Impact of particulate matter 2.5 exposure on severity of pediatric chronic rhinitis: A prospective cohort study Authors: Rinrada Naka, Ratchaneewan Sinitkul, Ussanai Nithirochananont, Watcharoot Kanchongkittiphon, Potjanee Kiewngam, Wanlapa Jotikasthira, Adithep Sawatchai, Chaiyawat Suppasilp, Wiparat Manuyakorn Published: 2026 ABSTRACT AbstractBackgroundParticulate matter with a diameter of ≤2.5 micrometers (PM2.5) is associated with respiratory and allergic diseases. However, the longitudinal impact on pediatric rhinitis control remains limited.ObjectivesTo determine the association between PM2.5 exposure and symptom control in pediatric patients with chronic rhinitis.MethodsThis prospective longitudinal study (July 2024–August 2025) followed children with chronic rhinitis in the Bangkok Metropolitan Region. Rhinitis symptom severity was assessed weekly using a visual analog scale (VAS), with covariates such as medication use and concurrent respiratory infections. A 1‐week average PM2.5 concentration preceding VAS assessment (via Inverse Distance Weighting interpolation) was estimated. Associations between PM2.5 exposure and VAS were analyzed using multivariable mixed‐effects and fixed‐effects linear regression models. Additional analyses were stratified by aeroallergen sensitization status (allergic rhinitis versus nonallergic rhinitis).ResultsAmong 146 children followed for a median of 52 weeks, higher PM2.5 exposure (per 10 μg/m3) was consistently associated with increased VAS (β coefficient; 95% confidence interval = 0.260; 0.215–0.306, p < .001). Wet season, concurrent respiratory infections, and medication use were also associated with increased VAS. These associations were observed in both allergic rhinitis and nonallergic rhinitis.ConclusionHigher short‐term ambient PM2.5 exposure was associated with greater rhinitis symptom severity in children with chronic rhinitis. These findings support consideration of ambient air pollution as an environmental factor in pediatric rhinitis management. Keywords: air pollution, allergic rhinitis, pediatric rhinitis, PM2.5 , Thailand Key message The PM2.5 exposure is associated with increased rhinitis symptoms in children. INTRODUCTION Fine particulate matter (PM2.5), which has a diameter of ≤2.5 micrometers, remains a major global health concern. 1 , 2 , 3 In Thailand, ambient PM2.5 levels generally exceed the World Health Organization (WHO) recommended thresholds of 5 μg/m3 for annual mean concentration and 15 μg/m3 for 24‐h mean concentration. 1 , 4 , 5 Due to its small size, PM2.5 can penetrate the lower airways and induce inflammatory responses. 1 , 2 , 3 Daily variations in PM2.5 concentrations occur due to traffic emissions, industrial activities, and seasonal meteorological changes. 1 , 2 , 3 Chronic rhinitis is a common nasal inflammatory disorder, defined by the presence of at least two symptoms—nasal obstruction, rhinorrhea, sneezing, or nasal, ocular, or oral pruritus—persisting for at least 1 h per episode and lasting more than 12 weeks per year. 6 It is broadly classified into allergic rhinitis (AR), an IgE‐mediated hypersensitivity condition, and nonallergic rhinitis (NAR), which involves inflammation without evidence of allergic sensitization. 6 Both phenotypes affect quality of life, sleep, school performance, and daily functioning. 6 , 7 Rhinitis severity is influenced by multiple factors, including respiratory infections, allergen exposure, medication adherence, environmental conditions, and air pollution, including ambient PM2.5 exposure. 6 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 A systematic review reported that prenatal and early‐life PM2.5 exposure increases childhood AR risk. 12 Furthermore, population‐based studies have shown that higher annual PM2.5 concentrations are associated with increased rhinitis prevalence and symptom severity. 13 In adults, long‐term PM2.5 exposure has been linked to persistent symptoms across multi‐country cohorts. 14 Although epidemiologic studies have linked PM2.5 exposure with rhinitis prevalence and symptom severity, evidence from prospective longitudinal studies of short‐term exposure and repeated symptom assessment in children remains limited. Children are particularly relevant because their environmental exposure patterns and symptom reporting may differ from those of adults. A recent study among university students in northern Thailand reported worsening symptoms and reduced well‐being during high PM2.5 seasons. 15 However, evidence regarding the effects of PM2.5 on pediatric rhinitis in urban settings remains limited. This gap is important because children are more vulnerable to particulate matter. 