Blatt, Amy J. PhD; Nakamoto, Jon M. MD, PhD; Kaufman, Harvey W. MD
Gestational diabetes mellitus (GDM), affecting between 4% and 12% of pregnancies, predicts the future development of type 2 diabetes mellitus.1,2 Approximately 10% of women with GDM will have postpartum diabetes, and another 3% to 35% will have impaired fasting glucose.1,3 The American College of Obstetrics and Gynecologists and the American Diabetes Association recommend that all pregnant women without diabetes diagnosed should be screened for GDM, by patient history, clinical risk factors, or a laboratory screening test, and that all women with GDM should be screened for diabetes 6 to 12 weeks postpartum.3–7
The Hyperglycemia and Adverse Pregnancy Outcomes Cooperative Research Group recently published a large, multicenter, blinded study that clarified the risks of adverse outcomes associated with GDM.8 The International Association of Diabetes and Pregnancy Study Groups reviewed available information and issued new diagnostic guidelines for hyperglycemia during pregnancy: one or more of the threshold values for the 75-g oral glucose tolerance test (OGTT) in Table 1 must be equaled or exceeded for GDM to be diagnosed.9
To date, published studies regarding GDM and postpartum diabetes screening are regional investigations confined to one hospital system or insurance network.10,11 In the current study, we assessed the screening rate and prevalence of GDM, as well as the postpartum diabetes screening and positivity rates, in a large, national population. We also assessed the effect of the new International Association of Diabetes and Pregnancy Study Groups recommendations on the current GDM positivity rate.
MATERIALS AND METHODS
Quest Diagnostics has more than 150 million patient encounters with individuals all across the United States, annually. We believe these patients to be representative of patients who seek medical care and testing in the United States. Test results are stored in the largest private clinical laboratory data warehouse in the United States, referred to as the Quest Diagnostics Informatics Data Warehouse. For the present study, we extracted testing data for pregnant and postpartum women as described; all data were de-identified before analysis. This study was found to be exempt by the Western Institutional Review Board.
To calculate the GDM screening rate, plasma glucose results from gestational diabetes screening and confirmatory tests for women aged 25 through 40 years, those not in a low-risk age group, were extracted from the Informatics Data Warehouse. Women aged 24 or younger were excluded from the screening rate calculations because women in this age range have been identified by the American College of Obstetrics and Gynecologists as having a low risk for GDM. Because of the lack of personal case histories for the patients in our database, we were not able to identify women who exhibited all of the factors indicating a low risk for GDM (such as body mass index of 25 or less, no history of abnormal glucose tolerance).4 The screening rate for GDM was calculated as the number of pregnant women who had a glucose tolerance test result (from the 50-g screen, 75-g or 100-g confirmatory test) divided by the total number of women identified as pregnant within the Informatics Data Warehouse population. A pregnant woman was identified as having had both a rubella test (either immunoglobulin G or immunoglobulin M; associated with the obstetric–gynecological panel of testing performed during the first prenatal visit) and any other laboratory test performed at Quest Diagnostics between weeks 30 and 45 of gestation; gestational week was based on an estimation of the timing of the first prenatal visit as week 8.
Two different age-inclusion criteria were used in this study. The first, used in the screening rate calculations described, includes women from ages 25 to 40 (those not in a low-risk age group). This allows us to estimate the percentage of women at potentially high risk who comply with the American College of Obstetrics and Gynecologists recommendations and undergo screening for GDM. The second, used in the positivity rate calculations described, includes women from ages 18 to 40 (adult women of child-bearing age). This allows us to estimate the number of women with positive test results for GDM from all of those who obtained an OGTT.
