Introduction
Irritable bowel syndrome (IBS) is defined by the Rome IV diagnostic criteria as a group of disorders characterized by abdominal pain associated with altered bowel habits[1]. IBS is classified into four subtypes according to stool characteristics: constipation-predominant (IBS-C), diarrhea-predominant (IBS-D), mixed type, and unclassified type[1-3]. IBS-D is characterized by frequent diarrhea and is associated with impaired quality of life[3].
The global prevalence of IBS in the general population is approximately 10%, although it is lower when using the latest Rome IV criteria, estimated at approximately 6%[1-3].
The pathogenesis of IBS is associated with increased mucosal permeability[4]. IBS has also been linked to alterations in the gut microbiota, as IBS develops in approximately 10% of patients after infectious enteritis[4]. In addition, mucosal microinflammation is considered to contribute to IBS pathogenesis[3,4].
Female sex, younger age, psychological stress, and the severity of acute enteritis have been reported as risk factors for IBS[4,5]. A systematic review of gut microbiota studies suggested that IBS is associated with gut microbial dysbiosis characterized by increased Enterobacteriaceae and Bacteroides and decreased Faecalibacterium prausnitzii and Bifidobacterium, although a consistent microbial signature has not yet been established[6].
Commonly used pharmacological treatments for IBS-D include 5-HT3 receptor antagonists, probiotics, and bulking polymers or dietary fibers[3,4,7]. However, currently available therapies often show limited clinical efficacy. Hangeshashinto (HST) is a Kampo medicine composed of extracts from seven medicinal herbs and has been approved by Japan's Ministry of Health, Labour and Welfare for the treatment of various gastrointestinal symptoms, including diarrhea, heartburn, gastroenteritis, and stomatitis[8-10].
Experimental studies have demonstrated that HST shows suppression of inducible prostaglandin E2 production and exerts antibacterial effects, contributing to the alleviation of oral ulcerative mucositis-associated pain[11-13]. These reports suggest the efficacy of HST for IBS-D; however, in the Japanese clinical guidelines for IBS, Kampo medicines, including HST, are only weakly recommended because of the limited number of studies and the small sample sizes of the reported clinical cases[3].
We previously conducted a single-arm observational study titled “Microbiota Improvement by Hangeshashinto in IBS-D: the MIHAN-R study,” in which we evaluated symptom improvement in IBS-D patients treated with HST and analyzed changes in gut microbiota before and after treatment[14]. Among 100 patients, the overall clinical improvement rate was 82.0%. Microbiota analysis revealed that Bacteroides and Ruminococcus significantly decreased, whereas Megasphaera and Subdoligranulum increased after HST treatment.
The present study aimed to further investigate the gut microbiota profiles in IBS-D and their relationship with clinical efficacy of HST: the GMIBS-HA study.
Methods
This was a retrospective post-hoc analysis of a previous study (MIHAN-R study)[14]. In the study, 100 patients diagnosed as IBS-D according to the Rome IV criteria treated with HST at five Japanese affiliated institutions were analyzed for clinical improvement and gut microbiota[1]. Among 100 patients, only 20 patients consented to the analysis of gut microbiota.
In the current study, these 20 IBS-D patients treated with HST at two affiliated institutions between March 2023 and December 2023 among 100 patients in our previous study were retrospectively reviewed[1,14]. Eligible patients were analyzed for symptom improvement of IBS-D and underwent gut microbiota analysis before and after HST treatment (Figure 1). HST (Tsumura, Tokyo, Japan) was administered at a dose of 2.5 g per package three times daily before meals (total 7.5 g/day) for 2–3 weeks.
HST is approved under the Japanese health insurance system for diarrhea, acute and chronic gastroenteritis, and oral mucositis, and patients with IBS-D are also treated with HST in daily clinical practice among our affiliated institutions. The inclusion criteria of the current study were as follows: (1) male or female patients aged ≥20 years and (2) patients newly diagnosed with IBS-D without prior treatment or patients previously diagnosed with IBS-D with poor symptom control. In principle, newly diagnosed patients with IBS-D received HST monotherapy.
For patients previously diagnosed with IBS-D, switching to HST or adding HST to existing IBS-D medications was permitted. Patients with inflammatory bowel disease, severe systemic disease, or those taking other herbal medicines were excluded.
