[Epidemiological study of maternal depression as a risk factor in the developing of early childhood psychosis]. / Etude épidémiologique sur la dépression maternelle comme facteur de risque dans la survenue d'une psychose infantile précoce.

ABSTRACT

After reviewing epidemiological literature on the relationships between depression in parents and onset of a psychiatric disturbance in children as well as investigations of the interactions between a depressed mother and her infant, the authors discuss the psychoanalytical concepts of these interactions and the findings of their research. This epidemiological research is based on the assumption that depression in the mother during pregnancy and the first few months of the post-partum constitutes a risk factor for the onset of an early psychosis in the child. This research is a comparative study between a group of mothers with early onset psychotic children and a group of mothers with non patient children, the age and sex of the children in both groups being equally distributed. The methodology includes two instruments aimed at a retrospective assessment of mothers' depression the SADS-LA questionnaire and a standardized scoring system for semi-structured interviews investigating the mother's feelings during pregnancy and early development of her baby. Results show that there is a statistically significant relationship between a major depressive condition in the mother during pregnancy and/or during the first year of the infant's life and the onset of an autism in the child. It might be that major depressive conditions starting before delivery could by themselves account for the risk. Depression in the mother therefore constitutes a risk factor for early psychosis, the relative risk being about four. This study also emphasizes particular features of the mother's depressive conditions difficulties to accept the real child, difficulties in perceiving the infant's psychic evolution, decrease of interactive skills. The statistically significant relationship in no way points to a linear causal relationship, which hypothesis seems to be negated by the statistical findings of this research.