Abstract
Objective. To determine if knee alignment measures differ between African Americans and Caucasians without radiographic knee osteoarthritis (rOA).
Methods. A single knee was randomly selected from 175 participants in the Johnston County Osteoarthritis Project without rOA in either knee. Anatomic axis, condylar, tibial plateau, and condylar plateau angles were measured by 1 radiologist; means were compared and adjusted for age and body mass index (BMI).
Results. There were no significant differences in knee alignment measurements between Caucasians and African Americans among men or women.
Conclusion. Observed differences in knee rOA occurrence between African Americans and Caucasians are not explained by differences in static knee alignment.
Static knee malalignment contributes to osteoarthritis (OA) progression at the knee, with varus alignment leading to increased odds of medial compartment progression, and valgus alignment to more lateral progression1. The extent of malalignment is related to magnitude of joint space narrowing (JSN), accelerated functional decline, and increased pain, over 18 months of followup1. There may be an association between malalignment and incident radiographic knee OA (knee rOA), although data are inconsistent2–4.
Known racial/ethnic differences in knee OA prevalence5,6 suggest that race/ethnicity and alignment might be related. Lateral knee rOA was more common among elderly Chinese compared to Caucasians; greater valgus alignment was identified as a potential explanatory factor in a subsample of 100 individuals without knee rOA5. Recently, a larger followup study (n = 570) evaluated static knee alignment indices among Chinese and Caucasian subjects without knee rOA, confirming greater valgus alignment among Chinese subjects7. We have previously identified increased odds of lateral JSN among African American men [odds ratio (OR) 2.19, 95% confidence interval (CI) 1.32–3.65] and African American women (OR 1.48, 95% CI 1.02–2.16) compared to their Caucasian counterparts8. The purpose of this study was to assess whether differences exist in indices of static knee alignment between African Americans and Caucasians without knee rOA that might contribute to the increased risk of lateral JSN at the knee in African Americans.
MATERIALS AND METHODS
This cross-sectional analysis used baseline visit (1991–97) data from the Johnston County Osteoarthritis Project, described in detail elsewhere6. In brief, this is a population-based study of OA in rural North Carolina among African Americans and Caucasians aged 45 years and older. All individuals underwent anteroposterior, weight-bearing, extended knee radiography, and all radiographs were read for Kellgren-Lawrence (K-L) grade by a single musculoskeletal radiologist (JBR) for whom inter- and intrarater reliability were high (kappas 0.86 and 0.89, respectively)9. Participants without knee rOA (n = 175) were randomly selected by sex and race; a single knee from each individual was randomly selected for study (115 knees with K-L grade = 0; 60 knees with K-L grade = 1). A difference of 2 degrees in anatomic axis was chosen to calculate sample size based on the observed differences in the Beijing and Framingham samples7. Using a 2-sided 0.05 significance level, a sample size of 43 knees per group (African American men and women, Caucasian men and women) was needed to achieve 90% power to detect a 2 degree difference in anatomic axis (this sample size provides 68% power to detect a 1.5 degree difference).
Alignment measurements were performed by a trained musculoskeletal radiologist (ABB), based on the method described by Harvey, et al7 and Cooke, et al10. Intrareader reliabilities (38 films) using intraclass correlation statistics were excellent (0.85 to 0.97). Measurements were made by hand directly on 14 × 17 inch knee films, a method comparable to full-length limb radiographs11 and to digital alignment measurements12. The following angles (degrees) were assessed (Figure 1): (a) Anatomic axis: angle between lines drawn from the visual center of the femur and tibia at a point 10 cm from the joint line, through visual midpoint of the tibial spines; (b) Condylar angle: angle between a line tangent to the distal end of the femur (condylar line) and the line through visual center of the femur; (c) Tibial plateau angle: angle between a line tangent to the lateral aspect of the tibial plateau (tibial plateau line) and the line through the visual center of the tibia; and (d) Condylar-plateau angle: angle formed by the condylar line and the tibial plateau line.
Statistical analyses used SAS version 9 (SAS Institute, Cary, NC, USA). Means for each continuous measurement were compared using analysis of covariance, adjusting for age and body mass index (BMI); all analyses were stratified due to known sex differences5.
RESULTS
There were 92 male knees and 83 female knees included in the study (Table 1). The mean age for men was 59, and for women 62 years; age did not differ by race within sexes. The mean BMI was 27 kg/m2 for men; Caucasian men were significantly heavier than African American men (p = 0.04). For women, the mean BMI was 28 kg/m2; African American women were significantly heavier than Caucasian women (p = 0.003). For Caucasian and African American knees, the proportion with K-L grade 1 was similar among men (27% vs 31%, respectively) and women (43% vs 37%).
The mean anatomic angle was > 180 degrees for all groups, and greater in men than in women (Table 1). No differences were seen by race, in men or women, in any of the knee angles or standard deviations (SD) (Figure 1, Table 1).
DISCUSSION
We did not identify any significant differences in measures of static knee alignment or their SD among African Americans and Caucasians without knee rOA. Our overall results were qualitatively similar to those reported by Harvey, et al for the Framingham cohort7. That group identified more valgus alignment among subjects from Beijing without knee rOA, with a larger SD, in comparison to Framingham subjects, and concluded that the Beijing subjects’ tendency toward valgus alignment may explain the observed increase in lateral rOA among Beijing compared to Framingham participants7. In our study, we found no significant differences by race within sexes. The differences observed in our study were extremely small, ranging from 0.1 to 0.5 degrees, and would be highly unlikely to be clinically meaningful or reach statistical significance even in a much larger sample. Therefore, it seems unlikely that variations in static knee alignment are responsible for the differences in knee rOA pattern or prevalence we have previously observed among African Americans and Caucasians. We did not assess symptoms in the current analysis, and full-length radiographs were not available, so variations in proximal or distal anatomy could have been missed13. However, the methods used were directly comparable to other published studies. Other potential factors, such as hip anatomic differences and dynamic alignment4, need to be explored further, ideally in a young, healthy population, to better understand the race differences in knee rOA patterns and prevalence.
Acknowledgments
Thanks to the Johnston County Osteoarthritis Project staff and participants, and to Erik Myers for his help with the figure.
Footnotes
-
The Johnston County Osteoarthritis Project is funded by the Centers for Disease Control and Prevention/Association of Schools of Public Health S043, and S3486. Dr. Nelson’s support is through the Rheumatology Fellowship Training Grant T-32-AR007416 and a John A. Hartford Foundation CoE Fellowship Grant. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
- Accepted for publication April 7, 2009.