Vegetarian Diets and Quality of Life: Cause or Effect?

Very few would argue that simply excluding flesh from the diet will guarantee optimal health and longevity. However, the CBS Atlanta recently featured a concerning article, Study: Vegetarians Less Healthy, Lower Quality Of Life Than Meat-Eaters, suggesting that diets that exclude flesh promotes poor health. This article which has gathered much attention describes the findings of a cross-sectional survey from Austria that was published in PLoS One.1 This study has previously been addressed by Don Matesz in an very informative post. However, due to very serious omissions made by the CBS Atlanta, I felt that it was necessary to also address this study.

Vegetarian Diets and Perceived Health: Cause or Effect? 

It cannot be emphasized enough how important it is to recognize that this study, based on the Austrian Health Interview Survey (AT-HIS) examined dietary patterns after the subjects had developed health problems. Many vegetarians are not born into vegetarianism, but adopt a vegetarian diet later in life. Therefore, it is important to address why the vegetarians in this study adopted a flesh free diet. This important limitation was acknowledged by the Austrian researchers, who asserted:

Potential limitations of our results are due to the fact that the survey was based on cross-sectional data. Therefore, no statements can be made whether the poorer health in vegetarians in our study is caused by their dietary habit or if they consume this form of diet due to their poorer health status. We cannot state whether a causal relationship exists, but describe ascertained associations.

More importantly, in regards to causation the researchers asserted:

Our results have shown that vegetarians report chronic conditions and poorer subjective health more frequently. This might indicate that the vegetarians in our study consume this form of diet as a consequence of their disorders, since a vegetarian diet is often recommended as a method to manage weight and health.

The researchers suggested that if anything, it was not a flesh free diet that caused a higher rate of a number of health problems, but rather that it was poor health that caused these subjects to adopt a flesh free diet. This is similar to the phenomenon where former smokers report poorer perceived health than current smokers, because they quit smoking with the intention of alleviating poor health.2 This phenomenon is often referred to as reverse causality. 

Unfortunately, Benjamin Fearnow, the author of the article in the CBS Atlanta ignored the evidence suggesting that these results were the result of reverse causality, and instead suggested that a flesh free diet was actually the cause of a number of health problems:

...the vegetarian diet — characterized by a low consumption of saturated fats and cholesterol that includes increased intake of fruits, vegetables and whole-grain products — carries elevated risks of cancer, allergies and mental health disorders.

It is important to note that the Austrian Health Interview Survey did not measure food intake in actual detail. Subjects who reported consuming a flesh free diet were simply assumed to be consuming a diet poor in dietary cholesterol and saturated fat. However, in this study 36% of the vegetarian subjects were classified as lacto-ovo vegetarians, and 55% pescetarians (allowing fish, dairy and eggs). Only 9% were classified as vegans.1 Therefore, up to 91% of the subjects classified as vegetarians consumed dairy and eggs, being the richest sources of saturated animal fat and cholesterol, respectively. The CBS Atlanta failed to mention even the definition of a vegetarian diet used in this study, yet alone the breakdown of subjects in each category of vegetarian diet.

Vegetarian Diets and Cancer

At the time of the report, it was observed that 4.8% of the subjects of the Austrian Health Interview Survey classified as vegetarians had cancer, as opposed to 1.8% of the subjects following an omnivorous diet rich in meat. Unfortunately, no details were provided as to what portion of the studied population adopted a flesh free diet after diagnosis. However, data from previous studies suggest that cancer patients are highly motivated to adopt a plant based diet. As described previously: 

The results of a recent study from the Netherlands illustrates the critical importance of considering reverse causality in research on plant-based diets. The researchers found that 75% of the vegetarian participants with cancer adopted a vegetarian diet after diagnosis, consistent with previous research which found that cancer survivors are highly motivated to adopt a more plant-based diet with the intention of improving poor health.3 4

If the 75% figure from the study from the Netherlands is to be considered representative of this Austrian population, this would suggest that only 1.2% of the vegetarians adopted a flesh free diet prior to diagnosis of cancer. This is lower than the 1.8% figure for omnivores following a meat rich diet, but similar to that of the omnivores following a diet low in meat. Unfortunately, due to the lack of reliable data these estimates should be taken with a grain of salt. 

Prospective (forward-looking) studies which measure diet before diseases are diagnosed are much less likely to be complicated by reverse causality than cross-sectional studies, and therefore considered to be more appropriate for determining causality. I previously carried out a meta-analysis of 5 prospective cohort studies comparing the rates of cancer incidence in vegetarians compared to health conscious omnivores. For this review, I updated the meta-analysis to include the rates of major cancers in the Adventist Mortality and Adventist Health studies. In addition, I limited the inclusion criteria to studies that provided estimates specifically for subjects classified as either vegans, or lacto-ovo vegetarians.

In a meta-analysis including 7 prospective cohort studies, vegetarians had a statistically significant 9% lower risk of cancer incidence compared to health conscious omnivores (Fig. 1).5 6 7 8 9 It is important to note that meat intake was relatively low in the omnivorous group in these studies, especially taking into account that a significant portion of the omnivorous subjects were actually classified as semi-vegetarians. This suggests the difference in cancer incidence may be greater when compared to regular meat eaters.

FIGURE 1. Risk ratios and 95% CIs for fully adjusted random-effects models examining associations between vegetarian diets in relation to cancer incidence. ¹Mortality from cancers of the breast, colorectal, lung, prostate and stomach combined. VEG, vegetarian diet.

The finding of a decreased risk of cancer in vegetarians may be explained, in part, by a diet devoid in heme iron. Controlled feeding trials have established that NOCs (N-nitroso compounds) arising from heme iron in meat forms potentially cancerous DNA adducts in the human digestive tract, likely in part, explaining the significant association between heme iron and an increased risk of colorectal cancer in recent meta-analyses of prospective cohort studies.10 11 12 Heme iron has also been associated with numerous other cancers.

Vegetarian Diets and Heart Disease

In the Austrian Health Interview Survey, it was suggested that subjects classified as vegetarians were more likely to have had a history of heart attacks. It is important to note however, that, plant-based diets, poor in saturated fat and cholesterol have for long been adopted by individuals at risk of coronary heart disease. For example, it is known that in studies carried out as far back as the late 1950s, subjects with unfavorable blood cholesterol levels tended to limit intake of dietary cholesterol and saturated fat in order to improve cardiovascular risk factor.13

I previously carried out a meta-analysis of 7 prospective cohort studies comparing the rate of death of coronary heart disease of vegetarians compared to health conscious omnivores. For this review, I examined the incidence of coronary heart disease, and limited the inclusion criteria to studies that provided estimates specifically for subjects classified as either vegans, or lacto-ovo vegetarians. In a meta-analysis including 7 prospective cohort studies, vegetarians had a statistically highly significant 24% lower risk of coronary heart disease compared to health conscious omnivores (Fig. 2).5 6 7 14 15  

FIGURE 2. Risk ratios and 95% CIs for fully adjusted random-effects models examining associations between vegetarian diets in relation to coronary heart disease incidence. VEG, vegetarian diet.

The degree of reduction in risk of mortality from coronary heart disease observed in vegetarians in these cohort studies was generally in proportion to the expected reduced risk based on the differences in levels of total and non-HDL cholesterol, and blood pressure. This is supported by evidence from prospective cohort studies which found that diets characterized as being low in saturated fat and rich in dietary fiber decrease the risk of death from coronary heart disease. These findings are also supported by a recent meta-analysis of clinical trials and observational studies that found that vegetarian diets are associated with lower blood pressure and a lower risk of hypertension.16 Interestingly, the rates of hypertension tended to be lower in the vegetarians in the Austrian Health Interview Survey, suggesting that if the subjects adopted a vegetarian diet as a means to control hypertension, they were likely successful doing so.  

