Abstract

The variability of voluntary isometric strength measurements has been assessed in normal subjects and patients with peripheral neuromuscular disorders. Knee extensor strength was measured in a muscle testing chair 13 times over 5 months in each of six normal subjects: coefficients of variation (CV) ranged from 4.5 to 14.0% (mean 8.5%) for individual legs in different subjects. Paired measurements of the strength of several clinically weak muscle groups were made 1-4 days apart in 20 patients using both a handheld dynamometer and the muscle chair technique: the test/retest correlation was high (r = 0.97, p less than 0.001). Visual biofeedback did not affect the strength recorded in most cases. Each of five patients had the strength of six or seven clinically weak muscle groups measured by five examiners within a 24 hour period: the CV for the five examiners ranged from 3.6-27.3% (mean 12.8%). A single examiner measuring the same groups on five occasions in three patients obtained a mean CV of 8.9%. Sources of variation are analysed and it is concluded that, with certain precautions, voluntary strength measurements offer a simple, reliable and acceptable method for monitoring change in patients.