A Response to Denise Minger – Part 1
by Garth Davis, MD on May 03, 2016“Brevity is the Soul of Wit”- Shakespeare
“Not”- Garth Davis and Denise Minger
(In Other Words, This Is Long)
PART I
In the interest of time I have decided to break down Denise’s review into parts and will publish them separately. I also do not think anybody is going to read an extremely long winded response. So every few weeks, as time permits, I will further address her critique page by page.
I have gotten some very interesting responses to Denise Minger’s review of my book. Many have suggested that I simply ignore her and not give her any acknowledgement. When it comes to nutritional science the blogosphere can be ridiculous. I often ignore ad hominem attacks thrown my way because, as they say, arguing with an idiot is ill advised because they will drag you down to their level and beat you with experience.
Denise, however, is certainly not an idiot. She may not have a PhD or MD, but she is far more knowledgeable than most physicians I know. I, like Denise, am a bit of a nerd about nutritional science, and I have to say that I love her blog.
She claims to read nutritional literature 5-7 hours a day. I simply do not have that kind of time, so I enjoy learning from her literature searches. She defends her abilities here.
She reminds me of another non non-physician, non-PhD blogger, who publishes anonymously at PlantPositive.com, who like Denise does an excellent job of reviewing and synthesizing complex and huge amounts of data in spite of lacking the classic credentials. (In fact, I am going to use his responses to Denise in a later part of my rebuttal.)
I agree with Dr. T. Colin Campbell, who said in his respectful rebuttal to Denise’s criticism of the China Study, “I am the first to admit that background and academic credentials are certainly not everything, and many interesting discoveries and contributions have been made by “outsiders” or newcomers in various fields.”
While I mainly read the scientific literature, I also love looking through blogs and social media of a variety of people, some with credentials and some without. Often I find that some of the least experienced offer some of the best assessments and critiques of the literature. In fact, most of what I read on the internet comes from people that have a very different view of nutritional science than I do. I find myself spending much of my free time reading low carb blogs, believe it or not.
Interestingly, I don’t view Denise’s blog as my antithesis. Denise and I agree far more than we disagree. Denise eats a predominately plant based diet with some meat, appropriately sourced. This is exactly what I prescribe for my patients.
She also bravely spoke before a Paleo diet convention and pointed out the benefits of low fat diets and acknowledged, as I have before, the magic results in the extremes of diet (ie vegan/very low fat vs very low carb). While she is championed by many low carbers and paleo followers, her opinions seem to not be rooted in dogma, which I greatly respect.
I have decided to go into depth on her critique because I want to be sure I made no crucial mistakes in writing this book, and if I did I would like to correct them. If you have read Proteinaholic, you know that in a previous book I supported a high protein diet and regretted that bad advice.
I tried to be as truthful and unbiased as possible when writing this book. I was petrified to make any false claims as I truly want to provide people with the best health advice possible. I reviewed thousands of articles for this book and I knew there would be mistakes.
I asked editors and contributing author Howard Jacobson to go through the book with a fine tooth comb, but the volume of articles referenced made the possibility of mistakes inevitable. In fact, on our website, we included a page for corrections in case my views changed or I found glaring errors. Up to this point, we had found few errors, mainly editing and not content. Denise offered the first content-based challenges, and I feel obligated to address them.
I also wanted to read her views with open mind to be sure I had not let bias get in the way of good science. Those of you who have followed my work know that I abhor bias in science. In Proteinaholic, I go into details about the role of industry funding in the corruption of scientific literature, but bias does not only come in the form of money.
While I am not funded by the big veggie lobby, I have become vegan. In other words, I do not eat animals, not just because I think it is healthier, but also because of the effects of consuming animals on the environment, not to mention the ethical implications of eating meat. It is therefore feasible to say that I am subject to a bias of sorts.
I initially changed my diet due to concerns for health. My hope for this book was to share what I learned about the health effects of animal protein. I have learned much about the nutritional benefits of plant based diets, but my goal was not to make people go vegan. When I counsel patients in my office I do not let my ethics affect my medical advice. As such, I advise that the majority of the diet be plant food, but I allow for some meat consumption, thereby steering my patients to a diet very similar to Denise’s.
