Underlying this thesis is the McKenzie school of thought, a physiotherapy approach that teaches clinicians to recommend particular exercises to their clients in an attempt to accelerate recovery/ prevent recurrence of disc prolapse. The recommendations are based on an untested clinical theory that movements opposite to those that cause disc prolapse can achieve reversal of disc prolapse. Little consideration has been given scientifically to the reversal of the failure process of the lumbar discs. Three in vitro and one in vivo study were designed to attain a greater understanding of both disc failure and the mechanics of its clinical treatment responses and thereby provide a foundation for evidence-based practice.

The first in vitro study in this thesis compared in vitro and in vivo herniated discs in an attempt to link the two and provide a more thorough understanding of the in vitro model proposed to test the mechanical theory underlying the McKenzie derangement approach. Ten C3/4 osteoligamentus porcine specimens were repeatedly flexed or flexed and side bent to result in posterior migration of the nucleus. Three of the 10 specimens had posterior migration of the nucleus. Statistically significant (p < 0.01) and clinically significant (>33%) disc height loss occurred in all 10 specimens. The results provide a sub-classification of in vitro herniated discs that is similar to the spectrum of herniated discs that occurs in vivo.

Continuing from the disc height loss sub-classification of post-herniated in vitro discs, the second in vitro study in this thesis pursues alternate methods of creating herniation with the goal of creating herniation without causing more than thirty three percent disc height loss of the specimens. Repeated flexion of porcine cervical specimens under a lower compression level (1kN) resulted in disc herniation but with loss of 50% of the pre-test disc height (p < 0.001). Re-hydrating specimens by injecting the disc after a period of failure testing with a barium sulphate nucleus mix (n = 5) or by placing the specimen in a saline bath for an extended period of time (n = 4) resulted in a significant increase of the disc height of the specimens. Further flexion testing of the specimens significantly reduced the disc height again. Intermittent saline injection of specimens (n = 3) during the failure procedure did not prevent or reduce the disc height loss that occurred in the absence of saline injections. Using higher compression levels (2 and 2.596kN, n = 4), failure testing under torque control (n = 3), non-physiologically starting the annular rupture (n = 5) and using hypolordotic thoracic porcine spines (n = 9) instead of porcine cervical spines were unsuccessful attempts at creating herniations. This study indicated that the in vitro model used in the first in vitro study displayed features from one end of the spectrum of damage seen clinically but was then the bestavailable. Combined these two studies provide a framework for interpretation of the results of the subsequent and third in vitro study in this thesis.

_ The focus of the third study is the mechanical investigation of the McKenzie clinical theory of the treatment response seen in vivo in prolapsed discs, which is that movements or positioning can alter the location of a displaced portion of nucleus in a prolapsed disc. This study is a proof of the principle on which this aspect of the McKenzie approach is based and provides, to the author’s knowledge, the first scientific evidence supporting the theory that repeating movements opposite to those that caused posterior migration of the nucleus can centralize the prolapsed material. The results indicate that the McKenzie approach works on some prolapsed discs and not on others. Consideration of the changes in disc height of the specimens during the testing procedures offers some understanding of the varied success of this approach and exposes a vast area of future research that will refine the clinical approach and mechanical understanding of this specific disc pathology.

_ The fourth study, an in vivo study, provides a first look at the kinematics and kinetics of the current in vivo application of this approach. Twenty asymptomatic subjects volunteered to participate in this study and performed frequently prescribed McKenzie exercises and a selection of activities of daily living during which a 3-SPACE Isotrak system measured their three dimensional lumbar kinemetics. One subject underwent a series of McKenzie exercises while electromyography and threedimensional lumbar motion were measured. Mean peak extension of extension in standing and extension in lying exercises were within 3% (SD 22.33%) of each other. An additional 6.75% (SD 11.18%) of extension occurred when the extension in lying exercise was combined with a Grade 3 Maitland extension mobilization to L3, a passive physiotherapy technique that involves the therapist applying intermittent low amplitude oscillations to, in this case, the posterior aspect of the spinous process with the goal of subsequently increasing the range of active motion in the direction of the mobilization. The peak extension during the extension in lying exercise was increased after the mobilization relative to the pre-mobilization range. The mean peak right side bend in the right side glide exercise, normalized to the full right side bend range, was 61% (SD 17.4%). The L4-5 forces at the position of peak extension in extension in lying and extension in standing were 828.97N and 1368.86N respectively. The peak flexion ranges of the activities of daily living investigated match those previously used to create disc prolapse when applied at high repetitions and under moderate axial compression. The lumbar spine ranges achieved in commonly prescribed McKenzie rehabilitative and preventative exercises and those that occur in seemingly non-problematic activities of daily living were quantified. The results of this study will enhance clinical practice by providing quantitative evidence of the relative peak motion of the McKenzie exercises as well as highlighting seemingly benign activities of daily living that involve levels of flexion, side bend and rotation sufficient to cause disc damage and even prolapse.

The macroscopic goal of this thesis was to attain a greater understanding of the mechanics of both disc failure and its clinical treatment responses and thereby provide a foundation for evidence-based practice, a goal that was successfully achieved. This thesis ultimately challenged and increased our understanding of pathological discs while simultaneously adding information to assist clinical decision making. Several new contributions to the existing knowledge of lumbar spine biomechanics and clinical concepts of treating disc prolapse have been made.