Introduction
Atopic dermatitis (AD) is a chronic relapsing inflammatory skin disease with peak incidence in the first 2 years of life 1. It affects approximately 8.4 million children (<18 years) in the US , with 1 in 3 AD cases exhibiting moderate to severe forms of disease2. Moreover, children with severe AD exhibit lower quality of life 3 and higher risk of comorbid chronic health conditions, such as asthma and allergic rhinitis, compared to mild AD, in addition to a two-fold higher healthcare utilization4. Early childhood caries (ECC), a chronic diet-mediated oral infectious disease, affecting approximately 12.4 million children (<18 years) 5, is ranked fifth amongst the top 30 global health conditions in terms of incidence rates 6. Although two large-scale cross-sectional studies in 0-17 year old American (N=79,667) 4 and 6-15 year old Japanese children (N=21,792) 7 reported no association between allergic diseases (including AD) and dental caries, our previous study using a longitudinal study design revealed a potential link with higher ECC risk at 2- and 3-years in infants with atopic dermatitis 8. However, validity of the diagnostic criteria employed previously and effect of clinical severity on the AD-ECC link remains unknown.
Large-scale population studies often employ caregiver-reported 1-year history of a doctor-diagnosed AD as a proxy measure to estimate AD prevalence in children 9, which has been reported to render low sensitivity (70%) in identifying AD in 0-17 year old children 10. As more than two-thirds of infants with moderate-to-severe forms of AD demonstrate allergic sensitization in the first 2 years of life 11, our previous study supplemented the caregiver-reported history with a positive skin prick test (SPT) for identification of AD 8. Nevertheless, use of SPT in AD diagnosis has not been validated yet. Furthermore, children with severe AD have been associated with poor oral hygiene/dietary control possibly due to a distraction by symptoms and management of AD 12; however the regression model testing the previous AD-ECC link did not account for these behavioural factors. Hence, the current study aimed to investigate if clinical severity of AD by 3 years is associated with increased ECC risk among 3-year old children with/without control of potential dietary/behavioural confounders. Additionally, diagnostic accuracy of caregiver-reported history of doctor-diagnosed AD, with/without SPT, was evaluated using physician diagnosis as the gold standard.