Lifetime air pollution exposure, cognitive deficits, and brain imaging outcomes: A systematic review

Abstract

As the amount of air pollution and human exposure has increased, the effects on human health have become an important public health issue. A field of growing interest is how air pollution exposure affects brain structure and function underlying cognitive deficits and if structural and connectivity changes mediate the relationship between the two. We conducted a systematic review to examine the literature on air pollution, brain structure and connectivity, and cognition studies. Eleven studies matched our inclusion criteria and were included in the qualitative analysis. Results suggest significant associations between air pollution and decreased volumes of specific brain structures, cortical thickness and surface area such as in the prefrontal cortex and temporal lobe, as well as the weakening of functional connectivity pathways, largely the Default Mode (DMN) and Frontal Parietal (FPN) networks, as detected by fMRI. Associations between air pollution and cognitive outcomes were found in most of the studies (n = 9), though some studies showed stronger associations than others. For children & adolescents, these deficiencies largely involved heavy reasoning, problem solving, and logic. For young and middle-aged adults, the associations were mostly seen for executive function and visuospatial cognitive domains. To our knowledge, this is the first systematic review to consolidate findings on the associations among air pollution, brain structure, and cognitive function. In the future, it will be important to conduct further longitudinal studies that follow children who have been exposed at a young age and examine associations with brain structure and cognition throughout adulthood.

1. Introduction

Air pollution is a large public health concern as a majority of the world experiences increasing exposure, and greater increases in exposure noted across the Middle East, Sub-Saharan Africa, and Central and Southern Asia (Shaddick et al., 2020). Sources of air pollutants include motor vehicles, industrial facilities, and forest fires. Elevated exposures to these air pollutants, ranging from carbon dioxide (CO₂) to sulfur dioxide (SO₂), can be detrimental to human health (WHO, 2021). Recently, an increasing number of studies have examined how air pollution exposures affect health outcomes across the lifespan, especially cognition. Much of the literature has shown an association between higher exposure to air pollution and worse performance on a range of cognitive domains (Sakhvidi et al., 2022, Zhang et al., 2018, Shehab et al., 2019).

Cognitive decline and even mild cognitive impairments (MCI), which is defined as older adults who have more cognitive and memory deficits than others that are also in their age group (National Institutes of Aging, 2021), have been shown to preface, accelerate, or occur alongside with other health issues, in diseases like Parkinson’s, Alzheimer’s, and dementia (Wilson et al., 2012, Watson et al., 2010, Aarsland et al., 2021, Pal et al., 2018, Campbell et al., 2013). Cognitive domains and their functions can be assessed through different ways, including testing sensation, motor skills, or global cognition and function (Harvey et al., 2019). Understanding the association between air pollution and cognitive functioning is necessary to determine how air pollution may affect future health outcomes and neurocognitive disease diagnoses.

Non-invasive neuroimaging techniques, such as magnetic resonance imaging (MRI), have been widely used for assessing both cognitive function and overall brain health across the lifespan, including in childhood (Copeland et al., 2021). Furthermore, MRI has been shown to be effective in identifying biomarkers of MCI (Chandra et al., 2019) such as structural brain changes and functional connectivity disruptions in the Default Mode Network (DMN) pathway. MRI and functional magnetic resonance imaging (fMRI) techniques have allowed researchers to study associations between air pollution and the structure and function of the brain. A recent longitudinal study showed that in children followed longitudinally from <6 months annually until 12 years of age, exposure to air pollution was associated with reduced gray matter volume and cortical thickness in several areas of their brain at 12 years of age (Beckwith et al., 2020). A recent systematic review focused on specific brain structures like the hippocampus and found that air pollution exposure was negatively correlated with hippocampal volume (Balboni et al., 2022).

A relatively new field of study has emerged connecting all three of these areas: air pollution, brain structure through MRI imaging, and cognitive outcomes; however, to the best of our knowledge, there has not been a comprehensive systematic review conducted to analyze these associations. This is an expanding and important field of health science; if we are able to draw connections between brain structural/connectivity changes and cognitive outcomes, and how both are linked to air pollution, we will have better information about possible mechanistic pathways between air pollution and health outcomes, which can be used for early diagnosis and intervention.

Thus, given the importance of understanding the mechanisms and neuroscience behind the air pollution and associated changes in cognitive outcomes, we conducted this systematic review in order to fill necessary gaps in the literature. Overall, the aim of our systematic review was to identify the associations among air pollution, brain structures and connectivity, and cognition across the life course and determine gaps in the literature for future research.

2. Methods

This systematic review was conducted by following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. We included studies examining associations between air pollution exposure, neuroimaging, and cognitive outcomes across the life course.

2.1. Search Strategy

We performed an extensive literature search using keywords and MeSH terms in 5 databases: PubMed, Cochrane, Web of Science, Scopus, and PsycINFO. We searched for literature that included key terms of associations between air pollution exposure, brain imaging, and cognition and cognitive outcomes. Key terms for our search included “air pollut*” or “PM10” or “PM2.5” “MRI” or “magnetic resonance imaging” or “neuroimaging” and “cognition” or “cognitive outcome”. The detailed search strategy is outlined in Supplemental Materials.

2.2. Eligibility

2.2.1. Inclusion criteria

We included all age groups in our study to better understand the association between air pollution exposure, brain structure and function, and cognitive function throughout the life course. Infants, children and adolescents, young adults, and older adult/senior populations were included. Cross-sectional, longitudinal, and cohort studies qualified as eligible too. Studies that examined air pollution and MRI outcomes, which were then linked to cognitive function of individuals in the study, were eligible for inclusion. Studies did not need to perform true mediation analysis to be included. This was done as there were too few studies testing true mediation, however many studies included air pollution, brain imaging, and cognition without testing mediation, which were included. We included any studies that tested at least one cognitive domain or function, for example, they could test only episodic memory, or could test all the cognitive functions globally.

2.2.2. Exclusion criteria

We excluded studies that did not focus on at least one air pollutant, such as articles studying polycyclic aromatic hydrocarbons. Furthermore, if studies analyzed cognitive outcomes with other neuroimaging techniques (e.g., EEG, PET, CT scans), but did not include MRI outcomes, they were excluded. This was done to make the results and implications of the included studies more comparable. If a study examined effects of air pollution on the brain through neuroimaging but did not assess cognitive outcomes at the same time, then it was excluded from this systematic review. Details on study selection, data extraction, and quality assessment can be viewed in Supplemental Materials.

