Reliability of lumbar dynamometry measurements in patients with chronic low back pain with test-retest measurements on different days.

Lumbar dynamometry is a potentially useful method for assessing the state of trunk muscles in low back pain (LBP) patients. The purpose of this study was to assess the reliability of lumbar dynamometry measurements in chronic LBP patients by conducting test-retest measurements on different days. Thirty-one men and 14 women with chronic LBP participated in this study. The experiments consisted of three sets of lumbar dynamometry measurements (Isostation B200) carried out on three different days with a 2- to 3-day interval. A standard protocol was administered to all subjects, consisting of a range-of-motion measurement about each axis, a 5 s maximum isometric trial about each axis and five dynamic repetitions about each axis against a resistance set at 25% and at 50% of the maximum isometric torque. Correlation coefficients and regression analysis were used to detect possible learning effects. One-way anova and regression analysis were used to assess the reliability of the measurements. High coefficients were found for the correlation between the first and second lumbar dynamometry measurements. Regression analysis showed that the differences between those measurements were not significant. This means that there was no learning effect operating between the first and second lumbar dynamometry measurements. One-way anova showed a reliability higher than 0.90 for the torque and velocity parameters. Reliability for the range-of-motion parameters was somewhat lower: between 0.76 and 0.94. Regression analysis showed no significant differences between the second and third measurements for the torque and velocity parameters. For range-of-motion parameters significant differences were found. From this study it can be concluded that the Isostation B200 provides reliable measures of torque and velocity parameters, but measures of the range-of-motion parameters are unreliable. No learning effect operates between the first and second lumbar dynamometry measurements, which means that a single measurement, with prior warming up and practice, is sufficient to assess the performance of the LBP patient.