INTRODUCTION
With the growth of the aging population, there is a concurrent increase in the prevalence of dementia [1]. This upward trend is notably amplified in South Korea due to its rapid population aging, which suggests a rise from 10.3% in 2020 to an anticipated 15.9% by 2050 among individuals aged 65 or older [2]. Given the lack of currently approved disease-modifying therapies for Alzheimer’s dementia (AD), the surge in the number of dementia patients has resulted in increasing socioeconomic burdens. In response, various non-pharmacological treatments (NPT) aimed at prevention, delaying onset, or slowing progression are being explored [3]. Among these, cognitive training and physical exercise have been considered beneficial in some patients with subjective cognitive decline (SCD), mild cognitive impairment (MCI) and very early stages of dementia [4]. In older adults with SCD, cognitive training interventions led to improvements in memory, attention, and executive function as well as physical exercise, particularly aerobic exercise, was associated with improvements in cognitive performance and a reduction in perceived cognitive decline in this population [5,6]. In the context of MCI and early stages of dementia, combined cognitive and physical interventions have demonstrated the potential to enhance global cognitive function, subsequently improving activities of daily living (ADLs) and mood in older adults with MCI or dementia [7]. A meta-analysis by Zhu et al. [8] highlighted the considerable benefits of combined cognitive and physical interventions on overall cognitive function in healthy older adults. Although cognitive-motor dual-task training that is performed simultaneously rather than sequentially has shown effectiveness in promoting cognitive and physical functions [9], successful execution of simultaneous training is often not practical. Moreover, most traditional interventions have required in-person guidance, entailing considerable time, cost, and resources. However, since the start of the coronavirus disease-2019 (COVID-19) pandemic, it exacerbated the situation, leaving dementia patients isolated at home, leading to a decline in cognitive, physical, or social function, and deteriorating dementia conditions [10]. Even beyond the COVID-19 pandemic, the requirement for remote interventions and remote care services remains prominent. In this context, home-based cognitive or physical training tools could emerge as attractive alternatives for intervention. Although there are individual cognitive training or physical exercise applications, dual-task methods enabling simultaneous cognitive and physical training for elderly people remain scarce yet. Moreover, they have limitations such as one-way content based on photos or videos, content complexity unsuited for elderly individuals, or the requirement for expensive motion recognition equipment. To address these gaps, we developed an elderly people-friendly, tablet-based, home-based dual-task service platform that enables concurrent cognitive and physical training with augmented reality (AR) and motion recognition. The aim of this study was to investigate the potential effectiveness and benefits of this novel home-based dual-task platform with AR, in enhancing both cognitive and physical abilities among elderly patients, particularly focusing on mild AD, MCI, and SCD.
METHODS
Participants
We consecutively enrolled a total of 57 patients in this 12-week observational study. The trial was registered with the Clinical Research Information Service (http://cris.nih.go.kr) with ID No. KCT0007765. The sponsor was not involved in the study design or operational procedures. Ethical approval was obtained from the institutional review boards at the participating institutions (IRB file no. PC22ONSI0137 at Eunpyeng St. Mary’s Hospital). Informed consent was obtained from all participants after a comprehensive explanation of the study’s objectives and protocols. Participants were also informed of their right to withdraw from the trial at any point during its duration.
Eligible participants were aged 60-85 years and diagnosed with SCD [11], MCI based on Petersen’s criteria [12], or probable AD based on the diagnostic NINCDS-ADRDA criteria [13]. Participants with SCD were those with self-perceived cognitive decline over time but normal cognitive performance and who had scores on the comprehensive neuropsychological tests that are -1.5 standard deviations (SD) or more from the mean [14]. An MCI diagnosis was characterized by 1) memory complaints by participants or caregivers; 2) objective memory impairment as evidenced by scores >-1.5 SD from the mean of age-appropriate norms on the delayed verbal-memory recall tests of either the Seoul Neuropsychological Screening Battery II [15] or the Literacy Independent Cognitive Assessment [16]; 3) normal general cognitive function as assessed by a score on the Korean Mini-Mental State Examination, 2nd edition (K-MMSE-2) [17] <-1.5 SD from the age- and education-adjusted normative means, and a clinical dementia rating (CDR) [18] score of 0.5; 4) preserved ADL; and 5) no dementia. For an AD diagnosis, mild cases (CDR score of 0.5 or 1) were additionally considered. We excluded participants who were unable to walk independently; those with aphasia or language problems; those with blindness or impaired visual acuity hindering a tablet use; those with hearing difficulties that would affect understanding tablet-based instructions; those who were uneducated or illiterate; and those with other neurological, psychological, or metabolic disorders that impacted cognitive function.
