As if we needed a reminder that diets mostly fail, The New England Journal of Medicine has published a new report on an intense, tightly controlled experiment involving more than 300 moderately obese people.
After two years of effort the dieters lost, on average, 6 to 10 pounds. The study, funded in part by the Atkins Research Foundation, seemed designed to prove that low-carb diets trump low-fat diets. But in the end, all it really showed is that dieters can put forth tremendous effort and reap very little benefit.
Dr. Dean Ornish, a proponent of low-fat diets, is critical of the study design, particularly the fact that the so-called “low-fat” diet group was really only a moderate-fat diet that included about 30 percent of calories from fat. He writes more about the issue in a Newsweek column here.
The New England Journal report also affirmed something many women have believed all along — that low-carb diets work better for men than women. Male low-carb dieters lost about 11 pounds, compared to about 9 pounds on a Mediterranean diet. Women low-carb dieters lost only about 5 pounds, compared to about 14 on the Mediterranean diet. To read more about the new research, click here.
Healthy living doesn't happen at the doctor's office. The road to better health is paved with the small decisions we make every day. It's about the choices we make when we buy groceries, drive our cars and hang out with our kids. Join columnist Tara Parker-Pope as she sifts through medical research and expert opinions for practical advice to help readers take control of their health and live well every day. You can reach Ms. Parker-Pope at 
The easiest and most pleasurable way to eat well is to cook. Recipes for Health offers recipes with an eye towards empowering you to cook healthy meals every day. 
From 1 to 25 of 120 Comments
Here we go again: another 600-post thread salted with insults and cries of victimization.
TPP, how do *you* interpret this study? Shall we all give up, eat whatever we please, and resign ourselves to a lifetime of being overweight?
— perraI’m testing this to check if it’s working. Don’t approve me!
— Sarah GrahamAfter reading the specifics of this study it was clear that there is no accurate conclusion regarding the “Atkins Diet.” A true Atkins diet would include saturated fat, including lots of cream and butter, creating marvelous satiety and resulting in eating less and controlling blood sugar. According to what I read, the diet used in this study emphasized plant sources of protein - not an Atkins diet by my definition! I can’t imagine what the “dieters” ate - Atkins was very opposed to the use of soy protein.I suspect the less than stellar conclusion was due to a “watering down” of the Atkins diet - too high in protein and carbs and not enough fat. A true Atkins diet is very effective for women and diabetics in its orginal form which includes a high percentage of fat from animal sources. I live on butter on cream and I’m very happy, healthy, and thin (and my entire family is diabetic!)
— PeggyThe most interesting thing to me in the linked article about this study was this:
“The relative reduction in the ratio of total cholesterol to high-density lipoprotein cholesterol was 20% in the low-carbohydrate group and 12% in the low-fat group (P=0.01).”
Remember all those years that the “experts” were insisting that a low-carb diet would inevitably RAISE cholesterol? This study indicates that for many people the best way to lower serum cholesterol may be to eat a low-carb (and therefore relatively high-protein and high-fat) diet. Yet more evidence that the physiology of weight, cholesterol, heart disease, etc., is far more complicated than we thought.
The researchers conclude: ” Mediterranean and low-carbohydrate diets may be effective alternatives to low-fat diets. The more favorable effects on lipids (with the low-carbohydrate diet) and on glycemic control (with the Mediterranean diet) suggest that personal preferences and metabolic considerations might inform individualized tailoring of dietary interventions.”
— eWhat is a diet? The only way to lose weight is to take in fewer calories than are expended. This is a “diet” isn’t it? If diets don’t work, then there can be no lasting weight loss.
— Red GauntletThis is certainly against my own experience.
After age 50 nothing seemed to take it off and keep it off. then a few years ago I discovered the South Beach book. Since I was more interested in bringing down the blood pressure and cholesterol count, I tried the maintenance phase, figuring that any weight that came off as a result was a bonus. I lost about 18 pounds, and most of it has stayed off. The secret there is not low carbs but slow carbs.
As Dr. Agatston says that people will break the good habits occasionally, I don’t feel too guilty about consuming bad carbs from time to time. However, I probably should pursue the straight and narrow for my heart’s sake and maybe lose the last 5-10 pounds.
— JoanDepends on what you mean by “work”. The metabolic biomarkers (LDL, HDL, etc) generally continued to improve modestly over the course of the 2 year follow up, across diets, even though some weight was regained. Thus, one could argue that eating a healthy diet (however one wishes to define that) as opposed to “dieting” can be beneficial even in the absence of significant weight loss. If we are really in the health promotion business (as opposed to the appearance promotion business) then we should take some heart from this.
