Skip to main content
Download PDF

Potential roles of cigarette smoking on gut microbiota profile among Chinese men

BMC Medicine volume 23, Article number: 25 (2025) Cite this article

Abstract

Background

Cigarette smoking is posited as a potential factor in disrupting the balance of the human gut microbiota. However, existing studies with limited sample size have yielded inconclusive results.

Methods

Here, we assessed the association between cigarette smoking and gut microbial profile among Chinese males from four independent studies (N total = 3308). Both 16S rRNA and shotgun metagenomic sequencing methods were employed, covering 206 genera and 237 species. Microbial diversity and abundance were compared among non-smokers, current smokers, and former smokers.

Results

Actinomyces[g], Atopobium[g], Haemophilus[g], Turicibacter[g], and Lachnospira[g] were found to be associated with smoking status (current smokers vs. non-smokers). Metagenomic data provided a higher resolution at the species level, particularly for the Actinomyces[g] branch. Additionally, serum γ-glutamylcysteine (γ-Glu-Cys) was found to have a potential role in connecting smoking and Actinomyces[g]. Furthermore, we revealed putative mediation roles of the gut microbiome in the associations between smoking and common diseases including cholecystitis and type 2 diabetes.

Conclusions

We characterized the gut microbiota profile in male smokers and further revealed their potential involvement in mediating the impact of smoking on health outcomes. These findings advance our understanding of the intricate association between cigarette smoking and the gut microbiome.

Peer Review reports

Background

The human gut harbors a complex and dynamic community of microorganisms, known as the gut microbiota, estimated to exceed 100 trillion [1]. The gut microbiota has co-evolved with its host, forming an intricate and mutually beneficial relationship that intimately influence the overall health of the host [2, 3]. In turn, the host conditions (like genetic background [4] and lifestyle habits [5]) also shape the composition and functionality of the gut microbiota.

Cigarette smoking is considered a major behavioral risk factor for various health outcomes [6]. Recent evidence has suggested that this might involve alterations in the gut microbiota [7,8,9]. For instance, smoking-induced dysbiosis of the gut microbiota played a pro-tumorigenic role in colorectal cancer [9]. Over the past decade, there has appeared a growing focus on understanding the connection between smoking and the gut microbiota [10, 11]. Two recent reviews have diligently compiled existing knowledge from rodent experiments, shedding light on the mechanistic effects of cigarette smoke on intestinal microbiota dysbiosis [12, 13]. This intricate procedure may involve a cascade of events, including elevated lipid peroxidation, pro-inflammatory response, as well as regulation of the intestinal microenvironment. Emerging population-level evidence has prompted another comprehensive review of 13 existing studies which mostly conducted among European Americans [14]. In addition to these 13 studies on healthy subjects, there were also a number of studies conducted in specific populations, including patients with coronary artery disease and hypertension [15,16,17,18,19]. In most of above studies, a consistent trend of decreased microbial diversity was observed among smokers [20,21,22,23,24,25]. However, there existed substantial interstudy heterogeneity for specific microbes, especially at the genus level. For example, three studies [22, 26, 27] reported a higher level of Bacteroides[g] in smokers, whereas two other studies [20, 21] found its enrichment in non-smokers. Generally, the existing studies encountered common challenges, including (1) small sample sizes (only three studies included more than 300 subjects), (2) insufficient consideration for confounding factors, particularly sex (only five studies had statistical adjustments), and (3) poor reproducibility of results across studies.

In our prior endeavor, we sought to establish a causal connection between tobacco smoking and gut microbiota using two-sample Mendelian randomization (MR) based on publicly available Genome-Wide Association Studies (GWASs) [28]. This analysis identified the impact of smoking on 16 taxa, exemplified by a reduction in the abundance of Haemophilus[g]. However, it is important to note that these findings are somewhat constrained to the European population. Promisingly, advancements in sequencing technology have facilitated the ongoing accumulation of microbiome data within the Chinese population. The integration of these data with large-scale cohort studies, exemplified by the Westlake Gut Project [29], has provided a unique opportunity to explore the potential role of the gut microbiota in mediating associations between various exposures (e.g., cigarette smoking) and human health in China.

In the present study, our primary purpose was to conduct a multi-study investigation into the gut microbial profile associated with cigarette smoking among only Chinese males, duo to the low proportion of Chinese female smokers (~ 5%). The study integrated microbial data of a total of 3308 participants from four independent studies: Zhejiang Metabolic Syndrome Cohort (ZMSC), Guangzhou Nutrition and Health Study (GNHS), Guangdong Gut Microbiome Project (GGMP), and Sir Run Run Shaw Hospital Study (SRRSHS). In addition, our preliminary investigation extended into the potential relationships of the identified smoking-associated genera with health-related outcomes and smoking-related metabolites. This study seeks to improve our comprehension of the intricate interplay among cigarette smoking, the gut microbiome, metabolites, and health outcomes.

Methods

Description of studies and participants

The current analysis was based on the four separate studies, with ZMSC (n = 1781, 2020 survey) and SRRSHS (n = 215) conducted in Zhejiang province, as well as GNHS (n = 4048, 2008-2013 baseline) and GGMP (n = 7009) in Guangdong province (Fig. 1). A detailed description regarding the study design, recruitment and selection criteria of participants, stool sample collection, and sequencing procedures for ZMSC [30], GNHS [31], GGMP [32], and SRRSHS [26] has been previously documented and was also collated in the Supplementary Methods (Additional File 1). Briefly, ZMSC and GNHS are two community-cluster prospective cohorts, while GGMP and SRRSHS follow the cross-sectional design. A subset of participant pool, including 1382 from ZMSC, 1933 from GNHS, 7009 from GGMP, and 116 from SRRSHS, underwent 16S rRNA sequencing on stool samples. Shotgun metagenomic data were also available in ZMSC and GNHS. Given the substantial gender imbalance between smokers and non-smokers, female participants were excluded in the final analysis. In addition, we also excluded recent antibiotic and aspirin users, former smokers or non-smokers with secondhand smoke exposure, and those with missing data on age and body mass index (BMI). A portion of ZMSC participants from the 2015 survey underwent widely targeted metabolomic measurements, involving 1912 blood metabolites. Health-related outcomes for the majority of ZMSC participants were followed up through the linkage with regional medical health records since 2010.

Fig. 1
figure 1

Overview of study design and analyses. We ultimately incorporated 3308 male participants from four independent studies: ZMSC (n = 383), GNHS (n = 554), GGMP (n = 2255), and SRRSHS (n = 116). These individuals underwent a smoking survey and provided stool samples for 16S rRNA and/or shotgun metagenomic sequencing. The analytical strategy was multifaceted, involving (1) a comparison of the gut microbial diversity among three groups (non-smokers, current smokers, and former smokers), (2) the identification of smoking-associated gut microbes at the genus/species/pathway level, and (3) the exploration of associations between the identified genera and smoking-associated blood metabolites, as well as health-related outcomes

Full size image

Metadata collection and definition of smoking status

Demographic information for participants in all four studies was collected via the questionnaire. Individuals were categorized into different groups based on their smoking status: non-smokers, current smokers, and former smokers. In both ZMSC and GNHS, individuals classified as current smokers were those who have smoked at least 100 cigarettes throughout their lifetime and were still smoking, whereas former smokers were defined as those who have refrained from smoking for a minimum of six months. Within GGMP, respondents indicating “currently smokes, every day” and “currently smokes, but not every day” were categorized as current smokers, while those who stated “never smoked” and “has quit smoking” fell into the non-smoker and former smoker categories, respectively. In SRRSHS, current smokers were individuals with a history of continuous smoking for six months or longer, and information on smoking cessation was not available. In the first three studies, current smokers were additionally asked various smoking metrics, such as the duration of smoking in years and the number of cigarettes smoked per day. In ZMSC and GNHS, former smokers also reported the number of years they had quitting smoking.

Stool sample collection, DNA extraction, and 16S rRNA/shotgun metagenomic sequencing

The collection of stool samples occurred either on-site or at participants’ homes, with subsequent swift transportation to the research laboratory utilizing a cold-chain vehicle, followed by storage in − 80 °C freezers until further processing. Stool samples not reaching the designated collection point within the specified time frame were discarded. The extraction of total bacterial DNA was performed using suitable fecal DNA isolation kits. For 16S rRNA analysis, the V3 and/or V4 variable regions were amplified using corresponding barcoded primers and subsequently subjected to sequencing. For shotgun metagenomic analysis, performed in ZMSC and GNHS, sequencing was conducted on the Illumina HiSeq platform (Illumina Inc., CA, USA) using the 2 × 150-bp paired-end read protocol. Detailed information on stool sample collection, processing, and 16S rRNA/shotgun metagenomic sequencing of each study is provided in Table S1 (Additional File 1).

