A large-scale genetic correlation scan between rheumatoid arthritis and human blood metabolites.

BACKGROUND

Rheumatoid arthritis (RA) is a complex disease with several risk factors. The effects of blood metabolites on RA remains elusive. We conducted a genetic correlation scan to explore the relevance of blood metabolism with RA.

METHOD

The genome-wide association study (GWAS) dataset of RA(2014) was obtained from a large scale meta-analysis, including 29,880 RA cases and 73,758 controls. The GWAS datasets of 529 blood metabolites were derived from a recently published study. Linkage disequilibrium score regression (LDSC) analysis was performed to evaluate the genetic correlation between each of the blood metabolite and RA(2014). Then we used another GWAS data of RA(2021) and blood metabolites for LDSC analysis to verify whether the same blood metabolites were genetically correlated with RA. Mendelian randomization (MR) analysis was then applied to assess the causal relationship between the significant blood metabolites identified by LDSC and RA(2014).

RESULT

Six suggestive blood metabolites were identified for RA(2014), including 10-Undecenoate (correlation coefficient = -0.1686, p value = 0.0394), isovalerylcarnitine (correlation coefficient = 0.1660 p value = 0.0273), proline (correlation coefficient = 0.1647, p value = 0.0145), pantothenate (correlation coefficient = -0.3311, p value = 0.0078), tyrosine (correlation coefficient = 0.1735, p value = 0.0010), X-14057 (correlation coefficient = 0.2695, p value = 0.0373). We identified four blood metabolites may have genetic correlations with RA(2021), including oleoylcarnitine (correlation coefficient = 0.1927, p value = 0.0432), levulinate (correlation coefficient = 0.1008, p value = 0.0413), pantothenate (correlation coefficient = -0.2311, p value = 0.0180), tyrosine (correlation coefficient = 0.1301, p value = 0.0078). There are two identical blood metabolites were found to be related with RA: pantothenate and tyrosine. It was found that there was a significant positive causal relationship between RA (exposure) and 10-Undecenoate (outcome) (β = 0.0077, SE = 0.0033, p = 0.0192) by MR analysis.

CONCLUSION

We investigated the genetic correlation and causal relationship between RA and blood metabolites by LDSC and MR analysis. These results may provide novel insights into the genetic mechanism of RA.