— MD,CABoy, you gave a different spin on this than did the AP or NBC TV news. They pointed out how well the Atkins diet worked and you came to the conclusion that no diets work. Is that because the results are not what mainstream medicine has been preaching for 25 years. I would think losing 14 and keeping off 10 after two years is a big achievement.
This is the second study in the past few years that said a low carb diet, ala Atkins, is the best for your heart health (at least as measured by the sacred cholesterol levels and ratios.
— RichYep, this study agrees with the other studies, only this one was better done, so it’s more convincing if you were a doubter.
Let us now look at the studies that include exercise as a weight-loss aid. Unfortunately, they don’t do any better. 5.5 kg on average is the best weight loss you can acheive.
Finally, to the studies of gastric surgery: here we see significant weight loss AND improved morbidity/mortality, despite the inherent risks of surgery.
Sad but true: you can’t lose weight without surgery. Unless you’re really exceptional.
— Anonymous MDWhy don’t you mention that the low-carb Atkins diet also lowered cholesterol the most. My mother has been on a high-carb vegan diet for years, and has sky-rocketing triglycerides and blood pressure. I am on a relatively low-carb diet, and my cholesterol levels and blood-pressure is extremely low. I will grant though, it is hard to stay on a low-carb diet. Those sweets are so tempting.
FROM TPP –It didn’t lower cholesterol the most. There were no statistically significant differences in bad cholesterol. The low-carb diet did, however, lower good-bad cholesterol ratios more than the low-fat diet, but the declines appeared to be similar for the Mediterranean diet.
— AtkinsHi Tara,
for your readers, a press release about the NEJM article from Ben-Gurion University is at
http://www.eurekalert.org/pub_releases/2008-07/aabu-ndn071108.php
and although you summarize the results accurately, you forgot to mention that the blood lipid control (TC/HDL and triglycerides) is best on the low-carb diet, although the Mediterranean and low-carb diets were equally effective with weight loss. In every category the low fat diet was clearly the least effective of the three, and was especially problematic for patients needing to control blood glucose levels.
Note too that there was no caloric restriction for the low-carb dieters, which is why many of us find it a much easier way to eat. When will conventional dietary advice change? And when will these ideas be tested rigorously on non-obese people?
You highlight an interesting difference between
the Mediterranean and low-carb results for men and women. Since the number of women was so small in this trial, someone should conduct a similar study, on a larger group of women to see if the result still holds.
— NickFROM TPP — I think the results were presented in a very misleading fashion. The low-carb diet was only compared to the low-fat diet (which wasn’t a true low fat diet) Overall, there were no statistical differences between LDL levels on any of the diets. The low-carb and Mediterranean diets resulted in virtually identical changes in good-bad cholesterol ratios. The low-fat diet also resulted in improvement in cholesterol ratios, but not as much. It would have been useful if the researchers had provided better data to compare the low-carb and Mediterranean diets.
From the reported data it is hard to assess the daily energy expenditure vs daily energy intake in these self-reporting obese subjects. But the lack of significant weight loss on these “diets” is not surprising. A similar study, A to Z, reported in JAMA last year
http://jama.ama-assn.org/cgi/content/abstract/297/9/969?ijkey=925fe6e7e29adc0c774ce4409561da8c68fdfd9d&keytype2=tf_ipsecsha
showed a major discrepancy between what the subjects were reporting and what actually happened. If they had really eaten what they said they did and exercised the way they did they would have lost 100 pounds per year, not 10. It is well known that obese people lie about what they eat and how much they exercise.
See my detailed analysis of A to Z.
http://medicalmyths.wordpress.com/2007/03/12/a-to-z-trial-atkins-tops-the-trial-proved-only-that-most-food-addicts-lie-about-food-consumption-and-exercise/
— Colin Rose MDI think a major problem with diet studies is that they only study diet. But the NYT itself reported on how things like gut flora can affect one’s tendency to be obese (Fat Factors, 2006), along with other
Personally, despite being female, I’ve lost over 30 lbs. over the past year on Atkins. *But* I’ve also done other things: reduced MSG consumption where possible (MSG can actually make rats on a starvation diet obese); worked to balance my gut flora with probiotics; and eaten organic when possible. I’ve been more active, although I have not been consistent about my exercise program. I did not track calories or the amounts of different food types, but I know I ate a lot of vegetables.
