Estimation Development and Validity for Maximal Oxygen Consumption based on Submaximal Exercise Responses and Body Index Variables for Adult Women |
Yoo-Joung Jeon, Byung-Kun Lee, Jae-Hyeng Im |
KAIST Sangmyung University |
Correspondence:
Jae-Hyeng Im, Email: imjh@kaist.ac.kr |
Received: 27 February 2014 • Accepted: 19 July 2014 |
|
Abstract |
PURPOSE The purpose of this study is to develop estimation for maximal oxygen consumption based on submaximal exercise responses and body index variables for adult woman, and to verify prediction model based on verification group. METHODS The subjects are consisted of 370 woman aged from 20’s to 50’s, and we separated them into sample group(n=270) and cross-validation group(n=100). The estimation models were developed from multiple regression analysis to the prediction group, and verified it with cross-validation group. The prediction group’s input variables were body indexes and metabolic variables measured at 3 minute and 6 minute of Bruce treadmill protocol.
RESULTS Model 1 was V˙O 2 max = 36.630 + 0.0198*(HR170) - 0.076*(age) + 0.017*(6min V˙O 2 ) - 0.371*(weight) - 0.008*(6min V˙CO 2 ) + 0.005*(3min V˙O 2 ) - 0.046*(6min HR), model 2 was V˙O 2 max = 27.076 + 0.025*(HR 170) - 0.072*(age) + 0.017*(6min V˙ O 2 ) - 0.361*(weight) - 0.009*(6min V˙CO 2 ) + 0.005*(3min V˙O 2 ), model 3 was V˙O 2 max = 27.166 + 0.024*(HR 170) - 0.063*(age) + 0.019*(6min V˙O 2 ) - 0.323*(weight) - 0.009*(6min V˙CO 2 ), model 4 was V˙O 2 max = 25.711 + 0.031*(HR 170) - 0.101*(age) + 0.013*(6min V˙O 2 ) - 0.403*(weight). All 4 models had high R value(R=0.77~0.82, p<.01), low SEE,(±2.8~±3.1), and SEE%(8.5~9.5, p<.01). Also, they did not show multicolinearity. From Cross-validation, there were significant correlation between predicted and measured V˙O 2 max((R=0.66~0.76, p<.01). In addition, both %error(-9.71~11.17) and %TE(2.8~3.3) were very low, confirming its accuracy and validity. CONCLUSION V˙O 2 max estimation 4 models for adult woman were developed based on submaximal exercise responses and body index variables. As a result of verification, these model show valuable cross-validation.
The models from this study can be effectively utilized in exercise prescription field, and in any hospitals that can conduct graded exercise tests. |
Keywords:
estimation of V |
|