Effectiveness of Smartphone Application-Based Interventions to Prevent Type 2 Diabetes Mellitus in Individuals With Prediabetes: A Systematic Review and Meta-Analysis.

Abstract

Smartphone application (app)-based interventions to prevent Type 2 diabetes (T2D) are becoming increasingly available. A thorough summary of their effectiveness is lacking. We synthesized evidence on the effectiveness of app-based interventions to prevent T2D targeting individuals with prediabetes. For this systematic review and meta-analysis, we searched Web of Science, Embase, Scopus, PubMed Central, and Medline between January 1, 2013, and January 31, 2024, to identify randomized controlled trials (RCTs) that assessed the effectiveness of app-based interventions to prevent T2D targeting individuals with prediabetes, published in English, without restrictions regarding the effectiveness outcome. We synthesized all outcomes graphically via effect directions and conducted meta-analyses for clinical outcomes, including the Risk of Bias 2 Tool. This study was prospectively registered with PROSPERO (CRD42023491693) and OSF (DOI 10.17605/OSF.IO/B89QP). Of 9703 articles, 18 RCTs were included in the systematic review, and 15 RCTs in the meta-analysis. We found statistically significant reductions in body weight (-1.35 kg, 95% CI: [-2.48; -0.23], N = 13 RCTs), body mass index (-0.53 kg/m², 95% CI: [-0.97; -0.09], N = 11 RCTs), and glycated hemoglobin (-0.08%, 95% CI: [-0.10; -0.05], N = 11 RCTs) and point estimates and/or effect directions predominantly suggesting improvements for additional outcomes. Stratified meta-analyses showed no statistically significant between-group differences and missing evidence for long-term effectiveness and equity-relevant subgroups. Our study shows that app-based interventions can improve outcomes (i.e., motivation, behavior, and clinical parameters) in individuals with prediabetes. The effect on clinical outcomes is small. Evidence on equity impacts, long-term effectiveness, and diabetes incidence is missing and remains to be investigated.