Advertisement

Diabetologia

, Volume 60, Issue 2, pp 249–257 | Cite as

A longitudinal study of iron status during pregnancy and the risk of gestational diabetes: findings from a prospective, multiracial cohort

  • Shristi Rawal
  • Stefanie N. Hinkle
  • Wei Bao
  • Yeyi Zhu
  • Jagteshwar Grewal
  • Paul S. Albert
  • Natalie L. Weir
  • Michael Y. Tsai
  • Cuilin Zhang
Article

Abstract

Aims/hypothesis

The aim of this study was to prospectively and longitudinally investigate maternal iron status during early to mid-pregnancy, and subsequent risk of gestational diabetes mellitus (GDM), using a comprehensive panel of conventional and novel iron biomarkers.

Methods

A case–control study of 107 women with GDM and 214 controls (matched on age, race/ethnicity and gestational week during blood collection) was conducted within the the Eunice Kennedy Shriver National Institute of Child Health and Human Development Fetal Growth Studies–Singleton Cohort (2009–2013), a prospective and multiracial pregnancy cohort. Plasma hepcidin, ferritin and soluble transferrin receptor (sTfR) were measured and sTfR:ferritin ratio was derived, twice before GDM diagnosis (gestational weeks 10–14 and 15–26) and at weeks 23–31 and 33–39. GDM diagnosis was ascertained from medical records. Adjusted ORs (aORs) for GDM were estimated using conditional logistic regression analysis, adjusting for demographics, prepregnancy BMI and other major risk factors.

Results

Hepcidin concentrations during weeks 15–26 were 16% higher among women with GDM vs controls (median 6.4 vs 5.5 ng/ml; p = 0.02 ), and were positively associated with GDM risk; the aOR (95% CI) for highest vs lowest quartile was 2.61 (1.07, 6.36). Ferritin levels were also positively associated with GDM risk; the aOR (95% CI) for highest vs lowest quartile was 2.43 (1.12, 5.28) at weeks 10–14 and 3.95 (1.38, 11.30) at weeks 15–26. The sTfR:ferritin ratio was inversely related to GDM risk; the aOR (95% CI) for highest vs lowest quartile was 0.33 (0.14, 0.80) at weeks 10–14 and 0.15 (0.05, 0.48) at weeks 15–26.

Conclusions/interpretation

Our findings suggest that elevated iron stores may be involved in the development of GDM from as early as the first trimester. This raises potential concerns for the recommendation of routine iron supplementation among iron-replete pregnant women.

Keywords

Ferritin Gestational diabetes mellitus Hepcidin Iron overload Soluble transferrin receptor sTfR:ferritin ratio 

Abbreviations

aOR

Adjusted OR

CRP

C-reactive protein

GDM

Gestational diabetes mellitus

LMP

Last menstrual period

NICHD

Eunice Kennedy Shriver National Institute of Child Health and Human Development

sTfR

Soluble transferrin receptor

Introduction

Iron is regarded as a double-edged sword in living systems, as both iron deficiency and overload can be harmful. Pregnant women are particularly vulnerable to iron deficiency and related adverse pregnancy outcomes [1]. While a few guidelines, including those from the American Congress of Obstetricians and Gynecologists, recommend screening and treatment for iron deficiency as required, several other groups, such as WHO and the Centers for Disease Control and Prevention, recommend routine iron supplementation in pregnant women [2]. Emerging findings from both animal and human studies, however, have raised critical concerns about significant links between larger iron stores and disturbances in glucose metabolism, including an increased risk of type 2 diabetes among non-pregnant individuals [3, 4, 5]. The evidence is unclear regarding the contribution of larger iron stores in pregnancy to the development of gestational diabetes mellitus (GDM), a common complication of pregnancy that is associated with a number of adverse health outcomes for both women and their children [6]. Advancing gestation is characterised by progressive increases in demands for iron [6] and insulin resistance [7], the latter being a central feature of GDM. Longitudinal studies that examine iron status across pregnancy are, hence, needed to comprehensively elucidate the role of iron status in the pathogenesis of GDM.

Prospective studies investigating the association between iron status in pregnancy and the risk of GDM are few and inconsistent in their findings [8, 9, 10, 11, 12], and none have examined these associations with longitudinal measurements of iron status. In addition, the vast majority of the existing literature centres on serum ferritin as a biomarker of iron stores. Ferritin levels, however, may not be the optimal indicator of iron status, as they may also increase in the presence of infection or inflammation [13]. Alternatively, soluble transferrin receptor (sTfR), a marker of tissue iron insufficiency, could provide useful and complementary information on iron status as it is considered to be less influenced by the acute-phase response than ferritin [13, 14, 15]. Furthermore, emerging evidence has identified hepcidin, a novel biomarker, as the master regulator of iron homeostasis [16]. In response to high body iron stores, increased hepatic secretion of hepcidin inhibits iron absorption from three main sources: dietary absorption in the gut, recycled iron from macrophages and stored iron released from hepatocytes [16]. In pregnancy, hepcidin may also play a critical role in the active transport of iron to the fetus [17]. To our knowledge, only one prior study has examined serum hepcidin levels in relation to GDM [18]. The study found a positive association, yet provided limited inference because of the small number of GDM cases (n = 30), the use of cross-sectional analysis with blood samples collected at the same time as GDM diagnosis and the lack of adjustment for potential confounders.

In the present study, we aimed to prospectively and longitudinally investigate body iron status during early to mid-pregnancy and its association with subsequent GDM risk using a comprehensive panel of conventional and novel iron biomarkers, including hepcidin, ferritin, sTfR and sTfR:ferritin ratio. Additionally, concurrent measurements of plasma C-reactive protein (CRP) levels were obtained to account for inflammation. The secondary objective of the study was to characterise longitudinal changes in these iron metabolism biomarkers throughout pregnancy.

