Three-dimensional knee joint contact forces during walking in unilateral transtibial amputees

Abstract

Individuals with unilateral transtibial amputations have greater prevalence of osteoarthritis in the intact knee joint relative to the residual leg and non-amputees, but the cause of this greater prevalence is unclear. The purpose of this study was to compare knee joint contact forces and the muscles contributing to these forces between amputees and non-amputees during walking using forward dynamics simulations. We predicted that the intact knee contact forces would be higher than those of the residual leg and non-amputees. In the axial and mediolateral directions, the intact and non-amputee legs had greater peak tibio-femoral contact forces and impulses relative to the residual leg. The peak axial contact force was greater in the intact leg relative to the non-amputee leg, but the stance phase impulse was greater in the non-amputee leg. The vasti and hamstrings muscles in early stance and gastrocnemius in late stance were the largest contributors to the joint contact forces in the non-amputee and intact legs. Through dynamic coupling, the soleus and gluteus medius also had large contributions, even though they do not span the knee joint. In the residual leg, the prosthesis had large contributions to the joint forces, similar to the soleus in the intact and non-amputee legs. These results identify the muscles that contribute to knee joint contact forces during transtibial amputee walking and suggest that the peak knee contact forces may be more important than the knee contact impulses in explaining the high prevalence of intact leg osteoarthritis.

Introduction

Individuals with unilateral transtibial amputations have altered gait mechanics and muscle coordination patterns relative to non-amputees (e.g., Fey et al., 2010, Silverman and Neptune, 2012), which may lead to the onset of joint disorders with prolonged use. For example, transtibial amputees have an increased prevalence and early onset of osteoarthritis (OA) and pain in the intact leg knee joint relative to the residual leg and non-amputees (Burke et al., 1978, Lemaire and Fisher, 1994, Melzer et al., 2001, Norvell et al., 2005, Struyf et al., 2009). However, the biomechanical mechanisms that contribute to the increased prevalence remain unclear.

The etiology of OA is not completely understood, but is partially attributed to increased and/or atypical joint loading (Maly, 2009, Morgenroth et al., 2012). Thus, the increased prevalence of OA in the intact knee of amputees may be a result of greater joint loading relative to the residual leg and non-amputees. Studies of amputee walking have shown elevated intact leg ground reaction forces (GRFs) and joint kinetics relative to the residual leg (Nolan et al., 2003, Royer and Koenig, 2005, Sanderson and Martin, 1997, Silverman et al., 2008) and non-amputee subjects (Nolan and Lees, 2000). Amputees also often have greater stance times on the intact leg relative to the residual leg (e.g., Isakov et al., 2000, Nolan et al., 2003), which may result in greater force impulses in the intact leg knee joint. Identifying differences in knee joint loading in amputees relative to non-amputees is important for understanding the potential biomechanical mechanisms that contribute to the high prevalence of OA in this population.

Recent work has investigated knee joint intersegmental forces across a range of walking speeds and found no significant differences between the intact and residual legs or between the intact and non-amputee legs (Fey and Neptune, 2012). However, inverse dynamics-based intersegmental forces often underestimate joint contact forces, as they do not account for the compressive forces from muscles (Zajac et al., 2002). In addition, altered muscle coordination patterns (Fey et al., 2010, Powers et al., 1998, Winter and Sienko, 1988) and increased co-contraction of the residual leg vasti and hamstring muscles (e.g., Culham et al., 1986, Isakov et al., 2001, Pinzur et al., 1991) in transtibial amputee walking likely influence the knee joint contact forces.

A significant challenge to investigating joint contact forces is the extreme difficulty, if not impossibility, of measuring them in vivo. In contrast, musculoskeletal modeling and simulation provide an ideal framework to estimate joint contact forces and quantify the contributions of individual muscles during dynamic movements (e.g., Sasaki and Neptune, 2010, Shelburne et al., 2006, Zajac et al., 2003). The gastrocnemius and soleus muscles have been shown to be large contributors to the knee joint contact force in non-amputee walking (Lin et al., 2010, Sasaki and Neptune, 2010, Shelburne et al., 2006), which may lead to asymmetric knee loading in unilateral transtibial amputees because they no longer have the functional use of these muscles in the residual leg.

The objective of this study was to investigate differences in knee joint contact forces in both the residual and intact legs relative to non-amputees during steady-state walking using three-dimensional musculoskeletal models and forward dynamics simulations. We expected that the peak knee contact forces and stance phase force impulses would be greater in the intact knee relative to the residual leg and non-amputees. In addition, individual muscle contributions to the knee contact forces were quantified to identify the potential biomechanical mechanisms that may contribute to the increased prevalence of OA in the intact knee.