From liraglutide (Saxenda) to tirzepatide (Mounjaro), new weight loss medications are reshaping the treatment of diabetes and obesity. But what’s their impact in psychiatry? In this article, we’ll look at the promise and peril of using these glucagon-like peptide-1 receptor agonists (GLP-1 RAs) to treat metabolic side effects and explore other potential uses for these medications in psychiatry.

How they work

GLP-1 is a short-lived hormone released in the gut when we eat. The new weight loss medications mimic this hormone but are engineered to last a lot longer. They slow stomach emptying, causing the vagal nerve to signal fullness to the brain. At the same time, these medications cross the blood-brain barrier and directly signal satiety at the hypothalamus. They stimulate the growth of new neurons in the hippocampus and improve brain meta­bolism, which has attracted the attention of Alzheimer’s researchers. Their anti-­inflammatory and insulin sensitization properties in the brain may help explain the antidepressant effects seen in some patients. They also promote the growth of dopaminergic neurons and have been tested in Parkinson’s disease. 

Obesity

Three of the GLP-1 RAs are FDA approved for obesity: liraglutide (Victoza, Saxenda), semaglutide (Ozempic, Wegovy), and tirzepatide (Mounjaro, Zepbound) (these have separate brands for diabetes and obesity, and we have listed the diabetes brand first). Specifically, they are approved in patients who have a BMI ≥30 kg/m² or who have a BMI ≥27 kg/m² and an obesity-­related health condition (eg, diabetes, ­hypertension, or dyslipidemia).

Among them, tirzepatide is the most effective, with average weight loss of about 20% of body weight (Jastreboff AM et al, N Engl J Med 2022;387(3):205–216). Tirzepatide targets both GLP-1R and the glucose-dependent insulinotropic peptide receptor (GIPR), while the older two medications focus on GLP-1R alone. Between these older two, semaglutide appears most effective with an average of 10%–15% of body weight loss over a year, followed by liraglutide with ~5%–10% of body weight lost.

Among the GLP-1 RAs that lack approval in weight loss, exenatide (Byetta) is slightly less effective than liraglutide for weight loss, and dulaglutide (Trulicity) lacks evidence for weight loss outside of trials in diabetes. 

Even stronger drugs may be on the horizon. Retatrutide, with triple-target effects (it works at the two receptors mentioned above and the glucagon receptor), caused people to lose about a quarter of their body weight after a year in a Phase II study (Jastreboff AM et al, N Engl J Med 2023;389(6):514–526), although a follow-up study found the weight loss more in line with tirzepatide.

Antipsychotic weight gain

What sets the GLP-1 RAs apart from alternatives for antipsychotic-induced weight gain (like metformin and olanzapine/samidorphan) is their ability to reverse weight gain after it has already started to get out of control. In a small open-label trial of exenatide in people with clozapine-associated obesity (BMI 30–45 kg/m²), those on exenatide had about 10 pounds more weight loss than placebo at 24 weeks (Siskind D et al, J Psychiatr Res 2020;124:9–12). Liraglutide worked even better, with a mean difference of about 12 pounds in a 16-week study in patients on clozapine or olanzapine. While more studies in psychiatric populations are needed, obesity and antipsychotic-induced obesity are more alike than different, and it is likely that other GLP-1 RAs will work in this population as well. 

By comparison, Lybalvi, the opioid antagonist combination of olanzapine and samidorphan, has only been tested for its ability to reduce weight gain in non-obese patients. Similarly, metformin is most effective when starting a new antipsychotic, before weight gain gets out of control. In one meta-analysis of RCTs, metformin worked about three times as well when starting a new antipsychotic in first-episode psychosis (about a 13-pound difference from placebo) compared with its use after chronic antipsychotic treatment (about a 4-pound difference from placebo) (De Silva VA et al, BMC Psychiatry 2016;16(1):341).

Between these two, metformin is our first-line preference as it helps not only with antipsychotic-related weight gain but also with the metabolic side effects of antipsychotics (insulin resistance, fasting glucose, triglycerides, and hyperprolactinemia), earning it a recommendation from the APA for this purpose. Metformin also has evidence in patients with a BMI ≥30 kg/m². But neither it nor olanzapine/samidorphan can match the effects of GLP-1 RAs in patients with high BMIs after chronic antipsychotic use. 

Psychiatric effects

Substance use

As the use of GLP-1 RAs expanded, clinicians noticed that some patients taking them had decreased cravings for alcohol. This made sense to researchers, who had seen the GLP-1 RAs dial down reward-response circuitry in animal models. So far, the clinical data are mixed but intriguing. Dulaglutide did not work in an RCT for smoking cessation in mostly obese adults, but alcohol consumption decreased in the treatment group (Lengsfeld S et al, eClinicalMedicine 2023;57:101865; Probst L et al, JCI Insight 2023;8(22):e170419). Exenatide did not work in cocaine use disorder, but it did reduce drinking in obese patients with alcohol use disorder (Klausen MK et al, JCI Insight 2022;7(19):e159863). These results were arrived at by ­data-fishing techniques, but a more recent, small RCT of people with alcohol use disorder found that nine weeks of low-dose semaglutide (0.25–1 mg/week) reduced the main outcome of interest: average alcohol consumption (Hendershot CS et al, JAMA Psychiatry; published online February 12, 2025). These medications may be particularly suitable for those looking to lose weight and cut down on drinking, but larger studies are needed to determine their place among strategies for treating alcohol use disorder.

