Talk:Glycemic index: Difference between revisions

Definitions

Okay ... but what is the Glycemic Index ? Where do these numbers come from? Are they arbitrarily assigned or do they represent the end product of some calculation? I think an encyclopedia entry would have to relate this! Zuytdorp Survivor 22:47, 8 Feb 2004 (UTC)

I believe they're measured by having a group of people eat 100 grams of <carbohydrate>, then taking blood samples every few minutes to measure the blood sugar level. The index value is either the height or the time of the average peak relative to glucose. -- wwoods 08:32, 18 Mar 2004 (UTC)

wwoods is on the right track. I'm not sure about the "every few minutes". I have read that the index value is based on the area under the graph, but that doesn't quite square in my mind with the information that the index is supposed to provide, so I want to research this more before (if nobody else does it sooner) editing the page with the info rightly requested by Zuytdorp Survivor.

Right now I'm doing a major edit, which includes some deletions, so I want to explain the main features. For context, I've just been diagnosed with diabetes and so I'm doing a lot of research to understand GI and other issues so that I can repair what is repairable and learn to manage and live with what is not.

Key changes:

Deleted erroneous information about sugars and starches/complex carbohydrates. Dietary advice used to favour complex carbohydrates over sugars because the notion made sense that the body takes longer to process complex carbohydrates than sugars. After all, the processing involves breaking them down into sugars. However, research proved conclusively that the notion was incorrect. The starch in many foods is converted to blood glucose faster than table sugar! These findings were what prompted the development of the GI as an indicator of carbohydrate quality.

Deleted the examples of GI values - too few to be useful. I'll be working on something useful.

Deleted the links. One was less informative than any of the many pages I've found in my research. The other, with a GI list, was (like nearly all the GI lists I've found) an ad for a book. There are a great many such lists on the web, and I can't see grounds for selecting a particular one to promote. Moreover, short GI lists, designed to be of practical help to individuals, are culturally restrictive. They list foods thought to be most likely to be used by their target audience.

I question the practice of linking to pages that provide info that should be in Wikipedia, but maybe it's valid as a stopgap, an aid to readers while content is in preparation. So I've linked to what I understand to be the most comprehensive and authoritative list of GI values, in three ways. One is a searchable database, presumably up to date. Then there's the research group's data published in the American Journal of Clinical Nutrition in 2002. Finally, there's a simplified version of that list published by Rick Mendosa.

My intention is to try to condense the information from these sources into a list that is still comprehensive and culturally unbiased, but of practical use to individuals who want to control their carbohydrate diet. --Richard Jones 11:37, 20 Mar 2004 (UTC)

---

I think the last paragraph is wrong. GI is the measurement of how the blood sugars raise: foods with fat do lower the rate of which they raise. So how is it a 'false' rating? Fineform, if that was the source at which you got the information, was trying to say that foods aren't healthy just because of a low GI; not that they don't have a false GI. That, in any cae, is biased in favor of a low fat diet. Juan Ponderas

---

The comment suggesting that a short fall of the GI is that it only measures glucose should be removed. While this is true that it measures only glucose, it is not a shortfall as it is *glucose* we are concerned about and high glucose levels (spikes) which have been implicated in disease aetiology. It actually does measure the glycemic response from fructose (after conversion to glucose) as low GI and is therfore extremely valid. This explains why honey has a low GI and table sugar a moderate one. The glycemic response correlates very strongly with insulin response. The statement written in the article is incorrect. I will add to this article soon hopefully when I have more time.--sd 03:58, 2 Feb 2005 (UTC)

Just wrote a major revision and tightened it up in many places. I have a good image to load but need to get clearance from the University of Sydney first.

Booze

The sentence "Alcoholic beverages are also low GI" seems like, at the very best, a gross oversimplification. Beer contains maltodextrin, which has a relatively high GI, right? Let's at least use a qualifying word like 'some,' or get specific and talk about red wine, which as a lower GI than beer.

---

Beer contains, on average, 10 g of carbohydrate per serve. The amount is negligible for raising blood glucose levels when drunk at a reasonable rate. If, on the other hand, someone were to drink 6 beers in an hour than it may become significant. It is therefore incorrect to say that red wine has a lower GI than beer - it is irrelevant.

Check the link to the GI database. They say that foods with very low carbohydrate content do not have well defined GI values. --Slashme 11:14, 26 October 2005 (UTC)

---

Vegetables

Carrots do NOT have a high GI. This was inaccurate and incorrectly reported data back in the 1980s. Unfortuantely the myth prevails. Most fruit and veg have a low GI by virtue of their very low carb content per serve. The rate of absorption in such foods is irrelevant.

---

From the GI database:

Food	GI	Serving	Carbs/serving	GL
Carrots, raw (Romania)	16	80	4.2	0.7
Carrots, peeled, boiled(Australia)	41	80	4.6	1.9
Carrots (Canada)	92	80	4.2	3.9

You're thinking of glycemic load. Carrots (and many other healthy vegetables) can indeed have realatively high GI values in spite of their low carb content, because GI is defined in terms of an amount of glucose equal to the amount of carbs in the food tested. However, this does not make them unhealthy foods, as their glycemic load is low. --Slashme 11:14, 26 October 2005 (UTC)

Notice the large difference between the values. This occurs because you're measuring such a small effect that small variations can completely throw the result.

Glycemic load

• The external link [1] lists Glycemic Load aside the Glycemic Index, which sounds very helpful. But why does it depend on the "serve size"?
• For example watermelon. "Its glycemic index is pretty high, about 72. According to the calculations by the people at the University of Sydney's Human Nutrition Unit, in a serving of 120 grams it has 6 grams of available carbohydrate per serving, so its glycemic load is pretty low, 72/100*6=4.32."
• Why not take the percentage of carbohydrates in it, 5% (6/120) in this case, and multiply the glycemic index by that? 0.05*72=3.6. To not use a percentage sounds pretty strange, why would anyone not do that? - Jerryseinfeld 18:04, 16 Oct 2004 (UTC)

Actually, the GL does not depend on the serving size. That 120 grams they quote makes absolutely no difference to the calculated GL. Dunno why they chose to express it that way. Wikipedia (obviously a superior site) doesn't. --Slashme 11:23, 26 October 2005 (UTC)

I have a puzzle for you

The effective ranking of foods into three distinct categories is important for functional use of GI. High=x range Med=y range Low=z range

If I change the reference food, why do these ranges stay the same?

You have to change the ranges if you change the reference food. For example, white bread is a popular reference food these days, but if you look at lists of GI these days, you'll sometimes find two lists, one for bread=100%, one for Glucose=100%, and different ranges. The range is not so important anyway, as "high" and "low" are relative concepts with grey areas. --Slashme 11:06, 26 October 2005 (UTC)

== Why is the link i add being removed constantly(and anonymously)? == And , on the somewhat strange grounds that calculating gi for and entire meal is not meaningful?? 1. The site doesn't do that. 2. It is meaningful and correct if you do it right. "Using this type of calculation, there is a good correlation between meal GI and the observed glycemic responses of meals of equal nutrient composition."

Also, i noticed an error in the article, the list of food that most use contains only 700-800 actual foods. Not 1500 as the "Searchable database of glycemic load values for over 1500 food items." claim. Most rows in the table are headings and groupings.