Jul 17 2008

New Diet Study Compares Low-Carb with Low-Fat

Published by under Uncategorized
Comments: 14

A new study just published in the New England Journal of Medicine followed three diets for two years: a low-fat diet, a low-carb diet, and a so-called Mediterranean diet which is moderately low fat and replaces red meat with poultry, fish, and nuts. The result are being touted in the media as vindication for the Atkins style low-carb diet, but a careful look at the study tells a more complex story.

Weight loss

There are different ways to assess the healthfulness of a diet. Most Americans are primarily interested in weight loss, but there is also heart-health, reduction in risk for DM, and overall nutrition to consider as well. But let’s start with weight loss.

The low-carbohydrate (sugars and starches) diet is promoted primarily as a weight loss strategy.  Proponents sometimes claim or infer that you can lose weight with a low-carb diet without reducing total calories – or, at any level of caloric intake, you will lose more weight. This has never been established and this new study does not establish it either. The real (and more plausible) question is whether or not a low-carb diet or a low-fat diet helps dieters achieve and maintain lower caloric intake to aid in weight loss.

This study looked at three groups of employees at an Israeli research center.  They were given lunch according to which diet they were on (lunch is the biggest meal of the day in Israel) and then were counseled how to eat for breakfast and dinner. The study reports that exercise was similar among the groups. The low-fat diet restricted fat to 30% and total calories to 1500 per day for women and 1800 per day for men. The Mediterranean diet restricted fat to 35% with the same calorie restrictions and replaced red meat with poultry and fish and included things like nuts and olive oil. The low-carb diet restricted carbohydrates to 20g per day for a two month induction period then 120g per day after that, without other restrictions.

The results indicate that the low-fat dieters after two years lost an average of 2.9kg (6.5 lbs), Mediterranean dieters 4.4 kg (10 lbs), and low-carb dieters 4.7kg (10.3 lbs).

A couple of things to note from this – the first of which is that (as is typical of such studies) this is very modest. The people in the study had a BMI of 27 at least at the start (25 is overweight, 30 is obese). So they had the weight to lose (I understand the limitations of BMI, but on average this is likely true). A healthy diet and exercise program should be able to shed 10 pounds in 4-5 weeks.

But the real story comes from looking at the weight loss curves in the study. We see that all of the weight loss for all three groups occurred in the first 5 months. After that the low-carb and low-fat groups gained back some weight and then stabilized. Both of these groups gained back weight and stabilized along very similar curves. The only real difference between the two is that the low-carb diet lost more weight in the first 5 months – after that there was no advantage to low carb. This is consistent with earlier studies.

The Mediterranean diet also only resulted in weight loss for about 5 months but then stabilized without a period of weight gain.

To my understanding of the relevant literature, this study changes nothing and only confirms prior studies. Most people who diet will initially lose weight, and most will gain some or all of that weight back. There is a short term (5-6) month advantage for low-carb diets, but no long term advantage. Overall in such studies total weight loss is very modest.  Weight loss comes from reduction of calories and increased exercise.


The media is reporting as surprising that the low-fat diet did not have an advantage to lipid profile over the low carb diet. This was surprising 20 years ago, but again this has already been established by previous trials. In this trial the ratio of LDL (bad cholesterol) to HDL (good cholesterol) improved by 20% for the low carb group, 17% for the Mediterranean, and 12% for the low-fat group. This is in rough keeping with the amount of weight loss.

This study does support prior research in establishing that a low-carb diet does not have a negative effect on lipid profile, as was feared by low-fat advocates. One interpretation of this is that weight loss is a dominant factor on lipid profile – trumping any differences in diet, at least as measured in these studies. Weight loss is a marker for level of exercise, which is known to increase HDL, and total caloric intake.  A low total caloric intake may be a better predictor of total fat intake then the ratio of fat to carbs and protein.


This study, which  appears to be a good study in terms of design and execution, and has the advantage of a 2 year follow up, generally supports the findings of prior studies. It reinforces my prior conclusions about different diet types and weight loss and overall health.

