Background: In issuing an advisory about a relationship of suicidality emerging during antidepressant use in mood and other psychiatric disorders in juvenile and, possibly patients of all ages, the FDA listed such warning symptomatologic correlates as psychomotor agitation, panic, insomnia, irritability, aggressiveness, and impulsivity, and additionally urged screening for personal and family history for bipolar disorder. It must be recalled that the FDA warning was based on post hoc examination of antidepressant trial data pertaining largely to emergent suicidal ideation rather than attempts. In this report we examine the relevance of the "FDA list" for adults.
Methods: Consecutive 650 outpatients presenting with a major depressive episode (MDE) in the second author's (FB) private practice were interviewed using the Structured Clinical Interview for DSM-IV, the Hypomania Interview Guide, and the Family History Screen. The threshold for a bipolar II (BP-II) diagnosis was broadened by bypassing the stem-A probes on hypomanic mood, lowering the duration of hypomanic episodes to as short as 2 days and, most importantly, eliciting hypomanic intrusions into depressive episodes. Beyond the FDA proposed list and in line with previous work by us, we considered symptoms of depressive mixed state (DMX, an MDE plus 3 or more DSM-IV intradepressive non-euphoric hypomanic symptoms) as possible substrates for suicidality.
Results: Suicidal ideation was present in 49.3% of the sample, and significantly greater in BP-II versus MDD (OR=1.46). Odds ratio for patients with suicidal ideas, versus those without suicidal ideas in univariate analyses were as follows: irritability (1.32), psychomotor agitation (1.48) and racing/crowded thoughts (2.0). In DMX, irritability was present in 76.4% and in non-DMX in 23.8% (OR=10.3), psychomotor agitation in 51.3% versus 4.4%, (OR=22.6), and racing/crowded thoughts in 89.2% versus 45.3% (OR=9.9). Insomnia and family history for bipolarity were not associated with suicidal ideation. Racing/crowded thoughts emerged as the only significant independent predictor in multiple logistic regression.
Limitation: Panic was not specifically tested, because it was not systematically entered into our clinical database. Furthermore, we studied a population which was not taking antidepressants. Our operating hypothesis was that the FDA listed correlates of suicidality are pre-existing baseline substrates that might be potentially further activated by antidepressants.
Conclusion: Going beyond previous work by us, the present data further refine the "FDA list" under scrutiny. While not designed to test whether antidepressants induce suicidality, the present report shows that the type of depression associated with suicidal ideas is an agitated, irritable, and especially mentally overactive syndrome (i.e. a DMX). Bipolarity falling short of the DSM-IV schema (represented by BP-II with brief hypomania and DMX) appears to represent the main diagnostic substrate for the occurrence of suicidality. Although controlled trials of antidepressants in adults have not been specifically conducted in DMX patients, it would be clinically prudent to avoid the prescription of antidepressant monotherapy to such patients to avert any potential iatrogenic aggravation of their condition.