Electroacupuncture alleviates symptoms and identifies a potential microbial biomarker in patients with constipation-predominant irritable bowel syndrome

Abstract

BACKGROUND

Irritable bowel syndrome with predominant constipation (IBS-C) is a chronic gastrointestinal disorder that significantly impacts the quality of life of patients and currently lacks a definitive treatment. The use of electroacupuncture (EA) has demonstrated clinical efficacy in treating IBS-C and the gut-brain axis modulation, though its mechanisms remain unclear.

AIM

To investigate gut-brain-microbiota axis alteration and EA-associated microbial changes in IBS-C patients and treatment responders.

METHODS

This study consisted of two phases. The first phase was a cross-sectional study recruiting sixteen IBS-C patients and 16 healthy controls. Baseline fecal samples were collected to assess gut microbiota profiles between the two groups. The second phase was an observational longitudinal study in which the 16 IBS-C patients underwent nine EA sessions over one month. Gut microbiota profiles and clinical outcomes were assessed post-treatment course and at a one-month follow-up.

RESULTS

IBS-C patients exhibited significant gut dysbiosis, as indicated by altered beta diversity compared to healthy controls. EA significantly improved clinical outcomes and gut dysbiosis, with sustained therapeutic effects and normalization of neurotransmitter-related metabolic pathways observed at one-month follow-up. Notably, the gut bacterium Senegalimassilia was positively associated with symptom improvement, suggesting its potential as a predictive biomarker of EA responsiveness.

CONCLUSION

These findings support the integration of EA into IBS-C management and highlight Senegalimassilia as a candidate microbial biomarker for treatment response.

Key Words: Irritable bowel syndrome; Traditional Chinese Medicine; Gut microbiota; Constipation; Microbial biomarker

Core Tip: Given the complex pathophysiology of irritable bowel syndrome (IBS), inconsistency of gut microbiota profiles, and the scarce evidence regarding constipation subtype, current treatments often focus on symptom management. Electroacupuncture (EA), as an alternative or complementary approach, has shown clinical efficacy. Our study demonstrated that EA not only alleviated constipation-predominant type (IBS-C) symptoms but also modulated gut dysbiosis and normalized neurotransmitter-related metabolic pathways. Notably, Senegalimassilia was positively linked with clinical improvement, suggesting its potential as a predictive biomarker of EA responsiveness. Additionally, EA offered sustained beneficial therapeutic effects in IBS-C through gut microbiota modulation, supporting its role in clinical practice.

INTRODUCTION

Irritable bowel syndrome (IBS) is a common disorder of the gastrointestinal tract, characterized by recurrent abdominal pain or discomfort along with various abnormal gastrointestinal motility symptoms. According to the Rome IV criteria, IBS is categorized into four subtypes: Diarrhea-predominant type (IBS-D), constipation-predominant type (IBS-C), mixed/alternating type (IBS-M), or an undefined type (IBS-U), depending on the predominant stool pattern[1]. The pathogenesis of IBS remains unclear but is thought to be related to immune dysregulation, neurological dysfunction, and psychological disorders[2]. Several studies have suggested that increased oxidative stress, indicated by elevated serum malondialdehyde (MDA) levels, along with reduced antioxidant capacity, may contribute to the pathophysiology of IBS[3,4]. Dysregulation of the gut-brain axis is also implicated, as alterations in autonomic and endocrine systems—such as altered cortisol or serotonin levels—and gut dysbiosis have been observed in IBS patients[3,5,6].

The pathogenesis of each IBS subtype differs. IBS-C patients have been shown to exhibit lower levels of proinflammatory cytokine release, reduced parasympathetic tone, and higher autonomic nervous system (ANS) activity compared to IBS-D patients[4,7]. Several studies on the gut microbiota in IBS have been conducted in IBS-D patients[8,9], while the gut microbiota profiles of IBS-C patients remain underexplored. The management of IBS primarily aims at symptomatic relief through pharmacological therapies and psychological interventions[1]. Recently, non-pharmacological approaches, such as acupuncture, have gained increasing attention. A previous study suggests that acupuncture may offer comparable therapeutic effects with fewer adverse effects than traditional pharmacological therapies[10].

Acupuncture is widely used in Traditional Chinese Medicine (TCM) to harmonize functions of the body’s systems by balancing energy flows, known as Qi, and modulating blood flow[11]. In the context of TCM diagnosis for IBS[12], IBS-C may present with either a ‘deficiency syndrome’ or ‘excess syndrome’, each influencing bowel patterns differently. Therapeutic techniques such as manual acupuncture, moxibustion, and electroacupuncture (EA) have been employed in the treatment of IBS. Tianshu (ST25), Zusanli (ST36), and Shangjuxu (ST37) are specific acupoints commonly used to alleviate symptoms of IBS, particularly by reducing visceral hypersensitivity and enhancing gut motility[11,13-17]. Acupuncture can enhance intestinal motility and visceral sensitivity by regulating gut-brain peptide levels in both enteric and central nervous systems[13,17]. Previous studies have demonstrated that EA ameliorated gastrointestinal symptoms by balancing the gut-brain axis[13,16,17], and has been reported to be more effective in treating IBS-C than other acupuncture techniques[16,17]. However, no studies have yet investigated the effects of EA on gut microbiota changes and clinical improvements in IBS-C patients. The present study aimed to: (1) Investigate the alterations in the gut-brain-microbiota axis in patients with IBS-C compared to healthy subjects; (2) Assess the alteration of gut microbiota composition in patients with IBS-C who responded to EA treatment; and (3) Identify gut microbial signatures as potential biomarkers for predicting EA treatment responsiveness.

