Saturated Fat and Heart Disease Meta-Analyses: Were Scientists Wrong?

Was it Ancel Keys or the low carb advocates
who fabricated data on saturated fat?

In a previous review, Clearing up the Confusion Surrounding Saturated Fat, I examined several important flaws and omissions in two meta-analysis studies which are frequently cited as evidence against the hypothesis that dietary saturated fat increases the risk of heart disease. The earlier study was published in 2010 by Siri-Tarino and colleagues in the American Journal of Clinical Nutrition, and the more recent study published by Chowdhury and colleagues in the Annals of Internal Medicine. These studies were both cited in a recent Time magazine article by Bryan Walsh as evidence to "eat butter" for health. In this earlier review I focused primarily on how the studies included in these meta-analyses, as well many other studies actually provide support for the hypothesis that saturated fat increases the risk of fatal forms of heart disease.

Although a number of prominent diet-heart researchers identified many serious flaws and omissions in these meta-analyses that cast doubt on the validity of the researchers conclusions, there are several other critical flaws related to how the estimates were derived for these meta-analyses that have not received adequate attention.1 2 3 4 5 This review will examine how a number of erroneous estimates may have biased the results and primary findings of these meta-analyses. Of particular concern, is in regards to how Siri-Tarino and colleagues derived negative estimates for three studies for which the original peer reviewed publications found positive associations between saturated fat and risk of heart disease. These studies and the estimates published in the meta-analysis by Siri-Tarino and colleagues are described in the table below. 

Study Name                                Risk Ratio (RR)    95% Confidence Interval (95% CI)

Framingham Heart Study6         0.92                       0.68, 1.24

Honolulu Heart Study7               0.86                       0.67, 1.12

Lipid Research Clinics Study8    0.97                       0.80, 1.18

Note: The risk ratio represents high vs low intake of saturated fat in relation to coronary heart disease. An estimate below 1 suggests a benefit of higher intake, and vice-versa. The estimates are considered statistically significant only if both the lower and upper 95% confidence intervals are on either side of 1.

The estimates published in the meta-analysis by Siri-Tarino and colleagues for all three of the abovementioned studies suggest a non-significant trend towards a benefit of saturated fat intake, greatly contrasting the findings from the original peer reviewed publications.9 This warrants an investigation into how Siri-Tarino and colleagues derived these negative estimates.

The Framingham Heart Study

Contrary to other studies included in the meta-analysis, the Framingham study reported estimated risk ratios for low vs high intake, rather than high vs low intake of saturated fat. Among the younger cohort in this study, 10% vs 15% of intake was associated with a 22% decreased risk of coronary heart disease (RR=0.78 [95% CI, 0.61-1.00]).6 Conversely this benefit of low intake was not observed among the older cohort (RR=1.06 [95% CI, 0.86-1.30]). Nevertheless, both of these estimates were controlled for serum cholesterol, which is expected to have biased these findings against showing an adverse effect of saturated fat. 

The tables in the meta-analysis clearly show that rather than converting the risk ratios for low vs high intake to appropriately reflect high vs low intake, Siri-Tarino and colleagues simply reported the estimates for 10% vs 15% of intake as representing 15% vs 10% of intake of saturated fat. The combined estimate for the younger and older groups in this cohort should therefore have been reported as RR=1.09 [95% CI, 0.81-1.47], and not as RR=0.92 [95% CI, 0.68-1.24], which was the estimate reported in the meta-analysis. The estimate provided in the meta-analysis for this study is clearly erroneous and biased the meta-analysis against showing an adverse effect of saturated fat.

The Honolulu Heart Study

In the original peer reviewed publication from the 10 year follow-up of the Honolulu Heart Study, a high intake of saturated fat was associated with a highly statistically significant increased risk of incidence of combined myocardial infarction (heart attack) and coronary heart death.7 This was despite over-adjustment for serum cholesterol. However, when combined with the incidence of the soft end points, angina and coronary insufficiency, the association was only statistically significant in the univariate analysis which did not control for serum cholesterol. The authors of this paper however, warned about misinterpreting the results for these soft end points, asserting:

When interpreting these results about whether the nutrients relate differently to the different manifestations of coronary heart disease, one should keep several problems in mind. The diagnostic certainty of the soft end points (angina pectoris or coronary insufficiency) is much less than that of the hard end points (myocardial infarction or coronary heart disease death).This could result in attenuation of a true relationship.

In the meta-analysis by Siri-Tarino and colleagues, the estimates for the 10 year follow-up of the Honolulu Heart Study were derived not from a peer reviewed publication, but rather from a separate data set obtained from the NHLBI. The primary reason that these estimates provided in the meta-analysis greatly contrasted with that of the peer reviewed publication can likely be explained by difference in the inclusion criteria of the participants. The tables in the meta-analysis show that the data used was based on 1177 cases of coronary heart disease from 8006 participants. However, in the peer reviewed publication there was a total of only 456 cases of coronary heart disease from 7088 participants.  Due to several important factors, the researchers of the original peer reviewed publication excluded 918 of the total 8006 participants from their analysis, asserting:

Men who reported that their intake was atypical for the day covered by the 24 hour recall, or who could not recall their intake, were excluded from the present analyses (n=502). Additionally, all men assessed as being prevalent cases of coronary heart disease (n=301), stroke (n=111), or cancer (n=49) were excluded from the analyses.

It is therefore evident that more than 60% of coronary events included in the meta-analysis for this particular study occurred in participants who had pre-existing disease at study baseline and/or reported unreliable dietary intake. As these additional participants made up less than 12% of the population, yet experienced more than 60% of the total events, this suggests the likelihood that these events occurred primarily among the participants with pre-existing cardiovascular disease, rather than the healthy participants reporting unreliable dietary intake. It should be noted that Siri-Tarino and colleagues clearly stated that one of the inclusion criteria for the meta-analysis was that “study participants were generally healthy adults at study baseline”, suggesting that the data they selected did not meet their own criteria.

When considering the fact that the estimate for this study used in the meta-analysis was negatively associated with coronary heart disease, rather than positively associated as was the case in the peer reviewed publication, this suggests that the participants excluded in the peer reviewed publication who experienced coronary events, largely being those with pre-existing disease, were reporting a lower intake of saturated fat. This is concerning considering the evidence that people who are diagnosed of being at risk of cardiovascular disease tend to limit saturated fat intake post-diagnosis. For example, in the earliest study included in the meta-analysis, participants diagnosed with high cholesterol had a tendency to reduce intake of saturated fat in order to improve risk factors.10 It should therefore not be surprising that participants with established cardiovascular disease likely reported a lower intake of saturated fat in this study, suggesting that a lower reported intake was only a marker of more disease, rather than a cause of it. This phenomenon is known as reverse causality, and is one of the important reasons why participants with pre-existing disease are either excluded, or examined separately from healthy participants in such studies. The inclusion of such data in a meta-analysis could significantly distort the findings. 

