Abstract
Objective
To investigate the association between oral microbiome characteristics and suicidal ideation (SI) in vocational high school students.
Methods
The oral microbiota composition and abundance of the SI group (N = 45) and the matched healthy control group (N = 45) were analyzed using 16S rRNA sequencing. Demographic variables and SI assessments (measured by the Columbia Suicide Severity Rating Scale) were obtained. Bioinformatics analyses (e.g., alpha and beta diversity indices, PCoA, NMDS, and LEfSe) were performed to compare oral microbiota characteristics between the SI group and the control group.
Results
The vocational high school students with SI exhibited significantly lower alpha diversity indices in oral microbiota compared to the control group, including PD whole tree (20.56 [19.45, 21.38] vs. 21.37 [20.17, 23.27]), Chao1 (621.82 ± 66.99 vs. 672.24 ± 68.52), and observed species (426.91 ± 54.59 vs. 466.42 ± 69.75) (all p values < 0.01). Beta diversity analysis revealed distinct microbial composition between groups ( p < 0.05), with greater homogeneity in the SI group. LEfSe analysis identified increased abundances of Enterobacter, Escherichia_Shigella, and Parabacteroides in the SI group, while controls showed enrichment of Rikenellaceae and Alistipes.
Conclusion
A decrease in oral microbiota diversity and alterations in the abundance of specific bacterial genera (Enterobacter, Escherichia_Shigella, and Parabacteroides) are associated with SI among vocational high school students. Longitudinal studies are needed to elucidate the underlying mechanisms and potential health impacts.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12903-026-08317-3.
Keywords: Vocational high school students, Suicidal ideation, Oral microbiota, Alpha diversity, Beta diversity
Introduction
Adolescent suicide has become a global public health crisis. The latest WHO report indicates that suicide has become the third leading cause of death among adolescents and young adults aged 15–29 years [1]. In China, the suicide mortality rate among adolescents aged 10–24 years was 3.40 per 100,000 population. In Hubei Province, the rate was approximately 3.33 per 100,000, with suicide ranking as the second leading cause of injury-related deaths among adolescents aged 10–19 years and the leading cause among those aged 20–24 years [2]. Suicidal ideation (SI), defined as a psychological state characterized by explicit suicide plans or persistent death wishes without overt suicidal behaviors, is a key precursor to suicidal behaviors [3]. Studies have confirmed a significant developmental continuity between SI and completed suicide (CS), underscoring the essential early detection value of SI for suicide prevention [4]. The lifetime prevalence of SI in China’s general population is 3.9% [5], while adolescents exhibit markedly higher rates of 15.7%–19.7% [6]. Notably, vocational high school students, as a special subgroup of adolescents, are often at a disadvantage in terms of social expectations and family background. They are more prone to anxiety, impulsive behaviors, and somatic symptoms, coupled with relatively lower self-esteem levels and diminished psychological adaptability and emotional regulation capacity, thus exhibiting a higher risk of SI compared with adolescents in general [7–9].
The oral cavity, harboring hundreds of microbial species on mucosal and dental surfaces, constitutes one of the most complex microbial ecosystems in the human body, maintaining dynamic equilibrium. Emerging evidence reveals that oral microbiota modulate central nervous system functions via the “oral-gut-brain axis” through neuroendocrine, immunoregulatory, and metabolic pathways, demonstrating significant associations with periodontitis, cardiovascular diseases, and neuropsychiatric disorders [10]. Although existing studies have confirmed characteristic microbial alterations in patients with depression, such as abnormal abundances of Neisseria and Prevotella nigrescens [11], research on the association between SI and oral microbiota remains scarce [12], particularly lacking targeted investigations in high-risk groups like vocational high school students. This study employs 16 S rRNA high-throughput sequencing technology to compare the salivary microbiome structures between vocational students with and without SI, systematically analyzes their diversity and core genus abundance differences, and aims to explore potential biological correlates between oral microbiome ecology and SI.
Materials and methods
Participants and sampling
This study was part of the baseline survey of a multi-center adolescent cohort study, which employed cluster sampling at a vocational high school in Wuhan, China, and started in March 2023. The inclusion criteria were as follows: (1) officially enrolled students; (2) continuous residency in Wuhan for ≥ 6 months; (3) written informed consent from both participants and legal guardians. Students with cognitive impairment or psychiatric history preventing independent questionnaire completion were excluded.
This study included 99 participants, of whom 49 reported SI and 50 did not. A case-control study design using propensity score matching (PSM) was subsequently employed. After matching, both the SI group and the control group comprised 45 individuals each. The study was approved by the Ethics Committee at the School of Medicine, Wuhan University. Clinical trial number: not applicable.
Data collection
Trained investigators then distributed self-administered questionnaires with detailed instructions to each student in the classroom. Emphasis was placed on the confidentiality of information, assuring students that all data would be used solely for research purposes, and encouraging honest completion. Investigators thoroughly explained the questionnaire requirements and collected the completed questionnaires. Concurrently, unstimulated saliva samples were uniformly collected at 3:00 PM for all participating students. Participants were instructed to abstain from eating or smoking for 60 min and avoid oral hygiene practices for 12 h prior to saliva collection. Saliva (2 mL) was collected using the OMNIgene·ORAL saliva collection kit and immediately stored at -80 °C for subsequent analysis.
Measures
Suicidal ideation
SI was assessed using the Suicidal Ideation subscale of the Columbia-Suicide Severity Rating Scale (C-SSRS), specifically the formally validated Chinese version [13]. The scale consists of four core items regarding SI: ‘Have you wished you were dead or wished you could go to sleep and never wake up?’, ‘Have you had any thoughts about killing yourself?’, ‘Have you thought about how you would do that or how you would make yourself not alive anymore (kill yourself)?’, and ‘When you thought about making yourself not alive anymore (or killing yourself), did you think that this was something you might actually do?’. Participants were categorized into the SI group if they responded “yes” to any of these four items; otherwise, they were assigned to the non-SI (control) group. In our sample, this Chinese version demonstrated good internal consistency, with a Cronbach’s α coefficient of 0.858.
Oral microbiota
Saliva samples were processed to extract microbial DNA using the StoolGen Kit, followed by verification via agarose gel electrophoresis and dilution. The V3-V4 region of the 16 S rRNA gene was PCR-amplified using 515 F/806R primers and 30 ng of DNA template with Phusion Master Mix on a T100 thermal cycler. Libraries were prepared using the TruSeq Nano DNA LT Kit, quality-checked by Agilent 2100 and Promega QuantiFluor, and subjected to paired-end sequencing on an Illumina MiSeq. Sequencing data were processed using an OTU-based bioinformatic pipeline. Paired-end reads were first merged using FLASH (v1.2.11) to generate full-length amplicon sequences. The merged sequences were then quality-filtered using the split_libraries script in QIIME (v1.8.0) to remove reads containing ambiguous bases or low-quality scores, yielding high-quality clean tags. Chimeric sequences were detected and removed using UCHIME (v2.4.2). The resulting valid tags were clustered into Operational Taxonomic Units (OTUs) at 97% sequence similarity using VSEARCH (v2.4.2). Taxonomy was assigned using the RDP classifier (confidence 0.8, Greengenes database), and unassigned OTUs were removed.
Demographic variables
General demographic information was collected using a self-designed questionnaire (see Supplementary Material for the English version of the questionnaire). The demographic variables included age (in years), gender (boys/girls), household registration (urban/rural/unclear), left-behind experience (no/yes), family economic status (good/average/poor), parental relationship (very harmonious/relatively harmonious/average/relatively tense/very tense), parenting style (free-range parenting/relatively democratic/relatively coercive), enrollment in a key class(yes/no), class leadership experience (yes/no), and class academic ranking (top 0%-10%/top 10%-20%/top 20%-30%/top 30%-40%/top 40%-50%/after 50%).
Statistical analysis
To mitigate potential confounding from demographic variables, Propensity Score Matching (PSM) was employed to balance the SI group and the control group. A one-to-one nearest-neighbor matching algorithm was utilized, based on propensity scores calculated from baseline covariates including age and sex. The matching process aimed to achieve a balanced distribution of these key variables between the two groups, ensuring comparability for subsequent analyses. The robustness of the matching results was further examined by testing different caliper widths (0.1, 0.2, and 0.3 times the standard deviation of the propensity score), which showed consistent balance across these settings.
Alpha diversity indices (PD whole tree, Chao1, Shannon, Simpson, observed species, Good’s coverage) were calculated using QIIME. Alpha diversity metrics are presented as mean ± standard deviation (SD) for normally distributed data or median (interquartile range, IQR) for non-normally distributed data. Normality of distribution was assessed using the Shapiro-Wilk test, and homogeneity of variance was evaluated using Levene’s test. For indices meeting the assumptions of normality and homogeneity of variance, group comparisons were performed using independent samples t-test. For indices violating these assumptions, the non-parametric Mann-Whitney U test (Wilcoxon rank-sum test) was applied. All tests were two-sided with a significance level of α = 0.05. Beta diversity was assessed via Principal Co-ordinates Analysis (PCoA) and Nonmetric Multidimensional Scaling (NMDS) based on unweighted UniFrac distance. Statistical significance of between-group differences was evaluated by PERMANOVA with 999 permutations. Homogeneity of multivariate dispersion was assessed using PERMDISP with 999 permutations to confirm that observed differences were not driven by group dispersion heterogeneity. Additionally, clustering analysis was performed using the Unweighted Pair Group Method with Arithmetic Mean (UPGMA) based on unweighted UniFrac distance to group samples according to their similarities or distances. The Linear Discriminant Analysis Effect Size (LEfSe) method was applied to analyze the microbial groups with statistically significant differences in relative abundance between groups (LDA score > 2.0 and p < 0.05). All statistical analyses were completed using R (version 4.5.0, The R Foundation).
Results
Characteristics of the study sample
Table 1 shows the basic characteristics of the samples after propensity score matching. The gel electrophoresis results of saliva samples from 99 participants are shown in Supplementary Figure S1. The basic characteristics of the samples before matching are shown in Supplementary Table S1, and the matching balance is shown in Supplementary Figure S2. The results showed that among the 90 participants included in the study, the average age was 17.23 (0.45) years, 11.11% were boys and 88.89% were girls. More than half (65.56%) were rural household registration. 26.67% had left-behind experiences, 71.11% were average family economy, 62.22% reported very or relatively harmonious parental relationships, and 81.11% experienced democratic parenting styles. In terms of education, 87.78% were not admitted to key class, 73.33% had no class leadership experience, and 28.89% had academic performance ranking in the bottom 50% of their grade.
Table 1.
Characteristics of the study sample
| Variable | Case group | Control group | Total |
|---|---|---|---|
| Total | 45 (50.00) | 45 (50.00) | 90 (100.00) |
| Age | 17.24 (0.43) | 17.22 (0.47) | 17.23 (0.45) |
| Gender | |||
| Boys | 5 (11.11) | 5 (11.11) | 10 (11.11) |
| Girls | 40 (88.89) | 40 (88.89) | 80 (88.89) |
| Household Registration | |||
| Urban | 12 (26.67) | 12 (26.67) | 24 (26.67) |
| Rural | 28 (62.22) | 31 (68.89) | 59 (65.56) |
| Unclear | 5 (11.11) | 2 (4.44) | 7 (7.78) |
| Left-behind Experience | |||
| No | 30 (66.67) | 36 (80.00) | 66 (73.33) |
| Yes | 15 (33.33) | 9 (20.00) | 24 (26.67) |
| Family Economic Status | |||
| Good | 5 (11.11) | 10 (22.22) | 15 (16.67) |
| Average | 32 (71.11) | 32 (71.11) | 64 (71.11) |
| Poor | 8 (17.78) | 3 (6.67) | 11 (12.22) |
| Parental Relationship | |||
| Very Harmonious | 6 (13.33) | 21 (46.67) | 27 (30.00) |
| Relatively Harmonious | 13 (28.89) | 16 (35.56) | 29 (32.22) |
| Average | 10 (22.22) | 6 (13.33) | 16 (17.78) |
| Relatively Tense | 11 (24.44) | 2 (4.44) | 13 (14.44) |
| Very Tense | 5 (11.11) | 0 (0.00) | 5 (5.56) |
| Parenting Style | |||
| Free-range parenting | 4 (8.89) | 0 (0.00) | 4 (4.44) |
| Relatively Democratic | 29 (64.44) | 44 (97.78) | 73 (81.11) |
| Relatively Coercive | 12 (26.67) | 1 (2.22) | 13 (14.44) |
| Key Class | |||
| Yes | 5 (11.11) | 6 (13.33) | 11 (12.22) |
| No | 40 (88.89) | 39 (86.67) | 79 (87.78) |
| Class leadership experience | |||
| Yes | 10 (22.22) | 14 (31.11) | 24 (26.67) |
| No | 35 (77.78) | 31 (68.89) | 66 (73.33) |
| Class Academic Ranking | |||
| Top 0%-10% | 1 (2.22) | 7 (15.56) | 8 (8.89) |
| Top 10%-20% | 2 (4.44) | 5 (11.11) | 7 (7.78) |
| Top 20%-30% | 3 (6.67) | 7 (15.56) | 10 (11.11) |
| Top 30%-40% | 3 (6.67) | 8 (17.78) | 11 (12.22) |
| Top 40%-50% | 17 (37.78) | 11 (24.44) | 28 (31.11) |
| After 50% | 19 (42.22) | 7 (15.56) | 26 (28.89) |
Continuous variables are presented as mean ± standard deviation, and categorical variables are presented as frequency and percentage
Differences in alpha diversity between case and control group
The rarefaction curves in Supplementary Figure S3 show that the detection of new species increases rapidly at low sequencing depths. After a turning point around 20,000 reads, the growth rate slows down, indicating that the main microbial communities have been largely covered and the sample diversity tends to stabilize. In Supplementary Figure S4, both the Chao1 and Shannon curves tend to level off, suggesting that the current sequencing depth is sufficient to comprehensively reflect species richness and diversity, and further increasing the sequencing volume would not be meaningful. Good’s coverage values were high in both groups (Case: 0.9978; Control: 0.9976), confirming that the sequencing depth was adequate to capture the majority of microbial diversity in all samples.
Results in Table 2 show that the alpha diversity indices, including PD whole tree, Chao1, Observed-species, and Goods-coverage were differed significantly between the case and control groups. The PD whole tree index was lower in the case group than the control group (20.56 [19.45, 21.38] vs. 21.37 [20.17, 23.27], p = 0.003). The Chao1 index (621.82 (66.99) vs. 672.24 (68.52), p = 0.001) and Observed species (426.91 (54.59) vs. 466.42 (69.75), p = 0.004) were significantly lower in the case group. No significant differences in Shannon index and Simpson index were found between the two groups. As shown in Fig. 1, alpha diversity indices (PD whole tree, Chao1, and Observed species) were significantly lower in the case group compared to the control group, while no significant differences were observed in Shannon and Simpson indices.
Table 2.
Differences of alpha diversity indices between the case and the control group, mean (SD)
| Alpha index | Case group | Control group | t / Z | p |
|---|---|---|---|---|
| PD_whole_treea | 20.56 (19.45, 21.38) | 21.37 (20.17, 23.27) | -3.01 | 0.003 |
| Chao1 | 621.82 (66.99) | 672.24 (68.52) | -3.53 | < 0.001 |
| Shannon | 4.89 (0.56) | 4.79 (0.73) | 0.68 | 0.497 |
| Simpsona | 0.92 (0.89, 0.93) | 0.92 (0.87, 0.94) | 0.28 | 0.781 |
| Observed-species | 426.91 (54.59) | 466.42 (69.75) | -2.99 | 0.004 |
Superscript “a” indicates that the index did not meet the normality assumption and was analyzed using the Wilcoxon rank-sum test
Bold values indicate statistical significance
Fig. 1.
