Virtual reality interventions to reduce social isolation in older adults: a protocol for systematic review

Abstract

Introduction

Social isolation and loneliness are prevalent among older adults and associated with negative health outcomes. Virtual reality (VR) interventions have emerged as a potential approach to address this problem, but their effectiveness remains unclear. This systematic review aims to synthesise evidence on the effects of VR interventions on social isolation and loneliness in adults aged 60 years and older.

Methods and analysis

We will search PubMed, Web of Science, Embase, CINAHL, MEDLINE, The Cochrane Library, PsycINFO and Scopus from inception to February 2025 for randomised controlled trials, quasi-experimental studies and before-after studies that evaluate VR interventions compared with usual care, wait-list, no treatment or other active interventions in older adults. The primary outcomes will be measures of social isolation and loneliness assessed with validated scales. Secondary outcomes will include depression, quality of life, cognitive function, physical function and adverse events. Two reviewers will independently screen, select and extract data from studies. Risk of bias will be evaluated using the Cochrane Risk of Bias Tool 2 for randomised trials and ROBINS-I for non-randomised studies. If feasible, meta-analysis will be performed; otherwise, a narrative synthesis will be conducted. The quality of evidence will be assessed using GRADE.

Ethics and dissemination

Ethical approval is not required for this systematic review, as it will only include published data. The review findings will be disseminated through a peer-reviewed publication and conference presentations.

PROSPERO registration number

CRD42025637230.

STRENGTHS AND LIMITATIONS OF THIS STUDY

• Rigorous Cochrane-based methodology including pre-specified protocol, comprehensive search, independent study selection and data extraction, standardised risk of bias assessment (RoB 2 and ROBINS-I) and PRISMA reporting.

• Addresses a timely evidence gap on the effectiveness of virtual reality interventions for social isolation and loneliness in a rapidly ageing global population.

• Comprehensive inclusion criteria encompassing multiple study designs and validated outcome measures across diverse settings.

• Potential heterogeneity in specific VR technologies (hardware, software, content), intervention designs, comparators and outcomes may limit comparability, which will be explored through subgroup and sensitivity analyses.

• Focusing on quantitative effectiveness data means acceptability, feasibility and implementation issues require further qualitative and mixed-methods research to guide translation.

Introduction

Social isolation and loneliness are major public health concerns that disproportionately affect older adult populations. Social isolation refers to an objective lack of social contacts and relationships, while loneliness is the subjective feeling of being isolated or lacking companionship.¹ Both social isolation and loneliness are prevalent among adults aged 60 years and older, with estimates ranging from 10% to 50% depending on the population and measures used.^{2 3}

The health risks associated with social isolation and loneliness are well-established. Numerous studies have linked these conditions to adverse physical, mental and cognitive outcomes in older adults, including increased risk of cardiovascular disease, stroke, depression, cognitive decline, Alzheimer’s disease and premature mortality.4,8 The impact of social isolation and loneliness may be comparable to or greater than that of traditional risk factors like obesity, physical inactivity and smoking.⁹ Moreover, the negative effects can be mutually reinforcing, creating a downward spiral of isolation, loneliness and health problems.¹⁰

The challenge of addressing social isolation and loneliness in older adults is growing as the global population rapidly ages. The United Nations projects that the number of adults aged 60 and older will rise from 962 million in 2017 to 2.1 billion by 2050, comprising 22% of the world population.¹¹ As more people live into advanced old age, often with chronic illnesses, mobility limitations or diminished social networks, the risk of isolation and loneliness will likely increase.¹² The COVID-19 pandemic has further exacerbated these issues, as older adults are particularly vulnerable to the virus and have faced prolonged periods of physical distancing and social restrictions.^{13 14}

Given the significant prevalence and impacts of social isolation and loneliness, there is an urgent need for effective interventions to promote social connectedness and support among older adults. While traditional approaches have focused on in-person social activities and support groups, these may not be accessible or appealing to all older adults, particularly those with mobility issues, health concerns or limited community resources.¹⁵ Technology-based interventions have emerged as promising alternatives, offering new ways for older adults to connect with others and engage in meaningful activities from their own homes.¹⁶

One such technology is virtual reality (VR), which involves computer-generated simulations of three-dimensional environments that users can interact with using specialised electronic devices, such as head-mounted displays or motion tracking systems.¹⁷ By immersing users in realistic virtual environments and avatars, VR can create opportunities for social interaction, learning and enjoyment that may be lacking in real life. VR has been used for a variety of health applications, including pain management, physical rehabilitation, cognitive training and psychotherapy.^{18 19} In recent years, researchers have begun exploring the potential of VR for reducing social isolation and loneliness.