Methods

Study design

We conducted a case–control study within the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) Fetal Growth Studies–Singleton Cohort (2009–2013), which was a prospective and multiracial pregnancy cohort. Briefly, 2334 non-obese [19] and 468 obese women, aged 18–40 years, were recruited between 8 and 13 weeks of gestation, from 12 US clinical centres, and followed throughout pregnancy [19]. Eligible women had a known date of last menstrual period (LMP), which was confirmed by ultrasound screening at enrolment [19]. Women whose LMP and ultrasound dates did not match within a specific range (±5–7 days) were excluded from the study. Women with a history of pre-existing hypertension, diabetes or other major chronic diseases were also excluded [19]. Research approval was obtained from all participating institutions and the participants provided written informed consent.

In the NICHD Fetal Growth Studies singleton cohort, 107 women with GDM (cases) were identified by medical record review. The GDM diagnosis was based on Carpenter and Coustan’s diagnostic criteria [20]. Each case was matched with two randomly selected non-GDM controls, based on age (±2 years), race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic or Asian/Pacific Islander) and gestational week of blood collection (±2 weeks). Thus, a total of 321 women (107 GDM cases and 214 non-GDM controls) from the original cohort were included in this case–control study. Following a standardised protocol, blood specimens were collected at four study visits during pregnancy. The visits were targeted at gestational weeks 8–13 (enrolment), 16–22, 24–29 and 34–37, but the actual ranges were gestational weeks 10–14, 15–26, 23–31 and 33–39, respectively. Blood specimens were collected at the second visit (week 15–26), after an overnight fast. All biospecimens were immediately processed and stored at –80°C until being thawed for laboratory analysis.

Exposure assessment

Concentrations of plasma hepcidin (ng/ml) were measured using a competitive binding ELISA kit (Human Hepcidin-25 [bioactive] ELISA kit; DRG Diagnostics, Marburg, Germany). Concentrations of plasma ferritin (pmol/l) and sTfR (nmol/l) were measured using Roche reagents on the Roche Modular P chemistry analyser (Roche Diagnostics, Indianapolis, IN, USA). sTfR:ferritin ratio was derived by dividing the plasma concentration of sTfR (μg/l) by that of ferritin (μg/l). Plasma CRP (nmol/l) was measured using a highly sensitive latex-particle-enhanced immunoturbidimetric assay kit on the Roche Modular P chemistry analyser. All assays were performed blinded to case or control status, and the inter- and intra-assay coefficients of variation were ≤13%. For women with GDM and one of their matched controls, the assays were performed using samples from all four time points of blood collection. For the remaining control participants, biomarkers were only measured in the blood specimens collected at the two visits prior to GDM diagnosis (i.e. gestational weeks 10–14 and 15–26).

Assessment of covariates

A structured questionnaire was administered at enrolment and collected information on several GDM risk factors, including maternal age (years), race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic or Asian/Pacific Islander), nulliparity (yes/no), family history of diabetes (yes/no), education (less than, equal to or more than high-school), smoking in the 6 months preceding pregnancy (yes/no) and alcohol consumption in the 3 months preceding pregnancy (yes/no). Prepregnancy BMI (<25.0, 25.0–29.9 or ≥30.0 kg/m2) was calculated from prepregnancy weight (self-reported) and height (measured at enrolment). Information on GDM treatment (diet, lifestyle modification and/or medication) was obtained by reviewing medical records. Gestational age (weeks) at each blood collection was calculated from the reported LMP date. At each study visit, women also reported medication and supplement use, including iron supplements.

Statistical methods

Descriptive statistics are presented as mean (SD) for parametric continuous variables, median (25th, 75th percentile) for non-parametric continuous variables and frequencies for categorical variables. Differences between cases and controls were analysed using generalised linear mixed-effects models for continuous variables and binomial/multinomial logistic regression with generalised estimating equations for categorical variables, accounting for matched case–control pairs. To examine the longitudinal trajectory of each iron biomarker over pregnancy, median concentrations of the biomarkers were plotted by study visit.

For each iron marker, crude and adjusted quartile-specific ORs of GDM were estimated at weeks 10–14 and 15–26 (visits prior to the typical screening for GDM) using conditional logistic regression models with the lowest quartile as the reference group. Multivariable models were adjusted for a priori selected covariates including parity, education, family history of diabetes and prepregnancy BMI. Since maternal age and gestational age at blood collection were only matched between cases and controls within a specified range, both were included as covariates to derive conservative risk estimates. In addition, for each time-point, we adjusted for CRP levels at the corresponding visit to limit the potential influence of inflammation status. Tests of linear trend were conducted using the median value for each quartile and fitting it as a continuous variable in the conditional logistic regression models. We excluded one case at weeks 10–14 and five cases at weeks 15–26 from the final analyses, as these women were diagnosed with GDM before the respective visit.

In order to test the robustness of our findings, we conducted sensitivity analyses by excluding women with a history of GDM (n = 6), a history of chronic anaemia and other haematological disorders (n = 2), smoking before the current pregnancy (n = 5), self-reported anaemia during the current pregnancy (n = 16) or suspected iron deficiency (defined as plasma ferritin <26.9 pmol/l [9]) at weeks 10–14 (n = 3) or 15–26 (n = 13). To evaluate whether the findings were modified by major risk factors for GDM, we also conducted stratified analyses by prepregnancy obesity status (BMI <30.0 vs ≥30.0 kg/m2), median CRP level (<48.3 vs ≥48.3 nmol/l at weeks 10–14 and <49.5 vs ≥49.5 nmol/l l at weeks 15–26), family history of diabetes (yes vs no) and GDM severity as indicated by treatment (diet/lifestyle modification vs medication use). Statistical analyses were conducted using SAS version 9.4 (SAS Institute, Cary, NC, USA). Significance was set at two-tailed p values of <0.05.

