Portable hand-held dynamometers (Manual Muscle Test System, Model 01163; Lafayette Instruments, Lafayette, IN, USA) were used to assess arm abduction, flexion and extension, hip flexion, knee extension, plantar flexion, and ankle dorsiflexion. Grip and pinch strength were measured bilaterally using a hydraulic dynamometer (Jamar®; Lafayette Instruments). Motor performances were tested in both arms and legs, using seven performance-based tests of leg function and two tests of motor performance of the arms, as previously described [9 (link)].
We scored each of the performance measures so that higher scores were associated with better performance. To ensure the same directionality in all the performance measures, we first reciprocated the recorded values for five variables (the time and number of steps for the walking and turning tasks, and the number of steps off the line for tandem walking). A score of 0 was recorded if a participant was unable to perform a particular task. Each score was then scaled by dividing by the sex-specific median value of the non-zero values at baseline [24 (link)-26 (link)]. The scaled scores for each measure were then averaged to obtain a composite global motor score for each subject. In previous publications from this cohort, measures were centered before scaling when calculating an alternative global motor score using z scores [9 (link)]. The correlation of our global motor score with the previously published score was high (r = 0.89, P < 0.001).
We scored each of the performance measures so that higher scores were associated with better performance. To ensure the same directionality in all the performance measures, we first reciprocated the recorded values for five variables (the time and number of steps for the walking and turning tasks, and the number of steps off the line for tandem walking). A score of 0 was recorded if a participant was unable to perform a particular task. Each score was then scaled by dividing by the sex-specific median value of the non-zero values at baseline [24 (link)-26 (link)]. The scaled scores for each measure were then averaged to obtain a composite global motor score for each subject. In previous publications from this cohort, measures were centered before scaling when calculating an alternative global motor score using z scores [9 (link)]. The correlation of our global motor score with the previously published score was high (r = 0.89, P < 0.001).
Full text: Click here