Separate Bonferroni corrections for multiple comparisons were applied for each family of tests. For the 2 fluency tasks, alpha was set at <.025 (.05/2). For the 4 SSQ scores, alpha was set at <.013 (.05/4). For the pre-expectations and enjoyment scale, alpha was set at <.05 (.05/1). Assumptions of normality using the Shapiro-Wilk test found both oculomotor and disorientation subscale SSQ scores to significantly deviate from a normal distribution; therefore, the Wilcoxon signed-rank test was performed for all SSQ comparisons. Two-tailed paired samples t tests were used for the remaining comparisons. Means and standard deviations for all pre- and postmeasures are shown in Table 1. Two-tailed paired sample t tests revealed that semantic verbal fluency scores were significantly higher post-VR, which was the only significant difference (Table 1).

Neither pre- nor post-VR comparison between phonemic and semantic fluency reached statistical significance in a two-tailed paired samples t test (pre-VR: mean 0.40 (SD 5.10); 95% CI −2.42 to 3.22; t₁₄=0.30; P=.77. post-VR: mean −2.47 (SD 5.21); 95% CI −5.35 to 0.42; t₁₄=−1.83; P=.09).

Bivariate (Pearson) correlations between the AES scores with either phonemic or semantic verbal fluency taken at baseline were not significant (AES and phonemic: r=0.309; 95% CI −0.241 to 0.709; P=.26; AES and semantic: r=−0.104; 95% CI −0.585 to 0.432; P=.71).

Bivariate (Pearson) correlations between the AES scores taken at baseline and phonemic verbal fluency change scores postminus pre-VR were not significant (r=−0.195; 95% CI −0.643 to 0.352; P=.49. There was a strong positive relationship between the AES and semantic verbal fluency change scores postminus pre-VR (r=0.719; 95% CI 0.327 to 0.900; P=.003; Figure 1).

Correlations between the AES scores and either phonemic or semantic change scores were significant (P=.03). Statistical differences between correlations were calculated using the Steiger method [84]. Figure 2 shows a breakdown of the SSQ scores for each participant pre- and post-VR using SSQ score cutoffs [85]. Figure 3 shows a summary of the total and subscale SSQ scores. Figure 4 shows a breakdown of the pre-expectation/postenjoyment scores for each participant.

Most participants had some difficulty with the SUS questionnaire; question number 5 had to be repeated or further explained. Four participants did not attempt answering question number 5, and 2 participants could not complete any of the questions. Therefore, data from the SUS were not analyzed.

After the VR session, the participants were asked a set of 8 questions to gain feedback about their experiences. Although some of the questions could be answered with a yes or no response, participants were given the opportunity to elaborate on their answers, if required.

All the study participants found the experience enjoyable. When asked what they liked about the experience, 11 participants said they enjoyed reminiscing, 4 participants reported it was nice to do something different, 1 participant reported how realistic the experience was, and finally, 1 participant liked the colors and clarity of the images shown. Comments included “liked the accuracy, feeling of being there and sites that I wanted to see again,” “taken back to somewhere that I know I will never see again,” “better than looking at a computer,” and “I had a feeling of being there.”

In reference to acceptability, 1 participant reported that the HMD was heavy; however, this was rectified by adjusting the head strap. All remaining participants found everything acceptable. When asked if they would do it again, all responded yes, with 1 participant stating, “only for research purposes.”

There were instances where participants reported negative symptoms or side effects. This included 1 participant with discomfort around the cheekbone and forehead areas. Another participant reported dizziness and the feeling that they were “going to fall off the edge;” a further 2 participants reported the feeling that they were going to fall off the edge. One participant reported eyestrain, and another participant reported nausea 10 min after the VR session, which subsided within 10 min. Therefore, 6 out of 17 participants reported some type of negative symptoms or side effects.

Participants were asked to offer suggestions to improve their experience. One participant stated that they would have liked to use the controller to navigate in the virtual environment on their own; another participant suggested having some type of interaction; a third participant said it would have been good to “show the difference between then and now,” referring to the Wander app; and finally, one participant would have preferred to see the places that they had not been to before. All participants said they would recommend the VR experience to a friend.

Finally, the participants were asked what they thought about the questionnaires. A total of 12 participants reported that they were all acceptable; 1 participant did not like them without giving a reason; 1 participant could not remember enough to comment; 1 participant said there were too many questions; 1 participant thought there was not a lot of relevance; and 1 participant said that the pre-VR measures did “cause some tension.”

