Reliability And Validity Study Of Smart Sports Watches In Measuring TRIMP During Running

Purpose:Wearable devices make load monitoring during exercise more convenient,and smart sports watches based on PPG technology are popular among runners.Training impulse is the main load indicator provided by sports watches.However,the calculation of TRIMP is related to heart rate zones.The heart rate fluctuation range during exercise is large,and the effectiveness and accuracy of smart sports watches’testing need to be explored for different heart rate zones.Methods:Seventeen male university runners were recruited to participate in one incremental load exhaustion test and three field running tests of different intensities.During the four tests,the subjects were required to wear four watches（2 of each brand,brand A and brand B）throughout the entire testing process.During the incremental load exhaustion test,the subjects were also required to wear a gas metabolism analyzer to measure the true HR_max and HR_max recorded by the four watches.The three field running tests included a low-intensity 8km self-paced running test(intensity:50-70%HR_max),a medium-intensity 30-minute self-paced running test(intensity:70-80%HR_max),and a high-intensity 5×800m self-paced interval running test(intensity:80-100%HR_max).In the experiment,the POLAR TEAM pro2 team system displayed the subjects’heart rate in real-time via a Bluetooth terminal to ensure that their heart rate remained within the required range.The reliability test used the intra-class correlation coefficient（ICC）and standard error of the mean（SEM）of the two identical sports watches to evaluate the reliability of the maximum heart rate and training impulse recorded by the watches.The validity test used paired sample T-tests to compare the differences between the true maximum heart rate and the predicted maximum heart rate using a formula,and the maximum heart rate recorded by the two watches.It also tested the differences between the training impulse calculated by the two watches and the training impulse calculated by the heart rate monitor.The effect size was used to indicate the size of the difference,and the correlation and consistency were evaluated through linear regression and Bland-Altman consistency analysis.Results:Validity results showed that in the maximum heart rate test,the formula for predicting maximum heart rate had a moderate effect size（ES=0.69）,but had very weak correlations（r=0.149）and consistency（r²=0.802,Lo A=Lo A=-12.7～19.7）.Brand A watches had a small effect size（ES<0.2）and extremely strong correlations（r=0.909）,but lower consistency（r²<0.001,Lo A=-6.83～7.18）.Brand B watches had a small effect size（ES<0.2）and moderate correlations（r=0.691）,but lower consistency（r²=0.022,Lo A=-6.83～7.18）.In the 50%-70%HR_max intensity range,brand A watches had a small effect size（ES<0.2）but extremely high correlations（r=0.989）and consistency（r²=0.134,Lo A=-9.56～6.93）,while brand B watches had a small effect size（ES<0.2）but extremely high correlations（r=0.977）and consistency（r²=0.50,Lo A=-10.70～13.58）.In the 70%-80%HR_max intensity range,brand A watches had a larger effect size（ES>0.8）but very low correlations（r=0.196）and consistency（r²=0.049,Lo A=-2.52～6.06）,while brand B watches had a small effect size（ES<0.2）but moderate correlations（r=0.419）and consistency（r²=0.53,Lo A=-4.67～8.12）.In the80%-100%HR_max intensity range,brand A watches had a smaller effect size（ES<0.2）but extremely high correlations（r=0.971）,but lower consistency（r²<0.001,Lo A=-7.54～15.98）,while brand B watches had a small effect size（ES<0.2）but extremely high correlations（r=0.965）but lower consistency（r²=0.97,Lo A=-25.94～21.15）.Conclusion:Two watches were tested for inter-device reliability of HR_max and were found to be moderate to high,but both watches had shortcomings in the validity of HR_max measurement.The formula"HR_max=220-age"underestimated the HR_max of the subjects.In actual running tests,both watches had high reliability and validity in measuring TRIMP in the low-intensity zone.In the medium to high-intensity zone,both watches had moderate to high reliability in measuring TRIMP,but both had shortcomings in the validity of TRIMP measurement.