3. Results

The initial search yielded 5,365 papers. Duplicates were removed and the remaining 4,169 papers were screened using our inclusion and exclusion criteria. After title and abstract screening, 21 full text papers remained. Full text screening resulted in 11 empirical articles (7 adults and 4 child and adolescent papers) to be included in this systematic review. The other ten studies were excluded for several different reasons, some of which include a lack of testing on all three variables of air pollution, brain imaging, and cognitive outcomes (n = 10). Other studies were excluded for not using the target neuroimaging technique of MRI. Details on the screening and inclusion/exclusion process can be viewed in Figure 1 below.

Figure 1.

PRISMA flowchart (Page et al., 2021) detailing the search process and identification of studies to include in our systematic review.

3.1. Study Populations

Study characteristics are described in Table 1. Out of the total 11 studies, 4 of the studies included participants under the age of 18. Three of the 4 studies included participants that were in their earlier years of childhood, with a mean age ranging from 7–10 years old (Calderón-Garcidueñas et al., 2011, Pujol et al., 2016, Cserbik et al., 2020). One study examined fetal air pollution exposure and then performed both the MRI scans and cognitive testing on the children during the ages of 6–10 years (Guxens et al., 2018). The other seven studies included adults; within this age group 1 study included young adults ranging from 18–49 years (Calderón-Garcidueñas et al., 2022), 2 studies included middle-aged adults ranging from 50–65 years (Crous-Bou et al., 2022; Nußbaum et al., 2020), and 4 studies included older adults, whose average age was greater than 65 years old and were studied for cognitive decline (Younan et al., 2020; Chen et al., 2017; Wong et al., 2020; Semmens et al., 2013). A detailed summary of major findings as well as similarities and differential findings in each of these age groups can be found in Tables 2 and 3.

Table 1.

Characteristics of studies included in the systematic review

Age Group	Authors and study name	Population Description	Country	Study Design	Air pollution exposure	Neuroimaging methods	Cognition measurement methods	Covariates
Children	Calderón-Garcidueñas et al. 2011 Exposure to severe urban air pollution influences cognitive outcomes, brain volume and systemic inflammation in clinically healthy children	n = 30 20 from Mexico City (mean age = 7.1 years), 10 from Polotitlán (mean age = 6.8 years)	Mexico	Cohort	Monitoring station levels PM2.5 levels for MC: 35.89 μm/m3 Control city of Polotitlán has levels below the current US standards.	All 30 underwent MRI scan 1.5 Tesla 5T Signa Excite HD MR 2 MRI scans taken in baseline and follow-up year	Wechsler Intelligence Scale for Children which includes subscales that test for Information, Arithmetic, Vocabulary, Digit Span, Picture Completion, Coding	Age, socioeconomic status, maternal education
	Guxens et al. 2018 Air Pollution Exposure During Fetal Life, Brain Morphology, and Cognitive Function in School-Age Children	n = 783 Population based birth cohort	The Netherlands	Longitudinal	Calculated using land-use regression models NO2 average: 39.3 μg/m³ Fine particles average: 20.2 μg/m³	Structural MRI scans using 3T scanner Scans performed between 6–10 years of age	Developmental Neuropsychological Assessment (Dutch version) Subtasks to test attention and executive functioning, language, memory and learning, sensorimotor function, visuospatial processing	Age, socioeconomic status
	Pujol et al. 2016 Traffic pollution exposure is associated with altered brain connectivity in school children	n = 263 BREATHE project Average age = 9.7 years	Spain	Case control	Elemental carbon in particulate matter, and nitrogen dioxide measured using dosimeter	1.5 Tesla Signa Excite system Both MRI, fMRI and Diffusion tensor imaging (DTI) obtained	Assessment included working memory, motor response speed, attention	Level of parental education, unemployment, and occupation
	Cserbik et al. 2020 Fine particulate matter exposure during childhood relates to hemispheric specific differences in brain structure	n = 11,875 Adolescent Brain Cognitive Development (ABCD) 9.93 average age 21 study sites	USA	Longitudinal	Measured PM2.5 pollutant levels Spatiotempor al PM2.5 models using residential addresses and ensemble-based model approach	3T scanners Cortical surface reconstruction and subcortical segmentation completed with FreeSurfer	NIH Toolbox Cognition Battery (NIHT) List Sorting Working Memory Test, Flanker Attention Test, Dimensional Card Sorting Task, Picture Vocabulary Test, Oral Reading Recognition Test, Picture Sequence Memory Test, Pattern Comparison Processing Speed Test	Age, sex, race and ethnicity, parental higher education, total family income
Young Adults (18–49 years)	Calderón-Garcidueñas et al. 2022 Hemispheric Cortical, Cerebellar, and Caudate Atrophy Associated to Cognitive Impairment in Metropolitan Mexico City Young Adults Exposed to Fine Particulate Matter Air Pollution	n = 302 Average age of 32.7 years old	Mexico	Longitudinal	Spatiotemporal modeling from monitoring stations Residents exposed to PM2.5 Average exposure was 30.9 μg/m3	1.5 Tesla 5T Signa Excite HD MR Scanned the entire surface of the brain, including white matter, gray matter, cerebrospinal fluid, subcortical structures	Montreal Cognitive Assessment in the Spanish version	Age, sex, years of formal education, BMI MoCA scores, cognition indexes, socioeconomic status
Middle-Aged Adults (50–65 years)	Crous-Bou et al. 2020 Impact of urban environmental exposures on cognitive performance and brain structure of healthy individuals at risk for Alzheimer’s dementia	n = 958 Alzheimer and Families study Middle age, cognitively unimpaired subjects Average age = 56.3 years	Spain	Prospective cohort	Land Use Regression models to estimate residential exposure to air pollutants	Structural brain imaging performed in a subsample of participants (n = 228) 3.0 T scanner	Episodic memory and executive function tests MBT (Memory Binding test), WAIS (Wechsler Adult Intelligence Scale)	Age, sex, socioeconomic factors
	Nußbaum et al. 2020 Associations of Air Pollution and Noise with Local Brain Structure in a Cohort of Older Adults	n = 615 1000BRAINS Study 61.5 years old average	Germany	Cohort	Land use regression models PM10 exposure: 27.5 PM2.5: 18.3	3T Siemens MRI scanner	Testing attention, executive functions, memory, short term and working memory, language functioning	Socioeconomic status, lifestyle variables, physical activity, age, sex, education
Older Adults	Younan et al. 2020 Particulate matter and episodic memory decline mediated by early neuroanatomic biomarkers of Alzheimer’s disease	n = 998 All older females, ages ranged from 73–87 Women’s Health Initiative Study of Cognitive Aging and Women’s Health Initiative Memory Study of Magnetic Resonance Imaging	USA	Prospective cohort study	Estimated a 3-year average of PM2.5 exposure using residential address and geocoding	MRI scanner	California Verbal Learning Test- free recall and new learning, and delayed recall	Race/ethnicity, socioeconomic factors, family income, employment status, lifestyle factors, smoking status, physical activity, clinical characteristics
	Chen et al. 2017 Particulate Air Pollutants, Brain Structure, and Neurocognitive Disorders in Older Women	n = 7,479 All older women Average age = 71.0 years	USA	Longitudinal	Tested with both PM2.5 and Diesel PM exposure Spatiotemporal modeling	WHIMS-MRI Study	Modified mini-mental State Examination and a modified Consortium to Establish a Registry for Alzheimer’s Disease (CERAD) battery. Tested neurocognitive outcomes, cognitive impairment and dementia, brain volume measures	Age, race, socioeconomic status, smoking, alcohol use, physical activity, body mass index, clinical characteristics
	Wong et al. 2020 Indoor incense burning impacts cognitive functions and brain functional connectivity in community older adults	n = 515 Community older adults without stroke of dementia Greater than 65 years	Hong Kong	Prospective, longitudinal study	Measure levels of 6 outdoor air pollutants: NO2, O3, SO2, nitrogen oxides from incense	3 Tesla Philips MRI scanner with 8-channel head coil fMRI images obtained for select number of users who matched criteria	Hong Kong version of MoCA- covers learning and memory, executive and visuospatial functions, language, attention, working memory, abstraction/orientation Total score - index of global cognitive functions	Age, sex, education, vascular
	Semmens et al. 2013 Effects of Traffic Related Air Pollution on Cognitive Function, Dementia Risk, and Brain MRI Findings in the Cardiovascular Health Study	n = 5,888 participants n = 2,116 for brain imaging Cardiovascular Health Study (CHS) 65+ years old	USA	Cohort	Geocoding model PM10 and NO2	T1 weighted localizer images, axial T1, spin density and T2 weighted images All participants invited to undergo brain MRI scans twice during follow up (n = 2,116 underwent)	Assessed cognitive decline with Modified Mini-Mental State Examination and Digit Symbol Substitution Test	Age, gender, enrollment center, education, race, income, BMI, physical activity, physical function, depression, smoking status, other health data (hypertension, diabetes, prevalent cardiovascular disease at baseline)

