Concurrent validity of the step time and walking speed obtained from the smartphone application CareCoaching in independent, community-dwelling older adults

[Purpose] In this study, we verified the validity of the step time and walking speed obtained from the smartphone gait analysis application CareCoaching. [Participants and Methods] The participants were 66 independent, community-dwelling adults aged 65 years or older who performed a 10-m walking test twice each under preferred- and slow-speed conditions. We concurrently measured gait motions using CareCoaching and the OptoGait system for reference data. Both systems compute walking speed and step time as gait parameters. We examined the concurrent validity of these parameters by using intra-class correlation coefficients (ICCs) and limits of agreement (LOAs) with Bland−Altman analyses. [Results] In the preferred walking speed condition, the ICCs of walking speed and step times between the CareCoaching and the OptoGait system were 0.67 and 0.93, respectively. In the slow walking speed condition, the ICCs for walking speed and step time were 0.78 and 0.97, respectively. In addition, the LOAs for step time were −0.0941 to 0.1160 for preferred walking speed and −0.0596 to 0.0883 for slow walking speed. The LOAs for walking speed were −0.4158 to 0.0568 for preferred walking speed and −0.3348 to 0.0523 for slow walking speed. [Conclusion] CareCoaching showed excellent agreement for step time and moderate-to-good agreement for walking speed in independent, community-dwelling older adults.