Current issues of ACP Journal Club are published in Annals of Internal Medicine


Diagnosis

Gliadin antibody test had moderate sensitivity for celiac disease in adults

ACP J Club. 1994 Sept-Oct;121:51. doi:10.7326/ACPJC-1994-121-2-051


Source Citation

Bodé S, Gudmand-Høyer E. Evaluation of the gliadin antibody test for diagnosing coeliac disease. Scand J Gastroenterol. 1994 Feb;29:148-52.


Abstract

Objective

To evaluate the diagnostic value of gliadin antibody (GA) in patients with suspected celiac disease.

Design

Blinded comparison of serum IgA and IgG GA with small-intestinal biopsy in 2 populations: one with low and one with high incidence of celiac disease.

Setting

A hospital in Denmark.

Patients

100 consecutive patients (median age 51 y, 64% women) admitted for small-intestinal biopsy because of suspected celiac disease. The positive predictive value was also measured in 118 adults (median age 48 y, 72% women) and 55 children (median age 4 y, 56% girls) with increased or borderline GA.

Description of test and diagnostic standard

Serum IgA and IgG GA were determined in duplicate with diffusion in gel enzyme-linked immunosorbent assay. Increased (positive) GA was defined as IgA > 10.5 mm, IgG > 14 mm, or both. Borderline values were defined as 9.5 ≤ IgA ≤ 10.5 mm or 13 ≤ IgG ≤ 14 mm, or both. The diagnostic standard was a small-intestinal biopsy specimen with crypt hypertrophic villous atrophy and increased numbers of inflammatory cells.

Main outcome measures

Sensitivity, specificity, and positive and negative predictive values for IgA and IgG GA.

Main results

Among the 100 consecutive patients, 13 had untreated celiac disease, 35 had chronic diarrhea, 22 had dyspepsia or the irritable bowel syndrome, 6 had acute gastroenteritis, 5 had chronic pancreatitis, 3 had alcoholic cirrhosis, 3 had Crohn disease or ulcerative colitis, 7 had other gastrointestinal diseases, and 6 had nongastrointestinal diseases. Sensitivities and specificities are shown in the Table. Positive predictive value and negative predictive value for IgA were 75% and 92%; for IgG, values were 73% and 94%; and for IgA and IgG combined, values were 71% and 97%. In the second sample, 57 adults and 19 children had untreated celiac disease. The positive predictive value for adults was 90% and for children was 74%.

Conclusions

Gliadin antibody had 77% sensitivity as a noninvasive screening test for patients with suspected celiac disease. A small-intestinal biopsy was still necessary for confirming celiac disease in those patients with positive gliadin antibody tests.

Source of Funding: Danish Medical Research Council.

For article reprint: Dr. S. Bodé, Department of Medicine F, Gentofte Hospital, DK-2900 Hellerup, Denmark. FAX 45-397-77631.


Table. Test characterics of gliadin antibody (GA) for detecting celiac disease*

Test Sensitivity Specificity +LR -LR
Serum lgA GA > 10.5 mm 46% (19 to 75) 98% (92 to 100) 20 0.5
Serum IgG GA >14 mm 62% (32 to 86) 97% (90 to 99) 18 0.4
Combined serum lgA and lgC 77% (46 to 95) 95% (89 to 99) 17 0.2

*Other abbreviations defined in Glossary; LRs and CIs calculated from data in article.


Commentary

Noninvasive screening tests that are sensitive and specific would be a useful adjunct for the diagnosis of celiac disease. None of the currently available tests meets sufficiently stringent criteria for sensitivity, specificity, and positive and negative predictive values to warrant widespread use in screening populations at risk. The small-bowel biopsy remains the "gold standard" in establishing diagnosis. Serologic tests, including the IgA and IgG serum antigliadin antibody and the IgA antiendomysial antibody, may be more useful in measuring patient compliance with a gluten-free diet.

Screening tests work best in populations where disease prevalence is high. Among children, such populations include those with short stature (1), Down syndrome, and diabetes who fail to thrive (2). Additionally, patients with unexplained iron deficiency anemia, weight loss, osteomalacia, and fatigue, and asymptomatic relatives of patients with gluten-sensitive enteropathy-celiac disease (GSE-CD) may produce a higher diagnostic yield from screening tests (3). The prevalence of GSE-CD in most urban populations in the United States is between 1:800 and 1:10 000 but is virtually nonexistent in Chinese and African populations.

The study by Bodé and Gudmand-Høyer, using IgA and IgG antibodies, failed to diagnose celiac disease in 23% of patients. In view of the considerable morbidity and some mortality associated with the untreated condition and the high success rate and low morbidity of treatment, this false-negative rate remains unacceptably high.

James P. Keating, MD
St. Louis Children's HospitalSt. Louis, Missouri, USA


References

1. Knudtzon J, Fluge G, Aksnes L. Routine measurements of gluten antibodies in children of short stature. J Pediatr Gastroenterol Nutr. 1991;12:190-4.

2. Cacciari G, Salardi S, Volta U, et al. Prevalence and characteristics of coeliac disease in type 1 diabetes mellitus. Acta Paediatr Scand. 1987;76:671-2.

3. Corazza G, Valentini RA, Frisoni M, et al. Gliadin immune reactivity is associated with overt and latent enteropathy in relatives of celiac patients. Gastroenterology. 1992;103:1517-22.