2022 Volume 10 Issue 4 Pages 108-116
Aim: To clarify the relationship between the incidence of gestational diabetes mellitus (GDM) with or without hypertensive disorders of pregnancy (HDP) and the incidence of small-for-gestational-age (SGA) infants.
Methods: We conducted a 75-g oral glucose tolerance test between 24 and 28 weeks of gestation in 455 pregnant women who were classified into four groups according to their GDM and HDP status retrospectively.
Results: The number of SGA infants born to women with GDM+HDP was similar to that born to women with HDP alone. In the GDM group, the SGA rate in women with GDM+HDP was similar to that in women with GDM alone. However, in the non-GDM group, significantly more SGA infants were born to women with HDP alone than to women with no pathology. The mean birthweight SD in women with GDM+HDP was comparable to that in women with GDM alone. However, the birthweight SD in women with HDP alone was significantly lower than that in women with no pathology. By multivariate analysis, birthweight SD<−0.564 was a predictor of HDP in the non-GDM group, but not in the GDM group.
Conclusions: This study showed that GDM might offset the number of SGA infants induced by HDP.
Pedersen proposed a hyperglycemia–hyperinsulinemia theory, which states that hyperglycemia in pregnant women induces fetal hyperglycemia and hyperinsulinemia with insulin hypersecretion. Fetal hyperinsulinemia results in fetal abnormalities, excessive weight (macrosomia), and complication(s) after birth associated with immature fetal organs (left side of Figure 1).1,2,3) Thus, pregnant women with gestational diabetes mellitus (GDM) are at an increased risk of having large-for-gestational-age (LGA) infants. In a previous study of nulliparous women with singleton pregnancies who delivered at 37–40 gestational weeks (GW), women with either GDM or comorbid hypertensive disorders of pregnancy (HDP)/preeclampsia and GDM had significantly higher odds ratios of having LGA infants (1.51 and 1.85, respectively).4) Both HDP/preeclampsia and GDM were suggested to represent insulin-resistant states.
Hypothesis of gestational diabetes mellitus (GDM; left side) and hypertensive disorder of pregnancy (HDP; right side).
Pregnant women with HDP or preeclampsia are reportedly at an increased risk of having small-for-gestational-age (SGA) infants. According to the two-stage model of preeclampsia (Figure 1, right side),5,6) this begins with a poorly perfused placenta caused by insufficient remodeling of uterine spiral arteries at 8–18 GW. In normal pregnancies, the remodeling of these arteries increases blood flow from the mother to the fetus via the placenta, ensuring normal growth. With preeclampsia, however, the remodeling is dysfunctional and leads to decreased blood flow and SGA infants. Approximately 15% of women diagnosed with SGA will develop preeclampsia.7) Therefore, the 2018 revision of the Japanese classification of HDP includes peripartum hypertension with SGA infants, but without fetal abnormalities, in the diagnosis of preeclampsia.8)
The present study aimed to clarify the relationship between the incidence of GDM with or without HDP and the incidence of SGA infants.
This retrospective, single-hospital, case-control study was conducted between January 2011 and December 2017. We collected data on the characteristics of pregnant women who underwent a 75-g oral glucose tolerance test (OGTT) between 24 and 28 GW at the Hokkaido University Hospital if they delivered from 28 GW onward. We excluded women with diabetes mellitus, including those with overt diabetes in pregnancy, chronic hypertension (CH), and fetal abnormalities. Women were classified into the following groups: (1) GDM group, comprising subgroups with GDM only (GDM alone group) and both pathologies (GDM+HDP group), and (2) non-GDM group, comprising subgroups with HDP only (HDP alone group) and no pathology (non-pathology group) as a control.
