First-trimester maternal triglycerides and atherogenic index of plasma in relation to large for gestational age and macrosomia among non-diabetic pregnant women: A propensity score-matched retrospective cohort study.

BACKGROUND: Fetal overgrowth can occur without pregestational diabetes or GDM, indicating a potential role of non-glucose metabolic factors. This study examined first-trimester TG status and AIP in relation to LGA and macrosomia among non-diabetic pregnant women. METHODS: This single-center retrospective cohort study included women with singleton term live births and fasting lipid profiles obtained at 10 + 0 to 14 + 0 gestational weeks between January 2023 and December 2025. Women who had pregestational diabetes, GDM, multiple gestation, significant maternal disease based on lipids, fetal anomalies, stillbirth or lack of essential data were excluded. High TG was defined as TG >1.72 mmol/L, corresponding to the 75th percentile of the cohort-specific distribution. LGA was defined as birth weight ≥90th percentile for gestational age and neonatal sex, and macrosomia was defined as birth weight ≥4000 g. Propensity score matching was performed at a 1:1 ratio, and logistic regression was used to estimate associations with LGA and macrosomia. AIP was estimated to be log10 (TG/HDL-C) and examined using quartiles and restricted cubic splines. RESULTS: There were 3433 women, 858 with high TG and 2575 with normal TG; 831 pairs were matched. High TG was associated with higher odds of LGA (15.2% vs. 9.9%; OR = 1.55, 95% CI: 1.15-2.09) and macrosomia (9.3% vs. 5.8%; OR = 1.58, 95% CI: 1.07-2.32). Similar associations were observed when TG was analyzed as a continuous variable. Compared with the lowest AIP quartile, the highest quartile was associated with LGA (OR = 1.73, 95% CI: 1.25-2.40) and macrosomia (OR = 1.68, 95% CI: 1.12-2.53). The primary LGA findings were confirmed by sensitivity and subgroup analyses. CONCLUSIONS: Higher first-trimester TG levels and AIP were associated with greater odds of LGA and macrosomia among non-diabetic pregnant women. These lipid-related markers warrant further evaluation as potential components of future fetal-overgrowth prediction models, but the present study did not assess clinical prediction or screening performance.