Original Article
771
Interaction between Maternal Obesity and 1-Hour Glucose Challenge Test Results on Maternal and Perinatal Outcomes Akila Subramaniam, MD, MPH1 Lorie M. Harper, MD, MSCI1
Victoria C. Jauk, MPH, MSN1
1 Division of Maternal-Fetal Medicine, Department of Obstetrics and
Gynecology, University of Alabama at Birmingham, Birmingham, Alabama
Alan Tita, MD, PhD1
Address for correspondence Akila Subramaniam, MD, MPH, Women and Infants Center Room 10270, 619 19th Street South, Birmingham, AL 35249-7333 (e-mail: [email protected]).
Abstract
Keywords
► maternal obesity ► gestational diabetes screening ► pregnancy outcomes
Objective The objective of this study was to examine the relationship between positive glucose challenge test (GCT) values and perinatal outcomes stratified by maternal body mass index (BMI). Study Design Retrospective cohort of singleton gestations with a GCT performed at >20 weeks and documented BMI at entry to care. Subjects were classified by GCT level and BMI. Primary outcomes included large for gestational age (LGA), macrosomia, shoulder dystocia, and pregnancy-induced hypertension. Cochran–Armitage tests for trend and logistic regression were used to compare the GCT categories. Results A total of 14,525 women met enrollment criteria—8,521 with a GCT < 120 mg/dL and 6,004 with a GCT 120 mg/dL. When BMI < 25 kg/m2 was considered, the risks were not increased at any level of GCT for any outcome. However, for subjects with BMI 25 kg/m2, the risk of LGA for a GCT 130 to 134 mg/dL was increased, but not at GCT of 135 to 139 mg/dL (p < 0.001). Similar, but nonsignificant, trends were observed for macrosomia and shoulder dystocia. Conclusion Increasing GCT is associated with adverse outcomes primarily in women with a BMI 25 kg/m2. Women with a BMI 25 kg/m2 and a GCT 135 to 140 mg/dL appear to have less risk of LGA than women with GCT 130 to 134 mg/dL, suggesting a possible effect of diagnosing and treating gestational diabetes mellitus in this group.
Gestational diabetes mellitus (GDM), or carbohydrate intolerance in pregnancy, complicates 2 to 14% of pregnancies in the United States.1,2 GDM increases the risk of fetal macrosomia, neonatal hypoglycemia, shoulder dystocia, birth injury, and perinatal death as well as the maternal risks of preeclampsia, operative delivery, and future diabetes mellitus.1–6 Given these adverse maternal and perinatal outcomes, screening, diagnosing, and treating gravidas with GDM is the standard of care in the United States. However, as studies suggest that the increased risks of GDM directly correlate
with glucose control and glycemic levels in a continuous fashion, optimal screening, and diagnostic criteria for GDM remain controversial.3,7–9 The most widely accepted strategy for GDM screening and diagnosis is the 1-hour 50-g glucose challenge test (GCT) at 24 to 28 weeks’ gestation, followed by a 3-hour 100 g oral glucose tolerance test (GTT) in GCT positive patients. Screening cutoff values for the GCT vary from 130 to 140 mg/dL; higher cutoff values decrease the number of GTT performed, but fail to detect some cases of GDM (i.e., increased specificity with decreased sensitivity).10
received June 21, 2014 accepted after revision October 2, 2014 published online December 29, 2014
Copyright © 2015 by Thieme Medical Publishers, Inc., 333 Seventh Avenue, New York, NY 10001, USA. Tel: +1(212) 584-4662.
DOI http://dx.doi.org/ 10.1055/s-0034-1396695. ISSN 0735-1631.
Downloaded by: University of Pittsburgh. Copyrighted material.
Am J Perinatol 2015;32:771–778.
