Document resource
WHAT IS ALREADY KNOWN ON THIS TOPIC Gestational diabetes mellitus (GDM) increases the risk of adverse perinatal outcomes and future metabolic disorders. Although screening at 24–28 weeks is typically recommended, the benefits of early diagnosis and treatment of GDM remain uncertain. Previous studies have reported inconsistent results regarding the effect of early-onset GDM (EGDM) intervention on pregnancy outcomes and postpartum metabolic risk.WHAT THIS STUDY ADDS EGDM (before 24 weeks) in non-obese Japanese pregnant women is associated with decreased early postpartum initial insulin secretion capacity, and they are at high risk of postpartum glucose intolerance.HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY Non-obese EGDM women should be carefully managed and monitored during pregnancy to prevent impaired insulin secretion and glucose intolerance in the postpartum period. Therefore, identifying the features and early diagnosis of GDM may aid in the prevention, prediction, and appropriate management of GDM-related adverse events. Moreover, it may be necessary to reconsider the initial screening methods for GDM.Introduction Gestational diabetes mellitus (GDM) is one of the most common metabolic complications of pregnancy. It is defined as glucose tolerance abnormalities with hyperglycemia that occur during pregnancy. It is associated with increased perinatal complications in both the fetus and the mother, such as cesarean section, preterm birth, macrosomia and large for gestational age (LGA), with high prospective metabolic risks in the mother and offspring. 1 2 The diagnostic criteria for GDM proposed by the International Association of Diabetes and Pregnancy Study Group (IADPSG)3 are based on the results of the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study,4 which analyzed data from pregnant women in the second or third trimester of pregnancy (24–32 weeks), and therefore recommended that the timing for screening and initiation of treatment for GDM is from weeks 24–28 of gestation.5 6 Despite the importance of screening in early pregnancy (<20 weeks gestation) for previously undiagnosed DM, evidence supporting the use of the IADPSG criteria for the diagnosis of early-onset GDM (EGDM) has been insufficient. Furthermore, evidence supporting the benefits of therapeutic intervention during early pregnancy has been limited. In recent years, systematic reviews have reported an increase in macrosomia, cesarean section, labor induction, hypertension, preterm birth, and shoulder dystocia in early pregnancy GDM, as well as an increase in postpartum glucose intolerance and insulin requirements.7 In addition, some randomized controlled trials (RCTs) have examined the effectiveness of therapeutic intervention for GDM diagnosed in early pregnancy: Harper et al8 reported that screening for EGDM in American pregnant women who were obese did not result in the reduction of composite perinatal outcomes, and Simmons et al9 reported that therapeutic intervention for EGDM was associated with decreased risk of neonatal respiratory distress syndrome. However, it did not affect the incidence of pregnancy-related hypertension, LGA, or neonatal lean body mass. A recent systematic review and meta-analysis of RCTs comparing early treatment versus observation in EGDM reported that the early treatment group did not reduce the primary outcomes of LGA, macrosomia or pregnancy-induced hypertension, but only significantly reduced the secondary neonatal outcome of neonatal respiratory distress.10 In agreement with these observations, in Japan, Nakanishi et al11 reported that uniform therapeutic intervention for EGDM did not reduce perinatal complications, and Yokoyama et al12 reported a lack of difference in pregnancy outcomes between EGDM and late-onset GDM (LGDM), although EGDM was associated with a high risk of developing postpartum