Model choice for estimating the association between exposure to chemical mixtures and health outcomes: A simulation study

Endocrine-Disrupting Chemicals and Persistent Nausea among Pregnant Women Enrolled in the Illinois Kids Development Study (I-KIDS)

BACKGROUND

Pregnant women are exposed to numerous endocrine-disrupting chemicals (EDCs). Pregnancy-related nausea likely has hormonal etiology and may persist beyond the first trimester.

OBJECTIVES

Therefore, we aimed to determine the relationship between EDC biomarkers and pregnancy nausea characteristics.

METHODS

Illinois Kids Development Study (I-KIDS) pregnant women (n = 467 ) reported nausea symptoms monthly from conception to delivery. We categorized women as never having nausea (9%) or as having typical (ends by 17 wk gestation; 42%), persistent (ends after 17 wk gestation; 25%), or irregular (24%) nausea. Women provided five urine samples across pregnancy, which we pooled and analyzed for phthalate/replacement, phenol, and triclocarban biomarkers. Using covariate-adjusted logistic regression, we evaluated relationships of EDCs with nausea and used quantile-based g-computation (QGComp) and Bayesian kernel machine regression (BKMR) to evaluate joint associations of EDCs with nausea symptoms. We also considered differences in associations by fetal sex.

RESULTS

Only the sum of urinary biomarkers of di(isononyl) cyclohexane-1,2-dicarboxylate (Σ DiNCH ) was associated with higher risk of persistent nausea compared to typical nausea [odds ratio (OR) = 1.18 ; 95% CI: 1.01, 1.37] in all women. However, using QGComp, a 10% higher concentration of the EDC mixture was associated with 14% higher risk of persistent nausea [relative risk (RR) = 1.14 ; 95% CI: 1.01, 1.30], due to Σ DiNCH , ethylparaben, and the sum of di-2-ethylhexyl phthalate (Σ DEHP ) metabolites. Similarly, using BMKR, the EDC mixture was associated with greater odds of persistent nausea in all women. In women carrying male offspring, ethylparaben was associated with persistent nausea, and a 10% higher concentration of the QGComp mixture was associated with 26% higher risk of persistent nausea (RR = 1.26 ; 95% CI:1.13, 1.41), driven by ethylparaben and Σ DiNCH . Consistently, using BKMR, EDCs were positively associated with persistent nausea in women carrying males. We did not identify associations between EDC biomarkers and persistent nausea in women carrying females or between EDC biomarkers and other nausea patterns.

DISCUSSION

Nonpersistent EDCs, modeled as a mixture, are associated with persistent nausea in pregnancy, primarily in women carrying males. Future work should explore possible mechanisms, clinical implications, and interventions to reduce exposures and symptoms. https://doi.org/10.1289/EHP15547.

Associations between single and combined exposures to environmental phenols and ulcerative colitis in American adults

OBJECTIVE

The etiology of ulcerative colitis (UC) is complex and involves multiple factors, with exposure to environmental toxins potentially contributing greatly to its pathogenesis. Therefore, this study was carried out with the purpose of delving into the associations between single and combined exposures to environmental phenols and UC among American adults.

METHODS

Survey data from the 2009-2010 National Health and Nutrition Examination Survey were selected for our research. The associations between single and combined exposures to environmental phenols and the prevalence of UC were analyzed using weighted multivariate logistic regression models as well as Bayesian kernel machine regression (BKMR).

RESULTS

A total of 1,422 adults aged 20 years old and above were included in this study, 17 of whom had UC. The correlation matrix showed strong associations between 2,4-dichlorophenol (2,4-DCP) and 2,5-dichlorophenol (2,5-DCP) (R = 0.81), as well as between 2,4,5-trichlorophenol (2,4,5-TCP) and 2,4,6-trichlorophenol (2,4,6-TCP) (R = 0.73). The logistic regression model revealed that, after adjusting for confounders, exposure to environmental phenols was positively associated with the prevalence of UC, with 2,4,6-TCP showing a significant association (OR = 2.37, 95 % CI = 1.10, 5.09, P = 0.037). The BKMR analysis indicated an upward trend in the overall effect of combined exposures to environmental phenols on UC. All five phenols contributed to this effect, with 2,4,6-TCP exhibiting the most pronounced effect. When other compounds were fixed at the 50th percentile, the impact of the five phenols on UC demonstrated a positive association, without any noteworthy interaction among the compounds.

CONCLUSION

Our findings suggested that exposure to environmental phenols may contribute to the occurrence of UC among American adults.

