ABOVE: © Lynn scurfield

We live in a medicated era. Recent data indicate that more than half of Americans are currently taking prescription drugs. Among pregnant women this number skyrockets to more than 80 percent. One of these women was a 24-year-old from California named Carol, whom I met and befriended through an online drug research forum. After weeks of debilitating morning sickness, persistent pain in her back and hips, and chronic anxiety about becoming a mother, Carol was taking a tranquilizer called alprazolam as needed, plus daily doses of acetaminophen and an anti-nausea drug called metoclopramide.

Carol felt uneasy using the medications. Like many Americans and an even greater proportion of Europeans, Carol (who asked that I not use her surname) favors home remedies over pharmaceutical treatments. “I’ll always choose a tea over a pill,” she says. And so, as she sought relief during her pregnancy, she turned...

In the summer of 2007, Carol was surrounded by people touting the wonders of cannabis as a panacea for diseases from depression to glaucoma and myriad ailments in between—including nausea, pain, and anxiety. Worried that her suboptimal diet and poor sleep could be affecting the development of her child, she considered using small amounts of cannabis instead of the multiple prescription medications suggested by her doctor.

Seventy percent of women in the United States believe that there is “slight or no risk of harm” in using can­nabis during pregnancy.

“Instead of two or three pills per day, I could have a little bit of pot in the morning,” she recalls thinking at the time, “and everyone agreed it was harmless.” Weighing the options, Carol decided that while she would prefer to consume neither prescription drugs nor cannabis, the latter was the lesser of two evils. “It seemed natural,” she says. Now, with a 10-year-old son who suffers from attention-deficit/hyperactivity disorder and behavioral problems at school, she wonders if the choice she made has come back to haunt her.

Epidemiologists are looking into the concerns of parents such as Carol. Various large-scale longitudinal research projects in both North America and Europe, ranging from several hundred to thousands of subjects, on cannabis use during pregnancy point to a number of potential consequences, including hyperactivity, in children. The problems start early—exposed infants are more likely than unexposed babies to have low birth weights and to spend time in neonatal intensive care. And troubles can last into adulthood. Higher rates of depression and drug abuse are among the health issues most commonly linked with maternal cannabis use.

Research that can address the causal nature of these links is still in its infancy, however. The status of cannabis as a Schedule I substance in the US makes it tough to get approval for experiments. For this and other reasons, research on the incorporation of the drug in Western medicine is relatively new as well. This means that conventional health practitioners receive little, if any, information that they can pass onto their patients, including those considering the use of cannabis during pregnancy.

It’s no wonder, then, that 70 percent of women in the United States believe that there is “slight or no risk of harm” in using cannabis during pregnancy. And about 4 percent of pregnant women in the US report using the drug during gestation, just like Carol. Of expectant moms between the ages of 18 and 25, this number is nearly 7.5 percent. With a growing national and global acceptance of marijuana use, research on the effects of cannabis exposure during pregnancy is more urgent now than ever.

Cannabis and the prenatal brain

Researchers identified cannabinoids as one group of pharmacologically active compounds in marijuana in 1940, but it was another half a century before they confirmed the existence of an endocannabinoid system (ECS). Over the course of about three years in the early 1990s, neuroscientists discovered the first cannabinoid receptor (CB1R) in mammals, cloned both the rat and human variants, and identified a second cannabinoid receptor (CB2R). Later, the first endogenous cannabinoid was identified and named anandamide after the Sanskrit word for bliss.

Researchers have since characterized a second primary endocannabinoid, 2-AG, known to bind these receptors, plus a host of additional endogenous molecules that interact in other ways with the ECS. Scientists have also identified the enzymes that synthesize and degrade these compounds along with a number of additional putative cannabinoid receptors.

Identified as the brain’s most widely expressed G-protein coupled receptor, CB1R acts to regulate body temperature, relay hunger signals, and process sensory input, with a hand in countless other physiological and cognitive states. (See “Your Body Is Teeming with Weed Receptors,” The Scientist, July/August 2017.) Cannabis contains at least 108 exogenous cannabinoids—including Δ9-THC, which binds to CB1R to cause the high associated with the drug—along with dozens of other pharmacologically active compounds, such as the terpenes and flavonoids. These interact with the ECS to cause an altered psychological state, and mediate myriad other effects the drug in both the brain and the body.

When marijuana is ingested or smoked during pregnancy, exogenous cannabinoids enter the blood and cross easily through the placental barrier due to their highly lipophilic nature. Pairing this ready availability with slow pharmacokinetics—active metabolites continue circulating for up to five days depending on dosage and frequency of use—fetal exposure to the active compounds in cannabis is both efficient and prolonged. It should therefore be expected that this exposure can profoundly influence the development of the ECS.

