Introduction

Dementia is a degenerative syndrome that characterizes cognitive and functional capacity significant to perform activities of daily life (ADL) [1]. Exercise often provides some physical and psychosocial benefits, such as reducing loss of physical function and reducing pain and improving function and depressive mood, reducing agitation in older people with dementia [2] or MCI [3]. Although people with dementia may face unique challenges (e.g., changes in attention span, judgment, memory, communication), research has shown that they can successfully engage in regular exercise programs with appropriate support. For example, a study [4] showed that participation in exercise by people with dementia mediated the relationship between cognitive impairment and rehabilitative outcomes. People with cognitive impairment enjoy exercise and believe that it is beneficial for their health and cognition [5]. Exercise in this scoping review paper was defined as physical activity as any planned, structured, and repetitive bodily movement produced by skeletal muscle contraction [6, 7]. Group exercise in particular has been shown to have benefits for people with dementia. They tend to prefer group exercise [5], as it fulfils an important need for “building relationships” [8]. Further, group exercise has been seen as a helpful coping strategy, an effort to “keep the focus off the dementia” by having fun through physical activities with others [9].

Unfortunately, the COVID-19 pandemic caused disruption to community-based exercise programs across the world [10]. Various restrictions were put in place, including closure of important services and activities (e.g., community exercise and wellness programs). These restrictions negatively affected people with dementia and their family caregivers [11]. Even as many facets of society begin to open and lift restrictions, people with dementia and their caregivers continue to exercise precaution and are hesitant to meet in groups in person due to vulnerability to COVID-19 and associated complications [12]. People with dementia are more likely than people without cognitive impairment to experience social isolation and loneliness [13]. Social isolation has been associated with increased risk of depression, cardiovascular complications, and overall adverse effects to quality of life [14].

Technology can serve as a pivotal tool in enhancing access to healthcare for individuals with dementia living in the community, by reducing travel expenses and saving on transportation time for health professional visits [15]. Additionally, it offers innovative alternative exercise options, such as online yoga sessions and virtual reality exercise games, catering to the unique needs of this population [16, 17]. The usage of technology by older people has increased rapidly in recent years [18] During the COVID-19 pandemic and beyond, technology has become essential to maintaining the health of persons living with dementia. Telemedicine by video conference during the COVID-19 pandemic was associated with improved resilience and well-being in people with cognitive impairment [19]. Technology-based group exercise programs, or group exercise programs that utilize technology, including chair yoga (CY), benefit the physical well-being of people with dementia by improving mobility and function and support socialization through social engagement and networking, while simultaneously reducing social isolation and loneliness [16].

To date, no systematic review, meta-analysis or scoping review has been conducted on technology-based group interventions that include virtual reality, exergames, or other Internet-based interventions (synchronous or asynchronous) to support people with dementia. Little is known about whether technology-based exercise interventions provide persons with dementia effective means to partake in group exercise. If so, what types of technology-based exercise programs were offered, what worked, how and why, and what outcome measures and lessons were learned? This scoping review addresses this gap by focusing on various types of technology-based interventions across dementia and MCI and their effects. The format of a scoping review is often used for burgeoning fields with only small amounts of research published [20]. Scoping reviews attempt to map out volume, nature, and characteristics of a topic of research and do not assess quality of the evidence [20]. Since this field is relatively new, with little published research, a scoping review was considered an appropriate framework.

Methods

The protocol for this study can be found at https://doi.org/10.1136/bmjopen-2021-055990. The scoping review follows the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analysis Extension for Scoping Reviews) checklist as shown in the S1 Table [21].

Study selection and reviewing results

Identified articles were collated and uploaded into Mendeley, a bibliographic reference management tool, to ensure that references and articles were systematically managed. Once all identified articles were uploaded, duplicate articles were removed. The review process was a two-step process. First, two researchers independently performed a title and abstract review, screening for relevancy; third reviewer resolved any conflicts. Second, relevant articles underwent full text review with examination against inclusion criteria: (a) focusing on people with dementia or MCI, (b) setting at home or in ALFs, (c) included group physical exercise, and (d) used a form of technology to implement exercise. We included studies published in English, with no time limit on publication date. A range of study designs was included, including qualitative and quantitative designs, randomized controlled trials (RCT) and descriptive studies. Articles were excluded if they (a) included individual physical exercise, not group exercise, (b) were not intervention trials, (c) were not reported in English, or (d) consisted of a single case report or study protocol that did not report results.

