Wednesday, August 17, 2011

About the Glycemic Index - I Don't Trust It

The Glycemic Index is one of the most touted ways to judge the impact of foods on people's blood glucose. Diabetics, in particular, are told to eat foods "low on the glycemic index" in order to better control their blood sugar.

It sounds, at first, to be an elegant thing. Judge foods not just by how many carbs are in them but by what their actual impact on blood sugar is. How could that be wrong?

Well, in reading Dr. Richard Bernstein's book, Diabetes Solution, I've learned some very interesting things about the glycemic index and how it was developed that make me much less likely to ever rely on it for judging what foods I (as a pre-diabetic) can eat and still maintain good blood sugar control.

The Glycemic Index was pioneered in 1981 by Dr. David Jenkins, a professor of nutrition at the University of Toronto. The scale goes from 0 to 100, with pure glucose being 100. A glycemic index for any given food is determined by:

  1. Gather a pool of at least 10 volunteers and have them fast overnight.
  2. Feed these volunteers a sample of the food being tested containing 50g of available carbs.
  3. Take blood samples at 15 minute intervals for the first hour and 30 minute intervals for the second hour.
  4. Measure the blood glucose level in each sample for each volunteer.
  5. Average the measurements at each increment between all the volunteers.
  6. Map out these averaged measurements over time (this is called an AUC or a glucose response curve).
  7. Calculate the area under this AUC curve and divide it by the area achieve with a control (usually white bread or sugar).
  8. Multiple this number by 100 to obtain the test food's glycemic index.
Now, I'm an admitted geek and, especially after reading Diabetes Solution, I have issues with this system as the "holy grail" it's often treated as:
  1. These tests are carried out on "normal" subjects. By definition, the insuline and glucose-regulating systems of a diabetic or even someone who is insulin resistant are NOT normal.
  2. A pool of 10 individuals is a damned small pool and the results are averaged, which further casts doubt.
  3. Measurements are only carried out for two hours and the glucose response can last longer than two hours, depending on the person and the other factors involved.
  4. Each food is tested in isolation. While this is an attempt to not allow other factors to interfere, we don't tend to EAT these foods in isolation.
What I do (and strongly advocate) is test foods on myself to determine my own, individual, reaction to them. There are already foods that raise my blood sugar a lot but barely affect my husband and vice versa. What I do is fast for four hours, take a blood surgar measurement, then eat the food and start taking measurements at 15 minute intervals for two hours, then 30 minute intervals for two more hours.

Doing this means I can compare what a particular food does TO ME and make my own decision about what actions to take.

No comments:

Post a Comment