By Ghazaleh Barghgir, MD, and Nancy Selfridge, MD

Dr. Barghgir is a Clinical Skills Facilitator, Clinical Foundations Department, Ross University School of Medicine, Barbados, West Indies

Dr. Selfridge is Professor, Clinical Foundations Department, Ross University School of Medicine, Barbados, West Indies

Dr. Barghgir and Dr. Selfridge report no financial relationships relevant to this field of study.

SYNOPSIS: Data analysis from The Environmental Determinants of Diabetes in the Young (TEDDY) study, to evaluate risk of celiac autoimmunity and celiac disease in children who screened positive for at-risk human leukocyte antigen genotypes, demonstrated increased risk for both outcomes in genetically predisposed children correlating with increasing quantities of daily gluten intake during the first five years of life.

SOURCE: Aronsson CA, Lee H-S, Segerstad EMHA, et al. Association of gluten intake during the first 5 years of life with incidence of celiac disease autoimmunity and celiac disease among children at increased risk. JAMA 2019;322:514-523.


Increased intake of gluten during the first five years of life was associated with significantly increased risk of celiac disease autoimmunity and celiac disease in genetically predisposed children.

Incidence was most notable between 2 and 3 years of age.

Risk increased for every 1 g/day incremental increase in gluten intake over an average gluten reference intake determined at 2 years of age.

Post-hoc analysis of data suggested a cut-off intake of 2 g/day or less at age 2 years to reduce the risk of celiac autoimmunity and celiac disease.

Celiac disease is a gastrointestinal autoimmune disease
that affects 1% of the world population. Often, it manifests in early childhood. The pathophysiology of celiac disease autoimmunity involves both genetic and environmental factors. Individuals who are positive for certain inherited cell surface human leukocyte antigen (HLA) types (primarily HLA class II genes: DR3 in haplotypic association with DQ2 and DR4 in haplotypic association with DQ8) are at an increased risk of developing immunoglobulin A (IgA) autoantibodies to gliadin and tissue transglutaminase (tTG) upon exposure to dietary gluten, which in turn may attack intestinal and dermal epithelium.1

Symptoms associated with celiac disease are characteristic of malabsorption syndrome due to destruction of intestinal villi: diarrhea, bloating, iron-deficiency anemia, weight loss, malnutrition, and failure to thrive in young children. Celiac disease also is associated with a long-term risk of T-cell and non-Hodgkin lymphomas and small intestinal adenocarcinoma. Once celiac disease is diagnosed, current management guidelines include a strictly gluten-free diet, which is helpful for symptom management for many patients. However, when the typical Western diet contains 5 g to 15 g of gluten per day and gluten is present in rye, wheat, and barley flours, this is a difficult recommendation to implement and maintain, especially for children.2

Further, can celiac disease risk be mitigated by delaying or limiting gluten exposure in a child’s diet? The European Society for Pediatric Gastroenterology recently updated guidelines for infant and childhood feeding practices to reduce the risk of gluten sensitivity.3 One of the aims of this study was to bring us closer to an understanding of the association between celiac autoimmunity/celiac disease related to the timing of introduction and quantity of gluten intake in genetically at-risk children toward refining dietary recommendations.

Aronsson et al analyzed data from The Environmental Determinants of Diabetes in the Young (TEDDY) study to identify whether increased gluten intake in the first five years of life was associated with an increased risk of celiac disease or celiac disease autoimmunity in at-risk children. TEDDY is a multinational, prospective cohort study following children from birth to age 15 years at six clinical research centers in Germany, Finland, Sweden, and the United States. Their primary goal is to identify genetic and environmental risk factors associated with type 1 diabetes, celiac disease, or both. The TEDDY researchers enrolled participants from September 2004 to February 2010 with gluten intake data available in September 2017. Study participants were enrolled before the age of 4 months and consisted of 8,676 (40%) out of 21,589 eligible children who tested positive as neonates for HLA DR and DQ antigens known to be associated with type 1 diabetes and celiac disease.

