INTRODUCTION
Advances in antiretroviral treatment accessibility and effectiveness mean that HIV infection can now be considered a chronic disease. This brings new priorities for clinical care, including quality of life (QOL) and everyday function. Optimizing these outcomes for people aging with HIV infection requires an understanding of both biological and psychosocial contributors.1–3
Even with good viral suppression, cognitive impairment is reported in 30%–50% of people with HIV in research cohorts, with prevalence increasing with age. This impairment is usually mild, but still can have consequences in everyday life.3–5 The underlying causes are not fully understood and the potential for reversibility is unknown. Current research has focused on biological factors, such as viral, cerebrovascular, and accelerated aging effects.3,5 However, psychosocial contributors are also likely. Mental health problems are common in this population and may share causal pathways or interact with cognitive impairment. Furthermore, although brain health indicators are typically conceived of as characteristics of the affected individual, the environment in which people live with HIV likely contributes to both cognitive performance and mental health.6
Among potentially relevant factors, stigma stands out. Operationalized as negative attitudes toward a person held by an individual, a group, or society at large, the experience of stigma remains an important aspect of living with HIV.7 In a recent systematic review of stigma in the combined anti-retroviral therapy (cART) era, over 50% of people with HIV reported experiencing stigma.8 Different facets of stigma are recognized, such as experienced, anticipated, and internalized stigma.9 Although these are important to understand for stigma-reduction purposes, they tend to show at least moderate intercorrelation.10 In general, stigma acts as a barrier to full participation of the individual in personal, family, and societal roles.11 In a recent study of QOL in a large sample of older people living in Canada with HIV infection,12 stigma was spontaneously reported as a priority area affecting QOL.13 That study also highlighted the special importance of stigma in HIV, as it was not reported on the same open-ended QOL questionnaire by people living with other serious chronic conditions, including stroke, multiple sclerosis, and cancer. The importance of stigma to QOL is well recognized in HIV research and care: Stigma is included as a domain in the World Health Organization's HIV-specific measure of QOL (WHOQOL-HIV).14
The relationship of stigma and mental health in people living with HIV has been extensively studied, as summarized in recent meta-analyses and systematic reviews.1,8,15 However, the impact of stigma on other aspects of brain function, notably cognition, has not been addressed. There are several potential mechanisms by which stigma could affect cognition. First, it could act through its impact on mental health, given that depression and anxiety may themselves influence cognition. Second, by affecting social experience, stigma might affect brain structure and function directly. Variation in social network size in healthy older adults has been associated with variation in the structure of specific brain regions and their interconnections,16 replicating findings in nonhuman primates assigned to social groups of different sizes.17 This work argues that the social environment can change brain structure. Third, as a chronic stressor, stigma may also affect cognition through other neurobiological mechanisms, with effects on the hypothalamic–pituitary axis,18,19 neuroinflammation,20,21 and cerebrovascular risk.22 Finally, cognitive performance can be affected by internalized stigma: that is, people respond to internalized negative stereotype expectations, performing worse in a testing situation than others not belonging to a stigmatized group.23
Here, we propose that HIV-related stigma has negative effects on both mental health and cognition, and that these effects will, in turn influence everyday functioning in people with HIV. Potentially complex relationships between stigma, cognition, and mental health have been suggested,21 but this view has yet to be tested empirically. Given this complexity, an understanding of the paths by which these variables affect real-world function is needed to guide work aiming to preserve or improve brain health in HIV. Here, we apply structural equation modeling (SEM) and a well-established conceptual framework, the World Health Organization's biopsychosocial model from the International Classification of Functioning, Disability and Health (ICF)11 to systematically address this complexity in a large, well-characterized sample drawn from the Positive Brain Health Now (BHN) cohort. This longitudinal cohort study of aging with HIV aims to characterize the contributors to and consequences of cognitive and mental health difficulties in older individuals on antiretroviral treatment, with well-controlled infection. The specific objective of the present analysis was to identify direct and indirect relationships among stigma, cognition, anxiety, depression, and everyday function in older white men living with HIV in Canada.
METHODS
Source of Data
The data for this analysis came from the BHN cohort, a prospective study involving 856 older persons living with HIV recruited between 2014 and 2016 from 5 clinics in Canada. The study protocol has been published.12 The project was approved by the research ethics boards of all participating institutions, and all participants provided written informed consent.
