Low levels of vitamin D and worsening of knee osteoarthritis: results of two longitudinal studies.
OBJECTIVE: To confirm reports that 25-hydroxyvitamin D (25[OH]D) deficiency is associated with an increased risk of joint space narrowing or cartilage loss in osteoarthritis (OA). METHODS: We measured 25(OH)D levels in subjects from 2 longitudinal cohort studies, the Framingham Osteoarthritis Study and the Boston Osteoarthritis of the Knee Study (BOKS). In the first, weight-bearing anteroposterior (AP) and lateral knee radiographs were obtained on subjects in 1993-1994 and again in 2002-2005 (mean interval 9 years); blood was drawn for measurement of vitamin D status in 1996-2000. In the second, subjects with symptomatic knee OA participating in a natural history study had fluoroscopically positioned semiflexed posteroanterior (PA) and lateral radiography of both knees and magnetic resonance imaging (MRI) of the more symptomatic knee performed at baseline and at 15 and 30 months. Blood was drawn at all visits, and the baseline specimen was used when available. In both studies, we defined radiographic worsening based on joint space loss in the tibiofemoral joint on either AP/PA or lateral weight-bearing views, using a semiquantitative scale (worsening defined as increase by > or =1 on a 0-3 scale). In the BOKS, we evaluated cartilage loss semiquantitatively, using the Whole-Organ Magnetic Resonance Imaging Score. In both studies, 25(OH)D levels were measured by radioimmunoassay. Analyses focused on whether vitamin D levels, defined in tertiles or as deficient (25[OH]D <20 ng/ml) versus nondeficient, predicted worsening of OA. Logistic regression analysis adjusted for age, body mass index, sex, and baseline OA level was used. RESULTS: The 715 subjects in the Framingham Study had a mean 25(OH)D level of 20 ng/ml at baseline, and 20.3% of the knees showed worsening, during the course of the study, with most knees having had no evidence of OA at baseline. The 277 subjects with OA in the BOKS had a mean 25(OH)D level of 20 ng/ml at baseline with 23.6% of knees showing radiographic worsening. We found no association of baseline 25(OH)D levels with radiographic worsening in either cohort, and confidence limits in the analyses of vitamin D deficiency were narrow, suggesting that results were not based on insufficient power. In fact, the risk of worsening was slightly, but not significantly, lower in persons with low levels of vitamin D than in persons with higher levels. In the BOKS, vitamin D levels were unrelated to cartilage loss seen on MRI. CONCLUSION: The findings indicate that vitamin D status is unrelated to the risk of joint space or cartilage loss in knee OA.