Due to decades of structural and institutional racism, minoritized individuals in the US are more likely to live in low socioeconomic neighborhoods, which may underlie the observed greater risk for neurocognitive impairment as they age. However, these relationships have not been examined among people aging with HIV. To investigate neurocognitive disparities among middle- and older-aged Latino and non-Latino White people living with HIV (PWH), and whether neighborhood socioeconomic deprivation may partially mediate these relationships.

Participants were 372 adults ages 40-85 living in southern California, including 186 Latinos (94 PWH, 92 without HIV) and 186 non-Latino (NL) Whites (94 PWH, 92 without HIV) age-matched to the Latino group (for the overall cohort: Age M=57.0, SD=9.1, Education: M=12.7, SD=3.9, 38% female; for the group of PWH: 66% AIDS, 88% on antiretroviral therapy [ART]; 98% undetectable plasma RNA [among those on ART]). Participants completed psychiatric and neuromedical evaluations and neuropsychological tests of verbal fluency, learning and memory in person or remotely. Neuropsychological results were converted to demographically-unadjusted global scaled scores for our primary outcome. A neighborhood socioeconomic deprivation variable (SESDep) was generated for census tracts in San Diego County using American Community Survey 2013-2017 data. Principal components analysis was used to create one measure using nine variables comprising educational (% with high school diploma), occupational (% unemployed), economic (rent to income ratio, % in poverty, (% female-headed households with dependent children, % with no car, % on public assistance), and housing (% rented housing, % crowded rooms) factors. Census tract SESDep values were averaged for a 1km radius buffer around participants’ home addresses.

Univariable analyses (independent samples t-tests and Chi-square tests) indicated Latinos were more likely to be female and had fewer years of formal education than NL-Whites (ps<.05). Latino PWH had higher nadir CD4 than White PWH (p=.02). Separate multivariable regression models in the overall sample, controlling for demographics and HIV status, showed Latinos had significantly lower global scaled scores than Whites (b=-0.59; 95%CI-1.13, -0.06; p=.03) and lived in more deprived neighborhoods (b=0.62; 95%CI=0.36, 0.88; p<.001). More SES deprivation was significant associated with worse global neurocognition in an unadjusted linear regression (b=-0.55; 95%CI=-0.82, -0.28; p<.001), but similar analyses controlling for demographics and HIV status, showed SESDep was not significantly related to global scaled scores (b=-0.11; 95%CI= -0.36, 0.14; p=.40). Exploratory analyses examined primary language (i.e., English vs Spanish) as a marker of Hispanic heterogeneity and its association with neurocognition and SESDep. Controlling for demographics and HIV status, both English-speaking (b=0.33; 95%CI=0.01. 0.64; p=.04) and Spanish-speaking Latinos (b=0.88; 95%CI=0.58, 1.18; p<.001) lived in significantly greater SESDep neighborhoods than Whites, with SESDep greater for Spanish-speakers than English-speakers (p<.001). However, only English-speaking Latinos had significantly lower neurocognition than Whites (b=-0.91; 95%CI=0-1.57, -0.26; p<.01; Spanish-speakers: b=-0.27; 95%CI=-0.93, 0.38; p=.41).

Among our sample of diverse older adults living with and without HIV, English-speaking Latinos showed worse neurocognition than Whites. Though SES neighborhood deprivation was worse among Latinos (particularly Spanish-speakers) it was not associated with neurocognitive scores after adjusting for demographics. Further studies investigating other neighborhood characteristics and more nuanced markers of Hispanic heterogeneity (e.g., acculturation) are warranted to understand factors underlying aging and HIV-related neurocognitive disparities among diverse older adults.