Standard anatomic classifications such as "trunk", "lower limbs", "upper limbs" can be misleading regarding the functional role and influence that the tissues in these body regions may play in adjacent body regions. In particular, much of the spine biomechanics literature has considered the lumbar spine in isolation, neglecting to account for the influence of the tissues of the lower extremities (muscles, ligaments and fascia) on the performance of the lumbar region of the torso. Some previous literature supports a systems level (i.e., trunk, pelvis and lower extremities) approach for better understanding of trunk stability during flexion-extension motions. The current study presents a new musculoskeletal model of the active spinal stability system that includes the local system (e.g., multifidus muscles) and global system (e.g., lateral erector spinae, rectus abdominis muscles etc.) as proposed by Bergmark (1989), but then adds a super global system that considers the influence of the lower extremity tissues on the responses of the lumbar region. This innovative model was verified throughout in vivo experiments involving human subjects that included three different physical exertion tasks that stressed the low back and the lower extremities in different ways to explore these important interactions.

The empirical work in this dissertation focused on gathering data from the local, global and super global biomechanical systems before and after three 10 minute exercise protocols and then during a 40 minute recovery session. Twelve participants performed three separate experiments (three protocols) on different days: Protocol A- alternately perform 25 seconds of full trunk flexion and 5 seconds upright, relaxed posture; Protocol B- alternately perform 25 seconds of isometric exertion in a 45 degree trunk flexion posture and 5 seconds upright, relaxed posture; and Protocol C- consecutively perform 25 seconds of full trunk flexion followed by 5 seconds of upright, relaxed posture followed by 25 seconds of isometric exertion in a 45 degree trunk flexion posture and 5 seconds upright, relaxed posture. Kinematic and physiological measures were recorded before during after these protocols as well as during the recovery period. In addition, a variable describing the level of fixation of the pelvis was considered to allow for a direct evaluation of the role of the pelvis/lower extremities on the performance of the lumbar region during these exertions: 1) lower extremity restricted stooping posture (pelvis and thigh restriction) and 2) free stooping posture. The data collected in these experimental trials included the peak lumbar flexion angle, the peak hip flexion angle, the peak trunk flexion angle, the EMG-off angle (i.e., flexion-relaxation), and the average normalized integrated electromyography (NIEMG) for the agonist muscles (lumbar extensors (multifidus and iliocostalis)), the antagonist muscles (lumbar flexors (rectus abdominis and external obliques)) and the lower extremity synergistic muscles (gluteus maximus and biceps femoris).

The results of in vivo experiments, focused on the role of the pelvis/lower extremities in trunk flexion-extension, showed a 6.4% greater lumbar flexion angle (36y vs. 38.3y), a 10.2% greater (or later) EMG-off angle in multifidus (31.6y vs. 34.8y), and a 8% greater EMG-off angle in the iliocostalis (30.6y vs. 33y) in the restricted stooping posture than in the free stooping posture. Collectively, these results suggest that additional passive moments about the lumbar spine are generated in the restricted stooping posture because of the relative fixation of the pelvis that is seen during the restricted stooping condition. Consistent with these results, 22% greater lower extremity activation (10.5% MVC vs. 8.2% MVC) was observed in the free stooping posture, as compared to the restricted stooping posture. This additional lower extremity muscle activation acts to stabilize the pelvis (the foundation of the spinal column) and generate passive moments in low back through the lumbodorsal fascia. Consequently, the enhanced pelvic stability and passive moments in the low back generated by the lower extremity active system (i.e. the super global system) led to the an 8% lower low back muscle activation level (15.1% MVC vs. 16. 3% MVC) in the free stooping condition. In addition, under the abnormal low back conditions (after protocols), the agonist muscles showed significant increases in both the free stooping posture and the restricted stooping posture (15% in both) to maintain spinal stability, but the synergist only increased in the free stooping (22%, 11.2% MVC vs. 13.7% MVC) (no difference in the restricted stooping posture). To summarize, these results indicate a significant role of the tissues of the larger super global system as a trunk stabilizer by immobilizing the pelvis during trunk flexion-extension motions and increasing the stiffness of the trunk systems by enhancing tension of the lumbodorsal fascia.

Regarding the effects of the 10 minute protocols on the biomechanical responses, results showed greater full lumbar flexion and deeper biomechanical equilibrium point between passive tissues and external moment (i.e., EMG-off angles) than the baseline (initial measure) after Protocol A: full lumbar flexion increased 7%; EMG-off angle increased 7.2% in multifidus and increased 7.8% in iliocostalis. In Protocol B the trends in the dependent variables were opposite to those seen in Protocols A: full lumbar flexion angle decreased by 4% and the EMG-off angles decreased by 4.9% in the multifidus and by 6.3% in iliocostalis. Protocol C (the mixed protocol) generated similar, but less pronounced results as compared to Protocol A: full lumbar flexion increased by 3.7%; EMG-off angles increased by 3.7% in multifidus and by 5.9% in iliocostalis. The results of Protocol A and B are consistent with the results of previous studies of these responses and demonstrate important biomechanical effects that need to be considered when modeling the lumbar spine in full or near full-flexion postures. Protocol C was a condition that had not been considered in previous studies and these results indicate that the result of a mixed effort protocol may depend on the relative intensity of the passive vs. the active fatigue. In the current study the passive tissue fatigue appears to have dominated since the results of Protocol C are somewhat similar to those seen in Protocol A. In all three protocols there appears to have been significant compromise of the passive spinal stability system, as the muscle activities in agonist muscles and synergist muscles were significantly increased in all three protocols illustrating an increased need for active control of the lumbar region.

In terms of the recovery process, the in vivo experiment, comparing characteristics of the recovery phase in three protocols, showed longer recovery time after the passive tissue elongation protocol (not fully recovered until 40 minutes of rest in all variables) than the muscle fatigue protocol (recovered after 5 minutes of resting in all variables) and the combined protocol (not fully recovered until 40 minutes of resting for the full lumbar flexion angle and the EMG-off angle; fully recovered in agonist muscle activation after 40 minutes of resting; and fully recovered in the synergist muscles after 5 minutes of resting). The results suggest that the slow recovery of the viscoelastic tissues caused by the prolonged stooping of Protocols A and C may lead to periods of spinal instability because of the abnormally lax passive tissues. While not a direct results of this study, these results may indicate an increased risk of injury during this period of passive tissue remodeling. Also, the enhanced activation in the synergist muscles (i.e., super global system) and depression in the antagonist muscles during the recovery session suggest an interaction mechanism between antagonist and synergist which may be planned in skilled motor programs before the initiation of the movement. Meanwhile, contrary to the results of passive tissues elongation protocol, the muscle fatigue protocol showed relatively quick recovery in all responses measures, but higher levels of muscle activity increase immediately after the protocol: Protocol B (agonist: 14.2%; synergist: 12.5%) vs. Protocol A (agonist: 9.2%; synergist: 4.7%) and Protocol C (agonist: 11.5%; synergist: 5.1%)). In all three protocols, the super global system (i.e., synergist) showed a recovery pattern that was quite similar to the agonist muscle response.

The results of the theoretical modeling and experimental validation components of the current study indicate that a new musculoskeletal model with a more "systems-level" perspective is necessary to fully understand the biomechanical response of the lumbar spine during full flexion and near full flexion exertion. This study has filled a void in the literature in that it addresses 1) the role of the super global system (i.e., lower extremity) in both normal and abnormal condition, 2) the effect of combined effect protocol (both laxity of the passive tissues and fatigue of the active tissues), 3) differences in the biomechanical responses as a function of the type of fatigue developed (passive tissue, active tissue, combined passive and active tissue fatigue), and 4) dynamic and variable responses of the chosen biomechanical measures during recovery. The results of this new systems-level biomechanical model can be used to develop a new EMG-assisted model of spinal loading and spinal stability as well as guidelines for designing safer working environments that can lower the risks of musculoskeletal injury to the low back.