Thursday, February 21, 2013

Harriet Hall's "Critique" of Greger's "Uprooting the Leading Causes of Death"

Harriet Hall, MD, writes a blog titled “Science–Based Medicine.”  On Feb 12, 2013, she posted “Death as a Foodborne Illness Curable by Veganism” in which she purports to critique Dr. Michael Greger’s presentation “Uprooting the Leading Causes of Death.”

Hall starts her blog by citing “some comments recently posted by vegans on Facebook.”  If your interest is science-based health information, or you have the intent of critiquing a particular individual’s presentation, I don’t see the relevance of these comments to the discussion.  I could find hundreds of ridiculous comments on Facebook made by non-vegans, but these would provide no information relevant to Hall’s perspective.  To me, this looks like a pre-emptive character assassination, designed to lead the reader into an opinion about “vegans” which she would like her reader to then transfer to Greger. 

Among the comments she attributes to “vegans on Facebook,” she produced this:

Humans are not omnivores; they are herbivores.

To which Hall responds: [Most biologists would disagree.]

I feel surprised that a person claiming to have “science-based” views responds to that assertion by citing her opinion that “most” biologists would not agree with it.  Even if some survey has determined that “most” biologists would disagree with that statement, we do not determine the truth or falsity of a statement solely on whether “most” experts agree or disagree with it, because it remains possible that “most” have an incorrect, albeit evidence-based, belief.   At one time, “most” astronomers believed that the sun orbited around the earth, and could marshal many evidence-based arguments in favor of this belief, and only a few, like Kepler and Galileo, believed otherwise, because they had produced new evidence which ultimately refuted the idea that the solar system is geocentric.

In my view, the statement that Hall attributes to “vegans on Facebook” (what kind of reference is that?) presents a false dichotomy, since many de facto omnivorous species eat primarily plant-based diets and thus we can also call them herbivorous, such as chimpanzees, pigs, many monkeys, and many rodents. 

I find that many people (biologists included) mistakenly believe that the statement “Humans are omnivores,” which more or less correctly reports the behavior of many (but not all) humans, also implies that human physiology not only tolerates (to some extent) but requires consumption of animal flesh.  However, I have yet to see anyone specify the heritable characteristic of humans (such as a requirement for some substance found only in animal flesh) that unequivocally indicates that humans require inclusion of animal flesh in their diets. 

Anyway, leaving aside Hall’s other preliminaries, when she finally gets around to critiquing Greger’s presentation, she links to this page and writes this:

“Cardiovascular Disease (CVD)

“He cites a reference showing that “a plant-based diet of primarily whole grains, fruits, vegetables, and legumes can completely prevent heart attacks.” This is a quotation from an article in the Food and Drug Law Journal, and the footnotes there only send us to Caldwell Esselstyn’s flawed research.  Esselstyn studied only a small number of patients who already had heart disease, and he treated them with statin drugs in addition to diet, and their diet included skim milk and low-fat yogurt. You can read my criticism of his research here.  It is ludicrous to interpret that research as showing that a plant-based diet can completely prevent heart attacks.  A more accurate interpretation is that patients (only a few patients in one study) who had already had a heart attack did not have a second heart attack while being treated with cholesterol-lowering medications and a diet that was largely plant-based but also included foods derived from animals.”

I did not remember Greger discussing that article from the Food and Drug Law Journal, nor Esselstyn’s research, at all, in the portion of the "Uprooting" video wherein he discusses cardiovascular disease.   I went back to listen a second time to Greger’s presentation.  He discusses cardiovascular disease  in a segment starting at about 1:10 and ending at about 9:19.

As anyone can hear or see, Greger didn’t base his discussion of cardiovascular disease on  an article in the Food and Drug Law Journal, as Hall claims.

So, at this point, I had to wonder, did Hall actually watch this video, or does she just want to mislead anyone reading her blog by inserting another diversion, or what?

Greger provides the references he used below the video, on the page to which Hall links, as follows:

Risk Factors for Mortality in the Nurses’ Health Study: A Competing Risks Analysis

Use of intensive lipid-lowering therapy in patients hospitalized with acute coronary syndrome: An analysis of 65,396 hospitalizations from 344 hospitals participating in Get With The Guidelines (GWTG)

Shifting from decreasing risk to actually preventing and arresting atherosclerosis.

