Longitudinal Iron Status During Pregnancy And Its Association With Pregnancy Outcomes: A Prospective Cohort Study

Background:Physiologic iron requirements increase substantially during pregnancy.Iron supplementation has been recommended widely for pregnant women to prevent iron deficiency;however,unnecessary iron supplementation has raised concerns about iron excess and its detrimental effects.Accumulating evidence has revealed a positive association between plasma ferritin and gestational diabetes mellitus(GDM),nevertheless,dynamic feature of plasma ferritin during pregnancy has not been fully taken account.Although studies support the effect of iron supplementation on improving hematologic indices for pregnant women,it is still controversial about its effect on birth outcomes.TMPRSS6 gene,which plays an important role in the regulation of iron absorption,has been related to ferritin concentrations and type 2 diabetes in aging population.More data are needed to evaluate the impact of TMPRSS6 single nucleotide polymorphisms on plasma ferritin and GDM among pregnant women.Objective:1.To evaluate the association between iron supplement use and longitudinal changes in iron indicators(plasma ferritin and plasma iron).2.To investigate the association of plasma ferritin change and GDM,anemia in late pregnancy,and adverse birth outcomes(i.e.,preterm birth,small for gestational age,large for gestational age).3.To explore the effect of TMPRSS6 genotypes and iron supplement use on plasma ferritin and pregnancy outcomes.Methods:This cohort study was conducted within the Tongji Maternal and Child Health Cohort.The demographic information,medical and obstetrical history,prepregnancy weight and lifestyle factors were collected using structured questionnaires,and current weight and height were measured at enrollment in early pregnancy.Information on supplement use during pregnancy was obtained by questionnaires at enrollment and the follow-up visits.Dietary intake during pregnancy was collected using a validated food frequency questionnaire(FFQ).Iron supplement use in early and mid-pregnancy were separately categorized into no intake,low intake(iron <30 mg/day)and high intake(iron ≥30 mg/day).Hemoglobin concentrations were detected at enrollment and before delivery for screening anemia.A 75-g,2-h oral glucose tolerance test was conducted during gestational week 24–28 for GDM diagnosis.Peripheral venous blood samples were collected for further measurements.Information on birth outcomes was obtained by medical records after delivery,including gestational weeks at delivery and birth weight.Plasma ferritin concentrations at both gestational week 14–18(Collection 1)and 24–28(Collection 2)were quantified using a commercial enzyme linked immunosorbent assay kit,and plasma iron concentrations were measured by inductively coupled plasma mass spectrometry.Allelic discrimination for TMPRSS6 gene(rs855791,rs4820268,rs2235321)was conducted by a genome-wide beadchip.Generalized linear mixed models were conducted to estimate the effects of iron supplement use on the longitudinal plasma ferritin and plasma iron levels.Plasma ferritin change was categorized into 6 groups based on ferritin levels at Collection1(<30,30-70,or ≥70 μg/L)and following decrease(using cut-offs of the medians of percentage decrease).The Log-Poisson regression models were used to evaluate the relative risk(RR)and 95% confidence interval(CI)for GDM,anemia in late pregnancy,preterm birth,small for gestational age and large for gestational age according to categories of plasma ferritin change.The jointly effect of TMPRSS6 single nucleotide polymorphisms and iron supplement use during pregnancy on plasma ferritin levels and the risk for GDM,anemia in late pregnancy,small for gestational age,large for gestational age was examined by generalized linear models and Log-Poisson regression models.Result:1.Among 1604 women enrolled in the present study,the mean age was 28.0 ± 3.1 years,and mean prepregnancy BMI was 20.8 ± 2.7 kg/m2.The proportion of nulliparous women was 85.5%.The medians and interquartile ranges(IQR)of plasma ferritin concentrations were 55.3(31.3,93.4)μg/L at