A Possible Metabolic Advantage … But It Will Kill You

Two studies involving low-carb diets made headlines this week.  One suggested a low-carb diet may provide a metabolic advantage, while the other suggested a low-carb diet may kill you.

Let’s start with the good news:

Low-carb diet burns the most calories in small study

A new study is raising questions about the age-old belief that a calorie is a calorie.

The research finds that dieters who were trying to maintain their weight loss burned significantly more calories eating a low-carb diet than they did eating a low-fat diet.

The study, funded by the National Institutes of Health, was designed to see if changing the type of diet people consumed helped with weight maintenance because dieters often regain lost weight.

So scientists had 21 obese participants, ages 18 to 40, lose 10% to 15% of their initial body weight (about 30 pounds). After their weight had stabilized, each participant followed one of three different diets for four weeks. Participants were fed food that was prepared for them by diet experts. The dieters were admitted to the hospital four times for medical and metabolic testing.

The diets had the same number of calories, but the fat, protein and carbohydrate content varied. Those diets:

•A low-fat diet which was about 20% of calories from fat and emphasized whole-grain products and fruits and vegetables.

•A low-carb diet, similar to the Atkins diet, with only 10% of calories from carbohydrates. It emphasized fish, chicken, beef, eggs, cheese, some vegetables and fruits while eliminating foods such as breads, pasta, potatoes and starchy vegetables.

•A low-glycemic index diet, similar to a Mediterranean diet, made up of vegetables, fruit, beans, healthy fats (olive oil, nuts) and mostly healthy grains (old-fashioned oats, brown rice). These foods digest more slowly, helping to keep blood sugar and hormones stable after the meal.

Findings, published in this week’s Journal of the American Medical Association: Participants burned about 300 calories more a day on a low-carb diet than they did on a low-fat diet. “That’s the amount you’d burn off in an hour of moderate intensity physical activity without lifting a finger,” says senior author David Ludwig, director of the New Balance Foundation Obesity Prevention Center at Boston Children’s Hospital.

Now, before we all jump up and down and yell “Hallelujah!” we should keep in mind that a study population of 21 is pretty small.  The smaller the study population, the more likely a significant difference in the results is due to chance.  That’s just basic statistics.  If I flip a coin four times and it comes up heads three times, I can’t declare that my coin will come up heads 75% of the time.  My sample size was too small.  If it comes up heads 750 times out of a thousand, now we’re talking.

With that caveat out of the way, it was a controlled clinical study and does suggest a low-carb diet might provide a metabolic advantage.   That would certainly fit with my experience.  I always found it difficult to lose more than 10 pounds on a low-fat diet, and I had to put up with being hungry most of the time to achieve even that modest result.

I wasn’t hungry while losing weight on a low-carb diet, and this study hints at what I believe is the reason:  fuel availability.  If you burn more calories on a low-carb diet than on a low-fat diet, even at the same caloric intake, then your body is either storing less of what you eat or tapping more of what you’ve already stored.  Either way, your cells are getting more fuel, which means your body is less likely to slow your metabolism in response to what it considers a fuel shortage.

I think it’s significant that the dieters had already lost an average of 30 pounds before being divided into the three diet groups.  To keep your blood sugar within the biologically safe range, your body needs to switch between storing fatty acids when blood sugar goes up (so you’ll burn the excess glucose first), then releasing fatty acids as blood sugar goes down (so it doesn’t drop too low).

I believe what some people call a set point is related to our ability to release fatty acids at a rate sufficient to supply us with fuel when blood sugar begins to fall.  Obese people release as many fatty acids on average as non-obese people, but here’s the kicker:  they require more fat mass to do so.  Each unit of fat releases fewer fatty acids, so they need more body fat to release the same number of fatty acids as a thin person.  So at some point, shrinking the fat mass means releasing fewer fatty acids than the body needs to keep blood sugar stable.  That’s when the body rebels and slows down the metabolism (and ramps up hunger) instead of allowing the glucose/fatty-acid balancing act to go all out whack.

Obese people who’ve already lost 30 pounds are more likely to be approaching the point of biological rebellion than people who just started dieting.  That’s where the low-carb diet might provide a metabolic advantage.  By keeping insulin levels lower, the low-carb diet would enable a dieter’s shrinking fat mass to continue releasing fatty acids at the necessary rate.  No fuel shortage, no cellular-starvation response, no slowing of the metabolism.

Naturally, The Usual Suspects don’t agree:

George Bray, an obesity researcher at Pennington Biomedical Research Center in Baton Rouge who has also studied this topic and who wrote the accompanying editorial in JAMA, says that other studies “show that you can do well on any diet as long as you stick to it. Adherence is the major key for weight loss and maintenance. There is no magic in any diet.”

Fuel availability isn’t about magic, Dr. Bray.  It’s about biochemistry.

George Bray is what I’d call an intelligent imbecile. He’s been promoting high-carb and low-salt diets for years and plays games with his study designs and data to support his beliefs.  It was Bray, for example, who declared that macronutrient ratios make no difference for weight loss —  after he conducted a study in which protein and fat intake were altered, but carbohydrate intake stayed the same.  (Uh … wouldn’t you want to try messing with that variable too?)

