Sage Journals: Discover world-class research

Abstract

Background:

Patients with ulcerative colitis (UC) usually experience anxiety symptoms that seriously affect their quality of life, treatment, and prognosis. Dysbiosis of the gut microbiota plays an important role in UC and mental illness. However, little is known about the role of the gut microbiota in UC patients with anxiety.

Objectives:

To identify the gut-microbiome and fecal metabolome profiles uniquely associated with comorbid anxiety in UC patients and to explore potential biomarkers for diagnosis.

Design:

A cross-sectional, two-group comparative study.

Methods:

To study the underlying association between them, we recruited 126 UC patients in this study, including 78 with anxiety and 48 without anxiety. A total of 102 fecal samples were collected for metagenomic sequencing and metabolome sequencing. Microbial diversity, differential gut microbiota, functional pathways, and metabolites were analyzed. Multivariable logistic regression was used to identify independent risk factors associated with anxiety in UC patients, while Spearman correlation was employed to explore microbe-metabolite interactions and the performance of potential biomarkers.

Results:

We found that disease severity, steroid usage, and abdominal pain may promote the occurrence of anxiety. Compared to UC patients without anxiety, UC patients with anxiety had low fecal microbial community diversity, with an increase in the species Haemophilus sp. HMSC71H05 and Corynebacterium durum, and a decrease in the species Roseburia intestinalis (RI), Bifidobacterium longum (BL), and Enterococcus hirae. The metabolic pathways driven by the gut microbiota were disrupted. Moreover, the levels of most metabolites (such as L-kynurenine) were increased in the feces, while the levels of a few metabolites decreased, including indole-2-carboxylic acid, N-demethylmirtazapine, and tauroursodeoxycholic acid.

Conclusion:

Our research further revealed that these gut microbiota and metabolites are highly correlated. This study provides a new perspective for understanding the occurrence and development of anxiety in UC patients, suggesting that RI and BL may serve as potential candidate biomarkers to diagnose UC patients with anxiety.

Keywords

abdominal pain anxiety gut microbiota ulcerative colitis

Introduction

Ulcerative colitis (UC) is a chronic immune-mediated intestinal inflammation¹ that often invades the rectum and colon, and is characterized by diarrhea, a bloody, purulent stool, a sense of urgency during defecation, and tenesmus.²

Multiple studies have shown that UC patients are prone to experiencing mental symptoms, such as anxiety, which seriously affects patients’ daily lives.^3,4 A recent study reported that the prevalence of anxiety was 34.2% in patients with UC,⁵ accounting for a significant proportion of UC patients. However, the anxiety symptoms of UC patients do not attract enough attention in the clinic, and about two-thirds of UC patients with anxiety have not been diagnosed and treated.³ In addition, anxiety can, in turn, exacerbate UC, affecting the treatment and prognosis of UC patients.⁶ Thus, understanding the relationship between UC and anxiety, and exploring the critical factors that cause anxiety in UC patients, is of great guiding significance for the treatment of UC combined with anxiety.

Several recent studies have shown that the gut microbiota is associated with many mental illnesses, including autism, anxiety, depression, and schizophrenia.^7–9 The importance of the gut–brain axis in mental illness has been recognized,^10–12 and the microbiota, as a key regulator of the gut-brain axis,^7,13,14 can participate in crosstalk with the brain via metabolic and neuroimmune pathways.^15,16 For instance, the bacterial metabolite 4-ethylphenyl sulfate has been reported to induce anxiety-like behavior in mice.¹⁴ Bifidobacterium breve and Lactobacillus plantarum relieve anxiety or depression symptoms by modulating the gut microbiota and its metabolites.^17–19 Bacteroides vulgatus was reported to alleviate depression-like behavior in mice with colitis via the gut–brain axis mediated by the metabolite p-hydroxyphenylacetic acid.²⁰ These studies provide a research basis and direction for exploring the gut microbiome-mediated gut–brain axis regulation of anxiety. Furthermore, a previous study has demonstrated that the richness and diversity of the fecal microbiota in UC patients with anxiety are reduced and that intestinal microbiota disorders, including increased Lactobacillales, Sellimonas, and Streptococcus, and depletion of Lachnospira.⁴ Most metabolites are altered, and the administration of downregulated metabolites, such as L-pipecolic acid, reduces depression-like behavior in mice.⁴ However, at present, the use of metagenomics to deeply analyze the species and functional levels, combined with broad-target metabolomics to identify potential diagnostic biomarkers related to anxiety status, is not sufficient.

Therefore, through this observational research, we summarized the clinical characteristics of UC patients with anxiety and further utilized a multi-omics approach involving the intestinal microbiota and metabolome to characterize UC patients with anxiety. This study aimed to explore and identify biomarkers of anxiety in UC patients, provide guidance strategies for the treatment of UC combined with anxiety, and improve patients’ quality of life.

Materials and methods

Study subjects

In this study, 126 UC patients who visited the Department of Gastroenterology, Army Medical Center, Army Medical University from September 2021 to April 2022 were consecutively recruited. All patients were diagnosed with UC and satisfied the following criteria: (i) met the JSGE Evidence-based Clinical Practice Guidelines for Inflammatory Bowel Disease²¹; (ii) were gender unlimited, aged between 18 and 70 years old; (iii) had a disease duration exceeding 6 months; (iv) had no previous history of malignancy or mental illness, with normal communication skills; (v) did not have hypertension, diabetes, heart disease, stroke, or serious chronic diseases (infectious, hereditary, metabolic, or endocrine diseases); and (vi) non-pregnant and lactating patients. Disease activity was assessed using the modified Mayo (mMayo) score^22,23 for UC at the time of the interview. Active disease was attested when the mMayo score exceeded 2 in UC patients. All participants provided written informed consent before enrollment.

The reporting of this study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement.^24,25

Sample processing and storage

The included subjects were instructed to provide first stool samples of the day for metagenomic and metabolomics sequencing. All participants were required to stop using antibiotics, probiotics, prebiotics, and other microbiota-related supplements at least 1 month before fecal sampling. Fecal samples were self-collected in a DNA preservation tube, transferred to the laboratory on dry ice within 2 h, and stored at −80°C until DNA was extracted. Finally, 102 fecal samples were collected.

