Antipsychotics linked to higher risk of pneumonia in Alzheimer's patients

People with Alzheimer’s disease who take antipsychotic medications have twice the risk of pneumonia compared with people not taking antipsychotics, according to a study published online in Chest.

Because this was an observational study, the researchers can’t say whether antipsychotics caused pneumonia, “but the risk-benefit balance should be considered when antipsychotics are prescribed,” they advised.

Several studies in the past decade have linked antipsychotics to an increased risk of pneumonia but only one small study included people with Alzheimer’s disease, even though a significant proportion of people with Alzheimer’s use antipsychotics.

For this study, researchers in Finland investigated whether antipsychotic use was related to higher risk of hospitalization or death due to pneumonia in the MEDALZ cohort. This cohort included 60,584 people in Finland who received a clinically-verified diagnosis of Alzheimer’s disease between 2005 and 2011. The researchers used a matched comparison cohort of 60,584 people without Alzheimer’s disease to compare the magnitude of risk.

Results showed that antipsychotic use was associated with a two-fold increased risk of pneumonia among patients in the Alzheimer’s cohort, and an even greater risk (3.43-fold) in people who did not have Alzheimer’s disease. The three most commonly-used antipsychotics—quetiapine, risperidone, and haloperidol—all had similar associations with pneumonia risk.

The researchers concluded that antipsychotic use was associated with a higher risk of pneumonia regardless of age, applied study design, treatment duration, or the choice of drug. Also, because the study included only cases of pneumonia that involved hospitalization or death, the actual risk increase may be even higher, the researchers noted.

Thus, they recommend that clinicians carefully weigh the risks vs. benefits when considering whether to prescribe antipsychotics, and to make the treatment period as short as clinically possible.