Chenyang Fan ^1,2 Pengying Yuan ^3* Xiangdong Yang ^1,2* Weifeng Zhang ^1,2* Xingli Wang ^1,2* Juan Xie ^1,2* Jing He ^1,2* Haijing Chen ^1,2* Lixiang Yan ^1,2* Zhexin Shi ^1,2*

• ¹ First Teaching Hospital of Tianjin University of Traditional Chinese Medicine, Nankai District, China
• ² National Clinical Research Center for Chinese Medicine Acupuncture and Moxibustion, tianjing, China
• ³ University Hospital of International Business and Economics, Beijing, China

Recent studies have confirmed that metabolites and immunocyte phenotype may be associated with the risk of lymphoma. However, the bidirectional causality between metabolites, immunocyte phenotype, disease risk, and whether immunity is an intermediate mediator between metabolism and lymphoma causality is still unclear.To elucidate the causal relationship between metabolites, immune cell phenotypes, and lymphomas, we used two-sample Mendelian randomization (MR) and two-step MR analysis.Applying large-scale genome-wide association studies (GWAS) pooled data, we selected 1400 metabolites and 731 immunocyte phenotypes with eight lymphoma subtypes for two-sample bi-directional MR analysis. In addition, we used two-step MR to quantify the proportion of metabolite effects on lymphomas mediated by immunocyte phenotype.This study yielded a bidirectional causal relationship between 17 metabolites and lymphoma and a bidirectional causal relationship between 12 immunocyte phenotypes and lymphoma.In addition, we found causal associations between metabolites and lymphomas, three groups of which were mediated by immunocyte phenotypes. Among them, CD27 on plasmablast/plasma cell (PB/PC) was a mediator of the positive association of arginine to glutamate ratio with chronic lymphocytic leukemia, with a mediator ratio of 14.60% (95% 3 CI=1.29-28.00%, P=3.17 × 10-2). Natural killer (NK) cells as a percentage of all lymphocytes(NK %lymphocyte) was a mediator of the negative association of X-18922 ( unknown metabolite ) levels with diffuse large B-cell lymphoma, with a mediation proportion of -8.940% (95% CI=-0.063-(-17.800) %, P=4.84 × 10-2). CD25 on IgD-CD24-B cell was the mediator of the positive association between X-24531 ( unknown metabolite ) levels and diffuse large B-cell lymphoma, with a mediation proportion of 13.200% (95% CI=-0.156-26.200%, P=4.73 × 10-2).In the present study, we identified a causal relationship between metabolites and lymphoma, in which immunocyte phenotypes as mediators are involved in only a minor part. The mediators by which most metabolites affect the risk of lymphoma development remain unclear and require further exploration in the future.