradient boosting models were fitted using 59 predictors from 11 predictor groups, or each group separately, with relative influence analysis to identify the top predictors. Model stability of selected hyperparameters was assessed across 30 random train-test splits (80/20; n = 550/137), with 3-fold cross-validation; the optimal model achieved a cross-validation AUC of 0.647 and a test AUC of 0.704. Mental health during late pregnancy was the single greatest contributor to poor sleep quality, followed by sociodemographic and noise indicators (range of AUC: 0.53-0.72). Additional environmental exposures were also identified as top 10 predictors (e.g., ozone and organophosphate esters). We revealed that mental health interacted with environmental factors; mothers had an even higher risk of poor sleep quality if they had both higher-than-average mental health scores and environmental exposure. Late-pregnancy sleep quality was influenced not only by traditionally identified personal-level factors but also by a complex interplay with environmental exposures such as noise, air pollution, and chemical exposures.

Accurate prediction of atmospheric air pollutants is critical for public health protection and environmental management. Traditional machine learning (ML) methods achieve high spatial resolution but lack physicochemical constraints, leading to systematic biases that compromise exposure estimates for epidemiological studies. Chemical transport models incorporate atmospheric physics but require expensive parameterization and often fail to capture local-scale variability crucial for health impact assessment. This gap between data-driven accuracy and physical realism presents a major obstacle to advancing air quality science. We address this challenge through a novel physics-informed deep learning framework that integrates advection-diffusion equations and fluid dynamics constraints directly into neural network architectures for multi-pollutant prediction. Our approach models air pollutant pairs across geographically distinct domains (NO2/NOx for California; PM2.5/PM10 for mainland China), providing a comprehensive framework for physics-constrained atmospheric modeling at high resolution. Through an efficient framework, our methodology demonstrates that incorporating proxy advection and diffusion fields as physical constraints fundamentally alters learning dynamics, reducing generalization error and eliminating systematic bias inherent in data-driven approaches while improving computational efficiency compared to graph networks. Site-based validation reveals unprecedented bias reduction: 21%-42% for nitrogen oxides and 16%-17% for particulate matter compared to the baseline deep learning methods. Our methodology uniquely generates physically interpretable parameters while providing explicit uncertainty quantification through ensemble techniques. The substantial bias reduction coupled with physically interpretable parameters has immediate implications for improving air pollutant exposure assessment and understanding in epidemiological research, potentially transforming health effect evaluations that rely on accurate spatial predictions.

We investigated associations between preconception and prenatal heat stress and wildfire (WF) smoke exposures on adverse birth outcomes and whether neighborhood climate vulnerability is an effect modifier in the Maternal And Developmental Risks from Environmental and Social stressors cohort (N = 713). Generalized linear models were fit to test the association between exposures and small-for-gestational-age (SGA), low birthweight (LBW), and Fenton growth z-score outcomes, adjusting for confounders. Living in a high climate vulnerability index neighborhood was tested as an effect modifier. During preconception, increases in heat stress and WF measures were associated with higher odds of SGA. Living in the most climate-vulnerable neighborhoods during preconception significantly modified and nearly doubled the odds of SGA with exposure to heat stress. Similarly, heat stress and WF exposure in trimester-specific time periods were associated with adverse birth outcomes. Conversely, third-trimester exposures were associated with lower odds of LBW. Throughout pregnancy, two measures of infant size (SGA and Fenton z-scores) were lower among those with greater exposure to multiple WF exposures. This study highlights how living in more climate-vulnerable neighborhoods significantly modifies the effect of heat stress on SGA, suggesting that the increasing adaptation capacity of communities may strengthen climate change resilience.

Exposure to air pollution during pregnancy that disrupts thyroid function can lead to adverse health outcomes in mother and child. We evaluated the overall effect and critical exposure window of residential ambient air pollution exposures on thyroid function in the MADRES pregnancy cohort. We also investigated whether these associations varied by iodine deficiency status and neighborhood deprivation. Early pregnancy (6-20 weeks) serum thyroid stimulating hormone (TSH) and free thyroxine (FT4) were measured for 217 mothers. Daily residential ambient air pollution exposures (PM2.5, PM10, NO2 and O3 8hr max) were estimated using inverse-distance squared spatial interpolation from regulatory monitors. We used linear regression to assess effects of single ambient air pollutants on thyroid function, including exploration of effect modification by iodine deficiency and neighborhood deprivation (Area Deprivation Index and Gini Index of income inequality, dichotomized at the median). Distributed lag models (DLM) were used to assess critical windows of exposure for ambient air pollutants from 12 weeks preconception to first trimester. We found that one SD increase in PM2.5 (2.4 μg/m3) and PM10 (5.8 μg/m3) were associated with 18.9 % (95 % CI: 2.7, 37.8 %) and 16.8 % (95 % CI: 0.7, 35.6 %) higher TSH levels, respectively, with significant windows of susceptibility in the first trimester (GW 5-8 or 6-8). These associations were also modified by neighborhood deprivation, and iodine status. Our findings indicate that relatively low levels of PM exposures in early pregnancy are associated with increased TSH levels particularly among women with replete iodine levels and women living in neighborhoods with greater deprivation.

