Skip to content

The EAT Study: More food for thought on earlier introduction of solids to prevent food allergy

Would you offer your 3-month-old baby a taste of peanut butter? Hard-boiled egg? How about a little tilapia with tahini sauce? That’s exactly what parents did in a study published in May in The New England Journal of Medicine.1 The study found that early introduction to these foods might protect children from development of food allergies, but many parents had a difficult time following the study instructions.

The study, called the Enquiring About Tolerance (EAT) study, was conducted in the U.K. and led by Dr. Gideon Lack. Lack was also the lead author on the LEAP study published in 2015, which found that earlier introduction of peanut to high risk infants prevented peanut allergy.2 Previous observational studies have found a lower rate of peanut, egg, cow’s milk, and wheat allergy in children with earlier introduction to these foods.3–6 However, LEAP and EAT are the first published randomized controlled trials of this question, giving us higher quality evidence than the earlier observational studies.

In the EAT study, 1303 exclusively breastfed 3-month-old babies were randomly divided into two groups. One was instructed to follow current U.K. recommendations to continue with exclusive breastfeeding until 6 months of age and then to introduce solid foods, leaving the types and amounts of foods up to the parents. The other group was instructed to begin introducing solids as early as 3 months, including six commonly allergenic foods: cow’s milk yogurt, peanut, cooked egg, sesame (in the form of tahini paste), whitefish, and wheat. Mothers in the early solids group were also asked to continue breastfeeding through at least 6 months.

Here is what parents in the early introduction group were asked to do (from the study protocol):

-Begin with rice cereal mixed with breast milk or water. Average age of enrollment was actually 3.4 months, so babies were starting foods sometime between 3 and 4 months.

-Next, offer cow’s milk yogurt with a goal of 4 grams of protein per week by 5 months of age. This comes out to a heaping one-third of a cup of whole milk yogurt, fed over the course of the week.

-Next, offer sesame, egg, fish, peanut, and wheat. Each family was given a randomly assigned order of introduction to these foods. The target amount was 4 grams of protein from each food per week. I show how that translates to amounts of actual foods in the photo below. Parents were asked to not introduce wheat before 4 months of age, because previous research has found an association between celiac disease and earlier introduction of wheat.

An earlier paper on the EAT study provides an example of a weekly calendar [PDF] for food introduction given to parents.7 (Please, please, please… don’t get caught up in following this protocol without reading to the end of this post!)


I happened to have all of these foods on hand, so I dished them up so you can see what a baby in the early introduction group of the EAT study should have been eating by 5 months. (Again, this is a week’s worth of food.) If you think it sound challenging to incorporate all of these foods into a young infant’s diet, then you are not alone. One of the big limitations of the study was that many babies (and/or their parents) didn’t actually follow these dietary instructions.


So, did timing of introduction of these foods make a difference to the development of food allergy?

The researchers looked at differences in food allergy between the two groups in a couple of different ways. The first was an intention-to-treat analysis, comparing the incidence of food allergies regardless of whether or not subjects had actually followed the study instructions. Looked at this way, 5.6% of the early introduction group and 7.1% of the standard introduction group developed an allergy to at least one food. This difference was not significant.

Next, the researchers looked more closely at how well the families actually followed the dietary advice they were given. They defined “adequate adherence” in the early introduction group as consuming at least 3 grams of protein per week of at least 5 of the 6 assigned foods for at least 5 weeks sometime between 3 and 6 months of age. Just 43% of babies met these criteria.

But those babies, it turned out, did enjoy significant protection from food allergies. Only 2.5% of them developed a food allergy, compared to 7.3% of the standard group that stuck to their instructions. For peanut, there were no cases of allergy in the early introduction group compared with 2.5% in the standard group. For egg, 1.4% of the early introduction group was allergic compared with 5.5% of the standard group. These differences were all statistically significant.

Babies in the early introduction who ate less than the prescribed amounts also seemed to benefit, particularly for peanut. Those who ate just 1.5 teaspoons of peanut butter per week for 4 weeks (prior to 6 months of age) had a 10-fold reduction in the risk of developing peanut allergy compared to the standard group.

