Interindividual Variation in the Relationship of Different Intensity Markers—A Challenge for Targeted Training Prescriptions

• Published in: PloS one Vol. 11; no. 10; p. e0165010
• Main Authors: Egger, Florian, Meyer, Tim, Hecksteden, Anne
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 27.10.2016; Public Library of Science (PLoS)
• Subjects: Accuracy; Adult; Bicycling - physiology; Bicycling - statistics & numerical data; Biology and Life Sciences; Biomarkers; Comparative analysis; Cycling; Cyclists; Exercise intensity; Heart Rate - physiology; Homeostasis; Humans; Indicators; Individuality; Lactates; Lactic acid; Load tests; Male; Markers; Medicine and Health Sciences; Metabolism; Model accuracy; Oxygen; Oxygen Consumption - physiology; Oxygen uptake; Parameters; Physical Conditioning, Human - methods; Physical Sciences; Predictions; Preventive medicine; Reference Values; Regression analysis; Regression models; Social Sciences; Training; Variability; Germany
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0165010

Training intensities are frequently prescribed as relative workloads based on a single reference value (e.g. maximum oxygen uptake). However, exercise-induced physical strain is multifaceted and large interindividual variability in intensity markers has been reported for constant load exercise with standardized relative intensity. This questions the accuracy of (univariate) relative intensities in targeting specific training stimuli. The present trial aims to investigate interindividual variability in the relationship of strain indicators using interpolated performance curves derived from constant load tests at different workloads. This approach enables the prediction of other indicators based on a chosen reference and subsequent comparison of predictive accuracy between group-based and individualized regression models. 15 competitive cyclists completed a stepwise incremental cycling test followed by 5 constant load tests with the same absolute workloads as in the stepwise incremental test. The highest of theses workloads which yielded a lactate (BLa) steady state was repeated enabling estimation of intraindividual variability. From constant load tests, the courses of BLa relative to the respective reference value (e.g. %VO2peak) were interpolated by polynomial regression. Variability between individual regression curves was analyzed by mixed modeling. Predictive accuracy was estimated as the sum of squared differences between predicted and observed values. The proportion of total variation in the course of BLa relative to the respective reference parameter accounted for by subject identity ranged between 36 and 51%. A significant increase in predictive accuracy was observed for VO2peak and HRmax, respectively, as predicting parameters. These results are in support of a multivariable, individualized approach to intensity prescriptions when aiming at accurately targeted perturbations of homeostasis.