When marijuana is ingested or smoked during pregnancy, exogenous cannabinoids enter the blood and cross easily through the pla­cental barrier.

To date, the three largest longitudinal studies of the children of women who smoked marijuana once a week or more during their pregnancies have identified remarkably consistent outcomes during early development and through young adulthood. In infants, these include increased impulsivity, hyperactivity, and delinquent behaviors, as well as memory dysfunction and decreased IQ scores. During adolescence and early adulthood, fetal cannabis exposure has been linked to persistent reduction in memory and concentration, higher rates of drug use, and an increased incidence of hyperactivity, signs of depression, and psychotic and schizophrenic-like symptoms. These mental health issues are further evidenced by increased reports from both parents and schoolteachers of problematic behavior and delinquency in cannabis-exposed kids.

One of the three large-scale projects, the Ottawa Prenatal Prospective Study, which began following approximately 700 pregnant cannabis users in 1978, has tracked nearly 200 of the offspring from the neonatal period into adulthood, and has identified persistent effects ranging from changes in cortical function to higher rates of drug abuse even in maturity, compared with controls. In men, for example, regular use of marijuana as well as daily tobacco use are both more than twice as likely for those exposed to cannabis in the womb. Data on adults from the two other longitudinal studies—the Generation R study in the Netherlands that is currently tracking nearly 8,000 children and the Maternal Health Practices and Child Development Study at the Western Psychiatric Institute and Clinic in Pittsburgh that is following 580 teenagers—have yet to emerge. But with the persistent and significant changes already seen at multiple ages across nations, spanning economic and social strata, more links between human health and maternal cannabis use would not be surprising.

The nature of the relationship remains unclear, however. Moreover, quantifying marijuana intake is riddled with challenges—the potency of cannabis, as well as the ratio of various active cannabinoids that it contains, is extremely variable, for example—so researchers cannot yet say with confidence how doses of the drug influence these correlations. Now, researchers in the laboratory are linking observations concerning mental health with biological mechanisms involved in cannabis use to get a better handle on the risks of using the drug while pregnant.

Exogenous cannabinoids in an endogenous system

To root out the cellular and molecular mechanisms that might underlie the epidemiological patterns seen in humans exposed to cannabis in utero, researchers are turning to experiments with rodents. Researchers now know that the ECS plays a significant role in the development of the central nervous system, and perturbations to this system in developing mice and rats are associated with lasting disruptions to cell differentiation and neuronal migration, critical steps in the formation of a functional brain. CB1Rs are present in the human cerebrum by the first weeks of the second trimester, and many studies have shown that CB1R knockout mice show significant behavioral problems.

A Lifetime of Consequences?

Large-scale, longitudinal studies of humans whose mothers smoked marijuana once or more per week and experimental work on rodents exposed to cannabinoids in utero have yielded remarkably consistent intellectual and behavioral correlates of fetal exposure to this drug. Some exposed individuals exhibit deficits in memory, cognition, and measures of sociability. These aberrations appear during infancy and persist through adulthood and are tied to changes in the expression of multiple gene families, as well as more global measures of brain responsiveness and plasticity.  Researchers currently consider these perturbations to be mediated by changes to the endocannabinoid system caused by the active compounds in cannabis. 

See full infographic: WEB | PDF
© Laurie o’keefe

In 2010, Hui-Chen Lu of Baylor College of Medicine and colleagues investigated the role of endocannabinoids in the so-called handshake hypothesis of cortical development, which posits that developing brain regions that form reciprocal connections produce a signal indicating their meeting. This process, which is now well characterized, allows neurons in these areas to act as scaffolds, guiding each other to their respective destinations and linkage partners.

Using mice that had CB1R knocked out of specific brain regions, Lu and collaborators found that axons connecting the cortex to the thalamus or vice versa require the activation of CB1Rs to arrive at their respective destinations. Deletion of CB1R at either end of this reciprocal connective circuit led to a mis­wiring of the pathway, which is crucial to such complex functions as goal-directed behavior and decision making.1 The same group has since found that synaptic plasticity in these brain regions is also disrupted in CB1R knockout mice.2

Although it might appear counterintuitive, stimulation of CB1R by the exogenous cannabinoids in marijuana might eventually result in less cannabinoid signaling by that cell. As with many neurotransmitters, endocannabinoids’ activation of cannabinoid receptors is under constant homeostatic regulation, and stimulation by potent, high-affinity agonists such as THC can disrupt this regulation. For example, exogenous cannabinoids cause CB1Rs to be rapidly internalized by the cell or otherwise to be desensitized in a more drastic manner than is seen with endocannabinoids, in order to avoid overactivating downstream signaling pathways. Embryonic exposure to cannabis is thus expected to downregulate CB1R function, leading to a reduction in the efficacy of normal endocannabinoid functions. Indeed, Tibor Harkany at the Karolinska Institute in Stockholm and colleagues found that THC displaces endocannabinoids from CB1Rs in the brains of rat pups in utero, resulting in temporarily reduced CB1R function and altered levels of nearly three dozen proteins in the fetal brain.3

Estimates for transfer efficiency through breast milk range from less than 1 percent to more than 10 percent of the lev­els found in maternal blood.

