Anorexia nervosa is associated with a markedly increased risk of sudden cardiac death, but the mechanism has not been elucidated. Whether QT prolongation is an intrinsic feature of eating disorders is uncertain because previous studies are limited by small sample size, and extrinsic factors associated with QT prolongation were inconsistently reported. This study set to determine population-mean heart-rate–corrected QT interval (QTc) in an unselected cohort of patients with eating disorders.

Electrocardiogram (ECG) data from 1026 consecutive adults admitted into residential treatment were stratified by subtype: anorexia nervosa (caloric restriction only), anorexia nervosa binge-purge, and bulimia nervosa. Eating disorders not otherwise specified were excluded. Population-mean Fridericia-corrected QTc and categorical QTc threshold analysis were performed. Multivariable regression, controlling for age sex, duration of illness, body mass index (BMI), hypokalemia, QTc-prolonging drugs, purging behaviors, and laxatives was assessed.

Among 906 patients, population-mean QTc (424 ± 25 standard deviation [SD]) was normal and lowest among patients with anorexia nervosa (417.3 ± 22.3, P <0.001 vs other subgroups). Only 1.2% (N = 11) had marked QTc prolongation (QTc >500 ms); all 11 patients had hypokalemia and were receiving QTc-prolonging medications or laxatives. After controlling for clinically relevant covariates, differences in mean QTc across eating disorder subtypes diminished yet persisted (P = 0.048).

In the largest study of patients with eating disorders, population-mean QTc was normal and varied by subtype. Marked QTc prolongation occurred solely in the presence of extrinsic factors, suggesting that QTc prolongation is not intrinsic to eating disorders. Therefore, further study is needed to define the etiology of sudden death in patients with eating disorders.