Why I don’t use Glycemic Index, or even Glycemic Load

The glycemic index is a completely confusing tool that I kind of wish was never created. It started out as a great idea from researchers at the University of Toronto in the 1980s as a way for people with diabetes to make food choices that would help control their blood glucose (sugar). The glycemic index (GI) is a number rating for foods from 0-100 based on how much the food raises blood sugar up to two hours after eating it. The food portion to measure the GI for all foods is based on 50 grams of carbohydrate.

It seems simple: limit high GI foods and eat mostly low GI foods and your blood sugar stays in a healthy range. But confusion starts when looking at foods on the high and low GI lists. You’d expect candy, pastries, and white bread to have a high GI, which they do. But maybe you wouldn’t expect white pasta to have a low GI. Or wheat bread to have a high GI. I credit GI as the main reason why carrots and watermelon have been demonized…when they actually have a low impact on blood sugar.

Six flaws with the glycemic index:

  • The GI of a food can change. Factors can change a food’s GI such as how much it’s processed (steel-cut versus instant oats), how much it’s cooked (al dente versus soft pasta), how ripe it is (green versus ripe brown banana), and what it is eaten with (protein, fat, and fiber can all lower a food’s glycemic response).
  • It doesn’t account for serving sizes that people actually eat. GI is based on the same food portion of 50 grams of carbohydrate. That amount will look very different among foods. For example, 50 carb grams of chopped carrots is about 4 cups, whereas 50 carb grams of a super-premium chocolate ice cream is about 1 cup. Both foods have a high GI greater than 70, but most people don’t eat 4 cups of carrots at one meal, whereas it would be quite easy to devour 1 cup or more of ice cream!
  • Portion size still matters. If you eat a very large portion of a lower GI food, such as pasta, the total calories and carbohydrate can still cause a rise in blood sugar due to the sheer amount.
  • It doesn’t account for the nutrient value of a food. Whole dairy milk and cream have a low GI rating but are high in saturated fat, so not great choices if you’re trying to lower your cholesterol. M&M peanut candies also have a low GI, placing it in the same category as high-fiber bran cereal!  
  • Variability in glycemic index tables. If you look at different GI tables, chances are that the GI number will be different for the same food. An example could be potatoes. Authors of different GI tables may be examining a specific variety of potato from a specific country and using a certain testing process. All of these factors can change the GI rating significantly.
  • It measures total carbohydrate, not the amount of carbohydrate actually digested and absorbed. Whole wheat bread, yogurt, and certain fresh fruits like watermelon or pineapple have a high GI, but they also contain more fiber and/or water. Because a cup of watermelon is mostly water without much absorbable carbohydrate, in reality it will not have much of an impact on your blood sugar. This is where the glycemic load comes in.

The glycemic load was created to improve upon the glycemic index. It’s also a rating scale but which multiplies the total useable carbohydrate in a food (minus the fiber because fiber isn’t absorbed by the body) with its glycemic index. When using this rating system, carrots and watermelon are exonerated as both contain a fairly low amount of carbohydrate per serving.

I could refer you to glycemic index and glycemic load tables but I don’t want to promote them because of the confusing points mentioned above. They’re still used by nutrition researchers as one way to evaluate people’s diets and look for assocations with chronic conditions like diabetes and cardiovascular diseases. Some studies do show that a diet containing a lot of high GI foods is associated with these diseases. But I feel it’s a flawed measure because of so much potential variability, and there are other tools that can better calculate the quality of one’s diet. I also feel that it’s best used in a research setting, and not for the practical everyday meal planning of an individual. Healthy meal planning is already challenging enough!

Some patients come into the clinic saying that their well-meaning endocrinologists (medical doctors who specialize in diabetes and other hormonal disturbances) advised them to start a low GI diet but they’re confused by it. I tell them I am too. I explain the drawbacks to using GI and that glycemic load could be a better method except that even the best glycemic load tables aren’t exhaustive and don’t list many new commercial food products. Unfortunately it wouldn’t be possible to estimate the glycemic load based on similar products for popular new items like the Impossible Burger or oat milks because of the degree of processing and additives.

So what to do if you’re trying to follow an eating plan that controls your blood sugar?

In my experience working with people who have diabetes and are tesing their blood sugar regularly with meters or continuous glucose monitors, I find that methods such as eating more fiber, lowering the ratio of high-carb to low-carb foods on the plate to 1:3 or 1:4, and watching portion sizes most successfully controls blood sugar. Try these tips:

  1. Follow the balance plate model and add in as many fresh whole foods as possible. I know there are limitations with food supply and budgets, but if the choice is available, do it. You’ll likely get more fiber, more nutrients, and more flavor = greater satisfaction from the meal.
  2. When building a meal, focus first on including a lean protein and a nonstarchy vegetable. Allow some healthy fats like olive oil, nuts or seeds, nut butters, avocado. Protein, vegetables, and healthy fats contain no or minimal carbohydrates, so you’re guaranteed a lower blood sugar after eating these foods. From there, allow one portion of starch or starchy vegetable (on average, 1 cup or 1 fistful of cooked starch or 2 slices high-fiber bread).
  3. Limit between-meal snacking. Snacking is definitely ok but be careful how much you’re eating because a “snack” can easily become another meal! If possible, reach first for lower carbohydrate snacks like mixed nuts, cheese, or low-sugar or plain yogurt.

5 thoughts on “Why I don’t use Glycemic Index, or even Glycemic Load

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s