RESUMO
Several studies have investigated the association between selenium levels and gestational diabetes mellitus (GDM); however, their results are not conclusive. This systematic review and meta-analysis aimed to update and draw conclusions regarding the evidence from published studies that investigated selenium levels in relation to GDM. PubMed, Google Scholar, Cochrane Library and ScienceDirect were searched for studies related to selenium and GDM, published from the inception of each database through to July 2022. The meta-analysis was conducted by measuring the standardized mean difference (SMD) between the selenium levels of women with GDM and those pregnant without GDM (control group). Stratified meta-analysis, meta-regression analysis and reporting bias were applied. The "meta" package in the open-access software R was used to analyze all of the data. A total of 12 studies, including 940 pregnant women with GDM and 1749 controls met this study's inclusion criteria. The selenium levels were significantly lower in women with GDM compared with the control group (SMD = -0.66; 95% confidence interval (CI): (-1.04, -0.28); p ≤ 0.001). Due to significant heterogeneity (I2 = 94%, Cochrane Q = 186.7; p ≤ 0.0001), the random-effects model was followed. The stratified meta-analysis showed that the selenium levels were lower in the cases compared with the normal controls in the third trimester (SMD = -1.85 (-3.03, -0.66); p ≤ 0.01). The same trend was observed in the studies published before the year 2014 (SMD = -0.99 (-1.70, -0.28); p ≤0.01) and those published in or after 2014 (SMD = -0.45 (-0.90, 0.00); p = 0.05). None of the investigated covariates in the meta-regression analysis (each study's geographic location, trimester of selenium quantification, World Bank economic classification, method of selenium determination, study design, study quality score, publication year and study's sample size) were significantly associated with the selenium SMD. The current evidence indicates that selenium levels are lower among women with GDM in comparison to those without GDM; however, after the correction of the reporting bias, the result was no longer significant. Further studies with more prospective designs are needed to confirm this evidence and explain the function of selenium in GDM throughout pregnancy.