This dissertation evaluates how workspace design impacts muscular demand and joint loading during dynamic submaximal push-pull tasks. Work involving extensive pushing and pulling is associated with higher frequency of shoulder complaints. While reports of shoulder muscle demand during submaximal isometric tasks are abundant, dynamic submaximal push-pull exertions are not well understood. First, the effects of task type and task target on muscle demand (weighted EMG average) of surface glenohumeral muscles were evaluated. Seventeen healthy young adults performed unimanual and bimanual pushes and pulls to 3 thoracohumeral elevations (20º, 90º, 170º) and 4 elevation planes (0º, 45º, 90º, 135º) with loading at 15% of isometric push-pull capacity. Pulling required less demand than pushing (p<0.0001). Muscle demand varied more with elevation than elevation plane. The lowest target had highest demand for pulling (p<0.01), and the most elevated target had highest demand for pushing (p<0.0001). Working above the shoulder is known to increase demand during isometric tasks, however, these results suggest that for dynamic tasks working against gravity has a larger effect on demand than task target.

Next, experimental data was used to inform computational simulations of push-pull tasks to evaluate the spatial dependency of glenohumeral stability. Degenerative wear to the glenoid from repetitive loading can reduce effective concavity depth and lead to future instability. Therefore, workspace design should consider glenohumeral stability to prevent initial wear. While glenohumeral stability has been previously explored for activities of daily living including push-pull tasks, whether stability is spatially dependent is unexplored. A subset of the bimanual and unimanual push-pull tasks were simulated including tasks to 4 horizontal targets (planes of elevation:​ 0º, 45º, 90º, and 135º) at 90º thoracohumeral elevation and 3 elevation targets (thoracohumeral elevations:​ 20º, 90º, 170º) at 90º plane of elevation. The lateral 45º horizontal target was most stable regardless of exertion type and would be the ideal target placement when considering stability. This target is likely more stable because the applied load acts essentially perpendicular to the glenoid, limiting shear force production. The cross-body 135º horizontal target was particularly unstable for unimanual pushing, and the applied force direction for this task is essentially parallel to the glenoid, likely creating shear forces. Pushing was less stable than pulling (all targets except sagittal 170º for both unimanual and bimanual and horizontal 45º for bimanual) (p<0.01), which is consistent with prior reports. There were limited stability benefits to task placement for pushing, and larger stability benefits may be seen from converting push tasks to pull rather than optimizing task layout. There was no difference in stability between bimanual and unimanual tasks, suggesting no stability benefit to bimanual operation.

Lastly, a comparison was made among modeling techniques that attempt to account for active glenohumeral stabilization in inherently stable musculoskeletal models. Most upper extremity musculoskeletal models represent the glenohumeral joint with an inherently stable ball-and-socket, but the biological joint requires active muscle coordination for stability. Sensitivity of common predicted outcomes (net glenohumeral reaction force, rotator cuff activations, and instability) to different implementations of active stabilizing mechanisms (constraining net joint reaction direction and incorporating surface EMG) were evaluated. Both EMG and reaction force constraints successfully reduced joint instability, however other outcomes were more sensitive to EMG constraints, which also tended to overconstrain the model, leading to poor tracking and simulation failure. Therefore, force constraints may be more robust when representing stability.

This dissertation provides new information characterizing the spatial dependency of muscular demand and glenohumeral stability during dynamic submaximal push-pull tasks. Furthermore, the evaluation of modeling techniques lays the foundation for future studies evaluating glenohumeral instability for other functional tasks.