Several studies have investigated the feasibility and acceptability of VR interventions for socially isolated older adults. For example, Lin et al²⁰ developed a VR system for older adults to visit virtual destinations and engage in social interactions with family members and friends. The system was well-received by participants, who reported high levels of enjoyment, perceived usefulness and intention to use. Similarly, Mao et al²¹ piloted a VR programme in which older adults could explore virtual environments and participate in guided reminiscence and discussion groups. Participants found the programme engaging and felt it helped them connect with others and recall positive memories. Other studies have used VR to simulate family gatherings,²² travel experiences^{23 24} and cultural activities²⁵ for older adults.

Some studies have also examined the impacts of VR on older adults’ psychosocial well-being. Baez et al²⁶ conducted a pre-post study of a VR-based social activities programme for older adults in long-term care. After 12 weeks, participants showed significant improvements in loneliness, social support and depression compared with a control group. In another study, Appel et al²⁷ found that a VR-based reminiscence therapy programme reduced loneliness and improved quality of life among older adults with cognitive impairment. A case study by Chiu et al²⁸ reported that a VR visit programme helped reduce an older adult’s social isolation and depression following a stroke.

Despite these promising findings, the overall evidence bases for VR interventions to address social isolation and loneliness in older adults remain limited. Most studies to date have been small-scale feasibility or pilot trials, with varied intervention designs, outcome measures and populations.^{29 30} There is a need to systematically review the available evidence to determine the effectiveness of VR for this purpose, identify key intervention components and contextual factors and guide further research and development in this area.

Therefore, the objective of this systematic review is to synthesise and appraise the evidence on the effectiveness of VR interventions for reducing social isolation and loneliness among adults aged 60 years and older. Specifically, the review will address the following research questions:

1. What are the characteristics of VR interventions that have been used to target social isolation and/or loneliness in adults aged 60 years and older?

2. What is the effectiveness of VR interventions in reducing social isolation and/or loneliness in older adults, compared with control conditions?

3. Do the effects of VR interventions on social isolation and/or loneliness vary by participant characteristics, intervention features or study-level variables?

4. What are the effects of VR interventions on secondary outcomes of interest, such as depression, quality of life, cognitive function, physical function and adverse events?

By addressing these questions, the review will provide a comprehensive and up-to-date synthesis of the evidence to inform future research, policy and practice regarding the use of VR to enhance social connectedness and well-being among older adults. The findings could help identify effective VR intervention components and delivery approaches, prioritise research gaps and methodological issues and guide the development and implementation of VR programmes for socially isolated older adults in various settings. Ultimately, this could have important implications for promoting healthy ageing and social inclusion in the growing population of older people worldwide.

Methods

Review framework

This systematic review will be conducted following the framework provided by the Cochrane Handbook for Systematic Reviews of Interventions³¹ and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.³² The review protocol has been developed using the PRISMA-P checklist³³ and registered in the International Prospective Register of Systematic Reviews (PROSPERO: CRD42025637230). This protocol paper presents the methodology for a planned systematic review. No results are presented as data collection and analysis have not yet been conducted. The completed systematic review with results will be submitted for publication following completion of the review process.

To formulate the review questions and eligibility criteria, we used the PICO (Population, Intervention, Comparison, Outcome) framework,³⁴ as follows:

• Population: adults aged 60 years or older, living in any setting.

• Intervention: VR interventions aimed at reducing social isolation and/or loneliness.

• Comparison: usual care, wait-list control, no treatment or other active interventions.

• Outcomes:

  • Primary: measures of social isolation and/or loneliness.

  • Secondary: measures of depression, quality of life, cognitive function, physical function and adverse events.

Eligibility criteria

Types of studies

We will include randomised controlled trials (RCTs), cluster RCTs, quasi-experimental studies and before-after studies (with or without a control group) that evaluate the effectiveness of VR interventions for reducing social isolation and/or loneliness in older adults. Only studies with a clear pre- and post-intervention evaluation of social isolation and/or loneliness using a validated measure will be included.

