Introduction

Cognitive impairment (CI) is a prominent issue facing people living with HIV, which occurs more frequently and at younger ages compared to those who are HIV-negative [1]. While severe dementia associated with HIV is uncommon due to the effectiveness of combination antiretroviral therapy (cART), milder forms of cognitive impairment are now more prevalent [2]. Approximately 14% of people living with HIV over 50 years of age have been shown to experience objective CI [3], and older age is identified as a risk factor in both HIV and non-HIV populations [4, 5]. This suggests that the prevalence of CI may rise as the HIV population ages [6].

CI in the context of HIV is commonly referred to as HIV-associated neurocognitive disorder (HAND) [7]. However, over the past decade, there has been growing recognition that CIs in people living with HIV are often multifactorial [8]. For many, these impairments are not solely due to brain injury directly caused by HIV, suggesting that the term "HIV-associated" might not fully encompass the range of factors involved. The International HIV-Cognition Working Group has advocated for distinguishing between brain injuries directly associated with HIV and other causes of brain injury in individuals living with HIV, shifting the focus from strictly neuropsychological testing to considering the broader clinical context [9].

Indeed, while the underlying pathogenic mechanisms involved in the development of CI remain under investigation, aetiology, in the era of effective cART, likely involves a combination of irreversible damage to the central nervous system (CNS) from HIV replication prior to cART initiation, neuroinflammation, ongoing HIV replication in the CNS compartment despite systemic control and the presence of comorbid conditions, lifestyle factors and ageing which have cumulative impacts on cognitive function [8].

Identifying the underlying causes of cognitive decline in people living with HIV is crucial for effective management: it is essential to control HIV infection comprehensively, both in the peripheral and central nervous systems, and to consider the potential for HIV to persist in the cerebrospinal fluid (CSF), termed as CSF escape, which should be investigated and treated if present. Reviewing, modifying, and optimising cART is necessary to address issues such as drug resistance, antiretroviral neurotoxicity, CNS penetration, and treatment adherence. However, when individuals experience cognitive decline despite effective HIV treatment, the causes are often multifactorial. Thus, a holistic approach to management is typically required to address and manage comorbidities and their associated risk factors (i.e. cardiovascular risk factors like smoking, hypertension, diabetes, dyslipidaemia, and obesity to maintain brain health; ageing people living with HIV often face multimorbidity and polypharmacy, increasing the risks of frailty, falls, drug interactions etc.). This approach further emphasizes the importance of lifestyle, social, mental health and educational factors when investigating low cognitive performance [9, 10].

Importantly, the CI seen in the majority of people living with HIV appear to be stable [11, 12], however improvement in cognition has proved difficult to achieve. No specific pharmaceutical treatments are currently available for HIV-associated brain injury and numerous small clinical trials of various compounds have not shown a clear beneficial effect on neurocognitive function [13]. While, optimising the management of comorbidities is an essential requisite for cognitive stability and potential improvement, alternative non-pharmaceutical approaches warrant greater consideration. Indeed, in populations of people without HIV with mild CI, cognitive rehabilitation programmes have been found to improve various cognitive domains including verbal, visual and working memory [14] through posited improved resilience to neurodegeneration [15] and enhanced neuroplasticity [16].

Alternative approaches that target broader indicators of wellbeing, such as quality of life are also important to patients [17]. Research indicates there is a direct link between changes in cognition and in health-related quality of life (HRQL), because CI can be a significant source of anxiety and stress, and perpetuates social isolation [18]. Indeed, people living with HIV and CI report particularly low HRQL: lower than people living with HIV without cognitive issues and lower than people without HIV attending generic memory clinics with mild to moderate dementia [10, 19]. People living with HIV with cognitive issues have reported the key factors driving HRQL to be difficulties performing activities of daily living, high levels of social isolation and low levels of self-efficacy to cope with CI [17].