glucose intolerance. Hence, although many studies have examined the effect of EGDM on perinatal outcomes, current evidence from RCTs only partially supports the effectiveness of early diagnosis and intervention. No consensus has been reached regarding the benefits of early intervention on perinatal outcomes, diagnostic, or therapeutic strategies. In this study, we aimed to compare the perinatal outcomes and incidence of postpartum glucose intolerance between early and LGDM in Japan. To our knowledge, this is the first study that compared postpartum insulin secretion capacity between EGDM and LGDM.Research design and methods Study design The study was conducted in accordance with STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) criteria for retrospective observational studies and adhered to the ethical principles of the Declaration of Helsinki.Participants This study included women with singleton pregnancy and GDM who delivered at National Hospital Organization Okayama Medical Center between July 2010 and April 2024. The diagnosis of GDM was based on the results of 75 g oral glucose tolerance tests (OGTTs) as proposed by the IADPSG. 3 EGDM was defined as GDM diagnosed at<24 weeks of gestation and LGDM as that diagnosed after 24 weeks. After excluding those with unavailable 75 g OGTT results at diagnosis and post partum, 107 women with EGDM and 109 with LGDM were analyzed (figure 1).Figure 1Flow chart of this study. EGDM, GDM diagnosed at<24 weeks of gestation; GDM, gestational diabetes mellitus; LGDM, GDM diagnosed at≥24 weeks of gestation; NHO, National Hospital Organization; OGTT, oral glucose tolerance test.A 75 g OGTT at diagnosis was performed for pregnant women with a random glucose level of≥100 mg/dL at early or mid-term screening. Random blood glucose levels were measured early during pregnancy. Those who had glucose levels≥100 mg/dL underwent a 75 g OGTT, with those satisfying the IADPSG diagnostic criteria for GDM being diagnosed with EGDM. When glucose levels during the initial screening were below the cut-off value, random blood glucose levels were reevaluated after the second trimester (after 24 weeks). Additionally, women who had elevated random glucose levels (≥100 mg/dL) but normal OGTT results in early pregnancy were also retested in mid-pregnancy. Those who had glucose levels≥100 mg/dL underwent a 75 g OGTT, with those satisfying the criteria being diagnosed with LGDM. Almost all pregnant women at our hospital and other medical facilities underwent both early and mid-pregnancy screenings.Pregnant women who had previously been diagnosed with type 1 or type 2 diabetes before pregnancy or who had confirmed diabetic retinopathy were excluded from this study for having pregestational diabetes. Additionally, based on the results of the 75 g OGTT, the following cases were excluded due to overt diabetes during pregnancy: (1) fasting blood glucose levels≥126 mg/dL or (2) a glycosylated hemoglobin (HbA1c)≥6.5% on blood tests. Furthermore, pregnant women whose random or 2-hour blood glucose levels during the 75 g OGTT were≥200 mg/dL and those who satisfied either criterion were considered highly likely to have overt diabetes and were excluded from the study.Although Japanese clinical guidelines13 recommend performing a 75 g OGTT 6–12 weeks post partum, in actual clinical practice, it is often performed slightly earlier or later than the recommendation. Thus, in this study, 75 g OGTTs were performed 4–16 weeks postpartum. Subgroup analysis was also performed by classifying participants into obese (body mass index (BMI) ≥25 kg/m2) and non-obese (BMI<25 kg/m2) groups.Treatment for GDM In both groups, HbA1c and glycated albumin were measured once every month starting from the initial diagnosis of GDM, and self-monitoring of blood glucose was performed at home. 14 Regarding treatment, non-pharmacological and dietary therapy