In the United States (US), significant progress has occurred in HIV prevention, treatment and care resulting in decreased incidence, morbidity, and mortality associated with HIV infection. The annual rate of HIV diagnoses per 100,000 population among US adults and adolescents decreased from 14.1 in 2017 to 12.7 in 2021.¹ However, some populations remain disproportionately affected. In 2021, the highest rates of HIV diagnoses were among Black or African American (hereafter referred to as Black) (41.8 per 100,000) and Hispanic/Latino (20.4 per 100,000) adults and adolescents, compared to White (5.3 per 100,000) adults and adolescents.¹ Reducing these disparities and achieving health equity are important for reaching public health goals. Federal initiatives, such as the National HIV/AIDS Strategy (NHAS) (2022‒2025)² and Ending the HIV Epidemic in the U.S. (EHE),³ place achieving health equity and reducing HIV-related racial and ethnic disparities at the core of ending the HIV epidemic.

Racial and ethnic disparities in HIV diagnoses are associated with an interplay of individual-level factors; systemic interpersonal practices; and structural-level social, economic, and political factors.^4–6 Although behavioral factors, for example, sexual behaviors that increase the chances of HIV acquisition and injection drug use, affect HIV transmission, underlying social and structural factors create conditions that lead to HIV-related disparities.^4,7,8 These factors are referred to as “social determinants of health (SDOH),” a host of overlapping structural factors and economic systems that shape the conditions in which people are born, live, work, and access resources.⁹ Racial and ethnic disparities in HIV diagnosis rates occur because of underlying root causes including poverty, experiences with homelessness, stigma, discrimination, residential segregation, systemic racism, and lack of or limited access to health care.^10,11 Race is a social construct and not a biologic driver of disparities in HIV diagnosis. Racial and ethnic disparities in health care are more likely because of differences in people’s experiences based on the relatively advantaged or disadvantaged position in society into which they are born. The racial and ethnic disparities examined in this analysis are markers of disparities. The unfair distribution of the SDOH factors in the analysis represents drivers of disparities.^12,13

Previous studies described the associations between HIV-related disparities and a wide range of SDOH such as income inequality, experiences with homelessness and unstable housing, stigma, discrimination, lack of healthcare insurance/access, poverty, and low education levels.^14–16 Some racial and ethnic groups in the US are disproportionately affected by these SDOH and these racial and ethnic groups also experience increased risk for HIV infection. For example, 19.5% of Black and 17.1% of Hispanic/Latino persons lived in poverty in 2021, compared to 8.1% of White persons.¹⁷ The literature is abundant with studies describing the associations between HIV rates, SDOH, and race and ethnicity.^7,14,18 However, there is paucity in research examining the magnitude to which variation in SDOH factors relate to differences in the racial and ethnic disparities in HIV diagnoses. This research gap could be addressed by using disparity measures to capture the relationship between SDOH factors and racial and ethnic disparities in HIV diagnoses.¹⁹

Previous studies have used multiple disparity measures to highlight the disproportionate distribution of HIV among racial and ethnic minority groups.^20–22 We used 8 relative and 4 absolute disparity measures to calculate racial and ethnic disparities in reported HIV diagnosis rates among adults in 2021 by selected SDOH stratified by sex assigned at birth (sex). We compared racial and ethnic disparity measures in census tracts in the most advantaged quartiles to racial and ethnic disparity measures in census tracts in the most disadvantaged quartiles. Findings from this study may inform assessments of the impact of addressing SDOH as a strategy to reduce HIV-related disparities and eliminate new HIV infections in the US.

Table 2 provides the numbers and rates of HIV diagnosis by sex, race and ethnicity, and SDOH for the most disadvantaged quartiles (Q1) and most advantaged quartiles (Q4). During 2021, a total of 32,907 adults were diagnosed with HIV the US, including 26,859 males (81.6%) and 6,048 females (18.4%). Nearly 95% of the HIV diagnoses in males were reported among Black (n=10,059), Hispanic/Latino (n=8,491), and White (6,763) males. Similarly, 95% of the HIV diagnoses in females were reported among Black (n=3,288), White (n=1,304), and Hispanic/Latino (n=1,136) females.

This analysis examined the relationship between racial and ethnic disparities in HIV diagnosis rates and selected SDOH factors—federal poverty level, education level, median household income, and health insurance coverage—within the most disadvantaged and most advantaged US census tract quartiles among adult males and females using four absolute and eight relative disparity measures. Overall, racial and ethnic disparities in HIV diagnosis rates were generally higher in absolute terms in the most disadvantaged census tracts but were generally higher in relative terms in the most advantaged census tracts.

When moving from the most disadvantaged to the most advantaged SDOH areas, the absolute disparity measures decreased while relative disparity measures generally increased. When HIV diagnosis rates decline for all racial and ethnic groups, decreases in absolute disparity coupled with increases in relative disparity can arise if the absolute decreases in HIV diagnoses are higher among Black and Hispanic/Latino persons compared with White persons, even though the proportional decreases are higher among White persons.²⁸ Reductions in relative disparity measures can occur when the proportional reduction in the HIV diagnosis rate in a group or groups with higher HIV diagnosis rates exceeds the proportional reduction in the HIV diagnosis rate in a group or groups with lower HIV diagnosis rates. For example, the Black-to-White rate ratio will decrease if the proportional reduction in the HIV diagnosis rate is greater for Black people than for White people. Policies that combat systemic racism and discrimination could achieve the magnitude of reduction in HIV diagnosis rates among Black and Hispanic/Latino persons that is required to reduce the relative disparities.

