This study used a descriptive qualitative design to explore community-dwelling older adults’ experiences and perceptions of VR-assisted cognitive health education in relation to MCI screening. The qualitative component was initiated during the recruitment and health education phase of a broader community-based cluster-randomized study after the research team observed that some older adults were hesitant to participate in cognitive screening. It focused on participants’ accounts, meanings, and contextual factors and was not designed to test intervention effectiveness or estimate causal effects on screening uptake [32,33]. Reporting was guided by the COREQ (Consolidated Criteria for Reporting Qualitative Research) checklist [34] (Multimedia Appendix 1).

The broader community-based cluster-randomized study comprised 3 study arms: VR-based adaptive-difficulty cognitive training, VR-based fixed-difficulty cognitive training, and a control condition without additional cognitive training. The community was the unit of randomization, and the 9 participating communities were allocated to the 3 study arms using a random-number-table method. A total of 178 older adults who met the parent-trial eligibility criteria, including confirmation of MCI, were formally enrolled in the 3 randomized trial groups. The shared VR-assisted cognitive health education component was provided across the study conditions, whereas the groups differed in the subsequent cognitive training component. The present qualitative sample was recruited earlier, during the common recruitment and health education phase, and participation in the qualitative interviews did not require a confirmed MCI diagnosis or depend on subsequent randomized-group allocation.

The shared VR-assisted cognitive health education session lasted approximately 40 minutes and comprised 2 components. First, participants viewed a 10-minute Chinese-language immersive 360° video using a Meta Quest 3 head-mounted display. Viewing was individual and seated. The experience was observational rather than interactive: participants could explore the 360° visual environment through head movements but did not manipulate virtual objects. The video presented first-person scenarios intended to depict cognitive change, including a familiar living room transforming into a floating island and the intermittent appearance of relatives’ faces in a “memory maze.” A local person living with dementia and a family member also described experiences of early screening, result notification, diagnosis, and treatment. Chinese subtitles were provided. Immediately after the VR session, participants received a 30-minute face-to-face educational session delivered by a nurse, who provided additional information about MCI and dementia and explained cognitive screening and the screening process.

Semistructured interviews were conducted with 22 older adults who had completed the shared VR-assisted cognitive health education session. The qualitative study specifically examined participants’ experiences of this common educational component and was not designed to compare experiences or outcomes across the randomized cognitive training groups. The interviews explored participants’ prior experiences of cognitive health education, reactions to the VR format, understanding and concerns regarding MCI and dementia, and views about screening and follow-through. The interview guide (Table 1) was informed by the IMB model and covered information-related understanding and misconceptions, motivational factors such as fear and stigma, and behavioral skills such as confidence and service navigation.

Participants were recruited from 9 participating communities in Nanjing, Jiangsu Province, China, between August 23 and September 17, 2025, using purposive sampling supplemented by snowball sampling. Community staff posted standardized recruitment notices in community WeChat (Tencent Holdings Limited) group chats (WeChat is a widely used messaging and social media platform in China), including brief information about the study and the research team’s contact details. Interested older adults contacted the researchers directly and voluntarily. Among eligible individuals who expressed interest, the research team purposively selected participants with attention to variation in age and educational attainment to capture a range of perspectives. Snowball sampling was used as a supplementary strategy, whereby some participants learned about the study through older adults who had already participated. Eligibility criteria for the qualitative study differed from those of the parent trial because the qualitative study focused on experiences of cognitive health education and screening rather than cognitive-training outcomes among people with confirmed MCI. Eligibility criteria for the qualitative study were an age of 60-85 years, completion of the shared VR-assisted cognitive health education session, the ability to communicate sufficiently to participate in an in-depth interview, and willingness to discuss perceptions and experiences concerning MCI, dementia, and cognitive screening. A confirmed MCI diagnosis was not required. Individuals who had previously undergone MCI or other cognitive impairment screening at another institution and medical professionals were excluded. Participants provided written informed consent before the interview and received a small gift worth approximately CNY 10 (US $1.50) after participation.