Vegetarian Diets and Mental Heath

In the Austrian Health Interview Survey, it was observed that subjects classified as vegetarians had a higher rate mental illnesses, defined as anxiety disorder or depression. Unfortunately, no data was provided as to what portion of the subjects adopted a vegetarian diet after developing these conditions. These findings have appealed to proponents of Paleoloithic diets who hypothesize that humans have a dietary requirement for meat in order to maintain large brains and mental health. However, in Powered By Plants: Natural Selection & Human Nutrition, Don Matesz examines an extensive body of research that casts considerable doubt on the hypothesis that meat is required to maintain mental health and is responsible for the evolution of the large human brain.

The findings from a number of clinical trials cast doubt on the hypothesis that an appropriately designed flesh free diet has adverse effects on, and that flesh rich diets, poor in carbohydrate have beneficial effects on overall mental health.

• Sacks and colleagues carried out a crossover trial to examine the effects of adding 250 g/day of beef isocalorically to the diet on blood cholesterol of vegetarians. As expected, during the meat phase total cholesterol and systolic blood pressure increased significantly. However, it was also observed that the participants experienced increased anger, anxiety, confusion, depression, and fatigue and less vigor compared to the vegetarian phase.17
• Beezhold and Johnston compared the mood scores of participants assigned to either a vegetarian diet, excluding all animal foods except dairy to participants assigned to either a omnivorous diet, or a diet that included fish, but excluded meat and poultry. The researchers found that the vegetarian group demonstrated significantly improved mood scores compared to both the omnivorous and fish groups.18
• Schweiger and colleagues compared the effects of a vegetarian diet and an omnivorous diet on global mood scores. They found that the vegetarian group demonstrated significantly better global mood, and that carbohydrate intake associated with better global mood.19
• Kieldsen-Kragh examined the effects of a vegetarian diet on rheumatoid arthritis. The researchers hypothesized that the participants may find the vegetarian diet too restrictive, and that therefore adherence to the diet would impose psychological distress on the them. However, contrary to their expectations, the vegetarian group demonstrated significantly improved physiological health, and were less anxious and depressed compared to the omnivorous group.20
• Brinkworth and colleagues examined the effects of a very low-carbohydrate diet and a low-fat diet on body weight and mood and cognitive function. Although there was no statistical difference in terms of weight loss between the groups, the participants assigned to the low-fat group demonstrated significantly improved mood scores compared to the participants assigned to the low-carbohydrate diet.21
• Holloway and colleagues carried out a crossover trial to examine the effects of a high-fat, low-carbohydrate diet on alterations to heart and brain function. The researchers found that the participants not only demonstrated significantly impaired cardiac health, but also impaired attention, memory recall speed, and mood while following the high-fat, low-carbohydrate diet.22
• Halyburton and colleagues examined the effects of a low and high-carbohydrate diet on mood and cognitive function. Although, unlike other studies, the researchers found that mood was similar in both groups, participants assigned to the low-fat diet demonstrated improved speed of processing compared to the participants assigned to the low-carbohydrate group.23

Mass Media as a Source of Health Information

The article featured in the CBS Atlanta is just one example of many studies that are misinterpreted, likely intentionally by the mass media. Unfortunately, the mass media is certainly not a reliable source for health information, as their primarily concern is to publish news that appeal to their targeted audience. In this case it was meat eaters who desired to hear negative things about vegetarian diets. This is likely why many important studies do not receive appropriate media attention,  and why consumers are either left in the dark or simply confused about health information.

Although there is convincing evidence of the health benefits of an appropriately planned diet that either excludes or significantly limits the intake of flesh, such findings cannot be extrapolated to all diets that exclude flesh. The definition of a vegetarian diet only provides information as to what foods an individual restricts, and not which foods are included. This is why the emphasis of a healthy diet also needs be on which foods are included, not only on those that are excluded. Future research in this area should address what foods vegetarians are substituting meat with,  the length of adherence to a vegetarian diet, and whether subjects adopted a vegetarian diet in order to alleviate poor health. This would allow for a considerably more meaningful interpretation of the effects of vegetarian diets.

Cracking Down on Eggs and Cholesterol: Part II

Recently two meta-analysis papers were published addressing the findings from population studies of the association between egg intake and the risk of cardiovascular disease.1 2 Unfortunately the authors of these two review papers reached contradictory conclusions regarding the dangers of egg intake which is likely to lead to unnecessary public confusion. The authors of the most recent meta-analysis paper reviewed studies on coronary heart disease, heart failure, diabetes and all cardiovascular diseases (CVD) combined and concluded:

Our study suggests that there is a dose-response positive association between egg consumption and the risk of CVD and diabetes.

In contrast to this conclusion, the authors of the earlier meta-analysis paper limited their review to studies that specifically addressed coronary heart disease and stroke and concluded:

Higher consumption of eggs (up to one egg per day) is not associated with increased risk of coronary heart disease or stroke. The increased risk of coronary heart disease among diabetic patients and reduced risk of hemorrhagic stroke associated with higher egg consumption in subgroup analyses warrant further studies.

The second meta-analysis paper is problematic in part because the authors failed to consider the relevant findings from dozens of rigorously controlled feeding experiments on humans and thousands of experiments on animals, including nonhuman primates that strongly support the recommendations to limit the intake of eggs and cholesterol [reviewed previously]. This paper is also problematic in part because the authors failed to consider many other relevant findings from prospective cohort studies which suggest that egg and cholesterol intake increases the risk of coronary heart disease, diabetes, heart failure, cardiovascular disease and all-cause mortality.

Firstly, the association between egg intake and the risk of cardiovascular disease is meaningless without considering suitable substitutes for eggs. As a lower intake of eggs implies a higher intake of other foods in order to maintain caloric balance, the effect that egg intake has on coronary heart disease depends on which foods eggs are substituted for. For example, data from the Nurses’ Health Study, one of the largest studies included in these meta-analyses suggested that replacing one serving of nuts, but not red meat and dairy with one serving of eggs per day is associated with a significantly increased risk of coronary heart disease.3 The authors of both meta-analyses failed to address this factor despite the fact that the importance of evaluating suitable food alternatives has been strongly emphasized by many prominent diet-heart researchers.4 The findings from these meta-analyses should therefore be interpreted with caution as eggs may have been primarily compared to processed foods and other animal foods which make up the majority of caloric intake in developed nations.4 5

Eggs, Cholesterol and Diabetics

The authors of the most recent meta-analysis paper found that among diabetics, frequent egg intake was associated with a 83% increased risk of cardiovascular disease, whereas the authors of the earlier meta-analysis paper found that frequent intake was associated with a 54% increased risk of coronary heart disease. The authors of the most recent meta-analysis paper excluded one, while the authors of the earlier meta-analysis paper excluded two additional cohort studies that found that among diabetics, high compared to low intake of eggs was associated with an approximately five-fold increased risk of cardiovascular disease.6 7 These additional studies had they been addressed by these authors would have potentially strengthened the association between egg intake and an increased risk of cardiovascular disease in diabetics.