We will exclude observational studies (eg, cross-sectional, case-control, cohort studies without a pre-post design), qualitative studies, case reports or series and reviews. We will also exclude conference abstracts, editorials, letters and opinion pieces.

Types of participants

We will include studies involving community-dwelling or institutionalised adults aged 60 years or older, with or without existing social isolation or loneliness. Studies that include some participants younger than 60 years will be included if the mean age of the sample is ≥60 years or if subgroup data are available for those aged ≥60 years.

We will exclude studies that focus exclusively on older adults with specific clinical conditions or diagnoses that may significantly impact social interaction or technology use (eg, severe dementia, acute stroke with communication deficits, advanced Parkinson’s disease with severe motor impairment, severe depression requiring hospitalisation or other conditions requiring specialised clinical management), as their unique clinical characteristics and specialised care needs may limit the generalisability of the findings to the broader population of older adults. By ‘clinical subpopulations’ we refer to samples recruited primarily based on a specific medical diagnosis or condition rather than general ageing or social isolation criteria.

Types of interventions

We will include studies that evaluate interventions using VR technology with the primary aim of reducing social isolation and/or loneliness among older adults. The VR component must be the core element of the intervention, rather than an adjunct to other activities. We will include both immersive VR (using head-mounted displays or cave systems) and non-immersive VR (using desktop displays or projectors) interventions.

Examples of eligible VR interventions may include, but are not limited to:

• Virtually visiting places of interest (eg, museums, parks, cities).

• Participating in virtual social activities or support groups.

• Interacting with avatars representing family members, friends or other participants.

• Engaging in virtual exercises, games or hobbies with others.

• Creating and sharing virtual experiences or content.

We will exclude interventions that use other technologies (eg, telephone or video calls, online chat rooms, social media) without a specific VR component. We will also exclude multifaceted interventions in which VR is not the main component and single-session VR exposures without a clear pre-post assessment.

There will be no restrictions on the setting, intensity, duration or delivery method of the VR interventions. Both individual and group-based interventions will be included.

Types of comparators

Eligible comparators will include:

• Usual care (ie, no specific intervention).

• Wait-list control (ie, offered the intervention after the study period).

• No treatment (ie, no activity).

• Other active interventions (eg, face-to-face social groups, befriending services, internet-based programmes).

We will include studies that compare a VR intervention to one or more of these comparators, as well as studies that compare different types of VR interventions to each other (ie, head-to-head trials).

Types of outcome measures

The primary outcomes of interest will be measures of social isolation and/or loneliness, assessed using validated scales or questionnaires. These may include, but are not limited to:

• Lubben Social Network Scale (LSNS).

• De Jong Gierveld Loneliness Scale.

• UCLA Loneliness Scale.

• Campaign to End Loneliness Measurement Tool.

• Duke Social Support Index (DSSI).

• Revised Social Connectedness Scale.

• Social Disconnectedness Scale.

• Hawthorne Friendship Scale.

We will include studies that report quantitative data on at least one of these measures at both baseline and post-intervention, regardless of the time points used. If a study uses multiple measures of social isolation or loneliness, we will prioritise the one that is most commonly reported across studies or has the strongest psychometric properties.

Secondary outcomes will include measures of:

• Depression (eg, Geriatric Depression Scale, Patient Health Questionnaire-9).

• Quality of life (eg, WHO Quality of Life Scale-BREF, 36-Item Short Form Survey).

• Cognitive function (eg, Mini-Mental State Examination, Montreal Cognitive Assessment).

• Physical function (eg, Short Physical Performance Battery, Timed Up and Go test).

• Adverse events (eg, motion sickness, falls, injuries).

We will include studies that report any of these secondary outcomes using validated instruments at baseline and follow-up. There will be no restrictions on the timing or number of outcome assessments.

Summary of inclusion and exclusion criteria

Inclusion criteria:

• RCTs, cluster RCTs, quasi-experimental studies, before-after studies.

• Participants aged ≥60 years, community-dwelling or institutionalised, with or without existing social isolation/loneliness.

• VR interventions primarily aimed at reducing social isolation and/or loneliness

• Immersive or non-immersive VR.

• Any setting, intensity, duration, delivery format.

• Comparators: usual care, wait-list, no treatment, other active interventions.

• Primary outcome: validated measures of social isolation and/or loneliness at baseline and post-intervention.

• Secondary outcomes: measures of depression, quality of life, cognitive function, physical function, adverse events.

