Development of a smartphone application to measure reaction times during walking

Dual-task methodology is useful to assess walking ability. We developed a smartphone application to measure reaction times (RTs) during walking. We can assess the attentional demands for a task from the RTs. In experiment 1, the accuracy and precision of the RT application were evaluated in two subjects. We investigated the agreement between the RTs calculated based on the external inertial sensor and the RT application; the application was installed in two smartphone models with different levels of performance. Additionally, in experiment 2, we investigated the RTs under 4 conditions: sitting, overground walking, treadmill walking, and auditory cued overground walking (n=19). The constant systematic error and low standard deviation of the difference between the two methods was demonstrated; this depended on the sampling interval of each sensor. The RTs in overground walking were increased compared with sitting and decreased compared with treadmill walking and auditory cued overground walking. Overall, the RTs were more decreased in the smartphone model with the shorter sampling interval. The RT application would be valid within a smartphone with a similar level of performance, because bias and precision are strongly dependent on the sampling interval. In field tests under different walking conditions, the RT application obtained results similar to those seen in previous studies and could identify even slight differences if there were many trials. The developed RT application will be a simple tool that is able to assess attentional demands during dual-task walking.