MATERIALS AND METHODS

Study design

The study protocol was approved by the Institutional Ethics Committee of the Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand (Ethical approval number: 296/2021). This study was designed as an observational longitudinal study and consisted of two phases: A cross-sectional study comparing gut-brain-microbiota profiles between IBS-C patients and healthy controls (Phase 1) and a single-arm, before-after observational study evaluating the effects of EA on IBS-C patients (Phase 2). Both phases of the study were conducted at Sriphat Medical Center, Faculty of Medicine, Chiang Mai University, from September 2, 2021 to July 7, 2022.

In Phase 1, 16 IBS-C patients (diagnosed according to the Rome IV criteria) and 16 healthy controls, aged 18 to 65 years, were recruited. In Phase 2, patients from Phase 1 received nine EA sessions within one month. This before-after observational study assessed clinical and biological parameters immediately after treatment and one-month follow-up to evaluate both the immediate and sustained effects of EA. Clinical outcomes included IBS symptom severity scale scores (IBS-SSS), IBS quality of life (IBS-QoL), Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA) scores. Biological measures, including plasma cortisol, MDA, serotonin levels, and gut microbiota profiles, were evaluated to assess changes following treatment (Supplementary material and Supplementary Figure 1).

Heart rate variability

A variation in time between adjacent heartbeats, or heart rate variability (HRV), was used to evaluate the balance of cardiac autonomic regulation. An electrocardiogram was recorded continuously using the SEER Light Holter system (GE Healthcare, Milwaukee, WI, United States) for 90 minutes in both phases. The HF power mainly reflects parasympathetic tone, whereas the LF power and LF/HF ratio mainly indicate sympathovagal interaction and sympathetic tone, respectively.

Determination of plasma morning cortisol and serotonin levels

Whole blood was collected from subjects between 8 A.M. and 9 A.M. The sample was centrifuged at 3000 rpm for 10 minutes, and plasma was collected. Plasma cortisol levels were determined using an automatic enzyme-linked immunosorbent assay (Roche, Basel, Switzerland). Plasma serotonin levels were measured using a commercially available ELISA kit (#ab133053, Abcam, United Kingdom).

Determination of plasma MDA levels

Plasma MDA levels were quantified by high-performance liquid chromatography (Thermo Scientific, Massachusetts, United States) to investigate systemic oxidative stress, as previously mentioned. The absorption spectra of TBARS exposed at 532 nm were measured by chromatography, and plasma MDA levels were determined directly from a standard curve.

Collection and analysis of fecal samples

Fecal samples (approximately 50 g) were collected in sterile plastic cups and stored in RNAlater (Invitrogen, Massachusetts, United States) before being kept at -80 ^oC. Fecal samples were obtained, and microbiome evaluated in both controls and IBS-C cohort at baseline, followed by testing post-treatment, and at the one-month follow up. A one-month interval was chosen to assess the long-term effects of EA therapy on gut microbiota. This interval allowed time for the gut microbiota to stabilize and provided insights into whether the observed changes were sustained beyond the immediate effect of treatment. Additionally, this approach minimized the influence of acute responses, ensuring that the results reflected the true impact of EA on gut dysbiosis. Genomic DNA was extracted from fecal samples using a commercial DNA extraction kit (QIAamp Power Fecal Pro DNA Kit, Qiagen, Germany). The 16S rRNA hypervariable region V3-V4 of extracted bacterial genomic DNA was sequenced using the Illumina NovaSeq 6000 platform (Novogene Inc., Singapore). Sequence data were processed using Quantitative Insights Into Microbial Ecology 2 (QIIME2-2024.10). The paired-end reads data were merged, denoised, trimmed according to quality scores, and assigned to amplicon sequence variants (ASVs) were inferred using the q2-dada2 plugin. A feature table was then generated based on ASVs. Multiple Sequence Alignment was performed using MAFFT, and the resulting alignments were used to reconstruct a phylogenetic tree for downstream analysis. Taxonomy classification was assigned to ASVs using the q2-feature-classifier plugin with the classify-sklearn naïve Bayes classifier, referencing the SILVA database (version 138). Differences in taxa abundance between groups and associations between taxa and IBS symptom scores were estimated at each taxonomic level with a statistical framework. These were determined by Analysis of the Composition of Microbiomes with Bias Correction (ANCOM-BC). Gut microbial diversity and composition were analyzed to yield alpha diversity indices and beta diversity indices, which are presented as boxplots and principal coordinate analysis (PCoA) plots. The findings were visualized in R version 4.3.1 Beagle Scouts with the package qiime2R. Phylogenetic Investigation of Communities by Reconstruction of Unobserved States 2 version 2.5.2 inferred metabolic pathways based on the MetaCyc database. For this analysis, we considered only pathways with a P value < 0.05 to ensure precision and specificity in the interpretation of the data.

Statistical analysis

Statistical analyses were performed using Stata 16, with P value of less than 0.05 considered statistically significant. Normality was assessed using the Shapiro-Wilk test. Data are presented as mean ± SD for parametric variables and as median with interquartile range (IQR) for non-parametric variables. In the initial cross-sectional analysis, comparisons between two groups were conducted using Student’s t-test for parametric data, the Mann-Whitney U test for non-parametric data, and the χ² test for categorical variables, as appropriate. For multiple comparisons of continuous parameters across different time points within the IBS-C group, repeated measures ANOVA with Bonferroni correction was employed for parametric variables, while the Kruskal-Wallis test with Dunn’s post hoc test was employed for non-parametric variables.

