QTc Interval and Risk of Cardiac Events in Adults With Anorexia Nervosa: A Long-Term Follow-Up Study

BACKGROUND: The literature contains several cases of anorexia nervosa (AN) patients with prolonged QTc interval. However, the risk of prolonged QTc interval is controversial and the risk of cardiac events in AN patients has yet to be investigated. METHODS: We estimated the difference in mean QTc interval and relative risk of borderline prolonged QTc (>440 ms) and prolonged QTc (>460 ms) between 430 adult women AN patients and 123 healthy controls using 3 correction formulas. In a follow-up study, we estimated the risk of a primary end point (a composite of ventricular tachycardia, aborted cardiac arrest, and cardiac arrest) in AN patients compared with a population-based cohort derived from the Danish Civil Register. RESULTS: Mean QTc for AN patients was 408 ms (Hodges), 402 ms (Fridericia), and 399 ms (Bazett). Hodges' found a slightly increased mean QTc (6.8 ms, 95% confidence interval, 1.6-12.0; P=0.01) and percentage with QTc >440 ms in AN patients (relative risk, 3.7, 95% confidence interval, 1.4-10.3; P=0.01), not observed with Fridericia's and Bazett's formulas. There was no difference in the risk of QTc >460 ms between AN patients and healthy controls. During a median follow-up of 10.1 years, AN patients had an increased risk of the primary end point compared with the population-based cohort (hazard ratio, 10.4, 95% confidence interval, 2.6-41.6; P=0.001). However, absolute numbers were small with cumulative incidences of 0.5% and 0.07%, respectively, after 10 years. No events occurred in any AN patient with QTc >440 ms. All-cause mortality was also significantly increased in AN patients compared with the population-based cohort (hazard ratio, 11.2, 95% confidence interval, 5.1-24.5; P<0.001). CONCLUSIONS: Overall, there was no difference in mean QTc interval or risk of prolonged QTc between AN patients and healthy controls. However, AN patients had a notably increased all-cause mortality, as well as an increased risk of cardiac events, which was not related to the baseline QTc interval.