Allergic Diseases and Risk of Incident Autoimmune Diseases: Phenotype‐Specific Patterns and Multimorbidity Effects in a Nationwide Cohort

Summary

• Allergic rhinitis, asthma and atopic dermatitis show distinct, phenotype‐specific patterns of incident autoimmune disease risk.

• Asthma showed the strongest associations; multimorbidity marked a higher autoimmune burden.

To the Editor,

The classical Th1–Th2 paradigm has long cast allergic and autoimmune diseases as immunological opposites, yet this dichotomy is increasingly difficult to sustain. Accumulating evidence points to shared epithelial‐barrier dysfunction, chronic tissue inflammation and impaired immune regulation [1, 2, 3], and several population‐based studies have linked individual allergic diagnoses to a range of autoimmune outcomes [4, 5, 6]. However, most evidence derives from single‐exposure or single‐outcome designs, leaving open whether the breadth and magnitude of these associations differ across allergic phenotypes and whether allergic multimorbidity confers greater autoimmune risk.

We addressed these gaps using the Korean National Health Insurance Service–National Health Screening Cohort, evaluating 269,541 adults aged ≥ 40 years screened in 2004–2005 and followed from 2006 to 2013. Allergic rhinitis, asthma and atopic dermatitis were identified during 2002–2005; seven incident autoimmune diseases, including rheumatoid arthritis, autoimmune thyroid disease, ankylosing spondylitis, inflammatory bowel disease and systemic lupus erythematosus, were ascertained during follow‐up. Cox proportional hazards models were fitted; the fully adjusted model, prespecified as primary, adjusted for demographic, metabolic, lifestyle and comorbidity factors, with Benjamini–Hochberg false‐discovery‐rate correction across 28 exposure–outcome pairs. Prespecified sensitivity analyses addressed healthcare utilisation (surveillance bias), reverse causation (3‐year washout), competing risk of death (Fine–Gray), exposure updating, stricter case definitions and non‐proportional hazards (time‐split Cox); stratified analyses by sex, age, smoking, alcohol consumption and physical activity were additionally performed. Full methods, ICD‐10 codes and complete results (including supplementary tables and figures) are deposited in the Zenodo repository (https://doi.org/10.5281/zenodo.20047418). The study was approved by the Institutional Review Board of Korea University Guro Hospital (No. 2024GR0390). No written consent has been obtained from the patients as there is no patient‐identifiable data included.

Among 269,541 participants, 105,437 (39.1%) had at least one allergic disease. In the primary fully adjusted model, allergic disease was associated with higher risks of six autoimmune outcomes—rheumatoid arthritis, Hashimoto's thyroiditis, Graves' disease, ankylosing spondylitis, ulcerative colitis and systemic lupus erythematosus—but not Crohn's disease (Figure 1A). Three patterns were notable. First, risk profiles differed by allergic phenotype: allergic rhinitis showed the broadest associations; asthma showed the largest estimates, especially for ankylosing spondylitis (HR 1.36, 95% CI 1.20–1.53) and ulcerative colitis (HR 1.45, 95% CI 1.25–1.68); and atopic dermatitis was more selectively associated with rheumatoid arthritis, Hashimoto's thyroiditis and ulcerative colitis. Second, allergic multimorbidity generally showed the highest cumulative incidence for most autoimmune outcomes (Figure 1B). Third, time‐split Cox models showed that asthma– and atopic dermatitis–Hashimoto's thyroiditis associations were restricted to the first 4 years and attenuated thereafter (Figure 1C).

FIGURE 1.

Associations between allergic diseases and incident autoimmune diseases. (A) Fully adjusted hazard ratios (HR, 95% CI) from Cox proportional hazards models for four allergic‐disease exposures across seven autoimmune outcomes (28 pairs). Filled dark circles, p < 0.05 and FDR q < 0.05; filled light circles, p < 0.05 but FDR q ≥ 0.05; hollow grey circles, p ≥ 0.05. CI, confidence interval; SLE, systemic lupus erythematosus. (B) Kaplan–Meier cumulative incidence of three autoimmune outcomes across five mutually exclusive allergic‐disease groups. Shaded areas, 95% confidence intervals. (C) Time‐split Cox estimates for three exposure–outcome pairs with non‐proportional hazards (exposure × log[follow‐up time] interaction, p < 0.05). Blue circles, 0–4 years; orange squares, 4–8 years; filled markers, p < 0.05; faded markers, p ≥ 0.05.

Across additional analyses, the overall direction of association was largely preserved, with rheumatoid arthritis and ulcerative colitis emerging as the most reproducible outcomes. Adjustment for healthcare utilisation attenuated several associations, particularly for thyroid disease, and only the asthma–rheumatoid arthritis and asthma–ulcerative colitis associations remained statistically significant. These findings suggest that differential ascertainment may explain part of the observed excess risk and support cautious interpretation of the claims‐based associations. Complete additional analyses are available in the Zenodo repository (https://doi.org/10.5281/zenodo.20047418).

The principal strength of this study is its integrated comparative approach, in which three common allergic diagnoses and seven autoimmune outcomes were assessed in parallel under formal multiple‐testing control and an extensive prespecified sensitivity‐analysis programme, enabling phenotype‐specific allergy–autoimmunity patterns to be examined as the central question. Limitations include reliance on claims‐based exposure and outcome definitions, absence of laboratory data and asthma‐endotype information, descriptive cross‐phenotype comparisons because diagnosis‐specific models were not mutually adjusted, and restriction to adults aged ≥ 40 years, which may limit generalisability. In summary, allergic diseases in midlife and older Korean adults were associated with increased incidence of several autoimmune diseases, with phenotype‐specific patterns and greater cumulative autoimmune burden among individuals with allergic multimorbidity. These findings support heightened clinical vigilance for autoimmune comorbidity, particularly in adults with asthma or multiple allergic conditions, and provide a more nuanced framework for interpreting the allergy–autoimmunity link beyond the classical Th1–Th2 paradigm.

Author Contributions

H.L. conceived and supervised the study, had full access to all data, and serves as guarantor. K.M.K. and J.M.L. contributed equally as first authors. K.M.K., J.M.L., S.H.K. and S.J. acquired and analysed the data; K.M.K., J.M.L. and S.J. performed the statistical analysis. K.M.K. and J.M.L. drafted the manuscript; all authors critically revised the manuscript for important intellectual content and approved the final version. The corresponding author attests that all listed authors meet the authorship criteria.

Funding

This work was partly supported by the Institute of Information & Communications Technology Planning & Evaluation(IITP)‐ICAN (ICT Challenge and Advanced Network of HRD) grant funded by the Korea government (Ministry of Science and ICT) (IITP‐2026‐RS‐2022‐00156439) and the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (RS‐2026‐25477844).

Ethics Statement

The use of anonymized personal data for research is permitted under the current National Health Insurance Act, and the study protocol was reviewed and approved by the Institutional Review Board of Korea University Guro Hospital (No. 2024GR0390). No written consent has been obtained from the patients as there is no patient‐identifiable data included.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

This study and its reporting follow the STROBE guideline for cohort studies. The completed STROBE checklist, full Methods, ICD‐10 code lists, statistical code and supplementary results are openly available in the Zenodo repository (https://doi.org/10.5281/zenodo.20047418). Individual‐level NHIS‐HEALS data are not publicly available; qualified researchers may apply for access at https://nhiss.nhis.or.kr, subject to institutional review board approval and a data‐use agreement.