Study links stopping GLP-1 drugs with poorer outcomes but cannot prove cause

An observational study of more than 80,000 people with type 2 diabetes found an association between stopping GLP-1 medicines and poorer health outcomes than continuing treatment. Researchers followed patients in Italy without established cardiovascular disease, using records spanning 2004 to 2024.

About four in every hundred people who stopped treatment had a heart attack, stroke or died from any cause, compared with about two in every hundred who continued. The comparison is a combined outcome, not a finding that stopping treatment doubled every participant's risk of each separate event.

The research did not record why patients stopped. Published reporting experts warned that illness or other problems could explain both discontinuation and poorer outcomes. The study cannot prove that stopping the medicine caused those events or show whether starting and stopping is worse than never taking it.

Most participants used liraglutide or dulaglutide. Few used semaglutide and none used tirzepatide, limiting how widely the findings apply across newer treatments. The results add to questions about long-term use but are not a reason for individuals to change prescribed treatment without clinical advice.