It's the cholesterol, stupid!

FDA announces safety changes in labeling for some cholesterol-lowering drugs

FDA Warns on Statin Drugs

Effect of a single high-fat meal on endothelial function in healthy subjects.

Effects of a high-fat meal on pulmonary function in healthy subjects.

Increase in plasma endotoxin concentrations and the expression of Toll-like receptors and suppressor of cytokine signaling-3 in mononuclear cells after a high-fat, high-carbohydrate meal: implications for insulin resistance.

The capacity of foodstuffs to induce innate immune activation of human monocytes in vitro is dependent on food content of stimulants of Toll-like receptors 2 and 4.

High fat intake leads to acute postprandial exposure to circulating endotoxin in type 2 diabetic subjects.

In addition, Hall implies that Esselstyn treated all of his patients with statin drugs, and that he had them consume low-fat dairy products.  In fact, Esselstyn only used statins for the few patients who did not achieve a total cholesterol below 150 by diet alone, while some patients were incapable of taking statins and others refused.  In the Q&A section of his website, under the heading “Statins–Should I take statins or not?” Esselstyn reports “Clearly some of our most profound successes in arresting and reversing disease were with patients who either refused or were incapable of taking statins.”

In addition,  some fungi (mushrooms and yeasts) produce statins (lovastatin and mevastatin), emphasizing the importance of compounds produced by the "fruits" of  plant-like organisms in the correction of cardiovascular disease.

As for the low-fat dairy products, although Esselstyn initially included them in the foods allowed to his patients, he later asked them not to consume these foods.   In the Q&A section of his website, he writes:

“Egg whites, fat free milk, yogurt – So What is wrong with egg whites, fat free yogurt, skim milk?
Egg whites, fat free milk and yogurt are ALL animal protein, and animal protein injures the lining of the arteries. Do not eat.”

Hall declares “It is ludicrous to interpret that [Esselstyn’s] research as showing that a plant-based diet can completely prevent heart attacks...etc.”  This constitutes a straw man since Greger didn’t even cite Esselstyn’s research, let alone interpret it as alone sufficient to show that a plant-based diet can "completely prevent" heart attacks. 

Nevertheless, Esselstyn’s peer-reviewed papers reference the body of evidence that supports the idea that a whole foods plant-based diet that lowers total cholesterol to less than 150 mg/dL can prevent almost all of the heart attacks that occur in industrialized nations:

“However, coronary artery disease is virtually absent in cultures that eat plant-based diets, such as the Tarahumara Indians of northern Mexico,8 the Papua highlanders of New Guinea,9 and the inhabitants of rural China10 and central Africa.11 Hundreds of thousands of rural Chinese go for years without a single documented myocardial infarction.10” [1]
 "Although coronary artery disease is the leading killer of men and women in the USA, it is rarely encountered in cultures that base their nutrition primarily on grains, legumes, lentils, vegetables, and fruit.3,4 In the Framingham study, people with cholesterol levels between 150 and 200 mg/dl accounted for 35% of those with coronary heart disease, but among those with levels <150 disease="" encountered="" face="Arial, Helvetica, sans-serif" font="" rarely="" the="" was="">.  5, 6
" [2]

Thus, as part of reporting the results of his own clinical study, Esselstyn provided references to other research supporting his protocol and conclusions.  As expected in science, Esselstyn took into account a large body of evidence when formulating his hypothesis and experimental design. 

Hall ignores Greger’s reference to Risk Factors for Mortality in the Nurses’ Health Study: A Competing Risks Analysis which, as Greger notes, found that the single factor most strongly linked to heart disease mortality among nurses studied was cholesterol, only found in animal products, while the factor most protective was fiber, found only in plants. 

She also ignores his reference to the 2010 article  It's the cholesterol, stupid! in which William C Roberts, M.D., who has has authored several books on cardiovascular disease, with his colleagues published more than 1,150 articles on cardiovascular disease in medical journals, and serves as editor-in-chief of The American Journal of Cardiology, states that, based on the evidence available, virtually complete prevention of cardiovascular disease requires reduction of total cholesterol to 150 mg/dL, the average level of people who completely avoid or only every occasionally consume dietary animal products, or who host cholesterol-lowering parasites.

