Somewhere in the last two years, the conversation about weight-loss drugs moved into the therapy room. About one in eight American adults is currently taking a GLP-1 medication — semaglutide, tirzepatide, liraglutide, sold as Ozempic, Wegovy, Mounjaro, Zepbound — and nearly one in five has taken one at some point, according to KFF’s November 2025 tracking poll.
So the question of how GLP-1 drugs and mental health interact is no longer academic. It’s the client who feels strangely flat around month three and can’t tell whether that’s the medication, the weight, or the year she’s had. The evidence here has whipsawed hard, and it’s worth knowing where it stands.
The scare started with about 150 reports
In July 2023, the European Medicines Agency announced its safety committee was reviewing whether these medicines caused suicidal thoughts and thoughts of self-harm. Iceland’s medicines agency triggered the review, and regulators had retrieved roughly 150 reports of possible self-injury and suicidal ideation among people taking liraglutide and semaglutide.
That number sounds alarming until you put a denominator under it. In the same statement, the EMA noted these drugs had already accumulated over 20 million patient-years of exposure. A safety signal is a question, not a finding — as the agency put it, “The presence of a signal does not necessarily mean that a medicine caused the adverse event in question.”
The headlines had already left the building by then.
Then the large studies arrived
In September 2024, researchers at Karolinska Institutet working with Danish colleagues published a nationwide cohort study in JAMA Internal Medicine. They followed roughly 300,000 adults who started either a GLP-1 or a different class of diabetes drug between 2013 and 2021. After a mean follow-up of just over two years, they found no clear link between GLP-1 use and suicide death, self-harm, or depression- and anxiety-related disorders. Co-author Peter Ueda added a caveat: people with prior self-harm or suicidal thoughts need studying specifically, because the safety profile there might differ.
Then, in March 2026, The Lancet Psychiatry published something more pointed. Heidi Taipale and colleagues looked at 95,490 people in Sweden who already had a depression or anxiety diagnosis and were prescribed medication for diabetes; 22,480 had used a GLP-1. Rather than comparing users to non-users, they compared each person to themselves — treated periods against untreated periods in the same individual, which strips out much of the confounding that haunts this literature.
During periods of semaglutide use, the risk of psychiatric hospital care or sick leave was 42 per cent lower: 44 per cent lower for depression, 38 per cent for anxiety disorders, 47 per cent for substance use. Liraglutide came in at 18 per cent. Exenatide and dulaglutide showed nothing significant. As a class, GLP-1s were associated with a reduced risk of self-harm.
What research on GLP-1 drugs and mental health still can’t tell you
Before anyone starts prescribing Ozempic for depression, read the fine print — the researchers are holding it up themselves.
This is observational data. Nobody randomised anybody. In the expert reaction collected by the UK’s Science Media Centre, Prof Eduard Vieta of the University of Barcelona noted that residual confounding can’t be excluded and that the pandemic itself may have skewed the results. Prof Ian Maidment of Aston University pointed out the study ran in a single country, without data on ethnicity, symptom severity, or weight change. Prof David Nutt of Imperial College London was blunter: he thinks it unlikely GLP-1s alone will work as treatments for depression or anxiety.
There’s a subtler limitation worth sitting with. The study measured a lower risk of worsening — fewer hospitalisations, fewer sick days. That is not the same as feeling better. A medication can reduce your odds of a psychiatric admission without changing much about how Tuesday afternoon feels. If you’ve read our piece on what the research actually supports for anxiety supplements, the pattern is familiar: the honest answer is usually narrower than the headline.
The part the trials don’t measure
Here is where clinical experience runs ahead of the data. Rebecca Boswell, PhD, who directs the Penn Medicine Princeton Center for Eating Disorders, has raised concerns no cohort study captures. For weight loss, she points out, these medications are prescribed at two to five times the dose used for diabetes. And there’s a gap in the system: “There’s no protocol in place to screen for eating disorders prior to prescribing GLP-1 receptor agonists,” she says. Nor, she notes, are people routinely monitored for the psychological and medical effects of malnutrition — which can be an issue at any body size.
She adds something a therapist will recognise immediately: “These medications don’t allow people to experience their hunger cues and practice intuitive eating, which is an important goal in eating disorders treatment.” For someone with a history of restriction, a GLP-1 can quietly become a compensatory behaviour with a prescription attached.
Then there’s what surfaces around month two or three. If food was how you managed a long day — and for a great many people it was — the medication doesn’t only reduce appetite. It removes a coping strategy. What comes up afterward isn’t a side effect in the pharmacological sense. It’s the grief or boredom the strategy was covering. That’s not a reason to stop the drug. It’s a reason to have somewhere to put it.
And then there’s stopping
Most people do stop. In the KFF poll, 14 per cent of users said they’d quit over cost and 13 per cent over side effects.
The STEP 1 trial extension followed 327 people who completed 68 weeks on semaglutide and then came off it, along with the lifestyle programme. They had lost an average of 17.3 per cent of their body weight. A year later they had regained about two-thirds of it, landing at a net loss of 5.6 per cent. The authors read this as confirmation that obesity is a chronic condition requiring ongoing treatment.
That’s a hard landing if nobody warned you. If your sense of self was rebuilt on a loss that was being pharmacologically maintained, coming off the drug is not only a metabolic event. Have that conversation before you need it.
What to actually do with this
- Tell your therapist the date you started. Attribution is nearly impossible in hindsight; a timeline makes it tractable.
- Track your mood in writing for the first three months. Memory is a poor instrument for this.
- Disclose any history of disordered eating to the prescriber — even old, even resolved. Nobody may think to ask.
- Notice the “just a little more” thought. Wanting to lose more after hitting a healthy target is a signal, not a goal.
- Treat appetite loss as a mood variable. Barely eating because nothing appeals can flatten you on its own.
- Don’t stop abruptly on your own if mood is the reason. Call the prescriber first.
And if suicidal thoughts show up at any point — on the medication, off it, or unrelated to it — that isn’t something to wait out until your next appointment. In the US you can call or text 988 to reach the Suicide & Crisis Lifeline, free and 24/7. Tell your prescriber too. Contradictory headlines are exactly the conditions under which people talk themselves out of asking for help.
The reassuring reading of the evidence is real: two large studies failed to find the harm the 2023 signal suggested, and the more recent one found the opposite. But “reassuring at the population level” and “fine for you specifically” are different claims. Your own history and your reasons for wanting this are the data that matter — worth bringing to a prescriber who knows your psychiatric history, and to a therapist who can hear what the weight was carrying. If you’re unsure whether it’s time, our guide to the signs it may be worth talking to a therapist is a start, and finding the right therapist matters more than starting quickly.
This article is for general information only and is not a substitute for medical or mental health advice, diagnosis, or treatment. Do not start, change, or stop any medication based on what you read here — talk to the clinician who prescribed it. If you are having thoughts of suicide or self-harm, call or text 988 in the US to reach the Suicide & Crisis Lifeline, available free and 24/7, or contact your local emergency services.