While extreme temperatures have been linked to adverse mental health outcomes, the impact of prenatal heat stress on postpartum depression remains understudied. We evaluated associations between prenatal heat stress and depressive symptoms one year postpartum, identified susceptible gestational windows, and assessed whether neighborhood climate vulnerability modifies maternal susceptibility to heat-related postpartum depression. We included 275 predominantly low-income Hispanic/Latina participants in the MADRES cohort. Residential wet bulb globe temperature (WBGT), is more reflective of physiological responses to heat stress compared to temperature alone, and was estimated throughout pregnancy at residential addresses. Depressive symptoms at 12 months postpartum were assessed using the Center for Epidemiologic Studies-Depression (CES-D) scale. The Climate Vulnerability Index (CVI) was linked at census tract level. We used multivariable linear regression to estimate associations between trimester-specific heat stress and log-transformed CES-D scores and examined effect modification by CVI (≥75th vs. <75th percentile). Distributed lag models (DLM) were used to identify susceptible exposure windows to WBGT, and Bayesian distributed lag interaction models (BDLIM) were used to assess differences in associations by CVI. We found that each interquartile range increase (IQR) in first-trimester WBGT was associated with a 18.93 % (95 % CI: 0.03 %-41.75 %) increase in postpartum CES-D scores. Neighborhood-level CVI modified second-trimester effects of heat stress on postpartum depressive symptoms, with participants in high-CVI neighborhoods experiencing stronger positive associations (31.21 %, 95 % CI: -11.80 %-95.21 %) compared to those in low-CVI neighborhoods (-16.70 %, 95 % CI: -32.29 %-3.55 %) for WBGT. These findings provide novel evidence linking early-pregnancy heat exposure to postpartum depression risk, with effects amplified in climate-vulnerable neighborhoods during mid-pregnancy. Our results suggest that heat mitigation strategies, including providing heat-health education during prenatal visits, may be important for postpartum mental health.

Past studies support the hypothesis that the prenatal period influences childhood growth. However, few studies explore the joint effects of exposures that occur simultaneously during pregnancy. To explore the feasibility of using mixtures methods with neighborhood-level environmental exposures, we assessed the effects of multiple prenatal exposures on body mass index (BMI) from birth to age 24 months. We used data from two cohorts: Healthy Start (n = 977) and Maternal and Developmental Risks from Environmental and Social Stressors (MADRES; n = 303). BMI was measured at delivery and 6, 12, and 24 months and standardized as z-scores. We included variables for air pollutants, built and natural environments, food access, and neighborhood socioeconomic status (SES). We used two complementary statistical approaches: single-exposure linear regression and quantile-based g-computation. Models were fit separately for each cohort and time point and were adjusted for relevant covariates. Single-exposure models identified negative associations between NO2 and distance to parks and positive associations between low neighborhood SES and BMI z-scores for Healthy Start participants; for MADRES participants, we observed negative associations between O3 and distance to parks and BMI z-scores. G-computations models produced comparable results for each cohort: higher exposures were generally associated with lower BMI, although results were not significant. Results from the g-computation models, which do not require a priori knowledge of the direction of associations, indicated that the direction of associations between mixture components and BMI varied by cohort and time point. Our study highlights challenges in assessing mixtures effects at the neighborhood level and in harmonizing exposure data across cohorts. For example, geospatial data of neighborhood-level exposures may not fully capture the qualities that might influence health behavior. Studies aiming to harmonize geospatial data from different geographical regions should consider contextual factors when operationalizing exposure variables.

Prenatal exposures are associated with childhood asthma, and risk may increase with simultaneous exposures. Pregnant women living in lower-income communities tend to have elevated exposures to a range of potential asthma risk factors, which may interact in complex ways. We examined the association between prenatal exposures and the risk of childhood acute-care clinical encounters for asthma (hospitalizations, emergency department visits, observational stays) using conditional logistic regression with a multivariable smoothing term to model the interaction between continuous variables, adjusted for maternal characteristics and stratified by sex. All births near the New Bedford Harbor (NBH) Superfund site (2000-2006) in New Bedford, Massachusetts, were followed through 2011 using the Massachusetts Pregnancy to Early Life Longitudinal (PELL) Data System to identify children aged 5-11 years with acute-care clinical asthma encounters (265 cases among 7787 children with follow-up). Hazard ratios (HRs) were higher for children living closer to the NBH site with higher umbilical cord blood lead levels than in children living further away from the NBH site with lower lead levels (P <.001). HRs were higher for girls (HR = 4.17; 95% CI, 3.60-4.82) than for boys (HR = 1.72; 95% CI, 1.46-2.02). Our results suggest that prenatal lead exposure in combination with residential proximity to the NBH Superfund site is associated with childhood asthma acute-care clinical encounters. This article is part of a Special Collection on Environmental Epidemiology.

