A two-factor model of disordered eating

General psychopathology (i.e., personality disturbance, mood/affect dysregulation) has been identified as common and perhaps etiologically important in eating disorder (ED) patients. In this context, we examined a two-factor model of disordered eating which implicated the independent and interactive contribution of (a) general psychopathology (personality) and (b) eating-related psychopathology (body esteem) in explaining deviant eating patterns. A sample of 266 female college/university students (M age=22.1) and 76 women with a clinically diagnosed ED (M age=28.09) completed paper and pencil questionnaires of deviant eating patterns, body esteem, and personality pathology. First, a K-means cluster procedure revealed a compelling three-cluster solution among the nonclinical women based on deviant eating variables: <>(n=61), <> (n=92), and <> (n=103). The ED women were classified as Cluster 4; <> (n=76). Second, to evaluate our two-factor model, multinomial logistic regression (MLR) was used with cluster membership as the outcome variable and the following variables as predictors: body esteem, personality pathology, and all two-way interactions. In addition to several significant main effects, three interaction terms were marginally significant: Body esteem appearance x Narcissism (P=.047), Body esteem weight x Narcissism (P=.044), and Body esteem attribution x Stimulus seeking(P=.051). The overall extent of correct cluster classification was 63%. These results indicate that the presence of both low body esteem and maladaptive personality, and the interactive operation of these two factors, seems to contribute to the likelihood of having an eating disorder, beyond the independent contribution of either factor alone. Results are considered in the context of etiological models in which general psychopathology presents a vulnerability factor in ED development.