Sustained Impact of Lifestyle and Metformin Interventions on Type 2 Diabetes Prevention: Insights from 21 Years of the US DPP Trial

Highlight

The landmark US Diabetes Prevention Program (DPP) and its extended Outcomes Study (DPPOS) demonstrate sustained reductions in type 2 diabetes incidence over 21 years through intensive lifestyle intervention and metformin. The largest early benefits from lifestyle changes were followed by lasting decreases in diabetes risk. Effectiveness varied by baseline risk profiles, supporting targeted precision approaches.

Study Background

Type 2 diabetes mellitus (T2DM) represents a major global public health challenge due to its increasing prevalence, chronic complications, and health system burden. Prediabetes—a state of impaired glucose regulation—confers a heightened risk for progression to T2DM. Interventions that prevent or delay this progression have substantial clinical and population health importance.

The US Diabetes Prevention Program trial (DPP), initiated in the 1990s, provided robust evidence that intensive lifestyle intervention (ILS) involving diet and exercise and pharmacologic treatment with metformin both reduced the incidence of T2DM in adults with prediabetes, compared with placebo, during a 3-year follow-up. However, the long-term durability of these interventions, their comparative long-term effectiveness, and potential heterogeneity of response across individuals remained important knowledge gaps that could inform clinical practice and population strategies for diabetes prevention.

Study Design

The original DPP was a randomized clinical trial enrolling 3,234 adults with prediabetes in the United States, testing three parallel arms: intensive lifestyle intervention (targeting weight loss and physical activity), metformin therapy (850 mg twice daily), and placebo control.

Following the initial 3-year trial, the study continued under modified protocols as the Diabetes Prevention Program Outcomes Study (DPPOS), extending follow-up to approximately 21 years. Key modifications included discontinuation of the placebo arm, continuation of metformin in the metformin group, and periodic ‘booster’ group lifestyle intervention sessions for the lifestyle group, with lifestyle classes offered to all participants. The primary endpoint remained the incidence of T2DM as defined by American Diabetes Association criteria.

Data cutoff for the present analysis was February 23, 2020, chosen to avoid confounding by disruptions from the COVID-19 pandemic. The analysis followed an intention-to-treat principle, including 3,195 participants with diverse backgrounds and a median follow-up of 8 years (IQR 3.0 to 18.0), with meaningful data up to 21 years.

Key Findings

Long-term Effects on Diabetes Incidence: Over 21 years, both intensive lifestyle intervention and metformin groups demonstrated sustained reductions in the incidence of diabetes compared to placebo, albeit with attenuated effect sizes relative to the initial 3-year trial period.

• Compared to placebo, the lifestyle group had a hazard ratio (HR) for diabetes incidence of 0.76 (95% CI 0.68 to 0.85) and an absolute rate difference (RD) of -1.59 cases per 100 person-years (95% CI -2.25 to -0.93).
• The metformin group showed a HR of 0.83 (95% CI 0.74 to 0.93) with an RD of -1.17 cases per 100 person-years (95% CI -1.85 to -0.49).
• Median diabetes-free survival increased by 3.5 years with lifestyle intervention and 2.5 years with metformin, and 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.

The cumulative incidence curves separated notably within the first 3 years after randomization, evidencing large early effects. Over time, the incidence rates in the lifestyle and metformin groups converged somewhat but remained significantly lower than placebo even at 21 years.

Effect Heterogeneity: The study uncovered important differences in intervention effectiveness according to baseline risk factors:

• Lifestyle intervention produced larger absolute benefits in participants with higher baseline fasting glucose, higher HbA1c levels, and elevated composite clinical and physiological risk scores.
• Metformin benefits were more pronounced in younger participants.

These patterns suggest that individual baseline characteristics modulate long-term response to interventions, emphasizing the potential for precision prevention strategies.

Expert Commentary

The DPP and DPPOS data strongly support the role of both lifestyle modification and metformin as effective, durable strategies to reduce diabetes incidence in people at high risk. The advantages of lifestyle intervention were particularly compelling with sustained absolute risk reduction and prolonged diabetes-free survival. The modest but consistent benefits of metformin, especially in younger adults, underscore its utility as an adjunct or alternative where lifestyle changes are challenging.

From a mechanistic standpoint, lifestyle interventions improve insulin sensitivity and promote weight loss, directly addressing key pathophysiological drivers of T2DM. Metformin acts primarily by reducing hepatic glucose production and improving insulin sensitivity, which may explain differential benefits among demographic and metabolic subgroups.

Nevertheless, it is notable that treatment effect attenuation over time likely results from convergence of behaviors across groups and natural disease progression. The offer of lifestyle classes to all participants during DPPOS and discontinuation of placebo could dilute relative group differences.

Limitations include some loss to follow-up, administrative censoring beyond 21 years limiting longer-term data, and potential challenges in generalizing findings beyond the US population and clinical trial setting. However, the rigorously obtained long-term data represent an unprecedented resource.

Conclusion

The US DPP and its Outcomes Study provide definitive evidence that intensive lifestyle changes and metformin confer substantial and sustained protection against progression to type 2 diabetes across two decades. The magnitude and durability of these effects, combined with observed heterogeneity by baseline risk factors, inform a precision prevention paradigm—targeting interventions to individuals most likely to benefit maximizes clinical effectiveness and resource utilization.

Future research should explore implementation strategies to optimize adherence, further delineate subgroups for tailored prevention, and assess cost-effectiveness in diverse healthcare settings. Addressing the global diabetes epidemic necessitates translating these proven interventions into broad, equitable public health initiatives.

Funding and Clinical Trial Registration

This study was funded by the US National Institute of Diabetes and Digestive and Kidney Diseases and other agencies. The original DPP is registered under ClinicalTrials.gov identifiers NCT00004992 and NCT00038727 for DPPOS.

References

Knowler WC, Doherty L, Edelstein SL, Bennett PH, Dabelea D, Hoskin M, Kahn SE, Kalyani RR, Kim C, Pi-Sunyer FX, Raghavan S, Shah VO, Temprosa M, Venditti EM, Nathan DM; DPP/DPPOS Research Group. 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. Lancet Diabetes Endocrinol. 2025 Jun;13(6):469-481. doi: 10.1016/S2213-8587(25)00022-1. Epub 2025 Apr 28. PMID: 40311647; PMCID: PMC12414453.