RESULTS

Differences in clinical characteristics between the IBS-C and control groups

This study included 32 participants (16 controls and 16 IBS-C patients). The mean age (control: 34.37 ± 8.81 years; IBS-C: 35.50 ± 7.52 years) and BMI (control: 23.50, IQR: 20.56-30.34 kg/m²; IBS-C: 20.96, IQR 19.20-25.67 kg/m²) did not differ significantly between the groups. The majority of participants in both groups were female, with no significant differences in overweight or obesity status, or basic laboratory parameters used to exclude organic conditions (Table 1). Dietary intake across 16 food categories records also showed no significant differences between the healthy control group and patients with IBS-C (Supplementary Table 1).

Table 1 Demographic data.

Characteristic	Control (n = 16)	IBS-C (n = 16)	95%CI	P value
Age (years)	34.37 ± 8.81	35.50 ± 7.52	-7.037, 4.787	0.700
Gender (female/male)	10/6	13/3		0.433
BMI (kg/m2)	23.50 (20.57-30.34)	20.96 (19.20-25.67)		0.247
Overweight, %	20	14.28		0.352
Obesity, %	33.33	21.43
Laboratory
Hb (g/dL)	13.44 ± 1.98	13.20 ± 0.58	-0.893, 1.369	0.669
Hct (%)	41.65 ± 5.00	40.21 ± 2.36	-1.576, 4.459	0.335
Platelet (cell/mL)	291285.70 ± 68977.62	286678.60 ± 59053.38	-45276.790, 54491.070	0.851
FBS (mg/dL)	88.00 (83.00-92.00)	87.50 (85.25-94.00)		0.642
Total cholesterol (mg/dL)	209.57 ± 38.06	195.83 ± 35.17	-16.115, 43.591	0.606
LDL-c (mg/dL)	135.07 ± 30.12	120.26 ± 31.65	-10.230, 39.840	0.234
HDL-c (mg/dL)	56.57 ± 10.82	64.75 ± 13.23	-17.912, 1.555	0.095
LDL/HDL ratio	2.49 ± 0.78	1.96 ± 0.70	-0.075, 1.136	0.083
Triglycerides (mg/dL)	85.00 (58.50-119.75)	74.00 (51.00-183.75)		0.572
AST (U/L)	19.00 (17.75-24.25)	17.00 (16.50-23.25)		0.179
ALT (U/L)	20.00 (12.75-25.50)	13.00 (9.50-18.00)		0.092
AST/ALT ratio	1.18 ± 0.34	1.41 ± 0.57	-0.743, -0.0008	0.202

Data are presented as mean ± SD for parametric variables (numerical data), median (IQR) for non-parametric variables (numerical data), and proportion (categorical data). The groups were compared using Student’s t-test, Mann-Whitney U test or χ² test as appropriate. ALT: Alanine transaminase; AST: Aspartate transaminase; BMI: Body mass index; Hb: Hemoglobin; Hct: Hematocrit; HDL-c: High-density lipoprotein cholesterol; FBS: Fasting blood sugar; IBS: Irritable bowel syndrome; LDL-c: Low-density lipoprotein cholesterol.

IBS-C patients had disease durations ranged from 1 year to more than 15 years and significantly higher IBS-SSS scores (control: 25.94 ± 34.31; IBS-C: 251.19 ± 57.19) and HAD-anxiety scores (control: 6.25 ± 3.19; IBS-C: 8.87 ± 3.95). IBS-QoL scores were significantly higher in the IBS-C group (control: 40.75 ± 7.66; IBS-C: 88.81 ± 17.21), indicating worse quality of life. Plasma cortisol levels were higher in IBS-C patients (control: 8.25, IQR 7.04-9.32 μg/dL; IBS-C: 10.90, IQR 8.92-16.05 μg/dL), while plasma MDA levels, serotonin levels, and HRV parameters (HF, LF, LF/HF ratio) showed no significant differences. Cognitive and mood scores (MoCA, MMSE, HAD-depression scores) were also comparable between groups (Table 2).

Table 2 The comparison in clinical parameters, serum biomarkers and autonomic nervous systems parameters between healthy controls and patients with irritable bowel syndrome with predominant constipation.

Characteristic	Control (n = 16)	IBS-C (n = 16)	P value
Clinical score
IBS-SSS score	25.94 ± 34.31	251.19 ± 57.19	< 0.0001
IBS-QoL	40.75 ± 7.66	88.81 ± 17.21	< 0.0001
Memory assessment
MoCA	25.50 ± 3.05	25.62 ± 2.47	0.899
MMSE	25.81 ± 2.64	26.75 ± 1.69	0.241
Emotional assessment
Anxiety	6.25 ± 3.19	8.87 ± 3.95	0.047
Depression	3.62 ± 2.94	4.56 ± 2.39	0.331
Laboratory
Cortisol level (μg/dL)	8.25 (7.04-9.32)	10.90 (8.92-16.05)	0.018
MDA level (nmol/mL)	393.59 ± 127.89	377.17 ± 86.57	0.670
Serotonin level (μg/dL)	0.80 (0.43-3.49)	1.45 (0.51-3.14)	0.555
Heart rate variability
LF	22.97 ± 11.89	23.15 ± 5.38	0.958
HF	11.03 (7.68-17.75)	14.38 (10.33-19.91)	0.112
LF/HF	1.75 (1.43-2.08)	1.48 (1.24-1.53)	0.153

