The changes of bone mineral density on the risk of progression of osteoarthritis of the knee

Aim. To find the relationship between bone mineral density (BMD) and risk of knee OA progression in a 5-year prospective study. Materials and methods. 110 females with knee OA were examined twice with 5-year interval. Examination included filling questionnaires, VAS pain assessment, plain knee radiography and axial skeleton densitometry. I stage knee OA was established in 33 (30%) patients, II stage - in 46 (41.8%), III stage - in 26 (23.6%), and IV - in 5 (4.5%). Normal lumbar vertebrae densitometry BMD values were found in 45 patients (40.9%), osteopenia-corresponding BMD values - in 33 (30.0%), and osteoporosis - in 32 (29.1%). Normal femoral neck BMD values were identified in 60 (54.5%) patients, osteopenia-level BMD - in 48 (43.7%), osteoporosis - in 2 (1.8%). In all premenopausal patients (n=15) axial skeleton BMD values were normal. Results. In 5-year interval radiographic progression was established in 40 patients (Group 2), while in 70 (Group 1) patients no progression occurred. Both groups were comparable in terms of age and disease duration, although, more patients from Group 2 tended to have normal baseline densitometry BMD values - both in lumbar vertebrae and femoral neck: 47.5% vs 37.1%, and 62.5% vs 44.3% as compared to Group 1 patients. Patients from Group 1 more often had BMD values corresponding to osteoporosis and osteopenia: 32.9% vs 22.5%, and 55.7% vs 37.5%, respectively, as compared to Group 2 patients, although not achieving statistical significance. These differences were still identifiable after 5-year interval. Absolute BMD values at the second examination in 5 years were indicative of statistically significant increase in femoral neck and total hip BMD in Group 2 patients with knee OA progression: 0.79 +/- 0.11 vs 0.73 +/- 0.16, p<0.01, and 0.93 +/- 0.14 vs 0.84 +/- 0.25, p<0.05, respectively. Thorough analysis of lumbar vertebrae BMD (g/cm(2)) relationship with OA stages revealed that in patients with stage IV OA lumbar BMD values were significantly higher than in patients with stages I-III OA (stage I OA - BMD 0.87 +/- 0.12 g/cm(2); stage II OA - 0.92 +/- 0.21 g/cm(2) ; stage III OA - 0.88 +/- 0.13 g/cm(2), stage IV OA - BMD 1.07 +/- 0.17 g/cm(2)). Femoral BMD values didn't show evident correlation with knee OA stage, although there was a trend towards higher BMD values in patients with stage IV OA compared to stage III OA (p=0.06). Total hip BMD values were quite similar to lumbar BMD values (p=0.01). BMD values were statistically significantly higher in patients with stage IV OA, than in patients with stages I and III (respectively, IV - 0.98 +/- 0.13 g/cm(2), I - 0.85 +/- 0.10 g/cm(2) and III - 0.86 +/- 0.16 g/cm(2)). Correlation analysis also confirmed direct correlation between knee OA stage and BMD values in all evaluated compartments (p<0.05). Conclusion. Increasing during the 5-eyar follow up period femoral neck and total hip BMD values can be interpreted as the predictor of knee PA progression. More advanced OA stages are associated with higher BMD values.