Long-term benefits of medications for schizophrenia called into question

Antipsychotics, antidepressants, and benzodiazepines do not appear to have a protective association with all-cause mortality in patients with schizophrenia when controlling for immortal time bias (ITB), suggesting that the effects of these drugs on mortality may be overestimated in analyses not adjusting for ITB, according to a study.

Researchers used administrative data from Québec in Canada and identified 32,240 patients aged 17 to 64 years (mean age 46.1 years, 61.3 percent men) who had a diagnosis of schizophrenia. The primary outcome of all-cause mortality was assessed in relation to levels of exposure (low, moderate, high) to antipsychotics, antidepressants, and benzodiazepines. Two separate analyses were performed, with and without controlling for ITB, respectively.

Of the patients, 1,941 (6.0 percent) died during follow-up. For antipsychotics, there was no dose-response association with mortality observed using the time-fixed method. However, high-dose exposure to antipsychotic was associated with increased mortality after correcting for ITB (adjusted hazard ratio [aHR], 1.28, 95 percent confidence interval [CI], 1.07–1.55; p=0.008).

Antidepressants were associated with a reduced risk of mortality using the time-fixed method, but this protective association was only observed at high doses when correcting for ITB (aHR, 0.86, 95 percent CI, 0.74–1.00; p=0.047).

Finally, benzodiazepines were associated with increased mortality risk, irrespective of the adjustment for ITB.

The findings highlight the need for appropriate analytical techniques.