[Influence of depressive history on biological parameters in major depression]

Major depressive disorder is associated with several biological abnormalities. Among them, sleep disturbances and alterations of hypothalamic structures and cortisol system have been largely studied. Most studies have found a relationship between depression and alteration of sleep. Electroencephalic sleep profiles during depression demonstrate abnormalities of sleep continuity, reduction of slow waves sleep and REM pressure (27). The cortisol system, investigated by the Dexamethasone Suppression Test (DST), is abnormal in about an half of the depressed subjects. We confirm a cortisol escape from suppression by dexamethasone (21). A complex dysregulation of the Hypothalamic Pituitary Adrenal axis (HPA) is thought to explain this escape (15, 33). The HPA has been first involved in the theory of stress. There are two ways to study this. First by looking at early adversities and genetic susceptibility to stress, and second by studying acute stressors and depressive reactions (8). The sensitization model postulated that the acute abnormalities of depression may leave biological scares. Those scares could make people more vulnerable to latter depressive triggers (34). We could then suppose that biological correlates of depression become more severe during the course of the illness. The present study further examines relationships between DST, polysomnography and some clinical and epidemiological characteristics of the depressive illness. We tried to examine if there were increasing biological disturbances during the course of the illness. We also examined the effect of the history of illness on the psychosocial stressors, and the effects of those stressors on the biological correlates of depression.

The DST and polysomnographic recordings were performed in a sample of 130 inpatients with primary major depressive disorder, unipolar form, as defined with the RDC (41). All of those patients have been consecutively admitted to the Sleep Laboratory of the Department of Psychiatry, Erasme Hospital, between 1981 and 1992. This population has been previously reported elsewhere (19, 20, 21). The depressive symptom severity was assessed with the 24-items HRSD (17). The Newcastle Endogenous Depression Diagnostic Index was used to assess the endogenous character of the depressive episode (5). The history of the illness and the impact of psychosocial stressors were assessed retrospectively using the interview associated with the RDC (41), the SADS (7). Psychosocial stressor has been assessed using the item 216 of the SADS. We must remind that it is not a precise measurement of stress.

Table I shows clinical characteristics and biological variables in our 130 depressed patients. No correlation were found between number of depressive episodes and DST or EEG sleep records (table II). Age of onset was correlated with DST and all sleep EEG parameters (REM latency, REM density, awakening and slow wave sleep). But the age was a major confounding factor. When corrected the results for age, a significant correlation between age of onset and DST still remained, but no correlation between age of onset and EEG sleep results (table III). The psychosocial stressors were correlated only with awakening. A positive trend was found between an augmentation of psychosocial stressors and the number of episodes (p = 0.06).

This study does not support the view that the biological correlates of depression are worsening with the course of the illness. We found only correlation between age of onset and DST, but a possible confounding effect of age cannot formerly be excluded. The impact of psychosocial stressors on the biological correlates of depression was minimal. The only significant correlation found was between awakening and psychosocial stressors. We found no correlation between psychosocial stressors and the course of depression. Those results do not support the view of sensitization of the illness, but it should be remember that evaluation of psychosocial stressors by item 216 of the SADS was probably not a sufficient sensitive measure. We suggest thus that the impact of the history of depression on biological correlates of depression is not very strong. An alternative explanation of the lack of correlation found is that we used inaccurate measurement of the course of depression. For example, we had no evaluation of the quality of remission between different episodes of depression and of subsyndromic depression. Recent works show the importance of this (6, 9). The major limitations of this work were the retrospective character of the study and the low precision of the evaluation of psychosocial stressors used.