Abstract: Objective East Asia exhibits one of the highest global incidence rates of gastriccancer. The intricate relationship between peripheral blood immune cell counts and the onset and progression of gastric cancer is often obscured by reverse causality and confounders. This study employs Mendelian randomization analysis to mitigate associated biases, aiming to elucidate the causalconnection between peripheral immune cell counts and gastric cancer. Methods Both univariateand multivariate Mendelian randomization methods were employed to explore the causal relationship between diverse immune cell subtypes in peripheral blood and the risk of gastric cancer using data from Genome-wide association studies ( GWAS). Exposure factors included counts of white blood cells, granulocytes, neutrophils, eosinophils, basophils, lymphocytes, and monocytes. The primary approach for univariate Mendelian randomization analysis involved the inverse variance-weighted method. Afterwards, a two-step mediation analysis was conducted to investigate the potential roleof plasma proteins in this process. Results Mendelian randomization analysis revealed a causal relationship between lymphocytes ( OR = 1. 094, 95 % CI: 1. 009-1. 185, P = 0. 029) and eosinophils (OR = 0. 881, 95 % CI: 0. 813-0. 955, P = 0. 002 ) and the incidence of gastric cancer. Sensitivity analysis corroborated the reliability of these results. Furthermore, the association between decreased eosinophil counts and a reduced risk of gastric cancer persisted in multivariate Mendelian randomization analysis ( OR = 0. 807, 95 % CI: 0. 671-0. 970, P = 0. 023). Reverse causality was not significant. The secreted protein acidic and rich in cysteine (SPARC) is an extracellular matrix protein that involves in various cell-matrix interactions and cellular signaling pathways, playing a significant role in tumor development. The mediation analysis of 1 124 plasma proteins suggested that plasma SPARC might mediate the protective effect of eosinophils against gastric cancer(Beta = - 0. 030, 95 % CI: - 0. 072-0. 000, P = 0. 024). Conclusion Mendelian randomizationanalysis establishes eosinophil counts as an independent factor in diminishing the incidence of gastric cancer. Plasma SPARC acts as a mediator in this process, suggesting that circulating eosinophil count and plasma SPARC have the potential to become early clinical indicators and therapeutic targets for gastric cancer. The Mendelian randomization method effectively mitigates biases stemming from reverse causality and psychosocial factors. Nevertheless, additional research is imperative to validate its underlying biological mechanisms.

Key words: