History of large-for-gestational-age birth is independently associated with subsequent gestational diabetes in Chinese multiparous women: A retrospective cohort study

Juan Yang^1,2,3Shilin Zhong^1,2,3*Min Wang^1,2,3Yuzhen Liu^1,2,3Ying Wang^1,2,3Yuqing Deng^1,2,3Kaidong Ma^1,2,3Ao Yang^1,2,3Yanmei Li⁴Chang Xu⁵

• ¹Center of Obstetrics and Gynecology, Peking University Shenzhen Hospital, Shenzhen, China
• ²Institute of Obstetrics and Gynecology, Shenzhen PKU-HKUST Medical Center, Shenzhen, China
• ³Shenzhen Key Laboratory on Technology for Early Diagnosis of Major Gynecologic Diseases, Shenzhen, China
• ⁴Department of Obstetrics and Gynecology, Guangzhou Women and Children's Medical Center, Guangzhou Medical University, Guangzhou, China
• ⁵Intelligent Hospital Research Academy, Peking University Shenzhen Hospital, Shenzhen, China

Background: While prior macrosomia has been linked to subsequent gestational diabetes mellitus (GDM), the association between a history of large-for-gestational-age (LGA) births and subsequent GDM, as well as its comparative relevance versus macrosomia, remain underexplored in Chinese populations, highlighting a critical gap in perinatal risk stratification. Aim: To determine whether a history of LGA (f-LGA) birth is independently associated with subsequent GDM (s-GDM) and compare its associative relevance versus macrosomia in Chinese multiparous women. Method: This retrospective cohort study included women with consecutive singleton deliveries (2002-2024) at Peking University Shenzhen Hospital. Exposures were LGA (birth weight ≥90th percentile, China 2020 standards) or macrosomia (≥4000 g) in the first pregnancy; outcome was s-GDM (IADPSG criteria). Associations were evaluated using restricted cubic splines (RCS) and multivariable logistic regression with sequential adjustment (unadjusted, minimally adjusted, and fully adjusted). Stratified analyses assessed effect modification by prior GDM, maternal age, and pre-pregnancy BMI in the first pregnancy. Results: Among 3,110 women analyzed, RCS analyses revealed a significant monotonic increase in s-GDM risk with higher birth weight percentiles (P for association < 0.001, P for nonlinearity = 0.010 after full adjustment). Although absolute birth weight categories showed a statistically significant overall association (P for association = 0.033), no dose-response pattern was observed.. After full adjustment, prior LGA remained significantly associated with s-GDM (aOR = 1.458, 95% CI: 1.045–2.035), whereas prior macrosomia did not (aOR = 1.271, 95% CI: 0.782–2.067). Significant positive trends existed across percentile categories (P for trend = 0.025) but not absolute weight categories (P for trend = 0.310). Stratified analyses demonstrated robust associations of f-LGA with s-GDM specifically in women without prior GDM (aOR = 1.544, 95% CI: 1.069–2.232), aged <35 years (aOR=1.668, 95% CI: 1.095–2.542), and with normal pre-pregnancy BMI (aOR=1.757, 95% CI: 1.169–2.641) (P for trend < 0.001 for all strata). Conclusion A history of LGA birth is independently associated with s-GDM risk in Chinese multiparous women, particularly among those without prior dysglycemia, aged <35 years, or with normal BMI. The percentile-defined LGA definition demonstrates superior associative relevance compared to absolute macrosomia, supporting its application in precision screening for multiparous pregnancies.