Prevailing reasons for not enrolling eligible HLA-positive children included parental concerns about the demanding protocol, repeated blood draws, or other family circumstances. Out of the original 8,676 enrollees, 6,605 participants were determined to have adequate screening for tissue transglutaminase antibodies and sufficient dietary records of gluten intake to be included in analysis. Gluten intake was estimated using three-day food records collected by participants’ parents at ages 6, 9, and 12 months and biannually at 18, 24, 30, and 36 months thereafter, until age 5 years. During the three-day food intake recordings, parents were encouraged to maintain their child’s normal food habits. Dietary intake was analyzed using food composition databases from each country and mean gluten intake in grams/day was calculated from the estimated total intake of gluten-containing flours in foods reported in the three-day food diary. Subsequently, the gluten intake was calculated and categorized as absolute daily intake, residual intake, and gluten intake per 10 kg of body weight at a given age. Residual intake calculations use regression analyses and are commonly applied in epidemiologic nutritional studies to determine nutrient intake gluten in this instance adjusted for total energy intake, so that any association with a measured disease outcome can be attributed more robustly to the nutrient alone.

Testing for tTG autoantibodies started at 24 months and continued yearly thereafter. The primary outcome was celiac disease autoimmunity, defined as positive tTG autoantibodies found in two consecutive serum samples. The secondary outcome was celiac disease confirmed by intestinal biopsy or persistently high tTG autoantibody levels > 100 U in an average of two consecutive samples. Joint modeling was used to assess the time-to-event relationship between gluten intake and the primary and secondary outcomes, a widely accepted statistical modeling tool that considers multiple covariates, random effects, and imputations for missing data to improve individual-specific predictions. This longitudinal modeling included adjustments for energy intake (kilocalories/day), HLA genotype, sex, country of residence, and immediate family history of celiac disease.

In addition, two Cox regression analyses, based on observed data only, were performed to assess risk of celiac autoimmunity and celiac disease that included the most recent gluten intake and energy intake prior to the outcome event as time-dependent covariates. One analysis included all children, and the other analysis included children with gluten intake data available within one year prior to the outcome event to control for various lag times between gluten exposure and the event. Risk assessment results were similar for both the joint modeling and Cox regression analyses.

Average gluten intake at age 2 years, when tTG assessment commenced, was considered the reference level for comparisons. Results indicated that the incidence of both outcomes peaked between 2-3 years of age. Daily gluten intake was associated with higher risk of celiac disease autoimmunity and increased for every 1 g/day increase in gluten consumption over the reference level (hazard ratio [HR], 1.30; 95% confidence interval [CI], 1.22-1.38). Findings based on mean absolute daily gluten intake at 3 years of age indicated an absolute risk of 28.1% that rose to 34.2% if gluten intake was 1 g/day higher than the reference amount, an absolute risk difference of 6.1%. Daily gluten intake also was associated with higher risk of celiac disease for every 1 g/day increase in gluten consumption (HR, 1.50; 95% CI, 1.35-1.66). Absolute risk by age 3 years was calculated at 20.7% if the mean absolute daily reference amount of gluten was consumed, and rose to 27.9% if gluten intake was 1 g/day higher than this reference amount, an absolute risk difference of 7.2%. Results for the various daily gluten intake calculation methods (absolute, residual, and intake per 10 kg body weight), including absolute risk differences and 95% confidence intervals, are summarized in Table 1. In a post-hoc analysis, the authors noted a gluten intake cutoff point of 2 g/day in children at 2 years of age, over which a higher risk of celiac disease autoimmunity and celiac disease was statistically apparent.

Table 1: Absolute Risk for Developing Celiac Disease Autoimmunity and Celiac Disease in Children With At-Risk HLA Genotypes


Absolute Risk for Developing Celiac Disease Autoimmunity by Age 3 Years

Absolute Risk for Developing Celiac Disease by Age 3 Years

Measurements of Gluten Intake

If Reference Amount of Gluten Was Consumed, %a

If Amount of Gluten Consumed Was 1 g/day Higher Than Reference Amount, %

If Reference Amount of Gluten Was Consumed, %a

If Amount of Gluten Consumed Was 1 g/d Higher Than Reference Amount, %

Absolute intake, g/day





Residual intake, g/dayb





Intake/10 kg of body weight





aAssessed in relation to average daily gluten intake at 2 years of age.

bAdjusted for age and energy intake using the residual method.


This study adds to a body of population data from recent published research concerning early childhood gluten intake for children at risk for celiac disease, focusing on the timing of introduction and quantity.3,4,5 For children with known at-risk HLA genotypes, 2 g of gluten intake per day appears to be a cut-off point, over which the risk of celiac autoimmunity and celiac disease is increased significantly.

Strengths include the prospective cohort design, a large multinational study population, and dietary monitoring and data recording that accounted for the variations in dietary content and composition characteristics of growing children. The study included a population of children from four different countries, offering a large genetic pool of at-risk children. It also employed a standardized process of accounting for variations in diet and the amount of gluten intake in these culturally diverse populations. Parents were not aware of their child’s antibody status and, thus, parental changes in their children’s dietary intake were minimized. The results not only were statistically significant, but also were clinically relevant as well. One gram of gluten, the increment associated with increased risk over the reference intake at age 2 years, is equivalent to about half a slice of white bread.

Nonetheless, the study had some limitations. Although the study design was created to minimize inaccuracies in dietary self-reporting, missing and inaccurate data on daily gluten intake is still possible with a self-report protocol. Despite the fact that this study gathered participants from four different countries, including four different U.S. locations, the study population might not be optimally representative of all at-risk groups. Including infants from other high prevalence countries such as Ireland, Australia, northern Africa, and northern India might be valuable, especially given the potential for varying infant feeding practices and exposure to processed cereal grains. Although an association between time and quantity of gluten intake and onset of celiac autoimmunity/disease was noted, other environmental factors may influence risk, including GI infections and rotavirus vaccination status. This is suggested by the finding that a subset of Swedish children appeared more prone to develop celiac disease in this study; although rotavirus vaccine was introduced in Sweden in 2014, it was not included in the Swedish national immunization program until 2019.

How best, then, to use these study results to counsel parents concerning infant and pediatric diet? Although early gluten intake is known to increase celiac autoimmunity/disease risk only in children with at-risk HLA genotypes, neonatal screening for at-risk HLA genotypes is not performed routinely, and physicians likely will not know their pediatric patient’s genetic risk status. This study supports some specific recommendations for optimum pediatric gluten intake to be added to the current guidelines,3 although the authors suggest that a randomized, placebo-controlled trial assessing varying amounts of gluten intake would be needed to confirm their findings and to establish gluten intake guidelines for patient counseling. Physicians can inform parents who express concern about their children’s risk of celiac disease that preliminary results from this study indicate gluten intake of 2 g/day or less (the equivalent of one full slice of white bread or one 150 g serving of pasta) seems to be associated with lower risk of developing celiac autoimmunity and celiac disease. This dietary limit can be applied reasonably when a child has one of the at-risk genotypes or has a family history of celiac disease or type 1 diabetes. There may be no increased risk for children with negative family histories and without at-risk HLA genotypes and, therefore, gluten restriction for this group cannot be recommended. In the meantime, we await further studies to support our clinical practices in this domain.


  1. Lebwohl B, Sanders DS, Green PHR. Coeliac disease. Lancet 2018;391:70-81.
  2. Hoppe C, Gøbel R, Kristensen M, et al. Intake and sources of gluten in 20- to 75-year-old Danish adults: A national dietary survey. Eur J Nutr 2017;56:107-117.
  3. Szajewska H, Shamir R, Mearin L, et al. Gluten introduction and the risk of coeliac disease: A position paper by the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition. J Pediatr Gastroenterol Nutr 2016;62:507-513.
  4. Marild K, Dong F, Lund-Blix NA, et al. Gluten intake and risk of celiac disease: Long-term follow-up of an at-risk birth cohort. Am J Gastroenterol 2019;114:1307-1314.
  5. Lund-Blix NA, Mårild K, Tapia G, et al. Gluten intake in early childhood and risk of celiac disease in childhood: A nationwide cohort study. Am J Gastroenterol 2019;114:1299-1306.