Affiliations: [a] Institute of Movement and Neuroscience, German Sport
University, Cologne, Germany
Note: [] Corresponding author: Tobias Alt, Institute of Movement and
Neuroscience, German Sport University, Am Sportpark Müngersdorf 6, 50933
Cologne, Germany. Tel.: +49 221 4982 3322; Fax: +49 221 4973 454; E-mail:
[email protected]
Abstract: BACKGROUND: Although the reliability of isokinetic strength tests of
knee flexors (flex) and extensors (ext) has been examined several times,
statistical evidence about the influence of internal and external factors is
missing. OBJECTIVE: This study aims to examine the impact of familiarisation,
muscle group, contraction mode, angular velocity and test parameters on the
reproducibility of findings derived from an isokinetic dynamometer (IsoMed
2000). METHODS: Thirty-three male subjects (mean age: 22.3 years) with no
prior experience of isokinetic exercise participated in three identical test
sessions (T1, T2, T3), each separated by 48–72 h. Peak moment (PM), angle
of peak moment (APM) and contractional work (CW) were determined unilaterally
(left and right) during maximum concentric (con) and eccentric (ecc) knee
flexion (abdominal position) and extension (supine position) at 30, 90 and
150°/s, respectively. An ANOVA with repeated measures confirmed systematic
bias. Reproducibility of consecutive tests (T1–T2, T2–T3) was
assessed by calculating the intra-class correlation coefficient (ICC 2,1)
(relative reliability) as well as the standard error of measurement (SEM)
(absolute reliability). ICC values were averaged according to respective
factors (Fisher's z-transformation) and tested for significant differences by
Steiger's formulas. RESULTS: PM and CW demonstrated a high absolute reliability (SEM:
4.7–10.5%). Relative reproducibility varied considerably (p⩽ 0.05)
between muscles (ecc flex > ecc ext), contraction modes (con ext > ecc
ext) and test parameters (PM=CW > APM), but did not depend on angular
velocity (30=90=150°/s). Due to familiarisation the reliability of PM
obtained from eccentric knee extensions significantly increased (T2–T3
> T1–T2). CONCLUSIONS: These results improve the development and execution of
reliable isokinetic strength testing protocols for unilateral knee flexion and
extension together with the interpretation of different test parameters.
Keywords: Reproducibility, intra-class correlation coefficient, standard error of measurement, maximum strength testing, familiarisation, IsoMed 2000
