Contextual and environmental factors that influence health: A within-subjects field experiment protocol.

Background: Despite the growing research on environment-physical activity (PA) relationships, field experimental studies are limited. Such studies offer opportunities to focus on real-world environmental exposure and related PA and health outcomes, allowing researchers to better isolate the causal effect of exposures/interventions. Focusing on the street/pedestrian environment as a routine setting for people's daily activities, this research aims to develop and test a field experiment protocol that integrates instantaneous assessments of the environment, PA, and health outcomes. The protocol involves the use of state-of-the-art environmental monitoring and biosensing techniques and focuses on physically active road users (pedestrians and bicyclists) who are more directly exposed to their surrounding environment than others such as drivers. Methods/Design: An interdisciplinary research team first identified the target measurement domains for the health outcomes (e.g., stress, thermal comfort, PA) and the street-level environmental exposures (e.g., land use, greenery, infrastructure conditions, air quality, weather) guided by the previous literature which was primarily observational. Portable or wearable measurement instruments (e.g., GPS, accelerometer, biosensor, mini camera, smartphone app, weather station, air quality sensor) were identified, pilot tested, and selected for the identified measures. We ensured that these measures are readily linkable using the time stamp and include eye-level exposures as they impact the users' experiences more directly yet missing in most prior studies relying on secondary, aerial-level measures. A 50-min experimental route was then determined to include typical everyday environments in park and mixed-use settings and to engage participants in three common modes of transportation (walking, bicycling, and driving). Finally, a detailed staff protocol was developed, pilot-tested, and used in a 36-participant within-subject field experiment in College Station, TX. The experiment was successfully executed, showing its potential to support future field experiments that can provide more accurate real-time, real-environment, and multi-dimensional information. Discussion: Our study demonstrates the feasibility of capturing the multifold health benefits/harms related to walking and bicycling in varying urban environments by combining field experiments with environmental, behavioral, and physiological sensing. Our study protocol and reflections can be helpful for a broad spectrum of research addressing the complex and multi-level pathways between the environment, behavior, and health outcomes.