Cow's milk protein allergy (CMPA) is an abnormal immunological response(s) to cow's milk protein(s) [CMP(s)] that commonly occurs in infants and young children (1). Since no single laboratory test is diagnostic, the diagnosis still has to be clinical, based on indirect milk elimination and provocation procedures, and a care provider's awareness becomes extremely important in making it diagnosed. Before 1950, CMPA was rarely diagnosed. Since 1970, this condition has been well documented. Now, it is estimated to occur in 1.8–17% of formula-fed infants and 0.5% breastfed infants (2, 3). For CMPA to occur, an individual must be first exposed and subsequently develop an immunological response to cow's milk allergen, and this process will take weeks or months to occur. As a result, CMPA most commonly occurs within weeks or months after birth. Importantly, cow's milk allergens are not only transported orally and through a mother's breastmilk into infants' circulation and induce CMPA; earlier studies suggest that they can also pass through the placenta and amniotic fluid to sensitize fetuses (4, 5) and cause allergy (6–9). For example, the proliferative responses of cord blood lymphocytes to cow's milk allergens such as α-lactalbumin, ß-lactoglobulin, and α-casein were found in full-term neonates (7–9). This means that CMPA can also happen in neonate infants even before their first postnatal feeds occur. However, while the former is now well known to care providers, the latter is often overlooked, particularly when this early-onset CMPA presents with atypical symptoms in the absence of hematochezia as in the vignettes of this communication. In the latter case, most pediatricians including pediatric gastroenterologists neither recognize nor are they even aware of it. As a result, many atypical cases of neonate infants with early-onset CMPA were either left undiagnosed, resulting in a delay in initiating the appropriate treatment for this very treatable condition, or misdiagnosed with other diseases, leading to many unnecessary workups. Since this early-onset CMPA is not uncommon, in order to raise an awareness and improve the clinical care of this unique clinical entity, we feel that it is necessary to report it again to pediatric communities.

In this communication, we report three cases of infants whose mothers persistently consumed dairy products during pregnancy, who then developed a severely constipated bowel, both pre- and postnatally. Their passages of meconium and stools were significantly delayed. They were so severely constipated that it was intractable to conventional laxative therapies. Interestingly, when milk-based formula or breastmilk was withheld from them, constipation resolved; when milk was reintroduced to them, constipation recurred; when their diet was switched to an extensively hydrolyzed or an amino acid-based infant formula, constipation resolved again. Based on this, we diagnose that these infants have CMPA that developed through prenatal exposure and is now evident through perinatal life.

We present three unconventional cases of CMPA that presented at birth with a delayed passage of meconium/stool. Table 1 summarizes the main demographics and clinical symptoms of the three cases. Figure 1 summarizes the relationship between the exposure/elimination of CMP and the expression/resolution of constipation in case 1 (Figure 1A), case 2 (Figure 1B), and case 3 (Figure 1C), establishing the diagnosis of CMPA in these patients. In these three cases, the elimination of CMP was done through an extensively hydrolyzed or amino acid-based formula; however an alternative would be to eliminate CMP from maternal diet and thus from the breastmilk, allowing families to continue to breastfeed. In our cases, the families chose not to pursue this option. The infants' delayed passage of meconium/stool is not simply due to hard stool as the problem had remained unsolved when stool softeners were applied. Instead, their primary pathophysiology seems to be unable to move their bowel as in those with Hirschsprung disease. None of them had spontaneous bowel movement, and all required rectal stimulation for bowel movement. The facts that these infants did not respond to regular formulas but responded to an amino acid-based formula (cases 1 and 3), extensively hydrolyzed formula (case 2), or milk avoidance (case 1) suggest that the culprit is not the fats, carbohydrates, or other non-protein components in formulas but proteins. Indeed, symptoms remitted when CMPs were withheld (Case 1), recurred when CMPs were reintroduced (cases 1 and 3), and resolved again when CMPs were eliminated (cases 1–3), thus establishing the relationship between CMP exposure and the development of the disease.

Although our cases have established an association between CMP exposure and the development of slow-transit enteropathy, the mechanism of this association is unclear. Clinically, CMP can cause a slow-transit bowel at least in two ways: (1) nonspecifically through CMP constipation and (2) specifically through CMPA. CMPA, rather than CMP constipation, was diagnosed because for the latter to occur, patients require a prolonged consumption of large amounts of CMP similar to the changes in stool consistency caused by other dietary alterations, and this is unlikely to occur in our patients, neither in utero nor at day one of their lives. In addition, for the latter (constipation) to resolve, patients require the withdrawal of CMP for a prolonged period of time of at least 2 weeks, not 2 or 3 days as in our cases. Finally, infants with CMPA will usually also experience symptoms other than just constipation. In keeping with this, our patients had vomiting and a failure to thrive besides constipation, all resolved upon milk protein elimination, consistent with the diagnosis of CMPA.