Obesity, Gestational Diabetes Screening, and Pregnancy Outcomes Moreover, studies show that higher values of the GCT (with or without positive GTT) are associated with increasing adverse perinatal morbidity.5,11,12 Although the American Congress of Obstetricians and Gynecologists suggests GCT cutoff values between 130 and 140 mg/dL, it is unclear if the actual cutoff value may be refined by patient characteristics. Although most widely used, this GDM screening strategy was initially developed decades ago for utilization in a largely Caucasian population with a normal body mass index (BMI). In a modern population with significantly higher rates of obesity, these screening and diagnostic criteria may be suboptimal. Maternal obesity—an independent risk factor associated with adverse maternal and perinatal morbidity—is itself a risk factor for developing GDM.13,14 As obese women with GDM have worse pregnancy outcomes than normal weight women with GDM, it is clear that the combination of both obesity and GDM have an even greater impact on maternal and fetal outcomes.15,16 Given the potential interaction between carbohydrate intolerance, maternal obesity, and adverse perinatal outcomes, it is possible that obese women may benefit from a lower threshold for GDM screening (e.g., 135 vs. 140 mg/dL). As such, our objective in this study was to analyze the relationship between positive GCT values and perinatal outcomes stratified by maternal weight to explore if obese women require a lower GCT threshold for GDM screening.
Methods We performed a retrospective cohort study of all singleton pregnancies delivered at the University of Alabama at Birmingham from 2000 to 2010. Institutional review board approval was obtained from the University of Alabama at Birmingham. Women were included in this analysis if they delivered a singleton gestation, had a GCT performed after 20 weeks’ gestation, and had a documented BMI at entry to care. The first GCT performed after 20 weeks was used in the analysis; no early GCT results were included. Women were excluded for multifetal gestations, delivery before 20 weeks, fetal anomalies, and maternal medical comorbidities (including but not limited to pregestational diabetes, cardiac disease, HIV, hypertension, autoimmune disease, and placenta previa). In addition, women with a GCT < 40 mg/dL were excluded as this was likely due to laboratory error; women with a GCT > 200 mg/dL were also excluded as these women did not undergo confirmatory GTT. Only the first pregnancy during the ascertainment period was considered. At our institution, women with a GCT 135 mg/dL undergo a GTT. Subjects were classified by GCT level (GCT < 120, 120–129, 130–134, 135–139, 140–144, and 145–199 mg/dL) and stratified by BMI (BMI < 25, 25–29.9, 30–39.9, and 40 kg/m2). Increments of 5 mg/dL were used to define GCT categories between 130 and 144 mg/dL because of the variety of GCT cutoff values used in this range in clinical practice. Women with GCT < 120 mg/dL were used as the reference group. Primary outcomes considered were large for gestational age (LGA, > 90th percentile birth weight adjusted for gestational age by Brenner nomogram), macrosomia (4,000 g), American Journal of Perinatology
Vol. 32
No. 8/2015
Subramaniam et al.
shoulder dystocia as documented by the delivery physician, and pregnancy-induced hypertension (PIH), which includes both gestational hypertension and preeclampsia.17 Macrosomia and LGA were ascertained from the birth weight documented in the medical record. PIH and shoulder dystocia were based on documentation by the delivery physician in the medical record. Baseline maternal characteristics were compared across GCT categories using the chi-squared test for categorical variables and Student t-test or Wilcoxon rank sum test for continuous variables. The incidence of the study outcomes (calculated in the total study population except for shoulder dystocia calculated in those with vaginal delivery) was compared for increasing categories of GCT and BMI using the Cochran–Armitage test for trend. The population was stratified by GCT and BMI category to examine for an interaction effect between GCT level and maternal BMI. Logistic regression was used to estimate the independent associations between these categories while adjusting for potential confounders, including gestational diabetes (diagnosed with a 3-hour GTT using either Carpenter–Coustan or National Diabetes Data Group criteria), gestational age at screening and delivery, parity, and self-reported race or ethnicity. Adjusted odds ratios and 95% confidence intervals were computed relative to the GCT < 120 mg/dL category. Where relevant, p < 0.05 was considered statistically significant. All analyses were performed with SAS 9.2 statistical software (SAS Institute Inc, Cary, NC).