To calculate the GDM positivity rate, plasma glucose results from GDM screening and confirmatory tests for women aged 18 through 40 years were extracted from the Informatics Data Warehouse. Patients were identified as having GDM according to the American Diabetes Association/Carpenter-Coustan criteria developed in 1982 if two or more results from the 75-g or 100-g confirmatory test met or exceeded the threshold criteria (Table 1).12
Using results from the Hyperglycemia and Adverse Pregnancy Outcomes study, the International Association of Diabetes and Pregnancy Study Groups recommended thresholds for OGTT testing that correspond with adverse perinatal outcomes. These criteria, representing the first outcome-based thresholds published, were based on the average glucose levels at which the odds of the following complications were increased by a factor of 1.75 relative to the odds at mean glucose levels: birth weight above the 90th percentile for gestational age, cord-blood serum C-peptide level above 90th percentile, and percent body fat above the 90th percentile. Patients were identified as having GDM, according to the International Association of Diabetes and Pregnancy Study Groups recommendations, if one or more results from the 75-g confirmatory test met or exceeded the threshold criteria (Table 1).9
The positivity rate for GDM was calculated as the number of pregnant women with a positive GDM result divided by the total number of pregnant women who had a glucose tolerance test result (ie, from the 50-g screen, 75-g or 100-g confirmatory test).
To calculate the postpartum diabetes mellitus screening rate, we first limited the population to women who had GDM and continued to receive testing at Quest Diagnostics after pregnancy. These patients had additional laboratory testing at Quest Diagnostics during the period between the estimated delivery due date (based on an estimation of the timing of the first glucose tolerance test performed at week 24) and 6 months postpartum. This range is wider than the 6 to 12 weeks recommended by the American College of Obstetrics and Gynecologists and the American Diabetes Association to allow for any variations of this window and to capture more data of women who are offered postpartum testing beyond week 12 postpartum. In addition, this method of estimating the postpartum screening rate ensures that those who had postpartum screens are drawn from a population of women with GDM who had continued laboratory testing at Quest Diagnostics. Because a limitation of this study is that we are not able to follow-up women who did not return to our laboratory system, it does not seem reasonable to estimate the postpartum screening rate using all women with positive results for GDM testing. Therefore, in this study, we report the best-case scenario for postpartum screening.
Results indicating postpartum screening were extracted from the following four tests: 75-g glucose tolerance test, 100-g glucose tolerance test, serum glucose test (fasting and nonfasting), and the hemoglobin A1C test. A woman with GDM previously diagnosed was considered to have a postpartum screening result if we found records of at least one of the four screening tests being performed within 6 months after the woman's estimated delivery due date.
The postpartum screening rate was calculated as the number of women with a postpartum screening result divided by the number of women who had GDM and continued to receive testing at Quest Diagnostics 6 months after pregnancy. The postpartum diabetes mellitus positivity rate was calculated as the number of women with a positive diabetes result (based on the criteria in Table 2) divided by the number of women with a postpartum screening result.
Statistical comparisons were made using the Pearson χ2 test to assess the difference between proportions. SAS 9.2 was used for all data analyses.
There were 924,873 pregnant women aged 25 to 40 years in the study population (Table 3). Of these, 632,820 (68%) were screened for GDM (ie, had a 50-g glucose screen, 75-g or 100-g confirmatory test). Asians had the highest screening rate, more than 77%, whereas whites had the lowest screening rate, at 71% (Table 4). Multiple logistic regression analysis was performed to examine whether age, race group, or maternal weight could predict whether a woman is likely to be screened for GDM (Table 5). Women ages 35 through 40 are 21% more likely to be screened for GDM than are those ages 25 through 29. In addition, women weighing more than 275 pounds are 12% less likely to be screened for GDM than are those weighing between 100 and 124 pounds. Asians are 38% more likely to be screened for GDM than are whites.
Of the 842,993 pregnant women aged 18 to 40 years who were screened for glucose intolerance, 40,955 (4.9%) had positive test results for GDM under the Carpenter-Coustan criteria (Table 6). Multiple logistic regression analysis was performed to examine whether age, race group, or maternal weight could predict whether a woman had GDM (Table 7). Women ages 35 through 40 are 183% more likely to have GDM than those ages 18 through 24. In addition, women weighing more than 275 pounds are 348% more likely to have GDM than those weighing between 100 and 124 pounds. Asians are 177% more likely to have GDM than whites.
Using the new International Association of Diabetes and Pregnancy Study Groups recommendations, the GDM positivity rate increased to 5.3% (44,761), mainly because the number of patients with positive results after receiving the 75-g OGTT nearly doubled, from 3,842 (under the Carpenter-Coustan criteria, in which two or more results of 75-g OGTT met or exceeded the threshold criteria in Table 1) to 7,268 (under the International Association of Diabetes and Pregnancy Study Groups criteria, in which one or more results of the 75-g OGTT met or exceeded the threshold criteria in Table 1). The relative risk for a diagnosis of GDM, comparing the GDM positivity rates from the Carpenter-Coustan and International Association of Diabetes and Pregnancy Study Groups criteria, was 1.09 (95% confidence interval 1.078–1.107; P<.001).
The GDM positivity rates varied with age; rates were less than 3% for young adults (age 18–25 years) and approaching 9% for older adults (age 35–40 years). For all ages, rates were higher with the International Association of Diabetes and Pregnancy Study Groups criteria than with the Carpenter-Coustan criteria (Fig. 1). GDM positivity rates also varied with ethnicity under the Carpenter-Coustan criteria, being highest for Asians (8.7%) and lowest for African Americans (4.1%; Table 6). For all races, the GDM positivity rates were higher under the International Association of Diabetes and Pregnancy Study Groups recommendations than the Carpenter-Coustan criteria (Table 6). Maternal weight (as recorded on the maternal serum screen) was lowest among Asians, at 137 pounds, and highest among African Americans, at 175 pounds.
During the analysis period, 166,085 pregnant women had either the 75-g or the 100-g confirmatory test performed at Quest Diagnostics. Nearly 90% of the tests were the 100-g OGTT, and more than 10% were the 75-g OGTT. This trend was seen across all ages (data not shown). The GDM positivity rate for the 100-g test population was 24%. The GDM positivity rate for the 75-g test population was 22% (relative risk 1.08; 95% confidence interval 1.048–1.111; P<.001).
A distribution of the positive results from the 75-g and 100-g OGTT indicated that most occurred at either the 1-hour or 2-hour blood draws using the Carpenter-Coustan criteria (Table 8). The International Association of Diabetes and Pregnancy Study Groups criteria utilizes slightly lower threshold values for the fasting and 2-hour blood draws, and this was responsible for a 118% increase in the number of positive fasting results, relative to the Carpenter-Coustan criteria (Table 8).
Of the women who had test results indicative of GDM using the Carpenter-Coustan criteria, 23,299 women had continued laboratory care with Quest Diagnostics within 6 months after the estimated delivery due date. Of these, 4,486 (19.3%) returned for postpartum diabetes monitoring (Tables 4 and 9). More than half of the women who returned for postpartum diabetes monitoring were tested by postpartum week 6, and three-fourths were tested by postpartum week 13 (Fig. 2). This trend was seen for all of the race groups (data not shown). This indicates that only 20% of the women adhered to the American College of Obstetrics and Gynecologists and the American Diabetes Association guidelines to be screened between weeks 6 and 12 postpartum. Postpartum testing rates varied by ethnicity, being the highest for Asians and lowest for whites (Table 4), and also increased with age (Fig. 3).
Using the results from the 2-hour blood draw, we found that the population of pregnant women who did not undergo postpartum testing had a significantly lower glucose concentration than those who underwent postpartum testing (178.3 compared with 180.1; t=−3.39; P<.001). Although this difference is statistically significant based on the large population evaluated, the difference of 1.8 mg/dL is not of clinical significance on an individual patient basis.
Of the 4,486 women who returned for postpartum diabetes screening, 62 (1.4%) had postpartum diabetes mellitus diagnosed. Of these, 23 (37%) met diagnostic criteria with the serum glucose test (fasting or nonfasting), and 39 (63%) met diagnostic criteria with a hemoglobin A1C test.
