SkepticblogSkepticblog logo banner

top navigation:

The Nihilism Diet

by Steven Novella, Jan 03 2011

In a recent New York Times article, Abigail Zuger takes a crack at the diet wars, and does a reasonable job. There are many points left to be made, however.

The question is, essentially, how to maintain a healthful and desired weight. In broad brushstrokes, the variable are — how many calories people consume, what kind of calories are consumed (the proportion of macronutrients — fat, protein, and carbohydrates) and how much people exercise. These are the (allegedly) controllable variables. The variable we cannot control is our genetics, which largely determines our inherent metabolism and hormonal function.

Different people have different interpretations of the evidence. There are also countless popular diets that don't even attempt to square with the evidence, but I will largely ignore them for this post.

Zuger spends much of the article discussing the writing of Gary Taubes – a science journalist and author of Good Calories, Bad Calories, and now an updated and streamlined version titled, Why We Get Fat: And What to Do About It. I have not yet read the book, but I have read many of Taubes' articles and lectures. Zuger summarizes his main points:

We’ve got the whole thing backward, he argues. The overweight are not lazy hogs who eat too much and exercise too little. The thin are not virtuous and disciplined. Rather, all of us are fulfilling a fixed biological mandate, just as growing children are. Our bodies have a nonnegotiable agenda, and our behavior evolves to make that agenda happen, he writes: “Eating in moderation and being physically active (literally, having the energy to exercise) are not evidence of moral rectitude. Rather, they’re the metabolic benefits of a body that’s programmed to remain lean.”

This all sounds nihilistic – our weight is genetically determined, so why bother trying to control your eating and exercise regularly? But Taubes goes on to endorse the low-carb diet. He argues that insulin is the one hormone we can affect, by eating a low-carb diet. Carbohydrates are the villain – the cause of the obesity epidemic.

Taubes knows that his opinions go against the consensus of expert opinion, and he too casually tosses the consensus aside, in my opinion. I do not maintain that the experts are always right, but it's rare that they are systematically wrong in a field with a lot of evidence. At the very least, I would be very concerned if my read of the evidence was not shared by the experts – I would assume that I was missing something, and not the experts.

Here is what I think Taubes is missing: Interpreting the research is complex – and you have to balance basic science with clinical research. I think Taubes is overemphasizing the basic science (and putting it in the wrong context) and is dismissing the clinical research.

For example, Zuger makes her best point when she writes:

How to account for the fact that in virtually all head-to-head comparisons of various diet plans, the average long-term results have invariably been quite similar — mediocre all around? The party line holds that backsliding is universal. Mr. Taubes makes much of the addictive effect of carbohydrates: once you taste them you never forget them.

Zuger is correct – if low carb dieting were the answer, then the clinical research would show it. What the research shows is that any advantage to low carb diets are minimal (if they exist), and short lived. They are  not the answer to weight control. Zuger then lets Taubes off the hook too easily with his reply that carbs are addictive. This is a massive bit of special pleading, a casual dismissal of a large body of clinical evidence that cuts the heart out of Taubes' interpretation of the research.

In addition  Taubes and other low-carb enthusiasts get the context of basic science research wrong. First, I agree that genetics largely determine our inherent tendency to maintain a certain weight. Most people who are obese are so because of their genetically determines hormonal function and appetite control. But genetics are not destiny. While our genetics strongly influence our behavior, they do not control it. We have the ability to hierarchicaly supersede our default behavior. It's not easy, as the research also indicates, but it is possible.

So far we have not found a safe and effective way to alter our hormones. Low carb diets do not accomplish this (as an aside, they are useful for avoiding and treating diabetes, but don't seem to help in weight control). There are appetite suppressants, but these are of limited utility. Maybe one day we'll have an obesity pill, but we don't now.

So – bottom line, we can't change our inherent metabolism or appetite. It is also true that if we decrease our caloric intake we will tend to exercise less, and if we exercise we will get more hungry and tend to eat more. Our bodies manage to balance our caloric intake and output fairly well over long periods of time. But if one exercises regularly and limits their caloric intake at the same time that would avoid this automatic compensation.

Another factor that has recently received research attention is that lack of sleep results in hormonal changes that also correlates with obesity, either through increased appetite or decreased physical activity. Getting a good-night's=sleep is one way to alter our hormonal status in a favorable way.

As evidence that genetics are not destiny is the fact that obesity has been steadily increasing over the last thirty years, and not just in the US but in many industrialized countries. There is a great deal of argument over what, exactly, is causing the increase. It does seem to correlate with an increase in overall caloric consumption. Others, however, point to computers and increased time spent on sedentary activities.

Also (and here is where I think Taubes gets the context wrong) focusing on the hormonal and metabolic differences among people may help explain why person A is overweight and person B is not, but it does not help person A maintain a healthier weight. For that we need to focus on those variable that can be changed. Because the body is a horrifically complex system, even though we can explore and understand many aspects of the system, in order to understand the net effect of making specific changes to one's behavior, these have to be studied clinically.

The clinical research shows various things. First – there is no significant difference among the various diets based upon macronutrient ratios. Adjusting macronutrients does not seem to have a significant net effect on weight control. Finding a way to control overall caloric intake does work. To be clear, nothing works for most people, since most people fail most attempts at long term weight control. But the variables that do seem to help are utilizing convenient methods for estimating caloric intake. Even just writing down what one eats every day helps. Also, healthy eating should, as much as possible, be the default behavior. Having low calorie healthful snacks around the house, rather than tempting calorie dense snacks, is helpful. Eating at restaurants that publish the caloric content of their meals also makes sense, but I'm not aware of any data on this yet. Meal replacements are helpful, probably because they help estimate caloric intake and help avoid accidentally overeating.

There is recent evidence for a strategy of eating only when their is initial hunger, which correlates with low blood sugar. The idea here is that overall caloric intake will be decreased if one reinforces the pattern of eating only when there is a sensation of hunger.

