Many important questions about depression in young people can only be answered from general population studies (as opposed to studies of clinical samples). First, given that only a small minority of disturbed children are ever referred for psychiatric treatment (Costello et al., 1993), estimates of the rates of depression in children and adolescents cannot be determined from clinical data. Second, although clinical studies have often provided important leads to be followed up in epidemiological studies, the undeniable presence of referral biases (Berkson, 1946; Costello & Janiszewski, 1990; Cohen & Hesselbart, 1992; Cohen & Hesselbart, 1993; Goodman et al., 1997; Angold et al., 1998b) vitiates their use in describing patterns of diagnostic comorbidity or the sizes of impact of risk factors, or the level of need for services. Epidemiological studies are, therefore, important from both the administrative point of view (in determining needs for service provision or preventive interventions) and from the perspective of aetiological research.

Prevalence of depressive disorders in children and adolescents

Unipolar disorders

Table 6.1 presents prevalence estimates for unipolar depression from a number of general population studies that used the DSM diagnostic system (American Psychiatric Association, 1980, 1987, 1994). We have not included studies that used the ICD system because the existence of the categories of mixed disorders of conduct and emotions and depressive conduct disorder means that overall rates of depressive disorders are not usually ascertainable from reports from such studies. Where data are available from multiple waves of a longitudinal study, figures are reported for each wave separately.