It was also Bray who conducted the study of salt and hypertension that I mentioned in my Science For Smart People speech.  He put two groups of people on two different diets.  One diet was high in sugar and other junk, the other was more like a Mediterranean diet with lots of vegetables and no sugar.  Bray had each group consume a lot of salt with their diets for several weeks, then measured their blood pressure.  Then he had each group consume very little salt with their diets for several weeks and measured their blood pressure again.  Within each group, the average drop in blood pressure after severely restricting salt was slightly less than three points.  Whoopie.

That wasn’t the result Bray wanted, so in the study’s conclusions he compared blood pressure readings for the junk-food dieters after their high-salt phase to the blood pressure readings for the Mediterranean dieters after their low-salt phase, found a significant difference, and declared it evidence that salt causes high blood pressure.   That’s like comparing the livers of heavy drinkers who consume a lot of salt to the livers of moderate drinkers who consume very little salt and declaring that salt causes liver damage.

But enough about Bray.  There’s more bad science to dissect in the other study that hit the news this week:

‘Atkins’-Type Diets May Raise Risk of Heart Problems

Women who regularly eat a high-protein, low-carbohydrate diet may be raising their risk of heart disease and stroke by as much as 28 percent, a new study suggests.

Although the absolute increase in risk is small — four or five extra cases per 10,000 women — many young women try the Atkins diet or similar regimens and could be setting themselves up for cardiovascular problems later in life, the researchers noted.

“Low-carbohydrate, high-protein diets are frequently used for body-weight control,” said lead researcher Dr. Pagona Lagiou, assistant professor of epidemiology at the University of Athens Medical School in Greece. “Although [the diets] may be nutritionally acceptable if the protein is mainly of plant origin, such as nuts, and the reduction of carbohydrates applies mainly to simple and refined [carbohydrates] like unhealthy sweeteners, drinks and snacks, the general public does not always recognize and act on this guidance.”

That’s because we recognize the guidance as bull@#$%, Dr. Lagiou.  People weren’t living mainly on plant proteins back when rates of heart disease were much lower.

For the study, Lagiou’s team collected data on the diets of more than 43,000 Swedish women who were between the ages of 30 and 49 at the start of the study.

Over an average of 15 years of follow-up, there were more than 1,200 cardiovascular events, including heart disease and stroke. There were more of these events among the women who followed a high-protein, low-carbohydrate diet than among women who didn’t, the researchers found.

Compared with women who veered furthest from the high-protein, low-carbohydrate diet, women who followed the diet most closely increased their risk by 28 percent, even after other risk factors, such as smoking, drinking, hypertension, exercise and fat intake, were taken into account, the researchers noted.

I’m sure you’ve already spotted this study for what it is:  an observational study (strike one) based on food-recall questionnaires (strike two).  We don’t know if the participants accurately recalled what they’d been eating over the years (not likely), we don’t know what kinds of fats the so-called Atkins dieters consumed, and we don’t know if the people on the “high protein” diet ate a lot of meat because they believed it was good for them, or if they’re don’t-give-hoot types who believed red meat is bad for them but ate it anyway.  Those are two totally different types of people.  Since this study began 15 years ago, long before the low-carb/high-fat craze took off among health-conscious Swedes, I vote for the latter.

Another expert, Samantha Heller, an exercise physiologist and clinical nutrition coordinator at the Center for Cancer Care at Griffin Hospital in Derby, Conn., said “the results of this study are not surprising.”

Popular high-protein diets inevitably include an abundance of cheese and red and processed meats, and a dearth of healthy carbohydrates such as whole grains, vegetables, legumes and fruits, she said.

First off, Ms. Heller, the low-carb diets most advocates recommend aren’t “high-protein.”  They’re high-fat.  And secondly, why it is “inevitable” that those diets will include an abundance of cheese and processed meats?  I eat a bit of cheese and almost no processed meat whatsoever.  And if you believe whole grains are “healthy carbohydrates,” I suggest you read Wheat Belly.

Someone sent me a link to the full study, so I read it.  Here are some interesting numbers if we choose to assume (for no apparent reason) that the food-recall questionnaires were an accurate accounting of what people ate.

The researchers divided the women into groups based on carbohydrate and protein intake, but not fat intake.  So their scale went from high-carb/low-protein to low-carb/high-protein.  The researchers declared that the Atkins-type diet increased risk of heart disease and stroke, but the figures for strokes are laughable.  In the high-carb diet group, there were 3.1 ischemic (blood clot) strokes per 10,000 women.  In the high-protein group, there were 5.2 per 10,000 women.  So the actual difference (as I like to pound home in Science For Smart People) is 2.1 ischemic strokes for every 10,000 women.  If we take those odds and express them as percent, here’s what we get:

High-carb diet:  0.031%
High-protein diet:  0.052%

Yee-ikes, my Swedish lady friends.  (Yes, Katarina, I’m talking to you.)  Better lay off the “high-protein” foods.  Your absolute odds of an ischemic stroke will go up by two one-hundredths of one percent.  Oh, and your absolute odds of a hemorrhagic (vessel-bursting) stroke will also go up by one one-thousandth of one percent, according to the study data.  Expressed as the relative risk that researchers love to toss around, that’s a 12% increase (0.9 per 10,000 vs. 0.8 per 10,000), so I’d be worried if I were you.

Since the headlines were about the increase in risk for heart disease, let’s look at those numbers.  In the high-carb group, there were 8.4 cases of heart disease per 10,000 women.  In the high-protein group, there were 12.6 cases per 10,000 women.  So expressing the odds as percentages:

High-carb diet:  0.084%
High-protein diet:  0.126%

Even if you accept that the food-recall questionnaires were accurate, that the “high-protein” diet was actually an “Atkins-like” diet (not just people eating a lot of processed meat), and that these figures are actually telling us something about cause and effect (which an observational study cannot do), we’re looking at an absolute difference in the odds of about four one-hundredths of one percent.

In order to stay lean and continue feeling strong and energetic, I’ll roll the dice and take those odds.  I have a feeling I can beat them.


If you enjoy my posts, please consider a small donation to the Fat Head Kids GoFundMe campaign.
Share

122 thoughts on “A Possible Metabolic Advantage … But It Will Kill You

  1. Pai

    “Popular high-protein diets inevitably include an abundance of cheese and red and processed meats, and a dearth of healthy carbohydrates such as whole grains, vegetables, legumes and fruits, she said.”

    Obviously she knows jack-squat about low-carb diets. On what planet do Paleo/Primal people eat a ‘dearth’ of vegetables?

    There’s no shortage of vegetables on the plate in this house. Without the bread, rice, white potatoes, etc., vegetables are what fill the non-meat side of our plates.

    Reply
  2. Claude

    One point worth noting is that the researcher claim to have adjusted for fat intake risk factor. This leads to one question, one separate conclusion and one observation.

    Question: what was the fat consisting of? Was it saturated, mono, poly, trans, rich or poor in Omega3?

    Conclusion: If they really adjusted the fat intake risk factor as I think they did, the conclusion is that the lower fat intake ratio actually increase your risk of contracting a stroke. Perhaps it would be helpful if you will explain one day in layman terms how risk factors are adjusted.

    Observation: I’ often puzzled as to why researchers don’t include data from kidney disease when researching the occurrence of IHD. The two systems are closely linked with one affecting the other and speaking to some nephrologists I’m getting the impression that kidney diseases are getting more rather than less frequent.

    Reply
  3. Paleo Suz

    I wish they’d stop with the sensationalist headlines based on poor interpretations of studies. My non-paleo (fat-fearing) friends and co-workers don’t look past the headlines…

    Newspapers need dramatic headlines to sell.

    Reply
  4. Rebecca Foxworth

    Huh. I assume, however, that when comparing a person who eats high carb and is 100 pounds overweight with others who switched to low carb and LOST THE EXCESS WEIGHT, you might find a completely different picture when it comes to strokes.

    I would suspect as much.

    Reply
  5. mezzo

    i would appear that the researchers extracted the data and turned them into a score. But they only looked at protein and fat and neglected the carbs. I think Denise Minger is on to that study as well. I am sure she will dissect it in her usual precise and humorous manner.

    They counted protein and carbohydrate, but not fat.

    Reply
  6. Jørg V.

    I also love the fact that the authors of the last paper you talk about says this in the discussion: “Finally, we did not have data on blood cholesterol, an important risk factor for cardiovascular diseases, but even if such values were available, they would probably be, at least partly, intermediates in the association between diet and incidence of cardiovascular disease”

    So, cholesterol is really important, but in THIS case it wouldn’t matter?

    Also “data on drugs for cardiovascular diseases were not available (the relevant registry was not operational during most of the follow-up period)”

    Guess that’s not relevant in this study… 😛

    The whole study is irrelevant, so why should they worry about those factors?

    Reply
  7. Guy Incognito

    These studies never seem to go into detail about exactly what types of food are consumed. For example oils, we were never meant to consume seed oils yet I’m sure these items where cooked with “healthy” unsaturated fats. What were the animals consumed fed? Free range chicken is much healthier than chickens fed those “heart healthy” grains.

    They should also do further studies on the c-reactive protein (the inflammation hormone they talk about) and cortisol (an anti-inflammation hormone). Do the both of them together cancel each other out and it’s just Chicken Little they sky is falling? Don’t all those grains consumed on low fat diets cause inflammation too?

    Reply
  8. Swedo

    The Swedish study is a rather striking example of most things that are wrong with dietary “science” (term used loosely).

    First the mundane stuff, that we can check off as we go along (it can almost be assumed, but let´s at least tick the boxes):

    – Observational study? (Of course)
    – Mail-in questionnarie-based study? (Naturally!)
    – Relative risk ratios below 2? (Checkitycheck!)

    The above alone lands the study in “low” grade of evidence land even before we move on to the juicy stuff, according to GRADE evidence criteria (I.e. http://www.gradeworkinggroup.org/FAQ/evidence_qual.htm)

    Now for the more out-of-the-ordinary stuff:

    – First, let´s note that no person in this study can be said to have been on any particular “diet”. It´s a questionnaire-based observational trial. The “diet” comparison is a result of data sorting conducted by the researchers themselves.

    – Hence, it´s not surprising that the extreme “low carb” decile (the 10 percent of the study population with the lowest carbohydrate intake) had an intake of… 124 grams of carbohydrate per day.

