Interventions for prediabetes: an umbrella review of systematic reviews and meta-analyses of randomized controlled trials

Aims: This umbrella review synthesized evidence from systematic reviews and meta-analyses of randomized controlled trials (RCTs) to assess the effectiveness of interventions for preventing adverse outcomes in individuals with prediabetes. Methods: A total of 23 meta-analyses comprising 602 RCTs and over 30,000 participants were analyzed from databases including Medline, Embase, Web of Science, and Cochrane through February 1, 2025. Interventions evaluated included both pharmacological and non-pharmacological strategies compared to placebo or usual care. Studies reporting only bio-humoral markers were excluded. The GRADE approach was used to assess the certainty of evidence. Results: Among 15 evaluated interventions, four were supported by high-certainty evidence for diabetes prevention: GLP-1 receptor agonists (RR = 0.26), liraglutide/exenatide/semaglutide (OR = 0.29), orlistat (OR = 0.67), and structured self-care programs (OR = 0.58). High-certainty evidence also supported modest reductions in systolic blood pressure (∼2-3 mmHg) from lifestyle interventions, digital health tools, and liraglutide. Significant reductions in BMI and waist circumference were observed with pharmacologic agents, aerobic exercise, and digital platforms. However, metformin combined with lifestyle changes did not yield notable anthropometric benefits. Conclusion: These findings underscore the value of a multidimensional and personalized approach to prediabetes management, emphasizing evidence-based pharmacological options alongside behavioral and digital health strategies. Systematic review registration: PROSPERO CRD42025649619.