Abstract
Maternal lipid levels change dynamically during gestation to support normal fetal growth. To obtain a detailed footprint of these changes and their differences in pregnancies with preterm or small-for-gestational-age (SGA) neonates, we analyzed 641 lipids and 639 metabolites in plasma from women by 24 weeks of pregnancy from three cohorts from low- and middle-income countries: Bangladesh, Zimbabwe, and Kenya. We consistently found a significant lipid imbalance with increased lipid levels that preceded preterm birth and decreased levels that preceded SGA births. Changes were most pronounced in triglycerides, including triglycerides containing proinflammatory omega-6 polyunsaturated fatty acids (PUFAs) in pregnancies with preterm infants. A machine learning model for prediction of preterm birth had modest performance [area under the receiver operator curve (AUC) = 0.69, 95% confidence interval (CI) = (0.68, 0.70)] and lower performance for predicting SGA [AUC = 0.64, CI 95% = (0.62, 0.65)]. Increased triglycerides containing proinflammatory PUFAs provide further evidence in favor of a previously considered dietary supplementation with the long-chain fatty acids.