The primary lesson to take away, in my opinion, is that calorie control and exercise are of primary importance in losing weight, keeping it off, and maintaining a healthy lipid profile. Focusing on what kinds of food one eats makes little difference in the long run. Think about it – all the media hoopla is ultimately about 4 pounds difference over two years of dieting. That is nothing.

It’s even worse because the only difference really came from the induction phase of very low carbohydrates (20g per day) –  after that there was no advantage to low-carb. This is a short-term, and ultimately failed strategy for weight control.

To be clear, the kinds of foods one eats does matter for things like overall nutrition. There is still evidence that it also mattes for lipid profile and diabetic health – but even for these issues overall weight and regular exercise are the dominant factors. They overwhelm the modest differences in diet types.

I agree with those who argue that studies like this that fuel the public’s obsession with diet strategies is partly to blame for the obesity epidemic. It is taking our eyes off the ball. What works long term is lifestyle changes, making good food choices easier for more people, increasing exercise, and portion control. Basic lifestyle factors.

What this study really shows is that tweaking the ratio of macronutrients in one’s diet is as effective as rearranging the deck furniture on the Titanic.

14 responses so far

14 thoughts on “New Diet Study Compares Low-Carb with Low-Fat”

  1. Niobe says:

    I think the most basic mistake people make is, after losing weight and being at a lower weight that can be sustained with less calories, they go back to their old portions and thus overeat for their weight and gain back.

    Sure you may plateau at 160 at 3000 calories a day, but going down to 120 you need way less to maintain plateau, unless you’re an athlete.

    Of course we live in a world where 160 is the target weight.

  2. Jim Shaver says:

    What this study really shows is that tweaking the ratio of macronutrients in one’s diet is as effective as rearranging the deck furniture on the Titanic.

    Actually, in a literal sense, rearranging deck furniture on the Titanic would be more effective for weight loss than the diet schemes, since it would probably be good exercise (assuming you could have done it for longer than just one voyage).


  3. richdiet says:

    Slightly on-topic question: What *is* the actually utility of BMI as a predictor of anything to do with health? Do higher scores really correlate with negative health outcomes? Are the boundaries between ‘normal’ and ‘obese’ on the scale well-calibrated?

    I am skeptical because of past criticisms (like Penn and Teller’s) of BMI I’ve heard, plus my own experiences as a weightlifter (don’t worry, my personal anecdotal evidence would be cheerfully discarded given some decent data). Maybe body composition doesn’t matter … perhaps 20 extra pounds is 20 extra pounds regardless if it’s fat or muscle, but I think it’s a question worth asking.

    Is there any research that addresses these kinds of questions?

  4. tooth fairy says:

    BMI is certainly not perfect, it can be applied to the general public, who aren’t more than one or two standard deviations from the average weight height etc. if you’re a body bulider and you’re 160cm your BMI will be quite large and if your quite tall and well built it will also be innacurate, skin folds are a better way but they are more obstrusive adn also you have to be consistent with your measurements, alot of gyms will have someone do a clients skin fold and in 6 weeks for a follow up another person will do the measurements and most likely wont measure in exactly the same plcaces, the weight scales that bounce elsectric signals through the body are OKAY because they read consistantly and give a reasonably accurate measure in terms of overall fat % adn water etc. I’ve been a PT for 10 years now and in my experience a net drop in calorie intake is the way to go-here’s a calculation 24 times your weight will give you your BMR-Base Metabolic Rate, for a relatively normal sized person this is the ammount of calories that the body will burn everyday compleating normal mundane tasks(for an over weight person times by 21 or so becasue alot of the weight is fat and fat as apposed to muscle doesn’t have busy active cells, not much blood passes through so reactions within the cell are far less) so if you can intake this number or 50-200 calories less(depends on your size) than your BMR and do at least 20 mins of excersise a day you’ll be on your way to a slow and healh body adaptation.

  5. tooth fairy says:


    check this link out by the way!! ***ALARM BELLS***

  6. superdave says:

    amazing how often woo seems to claim something that normally takes hard work is actually really easy…

  7. tooth fairy says:

    i know, why the hell would doctors not want people to loose weight? such a crank!

