Research in pregnancy and mental illness: testing old wives' hypotheses

Methodological challenges encountered in evaluating the relationships between life stress, mental illness and pregnancy outcome are identified and several studies on the relationships between pregnancy and childbirth and psychosocial factors, including life stresses, major psychiatric disorders, and puerperal depression, are reviewed. Certain methodological pitfalls are illustrated by showing how relationships between psychiatric diagnosis, severity of psychiatric symptomatology, life stresses, strains specific to the pregnancy, amount of intervention received and outcome of keeping or losing the infant have been explored. Five stages of research are identified and particular difficulties encountered at each stage are described with application to the authors' longitudinal study.