Safety of herbal medicines used in early gestations among the Chinese population: A population-based cohort study

Background: Herbal medicines have a long history of use for pregnant women around the world. However, their use in the early pregnancy is often questioned in terms of safety on offspring. Purpose: To investigate whether herbal medicines used at early pregnancy are associated with an increased risk of birth defects. Study design: The population-based retrospective cohort study included pregnancies registered in a populationbased and long-term follow-up database (i.e., REPRESENT) with live births between January 2013 and December 2018, by linking a population-based drug prescription database covering all the hospitals in Xiamen, China. Methods: The exposure of interest was defined as the consecutive administration of one or more prescriptions of herbal medicines to prevent miscarriage during the first trimester. We examined the overall birth defects (excluding chromosomal malformations) and 23 individual types by national surveillance. To facilitate comparison, we included blank (non-use) and active controls. Poisson regression based on propensity score matching (PSM) was applied to control for the confounders. Multiple sensitivity analyses and negative control analyses were conducted to examine the robustness. Results: The final cohort included 195,824 pregnancies, and 29,063 (14.8 %) were prescribed with herbal medicines at early pregnancy, of which 3,024 (1.5 %) received herbal medicine monotherapy. These herbal medicines were often used in combination with other medicines (especially pharmaceutical medicines) and peaked between the fifth and eighth gestational weeks. A total of 2,795 birth defects were identified. Compared to progesterone monotherapy (i.e., active control), herbal medicine monotherapy was not associated with a higher risk of overall birth defects (relative risk [RR] 1.28, 95 %CI 0.57-2.90). Compared to no use of drug (i.e., blank control), pregnancies with herbal medicine monotherapy showed a similar risk (RR 1.25, 95 %CI 0.69-2.29). Consistent results were shown in sensitivity analyses. Conclusion: The current evidence dose not suggest an increased risk of birth defects when using herbal medicines in early pregnancy, although larger studies with wider pregnancy populations are needed for further confirmation.