As we approach the millennium, dramatic progress in our understanding and in our treatment of mood disorders, among the oldest diseases known to mankind, is now a reality. Therapeutic strategies developed within the last 30 years now provide significantly better treatment outcomes for symptomatic recovery of acute episodes. Furthermore, strategies for prevention of subsequent episodes have now become a regular part of our clinical armamentarium. In short, we have begun to alter the prognosis of these disorders across the life span. Nevertheless, we need to be aware that mood disorders are currently considered a serious global health burden. By 2020 it is expected that depression will be the second most serious medical disease with respect to global disease burden. Epidemiological studies continue to point to the high prevalence of mood disorders across the life span and, just as important, the strong likelihood of recurrence.

The various iterations of DSM have led to improved diagnostic clarity of mood disorders and have helped define therapeutic options more precisely. These diagnostic advances have contributed to a general change in the ease of diagnosis, our ability to intervene and to develop appropriate adherence strategies, and follow-up with mood disorder patients. In fact, the treatment of mood disorders, once the province of psychiatrists, is now adequately managed in many cases by our primary care colleagues.

Part of our general success has come about over the last 30 years because of the introduction and success of several "generations" of antidepressant compounds. The tricyclic antidepressants have been largely superseded by the success of the so-called selective serotonin reuptake inhibitors (SSRIs), which, because of their ease of administration and side effect profile, have increased the number of depressed patients who receive active treatment. The introduction of these new compounds has also led to therapeutic success in various subtypes of recurrent mood disorders, including chronic depression and dysthymia. As evidenced in this issue, the work of Reynolds and his colleagues demonstrates the successful implementation of systematic drug administration coupled with a standardized psychotherapeutic intervention. This report on the management of bereavement-related mood disorder also describes the use of treatments to disaggregate specific symptom clusters and specific treatment responsivity.

While treatment is more available than in the past, the report by Oquendo et al. in this issue demonstrates that in practice our treatment strategy for depression appears to be inadequate. In this study both high-risk depressed patients, i.e., those with a history of suicide attempt, and a low-risk group received inadequate treatment. This pattern of insufficient dosage and duration, therefore, has not changed since comparable naturalistic studies appeared over a decade ago (1). The authors conclude that education for health professionals is a necessity, probably for both general practitioners and psychiatrists.

The essential tools for clinical psychiatry include appropriate diagnostic methods and improved clinical trial methodology. If we are to improve our understanding of pathogenesis, then we must take advantage of both the genetic and neuroimaging tools that are now available. The use of genetic strategies to increase our understanding of complex diseases and multifactorial inheritance leads us to conclude that models of suicidal behavior such as the one advocated by Mann et al. in this issue require a precise grasp of genetic vulnerability issues.

In a similar way, neuroimaging approaches allow us to understand in vivo brain functioning and offer unheralded and unique opportunities to study aspects of brain functioning and behavior. Two reports in this issue provide important clues in our search for relating cognition, affect, memory, and brain functioning. For example, Mori et al. used a creative method to show that atrophic abnormalities in the amygdala were associated with a failure to learn/remember powerful emotional experiences in patients with probable Alzheimer’s disease. They achieved this aim by showing a relation between amygdala volume determined with magnetic resonance imaging and emotional memory for events after the traumatic 1995 Kobe earthquake in Japan. Their data indicate that disturbances in emotional influences on memory processes were related to lower amygdala volume but not to low hippocampal volume or general cognitive status. These observations suggest that disease-related changes in the amygdala may be a cause of a particular form of memory disturbance in Alzheimer’s disease affecting the ability of emotional salience to facilitate memory processes.

The study by Teasdale et al. is an important step forward in addressing questions about the neural substrate of higher-order cognitive controls of emotions. While their study group was small (N=6) and more studies are needed to validate their conclusions, the strategy of using verbal cues to define the context and affective significance of pictures was a strategy for identifying the sources of cognitive inputs that modulate immediate emotional experience. Their findings indicate that mesial prefrontal cortical areas, including medial superior frontal gyrus (area 9) and anterior cingulate cortex (areas 24 and 32), may be important sources of cortical inputs that modulate emotional reactions. It is important that, through similar mechanisms, these regions might play a role in regulating general mood state as well. The combination of creative psychological paradigms and sophisticated brain imaging methods will likely prove to be a promising strategy for gaining a clearer understanding of the basic brain mechanisms underlying emotional experience and of their pathology in psychiatric disorders.

While we can feel both proud and excited about the ongoing advances in mood disorder research and its application to treatment strategies, we cannot afford complacency. Certainly, a number of important questions remain unanswered. For example, we still do not know the true mechanism of action of antidepressant treatments, nor whether there are common neurochemical pathways independent of the specific treatment modality. While treatment response may approach 70% or sometimes higher, failure to achieve full recovery and, therefore, greater protection against future episodes, is still common. Given the estimation of the global burden of depression by 2020, considerable effort must be directed toward this problem as soon as possible. With respect to other treatment issues, bipolar depression continues to remain an enigma. Finally, our inadequate understanding of suicidal behavior and the likelihood of future suicidal behavior is a critical challenge. Also in this issue, Leon and colleagues, using the large National Institute of Mental Health Collaborative Depression Study data set, point out that there is no support for the speculation that intervention itself (e.g., using fluoxetine) will increase the risk of suicide. In fact, treatment generally decreases the risk. Mann and colleagues review their own data on suicide attempts and develop a model of suicidal behavior that can provide a template for further biological and behavioral research. I would only recommend that other important risk paradigms, such as those advocated by Kraemer et al. 2 years ago (2), be included in this conceptualization.

The questions that we asked about mood disorder 25–30 years ago have not been completely answered, but enormous progress has occurred, for which we as psychiatrists can feel justifiably proud. Our tools for answering these questions have improved dramatically, our treatment options have increased, and the success of treatment is being reflected in the greater number of depressed individuals seeking and receiving treatment. Nevertheless, the challenge of understanding the mechanism of action of our treatments, the underlying biologic pathogenesis of these disorders, and its relationship to the environment remain unsolved puzzles. Addressing this challenge is our task during future decades.

Address reprint requests to Dr. Kupfer, Department of Psychiatry, University of Pittsburgh School of Medicine, Western Psychiatric Institute and Clinic, 3811 O’Hara St., Pittsburgh, PA 15213; kupferdj@msx.upmc.edu (e-mail).