Eating disorders, a public health problem because of their chronicity and impact on health, typically begin during adolescence. The goal of this study is to identify potential risk and protective factors in adolescents, that may be useful in screening for typical and atypical eating disorders in pediatric practice, and that may generate hypotheses about the etiology of these disorders. A secondary analysis of a two-stage epidemiological study of a countywide high school population (n=5596) completed in 1984-1985, is proposed. The first stage consisted of a survey questionnaire, which provided dimensional assessments of eating attitudes, behaviors, other psychopathology, as well as weight history, medical problems, and menstrual history. The second stage consisted of semi-structured diagnostic interviews, conducted by experienced clinicians, which yielded psychiatric diagnoses for eating disorders, other selected disorders, as well as more detailed information on correlates of eating disorders. The primary data analysis focused on the prevalence of psychiatric symptoms and disorders in adolescents, and data regarding the relation of potential risk factors to eating disorders was explored only in a preliminary way. Given advances in the nosology of eating disorders and in statistical methods, a second look at this data set is important and timely. In a series of analyses which will include adequate sophisticated statistical techniques, we propose to accomplish the following specific aims: (1) to determine, in boys and girls separately, the relation of potential risk and protective factors for Anorexia Nervosa (AN) and Bulimia Nervosa (BN); factors to be examined include developmental, dieting, weight, health, other psychopathology, family, and demographic factors; (2) to determine if DSM IV eating syndromes are distinct from AN and BN, such as Binge Eating Disorder can be identified in this population and, if so, to determine potential risk factors for these syndromes; (3) to develop efficient screening protocols for eating syndromes, which will exploit not only the cardinal features of each syndrome but also potential risk factors identified above. The strengths of this application lie in the application of both recent advances in the nosology of eating disorders and sophisticated statistical techniques applied to an existing data set. This unique epidemiological dataset, although cross-sectional, contains unparalleled information on the correlates of eating pathology. Findings from this study will contribute to the description and detection of eating disorders in adolescents, will identify potential risk and protective factors for eating pathology and will help generate hypotheses about the etiology of adolescent eating disorders.
