_{2max}in Well-Trained Rowing Athletes

^{1}

^{2}

^{3}

^{1}

^{1}

^{2}

^{*}

^{1}

^{2}

^{3}

Edited by: Stefanos Volianitis, Qatar University, Qatar

Reviewed by: William Sheel, University of British Columbia, Canada; Steffen Held, German Sport University Cologne, Germany

This article was submitted to Exercise Physiology, a section of the journal Frontiers in Physiology

This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

This study was designed to investigate the validity of maximal oxygen consumption (VO_{2max}) estimation through the Firstbeat fitness test (FFT) method when using submaximal rowing and running programs for well-trained athletes.

Well-trained flatwater rowers (^{–1}⋅min^{–1}) and paddlers (^{–1}⋅min^{–1}) completed the FFT and maximal graded exercise test (GXT) programs of rowing and running, respectively. The estimated VO_{2max} was calculated using the FFT system, and the measured VO_{2max} was obtained from the GXT programs. Differences between the estimated and measured VO_{2max} values were analyzed to assess the accuracy and agreement of the predictions. Equations from the previous study were also used to predict the VO_{2max} in the submaximal programs to compare the accuracy of prediction with the FFT method.

The FFT method was in good agreement with the measured VO_{2max} in both groups based on the intraclass correlation coefficients (>0.8). Additionally, the FFT method had considerable accuracy in VO_{2max} estimation as the mean absolute percentage error (≤5.0%) and mean absolute error (<3.0 mL⋅kg^{–1}⋅min^{–1}) were fairly low. Furthermore, the FFT method seemed more accurate in the estimation of VO_{2max} than previously reported equations, especially in the rowing test program.

This study revealed that the FFT method provides a considerably accurate estimation of VO_{2max} in well-trained athletes.

Maximal oxygen consumption (VO_{2max}) is defined as the maximal capacity of the pulmonary, cardiovascular, and muscular systems to deliver and utilize oxygen, which can reflect an individual’s cardiorespiratory fitness (_{2max} provides important outcomes for both physical performance and health status in general (_{2max} is often used in endurance sports to provide training and athletes’ performance information to coaches (

VO_{2max} can be measured through direct methods with a metabolic gas measurement system, with the athlete performing a maximal graded exercise test (GXT) until exhaustion. This is regarded as the gold standard as it can obtain an accurate value of VO_{2max} (_{2max} based on the submaximal exercise program seem to be a good choice for athletes or teams, and these can frequently be used during the training season.

Previous studies have reported several indirect methods for estimating VO_{2max} in athletes based on running programs (_{2max} well, and the equation was fairly accurate [standard error of estimate (SEE) = 3.98-4.08 mL⋅kg^{–1}⋅min^{–1}; _{2max} predictive equations [(SEE) = 2.52–3.51 mL⋅kg^{–1}⋅min^{–1};

^{2} values of 0.79 and 0.64 in male and female rowers, respectively, _{2max} estimation for both male and female rowing athletes.

Recently, a new system [Firstbeat fitness test (FFT)] was used for the indirect estimation of VO_{2max}. The estimated VO_{2max} value is automatically generated after collecting heart rate (HR) data from a configurable test program (rowing or running) using Firstbeat sports software (Firstbeat, Jyvaskyla, Finland). Studies have revealed that the Firstbeat VO_{2max} estimation system is valid in nonathletic populations (_{2max} in well-trained athletes. Therefore, this study was designed to investigate the accuracy of VO_{2max} estimation based on the FFT system when using submaximal rowing and running programs. In addition, this study aimed to evaluate the cross-validation of previous VO_{2max} predictive equations in both submaximal running (_{2max} estimated by the FFT system to coaches and sports scientists. The hypothesis for this study was that the FFT method could provide accurate VO_{2max} estimation in well-trained athletes, which would varify a new accurate option for estimating VO_{2max} in athletic population.

The FFT system was used to estimate the participants’ VO_{2max} values, which were also compared to the VO_{2max} values from the direct method measurement of the GXT programs. Additionally, a cross-validation design was used to evaluate the validation of the VO_{2max} estimation when compared to other classical predictive equations based on submaximal rowing (

A total of 90 well-trained athletes were recruited from Zhejiang Water Sports Training Center and divided into two groups based on the sports items (i.e., rowers and paddlers): 45 flatwater rowers (23 males and 22 females) in the ROW group and 45 flatwater paddlers (29 males and 16 females) in the RUN group (participant characteristics are shown in

Descriptive characteristics of the participants.

