Abstract
BACKGROUND: The prevalence of prediabetes and, consequently, type 2 diabetes is increasing around the world. Previous meta-analyses reported controversial findings regarding the association between vitamin D supplementation with glycemic control and diabetes risk. This comprehensive meta-analysis summarized existing research to provide an estimate of the impact of vitamin D supplementation on metabolic parameters and diabetes risk in individuals with prediabetes. METHOD: A comprehensive systematic search was conducted across the Web of Science, Scopus, PubMed, and Cochrane databases, and Google Scholar using relevant keywords until 22 July 2025. The AMSTAR2 scale was used to evaluate the methodological quality of the included articles. Moreover, the certainty of the evidence was assessed using the GRADE tool. Stata 17 software was used for data analysis. RESULTS: Fourteen meta-analyses comprising 31 randomized controlled trials (RCTs) of 3856 prediabetic patients were included in this review. Combining the findings of RCTs revealed that vitamin D supplementation significantly reduced the levels of fasting blood sugar (WMD= -0.377 mg/dl, 95% CI (-0.589, -0.165), p < 0.001), insulin (WMD = -0.174 µU/mL, 95% CI (-0.274, 0.074), p < 0.001), hemoglobin A1c (WMD = -0.479%, 95% CI (-0.714, -0.245), p < 0.001), and serum triglyceride (TG) (WMD = -0.385 mg/dl, 95% CI (-0.622, -0.147), p = 0.002) in comparison with the control group. The effects of vitamin D on insulin resistance by homeostasis model assessment, 2-hour oral glucose tolerance test plasma glucose, homeostasis model assessment of β-cell function, body mass index, and diabetes risk of participants were not significant. CONCLUSION: The findings of this umbrella review suggested that vitamin D supplementation could help to improve some glycemic indices and TG levels. However, due to discrepancies among the results, more well-designed RCTs are warranted to confirm and clarify the impacts of vitamin D supplementation in prediabetic patients. CLINICAL TRIAL NUMBER: Not applicable.