**4. Discussion**

The purpose of the current study was to validate the physiological and biomechanical parameters corresponding to a fixed lactate concentration of 4 mmol·L−<sup>1</sup> with those measured in an intermittent 5 × 400-m swimming training set at constant speed. A constant swimming speed corresponding to V4 was successfully maintained and BL was no different from that at 4 mmol·L−<sup>1</sup> during the 5 × 400-m training set. However, during the training set, a lower SL and a higher HR and SR compared to predicted values were recorded. Despite the Bland and Altman plot indicating agreement, a grea<sup>t</sup> individual variation was observed for all physiological and biomechanical variables.

A similar speed compared to V4 was expected during the 5 × 400-m training set as this was predesigned in our experimental design. This is because swimmers were guided to follow this speed using audio signals. Considering that swimmers maintained the prescribed speed, we should explain the observed metabolic and biomechanical alterations required by the swimmers to maintain this speed. No difference was found between 4 mmol·L−<sup>1</sup> and BL-5×400-m, although a large effect size (*d* = 0.75) and a grea<sup>t</sup> variation in measured values was observed between swimmers. Using a fixed value of 4 mmol·L−<sup>1</sup> has been criticized in previous studies since it does not taking into account an individual approach to estimate BL [3,12]. Specifically, it was proposed that the fixed

value did not take into account the individual blood lactate concentration curve kinetics and can be a ffected by the muscle glycogen content [12,13]. This is confirmed by studies reporting a higher V4 or BL-V4 during a continuous swimming trial but not in an intermittent swimming set as in the current study [19,20]. However, it should be considered that in the aforementioned studies [19,20], well-trained long-distance swimmers or young swimmers were included. Well-trained swimmers specializing in various distances participated in the current study, and they may present di fferent individual characteristics compared to swimmers participating in previous studies. The swimming stroke specialty or athlete specializations may di fferentiate BL e fflux under a constant speed [20,21]. Specifically, it has been shown that sprint-oriented athletes or those with limited aerobic potential may present higher BL values compared with long distance athletes during a continuous swimming training set under a constant speed at V4 [21,22], as was observed in the current study. Despite the observed variation, BL concentration prescribed by the progressively increasing speed test was similar to that measured during the intermittent training set. Lactate responses are reflected by RPE, an index of internal training load, during intermittent swimming [23]; RPE calculated by the 5 × 200-m test was found to be similar to that measured during the 5 × 400-m training set.

Individual variation observed in measured BL and RPE, as well as di fferences in HR, may be attributed to equations used for calculation of these parameters. The second-order polynomial function used to calculate V4 presented a low error of estimation (r = 0.98, SE = 0.07) and a high accuracy of the predicted BL value. In contrast with BL, the measured HR values were higher during the 5 × 400-m intermittent training set compared to the value calculated by the 5 × 200-m progressively increasing speed test. This may be explained by a higher sympathetic activation during the 5 × 400-m intermittent training set [24]. Furthermore, the higher HR during the 5 × 400-m training set may indicate a higher effort of swimmers to maintain the prescribed V4 swimming speed. It should be noted that HR-V4 was calculated using an appropriate linear fit of data (r = 0.98) and low standard error of estimate (3 <sup>b</sup>·min−1) that may partly explain the di fferences during 5 × 400-m training set and HR-V4. Possibly, di fferences in HR were dependent on exercise duration between progressively increasing speed testing protocol (200-m repetitions) and training set (400-m repetitions; [5]). In this case, HR may have not reached steady values within the shorter 200-m distance, thus underestimating the values during a longer duration 400-m distance.

Besides the physiological load required in maintaining constant V4, some di fferences in biomechanical parameters were also noticed. Increased SR-V4 (large e ffect size) and decreased SL-V4 (medium e ffect size) were observed during the 5 × 400-m intermittent training set compared to the calculated values by the progressively increasing speed 5 × 200-m test. Increased SR and decreased SL during swimming are connected to increased energy cost [25]. These changes occur when swimmers manage to maintain the prescribed speed. Possibly, well-trained swimmers who participated in the current study applied less force (decreased SL) to maintain an e fficient arm stroke [26]. However, these changes were not severe enough to induce fatigue manifested as an inability to maintain constant speed during the 5 × 400-m. A further explanation for these di fferences may be that swimmers completed a higher number of arm-stroke cycles in the longer distance of 400-m compared to 200-m, thus altering their mechanics to compensate for the longer distance [27]. Possibly, the di fferent distances used for testing (400-m vs. 200-m) as well as the swimmer's specialty combined with the large interindividual variation between swimmers may have led to the aforementioned di fferences in biomechanical parameters.

The characteristics of the progressively increasing speed testing protocols (number of repetitions, duration of each repetition) that have been used to estimate physiological and biomechanical variables may lead to di fferent responses during an intermittent or a continuous training set [28,29]. However, this is controversial throughout the literature. Specifically, Madsen et al. [30], found that a progressively increasing speed test consisted of 200-m repetitions overestimates the physiological variables obtained during continuous swimming training, while recently it has been found that 200-m and 300-m testing protocols showed similar physiological and biomechanical variables during intermittent swimming training [31]. Whatever the case, coaches should be aware that the predicted parameters are dependent on testing protocol used for their calculation and may not be similar to those expected during an intermittent training set.
