Asthma, a chronic respiratory disease, manifests a spectrum of symptoms ranging from wheezing, shortness of breath, cough and chest tightness, all accompanied by variable airflow obstruction. It is a multifarious disease, characterized by a myriad of endotypes and phenotypes. In brief, asthma manifests predominantly in two major endotypes: T-helper cell type 2 (Th2)-high (eosinophilic) and Th2-low (non-eosinophilic), also known as type 2 (T2) asthma and non-T2 asthma, respectively. Th2-high phenotypes are delineated into three distinct subgroups: early-onset allergic asthma, late-onset eosinophilic asthma, and aspirin-exacerbated respiratory disease. Conversely, Th2-low phenotypes are classified based on clinical attributes encompassing obesity, smoking and age (Kuruvilla et al., 2019).

The multifarious nature of asthma yields varying responses to therapy, resulting in severe cases that persist despite potent treatments. Severe asthma is observed in up to 10% of adults and 2.5% of children diagnosed with asthma, bearing a greater burden of asthma-related morbidity and mortality (Kuruvilla et al., 2019; Brusselle and Koppelman, 2022). Therefore, there is a necessity for an in-depth study of the intricate mechanisms underlying the diverse asthma endotypes and phenotypes.

Among the multitude factors that contribute to the development of asthma, the gut microbiota, a pivotal agent in immune system development, has recently emerged as a compelling area of investigation. Recent discoveries indicate that alterations in gut microbial composition and function are closely linked to fluctuations in immune responses and the occurrence of respiratory diseases, including asthma (Budden et al., 2017). Disruptions in the gut microbiota during early life have been associated with an increased susceptibility to asthma (Fujimura et al., 2016). Although the overall composition of the gut microbial community tends to remain relatively stable in both children and adults, specific differences in certain taxa are believed to play a potential role in the development of asthma (Budden et al., 2017; Barcik et al., 2020).

Microecological studies typically involve collecting fecal samples from patients and analyzing them through sequencing techniques like 16S rRNA gene sequencing. These studies aim to compare the alterations in the microbial composition between individuals with a specific disease and healthy individuals. However, conflicting results are often observed in similar studies due to the absence of standardized sampling methods and longitudinal study designs (Knight et al., 2018; Combrink et al., 2023). Moreover, these observational studies are often hindered by confounding factors (e.g., antibiotic or corticosteroid use) and reverse causality. Thus, establishing a causal relationship between gut microbiota and asthma is challenging.

To overcome these obstacles, we employed a two-sample Mendelian randomization (MR) study, an effective analytical tool that utilizes genetic variants as instrumental variables (IVs) to infer causality between potentially modifiable risk factors and diseases. MR uses genetic variation to mimic randomized controlled trials (RCTs). Suppose there is a single nucleotide polymorphism (SNP) that is known to influence a certain phenotype (exposure). By following Mendel’s laws of inheritance and the fixed nature of germline genotypes, the alleles received by an individual at this SNP are expected to be random. In this “natural experiment,” the SNP is considered an IV, and observing an individual’s genotype at this SNP is similar to randomly assigning them to a treatment or control group in an RCT (Hemani et al., 2018). MR analysis evaluate the causal relationship between the exposure and outcome by comparing the impact of SNPs on the outcome to their impact on the exposure (Davies et al., 2018). SNPs are treated as independent RCTs, and meta-analysis tools can be adapted to combine the results from each SNP, providing a more powerful overall causal estimate. Two-sample MR utilizes data from separate GWAS studies for the exposure and the outcome. The independent transmission of alleles from parents to offspring ensures that confounding factors do not affect the study of health outcomes. Additionally, germline genotypes remain unaffected by disease processes, preventing reverse causality from impacting associations between genotype and disease outcomes (Lawlor et al., 2008).

In previous MR studies, Shi and colleagues have revealed the causal interplay between gut microbiota and various chronic respiratory diseases, including unspecified asthma (Shi et al., 2023). As the scientific spotlight on asthma moves towards understanding the etiologic molecular mechanisms, our objective was to further investigate the causal relationships between gut microbiota and distinct asthma phenotypes. This study aimed to strengthen the basis for precision gut-microbiota-mediated therapeutic strategies, thereby adding genetic evidence to the evolving landscape of asthma management.

