My take on the most recent paper on cannabis and human milk, "Marijuana Use by Breastfeeding Mothers and Cannabinoid Concentrations in Breastmilk."

2018 has been relatively prolific for research on cannabis and human milk. The InfantRisk Center published a paper in April 2018 examining THC pharmacokinetics in human milk in 8 subjects from Colorado. I reviewed the data presented in Baker paper (and lecture) in a condensed version here and a longer version here.

This most recent paper on cannabis use and human milk in the August issue of Pediatrics comes from California and adds a fair bit of detailed information to the body of information.  You can find the full paper here.

 I begin this post with my summary of the paper and what I think are the most important take-away messages. At the end of the post, I integrate what we learn from this paper with the rest of the literature. You can find a downloadable Elephant Circle fact-sheet with a similar integration here.

Join me as we dive into

“Marijuana Use by Breastfeeding Mothers and Cannabinoid Concentrations in Breastmilk”

by K. Bertrand et al. in the journal Pediatrics, August 2018

 The samples:

The milk samples came from the Mommy’s Milk Human Milk Research Biorepository.  This repository “strives to understand the numerous benefits human milk offers at a molecular level, and use these findings to improve the health and development of all children.”

The biorepository folks go on to say, “This [biorepository] demonstrates our place at the forefront of human milk and lactation discoveries. We are making it easier for scientists to perform research with standardized, sterile and indexed human breast milk samples. We are the organizing entity, but we need inquisitive researchers and the open-handedness of lactating moms. With your help, we can all demonstrate our commitment to the future improvements in children’s health and development.”



This particular research was grant funded by NIH and Gerber.


  1. A subset of samples from the repository were used in this study

  2. The samples selected for inclusion in this paper were samples from folks who reported cannabis use in the last 14 days

  3. 88% reported at least daily use (This is important to remember!)

  4. 64% used inhalation as their primary consumption method

A few important technical notes:

  1. The limit of detection for the quantitative analysis: > 1ng/mL (as compared to > 5ng/mL in Baker, 2018).

  2. This study examined single milk samples from individuals.  Unlike Baker 2018, there was no dosing with cannabis to determine THC levels.  The time since last consumption was self-reported and these data were used to evaluate (and demonstrate) a relationship between time of last consumption and THC levels.

  3. This research was looking to describe concentrations of the psychoactive cannabis molecules in addition to two other well-known cannabinoids found commonly in cannabis products. It is important to distinguish this from the scenario in which parents or infants are tested for exposure to cannabis. IN this detection situation, THC is not the molecule being measured.  Instead, COOH-THC is the molecule used in the detection of the exposure to cannabis.  This research did not look for COOH-THC.  For more information on the molecules involved in the metabolism and detection of cannabis in humans, please check out our Molecules 101 fact sheet.


  1. 20/54 samples did not show any detectable THC

  2. When THC was detected (n = 34), the median milk THC concentration = 9.47 ng/mL (range = 1.01 – 323)

  3. 11-OH-THC was detected in 5 samples

  4. CBD was detected in 5 samples 

  5. CBN was not detected in any samples

  6. In 76.5% of the samples with detectable THC, exclusive inhalation was the method of consumption (vs 36.8% in the no THC samples)

  7. Calculated a ½ life of ~ 27 hours

  8. The longest duration between last use and THC detection was 6 days. This was seen in only one sample.

  9. Only one sample had all three molecules: THC, 11-OH-THC and CBD

  10. This sample also had the highest level of THC detected in this study = 323 ng/mL

  11. The sample with the highest CBD had no detectable THC (11-OH-THC was detected)

  12. The cohort with no detectable THC had fewer daily users and a longer average time since the last use (53h vs 24h).  This suggests that daily use increases THC levels in human milk and that time allows for the metabolism and subsequent decrease of THC levels.

  13. See Figure 1 below for a visual of the relationship between time and THC levels.  As time since last consumption increases, THC levels decrease. (Note: This paper used log concentrations for their analysis because the absolute amounts of THC were very low. This mathematical transformation allowed for numbers that were easier to analyze statistically.)

Bertrand figure 1.png

Incorporating 2018 data with the pre-existing literature. What can we say we know?

  1. Nursing infants are exposed to an Relative Infant Dose of 2.5 (Baker, 2018), or, as stated in Bertrand 2018, “the estimated infant dose ingested would be 1000 times lower than the adult dose.”  Remember, a baby is exposed to THC orally. The bioavailability of THC when consumed by mouth is between 1-5%.

  2. There appears to be a curve for THC metabolism in human milk

    1. The shape of the metabolic curve is quite consistent between individuals

    2. The THC concentration at the peak of this curve varies dramatically between individuals

    3. Peak THC levels may occur between 60 and 120 minutes after use

  3. The half-life of THC in human milk is ~ 1d in daily users

  4. THC detection and metabolism in human milk appears to vary dramatically between daily and more occasional users

  5. Clinical advice should distinguish between these two types of users in both perspective and advice

So there you have it! My thoughts on the state of the science on cannabis and human milk.

I believe that 2018 provided us with some compelling data that can be used in personalized, clinical, conversations about cannabis and lactation. Whether we stand in the role of parent, practitioner or both, holding space for careful consideration of what we know, what we don’t know, and what we believe is essential. Centering these ideas help keep the conversation curious and empathetic - creating space for a non-judgemental risk/benefit analysis of cannabis use during breastfeeding In my view, that is the ideal conversation for this tricky topic.