Abstract
Obesity is a major threat to health, but the etiology of obesity is incompletely understood. Phthalates, synthetic chemicals ubiquitous in the environment, are suspected to have obesogenic effects, but the relationship of phthalates and obesity in humans remains uncertain. We examined whether phthalate exposure was associated with body fat gain in midlife women. We analyzed data from 1369 women in the Study of Women's Health Across the Nation Multi-Pollutant Study. Eleven phthalate metabolites measured in spot urine samples at baseline (1999/2000) were standardized with covariate-adjusted creatinine. Body weight (BW), fat mass (FM) from dual-energy X-ray absorptiometry (DXA), and body fat percentage (BF%) from DXA were measured near-annually until 2016/2017. For each metabolite, linear mixed effects models with time and log(2)(metabolite) interactions were examined, adjusting for demographic, lifestyle, and menopause-related factors. Analyses were conducted overall and stratified by baseline obesity status. As sensitivity analyses, all analyses were repeated using a second set of metabolites measured in 2002/2003. Higher levels of all metabolites except mono-carboxy-isononyl phthalate were associated with faster increases in BF%. Per doubling of metabolite concentrations, differences in five-year BF% change ranged from 0.03 percentage point (ppt) (95% confidence interval (CI): -0.03, 0.09) for mono-isobutyl phthalate to 0.09 ppt (95% CI: 0.02, 0.16) for mono(3-carboxypropyl) phthalate. Results were similar for FM change, but associations with BW change were mostly null. In stratified analyses by baseline obesity status, positive associations were strongest in women who were normal/underweight at baseline. When metabolites from 2002/2003 were used as exposures, most associations were attenuated and not statistically significant, but they remained positive for normal/underweight women. In conclusion, phthalate metabolites were associated with more rapid body fat gain in midlife women, but our results need confirmation given attenuation of estimates in the sensitivity analyses.