In a large, national population of more than 800,000 pregnant women, screening rates for GDM (68%) and follow-up screening rates for postpartum diabetes (19%) were much less than the screening rates for those without preexisting diabetes, as suggested by the current American College of Obstetrics and Gynecologists and American Diabetes Association guidelines.4,5,6 Note that women with preexisting diabetes were not excluded from this study. The effect of their inclusion is that the screening rate for gestational diabetes is slightly understated. The effect is quite modest because the incidence of preexisting diabetes is so low and does not affect our observations or conclusions.
Screening for GDM is important because studies have shown that treating even mild GDM reduces morbidity for both the mother and newborn.2 Asians had the highest GDM rate and the lowest mean maternal weight, whereas African Americans had the lowest GDM rate and the highest mean maternal weight. This difference in GDM rates may be partially explained by factors other than weight or obesity, such as genetic predisposition to GDM or ability to metabolize lipids.13–16 The screening rates for both GDM and postpartum diabetes were higher for Asian women than for other racial groups, and this also has been previously reported in smaller, regional studies.10,16 However, screening rates were low in all race groups, a finding that could be explained by the lack of familiarity with recommendations among medical care providers, a lack of general understanding about the risk of diabetes mellitus after a pregnancy complicated by GDM, or perceived confusion among the recommendations for screening.3
According to a recent position statement on the standards of medical care for diabetes, the American Diabetes Association is planning to work with U.S. obstetrical organizations to consider adoption of the International Association of Diabetes and Pregnancy Study Groups recommendations.6 We quantified the effects that the new International Association of Diabetes and Pregnancy Study Groups recommendations would have on current testing practices and GDM positivity rates. Nearly 90% of the 166,085 pregnant women undergoing a confirmatory test were tested using the 100-g test. This means that a significant adjustment from medical providers is needed for the International Association of Diabetes and Pregnancy Study Groups recommendations to be fully adopted, because the recommendations indicate the use of the 75-g OGTT to diagnose GDM.8 Some regional clinics currently do not offer a 75-g OGTT to their pregnant patients.17
According to a recent editorial, the new International Association of Diabetes and Pregnancy Study Groups recommended criteria would result in 18% of all pregnant women having GDM diagnosed, which is approximately double the proportion of women with GDM diagnosed.18 We found that use of the International Association of Diabetes and Pregnancy Study Groups recommended criteria almost doubled the number of positive GDM results in the 75-g test.
Limitations of this study include an estimation of the screening rate that may be biased, because our test population consisted of patients who utilize the services of Quest Diagnostics and, hence, are seeking medical care. However, previous analysis (unpublished) compared the distribution of patients in the Informatics Data Warehouse with that of the national census and demonstrated that the population in the Informatics Data Warehouse is representative of the overall national population in terms of age, gender, and race (as indicated for pregnant women undergoing the maternal serum screen). Another limitation of this study is the estimation of the postpartum diabetes positivity rate. The low prevalence of postpartum diabetes in this study should not imply that postpartum monitoring is not important. Given that 19% of women with GDM were screened postpartum and 1% of these women had positive results for diabetes mellitus, 1% of the remaining 81% of those women who did not undergo screening postpartum, approximately additional 188 women, could also have postpartum diabetes develop. Another reason that the estimated postpartum diabetes positivity rate may not reflect the actual number of women who have postpartum diabetes is that we were able to follow-up a pregnant woman for only 6 months after the estimated delivery due date. Future extensions of this study investigating the prevalence of postpartum diabetes will include 3- to 5-year follow-ups of the women who have had GDM.
We examined the laboratory test results of nearly one million pregnant women. This was not a population of pregnant women restricted to a certain geography or managed health care organization, as is the case with several other studies.3,9,10 Our nationally based study demonstrates that the pregnancy and postpartum screening rates of pregnant women are much less than the recommended guidelines from the American College of Obstetrics and Gynecologists and the American Diabetes Association. Because the major national diabetic and obstetrical organizations are considering the implications of the new International Association of Diabetes and Pregnancy Study Groups recommended criteria, it is also important to note that the proportion of women with GDM diagnosed returning for postpartum testing is extremely low.
