Prenatal THC Exposure Linked to Child Attention Problems

Cannabis research during pregnancy often produces headlines that sound more certain than the evidence behind them. For parents, that can turn an already difficult subject into a choice between alarm and reassurance, with little guidance about what researchers actually measured.

A new study offers a more useful question: does the relationship between prenatal THC exposure and childhood development depend on when exposure occurs? Researchers following families in Colorado found that several behavioral associations were strongest when THC exposure continued through delivery.¹

The findings warrant attention, but they do not establish that cannabis causes ADHD or predict an individual child’s future. Their larger contribution is showing how repeated biological measurements can improve research, while highlighting why supportive conversations about cannabis should remain available throughout pregnancy.

What the Prenatal THC Study Measured

The research, accepted by The Journal of Pediatrics and available as a journal pre-proof, examined 460 mother-child pairs participating in the Healthy Start study. Researchers measured THC and its metabolites in maternal blood at approximately 17 and 27 weeks of pregnancy, and in umbilical cord blood at delivery.

Available samples differed across those timepoints: 460 participants had measurements at 17 weeks, 434 at 27 weeks, and 238 at delivery. Exposure was detected in approximately 8%, 6%, and 5%, respectively.

At around age five, parents or caregivers completed the Child Behavior Checklist, a standardized questionnaire covering attention, emotional regulation, aggression, and other behaviors. Children also completed tablet-based assessments of cognitive flexibility, inhibitory control, and receptive language.

This distinction matters. Behavioral questionnaires and cognitive tasks examine different aspects of development. A child can experience difficulty regulating attention without showing a measurable difference in vocabulary or performance on a particular cognitive task.

What Researchers Found About Attention and Behavior

The strongest associations generally appeared among children with THC exposure detected at delivery. Compared with children without detected exposure at that time, their adjusted attention-problem scores were 5.2 points higher. ADHD symptom scores were 2.4 points higher, while externalizing behavior scores, which capture outwardly directed difficulties, were 7.6 points higher.

These were differences in standardized T-scores, not percentages or increases in the probability of a diagnosis. The study describes these scores as having a reference mean of 50 and a standard deviation of 10.

The confidence intervals indicate uncertainty around each estimated difference. Importantly, adjusted average behavioral scores among exposed children remained below established thresholds for referral for further evaluation. That does not make the differences irrelevant, but it prevents interpreting them as evidence that the average exposed child had a clinical disorder.

The primary analyses also found no clear associations with cognitive flexibility, inhibitory control, or receptive language. Those findings should appear alongside the behavioral results, rather than disappearing beneath a broad claim about impaired brain development.

Does Late Pregnancy Represent a Critical Window?

There is a biological reason to investigate timing. The endocannabinoid system helps guide brain development, and THC can interact with receptors involved in that process. Late pregnancy includes rapid maturation of brain circuits, making it a plausible period of heightened sensitivity.

However, a plausible explanation is not a demonstrated mechanism. This study did not directly measure changes in fetal brain circuitry.

There is also an important alternative explanation: exposure at delivery may identify pregnancies with longer-lasting exposure. A stronger association could reflect cumulative exposure, a particularly sensitive late window, or both. Because exposure measurements were correlated across pregnancy, the study cannot cleanly separate these possibilities.

That distinction improves the article’s practical meaning. The results do not establish an earlier safe window or suggest that occasional use before a certain week is harmless. They show why treating prenatal cannabis exposure as a single yes-or-no event can conceal potentially important differences.

Why Blood Biomarkers Improve Cannabis Research

Many pregnancy studies depend on people remembering and disclosing cannabis use. Researchers can miss exposure when participants forget details, interpret questions differently, or fear judgment.

Blood biomarkers provide an objective indication that THC or its metabolites were present above the laboratory’s detection threshold. Repeated measurements add information about persistence that a single questionnaire or sample cannot provide.

Nevertheless, biomarkers are not a complete exposure history. This study lacked detailed information about product potency, frequency of use, and consumption method. Its relatively small exposed groups also prevented a meaningful analysis of how outcomes varied with biomarker concentrations.

MyCannabis’s recent examination of barriers to stronger cannabis research explains why better evidence requires more than identifying a possible biological mechanism. Here, progress means combining laboratory measurements with better information about products, symptoms, and the circumstances surrounding use.

The Limitations Behind the Strongest Findings

The most consequential limitation is the small group behind the delivery results: only 12 children had exposure detected in cord blood. That makes replication in larger samples essential.

Exposure groups also differed in income, education, tobacco use, diet, and other characteristics. Researchers adjusted for several potential confounders, including maternal age, education, psychiatric illness, alcohol, and smoking. Nevertheless, statistical adjustment cannot guarantee that every relevant difference has been removed.

Other uncertainties included substances beyond tobacco and alcohol, incomplete information about genetic conditions, and reliance on caregiver reports of behavior. These limitations do not erase the findings. They define how confidently the findings can be interpreted.

Recent MyCannabis coverage of prenatal cannabis exposure and child brain connectivity illustrates a complementary research approach. Imaging can explore possible developmental pathways, while behavioral follow-up examines everyday functioning. Neither approach, by itself, establishes causation or determines a child’s outcome.

Turning Research Into Supportive Prenatal Care

The authors suggest that stopping cannabis use later in pregnancy could still be valuable. That is a constructive implication, but this was not a cessation trial. It did not establish how much stopping at a particular week changes developmental outcomes.

The stronger practical message is that an earlier exposure should not make someone feel that subsequent choices are pointless. Prenatal care offers repeated opportunities to discuss symptoms, reassess treatment, and obtain support.

The American College of Obstetricians and Gynecologists’ 2025 cannabis guidance recommends cessation during pregnancy and emphasizes supportive approaches. It also distinguishes conversational screening from biological testing, advising that biological tests should not serve as cannabis screening assessments.

That distinction matters: laboratory testing can strengthen a consented research study without becoming the preferred foundation for routine clinical conversations.

A useful discussion with a prenatal care provider can cover:

• What cannabis is being used to manage, including nausea, sleep difficulties, or anxiety.
• Which alternatives have better-established safety information during pregnancy.
• What practical support could make reducing or stopping use more achievable.

This approach addresses the reason for use alongside the exposure itself. Simply advising someone to stop leaves an important gap if the symptom that prompted use remains untreated.

Why CBD and Edibles Are Not Established Pregnancy Alternatives

The Healthy Start analysis focused on THC exposure. It did not demonstrate that switching to CBD, gummies, or another product would protect fetal development.

Health Canada’s pregnancy guidance, updated in October 2026, states that a safe amount of cannabis has not been established. It also explains that THC and CBD cross the placenta, and that edibles and vaping are not safer pregnancy alternatives to smoking.

The opportunity emerging from this research is better precision: clearer exposure measurements, more careful distinctions between behavior and cognition, and support that recognizes why people use cannabis. For families, the findings support informed conversations and continued care, without treating a group-level association as a forecast for an individual child.

References:

¹ Moore, B. F., Crume, T., Wilkening, G., Mueller, N. T., Dabelea, D., Sempio, C., Klawitter, J., & Christians, U. (2026). Prenatal exposure to delta 9-tetrahydrocannabinol and child neurodevelopment: A critical windows analysis. The Journal of Pediatrics. Advance online publication. https://doi.org/10.1016/j.jpeds.2026.115363