Before each interview, the researcher collected participant characteristics including age, retirement status, marital status, educational level, family history of dementia, and prior MCI screening. The semistructured interview guide was developed in Mandarin Chinese and pilot-tested with 3 older adults, who were included in the final sample of 22. No substantive changes were made to the interview guide after pilot testing; therefore, the 3 pilot interviews were retained in the final analytic sample. During the pilot stage, interview notes were reviewed with the 3 pilot participants to confirm that the recorded content accurately reflected their intended meaning. Interviews were conducted in Mandarin by TY, a researcher trained in qualitative methods who was involved in volunteer recruitment and intervention delivery in the broader study. Before the qualitative interviews, TY had prior study-related contact with 7 of the 22 participants; she had no prior relationship with the remaining 15. All interviews were conducted face-to-face in private, quiet indoor rooms provided by the participating communities, with only the interviewer and participant present. Interviews lasted 39-152 minutes. The variation reflected differences in the depth and richness of participants’ responses; some participants provided particularly detailed and reflective accounts, leading to additional probing and clarification and consequently longer interviews. With participants’ consent, interviews were audio-recorded and transcribed verbatim in Mandarin within 24 hours after each interview. Field notes were recorded during and immediately after the interviews to document contextual observations, relevant nonverbal responses, and issues requiring clarification; these notes were considered alongside the transcripts during coding and interpretation. Thematic saturation was initially judged to have been reached after the 19th interview. Three additional interviews were conducted to confirm saturation, and no new themes emerged in interviews 20-22. Full verbatim transcripts from the 22 formal interviews were not returned to participants for checking.

Coding, theme development, and interpretation were conducted using the original Mandarin transcripts. All 4 researchers (TY, BY, JY, and YW) coded the full set of 22 transcripts using Microsoft Word and NVivo (version 15; Lumivero). Codes, subthemes, and themes were developed iteratively from the interview data. Coding decisions and emerging interpretations were compared regularly among the 4 researchers; disagreements regarding coding, code definitions, theme boundaries, or interpretation were discussed until consensus was reached. Related codes were grouped into candidate subthemes, which were reviewed against the coded extracts and original transcripts and refined until the research team agreed on the final thematic structure. After the subthemes had been developed from the data, they were mapped to the IMB model domains according to their predominant conceptual meaning. Thus, the IMB model served as a theoretical framework for organizing and interpreting the final thematic structure rather than as a rigid set of predetermined themes. English translations were prepared for the participant quotations selected for reporting after the qualitative analysis had been completed.

Patient and public involvement occurred at different stages of the broader research program and was separate from the 22 semistructured interviews reported in this study. First, formative qualitative work was conducted before this interview study to characterize community perceptions and stigma related to MCI and dementia; findings informed development of the VR-assisted cognitive health education content. During intervention development, community-dwelling older adults and community staff viewed the VR educational materials and provided feedback on the content and presentation. Their comments were considered by the research team when refining the materials before implementation. During recruitment for the present qualitative study, community health coordinators supported logistical organization and dissemination of recruitment information.

This study was approved by the Ethics Committee of the First Affiliated Hospital of Nanjing Medical University (number 2025-SR-015) and conducted in accordance with the Declaration of Helsinki [35]. Participants provided written informed consent before the semistructured interviews and received the stated compensation after participation.

Twenty-two community-dwelling older adults from participating communities participated in the interviews. Their mean age was 73.27 (SD 4.95) years, and 15 (68.18%) were women. Most participants had completed secondary education or higher (Table 2). The relationship between the qualitative interview sample and the formally enrolled parent cluster-randomized trial sample is summarized in Table S1 in Multimedia Appendix 2. The two samples were recruited at different stages and under different eligibility criteria and were not statistically compared.

Participants described varied prior understanding of MCI and dementia, concerns about screening, and experiences of the VR-assisted education session. Several participants perceived the VR content as making concepts easier to understand or less threatening, whereas others described information overload, distraction by delivery details, or persistent uncertainty. Across interviews, participants also emphasized practical barriers to acting on screening intentions, particularly uncertainty about where to go and how to obtain referral support. Sixteen subthemes were developed from the interview data and subsequently mapped to the 3 IMB domains (Figure 1).