was initiated on all participants. In addition, insulin therapy was initiated when the fasting glucose level exceeded 95 mg/dL or the 2-hour postprandial glucose level exceeded 120 mg/dL during pregnancy.5Measures of pregnancy outcomes Maternal pregnancy outcomes included gestational weight gain (GWG), preterm birth, insulin use, pregnancy-induced hypertension, emergency cesarean section and instrumental delivery. Neonatal outcomes included birth weight, Apgar score at 5 min<7, small for gestational age (SGA), LGA, macrosomia, neonatal intensive care unit admission, neonatal hyperbilirubinemia and neonatal hypoglycemia. SGA and LGA were defined as birth weight below and above the 90th percentiles for sex and gestational age, respectively, in the Japanese population. 15Measures of postpartum glucose tolerance and insulin secretion capacity Glucose tolerance and insulin secretion capacity were assessed based on the results of 75 g OGTTs performed 4–16 weeks post partum. Postpartum glucose intolerance was categorized into three groups based on postpartum 75 g OGTT results: DM (fasting glucose level of≥126 mg/dL and/or glucose level of≥200 mg/dL 2 hours after glucose loading), impaired fasting glucose (fasting glucose level of 110–125 mg/dL and glucose level<140 mg/dL 2 hours after glucose loading) and impaired glucose tolerance (IGT) (glucose level of 140–199 mg/dL 2 hours after glucose loading). To evaluate postpartum insulin secretion capacity, immunoreactive insulin (IRI) levels were measured before and after glucose loading, and the insulinogenic index (IGI) was calculated; 16 an IGI of<0.4 was considered a decrease in initial insulin secretion capacity.17 Homeostatic model assessments of insulin resistance (HOMA-IR) and β-cell function (HOMA-β) were also calculated from the 75 g OGTT results.Statistical analysis Categorical data are presented as numbers (percentages). Numerical data are expressed as median (IQR). Continuous variables were compared between the two groups by Mann-Whitney U test, and categorical variables were compared using the χ 2 test or Fisher’s exact test. Logistic regression analysis was used to calculate the OR with 95% CIs for confounders. Multivariate analysis was performed by adjusting for confounders of maternal age>35 years at delivery and last antepartum HbA1c level. All statistical analyses were performed with IBM SPSS Statistics V.28 (IBM, Armonk, New York, USA). A p value of<0.05 was considered significant.Results Table 1 presents the maternal characteristics. Although no significant differences in age, pre-pregnancy BMI and 75 g OGTT results were found between the EGDM and LGDM groups at diagnosis, a significantly higher proportion of women in the EGDM group had two or more positive values on the 75 g OGTT. Moreover, no significant difference in the last antepartum HbA1c level was noted between the two groups. Pregnancy outcomes and the results of postpartum 75 g OGTT for the EGDM and LGDM groups are shown in table 2. GWG was significantly lower in the EGDM group; however, no significant differences in neonatal birth weight were observed between the two groups. The incidence of SGA and LGA did not significantly differ between the two groups, and no difference in the incidence of macrosomia was noted. Regarding postpartum 75 g OGTT results, blood glucose levels were significantly higher 2 hours after loading, and IRI levels were significantly lower 30 min and 1 hour after loading in the EGDM group than in the LGDM group. However, the IGI and proportions with IGI<0.4 were comparable in both groups. Postpartum glucose intolerance was more prevalent in the EGDM group; however, the differences were not significant.Table 1Maternal characteristicsEGDM (n=107)LGDM (n=109)P valueMaternal age (years)36 (31–38)34 (30–39)0.624Pre-pregnancy BMI (kg/m2)23.1 (19.6–27.8)22.4 (19.9–27.0)0.785Nullipara, n (%)50 (46.7)54 (49.5)0.686Smoking, n (%)0 (0)1 (0.9)1.000History of GDM, n (%)7 (6.5)1 (0.9)0.034Gestational age atdiagnostic OGTT (weeks)13 (12–16)28 (27–29)<0.0001Plasma glucose of diagnostic OGTT (mg/dL) 0 min (fasting)88 (83–95)86 (79–96)0.452 1 hour182 (158–203)180 (159–192)0.417 2 hours163 (143–180)159 (135–170)0.101Number of positive glucose values on OGTT, n (%) 149 (45.8)66 (60.6)0.041 2 or 358 (54.2)43 (39.4)0.041Last antepartum HbA1c (%)5.5 (5.3–5.7)5.6 (5.3–5.8)0.064Values are expressed as median (IQR) or n (%).BMI, body mass index; EGDM, GDM diagnosed at<24 weeks of gestation; GDM, gestational diabetes mellitus; HbA1c, glycosylated hemoglobin; LGDM, GDM diagnosed at≥24 weeks of gestation; OGTT, oral glucose tolerance test.Table 2Pregnancy outcomes and postpartum OGTT resultsEGDM (n=107)LGDM (n=109)P valueGestational weight gain (kg)6.3 (3.5–9.7)8.4 (4.0–11.0)0.048Insulin therapy, n (%)79 (73.8)75 (68.8)0.454Maximum daily insulin dose during pregnancy, unit (N)29.5 (14.5–45.5)(n=78)18.0 (10.8–42.3)(n=74)0.083Gestational age (weeks)38 (37–39)38 (37–39)0.662Preterm birth, n (%)15 (14.0)14 (12.8)0.844Pregnancy-related hypertension, n (%)9 (8.4)9 (8.3)1.000Neonatal birth weight (g)2888 (2536–3126)2898 (2564–3230)0.635LGA, n (%)10 (9.3)10 (9.2)1.000SGA, n (%)12 (11.2)15 (13.8)0.682Macrosomia, n (%)0 (0.0)2 (1.8)0.498Apgar score<7 at 5 min, n (%)5 (4.7)5 (4.6)1.000Pre-eclampsia, n (%)0 (0.0)0 (0.0)–Emergency cesarean section, n (%)17 (15.9)19 (17.4)0.454Instrumental delivery, n (%)10 (9.3)7 (6.4)0.459NICU admission, n (%)16 (15.0)19 (17.4)0.713Neonatal hypoglycemia, n (%)1 (0.9)2 (1.8)1.000Neonatal hyperbilirubinemia, n (%)16 (15.0)11 (10.1)0.309Postpartum weeks at OGTT (weeks)8 (7–10)8 (7–11)0.068Plasma glucose of the postpartum OGTT (mg/dL)0 min (fasting)90 (85–98)90 (86–97)0.74530 min151 (140–167)155 (139–176)0.3911 hour163 (141–189)158 (140–181)0.5352 hours134 (115–155)126 (110–147)0.047IRI of the postpartum OGTT (µIU/mL)0 min (fasting)4.4 (2.8–6.5)4.9 (3.1–7.2)0.14230 min30.2 (21.1–46.3)38.1 (28.2–54.7)0.0041 hour37.1 (25.5–51.7)42.8 (29.3–62.8)0.0392 hours29.8 (19.0–45.2)29.1 (21.7–49.6)0.351IGI0.43 (0.27–0.73)0.51 (0.33–0.80)0.069IGI<0.4, n (%)48 (44.9)37 (33.9)0.125HOMA-β56.6 (36.8–82.5)63.4 (43.5–85.1)0.140HOMA-IR0.99 (0.60–1.51)1.09 (0.65–1.69)0.173Glucose tolerance 4–16 weeks postpartum, n (%)NGT59 (55.1)73 (67.0)0.094IGT/IFG45 (42.1)35 (32.1)0.159DM3 (2.8)1 (0.9)0.367Values are expressed as median (IQR) or n (%).Postpartum 75 g OGTT was performed at 4–16 weeks.DM, diabetes mellitus; EGDM, GDM diagnosed at<24 weeks of gestation; GDM, gestational diabetes mellitus; HOMA-IR, homeostatic model assessment for insulin resistance; HOMA-β, homeostatic model assessment of β-cell function; IFG, impaired fasting glucose; IGI, insulinogenic index; IGT, impaired glucose tolerance; IRI, immunoreactive insulin; LGA, large for gestational age; LGDM, GDM diagnosed at≥24 weeks of gestation; NGT, normal glucose tolerance; NICU, neonatal intensive care unit; OGTT, oral glucose tolerance test; SGA, small for gestational age.Non-obese pregnant women who had BMI<25 kg/m2 (EGDM group, n=70; LGDM group, n=71) were analyzed. The results revealed no significant differences in maternal parameters or 75 g OGTT results at diagnosis between the two groups, although the EGDM group tended to have a higher proportion of women with two or more positive values on the 75 g OGTT (data not shown). However, when only comparing the non-obese EGDM group with the non-obese LGDM group, GWG was significantly lower (p=0.020) in the EGDM group (7.1 kg) than in the LGDM group (9.7 kg), and neonatal birth weight tended to be lower (p=0.063) in the EGDM group (2814 g) than in the LGDM group (2925 g), with no apparent differences in neonatal outcomes, including SGA, LGA or macrosomia. Postpartum 75 g OGTT results and blood glucose levels were comparable; however, the 30 min and 