Should the Diagnosis of Melancholia Be Revived?

Relevance for research assessment

Researchers already acknowledge different response rates for psychotic and nonpsychotic depressed patients and exclude those with psychotic features from clinical trials. Drug dependence complicates compliance, and concomitant use of medications for medical disorders risks mixing the systemic effects of medications with those of psychotropic agents. Patients with drug dependence and those who are taking drugs for medical illness are usually excluded from experimental trials (leaving clinicians who have to treat such populations with unstudied risks and an absence of guidelines). Stratifying populations in depression research studies by the presence of melancholia will ensure a more homogeneous population for evaluating treatment efficacy.

Relevance for clinical practice¹

Few studies specifically assess the efficacy of modern treatments for patients with melancholia, so we are limited to interpreting experiences before the introduction of DSM-III. When convulsive therapy was introduced in the1930s, its ease of use encouraged widespread testing. Within a few years, its efficacy in patients with manic-depressive illness, endogenous depression, and high suicide risk was recognized. The introduction of chlorpromazine(Drug information on chlorpromazine) and imipramine(Drug information on imipramine) followed, with both treatments being found effective for these patients. Lithium(Drug information on lithium), anxiolytics, anticon anticonvulsants, and the spate of SSRI antidepressants have been poorly assessed in melancholic patients, and we are therefore insecure in their use.¹

A useful algorithm starts with considering the possibility of melancholia and establishing a symptom baseline with a depression rating scale. A formal DST or a 4 PM cortisol level marks the severity of the illness, and repeated tests guide decisions as to remission and relapse. Treatment with a tricyclic antidepressant has better evidence for efficacy than treatment with more recently introduced medications1 the formal treatment algorithm for melancholia calls for dosages equivalent to 200 to 300 mg/d of imipramine, established as rapidly as can be tolerated.

In severely suicidal patients who require continuing observation or who are in such a state of inanition that their condition is considered to be life-threatening, ECT is the preferred treatment. It is also the treatment of choice when medications fail. Periodic monitoring of cortisol levels and use of rating scales offer clinicians a guide to the course of treatment and the schedule of continuation treatments.

Conclusion

DSM-III and DSM-IV diagnostic criteria are too imprecise to define homogeneous populations in treatment assessments. Discarding serum cortisol measurements served our patients and our science poorly. A parallel example is the DSM's consideration of catatonia, a syndrome that is often seen in patients with mood disorders or as a toxic response to medications. In the DSM, catatonia is defined only as a subtype of schizophrenia, a narrow view that has handicapped the recognition of its many forms and inhibited the development of effective therapeutics. This deficiency has recently been redressed in 3 useful texts that offer effective guidelines to diagnosis and treatment.^12-14

An appreciation of melancholia as the principal definable mood disorder offers a better guide to diagnosis and treatment than does DSM-identified “major depression.” It offers more homogeneous populations of depressive patients, avoiding the difficulties of treatment assessments in mixed populations, and ensuring better outcomes for more patients even with the limited means that are now at hand.