RESUMO
AIMS: Direct and indirect evidence were combined in this systematic-review and network meta-analysis (NMA) to assess and compare the effect of nutritional supplements on glycemic control, and rank the supplements accordingly. METHODS: PubMed, Scopus, and Web of Science were searched up to April 2021. We included randomized controlled trials that investigated the effect of vitamins D, C, and E, magnesium, zinc, calcium, selenium, and omega-3 on at least one glycemic marker, including glycated hemoglobin (HbA1c), fasting blood sugar (FBS), homeostasis model assessment-estimated insulin resistance (HOMA-IR), HOMA-B, and insulin, in adults with type 2 diabetes. To estimate effectiveness of supplements, a random-effects NMA in the Bayesian framework was applied. To assess risk of bias, Cochrane Collaboration Tool was used. RESULTS: Analysis of 178 studies indicated that zinc, vitamin D, omega-3, vitamin C, and vitamin E were effective in reducing HbA1c with low certainty. For reduction of FBS, zinc, vitamin D, and vitamin C, and for HOMA-IR, vitamin D were effective with low certainty. None of the supplements were effective in the reduction of insulin and HOMA-B with low certainty. After excluding poor-quality studies, only vitamin D was significantly effective in reducing all of the markers. Consistently, when the analysis was restricted to studies with a duration of ≥12-weeks, vitamin D reduced HbA1c, FBS, and HOMA-IR. CONCLUSIONS: Vitamin D supplementation was more effective compared to other supplements in improving HbA1c, FBS, and HOMA-IR, albeit with low certainty of evidence. This result was confirmed by low-risk of bias studies. REGISTRATION: CRD42021240691.