This paper is in the following e-collection/theme issue:

Cognitive Training for the Elderly (95) Serious Games for Health and Medicine (683) Dementia and Cognitive Decline (481) Cognitive and Neurorehabilitation (262) Methods and Feasibility Studies (379) Games for Cognitive Assessment (61) Serious Games for Health and Medicine (683) Games and Gamification for Health (650) Web-based and Mobile Health Interventions (3297) Games and Gamification for Health (650) Methods and Feasibility Studies (379) Games for Cognitive Assessment (61) Web-based and Mobile Health Interventions (3297) Dementia and Cognitive Decline (481) Cognitive and Neurorehabilitation (262)

Published on in Vol 6 (2023)

Preprints (earlier versions) of this paper are available at https://preprints.jmir.org/preprint/46177, first published .
Tablet-Based Puzzle Game Intervention for Cognitive Function and Well-Being in Healthy Adults: Pilot Feasibility Randomized Controlled Trial

Tablet-Based Puzzle Game Intervention for Cognitive Function and Well-Being in Healthy Adults: Pilot Feasibility Randomized Controlled Trial

Tablet-Based Puzzle Game Intervention for Cognitive Function and Well-Being in Healthy Adults: Pilot Feasibility Randomized Controlled Trial

Authors of this article:

Prabitha Urwyler1, 2, 3 Author Orcid Image ;   Rajnish Kumar Gupta1, 4 Author Orcid Image ;   Michael Falkner1, 2, 5 Author Orcid Image ;   Joel Niklaus1, 2 Author Orcid Image ;   René Martin Müri1, 3 Author Orcid Image ;   Tobias Nef1, 2 Author Orcid Image
Article Authors Cited by (1) Tweetations (3) Metrics

    1Gerontechnology & Rehabilitation Group, University of Bern, , Bern, , Switzerland

    2ARTORG Center for Biomedical Engineering Research, University of Bern, , Bern, , Switzerland

    3Department of Neurology, University Neurorehabilitation, University Hospital Bern, Inselspital, University of Bern, , Bern, , Switzerland

    4Department of Psychology, Manipal University Jaipur, , Jaipur, , India

    5University Hospital of Old Age Psychiatry and Psychotherapy, University of Bern, , Bern, , Switzerland

    *these authors contributed equally

    Corresponding Author:

    Prabitha Urwyler, PhD


    Background: Promoting cognitive health is key to maintaining cognitive and everyday functions and preventing the risk of cognitive impairment or dementia. Existing scientific evidence shows the benefits of various training modalities on cognition. One way to promote cognitive health is through engagement in cognitive activities (eg, board and video games).

    Objective: This study aims to investigate the benefits of dynamic adaptive casual puzzle games on cognitive function and well-being in healthy adults and older people.

    Methods: A total of 12 adults and older people (female participants: n=6; mean age 58.92, SD 10.28 years; range 46-75 years) were included in this pilot randomized controlled trial. This study used a crossover design with two phases (8 weeks each) and three measurement waves (pretest, midtest, and posttest). The participants were randomly allocated either to the control or experimental group. In the control group, participants read newspapers between the pre- and midtest, then switched to cognitive training with puzzle games. In the experimental group, the interventions were reversed. Baseline measurements (pretest) were collected before the intervention. The interventions were delivered on tablet computers and took place unsupervised at participants’ homes.

    Results: The outcome measures included global cognitive function, higher cognitive function, and emotional well-being at 3 time points (pretest, midtest, and posttest) using standardized neuropsychological tests. The participants showed improvements in their visual attention and visuospatial measures after the puzzle game intervention.

    Conclusions: The study showed that digital games are a feasible way to train cognition in healthy adults and older people. The algorithm-based dynamic adaption allows accommodations for persons with different cognitive levels of skill. The results of the study will guide future prevention efforts and trials in high-risk populations.

    Trial Registration: ClinicalTrials.gov NCT03139799; https://clinicaltrials.gov/study/NCT03139799

    JMIR Aging 2023;6:e46177

    doi:10.2196/46177

    Keywords

    puzzle games; aging; cognitive assessment; visual attention; adults; elderly; well-being; randomized controlled trial; RCT; older adult 


    The world’s older adult population is increasing in both size and proportion. By 2050, the world’s population of older people will nearly double [1]. The aging population inevitably requires better health care services for age-related diseases and the prevention of physical and cognitive health decline. Age-related cognitive decline affects memory, attention, orientation, perception, and executive and motor function that impact everyday activities and quality of life [2,3]. Considering population aging as a dominant worldwide phenomenon, preventing or reversing cognitive decline are clinical and public health priorities [3].

    The use of video games for training in strategic control in structured conditions may enhance the cognition of both adults and older people [4-7]. Ample research suggests that playing video games enhances a variety of cognitive skills such as visual selective attention, speed of processing, memory, reasoning, problem-solving, learning, and executive functioning [8-13]. The use of video games for cognitive assessment may facilitate reliable, personalized, and cost-effective tests [14]. Video games provide a variety of performance data (eg, number of errors, reaction time, and other task-based measures) that can make cognitive function assessments more accurate [15].

