Abstract
Background
Environmental exposures such as perfluoroalkyl substances (PFASs) were considered potential risks for bone mineral density (BMD).
Objective
To examine the associations between PFASs and BMD among the U.S. population.
Methods
This study included a total of 6,817 participants from the National Health and Nutrition Examination Survey (NHANES 2005-2014). Multiple linear regression models were used to analyze the associations between serum PFASs and BMD. Covariates such as age, race, BMI, smoking, alcohol intake, milk intake, physical activity, and previous prednisone treatment were adjusted in these models. Additionally, gender and menopausal period were considered in further subgroup analyses.
Results
Five PFASs (PFOA, PFOS, PFHxS, PFDE and PFNA) with detection rate (DR) higher than 70% were selected in the analyses of the associations with BMD. Using combined data of NHANES 2005-2014, the effects on BMD were found significant gender differences. Negative associations for PFOA (β = -0.0110; 95% CI: -0.0200, -0.0020), PFOS (β = -0.0116; 95% CI: -0.0191, -0.0041), PFHxS (β = -0.0081; 95% CI: -0.0140, -0.0021) and PFDE (β = -0.0134; 95% CI: -0.0219, -0.0048) were found in premenopausal women, while positive association for PFOA(β = 0.0112; 95% CI: 0.0027, 0.0196), PFOS (β = 0.0084; 95% CI: 0.0014, 0.0153), PFHxS(β = 0.0063; 95% CI: 0.0003, 0.0123), PFDE(β = 0.0078; 95% CI: 0.0009, 0.0148) and PFNA(β = 0.0112; 95% CI: 0.0034, 0.0190) were found in men. In further subgroup analyses, women in pre-menopause statues showed consistent negative associations.
Conclusion
Serum PFAS concentrations and femur BMD were associated with gender and menopause status differences. Most negative associations were limited to premenopausal women, and most positive associations with femur BMD were found in men. Further epidemiological evidence and studies about the mechanisms are needed.