Abstract
Background
Maternal iron deficiency anemia has been linked to reduced fetal growth assessed at birth, but research on elevated maternal iron levels and their impact on fetal growth during pregnancy is limited. In this study, we analysed the association between maternal serum ferritin (SF) levels without anemia during pregnancy and fetal growth in each trimester up until birth.
Methods
We prospectively analysed data from 713 mother-child pairs from the ECLIPSES cohort. Maternal SF was measured in the 1st and 3rd trimester (T) of pregnancy. SF trajectories were calculated considering the SF tertiles (low, medium, high) from T1 to T3: low FS, medium SF (reference), and high SF trajectories. Fetal biometry during gestation (estimated fetal weight (EFW), femur length, head circumference (HC); and anthropometric measurements at birth (weight, length and HC) were estimated.
Results
In the fetus, increased maternal SF levels without anemia at T1 and/or T3 were associated with lower fetal biometrics (EFW, femur length, HC) at T3. The high SF levels in both, tertiles and trajectories, were associated with lower femur length. Further, high SF levels were positively associated with greater risk of low (< 10th percentile) EFW (OR: 1.58; 95%IC, 1.14–2.18) and low femur length (when SF was modeled as either continuous, tertile, or trajectory exposure), with a doubling of the risk of low femur with high FS trajectory (OR: 4.26; 95%IC, 1.84–9.83). Conversely, low SF tertiles, without anemia, at both T1 (β:34.80, p = 0.024) and T3 (β:38.42, p = 0.040) were associated with a higher EFW. In newborns, high maternal SF levels without anemia, but not low levels, was negatively associated with birth weight and HC. A high SF trajectory doubled this negative effect on birth weight (β:-107.96, p = 0.011). The risk of having a low birth weight (< 2.500 g) was consistent with high SF levels at both T1 and T3, and was markedly tripled in the high FS trajectory (OR: 13.56; 95%IC, 2.27–81.17).
Conclusions
Our findings suggest that high levels of maternal SF during pregnancy are associated with reduced fetal growth until birth, whereas low levels without anemia do not appear to be detrimental.
Trial registration
This clinical trial is registered at www.clinicaltrialsregister.eu as EudraCT number 2012-005480-28. Registration date: April 04, 2013.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12884-026-09346-0.
Keywords: Serum ferritin, Iron status, Pregnancy, Fetal growth, Birthweight
Introduction
Anemia represents a public health concern that affects approximately 25% of pregnant women in Europe, with iron deficiency (ID) being the predominant cause [1]. However, ID without anemia often goes undiagnosed and unmonitored in daily clinical practice due to the lack of complementary measures of iron status, such as determination of serum ferritin (SF), in addition to hemoglobin (Hb) [2]. In fact, SF is considered the most reliable indicator to estimate body iron levels, before anemia appears [3]. Depending on the SF cutoffs points used and the type of population, the percentage of pregnant women in Europe suffering from ID without anemia is also frequent (10–32%) [4].
In order to meet the increased iron needs of pregnancy and prevent the onset of anemia, systematic preventive iron supplementation for all pregnant women has long been proposed. However, this systematic iron supplementation may be excessive for some women. This is particularly important for those who have good iron stores at the beginning of pregnancy and/or carry the HFE gene mutation, which increases iron absorption in the intestine and is common in our population (43,4%) [5]. As a result of systematic preventive supplementation, a certain group of women may present iron excess during pregnancy [6]. In this regard, our research group found that approximately 24% of pregnant women with elevated Hb levels at the beginning of pregnancy presented a risk of hemoconcentration at the end of gestation [7].
Both iron deficiency anemia (IDA) and excess iron stores during pregnancy are common imbalances seemingly associated to adverse maternal and neonatal consequences, including reduced fetal growth [8–10]. On one hand, maternal anemia impairs placental development, and reduces oxygen and nutrient supply to the fetus [11], while also weakening immune function and increasing infection risk, which all negatively affect fetal growth [12]. Conversely, excess maternal iron increases blood viscosity, disrupting uteroplacental blood flow and impairing fetal development [13]. This duality in iron status is supported by a recent systematic review, which confirms that the detrimental effect of maternal iron levels on offspring outcomes is U-shaped [14].
Optimal fetal growth is crucial for immediate and long-term health. Newborns with low birth weight (LBW) or small-for-gestational-age (SGA) are at greater risk of perinatal problems and are more likely to develop chronic diseases in adulthood, such as type 2 diabetes and cardiovascular diseases [15], being one of the leading causes of neonatal morbidity and mortality [16].
Notably, studies done so far have looked the effect of iron imbalances prenatally on fetal growth, focusing on anemia as exposure [10, 17], whereas ID in absence of anemia [18] or high iron status during pregnancy [19–23] have received relatively less attention. Furthermore, many studies in this field use only one determination of iron status at a single time point in gestation only one time point during gestation [20, 21, 24]. It is therefore important to determine the impact of maternal trimester-specific iron levels and longitudinal trajectories prenatally on fetal growth. In addition, although the detrimental effects of iron imbalances on fetal growth occurs prenatally, most current studies assess fetal growth using newborn anthropometric indicators such as birth weight [10, 14, 17, 25], which is an important limitation for knowing the moment at which fetal growth disturbance begins. To our knowledge, only two studies have assessed fetal growth using fetal biometry through ultrasound throughout pregnancy [21, 26]. Therefore, we aimed was to examine associations of maternal SF levels without anemia from early to late gestation with fetal growth during the trimesters of gestation until birth in a population of healthy pregnant women from the Mediterranean area.
Methods
Study design and participants
A total of 713 mother-child pairs participated in this longitudinal study, conducted within the framework of the ECLIPSES study, a community-based research initiative conducted at sexual and reproductive health care services (ASSIR) of the Catalan Institute of Health in the province of Tarragona (Catalonia, Spain) between 2013 and 2017. A description of ECLIPSES has been published elsewhere [27]. Briefly, the ECLIPSES study recruited 791 healthy pregnant women during their initial routine control visit, and they were closely monitored at weeks 12, 24, and 36 of gestation and after delivery. The main inclusion criteria were over 18 years old, within ≤ 12 weeks of gestation, and not having anemia (Hb ≥ 110 g/L on week 12). Women who had taken > 10 mg iron daily for three months prior to the 12th week of gestation and those with multiple pregnancies were excluded. Further details on the inclusion/exclusion criteria can be found elsewhere [27]. Out of the initial cohort of non-anaemic pregnant women recruited (n = 791), participants with missing data on SF during the 1st (T1) and/or 3rd (T3) trimesters and those who had levels of Hb < 105 g/l at T3 were also excluded for the current analysis (see Fig. 1). The ECLIPSES trial is registered at www.clinicaltrialsregister.eu as EudraCT number 2012-005480-28. This study was approved by the Ethics Committee of the Pere Virgili Health Research Institute (IISPV) and complies with the principles of the Declaration of Helsinki. All participants signed an informed consent form.
Fig. 1.
Flowchart of the study. Abbreviations: SF, serum ferritin; Hb, hemoglobin. * Missing data on fetal outcomes at 2nd trimester: estimated weight (n = 26), femur length and head circumference (n = 116). ** Missing data on fetal outcomes at 3rd trimester: estimated weight (n = 15), femur length and head circumference (n = 48). ***Missing data on birth outcomes: weight (n = 65), length (n = 192) and head circumference (n = 295). † Missing data on fetal outcomes at 3rd trimester: estimated weight (n = 7), femur length and head circumference (n = 32). ‡Missing data on birth outcomes: weight (n = 13), length (n = 57) and head circumference (n = 149)
Exposure assessment
In this current research, the main exposure variable was the maternal serum levels of SF during pregnancy, determined by turbidimetric immunoassay. For each trimester (T1 and T3), SF was analyzed as a continuous exposure variable and as a categorical variable in tertiles. In addition, three trajectories were constructed from the SF tertiles (low, medium, high) from T1 and T3 of pregnancy to examine the effects of persistent antenatal exposure to both low and high levels of FS: (1) Low SF trajectory, including two scenarios: low-low and medium-low tertiles; (2) High SF trajectory, including two scenarios: high-high and medium-high tertiles; (3) Medium SF trajectory, used as the reference category, including all other scenarios.
Outcome assessment
The main outcome variable was fetal growth estimated by ultrasound and anthropometric measurements at birth. Fetal ultrasound scans were performed following the recommendations of the International Society of Ultrasound in Obstetrics and Gynecology in T1 (81.2% before 13 weeks of gestation), T2 and T3. In T1 the following fetal measurements were taken in mm: crown-rump length (CRL) and biparietal diameter. In T2 and T3, femur length (FL), head circumference (HC), and abdominal circumference (AC) were obtained, and estimated fetal weight (EFW) was calculated at both time points using Hadlock’s three-parameter formula (HC, AC and FL): EFW = 10^(1.326 + 0.0107 × HC + 0.0438 × AC + 0.158 × FL − 0.00326 × AC × FL) [28]. All fetal growth parameter values were adjusted for gestational age (weeks) at the time of ultrasound measurement using the residual method [29]. Gestational age was determined by assessing the coincidence (within 7 days) between the last menstrual period reported at recruitment and the ultrasound estimate, otherwise the ultrasound estimate was used. Babies were assigned a sex at birth based on the observation of their external genitalia.
Fetal biometrics in T2 and T3 (EFW, FL and HC) were analyzed as continuous variables. Subsequently, to classify the group at the highest fetal growth risk, fetuses with lower growth than expected were defined as a gestational age-and sex-adjusted EFW, FL and HC below the 10th percentile in the study cohort [30]. Fetuses with growth above the 10th percentile were considered appropriate-for-gestational-age and used as the reference group.
In the newborn, weight (g), length (cm) and HC (cm) were measured by the obstetrician or midwife using standard procedures. Low birth weight (LBW) was defined as less than 2500 g. Low size for length and HC, adjusted for gestational age and sex, was defined as being below the 10th percentile based on z-scores according to the IN-TERGROWTH-21st standards as reference [31]. It is opportune to mention that all newborns included in the present study were born at or beyond 35 weeks of gestation—1.1% (n = 8) were late preterm (≥ 35 to < 37 weeks), and 98.9% (n = 705) were term births (≥ 37–42 weeks).
Covariate assessment
Regarding other variables, at enrolment and once during each pregnancy trimester, data on maternal age, socio-demographic status, lifestyle habits, and medical and obstetric history were collected through individual interviews. Family socioeconomic status was estimated with the Hollingshead index [32] by combining data about the mother’s and her partner’s (if there is one) level of education and occupational status, classified according to the Catalan Classification of Occupations (CCO-2011) [33]. It was then classified as low, medium or high. The Fagerström questionnaire [34] was used to evaluate smoking during pregnancy and the women were classified as smokers and non-smokers. The International Physical Activity Questionnaire (IPAQ) [35] was used to record participants’ physical activity (PA). PA was derived from total metabolic equivalent (MET) values based on the frequency and duration of walking and moderate and vigorous intensity activity, and the women were divided into tertiles based on their METs-min/week: tertile 1: ≤1158; tertile 2: 1159–3323; and tertile 3: ≥3324 METs-min/week.
In addition, dietitians used a 45-item self-administered food frequency questionnaire validated in our population [36] to produce a relative Mediterranean Diet (rMedDiet) score evaluating maternal whole diet quality [37]. For this purpose, the rMedDiet score (ranging from 0 to 18 points) was divided into tertiles: tertile 1: ≤ 8; tertile 2: 9–12; and tertile 3: ≥12 points.
Maternal weight (kg, to the nearest 0.1 kg) and height (cm, to the nearest 0.1 cm) were measured at the beginning of pregnancy and during follow-up using a calibrated SECA® scale with stadiometer. Body mass index (BMI: weight (kg)/height (m)2) was classified as normal weight (BMI 18.5–24.9 kg/m2), overweight (BMI 25.0–29.9 kg/m2) or obese (BMI ≥ 30 kg/m2) [38]. Total gestational weight gain (GWG: T3-T1 weights) was calculated.
The Hb levels were measured using a Coulter Gen-S analyzer (Coulter, Hia-leah, FL, USA). Genetic determinations of HFE gene mutations (C282Y, H63D and S65C) were performed using DNA extracted from leukocytes with the polymerase chain reaction technique determined by enzyme immunoassay (ELISA) expressed in mg/L. C-reactive protein (CRP) levels were also measured in serum samples collected during the first trimester using an immunoturbidimetric method. Iron supplementation groups from the ECLISPES trial were also considered as a categorical covariate: stratum 1 (baseline Hb: 110–130 g/L) received 40 or 80 mg iron per day, while stratum 2 (baseline Hb > 130 g/L) received 40 or 20 mg iron per day during pregnancy.
Statistical analysis
Statistical analyses were performed using STATA, version 15.0 (StataCorp LP, College Station, TX, USA). Descriptive data are presented as mean (SD), median (IQR) or numbers (%). Comparisons according to SF tertiles in T1 of pregnancy were performed using ANOVA and chi-square tests as appropriate. SF was assessed for normality using both the Shapiro-Wilk test and visual inspection (quantile-quantile plot). As SF was right-skewed, it was log-transformed prior to analyses. Associations between SF levels in T1 and T3 and the SF trajectory from T1 and T3 with fetal biometrics (as continuous outcomes) in T2 and T3 and at birth anthropometric parameters were evaluated using separate unadjusted linear regression analyses. For these analyses, maternal SF was treated both as a continuous variable (per 1-SD increase in log-transformed SF concentration) and as a categorical variable using tertiles, with the lowest tertile as reference. Estimates were reported as β coefficient with 95% confidence intervals (CIs). In addition to the unadjusted model, a multivariable model was fitted including confounding factors selected a priori based on biological plausibility, prior literature, and established expertise in the field. These covariates included iron supplementation group (stratum 1: 40 mg/d (ref.) and 80 mg/d; stratum 2: 20 mg/d and 40 mg/d), maternal age (years), maternal early-pregnancy BMI categories (normal weight (ref.), overweight, obesity), family SES (low (ref.), medium, high), smoking (no (ref.), yes), Hb at T1 (g/L, as an indicator of baseline hematological status, not as a mediator of the main causal pathway), CRP ≥ 1 mg/L (80th percentile) at T1 (no (ref.), yes), carriers of HFE gene mutation (no (ref.), yes), sex of baby, and gestational weight gain (kg, only for fetal growth parameter analyses in T3 and at birth). Multiple imputation by the Markov chain Monte Carlo method (20 copies) was used in all multivariable analyses to impute missing values for covariates HFE gene mutation (n = 136) and CRP (n = 99). Estimates were combined using Rubin’s rule.
Additionally, separate multivariable logistic regression models, adjusted for the above-mentioned confounders, were used to estimate the odds ratios (ORs) and 95% CIs for low fetal growth (< 10th percentile) compared with normal fetal growth (defined as the 10th-90th percentiles) in terms of weight, length and HC as binary outcomes in T2, T3 and at birth. For birth weight, LBW (< 2.500 g) was compared with normal birth weight (2,500–4,000 g). These models were fitted according to trimester-specific SF levels and trajectories of SF tertiles.
Finally, restricted cubic spline regressions models with four knots, using the 50th percentile of SF as the reference, were fitted to explore potential non-linear adjusted associations between maternal SF concentrations in both T1 and T3 and fetal growth outcomes assessed at T3 and at birth.
In all analyses, a two-sided p < 0.05 was considered significant. The data were analyzed using the STATA application, version 15.0 (StataCorp LP, College Station, TX, USA).
Results
Table 1 presents the general characteristics of the studied population by T1 SF tertiles (low: SF < 26 µg/L, median 15.4 (1.4); medium SF 26–42.4 µg/L, median 36.0 (1.2); and high SF ≥ 42.5 µg/L, median 67.0 (1.4)). Maternal characteristics include sociodemographic status, lifestyle, amount of iron supplemented in each intervention group, biochemical parameters of iron status, and presence of genetic alterations in the HFE gene. Fetal growth parameters were described in T1, T2 and T3 of gestation, with higher scores observed in T3 in EFW (p = 0.015) and femur length (p = 0.065) in the high tertile of SF. No significant differences in birth anthropometric measures were observed in the newborn according to the T1 SF tertiles.
Table 1.
