Physiological Features of Olympic-Distance Amateur Triathletes, as Well as Their Associations with Performance in Women and Men: A Cross-Sectional Study

his study was to verify the physiological and anthropometric determinants of triathlon performance in female and male athletes. This study included 40 triathletes (20 male and 20 female). Dual-energy X-ray absorptiometry (DEXA) was used to assess body composition, and an incremental cardiopulmonary test was used to assess physiological variables. A questionnaire about physical training habits was also completed by the athletes. Athletes competed in the Olympic- distance

and anthropometric determinants of triathlon performance in female and male athletes. This study included 40 triathletes (20 male and 20 female). Dual-energy X-ray absorptiometry (DEXA) was used to assess body composition, and an incremental cardiopulmonary test was used to assess physiological variables. A questionnaire about physical training habits was also completed by the athletes. Athletes competed in the Olympic- distance triathlon race. For the female group, the total race time can be predicted by VO 2max ( = 131 , t = 6.61,p< 0.001), lean mass ( = 61.4, t = 2.66,p= 0.018), and triathlon experience ( = 886.1, t = 3.01,p= 0.009) (r 2 = 0.825,p< 0.05). For the male group, the total race time can be predicted by maximal aerobic speed ( = 294.1, t = 2.89,p= 0.010) and percentage of body fat ( = 53.6, t = 2.20,p= 0.042) (r 2 = 0.578,p< 0.05). The variables that can predict the performance of men are not the same as those that can predict the triathlon performance of women. These data can help athletes and coaches develop performance-enhancing strategies. Keywords:triathlon; ports physiology; performance; women; maximal aerobic speed 1. Introduction Although female participation in triathlon is still lower than male participation (25–40%), there has been a signi cant increase in female participation in this sport since 1990 [1–3]. Female participation has increased not only in triathlon, but also in several other sports. In running, female participation reached the same percentage as male participation in 2018, and at the most recent Olympic Games (Tokyo 2020), female participation set a new record, reaching 49% [4]. Despite a recent surge of female participation in sports, most scienti c studies on sports sciences continue to focus on men [4]. Therefore, the results of the studies conducted on male athletes on sports training are applied to both male and female athletes, despite the lack of a reasonable scienti c justi cation [5,6]. The literature's consensus that maximal oxygen uptake ( VO2max), the percentage of VO2max that can be sustained for an extended period of time, the running economy, and body composition are important variables associated

studies conducted on male athletes on sports training are applied to both male and female athletes, despite the lack of a reasonable scienti c justi cation [5,6]. The literature's consensus that maximal oxygen uptake ( VO2max), the percentage of VO2max that can be sustained for an extended period of time, the running economy, and body composition are important variables associated with performance in long-distance events [7]. However, the relative importance of each one can differ for male or female Healthcare2023,11, 622.

Healthcare2023,11, 622 2 of 12 performance, once there are several physiological (red cell mass, hemoglobin, muscular ber type percentage, muscle capillarization, vasodilatory capacity, and energetic substrate use) and body composition (fat mass percentage, lean mass) differences between sexes [8]. The variables associated with triathlon performance have previously been studied. There- fore, prediction equations for triathlon performance have also been developed; however, previous studies have only included male athletes, or when female athletes were included, the sex was not evaluated as a biological variable [9,10]. The knowledge of the main pre- dictive variables of performance for each sex separately, can help coaches in directing the training sessions to obtain adaptive responses of the most important predictive variables of performance and optimize the improvement of sports performance. Therefore, the purpose of this study was to con rm the levels of association between physiological and body composition variables and performance in an Olympic-distance triathlon test for each sex and, later, to describe a performance prediction equation for the modality based on the measured variables for each sex. Second, the study aimed to compare physiological and body composition variables between sexes. We hypothesized that the level of association between the measured variables and triathlon race results would differ between sexes, and thus the Olympic-distance triathlon race prediction equations would differ for each sex. 2. Materials and Methods 2.1. Participants Triathletes were invited to take part in the study via social media (WhatsApp, Insta- gram, and email), as well as folders distributed at triathlon competitions. The inclusion criteria for participating in the study were being enrolled in the 30th Santos International Olympic Triathlon in February 2022, between the ages of 18 and no more than 61 years, both sexes, training triathlon regularly for at least 6 months, and having a medical al- lowance. The exclusion criteria included being pregnant, having competed in the alternate modality, failure to nish the race, or failure to submit to laboratory tests for any reason. Initially, 42 athletes were selected to participate in the study (22 men and 20 women). Two male athletes were excluded from the study, one was

