Children exposed to both cannabis and tobacco before birth scored slightly lower on tests measuring reasoning, attention, memory and processing speed at ages 9 to 11, according to a large U.S. study.
Researchers analyzed data from 8,706 children enrolled in the Adolescent Brain Cognitive Development, or ABCD, Study and compared cognitive performance across four prenatal-exposure groups.
The study published in Addiction found that children with reported prenatal exposure to both cannabis and tobacco had modestly lower fluid cognition scores than children exposed only to cannabis or to neither substance.
However, the researchers found no significant difference in crystallized cognition, which covers accumulated knowledge such as reading and vocabulary.
The results also do not prove that combined cannabis and tobacco exposure caused the cognitive differences. The study was observational, exposures were reported retrospectively years later, and the authors describe the findings as exploratory.
Table of Contents
More Than 8,700 Children Were Included
Researchers used baseline data from the U.S. ABCD Study, one of the largest long-term investigations of brain development and child health.
The children were between 9 and 11 years old.
They were divided into four groups according to reported prenatal exposure:
| Prenatal exposure | Children |
|---|---|
| Cannabis + tobacco | 236 |
| Cannabis only | 195 |
| Tobacco only | 803 |
| Neither substance | 7,472 |
| Total | 8,706 |
Researchers then compared two broad types of cognitive performance.
What Is Fluid Cognition?
Fluid cognition describes abilities used to process new information and solve unfamiliar problems.
The study measured it using the NIH Toolbox Cognition Battery, including tasks related to:
- working memory;
- attention;
- inhibitory control;
- cognitive flexibility;
- episodic memory;
- processing speed.
These abilities are important for learning and day-to-day problem solving.
The second category, crystallized cognition, reflects knowledge accumulated through learning and experience, including reading and vocabulary skills.
Combined Exposure Was Linked to Lower Fluid Cognition
In the main statistical analysis, prenatal cannabis exposure alone was not significantly associated with lower fluid cognition.
Prenatal tobacco exposure alone was also not significantly associated with lower scores in that model.
But when the researchers examined the interaction between cannabis and tobacco exposure, a statistically significant association emerged.
The interaction estimate was:
β = −2.08, 95% CI −4.13 to −0.03.
The confidence interval was wide and close to zero, meaning the size of the effect remains uncertain.
The authors therefore recommend cautious interpretation.
Children With No Exposure Scored About 1.4 Points Higher
When the four exposure groups were compared directly, children exposed to neither cannabis nor tobacco scored approximately 1.4 points higher on fluid cognition than children exposed to both.
Children exposed only to cannabis scored about 2.1 points higher than the co-exposed group.
Those exposed only to tobacco scored around 1.3 points higher, but that difference was not statistically significant.
Importantly, standardized analyses indicated that the differences were small in magnitude.
This does not support claims that prenatal co-exposure causes severe cognitive impairment or dramatically lowers intelligence.
Reading and Vocabulary Scores Were Not Significantly Different
The researchers found no significant cannabis-tobacco interaction for crystallized cognition.
Children in the combined-exposure group did not differ significantly from:
- cannabis-only children;
- tobacco-only children;
- unexposed children
on the crystallized cognition composite.
That means the study does not support the broader claim that prenatal cannabis and tobacco exposure reduced performance across all cognitive domains.
The association was specific to the fluid cognition measure examined.
Why Researchers Studied Cannabis and Tobacco Together
Cannabis and tobacco are frequently used together.
The original paper notes that in the United States, a substantial proportion of pregnant people who report cannabis use also report tobacco use.
That makes it difficult to understand the potential developmental effects of cannabis by examining it in isolation.
Cannabis contains tetrahydrocannabinol, or THC, while tobacco exposes the fetus to nicotine and numerous other compounds.
Both THC and nicotine can cross the placenta.
Cannabinoid and nicotinic acetylcholine receptors are also present during early development and overlap in several brain systems, giving researchers a biological reason to investigate whether combined exposure could differ from either exposure alone.
The study does not demonstrate that the two substances act synergistically in the developing human brain, however.
The Study Cannot Establish Cause and Effect
This is the most important limitation.
The research was a cross-sectional exploratory analysis of baseline ABCD data.
Researchers identified an association between reported prenatal co-exposure and cognitive performance years later, but several other factors could influence that relationship.
The statistical models accounted for variables including:
- maternal age;
- education;
- family income;
- prenatal alcohol exposure;
- whether the pregnancy was planned;
- parental mental-health symptoms.
But observational studies cannot account perfectly for every social, environmental or genetic difference between families.
The authors specifically cite residual genetic confounding, selection bias and exposure misclassification as important limitations.
Exposure Was Reported Nearly a Decade Later
Another major issue is how prenatal exposure was measured.
Mothers or caregivers were asked to recall cannabis and tobacco use during pregnancy when the children were already 9 to 11 years old.
That creates the possibility of inaccurate recall or underreporting.
The study also lacked detailed information about:
- cannabis dose;
- frequency of use;
- THC potency;
- tobacco quantity;
- precise timing during pregnancy;
- secondhand smoke exposure.
Researchers therefore could not determine whether greater exposure was associated with larger cognitive differences.
‘Co-Exposure’ Did Not Necessarily Mean Both Were Used at the Same Time
Children classified as co-exposed had prenatal histories involving both cannabis and tobacco.
But the available data could not establish that the two substances were always used simultaneously.
For example, one substance may have been used before the pregnancy was recognized and another later.
That limits the ability to infer a direct biological interaction between THC and nicotine.
The authors consequently describe the cannabis-tobacco interaction as a statistical interaction, not proof of a biological synergistic effect.
The Analysis Was Not Preregistered
The research team also notes that the analysis was not preregistered before the data were examined.
This means the findings should be viewed as exploratory and need independent replication.
The result also only narrowly crossed the conventional threshold for statistical significance.
Together with the relatively small number of co-exposed children, just 236 out of 8,706, this reinforces the need for caution.
Earlier Research Has Produced Mixed Results
Studies of prenatal cannabis exposure and later cognition have not produced completely consistent findings.
Some research has reported associations with attention or executive-function outcomes, while other studies have found little evidence of meaningful cognitive differences after accounting for tobacco exposure and socioeconomic factors.
The new study contributes something different by directly asking whether combined cannabis and tobacco exposure is associated with cognitive outcomes beyond either substance alone.
Its findings suggest a possible effect on fluid cognition, but replication in prospective studies with more accurate exposure measurements is needed.
What the Study Really Shows
The most defensible conclusion is relatively narrow.
Among 8,706 U.S. children aged 9 to 11, those with reported prenatal exposure to both cannabis and tobacco had slightly lower fluid cognition scores than children exposed only to cannabis or to neither substance.
The researchers did not find a significant effect on crystallized cognition.
They also did not demonstrate that prenatal cannabis alone lowered fluid cognition in this analysis.
The study therefore adds evidence that co-exposure deserves closer investigation, but it does not establish that cannabis and tobacco together directly caused the observed differences.
Future prospective studies will need to measure substance exposure during pregnancy itself, collect detailed dose and timing information, and follow children over time before the developmental implications can be understood more confidently.