A home-based, AR dual-task platform
The home-based, AR dual-task platform, named the COGNIMO program, which is still unavailable for commercial use, was employed for this study. Developed by a multidisciplinary team consisting of two neurologists, a physiatrist, a psychologist, and a computer graphics team, the COGNIMO program aims to effectively target cognitive training and physical exercise domains. This collaboration was established in January 2021, focusing on designing engaging cognitive and physical exercises that are safe for elderly patients. The program regimen includes warming-up exercises, aerobic exercises, cognitive training tasks, strength and resistance exercise, and cool-down activities, referencing the previous study [19]. Careful consideration was given to selecting movements that were both effective and safe as well as easy to follow for elderly patients. Cognitive training tasks encompassed diverse domains such as attention, working memory, visuospatial perception and construction, episodic memory, frontal executive function, and language. For instance, the tasks of the program include matching color and shape pairs, identifying the identical card images, finding another one of four cards, selecting objects’ names, remembering a picture, drawing a clock, a Stroop Test, and Trail Making Tests (Figure 1). The program spans a total of 42 minutes. A detailed description of the program content is outlined in Table 1.
Upon obtaining informed written consent, participants were comprehensively acquainted with the procedure details and provided guidance on how to use the tablet and access the program. Baseline cognitive and physical assessments were conducted before participants took the personal tablet home. We strongly encouraged them to engage in the training program for a minimum of 3 days per week over the12-week period. Upon study completion, tablets were returned, and the same cognitive and physical assessments were re-evaluated as part of the 12-week follow-up. All data of the COGNIMO program were securely stored on an independent server managed by AI.Ble therapeutics, ensuring patient-deidentification for data privacy and confidentiality.
Outcome measures
The primary outcome measures encompassed cognitive parameters, specifically changes in scores for both the K-MMSE-2 and the Korean-Montreal Cognitive Assessment (K-MoCA) [20], between baseline and the 12th week. The secondary outcome measures included changes in scores for the Berg Balance Scale (BBS) [21], the Six-Minute Walking Test (6MWT) [22,23], the Five Times Sit-to-Stand test (5xSTS) [24], and the short form of the Geriatric Depression Scale (sGDepS) [25,26] between baseline and week 12. The BBS objectively assesses a patient’s ability (or inability) to safely maintain balance during a series of predetermined tasks. The BBS comprises 14 items requiring approximately 20 minutes to complete, and each item is rated on a 5-point ordinal scale ranging from 0 to 4, which quantifies functional levels from the lowest (0) to the highest (4). The 6MWT is a submaximal exercise evaluation of aerobic capacity and endurance [27], that measures the gait speed covered within six minutes. We counted the number of times that participants engaged in the program and considered it as ‘1 training session’ when performed once.
Statistical analysis
To determine program adherence, participants who completed at least one training session per week were considered compliant. Based on this criterion, participants were categorized into the active group (≥12 training sessions) and the control group (<12 sessions) during the 12-week study period. We employed chi-squared tests for categorical variables and Mann-Whitney U test for continuous variables to compare demographics and clinical characteristics between the two groups. Bivariate correlation analyses using Pearson’s correlation coefficient explored the relationship between training time and outcomes. Linear regression models were employed for outcome variables significantly correlated with training time. In addition, we performed subgroup analyses by diagnosis. Statistical analyses utilized the R software version 4.0.5 (R Foundation for Statistical Computing, Vienna, Austria; www.r-project.org). A p-value <0.05 was considered statistically significant in all two-sided tests.
