Background: The effects of nonsteroidal anti-inflammatory drugs (NSAIDs) on knee osteoarthritis progression are unclear. The aim of this longitudinal study was to determine the associations between use of NSAIDs and changes in knee cartilage volume and knee cartilage defects over 2.9 years in older adults.
Methods: T(1)-weighted fat-suppressed magnetic resonance imaging on the right knee was performed in a total of 395 randomly selected subjects (mean age 62 years, range 51-80 years, and 50% female) to assess knee cartilage volume at tibial sites and knee cartilage defects (0-4 scale) at baseline and 2.9 years later. Medication use in the last month was recorded by questionnaire.
Results: Compared with nonusers of NSAIDs (n = 334), users of cyclooxygenase (COX)-2 inhibitors (n = 40) had decreased knee cartilage defect development in the medial tibiofemoral compartment (odds ratio [OR] 0.4, 95% confidence interval [CI], 0.2-0.99), whereas users of conventional NSAIDs (n = 21) had increased knee cartilage defect development in both medial (OR 3.1, 95% CI, 1.0-9.1) and lateral (OR 2.6, 95% CI, 1.0-6.7) tibiofemoral compartments. Comparing users of COX-2 inhibitors with users of conventional NSAIDs, the latter had higher knee cartilage volume loss (-5.3% vs -3.1% at medial tibia and -3.6% vs -1.1% at lateral tibia; all P <.05). All associations were adjusted for potential confounders including knee pain and radiographic osteoarthritis.
Conclusions: This study suggests that nonselective NSAIDs may have deleterious effects, while selective COX-2 inhibitors might have beneficial effects on knee cartilage. Randomized controlled trials examining knee structure to confirm this finding are warranted.