Objective: The prevalence of hip osteoarthritis (OA) increases significantly with age. Although it is not clear whether joint space loss at the hip is a feature of normal aging or a reflection of the OA process, epidemiologic criteria for OA are based on narrowing alone. The aim of this study was to determine whether changes in joint space width occur with age, and whether there are sex differences, in asymptomatic subjects without hip OA.
Methods: We identified a total of 1,806 subjects who had undergone intravenous urography between 1994 and 1996 and sent a questionnaire to the 1,527 of these subjects who were alive in 1998; 1,031 replies (68%) were received. All radiographs were read by an observer blinded to age, sex, and pain status. Individual radiographic features of OA (narrowing, osteophyte, sclerosis, and cysts) were graded, and an overall qualitative grade was allocated, according to a standard atlas. Minimum joint space width (JSW) was measured by metered caliper to within 0.1 mm. A total of 276 women (mean age 63 years) and 257 men (mean age 64 years) were identified who had never had hip pain (defined as having ever had pain on most days for at least 1 month) and who had no evidence of either joint space narrowing or osteophyte (grade 0, no structural changes). The minimum JSW in either hip was tabulated according to age.
Results: JSW measurement was reproducible (95% confidence limits of agreement) to within +/-0.5 mm. At all ages, men had larger JSW than women (3.85 mm in women, 4.19 mm in men, mean difference 0.34 mm; 95% confidence interval [95% CI] 0.24, 0.44). A significant decline in JSW with age was seen in women, with a mean difference between ages 45-54 and 75-84 years of 0.36 mm (95% CI 0.15, 0.58; P = 0.001). No significant change in JSW with age was seen in men (mean difference 0.16 mm; 95% CI -0.11, 0.43). Analysis of an additional 64 women and 61 men who were without hip pain and had overall qualitative grade 1-2 changes gave similar results. Implementing these results to alter the threshold for definition of hip OA in women from < or =2.5 mm to < or =2.2 mm reduced the prevalence of hip OA from 10.6% to 5.6%.
Conclusion: These sex differences in joint space have significant implications in terms of the major emphasis on joint space narrowing in definitions of hip OA. Women also have a significant progressive decline in joint space with age that is not seen in men. This suggests that in women, loss of cartilage may be an age-related phenomenon that is independent of other aspects of structural change. Consideration should be given to the development of sex-specific definitions of hip "OA."