Session S64.3

Fractal Dimension of Heart Rate Variability for Screening of Obstructive Sleep Apnoea Syndrome

M Baumert*, H Dimitri, A Fitch, D Leong, A Thornton, DH Lau, MK Stiles,
B John, E Nalivaiko, G Young, R Antic, P Sanders

The University of Adelaide
Adelaide, Australia

The obstructive sleep apnoea syndrome (OSAS) is a disorder of breathing during sleep affecting ~10-15% of men and 4-7% of women. It has been associated with increased cardiac morbidity and mortality. The analysis of heart rate variability (HRV) provides information on autonomic nervous system (ANS) function and has been shown to be an independent predictor of cardiovascular mortality. Repetitive episodes of hypopneas/apnoeas trigger ANS responses affecting heart rate. In this study we hypothesized that the analyses of HRV might provide a screening tool for OSAS.
Standard overnight polysomnography was performed on 239 consecutive patients with suspected OSAS. For HRV analysis, we recorded ECG (lead II) at a sampling frequency of 128 Hz. A template matching algorithm was used to extract RR intervals throughout the night. Artefacts and ectopic beats were filtered out and remaining gaps interpolated using a local variance estimation algorithm. HRV was quantified with standard time and frequency measures as described by the HRV task force. In addition, the fractal dimension was computed using Higuchi’s algorithm.
We developed a binary logistic regression model for the screening of moderate-severe OSAS (AHI>30). The model was controlled for age, gender, BMI, smoking, CAD, DCM, HT, DM, alcohol consumption. HRV measures were stepwise entered in the model using forward likelihood ratio procedure. HRV measures that were predictive of OSAS were meanNN, high frequency power (HF), and fractal dimension. The sensitivity and specificity of the model were 50% and 93%, respectively, with a correct overall classification of 82%.
In conclusion, HRV analysis of cardiac Holter monitoring may provide a screening tool for OSAS.

(Abstract Control Number: 87)