Muscle Oxygenation Measured with Near-Infrared Spectroscopy Following Different Intermittent Training Protocols in a World-Class Kayaker—A Case Study

Training elite kayakers at a distance of 1000 m is associated with aerobic and anaerobic metabolism, while intermittent training, in a variety of forms, is one of the effective ways to improve cardiorespiratory and metabolic function. Thus, this study aimed to investigate muscle oxygenation responses during repetition training (RT), interval training (IT), and sprint interval training (SIT). Near-infrared spectroscopy (NIRS) monitors were placed on the latissimus dorsi (LD), pectoralis major (PM), and vastus lateralis (VL) of a world-class kayaker during their preparatory period. The intensity of work, relief, and recovery intervals were the independent variables that were manipulated using three different training protocols. The inferential analysis between intermittent training protocols showed significant differences for all variables except total the hemoglobin (tHb) index in LD during bout 2 (F = 2.83, p = 0.1, ηp2 = 0.205); bout 3 (F = 2.7, p = 0.125, ηp2 = 0.193); bout 4 (F = 1.8, p = 0.202, ηp2 = 0.141); and bout 6 (F = 1.1, p = 0.327, ηp2 = 0.092). During the rest bouts, all training protocols showed significant differences for all variables except muscle oxygen saturation (SmO2) in the VL during bout 5 (F = 4.4, p = 0.053, ηp2 = 0.286) and tHb in VL during bout 1 (F = 2.28, p = 0.132, ηp2 = 0.172); bout 2 (F = 0.564, p = 0.561, ηp2 = 0.049); bout 3 (F = 1.752, p = 0.205, ηp2 = 0.137); bout 4 (F = 1.216, p = 0.301, ηp2 = 0.1); and bout 6 (F = 4.146, p = 0.053, ηp2 = 0.274). The comparison between IT protocols RT and SIT presented similar results. All variables presented higher values during SIT, except HR results. Finally, the comparison between IT and SIT showed significant differences in several variables, and a clear trend was identified. The results of this study suggest that the application of different intermittent exercise protocols promotes distinct and significant changes in the peripheral effect of muscle oxygenation in response to training stimuli and may be internal predictors of hemodynamic and metabolic changes.


Introduction
To achieve high-standard-specific goals, athletes are constantly looking for ways to optimize skeletal muscle function and its monitoring process through different types of physical training methodologies and evaluation procedures. Based on oxygen-dependent characteristics, NIRS is one of the non-invasive methods that can provide information about the changes in the oxygen saturation of muscle tissue during various sports exercises [1]. For that reason, the popularity of the NIRS method in sports research and real-world scenarios has been growing in recent years [2]. Wearable and wireless devices are fixed (AD), triceps brachii (TB), LD, and VL [30]. Thus far, there is only one study in junior male athletes that displayed a moderate correlation between the maximal O 2 extraction in the LD during an incremental test on a kayak ergometer and both 200 and 1000 m performances [15]. It is understood that some muscle oxygenation studies are limited by technical conditions that cause inconvenience in attaching sensors and performing paddling movements. However, there is a need to determine the contribution of other muscles, such as the PM, that have not been previously studied in any kayak training protocol. Depending on the demands of the training, several approaches exist to control and individualize intermittent exercise intensities [31,32]. Individual incremental test parameters are much more objective, practical, and likely the more accurate and effective criteria for achieving the desired performance results [33][34][35]. Paquette and Bieuzen [36] studied thirteen canoe kayak athletes of different genders and different levels to determine their muscle oxygenation and cardiac output responses to various HIIT sessions with the intensity ranging from 110% peak power output to all-out, suggesting that the muscle demand for O 2 is high, especially with the increase in the number and targeting intensity. However, it is unknown whether moderate-intensity repetition work can cause muscle oxygenation and cardiac output in world-class kayakers, compared with moderate interval and sprint interval training. Thus, the purpose of this study is to assess muscle oxygen responses during the RT, IT, and SIT in a world-class kayaker and to determine oxygenation parameters in the VL, PM, and LD muscles in each training workout. We hypothesized that the SIT training protocol activates the muscle's O 2 dynamics and oxidative energy metabolism more than the IT and RT, and this response would be detected using NIRS.

