Here is a discussion I first posted in my Wheat Belly Total Health book, chapter 7: Grainless Living Day-to-Day.
Glycemic index, or GI, describes how high blood sugar climbs over 90 minutes after consuming a food compared to glucose.
The GI of a chicken drumstick? Zero: No impact on blood sugar. How about three fried eggs? Zero, too. This is true for other meats, oils and fats, seeds, mushrooms, and non-starchy vegetables. You eat any of these foods and blood sugar doesn’t budge, no glycation phenomena follow, no glucotoxic or lipotoxic damage to such things as your pancreas.
There is nothing intrinsically wrong with the concept of GI nor of the related concept, glycemic load, GL, a measure that also factors in the quantity of food. The problem is how the values for GI and GL are interpreted. For instance, categories of GI are arbitrarily broken down into:
High glycemic index: 70 or greater
Moderate glycemic index: 56-69
Low glycemic index: 55 or less
This is like being a little bit more or less pregnant. By this scheme, cornflakes, puffed rice, and pretzels have “high“ GIs above 70, thereby sending your blood sugar through the roof, while whole grain bread, oatmeal, and rice have “low” GIs. A typical non-diabetic person consuming a typical serving of cornflakes, e.g., 1 cup cereal in ½ cup milk, will thereby experience a blood sugar in the neighborhood of 180 mg/dl—very high and more than sufficient to set the process of glycation and glucotoxicity on fire, add to adrenal disruption, cataract formation, destruction of cartilage, hypertension, heart disease, and neurological deterioration or dementia. (Blood sugars will vary, of course, depending on body weight, degree of overweight, insulin sensitivity, time of day, and other factors, but this would be typical. Someone with pre-diabetes or diabetes will have a higher blood sugar.)
How about a low-glycemic index food, such as a bowl of oatmeal, 1 cup cooked, in ½ cup milk? A typical response: blood sugar 170 mg/dl—lower, yes, but still quite awful, triggering all the same undesirable phenomena triggered by the high-glycemic cornflakes. This is why I believe “low” GI is more accurately labeled “less-high” GI, not “low.” Alternatively, we could just recognize that any GI above single digits should be regarded as high because it’s not until you get to single digits or zero that blood sugars no longer range into destructive levels.
The concept of “glycemic load” tries to take this into account by factoring in portion size. Thus the GL of cornflakes is 23, while the GL of oatmeal is 13 and that of whole wheat bread is 10. GL is usually interpreted as:
High glycemic load: 20 or greater
Moderate glycemic load: 11-19
Low glycemic load: 10 or less
Once again, this lulls you into thinking that foods like oatmeal or whole wheat bread don’t raise blood sugar—but they do. They are not low glycemic load; they have less high glycemic loads.
The value that truly appears to count and predict whether or not we will have a blood sugar rise? Grams of carbohydrate. Specifically, “net” grams of carbohydrate calculated by subtracting fiber:
“Net” carbohydrates = total carbohydrates – fiber
Net carbohydrates is a concept popularized by the late Dr. Robert Atkins, who recognized that fiber has no impact on blood sugar despite being lumped together with other carbohydrates. (Fiber is technically a carbohydrate, or polysaccharide, but humans lack the enzymes to digest fibers into sugars.) In other words, there is really no need for manipulations such as glycemic index or glycemic load and their misleading classifications.
If you were to test blood sugars with a common fingerstick glucose meter (as many of us, diabetic and non-diabetic, often do to gauge the effect of different foods) 30-60 minutes after consuming a food, you would see that it takes most of us around 15 grams of net carbohydrates before we begin to see a rise above the starting value. We check blood sugars at 30-60 minutes after consuming a food. The peak can actually occur before or after 30-60 minutes, depending on the mix of protein, fat, fiber, the amount of water or other liquids, pH of the food, and other factors. This is just an approximation of peak blood sugar that allows you to perform a single fingerstick, rather than every few minutes. What we don’t do is check blood sugar two hours after consumption, as advised by most physicians interested in blood sugar control on insulin and other diabetes medications. This seems obvious, but is a common tripping point when discussed with doctors.
Ideally, allow little to no rise in blood sugar. In this way, you have turned off any excess levels of glycation and glucotoxicity, undo the effects of high insulin and insulin resistance, and allow fasting blood sugars to head downward over time, while also accelerating weight loss.
There is another common fiction—or perhaps half-truth might be a better term—prevailing in nutritional thinking offered by the dietary community that tells us that, if a high-glycemic index food is consumed along with proteins, fats, or fiber, i.e., foods with low- or zero- glycemic index, that the net glycemic effect will be much improved. For this reason, dietitians often advise people to consume, say, bread with peanut butter: the high blood sugar potential of the bread is blunted by the low-glycemic protein, fat, and fiber of the peanut butter. As often occurs in the flawed logic of nutrition, this is another example of something being less bad, not necessarily good. For instance, a typical blood sugar in a middle-aged mildly overweight male after consuming two slices of multigrain bread made with whole wheat flour, oats, and millet on an empty stomach might be 170 mg/dl—high enough to provoke insulin, cortisol, insulin resistance, visceral fat accumulation, inflammation, glycation and glucotoxicity, and add to dementia risk. In another session, again starting with an empty stomach, the same male consumes two slices of multigrain bread, but this time with several slices of turkey (mostly protein), mayonnaise (mostly fat), and lettuce (mostly fiber and water). Blood sugar: 160 mg/dl—better, yes, but still pretty awful and more than sufficient to generate all the phenomena generated at a blood sugar of 170 mg/dl, including brain atrophy.
Less bad is not necessarily good. Feel free to count your carbs, but ignore the misleading concepts of glycemic index and glycemic load. Use those tables of glycemic index you might have to line your box of cat litter, but don’t use them to construct a healthy diet.