Introduction

Iodine, an essential trace element for synthesis of thyroid hormones, plays a vital role in early growth and development of most organs, especially in the central nervous and reproductive systems [1,2,3]. A active transportation of iodine from the blood into the thyroid is regulated by the thyroid-stimulating hormone (TSH) from the pituitary gland and by the concentration of iodine in the blood [4]. Iodine status may influence growth through its effects on the thyroid axis [5,6,7], and both iodine deficiency and its excess affect thyroid hormones. Sufficient maternal iodine is required for the production of thyroid hormones for the normal development of the fetus and the neonate [8].

The need for iodine in pregnancy is increased [9] due to an increase in maternal thyroxin production to maintain maternal euthyroidism and for transfer of thyroid hormones to the fetus in early pregnancy, before the fetal thyroid begins functioning [10]. Also, maternal glomerular filtration rate (GFR) is increased during pregnancy resulting in increased renal loss of ingested iodine [11]. Adequate iodine consumption is essential for production of thyroid hormones and prevention of any possible feto-maternal complications [1, 12]. Women with iodine deficiency may start pregnancy with inadequate intrathyroidal iodine stores that cannot meet the increasing demands of pregnancy [13,14,15], and this unmet need for iodine leads to pathological changes that may adversely affect maternal and fetal health [16]. Even mild-to-moderate iodine deficiency during pregnancy can cause maternal and fetal hypothyroidism, resulting in adverse feto-maternal and neonatal effects [4, 16,17,18]. Severe iodine deficiency in pregnant women has been associated with serious adverse health effects, including cretinism and growth retardation and impaired neurological development of the fetus and increased rates of pregnancy loss, stillbirth, and perinatal, neonatal, and infant mortality [14, 19, 20].

Iodine deficiency is one of the most prevalent disorders, especially during pregnancy [17, 18]. Severe maternal iodine deficiency may be associated with both maternal and fetal hypothyroidism and with adverse pregnancy outcomes, including spontaneous abortion, congenital anomalies, growth retardation, stillbirth and perinatal mortality, recurrent miscarriage, preterm delivery, and perinatal and infant mortality; significant improvements however have been documented observed with iodine supplementation [16,17,18, 21,22,23,24]. Potential adverse effects of severe iodine deficiency in pregnancy are believed to be due to alteration of maternal thyroid function and possibly fetal thyroid function as well, since iodine is an essential component of thyroid hormones [19, 25].

Following implementation of universal salt iodization in 1989, the Islamic Republic of Iran was declared to be a country free of iodine deficiency disorders (IDD) [26]. Despite consumption of iodized salt in Iran, studies conducted in 2013 showed inadequate iodine nutrition both in pregnant and in non-pregnant women in Tehran [27, 28]. Given the lack of sufficient data regarding the impact of mild-to-moderate iodine deficiency on pregnancy outcomes, we aimed to investigate the relationship of maternal iodine and TSH status with pregnancy outcomes in these women.

Materials and Methods

Study Design and Participants

We used data from the Tehran Thyroid and Pregnancy Study, a two-phase population-based study, conducted from September 2013 through February 2016. Details of the study protocol have previously been published [29]. Using a cluster sampling method, a total of 1671 pregnant women in their first trimester were selected (calculated by the date of last menstrual period or sonography), from among those receiving prenatal care in centers under coverage of Shahid Beheshti University of Medical Sciences, which provides health services to over two-thirds of Tehran’s population.

After obtaining a written informed consent from participants, a comprehensive questionnaire including demographic, reproductive, medical, and prenatal histories was completed during face-to-face interviews. Fasting blood samples were collected for thyroid hormonal assessment including thyroxine (T4), Resin T-uptake (RTU), thyrotropin (TSH), and thyroid peroxidase antibody (TPOAb) to determine the thyroid status of participants. Two additional fasting blood samples were collected in the second (20–24 weeks of gestation) and third (30–34 weeks of gestation) trimesters. Following centrifugation, samples were stored at − 80 °C till the end of the study to measure serum levels of TSH, T4, and RTU.

Women with subclinical hypothyroidism (TPO-Ab+ or TPO-Ab−) were invited for the second phase of the study and randomly assigned into two groups (treated with levothyroxine (LT4) and without treatment). Euthyroid TPOAb− women served as the healthy controls and were followed till delivery. Pregnancy outcomes in terms of height, weight, and head circumference of newborns, neonatal TSH levels and preterm delivery, placenta abruption, stillbirth, and neonatal admission were compared between groups.

