Association Between Gut Microbiota and Chronic Kidney Disease: A Two-Sample Mendelian Randomization Study in a Chinese Population

Abstract

Background: Population differences in gut microbiota composition and related metabolites may influence their potential causal relationship with chronic kidney disease (CKD); however, this relationship remains poorly understood in the Chinese population. Materials and Methods: We conducted a two-sample Mendelian randomization (MR) study using summary statistics of 500 gut microbial features (9 phyla, 3 classes, 14 orders, 32 families, 95 genera, 248 species, and 99 gut metabolic modules (GMMs)) from the 4D-SZ (from Shenzhen, China) discovery cohort (n = 1539). CKD summary statistics were obtained from the China Kadoorie Biobank (CKB) (489 cases and 75,531 controls). Associations between gut microbiota and CKD were evaluated via inverse variance weighted, MR-Egger, weighted median, and MR-PRESSO. To validate our findings, we replicated the analyses in two independent East Asian CKD GWAS datasets: the Biobank of Japan (BBJ) dataset (2117 cases and 174,345 controls) and the J-Kidney-Biobank (JKB) dataset (382 cases and 3471 controls). We further validated the results via a meta-GWAS of BUN and eGFR in Biobank Japan (BBJ) and the Taiwan Biobank (TWB). Additionally, we analyzed 304 serum proteins from the Guangzhou Nutrition and Health Study (GNHS) and conducted mediation MR analyses to explore potential mediators. Result: At the locus-wide significance threshold, we identified 18 gut microbiome features associated with CKD onset in the China Kadoorie Biobank (CKB). Genus Alistipes (OR 1.02, 95% CI 1.00-1.03, p = 0.03) was associated with incident CKD risk in the JKB cohort. Species Bifidobacterium catenulatum-Bifidobacterium pseudocatenulatum complex (OR 1.0074, 95% CI 1.0070-1.0142, p = 0.01) was associated with incident CKD risk in a meta-GWAS of BUN. Sensitivity analyses, including Cochran's Q test, MR-Egger intercept analysis, leave-one-out analysis, and funnel plots, yielded consistent results. Mediation analysis revealed that 26.7% (95% CI: 0.006-0.6700, p = 0.04) of the effect of Alistipes on CKD risk was mediated through the serum protein FBLN1. Conclusions: Our study provides Mendelian randomization-based evidence supporting a potential causal relationship between gut microbiota and CKD, highlighting the potential mediating role of FBLN1 in the association between genus Alistipes and CKD. Further studies are needed to explore whether and how genus Alistipes and FBLN1 contribute to CKD development.

Keywords: Mendelian randomization analysis; chronic kidney disease; gut microbiota; mediation analysis; serum proteins.

Figure 1

Study design. This flowchart illustrates the framework of our investigation, which is structured into three main phases. Phase 1 involved conducting a two-sample Mendelian randomization (MR) analysis utilizing the inverse variance weighted method along with multiple sensitivity analyses to identify potential causal gut microbial taxa linked to chronic kidney disease in the Chinese population. Phase 2 focused on validating the significant gut microbial taxa identified in the first phase by using two summary-level datasets related to chronic kidney disease and chronic renal failure, as well as meta-GWAS data for BUN and eGFR sourced from BBJ and TWB. Phase 3 comprised a mediation MR analysis, where we assessed the causal associations between gut microbial taxa and various serum proteins. Additionally, we determined the extent to which these microbial taxa influence atrial fibrillation through mediation by these risk factors. MR-PRESSO: MR Pleiotropy RESidual Sum and Outlier. MR-Egger: Mendelian randomization Egger regression. BBJ: BioBank Japan. JMOP: Japanese Multi-Omics Reference Panel. eGFR: estimated glomerular filtration rate. BUN: blood urea nitrogen. GNHS: Guangzhou Nutrition and Health Study.

Figure 2

Preliminary MR estimates for gut microbiota and CKD risk. This figure presents the initial Mendelian randomization (MR) estimates assessing the relationship between gut microbiota and the risk of chronic kidney disease (CKD). The concentric circles represent estimates derived from different analytical methods in the following order, from the innermost to the outermost layer: PRESSO, MR-Egger, weighted median, and inverse variance weighted methods. The color intensity reflects statistical significance, with darker shades indicating smaller p-values.We identified 18 microbial features significantly associated with CKD, including one family, five genera, nine species, and three gut metabolic modules (GMMs) (Figure 3A). Specifically, the family Moraxellaceae (OR 0.87, 95% CI 0.76–1.00, p = 0.05), genus Alistipes (OR 1.20, 95% CI 1.03–1.40, p = 0.02), genus Dickeya (OR 1.22, 95% CI 1.02–1.46, p = 0.03), genus Dorea (OR 1.32, 95% CI 1.01–1.72, p = 0.04), genus Klebsiella (OR 1.16, 95% CI 1.01–1.33, p = 0.04), and genus Parabacteroides (OR 1.37, 95% CI 1.07–1.75, p = 0.01) were found to be significantly associated with CKD.

Figure 3

(A) Mendelian randomization analysis of significant gut microbial taxa and chronic kidney disease in the Chinese population. SNP: single nucleotide polymorphism. N. SNPs: Number of SNPs used for the estimation of the causal effects. (B) Replication of gut microbial taxa and chronic kidney disease using GWAS data from J-Kidney-Biobank. (C) Replication of gut microbial taxa and chronic kidney disease using blood urea nitrogen (BUN) GWAS data from meta data of TW and BBJ. Odds ratios (ORs), 95% confidence interval (95% CI), and p-values were calculated using the inverse variance weighted method. The prefixes “g.”, “s.”, “f.”, and “MF” in the taxa column represent genus, species, family, and the metabolic feature (gut microbiome modules, GMMs), respectively. The size of blue dot represents the stand error of the Odds ratios.

Figure 4

Mediation MR analysis of the causal effect of gut microbiota on chronic kidney disease via serum proteins. (A) Mendelian randomization (MR) analysis of serum protein on chronic kidney disease; (B) MR analysis of significant gut microbiota taxa on significant CKD-related serum proteins; (C) estimates of the effect of gut microbiota on chronic kidney disease explained by risk factors. β1, the effect of the serum protein on chronic kidney disease. β2, the effect of the gut microbial taxon on the serum protein. p-values were calculated from the inverse variance weighted method. The prefixes “g.”, “s.”, and “MF” in the taxa column represent genus, species, and the metabolic feature (gut microbiome modules, GMMs), respectively.

Figure 5

Summary of the study. We conducted a two-sample Mendelian randomization (MR) analysis using summary statistics from genome-wide association studies (GWAS) of 500 gut microbial traits (4D-SZ) and large-scale GWAS of chronic kidney disease (CKD). We identified two microbial taxa causally associated with CKD: the Bifidobacterium catenulatum–Bifidobacterium pseudocatenulatum complex and the genus Alistipes. Mediation MR analyses were performed for plasma protein levels. Our findings suggest that the protein FBLN1 may mediate the effect of Alistipes on CKD. OR: odds ratio. The prefixes “g.” and “s.” refer to genus and species, respectively.