Comparison of alpha diversity indices between Case and Control groups. Note: Violin plots show the distribution of (a) PD whole tree, (b) Chao1, (c) Shannon, (d) Simpson, and (e) Observed species. Statistical significance was evaluated by Wilcoxon rank-sum test or t-test as appropriate. **p < 0.01, ***p < 0.001
Differences in beta diversity between case and control group
PCoA based on unweighted UniFrac distances (Fig. 2a) showed partial separation between groups, with PCoA1 and PCoA2 explaining 7.66% and 6.34% of the total variation, respectively. Considerable overlap was observed, though case group samples showed a tendency to cluster more densely. The NMDS analysis in Fig. 2b (stress value = 0.166) revealed a similar pattern of community distribution between the two groups. A stress value below 0.2 indicates biological relevance. PERMANOVA analysis based on unweighted UniFrac distances detected a statistically significant but modest between-group difference (R² = 0.047, F = 4.366, p = 0.001, 999 permutations), with group membership explaining approximately 4.7% of the total variation in community structure. Homogeneity of multivariate dispersion was confirmed by PERMDISP (F = 0.654, p = 0.410, 999 permutations). Hierarchical clustering (UPGMA) based on unweighted UniFrac distances (Fig. 2c) provided a descriptive visualization of sample relationships, showing a general tendency for case group samples to cluster together, though with considerable mixing between groups.
Fig. 2.
Beta diversity analysis comparing the microbial communities between the case group and the control group. Note: (a) PCoA; (b) NMDS; (c) UPGMA clustering
Differences in microbial community composition between the case and control group
Figure 3 presents the distribution of LDA scores from LEfSe analysis comparing the case group versus the control group. The specific data can be found in Supplementary Table S2. The results suggest differences in relative abundance of certain taxa between the case and control groups at multiple taxonomic levels. Taxa with higher relative abundance in the case group mainly included the Enterobacteriales and Bacteroidales orders, along with their respective families and genera such as Enterobacteriaceae, Escherichia-Shigella, and Parabacteroides, with LDA scores exceeding the threshold of 4.0 Taxa with higher relative abundance in the control group included Rikenellaceae, and Alistipes, among others.
Fig. 3.
LDA scores identifying differentially abundant taxa between case and control groups
Figure 4 shows the taxonomic distribution of differential microbial taxa based on a cladogram. The cladogram illustrates the taxonomic levels at which differential abundance was observed, with Enterobacteriales and related taxa more abundant in the case group, while Rikenellaceae and Lachnospiraceae were more abundant in the control group.
Fig. 4.
Cladogram of differentially abundant operational taxonomic units between the case group and the control group. Note: Different colors represent different groups. Red nodes indicate species with significantly higher relative abundance in the case group, while green nodes indicate species with significantly higher relative abundance in the control group. Yellow nodes represent species with no significant differences between the two groups. The diameter of each node is proportional to its relative abundance. The nodes are arranged in layers from the inside out, representing phylum/class/order/family/genus, respectively
Discussion
This study found that the species richness of oral microbiota in vocational high school students with SI was significantly lower than that in the control group, as evidenced by the decreased Chao1 and Observed_species indices, as well as the reduced phylogenetic diversity indicated by the PD_whole_tree index. These results suggest that SI may be associated with a decline in the species richness and phylogenetic diversity of oral microbiota. Additionally, beta-diversity analysis revealed significant differences in the composition of oral microbiota between the two groups. Compared with the study by Ahrens [12], no significant differences were found in the Shannon and Simpson indices in this study, indicating similar evenness of microbial communities between the groups. This may be attributed to the fact that the subjects in this study were from the same vocational school, with uniform schedules and diets, which reduced the interference of confounding factors. Therefore, the microbial diversity changes associated with SI may primarily manifest in species richness rather than evenness. Unlike the findings of Chen [14] in patients with major depressive disorder (MDD), this study observed significant differences in alpha diversity between the SI group and the control group. This discrepancy may be due to differences in sample types, as oral microbiota may be more sensitive to diversity changes, whereas gut microbiota is relatively stable. Moreover, vocational high school students, as a distinct adolescent population, experience different psychological stress and social environments compared with college students and MDD patients, which may contribute to the distinct patterns of microbial diversity. The pathological mechanisms of SI and depression overlap but are not identical, which may also account for the differences in microbial diversity.
Differences in the abundance of specific oral microbiota are key features distinguishing vocational high school students with SI from those without. Through LEfSe analysis, this study identified differentially abundant microbial taxa at the genus level between the SI group and the control group. Enterobacter, Escherichia_Shigella, and Parabacteroides were significantly enriched in the SI group, whereas Rikenellaceae and Alistipes were significantly enriched in the control group. Compared with the study by Ahrens et al., no significant differences were found in Prevotella in this study. However, the significant enrichment of Enterobacter and Escherichia_Shigella in the SI group was observed, which is consistent with the inflammation hypothesis that the overgrowth of certain taxa may induce inflammatory responses, thereby affecting the central nervous system and increasing the risk of suicide [15]. Chen found that Alloprevotella rava was absent in individuals with SI. Gao [16] studied the gut microbiota of patients with depression and found that the levels of beneficial genera decreased while the levels of potentially pathogenic genera increased. This is similar to the decreased Paeniclostridium and increased Enterobacteriaceae observed in this study, suggesting that the reduction of beneficial bacteria and the increase of pathogenic bacteria may be associated with SI.
