- Comparing Apples to Apples: Addressing the “Healthy-User Bias”
- Dying from Cancer Isn’t Better Than Dying from Heart Disease: Why Total Mortality Is the Most Important Endpoint in Longevity Studies
- The Seventh Day Adventist Studies: The Healthy-User Bias Strikes Again!
- The “Apples to Apples” Studies: What Research Comparing Healthy Omnivores to Vegetarians and Vegans Tells Us
- Meta-Analyses: What Have Reviews of Individual Studies Added?
- The Mormon Studies: Another Group of “Healthy Omnivores” That Lives Longer Than the General Population
- There’s No Real Evidence That Vegetarians and Vegans Live Longer Than Meat Eaters
In this post, I’m going to do a deep dive into the research into whether vegetarians and vegans live longer than omnivores—people who consume a mixed diet including meat. It’s a long article, because I wanted to cover all of the available studies on the topic and be as thorough as possible.
For the time-challenged among you, I’ll spill the beans right up front: while early studies did suggest a survival advantage for vegetarians and vegans, more recent and much higher-quality evidence has found no difference in lifespan between omnivores and vegetarians and vegans.
That’s the TL;DR. Let’s take a closer look at how researchers arrived at this conclusion.
Comparing Apples to Apples: Addressing the “Healthy-User Bias”
One of the biggest issues with studies that compare the health and longevity of vegetarians with omnivores is that they’re not comparing “apples to apples.”
The average vegetarian tends to be more health-conscious than the average omnivore: the very fact that a person has chosen to be a vegetarian indicates that she’s thinking more about her health than a “SAD omnivore”—someone following a Standard American Diet (SAD) and lifestyle.
This explains why studies show that vegetarians typically engage in a healthier lifestyle overall than SAD omnivores: they smoke and drink less, are less likely to be overweight or have conditions like diabetes, have higher levels of physical activity, and eat more fruits and vegetables, among other factors. (1, 2)
If scientists do a study to compare lifespan between SAD omnivores and vegetarians, and they observe that vegetarians live longer, how do they know that it was abstaining from eating meat that led to that survival advantage? How do they know that it wasn’t exercising more, drinking and smoking less, and/or other healthy lifestyle choices that extended the vegetarians’ lives?
These lifestyle factors are highly significant—more on this below—and are almost certainly responsible for the results of the few studies that did find a survival advantage in vegetarians.
This issue is known as the healthy-user bias, and it’s a huge problem that plagues most nutrition studies. I discussed it in detail in a podcast called “Heart Attacks and Red Meat—Correlation or Causation?,” but here’s the short version: People who engage in a behavior perceived as healthy are more likely to engage in other behaviors that are also perceived as healthy, and vice versa.
No, vegetarians and vegans don’t live longer than omnivores. But how did that myth get started? It turns out that a handful of poorly designed but well-publicized studies are to blame. Check out what the data really shows about the impacts of a plant-based diet.
Since eating meat has (wrongly) been perceived as unhealthy for the past five decades, people who eat less meat are less likely to smoke, drink excessively, or be overweight, and are more likely to exercise, eat fruits and vegetables, and get enough sleep. Can you see how this might cause problems?
If we really want to know whether removing meat and animal products from the diet extends lifespan, we’d need to study two groups of people:
A nutrivore would be someone who lives a healthy lifestyle (doesn’t smoke, doesn’t drink excessively, gets plenty of exercise and sleep, etc.) and consumes nutrient-dense, anti-inflammatory, whole foods like:
- Organ meat
- Bone broth
- A wide variety of fresh fruits and vegetables (including sea vegetables)
Comparing vegetarians with nutrivores would be more like comparing apples to apples.
To date, no studies have done this in a rigorous way. However, there are four studies that at least attempted to select a population of more health-conscious omnivores to compare against vegetarians and vegans:
- The Health Food Shoppers Study
- The Oxford Vegetarians Study
- The EPIC-Oxford Cohort
- The Heidelberg Study
We also have the 45 and Up Study from Australia. While this study didn’t select a healthier omnivore population, as the four studies I mentioned above did, the researchers did a much better job of controlling for confounding factors—like obesity, diabetes, smoking, drinking, and socioeconomic status—that would be likely to influence lifespan.
I’m going to review these studies in more detail below. But before I do that, let’s first discuss why total mortality is always the most important endpoint to consider in this kind of research.
