There is growing concern regarding teratogenic effect of antipsychotics. Previous research assessing association between antipsychotics and congenital malformations (CMs) yielded mixed results and were all derived from Western countries. We aimed to examine risk of major and organ/system-specific CMs associated with prenatal antipsychotic exposure in Hong Kong.

This population-based study identified women aged 15–50 years who delivered their first/singleton child between 2003–2018 from public healthcare service database. Propensity score (PS)-weighted logistic-regression analyses were performed to examine risk of CMs following first-trimester exposure to antipsychotic classes (second- and first-generation antipsychotic; SGA and FGA) and six most frequently-prescribed individual antipsychotics.

Of 465,069 women, 419 and 420 redeemed ≥1 prescription of SGA and FGA during first-trimester, respectively. Prevalence of any CMs was 4.9% (95%CI:4.9–5.0%) in unexposed-infants, 9.1% (6.7–12.3%) in SGA-exposed infants, and 6.2% (4.3–9.0%) in FGA-exposed infants. SGA exposure (adjusted-odds-ratio: 2.11 [95%CI:1.19–3.86]) was associated with increased risk of CMs. This finding was consistent with sensitivity analyses addressing exposure misclassification and confounding by treatment indication, but not with PS-matched sensitivity analysis. Elevated risk of CMs was observed in infants exposed to high-dose olanzapine (7.50 [1.65–36.13]) and high-dose quetiapine (15.03 [4.86–56.72]), but with wide-CIs. Organ/system-specific malformations were not associated with SGA, FGA or individual antipsychotics.

We observed a small increased risk of major malformations associated with SGA, but was not consistently affirmed in sensitivity analyses, precluding firm conclusions. Research with large sample size clarifying comparative safety of individual antipsychotics on specific malformations is warranted.