Factors influencing QT prolongation in patients hospitalized with severe anorexia nervosa

Anorexia nervosa (AN) carries the highest mortality of any psychiatric disorder largely attributable to sudden cardiac death and suicide. Controversy exists regarding the underlying mechanism of cardiac risk, whether QT prolongation is a consistent feature of the disorder and whether repolarization varies by disease severity. Some of the uncertainty may relate to a lack of standardized electrocardiography (ECG). To date, studies have not utilized centrally adjudicated digital ECG, and most have relied on the Bazett formula for rate-correction, which is suboptimal at the extremes of heart rate often observed in AN.

We evaluated a hospitalized cohort of medically compromised, very-low-body mass index (BMI) AN patients. The QT interval was measured with high-precision calipers by a single, blinded electrophysiologist and rate corrected (QTc) using the Fridericia formula. Anatomically corrected left ventricular (LV) mass and resting energy expenditure (REE) were calculated as proxies for disease severity. Proportions exceeding categorical thresholds for QTc prolongation and correlations between admission QTc and disease severity were performed.

Among 19 patients, mean BMI was 12.3 kg/m(2), and 95% were female. The majority (68%) of patients were receiving QT-prolonging drugs. Four patients (21%) had QTc prolongation. Two of these patients (10.5%) exceeded the 500 ms threshold for marked QTc-prolongation, though each had concomitant factors contributing to delayed repolarization. The QTc interval was not significantly correlated with LV mass, LV mass index, BMI or REE.

Although delayed cardiac repolarization was observed among a medically compromised cohort of patients with anorexia nervosa, the QTc interval was not a reliable correlate of disease severity despite digital ECG adjudication and optimal rate correction.