Quetiapine edges out rival antipsychotics in Alzheimer’s behavioral symptoms, with far fewer movement side effects

When Alzheimer’s disease drives a patient into agitation, aggression, or psychosis, families and clinicians face one of geriatric medicine’s most uncomfortable decisions: whether to reach for an antipsychotic drug whose risks in older adults include stroke, metabolic disturbance, and a nearly doubled death rate. A new systematic review and meta-analysis published in Frontiers in Psychiatry on 26 October 2026 argues that if that decision must be made, quetiapine may deserve a closer look than it has traditionally received. Pooling data from 25 randomized controlled trials involving 2,143 patients with Alzheimer’s disease, researchers from Quzhou, China, found that quetiapine was associated with significantly lower odds of treatment failure than other atypical antipsychotics, and with a dramatically reduced risk of the movement disorders that make these drugs so hazardous in the elderly.

The behavioral and psychological symptoms of dementia, abbreviated BPSD by clinicians, are not a side note to Alzheimer’s disease; they are arguably its most burdensome face. Up to 90 percent of patients develop them over the course of the illness, spanning agitation, aggression, delusions, hallucinations, apathy, depression, anxiety, and aberrant motor activity. These symptoms are a primary driver of caregiver distress, accelerated placement in nursing homes, reduced quality of life, and rising healthcare costs. Clinical guidelines place non-pharmacological approaches first: person-centered care, caregiver training, environmental modification, and structured activity therapies. But when symptoms become severe, persistent, or dangerous, those approaches often fall short, and the pharmacological gap left by the modest effects of cholinesterase inhibitors and memantine has long been filled, off-label, by antipsychotics.

That fill comes with a steep price. Meta-analyses and regulatory warnings have consistently linked antipsychotic use in older adults with dementia to cerebrovascular events, sedation, metabolic disturbance, and roughly a two-fold increase in all-cause mortality, prompting black-box warnings from the U.S. Food and Drug Administration and universal calls for judicious, time-limited prescribing at the lowest effective dose. Within the class, risperidone and olanzapine carry the strongest efficacy evidence for aggression and psychosis in dementia, but risperidone is tied to extrapyramidal symptoms, the tremor, rigidity, and restlessness of drug-induced parkinsonism, while olanzapine is notorious for metabolic side effects. Quetiapine, with its lower affinity for dopamine D2 receptors and stronger blockade of histamine H1 receptors, has long been theorized to carry a gentler neurological profile, yet individual trials of its efficacy in dementia have returned mixed results, and earlier syntheses lumped together different dementia types or lacked direct head-to-head comparisons.

To resolve that ambiguity, the research team, led by Jun Wang of the Quzhou Affiliated Hospital of Wenzhou Medical University, searched six databases, including PubMed, Embase, the Cochrane Library, Web of Science, CNKI, and WanFang, from inception to 15 March 2024, following PRISMA 2020 reporting guidelines. From 454 identified records, screening and full-text review distilled 25 eligible randomized controlled trials published between 2005 and 2022, conducted in China, the United Kingdom, the United States, Switzerland, and Israel. Study sizes ranged from 22 to 194 participants, mean ages clustered between 70 and 80 years, and treatment durations ran from 4 to 36 weeks. The trials compared oral quetiapine, given alone or added to a stable cholinesterase inhibitor such as donepezil or galantamine, against risperidone, olanzapine, haloperidol, rivastigmine, or placebo.

The primary efficacy outcome was deliberately pragmatic: treatment failure, meaning the absence of a clinical improvement defined as a reduction of 30 percent or more from baseline on the Behavioral Pathology in Alzheimer’s Disease Rating Scale, or the equivalent responder threshold on the Neuropsychiatric Inventory. Because the large CATIE-AD trial used the Clinical Global Impression of Change scale instead, the authors harmonized its data by classifying a score of 4 or higher, indicating no change or worsening, as treatment failure. Pooling 16 studies with 792 patients on quetiapine and 786 on comparators under a random-effects model, they calculated an odds ratio of 0.66 for treatment failure, with a 95 percent confidence interval of 0.49 to 0.89 and a P value of 0.006. In plain terms, patients on quetiapine were about two-thirds as likely to fail to improve as patients on the other antipsychotics, and the absence of statistical heterogeneity, with an I² of 0 percent, lent the pooled estimate unusual coherence.

The subgroup analyses sharpened the picture considerably. Against risperidone, across 10 studies covering 469 versus 457 patients, quetiapine showed a statistically significant superiority, with an odds ratio of 0.50 for treatment failure, a 95 percent confidence interval of 0.33 to 0.75, and P equal to 0.0009. Against olanzapine, across 6 studies with 323 versus 329 patients, the point estimate still favored quetiapine at 0.90, but the confidence interval of 0.58 to 1.38 crossed unity, so the difference did not reach conventional significance. The authors suggest that quetiapine’s broader receptor activity, spanning serotonergic 5-HT2A and adrenergic α1 blockade, may address the multifaceted symptom cluster of BPSD more comprehensively, while risperidone’s higher rate of extrapyramidal symptoms may force dose reductions, discontinuation, or anticholinergic rescue medications that can themselves worsen cognition and erode apparent effectiveness in real-world trials.

