Introduction

Air pollution is a mixture of physical, biological, and chemicals substances emitted from both artificial (e.g., vehicles, industries, and power plants) and natural (e.g., volcanic eruptions and wildfires) sources that contaminate the air that we breathe and can be harmful to both the environment and human health1. The World Health Organization (WHO) reported that about 2.4 billion people cook and heat their homes with polluting fuels and every year 3.2 million people prematurely die from household air pollution. The WHO estimates that 99% of people lived in areas where the air quality was not up to the WHO guidelines and outdoor pollution contributed to approximately 4.2 million premature deaths worldwide2. Moreover, WHO also highlights that air pollution causes 6.7 million premature deaths every year. These facts and figures are important to understand the significance of air pollution as a public health issue affecting developing and developed nations alike. Air pollution has been linked to increased hospitalization and mortality rates due to its impact on various health issues, especially the lungs and the heart1,3. Particulate matter (PM2.5 and PM10), ozone (O3), sulfur dioxide (SO2), carbon monoxide (CO), and nitrogen dioxide (NO2) have been identified and reported as the most common air pollutants in the environment2.

Recent evidence has suggested the possible association of air pollution and the brain as a decline in cognitive functions has been noted in individuals in polluted regions. Cognitive function is a broad term that includes attention, perception, memory, learning, and executive functions domains. It is critical in determining individuals' quality of life, health, and mortality. Declines in cognitive function are linked to heightened dementia risks, diminished life quality, and increased healthcare costs. Environmental factors, notably air pollution, have emerged as significant influencers of cognitive abilities. Research indicates that air pollutant exposure can trigger processes that cause cognitive deterioration, such as inflammation (systematic and neuronal), oxidative stress, alterations in the blood–brain barrier, and cerebrovascular dysfunction4. The ever-expanding body of epidemiological research suggests a link between air pollution exposure and different cognitive outcomes with varied and sometimes contradictory results, warranting a comprehensive synthesis of evidence to explore the nature and extent of this association.

Previously, few systematic reviews5,6,7,8 have been done exploring the association between global cognition and air pollutants, and even fewer have been done with meta-analyses9,10. With the global increase in air pollutant levels, the need for conclusive evidence is paramount. This research, therefore, aims to synthesize data from multiple cohort and cross-sectional studies. The goal was to provide public health officials, researchers, and healthcare professionals with comprehensive insights into the impact of air pollution on cognitive health. Our main objective was to assess the effect of air pollutants, particulate matter (PM2.5, PM10), sulfur dioxide (SO2), and ground-level ozone (O3), on global cognitive functions. By focusing on global cognition rather than clinically diagnosed diseases, such as Alzheimer’s or neurodegenerative diseases, we aim to capture a broader spectrum of cognitive changes that might not yet meet the clinical threshold for a specific diagnosis.

Methods

This study was conducted in the “Department of Physiology, College of Medicine, King Saud University, Riyadh, Saudi Arabia.”

Data search strategy

For the selection of documents, we followed the PRISMA “Preferred Reporting Items for Systematic Reviews and Meta-Analyses” protocol. The literature search was performed, and data were collected using electronic platforms such as PubMed, Embase, Google Scholar, Web of Science, and Scopus to search the literature on the relationship between air pollutants and cognitive function. The keywords included a combination of two main items: exposure (air pollutants) and outcome (global cognitive function). We filtered the data using the key term’s “environmental pollution, air pollution, particulate matter, PM2.5, PM10, Ozone (O3), and Sulfur Dioxide (SO2), and cognitive function”. While initially using the term “environmental pollution and cognitive function,” 790 documents were identified; after screening and going through the abstracts and studying the detailed article, 21 studies were finally selected for a detailed analysis and discussion (Fig. 1).

Fig. 1
figure 1

PRISMA Flow Diagram for the selection of documents.

Full size image

Inclusion and exclusion criteria

The inclusion criteria were set as follows: Studies with exposures to air pollution (PM2.5, PM10, O3, and SO2); Outcome was global cognitive function measured by standardized cognitive tools; Effect size was reported in odds ratio (OR) or hazard ratio (HR); Original research; and the article language was written in English. The exclusion criteria were set as, studies on other air pollutants like smoking, or airborne metals, studies other than original articles, such as letters, editorials, brief communications, review articles or meta-analyses; reported effect sizes other than OR/HR and studies for which full article could not be found or accessed. Since we focused on global cognition rather than clinically diagnosed diseases, studies that reported the outcome of clinical behavioral, neurodevelopment, or central nervous system disorders like dementia, autism, Attention-deficit/hyperactivity disorder (ADHD), Alzheimer’s, and Parkinson’s were also excluded.

