Air Pollution and Atopic Dermatitis, from Molecular Mechanisms to Population-Level Evidence: A Review

Atopic dermatitis (AD) has increased in prevalence to become the most common inflammatory skin condition globally, and geographic variation and migration studies suggest an important role for environmental triggers. Air pollution, especially due to industrialization and wildfires, may contribute to the development and exacerbation of AD. We provide a comprehensive, multidisciplinary review of existing molecular and epidemiologic studies on the associations of air pollutants and AD symptoms, prevalence, incidence, severity, and clinic visits. Cell and animal studies demonstrated that air pollutants contribute to AD symptoms and disease by activating the aryl hydrocarbon receptor pathway, promoting oxidative stress, initiating a proinflammatory response, and disrupting the skin barrier function. Epidemiologic studies overall report that air pollution is associated with AD among both children and adults, though the results are not consistent among cross-sectional studies. Studies on healthcare use for AD found positive correlations between medical visits for AD and air pollutants. As the air quality worsens in many areas globally, it is important to recognize how this can increase the risk for AD, to be aware of the increased demand for AD-related medical care, and to understand how to counsel patients regarding their skin health. Further research is needed to develop treatments that prevent or mitigate air pollution-related AD symptoms.


Introduction
Air pollution is a complex mixture that includes certain solid particles, liquid droplets, and gaseous molecules; the U.S. Environmental Protection Agency has identified criteria air pollutants that are common in the U.S. and have well-known adverse human health effects, including ozone (O 3 ), particulate matter (PM), carbon monoxide (CO), sulfur dioxide (SO 2 ), and nitrogen dioxide (NO 2 ) [1]. Solid particles, often including ultra-fine particles smaller than 0.1 microns, are found in dust, smoke, and soot [1]. Air pollutants negatively affect the functions of multiple organs, notably the heart and lungs [2]. Concentrations of certain air pollutants are increasing in areas around the world due to environmental changes, such as increased wildfires, and sociocultural trends, such as urbanization and industrialization [2]; much of the global increase in air pollution is due to more motor vehicles and coal-fired power generation, and it is a pervasive public health issue due to its many negative effects on human health. Additionally, air pollution from a California wildfire was recently found to be associated with increased clinic visit rates for itch and atopic dermatitis (AD), a common inflammatory skin disease that affects up to 1 in 5 children and 1 in 10 adults worldwide [3][4][5].
Human skin is in constant contact with the environment, and air pollutants can directly harm skin barrier function and homeostasis to contribute to the development and exacerbation of cutaneous diseases [6][7][8]. AD, often referred to as eczema, is an inflammatory skin disease in which patients have underlying skin barrier defects and a heightened immune response to irritants and allergens [9,10]. Climatic variables-such as temperature, humidity, pollen load, and sun exposure-have been shown to affect AD symptoms [11][12][13].
The effects of air pollution on AD have been less studied compared to their impact on respiratory and cardiovascular diseases, but some of the underlying pathologic mechanisms, such as triggering inflammatory responses, are similar [14][15][16]. There have been many reviews on the topic of air pollution and skin health, but most were not specific to AD, did not include recent studies, or focused on either only biological or only epidemiologic evidence [6][7][8]15,[17][18][19][20][21][22]. Here, we provide an updated, comprehensive, and synthesized overview of the molecular, cellular, and epidemiologic literature focusing on the relationship between air pollution and AD. The aims of this manuscript are to discuss the connection between air pollution and AD through (i) reviewing potential biological mechanisms, (ii) evaluating population-level evidence, including impacts on healthcare utilization that were not included in previous reviews, and (iii) discussing avenues for future research.

Review of the Evidence
Many studies have been conducted around the world using a variety of exposure and outcome assessment methodologies to investigate the effects of air pollution on AD. In aggregate, they suggest a higher risk for AD symptoms and disease incidence associated with air pollution exposure. Characteristics of air pollution exposure, including the type of air pollutant, source of air pollution, and concentration of exposure (all of which can affect its impact on the skin), have been found to vary across studies.

