Introduction

Endometrial cancer (EC) is the most common gynecological malignancy in industrialized countries1. Notably, its incidence has risen in recent years among women in their 30s2. The precursor lesion for the predominant endometrioid adenocarcinoma subtype is endometrial hyperplasia (EH)1.

Excessive estrogen exposure without the protective counteracting effect of progestin is a significant risk factor for EH and EC3. Factors contributing to excessive estrogen exposure include obesity, chronic anovulation, early menarche, late menopause, and estrogen-secreting tumors1.

In 2014, the World Health Organization classified EH based on cellular atypia into two types: Endometrial hyperplasia without atypia and endometrial atypical hyperplasia4. The cancer risk of EH without atypia is 1–2% but can reach up to 60% for atypical cases of endometrial intraepithelial neoplasia (EIN)5. The definitive treatment for EIN is total hysterectomy, which causes loss of fertility. However, hormonal therapy has emerged as an alternative for young women who have not yet completed their fertility plans6. With the increasing diagnosis of EH in younger women and the proven efficacy of hormonal treatments, more women are opting to preserve their fertility and to attempt pregnancy later in life7.

Even after EH shows a complete response to hormonal therapy, underlying factors causing hormonal imbalances, such as obesity, may persist in women diagnosed with EH. Previous studies have predominantly focused on the risk of progression to EC or the pregnancy success rate in women diagnosed with EH8,9.

Therefore, in this study, we aimed to analyze pregnancy outcomes in women diagnosed with EH who are likely to become pregnant in the future.

Methods

Data characteristics

The Korean National Health Insurance (KNHI) claims database encompasses health information for the majority of Koreans, except for 3% of the population covered by the Medical Aid Program. This comprehensive database includes data on women’s health conditions, particularly obstetric and gynecological diagnoses, identified using the International Classification of Diseases, Tenth Edition (ICD-10) codes, except for uninsured procedures, such as plastic surgeries.

Dataset

The study analyzed the KNHI claims data from 2006 to 2023. This study was approved by the Institutional Review Boards of the Catholic Medical Centre (OC23ZISI0058). Because this was a retrospective cohort study and because all data were anonymized, the need for informed consent was waived by the Institutional Review Board of The Catholic University of Korea. This study was conducted in accordance with the guidelines of the Declaration of Helsinki, and the rights of all patients were protected. We included women diagnosed with EH before pregnancy and excluded those with missing data. The study population consisted of 1 778 529 women who gave birth from 2006 to 2023. After applying the exclusion criteria, 199 143 women were included in the study (Fig. 1). Of these, 1655 (0.83%) had EH with or without atypia and 197 488 (99.17%) women were in the non-EH group.

Fig. 1
figure 1

Participant flow chart of the total population.

Full size image

The baseline characteristics are shown in Table 1.. Obstetric diagnoses were identified using specific ICD-10 codes: Pregnancy associated hypertension (PAH; O13–O16), gestational diabetes mellitus (GDM; O24.4, O24.9), placenta previa (O44), placenta abruption (O45), placenta accreta spectrum (O432, O720, O73), oligohydramnios (O41), polyhydramnios (O40), postpartum hemorrhage (O72), preterm premature rupture of membranes (PPROM; O42), intrauterine growth restriction (IUGR; O365) and fetal death in uterus (FDIU) (O364). Neonatal outcomes, including preterm birth and birth weight, were extracted from the National Health Screening Program for Infants and Children database. PAH included gestational hypertension, pre-eclampsia, eclampsia, and superimposed pre-eclampsia. Preterm birth was defined as delivery before 37 weeks of gestation.

Table 1 Baseline characteristic of population.
Full size table

The diagnostic criteria for obstetrics and gynecological conditions in Korea are based on the standards outlined in Williams Obstetrics.

Statistical analysis

Continuous variables were expressed as mean ± standard deviation, and categorical variables as percentages. Clinical characteristics were compared using analysis of variance for continuous variables and the chi-squared test for categorical variables. A multivariate logistic regression analysis model was used to estimate the adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for adverse pregnancy outcomes. Maternal age (≥ 35years), BMI (≥25kg/m2), maternal disease (high blood pressure, diabetes mellitus, thyroid disease), previous uterine surgery (myomectomy, hysteroscopy, dilatation and curettage) were adjusted as potential confounders.

Statistical analyses were performed using statistical analysis system software for Windows (version 9.4, SAS Inc., Cary, NC, USA). The study protocol was approved by the Institutional Review Board of the Catholic University of Korea at Incheon St. Mary’s Hospital (approval no. OC23ZISI0058).

Results

Baseline characteristics of the study population

The EH group exhibited significantly higher maternal age and body mass index (BMI; P < 0.05) compared with the non-EH group. Additionally, the proportion of women with maternal age ≥ 35 years and BMI ≥ 25 kg/m2 was significantly greater in the EH group versus the non-EH group (P <0.001). The prevalence of pre-existing maternal diseases, including polycystic ovary syndrome (PCOS), hypertension, diabetes, thyroid disease, and autoimmune disease was significantly higher in the EH group than in the non-EH group (P < 0.001). The EH group showed a significantly higher rate of previous uterine surgery compared to the non-EH group.

Obstetric outcomes according to maternal EH

Data on obstetric outcomes are presented in Table 2.. The EH group had significantly higher rates of pregnancy-associated hypertension and GDM, compared with the non-EH group (P < 0.001). Additionally, the EH group showed significantly increased incidences of threatened abortion, preterm labor, PPROM, polyhydramnios, oligohydramnios, placenta previa, and placenta accreta, compared with the non-EH group (P < 0.001). Rates of IUGR and FDIU were also significantly elevated in the EH group than in the non-EH group (P < 0.001). Furthermore, the EH group also experienced higher rates of postpartum hemorrhage than the non-EH group (P < 0.001), whereas no significant intergroup differences were observed in the frequencies of placental abruption and uterine rupture.