Participants
Participation in the BHN cohort was restricted to people aged 35 years and older, HIV+ for at least 1 year, and on stable cART for at least 6 months. Exclusion criteria included dementia of stage 3 or worse on the Memorial Sloan Kettering dementia severity scale,24 as judged by the treating physician (ie, a severity that precludes informed consent), a non–HIV-related neurological disorder likely to affect cognition, active central nervous system opportunistic infection, psychotic disorder, substance dependence or abuse within the past 12 months, or life expectancy of <3 years as judged by the treating physician (see Ref. 12 for details).
Of the 856 subjects in the BHN cohort, only data from the 512 white men were analyzed here to avoid the confounding effects of gender and race, which are known to also influence stigma in people living with HIV.15,25 White men were by far the largest demographic subgroup represented in this cohort, reflecting the demographics of the recruitment sources. Sample sizes of other racial subgroups or women in this data set were too small to support the SEM approach.
This project received funding from the Canadian Institutes of Health Research. The funding agency had no role in the design, data collection, analysis, or interpretation of the study, nor in the writing of the report or decision to submit the article for publication.
Measurement
The measurement framework for this study was the ICF. This provided an a priori model, avoiding the pitfalls arising from analyzing high-dimensional correlated data without specific hypotheses concerning the paths among variables. This model links physiological and structural variables leading to symptoms or impairments, in turn affecting activities and participation common in everyday function, all of which relate to health perception and QOL. We focused on the paths between stigma, cognitive, and mental health variables, but did so within this comprehensive model, systematically assessing the variables that influence stigma, tracing the downstream, real-world impact of stigma on brain health, and contextualizing stigma effects on real-world function relative to other paths in the model.
The data were collected from direct measurement, self-report questionnaires, and chart review. Table 1 presents the structure of the measurement model and an overview of the measures, as well as the demographic, clinical, and environmental factors included. All self-report measures are well-known instruments with strong psychometric properties.26–28 Cognitive performance was directly measured with a short battery of computerized tests of processing speed, attention, memory, and executive function [brief cognitive ability measure (B-CAM)].29 This battery was developed to measure cognitive ability in older people with HIV. Although multiple domains are assessed, empirical work guided by item-response theory has established that these tests reflect a single underlying latent variable in HIV. Rasch analysis allows a single score measuring overall cognitive ability to be assigned.30 Depression and anxiety symptoms were elicited with the Mental Health Index31 and the Hospital Anxiety and Depression Scale (HADS).32 Self-reported cognitive function limitations were ascertained with the 20-item Perceived Deficits Questionnaire.33 HIV-related experienced stigma was assessed using a single item from the Beliefs Domain of the WHOQOL-HIV BREF questionnaire “To what extent are you bothered by people blaming you for your HIV status?” which has 5 response options (“not at all” to “an extreme amount”).34 Details of all self-report measures, including example items, are provided as Supplementary Material (see Table S1, Supplemental Digital Content, https://links.lww.com/QAI/B234).
;)
TABLE 1.: Structure of the Measurement Model and Overview of the Measures
Statistical Methods
SEM was used to test the theoretical model against the observed data. SEM encompasses factor analysis, path analysis, and regression.35 Measured and latent variables (ie, variables for which no single measure can reflect the construct adequately) were included in the structural model; pathways between the variables were used to calculate direct effects. Single imputation was done to address potential bias arising from incomplete data using SAS 9.3 proc mi. Factor analysis was used to define anxiety and depression latent variables, both of which drew on the HADS and RAND-36 Mental Health Index.12
The SEM model for HIV-related stigma was developed sequentially, with the ICF model providing the theoretical framework. Generally, the strategy was to allow variables within the ICF rubrics to correlate, and to apply paths across the rubrics of the ICF model (Table 1). Paths were applied from personal factors, environmental factors, and biological factors to the rubrics of impairments, activity limitations, and participation restrictions. Under the impairments rubric, we focused on HIV-related symptoms, stigma, cognition, and mental health. Initially, the model only included adjacent paths between the ICF rubrics of impairments, activity limitations, and participation restrictions. Paths that were not statistically significant were removed, unless deemed theoretically relevant. Paths across more distant ICF rubrics were added later in model development and were retained when fit was improved. Although SEM assumes multivariate normality, many variable-specific measures were not normally distributed; so, robust maximum likelihood estimation was used. Model fit was determined by the Satorra–Bentler scaled χ2 and measures of approximate fit, which are more informative when sample sizes are large.