To compare gut microbiota profiles with those of healthy individuals from the same country, each IBS-D case was matched to one healthy control who had neither undergone surgical procedures nor taken medications and had no symptoms suggestive of IBS-D. Matching was performed using a commercial database (Cykinso Inc., Tokyo, Japan) according to residential country, sex, and age, based on a previous study[15]. As a result, 20 healthy controls were selected.
The primary outcome was the identification of gut microbiota associated with IBS-D (Figure 1). The secondary outcomes were as follows: 1. Differences in pre-HST gut microbiota between effective and non-effective cases, 2.
Changes in post HST gut microbiota between effective and non-effective cases and 3. Symptom improvement according to bacterial changes.
The clinical improvements for three symptoms of IBS-D such as stool frequency, stool consistency, and abdominal pain were defined according to our previous study[14]. Regarding stool frequency, the improvement was defined as a decrease in stool frequency of at least once a day. Stool consistency was scored using the Bristol stool shape score (1–7) before and after administration, where a change score to 3–5 was defined as improvement[16].
Abdominal pain was evaluated as improvement and non-improvement based on the patient's complaints. The overall improvement rate was defined as improvement in stool frequencies, stool consistency, or abdominal pain. In the related factors analysis of efficacy, patients were divided into two groups: effective and non-effective due to overall improvement.
Adverse events were also evaluated.
Gut microbiota profiles before and after HST treatment were analyzed in 20 patients who had consented to microbiota analysis in the previous study and in the current post-hoc analysis[14]. Alpha diversity between healthy controls and IBS-D patients before HST administration was evaluated using observed features, Shannon index, and Faith's phylogenetic diversity. Beta diversity was assessed using unweighted and weighted UniFrac distances.
Taxonomic composition was analyzed at both the phylum and genus levels for healthy controls and IBS-D patients. Changes in gut microbiota before and after HST treatment in effective and non-effective IBS-D cases were previously analyzed in the MIHAN-R study[14]. In the present study, we further analyzed bacteria that showed significant differences before and after HST treatment, as well as those that differed significantly between healthy controls and IBS-D patients.
Additionally, the abundance changes of these specific bacteria were categorized as ≤-30% decrease, -30 ～ -5% decrease, -5%< ～ <5% change (defined as no change), 5 ～ 30% increase, and 30%≤ increase.
Gut microbiota analysis
The analysis of gut microbiota was performed according to a previously reported method with the assistance of an analysis company (Cykinso Inc., Tokyo, Japan)[17]. Fecal samples were collected either at the hospital or at home using a fecal collection kit containing guanidine thiocyanate solution. Samples were transported to the company at ambient temperature and subsequently stored at 4 °C.
Previous studies have demonstrated that storing fecal samples in test tubes containing guanidine thiocyanate solution at room temperature preserves microbiota composition comparably to immediate deep freezing[18].
Deoxyribonucleic acid (DNA) was extracted from fecal samples using an automated DNA extraction system (GENE PREP STAR PI-480, Kurabo Industries Ltd., Osaka, Japan). For sequencing, amplicons targeting the V1–V2 region of the bacterial 16S rRNA gene were prepared using the forward primer (16S_27Fmod: TCG TCG GCA GCG TCA GAT GTG TAT AAG AGA CAG AGR GTT TGA TYM TGG CTC AG) and reverse primer (16S_338R: GTC TCG TGG GCT CGG AGA TGT GTA TAA GAG ACA GTG CTG CCT CCC GTA GGA GT). Although the V3–V4 region is commonly used for 16S rRNA sequencing, several studies have reported that the V1–V2 primer set provides better characterization of the Japanese gut microbiota[17,19,20].
The libraries were sequenced using a MiSeq Reagent Kit v2 (Illumina, San Diego, CA, USA) with a 250-bp paired-end protocol (500 cycles). Data processing and taxonomic assignment were conducted using the QIIME2 pipeline (version 2021.2) as follows: (1) paired-end reads were joined, filtered, and denoised using a divisive amplicon denoising algorithm, and (2) taxonomic classification of amplicon sequence variants (ASVs) was performed using a naïve Bayes classifier implemented in QIIME2[21].
