Thigh–calf contact force measurements in deep knee flexion

Abstract

Background

Knee models often do not contain thigh–calf contact which occurs in deep knee flexion. Thigh–calf contact is expected to reduce muscle forces and thereby affects internal stresses in the knee joint. The purpose of this study was to measure thigh–calf contact forces. Two deep knee flexion activities were selected: squatting and kneeling.

Methods

Ten healthy subjects participated in the experiment. Contact pressures between the thigh and calf were measured using the Tekscan Conformat pressure mapping sensor. Knee flexion angles were measured unilaterally using an infrared motion capture system. Contact forces were averaged in terms of means and standard deviations. The magnitude and location of the resultant contact force were calculated. Correlations between anthropometric subject parameters and experimental outcome were studied.

Findings

In general, thigh–calf contact did not take place below 130° knee flexion. The average maximal contact forces for each leg were 34.2% bodyweight during squatting and 30.9% bodyweight during kneeling. Corresponding average maximal knee flexion angles were 151.8° during squatting and 156.4° during kneeling. Thigh and calf circumferences were correlated with the contact force measurements.

Interpretation

The current study shows that thigh–calf contact is substantial (>30% bodyweight on one leg) and likely reduces the forces inside the knee during deep knee flexion. Subsequently, total knee replacements may be subjected to lower loads than assumed before, which reduces the risk of implant failure at large flexion angles. Results presented in this study can be utilized in knee models that focus on deep knee flexion.

Introduction

Total knee replacement (TKR) is a widely used and successful surgical procedure. In 2003 a total of 418 000 TKR and 33 000 revision procedures were performed in the USA (National Hospital Discharge Survey, 2003). New developments are continuously made to improve implant performance. One of the latest developments is the so-called high-flexion knee prosthesis, which allows flexion angles larger than 120°. The development of this type of prosthesis is mainly a result of two concurrent trends. Firstly, due to its success, TKR is applied to younger and more active patients. The more active way of living demands a prosthesis with a larger range of motion. Secondly, the number of patients undergoing TKR surgery in non-Western countries is growing steadily. Studies on implants used in non-Western cultures report the necessity for high-flexion knee implants due to local daily living activities like kneeling and squatting (Mulholland and Wyss, 2001).

The development of high-flexion knee implants puts higher demands on implants, both in a kinetic and a kinematic perspective. Kinematic studies have been performed to investigate high-flexion knee behavior. As an example, several studies reported an asymmetric femoral roll-back mechanism during squatting and kneeling, which was caused by an internal tibial rotation (Hefzy et al., 1998, Conditt et al., 2006). Other studies reported the separation of the medial condyle and medial tibial plateau occurring at high knee flexion angles (Nakagawa et al., 2000, Conditt et al., 2006).

Over the last decades the finite element (FE) method has proven its value for TKR research and has recently been utilized in high-flexion knee research. Morra and Greenwald (2005) used FE models to calculate tibio-femoral contact stresses in different high-flexion knee implants. For some types of implants, they reported stresses above the yield point of the polyethylene insert at high knee flexion angles. Barink et al. (Submitted for publication) used FE models to compare a conventional knee prosthesis with a high-flexion knee prosthesis of the same manufacturer. In general, they found higher implant stresses with increasing flexion angles, which was primarily caused by higher quadriceps forces occurring at these higher flexion angles. They also showed that the high-flexion knee prosthesis did outperform the conventional prosthesis in the high-flexion range.

The outcome reliability of finite element analyses depends on input parameters such as joint forces applied to the finite element models. Joint forces are often estimated by simplified musculo-skeletal models using inverse dynamics. Most of these models do not include thigh–calf contact, which occurs in high knee flexion. Nagura et al. (2002) measured high knee flexion kinematics and calculated corresponding net knee joint forces and moments using an inverse dynamics model. This model contained estimations of segment dimensions and mass distributions. However, the model did not contain thigh–calf contact, which they reported as a shortcoming of their model. Caruntu et al. (2003) calculated knee joint forces during deep knee flexion using a mathematical model, which contained thigh–calf contact. With this model they showed that thigh–calf contact could lead to a considerable reduction of quadriceps and hamstring forces. Nonetheless, the model used in their study was a simplified and two-dimensional representation of the human knee and was not validated.

Because thigh–calf contact is often neglected, musculo-skeletal models would typically predict higher knee joint forces with increasing flexion angles, even in the high-flexion range. This seems to contradict the fact that people can squat for long periods of time in a relaxed manner, which is possibly caused by contact between the thigh and calf. Our hypothesis is that thigh–calf contact is substantial, reduces muscle forces in the knee during high knee flexion and should in that case not be neglected in models that focus on deep knee flexion.

Thus far, no prior studies were found which actually quantified the thigh–calf contact characteristics. The purpose of the current study was to gather information on thigh–calf contact by measuring its pressure distribution in relation to the knee flexion angle. Two high-flexion activities were included in this study: squatting and kneeling. The pressure distributions were used to calculate the magnitude and location of the resultant contact force on the calf, which can be used in further research. Furthermore, we hypothesized that anthropometric properties affect thigh–calf contact characteristics. Hence, we investigated whether we could detect any trends between body mass index related subject properties and the thigh–calf contact characteristics.