Results A total of 37,130 women with singleton deliveries at our institution were identified from our obstetric database, of which 14,525 women met inclusion criteria (►Fig. 1)—8,521 patients with a GCT < 120 mg/dL and 6,004 with a GCT 120 mg/dL were included in this study. Baseline demographics between these GCT groups are presented in ►Table 1. Overall, patients with a higher GCT tended to be older, parous, and have greater BMI. Gestational age at the time of GCT and at BMI ascertainment was similar between groups. The primary outcomes by major GCT category are presented in ►Table 2. The incidence of LGA was 4.4% in women with a GCT < 120 mg/dL increasing to 7.3% for GCT 130 to 134 mg/dL, decreasing to 5.0% for GCT 135 to 139 mg/dL, and then again rising to 9.8% for GCT 145 to 199 mg/dL (p-value for trend < 0.005). For macrosomia and shoulder dystocia, there were similar patterns by GCT category with either a decrease or a plateau at the 135 to 139 mg/dL GCT range (p-value for trend < 0.005 and p ¼ 0.04 respectively). For PIH, the decreased incidence was noted at a GCT of 130 to 134 mg/dL (p < 0.005). The four primary outcomes by major BMI category are presented in ►Table 3. The incidence of all four primary outcomes increased significantly as BMI increased (p 0.01). For each primary outcome, the unadjusted and adjusted interactions between BMI and GCT category are presented in the stratified analysis in ►Table 4 and ►Fig. 2A–D. For LGA,
Downloaded by: University of Pittsburgh. Copyrighted material.
772
Subramaniam et al.
Fig. 1 Flow diagram of patients included in the study.
when only BMI < 25 kg/m2 was considered, the risk of LGA was not significantly increased at any GCT level (p-value for trend ¼ 0.16). In all BMI categories with BMI 25 kg/m2, the incidence of LGA increased with increasing GCT, except for a decreased incidence of LGA at 135 to 139 mg/dL (p-value for trend < 0.005). Similarly, for patients with a BMI < 25 kg/m2, the risk of macrosomia, shoulder dystocia, and PIH was not significantly increased at any GCT level (p-value for trend ¼ 0.39, p ¼ 0.63, and p ¼ 0.14, respectively). In all BMI categories with BMI 25 kg/m2, the incidence of macrosomia significantly increased with increasing GCT, except for a decreased incidence of LGA at 135 to 139 mg/dL. This pattern in incidence between GCT 130 to 134 and 135 to 139 mg/dL was not as clearly observed with the risk of PIH or shoulder dystocia.
Discussion Our study shows that both an increasing GCT and an increasing BMI are strongly associated with increasing risk of LGA, macrosomia, shoulder dystocia, and PIH. In addition, there is
an interaction between GCT and BMI—the relationship between GCT and outcomes appears to differ by whether or not BMI is less than 25. For women with a BMI < 25 kg/m2, the risk of the studied adverse outcomes did not significantly increase at any level of GCT < 200 mg/dL. For women with a BMI 25 kg/m2, increasing GCT resulted in an overall increasing incidence of all outcomes. Interestingly, a plateau or decrease in many of these adverse outcomes was seen at a GCT level of 135 to 139 mg/dL. Given that the GCT screenpositive cutoff used at our institution is 135 mg/dL, a portion of patients about this GCT cutoff would have been diagnosed with GDM and treated resulting in a downward adjustment of outcome risks. Our results suggest a possible effect of treating and diagnosing GDM with a subsequent adjustment of the increasing risks of adverse pregnancy outcomes. Studies have shown that increasing glycemia is associated with adverse perinatal outcomes.5,7,11,12,18 In the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study of 23,316 women with maternal hyperglycemia less severe than in diabetes mellitus, a strong continuous association was demonstrated between maternal glucose levels and American Journal of Perinatology
Vol. 32
No. 8/2015
773
Downloaded by: University of Pittsburgh. Copyrighted material.
Obesity, Gestational Diabetes Screening, and Pregnancy Outcomes
Obesity, Gestational Diabetes Screening, and Pregnancy Outcomes
Subramaniam et al.