RESULTS
Table 2 presents the characteristics of the sample of 512 men (mean age 54 years, SD 8) on variables under the rubrics of the ICF model. As the imputed data were very close to the observed data, only the observed data (mean values and SDs) are provided. Mean duration of HIV infection was 17.4 years (SD 8). All participants were treated with cART. Ninety-two percent had complete viral suppression at the baseline visit and 96% had complete viral suppression at the first follow-up visit 9 months later, consistent with effective treatment and excellent cART adherence in this sample. This likely reflects the inclusion criteria for the study (stable cART for the previous 6 months at least) and a selection bias for good adherence in these older long-term survivors.
TABLE 2.: Description of the Sample Under the ICF Rubrics of Personal, Environmental, and Biological Factors
The final model is shown in Figure 1, emphasizing paths related to stigma and its downstream impact on brain health and real-world function. The values for all the direct effects are given in Table S2, Supplemental Digital Content, https://links.lww.com/QAI/B234. Table 3 shows the direct effects for the stigma and brain health variables that were the conceptual focus of this study. Figure 2 illustrates the relative impact of these variables. The final model fit the data well: χ2 (χ2 test of exact fit, using the Satorra–Bentler correction for non-normality, and associated degrees of freedom) =128.7, df = 70, P < 0.05 [although the Satorra–Bentler χ2 was statistically significant, this was offset by the very large sample size and low ratio of the χ2 to its degrees of freedom (1.8)]; root mean square error of approximation (a measure of global close fit where values less than 0.05 represent good fit) =0.040; standardized root mean square residual (a measure of badness of fit based on fitted residuals; values less than 0.05 represent good fit and values to 0.10 represent reasonable fit) =0.027; comparative fit index (CFI) =0.980; Tucker-Lewis index (TLI) =0.963 (CFI and TLI values greater than 0.95 indicate acceptable fit, 0.97 indicate good fit; both measure fit relative to an independent model, but the TLI includes a correction for model complexity). HIV-related stigma had direct effects on cognitive test performance and anxiety. There was also a direct but weaker path from stigma to depression (P < 0.1; shown by the dotted line in Fig. 1), retained in the model because the existing literature argues for links between stigma and depression, and between depression and cognition.
FIGURE 1.: Final SEM model (N = 512) for HIV-related stigma based on the ICF model. While the variables within each ICF rubric were allowed to correlate, these correlations are not shown. Each rubric is coded in a different color. Paths from stigma are emphasized; all other paths are reported in Table S2, Supplemental Digital Content,
https://links.lww.com/QAI/B234. Solid lines indicate significance at
P < 0.05; the dashed line indicates significance at
P < 0.10. This weaker relationship was retained as theoretically relevant.
Direct Effects of Model Variables, Focusing on the Paths to and From Stigma, Organized by the Rubrics of the ICF Model
FIGURE 2.: Magnitudes of the effects of the key path parameters relating stigma, brain health, and everyday function. These are expressed in standardized units (stdXY), allowing direct comparison.
Stigma was influenced by HIV-related variables (duration of infection and HIV-specific signs and symptoms). Figure 1 also shows that duration of HIV and presence of HIV-specific signs (including physical indicators of HIV infection such as changes in body shape) have an impact on other variables in the model such as depression and physical function, separately from their effect on stigma. Poor social support was also a contributor to stigma as well as to depression, anxiety, and worse physical function. The influence of stigma on cognitive test performance and mood in turn had widespread downstream effects on real-life function and participation. This included distinct paths from cognitive test performance and anxiety to self-reported cognitive difficulties, and from depression to physical function. Quality of the environment and age affected all ICF model variables, represented in Figure 1 by an arrow to the label “ICF Model” (see Table S2, Supplemental Digital Content, https://links.lww.com/QAI/B234).