Outcomes

The primary outcome was screening for fecal biomarkers of the gut microbiota. The secondary outcome was alterations in gut microbiota composition.

Hamilton Anxiety Scale

The Hamilton Anxiety Scale (HAM-A),²⁶ developed in 1959, is a clinician-rated instrument for assessing the severity of anxiety symptoms in patients with neurosis and other disorders.^27,28 It contains 14 items, divided into mental anxiety and somatic anxiety. Each item is scored on a 5-point scale ranging from 0 to 4 points.²⁹ The higher the score, the more severe the symptoms. According to the data provided by the Scale Collaboration Group of China, the HAM-A total score of less than 7 is commonly considered as no anxiety. Therefore, based on the cutoff scores of HAM-A without anxiety, we classified participants with HAM-A scores ⩾7 into the UC with anxiety (UC-A) group. It is worth noting that this study was conducted exclusively in a Chinese population, and methodological thresholds may vary across populations.

Metagenomic sequencing and data analysis

The libraries were prepared using the MGIEasy Fast FS DNA Library Prep Set (MGI, Wuhan, China) according to the manufacturer’s protocols, and PE150 sequencing was performed using the MGI DEBSEQ2000 platform. KneadData software (v0.7.4) was run by calling Bowtie2 v2.4.2 (Baltimore, MD, USA) to process the raw sequencing data and to remove host genome contamination from the samples. The HMP Unified Metabolic Analysis Network 3.0 (HUMAnN 3.0, Boston, MA, USA) was used to obtain MetaCyc pathways, and MetaPhlAn3 was used for taxonomic annotations.

Alpha diversities based on the taxonomy profile were calculated using the R package vegan v2.6.4. Beta diversities based on the Bray-Curtis distance matrix were calculated and visualized using R built-in functions and the R package ggplot2. The Wilcoxon rank sum test was used to identify differences in species abundance and the MetaCyc pathway. Spearman’s correlations and FDR were evaluated using the psych R package.

Metabolomic analysis

The feces of 102 UC patients saved for metabolite analyses were delivered to METWARE BIO (Wuhan, China) for “TM” widely targeted metabolomics analysis. All chromatographic separations were performed using an ultra-performance liquid chromatography system (ExionLC AD, AB SCIEX, Framingham, MA, USA). A high-resolution tandem mass spectrometer TripleTOF 6600 (AB SCIEX, Framingham, MA, USA) was used to detect eluted metabolites in the column. Partial least squares discriminant analysis (PLS-DA) was performed using the R package ropls to evaluate the difference in metabolic profiles between the UC-A and UC-NA groups. The significant metabolites with variable importance in projection ⩾1, p value (Wilcoxon rank sum test) <0.05, and with >2-fold change.

Statistical analysis

Analyses of clinical features were performed with IBM SPSS Statistics for Windows, version 26.0. The Shapiro-Wilk test was used to evaluate whether the continuous data conformed to a normal distribution. Unpaired Student’s t test or Mann-Whitney U test was utilized to estimate differences between quantitative data, as appropriate. In this study, the two-sided Wilcoxon rank-sum test was used to compare the continuous variables. The categorical variables were compared via Pearson’s chi-square test or Fisher’s exact test. A multivariate binary logistic regression model was used to examine potential factors affecting the occurrence of UC-A. For all analyses, differences were considered significant at p < 0.05.

Results

General information about patients

As shown in Figure 1, 178 patients were screened, and 126 patients were enrolled in this observational study at the Department of Gastroenterology at the Army Medical Center of the Army Medical University from September 2021 to April 2022. A total of 126 patients (78 in the UC-A group and 48 in the UC without anxiety (UC-NA) group) completed the questionnaires, and 102 patients provided fecal samples (Figure 1). The information on demographics and clinical manifestations of these patients is presented in Table 1. Among them, 58.7% (n = 74) were males, 41.3% (n = 52) were females, and 34.1% had a history of smoking. Moreover, 94.44% of UC patients were in the disease-active phase, of which 33.3%, 54.0%, and 7.1% were in the mild, moderate, and severe active stages, respectively. The course of disease ranged from 6 months to 21 years, with a median duration of 10.75 years and a median onset age of 47 years (Table 1).

Figure 1.

Flow chart of patients included and excluded in this study.

Table 1.

Summary of demographic characteristics of all cohorts.

Characteristic	UC	UC-A	UC-NA	p Value
Number	126	78	48	–
Female^a	52(41.27)	27(34.62)	25(52.08)	0.053^b
Age^c	47(18, 69)	48.5(18, 69)	42.5(18, 66)	0.072^d
BMI^c	21.48(15.35, 34.6)	21.74(16.23, 34.6)	20.98(15.35, 28.3)	0.077^d
Smokers^a	43(34.1)	30(38.5)	13(27.1)	0.191^b
Course^c	3(0.08, 21)	3(0.08, 21)	2(0.08, 20)	0.092^d
mMayo score^c	6(5, 9)	7(5, 9)	5.5(4, 8)	0.01^d
Bristol score^c	6(4, 6.75)	6(4.25, 7)	5(4, 6)	<0.013^d
Frequency of defecation^e	4.79 ± 4.84	5.67 ± 5.78	3.35 ± 1.96	<0.01^f
GSRS score^e	30.98 ± 9.70	33.79 ± 9.82	26.40 ± 7.48	0.000^d
IBDQ score^e	162.47 ± 38.50	150.74 ± 37.51	181.52 ± 32.70	0.000^d
HAM-A^e	8.79 ± 6.23	12.36 ± 5.14	2.98 ± 2.13	<0.0005^f
No anxiety, 0–6	48(38.10)	0(0)	48(100)
There may be anxiety, 7–13	56(44.44)	56(71.79)	0(0)
There must be anxiety, 14–20	14(11.11)	14(17.95)	0(0)
Obvious anxiety, 21–28	8(6.35)	8(10.26)	0(0)
Severe, ⩾29	0(0)	0(0)	0(0)
Gastroenterology^a
Abdominal pain	86 (68.25)	64 (82.05)	22 (45.83)	0.000^b
Diarrhea	76 (60.32)	52 (66.67)	24 (50.00)	0.063^b
Bloating	37 (29.37)	23 (29.49)	14 (29.17)	0.969^b
Mucus stool	72 (57.14)	43 (55.13)	29 (60.42)	0.560^b
Hematochezia	92 (73.02)	59 (75.64)	33 (68.75)	0.397^b
Extraintestinal manifestations	18 (14.29)	11 (14.10)	7 (14.58)	0.940^b
Other symptoms^g	38 (30.16)	25 (32.05)	13 (27.08)	0.555^b
Drugs^a
5-ASA^h	116 (92.06)	72 (92.31)	44 (91.67)	0.897^b
Steroid	23 (18.25)	20 (25.64)	3 (6.25)	0.012^b
Immunosuppression	3 (2.38)	3 (3.85)	0 (0)	0.439^b
Biologic	17 (14.49)	8 (10.26)	9 (18.75)	0.175^b
Enema	31 (24.60)	22 (28.21)	9 (18.75)	0.231^b
Other drugsⁱ	32 (25.40)	23 (29.49)	9 (18.75)	0.179^b

n (%).

Statistics were based on the Chi-squared test.

Median (interquartile range).

Statistics were based on the Mann-Whitney U test.

Mean ± standard deviation.

Statistics were based on a summary independent-samples T test.

Tenesmus, anal prolapse, and distension, unformed stool.

5-Aminosalicylic acid drugs.

Including traditional Chinese medicine, macrobiotics, etc. The GSRS scale has two missing values, and the mean substitution method is followed by the Mann-Whitney U test.

GSRS, Gastrointestinal Symptoms Rating Scales; IBDQ, Inflammatory Bowel Disease Questionnaire; HAM-A, the Hamilton Anxiety Scale; mMayo, modified Mayo; UC, ulcerative colitis; UC-A, UC patients with anxiety; UC-NA, UC patients without anxiety.

Clinical characteristics of UC patients with anxiety

Subsequently, we observed the clinical manifestations of UC-A and UC-NA groups. We found that patients with gastrointestinal symptoms have a higher rate of anxiety, with abdominal pain being the most significant symptom (Table 1). In the UC-A group, 82.05% (n = 64) of patients had abdominal pain symptoms, 75.64% (n = 59) had hematochezia, and 66.67% (n = 52) had diarrhea, while UC-NA group, they were 45.83% (n = 22), 68.75% (n = 33), and 50% (n = 24), respectively. The incidence of these symptoms was higher in the UC-A group than in the UC-NA group. However, the effect of extraintestinal manifestations on anxiety was not obvious, with 14.1% (n = 11) and 14.58% (n = 7) in the UC-A and UC-NA groups, respectively.

In addition, the mMayo score of the UC-A group was significantly higher than that of the UC-NA group (p = 0.01), indicating that the UC-A group had higher disease activity (Table 1). Patients in the remission phase of UC were more likely to have no symptoms of anxiety, with 10.42% (n = 5) in remission in the UC-NA group and only 2.56% (n = 2) in the UC-A group. Although there were fewer patients in remission in our study, this trend can still be demonstrated.

Subsequently, to further explore the potential influencing factors resulting in anxiety in UC patients, we employed multivariate logistic regression analysis. mMayo score (odds ratio (OR) 1.199, p = 0.025), abdominal pain (OR 5.466, p < 0.001), steroid usage (OR 5.742, p = 0.009), and the Bristol score (OR 1.451, p = 0.013) were still found to be independent predictors of anxiety (Table 2), after adjusting for age, BMI, and smoking status. However, there was no significant correlation between gender and anxiety.

Table 2.

Factors influencing UC in anxiety.

Factor	Odds ratio	95% Confidence interval	p Value
Gender	2.069	0.869–5.074	0.1
Abdominal pain	5.466	2.372–13.234	<0.001
Diarrhea	1.946	0.919–4.162	0.083
Mayo score	1.199	1.027–1.412	0.025
Bristol score	1.439	1.085–1.944	0.014
Frequency of defecation	1.239	1.062–1.489	0.015
GSRS score	1.118	1.064–1.183	<0.001
IBDQ score	0.970	0.956–0.983	<0.001
Steroid	5.742	1.774–26.01	0.0085

GSRS, Gastrointestinal Symptoms Rating Scales; UC, ulcerative colitis.

Fecal microbiota features in patients of UC-A

To investigate the characteristics of the intestinal microflora in patients of UC-A, metagenomic sequencing was applied to the fecal samples of 63 UC-A and 39 UC-NA. Compared to patients of UC-NA, the patients of UC-A had lower fecal microbial diversity, as represented by the Shannon index (Wilcoxon rank-sum test; Figure 2(a)). However, both UC-A and UC-NA were clustered together according to principal coordinate analysis (PCoA) based on the Bray–Curtis distance matrix, indicating that fecal data may have no power to distinguish UC-A (permutational multivariate analysis of variance (PERMANOVA), p = 0.332; Figure 2(b)). Furthermore, we found that the UC-A group exhibited an increase in Firmicutes and Proteobacteria abundance and a decrease in Bacteroidetes abundance, further indicating changes in the microbial community structure of the UC-A group (Figure 2(c)).

Figure 2.

The variations in the fecal microbiome of UC-A and UC-NA. (a) Alpha diversity index of species for UC-A and UC-NA groups. (b) PCoA of the microbiota based on the Euclidean distance metrics for UC-A and UC-NA. PERMANOVA, p = 0.332. (c) The relative abundance of the dominant phylum in the fecal microbiome of UC-A and UC-NA groups. (d) The bar plot showed a relative abundance of the 13 differential expression species across two groups. (e) Correlations between 13 species and clinical characteristics. (f) The volcano plot of differently enriched MetaCyc pathways in UC-A and UC-NA groups. (g) Correlations between 7 MetaCyc pathways and clinical characteristics.

Next, we compared the differences in the bacterial profiles between the UC-A and UC-NA groups. A total of 13 species showed differential expression between the two groups (Wilcoxon rank-sum test, p < 0.05; Figure 2(d)). The abundances of the species Roseburia intestinalis (RI), Bifidobacterium longum (BL), and Enterococcus hirae decreased significantly in UC-A, but the species Actinomyces sp. oral taxon 414, Haemophilus sp. HMSC71H05 and Corynebacterium durum were increased significantly. Then, we investigated whether these changes correlated with the clinical characteristics to assess the potential role of the gut microbiota in UC-A. We found that the species RI, Anaerotignum lactatifermentans, and Aggregatibacter segnis were negatively correlated with HAM-A; however, Actinomyces sp. oral taxon 414 and C. durum were positively correlated with HAM-A. The results suggested that the decrease in the species RI and the increase in C. durum may be associated with anxiety in UC-A patients (Spearman correlation analysis, *p < 0.05, **p < 0.01; Figure 2(e)).

To explore the functional implications of microbial shifts that drive anxiety in patients with UC, the MetaCyc pathways were studied. Compared with those in UC-NA, only seven MetaCyc pathways were downregulated in the UC-A group (Wilcoxon rank-sum test, p < 0.05; Figure 2(f)), which were primarily enriched in amino acid degradation, fatty acid and lipid biosynthesis, and the generation of precursor metabolites and energy. In addition, all of these pathways tended to be negatively correlated with HAM-A (Figure 2(g)). These results indicated that UC-A may be caused by dysregulation of these pathways.

The gut microbiota characteristics in patients with active UC-A

Next, we further explored the influence of gut microbiota on anxiety symptoms at different stages of active UC. We found that the fecal microbial diversity of UC-A patients in mild (UC-A, n = 18; UC-NA, n = 14) and severe (UC-A, n = 6; UC-NA, n = 2) stages was lower than that of UC-NA patients (Figure 3(a)). However, in mild and severe stages, PCoA showed that there was no significant separation between patients with UC-A and UC-NA (PERMANOVA, p = 0.365, p = 0.437; Figure 3(b) and (d)). Similarly, the alpha-diversity and beta-diversity of UC-A (n = 37) and UC-NA (n = 18) patients did not show differentiation in the moderate stage (PERMANOVA, p = 0.12; Figure 3(a) and (c)). Subsequently, we observed that the relative abundance of Bacteroidetes, Firmicutes, Proteobacteria, and Actinobacteria changed in two groups at different stages of active UC, and also showed significant differences between different stages (Figure 3(e)).

Figure 3.

The variations in the fecal microbiome of active UC-A and UC-NA at different stages. (a) Alpha diversity index of species for UC-A and UC-NA groups at mild, moderate, and severe stages. (b–d) PCoA of the microbiota based on the Euclidean distance metrics for UC-A and UC-NA groups at mild, moderate, and severe stages, respectively. PERMANOVA, p = 0.365, p = 0.437, p = 0.12. (e) The relative abundance of the dominant phylum in the fecal microbiome of UC-A and UC-NA groups at different phases. (f) The relative abundances of the species with statistical differences between UC-A and UC-NA groups based on the Wilcoxon rank-sum test (p < 0.05) were shown in a heatmap. (g) Spearman’s correlations between anxiety levels and species.

Similarly, we also compared the differences in bacterial profiles between active UC-A and UC-NA at different stages. It is worth noting that the abundance of species RI was still significantly reduced in UC-A, BL, Clostridium paraputrificum, Escherichia marmotae, and C. paraputrificum were significantly reduced at the moderate stage of UC-A (Figure 3(f)). The abundances of species Atopobium rimae and Clostridium citroniae were significantly increased in UC-A (Figure 3(f)). The correlation with HAM-A scores indicated that some species were associated with anxiety levels, such as A. rimae was positively correlated with anxiety levels, while RI, BL, and C. paraputrificum were negatively correlated with anxiety levels (Figure 3(g)).

These results suggest that there is indeed an imbalance in gut microbiota between patients of active UC-A and UC-NA at different phases, and these gut microbiota are associated with anxiety. The occurrence of anxiety symptoms in UC patients may be due to the imbalance of these gut microbiota.

Metabolomic alterations in patients with UC and anxiety

To clarify the metabolic changes in UC-A or UC-NA, we analyzed their fecal metabolome data and examined the relationships between the microbiota and metabolites. According to the PCoA and PLS-DA, the two groups of fecal samples were largely separated, indicating that UC-A and UC-NA had dissimilar metabolic patterns (PERMANOVA, p = 0.004; Figure 4(a) and (b), Supplemental Figure 1(A)–(F)). In total, 11 KEGG pathways were significantly dysregulated when comparing the two groups (Figure 4(c)). The top three enriched pathways were involved in glycerophospholipid metabolism, oxidative phosphorylation, and choline metabolism in cancer.

Figure 4.

Metabolome alterations across UC-A and UC-NA groups. (a) The PCoA of the metabolites based on the Euclidean distance metrics for UC-A (n = 63) and UC-NA (n = 39) groups. PERMANOVA, p = 0.004. (b) The clustering analyses of PLS-DA for UC-A and UC-NA groups. (c) 11 KEGG pathways were significantly different in UC-A and UC-NA. (d) The heatmap showed a relative abundance of the top 45 differential expression metabolites across the two groups. Metabolites >2-fold changes, VIP ⩾1, p < 0.05 (Wilcoxon rank-sum test). (e) The network revealed representatively significant and suggestive associations among differentially abundant bacterial species and metabolites in UC-A and UC-NA. The edge width corresponds to the size of the Spearman correlation coefficients. The edge color indicates the signs of correlation coefficients; red means positive correlation, while blue means negative correlation. Only connections with Spearman’s correlation coefficients |R| > 0.3 and p < 0.05 are shown.

There were 169 metabolites with larger than twofold changes in the UC-A group, including 118 increased and 51 depleted (Figure 4(d)). The levels of some endogenous eicosanoids, such as 5-hydroxyeicosatetraenoic acid (5-HETE), 8-HETE, and 9-HETE, were increased in the UC-A group. The levels of 5-oxoETE, leukotriene B4 (LTB4), and palmitoyl 3-carbocyclic phosphatidic acid, which are organic compounds known to be long-chain fatty acids, significantly differed between UC-A and UC-NA. The level of tryptophan metabolites, such as indole-2-carboxylic acid, decreased in UC-A, while the level of L-kynurenine increased in UC-A. Moreover, some intermediate metabolites of amino acids and lysophospholipids were also significantly dysregulated in the UC-A group.