Subject
A male world-class kayaker (World Championship silver and bronze medal winner and European Championship bronze medal winner in 1000 m kayak flat water race event), during the preparatory period, participated in this study. At the start of the data collection, the participant's age was 32 years, with a height of 184.5 cm, body mass of 89 kg, and training volume of 18 h·week −1 . Physical characteristics are presented in Table 1. Note: VO 2 max-maximal oxygen uptake; CIL-critical intensity limit; VT2-2nd ventilatory threshold; HR-heart rate; W-watts.

Design
During the study, the athlete was encouraged to undertake their normal training but not to train on the day before each test. The athlete was acquainted with the experimental procedures prior to testing and gave written informed consent to participate in the study. All the experimental procedures involved in this study were approved by the Bioethics Research Committee of Vilnius Region (#158200-18 /11-1040-573). This study adhered to ethical principles under the Declaration of Helsinki.
The participant performed three randomized separated training sessions, RT, IT, and SIT, upon a Dansprint PRO, KE001 ergo, Denmark kayak ergometer at air brake resistance level 7. The ergometer was calibrated before each test according to the manufacturer's recommendations, and the tension in the ergometer's ropes was verified regularly [37]. All sessions were performed under similar environment conditions (relative humidity 60%) and circumstances (from 11.00 to 12.30 h).
The study protocol practice sessions started with a 15 min standard warm-up comprising rowing exercises and 5 min of recovery (Table 2). Table 2. Experimental conditions of the protocol depicting the three IT intensity modes.

Duration 15 6 Bouts
RT Warm-Up S- 6 6 Research Committee of Vilnius Region (#158200-18/11-1040-573). This study adhe ethical principles under the Declaration of Helsinki. The participant performed three randomized separated training sessions, RT, I SIT, upon a Dansprint PRO, KE001 ergo, Denmark kayak ergometer at air brake resi level 7. The ergometer was calibrated before each test according to the manufac recommendations, and the tension in the ergometer's ropes was verified regularl All sessions were performed under similar environment conditions (relative hu 60%) and circumstances (from 11.00 to 12.30 h).
The study protocol practice sessions started with a 15 min standard warm-up prising rowing exercises and 5 min of recovery (Table 2). For all tests, the athlete started in the kayak ergometer, waited 3 min for the s signal, and established the baseline responses for NIRS and HR monitor. The parti was asked to perform the same volume of 6 bouts of activity interspersed with 6 passive recovery (Table 2). To program the RT, IT, and SIT intensities, the subject' vidual power output characteristics were used at the critical intensity limit (CIL) the second ventilatory threshold (VT2) associated with physiological markers s VO2max, VO2, and HR (Table 1) [38]. The participant received stroke power outp heart rate (HR) feedback during the test and was asked to maintain the right intensi in each stage of the exercises. Blood lactate (Bla) concentration (mmol·L −1 ), as a pro metabolic anaerobic demand, was determined 3 min after the end of each interm training.

RT
The RT protocol comprised 6 bouts of 6 min of ergometer paddling at 200 wa tensity interspersed with 6 min of passive recovery. The participant was instructed sume the ready position, and after the starting signal, the activity lasted for 72 min

IT
The IT protocol comprised 6 bouts of 6 min that consisted of interspersing per 1-min ergometer paddling at the intensity of 200 watts and 1 min relief at 40 watt dling intensity. The 6 min of activity were interspersed with 6 min of passive rec Research Committee of Vilnius Region (#158200-18/11-1040-573). This study ad ethical principles under the Declaration of Helsinki.
The participant performed three randomized separated training sessions, RT SIT, upon a Dansprint PRO, KE001 ergo, Denmark kayak ergometer at air brake re level 7. The ergometer was calibrated before each test according to the manuf recommendations, and the tension in the ergometer's ropes was verified regula All sessions were performed under similar environment conditions (relative h 60%) and circumstances (from 11.00 to 12.30 h).
The study protocol practice sessions started with a 15 min standard warmprising rowing exercises and 5 min of recovery (Table 2). For all tests, the athlete started in the kayak ergometer, waited 3 min for the signal, and established the baseline responses for NIRS and HR monitor. The pa was asked to perform the same volume of 6 bouts of activity interspersed with passive recovery (Table 2). To program the RT, IT, and SIT intensities, the subje vidual power output characteristics were used at the critical intensity limit (CIL the second ventilatory threshold (VT2) associated with physiological markers VO2max, VO2, and HR (Table 1) [38]. The participant received stroke power ou heart rate (HR) feedback during the test and was asked to maintain the right inten in each stage of the exercises. Blood lactate (Bla) concentration (mmol·L −1 ), as a p metabolic anaerobic demand, was determined 3 min after the end of each inte training.

