Using network analysis to personalize treatment for individuals with co-occurring restrictive eating disorders and suicidality: a proof-of-concept study.

Abstract

Background: Nomothetic (i.e., on average) eating disorder interventions generally provide insufficient guidance for managing suicidality. The present proof-of-concept study demonstrates how idiographic network models can be used to inform a modular, highly personalized approach to treatment for individuals experiencing suicidality in the context of Anorexia Nervosa spectrum disorders (ANSD).

Methods: Using 21 days of ecological momentary assessment data (105 assessment points), contemporaneous and temporal idiographic symptom networks were generated for three patients with unique clinical presentations of ANSD. For each patient, we identify the most central symptoms in their network, as well as potentially important bridge symptoms linking eating pathology and suicidality. We then provide guidelines for using this information to guide the delivery of evidence-based intervention strategies.

Results: Intervention strategies may vary substantially depending upon which network statistics are used to guide treatment target selection. Bridge symptoms, or symptoms that serve as links between eating pathology and suicidality, may represent particularly promising intervention targets for individuals experiencing these conditions concurrently. Interventions which target the symptoms with the highest strength centrality may also yield symptom improvement throughout the entire network.

Conclusions: Although the viability of network-informed, personalized treatment is contingent upon continued intervention development research, this approach has the potential to improve treatment outcomes for individuals with co-occurring eating disorders and suicidality.