Exclusion criteria:

• Observational studies, qualitative studies, case reports/series, reviews.

• Samples with only participants aged <60 years or clinical subpopulations.

• VR as an adjunct rather than primary intervention component.

• Multifaceted interventions with VR as a minor element.

• Single-session VR exposures without pre-post assessment.

• Non-validated outcome measures.

• No measure of social isolation or loneliness.

Search strategy

Electronic database search

We will systematically search PubMed, Web of Science, Embase, CINAHL, MEDLINE, the Cochrane Library, PsycINFO and Scopus from their inception through February 2025. Our search strategy will use a combination of keywords and controlled vocabulary terms (eg, MeSH) related to the main concepts of the review:

1. VR technology: VR, head-mounted display, HMD, computer-generated, avatar, virtual environment.

2. Social isolation and loneliness: social isolation, social withdrawal, social connectedness, social network, loneliness, lonely, solitude, aloneness.

3. Older adults: aged, elderly, older, senior, geriatric, 60 and over.

4. Study design: randomised controlled trial, RCT, quasi-experiment, pre-post, before-after.

An example search string for PubMed is:

((virtual reality[tiab] OR VR[tiab] OR head mounted display*[tiab] OR HMD[tiab] OR computer generated[tiab] OR avatar[tiab] OR virtual environment*[tiab]) AND (social isolation[tiab] OR social withdraw*[tiab] OR social connect*[tiab] OR social network*[tiab] OR lonel*[tiab] OR solitude[tiab] OR aloneness[tiab]) AND (aged[tiab] OR elderly[tiab] OR older[tiab] OR senior*[tiab] OR geriatric*[tiab] OR “60 and over”[tiab]) AND (randomized controlled trial[pt] OR randomized[tiab] OR randomized[tiab] OR RCT[tiab] OR quasi-experiment*[tiab] OR pre-post[tiab] OR before-after[tiab]))

We will adapt this search strategy for each database, using appropriate syntax and terminology. We will apply no limits on date, language or publication status. For studies published in languages other than English that cannot be assessed by the research team, we will first attempt to identify bilingual researchers or professional translators within our institutional networks. If translation resources are unavailable, we will exclude these studies but document them separately and discuss this as a limitation. We will report the number and languages of excluded studies to ensure transparency about potential selection bias.

Other search methods

To ensure a comprehensive search, we will supplement the electronic database search with the following methods:

• Manual screening of reference lists of included studies and relevant reviews.

• Citation tracking of included studies using Google Scholar.

• Hand searching of key journals (eg, Cyberpsychology, Behavior and Social Networking; Journal of Medical Internet Research; Gerontechnology).

• Searching trial registries (ClinicalTrials.gov, WHO ICTRP) for ongoing or unpublished studies.

• Contacting experts in the field for additional studies or data.

We will document all search methods and results in a PRISMA flow diagram³² (figure 1).

Figure 1. PRISMA flow diagram.

Study selection

Screening process

We will export all search results into Covidence systematic review software (www.covidence.org) for deduplication and screening. Two reviewers will independently screen the titles and abstracts of all unique records against the eligibility criteria. Records that clearly do not meet the criteria will be excluded at this stage. Any disagreements will be resolved through discussion or consultation with a third reviewer.

We will then retrieve the full texts of all records that potentially meet the eligibility criteria or lack sufficient information to judge eligibility. Two reviewers will independently assess the full texts against the criteria and record reasons for exclusion. Disagreements will be resolved through consensus or arbitration by a third reviewer.

Data extraction

For each included study, two reviewers will independently extract data using a structured form in Covidence. The data extraction form will be pilot tested on a sample of included studies and refined as needed.

The extracted data will encompass various aspects, including study characteristics such as authors, year of publication, country and funding source. It will also cover methodological details, including study design, setting, duration, sequence generation, allocation concealment and blinding. Information on participants will be collected, including sample size, age, gender, ethnicity, study setting and baseline levels of social isolation or loneliness. Additionally, data on interventions will be extracted, specifying the type of VR used, its content, duration, frequency, delivery method and any co-interventions. Comparators will be documented, including their type, duration and delivery method. Outcome measures will be assessed based on the tools used, timing of assessments, results and any reported adverse events. Lastly, risk of bias information will also be evaluated to ensure the reliability of the findings (Appendix 1 in online supplemental file).

