Measuring physical performance in later life: reliability of protocol variations for common performance-based mobility tests

Background and aims: Performance-based tests of mobility or physical function such as the Timed Up and Go (TUG), gait speed, chair-rise, and single-leg stance (SLS) are often administered using different protocols in aging populations, however, the reliability of their assessment protocols is not often considered. The purpose of this study was to examine the reliabilities of frequently used assessment protocols for the TUG, gait speed, chair-rise, and SLS in different age groups. Methods: We administered the following assessment protocols in an age-stratified (50–64, 65–74, 75+ years) sample of participants (N= 147) from the Canadian Longitudinal Study on Aging (CLSA): TUG fast pace and TUG normal pace: TUG-cognitive counting backwards by ones and counting back by threes, gait speed with 3-m and 4-m course, chair-rise with arms crossed and allowing the use of arms, and SLS using preferred leg or both legs—on two occasions within 1 week. We assessed the relative (intra-class correlation) and absolute reliability (standard error of measurement, SEM and minimal detectable change, MDC) for each protocol variation and provided recommendations based on relative reliability. Results: For participants aged 50–64 years, our results suggest better reliability for TUG fast-pace compared with normal-pace (ICC and 95% CI 0.70; 0.41–0.85 versus 0.38; 0.12–0.59). The reliability values for 3-m gait speed were potentially higher than for 4-m gait speed (ICC 0.75; 0.67–0.82 versus 0.64; 0.54–0.73) and values for chair-rise suggested better reliability allowing participants to use their arms than with arms crossed (ICC 0.79; 0.66–0.86 versus 0.64; 0.45–0.77) for participants overall. For participants aged 75+ years, ICCs for SLS with the preferred leg showed better reliability than for both legs (ICC = 0.62–0.79 versus 0.30–0.39). Conclusions and discussion: These reliability data and the recommendations can help guide the selection of the most appropriate performance-based test protocols for measuring mobility in middle-aged and older community-dwelling adults.