A long-running diabetes trial found lifestyle changes prevented more than diabetes

A landmark diabetes-prevention program appears to have delivered benefits beyond delaying type 2 diabetes. More than two decades of follow-up found that participants originally assigned to intensive lifestyle changes developed multiple chronic conditions more slowly than those assigned to placebo.

The result adds a wider measure of health to a trial already known for diabetes prevention. It does not show that every diet or exercise plan will produce the same outcome, and the long-term analysis applies to a specific group of older adults with prediabetes whose Medicare records were available.

The original program targeted weight and activity

The Diabetes Prevention Program enrolled 3,234 adults at high risk for type 2 diabetes at 27 U.S. sites between 1996 and 1999. Participants were randomly assigned to intensive lifestyle intervention, metformin or placebo. The lifestyle curriculum aimed for at least 7% weight loss and 150 minutes of moderate physical activity each week, supported by 16 individual sessions and continuing contact.

After the initial trial, participants entered the Diabetes Prevention Program Outcomes Study. All groups were offered some lifestyle education, while the original lifestyle group continued to receive extra booster sessions. That transition means the sharply separated randomized treatments of the first three years became less distinct during long-term follow-up.

Medicare records tracked fifteen chronic conditions

The new analysis focused on 1,173 participants who were enrolled in Medicare and consented to linkage with federal claims. Researchers followed records through 2021 and examined 15 conditions, including hypertension, heart disease, stroke, chronic kidney disease, chronic obstructive pulmonary disease, arthritis, cancer, depression, dementia, osteoporosis and diabetes.

By the end of follow-up, multimorbidity was common: 85% of the analyzed participants had at least two conditions. The proportions were 82% in the original lifestyle group, 85% in the metformin group and 87% in the placebo group. Participants had a median age of 74 in the analysis, and 68% were women, according to the JAMA study record.

Lifestyle assignment was linked to a lower burden

Compared with placebo, assignment to the intensive lifestyle program was associated with a 21% lower risk of developing two chronic conditions and a 25% lower risk of developing three. The relationship persisted when diabetes itself was removed from the definition, supporting the conclusion that the result was not merely a restatement of the program’s known effect on diabetes.

Those percentages describe relative differences in the rate at which participants reached the specified disease counts. The absolute proportions remained high in every group, with more than four in five participants eventually developing at least two tracked conditions. The intervention therefore delayed or reduced part of the accumulated burden; it did not make aging free of chronic illness. That distinction keeps a meaningful result from sounding like complete prevention.

Metformin did not produce a statistically significant reduction in multimorbidity risk in this analysis. That finding should not be read as a judgment that the drug lacks value. Metformin is widely used for glucose management, and the original program found that it delayed diabetes in some high-risk adults. The narrower conclusion is that this long-term claims analysis did not detect the same broader reduction in accumulated chronic conditions.

A randomized start became observational follow-up

The National Institutes of Health summary describes the work as follow-up to a clinical trial, but the later decades require careful interpretation. Participants retained their original assignment for analysis, yet treatments were unmasked after the first phase. Placebo stopped, metformin continued, and lifestyle classes became available across groups.

Claims data are useful for following diagnoses at scale, but they were available for only part of the original cohort. The selected participants had survived to Medicare eligibility, enrolled in coverage represented in the dataset and agreed to linkage. Those conditions can produce differences between the analyzed group and the full trial population. Claims also record billed diagnoses rather than every detail of disease severity or daily behavior.

Researchers used time-to-event models to compare how quickly participants accumulated conditions, rather than relying only on the final percentages. That method respects differences in follow-up time, but it still depends on accurate coding and assumptions about how group risks behave over the study period. The published result is an association with original random assignment during a long observational extension, not a fresh decades-long randomized comparison with fully separated treatments.

The intervention was more than generic advice

The findings do not establish that occasional exercise or a brief diet attempt prevents multimorbidity. The original intervention was structured and intensive. It paired specific weight and activity goals with individual coaching, repeated sessions and long-term reinforcement. Reduced calorie and fat intake, regular physical activity and behavioral support worked as a package.

That distinction matters for public-health translation. Community programs can adapt the model, but lowering cost or contact time may also change effectiveness. Future research can test which elements are necessary to preserve the long-term benefit and whether similar results appear in younger adults, people without prediabetes and populations not captured through Medicare claims.

The study’s larger message is that prevention can be judged by more than the first diagnosis it delays. A sustained lifestyle program may influence several pathways that contribute to chronic illness across aging. The evidence is strongest for the program and population actually studied, not for a universal prescription.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

More from Morning Overview