Background Ballet loads the hip joint in extreme ranges of movement (ROM). Abnormal bony morphology and joint laxity may enable this movement, but predispose dancers to hip injury.

Objective To compare hip injury and bony morphology between ballet dancers and non-dancing athletes, and to investigate relationships between these factors and clinical features.

Design Blinded, case-control study.

Participants 49 professional ballet dancers and 49 age- and sex-matched non-dancing athletes.

Setting Professional ballet company and sporting clubs.

Assessment of Risk Factors Lateral centre edge angles, anterior and superior alpha angles, femoral neck-shaft angles and acetabular version angles measured with 3.0-Tesla magnetic resonance imaging (3-T MRI), ROM and hypermobility testing.

Main Outcome Measurements Cartilage, labral, ligamentum teres (LT) lesions and effusion scored on MRI, and HAGOS (Copenhagen Hip and Groin Outcome Score).

Results Cartilage defect and labral tear prevalence was similar in dancers and athletes (p>0.05); LT tears were more prevalent in dancers compared to athletes (p=0.001). Dancers had higher neck-shaft angles (p=0.002); lower acetabular version angles (p=0.003); lower superior alpha angles (p<0.001); similar anterior alpha angles (p=0.2); and similar lateral acetabular coverage (p=0.07), compared to athletes. There was no significant association between the morphological parameters and cartilage, labral, LT injury or effusion (p>0.01 for all). All 6 dancers with cam morphology (α angle >50°) measured anteriorly had cartilage defects, and 5 had labral tears; a trend not observed in athletes. Joint effusions were larger in dancers compared to athletes (p=0.009), and the only factor related to lower HAGOS pain and sports/recreation scores in female dancers (p=0.001 for both).

Conclusions Professional ballet dancers had a high incidence of LT tear, and hip bony morphology that may allow extreme motion. Dancers with cam morphology had cartilage defects, but joint effusion was the only correlate with symptoms. Longitudinal studies are required to determine the clinical sequelae.