6 , 11 , 13 The Bangkok Metropolitan Region (BMR) represents a distinct exposure context, characterized by relatively lower seasonal peaks but sustained year‐round PM2.5 levels and a different chemical composition dominated by traffic‐related and urban emission sources, in contrast to the biomass‐burning pollution commonly observed in northern Thailand. 4 , 5 Therefore, this prospective longitudinal study aimed to examine the association between short‐term PM2.5 exposure and the severity of chronic rhinitis symptoms in children living in the BMR. METHODS Study design This prospective observational cohort study was conducted at the Pediatric Allergy Clinic, Ramathibodi Hospital, Mahidol University in Bangkok, Thailand, from July 2024 to August 2025. Eligible participants were children aged 2–18 years diagnosed with chronic rhinitis by pediatric allergists. Both their residential areas and schools were located within the BMR. Children were excluded if they had a critical illness, an active malignancy, an autoimmune disease, or a history of a respiratory tract infection within 4 weeks prior to enrollment. Aeroallergen sensitization was evaluated by the skin prick test (SPT). Participants with a positive skin prick test to at least one aeroallergen were categorized as having allergic rhinitis (AR), whereas participants without evidence of aeroallergen sensitization were categorized as having nonallergic rhinitis (NAR). 6 , 16 , 17 The SPT was performed after participants discontinued their oral antihistamine for 1 week using a blood lancet (Feather Safety Razor) on the volar aspect of the forearm with a positive control (0.1% histamine solution), normal saline as a negative control, and 9 commercial aeroallergen extracts which were Dermatophagoides pteronyssinus (Der p), Dermatophagoides farini (Der f), cat hair, Careless Weed, Mixed cockroach (produced by a partnership between Faculty of Medicine Siriraj Hospital, Mahidol University and Greater Pharma Manufacturing Co., Ltd.), Johnson grass, dog pelt (ALK‐AbelloPharm., Inc., Mississauga, On), Bermuda grass, and Mixed mold (Cladosporium Sphaerospermum, Alternaria, Aspergillus fumigatus, and Curvularia), (AllerMed SanDiego, CA). Assessment of individual PM2.5 exposure Individual PM2.5 exposure was estimated by linking geocoded residential and school addresses to interpolated ambient PM2.5 concentration. Hourly PM2.5 concentrations were obtained from 171 ambient air‐monitoring stations across Thailand, including 95 within the BMR, operated by the Thailand Pollution Control Department (PCD) and the Bangkok Metropolitan Administration (BMA). Data were quality‐assured and verified by the PCD before public release. Hourly PM2.5 concentrations were spatially interpolated using the Inverse Distance Weighting (IDW) method in Quantum Geographic Information System (QGIS) to generate raster surfaces. PM2.5 concentrations corresponding to each participant's residential and school locations were extracted from these raster surfaces to estimate individual exposure. As Visual Analog Scale (VAS) symptom severity was assessed weekly, predefined 1‐week PM2.5 exposure windows were used. The 1‐week average PM2.5 concentration was the mean 24‐h Time‐Weighted Average (TWA) ambient PM2.5 concentration during the 7 days preceding each VAS assessment. 24‐h TWA ambient PM2.5 exposure was calculated according to estimated time spent at each location (Data S1). Preschool children were assumed to spend the entire day at their residence; therefore, daily exposure was based on residential ambient PM2.5. School‐aged children were assumed to spend 8 h per weekday (08:00–16:00) at school and 16 h at home, consistent with the standard Thai school schedule. Outcome assessment (VAS of rhinitis symptoms) Participants or their guardians completed weekly questionnaires on rhinitis symptoms, including nasal congestion, rhinorrhea, sneezing, nasal itching, watery eyes, and ocular pruritus, via Google Forms. Symptom severity was quantified using a VAS from 0 to 10, where 0 indicated no symptoms, and 10 indicated the most severe symptoms. For participants younger than 12 years, who may not have reached the formal operational stage of cognitive development, the questionnaires were completed by their guardians to ensure reliability. 