Analysis of PFAS Levels and Associations with Adolescent Anthropometric Outcomes in NHANES

Abstract

Perfluoroalkyl and polyfluoroalkyl substances (PFAS) are found in some consumer products due to their heat resistance and durability. However, there is potential for these substances to bioaccumulate in humans. It is relevant to investigate biological effects of these chemicals, as studies have suggested early life exposure may impact human developmental outcomes such as infant birth weight and youth adiposity. The objective of the current study was to determine if a relationship exists between increasing levels of certain PFAS and anthropometrics in adolescents ages 12 - 18. The three PFAS examined were: perfluorodecanoic acid (PFDeA), 2-(N-methyl-perfluoroctane sulfonamido) acetic acid (Me-PFOSA-AcOH), and perfluoroundecanoic acid (PFUA). The data was obtained from the National Health and Nutrition Examination Survey (NHANES) from the years 2011-2012 (N = 287) and 2013-2014 (N = 344). An additional analysis combined data from 3 NHANES survey cycles using sampling weights for the years 2011-2016 (N = 875) to generate a larger sample size of detectable PFAS. PFAS concentrations were classified as above or below the lower limit of detection (LLOD) to evaluate differences in weight, waist circumference, BMI (body mass index), and height using Student’s t-tests. These same anthropometric outcomes were examined as continuous variables in linear regression models and were stratified by sex. In the 2013-2014 dataset, there were significant inverse associations between female concentrations of PFUA and PFDeA with waist circumference (PFUA β = −0.056; 95% CI, −0.106, −0.005; PFDeA β = −0.06; 95% CI, −0.10, −0.02), weight-for-age z-score (PFUA β = −0.40; 95% CI, −0.74, −0.05; PFDeA β = −0.38; 95% CI, −0.64, −0.12), and BMI-for-age z-score (PFUA β = −0.48; 95% CI, −0.86, −0.10; PFDeA β = −0.45; 95% CI, −0.73, −0.16). In the 2011-2012 dataset, males displayed a significant inverse relationship between PFDeA and waist circumference (β = −0.08; 95% CI, −0.14, −0.02), weight-for-age z-score (β = −0.49; 95% CI, −0.88, −0.11), and BMI-for-age z-score (β = −0.44; 95% CI, −0.84, −0.05). In the combined analysis of NHANES years 2011-2016, there were significant inverse associations with PFUA and PFDeA and weight-for-age z-score, waist circumference, and BMI-for-age z-score. In the given sample years, there was no compelling evidence for a relationship between any of the perfluoroalkyl chemicals and height, nor between Me-PFOSA-AcOH and any of the body measures after adjusting for age, sex, and race/ethnicity. This suggests that PFUA and PFDeA exposure in adolescents may be related to smaller waist circumference, weight, and BMI, but longitudinal studies are recommended to confirm these findings.