It is not entirely clear why a null estimate was reported for this study in the more recent meta-analysis published by Chowdhury and colleagues. However, the sample size described in supplement section of the meta-analysis paper suggests that the estimates were based on the same misleading data used by Siri-Tarino and colleagues, and not the original peer reviewed publication.11

The Lipid Research Clinics Study

In the Lipid Research Clinics study, it was observed that among the younger cohort that a 1% increased intake of energy from saturated fat was associated was a statistically significant 11% increased risk of death from coronary heart disease (RR=1.11 [95% CI, 1.04-1.18]).8 However, similar to the Framingham study, this study also included over-adjustments for serum cholesterol and the adverse effect of a high intake of saturated fat was not apparent among the older cohort (RR=0.96 [95% CI, 0.88-1.05]).

In the meta-analysis by Siri-Tarino and colleagues, the estimate for both the younger and older cohorts combined using the random-effects model was reported as RR=0.97 [95% CI, 0.80-1.18]. It is clear that this estimate is erroneous. When using the random-effects model for two estimates, the estimate with the narrowest confidence intervals should carry more weight, and therefore the combined estimate should have been much closer to the estimate for that of the younger cohort. The estimate for a 1% increase of energy from saturated fat should have been reported as RR=1.04 [95% CI, 0.90-1.19]. Similar to the abovementioned studies, this erroneous estimate biased the meta-analysis against showing an adverse effect of saturated fat.

Reanalysis of the Data

The use of erroneous estimates for several studies included in the meta-analysis by Siri-Tarino and colleagues warrants a reanalysis of the original data. I therefore performed a revised meta-analysis of the same 16 prospective cohort studies included in this meta-analysis. The methods used to derive the estimates have been described previously, and were mostly consistent with those used by Siri-Tarino and colleagues. For the Honolulu Heart Study, I derived the estimates based on only the hard coronary end points due to the concern of the accuracy of the estimates for the soft end points being less certain. As the exact P-value was not available for this publication, and indicated only as being between 0.01 and 0.001, I chose to derive the corresponding standard error using a P-value of 0.01, as this was the most conservative estimate.7

In a meta-analysis based on the 16 studies included in the meta-analysis published by Siri-Tarino and colleagues, dietary saturated fat intake was associated with a statistically significant 16% increased risk of coronary heart disease (Fig. 1).

FIGURE 1. Risk ratios and 95% CIs for fully adjusted random-effects models examining associations between saturated fat intake in relation to incidence of coronary heart disease. ¹Studies that included adjustments for serum or LDL cholesterol. SAT, saturated fat intake.

I also carried out an updated meta-analysis, including prospective cohort studies published up until November 2014 that provided the necessary data to derive risk ratios and the corresponding 95% confidence intervals. If a study was published multiple times, the estimate for the longest period of follow-up was used. A total of 21 studies were included, including recent publications not included in either the meta-analyses by Siri-Tarino and colleagues, and Chowdhury and colleagues.7 8 9 12 13 14 15 16 17 18 19 20 21 22 23 24 25 In a meta-analysis of 21 studies, dietary saturated fat was associated with a statistically significant 15% increased risk of coronary heart disease (Fig. 2).

FIGURE 2. Risk ratios and 95% CIs for fully adjusted random-effects models examining associations between saturated fat intake in relation to incidence of coronary heart disease. ¹Studies that included adjustments for serum or LDL cholesterol. SAT, saturated fat intake.

The findings from these meta-analyses presented here are compatible with the findings from a broad range of evidence described previously. This includes findings from randomized controlled trials showing an adverse effect of saturated fat on blood cholesterol and arterial function, as well as the demonstrated unequivocal causal relationship between diets rich in cholesterol and saturated fat, and the development of atherosclerosis in nonhuman primates, among dozens of other animal species. Furthermore, these findings are supported by numerous ecological studies, including the Seven Countries Study. 

The estimates of this meta-analysis are clearly stronger than that of the Siri-Tarino and Chowdhury meta-analyses, neither of which produced statistically significant estimates. One of the primary reasons for these contrasting estimates was due to the correction of the estimates for several studies described above. These contrasting estimates can also in part be explained by the correction of erroneous estimates for the confidence intervals reported in the Siri-Tarino meta-analysis for the Ireland-Boston Diet Heart Study, and in the Chowdhury meta-analysis for the EPIC-Greece study. It is clear that in the original peer reviewed publications that the estimates in these studies were statistically significant to the 95% confidence level, yet non-statistically significant estimates were reported in the meta-analyses for these studies, biasing against showing an adverse effect of saturated fat.13 20

Another important difference in the meta-analyses presented here, is that the estimates for studies reporting estimates as a 1% increase of energy were transformed to represent a 5% increase of energy from saturated fat. In the Siri-Tarino meta-analysis the researchers either simply reported the estimates for a 1% increase, or transformed the estimates to represent a similarly small increase of energy. Similarly, Chowdhury and colleagues multiplied the estimates from these studies by the power of 2.18, effectively representing the effect of an increase of only a 2.18% increase of energy. The researchers rational for this was that this equation would be expected to show the effect of mean top vs bottom third of intake. While this equation may be suitable for some studies using different scales of measurement, it requires a leap of faith to assume a difference of only 2.18% of energy represents high vs low intake in these studies. Reporting estimates as a 5% increase of energy, as was done here would make differences for high vs low intake much more comparable to that of the other included studies, while still being moderately conservative. As the studies that reported estimates as a 1% increase of energy were more likely to show a positive relationship between saturated fat and coronary heart disease, failing to transform the estimates to represent a sufficient change in intake may result in minimizing the statistical power of these studies, and in-turn biasing a meta-analysis against showing an adverse effect of saturated fat.

Conflicts of Interest

The meta-analyses published by Siri-Tarino and colleagues and Chowdhury and colleagues contained erroneous estimates for several positive studies which in-turn biased against showing an adverse effect of saturated fat. Given the fact that these researchers were well informed in this area of research, it is difficult to accept that they were simply unaware of any of the issues described here. In the meta-analysis by Siri-Tarino and colleagues there was clear evidence of potential conflicts of interest. The meta-analysis was funded in part by the National Dairy Council, and the senior researcher, Ronald Krauss had reported receiving grants from the National Dairy Council, the National Cattleman’s Beef Association and the Robert C. and Veronica Atkins Foundation. Similarly, several of the researchers of the meta-analysis by Chowdhury and colleagues have reported receiving grants from the food industry. Suggestive evidence of these researchers intention to downplay the role of saturated fat on fatal heart disease have also been described previously.

Although receiving grants from a particular industry does not necessarily negate the findings of a study, when errors are made that bias the study results in favor of the concerned industry, the intentions of the researchers should be questioned. The lines of evidence described here lends support to the likelihood that these researchers put their own interests before that of the general public, driving the public to follow dangerous dietary patterns at the hands of fad diet advocates who promote these studies.