There is value in examining the literature to identify the types of non-pharmaceutical interventions being used along with the evidence pertaining to each intervention type. Despite some HIV guidelines directives advocating for rehabilitation strategies or services to be offered [20], detailed recommendations do not exist. While an interesting review paper by Chan et al. [21] has described both pharmaceutical and non-pharmaceutical strategies for the management of CI in people living with HIV, to the best of our knowledge no systematically synthesised review exists which specifically examines this topic exists. Mapping or scoping what is known about non-pharmaceutical interventions for people living with HIV with CI will assist in identifying gaps in knowledge and stimulate further consideration and development of these alternate management strategies. Here we describe the extent, range and nature of the body of knowledge on non-pharmaceutical interventions for people living with HIV who experience CI, for the purpose of providing a systematic, synthesised summary of the evidence. With this in mind, this review aims to:

1. Identify the intervention types developed and outcomes assessed by studies examining non-pharmaceutical interventions
2. Provide a picture of the evidence related to specific intervention types and describe the acceptability of interventions to people living with HIV with CI

Materials and methods

A systematic scoping review methodology was chosen given the paucity of research surrounding non-pharmaceutical interventions specifically for this population. This was based on initial searches of Google Scholar and PubMed, which found that most studies looked at pharmaceutical management options only for people living with HIV. Indeed, studies looking at non-pharmaceutical interventions for CI were often examining these interventions in other patient populations, such as those with dementia or depression. Additionally, we found considerable heterogeneity in the type and design of non-pharmaceutical interventions described for people living with HIV with CI, along with a wide range of reporting outcomes, making a systematic review or meta-analysis methodology inappropriate.

This scoping review utilised the framework outlined by Arksey and O’Malley [22], and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews (PRISMA-ScR) [23] was used to inform the structure and design of the review [S1 Appendix].

Data sources

Five electronic databases were searched to identify the relevant studies: Medline, Embase, PubMed, PsycINFO and CIANHL. Databases were searched on 25^th November 2023, with date limits applied from 1^st January 2000 to the current day. This timeframe was chosen given the introduction and access of cART from approximately this timepoint onwards [24]. Screening of studies was first completed via analysis of the title and abstracts, with a second stage of screening completed via full-text analysis. The search terms used for Ovid Medline is described in Fig 1, and this strategy was used with some adaptation for the other four databases.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Search strategy used for Ovid Medline, and PRISMA flowchart describing the identification, screening and inclusion of studies in this scoping review.

https://doi.org/10.1371/journal.pone.0314185.g001

Search results.

Our search strategy yielded 2221 articles from the five databases, with a further 7 articles identified from hand-searching of reference lists. These search results were imported to RefWorks for duplicate screening, where a further 1037 articles were removed, leaving 1184 articles for title and abstract screening. This screening was performed by one author (LS) with additional support and clarification from KA, leading to 52 articles being deemed as suitable for full-text screening. From these articles, 14 were included in the final scoping review, as detailed in Fig 1. Thirty eight articles were excluded; studies were excluded if CI was not observed or objectively assessed in participants at baseline, or where participants had significant co-morbidities or substance misuse disorders. Studies were also excluded if outcomes were not reported separately for those with and without HIV completing an intervention.

Results

A total of 14 studies were included in this review, and the key characteristics for each study are reported in detail in Table 1. Ten studies were conducted in the United States [25, 28–30, 32, 34–38], one in Italy [26], one in Iran [27], one in Canada [31], and one in Nigeria [33]. There was a vast heterogeneity of study designs; studies were most commonly single-blinded randomised controlled trials (RCT) [27, 33–35], but also included three pilot RCTs [28, 32, 36], two non-randomised cohort studies [26, 31], one case-comparison study [30], one descriptive cross-sectional study [25], one pre-post experimental design study [37] and one longitudinal RCT [38]. One qualitative study looking at the preliminary experiences of those involved in an ongoing RCT of an exercise intervention was included [29]. Recruitment settings varied from outpatient clinics [26, 32, 37, 38], existing trial cohorts [29, 30, 31, 33], research centres [27], or a combination of settings [25, 28, 35, 36]. Byun et al. [25] looked specifically at participants who had completed the TOPS study, conducted by Vance et al. [37], and Hossain et al. [30] undertook a case-control study on the first three participants to complete each arm of the Think Fast study, performed by Vance et al. [38].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Study characteristics of the 14 studies included in this review.

https://doi.org/10.1371/journal.pone.0314185.t001

Across all 14 studies, the reported total number of participants randomised was 709, with 589 participants completing the intervention and follow-up periods across studies. The mean age of participants reported across studies was 52.2 years (SD = 5.53). One study [33] did not report demographic information, and four studies [29, 31, 32, 35] reported the mean age per subgroup and not as a total cohort; for these studies, the mean age of the primary intervention subgroup was used to calculate the mean age of all participants in this review, and are included in Table 1. The majority of participants across studies were male, and African-American or Caucasian ethnicities were the most commonly represented across all studies. The median intervention length was 12 weeks (IQR = 6.5), and this ranged from 2 to 24 weeks across studies. The exact details of the intervention duration, regularity and intensity are detailed in Table 1, and full demographic data is available in S1 Table. The median follow-up timepoint was 12 weeks (IQR = 14) following intervention initiation, but follow-up timepoints varied from 3 weeks to 2 years from baseline.

