Long-term effects and effect heterogeneity of lifestyle and metformin interventions on type 2 diabetes incidence over 21 years in the US Diabetes Prevention Program randomised clinical trial.
Level 2 - randomized trial
Long-term follow-up analysis of an individual randomized controlled trial cohort.
PubMed 40311647 · doi:10.1016/S2213-8587(25)00022-1
What was done
Long-term intention-to-treat analysis over up to 21 years in 3,195 adults with prediabetes (67.9% female, mean baseline age 50.6 years) originally randomized in the US Diabetes Prevention Program (DPP) to intensive lifestyle intervention (ILS), metformin (850 mg twice daily), or placebo. In the subsequent DPP Outcomes Study (DPPOS), placebo was stopped, metformin was unmasked and continued, ILS participants received bi-annual booster classes, and all groups were offered quarterly lifestyle sessions. The primary outcome was type 2 diabetes incidence defined by American Diabetes Association criteria, analyzed through February 23, 2020.
What was found
Over follow-up (median 8.0 years, IQR 3.0 to 18.0; up to 21 years analyzed), diabetes incidence rate was reduced compared with the original placebo group in the ILS group (hazard ratio [HR] 0.76 [95% CI 0.68 to 0.85]; rate difference [RD] -1.59 cases per 100 person-years [95% CI -2.25 to -0.93]) and the metformin group (HR 0.83 [95% CI 0.74 to 0.93]; RD -1.17 cases per 100 person-years [95% CI -1.85 to -0.49]). Median diabetes-free survival increased by 3.5 years with ILS and 2.5 years with metformin; mean diabetes-free survival increased by 2.0 years (95% CI 1.2 to 2.8) and 1.2 years (95% CI 0.4 to 2.0), respectively. Incidence curves separated early (first 3 years) and converged over time. Absolute risk reductions were greater with ILS in participants with higher baseline fasting glucose, HbA1c, and composite risk indices, and greater with metformin in younger participants.
Why it matters
This study demonstrates that the early preventive benefits of lifestyle modification and metformin in individuals with prediabetes produce durable, multi-decade delays in diabetes onset.
Limits
Protocol modifications after the core trial—including unmasking, discontinuation of placebo, and provision of lifestyle classes to all participants—complicate direct long-term comparisons. Median follow-up was 8.0 years despite the 21-year analysis horizon, and numbers at risk decreased sharply after 21 years.
Cited by
- context In the Diabetes Prevention Program (DPP), a gentle lifestyle intervention was three times more effective than metformin.