Associated Professor of Global Public Health Dominican University of California Fairfax, California, United States
West Nile Virus (WNV) has emerged as a persistent source of vector-borne disease risk throughout California. Focal points of elevated WNV transmission risk are intrinsically spatial, resulting from a complex interplay of social and ecological factors. Within select urban environments, the geographic distribution of these environmental risk factors is driven, in part, by historical discriminatory housing policies (i.e. redlining) created by the Home Owner’s Loan Corporation (HOLC). In California, these cities include Fresno, Los Angeles, Oakland, Sacramento, San Diego, San Fransisco, San Jose, and Stockton. In this study, we sought to determine the association between these historically redlined communities and the risk of WNV emergence. We conducted a spatiotemporal analysis of WNV risk across eight historically redlined cities in California. We obtained georeferenced annual mosquito surveillance data from the California Department of Public Health vector sampling program between 2004-2021. WNV risk was quantified by calculating the mosquito infection rate (MIR), a metric that considers the number of WNV-positive pools and the number of mosquitos tested in a given location. We then aggregated the annual MIR at the census tract level within each city. We compared the annual WNV MIR between tracts historically graded on a four-point HOLC scale: A (‘most desirable’), B (‘still desirable’), C (‘definitely declining’), and D (hazardous, ‘redlined’). We also gathered epidemiologically meaningful socio-ecological data, including the proportion of poverty, impervious surface, and greenspace per census tract, from the California Health Places Index (HPI). Across the eight cities in our study area, we found clear evidence of geographic variation in neighborhood-level WNV risk. Census tracts that were previously redlined (‘D’ grade) had significantly higher cumulative WNV MIR (0.34) compared to non-redlined tracts (MIR = 0.21). The average MIR in census tracts given an ‘A’ grade was 0.25, 0.21 in ‘B’ grade tracts, and 0.18 in ‘C’ grade tracts. These differences extended to the HPI variables, with poverty rates found to be 3.4 times higher in redlined tracts compared to A-grade tracts. In conclusion, this study adds to the growing body of evidence linking historical redlining to health inequalities in urban environments. To our knowledge, this is the first study to quantify associations between discriminatory housing policies to emerging WNV risk.
