Wire open Last filed CagriSema REDEFINE 9 lower doses cut weight by about 21 percent in Phase 3 trial RSS
Peptide Zone

Peptide medicine news,
filed as it happens.

Morning wire
RoutineSAFNews

Stopping GLP 1 drugs tracks with twice the heart event rate in Italian diabetes cohort

Among 57,000 matched adults with type 2 diabetes, heart attack, stroke or death came about twice as often after GLP 1 drugs were stopped. The authors say illness behind stopping may explain much of the gap.

Editorial illustration for the story: Stopping GLP 1 drugs tracks with twice the heart event rate in Italian diabetes cohort
Illustration: Peptide Zone. Editorial drawing, not data.

What's new

An Italian cohort of 57,426 matched adults with type 2 diabetes found heart attack, stroke or death about twice as often after GLP 1 drugs were stopped, and deaths about 4.6 times as often.

Why it matters

Up to half of users stop within a year, but the study is observational and the authors say illness behind stopping may explain much of the gap.

What's next

Only a randomized test of stopping versus continuing could measure the true effect, and the cohort did not include tirzepatide.

Adults with type 2 diabetes who stopped GLP 1 receptor agonists had about twice the rate of heart attack, stroke or death as those who stayed on them, a large Italian cohort study reports. The analysis of stopping GLP 1 drugs, published in The Lancet Regional Health Europe, drew on records from 301 diabetes clinics. The authors call the result hypothesis generating. They also found that blood sugar and weight barely worsened after people stopped, which leaves open what drives the gap.

How the Italian cohort was built

The research team, with Ceriello as first author, used routine clinical records collected between 2004 and 2024 at diabetes clinics across Italy. They included adults with type 2 diabetes who received at least one prescription for a GLP 1 receptor agonist and excluded anyone who had already had a major cardiovascular event. A person counted as having stopped when no GLP 1 prescription appeared at two or more consecutive visits. People still receiving the drug at their last visit counted as persistent users. Of 80,302 users, 31,687 stayed on treatment and 48,615 stopped. Persistent users had been exposed for an average of 2.97 years and those who stopped for 1.21 years, according to a summary by News Medical.

The two groups differed at the start. People who stopped were more often women and more often also took sulfonylureas, an older class of glucose lowering pills, while other features looked similar. The team matched the groups, leaving 28,713 people in each. The main outcome combined non fatal heart attack, non fatal stroke and death from any cause. Death from any cause was also analysed on its own. Further analyses used a different definition of stopping and looked separately at people with and without obesity. Tirzepatide was not included because it reached Italy after the data were collected, so the findings apply only to older GLP 1 drugs.

Stopping GLP 1 drugs and the event rates

During treatment the matched cohort logged 3,143 primary outcome events, and after people stopped it logged 2,122. Because the time spent off treatment was shorter, the rate tells the story better than the count. The combined outcome occurred at 3.94 events per 100 person years after stopping, against 1.96 while on treatment, roughly double. Deaths from any cause rose from 0.48 to 2.23 per 100 person years, about 4.6 times higher. The direction held when the authors changed the definition of stopping, and the gap was larger among people with obesity. These are relative comparisons inside one observational cohort, not results from a randomized test of stopping.

Deaths from any cause rose from 0.48 to 2.23 per 100 person years, about 4.6 times higher.

The more surprising finding concerns the usual risk factors. Most of those measured at the point of stopping were similar to or better than at baseline, the authors report. Weight regain was modest, and glucose control did not get worse in the period the records cover. That pattern fits poorly with the idea that people lose the drug's benefit only through rebound in weight, blood sugar, cholesterol or blood pressure. The authors raise the possibility that GLP 1 drugs protect the heart through other routes, but they describe that idea as speculative. The published Lancet regional journal paper sets out the analyses and limits.

Measure per 100 person yearsOn treatmentAfter stopping
Heart attack, stroke or death1.963.94
Death from any cause0.482.23
Matched people per group28,71328,713
Mean time on GLP 1 drug, years2.971.21

The authors are direct about the limits. The design cannot prove that stopping GLP 1 drugs caused the extra deaths, and it cannot rule out reverse causation. People may have stopped because they became frail or developed cancer, and those illnesses carry their own risk of death. The records did not hold the reasons for stopping or any socioeconomic data, both of which affect heart health. Heart deaths could not be separated from other deaths, and data on heart failure admissions were too thin to use. The study also could not say whether starting and then stopping a GLP 1 drug is worse than never taking one.

The size of the death signal is itself a warning sign. The authors note that the mortality gap is much larger than the cardiovascular benefit seen in earlier randomized trials of these drugs. They conclude that withdrawal of treatment alone is unlikely to explain it and that unmeasured factors, perhaps reflecting poorer underlying health, probably contributed. Readers should weigh the result against that caveat. Observational data of this kind are useful for raising questions, and our editorial standards explain how this site labels such evidence. A randomized trial of stopping versus continuing treatment would be needed to measure the true effect, and none was reported alongside this paper.

What the findings mean for people on GLP 1 drugs

GLP 1 drugs are often stopped. Between 20% and 50% of people discontinue within the first year, the paper's summary notes, with stomach side effects, doubts about the benefit, cost and worries about safety among the reasons. The authors suggest that more flexible dose increases could improve tolerability and keep more people on treatment. They do not propose any specific schedule. For a person with type 2 diabetes, the study is a reason to discuss any plan to stop with the prescriber rather than a finding that stopping is dangerous in itself. Decisions about treatment depend on the person's own risks, side effects and access to the medicine.

The cohort also has a narrow scope. It covered only people with type 2 diabetes and no earlier heart attack or stroke, so it says little about people who take GLP 1 drugs for weight alone or who already have heart disease. It does not cover tirzepatide. People who cannot afford or obtain approved GLP 1 drugs sometimes turn to compounded or online products, which carry their own risks of wrong content and contamination. More reviews of side effects and the unapproved market are collected on the safety desk. Further work will need data on why people stop and on outcomes for newer drugs.

Conclusion

The Italian cohort links stopping GLP 1 drugs in type 2 diabetes with about twice the rate of heart attack, stroke or death and a far higher death rate. The authors themselves say the size of the death signal points to factors beyond the drug, such as illness that led people to stop. The study raises a question for trials to answer rather than giving a reason to change treatment on its own.

More from the wire

Same thread first, then the same desk.