SEX-SPECIFIC CONTINUOUS JOINT SPACE WIDTH: AN ALTERNATIVE TO RHOA GRADING

Abstract

INTRODUCTION

The reported prevalence of radiographic hip OA (RHOA) varies widely in literature and depends on the specific study population. The KLG and (modified) Croft grade are commonly used to quantify RHOA. Both these scoring systems are inherently subjective, and the reproducibility is largely dependent on the expertise of the reader. Furthermore, both of these RHOA grading system emphasize different features of RHOA, making them difficult to compare. Using automated RHOA grade would reduce subjectivity and allow for fast, reproducible, and reliable assessment of radiographs. Since JSW currently demonstrates the highest reliability as a ROA describing feature, utilizing continuous JSW measurements could be a promising step towards achieving an automated RHOA grade.

OBJECTIVE

To investigate the association between baseline demographics, RHOA, and automated, continuous JSW.

METHODS

We pooled individual participant data from two prospective cohort studies within the Worldwide Collaboration on OsteoArthritis prediCtion for the Hip (World COACH consortium). Both cohorts have standardized weight-bearing anteroposterior (AP) pelvic radiographs available at baseline, 4-5 years, and 8 years follow-up. JSW measurements were automatically determined on the AP radiographs based on landmarks on the acetabular sourcil and the femoral head contour. Four different JSW measurements were determined for each hip, namely at the most medial point, in the center and at the most lateral point of the sourcil, and the minimal JSW (Fig 1). RHOA was scored by KLG or modified Croft grade. Based on the baseline and follow-up RHOA grades, the RHOA pattern of the hip was defined as “no definite RHOA” (KLG/Croft < 2 at all timepoints), “baseline RHOA” (KLG/Croft ≥ 2 at baseline), or “incident RHOA” (KLG/Croft ≥ 2 at follow-up). Hips were included for analysis if they had JSW measurements available at all three time points, and RHOA grades available at baseline and follow-up. The association between baseline age, body mass index (BMI), and the RHOA pattern, and each definition of JSW over time was estimated using linear mixed-effects models (LMMs). The analyses were stratified by sex due to known differences in JSW and OA risk in males and females. The random effects included follow-up time, cohort, and participant, accounting for the repeated measurements and cohort clustering. No RHOA was defined as the reference category for RHOA pattern. The resulting model coefficients with 95% confidence intervals (CI) were presented.

RESULTS

A total of 2,895 participants were included in the current study. 3,368 hips of 1,698 females were included, with a mean baseline age of 60 ± 8 years, a mean baseline BMI of 27.8 ± 5.0 kg/m², 4.3% had baseline RHOA, and 3.9% had incident RHOA at follow-up. The JSW narrowed on average in all four locations, and the highest prevalence of narrowing of >1 mm of 9.8% was observed for the lateral JSW (4.5% medial; 4.6% central; 3.2% minimal JSW). 2,379 hips of 1,197 males were included, with a mean baseline age of 60 ± 9 years, a mean baseline BMI of 28.4 ± 3.8 kg/m², 7.4% had baseline RHOA, and 2.0% had incident RHOA at follow-up. The JSW narrowed on average in all four locations, and again the highest prevalence was observed in the lateral JSW (5.3% medial; 5.0% central; 10.7% lateral; and 2.7% minimal JSW). The results of the four LLMs in males and females are presented in Figure 2. Older baseline age in females and lower baseline BMI in males were associated with narrower JSW over time at all locations. Baseline and incident RHOA were associated with narrower JSW over time compared to no definite RHOA for all locations except for the medial JSW.

CONCLUSION

The lateral JSW was the most susceptible to narrowing over time and showed consistent associations with baseline and incident RHOA in both males and females. These results suggest that automated, continuous JSW measurements might be a good alternative for RHOA grades. However, seeing the marked sex differences in JSW measurements of the hip, a different interpretation of JSW measurements in males and females is warranted.