Taxonomic profiling

Detailed information on bioinformatics analysis of the microbial sequencing data could be found in the previous publications [26, 32,33,34,35] and Supplementary Methods (Additional File 1). For 16S rRNA data, the QIIME2 software (Quantitative Insights into Microbial Ecology 2) [36] was employed for downstream analysis, including merging paired reads, filtering reads, de-noising to obtain representative sequences in the form of Amplicon Sequence Variants (ASVs, in ZMSC and GNHS) or Operational Taxonomic Units (OTUs, in GGMP and SRRSHS), and subsequently taxonomic assignment. In the taxonomic assignment, the Silva database [37] was used as a reference for annotation in ZMSC, GNHS and SRRSHS, while GreenGenes database [38] was used in GGMP. For metagenomic data, consistent across ZMSC and GNHS, we conducted data quality control using PRINSEQ (version 0.20.4) [39]. Afterward, the metagenomic taxonomy was employed by MetaPhlAn2 (version 2.6.02) with default settings [40]. The functional profiling was carried out using HUMAnN2 (version 2.8.1) [41], which reconstructed microbial pathway according to MetaCyc metabolic pathway database [42, 43].

Genera (16S rRNA) and species (metagenome) with a prevalence of less than 10% in participants and an average relative abundance below 0.01% were excluded. This led to the inclusion of 133 genera in ZMSC, 102 genera in GNHS, 81 genera in GGMP, and 99 genera in SRRSHS for subsequent analyses (Additional file 1: Fig. S1; Additional file 2: Table S3-S6). It is important to note that applying the same criterion to species (metagenome) with lower relative abundance might be excessively stringent. Therefore, we conducted a parallel analysis on species without the restriction of requiring an average relative abundance greater than 0.01%. Microbial pathways (metagenome) were excluded solely based on their prevalence of less than 10%.

Statistical analysis

An outline of the study’s analytical strategy and methodology is depicted in Fig. 1. The analyses were structured into three steps.

The comparison of the gut microbial diversity (step 1)

The alpha diversity at the genus/species level was estimated to reflect the richness and evenness of the gut microbiota within a participant. Indicators included Richness (number of genera/species), Shannon index, Simpson index, and Pielou index. The differences of four alpha diversity parameters between groups with different smoking status were evaluated using a Wilcoxon test in each study. Beta diversity, quantified by the Bray-Curtis distance between participants, was estimated based on the relative abundance matrices of genera/species and visualized using principal coordinate analysis (PCoA) plots. Permutational multivariate analysis of variance (PERMANOVA) with 999 permutations was applied to discern whether and to what extent the dissimilarity in the beta diversity could be attributed to cigarette smoking, with adjustments for age and BMI.

The identification of smoking-associated microbes and pathways (step 2)

To identify the gut microbe at the genus level, we employed Microbiome Multivariable Association with Linear Models (MaAsLin2) in each study, a method demonstrated to be considerably effective in analyzing microbial data [44]. The relative abundance of genera was modeled as the dependent variable in log-transformed form, while three smoking statuses as the independent variables in comparison with each other. Before log-transformation, zeros in microbial data were substituted with a small value (one-tenth of the minimum abundance of the corresponding genera), which had a negligible impact on the data distribution. The models were adjusted for age and BMI across all studies. Following this, a meta-analysis with the fixed-effects model was conducted to integrate the MaAsLin2’s results of the four studies. Given the execution of high-dimensional tests, we applied the Bonferroni correction (totally 206 genera), deeming associations with a Bonferroni-corrected p less than 0.05 as statistically significant. The methods mentioned above were repeated for metagenomic data at the species level and for microbial pathway analysis.

To verify the robustness of the main results, we conducted sensitivity analyses that included (1) age-stratified analysis (≤ 60 years and > 60 years), (2) further adjustment for alcohol consumption, (3) the adoption of the random-effects model in meta-analysis, (4) the use of smoking information from the 2015 survey in ZMSC, and (5) a comparison of main results with those obtained by excluding subjects with common smoking-related diseases (stroke, coronary heart disease, lung cancer, chronic obstructive pulmonary disease, bronchiectasis, emphysema, and chronic bronchitis) in ZMSC. Additionally, we investigated potential dose-response relationships between smoking-associated genera and both the duration of smoking in years and the number of cigarettes smoked per day among current smokers as well as and the number of years since quitting smoking among former smokers.

The exploration of the associations between identified genera and blood metabolites as well as health outcomes (step 3)

In this regard, we conducted regression analysis based on ZMSC observational data and two-sample MR analysis based on GWAS summary data to offer epidemiological and genetic-based evidence. In regression analysis with smoking-associated microbes, logistic regression was applied for binary diseases (dependent variables), while linear regression was applied for continuous biochemical indicators (dependent variables) and blood metabolites (independent variables). We examined 14 common chronic diseases with at least 10 cases each, 22 biochemical markers, and 1912 blood metabolites available in ZMSC. The basic models were adjusted for age and BMI. In smoking-metabolite association analysis, we further adjusted for dietary intake as a sensitivity analysis. In MR analysis, the causal links between gut microbiota (exposures) and health-related phenotypes (outcomes) were mainly estimated by inverse-variance weighted (IVW) method. The summary data on gut microbiota was derived from GWASs containing over 7000 individuals of Asian descent (unpublish, the GWAS summary statistics will be available at https://omics.lab.westlake.edu.cn/data.html). The summary data for the outcomes were obtained from BioBank Japan (BBJ), where we used continuous variables and disease phenotypes (excluding those specific to female) with a case count greater than 1800 (approximately equivalent to a prevalence of 0.01%) [45]. The usage and interpretation of our MR analysis followed the STROBE-MR (Strengthening the Reporting of Observational Studies in Epidemiology-Mendelian Randomization) checklist [46] (Additional file 1: Table S2).

Here, false discovery rate (FDR) using Benjamini-Hochberg method was calculated. Results with FDR-corrected p less than 0.05 in one type of analysis were considered as credible evidence, while those with nominal p less than 0.05 were deemed suggestive evidence. More details about disease data, metabolic data, and MR analysis are available in the Supplementary Methods (Additional file 1).

Results

Characteristics of study participants

This study included a total of 3308 male participants: 383 from ZMSC, 554 from GNHS, 2255 from GGMP, and 116 from SRRSHS (Table 1). Of these, 2109 were current smokers, accounting for 63.75%. ZMSC participants were relatively older (65.46 ± 10.00 years) and SRRSHS participants were the youngest (49.90 ± 11.83 years). Both the number of smoking years and cigarettes per day among current smokers in ZMSC were numerically higher than those in GNHS and GGMP. Generally, age varied among the three different smoking statuses (nominal p < 0.05), while BMI showed no significant difference.

Table 1 Basic characteristics of the participants in the four independent studies
Full size table

Cigarette smoking and microbial diversity

Upon evaluating the results of alpha and beta diversity, no robust evidence supported the significant difference in microbial community structure between smokers and non-smokers. Specifically, the reduction in Pielou index in current smokers was noticed only in two studies from Zhejiang (nominal p < 0.05, Fig. 2a). Other alpha diversity metrics did not exhibit significant differences among the groups (Additional file 1: Fig. S2). As shown in the PCoA plots of beta diversity, the distribution of samples from the three groups did not show a significant separation (Fig. 2b). The group differences in beta diversity were observed only in GGMP, which has the largest sample size (PERMANOVA p = 0.006). It is worth mentioning that the contribution of smoking status to the microbial structure variation was limited.

Fig. 2
figure 2

Variations in the gut microbial diversity across groups with different smoking status. a The difference of alpha diversity parameter Pielou between three/two groups was evaluated using a Wilcoxon test in each study. The box plots depict its median and interquartile range, while gray dots represent individual samples, and the violin plots illustrate the distribution of these samples. Additional results for Richness, Shannon index, and Simpson index can be found in Fig. S2 (Additional file 1). b The difference of beta diversity was visualized using PCoA plot based on Bray-Cutis distance at the genus level. PERMANOVA with 999 permutations was employed to identify the variation of beta diversity among the three/two groups, adjusting for age and BMI

Full size image

Cigarette smoking and gut microbes at the genus level

A total of 206 eligible genera were finally analyzed in this study, of which 113 (54.9%) were shared across two or more studies (Additional file 1: Fig. S1). All results of MaAsLin2 on the association between smoking status and eligible genera are provided in Table S7-S9 (Additional file 2). When comparing current smokers to non-smokers, we identified five (ZMSC), six (GNHS), 18 (GGMP), and nine (SRRSHS) genera with nominal p less than 0.05 among 133, 102, 81, and 99 genera, respectively (Fig. 3a). After meta-analysis, four genera were able to withstand the Bonferroni correction, i.e., Actinomyces[g] (β = 0.748, se = 0.108, p = 3.66E-12), Atopobium[g] (β = 0.833, se = 0.146, p = 1.23E-08), Haemophilus[g] (β = − 0.923, se = 0.174, p = 1.10E-07), Turicibacter[g] (β = 0.674, se = 0.156, p = 1.52E-05) (Fig. 3b; Additional file 2: Table S11). We also identified Lachnospira[g] (β = − 0.492, se = 0.144, p = 6.59E-04) whose association was identified in ZMSC and replicated in GGMP (Fig. 3a), although the meta-analyzed p did not reach the threshold for Bonferroni correction (Bonferroni-corrected p = 0.136). We identified a total of five genera that were associated with smoking. Follow-up analyses were mainly based on these five genera. Among current smokers, we observed the dose-response relationships between smoking duration and the decline in Lachnospira[g] abundance (β = − 0.442, se = 0.156, p = 0.005) as well as the increase in Atopobium[g] abundance (β = 0.389, se = 0.163, p = 0.017) (Fig. 3c). In ZMSC, a nominally significant association between cigarettes smoked per day and decline in Haemophilus[g] abundance was also observed (β = − 0.789, se = 0.392, p = 0.044) (Fig. 3d).

