The epidemiology of child and adolescent psychiatric disorders: recent developments and issues.

Before 1960, few systematic surveys of behavioural and emotional problems in children were available in the literature. Most of psychopathological data therefore derived from the observation of clinical samples which are known to be sensitive to referral and other selection biases. As in many other fields of medical research, it was therefore imperative to generate knowledge about the rates, patterns and correlates of behavioural problems in non-referred children in order to derive a proper understanding of psychopathology. Epidemiology is the science which studies the repartition of disorders in human populations and the factors which influence this distribution. Epidemiological methods and designs were employed by psychiatric researchers in the 60s and 70s and, progressively, a specific methodology for use in child psychiatric surveys was developed. The Isle of Wight surveys conducted in Britain in the midand late-60s (Rutter et al., 1970; Rutter, 1989) represent a major development in child psychiatric epidemiology. The particular features of this survey were: a large sample size and a high participation rate; reliance on a two-stage survey design with a first screening stage based on questionnaires and a second interview stage of random samples of screened positive and negative children; the use of standardised diagnostic interviews of known reliability and validity; the use of impairment criteria to define psychiatric disorders; the combination of crosssectional design with longitudinal research strategies; and an examination of a wide range of biological and psychosocial risk factors (Rutter, 1989). Since then, numerous psychiatric surveys have been un-