Abstract
BACKGROUND: Composite mobility patterns may better model built environment exposures, yet are rarely implemented to understand the prevalence of cardiometabolic disease (CMD). OBJECTIVES: The purpose of this study was to investigate the association between a novel Wireless Movement Index (WMI) and prevalent CMD redlined U.S. census tracts. METHODS: The Homeowners Loan Corporation (HOLC) was used to identify census tracts graded A-D and their age-adjusted prevalence (2019) for systolic hypertension, coronary heart disease, diabetes, obesity, chronic kidney disease, and stroke were collected. From nationally representative cell-phone tracking data, the WMI was constructed to identify population-level movement patterns and visits to points of interest. The association between WMI and disease prevalence was investigated using multivariable linear regression models across HOLC grades. RESULTS: Among 16,352 tracts, 4,458 were classified as HOLC grades A-B, 7,572 as grade C, and 4,322 as grade D. Grade D tract residents reported 55% of their visits to other grade D census tracts, with only 9% to grade A/B census tracts. The WMI was negatively associated with CMD prevalence across all HOLC grades, but this protective association was most pronounced in redlined areas. In grade D tracts, each unit increase in WMI was associated with a -2.33 (95% CI: -2.79 to -1.86), -2.93 (95% CI: -3.42 to -2.45), and -0.99 (95% CI: -1.30 to -0.68) decrease in prevalent hypertension, obesity, and diabetes. CONCLUSIONS: Even among redlined census tracts, those that reported a higher WMI (indicative of more frequent and diverse mobility) were likely to have better population-level cardiometabolic health. WMI may serve as a scalable, dynamic proxy for environmental opportunity and structural inequity.