Abstract: Objective To explore the causal relationship between immune cell phenotypes and autism spectrum disorder. Methods The present study utilised a bidirectional two-sample mendelian randomization analysis. The present study utilised publicly available genetic data from genome-wide association studies(GWAS)in European populations. Single-nucleotide polymorphisms(SNPs)were utilised as instrumental variables(IVs). The causal effect analysis was primarily executed using the inverse variance weighted(IVW)method, with additional validation conducted via MR-Egger regression, weighted median, weighted mode, and simple mode. In order to control for the possibility of false positive results resulting from multiple comparisons, false discovery rate(FDR)correction was performed using the Benjamini-Hochberg method. The heterogeneity of the IVs was assessed by the Cochran Q-test, while the horizontal multiple validity was tested using the MR-Egger intercept analysis and the MR-PRESSO global test. Finally, sensitivity analyses were performed using the leave-one-out method. Results A total of 13,092 SNPs were included in the study as IVs. Following FDR correction(P_FDR>0.05), no statistically significant association was identified between immune cell phenotypes and ASD. Nevertheless, six low P-value phenotypes merit further discussion in the uncorrected analysis. Specifically, the percentage of CD8⁺T cells in leukocytes(OR=1.099,95%CI:1.039-1.163, P_IVW=0.001), the expression level of CD20 in lgD⁺CD38^-B cells(OR=1.064, 95%CI:1.019-1.110, P_IVW=0.005)and the expression level of CD45 in immature MDSCs(OR=1.056,95%CI: 1.021-1.093, P_IVW=0.001)may incease ASD risk. In addition, CD45 expression level in HLA DR⁺ T cells(OR=0.945, 95%CI:0.906-0.986, P_IVW=0.009), CD14 expression level in CD33⁺HLA DR⁺CD14 dim(OR=0.955, 95%CI:0.925-0.987, P_IVW=0.006)and CD25 expression level in CD4 regulatory T cells(OR=0.963, 95% CI:0.939-0.989, P_IVW=0.005)may have a protective effect against ASD. The analyses did not reveal any substantial horizontal pleiotropy or heterogeneity. Sensitivity analyses indicated robust results. Conversely, reverse MR analysis did not demonstrate a substantial impact of ASD on the aforementioned immunophenotype. Conclusion The present study suggests that specific immune cell phenotypes may have a potential causal relationship with the risk of ASD development, and provides a research direction for future exploration of immune-related biomarkers and potential therapeutic targets for ASD.

Key words: Autism spectrum disorder, Immune cell, Immune cell phenotype, Mendelian randomization