Screening for coeliac disease in families of adults with Type 1 diabetes based on serological markers

Abstract

The prevalence of coeliac disease (CD) in the adult population is unknown because silent and latent stages do exist. Type 1 diabetes mellitus may be associated with CD because of common genetic background and/or shared pathogenetic mechanisms. We investigated 74 adults with type 1 diabetes (32+/-11 yr, disease duration 13+/-9 yr), 69 parents of diabetic probands (56+/-10 yr), 59 siblings (30+/-11 yr) and 50 healthy controls (35+/-10 yr) for the presence of circulating islet cell antibodies (ICA), anti-glutamic acid decarboxylase antibodies (GADA65), anti-gliadin immunoglobulins A and G (IgA- and IgG-AGA). All patients with raised AGA, performed also IgA anti-endomysium antibody (EmA) indirect immunofluorescence assay. Samples were positive for ICA in 19 diabetics (26%), 4 parents (6%), 4 siblings (7%), 0 controls (p<0.001); for GADA in 34 diabetics (46%), 4 parents (6%), 1 sibling (2%), 0 controls (p<0.001). Twenty-five diabetic patients (34%), 10 parents (14%), 5 siblings (8%), 3 controls (6%) (p<0.001) had raised IgA-AGA (>4.4 mg/l). Four diabetic patients (5%), 5 parents (7%), 0 siblings (0%), 4 controls (8%) had raised IgG-AGA (>18 mg/l). Both IgA- and IgG-AGA were detected in 1 diabetic and 2 parents. The prevalence of ICA, GADA, and IgA-AGA positivity in Type 1 diabetes patients was significantly higher than in controls (p<0.001). Finally, 50 AGA-positive subjects performed EmA test: only 2 of them resulted EmA-positive, a diabetic patient and a sibling. The patient with Type 1 diabetes had a small-bowel biopsy specimen consistent with CD and, as sole evidence of malabsorption, sideropenic anaemia. EmA-positive sibling also showed severe iron deficiency, yet refused endoscopy. We conclude that: 1) CD cannot be diagnosed on the basis of associated IgA- and IgG-AGA alone. Nevertheless, detection of such antibodies is useful, in combination with EmA, in screening for endoscopic biopsy; 2) too high rate of detection of IgA-AGA in Type 1 diabetic patients in comparison with other groups excludes a false positivity of the test itself, while suggests a pathogenetic association of both immunological disorders, perhaps related to abnormal gammadelta TCR-bearing intraepithelial lymphocytes.