Large VA study links GLP-1 drugs to lower risk of substance use disorders

The BMJ published a study on March 4, 2026 comparing US veterans with type 2 diabetes who started a GLP-1 receptor agonist with similar patients who started a different diabetes drug, an SGLT-2 inhibitor. The veterans who started GLP-1 drugs were less likely to be diagnosed with substance use disorders, and those who already had one had lower rates of related emergencies, hospital admissions and deaths ^[1].

Researchers used Department of Veterans Affairs electronic health records to emulate eight parallel target trials, a method that tries to mimic a randomized trial using observational data. The starting population was 606,434 veterans with type 2 diabetes, followed for up to three years. The main comparison included 124,001 people who started a GLP-1 receptor agonist and 400,816 who started an SGLT-2 inhibitor, for a total of 524,817. A separate group of 81,617 people who already had a substance use disorder included 16,768 GLP-1 initiators and 64,849 SGLT-2 initiators ^[1].

Among people with no history of substance use disorder, starting a GLP-1 drug was associated with lower risk of:

• Alcohol use disorder: hazard ratio 0.82 (95% confidence interval 0.78 to 0.85)
• Cannabis use disorder: 0.86 (0.81 to 0.90)
• Cocaine use disorder: 0.80 (0.72 to 0.88)
• Nicotine use disorder: 0.80 (0.74 to 0.87)
• Opioid use disorder: 0.75 (0.67 to 0.85)
• Other substance use disorders: 0.87 (0.81 to 0.94)
• Any of these combined: 0.86 (0.83 to 0.88) ^[1]

The researchers also reported net three-year risk differences per 1,000 people. For alcohol use disorder, the difference was −5.57 (95% CI −6.61 to −4.53); for the combined outcome, −6.61 (−7.95 to −5.26). Differences were smaller for cocaine use disorder (−0.97), opioid use disorder (−0.86) and other substance use disorders (−1.12) ^[1].

Among the 81,617 people who already had a substance use disorder, starting a GLP-1 drug was associated with lower risk of substance-related emergency department visits (hazard ratio 0.69, 95% CI 0.61 to 0.78), substance-related hospital admissions (0.74, 0.65 to 0.85) and substance-related death (0.50, 0.32 to 0.79). The study also reported lower risk of drug overdose (0.61, 0.42 to 0.88) and suicidal ideation or attempt (0.75, 0.67 to 0.83). Net three-year risk differences per 1,000 people were −8.92 for emergency visits, −6.23 for admissions, −1.52 for deaths, −1.49 for overdose and −9.95 for suicidal ideation or attempt ^[1].

This was an observational study, so it shows associations rather than proof that GLP-1 drugs cause these outcomes. The authors write that the data "suggest a potential role" for the drugs in preventing and treating substance use disorders and "warrant further evaluation" ^[1]. The paper reports results for GLP-1 receptor agonists as a class and does not say which individual drugs the veterans received. Randomized trial evidence so far is limited and mixed: some small trials found reduced alcohol craving and consumption, one found no significant reduction in heavy drinking except in people with a body mass index over 30, and smoking cessation trials have produced conflicting results ^[1].

Why it matters for patients

The comparison was against another diabetes drug, not a placebo, and the population was veterans with type 2 diabetes. Whether the same pattern applies to other groups, including people who take GLP-1 drugs only for weight loss, has not been reported ^[1].

For several substances the absolute difference was small: about one fewer case per 1,000 people over three years for cocaine use disorder and opioid use disorder, and about two fewer for cannabis use disorder ^[1]. Larger differences appeared in the group that already had a substance use disorder, where the study counted about nine fewer substance-related emergency visits and about ten fewer cases of suicidal ideation or attempt per 1,000 people over three years ^[1].

The study does not establish that GLP-1 drugs can treat or prevent addiction, and it does not identify a specific drug, dose or treatment duration for substance use. The results do not change the approved uses of these medicines.

What happens next

The BMJ published a linked editorial and a linked opinion piece alongside the study ^[1]. The authors call for further evaluation of GLP-1 receptor agonists for substance use disorders. No follow-up trial or timeline is reported in the study ^[1].