These types of strategies work for some people, but if we want to have a societal impact on weight then there probably has to be regulations that make default behavior more healthful. We don't need to take choice away from people – but just make the lower calorie or more healthful choice easier.

Also, there is no question that there are many health benefits to regular exercise. Exercise alone is probably not enough to maintain weight. But if one engages in regular scheduled exercise, that would override any inherent tendency to reduce calorie expenditure, either from reducing caloric intake, or just with age or a sedentary job or lifestyle. It doesn't matter how much your genetics tell you to exercise, if you stick to a schedule of regular exercise.

I don't think there is any simple, easy, or universal solution to the problem of weight control.  Reversing the obesity trends will likely take societal changes as well as a focus on the many small things that all help. Perhaps they can add up to a significant effect. I also think that a large part of moving in the right direction is to get away from the distraction of focusing on gimmicks. Ultimately, the low-carb diet is just another weight-loss gimmick. There may be some interesting basic science behind it  – but bottom line, it doesn't work.

70 Responses to “The Nihilism Diet”

  1. HB says:

    I’m interested in the way mr Novella or readers o/c view the term ‘exercise’. The connotation with ‘running on a threadmill’ always creeps up on me and there isn’t a thing that involves moving my body I could think of that annoys me more. Sometimes I get the idea that exercise means exhausting yourself, or at least my people pick that up.

    Is there evidence that (very) strenuous exercise actually benefits more than let’s say; getting on your bike to work, or taking a walk for an hour a day?

  2. Somite says:

    Excellent article. I’d like to mention the book “The end of overeating”. It discusses the research into the addictive properties of some foodstuffs and how the food industry actively researches how to induce more and more eating by its costumers. Sadly, this increased in consumption and invoking addiction require the high caloric density of processed foods.

    I was surprised to see how some skeptics tackled the “issue” of organic foods which are not harmful, but only possibly more expensive and seem to give a pass to the food industry calorie bombs which are literally killing people and actively seek to promote addiction.

    • If you’re talking about Kessler’s book, yes, it’s good. That said, besides “sweet,” he notes “salty” and “fatty” are the two other “hooks” used for our taste buds. And a low-carb fat that’s meat-high is going to have the “fat” hook, unless it’s super-lean meat.

      That said, I’m worried about using the word “addictive.” I don’t think straight sugar, even, and its insulin “hit,” is addictive in quite the same way as alcohol, nicotine, etc. Sure, both produce body chemistry changes. But the insulin change isn’t as targeted on the brain as are traditionally addictive substances.

  3. sunny says:

    If happy meals or fast food in general were the culprit then all our kids would be obese. Go to a fast food place near a high school at noon and observe the weight of the teens who come in. You will find skinny kids eating huge meals and the heavier and even obese kids eating less. Go figure! But the fast food bigots would have you believe it is as easy as getting rid of fast food.

    In theory it is true that anyone, even someone who is genetically predisposed to be obese could lose weight and achieve a “normal” weight. Proof of this is the German concentration camps. But in fact people with a gentic predisposition to be obese will be obese (short of being in a concentration camp) and the more they diet the harder their body works against them.

    One important fact in this discussion is that in the 1990’s the government revised the way they determine obesity and being overweight and overnight the number of people who were obese doubled. Those who want to use this as “proof” that obesity is rampant do so fully aware that they are misrepresenting the facts. In fact obesity has not statistically increased in the dozen or so years since that change and for men has decreased slightly.

    Another important consideration is the BMI. It is a totally arbitrary measure which fails horrible for tall, short or athletic people. It’s arbitrariness became quiet clear after it was implemented and the “perfect BMI range” had been set the empirical evidence proved that in all ranges it was healthier to be heavier then the so-called perfect BMI then it is to be lighter. It is called “the obesity paradox”. Recently there has been some movement to revise this so-called ideal BMI upwards to 27.5 which is the ideal BMI as determined empirically.

    • Max says:

      The military has been lowering its standards, and still finds a growing percentage of potential recruits who are too fat to serve.

      More than three-quarters of the nation’s 17- to 24-year-olds couldn’t serve in the military, even if they wanted to. They’re too fat, too sickly, too dumb, have too many kids, or have copped to using illegal drugs.

      The armed services are willing to grant waivers for some of those conditions – asthma, or a little bit of weed. But the military’s biggest concern is how big and how weak its potential recruits have become.

      “The major component of this is obesity,” Curt Gilroy, the Pentagon’s director of accessions, tells Army Times‘ William McMichael. “Kids are just not able to do push-ups… And they can’t do pull-ups. And they can’t run.”

      23 percent of 18- to 34-year-old are now obese, up from just six percent in 1987.

    • Sunny – you are misrepresenting the situation with the definition of obesity. There was a change in 1998, but when data are examined by the CDC they use a consistent definition across time.

      There is a consistent increase in overweight and obesity for the last few decades – not a flat line with a blip in 1998 as one might think from what you wrote:

      Here is a graph from the CDC:

      Also – everyone recognizes that BMI has significant limitations when applied to individuals. BMI is a measure of weight, not fat. But – most people with a high BMI do, in fact, have excess adiposity. So it’s a reasonable estimate for populations. Studies show that the increasing BMI over the last few decades is not due to Americans becoming more athletic – it’s because we have greater adiposity.

      But yes, when dealing with an individual you have to consider other factors (sex, race, muscular development).

      • Max says:

        In contrast, the prevalence of autism is constant, while the prevalence of diagnosed cases increased, correct?

      • sunny says:

        My goal in all cases is to know and speak the truth. I firmly believe that you cannot solve problems if you allow yourself to accept and repeat politically correct. So I would never intentionally misrepresent data or a situation. The simple fact is obesity has been redefined. In 1998, the U.S. Government changed the standards by which body mass index is measured. As a result, close to 30 million Americans were shifted from a government-approved weight to the overweight and obese category, without gaining an ounce.