Fig. 3
figure 3

The association between cigarette smoking and gut microbes at the genus level (16S rRNA). a The Venn plot illustrates the number of genera with nominal p less than 0.05 in each of the four studies using MaAsLin2. The estimations involved the comparison between current smokers and non-smokers, with adjustments for age and BMI. b-d The forest plots display the meta-analysis results of association between b smoking status (current smokers vs. non-smokers), c the duration of smoking in years, and d the number of cigarettes smoked per day and genera. Genera with Bonferroni-corrected p less than 0.05 or showing overlap across studies are showed. The fixed-effects model was utilized, and heterogeneity was assessed using I2 and Cochran-Q tests

Full size image

When former smokers were compared to current smokers, we identified seven (ZMSC), six (GNHS), and 14 (GGMP) genera with nominal p less than 0.05 (Fig. 4a; Additional file 2: Table S7-S10). Through meta-analysis, Actinomyces[g] (β = − 0.542, se = 0.155, p = 4.68E-04) exhibited consistency across three studies, and Turicibacter[g] (β = − 0.903, se = 0.236, p = 1.27E-04) withstood Bonferroni correction (Fig. 4b; Additional file 2: Table S12). No significant dose-response relationship was observed between these two genera and the years of smoking cessation (Fig. 4c).

Fig. 4
figure 4

The relationship between smoking cessation and gut microbes at the genus level (16S rRNA). a The Venn plot illustrates the number of genera with nominal p less than 0.05 in each of the three studies using MaAsLin2. The estimations involved the comparison between former smokers and current smokers, with adjustments for age and BMI. b, c The forest plots display the meta-analysis results of association between b smoking status (former smokers vs. current smokers) as well as c the years of smoking cessation and genera. Genera with Bonferroni-corrected p less than 0.05 or showing overlap across studies are showed. The fixed-effects model was utilized, and heterogeneity was assessed using I2 and Cochran-Q tests. d The abundance of identified smoking-associated genera in groups with different smoking status in each study. The differences in abundance were evaluated using a Wilcoxon test. The box plots depict their median and interquartile range, while the violin plots illustrate the distribution of participants

Full size image

When former smokers were compared to non-smokers, we identified four (ZMSC), ten (GNHS), and five (GGMP) genera with nominal p less than 0.05 (Additional file 2: Table S7-S10). However, no genus overlapped across studies or reached the Bonferroni correction threshold in meta-analysis (Additional file 2: Table S13).

Taking all three comparisons together, we found that the abundance of two genera (Actinomyces[g] and Turicibacter[g]) elevated in current smokers comparing to non-smokers and decreased in former smokers relative to current smokers. However, simply from numerical terms, their abundance did not fully return to the levels observed in non-smokers. The intuitive comparison of their abundance between groups in each study is shown in Fig. 4d.

The majority results of sensitivity analyses were not substantially different from the main results, including age-stratified analysis (Additional file 1: Fig. S3a-S3c), additional adjustment for drinking status (Additional file 1: Fig. S3d), adopting the random-effects model in meta-analysis (Additional file 1: Fig. S3e), and using smoking information from the ZMSC 2015 survey (Additional file 1: Fig. S3f). Of note, heterogeneity analysis indicated a significant difference between the two age groups (≤ 60 years and > 60 years) in the effect of smoking on Actinomyces[g] (heterogeneity p = 0.02). In ZMSC, the trend of effect sizes remained concordant between the main results and those obtained by excluding patients with smoking-related diseases (Additional file 1: Fig. S3g).

Cigarette smoking and gut microbes at the species level

Based on conventional exclusion criteria, our study finally included 238 species from ZMSC and GNHS (Additional file 1: Fig. S4; Additional file 2: Table S14-S15). The results for alpha and beta diversity were consistent with the pattern from 16S data, with no significant differences among the three groups (Additional file 1: Fig. S5). All results of MaAsLin2 and meta-analysis on the association between smoking status and eligible species are provided in Table S16-S20 (Additional file 2) and Fig. S6 (Additional file 1). Considering the lower abundance observed at the species level, we performed another analysis that no longer excluded species with an average relative abundance of less than 0.01% (Additional file 2: Table S21-S22). Based on this data, we identified overlapping species when comparing current smokers to non-smokers, including Actinomyces sp. HMSC035G02[s], Actinomyces sp. oral taxon 181[s], Actinomyces graevenitzii[s], and Actinomyces sp. S6 Spd3[s], between the two studies with nominal p less than 0.05 (Fig. 5a). Together with the four above, the meta-analysis suggested ten significant species (Bonferroni-corrected p < 0.05) (Fig. 5b; Additional file 2: Table S23). Notably, eight of them fall under the branch of Actinomyces[g]. Five of the eight species demonstrated their effect directions consistent with Actinomyces[g] and were prominent in a large proportion, indicating that smoking-induced elevation in Actinomyces[g] abundance may be driven by these species (Fig. 5c). The results of meta-analysis comparing former smokers to current smokers and non-smokers, respectively, were presented in Table S24-S25 (Additional file 2).

Fig. 5
figure 5

The relationship between cigarette smoking and gut microbes at the species level (metagenome including species with an average relative abundance less than 0.01). a The Venn plot illustrates the number of species with nominal p less than 0.05 in ZMSC and GNHS using MaAsLin2. The estimation involved the comparison between current smokers and non-smokers, with adjustments for age and BMI. b The forest plot displays the meta-analysis results of association between smoking status (current smokers vs. non-smokers) and species. Species with Bonferroni-corrected p less than 0.05 are showed. The fixed-effects model was utilized, and heterogeneity was assessed using I2 and Cochran-Q tests. c The Sankey plot provides a visual representation of the quantitative relationships among species under the Actinomyces[g] branch. The blue boxes and arrows signify the species identified from the metagenomic analysis, with Bonferroni-corrected p less than 0.05 as well as the same direction of estimate as the Actinomyces[g] identified from the 16S rRNA analysis. The thickness of the central grey connecting line reflects the cumulative relative abundance of these species

Full size image

Cigarette smoking and microbial pathway

461 and 466 microbial pathways, respectively, were included for analysis in ZMSC and GNHS (Additional file 1: Fig. S7; Additional file 2: Table S26-S27). Within ZMSC, we observed some preliminary evidence that lysine synthesis-related pathways (PWY-5097, PWY-2942, PWY-724) were reduced in current smokers (nominal p < 0.05, Additional file 2: Table S28). Notably, by analyzing the ZMSC blood metabolomic data, we also found that current smokers tend to have a lower abundance of serum peptides including lysine (Lys-Glu-Glu, Lys-Ile-Val-Lys, Ser-Val-Lys-Arg, and Thr-Lys-Gln-Lys) than non-smokers (nominal p < 0.05, Additional file 2: Table S32). These lines of evidence indicated a link between smoking and lysine synthesis and metabolism. All results of MaAsLin2 on the association between smoking status and eligible pathways are provided in Table S28-S29 (Additional file 2).

Smoking-associated genera and health-related outcomes

To understand the phenotypic consequence of smoking-associated genera, we carried out regression analysis based on ZMSC observational data and two-sample MR analysis based on GWAS summary data. Here, results with FDR-corrected p less than 0.05 in one type of analysis were considered credible evidence, while those with nominal p less than 0.05 were deemed suggestive evidence. In ZMSC, three associations between specific microbes and traits remained statistically significant after applying FDR correction, with one notable pair being Haemophilus[g] and chronic enteritis (β = − 0.092, se = 0.031, p = 2.84E-03) (Fig. 6a). Regarding the MR analysis estimated by the IVW method, only the positive association between Atopobium[g] and chronic hepatitis C passed the significance threshold after FDR correction (β = 0.098, se = 0.028, p = 5.15E-04) (Fig. 6b). The sensitivity analyses showed similar evidence and no horizontal pleiotropy nor heterogeneity (among instrumental variables). Remarkably, both the genetics-based MR analysis and regression analysis pointed to an inverse association between Haemophilus[g] and T2D as well as its related traits (nominal p < 0.05, Fig. 6a-c). Meanwhile, this examination revealed potential smoking-genera-disease links wherein these genera may serve as mediators for the effects of smoking on these outcomes, like T2D (possibly mediated by Haemophilus[g]) and cholecystitis (possibly by Actinomyces[g]) (Fig. 6c). The comprehensive results concerning the association between the identified genera and health-related outcomes can be found in Table S30-S31 (Additional file 2).