Then she also largely ignores Greger’s presentation of the evidence that the combination of saturated fat and bacterial endotoxins in dietary animal products promotes inflammation involved in cardiovascular disease.  She tries to discredit only one of the several studies he cites, Effect of a single high-fat meal on endothelial function in healthy subjects with the remark that “This was a small study of 10 volunteers with no control group.” 

First of all, this is an experimental study, which involved feeding the subjects two different meals at different times, and closely monitoring their vasoreactivity for a time period after the meals, so can only be performed on a small group of people.  Does she think that a research team can do this type of study on 100 or 1000 or 10,000 people at a time?  Does she have any idea what it would cost in labor, funds and time to perform such a study on even 100, let alone 1000 or 10,000 people?  And what would be the point anyway?  Does she really think that human biology varies so much among individuals that the results would be markedly different in a larger study?

Next, she says that this study had no control group.  In fact, the abstract of the study states the following:

“To assess the direct effect of postprandial triglyceride-rich lipoproteins on endothelial function, an early factor in atherogenesis--10 healthy, normocholesterolemic volunteers--were studied before and for 6 hours after single isocaloric high- and low-fat meals (900 calorie; 50 and 0 g fat, respectively).”

In other words, since all 10 subjects received both treatments at different times, each of the 10 subjects served as his/her own control.  

Some Interesting Claims

Hall takes Greger to task for his claim that meat-eaters have 2–3 times the risk of dementia compared to vegetarians.  Her comment:

“This claim is based on an old Adventist health study that has not been replicated. It studied two groups: matched and unmatched subjects. The data he cites are from the matched group. There was no difference in incidence of dementia between meat eaters and vegetarians in the unmatched study.  Adventists are lacto-ovo-vegetarians who eat milk and eggs. And they are also a rather unique group with other healthy lifestyle practices. So it is disingenuous to claim this study as definitive evidence for veganism.”
Where to start?  First, not all Adventists are lacto-ovo-vegetarians; some Adventists eat nonvegetarian diets, some eat pesco-vegetarian diets, some eat lacto-ovo vegetarian diets, and some eat vegan diets.  Researchers like to study the Adventists because, regardless of diet, Adventists all share similar “healthy lifestyle practices” such as avoiding smoking and alcohol, making it easier to trace the effects of diet.  In other words, the "other healthy lifestyle practices" that she cites as if to cast doubt on the effect of diet found in this study actually factor out because this study compared meat-eating Adventists to vegetarian Adventists. 

In this study, when vegetarians and meat-eaters were matched for age, sex, and zip code (i.e. similar socio-economic status), meat-eaters had a risk of dementia 2–3 times that of vegetarians. When not matched, they found no difference in dementia incidence.  Hall seems to think that the results from the unmatched substudy cast significant doubt on the results from the matched substudy.

I don't know why.  Since dementia is age-related and may also be related to education and socioeconomic status, not matching subjects for age and zip code could result in comparing the dementia rates of young, college-educated, upper middle class meat-eaters to the dementia rates of elderly vegetetarians who never graduated from high school and had life-long low income.  Can you guess which of these two groups would have the higher incidence of dementia? Of course you can.  Matching provides protection against confounders that would obscure the effect of the variable under study, which in this case was diet.  Hence, I don’t know why Hall seems to think that the unmatched data in this study is even worthy of mention as if it somehow seriously contradicted the matched data.  
As for Hall's remark that it is disingenuous to claim this study as "definitive evidence" for veganism, here we have another straw man.  Greger did not claim that this study alone provides "definitive evidence" for veganism.  He cites it as part of a large and growing body of evidence indicating that consumption of various animal products promotes many degenerative diseases and limiting intake of animal products may reduce the risk of death from those diseases.

Scientists typically do not rely on any single study as "definitive evidence" for any hypothesis and value most a hypothesis capable of explaining a large body of diverse evidence.  Consequently, when formulating or providing an argument in favor of an hypothesis, they will rely upon multiple lines of evidence, as Greger did in his presentation by referring to research showing links between animal product consumption and more than one dozen conditions. 

