A Systematic Review and Meta-Analysis Examining the Risk of Adverse Pregnancy and Neonatal Outcomes in Women with Isolated Hypothyroxinemia in Pregnancy

Abstract

Background:

The relationship between isolated hypothyroxinemia (IH) in pregnancy and adverse pregnancy outcomes is controversial, with no consensus on the need for treatment.

Summary:

We conducted a systematic review and meta-analysis examining adverse pregnancy and neonatal outcomes in women with IH in pregnancy. We searched PubMed, Embase, Web of Science, and the Cochrane Central Register of Controlled Trials for publications from inception to December 2022. Randomized clinical trials and cohort studies were included. Random-effects meta-analyses were used to estimate pooled relative risks (RRs) for each outcome. We included 21 articles, of which 19 investigated the relationship between IH and maternal and neonatal outcomes and 4 investigated the efficacy of levothyroxine (LT4) treatment. Compared with euthyroid pregnancies, IH pregnancies were associated with an increased risk of preterm birth (RR 1.35 [confidence interval, CI, 1.16–1.56]; I ² = 9%), premature rupture of membranes (RR 1.41 [CI 1.08–1.84]; I ² = 0%), gestational diabetes (RR 1.34 [CI 1.07–1.67]; I ² = 76%), macrosomia (RR 1.62 [CI 1.31–2.02]; I ² = 42%), and fetal distress (RR 1.72 [CI 1.15–2.56]; I ² = 0%). However, no statistically significant differences were noted in adverse outcomes according to LT4 treatment status.

Conclusions:

There is evidence suggesting that IH in pregnancy may be associated with an increased risk of adverse pregnancy and neonatal outcomes. However, it is unclear whether LT4 may mitigate the risk of these adverse outcomes.

Introduction

Thyroid hormones are essential for maintaining normal pregnancy and fetal development. Isolated hypothyroxinemia (IH) was defined as a normal thyrotropin (TSH) concentration and low free thyroxine (fT4).¹ IH is relatively common in pregnant women, with a prevalence of 1.3% to 4.3%, and up to 23.9% in iodine-deficient areas.^2,3 Differences in prevalence are related to iodine intake, fT4 diagnostic cutoff point, gestational week of diagnosis, and fT4 testing methods. Iodine deficiency is one of the most common causes of IH.⁴ Environmental pollution (thiocyanate, polychlorinated biphenyls), obesity, advanced maternal age, iron deficiency, vitamin D deficiency, and insulin resistance are also associated with IH.⁵

Numerous studies have reported that hypothyroidism and subclinical hypothyroidism are associated with adverse pregnancy and neonatal outcomes.^6
–8 However, research on IH is more focused on the impact on the cognitive development of offspring,⁹ and it is not clear whether it may be associated with adverse pregnancy outcomes. The Generation R study in the Netherlands showed that IH was associated with an increased risk of preterm birth.¹⁰ Su et al. found that women with IH were at increased risk of fetal distress, small-for-gestational-age infants, and skeletal deformities.¹¹ However, Chen et al. found that there was no significant difference in the incidence of adverse maternal outcomes or perinatal complications between IH patients and pregnant women with normal thyroid function.¹²

In addition, there is uncertainty about whether levothyroxine (LT4) mitigates adverse outcomes in pregnant women with IH. In particular, a large randomized clinical trial (RCT) found that LT4 intervention did not improve adverse pregnancy outcomes.¹³ However, a recent study by Li et al. found that LT4 treatment could reduce the risk of miscarriage and neonatal intensive care unit (NICU) admission among pregnant women with IH.¹⁴ We thus conducted a systematic review summarizing the evidence on the association between IH and adverse pregnancy outcomes as well as the effect of LT4 therapy.

Methods

We performed a systematic review and meta-analysis to assess (1) the association of IH on maternal and neonatal outcomes, and (2) the efficacy of LT4 treatment of IH in preventing maternal and neonatal adverse outcomes. This report follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) standards for reporting of systematic reviews.¹⁵

Study search

We systematically searched the literature published from inception to June 1, 2022, in PubMed, EMBASE, Web of Science, and the Cochrane Central Register of Controlled Trials, with an updated search conducted on December 1, 2022. A librarian with experience in database designed the search strategy used the combination of the following four concept keywords: (1) mothers and pregnancy, (2) hypothyroxinemia, (3) maternal and fetal adverse outcomes, and (4) LT4 therapy. Supplementary Table S1 provides the complete search strategy. In addition, we manually screened the references of related articles and current review articles for potentially relevant articles.

Study selection and data extraction

Studies were independently assessed by two authors (X.G. and X.W.), and any disagreements were resolved by discussion with a third author (Z.S.). This systematic review and meta-analysis included only RCTs and cohort studies comparing IH pregnant women with euthyroid pregnant women or LT4 therapy in IH patients with untreated IH patients. Thyroid function tests were required to be performed in all participants during pregnancy. We included IH patients with normal TSH and decreased fT4.

We examined the following respective adverse pregnancy and neonatal outcomes: preterm birth, placental abruption, premature rupture of membranes (PROM), gestational diabetes, gestational hypertension, preeclampsia, low birth weight (LBW), macrosomia, small for gestational age (SGA), NICU admission, malformation, and fetal distress. Gestational diabetes was defined as fasting plasma glucose ≥5.1 mmol/L or abnormal oral 75 g glucose tolerance test (T1h ≥ 10 mmol/L, T2h ≥ 8.5 mmol/L).¹⁶ Gestational hypertension was defined as blood pressure more than 140/90 mmHg on at least two occasions after 20 weeks of gestation, no history of hypertension before pregnancy.¹⁷ Preeclampsia was diagnosed as gestational hypertension and coexistence of proteinuria. Fetal distress was diagnosed on fetal heart rate variability analysis. Nonhuman studies, reviews, guidelines, meeting abstracts, case reports, populations not meeting inclusion criteria, outcomes that were not predefined in our list, studies lacking necessary information, and non-English studies were excluded.

