Image
Loading...
GLP-1 drugs such as Ozempic and Wegovy have been hailed as breakthrough treatments for diabetes and obesity. A new study co-authored by Jonathan Zhang, a professor at Duke University’s Sanford School of Public Policy, finds that the drugs live up to that promise in everyday medical care. The study also finds that, at least in the short run, they do not reduce overall health care spending.
The research, based on medical records from nearly 1.4 million U.S. veterans, shows that patients who were more likely to receive GLP-1 medications experienced meaningful improvements in blood sugar control and weight loss. Those improvements closely match results seen in clinical trials. But the study finds no evidence that GLP-1 use leads to fewer emergency room visits or lower total medical costs through 2024.
“GLP-1 drugs clearly improve key health outcomes,” said Jonathan Zhang, assistant professor at Sanford, a core faculty member at Margolis Institute for Health Policy, and a faculty research fellow at the National Bureau of Economic Research. “What surprised many people is that those health gains do not quickly translate into lower overall medical spending.”
What the researchers studied
GLP-1 medications were originally developed to treat type 2 diabetes and are now widely prescribed for weight loss as well. Their rapid rise has sparked debate over whether their high price might be offset by better health and lower use of other medical care.
To study that question, Zhang and his co-authors examined care delivered through the Veterans Health Administration, the nation’s largest integrated health system. When semaglutide was added to the VA’s formulary in late 2020, some primary care providers began prescribing GLP-1 drugs quickly, while others did not. Because veterans were already assigned to providers, researchers could compare patient outcomes based on their provider’s prescribing behavior, after GLP-1 adoption.
That approach allowed the team to measure real-world impacts for a large population of patients, beyond clinical trials.
What they found
Patients whose providers were more likely to prescribe GLP-1 drugs saw clear health benefits, including:
- Lower blood sugar A1C levels
- Sustained weight loss over several years
- Health improvements---such as fewer major adverse cardiovascular events and a reduction in moderate alcohol consumption---comparable to those seen in major clinical trials
At the same time, the study found no statistically significant change in overall emergency department use, mental health outcomes, other substance use outcomes, or total non-drug medical spending.
In short, GLP-1 drugs worked as advertised for metabolic health, but they did not reduce broader health care use in the near term.
Why this matters for policy
As lawmakers and insurers debate whether to expand coverage for GLP-1 drugs, the findings offer a clear message.
“These medications deliver real health benefits, but policymakers should be cautious about assuming they will immediately pay for themselves,” Zhang said. “Many cost projection models assume that metabolic improvements will directly translate into cost savings. Our study, as well as other studies emerging that analyze real-world medical expenditures indicate that this assumption may not hold in practice.
How patient populations shape outcomes
The study also shows that the effects of GLP-1 drugs depend on who receives them. Within the VA, where medication costs are relatively low and access barriers are limited, patients who initiated GLP-1 therapy later tended to be healthier than earlier recipients.. As a result, the average health benefit per patient declined as treatment reached broader segments of the population.
That pattern may not hold in other settings. In patient populations such as Medicare or Medicaid, higher financial or administrative barriers may initially limit GLP-1 use to patients with more severe health needs. Reducing those barriers—such as recent cost reductions—could draw in sicker or more disadvantaged patients, who may experience larger health gains. As access expands nationally, the distribution of benefits will depend on how coverage rules shape who is able to receive these medications.
The takeaway for policymakers
GLP-1 drugs can significantly improve health for people with diabetes and obesity, but decisions about coverage should be based on their clinical value, not expectations of short-term cost savings. Any financial benefit may depend on longer-term outcomes or socioeconomic benefits beyond the health care system.
The paper, Weighing the Impacts of GLP-1s: Quasi-Experimental Evidence From Provider Adoption, is co-authored by Sam Bock of Duke University and the VA Office of Mental Health, and Jasmin Moshfegh of Imperial College London and the Veterans Health Administration. It is available as a National Bureau of Economic Research working paper.