Objectives: The transglutaminase (TG) antibody test is accurate in identify
ing celiac disease in symptomatic children. We sought to determine the posi
tive predictive value of this test in asymptomatic children at genetic risk
for celiac disease.
Study design: Asymptomatic children with a genetic risk for celiac disease
were studied to investigate the relationships between TC antibody titer, sm
all bowel histology, growth, and clinical features. Small bowel biopsy hist
ology was graded by using the system of Marsh.
Results: Of 30 children with a positive TG antibody test result, 21 (70%) h
ad definite (Marsh score 2 or 3) and 4 (13%) had possible (Marsh score 1) b
iopsy evidence of celiac disease. TG antibody titer correlated with Marsh s
core (r = 0.569, P < .01). There was an inverse correlation between Marsh s
core and height z score (r = -0.361, P = .05).
Conclusions: In this group of asymptomatic children screened because of a g
enetic risk, TG antibodies have a positive predictive value of 70% to 83% f
or biopsy evidence of celiac disease and may identify children before clini
cal features of celiac disease develop.