Influence of muscle forces on femoral strain distribution

Abstract

Musculoskeletal loading influences the stresses and strains within the human femur and thereby affects the processes of bone modeling and remodeling. It is essential for implant design and simulations of bone modeling processes to identify locally high or low strain values which may lead to bone resorption and thereby affect the clinical outcome. Using a finite element model the stresses and strains of a femur with all thigh muscle and joint contact forces were calculated for four phases of a gait cycle. Reduced load sets with only a few major muscles were analyzed alternatively. In a completely balanced femur with all thigh muscles the stress and strain patterns are characterized by combined bending and torsion throughout the bone. Similar to in vivo recordings, the model with all thigh muscles showed peak surface strains below 2000 με (45% gait cycle). Under simplified load regimes surface strains reached values close to 3000 με. Within the proximal femur, the simplified load regimes produced differences in strain as high as 26% in comparison to those with all thigh muscles included. This difference is reduced to 5% if the adductors are added to a loading consisting of hip contact, abductors and ilio-tibial band. This study demonstrates the importance of an ensemble of muscle forces to reproduce a physiological strain distribution in the femur. Analytical attempts to simulate bone modeling, remodeling or bone density distributions should therefore rely on fully balanced external load regimes which account for the role of the various soft tissue forces.

Introduction

Bone in healthy subjects adapts to its mechanical environment (Wolff, 1892). In some patients with endoprosthetic joint replacements or fracture fixation devices, local strains and stresses may exceed biological limits (Frost, 1964; Cowin and Hart, 1985), leading to bone resorption or remodeling and possible implant loosening (Harrigan et al., 1996; van Rietbergen et al., 1997). Various mathematical methods have been employed to describe these adaptation processes (Weinans et al., 1992; Turner et al., 1997). Thus, in order to achieve the physiological relevance of such simulations, it is essential to utilize realistic strain distributions which correlate to those reported from in vivo measurements (Lanyon and Smith, 1969; Cochran, 1972; Brennwald and Perren, 1974; Lanyon, 1976; Schatzker et al., 1980; Carter et al., 1981; Weinans et al., 1992).

In most finite element analyses of the human femur, the physiological loading is approximated using the abductor muscles and the ilio-tibial band (Rybicki et al., 1972; Crowninshield et al., 1980; Huiskes et al., 1987; Huiskes, 1990; Merz et al., 1992; Lu et al., 1996). The particular importance of the ilio-tibial tract and the abductors to the femoral loading condition were described by Pauwels (1951)and later confirmed in a continuum mechanics’ approach (Rohlmann et al., 1980, Rohlmann et al., 1982). Given the relative contribution of muscle activity to the loading situation (Finlay et al., 1991; Duda et al., 1997) it may be important to consider muscle activity other than the abductors and ilio-tibial band (Taylor et al., 1996). To the authors’ knowledge, no study has used a complete and balanced set of thigh muscle and joint contact loads to approximate a more comprehensive loading situation in the femur.

Thus, the goal of this study was to determine the strain distribution during gait based on a loading situation including all thigh muscles, to compare the resulting strain distributions with those obtained using simplified load regimes and to determine which muscle forces should be included in analytical investigations in order to appropriately simulate the loading conditions for the proximal femur with maximal physiological relevance.