--I feel that one has to try to get such testing done in an environment that is not likely to skew the results.

To that end, I don't necessarily think eating a carbohydrate rich diet a few days before the test is a good idea, if that is not what one normally does. Then, too, I don't think eating a very carbohydrate restrictive diet a few days before the test is a good idea, either, if that is not what one normally does. I think either might skew the test results somewhat. I do recognize that there are many people out there on special diets who might need to explain this to their physicians before the test.

There are various meds that do skew the results, and these should also be mentioned to physicians before the test begins. Anything that might interact with the rate or completeness of glucose absorption might alter the results.

One thing I do know--during the test, one should not do strenuous, or even not so strenuous, physical activity. That DEFINITELY alters results. (I'm amazed they let you go out and return, Doc, as they'd have no idea what you might have been doing during the absences. When I get the tests done, I know to bring a book or an ipod or something for the duration--I'm not going anywhere until the test is done.)

I still think that a more optimal glucose tolerance test is longer, with more frequent blood draws (half hour instead of an hour, in order not to miss rising/falling levels/patterns) and with insulin levels measured along with glucose levels. I have mine done for 4-5 hours. I admit this is not easy to tolerate for most, and the constant sticking--one can get stuck 9-10 total times--is a drag. I do alternate arms, and try to come in well hydrated, so there are different spots to access veins as time goes on.

What one is looking for is not only the glucose levels over time, but how they rise and fall in consonance with insulin levels. Even a fasting insulin level may be an interesting piece of info.