Mood symptoms

Animal studies, case reports, and a meta-analysis of RCTs of GLP-1 RAs suggest that these medications might help with depression (Chen X et al, Am J Geriatr Psychiatry 2024;32(1):117–127). At the same time, they worsened mood symptoms in case reports. We have not yet seen any actual clinical trials of a GLP-1 RA for depression to help clear up their role here. 

Binge eating disorder (BED)

There is currently only one FDA-­approved medication for BED: lisdexamfetamine. Case reports and a cohort study suggest GLP-1 RAs may work for BED (Richards J et al, Obes Pillars 2023;7:100080). Liraglutide 3 mg/day reduced binge eating in a small RCT of stable bipolar patients with obesity, both on and off antipsychotics, without changing mood symptoms (McElroy SL et al, J Clin Psychopharmacol; published online January 16, 2024).

Cognition

Animal studies suggested the GLP-1 RAs may improve cognition, but results in humans have been inconclusive. Liraglutide improved cognition in an open-­label study of mood disorders, but exenatide failed to do the same in a small RCT of schizophrenia with obesity. Likewise, studies in dementia have been mixed, but more definitive trials are underway (Mansur RB et al, J Affect Disord 2017;207:114–120; Ishøy PL et al, Acta Psychiatr Scand 2017;136(1):52–62).

Risks 

Nausea, vomiting, constipation, and diarrhea are the most common side effects, but they tend to improve with time. In trials, these side effects occurred in about a third of patients on the drug, approximately twice the rate on placebo. Muscle loss is another problem, and it appears to be an effect of the weight loss rather than the GLP-1 RA itself. To reduce this risk, titrate slowly and advise patients to increase protein intake (eg, 20–40 g of protein at each meal). 

Patients with a history of pancreatitis or medullary thyroid cancer, or with a personal or family history of multiple endocrine neoplasia 2 (MEN2), are advised not to take these drugs. The concern about thyroid tumors comes from animal data, and it’s not clear whether there is an elevated risk in people. 

Some people have reported worsening suicidal ideation, and given the known association between bariatric surgery and suicidality, these reports should be taken seriously and discussed with patients while we await further research on this potential connection. However, recent studies have started to allay these fears; if anything, the GLP-1 RAs may be protective (McIntyre RS et al, Expert Opin Drug Saf 2024;23(1):47–55; Strumila R et al, Eur Neuropsychopharmacol 2024;87:29–34). The risk of gastroparesis has prompted surgeons to recommend stopping GLP-1 RAs before certain procedures. 

How to use them

So which should you prescribe: liraglutide, semaglutide, or tirzepatide? The choice often comes down to cost and availability. Liraglutide is the best studied in psychiatric populations and came off patent in December 2023, but the downside is that it requires daily injections. Semaglutide and tirzepatide are more effective for weight loss and are weekly injections. Educate your patients in advance about the GI side effects, and titrate slowly if those symptoms come up. If slow titration does not work to minimize side effects, suggest small food portions and make sure any reflux symptoms are appropriately treated (a short course of a proton pump inhibitor is most commonly recommended) (Gorgojo-Martínez JJ et al, J Clin Med 2022;12(1):145). If all that fails, you can switch to another GLP-1 RA. In the general population, most people come off these drugs after a year (in line with the duration of most trials), although when effective, patients can stay on them for longer—which may be necessary for patients on long-term antipsychotic therapy. 

When medications are added to the FDA’s shortage list (which was recently the case for semaglutide and tirzepatide), drug compounders are allowed to prepare them at a lower cost. This means that some specialty pharmacies are selling replications of the patented medications, with an unclear source. Given how hard it is to get these medications for some patients, this can be an attractive alternative, but the FDA has expressed concern at the frequency of -glutide salts turning up in compounded formulations, which are not present in any of the FDA-approved drugs and thus have an unknown safety and efficacy profile. 

Many psychiatrists—particularly those treating serious mental illness in patients on long-term antipsychotic therapy—are now  prescribing GLP-1 RAs. For mental health practitioners with less experience prescribing antidiabetic and weight loss medications, it may be appropriate to coordinate with primary care colleagues.

Carlat Verdict: To prevent metabolic side effects when beginning an antipsychotic, start with metformin, but consider a GLP-1 RA for patients who have already been on antipsychotics for a while with a BMI ≥30 kg/m² (or ≥27 kg/m² and obesity-related health problems). GLP-1 RAs can be safely combined with metformin or used on their own.