Reliability Of Lower Extremity Torque Production In Persons With Post-stroke Hemiparesis

Abstract

A growing base of evidence suggests that weakness is the primary source of motor impairment following stroke and therefore, that resistance training may be a critical element of effective rehabilitation. The velocity-dependent nature of hemiparetic weakness has only recently been recognized and remains to be fully explored. Prior to conducting longitudinal investigation of resistance training in hemiparetic adults, it is necessary to identify objective and quantitative indicators of muscle performance including reliability across the full range of functional movement speeds. PURPOSE To establish the reproducibility of dynamic torque measurements in persons affected with post-stroke hemiparesis. METHODS Maximum voluntary isometric and isokinetic concentric and eccentric leg extensor torques were obtained on the contralesional leg in 17 subjects with hemiparesis (mn 58 ± 7.2 years, mn 6.1 ± 4.8 years post-stroke) and on a randomly selected leg in 10 age-matched control subjects (mn 63 ± 11.7 years) using a commercially available dynamometer. Data were obtained at 14 isokinetic speeds ranging from −180 deg/s to 240 deg/s and under isometric conditions. Each subject repeated the protocol on three separate days (days 1, 3 and 10) performing five trials of each condition. For each trial, torque was averaged over a 100 ms window centered at the highest value while at criterion speed. Data from trials 2–4 were averaged. Intra-class correlation (ICC) analysis was used to examine the reliability of force production across the range of criterion speeds and repeated testing sessions. RESULTS Intra-class correlation coefficients (ICCs) obtained from a random effects ANOVA model were excellent for both control (.989) and hemiparetic subjects (.957). ICCs for each criterion speed, obtained using a mixed model ANOVA, were also consistently high for both groups (> .90) with the exception of 240 deg/s in hemiparetic subjects, where the value was .77. Control subjects demonstrated somewhat higher reliability across sessions than hemiparetic subjects, with ICCs ranging from .947–.993 and .767–.954, respectively. In the hemiparetic subjects, reliability was somewhat higher in eccentric (.902–.954) compared to concentric (.767–.925) contractions. CONCLUSION In persons with chronic hemiparesis and stable motor impairment excellent reliability of lower extremity torque production was obtained under constrained task conditions across a range of criterion speeds and multiple testing sessions. These observations can be utilized for determining the effect sizes necessary to demonstrate positive effects of interventions investigating mechanisms of motor recovery.