This site is intended for health professionals only

How to diagnose and manage cow’s milk protein allergy

How to diagnose and manage cow’s milk protein allergy

Dr Toni Hazell offers advice on how to ensure cow’s milk protein allergy is addressed appropriately in primary care

Cow’s milk allergy (CMA, also known as cow’s milk protein allergy or CMPA) is defined by NICE as ‘a reproducible immune-mediated allergic response to one or more proteins in cow’s milk’.1 To understand the two different types of CMPA (IgE and non-IgE mediated) we need to touch briefly on some of the basics of allergy. 

When a person with IgE-mediated CMPA ingests milk, an allergy-specific immunoglobulin (IgE) binds to receptors on a mast cell. That mast cell degranulates and releases chemicals such as histamine, which cause symptoms of allergy. The same mechanism causes allergies such as hayfever (when a person comes into contact with pollen) and other food allergies. IgE-mediated allergy is fairly easy to diagnose, because the reaction usually comes quickly after contact with the allergen (minutes to a maximum of two hours), and because raised IgE levels can be measured in a blood test. 

Non-IgE-mediated allergy is more complicated. The immune process is less clear cut, there are no diagnostic tests and the reaction can come hours after contact with the allergen – sometimes up to 72 hours later, making it much harder to link the two.¹ The list of symptoms used by NICE of IgE- and non-IgE-mediated allergy is shown in the table (link below).1 While there is some overlap, broadly speaking IgE-mediated allergy presents with symptoms that a lay person would recognise (rash, wheeze, itch, swelling) whereas non-IgE-mediated allergy can present with more vague symptoms, such as loose or bloody stools, food refusal, constipation and failure to thrive.

Careful history and investigation
Diagnosis of CMPA depends on a healthcare professional listening carefully to the patient’s history; the symptoms may present during a formal appointment but may just be mentioned in passing when discussing feeding, sleep or other issues. The international Milk Allergy in Primary Care (iMAP) guideline2 is widely used and has easy-access flowcharts that outline the diagnostic process for both types of CMPA. IgE-mediated allergy is diagnosed in the same way as any other condition – take a history, examine the child and then investigate. This will usually involve a specific IgE blood test, done in primary care, but in some areas your pathway may be referral to an allergy clinic for skin-prick tests. 

If non-IgE-mediated allergy is suspected, the only way to make a diagnosis is by implementing a milk-free diet. Formula-fed babies (whether exclusively or in combination with breastmilk) will require a prescribed hypoallergenic formula. An extensively hydrolysed formula (eHF) will be suitable for most children. However, for those who have severe symptoms of allergy (such as anaphylaxis), or where there is no improvement with an eHF, an amino acid formula (AAF) should be tried. Follow local guidelines on choice of brand. Parents may wish to buy their own alternative milk, but this is not wise. Children with CMPA are often also allergic to other mammalian milks, such as goat and sheep, as well as soy milk. Soy milk provides insufficient nutrients and also has a weak oestrogenic effect that could theoretically affect the reproductive system. 3,4 Milks such as almond, oat and coconut have poor nutritional value, while rice milk has a natural arsenic content that makes it unsuitable for use below the age of five.1

For babies who are breastfed (exclusively or in combination with formula), the mother should exclude all cow’s milk products from her own diet. This is not always easy and she should be offered dietitian support and a calcium supplement. If the child’s symptoms improve with milk exclusion then it is crucial to reintroduce the milk and monitor to see if symptoms recur. This is not always popular with parents, as no one wants to deliberately make their child unwell, but if it isn’t done we risk overdiagnosis, as symptoms may have coincidentally improved when milk was excluded. CMPA is much less common in exclusively breastfed infants, with one study of 1,749 infants identifying 39 with CMPA, of whom only nine were exclusively breastfed.5

Ongoing management
In most areas, rules on medicines management indicate that ongoing provision of hypoallergenic formula needs to be accompanied by a referral to paediatrics. Whether or not this is the case, the child should be kept on a cow’s milk-free diet for at least six months, and as a minimum until they are nine to 12 months old, at which point there can be a gradual, supervised reintroduction of cow’s milk. This should be done in hospital if there was anaphylaxis, but otherwise can be done at home using the milk ladder to guide reintroduction. Referral to paediatrics may be done as well as, or instead of, referral to dietetics, depending on local pathways. 

Be mindful of overdiagnosis concerns
Parents who have done background reading or discussed their child’s CMPA online might ask if you are aware of overdiagnosis of the allergy. This has received increased attention since a 2018 report, which included criticism of the iMAP guidance.