    Traditional cognitive training programs require an in-person meeting, which entails setting a meeting location, traveling, and scheduling an appointment. For traditional training programs, recruiting older adults who are homebound or live in assisted living is more challenging. Computer-based cognitive interventions are potentially a more efficient alternative to traditional training programs that enhance their users’ accessibility, affordability, and applicability. Computerized training provides real-time feedback and can be personalized, restoring users’ activity engagement and motivation for the program [16,17] and eliminating staff load and training requirements, as well as errors and deliverer and observational biases in patient performance and performance anxiety. For a long-term effect (improvement in a broad range of cognitive domains), it is important to maximize player engagement.

    Computerized training can be deployed either as virtual reality games, augmented reality games, serious games, or casual video games (CVGs). CVGs are primarily designed for entertainment [18] and share features that closely match recommendations for video games for older adults [19]. CVGs have simple and minimized game elements, rules, and goals (simplicity); can easily be stopped and replayed, are error-forgiving, and use flexible difficulty levels (flexibility); and allow players to quickly learn the game (accessibility). One of the main advantages of using CVG-based tools is in their motivational properties, which are crucial in making the interaction enjoyable and engaging [20]. Previous studies have demonstrated artificial intelligence (AI)–based interventions as a promising tool for enhancing cognition, quality of life, or well-being among older adults [21-23]. A recent large-scale study showed that regular engagement in Sudoku and similar puzzles represent a cognitively enriching leisure activity that prevents and delays age-related cognitive decline [24]. Several studies have suggested puzzle video games are sensitive to the cognitive or motor alternations of normal aging [25,26]. In particular, the mazelike numberlink and match-3 puzzle video games’ performance was shown as a strong predictor of assessing cognitive or motor variabilities in older adults [25,27]. These games can be varied in difficulty to match the user’s level of cognitive ability and can help prevent practice effects during repeated administration and reduce ceiling and flooring effects by continuously matching the task difficulty to the participant’s cognitive ability [28,29].

    The study aimed to conduct a pilot randomized controlled trial (RCT) to evaluate the potential benefits of the puzzle game intervention in healthy adults including older adults. The primary objective was to examine whether a puzzle game supported by AI personalization significantly improves attentional function (visual search attention) and leads to in-game learning effects. Other secondary objectives were to investigate improvements in further cognitive outcome measures proposed to be engaged by the puzzle game (attention, processing speed, working memory, and spatial reasoning) and the efficacy of the puzzle game intervention in reducing symptoms of depression, anxiety, and stress and improving everyday function and quality of life. First, we expected significant improvement in attentional and executive function (near transfer) and other cognitive functions engaged by the game (far transfer). Second, we generalized that well-being (mood and stress) will show significant improvement.


    Ethical Considerations

    Ethics approval was granted by the University of Bern Ethics Committee and the Ethics Committee Northwest/Central Switzerland (ID 2016-01281). The study was carried out following the current version of the Declaration of Helsinki and the Good Clinical Practice guidelines. A written informed consent form was sought from each participant.

    Participants

    The sample included healthy adults and older adults (N=12; female: n=6, 50%; mean age 58.92, SD 10.28 years; range 46-75 years) recruited from the local community. The inclusion criteria for participation were adults aged between 45 and 75 years and with a Montreal Cognitive Assessment (MoCA) score ≥24. The exclusion criteria for participation were any previous history of comorbid neurological or psychiatric deficits; any previous diagnosis of mild or major neurocognitive disorder; and insufficient coordinative, motor, and perceptual ability to handle a tablet computer. All participants had normal or corrected-to-normal vision.

    The data of a participant (ID 09) was excluded due to the incomplete measurement of the participant’s cognitive function and emotional well-being at the 3 time points (pretest, midtest, and posttest).

    Study Design

    The pilot was designed as a 16-week randomized controlled trial using a crossover design with two phases (8 weeks each) and three waves (pretest, midtest, and posttest) of measurement. The participants were randomly allocated either to the control or experimental group. Baseline measurements (pretest) were collected before the intervention. In the control group, participants read newspapers at least 3 times a week for 8 weeks between the pre- and midtest, then switched to cognitive training with puzzle games after the midtest. In the experimental group, the interventions were reversed compared to the control group. The interventions were delivered on tablet computers (10.2” Apple iPad 2019 model, 32 GB, 4G edition [for SIM card], Apple Inc., Cupertino, CA) and took place unsupervised at participants’ homes (Figure 1). After completing recruitment formalities, participants were given an introduction to the study followed by a training session of both games consisting of 4 trials each of three grid sizes (4 × 4, 5 × 5, and 6 × 6). In addition, a manual/handbook was given to each of the participants.

    ‎
    Figure 1. Study timeline. Participants were randomly allotted either to the control or experimental group. To test the baseline, a pretest was conducted for each participant; after 8 weeks, a midtest was conducted; and then after 8 weeks of the second intervention, a posttest was performed. In the experimental group, interventions were reversed.