Data extraction

From each eligible study, we extracted the following data: author’s names, year of publication, type of study, population size (both overall and gender-specific), mean age of the participants, the cognitive tool used to assess the outcome, and the outcome measured by each study. Cognitive function impairment is a broad term; therefore, we also extracted information regarding how each study defined the criteria for cognitive impairment.

Information on the type of pollutant studied, the concentration of the pollutant for which the estimate was reported, and the risk estimates in the forms of odds ratio or hazard ratios, along with their subsequent 95% confidence interval were also recorded. When data for multiple models was presented within the study, we chose the fully adjusted models. If there was a choice of single pollutant and multiple pollutant models, we selected the single-pollutant model. The data was extracted and organized in an excel file.

Statistical analysis

The statistical analyses were performed using RStudio version 4.3.2 and package ‘meta. We focused on the odds ratio (OR) as our estimate for the association between air pollutants and cognitive function. Some of the studies reported their results as hazard ratios. As done in another study11, we considered them as OR and then performed subgroup analysis based on study design (cohort and cross-sectional). Studies reported ORs with different increments of the pollutants. We pooled all effects estimates for a fixed 10 μg/m3 increase in air pollutant concentration. Any other units and scales were harmonized using the conversion outlined as follows11: 1 ppm = 1000 ppb; SO2: 1 ppb = 64/22.4 mg/m3; O3: 1 ppb = 48/22.4 mg/m3. The following formula was used to standardize an odds ratio (OR) to a different unit of measurement to estimate the standardized risks: OR (standardized) = OR (original) Increment (10)/Increment (Original)11.

The overall effect was significant if the p value was < 0.05. Heterogenicity among the pooled studies was evaluated using the Cochrane chi-square test (Q) and I2. Moderate to a high degree of heterogenicity was indicated when the p value of the chi-square test was < 0.05 and the I2 value was ≥ 50%, and in that case, the random effect model was used12. The existence of publication bias was assessed by Egger’s Regression test and funnel plot. Sensitivity analysis was done to evaluate the reliability of this meta-analysis. RStudio version 4.3.2 was used to generate both.

Ethics approval

The data were recorded from publicly available data-based web sources and had no direct involvement of animals or humans; hence, ethical approval is not required.

Results

A total of 21 cohort studies13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33 focusing on various pollutants were identified. The combined population of these studies was 617,594 individuals. For gender-specific proportions within the studies that reported this measure, 51% of all the participants were male. For most of the studies, the participants belonged to the older age group (Average age: 71.78 years) (Table 1). The geographical distribution of the studies was diverse and included studies from multiple countries worldwide.

Table 1 Impact of PM2.5, PM10, SO2 and O3 on cognitive functions.
Full size table

Table 1 provides detailed information for all the studies including the author’s name, year of publication, sample size, a cognitive assessment tool used, evaluation of cognitive function impairment by each study, and the effect on cognitive functions. The ORs mentioned in the table are the original effect sizes extracted from the studies before standardization to 10 μg/m3 if it was needed.

Particulate matter (PM 2.5 µm)

17 studies encompassing both cohort and cross-sectional design, assessed the link between PM2.5 and the effect on cognitive function. These studies have been summarized in Table 1. An overall analysis of all study types was performed (Fig. 2), followed by separate cohort and cross-sectional studies analysis (Fig. 3). The Cochrane chi-squared test (Q) and I2 statistic revealed a significant heterogeneity for the overall analysis (Q = 128.48, p < 0.01, I2 = 88%), therefore, a random model was applied. Our forest plot for combined studies showed a significantly heightened risk between increased exposure to PM2.5 and incidence of cognitive impairment (OR 1.49; 95% CI 1.11, 1.99; p = 0.001) (Fig. 2). Results for publication bias, assessed using funnel plots and Egger’s regression test, are in the supplementary file.

Fig. 2
figure 2

Forest Plot: Impact of PM2.5 on global cognitive functions for all study designs.

Full size image
Fig. 3
figure 3

Study design stratified analysis for PM2.5 impact on cognitive function. (A) Cohort studies (B) Cross-sectional studies.

Full size image

For study design-specific analysis, we again used a random model due to significant heterogenicity (Fig. 3). There were 10 cohort studies and 7 cross-sectional studies. Cohort-only study analysis revealed a significantly increased risk of impaired cognition due to PM2.5 (OR 1.35, 95% CI 1.02–1.79). However, cross-sectional studies showed an increased but non-significant risk of impaired cognition following PM2.5 exposure (OR 2.66, 95% CI 0.79–8.90

Particulate matter (PM10 µm)

A total of 9 studies reported the impact of PM10 on cognitive impairment, the characteristics of which are all mentioned in Table 1. The Cochrane chi-squared test (Q) and I2 statistic revealed significant heterogeneity in the combined study design analysis (Q = 36.05 p < 0.01, I2 = 78%). Therefore, a random model was applied. The forest plot analysis for the overall studies suggested that increased exposure to PM10 was positively and significantly associated with an increased risk of cognitive impairment (OR 1.30; 95% CI 1.00, 1.70; p = 0.05) (Fig. 4). The statistical evidence from Egger’s test and the funnel plot regarding publication bias are mentioned in the supplementary file.

