The crooked house of GI

Krzywy Domek, Sopot, Poland

GI is in. Considering its young age, it has managed to get ingrained into our subconsciousness much like “fat makes you fat” and “it takes 5 seconds to contaminate a food particle once it has fallen on the ground”. But what exactly is it?

According to the official glycemic index website, GI (glycemic index) measures the extent to which carbohydrate foods raise blood sugar level (BSL from now). Glucose was nominated to have a GI of a 100. High GI (>70) means that this particular food raises BSL fast, giving you a “glucose spike”. Anybody with a brain knows that spikes are not good: they are nasty sharp things. This is the beauty of language: you can use a word that evokes vivid images in your head, helping you make an immediate judgement. But I digress. Low GI (55 and under) is supposedly better because these foods result in slow sustained release of blood glucose (mind picture: a calm and steady flowing river). Low GI was initially recommended to diabetics, then the advice was expanded for weight loss, followed by the publication of a very successful book “The New Glucose Revolution” . Now food manufacturers fall over themselves to have a Low GI symbol on their product (for a fee, of course)

I have quite a few problems with recommending a Low GI diet, as advocated by Jennie Brand-Miller, the author of the book.

The initial studies showed a great variation in glucose response to common foods. For example carrots were shown to have a glycemic index of 92 in the first study, then 37 in a later study. This fact becomes a little less surprising when you find out that the testing groups consisted of 5-10 people (!). 

The studies measured the effect of carbohydrate foods alone. Even the authors of the book concede that we do not eat “carbohydrates” for dinner, we have some spag bol. Other components of the meal affect the glycemic index by regulating the rate of absorption. Protein, fat and fiber all slow down the Glucose release effectively lowering the GI. This is why the GI of super supreme pizza is lower than the GI of watermelon: pizza is high in fat and protein, watermelon is pure carb. In addition, the composition of your previous meal also affects GI, this is why the studies were done after an overnight fast.

One of the major flaws of the diet, in my opinion, is that it makes two assumptions: 1. blood glucose levels are easily compared between individuals 2. insulin release predictably follows the rise in blood glucose. Even between healthy individuals (free from diabetes type I or II), the variation in the amount of insulin released after the same meal can be enormous. However this diet is recommended to DIABETICS. Their glucose metabolism is screwed up already. Type II diabetics and those with pre-diabetes have a higher BSL reading and therefore more insulin required after consuming the same meal as a healthy person. Let’s say I have a bowl of “heart-healthy” muesli (you will have to pay me a lot of money first!) and measure my blood glucose after 1 hour. My fasting BSL is around 4.2, my BSL after the cereal will be lucky to hit 5. Type II diabetics will start with a higher fasting BSL and their response to muesli might range anywhere between 11-20mmol. 

Can this get any more confusing? A bowl of All-Bran cereal bought in Australia (caveat 1), weighing at 30 g (2) has a GI of 30 if eaten in isolation (3), by a healthy individual (4), on an empty stomach (5), for breakfast (6). If you happen to be Canadian, your All-Bran has a GI of 51. There is no way of telling what the GI would be if you change any of the other 5 variables.

GI supporters are probably shaking their heads at me. I must misunderstand the principles of low GI eating. People are just supposed to swap their unhealthy high GI meals for healthier low GI version. Like ditching their Cornflakes (77 in Oz) for Frosties (55)?  Yay! I am liking my new diet!