The diagnostic method used to identify CI at baseline, and determine the eligibility of participants, varied across all studies. Four studies used the Frascati criteria to detect CI; other studies used global clinical rating scores or clinical cut-off scores from each neurocognitive battery to detect impairment in cognitive function. Other tools used are detailed in Table 2. Thirteen studies included a control group, and the details of comparators are also included within Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Neurocognitive batteries used across studies in this scoping review, where relevant, with a description of the assessment tools used to assess specific cognitive domains.

https://doi.org/10.1371/journal.pone.0314185.t002

The majority of studies (n = 11) examined cognitive function as the primary outcome. A detailed description of the neurocognitive batteries used across studies, including the assessment tools used to assess different cognitive domains, are outlined in Table 2. Other outcomes examined by studies included QoL, depressive symptomatology, intervention acceptability and cART adherence; these are detailed in Table 3.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Further outcomes and assessment tools detailed in studies in this review.

https://doi.org/10.1371/journal.pone.0314185.t003

Discussion

As a greater proportion of people living with HIV reach older age, addressing CI in this cohort will become increasingly important. Beyond adjusting medication regimes and managing the impact of comorbidities to reduce adverse cognitive effects, there are no pharmaceutical interventions that directly improve cognition or ameliorate QoL for people living with HIV who experience CI. This indicates a role for non-pharmaceutical interventions, which aside from potentially improving neuropsychological outcomes may be effective at improving self-reported functional abilities and QoL. The objective of this scoping review was to describe the non-pharmaceutical interventions for CI in people living with HIV, and describe the current evidence relating to these varied strategies.

CCT appears to have high levels of satisfaction and acceptability in this cohort, although the evidence pertaining to neuropsychological improvement is somewhat mixed in people living with HIV with CI. Interestingly, a recent meta-analysis of CCT in those with mild cognitive impairment (without HIV) found improvements in verbal, visual and working memory [14] with similar findings reported in those with major depression [42], and in dementia [43]. CCT has many practical benefits including the ability for standardisation, precise monitoring of adherence and individual personalisation [44], possibly indicating ease of widespread implantation. Whilst CCT is often carried out with supervision from a trained professional, unsupervised CCT interventions could be administered in home settings to further facilitate access to treatment and reduce costs and resource usage [45]. Further, large scale RCTs are required to clarify the efficacy of CCT in those with HIV and CI and to ascertain whether findings documented in non-HIV populations can be replicated. Additionally, studies should consider incorporating a tDCS arm into research designs given tentative indications that it may boost the effects of CCT.

The review highlights the importance of identifying specific cognitive training targets to enhance cognitive benefits. The ACTIVE study showed that speed of processing (SOP) training in older adults not only improved cognitive function but also greatly reduced the risk of depressive symptoms for several years post intervention [46]. This is relevant for people living with HIV who often exhibit SOP deficits [38] and are far more likely to experience depression than the general population [47]. Moreover, the deterioration of SOP ability with ageing places greater burden upon other cognitive functions such as executive functioning [48]; as a result, any improvement in SOP derived from CCT interventions may span across multiple cognitive domains. While several studies focused on SOP as a therapeutic target, further research is needed to optimize these interventions for people living with HIV.

Exercise may be a key target area of non-pharmaceutical interventions for CI, given that higher levels of physical activity have been associated with less CI in people living with HIV [49–52], and that this review demonstrated beneficial impacts of exercise on QoL and medication adherence. Depression is known to be up to three times higher in people living with HIV compared with the general population [47], which is significant given the reduced depressive symptomatology observed following an exercise intervention in this review [33]. This highlights a role for exercise, as a tool to not only improve cognitive function, but also alleviate depressive symptoms and improve cardiovascular health. The positive effects of exercise interventions on cognitive function has already been demonstrated for various subtypes of dementia [53], however acceptability and long-term adherence plague these intervention types. Further studies which incorporate implementation methodologies are required to investigate the effects of exercise in this population and provide guidance to maximise impact and adherence longevity. Studies examining the use of probiotics and specific diets were limited, making conclusions and recommendation difficult, however, given research indicating the role of probiotics in neuroinflammation [54] and neuropathophysiological changes observed in Alzheimer’s disease in those on a ketogenic diet [55], these research avenues should be further explored.

