Hypoglycemic Effect of Resveratrol: A Systematic Review and Meta-Analysis

García-Martínez BI, Ruiz-Ramos M, Pedraza-Chaverri J, Santiago-Osorio E, Mendoza-Núñez VM · Antioxidants

Pooled across everyone, glucose fell 5.24 mg/dL (p=0.002) and insulin 1.23 mIU/L (p=0.0003), while HbA1c and HOMA-IR did not move. Stratified by health status the pattern separates: in participants with type 2 diabetes all four measures improved, glucose by 13.36 mg/dL (p=0.0007), insulin by 0.94 mIU/L (p=0.007), HbA1c by 0.22 percentage points (p=0.02) and HOMA-IR by 0.83 (p=0.04). In participants without diabetes the discussion reports no significant change in any glycaemic measure, which the authors read as resveratrol acting on a pathway that is only disturbed in the first group. By dose, glucose improved at 500 to 1,000 mg/day (7.54 mg/dL, p=0.002) and insulin at under 500 and over 1,000 mg/day, a pattern the authors decline to convert into a therapeutic dose.

Population

Thirty randomised placebo-controlled trials of at least two weeks in adults aged 20 and over, with and without type 2 diabetes

Intervention

Oral resveratrol, doses from 10 mg/day to 3 g/day

Comparator

Placebo

Limitations

Heterogeneity was the reason the subgroups were cut at all, so each stratum is smaller and less balanced than the whole. The non-diabetic null is reported in the discussion rather than as a headline pooled estimate. Trials ran from two weeks to six months, and no glycaemic measure here is a joint measure.