    – However, the most striking thing is that they decided to come up with a new “diet” of their own that they defined according to (obscure) criteria.

    Rather than trying to post-construct a “diet” high in fat and low in carbohydrate (what the term LCHF* implies), they decided to sort the data according to “low carbohydrate high protein score”.

    At the same time, the article is obviously (going by the discussion) intended as a blow at the current low-carb high-fat craze in Sweden, that is irking the dietary establishment greatly.

    Which raises the fascinating question: Did they actually attempt to test a high-fat diet first, and fail to achieve desirable results? Or are they just honestly oblivious to what they are claiming to be “testing”?

    All good points. I think we’re going to see a lot of desperation among the grain-pushers as more and more people go low-carb or paleo.

    Reply
  9. Swedo

    Re: Small study size in the Metabolic Advantage study.

    What’s interesting is that the results are highly significant despite the small study size.

    So, if there is no design error, or large fluke (easy to check through replication / expansion), the small study size on the flip side implies that the results have large real-world effects. (I.e. a study with tens of thousands of participants can get “statistical significance” for results that have little clinical significance for any particular individual).

    I think it’s an encouraging study and matches my experience and the experiences of quite a few others I know. I’d like a larger version so it can’t be written off as a fluke.

    Reply
  10. Mats the swede

    When I first read the newspaper in sweden that published the result of the stydy, I was kind of sceptical… I did as you said in your filmclip science for smart people and check the study myself and as you said, observational study, people answering essays about what they ate. Not quite quality assured. Conclusion most likely, the result from the study is at best questionable and at the worst pure BS.

    Thank you Tom for showing me how to check my sources especially when it is reasearch as this.
    /Mats

    Reply
  11. NM

    Tom, the second study is even more shoddy than you suggest. The “scoring” system used to determine whether someone has a propensity of protein or carbs in their diet is fundamentally broken and hopelessly ambiguous. Indeed, it is so moronic, it takes one’s breath away that professional researchers could have the audacity to propose it.

    Zoe Harcombe deals with the study in full here:
    http://www.zoeharcombe.com/2012/06/can-atkins-diet-raise-heart-attack-risk-for-women/

    In particular, the scoring system adds up two figures for their final score – one component of which deals with protein, the other carbs, and then munged into one variable. As a computer programmer, you will immediately see the problem with trying to turn a two-part array into a simple scalar integer:
    “Low protein is 1, high protein is 10, low carbohydrates are 10, high carbohydrates is 1.
    So to have sum score of 11 it can be low protein (1) and low carbs (10) or high protein (10) and low carbs (1) or any sum between: 1+10, 2+9, 3+8, 4+7, 5+6, 6+5,7+4, 8+3, 9+2 or 10+1, all give 11 as score.”

    Yup, you’ve got to wonder if they devised different scales until they found one that produced the results they wanted.

    Reply
  12. Pai

    “Popular high-protein diets inevitably include an abundance of cheese and red and processed meats, and a dearth of healthy carbohydrates such as whole grains, vegetables, legumes and fruits, she said.”

    Obviously she knows jack-squat about low-carb diets. On what planet do Paleo/Primal people eat a ‘dearth’ of vegetables?

    There’s no shortage of vegetables on the plate in this house. Without the bread, rice, white potatoes, etc., vegetables are what fill the non-meat side of our plates.

    Reply
  13. Megan

    In my small study of one (namely, me!) I discovered that after losing weight (50 pounds) eating a low carb and high fat diet allows me to maintain my new healthy weight easily. What is more important is that I can eat around 500 calories more following this low carb regime than I could prior to losing weight before I start to put on weight.

    Now I can eat up to 2,000 calories before my weight increases, before I lost the weight I could only eat 1500. The difference? – the type of diet. Not magic! Looks like my metabolism survived losing 50 pounds on a ketogenic diet.

    So I am maintaining my weight on a higher calorie intake despite having lost weight and therefore, not requiring the same number of calories.

    That sounds like a metabolic advantage to me! And I will take advantage of it – thank you very much!

    And no laws of physics were violated in the process. I’ve found I can eat a LOT of food without regaining weight if I keep the carb count low.

    Reply
  14. Claude

    One point worth noting is that the researcher claim to have adjusted for fat intake risk factor. This leads to one question, one separate conclusion and one observation.

    Question: what was the fat consisting of? Was it saturated, mono, poly, trans, rich or poor in Omega3?

    Conclusion: If they really adjusted the fat intake risk factor as I think they did, the conclusion is that the lower fat intake ratio actually increase your risk of contracting a stroke. Perhaps it would be helpful if you will explain one day in layman terms how risk factors are adjusted.

    Observation: I’ often puzzled as to why researchers don’t include data from kidney disease when researching the occurrence of IHD. The two systems are closely linked with one affecting the other and speaking to some nephrologists I’m getting the impression that kidney diseases are getting more rather than less frequent.

    Reply
  15. Paleo Suz

    I wish they’d stop with the sensationalist headlines based on poor interpretations of studies. My non-paleo (fat-fearing) friends and co-workers don’t look past the headlines…

    Newspapers need dramatic headlines to sell.