Table 2.

Major Findings for Included Studies

Authors and Study Name	Major Findings	Strengths	Limitations
Younan et al. 2020 Particulate matter and episodic memory decline mediated by early neuroanatomic biomarkers of Alzheimer’s disease	Imaging outcomes: Air pollution was significantly associated with increasing AP-PS scores which measure gray matter loss in areas that are vulnerable to AD including amygdala, hippocampus, parahippocampal gyrus, thalamus, inferior temporal lobe areas, and the midbrain. Cognitive outcomes: Air pollution was significantly associated with declines in episodic memory (immediate recall, new learning) - precursor to dementia risk.	Participants were followed for a long period of time with annual assessments Advanced parallel-process mediation SEM approach Alzheimer’s-PS scores could capture the spatial patterns for AD more specifically The cohort of women was very geographically diverse and the data was comprehensive	Could not estimate early or mid-life exposures to PM2.5 Measurement errors since individual data not available on the time-activity patterns Only focused on regional PM2.5 levels Did not examine multiple mediation pathways Cannot rule out other confounding variables from environmental factors Not generalizable to other population groups
Chen et al. 2017 Particulate Air Pollutants, Brain Structure, and Neurocognitive Disorders in Older Women	Imaging outcomes: Cumulative PM2.5 exposure significantly associated with smaller brain volumes in frontal, parietal, temporal regions and corpus callosum, but not in gray matter and not for the hippocampus. Participants in highest areas of exposure had smaller gray matter volumes. Cognitive outcomes: No significant associations between air pollution and development of MCI or dementia.	Included the WHIMS study, which is very comprehensive and large WHIMS protocols were geographically diverse, comprehensive, and overall high quality	Brain volume assessment only occurred once Only studied older women Only able to use geostatistical modeling for PM2.5 Did not include genetic determinants of brain structure and dementia Selective and small sample Only focused on certain regions of interest for the brain structures
Calderón-Garcidueñas et al. 2022 Hemispheric Cortical, Cerebellar, and Caudate Atrophy Associated to Cognitive Impairment in Metropolitan Mexico City Young Adults Exposed to Fine Particulate Matter Air Pollution	Imaging outcomes: Air pollution significantly associated with decreased brain volumes within the temporal lobe, cingulate cortex, frontal superior lobes. Additional significant associations with decreases in gray and white matter volume in right cerebellar, left cerebellar, right and left caudate. These findings were only statistically significant for the left brain regions. Cognitive outcomes: Significant inverse associations between air pollution and language domain cognitive functions; lower verbal working memory, but higher visual-spatial working memory.	Were able to establish associations between Alzheimer’s, Parkinson’s, ad TDP-43 pathology The cohort was comprehensive on the population of interest. Individuals were clinically healthy and educated	Small sample size Did not include the control group for MoCA
Wong et al. 2020 Indoor incense burning impacts cognitive functions and brain functional connectivity in community older adults	Imaging outcomes: Functional connectivity weakness in the DMN specifically precuneus, medial frontal gyrus, left angular/right middle temporal gyrus - no specific structural brain changes of biomarkers related to AD. Cognitive outcomes: MoCA total score, executive functions, visuospatial functions, memory were all significantly negatively associated with air pollution.	Large, well-defined sample Combined cognitive, structural, and functional imaging data Confounding variables were comprehensive and included the influence of outdoor air pollutants	Specific, detailed information for each individual on incense use and exposure was not recorded Past exposure to incense smoke not recorded No real-time indoor air quality monitoring Neuropsychological and neuroimaging not repeated at follow-up
Crous-Bou et al. 2020 Impact of urban environmental exposures on cognitive performance and brain structure of healthy individuals at risk for Alzheimer’s dementia	Imaging outcomes: Nitrogen oxide exposure and PM exposure significantly associated with decreased volume in entorhinal, inferior and middle temporal, fusiform, posterior cingulate, precuneus and supramarginal gyri. No significant associations with noise. Cognitive outcomes: Marginal negative associations between air pollution and episodic memory and executive function.	Participants were from very well characterized study, detailed outcomes and information on covariables Spatial distribution estimates of air pollution in Barcelona have been steady for the past 20 years	Relatively small sample size, possibly not fully representative of Barcelona Participants mostly had high range scores in the different tests, detection on weak effects may be difficult No access to workplace exposures and daily mobility Focus only on multiple exposure analyses