Data are presented as mean ± SD for parametric variables (numerical data), and median (IQR) for non-parametric variables. The groups were compared using Student’s t-test or Mann-Whitney U test or χ² test as appropriate. HF: High frequency; IBS: Irritable bowel syndrome; IBS-SSS: Irritable bowel syndrome symptom severity scale scores; IBS-QoL: Irritable bowel syndrome quality of life; LF: Low frequency; MoCA: Montreal Cognitive Assessment; MMSE: Mini-Mental State Examination; HADs: Hospital Anxiety and Depression Scale; MDA: Malondialdehyde.

Differences in gut microbiota characteristics between the IBS-C and control groups

Alpha diversity showed no significant difference between the IBS-C and control groups (Figure 1A). However, beta diversity differed significantly, as shown in PCoA plots (Figure 1B). The orders Eubacteriales and Actinomycetales were significantly increased in the IBS-C group, while the orders Rhodobacterales, Rhizobiales, Cardiobacteriales, and Aeromonadales were markedly decreased in IBS-C group compared to healthy control group (Figure 1C). Similarly in the family and genus levels, the gut microbiota composition in IBS-C group was significantly distinct from that of the control group.

Figure 1 Diversity of gut microbiota and their composition in healthy controls, patients with irritable bowel syndrome with predominant constipation at baseline and after electroacupuncture. A: Alpha diversity is presented as boxplots; B: Beta diversities were presented by the principal coordinate analysis plot; C: Differential abundance calculated by Analysis of the Composition of Microbiomes with Bias Correction are presented as bar plots of beta coefficients and standard error of beta. Orange and blue colors represent decrease and increase of gut microbiota respectively, with the corresponding columns. Only taxa with significant associations after being adjusted with the Bonferroni correction are presented. IBS: Irritable bowel syndrome patients; ACP: Irritable bowel syndrome patients after electroacupuncture; PERMANOVA: Permutational multivariate analysis of variance.

Effects of EA on clinical, biological, and HRV outcomes of IBS-C patients

All IBS-C patients completed EA treatment and one-month follow-up. Food diary data showed no significant differences across time points (Supplementary Table 2). Immediately post-treatment, a significant reduction in IBS-SSS was observed, followed by variable changes at one-month follow-up (Supplementary Figure 2). Primary clinical assessments revealed that most IBS-C patients experienced sustained clinical improvement throughout the treatment period and at one-month follow-up, as reflected by reductions in IBS-SSS and IBS-QoL scores (Supplementary material and Supplementary Figure 3). At the one-month follow-up, IBS-C patients were classified as responders (IBS-SSS score < 150, indicating mild symptoms) and non-responders (IBS-SSS score > 150, indicating moderate to severe symptoms). Significant differences in IBS-SSS and IBS-QoL score changes were observed between these groups at one-month follow-up (Table 3). However, no significant differences over time were observed in changes in cognitive and emotional assessment, or in serum biomarkers between groups (Table 3).

Table 3 The comparison in clinical parameters, and serum biomarkers between responder and non-responder in patients with irritable bowel syndrome with predominant constipation after electroacupuncture treatment course.

Characteristic	IBS-C		P value
	Responder (n = 13)	Non-responder (n = 3)
Clinical score
IBS-SSS score	280.00 (205.00, 314.50)	200.00 (180.00, 270.00)	0.203
Delta 1	210.00 (170.00, 264.50)	130.00 (30.00, 220.00)	0.160
Delta 2	220.00 (137.50, 269.50)	-40.00 (-70.00, 0.00)	0.003
IBS-QoL	89.00 (77.00, 103.00)	51.00 (94.00, 110.00)	0.975
Delta 1	33.00 (10.00, 45.50)	14.00 (4.00, 28.00)	0.278
Delta 2	36.00 (23.00, 54.50)	0.00 (-1.00, 18.00)	0.014
Cognitive assessment
MoCA	26.00 (23.50, 27.50)	27.00 (26.00, 28.00)	0.332
Delta 1	-2.00 (-4.50, -1.00)	-2.00 (-3.00, -2.00)	0.728
Delta 2	-2.00 (-4.50, -1.00)	-3.00 (-3.00, -1.00)	0.945
MMSE	27.00 (26.00, 28.00)	27.00 (25.00, 28.00)	0.871
Delta 1	0.00 (-2.00, 1.00)	-1.00 (-4.00, -1.00)	0.210
Delta 2	0.00 (-2.00, 0.50)	-2.00 (-3.00, 0.00)	0.492
Emotional assessment
Anxiety	9.00 (7.50, 12.50)	6.00 (2.00, 13.00)	0.471
Delta 1	4.00 (2.50, 6.50)	-1.00 (-1.00, 3.00)	0.064
Delta 2	3.00 (0.50, 9.50)	-2.00 (-2.00, 3.00)	0.150
Depression	4.00 (3.00, 5.50)	4.00 (3.00, 7.00)	0.846
Delta 1	2.00 (-0.50, 4.00)	0.00 (-6.00, 1.00)	0.167
Delta 2	2.00 (0.50, 3.00)	0.00 (-7.00, 1.00)	0.096
Laboratory
Cortisol level (μg/dL)	10.70 (8.60, 14.45)	14.10 (10.90, 24.40)	0.228
Delta 1	-2.00 (-4.78, 1.94)	3.19 (-6.90, 4.82)	0.610
Delta 2	0.89 (-7.36, 1.89)	-3.40 (-4.30, 2.60)	0.946
MDA level (nmol/mL)	364.54 (323.68, 441.86)	354.19 (251.30, 453.51)	0.703
Delta 1	-207.08 (-350.18, -63.88)	-465.16 (-536.37, -165.54)	0.296
Delta 2	-83.22 (-230.66, 3.74)	-169.65 (-178.08, 54.15)	0.900
Serotonin level (μg/dL)	1.60 (0.62, 2.78)	0.858 (0.33, 16.36)	0.754
Delta 1	-0.25 (-2.42, 0.25)	0.19 (-2.53, 16.37)	0.481
Delta 2	0.29 (-0.84, 1.31)	-0.37 (-0.95, 15.37)	0.863