    At 3 time points (pretest, midtest, and posttest), global cognitive function, higher cognitive function, and emotional well-being were assessed using standardized neuropsychological tests (computerized visual-scanning Test of Attentional Performance [TAP] task [30], MoCA [31], Trail Making Test [TMT] [32], and Snellgrove Maze Test [33]) and questionnaires (Profile of Mood States [34-36], State-Trait Anxiety Inventory [37,38], and a quality of life questionnaire [39]). To remove bias in too-close measurements, alternative versions of MoCA were used for the pretest, midtest, and posttest. Additionally, information on the cognitive load (NASA Task Load Index [40]), motivation (adapted version of Intrinsic Motivation Inventory [IMI] [41]), and familiarity with using a tablet (adapted Tablet Familiarity Questionnaire [42]) was collected at the end of the study. All assessments were administered in paper-and-pencil format except for the computerized visual-scanning TAP task presented on a laptop. In the visual-scanning TAP task, participants actively scanned a 5 × 5 matrix and indicated whether a specific target stimulus (square with a top opening) was present among 3 types of similar distractor stimuli (squares with openings on the left, right, or bottom).

    Newspaper-Reading Task

    The participants were instructed to read a newspaper on the iPad for about 8 weeks at least 3 times a week for about 20 minutes a day. Some of the locally known news apps (BBC, Tages-Anzeiger, 20 Minuten, NZZ, etc) were preinstalled, and participants were free to choose a news app of their choice.

    Puzzle Game Task

    Participants were instructed to play both puzzle games (numberlink and match-3), delivered to them on an iPad, a minimum of 3 times per week (maximum of 10 min/game). A time tracker in the game limited the participants’ daily training time to 20 minutes, while a scoreboarder displayed stars corresponding to the number of levels completed by the participants. The training duration of 20 minutes a day was selected to avoid any fatigue or complaints of addiction. We also hypothesized that playing for shorter sessions over a longer study duration had better transfer effects and avoided early ceiling effects. Previous studies on game interventions in adults also used a similar training duration [43,44]. To begin with, participants were assigned an initial difficulty level depending on their scores from the pretest. The difficulty level for the next day was updated depending on the game performance of the participants on the previous day by the AI-based method.

    The match-3 puzzle [27] (Figure 2 left) requires visual search, visuospatial working memory, and pattern recognition skills to swap game objects into a row of three identical objects. The numberlink puzzle [25] (Figure 2 right) requires visuospatial, visuo-constructional, and executive (planning, foresight, and reasoning) functions to find paths between identically colored pairs of game objects, filling all empty cells. A range of difficulty levels for both games was implemented using the height and width of the puzzle board (from 4 × 4 to 8 × 8) and the number of unique game objects (from 4 to 8).

    ‎
    Figure 2. Left: A match-3 puzzle is played by swapping objects to make a row of three identical objects. Right: A numberlink puzzle is played by finding paths to connect pairs of objects [45].

    Puzzle Game Difficulty

    To dynamically adjust the puzzle game difficulty during the cognitive intervention, an AI-based method was developed. We hypothesized that the AI method could continuously model the participant’s ability through a multidimensional combination of puzzle difficulty parameters (ability assessment) and could adjust the puzzle difficulty as the participant progresses within the intervention (adaptivity assessment). To begin with, the participant’s initial level is set using predefined variables collected at the pretest before the game intervention. When a puzzle level is solved, the AI service receives the solving time (state, tm) and decides whether this lies within a threshold (reward). The AI agent then learns whether to increase or decrease the puzzle’s difficulty.

    The model is built on a database of input features, a regressor, the time predictor, and the level predictor. The end points are match_three/ putInitialLevel, match_three/ putNextLevel, number_link/ putInitialLevel, and number_link/ putNextLevel. The database of input features is a collection of sorted participant training data and difficulty level (determined by the set size and the number of gems/tiles) data for the two games. The participants’ data were cleaned and prepared with pandas 1.0.3 and numpy 1.18.1. The regressor was selected depending on the adjusted R2 value on the training data set. Accordingly, a voting regressor was chosen for the match-3 puzzle and an MLPRegressor for the numberlink puzzle. The features of just completed or played levels are sent to a timing predictor to calculate the predicted time tp. The level predictor uses solved time tm and predicted time tp as inputs to provide an output of –1 (retreat one level), 0 (stay on the same level), or 1 (advance one level).

    where c is a constant between 0 and 1, and σ is the SD of all the games played on the given difficulty level in the data set.

    Implementation

    The AI service was implemented in Python 3.8.2 (Python Software Foundation) and deployed on a CentOS cloud instance. Every entity in the architecture of the AI service corresponds to a Python file. The AI service is queried via a representational state transfer (REST) application programming interface (FastAPI 0.52.0 and uvicorn 0.11.3). The pilot RCT tablet computers were equipped with SIM cards to ensure a continuous connection with the AI server. A mail-alerting service using the free tier of Mailgun was implemented to follow up with the daily progress of the participants’ gameplay.

    Data Analysis

    Time-based performance indicators such as overall solving time (minutes), average target search time (seconds), and processing time per item (seconds) were calculated from the files stored on the tablet computer. The improvement in the match-3 puzzle was calculated as the slope of the search time over the intervention period; similarly, the slope of path completion time was used for the numberlink puzzle.

