Bacterial pneumonia, Corona Virus Disease 2019 (COVID-19), acute upper respiratory infections, and intestinal infections are common causes of hospital admission and important contributors to death (1). Among respiratory infectious diseases, lower respiratory tract infections (LRTIs) and pneumonia, in particular, rank in highest in terms of mortality (2). Several factors are implicated in heightening the risk of infection, including age, vaccination coverage, antibiotic therapy, seasonal fluctuations, and air pollution (3).

Notably, the persistently rising levels of air pollution worldwide have yielded dire consequences, leading to an alarming number of premature deaths. This increase in pollution is linked to a higher incidence of infections, especially respiratory infections (4, 5). Consequently, the role of respiratory infections as a driver of human mortality has assumed a position of paramount concern and it is now more crucial than ever to gain a comprehensive understanding of the intricate interplay between air pollution and infections. Particulate matter (PM) serves as a key indicator of air pollution, resulting from a variety of natural and human activities (6). Traditional observational studies have explored the association between PM and infections. Evidence from China by Yongjian Zhu indicated significantly positive associations of PM2.5, PM10, and Nitrogen dioxide (NO₂₎ over the past 2 weeks with newly confirmed COVID-19 cases (7). However, this study did not include gender- or age-specific cases. Another study in Italy found that long-term air quality data with cases of COVID-19 in up to 71 Italian provinces, further indicating that chronic exposure to air pollution could facilitate virus spread (8). A systematic review with 15 studies by Chiara Copat indicated PM2.5 and NO₂ are more closely correlated to COVID-19 than PM10 (9). Notably, these studies included limited sample sizes with potential confounders.

Investigating the causal association between air pollution and the risk of infections is challenging due to reverse causation and confounding. Mendelian randomization (MR) has emerged as a potent method for determining causation between risk factors and diseases by using genetic variants as instrument variables (IVs) (10). MR analysis can largely overcome the confounders with random assignment of an individual's genetic variants at conception. Moreover, it minimizes the risk of reverse causation, since the presence of a disease cannot affect individuals' genotypes (11).

In our study, we applied a two-sample MR analysis to explore the potential causal association between air pollution (including PM2.5, PM2.5–10, PM10, nitrogen dioxide, and nitrogen oxides) and risk of infections (intestinal infections, acute upper respiratory infections, bacterial pneumonia, and Corona Virus Disease 2019) using the summary statistics derived from the publicly available GWAS data.

In this two-sample MR study, we found that PM2.5, PM2.5–10, and nitrogen oxides increase the risk of Coronavirus Disease 2019. Furthermore, PM2.5, PM2.5–10, PM10, and nitrogen oxides are associated with an increased risk of bacterial pneumonia. However, our analysis did not reveal a suggestively association between air pollution and the risk of acute upper respiratory infections and intestinal infections, except for nitrogen dioxide.

As mentioned, exposure to PM is associated with upper and lower respiratory tract infections. The COVID-19 pandemic, caused by the novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has resulted in historic numbers of infections and deaths worldwide over the last 3 years (2022). Consequently, an increasing number of studies have investigated associations between PM and COVID-19 infection in China, Italy, and the USA (7–9, 38). The mechanism of COVID-19 infection related to PM can be summarized as follows. (1) PM may be involved in different life cycle stages of COVID-19, including alteration of SARS-CoV-2 viral receptors and proteases required for entry (e.g., angiotensin-converting enzyme 2 (ACE2) and transmembrane protease serine type 2 (TMPRSS2), proteins and protease critical to SARS-CoV-2 entry into host cells) (39), inhibition of mucosal ciliary clearance, alteration of antiviral interferon production and viral replication (40). (2) PM may impair the immune system. PM-induced pro-inflammatory cytokine production, oxidative stress, and impaired airway immune function may lead to increased susceptibility to respiratory pathogens, which can increase the risk of COVID-19 pneumonia (41). For example, exposure to chemicals in PM damages lung epithelial cells, interferes with tight junctions between epithelial cells, increases permeability of airway and lung epithelial cells, and decreases protection against viral infections (42). Furthermore, substances such as heavy metals and polycyclic aromatic hydrocarbons promote the production of reactive oxygen species in lung cells (40), increasing susceptibility to further oxidative damage due to oxidative stress. Interestingly, our results, consistent with findings by Bontempi (43), revealed a closer correlation between PM2.5 and COVID-19 than with PM10. This may be because PM10 larger than 5 μm is unlikely to reach type II alveolar cells, where the ACE2 receptor for cellular entry is primarily located (44).

