UMass researchers receive grant to study effects of exercise
The two-year, $985,000 grant was awarded to a team led by Professor Patty Freedson of the Kinesiology Department and John Staudenmeyer, an associate professor with the mathematics and statistics department. The group is developing new techniques to study why sometimes people on the same exercise regimen lose weight at different rates.
Having previously developed a wearable sensor called an accelerometer to measure activity levels, Freedson said that they are going to use this sensor to predict the amount of energy a person uses during an activity.
“When a person walks or runs, the accelerometer is going to measure the acceleration of the center of mass,” she said. “It’s going to look different depending on the type of activity.”
“The data collection is just starting,” Staudenmeyer said in an e-mail, “and all the data collection won’t be in for two years. We won’t have results or conclusions for a while.”
The information from the sensor is analyzed by a neural network – a kind of software that can learn.
“We train the neural network on a set of known data and apply it on a set of unknown data,” Freedson said. “Taking the accelerometer signal, we extract features to develop a pattern, and based on that, it predicts the amount of energy expenditure and identifies activity type.”
According to Freedson, the purpose of the grant is to see how well the neural network can differentiate between activity levels using data collected by two studies. In the first, 15 volunteers will wear both an accelerometer and a pedometer for three weeks. Each person will be assigned different levels of activity for each week, with one week of limited activity and two weeks of more strenuous activity.
“The biggest surprise is that it got funded,” Staudenmeyer said in an e-mail. “We thought it was a strong proposal, but we knew that NIH was only going to fund about 200 out of twenty-one thousand proposals.”
In the second study, 45 volunteers in three groups of 15 will also have different daily activity levels assigned. The groups will also maintain a weekly exercise regimen. One group will be assigned a moderate activity level, another a higher level, while the third will be the control group and be assigned a sedentary level. The study will last 12 weeks.
The conventional explanation for the different weight loss patterns is that they are caused by genetics, but the research group believes that there could be an alternate explanation.
“We predict that a person who sat around a lot will respond more poorly to training than a person who moves around a lot,” said Freedson.
The NIH grant money is part of the stimulus package provided by the American Recovery and Reinvestment Act
Along with Freedson and Staudenmayer, the group contains Associate Professor Barry Braun and two Kinesiology graduate students, Kate Lyden and Sarah Kozey.
“We’re also hiring some people to work on this project,” Freedson said. “We’re creating jobs.”
Matthew M. Robare can be reached at firstname.lastname@example.org