Redefining Aging in HIV Infection Using Phenotypes

Comorbidities in HIV infection, in the broadest sense, consist of a spectrum of conditions including medical (e.g., cardiovascular disease, diabetes, hypertension, osteoporosis, hepatic and renal disease, chronic respiratory disorders, cancers, HCV co-infection), neuropsychiatric comorbidities (e.g., substance use, depressive, and sleep disorders), and social comorbidities (e.g., educational disadvantage, poverty, chronic trauma, illiteracy) [19]. The extended exposure of older adults to both HIV and antiretroviral drugs appears to increase their risk of illness and death from HIV-associated non-AIDS (HANA) comorbidities, such as cardiovascular disease (including myocardial infarction [MI]), cerebrovascular accident [CVA], and other aforementioned illnesses not directly associated with HIV infection [20]. HANA comorbidities affect timely initiation of ART in older adults, suggesting the need for improved coordination and integration of care and management of prescription practices [21]. Previously described findings of racial disparities in the quality of care for persons with HIV infection [22] have been extended to common comorbidities suggesting the need for implementation of cross-cutting multilevel interventions [23].

The largest epidemiological study (Veterans Administration Cohort Study, VACS) addressing the prevalence of HANA comorbidities in HIV indicated a prevalence of 60–63% [20]. Some investigators have reported a yet higher prevalence of 94%, with an average of three comorbid conditions in HIV-infected persons over age 50 living in the USA [12]. Furthermore, among older HIV-infected individuals, increased psychiatric comorbidities are linked to decreased quality of life [24]. The neuropsychiatric aspects of HIV and aging have been reviewed elsewhere [25].

Major depressive disorder (MDD) is the most common psychiatric diagnosis in HIV/AIDS at a prevalence over twice the rates seen in the general population [26, 27]. However, whether older HIV-infected adults experience higher rates of MDD than younger adults is unclear. Some studies report depression rates increased with age [28], while others report absence of the usual age-dependent decline in depression for older HIV-infected adults [29]. Further, it has been suggested that MDD is an independent risk factor for heart failure in HIV+ adults reinforcing the importance of identifying and managing MDD among HIV+ patients and of identifying interventions to reduce cardiovascular morbidity and mortality in those with both HIV and MDD [30].

Substance use disorder comorbidities are also common in older patients and put them at higher risk for HIV infection. Substance use accounts for over 16% of new HIV infection in people aged 50 and older [31]. Unlike the general population where substance use disorder rates decline in people age over 50, older HIV-infected patients maintain steady rates of substance use disorder [28, 29]. In more detailed analyses, patterns of recent substance use were associated with varying levels of HIV medication adherence and perceived substance use impairment, indicating that substance type matters when considering the health of aging adults living with HIV and that multiple-substance use needs to be addressed by interventions aimed at improving medication adherence [32]. Current and lifetime alcohol and substance use disorders may be compounding brain dysfunction, specifically in older HIV-infected individuals. Thus, a study [33] found that there was a legacy effect in older methamphetamine-using HIV seropositive individuals compared to their younger counterparts. Another study [34] found that lifetime alcohol use disorder and old age were associated with significant reduction in brain volumes.

