Friday, 23 January 2015

Glycemic index makes fools out of people

(Originally posted on Thursday, 13 July 2017)

Recently I “discovered” what glycemic index really means and I am pissed off because nobody has pointed it out clearly to me. It is NOT explained properly even on the Internet sites where there are whole tables with glycemic indexes for many different foods!

My wife suffered from gestational diabetes (diabetes during pregnancy) and she ate mostly foods with low glycemic index. Unfortunately, even though she had changed her diet very significantly she still ended up using insulin. She did her own thorough search about diabetes on the Internet then, so I didn't analyzed the topic myself. The only thing I remember from that time is that “white bread is bad because it has a high glycemic index”. I didn't believe it then and I don't believe it now. To make a definite opinion I had to understand what the glycemic index really is.

The problem is that hardly anybody explains the methodology behind the glycemic index OR sometimes the methodology is explained incorrectly!!! What’s worse some Internet sites don’t give precise numbers and just group different foods into 3 categories based on their glycemic indexes (good, medium or bad). Even worse, some isolated results for some foods are so strange (compared to other sites) that there had to be a mistake somewhere (as always, you shouldn’t base your knowledge on just one Internet site).

Below there are glycemic indexes for several foods that are rather universal, except for honey. It turns out that the glycemic index of honey depends heavily on what kind of honey it is. Some honeys are mainly glucose, which makes the index very high, and some other honeys have much more fructose in them, which makes the index much lower.

Glycemic index (GI):
1. Glucose (the base for all the glycemic indexes): 100
2. Baguette French bread: 95
3. White wheat flour bread (average): 75
4. Honey (average): 70
5. Sucrose (white sugar, table sugar): 65 (!!!)
6. 100% Whole Grain bread: 50
7. Pumpkin seeds: 25
8. Fructose: 20
9. Mayonnaise (classic, homemade): 0

Looking at the results above some people may come to some strange conclusions, for example that white wheat flour bread is worse than white sugar. Or that any amount of such bread is bad just because it has a “high glycemic index”. Or that it is safe to eat lots of fructose (pure sugar from fruits) just because it has a “low glycemic index”. Some Internet sites really say such things, but I always felt that there is something wrong there. Now I know what the problem is – it is the methodology behind the glycemic index or rather the meaning of the results given by this methodology.

Buried somewhere on this site:
https://en.wikipedia.org/wiki/Glycemic_index
there is a rough description of the methodology behind the glycemic index:

“The glycemic index of a food is defined as the incremental area under the two-hour blood glucose response curve (AUC) following a 12-hour fast and ingestion of a food with a certain quantity of available carbohydrate (usually 50 g).”

Basically it’s a relative comparison of one kind of food to glucose as far as blood results are concerned measured after 2 hours after eating the food. The problem is that it does NOT compare the same amounts of foods!!!

For example let’s analyze the glycemic index of bread calculated in reference to 50 g of glucose. In a typical bread there are around 50 % carbohydrates, so to deliver 50 g of carbohydrates a person needs to eat at least 100 g of bread. And this is how the glycemic index of bread is calculated – after eating at least 100 g of bread. But it is calculated relatively to the effect of 50 g of glucose! And this is exactly what makes glycemic index so unnatural. And vastly misunderstood.

What’s worse the methodology uses the term “available carbohydrate”, which means that the amount of eaten food was even bigger than it may seem at first. Let’s ASSUME that 20 % of carbohydrates in bread can’t be digested, so only 80 % of carbohydrates in bread can be labelled as “available carbohydrates”. It means that there are only 40 % “available carbohydrates” in bread. So to calculated the glycemic index of bread a person needs to eat 125 g of bread (125 * 40 % = 50 g). So, the glycemic index of bread actually shows the effect of eating MUCH more bread (125 g) than 50 g of glucose that it is referred to. Strange isn’t it?

On top of that, or should I say at the base of that, there is a problem that is completely ignored: it is hard to precisely calculate the percentage of carbohydrates in every food and it is even harder to precisely calculate the percentage of “available carbohydrates” in every food. Different calculation of a particular glycemic index may be carried out with different assumptions about the amount of “available carbohydrates” in the analyzed food. It would explain why some glycemic indexes of the same food are sometimes so different from each other.

The truth is that it is MUCH better to eat 50 g of ANY bread than 50 g of white sugar just because there are only around 25 g of carbohydrates in 50 g of bread (compared to 50 g carbohydrates in 50 g of white sugar).

Some people might say that my way of thinking is wrong because the amount of delivered carbohydrates is different. I disagree. I believe that we SHOULD compare foods this exact way. The problem in most of today’s societies is that people weigh too much and it is caused by them eating too much, especially too much carbohydrates. When people start eating less carbohydrates it will be good for them. Of course we shouldn't avoid carbohydrates completely, because it is the main source of energy for people, but on the other hand we should not “maintain” the same level of carbohydrates consumption when we see that our weight is not what it should be.

