Does semaglutide affect mental health? A 15-year Swedish records study of 14,694 people with bipolar disorder suggests it might: periods spent on semaglutide came with 21% fewer psychiatric hospital stays than periods off the drug, while two older GLP-1 drugs showed no such link. The finding is an association, not proof of cause, and the researchers themselves say a randomised trial is needed before anyone draws treatment conclusions.
Semaglutide is the active ingredient in Ozempic and Wegovy, Novo Nordisk’s GLP-1 receptor drugs, first developed for type 2 diabetes and now widely used for weight management. Bipolar disorder brings recurring episodes of mania or hypomania and depression, and severe episodes often need hospital care. The World Health Organization estimates that about 37 million people live with bipolar disorder worldwide. The two conditions collide often: diabetes and obesity are common in people with bipolar disorder, partly because several mood medicines drive weight gain, and nobody had previously tested whether GLP-1 drugs change the course of the illness.
What did the Swedish study actually find?
Led by Professor Mark Taylor of the School of Medicine and Dentistry at Griffith University in Australia, the research team used Sweden’s national health registers to follow 14,694 people with bipolar disorder who used diabetes medication between 2009 and 2024. The average age was 53.9 years and 60.5% were women. Of the whole group, 5,200 used a GLP-1 drug at some point: semaglutide in 70.8%, liraglutide in 39.4% and dulaglutide in 15.9%, with some people using more than one across the 15 years.
During follow-up, 5,288 people in the cohort were hospitalised for psychiatric reasons. The headline result: semaglutide use corresponded to a 21% lower risk of psychiatric hospitalisation (adjusted hazard ratio 0.79; 95% confidence interval 0.69 to 0.91) compared with periods when the same individuals were not using GLP-1 drugs. Medscape’s report on the study adds a second result: a 17% lower risk of hospitalisation specifically for bipolar relapse.
The paper, “Use of Glucagon-Like Peptide 1 Receptor Agonists and the Associated Risk of Hospitalisation in Bipolar Disorder, From a Nationwide Cohort, 2009 to 2024”, was published in Acta Psychiatrica Scandinavica in 2026.
How does a “within-individual” study work, and why does it matter?
Instead of comparing semaglutide users with non-users, the researchers compared each person with themselves: time spent on the drug versus time spent off it. This within-individual design automatically accounts for everything stable about a person, from genetics to early-life history, because those factors do not change between the two periods. It is one of the strongest designs available in observational research, and it is a large part of why this finding is being taken seriously.
Even so, it cannot rule out factors that change over time. Someone might start semaglutide during a period when their life is more stable for unrelated reasons, and that stability, rather than the drug, could explain some of the difference. Changing circumstances could still explain part of the result, which is exactly why the authors want the finding tested in a randomised controlled trial next.
Does this prove that Ozempic treats bipolar disorder?
No. A 21% lower hospitalisation risk is a striking association, but observational studies cannot prove that semaglutide itself caused it. What the study does provide is the first large-scale, long-term test of GLP-1 drugs in bipolar disorder, with a design that strips out many of the usual sources of bias. That makes it a genuine signal worth testing properly, not a treatment recommendation. Anyone hoping the answer is yes should wait for the trial.
Why would a weight-loss jab affect the brain at all?
GLP-1 receptors are not only found in the gut and pancreas; they are also present in the brain. The working hypothesis is that GLP-1 signalling could protect the brain by reducing inflammation, cellular stress and other biological processes that have been linked to bipolar disorder, potentially improving mood stability and lowering relapse risk. Diabetes, obesity and bipolar disorder often occur together and may share underlying biological pathways involving metabolism, which is what motivated the study in the first place.
There is a practical angle too. Weight gain from mood medicines is one of the most common reasons people stop taking them, and stopping medication is one of the fastest routes to relapse. A diabetes or weight drug that is metabolically helpful and possibly mood-neutral, or better, would be genuinely useful in bipolar care, even before any direct brain effect is proven.