I can do all of the above on a Mediterranean diet, or macrobiotic, or a carefully-balanced vegan diet, and gain about a pound every week or two at 1200 calories a day. For some reason my metabolism does not work well with carbs. This may be because my metabolism has been adversely affected by far too many antibiotics, or because of chemicals I was exposed to, or whatever. That’s not the point. The point is that everyone’s metabolism is unique, with a different background, and any study that doesn’t track and measure these factors is wasting its time.
P.S. The big difference between Atkins and other diets for me is that I’m a carb addict and it stops the craving. When I eat the wrong kind of carb, my appetite kicks in and I don’t think about anything but food. I know it’s the carbs and not my brain, because when I’ve accidentally been given carbs by another person (say, a sugared cup of coffee instead of a splenda’d one) the same effect happens. Low-carb - with the other measures! - not only works for me, but by greatly reducing my hunger pangs it frees up my brain to enjoy the rest of my life.
— Alex O’NealWell, this is depressing.
As a personal trainer, I am constantly meeting people who have tried every diet in the book, with little long term success. They lose a few, re-gain a few - the typical yo-yo effect that we see in this study.
Lose the weight initially, plateau for a while and then gradually start the inevitable back towards obesity.
The National Weight Control Registry - http://www.nwcr.ws/ - is the largest prospective investigation of long-term successful weight loss maintenance.
Given the prevailing belief that few individuals succeed at long-term weight loss, the NWCR was developed to identify and investigate the characteristics of individuals who have succeeded at long-term weight loss.
The NWCR is tracking over 5,000 individuals who have lost significant amounts of weight and kept it off for long periods of time.
Registry members have lost an average of 66 lbs and kept it off for 5.5 years.
* Weight losses have ranged from 30 to 300 lbs.
* Duration of successful weight loss has ranged from 1 year to 66 years!
* Some have lost the weight rapidly, while others have lost weight very slowly–over as many as 14 years.
Here is the most important info that we should take away from this program:
* 98% of Registry participants report that they modified their food intake in some way to lose weight.
* 94% increased their physical activity, with the most frequently reported form of activity being walking.
* There is variety in how NWCR members keep the weight off. Most report continuing to maintain a low calorie, low fat diet and doing high levels of activity.
78% eat breakfast every day.
Research on why breakfast is vital - http://healthhabits.wordpress.com/2008/06/24/the-big-breakfast-diet/
75% weigh them self at least once a week.
62% watch less than 10 hours of TV per week.
90% exercise, on average, about 1 hour per day.
Remember these 4 tips -
Watch less tv (that should include internet surfing and video games as well)
Replace some of that tv time with exercise (1 hour each and every day)
Find some way of monitoring your weight / food intake / other health markers
Use a combined food / exercise log, weigh yourself, check yourself out in the mirror - don’t obsess, but don’t forget that it is way too easy to slip back into old habits.
Eat breakfast. I can’t believe how this simple advice is vital in preventing afternoon cravings.
There are lots of other tips that you can employ to make your weight loss more successful, but if you want long term success, ask someone who has lost the weight and kept it off.
DR
— DRhttp://healthhabits.wordpress.com
Low or no carbs is the ONLY way to go. It takes some getting used to at first, but you REALLY see a difference rather quickly. I lost 55 pounds in 3 months. From 255 to 200. No rice, no potatoes… etc. Anything high in carbs I cut out. Stayed between 0 and 50 a day. And yes, you CAN pig out on meat, fish, anything carbfree if you like. Without the carbs, your body can’t convert them into sugar or the bad type of cholesterol. I highly recommend it. I’m no doctor, but it sure worked for me! On a side note, might consider taking some fish oil supplements if you go on the diet. It helps. Once you lose your desired weight, treat yourself to whatever you’d like to eat once a week. It won’t affect your weight. The part that takes some getting used to is the fact that it’s the sugars or carbs that give you that “full” feeling. But if you can keep them (carbs) under 30 grams a day. You’ll definitely drop the pounds.
— JeffIt’s interesting, other news sources on this report are claiming a victory for low-carb when it isn’t that decisive for women or diabetics and you are reporting a general wash and overall failure of all long term diets.
I couldn’t find a source where Dr. Dean Ornish commented on the study but really he is more a proponent of a plant-based diet that a low-fat diet. He doesn’t really advocate lean meat, low-fat dairy or low-fat processed foods.
Of the three diets presented it would be better to categorize Ornish in the Mediterranean Diet category than with the American Heart Association guidelines. Realistically, (and perhaps unfortunately) Ornish would probably be categorized under “extreme diet protocols;” Mediterranean, without animal sources and a reduction of vegetable oil.
