Rethinking strategies to screen for celiac disease.

Celiac disease (CD) is likely an ancient disease. It is thought that CD became a health problem as people moved from a hunter-gatherer economy to farming-based economy some time during the Neolithic period and geographically in the Fertile Crescent. Indeed, the earliest description of CD is attributed to Aretaeus the Cappadocian. Cappaddocia is adjacent to the Fertile Crescent. Recent studies on the introduction of farming, however, demonstrate a complex coexistence of hunter-gatherers and farmers over 2 millennia, and this new observation may affect how we understand the history of CD. CD is not a benign disease; it can be associated with malnutrition, malabsorption, osteoporosis, liver disease, skin disease, mental health issues, and other health problems. In addition, and of particular concern, undiagnosed CD is associated with a fourfold increased risk of death. CD is not a rare disease; prevalence ranges from 0.3% to 3%. These observations make it imperative that a diagnosis of CD be definitively made, that consideration be given to screening tools and the populations to whom they should be applied, and that parameters are carefully defined to correctly interpret the results of the screen. Currently there are at least 3 guidelines, 2 written for children and 1 for adults, on how to evaluate for and treat CD. These guidelines recommend screening based on family history, health status, and symptoms and rely on the immunoglobulin A (IgA) antibody for transglutaminase (tTG). For IgA-deficient people, the immunoglobulin G tTG is often used. All of the guidelines focus on symptoms as the factor used to initiate screening for CD, unless a first-degree relative has CD or the patient suffers with one of the diseases closely associated with CD, such as type 1 diabetes or Down syndrome, among others. The large and carefully designed study by Rosen et al in this issue calls into question the validity of using symptoms to identify children for screening. The authors invited 10 041 children and their parents to participate in a population-based CD screening of 12-year-olds. Children known to have CD were excluded. Before learning their CD marker results, children completed a questionnaire on symptoms in their classrooms, supervised by a teacher or nurse. Parents reported on the child’s CD-associated conditions. Serology was then obtained. Children underwent a small-bowel biopsy if their tTG-IgA or tTGimmunoglobulin levels were elevated. Surprisingly, the authors found there was no correlation between the presence of symptoms and the diagnosis of CD; symptoms did not discriminate unrecognized CD children from their non-CD peers. In addition, there was no difference in CD prevalence between those with and without CD-associated conditions. These results call into question the recommendations of the guidelines. Importantly, they raise concern for the guideline that recommends a gluten-free diet be initiated without a biopsy when “signs or symptoms suggestive of CD and high anti-TG2 titers with levels .10 times ULN” are present. AUTHOR: Susan S. Baker, MD, PhD