To review current practice in centers that use the IVGTT for prediction of IDDM. To establish consensus protocol for performance of the test.


Postal questionnaires were delivered to 12 centers.


Eleven centers used a glucose dose of 0.5 g/kg and 1 used 0.3 g/kg; the dosage in adults was limited to a maximum of 25–50 g in some centers but others applied no upper limit. The glucose concentration of the infusate varied between 20 and 66%. Eight centers injected glucose manually, two used a syringe pump, and two used gravity infusion. The period of infusion ranged from 30 ± 10 s to 4 ± 2 min, and time zero was taken as the start (1 center), middle (1 center), or end (10 centers) of the infusion. The potential range in timing of the + 1-min sample varied between 1 and 7 min from the start of the infusion. Quality-assurance standards for the insulin assays used were not always appropriate for the fasting and low stimulated range of insulin levels.


The first-phase insulin response to the IVGTT is widely measured as an index of risk of progression to IDDM. We established that methodology varies widely. Because of this, a new standard protocol for use in prediction of IDDM was agreed by an ICARUS working group and is described herein.

This content is only available via PDF.