The association of dietary patterns with knee symptoms and MRI detected structures in patients with knee osteoarthritis (OA)

Abstract

Background/Purpose:

This study aimed to examine the cross-sectional and longitudinal associations of dietary patterns with knee symptoms and structures in knee OA patients.

Methods:

Participants were from a subset of a randomised, placebo-controlled trial in Tasmania , who had symptomatic knee OA and vitamin D deficiency at baseline and received 50,000IU vitamin D₃ (N= 129) or placebo (N= 132) monthly for 24 months (aged 63.0 ± 7.2 years). Baseline diet was assessed by the Anti-Cancer Council of Victoria food frequency questionnaire. At baseline and 24 months, knee symptoms were assessed using WOMAC index and knee structures using MRI. Factor analysis was used to identify dietary patterns. Each participant received a score for each dietary pattern, with a higher score indicating a great intake of food composing that pattern. Associations between dietary pattern scores and knee OA outcomes were examined using multivariable linear regressions with adjustment for age, sex, BMI/change in BMI, treatment arm, total energy intake and physical activity.  

Results:

Three dietary patterns were identified: “western pattern”, characterised by high intakes of processed food, chips and wine, “vegetable and meat pattern”, characterised by high intakes of vegetables, red meat and beers, and “healthy pattern”, characterised by high intakes of vegetables, legumes, nuts, fish, fruits and wholegrain. Participants with higher healthy pattern or vegetable and meat pattern scores had lower baseline WOMAC function scores (β: -54.0, 95% CI: -99.9 to -8.1 and -58.7, -10.9 to -0.8, respectively), and also had lower baseline total WOMAC scores (β: -69.4, 95% CI: -131.4 to -7.4 and -76.5, -144.7 to -8.3, respectively). Baseline western pattern scores were not associated with either total WOMAC or WOMAC function scores. Healthy pattern or vegetable and meat pattern scores were not associated in WOMAC scores over 24 months, but  higher western pattern scores were associated with increased total WOMAC scores and WOMAC function scores (β: 118.4, 95% CI: 25.2 to 111.6 and 101.8, 36.2 to 167.4, respectively) over 24 months. In multivariable analyses, dietary patterns were not significantly  associated with any knee structure cross-sectionally or longitudinally.

Conclusion:

These results suggest that maintaining a healthy dietary pattern may have beneficial effects on knee function, whereas maintaining a western pattern may contribute to increased functional disability over time in knee OA patients. The evidence does not suggest that dietary patterns affect knee structure.

Figure 1. Rotated factor loading for the three dietary patterns.