Twelve days ago, a team at Washington University in St. Louis published a study of more than 600,000 veterans that arrived at a number the GLP-1 industry had quietly been dreading: stopping the drugs for two years raised cardiovascular risk by 22 percent compared with continuous use. [1] [2]

This paper reported the study's central finding as it broke — the progressive erosion of protection, the 4 percent risk increase at six months, 14 percent at one year, 22 percent at two years. Today's question is what the field does with that number.

The question is not purely clinical. It is structural. GLP-1 drugs — semaglutide, tirzepatide, liraglutide — were developed as diabetes medications and subsequently approved for weight loss. Their cardiovascular benefits emerged from large outcome trials as a secondary finding and then became a primary selling point. The SELECT trial's results, published in 2023, showed semaglutide reducing major cardiovascular events by 20 percent in non-diabetic patients with overweight and established heart disease. The Washington University study adds a temporal dimension: the protection is not durable. It is contingent on continued use. [1] [3]

For the estimated 15 million Americans now on GLP-1 drugs, that conditional architecture raises several questions that medicine has not yet answered. Who should be prioritized for continued access when shortages hit? What is the clinical protocol for patients who lose insurance coverage mid-treatment? The on-again, off-again pattern that characterized 2024 supply disruptions — patients stopping for months while manufacturers caught up to demand — may have produced a wave of cardiac exposure that will take years to appear in mortality data. [4]

Dr. Ziyad Al-Aly, the study's lead author, has been careful not to overstate the mechanism. The study is observational. It cannot establish causality — it can only say that patients who stopped showed worse outcomes than those who continued. Confounders exist: patients who stop are often stopping for reasons of cost, illness, or tolerability that may themselves correlate with cardiac risk. But the gradient — linear, consistent across the full 600,000-person cohort — is not easy to explain away. [2] [3]

The pharmaceutical calculus is brutal. Novo Nordisk's share price rose on the day the study published. A drug that patients cannot safely discontinue is a drug with a lifetime revenue curve, and the market priced that in within hours. For patients, the same logic lands differently: a medication you start in your fifties may be one you take at eighty, or face the medical consequences of stopping. [1]

The clinical community's response has been measured but serious. The American Diabetes Association's rapid response encouraged clinicians to "treat discontinuation as deliberately as initiation" — meaning a conversation about risks, not a prescription renewal declined without explanation. What that looks like in a ten-minute primary care appointment, against a backdrop of prior authorization requirements and formulary changes, remains unresolved.

GLP-1 drugs are the fastest-adopted pharmacological class since statins. Statins produced a similar reckoning: a drug that most patients stayed on for life, whose benefits were partially contingent on continued use, and whose interruptions became a category of clinical concern. That analogy suggests the field is not at the end of a discovery but the beginning of a much longer conversation about what these drugs actually require of the healthcare system.

The exit problem is now visible. The exit protocol does not yet exist.