By X. C. Liu, MD, PhD, J. Thometz, MD, G. Fabry, MD, PhD, G. Molenaers, MD, J. Lammens, MD, and P. Moens, MD
The literature provides numerous studies on leg length discrepancy (LLD). Clinical examinations, indications for treatment, and biomechanical influences on the long and short legs and the low back have all been investigated.1-7
Various authors1,6,8,9 agree that LLD less than 2 to 2.5 cm in children can be left untreated. Amstutz et al1 stated that minor discrepancies (between 0 and 3 cm) in patients with normal stature do not need surgical intervention. Others have suggested that differences of 1 to 2.5 cm are rarely symptomatic and are acceptable with or without a shoe lift.2 These opinions have been based on symptoms, cosmesis, or clinical experience.
Published biomechanical studies include measurements of the kinematics of the pelvis, hip, knee, ankle, and lower back3,4 and force-plate measurements for detecting changes in the center of pressure (COP) in postural sway or in ground reaction forces.7,10 A study has also been conducted evaluating gait in patients with LLD secondary to complete congenital dislocation.5 Few studies have been published on the effect of different magnitudes of LLD on the mechanics of the lower limbs during gait.
We addressed three questions:
We tested 30 subjects (20 male, 10 female) with a mean age of 14.6 years. The subjects had inequalities of the lower limbs ranging from 0.9 to 4.7 cm. All had a normal range of motion in the joints of the lower limbs. Patients with neuromuscular disease were excluded.
According to their etiology, the patients could be classified into three categories: idiopathic LLD (15 cases); congenital growth disorders such as proximal femoral focal deficiency (PFFD), multiple epiphysial dysplasia, hypertrophy of the femur or the tibia, or Klippel-Trenaunay syndrome (10 cases); and miscellaneous cases such as femur or tibia fracture, Legg-Calvé-Perthes disease, and a tumor (five total). Patients were divided into three groups depending on the degree of inequality as measured radiographically (see Table 1).
The amount of correction to be provided by a heel lift was determined through a so-called indirect clinical method (also called the block test): We stood behind patients and placed an index finger on the top of each of their iliac crests, from which a discrepancy of the lower extremity was determined by measuring the distance between levels of our two fingers. Then we inserted a block under the heel of the shorter leg until the pelvis was leveled, which means that the iliac crest heights were symmetric. The heights of the heel lifts selected ranged from 0.5 to 4.5 cm.
Controversy over the amount of LLD correction remains. Some physicians prefer a complete correction of LLD (< 0.5 cm), others think it is better to leave 1 to 2 cm of LLD. So patients in our study were organized into two subgroups by the degree of heel-lift correction.
Subgroup 1 consisted of patients from groups I, II, and III who have a near-complete correction. Subgroup 2 had patients from groups II and III who have a partial correction. For subgroup 1, the level of the top iliac crest ranged from 0.2 to 0.9 cm with a mean value of 0.51 cm after correction with the heel block. In subgroup 2 , the range was from 1.1 to 1.8 cm with a mean of 1.39 cm (see Table 1).
Thirteen reflective markers were attached to anatomical landmarks including the inferior anterior iliac spine, greater trochanter, midpoint of the lateral femoral condyle, lateral malleolus, head of the fifth metatarsal, and middle of the posterior superior iliac spine. The patients were allowed to walk at a self-selected speed. Prior to and following heel lift in the short leg, motions in the joints of the lower extremity were recorded as patients walked on the 10-m-long level walkway. The time for recording was four seconds each using a Mac Reflex motion analysis system with four cameras, video process, and Macintosh computer operating at 50 Hz. The cameras were set up to capture the front, back, and both sides of the subject.
The motion analysis system recorded joint motion in the ankle, knee, and hip in the sagittal plane. Parameters H1 and H3 refer to peak hip flexion at heel contact and toe-off, respectively. H2 is the minimum flexion at midstance.
K1 and K4 refer to knee flexion at heel contact and toe-off, respectively. K2 is the peak flexion of the knee at midstance, while K3 is the minimum flexion at midstance.
A1 indicates the peak plantar flexion from heel contact to foot flat. A2 is the maximum dorsiflexion during midstance. A3 is the maximum plantar flexion at toe-off.