Shall We Begin?
I will go through Denise’s criticisms one by one in systematic fashion, but first let me give an overview.
I wrote this book after doing an extensive amount of study as to the optimal diet. If you have read Proteinaholic you know, and Denise agrees, that I did an enormous amount of research.
My goal was NOT to say that protein is not needed, nor was it to say you must be vegan. Rather, I have come to the conclusion that a predominately plant based diet is optimal for human health, but when I suggest people eat more plants I am constantly faced with the inevitable question, “But where do I get protein?”
As I point out in Proteinaholic, people are constantly looking for more and more protein. Food companies now advertise protein like crazy, knowing that it will lure people in. NHANES (National Health and Nutrition Examination Survey) data shows that we are not even close to having diets deficient in protein, yet we are grievously lacking in fiber. On top of that, and despite our relatively high protein consumption, we continue to get sicker and fatter. This is obviously not only due to protein consumption, but why would that individual macronutrient get so much glory?
I have mentioned that Denise’s lack of experience does not discredit her but, as Dr. Campbell pointed out, “I am the first to admit that background and academic credentials are certainly not everything, and many interesting discoveries and contributions have been made by “outsiders” or newcomers in various fields. On the other hand, background, time in the field, and especially peer review, all do give one a kind of perspective and insight that is, in my experience, not attainable in any other way.”
To take it even further, experience in a clinical setting is the real testing ground for any nutritional hypothesis. Even the most respected PhD researchers often miss the forest for the trees.
I wrote this book because I had extensive experience treating people suffering from obesity and diseases of western civilization. I have used low carb, high protein, ketogenic diets extensively in the past, and still use them today prior to surgery. I have seen how people respond and how often they fail.
My reputation depends on success and I have seen the huge benefits of a plant based diet, and the amazing successes my patients achieve when adapting a plant based lifestyle.
Patients who heed my advice have incredible weight loss, but you can lose weight on any diet. They also have incredible improvement on markers of disease, but you can see this too on many low carb diets.
But here’s the thing: as a doctor, seeing real patients, there’s more to the story than just numbers. My plant-based patients lose weight more easily. They aren’t hungry. They don’t feel deprived. And they radiate a vitality that can’t be reduced to lost pounds or A1C levels or lipid profiles.
So you can imagine my utter frustration when I give someone a plant based diet (NOT vegan) and then they come back and their diet log is filled with the same old meat and dairy. When I ask them why they chose beef jerky as a snack instead of the suggested apple, they invariably respond that they needed to get their protein! It was my daily experiences in the real world, treating real people, that led me to write this book.
Denise seems to lose this narrative theme throughout her discussion. At first, I was quite impressed with the depth of Denise’s research; however, I came to find that she was recycling many of her previous blog posts. She has a limited number of boilerplate rebuttals to common arguments pertaining to different dietary lifestyles, and she applied these, often out of context, to her critique of Proteinaholic.
As a prior raw vegan, she has researched the common arguments, both pro and con, about paleo vs vegan diets. In her review of my book she seems to gloss over much of the book and focuses only on those familiar arguments, recycling her canned answers from prior blog posts.
Each of the “disease chapters” in my book was written according to a precise formula. I first wanted to see if we can take a disease process, for example hypertension, and see if there is epidemiologic evidence as to the effect that animal protein may have on that disease process.
As Denise loves to point out, correlation is not causation, but I tried to show that there was not just one study but multiple epidemiologic studies that would demonstrate the relationship, suggesting that where there is smoke there is fire. I would then try and establish a mechanism of action.
In other words, if I found that population studies show that people on a plant based diet tend to have lower blood pressure, I would then have to prove that there was some biologic mechanism at play that can explain such an effect. So in the hypertension section I showed several epidemiologic studies, then followed with evidence that fiber can reduce blood pressure, and that glutamic acid, abundant in plants, may lower blood pressure. I followed up by sharing the results of randomized control trials to test the theory that the delineated mechanism did in fact explain the epidemiologic results.
