The criteria for prescribing insulin for women with gestational diabetes mellitus (GDM) remain controversial. An ACOG Technical Bulletin1 states that all women with diagnoses of gestational diabetes should be prescribed dietary therapy. Insulin should be prescribed when fasting plasma glucose exceeds 105 mg/dL, or 2-hour postprandial levels exceed 120 mg/dL.2 However, duration of dietary therapy is not specified.
Surveys of obstetricians and maternal-fetal subspecialists in practice3 and training4 show the need for clarification of treatment guidelines. These surveys found that practitioners had widely varied management practices for insulin prescription. For example, ACOG recommends1 that fasting and 2-hour postprandial blood glucose levels should be monitored “at least weekly.” However, ACOG's criteria were devised before the widespread use of more frequent blood glucose self-monitoring, which was used by 44% of practitioners4 for women with gestational diabetes treated with diet and by 76% of practitioners prescribing insulin.
Longitudinal ultrasonographic studies5,6 found that accelerated growth in fetuses of diabetic mothers might occur early in the third trimester, by 28–30 weeks' gestation. Most cases of GDM are diagnosed during this time, so it is important that quick, effective treatment of GDM be given to avoid fetal consequences. It is important to know how much time is needed to achieve good glycemic control with diet alone and to develop predictors for failure of dietary therapy alone at diagnosis. We sought to determine how much time it takes for dietary therapy alone to effect good glycemic control in GDM and whether the need for insulin could be predicted at diagnosis.
Materials and Methods
This prospective study included all women with GDM diagnosed at our institution from February 1994 to January 1995. All pregnant women in the University Health System and Metropolitan Health District maternal health clinics in the San Antonio, Texas, metropolitan area were screened for carbohydrate intolerance with a 50-g 1-hour glucose challenge, between 24 and 28 weeks' gestation. Those with risk factors for GDM, including obesity (prepregnancy body mass index [BMI] at least 27), histories of GDM, macrosomic infants (more than 4 kg), stillbirths, or fetuses with congenital anomalies were screened at their initial prenatal visits. If plasma glucose was at or above 130 mg/dL, a 3-hour, 100-g oral glucose tolerance test (GTT) was done. Gestational diabetes was diagnosed using the National Diabetes Data Group7 thresholds (equal to or exceeding 105, 190, 165, and 145 mg/dL), with one or more elevated values considered abnormal.8,9
Women with histories of diabetes outside pregnancy or those with fasting glucose levels on the 3-hour GTT above 140 mg/dL were excluded. Subjects were prescribed diets for 4 weeks, after which those in poor glycemic control, defined as mean blood glucose exceeding 105 mg/dL, were prescribed insulin therapy. According to our protocol, women with weekly mean glucose levels above 200 mg/dL at any time during the study were prescribed insulin immediately.
Caloric assignment was calculated based on actual weight at diagnosis. Obesity was defined by BMI, calculated by prepregnancy weight (kg) divided by height (m, squared). Maternal height was measured at first prenatal visit. Prepregnancy weight was taken from records or by subject reports if records were not available. Obese women (BMI at least 27) received 25 kcal/kg of actual body weight, those with BMI 20–26 received 30 kcal/kg, and those with BMI less than 20 received 35 kcal/kg.
Nutrients were distributed as follows: 50–55% carbohydrates, 20–25% protein (minimum 75 g per day), and 20–25% fat (less than 10% saturated fat). Energy was distributed as follows: breakfast 20%, lunch 25%, supper 25%, postmeal snacks 5% each, and bedtime snack 15%. All women received extensive dietary counseling by a registered dietician, with emphasis on developing diets consistent with their living situations and cultural norms. Subjects recorded all intake in diaries, which were reviewed on at least one other occasion by a dietician, who provided appropriate counseling. The importance of dietary compliance was reinforced by the obstetric care provider and a diabetes nurse-educator at each weekly clinic visit. All women checked for morning ketonuria daily, and caloric intake was adjusted as necessary during the clinic visits.
Subjects were instructed to monitor their blood glucose using memory reflectance meters seven times daily, including fasting, preprandial, 2-hour postprandial, and bedtime readings. The overall mean, fasting, preprandial, and postprandial averages were calculated using all values for the specified week of treatment.