Dying from Cancer Isn’t Better Than Dying from Heart Disease: Why Total Mortality Is the Most Important Endpoint in Longevity Studies
Let’s say a study finds vegetarians have a 30 percent lower risk of death from heart disease than omnivores. Sounds impressive, right? This difference will be what is reported in splashy media headlines, and people scanning through their newsfeeds will come away with the idea that vegetarians do live longer than meat eaters.
But then we take a closer look at the data, and we find that there was no difference in the number of deaths overall between the two groups. In medical research, the term “total mortality” refers to deaths from all causes—and it’s the most important endpoint to consider in studies comparing lifespan between groups.
Why? Because a reduction in deaths from one condition doesn’t mean much if it comes with an equal or greater increase in deaths from other conditions. The overall risk of death would be the same. Researchers call this “disease substitution”—because one cause of death is simply substituted for another.
You’ll see why this is such an important concept to understand as we review the studies below.
The Seventh Day Adventist Studies: The Healthy-User Bias Strikes Again!
Seventh Day Adventists (SDA) is a Protestant Christian denomination that grew out of the Millerite movement in the United States. Health has been a focus of the SDA teachings since the inception of the church in the 1860s. According to Wikipedia:
Adventists are known for presenting a “health message” that advocates vegetarianism and expects adherence to the kosher laws, particularly the kosher foods described in Leviticus 11, meaning abstinence from pork, shellfish, and other animals proscribed as “unclean.” The church discourages its members from consuming alcoholic beverages, tobacco or illegal drugs. … In addition, some Adventists avoid coffee, tea, cola, and other beverages containing caffeine.
Three studies have been done comparing the lifespan of SDAs to the general population:
- Adventist Netherlands (1968–1977)
- The U.S. Adventist Health Study (1977–1982)
- The U.S. Adventist Health Study 2 (2002–2009)
I’ve summarized the relevant findings of each of these studies below.
This study focused on 522 Dutch SDAs. Compared to omnivores, vegetarians had: (3)
- 59 percent fewer deaths from cardiovascular disease
- 57 percent fewer deaths from ischemic heart disease
- 46 percent fewer deaths from cerebrovascular disease
- 55 percent fewer deaths from all causes
The Adventist Health Study
This study examined 34,198 SDAs in the United States. Compared to omnivores, vegetarians had: (4)
- 38 percent fewer deaths from ischemic heart disease
- 7 percent fewer deaths from cerebrovascular disease
- 20 percent fewer deaths from all causes
The Adventist Health Study 2
In this study of 73,308 SDAs in the United States, compared to omnivores, vegetarians had: (5)
- 19 percent fewer deaths from ischemic heart disease
- 13 percent fewer deaths from cardiovascular disease
- 12 percent fewer deaths from all causes
These results certainly do seem impressive on the surface!
But, before you trade in your steak and salmon for tofu and lentils, let’s consider the serious shortcomings of these studies.
The Adventist Netherlands study compared SDA with the general population—not nutrivores! That introduces serious potential for healthy-user bias, because the members of the SDA church engage in lifestyle behaviors—like not smoking or drinking alcohol, eating more fresh fruits and vegetables, and getting more exercise—that have been shown to reduce the risk of death from cardiovascular disease and all causes. So, we can’t possibly know whether the reduction in deaths observed in this study was related to the vegetarian diet or these other causes, and thus the findings are not generalizable to the wider population.
Both Adventist Health Studies in the United States had better study designs. They recruited both the vegetarians and non-vegetarians from the SDA population. This reduced the healthy-user bias at least somewhat, since SDA omnivores would be expected to be healthier overall than SAD omnivores.
However, the first U.S. SDA study from the late 1970s and early 1980s did not control for smoking, body mass index, and other confounding factors. So, did the vegetarians in that study enjoy a longer lifespan because they didn’t eat meat, or because they smoked and weighed less than their omnivorous counterparts? We simply don’t know, which means the reduced risk of mortality among vegetarians in this study is also not generalizable to the wider population.
The second U.S. SDA study did a better job of controlling for confounding factors such as smoking and body mass index. But that doesn’t remove the potential for confounding altogether. On the contrary, as the authors of the study observed: (6)
Observed mortality benefits may be affected by factors related to the conscious lifestyle choice of a vegetarian diet other than dietary components. Potential for uncontrolled confounding remains.
