Has the term “depression” reached the end of its usefulness as a clinical descriptor in physical illness states? The reviews in these three small volumes create the uneasy sense that perhaps it has, especially in the settings of cancer, heart disease, and diabetes, respectively.
None of the reviewers formulate, let alone answer, at least one key question that I hear regularly from my oncologist colleagues. What part of the variation does depression account for in the outcome of cancer morbidity and survival? Perhaps this is because the term depression itself is so diffuse and poorly controlled in clinical studies that reasonable quantization cannot be conducted. This is probably true in heart disease and diabetes as well. The suggestion is that our current concepts of depression are inadequate to the task.
Depression means many different things in the world of general clinical psychiatry, exclusive of physical disorders. These can range from major depressive disorder, through the downward swings of bipolar disease, to the vacuous formulations of depressive dysthymia and even to depressive character diagnosis when people approach human experience as one long depression. Subgroups and subtypes deserve our attention because their natural history may be very different. None of the three volumes reviewed focus a sharp light on this issue.
Does antidepressant medication treatment work in the depression seen in cancer, heart disease, and diabetes? The answer is a guarded “possibly” according to the reviews here. Only Katon and van der Feltz-Cornelis, in their review of depression in diabetes, attempt cross-study comparisons using Cohen's d statistic, an assessment of the probability of effect size. While this is only one ingredient in a critical assessment of clinical outcome trials, it offers a useful tool in comparing effects reported in different studies. The conclusion with respect to diabetes is a mild to moderate effect across antidepressant medication studies. However, it is not clear which subgroups might account for this. The authors of the reviews in the volumes on cancer and heart disease do not present effect size comparisons.
The three books list their lead authors as Dr. Glassman, who is well-known in heart disease research, Dr. Katon, who is best known for work in depression treatment in primary care, and Dr. Kissane, who is from the Sloan-Kettering Cancer Center. In fact, each one is a co-author of one chapter of each book. The majority of the research appears to have been done by the two European co-authors who are listed on the three respective title pages.
The summative effect of the three volumes, taken as reviews in the field of psychosomatic medicine—that most unfortunate of terms—points toward the need to view the phenomenon of mental and emotional responses to physical illness, such as cancer, heart disease, and diabetes, in new and more effective terms. For example, to question the standard wisdom of depression in oncology patients, my research group compared the survival differences between depressive symptoms and ego adaptive styles in a sample of advanced-stage cancer patients (
1). The maturity of adaptive style predicted 5-year survival far more dramatically than the frequency of depressive symptoms. Do we need a paradigm shift in our thinking with respect to mental and emotional functions and physical illness? Both the reviews in these three volumes and our own explorations contribute to saying yes. Will this paradigm shift occur from the subspecialty of psychosomatic medicine, the field that lately seems to offer a final, although unconvincing, common pathway for all of human misery via the word depression? Probably not, until this subspecialty finds its way. Will it occur in the pages of the
Journal or the
Archives of General Psychiatry? That too raises considerable doubt in the new age of medical journalism where novel high-risk/high-gain formulations may lower an impact factor. Does it need to be investigated anyway in the interests of better diagnosis and treatment? Without a doubt.