For diversity analysis, alpha diversity was evaluated at the ASV level using observed features, the Shannon index, and Faith's phylogenetic diversity (PD) in comparisons between pre-HST and post-HST samples. Observed features represent a quantitative measure of taxonomic richness. The Shannon index reflects both richness and evenness of the microbial community, whereas Faith's PD is a phylogenetic measure of taxon richness expressed as the total branch length of the phylogenetic tree observed in a sample.
Alpha diversity metrics were calculated using the QIIME2 “diversity” plugin.
Beta diversity was used to evaluate differences in microbial community composition between samples using the Bray–Curtis distance metric. Pairwise distances between samples were calculated using a Bray–Curtis distance matrix and visualized using principal coordinate analysis (PCoA), where each point represents the overall microbiota composition of an individual sample.
In microbiota composition analysis, differential abundance analysis is challenging due to the compositional nature of sequencing data, which depends on sequencing depth. The number of reads assigned to each ASV must therefore be interpreted relative to the total reads within each sample. Bias-corrected compositional analysis was applied to address this issue by estimating and removing biases introduced by differences in sampling fractions.
Ethical considerations
This study was approved by the Ethics Committee of the representative center (Approval No. ERB-C-2284, March 17, 2022). The study was conducted in accordance with the Declaration of Helsinki of the World Medical Association.
Because this was a retrospective analysis of clinical data obtained from a prospective observational study, the evaluation of clinical efficacy was approved using an opt-out consent process. Analysis of the intestinal microbiota was performed after obtaining informed consent from each patient.
Statistical analysis
All statistical analyses were performed using SPSS software (version 22.0; IBM Japan, Tokyo, Japan), R version 4.2.2 (The R Foundation for Statistical Computing, Vienna, Austria), JMP Pro version 15.2 (SAS Institute Inc., Cary, NC, USA), and GraphPad Prism version 9.00 (GraphPad Software, La Jolla, CA, USA). Continuous variables such as age, height, and weight were analyzed using the Mann–Whitney U test. Normally distributed variables are presented as mean ± standard deviation (SD), whereas non-normally distributed variables are expressed as median with interquartile range (IQR, 25–75%).
Categorical variables were analyzed using the chi-square test with Yates'continuity correction. Alpha diversity indices were compared using the Wilcoxon test. For beta diversity analysis, both unweighted and weighted UniFrac distances were evaluated, and statistical significance was assessed using permutational multivariate analysis of variance (PERMANOVA), in accordance with previous studies.
Weighted UniFrac represents a quantitative assessment that accounts for the relative abundance of taxa, whereas unweighted UniFrac represents a qualitative assessment based solely on the presence or absence of taxa. To evaluate whether significant PERMANOVA results were influenced by differences in within-group dispersion, PERMDISP analysis was additionally performed using the betadisper() function in the vegan package in R (Version 4.4.2). Statistical significance was defined as p<0.05.
Benjamini-Hochberg false discovery rate correction was performed for the analysis of taxa comparing IBS-D patients to healthy control. Then, adjusted q values were calculated. Statistical significance was defined as q values <0.05.
Results
Twenty patients (11 males and 9 females) with IBS-D analyzed in our previous study (MIHAN-R study) were reviewed, with a median age [IQR] of 69.5 [62.7–76.0] years; 35.0% were aged ≥75 years (Table 1). The median BMI [IQR] was 22.3 [18.5–23.8]. There were 14 effective (9 males and 5 females) and 6 non-effective cases (2 males and 4 females).
Their median age [IQR] were 69.5 [54.5–75.8] years and 69.5 [65.3–77.5] years. Overall, functional dyspepsia was present in 15.0% of patients, and six patients had previously been treated for IBS-D. Two patients (10.0%) continued probiotic therapy.
One adverse event (5.0%) was observed, consisting of mild liver injury and hypokalemia, both resolving without treatment.
In the comparison of gut microbiota between 20 healthy controls and 20 IBS-D patients, alpha diversity showed no significant differences for observed features (p=0.909) or the Shannon index (p=0.966), whereas Faith's PD was significantly higher in IBS-D patients than in healthy controls, indicating increased microbiota diversity in IBS-D patients. (p=0.016) (Figure 1). The sequencing depth for alpha diversity analysis was 10,000 reads.
