Passive and dynamic rotation of the lower limbs in children with diplegic cerebral palsy

Transcription

1 Passive and dynamic rotation of the lower limbs in children with diplegic cerebral palsy Mariëtta L van der Linden* PhD; M Elizabeth Hazlewood MCSP; Susan J Hillman MSc CEng, Anderson Gait Analysis Laboratory, Eastern General Hospital; James E Robb MD FRCS, Royal Hospital for Sick Children, Edinburgh, UK. *Correspondence to first author at Anderson Gait Analysis Laboratory, Rehabilitation Engineering Services, Eastern General Hospital, EH6 7LN Edinburgh, Scotland, UK. MVandDerLinden@qmuc.ac.uk Rotation characteristics in gait and passive rotation of the lower limbs were evaluated retrospectively in 105 patients with diplegic cerebral palsy (65 males, 40 females; mean age 13y [SD 6y 9mo]; range 4y 4mo 40y 5mo). Of 105 patients, 22 (20.9%) required crutches, sticks, tripods, or a K-walker for their daily ambulation. Twelve (11.5%) patients used a wheelchair or buggy for community distances, e.g. shopping. Significant differences in rotational characteristics were found at the pelvis, hip, knee, and foot between left and right legs. Patients who were more affected on the right (group R, n=33) or the left side (group L, n=39) were re-evaluated. There was also a group of patients who were not asymmetrically affected (group S, n=33). In group L, maximum passive internal rotation was significantly greater on the left side, while no difference between the sides was found in group R. Peak internal rotation in gait was significantly higher on the right side in group R, but did not differ significantly between the sides in group L. Right hindfoot thigh angle and transmalleolar axis were more external on the right, irrespective of which leg was more affected. These findings may have implications for the early non-operative management of limb posture in infants with diplegia and the surgical management of established lower extremity malrotation. Children with cerebral palsy (CP) often walk with excessive internal rotation of one or both hips (Gage 1991). The exact mechanisms for this excessive internal rotation are still not well understood. Majestro and Frost (1971) identified two separate causal elements: dynamic and structural. They stated that the primary cause of femoral internal torsion is dynamic, as it arises from (abnormal) muscular contraction. The structural element, persistent femoral anteversion (Fabry et al. 1973), arises as a secondary consequence of the dynamic element. Brunner et al. (2000) identified two components of internal rotation of the hip: (1) a functional part, which can be corrected by weakening and lengthening the muscles responsible for increased internal rotation of the hip; and (2) a structural component, persistent femoral anteversion. Aktas et al. (2000) stated that muscle imbalance causes persistence of femoral deformity, which may contribute to rotational asymmetry. The position of the lower limbs before and after birth also influences hip rotation characteristics; there is some evidence that asymmetries can develop between the left and right leg as a result of incorrect positioning of the lower limbs. For example, Tachdajian (1990) stated that of all congenital dislocations of the hip, the left hip is dislocated at birth three times as much as the right hip (60% are on the left, 20% are on the right, and 20% is bilateral). The author explains this greater frequency of left-side involvement by stating that the most common intrauterine position is with the left leg adducted against the mother s sacrum. There is evidence that the postnatal position of the infant may be of importance. Fulford and Brown (1976) examined 20 children with CP (10 diplegia: one had superimposed right hemiplegia; two right hemiplegia; five bilateral hemiplegia; one dyskinesia; and one ataxia). Nineteen of the 20 children had windswept legs with asymmetrical hip position. The majority (n=13) was more windswept to the right, which is associated with a more adducted left leg. In a study of hip dislocation in 19 non-ambulant children with severe dystonic CP, Owers et al. (2001) reported that five were windswept to the left and 14 to the right. The