The authors of the most recent meta-analysis found that frequent egg intake was associated with a 68% increased risk of type II diabetes, a major risk factor for cardiovascular disease. However, the authors of the earlier meta-analysis largely failed to address this evidence. A literature search I performed produced papers from 5 separate prospective cohort studies addressing egg intake and the risk of developing type II diabetes, including two additional studies that were not addressed in both meta-analyses papers.8 9 10 11 In addition, I also found one additional cohort study addressing egg intake and the risk of developing gestational diabetes.12 All except one smaller cohort found a statistically significant association after adjusting for potential confounders. These cohorts also found suggestive evidence that the increased risk persisted regardless of whether eggs were consumed in the presence of a higher or lower carbohydrate diet, and that the association was even stronger when repeated measurements of egg intake were considered.9 In addition, these cohorts also found suggestive evidence that the increased risk could partly be explained by the dietary cholesterol and protein content of eggs, and that substituting eggs with carbohydrate-rich foods, especially fiber-rich bread and cereals significantly decreases the risk of developing type II diabetes.8 9 11 12

In the one cohort that did not find a statistically significant association, average egg intake was relatively low and there was suggestive evidence of an increased risk when a follow-up measurement of egg intake was used to update exposure overtime.10 In addition to these findings, a paper from the Health Professionals Follow-Up Study also found suggestive evidence that egg intake is associated with an increased risk of type II diabetes.13 Furthermore, papers from an additional 5 cohort studies found that dietary cholesterol was associated with a significantly increased risk of developing either type II diabetes or gestational diabetes.14 15 16

Overall findings from 12 prospective cohort studies with 265,675 participants and 14,497 cases of type II diabetes and gestational diabetes strongly implies that egg and cholesterol intake are significant risk factors in the development of diabetes. In addition to the findings from cohort studies, 4 cross-sectional studies found that egg or cholesterol intake was associated with between a nearly two-fold and greater than four-fold increased risk of developing type II diabetes and gestational diabetes.12 17 18 19 Also consistent with these findings, in the Adventist Health Study 2 it was observed that vegans had a lower risk of developing type II diabetes compared to lacto-ovo vegetarians, and especially non-vegetarians.20

One cohort included in these meta-analyses that used repeated egg intake measurements to update exposure over time found that in diabetics, intake of at least 7 eggs compared to less than 1 egg per week was associated with a two-fold increased risk of all-cause mortality, whereas another cohort that did not use repeated measurements found suggestive evidence of a 30% increased risk of all-cause mortality.21 22 The authors of the first study stated:

…among male physicians with diabetes, any egg consumption is associated with a greater risk of all-cause mortality, and there was suggestive evidence for a greater risk of MI [heart attack] and stroke.

An additional study found that in diabetics, an increment of one egg per day was associated with a greater than three-fold increased risk of all-cause mortality.6

According to the International Diabetes Federation, globally approximately 183 million people, or half of those who have diabetes have not been diagnosed. Even in high-income countries about one-third of people with diabetes have not been diagnosed.23 Given this data and the data that egg and cholesterol intake is associated with a significantly increased risk of developing diabetes, and that in diabetics egg intake is associated with a significantly increased risk of coronary heart disease, cardiovascular disease and all-cause mortality, there is likely a significantly greater number of people at risk than suggested by the authors of these recent meta-analyses.

Eggs, Cholesterol and Non-Diabetics

The Nurses’ Health Study found that an increment of cholesterol equivalent to one medium size egg per day was associated with a 17% increased risk of all-cause mortality, consistent with the findings from several other studies.24 25 26 Another study included in these meta-analyses found that in non-diabetics, intake of at least 7 eggs compared to less than 1 egg per week was associated with a 22% increased risk of all-cause mortality.21 Also, another cohort from Japan found that frequent egg intake was associated with an increased risk of all-cause mortality in women, consistent with the findings from the Adventists Mortality Study.27 28 In addition, a cohort of elderly found suggestive evidence that egg intake was associated with a significantly increased risk of all-cause mortality, and that substituting eggs with fruits, vegetables and grains significantly decreases risk.29

The authors of the most recent meta-analysis paper found that in largely non-diabetic populations that frequent egg intake was associated with 19% increased risk of cardiovascular disease compared to all other sources of calories combined, which is predominantly processed foods and other animal foods. The authors of the earlier meta-analysis that did not reach this conclusion suggested that their findings are relevant for total cardiovascular disease but failed to address the findings from prospective cohort studies regarding the risk for heart failure. For example, two cohort studies which were included in the most recent meta-analyses found that intake of at least 7 eggs compared to less than 1 egg per week was associated with an approximately 30% increased risk of heart failure.30 31

Another potential important finding that has contributed to the knowledge of the dangers of eggs are the results from studies that were carried out on populations with a low habitual cholesterol intake, such as vegetarian populations. The authors of the most recent meta-analysis paper excluded one, while the authors of the earlier meta-analysis paper excluded two cohort studies that were carried out on largely vegetarian populations. Frequent consumption of eggs was associated with a more than 2.5 increased risk of fatal coronary heart disease in the Oxford Vegetarian Study and also an increased risk in females in the Adventists Mortality Study.28 32 The characteristics of the participants in these studies differ from that of most other studies, not only because of the their lower habitual intake of dietary cholesterol, but also because of their lower rates of obesity and typically healthier overall diet. Therefore separately analyzing egg intake in this subgroup of the population may be of significant importance. The authors of a paper from the Nurses’ Health Study and the Health Professionals Follow-Up Study cited in these meta-analyses described the potential importance of addressing egg intake in people with very low habitual cholesterol intake and how their study may have been inadequate to test this hypothesis: 33

One potential alternative explanation for the null finding is that background dietary cholesterol may be so high in the usual Western diet that adding somewhat more has little further effect on blood cholesterol. In a randomized trial, Sacks et al found that adding 1 egg per day to the usual diet of 17 lactovegetarians whose habitual cholesterol intake was very low (97 mg/d) significantly increased LDL cholesterol level by 12%. In our analyses, differences in non-egg cholesterol intake did not appear to be an explanation for the null association between egg consumption and risk of CHD. However, we cannot exclude the possibility that egg consumption may increase the risk among participants with very low background cholesterol intake.

As it is well documented that cholesterol intake has a much greater effect of raising serum cholesterol when baseline intake is very low, this may in part explain why egg and cholesterol intake was more strongly associated with coronary heart disease in studies on largely vegetarian populations.34 35 Another explanation for a possibly stronger association in vegetarian populations is that egg intake may have a greater effect in leaner people, and it has been well documented that vegetarians are generally leaner than their omnivorous counterparts [reviewed previously]. This hypothesis is supported by several dietary experiments which found that dietary cholesterol had a greater effect of raising serum cholesterol among leaner compared to overweight participants.36 37 This hypothesis is also supported by the findings from the Chicago Western Electric Study which found that while dietary cholesterol was associated with a significantly increased risk of coronary heart disease in lean men over and above the adverse effects it has on serum cholesterol, increased intake had little appreciable effect on men with a greater BMI and body fatness.38 Another explanation for these findings is that vegetarians may choose healthier substitutes for eggs, such as nuts which was associated with a significantly lower risk of coronary heart disease compared to eggs in the Nurses’ Health Study.3

It was found in a sub-analysis based on 4 cohorts included in the earlier meta-analyses that egg intake was associated with an 18% non-significant increased risk of fatal coronary heart disease. The addition of the mortality findings from the two largely vegetarian cohorts that were excluded from this meta-analysis would have likely strengthened this association.28 32 This suggests that similar to saturated fat intake, egg intake may increase the risk of fatal coronary heart disease more than non-fatal coronary heart disease [reviewed previously]. The lack of a significant association likely reflects the fact that eggs were not compared to healthy foods, and also likely due to misclassification of participants into ranges of usual dietary intake as the result of measurement error [reviewed previously].

