Resources for Medical Professionals
Anti-depressants & Children (The issue of increased thoughts of suicide.)
The following article was published in the newsletter Journal Watch in January 2007.
Depression and Antidepressants in Children: Scylla and Charybdis? By Barbara Geller, M.D.
Many clinicians struggle with determining how to treat depressed children and adolescents at risk for suicide. In 2006, concerns that antidepressants might increase suicidality in these patients continued to complicate clinical decisions.
To understand the recent controversies, it may be useful to examine the history of research in the area, starting in the 1990’s. Researchers found high prevalence of suicidal ideation, completed suicides, and serious problems into adulthood among children and adolescents with depressive disorders (JW Psych Feb 2001, p. 13, and Am J Psychiatry 2001 Jan 158:125; Br J Psychiatry Sep; 179:218-23; JW Psych Nov 1999, p. 85, and Arch Gen Psychiatry 1999 Sep; 56:794; JW Psych Jul 1999, P. 58, and JAMA 1999 May 12; 281:1707; J Am Acad Child Adolesc Psychiatry 1993 Jan; 32:8-20; J AmAcad Child Adolesc Psychiatry 1993 Jan; 32:21-7). Investigators found that the prevalence of serious suicidal ideation in depressed pediatric patients ranged from 22% to 32% (see JW Psych Jul 1999, p. 58, and JAMA 1999 May 12; 281:1707; J Am Acad Child Adolesc Psychiatry 1993 Jan; 32:8-20) and that comorbid conduct disorder greatly worsened suicidality (J Am Acad Child Adolesc Psychiatry 1993 Han; 32:8-20).
These data compelled aggressive intervention by clinicians to prevent the horrific circumstance of teenagers taking their own lives, and use of pharmacologic interventions was deemed appropriate. After NIMH-funded studies of fluoxetine for depression in teens and children, there was a spate of largely industry-sponsored trials (see J Child Psychol Psychiatry 2005 Jul: 46:735-54). Overall, these data supported efficacy for fluoxetine and sertraline, but not for paroxetine.
Paradoxically, antidepressants have now come under suspicion as a potential cause of the same suicidal behaviors that the medications were given to treat. (e.g., JW Psych Oct, p.84, and Arch Gen Psychiatry Aug; 63:865; JW Psych Sep 2004, p69, and JAMA 2004 Jul 21; 292:338). Finding a scientific basis for supporting or rejecting claims of a relationship between antidepressants and suicide has been fraught with methodological issues, largely because drug-efficacy studies have not fully addressed suicidal ideation and behaviors. A meta-analysis (JW Psych May p. 44, and Arch Gen Psychiatry Mar; 63:332) from the FDA showed a non-significantly increased risk for suicidality among medicated patients in drug-efficacy trials.
The question of increased suicidality upon initiation of treatment has recently become a research focus, and case-control studies have yeilded conflicting results (e.g., JW Psych Mar, p. 21, and Am J Psychiatry Jan; 163:41; JW Psych Dec 2003, p. 93, and Arch Gen Psychiatry 2003 Oct; 60:978). For example, a large case-control investigation (JW Psych Oct, p. 84 and Arch Gen Psychiatry Aug; 63:865) has shown elevated rates of suicide completion for the first 1 to 4 months of treatment. (An earlier, very large, epidemiologic study showed a 38-fold increase within the first 9 days after therapy was initiated [JW Psych Sep 2004, p.69, and JAMA 2004 Jul 21; 292:338].) Other researchers, however, have not seen such increases (e.g., JW Psych Mar. p. 21, and Am J Psychiatry Jan; 163:41).
Given the methodologic issues, disparate results are not unexpected, but they beg the question of what clinicians and families should do until better data arrive. Some precautions are commonsensical. Physicians could avoid hospital discharge of patients with suicidality on Fridays or before holidays, when fewer services are available. It would be prudent to have suicide precautions for the first month of antidepressant administration. In addition, parents should keep the medications supply with them or in a locked cabinets to discourage teenagers from overdosing on their antidepressants.
Diligent education of families about removing firearms from the home remains an important, data-based strategy for lowering suicide-completion rates. Both epidemiologic and case-control data have strongly supported gun removal. Guns in the home were associated with increased suicide completion among psychiatric patients and healthy individuals (Am J Dis Child 1993 Oct; 147:1066-71), and teenage suicides decreased by 11% in states with gun safety laws (JW Psych Oct 2004, p. 79, and JAMA 2004 Aug 4; 292:594). Regrettably, 39% of parents were unaware of their children’s knowledge about guns in the home (JW Psych Jul, P. 60, and Arch Pediatr Adolesc Med May; 160:542), and only 27% of parents of depressed teens followed advice to remove home firearms over a 2-year follow-up (J Am Acad Child Adolesc Psychiatry 2000 Oct; 39:1220-6).
Another possibly relevant issue is that children and adolescents have much high switch rates from depressive disorders to bipolar disorder (≤33%) than do adults (JW Psych Feb 2001, p. 13, and Am J Psychiatry 2001 Jan; 158:125; JW Psych Nov 1999, p. 85, and Arch Gen Psychiatry 1999 Sep; 56:794). This difference is expected, because adults are past the typical age of bipolar disorder onset. Because bipolar diagnoses carry a substantially higher risk for suicide completion than depression diagnoses, clinicians should be especially cautious in prescribing antidepressants to pediatric patients who have first- or second-degree relatives with bipolar I disorders.
Someday, data might enable better identification of patients at risk for suicidality. Until then, families need to be informed of the controversies about the relationship among depression, its treatment, and suicide risk, so that they can knowledgably participate in individualized decisions for their children.
--Barbara Geller, MD
