Chronic stress is associated with altered gut microbiota profile and relevant metabolites in adolescents

Abstract

Background

Gut microbiota and microbiota-derived metabolites have been implicated in the regulation of stress-related diseases, yet their associations with chronic stress in adolescents remain unclear. Multi-omics studies on this topic in adolescents are still limited. This study aimed to characterize gut microbiota and metabolites in adolescents under chronic stress.

Methods

In this cross-sectional study, we assessed chronic stress in 124 adolescents aged 12–16 years using the Adolescent Life Events Scale and the Study Stress Scale. Participants were stratified by stress level into low (n = 42), medium (n = 41), and high stress (n = 41) groups. Fecal samples were collected from all participants for 16S rRNA gene sequencing. Subsequently, a subset of 30 adolescents with high stress and 29 low stress adolescents underwent metagenomic sequencing and untargeted metabolomics.

Results

Adolescents experiencing high-chronic stress showed lower alpha diversity, differential beta diversity, and a more complicated microbial network compared to those experiencing lower stress. Spearman’s rank correlation and Kruskal-Wallis test identified five genera with decreased abundances in high stress adolescents, including Faecalibacterium, Bacteroides, Akkermansia, Lachnospiraceae unclassified, and Ruminococcus (P_fdr<0.05). Additionally, 12 species showed decreased abundances and 5 increased abundances, and logistic regression analysis further revealed that the relative abundances of Bifidobacterium catenulatum, Streptococcus suis, Ruminococcus sp. CAG 108, and Phascolarctobacterium faecium were associated with chronic stress (P_fdr<0.05), after adjusting for sex, age, fruit consumption, and body mass index. We identified 21 differential metabolites, predominantly enriched in metabolic pathways based on KEGG analysis. Moreover, 19 out of these metabolites were significantly correlated with at least one of the four species significantly associated with chronic stress. These metabolites may explain health effects of species associated with chronic stress.

Conclusions

Chronic stress in adolescents is associated with altered gut microbiota composition and metabolite profiles, providing insights into possible mechanisms underlying stress-related diseases and highlighting the importance of longitudinal studies to clarify temporal and causal relationships.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12866-025-04094-1.

Background

Globally, 14.3% of adolescents suffer from mental health disorders, accounting for 13% of the global disease burden in this age group [1]. In response to the mental health challenges faced by adolescents, the World Health Organization launched the “Global Accelerated Action for the Health of Adolescents (AA-HA!)” initiative. Chronic stress has emerged as a major risk factor for adolescent mental health disorders [1]. Adolescents are subjected to multiple sources of chronic stress such as family, peers, and community-related stress. The prevalence of chronic stress varies globally: approximately 10.5% of adolescents in China report moderate- to high-level of chronic stress [2], compared to 26.9% in Canada and 61.5% in Bangladesh [3]. The COVID-19 pandemic has added further stress due to disruptions in daily life, study routines, and social interactions caused by social distancing measures [4]. Importantly, adolescence is a critical period during which chronic stress can result in lasting mental health effects.

Emerging evidence from the microbiota-gut-brain axis suggests that chronic stress-induced alterations in gut microbiota composition may play a pivotal role in the pathophysiology of stress-related diseases [5, 6]. However, current evidence on this relationship derives predominantly from animal studies and research in adult populations. Data from adolescents remain limited. To date, only two studies have investigated the association between chronic stress and gut microbiota composition in adolescents. One study found that adolescents under high stress had a lower alpha diversity [7]. Chronic stress was associated with increased abundances of Bacteroides, Parabacteroides, Rhodococcus, Methanobrevibacter, and Roseburia but decreased Phascolarctobacterium at the genus level [7]. Another study on youth patients with inflammatory bowel disease found that the relative abundances of Anaerostipes and Parabacteroides varied by stress levels [8]. However, both studies used 16 S rRNA sequencing, which precludes resolution of microbial changes at the species level and may obscure stress-specific taxonomic signatures.

Further, chronic stress may affect gut microbiota-derived metabolites by altering gut microbiota composition [9, 10]. However, current research examining the relationship between chronic stress and microbiota metabolites has primarily been conducted in animal models, youth patients with medical conditions, and adults. For example, animal studies have demonstrated that chronic stress significantly changed many metabolites in the jejunum, such as L-glutamine and L-tyrosine [11]. Metabolomic analysis revealed alterations in 5-hydroxytryptamine and kynurenine metabolites in mice under chronic stress [9, 12]. Additionally, metabolites such as alanine, nictate, lysine, and phenylalanine differed between Crohn’s disease patients under high- and low-perceived stress [8]. In adults, Coley et al. found that early-life adversity was associated with alterations in glutamate pathway metabolites (e.g., 5-oxoproline and malate) [13]. Nevertheless, whether chronic stress affects gut microbiota metabolites in adolescents remains unclear.

To address these gaps, we investigated the associations among chronic stress, gut microbiota, and their metabolites in adolescents, aiming to identify key microbial and metabolic biomarkers linked to chronic stress. Our findings may offer valuable insights into the gut-brain axis pathogenesis in adolescents exposed to high chronic stress.