Two review authors (B.P. and J.L.) extracted data from eligible studies independently. The extracted data included: basic study information (name of first author, year of publication, country, and study design), sample size, gestational age at diagnosis, diagnostic criteria for IH, thyroid function detection assay, study outcomes, gestational age at LT4 initiation and LT4 treatment dose, and prevalence of different types of adverse outcomes in each group.

Study quality assessment

Two authors independently assessed the study quality (C.Z. and B.G.). We used the Newcastle–Ottawa scale¹⁸ for cohort studies (score of 6 or higher considered as high quality) and the Cochrane Collaboration tool¹⁹ for RCTs. The authors' rated the risk of bias of studies as “low risk,” “high risk,” or “unclear risk.”

Statistical analysis

We performed meta-analyses when there were data from two or more studies investigating the same outcomes, with more than 10 patients in the IH group. We performed random-effects meta-analyses using Review Manager (RevMan 5.3; Cochrane Collaboration, Oxford, United Kingdom) to estimate pooled relative risks (RRs) and their associated confidence intervals (CIs) for each outcome. Cochran's Q test (p < 0.05, statistically significant) and I ² tests were used to assess heterogeneity. The I ² values of 25%, 50%, and 75% were considered as having low, moderate, and high heterogeneity, respectively.²⁰ Publication bias was assessed using Egger test plots using Stata 16 (StataCorp, College Station, TX).

To identify major sources of heterogeneity and to assess the robustness of pooled results, prespecified subgroup analyses were performed based on different gestational ages, fT4 cut-point values for diagnosing IH, and the study design.

Results

Study selection

A total of 9582 citations were initially identified. After excluding duplicate entries, 7209 titles and abstracts were screened, of which 58 full-text articles were eligible for evaluation. Twenty-one articles were included in this study. The reasons for exclusion of each study at the full-text review stage are reported in Supplementary Table S2. Among the 21 studies included, 19 studies investigated the effect of IH on maternal and neonatal outcomes^{9,11,12,21

–36} and 4 studies investigated the efficacy of LT4 treatment in preventing maternal and neonatal adverse outcomes.^13,14,29,31 The PRISMA flowchart is shown in Figure 1.

FIG. 1.

PRISMA flow diagram.

Characteristics of the included studies

We included 1 RCT, 16 prospective cohort studies, and 4 retrospective cohort studies, all of which were considered high quality (Supplementary Table S3). The characteristics of the studies are summarized in Table 1 for observational studies and Table 2 for interventional studies. In part I, namely, the analysis of the association of IH on maternal and neonatal outcomes, 19 articles were published from 2007 to 2022 from 10 countries, including a total of 12,129 pregnant women with IH. In part II, namely, the analysis of the impact of LT4 treatment on maternal and neonatal outcomes, 4 articles, 1 from the United States and 3 from China, were published from 2017 to 2022, including a total of 895 pregnant women treated with LT4 and 1247 pregnant women without LT4 treatment.

Table 1.

Study Characteristics of Observational Studies

Author (year)	Country	Study design	Sample size	Gestational age	IH definition	Detection assay (manufacturer)	Outcomes
Casey et al. (2007)²¹	United States	Prospective study	IH: 233 Control: 16,011	<20 Weeks of gestation	TSH 0.08–2.99 mU/L, fT4 < 0.86 ng/dL	Chemiluminescence immunoassays (DPC)	Preterm birth, gestational hypertension, preeclampsia, gestational diabetes, placental abruption
Cleary-Goldman et al. (2008)²²	United States	Prospective study	T1 IH: 232 Control: 10,021 T2 IH: 243 Control: 9981	First and second trimesters	TSH 2.5th–97.5th percentiles, fT4 < 2.5th percentile (0.71 ng/dL)	Chemiluminescence immunoassays (Siemens)	Preterm birth, placental abruption, gestational hypertension, preeclampsia, gestational diabetes, PROM, LBW, macrosomia
Hamm et al. (2009)²³	Canada	Prospective study	IH: 89 Control: 759	Second trimester	TSH 0.15–4.0 mU/L, fT4 < 10th percentile (8.5 pmol/L)	Chemiluminescence technology (Bayer)	Preterm birth, SGA
Männistö et al. (2010)²⁴	Finland	Prospective study	IH: 227 Control: 4719	<20 Weeks of gestation	TSH 5th–95th percentiles, fT4 < 5th percentile (11.96 pmol/L)	Chemiluminescence microparticle immunoassays (Abbott)	Placental abruption, gestational hypertension, preeclampsia, gestational diabetes
Su et al. (2011)¹¹	China	Prospective study	IH: 43 Control: 845	<20 Weeks of gestation	TSH 5th–95th percentiles, fT4 < 5th percentile	Chemiluminescence immunoassays (DiaSorin)	Preterm birth, LBW, macrosomia, SGA, malformation
Breathnach et al. (2013)²⁵	Ireland	Prospective study	IH: 18 Control: 870	<20 Weeks of gestation	TSH 2.5th–97.5th percentiles, fT4 < 2.5th percentile (10.7 pmol/L)	Electrochemiluminescence technology (not mentioned)	Preterm birth, placental abruption, gestational hypertension, gestational diabetes, macrosomia, NICU admission
Korevaar et al. (2013)¹⁰	Netherlands	Prospective study	IH: 145 Control: 4970	<20 Weeks of gestation	TSH 2.5th–97.5th percentiles, fT4 < 2.5th percentile (10.4 pmol/L)	Chemiluminescence assays (Ortho)	Preterm birth, PROM
Medici et al. (2014)²⁶	Netherlands	Prospective study	IH: 129 Control: 4451	<20 Weeks of gestation	TSH 2.5th–97.5th percentiles, fT4 < 2.5th percentile	Chemiluminescence assays (Ortho)	Gestational hypertension, preeclampsia
Furnica et al. (2017)²⁷	Belgium	Retrospective study	IH: 49 Control: 165	First trimester	TSH 0.2–2.5 mU/L, fT4 < 5th percentile (13.0 pmol/L)	Electrochemiluminescence automated immunoassays (Roche)	Preterm birth, LBW, macrosomia, NICU admission, malformation, fetal distress
Rosario et al. (2018)²⁸	Brazil	Prospective study	IH: 54 Control: 542	First trimester	TSH 0.1–2.5 mIU/L, fT4 < 10th percentile (0.92 ng/dL)	Chemiluminescence assay (DPC)	Preterm birth, gestational diabetes, gestational hypertension
Gong et al. (2019)²⁹	China	Prospective study	T1 IH: 106 Control: 2554 T2 IH: 121 Control: 522	First and second trimesters	TSH 2.5th–97.5th percentiles, fT4 < 2.5th percentile (13.35 pmol/L)	Chemiluminescence immunoassay (Roche)	Preterm birth, gestational hypertension, preeclampsia, gestational diabetes, placental abruption, PROM, LBW, macrosomia
Su et al. (2019)³⁰	China	Retrospective study	IH: 342 Control: 7831	<20 Weeks of gestation	TSH 0.06–3.38 mIU/L, fT4 < 2.5th percentile	Chemiluminescence immunoassay (Siemens)	Preterm birth, gestational hypertension, preeclampsia, gestational diabetes, placental abruption, LBW, macrosomia, SGA
Zhou et al. (2019)³¹	China	Prospective study	IH: 379 Control: 513	All trimesters	TSH 0.27–4.2 mIU/L, fT4 < 5th percentile (12.87 pmol/L)	Electrochemiluminescence immunoassays (Roche)	Preterm birth, gestational hypertension, gestational diabetes, placental abruption, PROM, LBW, fetal distress, malformation
Chen et al. (2020)¹²	China	Prospective study	T1 IH: 67 Control: 784 T2 IH: 70 Control: 1943	First and second trimesters	TSH 2.5th–97.5th percentiles, fT4 < 2.5th percentile	Chemiluminescence immunoassays (Beckman)	Preterm birth, gestational hypertension, gestational diabetes, placental abruption, PROM, LBW, fetal distress, malformation
Yang et al. (2020)³²	China	Prospective study	IH: 963 Control: 40,948	First trimester	TSH 2.5th–97.5th percentiles, fT4 < 2.5th percentile (11.7 pmol/L)	Chemiluminescence assays (Abbott)	Preterm birth, gestational hypertension, gestational diabetes
Avramovska et al. (2021)³³	Macedonia	Prospective study	IH: 131 Control: 218	All trimesters	TSH and TT4 were adapted by trimester and gestational weeks at 0.1–3.0/3.7 mIU/L and 65/97.5–165/247.5 nmol/L, respectively	Fluoroimmunoassay method (PerkinElmer)	Preterm birth, LBW
Su et al. (2021)³⁴	China	Retrospective study	IH: 6556 Control: 52,907	All trimesters	TSH normal reference range, fT4 < 10th percentile	Chemiluminescence assays (Siemens)	Gestational hypertension, preeclampsia
Wang et al. (2021)³⁵	China	Retrospective study	IH: 1790 Control: 34,077	<20 weeks of gestation	TSH 5th–95th percentiles, fT4 < 5th percentile	Electrochemiluminescence immunoassay (Roche)	Preeclampsia, gestational diabetes
Nazarpour et al. (2022)³⁶	Iran	Prospective study	IH: 142 Control: 1701	First trimester	TSH normal reference range, fT4 index <10th percentile (2.04)	T4: radioimmunoassay, (Izotop), TSH: immunoradiometric assay	Preterm birth, PROM, LBW, NICU admission

DPC, Diagnostic Products Corporation; fT4, free thyroxine; IH, isolated hypothyroxinemia; LBW, low birth weight; NICU, neonatal intensive care unit; PROM, premature rupture of membranes; SGA, small for gestational age; T1, first trimester; T2, second trimester; TSH, thyrotropin; TT4, total thyroxine.

Table 2.

Study Characteristics of Interventional Trials

Author (year)	Country	Study design	Sample size	Gestational week to start treatment	IH definition	Thyroid test assay (manufacturer)	LT4 initial dose	Outcomes
Casey et al. (2017)¹³	United States	Randomized clinical trial	Treated: 263 Untreated: 261	Mean 18 weeks	TSH normal reference range (0.08 to 3.99 mU/L) and fT4 < 0.86 ng/dL	Chemiluminescence assays (Siemens)	50 μg/day	Preterm birth, placental abruption, gestational hypertension, preeclampsia, gestational diabetes, NICU admission
Gong et al. (2019)²⁹	China	Prospective study	Treated: 95 Untreated: 106	Mean 6–7 weeks	TSH 2.5th–97.5th percentiles, fT4 < 2.5th percentile (13.35 pmol/L)	Chemiluminescence immunoassay (Roche)	50 μg/day	Preterm birth, gestational hypertension, preeclampsia, gestational diabetes, placental abruption, PROM, LBW, macrosomia
Zhou et al. (2019)³¹	China	Prospective study	Treated: 74 Untreated: 379	Within the 2nd and/or 3rd trimester	TSH 0.27–4.2 mIU/L, fT4 < 5th percentile (12.87 pmol/L)	Electrochemiluminescence immunoassays (Roche)	1.5–2.5 μg/kg/day	Preterm birth, gestational hypertension, gestational diabetes, PROM, LBW, macrosomia, fetal distress
Li et al. (2022)¹⁴	China	Prospective study	Treated: 463 Untreated: 501	<13 weeks of gestation	TSH 2.5th–97.5th percentiles, fT4 < 5th percentile	Chemiluminescence assays (Siemens)	25 μg/day	Preterm birth, gestational hypertension, gestational diabetes, placental abruption PROM, macrosomia, NICU admission, fetal distress

LT4, levothyroxine.

