View RSS Feed


Dietary Adventures: where we dissect science!

Rate this Entry
Okay, so not all scientists suck, but roll with me on this for a minute.

So I've been waiting weeks for a study to come out that I could really sink my teeth into in this blog. Any study would do since most nutrition studies are pretty bad, nutrition study reporting by news media is even worse, and I wanted an recent example to share and dissect to demonstrate the point. I wanted to do this maybe a month or two ago. Of course, because I wanted to do it, I didn't see any until last week because that's how the world seems to work. As soon as you want a bad piece of nutrition and health reporting you can't find squat. Luckily, the NIH were kind enough to oblige last week, and the Washington Post decided to report on it. I'll apologize up front because this will probably be a long on, but I promise it will be worth it.

Before I get started, I also want to give some credit to Tom Naughton at the Fat Head blog for being the first place I saw this study pop up, and doing some analysis of this article initially, and give him due credit for initially raising some of the questions I'll be covering here. I would have asked many of the same questions after years of reading up on this stuff, but I read his analysis when first hearing of the article, so it's only fair to give him credit so I don't come off as simply plagiarizing his work. So here we go (finally!).

(SPOILER)The Washington Post

Losing a small amount of weight doesn’t appear to lower the risk of heart attacks and strokes in people with diabetes who are already getting good medical care, according to a long and expensive clinical experiment whose results were announced Friday.

While modest weight loss has benefits in how overweight diabetics feel, sleep and move, whatever benefit it may confer in preventing cardiovascular disease — which is what most diabetics die from — is too small to measure, the study found.


The researchers recruited 5,145 people with Type 2 diabetes, which is the form that generally comes on in adulthood and is strongly associated with being overweight. Type 1 diabetes, an auto-immune disease, is the opposite. It comes on in childhood, and its sufferers are usually thin.

About 60 percent of the volunteers were women and slightly more than one-third were minorities. The average age was 58.

Half of the people were randomly assigned to get intensive dietary counseling on how to limit their calories. They were provided meal substitutes such as Slimfast drink. They were also urged to exercise more and instructed how to do so safely. They met weekly in support groups and once a month with a counselor, although that amount of attention was eventually reduced to a monthly visit and a monthly phone call.

The other half of the participants got less intensive diet and exercise advice, at first four times a year and later just once a year. Everyone was treated by her or his own doctor, who got an annual summary of the patient’s progress.

In the first year, the people with the intensive counseling lost 8 percent of their weight. They gained some back, but over the decade of the study maintained an average 5 percent reduction from their starting weight — about 10 pounds. The people in the less-intensive “arm” of the study lost, on average, about 1 percent of their body weight.

By the end of 11 years, there was no difference between the two groups in the rate of heart attack or stroke.
Okay, so that's in spoiler tags for obvious reasons. It's pretty lengthy. Here's the TLDR version: the study was done with type 2 diabetics. Half the participants got intensive dietary and exercise counseling over 11 years. The other half got less intensive counseling. The people in the intensive counseling group lost maintained an average weight loss of 5% or about 10 pounds. The other half maintained about 1%, or 2 pounds based on the math for the intensive group. The researchers expected to see a drop in heart attack and stroke risk for the more intensive group compared to the control. They didn't. There was no substantial (meaning a decrease not likely due to random error in this case rather than magnitude) decrease in risk between the two groups.

Now believe it or not, this is not the end of the story, despite what the scientists publishing the results, and the person who wrote the article may believe. The first big question that needs to come up is what this intensive dietary advice included? What diets were everyone eating? This is a pretty important question to ask for future studies, because maybe a different diet could have different results.

Luckily for myself and for all of you, Tom was kind enough to find the link to the study protocol for us, saving me some google searching.

