Screening for gestational diabetes by measuring fasting plasma glucose levels SQU JOURNAL FOR SCIENTIFIC RESEARCH: MEDICAL SCIENCES 2003 VOL 5, NO. –2, 5–8 ©SULTAN QABOOS UNIVERSIT Y Screening for gestational diabetes by measuring fasting plasma glucose levels *Khalil E. Rajab1, Jonathan H. Skerman2, Abdulla A. Issa1 1Department of Obstetrics and Gynaecology, Department of Anaesthesia and Intensive Care, Arabian Gulf University, College of Medicine and Medical Sciences, Salmaniya Medical Complex, P.O.Box 26752, Manama, Kingdom of Bahrain. *To whom correspondence should be addressed. Email: yonrajab@batelco.com.bh ABSTRACT. Objectives: (a) To test the sensitivity and specificity of measuring fasting plasma glucose levels (FPG) as a screening test for gestational diabetes mellitus (GDM). (b) To compare predicting levels of FPG levels with the one-hour, oral 50g non-fasting glucose chal- lenge test (GCT) for predicting GDM. Methods: One thousand and six hundred pregnant women from the Health Centres, antenatal clinics and Salmaniya Medical Complex were screened by the GCT after 50g of oral glucose during 26–32 weeks gestation, giving a 13.5% incidence of GDM (using the Third International Workshop cutoff values of 7.8 mmol /l). All patients also had an FPG estimation followed by the three- hour oral glucose tolerance test (oGTT). Seventy eight percent of the patients were Bahraini, 19% Asian and 3% other nationalities. Their mean age was 27.2+0.2 years. Receiver-operating curves (ROC) were used to test the ability of the FPG and the oGTT to differentiate patients with GDM and identify the cut off values for predicting a diagnosis of GDM. Results: FPG levels of 5.6 mmol /l and 5.4 mmol /l yielded sensi- tivities and specificities of 94% and 93% respectively. Measuring FPG as a screening test required a diagnostic oGTT in 32% compared with 13% when the GCT was used. Conclusion: Using FPG levels at a cutoff value of ≥ 5.5 mmol /l is an easier, more acceptable test for patients compared to the GCT. Using the FPG levels is also more cost effective and allows nearly 70% of women to avoid the oGTT. Key words: gestational diabetes, pregnancy, screening, fasting plasma glucose levels. THERE IS A WIDE DIVERSITY OF OPINION regard-ing the screening for gestational diabetes melli-tus (GDM). There is no consensus on whether screening is justified,1 who should be screened,2 and what is the optimal method for management after diag- nosis.3 In Bahrain’s Government Maternity clinics we have been following the recommendations of the Fourth International Workshop Conference4 for the universal screening of all pregnant women between 26–32 weeks of gestation by use of the non-fasting 50 g oral glucose test (GCT). This is followed by the confirmatory 75 g three-hour oral glucose tolerance test (oGTT) in cases of a positive screen which includes fasting plasma glu- cose levels (FPG). Because of the complexity of this test (which requires prior appointments, ingestion of oral glucose, prolonged waiting time and ever increasing costs and pressure on an already over-stretched labora- tory service), a search for an easier and less expensive method has been studied. In 997, the American Diabetic Association5,6 announced new criteria for screening diabetes in non- pregnant patients by measuring a fasting plasma glucose Skerman ��������������������������������������������������� ��������������������� �������������������������������������� ��������:�����:�-����������������������������������������������������������������������.�-������ �������������������������������������������������������������������������������������������������)������� �������������������������������������������������������������������������������� .(�������:�������������� ��������������������������������������������������������������������������������������������������� ��������)������������������������������������������������������������ (���������������������������� ��������*%������������� .����������������������������������������������������������������������������� ������������������������������ .������ %��������������������������������� %������������ %��������������� �������������������+������.����������������������������������������������������������������������� ����������������������������������������������������������������� .������� :����������������������� / ������� ��������������������������������������� %�������������� %�������������������������������������� �������������������������������� %��������� %����������������������������������������������������:�� ��������������������� / ���������������������������������������������������������������������������� ���������������������������������������������������������������������������������������������������� %�� ������.