Objective Data
The purpose of this study was to determine whether women with 1 abnormal value on 3-hour 100-g oral glucose tolerance test are at an increased risk for adverse pregnancy outcomes.
Study
Gestational diabetes mellitus is diagnosed by a 2-step method, with a 3-hour, 100-g oral glucose tolerance test that is reserved for women with an abnormal 1-hour, 50-g glucose challenge test. Although the increased maternal-fetal morbidity with gestational diabetes mellitus is well established, controversy remains about the risk that is associated with an isolated abnormal value during a 3-hour, 100-g oral glucose tolerance test.
Study Appraisal and Synthesis Methods
Prospective and retrospective studies that evaluated the maternal and perinatal impact of 1 abnormal glucose value during a 3-hour, 100-g oral glucose tolerance test were identified with the use of computerized databases. Data were extracted and quantitative analyses were performed.
Results
Twenty-five studies (7 prospective and 18 retrospective) that met criteria for metaanalysis included 4466 women with 1 abnormal glucose value on oral glucose tolerance test. Patients with 1 abnormal glucose value had significantly worse pregnancy outcomes compared with women with zero abnormal values with the following pooled odds ratios: macrosomia, 1.59 (95% confidence interval, 1.16–2.19); large for gestational age, 1.38 (95% confidence interval, 1.09–1.76); increased mean birthweight, 44.5 g (95% confidence interval, 8.10–80.80 g); neonatal hypoglycemia, 1.88 (95% confidence interval, 1.05–3.38); total cesarean delivery, 1.69 (95% confidence interval, 1.40–2.05); pregnancy-induced hypertension, 1.55 (95% confidence interval, 1.31–1.83), and Apgar score of <7 at 5 minutes, 6.10 (95% confidence interval, 2.65–14.02). There was also an increase in neonatal intensive care unit admission and respiratory distress syndrome. Similar results were seen that compared 1 abnormal glucose value to a population with a normal 1-hour 50-g glucose challenge test (normal glucose screen). With the exception of birthweight, outcomes of patients with 1 abnormal glucose value were similar to outcomes of patients with gestational diabetes mellitus.
Conclusion
Women with 1 abnormal value on 3-hour, 100-g oral glucose tolerance test have a significantly increased risk for poor outcomes comparable with women who have gestational diabetes mellitus.
Gestational diabetes mellitus (GDM), a carbohydrate metabolism disorder with onset or first recognition during pregnancy, is linked to maternal and perinatal morbidity. Although a 75-g oral glucose tolerance test (OGTT) is endorsed by the International Association of Diabetes in Pregnancy Study Group (IADPSG) and is used in many parts of the world, in the United States GDM usually is diagnosed with the use of a 2-step method with a 3-hour, 100-g OGTT reserved for women with an abnormal 1-hour, 50-g glucose challenge test (GCT). Abnormal values for the 1-hour, 50-g glucose screening test range from 130–140 mg/dL. Based on work done in the 1960s, GDM is diagnosed by 2 abnormal values on the OGTT with the use of either Carpenter and Coustan (CC) or National Diabetes Data Group (NDDG) cut-offs.
Although the increased maternal-fetal morbidity with GDM is well established, controversy remains regarding the risk of an isolated abnormal value on the OGTT. These patients fall into a category of glucose intolerance that has been labeled as borderline GDM, impaired glucose tolerance, or mild gestational hyperglycemia. Since the 1980s, the significance of 1 abnormal value on the 3-hour OGTT has been investigated with mixed results. Although some investigators have found that impaired glucose tolerance is associated with increased risk of large-for-gestational-age (LGA) infants, macrosomia, cesarean delivery, preeclampsia, and increased risk of diabetes mellitus later in life ; other investigators have not found increased risk, and trials of diet have found mixed results. Screening, early prediction, and treatment can result in unnecessary and costly intervention.
The specific goal of this systematic review and metaanalysis is to determine whether women with 1 abnormal value on the 3-hour, 100-g OGTT are at an increased risk for poor pregnancy outcomes that include LGA, macrosomia, increased birthweight, cesarean delivery, hypertensive disorders that are related to pregnancy, hypoglycemia, low Apgar scores, and respiratory distress when compared with (1) pregnant women with all normal values on OGTT, (2) women with a normal 1-hour 50-g screening test, and (3) women with GDM.