    Statistical analysis was performed in SPSS v20.0 (IBM Corp). The normality of all values of interest was checked using the Shapiro-Wilk test [46]. An α value of .05 was used to determine significance. To evaluate the intervention effect on participants’ cognitive function and emotional well-being at 3 time points (pretest, midtest, and posttest), repeated-measure ANOVA was performed. Paired t tests were used to compare the visual search attention task (visual-scanning TAP) before and after the intervention wave. Correlation analyses were used to examine the intervention performance and efficiency and the cognitive and emotional measures.


    The demographics and characteristics of the participants included in this pilot study are shown in Table 1.

    Table 1. Participant demographics and characteristics.
    Group and subject IDAge (years)GenderHandednessMoCAaTMTb-A (s)TMT-B (s)SMTc (s)
    Control group
    0175MaleRight263811927
    0251FemaleLeft28296713
    0346FemaleLeft29234610
    0466FemaleRight29439531
    0553FemaleRight30184313
    0656MaleLeft28457111
    0774MaleRight255126834
    Experimental group
    0848FemaleRight29294913
    1059MaleRight242212218
    1172MaleRight254112029
    1252FemaleRight28204515
    1355MaleRight30285716

    aMoCA: Montreal Cognitive Assessment (score: 1-30).

    bTMT: Trail Making Test. TMT-A is the selective attention test set. TMT-B is the divided attention test set.

    cSMT: Snellgrove Maze Task.

    The participants assigned to the control group read newspapers (20 min/d) between the first and second waves of assessments, then switched to cognitive training with the match-3 (10 min/d) and numberlink (10 min/d) puzzles after the second wave of assessments. In the experimental group, the interventions were reversed compared to the control group. The AI server logged every request sent from the iPad to the server. Figure 3 shows the duration of the game training intervention for the numberlink puzzle and the progress in the difficulty level.

    ‎
    Figure 3. Numberlink difficulty level of subjects S01 to S10 over intervention duration.

    There was a significant difference between the pre- and postintervention primary outcome measure (visual-scanning TAP; total trials: mean 5.04, SD 0.82 vs mean 4.41, SD 0.63; t11=3.5; P=.005; trials with targets present: mean 4.98, SD 0.76 vs mean 4.41, SD 0.73; t11=4.23; P=.001). No significant differences were found with the measures of global cognition (MoCA), selective (TMT-A completion time) and divided (TMT-B completion time) attention, mood disturbances, anxiety, and quality of life when compared between the 3 time points (pretest, midtest, and posttest), as shown in Table 2. However, there was a significant difference in the MoCA visuospatial executive subscore for the midtest and posttest (mean 3.83, SD 1.03 vs mean 4.58, SD 0.67; P=.04).

    Table 2. Neuropsychological, emotional, and quality of life scores.
    Variables and time pointScore, mean (SD)P valuea
    Montreal Cognitive Assessment (score: 1-30).40
    Pre27.58 (2.06)
    Mid26.92 (2.43)
    Post27.58 (2.64)
    Trail Making Test–Ab (s).44
    Pre32.25 (10.96)
    Mid31.50 (12.99)
    Post30.08 (10.88)
    Trail Making Test–Bc (s).13
    Pre91.83 (63.43)
    Mid63.75 (30.36)
    Post66.75 (27.48)
    Snellgrove Maze Task (s).02
    Pre19.17 (8.59)
    Mid19.75 (9.55)
    Post16.50 (8.12)
    Total mood disturbance (from Profile of Mood States).35
    Pre47.58 (7.34)
    Mid47.25 (6.05)
    Post43.42 (10.83)
    Trait anxiety.51
    Pre30.48 (17.11)
    Mid27.46 (14.94)
    Post24.53 (11.76)
    State anxiety.99
    Pre26.11 (14.38)
    Mid25.68 (18.04)
    Post25.72 (22.77)
    EQ-5D-L–health.38
    Pre90.42 (5.42)
    Mid86.33 (9.95)
    Post88.75 (6.78)
    Test of Attentional Performance (total).005
    Pre5.04 (0.82)
    Post4.41 (0.63) 
    Test of Attentional Performance (positive).001
    Pre4.98 (0.76)
    Post4.41 (0.73) 
    Test of Attentional Performance (negative).71
    Pre5.44 (2.47)
    Post5.19 (1.53) 

    aP value <.05 is significant.

    bTrail Making Test–A is the selective attention test set.

    cTrail Making Test–B is the divided attention test set.

    Correlation analyses showed a significant positive association between participants’ fondness to read computer magazines and the search time slope (r=0.97; P<.001) and the slope of path time (r=0.75; P=.01). Both the search time slope (r=−0.83; P=.003) and the slope of path time (r=−0.78; P=.007) were found to be negatively correlated with the IMI subscale “pressure/tension.”

    Of the 12 participants, 8 reported that they had used tablet computers before and could perform tasks successfully with them (Figure 4). Participants rated the IMI subscale “interest/enjoyment” with the highest score, while the “value/usefulness” was scored the lowest (Table 3). Participants scored physical demand the lowest compared to the other task load domains for the game intervention (Table 4).

