Lead
A large US observational study published in the British Medical Journal reports that glucagon-like peptide-1 (GLP-1) receptor agonists — drugs marketed for type 2 diabetes and obesity such as Mounjaro and Ozempic — were associated with lower rates of new substance use disorders and with reduced overdose, emergency attendance and death among people already using substances. The analysis followed 606,434 US veterans with type 2 diabetes for up to three years and compared GLP-1 users with patients prescribed sodium-glucose cotransporter-2 (SGLT2) drugs. Researchers found relative reductions in new alcohol-, cannabis-, cocaine-, nicotine- and opioid-related disorders and substantial declines in overdose, A&E visits and mortality for those with existing substance use. The authors and independent pharmacy bodies emphasised the observational nature of the findings and called for clinical trials to test causality.
Key Takeaways
- Study population: 606,434 US veterans with type 2 diabetes were followed for up to three years in an observational analysis published in the BMJ.
- Lower incidence among non-users: GLP-1 users showed an 18% reduced risk of alcohol-related disorders, 14% for cannabis, 20% for cocaine, 20% for nicotine and 25% for opioids versus patients on SGLT2 drugs.
- Reduced harms among people already using substances: GLP-1 treatment was associated with a 39% lower risk of overdose, a 31% lower risk of emergency A&E care and a 50% lower risk of death compared with the comparator group.
- Mechanism hypothesised: GLP-1 drugs are thought to influence brain reward pathways and reduce cravings by mimicking post-meal gut hormones that signal fullness.
- Evidence limits: the study is observational; authors and professional bodies warn it does not prove prevention or treatment of addiction and call for randomized trials.
- Related findings on weight permanence: a Cambridge meta-analysis of 48 studies in eClinical Medicine found most people regain 60% of lost weight within a year and 75% eventually, leaving about 25% of initial loss maintained long-term.
- Adherence and stigma: about half of GLP-1 users stop within a year and 75% by two years, often due to side-effects like nausea and costs; a survey of >3,000 users found two-thirds hide use and 38% report criticism.
Background
GLP-1 receptor agonists, including semaglutide (Ozempic) and tirzepatide (Mounjaro), were originally developed to improve blood glucose control in type 2 diabetes and later approved or prescribed off-label for weight management. They act on receptors that modulate insulin secretion and appetite, and mounting preclinical and clinical research suggests they also affect mesolimbic reward circuits implicated in craving and addictive behaviour. Use of these drugs has expanded rapidly in recent years, provoking debates about access, cost and long-term safety as well as interest in secondary benefits beyond glycaemic control and weight loss.
Parallel research has raised questions about durability and adherence: a meta-analysis of 48 trials reported by Cambridge researchers and published in eClinical Medicine indicates substantial weight regain after treatment cessation — about 60% regained within a year and 75% eventually, leaving roughly 25% of initial loss sustained long-term. Real-world discontinuation is high: roughly 50% stop within a year and 75% within two years, commonly because of nausea or financial burden. Social stigma is also prominent: a survey of more than 3,000 users found many conceal treatment and report criticism from peers and family.
Main Event
The US team analysed health records for 606,434 veterans with type 2 diabetes who initiated either a GLP-1 receptor agonist or an SGLT2 inhibitor and tracked outcomes for up to three years. Investigators used the SGLT2 group as an active comparator to reduce confounding by indication, and applied statistical adjustments for measured baseline differences. Primary outcomes included new diagnoses or treatment codes for alcohol-, cannabis-, cocaine-, nicotine- and opioid-related disorders; secondary outcomes focused on overdose, emergency attendance and death among those with prior substance use.
Compared with the SGLT2-treated group, GLP-1 users without prior substance diagnoses experienced lower risks across five substance categories: an 18% lower risk for alcohol-related disorders, 14% for cannabis, 20% for both cocaine and nicotine, and 25% for opioids. Among participants with existing substance use, GLP-1 exposure correlated with a 39% lower adjusted risk of overdose, a 31% lower risk of emergency department presentation and a 50% lower risk of mortality over the follow-up period.
Authors emphasised that the study design cannot prove causation: residual confounding, differences in health-seeking behaviour, or prescribing patterns could explain some or all of the associations. The paper notes biological plausibility — GLP-1 pathways intersect with reward circuitry — but stresses that randomized controlled trials are necessary to determine whether these agents directly reduce addiction risk or related harms. Professional bodies responding to the study urged cautious interpretation and further research.