I also emphasized in the book that no single study can support or debunk a hypothesis. Denise, as she did in her critique of the China Study, concentrates on isolated studies, thereby missing the whole structure of the argument. She criticizes a few studies in Proteinaholic from a huge bibliography and then concludes that “avoiding animal protein in order to get healthier has virtually no supportive evidence.”
What?
I provided long term huge epidemiologic studies. I showed randomized control trials that demonstrated that avoiding animal protein reduces hypertension, diabetes, heart disease, and certain cancers. I show studies that demonstrate that avoiding animal protein can actually change your chromosomes and reverse heart disease.
My thesis was not built on the dozen or so studies that Denise highlights, but rather on thousands of scientific articles that address population studies, basic science, and randomized control trials. Did she miss all of that? Does her problems with a few select studies, which I will address, completely debunk years of study? While she writes a very long rebuttal, the subject matter she addresses is a very small part of the book.
I might also add that Denise seems to miss another theme of Proteinaholic in that she seems to believe that the book is … “stoking the flames of our next macronutrient centric shift, bringing a new dietary villain to the fore.” She further states that she was expecting that the “world would eventually turn its dietary lynch mob on protein.” My goal was NOT to create a new macro nutrient fear, but rather to demonstrate that reliance on macronutrient breakdowns is erroneous. As I state in my introduction:
“Obviously we need protein. My concern is more with the fact that we no longer talk about food as food. Rather we are obsessed with breaking food down to its component parts and, in so doing, have developed an unhealthy obsession for one particular macronutrient. . . Believe me, I am not writing this book because I dare to be different and buck the norm. I am not looking to be sensational, and I certainly hate to further confuse the public. However, given my experience, I am in a unique position to see that we have missed the forest for the trees. . . . we may nod our heads when experts argue that you should eat whole foods like a carrot or an apple, but in the back of our mind we are still questioning the protein content of the apple. Given this calculus we would rather eat a chicken breast or a protein shake. This incorrect emphasis on protein results in the fact that we only eat about 5-7% of our calories from fruits and vegetables!
“We should be talking more about whole foods. The reductionist practice of breaking foods down to its component parts has really confused the situation. However, during the writing of this book I must be a reductionist. I cannot argue that the emphasis on protein is harmful if I do not roll up my sleeves and immerse myself into the world of reductionist nutrition. I need to take on protein mano-a-mano. Otherwise, no matter how much I tell you to eat an apple, you will always reach for the Beef Jerky –now available absolutely everywhere. More importantly, I also have to show you that there is a dramatic difference between the physiologic effects of plant vs animal protein.”
You see that right from the start of the book, I have to discuss macronutrients in order to demonstrate that reliance on such nutritional reductionism is harmful.
However – to suggest that the book is part of a “lynch mob” on protein is not accurate; Proteinaholic is not a vegan manifesto. I describe the health benefits of a vegan diet mainly to rebut the charges that plant based diets cannot provide sufficient nutrients, and that plants are inferior sources of protein.
A Section By Section Critique
RDA Confusion
Denise begins by pointing out that the RDA for protein is “the minimal amount to meet the nutrient requirement, rather than the optimal”, which is what I claim in Proteinaholic. This is a bit of semantics, but I hold by my statements, though a slight rewording may be called for.
The Food and Nutrition Board of the US National Academy of Sciences arrived at their protein recommendation after reviewing many different studies. They have re-reviewed and revised their recommendations several times. They calculate recommended protein consumption based on the amount of protein that is lost daily and needs to be replaced. The “obligatory losses” such as found in stool, urine, feces, skin and hair were estimated to be, in the average male, 24 gm of protein a day.
The RDA is designed to ensure that the vast majority of people eat adequate protein, so they assumed a normal curve, calculated standard deviation, and then added 2x standard deviation.
In other words, the average man’s protein requirement is right at the top of the graph, at the end of the red arrow. Half of the male population requires less protein (the area to the left of that center line), and half requires more.
So what does it mean to add a “standard deviation” to that requirement? Without getting too technical, you can see standard deviations in the graph above. Each vertical segment represents one standard deviation. So let’s look at the graph again, this time with the RDA set at two standard deviations above the average:
The percentages within each vertical segment refer to the percentage of American males whose protein requirements are matched by that segment. In other words, the two dark blue segments, representing 68.2% of the US male population, are within one standard deviation of the average.