Sociodemographic, lifestyle, maternal iron status, and fetal/newborn growth according to first trimester ferritin tertiles
| Maternal general characteristics | All | Tertiles of SF in the first trimester | p** | ||
|---|---|---|---|---|---|
| Low (T1, < 26 µg/L) |
Medium (T2, 26–42.4 µg/L) |
High (T3, ≥ 42.5 µg/L) |
|||
| No. (%) | 713 (100) | 236 (33.1) | 260 (36.5) | 217 (30.4) | |
| Age, mean (SD), year | 30.4 (5.1) | 30.3 (5.3) | 30.6 (5.2) | 30.1 (4.8) | 0.5132 |
| Iron supplementation group, no. (%) | |||||
| Stratum 1/40 g/d | 229 (32.1) | 75 (31.8) | 91 (35.0) | 63 (29.0) | |
| Stratum 1/80 g/d | 236 (33.1) | 89 (37.7) | 79 (30.4) | 68 (31.3) | 0.363 |
| Stratum 2/20 g/d | 122 (17.1) | 35 (14.8) | 43 (16.5) | 44 (20.3) | |
| Stratum 2/40 g/d | 126 (17.7) | 37 (15.7) | 47 (18.1) | 42 (19.4) | |
| BMI, mean (SD), kg/m2 | 25.1 (4.5) | 25.0 (4.3) | 25.3 (4.5) | 24.9 (4.6) | 0.672 |
| GWG, mean (SD), kg | 10.3 (3.7) | 10.4 (3.9) | 10.1 (3.7) | 10.5 (3.4) | 0.515 |
| Family SES, no. (%) | |||||
| Low | 115 (16.1) | 41 (17.4) | 41 (15.8) | 33 (15.2) | 0.684 |
| Medium | 477 (66.9) | 161 (68.2) | 169 (65.0) | 147 (67.7) | |
| High | 121 (17.0) | 34 (14.4) | 50 (19.2) | 37 (17.1) | |
| Smoking, no. (%) | 125 (17.5) | 30 (12.7) | 45 (17.3) | 50 (23.0) | 0.015 |
| PA (METs-min/week), no. (%) | |||||
| Tertile 1 (≤ 1158) | 236 (33.1) | 75 (31.8) | 88 (33.9) | 73 (33.7) | 0.966 |
| Tertile 2 (1159–3323) | 325 (45.6) | 112 (47.5) | 117 (45.0) | 96 (44.2) | |
| Tertile 3 (≥ 3324) | 152 (21.3) | 49 (20.7) | 55 (21.1) | 48 (22.1) | |
| rMedDiet score (point), no. (%) | |||||
| Tertile 1 (≤ 8) | 236 (34.0) | 74 (33.2) | 86 (36.3) | 66 (32.4) | 0.855 |
| Tertile 2 (9–11) | 285 (42.9) | 94 (42.2) | 101 (42.6) | 90 (44.1) | |
| Tertile 3 (≥ 12) | 153 (23.1) | 55 (24.6) | 50 (21.1) | 48 (23.5) | |
| Missing | 49 | 13 | 23 | 13 | |
| Hb in 1st trimester, mean (SD), g/L | 127.9 (7.8) | 126.6 (7.7) | 128.3 (7.6) | 128.8 (8.0) | 0.007 |
| SF, median (IQR), µg/L | |||||
| 1st trimester | 37.5 (21.3, 46.9) | 16.5 (11.9, 21.0) | 38.2 (31.3, 41.5) | 63.0 (51.2, 84.3) | < 0.001 |
| 3rd trimester | 14.6 (10.1, 19.2) | 11.2 (9.2, 16.8) | 14.6 (10.1, 17.2) | 18.9 (12.7, 26.6) | < 0.001 |
| H63D HFE gene mutation, n (%) | 162 (27.9) | 61 (29.8) | 48 (25.7) | 53 (28.0) | 0.665 |
| Missing | 132 | 31 | 73 | 28 | |
| C282Y HFE gene mutation, n (%) | 26 (4.5) | 10 (4.9) | 6 (3.2) | 10 (5.3) | 0.585 |
| Missing | 132 | 31 | 73 | 28 | |
| CRP ≥ 1 (80th percentile) 1st trimester, no. (%) | 128 (20.8) | 39 (16.7) | 34 (20.6) | 55 (25.4) | 0.076 |
| Missing | 97 | 2 | 95 | 0 | |
| Fetal growth parameters* | |||||
| First trimester | |||||
| GA, mean (SD), weeks | 12.0 (0.6) | 12.0 (0.6) | 12.0 (0.6) | 11.9 (0.6) | 0.687 |
| CRL, mean (SD), mm | 59.9 (5.1) | 60.2 (4.8) | 59.9 (5.1) | 59.7 (5.4) | 0.579 |
| BD, mean (SD), mm | 20.5 (1.9) | 20.4 (1.5) | 20.5 (1.9) | 20.7 (2.2) | 0.551 |
| Second trimester | |||||
| GA, mean (SD), weeks | 20.0 (0.8) | 20.0 (0.7) | 20.1 (0.8) | 20.0 (0.8) | 0.808 |
| EFW, mean (SD), g | 359.2 (35.6) | 360.9 (42.2) | 358.9 (33.8) | 357.8 (29.4) | 0.648 |
| FL, mean (SD), mm | 33.0 (1.7) | 32.9 (1.8) | 33.1 (1.6) | 33.0 (1.9) | 0.611 |
| HC, mean (SD), mm | 176.9 (6.3) | 177.0 (6.7) | 177.1 (6.2) | 176.5 (6.1) | 0.485 |
| Third trimester | |||||
| GA, mean (SD), weeks | 33.3 (1.3) | 33.3 (1.3) | 33.2 (1.3) | 33.4 (1.3) | 0.269 |
| EFW, mean (SD), g | 2190.0 (167.5) | 2215.2 (178.1) | 2180.6 (168.1) | 2173.0 (151.5) | 0.015 |
| FL, mean (SD), mm | 64.0 (2.3) | 64.2 (2.1) | 64.1 (2.3) | 63.7 (2.4) | 0.065 |
| HC, mean (SD), mm | 303.1 (8.6) | 303.9 (9.3) | 302.9 (8.5) | 302.2 (8.0) | 0.114 |
| Birth parameters* | |||||
| GA, mean (SD), weeks | 39.5 (1.5) | 39.6 (1.5) | 39.6 (1.5) | 39.5 (1.5) | 0.580 |
| Female sex, no. (%) | 362 (50.8) | 131 (55.5) | 123 (47.3) | 108 (49.8) | 0.178 |
| Birthweight, mean (SD), g | 3297.0 (410.8) | 3312.7 (421.7) | 3294.9 (401.4) | 3282.2 (410.7) | 0.747 |
| Length, mean (SD), cm | 50.0 (1.9) | 49.9 (2.0) | 50.2 (2.1) | 50.0 (1.6) | 0.514 |
| HC, mean (SD), cm | 35.0 (1.3) | 34.9 (1.2) | 35.1 (1.5) | 34.9 (1.3) | 0.404 |
Values are expressed in means ± SD or number (%)
Abbreviations: BMI body mass index, GWG gestational weight gain, PA physical activity, SES socioeconomic status, METs metabolic equivalents, rMedDiet Mediterranean diet, Hb hemoglobin, SF serum ferritin, CRP C-reactive protein, GA gestational age, SD standard deviation, IQR interquartile range, CRL crown–rump length, BD biparietal diameter, EFW estimated fetal weight, FL femur length, HC head circumference
*All fetal growth parameter values were adjusted for gestational age (weeks) at the time of ultrasound measurement or for gestational age (weeks) at birth by the residual method
The significance of numbers in bold is p-value < 0.05. **p-value for differences across SF level categories as derived from ANOVA or Chi-square test, as appropriate
Tables 2 and 3 describes the relationship of SF levels (modeled as continuous, tertiles and trajectories exposures) during gestation with fetal growth in T2 and T3, and anthropometric parameters at birth. After controlling for possible confounders, significant inverse associations were observed between high SF levels in both T1 and T3 and EFW, femur length and HC at T3, and a high SF trajectory, as opposed to a medium SF trajectory, was associated with lower fetal femur length (Table 2). While low tertile SF in T1 and T3 was associated with higher T3 EFW.
Table 2.
Associations of maternal SF levels and their trajectories during pregnancy with fetal growth parameters
| Fetal growth outcomes during pregnancy | ||||||
|---|---|---|---|---|---|---|
| Estimated fetal weight (g) | p | Femur length (mm) | p | Head C. (mm) | p | |
| β (95% CI) | β (95% CI) | β (95% CI) | ||||
| Second trimester | ||||||
| SF levels (µg/L) in the first trimester | ||||||
| Ln-SF-continuous (per 1-SD increase) | ||||||
| Crude model | 0.21 (-2.47, 2.89) | 0.879 | 0.02 (-0.12, 0.16) | 0.778 | -0.35 (-0.85, 0.15) | 0.170 |
| Adjusted model* | 0.15 (-2.58, 2.89) | 0.913 | 0.01 (-0.13, 0.15) | 0.884 | -0.41 (-0.91, 0.10) | 0.118 |
| SF-tertiles (adjusted model*) | ||||||
| Low (T1, < 26 µg/L) | 2.00 (-4.54, 8.56) | 0.548 | -0.18 (-0.52, 0.16) | 0.307 | 0.21 (-1.03, 1.45) | 0.741 |
| Medium (T2, 26–42.4 µg/L) | 0 Ref. | 0 Ref. | 0 Ref. | |||
| High (T3, ≥ 42.5 µg/L) | -1.35 (-7.99, 5.29) | 0.690 | -0.08 (-0.42, 0.27) | 0.660 | -0.73 (-1.99, 0.53) | 0.257 |
| Third trimester | ||||||
| SF levels (µg/L) in the first trimester | ||||||
| Ln-SF-continuous (per 1-SD increase) | ||||||
| Crude model | -20.29 (-32.81, -7.77) | 0.002 | -0.18 (-0.36, -0.02) | 0.033 | -0.94 (-1.59, -0.28) | 0.005 |
| Adjusted model* | -20.33 (-33.06, -7.59) | 0.002 | -0.20 (-0.37, -0.02) | 0.028 | -1.05 (-1.71, -0.39) | 0.002 |
| SF-tertiles (adjusted model*) | ||||||
| Low (T1, < 26 µg/L) | 34.80 (4.64, 64.96) | 0.024 | 0.10 (-0.32, 0.52) | 0.651 | 1.34 (-0.25, 2.94) | 0.098 |
| Medium (T2, 26–42.4 µg/L) | 0 Ref. | 0 Ref. | 0 Ref. | |||
| High (T3, ≥ 42.5 µg/L) | -6.76 (-37.57, 24.05) | 0.667 | -0.41 (-0.84, 0.02) | 0.065 | -0.67 (-2.29, 0.94) | 0.412 |
| SF levels (µg/L) in the third trimester | ||||||
| Ln-SF-continuous (per 1-SD increase) | ||||||
| Crude model | -21.71 (-36.60, -6.81) | 0.004 | -0.26 (-0.46, -0.06) | 0.010 | -1.01 (-1.74, -0.28) | 0.007 |
