Use of Cross-Correlation Analysis to Determine Heart Rate Kinetics During Non-Steady State, Fatiguing Exercise in Collegiate Female Soccer Athletes

For years, heart rate (HR) kinetics have been used as an indicator of training status and fatigue. Slowed kinetics indicate poor fitness and/or fatigue. In this study it was determined that HR kinetics can be reliably estimated during modeled, quasi-binary, and dynamic exercise using cross-correlation analysis of HR and external work rate. Heart rate and running speed were cross-correlated yielding a cross-correlation function (CCF) and analyzed for its peak (CCFmax) and time delay (CCFlag). Modeled exercise data where the time constants for HR (τ) increased from 2-120 sec. yielded linear decreases in CCFmax (r2 = 0.9949) and linear increases in CCFlag (r2=0.9996). A strong linear relationship existed between CCFmax and CCFlag (r2=0.9989). Steady-state exercise data produced strong relationships between the calculated τ and CCFmax (r2=0.8736) and CCFlag (r2=0.9061), and CCFmax and CCFlag showed a positive relationship (r2=0.7753). CCFmax between repeated sprint trials (R2=0.9123) and super set trails (R2=0.9227) were very similar. These results suggest good repeatability for both quasi-binary activity and random activity. To assess validity, CCFmax values during the sprint trials were compared to two standard field tests of fitness (Beep and Man U tests). There were strong relationships between CCFmax and distances covered during the Beep (r2=0.7911) and Mann U tests (r2=0.7770). Lastly, the applicability of the CCF method was applied to dynamic exercise, using data collected from competitive soccer matches. For the field players, significant reductions in CCFmax occurred during the first and second periods of the match. There was a significant relationship between the total distance covered during the match and the decline in CCFmax (r=-0.4297, p