    ‎
    Figure 4. Overview of the tablet computer familiarity scores.
    Table 3. Intrinsic Motivation Inventory subscale scores reported by the participants.
    Participant IDCompetence/effortInterest/enjoymentValue/usefulnessPressure/tension
    0113451321
    021341919
    03832221
    041441519
    051129321
    0610401119
    0713421220
    081340514
    101338918
    1114441321
    121343418
    131337414
    Table 4. Task Load Index subscale scores reported by the participants.
    Participant IDMental demandPhysical demandTemporal demandPerformanceEffortFrustration
    01611213
    02613423
    034118079
    04814235
    051026442
    06123122143
    071424242
    08512143
    10713443
    1115115111
    122152011
    13101101582

    Principal Findings

    Puzzle games have the potential to target the following cognitive domains: learning and memory (working memory), attention (visual search), executive functions (inhibition and flexibility), and perceptual motor function (visuospatial ability). This study aimed to assess the feasibility of using a tablet-based puzzle game intervention to improve cognitive function and well-being.

    We focused on associations between attention executive functions and game performance measures. The results of this pilot study show improvements in visual attention and visuospatial measures after the intervention.

    Limitations

    The main limitation is the small sample size, which poses a risk of providing false significant results. If rigorously tested and evaluated in larger cohorts, it will help to increase the methodological rigor, amplify the transferability, and thereby enhance the specificity and sensitivity of the puzzle game as a diagnostic tool. The other limitation was the duration of our pilot study, which was not long enough. The successive study should be longer to measure a far transfer as well as minimize novelty effects in the data. The IMI scale was assessed at posttest (in the control group after the puzzle game intervention and in the experimental group after the newspaper-reading task) and might be a potential bias in the study. The carryover design and having no washout period owing to a carryover bias can be a possible shortcoming. Although reading a newspaper is a good control for cognitive tasks, it can be counterintuitive, thereby increasing anxiety about mental health outcomes. Therefore, it would be advisable to have an alternate control group engaged in other control activities such as playing games, which are readily accessible on various tablet applications and easily downloadable.

    Comparison With Prior Work

    Visuospatial reasoning has been reported as a marker of Alzheimer disease and mild cognitive impairment [47]. Our results of improvement in the visual-scanning attentional and visuospatial measures also align with earlier studies where video games have been shown to benefit fluid cognitive abilities [48]. Fluid cognitive abilities are subject to age-related cognitive decline and crucial in activities of daily living, and hence need to be a focus in cognitive studies aiming for quality of life improvements [49].

    However, most RCTs only investigate improvements in cognitive function. In our pilot study, we have attempted to evaluate the efficacy of training on cognition, emotional well-being, and quality of life.

    A key aspect of motivational and enriching gameplay is the ability to continuously learn and adapt to new challenges. Dynamic adjustment systems can help to adapt to higher levels or to predict when the player is failing repeatedly to avoid such states. Random difficulty adjustment does not consider the player and can lead to unbalanced and unpredictable difficulty. Much of the existing dynamic level adjustment systems rely upon prediction. AI-based algorithms are used to follow these trends in predicting the time and level. The dynamic and personalized adjustment used in this study also uses a data-driven behavior that emerges from interaction with the game. The agent uses previous participant training data to learn and work out the prediction strategy. By using a person-centric adaption of the difficulty level, we were able to not only accommodate the heterogeneity in cognitive ability between participants but also continuously match the game difficulty to the skill level of the participant, thus benefitting the training effect and long-term motivation. Based on our results, it was also shown that participants with good visual search and executive skills progress faster to higher levels. Previous studies have demonstrated the effectiveness of video games for promoting multiple cognitive domains [5,50-53], quality of life [54], and emotional skills [55,56]. However, in our study, we didn’t find any significant difference in quality of life and well-being. The possible reason could be a small sample size, and future studies on a larger cohort might show significant changes in these variables.

    A significant positive association between the participant’s fondness for reading computer magazines and puzzle game improvement indicates a high level of interest in technology and gaming. When individuals engage in reading computer magazines, they are exposed to similar cognitive challenges such as understanding technical concepts or following complex instructions. This mental exercise can transfer to puzzle games, improving their ability to think critically, solve puzzles more efficiently, and adapt to different game scenarios. When individuals are passionate about a subject, they tend to invest more time and effort into it. Consequently, the more engaged individuals are, the more opportunities they have to practice and improve their puzzle-solving skills, which can lead to better performance over time.

    Research is lacking evidence of the long-term effects of such interventions, particularly if learning and improvement are meant to be lasting. The crossover design of this study minimizes the possibility of a carryover effect being a possible avenue of exploration. The sample size was heterogeneous, meaning the game applications were tested on a broad group of users. Previous studies have suggested cognitive decline begins at the age of 45 years [57,58]. Therefore, including a broader age range of adults (aged ≥45 years) in our study allowed for a better understanding of cognitive decline across different age groups and enhanced the research findings’ applicability to a wider range of individuals.

