Mechanical forces exerted on the knee contribute to the development of knee osteoarthritis (OA). Among these, the additional torque about the knee associated with increasing leg length may predispose people of taller stature to knee OA. However, this association has received little prior investigation. In contrast to body height, which may decrease with age, knee height is relatively constant, making knee height a more appropriate surrogate for stature in the elderly. Here, Hunter et al (Arthritis Rheum 2005;52(5):1418) investigate the association of knee height to the prevalence of knee pain and/or radiographic knee OA in an elderly population.
A random sample of male and female residents of Beijing, China of age > 60 years underwent the following assessments:
- Assessment for radiographic knee OA bilateral weight-bearing fully extended knee radiographs for tibiofemoral OA and skyline radiographs for patellofemoral OA. Tibiofemoral knee OA was defined as K/L grade > 2. Patellofemoral OA was defined as either > 1 osteophyte or joint space narrowing on skyline views.
- Assessment of knee pain in the prior 12 months via questionnaire
- Measurements knee height (seated, right knee only), body height, body weight, knee extension strength, and physical activity (via questionnaire)
1006 men and 1500 women were enrolled with a mean age of 68 years (men) and 67 years (women). Mean body mass index (BMI) was 25.3 for men and 26.0 for women.
|Men (n=1006)||Women (n=1500)|
|Mean knee height||49.6 + 2.2 cm||45.7 + 1.9 cm|
|Radiographic OA + knee pain||9.7%||20.3%|
Men in the highest quartile of knee height (mean knee height 52.5 cm) were significantly more likely to demonstrate patellofemoral and tibiofemoral OA than men in the lowest quartile of knee height (patellofemoral: OR 1.4(95% CI 1.1-1.8); p for trend = 0.002, tibiofemoral: OR 1.3 (95% CI 1.0-1.6; p for trend = 0.031). However, increasing knee height was not significantly associated with symptomatic knee OA in men.
In women, the association of knee height and radiographic OA was stronger than in men. Women in the highest quartile of knee height (mean knee height 48.2 cm) were significantly more likely to demonstrate patellofemoral and tibiofemoral OA than women in the lowest quartile of knee height (patellofemoral: OR 1.7(95% CI 1.5-2.0); p for trend = <0.0001, tibiofemoral: OR 1.7 (95% CI 1.5-2.0; p for trend = <0.0001). In contrast to men, increasing knee height was significantly associated with symptomatic knee OA in women, with women in the highest quartile of knee height demonstrating a 2.2 fold greater odds of symptomatic knee OA than women in the lowest quartile of knee height. Similar results were obtained when knee height as a proportion of body height was utilized.
In women without radiographic OA, increasing knee height was associated with increasing knee pain (OR 1.8 (95% CI 1.3-2.5) for the comparison of highest to lowest quartile of knee height. No significant association between increasing knee height and knee pain was found in men without radiographic OA.
Conclusions: Among elderly Chinese men and women, increasing knee height is associated with prevalent tibiofemoral and patellofemoral OA. In addition, increasing knee height is associated with prevalent symptomatic knee OA and knee pain with no radiographic evidence of OA in Chinese women.
Editorial Comment: This study is one of the first to confirm an association between taller stature and knee OA. Because the association is somewhat modest, a large sample size was required to detect differences between patients of different knee heights. In addition, as with any cross-sectional investigation, these findings fall short of proving any causal relationship between knee height and the development of knee OA. A longitudinal study is certainly appropriate to further examine the risk of taller stature on the development of OA, particularly as significant gender differences were seen here. In particular, the factors contributing to increased knee pain and symptomatic knee OA in taller women but not in men require further exploration, as the men in the cohort presumably had longer tibia and femur lengths than the women and thus would be expected to have much greater torque about the knee.
Chinese men and women have been shown to have a greater prevalence of knee OA than other ethnic groups. Cultural differences in stooping, kneeling, and portage may explain these difference and may disproportionately affect those of taller stature. Thus, the generalizability of these findings may not necessarily apply to other populations. Nevertheless, the results are compelling and contribute to the current understanding of the mechanics of knee OA.