Prospective Study of Fluoxetine Treatment and Suicidal Behavior in Affectively Ill Subjects

Abstract

OBJECTIVE: There has been speculation in the literature about a link between fluoxetine use and suicidal behavior. The authors of this study hypothesized that there is no elevation in risk of suicidal behavior associated with use of fluoxetine. METHOD: The data come from the National Institute of Mental Health Collaborative Depression Study, a prospective, naturalistic follow-up of persons who presented for treatment of affective disorders. The analyses included data on 643 subjects who were followed up after fluoxetine was approved by the Food and Drug Administration in December 1987 for the treatment of depression. RESULTS: Nearly 30% (N = 185) of the study group was treated with fluoxetine at some point during the follow-up period. Relative to the other subjects, those who were subsequently treated with fluoxetine had onset of affective illness at a younger age and, after intake into the study and before 1988, had elevated rates of suicide attempts before fluoxetine treatment. A mixed-effects survival analysis that incorporated treatment exposure time, multiple treatment trials, and multiple suicide attempts per subject showed that relative to no treatment, use of fluoxetine and use of other somatic antidepressants were associated with nonsignificant reductions in the likelihood of suicide attempts or completions. Severity of psychopathology was strongly associated with elevated risk, and each suicide attempt after intake into the Collaborative Depression Study was associated with a marginally significant increase in risk of suicidal behavior. CONCLUSIONS: The results do not support the speculation that fluoxetine increases the risk of suicide. Rather, there was a nonsignificant reduction in risk of suicidal behavior among patients treated with fluoxetine, even though those subjects were more severely ill before treatment with fluoxetine.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 92 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.