Background & Purpose

Low back pain (LBP) affects approximately 80% of the population at some point in their lives, with 65% of cases becoming chronic pain sufferers. Approximately 40% of LBP cases are attributed to intervertebral disc (IVD) disruption, which can be characterized by damage to the annulus fibrosus (AF) or cartilaginous endplates. Although the medical community strives to treat and rehabilitate these injuries, there is immense opportunity for primary prevention, given that most problems will continue to persist until the underlying mechanical risk factors are resolved. The most commonly accepted mechanical exposures linked to low back injury include:​ (i) high force demands, (ii) frequent repetition and (iii) awkward postures. However, these demands have typically been examined in isolation, and are often assumed to function independently regarding injury risk. As such our understanding of the combined effects of external task demands on internal joint loading, and their influence on fatigue injury pathways in the lumbar spine may be limited, overly simplified, and underestimate the potential risk of injury. Therefore, using a combination of both in vivo modeling of human movement and in vitro mechanical testing of porcine IVD tissue, the global purpose of this thesis was twofold:​ (i) explore interactions between external task demands on in vivo joint loading in the lumbar spine, and (ii) characterize the combined effects of each of these mechanical exposures on fatigue injury pathways in the IVD. In-line with these global purposes, four independent studies were conducted, each with their own focused objectives.

Study I:​ Exploring interactions between external task demands in lifting on low back joint loads in vivo.

Background:​ The most commonly accepted mechanical risk factors linked to low back injury include:​ (i) high force demands;​ (ii) frequent repetition;​ and (iii) deviated postures. However, our understanding of the interactions that exist between each of these factors on the resultant in vivo joint loads that occur in the lumbar spine during manual materials handling remains limited.

Objectives:​ The primary objective of this study was to explore interactions between:​ (i) the magnitude of the external load, (ii) speed of lifting movement, and (iii) symmetry of initial load placement (i.e. posture) on estimates of in vivo joint loading at L4/L5 during simulated occupational lifting.

Methods:​ Thirty-four participants with manual materials handling experience were recruited. Three-dimensional motion data were captured together with foot-ground contact forces and activation of six bilateral trunk muscle groups while participants performed lifts with two loads (“light” and “heavy”), at two movement speeds (“controlled” and “fast”), and using two initial load placements (“symmetric” and “asymmetric”;​ i.e. posture). Inverse dynamics analyses were conducted, and outputs were incorporated into a three-dimensional EMG-assisted musculoskeletal model of the lumbar spine to generate L4/L5 joint compression and shear force-time histories.

Results:​ Significant interactions between load, speed and posture emerged in virtually every dependent measure that was used to characterize the time-varying estimates of load in the lumbar spine. Significant two-way interactions between load and speed (p = 0.0035), as well as speed and posture (p = 0.0004) were revealed in peak measures of L4/L5 compressive forces. A significant three-way interaction between load, speed and posture (p = 0.0477) was revealed in the cumulative measure of compressive loading at L4/L5 during the ascending portion of the lifting trials. A significant main effect of load (p < 0.0001) and a significant two-way interaction between speed and posture (p = 0.0384) was revealed in peak measures of peak L4/L5 anterior/posterior (AP) joint shear.

Conclusions:​ Results from this investigation provide strong evidence that known mechanical risk factors linked to low back injury should not be viewed in isolation. Stated differently, the influence of external demands on in vivo estimates of joint loading was not additive. Therefore, future injury prevention efforts should consider the complex interactions that exist between external task demands and their combined influence on internal (musculoskeletal) joint loading.

Study II:​ Exploring the combined effects of force, repetition and posture on injury pathways in isolated functional spinal units from sub-acute-failure magnitudes of cyclic compressive force.

Background:​ Previous research suggests that when the magnitude of peak compressive force applied during cyclic loading exceeds 30% of a functional spinal unit’s (FSU) estimated ultimate compressive tolerance (UCT), endplate fracture will occur before intervertebral disc (IVD) herniation. However, the amount of tissue damage imposed to the IVD of “survivor” specimens from these investigations remains unknown, as this threshold has been established through the detection of fatigue injury or using radiographic measures used to track the nucleus pulposus through layers of the AF.

Objectives:​ The primary objective of this study was to explore interactions between:​ (i) the magnitude of the applied compressive force, (ii) cycle rate and (iii) degree of postural deviation on known fatigue injury pathways in FSUs at sub-acute-failure magnitudes of peak compressive force. The secondary purpose was to characterize the micro-structural damage imposed to the AF in “survivor” specimens using histological staining methods. Methods:​ A total of 126 FSUs were initially included in the study. Three levels of peak compressive force (10, 20 and 40% of specimens’ UCT), three cycle rates (5, 10 and 30 cycles per minute) and two dynamic postural conditions (100 and 300% neutral zone range;​ NZ) were examined using a full-factorial design. Cyclic compressive force was applied using a modified material testing apparatus, using a time-varying waveform with synchronous flexion/extension rotation that was based on in vivo estimates of lumbar compression and spine posture during an occupation lifting task. All FSUs were cyclically loaded for 5000 cycles or until fatigue failure occurred. AF tissue from 36 “survivor” FSUs was excised for histological analysis across experimental conditions. Three tissue samples, consisting of the outermost 10 lamellae, were obtained from the anterior and posteriorlateral regions of the IVD from each specimen. To characterize the micro-structural damage in the AF, frozen tissue samples mounted in OCT compound were serially sectioned using a cryostat microtome, H&E stained and visualized with a brightfield microscope linked to a digital camera. Characteristic images of the micro-structural damage were captures at 10X.

Results:​ Of the 123 specimens that were considered for analysis, 99 (80%) survived 5000 cycles of cyclic compressive loading. Twenty-four FSUs experienced a fatigue related injury, which could be classified into three primary injury pathways:​ (i) endplate fracture, (ii) avulsion of the superior endplate and (iii) fracture of the pars interarticularis. A marked difference of the magnitude of peak compressive force was noted in the Kaplan-Meier survival function of experimental conditions that induced fatigue injury in less than 5000 cycles, with a higher probability of survival at 20% UCT. Overall, in the 40% UCT load condition, the probability of survival was less than 67%. When considering the representative images across each of the 10 and 20% UCT experimental conditions, it is noteworthy that a qualitative depiction of emerging interactions could be detected. The micro-structural damage detected in excised samples of AF tissue consisted of clefts and fissures within the intra-lamellar matrix, as well as delamination within the inter-lamellar matrix. There was no consistent trend in the amount of damage that was observed between 10% and 20% UCT loading conditions. However, a moderate effect of posture was noted, with increased disruption observed in tissue samples exposed to 300% NZ postural deviation.

Conclusions:​ Consistent with previous research, our findings support a threshold of peak compressive force of 30% UCT, where cyclic loading above this level is likely to lead to fatigue injury of the cartilaginous endplate in less than 5000 cycles of in vitro mechanical loading. However, findings from our histological analyses demonstrate that considerable IVD disruption occurred in specimens that “survived” 5000 cycles of cyclic loading at 10 and 20% UCT. Therefore, it is strongly recommended that future efforts in injury prevention consider more than just the magnitude of the applied load when evaluating risk of low back injury.

Study III:​ Exploring interactions between force, repetition and posture on intervertebral disc height loss and bulging in isolated functional spinal units from sub-maximal cyclic compressive loading.

Background:​ Most in vitro studies are limited in the ability to partition intervertebral disc (IVD) height loss from total specimen height loss since the net changes in the actuator position of the materials testing system used for testing simply reflect net changes to the entire osteoligamentous system of FSUs used for testing. Moreover, previous research conducted to characterize changes in IVD bulging has been limited to static compressive loading for 15 minutes or less.

Objectives:​ To explore interactions between:​ (i) the magnitude of the applied compressive force, (ii) cycle rate and (ii) degree of postural deviation on IVD height loss, as well as pre/post changes in annulus fibrosus (AF) bulging in order to better understand the structural changes that occur in the IVD under cyclic loading conditions.

Methods:​ A total of 99 porcine cervical FSUs were included in the study. Three levels of peak compressive force (10, 20 and 40% UCT), three cycle rates (5, 10 and 30 cycles per minute) and two dynamic postural conditions (100 and 300% of specimens’ neutral zone range) were examined using a full-factorial design. Compressive force was applied using a time-varying waveform based on estimates of in vivo loading using a dynamic EMG-assisted model of the lumbar spine during a floor to waist height lift, which was synchronous paired with dynamic flexion/extension. All FSUs were cyclically loaded for 5000 cycles or until fatigue failure occurred. Surface scans from the anterior aspect of the IVD were recorded in a both a neutral and flexed posture before and after the cyclic loading protocol using a 3D laser scanner. The peak anterior bulge of the AF at the midline of the IVD in the frontal plane was computed from the 3D surface profiles, pre/post testing. To facilitate the comparison of AF bulging measurements between specimens and improve the anatomical interpretation of this measure, the maximum anterior bulge perpendicular to a vector defined by the endpoints of the superior and inferior endplates has been reported.

