Hyperthyroidism as a Potential Risk Factor for Placental Abruption: A Case Control Study


 BackgroundOvert hyperthyroidism and hypothyroidism are associated with pregnancy complications; however, most women with these conditions are diagnosed before conception and are under treatment during pregnancy, especially in the developed countries. The purpose of this study was to investigate pregnancy complications among these women.MethodsA retrospective case-control study was conducted, and data on 3824 pregnant women who gave birth at Nagoya University Hospital located in Japan from 2005 to 2014 was collected. The pregnancy outcomes were divided and compared among three groups: the control group (n = 3709), the hyperthyroidism group (n = 52) and the hypothyroidism group (n = 63). Risk factors for placental abruption were also evaluated in singleton pregnancies (n = 3588) by multivariable logistic regression analysis. Moreover, in hyperthyroidism, thyroid function was also compared between successful and failed placentation group, and the latter group included placental abruption and preeclampsia. ResultsThe incidence of placental abruption was significantly higher in hyperthyroidism than in control and hypothyroidism groups (p < 0.01). Hyperthyroidism was independently associated with an increased risk of placental abruption (adjusted odds ratio = 12.52, 95% confidence interval = 2.91–53.88). Thyroid stimulating hormone (TSH) was significantly lower in failed placentation group than in successful placentation group (p < 0.05). ConclusionAccording to the results of our study, pregnancy outcomes in women with treated hypothyroidism were comparable with those in women without thyroid disease. Conversely, women with treated hyperthyroidism showed an independent risk of placental abruption, which might be related with lower TSH level at early gestation. However, further research is required to validate our findings.


Abstract Background
Overt hyperthyroidism and hypothyroidism are associated with pregnancy complications; however, most women with these conditions are diagnosed before conception and are under treatment during pregnancy, especially in the developed countries. The purpose of this study was to investigate pregnancy complications among these women.

Methods
A retrospective case-control study was conducted, and data on 3824 pregnant women who gave birth at Nagoya University Hospital located in Japan from 2005 to 2014 was collected. The pregnancy outcomes were divided and compared among three groups: the control group (n = 3709), the hyperthyroidism group (n = 52) and the hypothyroidism group (n = 63). Risk factors for placental abruption were also evaluated in singleton pregnancies (n = 3588) by multivariable logistic regression analysis. Moreover, in hyperthyroidism, thyroid function was also compared between successful and failed placentation group, and the latter group included placental abruption and preeclampsia.

Results
The incidence of placental abruption was signi cantly higher in hyperthyroidism than in control and hypothyroidism groups (p < 0.01). Hyperthyroidism was independently associated with an increased risk of placental abruption (adjusted odds ratio = 12.52, 95% con dence interval = 2.91-53.88). Thyroid stimulating hormone (TSH) was signi cantly lower in failed placentation group than in successful placentation group (p < 0.05).

Conclusion
According to the results of our study, pregnancy outcomes in women with treated hypothyroidism were comparable with those in women without thyroid disease. Conversely, women with treated hyperthyroidism showed an independent risk of placental abruption, which might be related with lower TSH level at early gestation. However, further research is required to validate our ndings.
In the present study, we investigated the pregnancy outcomes of women with thyroid disease who had been diagnosed and treated prior to their pregnancy: birth weight, caesarean section, Apgar scores and the obstetric complications including preterm labour, preeclampsia, low birth weight, premature rupture of membranes (PROM), placenta previa, placenta abruption and preterm birth. We also focused on the placental abruption, which is a main cause of maternal and neonatal mortality and morbidity.

Study participants
The study samples were retrospectively recruited from medical records at a single tertiary care institute, Nagoya University Hospital in Japan. The inclusion criteria for this study was women who gave birth at Nagoya University Hospital located in Japan from 2005 to 2014. Patients who treated thyroid diseases at other hospitals were also included when their treatment information from the hospital were provided. The exclusion criteria for this study was as follows: less than 22 weeks of gestation at birth, and incomplete medical records. Subjects were divided into three groups, namely, the control group: those without thyroid disease (n = 3709); the hyperthyroidism group: those who were diagnosed and treated for hyperthyroidism (n = 52); and the hypothyroidism group: those who were diagnosed and treated for hypothyroidism (n = 63). In the hyperthyroidism, 23 patients (44.2%) were treated by oral propylthiouracil, and 12 patients (23.1%) were treated by oral methimazole. All of subjects in the control group were not screened for thyroid function before pregnancy.

