To assess whether children at risk for celiac disease should be screened systematically by comparing their baseline and follow-up characteristics to patients detected because of clinical suspicion.

Five hundred four children with celiac disease were divided into screen-detected (n = 145) and clinically detected cohorts (n = 359). The groups were compared for clinical, serologic, and histologic characteristics and laboratory values. Follow-up data regarding adherence and response to gluten-free diet were compared. Subgroup analyses were made between asymptomatic and symptomatic screen-detected patients.

Of screen-detected patients, 51.8% had symptoms at diagnosis, although these were milder than in clinically detected children (P < .001). Anemia (7.1% vs 22.9%, P < .001) and poor growth (15.7% vs 36.9%, P < .001) were more common, and hemoglobin (126 g/l vs 124 g/l, P = .008) and albumin (41.0 g/l vs 38.0 g/l, P = .016) were lower in clinically detected patients. There were no differences in serology or histology between the groups. Screen-detected children had better dietary adherence (91.2% vs 83.2%, P = .047). The groups showed equal clinical response (97.5% vs 96.2%, P = .766) to the gluten-free diet. In subgroup analysis among screen-detected children, asymptomatic patients were older than symptomatic (9.0  vs 5.8 years of age, P = .007), but the groups were comparable in other variables.

More than one-half of the screen-detected patients with celiac disease had symptoms unrecognized at diagnosis. The severity of histologic damage, antibody levels, dietary adherence, and response to treatment in screen-detected cases is comparable with those detected on a clinical basis. The results support active screening for celiac disease among at-risk children.