Table 1 Baseline demographics of patients by GCT result category Characteristics
GCT result category (mg/dL) Less than 120 (n ¼ 8,521)
120–129 (n ¼ 1,896)
130–134 (n ¼ 835)
135–139 (n ¼ 538)
140–144 (n ¼ 534)
145–199 (n ¼ 2201)
p
Age (y)a
21 (19–24)
23 (20–27)
23 (20–27
23 (20–29)
24 (20–29)
25 (21–30)
771
Interaction between Maternal Obesity and 1-Hour Glucose Challenge Test Results on Maternal and Perinatal Outcomes Akila Subramaniam, MD, MPH1 Lorie M. Harper, MD, MSCI1
Victoria C. Jauk, MPH, MSN1
1 Division of Maternal-Fetal Medicine, Department of Obstetrics and
Gynecology, University of Alabama at Birmingham, Birmingham, Alabama
Alan Tita, MD, PhD1
Address for correspondence Akila Subramaniam, MD, MPH, Women and Infants Center Room 10270, 619 19th Street South, Birmingham, AL 35249-7333 (e-mail: [email protected]).
Abstract
Keywords
► maternal obesity ► gestational diabetes screening ► pregnancy outcomes
Objective The objective of this study was to examine the relationship between positive glucose challenge test (GCT) values and perinatal outcomes stratified by maternal body mass index (BMI). Study Design Retrospective cohort of singleton gestations with a GCT performed at >20 weeks and documented BMI at entry to care. Subjects were classified by GCT level and BMI. Primary outcomes included large for gestational age (LGA), macrosomia, shoulder dystocia, and pregnancy-induced hypertension. Cochran–Armitage tests for trend and logistic regression were used to compare the GCT categories. Results A total of 14,525 women met enrollment criteria—8,521 with a GCT < 120 mg/dL and 6,004 with a GCT 120 mg/dL. When BMI < 25 kg/m2 was considered, the risks were not increased at any level of GCT for any outcome. However, for subjects with BMI 25 kg/m2, the risk of LGA for a GCT 130 to 134 mg/dL was increased, but not at GCT of 135 to 139 mg/dL (p < 0.001). Similar, but nonsignificant, trends were observed for macrosomia and shoulder dystocia. Conclusion Increasing GCT is associated with adverse outcomes primarily in women with a BMI 25 kg/m2. Women with a BMI 25 kg/m2 and a GCT 135 to 140 mg/dL appear to have less risk of LGA than women with GCT 130 to 134 mg/dL, suggesting a possible effect of diagnosing and treating gestational diabetes mellitus in this group.
Gestational diabetes mellitus (GDM), or carbohydrate intolerance in pregnancy, complicates 2 to 14% of pregnancies in the United States.1,2 GDM increases the risk of fetal macrosomia, neonatal hypoglycemia, shoulder dystocia, birth injury, and perinatal death as well as the maternal risks of preeclampsia, operative delivery, and future diabetes mellitus.1–6 Given these adverse maternal and perinatal outcomes, screening, diagnosing, and treating gravidas with GDM is the standard of care in the United States. However, as studies suggest that the increased risks of GDM directly correlate
with glucose control and glycemic levels in a continuous fashion, optimal screening, and diagnostic criteria for GDM remain controversial.3,7–9 The most widely accepted strategy for GDM screening and diagnosis is the 1-hour 50-g glucose challenge test (GCT) at 24 to 28 weeks’ gestation, followed by a 3-hour 100 g oral glucose tolerance test (GTT) in GCT positive patients. Screening cutoff values for the GCT vary from 130 to 140 mg/dL; higher cutoff values decrease the number of GTT performed, but fail to detect some cases of GDM (i.e., increased specificity with decreased sensitivity).10
received June 21, 2014 accepted after revision October 2, 2014 published online December 29, 2014
Copyright © 2015 by Thieme Medical Publishers, Inc., 333 Seventh Avenue, New York, NY 10001, USA. Tel: +1(212) 584-4662.
DOI http://dx.doi.org/ 10.1055/s-0034-1396695. ISSN 0735-1631.
Downloaded by: University of Pittsburgh. Copyrighted material.
Am J Perinatol 2015;32:771–778.