Fig. 6
figure 6

The relationship between the smoking-associated genera and health-related outcomes. a Multivariable logistic/linear regression was used to estimate the association between four identified genera and the array of binary and continuous phenotypes available in ZMSC. The model was adjusted for age and BMI. The continuous phenotypes were represented in blue and binary phenotypes in red. b IVW method was utilized to estimate the association between five identified genera and binary phenotypes in the BBJ. Statistical significance is denoted with asterisks (* nominal p < 0.05, ** nominal p < 0.01, # FDR-corrected p < 0.05). Additional results for continuous phenotypes are provided in Table S32 (Additional file 2). c Highlighted are potential pathways suggesting that these genera may mediate the effects of smoking on these outcomes. Solid lines indicate evidence from observational studies, solid lines with arrows indicate evidence from MR, and red highlights represent signals from both analyses

Full size image

Smoking-associated genera and smoking-associated metabolites

A total of 1912 serum metabolites were detected in ZMSC during the 2015 survey. We estimated the association between smoking and metabolites under a cross-sectional design. The results revealed that five metabolites (i.e., 4-hydroxytryptamine, trans-3-hydroxycotinine, N-1-naphthylbenzamide, γ-Glu-Cys, and cyclo(glu-glu)) were significantly associated with cigarette smoking, with FDR-corrected p less than 0.05 (Fig. 7a; Additional file 2: Table S32). After adjusting for dietary intake, the results for these five metabolites remained largely unchanged (Additional file 2: Table S33). Subsequent association analyses demonstrated the significant connections between these five metabolites and Actinomyces[g] (Fig. 7b; Additional file 2: Table S34). For example, the concentration of γ-glutamylcysteine (γ-Glu-Cys) was negatively correlated with Actinomyces[g] abundance (β = − 0.191, se = 0.072, p = 8.90E-03), a finding that is corroborated by evidence from the MetOrigin platform [47]. These results brought up a reasonable speculation that these five metabolites, influenced by smoking, potentially further modulated Actinomyces[g] as mediators (Fig. 7c). In addition to Actinomyces[g], we also observed suggestive findings related to Lachnospira[g] and Turicibacter[g] (nominal p < 0.05, but FDR-corrected p > 0.05).

Fig. 7
figure 7

The relationship between smoking-associated metabolites and smoking-associated genera. a The results of association analysis between smoking and blood metabolites in ZMSC (current smokers vs. non-smokers). The Y-axis corresponds to the minus logarithms of the p values. The direction of the triangle indicates the direction of the effect size. The dotted red lines correspond to the statistical significance level (FDR-corrected p < 0.05). Associations surviving the significance criteria are labeled by name. b Multivariable linear regression was used to estimate the associations between five significant metabolites and four identified genera in ZMSC. The model was adjusted for age and BMI. Statistical significance is denoted with asterisks (* nominal p < 0.05, # FDR-corrected p < 0.05). c Highlighted are potential pathways suggesting that these metabolites may mediate the effects of smoking on Actinomyces[g]

Full size image

Discussion

In the present population-based smoking-gut microbiota association study, we integrated four independent studies and compared gut microbiota profiles among Chinese men with different smoking status. We identified five genera (Actinomyces[g], Atopobium[g], Haemophilus[g], Turicibacter[g], and Lachnospira[g]) associated with cigarette smoking using the 16S rRNA data, and metagenomic data offered a higher resolution at the species levels and additional information from the microbial pathway perspective. These identified genera were subsequently linked to several prevalent chronic diseases (such as T2D), along with the smoking-associated metabolites (such as γ-Glu-Cys). This study contributes to our comprehension of the intricate interplay between cigarette smoking, the gut microbiome, metabolites, and health outcomes.

In our study, no clear evidence supported the significant changes in microbial diversity attributed to smoking. From our study and existing evidences, an intriguing observation surfaced: smaller-scale studies consistently reported a significant reduction in microbial diversity in smokers. Examples included studies by Stewart et al. (30 participants) [20], Curtis et al. (30 participants) [21], Zhang et al. (131 participants) [22], and Yan et al. (154 participants) [19]. However, in studies with larger sample size, this significance diminished, revealing merely a decreased trend. Notable instances were investigations by Prakash et al. (809 participants) [23], Nolan-Kenney et al. (249 participants) [24], and Chen et al. (118 participants) [25]. In our study, only SRRSHS showing a significant decline in alpha diversity among smokers. The aforementioned finding prompts the consideration of whether insufficient sample sizes could lead to false-positive results. The gut microbial structure is not only affected by sex and age [48, 49], but also largely by geographical location [32] and type of staple food [50]. The contribution of these factors on microbiome variation appeared more substantial compared to smoking status, which was also demonstrated in our study. In small-scale studies, balancing these factors between groups becomes more challenging, and unbalanced factors may confound the estimation. In addition, several specific explanations were also considered for our results. First, the genera analyzed in this study were those with relatively high prevalence (> 10%), which introduced a degree of artificial control for variability in genus richness between comparison groups. A typic fact was that the genera we analyzed were shared across all groups. Second, the smoking-associated microbes we identified exhibited relatively low abundance (0.01% ~ 0.2%). Thus, the alterations in their abundance induced by smoking may not have a remarkable impact on the global composition of the gut microbiota.

In addition to diversity, variations in specific microbes may be more informative and provide important insights [51]. Among five genera identified in the present study, the impact of smoking on Haemophilus[g] has been reported in our prior MR analysis [28]. The reduction abundance of Lachnospira[g] was also supported by findings from another multi-ethnic cohort [23]. As smoke directly traverses the oral cavity and upper aero-digestive tract, it induces alterations in the microbiota composition within these regions. Numerous studies have reported the enrichment of Actinomyces[g] [52, 53], Atopobium[g] [54,55,56], and the reduction of Haemophilus[g] [52, 57] in the oral cavity of smokers. However, the impacts on distal organs remain unclear. Our study suggested that this association may extend to the gut. Even the dose-response relationship for cumulative smoking exposure on Haemophilus[g], observed in oral [57], was evident in our gut microbiota study. Experimental evidence provided a possible reason that toxic substances from cigarette smoke could be detected in the gut, which triggered similar antibiotic effects [13]. As for Turicibacter[g], its association with smoking appeared to be controversial. A mouse study revealed a potentially time-dependent effect of smoking exposure on the increase in Turicibacter[g] abundance [58]. Nevertheless, two other mice studies have reported conflicting findings on this matter [59, 60]. In any case, our study contributed new evidence to this ongoing debate from a population perspective. As for Bacteroides[g] (a typical example mentioned in the introduction), our results suggested that smoking alone may not significantly alter its abundance. Notably, the statistical methods employed in previous reports (such as Linear discriminant analysis Effect Size (LefSe) [22, 26, 27] and naive Kruskal-Wallis test [20, 21]) overlooked the confounding variables. The instability of these results might be attributed to unaddressed confounding factors. In our study, after restricting to male participants, mitigating the possible effects from passive smoking and medication, and then adjusting for age and BMI using MaAsLin2, the results provided answers with a higher degree of confidence. Furthermore, previous studies on cigarette smoking and gut microbiota have been less thoroughly investigated at the species level. Our study, utilizing metagenomic data, further unveiled that the smoking-induced elevation of Actinomyces[g] was primarily driven by some species, notably Actinomyces graevenitzii[s], considering both compositional ratio and effect size. Case reports of pulmonary abscess indicated that this organism is an opportunistic human pathogen [61, 62].

From the established consequences of smoking (such as direct antibacterial activity [63], chronic low-grade inflammation [64], and alteration of the intestinal micro-environment including pH [13, 65] and intestinal barrier [66, 67]), we could propose some hypotheses related to the observed changes in the microbial profile. Toxicology indicates that heavy metals in cigarette smoke (such as cadmium) can exert a direct antibiotic effect on the Lachnospiraceae[f], the family to which Lachnospira[g] belongs [13, 68]. Cadmium, on the other hand, elevates Turicibacter[g], although this may be an indirect result [13, 68]. Exposure to smoke components could elevate intestinal pH, reducing the production of organic acids as one of the ways involved [13, 65]. This may favor certain bacterial populations, enabling the thriving of specific genera. Actinomyces[g] can be categorized into acidophilic and basophilic strains, with the majority likely being basophilic [69]. As a result, the Actinomyces[g] in smokers tends to increase, with its some species showing an increase while some decline. There was also a significant correlation observed between the presence of Atopobium[g] and an elevated pH [70].

This study also aimed to explore the potential impact of smoking-induced changes in bacteria on diseases and their potential modulation by blood metabolites. A noteworthy finding in this regard was the association between Haemophilus[g] and T2D. The consistent negative correlation reported in various observational studies suggested that Haemophilus[g] could serve as a valuable microbiological marker for T2D [71,72,73]. A recent MR study in individuals of European descent provided further confirmation, identifying Haemophilus[g] as a defense element against T2D [74]. Moreover, a multi-omics study has elucidated crucial mechanisms through which gut microbiota not only aid in carbohydrate digestion but also help to improve insulin resistance, thereby preventing the development of obesity and diabetes [75]. The association between Actinomyces[g] and cholecystitis could be direct, given that a specific type of cholecystitis named Actinomycosis of the gallbladder is caused by Actinomyces[g] [76]. The underlying mechanism is not clear between Actinomyces[g] and other types of cholecystitis. Our study also found the protective effect of Lachnospira[g] on chronic gastritis. Lachnospira[g] functions as a probiotic, offering diverse benefits including immune stimulation and intestinal acidification through short-chain fatty acids (SCFAs) production, as well as colonization resistance through antibiotic production [77]. In serval previous studies, Lachnospira[g] has been reported to offer protection against intestinal disease in humans [78, 79]. MetOrigin is a platform that provides a comprehensive pipeline for investigating the interactions between bacteria and specific metabolic reactions. In addition, it has identified numerous established bacteria-metabolite associations through the analysis of two studies (a pediatric study and the TwinsUK cohort) [47, 80]. The association between γ-Glu-Cys and Actinomyces[g] is further supported by results from MetOrigin. It seems to be related to glutamate-cysteine ligase, which serves as the primary rate-limiting enzyme of glutathione synthesis. However, the other four smoking-associated metabolites has not yet been studied directly with Actinomyces[g].