The initial database search yielded 1,585 publications; no additional publications were identified through Google search. After removal of duplicates (n = 435), 1,150 articles were screened by title and abstract and 73 articles were identified as potentially relevant. At full text screening, 59 articles were excluded further due to the following reasons: wrong setting (hospital or research facilities; n = 2), Wrong intervention (no exercise or innovative technology involved; n = 17), wrong study design (case report or protocol; n = 2), exercise not group-based (n = 25), wrong participants (older people without dementia or MCI; n = 13). After discussion of eligibility of the articles with participants and family partners, the final review included 14 publications. See Fig 1 for the PRISMA flow diagram [25] that details the review process.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. PRISMA flow diagram for the scoping review.

https://doi.org/10.1371/journal.pone.0305266.g001

Findings

Table 1 outlines the characteristics of 14 eligible studies, with eight originating from the United States [16, 27–33], two from Australia [17, 34], and one each from Germany [35], Korea [36], Taiwan [37], and Brazil [38]. Among these, nine were quantitative studies (including three RCTs [31, 36, 37], five pre-post intervention studies[16, 27, 28, 33, 34], and one cross-sectional design [38]), three were qualitative studies [29, 32, 35], and two were mixed-method studies, one of which was a pre-post intervention design [30], and the other a randomized controlled crossover design [17]. Seven studies were pilot studies [16, 17, 27–30, 33]. One co-design study involved three stages: a pre-study, technology design, and investigation of interaction with the technical artifact [35]. Six studies were identified by researchers as feasibility trials [16, 17, 27, 31, 33, 34]. Analysis of these studies yielded themes including design, population and setting, interventions, outcomes measured, feasibility and results.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Charting table of scoping review.

https://doi.org/10.1371/journal.pone.0305266.t001

Design

The majority of the studies included in the scoping review were primarily aimed at assessing feasibility and gathering preliminary data. Consequently, more quantitative or mixed studies were too small to include a control group. Specifically, six studies utilized a quasi-experimental pre-post intervention one-group design [16, 27, 28, 30, 33, 34]. Studies that had control groups often used a type of exercise that did not involve technology as the control; this allowed them to assess the effect of the technology. Of the studies with control groups, traditional Tai Chi (TC) [37], stretching [31], a seated group mobility exercise[17], home exercise [36] and no exercise [38] were provided as controls. Four trials utilized an RCT design. One was a mixed-methods randomized controlled, crossover feasibility study design, randomized to either virtual cycling or a control group of seated group mobility exercise sessions [17]. ALL three other RCT studies were single-blinded, randomized, parallel-group trials [31, 36, 37]. In total, six of the studies were researcher-identified as feasibility trials, including a feasibility pilot study in which participants were assigned to exercise sessions via video conferencing in addition to one education session with a health coach [33]. One case study involved three phases: (a) a pre-study that consisted of an analysis of existing practices in the field, (b) a design of the technological artifact based on findings from Phase 1, and (c) an investigation of interaction with the artifact over time [35].

Population, sample size, setting

Overall, the studies had small sample sizes, which is consistent with the fact that most were feasibility studies. The 14 selected studies recruited a total sample of 379 participants (Table 2), with mean age ranging from 69 (SD = 4.21) [30] to 87.07 (SD = 3.92) [29]. Study sample sizes ranged from 10 [17] to 60 [34]. In terms of the ratio of sex, 4 of the 14 studies had fewer women than men in terms of residents living with dementia or MCI [16, 28, 32, 38]. Several studies did not report race/ethnicity of participants [17, 34–38]. Studies that did report ethnicity/race generally had a majority of White/Caucasian participants [16, 27–29, 31–33]. Seven studies focused on participants with MCI [17, 27, 28, 30, 31, 36, 37]. One study focused on ALF residents and included persons with dementia [29]. One study focused on older adults with cognitive impairment but did not require a formal dementia diagnosis, utilizing a Montreal Cognitive Assessment (MoCA) score to screen for cognitive impairment [16]. Three studies broadly included all types of dementia [32, 34, 38]; participants in the other study included persons with AD and their caregivers [33]. In terms of setting, the majority of the studies took place in domestic settings, such as private households, ALFs, or senior living facilities. Several studies took place solely in homes [16, 28, 30–34, 36–38], while other studies took place in homes and senior living facilities [27] or homes and daycare centers [35]. Several studies took place only in senior living facilities, specifically in ALFs [29] and residential care facilities [17].

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Demographic characteristics of participants.

https://doi.org/10.1371/journal.pone.0305266.t002

Discussion and implications

This scoping review examined the types of technology-based group exercise interventions applied at home or long-term care facilities. Overall, they were found feasible for people with dementia or MCI. Most studies had small sample sizes, with a majority of participants being White and female. This finding of ethnic/racial disparity is consistent with previous trends in research [47]. While the higher participation rate of women is in line with the overall higher incidence of dementia in females than males [48], it is important to have racially diverse groups [47, 49]. As small sample sizes are consistent with the purpose of pilot studies [50], larger-scale studies are needed to provide more robust evidence.