Our group, led by Olivier Manzoni, is particularly interested in how in utero cannabis exposure alters the mammalian prefrontal cortex, a brain region critical for executive functions ranging from decision making to the moderation of social behavior. The prefrontal cortex starts developing during the first trimester of gestation in mammals and continues to mature into early adulthood, making it an area with one of the longest periods of vulnerability to developmental insult.

Earlier this year, our laboratory published work demonstrating that rats whose mothers were given low-dose THC (or an analogous synthetic cannabinoid) while pregnant showed significant changes in synaptic plasticity and altered levels of several important proteins lasting well into adulthood. We found that the consequences of these changes manifested as a reduced sociability in the exposed offspring. Male rats in particular were much less likely to approach, play with, or sniff other rats.4

These findings parallel sociobehavioral changes seen in young adult humans exposed to cannabis during gestation. And scientists are now linking those effects to changes in the brain that are similar to what we observed in rats. Using functional MRI technology, for example, researchers participating in the Ottawa Prenatal Prospective Study observed a reduction in activity in the prefrontal cortices of adult offspring of mothers who smoked marijuana during pregnancy. This drop was associated with decreased working memory, echoing the attentional problems and memory dysfunction seen as early as infancy.5

Similarly, the project’s investigators have noted decreased cerebellar activity and increased premotor cortex activation, which correlated with a significantly reduced ability to filter irrelevant information in favor of more-pertinent stimuli.5 These changes likely explain the documented increase in impulsivity and inattention reported in children whose mothers used marijuana while pregnant.

Cannabis is also likely to affect the amygdala, which is critical for emotional development. In 2004, Yasmin Hurd of the Karolinska Institute and colleagues identified a significant reduction in dopamine D2 receptor mRNA in the amygdalae of fetal brains that correlated with the reported quantity of cannabis consumed by their mothers.6 (The fetuses were all between 18 and 22 weeks of gestation and donated by women who underwent voluntary abortions.) Given the known role of amygdala dopamine signaling in the regulation of mood and emotion, these findings could explain the increased depressive-like symptoms observed in children following cannabis exposure in utero, as well as these kids’ propensity towards inattention and impulsivity.

How Cannabis Affects the Function of Neurons

The human body contains two primary cannabinoid receptors: CB1R and CB2R. CB1R is present in the human fetal cerebrum by the first weeks of the second trimester, and is the brain’s most abundant G-protein coupled receptor. Located at the presynaptic terminal of neurons, CB1R is activated by endocannabinoids, which are synthesized from fatty acids in the postsynaptic neuron. The receptors’ activation modulates the presynaptic release of neurotransmitters, thereby affecting synaptic function and a range of downstream signaling agents, from glutamate, dopamine, and serotonin to neuropeptides and hormones. The function of CB2Rs in the brain is still poorly understood, but there is some evidence that they exist both pre- and post-synaptically, as well as on glia and astrocytes. One recent paper suggests that, like CB1Rs, CB2Rs regulate neurotransmitter release (Synapse, 72:e22061, 2018).

When people smoke or ingest marijuana, exogenous cannabinoids enter the nervous system and activate these receptors. Stimulation by these high-affinity agonists results in stronger binding and greater activation of CB1R, triggering the process of receptor downregulation. Specifically, the greater binding causes the receptors to be internalized and degraded, such that they are no longer as available for cannabinoid signaling, and can thereby alter neuronal firing and other downstream events.

See full infographic: WEB | PDF
© Laurie o’keefe

While we do know for certain that THC and other cannabinoids in marijuana are transferred to the fetus, there is also the possibility that CB1R activation in the mother causes hormonal or other downstream signaling changes that affect the developing fetus.

Another question researchers are currently trying to answer is how cannabis affects males and females differently. Surprisingly, some of the documented effects of in utero cannabis exposure are sexually divergent. The drop in dopamine receptor mRNA in the amygdalae of cannabis-exposed fetuses, for example, was much greater in male offspring, though a similar, but not statistically significant trend was found in females, according to the data. And the recent paper from our lab found specific changes to social behavior only in male mice. Conversely, the Generation R study found that girls were more likely than boys to exhibit increased aggression and impaired attention if their mothers had smoked marijuana while pregnant. In rodents, other effects, such as persistent changes in synaptic function and mRNA expression in adulthood, are found in both sexes. More research is needed to understand what drives these differences.

