Tracking the Course of Eating Disorder Symptoms

Over time, the course of an eating disorder is often chronic and characterized by substantial changes in symptoms. Remission, relapse, and crossover data suggest that 20% to 50% of individuals with an initial diagnosis of anorexia nervosa (AN), for example, eventually develop bulimia nervosa (BN), and the crossover rate between subtypes of AN may be even higher than this (J Nerv Ment Dis 1997; 185:704; Int J Eat Disord 2002; 31:191).

In what is believed to be the first study of the individual pathways of multiple eating disorders symptoms, Dr. David Herzog and co-workers attempted to follow the course of three core eating disorders symptoms: low body weight, binge eating, and purging in women with AN and bulimia nervosa (BN) (Int J Eat Disord 2011;679). The researchers followed 246 treatment-seeking women with AN (n=136) or BN (n=110) with the Eating Disorders Longitudinal interval Follow-up Evaluation Interview every 6 months, and collected weekly eating disorder symptom data for a 5-year period. (The women were followed for a median of 9 years and more than 90% of the women were followed for 5 consecutive years.) A novel statistical method, semiparametric mixture modeling, was used to identify longitudinal trajectories for the three core symptoms.

Instability and fluctuation of symptoms

Four pathways were identified for each of the three symptoms: low body weight (non-low-weight, fluctuating weight, gaining from low weight, and persistent low weight); binge eating (non-binge eating, early decreasing binge eating, late decreasing binge eating, and persistent binge eating; and purging (non-purging, increasing purging, steadily decreasing purging, and persistent purging).

In some cases, the study results highlighted both the stability of eating disorder symptoms, and in others the results underscored the tendency of symptoms to fluctuate over time. For example, most persons who followed the route of “persistent low weight” also followed “non-binge eating” and “non-purging” pathways.

The authors noted that examining the unique paths of “increasing purging” and “fluctuating weight” may be particularly helpful for illustrating the phenomenon of diagnostic crossover as well as the reality of symptom fluctuation and periodic remission. As an example, the “increasing purging” trajectory may include both individuals whose purging behaviors accelerated over time as well as those who showed purging behavior at baseline and later experienced a period of remission.

The study results also had several clinical implications, according to the authors. If future research could identify clinical predictors of individual interventions might be developed or tailored to help patients with eating disorders move toward a route of decreasing symptoms. For example, learning what differentiates patients with symptoms that remit early might result in novel treatment strategies for those individuals with persistent symptoms that are more resistant to treatment.

The authors suggest that future studies might attempt to determine if the course of eating disorders among women who are not seeking treatment shows similar pathways to those they identified in their study. And, since the current study was limited to participants with AN or BN, the findings cannot be extended to patients with subthreshold eating disorders or eating disorders not otherwise specified, including binge eating disorder and purging disorder.