RT
The RT protocol comprised 6 bouts of 6 min of ergometer paddling at 200 w tensity interspersed with 6 min of passive recovery. The participant was instruct sume the ready position, and after the starting signal, the activity lasted for 72 m

IT
The IT protocol comprised 6 bouts of 6 min that consisted of interspersing p 1-min ergometer paddling at the intensity of 200 watts and 1 min relief at 40 wa dling intensity. The 6 min of activity were interspersed with 6 min of passive r Research Committee of Vilnius Region (#158200-18/11-1040-573). This study adhe ethical principles under the Declaration of Helsinki. The participant performed three randomized separated training sessions, RT, I SIT, upon a Dansprint PRO, KE001 ergo, Denmark kayak ergometer at air brake resi level 7. The ergometer was calibrated before each test according to the manufac recommendations, and the tension in the ergometer's ropes was verified regularl All sessions were performed under similar environment conditions (relative hu 60%) and circumstances (from 11.00 to 12.30 h).
The study protocol practice sessions started with a 15 min standard warm-up prising rowing exercises and 5 min of recovery (Table 2). For all tests, the athlete started in the kayak ergometer, waited 3 min for the s signal, and established the baseline responses for NIRS and HR monitor. The parti was asked to perform the same volume of 6 bouts of activity interspersed with 6 passive recovery (Table 2). To program the RT, IT, and SIT intensities, the subject' vidual power output characteristics were used at the critical intensity limit (CIL) the second ventilatory threshold (VT2) associated with physiological markers s VO2max, VO2, and HR (Table 1) [38]. The participant received stroke power outp heart rate (HR) feedback during the test and was asked to maintain the right intensi in each stage of the exercises. Blood lactate (Bla) concentration (mmol·L −1 ), as a pro metabolic anaerobic demand, was determined 3 min after the end of each interm training.

RT
The RT protocol comprised 6 bouts of 6 min of ergometer paddling at 200 wa tensity interspersed with 6 min of passive recovery. The participant was instructed sume the ready position, and after the starting signal, the activity lasted for 72 min

IT
The IT protocol comprised 6 bouts of 6 min that consisted of interspersing per 1-min ergometer paddling at the intensity of 200 watts and 1 min relief at 40 watt dling intensity. The 6 min of activity were interspersed with 6 min of passive rec ethical principles under the Declaration of Helsinki.
The participant performed three randomized separated training sessions, RT SIT, upon a Dansprint PRO, KE001 ergo, Denmark kayak ergometer at air brake re level 7. The ergometer was calibrated before each test according to the manufa recommendations, and the tension in the ergometer's ropes was verified regula All sessions were performed under similar environment conditions (relative h 60%) and circumstances (from 11.00 to 12.30 h).
The study protocol practice sessions started with a 15 min standard warmprising rowing exercises and 5 min of recovery (Table 2). For all tests, the athlete started in the kayak ergometer, waited 3 min for the signal, and established the baseline responses for NIRS and HR monitor. The pa was asked to perform the same volume of 6 bouts of activity interspersed with passive recovery (Table 2). To program the RT, IT, and SIT intensities, the subjec vidual power output characteristics were used at the critical intensity limit (CIL the second ventilatory threshold (VT2) associated with physiological markers VO2max, VO2, and HR (Table 1) [38]. The participant received stroke power out heart rate (HR) feedback during the test and was asked to maintain the right inten in each stage of the exercises. Blood lactate (Bla) concentration (mmol·L −1 ), as a p metabolic anaerobic demand, was determined 3 min after the end of each inte training.

