What Brain Scans Reveal About Prenatal Cannabis Exposure

When researchers study cannabis use during pregnancy, they often look for differences they can observe after birth: birth weight, attention, learning, or behavior. A new study in Translational Psychiatry¹ asks a question between those endpoints: do children with reported prenatal cannabis exposure also show differences in how their brain networks communicate?

Using brain scans and assessments from 10,838 children, the researchers found that reported exposure was associated with altered patterns of brain connectivity at around age 10. Some of those patterns resembled patterns associated with more behavioral symptoms and lower performance on certain cognitive measures. The findings offer a possible direction for understanding development, but they do not show that cannabis caused the differences or that a scan can predict a child’s future.

That distinction matters. A population study can detect a small average difference between groups even when the experiences of individual children overlap considerably. The practical story is therefore about what the research can add to an ongoing public health question, and what remains unresolved.

What the Researchers Measured

The study drew on baseline data from the Adolescent Brain Cognitive Development, or ABCD, Study. Of the 10,838 children included in this analysis, 751 were classified as having prenatal cannabis exposure based on retrospective reports of maternal use before or after pregnancy was recognized. The other 10,087 children formed the comparison group.

Researchers analyzed resting-state functional MRI scans. During this type of scan, a person is not asked to solve a problem or perform a task. Instead, investigators examine how activity in different parts of the brain rises and falls together. They used a framework called NeuroMark to organize the scans into 53 functional networks and estimate the relationships among them.

Think of the result as a map of coordination, rather than a photograph of damage. If two networks show a different pattern of coordinated activity, that observation does not establish what caused it, whether it is harmful, or how a particular child will function. The researchers compared these connectivity patterns with measures of behavior and cognition to see where the associations overlapped.

Why Brain Connectivity Adds to the Story

Previous coverage has examined outcomes closer to birth. For example, a review of prenatal cannabis use and neonatal outcomes addresses questions such as birth weight and preterm birth. The new study looks later in childhood and adds brain imaging to behavioral and cognitive assessments.

Its most useful contribution is the comparison between three sets of observations: reported exposure, connectivity, and measured outcomes. The researchers found connectivity differences associated with exposure across several brain systems. They then asked whether similar connectivity patterns were associated with children’s behavioral symptoms or cognitive performance. Some were.

One recurring pattern involved connections between subcortical and sensory networks. Lower connectivity in this area was associated with prenatal exposure, while stronger connectivity in these circuits generally aligned with fewer reported symptoms and better cognitive performance. That overlap gives researchers a more specific question to test as the children grow older.

It does not complete a chain of cause and effect. The scans, behavioral measures, and cognitive tests used here were assessed at the same childhood time point. A shared pattern cannot tell us whether prenatal exposure changed connectivity, whether connectivity contributed to behavior, or whether other influences affected all three. The authors explicitly describe these networks as candidates for further investigation, rather than proven pathways.

The Results Were Neither Uniform Nor Predictive

Children with reported exposure had higher scores across all 20 behavioral scales examined, but the differences varied in size. The largest reported behavioral difference was on the rule-breaking scale. A higher group average on that questionnaire does not mean an exposed child has a behavioral disorder, and the study cannot forecast an individual child’s conduct.

The cognitive findings also resist a simple headline. Exposed children performed less well on six of the 10 assessments. The largest difference was in crystallized cognition, which concerns learned knowledge and verbal skills. Yet the exposed group scored slightly higher on the composite measure of fluid cognition, which concerns reasoning through unfamiliar problems. Both the lower and higher differences were modest.

That mixed result is a reason to report the measures accurately, rather than describing cannabis exposure as producing a single, broad cognitive deficit. Children’s development spans many abilities, and a statistical difference in one score should not stand in for the whole child.

The imaging results require similar restraint. Although some associations were statistically significant in this large sample, the study’s strongest reported brain-to-outcome correlation was small. The researchers state that their connectivity patterns are not suitable as diagnostic or predictive biomarkers. An MRI scan based on these findings cannot tell parents whether their child was exposed to cannabis or will develop a mental health problem.

What the Study Cannot Separate

Pregnancy does not take place in a laboratory. Cannabis use may coincide with tobacco or alcohol exposure, stress, differences in income, family history, and other circumstances that can also relate to child development. The researchers adjusted for several such factors and ran additional checks, but they could not eliminate every competing explanation.

In particular, the study did not have a sufficiently large group exposed exclusively to cannabis to isolate its effects with confidence. Exposure information also came from retrospective caregiver reports, which may be affected by imperfect recall or reluctance to disclose use. The researchers did not have the precise, prospective exposure measurements needed to determine how dose, frequency, product type, and timing might change the results.

These limits do not make the observations meaningless. They define the next questions. Follow-up research could track connectivity and behavior across multiple ages, measure exposure during pregnancy rather than years afterward, and better distinguish cannabis from other exposures. Because the wider ABCD Study follows children over time, it provides an opportunity to examine whether the patterns identified here persist, change, or fade.

What This Means for Pregnancy Guidance

Scientific uncertainty about the size and cause of a long-term association is different from evidence that exposure is safe. THC can cross the placenta, and Health Canada’s pregnancy guidance advises avoiding cannabis during pregnancy. The U.S. Centers for Disease Control and Prevention likewise warns that cannabis use during pregnancy may affect a baby’s development.

For someone using cannabis to manage nausea, anxiety, pain, or another problem, that guidance raises a practical question: what can address the underlying need? A conversation with a healthcare professional can focus on the symptom, the products being used, and available alternatives. It can also make room for honest disclosure. As the study notes, fear of judgment may discourage people from discussing cannabis use, making both care and research less effective.

Three points are worth carrying forward:

• The study found group-level associations, not proof that cannabis caused an individual child’s outcome.
• The measured differences were generally modest and cannot be used to diagnose a child.
• Someone considering cannabis during pregnancy should discuss the reason for use and other options with a healthcare professional.

MyCannabis has also covered prenatal THC exposure in an animal study. That work explores a different question and cannot be treated as direct evidence of the human outcomes measured here. Likewise, recent laboratory research on CBD, THC, and developing immune cells illustrates why findings from cells, animals, and children must be interpreted within their respective settings.

The new brain imaging study moves the field beyond asking only whether a behavioral difference exists. It identifies network patterns that researchers can investigate over time. Its value lies in making the next studies more precise, while leaving room for the many biological, family, and environmental factors that shape a child’s development.

References:

¹ Fu, Z., Hutchison, K., Guha, A., & Calhoun, V. D. (2026). Prenatal cannabis exposure associated with altered brain connectivity: Neural correlates of cognitive and mental health variability in offspring. Translational Psychiatry. Advance online publication. https://doi.org/10.1038/s41398-026-04478-5