A total of 10 patients were selected; 8 with CI and 2 with IS. They were recruited from Tianjin Huanhu Hospital (Tianjin, China) between May 2015 and October 2015, and had been diagnosed by skull computed tomography examination. The group comprised6 males and 4 females. Additionally, 10 healthy volunteers, 5 male and 5 female, were selected as the normal group. The age of the individuals was from 53–82 years with a mean age of 64.3. Research subjects did not receive antibiotics in the week prior to the specimen collection. As required by the medical ethics committee, those in the normal group were examined to ensure that they had no metabolic, cardiovascular or cerebrovascular diseases or cancer. All participants signed informed consent prior to the experiment. The study was approved by the Ethics Committee of Tianjin Huanhu Hospital.

A 300 mg middle segment feces sample was collected in the morning and the genomic DNA was extracted using the Stool DNA kit (Tiangen Biotech Co., Ltd., Beijing, China). Bacteria cells were split at 95°C for 10 min and 4 µl RNase A was added and incubated at 70°C for 30 min. DNA concentrations were determined using a microplate reader (Multiskan FC, Thermo Fisher Scientific, Inc., Waltham, MA, USA) and stored at −80°C.

The bacterial genomic DNA was amplified with primers specific for the V4 hypervariable regions of the 16SrDNA gene. The sequences were forward, 5′-GTGCCAGCMGCCGCGGTAA-3′ and reverse, 5′-GGACTACHVGGGTWTCTAAT-3′, and the reaction was performed using Phusion High-Fidelity PCR master mix with GC buffer (New England Biolabs, Inc., Ipswich, MA, USA). Samples were sequenced by Illumina MiSeq platform provided by Illumina, Inc. (San Diego, CA, USA) with the following conditions: Denaturation 1 min at 98°C, 10 sec at 98°C, 30 sec at 50°C and 30 sec at 72°Cfor 30 cycles, and 72°C at 5 min.

A total of 5 ml venous blood was collected from the elbow of the subjects, early in the morning and before they had eaten, and placed in a coagulation-promoting tube then centrifuged at 300 × g for 10 min at room temperature. The serum was collected and the blood glucose and lipid concentrations ascertained using a Beckman AU5800 biochemical detector (Beckman Coulter, Inc., Brea, CA, USA).

Sequencing results analysis was performed by the QIIME software package 1.7.0 (qiime.org) (9) and UPARSE pipeline (10). Weighted/unweighted UniFrac analysis was performed by QIIME. Cluster samples analysis was performed by principal coordinate analysis (PCoA) and unweighted pair group method with arithmetic mean (UPGMA). Operational taxonomic units (OTUs) analysis was performed by partial least square discriminate analysis via Simca-p+11.5, values higher than 1.5 were considered as key OTUs. The heatmap was constructed with HemI version 1.0 (The CUCKOO Workgroup, Wuhan, Hubei, China). Data analysis was performed using SPSS 11.0 software (SPSS, Inc., Chicago, IL, USA) and one-way analysis of variance with Tukey's test was applied for the intergroup analysis. Values are presented as the mean ± standard deviation. P<0.05 was considered to indicate a statistically significant difference.

The sequencing results demonstrated that there were 274 (48key) OTUs that were significantly associated with CI and IS (Fig. 1A). A significant difference in the abundance of Firmicutes, Proteobacteria, Bacteroidetes, Actinobacteria, Cyanobacteria, Verrucomicrobia, Planctomycetes, Tenericutes, Euryarchaeota and Spirochaetesat the phylum level in the CI group compared with the normal group was observed using UPGMA analysis (P<0.05), however, no significant difference was observed in the abundance of the above phyla in the IS group compared with the normal group (Fig. 1B).

The heatmap demonstrates that there was a change in abundance at the genus level among the groups (Fig. 2A). Particularly, the results demonstrated that Escherichia, Megamonas, Dialister, Bifidobacterium and Ruminococcus were more abundant in the IS and CI groups. The abundance of Bacteroides, Parabacteroides, Akkermansia, Prevotella and Faecalibacterium was decreased in the CI group compared with the normal group (P<0.05). There was a greater abundance of Escherichia, Dialister and Bifidobacterium in the IS group compared with the normal group (P<0.05), although the abundance of Bacteroides, Megamonas, Parabacteroides, Akkermansia, Prevotella, Faecalibacterium and Ruminococcus was less. There was a greater abundance of Escherichia, Bacteroides, Megamonas, Prevotella and Ruminococcusin the CI compared with the IS group; although the abundance of Parabacteroides, Akkermansia, Faecalibacterium, Dialister and Bifidobacterium was less (P<0.05; Fig. 2B and C).

It can be seen from the principal component analysis (PCA) that the diversities were assessed by a net separation of the three groups produced by the intestinal gut between the healthy group and the CI and IS patients (P<0.05), which indicated less similarity in the microbial structure and composition of the feces between CI and IS patients and the healthy group (Fig. 3A). Notably, the same pattern was observed in the PCoA analysis and that there are also samples that were grouped into two distinct clusters in the intestinal microbial communities of CI patients (P<0.05); this may be associated with the process of CI and requires further investigation (Fig. 3B).

The present study demonstrated that the blood glucose of patients with CI and IS was significantly higher compared with the healthy group; however, no significant changes were observed in triglyceride, total cholesterol, high-density lipoprotein and low-density lipoprotein, apolipoprotein A, apolipoprotein B and uric acid (P<0.05; Fig. 4). Therefore, it is hypothesized that variations in the intestinal microbial communities in CI and IS patients may lead to a disorder of the blood glucose metabolism and thus a potential dangerous factor causing CI.