Associations between Human Chorionic Gonadotropin,Maternal Free Thyroxine,and Gestational Diabetes Mellitus

Abstract

Background:

Human chorionic gonadotropin (hCG) is a marker of placental function, which also stimulates the maternal thyroid gland. Maternal thyroid function can be associated with the pathophysiology of gestational diabetes mellitus (GDM). We aimed to study whether there is an association of hCG concentrations in early pregnancy with GDM and whether it is mediated through maternal thyroid hormones.

Methods:

This study included 18,683 pregnant women presenting at a tertiary hospital in Shanghai, China, between January 2015 and December 2016. GDM was diagnosed using a 2-hour, 75-g, oral glucose tolerance test (OGTT) according to the American Diabetes Association guidelines. Multivariable logistic or linear regression models were used to identify associations, adjusting for maternal age, education level, family history of diabetes, parity, fetal sex, thyroperoxidase antibody (TPOAb) status, and prepregnancy body–mass index.

Results:

Higher hCG concentrations were associated with a lower plasma glucose level during the OGTT, but not with fasting plasma glucose or hemoglobin A1c concentrations tested during early pregnancy. hCG in early pregnancy was negatively associated with GDM risk (p = 0.027). Mediation analysis identified that an estimated 21.4% of the association of hCG-associated GDM risk was mediated through changes in free thyroxine (fT4) concentrations (p < 0.05). In the sensitivity analysis restricted to TPOAb-positive women, hCG was not associated with GDM (p = 0.452).

Conclusions:

Higher hCG levels in early pregnancy are associated with a lower risk of GDM. Maternal fT4 may act as an important mediator in this association.

Introduction

Hyperglycemia resulting from glucose intolerance during pregnancy, known as gestational diabetes mellitus (GDM), is a common complication during pregnancy and has a major impact on public health worldwide (1). GDM is estimated to affect approximately one in six live births, with a prevalence of around 14.8% reported in mainland China (2,3). GDM is associated with a higher risk of adverse pregnancy as well as maternal and offspring outcomes such as preeclampsia, macrosomia, neonatal hypoglycemia, and a profound maternal and offspring predisposition to diabetes mellitus (4 –6). Typically, GDM is diagnosed at 24–28 weeks of gestation (7), but efforts for early identification of women who are at high risk of developing GDM can enable lifestyle and/or medical interventions, which may prevent the occurrence of GDM and its morbidities (8). For this purpose, researchers have been exploring routine first trimester biochemical markers, which may help identify pregnant women at risk of subsequent GDM (9,10).

Human chorionic gonadotropin (hCG) is routinely used for first trimester Down's syndrome screening programs worldwide (11,12). hCG could be a determinant of GDM risk as it has been shown to stimulate placental glycogenolysis through activation of adenylate cyclase and its immunomodulatory effects could ameliorate pancreatic autoimmunity (13,14). Furthermore, a potential path through which hCG could be associated with GDM is through the thyroid hormone. hCG has weak affinity for the thyrotropin (TSH) receptor and stimulates the thyroid gland, increasing serum free thyroxine (fT4) (15). The thyroid hormone is a regulator of insulin secretion and glucose homeostasis and may contribute to protecting and improving pancreatic β-cell function (16 –19). Decreased thyroid function has been found to be associated with a higher risk of GDM (20).

Previous studies have shown that mothers diagnosed with GDM have a lower hCG concentration in the first trimester (21 –23). It has been suggested that higher hCG concentrations during this period may reduce the risk of GDM (24,25). However, not all studies to date have replicated this (26,27), potentially due to different ethnic composition of patient cohorts, relatively small study sizes, and/or heterogeneous diagnostic criteria for GDM.

The aim of the current study was to confirm the association of hCG in early pregnancy with GDM and to examine whether it is mediated through changes in maternal thyroid function.

Materials and Methods

Patient enrollment

Participants were consecutively enrolled as part of a large, prospective, hospital-based cohort study at the International Peace Maternity and Child Health Hospital (IPMCH), a tertiary care hospital in Shanghai, China. Eligible participants were pregnant women who attended the first trimester antenatal visit at IPMCH between January 2015 and December 2016 and underwent hCG measurements. Of those eligible, we included women with available records of hCG, fT4, TSH, and thyroperoxidase antibody (TPOAb) in the first trimester and a completed 2-hour, 75-g, oral glucose tolerance test (OGTT) during 24–28 gestational weeks. Further exclusion criteria were pre-existing diabetes, in vitro fertilization, twin pregnancy, thyroid function altering medical treatment (levothyroxine, propylthiouracil, and methimazole, etc.) before or after baseline measurements, or a history of thyroid disease. Ethical approval was obtained from the IPMCH Ethics Committee (reference GKLW2012-49). Written informed consent was obtained from all study participants.

Data collection

Data were collected prospectively by medical staff, including nurses, residents, and gynecologists, during hospital visits. At the first clinical visit (9–13 weeks of pregnancy), maternal age, parity, last menstrual period (LMP), education level, and medical history data were collected through face-to-face interview. Body–mass index (BMI) before pregnancy was calculated using nurse-measured height in early pregnancy and patient-reported weight before pregnancy. Gestational age was estimated depending upon LMP and subsequently adjusted according to early pregnancy ultrasonography.

Fasting blood samples were obtained in the first trimester (between 9 and 13 weeks) from the median cubital vein; and maternal serum samples were collected in 10-mL vacutainer tubes, centrifuged, and stored at −80°C before analysis. Quantitative analyses of TSH and fT4 and TPOAb concentrations were measured with kits (ARCHITECT i2000; Abbott, Chicago, IL) according to the manufacturer's protocol. The intra- and interassay coefficients of variation were 1.6% and 3.59%, respectively, for TSH; 1.9% and 4.01%, respectively, for fT4; and 10% and 10%, respectively, for TPOAb. TPOAb concentrations >5.6 IU/mL were regarded as positive according to manufacturer-defined cutoff.

hCG serum concentrations were determined by a solid-phase, two-site, chemiluminescent immunometric assay on the Immulite 2000 XPi system (Siemens Healthcare Diagnostics, Deerfield, IL). The intra-assay variation coefficients were 8.0%, 6.3%, and 5.1% at concentrations of 9.7, 53.1, and 821.5 IU/L, respectively. Fasting plasma glucose (FPG) concentrations were tested at the first antenatal screening visit using a GOD-PAP kit (Human, Wiesbaden, Germany) with an intra-assay coefficient of variation of 4.3% and interassay coefficient of variation of 3.4%. Hemoglobin A1c (HbA1c) concentrations were measured using Roche Diagnostics HbA1c kits with an autoanalyzer (Cobas Integra 800; Roche Diagnostics, Mannheim, Germany); intra- and interassay coefficients of variation were 2.3% and 2.2%, respectively.

Diagnostic criteria

GDM was diagnosed with a 2-hour 75-g OGTT at 24–28 weeks gestation based on the criteria of the American Diabetes Association: a fasting glucose concentration (OGTT0) ≥92 mg/dL (5.1 mmol/L), 1-hour glucose concentration (OGTT1) ≥180 mg/dL (10.0 mmol/L), and/or 2-hour glucose concentration (OGTT2) ≥153 mg/dL (8.5 mmol/L) (28).

Statistical analyses

To compare concentrations throughout gestation, hCG values were transformed to week-specific multiples of the median (MoM) for analyses (29). Odds ratios and their associated 95% confidence intervals (CIs) derived from multivariable logistic regression models were applied to explore the associations of hCG, TSH, and fT4 with GDM. The beta coefficients (β) and their associated CIs derived from multivariable linear regression models were used to assess the associations between hCG and TSH, fT4, glucose levels, or HbA1c. To assess potential nonlinearity, we applied restricted cubic splines utilizing three knots. All models were adjusted for maternal age, education level, family history of diabetes, parity, fetal sex, TPOAb status, and prepregnancy BMI. Independent variables were added to the model as continuous variables. To avoid the influence of outliers, we removed the top 1% hCG MoM from further investigation.

Mediation analysis was used to determine potential mediation effects of TSH or fT4 on the association of hCG with GDM. Based on the prerequisites that the relationships between the exposure and mediator and the outcomes are all statistically significant, the total effect was separated into average direct effects and average causal mediation effects (ACMEs) (30). Total effect was the effect of hCG on GDM, and ACME was the effect mediated through the maternal thyroid hormone. The mediation proportion was obtained by calculating the ACME divided by the total effect.

Since TPOAb positivity may result in an attenuated hCG-induced increase in fT4 (31), we also performed sensitivity analysis of the association of hCG with GDM, restricting to TPOAb-positive cases, to confirm the mediation role of thyroid function. To evaluate potential selection bias, we compared characteristics between the included pregnant women (18,683 subjects) and excluded pregnant women without specific medical data (8792 subjects) by using the Wilcoxon two-sample test or chi-squared test.

Missing data on BMI (1.8% of cases) were dealt with multiple imputation using the Markov chain Monte Carlo method by creating and pooling five imputed data sets based on the missing at random assumption (32). Nonmissing variables (GDM, maternal age, education level, family history of diabetes, parity, fetal sex, TSH, fT4, hCG MoM, OGTT values, and TPOAb concentrations) were added to the model as prediction variables. There were no significant differences in descriptive characteristics between original and imputed data sets.

Analyses were performed using R Statistical Software, version 3.6.0 (R Foundation for Statistical Computing, Vienna, Austria), using the following packages: mice, rms, and mediation. Statistical tests were two-sided, with a p-value <0.05 considered statistically significant.

Results

After exclusions, the final study population was 18,683 women (Fig. 1), the descriptive data of which are shown in Table 1. In the study population, the median (interquartile range) maternal age was 29 (27 –32) years, median BMI was 20.5 (19.1–22.3) kg/m², 1086 (5.8%) women had a family history of diabetes, and 2214 (11.9%) women were defined as GDM cases. The median TSH was 1.15 (0.64–1.78) mU/L, median fT4 was 15.0 (13.9–16.2) pmol/L, and 1833 women (9.8%) were TPOAb positive. Compared with women excluded for lack of specific medical data, included participants were likely to have lower hCG concentrations, lower fT4 concentrations, and higher TSH concentrations (Supplementary Table S1).

FIG. 1.

Flow chart of the study population. IVF, in vitro fertilization; OGTT, oral glucose tolerance test.

Table 1.

Demographic Data of the Study Population

Characteristics	Value
Gestational age at sampling, median, (IQR), weeks	12.1 (11.7–13.2)
Age, median (IQR), years	29 (27 –32)
BMI, median (IQR), kg/m²	20.5 (19.1–22.3)
Parity, n (%)
Primiparous	157,78 (84.5)
Multiparous	2905 (15.5)
Education level, n (%)
Below college degree	5116 (27.4)
Bachelor's degree	10,278 (55.0)
Master's degree	3065 (16.4)
PhD degree	224 (1.2)
Fetal sex, n (%)
Male	9675 (51.8)
Female	9008 (48.2)
Family history of diabetes, n (%)
Yes	1086 (5.8)
No	17,597 (94.2)
GDM, n (%)
Yes	2214 (11.9)
No	16,469 (88.1)
hCG, median (IQR), IU/L	54,840 (36,843, 83060)
hCG, median (IQR), MoM	1.39 (0.95–2.10)
TSH, median (IQR), mU/L	1.15 (0.64–1.78)
fT4, median (IQR), pmol/L	15.0 (13.9–16.2)
TPOAb, n (%)
Positive	1833 (9.8)
Negative	16,850 (90.2)

BMI, body–mass index; fT4, free thyroxine; GDM, gestational diabetes mellitus; hCG, human chorionic gonadotropin; IQR, interquartile range; MoM, multiples of the median; TPOAb, thyroperoxidase antibody; TSH, thyrotropin.

Association of hCG with glucose measurements and GDM

A higher hCG level was associated with lower glucose measurements during the OGTT (β = −0.007 [CI −0.013 to −0.0006], p = 0.032, for fasting glucose concentration; β = −0.064 [CI −0.086 to −0.043], p < 0.001, for 1-hour glucose concentration; and β = −0.026 [CI −0.046 to −0.006], p = 0.011, for 2-hour glucose concentration) (Table 2). There were no associations of hCG with FPG and HbA1c, which were tested during early pregnancy. A higher hCG level was associated with a 6.4–11.8% lower risk of GDM, depending on the hCG cutoff value used (p = 0.027; Fig. 2).

FIG. 2.

The association of hCG with the risk of GDM. This figure shows the association of hCG concentration as MoM with GDM (p = 0.027). The graph (left) shows predicted mean with CI calculated using the multivariable logistic regression model. The table (right) shows various cutoff points for MoM variables to quantify the results from the graph more readily. All analyses were adjusted for maternal age, education level, family history of GDM, parity, fetal sex, TPOAb status, and prepregnancy BMI. BMI, body–mass index; CI, 95% confidence interval; GDM, gestational diabetes mellitus; hCG, human chorionic gonadotropin; MoM, multiples of the median; OR, odds ratio; TPOAb, thyroperoxidase antibody.

Table 2.

The Association of Human Chorionic Gonadotropin (Multiples of the Median) with Glucose Measurements

	β^*	CI	P
First trimester tests
FPG	−0.003	−0.010 to 0.003	0.270
HbA1c	−0.002	−0.008 to 0.003	0.380
Second trimester tests
OGTT0	−0.007	−0.013 to −0.0006	0.032
OGTT1	−0.064	−0.086 to −0.043	<0.001
OGTT2	−0.026	−0.046 to −0.006	0.011

^*β: The beta coefficients derived from multivariable linear models. Multivariable linear models were adjusted for maternal age, education level, family history of diabetes, parity, fetal sex, TPOAb status, and prepregnancy BMI.

CI, 95% confidence interval; FPG, fasting plasma glucose tested in early pregnancy; HbA1c, hemoglobin A1c; OGTT0, fasting level in the oral glucose tolerance test at 24–28 weeks; OGTT1, 1-hour level in the oral glucose tolerance test at 24–28 weeks; OGTT2, 2-hour level in the oral glucose tolerance test at 24–28 weeks.

Role of thyroid function

We verified the negative association of fT4 with GDM as well as the positive association of hCG with fT4 in the current study population. Although hCG was negatively associated with TSH, there was no association between TSH and GDM (Supplementary Fig. S1; Supplementary Table S2). The mediation analysis revealed a total effect of hCG on GDM of −0.0058 ([CI −0.0115 to −0.0016], p < 0.05), including a significant direct effect of −0.0046 ([CI −0.0106 to −0.0008], p < 0.05) (Table 3). A statistically significant mediation effect of hCG associated with GDM (mean indirect effect, 0.0013 [CI −0.0021 to −0.0004], p < 0.001) through maternal fT4 was found, and the estimated proportion of the mediation effect was 21.4% ([CI 8.0% to 54.2%], p < 0.05).

Table 3.

Mediation Analysis of the Association of Human Chorionic Gonadotropin (Multiples of the Median) with Gestational Diabetes Mellitus by Free Thyroxine Concentrations

Total effect (CI)	ADE (CI)	ACME (CI)	Proportion of mediation (%)
−0.0058 (−0.0115 to −0.0016)^*	−0.0046 (−0.0106 to −0.0008)^*	−0.0013 (−0.0021 to −0.0004)^**	21.4 (8.0 to 54.2)^*

Multivariable logistic models were adjusted for maternal age, education level, family history of diabetes, parity, fetal sex, TPOAb status, and prepregnancy BMI.

p-value <0.05.

p-value <0.001.

ACME, average causal mediation effect; ADE, average direct effect.

Sensitivity analysis

In the sensitivity analysis, selecting only TPOAb-positive women, there was no association of hCG with GDM (p = 0.452; Supplementary Fig. S2).

Discussion

In the current study, we showed that a higher hCG level in early pregnancy is associated with a lower glucose concentration during an OGTT at week 24–28 and a lower risk of GDM. Furthermore, we showed that an estimated 21.4% of the association of hCG with GDM was mediated through fT4.

GDM is a global health issue, affecting the short- and long-term health of both mothers and their offspring (4 –6). While various studies have tried to identify risk factors and pathophysiological mechanisms for GDM, the underlying mechanisms are complex and remain to be elucidated (9,10). Several studies have reported that mothers subsequently diagnosed with GDM had lower first trimester hCG concentrations compared with normoglycemic controls (21 –23), and two other studies reported that a higher first trimester hCG was associated with a lower risk of GDM (24,25). With other studies showing no association at all (26,27), the current literature remains inconsistent, which could partially be explained by interstudy differences in population ethnicity, sample sizes, and heterogeneous GDM diagnostic criteria. In the present study, we show that a higher hCG concentration is associated with a lower risk of GDM.

Sex hormone-binding globulin (SHBG) is a biomarker negatively associated with the risk of GDM (33), and it is proposed that the relationship between SHBG and GDM is attributed to its effect on insulin resistance (IR) (34). Recently, a study suggested that decreased SHBG expression might be involved in IR by upregulating the activity of the extracellular signal-regulated kinase (ERK) pathway in the placenta (35). Considering that activation of the ERK pathway is highly associated with hCG elevation in placental tissues (36), production of hCG may largely increase as an adaptive process to compete against lower SHBG-induced IR in this regard. Indeed, Bartha et al. showed that hCG in early pregnancy was negatively correlated with SHBG (37).

Although there is a time lag between measuring early pregnancy hCG and the OGTT, standardization of hCG according to gestational age using MoM transformation creates a variable that better represents relative overall hCG availability (29). Furthermore, hCG was measured at the gestational age, during which important placentation processes are regulated by both hCG and the thyroid hormone (38,39). Since placental endocrine IR is a major driver of GDM, it is possible that alterations in early pregnancy placentation become apparent during later pregnancy (40). hCG may maintain insulin sensitivity and glucose homeostasis by inhibiting interleukin-6 (IL-6) endometrial secretion in a concentration-dependent manner (40,41), which is one of several proinflammatory cytokines that can directly contribute to the development of IR and disturbances of glucose metabolism (42). In addition, hCG may also have a direct influence on glucose metabolism by regulation of glycogenolysis (13) or prevent autoimmune diabetes through immunomodulatory effects on effector cells (14).

The lack of associations of hCG with FPG or HbA1c concentrations tested during early pregnancy combined with the negative associations of hCG with OGTT values tested in the second trimester may suggest that effects of hCG on glucose homeostasis could be cumulative. hCG concentration rises from pregnancy implantation and peaks at around 10 weeks (29). Given the concentration-dependent inhibitive effect of hCG on IL-6 secretion, increasing hCG may cumulatively suppress IL-6 levels till the late first trimester (40,41). On the other hand, researchers have revealed that higher IL-6 levels in the late first trimester were significantly associated with the development of GDM diagnosed at 24–28 weeks (43). Therefore, early pregnancy hCG might influence the risk of GDM due to its cumulative control on low-grade inflammation, preventing the pathogenesis of IR before the clinical manifestation of this disease.

We hypothesized that early pregnancy hCG could also be associated with GDM by regulating maternal thyroid hormones. First, hCG is a major determinant of maternal thyroid hormone concentrations during early pregnancy. Owing to its affinity for the TSH receptor, hCG can exert thyrotropic activity and result in an increase in fT4 concentrations, which subsequently lead to a decrease in TSH concentrations (15,44). It is believed that early pregnancy hCG can be responsible for an up to 50% increase in maternal fT4 concentrations, which can not only facilitate the increased energy expenditure and metabolic demand in mothers but also safeguard adequate supply of thyroxine for fetal development (45). Second, early pregnancy maternal thyroid hormones have been found to be negatively associated with the risk of GDM (20). Although the specific biological mechanisms need to be clarified, thyroid hormones might be involved in the maintenance of glucose homeostasis as follows. Thyroid hormones may affect glucose metabolism through regulation of the type 2 glucose transporter in liver cell membranes and modification of catecholamine sensitivity, which can subsequently accelerate glycogenolysis (20). Thyroid hormones also regulate insulin secretion and sensitivity and exert islet β-cell-protective effects, which may thereby prevent the onset of IR and glucose intolerance (16 –19,46). Taken together, maternal thyroid hormones may act as a bridge between hCG and the risk of GDM.

Our mediation analysis suggests that maternal fT4 concentration accounts for 21.4% of the association of hCG with GDM. A previous study has reported that thyroid autoimmunity, such as TPOAb positivity, can attenuate the thyroid response to hCG stimulation, resulting in an impaired increase in fT4 caused by hCG (31). Therefore, we performed a sensitivity analysis and identified that in TPOAb-positive women, no association between hCG and GDM was found. Our findings may suggest that hCG could be involved in the pathophysiology of GDM, in which maternal fT4 might play a mediating role. Still, our findings should be further verified by more studies.

We were able to study a large group of women, with prospectively collected data available, including a wide range of possible confounding factors. However, the interpretation of our results is subject to certain limitations. First, we had only a single measurement of hCG and thyroid function available during the first trimester. Changes throughout pregnancy may better represent placental function, thyroid function, or hCG availability, and therefore future research should be performed using repeated measurements. Second, a number of women lacked specific medical data in our cohort because they refused to undergo an OGTT due to discomfort or due to a missed appointment. This may have led to a selection bias affecting the generalizability of our results. Third, as these data were solely collected among a predominant Han Chinese population, it remains unknown to what extent these results are generalizable to other populations. Finally, our results are based on observational data, which limit the inference of causality, and replications of our results, including the mediation analyses, in other large populations are warranted.

