The conceptualization of a Just-In-Time Adaptive Intervention (JITAI) for the reduction of sedentary behavior in older adults.

Abstract

Low physical activity and high sedentary behavior in older adults can be addressed with interventions that are delivered through modern technology. Just-In-Time Adaptive Interventions (JITAIs) are an emerging technology-driven behavior-change intervention type and capitalize on data that is collected via mobile sensing technology (e.g., smartphones) to trigger appropriate support in real-life. In this paper we integrated behavior change and aging theory and research as well as knowledge around older adult's technology use to conceptualize a JITAI targeting the reduction of sedentary behavior in older adults. The JITAIs ultimate goal is to encourage older adults to take regular activity breaks from prolonged sitting. As a proximal outcome, we suggest the number of daily activity breaks from sitting. Support provided to interrupt sitting time can be based on tailoring variables: (I) the current accumulated sitting time; (II) the location of the individual; (III) the time of the day; (IV) the frequency of daily support prompts; and (V) the response to previous support prompts. Data on these variables can be collected using sensors that are commonly inbuilt into smartphones (e.g., accelerometer, GPS). Support prompts might be best delivered via traditional text messages as older adults are usually familiar and comfortable with this function. The content of the prompts should encourage breaks from prolonged sitting by highlighting immediate benefits of sitting time interruptions. Additionally, light physical activities that could be done during the breaks should also be presented (e.g., walking into the kitchen to prepare a cup of tea). Although the conceptualized JITAI can be developed and implemented to test its efficacy, more work is required to identify ways to collect, aggregate, organize and immediately use dense data on the proposed and other potentially important tailoring variables. Machine learning and other computational modelling techniques commonly used by computer scientists and engineers appear promising. With this, to develop powerful JITAIs and to actualize the full potential of modern sensing technologies transdisciplinary approaches are required.