A longitudinal examination of the relationship between eating disorder symptoms and suicidal ideation

Cross-sectional research demonstrates significant correlations between eating disorders (EDs) and suicidal thoughts and behaviors. Although suicide ideation (SI) is a risk factor for suicidal behavior, longitudinal research investigating SI among EDs is limited. Thus, the aim of this study was to offer insight into the dynamic relationship between EDs and SI by investigating if these variables predicted one another at weekly time points.

Autoregressive cross-lagged modeling was used to test bidirectional relationships between ED symptoms and suicidal ideation among an ED patient sample (n = 92). Participants completed a measure of suicidal ideation and the Eating Disorder Examination-Questionnaire (EDE-Q) weekly for 5 weeks.

SI and ED symptoms were correlated with each other at each time point. Unexpectedly, the majority of cross-lagged pathways were nonsignificant. However, SI at Week 4 predicted ED symptoms at Week 5, while controlling for Week 4 ED symptoms. This pattern of results was found when the shape concerns, weight concerns, and eating concerns subscales of the EDE-Q were entered into the model. Moreover, Week 2 shape concerns predicted Week 3 SI and Week 3 eating concerns predicted Week 4 SI. No significant cross-lagged pathways were found with the dietary restraint subscale.

Nonsignificant cross-lagged pathways may indicate that third variables better explain the relations between certain ED symptoms and SI over time. However, there were instances where ED symptoms and SI predicted one another. Given this, targeting suicidal thoughts in therapy may help to reduce eating pathology and vice versa.