A shorter, summarized version of this article can be found here.
It has been several years since we last reported an update on results from AHS-2. We are forever grateful to the nearly 100,000 church members who completed that long questionnaire. Although the study has now resulted in more than 200 reports in the medical literature from AHS-2, you may reasonably say, “What does that mean for me?” Here we do our best to fill that gap. We will not detail well-known previous information showing that Adventists are particularly long-lived on average—part of a so-called “Blue Zone,” or that red meat consumption increases risk of several disorders, or that nut consumption protects against heart disease and possibly some other problems. More recently we found that meat protein is associated with higher risk and nut protein with lower risk of heart disease. Thus, it may not only be fats that are relevant to disease risk. We know that Adventists die of mainly the same disorders as other Americans, but whatever the cause, it is delayed for several years. Further, there is relatively good quality of life in those extra years. The nature of this research is that it does not allow predictions about the health of individual church members, but relates to “typical” Adventists who eat and live their lives in the specified ways. Thus, it does apply to most of us.
Ninety-six thousand Seventh-day Adventists, ages 30 to 112, from all 50 U.S. states and Canada were enrolled in the second Adventist Health Study (AHS-2) from 2002 to 2007, including 25 percent African American participants and smaller proportions of other racial and ethnic minorities. African Americans are more likely to have high blood pressure (hypertension) and prostate cancer. They are also at higher risk for diabetes and kidney disease, but have better bone health and much less skin cancer. We find all those trends in AHS-2. But we also find that African American study participants have approximately all of same trends with diet and health that we report below. Therefore, we have evidence that healthy dietary patterns are very important in reducing/addressing health disparities among African Americans also. We plan further work on this in the near future, under the direction of Dr. Fayth Miles-Butler.
Vegetarianism
In AHS-2 we have mainly taken the opportunity to focus on Adventists who eat differently from each other. These include: nonvegetarians (about 50 percent of our study Adventists) who nevertheless are generally more health-conscious than the average American. They eat meats only about three times per week, at least half of these being chicken or fish; then there are pesco-vegetarians (about 8 percent of AHS-2) who eat only fish as a flesh food; and lacto-ovo-vegetarians (about 32 percent of AHS-2) who consume no flesh foods, but as with the pesco-vegetarians may eat dairy foods and eggs; and finally total vegetarian Adventists (about 8-9 percent of AHS-2) who eat no animal products at all. We find that as Adventists get older most stay with their long-held dietary preferences, but to the extent that there is change, they tend to migrate to the stricter vegetarian categories.
We have often compared the health experience of the three types of vegetarians, with that of the nonvegetarian Adventists as a reference—a tough comparison as even the nonvegetarian Adventists are relatively healthy in their choices. This was evident when we compared the mortality experience of all Adventists to that of a U.S. Census population, which also took account of differences in past smoking history, and education. We found that even our Adventist nonvegetarians live longer than non-smoking non-Adventist Americans. This result was also true for Black Adventists as compared to U.S. Census Black Americans (we had insufficient numbers of other racial groups for useful comparisons).
Quite naturally, some may wonder how well even a long questionnaire about diet really measures what we eat. There are so many foods and so many different ways to prepare them. We looked at this very closely, and compared estimates from our questionnaire to those coming from a more accurate but difficult method of measuring diet. An amazing 1,100 study participants (who were randomly selected) talked 6 times on the phone to our nutritionists and reported everything that was eaten in the last 24 hours. A combination of these 6 (more accurate) recalls compared remarkably well with the questionnaire estimates. We also compared intakes of nutrients such as fatty acids, carotenoids, and many others, estimated from the questionnaire, to the same chemicals within the body. These chemicals were measured in blood, and from little nubs of fatty tissue (also provided by the amazing 1,100!). Again, our dietary estimates usually predicted well the relative levels of the same chemicals in body tissues. So, AHS-2 did much to investigate and prove the sufficient accuracy of our questionnaire.
