Effect of recording length and posture on the reliability of heart rate variability in horses

Heart rate variability (HRV) is applied in equine research as a biomarker for stress, athletic fitness, and health status. However, information on its reliability in horses is limited and improved understanding could strengthen future studies. The aim of this study was to quantify the effect of recording length and posture on the reliability of short-term HRV parameters. Fourteen horses underwent repeated night-time sessions with concurrent Holter and (infrared) video monitoring. The coefficient of variation (CV) and intraclass correlation coefficient (ICC) of time-domain and frequency domain parameters were calculated for recording lengths from 30 seconds to five minutes. The effect of posture on HRV outcomes was assessed by a mixed effects linear model. Intra-individual variability (ICC) accounted for 0.47-0.79 of overall variability for the various HRV parameters. Reliability increased with longer segment lengths. Segments shorter than 300s were not suitable for the LF and LF/HF HRV parameters. Posture affected the absolute values of HRV parameters but the size and direction of the effect differed between parameters. Respiratory rate was not recorded and the number of days between sessions was not standardised. Segment lengths of more than five minutes might have further improved the reliability of some parameters. Fair to moderate repeatability can be achieved for HRV parameters. Segments shorter than five minutes are not suitable for frequency domain analysis. Posture affects HRV outcomes.