������������ : ��������������������������������������������. mailto:yonrajab@batelco.com.bh 6 level (FPG) instead of the oral glucose tolerance test (oGTT). Between 999 and 200, progressive popula- tion studies by Peruccini et al7 in Switzerland, Agarwal from the UAE8 and Aguiar9 from Brazil have reported successes in the screening of pregnant women for GDM using the measurement of lower FPG levels. Our study aimed at evaluating the sensitivity and specificity of using FPG as a screening test for GDM compared to the 50g non-fasting glucose challenge test (GCT). M E T H O D S With the purpose of determining a FPG value with good sensitivity and specificity that could identify preg- nant women with GDM, we conducted a study between st January and 3st May 2002 in which ,600 pregnant women were screened by GCT followed by oGTT. These women were drawn from outpatients and inpatients in Salmaniya Medical Complex, the major referral hospi- tal in Bahrain. All patients who were diagnosed to have GDM (by GCT testing) had undergone the 75g oGTT which includes the FPG. The major ethnic distribution of the study was 78% Bahrainis, 9% Asians and 3% from other nationalities. The mean age distribution was 27.2 + 0.2 years, ranging from 7–48. We analysed the data using the Med/Calc statisti- cal package. A receiver-operating characteristics (ROC) curve was constructed in order to compare the ability of the FPG against the GCT in discriminating patients with a diagnosis of GDM, and to determine the best cut-off value for FPG levels, which would have the best predic- tive value for need of oGTT. R E S U LT S The incidence of GDM among the cohort of pregnant women in this study who had undergone the GCT fol- lowed by oGTT during 26–32 weeks of gestational age was 27 (3.5%) [Figure ]. Analysis of the results of the GCT group was carried out and is reported in Table . Table 2 shows the analysis of the results of those tested with the FPG. Figures 2 and 3 show the FPG level sensitivities and specificities of 94% and 93% respectively, along with the false positives encountered. The area under the curve in Figure 3 is 0.962 for FPG. The true positive rates (sensi- tivity) versus false positive rates (specificity) are plotted for determination of the cut-off value. D I S C U S S I O N Gestational diabetes mellitus (GDM) is defined as any degree of glucose intolerance with onset or first recogni- tion during pregnancy. This definition applies irrespec- tive of whether insulin or only diet modification is used for treatment, and whether or not the condition persists after pregnancy. It does not exclude the possibility that unrecognized glucose intolerance may have antedated or begun concomitantly with the pregnancy.10 In Bahrain, testing pregnant women with 75g three- hour oGTT revealed that approximately 7.8% of all preg- nancies are complicated by GDM.11 Prior to year 2000, R A J A B E T A L Table 1. Analysis of results of the fasting plasma glucose group using a receiver operating characteristics curve Positive Group Diagnosis “D” Sample Size 207 Negative Group Diagnosis “ND” Sample Size 1393 Area under ROC Curve 0.962 Standard error 0.017 95% Confidence interval 0.948–0.973 At 5.6 mmol/L: Sensitivity is 88.9 Specificity is 95.0 + LR 14.48 positive likelihood ratio - LR 0.0 negative likelihood ratio Figure 1. Distribution of negative and positive GDMs using fasting blood sugar testing under the receiving operating characteristics curve Rajab et al 3 Figure 1. >5.6 Sens: 88.9 Spec: 95.0 7 the use of the 50g GCT resulted in a higher yield of a 3.5% rate for GDM. Currently, approximately 900–200 pregnant women are diagnosed annually to have GDM.12 The prevalence may range from 7% to 4% of all pregnan- cies based on the ethnic group studied and the diagnos- tic criteria employed. In Bahrain’s Government antenatal clinics, the screening test currently used is the 50g GCT, performed between 26–32 weeks of gestation. This test involves a prior appointment, ingestion of glucose, nausea after the drink, a waiting time of one hour, and an ever increas- ing number of patients. More than 30% of those tested would eventually require a full oGTT. To explore new ways for reducing this burden on the laboratory, we should consider a test that is less expensive, more effi- cient, and more spe- cific. The other option would be the return to clinically selective cri- teria for patient screen- ing. Obviously, the first choice is more attrac- tive. In 997, the American Diabetic Association announced new diagnostic crite- ria for diabetic screen- ing, abolishing the use of oGTT, and shifting diagnosis exclusively to the use of fast- ing glucose.12 In 998, Ramachandran et al13 published a paper on using the new diagnos- tic criteria in an Asian population, suggesting the use of fasting glu- cose level as a screening S C R E E N I N G F O R G E S TAT I O N A L D I A B E T E S Table 2. Analysis of the results of the glucose challenge test group using a receiver operating characteristics curve Positive Group 217 Negative Group 1383 Area under ROC Curve 0.964–0.988 Sensitivity at 5.4 100% Specificity 93.1% Positive likelihood ratio 11.36 Negative likelihood ratio 0.00 Rajab et al 5 Figure 3. Figure 3. Receiver operating characteristics (ROC) curve of fasting plasma glucose concentration. Rajab et al 4 Figure 2. Figure 2. Sensitivity and specificity of fasting plasma glucose in detecting gestational diabetes mellitus 8 test for diabetes. There was some criticism to the use of fasting glucose instead of oGTT, based on observational studies of long term morbidity of GDM, showing that the glucose level is more accurate than fasting glucose levels in predicting