For the purpose of the present study, pregnant women with twin pregnancies (n = 21), those with overt thyroid dysfunction (hyperthyroidism or hypothyroidism) (n = 78), and those who had used levothyroxine during pregnancy (previous consumers or those selected as interventional group in the second phase of the study) or those using iodine-containing supplement (n = 286) were excluded; finally, a total of 1286 pregnant women were included in the present study; none of whom consumed iodine-containing supplement during pregnancy.

Since the World Health Organization (WHO) recommends analysis of urinary iodine concentration (UIC) on spot urine samples for assessment of iodine status of a population [23], at the first prenatal visit, participants were asked to collect three casual morning urine samples (5–10 mL) on an every other day basis for measurement of UIC. Spot urine samples were collected and stored at − 20 °C. The levels of serum hormones and urinary iodine were measured at the Research Institute of Endocrine Sciences of the Shahid Beheshti University of Medical Sciences. Urinary iodine was measured in three urine samples to obtain the median. Pregnant women were divided into four groups according to urine iodine status, and pregnancy outcomes were compared between groups.

The study was approved by the ethics committee of the Research Institute of Endocrine Sciences (RIES), approval no: IR.SBMU.ENDOCRINE.REC.1397.273.

Outcomes

In this study, the primary outcome was preterm delivery and secondary outcomes were height, weight, and head circumference of newborns, neonatal TSH levels, placental abruption, stillbirth, and neonatal admission.

Laboratory Determination

T4 was measured by radioimmunoassay (RIA) using commercial kits (Izotop Kit, Budapest Co, Hungary) and TSH was measured by the immunoradiometric assay (IRMA) using the gamma counter (Dream Gamma-10, Goyang-si, Gyeonggi-do, South Korea). TPOAb and RTU were measured by the immunoenzymometric assay (IEMA) (Monobind Kit, Costa Mesa, CA, USA) and enzyme immunoassay (EIA) (DiaPlus Kit, San Francisco, CA, USA), respectively, using a calibrated ELISA reader (Sunrise, Tecan Co., Salzburg, Austria). Intra- and inter-assay coefficients of variation (CVs) for T4, TSH, RTU, and TPOAb were 1.1% and 3.9%, 1.9% and 4.7%, 2.2% and 4.3%, and 1.0% and 1.6%, respectively. Since free T4 immunoassays may be affected by changes of serum thyroxine-binding globulin and albumin during pregnancy, Free Thyroxine Index (FT4I) was used [30].

UIC was determined using a manual method, based on the Sandell–Kolthoff technique [23]. The intra-assay CVs in three ranges of 3.4, 12.5, and 37.1 μg/L were 8.5, 7.2, and 9.6%, respectively, and inter-assay CV percentages were 9.1, 8.6, and 12.3%, respectively.

Definition of Terms

Overt hyperthyroidism was defined as TSH levels < 0.1 μIU/mL and FT4I > 4.5. Overt hypothyroidism was defined as TSH > 10 μIU/mL or TSH levels > 2.5 μIU/mL and FT4I < 1.

Subclinical hypothyroidism was defined by a normal FT4I (1–4.5), despite elevated TSH (4–10 mIU/L). TPOAb > 50 IU/mL was considered to be TPOAb positivity.

Group 1 included pregnant women with UIC < 100 μg/L; group 2, UIC 100–150 μg/L; group 3, UIC 150–250 μg/L; and group 4, UIC ≥ 250 μg/L.

Preterm delivery was defined as babies born alive before 37 weeks of pregnancy [31]; placental abruption was defined as the placenta separation from the wall of the uterus, prior to the birth of the baby, and was diagnosed by clinical or histological findings [32]. Fetal death or stillbirth was defined as a baby born with no signs of life after a given threshold (20 weeks gestation) [33]; newborn admission was defined as the admission of a newborn to the neonatal unit, mainly due to icterus or fetal distress.

Sample Size

The sample size is calculated using the following formula:\( {n}_i=2{\left(\frac{z_{1-\frac{\alpha }{2}}+{z}_{1-\beta }}{Es}\right)}^2 \); α = 0.05 is assumed and the test power is considered to be 1 − β = 0.8; the effect size is calculated as follows: \( Es=\frac{\mid {P}_1-{P}_2\mid }{\sqrt{P\left(1-P\right)}} \) and \( P=\frac{P_1+{P}_2}{2} \). P1 and P2 are the proportion of individuals in each group that is expected to produce the desired outcome, and P is the overall proportion.