Obesity, Gestational Diabetes Screening, and Pregnancy Outcomes Moreover, studies show that higher values of the GCT (with or without positive GTT) are associated with increasing adverse perinatal morbidity.5,11,12 Although the American Congress of Obstetricians and Gynecologists suggests GCT cutoff values between 130 and 140 mg/dL, it is unclear if the actual cutoff value may be refined by patient characteristics. Although most widely used, this GDM screening strategy was initially developed decades ago for utilization in a largely Caucasian population with a normal body mass index (BMI). In a modern population with significantly higher rates of obesity, these screening and diagnostic criteria may be suboptimal. Maternal obesity—an independent risk factor associated with adverse maternal and perinatal morbidity—is itself a risk factor for developing GDM.13,14 As obese women with GDM have worse pregnancy outcomes than normal weight women with GDM, it is clear that the combination of both obesity and GDM have an even greater impact on maternal and fetal outcomes.15,16 Given the potential interaction between carbohydrate intolerance, maternal obesity, and adverse perinatal outcomes, it is possible that obese women may benefit from a lower threshold for GDM screening (e.g., 135 vs. 140 mg/dL). As such, our objective in this study was to analyze the relationship between positive GCT values and perinatal outcomes stratified by maternal weight to explore if obese women require a lower GCT threshold for GDM screening.
Methods We performed a retrospective cohort study of all singleton pregnancies delivered at the University of Alabama at Birmingham from 2000 to 2010. Institutional review board approval was obtained from the University of Alabama at Birmingham. Women were included in this analysis if they delivered a singleton gestation, had a GCT performed after 20 weeks’ gestation, and had a documented BMI at entry to care. The first GCT performed after 20 weeks was used in the analysis; no early GCT results were included. Women were excluded for multifetal gestations, delivery before 20 weeks, fetal anomalies, and maternal medical comorbidities (including but not limited to pregestational diabetes, cardiac disease, HIV, hypertension, autoimmune disease, and placenta previa). In addition, women with a GCT < 40 mg/dL were excluded as this was likely due to laboratory error; women with a GCT > 200 mg/dL were also excluded as these women did not undergo confirmatory GTT. Only the first pregnancy during the ascertainment period was considered. At our institution, women with a GCT 135 mg/dL undergo a GTT. Subjects were classified by GCT level (GCT < 120, 120–129, 130–134, 135–139, 140–144, and 145–199 mg/dL) and stratified by BMI (BMI < 25, 25–29.9, 30–39.9, and 40 kg/m2). Increments of 5 mg/dL were used to define GCT categories between 130 and 144 mg/dL because of the variety of GCT cutoff values used in this range in clinical practice. Women with GCT < 120 mg/dL were used as the reference group. Primary outcomes considered were large for gestational age (LGA, > 90th percentile birth weight adjusted for gestational age by Brenner nomogram), macrosomia (4,000 g), American Journal of Perinatology
Vol. 32
No. 8/2015
Subramaniam et al.
shoulder dystocia as documented by the delivery physician, and pregnancy-induced hypertension (PIH), which includes both gestational hypertension and preeclampsia.17 Macrosomia and LGA were ascertained from the birth weight documented in the medical record. PIH and shoulder dystocia were based on documentation by the delivery physician in the medical record. Baseline maternal characteristics were compared across GCT categories using the chi-squared test for categorical variables and Student t-test or Wilcoxon rank sum test for continuous variables. The incidence of the study outcomes (calculated in the total study population except for shoulder dystocia calculated in those with vaginal delivery) was compared for increasing categories of GCT and BMI using the Cochran–Armitage test for trend. The population was stratified by GCT and BMI category to examine for an interaction effect between GCT level and maternal BMI. Logistic regression was used to estimate the independent associations between these categories while adjusting for potential confounders, including gestational diabetes (diagnosed with a 3-hour GTT using either Carpenter–Coustan or National Diabetes Data Group criteria), gestational age at screening and delivery, parity, and self-reported race or ethnicity. Adjusted odds ratios and 95% confidence intervals were computed relative to the GCT < 120 mg/dL category. Where relevant, p < 0.05 was considered statistically significant. All analyses were performed with SAS 9.2 statistical software (SAS Institute Inc, Cary, NC).
