Winston Churchill’s description of his depression as the "black dog" is well known among psychiatrists, who frequently encounter the morbidity and mortality associated with this common disorder in their patients (1). The prevalence of lifetime depression in the United States is approximately 17% overall, and the rate in women is even higher (2). Untreated or poorly treated depression is generally believed to underlie the unacceptably high rate of suicide worldwide. Suicide is currently ranked as the eighth leading cause of death in the United States (3). As if the disturbed sleep, hopelessness, cognitive dysfunction, and appetite disturbance associated with depression were not difficult enough, a plethora of data has revealed that major depression is a significant risk factor for the development of coronary artery disease and stroke (4).
Despite the availability of a myriad of antidepressants and psychotherapies that have been shown to possess unequivocal efficacy in the treatment of depression (5), a large number of depressed patients remain ill, either because they receive no treatment or, if they are treated, exhibit only minimal or no improvement. Indeed, much less than half of patients with major depression treated with the commonly prescribed newer antidepressants attain remission (6), i.e., the presence of minimal or no symptoms and no social or occupational dysfunction.
Clearly, advances in the management of depression would be welcome. Many developments are needed, including methods of predicting patients’ response to particular pharmacological agents or to psychotherapy, ways of accelerating the rate of response to antidepressants, and strategies for achieving a more complete overall response. The papers on mood disorders in this issue of the Journal contribute to a better understanding of the treatment of depression.
Recent data have suggested that increased rapidity of response to antidepressants is associated not only with less suffering but a greater likelihood of remission (7). For these reasons, among others, acceleration of antidepressant response is of considerable interest. In this issue, Altshuler and colleagues report the results of a meta-analysis of studies that followed the seminal observations of Prange and colleagues more than 30 years ago that triiodothyronine (T3), one of the two major thyroid hormones, accelerates the response to tricyclic antidepressants. The meta-analysis was conducted to determine whether the studies provide sufficient evidence to support clinical strategies based on these observations. Six studies comprising 125 patients were included. The results revealed a highly significant effect of T3, especially in women. These findings are congruent with those of a meta-analysis by Aronson et al. (8) indicating that T3 is also effective in converting antidepressant nonresponders to responders. Unfortunately, both of the meta-analyses focused on only tricyclic antidepressants, which are no longer a first-line treatment choice. Whether T3 accelerates responses to the selective serotonin reuptake inhibitors (SSRIs), venlafaxine, mirtazapine, or nefazodone is of great clinical interest, but enough data on this question are not yet available. Moreover, as the authors note, the tricyclic antidepressant doses used in these studies were modest.
Another challenging area of research is the treatment of women with postpartum depression, a common disorder with a 10% prevalence after live childbirth. The use of antidepressants in breast-feeding women has been a particularly controversial issue because of concerns about potential adverse effects of the medication on the developing neonate. Comprehensive reviews (e.g., 9) have summarized evidence that the newer antidepressants, particularly the SSRIs, produce no adverse effects on the infants of breast-feeding mothers. In this issue of the Journal, Epperson and colleagues used an indirect measure of serotonin reuptake blockade—the reduction in platelet serotonin (5-HT) concentration—to determine if the breastfed infants of women treated with sertraline sustain significant SSRI effects. The women were treated with 25–200 mg/day of sertraline, but the vast majority received ≤100 mg/day. The validity of the measure of 5-HT reuptake blockade was demonstrated by the marked reduction in maternal platelet 5-HT content after sertraline treatment. No such effect was observed in the blood of the breastfed infants, suggesting that the concentrations of sertraline present in breast milk are insufficient to produce significant SSRI effects in the neonate. These findings expand on those of previous studies of SSRI excretion in breast milk (10, 11) and reassure the practitioner of the safety of SSRI use in the postpartum period in breast-feeding women. Because remission often requires higher doses than those received by the majority of the women in this study, subsequent studies should include women treated with doses of 150–200 mg/day of sertraline and analogous doses of the other SSRIs.
It is not uncommon for patients treated for depression to exhibit persistent symptoms even after 6–8 weeks of receiving an adequate dose of antidepressants. This problem is particularly serious in primary care, and several investigators have developed organized treatment programs to improve clinical outcomes. Such programs incur additional cost and, therefore, must demonstrate incremental benefit. In this issue, Simon and colleagues report the results of their third study of collaborative care for depression. Primary care patients who began antidepressant treatment completed a telephone assessment 6–8 weeks after the initial prescription. Patients with persistent major depression or significant depressive symptoms were randomly assigned to receive usual care or a collaborative care treatment program. The collaborative care arm included systematic patient education (a book and videotape), a consultation visit with a consulting psychiatrist, a 2–4-month period of shared care (alternating visits with a psychiatrist and primary care physician), and systematic monitoring of follow-up visits and adherence to medication regimens. Clinical outcomes were assessed by blinded raters in telephone assessments at 1, 3, and 6 months. Over the 6-month period, the collaborative care treatment programs were associated with a mean of 16.7 additional depression-free days, compared with usual care. The program incurred an additional mean expense of $357. These findings indicate that such interventions, at a modestly increased cost, produce considerable improvement in outcome.
Finally, in this issue, Seidman and colleagues report surprising findings on the use of sildenafil in the treatment of erectile dysfunction in men with depressive symptoms. The randomized, 12-week, double-blind, placebo-controlled trial included 152 men who fulfilled DSM-IV criteria for depressive disorder not otherwise specified, had erectile dysfunction for ≥6 months, and were treated with sildenafil or placebo. Remarkably, the subjects’ mean duration of erectile dysfunction was 5.7 years. Their mean Hamilton Depression Rating Scale score at baseline was 16.9. The results for 136 subjects with at least one postrandomization assessment showed that sildenafil was effective in treating sexual dysfunction in 48 patients and ineffective in 18; placebo was effective in 10 patients and ineffective in 60 patients. Remarkably, the sildenafil responders exhibited a mean reduction in their Hamilton depression scale score of 10.6, and 76% had a ≥50% decline in the score. The nonresponders exhibited a mean reduction of 2.3 in their Hamilton depression scale score. The authors concluded that sildenafil is effective in treating erectile dysfunction in men with mild to moderate depression and, most remarkably, that the improvement in erectile dysfunction is associated with a marked improvement in depressive symptoms. As the authors point out, there are many potential explanations for their observations, ranging from a direct antidepressant effect of sildenafil to the resolution of depressive symptoms "secondary" to erectile dysfunction.
The progress in the area of affective disorders in the last two decades is impressive. The range of advances spans both pathophysiology and treatment. The reports on treatment of affective disorders in this issue of the Journal are representative of this progress. Such advances bode well for both patients and our profession and also serve to remind us that the rigorous scientific methods of controlled clinical trials and appropriate statistical analysis are essential for informing the field.
Address reprint requests to Dr. Nemeroff, Department of Psychiatry and Behavioral Sciences, Emory University School of Medicine, 1639 Pierce Dr., Suite 4000, Atlanta, GA 30322; firstname.lastname@example.org (e-mail).