Many common afflictions of middle and later life are caused by inflammation at the cellular level. These include heart disease, diabetes, many cancers, autoimmune disorders (such as rheumatoid arthritis and systemic lupus erythematosus [SLE]), and so, much of overall mortality. All categories of vegetarians (their diets determined from our questionnaire data) have lower levels of C-reactive protein (CRP) a marker of inflammation, especially the total vegetarians, this as compared to nonvegetarian Adventists. Further, all vegetarians have markedly lower body weights (adjusting for height) as compared to nonvegetarians. This is important as excess fatty tissue has often been called the “organ of inflammation,” in that it produces chemicals that enhance inflammation. Vegetarians have markedly lower rates of diabetes, an inflammation-mediated disorder where, in adult-onset diabetes, insulin does not work as well as it should. Our data shows that eating mainly plant foods is associated with better sensitivity to insulin, and that an omega-3 fatty acid from plant foods, alpha-linolenic acid (ALA, found especially in soy, flaxseed, chia, walnuts, and elsewhere), appears to be one probable cause of this.
Some of recent work compared more than 900 blood chemicals in total vegetarian to nonvegetarian Adventists. Amazingly, using appropriate statistical methods, we were able to show that nearly 600 of these were significantly different (not just a chance difference) between the two dietary groups. It appears that what we eat really does change what ends up in the blood and hence the fluids that bathe our cells. Thus, it is hardly surprising that diet has an effect on health!
We also focused on a small number of foods that account for some of the main distinctions between the different vegetarian dietary patterns. In particular, we have so far analyzed the apparent effects of meats and dairy. We have not documented any adverse effects of fish or poultry, except to say, for poultry at least, that it can matter greatly what one may choose to replace it after it is eliminated. Eating is not a habit (like say smoking) that we can stop and start. So, stopping one food must nearly always be balanced by replacing it with something else. We have evidence that vegetable sources of protein that may replace poultry, such as whole grains, legumes, nuts, will often have protective properties not possessed by the poultry. Fish may have some protective properties, and we are still working on this. Red meats, their dense calories, their lack of dietary fiber, their saturated fats, and probably some of their proteins, increase the risk of overweight, diabetes mellitus, heart disease, colorectal (and perhaps some other) cancers. They have little to recommend them.
We live longer on average, and with better quality of life.
Cancers
Three cancers are particularly common in Western societies. These are colorectal cancer, breast cancer in women and prostate cancers in men. Cancers in different organs all have some general similarities, but in many ways they are different diseases, with different risk factors. It would be unreasonable to expect that a lifestyle, such as that recommended by Adventists, would protect from all chronic diseases, and all of the different cancers, and that is what we find. However, it protects from many of these, sufficient to make it clearly beneficial.
There is good evidence from many sources that red meat, particularly processed red meat, increases risk of colorectal cancer. Results from AHS-2 are consistent with this. It is also worth noting that the pesco-vegetarians had a particularly lower risk of colorectal cancer, and that was unlikely to be a result of chance. So, that places a new focus on fish consumption that needs further confirmation. Was it the fish or some other characteristic of the pesco-vegetarians?
Dairy is a complex food, particularly in its association with different cancers. Many studies, including AHS-2, find that dairy consumers have lower risk of colorectal cancers, so total vegetarians who consume no dairy, get a little less protection here than other vegetarians. But our data, drilling a little deeper, suggest that this dairy effect on colorectal cancer is likely due to dairy’s calcium content, and calcium is also easily found in a wide variety of vegetable products.