Materials and Methods
This systematic review and metaanalysis was preceded by a prospectively written protocol and was conducted according to the Meta-analysis of Observational Studies in Epidemiology guidelines. The protocol was registered in Prospero (CRD42015032462). Institutional Review Board approval was not needed. A thorough and extensive search of published literature from January 1966 to December 2015 was conducted with PubMed, Medline, Google Scholar, Cochrane, and clinicaltrials.gov . The key words “gestational diabetes” and “glucose tolerance test” were searched independently as well as in conjunction with the following keywords: “1 abnormal value,” “large for gestational age,” “cesarean delivery,” “NICU admission,” “hypoglycemia,” and “macrosomia.” Prospective and observational studies that evaluated the maternal and perinatal impact of 1 abnormal value during a 3-hour, 100-g GTT were identified and selected. Each study assessed ≥1 of the following maternal and perinatal outcomes: macrosomia (at least 4000 g at birth), LGA based on a birthweight at least in the 90th percentile for gestational age, neonatal hypoglycemia, cesarean delivery, neonatal intensive care unit (NICU) admission, abnormal Apgar scores, or preeclampsia or pregnancy-induced hypertension (PIH) based on American College of Obstetricians and Gynecologists’ criteria. Randomized clinical control trials were included if data could be extracted from the noninterventional arm. Non–English-language studies were not excluded. Article references and syllabi from scientific meetings were screened for additional articles.
Because each report did not assess all outcomes of interest, specific outcome metaanalyses were performed based on a variable number of studies that were related to that outcome. Studies that used a 2-hour, 75-g OGTT were excluded. Studies were also excluded if they lacked a suitable comparison group (ie, combined patients with zero abnormal value [GTT-0] or 1 abnormal value [GTT-1] on OGTT), if they lacked sufficient quantitative data for extraction, if they lacked a comparison group, if they lacked assessment of neonatal outcomes of interest, if the authors failed to provide a breakdown of composite outcomes, or if the treatment was administered based on 1 abnormal value.
Each study was scored for quality by 2 authors (J.T.R. and R.J.-K.) who used the Newcastle-Ottawa Quality Assessment scale. Data that were collected from each study included first author, study design, publication year, comparison group definition, maternal body mass index (BMI), gestational age at birth of participants (when reported), neonatal outcomes that were measured, number of participants in the study and comparison group, criteria for abnormal GCT, and criteria for diagnosis of GDM. Raw data were extracted by the primary author who used 2×2 tables for each outcome that was measured.
Metaanalyses were performed for each maternal and neonatal outcome with a Stata statistical software package (version 11.0; Stata Corporation, College Station, TX). Odds ratios (ORs) for each outcome and 95% confidence intervals (CIs) were calculated for women in the GTT-1 group compared with GTT-0, normal 1-hour screen (normal glucose screen), and GDM when applicable. Summary effects estimates were calculated with fixed-effects (Mantel-Haenszel) and random-effects (DerSimonian and Laird) models. Random effects models were used whenever there was evidence of heterogeneity ( P < .10). Heterogeneity was assessed by implementation of the I 2 statistic, Breslow-Day method, and L’Abbe plots that were inspected visually to assess homogeneity across studies. The null hypothesis underlying the overall test of association was that the overall OR was equal to 1. Asymmetric funnel plots are caused by small trials that report greater effects than larger trials, which suggests publication or other biases. Publication bias was evaluated with the Egger test and by inspection of funnel plots, which plotted relative risk against study sample size.
Sensitivity analyses were performed by omitting each study sequentially and analyzing the overall impact of that particular study on the pooled results. Meta-regression analysis was performed to identify potential causes of heterogeneity among any statistically significant neonatal outcomes by an examination of the important covariables of BMI, gestational age at delivery, GDM classification systems (CC or NDDG), and 1-hour, 50-g GCT screening cut-off.
