Discussing carbohydrate content of food in terms of “simple versus complex carbohydrates” is inadequate in predicting how that food will impact your blood sugar
Glycemic index is a better predictor of blood glucose response to a given food but it does not take into account the amount of food eaten
Glycemic load is one of the best predictors of glucose response to a food because it does consider the amount of carbohydrates eaten
Carbohydrates are often described as either “simple” or “complex,” with the former defined as carbohydrates composed of one or two simple sugars (fructose, glucose, or sucrose) and the latter defined as carbohydrates composed of longer chains of sugars. A common dietary recommendation is to consume more complex carbohydrates while avoiding simple carbohydrates as a blood sugar control strategy. This advice, however, is too simple as some complex carbohydrates (ex. some whole grains) have a substantial impact on blood glucose levels. This is where it is important to understand the nuances of the glycemic index and the glycemic load of different foods.
The glycemic index is a range of values from 1 to 100 that defines the relative blood glucose response a given food is expected to cause, with 100 being the blood sugar response to ingestion of pure glucose. Individual foods have had their glycemic index number determined and published. Glycemic index studies track the effects of the ingestion of a given food in four impact areas :
1. Blood glucose level
2. Insulin secretion
3. Stimulation of lipoprotein lipase and fat storage
4. Effects on the pancreas
Generally, foods that have a high glycemic index are high in carbohydrates and are quickly digested, causing a larger spike in blood glucose with a concomitant spike in insulin. This large insulin response causes a relative rebound low blood sugar and resulting hunger only a few hours after eating. Meanwhile, foods that have lower glycemic indexes are more slowly digested, causing less blood glucose and insulin volatility. While the glycemic index has proven to be more helpful than the concept of simple and complex carbohydrates, it too has its limitations. Chiefly, the glycemic index does not consider the quantity of food ingested. Imagine eating a slice of an apple versus a baker’s dozen apples: their glycemic index is identical yet there would be a dramatic difference in the resulting blood glucose. Enter glycemic load.
Glycemic load is calculated by
(Glycemic Index)/100 x (Food’s carbohydrates in grams)
Stated in words, glycemic load is equal to glycemic index of a food divided by 100 multiplied by amount of carbohydrate ingested in grams. Glycemic load can be calculated for a single food or for an entire meal if the glycemic index and amount of carbohydrates is known. The factor of 100 is used to “convert” glycemic index from a percentage to a ratio out of 1.
The following divisions are useful for estimating food impacts on blood sugar:
Low glycemic index = 55 or less
Moderate glycemic index = 56 - 69
High glycemic index = 70 or more
Low glycemic load = 10 or less
Moderate glycemic load = 11 – 19
High glycemic load = 20 or more
The nutritiondata website is an excellent resource that has the glycemic index, glycemic load, macronutrient, and micronutrient data of many different types of foods.
To solidify these concepts, let’s imagine you are eating a bagel and you begin talking to yourself about the carbohydrate battle that is taking place inside of you at that moment. If you were talking in terms of glycemic index, you may say “If I eat exactly 50g of carbohydrates in this bagel, my glucose response will be 75% that of eating 50g of pure glucose.” Remember that you must eat exactly 50g of carbohydrates to use glycemic index correctly. This is complete mayhem: you might need to eat a tenth of a bagel or a dozen bagels (if they are low carbohydrate) in order to attain those 50g of carbohydrates. The next day, you decide to eat a full bagel and use glycemic load. You read the nutritional label and see that there are 75g of carbohydrates in a full bagel. You eat a full bagel and say “Those 75g of carbohydrates will cause a glucose response that will be 1.5 times that of eating 50g of pure glucose.” You can eat as much or as little as you want and still have an idea of what your glucose response will be using glycemic load and total intake of carbohydrates.
An example using watermelon can demonstrate how discordant glycemic index and glycemic load can be. The glycemic index of a watermelon is high at 72. A serving of 100g of watermelon has 8g of carbohydrates, so the glycemic load of 100g of watermelon would be (72/100)*8 = 5.8. Next, a 1 cup serving (150g) of watermelon has 12g of carbohydrates with a resulting glycemic load of (12*72)/100 = 8.6. These numbers demonstrate that while watermelon has a high glycemic index, the glycemic load better reflects the glucose response in terms of amount of carbohydrates per portion. A one cup serving of watermelon will not impact blood glucose significantly.
The next example demonstrates how brown rice, a moderate glycemic index food, can have a large glycemic load with a resultant high glucose burden. The glycemic index of brown rice is 50. A one cup serving of cooked brown rice weighs 195g with 44.8g of carbohydrates and has a resulting glycemic load of (50/100)*44.8g = 22.4. This is considered a high glycemic load and will cause a larger increase in blood sugar.
To recap: watermelon has a glycemic index of 72 but a one cup serving has a glycemic load of 8.6. Brown rice has a glycemic index of 50 but a one cup serving has a glycemic load of 24.
Finally, it must be noted that the postprandial glucose and insulin response will be affected by the entire composition of the meal, especially the fiber content, so merely using glycemic indexes and loads is not sufficient in designing healthy meals. In a crossover randomized trial 4 meals were tested: High Glycemic Index (GI) and High Fiber (HF), Low GI and HF, High GI and Low Fiber (LF), and Low GI and LF. The results demonstrated 
Plasma ghrelin levels were only decreased in both of the Low GI groups. Simplistically, ghrelin is the major centrally acting hormone that promotes hunger.
Blood glucose was higher in the high GI/LF diet as compared to the low GI/HF. These were the only 2 groups that showed a statistically significant different blood glucose response.
The area under the curve (AUC) of plasma insulin levels following a meal was lowest for low GI/HF, followed by HGI/HF, then low GI/LF, and lastly HGI/LF. This means the least amount of insulin was released over time for the low GI/HF.
This is all important because eating meals with higher glycemic load will cause a concomitant rise in insulin, eventually leading to increased insulin resistance. Increased insulin resistance is major risk factor in chronic diseases such as diabetes, obesity, Alzheimer’s disease and heart disease.2, 4 We will be chronicling our journey of eating both low and high glycemic load foods (although hopefully more of the former!) through residency using our continuous glucose monitors. That data is incoming soon!