Methods

Study design and participants

In this cross-sectional study, we recruited 124 adolescents aged 12–16 from a middle school in Xi’an, China, including 62 boys and 62 girls, during December 2022 and January 2023. Participants with pre-existing digestive disease or those who had taken antibiotics, probiotics, or prebiotics within the past three months were excluded. Additionally, participants who had difficulty in understanding the questionnaire were also excluded. Socio-demographic characteristics, physical activity, dietary habits, anthropometrics, chronic stress, and fecal samples were collected from all participant. Written informed consent was obtained from all adolescents and their parents. The study was approved by the Institutional Review Board at Xi’an Jiaotong University Health Science Center.

Measures

Chronic stress

Psychological stress was assessed by adolescent’s self-reported psychological stress and study stress [1]. Psychological stress was measured by the 27-item Adolescent Self-Rating Life Events Checklist (ASLEC), which assesses psychological stress over the past 12 months. The ASLEC includes five dimensions: punishment (e.g., “Scolded or beaten by parents”), interpersonal relationships (e.g., “Been misunderstood or blamed”), academic stress (e.g., “Heavy learning burden”), loss (e.g., “Relatives or friends died”), and adaptation problems (e.g., “Long-term separation from family”). Each item is rated on a 6-point Likert scale ranging from “not happened” to “extremely severe”. The total score ranges from 0 to 135, with higher scores indicating more severe life stress. The ASLEC has been widely used among Chinese adolescents and demonstrates good reliability and validity [14]. In this study, the scale’s Cronbach’s alpha was 0.91 [2]. Study stress was assessed using the Study Stress Scale [15]. Each item is self-reported on a 5-point scale ranging from “never true” to “very often true”. The total score ranges from 0 to 64, with a higher score indicating greater study stress. The scale a reliable and valid tool for measuring study-related stress [15]. In this study, the Cronbach’s alpha was 0.78. The chronic stress score was calculated as the sum of the standardized scores of psychological stress and study stress. Based on tertiles of this score, adolescents were categorized into low-, medium-, and high-chronic stress groups. The total chronic stress score was used as both a categorical and continuous variable in the data analyses.

Covariates

Covariates including age, sex, food consumption (vegetables, fruits, fish, meat, and milk), and body mass index (BMI) were adjusted for the logistic regression analyses examining the associations between gut microbial species and chronic stress. Each food category was adjusted separately. Age and sex were self-reported by participants. Food consumption frequency was assessed using a food frequency questionnaire (FFQ) [16]. BMI was calculated using the formula: BMI = weight (kg)/height (m) ², with body weight and height measured.

Gut microbiota composition and its metabolites

Fecal sample collection

Fecal collection containers were provided for each participant. The fecal samples were collected at school or home, temporarily stored in ice boxes, then transferred to the laboratory, and stored at −80℃ within 8 h until DNA extraction. The quality control details are shown in Table S1. DNA was extracted using the QIAamp DNA Stool Mini Kit (Qiagen, Hilden, Germany) following the manufacturer’s instructions.

16 S rRNA gene amplicon and sequencing

Feces from 42 adolescents with low stress, 41 with medium stress, and 41 with high stress were examined. Amplification of the 16 S ribosomal RNA (rRNA) gene was performed using primers (515 F-806R) targeting the V3-V4 high variant region of the 16 S rRNA gene. Ultra-pure water (Milli-Q, MerckMillipore) was used as the negative control. Sequencing libraries were generated using the NEBNext® Ultra™ IIDNA Library Prep Kit (Cat No. E7645) according to the manufacturer’s instructions. Library quality was evaluated using the Qubit@ 2.0 Fluorometer (Thermo Scientific) and Agilent Bioanalyzer 2100 system. Sequencing was performed on an Illumina NovaSeq platform, generating 250 bp paired-end reads. The raw data were analyzed using the QIIME2 platform, and sequencing reads were filtered using the DADA2 plugin. The final amplicon sequence variants (ASVs) were obtained by filtering out sequences with abundances below 5. Species annotation was performed using QIIME2 software to analyze the phylogenetic relationship of each ASV and to identify dominant species across samples. Each ASV was classified using a pre-trained Naive Bayes classifier with QIIME2’s classify-sklearn algorithm. ASVs are often used as a proxy for species or operational taxonomic units (OTUs) and represent distinct genetic variants within a microbial community.

Metagenomic sequencing

We selected a subset of participants for metagenomic sequencing based on their chronic stress scores. Specifically, the 30 adolescents with the highest chronic stress scores from the high stress group, and the 30 adolescents with the lowest scores from the low stress group were selected. A fecal sample from one adolescent in the low stress group was excluded due to insufficient DNA concentration, yielding final sample of 30 adolescents with high stress and 29 adolescents with low stress for metagenomic sequencing analysis.

After the fecal sample’s DNA detection was qualified, the DNA was fragmented, and end repair of the fragmented DNA was performed. An “A” nucleotide was added to the 3’end, sequencing splices were connected, and the linked product purification sheet screen was formed, followed by library amplification and product purification to form sequencing libraries. After passing the quality check, the library was sequenced with an Illumina MiSeq Sequencer. To ensure the accuracy of data analysis, fastp software was used to filter Raw Tags and obtain clean reads for subsequent analysis. MEGAHIT software was used for metagenomic assembly, and contigs shorter than 300 bp were filtered out. Gene annotation was performed using the Non-Redundant Protein Sequence Database (NR) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) databases.