The association between urban land use and depressive symptoms in young adulthood: a FinnTwin12 cohort study

BACKGROUND

Depressive symptoms lead to a serious public health burden and are considerably affected by the environment. Land use, describing the urban living environment, influences mental health, but complex relationship assessment is rare.

OBJECTIVE

We aimed to examine the complicated association between urban land use and depressive symptoms among young adults with differential land use environments, by applying multiple models.

METHODS

We included 1804 individual twins from the FinnTwin12 cohort, living in urban areas in 2012. There were eight types of land use exposures in three buffer radii. The depressive symptoms were assessed through the General Behavior Inventory (GBI) in young adulthood (mean age: 24.1). First, K-means clustering was performed to distinguish participants with differential land use environments. Then, linear elastic net penalized regression and eXtreme Gradient Boosting (XGBoost) were used to reduce dimensions or prioritize for importance and examine the linear and nonlinear relationships.

RESULTS

Two clusters were identified: one is more typical of city centers and another of suburban areas. A heterogeneous pattern in results was detected from the linear elastic net penalized regression model among the overall sample and the two separated clusters. Agricultural residential land use in a 100 m buffer contributed to GBI most (coefficient: 0.097) in the "suburban" cluster among 11 selected exposures after adjustment with demographic covariates. In the "city center" cluster, none of the land use exposures was associated with GBI, even after further adjustment with social indicators. From the XGBoost models, we observed that ranks of the importance of land use exposures on GBI and their nonlinear relationships are also heterogeneous in the two clusters.

IMPACT

This study examined the complex relationship between urban land use and depressive symptoms among young adults in Finland. Based on the FinnTwin12 cohort, two distinct clusters of participants were identified with different urban land use environments at first. We then employed two pluralistic models, elastic net penalized regression and XGBoost, and revealed both linear and nonlinear relationships between urban land use and depressive symptoms, which also varied in the two clusters. The findings suggest that analyses, involving land use and the broader environmental profile, should consider aspects such as population heterogeneity and linearity for comprehensive assessment in the future.

Statistical methods for chemical mixtures: a roadmap for practitioners

Quantitative characterization of the health impacts associated with exposure to chemical mixtures has received considerable attention in current environmental and epidemiological studies. With many existing statistical methods and emerging approaches, it is important for practitioners to understand when each method is best suited for their inferential goals. In this study, we conduct a review and comparison of 11 analytical methods available for use in mixtures research, through extensive simulation studies for continuous and binary outcomes. These methods fall in three different classes: identifying important components of a mixture, identifying interactions and creating a summary score for risk stratification and prediction. We carry out an illustrative data analysis in the PROTECT birth cohort from Puerto Rico. Most importantly we develop an integrated package "CompMix" that provides a platform for mixtures analysis where the practitioner can implement a pipeline for several types of mixtures analysis. Our simulation results suggest that the choice of methods depends on the goal of analysis and there is no clear winner across the board. For selection of important toxicants in the mixture and for identifying interactions, Elastic net by Zou et al. (Enet), Lasso for Hierarchical Interactions by Bien et al (HierNet), Selection of nonlinear interactions by a forward stepwise algorithm by Narisetty et al. (SNIF) have the most stable performance across simulation settings. Additionally, the predictive performance of the Super Learner ensembling method by Van de Laan et al. and HierNet are found to be superior to the rest of the methods. For overall summary or a cumulative measure, we find that using the Super Learner to combine multiple Environmental Risk Scores can lead to improved risk stratification properties. We have developed an R package "CompMix: A comprehensive toolkit for environmental mixtures analysis", allowing users to implement a variety of tasks under different settings and compare the findings. In summary, our study offers guidelines for selecting appropriate statistical methods for addressing specific scientific questions related to mixtures research. We identify critical gaps where new and better methods are needed.

Associations of urinary non-persistent endocrine disrupting chemical biomarkers with early-to-mid pregnancy plasma sex-steroid and thyroid hormones

BACKGROUND/OBJECTIVES

Pregnant women are exposed to numerous endocrine disrupting chemicals (EDCs) that can affect hormonal pathways regulating pregnancy outcomes and fetal development. Thus, we evaluated overall and fetal sex-specific associations of phthalate/replacement, paraben, and phenol biomarkers with sex-steroid and thyroid hormones.

METHODS

Illinois women (n = 302) provided plasma for progesterone, estradiol, testosterone, free T4 (FT4), total T4 (TT4), and thyroid stimulating hormone (TSH) at median 17 weeks gestation. Women also provided up-to-five first-morning urine samples monthly across pregnancy (8-40 weeks), which we pooled to measure 19 phthalate/replacement metabolites (reflecting ten parent compounds), three parabens, and six phenols. We used linear regression to evaluate overall and fetal sex-specific associations of biomarkers with hormones, as well as weighted quantile sum and Bayesian kernel machine regression (BKMR) to assess cumulative associations, non-linearities, and chemical interactions.

RESULTS

In women of relatively high socioeconomic status, several EDC biomarkers were associated with select hormones, without cumulative or non-linear associations with progesterone, FT4, or TT4. The biomarker mixture was negatively associated with estradiol (only at higher biomarker concentrations using BKMR), testosterone, and TSH, where each 10% mixture increase was associated with -5.65% (95% CI: -9.79, -1.28) lower testosterone and -0.09 μIU/mL (95% CI: -0.20, 0.00) lower TSH. Associations with progesterone, testosterone, and FT4 did not differ by fetal sex. However, in women carrying females, we identified an inverted u-shaped relationship of the mixture with estradiol. Additionally, in women carrying females, each 10% increase in the mixture was associated with 1.50% (95% CI: -0.15, 3.18) higher TT4, whereas in women carrying males, the mixture was associated with -1.77% (95% CI: -4.08, 0.58) lower TT4 and -0.18 μIU/mL (95% CI: -0.33, -0.03) lower TSH. We also identified select chemical interactions.

CONCLUSION

Some biomarkers were associated with early-to-mid pregnancy hormones. There were some sex-specific and non-linear associations. Future studies could consider how these findings relate to pregnancy/birth outcomes.

Methodological Challenges in Spatial and Contextual Exposome-Health Studies

The concept of the exposome encompasses the totality of exposures from a variety of external and internal sources across an individual's life course. The wealth of existing spatial and contextual data makes it appealing to characterize individuals' external exposome to advance our understanding of environmental determinants of health. However, the spatial and contextual exposome is very different from other exposome factors measured at the individual-level as spatial and contextual exposome data are more heterogenous with unique correlation structures and various spatiotemporal scales. These distinctive characteristics lead to multiple unique methodological challenges across different stages of a study. This article provides a review of the existing resources, methods, and tools in the new and developing field for spatial and contextual exposome-health studies focusing on four areas: (1) data engineering, (2) spatiotemporal data linkage, (3) statistical methods for exposome-health association studies, and (4) machine- and deep-learning methods to use spatial and contextual exposome data for disease prediction. A critical analysis of the methodological challenges involved in each of these areas is performed to identify knowledge gaps and address future research needs.

Using non-parametric Bayes shrinkage to assess relationships between multiple environmental and social stressors and neonatal size and body composition in the Healthy Start cohort

BACKGROUND

Both environmental and social factors have been linked to birth weight and adiposity at birth, but few studies consider the effects of exposure mixtures. Our objective was to identify which components of a mixture of neighborhood-level environmental and social exposures were driving associations with birth weight and adiposity at birth in the Healthy Start cohort.

METHODS

Exposures were assessed at the census tract level and included air pollution, built environment characteristics, and socioeconomic status. Prenatal exposures were assigned based on address at enrollment. Birth weight was measured at delivery and adiposity was measured using air displacement plethysmography within three days. We used non-parametric Bayes shrinkage (NPB) to identify exposures that were associated with our outcomes of interest. NPB models were compared to single-predictor linear regression. We also included generalized additive models (GAM) to assess nonlinear relationships. All regression models were adjusted for individual-level covariates, including maternal age, pre-pregnancy BMI, and smoking.

RESULTS

Results from NPB models showed most exposures were negatively associated with birth weight, though credible intervals were wide and generally contained zero. However, the NPB model identified an interaction between ozone and temperature on birth weight, and the GAM suggested potential non-linear relationships. For associations between ozone or temperature with birth weight, we observed effect modification by maternal race/ethnicity, where effects were stronger for mothers who identified as a race or ethnicity other than non-Hispanic White. No associations with adiposity at birth were observed.

CONCLUSIONS

NPB identified prenatal exposures to ozone and temperature as predictors of birth weight, and mothers who identify as a race or ethnicity other than non-Hispanic White might be disproportionately impacted. However, NPB models may have limited applicability when non-linear effects are present. Future work should consider a two-stage approach where NPB is used to reduce dimensionality and alternative approaches examine non-linear effects.