Adverse pregnancy outcome burden patterns and early postpartum cardiometabolic risk in gestational diabetes
Y W Fang, J G Li, R M Zhang, P F Sun, Y Chang, X Han, H K Liu, L S Zuo, L J Li, G R Aa, C Han, L J Lv, Q Yang, X ZhouAbstract
Background
Individual adverse pregnancy outcomes (APOs) confer increased postpartum cardiometabolic risk after gestational diabetes mellitus (GDM), yet the prognostic implications of co-occurring patterns remain unclear. We derived clinically interpretable APO phenotypes and integrated them with cumulative burden to stratify postpartum cardiometabolic risk.
Methods
In a multicenter retrospective cohort of 82,366 women with GDM in Tianjin, China (2008–2023), we used covariate-adjusted Cox models to examine associations of latent class-derived APO classifications, APO counts, and their joint framework with incident postpartum diabetes, cardiometabolic morbidity (incident diabetes, hypertension or dyslipidemia), and stage-4 cardiovascular-kidney-metabolic syndrome.
Results
Over a median 4.5-year follow-up, individual APOs conferred modest risk (most hazard ratios <2.0), whereas co-occurring combinations showed marked heterogeneity, ranging up to ~10-fold relative to GDM without APOs. Using prespecified (consensus-based) APOs, we identified five latent classes that replicated across GDM criteria and were confirmed externally (Guangzhou cohort). Mutual information and class projection identified preterm birth and fetal growth (small-for-gestational-age [SGA]/appropriate-for-gestational-age [AGA]/large-for-gestational-age [LGA]) as core structural determinants of class separation, motivating a six-category classification (preterm/term × SGA/AGA/LGA). Integrating this phenotype with APO counts into a two-dimensional framework revealed stepwise gradients attributable to cumulative burden and specific profiles, yielding a 2.5-fold gradient in 5-year diabetes risk (2.9%–7.3%) and 2.2-fold gradients for other cardiometabolic outcomes. Findings were consistent in Fine-Gray models.
Conclusions
APO burden patterns provide a scalable tool to stratify early postpartum cardiometabolic risk after GDM. Using routinely available delivery data, this framework supports standardized triage linking obstetric and cardiometabolic care for prevention.Graphical Abstract_new.jpg