        According to an American Medical Association report, 14.5 % of Americans in 1980 were obese, a total of 32,700,000 (based on a population of 226,000,000). The studies claim that the numbers of obese Americans have “doubled” in the past twenty years, this would mean there are now about 66 million of them. But thirty million of those fatties were created by a change in definition, so by the standards of the 1980 obesity percentage of 12.85 percent, there has been an actual decrease in obesity percentage since 1980.

        For whatever reason the CDC is complicit in the misinformation campaign. Like so many things in pure science there are special interest groups who benefit from spinning the data to support their particular belief system. When it comes to obesity there are few if any special interest groups with something to gain by providing the truth while there are many who will benefit from the statistical lie.

        Regarding your chart
        Does in fact show a jump in the period including 1998. A line graph like this one with 15 years between data points does indeed make it appear that there was a long slow increase. Was that intentional to hide the jump in 1998? Who knows but convenient never the less.

        Another point to consider is that obesity is genetic and some races are more prone to obesity then others. Coincidently the numbers of Hispanics and blacks over this same time period has increased dramatically. This accounts for most of the increase in obesity. However this is NOT the same as an actual increase in obesity. To put it simply had the numbers of both blacks and Hispanics decreased then the combined statistical obesity rate would have gone down and yet the actual obesity rate would have remained unchanged.

        The important thing to understand about BMI is that the “good BMI” was selected arbitrarily by non-scientists who had an agenda. They arbitrarily decided that a BMI between 20-25 was “good” i.e. “healthy”. Their bias was towards thin and thus their proclamation was a BMI that made you thin was “healthy”. Since the BMI was first anointed we have had the chance to gather empirical evidence and we now know that the ideal BMI in terms of health and longevity is 27.5. We also know that people who are “obese” are healthier then people who are skinny. An odd turn of events to be sure and one that grates on those whose bias favors skinny.

      • BMI killed JFK, too. The “second assassin” hit him in the front of the head with a book of CDC charts, and that’s why his head jerked the way it did.

      • tmac57 says:

        Sunny,do you have any sources that could be checked for your controversial claims such as “…the “good BMI” was selected arbitrarily by non-scientists who had an agenda.”
        and “For whatever reason the CDC is complicit in the misinformation campaign. Like so many things in pure science there are special interest groups who benefit from spinning the data to support their particular belief system.”
        That sounds like conspiracy theory thinking to me.Mistakes in science,I understand.Grand plots to make people obese by definition?…not so much.

      • sunny says:

        The obesity crisis is really a little bit of slight of hand. Between 1980 and 1988 the obesity rate increased by 60% but the actual change was much less. What actually happened in those 8 years is that the health officials who perform these studies and publish the reports decided to change the racial makeup of the participants in the study to more closely reflect the changing diversity of the U.S. Earlier surveys didn’t reflect much racial diversity so when they were redesigned, selected population groups were suddenly disproportionately sampled in large numbers: children under 5 years, elderly over 60, and minorities. Black and Mexican-Americans, which according to the CDC are twice as likely to be “overweight” than whites, each now represent 30% of the total sample populations. The second change occurred in 1998 when they changed the criteria for obesity and overnight created another 45% increase in obesity without anyone gaining a pound. So between 1980 and 1999 the CDC manipulated the method of measuring obesity and created a 130% increase in the rate of obesity.

        Now for the good news: we’ve seen modest increases in actual heights and weights among the population for half a century, during that same period, life expectancies have increased to the longest in our history, and age-adjusted rates for chronic diseases have dramatically fallen. More good news: the CDC reported in 2007 that “There has been no statistically significant change in obesity rates for the past 7 years.” I guess what they mean by that is if you don’t change the criteria and/or the sample population there is no change!!! Go figure!

      • sunny says:

        The origin of the BMI came from a Belgian mathematician Adolphe Quetelet (1796-1874). He published a book in 1846 on probability and social science that demonstrated as diverse a collection of human measurements as the heights of French conscripts and the chest circumferences of Scottish soldiers could be taken as approximately normally distributed.” The BMI was derived as a gross approximation to a set of observed weights and heights of recruits in a Western European army in the middle of the 19th century. In other words the BMI was an arbitrary statistical measure derived over 150 years ago, most likely, from a group of young men to determine the proper weight for men and women of any age. Mr Quetelet, being a mathematician, choose the formula for a parabola. If he had been a chef he might have used a recipe for béchamel sauce. One of the problems with the formula is it results in weight that decrease in relation to the height. This was not a problem in 1846 when most people were quite short, mostly because of inadequate diet, but it is a major problem when applied to people in the Western world who are relatively tall thanks to an adequate diet. At any rate the formula for BMI was arbitrary and based on a sample during a time period when a deficient diet was the norm.

        Here You can read between the lines as scientists admit that the so-called healthy or normal BMI had an unusually high cluster of sick people while the so-called overweight and obese BMI had an even larger representation of healthy people.

        Two more sources to show the uselessness of BMI in evaluating health:

      • Max says:

        Ok sunny, let’s read between the lines.

        Your selected paper says, “Among US adults 20 years and older, 23.5% (approximately 16.3 million adults) of normal-weight adults were metabolically abnormal, whereas 51.3% (approximately 35.9 million adults) of overweight adults and 31.7% (approximately 19.5 million adults) of obese adults were metabolically healthy.”

        Reading between the lines, I’ll assume that abnormal means not healthy.
        So 23.5% of normal-weight adults were abnormal, 48.7% of overweight adults were abnormal, and 68.3% of obese adults were abnormal. Yeah, there’s no correlation at all between BMI and health.

      • Max says:

        Oh, and that paper counts all BMI under 25 as normal-weight. The standard charts say that BMI under 18.5 is underweight, and only 18.5-25 is normal weight.

      • Sunny – you are grossly misrepresenting the data, and I recognize your prose from some ideological websites.

        The redefinition in 1998 applied to the NIH, and it did not change the range for obesity, only the overweight category. So that does not apply to obesity at all.

