Benzodiazepines in Pregnancy: Less Risk Than Feared

A massive Korean cohort study offers reassurance, with caveats

STUDY: Cho Y et al, BMJ 2026;393:e088671

STUDY TYPE: Population-based cohort study with sibling-controlled analysis

FUNDING: National Research Foundation of Korea

Background

Benzodiazepines and Z-hypnotics (like zolpidem/Ambien) are among the most frequently prescribed drugs in this pregnancy with insomnia or anxiety. Whether prenatal exposure harms children’s psychiatric development has been uncertain.

The Study

• Nationwide Korean cohort followed nearly 3.8 million children born 2010-22
• Comparing 94,482 exposed to benzodiazepines or Z-hypnotics in utero against unexposed children, siblings, and a “past user” group (mothers who had taken these drugs before but not during pregnancy)
• The sibling-controlled analysis was the key methodological strength: comparing exposed and unexposed siblings within the same family controls for shared genetic and environmental confounders

In the crude analysis, prenatal exposure was associated with higher rates of psychiatric disorders in children. But once family-level factors were accounted for in the sibling analysis, the risk disappeared entirely (hazard ratio 0.99, 95% CI 0.94–1.04). There was no higher risk of psychiatric disorders — including ADHD, autism spectrum disorder, anxiety, or mood disorder — with med exposure in the sibling-controlled model.

Some subgroups showed modestly elevated point estimates: exposure during the second half of pregnancy, exposure throughout both halves, and Z-hypnotic use for 30 days or more. None reached statistical significance, but the confidence intervals were wide and didn’t rule out a small effect.

Practice Implications

1. While reassuring, still best to keep exposures brief, avoid multiple med exposures, and avoid in the second half of pregnancy when possible.