Atmospheric nitrogen oxides (NOx) primarily from fuel combustion have recognized acute and chronic health and environmental effects. Machine learning (ML) methods have significantly enhanced our capacity to predict NOx concentrations at ground-level with high spatiotemporal resolution but may suffer from high estimation bias since they lack physical and chemical knowledge about air pollution dynamics. Chemical transport models (CTMs) leverage this knowledge; however, accurate predictions of ground-level concentrations typically necessitate extensive post-calibration. Here, we present a physics-informed deep learning framework that encodes advection-diffusion mechanisms and fluid dynamics constraints to jointly predict NO2 and NOx and reduce ML model bias by 21-42%. Our approach captures fine-scale transport of NO2 and NOx, generates robust spatial extrapolation, and provides explicit uncertainty estimation. The framework fuses knowledge-driven physicochemical principles of CTMs with the predictive power of ML for air quality exposure, health, and policy applications. Our approach offers significant improvements over purely data-driven ML methods and has unprecedented bias reduction in joint NO2 and NOx prediction.

Children’s prenatal exposure to multiple environmental chemicals may contribute to subsequent deficits in impulse control, predisposing them to risk-taking. Our goal was to investigate associations between prenatal exposure mixtures and risk of teen birth, a manifestation of high-risk sexual activity, among 5865 girls (1st generation) born in southeast Massachusetts from 1992–1998. Exposures included prenatal modeled polychlorinated biphenyls (PCBs), ρ,ρ′-dichlorodiphenyl dichloroethylene (DDE), hexachlorobenzene (HCB), lead (Pb), and mercury (Hg). We fit adjusted generalized additive models with multivariable smooths of exposure mixtures, 1st generation infant’s birth year, and maternal age at 1st generation birth. Predicted odds ratios (ORs) for teen birth were mapped as a function of joint exposures. We also conducted sensitivity analyses among 1st generation girls with measured exposure biomarkers (n = 371). The highest teen birth risk was associated with a mixture of high prenatal HCB, Hg, Pb, and PCB, but low DDE exposure, with similar associations in sensitivity analyses. The highest OR predicted for girls born in 1995 to mothers of median age (26 years) was at the 95th percentile of the HCB and PCB exposure distributions (OR = 3.09; 95% confidence interval: 0.29, 32.4). Additionally, girls born earlier in the study period or to teen mothers were at increased risk of teen birth. Prenatal environmental chemical exposures and sociodemographic characteristics may interact to substantially increase risk of teen births.

Background: Children born near New Bedford, Massachusetts, have been prenatally exposed to multiple environmental chemicals, in part due to an older housing stock, maternal diet, and proximity to the New Bedford Harbor (NBH) Superfund site. Chemical exposure measures are not available for all births, limiting epidemiologic investigations and potential interventions. Objective: We linked biomonitoring data from the New Bedford Cohort (NBC) and birth record data to predict prenatal exposures for all contemporaneous area births. Methods: We used prenatal exposure biomarker data from the NBC, a population-based cohort of 788 mother–infant pairs born during 1993–1998 to mothers living near the NBH, linked to their corresponding Massachusetts birth record data, to build predictive models for cord serum polychlorinated biphenyls (expressed as a sum, ΣPCBs), p,p′-dichlorodiphenyl dichloroethylene (DDE), hexachlorobenzene (HCB), cord blood lead (Pb), and maternal hair mercury (Hg). We applied the best fit models (highest pseudo R2), with multivariable smooths of continuous variables, to predict exposure biomarkers for all 10,270 births during 1993–1998 around the NBH. We used 10-fold cross validation to validate the exposure models and the bootstrap method to characterize sampling variability in the exposure predictions. Results: The 10-fold cross-validated R2 for the ΣPCBs, DDE, HCB, Pb, and Hg exposure models were 0.54, 0.40, 0.34, 0.46, and 0.40, respectively. For each exposure model, multivariable smooths of continuous variables improved the fit compared with linear models. Other variables with significant effects on exposure estimates were paternal education, maternal race/ethnicity, and maternal ancestry. The resulting exposure predictions for all births had variability consistent with the NBC measured exposures. Conclusions: Predictive models using multivariable smoothing explained reasonable amounts of variance in prenatal exposure biomarkers. Our analyses suggest that prenatal chemical exposures can be predicted for all contemporaneous births in the same geographic area by modeling available biomarker data for a subset of that population. https://doi.org/10.1289/EHP4849

After publication of the article [1], it was brought to our attention that a number in Table 1 is incorrect.