By placing the protein RDA two standard deviations above the average, the Food and Nutrition Board is recommending a daily protein intake that is actually higher than that required by 97.8% of the population.
This brings us to 31 gm of protein as being the most amount of protein lost per day by any given person. They then estimated that only a certain percentage of dietary protein is actually absorbed and utilized, which is how they arrived at the 56 gm recommendation for an average male. But this is NOT a minimum value. If you ate 50 gm of protein per day you would more than likely be fine. The 56 gm is a cushion to be sure nobody is truly deficient.
Now is it an optimum value, as I claim? I can see Denise’s point that it is not. The RDA is meant to be enough protein, which implies it is a minimal number, and for a population as a whole, I grant she is correct. However, that 56 gms of protein assures us that 98% (3 sigma rule) of the population is getting enough protein which is optimal for covering our bases. Or, explained differently, statistically you have a 98% chance that you can eat less than 56 gm of protein and yet still be consuming adequate amounts.
As I point out in my book, there have been numerous reviews that have demonstrated these values to be more than adequate: http://www.ncbi.nlm.nih.gov/pubmed/12499330 and http://www.ncbi.nlm.nih.gov/pubmed/18996869 are two such examples.
The RDA never attempts to delineate an upper limit of protein. Nor does it consider what constitutes an optimal level of protein. Built into the RDA is an implicit assumption that more, if not better, is definitely not worse.
Which is where Proteinaholic comes in. The whole point of my book is that there may be risks with higher levels of animal protein. When we perform a risk to benefits analysis, 56 gm consumption has very little risks and yet is completely adequate for day to day life, and is therefore an optimal level. I would not argue that it is optimal level for an athlete, but that is a whole other discussion.
2. Lester Morrison: Low Protein Pioneer or Misrepresented Researcher?
To be honest, when I read this section of Denise’s critique, I thought to myself, “Who the hell is Lester Morrision?” I could not remember studying Lester Morrison at all. When I looked at the section of the book she was referencing I realized I had never reviewed Lester Morrison’s work, and was actually discussing him as an influence on Nathan Pritikin. I never actually read the study and so did not reference it in the book but alluded to his study as I had read it discussed in Pritikin’s book.
I find it strange that this is the number two criticism of Denise’s and is demonstrative of her focusing on small issues and missing the big picture. This study was not part of my overall thesis and could have just as easy been left out of the book. That being said, for purposes of this response, I attained the actual study, at a cost of $30. Thanks Denise 🙁
This was not a great study, but Denise is right in that Morrison’s main arguments were that reduced fat diets could reduce heart disease. However, she claims that his low fat group was eating 120 gm/day of protein.
Maybe we are reading different things but his study explicitly states that his experimental group were eating 90 gm of protein and 320 gm of carbs. Denise lists various foods that were allowed, which makes it look like this was a high protein, low carb diet, but that just was not true.
Example diet given was: Breakfast- shredded wheat with skim milk, wheat toast and margarine, and orange juice. Lunch- pea soup with skim milk, 2 oz of meat or cheese (lamb), string beans, salad, wheat toast and canned pineapple, and dinner could be 2 oz meat, salad, veggies, bread (again) and lemon sherbet. Hardly high protein.
Secondly, they limited the fat to 25 gm and limited cholesterol intake to 50-70 mg. They don’t give us details on the actual diet of the controls but we are told it was their usual, customary diet and consisted of 80-160 gm fat and 50-70 mg cholesterol. They don’t tell us the usual protein BUT understand that cholesterol comes exclusively from animals.
Also most of the fat consumed at that time came from animals. So, it would be extremely difficult to reduce cholesterol and fat without reducing animal protein. I suppose if they ate a lot of egg whites and skim milk, but that is not what their sample diet looked like.
Several times in Proteinaholic I alluded to the fact that it is difficult to separate protein from fat to see an isolated cause effect relationship. I grant Denise that Lester was interested in fat, but the 90 gm of protein would hardly be classified as high protein, and I am not privy in this JAMA study from 1960 to the actual diet of the control group, but given the much higher cholesterol intake I must assume they ate more animal protein.
