December 17, 2008

More Bad Science: The "Low Glycemic" Diet in the JAMA Study is the ADA Low Fat Diet

The media are abuzz with the results of a diet study published in JAMA which supposedly "proves" that the Low Glycemic diet has benefits for people with diabetes.

It doesn't, but to see why, you have to download the entire, free, PDF, which includes the damning details that make it clear that the diet used in this study was not by any but the most twisted logic a "low glycemic diet."

To see these details and read what the participants ate in this diet study, click on the link below and then click on the PDF link to the right.

Effect of a Low–Glycemic Index or a High–Cereal Fiber Diet on Type 2 Diabetes: A Randomized Trial JAMA. 2008;300(23):2742-2753.

Despite the media spin, what this study actually did was attempt to answer a stupid question: Is a high carb diet/low fat made up predominantly of pasta, beans, carrots, pears, oranges and skim milk better for people with diabetes than a high carb/low fat diet made up of Wheetabix, potatoes, mangoes, skim milk and toast.

The answer turns out to be "yes," very slightly. This is one step away from asking if a diet made up of pasta and fruit is better than one made of hot fudge sundaes and soda. It is, but neither is a diet you would feed to someone with diabetes if you cared about their long term health.

Neither diet in the JAMA study lowered the blood sugar of the study subjects to the range that avoids complications. The average fasting glucose in the group eating the supposedly "low glycemic" diet ended up well over 130 mg/dl, and this was in a group of people taking Avandia, Actos, Metformin and sulfonylurea drugs.

With fasting blood sugars that high, the post-meal blood sugars of those eating either diet had to be spiking well over the 140 mg/dl level at which complications begin to occur and spending lots of time over the 200 mg/dl level where they become dangerous.

That the high carb/low fat diet raises blood sugars is not news. Nor is it news that it makes very little improvement in lipids compared to a lower carbohydrate diet. But what had me really scratching my head after reading this study was the definition of "low glycemic" used to craft the supposedly "low glycemic" diet used in this study. Because that diet was filled with stuff like skim milk, Carrot coins, Oranges, and tomato sauce, which even the most dedicated follower of the Glycemic Index would find baffling.

After all, doctors suggest people use orange juice to treat hypos because orange juice is one of the fastest carbs available and though oranges have fiber, any time I eat one, I see it hit my bloodstream within an hour and when it does, all 20 grams of carbohydrate make themselves known.

More "low glycemic" diet foods fed people on the low glycemic diet here were "low fat yogurt" and pears. Also jam. Also Balsamic vinaigrette salad dressing. I have eaten all these foods and tested afterwards I can tell you they will all hit my blood sugar with full force within one hour after eating.

Add to these super fast carbs the slower to digest but very high carb foods that fill out the diet, the rye pita, lentils and the full cup of spaghetti, and you have a diet that should keep blood sugars elevated for all day long without a break. No wonder that fasting blood sugars were well over 130 mg/dl despite a full load of meds. You can only imagine what blood sugars would have looked like without meds, too!

This is just plain nuts, people. This isn't a low glycemic index diet. This is the classic ADA low fat diet hiding behind a new name.

But it isn't 1998 anymore and the health claims for the low fat diet have been completely discredited. We now have a lot of high quality research showing that low carbohydrate diets with adequate protein and lots of healthy fat lower blood sugars much better than this tired old, hunger-provoking, misery inducing ADA low fat diet.

In fact, as little as I like the so-called "glycemic index" which is based on non-reproducible averages of measurements of the blood sugar of normal people and has little applicability to anyone with Diabetes who has lost part of their second phase insulin release, I have to admit that even a true low glycemic diet--the so-called Mediterranean diet, which is a medium fat/slow carb diet, will do a much better job than this one does.

The fact is, this really is not a low glycemic diet, it is low glycemic only if you you compare it with a fudge sundae and soda diet which may well have been what the subjects in the study were eating before they were recruited. It is also a high fructose diet, which is bad news for anyone who believes that step one in dealing with Type 2 diabetes is to reduce insulin resistance.

Fructose was believed to be good for people with diabetes a decade ago and the ADA was slow to stop recommending fructose to people with diabetes. But now we know that eating fructose contributes to fatty liver and increases insulin resistance.

You can read a brand new discussion of new research which looks at how eating fructose changes the way that genes in the liver express in this week's edition of the Diabetes in Control newsletter.

To those who say, but what about the findings that the diet the study promoted lowered cholesterol and A1c, there is only one answer: go look at the actual data. The changes in cholesterol were "significant" only in the statistical meaning of the word. The actual change in the lipid measurements was small and triglyceride level, which are more closely linked to heart risk than LDL did not change significantly.

The changes in body weight and A1c compared to baseline are slightly better, but are probably due to the diet being used here being one that severely restricted food choices, urged people to lower calories, and told people not to eat any of the junk food that adds calories to the Standard Diet. A lower calorie very high carbohydrate diet will have less carbs than a high calorie very high carbohydrate diet and that will reduce A1c, though in this case, not enough to provide lasting health benefits.

Pit this diet against a low carb diet of any kind and you'll see how pitiful this diet really was. For that matter, pit it against a truly "low glycemic" diet--one that also has a low glycemic load and you will similarly see the limitations of this antique, failed ADA low fat diet exposed.

You don't have to take my word for it. These studies have already been done. You can find them listed here:

Studies Proving the Safety and Efficacy of the Low Carb Diet

But you don't need to read studies to find out what is a safe and effective diet for your own, unique body. Get out your blood sugar meter and test after eating. If the foods you eat keep your blood sugar under 140 mg/dl, they're good for you. If they raise your blood sugar higher than that, they aren't. And if you really want to be healthy, shoot for a lower post meal blood sugar target. Many people stay under 120 mg/dl, which appears to be the true upper end of the normal range for post meal blood sugars. Some go even lower.

And write a letter to the publishers of JAMA asking why they gave valuable page space to a study that is so obviously flawed.


renegadediabetic said...

The spinmeisters are at it. Reminds me of the much ballyhooed vegan diet for diabetes. Sure, the vegan diet did better than the ADA diet, but neither diet lowered A1C below 7%.

At least debunks the notion that a high fiber diet is good for diabetics. I want to slap someone every time I hear that nonsense. If I ate the "Low GI" diet described in the article, my BG would be higher too.

Unknown said...

This drives me insane...just today a well intentioned person sent me a bunch of sugarfree recipees for diabetics...all extremely high carb. I am so weary of the food myths and diabetes...and the more I try to explain, the weirder people I know think I's useless trying to explain to nondiabetics and even to some diabetics if studies like this abound. Thanks Jenny.

Anonymous said...

This diet *is* very similar to what I ate for years and is the reason I ended up here. The only good thing about it is that it was not so rapidly and immediately damaging as the low fat even higher carb version I was put on by the dietician which wrecked my BG and made my already crap lipids worse.

The ADA claims Medical Nutrition Therapy can improve A1c by 1 - 2% in their Position Statement while on their very own forums we see people routinely achieving 5 - 8% and sometimes over 10% improvements by testing and eating appropriately low levels of carbs.

Papers like this are like rearranging the deckchairs on the Titanic.

Anne said...

You may want to include the new article in NEJM titled "Glucose Control and Vascular Complications in Veterans with Type 2 Diabetes" in the "More Bad Science" column.

This study concludes that tight blood sugar control does not lessen vascular complications. Of course the subjects were on multiple medications to for diabetes. All were on statins and the American Diabetic Association recommendations were used for diet. All the patients were on rosiglitazone, which is known to be associated with increased cardiovascular problems. Perhaps that is why there was no difference in the two groups???

When are we going to get the study that looks at tight control achieved with a truly low carb diet and meds used to cover only what the diet can't control?

I look forward to hearing your comments about this study. You can access the full paper at

Jenny said...

I am well aware of the Veterans study and very concerned about how its findings will be misused.

I plan to blog about it after the holidays. This time of year traffic to all health sites is way down.
At the beginning of January, traffic always picks up again. So that would be a better time to highlight it.

theoddbod said...

what a strange study, especially since they were all on a variety of drugs, might that not alter their results even further regarding the effects of the diets on blood sugar control?

Anonymous said...

Here is what we need to see more of:

And for kids a low carb diet is even more important.

Fred Hahn

Las said...

It's gotten to the point that I don't bother reading any studies on nutrition. Reading the ridiculous conclusions 'scientists' come to from their research makes me batty.

I just found this site...such a great source on information. I was wondering if you had any information about Type 1 diabetes. My brother-in-law has it, and while I've gently brought up low carbing to him and my sister I'd like to offer them some good research to back it up. His main problem, really, is his hypoglycemic episodes. This happens very often, at least once a week (happened on his wedding day!) and usually in the middle of the night or early morning. My sister is a nervous wreck over this, and since he has to inject insulin I'm not sure how a low-carb diet would work with that. I'm sure he has to lower his dosage, because these hypo episodes happen way too often. He's actually considering getting the pump so that his BG will be measured automatically and the right amount of insulin will be administered.

Jenny said...


I don't give advice to people with Type 1 beyond suggesting that they read the book, Dr. Bernstein's Diabetes Solution. There are some very helpful ideas in that book and even if a person does not adopts Dr. B's extremely low carb diet, reading the book will teach how to avoid hypo and think about insulin dosing.

I'd also suggest a visit to which has many Type 1s participating, many of whom are in excellent control and can help with fine tuning.

Anonymous said...

Another good site for Type 1s is

the How Tos are written by a doctor with a Type 1 son and AFAICR cover reducing carbs while on insulin pretty well