This qualitative study explored how community-dwelling older adults experienced VR-assisted cognitive health education and how their accounts related to the IMB model. Participants described the immersive educational content as making some information about MCI and screening more concrete and understandable, and some reported feeling less fearful or embarrassed about cognitive screening after the session. However, participants also described persistent uncertainty about how to proceed with screening, including where to go, how to register, and whom to approach for assistance. Taken together, these accounts suggest potential information-related, motivation-related, and behavioral skill–related processes through which VR-assisted cognitive health education may support screening readiness, while also highlighting practical conditions that may prevent willingness from translating into action. These findings represent participants’ experiences and perceived processes and should not be interpreted as evidence that VR caused an increase in screening uptake.

Participants described substantial gaps in basic knowledge about MCI, commonly conflating MCI with dementia or mental illness, and equating screening with being diagnosed and socially labeled. These accounts are consistent with studies documenting limited dementia knowledge in diverse community populations [13,14,36] and qualitative work showing that biomedical terminology, shame, and negative social meanings can shape how people understand dementia [16]. Help-seeking research further suggests that early cognitive changes are often discounted as normal aging, left uncertain, or misattributed, which can delay recognition of the need for assessment [17]. In our study, misconceptions were not only about disease concepts but also about the purpose and consequences of screening, with some participants anticipating social harm (loss of reputation) or family disruption (worrying adult children), which reduced perceived benefit.

Compared with their prior experiences of conventional health education, participants often described the VR session as easier to understand because it provided concrete visual explanations and could be followed at an individual pace. This perceived benefit is consistent with emerging intervention evidence showing that VR-based dementia education can support dementia-related knowledge, health literacy, or behavioral intentions in some settings [26,27]. In this study, however, participants’ descriptions of the immersive presentation should be interpreted as perceived clarity rather than evidence that knowledge objectively improved.

At the same time, participants described limitations related to novelty, peripheral details, and dense information. These concerns are consistent with systematic review evidence emphasizing the need to consider usability and acceptability when applying immersive VR with older adults [25]. Participants’ accounts in this study also suggested a preference for information that is paced and layered rather than densely presented. A practical implication is therefore to keep VR education brief and focused on a small set of core messages: what MCI is and is not, why screening may be considered, what screening involves, and what the next steps are in local services.

In the motivation domain, participants’ accounts highlighted fear of being labeled, anticipated discrimination, and concerns about face and reputation. This pattern is consistent with prior research identifying stigma, fear, and social consequences of a cognitive label as barriers to screening, diagnostic assessment, or dementia-service use [11,15,37]. Some participants described the community-based cases in the VR session as relatable and less stigmatizing, suggesting that familiar narratives may make screening feel more acceptable. Because these are retrospective qualitative accounts, however, the data do not establish the magnitude or durability of any change in stigma.

Several participants also expressed low perceived self-efficacy and fatalism, including skepticism that lifestyle changes or early identification could help. Pessimism about treatment and low self-prioritization were also evident in some accounts. These themes are consistent with prior research showing that reluctance toward cognitive screening or follow-up can be shaped by perceived limited benefits, psychological burden, stigma, and concerns about the consequences of diagnosis for patients and families [11,13,37]. Some participants described MCI as feeling more manageable after the session, particularly when it was framed as a risk state for which early management may still be meaningful; this should be interpreted as a perceived shift in understanding rather than a measured intervention effect.

A contribution of this study is the detailed description of reconciliation with aging as an adaptive psychological stance. Many participants normalized cognitive decline as part of aging and described acceptance as a way to maintain calm and identity. Similar narratives appear in help-seeking research, where memory changes are often discounted as normal aging or their significance is left uncertain [17]. In our data, this stance was double-edged: it protected well-being but reduced intrinsic motivation to pursue screening. Participants’ accounts suggest that the educational content did not necessarily require directly challenging acceptance of aging. For some participants, framing screening in relation to self-care, maintaining independence, and reducing potential family burden appeared to provide an alternative way of considering screening. For nursing practice, motivational work may therefore benefit from culturally sensitive and values-based communication rather than emphasizing disease threat alone.

