Association of Selenium Levels with Preeclampsia: A Systematic Review and Meta-analysis

Several observational studies have shown inconclusive findings on the association between selenium levels and preeclampsia. This systematic review and meta-analysis was conducted to clarify the association between selenium levels and preeclampsia. The databases PubMed, Google Scholar, ScienceDirect, and the Cochrane Library were searched for studies which investigated selenium levels with preeclampsia and which were published up to April 2022. The overall standardized mean differences (SMD) of selenium levels between cases and controls were measured. Sensitivity analysis, reporting bias, subgroup analysis, and meta-regression were performed for the estimate. The meta-analysis was calculated using the "meta" package in the open-source software R. A total of 26 studies with 1855 preeclampsia cases compared with 3728 healthy pregnant controls were included. The level of selenium was significantly lower in cases of preeclampsia compared with the controls [SMD =  - 0.85; 95% confidence interval (CI): (- 1.46, - 0.25); P =  < 0.01). As there was significant heterogeneity [I² = 96%, Cochran's Q = 620.47; P =  < 0.01], the random effects model was used. A stratified meta-analysis revealed that selenium levels were significantly lower in the cases compared with the controls among pregnant women from the African continent [SMD =  - 1.15 (- 1.65, - 0.65); P =  < 0.01]. Likewise, the same pattern was observed among women from middle- and low-income countries [SMD =  - 1.32 (- 2.22, - 0.42); P =  < 0.01]. None of the investigated factors (Modified Newcastle-Ottawa Scale quality score, year of publication, and sample size) showed significant association with the selenium SMD. The level of certainty of this evidence is "low certainty," as calculated by the GRADEpro GDT online tool. This meta-analysis with low level of evidence certainty revealed that low selenium level is associated significantly with preeclampsia. This pattern is also observed in women from the African continent and women from low- or middle-income countries. Further studies with different prospective designs and detailed patient characteristics are needed to consolidate this evidence.