Age (years) | 20.7 ± 3.6 | 19.0 ± 2.1 | 19.8 ± 3.0 | 18.7 ± 2.3 | 19.5 ± 3.0 | 19.0 ± 2.5 |

Height (cm) | 190.7 ± 5.4 | 176.6 ± 4.7 | 184 ± 8.7 | 184.0 ± 4.6 | 171.4 ± 5.0 | 180 ± 7.7 |

Body mass (kg) | 84.5 ± 3.6 | 67.0 ± 8.9 | 76 ± 12.9 | 78.7 ± 7.0 | 65.7 ± 7.0 | 74 ± 9.4 |

Training experience (years) | 4.7 ± 3.1 | 3.9 ± 2.1 | 4.3 ± 2.6 | 3.7 ± 1.9 | 4.6 ± 2.2 | 4.0 ± 2.1 |

_{2max} value was obtained from FFT program. After 3–5 days of recovery from the submaximal program, the participants performed a maximal graded rowing (ROW group athletes) or running program (RUN group athletes) to obtain the measuredVO_{2max} value using a breath-by-breath metabolic measurement system, which was regarded as the golden standard test of VO_{2max} (

Experimental tests diagram.

The ROW group athletes performed a submaximal incremental rowing test on the rowing ergometer (model E, Concept 2, Morrisville, VT, United States), while the RUN group athletes performed a submaximal running test on the treadmill (H/P/Cosmos, Nussdorf, Germany). According to previous studies (_{max}) was suggested to provide reliable predicted VO_{2max}. To determine the appropriate intensity for the submaximal programs, a pilot study of 10 athletes (five rowers and five paddlers) was conducted to obtain the physiological responses for different stages, especially the last stage on the rowing ergometer or running treadmill. The submaximal incremental exercise test program consisting of four 3-min rowing exercises (an initial workload of 160 W in male athletes and 120 W in female athletes) and treadmill running (an initial running speed of 9 km/h in male athletes and 8 km/h in female athletes) was designed based on the pilot study and then performed by the ROW and RUN group athletes, respectively. Detailed information on the two programs is shown in _{2max} values based on the collected HR data of the submaximal FFT program.

The ROW group athletes performed a maximal incremental rowing test program (_{2max} was identified when meeting at least two of the three following criteria (_{max} upon using the equation, 220 - age. The value of the measured VO_{2max} was defined as the highest 30-s average value of VO_{2} measured during GXT (

All data were presented as mean ± standard deviation. The Shapiro–Wilk test was performed to test the normality of the outcome variables. Then Pearson’s correlation between the estimated VO_{2max} values from the submaximal test programs and the measured VO_{2max} values from the direct method using the GXT program was performed to assess the correlation magnitude and coefficient of determination (^{2}). To assess the accuracy of the estimation, the mean absolute percentage error (MAPE) and mean absolute error were calculated. The intraclass correlation coefficient (ICC) was used to determine the agreement between the estimated VO_{2max} values and measured VO_{2max} values. Furthermore, the Bland-Altman plot was used to investigate the level of agreement with the 95% limits of agreement (

_{2max} values measured by GXT programs and the HR, RPE, and RER in the last stage of the GXT program, and _{2max} from the FFT and the analysis of correlations and differences between the two VO_{2max} values. The results showed that the estimated VO_{2max} was significantly overestimated in both ROW [constant error (CE) = 1.3 ± 3.5 mL⋅kg^{–1}⋅min^{–1}, ^{–1}⋅min^{–1}, _{2max} from the FFT had a good level of agreement with the directly measured VO_{2max} from the GXT in both the ROW (0.818, _{2max} estimation. Furthermore, linear regression plots demonstrated a good predictive model of the FFT method in both the rowing (^{2} = 0.695, ^{–1}⋅min^{–1}; ^{2} = 0.753, ^{–1}⋅min^{–1};

Results of the maximal graded exercise test.