The effect of maternal IH on pregnancy outcomes

Figure 2 and Supplementary Figure S1 show the outcome estimates for associations between IH and pregnancy outcomes. Compared with pregnancies of euthyroid women, pregnancies in women with IH were associated with an increased risk of preterm birth (RR 1.35 [CI 1.16–1.56]; I ² = 9%), PROM (RR 1.41 [CI 1.08–1.84]; I ² = 0%), gestational diabetes (RR 1.34 [CI 1.07–1.67]; I ² = 76%), macrosomia (RR 1.62 [CI 1.31–2.02]; I ² = 42%), and fetal distress (RR 1.72 [CI 1.15–2.56]; I ² = 0%). No association was found with placental abruption, gestational hypertension, preeclampsia, LBW, SGA, NICU admission, or malformations.

FIG. 2.

Pooled relative risks of adverse pregnancy outcomes in pregnant women with IH compared with pregnant euthyroid women. CI, confidence interval; IH, isolated hypothyroxinemia; PROM, premature rupture of membranes.

The effect of LT4 intervention on pregnancy outcomes of IH

Table 3 shows the pooled pregnancy and neonatal outcome estimates for women with IH according to LT4 treatment status. There was no significant association with LT4 treatment of IH and the outcomes of preterm birth, placental abruption, PROM, gestational diabetes, gestational hypertension, preeclampsia, LBW, macrosomia, and NICU admission.

Table 3.

Pooled Relative Risk Estimates of Adverse Pregnancy Outcomes in Women with Isolated Hypothyroxinemia According to Levothyroxine Treatment Status

Pregnancy outcome	Pooled RR [CI] in LT4-treated IH	I² (%)	Studies used for meta-analysis
Preterm birth	1.00 [0.45–2.23]	70	^13,14,29,31
Placental abruption	1.21 [0.32–4.53]	6	^13,14,29
PROM	1.88 [0.64–5.52]	88	^14,31
Gestational hypertension	1.05 [0.70–1.58]	6	^13,14,29,31
Preeclampsia	0.77 [0.33–1.77]	0	^13,29
Gestational diabetes	1.25 [0.97–1.62]	0	^13,14,29,31
LBW	0.72 [0.06–8.17]	72	^29,31
Macrosomia	0.97 [0.43–2.20]	67	^14,29
NICU admission	0.84 [0.69–1.01]	0	^13,14

CI, confidence interval; RR, relative risk.

Subgroup and sensitivity analyses

In a prespecified sensitivity analysis, the risk of preterm birth was significantly increased in IH diagnosed in the first trimester (RR 1.30 [CI 1.08–1.58]; I ² = 0%), but there was no significant association observed for IH diagnosed in the second trimester. In IH diagnosed in the second trimester, the risk of gestational diabetes (RR 1.61 [1.19–2.18]; I ² = 5%) and macrosomia (RR 1.73 [1.36–2.20]; I ² = 0%) were significantly increased (Table 4).

Table 4.

Association of Isolated Hypothyroxinemia with Adverse Pregnancy Outcomes According to Gestational Age

	First trimester		Second trimester		<20 Weeks of gestation
Pregnancy outcome	Pooled RR [CI]	I² (%)	Pooled RR [CI]	I² (%)	Pooled RR [CI]	I² (%)
Preterm birth	1.30 [1.08–1.58]	0	1.18 [0.85–1.65]	0	1.38 [0.93–2.04]	39
Placental abruption	0.72 [0.20–2.54]	0	1.70 [0.66–4.36]	0	1.74 [0.88–3.460]	0
PROM	1.00 [0.72–1.39]	0	1.30 [0.43–3.89]	18	1.96 [1.07–3.57]	0
Gestational hypertension	1.16 [0.91–1.49]	0	1.31 [0.53–3.24]	65	0.91 [0.63–1.33]	0
Preeclampsia	1.06 [0.34–3.32]	0	1.02 [0.35–2.95]	0	0.94 [0.59–1.48]	0
Gestational diabetes	1.01 [0.61–1.67]	72	1.61 [1.19–2.18]	5	1.45 [0.63–3.32]	76
LBW	1.09 [0.63–1.90]	23	0.70 [0.37–1.30]	0	0.86 [0.45–1.64]	38
Macrosomia	1.18 [0.46–3.04]	84	1.73 [1.36–2.20]	0	1.48 [0.99–2.21]	59
Malformations	1.98 [0.52–7.61]	0			0.75 [0.24–2.34]	0
NICU admission	1.03 [0.61–1.74]	0
Fetal distress	1.41 [0.50–3.94]	31

According to different fT4 cut-point values used for diagnosing IH, when using the fT4 < 2.5th percentile, gestational IH was associated with preterm birth (RR 1.42 [CI 1.18–1.70]; I ² = 7%), placental abruption (RR 1.92 [CI 1.05–3.51]; I ² = 0), PROM (RR 1.75 [CI 1.15–2.67]; I ² = 0%), gestational diabetes (RR 1.55 [CI 1.16–2.06]; I ² = 76%), and macrosomia (RR 1.74 [CI 1.46–2.06]; I ² = 84%). When using the fT4 < 5th percentile, IH was associated with preterm birth (RR 1.69 [CI 1.18–2.40]; I ² = 0%) and fetal distress (RR 1.84 [CI 1.16–2.91]; I ² = 0%). When using the fT4 < 10th percentile, we analyzed only preterm birth and gestational hypertension and found no significant association (Table 5).

Table 5.