Putting Data into Context

It is important to note that the influence that saturated fat has on the risk of disease is not primarily determined by intake per se, but by which foods saturated fat is substituted for. As the intake of dietary fiber was universally low among participants in these studies, this suggests that participants consuming diets lower in saturated fat were substituting saturated fat primarily with lean animal foods and heavily processed foods.26 In addition, as dietary fiber was associated with a decreased risk of coronary heart disease in a number of these studies, the estimates described here therefore could be a significant underestimation of the effect of replacing saturated fat with more healthful, fiber rich foods.26 Furthermore, in the studies included in this meta-analysis, the difference for high vs low intake of saturated fat was relatively low, often only ranging between about 5% and 10% of energy. This suggests that individuals following popular diets which emphasize far greater intakes of saturated fat than recommended levels may be at a much greater risk.

It is also important to note that the effect that a particular food has on the risk of coronary heart disease cannot be fully explained by the amount of one particular nutrient, but rather by multiple nutrients that likely operate together in a complex manner to modify the risk of disease. Therefore, it would be more appropriate to compare the substitution of different foods, rather than focusing entirely on substituting single nutrients. Focusing attention on recommending healthy dietary patterns that are naturally low in saturated fat, while rich in dietary fiber and other beneficial nutrients; primarily, minimally processed, plant-based diets would likely be a more effective measure to improve overall dietary quality, resulting in greater improvements to heart health compared to the more contemporary reductionist approach of focusing on modifying single nutrients. The effectiveness of such a diet was recently demonstrated again by Caldwell Esselstyn in a follow-up of 200 high-risk patients. In this study, coronary artery disease was either arrested or reversed in the great majority of adherent patients, clearly contrasting that of any other peer-reviewed study of similar size.27

This review demonstrates that the conclusions of several meta-analysis studies which suggest that dietary saturated fat unlikely increases the risk of heart disease are misleading, and that the current evidence supports the recommendations to replace foods rich in saturated fat with minimally refined plant based foods. Recommendations based on the findings of these meta-analyses made by the media and low carb advocates to consume more saturated fat rich foods are therefore unsubstantiated and likely dangerous. While it may make an interesting read being told that scientists, such as Ancel Keys have intentionally deceived us into believing that saturated fat-rich foods are unhealthy, it is appears that it may actually be the authors of such articles who lack in the way of honesty. 

Study acronyms: ATBC, Alpha-Tocopherol Beta Carotene Study; BLSA, Baltimore Longitudinal Study of Aging; EPIC-Greece, European Prospective Investigation into Cancer Greece; EUROASPIRE, European Action on Secondary and Primary Prevention through intervention to reduce events; FHS, Framingham Heart Study; GPS, Glostrup Population Study; HHS, Honolulu Heart Study; HLS, Health and Lifestyle Survey; HPFS, Health Professionals' Follow-Up Study; IBDH, Ireland-Boston Diet Heart Study; IIHD, Israeli Ischemic Heart Disease Study; JACC, Japan Collaborative Cohort Study; JPHC, Japan Public Health Center Based Study; KIHD, Kuopio Ischaemic Heart Disease Risk Factor Study; LRC, Lipid Research Clinics; MALMO, Malmo Diet and Cancer Study; NHS, Nurses' Health Study; SHS, Strong Heart Study; WES, Western Electric Study.

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.

Clearing Up The Confusion Surrounding Saturated Fat

In 2010, Siri-Tarino and colleagues published a meta-analysis of prospective cohort studies evaluating the association between dietary saturated fat and cardiovascular disease in the American Journal of Clinical Nutrition.1 Based on the results of this meta-analysis, these researchers concluded that there was insufficient evidence from prospective cohort studies to conclude that dietary saturated fat increases the risk of coronary heart disease. However, a number of prominent diet-heart researchers identified many serious flaws and omissions in this meta-analysis that cast doubt on the validity of these researchers conclusions.2 3 4 5 6

More recently, Chowdhury and colleagues published a separate meta-analysis in the Annals of Internal Medicine, and reached similar conclusions to that of Siri-Tarino and colleagues regarding the association between saturated fat and coronary heart disease.7 Unfortunately, this meta-analysis also failed to sufficiently address a number of important limitations that it shares with the meta-analysis by Siri-Tarino and colleagues. Furthermore, in this meta-analysis, although positively, but not significantly associated in the random-effects model, both dietary and total circulating concentrations of saturated fat were associated with a small, but statistically significant increased risk of coronary heart disease in the fixed effects model (RR=1.04 [95% CI, 1.01, 1.07] and RR=1.13 [95% CI, 1.03-1.25], respectively). These significant findings were however ignored in the conclusions of this study. Nevertheless, the media and proponents of popular Low-Carb and Paleo diets have repeatedly cited these meta-analyses as evidence to support a diet rich in saturated fat. 

Saturated Fat and Coronary Heart Disease Mortality

In the editorial to the Siri-Tarino meta-analysis, Jeremiah Stamler noted that saturated fat intake was more strongly associated with fatal than non-fatal incidence of coronary heart disease. Stamler calculated that based on the 11 studies included in the meta-analysis which provided estimates specifically for fatal cases, saturated fat was associated with a 32% increased risk of death from coronary heart disease, when weighted by person-years of exposure.2 Siri-Tarino and colleagues noted this concern in a follow-up paper, but instead downplayed these findings by asserting that in their own analysis of only 7 studies, saturated fat intake was associated with only a borderline significant 18% increased risk of death from coronary heart disease, when using the random effects model (RR=1.18 [95% CI 0.99-1.42]).8 Similarly, in the more recent meta-analysis, Chowdhury and colleagues found that in their sub-analysis of only 9 studies, saturated fat intake was associated with a borderline significant 7% increased risk of death from coronary heart disease (RR=1.07 [95% CI, 1.00-1.13]).7

It can be deduced by the estimates and the sample size of these sub-analyses by both Siri-Tarino and Chowdhury, that only the studies that provided estimates specifically only for fatal cases were included. Therefore these sub-analyses excluded studies that provided estimates for fatal cases in additional to that of total incidence of coronary heart disease. In total, 14 prospective cohort studies provided estimates for death from coronary heart disease, of which 3 were not included in the original meta-analysis by Siri-Tarino and colleagues.1 9 10 11 12 13 14 15 16 17 18 19 20 21 22

The exclusion of several studies in these analyses warrants a reanalysis of the studies evaluating the association between saturated fat and the risk of death from coronary heart disease. I therefore performed a meta-analysis including all 14 studies for which estimates were available specifically for death from coronary heart disease. Similar to the methods of Siri-Tarino and colleagues, I chose to compare extreme quantiles of saturated fat intake where possible. However, for the studies which estimates were provided as either a 1% increase of energy, or as a 1-unit increase, the estimates were transformed to represent roughly a 5% increase in energy from saturated fat, as this was similar to the difference for high vs low quantiles of intake in the other studies included in this meta-analysis. In order to ensure that the methods used for the statistical analysis were consistent with that used by Siri-Tarino and colleagues, I performed the meta-analysis in Review Manager (from The Cochrane Collaboration), and pooled the estimates using the random effects model for both within-study and between-study variation. Similarly, risk ratios and 95% confidence intervals were log transformed to derive the corresponding standard error for beta-coefficients by using Greenland’s formula.23 Otherwise, the exact P-value was used where available to derive the corresponding standard error.