Diagnosis and treatment of GDMIn Japan, all pregnant women undergo hyperglycemia screening during the first trimester via measurement of casual plasma glucose (PG) levels. Women with values higher than the cutoff levels determined by each institution (95–100 mg/dl) undergo a 75-g OGTT to detect GDM or overt diabetes in pregnancy as early as possible.9) In the present study, we excluded women with GDM or overt diabetes in pregnancy diagnosed at this first trimester screening. Next, the diagnosis of GDM is made beginning in the second trimester when women undergo a 50-g glucose challenge test at 24–28 GW and have a PG level ≥140 mg/dl (7.8 mmol/L) 1 h later and then have a positive test result after a 75-g OGTT at 24–28 GW.8) A GDM diagnosis was made based on the Japanese modification of the criteria of the International Association of Diabetes and Pregnancy Study Groups (IADPSG).10) At 0 (fasting), 1, and 2 h, the PG level is required to equal or exceed 92, 180, and 153 mg/dl, respectively. We also measured reference immunoreactive insulin at 0 (fasting), 1, and 2 h. Furthermore, we measured HbA1c levels of all pregnant women with or without GDM at 8–12 and 24–28 GW. Finally, we diagnosed GDM on the basis of a negative 75-g OGTT 12 weeks postpartum. If the test was positive, type 2 diabetes mellitus was diagnosed.
Target PG levels for GDM treatment are <100 mg/dl before meals, and <120 mg/dl 2 h after meals.9) In Japan, recommendations for the diagnosis and treatment of GDM stipulate that no medications used to treat GDM, except for insulin, should be used by pregnant women with poor control of their PG by self-measurement of blood glucose.9) Thus, we used insulin alone in the present study to treat all women with GDM.
Diagnosis of HDPHDP include diagnoses of gestational hypertension and preeclampsia.8) Thus, we excluded any women in the HDP group with CH. We defined hypertension as a systolic blood pressure ≥140 mmHg and/or a diastolic blood pressure ≥90 mmHg. We defined gestational hypertension as hypertension occurring on and after 20 GW in the absence of significant proteinuria defined as a spot urinary protein/creatinine ratio greater than 0.27 g or a 0.3 g protein loss per 24-h urine collection. We diagnosed preeclampsia in women who developed both hypertension and significant proteinuria at or after 20 GW. The 2018 Japanese criteria for HDP define preeclampsia as the onset of hypertension beginning at 20 GW with one or more other maternal complication(s) following proteinuria, including liver dysfunction without underlying liver diseases, perinatal progressive kidney dysfunction, perinatal cerebral stroke or neuropathy, blood coagulation disorder, and uterus/placental dysfunction (fetal growth restriction, stillbirth without fetal chromosomal abnormality, or fetal anomalies).8) As the present study aimed to investigate fetal growth, we defined preeclampsia as hypertension plus proteinuria.9)
Diagnosis of SGA and LGA infants and valuation basis of placental weight and the birthweight/placental weight ratio SDBirthweights and birth heights were measured immediately after delivery. The birthweight SD and birth height SD were calculated using normative data for Japanese infants.11) Birthweight SDs were calculated based on GW at birth, primipara or multipara, and male or female infants, and birth height SDs were calculated based on GW at birth. We defined SGA as a birthweight SD≤−1.5 plus a birth height SD≤−1.5. We defined LGA as a birthweight SD≥+1.5 plus a birth height SD≥+1.5. We measured placental weight immediately after delivery. The placental weight SD and birthweight/placental weight ratio SD were calculated using normative data for the birthweights and placental weights of normal Japanese pregnant women.12)
Statistical analysesData are presented as mean±SD or frequencies. Statistical analyses were conducted using JMP Pro, version 14.0 (SAS Institute Inc., Cary, NC, USA). We used the Tukey–Kramer honestly significant difference test or student’s t-test to compare means and Fisher’s exact test to compare categorical variables. Pearson’s product-moment correlation coefficient was used to measure linear correlations between two variables. In all analyses, P<0.05 indicated statistical significance. In linear correlations, a correlation coefficient (r) >0.25 or <−0.25 was required for statistical significance. A stepwise multiple logistic regression analysis was used in the multivariate analysis to compare parameters with P<0.1 in the univariate analysis. Receiver operating characteristic curves were used to evaluate the ability of parameters to identify the onset of perinatal/neonatal complications.
Ethics approval and consent to participateThis study was conducted with approval from the Hokkaido University Hospital Institutional Review Board (No. 018-297). All women provided informed consent to participate in this study.
Of the 2,314 women who delivered at ≥28 GW during the required period in the present study, 240 women (10.4%) had GDM, including 52 with CH and/or fetal abnormalities. The subject flow chart is shown in Figure 2. Of the 589 women who delivered at ≥28 GW, we excluded 16 women with CH (2.7%, including one woman with both CH and fetal abnormalities) and 118 women with fetal abnormalities (20.0%). The remaining 455 (77.2%) were included in the present study and divided into GDM (n=188, 41.3%) and non-GDM (n=267, 58.7%) groups. These groups were then subdivided by the presence or absence of HDP. The GDM group was divided into the GDM alone group (n=19, 10.1%) and GDM+HDP group (n=169, 89.9%), and the non-GDM group was divided into the HDP alone group (n=23, 8.6%) and non-pathology group (n=244, 91.4%).