What about liraglutide and dulaglutide?
Here the story gets specific, and it matters. Neither liraglutide nor dulaglutide was associated with a significant reduction in psychiatric hospitalisation, which suggests the effect may not extend across the whole GLP-1 drug class. The investigators noted that fewer people used these older drugs, which may have limited the statistical analyses, but the pattern is still notable: only semaglutide carried the signal. When GLP-1 drugs were evaluated as a group, their use was associated with a lower hospitalisation risk in this population, but the drug-by-drug breakdown tells the sharper story.
So “weight-loss jabs help bipolar disorder” would be the wrong summary of this study. The evidence, such as it is, points to semaglutide specifically, and generalising it to every drug in the class would go beyond what the data show.
Are GLP-1 drugs safe for people with mental health conditions?
The psychiatric safety of GLP-1 drugs has been debated, with early reports raising questions about possible effects on mood. On this evidence, Professor Taylor told Medscape that, contrary to initial reports, the GLP-1 drugs appear safe from a psychiatric perspective, and Healio’s report noted no worrying safety signals such as suicidality in the study population. That is reassuring, but it is not a green light to self-medicate.
Semaglutide interacts with diabetes treatment plans, can cause significant gastrointestinal side effects, and is not licensed for bipolar disorder. Anyone with a psychiatric condition who is taking, or considering, a GLP-1 drug should discuss it with their psychiatrist and their diabetes or weight-management clinician rather than making changes alone. The study is a reason for clinicians to pay attention, not a reason for patients to change prescriptions.
What should you take away from this finding?
First, if you have bipolar disorder and already take semaglutide for diabetes or weight, the finding is cautiously good news, but do tell your care team about any mood changes, good or bad, so they can track what is happening. Second, this is not a reason to seek out Ozempic or Wegovy off label for mood symptoms: the benefit is unproven, the trial evidence does not exist yet, and off-label use carries real risks. Third, watch for the randomised controlled trial the researchers are calling for. That is the study that could turn this signal into a genuine treatment option, possibly as an add-on to mood stabilisers for people with bipolar disorder who also have diabetes or obesity.
Records-based research like this is becoming a staple of modern medicine. On Acha Waqat we recently covered another example of health records revealing an unexpected drug bonus: the Shingrix shingles jab was linked to a lower dementia risk in large record analyses. Registry evidence also underpins this winter’s RSV guidance, where a single antibody dose cut infant hospital stays, and national surveillance keeps our measles coverage current as cases rise across Europe ahead of autumn travel.
Frequently asked questions
Does semaglutide affect mental health?
A 2026 Swedish study of 14,694 people with bipolar disorder found that periods on semaglutide were linked to 21% fewer psychiatric hospital stays than periods off the drug. It is an association, not proof, and a randomised trial is needed.
How many people were included in the Swedish study?
14,694 people with bipolar disorder who used diabetes medication, followed in Sweden’s national registers from 2009 to 2024. Of these, 5,200 used a GLP-1 drug, most commonly semaglutide (70.8%).
Does Ozempic treat bipolar disorder?
No. The study shows an association between semaglutide use and fewer hospitalisations, not a proven treatment effect. Semaglutide is not licensed for bipolar disorder.
Did other weight-loss jabs show the same effect?
No. Liraglutide and dulaglutide were not significantly linked to fewer psychiatric hospitalisations, suggesting the signal may be specific to semaglutide rather than the whole GLP-1 class.
Is semaglutide safe if I have a mental health condition?
The study found no worrying psychiatric safety signals, and the lead researcher described the drugs as appearing safe from a psychiatric perspective. Any change to medication should still go through your own clinicians.
Should I ask my doctor about semaglutide for my mood?
You can raise the study with your psychiatrist, especially if you already take a GLP-1 drug for diabetes or weight. It is not a basis for starting semaglutide off label for mood symptoms on your own.
