Introduction: Cardiorespiratory fitness (CRF) is used to assess the body's ability to perform daily activities and is a criterion of overall health besides endurance performance. Therefore, an estimation of CRF can provide valuable information for healthy people to better understand their risk of developing coronary diseases, as well as for athletes to design tailored training plans to optimize their performance. The gold standard for assessing CRF is a direct measurement of maximal oxygen uptake (VO2max) in laboratory settings, which quantifies oxygen consumption mostly during graded maximal exercise tests [1]. Although direct measurements provide high accuracy, they are time-consuming, costly, and require expertise. Several indirect methods have been developed to estimate VO2max, most of which are field tests. While these tests are more accessible, they often still require technical knowledge or specialized equipment, as most of them are designed primarily for athletes and administered by experts. Therefore, performing these tests and interpreting the results can be challenging for recreational individuals. Various studies have shown that deep learning models outperform linear regression models in predicting VO2max [2]. Previous studies also showed that, to accurately predict VO2max in a laboratory or through indirect tests, individuals need to reach their maximal effort rather than performing submaximal exercises [3]. In this study, we present an approach to predict VO2max from data collected during a maximal effort test on a 600-meter uphill course, using both linear regression and other machine learning models. Methods: The subjects consisted of 10 men and 8 women between the ages of 20 and 39 years. A maximal incremental treadmill test was completed by the 18 participants to assess VO2max values. The treadmill test began at an initial speed of 7 km.h-1 with a 7% incline and the speed was increased by 0.5 km.h-1 every 30 seconds until volitional exhaustion. After a rest period of 3 to 7 days, participants performed an all out test on a 600-meter uphill track with a mean gradient of 9.7%. Data were collected during this field test with participants wearing a Polar Vantage V2 watch and Polar H10 chest belt (Polar Electro Oy, Finland). In addition to the user’s anthropometric data (age, gender, body mass index) and time, we also used the speed to heart rate (HR) ratio as a predictor variable, since a previous study has shown its importance [4]. Multiple linear regression (MLR), XGboost, and a multilayer perceptron (MLP) were used to develop models to predict VO2max. The data set was randomly split into 80% and 20% as training and test set, respectively. For the purpose of overcoming multicollinearity among the predictor variables speed to HR ratio, time, and gender, principal component analysis with two components was applied before we fed the data into the multiple linear regression model. Results: The mean measured VO2max scores were 44.1 ± 4.7 mL.kg-1.min-1 and 55.8 ± 5.8 mL.kg-1.min-1 for women and men, respectively. The total distance was covered in an average time of 3:08 ± 0:41 minutes, with an average speed of 11.7 ± 2.5 km.h-1. Regarding the performance of VO2max estimation, the MLR model achieved an R2 of 0.77 with a standard error of the estimate (SEE) of 3.4 mL.kg-1.min-1, the XGBoost model an R2 of 0.82 with a SEE of 3.0 mL.kg-1.min-1, and the MLP model an R2 of 0.94 with a SEE of 1.6 mL.kg-1.min-1. Discussion/Conclusion: These results suggest that our short, high-intensity field test, when combined with a neural network model, can provide accurate predictions of VO2max. Within our data, we have uncovered a non-linear, complex relationship between VO2max and the predictor variables that a feed-forward neural network with one hidden layer can reliably approximate. This is consistent with the findings of the previous research that emphasizes the power of deep learning models to predict VO2max accurately. One potential limitation of this study is the size of the data, for having more data could unveil other relationships between VO2max and the predictor variables. Furthermore, the performance of the models may also suffer from overfitting due to the small and homogeneous sample population and potential discrepancies between the distributions of the training and the test data. Thus, future work should include increasing the sample size and exploring probabilistic models for VO2max prediction. References: [1] 10.1155/2016/3968393; [2] doi.org/10.1016/j.imu.2022.100863; [3] Buckley, D. J., & Rowe Jr, J. R. (2018). Actual Versus Predicted VO2max: A Comparison of 4 Different Methods. In International Journal of Exercise Science: Conference Proceedings (Vol. 2, No. 10, p. 41); [4] 10.1109/BHI.2017.7897252