This study found that VR using HMDs to deliver reminiscence therapy is feasible and may be beneficial for aged care residents with apathy. Of the 20 participants who were requested to participate, 17 agreed, indicating a willingness of residents in aged care to participate in research using VR technology. Therefore, the number of available participants for this study was easily achieved. The 3 participants who declined to participate were not interested. It is not known how often exclusion criteria had to be applied, as participants were selected by senior staff at the residential aged care facility following the exclusion criteria. A challenge for future research in apathy is a possible reluctance of those with high levels of apathy to participate because of a lack of interest and motivation, similar to reduced involvement in therapeutic activities found in those with higher levels of apathy [86]. The response rates to measures used in this study were positive. The SUS was the only measure that caused difficulties in understanding the questions, with 2 participants not completing any of the questions. The SUS was chosen because of the small number of items (6) and quick completion time compared with other measures of presence [81] to reduce participant burden. It was validated in a younger sample of university-based students with a technical background, who had no difficulties in completing the questionnaire [81]. As per the qualitative feedback received by this study, older adults may have difficulties in understanding some questions of the SUS. All other measures were fully completed, and all participants adhered to the procedure of the research. It should be noted that the reliability values for measures reported in the literature may not generalize to this study because of sample differences, that is, older adults with neurological and other chronic diseases.

The time taken to create content varied depending on the participant’s previous experiences. In this study, the researcher prepared the content before the VR session. It was found that by using readily available apps, the content could easily be sourced for all participants for a VR exposure time of 20 min. This finding is of importance for this study, as individualized reminiscence therapy can be resource intensive [87]. The accessibility of content utilizing the apps in this study reduces the resources required for creating content and places fewer demands on staff if applied in aged care lifestyle activities.

In this study, the preparation of VR content took approximately 90 min for each participant. This time can be substantially reduced once a library of 360-degree travel or music videos has been created. In addition, in this study, the content was compiled before the VR session to maximize the amount of time the participant would reminisce. Sourcing of VR content could also be done during a reminiscence session, working together with the participant while they are viewing images in the HMD, and an area for future research. Working together with the participant not only reduces the preparation time but also allows sessions to be tailored and flexible depending on participants’ responses [88]. Traditional methods, including making up a life history book, can require substantial resources and are also limited [88]. Using the Wander and YouTube apps reduces the preparation time and provides a wide range of content that can be easily changed. The use of a group reminiscence therapy format that is more generic can be a preferred approach because of the ability to improve social interaction [89] and can also be less resource intensive [90]. However, each person has a unique history, and using individualized content has been demonstrated to be more effective than using generic content [91]. In addition, an individualized approach is person-centered, focusing on a person’s strengths, thereby helping maintain a sense of self and identity [92].

No technical problems with conducting the sessions were encountered. To enable successful delivery of the content, it is recommended that the VR headset be mirrored to an external screen and controlled by the person delivering the content. Thus, the focus is on viewing the content rather than trying to learn to navigate or use the controllers. Although 2 participants would have liked to interact in some form with the virtual environment, this study was designed to provide a seamless experience. As the VR exposure consisted of one session, the intention was to reduce any frustration that could occur because of pressing the wrong buttons on the controller. We cannot draw any conclusions about the extent to which aged care residents could navigate within the app and use the controllers themselves, as this was not attempted. However, the feasibility of controller usage and navigation in VR in this population is an avenue for future research for studies conducted over a longer period. Providing the ability to interact allows the user to act independently and may improve the sense of presence. Recent research has found that the use of VR in providing interaction may be an alternative way of delivering stimulation to people with dementia who do not participate in other lifestyle activities [93].

Verbal fluency was used as a proxy measure to assess improvements in apathy. Both semantic and phonemic verbal fluency results were similar at baseline, despite phonemic fluency being known to be more difficult [94]. This could be because of the level of education in this group of participants (mean 13 years, SD 3), as higher education has been correlated with increased performance in phonemic fluency [95]. In pre-VR sessions, both phonemic and semantic fluency scores were not associated with the participants’ level of apathy. In post-VR sessions, not only did semantic fluency increase but the results were positively correlated with participants’ level of apathy at baseline. This suggests that the greatest cognitive improvements were seen in participants with higher levels of apathy. What is unclear is why these differences were seen only in semantic fluency. Similar improvements in verbal fluency have been reported in a study using traditional reminiscence therapy in a sample of dementia patients [96] in a short number of sessions. However, in this study, increases were seen in both phonemic and semantic fluency, and apathy was not measured. The executive function of initiation required for verbal fluency may be low in persons with apathy [97], and the reliving of autobiographical memories in an immersive environment may have stimulated the process of initiation.

