Virtual reality (VR) may appear to be a new phenomenon, with a surge of interest in the wake of the Kickstarter-funded release of two modern head-mounted displays (HMDs) Oculus DK1 and DK2 (Development Kit 1 and 2; McGill, 2016). However, academic attention to VR in the 1990s led to many of the theories that shape our understanding of the technology, including Milgram and Kishino’s virtuality-reality continuum (Milgram and Kishino, 1994). According to this continuum, VR is a manifestation of virtuality in which system users can view and interact with digital objects within a virtual environment (VE). Augmented virtuality (AV) occurs when real objects are represented in a virtual world. Augmented Reality (AR) falls closer to the reality end of the continuum, and consists of viewing virtual objects overlaid on a view of the real environment using pass-through devices (such as through a camera or specialized lenses; e.g., when someone on a video call turns on a “filter” and appears to have animated bunny ears added to their head). Mixed reality (MR) falls in the middle range of the continuum, allowing people to interact with both real and virtual objects together (e.g., exploring a virtual environment but using your actual smart phone while in it). In recent years, the many possible ways through which one can interact virtual systems has been summed up as extended reality (XR).

Prior to 2016, when Oculus released their first consumer VR device (Oculus Rift CV1), HMDs were expensive, saw attention in limited professional and academic settings, and had several performance and usability concerns (Hutchison, 2007). Because of the focus on developing usable and powerful devices, little research focused on users who do not fall within the innovation adopter categories of enthusiast or early adopters (Rogers, 2003).

Prior work highlights the potential of XR for diverse outcomes including increasing empathy (Bertrand et al., 2018; Barbot and Kaufman, 2020), training new skills (Johnson-Glenberg et al., 2014; Molka-Danielsen et al., 2018; Radianti et al., 2020), reducing loneliness (Gunkel et al., 2018; Smith and Neff, 2018), and improving mental health (Geraets et al., 2020), which are all of benefit to broad swaths of users. While research has begun to turn toward underserved populations, including older adults (Montana et al., 2020), more work is needed. Now that research surrounding XR is opening to include attention to the challenges of the “everyday user” (such as Gugenheimer et al., 2019, 2017), there is a need to uncover the gaps deserving of attention.

Reviews seeking to enumerate and close these gaps have focused on general use of VR in sports (Neumann et al., 2018), XR games for physical functioning among older adults (Miller et al., 2014; Molina et al., 2014). Beyond the physical uses of VR, recent reviews have focused on XR use in the support of persons with dementia (Hayhurst, 2018) and community-dwelling older adults (Dermody et al., 2020). Still other work has focused on the need for user-centered design including a framework of design recommendations for HMD-based VR exergames (Li et al., 2020) and the outcomes of a workshop on designing and using VR in sensitive settings (Waycott et al., 2018). The purpose of this systematic review is to guide future research on the use of XR by older adults, themselves, and in contexts beyond physical functioning.

For older adults, there exist concerns outside of the usability and transferability of previous findings in younger samples of XR users which intersect with the study of the prosocial benefits of the technology. In the United States of America, and indeed worldwide, falling birthrates combined with innovations in medical practices and technology are leading to the expansion of the older adult population (Ageing and health, 2018). While longer life is to be welcomed, it brings with it its own list of challenges, including planning for the costs of healthcare and combating isolation. One possible way to solve such problems is to leverage technology such as XR.

However, the focus of technology-centric research must expand to be able to offer broader offerings and accessibility to diverse users. Much of the development and design of innovative technology is conducted by younger adults, and it is common that developers and designers focus on creating technology that caters to people who are like them. This has been reflected in the lack of focus on accessibility in game consoles (until the Xbox Adaptive controller; Kujawski and Johnson, 2018) and XR HMDs (until the WalkinVR Driver; WalkinVR, 2020) and a reduced initial focus on narrow application areas. This naïveté coupled with the relative newness of XR has meant a focus on getting the technology to work for the “average” consumer of these new technology, as well as on making them more financially accessible (Lawton, 2006). While technical functionality is important to get right, this has meant the user experience and inclusive design needs have largely been neglected in HMDs.