future morbidity.14,15 The use of fasting blood sugar does not carry the same disadvantages, and will not obviate the need for a full GTT if this should be deemed necessary. A few stud- ies on the use of fasting glucose levels in pregnancy have since been published, pointing to the favourable use of FPG as a screening test in obstetrics.16–19 In our series we utilized the receiver-operating curve to test the ability of the fasting plasma sugar and oGTT to identify patients with GDM and find out the cut off value which predicts a GDM diagnosis. Fasting plasma glucose levels at 5.6 mmol/l yielded sensitivity and specificity of 94% and 93% respectively. When fasting plasma glucose was used, 3% of the positives required further testing while 32% needed this when the glucose challenge test was used. FPG at a value of 5.5 mmol/l is an easier, more accept- able test for patients compared with the 50g glucose challenge test. Using the FPG is more cost effective and allows 70% of women to avoid the challenge test. Future collaborative studies, pregnancy outcomes, and meta- analysis will answer all the questions related to the valid- ity of this test. C O N C L U S I O N Using fasting plasma glucose levels at cut off values of ≥ 5.5 mol/l (99 mg/dl) as the higher FPG threshold to rule out GDM with a specificity of 95% is an easier, more acceptable test for patients than the 50g glucose chal- lenge test. The fasting blood level test is more cost effec- tive and allows 70% of women to avoid the challenge test. Furthermore, 3% more cases of GDM were diagnosed using the FPG criteria compared to the glucose chal- lenge test. R E F E R E N C E S . Sacks DA, Greenspoon JS, Abu-Fadil S, Henry HM, Wolde- Tsadik G, Yao JEF. Toward universal criteria for gestational diabetes: the 75-gram glucose tolerance test in pregnancy. Am J Obstet Gynecol, 995, 72, 607–64. 2. Naylor CD, Sermer M, Chen E, Farine D. Selective screen- ing for gestational diabetes mellitus. N Engl J Med, 997, 337, 59–596. 3. Clapp JFIII. Effect of dietary carbohydrate on the glucose and insulin response to mixed caloric intake and exercise in both non-pregnant and pregnant women. Diabetes Care, 988, 2, (Suppl 2), B07–2. 4. Metzger BE, Coustan DM. Organizing Committee. Summary and recommendations of the Fourth International Workshop Conference on gestational diabetes mellitus. Diabetes Care, 998, 2, (suppl 2), B6–67. 5. American Diabetes Association. Report of the expert com- mittees on the diagnosis and classification of diabetes mellitus. Diabetic Care, 977, 20, 83–97. 6. Harris MI, Eastman RC, Cowie CC, Flegal KM, Eberhardt MS. Comparison of diabetes diagnostic categories in the US population according to 997 American Diabetes association and 980-985 World Health Organization diagnostic criteria. Diabetes Care 997, 20, 859–862. 7. Perucchini D, Fischer U, Spinas GA, Huch R, Huch A, Lehmann R. Using fasting plasma glucose concentrations to screen for gestational diabetes mellitus. Prospective popula- tion study. BMJ, 999, 39, 82–85. 8. Agarwal MM, Hughes PF, Ezimokhai H. Screening for gesta- tional diabetes in a high-risk population using fasting plasma glucose. Int J Gynaecol Obstet, 2000, 68, 47–48. 9. de Aguiar LG, de Matos HJ, de Brito Gomes M. Could fasting plasma glucose be used for screening high risk outpatients for gestational diabetes mellitus? Diabetes Care, 200, 24, 954– 955. 0. ACOG Practice Bulletin. Gestational Diabetes—Clinical management guideline for Obstetricians and Gynecologists 2000, 30, –3. . Das NF, Yousif W. Evaluation of gestational diabetes in Bah- rain. The Bahrain Conference on Diabetes in the in the Arab World. Abstracts Book, 2000, p. 80. 2. Department of Obstetrics and Gynaecology, Salmaniya Medical Complex, Kingdom of Bahrain. Annual Report, 2000, p. 2. 3. Ramachandran A, Snehalatha C, Lathe E, Vijay V. Evaluation of the use of fasting glucose as a new diagnostic criterion for diabetes in the Asian Indian population. Diabetes Care, 998, 2, 666–667. 4. The DECODE-study group. Is fasting glucose sufficient to define diabetes? Epidemiological data from 20 European studies. European Diabetes Epidemiology Group. Diabetes Epidemiology: Collaborative analysis of diagnostic criteria in Europe. Diabetologia, 999, 42, 647–654. 5. Gourdy P, Ruidavets JB, Ferrieres J, Ducimetiere P, Amouyel P, Arveiler D, et al. Prevalence of type 2 diabetes and impaired fasting glucose in the middle-aged population of three French regions—The MONICA study 995–97. Diabetes Metab, 200, 27, 347–358. 6. Agarwal MM, Hughes PF, Punnose J, Ezimokhai M, Thomas L. Gestational diabetes screening of a multiethnic, high-risk population using glycated proteins. Diabetes Res Clin Pract, 200, 5, 67–73. 7. Hanna FW, Peters JR. Screening for gestational diabetes, past, present and future. Diabet Med, 2002, 9, 35–358. 8. Reichelt AJ, Spichler ER, Branchtein L, Nucci LB, Franco LJ, Schmidt MI. Fasting plasma glucose is a useful test for detec- tion of gestational diabetes. Brazilian Study for Gestational Diabetes (EBDG) Working Group. Diabetes Care, 998, 2, 246–249. 9. Lucas MJ. Diabetes complicating pregnancy. Obstet Gynecol Clin North Am, 200, 28, 53–536. R A J A B E T A L Methods Results Discussion Conclusion References