Inactivity increases with age and also in individuals with conditions, such as arthritis, that limit physical movement. Health professionals want to help this population exercise more, but to develop effective interventions, they need to know specifics about activities they already engage in.
Scott Strath, a professor of kinesiology, is embarking on a first-of-its-kind data collection project aimed at providing the what, where, how and with whom as it pertains to activity behavior in those with physical limitations.
Backed by a $2.8 million grant from the National Institutes of Health, he and his research team are outfitting nearly 400 Milwaukee-area adults with wearable movement- and photography-trackers. The subjects will wear heart-rate monitors, camera and accelerometers, which are similar to Fitbits, for extended periods as they go about their daily routines.
“We know that people are unreliable in reporting how active they are,” Strath said. “So the results of this study will potentially enable specialized interventions to increase physical activity in populations that need it the most.”
While the Milwaukee researchers collect data, partners at the University of Massachusetts Amherst will develop machine-learning algorithms to create models that predict physical-activity behavior in people’s everyday lives.
“The algorithms will accomplish something analogous to when you are texting on your phone and it tries to guess the word you are typing,” said Strath. “One of our goals is to predict physical activity behavior with similar techniques.”
Participants will range in age and level of physical functional ability, and each will be assessed in their own home or in their community for four days with fully integrated movement and location monitoring.
Other research partners include the University of Colorado’s Anschutz Medical Campus and Marquette University.