Objective  To observe the changes in the composition of gut microbiota in stroke patients showing cognitive impairment within one month after the stroke, and to explore the correlation between bacteria presenting dissimilarity and cognitive functions and other clinical indicators.

  Methods  A cross-sectional study was conducted, involving 12 patients with post-stroke cognitive impairment (PSCI group), 12 stroke patients without cognitive impairment (Non-PSCI group), and 12 healthy volunteers in a normal control group (NC group). The demographic and clinical data were gathered. The abundance, diversity and dissimilarity of gut bacterial communities were determined by 16S rRNA gene sequencing. Then, we studied the correlation between gut microbiota and clinical characteristics and the effectiveness of using microbiome markers to identify cognitive decline.

  Results  The Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA) scores of the PSCI group were significantly lower than those the Non-PSCI group (P<0.001). There was no significant intergroup difference in the demographic data, the clinical data, and the Alpha diversity of gut microbiota among the three groups (P>0.05). Microbial composition analysis of the three groups revealed proportion alternations at the phylum, genus and species levels. At the phylum level, linear discriminant analysis (LDA) effect size (LEfSe) analysis suggested that the Actinomycetes had significantly increased relative abundance in the PSCI group (LDA score>2). At the genus and species levels, Firmicutes had the highest diversity among the top 10 bacteria in the three groups, while the relative abundance of Verrucomicrophyla presented an increasing trend in the Non-PSCI group and that of Actinobacteria showed an increasing trend in the PSCI group. Further LEfSe analysis revealed that there were different microbiome markers in each group, among which the Bifidobacterium, Alloscardovia, and Alloscardovia omnicolens of the phylum Actinomycetes and Lactobacillus gasseri and Anaerostipes hadrus of the phylum Firmicutes in the PSCI group increased significantly (LDA score>2). Correlation analysis indicated that Anaerostipes hadrus was negatively correlated with the MoCA scores, while Bifidobacterium was positively correlated with blood uric acid (UA). Bifidobacterium, Lactobacillus gasseri and Anaerostipes hadrus could be used to distinguish PSCI patients from Non-PSCI patients, presenting an area under the curve of 0.785, 0.792 and 0.750, respectively (P<0.05).

  Conclusion  Stroke patients with cognitive impairment in the early stage showed composition changes in their gut microbiota, and the bacteria exhibiting dissimilarity were correlated, to some degree, with cognitive function and related risk factors, which could provide new clues for the early management of PSCI.