Objective

Use classification tree analysis with lagged predictors to determine empirically derived cut-points for identifying adolescent girls at risk for future onset of threshold, subthreshold, and partial eating disorders and test for interactions between risk factors that may implicate qualitatively distinct risk pathways.

Method

Data were drawn from a prospective study of 496 adolescent girls who completed diagnostic interviews and surveys annually for 8 years.

Results

Body dissatisfaction emerged as the most potent predictor; adolescent girls in the upper 24% of body dissatisfaction showed a 4.0-fold increased incidence of eating disorder onset (24% vs. 6%). Among participants in the high body dissatisfaction branch, those in the upper 32% of depressive symptoms showed a 2.9-fold increased incidence of onset (43% vs. 15%). Among participants in the low body dissatisfaction branch, those in the upper 12% of dieting showed a 3.6-fold increased incidence onset (18% vs. 5%).

Conclusion

This three-way interaction suggests a body dissatisfaction pathway to eating disorder onset that is amplified by depressive symptoms, as well as a pathway characterized by self-reported dieting among young women who are more satisfied with their bodies. It may be possible to increase the effectiveness of prevention programs by targeting each of these qualitatively distinct risk groups, rather than only individuals with a single risk factor.