SerpApi.Org
Researchers who tracked roughly 165,000 dementia patients in the United Kingdom found that risperidone, an antipsychotic sedative prescribed to manage agitation and psychosis, raised the risk of stroke across every patient subgroup they examined. The elevated risk held even among people with no prior history of heart disease or stroke. The finding sharpens a safety signal that has trailed risperidone for two decades, raising hard questions about whether existing warnings reach prescribers and families quickly enough to change real-world practice.
Risperidone’s stroke signal in dementia patients and why it matters in 2026
Risperidone has carried a federal boxed warning since 2005, when the FDA flagged increased mortality in elderly patients with dementia-related psychosis who were given atypical antipsychotics. A post-warning analysis in JAMA Network Open found measurable drops in antipsychotic prescribing afterward, yet the drug has remained in wide off-label use for behavioral symptoms of dementia. Its U.S. labeling still states plainly that risperidone is not approved for dementia-related psychosis and that cerebrovascular adverse reactions, including stroke and transient ischemic attack, were reported in clinical trials involving elderly dementia patients.
What the new UK research adds is scope. Earlier trial data showed stroke risk in selected populations, but the matched cohort analysis covering approximately 165,000 people demonstrated that the hazard was consistent regardless of a patient’s baseline cardiovascular profile. That distinction matters because clinicians have sometimes treated patients without prior strokes or heart disease as relatively safer candidates for risperidone. The new data challenges that assumption directly and suggests that the risk is baked into the way the drug affects vulnerable brains and blood vessels, rather than being purely a function of pre-existing vascular disease.
A practical question follows from this finding. If hospital and clinic electronic health record systems embedded real-time stroke-risk alerts whenever a prescriber entered a risperidone order for a dementia patient, the intervention could reduce cerebrovascular events within a year, even without any change to the existing boxed warning. The logic is straightforward: the warning already exists on paper, but a point-of-care alert would force a clinical pause at the moment the prescription is written, not months later when a pharmacist or family member reads the label. No health system has publicly reported testing such an alert at scale, which itself signals a gap between the regulatory record and bedside practice.
What 165,000 patient records and trial-level data show
The core dataset comes from a UK matched cohort study conducted by researchers at Brunel University of London, who analyzed electronic health records of approximately 165,000 dementia patients. Their central finding was that stroke risk rises with risperidone even in patients with no history of heart disease or stroke. The consistency of that result across subgroups distinguishes this analysis from earlier, smaller studies that left room for clinicians to argue the risk was concentrated in patients who already had vascular problems.
Separate trial-level evidence reinforces the pattern. An individual participant data meta-analysis of risperidone dementia trials, summarized for clinicians in an educational review, quantified cerebrovascular event risk using hazard ratios and described how quickly events occurred after treatment began. That time-to-onset data is clinically significant because it suggests the danger is not limited to long-term use; strokes and transient ischemic attacks can emerge within weeks of initiation, particularly in frail older adults. For prescribers, that compresses the window in which they must monitor for neurological changes and reconsider the drug if benefits are not clear and immediate.
The U.S. prescribing label for risperidone, maintained by the National Library of Medicine through DailyMed, documents that serious cerebrovascular reactions including fatalities appeared in the original clinical trials of the drug in elderly patients with dementia-related psychosis. Those trial reports noted strokes, transient ischemic attacks, and sudden deaths occurring at higher rates in drug-treated groups than in placebo groups. While the label does not quantify risk for each individual patient, it underpins the existing boxed warning and confirms that the UK cohort findings are not an isolated statistical anomaly but rather an extension of a long-standing safety concern.
Together, the real-world cohort data and the trial-level meta-analysis point in the same direction: risperidone carries a stroke risk that does not spare any identifiable subgroup of dementia patients. The convergence of two independent methodologies, one observational and one experimental, strengthens the overall weight of the evidence. For clinicians, this convergence raises the bar for justifying risperidone use in dementia, pushing it further toward a last-resort option when other strategies have failed and symptoms remain dangerous or unmanageable.
Gaps in the evidence and what families should watch for next
Several questions remain open. The exact hazard ratios from the Brunel University cohort study have not been independently verified through peer replication in a second large population. No U.S.-based electronic health record analysis has yet quantified how many dementia patients are currently receiving risperidone prescriptions in the wake of these findings, or how many strokes could be averted if use declined. And neither the FDA nor the Centers for Medicare and Medicaid Services has publicly announced whether the across-subgroup results will prompt a label update, a prescribing audit, or new clinical decision support requirements in Medicare-participating systems.
The absence of a regulatory response so far does not mean one will not come, but it does mean the burden of awareness falls on patients, families, and individual prescribers in the near term. Families caring for someone with dementia who is taking risperidone should ask the prescribing physician directly whether the drug’s stroke risk has been weighed against the specific behavioral symptoms it is meant to control, and whether non-pharmacological approaches such as environmental changes, caregiver training, and structured routines have been tried first. They can also ask whether other medications with comparatively lower cerebrovascular risk might reasonably be substituted.
Warning signs of stroke or transient ischemic attack in someone on risperidone are the same as in the general population but may be harder to spot in a person with cognitive impairment. Sudden weakness on one side of the body, new trouble speaking or understanding words, facial drooping, sudden vision loss, or abrupt confusion that is different from the person’s usual baseline should prompt immediate emergency evaluation. Because the risk appears to be elevated even early in treatment, the first weeks after starting or increasing risperidone may warrant especially close observation by caregivers.
The next development to watch is whether any major health system or national regulator moves to embed automated stroke-risk prompts into prescribing workflows. An electronic alert that activates when a clinician orders risperidone for a patient with a dementia diagnosis could display a concise summary of the elevated stroke risk, link to current guidelines, and require the prescriber to confirm that non-drug measures have been attempted. Such a system would not ban risperidone outright but would make its risks harder to overlook in busy clinics and hospital wards, potentially narrowing use to situations where benefits clearly outweigh the dangers.
In parallel, researchers are likely to pursue further analyses that stratify risk by dose, duration, and co-prescribed medications, as well as by dementia subtype. Those details could eventually refine guidance by identifying combinations that are particularly hazardous or, conversely, scenarios in which the incremental risk is smaller. Until such data are available, the safest assumption, based on the current evidence, is that no dementia subgroup can be considered exempt from risperidone’s stroke risk.
For now, the practical takeaway is straightforward but sobering. Risperidone remains a powerful tool for calming severe agitation and psychosis in dementia, yet its use carries a consistent and sometimes fatal cerebrovascular hazard that extends even to patients without prior vascular disease. In the absence of new regulatory mandates, it falls to clinicians to integrate that knowledge into every prescribing decision, and to families to insist on clear explanations of why this particular drug is being chosen, what alternatives exist, and how stroke warning signs will be watched for in the weeks and months ahead.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
