| The National Health and Nutrition Examination Survey (NHANES), conducted by the Centers for Disease Control and Prevention, collects biomonitoring measurements on a representative sample of the U.S. population. This dissertation analyzed biomonitoring data on polyfluoroalkyl chemicals (PFCs) and bisphenol A (BPA), both endocrine disrupting chemicals, to address three key environmental health research questions.;Chapter 2 studied the relationship between serum concentrations of four PFCs and cholesterol levels, obesity, and insulin resistance in 2003-2004 data. In adults, three PFCs were positively associated with total and non-HDL cholesterol. We did not see consistent associations with obesity or insulin resistance. Though studies in workers and highly exposed communities found similar results, these are the first reported for the general population.;Chapter 3 analyzed 2003-2004 data to determine whether intake of certain foods predicted serum PFCs. We found that consumption of fast food, take-out hot beverages, red meat, potatoes, and snack foods was associated with certain PFCs. These findings indicate that PFCs may enter the food chain through both bioaccumulation and food packaging materials.;Chapter 4 investigated differences by socioeconomic position (SEP) and race/ethnicity in body burdens of PFCs and BPA in 2003-2006 data. Urinary BPA was higher in people with lower incomes, whereas the opposite was true for serum PFCs. Children with very low food security or who received emergency food had particularly high BPA levels. Mexican Americans had the lowest levels of both chemicals of any racial/ethnic group. The evidence that diet contributes to the SEP differences was stronger for BPA than PFCs. Income, education, occupation, and food security were not interchangeable as measures of SEP; environmental epidemiology studies should give more consideration to the choice of measures.;Biomonitoring surveillance programs such as NHANES present a tremendous resource to environmental health researchers. Though limited in their ability to examine exposure-disease relationships, these data can provide useful hypotheses for future study. The opportunity they offer to investigate predictors of and disparities in exposure is unparalleled. The Boston Consensus Conference on Biomonitoring, which gathered input from the lay public on questions surrounding biomonitoring, provides important additional insights about the conduct of biomonitoring surveillance programs. |