Association of Isolated Hypothyroxinemia with Adverse Pregnancy Outcomes According to Free Thyroxine Cut-Point Values

	fT4 < 2.5th		fT4 < 5th		fT4 < 10th
Pregnancy outcome	Pooled RR [CI]	I² (%)	Pooled RR [CI]	I² (%)	Pooled RR [CI]	I² (%)
Preterm birth	1.42 [1.18–1.70]	7	1.69 [1.18–2.40]	0	1.19 [0.75–1.88]	0
Placental abruption	1.92 [1.05–3.51]	0	1.09 [0.27–4.40]	0
PROM	1.75 [1.15–2.67]	20
Gestational hypertension	1.29 [0.95–1.75]	24	1.95 [0.54–6.97]	78	0.99 [0.89–1.11]	0
Preeclampsia	1.29 [0.67–2.48]	0	1.26 [0.77–2.07]	29
Gestational diabetes	1.55 [1.16–2.06]	76	1.14 [0.91–1.43]	17
LBW	0.99 [0.67–1.47]	0	1.57 [0.93–2.65]	0
Macrosomia	1.74 [1.46–2.06]	84	1.50 [0.37–6.08]	73
Malformations			0.97 [0.55–1.73]	0
Fetal distress			1.84 [1.16–2.91]	0

In prospective studies, there were significant associations between IH and preterm birth (RR 1.36 [CI 1.14–1.61]; I ² = 18%), PROM (RR 1.41 [CI 1.08–1.84]; I ² = 0%), gestational diabetes (RR 1.40 [CI 1.03–1.89]; I ² = 71%), gestational hypertension (RR 1.25 [CI 0.95–1.66]; I ² = 41%), and macrosomia (RR 1.66 [CI 1.19–2.31]; I ² = 52%). In retrospective studies, IH was only associated with macrosomia (RR 1.82 [CI 1.06–3.11]; I ² = 33%) (Supplementary Table S4).

In considering study country, gestational IH was associated with preterm birth (RR 1.28 [CI 1.01–1.63]; I ² = 16%), placental abruption (RR 2.03 [CI 1.06–3.88]; I ² = 0%), and macrosomia (RR 1.82 [CI 1.23–2.70]; I ² = 63%) in studies from outside of China. Gestational IH was associated with preterm birth (RR 1.41 [CI 1.17–1.56]; I ² = 4%), PROM (RR 1.47 [CI 1.04–2.06]; I ² = 0%), gestational diabetes (RR 1.30 [CI 1.04–1.63]; I ² = 80%), gestational hypertension (RR 1.51 [CI 1.01–2.72]; I ² = 76%), and macrosomia (RR 1.56 [CI 1.25–1.94]; I ² = 0%) in studies from China (Supplementary Table S5).

Discussion

In this systematic review and meta-analysis, we observed that maternal IH was significantly associated with an increased risk of premature birth, PROM, gestational diabetes, preeclampsia, macrosomia, and fetal distress. However, LT4 treatment was not associated with a reduction in the risk of preterm birth, placental abruption, PROM, gestational diabetes, gestational hypertension, preeclampsia, LBW, macrosomia, and NICU admission.

Several studies have reported on the relationship between IH and preterm birth, and three existing meta-analyses have had variable findings. A meta-analysis by Korevaar et al. used individual participant data to investigate the relationship between thyroid dysfunction in pregnancy and premature birth.⁸ A total of 904 pregnant women with IH were included with fT4 below the 2.5th percentile as the cutoff point. Korevaar et al. showed that pregnancy IH was related to the risk of premature delivery and early premature delivery.⁸ Similarly, Parizad Nasirkandy et al. reported that the a RR of 1.31 (95% CI: 1.04–1.66) for preterm birth in IH on (using data from 7 studies, including 1078 women with IH and 44,377 control women).³⁷

However, the meta-analysis of Sheehan et al. did not show a significant association between IH and preterm birth.³⁸ Only 5 related studies with a total of 603 pregnant women with IH were included in the meta-analysis by Sheehan et al.³⁸ Our meta-analysis included 3195 pregnant women with IH from 11 studies, so it is the most comprehensive review to date and the findings are consistent with the results of Korevaar et al.⁸ In a subgroup analysis, we observed that IH diagnosed in the first trimester, but not in the second, was associated with an increased risk of preterm birth. Furthermore, IH defined by fT4 levels below the 2.5th–5th percentile was associated with an increased risk of preterm birth. These results suggest that severe IH or IH occurring in the first trimester increases the risk of preterm birth.

In the analysis of IH and gestational diabetes, we included 11 studies and found that the incidence of gestational diabetes mellitus was significantly higher in patients with IH. Luo et al. included nine articles in their meta-analysis examining to analyze the relationship between IH and gestational diabetes mellitus.³⁹ IH was associated with a 1.45-fold increased risk of gestational diabetes, which was similar to our results. Bassols et al. also reported that fT4 was significantly inversely correlated with maternal HbA1c.⁴¹ Gu et al. observed that low levels of fT4 were associated with hyperglycemia and insulin resistance.⁴¹ In our subgroup analysis, only IH diagnosed in the second trimester was associated with an increased incidence of gestational diabetes. This may be due to the progression of glucose intolerance in pregnant women as pregnancy progresses.

Toloza et al. reported that in an individual participant meta-analysis including 933 pregnant women with fT4 below the 2.5th percentile, no significant association was observed between IH and gestational hypertension or preeclampsia.⁴² Similarly, no association between IH and gestational hypertension or preeclampsia was found either in the overall meta-analysis or in a subgroup analysis with fT4 < 2.5th percentile. In our meta-analysis, we included a larger number of patients, but we did not adjust for age, ethnicity, smoking, or other factors, which is a limitation of our analysis.

We observed that pregnant women with IH were at increased risk of macrosomia and fetal distress compared with pregnant women with normal thyroid function. A meta-analysis including data from 929 individuals with IH showed that IH was associated with a lower rate of SGA and higher mean birth weight and that a 1-standard deviation increase in fT4 was associated with a 21-g decrease in birth weight.⁴³ Gestational diabetes and fluctuations in blood glucose associated with IH may contribute to fetal weight gain. Furthermore, many studies have reported that pregnant women with IH have a higher body mass index, which may also play a mediating role between IH and macrosomia.^13,44 The adverse pregnancy outcomes, such as hypertension, placenta previa, and PROM, may also cause fetal distress.

