A new study shows that use of antipsychotic drugs is associated with an early and sustained increase in risk of death when used to treat disruptive behavior of older adults with dementia.

The study suggests that both newer “atypical” antipsychotics and older conventional antipsychotics are associated with increased mortality. The highest risk appears to involve use of the older conventional drugs.

Antipsychotic drugs are approved for treating psychotic conditions such as schizophrenia and mania. But they are widely prescribed to manage behavioral symptoms of dementia, such as agitation, verbal or physical aggression and delusions, hallucinations or paranoia.

“Our study adds to mounting concerns about the use of antipsychotic drugs in dementia,” said Sudeep Gill, MD, MSc, a geriatric medicine specialist and the study’s lead author.

The FDA and Health Canada both issued warnings in 2005 that use of newer atypical antipsychotics to treat elderly patients with dementia was associated with an increased risk of death. But Dr. Gill said, “Previous trials have generally been very short and could not provide information about the long-term impact of antipsychotics on death. And relatively little information has been available on the harms associated with the older conventional or typical antipsychotics.”

Investigators in the new study linked four administrative health care databases to look at health and death records of more than 27,000 people age 66 and older diagnosed with dementia between April 1, 1997, and March 31, 2002.

The investigators found:

Estimated mortality rate among study participants was high, especially in the long-term care setting.

Conventional antipsychotics were associated with higher risk for death than atypical antipsychotics, and atypical antipsychotics were associated with higher risk for death than no antipsychotic use.

Risk for death developed quickly, within one month of use, and persisted for up to six months.

“The clinical message is that even short-term use of these drugs can be associated with an increased risk of death, so physicians need to carefully weigh potential risks and benefits of using these drugs to manage symptoms of dementia, and they need to reassess the use soon after they’re initiated to see if they can be safely discontinued,” Dr. Gill said.

Gill did not recommend that the antipsychotic drugs never be used in dementia patients. “Sometimes they ease specific symptoms of aggression and hallucination,” he said. “But they are not appropriate for everyday use for everyone with dementia.”

Dr. Gill said the antipsychotics should not be initiated if effective non-drug treatments are available for specific problems or if symptoms, such as wandering, are unlikely to respond to antipsychotic treatment.

“This study shouldn't lead to a panic about these drugs,” said Dr. Gill. “The risk for an individual patient is relatively small. But our results are clinically important.”

“I hope this study encourages discussion between patients and their physicians and between caregivers and physicians about the appropriate time to use these drugs.”

Dr. Gill and his colleagues also call for better funding for non-drug interventions.

Finally, Dr. Gill suggests that regulatory agencies might consider looking at whether the warnings applied to atypical antipsychotics should be extended to the older conventional antipsychotics, to help clinicians avoid switching their patients from atypicals to conventional antipsychotics. “In our study, conventionals carried an even greater risk than atypicals, so it would be unfortunate if physicians started switching patients with dementia onto the conventional antipsychotics.”

The study, “Antipsychotic Drug Use and Mortality among Older Adults with Dementia,” is published in the June 5, 2007, issue of Annals of Internal Medicine. The research was funded by the Canadian Institutes of Health Research and the Ontario Ministry of Health and Long-Term Care.

Dr. Gill is a geriatric medicine specialist at Providence Care – St. Mary’s of the Lake Hospital and assistant professor of medicine at Queen’s University in Kingston, Ontario. He and his colleagues conducted the research at the Institute for Clinical Evaluative Sciences (ICES), an independent, non-profit organization that uses population-based health information to produce knowledge on a broad range of health care issues.