Benzodiazepine Use in Early Pregnancy Tied to Miscarriage

Batya Swift Yasgur, MA, LSW

May 20, 2019

Exposure to any benzodiazepine in early pregnancy is associated with an increased risk for spontaneous abortion (SA), new research suggests.

Using data from the Quebec Pregnancy Cohort (QPC), investigators matched each pregnancy with five randomized control persons and found that 1.4% of cases of SA took place in women who had been exposed to benzodiazepines in early pregnancy, compared to approximately 0.6% among the matched control persons.

Incident use of any benzodiazepine was associated with an 85% risk for SA. Short-acting agents conferred a greater risk than long-acting agents (81% and 73%, respectively).

"Anxiety and mood disorders need to be treated during pregnancy," Anick Bérard, PhD, professor, Faculty of Pharmacy, Université de Montréal, and director, Research Unit on Medications and Pregnancy, CHU Ste-Justine Research Center, Canada, told Medscape Medical News.

"Given the high prevalence of anxiety and mood disorders during pregnancy, physicians need to evaluate the benefits and risks of using benzodiazepines, given that alternative nonpharmacologic treatments exist," she said.

The study was published online May 15 in JAMA Psychiatry.

Birth Defect Risk

"Benzodiazepines, which are used to treat insomnia, anxiety, and mood disorders, cross the placental barrier and have been shown to alter fetal development in animal studies," Bérard said.

"They are also labeled 'pregnancy category D,' indicating that there is human evidence on the increased risk of birth defects — cleft lip and palate [and] urogenital defects," she added.

She explained that benzodiazepines "are anxiolytics and signaling molecules and are believed to disrupt binding and implantation when used in early pregnancy."

Bérard noted that previous studies have found that benzodiazepines "are associated with an increased risk of miscarriage, but no one has looked at the effect of specific benzodiazepines, and the duration of action — short or long acting."

To investigate the question, the researchers conducted a nested case-control study within the QPC, an online population-based cohort with prospective data collection on all pregnancies of women covered by the Quebec Public Prescription Drug Insurance Plan from January 1, 1998, to December 31, 2015 (n = 442,066 pregnancies in total; age, 15 – 45 years).

Individual health information was obtained from an array of other province-wide databases.

Pregnancies in women who were known to have been exposed to teratogens during the first trimester, as well as pregnancies in women who had a current or past history of epilepsy, or who had previously used benzodiazepines, or who had a history of planned/induced abortions were excluded.

SA was defined as pregnancy loss between the beginning of the 6th week of gestation and the 19th completed week of pregnancy.

The index date was defined as the calendar date of the SA diagnosis.

Potential Confounders

A pregnancy was regarded as being linked to "incident" benzodiazepine exposure if the mother had either filled ≥1 prescriptions for any type of benzodiazepine from the last menstrual period (LMP) date until the index date, or before becoming pregnant with a prescription duration that overlapped the LMP.

Benzodiazepine exposure was categorized on the basis of the agent's duration of action: short acting (half-life ≤24 hours) or long acting (half-life >24 hours).

The short-acting benzodiazepines included in the study were alprazolam (multiple brands), bromazepam (multiple brands), lorazepam (multiple brands), oxazepam (multiple brands), temazepam (multiple brands), and triazolam (Halcion, Pharmacia and Upjohn). The long-acting agents were chlordiazepoxide (multiple brands), clonazepam (Klonapin, Roche), diazepam (multiple brands), flurazepam hydrochloride (Dalmane, Valeant), and nitrazepam (multiple brands).

To control for potential confounding by the main indications for benzodiazepine prescription (mood and anxiety disorders or insomnia), the researches adjusted for the presence of physician-based diagnoses for these conditions during the year before pregnancy until the index date and also adjusted for exposure to antidepressant and antipsychotic medications between the LMP and the index date.

Additional potential confounders included maternal sociodemographic variables; pregnancy-associated variables; maternal chronic conditions; healthcare resource utilization during the year before the LMP and during pregnancy; diagnoses of tobacco, alcohol, and drug dependencies, when available; and exposure to folic acid during the months before the LMP and in early pregnancy.

Of 262,070 eligible pregnancies, the prevalence of SA was 7.0%; 27,149 pregnancies met inclusion criteria and were defined as cases, and 26,789 were paired with five matched control cases.

The mean age of the women who represented case pregnancies was 24.2 years (SD, 6.5 years). The mean gestational age at index date was 16.7 weeks (SD, 3.1 weeks).

There were several important differences between cases and matched control pregnancies.

Women whose pregnancies ended in SA were more likely to be older, welfare recipients, to have hypertension or asthma, and to have received diagnoses of drug, tobacco, and alcohol dependence in the year before or during early pregnancy (up to index date), compared to the matched control persons.

