Introduction
Gestational diabetes mellitus (GDM) has traditionally been defined as an abnormal glucose tolerance that is first identified during pregnancy and has been shown to be associated with adverse maternal and neonatal outcome when untreated, including accelerated fetal growth, macrosomia, and short- and long-term complications.
Women with twin pregnancies are at increased risk for pregnancy complications including preterm birth, hypertensive complications, as well as fetal growth restriction. The mechanisms responsible for the higher rate of fetal growth restriction in twin pregnancies are complex, but one of the factors that may play a role is the limited maternal resources that are being shared by >1 fetus. Based on that, it may be hypothesized that the fetal effects of the mild hyperglycemia associated with GDM with respect to accelerated fetal growth may be less pronounced in twin gestations or alternatively may even have a beneficial role in decreasing the risk of fetal growth restriction in these pregnancies.
This question regarding the effects of GDM on fetal growth and pregnancy outcome in twins has been addressed by a relatively small number of studies of small sample size that produced conflicting results, and the interpretation of these studies is further limited by several factors. First, one of the major difficulties in studying the effects of GDM on pregnancy outcome in twins is that many of the complications associated with GDM are less relevant for twin gestations due to the lower gestational age at delivery, increased risk for hypertensive complications, and high rate of malpresentation and cesarean delivery. Thus, adverse outcomes such as macrosomia and shoulder dystocia are rare in twin gestations, while other complications that have been reported by previous studies such as cesarean delivery, preeclampsia, and neonatal complications such as hypoglycemia and jaundice are more common in twins irrespective of GDM. We believe that focusing on fetal growth as the primary outcome can overcome this limitation as fetal growth is not affected by prematurity and may therefore reliably reflect the presence of fetal hyperglycemia and hyperinsulinemia in these pregnancies even in cases of preterm deliveries.
The second factor that limits the interpretation of available studies relates to the fact that all women with GDM included in these studies were monitored and treated to control maternal blood glucose levels, thereby leading to masking or at least underestimation of the true impact of GDM on pregnancy outcome in these cases. Since randomizing women with GDM to receive no treatment would be considered unethical given the presence of level I evidence regarding the benefits of treatment in these cases, the only way to address this limitation would be by assessing pregnancy outcome in women with evidence of glucose intolerance that is below the level that is consistent with the diagnosis of GDM. Therefore, in the current study we included a subgroup of women who did not meet 2008 Canadian Diabetes Association (CDA) criteria for the diagnosis of GDM (which were used to diagnose GDM during the study period) but did meet the more recent International Association of the Diabetes and Pregnancy Study Groups (IADPSG) criteria for the diagnosis of GDM, which are lower than the thresholds recommended by the CDA ( Table 1 ).
| 2008 CDA | IADPSG | |
|---|---|---|
| Strategy | 2 Steps | 1 Step |
| Screening test | 1-h 50-g GCT | None |
| Screen negative | <140 mg/dL (7.8 mmol/L) | |
| Screen positive (followed by OGTT) | 140–185 mg/dL (7.8–10.2 mmol/L) | |
| Diagnostic of GDM | ≥185 mg/dL (10.3 mmol/L) | |
| Definitive test | Fasting 2-h 75-g OGTT | |
| Fasting threshold | ≥95 mg/dL (5.3 mmol/L) | ≥92 mg/dL (5.1 mmol/L) |
| 1-h threshold | ≥191 mg/dL (10.6 mmol/L) | ≥180 mg/dL (10.0 mmol/L) |
| 2-h threshold | ≥160 mg/dL (8.9 mmol/L) | ≥153 mg/dL (8.5 mmol/L) |
| Diagnosis of GDM | ≥2 Abnormal values | ≥1 Abnormal values |
The aim of the current study was to assess the impact of GDM and milder degrees of glucose intolerance on fetal growth and pregnancy outcome in twin pregnancies while trying to overcome the limitations described above.
