Before weaning, baby allergies can be surprisingly hard to spot.

Mother holding her newborn baby in black and white.

When your baby hasn’t eaten a single solid food, allergy is probably not the first place your mind goes.

Babies get away with a remarkable amount in the name of being babies: vomiting over the only clean top you own, erupting in mysterious rashes and producing nappies that have you Googling things you never imagined would become part of your search history. Most of the time, it really is just babyhood.

But sometimes the pieces don’t quite fit. There may be no dramatic reaction, no obvious culprit and no helpful little sign announcing that food is the problem. Instead, there’s a baby who never seems quite right and a growing sense that something is being missed.

Food allergy in very young babies can be particularly difficult to untangle because, before weaning, the baby may never have eaten the food they are reacting to. If you are breastfeeding, the clues can be buried somewhere inside another person’s diet entirely. It is less “he ate this and this happened” and more detective work conducted on very little sleep.

Ten years ago, I knew none of this. When my second son was born, he seemed perfectly well. But as the weeks passed, that began to change. His eczema became severe, he vomited frequently, wheezed and seemed permanently under the weather. Rather than settling into newborn life, he was becoming more and more unwell. Food allergy wasn’t even on my radar. I was frightened there was something much more serious wrong with him.

I was also suffering from severe postnatal anxiety and panic attacks, which made trusting my own judgement incredibly difficult. I took him to the doctor, explained how worried I was and was reassured and sent home. When you are already wondering whether anxiety is making you see catastrophe everywhere, being told your baby is fine can be surprisingly persuasive. I tried to believe it. He just kept getting worse.

The turning point came from a friend in my original NCT group. She was a dietitian and had a child with severe allergies herself. When I described what was happening, she recognised a pattern I hadn’t. She suggested temporarily removing dairy, egg and soya from my diet, keeping a proper record of his symptoms and booking another doctor’s appointment while I was doing it. This time, she encouraged me to ask for a referral to a paediatric dietitian.

By the time I finally saw one, my son was around three months old and still exclusively breastfed. She was the first medical professional who made me feel I hadn’t imagined any of it. There was clearly an allergic picture here; the difficult part was working out what was causing it.

Why Breastfeeding Makes Allergy So Difficult To Spot

The problem was that he had never actually eaten any of the foods we were investigating. There was no suspicious spoonful of yoghurt or scrambled egg to point at. If food was behind his symptoms, he was encountering small amounts of those proteins through my breast milk.

That makes the whole thing considerably less tidy. Allergens from a breastfeeding mother’s diet can reach her baby through breast milk, which is why specialist guidance specifically says the maternal diet should be considered when allergy is suspected in a breastfed baby.

It also means you are effectively trying to decipher one person’s allergy through another person’s meals. Dairy at breakfast, egg in something baked, soya hidden in a label, nuts somewhere else entirely — while the baby is being exposed indirectly rather than sitting down to eat a full portion themselves.

“With a breastfed baby, the food causing the reaction may never have touched their lips.”

That was exactly what happened with my son. He had been well when he was born, then over the following weeks he became progressively more unwell. There was no single feed I could point to. Instead there was continuing exposure, worsening eczema, wheezing, vomiting and a baby who seemed less and less like the healthy newborn I had brought home.

Under the paediatric dietitian’s guidance, we stripped things right back. I removed the major suspects from my own diet — dairy, egg, soya and nuts — and waited for his symptoms to settle. Then, one by one, we began putting foods back in.

Because he was still exclusively breastfed, one meal in my diet wasn’t necessarily enough to tell us much. A food had to be back in long enough for us to see whether his symptoms started returning. If they did, it came out again, we waited for things to settle, and only then moved on to the next. He reacted to every one we tried.

It sounds almost absurdly simple written down — remove a food, wait, bring it back and watch — but that slow process is still one of the ways delayed food allergy is diagnosed. NICE recommends an elimination period of around two to six weeks followed by reintroduction when non-IgE allergy is suspected, while BSACI notes that some supervised exclusion diets may need six to eight weeks. Nice

Multiply that by several suspected foods and suddenly you are no longer talking about a quick dietary experiment. You are talking about months.

And you are trying to answer more than one question. What is the baby reacting to? What sort of reaction is it? Has removing that food genuinely changed things, or have you simply landed on one unusually good week?

Allergy Is Not The Same As Intolerance

This was another thing I didn’t understand at the beginning. A food allergy involves the immune system reacting to a food protein. An intolerance does not. The difference matters because an intolerance can make someone thoroughly miserable, but it cannot cause anaphylaxis.

Young baby with an older sibling kissing his cheek after allergy symptoms had improved. — Once the allergens were identified, his skin finally began to clear.

