Formula Type and Development of Type 1 Diabetes

William T. Basco, Jr., MD


December 10, 2010

Dietary Intervention in Infancy and Later Signs of Beta-Cell Autoimmunity

Knip M, Virtanen SM, Seppä K, et al; for the Finnish TRIGR (Trial to Reduce IDDM in the Genetically at Risk) Study Group
N Engl J Med. 2010;363:1900-1908

Study Summary

Knip and collaborators commented that 5 autoantibodies are associated with insulin-dependent (type 1) diabetes mellitus (IDDM), including antibodies to: islet cells, insulin, glutamic acid decarboxylase, insulinoma-associated 2 molecule, and zinc transporter. The question asked in this study was whether altering early food protein exposures could reduce the risk of developing some of these autoantibodies, theoretically reducing the risk for progression to IDDM. This study was a pilot study for a larger trial. Children were enrolled from 15 hospitals in Finland from 1995-1997. The investigators looked for newborns who had a first-degree relative with IDDM and identified 475 infants. HLA genotyping was performed to determine whether they appeared to be at risk for IDDM, and 230 (49%) had an eligible ("at risk") genotype and entered the randomized portion of the study. Of these 230 infants, 57% were boys, 37% had a mother with IDDM, 43% had a father with IDDM, and 15% had a sibling with IDDM. Four percent of the newborns had more than 1 first-degree relative with IDDM. The investigators obtained quarterly serum antibody samples during the first year of life, at 18 and 24 months, and then at 3, 5, 7, and 10 years of age (up to 10 total samples), and 208 (90%) of the children had serum drawn during at least 1 of the follow-up visits. Children were followed for a mean of 7.5 years (median 10 years).

The newborns were randomly assigned to be fed (when formula was introduced) with either the study formula, an extensively hydrolyzed, casein-based formula (n = 113), or the control formula, a formula with 20% of the protein hydrolyzed and 80% intact casein (n = 117). The formulas were not distinguishable by smell or taste. The investigators did not alter the maternal diets. The study was designed for the infants to use the study or control formulas for the first 6 months of life (if they were not breastfed) or in supplementation to breast milk if needed or desired. Infants of mothers who wanted to breastfeed were fed breast milk only for as long as the mother wanted to exclusively breastfeed. However, when formula was added to the infant's diet, it was the formula to which the child was assigned. If the mom breastfed exclusively until age 6 months, then the infant was allowed to use the study formula through 8 months. The investigators also expected the families to limit other foods in the infants' diets, especially proteins. The primary outcome was development of autoantibodies, but the investigators also followed the children to determine how many developed IDDM.

The demographics of the 2 groups were similar at enrollment. The hydrolyzed formula group had formula introduced to their diets later than the control group (2.6 months in the hydrolyzed formula group vs 1.1 months in the control group). In a similar manner, the intervention infants completed their intervention time at a median age of 7.4 months compared with 6.4 months for the control infants, a difference that did not reach statistical significance. However, the intervention infants (hydrolyzed formula group) received feedings of study formula for a median of 3.3 months compared with 4.9 months for control infants, and this difference was statistically significant. Development of at least 1 autoantibody was more frequent in the control group (30%) compared with the intervention group (17%); the frequency of developing 2 or more autoantibodies was 16% in the control group and 8% in the intervention group. Antibodies to islet cells and to insulin were by far the most common. Even after controlling for the difference in exposure time to study formulas, the intervention infants had lower hazard ratios for developing at least 1 autoantibody. Sixteen total children (7% of all participants) developed IDDM during the study and follow-up periods -- 6% of the intervention group and 8% of the control group -- but this difference was not statistically significant (hazard ratio 0.80, 95% confidence interval 0.30-2.14), even after adjusting for time exposed to study formula. The investigators concluded that dietary intervention, specifically hydrolyzed formula and delayed introduction of other foods, is associated with reduced risk for development of markers of beta-cell autoimmunity that might lead to type 1 diabetes.


The investigators emphasized that this pilot study was not powered to detect a difference in development of IDDM, but a larger, ongoing multinational study will be able to answer that question. In the meantime, this study does provide some compelling evidence that the immune response can potentially be altered with early dietary intervention. The data add evidence to the biologic plausibility that interactions between exposure to dietary proteins early in life and genetic predisposition can lead to IDDM in some individuals. As pointed out in an accompanying editorial,[1] the longer median exposure to study formula in the control group is concerning and introduces a bias that may not be fully eliminated by statistical correction. However, while waiting on the results of the full study, having genetically at-risk infants use hydrolyzed formula after nursing for an extended period seems to be a benign and potentially beneficial intervention.



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.
Post as: