Systematic review and meta-analysis shows increased mortality risk with antipsychotic drugs in dementia patients

Photo by 1981 Digital / Unsplash

Frontiers in Medicine Published May 8, 2026 Medically reviewed May 8, 2026 DOI ↗ By Dr. Ji-eun Park, MD · Brain, Mind & Pain

Key Takeaway

Note increased mortality risk with antipsychotics in dementia; prescribe cautiously.

This systematic review and meta-analysis examined the association between antipsychotic medication use and mortality in people living with dementia. The study included two million participants drawn from community-dwelling individuals and other settings where specific details were not reported. The primary outcome was mortality, while secondary outcomes included stroke, pneumonia, hip fractures, hospitalization, and cerebrovascular events.

The analysis found a significantly increased mortality risk associated with antipsychotic use. The pooled hazard ratio was 1.32 with a 95% confidence interval of 1.12 to 1.56. For cerebrovascular events, the pooled HR was 1.77 (95% CI 0.92-3.42), representing an attenuated but non-significant association. Evidence for other adverse outcomes such as stroke, pneumonia, hip fractures, and hospitalization was described as limited and heterogeneous.

The authors noted considerable heterogeneity with an I2 value of 98.86% for mortality. They also highlighted limited and heterogeneous evidence for other adverse outcomes. Given these limitations and the increased mortality risk, the authors emphasize the necessity of cautious prescribing, regular medication review, and close monitoring for people with dementia.

Study Details

Study typeMeta analysis

EvidenceLevel 1

PublishedMay 2026

View Original Abstract ↓

Antipsychotic drugs (APD) are commonly prescribed to people living with dementia (PwD) to manage behavioral and psychological symptoms of dementia (BPSD), despite well-documented safety concerns. International guidelines recommend non-pharmacological interventions as first-line treatment, yet APDs remain widely used. This systematic review and meta-analysis synthesize current evidence on the consequences of APD use in PwD. Following PRISMA guidelines, we conducted a comprehensive search of PubMed, EMBASE, Cochrane Library, and Web of Science for studies published between January 2010 and August 2024. The protocol was registered in PROSPERO on 25 March 2022 (CRD42022312570). Eligible studies included observational and interventional designs reporting APD use in PwD. Risk of bias was assessed using the Cochrane tool for randomized trials and the Newcastle-Ottawa Scale for observational studies. Hazard ratios (HRs) were pooled using a random-effects meta-analysis. Subgroup analyses were performed by APD type (typical vs. atypical), study design, risk of bias, and patient setting. The certainty of the evidence was evaluated using GRADE. Forty-five studies comprising two million participants were included. Twenty-five studies reported mortality outcomes. Meta-analysis showed APD use was associated with a significantly increased mortality risk (pooled HR = 1.32; 95% CI 1.12, 1.56), with considerable heterogeneity (I2 = 98.86%). Subgroup analysis indicated similar risk elevation for both typical and atypical APDs and higher hazards among community-dwelling individuals. Evidence for other adverse outcomes, such as stroke, pneumonia, hip fractures, and hospitalization, was limited and heterogeneous, though several individual studies indicated elevated risks. Meta-analysis of cerebrovascular events showed an attenuated but non-significant association (pooled HR = 1.77; 95% CI 0.92-3.42). Conclusion: APD use in PwD is consistently associated with increased mortality and may even elevate risks for serious non-fatal events. These findings emphasize the necessity of cautious prescribing, regular medication review and close monitoring for PWD. Future research can explore drug-specific effects, dose-response relationships, and long-term outcomes, particularly in low-resource settings.Systematic review registrationhttps://www.crd.york.ac.uk/prospero/, identifier CRD42022312570.