By John C. Hobbins, MD

Professor, Department of Obstetrics and Gynecology, University of Colorado School of Medicine, Aurora

Dr. Hobbins reports no financial relationships relevant to this field of study.

SYNOPSIS: A recent 25-study meta-analysis has shown that patients having only one elevation in a three-hour glucose tolerance test have similar maternal and fetal outcomes as patients diagnosed to have bona fide gestational diabetes.

SOURCE: Roeckner JT, Sanchez-Ramos L, Jijon-Knapp R, et al. Single abnormal value on three-hour oral glucose tolerance test during pregnancy is associated with adverse maternal and neonatal outcomes: A systematic review and meta-analysis. Am J Obstet Gynecol 2016;215:287-297.

Screening for diabetes has become complicated because methods in the literature vary according to different glucose challenges, glucose thresholds, and the timing for maternal blood sampling. The most common screening method in the United States has been to use a one-hour 50 g oral glucose load, followed by a blood sample taken one hour later. Published cutoffs vary between 130 and 140 mg/dL. (We use 140 mg/dL.) If the patient’s blood sugar exceeds the selected value, it is recommended that she have a 100 g, three-hour glucose tolerance test (GTT). There are two sets of upper limits commonly used in the United States: the Coustan/Carpenter method and one fashioned by the National Diabetes Data Group (NDDG). Any patient having two or more levels above these preset thresholds would be labeled as having gestational diabetes (GD), while one abnormal or none would exclude the diagnosis.

What happens to the patient who flunks the one-hour screen but does not meet the “glucose tolerance cut” because she has only one value that exceeds the limit? Actually, in most cases she is considered to be a nondiabetic and less attention generally is paid to tracking her progress and outcome. Fortunately, an ambitious attempt has been made to do just that.

Roeckner et al scoured the literature from 1966 through 2015 to find studies involving patients who had abnormal one-hour glucose screens followed by a full 100 g GTT. They focused on studies that satisfied their maternal and neonatal outcome criteria. Then they compared outcomes from those with only one abnormal value with others who had two or more (GDs), as well as those “normals” who had no elevated values in their GTTs.

After applying seemingly rigorous exclusion criteria, the authors were left with 25 studies that had adequate outcome data on 4,466 women. When compared with those with no GTT elevations, the “one onlys” had higher rates of macrosomia (odds ratio [OR], 1.59; 95% confidence interval [CI], 1.16-2.19), heavier mean birth weights by 44.5 g (95% CI, 8.10-80.8 g), higher rates of neonatal hypoglycemia (OR, 1.88; 95% CI, 1.05-3.38), total cesarean sections (OR, 1.69; 95% CI, 1.40-2.05), pregnancy-induced hypertension (OR, 1.55; 95% CI, 1.31-1.83), and low five-minute Apgar scores (OR, 6.10; 95% CI, 2.65-14.02). Increased rates in the same order also were found when the “one onlys” were compared with those with normal one-hour glucose screens and, therefore, never had a GTT. The clincher came when very similar rates of adverse outcome were found when the “one onlys” were compared with true GDs, diagnosed by their having two or more abnormal values.


Anecdotally, I am sure the readers of OB/GYN Clinical Alert frequently have encountered patients who have flunked their one-hour screen and passed their three-hour GTTs, but still progress through pregnancy with all the fetal and maternal earmarks of a bona fide gestational diabetic. This study suggests that their outcomes are no different than “true” gestational diabetics.

It is granted that there are so many inconsistent diagnostic variables that confuse the issue, such as glucose levels used in the one-hour test, the size of the glucose load in the screening process, as well as the different diagnostic thresholds chosen in the three-hour diagnostic test. However, any patient who passes her GTT after an abnormal one-hour screen, but whose fetus has body-to-head disproportion, an estimated fetal weight of greater than 90th percentile, and a generous amount of amniotic fluid should raise our antennas to the point of suggesting:

  1. Pursue a low glycemic (diabetic) diet,
  2. Order another ultrasound scan after 34 weeks to assess fetal growth trajectory, and
  3. Repeat at least the fasting and postprandial blood sugars after a month.

This information could become useful when these patients have reached 38 to 39 weeks, when macrosomia with body to head disproportion could affect labor and delivery management decisions. Also, any form of diabetes would put into question the benefit of any further time in utero for these patients and their babies. Last, these infants should be watched in the nursery for hypoglycemia.

This study suggests that patients in the intermediate glycemic category seem to have the same predilection for adverse neonatal outcomes as true gestational diabetics and, therefore, should not be ignored.