This Article Abstract Full Text Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Mäkikallio, T. H. Articles by Myerburg, R. J. Search for Related Content PubMed PubMed Citation Articles by Mäkikallio, T. H. Articles by Myerburg, R. J.

Prediction of sudden cardiac death by fractal analysis of heart rate variability in elderly subjects

Timo H. Mäkikallio, MD^* , Heikki V. Huikuri, MD, FACC, Anne Mäkikallio, MD, Leif B. Sourander, MD, Raul D. Mitrani, MD, FACC^*, Agustin Castellanos, MD, FACC^* and Robert J. Myerburg, MD, FACC^*

^* Division of Cardiology, University of Miami, School of Medicine, Miami, Florida, USA
Division of Cardiology, University of Oulu, Oulu, Finland
Department of Neurology, University of Oulu, Oulu, Finland
Department of Geriatrics, University of Turku, Turku, Finland

View larger version (30K): [in a new window]   Figure 1 Kaplan-Meier survival curves for the all-cause mortality, cardiac mortality, sudden cardiac deaths, as well as cerebrovascular mortality with the short-term fractal exponent of heart rate dynamics <1.0 or 1.0. Short-term fractal exponent was a particularly powerful predictor of sudden cardiac death with very high negative predictive accuracy (upper panel, right). Short-term fractal exponent did not predict cerebrovascular mortality as seen in the lower right panel.

View larger version (22K): [in a new window]   Figure 2 Receiver operating characteristic curves for the short-term fractal exponent, long-term scaling exponent and for SDNN in predicting cardiac death and sudden cardiac death. Short-term fractal exponent had higher sensitivity than SDNN or long-term scaling exponent in all specificity levels. AUC = area under the curve; SDNN = standard deviation of all N-N intervals.

View larger version (21K): [in a new window]   Figure 3 Examples of power spectra and different fractal scaling exponent values. High exponent values are seen when predominant low frequency fluctuation is present and high frequency fluctuation is reduced (left panel). Low exponent values are seen in cases with reduced low and very low frequency fluctuations and either reduced or preserved high frequency fluctuations (right panel). "Normal" fractal scaling value near 1.0 is seen when both low and high frequency power are relatively preserved with more power in the very low and low frequency areas than in the high frequency area (middle panel).