Causal relationship between gut microbes and cardiovascular protein expression

Abstract

Evidence supports associations between gut microbiota and cardiovascular protein levels in plasma. However, it is unclear whether these associations reflect a causal relationship. To reveal the causal relationship between gut microbiota and cardiovascular protein levels in plasma, we estimated their causal effects using two-sample Mendelian randomization (MR) analysis. Sensitivity analysis was also performed to assess the robustness of our results. Genome-wide association study (GWAS) of microbiomes in the MiBioGen study included 211 bacterial taxa (18,473 individuals), and GWAS of 90 cardiovascular proteins included 30,931 individuals. There were 196 bacterial taxa from five levels available for analysis. The following 14 causal relationships were identified: phylum Euryarchaeota and carbohydrate antigen 125 (β = 0.289), order Bacillales and CSF-1 (β = -0.211), genus Dorea and HSP-27 (β = 0.465), phylum Actinobacteria and IL-8 (β = 0.274), order Enterobacteriales and KIM-1 (β = -0.499), class Actinobacteria, genus Bifidobacterium, phylum Actinobacteria and LEP (β = -0.219, β = -0.201, and β = -0.221), genus Methanobrevibacter and NT-proBNP (β = 0.371), family Peptostreptococcaceae and SRC (β = 0.191), order Verrucomicrobiales, phylum Verrucomicrobia and TNF-R2 (β = 0.251 and β = 0.233), family Veillonellaceae and t-PA (β = 0.271), and class Erysipelotrichia and VEGF-D (β = 0.390). Sensitivity analysis showed no evidence of pleiotropy or heterogeneity. The results of the reverse MR analysis showed no reverse causality for any of the 13 gut microbes and 11 cardiovascular proteins. Mendelian randomization estimates provide strong evidence for a causal effect of gut microbiota-mediated alterations on cardiovascular protein expression.