16 The primary outcome was the highest score of VAS from the following six recorded rhinitis symptoms, which were nasal congestion, rhinorrhea, sneezing, nasal itching, watery eyes, and ocular itching, representing the most severe rhinitis symptoms during the week. Monthly telephone or message follow‐ups were performed to maintain adherence and verify data. Missing weekly assessments were assumed to be missing at random. All available observations were analyzed without data imputation. However, participants with completed questionnaires for less than 5 of 52 weeks (10%) were considered lost to follow‐up and excluded. Covariates Potential confounders included demographics (age, gender, and BMI z‐score), residential factors (household smoke exposure, mode of transportation, indoor cooking activity, printer use, incense use, pet ownership, and regular air purifier use), and time‐varying covariates (regular use of intranasal corticosteroids [INCS], rescue medication, and wet season). Medication use was considered a time‐varying marker of underlying disease activity and was therefore interpreted with caution, as it may reflect worsening symptoms. Rescue medication refers to the intermittent escalation of treatment during symptom aggravation, including antihistamines and INCS. 6 , 18 Seasons were classified into wet season (mid‐May to mid‐October) and dry season (mid‐October to mid‐May). 19 , 20 , 21 Ethical considerations This study was approved by the Institutional Review Board of the Faculty of Medicine, Ramathibodi Hospital, Mahidol University (MURA2024/355). Written informed consent was obtained from guardians, and informed assent was obtained from participants aged ≥ 7 years before enrollment. Statistical analysis All analyses were conducted using Stata version 18 (StataCorp LLC, College Station, TX, USA). Descriptive statistics were summarized as frequencies and percentages for categorical variables, means (SDs) for normally distributed continuous variables, and medians (IQRs) for non‐normally distributed continuous variables. Multivariable mixed‐effects linear regression models were used to assess the associations between the 1‐week average PM2.5 concentrations and VAS, with participant‐specific random intercepts fitted to additionally evaluate time‐invariant covariates, including age. Covariates were selected based on clinical knowledge and prior literature; variables with p < .15 in univariable analyses were considered for inclusion and subjected to backward elimination. Subgroup analyses were performed for AR and NAR groups. As sensitivity analyses, linear fixed‐effects models with standard errors clustered at the participant level were fitted. PM2.5 was analyzed primarily as a continuous exposure. Additional sensitivity analyses categorized PM2.5 exposure using both the standard WHO guideline threshold and a dataset‐specific, data‐driven threshold to assess the robustness of the findings. Results were reported as regression coefficients (β) with 95% confidence intervals (CIs), and a two‐sided p < .05 was considered statistically significant. RESULTS Baseline characteristics Of the 172 enrolled participants, 26 were excluded for loss to follow‐up. The mean VAS of the excluded participants before discontinuation was 2.58 ± 2.42, compared with 2.67 ± 1.55 among the analyzed participants. A total of 146 participants (101 with AR and 45 with NAR) were included. The mean age was 7.87 ± 3.41 years. The mean age of participants with AR and NAR was 8.38 ± 3.46 years and 6.73 ± 3.00 years, respectively (p = .007). One hundred and twenty‐four participants (84.93%) were younger than 12 years old, and therefore, the questionnaires were completed by their guardians. Male participants accounted for 54.79% of the sample. The mean BMI z‐score was 0.13 ± 1.22, with no significant difference between groups. Residential factors showed no significant differences between groups, with 23.29% of participants exposed to household smoke. For time‐varying covariates, concurrent respiratory infections accounted for 1.64% of person‐weeks. The median follow‐up duration was 52 ± 6 weeks. The full baseline characteristics are presented (Table 1). TABLE 1Baseline characteristics.Variables, n (%)Total n = 146 (100%)Allergic rhinitis n = 101 (69.18%)Nonallergic rhinitis n = 45 (30.82%)Demographic dataAge; years, mean (SD)7.87 (3.41)8.38 (3.46)6.73 (3.00)Gender; male80 (54.79)59 (58.42)21 (46.67)BMI; median (IQR)15.79 (4.19)15.73 (4.19)16.01 (3.51)<18.5109 (74.66)75 (74.26)34 (75.56)18.5–2327 (18.49)20 (19.80)7 (15.55)>2310 (6.85)6 (5.94)4 (8.89)BMI z‐score; mean (SD)0.13 (1.22)0.02 (1.42)0.37 (1.23)Environmental exposuresHousehold smoke exposure34 (23.29)20 (19.80)14 (31.11)Mode of transportationSolely private cars75 (51.37)55 (54.46)21 (44.44)Mixed modalities (public transport, motorcycles and private cars)71 (48.63)46 (45.54)25 (55.56)Indoor cooking activity81 (55.48)54 (53.47)27 (60.00)Household printer use23 (15.75)16 (15.84)7 (15.56)Household