Nußbaum et al. 2020 Associations of Air Pollution and Noise with Local Brain Structure in a Cohort of Older Adults	Imaging outcomes: Air pollution exposure was significantly negatively associated with posterior brain regions more than frontal brain regions, including dorsolateral prefrontal cortex. Significant negative associations between weakening fronto-parietal network (FPN) language and memory domains. Significant associations between air pollution exposure and decreasing Local gyrification index (loss of connection of the neurons in these brain regions). Both air pollution and noise were significantly associated with local atrophy of fronto-parietal network (FPN) in the right brain, but not the left brain. Cognitive outcomes Significant associations between air pollution and decreased cognitive performance, specifically when mediated by caudate, left orbital gyrus atrophy- specifically executive function, language, visuospatial, attention.	Included both sexes Comprehensive data for covariates, confounding variables, and the exposures First study looking at air pollution and 1GI for local brain atrophy Extensive sensitivity analyses Investigated structural differences during the period of brain aging	Small sample size Selection bias geared towards younger, well-educated men No information about the when cognitive function and brain morphology changes occurred during follow-up Some of the associations could have been due to chance because of the sheer number of analyses conducted
Semmens et al. 2013 Effects of Traffic Related Air Pollution on Cognitive Function, Dementia Risk, and Brain MRI Findings in the Cardiovascular Health Study	Imaging outcomes: White matter grade (lesions) significantly higher for those individuals exposed to longer periods of air pollution. Cognitive outcomes: Long-term exposure to PM10 and NO2 significantly associated with faster rates of cognitive decline and risk of vascular dementia.	Serial brain MRI imaging Individual level estimates of exposure Taking into consideration other possible confounding variables	Small sample size Population for the MRI scan may not be representative of the general population
Calderón-Garcidueñas et al. 2011 Exposure to severe urban air pollution influences cognitive outcomes, brain volume and systemic inflammation in clinically healthy children	Imaging outcomes: White matter hyperintensities significantly associated with high exposure to air pollution, and cognitive deficits. Greater white matter volume in right temporal/parietal and left temporal lobes for control groups not exposed to high levels of air pollution. Cognitive outcomes: Higher air pollution exposed children were more likely to have decreased abilities in tests that involved complex reasoning, diversity of forms of knowledge, basic verbal/working memory skills most affected including learning, short term memory, attention	Longitudinal study Tested cognitive outcomes twice, at two different time points, to get more comprehensive data. Statistical analysis conducted Included relatively even split of both sexes.	Small sample size.
Guxens et al. 2018 Air Pollution Exposure During Fetal Life, Brain Morphology, and Cognitive Function in School-Age Children	Imaging outcomes: Air pollution significantly associated with thinning of the gray matter of precuneus region (right temporal) rostral middle frontal region (dorsolateral prefrontal cortex), superior frontal region (large part of the frontal region, right), cuneus region (left occipital), fusiform region (temporal left region). Cognitive outcomes: Air pollution significantly associated with impaired inhibitory control.	Large sample size Prospective, longitudinal study Air pollution estimates were created at the individual level, they were thus more accurate	Residual confounding Bias due to the recruitment of children who had exposure/outcome data had mothers with higher socioeconomic status
Pujol et al. 2016 Traffic pollution exposure is associated with altered brain connectivity in school children	Imaging outcomes: Air pollution significantly associated with weaker functional connectivity within the DMN region. Increasing age was significantly negatively associated with functional connectivity strength at the boundaries of the DMN. Cognitive outcomes: Air pollution significantly associated with lower motor response speed. Those with faster reaction times had a stronger DMN network and weaker connectivity in the frontal operculum.	Very consistent functional connectivity findings Comprehensive cognitive assessment that included motor response speed. Included an even split of both sexes.	Introduced rigorous ways to control head motion, but this may have led to spurious changes Did not have the option of using 3-Tesla option for the MRI so a higher MRI signal was not obtained Potential confounding from sociodemographic variables
Cserbik et al. 2020 Fine particulate matter exposure during childhood relates to hemispheric specific differences in brain structure	Imaging outcomes: Increase in PM2.5 exposure significantly associated with smaller surface area in left cuneus and right frontal pole, but an increased surface area in the right lateral orbitofrontal areas. Increase in PM2.5 significantly associated with thinning of left superior frontal, left orbital frontal, left cingulate cortex, right inferior temporal, right parahippocampal, right insula, but an increase in thickness for right lateral orbitofrontal, right paracentral, right caudal anterior and posterior cingulate, left middle temporal cortex. Increase in PM2.5 exposure significantly associated with increase in right thalamic volumes, right pallidum, left accumbens, but a decrease in left putamen, left pallidum No significant associations for the whole-brain surface area, cortical thickness, and cortical or subcortical volumes. Cognitive outcomes: No significant associations.	Large, diverse sample Air pollution estimated at the individual level with precision Adjusted for demographic and socio-economic confounders Complete structural evaluation in the MRI	Cross-sectional design limited causal inference No prenatal exposure assessment No real-time air pollution exposure measurements No estimates of other regional pollutants other than PM2.5 or roadway proximities

Table 3.

