Longitudinal relationships between childhood, adolescent, and adult eating disorders.

OBJECTIVE: This study investigates the longitudinal course of eating problems from childhood though adulthood.The following questions are answered: (1) How stable are eating disorder symptoms and diagnoses over a 17-year interval from childhood to adulthood? (2) Do early childhood eating problems predict the occurrence of eating disorders in adulthood? METHOD: An epidemiologically selected sample of approximately 800 children and their mothers received DSM-based structured psychiatric assessments in 1975, 1983, 1985, and 1992. The stability of full DSM diagnostic criteria for anorexia nervosa and bulimia nervosa, symptom scales derived from DSM criteria, and individual symptoms such as binge eating or dieting between early adolescence, late adolescence, and young adulthood was examined. RESULTS: Early adolescent bulimia nervosa is associated with a 9-fold increase in risk for late adolescent bulimia nervosa and a 20-fold increase in risk for adult bulimia nervosa. Late adolescent bulimia nervosa is associated with a 35-fold increase in risk for adult bulimia nervosa. Symptom scale scores for anorexia nervosa and bulimia nervosa correlate in the 0.3 to 0.5 range from early to late adolescence and young adulthood. For both anorexia nervosa and bulimia nervosa, gender, as well as eating symptoms at early and late adolescence, all predict young-adult eating disorder symptoms. Risk factors for the later development of eating disorders comprise eating conflicts, struggles with food, and unpleasant meals in early childhood. CONCLUSION: The presence of eating problems in early childhood or an eating disorder in adolescence confers a strong risk for an eating disorder in young adulthood.