The Impact of Early-Life, Long-Term Exposure to Ambient Particulate Matter on Atopic Dermatitis Remission and Sensitization in Infants Authors: Sukyung Kim, Inbo Oh, Jaeyoung Choi, Jihyun Kim, Jiwon Kim, Jin-Hee Bang, Young-Min Kim, Kangmo Ahn Published: 2025 ABSTRACT AbstractBackgroundStudies that examine a link between long-term exposure to ambient particulate matter (PM) and atopic dermatitis (AD) in children are lacking. This longitudinal study aimed to investigate the effect of long-term exposure to PM on AD in infants.MethodsWe enrolled 150 infants aged 0–1 years with AD and followed them until age 3 in Seoul, Korea. We assessed long-term exposures to ambient particulate matter with an aerodiameter < 10 µm (PM10) from birth to age 3 using a fusion model based on each individual’s home address. Effects of long-term exposure to PM10 on AD persistence and sensitization to aeroallergens were evaluated using Cox proportional hazard regression models after adjusting for potential confounders.ResultsOf the 150 infants, 54 (36.0%) showed remission of AD by age 3. Overall, long-term exposure to PM10 was not significantly associated with AD persistence at age 3 (hazard ratio [HR], 1.04; 95% confidence interval [CI], 0.97−1.12; P = 0.267 per 1 µg/m3). However, when stratified by the baseline AD severity, increased exposure to PM10 significantly elevated the risk of persistence in children with moderate AD at enrollment (HR, 1.19; 95% CI, 1.05−1.36; P = 0.007). Long-term exposure to PM10 was also associated with an increased overall risk of sensitization to pollen (HR, 1.12; 95% CI, 1.00−1.24; P = 0.043), with a more pronounced effect in children with moderate AD (HR, 1.31; 95% CI, 1.08−1.58; P = 0.005). However, no significant association was observed between PM10 exposure and sensitization to house dust mite or pet allergens.ConclusionAn early-life long-term exposure to ambient PM10 in infants with AD may decrease the remission and increase the sensitization to pollen at age 3.Keywords: Atopic Dermatitis, Environmental Exposure, Particulate Matter, Remission, Pollen, Infant, Children INTRODUCTION Atopic dermatitis (AD) is a chronic relapsing inflammatory skin disorder mostly found in childhood.1 It has become a growing health issue as its prevalence is increased worldwide including Korea.2,3 Besides its grave influence on quality of life, AD also affects the entire family socially and economically.4 Moreover, as an initial manifestation of atopic march, more than 30% of AD patients has been reported to develop allergic rhinitis or asthma in their later lives, with the risk being higher in early-onset, severe, and persistent AD.5,6 AD is also associated with sensitization to inhalant allergens.7,8 Identifying risk factors for severe or persistent phenotype, sensitization, and allergic comorbidities might help establish a better management plan for AD patients. Among air pollutants, exposure to particulate matter (PM) is associated with adverse health outcomes, especially allergic diseases.9,10,11 Several studies have already reported that AD symptoms were aggravated by short-term exposure to PM. A longitudinal study involving children aged 1 to 5 years with AD has revealed a significant association between a 10 µg/m3 increase in ambient particulate matter with an aerodiameter < 2.5 µm (PM2.5) exposure and AD symptoms (adjusted odds ratio, 1.399; 95% confidence interval [CI], 1.216−1.610).12 In another panel study involving 125 AD children aged 0−6 years, the overall risk of AD symptoms significantly increased with higher levels of ambient PM2.5 and particulate matter with an aerodiameter < 10 µm (PM10) exposure.13 In that study, the risk of AD symptoms resulting from exposure to ambient PM was significantly higher during dry, moderate weather as determined by spatial synoptic classification. That study also found a lagged effect of PM2.5 up to 4 days on days with dry, moderate weather. To observe short-term effect of exposure to indoor PM2.5 on AD symptoms, a longitudinal study was conducted on 64 Korean children with moderate-to-severe AD.14 In that study, a 10 µg/m3 increase in indoor PM2.5 concentration led to a 16.5% increase in AD symptom scores in spring (95% CI, 6.5−27.5) and a 12.6% increase in winter (95% CI, 4.3−21.5). Compared to the short-term effect of PM on AD symptoms, studies on the impact of long-term exposure to PM on AD are lacking. A previous study has