Food allergies are on the rise in the U.S., so the idea that we might be able to at least partially reverse this trend by giving babies earlier exposure to certain foods is intriguing. However, the fact that so many infants in the EAT study didn’t follow the early introduction protocol makes it hard to interpret the results with certainty. For example, there may have been factors that affected both the early acceptance of solids and the incidence of allergies. Although the researchers did not think this was the case, more studies with an easier-to-follow protocol (maybe just focusing on two to three foods instead of six, for example) would help clarify these relationships. One such study is currently underway in Norway.

There were no cases of anaphylaxis in the early introduction group, and hospitalization and growth rates were similar between the two groups. Parents of the early introduction babies were a bit more likely to report vomiting and constipation, but not diarrhea, at 4-6 months. These differences were small.

In a paper published separately earlier this year, the EAT researchers reported no difference in the rates of breastfeeding at 6 months in the control and early introduction groups (98% and 97%, respectively – much higher than typical rates in the U.K., but this was part of the EAT protocol).7 It’s important to note that since the EAT study specifically focused on breastfed babies, these findings may not translate to formula-fed infants.

The NEJM produced a short video summary of the study, available here.

My Take:

This was a careful, well-designed study of an important question facing every parent: when should we start offering solids to our babies? The results suggest that giving babies solids (particularly common allergens) as early as 3-4 months might cut the risk of food allergies. Previous studies suggest that breastfeeding at the time of solids introduction may be protective against allergies, and earlier solids introduction might mean that more babies experience that combo.

We’re seeing more and more studies that show benefits of earlier introduction of solids with little risk. This flies in the face of current recommendations from the WHO, the AAP’s Section on Breastfeeding, and the U.K.’s NHS that babies should be exclusively breastfed for six months. I hope that medical organizations are reevaluating that advice, at least in developed countries. (You can read more of my analysis of this question here.)

I find the EAT study really promising, but I also feel worried when I think of parents trying to follow this prescriptive food introduction protocol, especially as early as 3 months. It is an important finding that more than half of babies and their parents failed to follow it. You can’t rush a baby in starting solid foods – they need to have key motor and oral developmental skills in place. It is rare to see these before 4 months, and many babies won’t be ready until 6 months. For example, a baby needs to have lost the tongue thrust reflex, allowing her to move food to the back of her mouth and swallow it with ease. She also needs to be able to sit upright comfortably. And she needs to be interested in eating!

Trying to push a baby without the interest or the developmental skills to eat may backfire. The baby may reiterate her displeasure with this weird stuff called food by spitting it out and putting up a fuss, and further pressure to eat might just make that negative response even stronger. Meanwhile, parents may feel more stressed that they are somehow failing in the feeding department, and that’s the last thing any of us need.

What we can do is watch for signs of readiness and give our babies opportunities to try new foods, including these potentially allergenic ones, as they are ready. I started offering tastes of solids to my babies around 4.5-5 months, but neither ate more than a couple of bites per day until around 6 months. For my second baby, who benefited from my hours of reading on this topic, those bites did intentionally include all of the foods in the EAT study, except for tahini.

CORRECTION: An earlier version of this post incorrectly stated the food amounts shown in the photograph. The current amounts match what was recommended to parents participating in the EAT study.


  1. Perkin, M. R. et al. Randomized Trial of Introduction of Allergenic Foods in Breast-Fed Infants. N. Engl. J. Med. 374, 1733–1743 (2016).
  2. Du Toit, G. et al. Randomized Trial of Peanut Consumption in Infants at Risk for Peanut Allergy. N. Engl. J. Med. 0, null (2015).
  3. Poole, J. A. et al. Timing of Initial Exposure to Cereal Grains and the Risk of Wheat Allergy. Pediatrics 117, 2175–2182 (2006).
  4. Koplin, J. J. et al. Can early introduction of egg prevent egg allergy in infants? A population-based study. J. Allergy Clin. Immunol. 126, 807–813 (2010).
  5. Du Toit, G. et al. Early consumption of peanuts in infancy is associated with a low prevalence of peanut allergy. J. Allergy Clin. Immunol. 122, 984–991 (2008).
  6. Katz, Y. et al. Early exposure to cow’s milk protein is protective against IgE-mediated cow’s milk protein allergy. J. Allergy Clin. Immunol. 126, 77–82.e1 (2010).
  7. Perkin, M. R. et al. Enquiring About Tolerance (EAT) study: Feasibility of an early allergenic food introduction regimen. J. Allergy Clin. Immunol. 137, 1477–1486.e8 (2016).




