Consistency and test-retest reliability of stepping tests designed to measure self-perceived and actual physical stepping ability in older adults.

Background: Older adults with an incorrect perception of their physical abilities may fall more often, suggesting a need for tests to quantify self-perceived and actual abilities.

Aims: To determine between-test consistency and test-retest reliability of three tests that measure self-perceived and actual stepping ability in older adults.

Methods: Older adults performed three stepping tests, covering high (bar test) and far steps (river and step tests). We studied between-test consistency in the perceived ability and actual ability of 269 participants at each task and in the difference between these two (degree of misjudgment). We also studied test-retest reliability in 21 participants.

Results: Perceived ability showed moderate consistency (r = 0.46-0.55, p < 0.001) and moderate-to-strong reliability [ICC(2,1) = 0.42-0.63, p < 0.03] for all tests. Actual ability showed strong consistency (r = 0.77, p < 0.001) and strong-to-excellent reliability [ICC(2,1) = 0.68-0.93, p < 0.001]. Degree of misjudgment was weakly consistent between two stepping far tests (r = 0.32, p < 0.001), but not consistent between stepping far and high tests (r = 0.05 and 0.06, p > 0.3). Test-retest reliability of the degree of misjudgment was poor-to-moderate [ICC(2,1) = 0.38 and 0.50, p < 0.05 on the two stepping far tests and ICC(2,1) = - 0.08, p = 0.63 on the stepping high test].

Conclusions: Actual and perceived ability can be consistently and reliably measured across tests, whereas the degree of misjudgment is less reliable and consistent within individuals.