JMIR Publications

ABSTRACT

Background:

Exergames are increasingly used in rehabilitation to simultaneously target physical and cognitive functions. Game scores are commonly used as indicators of performance, yet it remains unclear to what extent they reflect underlying movement and cognitive-control demands. Understanding this relationship is essential for interpreting what exergame performance scores reflect during gameplay.

Objective:

To investigate how movements and frontal theta activity, used as a proxy for cognitive-control demands, relates to game performance across game levels and age groups in a step-based exergame.

Methods:

This cross-sectional laboratory study included 32 participants (16 older adults: mean age: 76.5±4.5 years, range 71-86 years, and 16 young adults: mean age: 25±2.1 years, range 22-28 years) recruited using convenience sampling from a university and recreational exercise groups in Trondheim, Norway. Participants completed a step-based exergame (“The Mole”) at three difficulty levels (Easy, Medium and Hard). Movements were measured using accelerometers (vector magnitude and step count), and frontal theta activity was derived from mobile EEG as mean power spectral density in the 4–7 Hz range as a proxy for cognitive-control demands. Game scores were recorded as performance indicators. Repeated measures general linear models were used to test effects of difficulty and age group, and Pearson correlation analyses examined associations between variables.

Results:

Younger adults achieved higher game scores than older adults across all difficulty levels, and performance declined at higher task difficulty (all p<.001). The amount of movement (vector magnitude) showed strong and consistent positive correlations with game score across all levels and both age groups (r=0.67–0.80, all p<.001). Step count showed weaker and less consistent associations across groups. No statistically significant differences in frontal theta activity were observed between age groups or across task difficulty levels (all p>.15). Associations between frontal theta and game score were near zero in younger adults (r=–0.096 to 0.001), whereas older adults showed small-to-moderate correlations at lower difficulty levels (r=0.345–0.416) and a stronger correlation at the highest difficulty level (r=0.76). However, none of the correlations between frontal theta and game score were statistically significant after correction for multiple comparisons (p>.0028).

Conclusions:

In this controlled exergaming setting, game performance was strongly associated with the amount of movement performed, whereas associations with frontal theta activity, used as a proxy for cognitive-control demands, were limited and context-dependent. These findings suggest that game scores primarily reflect movement output rather than cognitive-control processes under the present game conditions. The results indicate that game scores alone may provide an incomplete representation of engagement and highlight the importance of considering both physical and cognitive aspects when designing and evaluating exergames.