Projected obstetric risks in an aging maternal population in Taiwan: a Bayesian age-period-cohort analysis

Delayed childbearing may alter maternal and neonatal risk profiles and potentially challenge current healthcare strategies. Therefore, we aimed to provide comprehensive Bayesian projections of pregnancy outcomes through 2050 to inform urgent healthcare system adaptation.

We analysed all singleton pregnancies in Taiwan using linked data from National Health Insurance and Birth Certificate Application databases from 2004 to 2021. Temporal trends and age-stratified prevalence were assessed for maternal (gestational hypertension [GHTN], gestational diabetes mellitus [GDM], mortality, sepsis, and postpartum hemorrhage) and neonatal (stillbirth, preterm birth, low birth weight, small for gestational age, and large for gestational age) outcomes. Bayesian age-period-cohort models were applied to project prevalence and absolute burden through 2050.

From 2004 to 2021, total pregnancies declined by 25.3%, while the proportion of pregnancies among women aged ≥35 years nearly tripled (11.8%-34.6%). GHTN prevalence rose from 0.4% to 2.1% and GDM from 2.0% to 9.3%. By 2050, GHTN and GDM are projected to reach 26.4% (95% credible interval [CrI], 5.2%-67.4%) and 60.9% (95% CrI, 28.9%-85.0%), respectively, while the GDM estimate should be interpreted in light of one-step screening adoption after 2010. Preterm birth and low birth weight are projected to increase modestly to 11.2% and 11.9%. BAPC modeling indicates these trends are associated with advancing maternal age and cohort patterns, reflecting evolving risk profiles shaped by shifting healthcare and behavioral practices.

Obstetric risk is shifting due to demographic and generational transitions, requiring prenatal care systems to adopt more proactive and routine management approaches.