A consort-guided randomized, blinded, controlled clinical trial on the effects of 6 weeks training on heart rate variability in thoroughbred horses.

Background: Fitness assessment of horses remains challenging. Heart rate variability (HRV) can be used to monitor human athlete's training, but its value is unknown in horses.

Hypothesis: The linear domain HRV variables are affected by fitness.

Animals: Twelve healthy untrained thoroughbreds were randomly split into a training group (6 weeks of incremental racetrack training) and a control group (no training).

Methods: Linear domain HRV variables were analyzed (high frequency [HF], low frequency [LF], their normalized units [Hf, Lf], root mean square of successive differences between beats [RMSSD], Poincaré plot features [SD1 and SD2]) while resting overnight before (baseline) and after 2, 4 and 6 weeks of training. V̇O and echocardiographic indexes were measured at baseline and after 6 weeks. Changes in HRV variables over time (ANOVA), and correlation with V̇O (Pearson's chi-squared test) were tested (P < .05 significance).

Results: V̇O, LF/HF ratio, and LF increased while HF decreased in the training group (before and after training mean [SD] values: V̇O 134 [12.8]-146 [16.5] mL/[kg min]; P < .001; LF/HF ratio 0.51 [0.2]-0.91 [0.3] [P = .02]; LF 37.5 [10.1]-46.8 [7.8] [P = .02]; HF 76.2 [7.9]-53.2 [7.7] [P < .001]). Training did not affect RMSSD, SD1, SD2, resting HR, or echocardiographic indexes. Strong correlations were found between V̇O and HRV variables (V̇O and LF [r = -0.59, P = .04]; ΔV̇O and the corresponding ΔLF [r = -0.88, P = .02]).

Conclusions And Clinical Importance: Six weeks of training affected some frequency domain HRV variables. Further studies are necessary to validate the use of HRV for monitoring horses' responses to training.