HR at VO_{2max} (bpm) |
189.0 ± 4.5 | 190.8 ± 5.3 | 189.9 ± 4.9 | 198.0 ± 5.2 | 193.3 ± 8.5 | 196.3 ± 6.9 |

RPE at VO_{2max} (6–20) |
18.2 ± 1.1 | 18.2 ± 0.9 | 18.2 ± 1.0 | 18.9 ± 0.8 | 18.8 ± 1.1 | 18.9 ± 0.9 |

RER at VO_{2max} |
1.13 ± 0.1 | 1.11 ± 0.1 | 1.12 ± 0.1 | 1.13 ± 0.1 | 1.08 ± 0.1 | 1.11 ± 0.1 |

VO_{2max} (mL⋅kg^{–1}⋅min^{–1}) |
60.7 ± 5.9 | 56.7 ± 5.5 | 58.7 ± 6.0 | 62.3 ± 3.4 | 55.6 ± 4.0 | 59.9 ± 4.8 |

_{2max}, maximal oxygen consumption.

The correlations and differences between the estimated VO_{2max} from FFT and the measured VO_{2max} from GXT.

_{2max} |
_{2max} |
||||||||

^{–1}⋅min^{–1}) |
^{–1}⋅min^{–1}) |
^{–1}⋅min^{–1}) |
^{–1}⋅min^{–1}) |
||||||

Male ( |
62.5 ± 6.0 | 60.7 ± 5.9 | 1.8 ± 3.6 | 2.429* | 0.798* | 0.736* | 3.0 ± 2.6 | 5.3 | |

Female ( |
57.5 ± 6.3 | 56.7 ± 5.5 | 0.7 ± 3.3 | 1.047 | 0.851* | 0.841* | 2.7 ± 2.0 | 4.8 | |

All ( |
60.0 ± 6.0 | 58.7 ± 6.0 | 1.3 ± 3.5 | 2.501* | 0.834* | 0.818* | 2.9 ± 2.3 | 5.0 | |

Male ( |
63.2 ± 4.8 | 62.3 ± 3.4 | 0.9 ± 2.6 | 1.838 | 0.851* | 0.787* | 2.2 ± 1.6 | 3.5 | |

Female ( |
58.0 ± 4.4 | 55.6 ± 4.0 | 2.4 ± 2.4 | 3.902* | 0.837* | 0.727* | 2.7 ± 2.0 | 5.0 | |

All ( |
61.3 ± 5.3 | 59.9 ± 4.8 | 1.4 ± 2.6 | 3.624* | 0.868* | 0.834* | 2.4 ± 1.7 | 4.1 |

_{2max};

The linear regression plots between the estimated VO_{2max} from FFT and measured VO_{2max} by GXT. ^{2}) and 95% confidence interval bounds (dotted line) are also depicted.

The Bland-Altman plots also demonstrated an agreement between the estimated VO_{2max} from the FFT and the directly measured VO_{2max} from the GXT program (_{2max} had fairly low mean differences (bias) in both the ROW (^{–1}⋅min^{–1} for rowing and 1.42 mL⋅kg^{–1}⋅min^{–1} for running). Furthermore, the FFT rowing program had a larger range of bias than that in the running program when estimating VO_{2max} from the FFT [upper to lower limits of agreement (ULoA-LLoA): 13.56 mL⋅kg^{–1}⋅min^{–1} vs. 10.32 mL⋅kg^{–1}⋅min^{–1}, respectively].

Bland-Altman plots between the estimated VO_{2max} from FFT and measured VO_{2max} by GXT.

We then examined previous equations that predict VO_{2max} based on _{2max} are shown in _{2max} from the GXT program. It was found that although Eq. 1 had a fairly accurate prediction of VO_{2max} in the rowing program, the FFT method had better accuracy and lower error terms for the overall ROW group (CE = 2.9 ± 2.3 mL⋅kg^{–1}⋅min^{–1} and ICC = 0.818), as well as in both male (CE = 3.0 ± 2.6 mL⋅kg^{–1}⋅min^{–1} and ICC = 0.736) and female (CE = 2.7 ± 2.0 mL⋅kg^{–1}⋅min^{–1} and ICC = 0.841) subgroups. In addition, in the RUN group, Eqs 2, 3 had similar validity coefficients (_{2max} from the GXT program. Moreover, Eq. 2 had a nonsignificant difference [^{–1}⋅min^{–1}, MAPE = 4.8%) than did Eq. 3.

Descriptive examination of the correlations and differences between other indirect methods and the measured VO_{2max}.