incense product use22 (15.07)15 (14.85)7 (15.56)Pet ownership38 (26.03)26 (25.74)12 (26.67)Lack of regular air purifier use62 (42.47)44 (43.56)18 (40.00)Time‐varying covariatesConcurrent respiratory infections (Time‐varying; Person‐weeks)115 (1.64)81 (1.68)34 (1.56)Medication (Time‐varying; Person‐weeks)No regular use of INCS without further rescue medication687 (9.82)475 (9.86)212 (9.73)No regular use of INCS but with rescue medication16 (0.23)1 (0.02)15 (0.69)Regular use of INCS without further rescue medication5979 (85.46)4111 (85.34)1868 (85.73)Regular use of INCS but with rescue medication314 (4.49)230 (4.77)84 (3.85)Duration of Follow‐up (weeks); median (IQR)52 (6)52 (6)52 (6)Abbreviations: BMI, body mass index (kg/m2); INCS, intranasal corticosteroids. Baseline characteristics. Abbreviations: BMI, body mass index (kg/m2); INCS, intranasal corticosteroids. Baseline PM2.5 exposure and longitudinal VAS of rhinitis symptoms Air‐monitoring stations, participants' residences, and schools were displayed (Figure 1A). The overall median PM2.5 concentration across the study was 20.32 μg/m3 (IQR 15.06–30.91). Distribution of the median 1‐week average PM2.5 concentrations primarily exceeded 15 μg/m3 during the dry season, particularly between December 2024 and March 2025 (Figure 1B). The median VAS showed a generally similar temporal pattern with PM2.5 concentrations. Higher median VAS was observed during periods of elevated PM2.5 exposure in the dry season, peaking during December 2024–January 2025. The lowest median VAS was presented during April–May 2025. However, a partial non‐concordance pattern was observed during the wet season, where the median VAS increased despite lower PM2.5 concentrations. (A) Locations of participants' residences and schools, air monitoring station in the study area. (B) Distribution of weekly median PM2.5 concentrations and median participant‐reported VAS across the 14‐month study period. VAS, visual analog scale. Association between PM2.5 exposure and VAS of rhinitis symptoms In univariable mixed‐effects analyses, 1‐week average PM2.5 concentrations (per 10 μg/m3 increase) preceding VAS assessment were significantly associated with higher VAS (β coefficient [β] = 0.112; 95% confidence interval [CI], 0.034–0.190; p .005) (Table 2). The wet season and concurrent respiratory infections were also significantly associated with increased VAS. Medication use patterns were also associated with symptom severity. Rescue medication use was associated with higher VAS, regardless of regular use of INCS. Older age was associated with lower VAS, whereas other baseline characteristics and residential factors were not significantly associated with VAS. TABLE 2Univariable multilevel mixed‐effects regression analysis of factors associated with VAS.VariablesCrude effects β (95% CI) p ValuePM2.5 concentrationsPM2.5 concentration (per 10 μg/m3)0.112 (0.034, 0.190).005Wet season0.283 (0.072, 0.495).009Demographic dataAge; years−0.072 (−0.142, −0.002).045Male−0.105 (−0.604, 0.393).679BMI; kg/m2, median (Q1, Q3)−0.018 (−0.098, 0.062).660BMI group; kg/m2 <18.50—18.5–23−0.577 (−1.229, 0.075).083>23−0.227 (−1.261, 0.807).667Residential and environmental exposuresHousehold smoke exposure0.558 (−0.012, 1.129).055Mixed modalities for transportation0.105 (−0.394, 0.603).681Indoor cooking activity0.305 (−0.189, 0.800).226Household printer use0.405 (−0.320, 1.129).274Household incense product use0.027 (−0.685, 0.740).940Pet ownership0.223 (−0.363, 0.809).456Lack of regular air purifier use−0.150 (−0.647, 0.348).555Concurrent respiratory infections (Time‐varying; Person‐weeks)3.191 (2.728, 3.654)<.001Medication (Time‐varying; Person‐weeks)No regular use of INCS without further rescue medication0—No regular use of INCS but with rescue medication3.729 (2.113, 5.345)<.001Regular use of INCS without further rescue medication0.911 (0.379, 1.443).001Regular use of INCS with further rescue medication3.088 (2.391, 3.785)<.001Abbreviations: BMI, body mass index (kg/m2); INCS, intranasal corticosteroids; VAS, visual analog scale. Univariable multilevel mixed‐effects regression analysis of factors associated with VAS. Abbreviations: BMI, body mass index (kg/m2); INCS, intranasal corticosteroids; VAS, visual analog scale. In multivariable mixed‐effects models, PM2.5 concentrations remained significantly associated with higher VAS after adjustment for significant covariates. Each 10 μg/m3 increase in 1‐week average PM2.5 concentration was associated with an estimated 0.260‐point increase in the adjusted VAS (β = 0.260; 95% CI, 0.215, 0.306; p < .001) (Table 3). Higher VAS were also observed during the wet season