Disturbed Eating and Body Dysmorphic Symptoms in a Young Adult Sample Are Separable Constructs That Each Show a Mixture of Distributions

Abstract

Eating disorder and body dysmorphic disorder symptoms overlap and frequently co-occur clinically, yet whether they represent one or more underlying constructs in the general population is unknown. We examined relationships between these symptoms on underlying factor structures and dimensional distributions in a young adult sample of 328 students using the Eating Disorder Examination-Questionnaire (EDE-Q 6.0) and the Dysmorphic Concern Questionnaire (DCQ). We performed factor and hierarchical cluster analyses on pooled items and Gaussian mixture modeling on score distributions. EDE-Q 6.0 and DCQ total scores were correlated (r = 0.53, p < .001). Pooled items demonstrated a three-factor solution; DCQ items separating from two EDE-Q 6.0 factors. Hierarchical clustering yielded a two-cluster solution that separated the two scales. Mixture modeling demonstrated that more than one underlying distribution best fit the data for each scale. These results suggest that the EDE-Q 6.0 and DCQ measure different sets of psychopathological features, despite their tendency to track together. Moreover, eating disorder and body dysmorphic phenotypes each show nonuniform variation from normal to abnormal. This argues against using linear dimensional applications of these scales to assess individuals ranging from mild to severe in symptom severity. Separate scales may be necessary to characterize lower and higher ranges of clinical severity.