1-hour IRI after loading and IGI were significantly lower (p=0.007) in the non-obese EGDM group (0.36) than in the non-obese LGDM group (0.50), and the proportion of women with IGI<0.4 was significantly higher in the EGDM group (54.3% vs 33.8%, p=0.018). HOMA-β and HOMA-IR showed no significant differences. The prevalence of postpartum glucose intolerance was significantly higher in the EGDM group than in the LGDM group (table 3).Table 3Pregnancy outcomes and postpartum OGTT results of pregnant women with pre-pregnancy BMI<25 kg/m2EGDM (n=70)LGDM (n=71)P valueGestational weight gain (kg)7.1 (4.9–11.1)9.7 (6.6–12.7)0.020Insulin therapy, n (%)51 (72.9)46 (64.8)0.364Maximum daily insulin dose during pregnancy, unit (N)22.5 (11.8–43.3)(n=50)16.0 (8.5–33.5)(n=45)0.088Gestational age (weeks)38 (37–39)38 (38–39)0.313Preterm birth, n (%)8 (11.4)7 (9.9)0.792Pregnancy-related hypertension, n (%)4 (5.7)3 (4.2)0.719Neonatal birth weight (g)2814 (2536–3012)2925 (2608–3222)0.063LGA, n (%)5 (7.1)7 (9.9)0.764SGA, n (%)11 (15.7)10 (14.1)0.817Apgar score<7 at 5 min, n (%)2 (2.9)3 (4.2)1.000Emergency cesarean section, n (%)11 (15.7)13 (18.3)0.823NICU admission, n (%)8 (11.4)10 (14.1)0.802Neonatal hypoglycemia, n (%)1 (1.4)0 (0.0)0.496Neonatal hyperbilirubinemia, n (%)8 (11.4)6 (8.5)0.586Postpartum weeks at OGTT (weeks)8 (7–10)8 (7–11)0.054Plasma glucose of the postpartum OGTT (mg/dL) 0 min (fasting)89 (84–95)89 (85–95)0.993 30 min148 (138–164)148 (138–166)0.706 1 hour163 (133–185)150 (133–177)0.191 2 hours132 (114–154)122 (108–138)0.062IRI of the postpartum OGTT (μIU/mL) 0 min (fasting)3.3 (2.3–4.8)3.9 (2.6–5.9)0.083 30 min25.2 (18.8–34.9)36.2 (26.2–46.7)<0.001 1 hour30.9 (23.3–42.6)39.0 (25.0–50.3)0.031 2 hours25.3 (17.6–34.2)27.3 (19.5–27.3)0.184IGI0.36 (0.26–0.62)0.50 (0.35–0.74)0.007IGI<0.4, n (%)38 (54.3)24 (33.%)0.018HOMA-β46.0 (34.0–68.1)55.3 (38.4–69.4)0.106HOMA-IR0.75 (0.49–1.04)0.90 (0.57–1.38)0.110Glucose tolerance at 4–16 weeks post partum, n (%)NGT42 (60.0)54 (76.1)0.048IGT/IFG27 (38.6)17 (23.9)0.071DM1 (1.4)0 (0.0)0.496Values are expressed as median (IQR) or n (%).Postpartum 75 g OGTT was performed at 4–16 weeks.BMI, body mass index; DM, diabetes mellitus; EGDM, GDM diagnosed at<24 weeks of gestation; GDM, gestational diabetes mellitus; HOMA-IR, homeostatic model assessment for insulin resistance; HOMA-β, homeostatic model assessment of β-cell function; IFG, impaired fasting glucose; IGI, insulinogenic index; IGT, impaired glucose tolerance; IRI, immunoreactive insulin; LGA, large for gestational age; LGDM, GDM diagnosed at≥24 weeks of gestation; NGT, normal glucose tolerance; NICU, neonatal intensive care unit; OGTT, oral glucose tolerance test; SGA, small for gestational age.Multivariate logistic regression analysis showed that early diagnosis of GDM was significantly associated with postpartum IGI<0.4 (OR, 2.131; 95% CI 1.015 to 4.474, p=0.046). Conversely, the timing of GDM diagnosis was not significantly associated with postpartum glucose intolerance. Postpartum glucose intolerance was most strongly associated with a 2-hour glucose level≥153 mg/dL on the 75 g OGTT (OR, 3.713; 95% CI 1.381 to 9.981, p=0.009; table 4).Table 4Association between IGI<0.4, postpartum glucose intolerance and EGDMUnivariate analysisMultivariate analysisOR95% CIP valueOR95% CIP valueIGI<0.4OGTT 1-hour value≥1803.1771.575 to 6.4100.0013.1551.491 to 6.6760.003OGTT 2-hour value≥1532.3821.087 to 5.2210.0302.1870.946 to 5.0570.067EGDM2.3261.178 to 4.5920.0152.1311.015 to 4.4740.046Postpartum glucose intoleranceOGTT 1-hour value≥1801.9450.944 to 4.0050.0711.9920.912 to 4.3510.084OGTT 2-hour value≥1533.8101.462 to 9.9330.0063.7131.381 to 9.9810.009EGDM2.1181.026 to 4.3730.0431.8790.861 to 4.1030.113EGDM defined as GDM diagnosed at<24 weeks of gestation. The logistic regression model was used to calculate the OR with 95% CIs. The multivariate analysis was performed and