UC San Diego

Rationale: Neighborhoods adversely impact neurocognitive impairment and depression among older adults without HIV. These associations are unknown among diverse older PWH, who have higher rates of neuropsychiatric outcomes. Negative repercussions of COVID-19 disproportionally impacted vulnerable communities. We aimed to 1) examine associations between neighborhood characteristics and neurocognitive decline and worsened mood during the COVID-19 pandemic among older PWH, and 2) determine if the neighborhood impact of COVID-19, as a severe community stressor, moderates these relationships.Design: Participants were 180 PWH 50+ years, enrolled in observational studies at the HIV Neurobehavioral Research Center. Seventy-nine completed neuropsychiatric assessments at two time-points: (1) March 2019-March 2020; and (2) March 2021-June 2022. Negative change in global mean scaled scores (GSS) reflected neurocognitive decline. Negative change in the Medical Outcomes Study HIV Health Survey Mental Health Summary T-scores (MSS) indicated worsened mood. Exploratory factor analyses of publicly available census and satellite data created ‘Sociodemographic’ ‘Economic’ and ‘Undeveloped’ neighborhood factors. Cumulative COVID-19 cases rates per zip code reflected the neighborhood impact of COVID-19. Participant’s home addresses were geocoded and linked to neighborhood data. Multivariable linear regression models investigated whether neighborhood factors were related to changes in GSS and MSS, and the moderating effect of COVID-19 case rates. Results: Average change in GSS was 0.05 (SD=0.99) scaled scores. Average MSS change was -1.76 (SD=6.91) T-scores. The average percent of COVID-19 cases was 23%. ‘Sociodemographic,’ ‘Economic,’ and ‘Undeveloped’ factors were not related to significantly greater neurocognitive decline or worsened mood over time (ps>.05). There was no significant moderating effect of COVID-19 cases on cognitive or emotional decline (ps>.05). Exploratory analyses suggested that living in neighborhoods with greater crowded households was significantly related to decline in MSS (b=-0.45, p=.04). Pooled estimates after multiple imputation procedures did not uphold this finding (b=-0.12, p=.49). Conclusions: Neighborhood features did not relate to neurocognitive or emotional change among older diverse PWH during a historically stressful time. Lack of substantial change may have contributed to null findings. Additional work with larger samples at risk for neuropsychiatric decline may elucidate how heterogenous environmental exposures may lead to positive and negative health outcomes in specific populations.