Risk Factors for Pregnancy Complications and Postpartum Glucose Intolerance in Women With Gestational Diabetes Mellitus

This prospective cohort study aimed to evaluate the risk factors for pregnancy complications and postpartum glucose intolerance (GI) in women with gestational diabetes mellitus (GDM). A total of 140 women with GDM were enrolled. Of these, 115 underwent a 75-g oral glucose tolerance test (OGTT) at 12 weeks after delivery. Clinical factors and parameters in the antepartum 75-g OGTT associated with pregnancy complications and postpartum GI were evaluated. Women with GDM experienced pregnancy complications, including hypertensive disorders of pregnancy (HDP, n=19), preterm delivery (PD, n=17), heavy-for-date (HFD, n=19), and light-for-date (LFD, n=12), and 22 of the 115 women with GDM developed postpartum GI. The univariate and multivariable logistic regression analyses revealed the following risk factors: histories of hypertension (odds ratio [OR], 23.8; 95% condence interval [CI], 4.2–134.7; p<0.01) for HDP; histories of hypertension (OR, 9.8; 95% CI, 2.5–38.9; p<0.01) for PD; HbA1c levels (OR, 7.6; 95% CI, 1.5–37.9; p<0.05) for HFD; and oral deposition index (DI) (OR, 0.1; 95% CI, 0.02–0.7; p<0.01) for postpartum GI. Higher HbA1c levels and lower oral DI on the antepartum 75-g OGTT may be useful markers for identifying GDM women who are at high risk for HFD and postpartum GI, respectively.

Previous studies have evaluated associations between maternal clinical or laboratory ndings of antepartum 75-g oral glucose tolerance test (OGTT) and pregnancy complications in women with GDM.
PD and HDP are associated with normal fasting and elevated post-load blood glucose (BG) levels 6 . HFD is associated with elevated fasting and normal post-load BG levels 6,7 and post-load hyperglycemia 8 .
Moreover, macrosomia is associated with post-load hyperglycemia 8 , fasting hyperglycemia, and excessive gestational weight gain 7,9 . A systematic review and meta-analysis has demonstrated that women with GDM have a 7.4-fold increased risk of developing type 2 diabetes mellitus after delivery compared with those without GDM 10 .
We have reported that, in women with GDM, the low insulinogenic index (II) levels on the antepartum 75-g OGTT is a risk factor for developing glucose intolerance (GI) during the early postpartum period 11 . This prospective cohort study aimed to assess predictive clinical factors and laboratory parameters in the antepartum 75-g OGTT for pregnancy complications and GI during the early postpartum period among women with GDM.

Page 3/10
Of 3,494 pregnant women who had singleton deliveries at the Kobe University Hospital, 140 (4.0%) were diagnosed with GDM from January 2011 to December 2018. The indications for the antepartum 75-g OGTT in the 140 pregnant women with GDM were as follows: a 1-hr BG level on a 50-g glucose challenge tests ≥140 mg/dL (n=99); casual BG level ≥100 mg/dL (n=13); suspicion of HFD and/or polyhydramnios on ultrasound examinations during pregnancy (n=10); and presence of other risk factors of GDM, including a history of GDM, obesity, and persistent glycosuria (n=18). Twenty-ve of the 140 women with GDM refused to receive a 75-g OGTT at 12 weeks after delivery. Therefore, 115 of the 140 (82.1%) women with GDM were included in the analyses of risks for GI during the early postpartum period.
Clinical factors and parameters in the antepartum 75-g OGTT associated with the occurrence of HDP in women with GDM Nineteen of the 140 (13.6%) pregnant women with GDM had HDP. Table 1 shows the clinical characteristics of participants and result of logistic regression analyses of factors associated with HDP.
The group of GDM women with HDP (HDP group) had a signi cantly higher body mass index (BMI) prior to pregnancy (p<0.05) and the frequency of a history of hypertension (p<0.01) than the group without HDP (non-HDP group). Furthermore, the HDP group had a signi cantly less weight gain during pregnancy (p<0.01) than the non-HDP group. No signi cant differences were observed in any parameter in the antepartum 75-g OGTT between the two groups.
Univariate logistic regression analyses demonstrated that the BMI prior to pregnancy (odds ratio [OR], 1.1;  Table 2 shows the clinical characteristics of participants and result of logistic regression analyses of factors associated with PD. The group of GDM women with PD (PD group) had a signi cantly higher parity (p<0.05) and frequency of the presence of a history of hypertension (p<0.01) than the group without PD (non-PD group).
Univariate logistic regression analyses demonstrated that the presence of a history of hypertension (OR, 9.8; 95% CI, 2.5-38.9; p<0.01) was a single independent factor associated with PD in pregnant women with GDM.
Clinical factors and parameters in the antepartum 75-g OGTT associated with the occurrence of HFD infants in women with GDM Nineteen of the 140 (13.6%) women with GDM had HFD newborns. Table 3 reveals the clinical characteristics of participants and results of logistic regression analyses of factors associated with HFD. The group of GDM women with HFD (HFD group) had a signi cantly higher weight gain during pregnancy (p<0.05) and HbA1c levels (p<0.05) than the group without HFD (non-HFD group).
Clinical factors and parameters in the antepartum 75-g OGTT associated with the occurrence of LFD infants in women with GDM Twelve of the 140 (8.6%) pregnant women with GDM had LFD newborns. Table 4 exhibits the clinical characteristics of participants and result of logistic regression analyses of factors associated with LFD. The group of GDM women with LFD (LFD group) had a signi cantly lower HbA1c levels (p<0.05) than the group without LFD (non-LFD group).
No clinical factors and parameters in the antepartum 75-g OGTT were selected as factors associated with LFD in pregnant women with GDM by univariate logistic regression analyses.
Clinical factors and parameters in the antepartum 75-g OGTT associated with the occurrence of postpartum GI in women with GDM Twenty-two of the 115 (19.1%) pregnant women with GDM had GI at 12 weeks after delivery, including one, two, and 19 women with diabetes mellitus (DM), impaired fasting glucose (IFG), and impaired glucose tolerance (IGT), respectively. Table 5 shows the clinical characteristics of participants and results of logistic regression analyses of factors associated with GI during the early postpartum period. The group of GDM women with postpartum GI (GI group) had a signi cantly lower oral DI (p<0.01) than the group without postpartum GI (non-GI group).