Study flow chart.
*: included one woman with both chronic hypertension and fetal abnormalities. Abbreviations: GW, gestational weeks; GDM, gestational diabetes mellitus; HDP, hypertensive disorders of pregnancy.
Characteristics of each group are shown in Table 1. The incidence of GW at HDP diagnosed <34 weeks was similar between GDM+HDP (n=3, 15.7%) and HDP alone (n=8, 34.8%) groups. Furthermore, the incidence of preeclampsia was similar between GDM+HDP (n=7, 36.8%) and non-GDM (n=13, 56.5%) groups.
| GDM | Non-GDM | |||||
|---|---|---|---|---|---|---|
| Overall (n=188; 100%) | GDM+HDP (n=19; 10.1%) | GDM alone (n=169; 89.9%) | Overall (n=267; 100%) | HDP alone (n=23; 8.6%) | Non‑pathology (n=244; 91.4%) | |
| Age (years) | 34.6±4.9 | 34.5±6.1 | 34.6±4.8 | 34.2±5.2 | 36.8±4.1† | 34.0±5.2 |
| Primipara (n) | 109 (58.9%) | 13 (68.4%) | 96 (56.8%) | 133 (49.8%) | 12 (52.2%) | 121 (49.6%) |
| Smoking | ||||||
| Before pregnancy (n) | 51 (27.1%) | 5 (26.3%) | 46 (27.2%) | 54 (20.2%) | 6 (26.1%) | 48 (19.7%) |
| During pregnancy (n) | 3 (1.6%) | 0 (0.0%) | 3 (1.8%) | 11 (4.1%) | 3 (13.0%) | 8 (3.3%) |
| BMI | ||||||
| Before pregnancy (kg/m2) | 22.9±5.1* | 25.8±7.3†‡ | 22.6±4.8 | 21.6±3.2 | 21.3±2.6 | 21.6±3.3 |
| <18.5 (n) | 25 (13.3%) | 0 (0.0%) | 25 (14.8%) | 32 (12.0%) | 4 (17.4%) | 28 (11.5%) |
| 18.5≤, <25.0 (n) | 115 (61.2%) | 9 (47.4%) | 106 (62.7%)§ | 203 (76.0%) | 17 (73.9%) | 186 (76.2%) |
| 25.0≤ (n) | 48 (25.5%)* | 10 (52.6%)† | 38 (22.5%)§ | 32 (12.0%) | 2 (8.7%) | 30 (12.3%) |
| At delivery (kg/m2) | 25.9±5.1 | 28.9±6.9†‡ | 25.6±4.7 | 25.3±3.5 | 25.6±3.2 | 25.3±3.5 |
| BMI gain during pregnancy (kg/m2) | 3.0±1.7* | 3.1±1.9 | 3.0±1.6‡ | 3.7±1.8 | 4.3±2.2 | 3.7±1.8 |
| Body weight gain during pregnancy (kg) | 7.6±4.1* | 7.8±4.5 | 7.6±4.1‡ | 9.3±4.6 | 10.7±7.8 | 9.2±4.5 |
| HbA1c | ||||||
| At 1st trimester (%) | 5.34±0.33* | 5.49±0.28† | 5.32±0.33 | 5.25±0.29 | 5.32±0.29 | 5.24±0.29 |
| At 2nd trimester (%) | 5.36±0.36* | 5.49±0.36 | 5.35±0.35 | 5.26±0.38 | 5.24±0.32 | 5.26±0.39 |
| At delivery (%) | 5.54±0.39 | 5.72±0.43 | 5.53±0.38 | No data | No data | No data |
| Treatment with insulin | 60 (31.9%) | 5 (26.3%) | 55 (32.5%) | |||
| Insulin dose at delivery (U/day) | 8.4±16.3 | 4.1±7.6 | 8.9±16.9 | |||
| GW at HDP diagnosis (weeks) | 36.6±3.2 | 36.6±3.2 | — | 34.5±4.2 | 34.5±4.2 | — |
| <34 GW (n) | 3 (1.5%) | 3 (15.7%) | — | 8 (3.0%) | 8 (34.8%) | — |
| Preeclampsia (n) | 7 (3.7%) | 7 (36.8%) | — | 13 (4.9%) | 13 (56.5%) | — |
| GW at delivery (weeks) | 38.3±1.7 | 38.2±1.6‡ | 38.3±1.7 | 38.1±2.3 | 35.9±3.3† | 38.3±2.1 |
| <37 GW (n) | 20 (10.6%) | 2 (10.5%) | 18 (10.7%) | 40 (15.0%) | 9 (39.1%)† | 31 (12.7%) |
| Cesarean section (n) | 99 (52.7%) | 11 (57.9%) | 88 (52.1%) | 136 (50.9%) | 14 (60.9%) | 122 (50.0%) |
| LGA infants (n) | 2 (1.1%) | 1 (5.3%) | 1 (0.6%) | 7 (2.6%) | 0 (0.0%) | 7 (2.9%) |
| SGA infants (n) | 13 (6.8%) | 3 (15.8%) | 10 (5.9%) | 15 (5.6%) | 4 (17.4%)† | 11 (4.5%) |
| GW when fetal growth restriction suspected (weeks) | 32.5±5.3 | 36.3±2.5 | 31.3±5.5 | 34.5±3.8 | 31.8±1.7 | 35.5±3.9 |
| <34 GW (n) | 8 (4.3%) | 0 (0.0%) | 8 (4.7%) | 7 (2.6%) | 3 (13.0%)† | 4 (1.6%) |
| Stillbirth or END (n) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) |
Data are shown as mean±SD or as number (percentage). The 1st trimester was from 8 to 12 gestational weeks. The 2nd trimester was from 24 to 28 gestational weeks. BMI, body mass index; GDM, gestational diabetes mellitus; HDP, hypertension disorders of pregnancy; TSH, thyroid-stimulating hormone; T4, thyroxine; IRI, immunoreactive insulin; OGTT, oral glucose tolerance test; PG, plasma glucose; GW, gestational weeks; SGA, small-for-gestational-age; LGA, large-for-gestational-age; END, early neonatal death
*: P<0.05, GDM vs. non-GDM group; †: P<0.05 vs. without HDP in same group; ‡: P<0.05 vs. HDP alone group; §: P<0.05 vs. non-pathology group.
Maternal age and GW at delivery were higher in the non-GDM group compared with the HDP alone group (P=0.0111 and P<0.0001, respectively, Figure 3A). The frequency of preterm delivery in the HDP alone group was significantly higher than that in the non-pathology group (P=0.0027). However, the frequency of preterm delivery was similar between GDM+HDP and GDM alone groups.
Parameters of women with and without HDP in GDM and non-GDM groups.
(A) Age (○) and gestational weeks at delivery (●), (B) BMI before pregnancy (○) and at delivery (●), (C) BMI gain during pregnancy (○) and body weight gain during pregnancy (●), (D) HbA1c at first trimester (○), second trimester (●), and delivery (■).
Data are shown as means. The first trimester was approximately 8–12 gestational weeks, the second trimester was approximately 24–28 gestational weeks, and delivery was approximately 37–39 gestational weeks.
Abbreviations: GW, gestational weeks; BMI, body mass index; GDM, gestational diabetes mellitus; HDP, hypertensive disorders of pregnancy; SD, standard deviation.
In the non-GDM group, first trimester HbA1c levels were significantly higher in the GDM+HDP group than in the GDM alone group (P=0.0444, Figure 3B), according to the univariate analysis. In both GDM and non-GDM groups, the incidence of infant hypoglycemia was higher in the HDP group than in the non-HDP group (data not shown).
In the GDM group, maternal body mass index (BMI) before pregnancy and at delivery was significantly higher in the GDM+HDP group than in the GDM alone group (P=0.0104 and P=0.0186, respectively, Figure 3C).
Women were categorized based on their pre-pregnancy BMI into the following groups: thin <18.5; 18.5≤ normal <25.0; 25.0≤ obese (Table 1). There were significantly more obese women in the GDM group than in the non-GDM group (P=0.0003). In the GDM group, there were significantly more obese women with HDP than without HDP (P=0.0097); this was not the case in the non-GDM group. Meanwhile, there were significantly fewer normal-weight women with HDP than without HDP (P=0.0042).
