Epidemiologists increasingly estimate associations between mixtures of correlated co-exposures and health outcomes. We propose an exploratory approach to help differentiate co-exposure confounding from co-exposure amplification of bias (CAB) for studies considering correlated co-exposures using a case example and simulations based on observed relationships. Our example uses associations observed in published studies between co-occurring per- and polyfluoroalkyl substance (PFAS) biomarkers and tetanus antibody concentrations but could apply to any context where statistical modeling is performed with and without inclusion of correlated co-exposures. Using directed acyclic graphs (DAGs), to define an assumed causal structure, and simulations of three PFAS predicting antibody titers, we show that if researchers can assume one component in the exposure mixture is correlated with the other components, but does not cause the outcome, then CAB can be detected by testing whether the estimate for the non-causal component in a multi-pollutant model is different from zero. In our simulation, the estimate for the non-causal component was zero (95%CI, -0.22, 0.25) when CAB was absent and -0.25 (95%CI, -0.48, 0.00) when present. These results illustrate a possible method to identify CAB and select statistical models that may produce less biased effect estimates if our assumed DAG is correct.

BACKGROUND:Due to transplacental exposure to perfluoroundecanoic acid (PFUnDA), there is a need to evaluate potential developmental effects in humans. We examined the association between PFUnDA exposure and birth weight (BW) using systematic review methodology. METHODS:We used a random effects model with inverse-variance weighting to meta-analyze associations between PFUnDA exposure biomarkers and changes in BW. We report a pooled estimate and stratified estimates based on factors including study confidence and timing of sample collection. RESULTS:We detected a pooled mean BW deficit (β) of 15.3 g (95%CI: -23.4, -7.1) per ln(ng/mL) PFUnDA increase in our primary analysis of 28 studies. We observed deficits in 5 low (β = -26.3 g, 95%CI: -56.5, 3.8), 14 medium (β = -16.5 g, 95%CI: -26.8, -6.1), and 9 high confidence studies (β = -10.3 g, 95%CI: -24.8, 4.1). The magnitude of the pooled BW deficit varied across sample timing categories, with the largest average deficit in eight studies with postpartum samples (β = -21.0 g, 95%CI: -43.8, 1.8) and the smallest among eight studies with early pregnancy samples (β = -6.8 g, 95%CI: -21.5, 7.9). CONCLUSIONS:We observed lower BW associated with higher PFUnDA exposure. Although confidence intervals overlapped across strata, we detected smaller BW deficits in higher quality studies and larger deficits among studies that collected samples in late pregnancy or postpartum. These trends may be partly explained by bias from pregnancy hemodynamics. Future studies conducted in populations with serial pregnancy PFUnDA measures would facilitate data interpretation and address some methodological challenges.

We used toxicologically informed relative potency factors (RPFs) to compare RPF-weighted versus unweighted epidemiologic analyses of five birth defects (atrial septal defect [ASD], ventricular septal defect [VSD], tetralogy of Fallot [TOF], cleft palate without cleft lip, and obstructive genitourinary defects [OGD]) and various disinfection byproduct (DBP) mixtures in drinking water. We derived RPFs for four trihalomethanes (THMs) and nine haloacetic acids (HAAs) from published animal developmental bioassays using benchmark dose models at 10% responses, estimated adjusted odds ratios (aORs) for RPF-weighted quintile 5 (vs 1) DBP mixture sums, and compared the RPF-weighted and unweighted aORs. RPFs were generally higher for brominated DBP species (e.g., HAA RPF10s from mouse whole embryo culture dysmorphogenesis data: monobromoacetic acid [MBAA] = 1.00, trichloroacetic acid [TCAA] = 0.0011, and dichloroacetic acid [DCAA] = 0.0005), which occurred at lower concentrations than chlorinated species. For THM mixtures, aORs changed <10% with RPF weighting, except for TOF (+19%) for the sum of brominated THMs (tribromomethane + bromodichloromethane + dibromochloromethane). We detected >10% aOR changes for HAA3 (TCAA + DCAA + dibromoacetic acid) with TOF and OGD; HAA5 (HAA3 + monochloroacetic acid + MBAA) with ASD, TOF, and OGD; HAA6 (all brominated HAAs) with TOF; and HAA9 (HAA5 + tribromoacetic acid + bromochloroacetic acid + bromodichloroacetic acid + dibromochloroacetic acid) with OGD. The results suggest that unweighted DBP metrics may reasonably estimate risks of regulated DBPs, but this should be examined elsewhere based on individual-level estimates, which should be prone to less exposure measurement error.