RT
The RT protocol comprised 6 bouts of 6 min of ergometer paddling at 200 w tensity interspersed with 6 min of passive recovery. The participant was instruct sume the ready position, and after the starting signal, the activity lasted for 72 m

IT
The IT protocol comprised 6 bouts of 6 min that consisted of interspersing pe 1-min ergometer paddling at the intensity of 200 watts and 1 min relief at 40 wa dling intensity. The 6 min of activity were interspersed with 6 min of passive r The participant performed three randomized separated training sessions, RT, I SIT, upon a Dansprint PRO, KE001 ergo, Denmark kayak ergometer at air brake resi level 7. The ergometer was calibrated before each test according to the manufac recommendations, and the tension in the ergometer's ropes was verified regularl All sessions were performed under similar environment conditions (relative hu 60%) and circumstances (from 11.00 to 12.30 h).
The study protocol practice sessions started with a 15 min standard warm-up prising rowing exercises and 5 min of recovery (Table 2). For all tests, the athlete started in the kayak ergometer, waited 3 min for the s signal, and established the baseline responses for NIRS and HR monitor. The parti was asked to perform the same volume of 6 bouts of activity interspersed with 6 passive recovery (Table 2). To program the RT, IT, and SIT intensities, the subject' vidual power output characteristics were used at the critical intensity limit (CIL) the second ventilatory threshold (VT2) associated with physiological markers s VO2max, VO2, and HR (Table 1) [38]. The participant received stroke power outp heart rate (HR) feedback during the test and was asked to maintain the right intensi in each stage of the exercises. Blood lactate (Bla) concentration (mmol·L −1 ), as a pro metabolic anaerobic demand, was determined 3 min after the end of each interm training.

RT
The RT protocol comprised 6 bouts of 6 min of ergometer paddling at 200 wa tensity interspersed with 6 min of passive recovery. The participant was instructed sume the ready position, and after the starting signal, the activity lasted for 72 min

IT
The IT protocol comprised 6 bouts of 6 min that consisted of interspersing per 1-min ergometer paddling at the intensity of 200 watts and 1 min relief at 40 watt dling intensity. The 6 min of activity were interspersed with 6 min of passive rec The participant performed three randomized separated training sessions, RT SIT, upon a Dansprint PRO, KE001 ergo, Denmark kayak ergometer at air brake re level 7. The ergometer was calibrated before each test according to the manuf recommendations, and the tension in the ergometer's ropes was verified regula All sessions were performed under similar environment conditions (relative h 60%) and circumstances (from 11.00 to 12.30 h).
The study protocol practice sessions started with a 15 min standard warmprising rowing exercises and 5 min of recovery (Table 2). For all tests, the athlete started in the kayak ergometer, waited 3 min for the signal, and established the baseline responses for NIRS and HR monitor. The pa was asked to perform the same volume of 6 bouts of activity interspersed with passive recovery (Table 2). To program the RT, IT, and SIT intensities, the subjec vidual power output characteristics were used at the critical intensity limit (CIL the second ventilatory threshold (VT2) associated with physiological markers VO2max, VO2, and HR (Table 1) [38]. The participant received stroke power ou heart rate (HR) feedback during the test and was asked to maintain the right inten in each stage of the exercises. Blood lactate (Bla) concentration (mmol·L −1 ), as a p metabolic anaerobic demand, was determined 3 min after the end of each inte training.

RT
The RT protocol comprised 6 bouts of 6 min of ergometer paddling at 200 w tensity interspersed with 6 min of passive recovery. The participant was instruct sume the ready position, and after the starting signal, the activity lasted for 72 m

IT
The IT protocol comprised 6 bouts of 6 min that consisted of interspersing pe 1-min ergometer paddling at the intensity of 200 watts and 1 min relief at 40 wa dling intensity. The 6 min of activity were interspersed with 6 min of passive r For all tests, the athlete started in the kayak ergometer, waited 3 min for the starting signal, and established the baseline responses for NIRS and HR monitor. The participant was asked to perform the same volume of 6 bouts of activity interspersed with 6 min of passive recovery ( Table 2). To program the RT, IT, and SIT intensities, the subject's individual power output characteristics were used at the critical intensity limit (CIL) and at the second ventilatory threshold (VT2) associated with physiological markers such as VO 2 max, VO 2 , and HR (Table 1) [38]. The participant received stroke power output and heart rate (HR) feedback during the test and was asked to maintain the right intensity (W) in each stage of the exercises. Blood lactate (Bla) concentration (mmol·L −1 ), as a proxy for metabolic anaerobic demand, was determined 3 min after the end of each intermittent training.

RT
The RT protocol comprised 6 bouts of 6 min of ergometer paddling at 200 watts intensity interspersed with 6 min of passive recovery. The participant was instructed to assume the ready position, and after the starting signal, the activity lasted for 72 min.