        The CDC figures over time use a consistent definition of BMI for the categories of overweight an obesity – so this is a red herring.

        Also there is data by race showing an increase in obesity – so again, the race issue is a red herring.

        These points are useful for sowing doubt, but not understanding the data.

        Further – the data has been replicated in other countries that are also getting more obese. The WHO uses a the current definitions, and did not change them (in 1998 the NIH adopted the WHO definition).

        There are thousands of studies linking obesity to increased risk of disease, morbidity, and mortality. The studies you refer to about overall mortality show that the highest mortality is in the obese category, then underweight, then normal weight, then overweight. So the overweight category did have the lowest mortality. This may be due to the fact that some people in the normal weight category have lost weight due to illness. But it probably also reflects the fact that being a little overweight is not a health problem.

        But this should not in any way confuse the data for obesity, which is clearly a medical problem.

        Your argument about diabetes is confused. Type II DM has a strong genetic predisposition, but manifests much more often with greater morbidity due to overweight and obesity. The predisposition does not go away, but many people can be under control without other treatment and avoid all the morbidity by weight control. Whether or not you call this a “cure” is irrelevant.

      • sunny says:

        I won’t bother simply restating what I already wrote but I clearly showed that indeed the change in race made an enormous difference in the resultant obesity rate. prior to the 1990’s almost 100% of the sample population were people of European heritage and when they made the change to reflect the diversity of the nation they made 30% of that sample black or hispanic whose obesity rate is roughly twice that of people of European descent.

        My point on diabetes is that the rate has not changed. What has changed is :1) we are better at identifying it early. 2)We have greater diversity of races that have a higher rate of diabetes. These two factors have allowed some to spin the theory that diabetes is on the rise. There is no diabetes epidemic and never was.

        The trick to relating diabetes to obesity is which is the cause and which is the effect. many in the medical community do not care. They believe that being overweight is bad and therefor if they scare you with a little white lie and you “eat healthier” then it is a good thing. I won’t try to argue about the morality of that I simply am interested in the truth. For most people with diabetes obesity is a symptom of their disease and not the cause. They were diabetic when they were 10 years old and looked like any other 10 year old; they were diabetic when they were 20 and got married but were more or less normal weight; and they were diabetic when they first discovered at age 35 when they were 50 lbs overweight and unable to lose it no matter what they did. Understandably the overweight diabetic and the doctor might both believe they simply “caught” diabetes from their body fat. I am thankful that our doctors spend years learning medicine and apparently zero time taking statistic classes. I don’t need them to understand how statistics work or don’t work but I do want them to be able to treat me and others when we are ill or injured. In summary you get fat from diabetes you do not get diabetes from your fat.

        You are absolutely correct that a diabetic can control their symptoms through diet and excercise (and of course in many cases insulin). But there is no cure! And that is important. We need a cure and if we want to find a cure then we need to know the facts about the disease and ignore the popular mythology.

      • Max says:


        One of the classic symptoms of diabetes is unexplained weight loss despite increased hunger. It’s right up there with increased thirst and frequent urination.

        Where’s your evidence that diabetes causes weight gain rather than weight loss? Next you’ll tell us it causes infrequent urination.

      • There are lots of reasons to infer that obesity causes Type II DM, rather than the other way around. For example, if people get gastric bypass and lose weight, their DM gets better.

        There is also a lot of basic science to support this. We’re not just guessing.

        Sunny seems to think that medical scientists are just guessing, and then he substitutes his own unsubstantiated guess.

        You also failed to respond to my point about confusing overweight and obesity, and CDC data using a consistent definition. I believe I blew your primary point out of the water – a complete lack of response from you on this is not very intellectually honest.

      • Steve, thanks for noting that Sunny is repeating canned information, sowing red herrings, and doesn’t understand Type II diabetes. And, below, Sunny, the rate of Type II diabetes has indeed changed. Doorknob, are you more accidentally ignorant, or deliberately ignorant?

        Oh, and Sunny? It’s “sleight,” not “slight.”

        And, this “I come to speak truth”? Well poisoning.

    • Somite says:

      Have we found the obesity denier?

      • Somite, if you commented at all on Shermer’s JFK blog post, you already know the answer.

        I’ll agree myself that BMI isn’t perfect. However, both obesity, measurable by a variety of techniques, and related illnesses, above all, Type II diabetes (which we used to call “adult onset diabetes but CAN’T anymore) have increased notably over the last couple of decades.

        Please, Steve Novella, do an autism post just so we can see what Sunny says.

      • sunny says:

        If you paid attention you would have read that the so-called obesity epidemic was a result of changes in the methodology of the survey and not as a result of actual increase in obesity rates. Reread it and maybe you will understand.

        Diabetes is genetic and the rate of diabetes has remained constant. However there are more blacks and hispanics in the population today then there was 20 years ago and they have a greater rate of diabetes and including them in the calculation does indeed make it appear that the rate has gone up. Another factor was that in 2008 the medical community made a major effort to identify undiagnosed diabetes which was successful in that they identified about 2 million of them. Following this the data was used to claim that there had been an actual increase in diabetes when in fact there had not been.

        Autism! Sure, I’ll play. It is a fact that since the government began paying people for being diagnosed with autism (SSI) the rate of diagnosed autism has gone up. Almost simultaneously the government began paying schools for diagnosing autism and they too have discovered new cases of autism. The science hasn’t caught up with the lawyers yet so until it does I will wait for the truth. But the science is working hard to find something, anything that might support the idea that autism is on the rise and caused environmentally and so far zero proof.

        I will add something else for you to blindly disagree with: Obesity does NOT cause diabetes is is a symptom of diabetes. Most diabetes can be “controlled” (not cured or “prevented”) by diet and excercise. That is if you HAVE diabetes you can reduce the symptoms and even delay or prevent some of the most serious symptoms with a diabetic diet and excercise. You do not “catch” diabetes, it is genetic and you either have it or you do not. You may well be symptom free but that does NOT mean you do not have diabetes. Furthermore if you do not have diabetes you will not “catch” it from eating sugar, fast food, HFCS, or shopping at Walmart.