In contrast, beta diversity analysis demonstrated significant differences in both unweighted (p=0.006) and weighted metrics (p=0.031) with the PERMANOVA analysis (Figure 2). Using PERMDISP analysis, the mean dispersion ± SD in IBS-D patients and healthy controls was 0.411±0.064 and 0.354±0.032, respectively, for unweighted distances (p<0.001), and 0.340±0.065 and 0.265±0.058, respectively, for weighted distances (p<0.001). These results suggest that IBS-D patients exhibited greater heterogeneity in gut microbiota composition than healthy controls.
Therefore, the significant PERMANOVA results may have been partially influenced by differences in within-group dispersion.
Taxonomic analysis demonstrated increased Escherichia–Shigella (p=0.003, q=0.035) and decreased Faecalibacterium (p=0.005, q=0.035) in IBS-D patients compared with healthy controls (Table 2).
In the comparison of pre-HST gut microbiota between 14 effective and 6 non-effective cases, Escherichia–Shigella was significantly increased in non-effective cases (p<0.001, q=0.014) (Table 3). In contrast, Fusobacterium (p=0.004, q=0.019) and Lactobacillus (p=0.002, q=0.014) were significantly decreased in non-effective cases.
Regarding changes in bacterial abundance from pre- to post-HST administration, a decrease in Escherichia–Shigella (3.00% vs. −3.94%, p=0.049, q=0.343) and an increase in Faecalibacterium (−0.65% vs. 2.11%, p=0.012, q=0.168) were observed in non-effective cases compared with effective cases (Table 4). However, these differences were significant only at the nominal p-value level and did not remain significant after BH-FDR correction.
The abundance changes of specific bacteria after HST varied widely (Figure 3). For Bacteroides, 40.0% showed a -30% ～ <-5% decrease and 35.0% showed <-30 decrease. For Escherichia–Shigella, 50% showed 30%≤ increase and 25% showed <-30% decrease increase.
For Faecalibacterium, 30% showed 30%≤ increase and 25% showed <-30% decrease increase.
Clinical improvement rates according to bacterial abundance changes were analyzed for the bacteria which showed significant difference in Analysis 1 (IBS-D vs. healthy controls) in the current study and significant changes after HST administration in our previous study (Table 5)[14]. In the previous analysis, decrease for Bacteroides and Ruminocococcus and increase of Megasphaaera and Subdoligranulum in effective cases were detected significantly after HST administration. Increase of Blautia in non-effective cases was also detected.
However, these changes after HST in the previous study was based on nominal p values. For Bacteroides, the improvement rates were 73.3% in patients with decrease and 75.0% for those with increase. The rates for Escherichia–Shigella were 66.7% and 72.7%.
For Faecalibacterium, the rates were 100% and 57.1%. Overall, the abundance change rates of post-HST for individual bacteria were not consistently associated with clinical efficacy. However, no significant differences were observed between cases with decreased and increased abundance for any bacteria.
Discussion
In this study, we identified significant differences in α diversity and β diversity between healthy controls and patients with IBS-D. Taxonomic analysis revealed significant increase in Escherichia–Shigella and significant decreases in Faecalibacterium in IBS-D patients. Additionally, regarding the comparison of pre-HST microbiota between effective and non-effective cases, there were significant increase in Escherichia–Shigella and significant decrease in Fusobacterium and Lactobacillus.
These results were suggested that this may be useful as a predictive factor for the clinical efficacy of HST. We further evaluated the relationship between bacterial abundance changes and clinical symptom improvement. Changes in the abundance of individual bacteria were not consistently associated with clinical efficacy.
IBS-D is associated with alterations in gut microbial balance, increased intestinal secretion, and mucosal microinflammation[11]. Several microbiota studies conducted in Western populations have demonstrated that IBS-D is associated with gut microbial dysbiosis characterized by increased Escherichia–Shigella, Bacteroides, Prevotella, and Ruminococcus, together with decreased levels of Faecalibacterium prausnitzii, Roseburia, and Bifidobacterium[6,22-25]. These alterations are not consistent in each report.
However, several studies have consistently reported increased abondance of Escherichia–Shigella and Bacteroides, which are associated with lipopolysaccharide production, mucosal immune activation, and intestinal inflammation, and decreased abondance of Faecalibacterium and Bifidobacterium, which are associated with anti-inflammatory bacteria[7,23,24]. Actually, Faecalibacterium prausnitzii is a major butyrate-producing bacterium with anti-inflammatory properties, while Bifidobacterium contributes to mucosal barrier integrity and immune modulation[22-25]. Studies conducted in Japanese populations have also reported gut microbial alterations in IBS patients, including increased Prevotella, Ruminococcus, Dorea, Klebsiella, and Bacteroides, together with decreased levels of Bifidobacterium, Lactobacillus, and Faecalibacterium[26,27].