IT
The IT protocol comprised 6 bouts of 6 min that consisted of interspersing periods of 1-min ergometer paddling at the intensity of 200 watts and 1 min relief at 40 watts paddling intensity. The 6 min of activity were interspersed with 6 min of passive recovery. The participant was instructed to assume the ready position, and after the starting signal, the activity lasted for 72 min.

SIT
The SIT protocol comprised 6 bouts of 6 min that consisted of interspersing periods of 10 s of ergometer paddling at 300 watts intensity with 30 s of relief paddling at 40 watts intensity. The 6 min of activity were interspersed with 6 min of passive recovery. The participant was instructed to assume the ready position, and after the starting signal, the activity lasted for 72 min. participant was instructed to assume the ready position, and after the starting signal, the activity lasted for 72 min.

NIRS Values
The oxygenation level of exercising muscles (oxygenated hemoglobin), SmO2 (%), and deoxygenated total hemoglobin, tHb (arbitrary units AU), were assessed with a NIRS device (Moxy Oxygen Monitor-USA, Hutchinson, MN, USA) ( Figure 1). Three NIRS monitors were placed and affixed using double-sided adhesive tape over the left (dominant) VL, PM, and LD muscles: for the VL, on the distal part of the VL muscle belly (10 cm above the proximal border of the patella); for the PM, on the center of the muscle belly along in the principal direction of the muscle fibers of the sternocostal head, and for the LD, on the midpoint between the inferior border of the scapula and the posterior axillary fold. The skinfold thickness at each site was measured using a skinfold caliper (Harpenden, C-136) to ensure that the skinfold thickness was less than half the distance between the emitter and the detector (25 mm). The raw muscle O2 saturation (SmO2) and total hemoglobin concentration (tHb) signals were captured at 10 Hz, and the data were smoothed using a 10th order low pass-zero phase Butterworth filter (cut-off frequency 0.1 Hz) provided by the recording Artinis Software (Oxysof, Artinis Medical System, Elst, The Netherlands) [39]. Black elastic bandages were used to shield the probes from ambient light and minimize movement during exercise. The values of muscles oxygenation at the baseline (averaging 30 s before exercise), during exercise (sample size of each bout n = 180), and during exercise recovery periods (sample size of each bout n = 180) were recorded in the Moxy PC software (Fortiori Design LLC, Minneapolis, MN, USA), which allowed for the calculation of the average of the recorded values and the lowest point of the SmO2 in each training. The variation between recovery and exercise in SmO2 (Δ SmO2) was calculated by evaluating the difference between the minimum SmO2 and baseline SmO2, Three NIRS monitors were placed and affixed using double-sided adhesive tape over the left (dominant) VL, PM, and LD muscles: for the VL, on the distal part of the VL muscle belly (10 cm above the proximal border of the patella); for the PM, on the center of the muscle belly along in the principal direction of the muscle fibers of the sternocostal head, and for the LD, on the midpoint between the inferior border of the scapula and the posterior axillary fold. The skinfold thickness at each site was measured using a skinfold caliper (Harpenden, C-136) to ensure that the skinfold thickness was less than half the distance between the emitter and the detector (25 mm). The raw muscle O 2 saturation (SmO 2 ) and total hemoglobin concentration (tHb) signals were captured at 10 Hz, and the data were smoothed using a 10th order low pass-zero phase Butterworth filter (cut-off frequency 0.1 Hz) provided by the recording Artinis Software (Oxysof, Artinis Medical System, Elst, The Netherlands) [39]. Black elastic bandages were used to shield the probes from ambient light and minimize movement during exercise. The values of muscles oxygenation at the baseline (averaging 30 s before exercise), during exercise (sample size of each bout n = 180), and during exercise recovery periods (sample size of each bout n = 180) were recorded in the Moxy PC software (Fortiori Design LLC, Minneapolis, MN, USA), which allowed for the calculation of the average of the recorded values and the lowest point of the SmO 2 in each training. The variation between recovery and exercise in SmO 2 (∆ SmO 2 ) was calculated by evaluating the difference between the minimum SmO 2 and baseline SmO 2 , and the tHb (∆ tHb) variation was also assessed by calculating the variation between the maximum tHb and baseline tHb [40].

Heart Rate Responses
HR responses were assessed with a telemetric HR monitor (Polar RS800 CX, Polar Electro Oy, Kempele, Finland). The HR (sample size of each bout n = 180) was measured during all the interval bouts, including during the rest. The HR signals were treated using a moderate filter, cleaning and replacing all irregular heartbreaks with interpolated, adjacent R-R interval values using the Polar Software (Pro Trainer 5, Polar Electro, Finland).