Molecular Pathogenesis
Multiple biological mechanisms by which air pollution may induce or exacerbate AD have been identified, including activation of the aryl hydrocarbon receptor pathway, promotion of oxidative stress, impairment of the skin barrier, and initiation of a proinflammatory response.

Aryl Hydrocarbon Receptor Pathway
Polycyclic aromatic hydrocarbons (PAHs) are produced from the combustion of any carbon-based fuel and are potent inflammation-inducing ligands of the aryl hydrocarbon receptor (AhR) [23]. AhR is a cytosolic ligand-activated transcription factor that can activate the expression of genes related to cell proliferation, detoxification, inflammation, and melanogenesis [23]. Transgenic mice that constitutively expressed AhR in keratinocytes had phenotypic characteristics similar to that of AD: pruritus, skin inflammation, and skin barrier dysfunction [9,24,25]. PAHs activated the AhR pathway to enhance artemin expression and inflammatory processes in murine epidermal cells [24]. Artemin is a neurotrophic factor that causes a hypersensitivity to itching, is coded by Artn-an AhR target gene-and expressed at high levels in patients with AD [26]. Human cultured epidermal cells exposed to diesel exhaust particles upregulated ARTN mRNA through activation of the AhR pathway [24]. In addition, cultured human keratinocytes exposed to O 3 and PAHs have an increased expression of cytochrome P450 isoforms-CYP1A1, CYP1A2, and CYP1B1-through an AhR-dependent mechanism [27], indicating that the AhR pathway may be implicated in inducing a response to adverse effects of air pollution on skin cells. On the other hand, PAH-containing coal tar has therapeutic benefits in treating AD and psoriasis, suggesting that AhR activation can be beneficial for treating some existing cutaneous symptoms [28]. This is due to the nature of AhR signaling in skin: patients with AD have dominant non-canonical AhR signaling that is pathologic, so AhR agonists are helpful for restoring the balance between canonical and non-canonical pathways in these patients [29]. The activation of the AhR pathway may also contribute to greater aldo-keto reductase expression, which affects mast cell activation and stimulates a Th2 response [30].

Oxidative Stress
Components of air pollution, such as nitrogen oxides, O 3 , and PAHs, can trigger the generation of reactive oxygen species (ROS) [31]. One mechanism involves intracellular metabolic pathways that convert PAHs into quinones, generating superoxide anion, hydrogen peroxide, and ROS [32]. Carbonyl moieties, a marker of direct oxidative damage of proteins, were analyzed in skin biopsies from 75 AD patients, and they were found to be significantly increased in AD lesions, especially in the superficial stratum corneum, compared to the skin of control patients [31]. The authors suggested that the skin of patients with AD is more susceptible to externally induced ROS damage, which can further disrupt the skin barrier function and exacerbate AD.
Another study assessed the impact of ambient O 3 on an indicator of oxidative damage, malondialdehyde (MDA), on mouse skins [33]. Higher MDA concentrations extracted from O 3 -exposed skin, compared with covered skin, indicated that O 3 can initiate lipid peroxidation reactions in cutaneous tissue and that O 3 exerts its antioxidant-depleting effects through direct action on skin cells as opposed to the action of secondary products of O 3 interactions formed in the lungs and then brought to the skin via blood [33]. Short-term O 3 exposure also depleted levels of the antioxidants vitamin C, uric acid, and glutathione in the murine stratum corneum, indicating oxidative damage [34].

Skin Barrier Function
Exposure to air pollution components can affect epidermal barrier function [35,36], as measured by transepidermal water loss (TEWL). Study subjects with AD exposed acutely to ambient NO 2 were found to have increased TEWL, an indication of barrier impairment, compared to both preexposure AD skin and to post-exposure healthy skin [37]. The mechanisms underlying skin barrier disruption may be the pollution-related generation of free radicals and lipid peroxidation of polyunsaturated fatty acids in cell membranes, as is found in other organ systems [37,38]. The skin barrier function is impaired by particulate matter exposure due to the decreased expression of E-cadherin and structural proteins in the stratum corneum, such as cytokeratin and filaggrin [39], as well as increased production of matrix metalloproteinases [40,41].

