GLP-1 receptor agonist use is associated with fewer psychiatric hospitalizations and emergency department (ED) visits in adults with schizophrenia and diabetes. 

A large retrospective US cohort study that included more than 70,000 adults with schizophrenia and type 2 diabetes showed that GLP-1s were associated with a 24% lower risk for psychiatric hospitalization and a 29% lower risk for psychiatric ED visits compared to their counterparts who did not use GLP-1s. 

“Our findings suggest that lower psychiatric acute care use might represent an additional benefit when GLP-1 receptor agonists are already clinically indicated for metabolic management in people with schizophrenia and type 2 diabetes,” study investigator Najmeh Davoodian, MD, PhD, Deakin Institute for Mental and Physical Health and Clinical Translation (Deakin IMPACT), Geelong, Australia, told Medscape Medical News. 

However, Davoodian, the study’s lead author, cautioned that the findings alone are “insufficient to change prescribing guidelines or justify prescribing GLP-1 receptor agonists specifically for psychiatric benefit. Replication in other large studies and confirmation in randomized controlled trials are needed before psychiatric benefits can become an established consideration in treatment selection.”

The study was published online October 14 in Lancet Psychiatry.

Fewer Hospitalizations, ED Visits

Patients with schizophrenia have a high burden of cardiometabolic disease, partly driven by antipsychotic-associated weight gain, insulin resistance, and other adverse metabolic effects. These complications can undermine treatment adherence and contribute indirectly to psychiatric relapse and increased healthcare utilization.

GLP-1s have demonstrated benefits for weight management and metabolic health in patients with schizophrenia receiving antipsychotics, without clear evidence of psychiatric destabilization. However, whether they are associated with improved psychiatric outcomes was unclear. 

To investigate, the researchers analyzed electronic health records for 70,257 adults (mean age, 53 years; 48.2% women) with schizophrenia and comorbid type 2 diabetes. Of these patients, 9093 (12.9%) initiated GLP-1s, 4968 (7.1%) initiated DPP-4 inhibitors, 5328 (7.6%) initiated SGLT2 inhibitors, and 50,868 (72.4%) did not initiate treatment with any of these drug classes.

Patients were followed for up to 2 years. The study’s primary outcomes were psychiatric hospitalization and psychiatric ED visits. The researchers used inverse probability weighting to balance demographic, clinical, psychiatric, and treatment-related factors across groups.

In weighted regression analyses, initiation of a GLP-1 receptor agonist was associated with a 24% lower risk for psychiatric hospitalization compared with non-initiation (hazard ratio [HR], 0.76). The risk was also significantly lower compared with initiation of DPP-4 inhibitors (HR, 0.82) and SGLT2 inhibitors (HR, 0.83).

Similar findings emerged for psychiatric ED presentations. GLP-1 receptor agonist initiation was associated with a 29% lower risk compared with non-initiation (HR, 0.71), a 13% lower risk compared with DPP-4 inhibitor initiation (HR, 0.87), and a 20% lower risk compared with SGLT2 inhibitor initiation (HR, 0.80).

The associations were generally strongest among adults younger than 55 years, men, and those receiving antipsychotic medications. Among patients receiving antipsychotics, the estimated number needed to treat (NNT) over 1 year was 32 to prevent one psychiatric hospitalization or ED visit, compared with compared with non-initiation.

A Foundation for Further Research

Exploratory weight-change analyses suggested that weight loss might not fully explain the association. Although the reduction in psychiatric acute care use was greatest among patients who lost at least 5% of their body weight, an association was also observed among those whose weight remained stable in some comparisons.

Analyses of individual agents showed associations for semaglutide, liraglutide, and tirzepatide, but not dulaglutide, although further research is needed to confirm this finding.

As an observational analysis, the study cannot establish causality, and residual confounding and selection bias are possible despite statistical adjustment, the researchers noted. 

In addition, medication initiation was identified from prescription records, but treatment adherence, sustained use, and subsequent medication changes could not be reliably assessed. Differences in access to healthcare or patterns of service use could also have influenced the findings. 

