Introduction
Autism spectrum disorder (ASD) is a neurodevelopmental disorder characterized by impairments in social interaction, communication, and the presence of restricted interests and repetitive behaviors [
1]. Recent epidemiological study has shown that 1 in 54 children in the United States is diagnosed with ASD, with a global prevalence of approximately 1% [
2]. The disease presents significant socio-economic burdens, often stemming from lifelong rehabilitation and care needs, lost productivity, and the challenges related to the integration of affected individuals into society [
2,
3]. The risk factors associated with ASD are not fully understood and appear to be related to genetic and environmental factors, such as family history, older parental age, pregnancy complications, and air pollution [
4].
Air pollution presents a significant economic and social challenge worldwide, affecting both developed and developing nations. It is linked to a wide array of adverse health outcomes [
5,
6]. Numerous studies have shown that air pollution is associated with a variety of diseases, such as cancer, respiratory diseases, cardiovascular diseases, and neurological diseases [
7‐
10]. This extensive impact underscores the urgent need for a comprehensive understanding of air pollution’s global health implications, including its role in the etiology of ASD. Recent comprehensive meta-analysis illuminated a concerning correlation: particulate matter 2.5 (PM
2.5) concentrations appear to increase the risk of ASD [
11]. However, some studies did not find a significant relationship between PM
2.5 and ASD [
12,
13]. Additionally, the evidence regarding the effects of PM
10 on ASD remains inconclusive [
14]. Therefore, no study to date has conclusively determined the causal relationship between PM and ASD risk.
Traditional observational studies, including cohort, case-control, and cross-sectional studies, are crucial in epidemiology and medical research for understanding health outcomes, disease prevalence, and associations between risk factors and diseases. However, they have several limitations that can affect the validity and interpretation of their findings, such as residual confounding, reverse causation, and measurement error [
15]. With the recent increase in availability of genome-wide association studies (GWAS) databases, Mendelian randomization (MR) is a novel method that utilizes genetic variants as instrumental variables (IVs) to infer causal relationships between potentially risk factors and diseases, based on the principles of Mendelian inheritance [
16]. The strength of Mendelian randomization lies in its resemblance to randomized controlled trial (RCT). In RCT, participants are randomly assigned to receive a treatment or a placebo to establish causal relationships. Similarly, in MR analysis, the random allocation of genetic variants at conception is utilized to mimic the randomization process, helping to reduce confounding and reverse causation issues that often affect traditional observational studies [
17,
18]. Consequently, MR provides a more robust approach to causal inference in epidemiological studies, potentially guiding public health interventions and therapeutic strategies. Additionally, it enables exploration of potential causality when RCTs are neither feasible nor ethical [
17]. For instance, assigning individuals to harmful pollution exposure is unethical in a trial. Furthermore, this method has been widely employed to investigate causal relationships between PM
2.5 and various diseases, including cancer [
19], cardiovascular disease [
20], thyroid diseases [
21], and gestational diabetes [
22].
In this study, we employed a two-sample MR analyses to investigate the potential association between PM and the risk of ASD.
Discussion
In this study, we used genetic variants as IVs to investigate the causal relationship between PM and the risk of ASD. Our results indicated that PM2.5 and PM2.5 absorbance might increase the risk of ASD. However, no causal association was observed between PM10 and ASD risk.
The reduction of air pollution has profound implications for public health, significantly enhancing respiratory, cardiovascular and neurological well-being across the general population, reducing healthcare costs, and contributing to overall societal health resilience [
6]. Within this broader context of improved general health, the specific impact on neurodevelopmental disorders, particularly ASD, is notable. Recent comprehensive meta-analyses analyzed 28 studies, aggregating data from over 750,000 newborns, revealing that every increase of 5 µg/m
3 in PM
2.5 consistently corresponded with a heightened risk of ASD across all analytical models. This risk was notably higher in relation to PM
2.5 exposure when compared to other pollutants like PM
10, NOx, or solvents [
14]. Several population-based investigations, like the study in Southern California involving 294,937 mother-child pairs, indicated that increased PM
2.5 exposure during the initial two trimesters was associated with a higher risk of ASD. The study went on to further emphasize stronger associations in male offspring compared to female [
39]. Additional research echoes these findings, pinpointing both prenatal and postnatal PM
2.5 exposures as risk factors for ASD in various geographical locations, ranging from Southwestern Pennsylvania to Israel [
40‐
42]. The meticulous examination of large datasets further confirmed this association. In Southern California, analyzing 318,750 mother-child pairs from 2001 to 2014, prenatal exposure to key components of PM
2.5 was linked to an increased risk of ASD in offspring [
43]. Similarly, a multi-site case-control study conducted on United States children born between 2003 and 2006 revealed an association between early life PM
2.5 exposure and ASD, further quantifying the risk with an odds ratio of 1.3 per 1.6 µg/m
3 increase [
44]. Yet, it is interesting to note that a cohort study in South Korea provided evidence suggesting both PM
2.5 and PM
10 exposures during the 4–10 trimester phase of pregnancy correlated with the onset of ASD [
45]. However, not all studies have found a direct association between PM and ASD. For instance, a collaborative study across various European countries, which observed 8,079 children, found no significant relationship between prenatal exposure to pollutants, including PM and NO
2, and childhood autistic traits [
12]. Another study of 132,256 births highlighted an association with prenatal NO
x exposure but didn’t establish a significant link for PM
2.5 [
13]. As to animal experiments, neonatal Sprague-Dawley rats exhibited ASD-like behavioral characteristics upon early neonatal exposure to PM
2.5 [
46]. Additionally, another experiment showed that gestational and early-life exposure to PM
2.5 led to notable behavioral and cognitive shifts in the offspring of juvenile male rats, underscoring the possible etiological role of PM
2.5 in the onset of ASD and related conditions [
47]. In light of these diverse findings, our research aligns with the predominant narrative suggesting the genetic association between PM
2.5 and elevated ASD, and provided the evidence that PM
10 was not significantly related to ASD.
