Prognostic ability of resting periodic breathing and ventilatory variation in closely matched patients with heart failure

PURPOSE: Periodic breathing (PB), characterized by waxing and waning oscillations of ventilation, has been reported in some patients with heart failure. Although PB is usually assessed in sleeping patients, PB has also been observed in awake patients. We tested the hypothesis that PB in awake, resting patients with heart failure predicts mortality. METHODS: Resting gas exchange data were collected in consecutive, awake patients with heart failure referred for possible cardiac transplantation. Periodic breathing was defined as oscillations of resting ventilation, that is, 30% or more during 2 consecutive cycles. Each PB patient was matched to a comparison patient by age, gender, heart failure etiology, peak oxygen consumption, peak respiratory exchange ratio, and ventilatory efficiency slope. RESULTS: Forty-four of 355 patients met the criteria for PB and were matched to 44 comparison patients. During an average follow-up Of 4.9 +/- 0.1 years, 30 patients with PB (68%) versus 23 comparisons (52%) died or were transplanted urgently (P = NS). However, among the 88 PB patients and comparisons, ventilatory variation was ail independent predictor of mortality by Cox regression analysis (P = .004). Resting ventilatory variation increased the explained variation in mortality by 44%, from 16% to 23%, in a mortality model that included ventilatory efficiency slope, peak oxygen Consumption, left ventricular ejection fraction, systolic blood pressure, sex, age, heart rate, and etiology. CONCLUSION: Resting ventilatory variation independently and powerfully predicted mortality in this cohort of patients with heart failure.