In the video below Dr. Michael Gregor addresses recent research on choline when consumed from eggs and other animal foods and the risk of cardiovascular disease and cancer.

Carnitine, Choline, Cancer and Cholesterol: The TMAO Connection

Egg Intake and Stroke

In regards to a sub-group analysis of 5 cohort studies, the authors of the earlier meta-analysis suggested that egg intake was associated with a lower risk of hemorrhagic stroke. The authors suggested that the inverse association between egg intake and hemorrhagic stroke is supported by findings of an inverse association between serum cholesterol and hemorrhagic stroke in several cohort studies. However, in the largest cohort study the authors cited, the inverse association was confined to participants with elevated blood pressure.39 A similar interaction between blood pressure and serum cholesterol and hemorrhagic stroke was observed in much larger cohort studies in both Asian and Western populations that the authors of this meta-analysis conveniently failed to cite.40 41 In a meta-analysis of 61 cohort studies it was found that among participants with near optimal systolic blood pressure (<125 mmHg), lower serum cholesterol was actually associated with a significantly lower risk of hemorrhagic, ischemic and total stroke mortality [reviewed previously]. Furthermore, most mammalian species have very low LDL levels (mean value of 42 mg/dl in 18 species), and there is very scant evidence that these animals are at high risk of having a stroke.42

This data demonstrates that continued emphasis should be placed on lowering both LDL cholesterol and blood pressure which have been proven in hundreds of randomized controlled trials to lower not only the risk of cardiovascular disease, but also all-cause mortality.43 44 Increasing the intake of eggs after achieving a near optimal blood pressure is unlikely to reduce the risk of hemorrhagic stroke and will likely increase the risk of dying of any cause.

Unwarranted Mediocre Health Recommendations

The conclusions of the earlier meta-analysis are misleading and inconsistent with the body of literature. What is more concerning is that these findings will likely be used in marketing campaigns to confuse the general population, of which the great majority are already at risk of cardiovascular disease. The most recent meta-analysis paper while being overall informative and more clearly demonstrating the dangers of eggs for both diabetics and non-diabetics, the authors still failed to address many important findings that have been addressed in this series of posts. A greater emphasis on the effects of replacing eggs with other suitable foods is required, and the available evidence suggests a significant benefit of replacing eggs with whole plant foods, including fruits, vegetables, whole grains and nuts.3 11 29 As Spence and colleagues pointed out in regards to recent controversy surrounding dietary cholesterol and eggs:45

…the only ones who could eat egg yolk regularly with impunity would be those who expect to die prematurely from nonvascular causes.

Diet-Heart Posts

Part I - Diet-Heart: A Problematic Revisit
Part II - Diet-Heart: Saturated Fat and Blood Cholesterol
Part III - Diet-Heart: The Role of Vegetarian Diets in the Hypothesis
Part IV - Cracking Down on Eggs and Cholesterol

Please post any comments in the Discussion Thread.

Cracking Down on Eggs and Cholesterol

Since the breakthrough led by Nikolai Anichkov a century ago, the feeding of cholesterol, and to an extent, dietary fat have been recognized as the sine qua nons for the dietary modification of experimental atherosclerosis, and have been used in thousands of experiments to successfully accelerate the development of atherosclerosis in mammalian, avian and fish species, not only of herbivorous, but also omnivorous and carnivorous nature.1 2 3 4 5 6 7 8 9 10 11 This includes the promotion of experimental atherosclerosis in over one dozen different species of nonhuman primates- New World monkeys, Old World monkeys, and great apes including the closest living relative to humans, the chimpanzee (Fig. 1).2 3 12 13 14 15 16 17 18 19 20 21 The atherosclerotic lesions induced by cholesterol feeding, including in the form of fresh eggs yolks in many opportunistic omnivores, such as various species of nonhuman primates, birds and pigs have been shown to closely resemble the disease in humans.1 2 3 4 22 23 24

Figure 1. Aortic atherosclerosis of a chimpanzee which died of a heart attack after long-term feeding of a diet rich in cholesterol and artery-clogging saturated fat

It has also been observed that the long-term feeding of cholesterol and saturated fat has resulted in heart attacks, sudden death, development of gangrene, softening on the bones and numerous other serious complications in nonhuman primates.2 3 25 26 27 28 For example, it has been shown that when diets rich in cholesterol and saturated fat are fed to monkeys of the genus Macaca, including the rhesus monkey and the crab-eating macaque, they experience heart attacks at approximately the same rate as high-risk populations living in developed nations.3

In species that are unlike humans, very resistant to dietary induced elevations in LDL cholesterol, such as the order of the carnivora, unless LDL-receptor deficient breeds are used atherosclerosis is typically induced by raising serum (blood) cholesterol with a diet with very large amounts of dietary cholesterol, and either containing thiouracil or deficient in essential fatty acids.9 10 29 As noted by Steinberg:30

The point is very clearly made: the arteries of virtually every animal species are susceptible to this disease if only the blood cholesterol level can be raised enough and maintained high enough a long enough period of time.

Long-term feeding of cholesterol in relatively small amounts has actually been shown to induce atherosclerosis in rabbits, chickens, pigeons and monkeys despite only small or insignificant increases in serum cholesterol.1 4 13 Armstrong and colleagues conducted an experiment ‘designed to demonstrate a null point of the effect of dietary cholesterol on the arterial intima’, by comparing a group of rhesus monkeys fed a cholesterol-free diet with a group fed cholesterol equivalent to that found in only half of a small egg in the average human diet of 2,000 calories per day (43µg/kcal). However, even when fed in very small amounts dietary cholesterol still had a significant adverse effect on these monkeys arteries after a period of only 18 months (Fig. 2).13 Armstrong and colleagues concluded:

No null point for the effect of dietary cholesterol on arterial intima was found even at an intake level far below that conventionally used for the induction of experimental atherosclerosis in the nonhuman primate. The intimal changes found in response to very low cholesterol intake imply that subtle qualitative alterations in lipoproteins are of critical importance to our understanding of lesion induction.

Figure 2. Subclavian artery from a rhesus monkey fed very small amounts of dietary cholesterol (43µg/kcal). Sudanophilia (black area) is intense in the area of major intimal thickening

It has also been demonstrated that the cessation of a cholesterol-rich diet and the subsequent lowering of serum cholesterol results in the regression of atherosclerosis in various mammalian and avian species, including herbivores, omnivores, carnivores and nonhuman primates.31 In one experiment Armstrong and colleagues induced severe atherosclerosis in rhesus monkeys by feeding a diet with 40% of calories from egg yolks for 17 months. The egg yolks were then removed from the monkeys diet and replaced with a cholesterol-free diet with either 40% of calories from corn oil or low-fat chow with 77% calories from sugar for three years, resulting in a reduction of serum cholesterol to <140 mg/dl and a marked regression of atherosclerosis.32 33

In a recently published study, Spence and colleagues observed that egg yolk consumption was associated with carotid plaque in high-risk patients.34 These findings should not come as a surprise considering the evidence accumulated from thousands of animal experiments over the last 100 years, which have demonstrated that the feeding of cholesterol and saturated fat accelerates the development of atherosclerosis in virtually every vertebrate that has been sufficiently challenged. These lines of evidence have been neglected by the egg industry and promoters of cholesterol laden diets (ie. Paleo, Primal and low-carb) who have attempted to discredit this study without considering the relevant evidence. As noted by Stamler:35

To neglect this fact in a review about humans is to imply that the Darwinian foundation of biomedical research is invalid and/or that there is a body of substantial contrary evidence in humans. Neither is the case. 

