Abstract
BACKGROUND: Recent studies have shown that the relationship between mercury exposure and diabetes is controversial. The aim of this study is to determine the relationship between mercury exposure and diabetes using a systematic review and meta-analysis approach. METHODS: We systematically searched PubMed, Web of Science, Cochrane, and Embase databases for cross-sectional, case-control, or cohort studies assessing the correlation between mercury exposure and diabetes in any population. Details of each included study were extracted using a pre-designed Excel spreadsheet. Quality assessment of cohort and case-control studies used the Newcastle-Ottawa Scale (NOS), whereas cross-sectional studies were assessed by the Agency for Healthcare Research and Quality (AHRQ) scale. Meta-analyses were performed using random-effects models to calculate the pooled odds ratio (OR), standardized mean difference (SMD), and their 95% confidence intervals (CIs). Subgroup and sensitivity analyses were employed to assess heterogeneity sources. Begg's and Egger's tests were used to evaluate publication bias. RESULTS: Our meta-analysis included 8 eligible articles, comprising a total of 40,891 subjects, reporting mercury OR and/or concentrations. Among the included studies, one was a case control, one was a cohort study, and the rest were cross-sectional studies. Two studies were rated as high quality and six as medium quality. The results revealed no link between mercury exposure and diabetes (OR: 1.11, 95% CI: 0.80, 1.55, n=6, I(2)=73.7%; and SMD: 0.41, 95% CI: -0.32, 1.14, n=3, I(2)=88.7%). In the stratified male and female subgroups, the pooled OR was 0.71 (95% CI: 0.57, 0.90, n=3, I(2)=0.0%), 1.11 (95% CI: 0.69, 1.79, n=3, I(2)=67.7%). The Begg's test results revealed no significant publication bias (P=0.06), but the Egger's test results did (P=0.013). The sensitivity analysis confirmed the stability of our results. CONCLUSIONS: No significant relationship was observed between mercury and diabetes mellitus. However, more well-designed studies on mercury exposure and diabetes risk are still needed, particularly on the type of mercury (i.e., elemental, inorganic, and organic), exposure time and dose, type of biological specimen, and the population's sex and age.