The underlying mechanism connecting PM
2.5 exposure and neurodevelopmental outcomes, including the potential risk for ASD, is complex and multi-faceted. The following are possible potential mechanisms: PM
2.5 contains a mixture of fine particles and droplets that consist of acids, organic chemicals, metals, and soil or dust. Once inhaled, these particles can lead to the release of pro-inflammatory cytokines and reactive oxygen species [
48,
49]. Chronic inflammation and oxidative stress can have harmful effects on both the mother and the fetus during pregnancy [
47,
50,
51]. It’s believed that excessive inflammation, especially during critical periods of fetal brain development, can lead to altered neural connectivity and increased susceptibility to ASD. In animal models, maternal immune activation has been shown to lead to behavioral and brain abnormalities in offspring, which are reminiscent of human neurodevelopmental disorders [
52]. Moreover, PM
2.5 can penetrate the blood-brain barrier (BBB), either directly through the olfactory bulb or through systemic circulation. Once in the brain, these particles can cause local inflammation [
53]. Neuroinflammation can disrupt the normal function and development of neural circuits, leading to abnormal patterns of neural connectivity and functionality associated with ASD [
54]. In the presence of chronic neuroinflammation, microglial cells can become overactive and produce inflammatory mediators that affect brain development [
55]. Furthermore, some components in PM
2.5, particularly polycyclic aromatic hydrocarbons (PAHs), are known to disrupt endocrine function [
56]. Prenatal exposure to certain endocrine disruptors has been shown to result in changes in social behavior, a core feature of ASD [
57]. Finally, epigenetic mechanisms control gene expression without altering the underlying DNA sequence. PM
2.5 exposure can lead to changes in DNA methylation patterns, histone modifications, and non-coding RNAs [
58‐
60]. Altered epigenetic regulation can influence brain development and function, potentially leading to an increased risk of neurodevelopmental disorders [
61,
62]. Recent studies have highlighted potential epigenetic modifications associated with ASD, suggesting this as a possible mechanism linking environmental exposures like PM
2.5 to the disorder [
62].
Urbanization and industrialization, while driving economic growth, pose significant challenges to global air quality [
63]. To address these challenges, scientific research plays a pivotal role in unraveling the complex interplay between air pollution and health, offering vital insights that inform strategies to combat the adverse effects of poor air quality. Current scientific efforts are directed towards both mitigation and adaptation. On the mitigation front, advancements in emission reduction technologies are critical. This includes the development of cleaner fuel sources, such as renewable energy, and the promotion of energy-efficient practices in industries and households [
64,
65]. Urban planning can also play a role by encouraging the use of public transportation, developing green spaces, and implementing zoning regulations that limit industrial activity in residential areas [
66]. For adaptation, enhancing air quality monitoring systems is vital to provide real-time data and enable prompt responses to pollution incidents. Public health initiatives that increase awareness about the impact of air pollution and promote behavioral changes, such as reducing car usage or advocating for cleaner cooking and heating solutions, are also essential [
67]. These strategies and methods could serve as references for various countries to adapt and apply according to their specific circumstances.
Our study has several strengths. The primary advantage is the employment of the MR design, it can mitigate confounding factors and reverse causation, and mimic randomized controlled trials. Secondly, we used the latest and largest GWAS database and rigorous screening to ensure the validity of the IVs, no significant heterogeneity and horizontal pleiotropy was found in the sensitivity analyses, and the Power value greater than 0.8 was also suggested the reliable of our study. Thirdly, the datasets utilized were of European populations, thereby minimizing the potential bias attributed to population stratification. Our research also presents several limitations. We did not assess the impact of PM2.5 on specific ASD subtypes, primarily attributable to the lack of the GWAS database for these subtypes. Additionally, despite leveraging the largest GWAS database available, the inclusion of SNPs significantly associated with PM2.5 was still limited. Lastly, we only obtained summary-level GWAS data and therefore could not analyze the detailed demographic information.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.