The next line of defence for the GI crowd is Glycemic Load, which now takes into account the quantity of carbohydrate in your meal (something you would like to know if you are diabetic and about to jab yourself with insulin). OK, let’s see how that works.
Glycemic load values for 150g of potato. (Low<10, Medium 11-19, High >20)
Potato, non-specified, boiled, Australia – 11
Potato, sebago, peeled, boiled for 35 min, Australia – 14
Potato, Pontiac, peeled, boiled for 30 min, Australia – 14
Potato, Pontiac, peeled, boiled for 35 min, Australia – 16
Potato, Desiree, peeled, boiled for 35 min – 17
Potato, boiled in salt water, India – 26

 WTF? (Why The Face, for the fans of the Modern Family)

 I think that the problem with GI  is that it is based on a faulty concept. Like a building sitting on a shaky foundation, you have to

Does this count as GREEN?

keep tinkering with a top structure just to keep it standing. Now the ceiling caves in, now the top story goes wonky. It is based on a premise that we have to eat 55%+ of our energy from carbs. The foundation is now laid and suddenly you realize that carbs raise blood sugar (not good for diabetics) and insulin (not good for weight loss). You have no choice but to play the cards you are given! So we are told to make the carbs low GI in an attempt to minimize the damage. Is eating vegetables (low GI) better than eating a muffin (medium GI)? Yes! Does it have anything to do with GI?  

I can make the same argument by telling you to only eat GREEN food. You will probably eat healthier, stabilise your blood sugar and lose weight. Does it mean that the colour GREEN has unique properties? Nope, it’s just most green foods happen to be vegetables, with an exception of an occasional peppermint candy.


How about taking down this crooked building of GI and starting with a new base? If you are a diabetic or weight loss is your goal, maybe eating so many carbs is not a great idea (a radical concept, I know). Do you know what has a GI of zero? Eggs, beef, lamb, pork, chicken. Most green vegetables have a GI<30. That’s what I call a true low GI diet!


13 thoughts on “The crooked house of GI

  1. Ok, I think I followed along with that. In short, eat fewer carbs. That’s EXACTLY how I lost weight, though what carbs I did get I get mainly from fruit & veg, of which I eat A LOT.

    • Sally, you have certainly done very well, not just in term of weight loss but also metabolic syndrome and insulin resistance. Basically, changing to low GI carbs is doing something “less bad”, a bit like smoking low nicotine cigarettes. Glad to see you here.

  2. Very very interesting read. It’s funny how I first learnt about the GI theory… via Les Mills!

    You may possibly remember from around the quarter of Jam 35/36 (circa 2005), there was an education session in all programs about which GI foods provided you with the most fuel to last through classes, and the consensus in that session was that Low GI foods consumed around 1-2 hours before class would help you endure and decrease the chances of the “spike”->lethargy feeling.

    I always get confused about what’s high and low GI and this post has helped somewhat to get me across the theory behind it!

    • Yes I do remember that education session! Les Mills and the fitness industry as a whole has come a long way in their attitude to diet. Check out their new food pyramid! It bears no resemblance to the conventional one pushed by the government: not a single “whole grain” in sight. While they won’t publicly admit it, they are definitely closer to paleo and lower-carb than ever.

  3. Anastasia , I would love you to post the topic we discussed that day when you drew me a little picture and explained about the glucose & fat and sugar already in our blood stream…if you follow.

  4. I think i’ve heard robb wolf say that its not true that eating protein, fat, and fiber will lower the glycemic index of a food (i.e. adding nuts to fruit). any additional thoughts on this?

    • I’m not sure whether this is what Robb Wolf meant. Don’t forget that GI is just the measure of how much glucose is detected in the blood after 1,2 hours. It is widely known that protein, fat and fibre reduce the blood glucose level, aka GI, when indigested together. But the total glucose is still the same and the pancreas still release the same amount of insulin (even higher insulin with the protein-carb meal). Here is a nice review explaining this

        • Oops my bad review here. Yes, that’s right. In this sense having just “safe” low GI carbs may not spike the blood glucose in a HEALTHY person but you still have the problem of hyperinsulinemia. Bad for weight loss, cancer and a bunch of other things. In diabetics all bets are off. They react to the weirdest things. I’ve seen a woman get sky-high on brown rice and stay low on orange juice. It’s funny because the GI system was designed for diabetics but it is only mildly applicable for non-diabetics.

  5. Pingback: What the @#$! do I feed my child? | primalmeded

Comments are closed.