Results

Compared with non-GDM controls, women who subsequently developed GDM were more likely to have a higher prepregnancy BMI and a family history of diabetes (Table 1). Figure 1 shows the longitudinal trajectory of each iron biomarker among women with GDM and their matched controls. Hepcidin and ferritin concentrations tended to decline through mid-pregnancy and then level off. In comparison, sTfR concentrations and the sTfR:ferritin ratio, both indicators of iron insufficiency, tended to increase with the progression of pregnancy.
Table 1

Characteristics of women with GDM and their matched controls in the NICHD Fetal Growth Studies, singleton cohort (2009–2013)

Characteristics

GDM cases

Non-GDM controls

p valueb

(n = 107)

(n = 214)

Age (years)

30.5 ± 5.7

30.4 ± 5.4

 

Race/ethnicity

  Non-Hispanic white

25 (23.4)

50 (23.4)

 

  Non-Hispanic black

15 (14.0)

30 (14.0)

 

  Hispanic

41 (38.3)

82 (38.3)

 

  Asian/Pacific Islander

26 (24.3)

52 (24.3)

 

Education

  

0.18

  Less than high-school

17 (15.9)

26 (12.1)

 

  High-school graduate or equivalent

15 (14.0)

23 (10.7)

 

  More than high-school

75 (70.1)

165 (77.1)

 

Married/living with a partner

92 (86.0)

167 (78.0)

0.12

Nulliparous

48 (44.9)

96 (44.9)

1

Family history of diabetes

40 (37.4)

48 (22.4)

0.003

Anaemia during pregnancy and before GDM diagnosis

6 (5.6)

11 (5.1)

0.85

Smoked before pregnancya

4 (3.7)

1 (0.5)

0.06

Consumed alcohol before pregnancy

61 (57.0)

137 (64.0)

0.22

Prepregnancy BMI (kg/m2)

  

<0.001

  18.38–24.99

37 (34.6)

123 (57.5)

 

  25.0–29.99

35 (32.7)

56 (26.2)

 

  30.0–45.11

35 (32.7)

33 (15.4)

 

Data are presented as n (%) for categorical variables and mean ± SD for continuous variables

aNon-obese women who smoked were not eligible for the study

b p values for differences between cases and controls were obtained by generalised linear mixed-effects models for continuous variables and binomial/multinomial logistic regression with generalised estimating equations for binary/multilevel categorical variables, accounting for matched case–control pairs. p values are not shown for matching variables (age, race/ethnicity)

Fig. 1

Median plasma concentrations of hepcidin (a), ferritin (b), sTfR (c) and sTfR:ferritin ratio (d) at each study visit among women with GDM and their matched controls. Solid line, non-GDM controls; dashed line, GDM cases. Visit 1 (weeks 10–14): 104 cases and 214 controls; visit 2 (weeks 15–26): 94 cases and 212 controls; visit 3 (weeks 23–31): 102 cases and 106 controls; visit 4 (weeks 33–39): 88 cases and 101 controls. *p < 0.05, **p < 0.01 for case–control comparisons at each study visit, obtained by generalised linear mixed-effects models accounting for matched case–control pairs

At the first visit (weeks 10–14), only ferritin levels differed significantly between GDM cases and controls. Ferritin levels were approximately 21% higher among women who subsequently developed GDM than those who did not (median levels 143.2 vs 118.7 pmol/l, p = 0.04; Table 2). At weeks 15–26, closer to the time of GDM diagnosis, both hepcidin and ferritin levels were significantly higher among cases than controls. In addition, the sTfR:ferritin ratio was lower among cases at both visits, but the difference was statistically significant only at the second visit. sTfR concentrations did not differ significantly between GDM cases and controls at either visit.
Table 2

Median plasma concentrations of iron biomarkers among women with GDM and their matched controls in the NICHD Fetal Growth Studies, singleton cohort (2009–2013)

Variable

GDM cases

Non-GDM controls

p valued

Gestational weeks 10–14 (first visit)a

  Hepcidin (ng/ml)

12.2 (8.3–21.4)

11.8 (7.0–18.7)

0.20

  Ferritin (pmol//l)

143.2 (92.0–206.2)

118.7 (76.8–173.9)

0.04

  sTfR (nmol/l)

28.2 (22.3–34.1)

27.1 (22.3–34.1)

0.63

  sTfR:ferritin ratiob

37.3 (26.0–61.9)

40.9 (27.6–77.8)

0.14

  Gestational age at blood collection (weeks)

13.0 (12.3–13.6)

13.0 (12.4–13.6)

 

Gestational weeks 15–26 (second visit)c

  Hepcidin (ng/ml)

6.4 (4.6–8.3)

5.5 (4.0–7.9)

0.02

  Ferritin (pmol/l)

94.5 (67.9–133.5)

78.1 (49.7–119.4)

0.007

  sTfR (nmol/l)

29.4 (24.7–35.3)

28.2 (24.7–35.3)

0.85

  sTfR:ferritin ratiob

60.0 (36.9–85.4)

69.4 (40.6–130.9)

0.02

  Gestational age at blood collection (weeks)

18.9 (17.3–21.0)

19.0 (17.7–21.1)

 

Data are presented as median (25th–75th percentile)

a Cases, n = 104; controls, n = 214

bObtained by dividing the plasma concentration of sTfR (μg/l) by the plasma ferritin level (μg/l)

cCases, n = 94; controls n = 212

d p values for differences between cases and controls were obtained by generalised linear mixed-effects models for continuous variables accounting for matched case–control pairs. p values are not shown for gestational age at blood collection as this was one of the matching variables