1. Kim C, Berger DK, Chamany S. Recurrence of gestational diabetes mellitus: a systematic review. Diabetes Care 2007;30:1314–9.
2. Landon MB, Spong CY, Thom E, Carpenter MW, Ramin SM, Casey B, et al. A multicenter, randomized trial of treatment for mild gestational diabetes. N Engl J Med 2009;361:1339–48.
3. Dietz PM, Vesco KK, Callaghan WM, Bachman DJ, Bruce FC, Berg CJ, et al. Postpartum screening for diabetes after a gestational diabetes mellitus-affected pregnancy. Obstet Gynecol 2008;112:868–74.
4. American College of Obstetricians and Gynecologists. Clinical management guidelines for obstetrician-gynecologists. ACOG Practice Bulletin 30. Washington, DC: ACOG; 2001.
5. Postpartum screening for abnormal glucose tolerance in women who had gestational diabetes mellitus. ACOG Committee Opinion No. 435. American College of Obstetricians and Gynecologists. Obstet Gynecol 2009;113:1419–21.
6. American Diabetes Association. Standards of medical care in diabetes–2010. Diabetes Care 2010;33:S11–61.
7. Pridjian G, Benjamin TD. Update on gestational diabetes. Obstet Gynecol Clin North Am 2010;37:255–67.
8. HAPO Study Cooperative Research Group, Metzger BE, Lowe LP, Dyer AR, Trimble R, Chovarindr U, et al. Hyperglycemia and adverse pregnancy outcomes. N Engl J Med 2008;358:1991–2002.
9. International Association of Diabetes and Pregnancy Study Group Consensus Panel, Metzger BE, Gabbe SG, Persson B, Buchanan TA, Catalano PA, et al. International association of diabetes and pregnancy study groups recommendations on the diagnosis and classification of hyperglycemia in pregnancy. Diabetes Care 2010;33:676–82.
10. Ferrara A, Peng T, Kim C. Trends in postpartum diabetes screening and subsequent diabetes and impaired fasting glucose among women with histories of gestational diabetes mellitus: A report from the Translating Research Into Action for Diabetes (TRIAD) Study. Diabetes Care 2009;32:269–74.
11. Lawrence JM, Black MH, Hsu JW, Chen W, Sacks DA. Prevalence and timing of postpartum glucose testing and sustained glucose dysregulation after gestational diabetes mellitus. Diabetes Care 2010;33:569–76.
12. Carpenter MW, Coustan DR. Criteria for screening tests for gestational diabetes. Am J Obstet Gynecol 1982;144:768–73.
13. Chu SY, Abe K, Hall LR, Kim SY, Njoroge T, Cheng Q. Gestational diabetes mellitus: All Asians are not alike. Preventative Med 2009;29:265–8.
14. Savitz DA, Janevic TM, Engel SM, Kaufman JS, Herring AH. Ethnicity and gestational diabetes in New York City, 1995–2003. BJOG 2008;115:969–78.
15. Shelley-Jones DC, Wein P, Nolan C, Beischer NA. Why do Asian-born women have a higher incidence of gestational diabetes? An analysis of racial differences in body habitus, lipid metabolism and the serum insulin response to an oral glucose load. Aust N Z J Obstet Gynaecol 1993;33:114–8.
16. Lawrence JM, Contreras R, Chen W, Sacks DA. Trends in the prevalence of preexisting diabetes and gestational diabetes mellitus among a racially/ethnically diverse population of pregnant women, 1999–2005. Diabetes Care 2008;31:899–904.
17. Rollins G. A new definition of gestational diabetes: will revised diagnostic criteria help or hurt? Clin Lab News 2010;36:1–7.
18. Moses RG. New consensus criteria for GDM: problem solved or a Pandora's box? Diabetes Care 2010;33:690–1.