Plasma lipidomics in early pregnancy and risk of gestational diabetes mellitus: a prospective nested case–control study in Chinese women
Yi Wang,Yichao Huang,Ping Wu,Yi Ye,Fengjiang Sun,Xue Yang,Qi Lu,Jiaying Yuan,Yan Liu,Huayan Zeng,Xingyue Song,Shijiao Yan,Xiaorong Qi,Chun-Xia Yang,Chuanzhu Lv,Jason H Y Wu,Gang Liu,Xiong-Fei Pan,Da Chen,An Pan,Yi Wang,Yichao Huang,Ping Wu,Yi Ye,Fengjiang Sun,Xue Yang,Qi Lu,Jiaying Yuan,Yan Liu,Huayan Zeng,Xingyue Song,Shijiao Yan,Xiaorong Qi,Chun-Xia Yang,Chuanzhu Lv,Jason H Y Wu,Gang Liu,Xiong-Fei Pan,Da Chen,An Pan
DOI: https://doi.org/10.1093/ajcn/nqab242
2021-09-03
Abstract:ABSTRACT Background Lipid metabolism plays an important role in the pathogenesis of diabetes. There is little evidence regarding the prospective association of the maternal lipidome with gestational diabetes mellitus (GDM), especially in Chinese populations. Objectives We aimed to identify novel lipid species associated with GDM risk in Chinese women, and assess the incremental predictive capacity of the lipids for GDM. Methods We conducted a nested case–control study using the Tongji-Shuangliu Birth Cohort with 336 GDM cases and 672 controls, 1:2 matched on age and week of gestation. Maternal blood samples were collected at 6–15 wk, and lipidomes were profiled by targeted ultra-HPLC–tandem MS. GDM was diagnosed by oral-glucose-tolerance test at 24–28 wk. The least absolute shrinkage and selection operator is a regression analysis method that was used to select novel biomarkers. Multivariable conditional logistic regression was used to estimate the associations. Results Of 366 detected lipids, 10 were selected and found to be significantly associated with GDM independently of confounders: there were positive associations with phosphatidylinositol 40:6, alkylphosphatidylcholine 36:1, phosphatidylethanolamine plasmalogen 38:6, diacylglyceride 18:0/18:1, and alkylphosphatidylethanolamine 40:5 (adjusted ORs per 1 log-SD increment range: 1.34–2.86), whereas there were inverse associations with sphingomyelin 34:1, dihexosyl ceramide 24:0, mono hexosyl ceramide 18:0, dihexosyl ceramide 24:1, and phosphatidylcholine 40:7 (adjusted ORs range: 0.48–0.68). Addition of these novel lipids to the classical GDM prediction model resulted in a significant improvement in the C-statistic (discriminatory power of the model) to 0.801 (95% CI: 0.772, 0.829). For every 1-point increase in the lipid risk score of the 10 lipids, the OR of GDM was 1.66 (95% CI: 1.50, 1.85). Mediation analysis suggested the associations between specific lipid species and GDM were partially explained by glycemic and insulin-related indicators. Conclusions Specific plasma lipid biomarkers in early pregnancy were associated with GDM in Chinese women, and significantly improved the prediction for GDM.
nutrition & dietetics