On Marijuana and Breastfeeding: A New Study

Determining the safety of any new medication is difficult. Generally, willing volunteers are given doses of experimental medications and monitored by medical professionals in a clinical setting. This, of course, after extensive testing in animal models first. While this approach generally works for the population in general, certain groups, like pregnant mothers, are excluded from research of this type. It’s easy to understand why; drugs that may be fine for adults can have serious, permanent effects on the unborn and infants. Because of this, only 10% of drugs approved by the FDA since 1980 have enough data to make recommendations for use by pregnant mothers. The safest course of action, then, is to discontinue any medications unless directed by a doctor. 

The effects of drinking or smoking tobacco during pregnancy are, sadly, well known. Even if a mother is able to abstain during pregnancy, infants can still be exposed to harmful chemicals via breastmilk if the mother chooses to start up again. Studies on alcohol and breastfeeding have shown that alcohol appears in breastmilk between 30-60 minutes after drinking, at concentrations similar to that found in the blood. In decades past, some doctors actually recommended alcohol for nursing mothers on the mistaken belief that it would stimulate milk production; this has since been proven false. In fact, alcohol can shorten the lactation period and reduce levels of oxytocin in the milk, increasing the odds of a child rejecting the milk. Today, most doctors recommend a “better safe than sorry” policy when it comes to drinking and breastfeeding.

When it comes to cannabis use and lactation, almost nothing is known. This is due in large part to the illegal status of cannabis in most states. New mothers who use cannabis are very unlikely to mention it to a doctor for fear of legal trouble and potential intervention by the state. However, with more states legalizing cannabis for medical and recreational use, we need a better understanding of the risks new mothers take when choosing to use cannabis while lactating. A new study in the journal Obstetrics and Gynecology tackles this issue, beginning with the most basic question: Does THC transfer into breast milk after smoking cannabis?


From a technical standpoint, answering this question is relatively straightforward. However, it would be completely unethical and unconscionable to ask lactating mothers to smoke cannabis. Therefore, the researchers sought out mothers who were already taking this risk upon themselves. They started by advertising for participants in the Denver, CO region. Every precaution was taken to ensure the mother’s absolute anonymity, to avoid exposing them to legal risks. They also did not offer any sort of monetary compensation to participants. The authors had hoped to recruit fifty subjects; they managed to recruit only eight.

Once the subjects had enrolled, they were given instructions for how to participate. The mothers were instructed to purchase a particular type of collection bottle from a known distributor. The researchers had previously determined that the plastic in these containers would not interfere with their measurements. Subjects were also directed to purchase a particular type of cannabis, all from the same dispensary, ensuring that the concentration of THC was consistent for all subjects. They were provided with a clean glass pipe for smoking. The subjects were asked to abstain from smoking for at least 24hours, then collect milk for a baseline comparison. Then, they were to smoke 0.1g of the purchased cannabis, and collect samples at 20 minutes, and at 1,2, and 4 hours.

The samples were then frozen and mailed to the research laboratory in Texas. Each sample was prepared and tested for the presence of THC and two of its most common metabolites. For six of the samples, no THC was detectable at baseline. At all time points after smoking, THC was present, with the highest concentrations detected at the one hour timepoint. There was a large range of values for each measurement, but on average, 2.5% of the smoked THC was transferred to breast milk.


From these results, it is clear that some level of smoked THC finds its way into breast milk. While this information is incredibly useful, it does have its limitations. For example, the researchers calculated the amount of transferred THC based on the initial dose. However, smoking habits and physiological differences affect how much THC makes its way into the blood, which in turn carries the THC to the mammary glands. Therefore, simultaneous blood sampling would have provided a better value for comparison. Because everything was done remotely, there was no way to verify cannabis use, the timing/correct labelling of each sample collection, or initial abstinence.

Regardless, this is an important first step in approaching this issue. There is a lot that is still unknown. For example, THC accumulates in the body following repeated, chronic use; does this lead to higher THC accumulation in breast milk? Does the method of ingestion (smoking vs vaping vs eating) make a difference? Do other cannabinoids, such as CBD, get transferred as well? Finally, it is currently unknown what the effects of THC exposure are on infants and newborns. Therefore, the best assumption is that no level of THC is safe for babies, and that lactating mothers should abstain from using cannabis.

2 thoughts on “On Marijuana and Breastfeeding: A New Study”

  1. This study raises important ethical questions for me as the researchers are from Texas and used a Texas IRB to approve the study even though the study was conducted using women in Colorado. Colorado IRB’s would not have approved this study because researchers in Colorado are directly barred from dosing (outside of NIDA) and from using any product purchased from a state producer. Unless an IRB in Colorado has ceded their oversight to Texas for this research then this study, while important and interesting, was done in an unethical manner.

    • Hi Greg, thanks for the comment. Let me start by saying that I’m not an ELSI expert, so my opinion on this is little more than educated guessing. My assumption is that they skirted this requirement by working completely anonymously. If they collected zero identifiable information, then it technically doesn’t count as “human subjects research”, and wouldn’t need IRB approval. I have to assume, then, that they did seek approval in order to step around the rules in the most ethical way possible. Perhaps they did speak to a Colorado IRB, and simply didn’t mention it in the manuscript.

      For the regular folks reading this:
      IRB = Institutional Review Board. They oversee all research on human subjects to ensure ethical and legal standards.
      ELSI = Ethical, Legal, and Social Implications. There’s a whole area of research devoted to these issues that helps determine policy.
      NIDA = National Institute on Drug Abuse. They fund and oversee research on illicit drug use and its impacts.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.