First-Trimester THC Exposure Changes Newborn Brain Connectivity, Study Finds

Early cannabis use in pregnancy may reshape the functional architecture of a newborn's brain before any behavioral signs emerge, according to preliminary research published in June 2026. The study, led by Nutor, Hendrix, and colleagues, detected significant associations between a biological marker of first-trimester cannabis exposure and altered resting-state connectivity across three brain regions in one-month-old infants. Because it focused exclusively on Black mother-infant dyads, a population historically absent from developmental neuroimaging, the work also begins filling a meaningful gap in the evidence base.

THC Metabolites in Early Pregnancy Linked to Infant Brain Connectivity

The team measured 11-nor-9-carboxy-delta-9-tetrahydrocannabinol (COOH-THC), a urinary metabolite that provides a biologically verified record of cannabis exposure, in maternal urine collected between 8 and 14 weeks of gestation. When researchers then imaged the babies' brains at roughly one month postpartum, higher COOH-THC levels in early pregnancy corresponded to distinctly altered connectivity patterns in three regions: the hippocampus, the insula, and the caudate. This is early, investigational-stage research into the neurodevelopmental consequences of prenatal cannabis exposure, and the findings do not establish causation. But the fact that associations held across three functionally distinct brain areas, and persisted after accounting for co-occurring maternal psychosocial factors, gives the results some weight despite the small sample.

The hippocampus handles memory formation and stress regulation. The insula integrates internal bodily signals with emotion and cognitive control. The caudate, part of the striatum, underpins reward processing, motivation, and motor learning. Altered connectivity in any one of these regions could matter for development; altered connectivity in all three at once is a pattern worth examining more closely.

Why Newborn Brain Wiring Matters Before Behavior Shows Up

One of the more striking aspects of the findings is that the altered connectivity patterns did not correlate with scores on the NICU Network Neurobehavioral Scale (NNNS), a standardized assessment used to gauge newborn neurological and behavioral function. In other words, the babies whose brains showed the most altered wiring did not look different on a behavioral exam at one month old.

That gap between imaging signal and behavioral measurement is not surprising, and it is not reassuring. The endocannabinoid system plays a central role in fetal brain development, guiding how neurons connect and how circuits stabilize. Disruptions during that window can leave traces that take years to surface as cognitive or emotional difficulties. Prior work in both animal models and clinical cohorts has linked prenatal cannabis exposure to lasting changes in cognitive function and behavior in older children; the current study suggests the biological substrate for those later problems may be detectable in the brain's resting-state organization within the first weeks of life, before any test can capture them behaviorally.

This framing is speculative, and the authors are careful not to overstate it. But it points toward a critical reason neuroimaging in newborns may serve as an early sentinel for developmental risk that behavioral screening would miss entirely.

How the Team Measured Exposure and Imaged Newborn Brains

Using biological rather than self-reported measures of cannabis exposure is a meaningful methodological choice. Self-report is prone to underestimation, particularly in populations where cannabis use during pregnancy may carry social stigma or legal risk. COOH-THC in urine is a direct metabolic product of THC: because THC is lipid-soluble and crosses the placenta readily, accumulating in fetal tissue with repeated maternal use, a urine assay collected in the first trimester offers a more objective window into fetal exposure than a questionnaire.

The imaging side relied on resting-state functional MRI in neonates, a non-invasive technique that maps temporal correlations in low-frequency blood-oxygen-level-dependent signal across brain regions. Infants do not perform tasks; they simply lie still (often while sleeping), and the scanner captures the spontaneous fluctuations that reveal which regions are functionally coupled. The team used seeds placed in bilateral hippocampus, insula, and caudate, then ran voxelwise connectivity analyses across the brain. The 42 dyads in the study were all Black mother-infant pairs, a deliberate focus that addresses the systematic underrepresentation of this population in developmental neuroimaging.

Key Limitations in a 42-Dyad Sample

With 42 participants, the study sits in genuinely preliminary territory. Small samples inflate the risk of false positives and make effect-size estimates unstable. The findings should not be interpreted as definitive evidence that first-trimester cannabis use causes lasting brain damage; they are a signal that warrants replication in a larger, more diverse cohort.

