Modifications of long-term heart rate variability produced in an experimental model of diet-induced metabolic syndrome

Abstract

Metabolic syndrome (MetS) has been linked to a higher prevalence of cardiac arrhythmias, the most frequent being atrial fibrillation, but the mechanisms are not well understood. One possible underlying mechanism may be an abnormal modulation of autonomic nervous system activity, which can be quantified by analysing heart rate variability (HRV). Our aim was to investigate the modifications of long-term HRV in an experimental model of diet-induced MetS to identify the early changes in HRV and the link between autonomic dysregulation and MetS components. NZW rabbits were randomly assigned to control (n = 10) or MetS (n = 10) groups, fed 28 weeks with high-fat, high-sucrose diet. 24-hour recordings were used to analyse HRV at week 28 using time-domain, frequency-domain and nonlinear analyses. Time-domain analysis showed a decrease in RR interval and triangular index (Ti). In the frequency domain, we found a decrease in the low frequency band. Nonlinear analyses showed a decrease in DFA-α1 and DFA-α2 (detrended fluctuations analysis) and maximum multiscale entropy. The strongest association between HRV parameters and markers of MetS was found between Ti and mean arterial pressure, and Ti and left atrial diameter, which could point towards the initial changes induced by the autonomic imbalance in MetS.

Keywords: cardiac autonomic dysfunction; frequency domain; heart rate variability; metabolic syndrome; nonlinear analysis; time domain.

Figure 1.

Analysis of long-term HRV (time domain). Quantification of mean RR interval during day and night is depicted in panel (a) and the hourly distribution during a period of 24 h in panel (b) from 08.00 to 19.59 exposure to light (day) and from 20.00 to 07.59 for total darkness (night). RR values measured in milliseconds (ms). Control, n = 10; MetS, n = 10. *p < 0.05 versus control. Error bars: SEM.

Figure 2.

Analysis of long-term HRV triangular index (time domain). Quantification of the mean triangular index in the day and night is shown in panel (a). Panel (b) shows hour by hour comparisons over a 24-hour period, from 08.00 to 19.59 exposure to light (day) and from 20.00 to 07.59 for total darkness (night). Control, n = 10; MetS, n = 10. *p < 0.05 versus control. Error bars: SEM.

Figure 3.

Analysis of long-term HRV (frequency domain). Quantification of the mean low frequency (LF) band in normalized units (n.u.) in the day and night is depicted in panel (a). Panel (b) shows hour by hour comparison over a 24-hour period, from 08.00 to 19.59 exposure to light (day) and from 20.00 to 07.59 for total darkness (night). Control, n = 10; MetS, n = 10. *p < 0.05 versus control; ^#p = 0.08. Error bars: SEM.

Figure 4.

Nonlinear analysis of long-term HRV (DFA-α1 and DFA-α2). Panels (a) and (c) show the mean values for both control and MetS during the day and night. Panels (b) and (d) show the comparison of the dynamics of DFA-α1 and DFA-α2, hour by hour, in a 24-h period from 08.00 to 19.59 for exposure to light (day) and from 20.00 to 07.59 for total darkness (night). Control, n = 10; MetS, n = 10. *p < 0.05 versus control. Error bars: SEM.

Figure 5.

Nonlinear analysis of long-term HRV (MSE_max). Quantification of the mean maximum multi-scale entropy (MSE_max), in a 24-h period from 08.00 to 19.59 for light exposure (day) and from 20.00 to 07.59 for total darkness (night) is shown in panel (a). Panel (b) depicts the dynamics of MSE_max hour by hour. Control, n = 10; MetS, n = 10. *p < 0.05 versus control. Error bars: SEM.

Figure 6.

Nonlinear analysis of long-term HRV (SD1/SD2). The comparison shows the SD1/SD2 index in arbitrary units, during day and night (a). Panel (b) shows hour by hour comparison over a 24-hour period, from 08.00 to 19.59 for exposure to light (day) and from 20.00–07.59 for total darkness (night). Control, n = 10; MetS, n = 10. Error bars: SEM.