This study needs to acknowledge the limitations. Firstly, microbiome studies have often found that microbiota-phenotype associations identified in one location did not necessarily hold true when used elsewhere underscoring the limitations of extrapolating such findings [32]. The same limitation applies to analyses of microbial pathways. Secondly, the differences (including the definition of smoking status, DNA extraction kits, 16S rRNA sequencing platform, clustering/de-noising methods, and taxonomic assignment reference) of the four sub-studies result in heterogeneities, which indicated that the results need to be interpreted with caution. Thirdly, unconsidered/unmeasured confounding factors are somehow inevitable. Because the microbial composition is influenced by various factors, precisely quantifying the impact of a particular factor on the microbes is difficult. More importantly, as discussed earlier, host location accompanying types of staple foods and urbanization showed the strongest associations and contributions with microbiota variations [50]. Traits with smaller contributions, like smoking, may be overshadowed or confounded by these factors. This may also affect the smoking-metabolite association analyses. Fourthly, the Bonferroni correction we applied may result in over-adjustment for p values. This correction method for multiple testing treats each genus as independent, overlooking the intricate inter-dependencies among bacterial organisms. Our study highlighted five genera with higher confidence levels, yet it remains plausible that there exist additional implicated genera. Fifthly, there was a risk of reversed causality when it comes to quitting smoking. Sixthly, both metabolites and the microbiome naturally change with aging. In ZMSC, metabolites were sampled five years prior to the microbiome. In the absence of repeated measurements, this time interval makes it difficult to draw definitive conclusions about the relationship between the two. Finally, the preliminary findings regarding microbial pathways and metabolites only showed nominal significance, which should be validated in an independent dataset.

Conclusions

In summary, we uncovered significant alterations in several gut microbes between smokers and non-smokers. Several smoking-affected genera were further identified to be associated with the risk of common diseases including cholecystitis and T2D, providing clues that gut microbiome as a potential mediator between smoking and common diseases.

Data availability

The raw sequence data of ZMSC are available in the Genome Sequence Archive (GSA) at https://ngdc.cncb.ac.cn/gsa/ under accession numbers (CRA008796: Ju-Sheng Zheng. 1617 raw sequence reads with gut metagenomic sequencing (2023); CRA010223: Ju-Sheng Zheng. 227 raw sequence data with gut metagenomic sequencing (2023)). The raw sequence data of GNHS are accessible in the CNGB Sequence Archive (CNSA) at https://db.cngb.org/cnsa/ under accession numbers (CNP0000829: Ju-Sheng Zheng. 16s rRNA of 2178 stool samples raw sequence reads (2020); CNP0001510: Ju-Sheng Zheng. Metagenomic sequencing of participants from GNHS study (2024)). The raw sequences data of GGMP are deposited in the European Nucleotide Archive at https://www.ebi.ac.uk/ena/ under accession number (PRJEB18535: Southern Medical University. GGMP (2017)). The raw sequences data of SRRSHS can be found in the NCBI Sequence Read Archive at https://www.ncbi.nlm.nih.gov/sra under accession number (SRP238779: Zhejiang University. The effect of smoking and drinking on intestinal microbiota (2020)).

The code utilized to execute the analyses described in this study can be accessed on the GitHub repository located at https://github.com/zdangm/smoking_microbiome.

Abbreviations

ASVs:

Amplicon Sequence Variants

BBJ:

BioBank Japan

BMI:

Body mass index

FDR:

False discovery rate

γ-Glu-Cys:

γ-glutamylcysteine

GGMP:

Guangdong Gut Microbiome Project

GNHS:

Guangzhou Nutrition and Health Study

GWASs:

Genome-Wide Association Studies

HUMAnN:

HMP Unified Metabolic Analysis Network

IVW:

Inverse-variance weighted

LefSe:

Linear discriminant analysis Effect Size

MaAsLin2:

Microbiome Multivariable Association with Linear Models

MR:

Mendelian randomization

MR-PRESSO:

MR pleiotropy residual sum and outlier

MetaPhlAn:

Metagenomic Phylogenetic Analysis

OTUs:

Operational Taxonomic Units

PCoA:

Principal coordinate analysis

PERMANOVA:

Permutational multivariate analysis of variance

PRINSEQ:

PReprocessing and INformation of SEQuence data

QIIME2:

Quantitative Insights into Microbial Ecology 2

SCFA:

Short-chain fatty acid production

SRRSHS:

Sir Run Run Shaw Hospital Study

STROBE-MR:

Strengthening the Reporting of Observational Studies in Epidemiology-Mendelian Randomization

T2D:

Type 2 diabetes

ZMSC:

Zhejiang Metabolic Syndrome Cohort

References

  1. Thursby E, Juge N. Introduction to the human gut microbiota. Biochem J. 2017;474(11):1823–36.

    Article  CAS  PubMed  Google Scholar 

  2. Hou K, Wu ZX, Chen XY, Wang JQ, Zhang D, Xiao C, Zhu D, Koya JB, Wei L, Li J, et al. Microbiota in health and diseases. Signal Transduct Target Ther. 2022;7(1):135.

    Article  PubMed  PubMed Central  Google Scholar 

  3. Valdes AM, Walter J, Segal E, Spector TD. Role of the gut microbiota in nutrition and health. BMJ. 2018;361: k2179.

    Article  PubMed  PubMed Central  Google Scholar 

  4. Zhernakova DV, Wang D, Liu L, Andreu-Sanchez S, Zhang Y, Ruiz-Moreno AJ, Peng H, Plomp N, Del Castillo-Izquierdo A, Gacesa R, et al. Host genetic regulation of human gut microbial structural variation. Nature. 2024;625(7996):813–21.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Valle Gottlieb MG, Closs VE, Junges VM, Schwanke CHA. Impact of human aging and modern lifestyle on gut microbiota. Crit Rev Food Sci Nutr. 2018;58(9):1557–64.

    Article  PubMed  Google Scholar 

  6. Dai X, Gil GF, Reitsma MB, Ahmad NS, Anderson JA, Bisignano C, Carr S, Feldman R, Hay SI, He J, et al. Health effects associated with smoking: a Burden of Proof study. Nat Med. 2022;28(10):2045–55.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Fluhr L, Mor U, Kolodziejczyk AA, Dori-Bachash M, Leshem A, Itav S, Cohen Y, Suez J, Zmora N, Moresi C, et al. Gut microbiota modulates weight gain in mice after discontinued smoke exposure. Nature. 2021;600(7890):713–9.

    Article  CAS  PubMed  Google Scholar 

  8. Chen B, Sun L, Zeng G, Shen Z, Wang K, Yin L, Xu F, Wang P, Ding Y, Nie Q, et al. Gut bacteria alleviate smoking-related NASH by degrading gut nicotine. Nature. 2022;610(7932):562–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Bai X, Wei H, Liu W, Coker OO, Gou H, Liu C, Zhao L, Li C, Zhou Y, Wang G, et al. Cigarette smoke promotes colorectal cancer through modulation of gut microbiota and related metabolites. Gut. 2022;71(12):2439–50.

    Article  CAS  PubMed  Google Scholar 

  10. Cicchinelli S, Rosa F, Manca F, Zanza C, Ojetti V, Covino M, Candelli M, Gasbarrini A, Franceschi F, Piccioni A. The impact of smoking on microbiota: a narrative review. Biomed. 2023;11(4):1144.

  11. Huang C, Shi G. Smoking and microbiome in oral, airway, gut and some systemic diseases. J Transl Med. 2019;17(1):225.

    Article  PubMed  PubMed Central  Google Scholar 

  12. Shapiro H, Goldenberg K, Ratiner K, Elinav E. Smoking-induced microbial dysbiosis in health and disease. Clin Sci (Lond). 2022;136(18):1371–87.

    Article  CAS  PubMed  Google Scholar 

  13. Gui X, Yang Z, Li MD. Effect of cigarette smoke on gut microbiota: state of knowledge. Front Physiol. 2021;12: 673341.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Antinozzi M, Giffi M, Sini N, Galle F, Valeriani F, De Vito C, Liguori G, Romano Spica V, Cattaruzza MS. Cigarette smoking and human gut microbiota in healthy adults: a systematic review. Biomed. 2022;10(2):510.

  15. Yang HT, Xiu WJ, Liu JK, Yang Y, Zhang YJ, Zheng YY, Wu TT, Hou XG, Wu CX, Ma YT, et al. Characteristics of the intestinal microorganisms in middle-aged and elderly patients: effects of smoking. ACS Omega. 2022;7(2):1628–38.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Hu X, Fan Y, Li H, Zhou R, Zhao X, Sun Y, Zhang S. Impacts of cigarette smoking status on metabolomic and gut microbiota profile in male patients with coronary artery disease: a multi-omics study. Front Cardiovasc Med. 2021;8: 766739.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Wang P, Dong Y, Jiao J, Zuo K, Han C, Zhao L, Ding S, Yang X, Chen M, Li J. Cigarette smoking status alters dysbiotic gut microbes in hypertensive patients. J Clin Hypertens (Greenwich). 2021;23(7):1431–46.