Receiver operating characteristic (ROC) curves were generated to determine predictability of fasting plasma glucose of GTT for poor glycemic control by mean blood glucose exceeding 105 mg/dL or the ACOG criterion of fasting glucose exceeding 105 mg/dL or postprandial glucose exceeding 120 mg/dL. For analysis, women were stratified by fasting values of GTT (at most 95 mg/dL versus above 95 mg/dL). Categoric data were analyzed with Fisher exact and χ2 tests using Bonferroni inequality to adjust P values for multiple comparisons when the overall χ2 statistic was significant. Continuous data were analyzed by one-way analysis of variance for repeated measures using the Bonferroni t test ad hoc to determine multiple comparisons when the F statistic was significant. P < .05 was statistically significant. Logistic regression analysis was used to assess the effect various factors had on the need for insulin.
Two hundred sixty-nine women met the study criteria: 122 had fasting plasma glucose of GTT at or below 95 mg/dL, 59 had 96–105 mg/dL, 56 had 106–115 mg/dL, and 32 had 116–140 mg/dL. Maternal age, parity, and race did not differ between groups, even when subjects were stratified by fasting glucose less or greater than 95 mg/dL (Table 1). However, maternal weight and percentage of obesity increased significantly as fasting glucose values increased, and gestational age at diagnosis was significantly lower in women with fasting levels above 95 mg/dL.
Following appropriate counseling, 85% of women were found to be compliant with their dietary prescriptions, monitoring their glucose an average of 4.7 times per day. Overall, 43% of women needed insulin in addition to dietary therapy. No woman was prescribed insulin before the end of 4 weeks. Among GTT values, fasting glucose best predicted the need for insulin. Receiver operating characteristic curves were generated to determine cutoff values of GTT fasting glucose to predict the need for insulin. Poor glycemic control was defined as mean blood glucose above 105 mg/dL (Figure 1) and using the ACOG criteria of fasting glucose above 105 mg/dL or 2-hour postprandial glucose above 120 mg/dL (Figure 2). The optimal operating point for both curves was a GTT fasting value between 91 and 95 mg/dL. The positive predictive value of a fasting value above 95 mg/dL for mean glucose above 105 mg/dL was 54%, whereas for the ACOG criteria it was 43%.
Mean glucose results by week of treatment are summarized in Figure 3. Analysis of variance found fasting glucoses significantly lower after the 2nd week of dietary treatment, whereas postprandial glucose improved in the 4th week. When all study subjects were analyzed, there were no significant differences in the overall mean and preprandial glucoses throughout the study period.
When subjects were stratified by fasting glucose values significantly more women with fasting values of 95 mg/dL or less were able to achieve good glycemic control after 2 weeks of dietary therapy alone. Those with fasting levels above 95 mg/dL did not improve their glycemic control during the 4-week study period (Figure 4). Significantly more women with fasting glucose values of GTT above 95 mg/dL were determined to need insulin on the basis of elevated mean glucose exceeding 105 mg/dL (54% in the group with values above 95 mg/dL versus 30.2% in the group with values of 95 mg/dL or less, P < .001).
Women with one abnormal value of GTT were likely to need insulin as those with two or more abnormal values (34% versus 46%, P = not significant). Thirty percent of those with one abnormal value and fasting values not more than 95 mg/dL needed insulin, compared with 31% of those with two or more abnormal values (P = not significant). In the women with fasting values above 95 mg/dL, 41% of those with one abnormal value needed insulin, compared with 58% of those with two or more abnormal values (P = not significant).
Logistic regression analysis was done using need for insulin as the dependent variable. Independent variables used were maternal weight, BMI, weight gain during study, gestational age at diagnosis, all values of GTT, and the number of abnormal values of GTT. The fasting (P = .01), 1-hour (P = .04), and 3-hour (P = .001) values of the GTT and the number of abnormal values (P = .03) were associated with need for insulin. The other independent variables were not significant risk factors.
The key finding in our study was that most women with GDM had good glycemic control early in the course of dietary therapy, if such therapy alone was likely to be effective. Gravidas with fasting glucose levels of GTT exceeding 95 mg/dL did not significantly improve their glycemic control after the 1st week of therapy, whereas those with fasting levels at or below 95 mg/dL showed continued improvement after 2 weeks of dietary treatment.