In this study, a 40% increase in the outcome ratio of iodine deficient individuals is clinically meaningful; hence, with a sample size of 543 people in each group and with a power of 80%, this difference can be identified in groups (mothers with different iodine statuses).

$$ Es=\frac{\mid 0.12-0.07\mid }{\sqrt{0.095\times 0.905}}=0.17\kern2.25em {n}_i=2{\left(\frac{1.96+0.84}{0.17}\right)}^2=543 $$

Statistical Analysis

Post hoc power calculation indicated that with at least a 240-sample size for each subgroup of UIC, this study had a 96% power to determine a 1.5% difference in the prevalence of our primary outcome (preterm labor) among pregnant women, of the four UIC categories.

Normality assumption was checked via the Kolmogorov–Smirnov and Shapiro–Wilk tests. Continuously distributed variables are reported as mean ± standard deviation for the normal and median (IQR) for the non-normal variables. ANOVA and Kruskal–Wallis tests were applied to detect significant differences in normally and non-normally distributed variables. Post hoc analysis was applied to make pairwise comparisons using the Dunn–Bonferroni and Tukey tests. Categorical variables are reported as N% of positive events and were tested using the chi-squared test.

Generalized linear regression model (GLM) via logit and linear link function for binary outcomes (preterm delivery and newborn admission) and continuous outcomes (head circumference, weight, and height) were applied to estimate odds ratio and mean difference in subgroups of UIC levels, respectively.

TSH-stratified multivariate-generalized estimating equation (GEE) method [34] was adjusted for age, body mass index (BMI), and TPOAb+ status, with linear link function and exchangeable working correlation; the matrix method was employed to estimate the overall effect of UIC on FTI and T4 and their trend over the pregnancy period. Statistical analyses were performed using SPSS version 19.0 (SPSS, Inc., Chicago, IL, USA).

Results

In the present study, median (interquartile range) UIC in a total of 1286 participants was 142.3 (133.6) μg/L; results showed 370 (28.8%), 315 (24.5%), 359 (27.9%), and 242 (18.8%) of participants had UIC < 100, 100–150, 150–250, and ≥ 250 μg/L, respectively. In these pregnant women, 6.6% (n = 85) and 4.8% (n = 62) had TSH ≥ 4.0 and TPO > 50, respectively. Ranges of neonatal TSH concentrations, head circumference, weight, and height of the newborns were 0.1–11.5 μIU/mL, 27–43 cm, 1470–4700 g, and 41–62 cm, respectively. Table 1 shows the characteristics of the study population based on their urinary iodine subgroups; we found no significant differences in maternal age among different subgroups of UIC, whereas maternal weight and BMI were significantly different between these subgroups (p value = 0.037 and 0.013, respectively). Preterm delivery in group 4 (UIC ≥ 250 μg/L) was significantly lower than in group 1 (UIC < 100) (8 [4%] vs. 29 [9%], p < 0.05), and neonatal admission in group 4 was significantly lower than in groups 1, 2, and 3 (6 [3%] vs. 22 [7%], 30 [12%], 28 [11%]; p = 0.005, respectively). There were no significant differences in the rates of placental abruption and stillbirth between the four study groups.

Table 1 Characteristics of the study population according to their first-trimester urinary iodine
Full size table

Table 2 presents the results of age, BMI, and TPOAb+ adjusted GLM analysis for pregnancy outcomes in the study groups based on the TSH cutoff value of 4.0 μIU/mL and different urinary iodine concentrations. Regardless of TSH levels, the results of GLM showed a higher odds ratio for preterm delivery in group 1 versus group 4 (OR 2.4; 95% CI 1.1, 5.0; p = 0.035); the odds ratio of preterm delivery in the subgroup of women with UIC < 100 μg/L and TSH ≥ 4 μIU/mL was 2.5 times higher than of those with similar urinary iodine but TSH < 4 μIU/mL (OR 2.5; 95% CI 1.1, 10.0; p = 0.024).

Table 2 Generalized linear regression model for pregnancy outcomes in study subgroups based on the first-trimester TSH cutoff value of 4.0 μIU/mL adjusted for age, body mass index, and TPO+ status
Full size table

Table 3 presents the results of GEE analysis for FTI and T4, considering cutoff values of 4.0 μIU/mL for TSH and different subgroups of UIC, after the adjustment for age, BMI, and TPOAb+ status. Regardless of TSH level, the interaction of UIC and gestational trimester revealed that in comparison with those with UIC 150–250 μg/L, FTI was decreased by 0.1 unit (95% CI − 0.2, − 0.004; p = 0.041). This decreasing trend for UIC < 100 μg/L decelerated in pregnant women with TSH < 4 μIU/mL by − 0.08 unit (95% CI − 0.183, 0.015; p = 0.097) and accelerated by − 0.49 (95% CI − 0.818, − 0.158; p = 0.004) in pregnant women with TSH ≥ 4 μIU/mL. Trends of FTI and T4 over gestational time in different subgroups of UIC based on a TSH cutoff value of 4.0 μIU/mL are presented in Figs. 1, 2, 3, and 4, respectively.