The objective of this study was to investigate the causal effects of specific gut microbial taxa on asthma phenotypes. ( Figure 1 ) A comprehensive set of 211 microbial traits were incorporated as exposure. As for the study outcomes, we included summary statistics related to various asthma phenotypes. These phenotypes comprised unspecified asthma, allergic asthma, childhood asthma, eosinophilic asthma, and obesity-related asthma. The triad of allergic asthma, childhood asthma, and suggestive for eosinophilic asthma adopt the T2 asthma classification, whereas obesity-related asthma adopts the non-T2 asthma classification.

Based on our current knowledge, this study represents a pioneering effort in exploring the causal connections between gut microbiota and diverse asthma phenotypes. Considering that asthma can be classified into type 2 asthma (including early-onset allergic asthma, late-onset eosinophilic asthma, and aspirin-exacerbated respiratory disease) and non-type 2 asthma (based on factors such as obesity, smoking, and age), we searched in the FinnGen biobank for GWAS data related to these phenotypes. We included five traits of asthma phenotypes that matched these classifications, namely unspecified asthma, allergic asthma, childhood asthma, suggestive for eosinophilic asthma, and obesity-related asthma. Then, we conducted MR analysis and sensitivity analysis for 211 gut microbial taxa and 5 traits of asthma. We identified a total of 58 causal relationships that met the nominal p-value significance threshold of 0.05 in our study. After performing sensitivity analysis, 40 associations remained robust. Following the multiple correction, 4 associations remained statistically significant. These causal relationships necessitate further validation and hold potential value for future studies. The immunomodulatory properties of different microbial taxa may be responsible for the observed causal relationships in our study. For example, gut microbiota-derived metabolites, specifically short-chain fatty acids (SCFAs), have shown protective effects against asthma by reducing atopic sensitization (Skrivankova et al., 2021), decreasing sputum eosinophilia (Bowden and Holmes, 2019), and alleviating airway (Verbanck et al., 2018) and lung inflammation (Cait et al., 2018; Skrivankova et al., 2021; Zou et al., 2021).

Our findings indicated that the genetically predicted abundance of three taxa, including genus Holdemanella, genus Oxalobacter, and genus Butyricimonas, were causally linked to an elevated risk of unspecified asthma. Additionally, the genetically predicted abundance of order NB1n is causally associated with a higher risk of obesity-related asthma. While we observed meaningful causal relationships in unspecified asthma and obesity-related asthma, no significant causal association was found in T2 asthma. This lack of significant findings in T2 asthma may be attributed to the relatively small number of cases within the three T2 asthma traits, as the bacterial taxa displaying significant causal relationships with unspecified asthma demonstrated potential effects on distinct asthma phenotypes. The limited sample size likely reduced the statistical power required to detect significant associations in T2 asthma.

Consistent with a recent MR study (Begley et al., 2018), we found that the genetically predicted abundance of genus Holdemanella demonstrated potential causal effects on four asthmatic traits, including unspecified asthma, allergic asthma, childhood asthma and obesity-related asthma. Interestingly, this genus was found to have a negative correlation with propionate levels in individuals with diabetes and cognitive impairment (Jin et al., 2023). This suggests that Holdemanella may contribute to the development of asthma through its impact on SCFAs like propionate. However, it is worth noting that Holdemanella biformis, a species of the Holdemanella genus, has exhibited protective effects against mouse colitis (Pujo et al., 2021) and type 2 diabetes (Du et al., 2022), potentially due to the production of beneficial long-chain fatty acids (LCFAs). Thus, it is crucial to remember that 16S rRNA sequencing can only classify microbial taxa at the genus level, and different species within the same genus may have contrasting effects on health. Similar to this case, it was reported that Butyricimonas, by decreasing the protective SCFA butyrate, had a protective effect on cystic fibrosis (Romaní-Pérez et al., 2021). However, it was observed that the abundance of this genus increased in patients with acute myocardial infarction. In the case of Oxalobacter, an MR study found that the genus had a causal effect on inflammatory bowel disease (Roduit et al., 2019). However, one species from this genus, Oxalobacter formigenes, has been used as a probiotic to prevent calcium oxalate kidney stone disease (McLoughlin et al., 2019). A conflicting study revealed that trauma patients who died had a relatively higher abundance of Oxalobacter formigenes ( Trompette et al., 2014 ).

