Cardiometabolic illness has quietly become the main killer in schizophrenia.

People living with schizophrenia spectrum disorders lose around 10–15 years of life expectancy, mostly due to cardiovascular disease, not suicide or violence. Obesity, insulin resistance, prediabetes and type 2 diabetes sit right at the centre of this excess mortality.

And some of our most effective antipsychotics – clozapine and olanzapine – are also the worst offenders metabolically.

We’ve all seen this in clinic:

• 10–20 kg weight gain in a couple of years

• HbA1c creeping up into the prediabetes range

• A young patient on clozapine whose lipid profile now looks like that of a 60-year-old with metabolic syndrome

Lifestyle advice, metformin and the occasional topiramate prescription help, but the effect sizes are modest and often hard to sustain in real-world practice.

A new randomized clinical trial in JAMA Psychiatry (online, December 2025) takes this problem head-on and asks a very specific question:

If we start semaglutide early in patients on clozapine/olanzapine who are just beginning to develop glycemic abnormalities, can we actually bend the metabolic curve?

The answer from this study is a cautious but solid yes.

The basic idea: Treat early, not after the damage is done

Instead of waiting for full-blown diabetes, the trial specifically targeted early-stage dysglycaemia in people who had:

• Schizophrenia spectrum disorders

• Started clozapine or olanzapine within the last 5 years

• HbA1c in the “high-risk” zone (roughly 5.4–7.4%)

• No current antidiabetic medication

Seventy-three adults (mean age 35, mean BMI ~36 kg/m²) from three Danish sites were randomized to:

• Semaglutide 1 mg once weekly (after slow titration from 0.25 → 0.5 → 1 mg), or

• Placebo once weekly

Both were added on top of their existing clozapine or olanzapine regimen, which was not switched or tapered.

The treatment duration was 26 weeks.

The primary outcome: change in HbA1c.
Key secondary outcomes: weight, waist circumference, fat mass, cardiometabolic markers, psychiatric symptoms and safety.

What actually happened?

1. Glycemic control: HbA1c moved in the right direction

By week 26:

• Semaglutide produced a clinically meaningful drop in HbA1c, while placebo showed essentially no change.

• The average difference between groups was about –0.25 percentage points.

• The more striking metric:

  • 43% of patients on semaglutide achieved “low-risk” HbA1c (<5.4%)

  • Only 3% of patients on placebo did so

This pattern held true in subgroup analyses for both clozapine-treated and olanzapine-treated participants.

It’s not the dramatic 1–1.5% HbA1c reduction we see in uncontrolled type 2 diabetes, but remember: this was early dysglycaemia. That small shift, occurring early, could translate into a big difference in long-term cardiovascular risk.

2. Weight and body composition: Big changes in a short time

As we’d expect with a GLP-1 receptor agonist, semaglutide also hit weight hard – in a good way:

Compared with placebo over 26 weeks, semaglutide led to:

• About 9.2 kg more weight loss

• Around 7 cm greater reduction in waist circumference

• A drop of 3 BMI units on average

• Roughly 6 kg reduction in total fat mass

Visceral fat also decreased numerically, though that particular measure didn’t reach statistical significance – likely due to sample size and variability.

For a young person on clozapine or olanzapine with BMI in the mid-30s, these numbers are not cosmetic. They are the difference between walking steadily toward diabetes and heart disease versus stepping back from the edge.

3. Cardiometabolic markers: Mixed but reassuring

On other metabolic parameters:

• Lipids (total cholesterol, LDL, HDL, triglycerides): no significant between-group differences

• Blood pressure: both groups had small reductions; no clear advantage for semaglutide

• Liver markers and FIB-4: small numerical changes but remained within normal limits

This is not very surprising given:

• The relatively short duration (26 weeks)

• The dose (1 mg weekly, lower than the 2.4 mg obesity dose)

• The young, early-risk profile of the cohort

The strong signal remains: glycemia + weight + fat mass.

4. Psychiatric symptoms and functioning: No destabilization

Understandably, many psychiatrists worry:

“Will adding semaglutide destabilize psychosis or worsen mood, especially in a fragile clozapine patient?”

In this trial:

• Psychotic symptom severity (PANSS-6) improved slightly in both groups with no significant difference between semaglutide and placebo.

• Clinical Global Impression, quality of life and psychosocial functioning did not diverge between groups.

• Rates of psychiatric hospitalisation, suicide attempts and self-harm were similar across arms.

At least over 6 months, there was no signal that semaglutide worsens psychiatric stability in this population.

5. Nicotine dependence: An intriguing bonus

Among participants who smoked:

• Semaglutide was associated with a reduction in nicotine dependence scores compared with placebo.

This fits with a growing preclinical and clinical literature suggesting GLP-1 receptor agonists may modulate reward circuitry and reduce drive for nicotine, alcohol and other substances in some individuals.

The current study excluded people with active substance use disorders, so we can’t extrapolate too far, but this is a tantalising bonus:

A drug that improves metabolic health and may make quitting smoking a little easier in a group where smoking prevalence is extremely high.

