HOME
GI Blues: What's wrong with the GI Diet
In March 2005, I was reading the latest copy of a popular woman's magazine, as you do when bored on holiday. It had a whole ten pages about the GI diet – billed on the cover as 'the healthiest low-carb plan around'. And it got it completely wrong!
The authors correctly stated that GI, the Glycaemic Index, is a measure of how much carbs raise blood glucose levels. But then, in lists of high-GI, medium-GI and low-GI foods they listed fatty foods as high-GI, when, of course, their GI is zero, zip, nil, nothing – as low GI as you can get. They also listed 'omega-3 eggs' as low-GI and 'eggs' as medium-GI, when neither type of egg has a GI at all; and low-fat cottage cheese was listed as low-GI, light cream cheese as medium-GI and full-fat cheese as high-GI when, again, none has a GI. In fact, because fats, meat, fish, cheese and eggs have little effect on blood glucose, they don't have a GI – or they have a GI of 0, depending how you look at it. And lastly, diet fizzy drinks were listed as both low-GI and high-GI depending on whether or not they contained caffeine – yet caffeine is not mentioned in the GI tables, published in the July 2002 edition of the American Journal of Clinical Nutrition, pages 5-56.
This looks to me like a classic case of ignorance trying to cash in on the latest fad.
The magazine article gave a recommended reading list of six recently-published GI diet books. If the article was based on information from these books, then those authors have got it all wrong as well.
So what is GI?
The glycaemic index (GI) was originally designed as an aid for diabetics; it is a measure of how quickly carbohydrates raise blood glucose and, thus, insulin levels.
To compile this index, scientists fed 50 grams of glucose to their test subjects and measured how much this raised their subjects' blood glucose. That became their reference point; they labelled it 100. Then they tested their subjects with other foods and measured blood glucose response relative to the initial reference. If, for example, one of those foods raised their test subjects' blood glucose level to 50% percent of the reference, then it had a glycaemic index of 50, and so on. So far, so good.
But glucose is a bit too sweet for many people. The testers didn't like to drink 50 grams of the stuff and so, later, white bread was substituted. White bread has a GI of about 70 compared to glucose. The people doing the eating preferred this but unfortunately it generated another index in which bread was rated at 100.
There were now two GIs: one based on glucose = 100; the other based on white bread = 100. This started the confusion as both indexes came into general use – and many publications failed to say which one they were using (as does this magazine article).
How useful is GI?
In a nutshell, not very.
A GI of 70 or more is classed as high; 56-69 is medium; 55 and below is low. But that doesn't that tell us much. For example, one grain of sugar has a GI of 64 – and a pound of sugar is also 64. So how much sugar can you eat? There is no way of telling. But as parsnip is 97, you can obviously eat a lot more sugar than parsnip – or can you? Strangely, although the article did rightly class parsnip as 'high-GI' it didn't list sugar as medium-GI. I wonder why not?
You will be told that white bread is high-GI and that wholemeal bread is low-GI, but the difference between their GIs is only 2: white bread is 71; wholemeal is 69. Big deal. By the way, the only whole-wheat bread made in the UK which is listed in the official International GI data is one made by Ryvita Co Ltd. This has a GI of 74 – which is higher than white bread!
Another problem is that the same food, made by the same manufacturer, but in a different plant can have widely differing GIs. Take Kellogg's All-Bran, for example, which has a GI of 30 in Australia, 38 in the USA and 51 in Canada. I have no idea what the GI of Kellogg's All-Bran is in Britain as it hasn't been tested.
Then there is wholemeal flour. This can be anything between 52 and 72 in Canada, and is as high as 78 in Australia and 87 in Kenya. Again the flour in UK hasn't been tested, so we don't know what the GI of wholemeal flour here is.
And there are some strange anomalies. For example, you might think that foods containing sugar would have a higher GI than the same food made without sugar. But Banana cake, made with sugar is 47, while Banana cake, made without sugar is 55.
Then the way a food is cooked or processed also makes a difference to its final glycaemic index, according to a trial conducted at Department of Dietetics, Queen Elizabeth Hospital in Hong Kong. (1)
And there is a last problem as far as diabetics are concerned. The GI of fructose (fruit sugar) is 22, very much lower than sucrose (table sugar) at 64, yet fructose is far more damaging to a diabetic's health than sugar.(2-4)
To sum up, the Glycaemic Index is a very weak index which is over simplified, over hyped, and over sold. While it may have some use in a clinical setting, it is really of very limited use to the general public.
What matters as far as your body is concerned is not the GI of a carbohydrate, but the total amount. A hundred grams of carbohydrate is a hundred grams of carbohydrate whatever its GI is.
By the way, as I mentioned, the GI diet was billed on the magazine's cover as a 'low-carb' diet, and so it should be, of course, as all the truly low-GI foods have little if any carb in them. However, in the recipes section, under the heading 'Putting it all into practice', readers were told to eat 6 portions of carb, 5 portions of fruit and veg (which are also carbs, of course, even though they are listed separately), 2-3 servings of protein and 3 portions of low-fat diary food. In other words it's the same dreary, old 'healthy' low-calorie, low-fat, high-carb diet that has consistently failed dieters and ruined their health for more than a century.
References
1. Chan EM, Cheng WM, Tiu SC, Wong LL. Postprandial glucose response to Chinese foods in patients with type 2 diabetes. J Am Diet Assoc. 2004; 104: 1854-8.
2. Bunn HF, Higgins PJ. Reaction of monosaccharides with proteins: possible evolutionary significance. Science 1981; 213: 222-9.
3. Bierman EL. George Lyman Duff Memorial Lecture. Atherogenesis in diabetes. Arterioscler Thromb 1992; 12: 647-56.
4. Swanson JE, Laine DC, Thomas W, Bantle JP. Metabolic effects of dietary fructose in healthy subjects. Am J Clin Nutr 1992; 55: 851-6
Last updated 19 March 2005
Related Articles
