Scientific deep-dive

Antipsychotics and Metabolic Harm

The drugs that work best for schizophrenia are the hardest on metabolism, and switching is often not an option. A Danish trial treated the consequence instead, and measured psychiatric symptoms as an outcome.

By Nora Bissett · Pricing Editor
Editorially reviewed. Not medical advice · How we verify contentLast reviewed
7 min read·2 citations

We covered a first attempt at this question in COaST, which ended at 31 participants. A second Danish trial has now answered it at more than twice the size, in people taking clozapine or olanzapine whose blood sugar had begun to drift. It reports the result this group needs most: metabolic improvement with no worsening of psychiatric symptoms.[1]

What this trial adds

Three things separate it from the earlier work. It is larger: 73 randomized against 31. It includes olanzapine alongside clozapine, widening the question beyond the single hardest case. And it selected for people whose blood sugar had already started to move but who were not yet on any diabetes treatment, which makes it a trial about prevention rather than about rescue.

That last point sets up the whole reading of the numbers below. The standard advice, to switch to something metabolically kinder, is unavailable to many of the people who most need it, because these drugs are prescribed when others have failed. Treating the metabolic consequence rather than removing the cause is not a compromise here. It is frequently the only route open, and not one that is reliably offered: across 49 studies, people with diabetes and a mental health diagnosis had a quarter of the odds of being prescribed a GLP-1.

What was done

Across three Danish sites between 2021 and 2024, adults aged 18 to 65 with a schizophrenia spectrum disorder, taking clozapine or olanzapine started within the previous five years, with an HbA1c between 5.4% and 7.4% and no diabetes treatment, were randomized to weekly semaglutide 1 mg or matching placebo for 26 weeks alongside their existing therapy. Of 104 screened, 73 were randomized and 57 (78%) finished. Average age was 35, 65% were women, and average BMI was 36.1.

Note the dose: 1 mg. That is a semaglutide dose used for type 2 diabetes, not the 2.4 mg used for weight management. Anyone setting the weight result below against figures from the obesity trials is comparing different doses as well as different populations.

What changed

Semaglutide against placebo at 26 weeks. Differences between groups, with 95% confidence intervals.[1]
OutcomeDifference95% CIP
HbA1c−0.25%−0.33 to −0.16<.001
Body weight−9.2 kg−13.3 to −5.1<.001
Waist circumference−7.0 cm−10.6 to −3.3<.001
Fat mass−6.1 kg−10.2 to −1.9.006

A quarter of a percentage point of HbA1c reads as unimpressive until you look at who was enrolled. Entry required an HbA1c between 5.4% and 7.4% with no diabetes medication, so this was a mostly prediabetic group with very little room to fall. In that setting the proportion crossing into low-risk territory is the better measure, and 43% of the semaglutide group (12 of 28) reached an HbA1c below 5.4%, against 3% on placebo.

A small mean difference in a population already near normal is not a small effect.

Lipids, liver function and blood pressure did not differ between groups. That is a reminder that 26 weeks is short for outcomes that move slowly, and that weight loss does not automatically improve everything measurable.

A third trial, and the mechanism question it answers

HISTORI is the largest of the three. It randomized 154 people with schizophrenia, prediabetes and excess weight, all taking second-generation antipsychotics, to semaglutide 1.0 mg or placebo for 30 weeks, and 91.5% completed.[2]

HISTORI at 30 weeks, against placebo.[2]
MeasureEffectP
Fasting glucose−0.87 mmol/L (95% CI −1.15 to −0.59)< 0.001
Insulin sensitivity+8.60 (5.82 to 13.65)0.001
Insulin resistance−0.69 (−1.00 to −0.20)0.006
Weight−9.2 kg—
β-cell function+8.10 — a modest increase0.19

The 9.2 kg weight loss matches the Danish trial above almost exactly, in a separate population, which is the kind of agreement that makes a result believable.

The last row answers a question the other trials could not: the drug is not repairing what the antipsychotics damaged. β-cell function did not significantly improve, and a mediation analysis attributed the gains in insulin sensitivity and insulin resistance to the weight loss itself.

That distinction has a practical edge. If the improvement depends on the weight loss, it depends on the weight loss persisting. Everything this register documents about what happens when these drugs stop applies here with particular force, in people who will remain on the antipsychotics that caused the problem.

The psychiatric result

For someone stable on clozapine, the question is not whether a drug will reduce their weight. It is whether adding it will destabilize an illness that took years to control. The trial found no difference in psychiatric symptoms between groups and no difference in psychiatric adverse events. Gastrointestinal effects were common, mild and temporary, as they are everywhere else in this class.

Read that reassurance at the size of the trial. Seventy-three people over 26 weeks can show that adding semaglutide did not commonly worsen psychiatric symptoms. It cannot rule out a rare psychiatric event, and it does not speak to what happens after six months. “No difference observed” in a trial this size is a statement about common effects.