Associations between quartiles of iron biomarkers and GDM status are shown in Table 3. During gestational weeks 10–14, ferritin levels were significantly and positively associated with GDM risk, with an adjusted OR (aOR) (95% CI) of 2.43 (1.12, 5.28) when comparing the highest vs lowest quartile.The odds of GDM tended to increase with increasing quartiles of ferritin at weeks 10–14 (p for trend = 0.03). In comparison, the sTfR:ferritin ratio was significantly and inversely associated with GDM risk at weeks 10–14 (p for trend = 0.02), with an aOR (95% CI) of 0.33 (0.14, 0.80) when comparing the highest vs the lowest quartile (Table 3). At 15–26 weeks, plasma concentrations of hepcidin and ferritin were significantly and positively associated with GDM risk, whereas the sTfR:ferritin ratio was inversely related to GDM risk (Table 3). Specifically, comparing the highest with the lowest quartile, the aOR (95% CI) associated with GDM risk was 2.61 (1.07, 6.36) for hepcidin, 3.95 (1.38, 11.30) for ferritin, and 0.15 (0.05, 0.48) for sTfR:ferritin ratio. The test for linear trend was only significant for sTfR:ferritin ratio (p for trend = 0.002). The associations between iron biomarkers and GDM risk remained practically unchanged between multivariable models with and without additional adjustment for CRP levels (Table 3; data not shown for the latter).
Table 3

Adjusted ORs (95% CI) for GDM according to quartiles of hepcidin, ferritin, sTfR and sTfR:ferritin ratio at gestational weeks 10–14 and 15–26 in the NICHD Fetal Growth Studies, singleton cohort (2009–2013)

Variable

Cases, n

Controls, n

Crude model

Multivariable modela

Gestational weeks 10–14b

  Hepcidin (ng/ml)

    Quartile 1: ≤6.97

21

53

1

1

    Quartile 2: 6.98–11.80

29

52

1.38 (0.73, 2.61)

1.12 (0.54, 2.32)

    Quartile 3: 11.81–18.70

25

53

1.19 (0.57, 2.46)

0.87 (0.38, 1.99)

    Quartile 4: ≥18.71

28

52

1.32 (0.66, 2.67)

1.11 (0.51, 2.44)

  Ferritin (pmol/l)*

    Quartile 1: ≤77.05

18

54

1

1

    Quartile 2: 77.06–118.70

23

53

1.36 (0.65, 2.84)

1.57 (0.69, 3.56)

    Quartile 3: 118.80– 173.90

26

54

1.57 (0.75, 3.27)

1.71 (0.75, 3.90)

    Quartile 4: ≥174.00

37

53

2.11 (1.06, 4.20)

2.43 (1.12, 5.28)

  sTfR (nmol/l)

    Quartile 1: ≤22.3

24

54

1

1

    Quartile 2: 22.4– 27.1

23

53

0.94 (0.47, 1.90)

0.99 (0.45, 2.17)

    Quartile 3: 27.2–33.8

29

54

1.21 (0.63, 2.35)

0.89 (0.41, 1.91)

    Quartile 4: ≥33.9

28

53

1.23 (0.62, 2.46)

1.00 (0.45, 2.20)

  sTfR:ferritin ratioc,*

    Quartile 1: ≤27.65

32

54

1

1

    Quartile 2: 27.66–40.90

26

53

0.83 (0.44, 1.57)

0.67 (0.32, 1.42)

    Quartile 3: 40.91–77.83

31

54

0.97 (0.51, 1.83)

0.88 (0.43, 1.81)

    Quartile 4: ≥77.84

15

53

0.47 (0.22, 1.00)

0.33 (0.14, 0.80)

Gestational weeks 15–26b

  Hepcidin (ng/ml)

    Quartile 1: ≤4.00

14

53

1

1

    Quartile 2: 4.01–5.46

19

53

1.63 (0.71, 3.77)

1.49 (0.57, 3.93)

    Quartile 3: 5.47–7.91

30

53

2.27 (1.01, 5.11)

2.97 (1.16, 7.60)

    Quartile 4: ≥7.92

31

52

2.40 (1.08, 5.35)

2.61 (1.07, 6.36)

  Ferritin (pmol/l)

    Quartile 1: ≤49.6

10

54

1

1

    Quartile 2: 49.7–77.9

28

52

2.76 (1.20, 6.33)

4.45 (1.61, 12.3)

    Quartile 3: 78.0–119.3

26

53

2.59 (1.11, 6.03)

3.42 (1.26, 9.30)

    Quartile 4: ≥119.4

30

53

3.06 (1.27, 7.34)

3.95 (1.38, 11.30)

  sTfR (nmol/l)

    Quartile 1: ≤24.2

23

53

1

1

    Quartile 2: 24.3–28.7

21

54

0.87 (0.43, 1.76)

0.89 (0.39, 2.04)

    Quartile 3: 28.8–34.9

24

53

1.08 (0.51, 2.29)

1.31 (0.55, 3.09)

    Quartile 4: ≥35.0

26

52

1.23 (0.56, 2.67)

1.17 (0.49, 2.82)

  sTfR:ferritin ratioc,**

    Quartile 1: ≤40.40

31

53

1

1

    Quartile 2: 40.41–69.36

23

53

0.66 (0.31, 1.37)

0.58 (0.25, 1.33)

    Quartile 3: 69.37–130.87

32

53

1.07 (0.51, 2.23)

1.14 (0.49, 2.65)

    Quartile 4: ≥130.88

8

53

0.25 (0.10, 0.64)

