OBJECTIVE: The causality between CMD and sepsis has remained largely unknown. To elucidate this, we conducted a Mendelian randomization (MR) and population study. METHODS: Firstly, we used univariable and multivariable MR analyses to investigate causal associations between CMD and sepsis-related outcomes. We obtained Genome-wide association study summary from both the MRC Integrative Epidemiology Unit and the FinnGen consortium. Subsequently, a two-step mediation MR analysis was performed to explore mediators. Afterward, we conducted an observational study using the MIMIC-IV database, in which multivariable logistic regression models were utilized to examine the relationship between CMD and sepsis-related outcomes. RESULTS: In the MR study, type 2 diabetes mellitus (T2DM) (OR = 1.058, 95%CI = 1.017-1.100, p = 0.005), obesity (OR = 1.113, 95%CI = 1.057-1.172, p <
0.001) and heart failure (HF) (OR = 1.178, 95% CI = 1.063-1.305, p = 0.002) were independently causally related to sepsis. Obesity (OR = 1.215, 95% CI = 1.027-1.437, p = 0.023) and HF (OR = 1.494, 95% CI = 1.080-2.065, p = 0.015) also showed independent causal associations with sepsis critical care admission. Mediation MR analysis identified 23 blood metabolites potentially causally linked to sepsis (p <
0.05), yet none mediated the relationship between CMD and sepsis. In the observational study, we found associations between sepsis and several conditions including T2DM, obesity, hypertension, stroke, HF, and hyperlipidemia after adjusting for confounding factors. Moreover, hypertension, stroke, HF, coronary artery disease, and hyperlipidemia were linked to sepsis critical care admission. CONCLUSION: This study has, for the first time, revealed indicative evidence of a causal relationship between CMD and sepsis through observational and genetic evidence. Taken together, clinical attention to sepsis may be warranted among patients with CMD.