This is the first study demonstrating that children with CD may have patchy villous atrophy of the duodenum. The bulb mucosa may be the only duodenal area involved, both at diagnosis and after gluten challenge. Therefore, multiple endoscopic biopsies should always be performed, not only in the distal duodenum, but also in the bulb.

On the basis of a carefully conducted study, CD prevalence in our series was seen as very high. These data suggest an accurate algorithm to select candidates for intestinal biopsy among CD high-risk subjects. First, an evaluation of the sensitive RIA TGAA and of total IgA (in IgA deficiency RIA IgG anti-tissue transglutaminase assay) should be performed. Then, an evaluation of the TGAA and the genetic study would be advisable 2 to 3 years later in negative subjects. Those carrying the DQ2/DQ8 heterodimers should continue the serologic follow-up; the others need a clinical follow-up.

This study demonstrated that CD-related histological lesions are always present in the bulb; sometimes this specific site is the only one affected. Therefore, we suggest taking 2 biopsies from the bulb and 2 from the distal duodenum for CD diagnosis.

Our aim was to evaluate body composition in a group of coeliac disease adolescents on a gluten-free diet and to re-examine them at the end of the adolescence spurt. We studied 48 patients (group 1A), 30 age-matched healthy controls (group 2A), 11 group 1A patients after 4 years (group 1B) and 11 adolescents who were age- and sex-matched with group 1B (group 2B). Weight, height, bone mineral content, fat mass, fat-free mass (FFM) and bone mineral density were evaluated using dual-energy X-ray absorptiometry. All parameters were lower in group 1A than in group 2A subjects ( p<0.001). After 4 years, the body compartments of group 1B coeliac disease patients normalised, except for weight and FFM which remained lower than in group 2B subjects ( p<0.005). In conclusion, our study demonstrates that adolescence is a period where some parameters of body composition can still be recovered.

Based on the positivity of culture or urease rapid test and histology, infection was detected in 47 patients. The statistical analysis showed that, for the detection of the infection, stool antigen assay is more effective in sensitivity and negative predictive value (91.5% and 96.5%), whereas specificity and positive predictive values appear slightly better in serology (89.6% and 76.0%). Correlations between serum IgG both with patients' age (r = 0.21, p < .05) and H. pylori stool antigen (r = 0.47, p < .01) were found. The search for bacterial DNA on oral samples proved to be very specific (99.1% on saliva and 98.2% on plaque), but insensitive (22.2% and 25.7%). CONCLUSIONS. In children H. pylori stool antigen represents a sensitive test, suitable for detecting H. pylori infection. Serum IgG proved to be more specific; the PCR on the oral cavity resulted as being a very specific, but insensitive test.