Abstract Background IBD is characterised by gut bacterial dysbiosis and secondary impaired conversion of primary to secondary bile acids (BA) by the gut microbiome, however, less is known about plasma BA profile1. We aimed to characterise associations between plasma BA and stool microbial diversity and composition in a cohort of patients with IBD Methods Patients with IBD in clinical remission and a group of healthy controls (HC) underwent collection of a fasting blood and stool sample. Plasma levels of 27 BA were analysed using liquid chromatography-mass spectrometry (LCMS). Targeted quantification of 3 BA and area under the curve (AUC) profiling of a further 24 BA was performed. Missing AUC metabolite values were imputed and data log 2 normalised as per accepted practice2. Stool samples underwent 16s rRNA gene amplicon sequencing. Unsupervised analysis of stool microbiota profile identified distinct clusters in microbiota profiles that reflected disease status. Supervised analysis of this clustering identified key taxa contributing to each community group. Univariate ANOVA analyses then determined if BA differed between this grouping and Pearson correlations investigated associations between the top 50 taxa separating this clustering. Results A total of 57 IBD (CD 31, UC 26) and 24 HC subjects participated. When supervised clustering was applied to BA levels by study group, differences did not reach statistical significance between IBD and HC groups (p 0.07). When participant gut microbiota profiles underwent unsupervised analysis we identified two distinct clusters which separated by study arm (Fig 1a). Assigned cluster groupings were then analysed for differences in BA. Seven BA differentiated between these IBD and HC predominant groups. (Fig 1b). The top 50 most abundant bacterial taxa contributing to the unsupervised clustering were then correlated with BA levels, with several significant findings (Fig 1c). On univariate analysis of the IBD group, 4 BA showed a negative correlation with stool microbial diversity parameters, while 5 BA, exhibited a positive correlation (Table 1). There were no associations between BA and IBD diagnosis (UC v CD), fecal calprotectin or risk of flare at follow up (median 18.5 months). Conclusion BA metabolomic profiling alone did not separate HC and IBD groups, however after cluster analysis by stool microbial composition was performed, clear groups formed, with one group consisting largely of IBD participants and the other of the HC. This emphasises the integral link between stool microbiota, plasma metabolomic profile and IBD and gives promise to the idea that a convenient and acceptable blood sample may provide useful insights into stool microbial diversity and composition. References: 1.Prast-Nielsen S, Granstrom A, Kiasat A, al e. Associations of the intestinal microbiota with plasma bile acids and inflammation markers in Crohn’s disease and ulcerative colitis. Scientific Reports. 2025;1535039. 2.Wei R, Jingye W, Su M, al e. Missing Value Imputation Approach for Mass Spectrometry-based Metabolomics Data. Scientific Reports. 2018;8663. Conflict of interest: Dr. Wark, Gabrielle: No conflict of interest Tillet, Bree: No conflict of interest O Kaakoush, Nadeem: No conflict of interest Samocha-Bonet, Dorit: No conflict of interest Ghaly, Simon: No conflict of interest Danta, Mark: No conflict of interest
Wark et al. (Thu,) studied this question.