A prediction model of multiple resource theory for dual task walking

Abstract A model of Multiple Resource Theory (MRT) was created to describe the influence of different concurrent tasks on driving but no such framework exists for DT walking. The objective of this study was to modify MRT model and computational DT framework to explore two task-related concepts—resource demand and resource conflict—in the context of gait and to illustrate the initial validity of the modified model. A literature search was conducted to identify DT walking-related elements related to the MRT model. An instructional guide for utilizing the computational framework was constructed using a modified e-Delphi method. Inter-rater reliability of the constructed guideline was analyzed by Kappa statistics. The initial validity of the modified model was tested using meta-regression analyses. Three new elements were included into the modified model: vestibular and somatosensory perceptions, and locomotor response. The spatial code from the original model was modified into a visuospatial domain. The computational framework was constructed, and two separate guidelines were created for walking and concurrent tasks. Using the guidelines and computational framework, task-related elements were quantified for meta-regression analysis. The analysis showed that the modified model can explain 79% of the variability in speed reduction during DT walking.