Discussion
This study used the International Association of Diabetes and Pregnancy Study Groups (IADPSG) criteria for diagnosing GDM 12 , and 140 of the 3,494 (4.0%) pregnant women who had singleton deliveries were diagnosed with GDM. Because medians of the prevalence of GDM in Japan were reported to be 2.8%-13.0% 13 , the prevalence of GDM in this study was thought to be valid. The incidences of HDP, PD, HFD newborns, and LFD newborns among women with GDM were 13.6%, 12.1%, 13.6%, and 8.6%, respectively.
The incidences of pregnancy complications in women with GDM that were diagnosed based on the IADPSG criteria have been reported as follows: HDP, 8.2% 14 ; PD, 6.5%-8.4% 15,16 ; large-for-gestational-age (LGA) newborns, 9.1%-21.4% 15,16 ; and small-for-gestational-age (SGA) newborns, 7.6%-8.0% 16,17 . Although the present study had higher incidences of HDP and PD than the previous studies, the incidence (19.1%) of GI during the early postpartum period in women with GDM was comparable to those (16.7%-36.6%) in previous studies 11,18,19 . Our study had higher incidences of HDP and PD than the previous studies because Kobe University Hospital has a maternal-fetal center where pregnant women with HDP and threatened premature labor are often referred from other hospitals and clinics.
For the rst time, this prospective cohort study of pregnant women with GDM simultaneously assessed both the clinical factors and parameters in the antepartum 75-g OGTT for pregnancy complications and GI at 12 weeks after delivery by logistic regression analyses using a stepwise approach, and revealed the following risk factors: the presence of histories of hypertension both for HDP and PD at 36 or less GW; higher HbA1c levels for HFD infants; and lower oral DI for postpartum GI.
Obesity, chronic hypertension, and a history of HDP have been reported as major risk factors for HDP 20,21 . Our result is comparable with those in previous studies.
Previous prospective cohort studies, including a general population of pregnant women, revealed that older maternal age, lower BMI, and GDM were risk factors for PD 22,23 . When subjects were limited to women with GDM as in our study, not maternal age and BMI prior to pregnancy, but a history of hypertension was selected as a risk factor for PD. In addition, we found that a history of hypertension was an independent risk factor for HDP in women with GDM. It is likely that pregnancies of GDM women with histories of hypertension result in preterm births due to recurrent HDP.
Several retrospective studies demonstrated that elevated fasting BG levels were associated with HFD 6,7 , and others suggested that post-load hyperglycemia was associated with HFD 8 . In this prospective cohort study, the univariate and multivariable logistic regression analyses demonstrated that the higher HbA1c level was a risk factor for HFD in pregnant women with GDM. This result indicated that HFD in women with GDM was more closely associated with continuous hyperglycemia, which higher HbA1c levels re ect, rather than higher levels of fasting or post-load BG on the antepartum 75-g OGTT.
The present study also found neither differences in clinical factors and parameters in the antepartum 75g OGTT between the LFD and non-LFD groups nor any factors associated with the occurrence of LFD in pregnant women with GDM. SGA/LFD newborns are likely to be associated with severe DM 7,24 rather than GDM.
Univariate and multivariable logistic regression analyses revealed for the rst time that oral DI was an independent risk factor for postpartum GI in pregnant women with GDM. Previous retrospective studies demonstrated that low II and II/fasting IRI were associated with postpartum GI in patients with GDM 18,25 . A previous prospective cohort study of 72 pregnant women with GDM revealed that a low II in the antepartum 75-g OGTT is an independent risk factor for developing GI during the early postpartum period 11 . It was reported that among the Japanese-American adults, including males and non-pregnant women, the low oral DI was predictive of developing DM in the future 26 . DI represents a hyperbolic relationship between insulin secretion and insulin sensitivity 27,28 . Therefore, this parameter represents the insulin secretory capacity of pancreatic β cells adjusted for insulin sensitivity 26 . An adequate insulin secretory response of pancreatic β cells adapting to changes in insulin sensitivity might be signi cant for the maintenance of normal glucose tolerance during the postpartum period. Pregnant women with low oral DI on the antepartum 75-g OGTT are at high risk not only for GI during the early postpartum period, but also for DM in the future.
This prospective cohort study demonstrated that a history of hypertension was a risk factor for HDP and PD in pregnant women with GDM, and higher HbA1c levels was a risk factor for HFD newborns. A low oral DI on the antepartum 75-g OGTT was an independent risk factor for GI during the early postpartum period in women with GDM. These ndings may enable clinicians to effectively identify and manage women with GDM who are at high risks for pregnancy complications and DM in the future.