Figure 2 presents the magnitudes of the effects of the key path parameters relating stigma, brain health, and everyday function. These are expressed in standardized units (stdXY) derived from the regression parameters in the path model (see Table 3 and Table S2, Supplemental Digital Content, https://links.lww.com/QAI/B234) to allow for direct comparison of effects across variables. Stigma was most highly associated with anxiety, with cognitive test performance the second strongest association. In turn, cognitive test performance was associated with 2 variables reflecting everyday cognitive functioning: degree of engagement in meaningful activities and self-reported cognitive difficulties. Of these associations, cognitive test performance had the strongest relationship with self-reported cognitive difficulties. Finally, self-reported cognitive difficulties had effects on social role and, to a somewhat greater extent, life-space mobility. To put the relative magnitude of these stigma and brain health path parameters in context, their strengths ranged from about one-third to one-half of the strongest relationship in the full model (ie, the path between depression and social role), which had a stdXY of 0.56 (see Table S2, Supplemental Digital Content, https://links.lww.com/QAI/B234).
DISCUSSION
As hypothesized, stigma had a direct effect on cognitive performance, in addition to its effects on mood [anxiety and (more weakly) depression] in this sample of older white men with well-controlled HIV infection. Through these variables, stigma affected everyday function, including physical function, self-reported cognitive difficulties, and engagement in meaningful activities. Downstream effects were observed for social role and life-space mobility (ie, the space in which people act, ranging from their own homes outward to the larger community).
These findings could suggest either that people who report feeling stigmatized due to their HIV status avoid social and community activities or that feelings of stigma arise from being excluded from these activities.36,37 These cross-sectional data do not allow the direction of the effects to be established. Indeed, the direction may differ across people. For example, in the model here, social support (which includes loneliness and social network variables) is shown as affecting stigma, but the relationship could be in the opposite direction (stigmatization leading to loneliness). Qualitative studies could clarify these directions and identify variation in experience.
The impact of stigma on mood and other health outcomes replicates the literature in other stigmatized populations38 as well as in HIV.1,22 Again, the directionality of these relationships is uncertain, as people with depression may pay more attention to experienced stigma, or ruminate about those events. Recent work has linked poorer cognition with loneliness in people with HIV,39 and stigma could contribute to loneliness.
Our study is novel in that we have shown a direct effect between stigma and cognitive performance in HIV. Multiple mechanisms likely underpin this association, opening avenues for future research on potentially modifiable social–environmental contributors to cognitive difficulties in older people with well-controlled HIV infection. These factors likely have impact beyond the psychosocial realm: Recurrent negative social experience or isolation can have direct effects on brain structure and function, through routes as varied as experience-driven neuroplasticity,17 chronic stress-related inflammation,21 and cerebrovascular injury.40 These insights suggest points of potential contact between the psychosocial and biological effects of HIV infection, 2 important but so far largely parallel themes of research in HIV. Stigma can affect cART adherence, in turn leading to greater HIV-related brain injury. However, this is unlikely to be a major factor here, as well over 90% of this sample had consistently undetectable viral load, suggesting excellent adherence.
Strengths of this study include the use of a strong theoretical model and a robust statistical approach ideally suited for high-dimensional, correlated data. Other work using a similar approach to stigma in black Caribbean women living with HIV in Canada reported similar interrelationships between stigma, social support, and depression, as well as with health perception, but did not address cognition.25 We found that quality of the environment, along with age, was important for all variables in the model, including stigma. Interestingly, the oldest men in our cohort expressed less stigma, perhaps reflecting a selection effect into the study, or a survival bias, or both.
This study has limitations. First, inclusion of only white men limits generalizability. This restriction was planned, to isolate HIV-related stigma from other well-known sources of stigma (gender and race); it seems likely that additional demographic sources of stigma would magnify the effects we observed in this restricted sample, but further work is needed to test this possibility. Guided by our findings, such work could take simpler statistical approaches, requiring smaller samples. Second, stigma was not a primary focus of the main BHN study and was only assessed with a single item. However, the use of an open-ended QOL measure was a planned feature of the main study, and this identified stigma as a uniquely important aspect of QOL in HIV,13 motivating the current analysis of the paths linking stigma and brain health. Future work would benefit from more extensive measurement of stigma; the item we used asked about the person's experience of stigma (ie, how distressed they are by perceived HIV-related social exclusion) rather than characterizing stigmatizing features of the environment. This item also does not disambiguate constructs such as internalized and anticipated stigma.41 It would be helpful to assess these facets of stigma in future work, as the literature shows that they may have distinct effects on physical and mental health,1,9 and they may require different stigma-reduction strategies. Finally, we assessed cognitive performance with a relatively brief battery of computerized tests. We have shown that these tests can be summarized as a global, continuous measure of cognition reflecting processing speed, attention, memory, and aspects of executive function relevant to HIV-associated cognitive impairment.3,29 However, they do not permit classification according to the current HIV-Associated Neurocognitive Disorder nosology.
