Patients taking clozapine routinely gain as much as 20 or 30 kg, with weight gain being the most common reason cited for discontinuation of the antipsychotic. Until recently, clinicians had little to offer them.

Then, in July 2025, researchers leading the phase 2 COaST trial in Australia reported that patients with schizophrenia taking clozapine showed a 13.9% body weight reduction over 36 weeks of treatment with the GLP-1 semaglutide, with no change in drug concentrations and no serious adverse events attributed to treatment. Earlier this year, a meta-analysis of COaST and two other trials showed an average weight loss of 11 kg across 258 patients — roughly four times what metformin delivers in this population.

Semaglutide’s effect was the largest weight reduction any pharmacologic intervention has achieved in clozapine patients. But that success raises a question: What are the effects when both drugs are prescribed long term?

Alone, clozapine’s most lethal side effect is gastrointestinal (GI) hypomotility, which can progress from constipation to paralytic ileus, bowel obstruction, and death. GLP-1s also cause hypomotility. Could that synergistic combination exacerbate these side effects?

Even though this question remains unanswered, the combination is already becoming more common.

“We’re in this perfect storm time period where more GLP-1 use is occurring,” said Dan Siskind, MD, PhD, professor of psychiatry at The University of Queensland in Brisbane, Australia, who led the COaST trial and co-authored the meta-analysis.

A global Delphi panel of 90 clozapine experts, published in The Lancet Psychiatry in January with Siskind as first author, identified clozapine-induced gut hypomotility as the leading cause of death among patients taking it. “The number-one killer of people on clozapine is not neutropenia. It’s not myocarditis. It is ileus, bowel rupture, and death,” Siskind said.

When seeing patients at Princess Alexandra Hospital in Brisbane, Australia, Siskind carries a Bristol Stool Chart clipped to his lanyard, the 7-point scale gastroenterologists use to classify stool form. His patients have learned to report their number as soon as they arrive in the office.

“They come into my office and say, ‘Hi, Dan, type 4.’ If there’s but one message, it is that managing people’s gut motility is the task of a clozapine prescriber.”

Experts say that most clinicians who prescribe concomitant clozapine and GLP-1s are not doing this, and at the moment, the research base that would tell them why they should do so barely exists.

A Minor Focus

Most clinical safety guidance for clozapine centers around blood counts due to agranulocytosis, not bowel function. Research on gut motility in this population has been “a minor focus, and very little research done on it,” said Anders Fink-Jensen, DMSci, clinical professor of psychiatry at the University of Copenhagen, Copenhagen, Denmark.

He led a 2025 trial published in JAMA Psychiatry examining the metabolic benefits of semaglutide and clozapine or olanzapine in 73 patients with schizophrenia and diabetes. The study showed constipation rates of 33% vs 19% with semaglutide vs placebo, higher than the rates reported in large general population semaglutide trials.

When asked how constipation was assessed in the trial, Fink-Jensen was straightforward: “You ask them if there’s any issues and they report yes or no, basically,” he said.

But a 2020 diagnostic accuracy study showed that even though 73% of clozapine patients show delayed colonic transit on motility testing, only 26% self-reported constipation, reported lead investigator Susanna Every-Palmer, PhD, forensic psychiatrist and head of the Department of Psychological Medicine at the University of Otago in Wellington, New Zealand.

That result is “worse than a toss of the coin,” said Every-Palmer, who has spent her career studying clozapine’s effects on the gut.

Clozapine’s anti-serotonergic effects dampen the colon’s visceral sensitivity, meaning patients don’t notice the colonic distension that would normally cause discomfort. As a result, Every-Palmer said patients don’t seek help.

A common misperception among clinicians is that clozapine-related constipation is a lifestyle issue, driven by poor diet, inactivity, or inadequate hydration. But Every-Palmer’s lab work has shown otherwise. In a 2017 study, when her group placed functioning rabbit colons in a bath and added clozapine, the colon stopped contracting.

“You didn’t need the mathematical modeling to be able to see it. You could see it in the bath,” she said.

The findings showed that the effect was pharmacologic and dose-dependent, not behavioral, which means this serious issue can’t be managed with more fiber and water.

Siskind’s COaST trial cited Every-Palmer’s diagnostic accuracy study in its methods and used what the paper described as “specific questioning regarding constipation.” The adverse event table still labeled the outcome “constipation (self-report).” COaST reported low rates of constipation and concluded that semaglutide “does not worsen clozapine-associated gastrointestinal hypomotility.”

But the data were still based on self-reports, which Every-Palmer said is a problem.

“I think it’s inadequate just to be asking people about their GI symptoms in this population, based on what we know from former research in this group,” she said.

When asked about that limitation in COaST, Siskind acknowledged the limitation. “I don’t think we can say that COaST showed that there wasn’t constipation,” he said. “We just weren’t able to show there was constipation.”

‘A Major Omission’

From outside psychiatry, there are similar concerns. Michael Horowitz, MBBS, PhD, professor of medicine at the University of Adelaide, Adelaide, Australia, and one of the world’s leading authorities on GLP-1 receptor agonists and gastric emptying, said GLP-1 agonists’ effects on the colon have barely been examined. This is despite a recent study that mapped GLP-1 receptors on inhibitory neurons in the myenteric plexus of the distal colon.

Clozapine silences this same neural network through its anticholinergic effects, echoing what Every-Palmer’s lab work in rabbit colons starkly illustrated.

Data are scarce on whether anyone has studied what happens when a GLP-1 is added to a gut already pharmacologically impaired by anticholinergic agents such as clozapine.