Metabolomics analysis

The subset of participants selected for metagenomic sequencing additionally underwent untargeted metabolomics analysis. Fecal samples were analyzed using the liquid chromatography-mass spectrometry (LC/MS) method. The LC/MS system for metabolomics analysis consists of a Waters Acquity I-Class PLUS ultra-high performance liquid chromatograph coupled with a Waters Xevo G2-XS QTof high-resolution mass spectrometer. The Waters Xevo G2-XS QTof high-resolution mass spectrometer can collect both primary and secondary mass spectrometry data in MSe mode under the control of the acquisition software. In each data acquisition cycle, dual-channel data acquisition can be performed on both low and high collision energies simultaneously. The raw data collected using MassLynx V4.2 were processed using Progenesis QI software for peak extraction, peak alignment, and other data processing operations, based on the Progenesis QI software online METLIN database and Biomark’s self-built library for identification.

After normalizing the peak area data by the total peak area, a follow-up analysis was performed. Principal component analysis and Spearman’s rank correlation analysis were used to assess the repeatability of intra-group samples and quality control samples. The identified compounds were searched for classification and pathway information in KEGG. Additionally, endogenous metabolites were screened in the Human Metabolome Database.

Statistical analysis

Background characteristics of adolescents and their parents

The Kolmogorov-Smirnov normality test was used to assess distributions of continuous variables for normality. Continuous variables were described as means (standard deviations) for normally distributed variables or medians (interquartile ranges) for skewed variables, while categorical variables were summarized using frequencies and percentages. The Analysis of Variance (ANOVA) was used to compare the differences in background characteristics across the three groups for normally distributed variables, while the Kruskal-Wallis test was used for skewed variables, and the Chi-square test was used for categorical variables.

Differences in the microbial community

The Kruskal-Wallis test, followed by Dunn’s multiple comparison tests, was used to assess differences in alpha diversities across the three groups. The Shannon index, Simpson’s diversity index, Chao1 index, and the number of observed ASVs were used to evaluate alpha diversity. The Shannon index measures both the richness and evenness of species within each sample. The Simpson’s diversity index measures species diversity, with higher values indicating greater diversity. The Chao1 index provides an estimate of the total species richness, with a particular focus on rare species. The number of observed ASVs directly reflects the diversity within a sample, indicating the variety of microbial species or genetic variants present. Beta diversity was calculated using Principal Coordinate Analysis (PCoA) of Unweighted UniFrac distances. Co-occurrence networks were constructed separately for the high stress, low stress, and medium stress groups using 16 S rRNA gene sequencing data. Spearman’s correlation coefficients were calculated based on the relative abundances of genera. Significant correlations (r ≥ 0.6, P < 0.05) were used to construct the networks, which were visualized using Cytoscape (version 3.9.1). To compare network complexity across groups, we quantified the number of nodes, edges, and average degree in each network. Linear Discriminant Analysis Effect Size (LEfSe) uses the Kruskal-Wallis test to identify features with significantly different abundances between taxa and estimate the effect size of each feature. To compare the relative abundance of bacteria at phylum and genus levels in adolescents under low-, medium-, and high-chronic stress, the Kruskal-Wallis test followed by Dunn’s multiple comparison tests was used. For comparisons at the species level, the Wilcoxon rank-sum test followed by Dunn’s multiple comparison tests was used. Spearman’s rank correlation test was conducted to evaluate correlations between chronic stress and the relative abundances of microbiota at the phylum, genus, and species levels, with the Benjamini-Hochberg method employed to control the False Discovery Rate (FDR). Logistic regression analysis was used to identify the key species associated with chronic stress, adjusting for sex (boys and girls), age (in years), food consumption (frequency/day), and BMI (kg/m2) as covariates. Covariates were selected based on biological relevance and univariate analysis results (P < 0.1). Multicollinearity among covariates was assessed using the variance inflation factor (VIF), and variables with VIF > 10 were excluded from the final models. Functional prediction for the fecal microbiome was performed using significant KEGG pathways.

Differences in microbial metabolites

The Wilcoxon rank-sum test was used to assess the statistical significance of differences in metabolite levels of each metabolite. Partial least squares discriminant analysis (PLS-DA) and orthogonal partial least squares discriminant analysis (OPLS-DA) were used to identify differential metabolites. PLS-DA and OPLS-DA are multivariate statistical methods used to model and discriminate between groups’ differences based on multiple variables [17, 18], which are particularly useful in metabolomics where the goal is to identify patterns that distinguish different groups based on their metabolic profiles. The R language package “ropls” was used to perform PLS-DA and OPLS-DA modeling with 200 permutation tests to validate the reliability of the model (https://bioconductor.org/packages/release/bioc/html/ropls.html). The VIP value of the model was calculated using multiple cross-validation. A method combining the fold change, the P-value, and the VIP of OPLS-DA model was adopted to screen the differential metabolites. The screening criteria were a fold change > 2 or fold change < 0.5, a P-value < 0.05, and a VIP > 1. The KEGG pathway enrichment analysis of identified differential metabolites was calculated using a hypergeometric distribution test. KEGG pathway enrichment analysis is a bioinformatics method used to identify biological pathways from the KEGG database that are overrepresented in a given set of genes or metabolites [19]. Spearman’s rank correlation analysis was used to examine the correlations between differential metabolites and differential species, as well as the correlations between chronic stress and the differential metabolites significantly correlated with at least one of the identified species, with the Benjamini-Hochberg method employed to control the FDR.Cite Reference.

