Two years off GLP-1s wiped out most of the heart benefit in a 333,000-veteran cohort

Benefit tracked how long patients stayed on the drug. Veterans on GLP-1 therapy for all three years had an 18% lower risk of major cardiovascular events than those prescribed sulfonylureas, another diabetes drug class that includes glipizide and glimepiride ^[7][18]. Stopping at two and a half years left a 15% reduction; stopping at two years left 7% ^[8]. For anyone who came off before 18 months, no significant reduction appeared ^[9].

Sicker or poorer patients quitting first could open a gap between the groups, but that kind of confounding does not obviously produce a staircase that follows exposure time. The absolute scale is modest. Continuous use came to about four fewer major cardiovascular events per 100 people over three years ^[7]. A patient who stopped at two years kept 7 of those 18 points, about 39% of the relative benefit ^[1], which at the same baseline risk is roughly 1.6 events per 100 ^[2].

In headcount, the interruptions were large: about 34,000 of the 132,551 GLP-1 patients quit outright, and about 30,000 went at least half a year without a dose before restarting ^[2][3]. Treatment status was reassessed every six months ^[12], so a two-month lapse does not appear in this analysis at all. The two-year risk figure is also a maximum, given as "up to 22%" in the text and stated flatly in the release's summary ^[4][5].

Ziyad Al-Aly, the senior author and chief of the research and development service at the VA Saint Louis Health Care System, took up the question after observing that about half of users stop not long after beginning treatment ^[16][17]. "There is enormous exuberance about starting GLP-1 drugs, but not nearly enough attention to what happens when people stop," he said ^[13]. People quit "because of cost, side effects or shortages", he said, and "Weight regain is visible; the metabolic reversal is not" ^[14].

Al-Aly reads the restart data as incomplete recovery. It "helped restore some protection, but only partially, showing that discontinuation leaves a lasting scar", he said ^[15]. The release does not describe how the analysis adjusted for the reasons patients came off treatment. Every veteran here had type 2 diabetes and a prescription for a glucose-lowering drug ^[1][2], so the study does not measure what happens to the heart when someone taking semaglutide for weight alone stops.