and during concurrent respiratory infections. Weeks requiring rescue medication showed higher VAS compared to weeks without rescue medication, regardless of regular INCS use. TABLE 3A comparative multivariable regression analysis of population‐average (mixed‐effects) and within‐individual (fixed‐effects) models.VariablesMixed‐effectsFixed‐effects β (95% CI) p Value β (95% CI) p ValueTotal; n = 146 (100%)PM2.5 concentrations (per 10 μg/m3)0.260 (0.215, 0.306)<.0010.260 (0.189, 0.332)<.001Wet season0.629 (0.519, 0.738)<.0010.629 (0.408, 0.85)<.001Age; years−0.072 (−0.141, −0.004).037Omitted—Concurrent respiratory infections2.804 (2.472, 3.137)<.0012.797 (2.294, 3.301)<.001MedicationNo regular use of INCS without further rescue medication0—0—No regular use of INCS but with rescue medication3.578 (2.636, 4.519)<.0013.577 (1.996, 5.158)<.001Regular use of INCS without further rescue medication0.812 (0.542, 1.082)<.0010.749 (0.109, 1.389).022Regular use of INCS with rescue medication2.786 (2.452, 3.120)<.0012.709 (1.934, 3.485)<.001AR; n = 101 (69.18%)PM2.5 concentrations (per 10 μg/m3)0.253 (0.16, 0.338)<.0010.253 (0.167, 0.340)<.001Wet season0.676 (0.451, 0.901)<.0010.678 (0.451, 0.905).035Age; years−0.094 (−0.164, −0.025).007Omitted—Concurrent respiratory infections2.793 (2.188, 3.398)<.0012.786 (2.175, 3.398)<.001MedicationNo regular use of INCS without further rescue medication0—0—No regular use of INCS but with rescue medication−0.014 (−0.190, 0.161).8730.005 (−0.180, 0.191).956Regular use of INCS without further rescue medication0.745 (0.259, 1.231).0030.661 (0.097, 1.224).022Regular use of INCS with rescue medication2.644 (1.928, 3.360)<.0012.548 (1.767, 3.330)<.001NAR; n = 45 (30.82%)PM2.5 concentrations (per 10 μg/m3)0.276 (0.147, 0.405)<.0010.275 (0.142, 0.409)<.001Wet season0.531 (0.036, 1.027).0350.530 (0.020, 1.039).042Age; years−0.037 (−0.188, 0.114).633Omitted—Concurrent respiratory infections2.814 (1.923, 3.705)<.0012.807 (1.888, 3.726)<.001MedicationNo regular use of INCS without further rescue medication0—0—No regular use of INCS but with rescue medication3.951 (1.964, 5.939)<.0013.966 (1.869, 6.063)<.001Regular use of INCS without further rescue medication1.003 (−0.737, 2.744).2590.999 (0.904, 3.106).344Regular use of INCS with rescue medication3.121 (1.256, 4.986).0013.100 (0.904, 5.295).007Abbreviations: AR, allergic rhinitis; INCS, intranasal steroids; NAR, nonallergic rhinitis. A comparative multivariable regression analysis of population‐average (mixed‐effects) and within‐individual (fixed‐effects) models. Abbreviations: AR, allergic rhinitis; INCS, intranasal steroids; NAR, nonallergic rhinitis. In subgroup analyses, 1‐week average PM2.5 concentration was associated with increased VAS in both the AR group (β = 0.253; 95% CI = 0.168, 0.338; p < .001) and the NAR group (β = 0.276; 95% CI = 0.147, 0.405; p < .001). Given the smaller subgroup sample sizes, particularly for NAR, these findings were interpreted descriptively and were not used to infer differences in susceptibility between rhinitis phenotypes. Wet season and concurrent respiratory infection showed a similar relationship with VAS in both the AR and NAR groups. Rescue medication use was significantly associated with higher VAS in the NAR group, whereas regular INCS use was associated with higher VAS in the AR group. For sensitivity analyses, a similar magnitude was observed in the fixed‐effects models (β = 0.260; 95% CI = 0.189, 0.332; p < .001). The predicted dose–response relationships between 1‐week average PM2.5 concentration and VAS from the final multivariable mixed‐effects and fixed‐effects models (Figure 2A,B) showed a steady increase in predicted VAS correlated with higher PM2.5 concentrations, similarly in both models. (A) Predicted dose–response relationship between PM2.5 exposure and VAS scores analyzed by mixed‐effects model. (B) Predicted dose–response relationship between PM2.5 exposure and VAS scores analyzed by fixed‐effects model. VAS, visual analog scale. Threshold analyses using the WHO guideline and data‐driven quartiles also supported a significant dose–response relationship between PM2.5 exposure and adjusted VAS (Table 4). In the WHO guideline‐based models, adjusted VAS was significantly elevated when PM2.5 levels exceeded 37.5 μg/m3 compared with the reference group (<15 μg/m3) (β = 0.381; 95% CI = 0.160, 0.603; p = .001). Similarly, quartile‐based analyses showed significant increases in adjusted VAS at the third quartile (20.32–30.91 μg/m3) (β = 0.304; 95% CI = 0.076, 0.532; p = .009) and peaking in the highest quartile (>30.91 μg/m3) (β = 0.862; 95% CI = 0.595, 1.128; p < .001). In