The relationship between NO₂ and COVID-19 has been reported as positive in China, Europe, and the United States (9). However, contrasting evidence suggesting a negative or insignificant association between NO₂ and COVID-19 also exists (45, 46). The potential effect of NO₂ exposure remains uncertain. Studies on nitrogen oxides (NO_x) and COVID-19 are scarce. We demonstrated that there is a significant association between NO_x and COVID-19. This may be determined by the next mentioned physicochemical properties of NO_x that are not readily soluble in water. In addition, a study by Pfeffer et al. discovered that higher environmental NO_x levels were linked to exacerbated viral lung infections, which increases the susceptibility to concurrent bacterial infections and enhances its severity (47).

Pneumonia has been a cause of morbidity and mortality throughout human history (48). Bacterial pneumonia represents the most prevalent manifestation of pneumonia, characterized by inflammation affecting the terminal airways, alveoli, or interstitial spaces because of pathogenic microorganism infections (49). Pathogenetic investigations have indicated Streptococcus pneumoniae as potentially the most widespread bacterial agent responsible for pneumonia on a global scale (50). Over the past few decades, the rise in mortality rates associated with bacterial pneumonia has been attributed, in part, to worsening air quality and the proliferation of environmental hazards. These factors facilitate pathogen transmission, often synergizing with particulate matter (PM) (51). Air pollution poses a significant challenge to public health, exacerbating the potential for pathogen epidemics.

Our study establishes a potential causal relationship between PM2.5, PM2.5–10, PM10, and nitrogen oxides (mainly nitric oxide), all of which exhibit associations with bacterial pneumonia. It is widely acknowledged that respirable particles encompass a complex amalgamation of ions, organic compounds, metals, carbonaceous matter, and other constituents (52). The composition and relative abundance of these components are primarily contingent upon factors such as source origin, climatic conditions, topography, and other environmental variables (53). Consequently, respirable particles may exhibit varying degrees of solubility, ranging from partial solubility to complete insolubility (54). Furthermore, nitrogen oxides (mainly nitric oxide) possess limited solubility and does not undergo reactions with water. These physical properties, to some extent, contribute to the characteristics of these air pollutants in the context of lower respiratory tract infections. Currently, a plethora of epidemiological and mechanistic studies lends support to our conclusion. Notably, a meta-analysis has demonstrated a causal relationship between PM2.5 exposure and acute lower respiratory tract infections (55).

In the case of Chile, respirable particles have been associated with an elevated frequency of emergency room visits among children under the age of 2 experiencing lower respiratory symptoms (56). Conversely, no significant association has been observed between indoor NO₂ concentrations and the incidence or severity of respiratory illnesses in infants (57). Furthermore, urban PM exposure can promote bacterial adherence to human respiratory epithelial cells by impairing mucus cilia activity in airway mucosa (58, 59), as well as promoting bacterial adherence to human respiratory epithelial cells (60) to enhance bacterial colonization of the host lower respiratory tract.

On the other hand, our findings indicate that only NO₂ suggestively constitutes a risk factor for upper respiratory tract infections, with its soluble nature dictating the characteristics of this air pollutant, primarily impacting the upper respiratory tract (54). In a study conducted by Arbex et al. (61), it was observed that a mere 10 μg/m³ rise in NO₂ concentration correlated with a 0.63% increase in visits for upper respiratory tract infections. Likewise, two researches conducted in China (62, 63) reported that each 10 μg/m³ increment in NO₂ levels corresponded to a 1.00% increase in emergency room visits for upper respiratory tract infections, along with a substantial 11.27% surge in pediatric emergency room visits related to upper respiratory tract infections. However, it is important to note that while NO₂ may not directly enter the pulmonary system, its potent oxidizing properties trigger an inflammatory response, subsequently activating the body's immune system and affecting the entire respiratory system (64).

