• bipolar disorder;
  • elderly;
  • antidepressants;
  • mania;
  • bipolar depression



The goal of this study was to examine the association between antidepressant use and hospitalization rate for mania or bipolar depression in a large, community-based sample of elderly bipolar patients.


Population-based retrospective cohort design. Administrative healthcare databases were linked for all individuals aged 66 years or older in the Province of Ontario, Canada. Bipolar subjects who received a prescription for an antidepressant medication (n = 1,072) between 1 April 1997 and 31 March 2001 comprised the antidepressant cohort. The control group (n = 3,000) consisted of randomly selected subjects from the eligible bipolar population who did not receive a prescription for an antidepressant medication during the same surveillance period. Primary outcomes were admission to hospital for a manic/mixed or depressive episode.


During a total of 5135 person-years of follow up, 113 admissions for a manic/mixed episode and 28 admissions for a depressive episode were identified. Model based estimates adjusted for a number of covariates revealed that, as compared with the control group, the antidepressant cohort had significantly lower likelihood of admissions for manic/mixed (adjusted rate ratio [aRR] = 0.5, 95% CI = 0.3–0.8) but not depressive episodes (aRR = 0.7, 95% CI = 0.2–2.2).


Antidepressant use among elderly bipolar patients was associated with decreased rates of hospitalization for manic/mixed episodes. This finding requires confirmation with further data of antidepressant use among elderly bipolars. Copyright © 2006 John Wiley & Sons, Ltd.