Cluster Analysis of Age-Dependent Allergen Sensitization and Allergic Disease Association To the Editor,Allergen sensitization patterns are closely linked to the prevalence and severity of allergic diseases across different age groups.1,2 In children, these patterns shift significantly with development, making it crucial to identify distinct sensitization profiles and their associations with allergic diseases at various developmental stages.1,3 While longitudinal studies have explored these associations in birth cohorts, real-world data on these relationships remain limited.2,4Consequently, studies leveraging real-world data are necessary to enhance generalizability and deepen our understanding of these associations in diverse populations.2,5 By identifying specific sensitization clusters, we can determine high-risk groups for allergic disorders, implement preventive strategies, and develop individualized treatment approaches based on endotypes.3,4 In this study, we investigated allergen sensitization patterns across age groups and their relationship with allergic diseases using a single-center common data model (CDM).This single-center retrospective observational cohort study utilized the Seoul National University Hospital (SNUH) CDM, formatted according to the Observational Medical Outcomes Partnership CDM. Patients who underwent allergy testing, including skin prick tests (SPT), multiple allergen simultaneous tests (MAST), and allergen-specific Immunoglobulin E (IgE) tests, between January 2002 and December 2021 were included. For SPT, a reaction was considered positive if the mean wheal diameter was ≥3 mm or exceeded the histamine control. For MAST, allergens classified as class 3 or higher were considered positive. For specific IgE tests, a value of ≥0.35 KU/L was considered positive.If a patient underwent multiple allergy tests, a positive sensitization result was recorded if at least one test was positive. Conversely, a patient was considered nonreactive only if all tests yielded negative results. If test results were unavailable, they were treated as missing data (NA values).Patients were categorized into five age groups: Group 1: <3 years, Group 2: 3–5 years, Group 3: 6–11 years, Group 4: 12–19 years, and Group 5: 20–29 years. Only patients who underwent allergy testing in at least two different age groups were included in the study. Hospital visit codes were retrieved and categorized for allergic disease classification (Supplementary Table). Allergy test results for each age group were integrated into 17 allergen categories.Cluster analysis was performed using latent class analysis (LCA),6,7 with the optimal number of clusters determined using the Gower distance. The prevalence of allergic diseases across age groups was compared. Continuous variables were expressed as means and standard deviations, and categorical variables were expressed as frequencies and percentages. Categorical variables were compared across age groups, conditions, and clusters using Pearson’s chi-square test.8 Statistical analyses were conducted using R version 4.3.3, and the gLCA package was used for LCA.The study included 1,656 patients, with a median age of 5 years (interquartile range [IQR], 2–11 years) at the first test. Among these patients, 65.9% were male. A total of 3,402 allergy tests were conducted at a median age of 7 years (IQR, 4–14 years). The types of allergy tests administered included SPT (16.2% of patients, 21% of the total tests), MAST (70% of patients, 56.6% of total tests), and specific IgE testing (13.8% of patients, 22.4% of total tests). The most frequently diagnosed allergic disease was rhinitis (47.2% of patients, 54.4% of total tests), followed by respiratory diseases (41.2% of patients, 39.4% of total tests). Eczema (24.2% of patients, 24% of total tests) and asthma (20% of patients, 23.5% of total tests) were also frequently diagnosed (Table 1).Allergy test results were analyzed to classify patients into four sensitization clusters (Figure 1). Cluster 1 had a high rate of sensitization to house dust mites (HDM). Cluster 2 was sensitized to common food allergens, including eggs, grains, legumes, meat, milk, and shellfish. Cluster 3 exhibited high sensitivity to a broad range of allergens. Finally, Cluster 4 had a high rate of sensitization to multiple aeroallergens, including HDM, mold, animal dander, trees, and weed allergens (Figure 1).The trajectory of allergic sensitization clusters across age groups 1–5 years was analyzed. The early infancy sensitization pattern was predominantly represented by Cluster 2, mostly sensitive to food allergens (44.7%), followed by Cluster 3, which exhibited polysensitization, and Cluster 4, which showed sensitization to aeroallergens, both appearing at similar proportions. In age Group 2, (3–5 years), Cluster 1 became the most dominant at 32.7%, followed by Cluster 4 at 31.7%, which was primarily associated with inhalant allergens, particularly HDM. In age groups 3 to 5, Cluster 1 remained the most dominant cluster, with sensitization rates of 53.9%, 55.4%, and 41.8%, respectively, HDM as the primary allergen (Supplementary Figure 1).The prevalence of allergic diseases, including asthma, eczema, food allergies, and rhinitis, was analyzed for each allergen sensitization cluster across age groups 1–5 years. Analysis of patients diagnosed with asthma showed that Cluster 1, which primarily exhibited sensitivity to HDM, was predominant among all age groups. In age Group 1, 32.4% of patients in Cluster 1 were diagnosed with asthma. This percentage increased to 44.6% in age Group 2, 46.7% in age Group 3, and then declined to 28.8% in age Group 4. In age Group 5, 22.9% of patients in Cluster 1 were diagnosed with asthma. In the case of eczema, Cluster 2, which exhibited high sensitivity to food allergens, was generally the most dominant in all age groups. Clusters 2 and 3 were the most prevalent clusters in eczema, indicating that polysensitized individuals could be at higher risk for multiple allergic diseases. Food allergy prevalence was highest in Cluster 3, which was characterized by polysensitization, in age Group 1, with 33 of 111 patients (29.7%) affected, followed by Cluster 2, which exhibited sensitivity to food allergens, with 26 of 203 patients (12.8%) affected. Similar patterns of cluster dominance were observed in age Groups 2 and 3. Clusters 1 and 4 demonstrated a higher prevalence of allergic rhinitis than other clusters, suggesting that patients with respiratory diseases were more likely to be sensitive to aeroallergens, as shown in Figure 2.This study demonstrated that allergen sensitization patterns evolve as children grow older, shifting from food allergens in infancy to aeroallergens in later childhood and adulthood. Distinct sensitization patterns were associated with specific allergic diseases, aiding in disease prediction and guiding treatment approaches. Understanding early-life sensitization patterns may help identify high-risk groups, enabling early interventions and targeted prevention strategies. Identifying distinct sensitization clusters supports more personalized and endotype-driven treatment approaches. Recognizing age-related changes in sensitization patterns is essential for long-term allergic disease management.However, this study has limitations. Conducted at a single center, its findings may not be generalizable to the broader Korean population. Including only patients who underwent allergy testing may have introduced selection bias toward individuals at higher risk for allergies. The retrospective study design limited causal inference and control over confounding factors. Although multiple age groups were analyzed, this is not true longitudinal study. Additionally, environmental factors such as allergen exposure, living conditions, and dietary habits may have influenced sensitization patterns but were not fully accounted for.In conclusion, this study provides valuable insights into allergen sensitization patterns and their relationship with allergic diseases across different age groups. These findings highlight the dynamic nature of allergic sensitization, reinforcing the importance of early identification, prevention, and tailored management strategies in clinical practice.Keywords : Age-dependent, allergic diseases, allergen, cluster analysis, sensitization