Celiac Disease: Which Children Should Be Tested?

Lara C. Pullen, PhD

March 23, 2015

Serological screening studies indicate that celiac disease (CD) has a prevalence of 1% to 2% in Western populations and that the incidence is increasing across all age groups. Although many individuals develop symptomatic CD, others do not. In fact, researchers estimate that just 10% to 30% of patients with CD are ever diagnosed, in part because they may be asymptomatic.

For children, that lack of diagnosis can be critical: Young children with undiagnosed CD are particularly vulnerable to the effects of malabsorption and failure to thrive.

Because of that danger, some researchers and clinicians are pushing for widespread screening. Some experts go even further, suggesting that the entire pediatric population should be routinely screened for CD. Yet the cost and the need for invasive confirmatory biopsies leave others skeptical of that approach unless better, cheaper diagnostics become available

New data published in the March issue of the American Journal of Gastroenterology by Rok Seon Choung, MD, PhD, from the Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, Minnesota, and colleagues, show that among a representative sample of nearly 15,000 Americans aged 6 years or older, the prevalence of confirmed CD was 0.8% (95% confidence interval [CI], 0.5% - 1.0%) in 2009 to 2012.

Extrapolating to the population at large, the data suggest that more than 1.5 million people in the United States have CD. Moreover, the analysis showed a near doubling in the prevalence among adults aged 50 years or older, rising from 0.17% (95% CI, 0.03 - 0.33%) in the 1988 to 1994 National Health and Nutrition Examination Survey to 0.44% (95% CI, 0.24% 0 0.81%) in the 2009 to 2012 survey. However, the adjusted prevalence is uneven among racial and ethnic groups, ranging from 1.0% of whites to 0.2% of blacks and 0.3% of Hispanics.

For Ritu Verma, MBChB, director of the Center for Celiac Disease and the Lustgarten Endowed Chair for Clinical Care of GI Motility Disorders at the Children's Hospital of Philadelphia in Pennsylvania, those numbers and the consequences of undiagnosed CD lead to a straightforward conclusion: "The biggest plea I have for pediatricians is to just do the panel," she told Medscape Medical News, referring to an antibody test panel that can help to determine whether a child likely has the disease.

However, other pediatric gastroenterologists, including Saeed Mohammad, MD, from Northwestern University Feinberg School of Medicine in Chicago, Illinois, feel that screening the entire population is just not feasible, and the cost of increased screening may not be justified.

"[CD] is more common in certain ethnic groups and less common in others," he told Medscape Medical News. "Unless the cost of genetic testing drastically decreases, I believe the current system of screening children with risk factors for [CD] such as diabetes, Down syndrome, or autoimmune diseases, as well as testing for celiac-specific antibodies in children whose growth is faltering [and] those who have diarrhea or chronic abdominal pain, is the most cost-effective option with the greatest sensitivity."

Moreover, he notes that children younger than 2 years should not be tested without high clinical suspicion, "and [diagnosis] should always be confirmed with endoscopy and biopsy," he continued.

When and How to Test?

Classic symptoms of CD include prolonged abdominal pain, poor growth, chronic diarrhea, and abdominal distention. When children present with these symptoms, physicians may think of testing for CD, but unfortunately, there is no standardized celiac panel.

The Children's Hospital of Philadelphia, where Dr Verma practices, recommends a screening panel that includes total immunoglobulin A (IgA), antitransglutaminase (anti-tTG), and antiendomysial. Deaminated gliadin can be ordered in conjunction with anti-tTG. In addition, for children younger than 3 years, it may be advisable to test for antigliadin, she says.

Dr Verma would prefer that all children be screened for CD, but if that is not possible, then she encourages pediatricians to screen when they have even the slightest suspicion that the child might have CD.

She notes that many children who are diagnosed with CD in her clinic do not have a classic presentation. Thus, Dr Verma actively educates primary care physicians about CD and encourages them to consider the possibility of CD if they are treating a patient who does not seem to get better. The primary care physician can then order the simple celiac blood test.

Meanwhile, Boston Children's Hospital in Massachusetts focuses their screening efforts on high-risk children. These would be children with classic symptoms of CD such as abdominal pain and chronic diarrhea, as well as those at known genetic risk for CD and children with type 1 diabetes.

Boston Children's Hospital uses a similar celiac panel to the one used by the Children's Hospital of Philadelphia. "We recommend that 'high-risk' children be screened with serologic celiac tests (specifically tissue transglutaminase IgA and a total IgA), even if they have no identifiable symptoms," Dascha C. Weir, MD, associate director of the Celiac Program at Boston Children's Hospital in Massachusetts, told Medscape Medical News.

A child who is strongly positive for all of these antibody tests likely has CD.

Dr Verma acknowledges that the tests can give false-negatives and provide a false sense of security to parents. They can also give a false-positive. This is why the definitive test for CD is upper endoscopy.

Dr Verma explained that physicians also need to take a dietary history. Some children do not eat a lot of gluten between the ages of 2 and 3 years, and therefore may not test positive on the celiac panel. Unfortunately, most pediatricians do not have the time to take a complete dietary history.

Moreover, the United States has a heterogeneous population, and some experts suggest that a one-size-fits-all screening approach may not be practical. That said, "care providers should have a low threshold to order celiac testing for children with gastrointestinal symptoms or nongastrointestinal symptoms," Dr Weir emphasized.

Classic Symptoms Inadequate Cues

Regrettably, many primary care physicians have the false impression that children who lack classic symptoms of CD should not be screened for the disease. Unfortunately, relying on symptoms to trigger CD testing could miss a substantial proportion of affected children, according to recently published data from Sweden.

Daniel Agardh, MD, PhD, from Lund University in Sweden, and colleagues published their subanalysis of data from The Environmental Determinants of Diabetes in the Young (TEDDY) study online March 2 in Pediatrics. Whereas the primary goal of TEDDY is to identify genetic and environmental factors that contribute to the development of type 1 diabetes, Dr Agardh and colleagues used the birth cohort data to compare children who were positive for anti-tTGA with those who were negative for anti-tTGA.

They found that, compared with children who were negative for anti-tTGA, children positive for anti-tTGA were more likely to be symptomatic at 2 and 3 years of age, but not at 4 years of age. In addition, the researchers report that anti-tTGA levels correlate with severity of mucosal lesion in symptomatic and asymptomatic children, and some asymptomatic children had quite elevated levels of anti-tTGA. These children were diagnosed with CD because they were enrolled in the study, but likely would have been missed in standard practice.

Results from the study have led the experts to propose a two-tiered screening system by which all newborns are first specifically screened for genetic susceptibility such as HLA type and, if positive, then screened for celiac-specific antibodies. This screening protocol would be a departure from that used in the United States and would include much more comprehensive genetic screening.

"With awareness and availability of gluten-free foods increasingly entrenched within the mainstream of the North American lifestyle, the burden lies on the identification of all children who may benefit from treatment. The prospective data from TEDDY effectively demonstrate the utility of 2-tiered screening and constitute a step forward in devising a population-screening strategy that best offers the appropriate treatment at a stage in life where it may yield the most lifelong benefit," writes Richard J. Noel, MD, PhD, from Duke University Medical Center in Durham, North Carolina, in an accompanying commentary in Pediatrics.

Dr Verma believes the study from Sweden advances the conversation, but does not take the concept of screening far enough. "I am glad that at least someone is thinking about screening people early," Dr Verma said. "[but] if you are between 2 and 3 years of age, you should just have a celiac panel done." She emphasized again that there are many atypical symptoms of CD, especially in children.

Genetic Risk Factors

The 2-tiered screening test may not yet be realistic for the United States.

Currently, children in the United States may be targeted for screening because they have a family history of CD or have Down syndrome, Turner syndrome, or Williams syndrome. The 2004 consensus guidelines from the National Institutes of Health indicate that patients with type 1 diabetes should also be screened for CD.

Although individuals who are homozygotic for HLA-DR3-DQ2 are at highest genetic risk for CD and tend to develop CD very early in life, HLA tests are not routinely performed in the United States, and therefore HLA status rarely triggers a CD test. According to Dr Weir, this is an evolving issue.

"Screening high-risk groups (such as family members or patients with predisposing conditions) for [CD] with serologic testing (tissue transglutaminase [immunoglobulin G and total immunoglobulin A] is widely accepted. However, in the USA, the use of HLA typing in screening children to identify who is at higher risk of [CD] has not been fully evaluated and is not currently recommended. However, as our understanding of the HLA typing and [CD] risk deepens, and as HLA testing becomes cheaper and more readily available, this may change," Dr Weir noted.

The genetic basis of CD does mean that a diagnosis of CD has implications for the entire family. If a child tests positive for CD, the family (siblings, parents, and grandparents) should be tested for CD. In addition, siblings should be periodically retested to look for the development of CD.

Poor Growth May Point to CD

Another newly proposed screening strategy suggests children should be tested for CD if they have poor growth. Although CD does not always affect growth, most children who are diagnosed experience a decrease or slowing of growth before diagnosis. A new study suggests these children could be diagnosed earlier if they were being followed in a well-established growth-monitoring program.

Antti Saari, MD, from the University of Eastern Finland in Kuopio, and colleagues published the results of their longitudinal retrospective study on the relationship between growth and CD online March 2 and in the March issue of JAMA Pediatrics. They propose that screening for CD would include the use of several growth-monitoring parameters in combination. Furthermore, the parameters should be integrated into a computerized screening algorithm associated with an electronic health record system.

"I have sort of mixed feelings about [the study].... I am personally of the belief that there should be mass screening of people," stated Dr Verma. She explained that because not all of the children with CD have growth issues, the approach proposed by Dr Saari and colleagues would miss many children with CD.

Atypical Symptoms

Dr Verma described some of the atypical symptoms of CD. For example, anemia may be a sign of CD. As a consequence, Dr Verma recommends that children with unexplained anemia be tested for CD.

She added that chronic headaches, seizures, chronic constipation, elevated liver enzymes, chronic pancreatitis, alopecia, and tiredness can all be signs and symptoms of CD. Moreover, many patients who are originally diagnosed with irritable bowel syndrome are ultimately diagnosed with CD.

Parents should be counseled to have the blood work done before experimenting with a gluten-free diet. A gluten-free diet will eventually lower the levels of antibodies, thereby making the celiac panel unreliable. Moreover, physicians should keep in mind that children who consume low levels of gluten or consume occasional gluten may have lower antibody levels, even though they have CD.

A patient who tests positive on the celiac panel should be referred to a gastroenterologist. The gastroenterologist can decide whether endoscopy is required and can counsel the patient on the importance of a gluten-free diet. Patients must understand that the gluten-free diet is for life, and they need encouragement to adopt a gluten-free lifestyle.

The authors, Dr Verma, Dr Weir, and Dr Mohammed have disclosed no relevant financial relationships.

Am J Gastroenterol. 2015;110:455-461. Choung abstract

Pediatrics. Published online March 2, 1015. Agardh full text, Commentary full text

JAMA Pediatr. Published online March 2, 2015. Abstract


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.