Inflammation
Studies conducted with keratinocytes from mice and humans illustrate how pollutants can facilitate the process of cutaneous inflammation. After exposure to diesel exhaust particles at high concentrations, mouse epidermal cells increased the expression of NF-κB [42], which can facilitate cytokine expression. Ushio et al. incubated cultured human keratinocytes with different concentrations of diesel exhaust particles and found a greater production of IL-1β at high exposure levels [43]. IL-1 is implicated in the pathogenesis of AD in patients with an associated FLG gene mutation [44]. In another study, researchers exposed cultured human epidermal cells to Asian dust storm particles that contained a mixture of particulate matter for 24 h [45]. The exposed cells significantly upregulated mRNA expression for several proinflammatory cytokines, including IL-6 [45], another cytokine implicated in the AD disease course [46][47][48]. These studies suggest that high levels of air pollution can induce a proinflammatory cytokine state in the skin due to cytokine release from keratinocytes [33,38,44,49].
In a study of children, researchers found a strong positive association between exposure to indoor volatile organic compounds (VOCs) and the percentage of IL-4-producing T cells in the blood of toddlers [50]. In another study, maternal exposure to VOCs was associated with a decreased percentage of interferon gamma (IFN-γ)-producing T cells and T-cell polarization toward the type 2 phenotype in neonates [51]. A reduced ability to secrete type 1 cytokines, such as IFN-γ, delays the maturation of type 1 sensitivity reactions and increases the risk for developing atopic diseases [50]. Taken together, the upregulation of proinflammatory cytokine expression and type 2 T-cell polarization may contribute to the development of symptomatic AD after air pollution exposure. In addition, the exposure of skin cells to airborne nanoparticles, such as carbonaceous pollutants, diesel exhaust particles, and tungsten carbide cobalt particles, can induce the production of proinflammatory cytokines and alter several cellular signaling pathways [52,53].

Population-Level Effects
Many epidemiologic studies have examined exposure to air pollutants and the development or exacerbation of AD. While the majority of studies have found a positive relationship ( Table 1) , some have reported null or inconclusive findings ( Table 2) [81][82][83][84][85][86][87][88][89][90]. The variance of the results across studies may be attributable to varying criteria for the severity and diagnosis of AD in different countries, age of the participants, and the approach to outcome assessment. In addition, people around the world are exposed to different components of air pollution at different magnitudes [91]. The heterogeneity in the type and concentration of air pollutant exposure, as well as varying cooccurring climatic factors, such as temperature and humidity, that may not be adjusted for in analyses can lead to inconsistent results across studies. A comparison of the study findings may also be limited by differences in the exposure assessment and characterization techniques; for example, ground-level monitor measurements may differ from satellite-based estimations [92]. The results across studies using self-reported data are mixed, but those with more direct measurements of the outcomes appear more likely to have positive associations. Most studies have been conducted with children as study participants, but almost all studies with adults, which are fewer in number, have found positive, though often slightly smaller, associations. Below, we focus on studies with eczema/AD as a primary outcome in children and adults, presenting data from cross-sectional studies and then longitudinal studies, followed by data on healthcare utilization.