0.15 (0.05, 0.48)

aAdjusted for maternal age (years), gestational age at blood collection (weeks), nulliparity (yes/no), education (less than, equal to or more than high-school), family history of diabetes (yes/no), prepregnancy BMI (<25.0, 25.0–29.9 or ≥30.0 kg/m2) and plasma CRP levels levels (nmol/l)

bTiming of blood sample collection preceded the diagnosis of GDM in all participants

cObtained by dividing the plasma concentration of sTfR (μg/l) by the plasma ferritin level (μg/l)

*p for trend <0.05; **p for trend <0.01

The majority of women in our study (83% at weeks 10–14; 87% in the second trimester during 15–26 gestational weeks) reported using supplemental iron, including both iron-only supplements and prenatal or standard multivitamins that contained iron (data not shown). In sensitivity analyses (data not shown), the associations between iron biomarkers and GDM risk were attenuated but did not appreciably change after additional adjustment for supplemental use of iron (yes/no) at the corresponding visit. The associations also persisted in sensitivity analyses after the exclusion of women with a history of prior GDM, self-reported anaemia during pregnancy, smoking before pregnancy, suspected iron deficiency (plasma ferritin <26.9 pmol/l [9]) or a history of chronic anaemia and other haematologic disorders. In addition, the directions of the associations remained similar when stratifying the analyses by prepregnancy obesity status, CRP level, family history of diabetes or severity of GDM (data not shown).

Discussion

In this prospective, longitudinal study of pregnant women without prepregnancy chronic disease, higher iron status in pregnancy (as indicated by higher hepcidin and ferritin concentrations and a lower sTfR:ferritin ratio) was significantly associated with an elevated risk of GDM, even after adjustment for plasma CRP levels, prepregnancy BMI and other major GDM risk factors. In addition, the magnitude of associations appeared greater in the second trimester, closer to GDM diagnosis, than in early pregnancy.

Among iron biomarkers, ferritin has been most often investigated in relation to GDM. Our finding of a significant and positive association between ferritin levels and GDM is consistent with most [9, 10, 11, 12, 18, 21, 22, 23, 24, 25, 26] but not all previous studies [8, 27, 28]. Notably, the majority of prior studies were cross-sectional, using only a single measurement of ferritin, typically assessed at the time of GDM diagnosis [18, 21, 22, 23, 24, 25, 26, 27]. Peripheral circulating ferritin concentrations are a good proxy for body iron stores but, being an acute-phase reactant, ferritin levels can also increase with subclinical systemic inflammation, which is associated with insulin resistance in GDM [13]. Hence, prospective studies that measure iron status well before GDM diagnosis are needed to establish temporality and to control for confounding by insulin resistance-mediated inflammation. Studies that have prospectively examined the association of ferritin with GDM risk have, however, been limited and inconsistent in their findings [8, 9, 10, 11, 12]. For instance, a study based in Lebanon found that high ferritin levels in early pregnancy were significantly associated with impaired glucose tolerance but not with GDM incidence, although the latter could be attributed to the small number of women with GDM in the study (n = 16) [8]. In another prospective study, the association between serum ferritin and GDM was found to be significant after accounting for several confounders, including ethnicity and a family history of diabetes, but was attenuated to non-significance after additionally accounting for prepregnancy BMI [11]. On the other hand, a recent case–control study based on data from the Danish National Birth Cohort (1996–2002) reported a positive and significant association between ferritin and GDM risk, even after adjustment for plasma CRP levels and several risk factors for GDM, including prepregnancy BMI [12]. Similarly to the Danish study [12], we minimised the influence of inflammatory status by controlling for CRP levels in our multivariable models and found that high ferritin was significantly associated with GDM risk at as early as 10–14 weeks of gestation. Our study offers several unique strengths compared with the Danish study [12], including a longitudinal design with two iron measurements prior to GDM diagnosis, a more contemporary and multiracial cohort, and the evaluation of hepcidin, a relatively novel biomarker that has recently been identified as the master regulator of iron homeostasis. Only one prior study has been carried out to examine hepicidin levels in relation to GDM [18]; in this study, positive associations were observed between hepcidin and GDM, similar to our study. However, inferences that can be drawn from the findings of this prior study are hindered by its cross-sectional examination of the data [18]. More importantly, the study did not control for potential confounding factors or account for the impact of inflammation status.

sTfR, a marker of tissue iron insufficiency, has also been understudied in the context of GDM risk [9, 12, 29]. In the present study, no significant association was observed between sTfR levels and GDM risk. Similarly, in an Australian cohort of pregnant women, first trimester sTfR levels were not found to be significantly associated with the risk of GDM [9, 29]. In a recent study conducted among Danish women, sTfR concentrations in early pregnancy (9.4 ± 3.2 weeks of gestation) were significantly and positively associated with GDM risk, but this was attenuated to non-significance after adjustment for prepregnancy BMI [12]. It is worth noting, however, that sTfR levels tend to increase only in the presence of a functional iron deficiency. In an experimental study in which iron-deficiency anaemia was gradually induced in healthy participants, serum ferritin levels declined as iron stores decreased, whereas circulating sTfR levels increased only once the iron stores had been significantly depleted. In our study sample, there were only a small number of women with suspected iron deficiency (three women at weeks 10–14 and 13 women at weeks 15–26), as defined by plasma ferritin levels less than 26.9 pmol/l [9]. Hence, sTfR levels alone might not have been an appropriate indicator of iron status in our sample.

Our study is unique in that we further examined the sTfR:ferritin ratio, a measure that captures both cellular iron demand and the availability of body iron stores [14]. The added value of using sTfR:ferritin ratio, as opposed to isolated measurements of sTfR and ferritin, is that it covers the full spectrum of iron homeostasis, from normal, healthy iron stores to mild or substantial functional iron deficiency [14, 15]. The sTfR:ferritin ratio offers greater sensitivity and specificity in characterising iron status than individual measurements of ferritin and sTfR and is considered to be particularly useful when individual measurements yield ambiguous results [13]. In the present study, we observed significant inverse associations between sTfR:ferritin ratio and GDM, despite having observed no association with sTfR levels alone. This suggests that the inverse association with sTfR:ferritin ratio may be driven primarily via levels of ferritin, hence this finding should be interpreted cautiously.

