Using inclinometer to enhance precision of hip motion measures
The idea of the use of the scoliometer to standardize pelvis position raised from the authors' observations, that medical students presented obvious difficulties in assessing the range of hip motion, even helped with a goniometer. Usually, the pelvis was not correctly stabilized, so it moved together with the moving lower limb. Moreover, the neutral pelvis rotation was erroneously assessed. We solved both technical pitfalls with the use of the scoliometer. We consider the classical way of pelvis stabilization (with one examiner's hand) very useful to detect the moment the pelvis begins to tilt or rotate. However, the neutral position of the pelvis cannot be precisely controlled, unless the scoliometer is used. Some technical points may be raised: (1) Shortening of the rectus femoris muscle may be responsible for elevation of a hemipelvis during knee flexion; we recommend performing the clinical test of Ely in order to detect rectus femoris shortening. (2) In the patient lying prone, the pelvis is not level spontaneously, usually the pelvis presents some degree of rotation, which is expressed by elevation of one side. (3) The iliac bones are susceptible to nutation and counter-nutation movements in the iliosacral junction, creating asymmetric position of the iliac spines. Thus, a part of the asymmetry found in this study could be attributed to the intrinsic pelvic deformation.
Hip joint asymmetry in patients with idiopathic scoliosis
The range of motion of the joints is normally symmetrical, and this serves in clinical practice to detect the affected limb, by comparing with the "healthy" limb. Generally, the amplitude of joint motion decreases in the presence of a pathology of the joints, bones or muscles. In our study, the global amplitude of the hip rotation motion, calculated as the sum of the internal rotation range and external rotation range, revealed no limitation. Thus, the traditional term of pathology seems not justified to describe the asymmetries found both in girls with scoliosis and in healthy controls. However, there was a significant asymmetry between left and right hip joint in the corresponding movements, namely medial rotation, lateral rotation, and adduction. We propose to apply the term of transposition of the sector of the hip joint motion.
Our results document numerous asymmetries around the pelvis in adolescent girls with idiopathic scoliosis. Both the prevalence of the hip asymmetry and its severity significantly exceeded the values of the control group of healthy adolescents sex and age matched. The range of left-right hip rotation asymmetry was higher in scoliotic girls comparing to controls; the difference being statistically significant. The left – right hip asymmetry of the range of rotation was extremely significant (as in the controls either). There was a pitfall to compare means, which were equal for the left and right side. Saji et al.  did not observed any left-right asymmetry of the femoral neck-shaft angle in patients with scoliosis, however their methodology comprised comparing the means, while calculating the absolute value of the left-right differences for each patient was not performed. In our study, only in 26 of 158 girls with scoliosis the range of rotation was symmetrical, in the remaining 132 girls the asymmetry ranged from 5 to 30 degrees. Increased internal rotation was usually accompanied by decreased external rotation and vice versa; the total range of rotation movement being the same. Thus, a static rotational offset of the pelvis was demonstrated.
Our results failed to demonstrate a significant difference in the values of the parameters of the hip joints range of rotation between the progressive and the non-progressive scoliosis (Table 2). Also, no relation of hip asymmetries to the radiological Lenke curve classification was found (Table 3).
In this study, the asymmetric range of hip rotation was documented in the cross-sectional pattern, with no longitudinal evaluation. This seems justified, unless the hip asymmetries are considered secondary to spinal curvatures. Thus, the evolution in the range of hip motion should be studied over time.
A specific asymmetry of the hip joint range of motion, namely the limitation of the adduction of the right hip, was hypothesized to be an etiologic factor for idiopathic scoliosis . Our findings regarding the right hip adduction in children with scoliosis did not follow the results reported by Karski. For a better precision, we checked the hip adduction in three positions: supine, lateral and prone. Our impression was that the prone position "at the end of the table" provided the best conditions for measurements, because: (1) the patient does not need to change the position during exam, (2) the pelvis is easily controlled with one hand while the other hand controls the flexion and rotation of the thigh.
The interpretation of our findings can be made not only under the Nottingham concept. Hip joint static asymmetries could be the consequence of muscle imbalance (primary or secondary), or of morphological asymmetries of the pelvis. Rigo  described an indirect radiological factor (BSIIa difference: bisacro-iliac-ischial angle difference) to assess the pelvis torsion. With this factor, the author was able to make the difference between the iliac rotation asymmetry (due to nutation/counter-nutation movements) and the structural intra-bone iliac deformity. He also combined patterns of pelvis asymmetry with structural changes inside the lumbar spine (wedging in L3/L4 disc). The author concluded that the pelvis asymmetries could be considered as a part of the whole body torsional phenomenon rather than cause-effect factors. Considering the fact that in the current study the asymmetries were observed in a significant number of normal subjects just with a lower prevalence and degree, our findings could be also interpreted following the theory of idiopathic scoliosis as a sign of developmental instability , where scoliotic subjects present a major degree of directional and fluctuating asymmetry , as a factor indicating their predisposition to develop scoliosis.
Implications for gait in scoliotics
The static rotational offset of the pelvis revealed more important values in girls with scoliosis, than in healthy adolescents. That means the pelvis is not as balanced in the transverse plane, as in controls. However, the difference seems moderate, and it was difficult for us to attribute a particular physiological meaning to this finding.
Debating the biomechanical factors of pathogenesis of idiopathic scoliosis, our findings correspond to the Nottingham concept. We found a significant asymmetry in the internal/external rotation between both hips, which might be the origin of the asymmetry of the "dinner plate mechanism". Kramers-de-Quervain  reported asymmetric rotation of the trunk during gait in between symmetrical rotation of the pelvis and the head. The magnitude of the torsional offset correlated with the degree of the thoracic component of the scoliotic deformity. Burwell et al. found a more important asymmetry of femoral anteversion in school screening referrals than in control subjects. Giakas et al  constructed the study to reveal eventual overloading of the lower limbs during gait due to the bending of the spine in one side. The authors failed to demonstrate such an effect and suspect that subjects with scoliosis have an unidentified functional difference that acts to prevent asymmetric loading of the ground reaction forces during gait.
There is a tendency to optimize the economy of the muscle effort in gait. Symmetrical motion of the pelvis is most economic; an abnormal pelvis swing provokes more important muscle effort. In scoliotic subjects the non ergonomic movements of the pelvis resulting from hip asymmetries usually are not observed, probably being corrected by asymmetric activation of back and lower limbs muscles. On the other hand, keeping a symmetric position of the head in gait is an obviousness. Thus, the effective walking function implies that the thoracic and lumbar spinal segments stay mobile in between actively stabilized upper and lower trunk, so they may be susceptible to asymmetric position/motion. Muscle activation pattern necessary to maintain the ergonomic gait may result in abnormal activation of trunk muscles, with subsequent asymmetric arms of force of the muscles attached to the vertebral column. Thus, the rotation-inducing mechanisms (mechanical component of the Nottingham concept) may provide an asymmetric action on the spine, even if they reveal a symmetrical motion in gait.
Does the "dinner plate" turns asymmetrically? Our study demonstrated a significantly higher static rotational offset of the pelvis in adolescent girls with idiopathic scoliosis comparing to age matched controls. During gait the 3D pelvis tilts are supposed to be actively minimized for a better gait ergonomics, the dynamic stabilization of the static pelvis rotational offset may be a source of abnormal (asymmetric) activation of back muscles. This provides arguments that the "dinner plate" may have an asymmetric influence on the spine, in spite it moves symmetrically during gait, by producing a "torsional offset" of muscle forces around the spine, a mechanism potentially contributive to scoliosis development/progression.