The incidence and prevalence of admissions for melancholia in two cohorts (1875–1924 and 1995–2005)

Abstract

Background and method

There have been recent proposals to have melancholia reinstated in classification systems as a disorder distinct from major depressive disorder. Data from two epidemiologically complete cohorts of patients presenting to mental health services in North Wales between 1875–1924 and 1995–2005 have been used to map the features of melancholia.

Results

The data point to a decline in the contemporary incidence of hospital admissions for depressive psychosis, and greater heterogeneity among hospitalized severe non-psychotic depressions today. They indicate that historically untreated episodes of “melancholia” had a sudden onset, an average duration of less than 6 months and a lesser likelihood of relapse than severe depressive disorders have today.

Limitations

This is a study of the hospitalized illness rather than the natural illness and the relationship between illness and being hospitalized remains at present poorly understood.

Conclusions

These are the first data on the comparative incidence and natural history of melancholia in historical and contemporary samples. They point to the distinctiveness of the condition and support arguments for its separation from other mood disorders.

Introduction

Melancholia has been in continuous use as a psychiatric diagnosis for centuries. For much of this time it did not refer to a depressive disorder. The restriction to denote a distinctive and generally severe depressive disorder took place in the middle to later years of the 19th century (Healy, 2008).

In 1899, Kraepelin amalgamated melancholic depressions and what would now be called bipolar depression into manic-depressive illness. Episodes of this new disorder were characterized by their acute onset and remission, ordinarily within a period of months (Kraepelin, 1899).

In the second decade of the 20th century Schneider focussed attention on the major psychopathological features of melancholic states, such as diurnal variation of mood, psychomotor retardation, and lack of mood reactivity (Schneider, 1950). These were the hallmarks of what was termed vital depression in contrast to neurotic depression or depressive personality disorders. Schneider's emphasis on clinical features laid the basis for the operational criteria that emerged in the 1980s.

In the inter-war and post-war years, vital depression also termed endogenous depression was widely thought to have a sudden onset, an apparent lack of precipitants and a propensity to remit. It responded to electroconvulsive therapy (ECT) and tricyclic antidepressants when these were introduced.

In 1980, although included as a specifier for depression in DSM III, melancholia (vital depression) vanished as a distinct diagnosis as its operational criteria overlapped so heavily with those for major depressive disorder. There have been proposals since to have it reinstated given its distinct phenotype, its linkage to biological markers such as raised cortisol levels and a more specific response to treatments such as ECT and tricyclic antidepressants than is found for major depression disorder (Parker et al., 2010, Schotte et al., 1997).

Many of the claims made for the natural history of depression, including rates of recurrence and periodicity, as well as co-morbidities with physical disorders such as cancer or cardiovascular disease (Simon et al., 2007, Steptoe, 2007), and risks such as suicide stem from studies on melancholia or endogenous depression antedating 1980 and major depressive disorder. At the same time, little has been published on incidence rates of or mortality from melancholia as such epidemiological studies post-date 1980.

This paper seeks to compare a sample of melancholic patients from a century ago with a contemporary sample of severe (putatively melancholic) depressive disorders to map the incidence of the disorder, and its natural history.