  1. Interesting, thanks for sharing! Do you know of any studies that differentiate between breast- and formula-feeding in terms of these specific allergens? I imagine formula will be kept “clean” of potential allergens, while with breastmilk it may depend on the mother’s diet (though I don’t really know how much of the dietary protein makes it into breastmilk…)

    Liked by 1 person

    July 19, 2016
    • Great questions. There are studies that show that food proteins make it into breast milk, but the data on whether or not breastfeeding actually protects against food allergies are really mixed. As always with breastfeeding research, there are a lot of confounding factors! As I wrote in the post, some studies show that breastfeeding *at the time of food introduction* might be the key, but again, we need more research on that. Babies get exposed to food allergens in other ways as well. A recent study found allergens in amniotic fluid, for example, and they also are exposed in the environment (dust, small food particles). The EAT study actually measured the presence of peanut dust in children’s beds as a measure of peanut exposure. It’s pretty complex and very interesting stuff!

      Liked by 1 person

      July 19, 2016
  2. Ana #

    I think that for many families in South America this early food introduction is part of our culture. Many different types of whole grains, fruits, eggs, dairy products are introduced at early stages in a baby’s diet, as early as 3-4 months.


    July 19, 2016
    • Thanks for saying this! I think many people have a notion that delaying the introduction of solids is the more traditional route, but early solids feeding is the cultural norm in many parts of the world and again, that may be associated with lower incidence of food allergies. We definitely see this with peanut in many parts of Africa, for example, as well as Israel. It’s a common food to be bed in infancy, and the incidence of peanut allergy is very low.


      July 19, 2016
  3. I had this conversation with my daughter’s doctor 5 years ago when we were about to start solids…
    I asked her if she thought that the seemingly higher incidence of food allergies in this generation was correlated to the fact that we waited until the immune system was more developed to introduced these common allergy type foods than we used to.
    I mean when I was a baby, they started introducing cereals at like 6 weeks or even less.
    She said she thought it was an interesting theory, and since we 1) had no food allergies running rampant in our family, and 2) were part of a national 5 year Child Allergy STudy and under the care of a pediatric allergist, to go ahead and introduce whatever we felt comfortable with.
    Neither one of our kids have allergies. Probably a coincidence, but glad to see all this research being done. It’s a fascinating topic.


    July 19, 2016
  4. Elke #

    This is super interesting, especially now that my baby is 3 months and 20 days old and I’m trying to read a lot about it. I was almost ready to open the calendar and try it out (until I read your comment right after that saying not to do so 🙂 ). But thinking it through, I’m still not super convinced. They compare the adequate adherence group with the control group. So if the adequate adherence group is in any way less likely to develop the allergies, it would invalidate their findings, right? Lets say for instance that babies in the adequate adherence group present better motor and development skills, they are bigger and stronger. And bigger and stronger babies are less likely to develop allergies (Just an ex., I have no idea what causes allergies in babies), then the we would have a “treated” group with babies that are less likely to develop allergies (and biased results…?). I haven’t read the paper though, maybe they explained something about it?


    July 21, 2016
    • Yes, this is definitely a limitation, and I’m glad you brought it up, because I only briefly mentioned it in the post. This is why we need more studies with protocols that are easier to follow. I would also like to see a study that began introduction a little later (maybe 5 months) so that developmental readiness is less of a factor in adherence.

      The authors do discuss this in the paper. They mention that their findings in the adherent group could be an example of reverse causation. For example, if a baby in the early introduction group tried a new food and it caused some discomfort, perhaps an early sign of an allergy, then the parents might be less likely to continue offering it, or the baby might refuse to eat it. That baby would not have been classified as “adherent” in the statistical analysis, but the fact that this baby was more likely to have a food allergy would not have been explained by the lack of early exposure to those foods. Meanwhile, the adherent group would have a lower risk of allergies from the start, unrelated to the introduction of solids. The authors say they don’t think this is the case, because if it was, they would expect the non-adherent early introduction infants to show a greater incidence of allergy than the control group, and that wasn’t the case. But still, you’re absolutely right that there are alternative explanations here, so we need to be cautious.

