Original paper-RISK ASSESSMENT OF DIABETES MELLITUS BY CHAOTIC GLOBALS TO HEART RATE VARIABILITY VIA SIX POWER SPECTRA - ABSTRACT
Authors: David M. Garner 1, Naiara Maria de Souza 2, Luiz Carlos M. Vanderlei 2
1 Cardiorespiratory Research Group, Department of Biological and Medical Sciences, Faculty of Health and Life Sciences, Oxford Brookes University, Oxford, United Kingdom
2 Department of Physiotherapy, UNESP - Univ Estadual Paulista - Presidente Prudente, Sao Paulo, Brazil
Background: The priniciple objective here is to analyze cardiovascular dynamics in diabetic subjects by actions related to heart rate variability (HRV). The correlation of chaotic globals is vital to evaluate the probability of dynamical diseases. Methods: Fortysix adults were split equally. The autonomic evaluation consisted of recording HRV for 30 minutes in supine position without any additional stimuli. “Chaotic globals” are then able to statistically determine which series of interbeat intervals are diabetic and which are not. Two of these chaotic globals, spectral Entropy and spectral Detrended fluctuation analysis were derived from six alternative power spectra: Welch, Multi-Taper Method, Covariance, Burg, Yule-Walker and the Periodogram. We then compared results to observe which power spectra provided the greatest significance by three statistical tests: One-way analysis of variance (ANOVA1); Kruskal-Wallis technique and the multivariate technique, principal component analysis (PCA). Results: The Chaotic Forward Parameter One (CFP1) applying all three parameters is proven the most robust algorithm with Welch and MTM spectra enforced. This was proven following two tests for normality where ANOVA1 (p=0.09) and Kruskal-Wallis (p=0.03). Multivariate analysis revealed that two principal components represented 99.8% of total variance, a steep scree plot, with CFP1 the most influential parameter. Conclusion: Diabetes reduced the chaotic response.
Key words: diabetes; power spectra; principal component analysis; complexity; chaos