To the Editor

The lack of adequate data on treatment adherence may be what is hampering the full understanding of the relationship between the use of selective serotonin reuptake inhibitor (SSRI) or newer-generation antidepressants and the risk of suicide in depressed children/adolescents. The clinical trial data have been analyzed and reanalyzed to answer the following question: do depressed children/adolescents treated with SSRIs and newer-generation antidepressants have a greater risk of suicide than those treated with placebo? However, adequate data do not exist to examine whether adherence mediates the relationship between antidepressant use and the risk of suicide.

This suggested mediating role of medication adherence is supported by an examination of the correlation between multiple-dose plasma elimination half-lives of the various antidepressants (venlafaxine, fluvoxamine, paroxetine, sertraline, mirtazapine, citalopram, and fluoxetine) and the risk ratio for suicide. There is a significant negative correlation between multiple-dose plasma elimination half-life and risk ratio (Spearman’s rho, r_s=–0.929, p<0.01). The risk ratio data used in this analysis are from Columbia University Outcome 3 (Food and Drug Administration’s Division of Neuropharmacological Drug Products analysis PID number D040495).

The significant negative correlation between plasma half-life and the risk of suicide strongly suggests that it is the effects of symptoms due to acute discontinuation from SSRIs and newer-generation antidepressants in children/adolescents, as opposed to the effective use of these medications by children/adolescents, that is related to the risk of suicide. We hypothesize that children/adolescents are less likely than adults to adequately articulate discontinuation symptoms and are more likely to react to these symptoms impulsively.

Dr. T. Hammad, senior medical reviewer at the Food and Drug Administration, presented his analysis of suicidality in pediatric trials of newer antidepressants to the Psychopharmacologic Drugs Advisory Committee and the Pediatric Advisory Committee on Sept. 13, 2004 (1). He found that although "medication compliance may have influenced the occurrence of the events of interest" (i.e., suicidal behavior or ideation), "the determination of noncompliance was suboptimal."

There is a paucity of research on treatment adherence in depressed children or adolescents. Most of the research on medication adherence in youth has investigated adherence to the treatment of medical disorders (e.g., asthma, arthritis, diabetes). Although the rates of medication adherence are suboptimal at all ages, Staples and Bravender (2) found that nonadherence "reaches a nadir during the adolescent years" (p. 504).

Based on the data presented, we strongly advocate for more sophisticated and systematic assessment of adherence in clinical trials of SSRIs and newer-generation antidepressants in depressed children or adolescents. We strongly encourage clinicians to assess the adherence of their child/adolescent patients taking antidepressants and to intervene with the children/adolescents and their caregivers to improve adherence when it is suboptimal.