There is no evidence of an inverse relationship between TH2-mediated atopy and TH1-mediated autoimmune disorders:: Lack of support for the hygiene hypothesis

Background: The findings of an inverse relationship between T(H)1- and T(H)2-mediated disorders would provide strong empiric support to the hygiene hypothesis. Objective: We sought to investigate the relationship between T(H)2-mediated atopic allergy and T(H)1-mediated autoimmune conditions in a nationally representative population. Methods: We used logistic regression to analyze adult data from the Third National Health and Nutrition Examination Survey. Data on allergic and autoimmune disease history were available for 20,050 subjects, and data on atopy were available for a subsample of 7304 subjects. Atopy was defined by one or more positive skin prick test responses (wheal of greater than or equal to3 mm) to 10 common aeroallergens. Allergic disease was defined by patient reports of physician-diagnosed asthma, hay fever, or both. T(H)1-mediated autoimmune disease was defined by patient reports of physician-diagnosed type 1 diabetes mellitus, thyroid disorders, and/or rheumatoid arthritis. Results: Adjusted for age and sex and taking into account the complex survey design, there was no relationship between atopy and a history of autoimmune disorders (adjusted odds ratio, 1.01; 95% CI, 0.61-1.67; P = .97). In contrast, physician diagnosis of allergic disorders was associated with a significant increased risk of physician-diagnosed autoimmune disorders (adjusted odds ratio, 1.67; 95% CI, 1.35-2.07; P < .001). Conclusion: We found no evidence of an inverse relationship between atopy and patient reports of physician-diagnosed common autoimmune disorders in the adult American population. Contrary to our initial hypothesis, reports of physician-diagnosed common allergic disorders are positively associated with reports of physician-diagnosed autoimmune disorders, with this possibly being caused by ascertainment bias. These findings suggest that the T(H)1/T(H)2 paradigm might be an oversimplification.