An 8 year old girl is referred to the paediatrician because of her stunted growth. Her height SDS is −2.3 (≈1st centile); her weight for height ratio is +0.9 SDS (≈80th centile). At the age of 3 years, her height SDS was +0.3 (≈60th centile). Her weight for height ratio had not changed considerably. At the age of 4–5 years, she had several episodes of constipation and anaemia. There were no other complaints. A diagnostic work-up was performed: IgA anti-endomysium (IgA EmA) antibodies were positive. A small intestinal biopsy (gold standard) showed total villous atrophy, consistent with coeliac disease. During follow up, she fulfilled the ESPGAN criteria for coeliac disease (finding of characteristic small bowel mucosa abnormalities in a small bowel biopsy, and clinical remission when placed on a gluten-free diet).

Structured clinical question

What is the prevalence of coeliac disease [outcome] in children with short stature and no gastrointestinal symptoms [patients]?

Search strategy and outcome

Pubmed—(body height AND (short OR little OR small OR abnormal) OR short stature OR dwarfism) OR failure to thrive AND coeliac disease (limited by: Ages: All child 0–18 year, Language: English);120 references of which 11 relevant and of sufficient quality (see table 1).

Embase—same search strategy; no additional relevant references.

Cochrane database—same search strategy; none relevant.

The way study groups were selected varied. Articles with the least selective study groups are reported higher in the table than studies with a more selective study group.

Commentary

Growth retardation in childhood may be one of the earliest signs of an underlying disease, such as coeliac disease. In the Netherlands, the growth of nearly every child is monitored. When growth is retarded, the child is referred to secondary health care. After referral it has been advised to perform a diagnostic work-up containing routine laboratory tests to search for diagnostic clues for, among others, coeliac disease. The tests presently used for coeliac disease are IgA EmA and IgA antitissue transglutaminase antibodies. The total immunoglobulin A count is determined as well, because coeliac disease is associated with IgA deficiency. It was questioned if diagnostic investigations for coeliac disease should be performed in all children with short stature, even without gastrointestinal complaints.

Studies 1–5 were based on study groups, in which no preliminary (endocrine) work-up to exclude other causes for short stature had been performed. The proportion of coeliac disease in children with short stature and no gastrointestinal symptoms in these studies ranged from 1.7% to 8.3%. When a group of children was studied, in which endocrine causes for short stature had been excluded (studies 6–11), the proportion of coeliac disease increased to a range of 18.6–59.1%. The characteristics of the preliminary work-up used in study 12 were not described.

The wide range of these percentages is probably mainly caused by the different methods of selecting the patients. The true variation in prevalence of coeliac disease throughout the world appears to be limited.¹³

Screening for coeliac disease in the general population shows a prevalence of 1:300 to 1:100. About 50% of these children are completely symptomless.¹³ In two British population based studies on short stature,^14,15 where coeliac disease was not specifically investigated, the prevalence of coeliac disease was 2:180 (one patient was already known with coeliac disease) and 0:149 respectively. In children with short stature and no gastrointestinal symptoms investigated for coeliac disease, the prevalence increases to 2–8%. When other (endocrine) causes for short stature are excluded, the prevalence might rise to even 59%.