PLoS One. 2023 Mar 14;18(3):e0280620. doi: 10.1371/journal.pone.0280620. eCollection 2023.
ABSTRACT
Food insecurity is prevalent and associated with poor health outcomes, but little is known about its geographical nature. The aim of this study is to utilize geospatial modeling of individual-level food insecurity screening data ascertained in health care settings to test for neighborhood hot and cold spots of food insecurity in a large metropolitan area, and then compare these hot spot neighborhoods to cold spot neighborhoods in terms of the CDC’s Social Vulnerability Index. In this cross-sectional secondary data analysis, we geocoded the home addresses of 6,749 unique participants screened for food insecurity at health care locations participating in CMS’s Accountable Health Communities (AHC) Model, as implemented in Houston, TX. Next, we created census-tract level incidence profiles of positive food insecurity screens per 1,000 people. We used Anselin’s Local Moran’s I statistic to test for statistically significant census tract-level hot/cold spots of food insecurity. Finally, we utilized a Mann-Whitney-U test to compare hot spot tracts to cold spot tracts in relation to the CDC’s Social Vulnerability Index. We found that hot spot tracts had higher overall social vulnerability index scores (P <0.001), higher subdomain scores, and higher percentages of individual variables like poverty (P <0.001), unemployment (P <0.001), limited English proficiency (P <0.001), and more. The combination of robust food insecurity screening data, geospatial modeling, and the CDC’s Social Vulnerability Index offers a solid method to understand neighborhood food insecurity.
PMID:36917592 | DOI:10.1371/journal.pone.0280620