The focus on Black mother-infant dyads is both a scientific strength and a constraint on generalizability. The patterns observed here may not extend to other populations, and conversely, findings from predominantly white cohorts that dominate the existing literature may not apply here either. Much of the evidence on perinatal cannabis exposure remains inconclusive, with conflicting results on cognitive outcomes across studies that used different exposure measures, sample compositions, and follow-up intervals. This study adds a rigorous biological-measurement approach and a historically underrepresented sample, but it cannot resolve those broader conflicts on its own.

The study also addressed only first-trimester exposure. Whether exposure in the second or third trimester produces similar or different connectivity patterns is an open question the current design cannot answer.

What Comes Next for Prenatal Cannabis Research

The authors have made their uncorrected statistical maps publicly available on Neurovault, which allows other research groups to incorporate the data into future meta-analyses. That kind of open-science practice is particularly valuable for small-sample neuroimaging studies: pooling data across sites can yield the statistical power that no single study at this scale can achieve alone.

Longer-term follow-up of this cohort would help clarify whether the connectivity differences observed at one month predict developmental outcomes at ages when behavioral and cognitive differences are measurable. NIH-funded work now underway is exploring biomarkers and screening tools related to prenatal cannabis, which suggests the field is moving toward the kind of prospective, multi-site infrastructure needed to answer the causal questions this study raises.

For now, the takeaway is appropriately cautious: biological traces of cannabis use in early pregnancy are associated with measurable differences in newborn brain organization, and those differences appear in regions critical to memory, emotion, and reward. Whether those differences matter for the child years down the road is exactly what the next generation of studies needs to determine.

Evidence rating

INVESTIGATIONAL — 57/100

Caveats:

• Very small sample (n=42), limiting statistical power and generalizability.
• Observational design; association does not establish causation.
• Findings are limited to Black mother-infant dyads and may not generalize to other populations.
• Altered connectivity patterns were not correlated with any measurable newborn neurobehavioral outcomes.
• Authors describe results as preliminary and note the study is not yet independently replicated.

Observational longitudinal study (n=42). Reports associations between first-trimester urinary THC metabolite levels and newborn resting-state fMRI connectivity patterns; no causal inference can be drawn. Findings are preliminary, limited to a single demographic cohort, and have not been independently replicated.

References

1. Chaela Nutor, Jocelyn Stanfield, Patricia A. Brennan, Dana Boyd Barr, Anne L. Dunlop, Cassandra L. Hendrix (2026). Maternal Early Pregnancy Tetrahydrocannabinol (THC) Metabolites Correlate With Newborn Resting‐State Functional Connectivity. Developmental Psychobiology. https://doi.org/10.1002/dev.70179
2. Cannabis Use during Pregnancy: An Update – PMC. NIH. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10595914/
3. Chia-Shan Wu, Christopher P Jew, Hui-Chen Lu (2011). Lasting Impacts of Prenatal Cannabis Exposure and the Role of Endogenous Cannabinoids In the Developing Brain. Future Neurology. https://doi.org/10.2217/fnl.11.27
4. Marijuana and Maternal, Perinatal, and Neonatal Outcomes – StatPearls – NCBI Bookshelf. NIH. https://www.ncbi.nlm.nih.gov/books/NBK555998/
5. Christopher D. Smyser, Jeffrey J. Neil (2015). Use of resting-state functional MRI to study brain development and injury in neonates. Seminars in Perinatology. https://doi.org/10.1053/j.semperi.2015.01.006
6. Chidimma Doris Azubuike, Oliver Grundmann, Amie J. Goodin (2025). Neurodevelopmental effects of perinatal exposure to cannabis on progeny: A narrative review. Drug and Alcohol Dependence Reports. https://doi.org/10.1016/j.dadr.2025.100372
7. UTHealth Houston study explores how prenatal cannabis exposure may affect fetal brain development through NIH grant. uth.edu. https://www.uth.edu/news/story.htm?id=c920556b-4e01-443b-810a-e24a5202a0a2