Abstract
AIMS: Nocturnal hypoxaemic burden, quantified as time spent with oxygen saturation below 90% (T90), is an established independent predictor of mortality in heart failure (HF) with reduced ejection fraction. The prognostic value of T90 in HF with preserved ejection fraction (HFpEF) is unknown. This study aims to determine the association of T90 with adverse outcomes in HFpEF. METHODS AND RESULTS: One hundred twenty-six patients prospectively included from our specialised HFpEF outpatient clinic underwent ambulatory home sleep monitoring to obtain oximetry data, including T90. We investigated the association between T90 and a composite endpoint of HF hospitalisations or all-cause mortality. Nocturnal hypoxaemic burden in this HFpEF population was high, with a median T90 of 13.7 min. In only 10 patients (7.9%), oxygen saturation was at no time point below 90%. After median 34 months [IQR 18.4-52.0] of follow-up, 32 patients (25%) reached the composite endpoint. T90 was significantly associated with the composite endpoint, also after adjusting for potential confounders (HR 1.004 (95% CI 1.001-1.007, P = 0.019) per 1 min T90 increase or HR 1.265 (95% CI 1.061-1.488) per 1 h T90 increase). Patients with HFpEF in the highest T90 tertile (T90 ≥ 31.4 min) had a significantly higher event rate compared to patients in the lowest two T90 tertiles, with 19 (45%) versus 13 (15%) events, respectively (P < 0.001). CONCLUSIONS: Nocturnal hypoxaemic burden is an independent prognostic marker for the composite of HF hospitalisations or all-cause mortality in HFpEF. Whether reduction of T90 improves the prognosis of patients with HFpEF warrants further research.