Any disagreements in extracted data will be resolved through discussion or referral to a third reviewer. If necessary, we will contact study authors to request missing or unclear data.

Risk of bias assessment

Two reviewers will independently assess the risk of bias of each included study using appropriate tools. For RCTs, we will use the Cochrane Risk of Bias Tool 2 (RoB 2),³⁵ which addresses five domains:

1. Bias arising from the randomisation process.

2. Bias due to deviations from intended interventions.

3. Bias due to missing outcome data.

4. Bias in measurement of the outcome.

5. Bias in selection of the reported result.

For each domain, studies will be judged as low risk, some concerns or high risk of bias. An overall risk of bias judgement will be made based on the highest risk level across domains.

For non-randomised studies (quasi-experimental and before-after studies), we will use the Risk Of Bias In Non-randomised Studies of Interventions (ROBINS-I) tool.³⁶ This assesses bias across seven domains:

1. Bias due to confounding.

2. Bias in the selection of participants into the study.

3. Bias in classification of interventions.

4. Bias due to deviations from intended interventions.

5. Bias due to missing data.

6. Bias in measurement of outcomes.

7. Bias in selection of the reported result.

Each domain will be judged as low, moderate, serious or critical risk of bias, with an overall judgement based on the severity of bias across domains.

Any disagreements in risk of bias assessments will be resolved through discussion or consultation with a third reviewer. Results will be presented in risk of bias graphs and summary tables.

Data synthesis

If two or more studies are sufficiently comparable in terms of design, population, interventions and outcomes, we will conduct a quantitative synthesis (meta-analysis)³⁷ using Cochrane Review Manager software (RevMan 5.4).

Measures of treatment effect

For continuous outcomes (eg, loneliness scores), we will calculate the mean difference (MD) with 95% CI if all studies use the same measurement scale. If studies use different scales, we will calculate the standardised mean difference (SMD) with 95% CI. For dichotomous outcomes (eg, proportions with improved social connectedness), we will calculate the risk ratio (RR) with 95% CI.

Assessment of heterogeneity

We will assess statistical heterogeneity across studies using the χ² test and quantify it using the I² statistic. We will consider I² values>50% to represent substantial heterogeneity. We will also assess clinical and methodological heterogeneity by comparing study characteristics and risk of bias.

Data synthesis methods

If there is no substantial statistical or clinical heterogeneity, we will use a fixed-effect meta-analysis model to pool the results. If there is substantial heterogeneity that cannot be explained by subgroup analyses (see the Confidence in cumulative evidence section), we will use a random-effects model, which assumes that the true treatment effect varies across studies.

We will undertake separate meta-analyses for each outcome and comparison. If studies report multiple follow-up times, we will conduct separate analyses for short-term (≤3 months), medium-term (>3 to ≤6 months) and long-term (>6 months) effects.

Unit of analysis issues

If cluster RCTs are included, we will check whether appropriate analysis methods were used (ie, accounting for clustering in analysis). If not, we will attempt to adjust the results by estimating the effective sample size based on the design effect.³¹ For studies with multiple treatment groups, we will combine all relevant intervention groups into a single group and all relevant control groups into a single control group to create a single pair-wise comparison.³¹

Dealing with missing data

If studies have missing or incomplete outcome data, we will attempt to contact the authors to request the data. If data remain unavailable, we will report the available results and judge the risk of bias due to missing data. We will undertake sensitivity analyses to assess the impact of including studies with high levels of missing data.

Assessment of reporting biases

If at least 10 studies are included in a meta-analysis, we will assess the risk of publication bias using funnel plots and Egger’s test.³⁸ Asymmetry in the funnel plot may indicate publication bias or other small study effects.

Subgroup and sensitivity analyses

If sufficient data are available, we will conduct subgroup analyses to explore possible sources of heterogeneity and effect modifiers. Pre-planned subgroups will include:

• Participant characteristics (eg, age, gender, baseline social isolation or loneliness levels).

• Intervention features (eg, type of VR technology, content, delivery format, intensity, duration).

• Study characteristics (eg, design, setting, comparator, risk of bias).

We will use random-effects meta-regression to test for subgroup differences if there are at least 10 studies per subgroup.

To test the robustness of the review findings, we will perform sensitivity analyses by repeating the meta-analyses under the following conditions:

• Excluding studies at high risk of bias.

• Excluding studies with missing data.

• Using different statistical methods (eg, fixed-effect vs random-effects models).

