Background: Intertrial and test-retest reliability of isokinetic knee extension and flexion strength measurements was studied in eighteen elderly women (mean age 68 +/- 5 yrs), using increasing and decreasing angular velocity testing procedures.
Experimental design: five reciprocal knee extensions and flexions at three different angular velocities (90, 120, and 180 deg/sec) were performed by means of a MERAC apparatus on two occasions fifteen days apart.
Measures: Data relative to torque, power and work were considered.
Results: Statistically significant differences were found often among the five trials and intraclass correlation coefficients, ranging from 0.03 to 0.90 for extensor muscles; ranging from 0.44 to 0.89 for flexor muscles, were shown among the best three outputs. Peak values were reached within the first three repetitions, though a 20% chance to produce peak outputs was shown in the last two repetitions. Increasing and decreasing test velocity procedures did not consistently show statistically significance for peak outputs. Generally, for each isokinetic parameter higher and more frequently significant correlation coefficients between test and retest experimental sessions were found for peak values (ranging from 0.36 to 0.80) with respect to mean values (ranging from 0.22 to 0.74).
Conclusions: These data suggest that a test procedure including five trials tends to increase the chance of producing the best peak outputs. Furthermore, considering the best peak and mean performances could be more appropriate when studying elderly women.