To our knowledge, this is the first study to prospectively demonstrate significant and positive associations between iron status during pregnancy and GDM risk using longitudinal measurements of a panel of conventional and novel iron biomarkers. Higher hepcidin levels, which indicate an iron surplus, were significantly related to an increased GDM risk only during the second trimester, before the diagnosis of GDM. Similarly, an elevated second-trimester iron status, whether based on higher ferritin levels or lower sTfR:ferritin ratio, was more strongly associated with GDM risk than an elevated first-trimester iron status.

In secondary analyses, we examined the longitudinal trajectory of several iron biomarkers over the entire course of pregnancy. Hepcidin and ferritin concentrations tended to decline with advancing gestation, whereas sTfR concentrations and the sTfR:ferritin ratio, both indicators of iron insufficiency, tended to increase with the progression of pregnancy. This was expected, as iron demands increase dramatically throughout pregnancy in order to support placental and fetal growth, sustain the expansion of erythrocyte mass and offset blood losses from delivery [30]. A similar decline in ferritin and hepcidin levels, and an increase in sTfR concentrations through the course of pregnancy have been previously reported by others, yet our study is the largest to date to provide data for changes in hepcidin levels during pregnancy [31, 32].

Our findings are biologically plausible; iron may play a role in the pathogenesis of GDM via several potential mechanisms. As a strong pro-oxidant, free iron can catalyse several cellular reactions that generate reactive oxygen species and increase the level of oxidative stress [3]. Oxidative stress induced from excess iron accumulation can cause beta cell damage and apoptosis and, consequently, contribute to impaired insulin synthesis and secretion [3]. In the liver, high iron stores may induce insulin resistance via impaired insulin signalling, as well as by attenuating the capacity of the liver to extract insulin [33, 34]. Excess iron deposition in the muscles may enhance NEFA oxidation and interfere with glucose uptake or disposal [33, 34]. Iron accumulation may also impair the action of insulin and interfere with insulin-induced glucose transport in adipocytes [33, 34]. Consistent with this notion, an increased prevalence of type 2 diabetes has been observed in patients with hereditary haemochromatosis, a genetic disorder characterised by excessive iron absorption and pathological increases in total body iron stores [33]. Similarly, increasing evidence from biological and epidemiological studies indicates that elevated body iron stores are associated with an increased risk of types 1 and 2 diabetes [3, 4, 5]. Furthermore, induction of iron depletion has been shown to improve insulin sensitivity measures in carbohydrate-intolerant participants [35, 36].

Our study has several strengths. First, longitudinal blood collection provided a unique opportunity to prospectively investigate levels of iron biomarkers across pregnancy and their trimester-specific associations with the subsequent risk of GDM. Second, the study participants represented various races/ethnicities and included a relatively high number of women with GDM who were well characterised based on medical records. More importantly, our study examined conventional and novel biomarkers of iron status, and measured them prospectively at multiple time points before GDM diagnosis. Finally, we considered the influence of a number of important confounding factors, including inflammatory status, in evaluating the relationship between iron status in pregnancy and GDM risk. It is important to note that the present study focused on examining the role of body iron status, as measured by plasma iron biomarkers, in the development of GDM; future studies investigating how dietary iron intake influences these biomarkers and ultimately impacts GDM risk are warranted.

In summary, the findings from this longitudinal and prospective study among multiracial, relatively healthy pregnant women without any prepregnancy chronic disease suggest that higher maternal iron stores may play a role in the development of GDM from as early as the first trimester. These findings are of clinical and public health importance since they extend the observation of an association between high body iron stores and an elevated risk of glucose intolerance among non-pregnant individuals to those who are pregnant, and raise potential concerns for the recommendation of routine iron supplementation among iron-replete pregnant women.

Notes

Acknowledgements

We would like to acknowledge the research teams at our participating clinical centres, including Christiana Care Health Systems, Wilmington, DE, USA; University of California, Irvine, CA, USA; Long Beach Memorial Medical Center, CA, USA; Northwestern University, Evanston, IL, USA; Medical University of South Carolina, Charleston, SC, USA; Columbia University, New York, NY, USA; New York Hospital Queens, NY, USA; St Peters’ University Hospital, New Brunswick, NJ, USA; University of Alabama at Birmingham, AL, USA; Women and Infants Hospital of Rhode Island, Providence, RI, USA; Fountain Valley Regional Hospital and Medical Center, CA, USA; and Tufts University, Medford, MA, USA. We would also like to thank the C-TASC Corporation, Owings Mill, MD, USA for providing data coordination, as well as the Department of Laboratory Medicine and Pathology, University of Minnesota, Minneapolis, MN, USA, for providing the laboratory support and resources for the analysis of blood samples and biomarkers. Some of the data from this study were presented as a poster at the 76th Scientific Sessions of the ADA, New Orleans, LA, USA (2016).

Funding

This research was supported by NICHD intramural funding and included American Recovery and Reinvestment Act funding via contract numbers: HHSN275200800013C, HHSN275200800002I, HHSN27500006, HHSN275200800003IC, HHSN275200800014C, HHSN275200800012C, HHSN275200800028C, HHSN275201000009C and HHSN275201000001Z.

Data availability

The datasets generated during and/or analysed during the current study are available from Eunice Kennedy Shriver National Institute of Child Health and Human Development, but restrictions apply to the availability of these data, and hence they are not publicly available yet.

Duality of interest

The authors declare that there is no duality of interest associated with this manuscript.

