Prolonged QTc interval predicts mortality in patients with Type 1 diabetesmellitus

Aims To evaluate prolonged QTc interval and QT dispersion as predictors of all-cause and cardiovascular mortality after adjustment for well-establishe d risk factors in Type 1 diabetic patients. Methods From a cohort of all adult Type 1 diabetic patients, duration of di abetes greater than or equal to5 years, attending the clinic in 1984 and fo llowed in an observational study for 10 years (n = 939), all subjects with resting baseline electrocardiograms were identified (n = 697, 360 males). T he QT length was measured and corrected for heart rate (QTc). Maximal QTc l ength (QTc max) and QT dispersion were determined. Results At baseline, 431 had normoalbuminuria (<30 mg/24 h), 138 had microa lbuminuria (30-299 mg/24 h) and 128 had macroalbuminuria (<greater than or equal to>300 mg/24 h) of whom 66 (15%), 35 (25%) and 61 (48%) died during f ollow-up, respectively (26 (6%), 14 (10%), 21 (16%) from cardiovascular dis ease). QTc max. was 442 (1.2) ms (mean (SEM)) for survivors and 457 (3.7) i n patients who died (P < 0.001). In a Cox proportional hazards model includ ing baseline values of putative risk factors, independent predictors of dea th were QTc max (P = 0.03), age (P < 0.001), presence of hypertension (P = 0.001), male sex (P < 0.001), log urinary albumin excretion (P < 0.001), sm oking (P = 0.04), log serum-creatinine (P < 0.001), height (P < 0.001), low social class (P = 0.04), whereas QT dispersion, heart rate, and HbA(1c) we re not included. In the subgroup with macroalbuminuria, but not for all pat ients, QTc max was an independent risk factor for cardiovascular mortality. Conclusion QTc prolongation, but not increased QT dispersion, is an indepen dent marker of increased mortality in patients with Type 1 diabetes mellitu s.