Van Leeuwen P;
Cysarz D; Edelhäuser F;
Grönemeyer Institute of Microtherapy, University of Witten/Herdecke, Bochum, Germany. Electronic address: email@example.com.
The change in fetal heart rate and its variability (HRV) during the course of gestation has been documented by numerous studies. The overall drop in heart rate and increase in fetal HRV is associated with fetal growth in general and with the increase in neural integration in particular. The increased complexity of the demands on the cardiovascular system leads to more variation in the temporal course of the heart rate. Most studies that document and interpret these changes are based on data acquired in groups of fetuses. The aim of this work was to investigate HRV within single fetuses. We acquired 213 5min fetal magnetocardiograms in 11 fetuses during the second and third trimesters (at least 10 data sets per fetus, median 17). From the magnetocardiograms we determined the fetal RR interval time series and calculated the standard deviation (SDNN), root mean square of successive differences (RMSSD), approximate entropy (ApEn) and temporal asymmetry (Irrev). For each subject and HRV measure, we performed regression analysis with respect to gestational age, alone and in combination with RR interval. The coefficient of determination R(2) was used to estimate goodness-of-fit. The coefficient of quartile dispersion (CQD) was used to compare the regression parameters for each HRV measure. Overall, the HRV measures increased with age and RR interval. The consistency of the HRV measures within the individual fetuses was greater than in the data pooled over all fetuses. The individual R(2) for the model including age and RR interval was best for ApEn (.79, .59-.94; median, 90% CI), followed by RMSSD (.71, .25-.88), SDNN (.55, .18-.90) and Irrev (.16, .01-.39). These values, except for Irrev, were higher than those calculated over all 213 data sets (R(2)=.65, .63, .35, .28, respectively). The slopes of the regressions of each individual's data were most consistent over all subjects for ApEn, followed by RMSSD and SDNN and Irrev. Interindividually, the time domain measures showed discrepancies and the within-fetus courses were more consistent than the course over all fetuses. On the other hand, the course of ApEn during gestation was not only very consistent within each fetus but also very similar between most of subjects. Complexity measures such as ApEn may thus more consistently reflect prenatal developmental factors influencing cardiovascular regulation.