It is reasonable to query a sudden increase in any diagnosis or treatment over a short period of time. Much of the concern has come from the observation that UK prescriptions for hypoallergenic formula increased markedly (by 500% between 2006 and 2016), without anything to suggest a true increase in allergy. However, closer scrutiny reveals a more complicated picture. The steep increase in formula prescriptions actually began in 2003, 10 years before the first iMAP guidance was published in 2013 and four years before there was any written guidance on CMPA in the UK.  Moreover, a key confounder here is that soy-based formulas were widely used in the UK until 2003, when guidance was issued advising against them (for the reasons discussed above).3,4 Prescriptions for soy-based formula fell steeply from 2003, the same point at which prescriptions for hypoallergenic formula started to increase, suggesting some infants would have been switched from one to the other. It is also possible that wider recognition of non-IgE-mediated CMPA contributed to the increase.

Another criticism of the guidelines has been use of non-specific symptoms. As any parent knows, children’s bowels seem to have a mind of their own. The contents of nappies changes from time to time. Likewise, infants’ burping, vomiting and colic can come and go for no obvious reason. There is clearly a risk that such normal variation could be mistaken for allergy and lead to overdiagnosis. However, the iMAP guidance has always advocated reintroduction of cow’s milk, so an improvement in symptoms when milk is removed and a worsening when it is reintroduced is needed for diagnosis. Provided this is carefully followed, we should avoid overdiagnosis due to a coincidental change in symptoms.   

While the formula industry has come under some legitimate criticism for promotional tactics, specific concerns about undue influence on the iMAP guidelines seem to be clouded by misconceptions. For example, the iMAP flowcharts have always been open access, with the unintended consequence that some formula manufacturers have used them in promotional material. While the iMAP authors have declared research grants or consultancy fees from the makers of formula milk, this is common under current funding systems; many doctors who write, present and conduct research inevitably receive at least some industry funding for their work.7

It should be noted that the 2019 iMAP update included unfunded input from parent and professional groups with no industry ties. 

Breastfeeding support
Concerns have also been raised around a reduction in breastfeeding rates, but this likely relates to wider issues around a lack of breastfeeding support. It is unusual for me to see a woman who is fully breastfeeding at the time of her six-week check and this is backed up by statistics – 81% of women give their baby some breastmilk after birth but only 24% are exclusively breastfeeding at six weeks, with this figure dropping to 1% at six months with 34% giving some milk at this point.8 

In the absence of any political will to, for example, adequately fund a network of accessible lactation advisors or to ensure all women have access to decent paid parental leave and support, it’s easy to blame this on guidelines that recommend breastfeeding women cut out dairy (which is difficult and may lead women to stop breastfeeding if not properly supported). The 2019 iMAP guideline does advise that exclusive breastfeeding until six months is a good thing, and recommends active support. Symptomatic CMPA is less common in infants who are exclusively breastfed than those given formula, but it is possible, so it is important that we provide quality information and support to this group of women, and that we are able to make a diagnosis if appropriate, while the woman continues to breastfeed.

I was always taught that 80% of a diagnosis comes from the history9 and that medicine is an art not a science. We are used to backing up initial impressions with tests, but where this isn’t possible (such as with non-IgE-mediated CMPA), doctors and nurses in primary care should be comfortable relying on clinical skills and our ‘spidey sense’ that something is wrong. If you listen, consider possible differentials, seek help if you need it and always re-challenge to confirm a non-IgE-mediated CMPA diagnosis, your patients will be in good hands.

Dr Toni Hazell is a GP in north London


  1.  NICE CKS. Cow’s milk allergy in children. 2021. Link
  2.  Fox A, Brown T, Walsh J et al. An update to the Milk Allergy in Primary Care guideline. Clin Transl Allergy. 2019 Aug 12;9:40.
  3.  Tuohy PG. Soy infant formula and phytoestrogens. J Paediatr Child Health. 2003 Aug;39(6):401-5
  4.  Committee on toxicity of chemicals in food, consumer products and the environment. Statement on the potential risks from high levels of soya phytoestrogens in the infant diet. 2003. Link
  5.  Høst A, Husby S, Osterballe O. A prospective study of cow’s milk allergy in exclusively breast-fed infants. Incidence, pathogenetic role of early inadvertent exposure to cow’s milk formula, and characterization of bovine milk protein in human milk. Acta Paediatr Scand. 1988 Sep;77(5):663-70.
  6.  Van Tulleken C. Overdiagnosis and industry influence: how cow’s milk protein allergy is extending the reach of infant formula manufacturers. BMJ 2018; 363: k5065
  7.  Sunshine UK. A register of doctors’ declared interests. Link 
  8.  UNICEF. Breastfeeding: a public health issue. Link 
  9.  Australian Family Physician. A is for aphorism – is it true that ‘a careful history will lead to the diagnosis 80% of the time’? 2012. Link


See how our symptom tool can help you make better sense of patient presentations
Click here to search a symptom