Explainer · Brain & Mental Health
GLP-1 treatment went with 45% fewer alcohol-related hospital stays, in 167,026 Swedes with diabetes
People on GLP-1 drugs often report drinking less. Swedish registry data compared the same individuals on and off treatment, and found fewer hospital admissions for alcohol problems while they took them.
- An observational study of 167,026 adults with type 2 diabetes in Stockholm County, followed to the end of 2024.
- Each person acted as their own comparison: periods on a GLP-1 drug against periods off it.
- During treatment, alcohol-related hospital admissions were 45% lower and substance-related ones 39% lower.
- The lower rate persisted for about six months after stopping, then faded.
- A comparison drug class, DPP-4 inhibitors, showed no consistent reduction.
One of the odder things people report on GLP-1 drugs is that they stop wanting a drink. Anecdotes became a research question, and researchers have been working through registry data while randomized trials run.
A study in The Lancet Psychiatry used Swedish records for 167,026 people with type 2 diabetes and compared each person with themselves, on treatment and off it. Hospital admissions for alcohol problems were markedly lower during GLP-1 treatment.
What alcohol use disorder is
MedlinePlus explains that alcohol misuse means that drinking causes distress and harm and that over time alcohol misuse can increase your risk of AUD, alcohol use disorder. Severe cases are sometimes called alcoholism or alcohol dependence, and the condition is treatable.
Hospital admissions are the severe end. The authors set out what was missing: evidence suggests that GLP-1 receptor agonists might reduce substance use across different substance use disorders, but data on what happens after people stop the drugs were absent.
How the Swedish GLP-1 and alcohol study worked
The team used national registers covering Stockholm County to find residents with a first registered diagnosis of type 2 diabetes between 2005 and 2024, then followed them to the end of 2024.
The design is the notable part. Rather than comparing people who took GLP-1 drugs with people who did not, they compared periods when the same person was on treatment against periods when they were not, which removes fixed differences between individuals. They did the same for DPP-4 inhibitors, another diabetes drug class, as a check.
What happened to hospital admissions
The admissions were rare but the pattern was clear. GLP-1 exposure was associated with lower rates of hospitalisations related to alcohol use disorder, 45% lower, and 39% lower for substance use disorder admissions overall, most of which were alcohol-related.
The after-effect was shorter. For alcohol use disorder, the reduction persisted during days 1-182 after discontinuation of the drugs, at about 30% lower, then disappeared in the following six months. DPP-4 inhibitors were not associated with consistent rate reductions, which argues against the finding being a general effect of diabetes treatment.
Why the alcohol signal matters
Alcohol use disorder is common, undertreated and hard to treat, and the existing medicines are used far less than they could be. A drug already taken by millions that also dampened drinking would matter.
There is a second, practical reading. If the benefit fades within months of stopping, then people with drinking problems who come off a GLP-1 drug, for cost, side effects or supply reasons, may be at risk of relapse at exactly that moment. The authors say their findings support clinical monitoring of substance use when GLP-1 receptor agonist treatment is stopped.
What a registry study of GLP-1 drugs cannot prove
Comparing people with themselves removes stable traits but not changing circumstances. Someone starting a GLP-1 drug may also be losing weight, seeing clinicians more often or making other changes, all of which could affect drinking.
Hospital admissions capture only severe episodes, so this says nothing about how many drinks people had. Everyone in the study had type 2 diabetes, in one Swedish county, and drinking problems were uncommon: 1,559 people had an alcohol-related admission. Randomized trials testing GLP-1 drugs specifically for alcohol use disorder are the evidence that would settle it.
What this changes for people on GLP-1 drugs who drink
Nothing about treatment today. If drinking is a problem, the route is the established one: talk to a doctor, ask about counseling and about the licensed medicines that reduce craving and relapse.
For people already on a GLP-1 drug who have noticed their drinking drop, this study offers a reason to watch for it creeping back if the treatment stops. For clinicians, the same applies: a stopped prescription is a moment to ask about alcohol.
People also ask
What did the study find?
GLP-1 exposure was associated with lower rates of hospitalisations related to alcohol use disorder (rate ratio 0.55, 95% CI 0.43 to 0.70) and substance use disorder (0.61, 0.50 to 0.74) compared with unexposed periods in the same individuals. For alcohol use disorder the reduction persisted during days 1 to 182 after stopping (0.70, 0.50 to 0.98) but not from day 183 onward (1.07, 0.71 to 1.63).
What is a within-individual design?
Each person is compared against themselves at different times, on and off the drug. That removes stable differences between people, such as genetics, personality or long-standing circumstances, though it cannot remove things that change over time.
Who was included?
167,026 adults in Stockholm County with a first recorded diagnosis of type 2 diabetes between 2005 and 2024. About a quarter received a GLP-1 drug at some point during follow-up.
How common were these admissions?
Uncommon: 0.9% of participants had at least one alcohol-related hospitalisation and 1.2% at least one substance-related admission. Two-thirds of the substance-related admissions were attributed to alcohol.
Are GLP-1 drugs a treatment for alcohol problems?
Not yet. Randomized trials are under way. This study is observational and involved people prescribed the drugs for diabetes, not for drinking.
What should someone drinking too much do?
Effective treatments for alcohol use disorder already exist, including counseling and licensed medicines. Talk to a doctor rather than waiting for a GLP-1 prescription. This is general information rather than medical advice.
References
- Association of GLP-1 receptor agonists with alcohol use disorder-related and substance use disorder-related hospital admissions during treatment and after discontinuation: a Swedish register-based within-individual observational study. The Lancet Psychiatry, 2026.
- MedlinePlus. Alcohol Use Disorder (AUD). US National Library of Medicine.