      However, this study fits with the patterns we see in other research. The LEAP peanut study, in particular, showed a very clear protective effect (although only in high risk infants) and they didn’t have the problems with compliance (I think this is probably because they started later, and with only one food to focus on, it was easier for parents to follow the protocol). From what we know so far, it seems like we shouldn’t be afraid of introducing these foods as early as about 4 months, but you still have to wait for your baby to be developmentally ready and interested.

      Enjoy feeding your baby, whenever you start!


      July 21, 2016
      • Elke #

        I often feel overwhelmed with all the information, so thanks so much for addressing my questions.

        I see how this study aligns with the previous studies on early food introduction for “normal” babies. But I heard that some exclusive breastfed babies present at a very early age allergies from animal proteins, commonly cow milk but also peanuts and other foods (particularly frequent in my social group). How would the findings above fit to these cases? Or are there other findings more related to allergic babies? Is the timing for food introduction in exclusive breastfed babies that are allergic other than for non-allergic babies?

        Thanks again!


        July 21, 2016
  5. Jessica's blog #

    I love this blog post.:) I know a lot of people worry about this all the time. I am one of them, even though I don’t have my own children yet. My mother told me she started giving my sister and I solids at age 6 months. She told me we both wanted to always eat, so she would give us some of her food that she was eating. My father always told us to try something we haven’t tried, for 1. to see if we like it and 2. to see if we were allergic. Of course we did this always with precaution just in case we were allergic, but everything was fine. (if you wanted to know, we both aren’t allergic to any food)
    I would also do this method to my own children when the time comes. I’d want them to start eating young so they won’t hopefully get food allergies.
    Thank you for sharing the research 🙂


    July 26, 2016
  6. Marta #

    Hi Alice,

    Thanks for this post, very interesting. We have been spontaneously doing something similar with our 4.5 month old baby (except we don’t spoon feed at all, we just let him pick and suck on whatever he likes that we are eating). My question is: if he starts getting more calories from solids and decreases breast milk, should we give him also some water to drink? I remember reading somewhere else (on your blog?) that at some point there may be a difference of 100 calories worth of breast milk for babies who are exclusively breastfed and babies who have breast milk + solids. Breast milk is more liquid so wouldn’t solid introduction risk causing some constipation?


    August 7, 2016
    • Yep, I think once you start solids, it’s good to offer a little water too. It’s really normal to see some constipation when you start solids – partly, it’s just the transition as the baby’s gut adapts to more solid foods. This is also one of the benefits of purees, because there is so much water in purees that this reduces the risk of constipation. One way to incorporate more liquids while still letting your baby self-feed is to make little popsicles from pureed fruits and veggies (and I often add yogurt). But yes, baby can also try drinking a little water from an open cup or straw cup – and if constipation is a problem, you can try a little prune juice too:)


      August 13, 2016
  7. Joy #

    This discussion is what most concerned mothers think about. Testing out which ones your baby is allergic to can be terrifying and liberating at the same time.


    October 2, 2016
  8. I found your article while I was searching for information on starting solids. I think this was a really interesting study with intriguing results (I’m very curious whether there were any further effects in early introduction group).
    With my first baby we started solid foods when she was 6 month old and I was still breastfeeding. Now I have almost 7 month old sons and I offered them solids for the first time when they were 4,5 m.o. However, I’m also breastfeeding. They don’t have any food allergies and I hope it’ll stay this way.
    Of course it is very important not to rush into it and not to put pressure on the baby when he/she is not ready (and interested:)).
    p.s. my whole family comes from a country area and everyone (including me, my siblings, parents and their siblings) started to eat solids early (like cow’s milk, yogurt, eggs, vegetables, cooked meat and rice) and no one has problems with food allergies! I guess few years ago no one thought about food allergies. However we were eating what we grew or produced in our farm:)


    February 15, 2017

Comments are closed.

%d bloggers like this: