Methods
Healthy men and women free of radiographic knee OA or recent knee pain were recruited through advertisements in newspapers, health clubs, and from hospital staff to undergo:
- Measures of body composition:
- Calculated body mass index (BMI)
- Quantification of total and regional body fat and lean tissue mass by total body DEXA scanning
- Quantification of knee cartilage volume of the dominant knee by MRI
A repeat knee MRI was obtained approximately 2 years after baseline MRI. The effect of baseline parameters of body composition on the change in medial- and lateral-tibial knee cartilage volume and on the rate of change in medial- and lateral-tibial knee cartilage volume at follow-up was determined with adjustment for potential confounding variables (age, gender, baseline BMI, medial/lateral tibial bone area, and physical activity).
Results
86 subjects (equally divided between men and women) underwent baseline evaluations and 70 (81%) of enrolled subjects underwent a follow-up assessment (mean 1.9 + 2.0 years after baseline). The mean age of participants was approximately 55 years with a mean BMI of approximately 26 kg/m2. Despite entry criteria excluding subjects with knee OA, 7 (8%) of enrolled subjects were designated as having knee OA at entry. While mean BMI was not significantly different between male and female participants, male subjects were generally heavier with higher mean muscle mass, lower mean fat mass, and larger mean knee cartilage volumes than female subjects.
| Parameters of body composition with significant associations in multivariate models | |||
| Parameter | Regression coefficient (95% CI) | P value | |
| Medial tibial cartilage volume at follow-up | Total lower limb muscle mass | 55.5 (21.7-87.2) | 0.001 |
| Total muscle mass in all limbs | 46.1 (18.8-73.3) | 0.001 | |
| Total body muscle mass | 20.8 (6.6-35.0) | 0.005 | |
| Change in medial-tibial cartilage volume | Total lower limb muscle mass | -46.5 (-68.8- -23.5) | <0.001 |
| Total muscle mass in all limbs | -39.2 (-57.5- -21.0) | <0.001 | |
| Total body muscle mass | -16.8 (-25.5- -8.2) | <0.001 | |
| Lateral-tibial cartilage volume at follow-up | No significan associations | — | — |
| Change in lateral-tibial cartilage volume | Total lower limb muscle mass | -36.0 (-58.8- -13.3) | 0.003 |
| Total muscle mass in all limbs | -29.4 (-47.7- -11.1) | 0.002 | |
| Total body muscle mass | -14.5 (-22.4- -6.7) | 0.001 | |
Other measures of body composition (BMI, total body fat mass, total lower limb fat mass, total fat mass in all limbs, and truncal fat mass) were not significantly associated with medial/lateral cartilage volume or change in medial/lateral cartilage volume in multivariate analysis.
Conclusions
Among healthy adults with a low prevalence of knee OA, increasing total and lower extremity muscle mass at baseline is associated with a greater medial (but not lateral) knee cartilage volume at follow-up. Increasing total and lower extremity muscle mass at baseline is also associated with a decreased rate of change in medial- and lateral-tibial cartilage volumes.
Editorial Comment
This study is one of the first to demonstrate strong associations between decreased total and lower extremity muscle mass and changes in knee cartilage that may lead to the development of knee OA. Importantly, neither BMI nor distributed body fat mass was found to be significantly associated with changes in knee cartilage volume. Although these findings suggest that interventions to increase total and lower extremity muscle mass before the development of knee OA might provide a protective effect, this study itself cannot provide proof of a causal relationship between reduced muscle mass and knee OA. However, these results do provide a foundation on which further controlled, interventional trials can be undertaken.
A notable strength of this study is the use of MRI for the measurement of knee cartilage volume. This imaging modality allows for accurate determinations of knee cartilage volumes while requiring fewer subjects and shorter follow-up than studies using measurements from plain radiographs as outcomes. However, these results make the assumption that declining knee cartilage volume in people without baseline knee OA predicts the development of knee OA. Although declining knee cartilage volume has been shown to be a strong predictor of disease progression in people with established OA, and it seems logical to assume that this is the same for people without OA, to date, this relationship has yet to be investigated.
