Fig 1.
Overview of the study framework.
All maps were created using ArcGIS Pro v3.3 by the authors.
Table 1.
Selected POIs categories related to CHD.
Table 2.
Demographic and Socioeconomic variables used.
Fig 2.
Neighborhood-level CHD prevalence in the US.
Map was produced using ArcGIS Pro v3.3 by the authors.
Fig 3.
Spatial distribution of the five health behavior measures (place visitation rate) at census tracts level in the US.
Visitation rate is the total number of visits to a type of POI per total number of mobile devices in a census tract. All maps were produced by using ArcGIS Pro v3.3 by the authors.
Table 3.
Evaluation metrics of the linear regression models.
Table 4.
Spatial error model regression results showing associations between demographic variables, socioeconomic variables, place visitation behavior, interaction terms, and coronary heart disease (CHD) prevalence at census tract level.