Associations of urinary concentrations of phthalate metabolites, bisphenol A, and parabens with obesity and diabetes mellitus in a Korean adult population: Korean National Environmental Health Survey (KoNEHS) 2015–2017

Abstract

Phthalates and bisphenol A (BPA) have been suspected as risk factors for obesity and diabetes mellitus (DM) among humans. However, associations between phthalates and environmental phenols are often inconsistent across different populations. In this study, we recruited the adult population (n = 3782) of the Korean National Environmental Health Survey (KoNEHS) 2015–2017 (Cycle 3) and assessed the associations between urinary biomarkers of phthalate, BPA, and paraben exposure with obesity and DM. A potential collider issue with the use of urinary creatinine (Cr) or specific gravity (SG) exists when adjusting urinary dilution; therefore, a covariate-adjusted standardization (CAS) was employed for adjustment, and the results were compared. In the present population, the direction of the association often varied depending on the choices made to adjust urinary dilution. When using CAS, the direction of association resembled those of previously reported experimental observations. With Cr or SG adjustment, ORs for obesity decreased in the highest quartiles of monocarboxyoctyl phthalate (MCOP) [OR (95% CI) = Cr: 0.71 (0.54, 0.93); SG: 0.68 (0.52, 0.90)], monocarboxy-isononyl phthalate (MCNP) [OR (95% CI) = Cr: 0.67 (0.52, 0.87); SG: 0.68 (0.52, 0.89)], and mono(3-carboxylpropyl) phthalate (MCPP) in the urine [OR (95% CI) = Cr: 0.60 (0.47, 0.76); SG: 0.61 (0.48, 0.77)]; however, with CAS, these negative associations disappeared. Instead, mono-benzyl phthalate (MBzP) [OR (95% CI) = 1.31 (1.03, 1.66)], BPA [OR (95% CI) = 1.62 (1.27, 2.06)], or ethyl paraben (EtP) [OR (95% CI) = 1.51 (1.19, 1.91)] concentrations in the highest quartile showed positive associations with a higher risk of obesity. On the other hand, for DM, an overall decrease in ORs was observed for phthalate metabolites and BPA following SG adjustment and disappeared with CAS adjustment. In addition, the highest quartiles of BPA, methyl paraben (MeP), and ethyl paraben (EtP) showed a significantly higher risk of DM than those in the lowest quartiles following CAS [OR (95% CI) = BPA: 1.65 (1.06, 2.59); MeP: 1.68 (1.08, 2.60); and EtP: 2.74 (1.77, 4.24), respectively]. The present observations outline the importance of using an appropriate adjustment method for urinary dilution in association studies on obesity and DM. In addition, several phthalates, BPA, and parabens were identified as potential chemical risk factors for these outcomes. Further studies are warranted in other populations to confirm these observations.