6 months, and having a medical al- lowance. The exclusion criteria included being pregnant, having competed in the alternate modality, failure to nish the race, or failure to submit to laboratory tests for any reason. Initially, 42 athletes were selected to participate in the study (22 men and 20 women). Two male athletes were excluded from the study, one was due to the fact that he did not nish the race and one was due to the fact that he failed to submit to laboratory tests. As a result, the study includes 20 male athletes and 20 female athletes. Data were collected during the pre-season. Table Table 1.Anthropometric data for men and women. Women (n = 20) Men (n = 20) p-Value Cohen's d Power Age (years) 42.7 7.3 (3.9–45.9) 43.7 9.3 (38.2–46.4) 0.880 0.05 0.052 Body mass (kg) 58.8 6.7 (55.9–61.8) 74.8 6.9 (71.8–77.9) <0.001 2.35 0.999 Height (cm) 165.0 5.7 (163.0–168.0) 175.0 8.2 (171.0–178.0) <0.001 1.35 0.986 Fat mass (kg) 13.3 7.2 (9.9–16.9) 12.8 5.0 (10.6–15.0) 0.826 0.07 0.055 Lean mass (kg) 42.2 6.5 (45.1–39.4) 58.3 5.8 (55.8–60.8) <0.001 2.60 1.000 % Body fat 23.3 11.3 (18.3–28.2) 17.8 6.3 (15.0–20.6) 0.066 0.60 0.453 % Gynoid fat 53.2 6.0 (50.6–55.9) 39.1 5.1 (36.8–41.3) <0.001 2.53 1.000 % Android fat 42.9 6.6 (40.0–45.8) 57.1 5.4 (54.7–59.5) <0.001 2.36 0.999 Mean standard deviation. Con dence interval: 95%.

Healthcare2023,11, 622 3 of 12 2.2. Procedures All experimental procedures followed the Declaration of Helsinki and the Recom- mendations for the Conduct, Reporting, Editing, and Publication of Scholarly Work in Medical Journals. The study was approved by the Human Ethics Committee of the Uni- versity Federal of S¢o Paulo-UNIFESP (approval number 5.059.538, 25 October 2021). The participants were given information about the purpose of the research, all of the proposed physiological laboratory tests, and the risks and bene ts. The researchers justi ed the prin- ciples of respect for the volunteers, as well as the guarantee of privacy, con dentiality, and anonymity rights. After all, every participant signed the informed consent form. Initially, the volunteers completed an online questionnaire from the Google Forms Platform about their training habits and medical conditions. Then, they attend the Exercise Physiology Laboratory at UNIFESP once, during the morning period. During the visit, volunteers were measured for height, body mass, and body composition. Thereafter, participants were submitted to a running economy evaluation. After 30 min of rest, participants were subjected to a cardiorespiratory maximal treadmill test. They were instructed to abstain from strenuous training in the last 24 h before the test and not consume hyper-stimulating foods on the day (e.g., caffeine). Wearing light clothes and comfortable running shoes was also recommended. The organizers provided the results of the total race time and split times of the competition, which were taken from the of cial website of the event (https://www. internacionaldesantos.com.br, accessed on 5 March 2022). All the data were collected in January 2022 (pre-competition phase), 1 month before the 30th Santos International Olympic Triathlon. 2.3. Questionnaire The questionnaire includes two open questions about their medical condition: Do you have any chronic diseases? Do you take any medications? The questionnaire also includes four open questions about their training habits: How many hours per week do you cycle train? How many hours per week do you train for running? How many hours per week do you train for swimming? How many days, months, or years have you been training for a triathlon? 2.4. Morphological Variables Body

take any medications? The questionnaire also includes four open questions about their training habits: How many hours per week do you cycle train? How many hours per week do you train for running? How many hours per week do you train for swimming? How many days, months, or years have you been training for a triathlon? 2.4. Morphological Variables Body mass and height were measured to the nearest 0.1 kg and 0.1 cm using a cali- brated stadiometer Filizola ® PL (Filizola, S¢o Paulo, SP, Brazil), respectively. Body composi- tion was determined using dual-energy X-ray absorptiometry (DEXA, software version 12.3, Lunar DPX, GE Healthcare, Madison, WI, USA). The volunteers were instructed to drink water ad libitum and were not given any instructions about fasting or taking any speci c feedings prior to the procedure. They were all evaluated after bladder voiding [11]. These procedures had been shown to be a reliable method for assessing body composition [12,13]. 2.5. Running Economy Test The volunteers were subjected to a treadmill running test for 4 min on a motorized treadmill (Inbrasport, ATL, Porto Alegre, Brazil) using a computer-based breath, via a breath gas exchange analyzer (Quark, Comedy, Italy) at a constant speed of 8 km/h, which was below the ventilatory threshold (VT) for all volunteers. Prior to each test, the calibration procedure was carried out according to the manufacturer's instructions. The last minute was taken into account when calculating the average oxygen uptake, CO2production, and respiratory exchange rate (RER). The RER should be lower than 1.0., as all the participants were exercising lower than the ventilatory threshold intensity. According to Silva et al. [14], these variables were used to calculate the oxygen cost and the energy cost of running.

