The main finding
In one sentence: Over a median of 21 years, adults originally assigned to intensive lifestyle change or metformin in the Diabetes Prevention Program did not have fewer heart attacks, strokes or deaths from heart disease than those assigned to placebo.
Group averages, not an individual prediction. A trial result describes what happened to a defined group over a defined time.
On this page
Why this study matters
Preventing type 2 diabetes is often expected to protect the heart, because diabetes raises the risk of heart attack and stroke. The Diabetes Prevention Program (DPP) showed that lifestyle change and metformin both delayed diabetes. This report asks the harder question: over more than two decades, did either approach lead to fewer major heart events? The answer matters for anyone weighing what diabetes prevention can and cannot do.
Who was in the study
The original DPP randomly assigned 3,234 adults with impaired glucose tolerance, a form of prediabetes where blood sugar rises higher than normal after a sugar drink. They were followed for three years in the DPP and then for an average of 18 more years in the follow-up study, called DPPOS. The median follow-up for this report was 21 years. The published summary does not state which countries took part.
What the researchers did
The DPP was a randomized trial, meaning chance decided each person's group. The three groups were metformin 850 mg twice daily, an intensive lifestyle program, and placebo, an inactive tablet.
After the first three years, everyone was offered a less intensive group lifestyle program. The metformin group kept taking metformin, now unmasked. People were analysed by their original group, an approach called intention to treat.
The main outcome was the first of a nonfatal heart attack, a stroke, or death from heart disease, each confirmed by standard criteria. A broader outcome added hospital stays for heart failure or unstable angina, procedures to open blocked arteries, coronary disease found on an angiogram, and silent heart attacks seen on an electrocardiogram. Heart risk factors and electrocardiograms were checked every year.
What they found
Neither metformin nor lifestyle change reduced the main heart outcome (R32-C101).
| Comparison | Hazard ratio | 95% confidence interval |
|---|---|---|
| Metformin versus placebo | 1.03 | 0.78 to 1.37 |
| Lifestyle versus placebo | 1.14 | 0.87 to 1.50 |
A hazard ratio compares how fast events happen in one group versus another. A value of 1.0 means no difference. The confidence interval is the range within which the true value would likely fall. Both intervals include 1.0, so the study did not find a difference. They are also wide enough that a small benefit or a small harm cannot be ruled out.
Adjusting for heart risk factors did not change the results. Neither intervention affected the broader heart outcome either.
The authors offer reasons the effect may have been diluted. Everyone was offered group lifestyle support after the DPP. Many participants used statins and blood pressure medicines outside the study. And study metformin use fell over time while metformin use from outside the study rose.
Side effects and people who stopped
The published summary we reviewed does not report side effects or how many people left each group during the long follow-up.
What this study does not tell you
- It did not show a heart benefit, but it also does not prove there is none. The confidence intervals are wide.
- After year three, all groups received some lifestyle support, which blurred the comparison.
- Widespread use of statins, blood pressure medicines and metformin outside the study likely reduced any difference between groups.
- Participants had prediabetes, not diabetes, and were enrolled in the late 1990s. Care has changed since then.
- This study did not test any weight-loss medicine. It cannot be compared with trials of medicines that measured heart outcomes.
- The summary we reviewed does not state who funded this report.
What it means in Canada
This report is a reminder that preventing diabetes and preventing heart events are separate questions. A program can do the first without showing the second, especially when many people receive other treatments over decades. It does not involve weight-loss medicines. Readers interested in medicines studied for heart outcomes can read about semaglutide in Canada and tirzepatide in Canada. For coverage, see public drug coverage by province. To find diabetes prevention programs, see compare services.
Common questions
Did the Diabetes Prevention Program reduce heart attacks and strokes?
No. Over a median 21 years, the hazard ratio was 1.03 for metformin versus placebo and 1.14 for lifestyle versus placebo. Neither showed fewer major heart events.
Why might the DPP not have shown a heart benefit?
The authors note that everyone was later offered lifestyle support, many used statins and blood pressure medicines outside the study, and metformin use changed over time.
Does this mean preventing diabetes does not help the heart?
The study cannot say that. The confidence intervals were wide enough that a benefit or harm could not be ruled out.
Related reading
Other studies in this library that bear on the same question.
Diabetes Prevention Program: lifestyle versus metformin
Diabetes Prevention Program: 15-year follow-up
Look AHEAD four-year risk factor results
SELECT: semaglutide and heart attack and stroke prevention
Source
Original paper: Ronald B Goldberg, DPP/DPPOS cardiovascular 21-year, Circulation, 2022. PubMed: https://pubmed.ncbi.nlm.nih.gov/35603600/. DOI: https://doi.org/10.1161/CIRCULATIONAHA.121.056756. Funding: The summary we reviewed does not state the funding source. How we summarized it: full text and extracted data.
This page explains a published study. It is not medical advice and does not describe whether a medicine is right for you. Talk to your doctor, nurse practitioner or pharmacist about your own situation. Reviewed and signed by two pharmacists registered in British Columbia, 2026-09-18.
Evidence records behind this page
Each record is a row in the clinical evidence file, taken from the published paper named above. The caution column is the limit the researcher recorded for that number.
This table scrolls sideways.
| Record | What the paper reports | Type | Population | Caution | Source |
|---|---|---|---|---|---|
| R32-C101 | Over a median 21 years in DPP/DPPOS, neither original metformin nor lifestyle assignment demonstrated fewer major cardiovascular events than placebo: hazard ratios 1.03 (95% CI 0.78 to 1.37) and 1.14 (0.87 to 1.50), respectively. | direct evidence | Original DPP adults with impaired glucose tolerance | Did not demonstrate cardiovascular event reduction; confidence intervals do not establish equivalence. Lifestyle for all, nonstudy treatments and crossover may dilute effects. Primary 11832527,15-year 26377054. | PubMed 35603600 |