Throwing with power and accuracy is a uniquely human behavior and a potentially important mode of early hunting. Chimpanzees, our closest living relatives, do occasionally throw, although with much less velocity. At some point in our evolutionary history, hominins developed the ability to produce high performance throws. The anatomical changes that enable increased throwing ability are poorly understood and the antiquity of this behavior is unknown. In this thesis, I examine how anatomical shifts in the upper body known to occur during human evolution affect throwing performance. I propose a new biomechanical model for how humans amplify power during high-speed throwing using elastic energy stored and released in the throwing shoulder. I also propose and experimentally test a series of functional hypotheses regarding how four key shifts in upper body anatomy affect throwing performance: increased torso rotational mobility, laterally oriented shoulders, lower humeral torsion, and increased wrist hyperextensability. These hypotheses are tested by collecting 3D body motion data during throws performed by human subjects in whom I varied anatomical parameters using restrictive braces to examine their effects on throwing kinematics. These data are broken down using inverse dynamics analysis into the individual motions, velocities, and forces acting around each joint axis. I compare performance at each joint across experimental conditions to test hypotheses regarding the relationship between skeletal features and throwing performance. I also developed and tested a method for predicting humeral torsion using range of motion data, allowing me to calculate torsion in my subjects and determine its effect on throwing performance. My results strongly support an important role for elastic energy storage in powering humans’ uniquely rapid throwing motion. I also found strong performance effects related to anatomical shifts in the torso, shoulder, and arm. When used to interpret the hominin fossil record, my data suggest high-speed throwing ability arose in a mosaic-like fashion, with all relevant features first present in Homo erectus. What drove the evolution of these anatomical shifts is unknown, but as a result the ability to produce high-speed throws was available for early hunting and likely provided an adaptive advantage in this context. / Anthropology

Background: Baseball players exhibit a more posteriorly oriented humeral head or humeral retrotorsion (HRT) in the dominant arm, likely representing an adaptive response to the stress of throwing. This adaptation is thought to occur while skeletally immature, however there is limited research detailing how throwing while young influences the development HRT. In addition, it is currently unclear how this changing osseous orientation influences shoulder motion within young athletes. Purpose: To determine the influence of throwing and age on the development of asymmetry in HRT and shoulder range of motion (ROM); and analyze the relationship between HRT and ROM. Study Design: Cross-sectional age matched study Methods: Healthy athletes (8-14 years-old) were categorized into two groups based upon sports participation; throwing group (n=85) and non-throwing group (n=68). Bilateral measurements of HRT, shoulder external (ER), internal rotation (IR) and total range of motion (TROM) at 90° were performed using diagnostic ultrasound and digital inclinometer. A two-way analysis of variance was performed with throwing status (yes/no) and age group (youth (8-10.5), junior (10.51-12) and senior (12.01-13.99)) as primary factors. Dependent variables were asymmetry (dominant-non-dominant) in HRT,ER, IR and TROM. The relationship between ROM and HRT was analyzed using Pearson correlation coefficients. Results: Throwing athletes demonstrated a larger degree of HRT on the dominant side, resulting in greater asymmetry (8.7° versus 4.6°). Throwing athletes demonstrated a gain of ER (5.2°), a loss of IR (6.0°) and no change in TROM when compared to the non-dominant shoulder. Pairwise comparisons identified altered HRT and shoulder ROM in all age groups of throwers. A significant but weak relationship between HRT and shoulder ROM existed. Conclusion: Throwing causes adaptive changes in HRT and shoulder ROM in youth baseball players at a very young age. Other factors in addition to HRT influence shoulder motion within this population. Clinical Relevance: In baseball players, an altered arc of motion can be expected at a young age. This adaptation is in part due to changes in osseous structures, however a larger component of change is likely due to other factors.

