MSD Manual

Please confirm that you are a health care professional

Commentary: Cow's Milk Exposure in Infancy and Recurrent Wheezing

12/14/2020 James Jeffrey Malatack, MD, Professor of Pediatrics, Thomas Jefferson University’s Sidney Kimmel College of Medicine

Human allergy to cow’s milk has been a topic of interest to medical professionals for many decades. Cow’s milk is primarily consumed by infants and children, and cow’s milk sensitization likely occurs in infancy, consequently, pediatrics has had a particular focus in understanding the mechanisms and the impact of cow’s milk protein as a cause of human allergic disease.

Hiroshi Tachimoto and colleagues weigh into this discussion with a recent publication (1). It is a laudable effort since prior studies have both implicated and dismissed cow’s milk as causal in human allergic disease and wheezing. The study was performed at a single Tokyo hospital as a randomized but unblinded (the study participants were infants but the parents were unblinded) study. The newborns studied were selected from families in which the risk of allergy was above the mean and then randomized into two equal groups to receive either breast feeding with or without supplemental elemental formula or breast feeding with cow’s milk supplemental feeding during the first 3 days of life. The participants were then assessed at five months and at two years of life by physicians blinded to the study design for the presence allergic disease including wheezing. Those patients who had clinical or laboratory evidence of allergic disease at two years were followed until age six. About half the patients met the criteria for the longer follow up.

The study, which was ultimately made up of 151 newborns in each of the two study limbs, found a significant difference in the outcome between the two groups. The infants who received breast feeding supplemented with cow’s milk in the first three days of life had a recurrent wheezing/asthma rate of 17.9% (27 newborns) in follow-up while those who avoided cow’s milk in those first three days of life (breast feeding with or without elemental formula supplementation) had a follow-up recurrent wheezing/asthma rate of 9.9% (15 newborns). The newborns who received supplemental cow’s milk were to continue this feeding regimen until 5 months while the group receiving breast feeding with or without elemental formula supplementation change diet. This latter group appears to have self-randomized to either quickly add cow’s milk supplementation to their diet before 14 days, to add cow’s milk to the diet after 14 days, or to maintain breast feeding with or without supplemental elemental formula and no cow’s milk for 5 months.

The authors believe that they controlled for all variables that might impact the study outcome, leaving exposure to cow’s milk as the only randomization that occurred during the first days of life. It is this presumption of variable control that warrants a closer look. Variable control along with a number of potential confounders may have impacted the study or its interpretation:

1. The failure of blinding the participants (actually, the participant’s parents) may have added error in the randomization. It should be noted that blinding the parents is not possible as the study was conceived. Elemental formula is readily recognized by differing appearance and odor in comparison to cow’s milk such that the parents, even if initially blinded, would rapidly appreciate what limb of the study their child was in. One needs to wonder whether in these families with known allergy, would parents, aware that their child did not get cow’s milk, be more likely to avoid cow’s milk in the future. When parents in the first five months of life either added cow’s milk before 14 days or after 14 days or maintained the cow’s milk-free regimen until 5 months, how would this knowledge have influenced the parents’ behavior after 5 months? How would this information impact the mothers’ own diet? Conversely, parents in the group continuing cow’s milk supplementation for five months who realized that their infants were already sensitized to cow’s milk, would mothers be more likely to continue cow’s milk after that and how would it impact on their own diet? How likely would they be to accept other potential food allergens as infant feedings?

2. There is no discussion of quantity of cow’s milk, and quantity may be important. If, at 14 days, parents decided to stop using elemental formula supplementation in favor of cow’s milk supplementation, but the breast milk volume was such that no supplement was needed, would any supplementation be given? If so, how much would be offered? If offered to a sated infant, how much would be consumed? This quantitative information may be critical.

3. The patient population was drawn from patients that had background information suggesting risk of atopy. The author explains that they came from an affluent area of Tokyo with a very homogeneous ethnicity that by the cleanliness hypothesis may be more likely to be sensitized by allergens. This study may not apply at all to the general population.

4. The authors admit only following the patients who met the study criteria of atopy at 2 years beyond that time until age six. Since patients who did not meet the atopy criteria at 2 years were not studied, one wonders how many “late wheezers” might be in the unstudied group which might negate the conclusions of the study. This is a very important consideration because the mean time to wheezing was longer than 2 years.

5. Could the failure to blind have influenced parents to change their behavior regarding other potential sensitizing agents?

The study also assesses the effectiveness of the first three-day intervention of cow’s milk or no cow’s milk supplementation as impacted by 25(OH)D levels. Levels of 25(OH)D were assessed at 5 months and at 2 years. Infants with a higher vitamin D level (> 29 ng/mL) at 5 months had a more dramatic reduction in the risk of recurrent wheezing/asthma in the no cow’s milk group than those infants who had a lower level for 25(OH)D (< 29 ng/mL) in which no beneficial effect of the absence of cow’s milk exposure was seen. No vitamin D effect was seen at two years.

IgE was assessed at 5 months and at 2 years as well and, while no effect of the IgE level was observed at 5 months, at two years those infants with the highest IgE levels had less wheezing if they had not received cow’s milk supplements in the first three days of life. Because of the concerns raised above, the impact of either Vitamin D or IgE level as related to the first three days of life randomization is unclear.

The study is complex, and the article is written with internal contradictions that add to the difficulty of reading and understanding it. As an example, The Trial Design section’s second sentence notes that the two study limbs adhered to the feeding design until 5 months of age. Later, in the results section, the reader is told that the infants in the breast feeding with or without elemental formula study limb were further divided within the first five months of life into three groups, two of which had not “adhered to the feeding design until 5 months.”

What does one make of this interesting if convoluted work? If the data hold up and overcome the concerns raised above, the question is why is the first three days of life so critical? Perhaps the newborn’s frequent but normal gastrointestinal reflux may be associated with microaspiration and direct sensitization of the lung to the allergen which sets up the process of recurrent wheezing. I believe, in the end, there are far too many concerns with the study to allow it to impact current practices, but the question raised by Tachimoto and colleagues about whether there is something very special about the first three days of life and sensitization to milk or for that matter other food allergens is intriguing and deserves more study.


Tachimoto H, Imanari E, Mezawa M, et al: effect of avoiding cow's milk formula at birth on prevention of asthma or recurrent wheeze among young children: Extended follow-upfrom the ABC randomized clinical trial. JAMA Netw Open Oct 1; 3(10):e2018534, 2020. doi: 10.1001/jamanetworkopen.2020.18534