Linear and non-linear properties of heart rate in postnatal maturation.

Objective: An investigation was made of how postnatal maturation of cardiac control can be described by linear and non-linear methods of time series analysis.

Methods: Sixteen healthy and term newborns were studied during their first 6 months of life. Power spectrum analysis including total power (TP), low frequency power (LF), high frequency power (HF) and LF/HF ratio were performed on the instantaneous heart rate (IHR) time series and mean heart rate (HR) was derived. The largest Lyapunov exponent (LLE) was calculated using a modified Wolf algorithm and checked by means of the surrogate-data test.

Results: There is an age dependency for the parameters LLE, HR, TP, HF, LF, LF/HF for active and LLE, HR, TP, HF, LF/HF for quiet sleep. HR is characterised by a steep increase between the 5th and 7th day. HF demonstrates a distinct development with high values around the first week and 90-180th day and low values around the 10-60th day. TP, LF and LF/HF show significantly higher values in active sleep in comparison to quiet sleep. For all ages and sleep stages, positive LLE were found, indicating sensitivity to initial conditions, a hall-mark of chaos. For the period between the 7 and 90th days of life, the LLE for active sleep took on larger values compared with the LLE of quiet sleep.

Conclusions: This study shows that by linear as well as non-linear analysis one can reveal the complexity of the IHR development in humans and may gain insight into the system controlling the heart during the period considered. The positive LLE indicate that there is a non-linear component in the heart rate control. There is no "straight line" development for the parameters analysed within the first 6 months. This may result from manifold influences on the autonomic system, due to structural and functional maturation in this period of life.