BMJ Study Links GLP‑1 Diabetes Drugs to Lower Addiction Risk
An NPR report details a new study in The BMJ of more than 600,000 U.S. veterans with type 2 diabetes finding that patients who started GLP‑1 drugs such as Ozempic were about 15–20% less likely to develop substance use disorders involving alcohol, nicotine, cannabis, cocaine or opioids than similar patients on other diabetes medications. Among veterans with a prior history of substance use disorder, GLP‑1 users had a 25–50% lower risk of emergency department visits, overdoses, drug-related hospitalizations, suicidal thoughts or attempts, and drug-related death, suggesting these drugs may blunt a common biological pathway underlying addiction. Lead author Dr. Ziyad Al‑Aly of Washington University in St. Louis and the VA system says the cross‑substance effect was a surprise and points to a shared "biologic signal," echoing anecdotal reports from clinicians that some GLP‑1 patients lose interest in alcohol or cigarettes. Outside experts at NIH and UNC call the results promising but emphasize that the observational design cannot prove causation and that GLP‑1s have not yet been properly tested as addiction treatments in people without diabetes or obesity, with several randomized clinical trials now underway. The findings sharpen debate in the U.S. over how far to extend use of these already widely prescribed and expensive drugs, potentially into addiction medicine, at the same time other studies are surfacing longer‑term safety questions such as bone‑health risks in older adults.
📌 Key Facts
- The study, published in The BMJ and reported March 10, 2026, analyzed health records from more than 600,000 U.S. veterans with type 2 diabetes.
- Veterans who initiated GLP‑1 drugs had roughly a 15–20% lower risk of being diagnosed with substance use disorders across multiple substances compared with those on non‑GLP‑1 diabetes medications.
- Among patients with pre‑existing substance use disorders, GLP‑1 users had an estimated 25–50% lower risk of emergency visits, hospitalizations, overdoses, suicidality and drug-related deaths.
- Researchers and outside experts caution the results are observational and say randomized clinical trials now underway are needed before GLP‑1s can be considered established treatments for addiction.
📊 Relevant Data
In 2022, the prevalence of past-year substance use disorder among people aged 12 or older ranged from 9.0% for Asian people to 17.2% for American Indian or Alaska Native people, with 16.8% for Native Hawaiian or Other Pacific Islander, 15.6% for people of two or more races, 14.9% for Hispanic or Latino, 11.9% for Black or African American, and 10.9% for White people.
Highlights by Race/Ethnicity for the 2022 National Survey on Drug Use and Health — SAMHSA
In 2023, the age-adjusted drug overdose death rate was highest for American Indian and Alaska Native non-Hispanic people at 65.0 per 100,000, followed by Black non-Hispanic at 45.9, White non-Hispanic at 32.9, Hispanic at 24.3, and Asian non-Hispanic at 6.8.
The prevalence of diagnosed diabetes among US adults is 15.7% for American Indian or Alaska Native, 12.2% for Black non-Hispanic, 11.8% for Hispanic, 9.7% for Asian non-Hispanic, and 7.1% for White non-Hispanic.
Diabetes in America: Prevalence, Statistics, and Economic Impact — American Diabetes Association
Black and Hispanic patients with diabetes are less likely to be prescribed GLP-1 receptor agonists compared to White patients, with adjusted odds ratios of 0.88 for Black and 0.85 for Hispanic patients.
Genome-wide association studies have identified shared genetic markers across multiple substance use disorders, with heritability estimates for substance use disorders ranging from 30-70% depending on the substance.
New NIH study reveals shared genetic markers underlying substance use disorders — NIDA
📰 Source Timeline (1)
Follow how coverage of this story developed over time