Adverse outcomes and mortality in individuals with eating disorder-related electrolyte abnormalities in Ontario, Canada: a population-based cohort study.

BACKGROUND

Individuals with eating disorders are at a higher risk of electrolyte abnormalities than the general population. We conducted the first representative cohort study assessing whether electrolyte abnormalities in people with eating disorders were associated with mortality and physical health outcomes.

METHODS

This was a retrospective population-based cohort study in Ontario including people aged 13 years or older with an eating disorder and an outpatient electrolyte measure within 1 year (between Jan 1, 2008 and June 30, 2019). An electrolyte abnormality was any of hypokalaemia, hyperkalaemia, hyponatraemia, hypernatraemia, hypomagnesaemia, hypophosphataemia, metabolic acidosis, or metabolic alkalosis. The primary outcome was all-cause mortality. Secondary outcomes were hospitalisation, a cardiac event, infection, acute or chronic kidney disease, fracture, and bowel obstruction. In additional analyses, we examined a younger cohort (<25 years old) and individuals with no previously diagnosed secondary outcome. We involved people with related lived or family experience in the study.

FINDINGS

6163 patients with an eating disorder and an electrolyte measure within 1 year since diagnosis (mean age 26·8 years [SD 17·5]; 5456 [88·5%] female, 707 [11·5%] male; median follow-up 6·4 years [IQR 4-9]) were included. Ethnicity data were not available. The most common electrolyte abnormalities were hypokalaemia (994/1987 [50·0%]), hyponatraemia (752/1987 [37·8%]), and hypernatraemia (420/1987 [21·1%]). Overall, mortality occurred in 311/1987 (15·7%) of those with an electrolyte abnormality versus 234/4176 (5·6%) in those without (absolute risk difference 10·1%; adjusted hazard ratio 1·23 [95% CI 1·03-1·48]). Hospitalisation (1202/1987 [60·5%] vs 1979/4176 [47·4%]; 1·35 [1·25-1·46]), acute kidney injury (206/1987 [10·4%] vs 124/4176 [3%]; 1·91 [1·50-2·43]), chronic kidney disease (245/1987 [12·3%] vs 181/4176 [4·3%]; 1·44 [1·17-1·77]), bone fracture (140/1987 [7·0%] vs 167/4176 [4·0%]; 1·40 [1·10-1·78]), and bowel obstruction (72/1987 [3·6%] vs 57/4176 [1·4%]; 1·62 [1·12-2·35]) were associated with an electrolyte abnormality, but not infection or a cardiovascular event. Findings were consistent in young individuals (<25 years old) and those without secondary outcomes at baseline, by eating disorder type, and by sex.

INTERPRETATION

Electrolyte abnormalities are associated with death and poor physical health outcomes, supporting the importance of monitoring and possible interventions to prevent adverse outcomes. Findings also call for a refinement of the definition of severity of eating disorder and replication of these findings in other jurisdictions.

FUNDING

None.