Beyond the interventions currently considered in the literature, improving sleep quality may be an important target to prevent cognitive decline, given the association with poor cognitive function in people living with HIV [56]. This is especially important given that sleep disturbance is commonly seen in both people living with HIV and older adults [7]. Interestingly, Vance et al. [57] conducted a focus group study looking at Cognitive Prescriptions, where lifestyle factors are addressed by daily goals in the form of a prescription for people living with HIV. This study showed a high acceptability and willingness to engage with such an intervention, suggesting a further avenue of future research.

It is of interest, that for most interventions included in this review, the main outcome was improvement in neuropsychological test scores. This is somewhat in contrast to recent community-based approaches that have documented a shift in patient priorities and made calls to develop local and feasible interventions which largely focus on improvement of HRQL and daily life for those with HIV and cognitive difficulties [17]. Consideration of the outcomes important to patients is an essential consideration for those designing interventions. Studies should incorporate designs which ensure patient perspectives and priorities are embedded throughout the research process. Improvements in neuropsychological test outcomes which do not translate to a meaningful improvement in patients’ lives can be considered redundant from a person-centred perspective if they fail to address or improve the practical, everyday issues that matter most to patients, emphasising the need for a holistic approach that prioritises QoL and functional wellbeing. Developing interventions which assist individuals with activities of daily living and encourage social inclusion are important to facilitate independence and overall wellbeing. Indeed, in an era where people living with HIV are living longer it is important to also approach person-centred care from the perspective of maintaining or improving HRQL. Importantly, these considerations align with growing calls to include QoL as a ‘fourth 90’ in the updated UNAIDS 90-90-90 Testing and Treatment targets [58, 59]. A key question regarding future work in this area relates to whether studies should focus on those with cognitive symptoms, or those with measurable impairment on neuropsychological testing, who may not necessarily have manifested clinical symptoms as of yet. Given the complexity of neuropsychological testing, and the heterogeneity of cognitive batteries currently. used across studies, this may make application to the clinical setting more challenging. Despite this, clinical manifestation may be a late presentation of CI, and given the lack of consensus over how to quantify CI in the HIV population, using precise neuropsychological testing to detect CI may be preferable to ensure studies capture all the possible participants affected. Establishing a standardised cognitive battery for use in experimental studies investigating CI in people living with HIV could standardise research to facilitate comparison between studies and help identify which interventions are the most valuable. A standardised battery could also be implemented into clinical practice for detecting those with any CI.

The scoping design of this review enabled the inclusion of diverse sources of evidence and provide an overview of the current evidence describing the non-pharmaceutical management of CI in people living with HIV. This review followed the PRISMA Sc-R [23] and the JBI Manual for Evidence Synthesis [60], ensuring gold standards for the design of a scoping review were met. Whilst there is certainly value in mapping the current literature, a limitation of this review was the heterogeneity of study designs, outcome measures and cognitive batteries used, making comparison within intervention type difficult. Additionally, although we aimed to include studies where objective CI detected at baseline was not primarily caused by co-morbidities, some studies identified mental health disorders such as depression amongst some participants. Whilst these were reported as secondary illnesses and not the cause of CI, the presence of these co-morbidities may have impacted the intervention outcomes. Most participants across studies were male, leading to an underrepresentation of female participants in this cohort. Many studies had small sample sizes, which limits the reliability of findings; further work should investigate the effects of promising interventions in larger populations.

Future studies should ensure implementation outcomes are assessed to ensure potential interventions are feasible, acceptable and appropriate. Cost effectiveness analyses may also be beneficial to determine which interventions may be suitable for widespread implementation. In conclusion, non-pharmaceutical interventions may represent a promising area for improving CI in people living with HIV. Work must be done to ensure interventions are effective and specific for people living with HIV, and crucially, embedded into HIV care settings to facilitate uptake.

Supporting information