The studies included in the review were focused on older adult populations, with mean ages of 69 years to 87.07 years. Given the fact that there are physiological changes associated with aging, such as loss of muscle mass and strength and decreased physical performance [51], research comparing the effects of these technology-based exercise interventions between younger and older adults with dementia should be examined in the future studies. The majority of participants in these interventions resided in their homes in the community. Very few studies were conducted in long-term care facilities. Individuals with dementia residing in long-term care facilities are more likely to have more functional symptoms of dementia compared to those living at home. It is important for future research to investigate strategies to assist these individuals in engaging with technology-based group exercise programs. The impact of workforce shortage in long-term care facilities offers opportunities and limitations for technology-based group exercise programs.

There was little focus in the studies on the effect of the interventions on various types of dementia or the severity of dementia. None of the studies analyzed results by type of dementia. Different types of dementia can have distinct clinical presentations and varied responses to specific therapies due to underlying differential mechanisms. For example, while cholinesterase inhibitors are first-line pharmacologic interventions to treat mild Alzheimer’s dementia, Parkinson’s disease dementia, and Lewy body dementia, there is differing strength of evidence to support the effectiveness of this medication class for each type of dementia [52, 53]. Likewise, nonpharmacologic interventions such as exercise could have a differential effect depending on the underlying disease process. In people with vascular dementia, in which a core component of treatment is vascular risk factor management, exercise may play a relatively stronger role [54]. Stage and severity of disease have also been shown to influence treatment efficacy; These findings indicate the need for research with a racially and ethnically diverse population, with large sample sizes, and with participants stratified by type of dementia or cognitive impairment.

In the synthesis process, the strategy of identifying key themes and concepts was employed to categorize interventions into three overarching themes. Technology-based group exercises often incorporate "game elements," "immersive experiences," and "remotely supervised" features, all aimed at motivating participants to maintain adherence to exercise. Under this common aim, integrating three key characteristics might potentially maximize engagement effects. Anderson et al. [27] reported that improved executive function and verbal memory can only be observed at a “full dose” (20–40 minutes per session and ≥2 sessions per week) instead of an incomplete dose, emphasizing the importance of ensuring exercise adherence by introducing innovative technology into exercise routines to attract and inspire interest. D’Cunha et al. [17] and Choi et al. [36] utilized filmed videos of lakes and mountains and 100-inch projectors to enhance immersive effects of virtual reality, reported improved postural balance, muscle performance and cognitive function, and positive experience on older adults with MCI. One study utilizing Oculus Rift head-mounted display and Oculus touch controllers to create a virtual reality environment reported that VR exergames could be a promising technology for engaging people living with dementia [55]. Another study attempted to develop new VR exergames called “Seas the Day” through multistakeholder co-design, combining exergaming, virtual reality, and online assessment to fully engage persons living with dementia [56]. A systematic review reported that VR exergames provided potentially positive influences on cognition, memory, and depression in older adult populations [57]. Innovative technology, especially integrating these three types of interventions, needs to be assessed in future research concerning older adults with dementia or MCI. Besides, based on observations from qualitative and mixed-methods studies, participants have articulated their preferences for exercise, highlighting several key characteristics: enjoyable, user-friendly (especially in home environments like CY), appropriately challenging (avoiding monotony), immersive, of suitable duration (around 2 minutes), fast-paced, unpredictable, requiring minimal reliance on memory, socially interactive, and well-structured. These insights serve as valuable guidelines for the development of innovative technology in the field of exercise.

Comparison and contrast strategies were utilized to analyze various resources focusing on outcomes categorized by different functions. This approach involved systematically examinging the similarities and differences in the reported outcomes across the diverse resources. By categorizing outcomes based on distinct functions, the synthesis aimed to highlight commonalities and discrepancies among the findings, thereby facilitating a comprehensive understanding of the subject matter. All quantitative studies assessed various cognitive function including memory and executive function, physical function, and psychosocial function (e.g., depression, loneliness), and neuropsychiatric symptoms [17, 31, 36, 37]. Some studies showed significant improvement in outcomes, while others showed nonsignificant results. Significant improvements in psychosocial, cognitive, and physical functions are consistent with findings of previous research on in-person exercise for persons with dementia [58, 59]. Other studies did not find significant changes in certain measures of cognitive function, psychosocial function, or physical function [17, 28, 31]. In order to resolve the inconsistent findings, a rigorous design with a larger sample size is needed. Within the literature included in this review, relatively few studies utilized a blinded RCT design, while many were quasi-experimental or pilot studies. In addition to the paucity of studies, this indicates the relative novelty of the field. Positive results of these pilot studies support the need for further research, with RCT design, to be performed to increase strength of evidence and impact.