In a meta-analysis including 14 studies, dietary saturated fat intake was associated with a statistically highly significant 24% increased risk of death from coronary heart disease (Fig. 1). Similarly, for the 11 studies included in the Siri-Tarino meta-analysis, saturated fat was associated with a statistically highly significant 26% increased risk of death from coronary heart disease (RR=1.26 [95% CI, 1.14-1.40]). 

FIGURE 1. Risk ratios and 95% CIs for fully adjusted random-effects models examining associations between saturated fat intake in relation to coronary heart disease mortality. ¹Studies that included adjustments for serum or LDL cholesterol. SAT, saturated fat intake.

It is well established that saturated fat raises serum and LDL cholesterol, and that these blood lipids increase the risk of coronary heart disease.24 25 26 27 However, in this meta-analysis, almost 40% of the weight was derived from studies that controlled for either serum or LDL cholesterol. Therefore, the inclusion of these studies would be expected to have bias these findings towards null.2 In a sub-analysis excluding the 6 studies that controlled for either serum or LDL cholesterol, saturated fat was associated with a statistically significant 30% increased risk of death from coronary heart disease (Fig. 2). Interestingly, in a sub-analysis including only the 6 studies which controlled for either serum or LDL cholesterol, saturated fat was associated with a statistically significant 18% increased risk of death from coronary heart disease (RR=1.18 [95% CI, 1.01-1.37]). This suggests that the adverse effects of saturated fat may extend beyond on simple measures of blood lipids.

FIGURE 2. Risk ratios and 95% CIs for fully adjusted random-effects models examining associations between saturated fat intake in relation to coronary heart disease mortality. SAT, saturated fat intake.

As coronary heart disease is the leading cause of death in the world, naturally these findings should be a cause for concern.28 Nevertheless, both the Siri-Tarino and Chowdhury meta-analyses are widely cited by proponents of Low-Carb and Paleo diets as providing compelling evidence in favor of a diet rich in saturated fat. It is important to note, however, that in the studies included in this meta-analysis, the difference for high vs low intake of saturated fat only ranged between about 5% and 10% of energy. This suggests that individuals following popular variants of these diets which often emphasize far higher intakes of saturated fat than recommended levels may be at a much greater risk of death.

It is important to note that the influence that saturated fat has on the risk of disease is not primarily determined by intake per se, but by which foods saturated fat is substituted for. As the intake of dietary fiber was universally low among subjects in these studies, this suggests that subjects consuming diets lower in saturated fat were substituting saturated fat primarily with lean animal foods and heavily processed foods.29 As dietary fiber was associated with a decreased risk of death from coronary heart disease in a number of these studies, this suggests that compared to fiber-rich foods, foods rich in saturated fat may be associated with an even stronger risk of coronary heart disease death.29

Although in this meta-analysis, the Israeli Ischemic Heart Disease Study appeared the least favorable of the hypothesis that saturated fat increases the risk of death from coronary heart disease, it should be noted that not only were the estimates controlled for serum cholesterol, in this study, saturated fat as a percentage of fat was actually associated with a statistically significant increased risk of death from coronary heart disease. In addition, subjects who were classified as being most adherent to religious Orthodoxy, which is typically accompanied by fasting periods in which the consumption of meat and other foods rich in saturated animal fat are prohibited, experienced a significantly lower death rate of coronary heart disease.11 This observation is supported by several other studies which found that Orthodox fasting is associated with improved cardiovascular risk factors, including blood lipids.30

The findings from this meta-analysis are in agreement with the demonstrated unequivocal causal relationship between diets rich in cholesterol and saturated fat, and the development of atherosclerosis in nonhuman primates, among dozens of other animal species. It had also been demonstrated that such diets cause heart attacks, and even cardiovascular related deaths in nonhuman primates at a rate similar to that of high-risk populations living in developed nations.31

The findings from this meta-analysis are also in agreement with numerous longitudinal ecological studies. For example, intake of saturated fat explained about 88% of the variance in death from coronary heart disease between the 16 cohorts in the 25-year follow-up of the Seven Countries Study.32 Similar estimates were also found for foods rich in saturated fat, including butter, meat, and animal foods combined.33 Similarly, in 1989, Epstein examined the changes in death from coronary heart disease in 27 countries during the previous 10 to 25 years, and noted that:

In almost all of the countries with major falls or rises in CHD mortality, there are, respectively, corresponding decreases or increases in animal fat consumption...

Epstein also noted that a number of other risk factors, such as smoking could not explain these findings, as the prevalence of smoking among women either remained largely unchanged or increased in most nations during this period, yet similar declines in death were often observed in both men and women.34 Epstein's findings are further supported by a number of studies that have incorporated the IMPACT CHD mortality model, which has been shown to adequately explain which risk factors and treatments that have contributed most significantly to the changes of rates of coronary heart disease mortality throughout most parts of the world.35

Dietary Patterns and Coronary Heart Disease Mortality

Dietary patterns characterized by high or low intakes of saturated fat may provide indirect evidence of the effect saturated fat has on the risk of death from coronary heart disease. For example, a recent meta-analysis of prospective cohort studies by Noto and colleagues found that both low-carbohydrate, and low-carbohydrate, high-protein diets, which highly correlated with saturated fat intake were associated with a statistically significant increased risk of death from all causes combined.36 Conversely, the findings for death from cardiovascular disease, although positive, were not statistically significant. However, several of the studies controlled for saturated intake, and sub-analyses in several of the studies found that the excess risk of death was greater when either saturated fat intake was above the median, or when the diets were classified as being animal based.37 38 Similarly, a sub-analysis in one of the studies found that the association with death was stronger after excluding nonadequate dietary reporters.39

These sub-analyses would likely allow for a clearer interpretation of the effects of a carbohydrate restricted diet rich in saturated fat. Unfortunately, estimates based on these sub-analyses were not provided in the meta-analysis by Noto and colleagues. Therefore, I performed a meta-analysis based on the studies included in the meta-analysis by Noto and colleagues, while using the estimates for the sub-analyses described above where possible.37 38 39 40 41 For the statistical analysis I used the same methods described by Noto and colleagues.36

In this meta-analysis, a high low-carbohydrate score was associated with a statistically significant 15% increased risk of death from cardiovascular disease, for which only 2 of 4 cohorts did not control for saturated fat intake (Fig. 3). On the other hand, a high low-carbohydrate, high-protein score was associated with a statistically significant 100% increased risk of death from cardiovascular disease, for which 3 of 4 studies did not control for saturated fat intake (Fig. 4). 

FIGURE 3. Risk ratios and 95% CIs for fully adjusted random-effects models examining associations between low-carbohydrate diets in relation to cardiovascular disease mortality. ¹Studies that included adjustments for saturated fat intake.

