Gestational hypertriglyceridemia and adverse pregnancy outcomes: A search for cutoffs using generalized additive models

Objective: To explore cutoffs of gestational hypertriglyceridemia based on the risk of adverse pregnancy outcomes. Methods: Pregnant women who visited National Taiwan University Hospital for prenatal care were included. Fasting plasma TG in the first and second trimesters were measured. Adverse pregnancy outcomes, including gestational diabetes and large for gestational age, were recorded and used in simple and multiple generalized additive models (GAM) to identify cutoffs for gestational hypertriglyceridemia. Results: We recruited 807 pregnant woman-newborn pairs. Using GAM analyses, we identified plasma TG at 95 or 140 mg/dL (1.07 or 1.58 mmol/L) in the first trimester, and 173 or 220 mg/dl (1.95 or 2.48 mmol/L) in the second trimester as potential cutoffs. Gestational hypertriglyceridemia defined by the higher cutoffs in both trimesters were associated with adverse pregnancy outcomes and had a more reasonable prevalence and better specificity than the lower cutoffs (First trimester plasma TG ≥ 140 mg/dL, adjusted OR 2.56, 95% CI 1.17–5.69, p = 0.019, prevalence 19%, specificity 83%; Second trimester plasma TG ≥ 220 mg/dL, adjusted OR 1.70, 95% CI 1.00–2.87, p = 0.049, prevalence 19%, specificity 81%). Conclusions: Fasting plasma TG ≥ 140 mg/dL in the first trimester and ≥ 220 mg/dL in the second trimester can be used as cutoffs of gestational hypertriglyceridemia.