Cannabis in Pregnancy: What the Science Actually Shows and What It Doesn’t
A new narrative review maps what researchers know about prenatal cannabis exposure, exposes critical gaps in the evidence base, and calls for better coordination between laboratory science, clinical practice, and policymaking to address rising use during pregnancy.
Why This Matters
Cannabis use during pregnancy is increasing in many countries, driven in part by expanding legalization and shifting public perceptions of safety. Pregnant patients frequently ask clinicians whether cannabis is harmful, and clinicians often lack clear, evidence-based answers. The endocannabinoid system plays a documented role in fetal brain development, making the biological plausibility of harm from exogenous cannabinoid exposure a legitimate scientific concern. This review arrives at a moment when the gap between what patients assume, what clinicians can confidently say, and what the research actually supports is widening rather than narrowing.
Clinical Summary
Prenatal cannabis exposure (PCE) has become a topic of increasing clinical urgency as prevalence data suggest rising use among pregnant populations across multiple countries. A 2024 narrative review published in Drug and Alcohol Dependence Reports by Cupo, Dominguez-Cancino, Nazif-Munoz, and Chakravarty synthesizes evidence from human observational studies and animal experiments to characterize what is known about PCE and offspring neurodevelopment. The mechanistic concern centers on delta-9-tetrahydrocannabinol (THC), which crosses the placental barrier and interacts with the endocannabinoid system during critical windows of fetal brain development. This system regulates neuronal proliferation, migration, and synaptogenesis, meaning exogenous cannabinoid exposure has a credible biological pathway through which it could disrupt normal neurodevelopmental trajectories.
The review reports that prevalence estimates for cannabis use during pregnancy vary dramatically, from 0.24% to 46.5%, depending on the population studied and the measurement method used, with biological assays consistently yielding higher rates than self-report. Data from California showed an increase from approximately 4% to 7.5% between 2009 and 2016, and cannabis use disorders among hospitalized pregnant individuals rose from 0.008 to 0.02 of hospitalizations between 2010 and 2018 across 35 U.S. states. Both human and animal literatures suggest potential cognitive, behavioral, and neurological effects in exposed offspring, but the authors emphasize that human evidence is observational and confounded by factors such as co-occurring substance use, socioeconomic status, and mental health conditions. Animal models, while allowing causal inference, have relied predominantly on THC alone, which does not reflect the complex composition of real-world cannabis products. The authors call for improved methodological alignment between laboratory paradigms and actual human exposure patterns before clinical or policy recommendations can be firmly grounded.
Dr. Caplan’s Take
This review does something genuinely useful: it names the translational gap between what animal labs study and what pregnant patients actually experience. That honesty matters. When a patient asks me whether cannabis is safe during pregnancy, the truthful answer is that we have biologically plausible reasons for concern, accumulating signals of potential neurodevelopmental harm, and no definitive human proof of causation. That is a difficult message to deliver clearly, and this paper helps articulate why the uncertainty persists. It is not that scientists have been negligent; it is that the research questions are genuinely hard to answer given confounding, stigma-driven underreporting, and rapidly evolving cannabis products.
In practice, I advise pregnant patients to avoid cannabis based on the precautionary principle, the plausible mechanism of harm, and the absence of evidence establishing safety. I also acknowledge the reasons some patients use cannabis during pregnancy, including nausea, anxiety, and chronic pain, and I work to ensure they have access to evidence-based alternatives. I do not treat this as a moral conversation. I treat it as a clinical one where the honest answer is that we cannot guarantee safety, and that alone warrants caution.
Clinical Perspective
This review sits early in the research arc for prenatal cannabis exposure, a field where observational human data and mechanistic animal data point in a concerning direction without converging on definitive causal conclusions. For clinicians, the key takeaway is that the evidence does not yet support specific dose-response thresholds, trimester-specific risk windows, or outcome-specific warnings that could be communicated to patients with confidence. What it does support is a general precautionary stance grounded in biological plausibility and the consistent direction of available findings. Importantly, the wide variability in prevalence estimates (0.24% to 46.5%) suggests that clinicians should not assume low rates of use in their patient populations, particularly in jurisdictions where legalization may reduce perceived risk.