aim of this retrospective study was to investigate the passive and dynamic rotation characteristics of the left and right leg separately to gain more insight into factors that might contribute to rotational abnormalities in the lower limbs of children with CP. Method PARTICIPANTS Kinematic and physical examination data collected between 1999 and 2002 during routine clinical gait analysis at the Anderson Gait Analysis Laboratory in Edinburgh, UK were studied retrospectively. Participants included 105 patients (65 males, 40 females; mean age 13y [SD 6y 9mo]; range 4y 4mo 40y 5mo) with diplegia. Twelve patients (11.4%) were older than 17 years. At the time of the clinical gait analysis, all patients or their parents had signed a form stating that they allowed their data to be used for research purposes. This retrospective study was approved by the Lothian Research Ethics Committee. Patients who had undergone derotation osteotomy of the femur or tibia were excluded. Thirty-four of the 105 patients had undergone other surgery 12 months or more preceding gait analysis. Surgical interventions included hamstring and gastrocnemius lengthening, rectus tendon transfer, adductor 176 Developmental Medicine & Child Neurology 2006, 48:

2 and psoas releases, and anterior and posterior tibial muscle transfer. Seven children had received botulinum toxin in the preceding 6 months. After initial analysis, patients were divided into three groups. Group L consisted of participants who were clearly more affected on the left side and group R consisted of participants who were clearly more affected on the right. Participants in group S were not clearly asymmetrically affected. Determining whether one side was clearly more affected than the other was based on information from data derived during the patient s gait analysis appointment. This included clinical diagnosis; information from the Gillette Functional Assessment Questionnaire (Novacheck et al. 2000), in particular, whether the patient could hop on either foot or only the left or the right foot; the presence of contractures; muscle strength; tone in the muscles of the lower limbs; leg length; and the Edinburgh Gait Score (Read et al. 2003). The Edinburgh Gait Score is a scoring system for use in individuals with CP. A score of 0 indicates a normal gait and the theoretical maximum score is 34. Left and right sides were scored separately. Of the 105 participants, 39 were more affected on the left side (group L) and 33 were more affected on the right side (group R); 33 were not asymmetrically affected (group S). Table I shows that all three groups were comparable with respect to age, Edinburgh Gait Score of the most affected leg, and number of preterm births: whether children are born at term may be of importance, as there is some evidence that otherwise healthy children born preterm are at higher risk of pronounced asymmetry (Konishi et al. 1987). KINEMATIC DATA Kinematic data were derived using a six-camera Vicon 512 camera system (Oxford Metrics, Oxford, UK). The marker set used was described in the Vicon Clinical Manager (VCM) manual. A knee alignment device was used to derive the threedimensional (3D) orientation of the knee flexion extension axis during the static trial. VCM software was used to derive 3D joint angles. For each participant, the average rotation data of at least three trials were used for analysis. Transverse plane hip motion data are very sensitive to the definition of the embedded coordination system of the thigh. A small change in the direction of the transverse axis through the femoral condyles can lead to a considerable offset of hip rotation angle (Kadaba et al. 1990). For this reason, the hip rotation data of any limb for which the range of knee ab/adduction exceeded 15 (averaged over at least three trials) was discarded from the analysis. It was considered that a range of more than 15 was a result of an inaccurate definition of the knee axis and not a real representation of the knee motion in the frontal plane. As a result, the hip rotation data of 83 right legs and 84 left legs were included from the 105 patients. For group R the hip rotation data for 32 right legs and 25 left legs were included, for group L the data for 23 right legs and 37 left legs, and for group S, 28 right legs and 22 left legs were included. Peak hip rotation in stance was used for analysis, as this parameter has been shown to correlate best with passive measures of hip rotation (Kerr et al. 2003). Mean pelvic rotation was derived by averaging the values over the whole gait cycle. Protraction of the ipsilateral side of the pelvis was defined as Table I: Mean (SD) of Edinburgh Gait Score (Read et al. 2003) Group L, n=39 Group R, n=33 Group S, n=33 Right Left Right Left Right Left Edinburgh Gait Score, mean (SD) 8.7 (5.2) 11.2 (5.2) a 11.8(5.3) 8.6 (4.3) a 11.8 (4.8) 11.4 (4.8) Age, mean (SD) 13y 8mo (7y 11mo) 12y 6mo (6y 5mo) 12y 4mo (6y 2mo) Preterm, n (%) a Significantly different (p<0.001) between right and left. Group L, more affected on left; group R, more affected on right; group S, not clearly asymmetrically affected. Edinburgh Gait Scores range from 0 to 34, with 0 indicating normal gait. Table II: Mean (SD) of kinematic and physical examination data for all patients (n=105) a Right leg Left leg t-test Mean (SD) Mean (SD) p Peak hip rotation in stance 12.5 (9.5) 9.8 (10.8) Mean pelvic rotation 1.1 (6.9) 1.6 (6.9) Mean knee progression angle 7.9 (9.6) 3.0 (10.4) <0.001 Mean foot progression angle 3.1 (14.0) 0.9 (16.8) Limit of passive internal rotation 54.1 (12.3) 64.4 (12.7) <0.001 Limit of passive external rotation 29.6 (14.8) 28.5 (14.9) Midpoint of rotation 12.4 (10.8) 18.0 (10.3) <0.001 Femoral anteversion 30.0 (8.5) 31.0 (8.6) Hindfoot thigh angle 8.6 (9.5) 3.5 (9.4) <0.001 Transmalleolar axis 19.1 (10.2) 12.0 (9.6) <0.001 a For peak internal rotation in stance and knee progression angle, only those trials with knee ab/adduction range <16 are included (83 right and 84 left legs). Significant differences are shown in bold. Rotation of Lower Limbs in Children with Diplegic CP Mariëtta L van der Linden et al. 177

3 positive and retraction was defined as negative. Mean knee progression angle was calculated by adding the mean hip rotation (internal positive) and the mean pelvic rotation angle (protraction positive): a positive knee progression angle signifies an internal knee progression angle. PHYSICAL EXAMINATION DATA The physical examination data analyzed included femoral anteversion, maximum passive internal and external hip rotation ranges, hindfoot thigh angles, and transmalleolar axes. Femoral anteversion was assessed using the method described by Ruwe et al. (1992). A separate assessor used a manual goniometer to measure the angle of rotation between the shank and the vertical. Internal and external hip rotation ranges were also measured with the participant in prone position and with the pelvis level, hips in neutral, and knees flexed to 90. Again, rotation was measured as the angle between the shank and the vertical. Hindfoot thigh angle, the angle between the lines bisecting the thigh and the hindfoot, was measured with the participant in prone position, the hips in neutral, knees flexed to 90, and with the subtalar joint neutral (Karol 1997). Transmalleolar axis was measured using a novel footprint method developed in our gait analysis laboratory. The foot was placed on a piece of lined paper with the participant sitting. The hip was held in neutral rotation and the tibial tubercle was facing forward to minimize rotation through the flexed knee. The paper was aligned with the long axis of the thigh so that the lines were parallel with the knee axis. Two marks were made vertically downwards from the centres of the malleoli using a small set square. A line was drawn between these two marks and the angle measured between this line and any line on the paper. All goniometric measurements were recorded to the nearest degree. The mid-point of passive hip rotation range (Kerr et al. 2003) was determined by mathematically combining the internal and external ranges, measured on the same numerical scale as the sum of internal and external rotation, with external rotation defined as negative and divided by two. Coefficients of intra-observer reliability (Bland and Altman 1986) were derived by two teams including one assessor and one measurer each. Both teams took two measurements one week apart in five children with CP. Coefficients of intra-observer reliability of the measurements ranged from 12.3 to 12.4 for internal rotation, from 12.4 to 20.3 for external rotation, and from 6.4 to 13.8 for femoral anteversion, depending on the team of assessor and measurer. Physical examinations were carried out by one of the two assessors in the repeatability study. STATISTICAL ANALYSIS Left and right sides were analyzed separately for both groups. An independent Student s t-test was used to evaluate differences between the left and right side, and between the most affected legs, with p<0.05 set as the level of significance. To assess the differences between the left and right leg in group S, a paired Student s t-test was used. Pearson s correlation coefficient (r) was determined between hip rotation in gait and the physical examination, as well as other kinematic parameters. From Pearson s correlation coefficients, coefficients of determination (r 2 ) were calculated. The value of r 2 represents the proportion of variance in hip rotation in gait that can be explained by physical examination parameters. Results ALL PATIENTS Of the 105 patients, 22 (20.9%) required crutches, sticks, tripods, or a K-walker for daily ambulation. Twelve patients (11.5%) used a wheelchair or buggy for community distances, Table III: Coefficient of determination (r 2 ) between hip rotation in gait and other kinematic parameters and physical examination measurements Right leg (n=83) Left leg (n=84) Limit of passive internal rotation a Limit of passive external rotation a a Midpoint of rotation a b Femoral anteversion a a Hindfoot thigh angle Transmalleolar axis a p<0.001; b p=0.04. Only those trials with a knee ab/adduction range of less than 16 (83 right and 84 left legs) are included. Table IV: Average physical examination and kinematic parameters for group L (n=39) and group R (n=33) Group L Group R Difference between Right Left p Right Left p affected legs Mean (SD) Mean (SD) Mean (SD) Mean (SD) p Max hip rotation (stance) 5.7 (8.4) 10.5 (9.8) (7.9) 5.9 (9.5) < Mean pelvic rotation 4.3 (6.8) 4.6 (7.0) (8.0) 1.5 (7.4) Mean knee progression angle 6.7 (10.0) 0.2 (9.8) (9.0) 2.7 (10.9) <0.001 Mean foot progression angle 5.6 (16.1) 2.0 (18.4) (11.8) 1.9 (15.3) Maximum passive internal rotation 49.6 (11.9) 65.7 (11.8) (12.8) 61.7 (15.8) Maximum passive external rotation 34.1 (13.2) 28.0 (13.7) (14.5) 27.2 (13.3) Midpoint of rotation 7.8 (9.9) 18.9 (9.6) < (10.2) 18.1 (10.4) Femoral anteversion 27.6 (6.9) 31.6 (8.5) (9.3) 31.7 (7.7) Hindfoot thigh angle 8.2 (10.8) 2.4 (9.7) (8.1) 7.1 (9.7) Transmalleolar axis 18.5 ( (8.8) (9.3) 12.9 (8.1) Only those trials with a knee ab/adduction range of less than 16 are included. Pelvic protraction, internal hip rotation, external hindfoot thigh angle, external transmalleolar axis, internal foot progression angle, and internal knee progression angle are positive. Most affected legs are in bold. 178 Developmental Medicine & Child Neurology 2006, 48:

4 e.g. shopping. Fifty-four patients (51.4%) usually wore one or two ankle foot orthoses, one (0.9%) patient used ground-reaction foot orthoses, and one (0.9%) patient used a knee ankle foot orthosis. Ten children (9.5%) normally used one or two dynamic ankle foot orthoses. The average score of the Gillette Functional Assessment Questionnaire (Novacheck et al. 2000) was 7.5 (median 8), ranging from 2 to 10 (with 10 the maximum score). Several significant differences between the left and right legs were found in the whole group of patients regarding gait and passive rotation characteristics (Table II). Gait results for all patients showed that the left pelvis was slightly but significantly more retracted (p=0.035), while knee progression on the right was clearly more internal on the right (7.9 ) than on the left (3 ). There was a trend towards more internal hip rotation on the right, but this was not statistically significant. Although the knee progression angle was clearly more internal on the right, foot progression angle tended towards being more external on the right; this trend was not significant. Although there was a tendency for