Twin studies and the etiology of eating disorders

Objective: To evaluate the major twin studies of eating disorders in order to consider the empirical standing of the suggestion that these disorders, notably bulimia nervosa, are largely genetically determined. Method: Examination of the relevant twin studies and the methods used. Results: The findings of the twin studies are inconsistent and difficult to interpret, with estimates for the heritability of liability to bulimia nervosa ranging from 0% to 83%, and from 0% to 70% for anorexia nervosa. Methodological reasons for this variability include issues of definition of phenotype, diagnostic reliability violation of the equal environments assumption, and small sample sizes. Many of these issues also apply to twin studies of other psychiatric disorders. Conclusions: The heritability of eating disorders remains unknown. A broad view on the etiology of eating disorders should be maintained, with a focus on environmental mechanisms and gene-environment interactions, as well as continuing genetic studies. (C) 1999 by John Wiley & Sons, Inc.