Using Wearable Sensors for Physical Activity Measurement and Promotion in Manual Wheelchair Users

Regular participation in physical activity (PA) is vital to good health. Despite numerous reported physical, physiological, and psychological benefits associated with regular PA for people with disabilities, many of them still lead a sedentary lifestyle, especially those who use manual wheelchairs as their primary means of mobility. Therefore, this research aims to provide tools to accurately measure and promote PA participation in manual wheelchair users (MWUs). In the first study, we established a PA compendium for MWUs with various diagnoses based on various daily and leisure PA performed by a total of 90 MWUs in the lab and home environments. The improved compendium allowed MWUs and/or clinicians to accurately determine the energy cost of various PA and develop a strategy to maintain energy balance. In the second study, we developed a set of custom algorithms for objectively predicting energy expenditure of MWUs and classifying time spent at different intensity levels during different PA using the commercial sensor, ActiGraph GT3X+, which was worn at the dominant upper arm and wrist. Algorithms were developed and validated using PA data from 45 MWUs in Pittsburgh, PA and Birmingham, AL. To achieve the overall goal of providing tools to measure and promote PA, the custom algorithms were made available for use through a smartphone app, WheelFit. In the third study, we conducted focus groups with MWUs and rehabilitation professionals who worked with MWUs to determine app content, features and functions that were most relevant, attractive, and which appropriately addressed the needs and desires of MWUs in a PA measurement and promotion app. As a result of the focus group inputs, we developed a functional prototype of WheelFit v1.0 including the following features: 1) goal setting; 2) real-time self-monitoring of daily PA; 3) progress tracking; 4) a workout library; 5) exercise reminder to trigger more PA; and 6) preliminary action planning to motivate users to perform target behavior. In the last study, we tested the usability of WheelFit v1.0 in 5 MWUs which showed that MWUs were generally satisfied with the app and were enthusiastic about using it to adopt and maintain an active lifestyle