evaluated the relationship between air pollution and the incidence of AD using the National Health Insurance Service-National Sample Cohort database, including 1,030,324 person-years and 3,203 incident cases. It found that long-term exposure to air pollutants significantly increased the development of AD.15 However, there is very limited evidence of whether long-term exposure to ambient PM affects AD remission or allergic march, particularly in young children. Thus, this longitudinal study aimed to investigate whether long-term exposure to PM could affect remission of AD at age 3 after the development of AD during infancy. Effect of long-term exposure to PM on aeroallergen sensitization in infants with AD was also evaluated. METHODS Study population and clinical data We enrolled 150 infants and young children (88 boys, 62 girls) aged 0–1 years with AD visiting our hospital, Samsung Medical Center, who were living in the Seoul metropolitan area of Korea. Locations of their residences are shown in Fig. 1. AD was diagnosed according to the Hanifin and Rajka criteria.16 Each patient was followed until they were 36 months old. CMAQ modeling in this study. (A) Locations of the Seoul metropolitan Area (green area) and the 9-km (D2) and 3-km (D3) domains for CMAQ modeling. (B) Locations of AD patients’ residences (circles) and AQMS (blue triangles) in the study area. The color tiles indicate one month-averaged PM10 concentrations predicted by the CMAQ model in May 2018. CMAQ = Community Multiscale Air Quality, AQMS = air quality monitoring system, PM10 = particulate matter with an aerodiameter < 10 µm. At enrollment, demographic data were obtained, including date of birth, age of AD onset, and body mass index (BMI). AD severity was assessed using the SCORing Atopic Dermatitis (SCORAD) index, developed by the European Task Force Group on AD in 1993.17 Subjects were classified into 3 groups, ‘mild (< 25),’ ‘moderate (25−49),’ and ‘severe (≥ 50),’ according to SCORAD at enrollment. Levels of total IgE and specific IgE against egg white, cow’s milk, soybean, wheat and peanut in the peripheral blood were measured using ImmunoCAP (ThermoFisher Scientific Inc., Waltham, MA, USA) and considered positive when levels were 0.69 kUA/L or above. The cut-off point was arbitrarily determined. To evaluate the status of AD, children were brought to the hospital at least once a year until the age of 3 years. AD remission was defined as absence of AD symptoms in the past 6 months without needing medical treatments including topical corticosteroid. When they visited the hospital at age 3, skin prick tests were performed to detect specific IgE against common inhalant allergens including Dermatophagoides pteronyssinus, D. farinae, tree pollen mixture, grass pollen mixture, weed pollen mixture, and cat and dog dander (Lofarma, Milan, Italy). Histamine was used as a positive control, and normal saline was used as a negative control. A positive response to allergens in the skin prick test was determined when wheal size was ≥ 3 mm with adequate reactions to controls. Assessment of PM exposure and other environments To assess the long-term exposure to ambient PM10 from birth to the age 3, we tracked the patients’ residential addresses upon relocation and utilized a fusion model that integrates the air quality monitoring sites (AQMSs) and Community Multiscale Air Quality (CMAQ) modeling data. This fusion technique can improve air pollution exposure estimation in a large and complex urban area like the Seoul Metropolitan Area where spatial heterogeneity of air pollution is significant and monitoring data are limited in space. The CMAQ model (version 5.0.2)18,19 was run for one year in 2018 to capture local gradient of PM10 concentrations in the Seoul metropolitan area in a nested mode at 27, 9, and 3 km horizontal grid dimensions; The fine-scale innermost domain (D3 in Fig. 1) is the target area of the study that covers the Seoul metropolitan area. Detailed CMAQ-ready meteorological and emission inputs, initial and boundary conditions, and physical and chemical options used in our CMAQ modeling were described previously by Oh et al.20 Hourly gridded concentrations predicted by the CMAQ model were averaged on a monthly basis to blend ambient monitored data. Monitored concentrations of PM10 and other air pollutants were obtained from 141 AQMS (Fig. 1B) in the study area. Neighbor AQMS surrounding each