The etiology of scoliosis.

Twenty years ago the late John Cobb first showed me that scoliosis could be interesting. By the very simplicity of his teaching he changed an obscure, incomprehensible deformity into a disease which has fascinated me ever since. I recall gratefully his ability to make the complex simple. I propose to look at this deformity in all its forms, to name the questions one must ask and answer if its etiology is to be established. At the end of this you will not know the etiology of scoliosis, for I do not myself, but it is my hope that you will comprehend the problem better. Although idiopathic scoliosis because of its frequency, its severity and the mystery of its origin is the main problem, it is important to look at scoliosis from all causes. Scoliosis is a deformity that may be caused by many diseases, and though in some the mechanism may be clear-as for example in congenital scoliosis-in most it is wholly beyond our understanding. After poliomyelitis, limb deformity occurs in children left with muscle imbalance while still growing; trunk muscle imbalance likewise causes paralytic scoliosis. Knowing this has singularly failed to help us to understand idiopathic scoliosis; children with idiopathic curves do not have muscle imbalance. Although there are differences between paralytic and idiopathic curves, the more remarkable feature is their similarity despite the dissimilarity of cause. It has been argued that unrecognised poliomyelitis is the cause of idiopathic scoliosis. It is now unnecessary to discuss all the reasons against this hypothesis because since the introduction of vaccine poliomyelitis has disappeared but idiopathic scoliosis has not. Cerebral palsy, the muscular dystrophies, arthrogryposis multiplex congenita: these muscular conditions cause scoliosis and perhaps they do so by the unequal pull of unequal muscles. Congenital scoliosis probably arises from asymmetrical growth in the abnormal vertebrae and growth plates. This seems evident, but as one cannot see the growth plates by radiography it is not known whether abnormal bone growth does cause these curves. Why is it that when one hemivertebra is present and causes a curve this rarely increases? Mechanically, once a curve is present it ought to increase. What happens to the growing areas on the concavity under pressure and those on the convexity released from pressure? It is a situation where our hypothesis suggests that the curve should always increase but observation proves that it does not. What of the neuropathies accompanied by scoliosis? Do we comprehend the mechanisms of these curves? In the general problem of why a curve develops in syringomyelia, Friedreich’s ataxia or neurouIbromatosis, not one single clue as to why deformity occurs is known. In neurofibromatosis 10 per cent of patients acquire a curve; although it is variable, there is a classical pattern (Fig. 1). Another characteristic is the acute angle at which the vertebral borders meet (Fig. 2). What is it that causes the deformity? It is not neurofibromata in relation to the vertebrae. Is it related to the disappearing bones of this disease? The vertebrae do not disappear; why indeed do bones disappear in this inherited, protean oddity? Of the scolioses due to known diseases it seems that a study of this particular condition, if it could explain how it gives rise to curves, would be invaluable in our understanding of idiopathic scoliosis. There is not yet even a working hypothesis.