Data are presented as the median (IQR). Comparisons between the groups were conducted using Mann-Whitney U test. IBS-SSS: Irritable bowel syndrome symptom severity scale scores; IBS-QoL: Irritable bowel syndrome quality of life; MoCA: Montreal Cognitive Assessment; MMSE: Mini-Mental State Examination; HADs: Hospital Anxiety and Depression Scale; MDA: Malondialdehyde; Delta 1 Δ (Pre-treatment) – (Post-treatment); Delta 2 Δ (Pre-treatment) – (Follow-up).

Among IBS-C responders, significant reductions were observed in IBS-SSS, IBS-QoL (where lower score indicates better quality of life), and HAD anxiety scores, along with a significant increase in MoCA scores. These improvements remained significant at the one-month follow-up. No significant changes were found in plasma cortisol and serotonin levels across time points. Plasma MDA levels significantly increased at post-treatment, followed by a decrease at the one-month follow-up, with no significant difference compared to pre-treatment levels. Additionally, no significant changes were observed in HRV frequency domain measures (HF, LF, and LF/HF ratio) across pre-treatment, post-treatment, and follow-up periods (Table 4).

Table 4 Effects of electroacupuncture on clinical, biological, and heart rate variability outcomes in irritable bowel syndrome with predominant constipation responders at baseline, post-treatment, and one-month follow-up.

Outcome measures	IBS-C responders (n = 13)			P value
	Pre-treatment	Post-treatment	Follow-up
Clinical score
IBS-SSS score	280.00 (205.00-314.50)	30.00a (0.00-85.00)	45.00a (27.50-82.50)	< 0.001
IBS-QoL	89.00 (77.00-103.00)	55.50a (49.25-80.50)	47.00a (44.00-56.50)	< 0.001
Memory part
MoCA	25.30 ± 2.62	27.76 ± 1.69a	28.00 ± 1.35a	< 0.001
MMSE	26.76 ± 1.78	27.00 ± 1.58	27.53 ± 1.45	0.284
Emotional part
Anxiety	9.30 ± 3.63	4.61 ± 2.18a	5.08 ± 3.67a	< 0.001
Depression	4.53 ± 2.53	2.76 ± 2.35	3.00 ± 3.16	0.060
Laboratory
Cortisol level (μg/dL)	10.70 (8.60-14.45)	12.30 (9.07-19.65)	9.81 (7.65-18.30)	0.774
MDA level (nmol/mL)	382.75 ± 86.50	608.63 ± 141.19a	484.17 ± 132.80	< 0.001
Serotonin level (μg/dL)	1.60 (0.62-2.78)	2.86 (1.50-3.73)	0.92 (0.77-2.23)	0.251
Heart rate variability
LF	22.08 ± 4.26	21.20 ± 5.51	20.16 ± 5.61	0.465
HF	15.20 ± 5.07	14.24 ± 5.10	13.05 ± 4.56	0.398
LF/HF	1.46 (1.26-1.53)	1.54 (1.13-1.94)	1.45 (1.34-1.81)	0.833

^aP < 0.05 vs pretreatment.

Data are presented as mean ± SD for parametric variables (numerical data), and median (IQR) for non-parametric variables. Multiple comparisons were conducted using repeated measures ANOVA with Bonferroni correction for parametric variables, and Kruskal-Wallis test with Dunn’s post hoc for non-parametric variables. HF: High frequency; IBS: Irritable bowel syndrome; IBS-SSS: Irritable bowel syndrome symptom severity scale scores; IBS-QoL: Irritable bowel syndrome quality of life; LF: Low frequency; MoCA: Montreal Cognitive Assessment; MMSE: Mini-Mental State Examination; MDA: Malondialdehyde; Pre-treatment: IBS-C patients before acupuncture; Post-treatment: IBS-C patients on the last day of acupuncture; Follow-up: IBS-C patients after acupuncture for one month.

