Cognitive costs in motor-cognitive performance assessments depend on movement complexity and cognitive task design

Cognitive performance is typically assessed using computer-based tests where participants respond via a simple upper extremity motor task such as a button press. This type of assessment has been criticised for its low ecological validity that does not consider the interaction between cognitive and more complex motor skills in sports and everyday life situations, which results in motor-cognitive interference. Consequently, motor-cognitive assessments integrating a more complex motor response into a cognitive test have gained popularity. However, the cognitive costs in motor-cognitive tests due to the interference of cognitive and motor processes have not yet been determined. Therefore, this study aimed to quantify the cognitive costs in motor-cognitive tests.

Thirty-three healthy athletes performed four cognitive tests (simple reaction, choice-reaction, working memory, cognitive flexibility) in a cognitive and motor-cognitive setting. For the cognitive task, participants performed a computer-based cognitive assessment by responding with a button press on a keyboard. In the motor-cognitive condition, participants conducted a stepping movement which was identified by a Light Detection and Ranging (LiDAR) system integrated into motor-cognitive testing and training technology (SKILLCOURT®). Cognitive costs were determined by comparing reaction time and error rate between conditions (cognitive vs. motor-cognitive) while controlling for differences in measurement technology, neuromuscular conduction delay, and movement amplitude. Correlation analyses quantified the relationship between cognitive and motor-cognitive performance.

There were cognitive costs, as indicated by slower reaction times in the motor-cognitive test, for the choice-reaction (p = 0.014) and working memory (p < 0.001) tests. There were inverse cognitive costs, denoted by faster reactions, in the motor-cognitive compared with the cognitive condition for the cognitive flexibility test (p < 0.001). There were strong correlations for the simple-reaction (r₃₁  = 0.79, p < 0.001), choice-reaction (r₃₁  = 0.60, p < 0.001), and cognitive flexibility (r₂₈  = 0.83, p < 0.001) tests. The working memory task revealed a moderate correlation (r₃₁  = 0.46, p = 0.009).

The results confirm the presence of cognitive costs in motor-cognitive assessments. The type of motor response and test design influence cognitive costs and test performance and can even result in inverse cognitive costs during motor-cognitive tasks. This must be considered when interpreting motor-cognitive tests and suggest that computer-based assessments cannot simply be replaced by motor-cognitive alternatives.