A better way to compare foods is the glycemic load (instead of glycemic index):
https://en.wikipedia.org/wiki/Glycemic_load

“The glycemic load (GL) of food is a number that estimates how much the food will raise a person's blood glucose level after eating it. One unit of glycemic load approximates the effect of consuming one gram of glucose.[1] Glycemic load accounts for how much carbohydrate is in the food and how much each gram of carbohydrate in the food raises blood glucose levels. Glycemic load is based on the glycemic index (GI), and is calculated by multiplying the grams of available carbohydrate in the food times the food's GI and then dividing by 100.”

The above description is strange, but the idea is to compare the same amount of food instead the same amount of “available carbohydrates”. The formula for the glycemic load is actually pretty obvious:
GL = GI * % of “available carbohydrates” in a particular food

In the above example (with the GI of bread) it would be like this:
GL = GI * 0.4,
which is consistent with the actual amounts of foods used – the GI of bread compared the effect of 125 g of bread with the effect of 50 g of glucose, so to make the foods comparable as far as they actual weights are concerned we have to makes this calculation:
GI * 50 g / 125 g = GI * 0.4.

To calculate glycemic loads for the foods mentioned above I needed the percentages of “available” carbohydrates in the foods, but I found on the Internet only tables with percentages of overall carbohydrates (sums of “available” and “unavailable” carbohydrates per 100 g). What's worse the percentages for the same foods were sometimes very different from each other (on different Internet sites). I decided to take the MAXIMUM overall carbohydrates percentages that I found on the net, which means that I calculated the WORST possible values of glycemic loads.

Overall carbohydrates percentages in particular foods:
1-3. Glucose (the base for all the glycemic indexes): 100 %
1-3. Sucrose (white sugar, table sugar): 100 %
1-3. Fructose: 100 %
4. Honey: 80 %
5. Baguette French bread: 60 %
6. White wheat flour bread: 56 %
7. 100% Whole Grain bread: 50 %
8. Pumpkin seeds: 18 %
9. Mayonnaise (classic, homemade): 0 %

Glycemic loads (relative to glucose):
1. Glucose (the base for all the glycemic indexes): 100
2. Sucrose (white sugar, table sugar): 65
3. Baguette French bread: 57
4. Honey (average): 56
5. White wheat flour bread (average): 42 (!!!)
6. 100% Whole Grain bread: 25
7. Fructose: 20
8. Pumpkin seeds: 4.5
9. Mayonnaise (classic, homemade): 0

The list for the glycemic load is significantly different than the list for glycemic index! Now it is clear that “white bread” is not so “lethal” after all. Its effect is only 42 % of the effect of glucose and 65 % of the effect of white sugar. Sure the 100% Whole Grain bread is better than “white bread” as far as glycemic load is concerned, but what do the numbers above actually mean? They mean that instead of eating 1 slice of white wheat flour bread you “can” eat 1.68 slices of 100% Whole Grain bread (42 / 25 = 1.68) AND the effect after 2 hours after eating will be exactly the same. However this “can” in fact should be said “can't”. Why? Because the glycemic loads regard ONLY what will happen exactly after 2 hours and ignore completely everything else. For example glycemic loads don't “show” what will happen after 3, 4 or more hours after eating! Remember that all those “available carbohydrates” don’t disappear, but are just digested at a different rate! They will simply “load” some other points in time. We'll get back to this issue later.

I noticed yet another strange thing – on the Internet there are some tables with glycemic loads for different foods, but they are calculated for “servings” of different weights! This is ridiculous! It makes the foods still incomparable! However, that made me come up with a cool idea – I decided to calculate glycemic loads for the weight equal to the weight of glucose that is usually used to calculate the glycemic index (50 g).

Glycemic loads (relative to glucose) for “servings” of 50 g:
1. Glucose (the base for all the glycemic indexes): 50
2. Sucrose (white sugar, table sugar): 32.5
3. Baguette French bread: 28.5
4. Honey (average): 28
5. White wheat flour bread (average): 21
6. 100% Whole Grain bread: 12.5
7. Fructose: 10
8. Pumpkin seeds: 2.25
9. Mayonnaise (classic, homemade): 0

Now we can calculate how many grams of food a person can eat to get the same effect after 2 hours after eating (relatively to 50 g of glucose):

1. Glucose (the base for all the glycemic indexes): 50 g
2. Sucrose (white sugar, table sugar): 77 g
3. Baguette French bread: 88 g
4. Honey (average): 89 g
5. White wheat flour bread (average): 119 g
6. 100% Whole Grain bread: 200 g
7. Fructose: 250 g (!!!)
8. Pumpkin seeds: 1111 g (over 1 kg!)
9. Mayonnaise (classic, homemade): infinity!?

