Different categories of VO 2 kinetics in the 'extreme' exercise intensity domain.

The aim of this study was to classify potential sub-zones within the extreme exercise domain. Eight well-trained male cyclists participated in this study. The upper boundary of the severe exercise domain (P upper-bound ) was estimated by constant-work-rate tests. Then three further extreme-work-rate tests were performed in discrete regions within the extreme domain: extreme-1) at a work-rate greater than the P upper-bound providing an 80-110-s time to task failure; extreme-2) a 30-s maximal sprint; and extreme-3) a 4-s maximal sprint. Different functions were used to describe the behaviour of the V ˙ O 2 kinetics over time. V ˙ O 2 on-kinetics during extreme-1 exercise was best described by a single-exponential model (R 2 ≥ 0.97; SEE ≤ 0.10; p  < 0.001), and recovery V ˙ O 2 decreased immediately after the termination of exercise. In contrast, V ˙ O 2 on-kinetics during extreme-2 exercise was best fitted by a linear function (R 2 ≥ 0.96; SEE ≤ 0.16; p  < 0.001), and V ˙ O 2 responses continued to increase during the first 10-20 s of recovery. During the extreme-3 exercise, V ˙ O 2 could not be modelled due to inadequate data, and there was an M-shape recovery V ˙ O 2 response with an exponential decay at the end. The V ˙ O 2 response to exercise across the extreme exercise domain has distinct features and must therefore be characterised with different fitting strategies in order to describe the responses accurately.