How Is CMPA Diagnosed? What to Expect at the Doctor

If you are searching for how CMPA is diagnosed, you are probably holding an uncomfortable baby and wondering whether milk is the reason. The process is more structured than it looks, and you can do a lot to make it faster. Here is what clinicians actually do, roughly in order, and what to bring.

Key takeaways

• There is no single test for CMPA. Diagnosis starts with an allergy-focused history and an exam that includes growth.^[1][2]
• For delayed (non-IgE) symptoms, the practical standard is a supervised 2 to 4 week milk-free trial followed by planned reintroduction.^[1][3]
• For rapid (IgE) reactions, skin prick or specific IgE blood tests help, but a positive result alone does not confirm allergy.^[1][2]
• A supervised oral food challenge (milk given back under a plan) is what confirms the diagnosis; the double-blind version is the research gold standard.^[1]
• IgG panels, hair analysis and kinesiology are not recommended by any major guideline.^[1][2][8][9]
• CoMiSS is an awareness tool, not a diagnostic test, and common baby symptoms are not diagnostic on their own.^[1][4]

Step one: the allergy-focused history

ESPGHAN (Europe's pediatric gastroenterology society), NICE (the UK's guideline body) and the international iMAP group agree: diagnosis begins with a detailed conversation, not a test.^[1][2][3] NICE tells clinicians not to order allergy tests without first taking an allergy-focused history, and to interpret any result in the light of it.^[2]

What your clinician will ask

• How old your baby was when symptoms started.^[2]
• How quickly symptoms appear after a feed: minutes, hours or days.^[2]
• What the symptoms look like: skin, tummy, breathing or general unsettledness.^[1]
• How your baby is fed, and any formula changes.^[1][3]
• Whether parents or siblings have eczema, asthma, hay fever or food allergies.^[1][3]
• What you have already tried, and what happened when milk was removed or brought back.^[2]

Timing is one of the most useful details. In IgE-mediated allergy (the "immediate" type, driven by IgE antibodies) symptoms usually start within minutes of a feed. In non-IgE-mediated allergy (the "delayed" type) they usually begin two or more hours later, most often between 6 and 72 hours.^[1] If the terms are new, our explainer on CMPA, CMA, CMPI and MSPI untangles them, and What is CMPA? covers the basics.

The growth check and physical exam

Expect your baby to be weighed, measured and examined. NICE asks clinicians to pay particular attention to growth and signs of malnutrition, and ESPGHAN includes a complete physical examination in every work-up.^[1][2] Faltering growth (weight gain that slows or stalls) together with gut symptoms is one reason NICE recommends specialist referral.^[2]

ESPGHAN notes that challenge-proven cow's milk allergy affects roughly 1% of infants, while troublesome crying, vomiting or rashes are each reported in 15 to 20% of babies.^[1] That is why guidelines insist on a short trial and reintroduction before any long-term milk-free diet, and why the NHS advises parents not to cut out a major food such as milk on their own.^[1][6] If lactose might be the issue instead, read CMPA vs lactose intolerance in babies.

Non-IgE CMPA: elimination and reintroduction, supervised

For delayed symptoms there is no reliable skin or blood test. ESPGHAN describes the response to a diagnostic elimination diet followed by an oral food challenge as the cornerstone of diagnosing cow's milk allergy.^[1] It is planned with your child's clinician and, ideally, a pediatric dietitian, who checks the diet is nutritionally adequate and sets the dates.^[2]

How long does the trial last?

ESPGHAN recommends 2 to 4 weeks for most infants; iMAP says up to 4 weeks with a minimum of 2; NICE, for the UK, allows 2 to 6 weeks.^[1][2][3] Clinicians look for a clear improvement, not necessarily complete resolution.^[3] If nothing changes, the guidance is to re-evaluate the diagnosis rather than extend the diet.^[1]

Why reintroduction is not optional

Improvement alone does not prove milk was the problem; babies also grow out of phases. iMAP says the reintroduction step "is of ultimate importance to confirm the diagnosis" and provides a written protocol for parents and clinicians.^[3] ESPGHAN warns that over-diagnosed children are exposed to the harms of an unnecessary elimination diet.^[1] For mild-to-moderate delayed allergy the reintroduction usually happens at home with a written plan, and a health professional interprets the result.^[1][3]

Breastfed and formula-fed babies

CMPA in exclusively breastfed babies is rare.^[1] When suspected, ESPGHAN says a maternal milk-free diet for 2 to 4 weeks while continuing to breastfeed may be considered, then milk is returned to the mother's diet to confirm.^[1] Because unwarranted maternal elimination diets can harm the mother's nutrition and cut breastfeeding short, this is done with a dietitian, not alone.^[1] For formula-fed babies the trial usually uses an extensively hydrolysed formula (eHF), in which milk proteins are broken into very small pieces; an amino acid formula (AAF) is reserved for severe cases; your clinician chooses.^[1] Lactose-free and partially hydrolysed (often labelled "HA") formulas still contain cow's milk protein and are not suitable for babies with CMPA, so tell your clinician if you have already tried one.^[1][11]

IgE-mediated CMPA: skin prick and specific IgE tests

If reactions are rapid (hives, swelling, vomiting or wheeze within minutes to two hours of a feed), NICE recommends a skin prick test and/or a blood test for specific IgE to milk.^[2] If a reaction involves breathing difficulty, swelling of the tongue or throat, or a baby who goes limp or floppy, that is an emergency: see When it is an emergency below and call 911 (US), 999 (UK) or your local emergency number.^[10] In a skin prick test a tiny amount of milk protein is placed on the skin, which is lightly scratched; a small raised bump, like an insect bite, suggests sensitization.^[7] This is usually done by an allergist in the US or a hospital allergy service in the UK.