Introduction The Swiss Armed Forces have released a fitness app for personalised physical training that takes into account the current endurance performance. This is done by means of an integrated digital self-test of physical performance. The Cooper test (12-minute run), which is widely used in the military setting, is too long for the digital selftest. This study aims to investigate whether a self-paced 4-minute all out outdoor run (4Minmax run) is a valid method to assess endurance performance for personalised training planning in the app. Methods On the same day, the subjects completed a 4Minmax run on a flat 300 m circular track and a maximal exercise test (VO2max-Test) using a graded protocol (figure 1). Average speed was calculated from the 4Minmax run (v4Min), the relative maximum oxygen uptake (VO2max) and the maximum speed (vend) from the VO2max-Test. Maximum heart rate (HRmax) was measured during the 4Minmax run and the VO2max-Test. Respiratory gases were analysed during the VO2max-Test using a mixing chamber (Cosmed srl, Rome, Italy). Pearson correlations and linear regressions were used to test whether v4Min predicted vend and the relative VO2max. Two regression analyses were performed separately for women and men. Results Table 1 shows the descriptive statistics of the 4Minmax run and the VO2max-Test of the 18 subjects. In both men and women, vend and v4Min were strongly correlated (r = 0.79, p = 0.001, 95% CI: 0.424 - 0.934 and r = 0.974, p = 0.005, 95% CI: 0.647 - 0.998, respectively; figure 2). Also the VO2max and the v4Min were strongly correlated in men and in women (r = 0.742, p = 0.004, 95% CI: 0.324 - 0.918 and r = 0.897, p = 0.039, 95% CI: 0.073 - 0.993, respectively; figure 3). The linear regression model of the relative VO2max explained by the v4Min was statistically significant in women (adjusted R2 = 0.74, F(1,3) = 12.41, p = 0.039) and in men (adjusted R2 = 0.51, F(1,11) = 13.504, p = 0.004). The regression equation for women was: VO2max [ml*min-1*kg-1] = -28.792+5.758 (v4Min [km*h-1]). The regression equation for men was: VO2max [ml*min-1*kg-1] = -21.627+4.709*(v4Min [km*h-1]). Conclusion The performance in the 4Minmax run can accurately estimate both VO2max and vend with a slightly higher adjusted R2 between v4Min and vend than between v4Min and VO2max. As the VO2max is comparable to reference values, it is recommended to calculate the estimated VO2max from the 4Minmax run to assess the endurance performance in the digital self-test. The regression equation for women should be further investigated with a larger number of subjects and should therefore be used with caution. Military Impact The Swiss Armed Forces offer their recruits and soldiers a digital solution for physical training before and between military service. The digital performance test implemented in this application, which can be carried out independently, regardless of time and place, and without any special equipment, enables the generation of a personalised physical training. The 4Minmax run is a promising solution to assess the endurance performance and is an alternative to conventional laboratory and field tests.
In military service, marching is an important, common, and physically demanding task. Minimizing dropouts, maintaining operational readiness during the march, and achieving a fast recovery are desirable because the soldiers have to be ready for duty, sometimes shortly after an exhausting task. The present field study investigated the influence of the soldiers' cardiorespiratory fitness on physiological responses during a long-lasting and challenging 34 km march.; Heart rate (HR), body core temperature (BCT), total energy expenditure (TEE), energy intake, motivation, and pain sensation were investigated in 44 soldiers (20.3 ± 1.3 years, 178.5 ± 7.0 cm, 74.8 ± 9.8 kg, body mass index: 23.4 ± 2.7 kg × m-2, peak oxygen uptake (VO2peak): 54.2 ± 7.9 mL × kg-1 × min-1) during almost 8 hours of marching. All soldiers were equipped with a portable electrocardiogram to record HR and an accelerometer on the hip, all swallowed a telemetry pill to record BCT, and all filled out a pre- and post-march questionnaire. The influence of aerobic capacity on the physiological responses during the march was examined by dividing the soldiers into three fitness groups according to their $\dot{\rm{V}}$O2peak.; The group with the lowest aerobic capacity (VO2peak: 44.9 ± 4.8 mL × kg-1 × min-1) compared to the group with the highest aerobic capacity (VO2peak: 61.7 ± 2.2 mL × kg-1 × min-1) showed a significantly higher (p
