knowing the total amount of carbs in foods was not enough to predict the effect of these

sugars on blood glucose fluctuations. Knowing the type of carbs contained in a food, simple

sugars or complex carbohydrates, provided some clues but was far from sufficient to

make a correct prediction. There are some foods that are rich in complex starches, such

as white bread, that raise blood glucose quickly and abruptly, whereas some fruits rich in

simple sugars, such as blueberries, barely raise blood glucose at all.

In 1981, Jenkins and coworkers published a landmark article in the American Journal of

Clinical Nutrition in which they proposed a new indicator that was able to measure the

speed at which the sugars contained in a specific food would raise blood glucose levels, and

they called this indicator “glycemic index”. Propelled by numerous findings about the detrimental

effects of repeated and elevated blood glucose and insulin peaks, and their risk for fat

accumulation, obesity, diabetes and cardiovascular disease, the concept of glycemic index soon

became very popular. The higher the glycemic index of a food, the

faster its carbs enter the bloodstream and raise blood glucose, causing the pancreas

to release insulin.

You may wonder how this index is calculated. In a few words, here’s how. You take a solution

of 50 grams of glucose in water, and you give it to a group of individuals. Then, for the

next two hours, you draw a drop of blood by finger puncture at regular intervals, and

you measure blood glucose concentrations. In this example, you see that blood glucose

peaks between 30 minutes and one hour, and after two hours is almost back to baseline.

If you measure the area below this curve, you know it’s proportional to the amount

of glucose that was in the bloodstream during those two hours.

Then, you take the same subjects and you have them eat a test food, say a bowl of lentils,

in such an amount that the quantity of carbs they contain is still 50 grams. Let me stress

this again: you are not giving them 50 grams of lentils, but a serving of lentils that

provides 50 grams of carbs. Now you do the same measurements, and in this example you

see that the peak is about at the same time but it is much lower, and it doesn’t decline

as fast. Of course different people will respond in different ways, but this doesn’t matter

because you can use the first measurement to standardize the result. So you calculate

the area under the curve for your lentils, you divide it by the area you obtained for

the solution of glucose, times a hundred to get rid of the decimals, and there you have

your glycemic index of lentils. If you have worked well, you will get very similar numbers

in all your different subjects. To complicate things, some labs have chosen

not to use glucose in water as a standard, but white bread. This has advantages and disadvantages,

whose discussion goes beyond the scope of our course. Just keep in mind that you can't

compare GI numbers from different tables if they are calculated using different standards.

The numbers I will show you here are all calculated using glucose as a standard.

The concept of glycemic index is often misunderstood, so let me clarify a very important point:

the glycemic index is a measure of the speed at which the carbs in a food raise blood glucose,

it is not a measure of the extent of this effect, because the GI is independent of the

amount of carbs contained in the food. Think of a man target shooting with a gun:

the glycemic index would tell us how fast he shoots his bullets, but it doesn’t tell

us how many bullets he has to shoot. If he only has a few bullets, he can shoot them

as fast he wants but he’ll still not be very deadly.

So the glycemic index only makes sense if we consider it together with another piece

of information, which is the amount of carbs contained in that food.

For example, carrots have a relatively high GI, but they also have very little sugar:

this means their sugar enters the bloodstream fast, but because it’s a very small amount,

the overall effect of carrots on blood glucose and insulin release is minimal. Conversely,

the glycemic index of potatoes is lower than that of carrots, but while an average carrot

has about 4 grams of carbs, an average potato has about 25 grams, so even if these carbs

are absorbed more slowly, they still have a significant effect on blood glucose and

insulin, because they are a lot.

To take into account both the glycemic index and the total amount of carbs, a group of

researchers from Harvard University proposed in 1997 the concept of glycemic load. It is

very simple: it is just the glycemic index of a food, times the amount of digestible

carbs contained in a standard serving of that food, divided by a hundred just to get rid

of some zeroes. This way, you have an index that considers both factors: if a food has

a low GI but a lot of sugars, its GL will still be high. Conversely, if it has a high

GI but very few sugars, its GL will still be relatively low.

Let’s go back to our previous example, and consider a GI of 72 for carrots, and of 63

for boiled potatoes. The GL of our carrot would be 72 * 4 grams of carbs, divided by

a hundred, 3. The GL of our potato would be 63, times 25 grams of carbs, divided by a

hundred, 16. So the GL of our potato is more than five fold the GL of our carrot, even

though its GI is lower. Watermelon also have a high GI of 72, but

a slice of 120 grams only has 6 of sugar, so its GL is 4. A banana of the same weight

has a much lower GI, 52, but it has 20 grams of carbs so its GL is 11, almost three times

that of watermelon. Of course the GL has one big disadvantage

compared to the GI: the GI is always the same independent of the amount of food. The GL

is always based on a specific amount of food. If I eat three slices of watermelon, the GL

would be three times higher, while the GI would still be 72.

Take this coke, its GI is 58. A small 100 mL glass, a 330 mLs can or a large 600 mLs

fast-food cup, all have a GI of 58. But the small glass has 11 grams of sugar, with a

GL of 6.4; the can has 37 grams of sugar, GI 21.5, and the large cup, 67 grams of sugar

and a GL of 38.9, and by the way, 266 calories.