Yes, the criticism is that people can’t stick with his diet, but for those who do, what are their outcomes?
It’s an important distinction because Ornish has always been more concerned with the advantages that have been somewhat documented with plant-based diets like reducing cardiovascular disease, diabetes, cancers and a host of “diseases of affluence” than just weight loss, though lower weight affects these outcomes of course. Overtime, low-carb and high fat advocates have begun asserting these types of benefits as well besides rapid weight loss, though their mainstream data sources haven’t been as strong. I’m a low-carb nutritionist somewhere will examine the low-carb diet in the study and point out that is wasn’t true low-carb based on the given intakes.
The Mediterranean Diet did fairly well (considering the outcomes), I’m just curious how an exclusive plant-based interpretation of the diet as Ornish advocates, would have faired, even in a small sample of participants. Is elimination of saturated fat and cholesterol alluded to in mainstream nutrition and epidemiological studies and advocated by Ornish and other plant-based nutritionists truly beneficial? There wasn’t much chance to find out in this study.
But I concur, losing an average of 6-10 pounds over two years isn’t very compelling. I easily lose 6 pounds of (water) weight after running a few miles on a warm day (unless the nutritional scientists correct for this factor.) On the other hand, would there have been a weight gain in these participants overtime if they didn’t follow any diet plan? Weight maintenance is as important, if not more so, as weight loss. Maybe it’s too late for the older generation of obese, but a younger generation who are approaching middle-aged spread may benefit from being on a preventative weight gain diet.
Good study though. I find these diet comparisons fascinating. Probably like most people, I always wish they could have gone a little farther and included more diets, data and participants. But, I’m sure that the undertaking of variables and recording of these three diet plans was already difficult enough.
FROM TPP — I’m troubled at the way this is being reported. This was not a clear victory for the Atkins diet by any stretch, although the researchers tried to present it that way. I’ve added a link to the post with comments from Dr. Ornish. In my view, the Mediterranean diet had the edge in this study, but the reality is that none of them was very successful.
— LexingtonI think the more important result is that low-carb is best for lowering cholesterol.
FROM TPP — The study did not conclude that at all. There was no statistical difference in LDL levels among the three diets.
— RobertThe key to weight loss is fairly simple–caloric expenditure HAS to exceed caloric intake, i.e. don’t eat like a pig, and start exercising–EVERYDAY.
— Matt Gangwer,M.D.after reading the articles about stomach fat i decided that i needed to lose some weight. i ate less. fewer calories. especially desserts. down 6 pounds in a few weeks (thats 163 to 157).
— bruceless fat, fewer carbs, lots of veges.
forget the ‘plans’. just eat well and less.
I would like to recommend the low-carbohydrate approach for most people. By switching from glucose-based fuel to ketones at various times over 24 hours, our bodies consume some of our fat reserves. Carbohydrates in your food become glucose in your blood, leading to insulin secretion, leading to storage of some of these calories as fat. Until you shut down the insulin, you can’t switch to ketone metabolism. And a diet like that permits some indiscretions can still work because once you’ve sinned, you can get back to eating low carb and reprime the ketosis metabolic pathway needed to resume burning your extra calories. And fat in the diet is the very best appetite suppressant. If you doubt me, try eating a whole stick of butter some time.
— Knight MDI had a closer look at the data and calculated that, if the participants really had done what they say they did they should have lost at least 100 pounds in two years on any of the “diets”. If a study on energy flows cannot accurately balance energy input and output it shouldn’t be published.
See my new blog on DIRECT.
http://medicalmyths.wordpress.com/2008/07/16/direct-more-obese-people-lying-about-food-intake/
— Colin Rose MDFor TPP’s comment in 17, the science seems to say
that the ratio of TC/HDL is a superior single cholesterol
marker for heart disease risk, and that LDL levels by themselves
are not a great marker. Here’s the original article:
http://www.annals.org/cgi/content/abstract/121/9/641
(My opinion is that you can thank Pharma for the misplaced
emphasis on LDL as a marker, and its synonym as the bad
cholesterol, because statins are so good at lowering LDL levels. And in any case, the most recent research on cholesterol shows that LDL itself
breaks down into harmless LDLs and bad LDLs.) In the current NEJM article
we’re discussing, the authors are justified in quoting the
statistically significant improvement in the scientifically valid
TC/HDL ratio for the low-carb group as compared to the low-fat group, and also the statistically significant improvements in triglyceride levels for
low carb vs low fat.
I’m not sure the authors are biased against the Mediterranean diet,
they are happy to quote its statistically significant improvement
in blood glucose levels as compared to the low fat diet. And that
both the Mediterranean and low carb diets lowered C-reactive
protein levels significantly, vs. the low fat diet.
I find the NEJM article very well written. The authors are very careful
to give statistical significance values for each of their comparisons,
and they also include all possible comparisons. What jumps out at
me, is that the “low” fat diet is the worst performer in every comparison,
even thought the numbers don’t always reach the level of statistical
significance.
Maybe Dr. Ornish has a point that a lower fat diet should also
— Nickbe studied than the one these authors tried, but on the other hand
the current research is a breath of fresh air to those of us who know that
our own bodies do not tolerate high carbohydrate loads, and that
we need to pretty much avoid all refined carbohydrates. Plus,
this NEJM published study is like the ones Gary Taubes calls
for in his excellent book “Good Calories Bad Calories,” although
as he says, there should also be trials with healthy, non-obese people.
TPP,
I’m puzzled by your comments. The NEJM article says:
“The relative reduction in the ratio of total cholesterol to high-density lipoprotein cholesterol was 20% in the low-carbohydrate group and 12% in the low-fat group (P=0.01).”
Is the difference between 20 percent and 12 percent really statistically insignificant? This seems like a clear difference to me, and a clear indication that low-carb improve cholesterol levels more than low-fat does.
I agree, it would be nice to see the percentages for the Mediterranean diet. I can’t find that in the article, but if you click on Figure 3, you’ll see in panel D that the Mediterranean diet’s success in improving the ratio of HDL and total cholesterol is right in between low-fat and low-carb, and low carb shows the greatest improvement in that ratio.
For triglycerides (Figure 3.B), the low-carb is best at 6 months and it and the Mediterranean are just about tied at 24 months. They both do MUCH better than the low-fat.
Nothing seemed to do much for LDL levels.
This may not be the stunning vindication of the Atkins diet that its supporters were hoping for, but it certainly looks to me like clear evidence that a low-carb diet really does significantly improve blood lipids, –much more than a low-fat diet does and marginally better than a Mediterranean diet does. It’s too bad it doesn’t help with LDL levels as well, but as they say “two out of three ain’t bad” (HLD/total ratio, triglycerides . . .)
Am I missing something in these graphs?
FROM TPP — The problem with this study is that it compared low-carb to low fat. But it didn’t compare low-carb to Mediterranean. So it’s very difficult to tell what’s really going on. You are correct that there was a statistically significant difference in cholesterol ratios between low-carb and low fat (low-carb wins). But the Mediterranean diet also improved cholesterol ratios, but the study doesn’t provide a statistical comparison against Atkins. My comment earlier was that the Atkins people are claiming a victory here but the victory is only over the moderate fat diet used in this study (The truth is, you can’t really call a 30% fat diet a low-fat diet). But it appears that the Atkins and Mediterranean diets were about the same in this study. But the real question is what is clinically meaningful here? Did the dieters lose very much weight at all? No they did not.
I have one other big complaint about the way this study was presented. The researchers counseled the low-carb dieters to choose vegetarian sources of fat and protein. WHat??? That’s not an Atkins diet. Atkins claim to fame is that you can eat all the meat,butter and animal products you want. What is described here is a lot closer to a Mediterranean diet or perhaps the South Beach diet, but certainly not Atkins.
— eFor those who are curious, I urge you to read the book, “Good Calories, Bad Calories,” by Gary Taubes, which directly addresses the conventional wisdom that calorie restriction does not work and that low carb diets do work for weight loss, etc. The book is very well documented.
I graduated with a degree in nutrition and was brought up in the world of conventional wisdom - i.e., a calorie is a calorie, low fat high carb diets are good, the low carb cadre of Atkins and his predecessors is evil, etc. I tend to view things that upset tradition with healthy skepticism, so I was quite surprised that this book forced me to re-evaluate the dietary theories and concepts that I was taught.
That being said, the results of the NEJM study should at minimum open the eyes of those whose thinking is stuck in tradition.
- The “healthy” low-fat approach did not prevail.
- The low-carb diet produced positive results (the book to which I alluded states that this is consistent with studies conducted in the past that were played down).
- Further study needs to occur.
As to the study’s supposed pro-Atkins bias, it should be noted that some scientists who established the dietary do’s and don’s we take for granted today were influenced by their funders too.
— PangaeaThat’s a horribly misleading headline- diets do work when followed. If you don’t follow the diet, then you really aren’t dieting, and you can’t criticize the diet for not working. “More Evidence that Diets Are Really Hard to Follow” would be a much better title.
— Chuckles McGee