contrast, no significant difference was observed for the second quartile (15.05–20.32 μg/m3) (β = −0.017; 95% CI = −0.171, 0.137; p = .830). Overall, similar patterns were seen in both mixed models and fixed models. TABLE 4Exploratory categorical analyses of 1‐week average PM2.5 concentrations preceding VAS assessment using WHO‐informed concentration categories and data‐driven quartiles.ModelsCrude effectsAdjusted effects β (95% CI) p β (95% CI) p ValueMixed‐effects a WHO cutoff<150—0—15–37.5−0.164 (−0.368, 0.039).1140.113 (−0.053, 0.279).181>37.50.076 (−0.192, 0.343).5790.381 (0.160, 0.603).001Quartile cutoff<15.0570—0—15.057–20.324−0.270 (−0.430, −0.111).001−0.017 (−0.171, 0.137).83020.324–30.912−0.318 (−0.567, −0.069).0120.304 (0.076, 0.532).009>30.9120.245 (−0.025, 0.515).0760.862 (0.595, 1.128)<.001Fixed‐effects b WHO cutoff<150—0—15–37.5−0.165 (−0.370, 0.041).1150.112 (−0.056, 0.280).188>37.50.075 (−0.195, 0.345).5820.380 (0.156, 0.603).001Quartile cutoff<15.0570—0—15.057–20.324−0.272 (−0.433, −0.110).001−0.018 (−0.173, 0.137).81920.324–30.912−0.318 (−0.569, −0.067).0130.303 (0.073, 0.534).010>30.9120.244 (−0.028, 0.517).0780.862 (0.593, 1.131)<.001Abbreviations: VAS, visual analog scale; WHO, World Health Organization. a Adjusted for wet season, infection, medication, age. b Adjusted for wet season, infection, medication. Exploratory categorical analyses of 1‐week average PM2.5 concentrations preceding VAS assessment using WHO‐informed concentration categories and data‐driven quartiles. Abbreviations: VAS, visual analog scale; WHO, World Health Organization. Adjusted for wet season, infection, medication, age. Adjusted for wet season, infection, medication. DISCUSSION This prospective observational study found that short‐term ambient PM2.5 exposure was associated with increased severity of rhinitis symptoms in children residing and attending school in the BMR. Each 10 μg/m3 increase in weekly PM2.5 exposure was associated with a 0.260‐point increase in VAS, with consistent findings across fixed‐effects and mixed‐effects analyses for within‐participant and population‐average levels. Although modest at the individual level, baseline symptom burden was relatively low (mean VAS, 2.67 ± 1.55), and the ambient PM2.5 may vary by substantially more than 10 μg/m3. Assuming a linear exposure–response relationship, differences of 20 and 30 μg/m3 would correspond to approximately 0.52‐ and 0.78‐point higher VAS, respectively. The observed PM2.5 interquartile range corresponded to approximately 0.41 points in VAS. Thus, larger real‐world exposure contrasts may be associated with more appreciable symptom differences. However, as a minimal clinically important difference has not been validated for the repeated pediatric rhinitis VAS used in this study, the clinical magnitude should be interpreted cautiously. Furthermore, sensitivity and threshold analyses supported a dose–response relationship, showing progressively higher symptom scores at increasing PM2.5 concentrations. Our findings align with previous literature indicating that PM2.5 is associated with rhinitis symptoms. 12 , 13 , 14 , 15 Similar correlations were reported in Chinese adolescents 13 and European adults. 14 In addition to the findings of previous studies, our results extend the evidence across a broader pediatric population. In Thailand, Pisithkul et al. (2024) observed worsening symptoms among university students in Chiang Mai during periods of high PM2.5 seasons. 15 Our study adds to this evidence by demonstrating these effects in a broader pediatric population within a traffic‐dominated urban setting. During the study period, PM2.5 concentrations and rhinitis symptoms shared a broadly concordant temporal trend. The median PM2.5 concentration across the study was 20.32 μg/m3, exceeding the 2021 WHO 24‐h recommended threshold of 15 μg/m3. Higher PM2.5 levels were observed during the dry season, aligned with air quality reports from PCD of Thailand, 4 , 5 and coincided with higher mean VAS. These findings suggested a concordant increase in rhinitis symptom burden during months with elevated ambient PM2.5 exposure, with the highest concentrations and VAS during December 2024–March 2025. Similarly, the lowest VAS was observed during March–April 2025 when PM2.5 concentrations were relatively lower. This finding supported a concordant temporal trend between PM2.5 exposure and rhinitis symptoms. However, a partial discordance was observed during the wet season. This pattern suggested that rhinitis symptoms may also be influenced by additional factors, including the wet season and respiratory infection triggers. In tropical settings such as Thailand, the wet season coincides with increased environmental humidity, which can elevate mold spore concentrations and increase the incidence of respiratory tract infections, particularly viral infections. 