It is not infrequent in the neonatal unit to be faced with an infant who has rectal bleeding. In some places the tendency has been to assume that the infant has cows milk protein intolerance (even among those who are breast fed!) and place the infants on a hydrolysed protein diet or even on an amino acid based diet. What is the evidence for the diagnosis and treatment of cows milk protein intolerance in the preterm, or term newborn?
There are very few reports of cow’s milk protein intolerance in the preterm infant. Even those that have reported have not usually confirmed the diagnosis, but how to do that? There are two methods which have been reported as being reliable, although their reliability is uncertain as we do not have a gold-standard. The most commonly used is response to exclusion of cow’s milk protein, but this must be followed by a challenge; the re-introduction of cow’s milk to see if the signs return and, if they do, the re-exclusion to make sure they go away again. If you don’t do the challenge then any case of transient clinical signs will be mis-diagnosed as cow’s milk protein intolerance.
One study in older infants found 2 things, (Xanthakos SA, Schwimmer JB, Melin-Aldana H, Rothenberg ME, Witte DP, Cohen MB: Prevalence and Outcome of Allergic Colitis in Healthy Infants with Rectal Bleeding: A Prospective Cohort Study. Journal of Pediatric Gastroenterology & Nutrition 2005, 41(1):16-22) firstly, most pediatric gastroenterologists don’t follow their own societies guidelines, and don’t do a challenge if a child presents with rectal bleeding that improves after exclusion. Which means that many infants end up with a diagnosis that has not been confirmed. Secondly, they also showed that using the other method of diagnosis, that is doing a rectal biopsy, less than half of the infants with rectal bleeding actually had evidence of allergic colitis. Another study, (Arvola T, Ruuska T, Keränen J, Hyöty H, Salminen S, Isolauri E: Rectal Bleeding in Infancy: Clinical, Allergological, and Microbiological Examination. Pediatrics 2006, 117(4):e760-e768) also in infants, noted that rectal bleeding was benign and self-limited, and despite rectal biopsies etc, were able to confirm the diagnosis in only 7 of 40 consecutive cases of rectal bleeding. Interestingly they noted a low number of bifidobacteria in the stools of the infants, suggesting that an imbalance in the microbiota may play a rôle.
Even a test as invasive as a biopsy is not certain, the findings of patchy eosinophilic colitis is supposed to be diagnostic, but the cause of the allergic colitis cannot be determined from the biopsy. As it may be found in infants who are exclusively breast-fed, even among those whose mother is on a cows milk free diet, the diagnosis of human milk protein allergy has been created. Which I find very doubtful! What is the evidence that human milk may contain cows milk proteins if the mother consumes them? Well the most common protein implicated is bovine beta lactoglobulin, in one older study one third of mothers consuming cows milk based products had detectable traces of bovine BLG in their breast milk. So it is possible that these trace amounts might trigger symptoms. That study also found that breast fed babies with signs of cows milk protein intolerance had all received supplements with a cows milk based formula in the neonatal period.
For full term infants does it matter if they have a diagnosis of cows milk protein intolerance which is not correct? It is actually pretty straight forward to find nutritionally adequate hydrolyzed protein formulae for the full term infant. (Not soy protein based formulae; a lot of infants proven to be intolerant of bovine proteins are also intolerant of soy proteins). On the other hand they are more expensive, and not all kids will tolerate even them; but amino acid based mixtures are much more expensive. Also bovine protein exclusion diets are a real pain for mothers who are breast feeding, and some of them are inappropriately discouraged from breast feeding. Having a diagnostic label that is incorrect is not the best situation, but some parents are happy to buy hydrolysed protein formulae, accept that the diagnosis is not certain, and live with that uncertainty for a few months, (more than half of proven cows milk protein intolerance disappears in the first year).
For preterm infants the situation is different, there are no artificial milk formulae which are nutritionally adequate for the preterm and yet are free of bovine proteins. So a diagnosis in the preterm has other implications for their nutrition. In addition all approved and appropriate fortifiers which are required to add protein, calories and minerals to human milk are based on cow’s milk.
So going back to the diagnosis of bloody stools, and now looking at the neonatal period, a recent study using new microarray methods found that newborn infants with rectal bleeding did not have findings consistent with allergy. (Ohtsuka Y, Jimbo K, Inage E, Mori M, Yamakawa Y, Aoyagi Y, Suzuki M, Kudo T, Suzuki R, Shimizu T: Microarray analysis of mucosal biopsy specimens in neonates with rectal bleeding: Is it really an allergic disease? Journal of Allergy and Clinical Immunology 2012, 129(6):1676-1678.) Other studies have found weird viruses in the stools of such babies (Chappé C, Minjolle S, Dabadie A, Morel L, Colimon R, Pladys P: Astrovirus and digestive disorders in neonatal units. Acta Paediatrica 2012, 101(5):e208-e212). I have certainly seen rectal bleeding associated with rotavirus in more than one preterm infant.
Rectal bleeding may also occur in the setting of strictures, either congenital strictures or post enterocolitis. This is probably the most common cause of rectal bleeding in the preterm after NEC, and is often misdiagnosed, often as cows milk protein intolerance.
In fact there is evidence that the pre-term period is a time when exposure to foreign proteins is more likely to lead to tolerance than to allergy. There is certainly no evidence of an increase in allergic diseases among preterm infants. There are a couple of large cohort studies showing a similar, and relatively low rate of allergic diseases, and specifically cows milk protein intolerance, among preterm infants to term infants. (Zachariassen G, Faerk J, Esberg BH, Fenger-Gron J, Mortensen S, Christesen HT, Halken S: Allergic diseases among very preterm infants according to nutrition after hospital discharge. Pediatric allergy and immunology 2011, 22(5):515-520.)
In general then a baby in the NICU who has rectal bleeding is more likely to have something other than cows milk protein intolerance, and it becomes even less likely to be that diagnosis if the infant is preterm, and very much less likely if they are breast fed.
Neonatal Research 