Identifying the characteristics of individuals with high body burdens of endocrine disrupting chemicals: An analysis of the National Health and Nutrition Examination Survey (NHANES) 1999-2002

Abstract

Characterizing adverse health affects associated with endocrine disruptors has proved difficult as few epidemiological studies have utilized validated exposure data or adequate knowledge of exposed individuals within the general population (Landrigan et al., 2003; WHO, 2002). Given the pervasive nature of endocrine disruptors and their ability to bioaccumulate in food chains, these compounds may constitute a serious public health problem with widespread consequences. In order to characterize which populations are at risk of the potentially detrimental effects of endocrine disruptors, a series of analytical tools, including classification and regression trees, descriptive analysis and stratified regression models, were applied to the bio-monitoring data collected by the US National Health and Nutrition Examination Survey (NHANES) for the years of 1999 to 2002. Overall, 7.8-33.7% of the variability in exposure to non-persistent endocrine disrupting chemicals could be accounted for by selected variables such age, race/ethnicity, country of birth, occupation, marital status, and rooms treated for pests. However, these results should be interpreted with caution as a linear approach to creatinine adjustment increased the r2 value in these models. In fact, very little of the variability could be accounted for using the aforementioned variables, indicating that exposures to non-persistent chemicals are difficult to predict using proxy measures. For persistent chemicals, models were successful in predicting between 22.53-51.21% of the variability in exposure levels, with age proving to be the strongest predictor. Other variables found to significantly predict exposures were race, country of birth, BMI, and marital status. Despite the fact that the predictors determined in this study are often tied to a wide range of risk factors for disease, other chemical exposures, and socioeconomic factors, these results will allow future epidemiological studies and public health interventions to effectively target exposed populations within the general public.