Hip shape is an important determinant of hip osteoarthritis (OA) and osteoporotic hip fracture; however, little is known about its development in childhood and adolescence. In this thesis, I aimed to address this question by exploring factors influencing hip shape in adolescence.
Hip DXA scans were collected in offspring from the Avon Longitudinal Study of Parents and Children, at mean ages of 13.8 and 17.8 years. These images were analysed in Shape software to quantify hip morphology using a 53-point Statistical Shape Model (SSM). Principal component analysis was used to generate independent modes of variation (hip shape modes (HSMs) and scores from an adult reference SSM (based on 19,379 images) were applied to adolescent data to aid comparability between the time points. The top ten HSMs were selected as outcomes and associations with sex, tempo (marker of pubertal timing) and BMD were examined, and genome-wide association study (GWAS) explored genetic influences on hip shape at both time points.
Several HSMs showed sex differences. At age 14, there was strong evidence of relationships between tempo and hip shape, particularly in males, whereas at age 18 these associations were weaker in both sexes. Consistent associations between BMD and hip shape were observed in males and females, and at both time points. GWAS identified five independent variants associated with hip shape (p<5x10-8).
These findings suggest that sexual dimorphism in hip shape can be discerned in early adolescence, and that puberty is a critical period for hip shape development. Observed relationships between BMD and hip shape could be partially explained by shared genetic factors underlying these traits. Finally, genetic variants implicated in endochondral bone formation, which were previously found to be associated with adult hip shape, appear to influence hip shape in adolescents, consistent with a contribution of longitudinal growth to hip shape.