RESULTS
Among the initial pool of 57 participants, 49 individuals successfully completed this study. Eight participants withdrew due to either withdrawal of consent or loss to follow-up. A comprehensive analysis comparing demographic and clinical variables was performed between the control group (n=19) and the active group (n=30), with detailed findings presented in Table 2. The active group had an average age of 73.8 years, while the control group had an average age of 75.9 years. Among the 49 participants, the clinical diagnosis distribution was as follows: 15 patients had AD, 19 had MCI, 15 had SCD. The distribution of these diagnoses across the two groups was not significantly different (p=0.307). The baseline characteristics were comparable between the active and control groups, with no statistically significant differences in the baseline scores of K-MMSE-2, K-MoCA, and sGDepS. However, baseline secondary outcomes assessing physical abilities such as BBB, 6MWT, and 5xSTS showed difference only in the BBS (Table 2). Concerning the changes in parameters, after the 12-week intervention period, the active group exhibited a notable improvement in K-MoCA scores (2.3±2.2), compared to the control group (0.2±2.6), indicating a statistically significant difference (p=0.008). In contrast, changes in K-MMSE-2 scores between the active group (0.3±2.3) and the control group (-0.1±1.9) were not statistically significant (p=0.506). Regarding changes of physical ability parameters, including BBB, 6MWT, and 5xSTS, no statistically significant differences were found between the two groups (Table 2).
Initial bivariate correlations demonstrated a positive association between the number of training sessions and changes in K-MoCA scores (r=0.31, p=0.038), while no significant correlations emerged between the number of training sessions and other parameters, including changes in K-MMSE-2 scores (r=-0.13, p=0.417). Regression analyses consistently revealed a positive association between changes of K-MoCA scores and the number of training sessions (B(SE)=0.04(0.02), β=0.32, t=2.28, f=5.19, R2=0.101, p=0.027), as depicted in Figure 2A. Upon further analysis by diagnosis, this positive association was clearly observed, particularly within the MCI group (Figure 2B). Considering subgroup analyses by group and diagnosis, apparent discrepancies not only in the changed scores of the K-MoCA but also in the K-MMSE-2 were observed among participants with MCI, distinguishing the control group from the active group (Figure 3).
DISCUSSION
This study introduced a novel home-based AR dual-task platform, targeting both cognitive and physical abilities in elderly patients with cognitive complaints. The positive impact observed in terms of improved K-MoCA scores suggested the potential benefit of the COGNIMO platform in enhancing cognitive function particularly in the MCI group. The positive correlation between the number of training sessions and changes in K-MoCA scores suggested a dose-response relationship, emphasizing that steady repetitive training with this platform is likely to be effective in improving cognitive function.
The NPTs have demonstrated applicability across a spectrum of cognitive impairment, encompassing dementia, MCI [28] and even cognitively unimpaired adults at risk for dementia [11]. In particular, the combination of physical exercise and cognitive training within an NPT framework shows promise in enhancing cognitive function. Our results aligned with those of meta-analyses reporting consistent benefits of combined cognitive and physical training for elderly individuals without cognitive impairment [8], patients with MCI [29-33], or those with dementia [7]. These studies highlighting the effectiveness of dual-task platforms have observed improvements across various cognitive domains, encompassing global cognition, memory, executive function, and functional status, regardless of the baseline cognitive status [34,35]. Furthermore, compared to single-task approaches, dual-task interventions necessitate greater cooperation and high concentration during tasks, leading to broader brain activation [36] and increased blood flow [37]. The positive effects of cognitive training on cognition in patients with MCI may be linked to enhanced neuroplasticity, potentially mediated by increased brain-derived neurotrophic factor (BDNF) levels or neuroimaging findings [38-41].
The absence of significant improvement in the physical parameters over the 12-week period could be attributed to several factors. The relatively short duration may have been insufficient to elicit substantial physical improvements. Generally, moderate-intensity aerobic exercise is recommended for at least 30 minutes, five days per week [42], however, aerobic exercise in one session of this program is 16 minutes, comparatively shorter than the recommended guidelines, which could potentially affect the outcomes. Additionally, the platform’s design prioritized user-friendliness for elderly populations, which might have inadvertently restricted the exercise intensity or exertion level required to elicit significant changes. There is also the possibility that the selected parameters were not optimally suited to detect subtle alterations in physical functions or lacked adequate sensitivity to detect such changes. However, given prior research suggesting that the implication of aerobic or physical exercise combined with cognitive training has the potential to enhance neuroplasticity in individuals with MCI [41,43], the absence of statistically significant improvements in physical parameters within this short-term trial is to be expected. Based on the results of this pilot study, it might be beneficial to extend the duration of aerobic exercise in future iterations of the program to align with current guidelines, thereby potentially enhancing both physical and cognitive outcomes.
