Objective: To assess the genetic contribution to radiographic hip osteoarthritis (OA) by measuring the distribution of disease features in monozygotic (MZ) and dizygotic (DZ) twins.
Methods: A population-based, cross-sectional study was conducted of 135 MZ and 277 DZ healthy female twin pairs, 50 years of age and older, who were recruited into the St. Thomas' UK Adult Twin Registry. Pelvic radiographs were read by a single observer who was blinded to the pairing and zygosity of the twins. The films were assessed for overall OA grade using a modification of the Kellgren and Lawrence scheme, and assessed for individual radiographic features.
Results: There was evidence of significant familial clustering for grade I and grade II OA changes, with an excess concordance in MZ twins compared with DZ twins, suggesting a genetic effect. The MZ versus DZ excess was also apparent for those classified as having more severe disease, although the number of pairs with these disease features was small. Familial clustering attributable to genetic factors was evident for joint space narrowing of <2.5 mm. Familial, but not genetic, clustering was seen for subchondral sclerosis. The number of pairs concordant for definite osteophytes in the sample was too low to assess this feature alone. These results translate into a significant heritability of 58% for OA overall and 64% for joint space narrowing. The heritability estimates decreased a little when the potential confounding influences of age, body mass index, and hip bone density were taken into account.
Conclusion: Genetic factors have a significant contribution to OA at the hip in women and account for approximately 60% of the variation in population liability to the disease.