Increased risk of somatic diseases following anorexia nervosa in a controlled nationwide cohort study

OBJECTIVE: To assess the risk of somatic diseases in connection with anorexia nervosa (AN).

METHOD: This matched cohort study was based on Danish registries of all patients born 1961-2008 with a first-time diagnosis of AN in 1994-2018 at age 8-32 and matched controls without an eating disorder. For 13 somatic disease categories, time from inclusion date to time of first somatic diagnosis, accounting for censoring, was studied by use of time-stratified Cox models.

RESULTS: A total of 9985 AN patients born 1961-2008 and 49,351 controls were followed for a median (interquartile range) of 9.0 (4.4-15.7) years. During the first 2 years after entry there was a 60% higher hazard for any somatic disease among patients with AN than among controls, while the ratio from three to 11 years was reduced to 1.18. Regardless of age at diagnosis, the hazard among patients and controls were no different at approximately a decade after diagnosis of AN and the cumulative risk for patients for 12 of 13 disease categories was always higher or no less that for controls. For all disease categories, the hazard ratio (HR) was higher when close to entry. For most disease categories, age at diagnosis of AN did not modify the effect.

DISCUSSION: While around 90% of all individuals had any somatic disease at the end of follow-up, the cumulative incidence over time was higher for patients with AN than for controls. Large HRs were seen in the early years after diagnosis during which patients require extensive medical interventions.

PUBLIC SIGNIFICANCE: Based on Danish registries, a large sample of almost 10,000 patients with AN born 1961-2008 and almost 50,000 matched controls were followed for a median of 9 years. While around 90% of all individuals had any somatic disease at the end of follow-up, the cumulative incidence over time was higher for patients with AN than for controls.