Regarding the differences in the observed efficacy between groups, one plausible reason for the differential effects could be the varying baseline cognitive reserves and progression rates of the conditions [44,45]. Patients with MCI may have a greater capacity for cognitive improvement due to less severe baseline impairments compared to those with AD [44,45]. Conversely, individuals with SCD might have minimal objective cognitive impairments, making significant improvements harder to detect with the given outcome measures. This suggests that the intermediate level of cognitive impairment in MCI allows for more noticeable improvements, and the possible heightened neuroplasticity response to the dual-task intervention might be more pronounced in this group.
The disparity in K-MMSE-2 and K-MoCA score improvements is perhaps because the assessment of K-MoCA involves more intricate cognitive functions, including frontal executive functions. This complexity might render the K-MoCA more sensitive to subtle cognitive changes. Encouragingly, among MCI patients, the improvement in K-MMSE-2 scores in the active group compared to the control group raises the potential for K-MMSE-2 improvement.
The dropout and compliance rates observed in our study could be indicative of several factors. The reasons for noncompliance may include the novelty of the AR platform, potential technical difficulties, or a lack of immediate perceived benefits, which could affect motivation. Future studies should aim to provide more detailed guidance and support to participants to enhance compliance and explore ways to increase engagement.
Home-based, self-administered computerized cognitive training enables individuals to independently access cognitive exercises from their personal devices at any time. Research indicates that this approach yields comparable benefits to cognitive function as supervised in-person training sessions in healthy elderly individuals [46-48]. Previous studies employing home-based cognitive training have shown cognitive benefits in patients with MCI [49-52].
AR involves overlaying virtual images onto the real world using visual devices such as cameras, which enables functional recovery through interactive manipulation of the patient’s environment and monitoring and enhances engagement, motivation, and immersion [53]. AR interventions can be applied comprehensively to patients of various ages, diseases, domains, and clinical patterns [54], promoting motivation and potentially improving outcomes [55]. AR interventions can be particularly valuable for patients with limited mobility, allowing interventions beyond the clinical setting to homes [56]. The growing application of AR in cognitive interventions for MCI and dementia shows its potential for enhancing daily life or reducing inconveniences [57,58]. From this perspective, our home-based, dual-task AR platform was useful in investigating whether it is suitable for elderly individuals, specifically those with cognitive decline.
The strength of our study lies in its distinctive features. Firstly, our program is home-based, a characteristic that enhances adherence to the intervention. This is particularly pivotal given the high costs and limited infrastructure for NPT in Korea, where many elderly patients encounter barriers to accessing such interventions. The home-based cognitive training program aims to enhance accessibility for a broader population of elderly individuals, especially those with cognitive decline. This potentially provides a cost-effective solution for cognitive and physical training in the elderly individuals. Utilizing human-computer interaction, our program fosters self-directed learning, aligning with the growing importance of personalized and self-directed interventions. Computerized tasks, being more interactive and engaging, may contribute to increased motivation levels within the active control group [47]. Another strength lies in the integration of AR in our program, introducing an innovative dimension that stimulates interest and enhances concentration during training sessions. This combination of dual-task training with AR technology suggests the potential for heightened engagement and effectiveness in improving cognitive and physical abilities among elderly individuals.
This observational study has several limitations. First, small sample size and short duration of this study necessitate cautious interpretation of the findings. Second, the observational study design, which lacks randomization and control group, gives rise to the potentials for confounding variables and selection biases, potentially affecting internal validity. The single-arm design, coupled with the limited sample size may also restrict the generalizability of the findings. Third, the heterogeneity within the MCI group adds complexity to the interpretation of outcomes. Despite these limitations, the findings from this pilot study provide valuable insights into the prospective efficacy of an AR-integrated, home-based dual task platform in enhancing cognitive function in elderly individuals, particularly those diagnosed with MCI. These promising results encourage the platform’s positive impact, suggesting a hopeful path for cognitive training interventions. To validate the true potential and efficacy of cognitive training and to amplify the present findings, further research involving larger, more extensive, randomized controlled trials, with a greater emphasis on individuals with MCI should be undertaken.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