patient’s residential locations (updated monthly for 3 years) within a 5 km radius were first identified. An inverse-distance weighted (IDW) average of observed values was obtained. Fused concentrations were then calculated by multiplying IDW average concentrations with the ratio of CMAQ modeling data for the destination grid-cell containing the patient’s residence to those containing the neighbor monitor. These fused concentrations were calculated on a monthly basis and averaged throughout the exposure assessment period of 3 years. Monthly exposure to PM was assessed by checking subjects’ addresses whenever they moved during the study period. Statistical analysis To investigate whether exposure to ambient PM affects the remission of AD in young children, we evaluated the associations of long-term exposure (from birth to age 3) to ambient PM10 with the persistence of AD and the incidences of allergen sensitizations at age 3 using Cox proportional hazard regression models. Ambient nitric dioxide (NO2), sulfur dioxide (SO2), ozone (O3), carbon monoxide (CO), relative humidity (RH), and temperature, sex, BMI, comorbid food allergy, and SCORAD at enrollment were included in the Cox models as confounding factors. All procedures were conducted using R version 4.0.2 (Comprehensive R Archive Network: http://cran.r-project.org). The “survival” package (version 3.1-12) was used for Cox proportional hazard regression model fitting. All tests were 2-sided, and an alpha level of less than 0.05 was considered significant. Ethics statement Written informed consent was obtained from the parent or guardian of each participating child. Study protocols were reviewed and approved by the Institutional Review Board (IRB) of Samsung Medical Center (IRB No. 2014-07-064). RESULTS After excluding AD patients who were not followed until 3 years of age or had missing data, the number of children enrolled in this study was 150. Table 1 shows characteristics of the 150 subjects at enrollment and the remission rate of AD at age 3. According to the SCORAD at enrollment, AD severity was ‘mild’ in 72 (48.0%), ‘moderate’ in 61 (41.0%), and ‘severe’ in 17 (11.0%) children. The remission rate of AD at age 3 was 36.0%. There was no significant difference in the remission rate between boys and girls (P = 0.529). Sensitization rates to house dust mite (HDM) and pollen at age 3 were 52.7% and 35.3%, respectively. Sensitization to food allergens were assessed at enrollment only in cases where food allergies were clinically suspected. Sensitization rates for egg white, cow’s milk, peanuts, soybeans, and wheat were observed in 64/99 (64.6%), 47/85 (55.3%), 34/94 (36.2%), 16/39 (41.0%), and 17/39 (43.6%) patients tested, respectively. Table 1Characteristics of the study populationCharacteristicsValuesNo. of subjects150 (100.0)Age at enrollment, yr1.3 ± 1.1BMI, kg/m2 15.7 ± 1.2SCORAD at enrollment22.8 ± 15.0Family history of allergic diseases88 (58.7)Severity at enrollmenta Mild72 (48.0)Moderate61 (41.0)Severe17 (11.0)Total IgE at enrollment, U/L332.5 ± 610.0Duration of follow-up, mon22.1 ± 13.7AD status at age 3Persistence96 (64.0)Remission54 (36.0)Sensitization at age 3Food allergensb 77 (51.3)Inhalant allergensHouse dust mitec 79 (52.7)Pollend 53 (35.3)Pete 35 (23.3)Data are expressed as mean ± standard deviation or number (%).BMI = body mass index, SCORAD = SCORing Atopic Dermatitis, AD = atopic dermatitis.aSeverity is categorized according to SCORAD at enrollment; ‘Mild’ indicates infants with SCORAD < 25, ‘Moderate’ with 25 ≤ SCORAD < 50, and ‘Severe’ with 50 ≤ SCORAD.bSensitized to at least one of 5 common food allergens, including egg white, cow’s milk, soybean, wheat, and peanut.cSensitized to house dust mite (Dermatophagoides pteronyssinus or D. farinae).dSensitized to pollen (at least one of pollen mixture allergens, including tree, grass, and weed).eSensitized to pet (cat or dog allergens). Characteristics of the study population Data are expressed as mean ± standard deviation or number (%). BMI = body mass index, SCORAD = SCORing Atopic Dermatitis, AD = atopic dermatitis. aSeverity is categorized according to SCORAD at enrollment; ‘Mild’ indicates infants with SCORAD < 25, ‘Moderate’ with 25 ≤ SCORAD < 50, and ‘Severe’ with 50 ≤ SCORAD. bSensitized to at least one of 5 common food allergens, including egg white, cow’s milk, soybean, wheat, and peanut. cSensitized to house dust mite (Dermatophagoides pteronyssinus or D. farinae). dSensitized to pollen (at least one of pollen mixture allergens, including tree, grass, and weed). eSensitized to pet (cat or dog allergens). The monthly average level of PM10 exposed to each subject was 47.8 ± 3.9 µg/m3 for the whole study period (Table 2). Average values of ambient NO2, SO2, O3, CO, temperature, and RH for the residential places of the young children were 32.1 ± 3.8 ppb, 5.4 ± 1.2 ppb, 22.8 ± 1.6 ppb, 564.5 ± 63.1 ppb, 12.9 ± 0.5°C, and 63.9 ± 5.1%, respectively. Table 2Monthly PM10 levels exposed to the infants with ADGroupaMonthly level of PM10, µg/m3P valuebAll47.8 ± 3.9Mild47.1 ± 3.3-Moderate48.3 ± 4.10.057Severe48.5 ± 5.00.264Data are expressed as mean ± standard deviation.PM10 = particulate matter with an aerodiameter < 10 µm, AD = atopic dermatitis, SCORAD = SCORing Atopic Dermatitis.aThe severity of AD was assessed using SCORAD and classified into ‘mild (< 25),’ ‘moderate (25–49),’ and ‘severe (≥ 50).’bThe t-test for means of PM10 concentration compared to that of the mild group. Monthly PM10 levels exposed to the infants with AD Data are expressed as mean ± standard deviation. PM10 = particulate matter with an aerodiameter < 10 µm, AD = atopic dermatitis, SCORAD = SCORing Atopic Dermatitis. aThe severity of AD was assessed using SCORAD and classified into ‘mild (< 25),’ ‘moderate (25–49),’ and ‘severe (≥ 50).’ bThe t-test for means of PM10 concentration compared to that of the mild group. Table 3 shows the long-term effects of PM10 exposure on AD persistence and inhalant allergen sensitization in young children. After adjusting for all confounding factors in the Cox proportional hazard regression model, the overall hazard ratio (HR) of AD persistence at age 3 was 1.04 (95% CI, 0.97−1.12) due to long-term exposure to PM10 by 1 µg/m3. In contrast, children with food allergy reduced the risk of AD persistence at age 3 (HR, 0.59; 95% CI, 0.37–0.93). There was no significant association between PM10 exposure and the incidence of sensitization to HDM or pet allergens at age 3. However, a 1 µg/m3 increase in PM10 exposure significantly increased pollen sensitization at age 3 (HR, 1.12; 95% CI, 1.00−1.24). Similarly, food allergy was significantly associated with an increased risk of pollen sensitization (HR, 2.27; 95% CI, 1.22–4.23). Table 3Effects of long-term exposure to PM10 on AD and allergen sensitization at age 3OutcomeHRd (95% CI)P valueAD persistence1.04 (0.97–1.12)0.267Sensitization to inhalant allergensHDMa (79/150)1.05 (0.96–1.14)0.290Pollenb (53/150) 1.12 (1.00–1.24) 0.043 Petc (35/150)1.00 (0.88–1.13)0.970For a statistical analysis, a Cox proportional hazards regression model was used, adjusting for confounders including ambient NO2, SO2, O3, CO, RH, temperature, sex, body mass index, comorbid food allergy, and baseline severity score. Bold values denote statistical significance at the P < 0.05 level.PM10 = particulate matter with an aerodiameter < 10 µm, AD = atopic dermatitis, HR = hazard ratio, CI = confidence interval, HDM = house dust mite, NO2 = nitric dioxide, SO2 = sulfur dioxide, O3 = ozone, CO = carbon monoxide, RH = relative humidity.aSensitized to Dermatophagoides pteronyssinus or D. farinae.bSensitized to at least one of pollen mixture allergens, including tree, grass, and weed.cSensitized to pet (cat or dog allergens).dHR due to PM10 exposure by 1 µg/m3. Effects of long-term exposure to PM10 on AD and allergen sensitization at age 3 For a statistical analysis, a Cox proportional hazards regression model was used, adjusting for confounders including ambient NO2, SO2, O3, CO, RH, temperature, sex, body mass index, comorbid food allergy, and baseline severity score. Bold values denote statistical significance at the P < 0.05 level. PM10 = particulate matter with an aerodiameter < 10 µm, AD = atopic dermatitis, HR = hazard ratio, CI = confidence interval, HDM = house dust mite, NO2 = nitric dioxide, SO2 = sulfur dioxide, O3 = ozone, CO = carbon monoxide, RH = relative humidity. aSensitized to Dermatophagoides pteronyssinus or D. farinae. bSensitized to at least one of pollen mixture allergens, including tree, grass, and weed. cSensitized to pet (cat or dog allergens). dHR due to PM10 exposure by 1 µg/m3. Effects of PM10 on AD persistence and inhalant allergen sensitizations stratified by severity were then evaluated, because the disease course over time was mostly affected by its severity (Table 4). In children with moderate AD, there was a significantly positive association between AD persistence and long-term exposure to PM10 (HR, 1.19; 95% CI, 1.05−1.36) for the moderate group by 1 µg/m3 increase of PM10 exposure. However, AD persistence in mild and severe groups was not affected by long-term exposure to PM10. Table 4Effects of long-term exposure to PM10 on AD and allergen sensitizations by severityOutcomeSeverity at enrollmentdPositive at age 3, n (%)HRe (95% CI)P valueAD persistenceMild39/72 (54.2)1.00 (0.89–1.13)0.958Moderate41/61 (67.1) 1.19 (1.05–1.36) 0.007 Severe16/17 (94.1)1.20 (0.93–1.55)0.172Allergen sensitizationHDMa Mild37/72 (51.4)0.97 (0.84–1.12)0.672Moderate29/61 (47.5)1.12 (0.97–1.29)0.119Severe13/17 (76.5)1.38 (0.91–2.09)0.127Pollenb Mild23/72 (31.9)1.02 (0.82–1.26)0.885Moderate23/61 (37.7) 1.31 (1.08–1.58) 0.005 Severe7/17 (41.2)NAPetc Mild17/72 (23.6)0.91 (0.74–1.12)0.367Moderate13/61 (21.3)1.09 (0.86–1.39)0.483Severe5/17 (29.4)NAFor a statistical analysis, a Cox proportional hazards regression model was used, adjusting for confounders including ambient NO2, SO2, O3, CO, RH, temperature, sex, body mass index, comorbid food allergy, and baseline severity score. Bold values denote statistical significance at the P < 0.05 level.PM10 = particulate matter with an aerodiameter < 10 µm, AD = atopic dermatitis, HR = hazard ratio, CI = confidence interval, HDM = house dust mite, NO2 = nitric dioxide, SO2 = sulfur dioxide, O3 = ozone, CO = carbon monoxide, RH = relative humidity, SCORAD = SCORing Atopic Dermatitis.aSensitized to Dermatophagoides pteronyssinus or D. farinae.bSensitized to at least one of pollen mixture allergens, including tree, grass, and weed.cSensitized to pet (cat or dog allergens).dThe severity of AD was assessed using SCORAD and classified into mild (< 25), moderate (25–49), and severe (≥ 50).eHR due to PM10 exposure by 1 µg/m3. Effects of long-term exposure to PM10 on AD and allergen sensitizations by severity For a statistical analysis, a Cox proportional hazards regression model was used, adjusting for confounders including ambient NO2, SO2, O3, CO, RH, temperature, sex, body mass index, comorbid food allergy, and baseline severity score. Bold values denote statistical significance at the P < 0.05 level. PM10 = particulate matter with an aerodiameter < 10 µm, AD = atopic dermatitis, HR = hazard ratio, CI = confidence interval, HDM = house dust mite, NO2 = nitric dioxide, SO2 = sulfur dioxide, O3 = ozone, CO = carbon monoxide, RH = relative humidity, SCORAD = SCORing Atopic Dermatitis. aSensitized to Dermatophagoides pteronyssinus or D. farinae. bSensitized to at least one of pollen mixture allergens, including tree, grass, and weed. cSensitized to pet (cat or dog allergens). dThe severity of AD was assessed using SCORAD and classified into mild (< 25), moderate (25–49), and severe (≥ 50). eHR due to PM10 exposure by 1 µg/m3. When effects of ambient PM10 exposure on incidence of HDM sensitization were stratified by severity, there were no statistically significant effects in any subgroups. PM exposure was also not associated with sensitization to pet allergens either, regardless of AD severity. Of interest, pollen sensitization was significantly associated with ambient PM10 exposure in children with moderate AD (HR, 1.31; 95% CI, 1.08−1.58), but not in those with mild AD (HR, 1.02; 95% CI, 0.82–1.26). In the mild AD group, comorbid food allergy showed a significant association with pollen sensitization (HR, 4.07; 95% CI, 1.20–13.79). In the severe group, the association of ambient PM10 exposure with pollen sensitization could not be evaluated statistically because the number of patients was too small. DISCUSSION The objective of this investigation was to elucidate the impact of chronic exposure to PM on the progression of AD in infants. Our study revealed that long-term PM10 exposure significantly raised the risk of AD persistence at age 3, particularly in patients with moderate AD. Additionally, this exposure notably increased sensitization to pollen, rather than to HDM or pet allergens. Skin damage attributed to PM primarily results from oxidative stress and subsequent inflammation.21 Upon exposure to PM with polycyclic aromatic hydrocarbons (PAHs), human keratinocytes increase the production of reactive oxygen species (ROS).22 PM2.5 can also stimulate interleukin-6 production in human keratinocytes via ROS production and oxidative stress.23 In addition, keratinocytes exposed to PM can upregulate cytokine expression, indirectly leading to diminished expression of skin barrier proteins and skin barrier dysfunction. PM2.5 exposure can induce the expression of tumor necrosis factor-α in human keratinocytes, which subsequently suppresses the expression of filaggrin (FLG) and loricrin through an aryl hydrocarbon receptor (AhR)-dependent pathway.24 PAHs as components of PM are known to penetrate the stratum corneum effectively due to their lipophilic properties, initiating skin inflammation through binding to the AhR, a ligand-dependent transcription factor.25 PM can also trigger the release of alarmins such as thymic stromal lymphopoietin from keratinocytes, which can then promote Th2-type immune response and reduce FLG expression.26,27 These findings indicate that PM can alter the molecular structure and function of the epidermal barrier, thereby exacerbating AD. A methodological challenge in our study was how to accurately quantify individual patient exposure to ambient PM. To address this, we utilized an advanced exposure assessment approach, integrating data from AQMS with the CMAQ modeling outputs and updating the relocation of patients’ residences for 3 years. This fusion technique of exposure assessment is more advantageous than AQMS alone in urban environments with spatial variability and limited monitoring resources, making our results more reliable. Consequently, by dynamically tracking PM10 exposure over the study period, individual exposure levels reflected variations in locations and relocation patterns, even within regions that appear to have similar overall air quality. In this study, prolonged exposure to PM was significantly associated with AD persistence in children with moderate AD. In contrast, no such associations were observed in the mild or severe AD group. This disparity may reflect greater susceptibility of the skin with moderate AD to PM penetration and subsequent inflammatory response, compared to the skin with mild AD. The lack of statistical significance in severe AD is likely due to the small sample size and the fact that severity itself is a strong risk factor for AD persistence. In other words, in patients with severe AD, the impact of long-term exposure to PM on AD persistence may be masked by the predominant influence of disease severity, which could explain the absence of a statistically significant association. Pollen sensitization is associated with PM exposure,28,29 and is increased in patients with AD.30 Exposure to pollen, in turn, leads to an exacerbation of AD symptoms.31 Indeed, bioaerosols from pollen adhere to PM and facilitate penetration into the human body, potentially exacerbating allergic responses.32 Consequently, the identification and mitigation of risk factors contributing to pollen sensitization in infants with AD are important for favorable disease prognosis. This study demonstrated the effect of long-term exposure to PM on pollen sensitization among infants with moderate AD. Our observation is consistent with the results from the previous meta-analysis of 5 birth cohort studies,28 where increased exposure to ambient PM2.5 was associated with sensitization to outdoor allergens, not indoor allergens. Our finding supports the implementation of proactive strategies aimed at reducing PM exposure in infants with AD to modify the disease course. In contrast to pollen exposed outdoors, the present study demonstrated that sensitization to HDM, an indoor allergen, did not have an association with PM exposure. This result is not consistent with a previous report showing that epicutaneous sensitization to various allergens including HDM could occur.8 Our observation might be attributed to the estimation of long-term PM exposure based on ambient PM concentrations. However, upon examining the relationship between PM exposure and HDM sensitization in patients with moderate or severe AD, the HR was 1.12 (95% CI, 0.97−1.29) and 1.38 (95% CI, 0.91−2.09), respectively. This suggests that an increase in the sample size of moderate or severe AD patients might yield statistically significant results. Regarding sensitization to pet allergens, 35 individuals (23.3%) demonstrated sensitization to either cat or dog. Specifically, 22 patients (14.7%) exhibited sensitization to cat allergens, and 21 patients (14.0%) to dog allergens. These figures are relatively lower than those observed for sensitization to HDM (52.7%) and pollen (35.3%), rendering it challenging to achieve statistical significance in the present study. Therefore, drawing conclusions from current findings proves to be complex. Food allergy is known to be an important factor influencing AD persistence or atopic march.33,34,35 In the present study, food allergy had a statistically significant effect on AD persistence after adjusting with all confounding factors. However, unlike previous studies, our findings indicated that food allergy was associated with a reduced risk of AD persistence. This may be because a greater proportion of patients with food allergy in our study belonged to the mild group compared to those without food allergy (53.3% vs. 44.4%). It is also possible that early diagnosis of food allergy led to better adherence to dietary restrictions under medical guidance, resulting in improved eczema control and reduced AD persistence. There are limitations in this study. Firstly, chronic exposure to PM was estimated from birth to age 3 years rather than commencing from AD onset to age 3. The reason was that the exact time of AD onset could not be identified. Rather, pathophysiological changes of AD might have begun immediately after birth. Secondly, we did not include indoor environments in exposure assessments, including indoor PM, temperature, relative humidity, volatile organic compounds, and so on. However, it is not feasible to monitor indoor environmental factors at each patient’s residence for such a long period covering over 3 years. Since indoor PM originates partly from the outside, the influence of indoor PM exposure was not entirely excluded in this study. Thirdly, we did not assess whether children were sensitized to inhalant allergens at birth or upon study enrollment, making it hard to exclude pre-existing sensitizations before age 3. Indeed, sensitization to food allergens mostly occurs in infancy, whereas inhalant allergen sensitization is uncommon at this age and increases over time.36,37 For instance, a German birth cohort study found inhalant allergen sensitization rates of 1.5% at age 1 and 26% at age 6.38 Thus, inhalant allergen sensitization observed at age 3 likely represents new sensitizations. Fourthly, socioeconomic status, which is known to influence health outcomes including AD, was not accounted for in this study due to a lack of available data. Finally, selection bias was also possible, because the present study was conducted in a single center and very mild patients were less likely to continue follow-up visits by age 3. In conclusion, chronic exposure to ambient PM10 in infants with AD may affect the disease course by decreasing remission and increasing sensitization to pollen at age 3. In infants with AD, efforts are needed to reduce short-term and long-term exposure to PM in early life to prevent acute exacerbation and improve disease prognosis. ACKNOWLEDGMENTS The authors would like to thank Bohyeon Kim, Younghee Chung, an audiovisual engineer at Samsung Medical Center Information & Medical Services, for designing for this work. Footnotes Funding: This work was supported by the Research Program funded Korea National Institute of Health (2021-ER120400-02). Disclosure: The authors have no potential conflicts of interest to disclose. Author Contributions: Conceptualization:Kim S, Oh I, Choi J, Kim J, Bang JH, Kim YM, Ahn K.Data curation:Kim S, Oh I, Choi J, Kim J, Bang JH, Kim YM, Ahn K.Formal analysis:Kim S, Oh I, Choi J, Kim J, Kim J, Bang JH, Kim YM, Ahn K.Supervision:Oh I, Kim J, Ahn K.Writing - original draft:Kim S, Oh I, Kim J, Kim YM, Ahn K.Writing - review & editing:Oh I, Kim J, Kim YM, Ahn K. Conceptualization:Kim S, Oh I, Choi J, Kim J, Bang JH, Kim YM, Ahn K. Conceptualization:Kim S, Oh I, Choi J, Kim J, Bang JH, Kim YM, Ahn K. Data curation:Kim S, Oh I, Choi J, Kim J, Bang JH, Kim YM, Ahn K. Data curation:Kim S, Oh I, Choi J, Kim J, Bang JH, Kim YM, Ahn K. Formal analysis:Kim S, Oh I, Choi J, Kim J, Kim J, Bang JH, Kim YM, Ahn K. Formal analysis:Kim S, Oh I, Choi J, Kim J, Kim J, Bang JH, Kim YM, Ahn K. Supervision:Oh I, Kim J, Ahn K. Supervision:Oh I, Kim J, Ahn K. Writing - original draft:Kim S, Oh I, Kim J, Kim YM, Ahn K. Writing - original draft:Kim S, Oh I, Kim J, Kim YM, Ahn K. Writing - review & editing:Oh I, Kim J, Kim YM, Ahn K. Writing - review & editing:Oh I, Kim J, Kim YM, Ahn K. References