These findings from Spence and colleagues are not only supported by the findings from animal experiments, but also by numerous previous human studies that found a positive association between dietary cholesterol and the severity of atherosclerosis.36 37 38 39

In the video below Dr. Michael Greger addresses the completely unethical measures that the egg industry resorted to in order to confuse the general public about these findings from Spence and colleagues, including attempts to bribe researchers.

Eggs vs. Cigarettes in Atherosclerosis

In the video below Plant Positive addresses various critiques of Spence and colleagues findings, as well as other relevant research on dietary cholesterol.

Cholesterol Confusion 6 Dietary Cholesterol (And the Magic Egg)

Eggs, Cholesterol and Xanthomatosis

In addition to developing atherosclerosis and gangrene, the feeding of egg yolks and cholesterol to various species of nonhuman primates has also resulted in the development of xanthomatosis, a condition where deposits of cholesterol develop underneath the skin and is associated with chronically elevated serum cholesterol.18 40 41 42 43 This condition has been shown to be cured in nonhuman primates upon the cessation of a cholesterol-rich diet.41 A case report found that a 30-year-old woman with a healthy body weight who had been following a carbohydrate restricted diet for three and a half years had developed xanthomas on her hands and a chronically elevated serum cholesterol level of 940 mg/dl.44 The composition of the woman’s diet was reported as follows:

Each day she had consumed eight to 12 eggs, one or two lean steaks or half a small chicken and, half to one litre of milk. Sometimes some cottage cheese or tomatoes enriched the menu and, on rare occasions, fruit. She completely avoided butter, bread, potatoes, rice, noodles, alcohol, or any other food or beverage containing carbohydrate. The daily cholesterol intake, which was mainly derived from the egg yolks, was about 3500 mg. The total calorie intake was about 8-4 MJ (2000 kcal) (35 % protein, 55 % fat, and 10 % carbohydrates, polyunsaturated fat:saturated fat (P:S) ratio=0 26).

The woman was advised to change her diet, and in particular to stop eating eggs. After 16 days her serum cholesterol dropped to 750 mg/dl, and after several years dropped to 188 mg/dl and the lipid deposits on her skin had cleared up. This woman’s diet induced xanthomas and chronically elevated cholesterol resemble the characteristics of people with homozygous familial hypercholesteromia, a rare genetic disorder that results in chronically elevated concentrations of predominantly large LDL cholesterol particles.45 People with this disorder are short lived and often experience heart attacks during childhood.46 Such unfavorable risk factors would normally be of great concern to any responsible physician. However, despite the overwhelming evidence of the danger of elevated serum total and LDL cholesterol,30 47 48 including for women,49 50 Sally Fallon and Mary Enig, the founders of the Weston A. Price Foundation claim that ‘For women, there is no greater risk for heart disease, even at levels as high as 1000 mg/d’.51 It is clear that this organization has little concern for the wellbeing of people.

Eggs, Cholesterol and Serum Lipids

It has been well established in rigorously controlled feeding experiments that adding dietary cholesterol to a diet that is low in cholesterol can significantly raise serum cholesterol in humans.52 An addition of 200 mg cholesterol per day to a cholesterol-free diet has been shown to raise serum cholesterol by as much as 20%.53 This may be largely explained by the strong evidence that dietary cholesterol down-regulates the LDL receptor.54 However, as Hopkins addressed in a meta-analysis of rigorously controlled feeding experiments, there exists a ceiling effect at which adding additional dietary cholesterol to a diet already rich in cholesterol has little appreciable effect on serum cholesterol (Fig. 3). Therefore, the fact that numerous studies carried out on populations with a relatively high baseline cholesterol intake failed to find a significant association between cholesterol intake and serum cholesterol does not negate the evidence that lowering intake to near zero will significantly lower serum cholesterol.

Figure 3. Effects of added dietary cholesterol on serum total cholesterol at different baseline levels of intake

Several controlled experiments have found that overweight compared to lean people, and insulin resistant compared to insulin sensitive people are less responsive to dietary cholesterol.55 56 This likely explains why researchers who have financial or personal connections with the egg industry have specifically selected overweight and insulin resistant participants with a modestly high baseline dietary cholesterol intake for controlled trials, as it can be pre-empted that this subgroup of the population will show little response when egg intake is increased.

It has been demonstrated in multiple meta-analyses of rigorously controlled feeding experiments that dietary cholesterol, including that from eggs yolks does have a modest adverse effect on the LDL:HDL cholesterol ratio.57 58 Furthermore, unlike for LDL cholesterol, there is limited causal evidence that simply raising HDL will lower the risk of coronary heart disease. For example, a meta-analysis of 108 randomized controlled trials found that while lowering LDL cholesterol significantly decreased the risk of coronary heart disease and all-cause mortality, modifying HDL had little appreciable effect after controlling for LDL cholesterol.47 In addition, a recent meta-analysis of mendelian randomization studies found that while genetically modified LDL significantly influenced the risk of coronary heart disease, genetically modified HDL had little appreciable influence.59 This evidence together with the evidence that dietary cholesterol adversely influences both concentrations of LDL as well as the LDL:HDL ratio, especially in healthy people reinforces the recommendations to limit egg and cholesterol intake.

Another contributor to confusion caused by studies typically influenced by the egg industry is the suggestion that dietary cholesterol does not increase the number of LDL particles, or only increases the concentration of large LDL particles, which is considered by some to be less atherogenic. However, as elaborated by Plant Positive, several studies not influenced by the egg industry have found that cholesterol intake does increase the total number of LDL particles in healthy people.60 61 In addition, a systematic review found that higher LDL particle number, but not other LDL subfractions was consistently associated with an increased risk for cardiovascular disease, independent of other lipid measurements.62 The National Lipid Association Expert Panel recently concluded that ‘All lipoprotein particles in the LDL fraction are atherogenic, independent of size’, and was unable to identify any patient subgroups in which LDL subfraction measurements are recommended. In specific, the panel provided the following evidence for these conclusions:63

Studies have linked large LDL particles to atherosclerosis in nonhuman primates, in patients with familial hypercholesterolemia (who have an elevated concentration of predominantly large LDL particles), in participants of the population-based MESA study, in normolipidemic men with CHD, and among patients after MI [heart attack] in the Cholesterol And Recurrent Events (CARE) study... Many studies document links between small dense LDL particles and atherosclerotic CVD. However, these statistical associations between small, dense LDL and CV [cardiovascular] outcomes are either significantly attenuated or abolished when the analyses are adjusted for the overall number of circulating LDL particles (LDL-P) either by adjustment for Apo B levels or by adjustment for nuclear magnetic resonance-derived LDL-P... To date, there is no evidence that the shift in LDL subfractions directly translates into change in disease progression or improved outcome.

More recently a meta-analysis of mendelian randomization studies with over 312,000 individuals found that inheriting any of nine studied genetic variants that modify lifelong LDL cholesterol concentrations, but not any other known risk factors predicted a 55% lower risk of coronary heart disease for each mmol/l (38.7 mg/dl) lower LDL cholesterol.48 Despite having significantly different effects on LDL particle sizes, all of the nine studied genetic variants predicted essentially the same decrease in coronary heart disease per unit lower LDL cholesterol, including the gene responsible for familial hypercholesterolemia which elevates predominantly large LDL particles.45 Therefore there is convincing evidence that large LDL particles promote atherosclerosis.