“There isn’t anything meaningful in that area — and it’s a major omission,” Horowitz said.

Large GLP-1 obesity trials have systematically excluded patients with psychiatric disorders and patients with gastroparesis, he noted, meaning the safety data that exist for hundreds of thousands of patients don’t include the population most vulnerable to GI harm.

Many GLP-1 studies also have underutilized validated techniques to measure what the drugs do to gastric emptying, Horowitz said. The gold-standard technique is scintigraphy, which uses a radiolabeled meal to directly image how fast food leaves the stomach. Instead, he said, most studies have relied on paracetamol absorption, a proxy technique that measures how fast an oral dose of acetaminophen reaches the bloodstream. Paracetamol absorption is an indirect proxy; scintigraphy directly images gastric contents.

Horowitz said the few studies that have used scintigraphy did not measure the effect in patients with preexisting gut dysfunction from other medications, including clozapine. “It’s inexcusable that a study is not done, which will provide major insights into clarifying the risk and also the benefits,” Horowitz said.

What Clinicians Should Do Now

An analysis coauthored by Every-Palmer of UK pharmacovigilance data spanning 31 years documented at least 527 cases of harmful clozapine-induced GI hypomotility and 172 deaths in the UK alone. Fatal cases peaked at 10-14 years of clozapine treatment, during which time 25% of reported events were fatal, compared with only two deaths documented in the first year of treatment.

Yet the semaglutide trials only ran for 26-36 weeks, not nearly long enough to capture the fatal events that could occur from this drug combination.

The absence of published serious harm from the combination should not be taken as evidence of safety, Every-Palmer cautioned. “These trials are small. They’re not powered to pick up on the serious or delayed side effects, and they were not looking particularly at transit time,” she said.

What the field needs, she argued, are studies that combine GLP-1s with clozapine and that measure colonic transit time objectively, using radiopaque markers or wireless motility capsules.

Sri Mahavir Agarwal, MD, PhD, psychiatrist who runs both a clozapine clinic and a metabolic clinic at the Centre for Addiction and Mental Health in Toronto, Ontario, Canada, agreed. “Historically, as a field, we have not done very well in managing metabolic health,” he said.

But he urged balance. Patients with severe mental illness “are usually the ones to get access to drugs last,” he said. “Clozapine itself is underprescribed because people are afraid that it might cause side effects, and that fear is as much in the heart of the prescriber as in the heart of the patient.”

Adding a GLP-1 to clozapine requires “a higher level of alert. It can’t be just business as usual if you are combining those two, but you don’t need to panic,” noted Agarwal, who said that he would decline to prescribe a GLP-1 for a clozapine patient who had a history of severe constipation requiring medical intervention, such as bowel evacuation, enemas, or hospitalization.

While consensus guidance exists for co-initiating metformin with clozapine to address weight gain, similar recommendations for the concomitant use of GLP-1s and clozapine are lacking. Agarwal said such a consensus would be useful and that one should eventually be developed. For now, he doesn’t want to scare clinicians or patients away from considering the combo.

“I just want it to be acknowledged as a clinical situation, not overblown,” Agarwal said. “The benefits are likely to be more than the risk as of now, based on the evidence.”

Adverse events from the combination should be reported to pharmacovigilance agencies, Every-Palmer said, because the data needed to quantify the risk will not come from the current trials. “We were using clozapine for about 20 years before that became recognized as a significant life-threatening event,” she said.

To date, there are no published case reports documenting a serious GI event in a patient taking both clozapine and a GLP-1 receptor agonist, and Siskind, Every-Palmer, and Agarwal said they haven’t encountered them in their own clinical practice. However, they still believe such events do occur, even if they have not been formally reported. “That’s because no one’s written it up,” Siskind said.

Every-Palmer treats every clozapine patient with prophylactic laxatives at the outset, the same way clinicians manage opioid-induced constipation. The resulting regimen, called the Porirua Protocol, which she helped develop, starts all clozapine patients on a stool softener (docusate) and a stimulant laxative (senna) from day 1, adding an osmotic laxative (macrogol, known as polyethylene glycol in the US) in resistant cases.

Before the protocol, roughly 8 in every 100 clozapine patients in her service developed serious GI complications each year. After implementation, that fell to about 1 per 100.

Clinicians should watch for overflow diarrhea — liquid stool leaking around a fecal impaction — which Siskind said can masquerade as a side effect of semaglutide but is actually a sign of severe constipation. Fink-Jensen added that psychiatrists who aren’t used to prescribing GLP-1 drugs should consider prescribing them with the help of a general practitioner or endocrinologist rather than managing the combination alone.

In Siskind’s clinic in Brisbane, the system is simple. Donning his laminated card from a lanyard around his neck, he asks the same question to every patient the second they walk through the door. What number are you today?

Siskind reported serving on the Viatris Australian Clozapine Quality Advisory Committee and the Douglas ANZ Clozapine Assurance Committee and receiving speaker and/or consultancy fees from Servier, Otsuka, Viatris Inc., Douglas, Teva, and Lundbeck. Agarwal reported receiving honoraria/speaker fees from HLS Therapeutics, Boehringer Ingelheim Canada, and Bristol Myers Squibb in the past 24 months. Fink-Jensen reported receiving an unrestricted grant from Novo Nordisk A/S in 2019 to investigate the effects of semaglutide on metabolic disturbances in participants with schizophrenia treated with antipsychotics, serving on a randomized controlled trial advisory panel for Novo Nordisk without receiving an honorarium, and serving as a consultant for PASA and Guidepoint. Every-Palmer reported consulting for Douglas Pharmaceuticals. Horowitz reported no financial disclosures.