BMI gain and body weight gain during pregnancy were significantly lower in the GDM group than in the non-GDM group (P <0.0001 and P <0.0001, respectively). Moreover, these were significantly lower in the GDM alone group than in the non-pathology group (P=0.0002 and P=0.0003, respectively, Figure 3D). Rates of smoking before and during pregnancy were similar across the four groups (Table 1).
Comparison of SGA rate, birthweight SD, birth height SD, placental weight SD, and birthweight/placental weight ratio SD between groupsIncidences of SGA and LGA are shown in Table 1. Incidences of LGA were similar across the four groups. The incidence of SGA concurrent with HDP in the GDM group (15.8%) and the non-GDM group (17.4%) were similar. In the GDM group, the incidence of SGA concurrent with HDP was similar to the incidence of SGA without HDP (5.9%, P=0.1303). However, in the non-GDM group, the incidence of SGA with HDP was significantly higher than SGA in the non-pathology group (4.5%, P=0.0303). Furthermore, in the non-GDM group, the incidence of GW at fetal growth restriction suspected by ultrasonography <34 weeks in the HDP alone group (13.0%) was significantly higher than that in the non-pathology group (1.6%, P=0.0156).
Birth height SD was similar between GDM+HDP and GDM alone groups (−0.29±1.09 vs. −0.42±0.94) (Figure 4A). However, birth height SD was significantly lower in the HDP alone group than that in the non-pathology group (−0.90±1.01 vs. −0.38±0.90, P<0.0001).
Parameters of infants and placentas of women with and without HDP in GDM and non-GDM groups.
(A) Birth height SD, (B) Birthweight SD, (C) Placental weight SD, and (D) Birthweight/Placental weight rate SD.
Abbreviations: GDM, gestational diabetes mellitus; HDP, hypertensive disorders of pregnancy; SD, standard deviation.
In the GDM group, birthweight SD was similar between GDM+HDP and GDM alone groups (0.28±1.26 vs. 0.12±1.03) (Figure 4B). However, in the non-GDM group, birthweight SD was significantly lower in the HDP alone group than in the non-pathology group (−0.65±1.04 vs. 0.28±1.00, P<0.0001).
In the GDM group, placental weight SD was similar between GDM+HDP and GDM alone groups (0.53±0.90 vs. 0.31±1.05) (Figure 4C). However, in the non-GDM group, placental weight SD was significantly lower in the HDP alone group than in the non-pathology group (−0.32±1.26 vs. 0.20±0.95, P=0.0125).
In the non-GDM group, birthweight/placental weight ratio was lower in the HDP alone group than in the non-pathology group (4.72±1.25 vs. 5.13±0.90, P=0.0465) (Figure 4D). In the GDM group, birthweight/placental weight ratio was similarly lower in the GDM+HDP group than in the GDM alone group (4.59±1.18 vs. 5.14±0.90, P=0.0096).
According to the multivariate analysis (with parameters: primipara vs. multipara, age ≥35 or <35 years, pre-pregnancy BMI ≥ or <25, and presence/absence of HDP), only the presence of HDP was correlated with the incidence of SGA in the non-GDM group (P=0.0323 and P=0.0286, respectively). However, none of the variables were associated with the incidence of SGA in the GDM group.
Predictors of HDP and/or SGA infants among women with or without GDMMultivariate analysis was conducted with the following variables: cutoff values for birth height SD, birthweight SD, placental weight SD, birthweight/placental weight SD, first trimester HbA1c levels, pre-pregnancy BMI, and body weight gain during pregnancy. The analysis yielded predictors of GDM+HDP among all 455 women, HDP among women with GDM, and HDP among women without GDM (Table 2). Cutoff values for the variables were determined by receiver operating characteristic curves.