The composition of a food is of course the first determinant of its GI. In particular,

two key factors are the type of carbs and the presence of other nutrients. As far as

simple sugars are concerned, glucose raises the GI much more than fructose. Sucrose is

in between since it mixes the two. As for complex carbs, the more branched starch is,

the lower the GI because it will take longer to break it down. As for other nutrients,

the presence of fiber, proteins and lipids, slows gastric emptying and intestinal absorption

of glucose, thus lowering the GI of a food. It is easy to understand why whole grains

and their derivative products have lower glycemic indexes than their refined counterparts.

But many other factors affect the GI of a food, including preparation, processing and

cooking.

Let’s consider our potato again. If we snack on it raw, its GI will be close to zero because

the starch of a raw potato is almost entirely undigestible so it will travel intact through

our small intestine without absorption. If we boil that potato, its GI will quickly raise

to 60, because starch gelatinizes with heat and becomes easily digestible. If however

we eat that same potato with its skin, its fiber will slow down glucose digestion and

absorption, and as a result, its GI will be lower. Conversely, if we turn that potato

into a mashed potato, its GI will further raise because starch becomes even more accessible

to our digestive enzymes, which will convert it to glucose even faster. Finally, if we

don’t eat our potato at all and just leave it there at room temp, within a few hours

its starch will start retrogradation becoming less and less digestible, and its GI will

progressively decrease - together with the desirability of our potato, which will become

less and less soft and buttery.

The degree of cooking always matters. If raw pasta has a glycemic index close to zero,

‘al dente’ pasta the way we Italians like it - still somewhat firm - has a GI of about

50, because part of its starch is still harder to digest. Conversely, overcooked pasta to

the point it becomes soft (that you somehow appear to like), has a GI about 10 points

higher because its starch turns to glucose much faster.

Some technological processes also fragment starch raising GI, such as extrusion (estrusione

a caldo) to make potato chips or corn flakes, or pressure-steaming (soffiatura) to make

puffed rice.

As far as fruit is concerned, the degree of ripeness influences the GI. As the fruit ripens,

its acids are converted to sugars, the texture becomes softer, and the GI tends to increase.

An overripe banana will have a higher GI than one that’s still green. And then of course,

peeling fruits and veggies always raises the GI because you remove a lot of fiber.

Whenever we cut, blend, puree or in any way turn a food into smaller pieces, the work

of digestive enzymes is easier and glucose absorption is faster, so the GI increases.

If you turn your bowl of lentils into a lentil puree, its GI will be higher. If you juice

a carrot, carrot juice will have a much higher GI than the whole carrot. And so on.

So far we have talked about individual foods, but it is important to remember that most

of the time we combine different food in a meal. When we do that, these foods mix in

our stomach and get digested and absorbed together, so what really matters in the end

is not the glycemic index or the glycemic load of the individual food items, but the

overall glycemic load of the whole meal.

Let’s suppose we eat our mashed potato together with two slices of white bread and a can of

coke. Or, let’s imagine eating the same mashed potato with just one slice of whole

wheat bread with turkey and salad, and a can of diet coke. What do you think would become

of the glycemic load of the whole meal? In the first scenario we have combined a high

GI food - our mashed potatoes - with two other GI foods, white bread and coke, so we have

turned our meal into a glycemic “bomb”, which will make our blood glucose and insulin

skyrocket. In the other scenario, we have compensated the high GI of mashed potatoes

with a fiber and protein rich, low GI whole wheat bread, green salad and turkey sandwich,

and a sugar-free glycemic neutral diet coke, thus lowering the GL of the whole meal, so

our pancreas will say thanks.

When we’ll go over food sources of carbs, I will show you some more glycemic index numerical

values, but make no mistake, I’m not suggesting that anybody should calculate the glycemic

index and glycemic loads of their meals. We said at the beginning of our course that no-one

should eat with tables and calculator. Besides, if the concept of GI is impeccable, its numerical

quantitation is rather imprecise and variable. It’s not just a problem of methodology:

food in itself is highly variable: a carrot grown in Italy can be extremely different

from a carrot grown in the US: different variety, different soil, different products used to

grow it, different weather... And then when it comes to food preparations,

such as cookies or bread, we also have different ingredients, different proportions, different

processing technologies, different cooking times or temperatures, you already know that

all these factors can greatly influence the glycemic index.

If we insist on the concept of glycemic index and glycemic load, it’s because they are

very useful to have an idea of the general metabolic effect of different foods on blood

glucose and insulin, but certainly not to make numerical calculations.

So to recap, the Glycemic Index is an indicator that tells us how fast the carbs in a food

are digested, absorbed and raise blood glucose levels. It is determined as the blood glucose

response after eating a specific food, compared to the response after eating a standard food

(glucose or white bread) providing an equivalent amount of carbohydrates (NOT an equivalent

amount of food). The GI only tells us how fast the carbs in

a food raise glycemia, NOT how much. For that, we must also know how much carb is inside

a serving of that food. Combining these two pieces of information, we can determine the

glycemic load of our meals. The general rule is that the glycemic load

of all of our meals should be as low as possible. We can accomplish this primarily in two ways:

one, by choosing low glycemic index foods, or two, by combining high glycemic index foods

with low glycemic index, fiber and protein rich foods to lower the glycemic load of the

overall meal.