The safety findings may prove the analysis’s most consequential contribution. Across 8 studies with 330 quetiapine and 336 control patients, the odds of extrapyramidal symptoms were 0.17, with a 95 percent confidence interval of 0.09 to 0.33 and P below 0.00001, an 83 percent relative reduction in the odds of these movement complications, with zero heterogeneity. This aligns squarely with quetiapine’s pharmacodynamics: its transient, low D2 receptor occupancy underlies its low EPS liability even at higher doses, whereas risperidone’s dose-dependent D2 occupancy is tightly linked to parkinsonian effects. Equally important, the feared trade-off did not materialize. Somnolence, driven by quetiapine’s H1 antagonism, showed no significant difference across 10 studies, with an odds ratio of 0.90. Dizziness across 12 studies yielded an odds ratio of 0.86, constipation across 10 studies 1.05, nausea across 7 studies a non-significant 0.56, and tremor across 7 studies 0.57. None reached statistical significance, countering the assumption that quetiapine’s neurological gentleness comes at the cost of heavier sedation.

The results sit in apparent tension with CATIE-AD, the landmark trial that found no significant advantage for quetiapine, olanzapine, or risperidone over placebo in outpatients with Alzheimer’s disease. The meta-analysis authors attribute the divergence to three methodological factors: CATIE-AD enrolled community-dwelling outpatients with relatively moderate symptom severity, whereas the pooled trials included more acutely agitated or hospitalized patients in whom antipsychotic stabilization is more pronounced; the mean quetiapine dose in CATIE-AD was a mere 56.5 milligrams per day, possibly subtherapeutic; and CATIE-AD’s primary endpoint, time to discontinuation for any reason, was a composite heavily influenced by early adverse events and caregiver burden rather than a direct measure of symptomatic control. A funnel plot for the primary outcome showed a symmetrical distribution, suggesting no substantial publication bias, and the team pre-specified a sensitivity analysis excluding combination-therapy trials in order to verify that the efficacy signal was not driven by concurrent cholinesterase inhibitors.

Caveats temper the enthusiasm. The constituent trials carried a moderate overall risk of bias on the revised Cochrane tool: random sequence generation and allocation concealment were frequently unclear, and most studies were open-label or single-blind, raising performance bias concerns, though detection, attrition, and reporting bias were generally rated low. Distinct side-effect profiles may have unblinded participants and assessors. The harmonization of CATIE-AD’s global impression scores relied on published aggregate totals rather than individual participant data, a source of potential ascertainment bias. Most critically, the analysis captured only short- to medium-term outcomes, leaving long-term risks such as weight gain, metabolic syndrome, cerebrovascular events, tardive dyskinesia, and mortality unassessed, and many trials tested quetiapine alongside cholinesterase inhibitors, so the findings cannot be read as an endorsement of quetiapine monotherapy in isolation.

The authors’ conclusion is correspondingly restrained: quetiapine appears to be a neurologically better-tolerated option when non-pharmacological strategies have failed and pharmacotherapy is unavoidable, not a first-line treatment in its own right. Prescribing should still begin at the lowest effective dose, be re-evaluated regularly for continued necessity, and proceed alongside ongoing non-drug support. For patients with parkinsonian features, high fall risk, or prior extrapyramidal reactions to other antipsychotics, however, the combination of a significant efficacy signal and a markedly lower movement-disorder risk gives quetiapine a defensible niche in a field with few good options. The researchers call for large, well-blinded, long-term comparative effectiveness trials with comprehensive metabolic and functional outcomes, and ideally individual participant data meta-analyses, to confirm whether this short-term advantage endures where it matters most: over years, not weeks.

Subject of Research: Comparative efficacy and safety of quetiapine versus other atypical antipsychotics for behavioral and psychological symptoms in Alzheimer's disease

Article Title: Comparative efficacy and safety of quetiapine versus other atypical antipsychotics for behavioral and psychological symptoms in Alzheimer’s disease: a systematic review and meta-analysis

Article References: Wang, J., Chen, W., & Zhang, Z. (2026). Comparative efficacy and safety of quetiapine versus other atypical antipsychotics for behavioral and psychological symptoms in Alzheimer’s disease: a systematic review and meta-analysis. Frontiers in Psychiatry, 17, Article 1887284. https://doi.org/10.3389/fpsyt.2026.1887284

Image Credits: AI Generated

DOI: 10.3389/fpsyt.2026.1887284

Keywords: Alzheimer's disease, quetiapine, atypical antipsychotics, behavioral and psychological symptoms of dementia, meta-analysis, extrapyramidal symptoms, risperidone, olanzapine, randomized controlled trials, drug safety, geriatric psychiatry, agitation