Materials and Methods
Study population
This was a retrospective cohort study of all women with twin pregnancies who underwent screening for GDM in a single tertiary referral center (Sunnybrook Health Sciences Center, Toronto, Ontario, Canada) from October 2003 through December 2014. Women with any of the following conditions were excluded from the study: pre-GDM, incomplete or nonstandard screening for GDM (eg, women who had the 2-hour 75-g oral glucose tolerance test [OGTT] with no prior screening with a nonfasting 1-hour 50-g glucose challenge test [GCT] or following a normal 50-g GCT finding), delivery <28 weeks of gestation (based on the assumption that any adverse effects of GDM will be minimal <28 weeks of gestation), complicated monochorionic-diamniotic twins (ie, twin-to-twin transfusion syndrome, selective intrauterine growth restriction, or twin-anemia-polycythemia sequence), monochorionic-monoamniotic twins, stillbirth or reduction of ≥1 fetuses, or genetic or structural fetal anomalies. The study was approved by the institutional research ethics board.
Definitions
The 2008 CDA guidelines for the screening and diagnosis of GDM were followed throughout the study period ( Table 1 ). These guidelines are based on a 2-step screening strategy. The first step involves universal screening at 24-28 weeks of gestation using the 1-hour 50-g GCT. The results of the 50-g GCT are interpreted as a negative screen, positive screen, or diagnostic for GDM ( Table 1 ). Patients with positive screen go on to have the diagnostic fasting 2-hour 75-g OGTT. The test is interpreted as positive for GDM if at least 2 of the values exceed their corresponding thresholds ( Table 1 ).
The IADPSG recommendations for the screening and diagnosis of GDM are based on a 1-step approach using a fasting state 2-hour 75-g OGTT ( Table 1 ), and the test is interpreted as positive for GDM if ≥1 of the values surpasses the corresponding threshold ( Table 1 ). All 3 thresholds of the 75-g OGTT recommended by the IADPSG are lower than those in the 2008 CDA recommendations ( Table 1 ).
All women diagnosed with GDM at our center are followed by a team that consists of a specialist in maternal fetal medicine, endocrinologist, diabetes nurse, and nutritional consultant experienced in the management of diabetes in pregnancy. Patients with GDM are seen in the diabetes clinic every 1-2 weeks from the time of diagnosis up to delivery. All patients monitor blood glucose levels 4 times a day using a glucometer. When patients do not meet the target glucose levels (fasting <95 mg/dL [<5.3 mmol/L] and 2-hour postprandial <121 mg/dL [<6.7 mmol/L]) after 1-2 weeks of diet management, treatment with insulin is started and the dose is titrated until adequate glycemic control is achieved. All women with twin pregnancies followed up in our medical center undergo ultrasound examination every 2 weeks during the third trimester until delivery. The routine sonographic evaluation on each of these examinations include presentation, amniotic fluid level (expressed as maximal vertical pocket), biophysical profile, and fetal biometry including biparietal diameter, head circumference (HC), abdominal circumference (AC), and femur length. The presence of asymmetric growth pattern, as reflected by high AC or AC/HC ratio, is used as an additional measure of the degree of glycemic control, and interpreted according to Campbell and Thoms.
Birthweight percentiles were calculated according to national growth curves based on sex-specific birthweights of singleton pregnancies. We chose to use singleton-based growth curves given the conflicting results regarding the benefit of twin-specific charts and the fact that the validation of currently available twin-specific charts is still limited.
Data collection
Data regarding all twin pregnancies were extracted from the institutional comprehensive perinatal database. This database contains information on demographic, medical, and obstetrical history, and information on the current pregnancy including initial and final number of fetuses, pregnancy dating, complications during pregnancy, mode of delivery, and neonatal outcome. The information on the results of the 50-g GCT and 75-g OGTT was extracted from our institutional laboratory database, which was cross-linked with the perinatal database. Finally, the results of the last third-trimester ultrasound were extracted from the database of our obstetric-ultrasound unit and were merged with the perinatal database.
