At its most basic level, the discipline of psychiatric epidemiology is the study of the patterns of mental disorders, including how frequently disorders occur, how they are distributed in populations, and what are associated risk factors. Psychiatric epidemiology also defines the time course of mental disorders including their onset, duration and recurrence. Recently, the field has greatly expanded and now includes detailed examinations of the natural history of psychiatric disorders, genetic epidemiology, the relationships between physical and mental disorders, and studies of the use and outcomes of mental health treatments. This expansion has required significant advances and developments in psychiatric epidemiologic methods (Tsuang and Tohen, 2002).
Important characteristics that distinguish psychiatric epi-demiological research from other clinical investigations are the in-clusion of representative samples and the application of systematic methods for determining diagnosis or outcome. The specific type of sample and choice of mental health measure depend on the goal of the study. Three types of samples are generally used in epidemiology. For studies aimed at establishing prevalence and incidence rates, the population-based survey is the optimal method. Complex sampling procedures have been developed to ensure random selection for both single-stage and two-stage studies. For studies of rare disorders, iden-tified patients are usually ascertained from registries or a represent-ative set of psychiatric treatment facilities. However, because only a minority of individuals with diagnosable disorders are ever treated for psychiatric problems within the mental health care system (Regier et al., 1993; Kessler et al., 1994; Wang et al., 2000), these sources may omit true case patients who do not present for treatment.
The development of structured diagnostic interview sched-ules tailored to clear operationalized diagnostic criteria was the crucial element underlying the recent progress in psychiatric epi-demiology. As a result of the development of structured diagnostic interview schedules, the need to establish the prevalence of specific disorders was finally realized, at least within the limits of our current ability to operationalize mental disorders and within the constraints inherent in interview data (Fennig and Bromet, 1992). Estimates suggest that approximately 12% of children (Institute of Medicine, 1989) and 15% of adults (Regier et al., 1988) currently meet criteria for one or more mental disorders. More precise estimates will be possible as more sensitive diagnostic tools become available.