Among more than 120,000 mothers identified in one of Japan’s largest health insurance claims databases, prescriptions for antipsychotic medications fell steadily during pregnancy and then rebounded sharply after delivery, according to a large retrospective study published in Pharmacology Research & Perspectives. The analysis, led by researchers using the JMDC claims database covering January 2005 through February 2022, found that roughly 67 of every 10,000 deliveries involved an antipsychotic prescription during pregnancy, a rate higher than reported in European cohorts but lower than in the United States. The study is the first comprehensive evaluation of antipsychotic prescribing before, during, and after pregnancy in Japan using a large administrative database.
The research team identified 4,381,720 individuals of reproductive age among 13,470,546 people in the database, then linked 335,321 of them to their children through shared health insurance enrollment. After applying eligibility criteria requiring estimated pregnancy onset and delivery dates and continuous insurance membership from 180 days before pregnancy through 180 days postpartum, 120,735 mothers remained for analysis. Because Japanese claims data lack exact dates of pregnancy onset and delivery, the researchers estimated these using validated algorithms, assigning delivery dates from diagnostic and procedural records for 52.5 percent of mothers and from the 15th day of the neonatal birth month for 39.7 percent.
Prescription prevalence was measured across defined windows: 180 to 91 days and 90 to 1 days before pregnancy, each trimester, and 1 to 90 days and 91 to 180 days postpartum. Before pregnancy, antipsychotics were prescribed to 818 mothers (67.8 per 10,000 deliveries) in the earlier window and 795 mothers (65.8 per 10,000) in the later window. During pregnancy, prevalence dropped to 656 mothers (54.3 per 10,000) in the first trimester and 415 mothers (34.4 per 10,000) in the second trimester, then leveled off in the third. After delivery, prescriptions rose again to 820 mothers (67.9 per 10,000) at 1 to 90 days postpartum and 714 mothers (59.1 per 10,000) at 91 to 180 days, returning to near pre-pregnancy levels.
The most frequently prescribed oral antipsychotics before pregnancy were aripiprazole at 28.3 per 10,000 deliveries, sulpiride at 17.1 per 10,000, and quetiapine at 14.2 per 10,000. During pregnancy, aripiprazole remained the leading agent at 23.5 per 10,000, followed by quetiapine at 14.5 per 10,000 and olanzapine at 11.6 per 10,000. Postpartum, sulpiride moved to the top at 24.8 per 10,000, with aripiprazole at 21.5 per 10,000 and olanzapine at 15.7 per 10,000. Oral formulations dominated over injectable and transdermal routes throughout the study period, with long-acting injectable antipsychotics prescribed at 0.4 per 10,000 deliveries or fewer across all windows.
Injectable antipsychotic prescriptions followed a distinct temporal pattern. Short-acting injectable formulations were prescribed less frequently during the first and second trimesters but increased in the third trimester, reaching 51 prescriptions per 10,000 deliveries for any injectable antipsychotic in that window compared with 8 per 10,000 in the first trimester. Chlorpromazine injections, recorded at 3.1 per 10,000 deliveries during pregnancy, were almost entirely concentrated in the third trimester at 3.0 per 10,000. The researchers suggest this pattern indicates short-acting injectables were used primarily for management of acute symptoms near delivery, while long-acting injectables, whose prescribing frequency changed little across the perinatal period, may have served as maintenance treatment similar to oral formulations.
New prescriptions, defined as the first antipsychotic dispensing after a baseline window of 180 to 91 days before pregnancy with no recorded antipsychotic use, were uncommon during pregnancy but increased markedly postpartum, particularly within 90 days of delivery. Most new postpartum prescriptions involved oral sulpiride, which the authors note may reflect its off-label use as a galactagogue to promote lactation rather than treatment of primary psychiatric indications. Although the overall number of prescriptions appeared similar before pregnancy and after delivery, the postpartum total included a substantially larger proportion of new users.
To examine prescribing behavior after pregnancy recognition, the researchers defined 5 to 13 weeks of gestation as the period when pregnancy is typically confirmed in clinical practice. Among 611 women who received antipsychotic prescriptions during 0 to 4 weeks of gestation, 482 (78.9 percent) continued prescriptions during 5 to 13 weeks, while 129 (21.1 percent) discontinued. By delivery, 334 women (54.7 percent) had continued prescriptions throughout, and 110 (18.0 percent) had discontinued before delivery. Some women who discontinued later resumed treatment before delivery, which the authors suggest may indicate symptom relapse or worsening necessitating renewed therapy.
Continuation patterns varied substantially by psychiatric diagnosis recorded before 4 weeks of gestation. Among women whose prescriptions continued until delivery, schizophrenia (ICD-10 code F20) was the most common diagnosis, present in 267 of 334 women (79.9 percent). Among those who discontinued, depressive episodes (F32) predominated, affecting 82 of 110 women (74.5 percent). Continuation rates until delivery were highest among women with schizophrenia at 63.6 percent, followed by bipolar disorder at 51.0 percent. However, nearly half of women with bipolar disorder and one-third of those with schizophrenia discontinued treatment before delivery, a pattern the authors flag as concerning because interruption of maintenance therapy in these conditions is strongly associated with relapse and hospitalization, particularly during pregnancy when exacerbation risk may be elevated.
Notably, oral and injectable haloperidol, which is contraindicated during pregnancy in Japan based on case reports of limb malformations and teratogenic findings in animal studies, was still prescribed to some women in the cohort. Haloperidol was recorded at 1.2 per 10,000 deliveries during pregnancy as an injectable and appeared among oral prescriptions as well. Bromperidol is similarly contraindicated in Japan, though haloperidol is not contraindicated during pregnancy in the United States. The authors state that this finding highlights the need for further accumulation of safety data and careful risk-benefit evaluation in clinical practice.
The study has several limitations. Claims data cannot capture disease severity, detailed clinical indications for each prescription, or medication adherence, and dispensed medications do not necessarily indicate actual ingestion, meaning continuation may have been overestimated. The database primarily covers employees of relatively large companies and their dependents, and mother-infant linkage required enrollment under the same insurer, introducing potential selection bias. The identified pregnancies represent approximately 1.1 percent of all live births in Japan during the study period. Mothers whose pregnancies ended in abortion or stillbirth could not be evaluated because the design relied on linkage to liveborn children. Clozapine, approved only for treatment-resistant schizophrenia in Japan and subject to strict monitoring requirements, was absent from the cohort, which the authors attribute to the low frequency of exposure rather than incomplete database capture. Despite these constraints, the researchers conclude that the findings provide the first comprehensive description of perinatal antipsychotic prescription patterns in Japan and may inform perinatal mental health care beyond the Japanese setting, with additional evidence needed to support risk-benefit assessments and shared decision-making.
Subject of Research: Perinatal antipsychotic prescription patterns in Japan using a large health insurance claims database
Article Title: Antipsychotic Prescriptions for Prenatal and Postpartum Women in Japan: A Large Claims Database Study
Article References: Obara, R., Sakai, T., Ishikawa, T., Morishita, K., Adomi, M., Hara, A., Kikuchi, S., Kobayashi, N., Iwama, N., Shinoda, G., Noda, A., Orui, M., Ishikuro, M., Tomita, H., Mano, N., Kuriyama, S., & Obara, T. (2026). Antipsychotic Prescriptions for Prenatal and Postpartum Women in Japan: A Large Claims Database Study. Pharmacology Research & Perspectives, 14(6), Article e70336. https://doi.org/10.1002/prp2.70336
Image Credits: AI Generated
DOI: 10.1002/prp2.70336
Keywords: antipsychotics, pregnancy, postpartum, Japan, claims database, schizophrenia, bipolar disorder, depressive episodes, aripiprazole, sulpiride, pharmacoepidemiology, perinatal mental health
News Source: Harold Sullivan. (November 6, 2026). Antipsychotic Prescriptions Drop in Pregnancy, Surge After Birth in Japanese Claims Study. Scienmag.
