John Sirard, Assistant Professor of Kinesiology, and John Staudenmayer, Professor of Math and Statistics, recently received a five-year, $2.2 million grant from the National Institutes of Health to develop methods for establishing more accurate assessments of physical activity and sedentary behavior in young people. Accurate and precise measures of these activities in youth are critical to researchers trying to provide evidence-based information related to health, Sirard explains.
He says a major problem now is that people don’t move in daily life as they do in a laboratory, where most accelerometer devices, or physical activity trackers, are tested. “In free-living settings, these devices don’t work as well and we know we’re misclassifying physical activity. We think our activity interventions are having an impact, but right now we might not be able to detect that with current methods of assessment. This work will allow us to advance the measurement techniques we use to make more accurate and precise estimates,” he says.
Sirard and Staudenmayer’s study, for which they hope to recruit 300 volunteers over five years, will measure physical activity levels in young people from 18 months to 20 years old using videotaped observations in natural settings including home, on the playground or sports field, in the community and in school. For this real-world calibration, participants will wear two research-grade accelerometers, one on the wrist and one on the hip, allowing the researchers to calibrate the accelerometer data to the video data using machine learning techniques. This will help to define such terms as “moderate” and “vigorous” activity, for example.
Sirard says others have done accelerometer calibration studies with young people but results are inconsistent. “Some label the output from the accelerometer device as moderate activity but the data are not consistent across studies,” he explains. “We really want to get some consistency so we can do a better job of defining different levels of physical activity. If we do that we can have faith in results across studies from different research groups.”
He adds, “Our research accelerometer does well at classifying activities performed in controlled laboratory settings. But, when we put them on children to measure physical activity in daily life, our accuracy decreases. The idea here is that performing the calibration outside the lab will produce better estimates of the intensity and type of activity in real-world settings.”
This work extends earlier research in adults by accelerometer research pioneer Patty Freedson at UMass Amherst, a co-investigator on the new project. Sirard is a physical activity epidemiologist in kinesiology in the School of Public Health and Health Sciences and has a joint appointment with the Commonwealth Honors College. Mathematics professor Staudenmayer is an expert in developing statistical techniques to identify and analyze physical activity and sedentary behavior patterns in accelerometer data. Their collaboration was fostered by the new Center for Personalized Health Monitoring at the campus’ Institute for Applied Life Sciences.