      • Max says:

        Where do you get that obesity is a symptom of diabetes? Weight loss is a symptom of diabetes.

      • Well, at least you’re not an antivaxxer. Of course, claiming autism is rising due to money-grubbing sounds sick. That said, I think whatever is included in the current definition of “autism” is on the rise, as well as questions about what “autism” is.

        “Asperger’s” was defined into existence by DSM-IV, on some of the “iffiness.”

        That said, autism as a legitimate syndrome, albeit still ill-defined, tracks parental age at time of a child’s birth – and both parents, not just mother.

        THAT is not an “artifact.” Science IS working right now.

  4. Chris Howard says:

    Nutrient dense calories, lots of fruits and veggies, carbs, some protien from lean meat, some cheese, butter, oils etc., exercise, H20, don’t smoke, drink in moderation. It’s not a guarantee, but what is? All we’re talking about here, is attempting to tip the odds in our favor. Sometimes, though, it just doesn’t work.

  5. Nyar says:

    “These types of strategies work for some people, but if we want to have a societal impact on weight then there probably has to be regulations that make default behavior more healthful.”

    So the Nannystate is the solution? If that is the case, then maybe the best thing to do about obesity is nothing.

  6. Brian says:

    Dr. Novella,

    One of the things that Taubes emphasizes is that the clinical studies that purport to evaluate low-carb diets are seriously flawed. The diets compared are either not truly low carb (say 30 percent carbohydrate rather than 10 percent) or compare low carb v. calorie reduced low fat. Or calorie-reduced low carb v. other calorie reduced diets. Studies like these contain multiple variables and cannot properly test the hypothesis.

    Since you have listened to Taubes’ lectures, I’m sure that you know that he has called for specific clinical trials to truly test the low-carb diet. He isn’t trying to rely on a simple interpretation of basic science. He is suggesting a trial that has not yet been conducted to properly test his hypothesis.

    A non-calorie-controlled intervention that compared a 10 percent carbohydrate diet to the standard American diet would be a good place to start.

    I understand that anecdotes do not constitute any sort of evidence. However, my personal experience with the diet caused me to change my opinion on it’s efficacy. I started out as a strong skeptic of the low-carb approach and gradually became convinced that it works.

    I began eating a non-calorie-controlled low-carb diet last November. In the course of three months I lost 20 pounds without hunger and without exercise. When I indulge at holidays or during business travel, my weight creeps upward slightly. When I resume low-carb eating, my weight effortlessly returns to 175 pounds. I had not weighed 175 in the 15 years prior to this change, despite various exercise and diet approaches to weight control.

    I have maintained this weight loss for more than 12 months by continuing to adhere to a low-carbohydrate eating program. Restriction of carbohydrates is the only significant change that I have made in my lifestyle over this time period (not entirely true, I have actually DECREASED my daily physical activity over the same time period).

    I understand that N=1 in this trial. And I suppose I could be accused of post-hoc reasoning, like the person who takes Vitamin C for his cold, has his symptoms resolve and then decides that the Vitamin C cured his cold.

    The difference is that there was a real, objective, measurable result (20 pounds of weight loss)as opposed to the subjective results that we normally see from placebo or flim-flam cures. I have not imagined the weight loss.

    It seems that those who are most convinced of the efficacy of low-carbohydrate diets are those who have had personal success with them. Taubes fits into this category. As do I and many others.

    Perhaps macro-nutrient ratio has nothing to do with our weight loss. But something has to be at work. Occam’s Razor suggests that the one major dietary change that I made in the past 12 months is the most likely cause.

    Or maybe I’m just another believer is woo. If that’s the case, so be it. Woo-woo weight loss is better than middle-age chubbiness.

    Many thanks,


    • Brian – You’re right – N=1 anecdotes are not helpful. I can give you mine, but what would be the point?

      Taubes is just engaging in more special pleading. The fact is – the clinical studies do not show that low-carb diets work or have any advantage over other diets with the same amount of calories. It is also not true that there are no clinical trials of very low carbs, and there are studies that control for calories. How many different ways to they need to study this question before the believers will be convinced. There is a remarkable consensus of the clinical data. Until someone provides a study that shows something different, there is no good reason to reject it.

      The safety of extremely low carb diets is also highly questionable, and short-term strategies are of no use to long term weight control. And besides, compliance rates long term are almost nil.

      • Brian says:

        Apologies for sharing the anecdote.

        I was incorrect. There has been a recent trial that incorporated a non-calorie-controlled low-carb intervention with a calorie restricted low-fat intervention. The trial was published in NEJM in 2008. Here is the link:

        Things to note: While not calorie-controlled, those in the low-carb group lost the most weight over the 2 year trial.

        Lipid panels improved for those on the low-carb intervention, despite (presumably) higher meat consumption.

        Gary Taubes has an interesting analysis of this article on his blog here:


  7. Brian says:

    Here is a somewhat different skeptical take on the book from the eSkeptic newsletter of Skeptic magazine:


    • Brian, Steven, all:

      How much do low-carb diets vs. “general” diets, or “general” diets vs. low-carb ones, distinguish between “simple” and “complex” carbs? And, “refined” vs. “unrefined” ones? Dietary fiber is not a cure-all, I know.

      That said, unless additional fiber is picked up from low-carb vegetable or legume sources, a low-carb diet is also going to be a low-fiber diet. In other words, if low-carb translates to high-meat, color me skeptical, Brian.

      • Oh, let me add one more thing. Many low-carb diet “toutings” are rooted in anecdotes about how much meat our ancestors ate. But, now that we know that Neanderthals ate cooked vegetables, I think there’s more reason to be skeptical about that claim, too.

