When glucose (a sugar) is absorbed from the gut into the bloodstream, a healthy pancreas will secrete the hormone insulin in response. One chief effect of insulin is to reduce blood sugar levels by facilitating the transfer of sugar into the body’s cells. This is a critically important function, and without it sugar levels can rise to dangerously high and even fatal levels. Insulin is essential to life. But, as is the way, we can always get too much of a good thing. In the long term, high levels of insulin can predispose the body to a range of health issues including obesity, heart disease and type 2 diabetes.
With this in mind it does make sense to avoid eating much in the way of foods that disrupt blood sugar, and therefore insulin, levels.
The ‘macronutrient’ most renowned for its effect on blood sugar is carbohydrate – a term that encompasses sugars and starch. Much of the sugar and starch we consume in our diets is digested down to glucose prior to absorption.
However, carbohydrate foods come in very wide array of forms, and they are most certainly not created equal. To begin with, some pack much more nutritional punch than others (e.g. blueberries or broccoli are more nutritious than white bread or cornflakes). But also the speed and extent to which they release sugar into the bloodstream (and also induce insulin) varies enormously too.
The standard way of gauging the extent to which foods raise blood-sugar levels is to compare it to the body’s response to consuming glucose (very fast sugar-releasing). In the world of the GI, glucose is assigned a GI value of 100, and all other foods are compared to it. Basically, the higher a food’s glycaemic index, the more disruptive its effects on blood-sugar and insulin levels are.
While the GI is an important measure, it’s not the sole arbiter of a food’s suitability for our consumption. How much we eat of that food is also critical. Foods with a relatively high GI will be most disruptive if we eat a lot of them. The eating of foods of high or medium GI matters less if we don’t tend to consume too much of them at any one time.
Let’s see how this may play out in real life. Imagine coming home ravenous and polishing off a bowlful or two of pasta in response. This is likely to cause considerable disruption in blood-sugar and insulin levels. On the other hand, it’s highly unlikely that you would find yourself gorging on mounds of beetroot or pineapple for dinner, however hungry you were. When eaten, such foods tend to be eaten in very moderate amounts, and thus are unlikely to disrupt sugar and insulin levels much.
This concept has spawned the development of another measure of the effect of food in the body, known as the ‘glycaemic load’. The glycaemic load of a food is calculated by multiplying its GI by the amount of carbohydrate contained in a standard portion of the food. This figure is then divided by 100.
While measuring GL values has some merit, the concept is based on standard portion sizes. However, people don’t tend to eat ‘standard portion sizes’, so applying GL values listed in a table has little relevance to individuals in the real world. Looking at GL values does, though, help us gain some idea of how disruptive starchy carbs generally are in comparison to, say, fruits and vegetables.
What constitutes a high or low GL is open to debate. But as a rough guide I recommend seeing a GL of 20 or more as high, and one of 10 or less as low. Anywhere in between can be classified as medium.
Considering GL values can cause a different picture to emerge than the one painted when judging foods by their GI alone. For example, many of the foods that have medium or high GIs turn out to have low GLs. Examples include kiwi fruit (GI 53, GL 6), pineapple (GI 59, GL 7), watermelon (GI 72, GL 4), cooked carrots (GI 58, GL 3) and beetroot (GI 64, GL 5).
On the other hand, the potato and many grain-based foods of relatively high GI have high GLs, too. Examples include bagels (GI 72, GL 25), white rice (GI 65, GL 23), rice cakes (GI 78, GL 17), cornflakes (GI 81, GL 21), baked potato (GI 85, GL 26) and pasta (GI 44, GL 21).
To my mind, the GL is a more useful measure than just knowing the carbohydrate content of a food. This is principally because, as we know, two foods of similar carbohydrate content can have wildly different effects on blood sugar balance. It stands to reason, therefore, that GL values are better than carbohydrate values in judging the effects of a food or meal on blood sugar and insulin levels.
I say this stands to reason, but until recently, this assumption had not been tested in any systematic way. This week, though, saw the publication of a study in which individuals were fed single foods and meals of mixed foods, after which their blood sugar and insulin levels were monitored .
The researchers then assessed the relationship of the GL values of the foods and meals and the levels of sugar and insulin subsequently. They did the same for carbohydrate content of the meals too.
In short, what they found is that GL values were a much better predictor of subsequent sugar and insulin levels after eating.
In other words, what theoretically made sense, does seem to translate into the real world on this occasion.
1. Bao J, et al. Prediction of postprandial glycemia and insulinemia in lean, young, healthy adults: glycemic load compared with carbohydrate content alone. Am J Clin Nutr 16 February 2011 [epub ahead of print].