Two caveats. A positive test shows sensitization, not necessarily allergy: ESPGHAN states that raised specific IgE and positive skin prick tests do not confirm cow's milk allergy, whose diagnosis is based on symptoms, and that these tests are better at ruling IgE allergy out than in.^[1] And total IgE (the overall antibody level) does not help.^[1] NICE adds that skin prick tests should only be done where anaphylaxis can be managed.^[2]

The oral food challenge: the gold standard

An oral food challenge (OFC) is a supervised feed in which milk is given in gradually increasing amounts under observation. A double-blind, placebo-controlled challenge is the gold standard for confirming cow's milk allergy; in everyday practice an open challenge is used, with at least 48 to 72 hours of observation recommended for delayed reactions.^[1] ESPGHAN considers a challenge mandatory except for babies who have had life-threatening reactions such as anaphylaxis alongside high specific IgE.^[1] The UK's BSACI guideline likewise lists history, IgE testing, a diagnostic elimination diet and an oral challenge as the elements that, together, reach a diagnosis in most cases.^[5]

Tests that are not recommended

These are widely advertised to parents. NICE says not to use vega testing, applied kinesiology, hair analysis or serum-specific IgG testing to diagnose food allergy.^[2] ESPGHAN adds bioresonance, iridology, cytotoxic tests and IgG or IgG4 levels to the list of unvalidated approaches.^[1] The AAAAI explains why IgG panels mislead: IgG is likely a normal immune response to eating a food, and higher IgG4 may indicate tolerance.^[8] ASCIA also strongly advises against online allergy 'testing' services and lists cytotoxic, kinesiology, hair analysis, vega and IgG tests among those that can lead to misdiagnosis and costly, often dangerous, dietary restriction.^[9]

What about the CoMiSS score?

CoMiSS stands for Cow's Milk-related Symptom Score. It scores crying, regurgitation, stools, skin and respiratory symptoms for a total of 0 to 33; the 2022 update lowered the cut-off suggesting milk-related symptoms from 12 to 10 and treats a score below 6 as unlikely to be milk-related.^[4] Its authors are explicit: "CoMiSS is an awareness tool and not a diagnostic test," and diagnosis "should always be the result of clinical interpretation of the outcome of a 2 to 4 week diagnostic elimination diet followed by a challenge test."^[4] ESPGHAN notes that CoMiSS and iMAP are non-specific: in one large cohort, 74% of babies had two or more mild-to-moderate symptoms at some point in their first year.^[1] A high CoMiSS is a reason to book an appointment, not a diagnosis.

Diagnostic tools at a glance

What to bring to the appointment

Because the history is the cornerstone, your notes are the most useful thing in the room.

• A feeding record: what, when and roughly how much, including formula changes.
• A symptom record with clock times, so the clinician can see how long after a feed things happen.^[2]
• Dated photos of rashes and diaper (nappy) contents.
• Recent weights or the growth chart.
• Family history of eczema, asthma, hay fever or food allergy.
• What you have already tried, for how long, and what changed.
• Labels of any formula or foods you are unsure about. Our free hidden milk ingredients label checklist shows the names milk hides behind.

A notebook works. If you would rather use your phone, the CMPA Baby app was built for keeping a record of feeds and symptoms between visits, so you can show the pattern rather than describe it from memory. It is a record-keeping tool, not a test or a treatment. Still at the "is this anything?" stage? 5 early signs of CMPA describes what parents commonly notice first.

When it is an emergency

Most CMPA symptoms are not emergencies, but anaphylaxis can happen with milk allergy and needs urgent treatment.^[6][10] Signs include swelling of the tongue or throat, difficulty breathing or very fast breathing, noisy breathing or wheeze, difficulty swallowing, skin, lips or tongue turning blue, grey or pale, or a baby who is limp, floppy or not responding the way they normally do.^[7][10] If you see any of these: use an epinephrine (adrenaline) auto-injector straight away if one has been prescribed, call 911 (US), 999 (UK) or your local emergency number and say you think it is anaphylaxis, lie your baby down, and if there is no improvement after 5 minutes and you have a second auto-injector, use it.^[10] Do not wait to see whether it passes.^[10]

Talk to your child's clinician

Nothing here replaces a conversation with your pediatrician, GP, allergist or dietitian, and none of it should be used to start or stop a diet on your own. It can help you walk in with the right records and questions: which type of reaction does this look like, what is the plan for elimination and reintroduction, and who will interpret the result? For why this diagnosis seems so common now, read Why does CMPA seem so common now?; for what comes after a confirmed diagnosis, see Do babies outgrow CMPA?