Effects of EA on diversity analysis of gut microbiota profiles in IBS-C responders

Alpha diversity, as measured by the Shannon index, showed no significant changes over time among IBS-C responders (Figure 2A). In terms of beta diversity, Jaccard similarity analysis revealed that gut microbiota profiles in IBS-C responders at post-treatment clustered more closely with their pre-treatment profiles. However, at the one-month follow-up, the gut microbiota composition shifted further, exhibiting a pattern more similar to that of the control group (Figure 2B). ANCOM-BC analysis detected minor compositional changes at post-treatment compared to baseline, including an increase in the genus Intestinimonas and decreases in Enterococcus, Butyrivibrio, and Eubacterium_siraeum_group. At one-month follow-up, several taxa exhibited reciprocal changes compared to pre-treatment, including increases in the orders Rhodobacterales and Cardiobacteriales; families Yersiniaceae, Rhodocyclaceae, Rhodobacteraceae, Microbacteriaceae, Comamonadaceae, and Cardiobacteriales; and genera Serratia, Rikenellaceae_RC9_gut_group, Cardiobacteriales, and C39. Furthermore, minimal decrease in genera Fournierella and Eubacterium_fissicatena_group at one-month follow-up compared to pre-treatment (Figure 2C).

Figure 2 Diversity of gut microbiota and their composition in healthy controls, patients with irritable bowel syndrome with predominant constipation at baseline, after electroacupuncture and one-month follow-up. A: Alpha diversity is presented as boxplots; B: Beta diversities were presented by the principal coordinate analysis plot; C: Differential abundance calculated by Analysis of the Composition of Microbiomes with Bias Correction are presented as bar plots of beta coefficients and standard error of beta. Orange and blue colors represent positive and negative associations respectively, with the corresponding columns; D: Correlation between microbial genera change in irritable bowel syndrome (IBS) symptom including IBS-SSS and IBS-QoL. Only taxa with significant associations after being adjusted with the Bonferroni correction are presented. ACP: Irritable bowel syndrome patients after electroacupuncture; ACP1M: Irritable bowel syndrome patients after electroacupuncture and one-month follow-up; IBS: Irritable bowel syndrome patients; IBS-SSS: Irritable bowel syndrome symptom severity scale scores; IBS-QoL: Irritable bowel syndrome quality of life; PERMANOVA: Permutational multivariate analysis of variance.

Correlation between microbial genera and IBS clinical symptom outcomes in IBS-C responders

In our study, several gut microbial families were found to be significantly correlated with changes in IBS-SSS scores. Positive correlations were observed for Yersiniaceae, UCG-010, Rhodocyclaceae, Rhodobacteraceae, Microbacteriaceae, Comamonadaceae, and Cardiobacteriales, while Akkermansiaceae was negatively correlated with IBS-SSS score changes (Figure 2D). Furthermore, only the family Enterobacteriaceae showed a negative correlation with changes in IBS-QoL scores (Figure 2D).

Predictive metabolism pathways of gut microbiota

The differential abundance of metabolic pathways in responders was analyzed across three time points: Pre-treatment IBS, post-treatment IBS, and the one-month follow-up groups. Seven neurotransmitter-related pathways were significantly altered in pre-treatment IBS patients compared to controls. These included folate transformations I, L-arginine degradation II, L-histidine degradation II, L-lysine biosynthesis V, and three superpathways related to arginine, putrescine, and tryptophan metabolism. Following EA treatment, the abundance of these pathways gradually shifted toward normal levels, with sustained normalization observed at the one-month follow-up (Figure 3).

Figure 3 Predicted neurotransmitter-related metabolic pathways analyzed by Phylogenetic Investigation of Communities by Reconstruction of Unobserved States 2, compared between healthy controls and irritable bowel syndrome with predominant constipation responders to electroacupuncture, indicating potential microbiome shifts associated with therapeutic response. ACP: Irritable bowel syndrome patients after electroacupuncture; ACP1M: Irritable bowel syndrome patients after electroacupuncture and one-month follow-up; IBS: Irritable bowel syndrome patients.

Gut microbiota abundance in EA responders and non-responders IBS-C patients

Differential abundant bacteria taxa in EA responders and non-responders, compared to healthy controls, revealed distinct gut microbiota profiles (Figure 4A). Comparative analysis of the gut microbiota across groups is illustrated using a Venn diagram. Notably, several bacterial taxa were found to be uniquely altered in both IBS-C and EA responders, including an increase in UBA1819, Eubacterium, DTU089, Aciaminococcus, and Selimonas, and a decrease in Lachnospiraceae_UCG-001, Mogibacterium, and Senegalimassilia (Figure 4B). Among these, Senegalimassilia showed a significant positive correlation with changes in both IBS-SSS and HAD anxiety scores (Figure 4C). Linear regression analysis revealed a significant association between log10-transformed baseline abundance of Senegalimassilia and clinical improvement in IBS-C symptoms (β = 94.41, P = 0.032), suggesting that higher abundance at baseline may predict greater clinical improvement following EA treatment.

Figure 4 Gut composition in irritable bowel syndrome with predominant constipation responders to electroacupuncture and irritable bowel syndrome with predominant constipation non-responders to electroacupuncture. A: Differential abundance calculated by Analysis of the Composition of Microbiomes with Bias Correction are presented as bar plots of beta coefficients and standard error of beta. Orange and blue colors represent positive and negative associations respectively, with the corresponding columns; B: Shared differentially abundant genera across responders and non-responder to electroacupuncture in patients with irritable bowel syndrome with predominant constipation by Venn diagram; C: Correlation between electroacupuncture-responsive bacterial and change in clinical scores. Only taxa with significant associations after being adjusted with the Bonferroni correction are presented. ^aP < 0.01. ACP: Irritable bowel syndrome patients after electroacupuncture; HAD: Hospital Anxiety and Depression Scale in the part of anxiety; IBS: Irritable bowel syndrome patients; IBSS: Irritable bowel syndrome symptom severity scale scores; IBSQ: Irritable bowel syndrome quality of life; MDA: Malondialdehyde.

