The abbreviated glucose tolerance test in screening for diabetes: the Islington Diabetes Survey.

The World Health Organization has recommended a single 2-h post-glucose load blood glucose level as a screening test for diabetes mellitus in epidemiological surveys. We have assessed its characteristics, when compared with a full supervised glucose tolerance test (OGTT), in estimating prevalence, and in diagnosing diabetes in the individual patient. A stratified sample of 223 of 1040 subjects who had participated in a diabetic survey that utilized a single capillary 2-h blood glucose estimation as a screening test were recalled for formal glucose tolerance testing. The numbers of subjects with diabetes at screening and at recall were similar (14/212, 6.6 %; 13/216, 6.0 %) but only 9 subjects were so classified on both occasions. Thirty-five subjects (16.5 %) were suspected of having impaired glucose tolerance (IGT) at screening, and 52 (24.1 %) at recall. There was substantial reclassification from screening IGT, with 3/35 worsening to diabetes, and 10/35 returning to normal. Capillary 2-h glucose levels gave an accurate assessment of the prevalence of diabetes but underestimated that of IGT. On the full OGTT, little difference in classification was found when the values of fasting and 1-h blood glucose were used in addition to those of the 2-h blood glucose used alone. The 2-h glucose had a within-subject coefficient of variation of 32.4 % which produced substantial reclassification of subjects with levels close to the diagnostic levels for diabetes, and this implies that such individuals should not be classified as having diabetes on the basis of a single glucose tolerance test. This variability is similar to that previously demonstrated in repeat OGTTs, suggesting that the unsupervised drink and the timing of the 2-h sample do not contribute substantially to the variability.