    Reply
  16. Tyler

    Can I just say thank you Tom for your years of great posts, movie, and talks?

    I saw the second study pop up on the CBC website a day or so ago and made all of the same points you made. After looking at the original study of course since the CBC article was making it sound like we’re all going to die of heart disease because we like eating meat. I’m not sure I would have been equipped with the tools I needed to do so before I found your movie and website and you got me thinking about things I hadn’t considered before.

    Of course one person replied to my comments on the CBC article saying the relative jump in risk was significant. Maybe statistically significant enough to direct further research, but with all of the other things that could kill me today, I’m not going to worry about the absolute jump (is jump appropriate? More like baby step really?) in my heart disease risk because I eat low carb and like ground beef.

    Statistically significant doesn’t necessarily translate to “meaningful.” Because of the large study population, the researchers can say their results (based on notoriously inaccurate food surveys) are “significant,” but an increase of four one-hundreds of one percent is practically nothing as far as absolute risk.

    Reply
  17. Rebecca Foxworth

    Huh. I assume, however, that when comparing a person who eats high carb and is 100 pounds overweight with others who switched to low carb and LOST THE EXCESS WEIGHT, you might find a completely different picture when it comes to strokes.

    I would suspect as much.

    Reply
  18. mezzo

    i would appear that the researchers extracted the data and turned them into a score. But they only looked at protein and fat and neglected the carbs. I think Denise Minger is on to that study as well. I am sure she will dissect it in her usual precise and humorous manner.

    They counted protein and carbohydrate, but not fat.

    Reply
  19. Jørg V.

    I also love the fact that the authors of the last paper you talk about says this in the discussion: “Finally, we did not have data on blood cholesterol, an important risk factor for cardiovascular diseases, but even if such values were available, they would probably be, at least partly, intermediates in the association between diet and incidence of cardiovascular disease”

    So, cholesterol is really important, but in THIS case it wouldn’t matter?

    Also “data on drugs for cardiovascular diseases were not available (the relevant registry was not operational during most of the follow-up period)”

    Guess that’s not relevant in this study… 😛

    The whole study is irrelevant, so why should they worry about those factors?

    Reply
  20. Guy Incognito

    These studies never seem to go into detail about exactly what types of food are consumed. For example oils, we were never meant to consume seed oils yet I’m sure these items where cooked with “healthy” unsaturated fats. What were the animals consumed fed? Free range chicken is much healthier than chickens fed those “heart healthy” grains.

    They should also do further studies on the c-reactive protein (the inflammation hormone they talk about) and cortisol (an anti-inflammation hormone). Do the both of them together cancel each other out and it’s just Chicken Little they sky is falling? Don’t all those grains consumed on low fat diets cause inflammation too?

    Reply
  21. Swedo

    The Swedish study is a rather striking example of most things that are wrong with dietary “science” (term used loosely).

    First the mundane stuff, that we can check off as we go along (it can almost be assumed, but let´s at least tick the boxes):

    – Observational study? (Of course)
    – Mail-in questionnarie-based study? (Naturally!)
    – Relative risk ratios below 2? (Checkitycheck!)

    The above alone lands the study in “low” grade of evidence land even before we move on to the juicy stuff, according to GRADE evidence criteria (I.e. http://www.gradeworkinggroup.org/FAQ/evidence_qual.htm)

    Now for the more out-of-the-ordinary stuff:

    – First, let´s note that no person in this study can be said to have been on any particular “diet”. It´s a questionnaire-based observational trial. The “diet” comparison is a result of data sorting conducted by the researchers themselves.

    – Hence, it´s not surprising that the extreme “low carb” decile (the 10 percent of the study population with the lowest carbohydrate intake) had an intake of… 124 grams of carbohydrate per day.

    – However, the most striking thing is that they decided to come up with a new “diet” of their own that they defined according to (obscure) criteria.

    Rather than trying to post-construct a “diet” high in fat and low in carbohydrate (what the term LCHF* implies), they decided to sort the data according to “low carbohydrate high protein score”.

    At the same time, the article is obviously (going by the discussion) intended as a blow at the current low-carb high-fat craze in Sweden, that is irking the dietary establishment greatly.

    Which raises the fascinating question: Did they actually attempt to test a high-fat diet first, and fail to achieve desirable results? Or are they just honestly oblivious to what they are claiming to be “testing”?

    All good points. I think we’re going to see a lot of desperation among the grain-pushers as more and more people go low-carb or paleo.

    Reply
  22. francis

    n=21 is NOT a small sample size if each person in the sample showed the same size effect. One needs very large samples to demonstrate statistical significance especially when the size of the effect is anticipated by the researchers to be very small. I will be interested to learn about that factor when I read the study. thanks for posting it. FWIW I started on Michael Eades’ 6 week cure in October of 2009 at 173#, migrated through Kurt Harris’ earlier Paleo and dropped to 165-168, and last October 2011, adopted William Davis’ Wheat belly as my go to fuel. I’m now stable at 158-162. I’m 74 btw w/ 9-13% body fat.

    Sure, size of the population and size of the effect both figure in. I’d still like to see something like this done with a larger population.

    Wow! You’re one cut 74-year-old.

