Several population studies have reported various obstetrical complications for hyperthyroid mothers. These studies have primarily been performed in the Nordic countries, which are considered to have a rather homogenous population and evaluated only a small number of complications. Our study, which is a large study of the American population, demonstrates that many pregnancy and delivery complications are increased in women with hyperthyroidism, from diverse ethnic backgrounds. Our study is unique in reporting the rates of intrapartum and postpartum conditions such as hypertensive disorders, venous thromboembolism, maternal infection, and wound complication, which have not previously been evaluated in the other published, population-sized studies. It is also unique given the size of the hyperthyroid population studied, and that it is a population-based study for which the women were not of Scandinavian origin.
Preeclampsia and Preterm Delivery
In a case–control study from 2014, Aggarawal et al. compared 208 cases and 403 controls. They found that hyperthyroidism was associated with increased risks of preeclampsia, preterm delivery, cesarean section, intrauterine growth restriction, low birth weight, and stillbirth. They also found that uncontrolled disease had worse outcomes as compared with women with medically managed hyperthyroidism. Lastly, they discovered that pregestational hyperthyroidism had better outcomes when compared with those initially diagnosed during pregnancy. Similar to Aggarawal et al., Millar et al. reported higher rates of preeclampsia and low birth weight in the offspring of poorly controlled hyperthyroid mothers, in a study of 181 subjects. Another recent study, which was conducted in Thailand and included 408 pregnant women with thyrotoxicosis, demonstrated increased complications, although it did not evaluate the degree of control of the disease. In our study, we demonstrated that hypertensive disorders during pregnancy were increased 20% in the hyperthyroid group when compared with the nonhyperthyroid controls. We demonstrated an increased risk for preterm premature rupture of membranes and preterm delivery in the hyperthyroid group. These findings further contribute to the results from another population-based study, which was conducted in Denmark. The Danish study reports an increased risk for preterm delivery in either hyper- or hypothyroidism mothers. The number of cases in this study who were first diagnosed with hyperthyroidism before or during pregnancy was 5643. In this study, the risks were also increased in women subsequently diagnosed with hyperthyroidism for up to 2 years after delivery.
Pregestational and Gestational Diabetes
Despite higher rates of pregestational diabetes and obesity for hyperthyroid mothers, we did not find a statistically significant difference in gestational diabetes in women with hyperthyroidism, when controlling for these factors. This finding is similar to the finding from a study conducted in Finland, of 2144 women with hyperthyroidism. They found that hyperthyroid mothers were more likely to be obese without an increased risk of gestational diabetes. However, it should be noted that when not controlling for the confounding factors women with hyperthyroidism were at increased risk of gestational diabetes with an OR of 1.41. Probably, this is partially attributable to the increased rate of obesity and its effect on insulin resistance. As such, the associated factors seen with hyperthyroidism may likely play a role in increasing risks of gestational diabetes in this population.
Small for Gestational age, Congenital Anomalies, and Stillbirth
In our study we found that maternal hyperthyroidism was associated with small for gestational age neonates, as was found in a previous study from Denmark. Regarding congenital anomalies, the literature suggests that birth defects may be increased by antithyroid medications,[5,20] even for the propylthiouracil, which was previously considered safer than methimazole.[16,20] However, in a recent meta-analysis, the risk of birth defects was attenuated, with adjusted risk ratio of 1.16 (95% CI: 1.08–1.25) for propylthiouracil, which was similar to 1.15 adjusted risk ratio (95% CI: 1.02–1.29) for untreated hyperthyroidism. In our study, we did not demonstrate an increased risk of congenital anomalies that was associated with hyperthyroidism. It should be noted that our database does not include information regarding treatment method, dosage, or duration. Therefore, the risk of congenital anomalies may be affected. On the other hand, the study from Finland, which was conducted by Turunen et al. did not find an association between hyperthyroidism and birth defects, even among patients who were taking anti-thyroid medications. As such, a definite conclusion cannot be drawn. Due to the conflicting results, it would be prudent to be cautious while considering the use of antithyroid medications during pregnancy, particularly as according to some studies the main risk factor for birth defects is the use of such medications. Further studies are needed to address this issue.
Another important outcome that we have investigated was the rate of stillbirth. Hamburger, in his review from 1992, summarized 11 older studies and noted a 5.6% stillbirth rate (14 cases out of 249 pregnancies). This rate is substantially higher than our findings with a stillbirth rate of 0.7%. Andersen et al., in their study from 2014, found that the rate for stillbirth in hyperthyroid Danish mothers with singleton pregnancies ranged between 0.25% and 0.68% (depending on the time of diagnosis). This reported rate of stillbirth closely resembled ours and the discrepancy between the summary from 1992 and our study may be attributed to improvements in prenatal care and medical management over time.
In our study, we found an association between hyperthyroidism and placenta previa. We can hypothesize that such association may be related to the role of thyroid hormone during implantation, which may also play a role in the subsequent development of preeclampsia or intrauterine growth restriction. We also noted increased risks of wound complications, blood transfusion, and venous thromboembolism, which may partially be explained by the underlying disease,[24,25] as patients with hyperthyroidism had higher levels of coagulation factors.
All our findings, as other findings that have been reported in the literature, may be attributed to hyperthyroidism per se, to the presence of maternal autoantibodies, or the antithyroid medications.[5,11] Treating hyperthyroidism during pregnancy is considered complicated due to the teratogenic effects of both antithyroid medications methimazole and propylthiouracil, and hepatotoxicity.
Our study has several limitations. The first is due to its retrospective design. Another limitation is the absence of laboratory findings or information regarding treatment. Given that we do not have access to serum TSH levels in pregnancy, we cannot be sure that subjects were compliant with their medications, and given that the database does not list different medications taken, we cannot control for the effect of different antithyroid medication types on outcomes. However, the results of this study are consistent with previous studies in the literature. Further limitations are that gestational thyrotoxicosis in the first trimester could not be excluded as a cause of the diagnosis of hyperthyroidism. Gestational thyrotoxicosis is due to increased human chorionic gonadotrophin levels, is usually mild or subclinical, and transient. It does not need antithyroid drugs and it has not been associated with adverse outcomes for the baby in several studies.[5,28] We cannot determine, due to the nature of the database, how many pregnant women had this form of hyperthyroidism in our group. However, it should be noted that in published studies able to exclude gestation thyrotoxicosis, results were very similar to ours and as such we do not expect that this was a major issue in our database. A final limitation is that the control population is not necessarily healthy and it likely has other diseases that may affect pregnancy outcomes.
The major strength of our study is the large sample size over the span of 10 years and the comparison of outcomes with nonhyperthyroid controls with similar age and background diseases that are potential confounders. Another strength of our study is that we analyzed an American database and, as such, it represents a different population type and more likely a more diverse group of subjects than evaluated in the previously published population-based studies.
Clin Endocrinol. 2022;97(3):347-354. © 2022 Blackwell Publishing