8 , 9 , 10 , 22 , 23 These conditions may aggravate nasal mucosal inflammation and contribute to worsening rhinitis symptoms despite lower PM2.5 exposure. 6 , 7 , 8 , 9 , 10 In addition to PM2.5 exposure, several other factors may contribute to increased symptom severity. Wet season and concurrent respiratory infections were independently associated with higher VAS, consistent with their etiology of nasal inflammation, even when air pollution levels were relatively low. 8 , 9 , 10 , 22 , 23 Additionally, younger age was associated with higher VAS, which may be related to more frequent respiratory infections in early childhood. Weeks requiring rescue medication were also associated with higher VAS, explained by the escalation of pharmacologic treatment in the presence of uncontrolled symptoms. Notably, PM2.5 exposure remained significantly associated with higher symptoms after adjusting for these factors. Our analyses also showed that medication use was associated with higher VAS scores but should be interpreted cautiously. Medication, particularly rescue treatment, was often initiated or escalated in response to greater symptom severity, creating potential confounding by indication. As medication use and symptoms were assessed within the same weekly period, their temporal sequence was unclear, and reverse causality cannot be excluded. Thus, this positive association likely reflects greater disease activity and treatment escalation rather than an adverse effect of medication itself. Despite adjustment for medication use as a time‐varying covariate, residual confounding by disease severity may remain. In exploratory subgroup analyses, elevated PM2.5 and respiratory infection were each associated with higher VAS scores in both the AR and NAR subgroups. Although the point estimate for PM2.5 was slightly larger in the NAR subgroup, this difference might not be statistically or clinically significant, and the study was not designed to formally compare the subgroups. Medication patterns also differed between groups, with rescue medication use significantly associated with higher VAS scores in the NAR group. This may reflect the episodic nature of NAR, where symptoms are often triggered by environmental irritants, leading to intermittent escalation of treatment. In contrast, AR is generally managed with regular INCS use to maintain baseline symptom control and prevent exacerbations from known allergen triggers. The smaller NAR sample further limited the statistical power for subgroup comparisons. As such, these findings are hypothesis‐generating rather than confirmatory and should not be interpreted as indicating that PM2.5 effects differ by rhinitis subtype. In the threshold analysis, adjusted VAS was higher during weeks when PM2.5 exceeded 37.5 μg/m3 than when it was below 15 μg/m3, consistent with the exposure–response relationship in the main model. However, these cut‐points should be interpreted cautiously because the WHO Air Quality Guideline (15 μg/m3) and Interim Target 3 (37.5 μg/m3) are defined as 24‐h means, whereas our exposure metric was a 1‐week average, which may smooth day‐to‐day variation and short‐term peaks. These cut‐offs were therefore used as exploratory benchmarks rather than indicators of guideline exceedance. They were selected as widely recognized reference concentrations that also frame Thailand's current and prospective national air‐quality standards, providing clinically and policy‐relevant anchors. A data‐driven quartile‐based analysis showed a similar exposure‐response pattern, with significantly higher adjusted VAS at higher exposures, suggesting that findings were not dependent on the specific cut‐points chosen. Notably, Thailand's national air‐quality control policy currently aligns with the WHO Interim Target 3 (≤37.5 μg/m3) and is moving toward a target of <25 μg/m3. 