_{2max} |
||||||||

^{–1}⋅min^{–1}) |
^{–1}⋅min^{–1}) |
^{–1}⋅min^{–1}) |
||||||

Equation 1 | ||||||||

Male ( |
65.7 ± 6.1 | 5.0 ± 5.6 | 4.310* | 0.568* | 0.427* | 6.0 ± 4.5 | 10.3 | |

Female ( |
59.5 ± 5.9 | 2.8 ± 3.5 | 3.701* | 0.807* | 0.723* | 3.9 ± 2.3 | 6.9 | |

All ( |
62.7 ± 6.7 | 3.9 ± 4.8 | 5.513* | 0.721* | 0.604* | 5.0 ± 3.8 | 8.6 | |

Equation 2 | ||||||||

Male ( |
62.1 ± 2.8 | -0.2 ± 2.8 | -0.333 | 0.602* | 0.600* | 2.2 ± 1.7 | 3.5 | |

Female ( |
57.8 ± 2.6 | 2.2 ± 4.5 | 1.954 | 0.125 | 0.099 | 3.6 ± 3.8 | 6.9 | |

All ( |
60.6 ± 3.4 | 0.7 ± 3.7 | 1.238 | 0.655* | 0.615* | 2.7 ± 2.5 | 4.8 | |

Equation 3 | ||||||||

Male ( |
64.0 ± 3.5 | 1.7 ± 2.3 | 4.116* | 0.787* | 0.704* | 2.3 ± 1.7 | 3.6 | |

Female ( |
59.2 ± 3.4 | 3.6 ± 5.2 | 2.728* | 0.008 | 0.006 | 4.8 ± 4.2 | 9.0 | |

All ( |
62.3 ± 4.2 | 2.4 ± 3.7 | 4.352* | 0.671* | 0.586* | 3.1 ± 3.0 | 5.6 |

_{2max};

_{2max}(mL⋅kg

^{–1}⋅min

^{–1}) in males = (3.2131 + 0.0076 × PWC170) / body mass, VO

_{2max}(mL⋅kg

^{–1}⋅min

^{–1}) in females = (2.4138 + 0. × PWC170) / body mass. Equation 2: VO

_{2}(mL⋅kg

^{–1}⋅min

^{–1}) = (speed (m⋅min

^{–1}) × 0.2) + (gradient × speed (m⋅min

^{–1}) × 0.9) + 3.5, where the estimated maximal speed was calculated as following steps, (1) linear regression was used based on steady-state heart rate values and running speed were obtained for each stage of the submaximal running test, (2) the linear line was then extrapolated to estimated maximal heart rate (220 – age) to determine the value of estimated maximal speed. Equation 3: VO

_{2max}(mL⋅kg

^{–1}⋅min

^{–1}) = VO

_{2}-stage 4 + b (HRmax – HR-stage 4), where the VO

_{2}-stage 4 was calculated based on the steady-state HR in stage 4 of the submaximal running test as the Eq. (2), HRmax refers to estimated maximal heart rate (220 – age), and the additional coefficient b is calculated from b = (VO2-stage 4 – VO2-stage 3) / (HR-stage 4 – HR-stage 3), which is the ratio of the difference between the estimated VO2 of last two stages of submaximal running test and corresponding change of steady-state heart rate.

This study aimed to investigate the accuracy of the FFT method for VO_{2max} estimation using submaximal programs, such as rowing or running, in well-trained athletes. Good levels of agreement between the estimated VO_{2max} from the FFT and the measured VO_{2max} from the GXT and < 10% MAPE were observed in the current study, which met the criteria suggested by _{2max} in the ROW and RUN groups, respectively, suggesting that the FFT method estimated VO_{2max} well. Furthermore, the results also illustrated that FFT methods were more accurate in predicting VO_{2max} than the previous predictive equations when using the same submaximal programs in well-trained athletes. These findings indicate that the FFT method can be a fairly accurate option for obtaining VO_{2max} in well-trained athletes. Such submaximal tests can be more widely applied in the sports setting, such as in an individualized training or intervention approach, and during a repeated baseline testing setting.

Previous studies have developed several equations for predicting VO_{2max} when using submaximal programs in rowing or running exercise (^{2} = 0.707, and %SEE = 4.6%] based on critical velocity and anaerobic rowing test trials. Other studies (_{2max} (^{–1}⋅min^{–1}; ^{–1}⋅min^{–1}) using only the single-stage submaximal treadmill jogging test in healthy adults. However, these equations were developed for the nonathletic population and should be used with caution for well-trained athletes. The findings of this study revealed that the FFT method had a fairly accurate VO_{2max} estimation for well-trained rowers and paddlers and that the FFT method has been proven to accurately estimate VO_{2max} in nonathletic populations (college students, healthy adults, and recreational runners; _{2max} in well-trained rowers using a specific submaximal rowing test program.

