Cardiorespiratory Coordination in Collegiate Rowing: A Network Approach to Cardiorespiratory Exercise Testing

legiate rowing performance is often assessed by a cardiopulmonary exercise test (CPET). Rowers' on-water performance involves non-linear dynamic interactions and synergetic recon gura- tions of the cardiorespiratory system. Cardiorespiratory coordination (CRC) method measures the co-variation among cardiorespiratory variables. Novice (n= 9) vs. Intermediate (n= 9) rowers' CRC (H 0: Novice CRC = Intermediate CRC; H A: Novice CRC < Intermediate CRC) was evaluated through principal components analysis (PCA). A female NCAA Division II team (N= 18) grouped based on their off-water performance on 6000 m time trial. Rowers completed a customized CPET to exhaus- tion and a variety of cardiorespiratory values were recorded. The number of principal components (PCs) and respective PC eigenvalues per group were computed on SPSS vs28. Intermediate (77%) and Novice (33%) groups showed one PC 1. Novice group formed an added PC 2due to the shift of expired fraction of oxygen or, alternatively, heart rate/ventilation, from the PC 1cluster of examined variables. Intermediate rowers

of cardiorespiratory values were recorded. The number of principal components (PCs) and respective PC eigenvalues per group were computed on SPSS vs28. Intermediate (77%) and Novice (33%) groups showed one PC 1. Novice group formed an added PC 2due to the shift of expired fraction of oxygen or, alternatively, heart rate/ventilation, from the PC 1cluster of examined variables. Intermediate rowers presented a higher degree of CRC, possible due to their increased ability to utilize the bicarbonate buffering system during the CPET. CRC may be an alternative measure to assess aerobic tness providing insights to the complex cardiorespiratory interactions involved in rowing during a CPET. Keywords: complex adaptive systems; coordinative variables; dynamic networks; network physiology; intra-individual co-variability; dynamic couplings; principal component analysis; time-series analysis; cardiovascular system; athlete's performance evaluation 1. Introduction Cardiopulmonary exercise testing (CPET) and the related concept of maximal oxygen consumption (VO2max) is considered as the most important indicator of endurance capacity, cardiorespiratory tness, and health in sports science [1–8]. It measures a variety of vari- ables (i.e., ventilation, VE; heart rate, HR; oxygen saturation, SpO2; ventilatory threshold, VT; expired fraction of oxygen, FeO2; expired fraction of carbon dioxide, FeCO2) linked to metabolic, cardiovascular and pulmonary responses during the CPET [7,9]. According to this concept, the human body has a limited capacity to utilize oxygen for muscle work as demonstrated by a plateau in VO2maxthat indicates a physiological ceiling in cardiorespira- tory capacity [10–12]. Since VO2maxis an important outcome for physical performance, a variety of tests designed to access VO2maxusing a treadmill or cycle ergometer have been used, with the most popular ones to be the ramp and graded maximal incremental exercise tests [13–16]. Rowing as a sport has different biomechanical characteristics compared to running and cycling, so there are few rowing-speci c VO2maxtests using work rate as a key element for the CPET setup on a Concept II rowing ergometer (i.e., stroke rate, critical velocity, Int. J. Environ. Res. Public Health2022,19, 13250.

are few rowing-speci c VO2maxtests using work rate as a key element for the CPET setup on a Concept II rowing ergometer (i.e., stroke rate, critical velocity, Int. J. Environ. Res. Public Health2022,19, 13250.

Int. J. Environ. Res. Public Health2022,19, 13250 2 of 11 power output) [17–20]. The most popular protocol to assess off-water VO2maxin rowers has been the simulation of the 2000 m distance time trial with maximal individual intensity with no differences observed compared to incremental exercise testing untilexhaustion [19–25] . Competitive collegiate rowing requires a high physical demand [26], with metabolic contri- bution to 70–88% aerobic and 12–30% anaerobic [22,27]. Due to the high contribution of the aerobic system in rowing performance, direct VO2maxmeasurement has been shown to be highly correlated (e.g., r = 0.85–0.88) with actual rowing performance and to be the most important physiological determinant of rowing performance [20,27–29]. Therefore, the gold standard to assess direct off-water VO2maxin rowing is through gas analysis in a rowing ergometer (e.g., Concept II). Such off-water VO2maxtest have reported maximum values of more of 6 L of oxygen per minute (L/min) or values close to 65 millimeters per kilogram of body weight per minute (mL/kg/min) [18,30–40], while collegiate female rowers have reported VO2maxvalues of 58 to 65 mL/kg/min [41]. Being able to measure and track the physical demands of rowers is crucial to both training and success in rowing [26]. Moreover, even though is dif cult to ensure a controlled load output on rowing ergometers during a CPET [42], such a CPET allows for observing and describing both external (i.e., maximal power output) [34,42–44] and internal loads (i.e., VO 2max, HR) [45–49]. The inability of the most commonly VO2maxused tests to capture maximum perfor- mances is due to their inability to capture the non-linear dynamic interactions among different physiological systems [50,51]. It has been proposed that the synergism of the cardiovascular and respiratory systems during a sub maximum or maximum exercise stim- ulus cannot be captured by the traditional VO2maxtests [52]. The framework of Network Physiology and, more speci cally, Network Physiology of Exercise (NPE) [51,53–58] utilizes non-linear modeling and time series analysis of coordinative variables to investigate how different physiological systems coordinate and synchronize as a network [59–62]. For ex- ample, cardiorespiratory coordination (CRC) has been recently proposed as an alternative method