Please notice that the numbers are consistent with the example with the breads: 200 g / 119 g = 1.68, which still means that instead of eating 1 slice of white wheat flour bread you “can” eat 1.68 slices of 100% Whole Grain bread AND the effect after 2 hours after eating will be exactly the same. As I wrote earlier this “can” in fact should be said “can't”. Let’s analyze a little more the example with the breads. Those 1.68 slices of 100% Whole Grain bread will actually make your carbohydrates consumption (50 % * 1.68 slices * the weight of 1 slice = 0.86 of the weight of 1 slice) significantly higher that when eating just 1 slice of “white bread” (56 % * 1 slice * the weight of 1 slice = 0.56 of the weight of 1 slice). I made similar calculations relatively to 50 g of glucose.

The amount of carbohydrates consumption leading to the same effect after 2 hours after eating (relatively to 50 g of glucose):

1. Fructose: 250 g (!!!)
2. Pumpkin seeds: 200 g
3. 100% Whole Grain bread: 100 g (!!!)
4. Sucrose (white sugar, table sugar): 77 g
5. Honey (range): 71 g
6. White wheat flour bread (average): 67 g
7. Baguette French bread: 53 g
8. Glucose (the base for all the glycemic indexes): 50 g
9. Mayonnaise (classic, homemade): 0

Interestingly after I made all those steps in my analysis I ended up with a list that could be calculated from the first list with the assumption that the base is 50 g of glucose. Bravo me! But I would never understand the meaning of the last list without all those little steps I made earlier, especially the list for the glycemic load.

Again the numbers are consistent with the example with the breads: 100 g / 67 g = 1.5 = 0.86 of the weight of 1 slice / 0.56 of the weight of 1 slice.

What this last list means? It means that eating lots of foods with low glycemic index or low glycemic load actually makes the overall situation worse! Don’t fix yourself on the glycemic index nor on the glycemic load! They have to be analyzed together with the overall carbohydrate consumption. In fact the glycemic index and the glycemic load are a secondary issue, behind controlling the overall carbohydrate consumption.

Basically the most important thing is the overall amount of eaten foods, especially foods with high carbohydrates percentages, mostly sugars. And the foods with low glycemic loads only help you prevent rapid increases and decreases of the level of glucose in your blood. And that’s it. Eating 1 slice of ANY bread every 2.5 or 3 hours is more than enough to keep you satiated, especially when it is eaten with a little cooked meat and/or a little vegetables. Consumption of big amounts of food should be generally avoided – even lunch or dinner should be eaten in modest amounts. The same goes for fruits and “sweet vegetables”. The rule “Eat little, but often” is still viable. In fact it is as viable as ever.

By the way, I found some interesting things about breads in general:
https://greatist.com/eat/best-healthy-bread

I have to also point out that fructose in high quantities is very BAD on its own:
http://ajcn.nutrition.org/content/88/5/1189.full

Overall, any “sweet water” is NOT healthy:
https://authoritynutrition.com/fruit-juice-is-just-as-bad-as-soda/

Obviously mayonnaise can't be eaten in big amounts either, even though both its glycemic index and its glycemic load are zero. I myself eat mayonnaise regularly in small amounts and I believe that people should NOT avoid it completely.

Some time ago I've become a “fan” of pumpkin seeds. Apparently they are one of the “super foods” and their low glycemic index and their very low glycemic load are only two of many of their virtues:
http://superfoodprofiles.com/raw-pumpkin-seeds-nutrition

PS. Analysing carbohydrates consumption is much more terrifying than thought. I bought a small kitchen scale that allows taring and weighted several foods and liquids. Then I made some calculations and I got one shocking result.

In one mug (250 g) of very “healthy” juice (based on 44 % carrots,
11 % apples and 9 % oranges, with some cane sugar) there were
26 grams of pure sugar! This may seem like not a big deal, but I checked how much is that compared to an average teaspoon – it's a little more than 4 (FOUR) heaped teaspoons or a little more than 8 (EIGHT) level teaspoons!!! That's around 10 times as much sugar as I put in tea in the same mug!!!

Let's compare the “healthy” juice to the “very unhealthy white bread”. There are only 20 grams of carbohydrates in one slice of “white bread” (that weights around 35 grams), compared to 26 grams of carbohydrates in one mug of the juice. More importantly one slice of “white bread” will definitely keep you more satiated than one mug of “sweet juice”. The truth is that “healthy” juices can in fact be a carbohydrate trap, not to mention other “sweet waters”.

I think that we should drink liquids just to deliver water to our bodies not carbohydrates. Sure there can (or even should) be a little sugar in the water (for example in the form of concentrated syrup) but ONLY
3 grams per mug. Not 26 grams per mug!

No comments:

Post a Comment