FIGURE 4. Risk ratios and 95% CIs for fully adjusted random-effects models examining associations between low-carbohydrate, high-protein diets in relation to cardiovascular disease mortality. ¹Studies that included adjustments for saturated fat intake. LCHP, low-carbohydrate, high-protein diet.

As coronary heart disease is the number one cause of cardiovascular death in the nations where these studies were carried out, this provides indirect evidence that diets rich in saturated fat, at least in the context of a carbohydrate restricted diet, increases the risk of coronary heart disease. Furthermore, the difference in intake of saturated fat between the low and high low-carbohydrate scores was generally smaller than the difference of intake between popular low-carbohydrate diets and recommended levels, suggesting that individuals who follow more extreme variants of these diets may be at an even greater risk of death. As reviewed previously, these findings may be explained, in part, by a number of adverse effects that carbohydrate restricted diets have been shown to exert on cardiovascular risk factors. For example, recent meta-analyses of randomized controlled trials have found that compared to diets rich in nutrient poor, low-fiber carbohydrates, carbohydrate restricted diets raise LDL cholesterol and impair flow-mediated dilatation.42

Findings from prospective cohort studies comparing vegetarians characterized by consuming moderately low saturated fat diets and health conscious omnivores may provide further indirect evidence of the adverse effects of saturated fat. I showed previously in a meta-analysis of 7 prospective cohort studies that compared to vegetarians, health conscious omnivores experienced a statistically highly significant 32% increased risk of death from coronary heart disease (Fig. 5).43 44 45 46 47

FIGURE 5. Risk ratios and 95% CIs for fully adjusted random-effects models examining associations between omnivorous diets in relation to coronary heart disease mortality.

It is important to note that the omnivores in these studies had a relatively low intake of meat, suggesting that individuals following popular meat based diets may be at a greater risk of death. This suggestion is supported by a recent meta-analysis of prospective cohort studies which found that an increment of 1 mg/day of heme iron, found only in animal tissue is associated with a 27% increased risk of coronary heart disease.48 As reviewed previously, in these studies, the degree of reduction in risk of death from coronary heart disease observed in vegetarians in these studies was generally in proportion to the expected reduced risk based on the differences in levels of total and non-HDL cholesterol. This suggests that these results may, in part, be explained by differences in intake of saturated fat.

The findings of a pooled-analysis of 11 prospective cohort studies by Jakobsen and colleagues suggested that replacing saturated fat with polyunsaturated fat, but not monounsaturated fat or carbohydrate, was associated with a significantly decreased risk of death from coronary heart disease.49 However, the researchers pointed out that these findings should be interpreted with caution, as the main dietary source of monounsaturated fat in these cohorts was animal fat, whereas the quality of carbohydrate was not considered. In this pooled-analysis, dietary fiber intake was controlled for, essentially removing a primary benefit of replacing foods rich in saturated fat with carbohydrate. In a different pooled-analysis including virtually the same studies, an increment of 10 g/day of dietary fiber was associated with a 27% decreased risk of death from coronary heart disease.23 This suggests that replacing saturated fat with the equivalent energy from fiber-rich carbohydrate would likely be associated with a significantly reduced risk of death from coronary heart disease. This suggestion is supported by a different meta-analysis which found that an increment of about 2 servings a day of whole grains was associated with a 22% decreased risk of death from cardiovascular disease.50 Interestingly, even Siri-Tarino and colleagues concluded in a more recent paper that saturated fat should be replaced with polyunsaturated fat and whole grains in order to reduce the risk of cardiovascular disease.7

Saturated Fat is a Major Problem

The findings reviewed here support the hypothesis that saturated fat increases the risk of coronary heart disease mortality. Furthermore, as reviewed previously, evidence also suggests that the hazardous effects of diets rich in saturated fat are also applicable to diets rich in organic, grass-fed animal foods. However, saturated fat is only one of a number of problems as far as chronic diseases are concerned. The effect that a particular food has on the risk of coronary heart disease cannot be fully explained by saturated fat content alone, but rather by multiple nutrients that likely operate together in a complex manner to modify the risk of disease. Therefore, it may be more appropriate to focus attention on recommending healthy dietary patterns that are naturally low in saturated fat, while rich in dietary fiber and other beneficial nutrients; primarily, minimally processed, plant-based diets. Such a focus may be more effective to help lower the intake of saturated fat, while simultaneously improving overall dietary quality compared to the more contemporary reductionist approach of focusing on modifying single nutrients.

In forthcoming parts of this review, I will examine both the effects of dietary and total circulating concentrations of saturated fat on the risk of total incidence of coronary heart disease. In addition, I will examine a number of other important limitations of the studies included in these meta-analysis that may have bias these findings towards null.2 3 4 5 6


Study acronyms: ATBC, Alpha-Tocopherol Beta Carotene Study; BLSA, Baltimore Longitudinal Study of Aging; EPIC-Greece, European Prospective Investigation into Cancer Greece; EUROASPIRE, European Action on Secondary and Primary Prevention through intervention to reduce events; FHS, Framingham Heart Study; HLS, Health and Lifestyle Survey; HPFS, Health Professionals' Follow-Up Study; IBDH, Ireland-Boston Diet Heart Study; IIHD, Israeli Ischemic Heart Disease Study; JACC, Japan Collaborative Cohort Study; LRC, Lipid Research Clinics; MALMO, Malmo Diet and Cancer Study; NHS, Nurses' Health Study; SHS, Strong Heart Study; SWLHC, Swedish Women’s Lifestyle and Health Cohort; ULSAM; Uppsala Longitudinal Study of Adult Men; VIP, Västerbotten Intervention Program; WES, Western Electric Study.


Please post any comments in the Discussion Thread.

Do Vegetarians Live Longer Than Health Conscious Omnivores?

Proponents of Paleo and Low-Carb diets suggest that plant-based diets, particularly those rich in grains and legumes, promote disease, ultimately resulting in premature death. However, there is a substantial amount of evidence casting doubt on such suggestions, with many studies providing evidence that plant-based diets increase longevity. This review will focus on studies examining the longevity of plant-based populations, and some of the criticisms of these studies, particularly in reference to Denise Minger's recently published book, Death By Food Pyramid.

It is not news that Denise Minger has a tendency to downplay the health benefits of plant foods and plant-based diets. In her critique of the China Study, Minger claimed that “as a plant-nosher”, she was hoping to find evidence to support Dr. T. Colin Campbell's findings from the China Study linking dietary fiber to lower rates of colorectal cancer.1 Somehow, however, despite her vegan bias apparently creeping into her critique, Minger suggested that she was unable to find sufficient evidence outside of the China Study supporting the hypothesis that dietary fiber protects against colorectal cancer. And yet, several months later the omnivorous panel of experts of the World Cancer Research Fund concluded based on a review of over 1,000 publications that there was convincing evidence that dietary fiber protects against colorectal cancer.2 In Death By Food Pyramid, Minger continues this trend of downplaying the health benefits of plant-based diets. 

