Towards sedentary lifestyle prevention: An autoregressive model for predicting sedentary behaviors

Sedentary lifestyles are ubiquitous in modern societies. Watching television, playing video games, and using the computer while sitting are examples of sedentary behaviors that are now common worldwide. Multiple research studies have found that the length of sedentary time is linked with an increased risk of obesity, diabetes, cardiovascular disease, and all-cause mortality. Medical researchers have identified time, daily rhythm, and past sedentary habits as three determinants (causes) of sedentary behaviors. Building on these insights, in this paper we propose an approach to automatically discover patients' temporal patterns of sedentary behaviors from raw activity logs by using an autoregressive model with maximum entropy method. With this approach, health care providers will be able to predict patients' deleterious sedentary behaviors as evidence to drive computer-generated preventive interventions based on behavior change theories such as the Theory of Planned Behavior.