UCSF

Background: Since the onset of the HIV epidemic, antiretroviral therapy (ART) has lengthened the lives of people living with HIV (PLWH). More than half of the population of PLWH are 50 years and older, and as the aging population of PLWH expands it is important to consider the environmental, socio-structural, and biological factors and pathways that impact aging with HIV (Centers for Disease Control and Prevention [CDC], 2018). It is estimated that about a third to one half of HIV-seropositive individuals have some degree of cognitive impairment (Heaton et al., 2010), though prevalence may be lower in populations with sustained viral suppression (V. Valcour, personal communication, December 2018). Studies show that various measures of loneliness and social isolation are associated with risk for cognitive impairment and dementias among the general population (Kuiper et al., 2015, 2019). Literature also indicates that, on average, PLWH experience more loneliness and social isolation than HIV-seronegative individuals (Greene et al., 2018; Grov et al., 2010; Poindexter & Shippy, 2008). In addition to the environmental and social factors that impact risk for cognitive impairment among PLWH, the study of inflammatory processes and biomarkers—and whether these are important for the development of HAND—is an important area of current research (Office of AIDS Research [OAR] Working Group on HIV and Aging, 2012). Current evidence demonstrates an overall increased risk of morbidity and mortality in HIV-positive individuals whose CD4/CD8 ratio fails to normalize above 1.0 (Mussini et al., 2015; Serrano-Villar et al., 2013, 2014). Additionally, a small number of studies show that inverted CD4/CD8 ratio is associated with the development of neurocognitive disorders (Correa et al., 2014; Grauer et al., 2015; Rawson et al., 2015; Vassallo et al., 2017). Problem Statement: The central hypotheses motivating this dissertation research is that loneliness and social isolation represent two distinct and uniquely important factors in the risk of cognitive impairment among people aging with HIV, and that CD4/CD8 ratio is an endophenotype that may help elucidate the pathways by which physical health, mental health, and environmental factors impact cognitive impairment among PLWH. Chapter 2: The 2nd chapter of this dissertation systematically examined the current body of quantitative literature about social support and loneliness and cognitive impairment among PLWH. We used meta-analysis to summarize the association of these variables and used meta-regression to identify moderating variables in N=11 studies. Among the 11 studies reviewed, many were limited by the use of un-validated measures of loneliness and/or social support as well as heterogeneous measures of cognitive symptoms and cognitive performance. The meta-analysis (n=10) showed a positive association between social isolation or loneliness and cognitive impairment. Though there was moderate heterogeneity among the studies analyzed, there was not substantial publication bias. Meta-regression showed moderation of the association by study quality, older age (≥55 years), and study country but not by sample mean CD4 cell count. This paper highlights knowledge gaps in the current body of research and reflects distinctions between performance-based and self-reported measures of cognitive impairment as well as various dimensions of social connectedness. Chapter 3: Building on the knowledge gaps identified in the second dissertation chapter, the 3rd chapter of this dissertation examined the association of both loneliness and social isolation with performance based cognitive impairment and subjective cognitive complaints in a sample of older adults living with HIV and confirmed HIV-Associated Neurocognitive Disorders (HAND). We performed a cross-sectional, secondary data analysis in a cohort of older adults living with HIV recruited at the Memory and Aging Center (N=171). This paper revealed that loneliness was correlated with mental health variables (depression, anxiety, and perceived stress) while social isolation was correlated with other marginalized conditions and socioeconomic factors (lower years of education, history of Hepatitis C (HCV), history of a substance use disorder, and Black/African American race, and area-level socioeconomic environment). This paper also showed that social isolation, but not loneliness, was associated with higher odds of impairment in two cognitive domains (Attention and Executive function [ATT] and Speed of Processing [SPD]). We concluded that social isolation may be conceptualized as a pathway of embodiment, reflecting experiences of marginality that could impact risk of HAND. Chapter 4: The 4th chapter of the dissertation used an exploratory, cross-sectional secondary analysis of the the Hawaii Aging with HIV Cardiovascular Study (HAHCS) cohort-Public Data Set. We aimed to examine the relationships between CD4/CD8 and biomarkers of inflammation among middle-aged and older adults living with HIV (N=103). This study revealed that the lowest tertile of CD4/CD8 ratio (Median: 0.346, range 0.123-0.501) was associated with a higher inflammation profile and higher concentration of mature monocytes in the blood. In elucidating the CD4/CD8 cell ratio among PLWH, this paper contributes to the body of evidence that suggests a link between CD4/CD8 ratio and established risk factors for HAND among middle-aged and older adults with HIV. Chapter 5: The 5th chapter of the dissertation was a longitudinal study of the Women’s Interagency Health Study (WIHS) public data set to examine the intra-individual variability of the CD4/CD8 ratio over 10 years as well as clinical and sociodemographic correlates of CD4/CD8 ratio among women living with HIV (n=1462). This study revealed that, over 10 years, the CD4/CD8 ratio remained relatively stable. Over time, decreased CD4/CD8 ratio was associated with higher age, detectable viral load, and sub-optimal ART adherence. We then examined associations of clinical and sociodemographic with “inverted” CD4/CD8 ratio (≤ 1.0 versus >1.0) and “low” CD4/CD8 ratio (≤ 0.70 versus >0.70). This paper revealed that inverted and low CD4/CD8 ratio were both associated with detectable viral load and fewer years of educational attainment. Because much of the WIHS sample had a CD4/CD8 ratio that was below 0.50, in addition to the findings from chapter 4 that indicate inflammation is association with the lowest tertile of CD4/CD8 ratio (0.123-0.501), we subsequently performed a follow-up analysis to examine variables associated with “very low” CD4/CD8 ratio (≤0.50 versus >0.50). We found that very low CD4/CD8 ratio was associated with older age, detectable viral load, past/current Hepatitis C, white racial group, and lower educational attainment. Overall, the findings from this paper show that CD4/CD8 ratio was relatively stable over time, but was significantly influenced by age, viral load, and adherence to ART medications. We also found that inverted, low, and very CD4/CD8 ratio were all associated with fewer years of education. In linking CD4/CD8 ratio to lower educational attainment, our study implicates immune function as a pathway by which root causes and social determinants of health (SDOH) may impact risk of cognitive impairment among women with HIV.