Attaining good glycemic control is the cornerstone of managing women with gestational diabetes.10 Women with poor control have a higher incidence of fetal macrosomia,11 with resultant birth trauma, and other metabolic complications. Ultrasonographic studies5,6 found that such fetal overgrowth might have begun already before diagnosis; therefore, the time to prevent related morbidity is limited, and the decision to prescribe insulin should be quick.
It is still controversial which women with GDM will benefit from insulin. Some investigators12,13 found benefit in prophylactic insulin in all women with GDM. Although an ACOG Technical Bulletin1 recommends prescribing insulin if fasting glucose levels exceed 105 mg/dL or 2-hour postprandial levels exceed 120 mg/dL, two single readings (fasting or postprandial) in 1 week might not adequately show the level of glucose control in women receiving no additional monitoring. Conversely, prescribing insulin based on two elevated readings in women showing otherwise good control, by intensified monitoring, might result in overtreatment.
We found that a fasting glucose level between 91 and 95 mg/dL can predict the need for insulin for good glycemic control. We also identified a subset of women with fasting glucose not more than 95 mg/dL who will benefit from dietary therapy alone. Such therapy should be continued for at least 2 weeks because significant improvement in glycemic control can be achieved during that time. Almost 60% of women with fasting glucose above 95 mg/dL failed to achieve good glycemic control even after 4 weeks of dietary therapy. Subjects in this group weighed more and had GDM diagnosed at earlier gestational ages (ostensibly because of risk factors), possibly indicating more severe disease. Because these women did not improve their glycemic control significantly given a longer treatment period, a shorter (1-week) trial of dietary therapy appears to be adequate. Consideration might be given to immediate insulin prescription in this subset, particularly if GDM is diagnosed late in gestation, or to a longer trial of dietary therapy if women show nearoptimal glycemic control early in treatment with diet alone.
1. American College of Obstetricians and Gynecologists. Diabetes and pregnancy. ACOG technical bulletin no. 200. Washington, DC: American College of Obstetricians and Gynecologists, 1994.
2. Metzger BE. Summary and recommendations of the Third International Workshop-Conference on Gestational Diabetes Mellitus. Diabetes 1991;40(Suppl 2):197–201.
3. Landon MA, Gabbe SG, Sachs L. Management of diabetes mellitus and pregnancy: A survey of obstetricians and maternal-fetal specialists. Obstet Gynecol 1990;75:635–40.
4. Owen J, Phelan ST, Landon MB, Gabbe SG. Gestational diabetes survey. Am J Obstet Gynecol 1995;172:615–20.
5. Ogata ES, Sabbagha R, Metzger BE, Phelps RL, Depp R, Freinkel N. Serial ultrasonography to assess evolving fetal macrosomia. JAMA 1980;243:2405–8.
6. Langer O, Kozlowski S, Brustman L. Abnormal growth patterns in diabetes in pregnancy: A longitudinal study. Israel J Med Sci 1991;27:516–23.
7. National Diabetes Data Group. Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 1979;28:1039–57.
8. Langer O, Anyaegbunam A, Brustman L, Divon MY. A prospective randomized study: Management of women with one abnormal glucose tolerance test value reduces adverse outcome in pregnancy. Am J Obstet Gynecol 1989;161:593–9.
9. Langer O, Brustman L, Anyaegbunam A, Mazze R. The significance of one abnormal glucose tolerance test value on adverse outcome in pregnancy. Am J Obstet Gynecol 1987;157:758–63.
10. Langer O. Management of gestational diabetes. Clin Perinatol 1993;20:603–17.
11. Langer O, Rodriguez DA, Xenakis EM, McFarland MB, Berkus MD, Arredondo F. Intensified versus conventional management of gestational diabetes. Am J Obstet Gynecol 1994;170:1036–47.
12. Coustan DR, Imarah J. Prophylactic insulin treatment of gestational diabetes reduces the incidence of macrosomia, operative delivery, and birth trauma. Am J Obstet Gynecol 1984;150:836–42.
13. Thompson DJ, Porter KB, Gunnells DJ, Wagner PC, Spinnato JA. Prophylactic insulin in the management of gestational diabetes. Obstet Gynecol 1990;75:960–4.