Validity of heart rate-based indices to measure training load and intensity in elite football players
This study aimed to identify the most sensible heart rate-based indices to physical measures of training load and intensity. Twenty professional football players competing in the Russian league and in the UEFA Champions League were monitored during 15 training sessions (270 individual records) using GPS devices (10 Hz) and heart rate telemetry. Expert knowledge and a collinearity r < .5 were used initially to select the external physical markers for the final analysis. A multivariate-adjusted within-subjects model was employed to quantify the correlations between heart rate indices with various measures of training intensity and load. The number of accelerations > 2.5 m/s2 and the number of high intensity bursts remained in the final multivariate model for training load. The adjusted correlations with Banister’s TRIMP were r = .49 and r = .3, respectively. For training intensity, the same previous variables expressed as per minute plus the volume of high speed running per minute remained in the final model. The adjusted correlations with the percentage of time spent above 80% of individual maximum heart rate (tHR80%) were, in the same order, r = .3, r = .22 and r = .18. The results of this study demonstrate the validity of TRIMP and tHR80% as measures of training load and intensity, respectively, and identified accelerations and high intensity repeated efforts (high intensity bursts) as being moderately predictive of heart rate responses.