Within that limit it is useful, because the alternative evidence is worse. Much of the public conversation about these drugs and mental health runs on registry signals and anecdote. A randomized trial conducted specifically in people with serious mental illness, measuring psychiatric symptoms as an outcome, is a better instrument than either, even at this size.

What this does not say

  • It is not an argument to change psychiatric medication. Clozapine and olanzapine are prescribed when other drugs have failed. Nothing here bears on that decision, which belongs entirely to a psychiatrist.
  • It is 26 weeks. Antipsychotic-associated metabolic risk accumulates over decades.
  • Everyone had early glycemic changes and no diabetes treatment. This is prevention in a group heading toward diabetes, not treatment of established disease.
  • It reports no cardiovascular outcomes. The mortality gap that motivates the whole question is not something a 73-person, six-month trial can address.

For what these drugs do to mood and motivation in other psychiatric settings, see semaglutide and motivation.

Frequently Asked Questions

In this randomized trial, adding weekly semaglutide 1 mg alongside those drugs for 26 weeks improved blood sugar, weight, waist and fat mass, with no difference in psychiatric symptoms or psychiatric adverse events. It remains a decision for the treating psychiatrist.
No difference was observed between the semaglutide and placebo groups in either psychiatric symptoms or psychiatric adverse events. With 73 participants over 26 weeks, that is meaningful evidence about common effects and cannot exclude rare ones.
Entry required an HbA1c between 5.4% and 7.4% with no diabetes medication, so most participants started near normal and had little room to fall. The more informative figure is that 43% on semaglutide reached an HbA1c below 5.4% against 3% on placebo.
9.2 kg more than placebo over 26 weeks, along with 7.0 cm of waist circumference and 6.1 kg of fat mass. That was on the 1 mg dose used for diabetes rather than the 2.4 mg used for weight management.
That is a question for a psychiatrist, and for many people it is not available: clozapine and olanzapine are typically prescribed precisely because other drugs did not work. This trial addresses treating the metabolic consequence, not replacing the cause.

References

  1. 1.Sass MR, Klausen MK, Schwarz CR, et al. Semaglutide and Early-Stage Metabolic Abnormalities in Individuals With Schizophrenia Spectrum Disorders: A Randomized Clinical Trial JAMA Psychiatry. 2026. PMID: 41335431.
  2. 2.Ganeshalingam AA, Uhrenholt N, Arnfred S, et al. Semaglutide Effects on Insulin Sensitivity and β-Cell Function in Patients With Schizophrenia, Prediabetes, and Obesity Treated With Second-Generation Antipsychotics: Findings From the HISTORI Trial Diabetes Care. 2026. PMID: 41778920.

A Quarter the Odds of Being Offered It

People with diabetes who also have a mental health diagnosis received worse care on almost every measure, and a quarter of the odds of being prescribed a GLP-1, while having half again the odds of being put on insulin.

7 min read

The Mental Health Score That Did Not Move

In the head-to-head trial, physical quality-of-life scores improved clearly over 72 weeks. The Mental Component Summary did not significantly improve in either arm, a result the paper's own conclusion omits.

6 min read

Real, and Below the Threshold

Semaglutide improves patient-reported physical functioning by 1.71 points, with an interval excluding zero. The threshold at which a patient notices a change is 3.7 points. Both are true at once.

6 min read

Eight Things Patients Said

Trials measure weight and adverse events. They do not measure being judged for taking a drug. Researchers interviewed 30 people across 15 states and sorted what they said into eight themes.

6 min read

Judged Worse Than Not Losing Weight

Researchers gave 607 people identical descriptions of a stranger, varying only how they had lost weight. The version who used a GLP-1 was judged more harshly than the version who had not lost weight at all.

6 min read

Does Ozempic or Mounjaro Cause Anxiety? What the Trials and Labels Show

In randomized trials, anxiety was no more common on semaglutide or tirzepatide than on placebo. Where cohort studies and adverse-event reports disagree, why low blood sugar and a faster pulse can feel like anxiety, and what is known about panic attacks.

12 min read

Where to get GLP-1 online, safely: sellers our editors have checked

These are telehealth sellers our editors have checked. For each one we hold a price, the form the drug comes in, and the states it reaches.

No insurance needed · vetted by our editors

Some of the links on this page earn us money. If you sign up with a provider after following one, that provider may pay GLP Watchdog a commission. Learn more

9.3

Embody

A flat compounded tirzepatide or semaglutide price at every dose, from a seller that names its pharmacies

8.7

Eden

Coverage in all 50 states, which most sellers will not confirm

7.9

Luvo Health

Moving between compounded and brand without changing seller

About this analysis

GLP Watchdog publishes independently. Each seller in our dataset is checked against that seller's own website, and the dataset is updated continuously. Read our full methodology →