Hall states that Greger  “compares raw meat to hand grenades, because of bacterial contamination.”   If you watch the video, at 39:01 you will find that it was a USDA poultry microbiologist named Nelson Cox who compared raw meat to a hand grenade, not Greger.  You can find the entire quote here  in a book published by the National Academies Press: Secret Agents:  The Menace of Emerging Infections by Madeline Drexler.  I find it perplexing that Hall had difficulty identifying the correct author of those words. 

Confused Much?

Hall used the word “veganism” in her title, and lampoons the “comments by vegans on Facebook,” but when it comes to trying to dispute the benefits of vegan diets, she conflates vegan with vegetarian and seems reluctant or unable to notice the possible variations of veganism.

She cites this study as evidence that “mortality from ischemic heart disease was 26% lower in vegans and 34% lower in lacto-ovo-vegetarians (in other words, it’s better not to eliminate milk and eggs)."

Careful reading of the study reveals this passage (p. 520S):

“The number of vegans was small (n= 753 subjects, 68 deaths), so the analyses in Table 7 were repeated with the inclusion of data from the Health Food Shoppers Study, making the assumptions that all nonvegetarians were regular meat eaters and that vegetarians who reported that they did not consume dairy products were vegans. This increased the number of vegans to 1146, of whom 165 died before age 90 y. However, the numbers of deaths from individual cancers among vegans remained small (range: 3–8). The death rate ratios for the vegans compared with the regular meat eaters from the other causes of death were: for ischemic heart disease, 0.89 (95% CI: 0.65, 1.24; NS); for cerebrovascular disease, 0.51 (95% CI: 0.26, 1.00; NS); for other causes, 1.39 (95% CI: 1.12, 1.72; P < 0.01); and for all causes, 1.06 (95% CI: 0.81, 1.38). However, these death rate ratios should be interpreted with caution because of the uncertainty of the dietary classification of subjects in the Health Food Shoppers Study.”

In other words, this study doesn’t provide any reliable evidence about vegan health because the survey methods used did not clearly distinguish vegetarians from vegans, and because the small number of vegans in toto led to very small numbers of deaths from each cause among vegans.  Moreover, this study may also suffer from selection bias problems, since, for example, the so-called Health Food Shoppers data it included would have only found vegans who shopped at specific health food shops.

Also, Healthy Longevity has discussed other limits of this study here, including the fact that a validity assessment of the survey used in the Health Food Shoppers cohort suggested that 34% of the individuals classified as vegetarians in that cohort actually ate meat on a regular basis. 

In addition and perhaps of more importance, since in this and other studies, the authors defined “vegan diet” by only one characteristic, namely exclusion of all animal products, their vegan group probably included some number of vegans consuming any number of highly processed and harmful plant-derived foods, such as hydrogenated oils, deep fried potato chips, cakes and cookies, highly saturated coconut and palm oils, and the like, as well as individuals who, for whatever reason, fail to ensure an adequate intake of fruits, vegetables, legumes, nuts, vitamin B12 ,or vitamin D, contrary to recommendations of Greger.  There exist many variations of vegan dieting, not all of them health protective. 

Hall claims that this study “showed that the healthiest people in Europe, the inhabitants of Iceland, Switzerland, and Scandinavia, consume large amounts of animal foods.”  The abstract of this opinion article does not give the criteria by which the author judges the inhabitants of those three nations as “the healthiest” in Europe.  Even if the heavy meat-eaters in Iceland, Switzerland, and Scandinavia are “the healthiest people in Europe,” they may be less healthy than they would be eating less animal foods, or less healthy than people living in some non-European nations, such as Japanese or Okinawans who eat much less animal-sourced foods.

By analogy, suppose I have 10 people in a room, 7 of which have chronic obstructive pulmonary disease, kidney failure, and congestive heart failure, while Tom, Dick, and Harry have angina, coronary atherosclerosis, and benign prostatic hypertrophy.  Perhaps I could correctly say that Tom, Dick, and Harry are the healthiest people in the room, but I don't think many people would consider their conditions desirable.

For example, Frassetto et al report that the elderly women of Switzerland, Denmark, Norway, and Sweden have hip fracture incidence more than 40 times greater than those of Nigeria, China, and New Guinea, and about twice that found in Japan, Italy, Holland, and Spain.  This data disputes the claim that the Swiss and Sandinavians qualify for the title "healthiest people in Europe" let alone healthiest people in the world.