The elevation of LDL cholesterol is not the only adverse effect that increased intake of eggs and cholesterol confers. As Spence and colleagues also pointed out in regards to recent controversy surrounding dietary cholesterol:40 64

Focusing on fasting serum cholesterol levels misses the bulk of the problem. Even though serum cholesterol rises very little after a meal, dietary cholesterol increases the susceptibility of LDL-C to oxidation, vascular inflammation, oxidative stress, and postprandial hyperlipemia and potentiates the harmful effects of saturated fat, impairs endothelial function, and increases cardiovascular events.

Classical Observations

Multiple international studies based on data from the World Health Organization have found the mean per capita dietary cholesterol levels are consistently associated with the rates of coronary heart disease mortality.39 65 This includes a large study of 40 countries.66 Similarly, it was found in the 25 year follow-up of the Seven Countries Study that dietary cholesterol was associated with a significantly increased risk of coronary heart disease across the 16 cohorts.67

In a review of the literature, Uffe Ravnskov, the spokesperson for The International Network of Cholesterol Skeptics reviewed 15 of the earliest prospective (longitudinal) cohort studies and inappropriately concluded that ‘Overall, longitudinal studies within population have found no difference between the diet of coronary patients and others’.68 Fourteen of these studies measured cholesterol intake, of which for the Chicago Western Electric Study Ravnskov inappropriately cited data from an earlier follow-up that found no association rather than the longer follow-up which found a significant association. Among the remaining thirteen studies, the participants who developed coronary heart disease actually had on average 13 mg/day greater intake of cholesterol for someone consuming on average 2,000 calories a day.

Considering the probable degree of measurement error dietary intake and the fact that these studies were carried out in largely homogenous populations where most people had similar diets, only relatively small differences in dietary composition would have been expected between participants with and without heart disease even if diet does play a major role in heart disease [reviewed previously]. Furthermore, Ravnskov failed to mention that four of the largest studies that he cited, including the Chicago Western Electric Study found on average that 200 mg/1,000 calories higher intake of cholesterol was associated with a 30% increased risk of coronary heart disease over and above the adverse effects it has on serum cholesterol.64

The next post in this series will focus on findings from more recent prospective cohort studies that addressed the intake eggs and cholesterol and the risk of coronary heart disease, diabetes, heart failure, cardiovascular disease and all-cause mortality. Many of these important findings have gone unaddressed in recent reviews of the literature.

Diet-Heart Posts

Part I - Diet-Heart: A Problematic Revisit

Part II - Diet-Heart: Saturated Fat and Blood Cholesterol

Part III - Diet-Heart: The Role of Vegetarian Diets in the Hypothesis

Part V - Cracking Down on Eggs and Cholesterol: Part II

Please post any comments in the Discussion Thread.

Forks Over Knives and Healthy Longevity: A Missed Opportunity for the Cholesterol Skeptics

This is the first part of a series of posts that addresses the science regarding plant based diets and the documentary Forks Over Knives and the very serious inaccuracies and omissions that compromise the critiques authored by the cholesterol skeptics, in particular Denise Minger. 

Food Shortages, Cardiovascular Disease and All-Cause Mortality in the World Wars

In Forks Over Knives, Dr. Caldwell Esselstyn described the classical findings from a paper authored by Strom and Jensen, who observed that in Norway between 1938 and 1948 there was a strong relationship between cardiovascular mortality and changes in intake of fat in the form of butter, milk, cheese and eggs, with the changes in mortality lagging behind dietary changes by approximately one year (Fig. 1).1 Denise Minger not only ignored these findings in her critique despite citing the mortality data from the same paper, but instead claimed in regards to a paper on rationing in Norway that animal foods did not decline until after cardiovascular disease mortality had already started declining.2 Minger misleads her readers by confusing the period when rationing was introduced with the period when the intake of animal foods declined. It can be deduced from the data from the Ministry of Supplies cited by Strom and Jensen that rationing was introduced as a result of a declining availability of such products, and therefore introduced after the intake of animal foods had already declined.1

Figure 1. Mortality from circulatory disease, correlated for age; consumption of fat in form of butter, milk, cheese and eggs, Norway 1938-48

Minger also misleads her readers into believing that there was almost an inverse relationship between the changes in animal protein intake and cardiovascular disease mortality in Norway during World War II by inaccurately reporting animal protein intake for the periods of 1936-37 and 1945. In order to verify Minger's interpretation of the statistics (Fig.3), please refer to the table below of macronutrient intake not present in Minger's post (Table 1), as well as the graph illustrating cardiovascular mortality rates (Fig. 2).1 2 In Minger’s own words, 'I pity da fool who doesn’t enlarge this image.'

Table 1. Macronutrient and micronutrient intake for Oslo men from 50 families, 1936-45 

Figure 2. Mortality from circulatory diseases, Norway 1927-48 

Figure 3. Denise Minger’s inaccurate interpretation of the Norwegian statistics 

Minger even posted a 'Fake Correlated Variable' graph, in an apparent attempt to ridicule Dr. Esselstyn, stating that 'For comparison’s sake, this is what a graph would look like if these variables were tightly linked'. Interestingly the 'Fake Correlated Variable' graph was actually remarkably consistent with the actual data (Figs. 4, 5).

Figure 4. Denise Minger’s 'Fake Correlated Variable' graph 

Figure 5. Actual animal protein intake and age-corrected circulatory disease mortality from the cited papers

In Minger’s critique she conveniently omitted the table from this study detailing animal protein intake despite posting the table of intake of individual food groups, and also failed to provide a free link to the paper claiming that she ‘couldn’t find any free copies to link’ despite one being easily locatable by googling the title of the paper, "Food Conditions in Norway during the War, 1939-45". These facts raise very serious questions as to whether Minger's inaccurate report of the data that appears to be heavily biased in favor of an agenda to promote animal foods was in fact intentional.

Minger also claimed that cardiovascular health did not actually improve in Norway during the war years, and that the decrease in cardiovascular mortality was obscured by an increase in mortality from infectious diseases. Minger appears to be either ignorant or unaware that Strom and Jensen provided additional data demonstrating that from over 15,000 operations carried out in Norway that were complicated by danger of thrombosis, the same surgeons found that the occurrence of these complications declined significantly during the period of deprivation of foods rich in animal fats, which then sharply increased after the resumption of intake.3 4 These findings provided strong evidence of actual improved cardiovascular health in Norway during the period of deprivation of animal foods. In Sweden where mortality from infectious diseases actually decreased during the war, there was a record decline in both cardiovascular disease and all-cause mortality during the war years when animal food intake decreased (Fig. 6).2 5 Other researchers also observed a striking decline in advanced atherosclerosis in Finland and Western Germany during the periods of deprivation of animal foods that returned to near pre-war levels after increasing intake.4 6

Figure 6. Percentage of energy from animal foods and mortality from arteriosclerosis and all-causes, Sweden 1940-1944

These observations from the World Wars are unlikely coincidental as they are consistent with the significant decline in serum cholesterol, and mortality from cardiovascular disease and all-causes in former communist nations of Eastern Europe, beginning in the early 1990's when the communist subsidies on meat and animal fats were abolished after the breakup of the Soviet Union (Fig. 7).7 8 Likewise, the significant decline in serum cholesterol, and mortality from cardiovascular disease and all-causes in the pre and early statin period of the second half of the century in developed nations throughout Western Europe, North America and Australasia is partly explained as a result of successful government policies that emphasized dietary changes, particularly a decreased intake of saturated animal fat. One of the best examples is Finland which experienced the most rapid decline of coronary mortality in the world, which was predominantly explained by a significant decline in serum cholesterol as the result of a large reduction in saturated animal fat and an increase in fruit and vegetable intake (Figs. 8, 9).9 10 