In the rowing program, although the results showed that the previous predictive equation (Eq. 1) by _{2max}, the FFT method showed even higher accuracy and lower error terms in all groups, as well as in both the male and female subgroups (_{2max} from both the FFT method and Eq. 1 showed a significant overestimation of VO_{2max} (FFT, 1.3 ± 3.5 mL⋅kg^{–1}⋅min^{–1}; Eq. 1, 3.9 ± 4.8 mL⋅kg^{–1}⋅min^{–1}). Well-trained athletes have a lower HR at the same intensity than the nonathletic population and have a relatively higher predicted VO_{2max} value based on the linear regression model, which may be the main reason. In addition, the estimation of VO_{2max} from the FFT method in the female subgroup had a nonsignificant difference and a lower MAPE than that in the male subgroup, indicating that the FFT method may provide a more accurate estimation of VO_{2max} in female athletes than that in male athletes when using the rowing program.

In the running program, _{2max} that were suitable for different populations. The present study used these two equations to estimate VO_{2max} in well-trained athletes in the submaximal treadmill running program and found that they all had acceptable accuracy for the overall group. The two equations had similar validity coefficients and agreement levels. However, both equations showed poor accuracy in the female subgroup, which was probably due to the fact that these equations were cross-validated in male athletes instead of in female athletes in the study by _{2max} than these two equations for the overall group. Additionally, the FFT method also had an accurate VO_{2max} estimation in the female subgroup, which was better than that from the two equations. The FFT method modifies equations in the software based on the relevant background variables (e.g., activity class, training zone, and HR variability) and then improves the accuracy of the VO_{2max} estimation, which could explain this phenomenon.

_{2max} because the workload is not high enough to promote adequate parasympathetic withdrawal and concomitant sympathetic activation. The tendency of underestimation was discovered in previous FFT-related studies (_{2max} in both the rowing and running programs. A pilot study to detect suitable workloads in submaximal programs may contribute to this phenomenon. Previous studies have indicated that individualized submaximal testing has been utilized in running (_{max} and maximal endurance performance. Other studies also concluded that an optimal submaximal test program includes a proper target intensity, and different workloads for different characteristics may yield a more accurate prediction (_{2max}.

Taken together, using the FFT method for VO_{2max} estimation has several practical advantages in the evaluation of aerobic capacity in well-trained athletes. First, only a wearable HR device is needed, and the HR data are recorded during the submaximal testing program available in the software; then, the estimated VO_{2max} value with acceptable accuracy would be automatically calculated. Additionally, the FFT method only requires submaximal tests, and multiple athletes can be tested simultaneously, making its use more feasible during the busy training schedule compared to the direct measurement method for VO_{2max} in the laboratory. Thus, the FFT method can be considered as a potentially convenient and cost-effective alternative to measure the maximal aerobic capacity of well-trained athletes, especially for rowing and running.

This study had some limitations. The first one is the lack of information regarding the underlying equation of VO_{2max} estimation for the reason that the exact equation can not be obtained from the company. Second, unlike rowers, the lack of sports event-specific testing (paddling ergometer) in paddlers may limit the applicability of the results. However, the running program was performed in this study for the reason that the achievement of VO_{2max} by treadmill running is consistent with paddling ergometer in well-trained paddlers (

The results of the present study indicate that the FFT method provides a considerably accurate estimation of VO_{2max} in well-trained rowers, kayakers, and canoeists, which can be considered as a potentially convenient and cost-effective alternative to measure the maximal aerobic capacity of well-trained athletes, especially for rowing and running.

The original contributions presented in the study are included in the article/

The studies involving human participants were reviewed and approved by Ethics Committee of Zhejiang Institute of Sports Science, Hangzhou, China. Written informed consent to participate in this study was provided by the participants’ legal guardian/next of kin.

WG, HF, and QC performed the material preparation, and data collection and analysis. WG and O-PN wrote the draft of the manuscript. WG, O-PN, and XC conducted the revision. XC supervise the whole program. All authors contributed to the conception and design of this study and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The Supplementary Material for this article can be found online at:

Submaximal FFT programs for both the ROW group and RUN group.

GXT grograms for both the ROW group and RUN group.