A future with cannabis

The problem of infant cannabis exposure extends well beyond pregnancy. THC and its lipophilic cannabinoid analogs are readily transferred through breast milk of humans and other mammals, and animal studies have pointed to these compounds’ influence on development throughout the pre-weaning period. Worse yet, given that these cannabinoids linger in the body for weeks, the “pump-and-dump” approach often employed to avoid feeding alcohol-laden milk to an infant isn’t as effective for cannabis users—a Friday night joint can continue to deliver active cannabinoids through breast milk throughout the weekend and into the next week.

Estimates for transfer efficiency through breast milk range from less than 1 percent to more than 10 percent of the levels found in maternal blood, leading to measurable quantities of THC in the urine of infants during the first weeks of feeding from a mother who has smoked or ingested cannabis. Once again, due to variable potency of cannabis in addition to other factors, knowing the exact amount of THC that enters the infant remains an open question. However, preliminary results from our lab indicate that even small quantities—such as those used for therapeutic rather than recreational purposes—cause significant and lasting changes in neuronal development in rats, including delayed maturation of prefrontal cortex networks.7 Earlier studies have documented increased ultrasonic vocalizations (an indication of elevated stress) in rat pups that ingested milk from THC-treated mothers, while multiple longitudinal studies in rodents have demonstrated that these offspring exhibit increased opioid-seeking behaviors, as well as higher rates of opioid self-administration.

As researchers work to untangle the nuances of how in utero and perinatal exposure to cannabis can lead to lasting changes in behavior and cognitive function, they are accumulating significant evidence pointing to aberrant changes in the developing brain. As the public and governments become increasingly tolerant of marijuana, we expect an uptick in research data, with cannabis users more likely to report their consumption.

Already, the legal status and availability of cannabis in the United States has led to the publication of human studies that were previously difficult or impossible to conduct. Just last year, researchers in Colorado, where recreational cannabis use by adults has been legal since 2014, studied the birth outcomes of more than 3,000 cannabis users and found a significantly higher rate of cannabis use during pregnancy than national estimates or than has been reported in regions where the drug remains illegal.8

Whether this is due to availability or changes in social acceptance is difficult to ascertain, but the trend is clear: as cannabis use becomes increasingly acceptable and widespread, maternal use is likely to follow suit. Research and public education must now be prioritized to inform future opinions on the safety, or lack thereof, of consuming cannabis during pregnancy, as well as during breastfeeding.

Andrew Scheyer is a postdoc focusing on the developmental consequences of perinatal exposure to cannabinoids in the laboratory of Olivier Manzoni at the Inmed Mediterranean Institute of Neurobiology in Marseille, France.


  1. J. Mulder et al., “Endocannabinoid signaling controls pyramidal cell specification and long-range axon patterning,” PNAS, 105:8760–65, 2008.
  2. C.-S. Wu et al., “Requirement of cannabinoid CB(1) receptors in cortical pyramidal neurons for appropriate development of corticothalamic and thalamocortical projections,” Eur J Neurosci, 32:693–706, 2010.
  3.  G. Tortoriello et al., “Miswiring the brain: Δ9-tetrahydrocannabinol disrupts cortical development by inducing an SCG10/stathmin-2 degradation pathway,” EMBO J, 33:668–85, 2014.
  4. A. Bara et al., “Sex-dependent effects of in utero cannabinoid exposure on cortical function,” eLife, 7:e36234, 2018.
  5. A.M. Smith et al., “Effects of prenatal marijuana on response inhibition: An fMRI study of young adults,” Neurotoxicol Teratol, 26:533–42, 2004.
  6. X. Wang et al., “In utero marijuana exposure associated with abnormal amygdala dopamine D2 gene expression in the human fetus,” Biol Psychiatry, 56:909–15, 2004.
  7. A.F. Scheyer et al., “Maternal cannabinoid exposure during lactation alters the developmental trajectory of prefrontal cortex GABA-currents in offspring,” bioRxiv, doi.org/10.1101/336735, 2018.
  8. T.L. Crume et al., “Cannabis use during the perinatal period in a state with legalized recreational and medical marijuana: The association between maternal characteristics, breastfeeding patterns, and neonatal outcomes,” J Pediatr, 197:90–96, 2018.Synapse

Interested in reading more?

Magaizne Cover

Become a Member of

Receive full access to digital editions of The Scientist, as well as TS Digest, feature stories, more than 35 years of archives, and much more!
Already a member?