Friday, December 14, 2007

Glycemic Index Doesn't Matter Much in Overall High-Carb Diet

Ever wonder what would happen if you place 19 obese, hyperinsulinemic women on a diet where calories aren't restricted, but glycemic index is tweaked a bit?

The short answer - not much when you look for differences between diets that are high or low glycemic index.

But y'all know I'm not about the short answer already, so how about we take a look at a newly published study, No effect of a diet with a reduced glycaemic index on satiety, energy intake and body weight in overweight and obese women, that has some very interesting findings that were notable but ignored in the paper.

Make a note of the paper title, it's important later!

Let's start with some background. For this study, researchers recruited otherwise healthy, obese women who had hyperinsulinemia. They were free of underlying conditions like diabetes, were not taking medications for cholesterol, hypertension or other medical conditions, were not pregnant or breastfeeding, and had not recently dieted to lose weight, nor recently lost or gained weight, thus deemed "weight stable" to participate.

The researchers then determined their habitual food intake and calories consumed each day, noting the subjects consumed an average of 1859-calories/day - 48.8% carbohydrate, 16.7% protein and 34.4% fat [227g carb, 77.6g protein, 71g fat].

The study was then underway, with the women consuming one of two diets for 12-weeks in a cross-over design, so the study lasted a total of 24-weeks. The objective of the study was to "investigate whether a diet with a reduced glycaemic index (GI) has effects on appetite, energy intake, body weight and composition in overweight and obese female subjects," so the researchers provided cereals, breads, rice, pasta and potatoes with different glycemic index for the women to consume while following either the low or the high glycemic diet.

The women were told to consume their foods as they did in their habitual diet and continue eating how they usually ate before starting the study.

In the paper, the researchers noted that "[t]here were no differences in energy intake, body weight or body composition between treatments. On laboratory investigation days, there were no differences in subjective ratings of hunger or fullness, or in energy intake at the snack or lunch meal."

They concluded, "This study provides no evidence to support an effect of a reduced GI diet on satiety, energy intake or body weight in overweight/obese women. Claims that the GI of the diet per se may have specific effects on body weight may therefore be misleading."

Yesterday, Sandy Szwarc at Junkfood Science blog penned her article Carbs humbug? — Are carbs really fattening? about the study, with her perspective being that this is one more piece of evidence that carbohydrates are not fattening.

She wrote, "Despite oft-repeated fears that refined “bad” carbs send our glucose and insulin levels soaring, these researchers found “no differences in glucose, insulin and non-esterfied fatty acid responses to the lower versus higher GI breakfasts, with no differences in either total area under the curve or single measurements at any time point.” Concerns that carbohydrates in the diet stimulate insulin production and are responsible for obesity and illness were recently examined here."

I point to the article at the Junkfood Science blog because it's a good example of seeing the forest while missing the trees.

Let's look at the data in the study now, comparing the dietary changes with their baseline starting values...with the baseline number first, followed by the measured effect of each diet (values are rounded up or down where appropriate).

Low-Glycemic Index Diet, 12-weeks
Baseline - Low-GI Diet
Weight: 87.5kg - 89.1kg
Waist: 103cm - 105cm
Fat Mass: 42.3kg - 44.54kg

Calories: 1859/day - 1928/day
Carbohydrate: 227g/day - 248g/day
Protein: 78g/day - 82g/day
Fat: 71g/day - 69g/day

High-Glycemic Index Diet, 12-weeks
Baseline - High-GI Diet
Weight: 87.5kg - 89.2kg
Waist: 103cm - 106cm
Fat Mass: 42.3kg - 42.9kg

Calories: 1859/day - 1874/day
Carbohydrate: 227g/day - 223g/day
Protein: 78g/day - 83g/day
Fat: 71g/day - 71g/day

The researchers didn't spend much time in the paper discussing the weight gain, fat mass increase, or change in waist circumference. In fact, they noted "[t]here were no differences in body weight, waist circumference or fatness between intervention periods. Weight increased during both intervention periods, although weight gain did not differ between treatments."

They didn't discuss the differences from baseline, nor seek to understand how it is that hyperinsulinemic women, who are obese, gained both weight and body fat while consuming diets that were not that much higher in calories than baseline - certainly not high enough to theoretically gain as much as they did!

But more importantly - look at the increase in waist circumference - more than one inch (3cm) is not something to sneeze at in just 3-months, nor is the scary gain of 4.9-pounds of body fat while consuming the low-GI diet for 12-weeks!

Why did that happen?

The researchers didn't discuss the possibility that high insulin played a role or that blood glucose from carbohydrates consumed had anything to do with it.

Perhaps it happened because insulin and blood glucose didn't budge?

The paper does not include baseline values of insulin and glucose in the table to compare it to each diet, but does provide a telling look at how dietary glycemic index, in the context of a high carbohydrate diet, when compared with each other had no effect on either:

And in case you missed it, what this showed, quite nicely I'll add is that 12-weeks of a high carb diet - whether the perceived junky high-GI or "healthy" low-GI carbs - can make you gain weight, body fat and see your waist grow bigger too...while doing absolutely nothing to resolve a state of hyperinsulinemia!

But yeah, the brilliant conclusion was that there was no difference between low-GI and high-GI...let's not confuse ourselves with the more important findings here!


  1. Great analysis, as usual. I noticed the JunkFood Science blog doesn't allow comments. Never a good sign when you don't want feedback.

    Here's the question I'd like to ask Sandy, if anybody knows how to contact her (or if maybe she reads this blog): the biochemistry of fat storage is well-understood (learned all of this from Gary Taubes' excellent lecture at UC Berkeley). Fat cells make alpha glycerol phosphate in order to bind free fatty acids into triglycerides, the form in which fat is stored. Alpha glycerol phosphate made from glucose. The only way to get a large amount of glucose into the fat cells is to raise insulin.

    Thus, the only way the fat cells can store a signficant amount of fat is if you consume an associated amount of glucose to a) provide the raw material for alpha glycerol phosphate, and b) to raise insulin enough to transport that glucose into the fat cells.

    So the question for Sandy(and others of her persuasion) is how can carbohydrates be anything but fattening? Or conversely, how can you possible gain fat if you don't eat any carbohydrates?

    I love this little piece of biochem, as it provides the molecular basis for what we low-carbers already know: eating fat does not make you fat. It makes you full.

  2. I'd hardly consider an 8.4 decrease in G-Index a significant reduction in G-Load. Not to mention Both GL's are still in the High Range 125.9 vs 145.

  3. Thanks for the explanation. When I read the post at Junkfood Science, I figured that both groups were eating a high carb diet, and that was why there was no weight loss, regardless of the glycemic index. Your explanation confirmed that.

  4. On this one it is important to note the women were not attempting to restrict calories or lose weight. The study was designed to see the effects of a maintenance diet with varying glycemic index foods...and I think the results speak to the effect!

  5. I'd hardly consider an 8.4 decrease in G-Index a significant reduction in G-Load. Not to mention Both GL's are still in the High Range 125.9 vs 145.

    I agree, but will add that I think this one was a good showing of what happens when you replace foods you currently eat (as we're repeatedly told is the way to go) with foods promoted as being "healthier" because they're lower for GI but are still, as we know, high carbohydrate (think whole grains).

  6. doctor bear10:53 AM

    Lies, damn lies, and statistics.

    These "scientists" are so desperate to prove their point that they lose all credibility.

    How could they avoid commenting on the overall weight gain of all the subjects on this calorie-modest diet? Good pick-up, Regina. Pity the researcher who tries to pull a fast one with a misleading abstract conclusion!! Regina will root you out!!! Beware the Weight of the Evidence!!!!!

    They would rearrange the deck chairs on the Titanic....pathetic fools.

  7. Great analysis, as usual. I noticed the JunkFood Science blog doesn't allow comments. Never a good sign when you don't want feedback.

    Agree on both counts.

    I read the Junkfood post's conclusions about how we can all eat sugar and flour and not get fat or diabetes. All I could do was laugh - I keep thinking, have you ever tried a low carb diet? Have you ever talked to anyone who has?

    So the question for Sandy(and others of her persuasion) is how can carbohydrates be anything but fattening?

    It's amazing, isn't it, how the basic biochemistry is ignored.

    Someday I want a job where I can research the effects of the law of gravity and publish repeated studies that "show" that balls thrown up in the air do not comeback down or if they do, it's with varying speed because of their mass. *grin*

    I wonder how long it will take review committees to catch up and start rejecting studies that ignore bio chem mechanisms. Do they even know the theories themselves?

  8. Great job picking that one apart, and thanks for going and getting the details for us!

    Interesting to think that even if already overweight, if you ate about 1800 calories a day of "healthy whole grains" you would be constantly getting fatter. Hmmn, and they wonder why obesity is an issue.

    But really, the important thing to take from this, sez the media, is that we might as well eat MORE junky carbs 'cause it's not much diff than the so-called good stuff!

  9. Anonymous11:24 AM

    You know how they use the ethical argument to avoid testing low-carb sometimes.... (because it would be unethical for people to eat such a high percent of fat!!! OMG!!!)

    How come putting obese women on a diet that makes them gain weight, is not unethical?

    Not one word about that?

  10. Anonymous10:03 AM

    I see the difference in carbs from low to high GI was only 25g. This is like 1 piece of candy. What a waste of time and money this test was.

  11. I am a diabetic and I have been tracking/ charting BG vs what I eat for a year. One of the things that was obvious in measuring BG after meals, is the fallacy of low GI <55 recommndations. It seemed to make no difference ... the graphs in your blog show why. The only kind of carb that truelly works for BG control for a diabetic (or fat accumulation for the rest)is ultra-low GI <15 (i.e. veg, ...) and zero GI (protien and fat!).

    All these studies, it would seem are still trying to keep the 200plus gms of carbs recommendations alive despite all the evidence to the contrary.

  12. Chris Repka, MS12:04 PM

    While there are certainly some interesting TRENDS associated with GI, if there was not statistical difference, you CANNOT say that there was a difference. BOTH groups increased overall waist circumference and weight (non-significant increase), which may reflect one of any number of things, including time of day, time of year (holiday time?) or the unusual circumstance under which they were eating breakfast(monitored in a laboratory setting). Additionally, if you took a look at the error bars on the graph, there was a significant amount of inter-individual variation. Saying that these researchers used statistics to mislead in this study is obtuse.

    As far as the issue on ethics, most institutional review boards like to see some sort of weight loss program (diet exercise counseling) for the subjects following the completion of the study, if so desired by the participant.

    Finally, the poster who is aghast at the realization that there was only a 25g difference in CHO has missed the point. They were investigating the effect of a diet's (specifically, the breakfast meal's) glycemic index, not the effect of varying the relative amount of different macronutrients.