    Reply
  23. Swedo

    Re: Small study size in the Metabolic Advantage study.

    What’s interesting is that the results are highly significant despite the small study size.

    So, if there is no design error, or large fluke (easy to check through replication / expansion), the small study size on the flip side implies that the results have large real-world effects. (I.e. a study with tens of thousands of participants can get “statistical significance” for results that have little clinical significance for any particular individual).

    I think it’s an encouraging study and matches my experience and the experiences of quite a few others I know. I’d like a larger version so it can’t be written off as a fluke.

    Reply
  24. Mats the swede

    When I first read the newspaper in sweden that published the result of the stydy, I was kind of sceptical… I did as you said in your filmclip science for smart people and check the study myself and as you said, observational study, people answering essays about what they ate. Not quite quality assured. Conclusion most likely, the result from the study is at best questionable and at the worst pure BS.

    Thank you Tom for showing me how to check my sources especially when it is reasearch as this.
    /Mats

    Reply
  25. NM

    Tom, the second study is even more shoddy than you suggest. The “scoring” system used to determine whether someone has a propensity of protein or carbs in their diet is fundamentally broken and hopelessly ambiguous. Indeed, it is so moronic, it takes one’s breath away that professional researchers could have the audacity to propose it.

    Zoe Harcombe deals with the study in full here:
    http://www.zoeharcombe.com/2012/06/can-atkins-diet-raise-heart-attack-risk-for-women/

    In particular, the scoring system adds up two figures for their final score – one component of which deals with protein, the other carbs, and then munged into one variable. As a computer programmer, you will immediately see the problem with trying to turn a two-part array into a simple scalar integer:
    “Low protein is 1, high protein is 10, low carbohydrates are 10, high carbohydrates is 1.
    So to have sum score of 11 it can be low protein (1) and low carbs (10) or high protein (10) and low carbs (1) or any sum between: 1+10, 2+9, 3+8, 4+7, 5+6, 6+5,7+4, 8+3, 9+2 or 10+1, all give 11 as score.”

    Yup, you’ve got to wonder if they devised different scales until they found one that produced the results they wanted.

    Reply
  26. Megan

    In my small study of one (namely, me!) I discovered that after losing weight (50 pounds) eating a low carb and high fat diet allows me to maintain my new healthy weight easily. What is more important is that I can eat around 500 calories more following this low carb regime than I could prior to losing weight before I start to put on weight.

    Now I can eat up to 2,000 calories before my weight increases, before I lost the weight I could only eat 1500. The difference? – the type of diet. Not magic! Looks like my metabolism survived losing 50 pounds on a ketogenic diet.

    So I am maintaining my weight on a higher calorie intake despite having lost weight and therefore, not requiring the same number of calories.

    That sounds like a metabolic advantage to me! And I will take advantage of it – thank you very much!

    And no laws of physics were violated in the process. I’ve found I can eat a LOT of food without regaining weight if I keep the carb count low.

    Reply
  27. Daytona

    I found a great analysis of the “low-carb causes heart disease” study. http://www.weightymatters.ca/2012/06/what-reading-that-low-carb-gives-you.html

    What’s really interesting it that the “low-carb” group was eating 157g of carbs (40%). So on top of all the other issues (like being asked to recall what you ate over the past 6 months and assume they continued to eat that way for 15 years), they weren’t even looking at what I would consider a true low-carb diet.

    Same old, same old. Anything below 50% is considered “low-carb” by these goofballs.

    Reply
  28. Zooko

    “even after other risk factors, such as smoking, drinking, hypertension, exercise and **fat intake**, were taken into account, the researchers noted.”

    Wait, so they factored out fat intake? So this is comparing people who ate the same amount of fat but more protein against those who ate the same amount of fat but less protein? And did they factor out absolute calories as well? It isn’t listed in the sentence above.

    I haven’t read the actual study yet.

    I wish I understood better how this “taken into account” stuff works. If I were doing it, my naïve approach would be stratification — find participants who had similar fat intakes as one another and look at differences just among them. I don’t know how you would generalize that to “take into account” five different factors like the ones they listed above, though, since there might not be enough people in each of the resulting categories.

    (I’ve heard of these things like “Analysis of Variance” — https://en.wikipedia.org/wiki/Analysis_of_variance — but I don’t understand them, or what effect they would have on these results if fat intake were one of the inputs.)

    They “take into account” those other variables by attempting to balance their potential impact across groups. It’s an area where researchers can get a little creative with their data if they’re so inclined.

    Reply
  29. Carolyn Gillham

    How many women were on birth control pills (or other hormonal meds) and what were their daily stress levels and sleep patterns?

    No idea. Those questions may not have been asked in the survey.

    Reply
  30. Trina

    Hi Tom, I’ve been following your blog since I watched FatHead (which was excellent by the way).

    I just read the following article this week http://www.theglobeandmail.com/life/health-and-fitness/health/how-to-keep-the-pounds-off-with-the-right-diet/article4371808/ and made note of their “warnings” about low carb (which I’ve been doing since Feb of this year – lost 32 pounds so far). Did you read this information in the study and what are your thoughts?

