To the Editor: The authors of the article on insulin resistance1 challenge users of serum insulin levels to demonstrate benefit in clinical practice.

The problem for those of us who see patients with the metabolic syndrome (and occasionally find time to take a history and perform an examination) is that we are spectacularly unsuccessful in getting patients to lose weight. Everyone is able to lose some weight in the first flush of enthusiasm or when taking part in a trial with lots of encouraging assistants monitoring them, but they almost invariably run out of steam, and the weight goes back on — with interest.

I use the “worse still” measurement of insulin responses during an oral glucose tolerance test not to diagnose the metabolic syndrome, but as a method of motivating patients about the seriousness of the condition. A graph that shows the insulin response “off the page” makes a very effective propaganda tool (Box). An elevated fasting insulin level or, even better, very high insulin levels during an otherwise normal glucose tolerance test, is also very effective in indicating to parents of well children with obesity that diabetes is threatened. Otherwise, we are left with vague references to increased risk, which means little to the parents. One picture is worth a thousand words.

A typical glucose tolerance test result showing the insulin response “off the page”

Dots = measured glucose levels; dotted lines = glucose normal range. Crosses = measured insulin levels; shaded area = insulin normal range.