(SPOILER)Look AHEAD protocol

Restriction of caloric intake is the primary method of achieving weight loss. In order to aim for a weight loss of 10% of initial weight, the calorie goals are 1200-1500 kcal/day for individuals weighing 250 lbs (114 kg) or less at baseline and 1500-1800 kcal/day for individuals who weigh more than 250 lbs. These goals can be reduced to 1000-1200 kcal/day and 1200-1500 kcal/day, respectively, if participants do not lose weight satisfactorily. These calorie levels should promote a weight loss of approximately one to two lbs/week.5

The composition of the diet is structured to enhance glycemic control and to minimize cardiovascular risk factors. The recommended diet is based on guidelines of the ADA and National Cholesterol Education program96,97 and includes a maximum of 30% of total calories from total fat, a maximum of 10% of total calories from saturated fat, and a minimum of 15% of total calories from protein.
So now we're getting into the meat of this a bit aren't we? The diet was 15% protein, 30% fat, leaving us with 55% of the diet being carbohydrates. That is a high carb diet if ever I saw one. So what did this study really demonstrate? Was it that weight loss doesn't affect heart attack or stroke risk, or is it that losing weight via caloric restriction on a high carb diet doesn't affect risk? It's really the latter, but this gets lost in the actual article. What would happen if the participants were losing weight on a low carb, high protein diet? Or a low carb, high fat diet? Or other combinations of macro nutrients? We can't say based on this study, but it's something worth pursuing.

Based on what we know about fat storage and insulin from my previous posts though, my bet would be that participants would have lost more weight on a low carb diet. Their risk of heart disease and stroke probably would have dropped as well, but we can save the reasons why for a future post I want to do specifically on heart disease.

But aside from simply debunking that weight loss on a high carb diet has any effect on heart attack risk, the study also works to debunk something else, almost unintentionally. Take a second to read the quoted sections above and try to figure it out before moving on. No shame in not figuring it out though because it's not immediately obvious, particularly if you've been told the wrong information your whole life as many of us are.


So here it is. In the dietary protocol above, calories were restricted to a level which the researchers believed, based on the caloric deficit, would promote a one to two pound per week weight loss. Do the math there and that should be about 50-100 lbs./year. But how much did participants actually lose on average? Remember, the people receiving the more intensive coaching and education maintained about 10 lbs. of weight loss over the 11 year study. The people receiving less intensive coaching maintained maybe 2 lbs.

Now there are two possible explanations which immediately spring to mind that could explain this. The first is that out of these more than 5,000 people who had the dedication and willpower to stick with this study for 11 years, many of them cheated on the protocol and lost little to nothing, or perhaps even gained, dragging the average down so low. I'll have to have a look around for the actual published results if they're freely available because the distribution of weight loss would help shine some light on this question, but realistically, I doubt that people ignored the study protocols in sufficient numbers, and to such a degree as to result in such underwhelming weight loss averages if they had the wherewithal to stick with a study for 11 years. If I do manage to track down the results later I'll edit them in.

The other option, which I'm sure anyone who was paying attention to my previous posts about fat storage and insulin, and the role carbohydrates play in the process might have picked up on, is that the old notion of weight loss being a matter of calories in minus calories out is bulltrout. These participants reduced calories to a degree which that hypothesis says should have resulted in them all being quite lean within a year or two. But they barely lost anything, and the reason is likely due to the amount of carbs they were eating. But again, this is the sort of stuff which should be investigated on the same level and scale as this NIH study. These are questions these researchers should be asking themselves as they move forward so they can design future research appropriately. Who knows if they actually will though.

Finally, I'll end with a quote from the article which demonstrates better than probably any other part of the article why I hate the present state of nutrition science and health reporting. I'll even let Mr. Naughton cut right to the heart of how terrible the following statement is:

Quote Originally Posted by Washington Post
The results will probably surprise many physicians and patients but are not likely to change the advice they give and get.
Quote Originally Posted by Tom Naughton
Well, of course not. You wouldn’t want failure to inspire a change in your beliefs, much less in your strategy.

Updated 10-29-2012 at 04:13 PM by Slothy

Tags: None Add / Edit Tags