In AHS-2, we noted that two hormone-responsive cancers, breast cancer in women and prostate cancer in men, were 25-35 percent less common in total vegetarians, but not at all less frequent in lacto-ovo- or pesco-vegetarians, when compared to nonvegetarians. This again threw a spotlight on dairy (as total vegetarians consume no dairy), and indeed we have now analyzed and reported apparently clear trends of increased risk of these two cancers with some dairy products. This particularly seems to involve dairy milk. We found little or no discernible effect for cheese and yogurt. The dairy milk association seemed almost pharmacologic, starting to accumulate at quite small amounts of regular, long-term consumption. The maximum effects (at least a 50 percent increase in risk) were achieved at only ¾ cup per day. At higher doses the apparent adverse effect remains but gets no greater. We speculate that this may be related either to sex hormones from the cows, which are found in small quantities in milk particularly, or to effects of milk protein on a hormone called IGF-1, which is thought to increase risk of both of these cancers. Low-fat and regular fat milks had almost identical effects, suggesting that it is not the milk fat that may be causing the problem. A word of caution, a number of studies find some evidence of a milk effect on risk of prostate cancer, but few studies have found the effect that we did for breast cancer. Yet in AHS-2 the association appeared strong, and was highly unlikely a result of chance. It is also true that few other studies have been able to examine milk effects at low doses, as we could, and that was where the greatest change in risk seemed to occur.
Overall vegetarian Adventists have about 10 percent lower risk of all cancers combined, when compared to nonvegetarian Adventists. Turning to the specific types of vegetarians, total vegetarians have about a 20 percent lower risk of all cancers combined, lacto-ovo-vegetarians about 10 percent lower risk, pesco-vegetarians about 15 percent lower risk, each when compared to nonvegetarian Adventists. But when we compare all Adventists to non-Adventists, there is at least a 30 percent lower risk. This tells us that even the relatively health-conscious nonvegetarian Adventists are gaining some benefit and have less cancer risk than non-Adventist Americans.
We have carefully studied the risk of more than 20 different forms of cancer, and whether dietary connections can be found. Above, we have already mentioned the three most common cancers, prostate, breast, and colorectal in relation to dietary patterns, red meat, fish, and dairy milk. Cancers in other bodily organs are less frequent which makes them more difficult to study, but there are some new (as yet unpublished) results from our ongoing work. As compared to nonvegetarian Adventists, vegetarians (as a group) have much lower risk of stomach cancer and moderately lower risk of lymphoma and lymphoid leukemia than nonvegetarian Adventists. In fact, vegetarians have also lower risk of some other less common cancers, even though our data does not allow us to definitely identify which they are. Some possibilities here are lung and pancreatic cancers, but more evidence is required.
Other Disorders
Although our focus, due to the National Institutes of Health funding agency, was mostly on cancer, we have taken the opportunity to explore possible effects of lifestyle on a small number of other disorders. There is little doubt that Adventist vegetarians do much better than Adventist nonvegetarians with respect to risk of diabetes, high blood pressures, and blood cholesterol levels. In all of these, the total vegetarians do best of all, followed by lacto-ovo-vegetarians and pesco-vegetarians. We have found however, that total vegetarians tend to have a higher risk of bone fractures (and thus presumably osteoporosis), but for them, this can be largely mitigated by adequate calcium and vitamin D supplements. Other data indicates, however, that for most people such supplements are of little value. Nonetheless, AHS-2, and other data, show that regular vigorous physical activity and adequate (but not excessive) animal or vegetable dietary protein are excellent supports for good bone health. Recent AHS-2 publications also indicate that the autoimmune disorder SLE is less prevalent in vegetarians, perhaps partially related to higher intake of the ALA omega-3 fatty acid. This is preliminary work that needs more study to further support causal connections.