Blood Lactate Concentration
Blood lactate (Bla) concentration (mmol·L −1 ) was calculated 3 min after the end of the protocols. The blood lactate samples were taken from the participant's fingertip and immediately analyzed with a validated lactate analyzer (Lactate Pro; Arkray, Tokyo, Japan).

Statistical Analysis
Descriptive analysis is presented in Tables 3 and 4, and data are presented as means (M) ± standard deviations (SD). Before using the parametrical statistical procedures, the assumptions of normality and sphericity were verified. A one-way repeated-measure ANOVA was performed to identify the differences in muscle oxygen saturation and the total hemoglobin in the VL, PM, and LD muscles, and the heart rate between the interval training modes. Bonferroni's corrections were used for the comparisons of more than two groups, and Cohen's d was calculated as the effect-size measure. The alpha level for all statistical tests was set a priori at α = 0.05, and the calculations were carried out using the SPSS software V24.0 (IBM SPSS Statistics for Windows, Armonk, NY, USA: IBM Corp.). The thresholds for effect-size statistics were <0.2, trivial; <0.6, small; <1.20, moderate; <2.0, large; and >2.0, very large. These statistical computations were processed with a specific post-only crossover spreadsheet for each age group [41].

Results
The results of the inferential analysis between the intermittent training protocols during the exercise bouts and rest bouts are presented in Tables 3 and 4 Figures 2 and 3 show the standardized (Cohen) differences for the pairwise comparisons. The comparison between the protocols showed that the RT protocol presented higher deoxygenation levels than the IT protocol. On the other hand, the SIT protocol presented higher deoxygenation levels than the RT and IT protocols but only in the LD muscle (Figure 2). the SIT protocol presented higher deoxygenation levels than the RT and IT protocols but only in the LD muscle ( Figure 2).  However, the SIT and IT protocols presented higher mean O2 saturation levels duri the passive recovery than the RT protocols.  However, the SIT and IT protocols presented higher mean O 2 saturation levels during the passive recovery than the RT protocols.

Discussion
Our research aimed to assess muscle oxygenation responses during the RT, IT, and SIT in a world-class kayaker and to determine their parameters in the VL, PM, and LD muscles. The findings of this case study only partially confirm our hypothesis: (1) the RT was characterized by a greater mean deoxygenation rate than the IT protocol, and the SIT mean deoxygenation was greater than the RT and IT workouts only in the LD muscle; however, (2) the mean O 2 saturation level during the passive rest period was higher in the SIT and IT protocols than in the RT; (3) oxygenation responses in the three active muscles suggest higher PM muscle recruitment than those of the LD and VL muscles as well as changes in the level of muscle contribution during the exercises of different intensities. This study shows the possibilities of using NIRS devices in the monitoring of elite kayak paddling performance and may provide complementary information to the HR and Bla concentration on a local muscle metabolism level.