DISCUSSION

The significant differences in IBS-SSS and IBS-QoL scores between the control and IBS-C groups confirmed the clinical severity of IBS-C, which aligns with previous studies linking psychological stress to IBS symptoms[18,19]. Prior research has suggested that IBS-C patients may experience greater psychological distress than those with other IBS subtypes, with anxiety being positively correlated with symptom severity[19]. Elevated cortisol levels observed in IBS-C patients support the presence of an active gut-brain axis, aligning with findings on stress-related hormonal changes in IBS[3,20].

Serotonin plays a crucial role in modulating gastrointestinal function via vagal afferent fibers[21]. However, in our study, no significant difference in plasma serotonin levels was found between IBS-C patients and control groups, nor was there any correlation with HAD scores. This lack of significance may reflect the limited sample size and resulting statistical power. Another potential limitation is that blood samples were collected only in the morning, which may not fully capture serotonin pathway dysfunction. Previous evidence has reported abnormal serotonin signaling has been reported in both IBS-D and IBS-C, with studies showing reduced postprandial plasma serotonin levels reported in IBS-C patients[5,22]. Additionally, gender differences may have influenced the results, as prior studies have shown significantly lower serotonin levels in female IBS patients[23]. Since the majority of participants in our study were female, this may have contributed to these findings.

No significant differences in HRV parameters were observed, suggesting that autonomic dysfunction may not contribute significantly to IBS-C. However, this contrasts with previous studies reporting ANS dysfunction in non-constipated IBS patients[6,20]. These discrepancies may stem from differences in IBS subtype, treatment durations, HRV protocols, or sample size limitation.

Although no significant differences in alpha diversity were observed, our study revealed significant distinctions in gut microbial composition (beta diversity), indicating that while the overall richness and evenness of gut microbial species remain comparable, specific taxa differed between IBS-C patients and healthy individuals, consistent with previous studies[9,24]. This supports the characterization of IBS-C as a functional gastrointestinal disorder, marked by physiological dysfunction without overt structural abnormalities or inflammation. Notably, we observed increased relative abundances of the orders Actinomycetales and Eubacteriales in IBS-C patients. Actinomycetales, belonging to the class Actinobacteria, had shown inconsistent associations with IBS in previous studies. However, some reports suggest that increased levels of Actinobacteria may correlate with the risk of IBS[25] and with altered gut metabolite profiles[26]. Furthermore, Eubacteriales, a member of the class Clostridia, play a crucial role in the production of short-chain fatty acids (SCFAs), particularly butyrate. Butyrate influences serotonin biosynthesis and gastrointestinal functions by improving intestinal barrier function. Sun et al[27] reported lower levels of these SCFAs, particularly butyrate in individuals with IBS-C. Reduced SCFA levels may contribute to decreased serotonin availability, as suggested by mechanistic studies on gut microbiota and enterochromaffin cell function[27]. Additionally, both Clostridiales and Eubacteriales have been implicated in bile acid metabolism, influencing stool consistency[28,29]. However, the role of the Eubacterium genus remains controversial, as it has been associated with both a protective effects and a potential risk factor for constipation[30]. As a key metabolic genus, Eubacterium involved in the breakdown cellulose and digesting of resistant starch. In our study, we observed an increase in genera Eubacterium and Eubacterium fissicatena in the IBS-C patients, suggesting a potentially compensatory or protective microbial shift. Further research is needed to elucidate their specific roles in IBS-C pathophysiology.

EA therapy significantly improved clinical outcomes, as evidenced by reductions in IBS-SSS, IBS-QoL, and HAD-anxiety scores observed immediately after treatment and sustained at one-month follow-up. These long-lasting effects suggest that EA may confer systemic benefits beyond the local treatment site, potentially through modulation of the nervous system, attenuation of inflammation, or enhancement of gastrointestinal function. However, as the clinical assessments were primarily subjective, further research is warranted using objective measures—such as fecal consistency scoring, defecography, or assessments of whole-gut or colon transit time (e.g., Sitzmarks test, scintigraphy or wireless motility capsule)[31]. In addition, MoCA scores improved following treatment and at the one-month follow-up, possibly reflecting EA’s influence on the gut-brain axis, consistent with prior studies[13,16]. EA also attenuated gut dysbiosis, which may contribute to cognitive preservation, supporting emerging evidence that a balanced gut microbiota promotes brain health and mitigates cognitive decline[32,33].

The increase in plasma MDA levels observed post-treatment contrasts with previous studies, which reported decreased MDA levels following EA in brain disorders[34,35]. This discrepancy may be explained by the underlying inflammatory processes in IBS, particularly in IBS-D patients[3,4]. In our study, EA appeared to induce an acute stress response, leading to increased inflammation and elevated plasma MDA levels immediately after treatment. This is further supported by the trend of increased plasma cortisol levels observed post-treatment and at the one-month follow-up.

EA did not significantly affect HRV parameters, suggesting no direct impact on cardiac autonomic regulation in IBS-C patients. This finding aligns with previous studies reporting subtype-specific differences in stress responsiveness and colonic mechanisms across IBS variants, which may account for the absence of HRV changes following treatment[4,7,8,20].