Air Pollution and Pediatric AD Prevalence, Incidence, and Severity
One cross-sectional study in Taiwan used a modified protocol from the International Study of Asthma and Allergies in Children (ISAAC) to examine atopic eczema symptoms in 23,980 school children using questionnaires and found that the risk of disease occurrence was strongly associated with perceived exposure to ambient air pollution, subjectively defined as not present, mild, or moderate to severe [61]. The same group later conducted a larger cross-sectional study with children and found that exposure to traffic-related air pollutants (TRAP), including CO and nitrogen oxides, was positively associated with the prevalence of flexural eczema in both boys and girls [62]. Larger associations between living near a traffic-heavy area and AD prevalence were detected in two other cross-sectional studies conducted in Bolivia [69] (odds ratio (OR) 1.4, 95% confidence interval (CI): 1.1-2.0) and Lebanon [68] (OR = 1.5, 95% CI: 1.1-2.0), though air pollution exposure was not quantified. These results are supported by studies in Korea that found significant but weaker associations for TRAP exposure among children [73,75]. Other cross-sectional analyses have found non-significant associations between various air pollutants and AD in children, many of which have also used ISAAC data and similar exposure assessment methods; however, some included AD as a secondary outcome and had less statistical power due to smaller sample sizes. Overall, these cross-sectional studies provide mixed evidence regarding the association between air pollution and AD in children; major limitations of these studies include their study design, which prevents causal inference, and self-reported outcomes, which could result in misclassification bias.
Longitudinal studies, which have a stronger methodology for assessing the temporality of exposure and outcomes, conducted with children with AD living in urban settings in South Korea provide evidence for the connection between air pollution and AD symptom severity. Song et al. found a significant association between pruritus severity and ultrafine particulate pollution exposure over 2 months in 41 children with AD [63]. Another study with 21 participants found that a 10 µg/m 3 increase in the daily mean particulate matter less than 2.5 microns in diameter (PM 2.5 ) concentration was associated with 40% increased odds (95% CI: 21-61%) of exacerbation of AD symptoms [71]. These investigations were supported by a longer 18-month study with 22 children with AD [64]. Of note, these studies involved small sample sizes, and misclassification bias may have occurred due to the subjective reporting of AD symptoms.
Infants, who have an immature skin barrier, living in urban environments may be particularly vulnerable to developing AD. Proximity to main roads and increased air pollution exposure during early life was associated with higher prevalence of AD in a German birth cohort study: PM 2.5 exposure, as determined by land-use regression, was associated with an adjusted relative risk of 1.69 (95% CI: 1.04-2.75) for doctor-diagnosed AD [55]. These results are in alignment with findings from another birth cohort study [56] showing that early life NO 2 exposure was associated with a higher occurrence of childhood AD: odds ratio = 1.18 (95% CI: 1.00-1.39). Of note, exposures to NO 2 (odds ratio = 1.35, 95% CI: 1.03-1.78), CO (odds ratio = 1.51, 95% CI: 1.16-1.97), and particulate matter less than 10 microns in diameter (PM 10 ) (odds ratio = 1.22, 95% CI: 1.02-1.45) before birth, especially in the first trimester when the fetus is rapidly developing, have been shown to increase the risk for the development of AD before 6 months of age [57,70]. In another study, NO 2 exposure throughout pregnancy was associated with the onset of childhood AD, and postnatal PM 10 (odds ratio = 0.82, 95% CI: 0.72-0.94) and NO 2 (odds ratio = 0.80, 95% CI: 0.65-0.98) exposures were associated with decreased AD remission [74]. Additionally, exposure to combinations of different types of air pollutants at varying concentrations in combination with favorable or non-favorable climatic factors such as temperature and humidity can impact the AD prevalence in children [77].
Regarding indoor air pollution exposure, two studies illustrated positive associations with the AD symptoms. A prospective study conducted with schoolchildren in Seoul, Korea, found that indoor exposure to pollutants, such as toluene, PM 10 , and nitrogen oxide, can increase risk for pruritus in patients with AD [65]. Additionally, a reduction in the indoor PM 10 concentrations following the implementation of a 7-month school program to improve the indoor air quality was strongly associated with a reduced prevalence of AD and mean eczema area and severity index (EASI) scores [66]. Another study showed that indoor air pollution exposure exacerbated the AD symptoms in children during the spring, winter, and cooler temperatures, with a greater risk in patients with inhalant allergen sensitization and pre-existing severe AD [96].