Caution must be used in generalizing results to other populations in which attitudes, motivations, and applications of vegetarian dietary patterns may differ. [emphasis added]
The “Apples to Apples” Studies: What Research Comparing Healthy Omnivores to Vegetarians and Vegans Tells Us
Let’s turn our attention now to the studies that compare a healthier omnivore population with vegetarians and vegans.
It’s important to point out that we’re still not talking about nutrivores here; as you’ll see below, all we know about these omnivores is that they shop in health food stores or read health magazines. That said, these studies are at least a step in the right direction toward comparing “apples to apples,” because they at least attempt to reduce the healthy-user bias.
There are four studies that were designed specifically to select a healthier omnivore population and one study that did not do this but was performed using a large sample of the general population (rather than a religious group) and did a much better job controlling for confounding factors than most other studies of this type do.
The Health Food Shoppers Study
The Health Food Shoppers Study was a prospective cohort study of 10,736 subjects in the United Kingdom recruited between 1973 and 1979. Here’s a description of how participants were recruited, directly from the study: (7)
Subjects were recruited by distributing a short questionnaire to customers of health food shops and clinics, subscribers to health food magazines and a Seventh Day Adventist publication, and members of vegetarian and health food societies.
The idea was that omnivores who shop at health food stores, belong to health food societies, or read health food and/or SDA magazines and publications would be more health-conscious than average omnivores. This is certainly a step in the right direction in terms of the likelihood that the study results would be thrown off by the “healthy-user bias.”
What did the researchers find? In a 1999 analysis of the data, both vegetarians and omnivores in the health food store group lived longer than people in the general population—not surprising given their higher level of health consciousness—but there was no survival difference between vegetarians or omnivores. (8)
The Oxford Vegetarians Study
The Oxford Vegetarians Study was a prospective cohort study of 11,045 subjects from the United Kingdom recruited between 1980 and 1984. (9) The researchers recruited vegetarians first through the Vegetarian Society of the United Kingdom. They then asked the vegetarians to invite friends and relatives who consumed animal products, with the assumption that the omnivorous friends of vegetarians are likely to be healthier than the general population of omnivores.
As in the Health Food Shoppers Study, both vegetarians and health-conscious omnivores had lower risk of early death than the general population, but there was no difference in lifespan between the two groups.
The EPIC-Oxford Cohort is a component of the European Prospective Investigation into Cancer and Nutrition and included 44,561 participants from the United Kingdom recruited between 2002 and 2007, with a follow-up in 2009. (10) They used the same method as recruiting participants as the Oxford Vegetarians Study mentioned above.
In this study, researchers found that the risk of death for both vegetarians/vegans and omnivores was 52 percent lower than in the general population—similar to findings from the two studies above. However, there was no difference in mortality between vegetarians and omnivores.
The Heidelberg Study was a prospective cohort study of 1,904 subjects from Germany recruited between 1976 and 1999. (11) They used a similar method for recruiting to that of the Oxford Vegetarian Study and the EPIC-Oxford Cohort: participants were recruited from readers of vegetarian magazines in Germany using a short questionnaire, and the vegetarians that agreed to participate were asked to invite family members that ate meat.
This study found that vegetarians had slightly higher (10 percent) total mortality than healthy omnivores. What’s more, the data suggested that non-dietary factors played a much greater role in predicting lifespan than diet:
In summary, we conclude that the recommended healthy lifestyle factors, particularly abstinence from smoking, a moderate or high level of physical activity, moderate alcohol intake, and absence of overweight, are important determinants of reduced mortality in both vegetarians and nonvegetarians who already follow a healthy lifestyle.
We’ll discuss the contribution of non-dietary factors to lifespan in more detail below.
The 45 and Up Australian Study
The 45 and Up Study is a longitudinal study that followed 243,096 participants in New South Wales, Australia, for six years. At baseline, the vegetarians in the study were:
… younger, less likely to be overweight or obese, more likely to be female, and less likely to have cardiovascular and metabolic diseases (including Type 2 diabetes, heart disease or stroke) and hypertension at the time of recruitment. They were also more likely to have healthy lifestyle behaviours such as a lower prevalence of smoking and risky alcohol intake. (12)
As you now understand, these significant lifestyle differences between the vegetarians and the omnivores in the study introduce huge potential for the healthy-user bias. However, the study authors were aware of this and mentioned the shortcomings of previous research (like the SDA studies) in this regard:
First, most of these studies were designed to recruit vegetarians or occurred in populations with higher proportions of vegetarians (such as the Adventists, who may have other lifestyle factors and health-enhancing behaviours responsible for the observed protective effects), therefore previous ﬁndings may have limited generalizability. Second, vegetarians usually engage in an overall healthier lifestyle compared with their non-vegetarian counterparts, such as a lower prevalence of smoking and excessive alcohol consumption (Key et al., 2009) and have higher levels of physical activity (Bedford and Barr, 2005). Therefore the protective effects observed could be due to the other concurrent behaviours.
