Contribution of diet and other factors to the levels of selected polyfluorinated compounds: data from NHANES 2003-2008.

Contribution of diet and selected risk factors to the levels of four polyfluorinated compounds was evaluated. Data from National Health and Nutrition Examination Survey for the years 2003-2008 were used. Dietary factors accounted for 10.4% to 21.2% of the explained variation. Amount of milk consumed was found to be positively associated (p<0.01) with perfluorononanoic acid (PFNA) but negatively associated with perfluorooctanoate (PFOA) and perfluorooctane sulfonate (PFOS) (p<0.01). Amount of meat and fish consumed was positively associated (p<0.01) with PFNA and PFOS. Amount of non-alcoholic beverages consumed was positively associated (p<0.01) with PFNA and PFOA. Levels of PFOS increased (p<0.01) with increase in the amount of alcoholic beverages consumed. Total amount of alcohol consumed was positively associated (p<0.01) with PFNA. Levels of both PFOA and PFOS decreased with increase in total amount of caffeine consumed. Total amount of fat consumed was negatively associated with PFNA and positively associated with PFOS. Total calories consumed were negatively associated with perfluorohexane sulfonate (PFHxS) and PFOS but positively associated with PFNA. New to this study, positive correlations (p<0.01) between serum cholesterol and PFNA, PFOA, and PFOS were found. Serum albumin levels were negatively correlated with PFHxS but positively correlated with PFOA and PFOS. Males had statistically significantly higher levels of all four PFCs as compared to females and Mexican Americans had the lowest levels of all four PFCs than other race/ethnic groups. Levels of all four PFCs increased with increase in family income. Body mass index was negatively correlated with PFNA but positively associated with PFOA. There was a statistically significant decrease in the levels of PFOS over survey years 2003-2008.