rethinkTHC Search
Menu
Study breakdown

Associations between prenatal cannabis use and child outcomes largely disappeared when paternal cannabis use was used as a control

Prospective CohortStrong evidence
The takeaway

In the ALSPAC cohort of 15,013 families, associations between maternal cannabis use in pregnancy and child outcomes were modest and not statistically different from paternal cannabis use associations, suggesting confounding rather than biological causation.

Perinatal epidemiologists, obstetricians, and public health officials evaluating prenatal cannabis guidance.

Paternal cannabis showed similar associations

What the researchers found

Maternal cannabis use during pregnancy was associated with decreased birth weight (-110.2 g), decreased birth length (-0.45 cm), neonatal special care admission (OR 1.64), and lower education achievement at age 16 (-19.2 points). However, most associations were attenuated after controlling for confounders including socioeconomic position, and associations were not quantitatively different from paternal cannabis use, suggesting residual confounding drives the observed relationships.

Why it matters

Many studies report associations between prenatal cannabis exposure and adverse outcomes, but these are often confounded by socioeconomic factors. By showing that paternal cannabis use (which cannot biologically affect the fetus in the same way) produces similar associations, this study suggests the observed harms may reflect social disadvantage rather than direct biological effects.

The numbers in context

N = 15,013 mother-father-child trios. 5% of mothers and 13% of fathers reported cannabis use. Birth weight decrease: -110.2 g (95% CI -185.1 to -35.3). Birth length decrease: -0.45 cm. Special care admission OR: 1.64. Age 16 education score decrease: -19.2 points. All attenuated after covariate adjustment.

How the study worked

Negative control exposure study using 15,013 mother-father-child trios from the ALSPAC birth cohort (1990-1992). Maternal cannabis use was the primary exposure and paternal cannabis use served as a negative control. Models adjusted for household SEP markers and parental tobacco, alcohol, and drug use. Wald tests compared maternal vs. paternal association strengths.

What this study cannot tell us

ALSPAC participants were recruited in 1990-1992 when cannabis potency was lower. Self-reported cannabis use may underestimate exposure. Negative control design assumes paternal cannabis has no biological effect on offspring, which may not be entirely true (e.g., epigenetic effects on sperm).

How to read the evidence

Large prospective cohort with innovative negative control design and comprehensive confounder adjustment, though historical cohort with potentially lower-potency cannabis exposure.

When this study was published

2025 publication using 1990-1992 cohort data

The bigger picture

This study challenges a common assumption in the prenatal cannabis literature. The negative control design is a powerful tool for detecting confounding, and its application here suggests that policies based on direct biological harm from prenatal cannabis may overstate the evidence.

Questions still open

  • Would modern high-potency cannabis produce different results? Are there dose-dependent biological effects that this study could not detect? How should prenatal cannabis guidance change if confounding explains most associations?

Common questions

Does cannabis use during pregnancy harm the baby?
This study found modest associations (lower birth weight, lower test scores), but crucially, paternal cannabis use showed similar-sized associations. Since fathers cannabis use cannot directly affect the fetus biologically in the same way, this suggests socioeconomic confounding rather than direct biological harm may explain the findings.
What is a negative control exposure?
Using paternal cannabis use as a negative control tests whether observed associations are biological or confounded. If the mother cannabis exposure harms the fetus directly, the association should be stronger for maternal than paternal use. Finding similar associations for both suggests confounding by shared family factors like socioeconomic position.

Read the original research

Maternal cannabis use in pregnancy, perinatal outcomes, and cognitive development in offspring: a longitudinal analysis of the ALSPAC cohort using paternal cannabis use as a negative control exposure.

European journal of epidemiology, 40(5), 549-562

Citation

Corsi, Daniel J; Morris, Tim T; Reed, Zoe E; Davey Smith, George. (2025). Maternal cannabis use in pregnancy, perinatal outcomes, and cognitive development in offspring: a longitudinal analysis of the ALSPAC cohort using paternal cannabis use as a negative control exposure.. European journal of epidemiology, 40(5), 549-562. https://doi.org/10.1007/s10654-025-01233-w

Explore the wider topic