Still, these technologies are being used by older adults who find useful applications for XR (for example, VR HMDs Baker et al., 2019; Roberts et al., 2019). Therefore, it is important that researchers work to enhance adoption of XR technology among older adult populations. This paper describes a systematic literature review that seeks to highlight extant research as well as areas that are lacking to in order to guide the continued advancement of scientific discovery in this area. Specifically, this review explores the current ways XR technologies are used with older adults, as well as how interaction is different as individuals age, summarizing the extant literature that centers on older adult use of XR in order to identify key themes, including benefits and barriers to XR use.

This review focuses on the use of XR technology by older adults; specifically of interest are the ways in which interaction is different as we age and how interaction with XR technology can impact psychosocial health and quality of life. However, these concepts deserve better clarity. As such, we delineate below what we mean by older adults, XR Technology – including a brief review of the history of XR and the interaction challenges inherent in the technology, and psychosocial health and quality of life.

Following the recommendations of PRISMA (Moher et al., 2009) and Cochrane (Higgins et al., 2019), the methods were set out prior to the initiation of the study. Keywords, inclusion criteria, exclusion criteria, and databases were identified and agreed upon by an interdisciplinary team of researchers prior to the start of the review. Web of Science, PSYCInfo, and Pubmed were searched on September 30, 2019. Searches were limited to articles on older adult humans published in the English language. The databases were searched using terms representing: age; psychosocial health; and XR (see Supplementary Appendix SA for a full list). Using these keywords, 2,018 citations were pulled from the databases and filtered to 1,513 citations after duplicate removal (see Figure 1).

Next, titles and abstracts were examined according to inclusion and exclusion criteria. We included published data that fulfilled the following inclusion criteria: 1) data were collected for samples with a minimum age of 60 years or older; 2) the psychosocial data were collected at the level of the individual; 3) data about the use of the XR were collected, rather than simply using the technology as a delivery mechanism; and 4) papers were published after 2013 because, for HMDs, the release of the Oculus Rift Development Kit 2 provided vast improvements of frames-per-second, weight, and resolution over previous HMDs.

Excluded were published data that: 1) took the form of an abstract-only conference proceedings, magazine article, or other non-peer-reviewed source (e.g., commentary, editorial report); or 2) were presented as a review or meta-analysis. Given that this research area is developing, theses and dissertations were included to reduce the publication bias that may have otherwise been present. All qualitative and quantitative studies meeting these criteria were included.

Next, abstracts and titles were reviewed by a team to determine whether papers fit the criteria defined above, resulting in 306 citations. The papers were downloaded and a second manual screening, which used the search feature of Mendeley to identify studies that did not meet inclusion criteria, reduced the corpus to 71 papers which used a virtual environment and included psychosocial or physical health measures. Given the intent to examine the use of 3D virtual environments, papers were not included if they did not feature an HMD or 3D visualizations (e.g., a 3D television was appropriate, but a projected image with no active or passive stereo lenses was not). Thus, the 71 papers were further reduced to 51. Through a thorough reading of each paper, 35 were determined to not meet inclusion criteria and were omitted from further review, while an additional paper 1) was added through citation chaining, bringing the final corpus, which is listed in Supplementary Appendix SB, to 17 papers.

The final corpus consisted of 17 papers that were published from 2015 to 2019. These publications included 19 studies which resulted from a dissertation including three studies (i.e., Almajid, 2019). Therefore, studies form the unit of analysis.

As a result of the range of outcomes evaluated and the relatively small number of studies, a meta-analysis was deemed unsuitable at this time. Instead, we describe the themes and commonalities and use these insights to offer suggestions for future research. First, the six identified focus areas of these studies are delineated. In addition, we identified eight technology types: four areas that had been studied in the XR literature at large and four that emerged during the literature review. The counts of occurrences of specific technologies in the corpus are also delineated below.

This paper presents research from the beginning of an important area of inquiry, with the hope that the insights will be useful to researchers and developers advancing XR technology. We have identified six focus areas of XR research in the extant literature: Assessment, Cognitive Training, Reminiscence, Driving Simulation, Physical Rehabilitation, and Socialization. Since we were interested in the use of XR by older adults specifically, information about use of XR by their caregivers was excluded from this analysis, although research does include this area – especially for training and informal caregiver support (Hayhurst, 2018).

