Serum glycolipids mediate the relationship of urinary bisphenols with NAFLD: analysis of a population-based, cross-sectional study

BACKGROUND: Bisphenol A (BPA) and its substitutes bisphenol S (BPS) and bisphenol F (BPF) are endocrine-disrupting chemicals widely used in consumer products, which have been proposed to induce various human diseases. In western countries, one of the most common liver diseases is non-alcoholic fatty liver disease (NAFLD). However, studies on the associations of the three bisphenols with NAFLD in human beings are scarce. METHODS: We included 960 participants aged ≥ 20 years from the NHANES 2013–16 who had available data on levels of urinary BPA, BPS and BPF. The hepatic steatosis index (HSI) > 36 was used to predict NAFLD. Logistic regression analysis and mediation effect analysis were used to evaluate the associations among bisphenols, glycolipid-related markers and NAFLD. RESULTS: A total of 540 individuals (56.3%) were diagnosed with NAFLD, who had higher concentrations of BPA and BPS but not BPF than those without NAFLD. An increasing trend in NAFLD risks and HSI levels was observed among BPA and BPS tertiles (p for trend < 0.05). After adjustment for confounders, elevated levels of BPA or BPS but not BPF were significantly associated with NAFLD. The odds ratio for NAFLD was 1.581 (95% confidence intervals [CI]: 1.1–2.274, p = 0.013) comparing the highest with the lowest tertile of BPA and 1.799 (95%CI: 1.2462.597, p = 0.002) for BPS. Mediation effect analysis indicated that serum high-density lipoprotein cholesterol and glucose had a mediating effect on the relationships between bisphenols and NAFLD. CONCLUSIONS: The present study showed that high exposure levels of BPA and BPS increased NAFLD incidence, which might be mediated through regulating glycolipids metabolism. Further studies on the role of bisphenols in NAFLD are warranted. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12940-022-00945-w.