Gait variability in older adults: a structured review of testing protocol and clinimetric properties.

Gait variability (stride-to-stride fluctuations) is used increasingly as a marker for gait performance and future mobility status, cognitive status, and falls. This structured review explicitly examined literature that reported on the reliability, validity and responsiveness of gait variability in older adults. We searched Medline, Embase, Web of Science, Scopus, CINAHL, PEDRO, Biomechanics, SportDiscus and PsycInfo databases. Two independent reviewers undertook data extraction, with adjudication by a third reviewer in cases of disagreement. Twenty-two full papers were screened and 10 met the predefined inclusion criteria, involving 1036 participants who were mainly community dwelling older adults in their 8th decade. A wide range of gait variability parameters, testing protocols and calculations of gait variability were reported. Reliability estimates varied, but were mostly fair to moderate. Concurrent validity was established for stance time variability and change estimates were reported for stance time and swing time. Standard of reporting was generally poor, with insufficient detail provided for aspects of measurement and testing protocols. Further research is required to standardise testing procedures and establish reliability, responsiveness and validity for confident use of gait variability as a robust measure.