O 2 Dynamics during Different Training
In this study, we identified oxygen muscle changes during the RT, IT, and SIT protocols. The RT induced greater mean oxygenation in the PM and VL and caused a greater HR response than other protocols we applied. In our design, the RT involved a constant intensity of around~65% of the critical intensity limit (CIL) when performing a 200 W workload. The increase in mean HR during the RT may reflect an increase in the cardiac output associated with a central cardiocirculatory component of the training [42]. IT intensities (200 W) were similar to RT intensities; however, relief intervals reduced the mean oxygenation level of the IT, which had the lowest ∆ SmO 2 (%) and the highest ∆ tHb during all exercise bouts when compared with the RT and SIT protocols ( Table 5).  The RT was distinguished by the duration of continuous work, while the SIT featured the increased intensity of short intervals. The reviewers of this type of RT response categorize them as metabolic, eliciting large requirements from the O 2 transport and utilization systems [43], and responses to protocols such as the SIT are considered metabolic but with a certain degree of neuromuscular strain [21]. Previous reports state that during moderate-intensity IT, the systolic volume of the heart increases during recovery intervals, causing myocardial metabolic load [21]. Therefore, hypothetically, peripheral metabolic changes were not expected in our study. Paquette and Bieuzen [40] considered that ∆ SmO 2 is a good performance predictor since SmO 2 represents the balance between O 2 delivery and extraction at the muscle level [4]. Thus, a decrease in SmO 2 may originate from both reduced delivery and/or increased extraction. However, it is difficult to draw a conclusion indicating the importance of the exercise mode to elicit the cardiovascular component. The control or adjustment of the intensity of the training sessions related to HR may be limited due to the well-known HR delay at exercise onset [20], which showed a slower response than the SmO 2 response during the IT protocol (Figure 4. As the oxygen demand in the working muscle is the driving force for oxygen delivery by the cardiovascular system [42], muscle deoxygenation responded even faster than the oxygen uptake to the onset of a time trial [44]. as impairments in neural drive and motor unit activation or metabolite accumulation [49]. The oxygenation in the muscles quickly adjusted post-exercise, indicating that the use of NIRS technology showed high sensitivity and may lead to discussion and further investigations as to whether oximetry and HR monitoring are more sensitive methods, especially in the IT ( Figure 4) and SIT ( Figure 6). The three training protocols elicited different increases in blood lactate concentrations during the exercise, showing the contribution of the anaerobic glycolytic system, inferred by blood lactate accumulation, to be numerically greater in the RT (3.5 mmol·L −1 ) ( Figure 5) than in the IT (1.4 mmol·L −1 ) ( Figure 4) and SIT (1.8 mmol·L −1 ) ( Figure 6). The benefit of the relief intensity has often been discussed via changes in blood lactate concentration [50]; however, neither blood [51] nor muscle lactate has a direct (nor linear) relationship with performance capacity [50]. It has also been shown that substantially different intermittent training modalities (as assessed by accumulated Bla-1 levels and the HR) may have relatively similar muscle mean peripheral O2 responses.  Judging by the NIRS indicators, in the LD, the peripheral effects on oxygen extraction during the SIT protocol were higher than during the RT. It has been previously established that increasing exercise intensity improves aerobic energy metabolism, which is primarily linked to increased skeletal muscle mitochondrial content and capillary density [45].
Another feature of our study was to monitor the mean muscle oxygenation of each training session in 6 min duration rest bouts (Figures 4-6). It was possible to observe after which training protocol O 2 returned faster to the pre-exercise conditions since recovery is an important component to improve physical training adaptations [46]. In the RT, oxygenation during the rest bouts in the PM ranged from 54.4% to 65.9%, in the IT from 72% to 79.5%, and in the SIT from 63.5% to 78.6% of SmO 2 ( Table 3). Our findings suggest that the link between the O 2 uptake recovery might be related to the exercise intensity and the nature of repeated sequences in the IT and SIT. The ability to resist fatigue (SmO 2 % decrement) and replenish the energy substrates (ATP and PCr) are oxygen-dependent processes [47]. In the present study, during the rest bouts, muscle oxygenation in different training protocols returned at a different pace to pre-exercise (~80%) levels (Figures 4-6), indicating a possible recovery of muscle PCr [48]. The VL muscle was the least affected, presenting oxygenation at its highest level, and in the RT, the mean values ranged from 71.1% to 76.5%, while in the IT, it ranged from 77% to 81%, and in the SIT, from 76.3% to 80.8% of SmO 2 (Table 4). Different levels of muscle recovery may be related to factors such as impairments in neural drive and motor unit activation or metabolite accumulation [49]. The oxygenation in the muscles quickly adjusted post-exercise, indicating that the use of NIRS technology showed high sensitivity and may lead to discussion and further investigations as to whether oximetry and HR monitoring are more sensitive methods, especially in the IT ( Figure 4) and SIT ( Figure 6). The three training protocols elicited different increases in blood lactate concentrations during the exercise, showing the contribution of the anaerobic glycolytic system, inferred by blood lactate accumulation, to be numerically greater in the RT (3.5 mmol·L −1 ) ( Figure 5) than in the IT (1.4 mmol·L −1 ) ( Figure 4) and SIT (1.8 mmol·L −1 ) ( Figure 6). The benefit of the relief intensity has often been discussed via changes in blood lactate concentration [50]; however, neither blood [51] nor muscle lactate has a direct (nor linear) relationship with performance capacity [50]. It has also been shown that substantially different intermittent training modalities (as assessed by accumulated Bla-1 levels and the HR) may have relatively similar muscle mean peripheral O 2 responses.   Muscle oxygen saturation (SmO2) kinetics in latissimus dorsi, pectoralis major, and vastus lateralis and heart rate (HR) response during RT protocol; blood lactate (Bla) concentration 3 min after the end of the protocol. Figure 6. Muscle oxygen saturation (SmO2) kinetics in latissimus dorsi, pectoralis major, and vastus lateralis and heart rate (HR) response during SIT protocol; blood lactate (Bla) concentration 3 min after the end of the protocol.

