| Summary: | Purpose: Variation in morphology in the proximal femur and pelvis (e.g. acetabular dysplasia, non-spherical femoral head) can biomechanically compromise the hip joint and predispose to hip osteoarthritis (OA). Such morphological variation may in part explain the heritability of hip OA. The objective of this study was to evaluate a range of 2-dimensional morphological measures on standard radiographs to determine: normal range, right: left symmetry, age and gender differences; and to investigate whether they are associated with the risk of hip OA. Methods: A nested case control study was undertaken in 566 unilateral hip OA cases and 1108 controls in the established Nottingham Genetics of Osteoarthritis and Lifestyle (GOAL) database. Unaffected hips of unilateral hip OA cases were compared to the normal controls, under the assumption that similar morphological features would be observed for the affected hips prior to the development of hip OA. Definition of radiographic hip OA was joint space width (JSW) ~ 2.5 mm. Standardized antero-posterior (AP) radiographs of the pelvis were used to measure the morphological features. Measurements were performed by a single observer and the reproducibility was evaluated at baseline, mid and end of the study. Normal values, thresholds (mean±1.96SD) and symmetry of the features were derived from the control subjects. The intra-observer reliability was examined using intra-class correlation coefficient (ICC). Odds ratio (OR) and 95% confidence interval (Cl) were calculated for association. Logistic regression was used to adjust for age, gender and body mass index (BM!). Measurements were divided into tertiles to examine dose response. ii Results: The intra-observer reliability
|
|---|
