Reciprocal relationships between personality disorders and eating disorders in a prospective 17-year follow-up study

Abstract

<p><b>Objective</b>

This study aimed to report the presence of categorical and dimensional personality disorders (PD) in adults with longstanding eating disorders (ED) over a period of 17 years and to investigate whether changes in PD predict changes in ED symptoms or vice versa.

<p><b>Methods</b>

In total, 62 of the 80 living patients (78% response rate) with anorexia nervosa (<i>n</i> = 23), bulimia nervosa (n = 25), or other specified feeding or ED (<i>n</i> = 14) at baseline were evaluated during hospital treatment and at 1-year, 2-year, 5-year, and 17-year follow-up. PD were assessed using the Structured Clinical Interview for DSM-IV Axis II disorders, and the eating disorder examination (EDE) interview was used to assess ED. Data were analyzed using multilevel modeling.

<p><b>Results</b>

From baseline to the 17-year follow-up, the number of patients with any PD decreased significantly from 74.2% to 24.2%, and the total number of PD diagnoses declined from 80 to 22. Mean EDE score was significantly reduced from 4.2 (SD: 1.1) to 2.0 (SD: 1.6). There was a positive association between ED and PD where the initial level of either disorder was followed by similar levels of the other disorder throughout the entire follow-up period. High baseline levels of borderline PD predicted less decrease in ED symptoms. No significant within-person effects were found.

<p><b>Conclusions</b>

Both ED and PD significantly declined over time. As the severity of either disorder seems to be associated with the other, thorough assessment and treatment that incorporates both the ED psychopathology and the personality disturbances are advisable.

<p><b>Public Significance Statement</b>

While personality disorders were highly prevalent in the sample of patients with longstanding eating disorders, both disorders were significantly reduced at the 17-year follow-up. The disorders are related in the sense that an initial high level of either disorder is associated with a high level of the other over time. A thorough assessment and attention to both illnesses are advisable in therapy.

<p><b>Clinical Trial Identifier</b>

NCT03968705.