  8. tooth fairy says:


  9. eiskrystal says:

    It’s the new “toxins”. Creatures with teeth…but no eyes…ooh. Definitely not to be trusted.

    I wonder if she would trust them if they did have eyes. Would she give them her security details i wonder?

    She is right about the Health professionals and diet fad foods though. However since she is no doubt touting herself as a health professional…

    I also can’t help noticing that the first person on the list of pics appears to have lost weight by shrinking. Since she looked perfectly healthy beforehand, perhaps this was the only way she could lose the weight.

  10. A scholarship athlete in college (ice hockey), I had remained in pretty good shape for maybe ten years after a blown out knee ended my playing days as a senior. During my 30s, however, I slowly lost muscle tone and added weight in small increments, a common enough trajectory for folks moving from youth to middle age. By my mid 40s I had grown into a pudge ball on a clear trajectory for outright obesity. Visions of my own mortality, also common for my then age, dictated I make changes I’d been putting off for years.

    Luckily my physician was an old school guy who told me that losing weight was a simple formula of decreasing calories and increasing exercise while maintaining nutrition. He also cautioned me not to confuse ‘simple’ with ‘easy’, that it would take a strong commitment and consistency in effort.

    I think the most important insight he gave me was that you cannot reasonably expect to take off -and keep off -in a few months all the excess weight that took years and years to accumulate. He advised I take the long view, a modest monthly weight loss goal of perhaps 5 lbs, with the idea that (1) success would translate into a 60 lbs per year weight loss, and (2) such a comparably long (to faddish diets) program of diet and exercise would require a literal change in how I lived my life in terms rather than a much er diet period after which I would most likely relapse and regain weight.

    I’m 52 now, and moved from 340 lbs at age 46 (hey, I’m 6’6″) to my current relatively svelte drift between 240 in the winter, 230 in the summer. I hit 265 by age 48 after 18 months of his plan, 240 by age 50 or so, and have managed to keep it off till now as I look at age 53 this September*. (Inexplicably, I also lost 1.5 inches in height – cumulative effect of gravity? lol)

    I also had knee surgery along the way, and have been skating again for three years. I have sent my player stats from 1973-77 and resume to our local NHL team, the Carolina Hurricanes, and await certain call up for training camp. I sent it 4 years ago and haven’t heard anything yet, but I’m sure this is just an oversight by Hurricanes management.

    *63 shopping days left till my birthday.

  11. BigSven says:

    You say that “What works long term is lifestyle changes, making good food choices easier for more people, increasing exercise, and portion control.”

    Given that this study did not measure the low-carb dieters caloric intake, please explain how this study supports your statement.

    It may well be that more exercise helps reduce weight long term- though there’s no peer reviewed evidence for that- but this study isn’t evidence either way. And as to “making good food choices,” well, that’s exactly what this study is looking at- what does “good food choices” mean?

    It’s disingenuous to argue that a 60% increase in sustained weight loss is “rearranging the deck furniture on the Titanic.” It points to a lack of understanding of how appetite, diet, and weight are related, and bears further investigation.

    What you should be skeptical about is the assumption that the AMA, AHA and diet industry understands what causes and cures obesity.

  12. tooth fairy says:

    No peer reviewed evidence????????? dude we can measure the physiological effects of excercise on the body and it’s acute outcomes and the chronic outcomes and you can have anyone do a sudy and you WILL find that in conjunction with a healthy eating plan you will loose weight. And i can see what you’re thinking-“in conjunction with health eating” well i’d hypothosise that even eating the same as what the subjects were used to, then an increase in excersise would still result in a net loss of weight.

  13. Owen says:

    BigSven: “Given that this study did not measure the low-carb dieters caloric intake, please explain how this study supports your statement.”

    What? But they did measure it. They just didn’t *restrict* it. Check out the study results, they show the low-carb types were on about 550 calories a day less than when they started.

Leave a Reply