Pregnancies complicated by diabetes, including pre-gestational diabetes mellitus (PGDM) and gestational diabetes mellitus (GDM), is one of the common pregnancy complications, which is increasing in recent years [1-3]. PGDM increases the risk of poor maternal and perinatal outcome, such as preeclampsia, fetal and neonatal death, macrosomia, preterm delivery, growth restriction, and so on [4-6]. Diabetes has become a major public health problem and its prevalence is increasing significantly among persons who are more than 20 years of age in China [7, 8]. However, due to the low proportion of regular physical examination, most diabetic patients are undiagnosed in China. In diabetic patients who have missed diagnosis before pregnancy, the control of blood glucose level is usually unsatisfactory, which will have adverse effects on maternal and perinatal outcome. However, there appears to be lack of studies on analysis of PGDM patients who have missed diagnosis before pregnancy.

Nowadays, the rising prevalence of diabetes increases the risk of diabetes for childbearing women in China, especially pregnant women [9, 10]. The study by Chan et al. demonstrated this epidemic is the result of the rapid societal transition, placing Chinese people at high risk of diabetes and multiple morbidities, and many are undiagnosed [9]. The pregnant women who had not been confirmed with diabetes before pregnancy with unsatisfactory blood glucose level management will lead to adverse pregnancy outcomes. For instance, Corrado et al. indicated the prevalence of fetal malformation and insulin requirements with PGDM first diagnosed during pregnancy were significantly higher compared with women with GDM [11]. Considering the high rate of missed diagnosis and risk of adverse pregnancy outcome, PGDM needs to be diagnosed in a timely fashion. The PGDM at different diagnosis time may affect the pregnancy outcome of pregnant women.

In the present study, the authors analyzed the clinical characteristics of PGDM pregnant women who were diagnosed during pregnancy (missed diagnosis before pregnancy). The maternal and perinatal outcomes were compared in women with PGDM who were diagnosed at different times.

A total of 822 pregnant women who were diagnosed with PGDM were included between January 2011 and December 2017 in Yidu Central Hospital of Weifang. According to the diagnosis time of PGDM, all the participants were classified into pre-pregnancy diagnosis (PPD) group and pregnancy diagnosis (PD) group. Then the pregnancy diagnosis group was divided into group A (PGDM patients were diagnosed before 24 weeks) and group B (PGDM patients were diagnosed at 24 weeks or later). This study protocol was approved by the Medical Ethics Committee of Yidu Central Hospital of Weifang. Written informed consent was obtained from each participant.

The diagnosis of PGDM with pregnancy was based on the diagnostic criteria for GDM [12, 13]. It can be confirmed as PGDM when it conforms to either of the following two criteria: (1) a patient who has been diagnosed with diabetes before pregnancy or (2) pregnant women who are at high risk for diabetes and have not had a blood glucose examination before pregnancy should be confirmed with PGDM at the initial examination according to either of the following conditions. Firstly, fasting blood glucose level ≥ 7.0 mmol/L (126 mg/dL). Secondly, after the 75-gram oral glucose tolerance test (OGTT), the one with plasma glucose value more than 11.1 mmol/ L after two hours was considered positive. Thirdly, the one accompanied by typical hyperglycemia or hyperglycemia crisis symptoms, as well as a random blood glucose level ≥ 11.1 mmol/ L. Finally, the glycosylated hemoglobin (HbA1c) ≥ 6.5%.

In this study, the authors analyzed the following maternal and pregnancy variables: delivery age, diagnosis time of PGDM, blood glucose control level, insulin treatment, delivery gestational week, delivery mode, the history of pregnancy loss, the average of HbA1c levels, and preeclampsia. Fetal variables analyzed were: preterm birth (before 37 completed weeks), macrosomia, newborns transfer to pediatrics (NTP), and neonatal birth weight. According to these data, the authors designed the retrospective study of women diagnosed with PGDM.

The statistical analysis was carried out using the software of SPSS 21.0. Data are expressed as the mean ± standard deviation (SD). The differences between two groups were analyzed by paired Student’s t-test. The association between PGDM and pregnant complication was analyzed by using the χ² test. Differences were considered significant when p < 0.05.

In this study, the authors compared the clinical characteristics of pre-pregnancy diagnosis group and pregnancy diagnosis group, as well as among pre-pregnancy diagnosis group, and pregnancy diagnosis groups A and B. Among 822 pregnant women, 31.9% (262/822) pregnant women with PGDM were diagnosed before pregnancy, and 68.1% were diagnosed in pregnancy, of whom the rate of missed diagnosis of PGDM was 68.1%.

