Association between Mean Heart Rate and Recurrence Quantification Analysis of Heart Rate Variability in End-Stage Renal Disease

Abstract

Linear heart rate variability (HRV) indices are dependent on the mean heart rate, which has been demonstrated in different models (from sinoatrial cells to humans). The association between nonlinear HRV indices, including those provided by recurrence plot quantitative analysis (RQA), and the mean heart rate (or the mean cardiac period, also called meanNN) has been scarcely studied. For this purpose, we analyzed RQA indices of five minute-long HRV time series obtained in the supine position and during active standing from 30 healthy subjects and 29 end-stage renal disease (ESRD) patients (before and after hemodialysis). In the supine position, ESRD patients showed shorter meanNN (i.e., faster heart rate) and decreased variability compared to healthy subjects. The healthy subjects responded to active standing by shortening the meanNN and decreasing HRV indices to reach similar values of ESRD patients. Bivariate correlations between all RQA indices and meanNN were significant in healthy subjects and ESRD after hemodialysis and for most RQA indices in ESRD patients before hemodialysis. Multiple linear regression analyses showed that RQA indices were also dependent on the position and the ESRD condition. Then, future studies should consider the association among RQA indices, meanNN, and these other factors for a correct interpretation of HRV.

Keywords: active standing; heart rate variability; hemodialysis; recurrence plot analysis.

Figure 1

Assessment of the autocorrelation function (upper panels) and false nearest neighbors method (lower panels) for each recording (thin lines) and averaged values per group (thick lines). Black lines correspond to healthy subjects, green lines correspond to ESRD patients before hemodialysis, and magenta lines correspond to ESRD patients after hemodialysis.

Figure 2

Examples of RR time series and recurrence plots from a healthy subject (upper panels) and an end-stage renal disease (ESRD) patient (lower panels). The recording from the healthy subject shows a reduction in RR intervals and their variability, as a consequence of changing to an active standing position. The recurrence plot shows extended global diagonal lines formed by short diagonals covering a regularly spaced pattern of the plot surface. In contrast, during the standing position, there is an increase in white zones (i.e., there are less accessible dynamic states in the central area). In the patient before hemodialysis, there is a marked reduction in the RR intervals and their associated variability, even in the supine position. Its recurrence plot quantitative analysis (RQA) plot has less regularly extended diagonals with more concentration around the identity line and its parallel lines. During the active standing condition, the plot is almost empty; there are very fewer accessible dynamical states. In the same patient after hemodialysis, there is a slight improvement in the RR intervals during the supine condition compared to those before hemodialysis, which is also reflected in the RQA plot. Again, during the active standing challenge, there is a reduction in RR intervals and reachable dynamical states.