Safety: What should clinicians watch for?

The safety profile was broadly consistent with what we already know about semaglutide.

Common adverse effects:

• Gastrointestinal issues:

  • Nausea

  • Vomiting

  • Constipation

  • Occasional diarrhoea, reflux, abdominal discomfort

More participants in the semaglutide arm experienced GI side effects, and four discontinued treatment because of adverse effects (none discontinued for this reason in the placebo arm).

The constipation signal deserves particular attention in clozapine-treated patients, who are already at risk of severe gastrointestinal hypomotility and ileus. Combining two agents that can slow gut transit means:

• You must monitor bowel habits closely

• Proactive bowel regimens (hydration, fibre, stool softeners/laxatives when needed) become non-negotiable

One patient on semaglutide died suddenly during the trial; autopsy did not attribute this to the drug, and overall serious adverse events were similar between arms. But as always, long-term, large-scale safety data in psychiatric populations will be important.

Overall:

• No new major safety signals

• No increase in psychiatric adverse events

• GI side effects manageable but require vigilance, especially with clozapine

How might this change day-to-day practice?

We’re still early, but several practical themes emerge:

1. Move from “damage control” to early protection

Instead of waiting until HbA1c crosses into frank diabetes or weight crosses 120–130 kg, this trial supports:

• Screening regularly (weight, waist, HbA1c) in clozapine/olanzapine users

• Considering semaglutide early, once HbA1c drifts above ~5.4–5.5% and weight is clearly in the overweight/obese range

This is preventive cardiometabolic medicine inside psychiatry, not post-hoc firefighting.

2. Accept that for many patients, clozapine stays

We all know patients for whom:

• Clozapine is the only drug that keeps psychosis controlled

• Olanzapine at modest doses is the only antipsychotic they tolerate

Switching to “metabolically cleaner” antipsychotics is often not clinically viable.

Semaglutide offers a way to keep the antipsychotic that works, while taking the metabolic consequences seriously and actively countering them.

3. Choose patients thoughtfully

Those most likely to benefit:

• On clozapine or olanzapine

• BMI clearly in the overweight/obese range

• HbA1c already in the high-risk band but not yet severely diabetic

• Family history of diabetes and CVD, or rapid weight gain after starting SGA

At the same time, you’d want to be extra careful in those with:

• Severe GI disorders

• History of pancreatitis

• Significant cardiovascular instability

4. Keep lifestyle and basic physical health care in the picture

Semaglutide is not a replacement for:

• Smoking cessation support

• Dietary counselling

• Physical activity interventions

• Blood pressure and lipid control

• Regular monitoring for clozapine-related side effects

The ideal model is integrated care: psychiatry + endocrinology/diabetology + primary care, with semaglutide as a powerful added tool, not a magic bullet.

5. Confront the access and cost problem

The most uncomfortable but important issue for real-world practice – especially in low- and middle-income settings – is cost.

GLP-1 receptor agonists remain expensive. Patients with severe mental illness already face multiple barriers:

• Lower income

• Patchy insurance

• Fragmented care

• Stigma at multiple levels

From a public health perspective, there is a strong case for preferential coverage of GLP-1 RAs in high-risk psychiatric populations on clozapine and high-dose olanzapine. But getting there will require:

• Advocacy

• Health-economics data

• Policy-level recognition that “psychiatric” patients are actually cardiometabolic high-risk patients

Limitations of the current evidence

We shouldn’t over-sell what one trial can do. Important limitations include:

• Sample size (73 randomized; 57 completers) – underpowered for some secondary outcomes

• Short duration (26 weeks) – long-term durability and cardiovascular outcomes are unknown

• Dose (1 mg weekly) – lower than the 2.4 mg obesity dose; higher doses might bring extra benefit and extra side effects

• A relatively young, mostly White, low-comorbidity cohort, which may not fully reflect complex real-world populations

• Potential unblinding due to GI side effects, which could influence subjective measures

So this is a strong proof-of-concept, not the final word.

The big picture: Protecting brain and body together

Clozapine and olanzapine remain among our best tools for severe psychosis. The price has long been paid in kilograms, centimeters of waistline, and points of HbA1c.

This trial suggests a different future:

• Continue using powerful antipsychotics when needed

• Pair them with equally serious metabolic protection early

• Monitor closely, treat side effects proactively, and keep psychiatric stability intact

In other words: stop choosing between the brain and the heart.

For now, semaglutide looks like one of the most promising candidates to help that happen in schizophrenia spectrum disorders. Larger, longer and more diverse trials will tell us how far this approach can go – but the direction of travel is clear.

Dr. Srinivas Rajkumar T, MD (AIIMS), DNB, MBA (BITS Pilani)
Consultant Psychiatrist & Neurofeedback Specialist
Mind & Memory Clinic, Apollo Clinic Velachery (Opp. Phoenix Mall)
✉ srinivasaiims@gmail.com 📞 +91-8595155808