adjusted by confounders of maternal age>35 years at delivery and last antepartum HbA1c.EGDM, GDM diagnosed at<24 weeks of gestation; GDM, gestational diabetes mellitus; HbA1c, glycosylated hemoglobin; IGI, insulinogenic index; OGTT, oral glucose tolerance test.In contrast, pregnant women with obesity (BMI>25 kg/m2; EGDM group, n=37; LGDM group, n=38) exhibited no significant differences in maternal parameters, pregnancy outcomes, postpartum insulin secretion or glucose intolerance (online supplemental table 1).SP110.1136/bmjdrc-2025-005114.supp1Supplementary dataDiscussion Currently, no consensus has been established on the appropriate diagnostic approach for EGDM. Moreover, although the effectiveness of screening and therapeutic interventions for GDM in early pregnancy has been partially demonstrated, 9 10 it still remains controversial whether this approach is more beneficial than screening in mid or late pregnancy.18 Although the WHO recommends the same diagnostic 75 g OGTT glucose thresholds for GDM in early pregnancy as the IADPSG criteria derived from HAPO,19 the prognostic value of these glucose levels in early pregnancy has not been established. A study by Yokoyama et al revealed that in Japan, universal screening for glucose intolerance in early pregnancy is performed more often than is risk-based selective screening, despite insufficient evidence. The study also showed that the same 75 g OGTT diagnostic criteria are generally applied in early and mid-to-late pregnancies.20 Similarly, a recent study by Simmons et al used these criteria for EGDM diagnosis.9 Based on these findings, our study applied the IADPSG 75 g OGTT criteria for both early and mid-to-late pregnancies. It also remains unclear whether using this criterion to predict postpartum glucose intolerance is better than using simpler screening methods, such as fasting glucose, random glucose, HbA1c, or glycated albumin. However, a Japanese study showed that women diagnosed with EGDM using this criterion had a higher rate of postpartum glucose intolerance than did those diagnosed later,12 suggesting that early diagnosis based on this criterion may be useful. Whether therapeutic intervention should be based on HbA1c in early pregnancy has also been explored; however, no clear view has emerged.21–23Several RCTs, systematic reviews, and meta-analyses have found no significant differences in most pregnancy outcomes between EGDM and LGDM. However, early diagnosis and intervention have been shown to significantly reduce the incidence of neonatal respiratory distress.8–10 In recent years, some comparative studies between EGDM and LGDM have been conducted in Japan. Nakanishi et al reported that early intervention in EGDM (diagnosed before 20 weeks of gestation) suppresses maternal GWG and increases the risk for SGA. However, they also stated that only in pregnant women with obesity, early treatment may reduce the risk for LGA.11 Therefore, the authors suggested that the current Japanese standard of early intervention for EGDM20 should be reconsidered.11 In contrast, a recent report using data from the Japanese prospective GDM registry did not find differences between EGDM (GDM diagnosed before 20 weeks of gestation) and LGDM (GDM diagnosed after 24 weeks of gestation) in pregnancy outcomes, including SGA and LGA. However, the incidence of postpartum glucose intolerance increased in women with EGDM.12 Postpartum insulin secretion capacity was not assessed in either study.The present study compared the differences in pregnancy outcomes, postpartum glucose intolerance, and insulin secretion capacity between EGDM and LGDM. The pregnant women included in this study, both in the EGDM and LGDM groups, underwent early screening and received appropriate and equivalent therapeutic interventions. Although maternal GWG was significantly reduced in EGDM, neonatal birth weight, and SGA incidence were not significantly different between the groups. Postpartum glucose intolerance was more common in the EGDM group than in the LGDM group; however, the difference was not significant. The results of this study, which showed no differences in most pregnancy outcomes, generally align with those of the RCT conducted by Simmons et al.9 In addition, the EGDM group had significantly lower initial insulin secretion than the LGDM group, whereas IGI showed comparable levels. Unlike Caucasians, in whom insulin resistance is thought to be predominant, Japanese pregnant women with GDM are thought to include both non-obese pregnant women with impaired insulin secretion capacity and obese pregnant women with insulin resistance. Therefore, we examined EGDM and LGDM in the non-obese and obese categories. In the obese subgroup, no differences in pregnancy outcomes, postpartum glucose intolerance, or insulin secretion capacity were found between the LGDM and EGDM groups. In the non-obese subgroup, the rate of postpartum initial insulin secretion was significantly lower, and the incidence of postpartum glucose intolerance was significantly higher in the EGDM group than in the LGDM group. Multivariate logistic regression analysis also showed that a diagnosis of GDM in early pregnancy was associated with decreased postpartum initial insulin secretion.Although none of the three studies included by Hannah et al7 in their systematic review of EGDM, all three have reported a higher incidence of postpartum glucose intolerance and DM in the EGDM group than in the LGDM group.24–26 However, the mechanism of increased glucose intolerance in EGDM is unclear. The recent CHIP-F (Cohort Study of Indian Women with Hyperglycemia in Pregnancy and their Families) study evaluated chronic (median of 32 months) postpartum glucose tolerance and found a higher incidence of DM, higher HOMA-IR and lower disposition index in EGDM than in LGDM.27 The results suggest insulin resistance and poor β-cell function in EGDM than in LGDM after delivery, leading to IGT. In addition, Quansah et al reported that the incidence of DM and pre-DM in the early postpartum period (6–8 weeks) was 7 and 3 times higher, respectively, in EGDM than in LGDM.28 The researchers thought that pre-existing glucose intolerance in EGDM may have contributed to this finding.Bozkurt et al studied insulin sensitivity and insulin secretion capacity in obese pregnant women with EGDM and LGDM in early pregnancy.29 They found that the EGDM group had reduced insulin sensitivity, whereas the LGDM group had insulin sensitivity comparable with that of the normal glucose tolerance group. Conversely, insulin secretion capacity was not different between the EGDM and LGDM groups, and both were reduced in comparison with the normal glucose tolerance group. Considering these findings, the authors reported that in addition to β-cell dysfunction during pregnancy, pre-existing insulin resistance may influence the timing of GDM onset in EGDM.29 In contrast, Lapolla et al calculated insulin sensitivity and β-cell function from 75 g OGTT results and reported that the EGDM group exhibited impaired β-cell function compared with the normal glucose tolerance group, whereas no β-cell dysfunction was detected in the LGDM group in early pregnancy.30In this study, the insulin secretory capacity and sensitivity were evaluated in the early postpartum period rather than during pregnancy. To our knowledge, this is the first study comparing early postpartum insulin secretion capacity between EGDM and LGDM. We found that among Japanese non-obese pregnant women, those with EGDM had lower insulin secretion capacity than those with LGDM and were more likely to have IGT. Early postpartum HOMA-IR was similar in the EGDM and LGDM groups, indicating comparable insulin sensitivity. This suggests that reduced insulin secretion capacity could