We focused on the 149 environmental pollutant variables that were measured in both stage 1 and 2 samples. The basic idea of an EWAS, like GWAS, is to conduct an agnostic search in a broad set of environmental compounds without any prior belief or hypothesis regarding the effects related to a given outcome. As our study was based on such a non-targeted approach and had no a priori assumption of the association directions, chemicals known to be less toxic, such as arsenosugars, were not screened out. For the concentrations below the National Centers for Health Statistics (NCHS) documented limit of detection (LOD), the values of each pollutant’s LOD/√2 were replaced. We eliminated 15 variables that had more than 90% of the observations missing (including missing due to below LOD), leaving 134 pollutants available for our analysis (Table S1). As stated above the four outcome variables included total cholesterol, HDL, LDL and triglycerides. Important covariates were chosen a priori and included age, sex, race/ethnicity (Mexican American, Other Hispanic, non-Hispanic white, non-Hispanic black, Other), education (categorized to less than high school diploma,