    Article  CAS  PubMed  Google Scholar 

  18. Kondo Y, Hashimoto Y, Hamaguchi M, Kaji A, Sakai R, Inoue R, Kashiwagi S, Mizushima K, Uchiyama K, Takagi T, et al. Effects of smoking on the gut microbiota in individuals with type 2 diabetes mellitus. Nutrients. 2022;14(22):4800.

  19. Yan S, Ma Z, Jiao M, Wang Y, Li A, Ding S. Effects of smoking on inflammatory markers in a healthy population as analyzed via the gut microbiota. Front Cell Infect Microbiol. 2021;11: 633242.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. Stewart CJ, Auchtung TA, Ajami NJ, Velasquez K, Smith DP, De La Garza R, 2nd, Salas R, Petrosino JF,. Effects of tobacco smoke and electronic cigarette vapor exposure on the oral and gut microbiota in humans: a pilot study. PeerJ. 2018;6: e4693.

  21. Curtis K, Stewart CJ, Robinson M, Molfese DL, Gosnell SN, Kosten TR, Petrosino JF, De La Garza R, 2nd, Salas R,. Insular resting state functional connectivity is associated with gut microbiota diversity. Eur J Neurosci. 2019;50(3):2446–52.

    Article  PubMed  PubMed Central  Google Scholar 

  22. Zhang W, Li J, Lu S, Han N, Miao J, Zhang T, Qiang Y, Kong Y, Wang H, Gao T, et al. Gut microbiota community characteristics and disease-related microorganism pattern in a population of healthy Chinese people. Sci Rep. 2019;9(1):1594.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Prakash A, Peters BA, Cobbs E, Beggs D, Choi H, Li H, Hayes RB, Ahn J. Tobacco smoking and the fecal microbiome in a large, multi-ethnic cohort. Cancer Epidemiol Biomarkers Prev. 2021;30(7):1328–35.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Nolan-Kenney R, Wu F, Hu J, Yang L, Kelly D, Li H, Jasmine F, Kibriya MG, Parvez F, Shaheen I, et al. The association between smoking and gut microbiome in Bangladesh. Nicotine Tob Res. 2020;22(8):1339–46.

    Article  CAS  PubMed  Google Scholar 

  25. Chen J, Ryu E, Hathcock M, Ballman K, Chia N, Olson JE, Nelson H. Impact of demographics on human gut microbial diversity in a US Midwest population. PeerJ. 2016;4: e1514.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Lin R, Zhang Y, Chen L, Qi Y, He J, Hu M, Zhang Y, Fan L, Yang T, Wang L, et al. The effects of cigarettes and alcohol on intestinal microbiota in healthy men. J Microbiol. 2020;58(11):926–37.

    Article  PubMed  Google Scholar 

  27. Harakeh S, Angelakis E, Karamitros T, Bachar D, Bahijri S, Ajabnoor G, Alfadul SM, Farraj SA, Al Amri T, Al-Hejin A, et al. Impact of smoking cessation, coffee and bread consumption on the intestinal microbial composition among Saudis: a cross-sectional study. PLoS ONE. 2020;15(4): e0230895.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Fan J, Zhou Y, Meng R, Tang J, Zhu J, Aldrich MC, Cox NJ, Zhu Y, Li Y, Zhou D. Cross-talks between gut microbiota and tobacco smoking: a two-sample Mendelian randomization study. BMC Med. 2023;21(1):163.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Gou W, Chen Y-m, Pan A, Wang H, Zhang K, Pan X-F, He Y, Fu Y, Jiang Z, Miao Z et al. Westlake Gut Project: a consortium of microbiome epidemiology for the gut microbiome and health research in China. Med Microecol. 2022, 14:100064.

  30. Zhou Y, Zhang X, Zhang L, Li Z, Wu Q, Jin Z, Chen S, He D, Wu S, Zhu Y. Increased stroke risk in metabolically abnormal normal weight: a 10-year follow-up of 102,037 participants in China. Transl Stroke Res. 2021;12(5):725–34.

    Article  CAS  PubMed  Google Scholar 

  31. Ling CW, Zhong H, Zeng FF, Chen G, Fu Y, Wang C, Zhang ZQ, Cao WT, Sun TY, Ding D et al. Cohort Profile: Guangzhou Nutrition and Health Study (GNHS): a population-based multi-omics study. J Epidemiol. 2023;34(6):301–6.

  32. He Y, Wu W, Zheng HM, Li P, McDonald D, Sheng HF, Chen MX, Chen ZH, Ji GY, Zheng ZD, et al. Regional variation limits applications of healthy gut microbiome reference ranges and disease models. Nat Med. 2018;24(10):1532–5.

    Article  CAS  PubMed  Google Scholar 

  33. Liang X, Fu Y, Cao WT, Wang Z, Zhang K, Jiang Z, Jia X, Liu CY, Lin HR, Zhong H, et al. Gut microbiome, cognitive function and brain structure: a multi-omics integration analysis. Transl Neurodegener. 2022;11(1):49.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Shuai M, Zhang G, Zeng FF, Fu Y, Liang X, Yuan L, Xu F, Gou W, Miao Z, Jiang Z, et al. Human gut antibiotic resistome and progression of diabetes. Adv Sci (Weinh). 2022;9(11): e2104965.

    Article  PubMed  Google Scholar 

  35. Deng K, Xu JJ, Shen L, Zhao H, Gou W, Xu F, Fu Y, Jiang Z, Shuai M, Li BY, et al. Comparison of fecal and blood metabolome reveals inconsistent associations of the gut microbiota with cardiometabolic diseases. Nat Commun. 2023;14(1):571.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, Fierer N, Pena AG, Goodrich JK, Gordon JI, et al. QIIME allows analysis of high-throughput community sequencing data. Nat Methods. 2010;7(5):335–6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Yilmaz P, Parfrey LW, Yarza P, Gerken J, Pruesse E, Quast C, Schweer T, Peplies J, Ludwig W, Glockner FO. The SILVA and “All-species Living Tree Project (LTP)” taxonomic frameworks. Nucleic Acids Res. 2014, 42(Database issue):D643–648.

  38. McDonald D, Price MN, Goodrich J, Nawrocki EP, DeSantis TZ, Probst A, Andersen GL, Knight R, Hugenholtz P. An improved Greengenes taxonomy with explicit ranks for ecological and evolutionary analyses of bacteria and archaea. ISME J. 2012;6(3):610–8.

    Article  CAS  PubMed  Google Scholar 

  39. Schmieder R, Edwards R. Quality control and preprocessing of metagenomic datasets. Bioinformatics. 2011;27(6):863–4.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Truong DT, Franzosa EA, Tickle TL, Scholz M, Weingart G, Pasolli E, Tett A, Huttenhower C, Segata N. MetaPhlAn2 for enhanced metagenomic taxonomic profiling. Nat Methods. 2015;12(10):902–3.

    Article  CAS  PubMed  Google Scholar 

  41. Abubucker S, Segata N, Goll J, Schubert AM, Izard J, Cantarel BL, Rodriguez-Mueller B, Zucker J, Thiagarajan M, Henrissat B, et al. Metabolic reconstruction for metagenomic data and its application to the human microbiome. PLoS Comput Biol. 2012;8(6): e1002358.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Caspi R, Billington R, Fulcher CA, Keseler IM, Kothari A, Krummenacker M, Latendresse M, Midford PE, Ong Q, Ong WK, et al. The MetaCyc database of metabolic pathways and enzymes. Nucleic Acids Res. 2018;46(D1):D633–9.

    Article  CAS  PubMed  Google Scholar 

  43. Caspi R, Billington R, Ferrer L, Foerster H, Fulcher CA, Keseler IM, Kothari A, Krummenacker M, Latendresse M, Mueller LA, et al. The MetaCyc database of metabolic pathways and enzymes and the BioCyc collection of pathway/genome databases. Nucleic Acids Res. 2016;44(D1):D471–480.

    Article  CAS  PubMed  Google Scholar 

  44. Mallick H, Rahnavard A, McIver LJ, Ma S, Zhang Y, Nguyen LH, Tickle TL, Weingart G, Ren B, Schwager EH, et al. Multivariable association discovery in population-scale meta-omics studies. PLoS Comput Biol. 2021;17(11): e1009442.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  45. Sakaue S, Kanai M, Tanigawa Y, Karjalainen J, Kurki M, Koshiba S, Narita A, Konuma T, Yamamoto K, Akiyama M, et al. A cross-population atlas of genetic associations for 220 human phenotypes. Nat Genet. 2021;53(10):1415–24.

    Article  CAS  PubMed  Google Scholar 

  46. Skrivankova VW, Richmond RC, Woolf BAR, Davies NM, Swanson SA, VanderWeele TJ, Timpson NJ, Higgins JPT, Dimou N, Langenberg C, et al. Strengthening the reporting of observational studies in epidemiology using mendelian randomisation (STROBE-MR): explanation and elaboration. BMJ. 2021;375: n2233.

    Article  PubMed  PubMed Central  Google Scholar 

  47. Yu G, Xu C, Wang X, Ju F, Fu J, Ni Y. MetOrigin 2.0: Advancing the discovery of microbial metabolites and their origins. Imeta. 2024;3(6):e246.

  48. Zhang X, Zhong H, Li Y, Shi Z, Ren H, Zhang Z, Zhou X, Tang S, Han X, Lin Y, et al. Sex- and age-related trajectories of the adult human gut microbiota shared across populations of different ethnicities. Nat Aging. 2021;1(1):87–100.