Also, people in Iceland, Switzerland, and Scandinavia may have other characteristics making them appear “healthier” than other people in Europe, such as low infant mortality, universal access to medical care,  less air pollution, or low rates of smoking or alcoholism.  A proper study of this would compare the health of, for example, Swiss meat-eaters to Swiss vegans, controlling for age, sex, smoking, alcohol use, exercise habits, economic status, and other potential non-dietary confounders, and also sort from vegans those who eat poorly and fail to ensure adequate vitamin B12 and D intakes.

Hall points to this study  as finding “no significant differences in mortality between vegetarians and nonvegetarians.”  This is irrelevant to the question of the effect of a long-term, healthful vegan diet on mortality, since vegetarians eat milk  products and eggs containing animal protein, saturated fats, and cholesterol (and some might eat meat, see above).  Sixty-five percent of the fat in dairy fat (whole milk, cheese, butter) consists of saturated fatty acids, compared to only about 40 percent in lard and  50 percent in tallow, and dairy fat also contains trans fats.  A single egg supplies 186 mg of cholesterol, the amount found in ten ounces of 90% lean ground beef.  An ounce of cheese or one egg has about the same amount of animal protein as an ounce of beef.  Thus, a “vegetarian” may have intakes of animal protein, saturated fat, or cholesterol as high or higher than non-vegetarians.

NonVegan Diets Can Lead To Deficiencies

Hall also remarks that “A vegan diet can lead to deficiencies in various nutrients: vitamin B12, vitamin D, calcium, iodine, iron and omega-3 fatty acids.”  Does she think that only vegans have a signficant risk for these deficiencies?

In fact, I can correctly say this:  Research shows that a nonvegetarian diet can lead to deficiencies in various nutrients:  vitamin B12, vitamin D, potassium, calcium, iodine, iron, and omega-3 fatty acids.  Here's the evidence:

According to the USDA, up to 40% of the meat-eating U.S. population may have marginal B12 status. 

According to the NHANES 2006, “The overall prevalence rate of vitamin D deficiency [in the USA] was 41.6%, with the highest rate seen in blacks (82.1%), followed by Hispanics (69.2%).”

According to the NHANES 2002, only 40% of Americans consume the estimated AI (adequate intake) for calcium, despite 48% of them taking calcium supplements:  "Many Americans--particularly men, ethnic minorities, and the socially disadvantaged--are not meeting the current recommendations for adequate calcium intake through diet alone or with supplements."

According to the NHANES 2008,  more than 98% of nonvegetarians fail to ingest adequate potassium.

In the early 20th century in the US, goiter from iodide deficiency was prevalent in the Great Lakes and Pacific Northwest regions, despite widespread consumption of animal flesh, eggs, and dairy products.  Since 1924, omnivores have been using salt fortified (supplemented) with iodine to prevent this deficiency.  These days, users of milk products also get iodide from those foods, because iodine is used to clean the teats of lactating cows. 

According to the CDC, about 9%–16% of US females aged 12–49 years of age exhibit iron deficiency.    Moreover, the incidence of iron deficiency in vegetarians does not differ significantly from that of nonvegetarians.  On Feb 19, 2013, I used to analyze the vegan diet I consumed that day, and found that I ingested 30 mg of iron primarily from fruits, vegetables, nuts, and legumes, an amount almost 3 times the RDA for people of my age and sex.

Omega-3 fatty acids?  Scientists have been proposing that many omnivores suffer from omega-3 deficiencies for about 20 years now; this is, in part, why the American Heart Association recommends two servings of fish per week: not all nonvegetarians eat omega-3 rich foods.  However, the only essential omega-3 fatty acid, alpha linolenic acid (ALA), occurs naturally in green leafy vegetables, flax seeds, soy beans, hemp seed, walnuts, and other foods.  As noted by Katherine Milton:

“A number of cultivated leafy vegetables Americans eat are rich (>50% of total fatty acid content) in ALA (e.g. chinese cabbage, white and red cabbage, kale, brussel sprouts, parsley).  But most Americans do not consume large quantities of these plant foods either fresh or cooked; cooking of these foods also tends to destroy ALA.  The diet of human ancestors (like the diets of extant monkeys and apes) was likely rich in both linoleic acid and ALA from fresh plant tissues and for this reason the fatty acid composition of such plant foods is likely to be most compatible with human biology.” 