Summary of Major Results Across Age Groups

Age Group	Sources	Imaging major findings/similarities	Cognitive outcomes major findings/similarities	Differential findings
Children	Calderón-Garcidueñas et al., 2011; Guxens et al., 2018; Pujol et al., 2016; Cserbik et al. 2020	White matter hyperintensities in the temporal and parietal lobes, as well as gray matter cortical thinning, volume and surface area decrease, especially in the frontal and prefrontal cortex. Functional connectivity also mirrored these trends, with weakening connectivity of the DMN region. Regions associated most with air pollution included precuneus region, prefrontal cortex, cingulate cortex, temporal and frontal cortical areas. There were negative associations between air pollution and brain regions responsible for reasoning, problem solving, learning such as response speeds and inhibitory control	Executive function, learning, and motor cognitive domains were the cognitive domains most associated with air pollution.	Cserbik et al., 2020 found no significant associations of air pollution exposure and neurocognitive performance, however, reported associations with structural changes which may predict future cognitive outcomes.
Young Adults	Calderón-Garcidueñas et al., 2022	Air pollution was significantly associated with decreases in cortical thickness and gray matter in left side, temporal regions of the brain. Correlated with signatures for AD.	Most impacted cognitive functions were in the language and working memory domains. Significant negative associations between air pollution and these specific functions were seen through extensive MoCA tests.	Results within the memory domain varied.
Middle-Aged Adults	Crous-Bou et al., 2020; Nuβaum et al., 2020	Significant associations between air pollution and loss of connection of neurons and weakening of frontal brain networks such as the FPN were seen, including areas in the frontal and cingulate cortices.	Visuospatial cognitive domain impairment (e.g., orientation) was observed in both studies.	Crous-Bou et al., 2020 reported that greater air pollution was associated with improved episodic memory, which could be partially explained by considering sample bias with all healthy, non-cognitively impaired middle-aged adults.
Older Adults/Elderly	Younan et al., 2020; Chen et al., 2017; Wong et al., 2020; Semmens et al., 2013	Air pollution significantly associated with gray and white matter volume loss in temporal regions (hippocampus and parahippocampal) and corpus callosum. White matter lesions were seen throughout these regions in addition to volume loss. Specific regions were measured as a precursor to understanding if the individual would develop AD. Similar functional connectivity disruptions in the frontal and temporal cortices through the DMN network significantly associated with air pollution.	Overall studies reported significant associations between air pollution and heightened risk of developing dementia and/or cognitive decline. Many studies reported significant associations between air pollution and declines in memory across several studies, specifically episodic memory. One study also showed decreases in visuospatial and executive functions.	Chen et al. 2017 did not find statistically significant increased likeliness of developing mild cognitive impairment or dementia with increasing PM exposure.

3.2. Air pollution

3.2.1. Air pollutants

Most of the studies measured only one form of air pollution exposure, the most common being particulate matter (PM), specifically PM_2.5 (Younan et al., 2020; Chen et al., 2017; Calderόn-Garcidueñas et al., 2022; Crous-Bou et al., 2020; Nußbaum et al., 2020; Calderόn-Garcidueñas et al., 2011; Guxens et al., 2018; Pujol et al., 2016; Cserbik et al., 2020). Two of the studies also measured PM₁₀ exposure (Nußbaum et al., 2020; Semmens et al., 2013), whereas Semmens et al. 2013 also studied NO₂. One study focused on indoor incense burning and measured common forms of both indoor and outdoor pollutants that included NO₂, O₃, SO₂, and other nitrogen oxides (Wong, 2020). Two of the studies focusing on children measured NO₂ pollutant levels in addition to PM (Guxens et al. 2018, Pujol et al. 2016), with one of these studies also including elemental carbon (Pujol et al. 2016). In addition to measuring PM2.5 levels, one study measured diesel PM exposure (Chen et al. 2017).

3.2.2. Air pollution exposure measurements

A majority of the included studies used residential addresses of the participants, satellite data, and geocoding in order to determine concentrations of air pollutants collected via government data on regional air quality. Different methods to map the data to residential addresses were used. For example, when using US Environmental Protection Agency (EPA) data on study employed the Bayesian maximum entropy modeling technique (Younan et al., 2020) and two studies utilized spatiotemporal modeling to estimate the exposure metrics (Chen et al., 2017, Cserbik et al., 2020). Another popular method of measuring air pollutant levels was using land use regression and chemistry transport models which are standardized based on protocols such as from the European Study of Cohorts for Air Pollution Effects (Nußbaum et al., 2020; Crous-Bou et al., 2020; Guxens et al., 2018; Cserbik et al., 2020). Wong et al., 2020 was a unique case as this study measured incense use as well as six outdoor pollutants. Incense use was recorded based on self-reports from participants, and outdoor air pollutant levels for the same study were generated using public data from the government on regional air quality. Finally, a passive individually worn dosimeter was used in one study, specifically measuring NO₂. (Pujol et al., 2016). One study estimated the concentration of PM_2.5 through a three-year period and averaged this exposure (Younan et al., 2020), and one study collected data on average 24-hour exposure (Calderón-Garcidueñas et al., 2022, Calderón-Garcidueñas et al., 2011).

3.2. Mediation studies

Out of the 11 studies, only 2 performed statistical analysis to examine MRI observed brain structures a mediator for associations between pollution and cognitive functioning task results (Guxens et al., 2018; Younan et al., 2020). One study in children (Age = 6 −10 years) reported that reduced cortical thickness in the frontal regions of the brain mediated the relationship between air pollution and inhibition errors (Guxens et al., 2018). The remaining study reported significant mediation, whereby structural changes and biomarkers signaling vulnerability to Alzheimer’s disease mediated the relationship between air pollution and worsening cognitive test scores (CVLT) (Younan et al., 2020).

3.3. Neuroimaging

3.3.1. Imaging methods

All studies included in this systematic review incorporated brain neuroimaging outcomes collected via MRI. Five used 3 Tesla (3T) scanners to obtain images (Cserbik et al., 2020, Guxens et al. 2018, Wong et al., 2020, Nußbaum et al., 2020, Crous-Bou et al., 2020), and two used 1.5T and 5T scanners (Calderόn-Garcidueñas et al., 2022, Calderόn-Garcidueñas et al., 2011) to obtain MRI data. Two studies additionally included fMRI data, in order to garner more data on the functional connectivity changes of the brain (Wong et al., 2020, Pujol et al., 2016). Three studies did not include the type of scanner they used because the scans were taken from previous published literature involving the same study (Younan et al., 2020; Chen et al., 2017; Semmens et al., 2013).

3.3.2. Neuroimaging results

3.3.2.1. Cortical thickness, surface area, and volume

Overall, 8 out of the 11 studies investigated cortical thickness, surface area, or brain volume. Five studies included adults and 3 included children. First, for the adult population, 3 in total (2 studies of older adults/elderly and 1 study of young adults) found that increased and long-term exposure to air pollutants resulted in reduced brain volume and cortical thickness, particularly in the frontal lobe (Calderόn-Garcidueñas et al., 2022, Chen et al., 2017, Younan et al., 2020). A similar trend was found in Calderόn-Garcidueñas et al., 2022 (Mean age = 32.7 years), which showed significant decreases in cortical thickness in the left brain regions, temporal, cingulate and frontal superior lobes as air pollution increased; these areas are also associated with the development of Alzheimer’s (AD) and Parkinson’s disease. This result was echoed in Crous-Bou et al., 2020 (Mean age = 56.3 years), with the observed lower cortical thickness in similar areas that are related to AD signatures. Two studies also measured white and gray matter volumes (Calderόn-Garcidueñas et al., 2022, Chen et al., 2017), while one study measured white matter grade (Semmens et al., 2013), with both finding overall decreases in white or gray matter volumes, and worse grades in white matter, as air pollution increased.