Introduction Exposures to various environmental contaminants have been associated with increased asthma risk among children. Prenatal exposures can trigger changes to the immune system in the developing fetus which can lead to an onset of childhood allergies including asthma. Children exposed to multiple environmental stressors may be more susceptible. We investigated the association between asthma risk and mixtures of environmental stressors among children born 2000-2005 to mothers living near a PCB-contaminated Superfund Site in New Bedford Harbor, Massachusetts. Methods We identified 10,517 births from four towns surrounding the Superfund Site who were followed until 2010. There were 428 asthma cases identified from emergency department and hospital discharge records. We examined the joint effects of modeled PCBs, DDE, distance to road, and maternal age on asthma risk by smoothing multiple exposures using generalized additive models (GAMs) adjusted for covariates including maternal education, prenatal smoking and alcohol use. Advantages of this approach include the ability to smooth variables of different units, generalizability to multiple covariates, and an overall test of the importance of the mixture. Results DDE, distance to road, and maternal age at birth were included in a multi-dimensional smooth, and the model was further adjusted parametrically for confounders. Results suggest that children of younger mothers who were also exposed to the highest levels of DDE and lived closest to major roads were at greatest risk of developing asthma (OR= 1.6). This distance to road and DDE pattern was not apparent for older mothers. PCBs, which were highly correlated with DDE, produced similar results in separate models. Conclusions This example illustrates the utility of using GAMs for visualizing mixtures; the importance of both chemical and non-chemical stressors in determining asthma risk would not likely have been discovered using more traditional epidemiologic models.

Arsenic exposure has been implicated as a risk factor for cardiovascular diseases, metabolic disorders, and cancer, yet the role mitochondrial dysfunction plays in the cellular mechanisms of pathology is largely unknown. To investigate arsenic-induced mitochondrial dysfunction in vascular smooth muscle cells (VSMCs), we exposed rat aortic smooth muscle cells (A7r5) to inorganic arsenic (iAs(III)) and its metabolite monomethylarsonous acid (MMA(III)) and compared their effects on mitochondrial function and oxidative stress. Our results indicate that MMA(III) is significantly more toxic to mitochondria than iAs(III). Exposure of VSMCs to MMA(III), but not iAs(III), significantly decreased basal and maximal oxygen consumption rates and concomitantly increased compensatory extracellular acidification rates, a proxy for glycolysis. Treatment with MMA(III) significantly increased hydrogen peroxide and superoxide levels compared to iAs(III). Exposure to MMA(III) resulted in significant decreases in mitochondrial ATP, aberrant perinuclear clustering of mitochondria, and decreased mitochondrial content. Mechanistically, we observed that mitochondrial superoxide and hydrogen peroxide contribute to mitochondrial toxicity, as treatment of cells with MnTBAP (a mitochondrial superoxide dismutase mimetic) and catalase significantly reduced mitochondrial respiration deficits and cell death induced by both arsenic compounds. Overall, our data demonstrates that MMA(III) is a mitochondria-specific toxicant that elevates mitochondrial and non-mitochondrial sources of ROS.

BackgroundArsenic (As) contamination of ground‐ and well‐water is common worldwide. Chronic As exposure can produce toxic effects including atherosclerosis and hypertension, yet mechanisms by which both inorganic As and As metabolites induce these effects are not well elucidated. Both arsenite (iAs3+) and its metabolite monomethylarsonous acid (MMAs3+) are believed to affect the activity of critical Vascular Smooth Muscle (VSM) components including the ‘L‐type’ calcium channel (LTCC), a key regulator of VSM tone, as well as the large‐conductance calcium‐activated potassium channel (BKCa), a known modulator of cell depolarization implicated in the development of hypertension and restenosis disease.Objectives and MethodsThis study examines effects of iAs3+ and MMAs3+ on expression and related activities of LTCC and BKCa in acutely isolated and primary / tissue cultured rat thoracic aorta, as well as the A7r5 rat thoracic aorta VSM cell line, using whole‐cell patch clamp, vascular contractility, RT‐QPCR, and light microscopy.ResultsInitial results indicate alterations in VSM cell morphology, viability, LTCC activity, and responsiveness to PE‐induced vasoconstriction on acute and subchronic exposure to both iAs3+ and MMA3+.ConclusionsBoth iAs3+ and MMA3+ affect activities of key effectors governing the maintenance of VSM tone. This work was supported by the Nevada INBRE and McNair Programs.