Eating Disorders and Later Incidence of Cancer: A Nationwide Longitudinal Study in Denmark

Background: We examined the incidence of cancer types among individuals with eating disorders (EDs). Methods: A nationwide longitudinal study of 6,807,731 individuals born between 1940 and 2015 was conducted using the Danish National Registries. Cox models with ED diagnosis as exposure and cancer di...

Full description

Saved in:
Bibliographic Details
Main Authors: Gabrielle E. Cooper, Natalie M. Papini, Katrine Holde, Cynthia M. Bulik, Zeynep Yilmaz, Liselotte V. Petersen
Format: Article
Language:English
Published: Elsevier 2025-07-01
Series:Biological Psychiatry Global Open Science
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2667174325000370
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Background: We examined the incidence of cancer types among individuals with eating disorders (EDs). Methods: A nationwide longitudinal study of 6,807,731 individuals born between 1940 and 2015 was conducted using the Danish National Registries. Cox models with ED diagnosis as exposure and cancer diagnoses as outcomes were used to estimate hazard ratios (HRs) and 95% CIs while adjusting for sex, birth year, and comorbidities. The primary analysis comprised ICD-8 and ICD-10 codes for anorexia nervosa (AN) and other ED (OED). The secondary analysis comprised ICD-10 codes and included AN, bulimia nervosa (BN), and eating disorders not otherwise specified (EDNOS). Results: AN was associated with a reduced incidence of breast cancer while adjusting for sex and birth year (HR, 0.80; 95% CI, 0.66–0.97) and elevated incidence of respiratory (HR, 1.59; 95% CI, 1.24–2.04), cervical (HR, 1.45; 95% CI, 1.05–1.98), and esophageal (HR, 4.77; 95% CI, 2.82–8.06) cancers. OED was associated with an elevated incidence of respiratory (HR, 1.57; 95% CI, 1.20–2.06) and cervical (HR, 1.60; 95% CI, 1.20–2.14) cancers. ICD-10–only analyses confirmed the association of AN with reduced incidence of breast cancer and elevated incidence of respiratory and cervical cancers. BN was associated with reduced incidence of breast cancer in sensitivity analysis. EDNOS was associated with reduced incidence of breast cancer and elevated incidence of respiratory and cervical cancers. Conclusions: All EDs were associated with a reduced incidence of breast cancer. All EDs except BN were associated with a higher incidence of respiratory and cervical cancers. AN was associated with a higher incidence of esophageal cancer.
ISSN:2667-1743