Results
In total, 25 studies were selected for inclusion. The study selection process is detailed in Figure 1 ; the characteristics of the included studies are detailed in Table 1 . The studies varied with respect to size; 1-hour, 50-g GCT cut-off; and classification of diabetes mellitus. The 25 studies that were included assessed a total of 4466 women with GTT-1 on OGTT.
| Study | Year | Country | Comparison groups a | 1-Hr glucose challenge test cut off (mg/dL) | Gestational diabetes mellitus diagnostic criteria | Quality score b |
|---|---|---|---|---|---|---|
| Berkus | 1993 | United States | A | 140 | NDDG | 6 |
| Bhat | 2005 | India | A | 140 | NDDG | 7 |
| Biri | 2009 | Turkey | A,B,C | 140 | NDDG | 8 |
| Carr | 2011 | United States | A,C | 140 | CC | 7 |
| Chico | 2005 | Spain | B | 140 | NDDG | 6 |
| Corrado | 2009 | Italy | A | 135 | CC | 7 |
| DiCianni | 2005 | Italy | A,C | 140 | CC | 6 |
| Forest | 1994 | Canada | B,C | 140 | Other | 6 |
| Kaufmann | 1992 | United States | B | 135 | CC, NDDG | 6 |
| Kim | 2002 | South Korea | A | 130 | NDDG | 8 |
| Kokanali | 2014 | Turkey | B | 140 | CC | 6 |
| Landon | 2011 | United States | A,B,C | 135 | CC | 8 |
| Langer | 1987 | United States | B,C | 135 | NDDG | 8 |
| Langer | 1989 | United States | B | 130 | NDDG | 7 |
| Lindsay | 1989 | United States | B | 135 | CC | 9 |
| McLaughlin | 2006 | United States | B | 140 | CC | 8 |
| Park | 2015 | Korea | A,C | 140 | CC | 8 |
| Retnakaran | 2008 | Canada | A,B | 140 | NDDG | 8 |
| Tamez Perez | 1993 | Mexico | A,C | ? | NDDG | 7 |
| Suresh | 2013 | India | A | 140 | NDDG | 6 |
| Thanasuan | 2006 | Thailand | B | 140 | NDDG | 7 |
| Vambergue | 2002 | France | A,C | 130 | CC | 8 |
| Vambergue | 2000 | France | B | 130 | CC | 7 |
| Wang | 2013 | Taiwan | A,B,C | 140 | CC, NDDG | 7 |
| Wang | 2009 | China | A,B | 140 | NDDG | 6 |
a A , 1 abnormal value vs zero abnormal values; B , 1 abnormal value vs normal glucose screen; C , 1 abnormal value vs gestational diabetes mellitus
Data on maternal and neonatal outcomes that compared GTT-1 with GTT-0 were pooled ( Table 2 ; Figure 2 ). Significant heterogeneity among studies was not present, except where noted. The risk of macrosomia was significantly increased in mothers with GTT-1 (OR, 1.59; 95% CI, 1.16–2.19) compared with mothers with GTT-0 ( Figure 2 ). Significant increased OR for LGA, hypoglycemia, birthweight, cesarean delivery, and Apgar score <7 at 5 minutes were noted ( Table 2 ; Figure 3 ). Although no significant differences were found, there was a trend toward increased rates of NICU admission or respiratory distress syndrome, when GTT-1 was compared with GTT-0.
| Outcome | References | Abnormal value | Pooled odds ratio (95% confidence interval) | |
|---|---|---|---|---|
| 1 | 0 | |||
| Macrosomia, n/N | 82/1116 | 123/3101 | 1.59 (1.16–2.19) a | |
| Large for gestational age, n/N | 157/972 | 271/2616 | 1.38 (1.09–1.76) a | |
| Mean birthweight, g b | 3303 ± 509 (n = 1022) | 3234 ± 442 (n = 3027) | 44.5 (8.10–80.80) c | |
| Total cesarean delivery, n/N | 564/1276 | 1308/3992 | 1.69 (1.40–2.05) d | |
| Pregnancy-induced hypertension, n/N | 229/2661 | 438/8000 | 1.55 (1.31–1.83) d | |
| Hypoglycemia, n/N | 33/702 | 33/1713 | 1.88 (1.05–3.38) d | |
| Apgar <7 at 5 min, n/N | 18/257 | 9/853 | 6.10 (2.65–14.02) d | |
| Neonatal intensive care unit admission, n/N | 61/656 | 153/2146 | 1.30 (0.95–1.78) d | |
| Respiratory distress syndrome, n/N | 18/550 | 34/1089 | 1.81 (0.99–3.30) d | |
a DerSimonian and Laird odds ratio, random effects model
b Data are given as mean ± standard deviation
c Inverse variance pooled weighted mean difference
When GTT-1 was compared with normal glucose screen, women with GTT-1 had a significantly higher rate of macrosomia, LGA, increased birthweight, hypoglycemia, cesarean delivery, and NICU admission ( Table 3 ). The OR for respiratory distress syndrome and Apgar scores <7 at 5 minutes were increased but did not achieve statistical significance.
