University of Mississippi researchers employ machine learning to understand what keeps people committed to exercise, aiming to improve public health strategies.
Maintaining a consistent exercise routine is a struggle for many, but a team of researchers at the University of Mississippi is leveraging machine learning to predict who will stick to their workouts. This innovative approach could revolutionize public health strategies by providing insights into physical activity behavior.
The research, published in the journal Scientific Reports, utilizes advanced data analytics to predict adherence to physical activity guidelines.
“Physical activity adherence to the guidelines is a public health concern because of its relationship to disease prevention and overall health patterns,” corresponding author Minsoo Kang, a professor of sport analytics in the Department of Health, Exercise Science and Recreation Management, said in a news release. “We wanted to use advanced data analytic techniques, like machine learning, to predict this behavior.”
The study, analyzed data from approximately 30,000 surveys sourced from the National Health and Nutrition Examination Survey (NHANES) spanning the years 2009 to 2018. Machine learning enabled the team to rapidly analyze this vast dataset and identify patterns that predict whether individuals are meeting recommended physical activity levels.
The Office of Disease Prevention and Health Promotion recommends adults get at least 150 minutes of moderate exercise or 75 minutes of vigorous exercise per week. However, research shows that the average American falls short, clocking in just two hours of physical activity weekly.
Lead author Ju-Pil Choe, a doctoral student in physical education, explained the team’s objective.
“We aimed to use machine learning to predict whether people follow physical activity guidelines based on questionnaire data, and find the best combination of variables for accurate predictions,” he said in the news release.
The team considered a variety of demographic and lifestyle factors, including gender, age, education level, marital status, income, BMI and waist circumference, as well as lifestyle habits like alcohol consumption, smoking, employment, sleep patterns and sedentary behavior.
The study revealed three primary predictors of exercise adherence: time spent sitting, gender and education level. Each model they developed highlighted these variables, despite differing in other aspects.
“I expected that factors like gender, BMI, race or age would be important for our prediction model, but I was surprised by how significant educational status was,” Choe added. “While factors like gender, BMI and age are more innate to the body, educational status is an external factor.”
The researchers excluded participants with certain diseases and those with incomplete physical activity data, narrowing their analysis to 11,683 subjects. Machine learning allowed the team to uncover patterns that traditional methods might miss, offering greater flexibility and depth in their analysis.
“One limitation of our study was using subjectively measured physical activity data, where participants recalled their activity from memory,” added Choe. “People tend to overestimate their physical activity when using questionnaires, so more accurate, objective data would improve the study’s reliability.”
Future studies might incorporate objective data and additional variables such as dietary supplements, enhancing the robustness of the predictions. Such research could help fitness professionals develop tailored workout plans that individuals are more likely to maintain, ultimately improving public health outcomes.
Seungbak Lee, also a doctoral student in physical education, co-authored the study.
Source: The University of Mississippi