The results from the debriefing questionnaire were positive and in agreement with previous research, finding acceptance of using HMDs in older adults [46-48]. This study further expands the knowledge of acceptability by demonstrating that VR using HMDs can be used to deliver reminiscence therapy with participants of varying levels of cognitive impairment, including minimal, mild, and moderate impairment, as assessed by the PAS. All participants reported enjoying the VR experience, despite 6 participants reporting a side effect. Another indicator of positive response was that all participants stated they would do it again (although 1 said “for research purposes only”). The most common response about what the participants liked about the experience was related to reminiscence and the enjoyment of seeing places from their past, indicating that this is something this age group wants to experience in a VR environment. An important aspect of reminiscence is maintaining a person’s sense of identity [87]. As loss of identity can be a key feature of Alzheimer disease that can increase apathy [98], participating in tailored reminiscence therapy may help in maintaining a person’s sense of self and identity and therefore improve levels of apathy. The expectations questions asked pre-VR compared with the enjoyment level post-VR, although in a positive direction, failed to reach a significant level. There were 2 participants in pain because of pre-existing conditions during the VR session, which may have contributed to this result.

Although quantitative findings of short-term side effects from VR were not significant, the debriefing questionnaire raised some concerns. Six participants reported side effects, 4 of which were symptoms related to items measured by the SSQ, including nausea, eyestrain, and dizziness. These symptoms dissipated quickly. In one instance where a participant experienced nausea, it occurred 10 min after the VR experience with symptoms subsiding within 10 min. This after-effect highlights the importance of monitoring participants both during and after VR to avoid health and safety implications.

Moreover, 3 participants had the feeling of “falling off the edge” in VR. This symptom was not captured by the SSQ and did create some discomfort, with 2 of the participants requesting to be moved back to their original position, when they had not actually moved. This could not be associated with the type of content. One explanation could be that sometimes the camera shooting the VR content is elevated, and a high sense of presence may have given a realistic feeling of their position in the virtual environment. Another factor that may have contributed to this symptom is the common occurrence of balance problems in older adults with neurological diseases [99]. This symptom will need to be monitored in future studies when working with this population.

The importance of taking pre- and post-SSQ measures was demonstrated in this study, with some participants reporting symptoms before the commencement of VR. This was usually because of pain from existing conditions. It is also interesting to note that in 4 participants, the SSQ scores decreased post-VR; this included a participant with the highest pre-VR SSQ score. This reduction in the SSQ scores could be a similar effect seen in the successful use of VR for distraction-based therapies [44,45]. Despite all participants finding the experience enjoyable, it is important to consider that 6 out of 17 participants (35%) had some type of negative response when viewing the VR content and the importance of monitoring symptoms. When using cutoff scores for the SSQ (negligible symptoms: <5, minimal symptoms: 5-10, significant symptoms: 10-15, concerning symptoms: 15-20, and problematic symptoms: >20 [85]), it can be seen in Figure 2 that the majority of participants scored >5, with 4 participants experiencing problematic symptoms (>20). However, these cutoffs were created with military personnel using flight simulators, and how these cutoff scores relate to the general population and, in the case of this study, older adults is unknown.

This study aimed to assess feasibility; therefore, a control group was not included, and we could not compare results using a traditional reminiscence therapy approach. However, the aim was first to establish if HMDs could be used in this population for reminiscence therapy. Although alternate forms of both phonemic and semantic verbal fluency were used to reduce practice effects, the lack of a control group receiving no VR intervention warrants a caveat when interpreting the results. It must remain open to what degree the pre-post comparisons are the result of the VR exposure versus a pure retest effect. It is also not known how long the verbal fluency improvements would last and if other aspects of apathy were reduced. Anecdotally, staff members providing services to participants involved in the research reported positive mood changes during the research process, and there were also instances where positive changes in behavior were reported by family members of participants involved in the research.

As the self-rated version of the AES was used, it needs to be considered that people with apathy will normally underestimate their condition [4]. Therefore, true scores may be higher than reported. This study did not specifically recruit participants with apathy, and a challenge with future research is the difficulty of recruiting participants who do have apathy because of their lack of interest and motivation.

This study demonstrated that VR using HMDs can be used to deliver tailored content about a person’s past and that residents in aged care enjoyed this experience. The aged care site in this study was in an area of high socioeconomic status, and study participants had a relatively high education level. Higher levels of education have been associated with an increased willingness to participate in research [100]. The sample was also selected by the staff and therefore was not a random sample, possibly introducing a selection bias [101]. How these factors would generalize to a random sample with an average or lower education level is unknown and needs to be considered when determining generalizability as participation rates, adherence to sessions, and therefore feasibility may differ. Immediate effects on cognition for those with higher levels of apathy after 1 session is an interesting finding and highlights the importance of continuing this research. The VR experience can be implemented and incorporated into current lifestyle activities in aged care with the correct protocols in place. VR is being increasingly used in older adults in areas including assessment, cognitive training, and rehabilitation; therefore, it is critical to monitor side effects, particularly after the VR experience, to avoid any health and safety implications and assess whether the benefits of using HMDs outweigh the risk of side effects. This study provides initial evidence that it is feasible to use VR with HMDs for therapy to treat symptoms of apathy in older adults in residential aged care. Further research over multiple sessions using an active control group to compare with traditional reminiscence therapy will help establish the advantages of using VR with HMDs.