Antipsychotic medicines are prescribed to hundreds of thousands of people with dementia to manage agitation, aggression and other distressing behaviors. A large study of British health records has added weight to a long-standing worry about that practice, linking the drugs to a broader set of serious harms than regulators have historically flagged, including a measurable increase in the risk of stroke. The findings do not prove the medicines directly cause those events, but they sharpen a picture that has been forming for two decades.
Researchers tracked what happened to people in the months after they first started an antipsychotic, comparing them with similar patients who did not. The pattern that emerged was consistent: elevated rates of several dangerous outcomes, concentrated in the earliest stretch of treatment, when families and clinicians are often still searching for a way to calm a crisis.
The BMJ cohort of nearly 174,000 dementia patients
The analysis, published in The BMJ, drew on the records of 173,910 people in England who were diagnosed with dementia at an average age of 82 between 1998 and 2018 and who had not been prescribed an antipsychotic in the year before diagnosis. Investigators from the universities of Manchester, Nottingham, Edinburgh and Dundee then matched patients who went on to receive a drug against comparable patients who did not, following both groups to see which serious health problems appeared.
The most commonly prescribed medicines in the group were risperidone, quetiapine, haloperidol and olanzapine, agents used across care homes and community settings when non-drug approaches fall short. Because the study observed real-world prescribing rather than assigning treatment at random, it can establish association but not direct causation.
Stroke, blood clots and a wider range of harms
Antipsychotic use was tied not only to stroke but to a cluster of other events, including venous blood clots, heart attack, heart failure, bone fracture, pneumonia and acute kidney injury, when set against non-use. That range is notably broader than the harms named in earlier safety alerts, which had focused most sharply on stroke and on overall death rates. The researchers framed the results as evidence that the risks accompanying these drugs in dementia have been underestimated rather than overstated.
Stroke sits at the center of the concern because it is both common in older adults and potentially devastating, capable of stealing mobility, speech and independence in a single event. An increase layered on top of an already high baseline risk translates, at the scale of a national health system, into a meaningful number of additional cases.
Why the danger clusters in the first weeks
One of the study’s more practical conclusions is about timing. The elevated risks were highest soon after patients began the medication, which argues for particular caution during the initial days and weeks of a prescription rather than treating the drugs as a routine, indefinite fixture of care. That early spike matters because antipsychotics are frequently started during an acute behavioral crisis, precisely the moment when the temptation to reach for a fast pharmacological answer is strongest.
The finding also has implications for how prescriptions are reviewed. If harm concentrates early, then short courses with prompt reassessment, rather than open-ended use, may limit exposure while still addressing the immediate problem that prompted treatment.
Existing warnings on antipsychotics in dementia
Regulators have cautioned about this drug class for years. In the United States, the Food and Drug Administration requires a boxed warning noting that older adults with dementia-related psychosis who are treated with antipsychotics face an increased risk of death, and no antipsychotic is approved specifically for dementia-related behavioral symptoms. The newer British data does not overturn that guidance so much as widen it, suggesting the list of downstream dangers is longer than the labels imply.
None of this means the medicines have no place. Severe agitation and aggression can endanger both patients and the people caring for them, and in some situations the alternatives are limited. The study’s authors position their work as a call for careful weighing of benefits and harms, not a blanket prohibition.
What clinicians weigh against the risk
Guidelines generally recommend that non-drug strategies come first: identifying and treating pain or infection that may be driving distress, adjusting the environment, and supporting caregivers who are managing difficult behavior around the clock. When a medicine is judged necessary, the emerging consensus favors the lowest effective dose, a clearly defined reason for use and regular review with an eye toward stopping.
For families navigating a dementia diagnosis, the takeaway is not to abruptly discontinue a prescribed drug, which can itself be risky, but to treat any antipsychotic as a decision worth revisiting with the prescribing clinician. The research reinforces that these medicines carry real trade-offs, that those trade-offs are heaviest at the start, and that the risk of stroke and other serious events belongs squarely in the conversation.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
More from Morning Overview
- The NSA is again telling phone owners to switch off one location setting
- A handful of car transmissions are so tough mechanics say they almost never fail
- A handful of SUVs keep hitting 300,000 miles, and they share one engine trait
- Supplements now rank as the fifth-leading cause of death from liver disease.
