Smartphone-based gait and balance accelerometry is sensitive to age and correlates with clinical and kinematic data

Abstract

BackgroundThe Gait&Balance (G&B) App has produced valid and reliable measures of gait and balance in young healthy adults but has not been tested in older adults. Research questionIn healthy middle-to-older aged adults, are G&B App measurements sensitive to age, valid against clinical and kinematic measures, and reliable? MethodHealthy participants (n = 34, 14 male, 42–94 years) completed the G&B App protocol three times within a single session. 3D kinematics were collected concurrently. Clinical balance measures were collected (Modified Clinical Test of Sensory Interaction in Balance (mCTSIB), Mini Balance Evaluation Systems Test (MBT), and Functional Gait Assessment (FGA)). Sensitivity to age was assessed with Pearson’s correlations. Validity tests included Pearson’s correlations and Bland-Altman limits of agreement. Reliability tests included intra-class correlation coefficients and standard error of the measure. ResultsDuring quiet stance on a compliant surface, the G&B App was sensitive to age-related differences not detectable with the mCTSIB. During walking tasks, there was adequate convergent validity between the MBT and G&B App measures of step length, and between the FGA and G&B App measures of walking speed, step length, and periodicity. The G&B App had moderate-to-excellent validity against 3D kinematics for postural stability during quiet stance (r 0.98 [0.98, 0.99]), step time (r 0.97 [0.96, 0.98]), walking speed (r 0.79 [0.7, 0.86]), and step length (r 0.73 [0.61, 0.81]). Test-retest reliability was moderate-to-excellent for G&B App measures of postural stability, walking speed, periodicity, step length, and step time. G&B App measures of step length asymmetry, step length variability, step time asymmetry, and step time variability had poor validity and reliability. SignificanceThe G&B App was sensitive to age-related differences in balance not detectable with clinical measurement. It provides valid and reliable measures of postural stability, step length, step time, and periodicity, which are not currently available in standard practice.