rethinkTHC Search
Menu
Study breakdown

Meta-analysis found inconsistent evidence linking prenatal cannabis to birth defects, with two specific exceptions

Meta AnalysisModerate evidence
The takeaway

A meta-analysis of 23 studies found a modest overall association between prenatal cannabis and birth defects (pooled OR 1.33), but most associations disappeared after adjustment, except for Ebstein anomaly and gastroschisis.

Obstetricians, teratology researchers, prenatal counselors

Most birth defect associations attenuated to non-significance after adjustment for confounders

What the researchers found

Across 23 studies spanning birth years 1968-2021, the pooled unadjusted odds ratio for any birth defect was 1.33 (95% CI 1.14-1.56), attenuating to 1.22 (95% CI 1.00-1.50) after adjustment. Most anatomic-group associations attenuated to non-significance after adjustment. Two specific defects retained significant adjusted associations: Ebstein anomaly (aOR 2.19, 95% CI 1.25-3.82) and gastroschisis (aOR 2.50, 95% CI 1.09-5.74), though both had limited supporting data.

Why it matters

This is the most comprehensive meta-analysis to date on cannabis and birth defects. The finding that most associations disappear after adjustment suggests confounding (tobacco, alcohol, socioeconomic factors) drives much of the apparent risk.

The numbers in context

23 studies; birth years 1968-2021; pooled OR 1.33 (95% CI 1.14-1.56); pooled aOR 1.22 (95% CI 1.00-1.50); Ebstein anomaly aOR 2.19 (95% CI 1.25-3.82); gastroschisis aOR 2.50 (95% CI 1.09-5.74)

How the study worked

Systematic review and meta-analysis searching Google Scholar, BIOSIS, PubMed/MEDLINE, EMBASE, CINAHL, and ClinicalTrials.gov. Screened with Abstrackr software. Included 23 observational studies examining birth defects with prenatal cannabis exposure. Meta-analyzed by anatomic group. Registered with PROSPERO.

What this study cannot tell us

Most included studies used unadjusted estimates and scored low on risk-of-bias assessment. Studies were heterogeneous in design and cannabis exposure measurement. The two specific defects with persistent associations had very limited data (2 and 5 studies respectively).

How to read the evidence

Registered systematic review and meta-analysis of 23 studies, but limited by poor quality of underlying studies and heavy reliance on unadjusted estimates.

When this study was published

Published 2023 analyzing studies from 1968-2021

The bigger picture

The inconsistency in this evidence base reflects the fundamental challenge of studying prenatal cannabis exposure: heavy confounding, self-reported use, and the impossibility of conducting randomized trials.

Questions still open

  • Are the Ebstein anomaly and gastroschisis signals real or artifacts of limited data? Would studies with better confounding control and validated exposure assessment change the overall picture?

Common questions

Does cannabis use during pregnancy cause birth defects?
Evidence is inconsistent. This meta-analysis of 23 studies found a modest overall association, but most specific birth defect associations disappeared after adjusting for confounders like tobacco and alcohol use. Two rare defects (Ebstein anomaly and gastroschisis) retained associations but had very limited data.
How strong is the evidence on cannabis and birth defects?
Weak overall. Most studies used unadjusted estimates and scored low on quality assessments. The fact that associations largely disappear after adjustment suggests confounding factors rather than cannabis itself may explain much of the observed risk.

Read the original research

Prenatal Exposure to Cannabis and Risk of Major Structural Birth Defects: A Systematic Review and Meta-analysis.

Obstetrics and gynecology, 142(2), 269-283

Citation

Delker, Erin; Hayes, Shana; Kelly, Ann E; Jones, Kenneth L; Chambers, Christina; Bandoli, Gretchen. (2023). Prenatal Exposure to Cannabis and Risk of Major Structural Birth Defects: A Systematic Review and Meta-analysis.. Obstetrics and gynecology, 142(2), 269-283. https://doi.org/10.1097/AOG.0000000000005252

Explore the wider topic