In pre-pregnancy diagnosis group and pregnancy diagnosis group, the delivery age of pregnant women was 32 ± 5 and 33 ± 4, the rate of pregnancy loss was 21.8% (57/262) and 17.8% (100/560), respectively. The information of delivery gestational weeks and neonatal birth weight are listed in Table 1. The clinical characteristic information, including delivery age, delivery gestational weeks, neonatal birth weight, and pregnancy loss, had no significant difference between pre-pregnancy group and pregnancy group (all p > 0.05, Table 1). Moreover, no positive difference in delivery age of pregnant women (32 ± 4, 33 ± 4, respectively), the rate of pregnancy loss (26/107, 74/453), delivery gestational weeks, and neonatal birth weight, was found between pregnancy diagnosis groups A and B (all p > 0.05, Table 1). In addition, the aforementioned clinical information in groups A and B of the pregnancy diagnosis group were compared with those in the pre-pregnancy diagnosis group, respectively, and both of the differences were not statistically significant (all p > 0.05).

In this study, insulin was used to control the blood glucose, and the level of HbA1c was detected to monitor the level of blood glucose control. Firstly, the authors compared the blood glucose control levels and pregnancy outcomes between pre-pregnancy diagnosis group and pregnancy diagnosis group. The average HbA1c level in pre-pregnancy diagnosis group was 6.3 ± 1.1%, which was higher than that in pregnancy diagnosis group. However, it had no statistically significant difference between the two groups. As shown in Table 2, the proportion of insulin treatment in pre-pregnancy diagnosis group was 89.7% (235/262), which was significantly higher than that in pregnancy diagnosis group (53.8%, 301/560, p < 0.001). The authors also compared the pregnancy outcomes between the two groups. As shown in Table 2, a statistically significant result was found in the rate of cesarean section in the pre-pregnancy group and pregnancy diagnosis group (p < 0.001). However there was no significant difference of other outcomes between the two groups, such as the rate of pre-term, macrosomia, newborns transfer to pediatrics, and preeclampsia (all p > 0.05, Table 2).

Secondly, the authors compared the related data between group A in pregnancy diagnosis group and pre-pregnancy diagnosis group. The average HbA1c level in group A was 6.4 ± 1.3%, which is higher than that in pre-pregnancy diagnosis group. However, as shown in Table 2, there was no significant difference between pre-pregnancy diagnosis group and pregnancy diagnosis group A (all p > 0.05). Then the pregnancy diagnosis group B was compared with the pre-pregnancy diagnosis group. As shown in Table 2, the average HbA1c level in group B was 6.0 ± 1.3%, and in pre-pregnancy diagnosis group it was 6.3 ± 1.1%. The blood glucose control level of group B was significantly lower than that of the pre-pregnancy diagnosis group, which had a significant difference between the two groups (p = 0.005). The comparison of insulin treatment between the two groups was also statistically significant (p < 0.001). The pregnancy outcome was also compared between the two groups. The rate of cesarean in group B (267/453, 58.9%) was lower than that in pre-pregnancy diagnosis group (190/262, 72.5%), which had a significant difference (p < 0.001). However, the pre-term rate, cesarean delivery rate, the incidence of macrosomia, preeclampsia, and NTP in the two groups were not statistically significant (all p > 0.05, Table 2).

As shown in Table 2, the blood glucose control level of pregnant women in group A was significantly higher than that in group B, and the average HbA1c level had a significant difference (p = 0.002). The proportion of pregnant women using insulin treatment in group A was significantly higher than that in group B, and the difference was statistically significant (p < 0.001). The pre-term rate, cesarean delivery rate, the incidence of macrosomia, preeclampsia, and NTP in groups A and B were also respectively compared, and the difference was not statistically significant (all p > 0.05, Table 2).

Pregnancy is a special stage of women’s experience, and a variety of hormones secreted by the placenta during pregnancy lead to an increase of insulin resistance, which can cause the changes of maternal glucose metabolism, lipid metabolism, and protein metabolism [14, 15]. Insulin resistance is the central pathological process of metabolic syndrome and GDM [16, 17]. Accumulated studies have reported that both GDM and PGDM in pregnancy can increase the incidence of maternal and fetal complications, such as fetal macrosomia, neonatal hypoglycemia, perinatal mortality, and increased risk of cesarean delivery [1, 18-20]. PGDM includes both types 1 and 2 diabetes mellitus occurring prior to pregnancy, which is associated with an increased risk for maternal and fetal adverse outcomes than GDM [21, 22]. For instance, a study by Fong et al. demonstrated that PGDM was associated with significantly higher mortality when compared to GDM [21]. In the Omani cohort, pregnant women with PGDM were also found at higher risk of developing obstetric and perinatal complications compared to GDM women [22]. Previous studies have shown that pre-pregnancy care for pregnant women with PGDM could improve the maternal and perinatal outcomes. For instance, Wahabi et al. demonstrated that pre-pregnancy care for women with pre-gestational type 1 or 2 diabetes mellitus was effective in improving rates of congenital malformations, perinatal mortality, and in reducing maternal HbA1c in the first trimester of pregnancy [23]. All these studies indicated the pivotal role of early screening and pre-pregnancy treatment of PGDM for maternal and perinatal outcomes.