internal hip rotation in gait to be higher on the right, the maximum passive internal rotation of the hip was significantly higher on the left than on the right (p<0.001). At the level of the tibia, the hindfoot thigh angle and the transmalleolar axis were significantly more external for the right leg (p<0.001): right hindfoot thigh angle was 5 more external and the transmalleolar axis was nearly 7 more external than left angles. The values of the coefficients of determination with internal hip rotation in gait were also found to differ between sides (Table III). Generally, there was a stronger relation between hip internal rotation in gait and the passive joint ranges on the right side than on the left side. Neither femoral anteversion, hindfoot thigh angle, nor transmalleolar axis was correlated with age. Pearson s correlation coefficients were all less than 0.1. PATIENTS CLEARLY ASYMMETRICALLY AFFECTED Because of differences between left and right, whether there were relatively more patients who were more affected on one Table V: Average physical examination and kinematic parameters for group S (n=33) Right Left Paired Mean Mean t-test (SD) (SD) p Max hip rotation (stance) 12.5 (10.8) 15.6 (11.0) Mean pelvic rotation 1.0 (8.6) 1.4 (9.1) Mean knee progression angle 7.7 (11.4) 8.4 (10.2) Mean foot progression angle 8.3 (15.4) 7.8 (17.0) Maximum passive internal rotation 54.2 (11.5) 65.4 (11.6) <0.001 Limit of passive external rotation 31.5 (15.6) 29.4 (15.9) Midpoint of rotation 11.5 (11.0) 18.0 (8.6) Femoral anteversion 29.0 (7.3) 30.4 (8.0) Hindfoot thigh angle 7.7 (10.6) 1.7 (9.1) Transmalleolar axis 21.0 (10.9) 11.6 (11.1) <0.001 Only those trials with a knee ab/adduction range of less than 16 included. Pelvic protraction, internal hip rotation, external hindfoot thigh angle, external transmalleolar axis, internal foot progression angle, and internal knee progression angle are positive. side was examined. Using the criteria described in the method section, there was no great difference in the number of patients who were clearly more affected on the right side (n=33) than those who were clearly affected on the left (n=39). Both groups also included approximately equal numbers of patients who were born preterm (Table I). As could be expected, the Edinburgh Gait Score (Read et al. 2003) of the left side was significantly higher in group L and the score on the right was significantly higher in group R (p<0.001; Table I). Group S had similar gait scores for both legs. Scores ranged from 1 to 25, out of a possible score of 34. To gain more insight into the differences between the left and right sides, only those patients who were clearly more affected on either the right (group R) or the left side (group L) were included in a separate analysis. Mean values for the physical examination and kinematic parameters are shown in Table IV. Results for group S are shown in Table V. Table IV shows significantly different values between groups L and R for the most affected leg. Maximum passive internal rotation of the hip was significantly higher for the left leg in group L (p<0.001). In group R the maximum passive internal rotation of the hip was the same for both sides. The affected legs in both groups showed similar characteristics for maximum passive external rotation of the hip and femoral anteversion. As expected, the most affected leg showed a significantly lower maximum passive external rotation and a significant higher femoral anteversion than the least affected leg. However, for both groups, the hindfoot thigh angle and the transmalleolar axis were significantly more external on the right side (p<0.001). Hence the most affected leg in group L had less external tibia torsion than the contralateral leg, while the most affected leg in group R had more external torsion than the contralateral leg. Comparing the most affected legs in both groups, the left hip of patients in group L was significantly less internally rotated in gait (p=0.003) but had a significantly higher maximum passive internal rotation than the right leg in group R (p=0.017). Left knee progression angle was also significantly less internal in group L: 0.2 compared with a right knee