| Adjusted model* | -22.95 (-39.18, -6.72) | 0.006 | -0.32 (-0.53, -0.11) | 0.003 | -1.36 (-2.15, -0.57) | 0.001 |
| SF-tertiles (adjusted model*) | ||||||
| Low (T1, < 11 µg/L) | 38.42 (1.71, 75.13) | 0.040 | 0.14 (-0.34, 0.63) | 0.519 | 0.89 (-0.92, 2.71) | 0.331 |
| Medium (T2, 11–17.1 µg/L) | 0 Ref. | 0 Ref. | 0 Ref. | |||
| High (T3, ≥ 17.2 µg/L) | -12.81 (-49.32, 23.69) | 0.491 | -0.64 (-1.11, -0.16) | 0.008 | -1.47 (-3.27, 0.32) | 0.107 |
| Trajectories of SF tertiles from first to third trimester | ||||||
| Low SF trajectory | 22.79 (-13.21, 58.79) | 0.214 | 0.27 (-0.19, 0.74) | 0.254 | 1.66 (-0.11, 3.44) | 0.066 |
| Medium SF trajectory (adjusted model*) | 0 Ref. | 0 Ref. | 0 Ref. | |||
| High SF trajectory | -32.31 (-69.49, 4.87) | 0.088 | -0.68 (-1.17, -0.20) | 0.005 | -0.70 (-2.53, 1.12) | 0.449 |
Values are regression β coefficient (β) and 95% confidence intervals (CIs). All fetal growth parameter values were adjusted for gestational age (weeks) at the time of ultrasound measurement or for gestational age (weeks) at birth by the residual method
*Model adjusted for iron supplementation group (stratum 1/40 mg/d (ref.), stratum 1/80 mg/d, stratum 2/20 mg/d, stratum 2/40 mg/d), sex of baby, maternal age (years), maternal early-pregnancy BMI categories (normal weight (ref.), overweight, obesity), family SES (low (ref.), medium, high), smoking (no (ref.), yes), Hb in 1st trimester (g/L, only for SF 1st trimester and trajectories analyses), Hb in 3rd trimester (g/L, only for SF 3rd trimester and trajectories analyses), CRP ≥ 1 (80th percentile) in 1st trimester (no (ref.), yes), carrier of HFE gene mutation (no (ref.), yes), and gestational weight gain (kg, only for fetal growth parameter 3rd trimester analyses)
Abbreviations: SF serum ferritin, Ref reference, Head C head circumference
The SF trajectories were calculated from the first to the third trimester, based on SF tertiles at both evaluations. The low SF trajectory included two scenarios: low-low and medium-low tertiles. The high SF trajectory included two scenarios: high-high and medium-high tertiles. The medium SF trajectory was used as the reference category and included all other scenarios: medium-medium, low-medium, high-medium, low-high and high-low tertiles. The significance of numbers in bold is p-value < 0.05
Table 3.
Associations of maternal SF levels and their trajectories during pregnancy with anthropometric parameters at birth
| Birth outcomes | ||||||
|---|---|---|---|---|---|---|
| Birthweight (g) | p | Length (cm) | p | Head C. (cm) | p | |
| β (95% CI) | β (95% CI) | β (95% CI) | ||||
| SF levels (µg/L) in the first trimester | ||||||
| Ln-SF-continuous (per 1-SD increase) | ||||||
| Crude model | -20.95 (-53.01, 11.11) | 0.200 | -0.04 (-0.20, 0.13) | 0.671 | -0.06 (-0.18, 0.06) | 0.312 |
| Adjusted model* | -20.38 (-51.88, 11.10) | 0.204 | -0.02 (-0.19, 0.14) | 0.784 | -0.07 (-0.19, 0.05) | 0.275 |
| SF-tertiles (adjusted model*) | ||||||
| Low (T1, < 26 µg/L) | 27.39 (-47.65, 102.43) | 0.474 | -0.21 (-0.60, 0.19) | 0.302 | -0.10 (-0.40, 0.21) | 0.530 |
| Medium (T2, 26–42.4 µg/L) | 0 Ref. | 0 Ref. | 0 Ref. | |||
| High (T3, ≥ 42.5 µg/L) | -1.81 (-78.37, 74.76) | 0.963 | -0.19 (-0.59, 0.21) | 0.345 | -0.21 (-0.52, 0.10) | 0.183 |
| SF levels (µg/L) in the third trimester | ||||||
| Ln-SF-continuous (per 1-SD increase) | ||||||
| Crude model | -57.20 (-91.70, -22.70) | 0.001 | -0.02 (-0.19, 0.15) | 0.810 | -0.18 (-0.30, -0.05) | 0.007 |
| Adjusted model* | -49.28 (-85.62, -12.94) | 0.008 | 0.00 (-0.18, 0.19) | 0.992 | -0.13 (-0.27, 0.00) | 0.050 |
| SF-tertiles (adjusted model*) | ||||||
| Low (T1, < 11 µg/L) | 19.86 (-63.21, 102.95) | 0.639 | -0.07 (-0.49, 0.35) | 0.748 | -0.08 (-0.41, 0.24) | 0.612 |
| Medium (T2, 11-0.17.1 µg/L) | 0 Ref. | 0 Ref. | 0 Ref. | |||
| High (T3, ≥ 17.2 µg/L) | -72.93 (-155.19, 9.31) | 0.082 | -0.13 (-0.55, 0.28) | 0.531 | -0.42 (-0.73, -0.12) | 0.007 |
| Trajectories of SF tertiles from first to third trimester | ||||||
| Low SF trajectory | -25.27 (-106.29, 55.74) | 0.540 | -0.14 (-0.55, 0.27) | 0.497 | 0.13 (-0.18, 0.45) | 0.409 |
| Medium SF trajectory (adjusted model*) | 0 Ref. | 0 Ref. | 0 Ref. | |||
| High SF trajectory | -107.96 (-191.37, -24.55) | 0.011 | -0.20 (-0.62, 0.22) | 0.345 | -0.21 (-0.52, 0.10) | 0.175 |
Values are regression β coefficient (β) and 95% confidence intervals (CIs). All anthropometric parameter values were adjusted for gestational age (weeks) at birth using residual method
*Model adjusted for iron supplementation group (stratum 1/40 mg/d (ref.), stratum 1/80 mg/d, stratum 2/20 mg/d, stratum 2/40 mg/d), sex of baby, maternal age (years), maternal early-pregnancy BMI categories (normal weight (ref.), overweight, obesity), family SES (low (ref.), medium, high), smoking (no (ref.), yes), Hb in 1st trimester (g/L, only for SF 1st trimester and trajectories analyses), Hb in 3rd trimester (g/L, only for SF 3rd trimester and trajectories analyses), CRP ≥ 1 (80th percentile) in 1st trimester (no (ref.), yes), carrier of HFE gene mutation (no (ref.), yes), and gestational weight gain (kg)
Abbreviations: SF serum ferritin, Ref reference, Head C head circumference
The SF trajectories were calculated from the first to the third trimester, based on SF tertiles at both evaluations. The low SF trajectory included two scenarios: low-low and medium-low tertiles. The high SF trajectory included two scenarios: high-high and medium-high tertiles. The medium SF trajectory was used as the reference category and included all other scenarios: medium-medium, low-medium, high-medium, low-high and high-low trajectory tertiles. The significance of numbers in bold is p-value < 0.05
For the birth outcomes, the T3 SF levels were liked to lower birth weight and HC at birth. Similarly, the high tertile of SF and a high SF trajectory were negatively associated with HC and lower birth weight, respectively (Table 3).
To further explore these associations, non-linear relationships were examined using restricted cubic spline regression models with four knots, with the 50th percentile of SF as the reference. No evidence of non-linearity was observed for fetal growth outcomes at T3 or at birth (all p > 0.05). Nevertheless, the spline curves showed patterns compatible with an inverse association between maternal SF concentrations in T1 and T3 and fetal growth parameters, particularly across the SF range above the 50th percentile, in line with the main linear regression findings (Supplementary Figs. 1 and 2).
Tables 4 and 5 describe the multivariable-adjusted associations between maternal SF levels during gestation (modeled as continuous, tertiles, and trajectories exposures) and the risk of low fetal growth and small size at birth. The SF levels in both T1 (OR: 1.58 (95% CI: 1.17, 2.15, p = 0.003) and T3 (OR: 2.01 (1.37, 3.03), p < 0.001) were associated with an increased risk of low fetal growth in terms of femur length in T3, with a high FS trajectory from T1 to T3 doubling the risk (OR: 4.26 (1.84, 9.83), p = 0.001). Furthermore, SF levels in T3 posed a greater risk of low fetal growth in terms of EFW in T3 (OR: 1.58 (1.14, 2.18), p = 0.005). While low SF tertile in T1 was associated with lower fetal growth for HC (OR: 0.54 (0.28, 1.05), p = 0.071) in T3 (Table 4). No associations were observed between maternal SF levels in T1 and fetal biometric outcomes assessed in T2 (Tables 2, 3 and 4).
Table 4.
Associations of maternal SF levels and trajectories with the risk of low fetal growth vs. normal fetal growth (10th–90th percentiles)
| Fetal growth outcomes during pregnancy | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Low estimated fetal weight (< 10th percentile) | p | Low femur length (< 10th percentile) |
p | Low head c. (< 10th percentile) | p | ||||
| Total/No. (%) | OR (95% CI) | Total/No. (%) | OR (95% CI) | Total/No. (%) | OR (95% CI) | ||||
| Second trimester | |||||||||
| SF levels (µg/L) in the first trimester | |||||||||
| Ln-SF-continuous (per 1-SD increase)* | 619/67 (10.8) | 1.01 (0.78, 1.31) | 0.958 | 543/48 (8.8) | 0.98 (0.72, 1.32) | 0.875 | 541/54 (10.1) | 1.01 (0.76, 1.33) | 0.941 |
| SF-tertiles (adjusted model*) | |||||||||
| Low (T1, < 26 µg/L) | 203/21 (10.3) | 0.98 (0.52, 1.86) | 0.965 | 182/19 (10.4) | 1.35 (0.65, 2.82) | 0.426 | 176/19 (10.8) | 1.24 (0.61, 2.53) | 0.555 |
| Medium (T2, 26–42.4 µg/L) | 218/24 (11.0) | 1 Ref. | 192/15 (7.8) | 1 Ref. | 190/17 (8.9) | 1 Ref. | |||
| High (T3, ≥ 42.5 µg/L) | 198/22 (11.1) | 0.96 (0.51, 1.80) | 0.914 | 169/14 (8.3) | 1.05 (0.48, 2.30) | 0.900 | 175/18 (10.3) | 1.27 (0.62, 2.61) | 0.510 |
| Third trimester | |||||||||
| SF levels (µg/L) in the first trimester | |||||||||
| Ln-SF-continuous (per 1-SD increase)* | 628/68 (10.8) | 1.07 (0.81, 1.40) | 0.633 | 603/53 (8.8) | 1.58 (1.17, 2.15) | 0.003 | 599/61 (10.2) | 1.21 (0.92, 1.58) | 0.181 |
| SF-tertiles (adjusted model*) | |||||||||
| Low (T1, < 26 µg/L) | 198/18 (9.1) | 0.61 (0.32, 1.15) | 0.127 | 205/12 (5.9) | 0.67 (0.31, 1.50) | 0.337 | 195/16 (8.2) | 0.54 (0.28, 1.05) | 0.071 |
| Medium (T2, 26–42.4 µg/L) | 233/31 (13.3) | 1 Ref. | 208/16 (7.7) | 1 Ref. | 213/28 (13.2) | 1 Ref. | |||
| High (T3, ≥ 42.5 µg/L) | 197/19 (9.6) | 0.67 (0.36, 1.24) | 0.202 | 190/25 (13.2) | 1.80 (0.92, 3.54) | 0.088 | 191/17 (8.9) | 0.64 (0.33, 1.22) | 0.176 |
| SF levels (µg/L) in the third trimester | |||||||||
| Ln-SF-continuous (per 1-SD increase)* | 474/51 (10.7) | 1.58 (1.14, 2.18) | 0.005 | 453/34 (7.5) | 2.01 (1.37, 3.03) | < 0.001 | 450/38 (8.4) | 1.40 (0.96, 2.03) | 0.081 |
| SF-tertiles (adjusted model*) | |||||||||
| Low (T1, < 11 µg/L) | 151/9 (6.0) | 0.47 (0.20, 1.12) | 0.091 | 147/5 (3.4) | 0.73 (0.23, 2.35) | 0.598 | 146/10 (6.9) | 0.64 (0.26, 1.59) | 0.340 |
| Medium (T2, 11-0.17.1 µg/L)* | 169/19 (11.2) | 1 Ref. | 158/8 (5.6) | 1 Ref. | 158/14 (8.9) | 1 Ref. | |||
| High (T3, ≥ 17.2 µg/L) | 154/23 (15.0) | 1.28 (0.65, 2.53) | 0.475 | 148/21 (14.2) | 2.85 (1.18, 6.83) | 0.019 | 146/14 (9.6) | 1.18 (0.52, 2.66) | 0.690 |
| Trajectories of SF tertiles from first to third trimester | |||||||||
| Low SF trajectory | 126/9 (7.1) | 0.67 (0.29, 1.53) | 0.342 | 123/4 (3.3) | 0.77 (0.23, 2.59) | 0.671 | 120/9 (7.5) | 0.78 (0.33, 1.85) | 0.574 |
| Medium SF trajectory (adjusted model*) | 228/23 (10.1) | 1 Ref. | 215/10 (4.7) | 1 Ref. | 217/20 (9.2) | 1 Ref. | |||
| High SF trajectory | 120/19 (15.8) | 1.61 (0.81, 3.21) | 0.171 | 115/20 (17.4) | 4.26 (1.84, 9.83) | 0.001 | 113/9 (8.0) | 0.88 (0.37, 2.11) | 0.780 |
Results are from logistic regression analyses, which were run separately for each outcome. Values are odds ratios (ORs) and 95% confidence intervals (CIs)
*Model adjusted for iron supplementation group (stratum 1/40 mg/d (ref.), stratum 1/80 mg/d, stratum 2/20 mg/d, stratum 2/40 mg/d), sex of baby, maternal age (years), maternal early-pregnancy BMI categories (normal weight (ref.), overweight, obesity), family SES (low (ref.), medium, high), smoking (no (ref.), yes), Hb in 1st trimester (g/L, only for SF 1st trimester and trajectories analyses), Hb in 3rd trimester (g/L, only for SF 3rd trimester and trajectories analyses), CRP ≥ 1 (80th percentile) in 1st trimester (no (ref.), yes), carrier of HFE gene mutation (no (ref.), yes), and gestational weight gain (kg, only for fetal growth parameter analyses in the third trimester)
Abbreviations: SF serum ferritin, Ref reference, Head C head circumference
The SF trajectories were calculated from the first to the third trimester, based on SF tertiles at both evaluations. The low SF trajectory included two scenarios: low-low and medium-low tertiles. The high SF trajectory included two scenarios: high-high and medium-high tertiles. The medium SF trajectory was used as the reference category and included all other scenarios: medium-medium, low-medium, high-medium, low-high and high-low trajectory tertiles. The significance of numbers in bold is p-value < 0.05
Table 5.
Associations of maternal SF levels and trajectories with the risk of low growth vs. normal fetal growth at birth (birth weight 2.500–4.000 g or 10th–90th percentiles)
| Birth outcomes | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Low birth weight (< 2.500 g) |
p | Low birth length (< 10th percentile) | p | Low head c. (< 10th percentile) | p | ||||
| Total/No. (%) | OR (95% CI) | Total/No. (%) | OR (95% CI) | Total/No. (%) | OR (95% CI) | ||||
| SF levels (µg/L) in the first trimester | |||||||||
| Ln-SF-continuous (per 1-SD increase)* | 612/27 (4.4) | 1.82 (1.11, 2.98) | 0.017 | 469/59 (12.6) | 0.89 (0.66, 1.19) | 0.421 | 332/27 (8.1) | 1.09 (0.73, 1.63) | 0.668 |
| SF-tertiles (adjusted model*) | |||||||||
| Low (T1, < 26 µg/L) | 203/5 (2.5) | 0.49 (0.13, 1.87) | 0.298 | 156/24 (15.4) | 1.47 (0.76, 2.84) | 0.253 | 118/8 (6.8) | 0.60 (0.22, 1.66) | 0.326 |
| Medium (T2, 26–42.4 µg/L) | 220/9 (4.1) | 1 Ref. | 164/20 (12.2) | 1 Ref. | 96/10 (10.4) | 1 Ref. | |||
| High (T3, ≥ 42.5 µg/L) | 189/13 (6.9) | 2.01 (0.72, 5.65) | 0.183 | 149/15 (10.1) | 0.81 (0.39, 1.67) | 0.561 | 118/9 (7.6) | 0.65 (0.24, 1.75) | 0.397 |
| SF levels (µg/L) in the third trimester | |||||||||
| Ln-SF-continuous (per 1-SD increase)* | 492/17 (3.5) | 3.87 (1.84, 8.13) | < 0.001 | 431/55 (12.8) | 0.99 (0.72, 1.35) | 0.932 | 307/26 (8.5) | 1.43 (0.92, 2.22) | 0.108 |
| SF-tertiles (adjusted model*) | |||||||||
| Low (T1, < 11 µg/L) | 163/4 (2.5) | 1.16 (0.20, 6.71) | 0.868 | 144/19 (13.2) | 1.09 (0.53, 2.27) | 0.811 | 101/6 (5.9) | 0.60 (0.19, 1.93) | 0.392 |
| Medium (T2, 11-0.17.1 µg/L) | 165/4 (2.4) | 1 Ref. | 143/18 (12.6) | 1 Ref. | 88/9 (10.2) | 1 Ref. | |||
| High (T3, ≥ 17.2 µg/L) | 164/9 (5.5) | 4.78 (1.03, 22.15) | 0.045 | 144/18 (12.5) | 0.97 (0.47, 1.98) | 0.925 | 118/11 (9.3) | 0.95 (0.36, 2.50) | 0.917 |
| Trajectories of SF tertiles from first to third trimester | |||||||||
| Low SF trajectory | 138/4 (2.9) | 3.06 (0.49, 18.89) | 0.229 | 122/19 (15.6) | 1.68 (0.83, 3.41) | 0.149 | 80/4 (5.0) | 0.51 (0.15, 1.74) | 0.283 |
| Medium SF trajectory (adjusted model*) | 229/4 (1.8) | 1 Ref. | 199/21 (10.6) | 1 Ref. | 142/14 (9.9) | 1 Ref. | |||
| High SF trajectory | 125/9 (7.2) | 13.56 (2.27, 81.17) | 0.004 | 110/15 (13.6) | 1.30 (0.63, 2.69) | 0.480 | 85/8 (9.4) | 0.93 (0.36, 2.44) | 0.925 |
Results are from logistic regression analyses, which were run separately for each outcome. Values are odds ratios (ORs) and 95% confidence intervals (CIs)
*Model adjusted for iron supplementation group (stratum 1/40 mg/d (ref.), stratum 1/80 mg/d, stratum 2/20 mg/d, stratum 2/40 mg/d), sex of baby, maternal age (years), maternal early-pregnancy BMI categories (normal weight (ref.), overweight, obesity), family SES (low (ref.), medium, high), smoking (no (ref.), yes), Hb in 1st trimester (g/L, only for SF 1st trimester and trajectories analyses), Hb in 3rd trimester (g/L, only for SF 3rd trimester and trajectories analyses), CRP ≥ 1 (80th percentile) in 1st trimester (no (ref.), yes), carrier of HFE gene mutation (no (ref.), yes), gestational weight gain (kg), and gestational age at birth (weeks, only for low birth weight analyses)
Abbreviations: SF serum ferritin, Ref reference, Head C head circumference
The SF trajectories were calculated from the first to the third trimester, based on SF tertiles at both evaluations. The low SF trajectory included two scenarios: low-low and medium-low tertiles. The high SF trajectory included two scenarios: high-high and medium-high tertiles. The medium SF trajectory was used as the reference category and included all other scenarios: medium-medium, low-medium, high-medium, low-high and high-low trajectory tertiles. The significance of numbers in bold is p-value < 0.05
Table 5 shows that the risk of having a low birth weight (< 2.500 g) was consistent with high SF levels at both T1 and T3, and was markedly increased in the high FS trajectory from T1 to T3 (OR: 13.56 (2.27, 81.17), p = 0.004).