Results:​ A significant three-way interaction (p = 0.0092) between the magnitude of peak compressive force, cycle rate and degree of postural deviation was observed in cyclevarying specimen height loss data. However, a significant main effect of peak compressive force (p=0.0003) was observed in IVD height loss obtained from the 3D surface profiles, pre/post testing. A wide range of Pearson product-moment correlation between total specimen height loss and IVD height loss (r = -0.54 to 0.95) was observed across experimental conditions, with the relative contribution of IVD height loss to total specimen height loss, ranging from 19 to 58%. Significant main effects of postural deviation (p=0.0016) and time (p=0.0423) were observed in peak measures of AF bulging. A wide range of Pearson product-moment correlation was also observed (r = -0.44 to 0.78) between IVD height loss and IVD bulging across experimental conditions.

Conclusions:​ This investigation provides evidence that total specimen height loss is not an accurate depiction of cycle-varying changes in the IVD across a range of in vivo scenarios that were replicated during in-vitro testing of FSUs. The magnitude of the applied compressive load was the only exposure variable that significantly influenced IVD height loss, as measured from the 3D surface profiles. Interestingly, postural deviation was the only factor that significantly affected the magnitude of peak AF anterior bulge, pre/post testing.

Study IV:​ Exploring interactions between the magnitude of tissue stretch and cycle rate on the mechanical properties of the annulus fibrosus during cyclic, biaxial tensile loading.

Background:​ Previous research has shown that the intervertebral disc (IVD) undergoes multidirectional tensile strain under in vivo loading conditions. In vitro testing has found that biaxial tensile testing of excised samples of AF tissue results in higher stresses at considerably lower magnitudes of strain compared to uniaxial testing methods. However, the cycle-varying changes in the mechanical properties of the AF have yet to be studied under cyclic biaxial tensile loading conditions.

Objectives:​ The primary objective of this study was to explore interactions between:​ (i) the magnitude of peak tissue stretch and (ii) cycle rate on cycle-varying changes in the mechanical properties of the AF tissue during cyclic biaxial tensile loading. A secondary purpose was to examine whether the mechanical response of multilayer AF tissue samples would be different about each axis of loading or across radial locations of the IVD.

Methods:​ Ninety-six AF tissue samples, consisting of 3-5 lamellae, were excised from two radial locations on the IVD, including:​ (i) the anterior and (ii) posterior-lateral regions. Each tissue sample was randomly assigned to 1 of 12 experimental conditions to examine interactions between the magnitude of peak tissue stretch and cycle rate on the mechanical properties of isolated multilayer AF tissue samples, during cyclic biaxial tensile loading. Three levels of peak biaxial tissue stretch (circumferential-axial), including:​ (i) 6-10%, (ii) 9- 15% and (iii) 12-20% and two cycle rates (5 and 10 cycles per minute) were examined using a full factorial design. Biaxial tensile load was applied in displacement control using a BioTester tensile loading apparatus in a temperature (29° C ± 1) and humidity (90% ± 5) controlled local testing environment. Top-down digital images of the tissue’s surface were captured during both the stretch portion of cycles 1, 10 and 100 using an overhead CCD camera.

Results:​ A significant three-way interaction between the radial location on the IVD, magnitude of tissue stretch and cycle rate (p = 0.0053) and a significant two-way interaction between the axis of loading and magnitude of tissue stretch (p < 0.0001) was observed in measures of peak tensile stress. Similarly, significant two-way interactions were observed in the S-S moduli between:​ (i) magnitude of tissue stretch and cycle rate (p = 0.0004), as well as (ii) magnitude of tissue stretch and axis of loading (p < 0.0001). Significant two-way interactions between:​ (i) the magnitude of tissue stretch and axis of tensile loading (p < 0.0001), (ii) cycle rate and axis of tensile loading (p = 0.0174) and (iii) transverse location of the IVD and cycle rate (p = 0.0005) were observed in cycle-varying measures of peak surface strains that were calculated using a virtual gauge region defined on the surface of each tissue sample.

Conclusion:​ Results from this study emphasize the cycle-varying tensile stress response of the AF, which changed across loading axis and region of the IVD. However, it is interesting that no significant changes were observed in S-S moduli after 100 cycles of cyclic biaxial tensile loading. These findings provide novel insight into the cycle-varying stress-relaxation that occurs in multilayer samples of AF tissue, which may be linked to the known mechanism of IVD tissue failure (i.e. development of fissures and clefts) under chronic loading conditions.

Global Summary:​

Collectively, the findings from this thesis indicate the combined effects of magnitude of the applied load, cycle rate (which inevitably affects movement speed) and postural deviation interact to modulate both the resultant load in the lumbar spine (in vivo), as well as the mechanical response (and resultant injury pathways) of the intervertebral disc under cyclic loading conditions. Stated differently, the influence of force repetition and posture on estimates of low back joint loading and intervertebral disc injury were not simply additive. Future injury prevention efforts need to consider the complex interactions that exist between mechanical risk factors that have been linked to musculoskeletal injury.