Studies with children reported similar results. Calderόn-Garcidueñas et al., 2011 (Mean age ~7 years) observed white matter volume differences in several regions of the brain, most prominently the temporal lobe on both sides of the brain and the parietal lobe on the right side when comparing children from different cities with differing values of air pollution. Specifically, overall decreased volumes were observed for increasing air pollution exposure. Guxens et al., 2018 (Age = 6–10 years) looked at cortical thickness and found thinning of the cortex in both hemispheric regions of the brain when exposed to higher levels of PM. Similarly, changes in cortical thickness were further evidenced by Cserbik et al., 2020 (Mean age = 9.93 years) which reported that higher exposure led to smaller cortical thickness in 22 out of 27 areas of interest regions in the brain. They also reported negative associations of air pollution exposure with overall surface area. This finding, however, was reported in fewer regions than for cortical thickness.

3.3.2.2. Functional connectivity

The remaining three studies examined functional connectivity and brain networks. Two of these studies focused primarily on the older adult population (Wong et al., 2020, Nußbaum et al., 2020). Functional connectivity was measured in Wong et al., 2020 and no group difference between those with and without exposure was found overall for white matter; however, there was shown to be functional connectivity changes that impacted global cognition results for participants. In Nußbaum et al., 2020, there was significant evidence to show that the frontal-parietal network, which is important for both language and memory cognitive domains (Nußbaum et al., 2020), had reduced connectivity associated with increased air pollution exposure. The final study in children echoed that of adult studies. Pujol and colleagues found that higher exposure to air pollution is significantly correlated with weaker functional connectivity, specifically for the DMN (Pujol et al., 2016), as found in the previous study from Nußbaum et al., 2020.

3.4. Cognition

3.4.1. Cognitive outcome measurements

The domains of cognitive outcomes examined in participants varied significantly across the included studies. For adults, outcomes focused primarily on cognitive decline. Younan et al., 2020 used the California Verbal Learning Test (CVLT) which examined recall and new learning, all falling under the umbrella term of episodic memory (Younan et al., 2020). The Memory Binding Test and Wechsler Adult Intelligence Scale were used in order to test episodic memory and executive function for one study (Crous-Bou et al., 2020). Nußbaum et al., 2020 had all participants complete a neuropsychological assessment based on paper/pencil. Two studies used a validated cognitive test, the Montreal Cognitive Assessment (MoCA). Calderόn-Garcidueñas et al., 2022 used the Spanish version and calculated the total MoCA scores for all participants, specifically focusing on orientation and executive. Finally, Semmens et al., 2013 utilized the Modified Mini-Mental State Examination and Digit Symbol Substitution Test to test cognition as well as diagnose vascular dementia.

When testing cognitive functions of children, assessments focused more on cognitive development and impairment tests. The Wechsler Intelligence Scale for Children was used by Calderόn-Garcidueñas and others; this instrument employs varying assessments for full-scale intelligence and cognitive functioning (Calderόn-Garcidueñas et al., 2011). Two similar tests, the Developmental Neuropsychological Assessment and the NIH Toolbox Cognition Battery (NIHT), were used in separate studies which investigated child populations, each focused-on areas of attention, memory, processing speed and learning (Guxens et al., 2018, Cserbik et al., 2020). Pujol et al., 2016 also assessed similar areas, in addition to motor response speed using the “Attentional Network Test.”

3.4.2. Cognitive results

Eight studies examined associations between air pollution and cognitive results in adults. Two of these studies showed significant associations between air pollution and episodic memory, in which episodic memory decreased as air pollution increased (Younan et al., 2020; Crous-Bou et al., 2020). Younan et al. 2020 found that immediate recall and new learning was significantly associated with increased long-term exposure to PM_2.5. Nußbaum et al. 2020 and Wong et al., 2020 found air pollution was associated with disruptions in the functional connectivity of the brain that thus impacted the overall global cognitive function, memory, executive functions, and visuospatial functions. Wong et al., 2020 and colleagues did not report any significant associations between air pollution and language and memory domains. However, Nußbaum et al., 2020 showed that higher air pollution exposure was associated with worse results in the language domain and lower verbal working memory. One study did not find a significant association between air pollution exposure and overall cognitive function (Chen et al., 2017). However, Semmens et al., 2013 reported a 10 μm/m³ increase in estimated PM₁₀ long-term exposure was associated with 2.45-fold increase in the risk of developing prevalent vascular dementia.

Four studies examined associations between air pollution and cognitive outcomes in children. Calderόn-Garcidueñas et al., 2011 showed that children exposed to higher levels of air pollution had more impairments in attention, short term memory, and learning domains of cognitive performance. Decreased inhibitory control, which is an important executive function, has been linked to issues of addiction or other mental health conditions (Jentsch et al., 2014). Increasing air pollution was associated with this decreased inhibitory control, as reported by Guxens et al., 2018. One study reported significant results for motor response speed, with a significant association between increasing air pollution exposure and slower reaction time (Pujol et al., 2016). Finally, one study did not report any significant associations between PM exposure during childhood and neurocognitive performance despite testing in several cognitive domains (Cserbik et al., 2020).

3.5. Quality Assessment Results (Supplemental Table 2)

Each source was assessed for risk of bias and quality of the study using frameworks from previous systematic reviews as well as guidelines from Cochrane (Higgins et al., 2011). None of our studies yielded a result of “high” risk of bias, while 4 were rated as “moderate” and the other 7 were rated as “low” risk.

4. Discussion

To the best of our knowledge, this is the first systematic review reporting on associations among air pollution, brain imaging findings measured via MRI, and cognitive outcomes. Eleven studies were included in the systematic review, including longitudinal (n = 5), case control (n = 1), and cohort (n = 5, with two prospective cohort studies). This systematic review examined studies that included adults (n = 7), including young and middle-aged adults (n = 3) and older adults (n = 4), as well as children and adolescents (n = 4). Few studies conducted a true mediation analysis to test how associations between air pollution and cognitive impairments were statistically mediated by brain structural and functional changes (n = 2). All included papers reported significant associations between air pollution and MRI findings for volumes and thickness of brain structures (n = 7) or functional connectivity (n = 4). Significant associations between air pollution and cognitive outcomes were found for a majority of the included studies, while two studies did not report significant cognitive associations, but reported significant associations with structural findings. Overall, we found probable negative associations between air pollution exposure, brain structure, and cognitive outcomes in adults as well as children.

4.1. Location and types of air pollution

Overall, significant results were found for both indoor and outdoor air pollution. Only one study looked at indoor air pollution in the form of incense burning (Wong et al., 2020). Significant differences for incense exposure were found specifically through fMRI and functional connectivity analysis. While differences in indoor and outdoor pollution results were not found, within different sources of outdoor pollution, traffic sources seemed to have more significant and impactful result. In urban environments, traffic pollution is one of the most significant sources of pollutants (Block et al., 2012). In both studies of traffic related air pollution, there were significant associations between pollution and impairments in cognitive function (Pujol et al., 2016; Semmens et al., 2013). Future research comparing effects of indoor and outdoor air pollution are needed.

Throughout all studies, PM2.5 was the most common air pollutant measured and demonstrated the most significant results with worsening brain structure and connectivity, as well as cognition. Other pollutants measured such as SO₂ and NO₂ were largely found not to be as significant—however, this could be due to reasons of being examined less. PM, especially finer matter such as PM_2.5 or PM_10, have been shown to be a very harmful air pollutant due to their small size and permeability that allows this matter to enter the bloodstream of the body, lungs and airways (WHO, 2021). A recent systematic review and meta-analysis showed the association between PM_2.5 exposure, as well as NO₂, and increased asthma exacerbations in both children and adults (Orellano et al., 2017). Other studies looking specifically at childhood asthma rates found that PM_2.5, PM₁₀, and NO₂ air pollutants (all fall under TRAP- traffic related air pollution), account for a significant portion of asthma incident rates (Alotaibi et al., 2019). Future studies should further integrate a wide spectrum of air pollutants to better understand the role specific air pollutants play in health. Additional studies focusing in on PM_2.5 and smaller PM pollutants would also be important to inform legislation and action on reducing rates of this harmful pollutant.

4.2. Timing of Air Pollution Exposure and Associated Outcomes

Air pollution in childhood has been established by previous literature to impact cognitive outcomes (Liu et al., 2014, Midouhas et al., 2018), which was echoed by child studies included in this review. Two of the 4 papers focusing on children reported motor speed and inhibitory control as two main cognitive domains negatively associated with air pollution exposure (Pujol et al., 2016, Guxens et al., 2018). Calderón-Garcidueñas et al., 2011 showed that white matter hyperintensities impacted young children’s complex reasoning and executive functioning domains which included verbal and working memory, and attention. Two longitudinal studies reported associations between air pollution during the fetal period and generalized thinning of the cortexes of both hemispheres of the brain at 6–10 years (Guxens et al. 2018) while adolescent volume and surface area of the cerebellum and subcortical were found to decrease with child air pollution exposure (Cserbik et al., 2020). These results suggest the effects of early life air pollution on cognition are long-lasting. However, only one of these studies reported a further association with a cognitive impairment, specifically for induced inhibitory control in children (Guxens et al., 2018). A potential explanation for these findings is the study conducted by Cserbik et al. 2020 examined a composite, overall cognitive score while Guxens et al. 2018 examined several individual domains of cognition. Higher order cognitive factors like memory, inhibitory control, etc. may be more impacted by air pollution in childhood and could be considered in future studies. Although cognitive impairment was most present in the memory domain, some significant negative associations were also seen for executive functions, global cognition, and visuospatial functions. This was mediated by weaker functioning and disrupted networks in the DMN region, which is important for goal-oriented tasks like response (Elton & Gao et al., 2015).

The significant association between air pollution in childhood and cognitive outcomes may be explained by that fact that children are generally more susceptible to the effects of air pollution because of their growing brains and that many of their immunities are not yet fully developed (UNICEF, 2017). Neurobehavioral outcomes such as attention were also especially impacted (Kicinski et al., 2015). Studies with other environmental toxins, such as lead exposure, have shown the detrimental effects of early childhood exposure especially on brain organization and cognitive functions, like language (Yuan et al., 2006). Even exposures during early childhood at levels below 10 ug/dl have a strong association with negative educational and behavioral outcomes (Chandramouli et al., 2009). In our included studies, Guxens et al., 2018 was unique because it measured exposure during the fetal period, which may have led to more significant associations considering the critical developmental period for exposure. On the other hand, Cserbik et al., 2020 only included a one-time geocoded measurement of air pollution based on the area in which the children lived at the time of the study. Air pollution throughout the first few years of life may be more impactful on cognitive outcomes later in life. Further longitudinal studies are important to help elucidate these associations.

Only one study examined young adults between the ages of 18–49 years. However, in this study there was an association between air pollution exposure and decreased cortical thickness in mostly the left-brain areas and structures, specifically in the frontal and prefrontal areas as well as temporal regions. Furthermore, MoCA cognitive tests performed showed a strong association between air pollution and worsening language and working memory cognition. Some of these results are similar to the ones found in the middle-aged adults age-group, possibly signifying that these are crucial years in which some of these effects take place and can be seen.

Two studies examined middle-aged adults. For this age group, the visuospatial cognitive domain seemed to be the most impaired in associations with air pollution in the majority of the included studies. In one study, this was mediated by disruptions in functional connectivity networks such as the frontal-parietal network, which includes the parietal lobe, a key area in visuospatial tasks (Seydell-Greenwald et al., 2017). There were also significant associations between air pollution and temporal lobe and cingulate cortex volumes as well as attention and executive function domains, though the memory domain had mixed results. Crous-Bou et al., 2020 partially attributed this to the fact that their sample included younger, healthy adults who were not cognitively impaired, but changes within the temporal lobe and other areas of the brain could indicate future declines in cognitive function.

Overall, the studies on older adults and the elderly found a probable strong, positive association between air pollution exposures, brain structure and connectivity, and declining cognition, which correlates with previous studies on this topic (Kulick et al., 2020; Chandra et al., 2022). Long term and cumulative air pollution has been suggested to have a significant effect on brain volume and structure of elderly individuals (Wilker et al., 2016; Chen et al., 2015). Importantly for older adults, there was a heavy emphasis on studying the risk of developing dementia in the future, and most (75%) of studies showed a cognitive impairment and probable diagnosis for dementia in the future, as mediated by changes specifically in the white & gray matter of the temporal lobe and corpus callosum. One of the precursors to dementia is memory impairment, which was a key way in which studies estimated how probable it was that these individuals would develop dementia in the future. Several of these studies also showed declines in memory, especially memory that involved immediate recall. These findings were supported by significant associations between air pollution exposure and gray/white matter volume loss, particularly in the temporal and midbrain part of the brainstem (Younan et al., 2020). In addition, Semmens et al. 2013 used white matter lesions to demonstrate how this was associated with an increased risk of developing vascular dementia, which may be important in developing new ways of finding brain neuroanatomic markers of this disease.

Variations in the significance of results in the adult/elderly population could be due to a variety of reasons. For example, especially for adult or older adult studies, exact amount or type of air pollution they have been exposed to over the life course is unknown. Studies only included current exposure, therefore, increased exposure to air pollutants earlier on in life may have influenced brain structure, connectivity, and cognitive results. In the future, studies examining exposures across the life course, especially during childhood, which measure brain structure and cognitive function throughout the lifetime would help to clarify associations.

4.3. Longitudinal Findings

Four of the included studies examined longitudinal associations (Guxens et al., 2018; Cserbik et al, 2020; Calderón-Garcidueñas et al., 2022; Chen et al., 2017). Out of these 4, 2 showed significant associations between air pollution, brain imaging results, and cognitive outcomes. All 4 showed that increased air pollution exposure in early life was associated with decreases in brain volume, thinning, or disruptions of functional connectivity later in life when the study results were collected. Three of the studies included were prospective cohort, which also followed the participants over a long period of time (Wong et al., 2020; Younan et al., 2020; Crous-Bou et al., 2020). The type of study did not seem to be associated with a stronger result or more significant associations with cognitive outcomes. However, this could be due to a variety of reasons especially for longitudinal studies as it could take a more time for exposure during young childhood to impact an individuals’ late in life cognition. Future longitudinal study that includes repeat measures of air pollution, brain imaging, and cognition over time would help to determine periods of greatest susceptibility.

4.4. Potential mechanisms

The mechanisms in which air pollution impacts cognitive outcomes and neuroimaging correlates is not fully understood. Several hypotheses have been proposed by included studies have shown that specific areas of the brain, specifically the frontal lobe, prefrontal cortex, and temporal lobes may have a great effect on cognitive domains. Previous studies have shown that neuroinflammation may have a role in this process Pollution-associated neuroinflammation is associated with cell loss in the central nervous system (Brockmeye et al., 2016). Specifically, air pollution degrades myelination in neurons, which is essential for the propagation of action potentials and, thus, the global functioning of the nervous system (Brockmeyer et al., 2016). Studies have also shown that air pollution, specifically exposure to PM2.5, is positively correlated to cell death specifically in neuronal cells, through the lysosome membrane permeabilization (LMP) pathway (Wei et al., 2022). Other studies have looked at how pollutants cross the blood-brain barrier (BBB). Studies have shown that the BBB is more easily permeated in younger children whose brains are still developing and growing (Calderón-Garcidueñas et al., 2014). Furthermore, neuroinflammation and the BBB are connected, as the breakdown of the BBB can be attributed to excess neuroinflammation, which can lead to hypoperfusion and deforming of white matter (Calderón-Garcidueñas et al., 2011). Similarly, oxidative stress could be another explanatory component (Crous-Bou et al., 2020). The combination of neuroinflammation and oxidative stress, leading to unregulated mediators in the brain, could lead to vascular dysfunction and problems with brain function (Wong et al., 2020).

Cserbik et al., 2020 suggest that the rapidly changing brains of children during development be a potential mechanism, where microglia cells could be responsible for disrupting the pruning, or elimination of extra synapses. Although pruning is a very typical process for children around the age of 10, disruptions could alter the development of the brain.

Finally, neuropathology explanations such as high amounts of amyloid-β plaque and phosphorylated tau have been shown to precede future cognitive declines. This may be a mechanism by which dementia occurs in the future even in cognitively healthy individuals (Younan et al., 2020; Calderón-Garcidueñas et al., 2022). Future studies should continue to investigate the mechanisms underlying how air pollution is able to disrupt brain structures and networks, and how that may lead to changes in cognition.

4.5. Strengths and Limitations

Our systematic review included several important strengths. Firstly, our question regarding associations between air pollution and brain structure and connectivity, and further associations with cognitive functioning is a part of a growing field and a more novel question in this area. Previous associations have been seen between increasing air pollution and impacts on cognition, both in children, adults, and the elderly, but fewer studies have examined brain structure and functional connectivity changes and their associations with cognitive functioning. Another strength of our systematic review was the extensive search strategy, to comprehensively review the literature.

Despite the many strengths, there were a few important limitations of our study. Although we included all studies that matched our inclusion criteria, there were still a small sample of studies included may result in biases. However, many of the studies included had several hundred or even thousands of participants, leading to lower bias and stronger reliability in results. Two of our studies included mediation analysis, which strengthened our assessment of the mediation of associations of air pollution and cognition by brain structure. While the remaining studies examined associations between air pollution, brain imaging, and cognition, as they did not conduct a true mediation analysis, the results on a mechanistic pathway are limited. We also could not perform a meta-analysis to add onto our systematic review due to the heterogeneity of air pollutants, areas of the brain imaged, and cognitive outcomes assessed by the eligible studies. For example, a few of the papers measured cortical thickness and/or volume changes of specific structures of the brain such as the hippocampus and prefrontal cortex, whereas others monitored larger regions of the brain as a whole or functional connectivity between regions. A quantitative meta-analysis would be inappropriate to conduct without a homogenous set of data to be compared with one another, as this is generally accepted to be required to conduct this sort of analysis (Kulinskaya et al., 2011). However, we have consulted with guidelines for systematic reviews that cannot conduct a quantitative analysis and followed these guidelines to the best of our abilities to synthesize using qualitative analysis only (Campbell et al., 2020).

5. Conclusion

This systematic review synthesized the current literature examining the intersection between air pollution exposure, MRI imaging results, and cognitive function. Our review consisted of studies from populations of adults and children and included a wide array of different cognitive outcomes. Overall, we report a possible significant association between air pollution and negative cognitive outcomes, specifically for executive function and memory cognitive domains. The results varied between our stratified age groups of children, young and middle-aged adults, and older adults. While results for the children and young/middle aged adults research was focused on executive functions, response times, and visuospatial functioning, studies on older adults looked at cognitive decline as a precursor for dementia. Significant neuroimaging results were present in every study included in this systematic review and suggested that air pollution was negatively associated with measures of brain structure and connectivity. This research is of public health importance in the clinical, policy, and research contexts and further underscores the need for an interdisciplinary strategy to address the detrimental effects air pollution has on brain health.

Highlights.

• Air pollution is known to be a detrimental environmental toxin

• Air pollution was associated with brain structural changes in all included studies

• Some studies showed that changes in brain structure mediated cognitive deficits

• The most affected cognitive domains varied between age groups

• Air pollution was associated with disruptions in functional connectivity networks