The Stata (version 15.1) and R software (version 4.1.2) were used for the statistical analysis, and data presentation was performed using GraphPad Prism (version 9.3.1) and R software (version 4.1.2).

Results

Background characteristics of participants

The scores of psychological stress, study stress, and their combined standardized scores significantly differed across the low-, medium-, and high-chronic stress groups. No significant differences were observed in the sex, ethnicity, age, BMI, physical activity, and dietary habits of adolescents across the three groups. Likewise, parental characteristics—including age, education level, and household income, demonstrated no significant difference across the three groups (Table 1). The characteristics of the adolescents selected for the metagenomic metabolomics study (N = 59) were consistent with the findings in the overall population (N = 124) (Table S2).

Table 1.

Characteristics of all adolescents and their parents

Characteristic	Low chronic stress (N = 42)	Medium chronic stress (N = 41)	High chronic stress (N = 41)	p-value
Adolescent’s characteristics
Sex (%)
Boys	20 (47.6)	19 (46.3)	23 (56.1)	0.63
Girls	22 (52.4)	22 (53.7)	18 (43.9)
Ethnicity (%)
Han	40 (95.2)	40 (97.6)	41 (100.0)	0.369
Others	2 (4.8)	1 (2.4)	0 (0.0)
Age	14.0 (13.0, 14.0)	14.0 (13.0, 14.0)	14.0 (13.0, 14.0)	0.314
Study stress	2.5 (0.0, 4.0)	11.0 (7.0, 13.0)	22.0 (17.0, 28.0)	< 0.001
Psychological stress	5.5 (4.0, 10.8)	19.0 (13.0, 25.0)	44.0 (36.0, 52.0)	< 0.001
Interpersonal relationship	2.0 (0.3, 3.0)	6.0 (4.0, 7.0)	11.0 (9.0, 14.0)	< 0.001
Academic stress	2.0 (1.0, 3.0)	5.0 (3.0, 6.0)	9.0 (7.0, 12.0)	< 0.001
Punishment	0.0 (0.0, 2.0)	4.0 (1.0, 6.0)	10.0 (8.0, 16.0)	< 0.001
Lose	0.0 (0.0, 1.0)	2.0 (0.0, 3.0)	4.0 (2.0, 6.0)	< 0.001
Adaption	1.0 (0.0, 2.0)	2.0 (1.0, 3.0)	6.0 (4.0, 9.0)	< 0.001
Physical activity (minute)	740.0 (465.0, 1,215.0)	580.0 (390.0, 830.0)	680.0 (475.0, 915.0)	0.245
Body mass index (kg/m²)	19.9 (18.4, 24.2)	20.0 (18.4, 23.5)	20.2 (18.3, 23.7)	0.899
Body fat mass (kg)	11.8 (7.7, 20.0)	12.5 (9.3, 18.9)	13.7 (8.4, 18.4)	0.843
Percent body fat (%)	24.0 (15.7, 31.9)	24.8 (19.6, 29.9)	26.3 (17.7, 31.0)	0.988
Food frequency questionnaire (times/month)
Vegetables	8.0 (7.0, 10.0)	8.0 (6.0, 10.0)	8.0 (6.0, 10.0)	0.396
Fruits	4.0 (4.0, 5.0)	4.0 (3.0, 5.0)	4.0 (4.0, 5.0)	0.326
Fish	4.0 (3.0, 4.0)	4.0 (3.0, 4.0)	4.0 (2.0, 4.0)	0.217
Meat	6.0 (5.3, 8.0)	6.0 (5.0, 7.0)	6.0 (5.0, 8.0)	0.434
Milk	4.0 (2.3, 5.0)	3.0 (2.0, 5.0)	3.0 (2.0, 5.0)	0.495
Parental characteristics
Maternal age (years)	41.0 (37.5, 44.0)	38.5 (36.0, 43.0)	38.0 (36.0, 42.0)	0.184
Paternal age (years)	42.0 (38.5, 45.0)	40.5 (37.3, 45.0)	40.0 (37.0, 43.5)	0.465
Paternal highest education (%)
Below high school	24 (61.5)	25 (65.8)	22 (62.9)	0.558
High school	12 (30.8)	11 (28.9)	13 (37.1)
College or above	3 (7.7)	2 (5.3)	0 (0.0)
Maternal highest education (%)				0.947
Below high school	29 (74.4)	25 (65.8)	24 (68.6)
High school	9 (23.1)	12 (31.6)	10 (28.6)
College or above	1 (2.6)	1 (2.6)	1 (2.9)
Household income (RMB, 10,000 yuan/year)	4.1 (2.9)	5.2 (4.2)	4.9 (3.2)	0.319

Continuous data are presented as medians (interquartile ranges) for skewed variables, while categorical variables are represented as frequency (percentage). The Kruskal-Wallis test or Chi-square test was used to compare group differences

Alterations in gut microbiota composition across different stress groups based on the 16 S rRNA data

Adolescents experiencing high chronic stress exhibited significantly reduced alpha diversity in the gut microbiome, characterized by reductions in the Chao1 index, observed ASVs, Simpson’s Diversity Index, and Shannon index levels (Figf. 1a–d; Figures S1a–d). PCoA of beta diversity revealed distinct clustering patterns among the three groups (Fig. 1e). The Venn diagram illustrated that 1,314 ASVs were shared among the three groups, while 3,955, 4,036, and 2,074 ASVs were unique to the adolescents under low stress, medium stress, and high stress, respectively (Fig. 1f). Participants under high stress exhibited a significantly higher Firmicutes/Bacteroidota ratio compared to those under medium stress (P < 0.05) (Figure S1e).