    Conclusions

    Digital games are a feasible way to train cognition while allowing additional data for assessment and monitoring progress. Gameplay can induce neuroplastic reorganization that leads to long-term retention and transfer of skill. However, the two games used in the training intervention targeted different cognitive domains. Some participants can have excellent spatial navigation skills, while others are good at visual divided attention tasks. The benefit of games on players who are already good in the domains targeted by the game is questionable. Future studies should be longitudinal and target dedicated domains to avoid ceiling effects.

    Acknowledgments

    The authors would like to thank the master students (Allet Noemi, Andrea Bächler, Saskia Bühler, Anja Winiger, Linda Etter, and Karolina Jankosva) for their valuable help with recruiting and testing participants. The study was partially supported by a Novartis-Freenovation grant to PU. The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

    Conflicts of Interest

    None declared.

    Checklist 1

    CONSORT-eHEALTH checklist (V 1.6.1).

    PDF File, 96 KB
    1. Ageing and health. World Health Organization. 2022 Oct 1. URL: www.who.int/news-room/fact-sheets/detail/ageing-and-health [accessed 2023-10-10]
    2. Deary IJ, Corley J, Gow AJ, Harris SE, Houlihan LM, Marioni RE, et al. Age-associated cognitive decline. Br Med Bull 2009 Dec;92:135-152 [CrossRef] [Medline]
    3. Wolinsky FD, Vander Weg MW, Howren MB, Jones MP, Dotson MM. A randomized controlled trial of cognitive training using a visual speed of processing intervention in middle aged and older adults. PLoS One 2013 May 1;8(5):e61624 [CrossRef] [Medline]
    4. Hertzog C, Kramer AF, Wilson RS, Lindenberger U. Enrichment effects on adult cognitive development: can the functional capacity of older adults be preserved and enhanced? Psychol Sci Public Interest 2008 Oct;9(1):1-65 [CrossRef] [Medline]
    5. Pallavicini F, Ferrari A, Mantovani F. Video games for well-being: a systematic review on the application of computer games for cognitive and emotional training in the adult population. Front Psychol 2018 Nov 7;9:2127 [CrossRef] [Medline]
    6. Flores-Gallegos R, Mayer RE. Learning cognitive skills by playing video games at home: testing the specific transfer of general skills theory. J Cogn Enhanc 2022 Dec;6(4):485-495 [CrossRef]
    7. Wang P, Liu HH, Zhu XT, Meng T, Li HJ, Zuo XN. Action video game training for healthy adults: a meta-analytic study. Front Psychol 2016 Jun 17;7:907 [CrossRef] [Medline]
    8. Green CS, Bavelier D. Action video game modifies visual selective attention. Nature 2003 May 29;423(6939):534-537 [CrossRef] [Medline]
    9. Dye MWG, Green CS, Bavelier D. Increasing speed of processing with action video games. Curr Dir Psychol Sci 2009;18(6):321-326 [CrossRef] [Medline]
    10. Clemenson GD, Stark SM, Rutledge SM, Stark CEL. Enriching hippocampal memory function in older adults through video games. Behav Brain Res 2020 Jul 15;390:112667 [CrossRef] [Medline]
    11. Wang P, Zhu XT, Qi Z, Huang S, Li HJ. Neural basis of enhanced executive function in older video game players: an fMRI study. Front Aging Neurosci 2017 Nov 21;9:382 [CrossRef] [Medline]
    12. Chen X. The relationship between video games, problem-solving skills, and academic performance from IT students’ perspective [master's thesis]. University of Oulu. 2019 Mar. URL: jultika.oulu.fi/files/nbnfioulu-201904051435.pdf [accessed 2023-10-11]
    13. Lamb RL, Annetta L, Firestone J, Etopio E. A meta-analysis with examination of moderators of student cognition, affect, and learning outcomes while using serious educational games, serious games, and simulations. Comput Hum Behav 2018 Mar;80:158-167 [CrossRef]
    14. Holmgard C, Togelius J, Henriksen L. Computational intelligence and cognitive performance assessment games. Presented at: 2016 IEEE Conference on Computational Intelligence and Games (CIG); Sep 20-23, 2016; Santorini, Greece [CrossRef]
    15. Areàn PA, Hoa Ly K, Andersson G. Mobile technology for mental health assessment. Dialogues Clin Neurosci 2016 Jun 30;18(2):163-169 [CrossRef] [Medline]
    16. Kueider AM, Parisi JM, Gross AL, Rebok GW. Computerized cognitive training with older adults: a systematic review. PLoS One 2012 Jul 11;7(7):e40588 [CrossRef] [Medline]
    17. Rebok GW, Carlson MC, Langbaum JBS. Training and maintaining memory abilities in healthy older adults: traditional and novel approaches. J Gerontol B Psychol Sci Soc Sci 2007 Jun 1;62 Spec No 1:53-61 [CrossRef] [Medline]