INTRODUCTION:Haloacetic acids (HAAs) are water disinfection byproducts (DBPs) regulated as a mixture of five species (HAA5) in the United States and Canada. To date, two brominated HAAs (BrHAAs) in HAA5 (monobromoacetic acid [MBAA], dibromoacetic acid [DBAA]) have been associated with birth defects in some epidemiologic studies, but the other four unregulated BrHAAs remain understudied. METHODS:We analyzed registry-based case-control data on 16 birth defect phenotypes in relation to temporally weighted first-trimester exposures to the four unregulated BrHAAs (tribromoacetic acid [TBAA], bromochloroacetic acid [BCAA], bromodichloroacetic acid [BDCAA], chlorodibromoacetic acid [CDBAA]) and the sum of all six BrHAAs (HAA6 = TBAA + BCAA + BDCAA + CDBAA + MBAA + DBAA). We matched cases to controls 1:10 on the week of conception and estimated adjusted odds ratios (aORs) via conditional logistic regression. RESULTS:We observed some elevated aORs for dichotomized BrHAA exposures with certain outcomes, such as cleft palate alone (BDCAA aOR = 1.75 [95% confidence interval: 1.15, 2.66]), ventricular septal defects (BDCAA aOR = 1.28 [0.97, 1.68]), tetralogy of Fallot (BDCAA OR = 1.57 [0.93, 2.64]), and obstructive genitourinary defects (CDBAA aOR = 1.65 [1.07, 2.53]), and reduced aORs for hypospadias (e.g., BCAA aOR = 0.58 [0.40, 0.84]). Most other associations were closer to the null, and many lacked precision. CONCLUSION:Our observations warrant further investigation given their novelty and the paucity of data on health impacts of prenatal BrHAA exposures overall; ours is the first epidemiological study to investigate most of these exposure-outcome relationships. Future work would benefit from a longer study period to ascertain additional birth defect cases and more direct exposure assessment in areas served by water systems with higher bromide levels.

We examined the association between mean birth weight (BW) differences and perfluorohexane sulfonate (PFHxS) exposure biomarkers.We fit a random effects model to estimate the overall pooled effect and for different strata based on biomarker sample timing and overall study confidence. We also conducted an analysis to examine the impact of a continuous measure of gestational age sample timing on the overall pooled effect.We detected a −7.9 g (95% CI −15.0 to –0.7; pQ=0.85; I2=0%) BW decrease per ln ng/mL PFHxS increase based on 27 studies. The 11 medium confidence studies (β=−10.0 g; 95% CI −21.1 to 1.1) showed larger deficits than 12 high (β=−6.8 g; 95% CI −16.3 to 2.8) and 4 low confidence studies (β=−1.5 g; 95% CI −51.6 to 48.7). 10 studies with mid-pregnancy to late-pregnancy sampling periods showed smaller deficits (β=−3.9 g; 95% CI −17.7 to 9.9) than 5 post-partum studies (β=−28.3 g; 95% CI −69.3 to 12.7) and 12 early sampling studies (β=−7.6 g; 95% CI −16.2 to 1.1). 6 of 12 studies with the earliest sampling timing showed results closer to the null.Overall, we detected a small but statistically significant BW deficit across 27 studies. We saw comparable BW deficit magnitudes in both the medium and high confidence studies as well as the early pregnancy group. Despite no definitive pattern by sample timing, larger deficits were seen in postpartum studies. We also saw results closer to the null for a subset of studies restricted to the earliest biomarker collection times. Serial pregnancy sampling, improved precision in gestational age estimates and more standardised reporting of sample variation and exposure units in future epidemiologic research may offer a greater understanding of the relationship between PFHxS on BW and any potential impact of pregnancy haemodynamics.

