Inverse relationship of subjective daytime sleepiness to mortality in heart failure patients with sleep apnoea

Takatoshi Kasai,Luigi Taranto Montemurro,Dai Yumino,Hanqiao Wang,John S Floras,Gary E Newton,Susanna Mak,Pimon Ruttanaumpawan,John D Parker,T Douglas Bradley

DOI: https://doi.org/10.1002/ehf2.12808

Abstract:Aims: Patients with sleep apnoea (SA) and heart failure (HF) are less sleepy than SA patients without HF. HF and SA both increase sympathetic nervous system activity (SNA). SNA can augment alertness. We previously showed that in HF patients, the degree of daytime sleepiness was not related to the severity of SA but was inversely related to SNA. Elevated SNA is associated with increased mortality in HF. Therefore, we hypothesized that in HF patients with SA, the degree of daytime sleepiness will be inversely related to mortality. Methods and results: In a prospective cohort study, 218 consecutive patients with systolic HF had overnight polysomnography. Among them, 80 subjects with SA (apnoea-hypopnoea index ≥15) were followed for a mean of 28 months to determine all-cause mortality rate. Subjective daytime sleepiness was assessed by the Epworth Sleepiness Scale (ESS). During follow-up, 20 patients died. The 5 year death rate in patients with ESS less than 6 (i.e. less sleepy) was significantly higher than in patients with an ESS at or above the median of 6 (i.e. sleepier) [21.3 deaths/100 patient-years vs. 6.2 deaths/100 patient-years, unadjusted hazard ratio (HR) 2.94, 95% confidence interval (CI) 1.20 to 7.20, P = 0.018]. After adjusting for confounding factors that included sex, history of hypertension, and mean arterial oxyhaemoglobin saturation, compared with the sleepier patients, less sleepy patients had greater risk of mortality (HR 2.56, 95% CI 1.01 to 6.47, P = 0.047). As a continuous variable, ESS scores were inversely related to mortality risk (HR 0.86, 95% CI 0.75 to 0.98, P = 0.022). Conclusions: In patients with HF and SA, the degree of subjective daytime sleepiness is inversely related to the mortality risk, suggesting that among HF patients with SA, those with the least daytime sleepiness are at greater risk of death. They may therefore have greater potential for mortality benefit from therapy of SA than those with greater daytime sleepiness.