The reviewed studies encompassed a spectrum of outcome measures, comprising qualitative psychosocial, physical, and cognitive assessments, alongside biomarkers. Although biomarkers were measured as frequently as other types of outcome markers, the types of biomarkers assessed were not conventional. Salivary biomarkers and heart rate variability were measured, whereas more conventional biomarkers for dementia such as structural brain imaging (CT or MRI), functional imaging (PET) and fluid markers in cerebrospinal fluid were not measured [60]. Even though the biomarkers that were measured were unconventional, one significant advantage was that both saliva and heart rate variability were collected remotely. Measuring biomarkers is crucial for monitoring changes and the progression of dementia. The utilization of remote biomarkers has the potential to eliminate the necessity for individuals with dementia to travel for disease progression monitoring. While both biomarkers are relatively recent, with limited collection and scope, and necessitating further study, the emerging practice of remote biomarker collection enhances accessibility for individuals who might otherwise lack opportunities to participate in research and monitor dementia progression.

However, remote data collection could also have several barriers to implementation. This type of data collection would either rely on researchers traveling to the participant’s location, which could be limited by distance, time and funding, or it could utilize caregivers, which could be limited by caregiver ability to learn how to adequately collect data. Further research is needed to fully understand the utility and reliability of these remote biomarkers.

Quantitative outcome measures are crucial for research and medicine, yet qualitative outcomes, including quality of life improvements, are equally if not more vital to individuals with dementia and their caregivers. Using techniques such as interviews, people with dementia were given the opportunity to discuss how the intervention made them feel and how it impacted their lives. Quality of life was also measured empirically. While QOL-AD can be used to evaluate quality of life on people living with dementia, more specific measures for these populations exist, such as Alzheimer Disease-Related Quality of Life (ADRQL), Quality of Life in Late-Stage Dementia (QUALID), QUALIDEM (a dementia-specific QOL tool), and DEMQOL (health-related QOL for people with dementia) [61]. These measures that are specific for people with dementia could be more sensitive in their assessment of quality of life.

Results from this review indicate that adaptations for people with dementia or MCI were necessary for technology-based interventions to be successful. Guidance was significant, as participants with dementia did not always remember how the technological intervention worked [35]. The need for this adaptation is supported by previously recognized symptoms of dementia, such as memory loss and decline in executive function [62]. Proper safety and preparation for the exercise intervention were recognized as concerns by the researchers. Previous research has identified that persons with MCI or dementia have increased fall risk [63]; thus, additional safety preparation may be warranted. Studies in the review showed that a “one-size-fits-all” game approach did not work; the intervention had to be customizable to the level of the individual participant. Future interventions for individuals with dementia should involve caregiver training for assistance and safety. Clear guidelines are needed for intervention adaptation to meet the needs of specific groups.

Dementia is associated with decreased attention span and other behavioral and psychological symptoms [64]. It is important to note that playing a game that is either too difficult or too easy may cause frustration, agitation, or loss of attention. This is consistent with the individualization and adaptation commonly used for in-person exercise interventions for people with dementia [65]. Future studies should actively involve individuals with dementia and their caregivers in the co-design of technology-based group exercise programs. This inclusive approach ensures that the interventions are tailored to the specific needs and capabilities of the end users [23]. Including caregivers and individuals with dementia in the development process can also identify practical implementation strategies, fostering a sense of ownership and increasing the likelihood of successful adoption in long-term care settings. In the review, participants in one study reported satisfaction with contributing meaningfully to research [17].

Overall, results from the scoping review indicate that technology-based group exercise interventions are feasible for persons with dementia or MCI. These feasibility studies are novel; however, research with similar populations, such as older adults [66] and adults with Parkinson’s disease [67], support these findings. Adherence to a technology-based intervention (range of 70–79%) [17, 31] is comparable to adherence to in-person exercise interventions by persons with dementia or MCI (mean of 70%) [68]. Establishing feasibility is important, as there is growing demand in this population for telehealth, technology-based, and remote interventions [69]. Such interventions circumvent current and common issues in accessibility faced by persons with dementia or MCI and their caregivers. For example, need for transportation to site or lack of community resources are common accessibility issues. The convenience and feasibility suggested for further research into the effects and efficacy of such interventions.

Conclusion

This scoping review suggests that group exercise interventions that utilize technology are feasible and have positive psychosocial, physical and cognitive benefits to persons with cognitive impairment. The use of technology allows for both the intervention and data collection to be conducted remotely, which is essential for maintaining the benefits of group exercise while introducing the convenience of home-based exercise. As technology continues to advance, it is important to investigate inclusive strategies for relevant deveoplement and feasible and sustainable implementation of technology-based group exercise. Future work should explore strategies to involve individuals with dementia and their caregivers in the design and implementation of technology-based group exercise programs.

Supporting information

Acknowledgments

The authors acknowledge and thank Senior Medical Librarian Michelle Keba Knecht at Florida Atlantic University Charles E. Schmidt College of Medicine for her assistance. We thank the family and partners of participants for their work and contribution to this scoping review.