Torque it up: reliability and variability of isokinetic peak torque and angle of peak torque

Introduction: Muscular performance is commonly tested using isokinetic dynamometry, with traditional protocols consisting of grouped repetitions and small rest periods between sets. Despite its popularity, current isokinetic testing assumes that the within-subject variation of angle of peak torque is low, and that most of the biological variability of peak torque and angle of peak torque is accounted for. Other assumptions of isokinetic testing include that the: (i) force output and behaviour of skeletal muscle is constant from one repetition to another, (ii) highest peak torque is achieved within five grouped repetitions, and (iii) torque-angle relationship is a static phenomenon and also represents sarcomere length. Previous and current isokinetic strength testing either ignores many factors of force production or assumes that they do not influence the biological variability and ultimately the within-subject variation of peak torque and angle of peak torque. Controlling for some factors of force production may yield a higher peak torque and a smaller Standard Error of Measurement (SEM) and Minimal Detectable Change (MDC) of peak torque and its angle, thus improving the reliability of such strength testing. The primary aim of this study was to determine and describe the within-session reliability of peak torque and angle of peak torque. This was tested using isokinetic concentric knee extensions at 60°/s across three different repetition protocols. Methods: Twenty four physically active male participants (23 ± 3 y, 77.9 ± 7.9 kg) performed 50 maximal concentric knee extensions of their dominant limb at 60°/s on a Biodex System II isokinetic dynamometer, in three different sessions 2 - 6 days apart. Prior to these sessions no warm up was performed. The three sessions consisted of: (i) ten sets of five grouped repetitions with a 120-s rest between sets (Gr-R), (ii) 50 interspersed repetitions with a 30-s rest between repetitions (IR-30), and (iii) 50 interspersed repetitions with a 60-s rest between repetitions (IR-60). Results: SEM (MDC) of peak torque over the first two repetitions were 12 (33) N.m, 11 (30) N.m and 13 (35) N.m for Gr-R, IR-30 and IR-60, respectively. SEM (MDC) of angle of peak torque over the first five repetitions was 4° (10°) for Gr-R, whereas over the first two repetitions it was 4° (11°) and 4° (10°) for IR-30 and IR-60 respectively. Mean peak torque across 50 repetitions was higher (p < 0.01) using IR-60 (243 N.m) and IR-30 (242 N.m) protocols than the Gr-R protocol (231 N.m). The obtained maximal peak torque of all 50 repetitions was higher compared to that of the first five repetitions, by 11%, 10%, and 7% for Gr-R, IR-30 and IR-60, respectively. 38% of participants achieved their highest peak torque between repetitions 11-20 and 1-10 for Gr-R and IR-60 respectively, and 33% of participants achieved their highest peak torque between repetitions 41-50 for IR-30. Conclusion: Only two grouped or interspersed repetitions of concentric knee extension tested isokinetically at 60°/s are required to obtain a reliable peak torque within a session. Whereas for angle of peak torque, five grouped and two interspersed repetitions of isokinetic knee extension are required to obtain a reliable measure. If the purpose of testing is to achieve the highest peak torque possible, then repetitions should be separated by a rest period (30-s or 60-s in the current study) rather than grouped in sets of consecutive repetitions. Individuals clearly differ as to when they achieve their highest peak torque, so more than the traditional five repetitions are necessary for most individuals. The transient increase of peak torque with repetitions, particularly for interspersed repetitions with a 30-s rest, may possibly be due to an augmentation of muscle properties or other muscular factors over time