progression angle of 10.1 in group R (p<0.001). Discussion Studies reporting the tibial torsion of the left and right leg separately have used a variety of methods such as X-ray, fluoroscopy (Clementz 1989), computerized tomography (CT; Reikeras et al. 2001), magnetic resonance imaging, ultrasound (Krishna et al. 1991), and physical examination (Staheli et al. 1985). The findings often do not agree; this may be because not only the methods but also the reference lines, age, and sex of participants differ among the studies. Staheli et al. (2003) stated that, as the human tibia tends to rotate to the right, unilateral medial or internal tibial torsion is more common on the left and lateral tibial or external torsion is more common on the right. A study using CT to measure the transmalleolar axes of 504 adult tibiae found a discrepancy of 3.4 between right and left sides, with the right being the more external (Strecker et al. 1997). Seber (2000) did not find any significant difference between left and right tibial torsion in 50 healthy male participants according to CT. Krishna et al. (1991) measured tibial torsion using ultrasound in 78 normal children aged 4 to 15 years and found no significant difference Rotation of Lower Limbs in Children with Diplegic CP Mariëtta L van der Linden et al. 179

5 between left (41.2 ) and right (39.3 ) legs. Using a mechanical method to measure tibial torsion, Campos and Maiques (1990) reported a significantly higher external tibial torsion in the right than in the left in healthy females aged 3 to 13 years old. In males aged between 3 and 17 years, the trend towards more external tibial torsion on the right was not significant. In a study of the effect of walking speed on the gait of 35 healthy children, the transmalleolar axis was routinely measured using the same method as in the current study (van der Linden et al. 2002). It was found that in this group of children, between the ages of 8 and 11 years (21 males, 14 females), the mean external tibial torsion on the right (22.4 ) was significantly greater than that on the left which was 18 (p=0.001). Children with CP in the current study (mean age 13y) had less external tibial torsion. Differences between left and right were more pronounced in the current study than in the previously studied group of healthy children. Differences between right and left found in this study may be related to the tendency for all neonates to lie preferentially rotated to the right, with the head rotated to the right and the limbs rolling into slight external rotation and abduction on the right, and internal rotation and adduction on the left. It may be that motor delay and difficulties associated with CP lead to children remaining in this position for longer than normal, resulting in noticeable gait deviations. Although a change in femoral and tibial torsion with age in healthy children has been reported by several authors (Staheli et al. 1985, Campos and Maiques 1990, Reikeras et al. 2001), no correlation with age was found in children with diplegic CP in the current study. Normal torsion development is the result of muscle forces on the immature skeleton during normal gait. The gait of children with CP is different from that of normally developing children. Children with CP often start walking at a later age than children without disabilities, hence the forces acting on the structures in the lower limbs of children with CP will also be different. The development of torsional characteristics, indicated when passive and in gait, is the result of a complex interaction between initial orientation of bony structures, muscle characteristics, and muscle forces during gait, and in frequently adapted positions, especially in the non-ambulatory phase. This study showed important differences between left and right legs. Maximum passive internal hip rotation tended to be greater on the left than on the right side, even when the right side was more affected. No correlation was found between maximum passive internal rotation of the hip and hip rotation in gait for the left leg. This may be important in the clinical evaluation and gait assessment of patients with diplegia and the subsequent surgical