In this study, IH was defined as a normal TSH concentration and low fT4, but there were variable cutoff point for the lower limit of fT4, including the use of the 2.5th/5th/10th percentiles.¹ In sensitivity analyses, the studies using fT4 < 2.5th percentile as the cutoff point were those where IH was associated with preterm birth, placental abruption, PROM, gestational hypertension, and macrosomia. When fT4 < 10th percentile was used as a cutoff, no significant association was observed. This suggests that the diagnostic definition of IH may influence the relationship between IH and adverse outcomes. The authors of a recent meta-analysis also suggest using the 2.5th percentile cutoff for fT4 levels to define IH,⁴⁵ and our findings are confirmatory.

Several studies have shown that iodine status affects pregnancy and neonatal outcomes,⁴⁶ but unfortunately, only four of the studies included in this article provided iodine nutrition levels, and we did not have sufficient information for a formal subgroup analysis. Among our included literature, there were nine studies from China, and they were all of high quality. China has successfully corrected iodine intake deficiencies since the salt iodization policies; iodine nutrition levels are currently appropriate nationwide.⁴⁷

However, considering the potential population differences, we performed subgroup analyses to analyze the results of the studies, according to whether they were performed in China or not. We observed an increased risk of preterm delivery and macrosomia for IH in pregnancy, regardless of whether the studies were performed in China or outside of China. IH was associated with more adverse pregnancy outcomes such as gestational diabetes and gestational hypertension in the studies from China. This suggests that multiple confounding factors associated with IH, such as population differences, iodine status, and management practices of pregnant women, may influence obstetric outcomes.

There is an increased risk of pregnancy complications associated with IH during pregnancy, but there is no consensus on the need for IH treatment in pregnancy. The American Thyroid Association guidelines do not recommend LT4 therapy for women with IH (weak recommendation, low-quality evidence).^1,48 RCTs have not demonstrated the beneficial effect of LT4 on adverse pregnancy outcomes and child cognition. In the evaluation of the effect of LT4 treatment on pregnancy outcomes, Casey et al.¹³ found no improvement in adverse pregnancy outcomes among 263 IH patients who received LT4 and 261 controls without IH who received placebo at a mean gestational age of 17 weeks. However, starting treatment in the second trimester may affect LT4 efficacy on maternal and neonatal outcomes. Li et al. conducted an interventional trial involving 463 IH pregnant women with IH in the first trimester who were treated with LT4 and 501 untreated women and found that LT4 treatment was significantly associated with a reduced the incidence of miscarriage and NICU admission.¹⁴

Zhou et al. initiated isolated IH treatment with LT4 in the second or third trimester and found that treatment was associated with a reduced the risk of preterm birth and LBW infants.³¹ In this systematic review, four studies on LT4 treatment were included, and the meta-analysis did not reveal a beneficial effect of LT4 treatment. A large birth cohort study from Spain suggested that LT4 treatment was safe and was associated with improved pregnancy outcomes of women with an unsatisfactory fT4 level (≤7.5 pg/mL) at before the 9th week of gestation;⁴⁹ however, we did not include this study in our meta-analysis because the authors only compared fT4 ≤ 7.5 pg/mL with those >7.5, without considering TSH level. Future well-designed RCTs with large sample sizes, enrolling women in early pregnancy, are urgently needed to inform whether thyroid function screening and LT4 treatment improve pregnancy and neonatal outcomes in women with IH. Risk factors and fT4 level thresholds should be explored.

To the best of our knowledge, this is the most comprehensive systematic review and meta-analysis to date that quantifies the risk of adverse maternal and neonatal outcomes in pregnant women with IH. We also examined the impact of LT4 therapy on adverse pregnancy and neonatal outcomes, advancing knowledge from previous systematic reviews.⁵⁰ The studies included in this review were considered of high quality. However, we did not have individual participant data, making it impossible to conduct individual participant data meta-analysis. Another limitation is that the definition of IH varied from study to study, thus influencing the heterogeneity of the overall assessment. For this reason, we conducted sensitivity analyses exploring various cutoffs for fT4 values. We were unable to evaluate the influence of other factors, such as differences in baseline indicators, iodine status, and underlying diseases, among the study population in the original literature, which may have explained the heterogeneity. Another limitation is that this systematic review was not registered in advance in the International Prospective Register of Systematic Reviews.

We believe that substantial evidence exists to support an association between IH in pregnancy and a variety of adverse maternal and infant outcomes, but the efficacy of levothyroxine treatment in preventing these adverse outcomes remains uncertain and more research is needed.

Footnotes

Authors' Contributions

Y.H.: Writing—original draft (lead). X.G. and X.W.: Data curation (equal). C.Z. and B.G.: Investigation (equal). B.P. and J.L.: Visualization (equal). A.L.: Conceptualization (lead). Z.S.: Writing—review and editing (lead).

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This study was funded by Chinese National Natural Science Foundation (81700697, 82100831).

Supplementary Material

Supplementary Figure S1

Supplementary Table S1

Supplementary Table S2

Supplementary Table S3

Supplementary Table S4

Supplementary Table S5

References

Alexander

, Pearce

, Brent

, et al. 2017 Guidelines of the American Thyroid Association for the diagnosis and management of thyroid disease during pregnancy and the postpartum. Thyroid, 2017; 27(3):315–389; doi:10.1089/thy.2016.0457

Taylor

, Muller

, Nana

, et al. Indications for treatment of subclinical hypothyroidism and isolated hypothyroxinaemia in pregnancy. Best Pract Res Clin Endocrinol Metab, 2020; 34(4):101436; doi:10.1016/j.beem.2020.101436

Dong

, Stagnaro-Green

. Differences in diagnostic criteria mask the true prevalence of thyroid disease in pregnancy: A systematic review and meta-analysis. Thyroid, 2019; 29(2):278–289; doi:10.1089/thy.2018.0475

Teng

, Shan

, Patil-Sisodia

, et al. Hypothyroidism in pregnancy. Lancet Diabetes Endocrinol, 2013; 1(3):228–237; doi:10.1016/s2213-8587(13)70109-8