HIV-related stigma has multiple contributors and widespread repercussions. Here, we show that it has distinct effects on different facets of brain health, notably including cognition. This argues for a broader view of the factors that affect cognition in HIV, pointing to toxic effects of an adverse social environment on the brain. This may not be unique to HIV, with extensive evidence for impact of loneliness and social exclusion on general health and cognitive decline in aging in the general population.40 This opens promising directions for research and program development aimed at supporting brain health in people living with HIV by intervening on societal factors that contribute to the experience of stigma and personal factors that bolster resilience in the face of such experiences.
REFERENCES
1. Rueda S, Mitra S, Chen S, et al. Examining the associations between HIV-related stigma and health outcomes in people living with HIV/AIDS: a series of meta-analyses. BMJ Open. 2016;6:e011453.
2. Chambers LA, Wilson MG, Rueda S, et al. Evidence informing the intersection of HIV, aging and health: a scoping review. AIDS Behav. 2014;18:661–675.
3. Heaton RK, Franklin DR, Deutsch R, et al. Neurocognitive change in the era of HIV combination antiretroviral therapy: the longitudinal CHARTER study. Clin Infect Dis. 2015;60:473–480.
4. Simioni S, Cavassini M, Annoni JM, et al. Cognitive dysfunction in HIV patients despite long-standing suppression of viremia. AIDS. 2010;24:1243–1250.
5. Clifford DB, Ances BM. HIV-associated neurocognitive disorder. Lancet Infect Dis. 2013;13:976–986.
6. Vance DE. The cognitive consequences of stigma, social withdrawal, and depression in adults aging with HIV. J Psychosoc Nurs Ment Health Serv. 2013;51:18–20.
7. Fredericksen RJ, Edwards TC, Merlin JS, et al. Patient and provider priorities for self-reported domains of HIV clinical care. AIDS Care. 2015;27:1255–1264.
8. Lowther K, Selman L, Harding R, et al. Experience of persistent psychological symptoms and perceived stigma among people with HIV on antiretroviral therapy (ART): a systematic review. Int J Nurs Stud. 2014;51:1171–1189.
9. Earnshaw VA, Smith LR, Chaudoir SR, et al. HIV stigma mechanisms and well-being among PLWH: a test of the HIV Stigma Framework. AIDS Behav. 2013;17:1785–1795.
10. Reinius M, Wettergren L, Wiklander M, et al. Development of a 12-item short version of the HIV stigma scale. Health Qual Life Outcomes. 2017;15:115.
11. WHO. The international classification of functioning, disability and health. World Heal Organ. 2001;18:237.
12. Mayo NE, Brouillette MJ, Fellows LK; Positive
Brain Health Now Investigators. Understanding and optimizing
brain health in HIV now: protocol for a longitudinal cohort study with multiple randomized controlled trials. BMC Neurol. 2016;16:8.
13. Mayo NE, Aburub AA, Brouillette MJ, et al. In support of an individualized approach to assessing quality of life: comparison between Patient Generated Index and standardized measures across four health conditions. Qual Life Res. 2017;26:601–609.
14. Skevington SM, Lotfy M, O'Connell KA. The World Health Organization's WHOQOL-BREF quality of life assessment: psychometric properties and results of the international field trial a report from the WHOQOL Group. Qual Life Res. 2004;13:299–310.
15. Logie C, Gadalla TM. Meta-analysis of health and demographic correlates of stigma towards people living with HIV. AIDS Care. 2009;21:742–753.
16. Noonan MPM, Mars RB, Sallet J, et al. The structural and functional
brain networks that support human social networks. Behav
Brain Res. 2018;355:12–23.
17. Sallet J, Mars RB, Noonan MP, et al. Social network size affects neural circuits in macaques. Science. 2011;334:697–700.