      • Max says:

        As I recall, low carb diets don’t limit vegetables, they only limit grains and fruits.

      • Somite says:

        Just fruit. Nuts are very low carb as well but very caloric dense. It’s best if you just eat your veggies, ok?

      • tmac57 says:

        I think grains are pretty restricted on most of these type diets.Some complex carbs are allowed,but at a much reduced rate compared to usual recommendations.White bread and rice is usually not allowed.Potatoes and similar starchy veggies are usually out as well.

      • So I thought, but wanted to clarify. And, that doesn’t make sense either. Our ancestors, pre-agriculture but as hunter-gathering evolved more, added wild-grass ancestors of modern grains to their diets, along with various starchy tubers, which technically aren’t grains but which usually are verboten on low-carb diets.

      • Brian says:

        Socratic Gadfly–

        Grains would have played a very limited role in the diet of evolving hominids in the historic evolutionary environment. They would have essentially been a food of last resort due to their low calorie density/high effort of foraging. Tubers would have played a larger role but meat was a major component.

        (PDF after the jump).

        The Expensive Tissue Hypothesis posits that the evolution of the human brain was only possible once early hominids began to eat significant quantities of meat:

        Eating meat allowed for the simplification of the gut through evolution at the same time our brains were evolving.

        Perhaps meat isn’t that bad for us after all. If it was a major part of the evolutionary diet, it seems strange that it would pose health problems now.

        Finally, many who advocate against meat-eating are quick to point to T. Colin Campbell’s The China Study as evidence for their arguments. The following critique of Campbell’s work suggests that Campbell got it wrong. Worth a skeptical look.

        There is a good breakout on meat in particular among the links on this page. And a comprehensive demolishing of Campbell’s methodology in general.

        Please see link to 2008 NEJM article comparing a variety of diets in the comments above.

        Grist for the mill.

        Thanks all,


      • Brian, thanks for the info. It’s true that the wild grass equivalents of modern grains probably wouldn’t have supplied too much.

        That said, pre-domestication, I’m skeptical about the meat-eating levels of homo sapiens, based in part on what we know about modern hunter-gatherer peoples. Oh, I’m sure we ate more than homo erectus or earlier, but how much earlier, I don’t know.

        That said, given the URL of your link, promoting a high-meat diet, I suppose, color me skeptical on that ground, too.

        Finally, a high-meat diet seems to fit with Pop Ev Psychers touting man the noble savage hunter-gatherer, when actually, he started as a scavenger-gatherer.

      • I want to comment further on the “meat changed our brains” idea.

        As any real nutritionist, especially one who has an M.D., etc., knows, its perfectly easy to get all the protein, and all the individual essential amino acids, we need, on a vegetarian diet, or even a vegan one.

        And, not just John/Jane Doe types.

        Greco-Roman Olympians actually usually ate a high-carb diet, NOT a high-meat one.

        As for caloric intake, that said, hunter-gatherers were eating all the animal, including marrow, innards and other high-fat areas. So, to extrapolate from that (and the calories they burned) to benefits of a diet high in lean meat, with protein intake far beyond what is necessary, is dicey.

        I’m not saying no extrapolation is possible, but let’s distinguish between the high fatty products + meat diet of ancestors vs. an Atkins-type diet today.

        Let’s also HUGELY distinguish between the calorie expenditures, again, too.

  8. sunny says:

    Max: One of the indicators of cardiometabolic abnormalities is being overweight. This does not necessarily mean they are not healthy.

    Maybe I did not explain my point well enough. I was trying to not be long winded. So let me try again. By convention above BMI 25 to 30 is considered overweight. In general below BMI 20 to 15 is considered underweight. When you compare the relative health of being overweight to being underweight those who are overweight live longer and have fewer health problems. A similar relationship exists between those who are obese, that is a BMI greater then 30 and those who are seriously underweight, that is a BMI less then 15. In other words people who are obese live longer and are healthier then those who are “skinny”. It is because of this “obesity paradox” that scientists have determined that the healthiest BMI is 27.5 and that “normal” or healthy range for BMI is between 25 and 30 NOT between 20 and 25. This is the empirically derived “ideal” BMI range.

    • Max says:

      The paper you cited defined overweight as BMI 25-30, and defined cardiometabolic abnormality as things like elevated blood pressure and triglycerides. It then said that 23.5% of adults with BMI below 25 (including underweight) were abnormal, compared with 48.7% of overweight adults, which tells me that overweight is less healthy on average.

      • sunny says:

        True, that’s what it said. By definition if you are catagorized as overweight you have cardiometabolic abnormality. Since the paper did NOT say they were actually unhrealthy we are left to wonder. But I thought it was ironic that so many people with “normal” BMI’s also had cardiometabolic abnormality.

      • Max says:

        Well yeah, there are nonsmokers who develop lung cancer and smokers who live to 100, but on average the smokers are much worse off.
        Although I’ve never seen an obese 100-year-old.

      • Max says:

        Here’s a hypothesis you won’t like. Maybe so many Americans with normal BMI have cardiometabolic abnormality because they get empty calories and saturated fat from junk food, just not so much that they gain weight.
        Still, the abnormality rate is more than double among the overweight and almost triple among the obese, so this is a solid correlation between BMI and metabolic abnormality.

        For anyone else reading this, we’re talking about this paper:

      • sunny says:

        Max: I don’t think you have gotten the irony yet. Let me spell it out; One criteria for determining if you have a cardiometabolic abnormality is if you are overweight. So to say “so this is a solid correlation between BMI and metabolic abnormality” is kind of redundant.

      • Max says:

        The paper says: “Cardiometabolic abnormalities included elevated blood pressure; elevated levels of triglycerides, fasting plasma glucose, and C-reactive protein; elevated homeostasis model assessment of insulin resistance value; and low high-density lipoprotein cholesterol level.”

        I don’t see being overweight in there, do you?

      • tmac57 says:

        sunny,can’t you even be bothered to read your own posted sources? You have failed miserably here.
        Max,good job!

