Published: Oct. 17, 2006
Updated: May 20, 2010
John March, MD, MPH, is a professor of psychiatry and chief of Child and Adolescent Psychiatry. He led the Treatment of Adolescents with Depression (TADS) study, published in the August 18, 2004 Journal of the American Medical Association (JAMA), as well as the Pediatric OCD Treatment Study (POTS), published in the October 27, 2004 JAMA.
In February 2004, the FDA convened an advisory panel and public hearing to discuss reports of increased rates of suicidality in pediatric patients with major depressive disorder treated with antidepressant medications. The hearing prompted a spate of articles in the lay press about the dangers of antidepressants in young people, as well as numerous phone calls and visits to physicians from concerned parents.
After a comprehensive data review, the FDA in October issued a "black box" directive for all antidepressants: drug manufacturers are required to add a warning (bordered in black) that describes the increased risk of suicidal thinking and behavior in children and adolescents and alerts caregivers about the risk. Prescribers are encouraged to balance risk with clinical need and to monitor patients on antidepressants closely. Finally, the FDA deemed that a Patient Medication Guide advising patients of the risks and offering precautions be distributed with every filled prescription and refill.
At this point, some perspective about the place of antidepressants in the treatment of mentally ill youth is warranted.
Twenty years ago -- when the first SSRI (selective serotonin reuptake inhibitor), fluoxetine (Prozac), received FDA approval -- some in the psychiatric community still doubted whether depression even existed in adolescents and children. The primary medications prescribed for depression were the tricyclic antidepressants -- like imipramine (Tofranil) -- that are side effect-ridden, dangerous in overdose, and in fact don't work for depression in young people.
We've made tremendous progress since then: clinicians concur that depression in young people is real, and some medication trials in patients under age 18 demonstrate the efficacy of newer-generation antidepressants for treating depression, anxiety, obsessive-compulsive disorder (OCD), and eating disorders. However, in the 1990s, case reports began to emerge hinting that SSRIs may be associated with the development of suicidal thoughts in youth. These were case reports, not controlled clinical studies. Only recently have enough patients in this age group participated in trials that the FDA could systematically review the data.
The FDA meta-analysis that brought about the black box directive included 24 trials of nine antidepressant drugs (SSRIs and others) involving more than 4,400 patients with major depressive disorder (MDD), OCD, or other psychiatric disorders.
Although no suicides occurred in the trials, the analysis showed a 4 percent risk of suicidality -- suicide events such as increased ideation, attempts, and interrupted attempts -- during the first few months in those receiving antidepressants. That was twice the placebo risk of 2 percent. In other words, a doctor treating 100 patients with an SSRI and 100 with placebo over one year would see six suicide events: two on placebo, two on active drugs that would have happened anyway, and two which could be attributed in part to the SSRI. Since most suicidal behavior occurs in the context of mental illness, major life stresses, and substance abuse, picking out the very small risk associated with medication likely is impossible -- in part because we don't know who is at risk or the mechanism by which this risk increases.
Similarly, a 2007 meta analysis found that antidepressants increased the risk of suicidality by a small margin, especially in clinically depressed patients. The study, led by Jeffrey A. Bridge, PhD, of Ohio State University, compared the rate of suicide events with the rate of effective outcomes from antidepressants for patients with MDD, OCD, and non-OCD anxiety disorders. Data showed that antidepressants increased the risk for suicidal events most among patients with MDD, followed by patients with non-OCD anxiety disorders and patients with OCD. The authors conclude, however, that benefits of antidepressants appear to be much greater than risks from suicidal ideation or attempt for each of the three categories. Gains from antidepressants among children younger than 12 were limited, but in teenagers, antidepressants showed strong positive benefits in non-OCD anxiety disorders, moderate benefits in OCD, and more modest benefits in MDD.
Nonetheless, there's now little doubt that these medications demonstrate a small but real adverse effect on suicide attempt rates. Since suicidality was uncommon in these trials, doctors, parents, and kids need to weigh the risk (very small) against the benefits (modest to moderate but still clinically meaningful) of the medications for most youth treated with them.
The results of the 2004 TADS trial (Treatment of Adolescents with Depression), conducted at multiple sites including Duke University, confirmed the FDA's analysis and made some additional noteworthy findings. The first of its kind to compare psychotherapy of any type to antidepressant medication in this age group, the study examined the effect of placebo, fluoxetine, cognitive behavioral therapy (CBT), and their combination in 439 patients ages 12-17 with major depressive disorder.