Oral microbiota plays a significant role in human microbial ecology and may influence suicidal behaviors in vocational high school students through various biological mechanisms. This study found that the alpha diversity of oral microbiota in the SI group was significantly lower than that in the control group, indicating a close association between SI and decreased microbial diversity. Reduced microbial diversity is generally considered a sign of microbial dysbiosis [17], which is not only associated with an increased risk of various diseases [18], but may also impair the immune regulation, nutritional metabolism, and barrier functions of oral microbiota. As a result, the microbiota may become more susceptible to dysbiosis under stressful events, exacerbate inflammatory responses, affect mental health, and increase the risk of suicide. Furthermore, this study found a significant increase in the abundance of Enterobacteriaceae in the SI group among vocational high school students. Enterobacteriaceae can produce pro-inflammatory substances such as lipopolysaccharides, which induce the release of pro-inflammatory cytokines (e.g., IL-6, TNF-α) and trigger systemic inflammatory responses [19]. These inflammatory responses are not only closely related to depressive symptoms but may also affect brain function through neuroinflammatory pathways, thereby increasing the risk of suicide [20–22]. The increase in Enterobacteriaceae may also influence psychological states through the gut-brain axis, creating a negative feedback loop [23]. Additionally, this study observed a significant decrease in the abundance of Paeniclostridium, a butyrate-producing genus, in the SI group. Butyrate, as a short-chain fatty acid with anti-inflammatory and neuroprotective properties, may reduce the risk of SI by regulating the synthesis of neurotransmitters [24]. Moreover, the trigeminal nerve, which connects the oral cavity and the brain directly, is also considered a potential pathway through which oral microbiota may affect mental health. Dysbiosis of oral microbiota may alter the brain’s response to stress and emotions through the trigeminal nerve, thereby exacerbating the formation of SI [25]. In summary, oral microbiota may be associated with suicidal behaviors in vocational high school students through multiple mechanisms, including inflammation, metabolic products, the gut-brain axis, and neural connections. These findings provide a scientific basis for the development of targeted intervention strategies.
Limitations
This study has limitations. First, the cross-sectional design employed in this study does not allow for a definitive inference of causality. Future research should attempt longitudinal designs to track the dynamic relationship between changes in oral microbiota and the development of SI among vocational high school students, thereby verifying the causal relationship. Second, the limited sample size included in this study may affect the generalizability of the results. Future validation in a broader population is needed. Third, propensity score matching was performed based on age and sex only, due to data availability constraints. We acknowledge that this approach may not fully account for other potential confounding variables, particularly psychosocial factors (e.g., depression/anxiety symptoms) and oral-health–related behaviors (e.g., tooth-brushing frequency, dietary sugar intake, smoking status, recent antibiotic use). Residual imbalance in socioeconomic and family-related variables remained after matching, as shown in Table 1. These unmeasured confounders may have influenced both SI and oral microbiota composition, and their effects cannot be ruled out in our analysis. Lastly, this study only analyzed the differential microbial taxa between the SI group and the control group. Future research should incorporate metabolomics and inflammatory marker analysis to further explore the potential mechanisms through which oral microbiota affects adolescent mental health.
Conclusion
The findings revealed a significant association between reduced oral microbiota diversity and SI among vocational high school students. Specifically, the SI group exhibited significantly higher abundances of Enterobacter, Escherichia_Shigella, and Parabacteroides. Further longitudinal studies are warranted to elucidate the underlying mechanisms and long-term health implications.
Supplementary Information
Acknowledgements
We thank all individuals who took part in the survey.
Authors’ contributions
YZ, CX, ZZ: Designed the study, conducted the data analysis and wrote the first draft. YZ, LX, LZ, KQ: conducted the data analysis and wrote the first draft. HY: Review the manuscript. BY: Designed the study and reviewed the manuscript.