Dietary habits remained consistent across study time points, confirming that EA’s therapeutic effects of EA were independent of dietary changes[36]. Instead, the observed benefits of EA were likely medicated through gut microbiota modulation and gut-brain axis interactions, consistent with finding from previous in vivo studies[14,15]. EA not only alleviates IBS symptoms but also contributed to the restoration of gut dysbiosis and neurotransmitter-related metabolic pathways, with sustained effects observed at the one-month follow-up. This was reflected in the gut microbiota composition of IBS-C responders, which became more similar to that of healthy controls after at one month. Among the taxa that exhibited reciprocal changes at one-month follow-up, several were significantly correlated with improvement in IBS-SSS and IBS-QoL scores. For example, Akkermansia, a key propionate-producing and mucin-degrading microorganisms[37], has been identified as a crucial modulator of gut health. Its increased abundance may reflect adaptive microbial shifts associated with symptom improvement. Additionally, predictive metabolic pathway analysis identified significant alterations in neurotransmitter-associated pathways in pre-treatment IBS-C, which showed a normalization trend following EA therapy. Seven key pathways involved in serotonin, GABA, and NO regulation were altered, linking oxidative stress and neurotransmitter interactions, particularly NO[38]. These findings suggest that EA-induced acute stress responses may modulate the gut microbiota, contributing to IBS symptom recovery. This interpretation is supported by the transient increase in plasma MDA levels post-treatment, which declined at the one-month follow-up. Alternatively, EA may exert therapeutic effects by restoring neurotransmitter-related metabolic pathways, potentially alleviating IBS-C symptoms such as abdominal pain and delayed gastrointestinal transit[39]. Additionally, EA could directly improve gut motility and involve in the alteration of bile acid and SCFA profiles which interact to lipid absorption and gut motility. Furthermore, EA might directly enhance gut motility and contribute to the modulation of bile acid and SCFA profiles, which are known to influence microbial metabolism, lipid absorption and gastrointestinal function[30,40]. However, these findings were based on computational inference rather than direct metabolite measurements. Therefore, further studies should incorporate targeted metabolomic or transcriptomic analyses (e.g., quantification of serotonin, GABA, or related neurotransmitters) to validate these predictions and elucidate the mechanistic roles of these pathways in the therapeutic effects of EA.

Comparative analysis of gut microbiota between IBS-C responders and non-responders revealed a distinct microbial composition in responders, which may serve as a potential biomarker for predicting favorable outcomes to EA treatment. Conversely, non-responders may benefit from adjunctive intervention such as pharmacotherapy or probiotics. Previous studies have demonstrated that acupuncture may outperform certain medications or probiotics in improving IBS symptom[13,41-44], suggesting that gut microbiota profiles could influence the efficacy of specific acupoints. Furthermore, variations in treatment outcomes reported in acupuncture studies have been attributed to acupoint selection, which should ideally be guided by TCM diagnostic principles[12]. Therefore, further in-depth research involving large-scale studies is warranted to validate these findings and to explore personalized acupoint selection strategies in relation to gut microbiota signatures and treatment response.

In exploring the correlation between EA-responsive gut microbiota and changes in clinical symptom, Senegalimassilia was found to be positive associated with improvements in IBS-SSS scores following EA treatment. This suggests that increased abundance of Senegalimassilia may contribute to symptom relief in IBS-C patients. Notably, previous studies have reported that Senegalimassilia is exclusively in IBS-C[45], supporting its potential as microbial marker for predicting clinical response to EA in this population. Although the mechanisms by which Senegalimassilia acts within the gastrointestinal tract remain poorly understood, its taxonomic affiliation with the Eggerthellaceae family suggest potential functional relevance. Members of this family are known to produce urolithin, an anti-inflammatory and antioxidant that enhances tight junction protein expression[46,47]. These findings support that a higher baseline abundance of Senegalimassilia in IBS-C patients may serve as a potential microbial biomarker for predicting EA responsiveness in IBS-C; however, its role remains correlative. To clarify its mechanistic contribution, further studies should employ in vitro co-culture systems, gnotobiotic or humanized mouse models, and metabolomic profiling to investigate host-microbe interaction, bioactive metabolites, and functional effects. These approaches will be essential to determine whether Senegalimassilia plays a causal role and holds therapeutic relevance.

This observational study had a small sample size and lacked a sham treatment control group. Additionally, short-duration HRV recordings may have limited the interpretation of ANS balance, and the analysis of only selected biomarkers may not comprehensively reflect the gut-brain-microbiota axis. Therefore, further studies with larger samples and more integrative multi-omics approaches are needed to validate and extend these findings.

CONCLUSION

This study demonstrates that EA is a promising therapeutic intervention for patients with IBS-C, providing sustained improvements in clinical symptoms, quality of life, and gut microbiota balance. EA treatment was associated with modulation of neurotransmitter-related metabolic pathways. Notably, changes in specific gut microbiota-particularly an increased abundance of Senegalimassilia-were significantly associated with clinical improvement, suggesting its potential as a predictive microbial biomarker for EA responsiveness. These findings underscore the importance of gut microbiota in mediating therapeutic effects of EA and support further exploration of microbiota-targeted strategies in IBS-C management. Large-scale, controlled studies are warranted to validate these results and clarify the underlying mechanisms.

ACKNOWLEDGEMENTS

The authors would like to thank Dr. Sirawit Sriwichaiin and Dr. Chotrawee Piriyakunthorn for their assistance with data analysis in the present study.