Finnish DPS at 18 Years: Why Prediabetes Remission Slips

Lifestyle change can bring glucose back into the normal range for many people with prediabetes. Keeping it there is harder. In the Finnish Diabetes Prevention Study, 32% of participants with impaired glucose tolerance reached normoglycemia at least once over 18 years, yet reported sustained remission was 13% at year 1, 12% at year 3, and 11% at the first post-intervention visit, which occurred after a median of five years.

The gap matters because the factors linked to early improvement differed from those linked to durability. Weight loss, diet quality, physical activity, and a smaller waist predicted short-term remission. Later on, insulin secretion and insulin sensitivity indices carried the predictive signal. Lifestyle still matters; it helps shape those physiological traits. But normal glucose after a successful stretch of dieting is not proof that the underlying metabolic problem has retired.

What the Finnish study actually followed

The Finnish Diabetes Prevention Study 18-year follow-up included 505 adults with impaired glucose tolerance, a prediabetes state diagnosed with an oral glucose tolerance test (OGTT). Participants underwent a median four-year active lifestyle intervention, then extended passive follow-up. The study tracked whether participants met its definition of normoglycemia at repeated assessments.

Its 32% figure is cumulative "ever-remission": a participant reached normoglycemia at one measured visit. That is useful, but it is not the same endpoint as uninterrupted remission. A single OGTT can vary within the same person, and the coefficient of variation for repeated two-hour glucose measures can exceed 10%. Someone near a diagnostic threshold can therefore cross it without a durable biological change.

The more conservative figures were 13% at year 1, 12% at year 3, and 11% at the first follow-up after the active intervention, at a median of five years (range, four to eight). The publication's available abstract does not report remission prevalence separately at years 9, 13, and 18, so it cannot describe the full late-course trajectory. It also does not provide hazard ratios, odds ratios, or confidence intervals for the individual predictors. That limits any attempt to rank which factor mattered most.

Early remission followed the usual lifestyle pattern

During the active intervention years, people were more likely to reach normoglycemia when they lost more weight, reduced waist circumference, ate more fiber, consumed less saturated fat, and were more physically active. Improved insulin sensitivity and improved insulin secretion also predicted short-term remission in the Finnish study.

This is practical, not mysterious. Reducing excess adiposity and remaining active can reduce the amount of insulin the body needs. The study supports treating weight loss, food quality, and activity as the main tools for improving glucose regulation in impaired glucose tolerance. It does not establish a universal dose of weight loss or exercise that guarantees remission.

Nor should "lifestyle" become a polite euphemism for blame. Two people can follow similar programs and arrive with different pancreatic reserve. The relevant question is whether the pancreas can increase insulin output enough to match the body's insulin demand.

Durability depended on secretion and sensitivity

After the active intervention period, the Finnish analysis found that the lifestyle variables no longer retained independent statistical significance as predictors of longer-term remission status. Measures of insulin secretion—the insulinogenic index and disposition index—and insulin sensitivity measured by the Matsuda index did. The study does not report the effect sizes or confidence intervals in the available abstract, so this is a change in the pattern of predictors, not a numerical leaderboard.

The disposition index is especially useful conceptually because it joins both sides of glucose control: how sensitive the body is to insulin and how adequately beta cells respond to that demand. Better insulin sensitivity lowers the amount of insulin required. Adequate early insulin secretion keeps glucose from climbing after a meal. A person can improve one side and still struggle if the other deteriorates.

A prospective Canadian cohort study by Retnakaran et al. supports that interpretation. It found that recovery of beta-cell function, rather than improvement in insulin resistance alone, was central to prediabetes remission. This finding provides a plausible explanation for the Finnish predictor shift, but it does not prove that beta-cell recovery caused every long-term outcome in Finland. The published abstract available here does not give the cohort size or effect estimates, which limits a fuller appraisal of precision and generalizability.

Why relapse can occur after a good start

Insulin sensitivity is not fixed at the level achieved during a successful intervention. In a 10-year longitudinal Korean cohort, Kim et al. found that participants who progressed to dysglycemia had greater declines in insulin sensitivity over time than non-progressors, even when beta-cell function appeared preserved at baseline. This was an observational association, not a lifestyle trial, but it identifies a credible route back toward high glucose: insulin demand rises while insulin output fails to keep pace.

A separate three-cohort longitudinal study in China found that insulin demand and insulin adequacy change across the life course in response to growth, development, and aging. Zhang et al. did not test the Finnish intervention, but the finding puts long follow-up in context. Maintaining the same disposition index for 18 years means adapting to an aging body, not merely repeating the habits that worked at age 45.

The Finnish data cannot separate behavioral drift from progressive biological decline during the later years. The available report does not provide participant-level trajectories for weight regain or adherence after the intervention ended. Both explanations may contribute, along with aging-related changes in insulin demand.

Beta-cell reserve is a constraint, not a verdict

It would be a mistake to read these studies as evidence that lifestyle cannot improve beta-cell function. In DOSE-EX, a randomized clinical trial of 82 people with established type 2 diabetes, diet-induced weight loss improved beta-cell function among participants with residual secretory capacity; exercise dose added benefit beyond diet alone. That trial measured beta-cell function, a physiological outcome, rather than long-term prediabetes remission.

A comparative clinical study in insulin-treated type 2 diabetes also found that diet alone improved beta-cell function in people with remaining secretory capacity. These studies differ from the Finnish cohort in population and endpoint: participants already had type 2 diabetes, not impaired glucose tolerance, and neither directly tells us how many people with prediabetes can maintain normoglycemia for 18 years.

Still, they argue against a hard biological ceiling. Beta-cell function can improve when reserve remains. The more defensible conclusion is that reserve sets part of the starting position, while weight loss, exercise, and diet can sometimes improve the position. No validated threshold for the insulinogenic index or disposition index currently tells an individual that lifestyle remission is impossible.

What to measure if remission is the goal

HbA1c and fasting glucose remain useful clinical tests, but they can miss early deterioration after meals. An OGTT can estimate the insulinogenic index, Matsuda insulin sensitivity index, and disposition index; these were the long-term predictors identified in the Finnish analysis. Their interpretation is specialized and depends on correctly timed glucose and insulin measurements, so they are best discussed with a clinician rather than treated as a home-scorecard project.

One-hour post-load glucose may offer another warning signal. In a 10-year Korean cohort of 6,144 people, Im et al. found that one-hour glucose of at least 155 mg/dL during an OGTT predicted higher future type 2 diabetes risk than one-hour glucose below that threshold among people with normal glucose tolerance. It is a risk marker, not a diagnosis or a guarantee of progression.

For people with prediabetes, the realistic target is sustained metabolic capacity: enough insulin sensitivity to reduce demand, and enough beta-cell function to meet it.

The Finnish study offers a useful corrective to both fatalism and victory laps. Meaningful lifestyle change can normalize glucose, often early. It may also improve beta-cell function when reserve remains. But an early normal test result is a checkpoint, not a discharge notice. The people most likely to sustain remission had preserved insulin secretion and insulin sensitivity; the study cannot tell us how much of that advantage came from biology, how much from long-term behavior, or how much a more intensive intervention could change.