This article was originally published by Undark Magazine.
For Taylor Arnold, a registered dietitian nutritionist, feeding her second baby was not easy. At eight weeks old, he screamed when he ate and wouldn’t gain much weight. Arnold brought him to a gastroenterologist, who diagnosed him with allergic proctocolitis—an immune response to the proteins found in certain foods, which she narrowed down to cow’s milk.
Cow’s-milk-protein allergies, or CMPA, might be on the rise—following a similar trend in other children’s food allergies—and they can upend a caregiver’s feeding plans: In many cases, a breastfeeding parent is told to eliminate dairy from their diet, or switch to a specialized hypoallergenic formula, which can be expensive.
But although some evidence suggests that CMPA rates are climbing, the source and extent of that increase remain unclear. Some experts say that the uptick is partly because doctors are getting better at recognizing symptoms. Others claim that the condition is overdiagnosed. And among those who believe that milk-allergy rates are inflated, some suspect that the global formula industry, valued at $55 billion according to a 2022 report from the World Health Organization and UNICEF, may have an undue influence.
Meanwhile, “no one has ever studied these kids in a systematic way,” Victoria Martin, a pediatric gastroenterologist and allergy researcher at Massachusetts General Hospital, told me. “It’s pretty unusual in disease that is this common, that has been going on for this long, that there hasn’t been more careful, controlled study.”
This lack of clarity can leave doctors in the dark about how to diagnose the condition and leave parents with more questions than answers about how best to treat it.
When Arnold’s son became sick with CMPA symptoms, it was “really, really stressful,” she told me. Plus, “I didn’t get a lot of support from the doctors, and that was frustrating.”
Though the gastroenterologist recommended that she switch to formula, Arnold ultimately used a lactation consultant and gave up dairy so she could continue breastfeeding. But she said she can understand why others might not make the same choice: “A lot of moms go to formula because there’s not a lot of support for how to manage the diet.”
Food allergies primarily come in two forms: One, called an IgE-mediated allergy, has symptoms that appear soon after ingesting a food—such as swelling, hives, or difficulty breathing—and may be confirmed by a skin-prick test. The second, which Arnold’s son was diagnosed with, is a non-IgE-mediated allergy, or food-protein-induced allergic proctocolitis, and is harder to diagnose.
With non-IgE allergies, symptom onset doesn’t tend to happen immediately after a person eats a triggering food, and there is no definitive test to confirm a diagnosis. (Some specialists don’t like to call the condition an allergy, because it doesn’t present with classic allergy symptoms.) Instead, physicians often rely on past training, online resources, or published guidelines written by experts in the field, which list symptoms and help doctors make a treatment plan.
Read: We’ve never been good at feeding babies
Numerous such guidelines exist to help providers diagnose milk allergies, but the process is not always straightforward. “It’s a perfect storm” of vague and common symptoms and no diagnostic test, Adam Fox, a pediatric allergist and a professor at King’s College London, told me, noting that commercial interests such as formula-company marketing can also be misleading. “It’s not really a surprise that you’ve got confused patients and, frankly, a lot of very confused doctors.”
Fox is the lead author of the International Milk Allergy in Primary Care, or iMAP, guidelines, one of many similar documents intended to help physicians diagnose CMPA. But some guidelines—including iMAP, which was known as the Milk Allergy in Primary Care Guideline until 2017—have been criticized for listing a broad range of symptoms, like colic, nonspecific rashes, and constipation, which can be common in healthy infants during the first year of their life.
“Lots of babies cry, or they [regurgitate milk], or they get a little minor rash or something,” Michael Perkin, a pediatric allergist based in the U.K., told me. “But that doesn’t mean they’ve got a pathological process going on.”
In a paper published online in December 2021, Perkin and colleagues found that in a food-allergy trial, nearly three-quarters of the infants’ parents reported at least two symptoms that matched the iMAP guidelines’ “mild-moderate” non-IgE-mediated cow’s-milk-allergy symptoms, such as vomiting. But another study, whose authors included Perkin and Robert Boyle, a children’s-allergy specialist at Imperial College London, reviewed available evidence and found estimated that only about 1 percent of babies have a milk allergy that has been proved by what’s called a “food challenge,” in which a person is exposed to the allergen and their reactions are monitored.
That same study reported that as many as 14 percent of families believe their baby has a milk allergy. Another study by Boyle and colleagues showed that milk-allergy formula prescriptions increased 2.8-fold in England from 2007 to 2018. Researchers at the University of Rochester found similar trends stateside: Hypoallergenic-formula sales rose from 4.9 percent of formula sold in the U.S. in 2017 to 7.6 percent in 2019.
Perkin and Boyle suspect that the formula industry has influenced diagnosis guidelines. In their 2020 report, published in JAMA Pediatrics, they found that 81 percent of authors who had worked on various physicians’ guidelines for the condition—including several for iMAP’s 2013 guidance—reported a financial conflict of interest with formula manufacturers.
The formula industry also sends representatives and promotional materials to some pediatric clinics. One recent study found that about 85 percent of U.S. pediatricians surveyed reported a visit by a representative, some of whom sponsored meals with them.
