Accounting for the uncertainty due to chemicals below the detection limit in mixture analysis

Association of urinary bisphenols with oxidative stress and inflammatory markers and their role in obesity

Bisphenol A (BPA) and its substitutes are widely used in daily life. Animal and cell line experiments have confirmed the effects of bisphenols on oxidative stress and inflammation. However, current population evidence for the effects of BPA alternatives, such as bisphenol F (BPF) and bisphenol S (BPS), on oxidative stress and inflammation is still sparse. Based on the National Health and Nutrition Examination Survey 2013-2016 data, our study used linear regression, weighted quantile sum model, and Bayesian kernel machine regression model to evaluate the effects of BPA, BPS, and BPF alone and in combination on oxidative stress (serum total bilirubin, and iron) and inflammation (alkaline phosphatase, C-reactive protein, γ-glutamyl transferase ferritin, neutrophil count, lymphocyte count, and neutrophil-to-lymphocyte ratio) markers. On this basis, the possible roles of oxidative stress and inflammation in obesity, which is associated with exposure to bisphenols (BPs), were initially explored. Based on the different covariates selected, a total of 3039 and 2258 participants were included in our study for models 1 and 2, respectively; the median age of participants was 48 years, and 48.7 % were male. Based on all models, our results showed that exposure to BPs alone or in combination was associated with downregulation of serum total bilirubin. Urinary BPF concentration was specifically associated with the neutrophil-to-lymphocyte ratio. Serum total bilirubin may play a role in the association between obesity and BP mixture exposure. Upregulation of the neutrophil-to-lymphocyte ratio was not associated with obesity. In conclusion, our study found that single or combined exposure to BPs, as measured in urine, may be associated with changes in oxidative stress and inflammatory markers, and a decrease in serum total bilirubin may play a mediating role in BP-induced obesity.

Modeling historic neighborhood deprivation and non-Hodgkin lymphoma risk

Many studies have identified associations between neighborhood deprivation and disease, emphasizing the importance of social determinants of health. However, when studying diseases with long latency periods such as cancers, considering the timing of exposures for deprivation becomes more important. In this study, we estimated the associations between neighborhood deprivation indices at several time points and risk of non-Hodgkin lymphoma (NHL) in a population-based case-control study at four study centers - Detroit, Iowa, Los Angeles County, and Seattle (1998-2000). We used the Bayesian index regression model and residential histories to estimate neighborhood deprivation index effects in crude models and adjusted for four chemical mixtures measured in house dust and individual-level covariates. We found that neighborhood deprivation in 1980, approximately twenty years before study entry, provided better model fit than did neighborhood deprivation at 1990 and 2000. We identified several statistically significant associations between neighborhood deprivation in 1980 and NHL risk in Iowa and among long-term (20+ years) residents of Detroit. The most important variables in these indices were median gross rent as a percentage of household income in Iowa and percent of single-parent households with at least one child and median household income in Detroit. Associations remained statistically significant after adjustment for individual-level covariates and chemical mixtures, providing evidence for historic neighborhood deprivation as a risk factor for NHL and motivating future research to uncover the specific carcinogens driving these associations in deprived areas.

Association between organophosphorus pesticide exposure and depression risk in adults: A cross-sectional study with NHANES data

Organophosphorus pesticides (OPPs) are widely used pesticides, and previous studies showed that OPPs can increase the risk of central nervous system disorders (e.g., Parkinson's and Alzheimer's disease). However, few studies have comprehensively explored their association with depression in general adults. We analyzed data from 5206 participants aged 20 years or more based on four National Health and Nutrition Examination Survey (NHANES) cycles. OPPs exposure was estimated using measures of urinary concentrations for six OPPs metabolites. Survey-weighted generalized linear regression model (SWGLM) was used to explore the association of OPPs metabolites with depression. Subgroup analyses were performed by age (≦60 years and >60 years) and gender. The weighted quantile sum (WQS) regression model was used to explore the overall association of six OPPs metabolites with depression. In addition, The Bayesian kernel machine regression (BKMR) model was applied to investigate the interaction and joint effects of multiple OPPs metabolites with depression. The SWGLM showed that dimethyl phosphate (DMP) and dimethyl thiophosphate (DMTP), whether taken as continuous or quartile variables, had a positive correlation with depression. Diethyl phosphate (DEP) and dimethyl dithiophosphate (DMDTP) in the highest quartile were positively associated with depression compared to the lowest quartile. In subgroup analysis, we found that the effects of the above chemicals on depression existed in the male and young middle-aged population, while DMP was present in the female. There was a significant combined overall effect of six OPPs metabolites with depression [OR = 1.232, 95%CI: (1.011, 1.504)] in WQS. Furthermore, the BKMR model also showed a positive trend in the overall effect of six OPPs metabolites with depression. In conclusion, our results suggest that exposure to OPPs may increase the risk of depression in US adults. Men and young and middle-aged populations are more vulnerable to OPPs and the mixture of OPPs metabolites may induce depression.

Estimating mixture effects and cumulative spatial risk over time simultaneously using a Bayesian index low-rank kriging multiple membership model

The exposome is an ideal in public health research that posits that individuals experience risk for adverse health outcomes from a wide variety of sources over their lifecourse. There have been increases in data collection in the various components of the exposome, but novel statistical methods are needed that capture multiple dimensions of risk at once. We introduce a Bayesian index low-rank kriging (LRK) multiple membership model (MMM) to simultaneously estimate the health effects of one or more groups of exposures, the relative importance of exposure components, and cumulative spatial risk over time using residential histories. The model employs an MMM to consider all residential locations for subjects weighted by duration and LRK to increase computational efficiency. We demonstrate the performance of the Bayesian index LRK-MMM through a simulation study, showing that the model accurately and consistently estimates the health effects of one or several group indices and has high power to identify a region of elevated spatial risk due to unmeasured environmental exposures. Finally, we apply our model to data from a multicenter case-control study of non-Hodgkin lymphoma (NHL), finding a significant positive association between one index of pesticides and risk for NHL in Iowa. Additionally, we find an area of significantly elevated spatial risk for NHL in Los Angeles. In conclusion, our Bayesian index LRK-MMM represents a step forward toward bringing the ideals of the exposome into practice for environmental risk analyzes.

Maternal Exposure to Per- and Polyfluoroalkyl Substances (PFAS) and Male Reproductive Function in Young Adulthood: Combined Exposure to Seven PFAS

BACKGROUND

Concerns remain about the human reproductive toxicity of the widespread per- and polyfluoroalkyl substances (PFAS) during early stages of development.

OBJECTIVES

We examined associations between maternal plasma PFAS levels during early pregnancy and male offspring reproductive function in adulthood.

METHODS

The study included 864 young men (age range:18.9-21.2 y) from the Fetal Programming of Semen Quality (FEPOS) cohort established between 2017 and 2019. Plasma samples from their mothers, primarily from the first trimester, were retrieved from the Danish National Biobank and levels of 15 PFAS were measured. Seven PFAS had detectable levels above the limit of detection in >80% of the samples and were included in analyses. Semen quality, testicular volume, and levels of reproductive hormones and PFAS were assessed in the young men. We used weighted quantile sum (WQS) regression to estimate the associations between combined exposure to maternal PFAS and reproductive function, and negative binomial regression to estimate the associations of single substances, while adjusting for a range of a priori-defined fetal and postnatal risk factors.

RESULTS

By a 1-unit increase in the WQS index, combined maternal PFAS exposure was associated with lower sperm concentration (-8%; 95% CI: -16%, -1%), total sperm count (-10%; 95% CI: -17%, -2%), and a higher proportion of nonprogressive and immotile sperm (5%; 95% CI: 1%, 8%) in the young men. Different PFAS contributed to the associations with varying strengths; however, perfluoroheptanoic acid was identified as the main contributor in the analyses of all three outcomes despite the low concentration. We saw no clear association between exposure to maternal PFAS and testicular volume or reproductive hormones.

DISCUSSION

In a sample of young men from the general Danish population, we observed consistent inverse associations between exposure to maternal PFAS and semen quality. The study needs to be replicated in other populations, taking combined exposure, as well as emerging short-chain PFAS, into consideration. https://doi.org/10.1289/EHP10285.

Association between exposure to organophosphorus pesticides and the risk of diabetes among US Adults: Cross-sectional findings from the National Health and Nutrition Examination Survey

OBJECTIVE

Organophosphorus pesticides (OPPs) are commonly used pesticides across the world, however there is little epidemiological evidence linking their exposure to diabetes. Hence, this study aimed at investigating the effect of OPP exposure on the prevalence of diabetes in American adults.

METHODS

Adults (≥20 years old) were eligible for this study from the National Health and Nutrition Examination Survey (NHANES). Multivariate logistic regression model was employed to explore the associations of six main urinary OPPs metabolites with diabetes. Subgroup analyses were performed by age and gender. Combined effect of OPPs metabolites on the overall association with diabetes was evaluated by weighted quantile sum regression (WQS). Furthermore, Bayesian kernel machine regression (BKMR) model was implemented to explore joint effect of multiple OPPs metabolites on diabetes.

RESULTS

Ultimately, 6,593 adults were included in our analysis. Of them, 1,044 participants were determined as diabetes patients. The results of logistic regression shown that urinary OPPs metabolites concentrations, whether taken as continuous variables or quantiles, were in positive correlation with diabetes. Notably, the p for trend of diethylphosphate (DEP), a kind of OPPs metabolites, was less than 0.05 indicated that a linear trend may exist between levels of DEP and prevalence of diabetes among adults while this trend was not obversed in other OPPs metabolites. In the WQS model, combined exposure of OPPs metabolites had a significantly positive association with diabetes (OR: 1.057; 95% CI: 1.002, 1.114) and diethylphosphate (36.84%) made the largest contributor to the WQS index. The result of BKMR also suggested a positive trend of association between mixed OPPs metabolites and diabetes.

CONCLUSION

Our results add credibility to the argument that OPP exposure might trigger diabetes. Certainly, prospective data are required to corroborate our findings.

A review of practical statistical methods used in epidemiological studies to estimate the health effects of multi-pollutant mixture

Environmental risk factors have been implicated in adverse health effects. Previous epidemiological studies on environmental risk factors mainly analyzed the impact of single pollutant exposure on health, while in fact, humans are constantly exposed to a complex mixture consisted of multiple pollutants/chemicals. In recent years, environmental epidemiologists have sought to assess adverse health effects of exposure to multi-pollutant mixtures based on the diversity of real-world environmental pollutants. However, the statistical challenges are considerable, for instance, multicollinearity and interaction among components of the mixture complicate the statistical analysis. There is currently no consensus on appropriate statistical methods. Here we summarized the practical statistical methods used in environmental epidemiology to estimate health effects of exposure to multi-pollutant mixture, such as Bayesian kernel machine regression (BKMR), weighted quantile sum (WQS) regressions, shrinkage methods (least absolute shrinkage and selection operator, elastic network model, adaptive elastic-net model, and principal component analysis), environment-wide association study (EWAS), etc. We sought to review these statistical methods and determine the application conditions, strengths, weaknesses, and result interpretability of each method, providing crucial insight and assistance for addressing epidemiological statistical issues regarding health effects from multi-pollutant mixture.

Imputation of Below Detection Limit Missing Data in Chemical Mixture Analysis with Bayesian Group Index Regression

There is growing scientific interest in identifying the multitude of chemical exposures related to human diseases through mixture analysis. In this paper, we address the issue of below detection limit (BDL) missing data in mixture analysis using Bayesian group index regression by treating both regression effects and missing BDL observations as parameters in a model estimated through a Markov chain Monte Carlo algorithm that we refer to as pseudo-Gibbs imputation. We compare this with other Bayesian imputation methods found in the literature (Multiple Imputation by Chained Equations and Sequential Full Bayes imputation) as well as with a non-Bayesian single-imputation method. To evaluate our proposed method, we conduct simulation studies with varying percentages of BDL missingness and strengths of association. We apply our method to the California Childhood Leukemia Study (CCLS) to estimate concentrations of chemicals in house dust in a mixture analysis of potential environmental risk factors for childhood leukemia. Our results indicate that pseudo-Gibbs imputation has superior power for exposure effects and sensitivity for identifying individual chemicals at high percentages of BDL missing data. In the CCLS, we found a significant positive association between concentrations of polycyclic aromatic hydrocarbons (PAHs) in homes and childhood leukemia as well as significant positive associations for polychlorinated biphenyls (PCBs) and herbicides among children from the highest quartile of household income. In conclusion, pseudo-Gibbs imputation addresses a commonly encountered problem in environmental epidemiology, providing practitioners the ability to jointly estimate the effects of multiple chemical exposures with high levels of BDL missingness.

Associations of single and multiple per- and polyfluoroalkyl substance (PFAS) exposure with vitamin D biomarkers in African American women during pregnancy

Vitamin D has been linked to various physiological functions in pregnant women and their fetuses. Previous studies have suggested that some per- and polyfluoroalkyl substances (PFAS) may alter serum vitamin D concentrations. However, no study has investigated the relationship between PFAS and vitamin D in pregnant women. This study aims to evaluate the associations of serum PFAS with serum total and free 25-hydroxyvitamin D (25(OH)D) during pregnancy in a cohort of African American women in Atlanta, GA. Blood samples from 442 participants were collected in early pregnancy (8-14 weeks of gestation) for PFAS and 25(OH)D measurements, and additional samples were collected in late pregnancy (24-30 weeks) for the second 25(OH)D measurements. We fit multivariable linear regressions and weighted quantile sum (WQS) regressions to estimate the associations of individual PFAS and their mixtures with 25(OH)D concentrations. We found mostly positive associations of total 25(OH)D with PFHxS (perfluorohexane sulfonic acid), PFOS (perfluorooctane sulfonic acid), PFDA (perfluorodecanoic acid), and NMeFOSAA (N-methyl perfluorooctane sulfonamido acetic acid), and negative associations with PFPeA (perfluoropentanoic acid). For free 25(OH)D, positive associations were observed with PFHxS, PFOS, PFOA (perfluorooctanoic acid), and PFDA, and a negative association with PFPeA among the women with male fetuses in the models using 25(OH)D measured in late pregnancy. In mixture models, a quartile increase in WQS index was associated with 2.88 ng/mL (95%CI 1.14-4.59) and 5.68 ng/mL (95%CI 3.31-8.04) increases in total 25(OH)D measured in the early and late pregnancy, respectively. NMeFOSAA, PFDA, and PFOS contributed the most to the overall effects among the eight PFAS. No association was found between free 25(OH)D and the PFAS mixture. These results suggest that PFAS may affect vitamin D biomarker concentrations in pregnant African American women, and some of the associations were modified by fetal sex.

Prenatal exposure to organophosphate and pyrethroid insecticides and the herbicide 2,4-dichlorophenoxyacetic acid and size at birth in urban pregnant women

BACKGROUND

Organophosphate insecticides and the herbicide, 2,4-dichlorophenoxyacetic acid (2,4-D) are used to protect crops or control weeds. Pyrethroids are used to manage pests both in agriculture and in residences, and to reduce the transmission of insect-borne diseases. Several studies have reported inverse associations between exposure to organophosphates (as a larger class) and birth outcomes but these associations have not been conclusive for pyrethroids or 2,4-D, specifically. We aimed to investigate the association between birth outcomes and urinary biomarkers of pyrethroids, organophosphates and 2,4-D among healthy pregnant women living in New York City.

METHODS

We quantified urinary biomarkers of 2,4-D and of organophosphate and pyrethroid insecticides from 269 women from two cohorts: a) Thyroid Disruption And Infant Development (TDID) and b) Sibling/Hermanos cohort (S/H). We used weighted quantile sum regression and multivariable linear regression models to evaluate the associations between a mixture of urinary creatinine-adjusted biomarker concentrations and birth outcomes of length, birthweight and head circumference, controlling for covariates. We also used linear regression models and further classified biomarkers concentrations into three categories (i: non-detectable; ii: between the limit of detection and median; and iii: above the median) to investigate single pesticides' association with these birth outcomes. Covariates considered were delivery mode, ethnicity, marital status, education, income, employment status, gestational age, maternal age and pre-pregnancy BMI. Analyses were conducted separately for each cohort and stratified by child sex within each cohort.

RESULTS

In TDID cohort, we found a significant inverse association between weighted quantile sum of mixture of pesticides and head circumference among boys. We found that the urinary biomarkers of organophosphate chlorpyrifos, TCPy, and 2,4-D had the largest contribution to the overall mixture effect in the TDID cohort among boys (b = -0.57, 95%CI: -0.92, -0.22) (weights = 0.81 and 0.16 respectively) but not among girls. In the multivariable linear regression models, we found that among boys, for each log unit increase in 3,5,6-trichloro-2-pyridinol (TCPy, metabolite of organophosphate chlorpyrifos) in maternal urine, there was a -0.56 cm decrease in head circumference (95%CI: -0.92, -0.19). Among boys in the TDID cohort, 2,4-D was associated with smaller head circumference in the second (b = -1.57; 95%CI: -2.74, -0.39) and third (b = -1.74, 95%CI: -2.98, -0.49) concentration categories compared to the first. No associations between pyrethroid and organophosphate biomarkers and birth outcomes were observed in girls analyzed in WQS regression or individually in linear regression models in TDID cohort. In the S/H cohort, head circumference increased with higher concentrations of 3-phenoxybenzoic acid (3-PBA, a biomarker of several pyrethroids) (b = 0.53, 95%CI: 0.03, 1.04) among boys and head circumference was lower among girls in the high compared to low category of 2,4-D (b = -2.27, 95%CI: - 3.98, -0.56). Birth length was also positively associated with the highest concentration of 2,4-D compared to the lowest among boys (b = 4.01, 95%CI: 0.02,8.00).

CONCLUSIONS

Weighted quantile sum of pesticides was negatively associated with head circumference among boys in one cohort. Nonetheless, due to directional homogeneity assumption in WQS no positive associations were detected. In linear regression models with individual pesticides, concentrations of TCPy were inversely associated with head circumference in boys and higher concentrations of 2,4-D was inversely associated with head circumference among girls; 2,4-D concentrations were also associated with higher birth length among boys. Concentrations of 3-PBA was positively associated with head circumference among boys.