Table 2 Obstetric outcomes according to maternal EH.
Full size table

Multivariate analysis of obstetric outcomes according to EH

Table 3. shows the risk of adverse obstetric outcomes associated with EH. In the unadjusted analysis, compared with the non-EH group, the EH group showed significantly higher risks for PAH (OR 2.182, 95% CI 1.792–2.656), GDM (OR 2.102, 95% CI 1.876–2.34), placenta previa (OR 2.650, 95% CI 1.951–3.599), placenta accreta (OR 4.795, 95% CI 2.855–8.052), preterm labor (OR 2.416, 95% CI 2.154–2.711), IUGR (OR 2.184, 95% CI 1.698–2.810), threatened abortion (OR 2.244, 95% CI 1.988–2.534), PPROM (OR 3.389, 95% CI 2.745–4.185), polyhydramnios (OR 2.541, 95% CI 1.49–4.323), oligohydramnios (OR 1.898, 95% CI 1.473–2.446), postpartum hemorrhage (OR 1.463, 95% CI 1.209–1.770) and FDIU (OR 3.345, 95% CI 2.334–4.796). In the adjusted analysis that controlled for maternal age, compared with the non-EH group, the EH group continued to demonstrate significantly higher risks for PAH (OR 1.675, 95% CI 1.374–1.2.042), GDM (OR 1.508, 95% CI 1.342–1.93), placenta previa (OR 1.707, 95% CI 1.253–2.325), placenta accreta (OR 3.020, 95% CI 1.790–5.095), preterm labor (OR 1.805, 95% CI 1.605–2.030), IUGR (OR 1.619, 95% CI 1.256–2.087), threatened abortion (OR 1.769, 95% CI 1.604–1.952), PPROM (OR 2.542, 95% CI 2.055–3.145), polyhydramnios (OR 1.721, 95% CI 1.009–2.938), oligohydramnios (OR 1.674, 95% CI 1.298–2.160), postpartum hemorrhage (OR 1.405, 95% CI 1.161–1.701) and FDIU (OR 2.383, 95% CI 1.657–3.425). In an adjusted analysis that controlled for maternal age and BMI, the EH group, compared with the non-EH group, showed a significantly higher risks for PAH (OR 1.459, 95% CI 1.192–1.786), GDM (OR 1.470, 95% CI 1.308–1.652), placenta previa (OR 1.730, 95% CI 1.270–2.357), placenta accreta (OR 3.014, 95% CI 1.785–5.088), preterm labor (OR 1.828, 95% CI 1.626–2.056), IUGR (OR 1.652, 95% CI 1.282–2.130), threatened abortion (OR 1.766, 95% CI 1.600–1.948), PPROM (OR 2.542, 95% CI 2.054–3.146), oligohydramnios (OR 1.660, 95% CI 1.286–2.141), postpartum hemorrhage (OR 1.404, 95% CI 1.160–1.700) and FDIU (OR 2.335, 95% CI 1.623–3.358). In the adjusted analysis that controlled for maternal age, BMI, hypertension and diabetes, the EH group, compared with the non-EH group, showed significantly higher risks for GDM (OR 1.249, 95% CI 1.107–1.409), placenta previa (OR 1.665, 95% CI 1.221–2.271), placenta accreta (OR 2.610, 95% CI 1.539–4.426), preterm labor (OR 1.664, 95% CI 1.478–1.874), IUGR (OR 1.473, 95% CI 1.140–1.902), threatened abortion (OR 1.711, 95% CI 1.550–1.888), PPROM (OR 2.325, 95% CI 1.876–2.881), oligohydramnios (OR 1.573, 95% CI 1.218–2.031), postpartum hemorrhage (OR 1.358, 95% CI 1.121–1.645) and FDIU (OR 2.112, 95% CI 1.465–3.046). In the adjusted analysis that controlled for maternal age, BMI, hypertension, diabetes, hyperthyroidism and hypothyroidism, the EH group, compared with the non-EH group, demonstrated significantly higher risks for GDM (OR 1.196, 95% CI 1.060–1.350), placenta previa (OR 1.610, 95% CI 1.180–2.197), placenta accreta (OR 2.361, 95% CI 1.390–4.009), preterm labor (OR 1.542, 95% CI 1.368–1.738), IUGR (OR 1.404, 95% CI 1.086–1.814), threatened abortion (OR 1.669, 95% CI 1.512–1.842), PPROM (OR 2.178, 95% CI 1.756–2.701), oligohydramnios (OR 1.543, 95% CI 1.195–1.994), postpartum hemorrhage (OR 1.343, 95% CI 1.109–1.627) and FDIU (OR 2.032, 95% CI 1.408–2.933). In the adjusted analysis that controlled for maternal age, BMI, hypertension, diabetes, hyperthyroidism and hypothyroidism and previous uterine surgery, the EH group, compared with the non-EH group, demonstrated significantly higher risks for GDM (OR 1.182, 95% CI 1.047–1.334), placenta previa (OR 1.555, 95% CI 1.139–2.122), placenta accreta (OR 2.277, 95% CI 1.341–3.865), preterm labor (OR 1.504, 95% CI 1.334–1.696), IUGR (OR 1.92, 95% CI 1.33–2.771), threatened abortion (OR 1.646, 95% CI 1.479–1.833), PPROM (OR 2.099, 95% CI 1.693–2.604), oligohydramnios (OR 1.535, 95% CI 1.181–1.984), postpartum hemorrhage (OR 1.275, 95% CI 1.052–1.545) and FDIU (OR 1.92, 95% CI 1.33–2.771).

Table 3 Multivariate analysis of obstetric outcomes according to EH.
Full size table

Discussion

This study revealed that EH is a significant risk factor for adverse obstetric outcomes. Women with EH exhibited clinical characteristics such as advanced maternal age, higher BMI, and higher incidence of PCOS, hypertension, diabetes, thyroid disorders, and autoimmune diseases. After adjusting for confounding factors, the EH group showed significantly higher rates of adverse obstetric outcomes, including GDM, placenta previa, placenta accreta, preterm labor, IUGR, threatened abortion, PPROM, oligohydramnios, postpartum hemorrhage, and FDIU, compared with the non-EH group.

Excessive estrogen stimulation of the endometrium, without the protective effects of progestins, induces proliferative glandular epithelial changes leading to hyperplasia. Moreover, many women receiving gynecological treatment for EH also have other underlying conditions, such as obesity, PCOS, hypertension, and type 2 diabetes10,11,12,13,14.

A systematic review and meta-analysis evaluating fertility in individuals with EH following progesterone therapy reported pregnancy rates ranging from 26.3 to 41.0%15,16. As many patients with EH also experience conditions that may result in chronic anovulation (e.g., polycystic ovary syndrome, obesity, and diabetes), the live birth rate achieved through assisted reproductive technology is higher than that achieved through spontaneous pregnancy (39.4% vs. 14.9%, P<0.001)17. Another study showed that the cumulative risk of progression to EC steadily increased over time in women with EH after completion of progesterone therapy (8.2% for 4 years versus 27.5% for 19 years)18. However, guidelines regarding the optimal duration for using levonorgestrel-releasing intrauterine devices or systemic progestogens to prevent long-term recurrence have not yet been established19. The American College of Obstetricians and Gynecologists recommends that medical management with progestogens be continued in patients at risk of EC after initial treatment with these drugs for the following reasons: Advanced age; late menopause; anovulation or chronic anovulation; high-risk conditions (Lynch syndrome, Cowden syndrome); modifiable risk factors, such as the use of unopposed estrogen therapy; and potentially modifiable risk factors, such as obesity and type 2 diabetes19. If these patient-specific risk factors remain unchanged, EH is likely to negatively affect obstetric outcomes, regardless of whether it regresses or not.

Maternal obesity increases the risk of complications such as miscarriage, gestational diabetes, preterm delivery, venous thromboembolism, induced labor, and caesarean section20. The effects of high-dose progestins on subsequent pregnancies remain unclear. High-dose progestins may downregulate the expression of progesterone receptors in EH, and reduced endometrial thickness (< 6 mm) is associated with decreased live birth rates21.

Recent research has determined that patients with EC or EIN experience significantly higher rates of GDM and postpartum hemorrhage. In addition, a previous study of Korean patients with EC or EIN reported significantly higher rates of twin pregnancy, preterm delivery, and caesarean section22.

Although EH with or without atypia shares risk factors with EC, previous studies have primarily focused on EC, leaving a gap in our understanding of the obstetric implications of EH.

Only a few studies have been conducted on the pregnancy outcome of endometrial hyperplasia. One propensity score matched study found that EH group has higher rate of abortion and preterm delivery23. Other studies have shown that the EH group had higher rates of induction of labor, cesarean section, and postpartum hemorrhage compared to the non-EH group24. However, research on endometrial hyperplasia and adverse obstetric outcomes is still lacking.

Our study demonstrated that not only EC but also EH with or without atypia are significant risk factors for adverse obstetric outcomes. Women with EH exhibited characteristics such as advanced maternal age, higher BMI, a greater prevalence of pre-existing disease and adverse obstetric outcomes. Notably, even after adjusting for confounding factors, differences in pregnancy outcomes persisted, except for preeclampsia and polyhydramnios. This study showed significantly higher rate of FDIU and placenta accreta in EH group than non-EH group. Previous studies have not investigated the association between EH and FDIU. However, several studies targeting the EH group showed that the rate of late pregnancy loss (13–28 weeks of pregnancy) was higher than that in the general population. Late pregnancy loss included FDIU that occurred after 20 weeks of pregnancy23,24. Another study found that EH group has a higher rate of placenta accreta than control group. They described that the EH group had a high frequency of dilatation and curettage (D&C), which may contribute to placenta accreta25. In our study, we analyzed the results by adding previous uterine surgery that may cause placental accreta (Model 5), and even after adjusting for previous uterine surgery, the EH group showed a higher risk of placental accreta than the control group. As the incidence of placenta accreta increases, the risk of postpartum hemorrhage also increases. High-dose progestin administration during EH treatment may result in thinning of endometrium, which may interfere with proper implantation and lead to placental dysfunction26,27. Thin endometrial thickness is related to adverse obstetric outcomes including threatened abortion, preterm birth, IUGR and placenta accreta25,28. These findings are consistent with those of the present study, suggesting the need for the close clinical monitoring of women with a history of EH. Nevertheless, further research is warranted to clarify these associations.

This study has some limitations. Control group was defined as individuals without any diagnostic codes for endometrial hyperplasia. However, we cannot completely exclude the possibility that some individuals in the control group may have had undiagnosed disease, which may introduce misclassification bias. Its retrospective design and the inclusion of multicenter hospitals from various regions made it challenging to standardize diagnostic tools and treatment protocols, potentially introducing detection bias. Moreover, the use of national claims data limited our access to detailed individual information regarding physical, underlying medical conditions and endometrial thickness through ultrasonography.

However, this study also has notable strengths. The large sample size, encompassing over 199 000 deliveries, and the utilization of data from the KNHI Program, which covers 97% of the Korean population, enhances the generalizability of our findings. Importantly, this study represents a large analysis of EH, addressing an area whose clinical importance has been relatively underexplored.

Our study indicates that EH is a significant risk factor for adverse obstetric outcomes. The results of this study suggest that women with EH require close obstetric surveillance, regardless of EH regression. Prospective, large-scale studies are required to further validate these hypotheses.