DO THE HAMSTRINGS AND ADDUCTORS CAUSE EXCESSIVE INTERNAL ROTATION OF THE HIP IN PERSONS WITH CEREBRAL PALSY?

Presented at NACOB 98:
North American Congress on Biomechanics
Canadian Society for Biomechanics - American Society of Biomechanics

University of Waterloo
Waterloo, Ontario, Canada
August 14-18, 1998

DO THE HAMSTRINGS AND ADDUCTORS CAUSE EXCESSIVE INTERNAL ROTATION
OF THE HIP IN PERSONS WITH CEREBRAL PALSY?

Allison S. Arnold and Scott L. Delp
Departments of Biomedical Engineering and Rehabilitation Medicine, Northwestern University
Sensory Motor Performance Program, Rehabilitation Institute of Chicago
345 E. Superior Street, Chicago, IL 60611

INTRODUCTION

Persons with cerebral palsy frequently walk with excessive internal rotation of the hip. Spastic hamstrings or adductors are often thought to be the cause of the excessive internal rotation based on evidence from electromyographic (EMG) recordings (e.g., Sutherland et al. 1969). For this reason, surgical lengthenings of the medial hamstrings and adductors are commonly performed in an attempt to decrease the internal rotation moment that is presumed to cause this abnormality. Whether these procedures are warranted, however, remains unclear. The surgical outcomes are unpredictable, the EMG evidence is not conclusive, and the rotational moment arms of the muscles during walking have not been adequately investigated. Hence, the capacity of the hamstrings and adductors to contribute to internal rotation is not known. Determination of hip rotation moment arms in cerebral palsy patients is difficult because rotational abnormalities of the hip are often accompanied by torsional deformities of the femur, which may alter the muscle lines of action. Furthermore, the moment arms must be evaluated over the range of body positions assumed by patients during walking, which frequently includes exaggerated flexion of the hips and knees in addition to increased internal rotation of the hip. We have developed a three-dimensional (3D) computer model that evaluates the moment arms of the hip muscles for a range of femoral deformities and limb positions. Analysis of this model revealed that the hamstrings and adductors do not have internal rotation moment arms for body positions and bone geometries typical of patients with rotational abnormalities, suggesting that other factors are more likely to cause the excessive internal rotation in these patients.

REVIEW AND THEORY

Persistent internal rotation of the hip is a troublesome problem that frequently accompanies crouch gait in persons with cerebral palsy. Excessive internal rotation causes tripping and may lead to lateral patellar subluxation or external tibial torsion. Hence, surgeries aimed at decreasing the excessive internal rotation are often performed.

Abnormal internal rotation of the hip is usually accompanied by excessive anteversion of the femur. Anteversion is defined as the angle (a) between the plane of the femoral neck axis (NA) and the plane of the condylar axis (CA, Fig. 1). This angle is generally 10°-20° in non-impaired adults (Fabry et al. 1973), but may be as large as 60° in persons with cerebral palsy (Ruwe et al. 1992, Laplaza et al. 1993).

We have developed a musculoskeletal model with a deformable femur to investigate whether the medial hamstrings or adductors are likely to cause hip internal rotation in patients with anteversion deformities who walk with a crouched, in-toed gait. Few investigators have attempted to quantify the rotational moment arms of these muscles (Maxwell et al. 1983, Dostal et al. 1986), and none have reported how the moment arms vary with hip flexion, knee flexion, or with femoral anteversion. We focused on the semimembranosus, semitendinosus, adductor longus, adductor brevis, and gracilis muscles in this study because these are the muscles most commonly lengthened in surgeries to treat excessive internal rotation of the hip.

Figure 1: Definition of anteversion (a). H is the center of the femoral head, O is the center of the femoral neck, P is the attachment of the posterior cruciate ligament, and L and M are the posterior aspects of the lateral and medial condyles. Figure adapted from Murphy (1987).

PROCEDURES

Our computer model characterizes the 3D geometry of the bones, the kinematics of the hip and knee, and the paths of the hamstrings and adductors for a nominal adult-sized subject (Delp et al. 1990). Anteversion deformities were represented in the model (Arnold et al. 1997) by rotating the bone vertices that make up the femoral head and neck about the femoral shaft axis (FA, Fig. 1). The position of the hip center was not changed. Muscle insertions were displaced with the bone vertices; hence, the moment arms of the muscles were potentially altered as a result of the deformities. Our undeformed model has an anteversion angle of 20°.

