Dynamic knee loads during gait predict proximal tibial bone distribution

Abstract

This study tested the validity of the prediction of dynamic knee loads based on gait measurements. The relationship between the predicted loads at the knee and the distribution of bone between the medial and lateral sides of the tibia was examined. The motion and external forces and moments at the knee were measured during gait and a statically determinate muscle model was used to predict the corresponding forces on the medial and lateral tibial plateaus. In particular, the relationship between the knee adduction moment during gait and the ratio or distribution of medial to lateral tibial bone mineral content was studied. Bone mineral content was measured with dual energy X-ray absorptiometry in four regions, two proximal regions 20 mm in height, one medial and one lateral and two distal regions 20 mm in height, one medial and one lateral. The best single predictor of the medial–lateral ratio of proximal bone mineral content (bone distribution) was the adduction moment ($\text{R}{}^2\text{=0.31,}\text{p=0.003}$). Adding weight (negative coefficient, p=0.0004) and the ratio of the average predicted peak force on the medial plateau to the predicted peak force on the lateral plateau (positive coefficient, p=0.0033) to the regression model significantly increased the ability to predict the proximal medial–lateral bone distribution ($\text{R}{}^2\text{=0.72,}\text{p=0.0001}$). Distally neither the subject characteristics nor the gait moments and predicted forces were significant predictors of the bone distribution. The lack of a correlation distally may be reflective of the forces being more evenly distributed further from the tibial plateau. While it has long been suggested that the adduction moment is the primary determinate of the distribution of load between the medial and lateral plateaus, this is the first evidence of its relationship to the underlying bone distribution.

Introduction

Statically determinate muscle modelsshow that the adduction moment is a major determinate of not just the total load across the knee joint but its distribution between the medial and lateral plateaus as well (Schipplein and Andriacchi, 1991). Recent studies using statically determinate muscle models show that during walking approximately 70% of the total load typically passes through the medial compartment of the knee joint. In addition, an increase in the adduction moment during gait is predicted to increase the medial compartment loading at the knee joint. The prediction of higher forces on the medial side relative to the lateral side in normal subjects is consistent with the fact that the medial side of the tibia has bone that is significantly denser or stronger than that of the lateral side (Akamatsu et al., 1997; Behrnes et al., 1974; Harada et al., 1988; Petersen et al., 1996).

Although the adduction moment during gait is very reproducible with a mean difference between tests on separate days of only 0.1% body weight×height (80% confidence interval, 0.55% body weight×height), it does vary substantially between normal subjects (Andrews et al., 1996). Thus, if model predictions based on gait measurements represent true joint loads, it would be expected that variations in the adduction moment would be related to the bone mineral content on the medial side relative to the lateral side of the proximal tibia. Variations in the adduction moment should be associated with the distribution of bone between the medial and lateral compartments of the proximal tibia because in other situations, such as injury induced disuse (Kannus et al., 1994; Karlsson et al., 1993) and joint replacement (Bryan et al., 1996; Kiratli et al., 1996), changes in loads applied to the bone are associated with changes in bone mineral content or density. In addition, some studies have found a close association between bone strength and bone mineral density as measured by dual energy absorptiometry (Edmondston et al., 1994; Singer et al., 1995).

Previous studies have only related a few static factors likely to affect the load distribution (tibial femoral knee angle and meniscectomy) to the tibia bone (Akamatsu et al., 1997; Christensen et al., 1982; Hvid, 1988; Hvid and Hansen, 1985; Muller-Gerbl et al., 1993; Petersen et al., 1996) and have not related the actual predicted loads to the tibia bone. The purpose of this study was to test the validity of the dynamic force predictions from gait analysis and a statically determinate muscle model by testing the hypothesis that the distribution of bone mineral content in the proximal tibia was significantly correlated with the knee adduction moment as well as the predicted forces on the medial and lateral tibial plateau during gait. Bone distribution was quantified by the ratio of the bone mineral content on the medial side of the tibia relative to that of the lateral side.