From a pharmacological standpoint, THC is lipophilic, crosses the placenta readily, and accumulates in fetal tissue, including brain tissue. Its interaction with the developing endocannabinoid system is not theoretical but well-characterized in animal models. Clinicians should be aware that patients may use a wide variety of cannabis products with differing THC concentrations, routes of administration, and co-occurring cannabinoids such as CBD, none of which are adequately represented in the current animal literature. Screening for cannabis use during prenatal care should be routine, nonjudgmental, and ideally supplemented by awareness that self-report alone significantly underestimates true prevalence. One concrete step every prenatal care provider can implement now is incorporating a standardized, destigmatized substance use screening tool at the first prenatal visit and revisiting the conversation at subsequent visits.
Study at a Glance
- Study Type
- Narrative review
- Population
- Pregnant individuals using cannabis; offspring exposed prenatally; nonhuman animal models
- Intervention
- Not applicable (observational synthesis of prenatal cannabis/THC exposure)
- Comparator
- Unexposed offspring in cited primary studies
- Primary Outcomes
- Prevalence of prenatal cannabis use; offspring neurodevelopmental outcomes; lab-to-clinic translational gaps
- Sample Size
- Not applicable (narrative review synthesizing multiple primary studies)
- Journal
- Drug and Alcohol Dependence Reports
- Year
- 2024
- DOI
- 10.1016/j.dadr.2024.100282
- Funding Source
- Not reported in available text
What Kind of Evidence Is This
This is a narrative literature review, which occupies a relatively low position in the evidence hierarchy compared to systematic reviews or meta-analyses. Narrative reviews synthesize expert interpretation of selected literature but do not employ formal search strategies, predefined inclusion and exclusion criteria, or risk-of-bias assessments. The single most important inference constraint this imposes is that the completeness and representativeness of the included studies cannot be independently verified, meaning the conclusions reflect the authors’ informed judgment rather than a reproducible, exhaustive survey of the evidence.
How This Fits With the Broader Literature
The findings of this review are broadly consistent with the direction of evidence from larger longitudinal efforts, particularly the Adolescent Brain Cognitive Development (ABCD) Study, which has reported associations between prenatal cannabis exposure and subtle differences in offspring brain structure and behavior. The review also aligns with earlier work by Gunn and colleagues (2016), whose systematic review found associations between prenatal cannabis exposure and decreased birth weight and increased neonatal intensive care admissions, though effect sizes were modest and confounding remained a concern.
Where this review adds value is in explicitly naming the translational gap between animal and human research as a primary barrier to progress. Most prior reviews have focused either on epidemiological data or on preclinical mechanisms in isolation. By framing the mismatch between THC-only animal dosing regimens and the complex, heterogeneous cannabis products used by pregnant individuals, the authors provide a roadmap for the kind of research most likely to move the field forward.
Common Misreadings
The most likely overinterpretation of this review is to treat it as confirmation that prenatal cannabis exposure definitively causes neurodevelopmental harm in human offspring. The review itself is careful to distinguish between biological plausibility, suggestive signals from observational data, and established causation, but selective citation of its more concerning findings could easily strip away those qualifications. Equally problematic would be reading the acknowledged uncertainty as evidence of safety. The absence of definitive proof of harm in observational human data does not mean harm has been ruled out; it means the studies conducted so far are not designed to provide that level of certainty, a distinction the authors make explicitly.
Bottom Line
This narrative review confirms that prenatal cannabis exposure is a legitimate and growing clinical concern supported by plausible biological mechanisms and suggestive observational evidence, but it does not establish definitive causation between PCE and neurodevelopmental harm in humans. Its most important contribution is identifying the translational gap between animal research paradigms and real-world human exposure as the critical bottleneck for the field. For clinicians now, the evidence supports a precautionary approach and routine, nonjudgmental screening, not definitive risk quantification.
References
- Cupo L, Dominguez-Cancino KA, Nazif-Munoz JI, Bhakta SG, Bhakta SG, Chakravarty MM. Prenatal cannabis exposure: Bridging the gap between preclinical research and clinical realities. Drug and Alcohol Dependence Reports. 2024. DOI: 10.1016/j.dadr.2024.100282
- Gunn JKL, Rosales CB, Center KE, et al. Prenatal exposure to cannabis and maternal and child health outcomes: A systematic review and meta-analysis. BMJ Open. 2016;6(4):e009986. DOI: 10.1136/bmjopen-2015-009986
- Paul SE, Hatoum AS, Fine JD, et al. Associations between prenatal cannabis exposure and childhood outcomes: Results from the ABCD Study. JAMA Psychiatry. 2021;78(1):64-76. DOI: 10.1001/jamapsychiatry.2020.2902
Have thoughts on this? Share it:
