Login / Signup

Inter-individual variability in the load-velocity relationship is detected by multilevel mixed regression models.

Eliseo Iglesias-SolerXian MayoJessica Rial-VázquezGuy Gregory Haff
Published in: Sports biomechanics (2018)
The main aim of this study was to explore the variability in the load-velocity relationship through the use of multilevel mixed regression models. The relationship between relative load (% of one repetition maximum: %1RM) and velocity was obtained in a sample of high-level judokas and rugby players (8 women and 13 men) for the bench press (BP) and parallel squat (SQ). The load-velocity relationship for the squat was obtained for the external load (barbell load) and for the system mass (barbell plus body mass). The data were fitted by different multilevel mixed regression models. Including the sex factor in the models improved the goodness of fit for the BP but not for the squat exercises. All the models detected significant inter-individual variability in both intercepts and slopes (p < 0.05 in all the cases). A decrease of 0.15, 0.10 and 0.16 m/s of velocity for each 10% of increment in the relative load were estimated for BP and squat considering the external load and the system mass, respectively. The multilevel mixed regression models detected significant inter-individual variability in the slope and intercept of the load-velocity relationship what entails differences in the velocity associated with a fixed percentage (%) of the one-repetition maximum load.
Keyphrases
  • blood flow
  • pregnant women
  • machine learning
  • polycystic ovary syndrome
  • skeletal muscle
  • insulin resistance
  • big data