--that the pattern is not suggestive of an insulinoma.

An insulinoma would be more likely if Alan had a low fasting blood sugar that continued to drop with continued fasting; the standard test in bygone years was a fast of up to 72 hours. If insulin secretion is not suppresed in the presence of falling blood sugar--they have to be measuring both insulin and glucose during the fast--the likelihood of insulinoma rises.

Since the insulin/glucose balance is a mutual feedback mechanism, in normal people, the body tries to maintain a baseline blood sugar level. If no additional food is being ingested, there is no need for insulin to be produced in appreciable quantities, as there is no need to push excess glucose into the cells (there is no excess glucose) and the insulin level drops. Normal people on fasts often have fairly in-range glucose levels (unless the fast is very long, or they have no fat stores) but low insulin levels until they start consuming food again. But in people with insulinoma, insulin levels stay high, as the tumor is producing excess insulin.

Moreover, insulinomas tend to produce high levels of "proinsulin", a slight chemical variant--generally, "normals" do not have more than about 20% of their serum insulin identifiable as proinsulin, whereas with insulinomas, generally there are percentages of 24 or higher. High c-peptide levels can also be indicative. (These can all be measured.)

While I agree that Alan's insulin levels should be measured (and I wish they did it during the 3-hour test), they should also measure his proinsulin and c-peptide levels. But, while I'm not a physician, I do not suspect insulinoma in him, primarily because he does not seem to exhibit the fasting hypoglycemia or the disorientation, palpitations, weakness, and weight gain associated with it.

Now, there are many other reasons his blood sugar may have gone that low in the third hour, and that should be investigated--these could range from other endocrine disorders to pre-diabetic insulin resistance. (I so wish they had taken samples every half-hour along with insulin levels during the three-hour test--would have been interesting, and likely instructive, to see the patterns in the rise and fall of both. If they do a multi-hour test again, insist they do it with half-hour measurements and insulin levels.)

More places to learn about this:

www.emedicine.com/med/topic2677.htm

http://www.indegene.com/JIACM/indJIA...glycaemia.html