Fig. 1.

Gut microbial alterations in the three stress groups that were divided by the tertile values of chronic stress Note: a, b, c, d: The correlation between the alpha diversity of gut microbiota and the total score of chronic stress. e Principal Coordinate Analysis (PCoA, based on unweighted unifrac distances) of beta diversity of the three groups. f: Venn diagram of the number of observed ASVs in participants with different levels of chronic stress

Significant differences were observed in the co-occurrence networks of genera among the three stress groups, based on the significant Spearman’s rank correlations (r > 0.6, P < 0.05). These genera primarily belonged to four phyla: Firmicutes, Bacteroidota, Proteobacteria, and Actinobacteria. The high stress group displayed a more complex microbial network, with significantly more nodes, edges, and average degrees than the low- and medium stress groups (Figure S1 f- h, and Table S3).

LEfSe identified substantial differences in microbiota composition across the three stress groups. At the genus level, adolescents under low stress had more abundant Faecalibacterium, Lachnospiraceae unclassified, Desulfovibric, Allobravorella, and Lachnospiraceae NK4 A136 group. In contrast, adolescents under high stress showed a higher abundance of Sphingomonas (Fig. 2a).

Fig. 2.

The difference in gut microbes’ abundance in adolescents with low-, medium-, and high-chronic stress based on the 16 S rRNA data Note: a: Linear discriminant analysis (LDA) scores for the bacterial taxa differentially abundant in adolescents with different chronic stress levels (LDA > 2). Red bars indicate taxa were enriched in the low stress group, green bars indicate taxa were enriched in the medium stress group, and blue bars indicate taxa were enriched in the high stress group. b, c Spearman’s rank correlations of the total score of chronic stress and the score of individual dimensions with the relative abundances of bacteria taxa at the phylum- (b) and genus-levels (c), respectively. The P-value was corrected by the False Discovery Rate (FDR). d, e The Kruskal-Wallis test was used to analyze differences in the relative abundances of bacteria taxa at the phylum- (d) and genus- (e) levels among adolescents with varying stress levels. These analyses were based on bacteria taxa that were significantly correlated with chronic stress, as identified by Spearman’s rank correlation in (b) and (c). The P-value was corrected by FDR. The PS-SS-Z is chronic stress, which was calculated by the sum of the standardized score of psychological stress and study stress

Adolescents under low-, medium-, and high stress displayed distinct microbial profiles at phylum and genus levels, as revealed by Spearman’s rank correlations and the Kruskal-Wallis test. At the phylum level, compared with adolescents under low stress, those under high stress showed significantly increased relative abundances of Chloroflexi, while decreased relative abundances of Bacteroidota, Verrucomicrobiota, and Desulfobacterota (P_fdr<0.05, Fig. 2b and c). At the genus level, participants under high stress had significantly lower relative abundances of Faecalibacterium, Bacteroides (belongs to Bacteroidota phylum), Akkermansia (belongs to Verrucomicrobiota phylum), Lachnospiraceae unclassified, and Ruminococcus (P_fdr<0.05, Fig. 2d and e). Similar correlations were found for the scores of study stress, psychological stress, and individual dimensions of psychological stress with the relative abundances of these genera (P_fdr<0.05, Fig. 2c).

Metagenomic sequencing revealed significant differences in gut microbial species in adolescents with low- and high stress

Chronic stress was associated with gut microbial composition at the species level, as revealed by metagenomic sequencing. Both the ACE and Chao1 index were significantly lower in the high stress group than in the low stress group (Fig. 3a and b). However, the PCoA of beta diversity did not reveal significant differences between the two groups (Figure S2a). The Venn diagram showed that 996,723 genes were shared between the two groups, while 11,611,695 and 864,930 genes were unique to the low stress and high stress groups, respectively (Figure S2b). We compared the relative abundances of the top 50 species between high stress and low stress, the relative abundance of Bifidobacterium catenulatum, Chitinophaga silvisoli, Akkermansia muciniphila, and Dialister invisus were significantly different (Figure S2c). Roseburia inulinivorans, Dialister invisus, Chitinophaga silvisoli, Roseburia faecis, Akkermansia muciniphila, and uncultured Ruminococcus sp were the most abundant species in adolescents under low stress (Fig. 3c).

Fig. 3.