Presentation of results

We will present the main results of the review in a ‘Summary of Findings’ table, including a summary of the amount of evidence, magnitude of effects and quality of evidence for each outcome.³⁹ We will also present forest plots, risk of bias summaries and GRADE evidence profiles where appropriate.

If a meta-analysis is not feasible due to insufficient comparable data, we will provide a systematic narrative synthesis, with studies grouped by intervention type, outcome and/or study design. We will summarise the characteristics, methods and findings of each relevant study in tables and describe patterns across studies.

Confidence in cumulative evidence

We will assess the overall quality of evidence for each main outcome using the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) approach.⁴⁰ This involves consideration of within-study risk of bias, directness of evidence, heterogeneity, precision of effect estimates and risk of publication bias. Based on these assessments, we will rate the quality of evidence for each outcome as high, moderate, low or very low. Two reviewers will independently assess each domain for each outcome and resolve disagreements by consensus.

Ethics and dissemination

Ethics approval

Ethical approval is not required for this systematic review as it involves analysis of previously published studies and does not involve primary data collection from human participants.

Dissemination plan

The findings of this systematic review will be disseminated through multiple channels to ensure broad reach to relevant stakeholders. We plan to submit the completed review to a peer-reviewed journal focused on ageing, geriatrics or digital health interventions. Results will also be presented at relevant academic conferences, including gerontology and digital health meetings. Additionally, we will prepare a plain language summary for policy makers and practitioners working with older adults. All materials will be made available through institutional repositories and research networks to maximise accessibility and impact.

Discussion

This systematic review will provide a comprehensive synthesis of the evidence on the effectiveness of VR interventions for reducing social isolation and loneliness in older adults. By systematically identifying, appraising and analysing all relevant studies, the review will address a key gap in the literature and inform future research and practice in this rapidly evolving field.

Strengths and limitations

Key strengths of this review include its pre-specified design based on rigorous Cochrane methodology, transparent reporting per PRISMA guidelines and comprehensive search across multiple databases and grey literature sources. The involvement of two independent reviewers in study selection, data extraction and quality assessment will help minimise bias and errors. Additionally, the use of established frameworks (PICO and GRADE) will enhance the validity and interpretability of the review findings.

However, some limitations should be noted. First, the review’s scope is limited to quantitative intervention studies with a validated measure of social isolation or loneliness. While this is necessary to assess effectiveness, it may exclude other types of relevant evidence, such as qualitative studies, case reports, or surveys of user experiences and perceptions. Second, the included studies may be heterogeneous in terms of VR technologies, intervention designs, comparators, outcomes and participant characteristics. This variability could limit the comparability of results and preclude meta-analysis for some outcomes. Nonetheless, the pre-planned subgroup and sensitivity analyses will help explore potential effect modifiers.

Third, the rapid evolution of VR hardware, software and applications means that the review findings may have limited generalisability to future systems and programmes. The capabilities and costs of VR are constantly changing, so what is feasible now may be quite different in a few years. Periodic updates of this review will be necessary to keep pace with the technology landscape.

Finally, the review will not directly address issues of cost-effectiveness, implementation or acceptability. While some included studies may report related data, a full economic evaluation or process evaluation is beyond the review’s scope. Further primary research will be needed to examine the affordability, feasibility and user experience of VR interventions in different contexts.

Implications for research and practice

Despite these limitations, this review will have important implications for advancing research and practice in the use of VR to combat social isolation and loneliness among older adults. By synthesising the available evidence, the review will help identify promising intervention approaches, key knowledge gaps and priorities for future research.

For researchers, the review will highlight methodological issues to be addressed in future studies, such as improving reporting standards, minimising bias and standardising outcome measures. It will also identify areas where more evidence is needed, such as comparing different VR technologies, testing long-term effects and exploring implementation factors.

For clinicians and policymakers, the review will provide a summary of the potential benefits, harms and uncertainties of VR interventions for socially isolated older adults in different settings. This will help inform decisions about whether and how to incorporate VR into existing programmes and services, and guide the development of evidence-based guidelines and policies.

Ultimately, by advancing the evidence base on VR for social isolation and loneliness, this review could help drive the development and adoption of innovative solutions to enhance social connectedness and well-being among the growing population of older adults worldwide. As the technology continues to evolve and become more accessible, it is critical to ensure that its potential benefits are realised through rigorous research and responsible implementation.