Table 3 Generalized estimating equation analysis for thyroid hormones in study subgroups according to the TSH cutoff value of 4.0 μIU/mL adjusted for age, body mass index, and TPOAb+
Full size table
Fig. 1
figure 1

Overall trends of FTI in categories of UIC over gestational time

Full size image
Fig. 2
figure 2

Trends of FTI in categories of UIC over gestational time by TSH cutoff values of 4.0 μIU/mL subgroups

Full size image
Fig. 3
figure 3

Overall trends of T4 in categories of UIC over gestational time

Full size image
Fig. 4
figure 4

Trends of T4 in categories of UIC over gestational time by TSH cutoff values of 4.0 μIU/mL subgroups

Full size image

Discussion

In this study conducted in an area with iodine sufficiency, despite the consumption of iodized salt, median UIC of 142.3 mIU/L, results showed mild iodine deficiency among our pregnant cohort, considering a WHO cutoff value of 150 mIU/L for the pregnancy period. We found that adverse pregnancy outcomes in terms of preterm delivery increased 2.5-fold (95% CI 1.1, 10.0; p = 0.024) in women with UIC < 100 μg/L and TSH ≥ 4 μIU/mL, compared with those with similar iodine concentrations and TSH < 4 μIU/mL; studies demonstrate that excessive iodine intake, especially in pregnant women with positive TPOAb, can induce a feedback loop to increase TSH levels and directly increase TSH levels by the action of the pituitary gland and the hypothalamus [35].

The present study showed an increased odds ratio of preterm delivery in pregnant women with deficiency of iodine (UIC < 100), in particular in those with TSH ≥ 4, compared with pregnant women with UIC ≥ 250, a finding consistent with the results of previous studies which show that adverse effects of iodine deficiency on pregnancy outcomes are higher in women with subclinical hypothyroidism [36, 37]. Charoenratana et al. demonstrated that in a severe iodine deficiency area, iodine status was an independent risk factor of preterm birth [38], and the Chaouki et al.’s study from Algeria showed that rates of abortion, stillbirth, and prematurity were significantly lower among women given oral iodized oil 1–3 months before conception or during pregnancy, than among untreated women in a severe iodine deficiency area [21]. On the contrary, Leon et al. (2015) reported UICs were not significantly related to preterm delivery in a mildly iodine deficient region [39].

Studies have documented associations between iodine deficiency, subclinical hypothyroidism, and adverse obstetric outcomes such as placental abruption and stillbirth, although the mechanisms behind these associations are unclear [40, 41]; our study did not show significant differences in the rates of these outcomes among different groups of iodine concentration. It has been suggested that a disturbance in thyroid hormones, viz. hypothyroidism, can lead to impaired endothelial function, increased arterial intimal media thickness, and insulin resistance, processes which may all result in placental insufficiency. Maternal thyroid status is important for trophoblastic function and to maintain pregnancy [40, 42]. However, in agreement with our study, Torlinska et al. [25] and Zhou et al. [43] also reported that maternal iodine status was not associated with adverse pregnancy outcomes in a mild-to-moderate iodine-deficient pregnant population. In contrast, Dillon et al. (2000) showed an increased risk of repeated miscarriages and stillbirth, which were related to severe iodine deficiency [41].

Differences in these results could be due to the method of assessing iodine status. Whereas we used three urine samples on different days of the week, Dillon et al. [41] and Leon et al. [39] assessed only one urine sample to determine iodine status in pregnant women. Considering the day-to-day and within-day variations in urinary iodine excretion, this method may not be reliable for determining iodine status. Even three urine samples, as we used, may not precisely define the individual iodine status, as demonstrated by König F et al., a reliable estimation of individual iodine status requires at least 10 urinary spot samples or 24-h samples [44]. These contradictory results may partly be explained by the timeline of recruitment. More than half of women in the present study were recruited after 8 weeks of gestation, and majority of miscarriages occur before 8 weeks of gestation. It is therefore possible that the study may have inadvertently excluded women with adverse pregnancy outcomes such as early miscarriages, and therefore, this data was not collected.