Clinical Information
Demographic data included maternal age, information on gravidity and parity, method of conception, and history of chronic maternal disease. In-vitro fertilization and embryo transfer and intracytoplasmic sperm injection were de ned as assisted reproductive technology (ART), and ovarian stimulation methods and intrauterine insemination (IUI) were de ned as other treatment. The clinical information on number of babies, gestational age at delivery, birth weight, foetal position at delivery, mode of delivery, Apgar scores at both 1 min and 5 min, and pregnancy complications including preterm labour, preeclampsia, PROM, placenta abruption, placenta previa, and preterm birth was also collected as maternal and neonatal outcomes. Birth weight was categorised into three groups: <1500 g; 1500-2500 g; and ≥2500 g and Apgar scores were divided into two groups: ≤6; ≥7.
Preeclampsia was de ned as persistent hypertension (> 140/90 mm Hg) and proteinuria (> 300 mg/24 h) after 20 weeks of gestation. Preterm birth was de ned as delivery between 22 and 36 weeks of gestation. Gestational ages were calculated based on the last menstrual period, and subsequently con rmed by crown-rump length, when the fertilisation date was unknown.

Sub-analysis
To determine whether hyperthyroidism was a risk factor of placental abruption, multiple pregnancies were excluded from study population for multiple logistic regression analysis. Additionally, 11 women were also excluded from the control group for the lack of data. Overall, the analysis was performed in singleton pregnancies including the control (n = 3540) and the hyperthyroidism (n = 48). Univariate and multiple logistic regression analysis were performed to estimate the ORs and 95% CI of placental abruption for binary prognostic variables including hyperthyroidism as follows: maternal age, parity, method of conception, gestational age at birth, birth weight, hypertension, diabetes mellitus, psychiatric disease, preterm labour, preeclampsia, premature rapture of membranes, and placenta previa.
The hyperthyroidism group was further divided into the two subgroups; the failed placentation group: those with preeclampsia or placental abruption and the successful placentation group: those who were not complicated with preeclampsia or placental abruption. To assess the effect of thyroid function including stimulating hormone (TSH), free T 3 (FT 3 ; triiodothyronine) and free T 4 (FT 4 ; thyroxine) and thyroid stimulating hormone receptor autoantibody (TRAb) positivity on placentation, these were compared in that two subgroups. However, these data were partially lacked in the part of the hyperthyroidism group, because they were treated at other institutes during pregnancy. In the hyperthyroidism group, 34 patients (65.4%) had data of TSH, FT 3 and FT 4 at the early gestation; from conception to 14 weeks of gestation, because it is thought to be the period of placental development. For the data of TRAb during the pregnancy, 36 patients (69.2%) had the results in the hyperthyroidism group.

Statistical analysis
All statistical analyses were performed using the SAS (version 9.4, SAS Institute Inc., Cary, North Carolina, USA) software program and JMP pro 14 (JMP, Tokyo, Japan). Continuous variables are presented as median (range) and p values were calculated by Kruskal-Wallis test. Categorical variables are presented as number (percentage) and p values were calculated by Fisher's exact test or chi-square test as appropriate. A p value of < 0.05 was considered to be signi cant.

Results
The demographic data of the control (n = 3709), hyperthyroidism (n = 52) and hypothyroidism (n = 63) groups are shown in Table 1. The percentage of ART in the hypothyroidism group was higher than those of the control and hyperthyroidism groups (p = 0.019). There were no signi cant differences in other characteristics like maternal age, gravidity, parity, and chronic maternal diseases among the control, hyperthyroidism, and hypothyroidism groups. Pregnancy outcomes were compared among the control, hyperthyroidism, and hypothyroidism groups (Table 2) and no signi cant differences were observed in the birth weight, foetal position at delivery, and mode of delivery. However, the incidence of placental abruption in the hyperthyroidism group was signi cantly higher than in the control and hypothyroidism groups (p < 0.01). The hyperthyroidism group also showed a higher incidence of preeclampsia as compared to the control and hypothyroidism groups, although the difference was not signi cant (p = 0.063). The number of cases with Apgar scores of ≤ 6 at 1 min and 5 min were signi cantly different among the control, hyperthyroidism, and hypothyroidism groups (p = 0.010 and p = 0.013, respectively), and those in the control were higher than the hyper-and hypothyroidism groups. The differences in the incidences of preterm labour, PROM, low birth weight, and placenta previa among the control, hyperthyroidism, and hypothyroidism groups were not signi cant. The incidence of preterm birth in the hypothyroidism group was lower than that in the control and hyperthyroidism groups, but this again was not signi cant (p = 0.064). To determine whether hyperthyroidism was a risk factor of placental abruption, other valuables were examined among singleton pregnancies (n = 3588, Fig. 1 (Fig. 1. upper panel). While gestational age ≥ 37 weeks (crude OR = 0.09, 95% CI: 0.03-0.22), birth weight ≥ 1500 g (1500-2500 g and ≥ 2500 g; crude OR = 0.22 and 0.06, 95% CI: 0.07-0.71 and 0.02-0.16, respectively) decreased placental abruption ( Fig. 1. upper panel). After multivariate analysis, hyperthyroidism (adjusted OR, aOR = 12.52, 95% CI: 2.91-53.88) was found to be independently associated with an increased risk of placental abruption, conversely, gestational age ≥ 37 weeks (aOR = 0.12, 95% CI: 0.04-0.40) was associated with a decreased risk ( Fig. 1. lower panel).
In the hyperthyroidism subgroup, serum TSH levels in the failed placentation group were signi cantly lower than those in the successful placentation group (p = 0.043, Fig. 2A). Additionally, although fT 3 and fT 4 were higher in the failed group, the difference was not signi cant ( Fig. 2B -C). In comparison of TRAb positivity, no signi cant difference was detected between the two groups (successful placentation − 16/28; 57.1%, failed placentation − 6/8; 75.0%, p = 0.441).

Discussion
The present study revealed that hyperthyroidism had a signi cantly higher incidence of placental abruption and a trend for increased preeclampsia compared to the control and the hypothyroidism.
Moreover, hyperthyroidism was shown to be an independent risk factor for placental abruption by multivariate analyses.
Previous reports on population-based and case-control studies demonstrated hyperthyroidism to be associated with a high risk of preeclampsia [5][6][7][8] and placental abruption [9]. The impact of interventions for hyperthyroidism before and during pregnancy on pregnancy outcomes remains uncertain [10]. Although one study had reported a higher occurrence of adverse pregnancy complications in an untreated hyperthyroidism group compared to a treated group [11], the study population was small and placental abruption was not evaluated. They also reported a higher occurrence of hypertensive disorders of pregnancy was seen in the treated group compared to control, which was consistent with the result in this study, because preeclampsia is included in hypertensive disorders of pregnancy.
Additionally, the present study revealed that the maternal and neonatal outcomes in women with treated hypothyroidism were comparable to those in women without thyroid disease. Both overt hyperthyroidism and hypothyroidism are well-known to have adverse effects on the mother and her child [12]. ART is known to be associated with increased adverse pregnancy outcomes [13]. In this study, despite a higher prevalence of ART, cases in the hypothyroidism group were not associated with a higher incidence of adverse pregnancy outcomes as compared to controls. This result is consistent with that of a previous report [3], and suggests that hypothyroidism treatment could decrease the risk of adverse pregnancy outcomes seen in overt hypothyroidism. In this study, the hypothyroidism group also had a lower incidence of preterm birth, although this nding was not statistically signi cant.
Moreover, multivariate analysis rstly demonstrated that hyperthyroidism, even though they were treated by endocrinologists, was an independent risk factor for placental abruption in the present study, although several retrospective cohort studies have also reported the association of placental abruption with hyperthyroidism [9,14]. However, whether treatment for hyperthyroidism could in uence the risk of placental abruption remains unknown, because the present study could not include population of untreated hyperthyroidism. As randomised controlled trials was lacked, the impact of antithyroid interventions for hyperthyroidism pre-pregnancy or during pregnancy on important pregnancy outcomes remains unknown [9]. Additionally, GA ≥ 37 weeks independently decreased the aOR of placental abruption. This indicates that preterm birth (GA < 37 weeks) was also an independent risk factor of placental abruption, which is consistent with previous reports [15,16].
As one of pathological mechanisms in the hyperthyroidism group, TSH levels at the early gestation were signi cantly lower in the subgroup with preeclampsia or placental abruption than the subgroup without preeclampsia or placental abruption. Recently, preeclampsia and placental abruption were reported to be associated with failed placentation due to ine cient invasion of trophoblasts after implantation [17]. Human chorionic gonadotropin from trophoblast stimulates thyroid hormones at early gestation. Thyroid hormones play a role in placentation, which is completed at almost 20 weeks of gestation [18], and T 3 is known to play a critical role in placentation by increasing trophoblasts invasion [19]. These ndings led us to a hypothesis that thyroid function at early gestation would cause failed placentation related disorders in treated hyperthyroidism. Serum FT 3 and FT 4 levels at early gestation in failed placentation group was higher than those in successful group, but the difference was not signi cant. These thyroid status at early gestation in failed placentation group might be similar to subclinical hyperthyroidism. The previous studies reported the association of placental abruption with subclinical hyperthyroidism at early gestation [14,20], which was consistent with the present study. Thus, these results suggested that subclinical hyperthyroidism at early gestation, in spite of treatment prior to or during pregnancy, would cause preeclampsia or placental abruption, although further investigation is required.

