Heart Rate response and cardiovascular risk during OSA: an easy biomarker derived from pulse oximetry

Sleep apnoea specific heart rate response (ΔHR) has been identified as a promising biomarker for stratifying cardiovascular (CV) risk, and predicting positive airway pressure (PAP) benefit in obstructive sleep apnoea (OSA). However, the need for prior manual scoring of respiratory events potentially limits the accessibility and reproducibility of ΔHR. We aimed to evaluate the association of pulse rate response to oxygen desaturations automatically derived from pulse oximetry (ΔHRoxi) with CV risk in OSA. ΔHRoxi and ΔHR were measured in OSA patients from the IRSR Pays de la Loire Sleep Cohort (PLSC; n=5,002) and the HypnoLaus cohort (n=1,307). The primary outcome was major adverse CV events (MACEs), a composite of mortality, stroke, and cardiac diseases. Cox regressions analyses were conducted to evaluate the association of ΔHRoxi and ΔHR categorized into low, midrange and high categories, with MACEs. MACEs occured in 768 patients from PLSC and 87 patients from HypnoLaus (median follow-up: 8.0 and 7.5 years respectively). Multivariable Cox models showed that subjects with high ΔHRoxi (vs midrange) had higher risk of MACEs in PLSC (hazard ratio, HR: 1.42 [95% CI:1.18-1.71]) and HypnoLaus (HR: 1.72 [1.03-2.87]). Similar findings were observed for high ΔHR. Among 2,718 patients from PLSC treated with PAP, the association of PAP adherence (PAP use ≥4 h/night, vs non adherence) with MACEs was modified by baseline ΔHR and ΔHRoxi (p for interaction<0.05). ΔHRoxi could constitute a reliable and easy to measure biomarker for stratifying CV risk and predicting CV benefit of PAP in OSA.