Davoodian noted that the study was not designed to establish how GLP-1s might influence psychiatric outcomes. 

“Direct effects on brain pathways and indirect benefits through improved metabolic health could both contribute and may interact,” she said. “It provides a foundation for future mechanistic studies to clarify whether, and through which pathways, GLP-1 receptor agonists might confer psychiatric benefits.”

Promising, Not Proven

The possibility that medications used to improve metabolic health might also be associated with better psychiatric outcomes “deserves serious attention,” Jon van Niekerk, MBChB, chair of the faculty of general adult psychiatry at the Royal College of Psychiatrists, said in a statement from the UK Science Media Centre. 

Van Niekerk, who was not involved in the research, noted that the large sample size and the comparisons with other diabetes medications were strengths of the study and that the reductions in psychiatric hospitalizations and emergency mental health care were clinically meaningful. 

However, he cautioned that the study cannot show GLP-1s directly improve psychiatric symptoms or prevent relapse, and he questioned whether the researchers adequately accounted for differences in illness severity, antipsychotic prescribing, and engagement with healthcare. It would be premature, he said, to change psychiatric prescribing practice on the basis of these findings alone.

Another expert sees the work as part of a broader trend. Roger McIntyre, MD, professor of psychiatry and pharmacology at the University of Toronto, who also was not involved in the research, told Medscape Medical News that the findings replicate earlier research linking GLP-1s to fewer psychiatric hospitalizations, including in those with schizophrenia and substance use disorders. 

He said the work reinforces emerging evidence that GLP-1s may have benefits beyond metabolic health in people with serious mental illness. 

Sameer Jauhar, MBChB, PhD, of Imperial College London, cautioned that patients prescribed GLP-1s in the US tend to be more engaged with healthcare and have better insurance than other patients, factors that can independently reduce psychiatric hospitalization.

In comments distributed by the Science Media Centre, Jauhar noted that the survival curves in the paper diverge almost immediately after treatment starts and then stay parallel throughout follow-up. That pattern, he suggested, is more consistent with preexisting differences between the groups than with a drug effect that emerges gradually over time.

Still, Jauhar said the findings at least indicate that GLP-1s do not appear to worsen psychiatric outcomes in this population, which he said is a worthwhile result in itself. Jauhar was not associated with the current study.

McIntyre cautioned that clinicians should watch for overlapping gastrointestinal side effects when combining GLP-1s with antipsychotics such as clozapine and olanzapine, or with the anticholinergic benztropine. 

Each can cause constipation, and GLP-1s slow gastrointestinal motility. McIntyre said the combination is not off-limits, but patients should be encouraged to stay hydrated, eat enough fiber, and exercise, and should be monitored for GI problems.

Sri Mahavir Agarwal, MBBS, MD, PhD, a psychiatrist who runs clozapine and metabolic clinics at the Centre for Addiction and Mental Health in Toronto, agreed that the overlap of these medications warrants attention.

“The risk is real; they are both likely to cause constipation,” he told Medscape Medical News. He recommended proactive management with laxatives and nonpharmacologic measures, along with closer monitoring.

However, Agarwal said the risk for GI side effects in concert with these psychiatric medications shouldn’t stop clinicians from offering GLP-1s when they are indicated. 

The study was supported by the Synergy for the Development of Innovative Clinical Treatments in Schizophrenia project, funded by the Australian National Health and Medical Research Council. The authors reported no relevant financial relationships. 

Agarwal reported receiving honoraria/speaker fees from HLS Therapeutics, Boehringer Ingelheim Canada, and Bristol Myers Squibb. McIntyre has received speaker/consultation fees from Lundbeck, Janssen, Alkermes, Neumora Therapeutics, Boehringer Ingelheim, Sage, Biogen, Mitsubishi Tanabe Pharma, Purdue, Pfizer, Otsuka, Takeda, and Neurocrine. Jauhar has received personal fees from Recordati, LB Pharmaceuticals, Boehringer Ingelheim, Accord Pharmaceuticals, Wellcome Trust, Lundbeck, Janssen, and Sunovion. Van Niekerk reported no relevant financial relationships.