Semaglutide in schizophrenia: HISTORI trial

People with schizophrenia die, on average, 15 years earlier than the general population^[1]. The main driver is cardiometabolic disease, not suicide or accidents. Second-generation antipsychotics that control psychosis also tend to drive weight gain, raise blood glucose, and push cholesterol in the wrong direction. For decades, clinicians had few good options for those secondary harms. A cluster of randomised clinical trials published between mid-2025 and early 2026 has shifted that picture. On 1 July 2026 JAMA Psychiatry editor Dost Ongur called the evidence a 'top clinical trial' moment for the field^[1].

This article explains what the trials actually found, what they did not find, and what the results mean for the broader question of semaglutide use in psychiatric settings.

Why cardiometabolic risk is so high in schizophrenia

Second-generation antipsychotics (SGAs), which include clozapine, olanzapine, quetiapine, and risperidone, are highly effective for psychosis. They are also among the medicines most associated with weight gain, insulin resistance, and dyslipidaemia. A patient who starts an SGA in their twenties and stays on it for decades faces a compound problem: the baseline cardiometabolic vulnerability associated with schizophrenia itself, layered on medication-driven metabolic changes that accumulate year after year.

Lifestyle interventions have shown limited and inconsistent results in this population. The condition itself, and the sedating profile of many antipsychotics, makes sustained exercise and dietary change harder to maintain than in the general obesity population.

The HISTORI trial: 154 patients, 30 weeks

The HISTORI trial is the anchor study in this cluster^[2]. Ganeshalingam, Frystyk, and colleagues recruited 154 adults with schizophrenia spectrum disorder who were on SGAs, had prediabetes (elevated HbA1c), and were overweight or obese. The trial ran from January 2022 to May 2024 across community mental health services in Southern Denmark and the Region of Zealand. Participants were randomised 1:1 to once-weekly subcutaneous semaglutide or matched placebo for 30 weeks. The completion rate was 91.5% (141 of 154), which is high for a psychiatric population.

The primary endpoint was change in HbA1c. Semaglutide reduced HbA1c by 0.46 percentage points compared with placebo (95% CI, -0.56 to -0.36)^[2]. On the categorical threshold, 81% of the semaglutide group versus 19% of the placebo group achieved HbA1c below 5.7% by week 30. Weight fell by 9.21 kg more in the semaglutide arm. Lipid profiles and triglycerides improved. Physical quality of life, measured on the SF-36v2, improved by 3.75 points, a margin exceeding the minimally important difference for that instrument.

Psychiatric symptoms did not worsen. The PANSS-6 scale (a six-item subset of the Positive and Negative Syndrome Scale) showed no difference between groups^[2]. Psychiatric adverse events were comparable. Gastrointestinal adverse events were more common in the semaglutide arm, consistent with the drug's known profile.

A secondary analysis of the HISTORI data, published separately in May 2026, examined quality of life in more detail^[3]. The physical component of quality of life improved, with roughly half of that benefit mediated through weight loss itself. The mental health component of quality of life did not change meaningfully at 30 weeks. The investigators noted that longer studies would be needed to determine whether mental health quality of life effects appear over time.

The Sass trial: intervening earlier

A second Danish trial, published in JAMA Psychiatry in February 2026, asked a different question^[4]. Sass, Klausen, Schwarz, and colleagues recruited 73 participants with schizophrenia spectrum disorders on clozapine or olanzapine started within the previous five years, and with early glycaemic abnormalities (HbA1c between 5.4% and 7.4%). The aim was to intervene before full prediabetes or type 2 diabetes was established.

After 26 weeks of once-weekly semaglutide 1 mg versus placebo, HbA1c fell by 0.25 percentage points more in the semaglutide group (95% CI, -0.33 to -0.16; p less than 0.001)^[4]. Weight fell by 9.2 kg. Waist circumference dropped 7.0 cm. Fat mass fell 6.1 kg. Psychiatric symptom scores were again comparable between groups.

Forty-three percent of semaglutide participants reached HbA1c levels in the low-risk range (below 5.4%), versus 3% on placebo. That is a reversal of early metabolic drift, not merely a slowing of it.

The COaST trial: clozapine-specific data

Clozapine is the most effective antipsychotic for treatment-resistant schizophrenia. It is also the most metabolically damaging. The COaST trial, published in Lancet Psychiatry in July 2025, ran at six Australian sites and enrolled 31 adults with schizophrenia or schizoaffective disorder on clozapine with obesity^[5]. The trial was stopped early when the investigational product became unavailable, so it enrolled 31 of the planned 80 participants.

Even in that smaller sample, the weight difference was notable: 13.88% body weight reduction in the semaglutide group versus 0.42% with placebo at 36 weeks (p less than 0.0001)^[5]. Clozapine blood levels did not change, which matters clinically because clozapine has a narrow therapeutic index and drug interactions are a real concern. No serious adverse events were attributed to the study treatment.

What the Ongur editorial adds

Ongur's commentary in JAMA Psychiatry^[1] frames the HISTORI trial as among the more clinically consequential psychiatric trials of recent years. His argument is structural: if the 15-year life expectancy gap in schizophrenia is driven primarily by cardiometabolic disease, and if SGAs compound that risk in the absence of effective countermeasures, then a drug that safely addresses the metabolic consequences without destabilising the underlying psychiatric condition is not purely metabolic medicine. It functions as a psychiatric intervention too.

The psychiatric system has historically treated metabolic monitoring in schizophrenia as secondary to symptom control. A high-quality trial body showing a meaningful cardiometabolic effect without psychiatric risk gives clinicians a grounded rationale to act on what was previously just a monitoring concern.

What is not yet known

The trials run 26 to 36 weeks. The cardiometabolic risk in schizophrenia accumulates over decades. Whether semaglutide's metabolic benefits persist over years, whether the HbA1c effect translates into reduced cardiovascular events, and whether the results replicate across more diverse populations than the Scandinavian and Australian samples are all open questions.

The mental quality of life question is also unresolved. Physical quality of life improved in HISTORI. Mental quality of life did not, at 30 weeks. Whether longer treatment changes that is not yet known.

There is also a regulatory and access gap. Semaglutide is licensed for type-2 diabetes and weight management. It is not licensed for antipsychotic-related metabolic disease as a standalone indication. Prescribing it in the schizophrenia context sits within clinical discretion in most jurisdictions, subject to the usual prescriber judgment, patient consent, and monitoring. It is a prescription medicine wherever PeptideMethods covers it.

This article is for educational and journalistic purposes only and does not constitute medical advice. Always consult a qualified healthcare professional before making any decision about medicines. PeptideMethods.com does not sell, distribute, or facilitate the sale of any medicine or peptide product.