Heart rate variability analysis in obstructive sleep apnea patients with daytime sleepiness
Seren Ucak, Hasthi U. Dissanayake, Philip de Chazal, Yu Sun Bin, Kate Sutherland, Bianca Setionago, B Tong, Brendon J. Yee, Kristina Kairaitis, John R. Wheatley, Amanda J. Piper, Peter A. Cistulli, Sydney Sleep Biobank Investigators, Peter A. Cistulli, Philip de Chazal, Kate Sutherland, N Sarkissian, Chin Moi Chow, Andrew S. L. Chan, Aimee B Lowth, Jacob T. Graham, William F. Wood, Gary Cohen, Callum Bennett, Mohammad Ahmadi, John R. Wheatley, Kristina Kairaitis, Stephen Lambert, R. Ginn, Tracey Burns, Brendon J. Yee, Amanda J. Piper, Keith Wong, Kerri Melehan, Margaret Chan, David Wang, Gislaine Gauthier
Abstract
STUDY OBJECTIVES: Recent studies suggest that sleepy patients with obstructive sleep apnea (OSA) are at higher risk for incident cardiovascular disease. This study assessed cardiac autonomic function in sleepy versus non-sleepy patients with OSA using heart rate variability (HRV) analysis. We hypothesized that HRV profiles of sleepy patients would indicate higher cardiovascular risk. METHODS: Electrocardiograms (ECG) derived from polysomnograms (PSG) collected by the Sydney Sleep Biobank were used to study HRV in groups of sleepy (ESS ≥ 10) and non-sleepy OSA patients (ESS < 10). HRV parameters were averaged across available ECG signals during N2 sleep. RESULTS: A total of 421 patients were evaluated, with a mean age of 54 (14) years, body mass index of 33 (9) kg/m2, apnea-hypopnea index of 21 (28) events/h, and 66% male. The sleepy group consisted of 119 patients and the non-sleepy group 302 patients. Sleepy patients exhibited lower HRV values for: root mean square successive difference (RMSSD, p = 0.028), total power (TP, p = 0.031), absolute low frequency (LF, p = 0.045), and high-frequency (HF, p = 0.010) power compared to non-sleepy patients. Sleepy patients with moderate-to-severe OSA exhibited lower HRV values for: (RMSSD, p = 0.045; TP, p = 0.052), absolute LF (p = 0.051), and HF power (p = 0.025). There were no differences in other time and frequency domain HRV markers. CONCLUSIONS: This study shows a trend toward parasympathetic withdrawal in sleepy OSA patients, particularly in moderate-to-severe cases, lending mechanistic support to the link between the sleepy phenotype and CVD risk in OSA.