Scientific Abstract
Introduction: Often in the view of environmental and public health, despite inconclusive body of literature, there is an implicit bias towards animal agriculture where it is villainized yet often neglected. Understanding the status quo of agriculture, environment, and public heath aspects collectively is needed for betterment of translational research that can support sustainable, evidence-based, actionable solutions, and finding middle grounds. Spatial analysis would facilitate understanding the heterogeneity of exposure landscape and would support areas to focus on collecting finer scale data related to CAFO land use, propose targeted changes and determine allowable water, air, soil quality thresholds per CAFO location in vulnerable geographical areas.
Objectives: To quantify the status quo how the human or environmental exposure to hog CAFOs may be measured, accounting for site-specific social, geographical, and environmental factors.
Materials & Methods: Three spatial indexes were created at the census tract level, in North Carolina, to measure the exposure to hog CAFOs: 1) number of households within 1-mile from a large hog farm, 2) hog density per square kilometer, and 3) Geostatistical interpolation using co-Kriging that account for both the number of animals and the number of manure lagoons (Supplementary figure 1). The latter determined the risk as a continuous risk that can affect isotropically in the surrounding areas of the farm and the lagoons. These three indices were then compared against the CDC’s Environmental Justice Index (EJI) and Social Vulnerability Index (SVI), the agricultural land area, number of public water sources, soil runoff characteristics, landfill, and air quality at the census tract; by calculating the correlation coefficient. Additionally a simple regression analysis was conducted to quantify the association between three indices and three selected health conditions (all-cause mortality, infant mortality, and diabetes; which was used as a control i.e. a health condition unlikely to be associated with exposure to hog farming), while adjusting for potential confounders recognized through the correlation analysis. In the absence of yearly data on core variables, this ecological study was done purely spatially.
Results: In the preliminary analysis, moderate correlation (0.11 – 0.31) was observed between all 3 indices and the CDC’s EJI and SVI suggesting the exposure to CAFOs is likely related to areas affected by social vulnerability or were determined to have environmental justice issues (Supplementary figure 2). The SVI was correlated with 1-mile buffer index at correlation coefficient 0.31, indicating socially vulnerable neighborhoods are on average within 1-mile buffer from CAFOs. Known public water sources, number of landfills, and crop-agricultural land areas were all moderately correlated with the indices. All three-health conditions indicated strong associations to social vulnerability compared to the three CAFO exposure indices; suggesting it’s applicability as a confounding factor when determining the association between CAFO exposure and the disease outcomes in the regression analysis. However, even after accounting for the social vulnerability, the association between CAFO exposure was statistically significant including for diabetes. However, when accounted for both social vulnerability and crop-agricultural lands the strength of associations seem to diminish (Supplementary Table 1).
Discussion: The indices provide finer scale exposure measure for hog farming areas compared to previous estimates that were at zip code level, which areoften subjected to change. Limitations of the analysis include variance of correlations, assumption that residents lived in the census tract all throughout their life and the exposures were due to the residential environment, and inability to tease apart the nature of exposure through an ecological study. Overall, regardless of how the exposure to hog CAFOs was measured, all suggested moderate association with CDC’s environmental justice, social vulnerability indices, and selected health conditions; indicating that these neighborhoods have collective environmental, socioeconomic, and health issues. However, associations does not imply causation and CAFOs are located in census tracts that have large crop-agricultural lands and landfill areas making it challenging to prove or disprove association to exposures that could contribute to potentially affect soil, water, or air in these census tracts; which is much more complex than the mere presence of hog CAFOs.