Approximately 50% of persons living with HIV suffer from impaired cognition, a frequency that increases with age [35]. Moreover, impaired cognition has adverse impact on everyday functioning and quality of life [36]. Clinically, HIV-associated neurocognitive impairment (NCI) is associated with a variety of symptoms, including problems in attention, concentration, learning, memory, psychomotor ability, and speed of information processing. There is considerable heterogeneity in the cognitive patterns among HIV-infected individuals, requiring a focus on distinguishing components of the cognitive aging phenotype. HIV infection often results in varying degrees of neurocognitive dysfunction, ranging from asymptomatic NCI to HIV-associated dementia (HAD), collectively termed HIV-associated neurocognitive disorders [37]. The nature of HIV-associated NCI has evolved considerably over time, particularly the spectrum of neurocognitive disorders in the era of combination antiretroviral therapy (cART) [38]. As widely reported elsewhere, there has been a significant decline in HAD but a higher persistence of the milder categories, with frequently cited percentages of 2% for HAD and as much as 50% for mild neurocognitive disorder taken together with asymptomatic neurocognitive disorder [37, 39]. Thus, further endophenotypic separation within the cognitive phenotype in HIV clinical conditions has assumed particular importance. Additionally, there has been a shift pertaining to types of cognitive domains that appear to be most notably impacted [38]. In the pre-cART era, HIV-infected individuals were more likely to demonstrate impairment in motor skills, information processing speed, and verbal fluency, whereas patients receiving cART today may be more likely to exhibit difficulties with memory and executive functioning [38]. Also relevant is the notion that there might be different trajectories of cognitive change over time among HIV seropositive persons. In a recent multi-site study of longitudinal neurocognitive change in HIV-infected persons living in the USA, most remained stable (61%), nearly a quarter declined (23%), and others improved (17%) cognitively [40]. Neurocognitive change appears to be driven by a complex set of risk factors involving HIV disease, its treatment, and comorbid conditions, and favorable neurocognitive outcomes may be achieved by instituting cART early [40]. Consistent with these findings, others have shown that cognitive declines may not necessarily occur in all HIV seropositive persons, suggesting that future interventions need to target distinct cognitive sub-groups. More specifically, two clusters have been described with a lower performing cluster exhibiting poorer performance across all domains (except psychomotor speed) and a “normal” cluster displaying similar performance as the HIV seronegative group [13]. Using longitudinal data from the MACS cohort, there were deleterious effects of aging and more severe stage of HIV infection on neurocognitive function, generally, with an additional deleterious interaction effect on episodic memory and motor performance [41].

Medical comorbidities occur earlier and more frequently in older persons living with HIV infection; some of its consequences include neurocognitive and functional impairment, frailty, organ system failure, and increased hospitalizations [78]. The highly prevalent medical comorbidities further complicate assessment, diagnosis, and interventions for HIV-associated neurocognitive disorder (19). More specifically, a greater medical comorbidity burden (mostly for diabetes and malignancy), assessed by the Charlson Comorbidity Index [79], in older HIV-infected adults was associated with global NCI and related functional declines and lower physical health-related quality of life [12]. Other data in older HIV-infected adults are supportive of this finding—reporting that these individuals develop metabolic dysfunction (and other comorbidities) disorders leading to type 2 diabetes and NCI, through depleting cognitive reserve, that can impair everyday functioning and reduce quality of life [80]. In line with these findings, changes in VACS Index scores, which is a composite marker of HIV disease severity, including age, HIV-related factors (current plasma HIV RNA load and CD4 cell counts) and “non-HIV-associated biomarkers” (indicators of renal and hepatic function, anemia, and HCV co-infection), correspond to longitudinal changes in neurocognitive function among middle/aged HIV-infected individuals [81].

