Temporal age-related changes in spectral, fractal and complexity characteristics of heart rate variability

Cross-sectional studies have suggested that heart rate (HR) variability, analysed using traditional time and frequency domain methods, is related to ageing, but no longitudinal studies have estimated the age dependence of HR fluctuation. This study evaluated temporal age-related changes in 12-h measures of HR variability among 109 patients with coronary artery disease (CAD), who underwent repeat Holter recordings at 32-month intervals. Time and frequency domain measures, along with fractal and complexity measures of HR variability, were determined at the baseline and after 32 months. Changes in HR dynamics were compared with various laboratory variables, exercise data and angiographic progression of CAD. Traditional time and frequency domain measures of HR variability did not change significantly during the follow-up, but the power-law scaling slope decreased from -1.29 +/- 0.20 to -1.36 +/- 0.23 (P<0.01) and the short-term fractal exponent (alpha1) of HR dynamics from 1.29 +/- 0.14-1.22 +/- 0.18 (P<0.001). The approximate entropy value also decreased from 1.00 +/- 0.19 to 0.95 +/- 0.18 (P<0.05). The changes in HR behaviour were not related to demographic data, laboratory values or angiographic progression of CAD. Only a weak correlation was observed between the change in the power-law slope and the baseline glucose value (P<0.05). This longitudinal study shows that the fractal characteristics of HR dynamics and the complexity properties of R-R intervals undergo rapid changes along with ageing, and that fractal and complexity analysis techniques are more sensitive than traditional analysis methods in documenting temporal age-related changes in HR behaviour.