Is 'hand preference' coded in the hominin skeleton? An in-vivo study of bilateral morphological variation

The presumed link between bilateral asymmetry and lateralized habitual activity in extinct hominins is the basis upon which inferences of 'hand preference' often derive. While this presumption is reasonable, in-vivo comparisons of skeletal asymmetries and self-reported handedness are rare, and as a result the accuracy of these inferences is questionable. To assess this relationship in living humans, reported 'handedness' was compared against peripheral quantitative computed tomography (pQCT) derived bilateral measurements of humeral, ulnar, and tibial midshaft cortical area (CA) and torsional rigidity (J). Significant bilateral differences were found in the humerus for all groups, and in the ulna for the cricketer and field hockey sub-samples. Additionally, cricketers' non-dominant tibiae were more robust than their dominant tibiae. An assessment of 'Dominance Asymmetry' revealed that measures of CA and J were higher in the dominant humeri in ∼90% of participants; in the ulna this was true in ∼75% of cases, and in the tibia CA and J were higher in the dominant limb less than 50% of the time. Comparisons of (self)'Reported' hand preference against 'Predicted' hand preference (based on the calculation of % Directional Asymmetry) revealed a low level of error for predictions based on both humeral (∼4-5% error) and ulnar (6-11% error) asymmetry. Error was decreased with the exclusion of individuals displaying less than 2.5-5% asymmetry. Contrarily, predictions based on tibial analyses had a much higher level of 'error' (∼45%). Overall, the results support currently accepted approaches for inferring 'hand preference' from measures of upper limb geometric asymmetry in the hominin skeleton.