Dodging a diabetes diagnosis did not save lives in big prevention trials

Millions of people sit in the grey zone of prediabetes, where blood glucose is raised through impaired glucose tolerance or impaired fasting glucose but has not crossed the diabetic line. The assumption behind screening and treating this state is that stopping progression to full diabetes will ultimately prevent heart attacks, strokes and early deaths. That assumption is testable: if interventions delay diabetes, hard outcomes in treated groups should eventually improve. This meta-analysis asked whether the randomised trials actually show that payoff, or only the intermediate win.

The authors pooled prospective randomised controlled trials of interventions in people with prediabetes, spanning diet, exercise and pharmacological therapy. Trials qualified if they reported all-cause mortality, recruited approximately 100 patients or more, and followed people for at least a year. Ten trials enrolling 23,152 patients met the criteria, running for an average of 3.75 years. Interventions were split into pharmacological and non-pharmacological approaches, and outcomes included progression to diabetes, all-cause and cardiovascular death, myocardial infarction and stroke.

The interventions worked for their immediate goal: diabetes was delayed or prevented, with a risk ratio of 0.83 against control. Non-drug approaches outperformed drug-based ones, with risk ratios of 0.52 versus 0.70. Yet all-cause mortality showed no difference between intervention and control groups, and cardiovascular death did not fall either. Fatal and non-fatal heart attack showed only a non-significant trend towards reduction, while fatal and non-fatal stroke was borderline reduced, the single outcome hinting at downstream benefit.

The analysis confirms what individual prevention trials had suggested, that progression from prediabetes to diabetes can be slowed, and it strengthens the case that lifestyle approaches beat medication for that specific goal. It complicates the comfortable assumption that preventing the diagnosis automatically prevents cardiovascular disease, at least within the timeframe trials have covered. An average follow-up of 3.75 years may simply be too short for survival benefits to emerge, which is the central open question. Longer trials, or extended follow-up of these cohorts, would be needed to settle it.

If you have been told your glucose sits in the prediabetic range, this is solid evidence that diet and exercise genuinely cut the odds of progressing to diabetes, and did so more effectively than drugs in these trials. It is also a reason to hold the benefit honestly: within the few years these trials ran, avoiding the diagnosis had not yet translated into fewer deaths or heart attacks, with stroke the possible exception. The case for lifestyle change rests on the diabetes outcome itself, and on patience.