This project investigas humeral torsion and activity-related change in the human upper limb. Increased humeral torsion angles have been identified in the professional throwing athlete and may be associated with strenuous activity. The nature of humeral torsion as an osteogenic response to the strain environment is investigated to identify its role in the behavioural morphology of the upper limb. These physical manifestations of strenuous physical activity provide an insight into the make-up of medieval armies prior to the establishment of standing armies. Populations analysed include two blade-injured samples, Towton and a subsample of blade-injured men from the Priory of St. Andrew, Fishergate, York. The men from the Mary Rose, a Tudor warship are also investigated. Other samples analysed include the rural sites of Wharram Percy and Hickleton, the urban cemeteries from the Priory of St. Andrew, Fishergate,York and the leprosarium of Sts. James and Mary Magdalene, Chichester, the modern cadaver-based Terry collection and non-human primates, Gorilla sp., Pan sp., Pongo sp., and Macaca sp.. Measurement of the humeral torsion angle and external measurements and indices of architecture, articulations and robusticity are employed. Cross-sectional geometric properties are investigated using CT imaging of the paired humeri from a sub-sample of blade-injured individuals and a comparative sample of those who were not. Bilateral asymmetry is investigated to identify the role of plasticity within the humerus and to reveal aspects of limb dominance. The results are compared with non-human primate species to obtain insight into inter-species differences. Results indicate the humeral torsion is not ontogenetically constrained, but is highly variable between and within populations, individuals and even between sides. Biomechanical analyses indicate that in the Towton population, humeral torsion may serve as part of a two-stage adaptation, in which the architecture is modified to enable greater biomechanical efficiency in distributing strain, reducing the need of increased cortical thickness. Changes in humeral torsion related to strenuous activity have been identified, although in the blade-injured samples it is decreased torsion angles, w hile in the comparative sample it is increased torsion angles that significantly correlate with limb hypertrophy. Humeral torsion appears to be influenced by other measurementd of humeral architecture, specifically, the amount of anterior bowing and anterior curvature to the distal humeral shaft. This work demonstrates the need for individual rather than population-based analyses, as the heterogeneity within population samples obscures individual variation in activity patterns. This analysis provides baseline data for typical populations of the Middle Ages. From this, it is then possible to investigate the individual within this baseline, to identify those who stand out from their samples through habitual, strenuous activity patterns. Movement patterns identified related to warfare include those consistent with the use of the longbow in the Towton sample and the use of a sword in the Fishergate blade-injured sample. These men, and those of the Mary Rose, appear to have either been selected for combat based on size, or benefited from a more nutritious diet during growth.

This project investigas humeral torsion and activity-related change in the human upper limb. Increased humeral torsion angles have been identified in the professional throwing athlete and may be associated with strenuous activity. The nature of humeral torsion as an osteogenic response to the strain environment is investigated to identify its role in the behavioural morphology of the upper limb. These physical manifestations of strenuous physical activity provide an insight into the make-up of medieval armies prior to the establishment of standing armies. Populations analysed include two blade-injured samples, Towton and a subsample of blade-injured men from the Priory of St. Andrew, Fishergate, York. The men from the Mary Rose, a Tudor warship are also investigated. Other samples analysed include the rural sites of Wharram Percy and Hickleton, the urban cemeteries from the Priory of St. Andrew, Fishergate,York and the leprosarium of Sts. James and Mary Magdalene, Chichester, the modern cadaver-based Terry collection and non-human primates, Gorilla sp., Pan sp., Pongo sp., and Macaca sp.. Measurement of the humeral torsion angle and external measurements and indices of architecture, articulations and robusticity are employed. Cross-sectional geometric properties are investigated using CT imaging of the paired humeri from a sub-sample of blade-injured individuals and a comparative sample of those who were not. Bilateral asymmetry is investigated to identify the role of plasticity within the humerus and to reveal aspects of limb dominance. The results are compared with non-human primate species to obtain insight into inter-species differences. Results indicate the humeral torsion is not ontogenetically constrained, but is highly variable between and within populations, individuals and even between sides. Biomechanical analyses indicate that in the Towton population, humeral torsion may serve as part of a two-stage adaptation, in which the architecture is modified to enable greater biomechanical efficiency in distributing strain, reducing the need of increased cortical thickness. Changes in humeral torsion related to strenuous activity have been identified, although in the blade-injured samples it is decreased torsion angles, w hile in the comparative sample it is increased torsion angles that significantly correlate with limb hypertrophy. Humeral torsion appears to be influenced by other measurementd of humeral architecture, specifically, the amount of anterior bowing and anterior curvature to the distal humeral shaft. This work demonstrates the need for individual rather than population-based analyses, as the heterogeneity within population samples obscures individual variation in activity patterns. This analysis provides baseline data for typical populations of the Middle Ages. From this, it is then possible to investigate the individual within this baseline, to identify those who stand out from their samples through habitual, strenuous activity patterns. Movement patterns identified related to warfare include those consistent with the use of the longbow in the Towton sample and the use of a sword in the Fishergate blade-injured sample. These men, and those of the Mary Rose, appear to have either been selected for combat based on size, or benefited from a more nutritious diet during growth. / Arts and Humanities Research Board (AHRB), Francis Raymond Hudson Fund, Andy Jagger Fund.