Fig. 4
figure 4

PM10 Forest plot of impact on global cognitive functions for all study designs.

Full size image

For study design stratified analysis, there were 5 cohort studies and four cross-sectional. For both, the random model was applied. Cohort study analysis and cross-sectional analysis both revealed a heightened but not significant impact of PM10 on cognitive functions (Fig. 5), with ORs and their subsequent 95% CI as being 1.37 (0.89–2.10) and 1.26 (0.68–2.33) respectively.

Fig. 5
figure 5

Study design stratified analysis for PM10 impact on cognitive function.

Full size image

Ozone (O3)

A total of 4 studies reported the impact of ozone (O3) on cognitive impairment, and all of them were cohort studies (Table 1). The Cochrane chi-squared test (Q) and I2 statistic revealed a significant heterogeneity (Q = 12.46 p < 0.01, I2 = 76%), and so a random model was used. The forest plot showed that increased exposures to O3 were not associated with an increased risk of cognitive impairment (OR 1.00; 95% CI 0.54, 1.87; p = 0.99). (Fig. 6). The supplementary file contains results from the Eggers regression test and funnel plot for publication bias.

Fig. 6
figure 6

Forest Plot: Impact of O3 on global cognitive functions.

Full size image

Sulphur dioxide (SO2)

A total of 3 cohort studies reported the impact of SO2 on cognitive impairment, as seen in Table 1. The Cochrane chi-squared test (Q) and I2 statistic revealed a non-significant heterogeneity (Q = 0.59 p = 0.75, I2 = 0%); therefore, a fixed model was used. The forest plot analysis showed that increased exposure to SO2 was positively and significantly associated with an increased risk of cognitive impairment (OR 1.39; 95% CI 1.27, 1.51; p < 0.01). (Fig. 7). A funnel plot was made for publication bias analysis, and Egger’s regression test was performed, as shown in the supplementary material file.

Fig. 7
figure 7

Forest Plot of Impact of SO2 on global cognitive functions.

Full size image

Sensitivity analysis

A leave-one-out sensitivity analysis was conducted to assess the overall results' robustness by determining each study's influence. For PM2.5 and PM10 this analysis was only done for the combined studies meta-analysis. The outcomes were similar when individual studies were excluded for each PM2.5, and SO2, which indicates that the meta-analysis findings were not heavily dependent on one study. The leave-one-out sensitivity analysis for O3 revealed that overall meta-analysis results were sensitive to the inclusion or exclusion of specific studies. The sensitivity analysis for the meta-analysis on PM10 shows moderate sensitivity to including particular studies. Excluding some studies leads to a non-significant pooled Odds Ratio (OR) with p values up to 0.14. Most of the other exclusions do not lead to a significant change, indicating robustness in the overall meta-analysis results. This implies that the meta-analysis is generally stable, with a few studies having a more pronounced effect on the findings. A detailed sensitivity analysis is outlined in the supplementary material file.

Discussion

The development of brain and cognitive functions starts from the gestation period and continues throughout life. The embryological period is very crucial, and growth of the brain begins in the early gestation period with the development of the neural tube. The neural tube then differentiates into three main divisions: the forebrain (prosencephalon), midbrain (mesencephalon), and hindbrain (rhombencephalon). These further divide into various structures of the brain34. Even after birth, the brain continues to grow and adapt. The brain significantly increases in size during pre-school; by age 6, it reaches about 90% of its adult volume34. There is a significant increase in the number and complexity of neural connections aiding cognitive, motor, and sensory development. The brain also undergoes extensive remodeling, pruning of synapses, and strengthening of critical neural pathways34. In adulthood, while the brain reaches its full size, it continues to change and adapt through a process known as neuroplasticity. Neuroplasticity is a general term for the structural and functional changes and adaptations the brain undergoes in response to experiences and learning35. As the individual ages and reaches late adulthood, there is a noticeable cognitive decline, often seen as worsened memory and struggle to perform complex tasks36. The brain shrinks in volume, especially the frontal cortex. Moreover, due to aged vasculature and a rise in blood pressure, there are increased incidences of stroke or brain ischemia, which cause lesions in white matter36.

Development and maintenance of normal cognitive functions is essential for a decent quality of life as it affects every aspect, including learning and education, work prospects, interpersonal relationships, and life experiences. While genetics are essential, environmental factors also play a detrimental role in the development. These factors include nutrition, physical activity, motor and sensory experiences, relationships, and stress, among other things. Another factor that is now gaining more recognition is air pollution. Due to rapid urbanization and industrialization, air pollution has emerged as a significant public health crisis globally. World Air Quality Report estimated that in 2019, 72.7% of Europeans and 98.8% of South Asians were exposed to PM2.5 levels that exceeded the standard levels of PM2.5 set in WHO’s guidelines37.

