The real-time relationship between the built environment and mental health among older adults in Hong Kong: A geographic ecological momentary assessment (GEMA) study based on wearable technologies

Numerous studies have revealed that the built environment can affect older adults' mental health. However, such studies have largely been based on static designs, such as questionnaires, which can be prone to recall bias. Recall bias can be minimized to a great extent by using objective and real-time measures providing more dynamic information about how the built environment affects older adults' mental health. This proposed study aims to explore this dynamic interaction. Objective and real-time indicators for the built environment and mental health among older adults will be developed based on geographic ecological momentary assessment (GEMA) and wearable technologies. These technologies will provide vivid information about the direct and indirect effects (i.e., based on a behavior-based model and a perception-based model) of the built environment on older adults' mental health. Participants will be recruited from three areas in Hong Kong, each area having at least 30 participants to ensure their environmental exposure includes all types of the built environment attributes in Hong Kong. They will be required to complete one 15-minute baseline survey, which will include basic person-level information. They will then be required to wear the watch and GPS device for 15 days. A research assistant will send participants real-time short questions four times daily to gather information about their real-time feeling and activities by using Qualtrics software. The real-time data will be geocoded and linked to geo environment dataset and real-time data set by address and time. The intra-class correlation coefficients (ICC) will be estimated to determine the total variance of mental health changes explained at the within-person and between-person levels. Latent growth curve modeling (LGCM) will be integrated with structural equation modeling (SEM) to provide a quantitative exploration of the longitudinal direct and indirect effect of the built environment on mental health. Methodoloically, this study will provide a new approach which can also be applied to other research topics. The wearable technologies will be helpful in reducing the re-call bias used in traditional approaches, i.e., questionnaire survey. From the policy perspective, the study findings have potential implications for urban planning and resident's welfare policy. Practically, the study findings will play a significant role in policy recommendations to create an environment that can improve older people's mental health.