Screening for coeliac disease: the meaning of low titers of anti-gliadin antibodies (AGA) in non-coeliac children.

Coeliac disease is diagnosed by means of jejunal biopsy, an invasive procedure. Anti-gliadin antibodies (AGA) have therefore been used in the first screening of the disease. On the other hand, low titers of AGA are widely detected also in normal subjects. In order to investigate if low levels of AGA could be correlated with laboratory and clinical data, we performed a study on 167 subjects with various illnesses, such as recurrent abdominal pain, failure to thrive, short stature, diarrhoea or constipation, cow-milk protein intolerance and/or food allergy, recurrent vomiting or previous gastroenteritis, all non coeliac conditions which have been associated with AGA presence. Seventy coeliac children, all biopsied, were selected as a control group. Among the 167 cases we found 60 subjects positive for AGA (35.9%), a high proportion as compared with the general population. Only 33/167 patients, all IgG and IgA AGA positive, fulfil our laboratory and clinical criteria to perform a 'confirming' biopsy. For the 134 residual cases (14 IgA, 13 only IgG AGA positive, 107 AGA negative) a diagnosis of coeliac disease has been excluded by clinical criteria (scoring). As a whole, the patients with coeliac disease had significantly higher levels of AGA of both IgG and IgA classes (p < 0.01). On the other hand, no significant difference emerged for all the anamnestic and laboratory parameters considered between AGA+ and AGA- non-coeliac subjects. However, laboratory parameters of IgG-AGA and/or IgA-AGA positive patients were similar to those of coeliac children for ion, Xylose, total IgA count. As no biopsied case showed mucosal atrophy, it is suggested that the presence of even low AGA levels in non-coeliac children may represent a highly sensitive index of intestinal alteration causing an increased permeability to macromolecules, but it is very unlikely that one could detect coeliac children by means of Ig-AGA among such illnesses and normal subjects. Strong clinical diagnosis and laboratory parameters are required to justify intestinal biopsies. In fact, the production of AGA seems to be a merely immunological phenomenon linked to an increased and probably transient permeability to macromolecules of the intestinal mucosa.