To compare the one-step procedure proposed by the World Health Organization (WHO) with the two-step procedure proposed by the National Diabetes Data Group (NDDG) for the identification of abnormalities of glucose tolerance during pregnancy.


One hundred twenty-seven non-diabetic Pima Indian women had a 75-g 2-h glucose tolerance test (WHO criteria). Those with an elevated 1-h glucose concentration (≥ 7.8 mmol/l) were referred for a 100-g 3-h glucose tolerance test (National Diabetes Data Group criteria). The effectiveness of the two test procedures was determined by comparing the frequency of macrosomia and cesarean section as outcomes of pregnancy.


Of 42 women with 1-h plasma glucose concentrations ≥ 7.8 mmol/l, 13 had no 100-g test, 27 had a normal test, and 2 had an abnormal test. Both women (100%) with abnormal two-step 100-g tests also had abnormal one-step 75-g tests, but only 2 of the 11 women (18%) with an abnormal one-step test had an abnormal two-step test. Sixteen of the 127 women delivered babies weighing ≥ 4,000 g. Six of these women (38%) were correctly identified as abnormal using the one-step test and one (6%) using the two-step test. Of seven women delivering by cesarean section, four (57%) had abnormal one-step tests, but none had an abnormal two-step test.


The one-step WHO test for glucose tolerance during pregnancy was abnormal in a greater percentage of women with adverse outcomes than the more cumbersome two-step NDDG test. The one-step test has the added advantage of being directly comparable to the standard glucose tolerance test used in nonpregnant women.

This content is only available via PDF.