| Outcome | References | 1 abnormal value | Normal glucose screen | Pooled odds ratio (95% confidence interval) |
|---|---|---|---|---|
| Macrosomia, n/N | 261/2128 | 3017/43,269 | 1.85 (1.59–2.15) a | |
| Large for gestational age, n/N | 204/1353 | 897/13,507 | 1.85 (1.32–2.60) b | |
| Weight, g c | 3346 ± 522 (n = 1154) | 3252 ± 469 (n = 18,907) | 69 (37–100) d | |
| Total cesarean delivery, n/N | 652/1875 | 9141/38,076 | 1.61 (1.45–1.80) a | |
| Pregnancy-induced hypertension, n/N | 140/1891 | 1233/32,544 | 1.84 (1.50–2.26) a | |
| Hypoglycemia, n/N | 67/822 | 466/12,814 | 1.77 (1.26–2.48) a | |
| Apgar score <7 at 5 min, n/N | 8/434 | 19/981 | 1.02 (0.40–2.60) a | |
| Neonatal intensive care unit admission, n/N | 122/1432 | 1718/27,258 | 1.47 (1.20–1.80) a | |
| Respiratory distress syndrome, n/N | 9/670 | 78/6,899 | 1.68 (0.78–3.66) a |
a Mantel-Haenszel pooled odds ratio, fixed effects model
b DerSimonian and Laird odds ratio, random effects model
c Data are given as mean ± standard deviation
There was no significant difference between GTT-1 and GDM for the outcomes of macrosomia, LGA, hypoglycemia, cesarean delivery, PIH, NICU admission, and respiratory distress syndrome ( Table 4 ; Figure 4 ). No studies of GTT-1 vs GDM included outcomes for Apgar scores <7 at 5 minutes. Mean birthweight was 92.4 g (95% CI, 41.7–143.2) higher in neonates who were born to mothers with GTT-1, compared with neonates born to mothers with GDM.
| Outcome | References | 1 abnormal value | Gestational diabetes mellitus | Pooled odds ratio (95% confidence interval) |
|---|---|---|---|---|
| Macrosomia, n/N | 70/1073 | 131/1837 | 0.98 (0.55–1.76) a | |
| Large for gestational age, n/N | 141/816 | 354/1943 | 1.12 (0.66–1.90) a | |
| Weight, g b | 3298 ± 489 (n = 672) | 3205 ± 534 (n = 1109) | 92.4 (41.7–143.2) c | |
| Total cesarean delivery, n/N | 377/832 | 696/1689 | 0.99 (0.83–1.18) d | |
| Pregnancy-induced hypertension, n/N | 188/2350 | 288/3083 | 0.88 (0.73–1.08) d | |
| Hypoglycemia | 33/526 | 120/1027 | 0.50 (0.24–1.05) a | |
| Apgar score <7 at 5 min | None | N/A | N/A | N/A |
| Neonatal intensive care unit admission, n/N | 61/656 | 128/859 | 0.71 (0.42–1.20) a | |
| Respiratory distress syndrome, n/N | 7/425 | 21/563 | 0.55 (0.22–1.41) d |
a DerSimonian and Laird odds ratio, random effects model
b Data are given as mean ± standard deviation
c Inverse variance pooled weighted mean difference
The results of the meta-regression analysis that used the covariates of BMI, 1-hour screening test cuff-off, and GDM classification method did not alter the summary effect estimates of statistically significant results. Visual inspection of funnel plots did not identify significant evidence of publication bias.
Results
In total, 25 studies were selected for inclusion. The study selection process is detailed in Figure 1 ; the characteristics of the included studies are detailed in Table 1 . The studies varied with respect to size; 1-hour, 50-g GCT cut-off; and classification of diabetes mellitus. The 25 studies that were included assessed a total of 4466 women with GTT-1 on OGTT.
