Measuring knee extensor muscle strength

OBJECTIVE: To compare manual muscle test with hand-held dynamometer measurements of knee extension strength. A secondary analysis of measurements (n = 256 knees) from 128 acute rehabilitation patients was performed. DESIGN: Knee extensor muscle testing was conducted according to the technique of Hislop and Montgomery; 0 to 5 grades were converted to an expanded 0 to 12 scale. Dynamometry was used to measure the isometric knee extension force with 'gravity eliminated.' RESULTS: Manual muscle test and dynamometer measures were highly correlated (r = 0.768; P < 0.001); the correlation was higher when the quadratic nature of the relationship was taken into account (R = 0.887; P < 0.001). Although the dynamometer forces that were associated with different manual muscle test grades differed overall (F = 67.736; P < 0.001), the forces associated with some of the higher grades did not differ statistically. CONCLUSIONS: These findings reinforce the convergent construct validity of the manual muscle test and dynamometry measurements but challenge the discriminant construct validity of manual muscle testing. An alternative manual muscle testing grading scheme is suggested that provides for discriminant validity and retains convergent validity.