This awareness of the high risk of bias in observational nutrition research is likely what led them to pay much more attention to confounding factors than other, similar studies. They controlled for age, gender, education, marital status, geographic remoteness, SEIFA (Socio-Economic Index for Area, a census-based ecologic measure of socio-economic disadvantage), smoking status, physical activity, alcohol intake, and comorbidities including cancer, hypertension, and cardiometabolic disease (CMD), which includes type 2 diabetes, stroke, and heart disease.
While this doesn’t completely eliminate the healthy-user bias, it does at least reduce it. What were the results? They found no significant difference in total mortality between vegetarians and omnivores. There was also no difference in mortality between vegetarians, pesco-vegetarians, and semi-vegetarians.
This finding is especially notable because the study had such a large sample size (around 250,000 participants). Because vegetarianism is relatively rare, a large sample size with sufficient numbers of vegetarians and different variations of vegetarian diets makes the results more likely to be accurate.
Meta-Analyses: What Have Reviews of Individual Studies Added?
There have been two meta-analyses (comprehensive reviews) of individual studies comparing mortality in vegetarians and omnivores, including all of the studies discussed above. Both found no difference in total mortality between vegetarians and omnivores.
The first was done in 2014. (13) It found no difference in total mortality between vegetarians/vegans and omnivores.
What’s more, it concluded that any benefit observed in previous analyses for vegetarian diets was driven by the SDA studies, which, as we’ve seen, suffer from multiple confounding factors that weren’t adequately controlled for:
… while all of the SDA studies demonstrate signiﬁcant reduction in all-cause mortality with vegetarian diet, this ﬁnding was not replicated in four of the non-Adventist studies.
In conclusion, the reduction in IHD [ischemic heart disease] and all-cause mortality with vegetarian diet stems mainly from the Adventist studies, and there is much less convincing evidence from studies conducted in other populations.
In view of these inconsistent ﬁndings, we conclude that the beneﬁts of vegetarian diet for reducing death and vascular events remain unproven. [emphasis added]
The study authors suggested several possibilities for why the SDA studies showed a difference in mortality whereas the other studies did not. The SDA diet is characterized by higher intake of fruits and vegetables compared with typical omnivores, and regular SDA churchgoers are more likely to abstain from smoking, have good health practices, and stay married (all of which contribute to longevity in other studies). In addition, they are advised to get sufficient rest, exercise regularly, and maintain healthy relationships. All of these differences explain why the findings from the SDA studies are not generalizable to the wider population.
The second meta-analysis was published in 2017 and reviewed 96 studies. (14) Although they found slight relative reductions in death from heart disease and cancer in vegetarians and vegans compared with omnivores, they found no difference in total mortality.
In the discussion section, the authors remark that the modest reductions in risk factors like BMI, glucose, and cholesterol observed among vegetarians and vegans need to be taken with a grain of salt:
The results of the present meta-analysis report that vegetarians and vegans show signiﬁcantly lower levels of the most relevant risk factor for chronic disease such as BMI, lipid variables and fasting glucose, when compared to nonvegetarians and nonvegans. These ﬁndings, however, are signiﬁcantly affected by the nature of the cross-sectional studies, which are highly susceptible to biases, as otherwise observed by the moderate-to-high risk of bias assessment in each included study. [emphasis added]
They also suggested that virtually all studies that claim to find health benefits from a vegetarian diet are highly subject to the healthy-user bias:
Indeed, generally speaking, vegetarians tend to be more conscious for the health aspects, slimmer, and in better health when compared with omnivores, and speciﬁc cohorts have been demonstrated to be not generalizable to the general population for the low prevalence of risk factors (Kwok et al., 2014). These ﬁndings might indicate the presence of ﬂaws in the analysis of possible health beneﬁts of vegetarian diet. [emphasis added]
The Mormon Studies: Another Group of “Healthy Omnivores” That Lives Longer Than the General Population
As we’ve learned, the findings from the SDA studies aren’t generalizable to the wider population because Adventists engage in many behaviors independent of their diet that could contribute to a longer lifespan, such as:
- Not smoking or drinking
- Exercising more
- Eating more fruits and vegetables
What if there were another population we could study that also engaged in these healthy lifestyle practices but followed an omnivorous diet? If such a population also enjoyed better health and a longer lifespan than the general population, that would be another strong piece of evidence that the benefits seen in the SDA are due to their lifestyle, not because they don’t eat meat.
