A factor mixture model for multivariate survival data: An application to the analysis of lifetime mental disorders

Abstract

The assessment of the lifetime prevalence of mental disorders under comorbidity conditions is an important area in mental health research. Because information on lifetime disorders is usually gathered retrospectively within cross-sectional studies, the information is necessarily right censored and this should be taken into account when setting up models for the estimation of lifetime prevalences. We propose a factor analytic discrete time survival model combining mixture item response theory and discrete time hazard functions to describe disorder associations while accounting for censoring. This model is used for describing the lifetime prevalence and comorbidity of eight depression and anxiety disorders from the European Study of the Epidemiology of Mental Disorders. © 2013 Royal Statistical Society.