    18. Baniqued PL, Lee H, Voss MW, Basak C, Cosman JD, Desouza S, et al. Selling points: what cognitive abilities are tapped by casual video games? Acta Psychol (Amst) 2013 Jan;142(1):74-86 [CrossRef] [Medline]
    19. Gerling KM, Schulte FP, Smeddinck J, Masuch M. Game design for older adults: effects of age-related changes on structural elements of Digital games. In: Herrlich M, Malaka R, Masuch M, editors. Entertainment Computing - ICEC 2012: Springer; 2016:235-242 [CrossRef]
    20. Stroud MJ, Whitbourne SK. Casual video games as training tools for attentional processes in everyday life. Cyberpsychol Behav Soc Netw 2015 Nov;18(11):654-660 [CrossRef] [Medline]
    21. Baragash RS, Aldowah H, Ghazal S. Virtual and augmented reality applications to improve older adults’ quality of life: a systematic mapping review and future directions. Digit Health 2022 Oct 31;8:20552076221132099 [CrossRef] [Medline]
    22. Mukaetova-Ladinska EB, Harwood T, Maltby J. Artificial intelligence in the healthcare of older people. Arch Psychiatr Ment Health 2020 Mar 20;4:007-013 [CrossRef]
    23. Ma B, Yang J, Wong FKY, Wong AKC, Ma T, Meng J, et al. Artificial intelligence in elderly healthcare: a scoping review. Ageing Res Rev 2023 Jan;83:101808 [CrossRef] [Medline]
    24. Ferreira N, Owen A, Mohan A, Corbett A, Ballard C. Associations between cognitively stimulating leisure activities, cognitive function and age-related cognitive decline. Int J Geriatr Psychiatry 2015 Apr;30(4):422-430 [CrossRef] [Medline]
    25. Nef T, Chesham A, Schütz N, Botros AA, Vanbellingen T, Burgunder JM, et al. Development and evaluation of maze-like puzzle games to assess cognitive and motor function in aging and neurodegenerative diseases. Front Aging Neurosci 2020 Apr 21;12:87 [CrossRef] [Medline]
    26. Sousa DE, Oliveira DC, Campos TF. Maze test: an instrument to evaluate age-related cognitive and motor changes in humans. Revista Brasileira Ciências Envelhecimento Humano 2013 Mar;10(1) [CrossRef]
    27. Chesham A, Gerber SM, Schütz N, Saner H, Gutbrod K, Müri RM, et al. Search and match task: development of a taskified match-3 puzzle game to assess and practice visual search. JMIR Serious Games 2019 May 9;7(2):e13620 [CrossRef] [Medline]
    28. Davis R, Libon DJ, Au R, Pitman D, Penney DL. Think: inferring cognitive status from subtle behaviors. Proc AAAI Conf Artif Intell 2014 Jul;2014:2898-2905 [Medline]
    29. Loe BS, Rust J. The perceptual maze test revisited: evaluating the difficulty of automatically generated mazes. Assessment 2019 Dec;26(8):1524-1539 [CrossRef] [Medline]
    30. Zimmermann P, Fimm B. A test battery for attentional performance. In: Applied Neuropsychology of Attention: Theory, Diagnosis and Rehabilitation: Psychology Press; 2004:124-165 [CrossRef]
    31. Nasreddine ZS, Phillips NA, Bédirian V, Charbonneau S, Whitehead V, Collin I, et al. The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc 2005 Apr;53(4):695-699 [CrossRef] [Medline]
    32. Reitan RM, Wolfson D. Neuropsychological evaluation of older children: Neuropsychology Press; 1992.
    33. Snellgrove CA. Cognitive screening for the safe driving competence of older people with mild cognitive impairment or early dementia. Department of Infrastructure, Transport, Regional Development, Communications and the Arts. 2005. URL: www.infrastructure.gov.au/​sites/​default/​files/​migrated/​roads/​safety/​publications/​2005/​pdf/​cog_screen_old.​pdf [accessed 2023-10-11]
    34. Albani C, Blaser G, Geyer M, Schmutzer G, Brähler E, Bailer H, et al. The German short version of “Profile of Mood States" (POMS): psychometric evaluation in a representative sample (article in German). Psychother Psychosom Med Psychol 2005 Jul;55(7):324-330 [CrossRef] [Medline]
    35. Shacham S. A shortened version of the profile of mood states. J Pers Assess 1983 Jun;47(3):305-306 [CrossRef] [Medline]
    36. Dalbert C. Young adults' subjective well-being: theoretical and empirical analyses of its structure and stability (article in German). Zeitschrift Differentielle Diagnostische Psychologie 1992;13(4):207-220
    37. Laux L. Das State-Trait-Angstinventar (STAI): Theoretische Grundlagen Und Handanweisung: Beltz; 1981.
    38. Spielberger CD. State-Trait Anxiety Inventory for Adults. APA PsycNet. 1983. URL: psycnet.apa.org/doiLanding?doi=10.1037%2Ft06496-000 [accessed 2023-10-23]
    39. EuroQol Group. EuroQol--a new facility for the measurement of health-related quality of life. Health Policy 1990 Dec;16(3):199-208 [CrossRef] [Medline]
    40. Hart SG. NASA-task load index (NASA-TLX); 20 years later. Proc Hum Factors Ergonomics Soc Annu Meeting 2006 Oct;50(9):904-908 [CrossRef]