Contribution statement

SR conducted the data analysis and wrote the first draft of the manuscript. SNH contributed to data interpretation and revised the manuscript. WB contributed to study coordination, data interpretation and manuscript reviewing. YZ contributed to data management, data analysis and interpretation, and reviewed the manuscript. JG contributed to data interpretation and reviewed the manuscript. PSA contributed to data analysis and interpretation, and reviewed the manuscript. NLW and MYT assisted with laboratory testing and reviewed the manuscript. CZ obtained funding, designed and oversaw the study, and revised the manuscript. All authors contributed to the critical interpretation of the results, reviewed the manuscript for important intellectual content and approved the final version of the manuscript, and have agreed to be accountable for their role in this manuscript. SR and CZ are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

References

  1. 1.
    Scholl TO, Reilly T (2000) Anemia, iron and pregnancy outcome. J Nutr 130:443S–447SPubMedGoogle Scholar
  2. 2.
    McDonagh M, Cantor A, Bougatsos C, Dana T, Blazina I (2015) Routine iron supplementation and screening for iron deficiency anemia in pregnant women: a systematic review to update the U.S. Preventive Services Task Force recommendation. Agency for Healthcare Research and Quality, Rockville, MD, USA. Available from www.ncbi.nlm.nih.gov/books/NBK285986, accessed 6 September 2016
  3. 3.
    Hansen JB, Moen IW, Mandrup-Poulsen T (2014) Iron: the hard player in diabetes pathophysiology. Acta Physiol (Oxf) 210:717–732CrossRefGoogle Scholar
  4. 4.
    Montonen J, Boeing H, Steffen A et al (2012) Body iron stores and risk of type 2 diabetes: results from the European Prospective Investigation into Cancer and Nutrition (EPIC)-Potsdam study. Diabetologia 55:2613–2621CrossRefPubMedPubMedCentralGoogle Scholar
  5. 5.
    Zhao Z, Li S, Liu G et al (2012) Body iron stores and heme-iron intake in relation to risk of type 2 diabetes: a systematic review and meta-analysis. PLoS One 7:e41641CrossRefPubMedPubMedCentralGoogle Scholar
  6. 6.
    Zein S, Rachidi S, Hininger-Favier I (2014) Is oxidative stress induced by iron status associated with gestational diabetes mellitus? J Trace Elem Med Biol 28:65–69CrossRefPubMedGoogle Scholar
  7. 7.
    Barbour LA, McCurdy CE, Hernandez TL, Kirwan JP, Catalano PM, Friedman JE (2007) Cellular mechanisms for insulin resistance in normal pregnancy and gestational diabetes. Diabetes Care 30(Suppl 2):S112–S119CrossRefPubMedGoogle Scholar
  8. 8.
    Zein S, Rachidi S, Awada S et al (2015) High iron level in early pregnancy increased glucose intolerance. J Trace Elem Med Biol 30:220–225CrossRefPubMedGoogle Scholar
  9. 9.
    Khambalia AZ, Collins CE, Roberts CL et al (2016) Iron deficiency in early pregnancy using serum ferritin and soluble transferrin receptor concentrations are associated with pregnancy and birth outcomes. Eur J Clin Nutr 70:358–363CrossRefPubMedGoogle Scholar
  10. 10.
    Tarim E, Kilicdag E, Bagis T, Ergin T (2004) High maternal hemoglobin and ferritin values as risk factors for gestational diabetes. Int J Gynaecol Obstet 84:259–261CrossRefPubMedGoogle Scholar
  11. 11.
    Chen X, Scholl TO, Stein TP (2006) Association of elevated serum ferritin levels and the risk of gestational diabetes mellitus in pregnant women: the Camden study. Diabetes Care 29:1077–1082CrossRefPubMedGoogle Scholar
  12. 12.
    Bowers KA, Olsen SF, Bao W, Halldorsson TI, Strom M, Zhang C (2016) Plasma concentrations of ferritin in early pregnancy are associated with risk of gestational diabetes mellitus in women in the Danish National Birth Cohort. J Nutr 146:1756–1761CrossRefPubMedGoogle Scholar
  13. 13.
    Harms K, Kaiser T (2015) Beyond soluble transferrin receptor: old challenges and new horizons. Best Pract Res Clin Endocrinol Metab 29:799–810CrossRefPubMedGoogle Scholar
  14. 14.
    Skikne BS, Flowers CH, Cook JD (1990) Serum transferrin receptor: a quantitative measure of tissue iron deficiency. Blood 75:1870–1876PubMedGoogle Scholar
  15. 15.
    Skikne BS (1998) Circulating transferrin receptor assay—coming of age. Clin Chem 44:7–9PubMedGoogle Scholar
  16. 16.
    Ganz T, Nemeth E (2012) Hepcidin and iron homeostasis. Biochim Biophys Acta 1823:1434–1443CrossRefPubMedPubMedCentralGoogle Scholar
  17. 17.
    Young MF, Griffin I, Pressman E et al (2012) Maternal hepcidin is associated with placental transfer of iron derived from dietary heme and nonheme sources. J Nutr 142:33–39CrossRefPubMedGoogle Scholar
  18. 18.
    Derbent AU, Simavli SA, Kaygusuz I et al (2013) Serum hepcidin is associated with parameters of glucose metabolism in women with gestational diabetes mellitus. J Matern Fetal Neonatal Med 26:1112–1115CrossRefPubMedGoogle Scholar
  19. 19.
    Buck Louis GM, Grewal J, Albert PS et al (2015) Racial/ethnic standards for fetal growth: the NICHD Fetal Growth Studies. Am J Obstet Gynecol 213:449.e441CrossRefGoogle Scholar
