The article you requested is

Cerebrovascular Events Among Elderly Nursing Home Patients Treated With Conventional or Atypical Antipsychotics

J Clin Psychiatry 2005;66(9):1090-1096

 Objective: Concern exists about a possible increased risk of cerebrovascular events (CVEs) among elderly patients receiving risperidone or olanzapine. We estimated the effect of atypical and conventional antipsychotics on the risk of CVEs among elderly nursing home patients with dementia.

Method: We conducted a case-control study on residents of nursing homes in 6 U.S. states by using the Systematic Assessment of Geriatric drug use via Epidemiology database, which includes data from the Minimum Data Set linked to Medicare inpatient claims. Participants were diagnosed with Alzheimer's disease or other forms of dementia on the basis of clinical criteria and medical history (including medical records and neuroradiologic documentation). Cases included patients hospitalized for stroke or transient ischemic attack between June 30, 1998, and December 27, 1999. For each case, we identified up to 5 controls hospitalized for septicemia or urinary tract infection residing in the same facility during the same time period. The sample consisted of 1130 cases and 3658 controls.

Results: After controlling for potential confounders, the odds ratio of being hospitalized for CVEs was 0.87 (95% CI = 0.67 to 1.12) for risperidone users, 1.32 (95% CI = 0.83 to 2.11) for olanzapine users, 1.57 (95% CI = 0.65 to 3.82) for users of other atypical agents, and 1.24 (95% CI = 0.95 to 1.63) for conventional antipsychotic users compared to nonusers of antipsychotics. A history of CVEs appeared to modify the effect of atypical antipsychotics other than risperidone on the risk of new events.

Conclusion: Overall, no increased risk of CVEs seems to be conferred by atypical or conventional antipsychotics. Preexisting cerebrovascular risk factors might interact with some atypical antipsychotics to increase the risk of events. These results should be interpreted in light of the limitations of the study and need to be confirmed.