did aripiprazole have the lowest sedation rate, 4%, of seven antipsychotics in the sino trial?
“aeropipersol only had four percent sedation. The lowest of all anti-psychotics, which is truly a novel finding that had not been reported previously using a good measure of sedation.”
a guest, a psychiatrist (co-director of the MGH Psychosis Clinical and Research program) said on Psychopharmacology and Psychiatry Updates, 2026-10-09, at 07:58. hear it in context → the speaker cited: the SINO trial.
the episode: Schizophrenia: Which of Seven Antipsychotics Wins? (at 07:58). we link the show's own page; we do not re-host audio.
first, the part that holds
the main point holds: in the SINO trial, which randomized 3,067 inpatients to one of seven antipsychotics, the published abstract says aripiprazole was the least sedating, with relative risks of 0.30 to 0.39 against each of the other six drugs. what we could not check is the specific 4% figure, because it is not in the abstract and the full paper was closed to us.
what we found when we went and looked
the SINO trial (American Journal of Psychiatry, 2026) randomized 3,067 acutely ill inpatients with schizophrenia across 32 hospitals in china to six weeks of olanzapine, risperidone, quetiapine, aripiprazole, ziprasidone, perphenazine or haloperidol. the abstract we read states "aripiprazole was least sedating (relative risks: 0.30-0.39)", which confirms the ranking. the abstract reports relative risks only, not the share of patients with sedation, and the journal's full text returned an access error, so we could not read the 4% figure in the primary record. that figure appears only in secondary summaries we are not allowed to cite. if 4% is correct, those relative risks imply sedation in roughly 10-13% of patients on the other six drugs, which is internally consistent, but it is our arithmetic, not a published number.
assertion by assertion
| assertion | verdict | what the record says |
|---|---|---|
| the SINO trial compared seven antipsychotics head to head, aripiprazole among them | confirmed | the abstract describes a 32-hospital, industry-independent, assessor-blinded, flexible-dose randomized trial of 3,067 inpatients aged 18-45 on six weeks of olanzapine, risperidone, quetiapine, aripiprazole, ziprasidone, perphenazine or haloperidol; 82% completed follow-up. [1] |
| aripiprazole had the lowest sedation rate of the seven antipsychotics in the trial | confirmed | the abstract says "aripiprazole was least sedating (relative risks: 0.30-0.39)", meaning a 61-70% lower relative risk of sedation than each comparator. the abstract does not name the sedation instrument, and we could not read that in the full text. [1] |
| 4% of patients on aripiprazole had sedation | unverifiable | the abstract gives relative risks, not per-arm percentages. the full text at the journal returned HTTP 403 every time we tried it, so we could not read the per-arm table. the number is plausible and consistent with the published relative risks, but we did not see it in the primary record. [1] |
| aripiprazole's sedation was "the lowest of all anti-psychotics" (if read as all antipsychotics, not just the seven tested) | overstated | the trial ranked only seven drugs. in a network meta-analysis of 32 oral antipsychotics, the abstract names pimozide as lowest for sedation compared with placebo (risk ratio 0.92), and an earlier 15-drug analysis named amisulpride. neither drug was tested in SINO. read in context, the speaker most likely meant the seven drugs in the trial. [2][3] |
| this was a novel finding not reported previously using a good measure of sedation | unverifiable | the abstracts of the two largest prior network meta-analyses do not name aripiprazole at the low end for sedation. they name pimozide and amisulpride. we did not read their full forest plots, so we cannot say whether aripiprazole's low ranking among these seven drugs was already visible there. "good measure" cannot be judged without the full methods, which we could not access. [2][3] |
both directions
for the speaker: the trial's own abstract states that aripiprazole was the least sedating of the seven drugs, with relative risks of 0.30 to 0.39 against every comparator. this is a large, industry-independent head-to-head trial with 3,067 patients, not a network meta-analysis that stitches together placebo-controlled trials, which gives the ranking weight. the 4% figure fits the published relative risks: it implies roughly 10-13% sedation on the other drugs, a range that sits near the 8.6% pooled rate of sedative adverse events in a 2025 individual-participant meta-analysis.
against: the exact 4% is not in any part of the primary record we could open. the trial was assessor-blinded, not double-blind, so patients and treating clinicians knew which drug was given, and sedation is a side effect that partly depends on reports. "lowest of all antipsychotics" holds only within the seven drugs tested. larger meta-analyses name other drugs, pimozide and amisulpride, at the bottom for sedation. the 2025 individual-participant meta-analysis found that 83% of sedation events began within the first two weeks and half had resolved within a week. a six-week snapshot rate therefore says more about early sedation than about how sedating a drug is over months.
what this cannot say
we read the abstract of the trial, not the full paper. the journal blocked access, so we could not see the per-arm sedation percentages, the sedation instrument, how sedation was defined, or dosing. the population was han chinese inpatients aged 18-45 during an acute exacerbation, treated for six weeks with flexible doses, so the rates may not carry over to outpatients, older adults, other populations or long-term treatment. our check on the novelty clause rests on meta-analysis abstracts, not their full drug-by-drug results. a relative risk shows how the drugs ranked, not how much sedation any one person will have.
sources, each one fetched and read
- SINO trial investigators. Efficacy and Tolerability of Seven Antipsychotic Drugs in Acutely Ill Patients With Schizophrenia: A Randomized, Multicenter, Assessor-Blinded Trial. American Journal of Psychiatry, 2026;183(2):112-123. doi:10.1176/appi.ajp.20250111. PMID 41152254 What we read there: 3,067 patients randomized, 82% completed follow-up; "aripiprazole was least sedating (relative risks: 0.30-0.39)"; olanzapine highest weight-gain risk (relative risk 1.44-3.22); aripiprazole lower hyperprolactinemia risk than all others (0.11-0.21). no per-arm sedation percentages in the abstract. Fetched 2026-10-15 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/41152254/
- network meta-analysis of 32 oral antipsychotics for acute multi-episode schizophrenia. Lancet, 2019;394(10202):939-951. PMID 31303314 What we read there: risk ratios versus placebo for sedation (30,770 participants) ranged "from 0·92 (0·17 to 2·03; pimozide) to 10·20 (4·72 to 29·41; zuclopenthixol)". Fetched 2026-10-15 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/31303314/
- multiple-treatments meta-analysis of 15 antipsychotic drugs in schizophrenia. Lancet, 2013;382(9896):951-962. PMID 23810019 What we read there: odds ratios for sedation ranged "1·42 (amisulpride) to 8·82 (clozapine)"; the abstract mentions aripiprazole only in connection with prolactin. Fetched 2026-10-15 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/23810019/
- individual-participant meta-analysis of the trajectory of sedative adverse events with antipsychotics. Lancet Psychiatry, 2025;12(4):266-275. doi:10.1016/S2215-0366(25)00025-2. PMID 40113354 What we read there: 6,791 participants in 19 RCTs; sedative adverse events in 8.6%; 418 of 505 (83%) sedation events began in the first 2 weeks; 50% resolved within 1 week; 24% still sedated after one month on oral agents. Fetched 2026-10-15 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/40113354/
speakers are described by role, not named. one sentence is quoted under fair use, with attribution and a link to the original at the timestamp; we do not re-host audio or video. nothing on this page is a reason to start, stop or change a medication. if you have the file behind anything marked unverifiable, the research board is where people can answer. adam