Student thesis: Doctoral ThesisPhD


The regular testing and monitoring of an athletes’ performance and well-being status are well-established key components of a successful training programme. However, the associated procedures are challenging to conduct in swimming practice. This is primarily due to the technological constraints of the aquatic environment, the multidisciplinary nature of swimming, and the limited time, resources, and physiological testing expertise often available to swimming coaches. The recent development of the 3-minute all-out test (3MT) and improvements in athlete heart rate variability (HRV) monitoring technology have helped to address some of these issues across a number of sports. However, their application in the sport of swimming remains to be investigated.

The first study (Chapter 3) of this thesis demonstrates that the 3MT is a valid (correlation with traditional method r=0.95 for CS, r=0.79 for D′, p<0.002), reliable (ICC=0.97, CV=0.9% for CS; ICC=0.87, CV=9.1% for D′), and feasible protocol for the assessment of the parameters describing the critical speed (CS) concept in competitive swimmers. In Chapter 4, the application of the 3MT is further extended to enable the demarcation of the remaining exercise intensity domains from CS only. In comparison to traditional step test, this method shows to represent a more valid whilst feasible approach (no statistically significant differences p=0.93-1.00, and nearly prefect correlations at lactate threshold r=0.92, and lactate turnpoint r=0.90, and very large correlations at maximal aerobic speed r=0.88; all p<0.0001), compared to the popular “beats below maximal heart rate” method, which produced significantly lower estimates than those established from a step test (all p<0.03), despite large-to-very large correlations (r=0.63-0.89, p<0.03) found between lactate threshold and lactate turnpoint (for speed and heart rate). Chapter 5 demonstrates that a seasonal training programme, individualised based on the data obtained from a regular 3MT, is associated with seasonal improvements in several important parameters of swimming performance (mean change ± 90% confidence limits: CS: +5.4 ± 1.6%, personal best time in 1st: -1.2 ± 1.3% and 2nd main event: -1.6 ± 0.9%, stroke rate: +6.4 ± 3.0% and stroke index: +4.2 ± 3.6% at CS), despite a substantial reduction (≥ 25%) in the overall training volume of highly trained swimmers. Finally, Chapter 6 describes swimmers’ HRV responses to a typical training season, reveals large associations between seasonal changes in key HRV parameters and CS (∆Ln rMSSDMEAN: r=0.51, p=0.13; ∆Ln rMSSDCV: r=-0.68, p=0.03), and demonstrates that the Banister Impulse-Response model and data collected via a novel smartphone application (HRV4Training) allow the effective monitoring and modelling of swimmers’ responses to swimming training and non-training related stressors (R2 values of 0.75-0.87).
Overall, this body of work focuses on bridging the gap between science and practice in the testing, training prescription, and monitoring of competitive swimmers and provides examples of approaches through which the existing gaps could be closed.
Date of Award18 Nov 2020
Original languageEnglish
Awarding Institution
  • University of Bath
SupervisorSean Williams (Supervisor), Javier Gonzalez (Supervisor) & Ana Sousa (Supervisor)

Cite this