Restricted cubic spline analyses revealed no significant non-linear associations between maternal SF at T1 and T3 and the risk of low fetal growth outcomes at T3 or at birth (< 10th percentile) (all p > 0.05), except for birth weight (Supplementary Figs. 3 and 4); a threshold effect was observed, whereby maternal SF levels above 38 µg/L in T1 (p for non-linearity = 0.008) and 27 µg/L in T3 (p for non-linearity = 0.016), relative to the median SF concentration in each period, were associated with an increased risk of low birth weight (< 2,500 g) (Supplementary Fig. 4), consistent with tertile-based analyses.
Discusion
Our findings indicate that high SF-based iron levels during pregnancy in non-anemic women from a developed country are associated with an increased risk of reduced fetal growth— assessed by ultrasound in the third trimester—and of LBW. Since the effect of iron deficiency anemia on fetal growth had already been widely studied [10], our study specifically focused on assessing the effect of iron status without anemia, and it was observed that although SF levels were low, because Hb production was not affected, fetal growth was not negatively affected. It is worth noting that the main contribution of this study is the evaluation of fetal growth through fetal biometry in each trimester of gestation until the newborn. Identifying critical exposure periods will be very useful, primarily for preventative purposes.
It is important to note that the SF values of our pregnant women (37.5 µg/L and 14.6 µg/L in T1 and T3, respectively) were similar to those previously reported among pregnant women supplemented with internationally accepted doses, where levels fluctuate between 15.5 and 30 µg/L, depending on the trimester of pregnancy, the type of population, among other factors [4]. The decrease observed from early to late gestation, as expected, is mainly due to increased fetal needs and the physiological hemodilution of pregnant women [39]. Similarly, the fetal biometrics and anthropometric measurements at birth in our study agree with the growth observed in others studies [31, 40].
To our knowledge, only a limited number of studies have investigated the association between maternal iron status and fetal growth using ultrasound-based assessments. With regard to the effect of low maternal SF concentrations on fetal growth, Chen et al. [21] conducted a study in a Chinese population including 914 mother–child pairs, in which fetal growth was assessed solely by femoral length measured shortly before delivery. The authors reported that low maternal SF concentrations (< 10th percentile, corresponding to < 15.5 µg/L, compared with the 10th–90th percentiles) were associated with an increased risk of LBW. In contrast, this association was not observed in the retrospective cohort study by Yang et al. [26], nor in our study, when low maternal SF concentrations were evaluated. The study by Yang et al. [26], also conducted in a Chinese population (n = 482), assessed fetal growth using ultrasound measurements of multiple parameters, similarly to our study, including BPD, FL, AC, and EFW during the second and third trimesters. Unlike Chen et al. [21], they observed a positive association between maternal SF concentrations in the first trimester (mean: 27.3 µg/L) and all fetal growth parameters evaluated in both the second and third trimesters. However, no association with fetal growth was observed when maternal SF concentrations were assessed in the third trimester. In our study, the effect of low maternal SF concentrations was evaluated by comparing pregnant women in the lowest SF tertile (< 26 µg/L in the first trimester and < 11 µg/L in the third trimester) with those in the middle tertile. Consistent with the findings reported by Yang et al. [26], SF concentrations in the lowest tertile were associated with higher estimated fetal weight in the third trimester (34.8 g and 38.42 g, based on first- and third-trimester SF concentrations, respectively). Differences between studies may be explained by variations in maternal SF levels and by the inclusion of pregnant women with anemia. In this context, the study by Chen et al. [21], characterized by very low maternal SF concentrations and the inclusion of women with anemia, may have identified adverse effects on fetal growth similar to those previously reported in populations of anemic pregnant women.
Regarding the impact of high maternal SF concentrations, Chen et al. [21] also evaluated this subgroup and reported that excessively high maternal iron levels (> 90th percentile, corresponding to > 343 µg/L, compared with the 10th–90th percentiles) at approximately 10 weeks of gestation were associated with shorter femoral length at 38 weeks. This finding is consistent with our results. In contrast, the study by Yang et al. [26] did not examine a subgroup of pregnant women with elevated SF concentrations, precluding comparison for this exposure category. Overall, further studies are warranted to clarify the potential adverse effects of both low and high maternal SF concentrations on fetal growth.
In the newborn, an association between elevated maternal SF concentrations and reduced fetal growth measured observed in our study has been supported by several other studies. A prospective observational study in a group of 488 Chinese women without anemia (i.e., Hb ≥ 10 g/l at any time during pregnancy) [18] showed significant progressive decrease in the newborn size—as indicated by lower birth weight, CRL, and placental weight—from the lowest quartile (SF ≤ 18 µg/L) to the highest quartile (SF ≥ 44 µg/L) of SF at 28–30 weeks gestation. Similarly, a prospective cohort study conducted in Vietnam, Hanieh et al. [20] found that maternal SF levels at 32 weeks gestation were inversely associated with birth weight. This association was also confirmed when SF data were analyzed in quartiles, with a lower birth weight observed for the SF highest quartile (SF 43–273 µg/L). Another retrospective cohort study conducted in China by Tao et al. [8] with the participation of 3,566 pregnant women found that maternal SF levels, as measured at any time during pregnancy, were linearly associated with the risk of LBW and SGA.
Information about the effect of iron levels across different trimesters of gestation is scarce. In a group of 580 low-income black women in the USA, Goldenberg et al. [41] observed that those with high SF levels at 19-, 26- and 36-weeks gestation (SF means, 95.8 µg/L, 55.4, µg/L and 41.4 µg/L, respectively) had babies with lower birth weight compared to a reference group with low SF levels. Similarly, another study by Rahman et al. [25] who evaluated 573 pregnant women of the MINIMat project in Bangladesh found that, around week 30, pregnant women with high SF levels (median 29 µg/L) had babies that weighed on average of 93 g less than those whose mothers had low SF levels (median 8 µg/L). However, the authors failed to find an association between high SF levels (median 64 µg/L vs. 17 µg/L) at 14 weeks gestation and LBW, unlike our findings, where we observed that per each 1-SD increase in T1 maternal SF levels, the odds of LBW were increased by 82%. Regarding to newborn length, in a prospective study of 797 rural women in India conducted by Rao et al. [22], only FS levels at 28 weeks gestation, but not at 18 weeks gestation was inversely related to birth length. In this regard, a recent review [14] provided some evidence that the detrimental effect on fetal growth was strongest in early pregnancy for low iron and consistent across all trimesters for high iron.
Regardless of the effect caused by iron levels in the different trimesters, the analysis of SF trajectories throughout pregnancy performed in our study indicates that maintaining elevated SF levels during pregnancy increases the risk of fetal growth abnormalities. Interestingly, being in the T3 highest SF tertile increases the risk of shorter femur length by 2.85 times, while a persistently high iron status doubles the risk, reaching 4.26 times. Similarly, in newborns, the highest T3 SF levels were associated with a 4.78-fold increase in the odds of LBW (< 2.500 g), and this risk was markedly tripled, reaching 13.56 times, for a high SF trajectory during pregnancy.
Overall, we found that low maternal iron levels without anemia on fetal growth are mild and not harmful, whereas the detrimental effect of moderately high iron levels during pregnancy—supported by elevated SF trajectories—appears to be consistently associated with reduced fetal growth from the second trimester to birth, also coinciding with many of the previous studies that have analyzed this topic. In our study, restricted cubic spline analyses indicated that maternal SF concentrations above approximately 38 µg/L in the first trimester and 27 µg/L in the third trimester of pregnancy were associated with an increased risk of reduced fetal growth and low birth weight (< 2,500 g). However, further studies are needed to better define the maternal SF thresholds at which iron levels may adversely affect fetal growth.
Therefore, we believe that while the FS upper risk limit is being confirmed, consideration should be given to the need to determine SF, together with Hb, in the follow-up of pregnancy should be considered, in order to prevent the state of anemia and to reconsider iron supplementation during pregnancy when SF levels are moderately high in the pregnant woman.
The pathophysiological mechanisms that would explain our results are that both maternal iron deficiency and excess contribute to fetal oxidative stress through distinct mechanisms. Dilute iron reduces antioxidant capacity and impairs mitochondrial function, while excess iron directly promotes the formation of free radicals and leads to their accumulation in tissues, posing a risk of toxicity. This also triggers a marked inflammatory response and the activation of immune cells, negatively impacting placental function and nutrient transport, ultimately affecting fetal growth [42–44]. Consistent with this evidence, our findings support the hypothesis that, in well-nourished populations without anemia, relatively high ferritin levels could reflect a pro-oxidative and pro-inflammatory intrauterine environment that contributes to impaired fetal growth.
One main strength of this study is its prospective design, which enables a more accurate assessment of the evolution of SF levels throughout pregnancy and its impact on the fetal growth until birth. This data provides additional power beyond what we could muster using birth measurements alone. Including multiple SF measurements during pregnancy also provides a solid basis for establishing temporal and potential causal relationships between iron levels and pregnancy outcomes. Moreover, the size of the study sample, comprising data from 713 mother-child pairs, increases the statistical power and improves generalizability of the findings to the general population. This is especially relevant in perinatal outcome studies, where large samples are needed to detect effects that may be subtle but are clinically relevant. The availability of multiple maternal factors as confounding variables for use in multivariate analyses strengthens the validity of the reported associations, thereby allowing a more detailed and specific analysis of the effects of SF on perinatal outcomes. This study has inherent limitations due to its observational design, which prevents establishing causality and does not rule out residual confounding despite extensive adjustment for relevant factors. However, it is a longitudinal study that allows for the identification of associations. The analysis was restricted to pregnant women without anemia, which improves the interpretation of the effect of elevated ferritin but limits generalizability to populations with a higher prevalence of anemia or malnutrition. Measuring ferritin only in the first and third trimesters prevents a more detailed characterization of iron status in the second trimester. Finally, ultrasound variability, the small size of some subgroups, and differences in genetic factors, lifestyles, and environmental conditions among populations may influence iron metabolism and pregnancy outcomes, potentially affecting the accuracy of the estimates and the extrapolation of the results.
Conclusion
This study shows that in healthy pregnant women without anemia, elevated maternal SF levels at the beginning and end of gestation are associated with a reduction in the fetal growth parameters assessed in T3 (i.e., EFW, femur length, and HC), as well as with an increased risk of LBW. Additionally, a persistently high SF status during pregnancy appears to further aggravate the risks of low fetal femur length and LBW. Whereas fetal growth did not appear to be adversely influenced by low SF levels in absence of anemia. These findings highlighted the importance of monitoring maternal iron status during pregnancy to allow for more personalized iron supplementation.
Supplementary Information
Acknowledgements
We thank to the Jordi Gol Research Institute in Primary Care [Institut d’Investigació en Atenció Primària; IDIAP] for their guidance regarding ethical matters. We thank to the entities and participants in the ECLIPSES study: Research Group in Nutrition and Mental Health (NUTRISAM), Universitat Rovira i Virgili (URV), Reus, Spain (Victoria Arija, Josefa Canals, Lucía Iglesias-Vázquez, Cristina Bedmar, Carmen Hernández-Martínez, Cristina Jardí, Núria Voltas); Sexual and Reproductive Health Care Services (ASSIR) of Tarragona, Spain (Francesc Fargas, Francisca Ruiz, Gemma March, Susana Abajo); the team of midwives recruited for the study (Irene Aguilar, Sònia Aguiles, Rosa Alzúria, Judit Bertrán, Carmen Burgos, Elisabet Bru, Montserrat Carreras, Beatriz Fernández, Carme Fonollosa, María Leiva, Demetria Patricio, Teresa Pinto, María Ramírez, Eusebia Romano, Inés Sombreo); the Central Unit-Barcelona (Josep Basora) of the Institut d’Atenció Primària IDIAP Jordi Gol, Institut Català de la Salut; and the Laboratory of Institut Català de la Salut, University Hospital of Tarragona Joan XXIII, Tarragona, Spain (Núria Serrat). The authors acknowledge the support of the Suport als Grups de Recerca (SGR) programme of the Agència de Gestió d’Ajuts Universitaris i de Recerca (AGAUR), Generalitat de Catalunya, through which the Nutrition and Mental Health research group has been recognized and funded (2021 SGR 00632).
Abbreviations
- ASSIR
Sexual and Reproductive Health Care Services
- BMI
Body Mass Index
- BD
Biparietal Diameter
- CI
Confidence Interval
- CRL
Crown-Rump Length
- EFW
Estimated Fetal Weight
- FL
Femur Length
- GWG
Gestational Weight Gain
- GA
Gestational Age
- Hb
Hemoglobin
- HC
Head Circumference
- ID
Iron Deficiency
- IDA
Iron Deficiency Anemia
- IPAQ
International Physical Activity Questionnaire
- IQR
Interquartile Range
- LBW
Low Birth Weight
- MET
Metabolic Equivalent
- OR
Odds Ratio
- PA
Physical Activity
- RCIU
Restriction of Intrauterine Growth
- rMedDiet
Relative Mediterranean Diet
- SF
Serum Ferritin
- SD
Standard Deviation
- SGA
Small for Gestational Age
- T1
First Trimester
- T2
Second Trimester
- T3
Third Trimester
Authors’ contributions
VA designed, conducted the research and performed data curation. SD-T, AD and VA wrote the article. SD-T, CJ and SA contributed to the collection and management of data. AD analyzed the data. All authors revised the manuscript for important intellectual content and read and approved the final version. The corresponding author attests that all listed authors meet authorship criteria and that no one who meets the criteria has been omitted. VA is the guarantor of this work, as such, she has had full access to all study data and takes responsibility for their integrity and for the accuracy of the data analysis. The data described in the manuscript, the codebook, and the analytical code will be available upon request.
Funding
The ECLIPSES trial was supported financially by grants (PI12/02777, PI17/01754) from the Health Research Fund of the Ministry of Health and Consumption (Madrid, Spain) [Instituto de Salud Carlos III, Fondo de Investigación Sanitaria, Ministerio de Sanidad y Consumo] and by the European Union (ERDF/ESF, “A way to make Europe”/”Investing in your future”). These funding bodies played no part in designing the study, collecting and interpreting the data, or deciding to publish. A.D. is a Serra Hunter Fellow, Spain.
Data availability
The datasets generated and/or analyzed during the current study are not publicly available due to subject confidentiality but are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
The study was approved by the Ethical Committee of Pere Virgili Institute for Health Research and complied with the tenets of the Helsinki declaration. All participants signed an informed consent form.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.The global prevalence of anaemia in 2011 WHO Library Cataloguing-in-Publication Data. 2015. Available from: https://www.who.int. Cited 2024 Feb 15.
- 2.Soppi ET. Iron deficiency without anemia – a clinical challenge. Clin Case Rep. 2018;6(6):1082–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Knovich MA, Storey JA, Coffman LG, Torti SV, Torti FM. Ferritin for the clinician. Blood Rev. 2009;23(3):95–104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Milman N, Taylor CL, Merkel J, Brannon PM. Iron status in pregnant women and women of reproductive age in Europe. Am J Clin Nutr. 2017;106(Suppl 6):S1655–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Aranda N, Viteri FE, Fernández-Ballart J, Murphy M, Arija V. Frequency of the hemochromatosis gene (HFE) 282C–>Y, 63H–>D, and 65S–>C mutations in a general Mediterranean population from Tarragona, Spain. Ann Hematol. 2007;86(1):17–21. [DOI] [PubMed] [Google Scholar]
- 6.The Global Library of Women’s Medicine | GLOWM. Available from: https://www.glowm.com/. Cited 2024 Feb 15.