The gut microbiota divergence in adolescents with low- and high-chronic stress based on the metagenomic sequencing data Note: a, b: Alpha diversity indices of the gut microbial species in adolescents with high- and low-chronic stress. ACE numbers (a), Chao1 index (b). c The specific bacteria taxa of two stress groups based on Linear Discriminant Analysis Effect Size (LEfSe) analysis. d Heatmap of the Spearman’s rank correlation coefficients between the relative abundance of gut microbial species and chronic stress. The P-value was corrected by the False Discovery Rate (FDR). e The Wilcoxon rank-sum test was used to analyze the difference in the relative abundances in adolescents with low- and high-chronic stress. These analyses were based on gut microbial species that were significantly correlated with chronic stress, as identified by Spearman’s rank correlation in Fig. 3(d). The P-value was corrected by FDR. f Logistic regression analysis was used to analyze the adjusted associations between chronic stress and gut microbial species, identified in Fig. 3 (e). The logistic regression analysis adjusted for sex, age, food consumption, and body mass index as covariates. g Metabolic pathway analysis based on differential metabolites using the Kyoto Encyclopedia of Genes and Genomes (KEGG). LS: Low stress, HS: High stress

At the species level, adolescents under high stress had lower abundances of Arenibacter certesii, Coprococcus sp. ART55 1, Denitrificimonas caeni, Desulfovibrio sp. An276, Halomonas halocynthiae, Oscillibacter sp. PEA192, Phascolarctobacterium faecium, Roseburia faecis, Ruminococcus sp. AF17 11, Ruminococcus sp. AM36 18, Ruminococcus sp. CAG 108, and Streptococcus suis, as revealed by the Kruskal-Wallis test and Spearman’s rank correlation (P_fdr<0.05, Fig. 3d and e). Most of the identified differential species belonged to the genus Ruminococcus, which was consistent with the findings of the 16 S rRNA analysis at the genus level. Adolescents under high stress demonstrated higher abundances of Achromobacter insuavis, Bifidobacterium catenulatum, Bifidobacterium pseudocatenulatum CAG 263, Desulfovibrio sp. 3 1 syn3, and Clostridium sp. AM25 23 AC, with most of these differential species belonging to the genus Bifidobacterium (P_fdr<0.05, Fig. 3d and e).

After adjusting for sex, age, food consumption, and BMI, logistic regression analysis revealed that four of these differential species remained significantly associated with chronic stress (Fig. 3f). Specifically, Bifidobacterium catenulatum was significantly associated with increased odds of chronic stress (OR = 2.7, 95% CI: 1.02, 7.05). Conversely, Streptococcus suis (OR = 0.12, 95%CI: 0.02, 0.73), Ruminococcus sp. CAG 108 (OR = 0.31, 95%CI: 0.12, 0.79), and Phascolarctobacterium faecium (OR = 0.48, 95%CI: 0.25, 0.91) were associated with decreased odds of chronic stress.

We further compared gut microbial functions across the two groups, and all the KEGG pathways at levels 2 and 3 were disrupted in the high stress group (Fig. 3g).

Metabolomics analysis revealed aberrant microbial metabolic patterns in adolescents with low- and high stress

No significant difference in the overall microbial metabolic profile was revealed by PLS-DA and OPLS-DA (Figure S2 d and S2e). A total of 21 metabolites exhibited alterations in adolescents experiencing high stress compared to those experiencing low stress (Fig. 4a). KEGG pathways analysis showed that these metabolites were mostly enriched in metabolic pathways (Fig. 4b).

Fig. 4.

Microbial metabolic patterns in adolescents with low- and high-chronic stress Note: a: Metabolites showed significant differences between the two stress groups. b Metabolic pathway analysis of differentially abundant metabolites between the two groups based on the Kyoto Encyclopedia of Genes and Genomes (KEGG). c Heatmap of Spearman’s rank correlations between differential metabolites and identified differential gut microbial species. d Heatmap of the Spearman’s rank correlations between chronic stress and identified metabolites significantly correlated with identified differential species

Among the identified differential metabolites, 19 were significantly correlated with at least one of the four species significantly associated with chronic stress (P_fdr<0.05, Fig. 4c). Spearman’s rank correlation further revealed that five of them were positively correlated with chronic stress, including hoduloside I, vinaginsenoside R5, epoxyeicosatrienoic acid, 4,8,12-trimethyl-1,3E,7E,11-tridecatetraene, and (+/-)-[R- (E)]−5-Isopropyl-8-methylnona-6,8-dien-2-one. These elevated metabolites were negatively correlated with the relative abundances of Ruminococcus sp. CAG 108. Four of them were negatively correlated with chronic stress, including acetylcarnitine, isohumbertiol, austalide J, and leukotriene A4. These decreased metabolites were positively correlated with the relative abundance of Ruminococcus sp. CAG 108 while being negatively correlated with Bifidobacterium catenulatum (P_fdr<0.05, Fig. 4c and d).

Discussion

As shown in Fig. 5, the gut microbiota composition of the high stress group differed from that of the low stress group, exhibiting lower alpha diversity and altered beta diversity. Our study identified five decreased genera in adolescents experiencing high-chronic stress compared to those with low chronic stress, including Faecalibacterium, Bacteroides, Akkermansia, Lachnospiraceae unclassified, and Ruminococcus. Further, we identified that the relative abundances of four species were associated with chronic stress, including Streptococcus suis, Ruminococcus sp. CAG 108, Bifidobacterium catenulatum, and Phascolarctobacterium faecium, after adjusting for covariates. Metabolomic profiling revealed 19 differential metabolites that were associated with at least one of the four altered species. Taken together, our findings illustrate a potential connection between chronic stress, bacterial composition and their metabolites in adolescents.