Liu

, Li

, Guo

, et al. Association between maternal characteristics and the risk of isolated maternal hypothyroxinemia. Front Endocrinol, 2022; 13:843324; doi:10.3389/fendo.2022.843324

van den Boogaard

, Vissenberg

, Land

, et al. Significance of (sub)clinical thyroid dysfunction and thyroid autoimmunity before conception and in early pregnancy: A systematic review. Hum Reprod Update, 2011; 17(5):605–619; doi:10.1093/humupd/dmr024

Maraka

, Ospina

, O'Keeffe

, et al. Subclinical hypothyroidism in pregnancy: A systematic review and meta-analysis. Thyroid, 2016; 26(4):580–590; doi:10.1089/thy.2015.0418

Korevaar

, Derakhshan

, Taylor

, et al. Association of thyroid function test abnormalities and thyroid autoimmunity with preterm birth: A systematic review and meta-analysis. JAMA, 2019; 322(7):632–641; doi:10.1001/jama.2019.10931

Korevaar

, Tiemeier

, Peeters

RJCe

. Clinical associations of maternal thyroid function with foetal brain development: Epidemiological interpretation and overview of available evidence. Clin Endocrinol, 2018; 89(2):129–138; doi:10.1111/cen.13724

10.

Korevaar

, Schalekamp-Timmermans

, de Rijke

, et al. Hypothyroxinemia and TPO-antibody positivity are risk factors for premature delivery: The generation R study. J Clin Endocrinol Metab, 2013; 98(11):4382–4390; doi:10.1210/jc.2013-2855

11.

, Huang

, Hao

, et al. Maternal thyroid function in the first twenty weeks of pregnancy and subsequent fetal and infant development: A prospective population-based cohort study in China. J Clin Endocrinol Metab, 2011; 96(10):3234–3241; doi:10.1210/jc.2011-0274

12.

Chen

, Yang

, Ye

, et al. Insignificant effect of isolated hypothyroxinemia on pregnancy outcomes during the first and second trimester of pregnancy. Front Endocrinol, 2020; 11:528146; doi:10.3389/fendo.2020.528146

13.

Casey

, Thom

, Peaceman

, et al. Treatment of subclinical hypothyroidism or hypothyroxinemia in pregnancy. N Engl J Med, 2017; 376(9):815–825; doi:10.1056/NEJMoa1606205

14.

, Liu

, Su

, et al. Effect of levothyroxine on pregnancy outcomes in pregnant women with hypothyroxinemia: An interventional study. Front Endocrinol, 2022; 13:874975; doi:10.3389/fendo.2022.874975

15.

Liberati

, Altman

, Tetzlaff

, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: Explanation and elaboration. J Clin Epidemiol, 2009; 62(10):e1–e34; doi:10.1016/j.jclinepi.2009.06.006

16.

Diagnostic criteria and classification of hyperglycaemia first detected in pregnancy: A World Health Organization Guideline. Diabetes Res Clin Pract, 2014; 103(3):341–363; doi:10.1016/j.diabres.2013.10.012

17.

Tranquilli

, Dekker

, Magee

, et al. The classification, diagnosis and management of the hypertensive disorders of pregnancy: A revised statement from the ISSHP. Pregnancy Hypertens, 2014; 4(2):97–104; doi:10.1016/j.preghy.2014.02.001

18.

Stang AJEjoe. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol, 2010; 25(9):603–605; doi:10.1007/s10654-010-9491-z

19.

Higgins

, Altman

, Gøtzsche

, et al. The Cochrane Collaboration's tool for assessing risk of bias in randomised trials. BMJ, 2011; 343:d5928; doi:10.1136/bmj.d5928

20.

Higgins

, Thompson

, Deeks

, et al. Measuring inconsistency in meta-analyses. BMJ, 2003; 327(7414):557–560; doi:10.1136/bmj.327.7414.557

21.

Casey

, Dashe

, Spong

, et al. Perinatal significance of isolated maternal hypothyroxinemia identified in the first half of pregnancy. Obstet Gynecol, 2007; 109(5):1129–1135; doi:10.1097/01.Aog.0000262054.03531.24

22.

Cleary-Goldman

, Malone

, Lambert-Messerlian

, et al. Maternal thyroid hypofunction and pregnancy outcome. Obstet Gynecol, 2008; 112(1):85–92; doi:10.1097/AOG.0b013e3181788dd7

23.

Hamm

, Cherry

, Martin

, et al. The impact of isolated maternal hypothyroxinemia on perinatal morbidity. J Obstet Gynaecol Can, 2009; 31(11):1015–1021; doi:10.1016/s1701-2163(16)34345-6

24.

Männistö

, Vääräsmäki

, Pouta

, et al. Thyroid dysfunction and autoantibodies during pregnancy as predictive factors of pregnancy complications and maternal morbidity in later life. J Clin Endocrinol Metab, 2010; 95(3):1084–1094; doi:10.1210/jc.2009-1904

25.

Breathnach

, Donnelly

, Cooley

, et al. Subclinical hypothyroidism as a risk factor for placental abruption: Evidence from a low-risk primigravid population. Aust N Z J Obstet Gynaecol, 2013; 53(6):553–560; doi:10.1111/ajo.12131

26.

Medici

, Korevaar

, Schalekamp-Timmermans

, et al. Maternal early-pregnancy thyroid function is associated with subsequent hypertensive disorders of pregnancy: The generation R study. J Clin Endocrinol Metab, 2014; 99(12):E2591–E2598; doi:10.1210/jc.2014-1505

27.

Furnica

, Gruson

, Lazarus

, et al. First trimester isolated maternal hypothyroxinaemia: Adverse maternal metabolic profile and impact on the obstetrical outcome. Clin Endocrinol, 2017; 86(4):576–583; doi:10.1111/cen.13301

28.

