Bidirectional Mendelian randomization explores the causal relationship between dietary habits and rheumatoid arthritis.

Abstract

Epidemiological and other studies have shown that the occurrence and progression of rheumatoid arthritis (RA) are closely related to diet. To further explore the causal association between dietary habits and RA, we performed a bidirectional Mendelian randomization (MR) analysis. The dataset related to dietary habits is from genome-wide association studies, including 143 dietary habits. The dataset of RA is from the FinnGen database. Inverse variance weighted (IVW), MR-Egger, simple mode, weighted median, and weighted mode were used for the 2-sample, 2-way MR analysis. At the same time, a variety of pleiotropic and heterogeneity tests were used to ensure the accuracy of the results. IVW results show that among current drinkers (drinks usually with meals yes + it varies vs no) was positively correlated with RA (β, 0.563 [95% confidence interval [CI], 0.286-0.840]; P = 6.7 × 10-5). Spread type (low fat spread vs any other) was negatively correlated with RA (β, -2.536 [95% CI, -3.725 to -1.346]; P = 2.9 × 10-5). In addition, the reverse MR results showed that RA was positively correlated with milk type (skimmed vs any other; β, 0.006 [95% CI, 0.000-0.011]; P = 4.5 × 10-2). RA was positively correlated with spread type (tub margarine vs never; β, 0.016 [95% CI, 0.002-0.029]; P = 2.5 × 10-2). The results of pleiotropy and heterogeneity tests showed that there was no pleiotropy (P > .05) in the obtained results. The analysis results of MR-Egger, simple mode, weighted median, and weighted mode are consistent with our IVW results. This study reveals a potential association between specific dietary habits and RA. Among current drinkers (drinks usually with meals yes + it varies vs no) was positively correlated with RA. Spread type (low fat spread vs any other) was negatively correlated with RA. RA was positively correlated with milk type (skimmed vs any other) and spread type (tub margarine vs never).