Two weight-friendlier antipsychotics, tested head to head, came out nearly even
- Adarsh Verma and colleagues at the All India Institute of Medical Sciences, Patna, ran a 12-week randomized trial in 98 adults with schizophrenia who had already been switched off olanzapine because of the weight it caused, assigning them 1:1 to cariprazine or aripiprazole.
- Weight, BMI, waist-to-hip ratio, lipid profile, blood glucose, HbA1c, PANSS and SOFAS were measured at baseline, week 6 and week 12.
- Weight fell over the 12 weeks in both arms (treatment × time interaction, p<.001), but which of the two drugs a patient received made no significant difference to that trajectory (main treatment effect, p=.249). Cariprazine outpaced aripiprazole by a mean of 0.82 kg over the 12 weeks.
- PANSS and SOFAS scores improved significantly within both arms over the 12 weeks, with no significant difference between the two drugs – the switch cost neither group ground on symptoms or on functioning.
Ninety-eight adults with schizophrenia at a tertiary hospital in Patna, India, had one thing in common before this trial began: their psychiatrists had already pulled them off olanzapine because of the weight it put on. Adarsh Verma and coauthors at the All India Institute of Medical Sciences randomized the group 1:1 to two drugs marketed as gentler on the metabolism, cariprazine and aripiprazole, and followed them for 12 weeks. Weight, BMI, waist-to-hip ratio, lipids, glucose, HbA1c, PANSS and SOFAS were checked at baseline, week 6 and week 12.
Both drugs did what they were switched in to do. Weight dropped over the 12 weeks in both arms, the interaction between treatment and time reaching significance (p<.001), and PANSS and SOFAS scores improved in both arms too, with no significant gap between drugs on either measure. What the trial could not show was a meaningful edge for one drug over the other: the main effect of treatment on weight fell short of significance (p=.249).
A narrow, real difference
Cariprazine did pull ahead – 0.82 kg more weight lost than aripiprazole over 12 weeks. The authors call the advantage modest and say any metabolic edge should be read cautiously, and a single-center, 12-week trial with roughly 49 patients per arm answers the question it is built to answer: both drugs work, by about the same amount. It was not built to answer the harder one, which is whether that difference holds up in a larger sample followed for longer than three months. Waist-to-hip ratio, lipids and glycemic markers moved the same direction in both arms, with no drug pulling clearly ahead on any of them. Nothing here argues for reaching past aripiprazole for cariprazine on cardiometabolic grounds alone, and nothing argues against it either.
Who notices first
The psychiatrist writes the prescription. The person who hears about the extra fifteen minutes on the treadmill, or watches a patient wave off a snack she used to eat without thinking, is often the psychologist or counselor seeing that patient every week, long before the next scheduled psychiatric review. This trial measured weight at week 6 and week 12; a treating clinician sees it move before either checkpoint arrives, in the ordinary detail of a session rather than on a chart. Should a specialist who watches the scale shift raise a switch with the prescriber before the numbers cross into metabolic syndrome, instead of waiting for that review to catch up?
Nothing in this trial says either drug makes that conversation unnecessary. Both groups needed the full 12 weeks to show the improvement they eventually showed, and both needed it because an earlier drug, olanzapine, had already done the damage the switch was meant to undo. A switch is not a substitute for catching the pattern early. It is what happens after someone already has.
Switched off olanzapine for weight gain, patients on cariprazine lost 0.82 kg more than those on aripiprazole over 12 weeks – a real difference, not a clear winner.
Single-center trial, 98 patients randomized roughly 49 per arm, analyzed per-protocol rather than intention-to-treat. No comparison arm continued on olanzapine or left unmedicated, so the trial cannot say how much of the improvement reflects the switch itself versus regression from an unusually high starting weight. Follow-up stopped at 12 weeks; nothing is known about durability beyond that point. The published time-effect statistic in the source abstract carries an F-value far outside the plausible range for this design; it could not be verified against the full text and is not used here.