Results A total of 37,130 women with singleton deliveries at our institution were identified from our obstetric database, of which 14,525 women met inclusion criteria (►Fig. 1)—8,521 patients with a GCT < 120 mg/dL and 6,004 with a GCT 120 mg/dL were included in this study. Baseline demographics between these GCT groups are presented in ►Table 1. Overall, patients with a higher GCT tended to be older, parous, and have greater BMI. Gestational age at the time of GCT and at BMI ascertainment was similar between groups. The primary outcomes by major GCT category are presented in ►Table 2. The incidence of LGA was 4.4% in women with a GCT < 120 mg/dL increasing to 7.3% for GCT 130 to 134 mg/dL, decreasing to 5.0% for GCT 135 to 139 mg/dL, and then again rising to 9.8% for GCT 145 to 199 mg/dL (p-value for trend < 0.005). For macrosomia and shoulder dystocia, there were similar patterns by GCT category with either a decrease or a plateau at the 135 to 139 mg/dL GCT range (p-value for trend < 0.005 and p ¼ 0.04 respectively). For PIH, the decreased incidence was noted at a GCT of 130 to 134 mg/dL (p < 0.005). The four primary outcomes by major BMI category are presented in ►Table 3. The incidence of all four primary outcomes increased significantly as BMI increased (p 0.01). For each primary outcome, the unadjusted and adjusted interactions between BMI and GCT category are presented in the stratified analysis in ►Table 4 and ►Fig. 2A–D. For LGA,
Downloaded by: University of Pittsburgh. Copyrighted material.
772
Subramaniam et al.
Fig. 1 Flow diagram of patients included in the study.
when only BMI < 25 kg/m2 was considered, the risk of LGA was not significantly increased at any GCT level (p-value for trend ¼ 0.16). In all BMI categories with BMI 25 kg/m2, the incidence of LGA increased with increasing GCT, except for a decreased incidence of LGA at 135 to 139 mg/dL (p-value for trend < 0.005). Similarly, for patients with a BMI < 25 kg/m2, the risk of macrosomia, shoulder dystocia, and PIH was not significantly increased at any GCT level (p-value for trend ¼ 0.39, p ¼ 0.63, and p ¼ 0.14, respectively). In all BMI categories with BMI 25 kg/m2, the incidence of macrosomia significantly increased with increasing GCT, except for a decreased incidence of LGA at 135 to 139 mg/dL. This pattern in incidence between GCT 130 to 134 and 135 to 139 mg/dL was not as clearly observed with the risk of PIH or shoulder dystocia.
Discussion Our study shows that both an increasing GCT and an increasing BMI are strongly associated with increasing risk of LGA, macrosomia, shoulder dystocia, and PIH. In addition, there is
an interaction between GCT and BMI—the relationship between GCT and outcomes appears to differ by whether or not BMI is less than 25. For women with a BMI < 25 kg/m2, the risk of the studied adverse outcomes did not significantly increase at any level of GCT < 200 mg/dL. For women with a BMI 25 kg/m2, increasing GCT resulted in an overall increasing incidence of all outcomes. Interestingly, a plateau or decrease in many of these adverse outcomes was seen at a GCT level of 135 to 139 mg/dL. Given that the GCT screenpositive cutoff used at our institution is 135 mg/dL, a portion of patients about this GCT cutoff would have been diagnosed with GDM and treated resulting in a downward adjustment of outcome risks. Our results suggest a possible effect of treating and diagnosing GDM with a subsequent adjustment of the increasing risks of adverse pregnancy outcomes. Studies have shown that increasing glycemia is associated with adverse perinatal outcomes.5,7,11,12,18 In the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study of 23,316 women with maternal hyperglycemia less severe than in diabetes mellitus, a strong continuous association was demonstrated between maternal glucose levels and American Journal of Perinatology
Vol. 32
No. 8/2015
773
Downloaded by: University of Pittsburgh. Copyrighted material.
Obesity, Gestational Diabetes Screening, and Pregnancy Outcomes
Obesity, Gestational Diabetes Screening, and Pregnancy Outcomes
Subramaniam et al.
Table 1 Baseline demographics of patients by GCT result category Characteristics
GCT result category (mg/dL) Less than 120 (n ¼ 8,521)
120–129 (n ¼ 1,896)
130–134 (n ¼ 835)
135–139 (n ¼ 538)
140–144 (n ¼ 534)
145–199 (n ¼ 2201)
p
Age (y)a
21 (19–24)
23 (20–27)
23 (20–27
23 (20–29)
24 (20–29)
25 (21–30)
