Objective: To summarize the evidence regarding the relationship between the circulating selenium level and stroke based on observational studies.Methods: The electronic databases of PubMed, Web of Science, and Embase were searched up to October 2020, for observational studies concerning the association between the circulating selenium level and stroke. The pooled relative risk (RR) of stroke for the highest versus lowest category of the circulating selenium level and the weighted mean difference (WMD) of the circulating selenium level for stroke versus control subjects, as well as their corresponding 95% confidence interval (CI) were calculated.Results: A total of 11 articles with 12 observational studies were included (5 prospective cohort, 4 case-control, and 3 cross-sectional). The overall multivariable-adjusted RR demonstrated that the circulating selenium level was negatively associated with stroke (RR = 0.48, 95%CI: 0.24 to 0.94; p = 0.033), which was further confirmed in cross-sectional/case-control (RR = 0.24, 95%CI: 0.10 to 0.63; p = 0.003) and whole blood selenium (RR = 0.41, 95%CI: 0.26 to 0.65; p = 0.004) studies, rather than prospective cohort (RR = 0.88, 95%CI: 0.52 to 1.48; p = 0.62) and serum/plasma selenium (RR = 0.55, 95%CI: 0.22 to 1.40; p = 0.21) studies. Moreover, the overall combined WMD showed that the circulating selenium level in stroke was lower than that in control subjects (WMD = -0.16, 95%CI: -0.29 to -0.03; p = 0.017), which was further confirmed in whole blood selenium (WMD = -0.26, 95%CI: -0.44 to -0.08; p = 0.005), rather than serum/plasma selenium (WMD = -0.10, 95%CI: -0.27 to 0.07; p = 0.25) studies.Conclusions: Although our results suggest that the circulating selenium level is inversely associated with stroke, current evidence is still insufficient to conclude their definite association. More well-designed prospective cohort studies with detailed selenium biomarker specification are needed to elaborate the concerned issues further.
Keywords: Circulating selenium level; clinical nutrition; meta-analysis; observational studies; stroke.