1056Maternal dyslipidaemia is associated with shorter gestational length in a UK cohort of 7440 pregnancies

Abstract

Abstract Background Mechanisms for human labour are unknown because of difficulties in human pregnancy experimentation, limiting our ability to prevent preterm birth. Maternal metabolism is potentially involved. This study aimed to explore associations of multiple maternal metabolic traits with gestational age at delivery (GA). Methods Women with singleton pregnancies recruited to the Born in Bradford cohort study were included. A total of 157 maternal blood metabolites sampled between 26-28 weeks were measured using high-throughput NMR metabolomics. Associations between each metabolite and GA was modelled using linear (GA continuous) and logistic (GA binary) regressions; adjusted for age, BMI, ethnicity, socioeconomical status, alcohol, smoking, parity, pre-existing and gestational diabetes and hypertension, pre-eclampsia, and labour onset. Results The complete case sample included 7440 pregnancies (12308 eligible; 4540 had missing data). 1SD increases in large and very large HDLs were associated with longer mean GAs of 0.5-1 day, including 1SD increased HDL mean diameter associated with +0.5 day mean GA (95%CI:0.2to0.8;p=9.1E-4). 1SD increases in VLDLs and LDLs were associated with shorter mean GAs of 0.5-1 day, including 1SD increase in large VLDL associated with -0.7 days difference in GA (95%CI:-1.021to-0.465;p=2.299E-07). Conclusions Our findings suggest for the first time that maternal dyslipidaemia is related to differences in GA even after adjusting for multiple key confounders. Further studies are needed to clarify whether lipoprotein metabolism is causally involved in human labour. Key messages Human labour may involve pathways related to lipoprotein metabolism.