    41. Ryan RM. Control and information in the intrapersonal sphere: an extension of cognitive evaluation theory. J Pers Soc Psychol 1982 Sep;43(3):450-461 [CrossRef]
    42. Schulenberg SE, Melton AMA. The computer aversion, attitudes, and familiarity index (CAAFI): a validity study. Comput Hum Behav 2008 Sep 17;24(6):2620-2638 [CrossRef]
    43. Sosa GW, Lagana L. The effects of video game training on the cognitive functioning of older adults: a community-based randomized controlled trial. Arch Gerontol Geriatr 2019;80:20-30 [CrossRef] [Medline]
    44. Parsons B, Magill T, Boucher A, Zhang M, Zogbo K, Bérubé S, et al. Enhancing cognitive function using perceptual-cognitive training. Clin EEG Neurosci 2016 Jan;47(1):37-47 [CrossRef] [Medline]
    45. Chesham A. Puzzling the mind: development and evaluation of puzzle games to assess cognitive and motor function in neurocognitive aging [dissertation]: Faculty of Medicine of the University of Bern; 2019.
    46. Shapiro SS, Wilk MB. An analysis of variance test for normality (complete samples). Biometrika 1965 Dec;52(3/4):591-611 [CrossRef]
    47. Hort J, Laczó J, Vyhnálek M, Bojar M, Bures J, Vlcek K. Spatial navigation deficit in amnestic mild cognitive impairment. Proc Natl Acad Sci U S A 2007 Mar 6;104(10):4042-4047 [CrossRef] [Medline]
    48. Jak AJ, Seelye AM, Jurick SM. Crosswords to computers: a critical review of popular approaches to cognitive enhancement. Neuropsychol Rev 2013 Mar;23(1):13-26 [CrossRef] [Medline]
    49. Arntzen AAB. Game based learning to enhance cognitive and physical capabilities of elderly people: concepts and requirements. Int J Inf Commun Eng 2011 Dec;5(12):1778-1782 [CrossRef]
    50. Oei AC, Patterson MD. Enhancing perceptual and attentional skills requires common demands between the action video games and transfer tasks. Front Psychol 2015 Feb 10;6:113 [CrossRef] [Medline]
    51. Schubert T, Finke K, Redel P, Kluckow S, Müller H, Strobach T. Video game experience and its influence on visual attention parameters: an investigation using the framework of the theory of visual attention (TVA). Acta Psychol (Amst) 2015 May;157:200-214 [CrossRef] [Medline]
    52. Rolle CE, Anguera JA, Skinner SN, Voytek B, Gazzaley A. Enhancing spatial attention and working memory in younger and older adults. J Cogn Neurosci 2017 Sep;29(9):1483-1497 [CrossRef] [Medline]
    53. Baniqued PL, Kranz MB, Voss MW, Lee H, Cosman JD, Severson J, et al. Cognitive training with casual video games: points to consider. Front Psychol 2014 Jan 7;4:1010 [CrossRef] [Medline]
    54. Tu JC, Yang CH, Liang CY, Chen HY. Exploring the impact of puzzle games for the elderly from experiential learning. Thinking Skills Creativity 2022 Sep;45:101107 [CrossRef]
    55. Naugle KE, Naugle KM, Wikstrom EA. Cardiovascular and affective outcomes of active gaming: using the Nintendo Wii as a cardiovascular training tool. J Strength Cond Res 2014 Feb;28(2):443-451 [CrossRef] [Medline]
    56. Bouchard S, Bernier F, Boivin E, Morin B, Robillard G. Using biofeedback while immersed in a stressful videogame increases the effectiveness of stress management skills in soldiers. PLoS One 2012;7(4):e36169 [CrossRef] [Medline]
    57. Singh-Manoux A, Kivimaki M, Glymour MM, Elbaz A, Berr C, Ebmeier KP, et al. Timing of onset of cognitive decline: results from Whitehall II prospective cohort study. BMJ 2012 Jan 5;344:d7622 [CrossRef] [Medline]
    58. Taylor CA, Bouldin ED, McGuire LC. Subjective cognitive decline among adults aged ≥45 years - United States, 2015-2016. MMWR Morb Mortal Wkly Rep 2018 Jul 13;67(27):753-757 [CrossRef] [Medline]

    ‎
    AI: artificial intelligence
    CVG: casual video game
    IMI: Intrinsic Motivation Inventory
    MoCA: Montreal Cognitive Assessment
    RCT: randomized controlled trial
    REST: representational state transfer
    TAP: Test of Attentional Performance
    TMT: Trail Making Test

    Edited by Haley LaMonica; submitted 01.02.23; peer-reviewed by Bin Huang, Huei-Ling Chiu, Liliana Mendes, Marcia Friesen; final revised version received 28.06.23; accepted 30.08.23; published 01.11.23

    Copyright

    © Prabitha Urwyler, Rajnish Kumar Gupta, Michael Falkner, Joel Niklaus, René Martin Müri, Tobias Nef. Originally published in JMIR Aging (https://aging.jmir.org), 1.11.2023.

    This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Aging, is properly cited. The complete bibliographic information, a link to the original publication on https://aging.jmir.org, as well as this copyright and license information must be included.