Background/Purpose: Few studies have examined the effect of diet on future risk of knee OA development. The association between overall dietary patterns and risk of OA is unknown. We aimed to examine the prospective association of major dietary patterns by principal component analysis (PCA) with risk of developing radiographic knee OA.

Methods:  In the Osteoarthritis Initiative (OAI), 2835 participants (4570 knees) without knee OA in at least one knee [Kellgren and Lawrence (KL) grade of 0 or 1] who had dietary data at baseline were followed up to 48 months. We defined knee OA incidence as KL grade ≥ 2. Dietary intake was assessed with a Block Brief Food Frequency Questionnaire completed at baseline. We identified two major underlying patterns of food consumption by PCA: “Western” pattern, characterized by high intakes of red / processed meats, refined grains, french fries, desserts, and sweets, and the “Prudent” pattern characterized by high intakes of fruit/ vegetables, legumes, poultry, and fish. Each subject receives a score for each dietary pattern, with a higher score indicating a higher adherence to the respective pattern. The derived patterns are statistically uncorrelated with each other. We grouped the dietary pattern scores using quartiles. We developed a Cox proportional hazards model (with discrete time) to assess the association between dietary patterns and incident knee OA. The robust covariance estimates were used to account for intra-subject correction.  

Results:   Among our sample, 414 knees developed knee OA within 48 months. In multivariable models after controlling for age, sex, race, injury/surgery, and other covariates, adherence to the Western pattern was associated with an increased risk of knee OA (HR_{Q4 vs Q1}: 1.60, 95% CI: 1.03 to 2.51, p trend: 0.03), while adherence to the Prudent pattern tended to be associated with a reduced risk of knee OA (HR_{Q4 vs Q1}: 0.72, 95% CI: 0.52 to 1.00, p trend: 0.11) (Table). The observed associations were attenuated after additionally adjusting for body mass index (BMI).

Conclusion: Following a Western diet may increase the future risk of knee OA, whereas following a Prudent pattern may be associated with a reduced risk of knee OA. The associations may be partially mediated through BMI. Replication of these novel findings in other prospective studies demonstrating that improvement of  dietary quality leads to delay in knee OA development are needed.

Key words: Western dietary pattern, Prudent dietary pattern, osteoarthritis.