The New England Journal of Medicine FEBRUARY 07, 2002

Reduction in the Incidence of Type 2 Diabetes with Lifestyle Intervention or Metformin

William C. Knowler, Elizabeth Barrett-Connor, Sarah E. Fowler, Richard F. Hamman, John M. Lachin, Elizabeth A. Walker, David M. Nathan; Diabetes Prevention Program Research Group

Bottom Line

The Diabetes Prevention Program (DPP) demonstrated that both intensive lifestyle modification and metformin therapy significantly reduce the incidence of type 2 diabetes in high-risk individuals compared to placebo, with lifestyle intervention proving more effective.

Key Findings

1. The intensive lifestyle intervention reduced the incidence of type 2 diabetes by 58% (95% CI, 48–66%) compared to the placebo group.
2. Metformin (850 mg twice daily) reduced the incidence of type 2 diabetes by 31% (95% CI, 17–43%) compared to the placebo group.
3. The incidence rates of diabetes were 11.0, 7.8, and 4.8 cases per 100 person-years in the placebo, metformin, and lifestyle intervention groups, respectively.
4. To prevent one case of diabetes over 3 years, 6.9 persons would need to participate in the lifestyle intervention, whereas 13.9 persons would need to receive metformin.
5. The intensive lifestyle intervention was significantly more effective than metformin (39% relative reduction in incidence).

Study Design

Design
RCT
Double-Blind
Sample
3,234
Patients
Duration
2.8 yr
Median
Setting
Multicenter, US
Population Nondiabetic persons with elevated fasting plasma glucose (95–125 mg/dl) and post-load plasma glucose (140–199 mg/dl) and a body-mass index of at least 24.
Intervention Intensive lifestyle-modification program (goal: ≥7% weight loss and ≥150 min physical activity/week) or metformin (850 mg twice daily).
Comparator Placebo group receiving standard lifestyle recommendations.
Outcome Development of type 2 diabetes.

Study Limitations

The trial was conducted in a specialized research setting, which may limit the generalizability of intensive lifestyle programs to real-world primary care settings.
The study duration (mean follow-up of 2.8 years) was relatively short, leaving initial questions about the long-term sustainability of the observed effects.
The study population was restricted to those with elevated fasting and post-load plasma glucose, potentially missing insights into broader prediabetic states.
The intensity of the lifestyle intervention (individual case managers, frequent contact) is resource-intensive and may be difficult to scale widely.

Clinical Significance

The DPP established that type 2 diabetes is a preventable disease. These findings shifted clinical practice, underscoring the efficacy of intensive lifestyle weight loss (aiming for ≥7% reduction) and the use of metformin as evidence-based prophylactic strategies for patients with impaired glucose tolerance.

Historical Context

Prior to the DPP, the efficacy of pharmacological or lifestyle interventions for the primary prevention of type 2 diabetes remained uncertain. This trial served as a landmark study, providing the first definitive large-scale evidence from the United States that reinforced and expanded upon the findings of previous studies like the Finnish Diabetes Prevention Study, thereby setting the standard for diabetes prevention guidelines.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

By what physiological mechanisms do intensive lifestyle interventions and metformin differ in their approach to preventing the progression from impaired glucose tolerance to type 2 diabetes?

Key Response

Lifestyle modification primarily improves insulin sensitivity via increased skeletal muscle glucose uptake and weight loss-induced reduction in adipokines, whereas metformin primarily inhibits hepatic gluconeogenesis and activates AMPK, targeting different facets of the metabolic syndrome to achieve a combined reduction in glycemic load.

Resident
Resident

According to the DPP subgroup analyses, which specific patient demographics showed a response to metformin that was nearly as effective as lifestyle modification, and how should this influence your prescribing habits for pre-diabetic patients?

Key Response

The DPP found that metformin was most effective in individuals with a BMI of 35 or greater and those younger than 60 years old. In these specific high-risk groups, the gap between metformin and lifestyle efficacy narrowed, suggesting metformin is an excellent adjunct or alternative when intensive lifestyle changes are not feasible.

