A large-scale UK study of over 165,000 dementia patients has revealed that the antipsychotic risperidone increases stroke risk across all patient groups, including those without prior heart disease. The findings challenge prior assumptions about safer candidates for the drug, which is often used as a last resort for severe agitation. Published in the British Journal of Psychiatry, the research calls for more informed prescribing practices.
Researchers from Brunel University of London analyzed anonymized NHS health records from 2004 to 2023, comparing dementia patients prescribed risperidone with similar patients not taking the drug. The study, involving more than 165,000 individuals, found elevated stroke rates in those using risperidone. For patients with a previous stroke, the annual rate reached 22.2 per 1,000 person-years on the drug, compared to 17.7 without it. Among those without prior strokes, rates were 2.9 percent versus 2.2 percent, respectively. Notably, the risk was higher for shorter treatment periods of 12 weeks.
Risperidone, the only antipsychotic licensed for dementia in the UK, is prescribed for severe agitation or aggression when non-drug strategies fail. About half of dementia patients experience such agitation, causing distress for patients and caregivers. "We knew risperidone causes stroke, but we didn't know whether some groups of people might be more at risk than others," said Dr. Byron Creese of Brunel University of London. "We thought if we might identify characteristics that make people more at risk, doctors could avoid prescribing to patients with those characteristics."
Current NHS guidance limits risperidone to six weeks for severe symptoms, though many patients use it longer. There is no specific monitoring guidance for dementia patients, and practices vary regionally. No alternative licensed drugs exist in the UK for this purpose. "These findings give clearer information about who is most at risk, which helps everyone make more informed choices," Dr. Creese added. "Every decision should be based on what is right for each person, through honest conversations between doctors, patients, and families."
The results, detailed in a 2025 British Journal of Psychiatry paper by Joshua Choma and colleagues, may prompt updates to clinical guidelines for person-centered care.
