Longitudinal designs, methods and analysis in psychiatric research

Abstract

Objective: To outline the strengths and limitations of longitudinal research designs in psychiatry, and to describe different types of longitudinal designs and methods for analyzing longitudinal data. Method: Key references on longitudinal methods were reviewed and examples drawn from literature in psychiatry and psychology. Results: Longitudinal studies provide important information regarding the incidence and developmental trajectories of mental disorders. They allow for identification of risk factors and developmental concomitants. Recent developments in statistical methods for analyzing longitudinal data provide efficient estimates of change and predictors of change over time, identification and characteristics of distinct subgroups defined by change pattern, and improved methods for obtaining unbiased population estimates when data are incomplete. Conclusion: Longitudinal designs, methods and analysis can contribute to psychiatric studies on risk factors for common mental disorders, studies of early intervention and prevention and treatment outcomes.