Figure 7. Trends in mortality from heart disease in former communist and western nations in men age ≤64

Figure 8. Observed and predicted declines in coronary mortality in males in Eastern Finland 

Figure 9. Observed and predicted decline in serum cholesterol based on dietary changes in Finish men and women without lipid-lowering medication (1, PUFA; 2, dietary cholesterol; 3, SFA; 4, PUFA + dietary cholesterol + SFA; 5, PUFA + dietary cholesterol + SFA + trans fatty acids; 6, observed serum cholesterol)

Randomized controlled trials provide further evidence of a causal association. A meta-analysis of 395 controlled feeding trials established that dietary cholesterol and isocaloric replacement of complex carbohydrates and unsaturated fat by saturated fat raises LDL and total cholesterol.11 In addition a meta-analysis of 108 randomized controlled trials of  various medical and dietary based lipid modifying interventions found that lowering LDL cholesterol significantly decreased the risk of coronary heart disease and all-cause mortality, while modifying HDL or triglycerides provided no clear benefit after controlling for LDL cholesterol.12

Not only does it appear that Denise Minger resorted to distorting the Norwegian data, she was even spineless enough to refer to the number of lives saved from cardiovascular mortality in Norway as being 'nothing to sneeze at' in an apparent attempt to downplay the importance of saving thousands of lives.

Dr. Caldwell Esselstyn and Treating the Cause of CAD

In regards to Dr. Esselstyn’s study of his initial coronary artery disease patients, Denise Minger misleads her readers into believing that 'half' the patients dropped out of the study by confusing the number of patients who had a follow-up angiogram with the number of adherent patients, simply ignoring the 7 patients who adhered to the diet but did not have a follow-up angiogram. There was actually a 75% adherence rate throughout most of this study, and in the more recent and larger decade long study of over 200 patients (known as Treating the Cause of CAD), there was an adherence rate of 91% (Vid. 1).13 14

Minger also suggested that Dr. Esselstyn’s results may have been due to luck as his study was an uncontrolled intervention study. Dr. Esselstyn however did compare the adherent and non-adherent patients. Despite having similar measurable amounts of disease at baseline as the other 18 patients, the 6 non-adherent patients had 13 new cardiac events within the first 12 years of the study despite the fact that they were still receiving standard care. On the other hand, the 18 compliant participants had no further cardiac events while being fully compliant, despite having 49 events during the 8 years prior to the study, of for which most of this time were receiving standard care.13 14 In the newer decade long study of over 200 patients, recurrent cardiac events only occurred in 0.5% of adherent participants, which is approximately 40 fold lower than other dietary or statin based trials (Vid. 1). Minger suggests that these results were due to luck but provided no evidence demonstrating that coronary artery disease can be spontaneously halted or reversed this frequently even when years of medical intervention have failed.

Video 1. TEDxCambridge - Caldwell Esselstyn on making heart attacks history

Dietary Cholesterol, Cardiovascular Disease and All-Cause Mortality

In regards to the information in Forks Over Knives about the disease promoting effects of dietary cholesterol, Denise Minger claimed that one of the reasons the consensus of the medical community that dietary cholesterol raises serum cholesterol and is unhealthy is due to experiments performed on ‘obligate herbivores’, primarily being ‘rabbits’. Minger is ignorant of the fact that literally hundreds of experiments on numerous different omnivorous species, the most relevant being non-human primates have demonstrated that dietary cholesterol has unfavorable effects on serum lipids and induces atherosclerotic lesions.15 Experiments on non-human primates have demonstrated that intake of even small amounts of dietary cholesterol as low as 43µg/kcal, the equivalent found in only half of a small egg in a human diet of 2,000 kcal induces atherosclerotic lesions. Furthermore, there was no evidence of a threshold for dietary cholesterol with respect to an adverse effect on arteries (Figs. 10, 11).16 [Click here for more information regarding study 16]

Figure 10. Subclavian artery from a Rhesus monkey supplementing 43µg/kcal dietary cholesterol. Sudanophilia (black area) is intense in the area of major intimal thickening.

Figure 11. Fermoral artery from a Rhesus monkey supplementing 43µg/kcal dietary cholesterol. Intimal fibrous thickening and disruption of internal elastic membrane differentiate this artery from control vessels of monkeys supplementing 0 dietary cholesterol.   

Minger also failed to mention that several large forward-looking prospective studies on humans found that dietary cholesterol was associated with a significantly increased risk of all-cause mortality, and that it has been consistently shown in studies on diabetic participants that intake of dietary cholesterol and eggs significantly increased the risk of cardiovascular disease and all-cause mortality.17 18 19 20 21 22

Protein Restriction and Healthy Longevity

Denise Minger suggested in regards to the original Indian study cited by Dr. Colin Campbell that in the presence of aflatoxins rats on low protein compared to high protein diets experience an increased risk of premature death. Minger appears to be ignorant in light of the fact that in the majority of studies on rats, especially those that have not been complicated by the administration of large doses carcinogens, protein restriction actually significantly increased maximum lifespan. For example, a review found that in 16 out of 18 studies protein restriction increased average maximum lifespan by approximately 20%, independent of caloric restriction.23 As for carbohydrate intake, increased intake has either been associated with no change or increased longevity.23 The association between protein restriction and longevity has been primarily attributed to methionine restriction, which has shown to increase both mean and maximum lifespan in rodents by on average up to 40%.23 24

Dietary restriction of methionine has also been shown to inhibit and even reverse human tumor growth in animal models and in culture demonstrating that tumors are methionine dependent, yet is relatively well tolerated by normal tissue.25 A review found that the benefit of replacing casein with soy protein on tumor suppression in the animal model was explained in part by the lower quantity of methionine and in part by numerous beneficial plant based compounds.26 For the sake of comparing 'apples and apples' as Minger put it, studies have found that casein is still far more cancer promoting compared to soy protein even when the casein and soy protein diets were formulated to contain equivalent amounts of the 'limiting amino acid' methionine (Fig. 12).26

Figure 12. Total number (A) and total weight (B) of mammory tumors in rats, 25 weeks after N-nitrosomethylurea injection. Diet Groups: Casein, 20% casein; SPI, 19% soy protein isolate; SPI +Met., 19% soy protein isolate formulated to contain the equivalent amount of methionine as the casein group

Compared to whole plant foods, both methionine content and bioavailability is significantly higher in most protein rich animal based foods, with little overlap.24 Therefore protein combining of unrefined plant foods will result in a quality sufficient to support normal tissue, but not the quality found in animal foods that promote cancer and premature death. These rodent studies are consistent with a number of prospective studies on humans that found that diets higher in protein and often fat, primarily of animal origin at the expense of vegetable protein or carbohydrates are associated with an increased risk of all-cause mortality.27 28 29 30

Minger suggested in regards to a study on non-human primates that in the presence of lower amounts of aflatoxins, higher compared to lower intakes of casein do not promote tumor growth. These findings are in disagreement with other studies that administered low amounts of aflatoxin cited by Dr. Campbell that Minger apparently ignored.31 However, the study on non-human primates did not test intermediate levels of protein intake or specific amino acids such as methionine, and Minger failed to cite any studies comparing casein with plant protein, therefore not allowing for a clear interpretation of these results. In studies on non-human primates, compared to casein, soy protein not only leads to genetic changes that are associated with a decreased risk of cancer, but also improvements in body weight, insulin sensitivity, lipid profile, and even decreases atherosclerosis plaques by on average up to 90% (Fig. 13).32 33 34 35

Figure 13. a, Proportion of each group of Cynomolgus monkeys with CAA plaques, defined as intimal thickness greater than half the medial thickness. b, Average lesion size for those monkeys with atherosclerotic plaques. Soy(-), Soy protein with phytoestrogens mostly extracted. Soy(+), Soy protein with phytoestrogens.

