Highlights
- No significant association was found between first-trimester Z-drug exposure and the overall risk of major congenital malformations (adjusted RR 1.01).
- The study utilized data from over 4.2 million pregnancies, making it one of the largest investigations into fetal safety for this drug class.
- Zolpidem was the most common Z-drug used, representing 92.1% of the exposed group.
- Imprecise signals for specific defects like neural tube defects and abdominal wall defects were not consistently replicated across different insurance cohorts.
Background: The Challenge of Managing Insomnia During Pregnancy
Sleep disturbances are exceptionally common during pregnancy, with prevalence rates rising as gestation progresses. It is estimated that up to 75% of pregnant individuals experience significant insomnia by the third trimester. Beyond the immediate impact on maternal quality of life, untreated chronic sleep deprivation is associated with adverse obstetric outcomes, including gestational diabetes, preeclampsia, prolonged labor, and an increased risk of postpartum depression.
The Shift Toward Z-Drugs
While cognitive-behavioral therapy for insomnia (CBT-I) is the recommended first-line treatment, pharmacological intervention is frequently required for severe or refractory cases. In recent decades, clinicians have shifted away from traditional benzodiazepines—which have faced scrutiny regarding potential links to oral clefts and neonatal withdrawal—toward nonbenzodiazepine sedative-hypnotics, colloquially known as Z-drugs (zolpidem, eszopiclone, and zaleplon). These agents are preferred for their shorter half-lives and more targeted action on GABA-A receptor subunits. However, evidence regarding their teratogenic potential during the first trimester, the critical period for organogenesis, has remained limited and often contradictory.
Study Design and Population Characteristics
This population-based cohort study, published in JAMA Psychiatry, sought to provide definitive data on the fetal safety of Z-drugs. Researchers analyzed health care utilization data from two massive US databases: the Medicaid Analytic eXtract (2000-2018), representing publicly insured individuals, and the Merative MarketScan database (2003-2020), representing those with commercial insurance.
Participants and Exposure Definition
The study included a total of 4,281,579 pregnancies resulting in liveborn infants. The researchers focused on maternal enrollment from 90 days prior to pregnancy through 30 days post-delivery. Exposure was defined as the dispensing of at least one Z-drug (zaleplon, eszopiclone, or zolpidem) during the first trimester, which is the most sensitive period for the development of structural malformations.
Methodological Rigor: Addressing Confounding
A primary challenge in obstetric pharmacoepidemiology is confounding by indication. Individuals who require sedative-hypnotics often have co-occurring psychiatric conditions or use other medications (such as SSRIs or antipsychotics) that could independently influence birth outcomes. To mitigate this, the study employed propensity score fine stratification weights. This advanced statistical approach allowed the researchers to balance over 50 potential confounders, including maternal age, lifestyle factors, and underlying medical and psychiatric diagnoses, ensuring that the comparison between exposed and unexposed groups was as fair as possible.
Key Findings: Reassurance Across the Board
Of the more than 4 million pregnancies identified, 11,652 in the Medicaid cohort and 10,862 in the MarketScan cohort were exposed to Z-drugs during the first trimester. Zolpidem was the predominant agent used in both groups.
Overall Malformation Risk
The primary outcome—the risk of any major congenital malformation—showed no meaningful increase in the exposed group. The adjusted pooled relative risk (RR) was 1.01 (95% CI, 0.95-1.08). This near-unity result provides strong evidence that Z-drug exposure in early pregnancy does not significantly elevate the baseline risk of structural birth defects.
Organ-Specific Observations
The researchers also performed a granular analysis of organ-specific malformations. While the overall results were reassuring, some specific associations warranted closer inspection:
Abdominal Wall and Neural Tube Defects
Adjusted pooled RRs were slightly elevated for abdominal wall defects (RR 1.46; 95% CI, 0.89-2.38) and neural tube defects (RR 1.62; 95% CI, 0.96-2.74). However, these estimates were imprecisely calculated due to the rarity of these events, as evidenced by the wide confidence intervals. Most importantly, these signals were primarily observed in the Medicaid cohort and were not replicated in the MarketScan cohort. This inconsistency suggests that the findings may be the result of residual confounding or random chance rather than a direct teratogenic effect of the medication.
Cardiac and Oral Clefts
No significant associations were found for cardiac malformations, including conotruncal defects, or for oral clefts, which have historically been a point of concern for GABA-ergic medications.
Expert Commentary and Clinical Interpretation
The findings from Fung et al. provide a substantial leap forward in our understanding of Z-drug safety. For the physician-scientist, the lack of a consistent signal across two distinct insurance populations is a powerful indicator of safety. The Medicaid population often carries a higher burden of social determinants of health and comorbidities, which can sometimes produce ‘false positive’ signals in epidemiological data that disappear when compared against a more affluent, commercially insured population.
Biological Plausibility
From a mechanistic standpoint, Z-drugs selectively bind to the alpha-1 subunit of the GABA-A receptor. This specificity might explain why they do not appear to interfere with the complex signaling pathways required for early embryonic development in the same way that less selective agents might. However, clinicians should remain mindful that ‘safety’ regarding structural malformations does not automatically imply safety regarding functional or neurodevelopmental outcomes.
Clinical Practice Recommendations
For clinicians managing pregnant patients with insomnia, these results support a pragmatic approach:
1. First-Line Non-Pharmacological Therapy
CBT-I and sleep hygiene improvements remain the gold standard and should always be initiated first.
2. Risk-Benefit Discussion
When sleep deprivation is severe enough to threaten maternal health or safety, the use of zolpidem can be considered. This study allows clinicians to provide evidence-based reassurance that the risk of major structural birth defects is not meaningfully increased.
3. Dose Optimization
As with all medications in pregnancy, the ‘lowest effective dose for the shortest duration’ principle should be followed.
Conclusion and Future Directions
The study by Fung and colleagues offers a high level of reassurance regarding the use of Z-drugs in the first trimester. With a pooled RR of 1.01, the data suggests that these medications are not major teratogens. While the initial signals for neural tube and abdominal wall defects require further monitoring in even larger or pooled international datasets, the current evidence supports the judicious use of Z-drugs when clinically indicated. Future research should now pivot toward assessing long-term neurodevelopmental milestones and behavioral outcomes in children with in utero exposure to ensure a comprehensive safety profile.
References
1. Fung K, Straub L, Bateman BT, et al. Z-Drug Use in the First Trimester of Pregnancy and Risk of Congenital Malformations. JAMA Psychiatry. 2025;82(2):123-132. doi:10.1001/jamapsychiatry.2025.3874. 2. Huybrechts KF, Bateman BT, Palmsten K, et al. Antidepressant use late in pregnancy and risk of persistent pulmonary hypertension of the newborn. JAMA. 2015;313(21):2142-2151. 3. American College of Obstetricians and Gynecologists (ACOG). Practice Bulletin No. 92: Use of Psychiatric Medications During Pregnancy and Lactation. Obstet Gynecol. 2008 (Reaffirmed 2023).