When we compared joint motions between three groups that did not receive heel lifts, significant differences were found at the hip (H1L, H2S, H3S) and knee (K1S), but not at the ankle when comparing group I with group II. Comparing groups I and III revealed significant differences only in ankle-joint flexion (A2S and A3S). No significant differences between group II and group III were noted (p > 0.05) (see Table 2), but results in group III showed slight increases in long-leg flexion and short-leg extension compared to group II.
There were no significant differences in joint motion between the long and the short legs in group I, while hip flexion at toe-off (H3) in the long leg was more pronounced in group II (H3L = 26.29° ± 4.49°, H3S = 20.36° ± 7.23°; p = 0.005). Dorsiflexion of the ankle at midstance in the long leg (A2L = 11.6° ± 4.92°) was significantly higher than its value in the short leg (A2S = 6.34° ± 3.77°) in group III (p = 0.018).
A heel lift was associated with less knee motion at initial contact (K1L) in subgroup 1 (7.81° ± 4.47°) than in all patients with uncorrected LLD (10.2° ± 6.82°), and with less knee motion at midstance (K3L) in subgroup 2 (7.58° ± 3.68°) compared to the patients with uncorrected LLD (10.9° ± 6.19°) from groups II and III. There were no significant changes in pelvic-, hip-, and ankle-joint flexions following the heel lift in either of the subgroups.
Although another study5 provided no information on how the amount of LLD affected gait patterns, it did show that LLD accounts for a short step length, lower pelvis, lateral shift of ground reaction force, and decreased maximum adduction moment of the hip and knee. However, our study measured joint motions of the lower extremity related to the different amounts of LLD.
From our results, it is clear that a number of compensatory mechanisms play a role in the gait of patients with LLD. The mechanism of compensation depends on the amount of LLD. As long as the difference remains below a mean of 2.64 cm (group II), the compensation is located in the hip and knee joints. Flexion in the hip joint at heel strike is augmented on the long side, while it is reduced in the short leg at midstance and toe-off, in combination with increased knee flexion at heel strike. Both these mechanisms aim to minimize pelvic tilt. Meanwhile, this reduced hip flexion (H2 and H3) and functional lengthening of the short leg provide necessary clearance for the contralateral (long) leg to swing through.
Once the leg length discrepancy reaches about 4.28 cm, the aforementioned compensations in the hip and knee joints diminish and the ankle joint compensates. Dorsiflexion during midstance is appreciably diminished, while plantar flexion at toe-off is dramatically increased. This again results in a functional lengthening of the short leg allowing for clearance of the long leg.
From the changes in kinematics (Table 2), it is impossible to decide on the ideal acceptable amount of leg-length discrepancy (LLD) because each group demonstrates significant biomechanical effects on different joints. We may say that with LLD ranging from a mean LLD of 2.6 to 4.2 cm (groups II and III), joint compensation can occur in attempts to reach symmetric gait pattern. As a result of symmetric gait patterns, the hip joint on the long side provides more flexion, the knee joint on the short side increases extension, and the ankle joint on the short side increases plantar flexion.
Using a heel lift to correct inequality to a mean LLD of 0.51 cm results in a reduction of knee flexion in the long leg, producing a relative functional shortening of that leg. Using a heel lift that leaves a mean LLD of 1.39 cm, the functional shortening of the long leg during stance still exists, but at different periods of stance phase. However, the heel lift did not affect the pelvis, the hip joint, or the ankle joint, which may mean the heel lift causes a forward inclination of the body and easily moves forward a vertical ground reaction force in front of the knee to reduce the flexion moment. Our kinematic results after a heel lift can’t predict precise amounts to be corrected.
Xue C. Liu, MD, PhD, and John Thometz, MD, are in the division of pediatric orthopedic surgery at Children’s Hospital of Wisconsin in Milwaukee. Guy Fabry, MD, PhD, Guy Molenaers, MD, John Lammen, MD, and Peear Moens, MD, are in the department of orthopedic surgery at the University of Pellenberg Hospital, Katholic University of Leuven, Leuven, Belgium
CancerNetwork | ConsultantLive | Diagnostic Imaging | Psychiatric Times | Applied Neurology
Copyright © 2008 CMP Healthcare Media Group LLC, a United Business Media company