Nonetheless, I had not subjected this study to the same rigorous investigation I did the actual studies used in the formulations of my arguments. Denise is correct in pointing out that the study was designed to look at lower fat NOT lower protein.
3. Do Nonhuman Primates Prove We Can Thrive on Low Protein Diets?
This is a famous argument in the nutritional science world. Specifically, does looking at the anatomy of various animals and comparing them to our own gives us any information on what we should eat? Denise challenged my agreement with Sylvester Graham’s assessment that orangutans and gorillas prove you don’t have to eat meat or dairy to build muscle.
We can write a whole other book on this isolated debate. First off, the section where I discuss comparative anatomy was used to show that, despite the common argument by meat eaters, we are not carnivores. I provide multiple examples of how we are nothing like carnivores, from our acid levels to our enzymes to our modes of mastication, etc. (read more in Proteinaholic).
Denise spends some time arguing that we are very different to apes because we have more small intestine and less colon, and she is correct in fact if not in intepretation.
My argument is our physiology is far more like a chimp than it is a lion. If you look at small intestine length compared to body length a carnivore has 3-6x the size of small intestine, the herbivore has 10-12x, the usual omnivore has 4-6x, and we have 10-11x. Comparing carnivores, herbivores and fruit eaters, we are far more like the fruit eaters.
I am not sure where Denise is going with the hindgut theory, because we too are hindgut fermenters. Maybe not as efficient as an ape, but hindgut fermentation is mainly good for production of volatile short chain fatty acids that can be absorbed and utilized as energy.
Here’s the key problem with her theory: hindgut fermenters do not absorb more protein. Even humans can have greatly varying sizes of colons, which may be genetic or adaptive based on diet, but you wouldn’t say one human is completely different from another based on size of the colon.
Denise does point out a valid fact that chimps and apes graze all day and consume huge amounts of food and therefore can get more protein but she seems to imply that they eat a lot of meat and insects. Jane Goodall was the first to note that chimps do organize hunts and that they use sticks to “fish” for termites, but she noted this was a small amount of what they eat.
In fact, observations of primates in the Congo show that they gravitate far more towards fruit and are NOT driven by protein consumption. When fruits are plentiful Apes will mainly eat fruit, especially figs. In Uganda, while they will eat 102 different species of plants, half of all food consumption comes from varieties of figs. Their percentage of protein consumption will drop down to 15% or less during plentiful fruit availability. When fruit is not available, their consumption of higher protein vegetation will dominate. Meat makes up only 3% of their diet or less, depending on which primate and what location. There is obviously a huge variation in diets but plants dominate regardless.
So when Graham says that apes and orangutans prove you can build muscle without meat and dairy, he is most certainly correct. They have far more muscle mass than we have and they eat far less animal protein. And, though we are different than the primates, they are BY FAR our closest relatives.
4. Sugar, Inflammation, and Funding Foibles
I am sorry to say it, but Denise demonstrates her inexperience in this section. First off she suggests that I didn’t read the article (which you can find at http://ajcn.nutrition.org/content/82/2/421.long). This is a bit ridiculous. Anybody who knows me or my writings knows I dissect these articles to pieces. She accuses me of only reading the abstract; however, my mode of evaluating the literature is to skip over the abstract and head straight for the methods and results. The data and the design are objective. The rest of the article is subject to interpretation.
Denise implies I am hypocritical because this study she says was funded by the sugar industry. The study was funded by the Danish Research and Development Program For Food Technology and a sugar company. Coke provided the soft drinks.
That being said, the whole design of the study was created to prove that sugar causes inflammation. In fact, the authors refer several times to a previous observational study that showed that people who ate a high glycemic diet had higher inflammatory scores, and this study was designed to test that hypothesis in a randomized control fashion. I actually felt that the authors massaged the data to try and find a correlation where one simply did not exist. I will explain later.
In addition, Denise obviously does not know the authors, but I do. Arne Astrup, the lead investigator, is a highly respected researcher, and someone I have found to be completely unbiased. He is certainly trustworthy. His design for this study was excellent. Double blind control trial where they were giving the experimental group 125-175 gm of sucrose a day. This group drank an estimated 1.3 liters of Coke a day!!!! The control group was matched perfectly but used artificial sweeteners. They went to great lengths to be sure that the two groups ate the same amount of protein, even checking their urine to be sure it correlated correctly with their reported consumption.
