Screening of obstructive sleep apnea syndrome from time-domain analysis and frequency-domain analysis of heart rate variability

Hai-cheng Zhang,Jian-ling SUN,Jing LI,Fang HAN,Xiao-song DONG,Ji-hong GUO
DOI: https://doi.org/10.3760/cma.j.issn.1007-6638.2005.01.007
2005-01-01
Abstract:Objective To evaluate the feasibility of screening obstructive sleep apnea syndrome ((OSAS)) from time-domain analysis and frequency-domain analysis of heart rate variability (HRV) via ambulatory electrocardiogram monitoring. Methods The overnight sleep investigation was administered to 95 subjects by polysomnogram (PSG) and 24 hours ambulatory electrocardiogram monitoring simultaneously during July through December, 2003. Forty-eight OSAS positive patients and 47 OSAS negative patients provided by sleep center of this hospital were enrolled into study group and control group,separately. However, the two electrocardiogram analyzers did not know the PSG results at all. They were asked to give the apnea scores by heart rate variability analysis. Results The average age, male gender, body mass index, history of hypertension and diabetes were higher in the study group than those of the control group. The PNN50 (night)、SDNN (day-night) and SDNNI (day-night) among time-domain analysis were more sensitive than other parameters. In frequency-domain analysis of HRV during nocturnal sleep, the mean nighttime Total Power, nighttime VLF Power, nighttime LF Power as well as their difference between day and night in study group were significanly higher than those of control group. The LF to HF ratio was significanly higher than that of control group whatever in daytime, nighttime and difference between day and night (P0.05). When sleep apnea risk score4, the sensitivity of screening diagnose of (OSAS) was 81.25%, with the specificity of 46.81%, and the positive predictive value of 64.21%. Whereas, when sleep apnea risk score 5,6 or 7 respectively, the specificity became gradually higher and the sensitivity and the positive predicting value pretty lower. Conclusion Time-domain analysis and frequency-domain analysis of HRV are helpful to screen the suspicious OSAS patients, and frequency-domain analysis is much more sensitive than time-domain analysis. Ambulatory electrocardiogram monitoring is a simple shortcut useful tool as a non-invasive means for screening OSAS.
What problem does this paper attempt to address?