The Latest NEJM Study.

Today’s New England Journal of Medicine had a special surprise for anyone interested in diet composition and weight loss.  By now, you’ve undoubtedly read or heard one of the headlines:

Want To Lose Weight?  Just Eat Less.

It’s Not What You Eat, It’s How Much.

Stick To a Low-Cal Diet and It Will Work. (This was my personal favorite.  What an awful headline.)

You’d be forgiven for thinking that this newest study proved, once and for all, weight loss ultimately comes down to calories in – calories out.  But it contains some glaring errors that skew the results, and you savvy readers out there will probably be able to identify them right away (particularly if you read the full text of the study – get it while you can!).

Here’s a summary of the study: 811 overweight adults were assigned to one of four diets: low-carb, low-fat, high-fat, and high-protein.  They were offered group and individual instruction in the diets and followed for two years.  The researchers measured weight loss and waist circumference.

The main problem I have with the study design – the nutritional breakdown of the diets:

* Low-fat, average-protein diet made up of 20 percent fat, 15 percent protein and 65 percent carbohydrates.
* Low-fat, high-protein diet of 20 percent fat, 25 percent protein and 55 percent carbohydrates.
* High-fat, average-protein plan containing 40 percent fat, 15 percent protein and 45 percent carbohydrates.
* High-fat, high-protein diet of 40 percent fat, 25 percent protein and 35 percent carbohydrates.

You can see where they went wrong here – based on the data in the study, none of the diets contain any less than 140 grams of carbohydrates (the so-called “low-carb” group).  And if you’re a regular reader of this blog or a low-carb aficionado, you know that 140 grams of carb intake is anything but “low-carb.”  At intakes of greater than 60-80 grams of carbs daily, you really don’t get much of the benefits of a low carb diet (insulin levels still trend too high).  But to achieve those levels of carb intake, the percentages would have to be much more heavily skewed towards fat – which, for researchers touting the current nutritional status quo – is a big NO-NO.

Consider this quote from the study:

Other goals for all groups were that the diets should include 8% or less of saturated fat, at least 20 g of dietary fiber per day, and 150 mg or less of cholesterol per 1000 kcal.

At 215 mg of cholesterol, I guess eggs are out (since at 1600 calories your daily limit is 240 mg of cholesterol).  You know, cholesterol and heart disease?  No connection. Saturated fat and heart disease?  No connection.  Someone should tell these guys.

No doubt low-fat experts like Dr. Joel Fuhrman will argue that his diet (at a whopping 10% of calories from fat) is grossly misrepresented as well and as such didn’t get a fair shake.  While I generally disagree with his diet conclusions, on this point, he’d be right.

Taking a cursory look at the composition of the different diets reveals that…well, they’re not all that different.  The Supplementary Appendix shows a listing of sample menus that the study participants were encouraged to follow.  Now, you tell me; does this look like a “low-carb diet” to you?

Breakfast:  Boiled or poached egg, whole wheat bagel w/lowfat cream cheese, cup of skim milk

Lunch: Turkey, whole wheat spaghetti, vegetable medley, one small banana

Dinner: Roast beef, one potato, mixed vegetables, cabbage salad w/walnuts, tiny box of raisins

Snack: Cup of skim milk and a graham cracker

If you’re eating bagels and crackers, let’s face it; you’re not eating a low-carb diet.  But before you take my rantings as a complete dismissal of all scientific inquiry, let me qualify everything and say I understand it’s not their fault.  The researchers are operating under incorrect premises (i.e., too much fat and cholesterol is bad for you) and as such are limited by that – the whole “Do No Harm” thing.  Think of it this way:  If they thought vegetables were toxic at a dose of greater than 500 calories per day, then they’d be ethically bound to limit the groups to 500 calories’ worth of vegetables (that might make for even more awkward study design).

But that’s enough quibbling.  Let’s take a look at some of the interesting data:

• As per usual, the groups had a tough time sticking to their respective programs but did better than most study participants.  They ended up with an average weight loss of 4kg (about 9 lbs).
• The lowest carb group increased their HDL level to a greater extent than other groups; the highest-carb group fared worst.
• “All the diets except the one with the highest carbohydrate content decreased fasting serum insulin levels by 6 to 12%; the decrease was larger with the high-protein diet than with the average-protein diet.”  This means that the greater you reduce carb intake, the greater the improvement in insulin sensitivity – and we all know that’s a good thing.
• An interesting finding:  The two groups that saw the least “improvement” in LDL levels were the highest-fat and the lowest-carb group.  This is probably due to an increase in the “large, fluffy” LDL particles and not the compact LDL (prone to oxidation).  If that was over your head, read this – or, just take my word that the smaller, more compact LDL particles are what predispose you to heart disease.

The last word:  The baseline data shows that virtually all participants were eating in excess of 2000 calories and 225 grams of carbs per day – yikes.  Reducing calories from 2000 to 1600 is going to make a positive impact health-wise when weight is lost.  And I’m not saying that reducing calories (independent of diet content) won’t result in weight loss – it will.  But a well-intentioned but poorly-designed study like this one does nothing but muddy the issue.  Sustaining insulin sensitivity is important.  Controlling blood sugar and hormonal tone (so you don’t get hungry; so fat-burning occurs as opposed to fat-storing) is important.  And you can do both those things better on a true low-carb diet than on a low-fat diet.

Edit (2/27/09):  Jeff points out that “both the “high protein” and “low carbohydrate” groups had favorable changes in waist size compared to the “low protein” and “high carbohydrate” groups respectively.”  Meaning, those groups probably lost more fat (as opposed to merely “weight”) than the higher carb groups.  Nice pick-up.