Collection 1,and 22.4(14.5,35.7)μg/L at Collection 2.The medians(IQR)of plasma iron concentrations were 26.9(21.1,34.0)μmol/L at Collection 1,and 13.5(10.9,16.8)μmol/L at Collection 2.The proportions of low intake and high intake of iron supplement use in early pregnancy were 24.3% and 34.1%,respectively.Low intake and high intake in mid-pregnancy were 20.2% and 23.9%,respectively.2.Percentage decrease of plasma ferritin increased with higher initial ferritin status.Medians(IQR)of percentage decrease in ferritin were 27.5%(0.8%,49.2%)for women with low ferritin status at Collection 1(<30 μg/L),58.8%(39.6%,70.2%)for those with medium ferritin status(30-70 μg/L),and 74.6%(58.6%,82.7%)for those with high ferritin status(≥70 μg/L).The differences among groups with different ferritin status at Collection1 were statistically significant.Decrease of plasma iron was not significantly associated with ferritin levels at Collection 1.Nevertheless,among women with equivalent ferritin status initially,longitudinal percentage changes in plasma ferritin and plasma iron did not differ significantly among different groups of iron supplement use in mid-pregnancy.3.A total of 10.9% participants developed GDM.Participants were categorized into 6groups according to a combination of ferritin levels at Collection 1 and following ferritin decrease.Compared with the group of medium ferritin at Collection 1 combined with high ferritin decrease,the RRs(95% CIs)for GDM were 1.50(0.95,2.37)for medium ferritin at Collection 1 combined with low ferritin decrease,and 1.61(1.02,2.53)for high ferritin at Collection 1 combined with low ferritin decrease.No obvious increase in GDM risk was found in the group of high ferritin at Collection 1 combined with high ferritin decrease.The positive relationship persisted in women with iron supplement use,but disappeared in those without iron supplement use.4.Anemia in late pregnancy was significantly inversely associated with plasma ferritin at Collection 1,whereas not associated with ferritin decrease.Compared with the group of medium ferritin at Collection 1 combined with high ferritin decrease,the RRs(95% CIs)for anemia in late pregnancy were 0.61(0.36,1.05)for high ferritin at Collection 1combined with high ferritin decrease,and 0.61(0.38,0.98)for high ferritin at Collection 1combined with low ferritin decrease.Furthermore,the RR(95% CI)for small for gestational age was 1.71(1.03,2.83)for high ferritin at Collection 1 combined with low ferritin decrease.No significant association was found between ferritin change and the risk for large for gestational age and preterm birth.5.TMPRSS6 rs855791 G and rs2235321 A were related to higher plasma ferritin at Collection 1.No significant difference on longitudinal ferritin changes was found among different genotypes.Among those with rs855791 G,or rs4820268 A,or rs2235321 A,high intake of iron supplements(iron ≥30 mg/day)during pregnancy was significantly associated with elevated risk for GDM.Specifically,compared to no intake of iron supplements,the RRs(95% CIs)of high intake for GDM were 1.61(0.98,2.65)for rs855791 AG/GG,1.55(0.96,2.49)for rs4820268 AG/AA,and 1.69(1.02,2.80)for rs2235321 AG/AA.Among women with rs855791 AA or rs4820268 GG or rs2235321 GG,no significant association was found between iron supplement use and GDM risk.Conclusions:Magnitude of longitudinal plasma ferritin decrease from early to mid-pregnancy was mainly related to initial ferritin levels.Higher decrease of plasma ferritin in mid-pregnancy was mainly driven by higher initial ferritin levels.Iron supplement use did not prevent the natural decrease of plasma ferritin and plasma iron.For women with initial ferritin over 30μg/L,the insufficient decrease of ferritin was significantly associated with higher risk for GDM.Complying with naturally decrease for ferritin is important for mitigating the negative effect by high iron stores and maintaining optimal hemoglobin during pregnancy.Iron supplementation is not necessary for women with plasma ferritin over 70 μg/L during early pregnancy.TMPRSS6 polymorphisms might play a role in the effect on health by iron supplementation during pregnancy.