cannot be captured by the traditional VO2maxtests [52]. The framework of Network Physiology and, more speci cally, Network Physiology of Exercise (NPE) [51,53–58] utilizes non-linear modeling and time series analysis of coordinative variables to investigate how different physiological systems coordinate and synchronize as a network [59–62]. For ex- ample, cardiorespiratory coordination (CRC) has been recently proposed as an alternative method to assess the synergism between the cardiorespiratory variables during an exer- cise stimulus [52,63–67]. This method requires the use of principal component statistical analysis (PCA) performed on time series cardiorespiratory data. Based on this statistical approach, the time synchronization as derived from different physiological systems re ects their shared co-variation between the involved physiological parameters, which at the end is represented through fewer principal components (PCs). The respective PCs are extracted in a decreasing fashion and represent the maximum covariance between the examined physiological variables, with the total number of PCs to represent the coordination level among them [67]. It is stated that the decrease in PCs and/or the increase PC eigenvalues re ects a higher level of cardiorespiratory ef ciency-coordination [52,68]. Along this line of thought, recent work from our lab analyzed postprandial network interactions between autonomic nervous system and lipemia data in response to acute partial sleep deprivation and high-intensity interval exercise under the NPE framework. Even though we did not perform a true network analysis due to a non-time series data col- lection methodology, we reported that negative links in short-sleep high-intensity interval exercise (HIIE) condition re ected the in uence of sleep on both the autonomic regulation and lipemia [69]. A time series analysis would have allowed us to both capture the syner- gistic recon guration of the autonomic nervous system and cardiometabolic lipemia and identify the causality between the physiological signals by creating a physiological post- prandial network [59,66,70]. In contrast, when the same research question was analyzed using in uential statistics, HIIE was cardioprotective regardless the impact of the short sleep [71,72]. It is clear that a research question analyzed under different frameworks may lead to different conclusions. Therefore, analyzing cardiorespiratory testing under the

the causality between the physiological signals by creating a physiological post- prandial network [59,66,70]. In contrast, when the same research question was analyzed using in uential statistics, HIIE was cardioprotective regardless the impact of the short sleep [71,72]. It is clear that a research question analyzed under different frameworks may lead to different conclusions. Therefore, analyzing cardiorespiratory testing under the NPE framework may reveal the non-liner dynamic interactions of the cardiorespiratory system. Evidence supports the notion that CRC may be implicated to higher training adapta- tions [52] and training load [64] than VO2maxand other markers of aerobic tness, pointing out CRC's potential to substitute the traditional markers of aerobic capacity. Although

Int. J. Environ. Res. Public Health2022,19, 13250 3 of 11 metabolic, cardiovascular, and pulmonary systems work in coordination during a CPET, there is scarce evidence related to CRC in collegiate rowing CPET. Therefore, this study investigated the measurement of CRC through PCA in collegiate rowers. We hypothesized that Intermediate rowers compared to Novice rowers will have higher CRC values. Authors seek to offer an alternative and more meaningful interpretation of rowers' performance and training for maximal adaptation and prevent undertraining and overtraining. Results from this study may enrich the information provided by the traditional rowing ergometry tests to assess VO2max[73–76]. 2. Materials and Methods 2.1. Study Design and Participants A cross-sectional observational study was conducted on a female rowing team that competes in NCAA Division II, as part of their annual pre-season physiological measure- ments. Rowers performed a preliminary rowing test to establish the initial rowing power for the customized discontinuous incremental rowing test to exhaustion. Based on the preliminary testing results, rowers were divided to Novice and Intermediate ones (Figure). Athletes' direct off-water VO2maxcapacity was measured through gas analysis using a Con- cept II rowing ergometer and a discontinuous incremental rowing protocol [77,78]. Besides measuring their VO2maxcapacity, demographics and anthropometrics were also recorded.Int. J. Environ. Res. Public Health 2022, 19, x 3 of 12 framework may reveal the non-liner dynamic interactions of the cardiorespiratory system. Evidence supports the notion that CRC may be implicated to higher training adaptations [52] and training load [64] than VO 2max and other markers of aerobic fitness, pointing out CRC’s potential to substitute the traditional markers of aerobic capacity. Although meta- bolic, cardiovascular, and pulmonary systems work in coordination during a CPET, there is scarce evidence related to CRC in collegiate rowing CPET. Therefore, this study investigated the measurement of CRC through PCA in colle- giate rowers. We hypothesized that Intermediate rowers compared to Novice rowers will have higher CRC values. Authors seek to offer an alternative and more meaningful inter- pretation of rowers’ performance and training for maximal adaptation and prevent un- dertraining and overtraining. Results from this study may enrich the information

