Shortcomings of the Glycemic Index

If you do low-carb long enough, someone will try to persuade you that you should not skip the “good carbs,” and when you press them further, they will say that they mean carbs that are low on the glycemic index.  The idea is, if low carb works by stabilizing blood sugar levels, wouldn’t it be better to follow a low-glycemic diet, which not only stablizes blood sugar, but also allows a greater range of “healthy” or “good” carbs like whole grains?  (I will comment only briefly on the circularity of that statement, assuming what you are trying to prove – that whole grains are healthy.)

I don’t want to dis the glycemic index too strongly - I myself rarely eat high-GI foods.  However, the GI is not the Holy Grail.  So today, I’d like to examine some shortcomings of the GI for managing healthy blood sugars and healthy body mass composition.  For anyone on the diabetes spectrum (reactive hypoglycemia, prediabetes, Type II, or Type I), the GI is a particularly flawed tool to be used cautiously, if at all.

Problems With How the GI Is Calculated

The GI is calculated by giving healthy volunteers carefully measured amounts of food and monitoring the rise in their blood glucose over a measured period of time.  The main problem is that this tells us nothing about how someone with disordered insulin metabolism will react.  Diabetics with no functioning pancreatic beta cells may see their blood sugar skyrocket and stay high in response to a food that causes a small to moderate rise in a healthy volunteer.  Diabetics with some functioning beta cells may see a smaller rise that also stays high.  Folks with insulin resistance but not much beta cell burnout may see initial high blood glucose readings followed by reactive hypoglycemia when the amount of insulin is finally high enough, and more than high enough, to remove the glucose from circulation.

Our second problem, according to Dr. Richard Bernstein’s excellent book Dr. Bernstein’s Diabetes Solution, is that the amount of carbohydrate in a food may vary by as much as 20% from what is shown on the label or in standard food count tables.  This is believable – we all know that summer tomatoes are sweeter than winter tomatoes, that the amount of nutrients in milk varies based on the cow’s diet, and that a ripe fruit has more sugar than an underripe one.  The GI researchers attempt to control for this by performing the test multiple times on each subject, then averaging the results.  But an average doesn’t predict how much blood sugar might rise if you eat a portion that is 20% off in either direction from the average.  For a diabetic trying to calibrate an insulin dosage, this variance could be a disaster.  According to Dr. Bernstein, minimizing this problem requires eating only very small amounts of carbohydrates, to lessen the impact of that 20% variance.

Third, the availability of carbohydrate varies with preparation method.  Cooking bursts cell walls, making starches and sugars more available.  Cooking and then cooling causes starches to become resistant to digestion, making them less available.  (Bix discusses resistant starch in great detail at Fanatic Cook.)  The GI researchers try to control for this, too, by taking separate measurements of foods prepared in multiple ways.  However, preparation methods, thoroughness of cooking, and serving temperature vary so much that the GI people can’t possibly test them all.

Junk Food In a Health Food Suit

Fourth, we have fructose.  Fructose does not cause blood sugar spikes; thus, foods containing fructose are typically low on the GI.  This does not mean that fructose may be consumed with impunity.  Fructose that is not immediately burned for energy is sent to the liver to be converted into triglycerides.  Triglycerides are then either stored in the fat tissue, leading to obesity, or float around in your blood, where they are a risk factor for heart disease.  (In fact, fructose is the single biggest reason I refer to fruit juice as junk food in a health food suit.  You get around 3 ounces of juice from a medium-sized orange.  That medium-sized orange is one serving, yet many people think nothing of polishing off a 16- or 20-oz bottle of orange juice, around 5 to 7 servings worth of fructose, just at breakfast!)

Beta Cell Burnout

Fifth, the GI does not measure how much insulin the body produces to cover a given rise in blood sugar.  This is important for those people on the diabetic spectrum who still have some beta cell function and want to avoid burning them out through overwork, thereby lessening their lifetime need for insulin injections.  As Dr. Bernstein points out, your pancreas has to make enough insulin to cover 100 grams of carbohydrate regardless of how fast or slow the carb is absorbed.  Whether it pumps it out quickly (when you’ve eaten a high GI food) or slowly (when you’ve eaten a low GI food) doesn’t matter much if you’re trying to preserve beta cells.

What Does It Mean?

As I said, I think the GI can be useful, but it is important to understand its flaws.  People with blood sugar issues may achieve better blood sugar control, need less insulin or oral medication, maintain a healthier body mass composition, avoid the risk of heart disease, and preserve their beta cells better if they limit the total amount of carbohydrate consumed than if they consume large amounts of low-glycemic foods.  Some people may need to combine the two methods, both limiting total carbohydrate and choosing low-GI carbohydrates, to achieve optimal control.  That is, choosing 10 grams worth of salad greens, broccoli, green beans, and cauliflower may be more helpful than choosing 10 grams worth of high-glycemic white rice.  (Furthermore, the flavor, nutrition, and eye-appeal of the green vegetables is greater.)

Most days the combined method is the one I choose, though I would be dishonest if I did not admit that I do, very rarely, trade part of my veggie allotment for a tablespoon of rice or a bite of bread.