Study design and participants
This prospective cohort study enrolled women with singleton pregnancies who were diagnosed with GDM by the 75-g OGTT during pregnancy and delivered at the Kobe University Hospital from January 2011 to December 2018. The study was approved by the Institutional Review Board of the Kobe University Hospital (reference number 200228), and informed consent was obtained from all participants. All research was performed in accordance with the relevant guidelines and regulations.

Procedures
All pregnant women who visited or were referred to the Kobe University Hospital underwent screening for GDM both at 10-14 and 24-28 GW. Pregnant women who had casual BG levels of ≥100 mg/dL (5.5 mmol/L) at 10-14 GW, or those who had 1-hr BG levels of ≥140 mg/dL (7.8 mmol/L) on 50-g glucose challenge tests (GCT) at 24-28 GW, or those with risk factors for GDM, including obesity, family history of DM , past history of macrosomia, presence of persistent glycosuria, polyhydramnios, and suspected HFD, underwent the 75-g OGTT. According to the IADPSG criteria 12 , the diagnosis of GDM is made when any of the following are met: FBG ≥92 mg/dL (5.1 mmol/L), 1-hr BG ≥180 mg/dL (10.0 mmol/L), or 2-hr BG ≥153 mg/dL (8.5 mmol/L). BG and immunoreactive insulin (IRI) levels at 0, 0.5, 1, 1.5, and 2 hr after the oral ingestion of 75-g glucose were also measured, and the total area under the curve (AUC) of glucose and insulin were calculated by the trapezoid method.
All pregnant women diagnosed with GDM were referred to diabetologists in the Kobe University Hospital and underwent self-monitoring of blood glucose (SMBG) and diet therapy. If FBG levels exceeded 100 mg/dL, or 2-hr BG levels exceeded 120 mg/dL in SMBG regardless of diet therapy, an insulin therapy was started. Insulin doses were adjusted to achieve both FBG levels of <100 mg/dL and 2-hr BG levels of <120 mg/dL. All pregnant women with GDM were instructed to undergo a 75-g OGTT at 12 weeks after delivery. Using the WHO's 1999 criteria 30 , DM was diagnosed by either FBG levels of ≥126 mg/dL (7.0 mmol/L) or 2-hr BG levels of ≥200 mg/dL (11.1 mmol/L). IFG was diagnosed by FBG levels of ≥110 mg/dL (6.1 mmol/L), and IGT was diagnosed by 2-hr BG levels of ≥140 mg/dL (7.8 mmol/L). GI was de ned by the presence of DM, IFG, or IGT. FBG levels of <110 mg/dL (6.1 mmol/L) and 2-hr BG levels of <140 mg/dL (7.8 mmol/L) were identi ed as normal.
This study assessed pregnancy complications, including HDP, PD at 36 or less GW, and HFD or LFD newborns. HFD and LFD were de ned as newborns with a birth weight >90th and <10th percentile for gestational age, respectively.

Statistical analysis
Clinical characteristics were compared between pregnancies with each pregnancy complication or with GI during the early postpartum period and pregnancies without them. Differences between the two groups were analyzed using the Mann-Whitney U test, Fisher exact test, and χ2 test. P values <0.05 were considered statistically signi cant. The stepwise approach was used to evaluate clinical factors and parameters in the antepartum 75-g OGTT associated with each pregnancy complication and GI during the early postpartum period. Variables with P values <0.05 in univariate logistic regression analyses were subjected to multivariable logistic regression analyses, and variables with P values <0.05 in multivariable logistic regression analyses were determined as clinical factors and parameters in the antepartum 75-g OGTT associated with each pregnancy complication or GI during the early postpartum period in women with GDM. All statistical analyses were performed using the SPSS software, version 19 (SPSS Inc., Chicago, Illinois).

Declarations Data availability
The datasets analyzed during the current study are available from the corresponding authors upon a reasonable request.