A simple, rapid, highly reproducible enzyme-linked immunosorbent assay detecting anti-gliadin antibodies in serum to screen for coeliac disease (CD) is described. By combining the results of anti-gliadin IgA and IgG determinations the overall sensitivity of the assay was found to be 100% and the specificity 96% for children and 99% for adults. Significantly elevated antigliadin IgA and IgG antibodies were detected in all 20 children and all 25 adults with untreated CD. False positive results were found in 1/79 histologically normal control and 5/86 disease control children, while for adults false positive rates were 0/74 and 1/34 for the healthy and disease control groups, respectively. Anti-gliadin IgA and IgG was measured in serum samples from 52 coeliac patients (11 children and 41 adults) treated with a gluten-free diet (GFD). Each of the children and 28 of the adults who followed a strict GFD had significantly lower IgA and IgG levels than untreated CD patients. The serum anti-gliadin IgA and IgG levels of the 13 adults not complying with a GFD were similar to those found for untreated CD patients. This assay is recommended as a screening test for CD as well as a tool for follow-up of treated patients.

The differential diagnostic potential of serum gliadin-specific IgG subclass antibodies was assessed by comparing the antigliadin IgG1, 2,3,4 profile at different stages of coeliac disease with that of gastro-intestinal infection and also conditions associated with increased intestinal permeability. The IgG subclass profile of untreated coeliac disease was found to be the same as in healthy controls (IgG1 ≈ IgG2 > IgG3 > IgG4), with only the magnitude of the individual subclass responses being increased in coeliac patients. Coeliac adults and children on gluten-free diets had different antigliadin IgG subclass profiles with IgG2 being elevated. Increased intestinal permeability or recent gastro-intestinal infection did not alter the antigliadin subclass profile from that observed in healthy individuals. Assessment of the diagnostic potential of antigliadin IgA1 and IgG1–4 measurements in screening for coeliac disease demonstrated that measurement of subclasses of gliadin-specific IgA and IgG was less sensitive and specific compared with the combined use of total antigliadin IgA and IgG. Therefore it is suggested that IgG subclasses should not be used for routine screening for coeliac disease.

No abstract