Low Thyroid Hormone in Early Pregnancy Is Associated With an Increased Risk of Gestational Diabetes Mellitus

Context:

Although thyroid dysfunction in early pregnancy may have adverse effects on pregnancy outcome and offspring, few prospective studies have evaluated these effects.

Objective:

Our aim was to evaluate the correlations between different thyroid hormone levels in early pregnancy and the incidence of gestational diabetes mellitus (GDM).

Setting and Participants:

The study comprised 27 513 mothers who provided early pregnancy serum samples for analyses of thyroid function. GDM was diagnosed using a 2 hours, 75-g oral glucose tolerance test, and the mothers were grouped and compared according to the results.

Main Outcome Measures:

We focused on GDM during the index pregnancy.

Results:

The incidence of GDM in pregnant women tended to increase with age (5.83%, 10.18%, 14.95%, and 22.40%; P < .0001). The incidence of GDM increased with increasing prepregnancy body mass index (P < .0001). Pregnant women with a family history of diabetes had a much higher incidence of GDM than those without a family history of diabetes (21.09% vs 12.92%; P < .0001). The level of free T₄ (FT₄) in early pregnancy in GDM women was lower than that in non GDM women (P < .0001). With increasing early pregnancy FT₄, the rate of incident GDM was decreasing (P < .0001).

Conclusions:

Low thyroid hormone levels in early pregnancy are a risk factor for GDM incidence.

Many studies have shown that gestational diabetes mellitus (GDM) and functional abnormalities in the thyroid can have a variety of adverse effects on pregnancy outcomes and offspring (1–9). Thyroid disease is known to impact pregnancy outcomes, and gestational diabetes is the most common obstetric metabolic disease. Both of these conditions can cause short- and long-term harm to the mother and child, and an increasing number of scholars have therefore begun to investigate whether there is a correlation between thyroid disease and GDM (10–17). Blood sugar levels and maternal thyroid function during pregnancy are influenced by various physiological hormones, such as estrogen, thyroid-binding globulin, human chorionic gonadotropin, placental lactogen, cortisol, and placental insulin enzyme (18, 19). A pregnant woman’s physiological characteristics are different from those in women who are not pregnant. Some studies have shown that there is a correlation between thyroid disease and GDM (10–12, 14, 15), whereas others have not found this association (16, 17). However, a recent metaanalysis (13) showed that the incidence of GDM in patients with subclinical hypothyroidism was 1.35-fold higher than the incidence in the control group. Because there are so many differences in race, region, genetics, environmental factors, diagnostic criteria for GDM, and gestation-specific thyroid function reference intervals, few studies have been conducted in a Chinese regional population of pregnant women (20).

Therefore, in this study, we aimed to evaluate the correlations between the levels of different thyroid hormones during early pregnancy and the incidence of GDM.

Materials and Methods

Study population and data collection

This study was performed in accordance with relevant guidelines and regulations. Ethical approval for this project was granted by the Ethics Committee of the International Peace Maternity and Child Health Hospital, School of Medicine, Shanghai Jiao Tong University.

For this retrospective study, we included pregnant women who underwent a first-trimester antenatal screening and who planned to deliver at the International Peace Maternity and Child Health Hospital from January 2013 to October 2015. The following inclusion criteria were used: 1) a single birth; 2) Chinese Han women; 3) residing in an iodine-sufficient area; 4) completion of a 2 hours, 75-g oral glucose tolerance test (OGTT) at 24–28 gestational weeks; and 5) delivery in the hospital. The following exclusion criteria were applied: 1) diabetes before pregnancy or diabetes diagnosed during the first prenatal examination; 2) polycystic ovary syndrome; 3) personal or family history of thyroid disease; 4) visible or palpable goiter; 5) taking hormone drugs that could affect thyroid function before or during pregnancy or receiving thyroid hormone replacement therapy; 6) history of autoimmune disease; 7) multiple pregnancy; and 8) incomplete medical history.

The characteristics and menstrual cycle parameters of the patients were obtained from databases. Approximately 36 389 pregnant women were initially included in this study, and the following patients were excluded: 43 patients with prepregnancy diabetes, 51 patients with a family history of thyroid disease, 1104 patients with a history of thyroid disease or surgical history related to the thyroid, 864 patients taking hormone drugs that affect thyroid function, 815 cases of twins, 65 patients with polycystic ovary syndrome, 5697 patients without OGTT results, and 237 patients with an incomplete medical record. Ultimately, 27 513 pregnant women (age <25 y, 926 cases; age 25–29 y, 12 244 cases; age 30–34 y, 10 973 cases; age ≥35 y, 3370 cases) were included.

The characteristics of the study population are shown in a flow chart in Figure 1.

Thyroid function assessment

At the first antenatal visit (gestational age, 9–13 wk), maternal serum samples were collected in 10-mL vacutainer tubes, centrifuged, and stored in aliquots at −80°C until assayed. Quantitative analyses of thyroid hormones (TSH and free T₄ [FT₄]) and autoantibodies associated with autoimmune thyroiditis (thyroid-peroxidase antibody [TPOAb]) were performed using chemiluminescent microparticle immunoassays in an Architect i2000 automatic analyzer (Abbott Diagnostics). The lower limits of detection and the intra- and interassay coefficients of variation were 0.0038 mIU/L and 1.6% and 3.59%, respectively, for TSH; 0.6200 pmol/L and 1.9% and 4.01%, respectively, for FT₄; and 1.0 IU/mL and 10% and 10%, respectively, for TPOAb.

Gestational diabetes mellitus

The women were screened for GDM at 24–28 weeks and classified at that time as having GDM in the index pregnancy if any abnormal plasma glucose values were obtained during the 2 hours, 75-g OGTT. Abnormal values were defined according to the following standard diagnostic criteria established by the American Diabetes Association (21): a fasting level of 92 mg/dL (5.1 mmol/L) or greater, a 1-hour value of 180 mg/dL (10.0 mmol/L) or greater, and a 2-hour value of 153 mg/dL (8.5 mmol/L) or greater. The intra- and interassay coefficients of variation were 5% and 10%, respectively.

Statistical analysis

All statistical analyses were performed using SAS software (SAS Institute, Inc). A Pearson’s χ² test and a 2-tailed Student’s t test, ANOVA, or nonparametric Wilcoxon rank test were used to evaluate the statistical significance of the numerical and measurement data, as appropriate. Multivariate logistic regression was used to evaluate the relationship between thyroid hormone levels and GDM. The results are represented as odds ratios (ORs) and 95% confidence intervals (95% CIs). Logistic regression models and χ² tests were used to analyze associations between FT₄, TSH, or TPOAb levels and the incidence of GDM. Differences were considered to be statistically significant at P < .05.

Results

Clinical characteristics of the participants

As shown in Table 1, the incidence of GDM in pregnant women tended to increase with age (5.83%, 10.18%, 14.95%, and 22.40%; P < .0001). The incidence of GDM also increased with increasing prepregnancy body mass index (BMI) (P < .0001). The cesarean section rate was higher in pregnant women with GDM than in those without GDM (15.46% vs 11.75%; P < .0001) and was higher in multiparous pregnancies (P < .0001). Pregnant women with a family history of diabetes had a much higher incidence of GDM than those without (21.09% vs 12.92%; P < .0001). However, the incidence of GDM was not related to employment status or education level (all P > .05).

Thyroid hormone levels in early pregnancy in GDM and non-GDM women

The levels of TSH and FT₄ in early pregnancy were both significantly lower in GDM women than in non-GDM women (P = .024 and P < .0001, respectively), whereas the difference in the level of TPOAb was not significantly different (Table 2).

Correlation analysis between different thyroid hormone levels and the risk of GDM in early pregnancy

As shown in Table 3 and Figure 2, we used a logistic regression analysis to investigate whether thyroid hormone levels in early pregnancy affect the risk of GDM. The results showed that before adjusting for age, BMI, delivery method, parity, a family history of diabetes, and FT₄ and TPOAb levels, have a level of TSH more than 1.96 mU/L reduced the risk of developing GDM (OR 0.88, 95%CI 0.79–0.99). After adjustment, TSH more than 1.36 mU/L was associated with a decreased risk of GDM. Using an FT₄ level between 9 and 13.7 pmol/L as a reference, we found that higher levels of FT₄ provided a protective effect against GDM before we adjusted for age, BMI, delivery method, parity, a family history of diabetes, and TSH and TPOAb levels. However, this effect was slightly decreased after we performed this adjustment. TPOAb had no effect on the incidence of GDM either before or after adjustment and regardless of positivity.

To further analyze the correlations between FT₄, TSH, and TPOAb levels and the incidence of GDM (Table 3 and Figure 3), we evaluated the incidence of GDM in patients with different levels of FT₄, TSH, or TPOAb and found that FT₄ played a role in predicting GDM in that the incidence of GDM gradually decreased as the level of FT₄ increased (P < .0001). In contrast, TSH and TPOAb did not predict GDM. These results indicated that a low level of FT₄ is an independent risk factor for GDM.

Discussion

Our study yielded the following results. First, the incidence of GDM in pregnant women tended to increase with maternal age. The incidence of GDM also increased with increasing prepregnancy BMI. Pregnant women with a family history of diabetes had a much higher incidence of GDM than those without a family history of diabetes. FT₄ levels in early pregnancy were lower in GDM women than in non-GDM women. Higher levels of FT₄ showed a protective effect against GDM, in that the incidence of GDM decreased as the level of FT₄ increased. Therefore, this study revealed that hypothyroxinemia is a risk factor for the development of diabetes during pregnancy.

In recent years, the incidence of diabetes has increased annually, especially in Asia (22). According to one survey, the incidence of diabetes in China increased from 1% in 1980 to approximately 9.7% in 2008 (23). Our study found that the incidence rate of GDM has gradually increased as prepregnancy BMI has increased. We found that the incidence rate of GDM is approximately 13.44%, which is significantly higher than the rate in European and American women. Studies have indicated that when the BMI is the same, Asian populations have a higher incidence of diabetes than other ethnicities. Possible explanations for this racial difference include the following. First, centripetal obesity diabetes is an independent predictive factor (24), and abdominal and visceral fat accumulates to a greater degree in Asian individuals than in European and American individuals with the same waist circumference. Second, although body weights are lower in Asian populations than in European and American populations, there is a higher incidence of insulin resistance in Asian populations (25). Third, mitochondrial dysfunction is involved in the development of central obesity and insulin resistance (26, 27). More common mitochondrial DNA 16 189 variants have also been shown to contribute to the onset of type 2 diabetes in East Asian but not in European populations (28). Hence, although diabetes is a global problem, more attention should be paid to diabetes in Asia.

As FT₄ levels increased, the incidence of GDM gradually decreased, which intuitively and powerfully illustrates that a low FT₄ level is an independent risk factor for GDM. Thyroid hormones play an important role in glucose metabolism. T₃ is the biologically active hormone that is primarily responsible for glucose metabolic activity (29), and 80% of circulating T₃ is converted peripherally via deiodinase activity and the mono-deiodination of T₄ (30). Some studies have found that there is a reciprocal relationship between FT₄/BMI and Log₁₀ FT₄/BMI and that there is a direct relationship between maternal free T₃ (FT₃) levels or Log₁₀FT₃ and BMI in pregnant women (31). Specifically, when the level of FT₄ decreased, the level of FT₃ and BMI correspondingly increased, and obesity increased the chances of developing GDM. Furthermore, another study found that the ratio between FT₃/FT₄ levels and BMI increased when the level of FT₄ was low in euthyroid pregnancies, suggesting an increase in peripheral deiodinase activity (32). Studies have also shown that the rate of peripheral transformation of T₄ to T₃ increases with excessive energy intake, suggesting that peripheral deiodinase activity is affected by energy intake (33). All of the above findings show that a low FT₄ level is associated with the occurrence of diabetes.

Studies of general populations have shown that several mechanisms are involved in the thyroid hormone-mediated regulation of glucose metabolism. Specifically, thyroid hormones can 1) reduce the half-life of insulin, accelerate the rate at which insulin is degraded, and increase the release of the inactive precursors of insulin; 2) promote hepatic glucose output by increasing the expression of glucose transporter 2 in liver cell membranes (34); and 3) activate β-adrenergic receptors via cAMP, which increases the sensitivity of catecholamines, powerful hormones that accelerate glycogenolysis (35). Studies have also shown that many of the pathways between the hypothalamic-pituitary axis and the T₃ receptor in thyroid cells are abnormal in patients with diabetes. Therefore, a variety of separate associations have been identified between thyroid function and glucose metabolism, and these provide common pathways through which thyroid dysfunction and diabetes can develop. However, there are currently no reports describing the effect of maternal thyroid hormones on the regulation of gestational glucose metabolism.

In accordance with the findings in our study, it has previously been shown that patients with either hypothyroidism or subclinical hypothyroidism can exhibit insulin resistance (36). In vivo and in vitro studies have shown that this resistance may be caused by a reduced need for insulin or glucose utilization in peripheral tissues that have been damaged by insulin (36, 37). The Fremantle Diabetes Study also found that serum TSH levels and insulin sensitivity were significantly negatively correlated with type 2 diabetes (38). In addition, Kapadia et al reported that insulin resistance indices were significantly higher in patients with subclinical hypothyroidism than that in a control group (39).

There are several advantages to our study. First, we excluded patients with a variety of thyroiditis and immune system diseases, and patients with diabetes caused by immune disorders were also excluded because of previous reports showing that autoimmune thyroiditis is closely associated with type 1 diabetes. Second, our study excluded patients who were taking hormone drugs, either before or during pregnancy, that could affect thyroid function in addition to patients who were receiving thyroid hormone replacement therapy. This allowed us to explore correlations between thyroid hormone levels in early pregnancy and GDM in a natural setting. However, this study did not examine FT₃ levels during early pregnancy. Because this is a retrospective study, this index cannot be included from the analysis, and we were therefore required to explain the relevant phenomenon we observed based on the findings of other researchers.

In summary, this study provides new evidence showing that low thyroid hormone levels increase the risk of developing GDM in early pregnancy. A low FT₄ level during early pregnancy was reciprocally associated with BMI, which to a certain extent explains why low FT₄ levels were found to contribute to the risk of developing diabetes. Nonetheless, because the direct biological activity of FT₃ was not measured in this study, we could fully explain this phenomenon, and further study is therefore needed. Our study results show that FT₄ is associated with GDM in early pregnancy. In the future, screenings aimed at determining thyroid function during early pregnancy should combine evaluations of TSH and FT₄.

Acknowledgments

We thank our patients as well as the doctors and nurses who assisted with recruiting patients into this study.

This work was supported by the National Natural Science Foundation of China Grant 81471516, the Shanghai Municipal Science and Technology Commission Medical Guide Project 134119a1100, and Shanghai Municipal Commission of Health and Family Planning Grants 15GWZK0701 and 201540084.

Disclosure Summary: The authors have nothing to disclose.

Abbreviations

 
• BMI

  body mass index

 
• 95% CI

  95% confidence interval

 
• FT₃

  free T₃

 
• FT₄

  free T₄

 
• GDM

  gestational diabetes mellitus

 
• OGTT

  oral glucose tolerance test

 
• OR

  odds ratio

 
• TPOAb

  thyroid-peroxidase antibody.

References

Author notes