Nocturnal hypoxemic burden has been shown to be a robust, independent predictor of all-cause mortality in patients with heart failure and reduced ejection fraction (HFrEF) and to occur in a severe form even in patients with low or negligible frequency of respiratory events (apneas/hypopneas). This suggests the existence of two components of hypoxemic burden: one unrelated to respiratory events and the other related. The aim of this study was to characterize these two components and to evaluate their prognostic value.

Nocturnal hypoxemic burden was assessed in a cohort of 280 patients with HFrEF by measuring the percentage of sleep with an oxygen saturation (SpO₂) <90% (T90), and the area of the SpO₂ curve below 90% (Area90). Both indices were also recalculated within the sleep segments associated with respiratory events (event-related component: T90_Eve, Area90_Eve) and outside these segments (nonspecific component: T90_Nspec, Area90_Nspec). The outcome of the survival analysis (Cox regression) was all-cause mortality.

During a median follow-up of 60 months, 87 patients died. T90, Area90, and their components were significant in univariate analysis (P < .05 all). However, when these indices were adjusted for known risk factors, T90, T90_Nspec, Area90, and Area90_Nspec remained statistically significant (P = .018, hazard ratio (HR)=1.12, 95%CI=(1.02, 1.23); P = .007, HR=1.20, 95% CI = [1.05, 1.37]; P = .020, HR = 1.05, 95% CI = [1.01, 1.10]; P = .0006, HR = 1.15, 95% CI = [1.06, 1.25]), whereas T90_Eve and Area90_Eve did not (P = .27, P = .28). These results were internally validated using bootstrap resampling.

By demonstrating a significant independent association of nonspecific hypoxemic burden with all-cause mortality, this study suggests that this component of total nocturnal hypoxemic burden may play an important prognostic role in patients with HFrEF.