Nursing home residents over 65 who take certain antipsychotics for dementia are at an increased risk of death, according to a study involving 75,445 patients.

In a paper published on, the Harvard Medical School study, the largest ever undertaken among US nursing home residents, looked at risks of mortality in these individuals during a six month period. Specifically, it assessed mortality risks associated with antipsychotics including Bristol-Myers Squibb/Otsuka's Abilify (aripiprazole), AstraZeneca's Seroquel (quetiapine), Eli Lilly's Zyprexa (olanzapine) and Pfizer’s Geodon (ziprasidone), plus Johnson & Johnson's Haldol (haloperidol) and Risperdal (risperidone).

Out of the whole study, some 6,598 died within the six months from non-cancer related causes. The study notes that patients treated with haloperidol had double the risk of death compared with those taking risperidone, while those taking quetiapine had a reduced risk. The effect of haloperidol was strongest during the first 40 days of treatment which did not change after a dose adjustment. Almost half of deaths (49%) were recorded as due to circulatory disorders, 10% to brain disorders and 15% to respiratory disorders.

The authors of the study suggested that not all antipsychotics carry the same risk of death in elderly populations but “clinicians may want to consider this evidence when evaluating…the best approach to treatment of behavioural problems”.

The US Food and Drug Administration warned in 2005 that atypical antipsychotics are associated with an increased risk of mortality in elderly patients with dementia, but the BMJ notes that questions still remain on whether risks differ by drugs. The agency's warning was expanded to include conventional antipsychotics in 2008, but the authors concluded by saying  the use of these drugs is likely to continue because of the “continued growth of the dementia population” and the need for some type of intervention.

In an accompanying editorial on, Jenny McCleery from the Oxford NHS Foundation Trust, argues that "the use of any antipsychotic in dementia is undesirable given the increased risk of death and the many other adverse effects of these drugs, in addition to their limited efficacy against target behavioural and psychological symptoms". She added that "future research should be pragmatic. It should focus on identifying the key components of non-drug based interventions and on establishing the service structures that can deliver them as simply and efficiently as possible".