Fig. 5.

The summary of the associations of chronic stress with gut microbiota composition and relevant metabolites in adolescents

Consistent with our findings in adolescents, mice exposed to chronic stress also had decreased relative abundances of genera including Faecalibacterium, Akkermansia, Ruminococcus, and Lachnospiraceae unclassified [9, 20, 21]. Faecalibacterium, which produces short-chain fatty acids (SCFAs) in the cecum, has been shown to attenuate the release of inflammatory cytokines release in rats under chronic unpredictable mild stress [22–25]. Akkermansia is regarded as a next-generation probiotic due to its ability to ameliorate metabolic dysfunction and intestinal inflammation [26–28]. Lachnospiraceae contributes to the degradation of dietary fiber and SCFAs production and exhibits anti-inflammatory effects [29, 30]. It is suggested that the increased inflammation level observed in adolescents under chronic stress [31–33] may be related to the decreased relative abundances of Faecalibacterium, Akkermansia, and Lachnospiraceae.

We observed that higher chronic stress was associated with the reduced relative abundance of Ruminococcus in adolescents. Specifically, one of its species, the relative abundance of Ruminococcus sp. CAG 108 remains lower in the high-chronic stress group after adjusting for covariates. We further found that the abundance of Ruminococcus sp. CAG 108 was positively correlated with the reduced levels of acetylcarnitine. Previous studies have also reported that adolescents with high stress had a lower level of acetylcarnitine [34]. These findings indicate that Ruminococcus sp. CAG 108 may influence health via acetylcarnitine. Acetylcarnitine, a central mitochondrial metabolite, plays a crucial role in fatty acid oxidation [35], aligning with our findings of lower fatty acid metabolism in the high-chronic stress group. Acetylcarnitine has antioxidant properties that may help combat oxidative stress associated with diabetes and obesity. These findings indicate that chronic stress may affect metabolism via Ruminococcus sp. CAG 108 and related metabolites. Further human studies are warranted to clarify the potential role of acetylcarnitine in chronic stress-related metabolic conditions. We found that Ruminococcus sp. CAG‑108 was negatively correlated with leukotriene C4 and positively correlated with leukotriene A4. These results suggest that lower abundance of Ruminococcus sp. CAG-108 in the high stress group is associated with increased leukotriene C4 and decreased leukotriene A4 levels, indicating a possible link with altered leukotriene biosynthesis that warrants further investigation. Given the pro-inflammatory nature of leukotriene C4, these microbial-metabolite associations may indicate a microbially mediated shift toward inflammation under chronic stress conditions in adolescents.

At the species level, we observed a negative association between the relative abundance of Phascolarctobacterium faecium and chronic stress. Phascolarctobacterium faecium is recognized for its ability to produce SCFAs through the fermenting of succinate [36]. Previous studies have indicated that Phascolarctobacterium faecium may possess anti-inflammatory properties, which could be beneficial for managing inflammatory gut disorders [37, 38]. Moreover, by generating SCFAs, Phascolarctobacterium faecium may indirectly modulate other anti-inflammatory mediators, such as prostanoids [5, 39, 40]. Our research provides further evidence of the anti-inflammatory properties of Phascolarctobacterium faecium, by discovering a negative correlation between Phascolarctobacterium faecium and leukotriene C4 (LTC4). LTC4 is a potent pro-inflammatory mediator derived from arachidonic acid via the 5-lipoxygenase pathway [41]. LTC4 is involved in various inflammatory conditions, such as asthma and allergic rhinitis. Emerging evidence suggests that LTC4 and its receptors have been implicated in neuroinflammation and cognitive impairment [42]. Overall, these findings highlight the anti-inflammatory properties of Phascolarctobacterium faecium and its relevant metabolite of LTC4.

Conversely, the relative abundance of Bifidobacterium catenulatum was positively associated with chronic stress in adolescents. From a metabolic standpoint, this species is known for its ability to catabolize monosaccharides [43]. Our KEGG enrichment pathway analysis found that adolescents with high stress showed higher fructose metabolism, fatty acid degradation, and phosphotransferase system, which may be related to the metabolic effects of Bifidobacterium catenulatum. We also found that Bifidobacterium is positively correlated with cholesterol levels. Previous studies have demonstrated that mice exposed to chronic stress exhibit metabolic disorders, with increased cholesterol levels [44, 45]. In contrast, Bifidobacterium catenulatum is a probiotic that contributes to maintaining gut microbiota balance, suppressing the growth of pathogenic bacteria, and promoting intestinal health in adults [46]. However, our study demonstrates that the relative abundance of Bifidobacterium catenulatum is higher in adolescents experiencing high stress level. The observed increase in Bifidobacterium catenulatum in high stress adolescents could represent a compensatory response rather than a pathogenic change. Several reasons may explain this phenomenon. First, Bifidobacterium catenulatum may affect the proliferation of other probiotics in the gut, thereby modifying the intestinal microecological balance [47]. Second, Bifidobacterium catenulatum has two subspecies that are distributed differently across adults and adolescents and have distinct functions [48]. Moreover, Bifidobacterium catenulatum——capable of fermenting non-digestible carbon sources—may produce substantial acetate in nutrient-limited environments, supporting acetate-dependent butyrate producers [49]. Butyrate, in turn, exerts anti-inflammatory effects by enhancing intestinal barrier function and modulating immune responses. Nonetheless, further studies are required to elucidate these potential mechanisms.

