Abstract
Melancholia has always been on the minds of philosophers and healers. Hippocrates, Sufi mystics, Romantic poets, Buddhists, and Ayurvedic practitioners have described it as an inexplicable despondency, attributed variously to a loss of religious faith, overinvestment in materialism, or the dysregulation of internal somatic processes. In a way, the history of melancholy is the history of humankind’s struggle to describe and treat suffering. This tradition has not ended with the modern-day equivalent of melancholy, major depressive disorder.
Despite the diversity of explanations, melancholia was always understood as a complex process through which the psychological experience of sadness was materialized. Similarly, major depressive disorder is understood as a psychobiological disorder that includes a sad or depressed mood with somatic disturbances of appetite, energy, and sleep. However, we continue to struggle to understand why people become depressed, in the hope we might better heal them. Complicating matters is the fact that how one’s suffering is expressed varies significantly across race, culture, and class.
Expressions of depression are overdetermined by the personal meanings individuals attach to their symptoms. Does the personal context of depressive symptoms confound today’s standardized treatments for depression? And the questions of evolutionary psychiatry also cloud the picture: Is depression an appropriate response to external conditions that provoke psychic pain? Does this imply the need for social treatments as opposed to individualized ones?
The article under review speaks to some of these ancient questions. The study is unique in that it is the first attempt to compare the efficacy of short-term dynamic therapy (specifically, supportive-expressive treatment) to contemporary antidepressant therapy, as well as to a placebo control for the treatment of depression. Also notable is that the study population was recruited from a poor urban population with a large minority component (52%), unlike previous studies of antidepressants and psychotherapy conducted in white middle-class populations. Most study participants also suffered from chronic depression (61.5%). The study was conducted between 2001 and 2007.
One hundred fifty-six individuals met the criteria for inclusion in the study. Exclusion criteria were substance dependence, bipolar disorder, psychosis, any Axis I disorder judged to be more severe than depression, high suicide risk, medical contraindications to study medications, and illiteracy. Participants had to meet criteria for depression on two separate evaluations one week apart using the Structured Clinical Interview for DSM-IV or via a score < 14 on the Hamilton Depression Rating Scale.
Comorbid Axis II disorders and substance abuse were allowed in the study. Participants were randomized to one of three groups for sixteen weeks: 55 were assigned to active medication treatment with sertraline (to be switched to venlafaxine ER if nonresponsive to sertraline by the eighth week of the study), 51 to supportive-expressive psychotherapy, and 50 to pill placebo. Participants in psychotherapy met with their therapist twice weekly for four weeks and then weekly for weeks 5-16. Participants receiving medication or placebo met with their psychopharmacologist weekly for the first six weeks, and then could be seen every two weeks thereafter at the discretion of the psychopharmacologist. Thirty-five percent of the study subjects dropped out before completing the sixteen weeks. Both the participants and the researchers in the medication arm of the study were unaware of whether medication or placebo had been prescribed. Thus, while the medication arm of the study was truly placebo-controlled, the psychotherapy arm was not. (Clearly both patient and therapist were aware of what treatment was being given, and there is no placebo situation for psychotherapy.) As a result, psychotherapy could not effectively be compared to placebo as intended in this study; an “open” trial of psychotherapy would need to have been compared to an “open” trial of medication.
Overall, the study results demonstrated that supportive-expressive therapy was no more effective than medication treatment for depression, and neither was more effective than placebo. Rates of response were 30.9% for those receiving medication, 27.5% for those receiving psychotherapy, and 24.0% for those receiving placebo. These differences were not found to be statistically significant (p = .73). However, when analyzed by gender and minority status, statistically significant differences in responses to treatment were found. Minority men who improved in the study improved most rapidly with supportive-expressive psychotherapy. White men improved more rapidly in the placebo arm of the study. Minority women responded at comparable rates across all three treatment arms, and white women responded more quickly to psychotherapy or medication than to placebo. Socioeconomic factors (income and education) did not influence the findings.
The authors speculate that the high attrition rate, small sample size, severity of illness, and study population demographics could explain the demonstrated lack of efficacy of both antidepressants and psychotherapy. They also point out that previous clinical trials of antidepressants may have overestimated their efficacy for the treatment of depression, especially in minority patients, and that pharmacological intervention for depression might be of more limited efficacy than is generally recognized. However, it is important to note that since this study was conducted, new research methodology has emerged that demonstrates how the therapeutic setting of an antidepressant clinical trial can inflate placebo effect (Rutherford and Roose 2013). In the study under discussion, medication may well have been more effective than placebo had the number of patient visits in the medication arm of the study been more reflective of naturalistic medication management conditions.
From a cross-cultural perspective, one might also postulate that depression might have different “pathways” among various populations. Thus, its expression in biological and/or psychological terms would vary, possibly explaining the differential response to various forms of treatment in this study. It is also important to note the difficulty of treating depression in a population subjected to the ongoing stresses of financial and housing instability and violence, in addition to interpersonal conflicts, substance abuse, and comorbid medical conditions. Ultimately, is it possible to completely treat depression in this population without also addressing its social challenges?
This study opens a door to further questions, particularly to the idea of an ethnographic approach to treatment formulation—would a treatment be more effective if it were based first on an individual’s explanatory model of illness? Perhaps in this way, the efficacy of psychodynamic psychotherapies will be better established in the future.