Failing to Equal the Seventh-day Adventists

In the chapter of her book, Herbivore’s Dilemma, Denise Minger provides a brief overview of the history and the growth of the popularity of vegetarian diets, bringing into picture the earliest of the studies on the Californian Seventh-day Adventists. Loma Linda, California which is highly concentrated by Adventists is considered to be a Blue Zone because of the greater life expectancy compared to other parts of North America. Loma Linda shares the title of Blue Zone with four other populations which are all characterized by traditionally consuming plant-based diets, typically rich in legumes and grains.3 These other Blue Zones include, Ikaria, Greece; Nicoya, Costa Rica; Okinawa, Japan; and Sardinia, Italy. It seems that Minger was not even able to get these simple details right in her book, claiming that the Greek island of Crete is considered a Blue Zone, while citing an article that clearly refers to Ikaria.

Minger hypothesizes that the longevity of the Adventists maybe unrelated to their low meat diet, and may rather reflect the discouraged use of tobacco and alcohol. To illustrate this, Minger points out that the Mormons who are also discouraged from the use of tobacco and alcohol, "but whose founder never endured any meat-abstinence visions"[p.194] have a greater life expectancy than the average population. Although Joseph Smith, Jr., the founder of Mormonism may not have demanded followers to completely abstain from meat, the Word of Wisdom (section 89 in The Doctrine and Covenants) which he delivered, apparently received as a revelation from god, states that:4

Yea, flesh also of beasts and of the fowls of the air, I, the Lord, have ordained for the use of man with thanksgiving; nevertheless they are to be used sparingly;

In regards to longevity, Minger then goes onto state that:

But what’s even more telling is the fact that meat-eating Mormons and vegetarian Adventists tend to live equally as long. When compared to ethnically matched folks outside their religious groups, both Adventist and Mormon men—once their birthday-cake candles start numbering in the thirties—can expect to live about seven years longer than the rest of the population.[p.194]

Unlike what Minger appears to have the reader believe, the 7 years greater life expectancy referred to in the study she cites is for the average Adventist male, and not specifically for vegetarian Adventist men. Less than one third of the men in the cited study were considered vegetarians, with more than half considered regular meat eaters.5 What these studies specifically found was that the life expectancy for active Mormon men who reached the age of 35 was about 7.5 years greater than the average U.S. white male, whereas the life expectancy of the average Californian Seventh-day Adventist male who reached the age of 30 was about 7.3 years greater than the average Californian white male.5 6 When specifically looking at the average vegetarian Californian Adventist male, their life expectancy was found to be about 9.5 years greater than the average Californian white male. It is important to note that the Californian whites that the Adventists were compared to have one of the highest life expectancies of any American state, and are expected to live up to a year longer than the average U.S. white that the Mormons were compared to.7

The more recent 25 year follow-up of the Mormons may allow for a more informative comparison, as like the Adventist study, it included both men and women from California, and examined the effects of other lifestyle factors on mortality. Mormon men and women over the age of 25 with four favorable lifestyle factors associated with significantly reduced mortality were expected to live about 9.8 and 5.6 years longer, respectively, compared to U.S. whites.8 In comparison, vegetarian Adventist men and women over the age of 30 with three favorable lifestyle factors were expected to live about 13.2 and 8.9 years longer, respectively, compared to non-Adventist Californians whites.5 9 Compared to the average U.S. white however, this difference in life expectancy would be expected to be closer to about 14 and 10 years.7 However, and more importantly, a later paper on the Californian Adventists found that those who adhered to a vegetarian diet for at least 17 years were expected to live 3.6 years longer than those who adhered for fewer years.10 This suggests that when restricting the analysis to long-term vegetarian Adventists, the difference in life expectancy compared to the Mormons would be even greater.

There are also other lines of evidence lending support to the observed greater life expectancy of the vegetarian Seventh-day Adventists compared to health conscious Mormons. For example, one study in which vegetarian Adventists and Mormons were matched for strength of religious affiliation, and consumption of tobacco, alcohol, tea and coffee, the vegetarian Adventists were found to have significantly lower levels of serum cholesterol, blood pressure and rates of obesity (Fig. 1).11 12 The difference in blood pressure remained significant even after controlling for BMI, and could not be explained by differences in sodium intake. Another study also found that vegetarian Adventists had lower blood pressure than Mormons, and that the difference increased with age, suggesting a greater favorable effect of long-term adherence to a flesh-free diet.12 

Figure 1. All percentiles of distribution of blood pressure were found to be lower in the vegetarian Adventists compared to health conscious omnivorous Mormons.

Although some of the many factual errors in Minger’s book may be passed off as sloppy research rather than as being intentional (one example perhaps being when she confuses Crete for Ikaria), given the number of occasions she has discussed the Adventist studies previously, it is difficult to believe that she was truly being honest and simply was not even aware of the data in the very studies she cites. Furthermore, Minger acknowledges in this book that she had others with knowledge in this field review her manuscript and assist her with the completion of this book (almost exclusively individuals who have demonstrated an anti-vegetarian stance), allowing for little excuse for these misleading statements.

Mortality in Vegetarians and Health Conscious Omnivores

As there are no published meta-analyses examining all of the current available prospective cohort studies comparing mortality in vegetarians and health conscious non-vegetarians, I performed a simple meta-analysis using the results for the fully adjusted model from the most recent follow-up of each cohort. For mortality from all-causes, based on 7 cohorts, the vegetarian group had a statistically significant 7%, and a borderline significant 6% reduced risk, using the fixed effects and random effects models, respectively (Fig. 2).10 13 14 15 16 The 6 cohorts that stratified data by sex suggested an even stronger protective effect of a vegetarian diet for men.16 17

Figure 2. All-cause mortality for vegetarians compared to health conscious non-vegetarians in a meta-analysis using the fixed effects model. 

*Indicates that participants classified as semi-vegetarians were included in the vegetarian group

It should be stressed that the meat intake in the non-vegetarian group in most of these studies was significantly lower than that of the general population. As described in a different review, the selection criteria for the non-vegetarians in these studies generally included being affiliated with vegetarians in some way or another, likely explaining their relatively low meat intakes. Furthermore, evidence suggests that many of the self-proclaimed vegetarians in these studies actually consumed meat on a regular basis, resulting in minimal differences in meat intake between the groups.10

Considering the lack of difference in meat intake between the vegetarian and non-vegetarian groups, it would only be expected that these studies would not have the statistical power to demonstrate a significant benefit of a vegetarian diet. Two studies which may be considered as especially having limited statistical power due to such limitations were the Heidelberg Study and Health Food Shoppers Study. In the Heidelberg Study, the non-vegetarian group were predominantly semi-vegetarians, while in the Health Food Shoppers Study, a validity assessment of the survey used to classify the participants vegetarian status suggested that 34% of the participants classified as vegetarians actually consumed meat. In fact, a slightly greater percentage of participants classified as vegetarians in the Health Food Shoppers Study were found to consume meat 3 or more times a week than the non-vegetarians in the Heidelberg Study (7.6% and 6.9%, respectively).14 18 Taking this into consideration, I performed a sensitivity analysis excluding either the Heidelberg Study or Health Food Shoppers Study from the meta-analysis. Excluding the Health Food Shopper Study alone reduced heterogeneity and strengthened the association between vegetarian status and a reduced risk of mortality from all-causes (RR 0.91 [95% CI, 0.87-0.94], and 0.92 [95% CI, 0.86-0.98] using the fixed effects and random effects model, respectively).

