Major depression is one of the most common psychiatric disorders of adolescence, and one whose many consequences can include mental health and substance abuse morbidity, academic and social derailment, unplanned pregnancy, and, perhaps most ominously of all, high rates of attempted and completed suicide. Epidemiological and longitudinal follow-up studies of adolescents with depression have documented high rates of recurrence, of progression into chronicity, and of continuity into adult forms of affective disorder.
These deeply concerning findings have generated important efforts in relation to treatment. As is well known, early studies on the efficacy of tricyclic antidepressants in the treatment of adolescents with depression were disappointing. More recently, psychotherapeutic and pharmacological interventions are beginning to show that there are promising therapeutic approaches. Nevertheless, there are enormous challenges for clinicians and researchers. Only a small proportion of adolescents with depression have their illness recognized, with an even smaller proportion receiving treatment. Moreover, there is considerable need for systematic, controlled treatment studies to clarify what treatments are most useful for individual children and adolescents, their impact on immediate symptoms and impairment, and their potential effect in improving natural history and long-term adaptation.
In addition to providing an expanded glimpse into the adult outcomes of adolescent depression, the article by Lewinsohn and colleagues in this issue provides an important attempt at both clarifying the phenomenology and natural history of adolescent depression as well as suggesting new hypotheses for future systematic research.
The prospect of psychiatric illness seems to loom heavily in the future of depressed adolescents. Only one in every four of the 274 teenagers with depression studied by Lewinsohn and colleagues remained free of illness by the time of their 24th birthday. Fully two-thirds of the adolescents with depression developed either depression alone, a nonmood disorder, or the comorbid combination of the two. Of special concern was the high rate of substance abuse disorders seen, which disproportionately accounted for the nonmood disorders that emerged during the young adult years. Substance abuse disorders also showed some of the most striking changes over time. Twenty percent of all depressed adolescents had a comorbid substance abuse disorder at the first time point. During the next several years, the rate increased, with substance abuse disorders representing almost 80% of nonmood outcomes by age 24 years. Given that over one-half of such nonmood outcomes represented new disorders that began after the age of 19, rather than the continuation of a preexisting condition, the sequence and timing of these disorders seems critical. These data, consistent with previous reports that have suggested a window of over 4 years between the onset of disorders (1), suggest a progression from depression into substance abuse, although causal pathways are, of course, probably complex rather than linear processes.
The observations on outcome offer an important source of hopefulness. Assuming timely identification and an effective array of interventions—perhaps including multiple modalities (familial, psychological/behavioral, and pharmacological)—it is possible that the trajectory from mood disorder to substance abuse and to other conditions can be averted. Thus, the early detection and effective treatment of adolescents with depression may provide an important window of therapeutic opportunity and help reduce the future burden of substance abuse as well as other conditions.
While the present study unfortunately provides no information regarding the effects of treatment interventions on eventual outcome or the recurrence of illness, the infrequency of any such interventions may be even more noteworthy. The fact that only 20% of depressed adolescents seen as part of this highly structured, longitudinal follow-up study received some form of treatment strongly suggests that in community settings, the adolescent window of therapeutic opportunity is more likely than not missed. Yet, even in those instances where adequate diagnostic identification is made and treatment is instituted, available data suggest a course perhaps more refractory than for adult-onset forms of depression. For example, 39% of adolescents successfully treated as part of the pivotal fluoxetine adolescent depression trial (2) had recurrences within 1 year of treatment; over one-half of those occurred within 6 months (3). Ongoing efforts such as the Treatment of Adolescents With Depression Study (4) may provide in coming years an empirically derived database to guide clinicians and to parsimoniously combine pharmacological and nonpharmacological interventions.
Identification of depressed adolescents at highest risk for illness recurrence will remain an important research goal while concurrent efforts at developing better treatment strategies proceed. The vulnerability factors identified in this study—namely, female gender, multiple depressive episodes during adolescence, high proportion of family members with recurrent depression, conflictual relationship with parents, and borderline personality traits—are all consistent with a conceptualization of depression that takes into account genetic, familial, personal, and environmental factors. As such, it argues for an equally broad and multifaceted treatment approach: one that includes not only pharmacotherapy but psychotherapy and family-based interventions as well. The high rates of comorbidity and the far-reaching consequences of adolescent depression further argue for a broad treatment approach that goes beyond the mere amelioration of circumscribed depressive symptoms.
As always, the study by Lewinsohn and colleagues has its own methodological limitations that should serve as a stimulus for further research. We note these problems mostly to indicate the great need for further elaboration of studies such as the important one in this issue. The limitations include the use of telephone interviews at the third time point, as opposed to the multiple-informant, face-to-face contacts used during the first two assessments. As convenient and useful as telephone interviews may be, they are likely to reduce the richness of data that can be obtained from more full, clinically rich data collecting. Second, compared to the investigation of risk factors, Lewinsohn and his colleagues did not provide much information about protective and resiliency variables. Third, there is relatively little discussion of suicidality and actual suicide attempts (or of completed suicides) in this cohort of adolescents, who were well within the range of high morbid risk for these outcomes. Fourth, the study appears to have omitted a domain of increasing salience in the study of depression—the powerful effect of early childhood exposure to abuse and neglect. Perhaps this domain could not be assessed as fully as one would like because of the legal and other problems that arise in querying about these issues, a real problem for the entire field. Finally, we are of course curious about whether there were any benefits from the interventions received by the treated 20% of this study group. What type of care did they receive, and did it make a difference? These shortcomings can set the stage for the next studies.
Dr. Lewinsohn and his colleagues have provided a remarkably useful description of the natural history and correlates of a major source of morbidity and death in adolescence and young adulthood. Further work along these lines will continue to lead toward a sharper definition of that window of opportunity that the timely identification and effective treatment of adolescent depression promises to be.
Address reprint requests to Dr. Martin, Child Study Center, Yale University School of Medicine, 230 South Frontage Rd., New Haven, CT 06520-7900; firstname.lastname@example.org (e-mail).