We observed a causal relationship between the order NBn1 and obesity-related asthma. Metabolic reconstruction revealed that NB1n lacks a tricarboxylic acid cycle and instead relies on anaerobic fermentation of simple sugars for substrate level phosphorylation (Zheng et al., 2022). This unique metabolic pathway sheds light on the potential mechanisms underlying the observed relationship. Furthermore, a recent MR study revealed a protective effect of NB1n against upper urinary urolithiasis (Liu et al., 2022). These contrasting findings highlight the complexity of NB1n’s interactions with human health and disease. To fully comprehend the role of this genus in maintaining health or contributing to disease development, further research is necessary.

Observational and longitudinal studies have identified biomarkers that can differentiate individuals with asthma from healthy controls. Consistently, our study revealed that some of these biomarkers, such as increased Lachnospiraceae (Burmeister et al., 2020), decreased Clostridia (Daniel et al., 2021) and decreased Veillonella (Dayama et al., 2020), have potential causal effects on asthma. In the study conducted by Begley et al., it was discovered that individuals with asthma had increased levels of Lachnospiraceae. Furthermore, these bacteria demonstrated an independent ability to differentiate between asthmatic and healthy individuals (Burmeister et al., 2020). Multiple studies have also documented higher abundance of Lachnospiraceae members in asthmatic mouse models compared to control groups, with a significant positive correlation observed between these bacteria and immunoglobulin E levels as well as the percentage of eosinophils (Skennerton et al., 2016; Han et al., 2021; Tsukuda et al., 2021; Zhang et al., 2023). Zheng et al. found that the relative abundance of members from Clostridia was lower in subjects with asthma (Daniel et al., 2021). Notably, a longitudinal study revealed that the abundance of Clostridiales had association to increased level of fecal butyrate in early life (Arrieta et al., 2015). Thus, it is highly possible that Clostridiales develops protective effects through producing beneficial butyrate and relieving airway and lung inflammation in asthma. In a study conducted by Arrieta et al., it was observed that children at risk of developing asthma exhibited a significant decrease in the relative abundance of Veillonella. This reduction was accompanied by a decrease in fecal acetate levels (Dayama et al., 2020), indicating a potential mechanism wherein the loss of protective fecal acetate could potentially contribute to the development of asthma. However, inconsistencies were also found when comparing our study to others. Fujimura et al. discovered that different compositions of neonatal human gut microbiota were linked to varying risks of childhood asthma. They observed that lower levels of Bifidobacteria, Lactobacillus, Faecalibacterium, and Akkermansia were associated with an increased likelihood of developing childhood asthma (Fujimura et al., 2016). In our study, we did not find any causal effects of these taxa on childhood asthma. To gain a more comprehensive understanding of the relationship between microbiota and asthma phenotypes, further studies from diverse perspectives are essential.

Our study yielded an intriguing finding: there was limited overlap among the taxa that exhibited potential causal relationships with distinct asthma phenotypes. This finding highlights the complexity of the relationship between gut microbial taxa and the development of asthma. The heterogeneity of various asthmatic phenotypes also exists in the gut microbiome. Various taxa may be involved in the development of distinct asthma phenotypes. This could involve different biological pathways, interactions with the immune system, or other factors that influence the development and progression of asthma phenotypes. Begley et al. identified three distinct groups of asthma patients based on their gut microbiota composition. Furthermore, these groups showed variations in inflammatory mediators, including sputum interleukin (IL) -8 and IL-1β (Burmeister et al., 2020). Zheng et al. discovered that allergic asthma patients and non-allergic asthma children exhibited notable differences in their gut microbiota composition and metabolites. This suggests that the unique gut microbiome and microbiome-derived metabolites may have a significant impact on the modulation of distinct asthma phenotypes (Daniel et al., 2021). Zou et al. made similar observations and demonstrated that variations in gut microbiota composition could effectively differentiate patients diagnosed with allergic asthma (IL-4 high) from non-allergic asthma (IL-4 low) patients (Hu et al., 2023). Taken together, there is a possible connection between gut microbiota composition and the presence of different types of inflammation, such as eosinophilic or neutrophilic, which could potentially contribute to the development of distinct asthma phenotypes. When targeting precision treatment for various asthmatic phenotypes, it is crucial to take into account the differences in the gut microbiome.

