The Risks of Codeine to Breastfed Infants- Facts and Fallacies

Motherisk Int J 2020;1:16


David Chitayat MD, FRCPC, FACCG (1) and Gideon Koren MD, FRCPC, FACMT (2)

The Prenatal Diagnosis and Medical  Genetics Program, Department of Obstetrics and Gynecology, Mount Sinai Hospital, The University of Toronto, ON, Canada (1); Motherisk International Program and Ariel University, Ariel, Israel (2)




It has been now almost 15 years since the Toronto Motherisk program described the neonatal risks of codeine exposure through breastmilk which has led the FDA and most other regulatory agencies to publish warnings and lead to contraindication of the drug during breastfeeding. I read now that Drs. Zipursky and Juurlink cast doubts whether babies can be at risk from exposure to codeine-morphine through breastfeeding (1). We are very interested in your take on this.


In 2006 we reported in Lancet the death of an exclusively- breastfed infant from morphine poisoning after his mother, who was an ultra- rapid metabolizer of codeine to morphine using Tylenol 3 for pain after delivery. He had very high post-mortem levels of morphine and the highest breastmilk level of morphine ever reported in the literature (2). We believe that the claim by Zipursky and Juurlink that this is an improbable cause of death  stems from fundamental flaws in their understanding of perinatal toxicology and their lack of recognition that the baby is not a small adult when it comes to morphine, codeine and paracetamol toxicology. Similar toxicities have been documented in many other infants, and we believe that hadn’t the FDA publish their warning, many more deaths would have occurred.


In 2006 we described the death of an exclusively- breastfed infant from morphine poisoning after his mother, who was an ultra- rapid metabolizer of codeine to morphine by Cytochrome P450 2D6 used Tylenol 3 (acetaminophen-codeine combination) for severe pain during delivery. He had very high post-mortem levels of morphine and the highest level of morphine ever described in the literature in neonates (2). Since then we have shown similar risks in ultra- rapid metabolizing babies given codeine after tonsillectomies (3). Zipursky and Juurlink (Z-J) tried to calculate the dose the baby was exposed to from the single high post-mortem level of morphine (80ng/ml).

The attempt of  Z -J to extrapolate a single serum level of morphine into a dose is inappropriate: To calculate a dose that results in a given serum concentration one needs to know a) the clearance rate in the specific individual, b) whether this constituted a steady state level, and c) whether there is post-mortem redistribution of the drug. Without having these parameters Z-J  declare that it is highly implausible that the breastfeeding caused the demise of this baby.

Luckily, other scientists, who are world experts in pharmacokinetics and PK modeling, took the time to examine this issue and came to sweepingly different conclusions than Z-J:

  1. In 2009 Willmann, Edgington and colleauges, a group of system biologists and pharmacokinteticists, investigated the risk of codeine poisoning to breast-fed neonates using coupled physiologically- based pharmacokinetic models for the mother and child (4). Neonatal morphine plasma concentrations were simulated for various combinations of cytochrome P450 2D6 (CYP2D6) genotype and morphine clearance, assuming a typical breast-feeding schedule and maternal codeine doses of <2.5 mg/kg/day. These simulations demonstrated that the mother’s codeine and morphine clearances and the neonate’s morphine clearance are the most critical determinants of morphine accumulation in the neonates. The cumulative doses ingested by the neonate over 14 days were 0.38 mg/kg codeine and 0.17 mg/kg morphine. Given the added effect of low neonatal elimination capacity for morphine, Willmann et al found that potentially toxic morphine plasma concentrations can be reached within 4 days in the neonate after repeated codeine dosing to the mother. Importantly, neonates of mothers with the ultrarapid CYP2D6 and neonates of mothers who are extensive metabolizers have comparable risks of opioid poisoning (4).


  1. In 2012, the Genome BC project conducted another study to predict codeine toxicity in infants and their mothers using a novel combination of maternal genetic markers( 5). They used symptoms of CNS depression in 111 breastfeeding mothers using codeine, and their infantsA genetic model combining the maternal genotypes in CYP2D6 and ABCB1 was significantly associated with the adverse outcomes in infants (OR 2.74, P=0.0002), and mothers (p=0.0005). A novel combination of genetic and clinical factors predicted 87% of the infant and maternal CNS depression cases with a sensitivity of 80% and specificity of 87% (5).

The serum concentration of morphine measured in the neonate was several- fold higher than previously reported, and it is not compatible with spontaneous breathing.  Similarly, breastmilk level of morphine was 5 fold higher than what had been previously described (2). They were probably much higher before the ultra- rapid metabolizing mother halved her dose due to her own adverse CNS effects. Z-J seem to disregard the obvious: neonatal opioid death in a breastfed infant of a CYP2D6 ultra rapid metabolizer mother, with a documented high concentration of morphine in milk.


Z-J also claimed that the codeine level measured was “too high” to stem from breastfeeding only. The same can be said about paracetamol. Unfortunately they ignore the proven dramatic post-mortem re-distribution of codeine and paracetamol (6-7). Such redistribution makes it futile to try to estimate doses.

Most critically, the prediction studies presented above (4-5) are consistent with documented reports of neonates exhibiting CNS depression and apnea following breastfeeding while on codeine (e.g.5, 8). Z-J also ignore the fact that neonates are much more sensitive than older children and adults to the toxicity of morphine due to higher permeability through the blood brain barrier (BBB). In a 2015 study completed by our  team, we have shown that the intensity of P Glycoprotein at the BBB, which is responsible for pumping morphine from the brain back to the blood, is much lower in the neonate, till around 6 months post- partum (9). This well explains the sensitivity of neonates to codeine during breastfeeding, because much more morphine penetrates the brain (9).

The  effects of codeine in a subset of breastfed infants and the fact that ethically one cannot conduct  prospective studies on this matter,  has led us to use the best available methods to establish causality in these observational studies. We conducted a systematic review and employed the Naranjo Adverse Reaction Probability Scale (NADRPS). In addition to our case reports, several abstracts and full length studies have described adverse drug reactions in infants exposed to codeine through breastmilk, Specific adverse drug effects were described as unexplained episodes of CNS depression,  apnea, bradycardia and cyanosis in suckling infants. Using the NADRPS, codeine was found to be a “definite” cause of CNS depression in breastfed infants (10).


If the baby in our case was not poisoned through breast milk,  then this had to be a deliberate poisoning because the baby was exclusively breastfed. That means that he was given orally codeine or morphine and paracetamol. We considered with the coroner the possibility of infanticide and it was utterly ruled out based on reports of the physicians who managed and knew the mother. Since this tragedy, the mother has had 3 more children.

The above exchange could have been perceived as differing interpretations of a toxicological event between J-Z and ourselves. However, in their paper they allude to “the previously unreported finding of a markedly elevated codeine concentration in post-mortem drug” (1). In reality this codeine concentration was reported by us in 2007(11), was brought up in a letter to Lancet by Bateman et al (12) and was fully addressed in our response. It is highly impossible that Z-J were not aware of these documents.


Suckling neonates are much more sensitive to the CNS effects of morphine, and in some of them it can be life-threatening and even lethal. Since mothers are not routinely genotyped for their CYP 2D6 status, all women should avoid codeine while breastfeeding.




  1. Zipursky J, Juurlink DN The implausibility of neonatal opioid toxicity from breastfeeding. Clin Pharmacol Ther 2020; May 7
  2. Koren G , Cairns J ,  Chitayat D, Gaedigk A,  Leeder SJ. Lancet 2006 Aug 19;368(9536):704.
  3. Ciszkowski C, Madadi P, Phillips MS, Lauwers AE, Koren G. Codeine, Ultrarapid-Metabolism Genotype, and Postoperative Death. N Engl J Med 2009; 361:827-828.
  4. S Willmann , A N Edginton, K Coboeken, G Ahr, J Lippert. Risk to the Breast-Fed Neonate From Codeine Treatment to the Mother: A Quantitative Mechanistic Modeling Study. Clin Pharmacol Ther 2009 Dec;86(6):634-43.
  5. Sistonen J, Madadi P, Ross CJ, Yazdanpanah M, Lee JW, Landsmeer ML, Nauta M, Carleton BC, Koren G, Hayden MR. Prediction of codeine toxicity in infants and their mothers using a novel combination of maternal genetic markers. Clin Pharmacol Ther. 2012 Apr;91(4):692-9.
  6. Yonemitsu K, Pounder DJ. Postmortem toxicokinetics of co proxamol Int J Legal Med 1992;104:347-53
  7. Tolliver SS, Lee Hearn W, Furton KG. Evaluating the relationship between post-mortem and antemortem morphine and codeine concentrations in whole blood. J Analyt Toxicol 2010;34:491-497
  8. Madadi P, Ross CJD,Hayden MR, Carleton BC,  Gaedigk A,  Leeder JS, Koren G. Pharmacogenetics of Neonatal Opioid Toxicity Following Maternal Use of Codeine During Breastfeeding: A Case–Control Study. Clin Pharmacol Ther 2009;85:31-35
  9. Lam L , Baello S , Iqbal M  , Kelly LE   , Shannon  PT, Chitayat D ,  Matthews SG  ,  Koren G . The Ontogeny of P-glycoprotein in the Developing Human Blood-Brain Barrier: Implication for Opioid Toxicity in Neonates.  Pediatr Res 2015 Oct;78(4):417-21.
  10. Madadi P, Shiraz F, Walter FG, Koren G: Establishing causality of CNS depression in breastfed infants following maternal codeine use.. Pediatr Drugs 2012;10:399-404
  11. Madadi P, Koren G, Cairns J, et al. Safety of codeine during breastfeeding. Can Family Phys 2007;53:33-35
  12. Bateman N, Eddelston M, Sandillans E: Codeine and breastfeeding. Lancet 2008; 372:625-626