Our previous study in Japanese populations also demonstrated that healthy individuals with loose stools exhibit distinct gut microbiota profiles compared with healthy individuals, including increased Streptococcus, Dorea, and Ruminococcus and decreased Bifidobacterium and Faecalibacterium[28]. In the current study, we identified significant differences in α diversity and β diversity between healthy controls and IBS-D patients. Taxonomic analysis revealed a significant increase of Escherichia–Shigella and decreases of Faecalibacterium and these findings were consistent with previous reports[22-28].
Regarding gut microbiota changes before and after HST administration, in vitro antimicrobial assays have shown that HST and its active constituents (e.g., baicalein, berberine, and coptisine) inhibit the growth of several oral Gram-negative bacteria[12]. The same study also reported no antibacterial activity against Escherichia coli and Bacteroides thetaiotaomicron. Our previous study showed the decrease of Bacteroides and Ruminococcus and increase of Megasphaera and Subdoligranulum before and after HST administration and this suggests the antibacterial activities of HST[14].
In the current study, we demonstrated changes in the abundance of various bacteria after HST administration in both effective and non-effective cases. Significant differences between the two groups were observed only at the nominal p-value level for Escherichia–Shigella and Faecalibacterium. Given the small sample size, these findings should be considered exploratory and interpreted with caution.
Further studies with larger sample sizes are warranted to clarify the antibacterial effects and underlying mechanisms of HST.
Regarding the comparison of pre-HST gut microbiota between effective and non-effective cases, significant differences were detected in increase of Escherichia–Shigella and decrease of Fusobacterium, and Lactobacillus in the non-effective cases. These results were suggested that this may be useful as a predictive factor for the effect of HST, particularly when combined with the gut microbiota findings in healthy controls, which was analyzed in the current study.
Our study uniquely demonstrated the relationship between bacterial abundance changes and clinical symptom improvement. Overall, the abundance increase and decrease of post-HST for individual bacteria were not consistently associated with clinical efficacy. This suggests that predicting the clinical efficacy of HST based solely on changes in single bacteria may be difficult.
Further investigations considering the combined effects of multiple bacterial taxa and other contributing factors are warranted.
This study has several limitations. The sample size was relatively small and included only Japanese patients. Stool frequency and stool characteristics were evaluated based on patient reports and may therefore be subjective.
Improvement in abdominal pain was assessed subjectively by both physicians and patients. The healthy controls were not taking medications for constipation or diarrhea. We did not analyze microbial metabolites that might be associated with mucosal permeability and microinflammation.
16S rRNA sequencing provides compositional data based on relative abundance. Therefore, a decrease in one bacteria may lead to an apparent relative increase in others without reflecting a true change in absolute abundance. We analyzed the gut microbiota of only 20 cases with IBS-D among the whole 100 cases, which agreed with the analysis of the microbiota in the original study[14].
The observed bacterial abundance changes in the current study may reflect the antibacterial activity of HST, which has been suggested in previous studies. However, the present analyses only demonstrate associations in bacterial abundance and do not provide direct evidence of antibacterial effects. Therefore, the relationship between the observed microbiota changes and the proposed antibacterial mechanism remains unclear, and further mechanistic studies are required to clarify this issue.
In conclusion, characteristic microbiota profiles of IBS-D patients were identified. Significant differences in bacterial abundance before HST administration were observed between effective and non-effective cases.
Author contributions
Yoshida N organized the study, performed data collection and analysis, and prepared the manuscript; Yasuda T and Uchiyama K, helped in the analysis of gut microbiota; Inagaki Y, Tomita Y, Hirose R, Kobayashi R, Iwai N, Inoue K, Dohi O, enrolled the cases; and Takagi T arranged the study plan and reviewed the manuscript.
IRB Number
ERB-C-2284 in the Ethics Committee of Kyoto Prefectural University of Medicine
Conflict of interests
Naohisa Yoshida received research grants and lecture fee from Tsumura Co.. The other author declares no conflict of interest for this article.