Durham VA Health Care System will recruit veterans for a national 18-site trial testing whether GLP-1 medications can treat alcohol use disorder, with enrollment beginning Aug. 17 and about 622 participants planned.
# Durham VA to test GLP-1 drugs against alcohol use disorder in 622 veterans
Durham VA Health Care System will recruit veterans for a national 18-site trial testing whether GLP-1 medications , already approved for diabetes and obesity, can treat alcohol use disorder . Enrollment is set to begin Aug. 17, with about 622 participants planned across 18 VA medical centers, roughly 35 eligible veterans per site. Final results are not expected for about three years.
The plan became public on Aug. 7, 2026; Akilah Davis reported the announcement at 9:49 p.m. that evening. The U.S. Department of Veterans Affairs estimates that more than 400,000 veterans are diagnosed with alcohol use disorder each year, which gives the trial both a large patient pool and a clear clinical target.
The Durham VA, part of the Department of Veterans Affairs health system, was chosen as one of 18 VA medical centers participating in the national trial. The Charles George VA Medical Center in Asheville is also a participating site. The public announcement names those two North Carolina locations but not the full list of participating centers.
Recruitment begins Aug. 17, 2026, about 10 days after the announcement. The VA gave the start date simply as Aug. 17, without an explicit year; the Aug. 7 announcement places the start in 2026. Veterans eligible for the study must be between 18 and 80 years old and have moderate to severe alcohol use disorder. They do not need to live in Durham, but they must be able to travel to the Durham VA medical center for four in-person visits across the study. That travel requirement is a genuine constraint on enrollment: veterans far from Durham, or without reliable transportation, may be effectively unable to participate even though the protocol does not restrict enrollment by residence.
Each site expects to enroll about 35 eligible veterans, for a planned total of about 622 nationwide. The two figures do not multiply cleanly: 18 sites times 35 per site is 630, not 622. The announcement does not reconcile them. Both numbers are presented as approximations, and 622 divided by 18 is roughly 34.6 veterans per site, which is what "about 35" describes; still, the VA has not said which figure is the firmer commitment. The scale is clear either way: a 622-participant trial of an approved peptide drug class in a population defined by a diagnosis of alcohol use disorder is a substantial test of a repurposing hypothesis.
GLP-1 medications are peptide drugs that mimic glucagon-like peptide-1, an incretin hormone released from the gut after meals. In approved use they lower blood glucose by potentiating insulin secretion, slow gastric emptying, and reduce food intake through effects on appetite centers. Their approval for diabetes and obesity is the platform this trial builds on; they are not approved for alcohol use disorder, and the trial, whatever it finds, does not change that.
The hypothesis under test comes from where GLP-1 receptors are located. The receptor is expressed not only in the pancreas but on neurons in reward-related regions of the midbrain and forebrain, including the ventral tegmental area and the striatum. Alcohol, like other reinforcing drugs, raises dopamine signaling in those circuits, and that signal is thought to encode the motivational pull of the drug. The working idea of the trial is that GLP-1 receptor activation dampens the dopamine response and weakens the drive to drink. Dr. Dave Oslin, the national study lead from Philadelphia, described it directly: "These medications seem to satiate that reward pathway and make it so that you don't have that strong desire anymore."
The convergence matters beyond veterans. Food, alcohol, and other abused substances engage overlapping reward circuitry, so a drug that reduces the reward value of food might plausibly reduce the reward value of alcohol. The announced endpoints follow that logic: reduction in alcohol cravings, changes in drinking behavior, improvements in health outcomes, and quality of life. The public description does not say whether the protocol measures dopamine signaling or other biomarkers directly. As described, the trial is a clinical test of a mechanistic hypothesis, not a mechanistic study.
Participants will receive weekly GLP-1 injections for six months while investigators monitor drinking behavior, health, and quality of life. The study requires four in-person visits at the Durham VA medical center per participant. The public description does not explain how the four visits relate to the weekly injection schedule, whether injections are administered at home or in a clinic, or how drinking is measured, whether by self-report, biological markers, or both.
The enrollment targets are modest per site, about 35 veterans, which spreads recruitment across 18 centers and keeps the study manageable at community-based VA facilities. The age range, 18 to 80, is broad for an addiction trial, and allowing veterans who do not live in Durham to participate effectively makes the site's catchment regional. Those choices also create a screening challenge: the announcement does not state which definition of moderate to severe alcohol use disorder will be used for eligibility, or how that determination will be made.