Healthcare2023,11, 622 4 of 12 2.6. Cardiorespiratory Maximal Treadmill Test After a 30-min recovery period from the running economy test, which was enough to return all volunteers' heart rates to rest levels, they were subjected to the cardiores- piratory maximal treadmill test. The same computer-based metabolic analyzer (Quark, Comedy, Italy) was used to measure VO2max, VT, and respiratory compensation point (RCP). The maximal aerobic speed (MAS) was also measured. VO2max was de ned as a stable increase in oxygen uptake (less than 2.1 mL/kg/min) even after increasing exercise intensity [15]. VT was calculated using the following criteria: An increase in the ventilatory equivalent for oxygen without an increase in the equivalent for carbon dioxide and an increase in end-tidal pressure of oxygen. The RCP was determined by increasing the CO2 equivalent ventilatory and decreasing the end-tidal pressure of CO2[16]. Two independent investigators determined VT and RCP, and a third researcher was consulted in the event of disagreement. The MAS was de ned as the lowest exercise intensity that produced VO2max [10]. 2.7. Total Race and Split Time Results Total race and split time results were provided by the organizers of the 30th Santos International Olympic Triathlon, which were accessed via the of cial website of the event (https://www.internacionaldesantos.com.br/, accessed on 5 March 2022). 2.8. Statistical Analysis The data were presented in the form of mean and standard deviations. According to the Kolmogorov–Smirnov and Levene's tests, all variables had a normal distribution and homogeneous variability. A Student'st-test for independent samples was used to compare the variables of male and female athletes. To determine the magnitude of the differences, between group effect sizes were computed for each outcome. Using Cohen's effect sizes, the magnitude of any change was judged according to the following criteria: d< 0.2 was considered as ignored; 0.2 d< 0.5 was considered as a “small” effect size; 0.5 d< 0.8 represented a “moderate” effect size; 0.8 d< 1.3 a “large” effect size; andd 1.3 a “very large” effect [17]. The Pearson linear correlation coef cient and dispersion diagrams were used to validate the level of association between each split

following criteria: d< 0.2 was considered as ignored; 0.2 d< 0.5 was considered as a “small” effect size; 0.5 d< 0.8 represented a “moderate” effect size; 0.8 d< 1.3 a “large” effect size; andd 1.3 a “very large” effect [17]. The Pearson linear correlation coef cient and dispersion diagrams were used to validate the level of association between each split time and total race time with other measured variables. To compare the triathlon race time to the previous triathlon experience, a one-way ANOVA was used. Thereafter, we considered the signi cant correlations for the stepwise adjustment of the multiple linear regression model. The formula of the regression model is x = + y + E, where x is the dependent variable, y is the independent variable, is the intercept, is the slope, and E is the residual. For each regression equation, the coef cient of determination (r 2 ), a number that measures how well a statistical model predicts an outcome, was presented. For all regression models presented, Durbin–Watson Test (to detect autocorrelation), variance in ation factor (VIF) and tolerance (to detect multicollinearity), the normality of the distribution of residuals, and Q-Q plot (to detect homoscedasticity) were presented. The G*Power version 3.1.9.2 (Franz, Universität Kiel, Germany) was used to determine the sample size and analyze the test power level. A sample size calculation for regression analysis for overall race time with two predictors, using previous published data from Puccinelli et al. [10] (r 2 = 0.607), showed that 20 athletes were needed to detect a relevant difference with 80% power and a signi cance level of 5%. The powers of the analyses were also calculated. The analyses were carried out using the IBM SPSS Statistics (version 22, USA) software, with the level of signi cance set atp< 0.05. 3. Results In terms of physiological variables, men had signi cantly higher values for VO2max (L/min) (p< 0.001,d= 2.74), VO2max (mL/kg/min) (p= 0.020,d= 0.769), MAS (p< 0.001, d= 1.30), VT speed (p= 0.011,d= 0.843), and RCP speed (p< 0.001,d= 1.78). The results showed no signi cant difference in the percentage of VO2max

the level of signi cance set atp< 0.05. 3. Results In terms of physiological variables, men had signi cantly higher values for VO2max (L/min) (p< 0.001,d= 2.74), VO2max (mL/kg/min) (p= 0.020,d= 0.769), MAS (p< 0.001, d= 1.30), VT speed (p= 0.011,d= 0.843), and RCP speed (p< 0.001,d= 1.78). The results showed no signi cant difference in the percentage of VO2max at VT (p= 0.129,d= 0.490)

