AUTHOR=Mi Jiarui , Liu Zhengye , Jiang Lingjuan , Li Meizi , Wu Xia , Zhao Nan , Wan Ziqi , Bai Xiaoyin , Feng Yunlu TITLE=Mendelian randomization in blood metabolites identifies triglycerides and fatty acids saturation level as associated traits linked to pancreatitis risk JOURNAL=Frontiers in Nutrition VOLUME=9 YEAR=2022 URL=https://www.frontiersin.org/journals/nutrition/articles/10.3389/fnut.2022.1021942 DOI=10.3389/fnut.2022.1021942 ISSN=2296-861X ABSTRACT=Background

There is very limited evidence on the causal effects of blood metabolites on pancreatitis risks. To reveal the causal associations between plasma metabolites and pancreatitis risks, we performed two-sample Mendelian randomization (MR) and Bayesian model averaging (MR-BMA) analyses in European ancestry.

Methods

The summary-level statistics from two genome-wide association studies with 249 and 123 metabolic traits derived from two separate cohorts involving ~115,000 (UK Biobank) and ~25,000 individuals from European ancestry were used for the analyses. The summary statistics of four pancreatitis datasets from FinnGen R5 and two pancreatitis datasets from UK Biobank were exploited as the outcome. We first performed univariable MR analysis with different metabolic GWAS data on multiple pancreatitis datasets to demonstrate the association pattern among different metabolites categories. Next, we exploited the MR-BMA method to pinpoint the dominating factors on the increased risk of pancreatitis.

Results

In the primary analysis with 249 traits, we found that plasma triglycerides were positively associated with pancreatitis risk. Intriguingly, a large number of traits associated with saturation or unsaturation of fatty acids also demonstrated causal associations. The replication study analyzing 123 metabolic traits suggested that bisallylic groups levels and omega-3 fatty acids were inversely correlated with pancreatitis risk. MR-BMA analyses indicated that the ratio of triglycerides to total lipid in various HDL particles played leading roles in pancreatitis susceptibility. In addition, the degree of unsaturation, the ratio of polyunsaturated fatty acids to monounsaturated fatty acids and the level of monounsaturated fatty acids showed causal associations with either decreased or increased pancreatitis susceptibility.

Conclusions

Our MR study provided an atlas of causal associations of genetically predicted blood metabolites on pancreatitis, and offered genetic insights showing intervention in triglycerides and the supplementation of unsaturated fatty acids are potential strategies in the primary prevention of pancreatitis.