Well, turns out there is a population like this: Mormons. Religiously active Mormons practice a healthy lifestyle, which includes:
- Not smoking or drinking
- Getting adequate sleep
- Having a strong family life
- Obtaining high levels of education (15)
All of these factors have been associated with reduced risk of disease and extended lifespan in other studies.
There have been three studies that examined the standardized mortality ratio (SMR) of Mormons to assess lifespan. The SMR is the ratio of deaths in a particular group compared to the general population. It can be expressed as a ratio (e.g., 0.75) or a percentage (e.g., 75 percent). If the SMR of a group studied is 75 percent, that means their death rate is 25 percent lower than the death rate in the general population.
Here are the results of the three studies:
- “Health Practices and Cancer Mortality among Active California Mormons” found that Mormons had an SMR of 66 percent for deaths from all causes.
- “Cancer and Total Mortality among Active Mormons” found that Mormons over 35 years of age had an SMR of 50 percent for deaths from all causes.
- “Lifestyle and Reduced Mortality among Active California Mormons 1980–2004” found that Mormons aged 25–99 had an SMR of 45 percent for deaths from all causes.
This means that, across these three studies, Mormons had a significantly longer lifespan than their counterparts in the general population. This provides even more evidence that a healthier lifestyle, rather than abstaining from meat, was responsible for the survival advantage of vegetarians observed in the SDA studies.
There’s No Real Evidence That Vegetarians and Vegans Live Longer Than Meat Eaters
Here’s what we can conclude based on the existing research:
- While the average vegetarian may live longer than a “SAD omnivore,” there is no evidence that they live longer than more health-conscious omnivores.
- Studies showing health benefits of vegetarian diets are highly susceptible to the healthy-user bias, and their findings are not generalizable to the wider population.
- Diet and lifestyle factors such as exercise, alcohol intake, smoking, BMI, sleep, and fruit and vegetable consumption play a strong role in predicting lifespan independently of whether meat or animal products are consumed.
It’s also worth noting that we still don’t have a study that compares “nutrivores” with vegetarians and vegans. Although the methods used in the “apples to apples” studies mentioned above are better than the SDA and most observational studies, they’re far from perfect.
These studies recruited subjects who either shop at health food stores or read certain magazines or are friends or family members of vegetarians and vegans. But is that a really a reliable method for selecting a healthy omnivorous population?
Why It’s Hard to Create an Accurate Study
A truly accurate comparison of health-conscious omnivores and vegetarians would require a randomized, controlled trial (RCT) that lasts for at least 25–30 years, locks people up in a metabolic ward, and tightly controls their diet, exercise, sleep, alcohol intake, smoking, and other relevant lifestyle factors so that the only variable that differs between the two groups is whether they consume meat. This will never happen as it would cost hundreds of millions of dollars (who’s going to pay for that?) and they’d never find people willing to do it.
In the meantime, what’s a concerned person to do? Since modern nutrition research is not sufficient to answer this question, we need to rely on other lines of evidence. For example, what is the natural, human diet? What do studies say about the nutrient density of meat and other animal products? What do RCTs tell us about a Paleo diet—a whole-foods, nutrient-dense diet that emphasizes fresh, whole vegetables and fruits, nuts and seeds, and starchy plants, but also includes animal products?
I’ve addressed many of these topics elsewhere on my blog, podcast, and in my first book, The Paleo Cure, and I’ll be revisiting them soon. For now, you can start with the links in the paragraph above and simply focus on all of the things we know are important from an ancestral health perspective:
- Eating fresh, whole foods and avoiding processed and refined food
- Getting enough exercise and sleep
- Managing your stress
- Having fun and making time for play
- Maintaining healthy relationships
- Spending time outdoors and making contact with nature
- Minimizing your exposure to environmental toxins
I hope you’ve enjoyed this article and that it has helped to clear up some of the confusion on this topic.
The term “nutrivore” comes from Dr. Sarah Ballantyne of The Paleo Mom.