Using GPS data to explore demographic predictors of life-space: a pooled analysis

Elyse Kadokura | 2016

Advisor: Mandy Fretts

Research Area(s): Aging & Neurodegenerative Diseases, Epidemiologic Methods, Psychiatric Epidemiology, Social Determinants of Health


Background: Life-space is defined as an individual’s spatial movement through the environment as covered in daily living. A restricted life-space is associated with cognitive decline, Alzheimer’s disease, frailty and depression. We sought to use GPS data to explore demographic predictors of a restricted life-space. Methods: An exploratory, cross-sectional analysis was conducted using GPS data from seven studies (n=669) provided by the Research in Environment, Active Aging & Community Health (REACH) center at University of California, San Diego. GPS data was converted into minutes spent in each life-space (within the home, within the immediate neighborhood, outside the neighborhood). Demographic variables (age, education, marital status, race, income, BMI, and employment) were standardized across studies. Associations between time spent in each life-space and demographics were analyzed within each study and in a pooled sample. Bivariate analyses and multivariable analyses using GEE models were conducted. Results: Employment was the strongest and most consistent predictor of time spent within each life-space (p<0.001). Participants who were employed full-time spent the most amount of time outside the neighborhood. Older age (p=018), being presently married (p=0.002), and being unemployed or retired (p<0.001) were positively associated with more time spent within the home. BMI, income, and education were not associated with time spent in each life-space. Conclusion: The use of technologies to monitor and develop methods to measure life-space mobility is an important step in identifying vulnerable populations that should be targeted for intervention.