The gut microbiota participates in host regulation through the regulation of metabolites and the fecal metabolome. Therefore, we further utilized Spearman correlation analysis to explore the correlation between altered gut microbiota and fecal metabolites (Figure 4(e)). The correlations with metabolites were mainly concentrated in BL and RI. We found that BL was significantly positively correlated with 1-methylhydantoin, indole-2-carboxylic acid, and tauroursodeoxycholic acid but negatively correlated with 8-HETE, L-kynurenine, and glycerophospho-N-oleoyl ethanolamine. The RI was significantly positively correlated with 4-hydroxydebrisoquine, ω-3 arachidonic acid, and N-desmethylmirtazapine. However, most lysophospholipids and intermediate metabolites were negatively correlated with bacterial species. In brief, the analysis further suggested that gut microbes might be involved in neural metabolism to influence patient anxiety.

Discussion

It has been demonstrated that the prevalence of anxiety in UC patients is greater than that in healthy individuals.^30–32 Furthermore, previous studies have shown that the disease activity of UC is associated with anxiety.^5,33,34 This study observed that disease activity and steroid use are significantly positively correlated with anxiety levels. Our results are consistent with previous studies. It has been reported that the disease severity was associated with symptoms of moderate to severe stress, severe depression, and anxiety.³⁵ Meanwhile, with the increase in UC disease activity, the abundance of anti-inflammatory bacteria such as Faecalibacterium prausnitzii and Lactobacilli declines, along with reduced levels of short-chain fatty acids and secondary bile acids, while the levels of pro-inflammatory bacteria and pro-inflammatory metabolites rise.^36,37 Corticosteroid use is associated with reduced quality of life in IBD, and increases the risk of psychological illness in IBD patients.^38–40 Steroids and the gut microbiota engage in bidirectional crosstalk; exogenous or endogenous steroids rapidly restructure the microbial community and modulate its metabolic function, thereby influencing host mental health via the gut–brain axis.^41,42 In addition, factors associated with UC-related anxiety have been found to include abdominal pain, defecation frequency, and Bristol score.

In recent years, a considerable amount of research has been conducted using a multi-omics approach to study a disease or the relationship between two concomitant diseases.^43–45 Herein, this study aimed to explore potential factors that affect the development of anxiety in UC patients and provide guidance strategies for their treatment by jointly analyzing the metagenomics and metabolomics data of 106 patients with and without anxiety and the correlation between the gut microbiota profiles and clinical features.

The gut–brain axis has been proven to be a key pathway affecting mental illness,^10,46 and disturbance of the intestinal flora plays a regulatory role in the occurrence and development of anxiety.^8,47 Moreover, bacteria-derived metabolites play an important role in mediating the gut–brain axis, affecting anxiety/depression-like behavior.^20,48,49 In this study, we found that the diversity of the gut microbiome in UC-A patients was significantly reduced, indicating a potential association between anxiety and microbiome dysregulation. Although there was no difference in alpha diversity between the UC-A and UC-NA groups at different stages of UC activity, which may be due to the small sample size, the microbial community structure still underwent significant changes, such as alterations in the relative abundance of Bacteroidetes and Actinobacteria, and these changes also varied at different stages. Several bacterial species that have been reported to be beneficial for human health, such as RI and BL, were significantly depleted in the UC-A group. RI has been shown to alleviate dextran sodium sulfate (DSS)-induced colitis and depression-like behavior in rats.⁵⁰ In addition, the abundance of RI is reduced in individuals with major depressive disorder,¹⁵ and RI has a beneficial effect on depression via the gut–brain axis.⁵¹ Nevertheless, few researchers have explored how the RI regulates anxiety behavior through the gut–brain axis. We found that the RI was significantly negatively correlated with the HAM-A. At the same time, N-demethylmirtazapine, ω-3 arachidonic acid, and LTB4 were disrupted in UC-A and exhibited the same or opposite regulatory effects as RI. Interestingly, the ω-3 arachidonic acid^52,53 is associated with neuroinflammation, and N-desmethylmirtazapine⁵⁴ is a benzazepine metabolite resulting from the demethylation of the antidepressant mirtazapine. LTB4 is an oxygenated metabolite of arachidonic acid that intensifies neuroinflammation.⁵³

Furthermore, the BL exhibited a significant negative correlation with the Gastrointestinal Symptoms Rating Scales (GSRS) as well as HAM-A, but the difference was not statistically significant, which may be related to the small sample size. A previous placebo-controlled trial revealed that probiotic BL reduced depression in irritable bowel syndrome (IBS) patients, unfortunately without alleviating anxiety scores.⁵⁵ Subsequently, BL NK46 was reported to significantly mitigate immobilization stress-induced anxiety-like behaviors and colitis by suppressing gut dysbiosis.⁵⁶ Moreover, BL has been shown to produce indoles,⁵⁷ some of which can act as arylhydrocarbon receptor ligands,⁵⁸ such as indole-3-lactic acid, which can reduce neuroinflammation and achieve antidepressant effects.⁵⁷ However, there is no report on whether indole-2-carboxylic acid can improve mental illness. L-kynurenine and tauroursodeoxycholic acid exhibit an antagonistic relationship, both of which are significantly correlated with BL. Increasing evidence suggests that an increase in kynurenine is closely related to emotional disorders.^51,59 Similarly, L-kynurenine can induce depression-like behavior in mice.⁶⁰ L-kynurenine is significantly increased in UC-A patients, which is consistent with previous studies. Tauroursodeoxycholic acid has been reported to be a potential antidepressant that may achieve antidepressant effects by inhibiting neuroinflammatory responses in the brain.^61–63 Meanwhile, BL was significantly positively correlated with 1-methylhydantoin but negatively correlated with 8-HETE. 1-Methylhydantoin⁶⁴ has been reported to inhibit inflammation and may be a metabolite affecting intelligence, while 8-HETE^65,66 has proinflammatory effects. We hypothesized that BL might alleviate anxiety in UC patients by decreasing the level of L-kynurenine, 8-HETE, and increasing the levels of indole-2-carboxylic and tauroursodeoxycholic acid.

In this study, the microbe-driven L-tyrosine degradation pathway, which was significantly inactivated in UC-A and strongly negatively correlated with HAM-A, also attracted our attention. Unfortunately, L-tyrosine levels did not change significantly between the two groups. Therefore, we hypothesize that anxiety may be related to the abnormal inactivation of the L-tyrosine degradation pathway. However, further studies are needed to verify this potential link. Moreover, A. lactatifermentans showed a strong negative correlation with HAM-A, abdominal pain, and GSRS, but a relationship between A. lactatifermentans and UC or anxiety behavior has not been reported. However, Actinomyces sp. oral tax 414 and C. durum were positively correlated with HAM-A, and these results suggest that dysbiosis of specific bacterial species may play a beneficial or detrimental role in the anxiety of UC patients. RI and BL may be key bacterial species for improving anxiety, but their specific mechanisms need further investigation.

