Wrist-worn fitness monitors may not be as accurate as once thought, a Brisbane study has found.
Subscribe now for unlimited access.
$0/
(min cost $0)
or signup to continue reading
University of Queensland PhD student Matthew Wallen led a study that put fitness monitors on 22 volunteers while they ran, cycled, walked and rested.
The study lasted for about one hour.
The Apple Watch, Fitbit Charge HR, Samsung Gear S and Mio Alpha were used in the study and were calibrated to each volunteer to reflect their age, gender, height and weight.
Electrocardiography readings that measure heart rate were taken every 15 seconds and a portable gas analysis system was used to measure the number of calories burnt.
Mr Wallen said the study found wrist-worn fitness bands used in the study undervalued how many calories each volunteer burned during the session.
"The heart rate application was actually quite good, it had a variation of one to nine per cent, so it was pretty good," he said.
"However with the energy expenditure, we found quite a large variation underestimating calories burnt."
The percentage error for energy expenditure was between nine and 43 per cent, Mr Wallen said.
"If someone is wearing a fitness band, if the device says they have burnt 100 calories, based on our data they could have burnt 150 calories depending on the variation," he said.
"If we are trying to work out how many calories we need to actually burn in a session, potentially using these devices may underestimate values and provide false information on what we are doing."
Mr Wallen said it was not known why the wrist-worn fitness monitors used in the study generated inaccurate data for calories burnt, but said the amount of movement involved may have been a factor.
"What we are thinking is that the movement may disrupt the monitor underneath the device ... this is a study we are currently analysing.
"In higher-intensity exercise, these devices may further underestimate the data," he said.
Findings from the study were published in the journal PLOS One.