 Get the Picture?

Hall seems to think the Inuit had / have good health on their meat-based diets.  Healthy Longevity has shown that evidence indicates the opposite, herePlant Positive also looked over the literature on the Inuit in this video:

Hall seems to think that the journalist Gary Taubes has more credibility than Greger because Taubes "has written a book (with far more references than this video)."  I feel surprised that someone interested in "science-based" perspective thinks that sheer quantity of references lends credibility to an author's work.  Too bad she didn't check up on Taubes's general credibility, discussed here in Reason Magazine, and follow up on the references cited by Taubes, to see if they actually support his position.  Plant Positive has done so in a few videos, starting with this: 

 I think the above gives an adequate idea of the quality of Hall’s critique of Greger.  I don’t have inclination to spend any more time on this project.


Healthy Longevity said...

Thank you for this informative critique of Hall.

The simple definition of an omnivore is not meaningful in the context of a healthy diet as it does define the most optimal ratio of animal to plant foods for a particular species. For example, rhesus monkeys have been observed to consume animal matter, yet it has been demonstrated that feeding even a minuscule amount of dietary cholesterol to this species is sufficient to induce atherosclerosis. Regardless of such findings, the Paleo and low carb crowd still make misleading claims suggesting cholesterol feeding can only produce experimental atherosclerosis in pure herbivores such as the rabbit and only at extreme intakes.

Frank said...

Don, thank you very much for this!

Don said...

I have added some things re. the "healthiest people in Europe" and nonvegan diets causing nutrient deficiencies.

Shane said...

Great critique. Very informative. Well done!

Healthy Longevity said...

In the EPIC-Elderly cohort with 74,607 men and women, it was found that greater adherence to a plant-based diet in European elders was associated with lower all-cause mortality.

Most Scandinavian countries experienced a decrease in coronary heart disease and all-cause mortality largely explained by a decline in total serum cholesterol as a result of a large decline in intake of saturated animal fat.

Don said...

Thanks for those additions, HL!

Cetamua said...

AHS-2 study conclusions ought to be part of this discussion.

The adjusted hazard ratio (HR) for all-cause mortality in all vegetarians combined vs nonvegetarians was 0.88 (95% CI, 0.80-0.97). The adjusted HR for all-cause mortality in vegans was 0.85 (95% CI, 0.73-1.01); in lacto-ovo-vegetarians, 0.91 (95% CI, 0.82-1.00); in pesco-vegetarians, 0.81 (95% CI, 0.69-0.94); and in semi-vegetarians, 0.92 (95% CI, 0.75-1.13) compared with nonvegetarians.

In other words, those who followed a diet closest to the Mediterranean one had the lowest mortality risk. the Universe, in our plate and within ourselves.

Don said...


How exactly does eating fish equate to "balance" in your mind?

If you read the full text of that study,

you will find that among men, vegans had the lowest all-cause, IHD, CVD, and cancer mortality, but pesco-veg had the lowest "other" cause mortality. In contrast, among women, the pesco-vegetarians had the lowest all-cause, IHD, CVD, and cancer mortality, but the vegans had the lowest "other" cause mortality.

Since the pesco-men had a 0.60 and for 'other' cause mortality, one has to wonder how in men only, eating fish prevents deaths not caused by IHD, CVD, or cancer, but increases IHD, CVD, and cancer risk.

Similarly, although vegan-men had the lowest risk of IHD (0.45), vegan women had the highest (1.39). So, how do you explain that eating fish protects women from IHD, but increases the risk of IHD in men? And it increases the risk of 'other' cause mortality in women, but decreases the risk in men?

Simply, these are correlations, not to be interpreted as a final proof that eating fish prevents heart disease in women and certainly not that it helps men.

This means we can't make too much of the numbers obtained when combining the male and female data. Since this study covered only 5 years of follow-up (2002 to 2007) and IHD, CVD, and cancer take years to manifest sufficiently for diagnosis, it is probable, for example, that the vegan group included some people who had only recently adopted the vegan diet to deal with some health issues (e.g. IHD). Further, no attempt was made to determine the B12 status of the vegans in this study, leaving open the possibility that this one nutrient may have influenced the results, i.e. perhaps vegan Adventist women had failed to meet their B12 requirements.