Association Between Eating Disorders and Subsequently Recorded Depression and Bipolar Disorder: A Retrospective Cohort Study
Vanessza Hadar, Battamir Ulambayar, Attila Csaba Nagy, Amr Sayed GhanemBackground/Objectives: Mental health is strongly affected by eating disorders, yet there are few longitudinal studies that have examined their association with subsequently recorded mood disorders. The present study aimed to investigate the association between exposure to eating disorders and first recorded diagnoses of depression and bipolar disorder during follow-up. Methods: This retrospective cohort study was based on the hospital database of the Clinical Center of the University of Debrecen, covering the period between 2007 and 2022. The dataset included 333,807 patients and 805,989 hospital records. Diagnoses were identified using physician-recorded ICD-10 codes. Eating disorders were defined as the exposure variable, while first recorded diagnoses of depression and bipolar disorder during follow-up were treated as outcomes. Patients with the relevant mood disorder recorded in their entry year were excluded from the corresponding analysis. Associations were evaluated using survival analysis to estimate hazard ratios (aHRs) and 95% confidence intervals (CIs). Results: In the respective Cox analysis samples, eight subsequent depression diagnoses occurred among 68 patients with baseline eating disorders, and seven subsequent bipolar disorder diagnoses occurred among 82 patients with baseline eating disorders. Eating disorders were associated with higher hazards of subsequently recorded depression (aHR = 4.23; 95% CI: 2.11–8.46) and bipolar disorder (aHR = 5.45; 95% CI: 2.59–11.44). Female sex was associated with a higher hazard of subsequently recorded depression and a lower hazard of subsequently recorded bipolar disorder. Conclusions: Recorded eating disorders were associated with subsequently recorded depression and bipolar disorder. These associations require cautious interpretation given the small number of exposed outcome events, imprecise estimates and potential residual confounding, and warrant further prospective investigation.