Association between multiple urinary metals and visceral obesity: a large cross-sectional study in South China

The majority of studies have predominantly used body mass index or waist circumference as measures to establish a link between urinary metals and obesity, leading to inconsistent outcomes. Visceral fat index was a simple, practical and non-invasive physical examination indicator for measuring visceral obesity, and the association between urinary metal and VFI is unclear.

This study utilized a cross-sectional design and based on the baseline data of the Prospective Cohort Study of Chronic Diseases in Ethnic Minority Natural Population in Guangxi from May 2019 to December 2019. Information on demographics, health status, lifestyle, and additional variables was obtained through structured face-to-face interviews. The study employed multiple statistical models, including lasso regression, logistic regression, restricted cubic spline, quantile g-computation, weighted quantile sum, and Bayesian kernel machine regression.

This study encompassed a total of 5794 participants, comprising 2641 males and 3153 females. Among the participants, 1423 (24.6%) were individuals with visceral obesity. The natural log-transformed urinary concentrations of metals are denoted as "ln" followed by the element symbol (e.g., lnSb for antimony, lnZn for zinc, lnRb for rubidium). LnSb (OR = 1.260, 95% CI: 1.149-1.383, P < 0.001) and lnZn (OR = 1.343, 95% CI: 1.134-1.591, P < 0.001) showed a positive association with visceral obesity. In contrast, lnRb (OR = 0.741, 95% CI: 0.624-0.881, P < 0.001) exhibited a negative association with this condition. A dose-response relationship was observed for lnSb (P-overall < 0.001, P-nonlinear = 0.244), lnZn (P-overall = 0.048, P-nonlinear = 0.592), lnRb (P-overall = 0.035, P-nonlinear = 0.482), and lnMg (female: P-overall = 0.005, P-nonlinear = 0.004; 60-74 years: P-overall = 0.089, P-nonlinear = 0.028) in relation to visceral obesity. The mixed effects on visceral obesity revealed significant differences in both positive (P = 0.003) and negative (P = 0.043) directions. In addition, an interaction effect was observed between lnSb (lnRb or lnZn) and the other six metals.

Urinary Sb and Zn as potential risk factors, and Rb as a potential protective factor, for visceral obesity, demonstrating significant overall and interactive effects arising from metal co-exposure. These findings underscored the critical role of environmental metal exposure, especially from mixtures, in the prevention and management of obesity and associated metabolic disorders.