Some allergies are what doctors call IgE-mediated. IgE is an antibody produced by the immune system, and this is the type associated with the quicker reactions most of us picture when we hear “food allergy”: hives, swelling, vomiting, coughing, wheezing and, in severe cases, anaphylaxis. Skin-prick tests and specific-IgE blood tests are used to help investigate this kind of allergy. BSACI

Then there are non-IgE allergies. These still involve the immune system, but the reaction is delayed and can be much harder to connect to one particular food. There is no simple validated skin or blood test for these reactions, which is why that laborious remove-and-reintroduce process can matter so much.

Babies make all of this harder because they cannot tell you what they are feeling. I have allergies myself; kiwi makes my tongue swell and my throat itch. I can immediately say that something is happening. A three-month-old cannot tell you their mouth feels strange, their throat itches or their stomach hurts. You are reading the whole thing from the outside.

By the time we reached weaning, my son was much better — but he still wasn’t completely well. That told us there were pieces missing.

I was still breastfeeding, so I was still following the same restricted diet, but now food was reaching him directly too. In some ways, that made things easier. Suddenly there was a food, a time and sometimes a reaction you could actually connect.

Weaning helped us identify more of what he was reacting to. Fish, tomatoes and berries joined dairy, egg, soya and nuts, and gradually we began to understand why, despite all the changes I had already made, he had never become completely well. Coconut came later, around his first birthday, after he reacted to an allergy-friendly cake made with it.

For babies without an existing allergy, current NHS guidance is to introduce common allergenic foods from around six months, one at a time and in small amounts so any reaction is easier to spot. Babies who already have eczema or a diagnosed food allergy may need more individual advice before those foods are introduced.

Once he turned one, we were finally able to start skin-prick testing in hospital. At one appointment he had around 40 tests — 20 on each tiny arm. After months of trying to decipher reactions through my own diet, there was something oddly reassuring about finally seeing the investigation laid out in front of us.

Skin-prick and specific-IgE blood tests can be extremely useful when an immediate allergy is suspected, but they still have to be read alongside the child’s history. A positive test shows sensitisation; on its own, it does not necessarily prove that eating the food will cause a clinical reaction.

By the end of that first year, though, the difference was extraordinary. Once we had finally identified what he was reacting to, he became a different child. His skin went from angry, severe eczema to smooth and completely clear. The wheeze disappeared. He stopped seeming permanently ill beyond the usual conveyor belt of childhood coughs and colds. More than anything, he was simply happier in himself.

The Skin-Barrier Lesson I Wish I’d Known

There is one thing I did during that first year that I would never do now. This was peak coconut-oil era, when it seemed to be the answer to everything — cooking, hair, skin, you name it. My son’s eczema was severe, his skin was angry and inflamed, and in an attempt to soothe it I would cover him in coconut oil because, at the time, “natural” felt reassuringly safe.

Then, on his first birthday, we had an allergy-friendly cake made with coconut and he reacted. He had never shown any sign of a coconut allergy before, yet it went on to become one of his most persistent. Years later, after discussing that history with his allergy specialists, I began to understand why those months of repeatedly putting a food-derived oil onto badly compromised skin might have mattered.

The science around this has moved on significantly. We now know that eczema is not just something happening on the surface; it means the skin barrier is damaged, and in susceptible babies that can give food proteins an opportunity to interact with the immune system through the skin. Research increasingly supports the idea that this kind of exposure can contribute to sensitisation, particularly in babies with significant or early eczema.

It is part of what researchers call the dual-allergen exposure theory: exposure through damaged skin may increase the risk of allergy, while eating a food appropriately can help the immune system learn tolerance. Which is why I now look at all those supposedly gentle, food-based “natural” products rather differently.

That doesn’t mean coconut oil causes coconut allergy, or that parents need to panic about every ingredient in the bathroom cabinet. But if a baby has severe eczema, broken or inflamed skin, I would be far more cautious now about repeatedly putting food-derived oils or ingredients onto it. “Natural” is a marketing word. A compromised skin barrier is a biological reality.

The List Can Change

My son is ten now and still returns to hospital for yearly testing, but the picture looks completely different. At one point we were dealing with roughly ten allergies. Today we are down to three. Even some of the nut allergies have disappeared.

That is perhaps the thing I most wish somebody had told me when I was standing in a supermarket reading every label while simultaneously wondering what on earth I was going to eat myself: the allergy list you have in infancy is not necessarily the one your child will carry forever. Some allergies, particularly milk and egg, are often outgrown, while others can be far more stubborn. Any reintroduction, though, should be done with guidance from your child’s allergy team.

The first year was hard. Managing a tiny baby’s allergies while breastfeeding meant managing my diet and his, then learning an entirely new set of rules once weaning started. But by the end of that year we finally understood enough to make him well.

What I wish I had known at three months old was that allergy in a very young baby does not always look like the dramatic reaction we recognise from adults. There may be no incriminating spoonful, no child old enough to tell you their throat feels funny and no single moment that solves the mystery. Sometimes the first clue is simply that your baby is not well. That feeling is not a diagnosis, but it is a very good reason to keep asking why.