Overall, this study provides new insights into the understanding of anxiety symptoms in UC patients. The result showed that the degree of disease progression, steroid usage, and presence of abdominal pain may promote anxiety in UC patients. In addition, dysregulation of the gut microbiome and abnormal metabolites may also play an important role in the development of anxiety in UC patients. The RI and BL bacterial species may serve as potential candidate biomarkers for the diagnosis of anxiety symptoms in UC patients. However, there were also shortcomings in this study. First, the HAM-A is a semi-structured, clinician-rated scale that is vulnerable to evaluator bias, and its somatic items overlap with UC symptoms, potentially yielding false positives. Second, without considering the confounding factors such as steroid use, diet, and lifestyle, and excluding irritable bowel syndrome using the Rome IV criteria, one may underestimate or overestimate the true HAMA–microbiota relationship. Third, the follow-up time is relatively short, and there was no long-term longitudinal monitoring of anxiety symptoms in UC patients, so it is impossible to evaluate the dynamic changes of anxiety with the course of the disease and its changes with gut microbiota. Fourth, the modest sample size may be underrepresented. Finally, observation of associations between HAM-A scores and specific microbial signatures has not been complemented by animal experiments, leaving the precise gut–brain axis pathways in UC-related anxiety undefined. Therefore, in the future, it is necessary to independently verify the screened biomarkers in a prospective, larger-scale multicenter cohort and systematically evaluate their diagnostic and predictive performance in combination with disease activity, diet, and other factors. At the same time, long-term longitudinal follow-up is also needed to map dynamic changes during remission and anxiety symptoms, clarifying the role of the microbiota–gut–brain axis across disease stages and informing precision interventions for UC patients with comorbid anxiety.

Conclusion

The HAM-A was used to investigate the anxiety status of UC patients and analyze its related influencing factors. At the same time, a preliminary exploration of the gut–brain axis relationship between anxiety and UC was performed. In this study, we provided a new perspective on the occurrence and development of anxiety in UC patients and pointed out that RI and BL may be potential candidate biomarkers for the diagnosis of anxiety symptoms in UC patients.

Supplemental Material

sj-docx-1-tag-10.1177_17562848251393419 – Supplemental material for Gut microbiota and metabolic signatures of anxiety in ulcerative colitis: a cross-sectional study

Supplemental material, sj-docx-1-tag-10.1177_17562848251393419 for Gut microbiota and metabolic signatures of anxiety in ulcerative colitis: a cross-sectional study by Yi Ping, Xuefei Zhao, Linling Lv, Wei Meng, Yue Meng, Guangcong Ruan, Yi Cheng, Zhifeng Xiao, Yuting Tian, Minjia Chen, Lu Chen, Ailin Yi, Zongyuan Tang, Ning Li, Dongfeng Chen and Yanling Wei in Therapeutic Advances in Gastroenterology

Supplemental Material

sj-docx-2-tag-10.1177_17562848251393419 – Supplemental material for Gut microbiota and metabolic signatures of anxiety in ulcerative colitis: a cross-sectional study

Supplemental material, sj-docx-2-tag-10.1177_17562848251393419 for Gut microbiota and metabolic signatures of anxiety in ulcerative colitis: a cross-sectional study by Yi Ping, Xuefei Zhao, Linling Lv, Wei Meng, Yue Meng, Guangcong Ruan, Yi Cheng, Zhifeng Xiao, Yuting Tian, Minjia Chen, Lu Chen, Ailin Yi, Zongyuan Tang, Ning Li, Dongfeng Chen and Yanling Wei in Therapeutic Advances in Gastroenterology

Supplemental Material

sj-docx-3-tag-10.1177_17562848251393419 – Supplemental material for Gut microbiota and metabolic signatures of anxiety in ulcerative colitis: a cross-sectional study

Supplemental material, sj-docx-3-tag-10.1177_17562848251393419 for Gut microbiota and metabolic signatures of anxiety in ulcerative colitis: a cross-sectional study by Yi Ping, Xuefei Zhao, Linling Lv, Wei Meng, Yue Meng, Guangcong Ruan, Yi Cheng, Zhifeng Xiao, Yuting Tian, Minjia Chen, Lu Chen, Ailin Yi, Zongyuan Tang, Ning Li, Dongfeng Chen and Yanling Wei in Therapeutic Advances in Gastroenterology

Footnotes

Appendix

Acknowledgements

The authors are grateful and appreciate the participation of our patients in this study. We would also like to thank the state authorities of the Chongqing Science and Health Joint Project.

Declarations

ORCID iD

Yanling Wei

Supplemental material

Supplemental material for this article is available online.

References

Agrawal

Spencer

Colombel

, et al. Approach to the management of recently diagnosed inflammatory bowel disease patients: a user’s guide for adult and pediatric gastroenterologists. Gastroenterology 2021; 161(1): 47–65.

Le Berre

Honap

Peyrin-Biroulet

. Ulcerative colitis. Lancet (London, England) 2023; 402(10401): 571–584.

Lewis

Marrie

Bernstein

, et al. The prevalence and risk factors of undiagnosed depression and anxiety disorders among patients with inflammatory bowel disease. Inflamm Bowel Dis 2019; 25(10): 1674–1680.

Yuan

Chen

Duan

, et al. Depression and anxiety in patients with active ulcerative colitis: crosstalk of gut microbiota, metabolomics and proteomics. Gut Microbes 2021; 13(1): 1987779.

Barberio

Zamani

Black

, et al. Prevalence of symptoms of anxiety and depression in patients with inflammatory bowel disease: a systematic review and meta-analysis. Lancet Gastroenterol Hepatol 2021; 6(5): 359–370.

Gracie

Guthrie

Hamlin

, et al. Bi-directionality of brain-gut interactions in patients with inflammatory bowel disease. Gastroenterology 2018; 154(6): 1635–1646.e3.

Cryan

O’Riordan

Cowan

CSM

, et al. The microbiota-gut-brain axis. Physiol Rev 2019; 99(4): 1877–2013.

Simpson

Diaz-Arteche

Eliby

, et al. The gut microbiota in anxiety and depression—a systematic review. Clin Psychol Rev 2021; 83: 101943.

Nikolova

Smith

MRB

Hall

, et al. Perturbations in gut microbiota composition in psychiatric disorders: a review and meta-analysis. JAMA Psychiatry 2021; 78(12): 1343–1354.

10.

Góralczyk-Bińkowska

Szmajda-Krygier

Kozłowska

The microbiota-gut-brain axis in psychiatric disorders. Int J Mol Sci 2022; 23(19): 11245.

11.

Socała

Doboszewska

Szopa

, et al. The role of microbiota-gut-brain axis in neuropsychiatric and neurological disorders. Pharmacol Res 2021; 172: 105840.

12.

Westfall

Caracci

Estill

, et al. Chronic stress-induced depression and anxiety priming modulated by gut-brain-axis immunity. Front Immunol 2021; 12: 670500.

13.

Gareau

Wine

Rodrigues

, et al. Bacterial infection causes stress-induced memory dysfunction in mice. Gut 2011; 60(3): 307–317.

14.

Needham

Funabashi

Adame

, et al. A gut-derived metabolite alters brain activity and anxiety behaviour in mice. Nature 2022; 602(7898): 647–653.

15.

Kovtun

Averina

Angelova

, et al. Alterations of the composition and neurometabolic profile of human gut microbiota in major depressive disorder. Biomedicines 2022; 10(9): 2162.

16.

Mayneris-Perxachs

Castells-Nobau

Arnoriaga-Rodríguez

, et al. Microbiota alterations in proline metabolism impact depression. Cell Metab 2022; 34(5): 681–701.e10.

17.

Tian

Chen

Zhu

, et al. Bifidobacterium breve CCFM1025 attenuates major depression disorder via regulating gut microbiome and tryptophan metabolism: a randomized clinical trial. Brain Behav Immun 2022; 100: 233–241.

18.

Rudzki

Ostrowska

Pawlak

, et al. Probiotic Lactobacillus plantarum 299v decreases kynurenine concentration and improves cognitive functions in patients with major depression: a double-blind, randomized, placebo controlled study. Psychoneuroendocrinology 2019; 100: 213–222.

19.

Zhu

Fang

, et al. Psychobiotic Lactobacillus plantarum JYLP-326 relieves anxiety, depression, and insomnia symptoms in test anxious college via modulating the gut microbiota and its metabolism. Front Immunol 2023; 14: 1158137.

20.

, et al. Bacteroides vulgatus alleviates dextran sodium sulfate-induced colitis and depression-like behaviour by facilitating gut-brain axis balance. Front Microbiol 2023; 14: 1287271.

21.

Nakase

Uchino

Shinzaki

, et al. Evidence-based clinical practice guidelines for inflammatory bowel disease 2020. J Gastroenterol 2021; 56(6): 489–526.

22.

D’Haens

Sandborn

Feagan

, et al. A review of activity indices and efficacy end points for clinical trials of medical therapy in adults with ulcerative colitis. Gastroenterology 2007; 132(2): 763–786.

23.

Sandborn

Sands

Vermeire

, et al. Modified Mayo score versus Mayo score for evaluation of treatment efficacy in patients with ulcerative colitis: data from the tofacitinib OCTAVE program. Therap Adv Gastroenterol 2022; 15: 17562848221136331.

24.

von Elm

Altman

Egger

, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol 2008; 61(4): 344–349.

25.

von Elm

Altman

Egger

, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med 2007; 147(8): 573–577.

26.

Hamilton

The assessment of anxiety states by rating. Br J Med Psychol 1959; 32(1): 50–55.

27.

Maier

Buller

Philipp

, et al. The Hamilton Anxiety Scale: reliability, validity and sensitivity to change in anxiety and depressive disorders. J Affect Disord 1988; 14(1): 61–68.

28.

Rabinowitz

Williams

JBW

Hefting

, et al. Consistency checks to improve measurement with the Hamilton Rating Scale for Anxiety (HAM-A). J Affect Disord 2023; 325: 429–436.

29.

Zhou

Kang

, et al. Psychosocial factors associated with anxious depression. J Affect Disord 2023; 322: 39–45.

30.

Bisgaard

Allin

Elmahdi

, et al. The bidirectional risk of inflammatory bowel disease and anxiety or depression: a systematic review and meta-analysis. Gen Hosp Psychiatry 2023; 83: 109–116.

31.

Kim

Lee

Han

, et al. Depression and anxiety are associated with poor outcomes in patients with inflammatory bowel disease: a nationwide population-based cohort study in South Korea. Gen Hosp Psychiatry 2023; 81: 68–75.

32.

Roderburg

Yaqubi

Konrad

, et al. Association between inflammatory bowel disease and subsequent depression or anxiety disorders—a retrospective cohort study of 31,728 outpatients. J Psychiatr Res 2024; 169: 231–237.

33.

Marrie

Graff

Fisk

, et al. The relationship between symptoms of depression and anxiety and disease activity in IBD over time. Inflamm Bowel Dis 2021; 27(8): 1285–1293.

34.

Wilkinson

Trick

Knight

, et al. Factors associated with depression in people with inflammatory bowel disease: the relationship between active disease and biases in neurocognitive processing. Neurogastroenterol Motil 2019; 31(8): e13647.

35.

Swaminathan

Fan

Borichevsky

, et al. The disease severity index for inflammatory bowel disease is associated with psychological symptoms and quality of life, and predicts a more complicated disease course. Aliment Pharmacol Therap 2022; 56(4): 664–674.

36.

Sharma

Agriantonis

Twelker

, et al. Gut microbiota serves as a crucial independent biomarker in inflammatory bowel disease (IBD). Int J Mol Sci 2025; 26(6): 2503.

37.

Basha

Hafez

Salem

, et al. Impact of gut microbiome alteration in ulcerative colitis patients on disease severity and outcome. Clin Exp Med 2023; 23(5): 1763–72.

38.

Mules

Swaminathan

Hirschfeld

, et al. The impact of disease activity on psychological symptoms and quality of life in patients with inflammatory bowel disease-results from the Stress, Anxiety and Depression with Disease Activity (SADD) Study. Aliment Pharmacol Ther 2022; 55(2): 201–211.