Limitations
The main limitations of this study consisted of the study design being a cohort study and the inclusion of a single tertiary care institution. Prevalence of low Apgar scores and higher preterm births in the control group compared to the thyroid disease groups, suggests the study population belonged to a high-risk category. This could cause a bias in detecting a signi cant difference in adverse pregnancy outcomes. However, the multivariate analysis on the risk of placental abruption should eliminate some bias by accounting for confounding factors. Secondly, all women in the control group were not screened for thyroid dysfunction. A small but signi cant proportion of these women could have undiagnosed subclinical thyroid dysfunction, which may in uence the results of the study. Subclinical hypothyroidism was reported to be associated with preterm birth [7], and these patients may have included in the control group. In addition, the information on the time to start treatment, the duration of treatment, and treatment adherence was also lacked, because a part of patients was treated at other hospitals before pregnancy and clinical information before pregnancy were limited. A few patients started to treatment during pregnancy, and the effect of the timing to start treatment might be minimal. Moreover, the sample size for the comparison between failed and successful placentation groups in hyperthyroidism was small. Thus, adequate TSH levels for successful placentation remains undetermined. Further prospective research with a large study population should be conducted on the effect of the intervention prior to pregnancy on pregnancy outcomes to determine the adequate management of women with hyperthyroidism who wish to pregnancy.

Conclusion
The risks of pregnancy outcomes among women with treated hypothyroidism are comparable with those in women without thyroid diseases. However, treated hyperthyroidism showed an independent risk of placental abruption. In hyperthyroidism, low serum TSH concentration at early gestation might be related to placental abruption, although further research is required in this regard.
Abbreviations aOR, adjusted odds ratio; CI, con dence intervals; FT 3, free T 3 / triiodothyronine; FT 4, free T 4 /thyroxine; GA, gestational age; OR, odds ratio; PROM, premature rupture of membrane; TSH, thyroid function including stimulating hormone. Risk factors for placental abruption of 3588 singleton pregnancies. Reference values for odds ratio of 1.00 are as follows. Maternal age; < 25 years, Parity; nulliparity, Conception; natural, Complication; subjects without complication, Gestational age (GA) at birth; < 37 weeks, Birth weight; < 1,500 g. Circles represent mean odds ratios with bars indicating 95% con dence intervals (CIs). GA at birth, preterm labour, preeclampsia, hyperthyroidism and birth weight showed signi cant differences by univariate analyses. GA at birth and hyperthyroidism also showed signi cant differences by multivariate analyses. OR, odds ratio; CI, con dence interval; IUI, intrauterine insemination; ART, assisted reproductive technology; GA, gestational age; DM, diabetes mellitus; PROM; premature rupture of membranes.

Figure 2
The thyroid function at early gestation was compared between successful and failed placentation group.