Quantifying the Relationship Among Socioeconomic Status and Prevalence of HIV Diagnoses Using 2018 Greater Atlanta Metropolitan Area Zip Codes

Introduction: HIV has been a public health concern for many decades. While cases have decreased slightly, public health professionals are still far from ending the HIV/AIDS epidemic. In order to properly address how to implement policy changes and promote healthy behaviors, dimensions of social determinants of health, specifically socioeconomic status, must be analyzed. By understanding the influence of social factors, conclusions can be made about health habits, decisions, and choices. While there is no cure for HIV, understanding social contributions can help decrease incidence rates. Aim: The purpose of this study is to quantify the relationship between socioeconomic status, a social determinant of health, and HIV diagnoses. This analysis aims to analyze which social variables are better predictors for prevalence of HIV diagnoses. This study can be duplicated to show social variables as predictors for any communicable or non-communicable disease. Methods: A 2018 dataset was downloaded from AIDSVu that included 133 zip codes in the greater Atlanta metropolitan area. The independent variables of interest were median household income, percent of the population living in poverty, percent of the population with a high school education, and percent of the population living with severe housing cost burden. The outcome of interest was HIV diagnoses per 100,000 in each zip code. County populations were obtained using the U.S. Census Bureau for the year 2020. Poisson regression was used to determine the association between each social variable and the total number of HIV cases per 100,000. Results: 128 zip codes were included in the analysis. The independent variables were dichotomized into high and low groups, with the cut off being equal to the Georgia state median. Both the univariate and multivariable regression models showed statistical significance between low median household incomes, high levels of people living in poverty, low levels of high school education, and higher percentages of the population living with severe housing cost burden. In the univariate model, percent of the population living in poverty held the highest HIV diagnoses prevalence ratio with an IRR of 3.39, followed by percent living in severe housing cost burden (IRR = 2.92), median household income (IRR = 2.66), and lastly high school education (IRR = 1.46). In the multivariable association, the prevalence ratios are attenuated due to confounding of variables. Discussion: The multivariable regression model shows a statistically significant relationship between individuals negatively impacted by social determinants of health when compared to individuals not negatively impacted. Social behavior within each population needs to be understood before effective health policies and interventions can be implemented. The built environment for zip codes with larger percentages of individuals living in poverty needs to be addressed before HIV preventative measures can be put in place. Educating individuals on HIV transmission is fundamental on slowing the spread, but population education levels must be considered first. Low-income areas (below the Georgia state average) are more likely to have individuals living below the poverty level, with severe housing cost burden, and less than a high school education. These areas should be targeted to HIV prevention before areas with higher rates