Air Pollution and Adult AD Prevalence and Incidence
Compared to studies conducted with children, there are fewer studies with adult participants, though they have primarily found positive associations. A survey study in Sweden and case-control study in Taiwan reported that exposure to air pollutants, such as PM 2.5 , was associated with a modestly increased AD prevalence (OR = 1.05, 95% CI: 1.02-1.08) [58,59]. In addition, a study in Australia found positive associations between PM 2.5 and NO 2 exposures, as calculated by land-use regression models and self-reported AD symptoms in adult men [76]. A cross-sectional study in the United Kingdom with a similar sample size and exposure assessment technique did not find a significant result for NO 2 , possibly because the outcome assessment was less accurate and did not include the skin prick test results for atopy similar to the Australian study [86].
In a longitudinal study, the incidence of AD symptoms in German women after age 55 was significantly associated with exposure to TRAP, nitrogen oxides, PM 2.5 , and PM 10 , as estimated by monitoring data and land-use regressions [60]. The risk associated with exposure to nitrogen oxides was almost three times higher for minor allele carriers of the aryl hydrocarbon receptor polymorphism rs2066853, which affects the transcriptional activation domain in the AHR gene compared to non-carriers, implicating the AhR pathway in adult onset AD associated with TRAP. A follow-up study showed that these associations were stronger in eczema patients without markers for atopy compared to those with atopic markers, suggesting that air pollution may play a bigger role in the development of a nonatopic form of eczema in older individuals, though further research is needed [67].

Air Pollution and Healthcare Utilization for AD
Increasing levels of air pollution can have broader impacts on the healthcare system by driving up service utilization and costs [100]. All 12 studies we found performed an exposure assessment with air pollution concentrations from air monitoring stations and measured clinic, hospital, and emergency department visits for AD based on International Classification of Diseases (ICD-10) codes. Eight epidemiologic studies conducted in China found positive associations between the poor air quality and medical visits for AD among children and adults [78,79,[93][94][95]98,101,102], findings that were supported by studies conducted in Korea and Turkey [97,99]. One study used a case-crossover analysis and found that combined adult and pediatric visits for eczema increased by 3.81%, 3.18%, 5.43%, and 5.57% per interquartile range increase in PM 2.5 , PM 10 , NO 2 , and SO 2 concentrations, respectively [93]. In another study, most air pollutants showed roughly linear exposureresponse relationships with total daily AD visits, suggesting the potential for no major threshold effect for air pollution on AD exacerbations [95]. Regarding a period of intense, short-term exposure to air pollution, a recent study found that wildfire air pollution was associated with significantly increased clinic visits for an AD: rate ratio of 1.49 (95% CI: 1.07-2.07) for pediatric patients and 1.15 (95% CI: 1.02-1.30) for adult patients [3]. A followup study found that the rates of clinic visits for AD for adults was higher among those age ≥ 65 years of age compared to that of younger adults, which suggests that the skin of older adults has a greater vulnerability to air pollution [80]. In combination, these studies suggest system-level impacts of poor air quality on both pediatric and adult visits for AD at healthcare centers.

