Association between erythrocyte polyunsaturated fatty acids and gestational diabetes mellitus in Chinese pregnant women.

Purpose: To explore the associations of erythrocyte polyunsaturated fatty acids (PUFAs) with the risk of gestational diabetes mellitus (GDM) in pregnant women.

Methods: A total of 951 women in mid-pregnancy (20-28 weeks of gestation) were recruited during 2017-2018. Erythrocyte PUFAs were measured by gas chromatography and expressed as a percentage of total fatty acids. A 75 g oral glucose tolerance test was used for GDM diagnosis. Logistic regression and restricted cubic spline models were conducted.

Results: Among the 951 pregnant women, 180 were diagnosed with GDM (18.93%). For n-3 PUFAs, each standard deviation (SD) increase in α-linolenic acid (ALA) and docosapentaenoic acid (DPA) was associated with multivariable-adjusted odds ratios (ORs) of 0.789 (95% CI: 0.649, 0.961; P = 0.018) and 0.782 (95% CI: 0.638, 0.957; P = 0.017), respectively. However, the associations became marginally significant after post hoc false-discovery rate (FDR) correction (both P_FDR=0.065). A significant nonlinear association was observed for ALA and GDM risk (P_{- nonlinearity}=0.001). For n-6 PUFAs, γ-linolenic acid (GLA) was significantly associated with a 46.0% higher risk of GDM [OR (95%CI): 1.460 (1.195, 1.785), P_FDR=0.003) per SD increase], with a significant non-linear relationship (P_{- nonlinearity}=0.031). Arachidonic acid (AA) showed a borderline significant association with lower GDM risk after FDR correction [OR (95%CI): 0.736 (0.568, 0.953), P = 0.020, P_FDR=0.065].

Conclusion: Erythrocyte GLA was significantly associated with an increased risk of GDM in pregnant women. Furthermore, we observed for the first time significant non-linear associations of erythrocyte ALA and GLA with GDM risk.

Clinical trial registry number: NCT03023293, 2019-01-11.