this study investigated the measurement of CRC through PCA in colle- giate rowers. We hypothesized that Intermediate rowers compared to Novice rowers will have higher CRC values. Authors seek to offer an alternative and more meaningful inter- pretation of rowers’ performance and training for maximal adaptation and prevent un- dertraining and overtraining. Results from this study may enrich the information pro- vided by the traditional rowing ergometry tests to assess VO 2max [73–76]. 2. Materials and Methods 2.1. Study Design and Participants A cross-sectional observational study was conducted on a female rowing team that competes in NCAA Division II, as part of their annual pre-season physiological measure- ments. Rowers performed a preliminary rowing test to establish the initial rowing power for the customized discontinuous incremental rowing test to exhaustion. Based on the preliminary testing results, rowers were divided to Novice and Intermediate ones (Figure 1). Athletes’ direct off-water VO 2max capacity was measured through gas analysis using a Concept II rowing ergometer and a discontinuous incremental rowing protocol [77,78]. Besides measuring their VO 2max capacity, demographics and anthropometrics were also recorded. As part of team ‘s requirements, all rowers (n = 18, age = 20.17 ± 2.28 SD years) sup- plied their consent to have both their VO 2max capacity and body composition assessed. All testing was performed on a single day by the same research personnel during morning hours in an air-conditioned levorotatory environment when no team practice was sched- uled for 24 h prior to testing. The study was approved by the Ethics Committee of Barry University’s institutional review board #1851725-2 based on established policies on class- room and student research. Figure 1. Research design. 2.2. Procedures 2.2.1. Anthropometrics and Body Composition Height and weight were determined using an electronic scale and stadiometer (Seca 703), with participants removing their shoes prior to stepping on the scale [79]. Body com- position was measured via a bipolar digital bioimpedance system with tatcile poles (OM- RON Body Fat Analyzer, HBF-306BL, Omron Healthcare Corporation, Kyoto, Japan) and body fat percentage (%) was calculcated following standard procedures as previously de- scribed. Briefly, participants

an electronic scale and stadiometer (Seca 703), with participants removing their shoes prior to stepping on the scale [79]. Body com- position was measured via a bipolar digital bioimpedance system with tatcile poles (OM- RON Body Fat Analyzer, HBF-306BL, Omron Healthcare Corporation, Kyoto, Japan) and body fat percentage (%) was calculcated following standard procedures as previously de- scribed. Briefly, participants visited lab in fasting condition, with no eating and consum- ing water 2 h before testing and abstain from exercise 24 h prior to testing. All Anthropometrics and Body Composition via Bioelectric Impedance Analysis Preliminary Rowing test to establish the initial power for the Customized Discontinuous Incremental Rowing Cardiopulmonary Exercise Testing protocol (CPET) Based on the results from the Preliminary Rowing Rest, athletes were grouped to Novice and Intermediates Customized Discontinuous Incremental Rowing Cardiopulmonary Exercise Testing protocol (CPET) Figure 1.Research design. As part of team `s requirements, all rowers (n= 18, age = 20.17 2.28 SD years) supplied their consent to have both their VO2maxcapacity and body composition assessed. All testing was performed on a single day by the same research personnel during morn- ing hours in an air-conditioned levorotatory environment when no team practice was scheduled for 24 h prior to testing. The study was approved by the Ethics Committee of Barry University's institutional review board #1851725-2 based on established policies on classroom and student research. 2.2. Procedures 2.2.1. Anthropometrics and Body Composition Height and weight were determined using an electronic scale and stadiometer (Seca 703), with participants removing their shoes prior to stepping on the scale [79]. Body composition was measured via a bipolar digital bioimpedance system with tatcile poles (OMRON Body Fat Analyzer, HBF-306BL, Omron Healthcare Corporation, Kyoto, Japan) and body fat per- centage (%) was calculcated following standard procedures as previously described. Briefly, participants visited lab in fasting condition, with no eating and consuming water 2 h before testing and abstain from exercise 24 h prior to testing. All measurements happened during noon time, before lunch, and approximately 20 min prior to the cardiopulomonary exercise testing [ 2.2.2. Preliminary Customized Discontinuous Incremental Rowing Protocol—Initial Rowing Power