18. Busse D, Yim IS, Campos B, et al. Discrimination and the HPA axis: current evidence and future directions. J Behav Med. 2017;40:539–552.
19. de Souza-Talarico JN, Marin MF, Sindi S, et al. Effects of stress hormones on the
brain and cognition: evidence from normal to pathological aging. Dement Neuropsychol. 2011;5:8–16.
20. Menard C, Pfau ML, Hodes GE, et al. Social stress induces neurovascular pathology promoting depression. Nat Neurosci. 2017;20:1752–1760.
21. Valdez AN, Rubin LH, Neigh GN. Untangling the Gordian knot of HIV, stress, and cognitive impairment. Neurobiol Stress. 2016;4:44–54.
22. Crockett KB, Rice WS, Turan B. Associations between multiple forms of discrimination and tobacco use among people living with hiv. J Acquir Immune Defic Syndr. 2018;78:9–15.
23. Smith JL. Understanding the process of stereotype threat: a review of mediational variables and new performance goal directions. Educ Psychol Rev. 2004;16:177–206.
24. Price RW, Brew BJ. The AIDS dementia complex. J Infect Dis. 1988;158:1079–1083.
25. Logie CH, Wang Y, Lacombe-Duncan A, et al. HIV-related stigma, racial discrimination, and gender discrimination: pathways to physical and
mental health-related quality of life among a national cohort of women living with HIV. Prev Med. 2018;107:36–44.
26. Community Healthy Activities Model Program for Seniors II (CHAMPS II): Program Manual
. San Francisco, CA: University of California, San Francisco: Institute for Health & Aging; 2003.
27. Peel C, Sawyer Baker P, Roth DL, et al. Assessing mobility in older adults: the UAB study of aging life-space assessment. Phys Ther. 2005;85:1008–1119.
28. Holzemer WL, Hudson A, Kirksey KM, et al. The revised sign and symptom check-list for HIV (SSC-HIVrev). J Assoc Nurses AIDS Care. 2001;12:60–70.
29. Brouillette MJ, Fellows LK, Palladini L, et al. Quantifying cognition at the bedside: a novel approach combining cognitive symptoms and signs in HIV. BMC Neurol. 2015;15:224.
30. Koski L, Brouillette MJ, Lalonde R, et al. Computerized testing augments pencil-and-paper tasks in measuring HIV-associated mild cognitive impairment. HIV Med. 2011;12:472–480.
31. Hays RD, Morales LS. The RAND-36 measure of health-related quality of life. Ann Med. 2001;33:350–357.
32. Zigmond AS, Snaith RP. The hospital anxiety and depression scale (HADS). Acta Psychiatr Scand. 1983;67:361–370.
33. Sullivan MJ, Edgley K, Dehoux E. A survey of multiple sclerosis: I. Perceived cognitive problems and compensatory strategy use. Can J Rehabil. 1990;4:99–105.
34. O'Connell KA, Skevington SM. An international quality of life instrument to assess wellbeing in adults who are HIV-positive: a short form of the WHOQOL-HIV (31 items). AIDS Behav. 2012;16:452–460.
35. Kline RBR. Principles and Practice of
Structural Equation Modeling. Vol. 156. 3rd ed. New York, NY: Guilford; 2011.
36. Emlet C. An examination of the social networks and social isolation in older and younger adults living with HIV/AIDS. Health Soc Work. 2006;31:299–308.
37. Webel AR, Longenecker CT, Gripshover B, et al. Age, stress, and isolation in older adults living with HIV. AIDS Care. 2014;26:523–531.
38. Wu YK, Berry DC. Impact of weight stigma on physiological and psychological health outcomes for overweight and obese adults: a systematic review. J Adv Nurs. 2018;74:1030–1042.
39. Han SD, Adeyemi O, Wilson RS, et al. Loneliness in older black adults with human immunodeficiency virus is associated with poorer cognition. Gerontology. 2017;63:253–262.
40. Leigh-Hunt N, Bagguley D, Bash K, et al. An overview of systematic reviews on the public health consequences of social isolation and loneliness. Public Health. 2017;152:157–171.
41. Earnshaw VA, Chaudoir SR. From conceptualizing to measuring HIV stigma: a review of HIV stigma mechanism measures. AIDS Behav. 2009;13:1160–1177.
