Almost all people (95%) with coeliac disease have either the variant HLA-DQ2 allele or (less commonly) the HLA-DQ8 allele. However, about 20–30% of people without coeliac disease have also inherited either of these alleles. This suggests additional factors are needed for coeliac disease to develop; that is, the predisposing HLA risk allele is necessary but not sufficient to develop coeliac disease. Furthermore, around 5% of those people who do develop coeliac disease do not have typical HLA-DQ2 or HLA-DQ8 alleles
There are several tests that can be used to assist in diagnosis. The level of symptoms may determine the order of the tests, but all tests lose their usefulness if the person is already eating a gluten-free diet. Intestinal damage begins to heal within weeks of gluten being removed from the diet, and antibody levels decline over months. For those who have already started on a gluten-free diet, it may be necessary to perform a rechallenge with some gluten-containing food in one meal a day over 6 weeks before repeating the investigations
Due to its high sensitivity, serology has been proposed as a screening measure, because the presence of antibodies would detect previously undiagnosed cases of coeliac disease and prevent its complications in those people. There is significant debate as to the benefits of screening. Some studies suggest that early detection would decrease the risk of osteoporosis and anaemia. In contrast, a cohort study in Cambridge suggested that people with undetected coeliac disease had a beneficial risk profile for cardiovascular disease (less overweight, lower cholesterol levels)
Since 2002, Finland's Social Insurance Institution had paid monthly compensation to people with newly diagnosed celiac disease for the additional cost of maintaining a gluten-free diet. To receive this compensation, people must submit proof of diagnosis, including biopsy findings, along with diagnostic criteria, in a statement from a physician. That information is kept in a national database. The researchers used the database to calculate incidence and prevalence rates of celiac disease through 2006.