      • You obviously didn’t read Max’s earlier response basically demolishing you, Sunny. Overweight people had a higher percentage of metabolically abnormals, and obese a higher percentage yet.

        So, in this thread, you’ve proven you don’t know basic math, let alone basic statistics. You don’t know Type II diabetes.

        Please, Steve Novella, now that Wakefield’s been exposed as a fraud and liar, do an autism post!

    • Max says:

      They determined the healthiest BMI by averaging the skinny with the fat? Doesn’t sound right. Wouldn’t it make more sense to determine it by examining the healthiest people?

      • sunny says:

        No max they compared the groups on both sides of the “normal” to determine if it was healther to be overweight or underweight. They did this because the hype was that being skinny was healthier then being fat. But the data refuted that claim.

      • Max says:

        Even if it’s better to be overweight than underweight, it’s still best to be normal weight.

      • sunny says:

        Absolutely true! But the “normal” or best weight is not the original WAG of BMI 20-25 it is a weight range that centers around 27.5. This “revelation” came about because of the obesity paradox.

      • tmac57 says:

        One of the confounding things that might make ‘normal’ weight appear to be associated with decreased longevity is that there are quite a few illnesses associated with sudden unexplained weight loss,such as:
        # Cancer.
        # Gastrointestinal diseases.
        # Endocrine disorders.
        # Infection.
        # Cardiovascular and lung disease.
        # Neurological illness.
        # Kidney disease.
        # Mood or mental health changes. Anxiety, stress and depression can affect your weight.

        Looking at the raw numbers only will not tell you why someone weighs what they do.

      • sunny says:

        You are absolutely correct and you have stumbled upon one of the weaknesses fo statistics.

      • tmac57 says:

        So are you conceding the possibility that overweight people may not be actually healthier based on the longevity stats? The slight advantage may be due to normal weight people suffering from a non-weight related disease which causes them to lose weight.Right?

  9. Homer says:

    But the big question is, do I have to give up beer and donuts?

  10. If Sunny were a Poe, she’d be halfway funny. Instead, she’s just a mindset-entrenched idiot. The JFK issue was one thing, but it’s clear this mindset transcends one issue. Don’t argue with her, it’s just like a creationist or IDer.

    • sunny says:

      tmac57: Not exactly. I never considered obesity as being healthy. My major complaint was that obesity and overweight have been redefined for political reasons. I only used the data that was available to show that the rash decision to define half the population as overweight or obese and would all die young was in fact wrong. I am very satisfied with the trend to look at the empirical evidence and correct the error in the way we have used the BMI.

      • tmac57 says:

        You have stated at least twice in this thread that obese people are healthier than ‘skinny’ people,which Dr Novella pointed out was disputed by the very studies that you cited:

        There are thousands of studies linking obesity to increased risk of disease, morbidity, and mortality. The studies you refer to about overall mortality show that the highest mortality is in the obese category, then underweight, then normal weight, then overweight.

        My point above was that the rates of mortality in underweight and normal weight individuals may be confounded by diseases that cause them to be at a lower weight,rather than the lower weight being the cause of their increased mortality.

      • sunny says:

        That is not what I said. What I said was that when they did a real study the results were that obese people did live longer and were healthier then “skinny” people and overweight people lived longer and were healthier then underweight people. This was called the obesity paradox. As a result the BMI index is being reevaluated with a BMI of 27.5 as optimal.

        I think it is ironic that you argiue that it isn’t really being underweight that cause early death it might simply be that their “underweight” condition is the result of a serious ilness and THAT is what cause early death. True enough, but ironic in that you could make the exact same arguement about being overweight or obese. I do believe that both of these are true. However my main interest was in the BMI and how it is an ineffective tool and has proven to be incorrect. Perhaps with the new changes it will become more meaningful.

      • tmac57 says:

        “obese people did live longer and were healthier then(sic) “skinny” people ”
        What is your source for this specific assertion?

      • GoneWithTheWind says:

        In 2006 at an American Heart Association meeting a study was reported which found that fatter cardiac patients were more likely to survive hospitalization and invasive treatments than thinner ones, even when adjusting for age and other contributing factors. In this analysis of 130,139 heart disease patients, 5.4% of “normal” weight patients died, as compared to 2.4% of “obese” and 3.1% of “overweight.” Yes, those whose were “obese” were more than two times more likely to survive!
        There is considerable evidence that found that fatness is not a risk factor for heart disease or premature death. This was confirmed nearly half a century ago upon examining 16 prospective studies in seven countries, as well as actual angiographic and autopsy examinations of 23,000 sets of coronary arteries which found no relationship between body fatness and the degree or progression of atherosclerotic build-up. And the most careful studies ever since have continued to support these findings. Dr. Paul Ernsberger, of Case Western Reserve School of Medicine in Cleveland, Ohio, led a review of nearly 400 studies that was published in the Journal of Obesity and Weight Regulation in 1987 which corroborated these results. “The idea that fat strains the heart has no scientific basis,” he said. “As far as I can tell, the idea comes from diet books, not scientific books.
        In 2006 researchers at Cedars-Sinai Medical Center in Los Angeles reported their clinical study of 14,739 patients with coronary artery disease, that had been confirmed on tomography, who were followed for over three years. They found that “obese” and “overweight” patients were at significantly lower risks for cardiac death than “normal” weight patients.
        A study published in 2006 in the journal Hemodialysis International, led by Kamyar Kalantar-Zadeh, MD, PhD, MPH, of UCLA David Geffen School of Medicine. It reported that among dialysis patients, “obese” patients are far more likely to survive than smaller patients. Since dialysis patients have protein-energy malnutrition and inflammation, termed Kidney Disease Wasting, obesity probably represents better overall nutrition and protective reserves that lower their risk of death, said Dr. Kalantar-Zadeh. According to these doctors, the popular belief that fatness is associated with heart disease among these patients has not been shown in any study, nor is the survival advantage of higher BMIs (body mass index) related to having greater muscle mass over fat. Fatness isn’t the only “paradoxical” association among favorable clinical outcomes of dialysis patients, they said in a 2005 issue of the American Journal of Clinical Nutrition:
        “High concentrations of total cholesterol have been associated with both a survival advantage in these patients, as has an inverse relation between blood pressure and outcome. These consistent findings across an array of cardiovascular risk factors in dialysis patients support the more inclusive term “reverse epidemiology.” “Reverse epidemiology has also been observed in heart failure patients, elderly persons, and patients with advanced malignancies, AIDS, and other chronic diseases. This means that 20 million persons—including almost half a million dialysis patients—in the United States alone may be subject to this reverse epidemiology. We believe this vulnerability to reverse epidemiology could have very important implications for public advice on health matters, because conventional recommendations for the management of CVD risk factors, such as weight reduction or aggressive treatment of hypercholesterolemia, may not be appropriate.

