Sticking to an exercise routine can be a difficult task for many individuals. However, a research team at the University of Mississippi is utilizing machine learning to uncover the key factors that keep people committed to their workouts.
The team, comprised of doctoral students Seungbak Lee and Ju-Pil Choe, along with Professor Minsoo Kang in the Department of Health, Exercise Science, and Recreation Management, is aiming to predict whether individuals are meeting physical activity guidelines based on various factors such as body measurements, demographics, and lifestyle choices.
With a dataset of approximately 30,000 surveys at their disposal, the team turned to machine learning to efficiently analyze the vast amount of information. By leveraging this technology, they were able to identify patterns and make predictions based on the collected data.
The results of their study, recently published in Scientific Reports, are particularly relevant given the public health concern surrounding physical activity adherence and its impact on disease prevention and overall health.
According to the U.S. Department of Health and Human Services, adults should strive for at least 150 minutes of moderate exercise or 75 minutes of vigorous exercise each week to maintain a healthy lifestyle. However, research indicates that the average American only dedicates two hours per week to physical activity, falling short of the recommended four hours by the Centers for Disease Control and Prevention.
The research team utilized public data from the National Health and Nutrition Examination Survey spanning from 2009 to 2018. By incorporating variables such as gender, age, race, educational status, BMI, waist circumference, alcohol consumption, smoking, employment, sleep patterns, and sedentary behavior, they aimed to predict adherence to physical activity guidelines.
Interestingly, the study revealed that factors such as sitting time, gender, and education level consistently appeared in the top-performing prediction models for exercise habits. These findings could provide valuable insights into understanding who is more likely to remain physically active and socially engaged.
The researchers acknowledged some limitations of their study, including the use of subjectively measured physical activity data. Moving forward, they plan to explore additional factors such as dietary supplements, utilize more machine learning algorithms, and rely on objective data rather than self-reported information to enhance the reliability of their research.
By leveraging machine learning techniques, the research team hopes to assist trainers and fitness consultants in developing workout regimens that individuals can maintain over the long term. This innovative approach has the potential to revolutionize how exercise adherence is predicted and supported in the future.
