Estimating psychiatric morbidity by logistic regression: application to post-natal depression in a community sample

The use of logistic regression to estimate the prevalence of psychiatric morbidity in community samples is illustrated here with data from a study of post-natal depression in 702 primiparous Cambridge mothers. The method is also used to validate the primary screening instrument (in this case the Edinburgh Postnatal Depression Scale--EPDS), and to estimate the effects of cofactors, such as maternal age, previous psychiatric history and social class.