contribute to the increased postpartum glucose intolerance in EGDM.Selective early screening for GDM is currently based on the notion that GDM results from insulin resistance associated with overweight or obesity. This screening targets women with established risk factors, such as obesity, a history of GDM, a family history of diabetes, or a history of delivering a macrosomic infant.5 31 However, the results of the current study suggest that non-obese pregnant women without these traditional risk factors may also develop GDM in early pregnancy, possibly due to primary β-cell dysfunction. These results raise concerns that current screening methods may fail to identify high-risk women with atypical GDM. Similarly, the International Federation of Gynecology and Obstetrics has pointed out that risk factors defined mainly in Western populations may not fully apply to Asian or other populations. In particular, many Asian women with GDM primarily exhibit impaired insulin secretion rather than insulin resistance,32 a concept supported by the findings of the present study.Well-designed multicenter studies are needed to confirm reproducibility and cost-effectiveness. The IADPSG has stated that screening for GDM at the first prenatal visit should depend on population characteristics and healthcare settings. Considering this and the present study’s findings, it may be reasonable to recommend that all pregnant women in Japan and other Asian countries undergo 75 g OGTT screening should they have a random blood glucose level of≥100 mg/dL before 24 weeks of gestation to enable EGDM diagnosis and intervention. In fact, numerous institutions across Japan have reportedly adopted this screening approach.20 Yokoyama et al demonstrated that EGDM in Japan is associated with a higher risk of postpartum glucose intolerance. Our study further revealed that non-obese women with EGDM tended to have reduced postpartum insulin secretion capacity. Although neither study found significant differences in maternal or neonatal outcomes, both studies showed that EGDM diagnosis enables the identification of women at risk for early postpartum glucose intolerance. While early therapeutic intervention has not been definitively associated with improved pregnancy outcomes, it may provide opportunities for patient education and lifestyle modifications that could help prevent the future development of type 2 diabetes. However, even within Asia, resource availability varies. Thus, adopting feasible, resource-appropriate strategies based on local experience is essential.This study has several limitations. First, this is a single-center, retrospective, observational study with a small number of cases. Second, this study was conducted in a perinatal medical center, which caters to high-risk pregnancies more than general medical facilities, which may have affected the pregnancy outcomes. Third, unlike Westerners, Japanese have lower initial insulin secretion.33 This makes direct comparisons with reports from other countries difficult, and the data may be more specific to Asians. However, even with these limitations, this study demonstrates the difference in clinical presentation between EGDM and LGDM in Japanese women. However, future multicenter studies are warranted to assess the effect of EGDM on perinatal outcomes.Conclusion To our knowledge, this is the first study that compared insulin secretion, pregnancy outcome, and postpartum glucose intolerance between EGDM and LGDM in Japanese women. Although further studies are needed to determine the appropriate diagnosis and timing of therapeutic intervention for EGDM and LGDM, the results of this study suggest that in non-obese pregnant women, EGDM may be associated with lower postpartum insulin secretion capacity compared with LGDM, which may increase the risk of developing IGT. Therefore, such risks should be considered during pregnancy.