In conclusion, we show that higher hCG levels in early pregnancy are associated with a lower risk of GDM, which could be mediated by maternal fT4 concentrations. This study provides new insights into the potential mechanism involved in the development of glucose intolerance.

Footnotes

Acknowledgments

The authors gratefully acknowledge the contributions and efforts of all pregnant women who participated in this study and the doctors and nurses involved in data collection and patient care.

Authors' Contributions

Y.L. and F.G. designed the study. Y.Z., C.Z., and J.F. supervised data collection. Y.L. and F.G. conducted statistical analyses. Y.L. drafted the original version of manuscript. All authors contributed to revision and approved the final version for publication. C.Z., T.I.M.K., and J.F. are the guarantors for this work and accept full responsibility for the conduct of the study, had access to the data, and controlled the decision to publish.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This work was supported by grants from the National Key Research and Development Program of China (2018YFC1004602) and the National Natural Science Foundation of China (81974235, 81501274). The study is also supported by the Chinese Academy of Medical Sciences Research Unit (No. 2019RU056), Shanghai Jiao Tong University, CAMS Innovation Fund for Medical Sciences (CIFMS) (No. 2019-I2M-5-064), and Shanghai Municipal Key Clinical Specialty, Shanghai, China.

Supplementary Material

Supplementary Table S1

Supplementary Table S2

Supplementary Figure S1

Supplementary Figure S3

References

World Health Organization 1999 Definition, Diagnosis and Classification of Diabetes Mellitus and Its Complications: Report of a WHO Consultation. Part 1, Diagnosis and Classification of Diabetes Mellitus. Geneva: World Health Organization.

International Diabetes Federation 2019 IDF Diabetes Atlas. Ninth edition. Brussels, Belgium: International Diabetes Federation.

Gao

, Sun

, Lu

, Liu

, Yuan

. 2019. Prevalence of gestational diabetes mellitus in mainland China: a systematic review and meta-analysis. J Diabetes Investig, 10:154–162.

Clausen

, Mathiesen

, Hansen

, Pedersen

, Jensen

, Lauenborg

, Damm

. 2008. High prevalence of type 2 diabetes and pre-diabetes in adult offspring of women with gestational diabetes mellitus or type 1 diabetes: the role of intrauterine hyperglycemia. Diabetes Care, 31:340–346.

Metzger

, Lowe

, Dyer

, Trimble

, Chaovarindr

, Coustan

, Hadden

, McCance

, Hod

, McIntyre

, Oats

, Persson

, Rogers

, Sacks

. 2008. Hyperglycemia and adverse pregnancy outcomes. N Engl J Med, 358:1991–2002.

Tutino

, Tam

, Ozaki

, Yuen

, So

, Chan

, Ko

, Yang

, Chan

, Tam

, Ma

. 2020. Long-term maternal cardiometabolic outcomes 22 years after gestational diabetes mellitus. J Diabetes Investig, 11:985–993.

American Diabetes

. 2018. 2. Classification and diagnosis of diabetes: standards of medical care in diabetes-2018. Diabetes Care, 41:S13–S27.

Artzi

, Shilo

, Hadar

, Rossman

, Barbash-Hazan

, Ben-Haroush

, Balicer

, Feldman

, Wiznitzer

, Segal

. 2020. Prediction of gestational diabetes based on nationwide electronic health records. Nat Med, 26:71–76.

Powe

. 2017. Early pregnancy biochemical predictors of gestational diabetes mellitus. Curr Diab Rep, 17:12.

10.

Savvidou

, Nelson

, Makgoba

, Messow

, Sattar

, Nicolaides

. 2010. First-trimester prediction of gestational diabetes mellitus: examining the potential of combining maternal characteristics and laboratory measures. Diabetes, 59:3017–3022.

11.

Kagan

, Wright

, Valencia

, Maiz

, Nicolaides

. 2008. Screening for trisomies 21, 18 and 13 by maternal age, fetal nuchal translucency, fetal heart rate, free beta-hCG and pregnancy-associated plasma protein-A. Hum Reprod (Oxford, England), 23:1968–1975.

12.

, Sahota

, Lao

, Xu

, Hu

, Zhang

, Liu

, Sun

, Tang

, Ma

. 2016. Applicability of first-trimester combined screening for fetal trisomy 21 in a resource-limited setting in mainland China. BJOG, 123(Suppl. 3):23–29.