Prenatal and postnatal exposure to pollutants like PM2.5 can potentially lead to neurodevelopmental disorders and cognitive deficits in children. The literature38,39,40,41 suggests that exposure during early life to pollutants can have adverse effects on the development of children, including potential impacts such as diminished intelligence levels and impairments in cognitive and psychomotor development and behavioral changes. Chronic exposure to pollutants has also been shown to affect adults as it accelerates cognitive decline and increases the risk of developing dementia or neurodegenerative diseases like Alzheimer’s42.

Our study enforced the conclusions of other pieces of literature that concluded that increased levels of PM2.5, PM10, and SO2 are significantly linked to a decline in cognition. For example, a study conducted in China43 measured the effects of improved air quality on the brain following the implementation of the 2013 Clean Air Act. The findings showed that improved air quality and decreased pollutant levels in China had a protective effect on cognitive function. The uniformity of the results across different regions and populations emphasizes the potential toxic effects of PM2.5 on the brain. However, the mechanisms by which this may happen are not fully understood. Particulate matter has been theorized to impact the brain through multiple pathways, which can result in neuroinflammation, oxidative stress, and protein aggregation, resulting in neuronal cell death4. PM2.5 particles are 2.5 µm or less in diameter, including ultrafine particles (< 0.1 µm) and nanoparticles. This makes it easier for PM2.5 to reach the brain through different entry routes including inhalation, ingestion, and the olfactory system4. Inhaled PM2.5 particles can cause respiratory inflammation, which triggers systematic inflammation. Proinflammatory cytokines that may disrupt the blood–brain barrier (BBB) are produced, allowing the inflammatory agents to enter. Inhaled PM2.5 can also enter the systematic circulation and directly get transported to the brain by eventually crossing the blood–brain barrier. Nasally inhaled PM particles may enter the nervous system through the olfactory bulb. Once the PM reaches the brain, it induces cellular and molecular mechanisms that result in neuronal injury4. Firstly, there is neuroinflammation, characterized by the activation of microglial cells, resulting in the production and release of inflammatory cascade mediators, including reactive oxygen species (ROS), proinflammatory cytokines, and nitric oxide, among others4. Secondly, there is oxidative stress. Air pollutants can trigger increased production of ROS, resulting in damage to cellular components like DNA, protein, and lipids4. Elevated levels of ROS can create a disparity between ROS production and the body’s ability to detoxify or repair cellular damage, resulting in oxidative stress. Responses to oxidative stress can lead to changes in mitochondria or other organelles within the cell, like the endoplasmic reticulum (ER), and eventually activate the cell death pathway4. Additionally, aggregation of misfolded proteins in the brain can also cause cognitive decline as they can lead to neurodegenerative diseases4. ER is a crucial organelle involved in protein synthesis and folding. If, for any reason, there is a loss of cellular calcium homeostasis, the ER stress response is activated, which can result in the unfolded protein response4. In one study44 human neuroblastoma cells SH-SY5Y were exposed to PM2.5, and it was observed that the ER stress-induced apoptosis in those cells. This can also be a potential mechanism behind the cognitive decline caused by PM2.5. However, more studies are needed to establish this association.

Study strengths and limitations

This study has some strengths. First, this study investigates the impact of four major air pollutants particulate matter (PM2.5, PM10), sulfur dioxide (SO2), and ground-level ozone (O3), on global cognitive health. Second: the study inclusion and exclusion criteria were highly standardized, and results were thoroughly analyzed while including the cohort and cross-sectional studies. Third: the study analysis is based on 21 cohort and cross-sectional studies from diverse geographical regions, multiple countries with a combined population of 617,594 individuals. Fourth: the findings are important for academicians, researchers, and policymakers.

Similar to other studies our study has certain limitations. First: some evidence may have been missed or overlooked due to human error. Second: We could not include the other contaminants like metals they may also be assessed for the association. Third: our study focused on global cognition, the impact of air pollution on specific cognitive domains may differ from those of global cognition. Fourth: it was challenging to compare the studies directly because they used different cognitive measuring tools, and the definition of cognitive decline also varied across the studies. In some studies, the follow-up might not have captured the full impact.

Conclusions

The study findings conclude that exposure to air pollutants PM2.5, PM10, and SO2 was associated with a decline in global cognition. Therefore, reducing these pollutants' levels could be a strategic approach to mitigate population cognitive health risks. Policies aimed at reducing emissions from industrial, power plants, and vehicular sources could be particularly effective. Furthermore, awareness of the association between cognitive function and air pollution should be raised among the global population. Appropriate protective measures should be taken in highly polluted areas where the pollutant levels exceed the standards.