    Article  PubMed  Google Scholar 

  49. de la Cuesta-Zuluaga J, Kelley ST, Chen Y, Escobar JS, Mueller NT, Ley RE, McDonald D, Huang S, Swafford AD, Knight R, et al. Age- and sex-dependent patterns of gut microbial diversity in human adults. mSystems. 2019;4(4):e00261–19.

  50. Lu J, Zhang L, Zhai Q, Zhao J, Zhang H, Lee YK, Lu W, Li M, Chen W. Chinese gut microbiota and its associations with staple food type, ethnicity, and urbanization. NPJ Biofilms Microbiomes. 2021;7(1):71.

    Article  PubMed  PubMed Central  Google Scholar 

  51. Johnson KV, Burnet PW. Microbiome: should we diversify from diversity? Gut Microbes. 2016;7(6):455–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  52. Jia YJ, Liao Y, He YQ, Zheng MQ, Tong XT, Xue WQ, Zhang JB, Yuan LL, Zhang WL, Jia WH. Association between oral microbiota and cigarette smoking in the Chinese population. Front Cell Infect Microbiol. 2021;11: 658203.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Al-Marzooq F, Al Kawas S, Rahman B, Shearston JA, Saad H, Benzina D, Weitzman M. Supragingival microbiome alternations as a consequence of smoking different tobacco types and its relation to dental caries. Sci Rep. 2022;12(1):2861.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Pfeiffer S, Herzmann C, Gaede KI, Kovacevic D, Krauss-Etschmann S, Schloter M. Different responses of the oral, nasal and lung microbiomes to cigarette smoke. Thorax. 2022;77(2):191–5.

    Article  PubMed  Google Scholar 

  55. Suzuki N, Nakano Y, Yoneda M, Hirofuji T, Hanioka T. The effects of cigarette smoking on the salivary and tongue microbiome. Clin Exp Dent Res. 2022;8(1):449–56.

    Article  PubMed  Google Scholar 

  56. Sato N, Kakuta M, Uchino E, Hasegawa T, Kojima R, Kobayashi W, Sawada K, Tamura Y, Tokuda I, Imoto S, et al. The relationship between cigarette smoking and the tongue microbiome in an East Asian population. J Oral Microbiol. 2020;12(1):1742527.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  57. Lim MY, Yoon HS, Rho M, Sung J, Song YM, Lee K, Ko G. Analysis of the association between host genetics, smoking, and sputum microbiota in healthy humans. Sci Rep. 2016;6:23745.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  58. Battey JND, Szostak J, Phillips B, Teng C, Tung CK, Lim WT, Yeo YS, Ouadi S, Baumer K, Thomas J, et al. Impact of 6-month exposure to aerosols from potential modified risk tobacco products relative to cigarette smoke on the rodent gastrointestinal tract. Front Microbiol. 2021;12: 587745.

    Article  PubMed  PubMed Central  Google Scholar 

  59. Wang X, Ye P, Fang L, Ge S, Huang F, Polverini PJ, Heng W, Zheng L, Hu Q, Yan F, et al. Active smoking induces aberrations in digestive tract microbiota of rats. Front Cell Infect Microbiol. 2021;11: 737204.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  60. Whitehead AK, Meyers MC, Taylor CM, Luo M, Dowd SE, Yue X, Byerley LO. Sex-dependent effects of inhaled nicotine on the gut microbiome. Nicotine Tob Res. 2022;24(9):1363–70.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  61. Gliga S, Devaux M, Gosset Woimant M, Mompoint D, Perronne C, Davido B. Actinomyces graevenitzii pulmonary abscess mimicking tuberculosis in a healthy young man. Can Respir J. 2014;21(6):e75–7.

    PubMed  PubMed Central  Google Scholar 

  62. Himeji D, Hara S, Kawaguchi T, Tanaka GI. Pulmonary Actinomyces graevenitzii infection diagnosed by bronchoscopy using endobronchial ultrasonography with a guide sheath. Intern Med. 2018;57(17):2547–51.

    Article  PubMed  PubMed Central  Google Scholar 

  63. Pavia CS, Pierre A, Nowakowski J. Antimicrobial activity of nicotine against a spectrum of bacterial and fungal pathogens. J Med Microbiol. 2000;49(7):675–6.

    Article  CAS  PubMed  Google Scholar 

  64. Arnson Y, Shoenfeld Y, Amital H. Effects of tobacco smoke on immunity, inflammation and autoimmunity. J Autoimmun. 2010;34(3):J258–265.

    Article  CAS  PubMed  Google Scholar 

  65. Tomoda K, Kubo K, Asahara T, Andoh A, Nomoto K, Nishii Y, Yamamoto Y, Yoshikawa M, Kimura H. Cigarette smoke decreases organic acids levels and population of bifidobacterium in the caecum of rats. J Toxicol Sci. 2011;36(3):261–6.

    Article  PubMed  Google Scholar 

  66. Papoutsopoulou S, Satsangi J, Campbell BJ, Probert CS. Review article: Impact of cigarette smoking on intestinal inflammation-direct and indirect mechanisms. Aliment Pharmacol Ther. 2020;51(12):1268–85.

    Article  CAS  PubMed  Google Scholar 

  67. Zuo L, Li Y, Wang H, Wu R, Zhu W, Zhang W, Cao L, Gu L, Gong J, Li N, et al. Cigarette smoking is associated with intestinal barrier dysfunction in the small intestine but not in the large intestine of mice. J Crohns Colitis. 2014;8(12):1710–22.

    Article  PubMed  Google Scholar 

  68. Breton J, Le Clere K, Daniel C, Sauty M, Nakab L, Chassat T, Dewulf J, Penet S, Carnoy C, Thomas P, et al. Chronic ingestion of cadmium and lead alters the bioavailability of essential and heavy metals, gene expression pathways and genotoxicity in mouse intestine. Arch Toxicol. 2013;87(10):1787–95.

    Article  CAS  PubMed  Google Scholar 

  69. Jiang Y, Li WJ, Xu P, Tang SK, Xu LH. Study on diversity of actinomycetes under salt and alkaline environments. Wei Sheng Wu Xue Bao. 2006;46(2):191–5.

    PubMed  Google Scholar 

  70. Schwiertz A, Taras D, Rusch K, Rusch V. Throwing the dice for the diagnosis of vaginal complaints? Ann Clin Microbiol Antimicrob. 2006;5:4.

    Article  PubMed  PubMed Central  Google Scholar 

  71. Letchumanan G, Abdullah N, Marlini M, Baharom N, Lawley B, Omar MR, Mohideen FBS, Addnan FH, Nur Fariha MM, Ismail Z, et al. Gut microbiota composition in prediabetes and newly diagnosed type 2 diabetes: a systematic review of observational studies. Front Cell Infect Microbiol. 2022;12: 943427.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  72. Qin J, Li Y, Cai Z, Li S, Zhu J, Zhang F, Liang S, Zhang W, Guan Y, Shen D, et al. A metagenome-wide association study of gut microbiota in type 2 diabetes. Nature. 2012;490(7418):55–60.

    Article  CAS  PubMed  Google Scholar 

  73. Wang X, Xu X, Xia Y. Further analysis reveals new gut microbiome markers of type 2 diabetes mellitus. Antonie Van Leeuwenhoek. 2017;110(3):445–53.

    Article  CAS  PubMed  Google Scholar 

  74. Sun K, Gao Y, Wu H, Huang X. The causal relationship between gut microbiota and type 2 diabetes: a two-sample Mendelian randomized study. Front Public Health. 2023;11:1255059.

    Article  PubMed  PubMed Central  Google Scholar 

  75. Takeuchi T, Kubota T, Nakanishi Y, Tsugawa H, Suda W, Kwon AT, Yazaki J, Ikeda K, Nemoto S, Mochizuki Y, et al. Gut microbial carbohydrate metabolism contributes to insulin resistance. Nature. 2023;621(7978):389–95.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  76. Freland C, Massoubre B, Horeau JM, Caillon J, Drugeon HB. Actinomycosis of the gallbladder due to Actinomyces naeslundii. J Infect. 1987;15(3):251–7.

    Article  CAS  PubMed  Google Scholar 

  77. Manzano M, Talavera-Rodriguez A, Moreno E, Madrid N, Gosalbes MJ, Ron R, Dronda F, Perez-Molina JA, Lanza VF, Diaz J et al. Relationship of diet to gut microbiota and inflammatory biomarkers in people with HIV. Nutrients. 2022;14(6):1221.

  78. Meehan CJ, Beiko RG. A phylogenomic view of ecological specialization in the Lachnospiraceae, a family of digestive tract-associated bacteria. Genome Biol Evol. 2014;6(3):703–13.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  79. Ai D, Pan H, Li X, Gao Y, Liu G, Xia LC. Identifying gut microbiota associated with colorectal cancer using a zero-inflated lognormal model. Front Microbiol. 2019;10:826.

    Article  PubMed  PubMed Central  Google Scholar 

  80. Yu G, Xu C, Zhang D, Ju F, Ni Y. MetOrigin: discriminating the origins of microbial metabolites for integrative analysis of the gut microbiome and metabolome. Imeta. 2022;1(1): e10.

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We thank all the researchers and participants involved in the four studies. We thank Zhoushan Municipal Medical Insurance Bureau, Precision Nutrition and Computational Medicine Lab, and Biobank Japan project for providing invaluable data. We thank Westlake University Supercomputer Center for the assistance in storing and computing microbial data.