A number of randomized controlled trials have demonstrated the damaging effects of animal protein in human cancers. For example, a randomized, placebo-controlled trial found that among men at high risk, those supplementing with milk protein were more than six times likely to develop prostate cancer compared to men supplementing with soy protein.36 Also, a number of tightly controlled feeding trials with human participants have established that heme iron from the protein portion of meat increases the production of NOCs (N-nitroso compounds) in the digestive tract to concentrations similar to that found in cigarette smoke, of which most are cancerous.37 38 Furthermore, a controlled feeding trial found that NOCs arising from heme iron in meat forms DNA adducts in the human digestive tract, and DNA adducts are a well-established marker of cancer.39 These findings are consistent with recent meta-analyses of prospective studies that found that intake of both fresh red meat and heme from meat is associated with a significant increased risk of colorectal cancer.37 40 Based partly on these lines of evidence, in 2011 the expert panel from the World Cancer Research Fund reviewed over 1,000 publications on colorectal cancer and concluded that there is convincing evidence that both fresh and processed red meats are a cause of colorectal cancer.41 Furthermore, a more recent prospective study with over 2.24 million men and women found that compared to participants who consumed less than 1 serving per week, consuming 2 or more servings of meat significantly increased the risk of colorectal cancer.42

The China Study

Denise Minger suggested in regards to the raw data from the China Study that the counties who had the lowest serum cholesterol levels and had the lowest intakes of animal foods had an increased risk of mortality. However, in the China Study animal protein intake was very strongly associated with numerous favorable socioeconomic factors, with household income explaining between 60% and 80% of the variance of intake between counties, likely biasing towards such findings.  Animal food intake was also associated with other favourable socioeconomic factors including access to doctors and hospitals for antenatal consultation and child births, immunisation, avoidance of famine, owning a fridge, a toilet and the ability to read, of which many were associated to some degree with a lower risk of mortality.43 Among the younger population studied in the China Study II, animal food intake was actually a significant predictor of an increased risk, and plant foods of a decreased risk of all-cause mortality despite the fact that the significant inverse relationship between mortality and household income would have biased these results towards the opposite direction (Tables. 2, 3).43 This resembles Dr. Campbell’s observations in the Philippines where the children from the wealthier families that consumed diets rich in animal foods were more likely to develop liver cancer.31

Table 2. Significant predictors of all-cause mortality in the raw data from the China Study II, ages 0-4

Table 3. Significant predictors of all-cause mortality in the raw data from the China Study II, ages 5-14

Forward-looking prospective studies that controlled for socioeconomic factors found that plant based dietary patterns are associated with a decreased risk of all-cause mortality.27 44 45 46 Furthermore, it has been well established from evidence from over 100 randomized controlled trials that lowering LDL cholesterol significantly reduces the risk of all-cause mortality, even in individuals who already have very low baseline LDL cholesterol concentrations similar to that observed in the rural Chinese.12 47 The great majority of the surge in coronary heart disease mortality in Beijing between 1984 and 1999 has been attributed to a significant increase in serum cholesterol explained largely by a 5-fold increase in red meat and egg intake as well as a decline in fruit and vegetable intake. Without improvements in medical interventions the increase in deaths would have been substantially higher.48

Minger also previously claimed that Dr. Campbell’s findings of an relationship between fat, a marker of animal food intake, and an increased risk of breast cancer mortality in the China Study was attributed to the intake of ‘hormone-injected livestock’. She however provided no evidence that consumption of such livestock was widespread in rural China long enough before the mortality data was collected almost four decades ago for this questionable claim to be plausible. However, she did agree that Dr. Campbell’s findings of early menarche as a risk factor for breast cancer as perhaps reflecting a causal relationship ‘given what we know about hormone exposure and breast cancer’. Not surprisingly she failed to mention that animal protein was associated with elevated circulating estrogen in the China Study, and has been associated with a higher risk of early menarche in numerous studies including a cohort of girls born during the 1930s and 1940s, before the widespread consumption of ‘hormone-injected livestock’.31 49 50 51

In addition, Minger previously criticized a number of Dr. Campbell's statements that he made apparently in regards to both the China Study I & II, yet she cited data only from the China Study I.52 With the addition of the data from the China Study II, the relationship between animal foods and an increased risk of breast cancer mortality became significantly stronger, as did plant foods with a decreased risk (Table 4).43

Table 4. Significant predictors of female breast cancer mortality in the raw data from the China Study II, ages 35-69

Furthermore, consistent with the findings from the China Study, the expert panel from the World Cancer Research Fund concluded in 2011 that there is convincing evidence that dietary fiber protects against colorectal cancer, clearly refuting Minger's claims that research on dietary fiber 'outside' of the China Study does not support Dr. Campbell's findings.41 52

Many of Dr. Campbell’s findings in regards to plant based diets and the risk of chronic diseases in China are consistent with much earlier studies from China and around the world. For example, Williams reviewed the medical literature and documentations on cancer from around the world in 1908 long before the widespread use of intensive farming practices, finding strong evidence of an association between plant based dietary patterns and exceptional longevity and very low rates of cancer. Williams also documented that compared to the less affluent parts of Asia that subsisted on plant based diets, cancer was relatively common in the affluent parts of China that could afford animal foods on a frequent basis. He asserted that:53

…cancer is comparatively uncommon in those parts of China where the bulk of the people live on an almost exclusively vegetarian diet, being too poor to purchase any of the various flesh foods which are there used for culinary purposes.

Consistent with Williams's findings on cancer, Snapper found a similar phenomenon for vascular disease. He asserted that:54

In 1940, I confirmed De Langen’s results... by the observation that in North China, coronary disease, cholesterol [gall]stones and thrombosis were practically nonexistent among the poorer classes. They lived on a cereal-vegetable diet consisting of bread baked from yellow corn, millet, soybean flour and vegetables sautéed in peanut and sesame oil. Since cholesterol is present only in animal food, their serum cholesterol content was often in the range of 100 mg. per cent. These findings paralleled the observation of De Langen that coronary artery disease was frequent among Chinese who had emigrated to the Dutch East Indies and followed the high fat diet of the European colonists.

Overall Impressions of Forks Over Knives

Overall, Forks Over Knives provides a lot of very useful information to help viewers make life saving and longevity promoting dietary changes, and best of all comes directly from the doctors who have actually reversed many of the chronic diseases which are leading causes of disability and death. Ignoring the preponderance of evidence favoring a predominately plant based diet, low in saturated fat that is recommended by virtually every respected healthy authority around the world, and instead blindly following the unfounded dietary advice of the cholesterol skeptics can result in absurd consequences and a missed opportunity for healthy longevity.55


Part II: Forks Over Knives and Health Longevity: Perhaps the Science is Legit After All

Please post any comments in the Discussion Thread.