One limitation of this study is the age-range defined in the inclusionary criteria. Many otherwise-relevant studies were excluded from our analysis because the youngest participants were in their 50s (Roberts et al., 2019), while we required at least one sample (to allow for age-comparison studies) to be aged 60 or older. A multi-disciplinary approach including researchers from gerontology is needed to define a clear range of age and abilities for studies with samples of “older adults” to allow better generalization within this population.

However, it is worth repeating that the group, “older adults,” should not be reduced to a single voice. As noted by participants in one study, older adults cannot speak to others’ experiences. This is a common mistake to be made, as members of minority groups and sensitive populations are often asked to play the role of “spokesperson” for the entire community as “token individuals.” To better avoid excluding voices or reducing their variability, we suggest a few converging paths: First, better measurement tools are needed to understand what components of individuals in the older adult population group are interesting to research; similar to the Big Five for understanding general trends in personality, a tool is needed to allow more fine-grained assessment of the aging process from a psychosocial and biological perspective. This is seemingly echoed in the finding that proficiency with navigating the VR platform in Brown (2019)’s study was reflective of individual experience with digital technology, rather than age. Second, more qualitative studies of adequate rigor are needed on the subject. Using these two approaches could lead to increased diversity in XR study sample populations both in 1) age and ability and 2) technology experience and proficiency.

Some of the studies presented methodological limitations, such as small quantitative or psychometric samples. One in particular noted that one of the two centers at which they conducted the study was not staffed by technologically-savvy individuals, stating that one center being evaluated had a Nintendo Wii^TM that had never been used because the staff “didn’t understand it” (Howes et al., 2019). The same study split their sample in two, with five participants using only a TV screen to complete guided exercise and the other four using both a TV screen and a 3DoF HMD. It is not clear if use of the HMD and TV was counterbalanced to account for order effects, nor is the split of participants in each condition between the two centers addressed in the paper.

Since the studies in the corpus do not center on specific outcomes, we have taken care to understand what has been studied and identify avenues for further research. Specifically, more work is needed to understand the specific usability and accessibility needs of the technology. A handful of studies commented on the need for better external awareness and reduced HMD weight, but relatively few user evaluative studies have been performed with this population of users. In need of evaluation are both the hardware considerations, such as the form factor of controllers and robust tracking, and software considerations like movement-smoothing and selection-interaction parameters.

Another concern raised by older adult participants was that of safety, both physical and emotional. Participants of these studies were hesitant to recommend the use of HMDs and immersive XR to those who may be emotionally or cognitively vulnerable, such as persons with dementia. Additionally, the usability concern for external awareness highlights the need for better support of safe use of XR HMDs. As noted by Yang et al. (2018), the physical play space may be used by other people or include movable objects that will pose safety challenges to the unaware HMD user.

One concern that was not directly raised in the samples of the reviewed studies is that of privacy. Whenever a person interacts with the Internet, they are exchanging data such as browsing and location history. In VR, and especially Social VR (Blackwell et al., 2019), privacy should be a concern – both understanding the transparency requirements of users and designing these systems and experiences with privacy in mind.

Overall, XR seems to have strong potential value as a technology for older adults. It has demonstrated efficacy in many areas including those explicitly tested with older adult users (Lee et al., 2017; Niki et al., 2019). In addition, older adults, themselves have largely preferred using VR over other methods, as discussed above. Lastly, there is no current indication that age alone affects cybersickness propensity, as only seven of the 17 studies in this corpus reported any indications of cybersickness.

Still, work is needed to verify these claims. In addition to creating more usable and accessible XR, attention to the prevalence of cybersickness is needed, and validated measures should be used to assess this. Relatively little work has focused on the psychosocial impacts of XR. While longitudinal studies have commented on the impacts on social health and quality of life, there have not been, to the authors’ knowledge, any that have addressed the topic of prolonged use of XR on psychosocial outcomes.

This review has presented information on the extant literature that centers on the use of XR by older adults. In total, 19 studies in a 5-years period met inclusion criteria and focused on six areas: Socialization; Physical Rehabilitation; Driving Simulation; Cognitive Training; Reminiscence; and Assessment. The contexts and outcomes of use, preferences for use, and reported occurrences of cybersickness, dropouts, and data exclusions are delineated, and conclusions are drawn from the data. We hope that this review will guide future research to address areas that are important to deployment of XR among older adult populations, while also examining the psychosocial impacts of the technology.