BACKGROUND:Some disinfection byproducts (DBPs) are teratogens based on toxicological evidence. Conventional use of predominant DBPs as proxies for complex mixtures may result in decreased ability to detect associations in epidemiological studies.OBJECTIVE:We assessed risks of obstructive genitourinary birth defects (OGDs) in relation to 12 DBP mixtures and 13 individual component DBPs.METHODS:We designed a nested registry-based case-control study (210 OGD cases; 2100 controls) in Massachusetts towns with complete quarterly 1999-2004 data on four trihalomethanes (THMs) and five haloacetic acids (HAAs). We estimated temporally-weighted average DBP exposures for the first trimester of pregnancy. We estimated adjusted odds ratios (aORs) and 95% confidence intervals (CIs) for OGD in relation to individual DBPs, unweighted mixtures, and weighted mixtures based on THM/HAA relative potency factors (RPF) from animal toxicology data for full-litter resorption, eye defects, and neural tube defects.RESULTS:We detected elevated aORs for OGDs for the highest of bromodichloromethane (aOR = 1.75; 95% CI: 1.15-2.65), dibromochloromethane (aOR = 1.71; 95% CI: 1.15-2.54), bromodichloroacetic acid (aOR = 1.56; 95%CI: 0.97-2.51), chlorodibromoacetic acid (aOR = 1.97, 95% CI: 1.23-3.15), and tribromoacetic acid (aOR = 1.90; 95%CI: 1.20-3.03). Across unweighted mixture sums, the highest aORs were for the sum of three brominated THMs (aOR = 1.74; 95% CI: 1.15-2.64), the sum of six brominated HAAs (aOR = 1.43; 95% CI: 0.89-2.31), and the sum of nine brominated DBPs (aOR = 1.80; 95% CI: 1.05-3.10). Comparing eight RPF-weighted to unweighted mixtures, the largest aOR differences were for two HAA metrics, which both were higher with RPF weighting; other metrics had reduced or minimally changed ORs in RPF-weighted models.

BackgroundResearchers have developed exposure assessment metrics for disinfection by-products (DBPs) utilizing drinking water monitoring data and accounting for spatial and temporal variability, water consumption, and showering and bathing time with an expectation of decreasing exposure misclassification compared to the use of measured concentrations at public water supply (PWS) monitoring locations alone.ObjectiveWe used exposure data collected for a previous study of DBPs to evaluate how different sources of information impact trihalomethane (THM) exposure estimates.MethodsWe compared gestational exposure estimates to THMs based on water utility monitoring data alone, statistical imputation of daily concentrations to incorporate temporal variability, and personal water consumption and use (bathing and showering). We used Spearman correlation coefficients and ranked kappa statistics to compare exposure classifications.ResultsExposure estimates based on measured or imputed daily THM concentrations, self-reported consumption, or bathing and showering differed substantially from estimates based solely on concentrations from PWS quarterly monitoring reports. Ranked exposure classifications, high to low quartiles or deciles, were generally consistent across each exposure metric (i.e., a subject with "high" exposure based on measured or imputed THM concentrations generally remained in the "high" category across exposure metrics.) The measured concentrations and imputed daily (i.e., spline regression) concentrations were highly correlated (r = 0.98). The weighted kappa statistics comparing exposure estimates using different exposure metrics ranged from 0.27 to 0.89, with the highest values for the ingestion + bathing/showering metrics compared to metrics for bathing/showering only (0.76 and 0.89). Bathing and showering contributed the most to "total" THM exposure estimates.Impact StatementWe compare exposure metrics capturing temporal variability and multiple estimates of personal THM exposure with THM concentrations from PWS monitoring data. Our results show exposure estimates based on imputed daily concentrations accounting for temporal variability were very similar to the measured THM concentrations. We observed low agreement between imputed daily concentrations and ingestion-based estimates. Considering additional routes of exposure (e.g., inhalation and dermal) slightly increased agreement with the measured PWS exposure estimate in this population. Overall, the comparison of exposure assessment metrics allows researchers to understand the added value of additional data collection for future epidemiologic analyses of DBPs.