  20. 20.
    American College of Obstetricians and Gynecologists Committee on Practice Bulletins—Obstetrics (2001) ACOG Practice Bulletin. Clinical management guidelines for obstetrician-gynecologists. Number 30, September 2001 (replaces Technical Bulletin Number 200, December 1994). Gestational diabetes. Obstet Gynecol 98:525–538CrossRefGoogle Scholar
  21. 21.
    Lao TT, Chan PL, Tam KF (2001) Gestational diabetes mellitus in the last trimester—a feature of maternal iron excess? Diabet Med 18:218–223CrossRefPubMedGoogle Scholar
  22. 22.
    Amiri FN, Basirat Z, Omidvar S, Sharbatdaran M, Tilaki KH, Pouramir M (2013) Comparison of the serum iron, ferritin levels and total iron-binding capacity between pregnant women with and without gestational diabetes. J Nat Sci Biol Med 4:302–305CrossRefPubMedPubMedCentralGoogle Scholar
  23. 23.
    Afkhami-Ardekani M, Rashidi M (2009) Iron status in women with and without gestational diabetes mellitus. J Diabetes Complicat 23:194–198CrossRefPubMedGoogle Scholar
  24. 24.
    Soubasi V, Petridou S, Sarafidis K et al (2010) Association of increased maternal ferritin levels with gestational diabetes and intra-uterine growth retardation. Diabetes Metab 36:58–63CrossRefPubMedGoogle Scholar
  25. 25.
    Sharifi F, Ziaee A, Feizi A, Mousavinasab N, Anjomshoaa A, Mokhtari P (2010) Serum ferritin concentration in gestational diabetes mellitus and risk of subsequent development of early postpartum diabetes mellitus. Diabetes Metab Syndr Obes 3:413–419CrossRefPubMedPubMedCentralGoogle Scholar
  26. 26.
    Javadian P, Alimohamadi S, Gharedaghi MH, Hantoushzadeh S (2014) Gestational diabetes mellitus and iron supplement; effects on pregnancy outcome. Acta Med Iran 52:385–389PubMedGoogle Scholar
  27. 27.
    Gungor ES, Danisman N, Mollamahmutoglu L (2007) Maternal serum ferritin and hemoglobin values in patients with gestational diabetes mellitus. Saudi Med J 28:478–480PubMedGoogle Scholar
  28. 28.
    Maitland RA, Seed PT, Briley AL et al (2014) Prediction of gestational diabetes in obese pregnant women from the UK Pregnancies Better Eating and Activity (UPBEAT) pilot trial. Diabet Med 31:963–970CrossRefPubMedGoogle Scholar
  29. 29.
    Khambalia AZ, Aimone A, Nagubandi P et al (2015) High maternal iron status, dietary iron intake and iron supplement use in pregnancy and risk of gestational diabetes mellitus: a prospective study and systematic review. Diabet Med 33:1211–1221CrossRefGoogle Scholar
  30. 30.
    Koenig MD, Tussing-Humphreys L, Day J, Cadwell B, Nemeth E (2014) Hepcidin and iron homeostasis during pregnancy. Nutrients 6:3062–3083CrossRefPubMedPubMedCentralGoogle Scholar
  31. 31.
    Walsh T, O’Broin SD, Cooley S et al (2011) Laboratory assessment of iron status in pregnancy. Clin Chem Lab Med 49:1225–1230CrossRefPubMedGoogle Scholar
  32. 32.
    van Santen S, Kroot JJ, Zijderveld G, Wiegerinck ET, Spaanderman ME, Swinkels DW (2013) The iron regulatory hormone hepcidin is decreased in pregnancy: a prospective longitudinal study. Clin Chem Lab Med 51:1395–1401PubMedGoogle Scholar
  33. 33.
    Rajpathak SN, Crandall JP, Wylie-Rosett J, Kabat GC, Rohan TE, Hu FB (2009) The role of iron in type 2 diabetes in humans. Biochim Biophys Acta 1790:671–681CrossRefPubMedGoogle Scholar
  34. 34.
    Fernandez-Real JM, McClain D, Manco M (2015) Mechanisms linking glucose homeostasis and iron metabolism toward the onset and progression of type 2 diabetes. Diabetes Care 38:2169–2176CrossRefPubMedGoogle Scholar
  35. 35.
    Facchini FS, Saylor KL (2002) Effect of iron depletion on cardiovascular risk factors: studies in carbohydrate-intolerant patients. Ann N Y Acad Sci 967:342–351CrossRefPubMedGoogle Scholar
  36. 36.
    Fernandez-Real JM, Penarroja G, Castro A, Garcia-Bragado F, Hernandez-Aguado I, Ricart W (2002) Blood letting in high-ferritin type 2 diabetes: effects on insulin sensitivity and beta-cell function. Diabetes 51:1000–1004CrossRefPubMedGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg (outside the USA)  2016

Authors and Affiliations

  • Shristi Rawal
    • 1
  • Stefanie N. Hinkle
    • 1
  • Wei Bao
    • 2
  • Yeyi Zhu
    • 1
  • Jagteshwar Grewal
    • 3
  • Paul S. Albert
    • 4
  • Natalie L. Weir
    • 5
  • Michael Y. Tsai
    • 5
  • Cuilin Zhang
    • 1
  1. 1.Epidemiology Branch, Division of Intramural Population Health Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of HealthBethesdaUSA
  2. 2.Department of EpidemiologyUniversity of Iowa College of Public HealthIowa CityUSA
  3. 3.Office of the Director, Division of Intramural Population Health Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of HealthBethesdaUSA
  4. 4.Biostatistics and Bioinformatics Branch, Division of Intramural Population Health Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of HealthBethesdaUSA
  5. 5.Department of Laboratory Medicine and PathologyUniversity of Minnesota Medical SchoolMinneapolisUSA

Personalised recommendations