        It’s probably shocking for some to hear that there even ARE health benefits to being fat. But as these doctors noted, kidney disease isn’t the only health problem where studies have shown that being fat appears protective and beneficial, especially as we age. It also includes infections, cancer, lung disease, heart disease, osteoporosis, anemia, high blood pressure, rheumatoid arthritis and type 2 diabetes.
        “Both U.S. government and Swedish studies indicate that fatness is not associated with increased doctor visits, medical procedures or hospitalizations…Fat men took no more days of sick leave…and were no more likely to have work-limiting health conditions,” wrote Drs. Paul Ernsberger and Paul Haskew in The Journal of Obesity and Weight Regulations. Among those over age 60, “‘obese’ men and women are no more likely to suffer from chronic diseases than ‘average-weight’” persons.” But they are more likely to survive longer.
        The Seven Country Study, for example, which has followed 13,000 men over the last 40 years, has found that the risks of dying from cancer and infections decrease with increasing weight. In long-term prospective studies, complications (like retinopathy) and mortality rates from type 2 diabetes are three times lower among heavier people. And people are much more likely to survive a hospitalization if they’re “overweight” than if they’re thin.
        UCLA researchers analyzed the Acute Decompensated Heart Failure National Registry and the more than 108,000 cases of acute heart failure that had been hospitalized in 263 hospitals across the country from October 2001 through December 2004. Despite adjusting for every contributing factor, age, gender, lab work and other health indice, they found that for every 5 unit increase in body mass index (BMI) the risk of dying dropped by 10 percent.
        Actual mortality rates for the “normal” weight patients was 6.3 compared with 2.4 for the most “obese.” This is not a fluke finding, as the researchers specifically note that “prior studies on chronic heart failure have demonstrated that body mass index is inversely associated with mortality, the so-called obesity paradox.”

      • tmac57 says:

        GWTW-This study from the Lancet ‘Body-mass index and cause-specific mortality in 900 000 adults: collaborative analyses of 57 prospective studies’ seems to directly contradict many of your assertions.Their findings show that a BMI of 22.5 – 25 was the apparent optimum,and that BMI was a strong predictor of overall mortality.As BMI goes up ,so do mortality rates.

  11. GoneWithTheWind says:

    What you cited was a “conclusion” from a collaborative analyses of 57 prospective studies. In general “studies” is code for a data dredge. A data dredge is where a special interest group looks at all the data from various sources in an attempt to find something that supports their preconceived idea. They throw out anything that contradicts their “theory” and draw conclusions on trends “dredged” up from disparate data. To make matters worse many of these data sources are unscientific in that they are already biased by how the data was collected or the population the data was collected from. So essentially what you have is conclusions made by questionable individuals based on questionable data which generally was collected improperly. Worse, (could it be worse) in these “studies” or the “analyses” of these studies there is no raw data which would allow a unbiased person to attempt to duplicate the results. The best that I can say for this is that I am highly skeptical of their analysis. You are free to decide this for yourself and I suspect you already have.

  12. tmac57 says:

    GWTW said

    “In general “studies” is code for a data dredge. A data dredge is where a special interest group looks at all the data from various sources in an attempt to find something that supports their preconceived idea.”

    So just for fun,I went back to GWTW’s last post to see how many “data dredges” were cited:

    “In 2006 at an American Heart Association meeting a study…”

    “And the most careful studies ever since have continued to support these findings.”

    “And the most careful studies ever since have continued to support these findings.”

    “…led a review of nearly 400 studies …”

    “Los Angeles reported their clinical study of 14,739 patients…”

    “A study published in 2006 in the journal Hemodialysis International…”

    “…disease among these patients has not been shown in any study…”

    “But as these doctors noted, kidney disease isn’t the only health problem where studies have shown…”

    “Both U.S. government and Swedish studies indicate that …”

    “The Seven Country Study, for example…”

    “In long-term prospective studies, complications …”

    “…the researchers specifically note that “prior studies on chronic heart failure…”

    So GWTW,what special interest group are you shilling for,and what is your preconceived idea?

  13. GoneWithTheWind says:

    Good work. You can cut and paste. What special interest group do you work for?

  14. ap says:

    My evidence is only anecdotal but I’d like to dispute the claim that doing exercise leads to increase energy (i.e. food) consumption–provided you are not an olympic athlete and are already consuming between 2-3000 calories/day. I am a lifetime athlete who has competed both nationally and internationally. Every athlete I speak to says the same thing. When we are in training we “naturally” want to eat healthy foods in appropriate amounts: in fact it what we crave. We DO not crave sweets or fats. When we are slacking, or injured, or not training for whatever reason, after about two weeks we start to eat progressively more food with progressively worse choices (i.e. ice cream, cake, etc….).
    I’ve heard that some research indicates that free radicals curb appetite. Free radicals are a byproduct of exercise. Maybe this has something to do with my anecdotal evidence that exercise actually helps to control appetite and cravings for “bad” foods… Thoughts?