Most of these studies did not provide separate data for length of adherence to a vegetarian diet. As already described earlier in this review, evidence from several of these studies suggest a stronger effect on mortality would have been observed if the analysis was restricted to long-term vegetarians.10 Another important limitation was that most studies also did not verify changes to vegetarian status of the participants throughout the follow-up, which may in part explain why the association between a vegetarian diet and a reduced risk of mortality weakened over time in several of the studies.9 10

Another important potential limitation of these studies described in detail in the next section, is that some of the participants likely adopted a vegetarian diet in order to improve poor health, such as symptoms of an undiagnosed or developing illness that would ultimately became fatal. Each of these limitations described are expected to have either biased these findings towards null, or even in favor of the non-vegetarian group, suggesting that the findings of this meta-analysis may have significantly underestimated the benefits of an appropriately planned vegetarian diet.

In order to determine which factors may have contributed to the observed reduced risk of death in vegetarians, I also performed separate meta-analyses for the major causes of death. This included mortality from coronary heart disease and cardiovascular disease, and the incidence of all cancers combined. For mortality from coronary heart disease, based on 7 cohorts, the vegetarian group had a statistically significant 25% and 24% reduced risk, using the fixed effects and random effects models, respectively (Fig. 3).13 14 15 16 17 For mortality from cardiovascular disease, based on 7 cohorts, the vegetarian group had a statistically significant 17% and 14% reduced risk, using the fixed effects and random effects models, respectively (Fig. 4).13 14 15 16 17 For the Adventist Mortality Study and Adventist Health Study, mortality from cardiovascular disease was derived from pooling the relative risk for mortality from coronary heart disease and stroke.17 Excluding the Adventist Mortality Study and the Adventist Health Study, the vegetarian group had a statistically significant 8% reduced risk of mortality from cardiovascular disease, using both the fixed effects or random effects model (0.92 [95% CI, 0.85-0.99]).

Figure 3. Coronary heart disease mortality for vegetarians compared to health conscious non-vegetarians in a meta-analysis using the fixed effects model. 

*Indicates that participants classified as semi-vegetarians were included in the vegetarian group

Figure 4. Cardiovascular disease mortality for vegetarians compared to health conscious non-vegetarians in a meta-analysis using the fixed effects model. 

*Indicates that participants classified as semi-vegetarians were included in the vegetation group

As described in a previous review, 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. There is a plethora of evidence, not only from epidemiological studies, but also clinical trials that plant-based diets and nutrients have favorable effects on total and LDL cholesterol, blood pressure, among several other factors which are established risk factors for cardiovascular and all-cause mortality.19 20 21 22

In the Oxford Vegetarian Study, high compared to low intake of saturated animal fat was associated with a nearly 3-fold increased risk of coronary heart disease mortality.23 Similarly, in a meta-analysis of 11 cohort studies, high compared to low intake of saturated fat was associated with a 32% increased risk of coronary heart disease mortality, despite the inclusion of over-adjustments for dietary and serum lipids.24 It was also found in the Oxford Vegetarian Study that high compared to low intake of total animal fat and dietary cholesterol was associated with a greater than 3-fold increased risk of coronary heart disease mortality.23 Furthermore, evidence from thousands of experiments carried out over the last century have shown that the feeding of dietary cholesterol and saturated fat has accelerated the development of atherosclerosis in virtually every animal species in which researchers were able to find a method to sufficiently elevate cholesterol concentrations. This includes herbivores, omnivores and carnivores from mammalian, avian and fish species, and over one dozen different species of nonhuman primates.

In both the Oxford Vegetarian Study and the Adventist Mortality Study, high compared to low intake of eggs was associated with an increased risk of coronary heart disease mortality.23 25 However, for ill-defined reasons, these studies were excluded from several recent meta-analyses. In the Adventist Mortality Study and Heidelberg Study, high compared to no intake of meat was associated with a 50% and almost 5-fold increased risk of coronary heart disease mortality, respectively.14 25 Similarly, in the Adventist Health Study, high compared to no intake of beef was associated with a greater than 2-fold increased risk for men.26 In addition, recent meta-analyses of prospective cohorts found that an increment of 1 mg/day of heme iron, found only in animal tissue, is associated with a 16% and 27% increased risk increased risk of type II diabetes and coronary heart disease, respectively.27 28 Therefore, the totality of evidence strongly suggests that the observed greater longevity of vegetarians can be explained, at least in part, by the reduced risk of cardiovascular disease as the result of the replacement of animal foods with minimally processed plant foods.

For incidence of all cancers combined, based on 5 cohorts, the vegetarian group had a statistically significant 8%, and borderline significant 6% reduced risk, using the fixed effects and random effects model, respectively (Fig. 5).13 14 29 30 Excluding the Health Food Shoppers Study removed evidence of heterogeneity and strengthened these findings (RR 0.90 [95% CI, 0.85-0.99] using both the fixed effects and random effects models).

Figure 5. Cancer incidence for vegetarians compared to health conscious non-vegetarians in a meta-analysis using the fixed effects model. 

*Indicates that participants classified as semi-vegetarians were included in the vegetation group

The finding of a decreased risk of cancer in vegetarians may also 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.31 32 33 Heme iron has also been associated with numerous other cancers. These lines of evidence also provide confidence in the validity of the findings of greater longevity in vegetarians.

Why Some People Choose to Become Health Conscious

Which came first,
vegetarianism or ill health?

In health research, the reasons why some people chose to become health conscious is critically important when interpreting data from observational studies. This is because it is possible that it may not have been the health conscious lifestyle that caused the examined outcome, but rather the outcome that caused the health conscious lifestyle, ie. reverse causality. As previously described in a different review, reverse causality occurs when the studied effect precedes the cause. An example in health research is the frequent paradoxical observation that former smokers have worse health outcomes than current smokers. These unfavorable outcomes are not explained as being caused by smoking cessation, but rather that those who quit smoking tend to have done so because they were showing symptoms of illness, illnesses that ultimately resulted in the observed unfavorable health outcomes. A similar phenomenon has been observed in nutritional research where sick people tend to adopt a more plant-based diet, suggesting that this would bias observational studies towards showing an unfavorable effect of plant-based diets and nutrients, and therefore a favorable effect of animal based diets and nutrients.34

Although Denise Minger suggests that the favorable health outcomes for vegetarians observed in many studies maybe unrelated to dietary factors, but rather explained by other healthy habits associated with vegetarianism, she fails to consider the possibility that these vegetarians may have become health conscious, including adopting a vegetarian diet, in order to improve poor health. 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.35 36

