The pendulum test as a tool to evaluate passive knee stiffness and viscosity of patients with rheumatoid arthritis

Abstract

Background: The pendulum test of Wartenberg is a technique commonly used to measure passive knee motion with the aim to assess spasticity. We used this test to evaluate changes of the knee angular displacement, passive stiffness and viscosity in rheumatoid arthritis patients. Stiffness and viscosity represent passive resistances to joint motion associated with the structural properties of the joint tissue and of the muscular-tendon complex. Stiffness can be considered an intrinsic property of the tissues to resist deformation, while viscosity is related to cohesive forces between adjacent layers of tissues. Both parameters may influence the joint range of motion affecting angular displacement.

Methods: Nine women with rheumatoid arthritis were compared with a group of healthy women. With the subject half-lying, the relaxed knee was dropped from near-full extension and the characteristics of the ensuring damped unsustained knee oscillation evaluated. The kinematics of leg oscillations was recorded using ultrasonic markers (Zebris CMS HS 10) and the kinetic data were calculated from kinematic and anthropometric measures.

Results: Knee stiffness significantly increased (p < 0.001) in patients with respect to the control group, while differences in viscosity were not significant. Moreover, the amplitudes of first knee flexion (the maximal flexion excursion after knee release) and first knee extension (the maximal extension excursion after the first knee flexion) were significantly decreased (p < 0.001). A regression analysis showed that disease severity correlated moderately with stiffness (R2 = 0.68) and first flexion (R2 = 0.78). Using a multivariate regression, we found that increasing stiffness was the main factor for the reduction of flexion and extension motions.

Conclusion: We showed that the Wartenberg test can be considered a practical tool to measure mechanical changes of knee caused by rheumatoid arthritis. This novel application of Wartenberg test could be useful to follow up the effects of pharmacological and rehabilitative interventions in this disease.

Figure 1

Limb oscillation during pendulum test. Solid line represents the leg at starting position (extended leg) and at final position (flexed leg). The numbers indicate location of skin reference markers: 1, 2/3 thigh; 2, lateral femoral condyle; 3, head of fibula; 4, lateral malleolus.

Figure 2

A. Typical normal knee flexion-extension angular response showing onset angle, resting angle, the first three peak flexion angles (F1, F2, F3), the first three peak extension angles (E1, E2, E3), the period of the first cycle (T); B. pendulum test angular response for a patient mildly affected (class I, ACR classification); C: pendulum test angular response for a patient severely affected (class III, ACR classification). Abbr.: ACR (American College of Rheumatology)

Figure 3

Mean values and standard errors of stiffness (A) and viscosity (B) computed for RA patients (thick line) and control group (fine line) during the first six half-cycles. KF = stiffness in flexion; KE = stiffness in extension; BF = viscosity in flexion; BE = viscosity in extension.

Figure 4

Mean values and standard errors of F1amp (amplitude of first flexion), E1amp (amplitude of first extension), RI (relaxation index) and ERI (extension relaxation index). * = p < 0.05, ** = p < 0.01.

Figure 5

Results of univariate linear analyses. The normal condition (N) and the pathology severity (ACR classification, [6]) were plotted against stiffness (A), viscosity (B), F1amp (C) and E1amp (D). R² = coefficient of determination; N = normal subjects.

Figure 6

Multivariate linear analysis to evaluate (A) the relationship between F1amp (amplitude of first flexion) and BF1 and KF1 (viscosity and stiffness during first flexion, respectively); (B) the relationship between E1amp (amplitude of first extension) and BE1 and KE1 (viscosity and stiffness during first extension, respectively). Regression model, represented by the surface grid, is superimposed on the observed data.