Home / Resources / Articles / Timing is Everything: A Focus on Gluten Initiation into Infants’ Diets

Timing is Everything: A Focus on Gluten Initiation into Infants’ Diets

Feb 10, 2018

A study examining the timing of foods and its associated risk to islet autoimmunity.

It is no secret that genetics and diet play an integral role in the onset of diabetes. As infants get older, additional nutrients are required that are not attainable from breastmilk.  The needed nutrients for growth and development are found in what is known as complementary foods. Previous studies have evaluated different infant feeding practices in populations at a higher risk for type 1 diabetes (T1D) or islet autoimmunity (IA). Islet autoimmunity in T1D is typically identified by insulitis, islet cell antibodies, activated b-cell-linked T lymphocytes and class II major histocompatibility complex (MHC) alleles.  The activated T lymphocytes proceed to attack pancreatic b-cells, leading to T1D.2  It has been reported that shorter breastfeeding durations, the timing of introductory infant rice/oat and gluten-containing cereals, cow’s milk and other whole foods may all play a role in an increased risk of IA and/or T1D. The reported results have varied due to late follow-up in older children, making the link between timing of food introduction and IA questionable. Current studies aim to analyze key mechanisms in this association, including immature and adverse immunological responses of the gut to foods, mucosal inflammation and increased GI permeability.


The Environmental Determinants of Diabetes in the Young (TEDDY) study set out to evaluate duration of breastfeeding, timing of initiating formula, regular milk, and solid food in relation to IA risk. As a prospective observational cohort study, the main objective was to identify environmental causes of T1D. Data from three U.S. clinical research centers and three European clinical research centers was used. 7,563 HLA-eligible infants, with at least 9 months of data about breastfeeding duration and timing of introductory complementary foods, were analyzed for this study.  IA status was assessed through blood samples collected every 3 months from 3-48 months of age, and every 6 months thereafter. IA was determined through the use of radiobinding assays to examine the presence of insulin autoantibodies (IAA), GAD antibodies (GAD-A), GAD65 or insulinoma antigen-2- autoantibody. Seroconversion or IA status was based on the date of the first collected positive sample, confirmed by an additional positive sample. Data was also stratified by different IA statuses, children with IAA alone or GADAs alone. The median ages for any IA, IAA alone or GADA alone were 33 months, 21 months, and 46 months, respectively.  Questionnaires and diaries, also known as the TEDDY Book, were used to obtain population characteristics, diet and health monitoring. Infant dietary exposures were assessed every 3 months up until 5 years of age. Data was documented as unknown if breastfeeding went beyond 5 years of age, complementary foods were not initiated by 24 months or timing could not be recalled. Breastfeeding was documented as exclusive or supplemented by other foods.  Infant formula that was studied was any milk foreign to the mother’s milk. Solid foods that were examined, included but were not limited to any type of cereal, rice, potatoes, root vegetables, fruits and berries, meets, eggs, and seafood. Cox proportional hazards models, martingale residuals and a change-point method were all used to assess the association between each dietary exposures and risk of IA.

Later integration of gluten containing foods showed an increased risk of any IA (1.05; 95% CI: 1.01-1.10; P=0.02) and IAA alone (1.08; 95%CI:1.00-1.16; P=0.04). When assessing the duration of breastfeeding, alone or supplemented with formula or solids, association could not be linked with risk of IA. The martingale residual analysis showed there was an increased risk of IA between 0-4 months of age and from ³9 months of age when introducing gluten-containing cereals. The change-point method separated duration of breastfeeding and age of introductory solid foods (foreign milk and solids).  When comparing the age of introductory solids at ages 4-9 months, giving gluten-containing cereals before 4 months of age showed a decreased risk of any IA (HR 0.68; 0.47-0.99); and an increased risk of any IA after 9 months of age (HR 1.57; 1.07-2.31) (P=0.04). Results also showed a plateau from 4-9 months of age.

In conclusion, the TEDDY study suggests that later introduction to gluten-containing cereals show an association with an increased risk of IA. Additionally, there was a strong association between what type of first solids were given and IA risk. In the U.S., cereals tend to be the first solids, while in European countries fruits and vegetables are preferred. The use of pre- and probiotics in the first year of life has become a widespread practice and can account for the positive response of the gut to foods.  Strengths of this study included consistency of reporting across countries, generalizability of the study, and the use of continuous exposures. Cumulative exposure to a new food was not reported, signifying a weakness in the study. Future studies should highlight the timing and portions of solid foods given.

Practice Pearls:

  • Incorporating gluten-containing cereals before 4 months of age is associated with a decreased risk of IA in infants.
  • Gluten is a factor in the activation of islet autoimmunity in type 1 diabetes.
  • Duration of breastfeeding with or without supplementation is not a significant determinant of IA risk.


Uusitalo, Ulla, et al. “Early Infant Diet and Islet Autoimmunity in the TEDDY Study.” Diabetes Care, 17 Jan. 2018, pp. 1–9., doi:10.2337/dc17-1983.

Boitard, Christian. “Pancreatic Islet Autoimmunity.” La Presse Medicale, vol. 41, no. 12, Dec. 2012, doi:10.1016/j.lpm.2012.10.003.

Adrianna Jackson, Doctor of Pharmacy Candidate: Class of 2018; LECOM College of Pharmacy