Torque, total work, power, torque acceleration energy and acceleration time assessed on a dynamometer: reliability of knee and elbow extensor and flexor strength measurements.

Isometric torque and isokinetic peak torque, total work, power, torque acceleration energy and acceleration time at 30, 120 and 240 degrees.s-1 of the knee and elbow extensors and flexors were measured using an isokinetic dynamometer in 24 healthy women. Intrasession variation of the measurements was evaluated and the short-term and long-term reliability was assessed by repeating all procedures after averages of 2 and 30 days, respectively. The effect of learning on peak torque during a session was also evaluated. Moreover, the effect of general warming-up on knee extensor and flexor strength was examined on a separate day. Using correlations, numerous studies have indicated that muscle strength measurements are reliable. Correlations, however, are inappropriate and misleading in studies on reliability. In the present study reliability of each strength variable was expressed as the coefficient of variation (CV). With the protocol used, neither learning nor warming-up had any significant effect on strength. As expected, intra-session variation tended to be less than short-term and long-term inter-session variation. The CVs for strength variables measured 30 days apart exceeded 5% for all variables and rose to 107% for acceleration time. Substantial between-subject variation of individual CVs were found. The study demonstrated that muscle strength measurements may be highly unreliable in the individual subject.