Madsen O R
Department of Rheumatology, Copenhagen Municipal Hospital, Denmark.
Spine (Phila Pa 1976). 1996 Dec 1;21(23):2770-6. doi: 10.1097/00007632-199612010-00012.
Isokinetic trunk extensor and flexor strength (peak torque, total work, power acceleration time, and torque acceleration energy) at 30 degrees/sec, 120 degrees/sec, and 180 degrees/sec and isometric extensor and flexor peak torque were measured in 24 healthy women. Measurements were performed in three separate sessions at the same time of day and by the same experienced examiner. The number of days between Session 1 and Session 2 and between Session 1 and Session 3 ranged from 1 to 7 (median, 2 days) and from 23 to 48 (median, 32 days), respectively. The test results were not revealed before all sessions were completed.
To determine the intrasession, short-term, and long-term reproducibility of several trunk strength variables measured by the Cybex 6000 isokinetic dynamometer (Lumex, Inc., Ronkonkoma, NY).
Only few and insufficient data regarding the reproducibility of trunk strength measurements exist. Previous studies have focused on coefficients of correlation to express reproducibility. Correlations, however, are misleading and inappropriate for this purpose. The critical difference is the difference between two results that would be statistically significant when applied to a reference group in steady state. The critical difference in percentage is calculated as 2.77 X coefficient of variation (%).
Isokinetic dynamometry (Cybex 6000). Reproducibility was expressed as pooled coefficients of variation, individual coefficients of variation, and critical differences.
Substantial between-subject variation of coefficients of variation was found for all strength variables; for example, the 80% central range for individual coefficients of variation for long-term reproducibility of extensor strength measurements at 30 degrees/sec ranged from 2.5% to 19.5%. Critical differences calculated using the pooled variances for long-term reproducibility were higher than 20% for all measures, mounting to 80% for acceleration time.
Strength testing is of limited value in the individual subject because of considerable day-to-day variability. This finding has important clinical implications.
对24名健康女性测量了其在30度/秒、120度/秒和180度/秒时的等速躯干伸肌和屈肌力量（峰值扭矩、总功、功率加速时间和扭矩加速能）以及等长伸肌和屈肌峰值扭矩。测量在一天中的同一时间由同一位经验丰富的检查者分三个不同时段进行。第1次和第2次测量之间以及第1次和第3次测量之间的天数分别为1至7天（中位数为2天）和23至48天（中位数为32天）。在所有测量时段完成之前，测试结果不予公布。
确定使用Cybex 6000等速测力计（Lumex公司，纽约州朗科纳科马）测量的几个躯干力量变量的测量时段内、短期和长期的可重复性。
关于躯干力量测量可重复性的数据很少且不充分。以往研究主要关注相关性系数来表示可重复性。然而，相关性用于此目的具有误导性且不合适。临界差异是指两个结果之间的差异，当应用于稳态参考组时该差异具有统计学意义。临界差异百分比的计算方法为2.77乘以变异系数（%）。
使用Cybex 6000进行等速测力。可重复性用合并变异系数、个体变异系数和临界差异来表示。
所有力量变量的变异系数在受试者之间存在显著差异；例如，30度/秒时伸肌力量测量的长期可重复性个体变异系数的80%中心范围为2.5%至19.5%。使用合并方差计算的长期可重复性临界差异对于所有测量指标均高于20%，加速时间的临界差异高达80%。
由于日常变异性较大，力量测试在个体受试者中的价值有限。这一发现具有重要的临床意义。