The experimental group was consuming a huge amount of sugar. The research I presented in Proteinaholic just before discussing this article showed that even small intake of cream will increase inflammatory markers. Can you imagine what 1.3 liters of cream would do. If sugar is the cause of inflammation in our bodies, then surely the intervention group should have huge inflammatory responses?
After 10 weeks of consuming 1.3 liters or more of Coke and eating even more sugar, the experimental group gained weight. Of course they did: the energy density of their food/drink was much higher and they were consuming far more calories. So not only were they consuming huge amounts of sugar, but they were fatter. We know fat actual secretes inflammatory mediators, so they MUST have more inflammation, right?
Well, after 10 weeks the CRP actually decreased in the sucrose group. The authors note that there were 4 subjects whose CRP results were not reliable, so they were removed from the study. After this adjustment they found that CRP increased only 6%, and the p value was .17. What this means is that the result is not significant and could be just from chance. They note that weight had no effect, which is strange to me.
Brief interlude for statistics geeks:
I think this is because of the use of ANCOVA vs multivariate regression and type 2 error of low power. The authors note that the low number of subjects may have caused the lack of statistical significance but I would argue that the more subjects they had the more you would see that the weight gain, more than the sucrose, would affect the CRP.
That being said this study, despite huge sugar intake, did not show sucrose significantly increases CRP.
Haptoglobin, another marker of inflammation, did increase but was found to correlate with energy intake, so it may have increased in sucrose group just because they were consuming more energy. Finally, transferrin in a negative acute phase reactant, meaning its level should drop if inflammation is occurring. The authors specifically note that it was “contrary to our expectations” that the transferrin value went up in the sucrose group, though this also could have been due to a relationship with energy intake.
Denise takes a small excerpt from the study which says that sugar sweetened beverages MAY increase inflammation but fails to convey their final thought. “The relative changes in inflammatory markers in the present study were small. Even though the between-group differences in haptoglobin and transferrin were statistically significant, it is doubtful that these differences are biologically important.” Not biologically important? Exactly!!!! Sugar, in this study did not increase inflammation significantly.
Oh – I almost forgot! Not only did the high sugar not cause high blood pressure, but it also did not increase insulin resistance. As I wrote in Proteinaholic, it’s the meat that makes you sweet, not the sugar.
Again, this was just one study in a whole line of reasoning that clearly showed that animal consumption increases inflammation. It was used as an aside to note that sugar, commonly believed to be the source of inflammation in the diet” does not appear to cause inflammation in a randomized control trial. This was a study performed by a top researcher, published in a top journal, and I stand by its relevance in this book 100%.
6. Wheat and The China Study
Again I made a quick, very small reference to Denise’s erroneous criticisms of the China Study. She jumps on that small reference with several pages of critique.
This has been hashed out countless times on the internet and I simply do not have the time to go into detail on this one. I refer you to Dr. T Colin Campbell’s response.
As mentioned previously, I also have enjoyed reading Plant Positive blog. Another blogger without the usual credentials, but who does a fantastic job analyzing data. He did a thorough review of Denise’s criticisms here (Like Denise and me, he is long-winded).
1) http://plantpositive.com/blog/2012/3/26/response-to-denise-minger-1-scrupulous.html
3) http://plantpositive.com/blog/2012/3/27/response-to-denise-minger-4-china-revisited.html
4) http://plantpositive.com/blog/2012/3/27/response-to-denise-minger-5-wheat-and-carbs.html. This one is more germane to the wheat question. (Hint, they were not eating whole grain; think dumplings.)
5) http://plantpositive.com/blog/2012/3/27/response-to-denise-minger-6-number-needed-to-treat.html
And Now for Intermission
As much as I enjoy reviewing the literature and engaging in high brow discussions, I am a very busy physician, training for triathlons, who also has two kids who want me to go play basketball. The next part of Denise’s criticism gets to the heart of her background and will provide some interesting debate. I will post further responses in the near future.