| Aims: In recent years,growing evidence suggests that gut-derived metabolites such as bile acids(BAs),trimethylamine N-oxide(TMAO)and its precursors were associated with metabolic pathways involved in glucose,lipoprotein,and energy metabolism.This study explored whether serum levels of these metabolites are associated with risks of gestational diabetes mellitus(GDM),macrosomia or large for gestational age infant(LGA),specifically to address associations:(1)between BA metabolites and risk of GDM;(2)between TMAO/its precursors and risk of GDM;(3)between BA metabolites and risk of macrosomia/LGA;and(4)between TMAO/its precursors and macrosomia/LGA.Methods: A birth cohort of over 20000 pregnant women and child pairs were established in Tianjin,China from 2010 to 2012.Of them,2764 pregnant women who had a clear GDM diagnosis had blood samples stored at early pregnancy.A 1:1 age-matched nested case control study was organized.The 243 women with diagnosed GDM were used as the cases and 243 women without GDM matched on +/-1 year of age of the case were used as the controls.Conditional binary Logistic regression was used to estimate the risks of BAs for GDM and restricted cubic spline(RCS)analysis nested in the logistic analysis was used to detect cutoff points of BA for the risk of GDM.At last,the area under the receiver operating characteristic curves(AUC)was performed to check whether BAs were able to be potential biomarker for early prediction of GDM.(2)The same methods above were used to check associations of TMAO/its precursors with the risk of GDM.Two new indicators,i.e.,TMA ratio and TMAO ratio,were defined to explore the significance of TMA and TMAO in the etiology of GDM.Additive interaction between high TMA ratio and low TMAO ratio,and mathematical interaction between high TMA ratio and high TMAO ratio were examined,to separately suggest whether TMA(suggested by significant additive interaction)or TMAO(suggested by significant mathematical interaction)is more likely to play a causal role in etiology of GDM.(3)Data of 467 pregnant women with delivery outcomes in the above case-control study were used to address associations of BAs with macrosomia/LGA.Binary Logistic regression with Firth penalty was used to estimate the OR and 95% CI of BAs for the risk of macrosomia/LGA.The additive interaction model was used to evaluate the interaction between BAs and GDM.In addition,subgroup analysis by lifestyle intervention was performed to check whether intervention intensity of GDM attenuated the associations of BAs with macrosomia/LGA if any;(4)Using the same sample above,the same methods were used to check the associations of TMAO/its precursors with the risk of macrosomia/LGA,and the additive interaction between TMAOs and GDM for macrosomia/LGA,and whether intervention of GDM attenuated these associations.Results:(1)A majority of BAs were associated with GDM in a non-linear manner,among which glycoursodeoxycholic acid(GUDCA)and deoxycholic acid(DCA)had a clear threshold effect on GDM.In multivariable analysis,women with GUDCA ≤0.070 nmol/mL had 6.84-fold(95%CI: 1.10-42.48)risks for GDM as compared to women with GUDCA >0.070 nmol/mL.DCA ≤0.280 nmol/mL had 2.06-fold(95%CI: 1.26-3.37)risks for GDM as compared to DCA >0.280 nmol/mL.Inclusion of GUDCA and DCA in the traditional risk factors model,the AUC increased from 0.69(95%CI: 0.64-0.74)to 0.76(95%CI: 0.71-0.80)(P<0.001).(2)TMAO/its precursors were associated with the risk of GDM in a non-linear manner.TMA ratio was positively associated with GDM with a clear threshold,and TMA ratio >0.35 associated with 6.54-fold(95% CI: 3.68-11.6)increased risk of GDM in multivariable analysis.However,TMAO ratio was inversely associated with GDM with a clear threshold,and TMAO ratio ≤0.15 associated with 4.56-fold(95% CI: 2.69-7.71)increased risk of GDM in multivariable analysis.Co-presence of higher TMA ratio(ratio >0.35)and lower TMAO ratio(ratio ≤0.15)had a synergistic effect on the risk of GDM with significant additive interaction,i.e.,an OR of 1.71(95% CI: 0.42-6.95)for high TMA only or 2.06(95% CI: 1.09-3.90)for low TMAO ratio only,increased to 11.16(95% CI: 5.45-22.8).However,the mathematical interaction between the two conversion ratios was non-significant.Inclusion of TMAO/its precursors in the traditional risk factors model,the AUC increased to 0.88(95% CI: 0.84-0.91)(P<0.001).(3)In multivariable analysis,the OR of DCA >0.31 nmol/mL for macrosomia was 2.19(95%CI: 1.10-4.39),and for LGA was 2.63(95%CI: 1.36-5.06)as compared with the lowest group.Lifestyle intervention tended to attenuate the association between DCA and macrosomia/LGA.(4)In multivariable analysis,the OR of TMAO>19.51 nmol/mL for macrosomia was 3.10(95% CI: 1.46-6.59),for LGA was 1.97(95% CI: 0.99-3.93);the OR of betaine >297.63 nmol/mL for macrosomia was 2.40(95% CI: 1.16-4.96),for LGA was 1.91(95% CI: 0.96-3.78);the OR of cholinechloride >165.45 nmol/mL for macrosomia was 2.25(95% CI: 1.09-4.66),for LGA was 2.52(95% CI: 1.25-5.08).Co-presence of betaine >297.63 nmol/mL and GDM markedly increased the ORs for macrosomia,i.e.the ORs of betaine >297.63 nmol/mL only(1.01,95%CI: 0.42-2.42)and GDM only(1.10,95% CI: 0.49-2.46)increased to 3.39(95%CI: 1.31-8.77),with significant interaction.Similarly,copresence of TMAO >19.51 nmol/mL and GDM also significantly increased the ORs for macrosomia from the high TMAO only(1.06,95%CI: 0.44-2.57)or GDM only(1.16,95%CI: 0.52-2.56)to 2.80(95%CI: 1.14-6.89),with significant interaction.While,the interaction of cholinechloride(or TMAO)and GDM for LGA was not significant.Lifestyle intervention attenuated the associations of betaine,cholinechloride and TMAO with macrosomia/LGA.Conclusions:(1)Serum BAs in early pregnancy were associated with GDM.GUDCA ≤0.07 nmol/mL and DCA ≤0.028 nmol/mL were associated with independently increased risks of GDM;(2)Serum TMAO/its precursors in early pregnancy were associated with GDM independently.Co-presence of TMA ratio >0.35 and TMAO ratio ≤0.15 had a synergistic effect towards a markedly increased risk of GDM,suggesting that TMA was more likely to play a causal role in etiology of GDM;(3)Serum DCA level in early pregnancy was associated with macrosomia and LGA;and(4)High levels of TMAO,betaine and cholinechloride in early pregnancy were associated with increased risks of macrosomia and LGA.These effects were exemplified by presence of GDM. |
PDF Full Text Request