Funding

This work was supported by the National Natural Sciences Foundation of China No. 82204118 (DZ), the Healthy Zhejiang One Million People Cohort K-20230085 (DZ), the Key Laboratory of Intelligent Preventive Medicine of Zhejiang Province 2020E10004 (DZ), and the National Natural Science Foundation of China No. 82473613 (FZ).

Author information

Author notes

  1. Jiayao Fan, Fangfang Zeng and Haili Zhong contributed equally to this work.

Authors and Affiliations

  1. The Second Affiliated Hospital, School of Public Health, Zhejiang University School of Medicine, 388 Yuhangtang Road, Hangzhou, 310058, Zhejiang, China

    Jiayao Fan, Chunxiao Cheng, Chunfeng He, Yuan Zhou & Dan Zhou

  2. Department of Public Health and Preventive Medicine, School of Medicine, Jinan University, Guangzhou, Guangdong, China

    Fangfang Zeng, Jun Cai, Tao Liu & Wenjun Ma

  3. Guangdong Provincial Key Laboratory of Food, Nutrition and Health, Department of Epidemiology, School of Public Health, Sun Yat-Sen University, Guangzhou, Guangdong, China

    Haili Zhong & Yu-Ming Chen

  4. Department of Gastroenterology, School of Medicine, Sir Run Run Shaw Hospital, Zhejiang University, Hangzhou, Zhejiang, China

    Wentao Shen, Shujie Chen & Lan Wang

  5. Institute of Gastroenterology, Zhejiang University, Hangzhou, Zhejiang, China

    Wentao Shen, Shujie Chen & Lan Wang

  6. Department of Epidemiology & Biostatistics, School of Public Health, Zhejiang University, 388 Yuhangtang Road, Hangzhou, 310058, Zhejiang, China

    Yuanjiao Liu & Yimin Zhu

  7. Institute of Basic Medical Sciences, Westlake Institute for Advanced Study, Hangzhou, Zhejiang, China

    Ju-Sheng Zheng

  8. Westlake Intelligent Biomarker Discovery Lab, Westlake Laboratory of Life Sciences and Biomedicine, Hangzhou, 310024, Zhejiang, China

    Ju-Sheng Zheng

  9. Key Laboratory of Growth Regulation and Translation Research of Zhejiang Province, School of Life Sciences, Westlake University, Hangzhou, Zhejiang, China

    Ju-Sheng Zheng

  10. The Key Laboratory of Intelligent Preventive Medicine of Zhejiang Province, Hangzhou, Zhejiang, China

    Dan Zhou

Authors
  1. Jiayao Fan
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Fangfang Zeng
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Haili Zhong
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Jun Cai
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Wentao Shen
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Chunxiao Cheng
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Chunfeng He
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Yuanjiao Liu
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Yuan Zhou
    View author publications

    You can also search for this author inPubMed Google Scholar

  10. Shujie Chen
    View author publications

    You can also search for this author inPubMed Google Scholar

  11. Yimin Zhu
    View author publications

    You can also search for this author inPubMed Google Scholar

  12. Tao Liu
    View author publications

    You can also search for this author inPubMed Google Scholar

  13. Ju-Sheng Zheng
    View author publications

    You can also search for this author inPubMed Google Scholar

  14. Lan Wang
    View author publications

    You can also search for this author inPubMed Google Scholar

  15. Yu-Ming Chen
    View author publications

    You can also search for this author inPubMed Google Scholar

  16. Wenjun Ma
    View author publications

    You can also search for this author inPubMed Google Scholar

  17. Dan Zhou
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

DZ, WM, YC and LW were the major contributors in design and conceptualization. JC and WS analyzed the data. CC, CH, YL and YZ1 verified the correctness of the data. JF, FZ and HZ were major contributors in writing the manuscript. SC, YZ2, TL and JZ critically revised the manuscript. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Lan Wang, Yu-Ming Chen, Wenjun Ma or Dan Zhou.

Ethics declarations

Ethics approval and consent to participate

The ZMSC protocol received approval from the Ethics Committee of Zhejiang University School of Medicine. The GNHS protocol received ethical approval from both the Ethics Committee of the School of Public Health, Sun Yat-sen University, and the Ethics Committee of Westlake University. The GGMP was approved by the Ethical Review Committee of the Chinese Center for Disease Control and Prevention under Approval Notice No. 201519-A. The SRRSHS was conducted under the guidelines of the Ethics Committee of Sir Run Run Shaw Hospital and the China Association for Clinical Research. Written consent was obtained from all participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

12916_2025_3852_MOESM1_ESM.docx

Additional file 1. Supplementary Methods included description of studies and participants, taxonomic profiling, and statistical analysis. Table S1. Detailed information on stool sample collection, processing, and 16S rRNA/shotgun metagenomic sequencing of each study. Table S2. STROBE-MR checklist of recommended items to address in reports of Mendelian randomization studies. Fig. S1. Details regarding eventually included genera (16S rRNA). Fig. S2. Variations in the microbial alpha diversity across groups with different smoking status. Fig. S3. The sensitivity analyses for the association between cigarette smoking and identified genera (16S rRNA, current smokers vs. non-smokers). Fig. S4. Details regarding eventually included species (metagenome). Fig. S5. Variations in the gut microbial diversity across groups with different smoking status (metagenome). Fig. S6. The association between cigarette smoking and gut microbes at the species level (metagenome, excluding species with an average relative abundance less than 0.01). Fig. S7. Details regarding eventually included microbial pathways (metagenome) and their associations with smoking.

12916_2025_3852_MOESM2_ESM.xlsx

Additional file 2. Table S3. Information on the genera eventually included in ZMSC. Table S4. Information on the genera eventually included in GNHS. Table S5. Information on the genera eventually included in GGMP. Table S6. Information on the genera eventually included in SRRSHS. Table S7. All results of MaAsLin2 on the association between smoking status and eligible genera in ZMSC. Table S8. All results of MaAsLin2 on the association between smoking status and eligible genera in GNHS. Table S9. All results of MaAsLin2 on the association between smoking status and eligible genera in GGMP. Table S10. All results of MaAsLin2 on the association between smoking status and eligible genera in SRRSHS. Table S11. The results of meta-analysis that integrates findings from the four studies at the genus level (current smokers vs. non-smokers). Table S12. The results of meta-analysis that integrates findings from the four studies at the genus level (former smokers vs. current smokers). Table S13. The results of meta-analysis that integrates findings from the four studies at the genus level (former smokers vs. non-smokers). Table S14. Information on the species eventually included in ZMSC (including species with an average relative abundance less than 0.01). Table S15. Information on the species eventually included in GNHS (including species with an average relative abundance less than 0.01). Table S16. All results of MaAsLin2 on the association between smoking status and eligible species in ZMSC (excluding species with an average relative abundance less than 0.01). Table S17. All results of MaAsLin2 on the association between smoking status and eligible species in GNHS (excluding species with an average relative abundance less than 0.01). Table S18. The results of meta-analysis that integrates findings from ZMSC and GNHS at the species level (current smokers vs. non-smokers, excluding species with an average relative abundance less than 0.01). Table S19. The results of meta-analysis that integrates findings ZMSC and GNHS at the species level (former smokers vs. current smokers, excluding species with an average relative abundance less than 0.01). Table S20. The results of meta-analysis that integrates findings from ZMSC and GNHS at the species level (former smokers vs. non-smokers, excluding species with an average relative abundance less than 0.01). Table S21. All results of MaAsLin2 on the association between smoking status and eligible species in ZMSC (including species with an average relative abundance less than 0.01). Table S22. All results of MaAsLin2 on the association between smoking status and eligible species in GNHS (including species with an average relative abundance less than 0.01). Table S23. The results of meta-analysis that integrates findings from ZMSC and GNHS at the species level (current smokers vs. non-smokers, including species with an average relative abundance less than 0.01). Table S24. The results of meta-analysis that integrates findings from ZMSC and GNHS at the species level (former smokers vs. current smokers, including species with an average relative abundance less than 0.01). Table S25. The results of meta-analysis that integrates findings from ZMSC and GNHS at the species level (former smokers vs. non-smokers, including species with an average relative abundance less than 0.01). Table S26. Information on the pathways eventually included in ZMSC. Table S27. Information on the pathways eventually included in GNHS. Table S28. All results of MaAsLin2 on the association between smoking status and eligible pathways in ZMSC. Table S29. All results of MaAsLin2 on the association between smoking status and eligible pathways in GNHS. Table S30. The results on the association between the four identified smoking-associated genera and available phenotypes in ZMSC. Table S31. The results of mendelian randomization on the association between five identified smoking-associated genera and available phenotypes in the Biobank Japan. Table S32. The results on the association between the smoking status and 1912 blood metabolites in ZMSC. Table S33. The results on the association between the smoking status and 1912 blood metabolites in ZMSC after further adjustment of dietary intake. Table S34. The results on the association between five significant metabolites and four identified genera in ZMSC.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Fan, J., Zeng, F., Zhong, H. et al. Potential roles of cigarette smoking on gut microbiota profile among Chinese men. BMC Med 23, 25 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12916-025-03852-2

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12916-025-03852-2

Keywords

BMC Medicine

ISSN: 1741-7015

Contact us