39.

Navabi

Gorrepati

Yadav

, et al. Influences and impact of anxiety and depression in the setting of inflammatory bowel disease. Inflamm Bowel Dis 2018; 24(11): 2303–2308.

40.

Gao

Tang

Lei

, et al. Symptoms of anxiety/depression is associated with more aggressive inflammatory bowel disease. Sci Rep 2021; 11(1): 1440.

41.

Petrullo

Ren

, et al. Glucocorticoids coordinate changes in gut microbiome composition in wild North American red squirrels. Sci Rep 2022; 12(1): 2605.

42.

Diviccaro

Giatti

Cioffi

, et al. The gut-microbiota-brain axis: focus on gut steroids. J Neuroendocrinol 2025; 37(7): e13471.

43.

Lloyd-Price

Arze

Ananthakrishnan

, et al. Multi-omics of the gut microbial ecosystem in inflammatory bowel diseases. Nature 2019; 569(7758): 655–662.

44.

Dan

Mao

Liu

, et al. Altered gut microbial profile is associated with abnormal metabolism activity of Autism Spectrum Disorder. Gut Microbes 2020; 11(5): 1246–1267.

45.

Yang

Hou

Wang

, et al. Unfavourable intrauterine environment contributes to abnormal gut microbiome and metabolome in twins. Gut 2022; 71(12): 2451–2462.

46.

Marano

Mazza

Lisci

, et al. The microbiota-gut-brain axis: psychoneuroimmunological insights. Nutrients 2023; 15(6): 1496.

47.

Bear

Dalziel

Coad

, et al. The microbiome-gut-brain axis and resilience to developing anxiety or depression under stress. Microorganisms 2021; 9(4): 723.

48.

Liu

Wang

Chen

, et al. Gut microbiota and its metabolites in depression: from pathogenesis to treatment. EBioMedicine 2023; 90: 104527.

49.

Lai

Zhang

, et al. Multi-omics analyses of serum metabolome, gut microbiome and brain function reveal dysregulated microbiota-gut-brain axis in bipolar depression. Mol Psychiatry 2022; 27(10): 4123–4135.

50.

Cheng

Ruan

, et al. New pathway ameliorating ulcerative colitis: focus on Roseburia intestinalis and the gut-brain axis. Therap Adv Gastroenterol 2021; 14: 17562848211004469.

51.

Zhou

Fan

, et al. Microbiome and tryptophan metabolomics analysis in adolescent depression: roles of the gut microbiota in the regulation of tryptophan-derived neurotransmitters and behaviors in human and mice. Microbiome 2023; 11(1): 145.

52.

Calder

PC.

Omega 3 polyunsaturated fatty acids, inflammation and immunity. World Rev Nutr Diet 2001; 88: 109–116.

53.

Gorica

Calderone

Arachidonic acid derivatives and neuroinflammation. CNS Neurol Disord Drug Targets 2022; 21(2): 118–129.

54.

Lind

A-B

Reis

Bengtsson

, et al. Steady-state concentrations of mirtazapine, N-desmethylmirtazapine, 8-hydroxymirtazapine and their enantiomers in relation to cytochrome P450 2D6 genotype, age and smoking behaviour. Clin Pharmacokinet 2009; 48(1): 63–70.

55.

Pinto-Sanchez

Hall

Ghajar

, et al. Probiotic Bifidobacterium longum NCC3001 reduces depression scores and alters brain activity: a pilot study in patients with irritable bowel syndrome. Gastroenterology 2017; 153(2): 448–459.e8.

56.

Han

S-K

Kim

DH.

Lactobacillus mucosae and Bifidobacterium longum synergistically alleviate immobilization stress-induced anxiety/depression in mice by suppressing gut dysbiosis. J Microbiol Biotechnol 2019; 29(9): 1369–1374.

57.

Qian

Zhu

, et al. Bifidobacteria with indole-3-lactic acid-producing capacity exhibit psychobiotic potential via reducing neuroinflammation. Cell Rep Med 2024; 5(11): 101798.

58.

Lavelle

Sokol

Gut microbiota-derived metabolites as key actors in inflammatory bowel disease. Nat Rev Gastroenterol Hepatol 2020; 17(4): 223–237.

59.

Savitz

The kynurenine pathway: a finger in every pie. Mol Psychiatry 2020; 25(1): 131–147.

60.

C-H

Chuang

H-C

Hong

C-J.

Physical exercise prevents mice from L-Kynurenine-induced depression-like behavior. Asian J Psychiatr 2020; 48: 101894.

61.

Cheng

Huang

Chen

Tauroursodeoxycholic acid ameliorates lipopolysaccharide-induced depression like behavior in mice via the inhibition of neuroinflammation and oxido-nitrosative stress. Pharmacology 2019; 103(1–2): 93–100.

62.

Yang

R-R

Zhang

J-L

, et al. Tauroursodeoxycholic acid produces antidepressant-like effects in a chronic unpredictable stress model of depression via attenuation of neuroinflammation, oxido-nitrosative stress, and endoplasmic reticulum stress. Fundam Clin Pharmacol 2018; 32(4): 363–377.

63.

Khalaf

Tornese

Cocco

, et al. Tauroursodeoxycholic acid: a potential therapeutic tool in neurodegenerative diseases. Transl Neurodegener 2022; 11(1): 33.

64.

Dell

Jacobi

Kamp

, et al. [1-Methylhydantoin, an unexpected metabolite of the intelligence-affecting substance dupracetam (author’s transl)]. Arch Pharm 1981; 314(8): 697–702.

65.

Maayah

Abdelhamid

El-Kadi

AO.

Development of cellular hypertrophy by 8-hydroxyeicosatetraenoic acid in the human ventricular cardiomyocyte, RL-14 cell line, is implicated by MAPK and NF-κB. Cell Biol Toxicol 2015; 31(4–5): 241–259.

66.

Zhang

Lin

Zhao

, et al. Structural characterization of extracellular polysaccharides from Phellinus igniarius SH-1 and their therapeutic effects on DSS induced colitis in mice. Int J Biol Macromol 2024; 275(Pt 2): 133654.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB

0.10 MB

0.04 MB