Bone mass density following developmental exposures to perfluoroalkyl substances (PFAS):a longitudinal cohort study [Inkl. Correction]

• Published in: Environmental Health
• Main Authors: Blomberg, Annelise, Mortensen, Jann, Weihe, Pál, Grandjean, Philippe
• Format: Article in Journal/Newspaper
• Language: English
• Published: 2022
• Subjects: Bone mass density; Childhood; DXA; Per- and polyfluoroalkyl substances; PFAS; Faroe Islands
• Online Access: https://curis.ku.dk/portal/da/publications/bone-mass-density-following-developmental-exposures-to-perfluoroalkyl-substances-pfas(5d4467ed-21b5-4882-bd9b-002f0cb17f13).html; https://doi.org/10.1186/s12940-022-00929-w; https://curis.ku.dk/ws/files/346048441/s12940_022_00929_w.pdf

Background: Environmental exposures to industrial chemicals, including perfluoroalkyl substances (PFAS), may play a role in bone development and future risk of osteoporosis. However, as prospective evidence is limited, the role of developmental PFAS exposures in bone density changes in childhood is unclear. The objective of this study was to estimate associations between serum-PFAS concentrations measured in infancy and early childhood and areal bone mineral density (aBMD) measured at age 9 years in a birth cohort of children from the Faroe Islands. Methods: We prospectively measured concentrations of five PFAS in cord serum and serum collected at 18 months, 5 years and 9 years, and conducted whole-body DXA scans at the 9-year clinical visit. Our study included 366 mother-child pairs with DXA scans and at least one PFAS measurement. We estimated covariate-adjusted associations of individual PFAS concentrations with age-, sex- and height-adjusted aBMD z-scores using multivariable regression models and applied formal mediation analysis to estimate the possible impact of by several measures of body composition. We also evaluated whether associations were modified by child sex. Results: We found PFAS exposures in childhood to be negatively associated with aBMD z-scores, with the strongest association seen for perfluorononanoic acid (PFNA) at age 5 years. A doubling in age-5 PFNA was associated with a 0.15 decrease in aBMD z-score (95% CI: − 0.26, − 0.039). The PFNA-aBMD association was significantly stronger in males than females, although effect modification by sex was not significant for other PFAS exposures. Results from the mediation analysis suggested that any potential associations between aBMD and 18-month PFAS concentrations may be mediated by total body fat and BMI, although most estimated total effects for PFAS exposures at age 18 months were non-significant. PFAS exposures at age 9 were not associated with age-9 aBMD z-scores. Conclusions: The PFAS-aBMD associations identified in this and previous studies ...