Related Experiment Video
Updated: Apr 18, 2026
A Rapidly Incremented Tethered-Swimming Maximal Protocol for Cardiorespiratory Assessment of Swimmers
Published on: January 28, 2020
Reliability of novel diagnostic protocols for metabolic performance determinants in swimming
Tinka Smolarek1, Patrick Wahl1, Sebastian Keller2
1Institute of Exercise Training and Sport Informatics, Section Exercise Physiology, German Sport University Cologne, Germany; The German Research Centre of Elite Sport Cologne, German Sport University Cologne, Germany.
Objectives:
While reliable performance diagnostics are crucial for guiding training and detecting small changes in swimming performance, the reliability of existing protocols remains insufficiently evaluated.
Design:
Test-retest study.
Methods:
Twenty-four swimmers and triathletes of both sexes completed a 20-s sprint to assess anaerobic power (i.e., maximal lactate accumulation rate [ċLamax], peak lactate concentration [Lapeak]) and a 3-min step test (ST) to evaluate aerobic performance indicators (i.e., post-exercise peak oxygen uptake [V̇O2peak], cost of swimming [Cs], speed at lactate thresholds [vLT1, vLT2] and in the final step [vpeak], alongside corresponding heart rate [HR]) on three separate days (within 3 weeks). Reliability and agreement were examined using (among others) intraclass correlation coefficient (ICC 2,1), typical error (TE), and TE to smallest worthwhile change (SWC) ratio, indicating sensitivity.
Results:
The sprint demonstrated moderate reliability for ċLamax (ICC: 0.625-0.728) and Lapeak (ICC: 0.705-0.749), with high TEs of 11.8-16.8% and 9.7-12.0%, respectively. Although familiarization improved reliability, sensitivity remained inadequate (TE/SWC: ≥1.23). The ST exhibited good to excellent reliability (mixed sensitivity) across all parameters. Uncorrected post-exercise V̇O2peak (ICC: 0.829-0.963, TE: 3.8-7.1%, TE/SWC: 1.02-1.83) and Cs (ICC: 0.884-0.952, TE: 3.8-5.7%, TE/SWC: 0.83-1.17) aligned with HR-corrected values. vLT1, vLT2, and vpeak showed excellent reliability (ICC: 0.971-0.990, TE: 0.8-1.4%) with adequate sensitivity (TE/SWC: 0.60-1.14), whereas HR measures were less reliable (ICC: 0.828-0.951, TE: 1.2-3.5%) and insufficiently sensitive (TE/SWC ≥ 1.60).
Conclusions:
The ST provides a reliable method for assessing aerobic performance, with high sensitivity for lactate-based training intensity markers and Cs, but limited sensitivity for V̇O2peak and HR. The 20-s sprint requires methodological refinement before practical application.
