How Is CMPA Diagnosed in Babies?
You've read the symptom lists, and you recognise too many of them — the mucus in the nappy, the eczema that won't settle, the feeds that end in screaming. Now you want someone to tell you, definitively, whether it's cow's milk protein allergy.
Here's what surprises most parents: for the most common kind of CMPA in babies, there's no test that gives that answer. Diagnosis works differently — and once you know how, you can make the process faster and far more conclusive.
Two Kinds of CMPA, Two Different Routes
CMPA comes in two broad forms, and they're diagnosed in completely different ways.
IgE-mediated CMPA is the fast kind. Reactions start within minutes to about two hours of milk — hives, swelling around the mouth or eyes, sudden vomiting. Because the immune system makes IgE antibodies to milk, this form can be tested: an allergy specialist may use a skin prick test or a blood test for milk-specific IgE, always read alongside what actually happened (North Yorkshire CMPA guideline).
Non-IgE-mediated CMPA is the slow kind, and the more common one in babies — especially breastfed babies. Symptoms build over roughly 2 to 72 hours: reflux, mucus or blood in stools, loose stools or straining, eczema, colic-like distress (Nottinghamshire CMPA guidance). For this form, skin prick and blood tests aren't recommended, because they can't detect it (Allergy UK).
Why There's No Blood Test for the Common Kind
"Can't they just test for it?" is the question parents ask most. For delayed CMPA, the honest answer is no: the reaction doesn't run through the IgE antibodies those tests measure, so a normal result doesn't rule it out.
That's why international guidance describes removing milk and then bringing it back as the only reliable way to diagnose the gut symptoms of non-IgE CMPA (WAO DRACMA guideline). In practice, your record of what happened becomes the test.
How the Milk-Free Trial Works
If your GP suspects mild-to-moderate non-IgE CMPA, the usual first step is a diagnostic trial of two to four weeks without cow's milk protein (North Yorkshire CMPA guideline):
- If you're breastfeeding: you remove cow's milk protein from your own diet and carry on breastfeeding. NHS guidance pairs this with calcium and vitamin D supplements for you (Nottinghamshire CMPA guidance).
- If your baby has formula: they're usually switched to an extensively hydrolysed formula, where the milk proteins are broken down small enough not to trigger a reaction in most babies.
- If your baby is on solids: dairy comes out of their food too.
The question isn't "do they seem a bit better?" It's whether symptoms clearly improve. If they don't — and the diet was strict — CMPA becomes unlikely and your GP will look for another explanation. What a strict milk-free diet actually involves is covered in CMPA and breastfeeding: what to avoid.
The Step Everyone Wants to Skip
When symptoms improve, it's tempting to stop there. But a planned reintroduction of cow's milk is what actually confirms the diagnosis (Nottinghamshire CMPA guidance). Babies improve for lots of reasons — a cold ending, a growth spurt passing, reflux easing with age. If symptoms come back when milk returns, the diagnosis is confirmed. If they don't, your baby has just avoided months of unnecessary restriction.
That matters because CMPA is over-diagnosed. In one large UK study, about 16% of parents reported a problem with cow's milk, but only about 1.4% of babies had a formal diagnosis (BEEP trial cohort).
Once it's confirmed, the usual plan is to stay milk-free until around 9–12 months old, and for at least six months after diagnosis, then reintroduce dairy gradually with a milk ladder (Nottinghamshire CMPA guidance). That stage is covered in reintroducing dairy while breastfeeding.
When It Shouldn't Wait for a Trial
Some situations skip the home trial and go straight to referral: faltering growth, severe eczema, or any immediate reaction after milk (Nottinghamshire CMPA guidance). Swelling of the lips, tongue or face, difficulty breathing, or a baby who becomes floppy or pale needs emergency care (999 or A&E) straight away.
What Makes the Trial Conclusive
A milk-free trial is only as good as the record behind it. Going into the review with "it seemed a bit better" leaves your GP guessing. Going in with symptoms, times and severity logged day by day — before, during and after — gives them something to decide on.
Worth recording through the whole process: every symptom with its time and how bad it was, everything you and your baby ate, photos of rashes and nappies, and anything else that could explain a bad day, like teething, vaccinations or a cold.
These observations are based only on patterns in your own log. They are not medical advice — always talk to your GP, health visitor, paediatrician or dietitian about suspected CMPA before changing your or your baby's diet.
Turning This Into a Record You Can Use
Whatever's behind your baby's symptoms, the thing that identifies it is the same: a written record of what was eaten and what happened, with times attached. Memory doesn't hold up across three days of this — and "I think it's dairy" isn't something a GP can act on. A log is.
On paper: A free printable food and symptom diary - one version for breastfeeding, which tracks your food and your baby's side by side, and one for formula and solids. One page per day, print at home, no signup and no email.
→ Get the free printable diary
In the app: LittleClues does the same thing with the timing built in: foods and symptoms logged on their own separate timelines, with the app ranking which foods are most likely behind the reactions. Free on iOS and Android.
Frequently Asked Questions
Is there a test for CMPA in babies?
For the fast, IgE-mediated type, an allergy specialist can use a skin prick test or a blood test for milk-specific IgE. For the more common delayed, non-IgE type there's no reliable skin or blood test — it's diagnosed by removing cow's milk protein and then reintroducing it.
Can a blood test rule out CMPA?
Not for the delayed type. Standard allergy blood tests measure IgE antibodies, and non-IgE CMPA doesn't involve them, so a normal result doesn't rule it out.
How long is the milk-free trial for CMPA?
Usually two to four weeks, though your GP or dietitian may set a different length. The aim is to see whether symptoms clearly improve without cow's milk protein.
Why reintroduce milk if my baby got better?
Because improvement on its own doesn't prove milk was the cause — babies get better for many reasons. If symptoms return when milk comes back, the diagnosis is confirmed; if they don't, your baby avoids an unnecessary restricted diet.
How do I know if my baby has CMPA?
Not from one symptom or one bad day. CMPA is suspected when several typical symptoms appear together and persist, and it's confirmed by clear improvement on a milk-free trial followed by symptoms returning when milk is reintroduced — agreed with your GP or dietitian.