1 Our findings suggest that merely avoiding extreme pollution episodes may be insufficient to reduce the health burden in susceptible children. Achieving 24‐h average PM2.5 concentrations below 15 μg/m3 appears to be important for protecting long‐term pediatric respiratory health. For clinical implications, the findings of this study suggest that ambient PM2.5 exposure was associated with increased rhinitis symptoms in children. These results highlight the importance of considering air quality as a potential environmental factor in the routine management of pediatric rhinitis, particularly during periods of elevated PM2.5 concentrations. Clinicians may consider implementing proactive strategies, including advising guardians, coordinating with schools to reduce outdoor activities, and promoting the use of physical protective barriers. In addition, preventing respiratory infections and medication adjustment may also reduce rhinitis symptoms. At the public health level, these findings provide pediatric‐specific evidence to support stricter PM2.5 control that aligns with the global literature. The strengths of this study include its prospective design and repeated assessments, which allowed us to observe temporal changes in symptom severity. Ambient PM2.5 exposure was estimated by using IDW interpolation based on data obtained from government‐operated air‐monitoring stations, improving the spatial estimation of exposure. TWA exposure equations were also used to estimate PM2.5 exposure based on both residence and school locations, providing a more precise estimate of participants' 24‐h average exposure. Sensitivity analyses could further support the consistency of the results. However, several limitations should be acknowledged. As an observational study, our findings reflect associations rather than causal relationships. Concurrent medication use may have been influenced by symptom severity, resulting in confounding by indication. PM2.5 exposure was estimated using ambient monitoring stations rather than personal monitoring, potentially introducing exposure misclassification. Residual confounding from indoor PM2.5, air purifier use, and individual outdoor activity may also remain. AR and NAR group analyses were exploratory, with the smaller NAR sample limiting precision; thus, subgroup findings are hypothesis‐generating and should be interpreted cautiously. Finally, recruitment from a single metropolitan region may limit generalizability. Larger multicenter studies incorporating personal exposure assessment would be beneficial in future research. CONCLUSION Short‐term increases in ambient PM2.5 exposure were independently associated with higher weekly rhinitis symptom scores among children with chronic rhinitis. These findings support minimizing exposure during periods of poor air quality and emphasize the need for prospective interventional studies to determine whether reducing PM2.5 exposure improves clinical outcomes. AUTHOR CONTRIBUTIONS Rinrada Naka: Conceptualization; methodology; data curation; writing – original draft; writing – review and editing; formal analysis; project administration; funding acquisition; validation; investigation. Ratchaneewan Sinitkul: Conceptualization; methodology; data curation; investigation; formal analysis; writing – review and editing. Ussanai Nithirochananont: Data curation; investigation; formal analysis. Watcharoot Kanchongkittiphon: Data curation; investigation. Potjanee Kiewngam: Data curation; investigation. Wanlapa Jotikasthira: Data curation; investigation. Adithep Sawatchai: Data curation; investigation. Chaiyawat Suppasilp: Visualization; formal analysis. Wiparat Manuyakorn: Conceptualization; methodology; supervision; data curation; investigation; formal analysis; writing – review and editing. FUNDING INFORMATION This study was supported by the Research Fund, Faculty of Medicine Ramathibodi Hospital, Mahidol University, Bangkok, Thailand (RF_67097). CONFLICT OF INTEREST STATEMENT All authors declare no conflicts of interest. DECLARATION OF AI USE The authors declare that Artificial Intelligence (AI) tools (ChatGPT‐3.5 and Google Gemini‐3.6 Flash) were used in the preparation of this manuscript only for language editing, to improve readability, and to assist in the design and generation of the graphical abstract based on the study abstract and reported findings. The authors took full responsibility for the content, interpretation, and conclusions of the work. Supporting information Data S1. ACKNOWLEDGMENTS This study was supported by the Research Fund, Faculty of Medicine Ramathibodi Hospital, Mahidol University. The authors thank the Pollution Control Department and the Bangkok Metropolitan Administration, Thailand, for providing ambient PM2.5 data. Naka R, Sinitkul R, Nithirochananont U, et al. Impact of particulate matter 2.5 exposure on severity of pediatric chronic rhinitis: A prospective cohort study. Pediatr Allergy Immunol. 2026;37:e70493. doi: 10.1111/pai.70493 Associate Editor: Ayobami Akenroye DATA AVAILABILITY STATEMENT The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. REFERENCES Associated Data Supplementary Materials Data S1. Data Availability Statement The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.