Data analysis
Pregnancy outcome, fetal biometry, and neonatal birthweight were compared among 4 groups of women with increasing degree of glucose intolerance. (1) GCT-NEGATIVE group included women who had a negative screening with 50-g GCT. This group represented women with normal glucose tolerance and was used as the control/reference group. (2) OGTT-NEGATIVE group included women who had a positive 50-g GCT screening but the subsequent 75-g OGTT was negative according to both the IADPSG and the 2008 CDA criteria. These women were previously shown to be at increased risk for pregnancy complications attributed to abnormal glucose tolerance. (3) GDM-IADPSG group included women with a positive 50-g GCT screening that was followed by a 75-g OGTT that was positive according to the IADPSG criteria but not the 2008 CDA criteria. These women were previously shown to have an abnormal glucose tolerance and an increased risk of adverse pregnancy outcome and considered to have GDM according to the IADPSG criteria. Because these women were not considered to have GDM during the study period they did not monitor their glucose level and were not subjected to any form of dietary or medical interventions to control blood glucose levels. Thus, the outcome of pregnancies in this group can be considered to reflect the true impact of untreated mild GDM in twin gestations. (4) GDM-CDA group included women with a positive 50-g GCT that was followed by a 75-g OGTT that was positive according to the 2008 CDA criteria as well as women with a diagnostic 50-g GCT. Women in this group were monitored and treated according to the GDM management protocol as described above, which may lead to masking or underestimation of the true impact of GDM on pregnancy outcome in twin gestations. In all the analyses, the GCT-NEGATIVE group, consisting of women with normal glucose tolerance, was used as the reference group.
The χ 2 and independent sample Student t test were used for comparison of categorical and continuous variables, respectively, among each of the 3 groups with abnormal glucose tolerance and the GCT-NEGATIVE reference group. Generalized logistic regression model was used to assess the association between the degree of glucose intolerance and accelerated fetal growth while adjusting for potential confounders and accounting for interclass correlation within sets of twins. We accounted for multiple comparisons between the groups using the Bonferroni correction. Differences were considered to be statistically significant for a P value < .002. The statistical analysis was performed using software (SPSS, Version 22; IBM Corp, Armonk, NY).
Materials and Methods
Study population
This was a retrospective cohort study of all women with twin pregnancies who underwent screening for GDM in a single tertiary referral center (Sunnybrook Health Sciences Center, Toronto, Ontario, Canada) from October 2003 through December 2014. Women with any of the following conditions were excluded from the study: pre-GDM, incomplete or nonstandard screening for GDM (eg, women who had the 2-hour 75-g oral glucose tolerance test [OGTT] with no prior screening with a nonfasting 1-hour 50-g glucose challenge test [GCT] or following a normal 50-g GCT finding), delivery <28 weeks of gestation (based on the assumption that any adverse effects of GDM will be minimal <28 weeks of gestation), complicated monochorionic-diamniotic twins (ie, twin-to-twin transfusion syndrome, selective intrauterine growth restriction, or twin-anemia-polycythemia sequence), monochorionic-monoamniotic twins, stillbirth or reduction of ≥1 fetuses, or genetic or structural fetal anomalies. The study was approved by the institutional research ethics board.
Definitions
The 2008 CDA guidelines for the screening and diagnosis of GDM were followed throughout the study period ( Table 1 ). These guidelines are based on a 2-step screening strategy. The first step involves universal screening at 24-28 weeks of gestation using the 1-hour 50-g GCT. The results of the 50-g GCT are interpreted as a negative screen, positive screen, or diagnostic for GDM ( Table 1 ). Patients with positive screen go on to have the diagnostic fasting 2-hour 75-g OGTT. The test is interpreted as positive for GDM if at least 2 of the values exceed their corresponding thresholds ( Table 1 ).
The IADPSG recommendations for the screening and diagnosis of GDM are based on a 1-step approach using a fasting state 2-hour 75-g OGTT ( Table 1 ), and the test is interpreted as positive for GDM if ≥1 of the values surpasses the corresponding threshold ( Table 1 ). All 3 thresholds of the 75-g OGTT recommended by the IADPSG are lower than those in the 2008 CDA recommendations ( Table 1 ).