The four required visits are the clearest burden on participants. Six months of weekly injections implies ongoing contact with the study team, yet the relationship between that schedule and the visit count is not spelled out. The announcement also does not say whether monitoring continues after the six-month injection period, a detail that matters because a treatment effect that persists after the drug is stopped would be more clinically valuable than one that requires continuous use.
According to the U.S. Department of Veterans Affairs, more than 400,000 veterans are diagnosed with alcohol use disorder each year. Chuck Sutton, an Army and National Guard veteran, described the prevalence from the inside: "Once you're in the military and you have to go through different environments and experience different things, drug and alcohol use is pervasive amongst veterans."
The VA is positioned to run a multisite study at a scale few other systems could match. It operates the largest integrated health system in the United States, with shared electronic health records, centralized research infrastructure, and care networks that span the country. A trial across 18 medical centers takes advantage of that structure: one protocol, consistent data systems, and clinical staff who already treat the population under study. That infrastructure is also why the trial can enroll veterans who live outside the local area.
The design choices reflect the population being served. The wide age range spans the full adult veteran spectrum. The travel requirement acknowledges that a veteran catchment area is wider than a single city. And the trial's central question, whether six months of GLP-1 exposure changes drinking behavior and cravings, is aimed at a condition that existing treatments have not resolved for many patients. The VA's own prevalence figure is the measure of that unmet need.
The most important limitation is time. Efficacy is not established, and it will not be established quickly. Recruitment begins Aug. 17, the six-month injection period follows, and final results are not expected for about three years, a timeline that puts the readout around 2029. The public description includes no protocol details and no registry number, so the design cannot be checked against a published document.
Design details that determine the strength of the evidence are also missing. The announcement names no specific GLP-1 medication and no dose. It does not state whether the trial is randomized, whether it is blinded, or whether it uses a placebo control. Those choices decide whether a positive result can be attributed to the drug or could reflect expectation effects, and they will determine how clinicians weigh the findings. The announced endpoints, reduction in cravings, changes in drinking behavior, health outcomes, and quality of life, are described only in general terms, with no primary outcome identified.
Open questions that remain:
Each of those questions has a concrete answer that would settle it. A protocol or registry entry would resolve the drug, dose, randomization, blinding, endpoints, and eligibility definition. A full site list would resolve the scope of the network. The enrollment arithmetic is consistent under rounding: 622 divided by 18 is roughly 34.6, which rounds to about 35 per site, though the announcement itself does not perform that calculation. The other items are substantive, and the absence of public detail is a limitation of the announcement, not a judgment on the trial.
If the trial succeeds, it would extend an approved peptide drug class into addiction medicine, a field dominated by small-molecule drugs and behavioral interventions. GLP-1 receptor agonists are peptide drugs, and an alcohol use disorder indication would expand peptide therapeutics beyond metabolic disease into a neuropsychiatric condition tied to the brain's reward system. Oslin connected the veteran-specific test to the general population: "If this biologically works in veterans, it's going to work in the community." The mechanism under test is biological, not military, so a positive result would speak to alcohol use disorder as a disease rather than as a service-connected problem.
Dan Blalock, a clinical research psychologist at the Durham VA, put the clinical stake plainly: "I think this would be a gamechanger to get more people toward effective treatments, including this one if it is effective." The conditional in his sentence is the honest summary of the moment. The trial has not been run, and the announcement explicitly does not claim efficacy. What it offers is a test large enough to matter, in a system that can deliver the result.
For researchers, the trial is an opportunity to study alcohol use disorder outcomes inside a health system that can link trial data to longitudinal electronic health records. For clinicians, the appeal of a GLP-1 medication is familiarity: the class is already prescribed widely for diabetes and obesity, its safety profile is established, and an added indication would not require learning a completely new pharmacology. For the supply chain, an additional indication would add demand to existing manufacturing and distribution networks for the peptide active ingredient, all of it building on production capacity that already exists for approved uses. None of that changes the three-year wait for efficacy data, and all of it depends on the trial's methods being sound.
The fastest way to resolve the open design questions is for the VA to publish the trial protocol or register the study in a public clinical trial registry. A registry entry would state the specific medication, the dose, the randomization scheme, the blinding, the comparator, the endpoints, and the eligibility criteria. Until then, the public description functions as a recruitment notice, not a methods document, and the strength of the…
Peptides referenced: Glucagon, GLP-1.
Related reading: Starbucks is ending GLP-1 coverage for weight loss as employer costs climb, No Link Between GLP-1 and Hypertensive Disorders of Pregnancy, Claims Data Lend Support to Tirzepatide's CV Effects in Diabetes, Amazon Pharmacy Launches GLP-1 Delivery for Medicare Patients.