| Odds ratio | 95% CI | Sensi- tivity | Speci- ficity | PPV | NPV | Uni‑variate analysis P-value | Multi‑variate analysis P-value | |
|---|---|---|---|---|---|---|---|---|
| GDM+HDP among all 455 women | ||||||||
| Birth height SD <0.188 | 2.38 | 0.93–6.08 | 0.421 | 0.766 | 0.073 | 0.968 | 0.524 | |
| Birthweight SD <0.881 | 2.76 | 1.08–7.06 | 0.421 | 0.791 | 0.081 | 0.969 | 0.043 | |
| Placental weight SD <0.067 | 2.98 | 0.97–9.12 | 0.789 | 0.444 | 0.058 | 0.980 | 0.058 | |
| Birthweight/Placental weight SD <0.031 | 3.34 | 1.09–10.2 | 0.789 | 0.471 | 0.061 | 0.981 | 0.033 | |
| First trimester HbA1c levels≥5.2% | 8.08 | 1.06–61.8 | 0.941 | 0.336 | 0.074 | 0.990 | 0.016 | 0.099 |
| Pre-pregnancy BMI≥22.2 | 5.29 | 1.87–15.0 | 0.737 | 0.664 | 0.085 | 0.983 | 0.001 | 0.082 |
| Body weight gain during pregnancy ≥1.6 kg | 6.19 | 1.87–20.6 | 0.737 | 0.408 | 0.123 | 0.932 | 0.393 | |
| HDP among women with GDM | ||||||||
| Birth height SD <0.188 | 2.42 | 0.91–6.45 | 0.421 | 0.769 | 0.170 | 0.922 | 0.560 | |
| Birthweight SD <0.881 | 3.11 | 1.16–8.37 | 0.421 | 0.811 | 0.200 | 0.926 | 0.034 | |
| Placental weight SD <0.067 | 2.59 | 0.82–8.13 | 0.789 | 0.408 | 0.130 | 0.945 | 0.135 | |
| Birthweight/placental weight SD <0.290 | 2.94 | 0.82–10.5 | 0.842 | 0.355 | 0.128 | 0.952 | 0.123 | |
| First trimester HbA1c levels ≥5.5% | 2.72 | 0.99–7.52 | 0.529 | 0.707 | 0.173 | 0.929 | 0.057 | |
| Pre-pregnancy BMI≥22.2 | 4.48 | 1.54–13.0 | 0.737 | 0.615 | 0.177 | 0.954 | 0.006 | |
| Body weight gain during pregnancy ≥6.2 kg | 1.93 | 0.67–5.61 | 0.737 | 0.408 | 0.123 | 0.932 | 0.864 | |
| HDP among women without GDM | ||||||||
| Birth height SD <−1.354 | 4.63 | 1.85–11.6 | 0.375 | 0.885 | 0.243 | 0.935 | 0.0107 | 0.3246 |
| Birthweight SD <−0.564 | 10.2 | 4.03–25.6 | 0.652 | 0.844 | 0.283 | 0.923 | <0.001 | 0.0010 |
| Placental weight SD <−0.313 | 3.64 | 1.51–8.80 | 0.609 | 0.701 | 0.161 | 0.950 | 0.004 | 0.0184 |
| Birthweight/Placental weight SD <0.805 | 2.48 | 0.99–6.22 | 0.348 | 0.823 | 0.157 | 0.930 | 0.056 | 0.0177 |
| First trimester HbA1c levels ≥5.5% | 2.04 | 0.64–6.53 | 0.357 | 0.786 | 0.143 | 0.924 | 0.312 | |
| Pre-pregnancy BMI ≥18.9 | 4.20 | 1.92–9.20 | 0.292 | 0.836 | 0.149 | 0.923 | 0.001 | |
| Body weight gain during pregnancy ≥8.3 kg | 3.01 | 1.08–8.35 | 0.783 | 0.455 | 0.119 | 0.957 | 0.139 |
GDM, gestational diabetes mellitus; HDP, hypertensive disorder of pregnancy; HbA1c, hemoglobin A1c; BMI, body mass index; SD, standard deviation; CI, confidence interval; PPV, positive predictive value; NPV, negative predictive value.
Multivariate analysis identified no predictors of GDM+HDP among all 455 women. Moreover, we found no predictors of HDP in the GDM group. However, birthweight SD<−0.564 was a predictor of HDP in the non-GDM group. Neither maternal pre-pregnancy BMI nor first trimester HbA1c levels predicted the incidence of HDP by multivariate analysis.
Our results indicate that mean birthweight SD was lower in women with HDP alone than in women without HDP in the non-GDM group; however, in the GDM group, the values were similar in women with and without HDP. Moreover, multivariate analysis found no predictors of HDP in the GDM group; however, birthweight SD<−0.564 was a positive predictor of HDP in the non-GDM group. Thus, we hypothesize that GDM might offset SGA induced by HDP (Figure 1). To our knowledge, the present study is the first to compare birthweight SD and birth height SD in women with GDM+HDP, GDM alone, or HDP alone.
In pregnant women with GDM+HDP, fetal growth might be determined by the balance between increased PG supply from the mother to the fetus via heavier placentas, according to the hyperglycemia–hyperinsulinemia theory,1,2,3) and decreased PG supply due to placental dysfunction in women with HDP, according to the two-stage model for HDP.5,6) Thus, we propose that the effects of GDM could counteract the incidence of SGA induced by HDP. Considering that the infants of women with HDP are at risk of being SGA, whereas those of women with GDM are at risk of being LGA, most infants of women with both pathologies may have normal birthweights. We describe a similar phenomenon occurring between diabetes mellitus and HDP in a previous study.13)
In a study of 93,034 women with vaginal deliveries of singleton infants in Japan, the incidence of low placental weight Z-scores was significantly higher in women who delivered SGA infants (52.3%) than in women who delivered appropriate-for-gestational-age infants (7.3%, P<0.0001) or LGA infants (0.4%, P<0.0001).14) Thus, women who delivered smaller infants had a higher risk of a small placenta. However, the rates of low birthweight/placental weight Z-scores were similar (10.7%, 6.9%, and 7.2%).14) We instead relied on the birthweight/placental weight ratio SD at delivery (Figure 4D). Placental weight SD was also significantly lower in women with HDP alone than in women with no pathology in the non-GDM group. Placental weight SD was similar in women with GDM+HDP and women with GDM alone (Figure 4C). Thus, GDM might offset the small placenta of SGA infants induced by HDP. However, based on the results of the present multivariate analysis, low placental weight SD is not a predictor of HDP.
The present study has three major strengths. First, pregnant women with CH and fetal abnormalities were excluded. Second, bias was not introduced by the various methods used to treat GDM. Third, all participants had not received daily low-dose aspirin to prevent the onset of preeclampsia. It was important to exclude pregnant women with CH because of the results of a previous multicenter study in the United States that reported a similar rate of SGA infants among women with CH and women with or without GDM.15) We also excluded women with fetal abnormalities because some abnormalities (e.g., trisomy 18 or 21) are known to result in LGA infants. We removed any bias that could have been introduced by different treatments for GDM by excluding women with preexisting diabetes. We followed the Japanese recommendations for diagnosing and treating GDM and used insulin alone.9)
The present study was limited by the small cohort sizes and retrospective cohort design. In particular, the numbers of women in GDM+HDP and HDP alone groups were small. All pregnant women in the case and control groups underwent a 75-g OGTT at 24–28 GW. However, women in the non-GDM group may also have had mild hyperglycemia based on the criteria for a positive 50-g glucose challenge test. The IADPSG reported a higher incidence of preeclampsia in women with GDM (9.1%) than in women without GDM (4.5%).10) In the present study, the incidence of preeclampsia was similar in GDM (10.1%) and non-GDM groups (8.6%). External validity of the study is limited for the following two reasons: 1) birthweight SD was calculated using normative data for Japanese newborns and 2) GDM was diagnosed according to the Japanese modified criteria of the IADPSG. Although the World Health Organization has recommended it as the international criteria for GDM, the United Kingdom,16) the United States,17,18) and other countries use their original criteria for GDM.
In conclusion, GDM might offset the number of SGA infants induced by HDP. The number of SGA infants born to women with HDP was significantly higher than that born to women without HDP in the non-GDM group; however, these incidences were similar in women with GDM.
BMI, body mass index; CH, chronic hypertension; CI, confidence interval; GDM, gestational diabetes mellitus; GW, gestational weeks; HDP, hypertensive disorders of pregnancy; IADPSG, International Association of Diabetes and Pregnancy Study Groups; LGA, large-for-gestational-age; OGTT, oral glucose tolerance test: PG, plasma glucose; SD, standard deviation; SGA, small-for-gestational-age.
The authors thank Enago (www.enago.jp) for the English language review.
The first author received support from the Manpei Suzuki Diabetes Foundation (Tokyo, Japan) to publish this article.
The authors declare that they have no conflict of interest regarding the publication of this study.