Women accounted for 15 of the 22 interview participants. Although several female participants discussed health education in relation to family roles and caregiving, the study was not designed to compare women and men, and the overrepresentation of women may reflect self-selection into health education or interview participation. These accounts therefore should not be interpreted as evidence of a sex or gender difference in interest in the intervention.

A practice-relevant pattern in the behavioral-skills domain was the gap participants described between considering screening worthwhile and knowing how to complete it. Participants frequently reported uncertainty about where to go, how to register, which department to choose, what screening would involve, and how to take the first step. In their accounts, this intention-action gap was closely tied to complex service navigation and low confidence in navigating hospital systems.

Consistent with research on primary care and community-based models of care, participants emphasized the value of a trusted and convenient referral route. Prior studies have highlighted the role of primary care in cognitive detection [21], the importance of coordination with specialist services in dementia diagnosis [38], and the value of integrated teams and navigation support for older adults [39,40]. In this sample, participants’ accounts suggest that education alone may be insufficient when there is no immediate, low-barrier next step; this is an implementation hypothesis that requires prospective evaluation.

For practice, the findings suggest a potential role for community nurses and primary care teams in linking education to action. Future VR-assisted education could be kept brief and structured, followed by facilitated discussion to check understanding and address concerns; multielement VR dementia education has previously combined immersive content with lectures and discussion [26]. It could also end with concrete service information: where screening is available, how to book, what to expect, and who can assist. Nurse-led navigation strategies such as appointment assistance, reminders, or accompaniment should be evaluated rather than assumed to improve uptake.

For education, nursing curricula and continuing professional development should strengthen skills in stigma-sensitive communication, motivational interviewing with older adults, and age-friendly digital health facilitation. Training should also cover how to recognize and respond to reconciliation-with-aging narratives in a way that respects values while supporting preventive action. For policy and service planning, investing in technology alone is unlikely to be sufficient. This emphasis is consistent with evidence supporting stronger coordination between primary care institutions and hospitals in China [20], navigation support for older adults using community-based primary care [39], and interdisciplinary team support in dementia primary care [40]. Future implementation strategies could evaluate standardized community-to-hospital referral processes, clear responsibility for follow-up, and protected staff time for navigation support. Program evaluation could consider not only knowledge-related outcomes but also screening completion and appropriate follow-up care.

A strength of this study is its theory-informed qualitative design conducted within a community intervention context, which allowed exploration of experiences and contextual factors that quantitative outcomes may not capture. Developing themes from the interview data and subsequently mapping the subthemes to the IMB model provided a structured way to distinguish information, motivation, and behavioral-skill issues while retaining patterns grounded in participants’ accounts.

Several limitations should be considered. The sample was drawn from communities in one province and included many participants with secondary education or higher, which may limit transferability to rural or lower-literacy populations who could face different barriers to VR use and screening access. Participants initially volunteered in response to community recruitment information and were then purposively selected with attention to age and educational variation; this process may still have favored older adults who were more willing to engage with cognitive health education or research. Women were overrepresented (15/22), and interview participation may therefore have been subject to self-selection; apparent gender-related patterns should be interpreted cautiously. Future research should prospectively evaluate integrated approaches that combine VR-assisted education with structured nurse-led clarification, referral, reminders, and follow-up and should assess equity, feasibility, and screening completion in underserved settings.

This qualitative study suggests that VR-assisted cognitive health education may be experienced as engaging and understandable by some community-dwelling older adults, while fear, stigma, uncertain self-appraisal, and practical service-navigation barriers can persist. Participants’ accounts particularly emphasized the value of trusted community-based guidance when considering what to do after receiving screening information. These findings support further evaluation of models that pair age-friendly digital education with clear referral and navigation support, but they do not establish that VR itself increases screening uptake.