Formula companies “like people getting the idea that whenever a baby cries, or does a runny poo, or anything,” it might be a milk allergy, Boyle told me.
In response to criticism that the guidelines have influenced the increase in specialized-formula sales, Fox, the lead author of the iMap guidelines, noted that the rise began in the early 2000s. One of the first diagnosis guidelines, meanwhile, was published in 2007. He also said that the symptoms listed in the iMAP guidelines are those outlined by the U.K.’s National Institute for Health and Care Excellence and the U.S.’s National Institute of Allergy and Infectious Diseases.
As for the conflicts of interest, Fox said: “We never made any money from this; there was never any money for the development of it. We’ve done this with best intentions. We absolutely recognize where that may not have turned out the way that we intended it; we have tried our best to address that.”
Following backlash over close ties between the formula industry and health-care professionals, including author conflicts of interest, iMAP updated its guidelines in 2019. The new version responded directly to criticism and said the guidelines received no direct industry funding, but it acknowledged “a potential risk of unconscious bias” related to research funding, educational grants, and consultant fees. The authors noted that the new guidelines had tried to mitigate such influence through independent patient input.
Read: What’s behind America’s shocking baby-formula shortage?
Fox also said he cut all formula ties in 2018, and led the British Society for Allergy & Clinical Immunology to do the same when he was president.
I reached out to the Infant Nutrition Council of America, an association of some of the largest U.S. manufacturers of infant formula, multiple times but did not receive any comment in response.
Though the guidelines have issues, Nigel Rollins, a pediatrician and researcher at the World Health Organization, told me, he sees the rise in diagnoses as driven by formula-industry marketing to parents, which can fuel the idea that fussiness or colic might be signs of a milk allergy. Parents then go to their pediatrician to talk about milk allergy, Rollins said, and “the family doctor isn’t actually well positioned to argue otherwise.”
Rollins led much of the research in the 2022 report from the WHO and UNICEF, which surveyed more than 8,500 pregnant and postpartum people in eight countries (not including the U.S.). Of those participants, 51 percent were exposed to aggressive formula-milk marketing, which the report states “represents one of the most underappreciated risks to infants and young children’s health.”
Amy Burris, a pediatric allergist and immunologist at the University of Rochester Medical Center, told me that there are many likely causes of overdiagnosis: “I don’t know that there’s one particular thing that stands out in my head as the reason it’s overdiagnosed.”
Some physicians rely on their own criteria, rather than the guidelines, to diagnose non-IgE milk allergy—for instance, conducting a test that detects microscopic blood in stool. But Burris and Rollins both pointed out that healthy infants, or infants who have recently had a virus or stomach bug, can have traces of blood in their stool too.
Martin, the allergy researcher at Massachusetts General Hospital, said the better way to confirm an infant dairy allergy is to reintroduce milk about a month after it has been eliminated: If the symptoms reappear, then the baby most likely has the allergy. The guidelines say to do this, but both Martin and Perkin told me that this almost never happens; parents can be reluctant to reintroduce a food if their baby seems better without it.
“I wish every physician followed the guidelines right now, until we write better guidelines, because, unequivocally, what folks are doing not following the guidelines is worse,” Martin said, adding that kids are on a restricted diet for a longer time than they should be.
Giving up potentially allergenic foods, including dairy, isn’t without consequences. “I think there’s a lot of potential risk in having moms unnecessarily avoid cow’s milk or other foods,” Burris said. “Also, you’re putting the breastfeeding relationship at risk.”
By the time Burris sees a baby, she said, the mother has in many cases already given up breastfeeding after a primary-care provider suggested a food allergy, and “at that point, it’s too late to restimulate the supply.” It also remains an open question whether allergens in breast milk actually trigger infant allergies. According to Perkin, the amount of cow’s-milk protein that enters breast milk is “tiny.”
For babies, Martin said, dietary elimination may affect sensitivity to other foods. She pointed to research indicating that early introduction of food allergens such as peanuts can reduce the likelihood of developing allergies.
Martin also said that some babies with a CMPA diagnosis may not have to give up milk entirely. She led a 2020 study suggesting that even when parents don’t elect to make any dietary changes for babies with a non-IgE-mediated food-allergy diagnosis, they later report an improvement in their baby’s symptoms by taking other steps, such as acid suppression. But when parents do make changes to their baby’s diet, in Martin’s experience, if they later reintroduce milk, “the vast majority of them do fine,” she said. “I think some people would argue that maybe you had the wrong diagnosis initially. But I think the other possibility is that it’s the right diagnosis; it just turns around pretty fast.”
Still, many parents who give up dairy or switch to a hypoallergenic formula report an improvement in their baby’s symptoms. Arnold said her son’s symptoms improved when she eliminated dairy. But when he was about eight months old, they reintroduced the food group to his diet, and he had no issues.
Whether that’s because the cow’s-milk-protein allergy was short-lived or because his symptoms were due to something else is unclear. But Arnold sees moms self-diagnosing their baby with food allergies on social media, and believes that many are experiencing a placebo effect when they say their baby improves. “Nobody’s immune to that. Even me,” she said. “There’s absolutely a chance that that was the case with my baby.”