Discussion
The results of in vitro, animal, and epidemiologic studies suggest that exposure to air pollutants increases the risk of the development and exacerbation of AD. Pollution exposure in utero, during childhood, and during adulthood has been shown to contribute to the incidence, prevalence, and severity of AD. The mechanistic pathways underlying the impact of air pollution on skin health include the activation of the AhR pathway, induction of oxidative stress, impairment of the skin barrier, and stimulation of an inflammatory response. Future research could focus on how individuals and mixtures of air pollutants can contribute to AD through the modulation of skin homeostasis and its microbiome alongside other environmental factors, such as ultraviolet radiation. Population-based studies demonstrate mixed results regarding the association between air pollution and AD, but the majority of longitudinal studies, which are more robust than cross-sectional studies, found that exposure to pollutants was positively associated with AD in pediatric and adult populations. Variations in the results from epidemiologic studies can be attributed to several potential factors, including types of air pollutants studied, sources of air pollution, magnitude of exposure, influence of concurrent climatic factors, precision of outcome assessment, and confounders included in statistical analyses. Future epidemiologic studies could be conducted with large, prospective cohorts with both meticulous exposure assessments of multiple air pollutants and precise outcome assessments with physician-reported diagnoses and symptom severity.
The climate crisis has contributed to a recent increase in the occurrence and severity of wildfires in the U.S. and around the world, leading to poor air quality. In California, wildfires burned approximately 1.6 million acres of land in 2018 and 4.3 million acres of land in 2020 [103], and millions of Australians and Brazilians experienced exposure to hazardous levels of air pollutants arising from the wildfires in 2019 and 2020 [104,105]. Since wildfire smoke can travel far distances, it will likely broadly affect patients' skin health and quality of life [3,106]. This is an important issue for clinicians who counsel patients with AD symptoms during periods of poor air quality as well as public health practitioners who produce public communications and policies that aim to reduce the risks for skin symptom exacerbations. To better understand the effects of air pollution on skin conditions, novel methodologies such as machine learning tree-based models and artificial neural networks are being used to predict AD development with air pollution data [107,108]. Of note, disparities in AD outcomes may be exacerbated by the roles of social determinants of health, such as race, income, education, and geography, in determining exposure to air pollution [109]. Social and structural factors that affect the generation of air pollution, geographic area of residence, and access to healthcare can contribute to environmental health inequities for AD [110][111][112].
As the understanding of air pollution and AD increases, new public health interventions and targeted treatments can be developed. Environmental policies that ban fossil fuel production and use, as well as promote sustainable land management practices, could decrease the amount of outdoor air pollution [113] and could reduce the risk for skin disease exacerbations. In addition, the installation of air filtration devices to improve the air quality in indoor settings could help to reduce the prevalence of AD and severity of the symptoms [66]. Regarding clinical management, specific emollients have been reported to improve the skin barrier function in AD [114,115], which could possibly also reduce pollution-induced symptoms. Additionally, patients could be advised to wear long sleeves and long pants to limit pollution exposure, similar to protection from ultraviolet radiation [116]. However, research on the effectiveness of preventative interventions is needed to develop evidence-based clinical recommendations for topical products and skin protection methods. Of note, topical AhR-modulating medications are being developed and tested for the management of AD and psoriasis, and future research can ascertain whether they can be used to alleviate pollution-related AD symptoms [117]. In addition, recent research has shown how nanomaterials can encapsulate antioxidants and other medications to improve their bioavailability and therapeutic effects [118,119]. Since antioxidants can help reduce oxidative stress and lipid peroxidation induced by environmental exposures, it is possible that this method of drug delivery to the skin may help manage pollution-related skin disease flares [118]. In addition, there are limited clinical studies that have investigated the effectiveness and safety of topical corticosteroids incorporated into lipid nanoparticles for the treatment of eczema [120,121]. Further research is needed to develop treatments that prevent or mitigate air pollution-related skin symptoms, and more robust epidemiological evidence would inform public health education.

Conclusions
Air pollution is a recognized public health issue, and it has been shown to have several negative impacts on the human body, including skin disease, obstructive pulmonary diseases, cardiovascular diseases, strokes, psychological stress, and poor obstetric outcomes. Overall, the scientific literature evaluated in this review paper found evidence in the support of an association between air pollution exposure and atopic dermatitis. Molecular, cellular, and animal studies demonstrated that the biological underpinnings of the pollution-AD relationship include activation of the AhR pathway, generation of reactive oxidative species, weakening of the skin barrier, and promotion of a proinflammatory response. The epidemiologic evidence was mixed, but most longitudinal studies found that exposure to air pollutants was positively associated with AD in both the adult and pediatric populations. Further areas of research needed in this topic include characterizing air pollution-related inequities in skin disease, air pollutants' interactions with the skin microbiome, longitudinal exposure to mixtures of air pollutants and the impacts on the incidence and severity of AD, and the effectiveness of clinical interventions in managing air pollution-induced skin exacerbations.    2.5 particulate matter less than 2.5 microns in diameter PM 10 particulate matter less than 10 micron in diameter ROS reactive oxygen species SO 2 sulfur dioxide TEWL transepidermal water loss TRAP traffic-related air pollutants