Although health researchers often attempt to partially control for reverse causality by excluding participants who were diagnosed with cancer, cardiovascular disease and other life threatening conditions prior to baseline of a study, it is nearly impossible to fully control for reverse causality, as participants may make dietary changes due to symptoms or unfavorable risk factors that act as markers of an undiagnosed or developing disease. For example, it is known that in studies carried out as far back as the late 1950s, participants with unfavorable blood cholesterol levels tended to reduce the intake of dietary cholesterol and saturated fat (ie. in part, by abstaining from animal foods). This resulted in paradoxical findings where participants who consumed more dietary cholesterol and saturated fat actually had lower serum cholesterol levels. These paradoxical findings were produced, in part, because those participants who continued to consume a diet rich in these lipids were choosing to do so because they were able to maintain lower cholesterol levels despite consuming such a diet (ie. due to favorable genetics).37 Many individuals who attempt to downplay the harmful effects of these lipids, such as Minger has done in her book have chosen to ignore this critical factor when reviewing research on the diet-heart hypothesis.24

In the chapter Herbivore’s Dilemma, Minger reviews several prospective cohort studies that compared the mortality rates of vegetarians to health conscious non-vegetarians, carefully selecting only those studies in which vegetarians were not found to live longer. These studies were the Oxford Vegetarian Study, the Health Food Shoppers Study and the Heidelberg Study, all included in my meta-analysis. Minger emphasizes the lack of reduced risk of mortality in the vegetarian groups, while neglecting to mention that the difference in intake of meat between the groups was relatively small. Minger especially focuses on the Heidelberg Study, happening to be the smallest study, which found a non-significant increased risk of all-cause mortality in the vegetarian group. Minger points out that in this particular cohort, the vegetarians had greater levels of physical activity, consumed less alcohol and smoked less. Minger then states “Mortality and disease rates, in this case, might be expected to turn up in favor of the vegetarian crowd, even though the opposite ended up happening”[p.270]. This statement is outright misleading as all these factors were controlled for in the analysis. One could also correctly point out that the vegetarians tended to be older than the meat eaters, but suggesting that this would be expected to turn up in favor of the meat eaters would also be misleading, as age was also controlled for. Either way, the differences in mortality between the groups did not even come close to being statistically significant.

There are several important findings that Minger neglected to mention that cast doubt on the suggestion that a vegetarian diet had a harmful effect on the participants in the Heidelberg Study. For example, it was shown in an earlier follow-up of this study that, similar to the Adventist studies, the participants who adhered to a very low meat diet for at least 20 years had a 29% lower risk of all-cause mortality compared to those who adhered to such a diet for fewer years.10 Excluding the first 5 years of follow-up significantly strengthened this association, resulting in a 45% lower risk of all-cause mortality, suggesting that reverse causation may have attenuated the results for the entire follow-up.38 This difference in mortality was considerably stronger than the mortality difference between vegetarians and non-vegetarians. Unfortunately the researchers appear to not have provided any data comparing mortality in long-term vegetarians and non-vegetarians. Indeed, in other cohorts it has been observed that compared to non-vegetarians, short-term vegetarians had a higher rate, whereas long-term vegetarians had a lower rate of all-cause mortality.34 These findings suggest that those who adopted a vegetarian diet more recently had done so due to deteriorating health, biasing the results in favor of the non-vegetarians. 

Also as already briefly mentioned, in this study it was also found that intake of meat 3 or more time per week was associated with an almost 5-fold increased risk of mortality from ischemic heart disease, whereas there was a greater than 2-fold risk increase for those who consumed fish more than once per month. One of the reasons that this did not translate into an increased risk of all-cause mortality in the non-vegetarian group may have been due to the very low intake of meat, with less than 7% of the non-vegetarians consuming meat 3 or more times a week.14 Another important finding was that there was a trend towards a favorable effect of a vegetarian diet on mortality in the early years of follow-up, which declined over time (Fig. 6).14 This suggests the likelihood of undocumented changes in vegetarian status in a portion of the participants throughout the follow-up, which would be expected to have attenuated a possible favorable effect of a vegetarian diet. A similar phenomenon was also observed in several other studies included in my meta-analysis.9

Figure 6. Standard Mortality Rates (SMRs) and their 95% CIs for all-cause mortality, dietary group, and 5-year follow-up period, separately for males and females. (○, female vegetarians; •, female nonvegetarians; □, male vegetarians; ▪, male nonvegetarians).

Minger also points out that there was a trend towards a higher rate of mortality among the vegans in the Heidelberg Study, which included only 60 vegans, while neglecting to mention that in the Adventist Health Study 2, which included almost 100 times many vegans found that compared to the non-vegetarians, the vegans, especially male vegans experienced a lower rate of mortality.16 In the Heidelberg Study, while current smokers had the same mortality rate from cancer as never smokers, past smokers suffered from a 70% greater risk of cancer mortality than current smokers. This strongly suggests that many of the past smokers in this cohort quit smoking because they had deteriorating health.14 This raises the question as to whether the participants in this cohort also had a tendency to adopt a vegetarian diet after becoming ill, possibly explaining the unfavorable mortality rates in the vegetarian group. Perhaps the reason why the vegetarians in this group tended to be older, was not because they were becoming more ethical with old age, but because they were becoming more health conscious, which included adopting a vegetarian diet due to deteriorating health with old age. As these lines of evidence described suggest, the trends towards elevated mortality in the vegetarian groups in the Heidelberg Study may be explained by reverse causation. It is more than possible that it was not the vegetarian diet that caused these unfavorable outcomes, but deteriorating health, which would ultimately resulted in these unfavorable outcomes that caused a portion of the participants to adopt a vegetarian diet.

Ignoring the difficult

There is strong evidence that plant-based diets, including those rich in whole-grains and legumes reduces the risk of premature death from chronic and degenerative diseases, which in-turn helps explain the longevity of a number of plant-based populations. Contrary to the findings for plant-based diets, there is a plethora of evidence demonstrating the harms of popular carbohydrate restricted diets. For example, meta-analyses of clinical trials have found that low-carbohydrate diets elevate LDL cholesterol and impair endothelial function.39 Furthermore, a recent meta-analysis of prospective cohort studies with more than 272,000 participants found that low-carbohydrate diets, particularly those rich in animal foods are associated with an increased risk of all-cause mortality.40 Evidence also shows that the hazardous effects of diets rich in animal foods are also applicable to that of organic, grass-fed animal foods. These, and other lines of evidence described throughout this review appear to have been almost entirely ignored by proponents of these diets, perhaps because they find it too difficult to explain. 

Although there is strong evidence that plant-based diets promote longevity, it is important to replace animal foods with minimally refined plant foods in order to achieve the maximum benefits. This may in part explain why the benefits of a vegetarian diet were more evident in the Adventists than other populations.9 In conclusion, the totality of evidence supports the hypothesis that appropriately planned whole foods, plant-based diets promote longevity. 


Please post any comments in the Discussion Thread.