Our study showed that the wedging of vertebral body and intervertebral disc was consistently present in both the thoracic spine and in lumbar spine. In the thoracic spine, the vertebral body wedging was more distinct as compared to the thoracic disc. However, in the lumbar spine the intervertebral disc wedging was more evident. The progress of the scoliotic curves is commonly thought to as per the Hueter-Volkmann's law . As per this law, epiphyseal growth is inhibited when compressive force act on it, and stimulated when distraction force is applied. Based on this law Roaf  has proposed a vicious cycle regarding progression of kyphosis. According to it, a minimal wedging of the vertebrae would produce abnormal compressive force on the vertebral end plate, which would further increase the wedging as per the Hueter-Volkmann's law and thus produce further abnormal forces. Using this principle, scoliotic curves have been reproduced on animal studies, like the one done by Braun et al . He created an idiopathic type of deformity in goats by applying forces across the spine. He also found wedging of the vertebrae similar to that seen in scoliosis in humans. Similarly, Mente et al [8, 9] and Stokes et al [10–12], in separate studies on rat-tail models, not only could they create a scoliosis like deformity, but were also able to correct it when the forces were reversed.
In the thoracic spine, the wedging pattern of growth modulation was thus according to the Hueter-Volkmann's law . It was seen more towards the apical region of the curves, gradually decreasing towards the end vertebrae supporting the law. In our study, the maximum wedging was seen in the apical vertebra. This result was consistent with the study of Stokes and Aronsson  suggesting disc and vertebral wedging in progressive scoliosis. The current study also shows that load distribution is always concentrated maximum at the apex on the assumption that wedging results from mechanically modulated growth. However, the role of ribs could not be neglected in thoracic spine for scoliosis. Numerous authors [17, 18] have reported role of rib deformity as pathogenesis of scoliosis or rib resection for the correction of scoliosis. Xiong and Sevastik  did shortening of three concave side ribs in a 7 years old girl with scoliosis and reported 36% correction. However, in a 7 years old child, role of spontaneous correction of the curve should not be forgotten. Sevastik et al  has measured rib asymmetry in 10 thoracic idiopathic scoliosis and found that convex rib vertebral angle (RVA) is smaller than concave RVA between T2–T8 and become reverse between T9–T12. Similarly Agostini et al  have established relationship between rib hump deformity and vertebral rotation in idiopathic scoliosis. However, in present, we did not measure the difference in RVA between convex and concave side to find out rib asymmetry, which may be a lacuna in this study.
In contrast to the thoracic spine, the lumbar spine showed no significant difference in vertebral body wedging between the apical and other vertebrae; however, it is significant finding that the intervertebral discs follow the same nature of wedging, as it is body in thoracic spine. Actually, in lumbar spine discs are more flexible than thoracic discs; so more stress concentration in disc explain more wedging than the vertebral bodies. According to Stokes and Aronsson , among the patients with idiopathic scoliosis who had a thoracic major curve, the wedging at the apex was greater in the vertebrae than in the discs, whereas the opposite was generally found at the apex of the major lumbar and thoracolumbar scoliosis curves. Therefore, the results of their study do not support the hypothesis of Taylor  that the wedge deformity begins predominately in the discs and subsequently, with curve progression, the vertebrae become wedged. The division of wedging between vertebrae and discs in thoracic and lumbar curves may be related to the different disc thickness (relative to vertebral height) in these two anatomic regions. Our findings also showed similar results and we therefore agree to their conclusion suggesting that in adolescent idiopathic scoliosis, the wedging in the disc and body will be different according to anatomic region even in same type of curve. Stokes and Morse  reported that muscle activation patterns generating spinal loading does not promote curve progression. They did their study in lumbar spine and they thought that scoliosis can adopt different muscle activation and so they did not support muscle role for curve progression. Puustjarvi et al  reported in their study that long distance running in digs causes reductions in proteoglycan content of cervical and thoracic discs but increases in lumbar discs. The differences depending on spinal region were attributed to different biomechanical demands, showing the characteristics of mechanical loading may influence disc component. Urban et al  noted that solute diffusion into the apical disc (measured by flux of nitrous oxide) was reduced due to abnormal mechanical stress on lumbar disc. They speculated that in scoliosis there is a combination of overload and reduced motion due to disc degeneration that results in curve development in elderly people. In addition, as disc is avascular in nature, reversal of load and stress cannot reverse the disc degeneration and wedging back and probably that is the reason why lumbar curve is difficult to treat with conservative treatment. Moreover, disc wedging is a consistent finding in lumbar scoliosis. This was again more towards the apex of the curve.
Grivas et al  has suggested that vertebral body wedging appears later when already Cobb angle increases and in small Cobb angle there is no vertebral wedging. Based on their findings they suggested that when the deformity is initiating, intervertebral disc is found wedged but not vertebra body, due to increased plasticity of disc. Recently they  proposed a theoretical model of idiopathic scoliosis pathogenesis describing the role of intervertebral disc in correction of scoliotic curves. They suggested that wedging of the elastic intervertebral disc in the immature scoliotic spine could be reversed by application of corrective forces on it either by bracing or staples, which ultimately create modulation of intervertebral disc composition. However comparing vertebral wedging to disc wedging, our study shows that disc wedging is a far more important component of lumbar scoliosis. Similarly the disc wedging in thoracic spine was far less than the vertebral wedging, stating that vertebral wedging was a much more important component of scoliosis in thoracic spine. The difference could be due to the fact that our study population comprised of established scoliotic patients not those who were initiating the curve. And therefore in thoracic spine we could observe more vertebral wedging than the discs confirming our hypothesis of differential wedging in lumbar and thoracic spine (Table 2). In a similar study, Stokes and Aronsson  also showed that in both idiopathic and neuromuscular scoliosis groups of patients, the mean vertebral wedging was more than the disc wedging in the thoracic region; the converse was found in curves in the lumbar and thoracolumbar regions.(greater vertebral body wedging in thoracic spine and disc wedging in lumbar spine.) However, their study contained a small sample of patients as compared to our sample group and the purpose of their study was mainly to document the spinal growth due to vertebral body after the age of ten years. In addition, we also noted more wedging angle in more degree of curve, which again explain that wedging increase with the progression of scoliosis. This finding again confirms the observations of Burwell et al  that severe and moderate thoracic idiopathic scoliosis the thoracic hump correlates with Cobb angle and the apical vertebral rotation and lateral asymmetry of the back is the major exterior aspect of scoliosis. Recently Stokes [14, 15] said that in a predictive model of the evolution of scoliosis simulating the 'vicious cycle' theory, and using published data, a small lateral curvature of the spine can produce asymmetrical spinal loading that causes asymmetrical growth and a self-perpetuating progressive deformity during skeletal growth. We also agree to these findings and we think our findings of differential wedging pattern in idiopathic scoliosis could be a possible mechanism for the 'vicious cycle' of progression of scoliosis curve.
In present paper we, however, could not study the effect of vertebral rotation according to severity of curve because we have measured the rotations Nash and Moe method which is not in measurement but in grades. Therefore we were not able to analyse statistically with severity of curve as most of the rotations were grade 1. We think that may be the weak point in our study that we could not analyze this ordinal data statistically.