Using GPS Tracking to Investigate Outdoor Navigation Patterns in Patients With Alzheimer Disease: Cross-sectional Study

Abstract

Background: Spatial disorientation is one of the earliest and most distressing symptoms seen in patients with Alzheimer disease (AD) and can lead to them getting lost in the community. Although it is a prevalent problem worldwide and is associated with various negative consequences, very little is known about the extent to which outdoor navigation patterns of patients with AD explain why spatial disorientation occurs for them even in familiar surroundings. Objective: This study aims to understand the outdoor navigation patterns of patients with AD in different conditions (alone vs accompanied; disoriented vs not disoriented during the study) and investigate whether patients with AD experienced spatial disorientation when navigating through environments with a high outdoor landmark density and complex road network structure (road intersection density, intersection complexity, and orientation entropy). Methods: We investigated the outdoor navigation patterns of community-dwelling patients with AD (n=15) and age-matched healthy controls (n=18) over a 2-week period using GPS tracking and trajectory mining analytical techniques. Here, for the patients, the occurrence of any spatial disorientation behavior during this tracking period was recorded. We also used a spatial buffer methodology to capture the outdoor landmark density and features of the road network in the environments that the participants visited during the tracking period. Results: The patients with AD had outdoor navigation patterns similar to those of the controls when they were accompanied; however, when they were alone, they had significantly fewer outings per day (total outings: P < .001; day outings: P = .003; night outings: P < .001), lower time spent moving per outing (P = .001), lower total distance covered per outing (P = .009), lower walking distance per outing (P = .02), and lower mean distance from home per outing (P = .004). Our results did not identify any mobility risk factors for spatial disorientation. We also found that the environments visited by patients who experienced disorientation versus those who maintained their orientation during the tracking period did not significantly differ in outdoor landmark density (P = .60) or road network structure (road intersection density: P = .43; intersection complexity: P = .45; orientation entropy: P = .89). Conclusions: Our findings suggest that when alone, patients with AD restrict the spatial and temporal extent of their outdoor navigation in the community to successfully reduce their perceived risk of spatial disorientation. Implications of this work highlight the importance for future research to identify which of these individuals may be at an actual high risk for spatial disorientation as well as to explore the implementation of health care measures to help maintain a balance between patients' right to safety and autonomy when making outings alone in the community.