Rosario

, Oliveira

, Calsolari

MJAoe

, et al. Maternal hypothyroxinemia in the first trimester of gestation and association with obstetric and neonatal outcomes and iron deficiency: A prospective Brazilian study. Arch Endocrinol Metab, 2018; 62(3):332–336; doi:10.20945/2359-3997000000043

29.

Gong

, Liu

, Li

, et al. The impact of isolated maternal hypothyroxinemia during the first and second trimester of gestation on pregnancy outcomes: An intervention and prospective cohort study in China. J Endocrinol Invest, 2019; 42(5):599–607; doi:10.1007/s40618-018-0960-7

30.

, Zhao

, Cao

, et al. Association between isolated hypothyroxinaemia in early pregnancy and perinatal outcomes. Endocr Connect, 2019; 8(4):435–441; doi:10.1530/ec-19-0088

31.

Zhou

, Wang

, Li

, et al. Effects of thyroid diseases on pregnancy outcomes. Exp Ther Med, 2019; 18(3):1807–1815; doi:10.3892/etm.2019.7739

32.

Yang

, Yu

, Zhang

, et al. The association between isolated maternal hypothyroxinemia in early pregnancy and preterm birth. Thyroid, 2020; 30(12):1724–1731; doi:10.1089/thy.2019.0818

33.

Avramovska

, Karanfilski

, Dimitrov

, et al. Isolated maternal hypothyroxinemia and its perinatal outcome in North Macedonia. Acta Clin Croat, 2021; 60(2):246–253; doi:10.20471/acc.2021.60.02.10

34.

, Liu

, Li

, et al. Associations of hypothyroxinemia with risk of preeclampsia-eclampsia and gestational hypertension. Front Endocrinol, 2021; 12:777152; doi:10.3389/fendo.2021.777152

35.

Wang

, Gong

, Peng

, et al. Association of thyroid function during pregnancy with the risk of pre-eclampsia and gestational diabetes mellitus. Endocr Pract, 2021; 27(8):819–825; doi:10.1016/j.eprac.2021.03.014

36.

Nazarpour

, Ramezani Tehrani

, Rahmati

, et al. Effects of isolated maternal hypothyroxinemia on adverse pregnancy outcomes. Arch Gynecol Obstet, 2022; 305(4):903–911; doi:10.1007/s00404-021-06226-0

37.

Parizad Nasirkandy

, Badfar

, Shohani

, et al. The relation of maternal hypothyroidism and hypothyroxinemia during pregnancy on preterm birth: An updated systematic review and meta-analysis. Int J Reprod Biomed, 2017; 15(9):543–552

38.

Sheehan

, Nankervis

, Araujo Júnior

, et al. Maternal thyroid disease and preterm birth: Systematic review and meta-analysis. J Clin Endocrinol Metab, 2015; 100(11):4325–4331; doi:10.1210/jc.2015-3074

39.

Luo

, Wang

, Yuan

, et al. Association of thyroid disorders with gestational diabetes mellitus: A meta-analysis. Endocrine, 2021; 73(3):550–560; doi:10.1007/s12020-021-02712-2

40.

Bassols

, Prats-Puig

, Soriano-Rodríguez

, et al. Lower free thyroxin associates with a less favorable metabolic phenotype in healthy pregnant women. J Clin Endocrinol Metab, 2011; 96(12):3717–3723; doi:10.1210/jc.2011-1784

41.

, Yang

, Gong

, et al. Lower free thyroid hormone levels are associated with high blood glucose and insulin resistance; these normalize with metabolic improvement of type 2 diabetes. J Diabetes, 2021; 13(4):318–329; doi:10.1111/1753-0407.13118

42.

Toloza

, Derakhshan

, Männistö

, et al. Association between maternal thyroid function and risk of gestational hypertension and pre-eclampsia: A systematic review and individual-participant data meta-analysis. Lancet Diabetes Endocrinol, 2022; 10(4):243–252; doi:10.1016/s2213-8587(22)00007-9

43.

Derakhshan

, Peeters

, Taylor

, et al. Association of maternal thyroid function with birthweight: A systematic review and individual-participant data meta-analysis. Lancet Diabetes Endocrinol, 2020; 8(6):501–510; doi:10.1016/s2213-8587(20)30061-9

44.

Han

, Li

, Mao

, et al. High body mass index is an indicator of maternal hypothyroidism, hypothyroxinemia, and thyroid-peroxidase antibody positivity during early pregnancy. Biomed Res Int, 2015; 2015:351831; doi:10.1155/2015/351831

45.

Osinga

JAJ

, Derakhshan

, Palomaki

, et al. TSH and FT4 reference intervals in pregnancy: A systematic review and individual participant data meta-analysis. J Clin Endocrinol Metabol, 2022; 107(10):2925–2933; doi:10.1210/clinem/dgac425

46.

Toloza

FJK

, Motahari

, Maraka

. Consequences of severe iodine deficiency in pregnancy: Evidence in humans. Front Endocrinol, 2020; 11:409; doi:10.3389/fendo.2020.00409

47.

, Teng

, Ba

, et al. Efficacy and safety of long-term universal salt iodization on thyroid disorders: Epidemiological evidence from 31 provinces of Mainland China. Thyroid, 2020; 30(4):568–579; doi:10.1089/thy.2019.0067

48.

Husslein

, Stoeger

. Clinical picture and therapy of Sheehan's syndrome. Wien Klin Wochenschr, 1974; 86(17):489–494

49.

Runkle

, de Miguel

, Barabash

, et al. Early levothyroxine treatment for subclinical hypothyroidism or hypothyroxinemia in pregnancy: The St Carlos gestational and thyroid protocol. Front Endocrinol, 2021; 12:743057; doi:10.3389/fendo.2021.743057

50.

Ramezani Tehrani

, Nazarpour

, Behboudi-Gandevani

. Isolated maternal hypothyroxinemia and adverse pregnancy outcomes: A systematic review. J Gynecol Obstet Hum Reprod, 2021; 50(7):102057; doi:10.1016/j.jogoh.2020.102057