O2 Responses in Different Muscles
Information about simultaneous oxygenation in different muscles provides a potential understanding of internal load. Paquette and Bieuzen [30] aimed to understand Figure 6. Muscle oxygen saturation (SmO 2 ) kinetics in latissimus dorsi, pectoralis major, and vastus lateralis and heart rate (HR) response during SIT protocol; blood lactate (Bla) concentration 3 min after the end of the protocol.

O 2 Responses in Different Muscles
Information about simultaneous oxygenation in different muscles provides a potential understanding of internal load. Paquette and Bieuzen [30] aimed to understand muscle oxygenation in more than one active muscle and suggested that the maximum O 2 extraction is independent and a better performance predictor than the VO 2 max in sprint canoeing and kayaking. Thus, our main results on muscle oxygenation during the RT show differences between SmO 2 in the LD, PM, and VL and between the tHb in the VL, PM, and LD during all workout intervals (Figure 2). This could suggest higher PM recruitment during ergometer paddling than the LD, and especially with the VL, in the applied intermittent training sessions. The ∆ tHb was lower in the VL than in the PM and LD across all the training protocols, suggesting a decrease in the leg muscle's blood volume (Table 4). This is in line with previous studies that showed a higher energy requirement of the fatigued muscle per unit of external work performed than the non-fatigued muscle [52]. The deoxygenation of the LD in the SIT was higher than during the other protocols, which was confirmed by a previous electromyography study on different muscle activation levels during kayak paddling, which showed that the LD muscle is highly active during the draw phase of the kayaking [52]. However, an increase in different muscle activation levels during different training protocols, which will likely produce an increase in O 2 extraction, may be associated with the technique required to cover the distances of different intensities. The tHb was lower in the VL (12.4 ± 0.1) than in the PM (13.2 ± 0.1) and LD (13.3 ± 0.1), suggesting a decrease in the muscle blood volume in lower body muscles. The drop in O 2 saturation in the less active muscles is explained by the sympathetic flow induced by exercise, promoting vasoconstriction in this tissue and consequently, a redirection of the blood flow to the more active muscles [53]. This way of explanation about the decreased muscle oxygenation in the non-exercising limb was already used during graded leg cycling exercises, by adopting ultrasound and NIRS methods [54]. At the same time, we did not observe any differences in the tHb between the RT, IT, and SIT in the LD during the exercise bouts and in the VL during the rest bouts, which should be considered in future studies.
Despite some limitations of the NIRS technique and its technology [55], this study was conducted during a real training scenario in the preparatory training period for the world-class kayak competition. Our study was limited to one participant to find out the individual response to single kayak training. Intermittent training is associated with aerobic and anaerobic metabolism; therefore, for practical reasons, it was not possible to invasively determine the accumulation of Bla after each exercise bout by measuring this level at the end of the training. However, previous studies of elite kayakers have shown [56] that the mean lactate threshold occurred at a blood lactate concentration of 2.7 mmol·L −1 , an HR of 170 beats·min −1 , and a VO 2 of 44.2 mL·kg −1 ·min −1 . The lactate threshold presented corresponded to a percentage of 89.6% of the maximum heart rate and 82.4% of the VO 2 peak. This shows that the characteristics of our subject are close to these indicators. Therefore, the relationships between oxygen kinetics and anaerobic metabolism should be further examined with experimental training studies.

Conclusions
The current results suggest that the observations of intermittent exercise performance and significant changes in the peripheral effect of muscle oxygenation in response to training stimuli are the internal predictors of the aerobic metabolism intensity related to work, relief, and recovery intensity. Differences in muscle oxygenation suggest muscle recruitment between the PM, LD, and VL during different exercises; however, this area is still poorly understood requiring further research. To our knowledge, this is the first study that shows the significant contribution of the PM muscle on individual performance in world-class kayakers following different modality intermittent kayak training. In addition to the HR, blood lactate, and VO 2 measurements, wearable NIRS technology is, therefore, a significant tool for monitoring muscle oxidative metabolism during different training modalities.