This study showed that regardless of TSH level, neonatal admission in pregnant women with UIC ≥ 250 μg/L was significantly lower than in those with UIC < 250 μg/L. Since most of the neonatal admissions can be due to prematurity, this finding seems logical. Similarly, Ozdemir et al. found that babies of pregnant women with hypothyroidism needed more frequent NICU admission, mainly due to preterm delivery [45]. In contrast, a randomized controlled trial conducted by Zhou et al. [43] did not detect any significant difference in the rate of NICU admissions between groups treated with 150 μg/L iodine supplement or placebo; their results however need to be interpreted with caution due to lack of adequate sample size. Another randomized, placebo-controlled trial conducted by Gowachirapant et al. [46] did not also detect any significant differences between the groups (receiving 200 μg iodine orally once a day or placebo until delivery) in the rate of serious adverse events in terms of death/hospital admission of either mother or baby for a cause other than delivery. Nevertheless, adequately powered randomized controlled trials with separate subgroups of iodine concentration are needed to provide conclusive evidence regarding the effect of iodine supplementation during pregnancy on neonatal admission rates.

The changing trend of FTI during pregnancy in the present study is in agreement with that reported by Azizi et al., markedly increasing in the first trimester, reaching a peak in the second trimester, and then falling in the third trimester [30]. Soldin et al. observed that in a region of iodine sufficiency, FT4 concentrations were significantly associated with a decline in the third trimester of pregnancy [47]. Khalil et al. also reported TSH values were lowest in the first-trimester group and increased in the second- and third-trimester groups, while FT4 levels showed the opposite trend [48], a finding similar to that of the current study. As pregnancy progressed, a decreasing trend of FTI was detected in women with UIC > 100 μg/L, compared with those with UIC < 100 μg/L per gestational trimester; although this negative trend was observed in both groups of pregnant women, regardless of baseline TSH cutoff value of 4 μIU/mL (− 0.1), it was more prominent in those with TSH ≥ 4 μIU/mL than in women with TSH < 4 μIU/mL (− 0.49 vs. − 0.08). These findings demonstrate the critical role of iodine deficiency in subclinical hypothyroidism in terms of decrease of FTI overtime throughout gestation.

According to WHO recommendations, the 2017 guidelines of the American Thyroid Association (ATA), and the European Thyroid Association (ETA), all pregnant women should ingest approximately 250 μg iodine daily, i.e., about 100 μg above their non-pregnant status. To ensure a daily consumption of 250 μg iodine, strategies may vary based on country of origin and iodine sufficiency levels [49]. Universal salt iodization is one of the most cost-effective ways of delivering iodine and improving maternal and infant health [20, 49]. In Iran, despite concerted attempts to ensure appropriate intake of adequate iodine in this population using strategies such as the universal iodization of salt and nutrition education programs, studies conducted on pregnant women shown that due to the increased need for iodine during pregnancy, urinary iodine concentration is below the recommended WHO level, demonstrating the need for increased iodine intake during pregnancy [50]. Similar to previous studies from Iran [28, 50, 51], in the current study, approximately a third (28.8%) of the pregnant population had urinary iodine concentrations < 100 μg/L (moderate-to-severe iodine deficiency), indicating that intakes of iodized salt in these women were inadequate and did not meet the iodine requirements of pregnancy.

The main strength of this study is its methodology, as a population-based study conducted mainly on first-trimester pregnant women with repeated assessments of thyroid status throughout the pregnancy period. Since in individuals, urinary iodine excretion can vary somewhat from day to day and even within a given day, median urinary iodine is the main indicator used to assess iodine status of a population [23]. In this study, collecting of urine samples on 3 different days of the week may partially remove the bias that may otherwise have occurred, in the precise evaluation of iodine status (due to day-to-day variations in urinary iodine excretion).

However, our study also had some limitations. First, the number of samples was insufficient to examine other rare pregnancy complications, e.g., preeclampsia and stillbirth. Second, not observing significant differences on miscarriage rates may be due to the non-timely recruitment of the study participants, and hence needs to be interpreted with caution. Third, lack of knowledge regarding some other risk factors of pregnancy outcomes could influence the results of this study. Fourth, changes in iodine status throughout the pregnancy cannot be assessed by use of single baseline assessment. Nutritional habits throughout pregnancy have not been evaluated, and their effects were not modified in analysis or interpretation of the results.

In conclusion, despite implementation of iodized salt strategies in Iran, mild iodine deficiency among our pregnant cohort indicates that iodine supplementation needs to be implemented for pregnant women. Considering the adverse pregnancy outcomes due to iodine deficiency, especially in women with TSH ≥ 4 μIU/mL, monitoring of urinary iodine concentration at the population level and iodine replacement may be needed during pregnancy, even in areas of iodine sufficiency.