Funding
The study was supported by the National Natural Science Foundation of China (#82204073, PI: Dr. Bin Yu), The Ministry of Education of Humanities and Social Science project (#22YJC840037, PI: Dr. Bin Yu), the Wuhan Health Science Research Funding (#WY22B09, PI: Dr. Bin Yu).
Data availability
Data are available by request from the researcher.
Declarations
Ethics approval and consent to participate
The study was approved by the Ethics Committee at School of Medicine, Wuhan University. Participants or the public were not involved in the study design, conduct, reporting or dissemination of our research. The study was conducted in accordance with the Declaration of Helsinki (of 1975, revised in 2013). Informed consent was obtained from all individual participants included in the study. For any participant under the age of 16, informed consent to participate was obtained from their parents or legal guardians.
Consent for publication
Written informed consent for publication was not required because this manuscript does not contain any individual person’s identifiable data.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Yuanzhong Zhu, Chenchang Xiao and Zilei Zhang contributed equally to this work.
Contributor Information
Bin Yu, Email: binyu1029@whu.edu.cn.
Hong Yan, Email: yanhmjxr@whu.edu.cn.
References
- 1.World Health Organization. Mental health of adolescents. [updated 2024/10/10; cited 2025/3/6]. Available from: https://www.who.int/news-room/fact-sheets/detail/adolescent-mental-health.
- 2.Pan J, Zhou M, Zhang L, Tang Y, Li Q, He T. Suicide mortality analysis in adolescents aged 10–24 years in Hubei, 2019. Disease Surveillance [Journal Article]. 2021;36(10):1086–91. [Google Scholar]
- 3.Goodfellow B, Kolves K, De Leo D. Contemporary Classifications of Suicidal Behaviors. Crisis [Journal Article; Syst Review]. 2020;41(3):179–86. [DOI] [PubMed] [Google Scholar]
- 4.Brent DA, Emslie GJ, Clarke GN, Asarnow J, Spirito A, Ritz L, et al. Predictors of spontaneous and systematically assessed suicidal adverse events in the treatment of SSRI-resistant depression in adolescents (TORDIA) study. Am J Psychiatry. 2009;166(4):418–26. [Journal Article; Multicenter Study; Randomized Controlled Trial]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Cao XL, Zhong BL, Xiang YT, Ungvari GS, Lai KY, Chiu HF, et al. Prevalence of suicidal ideation and suicide attempts in the general population of China: A meta-analysis. Int J Psychiatry Med. 2015;49(4):296–308. [Journal Article; Meta-Analysis; Research Support, N.I.H., Extramural; Research Support, U.S. Gov’t, P.H.S.]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Hu J, Dong Y, Chen X, Liu Y, Ma D, Liu X, et al. Prevalence of suicide attempts among Chinese adolescents: A meta-analysis of cross-sectional studies. Compr Psychiatry [Journal Article; Meta-Analysis]. 2015;61:78–89. [DOI] [PubMed] [Google Scholar]
- 7.Carli V, Wasserman C, Wasserman D, Sarchiapone M, Apter A, Balazs J, et al. The saving and empowering young lives in Europe (SEYLE) randomized controlled trial (RCT): methodological issues and participant characteristics. BMC Public Health. 2013;13:479. [Journal Article; Multicenter Study; Randomized Controlled Trial; Research Support, Non-U.S. Gov’t]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Wasserman D, Carli V, Wasserman C, Apter A, Balazs J, Bobes J, et al. Saving and empowering young lives in Europe (SEYLE): a randomized controlled trial. BMC Public Health. 2010;10:192. [Journal Article; Randomized Controlled Trial; Research Support, Non-U.S. Gov’t]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Swahn MH, Ali B, Bossarte RM, Van Dulmen M, Crosby A, Jones AC, et al. Self-harm and suicide attempts among high-risk, urban youth in the U.S.: shared and unique risk and protective factors. Int J Environ Res Public Health. 2012;9(1):178–91. [Journal Article; Research Support, U.S. Gov’t, P.H.S.]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Bui FQ, Almeida-da-Silva C, Huynh B, Trinh A, Liu J, Woodward J, et al. Association between periodontal pathogens and systemic disease. Biomed J [Journal Article; Review]. 2019;42(1):27–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Wingfield B, Lapsley C, McDowell A, Miliotis G, McLafferty M, O’Neill SM, et al. Variations in the oral microbiome are associated with depression in young adults. Sci Rep [Journal Article]. 2021;11(1):15009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Ahrens AP, Sanchez-Padilla DE, Drew JC, Oli MW, Roesch L, Triplett EW. Saliva microbiome, dietary, and genetic markers are associated with suicidal ideation in university students. Sci Rep. [Journal Article; Research Support, Non-U.S. Gov’t; Research Support, U.S. Gov’t, Non-P.H.S.]. 2022;12(1):14306. [DOI] [PMC free article] [PubMed]
- 13.Ji YJ, Reliability. Validity and Application of the Columbia-Suicide Severity Rating Scale in Patients with Major Depressive Disorder [master’s thesis]. Guangzhou: Southern Medical University; 2024. (in Chinese).