- 7.Vázquez LI, Arija V, Aranda N, Aparicio E, Serrat N, Fargas F, et al. The effectiveness of different doses of iron supplementation and the prenatal determinants of maternal iron status in pregnant spanish women: ECLIPSES study. Nutrients. 2019;11(10):2418. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Tao Y, Kang J, Liu J, Duan J, Wang F, Shi Y, et al. Association of low birthweight and small for gestational age with maternal ferritin levels: A retrospective cohort study in China. Front Nutr. 2022;9:1002702. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hou J, Cliver SP, Tamura T, Johnston KE, Goldenberg R. Maternal serum ferritin and fetal growth. Obstet Gynecol. 2000;95(3):447–52. [DOI] [PubMed] [Google Scholar]
- 10.Rodriguez-Bernal C, Rebagliato. Ballester. Maternal nutrition and fetal growth: the role of iron status and intake during pregnancy. Nutr Diet Suppl. 2012;4:25–37.
- 11.Hentze MW, Muckenthaler MU, Galy B, Camaschella C. Two to Tango: Regulation of Mammalian Iron Metabolism. Cell. 2010;142(1):24–38. [DOI] [PubMed] [Google Scholar]
- 12.Drakesmith H, Prentice AM. Hepcidin and the iron-infection axis. Science. 2012;338(6108):768–72. [DOI] [PubMed] [Google Scholar]
- 13.Ziaei S, Norrozi M, Faghihzadeh S, Jafarbegloo E. A randomised placebo-controlled trial to determine the effect of iron supplementation on pregnancy outcome in pregnant women with haemoglobin > or = 13.2 g/dl. BJOG. 2007;114(6):684–8. [DOI] [PubMed] [Google Scholar]
- 14.Dewey Dewey KG, Oaks BM. U-shaped curve for risk associated with maternal hemoglobin, iron status, or iron supplementation. Am J Clin Nutr. 2017;106(Suppl 6):1694S. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.de Mendonça ELSS, de Lima Macêna M, Bueno NB, de Oliveira ACM, Mello CS. Premature birth, low birth weight, small for gestational age and chronic non-communicable diseases in adult life: A systematic review with meta-analysis. Early Hum Dev. 2020;149:105154. [DOI] [PubMed] [Google Scholar]
- 16.Ausbeck EB, Allman PH, Szychowski JM, Subramaniam A, Katheria A. Neonatal Outcomes at Extreme Prematurity by Gestational Age Versus Birth Weight in a Contemporary Cohort. Am J Perinatol. 2021;38(9):880–8. [DOI] [PubMed] [Google Scholar]
- 17.Haider BA, Olofin I, Wang M, Spiegelman D, Ezzati M, Fawzi WW. Anaemia, prenatal iron use, and risk of adverse pregnancy outcomes: systematic review and meta-analysis. BMJ. 2013;346(7916):f3443. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Lao TT, Tam KF, Chan LY. Third trimester iron status and pregnancy outcome in non-anaemic women; pregnancy unfavourably affected by maternal iron excess. Hum Reprod. 2000;15(8):1843–8. [DOI] [PubMed] [Google Scholar]
- 19.Akkurt Akkurt MO, Akkurt I, Altay M, Coskun B, Erkaya S, Sezik M. Maternal serum ferritin as a clinical tool at 34–36 weeks’ gestation for distinguishing subgroups of fetal growth restriction. J Matern Fetal Neonatal Med. 2017;30(4):452–6. [DOI] [PubMed] [Google Scholar]
- 20.Hanieh S, Ha TT, De Livera AM, Simpson JA, Thuy TT, Khuong NC, et al. Antenatal and early infant predictors of postnatal growth in rural Vietnam: A prospective cohort study. Arch Dis Child. 2015;100(2):165–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Chen GD, Li PS, Zhou ZX, Wang HY, Gou XY, Ye SX, et al. Associations of maternal serum concentration of iron-related indicators with birth outcomes in Chinese: a pilot prospective cohort study. Ital J Pediatr. 2024;50(1):1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Rao S, Yajnik CS, Kanade A, Fall CHD, Margetts BM, Jackson AA, et al. Intake of micronutrient-rich foods in rural Indian mothers is associated with the size of their babies at birth: Pune Maternal Nutrition Study. J Nutr. 2001;131(4):1217–24. [DOI] [PubMed] [Google Scholar]
- 23.Scholl TO. High third-trimester ferritin concentration: associations with very preterm delivery, infection, and maternal nutritional status. Obstet Gynecol. 1998;92(2):161–6. [DOI] [PubMed] [Google Scholar]
- 24.Alwan N, Hamamy H. Maternal Iron Status in Pregnancy and Long-Term Health Outcomes in the Offspring. J Pediatr Genet. 2015;4(2):111–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Rahman SM, Siraj MS, Islam MR, Rahman A, Ekström EC. Association between maternal plasma ferritin level and infants’ size at birth: a prospective cohort study in rural Bangladesh. Glob Health Action. 2021;14(1):1870421. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Yang J, Chang Q, Du Q, Liu X, Dang S, Tian X. Maternal iron nutrition during pregnancy and fetal intrauterine growth. Nutr J. 2024;23(1):140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Arija V, Fargas F, March G, Abajo S, Basora J, Canals J, et al. Adapting iron dose supplementation in pregnancy for greater effectiveness on mother and child health: Protocol of the ECLIPSES randomized clinical trial. BMC Pregnancy Childbirth. 2014;18(1):33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Hadlock FP, Harrist RB, Sharman RS, Deter RL, Park SK. Estimation of fetal weight with the use of head, body, and femur measurements-A prospective study. Am J Obstet Gynecol. 1985;151(3):333–7. [DOI] [PubMed] [Google Scholar]
- 29.Willet W, Oxford. UK: 2012;50–67. Available online: https://books.google.es/books/about/Nutritional_Epidemiology.html?id=UKs3VaEtNukC&redir_esc=y. Accessed on 13 April 2024.
- 30.Zhang J, Merialdi M, Platt LD, Kramer MS. Defining normal and abnormal fetal growth: promises and challenges. Am J Obstet Gynecol. 2010;202(6):522–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Villar J, Ismail LC, Victora CG, Ohuma EO, Bertino E, Altman DG, et al. International standards for newborn weight, length, and head circumference by gestational age and sex: the Newborn Cross-Sectional Study of the INTERGROWTH-21st Project. Lancet. 2014;384(9946):857–68. [DOI] [PubMed] [Google Scholar]
- 32.Hollingshead A. Four Factor Index of Social Status. Yale J. Sociol. 2011;8:21–52.33.
- 33.De Catalunya G. Classificació catalana d’ocupacions 2011 (CCO-2011). Adaptació de la CNO-2011. Available from: http://idescat.cat/p/cco2011. Cited 2023 Apr 2.
- 34.Heatherton TF, Kozlowski LT, Frecker RC, Fagerstrom K-O. The Fagerström Test for Nicotine Dependence: a revision of the Fagerstrom Tolerance Questionnaire. Br J Addict. 1991;86(9):1119–27. [DOI] [PubMed] [Google Scholar]
- 35.Lee PH, Macfarlane DJ, Lam TH, Stewart SM. Validity of the international physical activity questionnaire short form (IPAQ-SF): A systematic review. Int J Behav Nutr Phys Activity. 2011;8:115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Validación de un cuestionario de frecuencia de consumo alimentario corto. reproducibilidad y validez.]. Available from: https://scielo.isciii.es/scielo.php?script=sci_arttext&pid=S0212-16112008000300011. Cited 2024 Jun 9. [PubMed]
- 37.Trichopoulou A, Costacou T, Bamia C, Trichopoulos D. Adherence to a Mediterranean diet and survival in a Greek population. N Engl J Med. 2003;348(26):2599–608. [DOI] [PubMed] [Google Scholar]
- 38.World Health Organization (WHO). Global Dabatase on Body Mass Index. 2006. Available from: https://www.who.int/data/gho/data/themes/topics/topic-details/GHO/body-mass-index. Cited 2024 Sep 15.
- 39.Sangkhae V, Fisher AL, Ganz T, Nemeth E. Iron Homeostasis During Pregnancy: Maternal, Placental, and Fetal Regulatory Mechanisms. Annu Rev Nutr. 2023;43:279–300. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Papageorghiou AT, Ohuma EO, Altman DG, Todros T, Ismail LC, Lambert A, et al. International standards for fetal growth based on serial ultrasound measurements: The Fetal Growth Longitudinal Study of the INTERGROWTH-21st Project. Lancet. 2014;384(9946):869–79. [DOI] [PubMed] [Google Scholar]
- 41.Goldenberg RL, Tamura T, DuBard M, Johnston KE, Copper RL, Neggers Y. Plasma ferritin and pregnancy outcome. Am J Obstet Gynecol. 1996;175(5):1356–9. [DOI] [PubMed] [Google Scholar]
- 42.Kell DB. Iron behaving badly: inappropriate iron chelation as a major contributor to the aetiology of vascular and other progressive inflammatory and degenerative diseases. BMC Med Genomics. 2009;2:2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Menon R. Oxidative Stress Damage as a Detrimental Factor in Preterm Birth Pathology. Front Immunol. 2014;5:567. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Sultana Z, Qiao Y, Maiti K, Smith R. Involvement of oxidative stress in placental dysfunction, the pathophysiology of fetal death and pregnancy disorders. Reproduction. 2023;166(2):R25–38. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets generated and/or analyzed during the current study are not publicly available due to subject confidentiality but are available from the corresponding author on reasonable request.