BACKGROUND:Manganese (Mn) is neurotoxic in adults and children. Current assessments are based on the more extensive adult epidemiological data, but the potential for greater childhood susceptibility remains a concern. To better understand potential lifestage-based variations, we compared susceptibilities to neurotoxicity in children and adults using Mn biomarker data. METHODS:We developed a literature search strategy based on a Population, Exposures, Comparators, and Outcomes statement focusing on inhalation exposures and neurological outcomes in humans. Screening was performed using DistillerSR. Hair biomarker studies were selected for evaluation because studies with air measurements were unavailable or considered inadequate for children. Studies were paired based on concordant Mn source, biomarker, and outcome. Comparisons were made based on reported dose-response slopes (children vs. adults). Study evaluation was conducted to understand the confidence in our comparisons. RESULTS:We identified five studies evaluating seven pairings of hair Mn and neurological outcomes (cognition and motor effects) in children and adults matched on sources of environmental Mn inhalation exposure. Two Brazilian studies of children and one of adults reported intelligent quotient (IQ) effects; effects in both comparisons were stronger in children (1.21 to 2.03-fold difference). In paired analyses of children and adults from the United States, children exhibited both stronger and weaker effects compared to adults (0.37 to 1.75-fold differences) on postural sway metrics. CONCLUSION:There is limited information on the comparative susceptibility of children and adults to inhaled Mn. We report that children may be 0.37 to 2.03 times as susceptible as adults to neurotoxic effects of Mn, thereby providing a quantitative estimate for some aspects of lifestage variation. Due to the limited number of paired studies available in the literature, this quantitative estimate should be interpreted with caution. Our analyses do not account for other sources of inter-individual variation. Additional studies of Mn-exposed children with direct air concentration measurements would improve the evidence base.

Maternal exposure to regulated disinfection by-products (DBPs) during pregnancy has been linked with adverse birth outcomes. However, no human studies have focused on drinking water nitrosamines, a group of emerging unregulated nitrogenous DBPs that exhibits genotoxicity and developmental toxicity in experimental studies. This cohort study included 2,457 mother-infant pairs from a single drinking water supply system in central China, and maternal trimester-specific and entire pregnancy exposure of drinking water nitrosamines were evaluated. Multivariable linear and Poisson regression models were used to estimate the associations between maternal exposure to nitrosamines in drinking water and birth outcomes [birth weight (BW), low birth weight (LBW), small for gestational age (SGA) and preterm delivery (PTD)]. Elevated maternal N-nitrosodimethylamine (NDMA) exposure in the second trimester and N-nitrosopiperidine (NPIP) exposure during the entire pregnancy were associated decreased BW (e.g., β = -88.6 g; 95% CI: -151.0, -26.1 for the highest vs. lowest tertile of NDMA; p for trend = 0.01) and increased risks of PTD [e.g., risk ratio (RR) = 2.16; 95% CI: 1.23, 3.79 and 1.66 for the highest vs. lowest tertile of NDMA; p for trend = 0.002]. Elevated maternal exposure of N-nitrosodiethylamine (NDEA) in the second trimester was associated with increased risk of SGA. Our study detected associations of maternal exposure to drinking water nitrosamines during pregnancy with decreased BW and increased risks of SGA and PTD. These findings are novel but require replication in other study populations.