The rotational moment arms of the muscles were examined for a range of hip flexion (0° to 90°), knee flexion (0° to 90°), and femoral anteversion (0 ° to 60°). In addition, 3D joint kinematics obtained from gait analysis were used in connection with the model to estimate the muscle moment arms for combinations of joint angles that correspond to walking.

Eighteen non-impaired subjects and one patient with cerebral palsy underwent gait analysis. The patient was 26 years old, had no previous orthopaedic surgery, and walked with severe flexion and internal rotation of his right hip. We used our undeformed model and the averaged gait kinematics of the non-impaired subjects to estimate the rotational moment arms of the muscles at limb positions corresponding to normal walking. We used our deformed model with 47° of anteversion to estimate the muscle moment arms at joint angles corresponding to the patient's gait. The anteversion angle of the patient's femur was accurately determined by constructing a 3D surface model of the bone from magnetic resonance (MR) images.

RESULTS AND DISCUSSION

In the anatomical position and at joint angles corresponding to normal walking, the adductor longus, adductor brevis, and medial hamstrings in our model have hip rotation moment arms that are very small or slightly internal (Figs. 2A and 3A). The gracilis in our model has a hip rotation moment arm that is slightly external. These results are consistent with EMG data and with moment arms published in the literature (Basmajian et al. 1985, Dostal et al. 1986). When the femur is anteverted more than 40°, however, or the hip is excessively flexed, the adductor longus and adductor brevis in our model have moment arms that are much less internal (e.g., Fig. 2B). Similarly, when the hip is internally rotated more than about 20°, the moment arms of the medial hamstrings in our model switch from internal to external (Fig. 2A). Flexion of the knee also causes the moment arms of the hamstrings and gracilis to become more external (not shown). When all of these effects are combined, the hamstrings and adductors in our model do not have internal rotation moment arms for anteversion angles and body positions commonly observed in persons with cerebral palsy who walk with excessive internal rotation of the hip (Fig. 3B). Currently we are in the process of verifying these findings by (1) measuring hip rotation moment arms in anatomical specimens, and (2) developing subject-specific models of patients in which the bone geometry and the muscle paths are derived from MR images.

Our results suggest that surgical lengthening of the medial hamstrings and adductors in an attempt to reduce the excessive internal rotation, in some patients, may be inappropriate. This study emphasizes the need to account for altered bone geometry and abnormal joint kinematics when hypothesizing the causes of movement abnormalities.

Figure 2: Hip rotation moment arm vs. hip rotation angle for the adductor brevis (AB), adductor longus (AL), semimembranosus (SM), semitendinosus (ST), and gracilis (GRA) muscles with 20° of anteversion (A), and for the adductor brevis with 20°, 40° and 60° of anteversion (B). Hip and knee flexion are 0°.

Figure 3: Hip rotation moment arm vs. gait cycle for the adductor brevis (AB), adductor longus (AL), semimembranosus (SM), semitendinosus (ST), and gracilis (GRA) muscles during normal walking (A), and for a cerebral palsy patient who walks with excessive flexion and internal rotation of the hip (B).

REFERENCES

Arnold et al. Dev. Med. Child Neuro., 39, 40-44, 1997.

Basmajian et al. Muscles Alive, 319-324, Williams & Wilkins, 1985.

Delp et al. IEEE Trans. Biomed. Eng., 37, 757-767, 1990.

Dostal et al. Phys. Ther., 66, 351-361, 1986.

Fabry et al. J. Bone Jnt. Surg., 55A, 1726-1738, 1973.

Laplaza et al. J. Ped. Orthop., 13, 192-199, 1993.

Maxwell et al. J. Biomed. Eng., 5, 253-256, 1983.

Murphy et al. J. Bone Jnt. Surg., 69A, 1169-1176, 1987.

Ruwe et al. J. Bone Jnt. Surg., 74A, 820-830, 1992.

Sutherland et al. J. Bone Jnt. Surg., 51A, 1070-1082, 1969.

ACKNOWLEDGMENTS

Special thanks to JoAnn Mason and the staff of the Gait Lab at the Children`s Memorial Medical Center in Chicago for assistance with the data collection. This study was funded by NIH grant #R01 HD33929 and the United Cerebral Palsy Foundation.