Abstract
<jats:p>Objective: Observational studies have provided evidence for positive associations between inflammatory dietary patterns (IDP) and mental health, which might be mediated by immune activation. However, a causal relationship has not yet been established. Here we aim to investigate the causal nature of associations between IDP and mental health traits (depressed affect, mood swings, neuroticism, feed-up feelings, worry, irritability) using Mendelian Randomization (MR) analyses. Method: In the UK Biobank dataset, IDP was identified by conducting a partial least squares regression (PLSR) on the items of the food frequency questionnaire along with three inflammatory biomarkers as response variables: C-reactive protein, platelets, and white blood cell (WBC) counts. An individual-level genome-wide association study (GWAS) of IDP was performed within an unrelated European subsample from the UK Biobank (n=320,137) and summary-level GWAS data for mental health traits were utilized for bi-directional two-step MR to test the association between genetically predicted IDP and mental health traits. Result: The first PLSR component was retained for subsequent analysis, with a higher IDP score indicating a more frequent consumption of processed meat, beef, pork, lamb/mutton, and poultry. Genome-wide association analysis identified 101 independent genomic loci. MR analyses indicated a uni-directional positive relationship from IDP to neuroticism and a positive bi-directional relationships between IDP and depressed affect, mood swings, fed-up feelings, and irritability. The mediation effect of total white blood cell count on neuroticism score was also significant (adjusted P<0.05). Conclusion: Our findings identified genetic loci and functional properties of IDP and provided evidence for causal pathways with depressed affect, mood swings, irritability, and fed-up feelings. Implementing dietary advice and interventions should become part of a public mental health approach. Keywords: Inflammatory dietary pattern; Partial least squares regression; Genome-wide association study; Mendelian Randomization; Mental health.</jats:p>