All women diagnosed with GDM at our center are followed by a team that consists of a specialist in maternal fetal medicine, endocrinologist, diabetes nurse, and nutritional consultant experienced in the management of diabetes in pregnancy. Patients with GDM are seen in the diabetes clinic every 1-2 weeks from the time of diagnosis up to delivery. All patients monitor blood glucose levels 4 times a day using a glucometer. When patients do not meet the target glucose levels (fasting <95 mg/dL [<5.3 mmol/L] and 2-hour postprandial <121 mg/dL [<6.7 mmol/L]) after 1-2 weeks of diet management, treatment with insulin is started and the dose is titrated until adequate glycemic control is achieved. All women with twin pregnancies followed up in our medical center undergo ultrasound examination every 2 weeks during the third trimester until delivery. The routine sonographic evaluation on each of these examinations include presentation, amniotic fluid level (expressed as maximal vertical pocket), biophysical profile, and fetal biometry including biparietal diameter, head circumference (HC), abdominal circumference (AC), and femur length. The presence of asymmetric growth pattern, as reflected by high AC or AC/HC ratio, is used as an additional measure of the degree of glycemic control, and interpreted according to Campbell and Thoms.
Birthweight percentiles were calculated according to national growth curves based on sex-specific birthweights of singleton pregnancies. We chose to use singleton-based growth curves given the conflicting results regarding the benefit of twin-specific charts and the fact that the validation of currently available twin-specific charts is still limited.
Data collection
Data regarding all twin pregnancies were extracted from the institutional comprehensive perinatal database. This database contains information on demographic, medical, and obstetrical history, and information on the current pregnancy including initial and final number of fetuses, pregnancy dating, complications during pregnancy, mode of delivery, and neonatal outcome. The information on the results of the 50-g GCT and 75-g OGTT was extracted from our institutional laboratory database, which was cross-linked with the perinatal database. Finally, the results of the last third-trimester ultrasound were extracted from the database of our obstetric-ultrasound unit and were merged with the perinatal database.
Data analysis
Pregnancy outcome, fetal biometry, and neonatal birthweight were compared among 4 groups of women with increasing degree of glucose intolerance. (1) GCT-NEGATIVE group included women who had a negative screening with 50-g GCT. This group represented women with normal glucose tolerance and was used as the control/reference group. (2) OGTT-NEGATIVE group included women who had a positive 50-g GCT screening but the subsequent 75-g OGTT was negative according to both the IADPSG and the 2008 CDA criteria. These women were previously shown to be at increased risk for pregnancy complications attributed to abnormal glucose tolerance. (3) GDM-IADPSG group included women with a positive 50-g GCT screening that was followed by a 75-g OGTT that was positive according to the IADPSG criteria but not the 2008 CDA criteria. These women were previously shown to have an abnormal glucose tolerance and an increased risk of adverse pregnancy outcome and considered to have GDM according to the IADPSG criteria. Because these women were not considered to have GDM during the study period they did not monitor their glucose level and were not subjected to any form of dietary or medical interventions to control blood glucose levels. Thus, the outcome of pregnancies in this group can be considered to reflect the true impact of untreated mild GDM in twin gestations. (4) GDM-CDA group included women with a positive 50-g GCT that was followed by a 75-g OGTT that was positive according to the 2008 CDA criteria as well as women with a diagnostic 50-g GCT. Women in this group were monitored and treated according to the GDM management protocol as described above, which may lead to masking or underestimation of the true impact of GDM on pregnancy outcome in twin gestations. In all the analyses, the GCT-NEGATIVE group, consisting of women with normal glucose tolerance, was used as the reference group.
The χ 2 and independent sample Student t test were used for comparison of categorical and continuous variables, respectively, among each of the 3 groups with abnormal glucose tolerance and the GCT-NEGATIVE reference group. Generalized logistic regression model was used to assess the association between the degree of glucose intolerance and accelerated fetal growth while adjusting for potential confounders and accounting for interclass correlation within sets of twins. We accounted for multiple comparisons between the groups using the Bonferroni correction. Differences were considered to be statistically significant for a P value < .002. The statistical analysis was performed using software (SPSS, Version 22; IBM Corp, Armonk, NY).
Results
Characteristics of the study population
A total of 2461 women with twin pregnancies were identified during the study period, of whom 1393 were eligible for the study ( Figure 1 ): 1021 (73.3%) in the GCT-NEGATIVE group, 184 (13.2%) in the OGTT-NEGATIVE group, 99 (7.1%) in the GDM-IADPSG group, and 89 (6.4%) in the GDM-CDA group ( Figure 1 ).
