Associations of infection burden with Kawasaki disease in a population-based setting during 30 years

The objective of this study was to investigate the role of infections in the pathogenesis of Kawasaki disease.

The investigation was a nationwide epidemiological case-control study, comprising all cases of Kawasaki disease diagnosed in Sweden 1987-2018. Controls were randomly sampled from the general population, matched on sex, age, and area of residency. Data on infections were obtained from the Swedish National Patient Register, which prospectively collects data on all Swedish residents. Infections were classified by organ system, infectious agent and by temporal proximity to Kawasaki disease diagnosis date. Prescription of antibiotics and infections in family members were also considered in separate analyses.

The study comprised n=1774 (61% male) cases and n=17 731 controls. Overall, a history of infections was associated with Kawasaki disease with an OR of 2.3 (95% CI 2.0 to 2.5). Respiratory, skin, urogenital and gastrointestinal tract infections were all associated with Kawasaki disease. Temporal stratification revealed a prominent clustering of infections during the weeks before a Kawasaki diagnosis, but also higher frequencies of infections several months preceding Kawasaki disease with OR ranging from 5.1 (95% CI 3.6 to 7.1) 15-28 days to 1.3 (95% CI 1.1 to 1.6) 181-365 days prior Kawasaki disease. A dose-response relationship was observed, with repeated infections associating with higher ORs of Kawasaki.

The findings suggest that infections are closely linked with Kawasaki disease, and with a wider temporal association than previously known. Further, the data imply that many different agents may induce the disease.