Facteurs psychosociaux associés à des épisodes dépressifs avec hospitalisation chez des femmes : étude cas-témoins

Abstract

The part played by psychosocial factors has frequently been studied in mental disorders whether as protective factors or as vulnerability factors, using variously adequate methods. A large body of research has shown that poor social support or poor self-esteem or presence of stressful life events could play a large part in triggering disorders. The importance of socioeconomic factors in mental illness is so great that such factors (unemployment, insecurity in employment, homelessness, lower social classes, low income) skew the studies in which they are not considered. In this study, in order to take into account these methodological problems, the study of psychosocial factors was undertaken in a standardized clinical manner and on a relatively socially privileged population. Two homogeneous samples for several variables, using a case-control approach have been formed. A group of hospitalized women for "neurotic" depressive disorders, aged 30-50 (n=59) was systematically recruited in a psychiatric hospital located in the Paris area depending on MGEN (Mutuelle Générale de l'Education Nationale). The control group (n=76) was recruited among the 75 000 individuals in the Paris area registered as members of MGEN. A large group of women received a physical and mental health questionnaire for initial screening, the CIDIS (Composite Diagnostic Interview Simplified), by post. Among the 395 women that did not show mental disorders, a group of 90 was examined a second time using a more discriminating tool: the SCAN (Schedule for Clinical Assessment in Neuropsychiatry). In fine, the control group was based on 76 women that did not show and had never shown any mental disorders. To assess neurotic mental disorders in a clinical standardized manner, the SCAN (Schedule for Clinical Assessment in Neuropsychiatry) was used for inpatients. Scores were processed by CATEGO software which enables subjects to be classified according to the ICD-10 system. Events and difficulties experienced by subjects were recorded using the LEDS (Life Events and Difficulties Schedule). The clinicians that interviewed subjects and collected date were trained beforehand by Harris. The Brown and Haris methodology was used to rate subjects' responses and to classify events and difficulties. To assess and measure self-esteem and social support two check-lists elaborated and implemented by Pearlin were used. Means were compared using Student t test and frequencies using the c2 method (Yates'correction or Fisher exact test when necessary) to analyse independent associated factors. A multivariate logistic analysis was performed to identify significant variables. The association between factors and mental pathology studied was expressed with an odds ratio (OR) with the 95% confidence interval. Compared to the control group, the hospitalized patients reported higher levels of exposure to six factors: practising an intermediate profession (p=0.051), living alone (OR=4, 38); low self-esteem (OR=40, 96); low social support (OR=6, 46); having experienced at least one severe event (OR=2,45), at least one difficulty lasting 6 Months or more (OR=25, 57) and at least one provoking agent (a severe event or a major difficulty) (OR=3,49). These six variables were considered as potential associated factors to "neurotic depressive disorders" and thus entered into a logistic regression analysis. From these six variables, four may be considered as psychosocial associated factors in "neurotic depressive disorders": poor self-esteem is the highest risk factor (OR=71,43), and having experienced at least one difficulty (OR=15,75). Having poor social support and having experienced one or more "provoking agents" (one severe event or one major difficulty lasting 24 Months or more) correspond to approximatively OR=3. Following this study, four psychosocial associated factors in depressive episodes can be considered as being risk factors for "neurotic depressive disorders". In the literature psychosocial factors are frequently considered to be factors that possess a certain independence. This idea is discussed in the full article.