Fellow
Fellow

The DPP used a 'masking' period for metformin before final testing to distinguish between a true preventive effect and a simple pharmacological lowering of glucose. How does this distinction impact our understanding of 'disease modification' versus 'symptomatic management' in pre-diabetes?

Key Response

If the drug merely masks hyperglycemia, glucose levels would rapidly rise upon discontinuation. The DPP and its follow-up (DPPOS) showed that while some benefit was due to active drug effect, there was a sustained delay in the onset of diabetes, suggesting that early intervention can fundamentally shift the metabolic trajectory, though lifestyle remains the more potent disease-modifier.

Attending
Attending

Given that the 'intensive lifestyle intervention' in the DPP involved 16 one-on-one sessions with case managers, how do we address the 'translational gap' when advising patients in a standard clinical setting where such resources are rarely available?

Key Response

This highlights the difference between efficacy (trial) and effectiveness (real-world). Attendings should emphasize the 'Diabetes Prevention Recognition Program' (DPRP) and community-based programs like the YMCA's adaptation of the DPP, which utilize group settings to achieve similar results at a fraction of the cost and intensity of the original trial.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

Critically evaluate the use of a binary biochemical threshold (e.g., fasting plasma glucose ≥126 mg/dL) as the primary endpoint in the DPP; how might a 'time-to-event' analysis for macrovascular complications have altered the statistical power requirements and the ultimate conclusions of the study?

Key Response

Using a glycemic threshold allows for a shorter study duration and smaller sample size but may not reflect clinically meaningful end-organ damage. A macrovascular endpoint would have required a much larger cohort and decades of follow-up (as seen in the transition to DPPOS), potentially revealing that the metabolic 'memory' or 'legacy effect' is more critical than the specific timing of a diagnosis based on an arbitrary glucose cutoff.

Journal Editor
Journal Editor

The DPP was terminated early for the lifestyle and metformin arms due to clear efficacy. As a peer reviewer, what concerns would you raise regarding the potential overestimation of effect sizes and the loss of long-term safety data that often accompanies early trial termination?

Key Response

Early termination for benefit frequently leads to 'the winner's curse,' where treatment effects are exaggerated. Furthermore, it truncates the observation period for rare side effects or the waning of efficacy over time, necessitating long-term observational extensions (like the DPPOS) to validate the initial findings' durability and safety.

Guideline Committee
Guideline Committee

The ADA Standards of Care currently give a Grade A recommendation to metformin for prevention of T2DM in specific subgroups. Based on the DPP evidence, should these guidelines be expanded to all patients with an A1c of 5.7–6.4%, or should they remain restricted to those with additional risk factors?

Key Response

Current ADA guidelines (Section 3) recommend metformin for those with pre-diabetes, especially those with BMI ≥35, age <60, and women with prior GDM. The DPP evidence supports this selectivity because the NNT (Number Needed to Treat) for metformin was significantly higher in older, less obese individuals, suggesting that a blanket recommendation would be less cost-effective and potentially expose more patients to gastrointestinal side effects without proportional benefit.

Clinical Landscape

Noteworthy Related Trials

1998

UK Prospective Diabetes Study (UKPDS)

n = 5,102 · Lancet

Tested

Intensive blood-glucose control with sulphonylurea or insulin vs. metformin

Population

Newly diagnosed type 2 diabetes patients

Comparator

Conventional treatment with diet

Endpoint

Any diabetes-related endpoint

Key result: Intensive glucose control reduced the risk of diabetes-related endpoints and microvascular complications.
2001

Finnish Diabetes Prevention Study (DPS)

n = 522 · NEJM

Tested

Intensive lifestyle intervention

Population

Overweight individuals with impaired glucose tolerance

Comparator

Standard care and brief advice

Endpoint

Incidence of type 2 diabetes

Key result: Lifestyle intervention resulted in a 58% reduction in the incidence of type 2 diabetes.
2013

Look AHEAD Trial

n = 5,145 · NEJM

Tested

Intensive lifestyle intervention focused on weight loss

Population

Overweight or obese patients with type 2 diabetes

Comparator

Diabetes support and education

Endpoint

Composite of cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, or hospitalized angina

Key result: Intensive lifestyle intervention did not reduce the rate of cardiovascular events despite successful weight loss and improved glycemic control.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis