Reproducibility and reliability of measurements using a linear isokinetic dynamometer, Aristokin

This study was performed to investigate the intra-observer reproducibility and reliability of measurements using a linear isokinetic dynamometer (Aristokin).

Sixteen female volunteers (age 18 to 23 years) participated in 4 test sessions, each consisting of 6 warming-up movements followed by 6 repetitions of 4 different movements. During each session, the subjects consecutively performed lifting movements by flexing elbows and shoulders (at 65 cm per sec), total lifting movements with arms and legs (at 65 cm per sec), and extension movements of the legs in sitting position (at 40 and 60 cm per sec). The first 3 test sessions were performed at a weekly interval, the 4th was executed 4 weeks after the 3rd session. Force, power and explosivity (force developed per sec) were recorded.

The intra-observer reproducibility was investigated of the 6 repetitions of the movements during each test session. The highest reproducibility for mean peak power and force was found in the combination of the 3rd, 4th, 5th and 6th measurement (intraclass correlation coefficient 0.85 to 0.99). Using this combination, the intra-observer reliability was investigated, defined as the consistency of results obtained during the 4 test sessions, performed at a weekly or longer interval. Comparison between these sessions showed that the results for mean force in the arm movement differed significantly (p=0.01) although a high between sessions correlation was found (0.96). Results for mean force and power for the 3 other movements were reliable, and no significant learning effect was observed. Lower reliability and reproducibility were observed for other parameters including explosivity and power and force at the first 0.25 sec of the test movement.

Even after standard warming-up movements, the first 2 out of 6 measurement repetitions during a test session are not reproducible and should not be used in calculations of the results. For the combination of the 3rd to 6th repetitions, a high reliability was found for the measurement of mean force and power of the total lifting movement and of the sitting leg extension movements repeated in 4 sessions over a period of 6 weeks.