The analysis indicated that the combination of fluoxetine plus CBT is better than fluoxetine alone, which is better than CBT alone, which is statistically no better than placebo. The finding that CBT is no better than placebo is surprising in light of its success in previous studies. (In fact, another 2004 Duke-led trial, POTS (Pediatric OCD Treatment Study), found that treatment with CBT alone was more effective than sertraline [Zoloft] alone for youth with OCD -- again, a combination was most effective. No POTS participants experienced suicidality.)
However, in the TADS study CBT did demonstrate an important benefit -- it reduced suicidal thoughts and behaviors, both alone and in conjunction with fluoxetine treatment. CBT -- which teaches patients how to overcome negative, pessimistic attitudes -- may help patients by providing coping skills for dealing with suicidal impulses and problem-solving skills when confronted with family conflict.
More recent data (2006 and 2007) from TADS confirm that medication accelerates recovery from depression and that the benefits are stronger when medication is combined with CBT than medicine alone. Moreover, combined treatment also shows a much greater impact on function, quality of life, and complete remission of symptoms than either CBT or medication alone.
However, with CBT alone, the results catch up with medication by 18 weeks of treatment. They don't catch up to combined treatment, which reaches maximum benefit at week 18, as contrasted to medication or CBT, which take another two to three months to reach maximum benefit. In addition, results from the TADS long-term study show very clearly that CBT protects against the slight risk of a suicidal event in patients treated with medication.
The TADS take home message: advantage combined treatment!
Another recent study made important findings about the relationship between SSRI prescriptions and rate of completed suicides. Although the FDA found that antidepressants slightly increased the suicide attempt rate among adolescents, data indicate that it actually decreases the rate of completed suicides among that group, by a significant margin. Earlier this year, Robert D. Gibbons, PhD, of the University of South Florida, led a study that compared the antidepressant prescription rate with the completed suicide rate in both the United States and the Netherlands.
As expected, the doctors in both countries prescribed SSRIs much less often following FDA public health warnings. But with that change came a spike in youth suicide rates in both countries. In the U.S., adolescent suicides increased by 14 percent between 2003 and 2004 -- the largest year-to-year change since systematic data collection began in 1979. The implications are clear: although antidepressants slightly increase the risk of a suicidal event for a vulnerable subpopulation, antidepressants also appear to decrease the completed suicide rate, presumably through effective treatment of depression.
So, how should our way of treating the adolescent with depression change in light of recent data and FDA directives?
In some respects, it doesn't. Every patient beginning a course of medication should be told that any hint of suicidal thoughts or self-harm needs to be reported to the physician right away. Moreover, along with being informed about the risks of medication, patients and families should receive a written list of symptoms that may foreshadow suicidal thoughts and behaviors -- including agitation, impulsivity, panic attacks, akathesia (restlessness), irritability, insomnia, hostility, hypomania, and mania. They should also be aware that major life stress as well as alcohol or drug use compounds the risk.
Understandably, primary care physicians may be wary of treating depressed youth with medication. Before treatment begins, young people with symptoms of depression need a comprehensive evaluation and accurate diagnosis. Not every patient needs medication -- those with milder cases often respond well to psychotherapy.
If the patient has mild depression and no prior or current suicidality, then the primary care physician may feel comfortable treating the patient with CBT alone, adding medication after four to six weeks if the patient shows no response to CBT. On the other hand, if the patient is very ill, has other mental disorders (such as ADHD or an anxiety disorder), or has a personal or family history of suicidality, then starting directly with combined treatment may make the most sense. Using medications alone is the least preferred option, but may be the best option if CBT is not available. In any case, the initial treatment frequent monitoring is essential, especially during the first weeks of treatment. The more severely depressed patient, particularly if suicidal, should be referred to someone who regularly deals with such patients -- a child psychiatrist, an adult psychiatrist who also specializes in adolescents, or a developmental pediatrician who treats children with severe mental illnesses.
Left untreated, childhood and adolescent depression can have serious, even fatal, consequences. Depressed young people have ongoing problems at school, with friends, and at home, and are at increased risk for substance abuse, eating disorders, and pregnancy. We now have very good evidence that SSRIs, particularly fluoxetine, accelerate recovery in depressed teens. Most clinicians agree that the benefits still outweigh the risks -- adding CBT seems to reduce the risk and very clearly enhances the effectiveness of medication. Thus, I would strongly encourage the use of combination therapy as the best treatment for teenagers with major depressive disorder, especially when there is a history of past or present suicidal ideation or behavior.