Regardless of the disparity measure used, racial and ethnic disparities in HIV diagnosis rates exist in both advantaged and disadvantaged census tract quartiles for all SDOH factors. These results are consistent with the robust literature on associations between poverty, lower education level, lower income, and insurance status and racial and ethnic disparities in HIV diagnosis rates. ^5,7,29 Our analysis contributes to the literature by showing that the disparities persist even in advantaged quartiles for the SDOH factors analyzed. Policies and practices that affect the inequitable distribution of resources and power among racial and ethnic populations and result in limited opportunities may partially explain the results. ^30–32 An important driver for observed disparities in the most advantaged census tracts could be the differential proportions of Black, White, and Hispanic/Latino residents in the census tracts. Historical policies that led to racial segregation resulted in lower proportions of racial and ethnic minoritized residents in the most advantaged census tracts, compared with White persons of similar income levels.^33,34 Residential segregation could also result in less access to quality HIV prevention, care and treatment services, and higher HIV diagnosis rates among racial and ethnic minoritized groups.^35,36

The finding regarding lack of health insurance coverage being associated with higher HIV diagnosis rates is important because insurance coverage may affect the racial and ethnic disparities in HIV diagnosis rates. The literature supports an association between the uptake of HIV prevention and treatment options and health insurance coverage.³² Results from a study that examined insurance coverage and viral suppression among persons with HIV found the percentage of people with sustained viral suppression was significantly lower among persons without insurance (54%), compared to those with private insurance (69%), Medicaid (59%), and Medicare (67%).³⁷ Additionally, Black and Hispanic/Latino persons have lower PrEP uptake compared to White persons, which could be due to lack of or inadequate health insurance coverage.³⁸ A study that compared PrEP use and health coverage between states with Medicaid expansion and states without Medicaid expansion found that men who have sex with men (MSM) in states with Medicaid expansion were more likely to have health coverage, use PrEP, and discuss PrEP with a provider.³⁹ Importantly, among MSM in states without Medicaid expansion, 35% were Black and 35% Hispanic/Latino, compared to 24% White.³⁹ Increasing access to health insurance may help in reducing HIV diagnosis rates and racial and ethnic disparities in these states.

Lower HIV diagnosis rates in the most advantaged census tract quartiles for the SDOH factors compared to the most disadvantaged quartiles highlight the association of SDOH factors with HIV diagnosis rates in the US. The fact that the relative disparities were higher and absolute disparities were lower in the most advantaged census tract quartiles compared to the most disadvantaged census tracts for the SDOH factors draw attention to the complex nature of racial and ethnic disparities. Possible explanations for these results include homophily in the selection of sex partners and the tendency of a person to seek a sex partner of the same race, both of which can increase HIV prevalence and viral load in sexual and social networks across geographic areas, possibly exacerbating racial and ethnic disparities in HIV incidence.^40,41 Further, the area-based SDOH factors examined in this analysis may not completely explain racial and ethnic disparities in HIV diagnosis rates. This analysis does not account for SDOH measured at individual-level, which might explain some of the observed disparities. Our analysis is descriptive in that we quantified racial and ethnic disparities by SDOH quartiles, without adjusting for census tract characteristics that might affect disparities. Controlling for variables that may be in causal pathways could obscure racial and ethnic disparities, particularly if there are racial and ethnic disparities in these variables.⁴² Findings support the need to address the root causes (e.g., systemic racism, stigma, and discrimination) of disparities in HIV diagnosis rates to eliminate these disparities.^30–32 These structural factors affect the inequitable distribution of SDOH factors among racial and ethnic minority groups across all quartiles of poverty, education, income, and access to health insurance. An equitable approach in the implementation of access to prevention, care, and treatment services focused on communities that experience disparities could help to address the racial and ethnic disparities in HIV. ^14,43 Findings from modeling studies suggest that substantial and equitable improvements in HIV prevention and treatment for the groups experiencing inequites are needed to reduce racial and ethinic disparities in HIV-related outcomes.^14,43,44 Continued efforts to ensure equitable allocation of HIV prevention and treatment resources, such as the EHE strategy of focusing on geographic areas with the highest burden of HIV, can help towards this effort.^45,46

The findings are subject to several limitations. First, the HIV diagnosis data are based on the national surveillance data and might not be representative of all persons with HIV because not all persons with HIV have been tested or tested at a time when the infection could be detected and diagnosed. Second, because the SDOH data were derived from information at the census tract level (e.g., Percentage of residents without health insurance coverage in the census tract of a person diagnosed with HIV) and not at the individual-level (e.g., Health insurance coverage status of a person diagnosed with HIV), the study findings are not suitable for individual-level interpretations due to ecological inference fallacy. Third, the ecological analysis precludes the examination of causality in the results. We are not able to establish that moving a census tract from the most disadvantaged quartile to the most advantaged quartile for a given SDOH factor would reduce absolute racial and ethnic disparities in HIV diagnosis rates in that census tract. Finally, we used the data based on the residence at the time of diagnosis of HIV, which could be different from the residence at the time of HIV acquisition.