Most of the empirical overlapping relationships between comorbid and cognitive aging are associative, with few studies demonstrating temporal relationships (Table 2). These correlational relationships may potentially be attributed to factors that are causal (i.e., one phenotype increases risk for the other), genetic (i.e., heritable genetic factors or common liability for familial aggregation), or phenomenological (i.e., shared behavioral features). Direct causality cannot be inferred from these correlational studies, as effects may be due to interactions of independent components. One hypothesis to explain the comorbidity/cognition interactions is that comorbidities pose an additional burden on the cognitive functioning of HIV-infected individuals so that cognitive deficits are exacerbated by the presence of comorbidities. This interaction may be evident in aging/HIV-infected individuals because of the co-occurrence of aging-related conditions. These comorbidities and associated cognitive impairment may also be attributable to the chronic HIV-induced inflammatory state [82]. HIV-associated comorbidities and inflammation may accelerate the aging process, which may detrimentally impact everyday functioning via declines in cognitive functioning [83]—highlighting the importance of the “accelerated aging” phenotype. In this view, comorbid and accelerated aging phenotypes may have causal consequences on other aging phenotypes. Other possibilities for the adverse impact of HIV-associated mental health comorbidities on cognition include (1) competing for cognitive resources through rumination (i.e., depressive or suicidal thoughts that prevent other thoughts from emerging), (2) displaying different cognitive profiles (e.g., executive dysfunction is common in schizophrenia), and (3) a convergence of compromising vascular effects (diabetes mellitus, hypertension, and HIV infection). Related to these possibilities, models have been proposed for the synergistic effects of HIV, type 2 diabetes, and aging on cognition [80]. To further elucidate the role of each of these mechanisms would require separating out components and domains of the cognitive phenotype.

Comorbid aging or cognitive aging also overlap with other phenotypes (i.e., frailty, disability, and accelerated aging), but they are distinctive in several respects. Correlated associative phenotypic findings are widespread, and few studies have explored their temporal likelihood of developing or how they arise and evolve (Table 2). The only exceptions to the rule of associative relationships are comorbidity predicting frailty [64] and cardiovascular comorbidities leading to cognitive decline [76]. Among a large cohort of men who have sex with men (MSM) from the Multicenter AIDS Cohort Study (MACS), medical comorbidities predicted the frailty phenotype with the strongest predictability for HCV co-infection, depressive symptoms, history of diabetes, and renal disease (in decreasing order of predictability) [64], leading to the conclusion that comorbidities and frailty are not synonymous; rather, comorbidities are potential etiologic agents for frailty, as originally hypothesized [9]. Thus, these phenotypes are most likely not orthogonal to one another but rather they will overlap in certain respects requiring further research to dissect them out and maintain construct validity across various older HIV-infected samples. Aging-related phenotypes may also be used to generate a multi-factorial vector toward deleterious aging outcomes of interest and are best studied in this combined fashion.

An issue of concern for HIV phenotype research is that these aging phenotypes have been identified, by and large, in non-HIV-infected populations, so that they must now be reassessed and evaluated in vulnerable HIV-infected populations (e.g., MSMs). This limits generalizability to other aging HIV-infected populations. To date, most studies examining comorbid and/or cognitive aging in the HIV-infected population have been cross-sectional in nature. The temporal relationship between aging-related phenotypes is not clearly understood, and longitudinal data are needed to better describe the trajectory of cognitive decline and how specific comorbidities contribute to other aging-related phenotypes. Finally, it is unclear from the current results how long-term cART may affect gait speed decline. This is an important question, as nearly all of the aging/HIV-infected individuals studied were receiving cART, and the majority were virologically suppressed—characteristics that are likely to be similar in most populations aging with HIV infection.

Until recently, the older HIV-infected population with comorbid conditions has been largely ignored in intervention research because patients with significant comorbid disease have frequently been excluded. Clinical practice guidelines are more often single-disease focused, usually based on results from clinical trials without significant comorbid disease [85]. Consequently, many questions remain [78] regarding the clinical care of older HIV-infected patients, tolerability of ART in patients with multiple comorbidities, pharmacodynamics and pharmacokinetics in older patients, and interactions between ART and co-administered drugs targeting comorbidities. HIV comorbidities might impact the disease process since these comorbidities may interfere with the receipt and timing of ART initiation, disrupt ART metabolism, or require drug treatment that may interact with ART, complicating HIV/AIDS disease treatment and survivability [86, 87]. While HANA conditions are common in HIV-related illness, in general, there are limited data to determine if the comorbidities (especially neuropsychiatric) are the result of the virus acting in the CNS, magnified by cART treatment, or comorbid with other chronic, inflammatory illness. The evidence favors an “accelerated aging” phenotype (i.e., organ- and disease-specific), but the question of whether the HANA conditions are accelerated or accentuated has not been answered [45, 68]. Thus, more research is needed addressing these possibilities to determine appropriate intervention targets.