For infants with a delayed passage of meconium and stool in the newborn period, we must consider Hirschsprung disease, the most frequently encountered congenital disorder of intestinal motility. However, the presence of rectal ganglion cells and resolution of the symptom in response to CMP elimination in our cases clearly speak against the diagnosis of this disorder. Other differential diagnoses include anorectal dyschezia, particularly the acquired form. However, infants with dyschezia will often have the 5–10 min period of fussing and red-contorted face straining prior to passing stool. They do so because their efforts to pass stool are uncoordinated due to a defective reflex for defecation. In addition, infants with dyschezia do not present with a delayed passage of meconium; they will not have gone 1 week without passing stool; they do not have to use a glycerin suppository or rectal stimulation to pass stool. Finally, infants with dyschezia should not respond to milk protein elimination.

Multiple mechanisms for CMPA have been suggested (3, 10–12). These include an immunoglobulin E (IgE)-mediated immediate hypersensitivity reaction through the mast cell release of histamine and a non-IgE-mediated delayed hypersensitivity reaction. The latter likely involves cytokine release from antigen-mediated T-cell stimulation or antigen-mediated immunoglobulin A (IgA) or immunoglobulin G (IgG) immune complex stimulation. Although the nature of reaction to CMPs in our patients is unknown, their symptoms and normal-appearing colorectal histology (case 3; not shown) did not suggest an IgE-mediated reaction. Therefore, the laboratory tests measuring CMP-specific IgE and skin test wheals are of limited diagnostic values and were not used in this study. In our rectal biopsy of the case 3 patient, no lamina propria/muscularis mucosa eosinophilia, lymphoid nodular hyperplasia, cryptitis, neutrophilic crypt abscess, glandular distortion, or Paneth cell metaplasia nor increased mast cell infiltration was observed. A more likely explanation would be a non-IgE-mediated mechanism.

Figure 2 illustrates how the maternal CMP exposure leads to the development of neonatal constipation. Cow's milk allergen is vertically transmitted during pregnancy from a mother to a fetus via the placenta and amniotic fluid to cause CMPA (Figure 2A). When the CMPA involves the colon, constipation results (Figure 2B). Although a detailed pathway(s) leading allergy to constipation is yet to be understood, it appears that the allergenic response involves neuromodulation of the enteric nervous system (ENS) (13), resulting in increased resting anal sphincter pressure and causing difficulties in defecation (14). In keeping with this, our three cases presented in this study all required anorectal maneuvers (suppository enema/thermometer insertion) to probably relieve anal pressure in order to defecate. Also, chronic functional constipation, once thought idiopathic, has now been regarded as neuropathologic due to a complex system of abnormalities of the ENS of these patients (15), and perhaps CMPA is the cause, at least in part, of such abnormalities.

CMP allergens can be transported orally and through a mother's breastmilk into infants' circulation and induce CMPA in infants. This is well established. Research also suggests that cow's milk allergens can pass through the placenta and amniotic fluid (7) to sensitize fetuses (16, 17), including fetal gut (16), at as early as 22 weeks of gestation (18) and cause CMPA. The latter includes conventional allergic colitis (19) and unconventional constipation (present study). This is important. This means that pediatricians must consider CMPA even in 1-day-old neonate infants as opposed to the currently taught manifestation at a few weeks or months of age. Were these neonates with early- onset CMPA left unidentified or untreated, they will become a more chronic clinical condition during childhood or adulthood. There are reports that up to 78% of chronic functional constipation later in life are probably CMPA-related (14).

The limitations of this report must be considered. The first is that sensitization and onset in utero were not shown directly. Second, CMPA was diagnosed clinically, rather than by laboratory tests such as an allergen-specific lymphocyte stimulation test (20) or prolonged, two-stage double-blind, placebo-controlled food challenges (18). In the future, we could consider performing cord blood cow's milk sensitization and/or allergen-specific lymphocyte stimulation tests in order to establish their correlation with the subsequent development of allergic disease. These tests, like other food allergy panels (e.g., skin prick tests and specific IgE tests), have a high negative predictive value; however, a positive result only indicates sensitization but not necessarily an allergy; in order to diagnose food allergy, the most appropriate testing remains clinically using an oral food challenge and elimination diet as in these vignettes (21, 22).

Given the discovery of the CMP hypersensitivity in utero and recognition of CMPA at birth, pediatricians should consider this condition in their differential diagnoses, particularly in newborns with a delayed passage of meconium and/or early-onset constipation suspicious of Hirschsprung disease or infant dyschezia, both congenital and acquired. Figure 3 proposes a new management algorithm for early-onset constipation in neonates with suspected CMPA. Considering its wide morbidity and simplicity in the treatment, we suggest that a 2-week trial of CMP avoidance be applied to these neonates and infants first. This way, an unnecessary invasive workup for Hirschsprung disease and others can perhaps be avoided.

The original contributions presented in the study are included in the article/supplementary materials, further inquiries can be directed to the corresponding author.

Written informed consent was obtained from the minor(s)' legal guardian/next of kin for the publication of any potentially identifiable images or data included in this article.

SXC: conceptualized and designed the work. AM, AL, and SXC: collected and analyzed the data and interpreted the data. AM and AL: drafted the manuscript. AL, SC, and SXC: revised the manuscript. SXC: finalized this manuscript. The artwork of the Figures 2A,B were hand created by SC, and all authors approved the final version and agreed to be accountable for the content of this work.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.