Acquired susceptibility to autoimmune diseases in pediatric patients with Escherichia coli infection: A population-matched retrospective cohort study.

BACKGROUND

Escherichia coli (E.coli) infection has been proposed to play an important role as an initial trigger in the development of autoimmunity via molecular mimicry. However, there has been no preliminary cohort study to establish the association of E.coli infection with autoimmune diseases. Therefore, we conducted a large scale, population-matched cohort study to determine the risk of autoimmune disease among patients with exposure to E.coli.

METHODS

Utilizing the National Health Insurance Service database, we retrospectively analyzed a total of 259,875 Korean children that consisted of 23,625 exposed and 236,250 unexposed persons from January 1, 2002 to December 31, 2017. The exposed cohort was defined as patients diagnosed with E.coli infection. Unexposed controls were matched by birth year and sex at a 1:10 ratio for each exposed patient, using incidence density sampling. The primary outcome was autoimmune disease development. We used the Cox model to estimate the risks of autoimmune diseases among patients diagnosed with E.coli infection.

RESULTS

Over a mean follow-up of 10 years, there were 1455 autoimmune disease cases among exposed patients (incidence rate, 63.6 per 10,000 person-years) and 11,646 autoimmune disease cases among unexposed persons (incidence rate, 50.4 per 10,000 person-years), with an adjusted hazard ratio (HR) of 1.254 (95% CI 1.187-1.325). E.coli infection was associated with increased risks of autoimmune diseases; Reactive arthritis, HR 1.487, 95% CI 1.131-1.956; Henoch Schönlein purpura, HR 1.265, 95% CI 1.050-1.524; Systemic lupus erythematosus, HR 1.838, 95% CI 1.165-2.898; Sjögren's syndrome, HR 2.002, 95% CI 1.342-2.987; IgA nephropathy, HR 1.613, 95% CI 1.388-1.874. Kaplan-Meier cumulative incidence curves also showed a significant association between E.coli infection and incident autoimmune disease (p < 0.0001). This relationship was not only independent of demographic variables, but also remained consistent across various sensitivity analyses. On the other hand, patients with longer hospital stay for E.coli infection were at a higher risk of autoimmune disease (p = 0.0003), and the risk of autoimmune disease also tended to increase, as the frequency of E.coli infection was higher. Moreover, the relative risk of autoimmune disease seemed to be attenuated by use of antibiotics and a history of intestinal infectious disease, but elevated by coexistence of other autoimmune diseases.

CONCLUSIONS

Our cohort study indicates that E.coli infection was significantly associated with increased susceptibility to autoimmune diseases, even after adjusting for different factors. Thus, among environmental factors, a previous history of E.coli infection could be a predisposing risk factor in the development of autoimmune diseases.