Abstract
The isokinetic strength of ankle inversion-eversion was twice tested, over a period of one week, by the same examiner in apparently healthy 15 women and 15 men at 30 and 120°/s. Test – retest measurements were examined for 1 and 3 sets of three repetitions. Intraclass correlation coefficients for the highest peak torque out of 1 set of 3 repetitions varied from 0.47–0.82 for men and 0.70–0.94 for women. The ICC for the mean-based PT (the 3 highest PTs out of 3 sets of 3 replicates) ranged 0.74–0.83 for men and 0.80–0.97 for women. Bland-Altman plots have indicated a largely homoscedastic distribution of the test-retest parameters. The standard error of measurement (SEM) varied from 3.6–7.00 Nm for men and 3.3–4.7 Nm for women (highest PT-based) and 2.9–5.7 Nm for men and 1.6–4.0 Nm for women (mean PT-based). Irrespective of either paradigm, there was much dispersion in the ICCs for both genders in other isokinetic parameters such as the time to PT and angle at PT. It is therefore concluded that the only reproducible isokinetic parameter under this test protocol is the peak torque.