Analysis & Implications
If the associations reflect a causal effect, GLP-1 drugs would represent a novel pharmacological tool that could reduce the incidence of several substance use disorders and lower the risk of overdose and death among people who use drugs. That could reshape prevention and harm-reduction strategies and expand the clinical considerations for prescribing GLP-1s beyond glycaemic and weight outcomes. However, translating observational associations into practice requires rigorous testing to avoid premature changes in clinical guidance.
Mechanistically, GLP-1 receptor agonists reduce appetite and alter dopaminergic signalling in brain regions tied to reward; animal and early human studies have shown diminished drug-seeking behaviour on some measures. Even so, the degree to which peripheral metabolic effects (reduced hunger, improved mood or sleep) versus direct central action account for the observed associations is unresolved. Distinguishing these pathways matters for designing targeted trials and for understanding which patient groups might benefit most.
There are public-health and equity considerations. GLP-1s are expensive and access varies; the NHS rollout for obesity has been described as slow, and many patients discontinue therapy within months or years. If benefits for addiction prevention or harm reduction were established, policymakers would face trade-offs between wider access and cost, and would need safeguards against widening health inequalities or diverting resources from established addiction services. Finally, stigma documented in patient surveys — many users hide treatment and report criticism — could shape uptake and adherence, complicating real-world effectiveness.
Comparison & Data
| Outcome | Relative reduction observed |
|---|---|
| Alcohol-related disorders (no prior use) | 18% |
| Cannabis (no prior use) | 14% |
| Cocaine (no prior use) | 20% |
| Nicotine (no prior use) | 20% |
| Opioids (no prior use) | 25% |
| Overdose (existing users) | 39% reduction |
| A&E attendance (existing users) | 31% reduction |
| Death (existing users) | 50% reduction |
The table summarises the study’s reported relative reductions; absolute risks and event rates were smaller and varied by outcome and subgroup. The analysis used statistical adjustment but cannot rule out residual confounding or selection biases. Follow-up lasted up to three years, a period sufficient to detect short- to medium-term associations but not long-term effects after discontinuation. The Cambridge meta-analysis of GLP-1 weight studies (48 trials) provides complementary context on durability: most patients regain weight after stopping, and discontinuation rates in real-world settings are high.
Reactions & Quotes
Senior figures in pharmacy welcomed the study as an important contribution while warning against overinterpretation. They highlighted biological plausibility but reiterated that observational work cannot demonstrate that these medicines prevent or treat addiction.
“This study adds to emerging research exploring whether GLP-1 medicines may influence brain pathways involved in reward and addiction.”
Prof Claire Anderson, Royal Pharmaceutical Society (statement)
The National Pharmacy Association noted potential therapeutic upsides and called for faster, evidence-based policy responses on access and prioritisation.
“This is a significant study and shows that weight loss treatment may have potential to give important additional therapeutic benefits.”
Gareth Jones, National Pharmacy Association (director of corporate affairs)
Unconfirmed
- Whether GLP-1 drugs causally prevent the onset of substance use disorders — the BMJ study is observational and does not establish direct causation.
- Generalisability beyond the study population — the cohort comprised US veterans with type 2 diabetes and results may not apply to women, younger people, or those without diabetes.
- Which mechanisms (direct central nervous system effects versus indirect metabolic or behavioural effects) explain the associations remain unproven.
Bottom Line
The BMJ analysis of 606,434 US veterans reports consistent associations between GLP-1 treatment and lower incidence of several substance use disorders plus markedly reduced overdose, emergency presentations and death among people already using drugs. These outcomes are biologically plausible given preclinical evidence of GLP-1 effects on reward circuits, but the observational design means causality has not been established. Clinicians, policymakers and researchers should view the results as hypothesis-generating: they justify investment in randomized clinical trials and mechanistic studies rather than immediate changes to prescribing guidelines for addiction prevention.
At the population level, any potential addiction-related benefits must be weighed against cost, access, discontinuation rates, side-effects and social stigma. If randomized trials confirm benefit, GLP-1s could become an adjunct in prevention or harm-reduction strategies, prompting debates on equity and resource allocation. For now, the prudent course is cautious optimism and rapid prioritisation of rigorous clinical research.
Sources
- The Guardian (news report summarising study) — independent news coverage
- British Medical Journal (BMJ) — academic journal (original observational study published in BMJ, study details)
- University of Cambridge / eClinical Medicine — academic press / journal (meta-analysis of 48 studies on weight regain)
- Simple Online Pharmacy survey — private survey provider (user-reported stigma and discontinuation statistics)
- Royal Pharmaceutical Society — professional body (statement from Prof Claire Anderson)
- National Pharmacy Association — trade body (comment from Gareth Jones)