    “While the very low-carbohydrate diet did the best job at preserving a participant’s resting energy expenditure, it had potentially harmful effects. Levels of cortisol, a stress hormone, were highest with this diet. Higher cortisol levels may promote fat storage and impair the body’s ability to use insulin, the hormone that regulates blood sugar.

    C-reactive protein, an inflammatory chemical, was also higher on the very low-carbohydrate diet. Ongoing low-grade inflammation contributes to many diseases, including heart disease, stroke, diabetes and Alzheimer’s.

    The findings suggest that the best diet strategy to maintain a weight loss is to cut the glycemic load of your diet rather than its fat content. The low-glycemic diet had similar, although smaller, metabolic benefits to the very low-carbohydrate diet but without the adverse effects on stress and inflammation.”

    The low-carb diet didn’t raise C-reactive protein levels. It lowered them slightly less than the other diets. Cortisol did go up a bit; my wild-@$$ guess is that it was related to the need for gluconeogensis — making glucose out of protein. Perhaps that would go down as the dieters become keto-adapted over time and rely less on glucose for brain fuel. I’d worry less about a minor rise in cortisol than a plunge in metabolism, which is what the low-fat dieters experienced.

    Reply
  31. Justin B

    Sadly, as long as one of these bogus studies hits the headlines at least once a month, it will maintain the general consensus religious belief that it is true. That’s all the USDA and pharmaceutical companies have at this point: hammer the untruth into people’s heads to keep it second-nature, so those people don’t even think to bring science into it.

    And most of the time we see a “meat kills!” headline, it’s based on an observational study.

    Reply
  32. Tyler

    Can I just say thank you Tom for your years of great posts, movie, and talks?

    I saw the second study pop up on the CBC website a day or so ago and made all of the same points you made. After looking at the original study of course since the CBC article was making it sound like we’re all going to die of heart disease because we like eating meat. I’m not sure I would have been equipped with the tools I needed to do so before I found your movie and website and you got me thinking about things I hadn’t considered before.

    Of course one person replied to my comments on the CBC article saying the relative jump in risk was significant. Maybe statistically significant enough to direct further research, but with all of the other things that could kill me today, I’m not going to worry about the absolute jump (is jump appropriate? More like baby step really?) in my heart disease risk because I eat low carb and like ground beef.

    Statistically significant doesn’t necessarily translate to “meaningful.” Because of the large study population, the researchers can say their results (based on notoriously inaccurate food surveys) are “significant,” but an increase of four one-hundreds of one percent is practically nothing as far as absolute risk.

    Reply
  33. Chris

    I am better for getting rid of sugar and flour and grains. My weight is down 40lbs…and my cholesterol went from 289…to 214..hdl was 34..now its’45.

    My doc was happy, BUT he was concerned I was eating too much fat!!!! Diet is like religion and politics..there is NO changing anyone’s mind.

    My vegan family and friends are sort of PISSED that I lost weight this way!

    It’s a cliche, but the best we can do is Be the Change…and rub in their face! 🙂

    A friend of mine experienced a major improvement in his lipid profile on the Atkins diet. His doctor’s attitude was something like, “Okay, if it’s working, keep doing it … but don’t tell me about it.”

    Reply
  34. Lissa

    I read somewhere that their idea of “low carb” was still somewhere north of 200 grams a day … and no telling what kind of carbs those were, whether grain, sugar, fruit or what.

    Reply
  35. Stacie

    What I find very telling is the fact that “they” keep pumping out studies to prove their point. Why the need to keep proving something it it is already supposedly true. The same can be said for the statin trials. They continue to talk about meta-analyses, etc. in order to prove their point,20 years after the fact. I think “they” are getting really nervous because the cat is out of the bag!! Too many people have too much money to lose.

    I think you’re exactly right. And they’ll keep trotting out these lousy observational studies because the clinical studies aren’t working in their favor.

    Reply
  36. Steve

    But but but but Tom, low carb will kill you! Haven’t you seen all of the headlines and news reporters talking about the dangers of low carb! You know those DANGERS right, right?! Im sure I’ve seen someone dying on low carb somewhere! I mean, can’t you see how starving yourself okay nutrient depleting low fat diet is better for you, especially with heart healthy whole grains!

    Yeah, I expect to keel over any day now.

    Reply
  37. Marilyn

    “. . .healthy carbohydrates such as whole grains, vegetables, legumes and fruits . . .” It’s as if these people get to a certain point, and then their brains go on “autofill” and they spout exactly the same words every time.

    Yeah, it’s like pushing the play button on a recorded message.

    Reply
  38. charles

    http://anthonycolpo.com/?p=3680

    Finally, a Study that Proves a Low-Carb Metabolic Advantage? Yeah, Right

    NO TIGHTLY CONTROLLED STUDY IN THE HISTORY OF HUMANKIND HAS EVER SHOWN GREATER FAT LOSS ON A TRULY ISOCALORIC LOW-CARB DIET.

    CRP (C-reactive protein) tends to be elevated during inflammatory states such as illness and infection. While CRP was lower on all three diets compared to the pre-weight loss baseline, it remained highest on the low-carb diet (CRP levels were almost identical during the low-fat and low-GI diets).

    For the three diets, cortisol excretion measured via 24-hour urine collection was 50 μg/d for low fat, 60 μg/d for low glycemic index; and 71 μg/d for very low carbohydrate. This adds to the numerous other studies showing carbohydrate restriction to increase cortisol levels (I’ve written about this here).

    So what we appear to have here, folks, is a metabolic milieu during the low-carb diet that indicated a more catabolic and inflammatory state.

    In other words, in the subjects that did experience lower declines in REE and TEE during the low-carb phase, the culprit may have been increased catabolism. If so, the lower drop in EE was not necessarily a good thing – it may have in fact been reflective of harmful rather than beneficial processes occurring in the body.

    Anthony’s “tightly controlled” studies were of semi-starvation diets of around 800 calories. As I’ve written before, I expect a metabolic advantage (if there is one) to show up at higher calorie levels, not levels so low that even a “high carb” diet translates to 100 grams per day or less. I saw a study recently of 1800-calorie diets with different macronutrient ratios. The low-carb group lost the most body fat by a wide margin.

    Reply
  39. Marilyn

    Just for kicks, I googled “healthy whole grains, fruits and vegetables” in quotes and got over 31,000 hits for that combination of words, the first two for DOG FOOD!

    Chareva talked to a dog-food representative at a store once. She said she wasn’t interested in dog foods that contain grains. The rep of course replied that dogs need carbohydrates for energy.

    Reply
  40. francis

    n=21 is NOT a small sample size if each person in the sample showed the same size effect. One needs very large samples to demonstrate statistical significance especially when the size of the effect is anticipated by the researchers to be very small. I will be interested to learn about that factor when I read the study. thanks for posting it. FWIW I started on Michael Eades’ 6 week cure in October of 2009 at 173#, migrated through Kurt Harris’ earlier Paleo and dropped to 165-168, and last October 2011, adopted William Davis’ Wheat belly as my go to fuel. I’m now stable at 158-162. I’m 74 btw w/ 9-13% body fat.

    Sure, size of the population and size of the effect both figure in. I’d still like to see something like this done with a larger population.

    Wow! You’re one cut 74-year-old.

    Reply
  41. Ari

    Tom, you wrote:

    I eat a bit of cheese and no almost processed meat whatsoever.

    How do you know if the meat you eat was almost processed or not? Just because the meat is not processed doesn’t mean that somebody wasn’t about to process it and just decided not to.

    I almost agree with you. I’ll fix the switched words, thanks.

    Reply
  42. Drew @ Willpower Is For Fat Pe

    Tom, Tom, Tom … [shakes head] You clearly don’t understand statistics. I know you follow politics, so let me put this in terms you’ll understand.

    When something is wrong with the country, say unemployment is too high, the TV pundits say, “Our rate of job creation is slower than our population growth. We are effectively losing jobs every day.”

    Someone from Party A will say, “My plan will slow the rate of loss by ten percent.”

    Someone from Party B will say, “Well my plan will slow the rate of loss by twenty percent.”

    The pundits will say, “Party B has a plan to increase employment twice as fast as Party A.”

    See? That’s just good math.

    Indeed. If a program was scheduled to increase by 10% and Congress reduces the increase to 5% instead, that’s a “cut in spending” in D.C.

    Reply
  43. AndreaLynnette

    I’m really, really grateful to people like you who willingly wade through these piles of bovine excrement so I don’t have to. I’m supposed to STOP eating LCHF because of a 0.05% increase in risk, but I SHOULD take scads of medicines with 30-70% risk of side effects instead?! ARGH!

    Excuse me, I’m going to go drink a coconut milk/heavy cream protein shake and eat some bacon.

    Give a toast in my honor when you drink that shake.

    Reply
  44. Daytona

    I found a great analysis of the “low-carb causes heart disease” study. http://www.weightymatters.ca/2012/06/what-reading-that-low-carb-gives-you.html

    What’s really interesting it that the “low-carb” group was eating 157g of carbs (40%). So on top of all the other issues (like being asked to recall what you ate over the past 6 months and assume they continued to eat that way for 15 years), they weren’t even looking at what I would consider a true low-carb diet.

    Same old, same old. Anything below 50% is considered “low-carb” by these goofballs.

    Reply
  45. Zooko

    “even after other risk factors, such as smoking, drinking, hypertension, exercise and **fat intake**, were taken into account, the researchers noted.”

    Wait, so they factored out fat intake? So this is comparing people who ate the same amount of fat but more protein against those who ate the same amount of fat but less protein? And did they factor out absolute calories as well? It isn’t listed in the sentence above.

    I haven’t read the actual study yet.

    I wish I understood better how this “taken into account” stuff works. If I were doing it, my naïve approach would be stratification — find participants who had similar fat intakes as one another and look at differences just among them. I don’t know how you would generalize that to “take into account” five different factors like the ones they listed above, though, since there might not be enough people in each of the resulting categories.

    (I’ve heard of these things like “Analysis of Variance” — https://en.wikipedia.org/wiki/Analysis_of_variance — but I don’t understand them, or what effect they would have on these results if fat intake were one of the inputs.)

    They “take into account” those other variables by attempting to balance their potential impact across groups. It’s an area where researchers can get a little creative with their data if they’re so inclined.

    Reply

Leave a Reply

Your email address will not be published.