The development and management of various asthma phenotypes are influenced by diverse factors. In this study, we have focused on investigating microecological differences as one aspect of understanding asthma’s pathogenesis. However, it is important to note that a comprehensive consideration of multiple factors is necessary to fully comprehend the intricate nature of asthma. Furthermore, it is crucial to consider the potential impact of different treatments, such as the use of antibiotics, on the gut microbiome. These treatments can have varying effects on the composition and function of the microbiome. Therefore, when targeting the gut microbiome for therapeutic purposes, it is necessary to take into account the potential consequences. The microecological differences observed in this study provide an entry point for exploring the complex interplay between the microbiome and asthma. By identifying and addressing the unique microbiome characteristics associated with different asthma phenotypes, we can potentially develop tailored approaches to effectively manage and treat the condition. We believe that targeting specific subtypes of the microbiome shows great promise for advancing precision diagnosis and treatment of asthma.

The utilization of a MR design is a notable advantage. By using genetic variants as IVs, MR analysis provides a way to explore causal relationships between exposures and outcomes. The random allocation of genetic variants during conception helps to mitigate potential confounding factors that may be present in observational studies. Moreover, our study benefits from the inclusion of the largest publicly available GWAS datasets, ensuring a robust and comprehensive analysis. The use of large datasets increases the statistical power of the MR analysis, enhancing our ability to detect and estimate causal effects accurately. Due to the absence of standardized sampling methods and longitudinal study, microecological studies often yield contradictory results (Knight et al., 2018; Combrink et al., 2023). Moreover, confounding factors and reverse causality can often hinder the results of observational studies. Hence, we hope to contribute to a better understanding of the intricate interplay between gut microbiota composition and the development of asthma phenotypes by employing a robust genetic framework.

We must acknowledge the limitations of our study. Firstly, it is important to recognize that the microbiome, being an exposure phenotype, is influenced by numerous variants with relatively small effect sizes. The complex and diverse nature of the gut microbiome means that it is influenced by numerous variants with small effects. To address this complexity and increase statistical power, we have used a less stringent p-value threshold in line with previous studies (Wang et al., 2021). By adopting this approach, we can incorporate a larger number of IVs into our analysis, thereby facilitating the use of sensitivity analysis and bolstering statistical power. However, this approach carries the risk of including false-positive variants. In order to mitigate the influence of weak instruments, we have excluded SNPs with an F-statistic below 10. Secondly, the microbiome GWAS used in our analysis did not target the complete 16S rRNA gene, which limits our ability to identify potential therapeutic targets with sufficient taxonomic resolution. Thirdly, our sample primarily consisted of individuals of European ancestry to reduce bias from population stratification. However, it is important to acknowledge that this restriction in sample diversity somewhat limits the generalizability of our findings to individuals of different ethnic groups.

Collectively, our findings underscore the diverse and complex nature of asthma and offer compelling genetic evidence supporting a causal association between gut microbiota and distinct asthma phenotypes. The causal relationships identified in our study highlighted the role of gut microbiota in the development of various asthma phenotypes. It is possible that different taxa play a role in the development of various asthma phenotypes. These findings underscore the importance of further research and replication studies to validate and understand the specific mechanisms underlying these observed causal relationships. By unraveling the distinct contributions of bacterial taxa to different asthma phenotypes, we can gain valuable insights into the complex pathogenesis of asthma and potentially identify targeted interventions for specific subgroups of patients.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

Ethical approval was not required for the study involving humans in accordance with the local legislation and institutional requirements. Written informed consent to participate in this study was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and the institutional requirements.

Z-XC: Data curation, Formal Analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing. Y-XW: Writing – original draft. Z-JJ: Conceptualization, Writing – review & editing. X-JL: Conceptualization, Writing – review & editing. JZ: Conceptualization, Funding acquisition, Supervision, Validation, Writing – review & editing.