Two decades of diabetes prevention: Sustained benefits, heterogeneous effects, and implications for precision prevention

The global burden of type 2 diabetes continues to rise, driven by population aging, urbanization, and the increasing prevalence of obesity and sedentary lifestyles. In Asia, including Japan, Korea, and China, the rapid pace of change in dietary and physical activity patterns has created a parallel surge in diabetes prevalence, making prevention an urgent priority¹. Despite significant progress in glucose-lowering therapies, the most cost-effective and sustainable approach remains preventing diabetes before it develops. Few studies have shaped this field more profoundly than the Diabetes Prevention Program (DPP) and its long-term follow-up, the Diabetes Prevention Program Outcomes Study (DPPOS). The DPP, published in 2002, demonstrated that intensive lifestyle intervention reduced the incidence of type 2 diabetes by 58% and metformin by 31% over 3 years compared with placebo². These results established lifestyle modification as the gold standard for diabetes prevention and validated metformin as an adjunct option in selected individuals. These findings became the cornerstone of diabetes prevention strategies worldwide. Yet a critical question has remained: Are the benefits of these interventions durable over the long term?

In their recent publication in The Lancet Diabetes & Endocrinology, Knowler et al.³ provide a long-term perspective of the DPP cohort, with follow-up extending approximately 21 years. Compared with placebo, diabetes incidence rate was reduced by 24% in the intensive lifestyle intervention group and by 17% in the metformin group, with a corresponding delay in diabetes onset by a median of 3.5 and 2.5 years (Table 1). Both interventions reduced cumulative diabetes incidence, although the absolute risk differences narrowed over time. The critical insight is that most of the benefit accrued during the first 3 years of the trial, when intervention intensity and adherence were highest. The curves diverged early and remained separated, demonstrating a “legacy effect” that persisted for two decades. This long-term durability reinforces a central principle of prevention that early and intensive interventions leave a lasting imprint on disease trajectories. Similar patterns were observed in the Da Qing study in China, where modest lifestyle changes implemented over 6 years translated into lower rates of diabetes and cardiovascular events over 30 years⁴. In that trial, 577 adults with impaired glucose tolerance were randomized to diet, exercise, diet plus exercise, or control. After 6 years of intervention, diabetes incidence was significantly reduced in the lifestyle groups compared with control. Importantly, long-term follow-up over 30 years revealed not only a persistent reduction in diabetes risk but also meaningful clinical outcomes: a 26% lower incidence of cardiovascular disease, a 35% reduction in microvascular complications, a 33% lower cardiovascular mortality rate, and a 26% lower all-cause mortality rate in the intervention groups. Together, these studies demonstrate that time-limited interventions can yield long-term benefits—a message that should encourage health systems to invest in structured, resource-intensive prevention efforts rather than settling for conventional maintenance strategies.

A second contribution of the 21-year report is its detailed exploration of effect heterogeneity. Lifestyle intervention conferred the greatest absolute benefit among individuals with higher baseline fasting glucose, glycated hemoglobin (HbA1c), and clinical and physiological risk indices, whereas metformin showed the clearest benefit in younger adults, particularly those under 60 years of age. In contrast, older adults appeared to derive little advantage from metformin³. These insights emphasize that “one-size-fits-all” approaches to prevention could be suboptimal. Instead, risk stratification can identify subgroups where absolute benefits are greatest, allowing for precision prevention, thereby optimizing resource allocation.

The study also reveals the persistent challenge of adherence. Intensive lifestyle intervention led to significant weight loss during the early years, but weight regain diminished its long-term impact. By 16 years, mean body weight was lower in the metformin group than in the lifestyle group, underscoring the difficulty of maintaining behavioral changes over decades³. Similarly, participants taking metformin declined steadily, with nearly one-third of participants discontinuing therapy before diabetes onset. These findings illustrate the dual challenge of behavior change and long-term pharmacotherapy, which both are difficult to sustain in otherwise healthy individuals who may feel little urgency to adhere to preventive strategies. This finding mirrors clinical practice, where the engagement of patients often wanes after initial counseling or when facing the prospect of indefinite medication use. Looking ahead, sustaining prevention will require innovative solutions. Digital health platforms can deliver ongoing coaching and interactive feedback. Community-based peer support, behavioral economics approaches such as financial incentives, and periodic booster interventions may help re-engage participants who lose momentum. Pharmacological innovation also offers new possibilities: Long-acting formulations, agents with additional benefits such as weight reduction or cardiovascular protection, and combination approaches that reduce treatment burden could all play a role.

The timing of the DPPOS publication is particularly relevant given the rise of powerful new anti-obesity and anti-diabetes therapies, including glucagon-like peptide-1 receptor agonists and dual incretin agonists⁵. These agents induce significant weight loss, improve glycemia, and appear to substantially reduce progression from prediabetes to diabetes. Their effects, at least in the short to medium term, exceed what was achievable with lifestyle programs or metformin in the DPP. Yet the DPPOS reminds us that any new therapy must be judged not only on short-term efficacy but also on long-term safety, durability, and cost-effectiveness. In contrast to high-cost injectables, metformin remains inexpensive and widely available, while lifestyle change yields broad health benefits—including cardiovascular, musculoskeletal, and mental health advantages—that no pill or injection can fully replicate. Thus, the findings of the DPPOS remain foundational, even in the era of novel therapies.

The clinical and policy implications are clear. First, diabetes prevention must begin early and with intensity. Intervening at the earliest stages of dysglycemia yields benefits that endure for decades, even if adherence later declines. Second, prevention should be targeted, not uniform. Risk stratification by age, glycemia, and metabolic profile allows us to identify who stands to gain the most. Third, sustained prevention requires strategies that extend beyond the initial intervention phase. Policymakers and clinicians must design programs that combine early intensity with innovative tools for long-term engagement.

As with any long-term trial, interpretation of these results requires caution. The study protocol evolved over time: placebo was discontinued, and all participants were eventually offered group lifestyle interventions, blunting between-group contrasts. Declining adherence and crossover may have further diluted treatment effects. Subgroup analyses may be underpowered, and the trial predated the widespread use of HbA1c as a diagnostic tool and the availability of modern pharmacotherapies, limiting direct applicability to contemporary practice. Nonetheless, the DPPOS remains unparalleled in its duration, methodological rigor, and relevance to clinical decision-making.

The 21-year follow-up of the DPP provides definitive evidence that both lifestyle modification and metformin delay diabetes onset, with benefits that persist for decades. The findings emphasize the enduring value of prevention, while also exposing the difficulty of sustaining adherence and the importance of targeting interventions to those at highest risk. As the field embraces powerful new therapies, the DPP serves as a benchmark: It shows that prevention is possible, durable, and most effective when implemented early, intensively, and strategically. New therapies will transform the landscape, but the fundamental truth endures. Preventing diabetes is as important as treating it, and prevention must be pursued with urgency, precision, and perseverance.

The author declares no conflict of interest.

Approval of the research protocol: N/A.

Informed consent: N/A.

Registry and the registration no. of the study/trial: N/A.

Animal studies: N/A.