The SNP rs12203592, located in the IRF4 gene intron, is linked to pigmentation traits, hematological traits, squamous cell carcinoma, and smoking cessation. A multi-ethnic GWAS identified it as a novel lung cancer locus. It also increases the risk of invasive aspergillosis post-hematopoietic stem cell transplantation by modulating IRF4 mRNA expression and immune responses (65, 66). Another SNP, rs1537371, in the CDKN2B-AS1 gene, is linked to cardiovascular diseases like coronary artery disease (67, 68). SNPs in the 9p21 region, such as rs1537371 and rs1333049, are significantly associated with an increased risk of cardiovascular disease (69). This study finds rs1537371 and rs12203592 significantly associated with air pollution-related COVID-19 and bacterial pneumonia risks, suggesting pleiotropic effects (70).

These findings offer potential causal relationship between air pollutant exposure and the incidence of respiratory infections. Notably, it is of concern that a staggering 91% of the world's population resides in regions where air pollution exceeds the limits recommended by the World Health Organization (WHO) (source: https://www.who.int/news-room/fact-sheets/detail/ambient-(outdoor)-air-quality-and-health). Therefore, taking steps to curtail air pollution has the potential to yield significant reductions in the global burden of respiratory infectious diseases. Consequently, our data underscore the importance of advocating for worldwide initiatives aimed at reducing air pollution and transitioning to more sustainable energy sources that yield cleaner air.

There are some strengths should be mentioned as following: Firstly, this is the first MR analysis to examine the casual effect of air pollution on infections using large-scale GWAS data. Comparing to the observational studies, MR analysis can largely overcome the confounders with random assignment of an individual's genetic variants at conception. Moreover, the risk of reverse causation could also be minimized. Secondly, we conducted sensitivity analysis both with and without outliers to detect any coincidental effects, and all models exhibited no directional pleiotropy. Finally, F statistics were consistently >10, indicating a robust correlation between SNPs and exposures with the sufficient statistical power and minimizing the possibility of weak instrumental variable bias.

There are some shortcomings in this MR study. Firstly, there was heterogeneity among our results. Due to the GWAS data, any potential non-linear relationships or stratification effects which varies by health status, age or sex cannot be examined which may be the resource of heterogeneity. Secondly, our study could not rule out the effect of canalization (i.e., dilution of the gene-exposure association) and thus the estimate might be inflated (71). Thirdly, the association between air pollution and different infections subtypes was not explored. Additionally, instances where air pollution levels exceed the standard cannot be entirely discounted, as such occurrences may introduce confounding factors that could bias our MR estimates. Finally, our datasets included the European populations, limiting the applicability of our results to non-European populations. Further studies are needed to verify the applicability of these results in other populations and ethnicities.

In conclusion, we demonstrated a causal association between PM2.5, PM2.5–10, and nitrogen oxides on Corona Virus Disease 2019. Additionally, PM2.5, PM2.5–10, PM10, and Nitrogen oxides will increase the risk of bacterial pneumonia. However, air pollution, except for nitrogen dioxide, was not associated with the risk of acute upper respiratory infections. Overall, more work is needed for policy formulation to reduce air pollution and the emission of toxic and harmful gases.

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

SY: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. TT: Conceptualization, Data curation, Formal analysis, Investigation, Project administration, Resources, Supervision, Validation, Writing – original draft, Writing – review & editing. HW: Conceptualization, Data curation, Formal analysis, Investigation, Project administration, Resources, Supervision, Validation, Writing – original draft, Writing – review & editing. ZL: Data curation, Resources, Supervision, Validation, Writing – review & editing. MW: Data curation, Resources, Supervision, Validation, Writing – review & editing. KN: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – review & editing.