Can’t Lose Weight Even with Exercise? UMass Amherst Researchers to Figure Out Why

(PhysOrg.com) -- Exercise researchers have long wondered why two people following the same exercise program can get different results, one losing weight and the other not, for example. This is one of the questions to be explored by Patty Freedson, chair of the department of kinesiology and leader of a research team at the University of Massachusetts Amherst, who recently received a two-year, $985,000 grant from the National Institutes of Health.

The NIH grant is “a real feather in the cap of our research group, the School of Public Health and Health Sciences and of the university,” says Freedson, because the proposal scored in the top two percent nationwide, one of only 200 requests funded from among 21,000 applications. She declines to take all the credit for this success, noting that “you don’t receive a grant like this without every team member’s hard work and full support.” Besides Freedson, the research group includes two other faculty members, statistician John Staudenmayer and exercise physiologist Barry Braun, plus doctoral students Sarah Kosey and Kate Lyden, and proposal editor Meg Bouvier.

The study begins just one year after the government issued the first-ever federally mandated Physical Activity Guidelines for Americans (PAGA) in October 2008, Freedson notes. “Members of the PAGA science committee point out that our limited knowledge of the dose-response relationship between physical activity and health is in part related to the poor measures researchers have used to assess physical activity dose,” she says. “Our study will address this.”

The first part of the investigation will focus on fine-tuning the interpretation of accelerometer patterns to improve estimates of what the wearer is actually doing, that is, the exercise dose. “If you’re gardening, we want the monitor to tell us the difference between that and walking. Each activity has a signature so the device can recognize what you just did and record it. We’re further developing the software that reads data from the device so we can provide a more accurate estimate of physical activity dose,” Freedson explains.

As statistician Staudenmayer explains, accelerometers produce rich measures of movement, but they are not interpretable in units familiar to kinesiologists. “We address this problem by using flexible regression models such as neural networks and other methods to translate the accelerometer information into METs (metabolic equivalent of task), a typical measure of physical activity energy expenditure.”

This was previously done using linear regression that assumes a simple, straight-line relationship between the average accelerometer measurement and energy expenditure, an approach pioneered by Freedson’s group. “It was a good first step and is widely used in applied public health research, but it can’t take advantage of the full richness of information in the accelerometer signal. The newer models can, and they do a better job of estimating energy expenditure from accelerometer measurements,” he adds.

The goal of the first study, led by doctoral student Kate Lyden, is to determine the sensitivity of neural networks and other methods of detecting differences in physical activity dose. Specifically, 15 subjects will wear an accelerometer and a pedometer for three weeks. Each will have a different activity level assigned, thus one sedentary week and two active weeks. For the sedentary week, subjects will be coached on how to take fewer than 5,000 steps per day. In the other two weeks, subjects will take part in a moderate amount of activity of 8,000 to 10,000 steps per day, then a high physical activity week requiring 12,000 to 14,000 steps per day.

The UMass Amherst researchers will use the new, more sophisticated accelerometer data processing techniques in the second study to explore how daily activity outside of exercise training affects responsiveness to a training program. Many kinesiologists say individual differences are due to genetics, but Sarah Kosey, a graduate student working on her PhD in Freedson’s lab, believes the answer could be that people naturally have different daily physical activity patterns that account for their different results. “With the improved accelerometer data processing techniques we can address this hypothesis,” she points out.

In this second study, a total of 45 subjects in three groups of 15 each will be assigned to one of three daily routine activity “doses” in addition to their exercise workout for 12 weeks. The first group will carry on light daily activity with their regular exercise training while a second group will carry out higher daily activity and a third, control group will lead a sedentary lifestyle.

The researchers will compare groups to see if individuals who spent time in a more active daily life routine will show greater improvements in cardiorespiratory fitness, insulin action, blood pressure and cholesterol level than the sedentary group. Among the other ideas to be explored is that sedentary behavior outside of exercise training may negate the benefits of an exercise program.

Kosey says the results “have the potential to affect how clinical exercise trials are conducted. We may show that physical activity outside of an exercise training program is as important as the program itself.”

Provided by UMass Amherst