- 14.Chen VC, Wu S. An exploratory analysis on the association between suicidal ideation and the microbiome in patients with or without major depressive disorder. J Affect Disord [Journal Article]. 2025;370:362–72. [DOI] [PubMed] [Google Scholar]
- 15.Deinzer R, Ruttermann S, Mobes O, Herforth A. Increase in gingival inflammation under academic stress. J Clin Periodontol [Journal Article]. 1998;25(5):431–3. [DOI] [PubMed] [Google Scholar]
- 16.Gao M, Wang J, Liu P, Tu H, Zhang R, Zhang Y et al. Gut microbiota composition in depressive disorder: a systematic review, meta-analysis, and meta-regression. Transl Psychiatry. [Journal Article; Meta-Analysis; Systematic Review]. 2023;13(1):379. [DOI] [PMC free article] [PubMed]
- 17.Shade A. Diversity is the question, not the answer. Isme J [Journal Article]. 2017;11(1):1–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Lozupone CA, Stombaugh JI, Gordon JI, Jansson JK, Knight R. Diversity, stability and resilience of the human gut microbiota. Nat [Journal Article; Review]. 2012;489(7415):220–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Maes M, Kubera M, Leunis J, Berk M. Increased IgA and IgM responses against gut commensals in chronic depression: further evidence for increased bacterial translocation or leaky gut. J Affect Disord [Journal Article]. 2012;141(1):55–62. [DOI] [PubMed] [Google Scholar]
- 20.Dantzer R, O’Connor JC, Freund GG, Johnson RW, Kelley KW. From inflammation to sickness and depression: when the immune system subjugates the brain. Nat Rev Neurosci. 2008;9(1):46–56. Journal Article; Research Support, N.I.H., Extramural; Review. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Dowlati Y, Herrmann N, Swardfager W, Liu H, Sham L, Reim EK, et al. A meta-analysis of cytokines in major depression. Biol Psychiatry. 2010;67(5):446–57. [Journal Article; Meta-Analysis; Research Support, Non-U.S. Gov’t]. [DOI] [PubMed]
- 22.Berk M, Williams LJ, Jacka FN, O’Neil A, Pasco JA, Moylan S et al. So depression is an inflammatory disease, but where does the inflammation come from? Bmc Med. [Journal Article; Research Support, Non-U.S. Gov’t; Review]. 2013;11:200. [DOI] [PMC free article] [PubMed]
- 23.Iwauchi M, Horigome A, Ishikawa K, Mikuni A, Nakano M, Xiao J, et al. Relationship between oral and gut microbiota in elderly people. Immun Inflamm Dis. 2019;7(3):229–36. [Journal Article; Research Support, Non-U.S. Gov’t]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Strandwitz P. Neurotransmitter modulation by the gut microbiota. Brain Res. [Journal Article; Research Support, N.I.H., Extramural; Review]. 2018;1693(Pt B):128 – 33. [DOI] [PMC free article] [PubMed]
- 25.Furness JB. The enteric nervous system and neurogastroenterology. Nat Rev Gastroenterol Hepatol. 2012;9(5):286–94. [Journal Article; Research Support, Non-U.S. Gov’t; Review]. [DOI] [PubMed]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data are available by request from the researcher.