following standard procedures as previously described. Briefly, participants visited lab in fasting condition, with no eating and consuming water 2 h before testing and abstain from exercise 24 h prior to testing. All measurements happened during noon time, before lunch, and approximately 20 min prior to the cardiopulomonary exercise testing [ 2.2.2. Preliminary Customized Discontinuous Incremental Rowing Protocol—Initial Rowing Power A customized discontinuous incremental rowing protocol based on rowers' 60 s rowing speed was employed. According to this, rowers had to perform a rowing sprint of 60 s

Int. J. Environ. Res. Public Health2022,19, 13250 4 of 11 as fast as possible they could on a Concept II rowing ergometer to determine the initial power output for the actual cardiopulmonary exercise test (CPET). A priori power output of 250 Watts for at least 10 strokes in 60 s was set as cutoff point in order to classify the rowers into the Intermediate (>250 W) or to Novice (<250 W) group [77,78]. There is no clear consensus in the literature on Novice/Freshman rowers or DII rowers regarding the testing protocols and how to establish the starting wattage [77,78]. Due to this, we used a practical eld approach to determine the starting power output. It is a widespread practice in rowing coaches to base the starting Wattage output on the performance of the 1-min all-out test. According to this common eld practice, if a rower could maintain a power output of ~250 Watts for 10 to 15 strokes, then it was expected to at least make it to the fth stage of the test and these values to represent a realistic oxygen consumption. However, if a rower could not achieve this initial output and strokes rate, then it was assumed that the lack of power is the reason why the test was terminated, without achieving a realistic maximum oxygen consumption [26,36,41,81,82]. 2.2.3. Cardiopulmonary Exercise Testing (CPET) Protocol Having set the initial power, rowers completed a customized discontinuous incre- mental rowing test to exhaustion. Prior to the customized CPET, a warmup of 5 min in the Concept II rowing ergometer was performed. Participants were instructed to perform 2 min of easy rowing at a power output of <70 Watts, and the intensity was increased every minute as follows: 1 min between 70–100 Watts, 1 min between 100–130 Watts, and last minute was divided in 2 intervals of 30 s in which the power output was 130–160 Watts and >160 Watts, respectively. Following the warmup, the incremental discontinued protocol was performed. The incremental stages were set at 30 Watts for all rowers, while Intermediate rowers started at 70 Watts and the advanced

70–100 Watts, 1 min between 100–130 Watts, and last minute was divided in 2 intervals of 30 s in which the power output was 130–160 Watts and >160 Watts, respectively. Following the warmup, the incremental discontinued protocol was performed. The incremental stages were set at 30 Watts for all rowers, while Intermediate rowers started at 70 Watts and the advanced ones at 100 Watts [77,78]. The duration of each stage was 2 min with 30 s rest in between. Since this was a maximum test until volitional fatigue exhaustion, rowers were expected to give their absolute best. Participants were encouraged to complete a maximum of 7 stages or until required power output could not be maintained. Following the CPET a cool down period of 3 min was performed, where participants were instructed to continue rowing at a power output of 50–70 Watts at their preferred stroke rate with heart rates to be below 100 beats per minute. 2.2.4. Principal Components Analysis (PCA) To analyze the CRC for each participant, we performed a PCA on the data series of the following selected cardiorespiratory variables: VE, FeO2, FeCO2, and HR. We excluded from the analysis VEqO2, VEqCO2, O2pulse, RER, VO2, etc., due to their known deterministic mathematical relation (linear combination) with the selected variables [52]. There is diverse evidence about the use of dimensionality reduction by PCA in small samples, which indicates certain robustness in the estimates of shared variance that [83] pointed out some time ago. In this sense, it should be noted that the estimates in small samples should be more descriptive than inferential considerations, but appropriate to our objectives [84]. Continuous blood pressure monitoring could not be provided in this study. However, non-published results of our lab have shown similar results while analyzing CRC with and without continuous blood pressure measurement. 2.3. Statistical Analysis To analyze the suitability of the PCA implementation, we calculated Bartlett's test for sphericity and the Kaiser-Mayer-Olkin (KMO) test for all participants. We determined the number of PCs using the Kaiser-Gutmann criterion and thus considered PCs with eigenval- ues 1.00 as signi cant [85]. Given

This study investigates cardiorespiratory coordination in collegiate rowers using a network approach.