Behavioral interventions that target comorbidities in the context of HIV have been limited to only one research group, employing tele-therapy or coping improvement and interpersonal support randomized controlled trial to relieve HIV-associated depressive symptomatology [88‐90]. This group has further analyzed and characterized tele-therapy treatment efficacy for older HIV-infected persons in several ways: moderating roles of gender [91], sexual identity [92], and baseline depressive symptoms [93]; longitudinal sexual behavior patterns [94]; and trajectories of depression symptom relief [95]. This type of moderator (and mediational) analytic research of phenotypically defined aging/HIVgroups suggests possible improvements to interventions through further study of participant characteristics and/or intervention components and conditions for successful intervention trials. However, the preceding studies await replication by others to address generalizability and ecological validity, paving the way to developing novel intervention targets. The efficacy of these psychosocial interventions might be strengthened by combining with biological interventions since chronic inflammation (through a variety of mechanisms) contributes to HANA comorbidities in the elderly [82, 96].

HAND confers an increased risk for early mortality, independent of medical predictors [97]. As people age with HIV infection, the synergistic effects of aging and HIV may translate into more vulnerability to developing cognitive deficits that impact everyday activities over time [83]. A common intervention strategy has been to develop interventions that target specific cognitive deficits involved in daily functional activities [98]. HIV cognitive aging intervention targets are also suggested by neurocognitive specificity, e.g., interaction effects between HIV infection and age on motor speed, orientation, registration, and recall [99]. Some neuroimaging studies also identified the interaction effects between HIV infection and aging in selective brain functions [100, 101].

Because cognitive functions begin to decline in middle age or even earlier [102], it will be important for future studies to initiate interventions during middle age, rather than rely upon interventions in—later age when the negative cognitive consequences of dementing pathologic changes are generally detectable. This is particularly important in HIV infection, where “older age” has been set by the CDC as 50 years of age or older and the majority of patients fall in the 50–59-year-old age range. Such declines may take a long time to present clinically, but it is important to recognize that sub-clinical changes like metabolic alterations, chronic inflammation, and oxidative stress may also lead to behavioral changes, which can include reduced physical activity, decreased nutritional status, and greater social isolation. Therefore, interventions in mid-life (or earlier) may help prevent cognitive problems in aging with HIV infection [103].

HIV behavioral interventions for cognitive aging have been limited and mostly focused on cognitive rehabilitation of the speed of information processing [104, 105]. The studies show proof of concept for this strategy, and although the intervention covered a short time span, findings are encouraging because of the specificity of this intervention for informational processing speed abilities (no benefits for executive function). Cognitive improvement by enhancing or re-establishing information processing speed represents a restorative approach that relies upon neuroplasticity to encourage more effective neural organization. Cognitive decline may be addressed by developing interventions aimed at improving cognitive reserve [104, 105]. Engaging in stimulating cognitive activities (reading, social engagement, employment) can promote neuroplasticity, which, over time, encourages better cognitive reserve and also facilitates better neurocognitive functioning in general. When cognitive function cannot be restored, alternative compensatory approaches may be useful which involve the utilization of external methods (e.g., mnemonics or spaced-retrieval) or means to work around a deficit with the goal of improving cognitive functioning by supporting intact cognitive processes. This might be particularly effective for HIV-associated cognitive aging deficits in view of evidence that compensatory methods are generally more easily generalized to daily activities than restorative ones [106]. Cognitive rehabilitation techniques, whether restorative or compensatory, are not based alone on issues related to cognitive decline but are also mediated by engagement in treatment, and what can be done to enhance patient adherence for cognitive rehabilitation therapies to obtain greater cognitive and functional benefits from the treatment itself.