Background While benefits of greenness to health have been reported, findings specific to child respiratory health are inconsistent. Methods We utilized a prospective birth cohort followed from birth to age 7 years ( n = 617). Residential surrounding greenness was quantified via Normalized Difference Vegetation Index (NDVI) within 200, 400, and 800 m distances from geocoded home addresses at birth, age 7 years, and across childhood. Respiratory health outcomes were assessed at age 7 years, including asthma and lung function [percent predicted forced expiratory volume in the first second (%FEV 1 ), percent predicted forced vital capacity (%FVC), and percent predicted ratio of forced expiratory volume in the first second to forced vital capacity (%FEV 1 /FVC)]. We assessed associations using linear and logistic regression models adjusted for community deprivation, household income, and traffic-related air pollution. We tested for effect measure modification by atopic status. Results We noted evidence of positive confounding as inverse associations were attenuated upon adjustment in the multivariable models. We found evidence of effect measure modification of NDVI and asthma within 400 m at age 7 years by atopic status ( p = 0.04), whereby children sensitized to common allergens were more likely to develop asthma as exposure to greenness increased (OR = 1.3, 95% CI: 0.9, 2.0) versus children not sensitized to common allergens (OR = 0.8, 95% CI: 0.5, 1.2). We found consistently positive associations between NDVI and %FEV 1 and %FVC which similarly evidenced positive confounding upon adjustment. In the adjusted regression models, NDVI at 7 years of age was associated with %FEV 1 (200 m: β = 2.1, 95% CI: 0.1, 3.3; 400 m: β = 1.6, 95% CI: 0.3, 2.9) and %FVC (200 m: β = 1.8, 95% CI: 0.7, 3.0; 400 m: β = 1.6, 95% CI: 0.3, 2.8; 800 m: β = 1.5, 95% CI: 0.1, 2.8). Adjusted results for %FEV 1 /FVC were non-significant except exposure at birth in the 400 m buffer (β = 0.81, 95% CI: 0.1, 1.5). We found no evidence of effect measure modification of NDVI by atopic status for objective measures of lung function. Conclusion Sensitivity to allergens may modify the effect of greenness on risk for asthma in children but greenness is likely beneficial for concurrent lung function regardless of allergic status.

Background: Per- and polyfluoroalkyl substances (PFAS) comprise a large class of chemicals with widespread use and persistence in the environment and in humans; however, most of the epidemiology research has focused on a small subset. Objectives: The aim of this systematic evidence map (SEM) is to summarize the epidemiology evidence on approximately 150 lesser studied PFAS prioritized by the EPA for tiered toxicity testing, facilitating interpretation of those results as well as identification of priorities for risk assessment and data gaps for future research. Methods: The Populations, Exposure, Comparators, and Outcomes (PECO) criteria were intentionally broad to identify studies of any health effects in humans with information on associations with exposure to the identified PFAS. Systematic review methods were used to search for literature that was screened using machine-learning software and manual review. Studies meeting the PECO criteria underwent quantitative data extraction and evaluation for risk of bias and sensitivity using the Integrated Risk Information System approach. Results: 193 epidemiology studies were identified, which included information on 15 of the PFAS of interest. The most commonly studied health effect categories were metabolic (n=37), endocrine (n=30), cardiovascular (30), female reproductive (n=27), developmental (n=26), immune (n=22), nervous (n=21), male reproductive (n=14), cancer (n=12), and urinary (n=11) effects. In study evaluation, 120 (62%) studies were considered High/Medium confidence for at least one outcome. Discussion: Most of the PFAS in this SEM have little to no epidemiology data available to inform evaluation of potential health effects. Although exposure to the 15 PFAS that had data was fairly low in most studies, these less-studied PFAS may be used as replacements for “legacy” PFAS, leading to potentially greater exposure. It is impractical to generate epidemiology evidence to fill the existing gaps for all potentially relevant PFAS. This SEM highlights some of the important research gaps that currently exist. https://doi.org/10.1289/EHP11185

BACKGROUND AND AIM: Air pollution comprises a mixture of gases and small particles, including metals, and has been associated with changes in mean birthweight (BW). Often, studies estimate exposure for a single metal in isolation or multiple metals simultaneously, but rarely evaluate both or examine the potential bias amplification. METHODS: We evaluated census tract level concentrations of 8 metals (arsenic, cadmium, chromium, cobalt, lead, manganese, mercury, and nickel) from the 2011 National Air Toxics Assessment linked with 431,929 infant-mother pairs with births between 2012-2015. Correlations were low to moderate (r0.62) between the metals. The change in BW (grams, g) and 95% confidence interval (CI) per one log-transformed unit of change in concentration was estimated with generalized estimation equation models, adjusted for race/ethnicity, age, marital status, education, and Medicaid status. Exposure to each metal was considered in isolation (IPM) and part of a multiple pollutant model (MPM) of 8 metals (mutually adjusted). RESULTS:We observed the largest decrease in BW for manganese [-8.0g(95% CI:-10.0,-6.1)] in the IPM and for arsenic [-11.5g(-19.0,-4.0)] in the MPM. The greatest increase in BW was observed with cadmium for both models [isolated: 16.5g(11.5,21.56); multiple: 13.7g(6.2,21.2)]. The direction of association changed with arsenic, chromium, and lead between the IPM and MPM. All metals, except manganese, were associated with more than a 10% change in BW per one log-transformed unit change in metal concentration when comparing the results from the IPM to the MPM, suggesting possible bias amplification. CONCLUSIONS:We observed changes in directionality and substantial changes in the magnitude of the associations between IPM and MPM. Evaluating MPM may over or under emphasize associations due to residual confounding or collinearity, leading to bias amplification. Considering the potential for bias amplification and its magnitude is useful in evaluating MPM. This abstract does not reflect EPA policy. KEYWORDS: Birth outcomes, Heavy metals, Multi-pollutant, Mixtures analysis, Air pollution