High chronic stress was associated with the change of several microbial species, such as Faecalibacterium and Phascolarctobacterium faecium, are known producers of SCFAs. SCFAs help regulate neuroinflammation, maintain gut barrier integrity, and influence brain function by stimulating enteroendocrine cells to release gut hormones (e.g., glucagon-like peptide 1, peptide YY) and neurotransmitters (e.g., gamma-aminobutyric acid), which signal to the brain via the circulation or vagus nerve [50, 51]. Altered SCFA levels have been linked to stress-related behaviors and mood disorders in both animals and humans. These findings suggest that microbial metabolites may mediate gut-brain communication under chronic stress. However, as this is a cross-sectional study, causal relationships cannot be established. Future longitudinal studies are needed to elucidate the directionality of these associations.

This study is the first to investigate the associations among chronic stress, gut microbiota, and metabolites in adolescents by integrating 16 S rRNA sequencing, metagenomics, and metabolomics. Our findings suggest that the brain-gut microbiome axis may play a role in chronic stress-related disorders in adolescents. However, our study also has several limitations. First, the generalizability of our findings is limited due to the small sample size and the specific characteristics of the study participants. The majority of participants were of Han ethnicity and exhibited relatively low chronic stress levels compared with those reported in previous studies [52]. Moreover, the majority of participants’ parents had low educational attainment. Second, chronic stress was measured by self-reported questionnaires, which may be subject to recall bias. However, the measurement of chronic stress lacks objective tools. We used the scale with the best current validity and reliability to minimize recall bias [14, 53]. Third, we did not adjust for physical activity in the analysis because of the small sample size. However, we found that there was no significant difference in physical activity across participants with different levels of chronic stress. Therefore, it can be considered that these factors may not affect the associations of interest. Fourth, we did not assess the pubertal stage of participants. Therefore, we could not evaluate how the pubertal stage might affect the associations between chronic stress and microbiota outcomes. Future research should incorporate measures such as the Tanner scale to assess pubertal development. Fifth, the causal associations between chronic stress, gut microbiota, and its metabolites could not be inferred due to the cross-sectional study design.

Conclusion

In conclusion, adolescents under high-chronic stress showed disturbances in their gut microbiota at the phylum, genus, and species levels, as well as in their metabolites. We identified five differential genera, four differential species, and 19 differential metabolites that were associated with at least one of the four differential species. The health effects of these bacteria may be explained by the altered metabolite. Our findings provide insights into potential links between the brain-gut microbiome axis and chronic stress-related conditions in adolescents.

Abbreviations

ASLEC

Adolescent Self-Rating Life Events Checklist

PCoA

Principal Coordinates Analysis

LC/MS

Liquid chromatography-mass spectrometry

LEfSe

Linear Discriminant Analysis Effect Size

KEGG

Kyoto Encyclopedia of Genes and Genomes

PLS-DA

Partial least squares discriminant analysis

OPLS-DA

Orthogonal partial least squares discriminant analysis

SCFAs

Short-chain fatty acids

LTC4

Leukotriene C4

Authors' contributions

YL: Formal analysis, Visualization, Writing original draft; YHW: Writing original draft; YFH: Formal analysis, Writing- review and editing; JL: Investigation; LND: Investigation; QZS: Investigation; CHY: Investigation; SXZ: Investigation; YXG: Methodology; MWS: Methodology; ZLC: Investigation; CCW: Investigation; ZHG: Investigation; XL: Writing-review and editing; LM: Conceptualization, Funding acquisition, Supervision, Writing-review and editing; LZ: Conceptualization, Funding acquisition, Supervision.

Funding

This study was supported by National Natural Science Foundation of China (82103868 (LM)), Key Research and Development Program of Shaanxi (2025SF-YBXM-364 (LM)), The Fundamental Research Funds for the Central Universities (xzy012025156 (LM)), Shaanxi Province Postdoctoral Research Project (2023BSHTBZZ03 (LM)), and Shaanxi Provincial Social Science Project (2023P098 (LM)), the National Natural Science Foundation of China (Grant number: 8191101420 (LZ)), Outstanding Young Scholars Funding (Grant number: 3111500001 (LZ)), Xi’an Jiaotong University Basic Research and Profession Grant (Grant number: xtr022019003 (LZ), xzy032020032 (LZ)), Xi’an Jiaotong University Young Talent Support Grant (Grant number: YX6 J004 (LZ)). The funders had no role in the study design, data collection, data analysis, interpretation, or writing of the manuscript.

Data availability

The raw data are deposited on the NCBI sequence read archive. (PRJNA1077629, PRJNA1077752)

Declarations

Ethics approval and consent to participate

The study was approved by the Institutional Review Board at Xi’an Jiaotong University Health Science Center (Approval No.: 2021 − 1389) and conducted in accordance with the ethical principles of the Declaration of Helsinki. Written informed consent was obtained from all adolescent participants and their parents/legal guardians, which specifically granted permission for fecal sample collection in both home and school settings.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.