management of pathological torsions in the lower extremities. The current findings also suggest that consideration should be given to a positioning programme for children with CP in the pre-ambulatory and early ambulatory phases of motor development to minimize lower extremity asymmetries. Longitudinal studies analyzing the torsional aspects of the lower limbs of young children with CP, when passive and during gait, may give further insight into the development of torsional abnormalities in children with CP. DOI: /S Accepted for publication 9th June Acknowledgements This research was funded by the James and Grace Anderson Trust. We thank Ms Alanah Kirby for her helpful advice. References Aktas S, Aiona MD, Orendurff M. (2000) Evaluation of rotational gait abnormality in the patients cerebral palsy. J Pediatr Orthop 20: Bland JM, Altman DG. (1986) Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 8: Brunner R, Krauspe R, Romkes J. (2000) Torsionfehler an den unteren Extrimitäten bei Patienten mit infantiler Zerebralparese. Orthopäde 29: (In German) Campos FF, Maiques JAP. (1990) The development of tibiofibular torsion. Surg Radiol Anat 12: Clementz B-G. (1989) Assessment of tibial torsion and rotational deformity with a new fluoroscopic technique. Clin Orthop Rel Res 245: Fabry G, MacEwen GD, Shands AR Jr. (1973) Torsion of the femur: a follow-up study in normal and abnormal conditions. J Bone Joint Surg (Am) 55: Fulford GE, Brown JK. (1976) Position as a cause of deformity in children with cerebral palsy. Dev Med Child Neurol 18: Gage JR (1991) Gait Analysis in Cerebral Palsy. Clinics in Developmental Medicine No.121. London: Mac Keith Press. p 111. Kadaba MP, Ramakrishnan HK, Wootten ME. (1990) Measurement of lower extremity kinematics during level walking. J Orthop Res 8: Karol LA. (1997) Rotational deformities in the lower extremities. Curr Opin Pediatr 9: Kerr AM, Kirtley SJ, Hillman SJ, Linden ML van der, Hazlewood ME, Robb JE. (2003) The mid-point of passive hip rotation range is an indicator of hip rotation in gait in cerebral palsy. Gait Posture 17: Konishi Y, Kuriyama M, Mikawa H, Suzuki J. (1987) Effect of body position on later postural and functional literalities of preterm infants. Dev Med Child Neurol 29: Krishna M, Evans R, Sprigg A, Taylor JF, Theis JC. (1991) Tibial torsion measured by ultrasound in children with talipes equinovarus. J Bone Joint Surg (Br) 73: Majestro TC, Frost HM. (1971) Spastic internal femoral torsion. Clin Orthop 79: Novacheck TF, Stout JL, Tervo R. (2000) Reliability and validity of the Gillette Functional Assessment Questionnaire as an outcome measure in children with walking disabilities. J Pediatr Orthop 20: Owers KL, Pyman J, Gargan, MF, Witherow PJ, Portinaro NMA. (2001) Bilateral hip surgery in severe cerebral palsy: a preliminary review. J Bone Joint Surg (Br) 83: Read HS, Hazlewood ME, Hillman SJ, Prescott R, Robb JE. (2003) Edinburgh visual gait score for use in cerebral palsy. J Pediatr Orthop 23: Reikeras O, Kristiansen LP, Gunderson R, Steen H. (2001) Reduced tibial torsion in congenital clubfoot. CT measurements in 24 patients. Acta Orthop Scand 72: Ruwe PA, Gage JR, Ozonoff MB, Deluca PA. (1992) Clinical determination of femoral anteversion. J Bone Joint Surg (Am) 74: Seber S, Hazer B, Kose N, Gokturk E, Gunal I, Turgut A. (2000) Rotational profile of the lower extremity and foot progression angle: computerized tomographic examination of 50 male adults. Arch Orthop Trauma Surg 120: Staheli LT, Corbett M, Wyss C, King H. (1985) Lower-extremity rotational problems in children. Normal values to guide management. J Bone Joint Surg (Am) 67: Staheli LT. (2003) Pediatric Orthopaedic Secrets. 2nd edn. Philadelphia: Hanley and Belfus. p 218. Strecker W, Keppler P, Kinzl L. (1997) Length and torsion of the lower limb. J Bone Joint Surg (Br) 79: Tachdajian MO. (1990) Pediatric Orthopaedics. Vol 1, 2nd edn. Philadelphia: WB Saunders Company. p 303. van der Linden ML, Kerr AM, Hazlewood ME, Hillman SJ, Robb JE. (2002) Kinematic and kinetic gait characteristics of normal children walking at a range of clinically relevant speeds. J Pediatr Orthop 22: Developmental Medicine & Child Neurology 2006, 48: