Do GLP-1 Drugs Reduce Repeat Overdoses? A Study of 680 Thousand People
Used to treat obesity and type 2 diabetes, GLP-1 receptor agonists — popularly known as "weight-loss injections" — have attracted attention in an unexpected area in recent years: addiction. A new record review examined 683.800 adults who had attended an emergency department for an overdose and found a striking association: in people who started treatment after the overdose event, GLP-1 use was associated, within three years, with a 39% lower risk of a repeat overdose.
The headline stops here. But the most interesting detail — and the one at the center of this article — is that even people who had stopped the drug had a 20% lower risk. The study interprets this as "a sign that benefit may extend beyond treatment." We think this finding may also carry precisely the opposite warning — and explain why in detail below. We also include randomized study data missing from the source news report that genuinely advance the discussion.
The researchers used records from nearly 2.200 hospitals and 50.000 clinics in the US, held in the Cosmos database. Adults who presented to an emergency department or as inpatients between January 2017 and March 2026 because of an overdose were included: 683.800 adults were examined. Overdoses included opioids, alcohol, stimulants, sedatives/sleeping medications, cannabinoids, cocaine, and hallucinogens.
A noteworthy methodological choice was made: people prescribed GLP-1 before the overdose event were excluded. Thus, all GLP-1 use began after the event. Demographics, social vulnerability, rural/urban residence, body mass index category, diabetes, mental health diagnoses, and opioid/alcohol/nicotine dependence medications were also taken into account. GLP-1 use was modeled as a time-varying variable: active use, discontinued use, or never used.
The endpoint was a further event following the initial overdose: between 30 days and 3 years, a new emergency department attendance or hospital admission for an overdose.
The finding also appeared consistent across subgroups:
| Subgroup | Risk reduction with active GLP-1 use | Risk reduction after discontinuation |
|---|---|---|
| People whose first overdose involved alcohol as the substance | %40 | %18 |
| People whose first overdose involved substances other than alcohol as the cause | %37 | %24 |
| People who never received addiction treatment medication | %44 | %28 |
| Entire group | %39 | %20 |
The attractive part of the headline is "benefit continues even after stopping the drug." The study interprets this as a sign that the effect may extend beyond active treatment. This is possible. However, the same finding has another explanation that is much simpler and more likely .
If a drug's effect persists after discontinuation, there are two possibilities. (1) The drug has caused a lasting biological change — possible, but unproven. (2) What we are measuring is not the drug's effect, but the fact that people taking it were different from the outset.
The second explanation is a well-known trap in epidemiology. Consider: who is prescribed GLP-1 after an overdose? Probably someone who attends follow-up, has a doctor, has insurance covering the drug, can fill the prescription, and has the capacity to continue treatment. All these characteristics are already associated with a lower repeat risk in their own right. Thus, the groups receiving and not receiving the drug may differ independently of the medication.
Persistence of benefit after discontinuation is exactly what this second explanation predicts. People's characteristics do not disappear with the drug. Thus, rather than telling us "the effect persists," the finding may warn us that "something other than the drug may also be playing a role here" .
To be fair, the researchers did not ignore this problem. In the analysis, outpatient attendance before and after the overdose was included among measures of contact with healthcare, and social vulnerability, mental health diagnoses, and addiction treatment medications were also added to the model. These are substantial and appropriate steps. However, statistical adjustment cannot eliminate unmeasured differences: motivation, social support, secure housing, and treatment adherence are not visible in records, but determine outcomes. This is why the study itself presents the result as an "association," not an "effect."
We have been critical so far. But honesty requires showing the other side of the evidence too. Possible effects of GLP-1 drugs on addiction are not speculation based solely on observational data. There are now randomized controlled studies in this field — and the source news report did not cover them.
- JAMA Psychiatry (2025) — the first randomized study: in adults with alcohol use disorder, low-dose semaglutide, compared with placebo, reduced alcohol craving, the amount consumed, and heavy drinking days. Consumption measured in a laboratory setting — based on observation rather than self-report — also fell. Daily cigarette numbers decreased in the smoking subgroup as well. Limitation: only 48 participants, 9 weeks, low doses, and a group not seeking treatment.
- The Lancet (2026) — a larger and stronger study: in people with alcohol use disorder and coexisting obesity, a study involving 108 people used a randomized, double-blind, placebo-controlled design; weekly semaglutide significantly reduced heavy drinking days compared with placebo (treatment difference about 13,7 points; p=0,0015). Side effects were mostly transient, mild-to-moderate gastrointestinal complaints.
- BMJ (2026) — a cohort study in veterans: a study examining the association between GLP-1 use and substance use disorder risk in US veterans with type 2 diabetes, with results in a similar direction, that underwent peer review .
This substantially changes the picture. We now have not just "an interesting correlation," but a placebo-controlled, replicated signal of an effect, at least for alcohol. This provides a strong reason to take seriously the hypothesis that GLP-1 drugs affect the brain's reward and craving circuits — a hypothesis with a solid evidential rationale.
However, the crucial distinction is that the randomized evidence concerns reducing alcohol consumption. The claim discussed in this article is different: preventing repeat overdoses. These are related, but not the same — and there is no randomized study of the overdose endpoint yet.
GLP-1 receptors are present not only in the pancreas and gastrointestinal tract, but also in the brain's reward and motivation circuits. The hypothesis is that, while suppressing appetite, these drugs may use the same circuits to blunt the feeling of "craving" itself — which may be directed not only at food, but also at alcohol and other substances.
Clinical observation fits this hypothesis: beyond weight loss, some people taking these drugs describe a change such as "I no longer enjoy alcohol." This appears different from satiety — a reduction in the substance's appeal. However, this effect does not occur in everyone; experts say it is pronounced in about a third of patients. The effects of GLP-1 drugs on sexual desire previously discussed on our site also provide an interesting parallel supporting this "general craving/reward system" hypothesis.
"Weight-loss injections also treat addiction — reducing overdose risk by 39% and protecting you even after you stop!"
This is an observational record review and has not undergone peer review. It does not demonstrate causality. The finding "protection even after stopping" may suggest confounding rather than lasting benefit. GLP-1 drugs have no approved indication for addiction. Moreover, they do not work for everyone and may adversely affect mood in some people.
- Adverse effects on mood: some people taking these drugs report feeling worse, more depressed, or more anxious. In a psychologically vulnerable group with a history of addiction, this is not an insignificant detail. Experts clearly emphasize listening to how the patient feels on the drug and reducing the dose if necessary.
- Abrupt discontinuation can cause problems: rather than stopping the drug suddenly, gradual dose reduction is recommended; abrupt discontinuation may make a person feel unwell or allow cravings to return.
- The study includes only overweight people: all participants had a body mass index of 25 or above. Findings therefore cannot be generalized directly to normal-weight patients with addiction.
- These drugs do not replace addiction treatment. Proven, approved medications and psychosocial treatments exist for opioid and alcohol use disorders. GLP-1 drugs are not alternatives to these.
- Obtaining them without oversight carries serious risks. For GLP-1 products purchased online or from unreliable sources, incorrect administration and accidental overdose reports exist. This is a bitter irony, especially given that the topic itself is overdose.
- It is observational; it does not prove causality. "Risk is lower in GLP-1 users" and "GLP-1 reduces risk" are not the same statement.
- It has not undergone peer review. Findings were shared on the publication platform of the organization operating the database. A large sample does not compensate for a lack of scrutiny.
- Selection bias is the main problem. Who receives GLP-1 after an overdose is not random, and these people may already be at lower risk.
- The finding "benefit even after stopping" can be read in two ways — a lasting biological effect or residual confounding. The latter is more likely.
- There is no randomized evidence for the overdose endpoint. There is such evidence for alcohol consumption; these are not the same claim.
- There is no approved indication for addiction — from FDA, EMA, or TİTCK. Use for this purpose is "off-label."
- This article does not recommend using any drug. GLP-1 drugs are used only for their approved indications, under medical assessment and supervision.
If you or a relative is struggling with alcohol or substance use, it is understandable that such news seems hopeful. However, we must be clear: GLP-1 drugs are not currently approved to treat addiction and should not be obtained on your own for this purpose. This is because efficacy has not been proven and these drugs have real side effects.
Instead, the good news is that, for alcohol and opioid use disorders, approved drug treatments and psychosocial support methods with efficacy demonstrated in randomized studies already exist — and are accessible. If you already use a GLP-1 drug and have noticed a change in alcohol/substance cravings, sharing this with your doctor provides valuable information. Similarly, if these drugs make you feel more depressed or anxious, be sure to report it; this should be discussed, not kept silent.
This article addresses a sensitive subject — substance use and overdose — within a scientific framework. If you or a relative is struggling with this, you are not alone, and help is available. In Turkey, free addiction counseling lines and specialist treatment services at AMATEM (Alcohol and Substance Addiction Treatment Centers) are available. In an emergency, 112 should be called.
Addiction is not a weakness of character, but a treatable medical condition. Taking the first step — telling a doctor, a center, or someone you trust — is often the hardest but most decisive step.
The hypothesis that GLP-1 drugs may affect the brain's reward and craving circuits beyond their metabolic effects is being taken increasingly seriously. An association of this magnitude with a severe endpoint such as overdose shows that the subject deserves investigation in randomized studies.
Cosmos data examined 683.800 adults; a 39% reduction in repeat overdose risk with active GLP-1 use and a 20% reduction after discontinuation were reported. Results were consistent across subgroups. Two randomized placebo-controlled studies in alcohol use disorder (JAMA Psychiatry 2025; Lancet 2026) also showed a significant reduction in heavy drinking days.
Observational; not peer-reviewed; susceptible to selection bias. "Benefit even after stopping" may indicate residual confounding. All participants had a body mass index of 25 or above — generalizability is limited. There is no randomized evidence for the overdose endpoint and no approved indication. Increased depression/anxiety was reported in some patients.
It does not change treatment practice today; GLP-1 is not prescribed for addiction. However, continuing the drug may gain an additional rationale in patients using GLP-1 for an approved indication who also have a substance use problem. The drug should not be stopped suddenly, but tapered; changes in mood should be actively assessed.
What exactly is the mechanism — reward circuits, inflammation, or another pathway? Which subgroup benefits most? Where do GLP-1 drugs stand compared with existing approved addiction medications, or should they be added to them? Will randomized evidence emerge for substances other than alcohol (opioids, stimulants)? Is there an effect in normal-weight people too?
- Epic Research. GLP-1 treatment associated with a lower risk of repeat substance overdose. Cosmos database analysis, July 2026. (Not peer-reviewed.) epicresearch.org.
- News: GLP-1s may reduce overdose risk, even after discontinuation. Medscape Medical News, July 2026.
- Hendershot CS, Bremmer MP, Paladino MB, et al. Once-weekly semaglutide in adults with alcohol use disorder: a randomized clinical trial. JAMA Psychiatry. 2025;82(4):395–405. doi:10.1001/jamapsychiatry.2024.4789.
- Once-weekly semaglutide versus placebo in patients with alcohol use disorder and comorbid obesity: a randomised, double-blind, placebo-controlled trial. The Lancet. 2026.
- Cai M, Choi T, Xie Y, Al-Aly Z. Glucagon-like peptide-1 receptor agonists and risk of substance use disorders among US veterans with type 2 diabetes: cohort study. BMJ. 2026;392:e086886.
- Wang W, Volkow ND, Berger NA, et al. Associations of semaglutide with incidence and recurrence of alcohol use disorder in real-world population. Nature Communications. 2024;15:4548.
- Related reading (drozdogan.com): New-generation weight-loss drugs: the history and future of GLP-1 agonists · What is semaglutide (Ozempic)? · What is Wegovy? · GLP-1 drugs and sexual desire · Weight-loss injections and cancer risk.
Editor's note: This content is for general information only; it is not medical advice or a recommendation for diagnosis or treatment. The data presented have been verified against primary sources and conveyed without direct quotation. The study underlying the main finding is an observational record review and has not undergone peer review; the findings demonstrate an association, as observed, not proof of causality. Because who receives GLP-1 after an overdose is not determined randomly, selection bias is an important limitation; benefit also seen after discontinuation may indicate residual confounding rather than a lasting effect. Only people with a body mass index of 25 or above were included. Although randomized evidence exists for reducing alcohol consumption, there is no direct randomized evidence for preventing repeat overdoses. GLP-1 receptor agonists have no approved indication for addiction treatment (FDA, EMA, TİTCK); use for this purpose is off-label and not recommended. These drugs may have serious side effects, and in some people they have been reported to cause increased depression or anxiety . This article does not recommend starting, stopping, or changing any medication; drugs are used only under medical supervision. This article addresses a sensitive subject; people struggling with substance use are advised to contact a healthcare facility, AMATEM centers, or a doctor they trust. For every decision about your health, consult a doctor who knows you.
Was this article helpful?
Your feedback helps us improve our content.