Healthcare2023,11, 622 5 of 12 or RCP (p= 0.558,d= 0.173) between men and women. Running economy, as measured by oxygen cost or energy cost, did not differ signi cantly between sexes (p= 0.540,d= 0.196 andp= 0.600,d= 0.167, respectively) (Table). Table 2.Measured variables in the cardiorespiratory maximal test and performance for each sex. Women (n = 20) Men (n = 20) p-Value Cohen's d Power Cardiorespiratory maximal test VO 2max (L/min) 2.90 0.39 (2.72–3.07) 4.08 0.47 (3.88–4.28) <0.001 2.74 1.00 VO 2max (mL/kg/min) 49.7 7.6 (46.3–53.0) 54.6 5.0 (52.4–56.8) 0.020 0.769 0.659 MAS (km/h) 14.9 1.8 (14.1–15.7) 17.1 1.6 (16.4–17.8) <0.001 1.30 0.979 VO 2at VT (mL/kg/min) 38.0 6.3 (35.3–40.0) 40.4 4.0 (38.6–42.2) 0.164 0.499 0.336 % VO 2max at VT 76.2 5.3 (73.9–78.6) 73.8 4.7 (71.7–75.9) 0.129 0.490 0.326 Speed at VT (km/h) 10.7 1.6 (10.0–11.4) 11.8 1.1 (11.4–12.3) 0.011 0.843 0.738 VO 2at RCP (mL/kg/min) 45.1 6.9 (42.1–48.1) 49.1 4.7 (47.4–51.1) 0.025 0.736 0.621 % VO 2max at RCP 90.5 3.9 (88.7–92.2) 89.8 3.7 (88.2–91.4) 0.588 0.173 0.083 Speed at RCP (km/h) 12.8 1.6 (12.0–13.5) 14.4 1,2 (13.9–15.0) <0.001 1.78 0.999 Oxygen cost (mL/kg/km) 226.0 20.3 (217.0–235.0) 222.0 18.1 (214.0–230.0) 0.540 0.196 0.092 Energy cost (Kcal/kg/km) 1.11 0.09 (1.07–1.15) 1.09 0.09 (1.05–1.13) 0.600 0.167 0.080 Race performance Swimming (seconds) 2072 518 (1844–2219) 1796 265 (1680–1912) 0.041 0.670 0.670 Cycling (seconds) 4274 405 (4097–4452) 3969 329 (3825–4114) 0.013 0.826 0.720 Running (seconds) 3143 546 (2904–3382) 3060 371 (2853–3266) 0.608 0.163 0.079 Total race time (seconds) 9489 1357 (8894–10083) 8825 920 (8421–9228) 0.078 0.573 0.423 Mean standard deviation. Con dence interval: 95%. MAS: Maximal aerobic speed; VO2at VT: VO2at ventilatory threshold; % VO2max at VT: % VO2max at ventilatory threshold; VO2at RCP: VO2at respiratory compensation point; % VO2max at RCP: % VO2max at respiratory compensation point. Men were signi cantly faster in the swimming split (p= 0.041,d= 0.670) and cycling split (p= 0.013,d= 0.826) of the Olympic-distance triathlon race, but there was no signi cant difference in the running split (p= 0.608,d= 0.163) or total race time (p= 0.078,d= 0.573) (Table). There was no signi cant difference in the years of

VO2max at respiratory compensation point. Men were signi cantly faster in the swimming split (p= 0.041,d= 0.670) and cycling split (p= 0.013,d= 0.826) of the Olympic-distance triathlon race, but there was no signi cant difference in the running split (p= 0.608,d= 0.163) or total race time (p= 0.078,d= 0.573) (Table). There was no signi cant difference in the years of triathlon experience (p= 0.807) between women [3(1–3)] and men [3(1–3)]. The level of association between each split and total race performance with body composition or physiologic variables was presented in Table.

This study investigates physiological features and performance associations in Olympic-distance amateur triathletes.