Research Associate National Taiwan University Taipei, Taiwan (Republic of China)
Background: Maternal use of benzodiazepines (BZDs) has raised serious concerns considering their potentially harmful effects on pregnant women and neonates. As these medications can easily cross the placenta and may accumulate substantially in the embryo and fetal tissues, evaluating the benefits and risks of their use during pregnancy is a public health priority. Previous studies have reported an increased risk of spontaneous abortion (SA) among BZDs users. However, most existing studies did not adequately control for potential confounders.
Objectives: This case-time control study aimed to quantify the risk of SA associated with BZDs use during pregnancy after controlling for unmeasured confounders and exposure time trends.
Methods: The data sources of this study were the National Birth Certificate Application (BCA) and National Health Insurance (NHI) data in Taiwan. Pregnancies resulting in SA between 2004 and 2018 were identified as cases. Discordant exposures to BZDs during the risk period (1-28 days before SA) and the reference period (31-58 days before LMP) were compared for each pregnancy. Exposure was categorized by use of BZDs (yes/no), long- or short-acting BZDs, and specific substances of BZDs agents (e.g. alprazolam). The time trend of BZDs use across pregnancy was adjusted by using a 1:1 disease risk score-matched controls, considering demographic characteristics and pre-pregnancy comorbidities. Conditional logistic regression models were used to estimate odds ratios (ORs) and 95% confidence intervals (CIs). Multiple sensitivity analyses, including negative control designs that redefined the risk periods, were performed to test the robustness of the study results.
Results: Use of BZDs during pregnancy was associated with a 1.81-fold increased risk (aOR 1.81; 95%CI 1.64-2.00) of SA, and similar results were obtained across multiple sensitivity analyses using different time windows. The risk was higher among pregnancies exposed to short-acting BZDs (aOR, 1.89; 95% CI, 1.65-2.17). All BZDs agents were independently associated with an increased risk of SA (range of aORs 1.16-2.52). The negative control, which redefined the risk periods to 91-118 days before conception, yielded a null result (aOR, 0.97 95% CI, 0.89-1.06).
Conclusions: This is the first nationwide case-time-control study that revealed an increased risk of SA associated with maternal use of BZDs considering unmeasured confounders. Given the high prevalence of insomnia, anxiety, and mood disorders during pregnancy, BZDs should only be used after careful consideration of the potential benefits and risks for both the mother and child.