In addition, women who experienced SA were more likely to have undergone hospitalizations, emergency department visits, general practitioner visits, or specialist visits; received a diagnosis of a mood or anxiety disorder in the 12 months before the LMP; experienced concomitant antidepressant or antipsychotic exposures; and have more pregnancies in the year before the LMP, compared to their control counterparts.

They were also less likely than control persons to have been exposed to folic acid before pregnancy.

Accurate Prescription Data

Most women (97.0%) who had been exposed to benzodiazepines during early pregnancy (n = 1163) received only one benzodiazepine agent; 77.1% filled only one prescription.

The most frequently prescribed benzodiazepines were lorazepam and clonazepam (44.8% and 23.4%, respectively).

Overall, 1.4% of the SA case patients had ≥1 prescriptions for benzodiazepines filled during early pregnancy, compared with 0.6% of the matched control persons (crude odds ratio [OR], 2.39; 95% confidence interval [CI], 2.10 – 2.73).

After the researchers adjusted for all potential confounding variables, benzodiazepine exposure in early pregnancy was found to be associated with an increased risk for SA (adjusted OR [aOR], 1.85; 95% CI, 1.61 – 2.12), which was independent of mood and anxiety disorders.

Both short-acting and long-acting benzodiazepines (aOR, 1.81; 95% CI, 1.55 – 2.12; and aOR, 1.73; 95% CI, 1.31 – 2.28, respectively) were associated with an increased risk for SA.

All of the benzodiazepines increased the risk for SA, with aORs ranging 1.13 for women exposed to flurazepam hydrochloride to an aOR of 3.43 for women exposed to diazepam.

The association between benzodiazepine use and SA grew stronger with increasing diazepam-equivalent daily dose (<5 mg: aOR, 1.73; 95% CI, 1.44 – 2.08; 6 – 20 mg: aOR, 1.96; 95% CI, 1.59 – 2.43; and >20 mg: aOR, 2.55 95% CI, 1.08 – 6.01; P < .01).

The authors note that a strength of their study was the ability to obtain accurate information on filled prescriptions, rather than having to rely on maternal recall; and the availability of physician-based diagnoses, which reduced the potential for biases regarding outcome status.

In addition, the association between benzodiazepine use and SA was adjusted for "the most important confounders (ie, concomitant exposure to antidepressants and/or antipsychotic medications)."

One limitation of the study was a lack of information about "potentially important confounders, such as smoking and alcohol intake," the authors point out.

Unmeasured Factor?

Commenting on the study for Medscape Medical News, Kimberly Yonkers, MD, professor of psychiatry, obstetrics, gynecology, reproductive sciences, and epidemiology, Yale School of Public Health, New Haven, Connecticut, who was not involved with the study, said that although the association between benzodiazepine use and SA is "important," the study "does not show cause" and the link should be regarded as a "connection."

Although the association might indeed be causative, it could also be due to an "unmeasured factor," said Yonkers, who is also director of the New Haven Yale Health Division of Psychological Medicine and the Center for Wellbeing of Women and Mothers.

She noted that the rate of SA was higher in older women and in women who had alcohol and tobacco use disorders.

"The authors were able to control for substance use disorders that are documented in the [patient's] chart, but not actual exposures from these and other substances, so that is a relative weakness," she said.

Moreover, "people are prescribed an anxiolytic for a reason, and it may be, as the authors discuss, there is some confounding by indication — ie, it could be the reason the medication was prescribed that at least, in part, contributes to the finding," Yonkers said.

Medical records and physicians' diagnoses "are not sufficient to tease this out, since some doctors may not chart conditions or do it accurately," she said.

Nevertheless, the study does make a contribution to the field by "suggesting caution in the prescription of benzodiazepine use among pregnant women or women planning on becoming pregnant," she said.

The take-home message is that "patients should be questioned about health habits and should be counseled about known and unknown risks of any medication taken in pregnancy," said Yonkers.

"Future research should replicate findings and compare use of benzodiazepines and other nonpharmacologic treatment for anxiety and mood disorders in pregnant women," Bérard added.

The study was funded by the Canadian Institutes of Health Research, a grant from the Canadian Network for Advanced Interdisciplinary Methods, Fonds de la Recherche du Québec–Santé, and the Réseau Québécois de Recherche sur les Medicaments. Bérard has been a consultant for plaintiffs in litigations involving antidepressants and birth defects. The other authors and Yonkers have disclosed no relevant financial relationships.

JAMA Psychiatry. Published online May 15, 2019. Abstract

For more Medscape Psychiatry news, join us on Facebook and Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.