BACKGROUND AND AIM: Recent evidence suggests that timing differences in PFOA biomarker sampling may result in reverse causality or confounding due to pregnancy-related hemodynamic changes. In a meta-analysis of epidemiologic data from 16 studies of birth weight identified in a systematic review of the health effects of PFHxS exposure, we examined whether results differed by study quality and by timing of exposure measurement. METHODS: Using a random effects model, we examined studies with maternal blood samples collected during first trimester (T1), second trimester (T2), third trimester (T3), post-partum (PP), or umbilical cord blood (UCB). Three grouping strategies were used: 1) early and mid-pregnancy (T1; T1/T2; T2, n=7) versus later sampling (T3; PP; UCB, n=9); 2) early pregnancy (T1; T1/T2, n=4) versus later sampling (T2; T3; PP; UCB, n=12); 3) early (T1; T1/T2, n=4), mid- and late-pregnancy (T2; T2/T3; T3, n=7), and post-pregnancy (UBC; PP, n=5). RESULTS:The summary mean birthweight deficit was -12 grams (95%CI: -25, 0.1; I2=0%) per ln-unit PFHxS increase with no differences across study confidence level (High/Medium/Low). We saw minimal differences in pooled deficits with the first two dichotomized groups (-11 vs -15 g for early/mid-pregnancy versus late/post-pregnancy; -10 vs -13 g for early pregnancy versus other periods). The third analysis showed comparable pooled estimates for maternal samples collected during pregnancy (early: -10 g, mid-/late-pregnancy: -8 g); but a larger deficit for UCB or PP (-28 g; 95%CI: -69, 13). CONCLUSIONS:Although we saw larger deficits for post-pregnancy samples based on few studies, we consistently saw birth weight deficits for pregnancy-based measures regardless of timing. Additionally, the few PFHxS studies adjusting for hemodynamics did not show evidence of confounding. Thus, more work to determine whether reverse causality is a plausible explanation of these differences and to determine the most relevant grouping strategies based on hemodynamic expectations will be informative. KEYWORDS: PFAS, PFC, PFHxS, Birthweight, Meta-analysis, Pregnancy

BACKGROUND AND AIM: Manganese (Mn) is an essential nutrient but a neurotoxicant at elevated concentrations. Inhalation exposure can bypass homeostatic regulations and enter the brain. The U.S. EPA's 1993 inhalation reference concentration and ATSDR's 2011 minimal risk level are based on neurological outcomes in adults. To understand whether more recent data could provide information on potential lifestage-based variations in susceptibility, we compared sensitivities to neurotoxicity in children and adults using Mn biomarker data. METHODS: We developed a literature search strategy and a Population, Exposures, Comparators, and Outcomes statement. The search covered 2011-2020, with supplemental backwards reference searching. Screening was performed using SWIFT and DistillerSR. Paired studies were identified based on concordant Mn source, biomarker, and outcome. Comparisons were made based on slope of the observed dose-response (children:adults). RESULTS:We identified five studies evaluating seven pairings of hair Mn and neurological outcomes in child and adult populations with environmental Mn inhalation exposure. Two Brazilian studies of children (ages 6-12) and one of adults reported effects on IQ; effects in both comparisons were stronger in children (203% and 121%). A paired study of children and adults in Ohio reported effects on five metrics of postural sway; children exhibited both stronger and weaker effects compared to adults (175% to 69%). CONCLUSIONS:Preliminary results from a small database of well-matched studies suggest that variability in neurological effects by lifestage is less than 10-fold. This conclusion should be interpreted with caution, given the limited available paired studies. Future work will consider the impact of study quality on these conclusions. Additional studies on Mn-exposed children with direct measurement of air concentrations would provide further understanding of the toxicity of inhaled Mn among susceptible lifestages. *The views expressed in this abstract are those of the authors and do not necessarily reflect the views or policies of the U.S. Environmental Protection Agency. KEYWORDS: Children's environmental health, Heavy metals, Risk assessment, Neurodevelopmental outcomes