In this study, among 822 pregnant women, 31.9% (262/ 822) pregnant women with PGDM were diagnosed before pregnancy, and 68.1% were diagnosed in pregnancy, of whom the rate of missed diagnosis of PGDM was 68.1%. The missed diagnosis rate of PGDM is consistent with the previous study by Yang et al. [7]. Even in developed countries with advanced medical technology and relatively complete regular physical examination system, such as the USA, among those with diabetes, the rate of undiagnosed diabetes was as high as 36.4% [24]. However, due to the incomplete periodic physical examination system in China, the rate of undiagnosed PGDM is more than 2/3 prior to pregnancy [25]. Due to the incidence of adverse pregnancy outcomes can be significantly reduced after blood glucose management before pregnancy, more strategies should be made to diagnose the PGDM.

To explore the effect of PGDM diagnosed at a different time on pregnancy outcome, the comparison among the pre-pregnancy diagnosis group, and pregnancy diagnosis groups A and B were analyzed. Firstly, the authors compared the pregnancy outcomes between pre-pregnancy diagnosis and pregnancy diagnosis groups. The average HbA1c level in pre-pregnancy diagnosis group was higher than that in pregnancy diagnosis group, but it had no statistically significant difference between the two groups. According to the results, the authors found that the proportion of insulin treatment and the rate of cesarean section were higher in pre-pregnancy diagnosis group than pregnancy diagnosis group, which had a significant difference. However, the rate of pre-term, macrosomia, newborns transfer to pediatrics, and preeclampsia showed no difference between the two groups. Further analysis revealed that the average HbA1c level in pregnancy diagnosis group A was higher than that in pre-pregnancy diagnosis group, but there was no significant difference between the two groups. Besides the average HbA1c level, the rate of insulin treatment, pre-term rate, cesarean delivery rate, the incidence of macrosomia, preeclampsia, and NTP were close to the pre-pregnancy diagnosis group. The reason might be that the pregnant women in the pregnancy diagnosis group A did not take good blood glucose control prior to pregnancy, and the risk of adverse pregnancy outcome was similar to that of in the pre-pregnancy diagnosis group.

The diagnosis of pregnancy diagnosis group B (PGDM pregnant women diagnosed at 24 weeks or after) was mainly based on the results of oral glucose tolerance test (OGTT). The results of the hyperglycemia and adverse pregnancy outcome study (HAPO) showed that center-to-center differences occurred in GDM frequency and relative diagnostic importance of fasting, one-, and two-hour glucose levels, for example, the proportion with only the two-hour value equal to or greater than threshold was just 6% in Bellflower but reached 29% in Hong Kong [26. 27]. The OGTT glucose metabolic is different in different regions or populations, whether the diagnosis of PGDM can be made by relying on OGTT two-hour blood glucose level remains to be further studied. The average level of HbA1c in pregnancy diagnosis group B was lower than those in pre-pregnancy diagnosis group and pregnancy diagnosis group A. At the same time, the proportion of pregnant women using insulin treatment was much lower in pregnancy group B than that of in pregnancy diagnosis group A, as well as pre-pregnancy diagnosis group. In addition, the rate of cesarean section was lower in pregnancy diagnosis group B than that in the pre-pregnancy diagnosis group. Taken together, there was a difference in pregnancy diagnosis group B when compared to the other two groups, and whether PGDM can be diagnosed according to the results of OGTT is still controversial. In considering the results of this study, the clinical data is limited, and more clinical parameters are needed to be involved in the further analyses. Furthermore, the postpartum follow-up data of PGDM pregnant women diagnosed at or later 24 weeks is needed to be analyzed in the further study.

In conclusion, the rate of undiagnosed PGDM was high, and PGDM was significantly associated with multiple adverse pregnancy outcomes. Fasting plasma glucose should be used as a screening test to identify PGDM at pre-pregnancy or first antenatal care. Using the abnormal value of two-hour plasma glucose after 24 gestational weeks as the only way to diagnose PGDM may not be suitable. The PGDM pregnant women who were diagnosed at or later 24 weeks were different from the PGDM pregnant women who were diagnosed at other times, whether the diagnosis of diabetes should be confirmed by a further postpartum review.