I’ve written about a 2009 New England Journal of Medicine article comparing weight-loss diets of various macronutrient (fat, protein, carbohydrate) composition. Its conclusion: Cut back on calories and you will lose weight, regardless of macrontrient percentages.
A blurry low-carb high-fat breakfast
A blog reader, Matt, brought up some interesting comments and questions. What follows will make little sense unless you read that prior post.
Matt writes:
Dr. Parker,
If the study folks didn’t do the real low carb diet because they “knew” that ketosis wouldn’t occur, couldn’t they at least have tried it, since what they were trying to prove was a calorie is a calorie?
Looking at the menus, the diet that they are purporting as low carb is really nothing close to a real low carb diet. It is a slightly lower carb diet, and not high enough in fat to prove anything. 35% carb is not Atkins phase anything. For a participant consuming 1600 calories, that’s 140g carb — too high for anyone attempting to restrict carbohydrates for health.
Please comment on the fact that the highest carb diet provided the worst lipid improvement.
Following up a little more, there really is no inference whatsoever that can be made with regard to a low carb diet with this study. Did you read the sample menu? No low carb diet phase would have any of the following as a typical meal. You can tell by looking at the menus that they had to be really PC about a “high fat” diet as well. I mean skim milk on a low carb / high fat diet? Note my level of surprise by the ? and ! in the parens with each “typical meal” option:
Breakfast:
1 poached egg
1/2 bagel (??)
4 oz apple juice (????!!!!)
skim (????) milk
Lunch:
1/2 cup spaghetti (??!!)
1/2 cup squash
1/2 cup peppers
1/2 cup mushrooms
1.5 T Olive Oil
1 small banana (????)
Dinner:
2 oz beef
1 small potato (????!!!!)
3/4 mixed veggies/legumes corn/carrots/lima/peas/green beans (???? since these are among the higher carb veggie choices)
1/2 cup cabbage
1 mini box raisins (??)
1 small apple (?????)
4 t Olive Oil
7 walnut halves
Snack
Skim (???) Milk
1 Graham cracker sheet (??????)
If you want a LC diet with what LC would consider a higher level of carbs (~60g) you need to do this:
Breakfast:
2-4 poached eggs
2 T olive oil
spinach
1 cup whole milk
Lunch:
1/2 cup squash
1/2 cup peppers
1/2 cup mushrooms
2 T Olive Oil
4-6 oz fish
Dinner:
4-6 oz beef
3/4 mixed lower carb (cruciferous/leafy) veggies such as broccoli, collards or other greens,
1/2 cup cabbage
2 T Olive Oil
20 walnut halves
1/2 cup low carb fruit such as cantaloupe
Snack
1/2 cup strawberries
1 cup whole milk yogurt ot cottage cheese
*******************************************
My response:
Thanks for your thoughtful comments/questions, Matt.
You’re right: The “low-carb” diet they studied indeed was not very low-carb, as succinctly illustrated by the sample menu you provided. (I didn’t read the supplementary appendix myself.)
You mention that the “highest carb diet provided the worst lipid improvement.” It’s not that clear-cut.
(Lipid changes are on pages 865-7 of the article, for anyone following along. Conventional wisdom is that better cardiovascular health is associated, generally, with lower total cholesterols, higher HDL chol, lower total LDL chol, and lower triglycerides.)
The study had two low-fat diets, with either 55 or 65% of total calories derived from carbohydrates. The two high fat diets had either 35 or 45% of total calories from carbohydrates.
Total cholesterol levels dropped by about 3 mg/dl in the low-fat diets compared to “no change” in the high-fat diets (2-year values). Measured at 6 months, total chol levels were down by about 5.5 mg/dl in the low-fat groups, and about 3 mg/dl in the high-fat groups. Baseline total chol levels for the whole group averaged 202 mg/dl.
The authors on page 865 write:
All the diets reduced risk factors for cardiovascular disease and diabetes at 6 months and 2 years. At 2 years, the two low-fat diets and the highest-carbohydrate diet decreased low-density lipoprotein [LDL] cholesterol levels more than did the high-fat diets or lowest-carbohydrate diet.
The lowest-carb diet increased HDL chol more than the highest carb diet, but we’re only talking about a 2 mg/dl difference measured at 2 years. HDL rose in all groups. Average baseline HDL level for the entire study group was 49 mg/dl.
All diets decreased triglycerides similarly, by 12-17%.
The magnitude of these changes is not great, and I question whether clinically important. The take-home point for me is that low-carb eating may not be (and probably isn’t) as atherogenic as warned by the medical community 15-20 years ago, judging purely from lipid changes. Other studies found similar numbers. But we’ve already agreed the this was not a serious trial of low-carb dieting.
The study authors write that HDL chol is a biomarker for carbohydrate intake: reducing dietary carbs tends to increase HDL chol levels, and vice versa.
If I understand “Nutrient Intake per Day” in Table 2 correctly, the participants who were told to increase their percentage of calories from fat really didn’t do it: they reduced it by 3.5% (!?). The low-fat cohorts had more success with compliance.
Clearly, it’s quite difficult to get free-living people to change their macronutrient intake and sustain the change for even six months, much less two years. Would compliance have been better if subjects had been allowed to choose a diet according to their natural inclinations? Maybe.
A recent study suggests that eating low-carb helps with prevention of weight regain because it burns an extra 300 calories a day compared to those eating low-fat. Dr. Barbara Berkeley took a close look at this research on June 30.
Steve Parker, M.D.
Here’s the traditional healthy Mediterranean diet.
Advanced Mediterranean Diet 