BACKGROUND AND AIM: Epidemiologic studies of disinfection byproducts (DBPs) and birth defects have employed a variety of exposure assessment methods. Because there is 1) no gold standard exposure assessment method for DBPs, and 2) documented variability in exposure based on spatial, temporal, and personal behavior patterns, it is important to compare methods and understand their potential impact on health effect estimates. This question is particularly relevant for studies measuring exposures for developmental outcomes such as birth defects, with a narrow critical window. METHODS: We reviewed epidemiologic studies of DBPs and birth defects published prior to 2021. Information on exposure assessment methods and critical windows of pregnancy captured by exposure assessment was extracted by one author, with quality control by two authors. Data were organized and summarized based on complexity of the method, including their ability to characterize and address temporal or spatial variability and personal water-use patterns. RESULTS:We identified exposure assessment methods utilized in 21 studies of DBPs and birth defects; 6 studies used non-specific DBP indicators, such as water type or disinfection method;17 studies developed spatial, temporal, or spatiotemporal-informed estimates based on data collected by public monitoring systems in conjunction with additional modeling; 7 studies incorporated personal behavior information to help address route-specific exposure; 4 of these incorporated DBP uptake factors. Most studies incorporated 1 or 2 exposure assessment methods, while 5 studies included 3 or more approaches; 14 studies developed exposure estimates based on the critical window for birth defects. CONCLUSIONS:There is heterogeneity in exposure assessment methods among studies of DBPs and birth defects. These methods vary in complexity and capture distinct sources of DBP variation. The choice of metric affects the anticipated magnitude of exposure misclassification. More consistent consideration of spatiotemporal patterns and the critical window of development will improve exposure assessment in future studies. *Abstract does not reflect EPA policy KEYWORDS: Exposures, Exposure assessment, Birth outcomes, Water quality

Disinfection of drinking water is a public health advance, but the potential risk of unintended consequences of disinfection byproduct (DBP) formation remains unclear. Over 600 DBPs have been identified to date, although most have not been analytically quantified in disinfected drinking water. Some DBP exposures occur through multiple routes which can complicate exposure assessment in epidemiological studies especially based on aggregate data derived from routinely-collected monitoring data. Therefore it is important to quantify the spatiotemporal variability in occurrence to minimize measurement error and exposure misclassification. In the absence of biomonitoring data, concentration-based exposure scores based on routinely-collected monitoring data can be spatial water system averages or location-specific estimates. However, intra-water system variation in DBP concentrations results from formation and degradation processes occurring over space and residence time. Temporal variability can also occur from diurnal variation to longer-term trends such as seasonality noted for trihalomethanes (THMs) and haloacetic acids (HAAs). Temporally averaged data may estimate lifetime exposure for cancer outcomes or trimester-specific windows for developmental effects, but they ill fully account for the potential impact of peak exposures. Although the DBP surrogacy issue is beyond the scope here, data availability limitations may preclude examination and delineation of potential risk for individual DBPs or DBP mixtures especially when the measurement error structures can vary due to use of aggregate exposure estimates. With few DBPs regulated or routinely measured, this review will characterize strengths and limitations of existing approaches based on HAAs and THMs and will offer evidence from simulations and sensitivity analyses and offer recommendations to help address the aforementioned challenges. Disclaimer: The views expressed in this abstract are those of the author and do not necessarily reflect the views or policies of the U.S. EPA. Keywords: Water, water quality, disinfection byproducts, environmental epidemiology