A Gamified N-back App for Identifying Mild-cognitive Impairment in Older Adults

Abstract

Introduction: Despite a dramatic increase in the incidence of mild-cognitive impairment (MCI) and early dementia, accessible and engaging screening methods for older adults are lacking. Gamification has gained attention in the self-management of various health conditions, making it a promising avenue for dementia screening. This study aimed to evaluate a gamified mobile application for the early detection of cognitive impairment associated with dementia.

Methods: The gamified app and the Mini-Mental State Examination (MMSE) were administered to 138 participants. The game, based on the N-back working memory task, simulates a restaurant scenario where players cook curries with hidden ingredients to fulfill customer orders, with the difficulty increasing in each round. The correlations between MMSE scores and game metrics were analyzed, and the game metrics were compared between the normal and impaired groups.

Results: Among the 138 older adult participants, the game metrics such as level reached, accuracy, response times, tap times, and swipe times exhibited significant correlations with scores on the MMSE, a standard cognitive screening tool (r = 0.42, 0.419, −0.575, −0.484, and −0.667, respectively; P < 0.05 for all). The participants were divided into the normal (≥28) and impaired (<28) groups based on the MMSE cutoff values. The impaired group had significantly worse performance on all game metrics. After multivariate adjustment, average swipe time emerged as the strongest predictor, achieving 70.8% sensitivity and 80.6% specificity in detecting impairment using a 3.31-s cutoff (area under the curve = 0.820).

Conclusions: This classification accuracy was comparable to standard dementia screening tests. These results indicate the potential use of gamification with joyous experience for older adults to enable scalable cognitive screening beyond conventional testing paradigms.