Distortion of complexity in oscillations of arterial blood pressure, due to aging
Abstract
Age both influences and limits an individual’s capability to withstand adverse circulatory distress such as sepsis or cardiac events, and is by itself considered a major risk factor for hypertension. The aim of this study was to find whether general degradation of integrated physiological regulatory systems, due to aging, would distort very low-frequency (VLF) oscillations in arterial blood pressure (AP). We postulated that by comparing two different age spans, we would find a loss of complexity of natural oscillations for AP pressure in the VLF-range area (0.0095–0.05 Hz). From an existing open-access (PhysioNet) dataset of 20 individuals, we extracted continuous multi-parameter recordings of circulation. Five young (21-34 yr) and five elderly (68-81 yr) subjects were chosen for further analysis. The continuous non-invasive AP signals were first analysed using Hilbert-Huang Transformations (HHT, provided by Signal Analysis Lab) which is fundamentally based on Entrinsic Mode Decompositions (EDM). The HHT produced several Intrinsic mode functions (IMFs), which then were graphically visualized in a time-frequency domain by using Hilbert spectrum analysis and thus identified classical blood pressure frequencies as shown in Fig 1.0. We also identified all systolic R-peaks in the raw data and then conducted both detrended fluctuation analysis (DFA) and continuous wavelet transformation of these new time series. Our findings show that there is a significant difference in complexity of the blood pressure signal assessed by DFA in the two age groups included in the study (n=6). p= 0.006728 with a 95 % CI of (0.04033657, 0.11903996) for the identified differences. However we were not able to determine whether the difference in complexity is caused by potential differences in the VLFrange. It seems that our spectrums from the HHT and Continuous Wavelet Analysis (CWT) also show differences between the two age groups in the VLF-range. Yet it poses a challenge to establish statistical tools to determine if such a difference indeed exist. Thus further research is needed in order to establish such a correlation.