13.

Desoye

, Shafrir

. 1994. Placental metabolism and its regulation in health and diabetes. Mol Aspects Med, 15:505–682.

14.

Khil

, Jun

, Kwon

, Yoo

, Kim

, Notkins

, Yoon

. 2007. Human chorionic gonadotropin is an immune modulator and can prevent autoimmune diabetes in NOD mice. Diabetologia, 50:2147–2155.

15.

Korevaar

TIM

, Medici

, Visser

, Peeters

. 2017. Thyroid disease in pregnancy: new insights in diagnosis and clinical management. Nat Rev Endocrinol, 13:610–622.

16.

Bertrand

, Blanchet

, Pessemesse

, Annicotte

, Feillet-Coudray

, Chabi

, Levin

, Fajas

, Cabello

, Wrutniak-Cabello

, Casas

. 2013. Mice lacking the p43 mitochondrial T3 receptor become glucose intolerant and insulin resistant during aging. PLoS One, 8:e75111.

17.

Crunkhorn

, Patti

. 2008. Links between thyroid hormone action, oxidative metabolism, and diabetes risk?. Thyroid, 18:227–237.

18.

Verga Falzacappa

, Mangialardo

, Raffa

, Mancuso

, Piergrossi

, Moriggi

, Piro

, Stigliano

, Torrisi

, Brunetti

, Toscano

, Misiti

. 2010. The thyroid hormone T3 improves function and survival of rat pancreatic islets during in vitro culture. Islets, 2:96–103.

19.

Verga Falzacappa

, Panacchia

, Bucci

, Stigliano

, Cavallo

, Brunetti

, Toscano

, Misiti

. 2006. 3,5,3'-triiodothyronine (T3) is a survival factor for pancreatic beta-cells undergoing apoptosis. J Cell Physiol, 206:309–321.

20.

Yang

, Shi

, Leung

, Huang

, Fan

. 2016. Low thyroid hormone in early pregnancy is associated with an increased risk of gestational diabetes mellitus. J Clin Endocrinol Metabol, 101:4237–4243.

21.

Ong

, Liao

, Spencer

, Munim

, Nicolaides

. 2000. First trimester maternal serum free beta human chorionic gonadotrophin and pregnancy associated plasma protein A as predictors of pregnancy complications. BJOG, 107:1265–1270.

22.

Spencer

, Cowans

. 2013. The association between gestational diabetes mellitus and first trimester aneuploidy screening markers. Ann Clin Biochem, 50:603–610.

23.

Donovan

, Nidey

, Jasper

, Robinson

, Bao

, Saftlas

, Ryckman

. 2018. First trimester prenatal screening biomarkers and gestational diabetes mellitus: a systematic review and meta-analysis. PLoS One, 13:e0201319.

24.

Sirikunalai

, Wanapirak

, Sirichotiyakul

, Tongprasert

, Srisupundit

, Luewan

, Traisrisilp

, Tongsong

. 2016. Associations between maternal serum free beta human chorionic gonadotropin (β-hCG) levels and adverse pregnancy outcomes. J Obstetr Gynaecol, 36:178–182.

25.

Visconti

, Quaresima

, Chiefari

, Caroleo

, Arcidiacono

, Puccio

, Mirabelli

, Foti

, Di Carlo

, Vero

, Brunetti

. 2019. First Trimester Combined Test (FTCT) as a predictor of gestational diabetes mellitus. Int J Environ Res Public Health, 16:3654.

26.

Savvidou

, Syngelaki

, Muhaisen

, Emelyanenko

, Nicolaides

. 2012. First trimester maternal serum free β-human chorionic gonadotropin and pregnancy-associated plasma protein A in pregnancies complicated by diabetes mellitus. BJOG, 119:410–416.

27.

Sweeting

, Wong

, Appelblom

, Ross

, Kouru

, Williams

, Sairanen

, Hyett

. 2018. A first trimester prediction model for gestational diabetes utilizing aneuploidy and pre-eclampsia screening markers. J Matern Fetal Neonatal Med, 31:2122–2130.

28.

American Diabetes Association. 2011. Standards of medical care in diabetes—2011. Diabetes Care, 34(Suppl. 1):S11–S61.

29.

