Introduction

The longitudinal perspective

The course of a disorder is examined by data collection at a minimum of two points in time, but an assessment on several occasions is required to identify a fluctuating course; if only two sets of data are available, their comparison is merely an outcome study. Many variables affect the course of PDs, including various interventions, other co-occurring disorders and secondary effects, such as a break-up of family relationships or the loss of employment.

The evaluation of the course of PDs is adversely affected by the complexity of PD assessment, so that even if a detailed evaluation has been carried out at one time-point, it may be difficult to obtain the subject's cooperation for further time-consuming interviews. Consequently, most longitudinal studies of PD have relied on relatively crude measures such as re-admission rates and suicide. Also, a high drop-out rate is a problem in any longitudinal study and, in relation to PD, has been associated with males and a diagnosis of antisocial PD.

Most PD traits appear to have a dimensional relationship with aspects of ‘normal’ personality in non-clinical populations, and it has been shown that several such traits show stability for the decades over the age of 30 (Zuckerman, 1991).