Korevaar

, Steegers

, de Rijke

, Schalekamp-Timmermans

, Visser

, Hofman

, Jaddoe

, Tiemeier

, Visser

, Medici

, Peeters

. 2015. Reference ranges and determinants of total hCG levels during pregnancy: the Generation R Study. Eur J Epidemiol, 30:1057–1066.

30.

Ato Garcia

, Vallejo Seco

, Ato Lozano

. 2014. Classical and causal inference approaches to statistical mediation analysis. Psicothema, 26:252–259.

31.

Korevaar

, Steegers

, Pop

, Broeren

, Chaker

, de Rijke

, Jaddoe

, Medici

, Visser

, Tiemeier

, Peeters

. 2017. Thyroid autoimmunity impairs the thyroidal response to human chorionic gonadotropin: two population-based prospective cohort studies. J Clin Endocrinol Metabol, 102:69–77.

32.

Sterne

, White

, Carlin

, Spratt

, Royston

, Kenward

, Wood

, Carpenter

. 2009. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ (Clin Res ed.), 338:b2393.

33.

Thadhani

, Wolf

, Hsu-Blatman

, Sandler

, Nathan

, Ecker

. 2003. First-trimester sex hormone binding globulin and subsequent gestational diabetes mellitus. Am J Obstetr Gynecol, 189:171–176.

34.

Zhang

, Jin

, Sun

, Zheng

, Jiang

, Feng

, Wang

. 2016. Expression and correlation of sex hormone-binding globulin and insulin signal transduction and glucose transporter proteins in gestational diabetes mellitus placental tissue. Diabetes Res Clin Pract, 119:106–117.

35.

Wang

, Chi

, Feng

, Zhang

, Jin

. 2020. Sex hormone-binding globulin regulates the activity of the ERK pathway in the placentas of patients with gestational diabetes mellitus. Biochem Biophys Res Commun, 532:613–619.

36.

Wang

, Zhao

, Wang

, Gao

, Yao

. 2013. The human leukocyte antigen G promotes trophoblast fusion and β-hCG production through the Erk1/2 pathway in human choriocarcinoma cell lines. Biochem Biophys Res Commun, 434:460–465.

37.

Bartha

, Comino-Delgado

, Romero-Carmona

, Gomez-Jaen

. 2000. Sex hormone-binding globulin in gestational diabetes. Acta Obstetr Gynecol Scand, 79:839–845.

38.

Adu-Gyamfi

, Wang

, Ding

. 2020. The interplay between thyroid hormones and the placenta: a comprehensive review. Biol Reprod, 102:8–17.

39.

Cole

. 2010. Biological functions of hCG and hCG-related molecules. Reprod Biol Endocrinol, 8:102.

40.

Desoye

, Hauguel-de Mouzon

. 2007. The human placenta in gestational diabetes mellitus. The insulin and cytokine network. Diabetes Care, 30(Suppl. 2):S120–S126.

41.

Perrier d'Hauterive

, Charlet-Renard

, Berndt

, Dubois

, Munaut

, Goffin

, Hagelstein

, Noel

, Hazout

, Foidart

, Geenen

. 2004. Human chorionic gonadotropin and growth factors at the embryonic-endometrial interface control leukemia inhibitory factor (LIF) and interleukin 6 (IL-6) secretion by human endometrial epithelium. Hum Reprod (Oxford, England), 19:2633–2643.

42.

Senn

, Klover

, Nowak

, Mooney

. 2002. Interleukin-6 induces cellular insulin resistance in hepatocytes. Diabetes, 51:3391–3399.

43.

Abell

, Shorakae

, Harrison

, Hiam

, Moreno-Asso

, Stepto

, De Courten

, Teede

. 2017. The association between dysregulated adipocytokines in early pregnancy and development of gestational diabetes. Diabetes Metabol Res Rev, 33:77–83.

44.

Glinoer

, de Nayer

, Bourdoux

, Lemone

, Robyn

, van Steirteghem

, Kinthaert

, Lejeune

. 1990. Regulation of maternal thyroid during pregnancy. J Clin Endocrinol Metabol, 71:276–287.

45.

Zhang

, Zhang

, Yang

, Meng

, Liu

, Peeters

, Huang

, Korevaar

TIM

, Fan

. 2019. Association of maternal thyroid function and thyroidal response to human chorionic gonadotropin with early fetal growth. Thyroid, 29:586–594.

46.

Mullur

, Liu

, Brent

. 2014. Thyroid hormone regulation of metabolism. Physiol Rev, 94:355–382.