| Table. Risk of dying from different contributing causes, according to diet: Adventists in AHS-2* (Red indicates more, and black less, certainty in the findings) | ||
| Cause of Dying† | Younger deaths | Older deaths |
| A. All vegetarians compared to nonvegetarian Adventists | ||
| Kidney Failure | 50% less | 33% less |
| Infectious diseases | 42% less | 12% less |
| Diabetes | 47% less | 31% less |
| Heart-related | 25% less | 12% less |
| Stroke | 2% more | 18% more |
| Dementia | 24% more | 15% more |
| Parkinson’s disease | 29% less | 43% more |
| B. Different types of vegetarians compared to nonvegetarian Adventists | ||
| Stroke in i) total vegetarians | 5% more | 30% more |
| ii) lacto-ovo-vegetarians | 4% less | 19% more |
| iii) pesco-vegetarians | 11% more | 5% more |
| Dementia in i) total vegetarians | 39% more | 19% more |
| ii) lacto-ovo-vegetarians | 17% more | 12% more |
| iii) pesco-vegetarians | 21% more | 19% more |
| Parkinson’s disease in i) total vegetarians | 32% less | 44% more |
| ii) lacto-ovo-vegetarians | 24% less | 51% more |
| iii) pesco-vegetarians | 77% more | 29% more |
| *The named causes of death come from any mention of this cause on the death certificate. Gray shading indicates deaths associated with neurological causes. †For deaths associated with other causes, our data (limited by low numbers in some cases) did not demonstrate clear differences in risk between vegetarians and nonvegetarians. | ||
One of our reports that got much attention in the press was that we found Adventists who spaced their meals such that there was a long (perhaps 16-hour) “fast” overnight had less weight gain over several decades of adult life. We also found that those preferring to make breakfast their main meal did well in this regard. It is noteworthy that the recently promoted “16/8” diet to lose weight has become popular and for many works well without great feelings of hunger. To have perhaps a little later breakfast and a little earlier supper about eight hours later is one option, although other plans with similar spacings are possible.
Total Mortality
This leads to the final focus of AHS-2 results, namely “how does diet relate to total mortality?” which is the risk of dying during a particular age. Often as doctors we encounter elderly total vegetarian, and other keenly health-conscious Adventists, who are discouraged, and sometimes angry, that they now suffer from a chronic disease. While we can and do sympathize with their distress, it is worth pointing out that protection afforded by a healthy lifestyle is that of a lower risk—not the absence of risk, and secondly that we do finally suffer from essentially the same disorders as others. The benefit is that they usually come at later ages. That is how we live longer on average, and with better quality of life.
Our very recent findings about deaths in relation to diet are both wonderfully supportive of the Adventist health message, but in one way perhaps a little concerning. At younger years (about 65 years of age), lower risk of dying from any cause is found in male vegetarians (about 16 percent lower, especially in total vegetarian males), and in women about 8 percent lower, this, as usual, in comparison to Adventist nonvegetarians. In later years, however, (say 75 years of age and above) we do not find clear evidence of a total mortality benefit for the vegetarians as a group. Comparing the various dietary patterns, pesco and lacto-ovo-vegetarians seem to do the best for overall total mortality. So it seems that vegetarian diets, as we currently practice them, are especially tending to prevent premature deaths at younger ages, allowing more of us to achieve older ages. In those who have survived that long, however, the advantage in total vegetarians and lacto-ovo-vegetarians for total mortality then becomes harder to detect. Some advantage seems to persist at older ages, however, for pesco-vegetarians.
A little more clarity comes if we look at the causes contributing to deaths in Adventists (please refer to the Table). You can see that a vegetarian lifestyle does offer substantial (30-50 percent) protection against deaths from cardiovascular disease, infectious diseases, diabetes, and kidney failure, particularly at younger ages, but to some extent, also in older age. Remember that this Table is comparing vegetarian Adventists to relatively health conscious, mainly low meat-consuming, nonvegetarian Adventists—a hard comparison. A comparison to nonvegetarian non-Adventists would undoubtedly be even more impressive.
One area of potential concern is a signal of increased deaths in vegetarians related to neurological diseases (such as stroke, dementia, and Parkinson’s disease),but only in old age—where the words “more than” outnumber “less than” in the table above. More research is under way to give more clarity, especially in regard to potential nutrients that could explain this difference and may be targets for supplementation if there is evidence of deficiency. Please note also that it is still quite possible that all Adventists, as a group, are doing relatively well for these neurological diseases, even in old age, as compared to the general American population; we do not at present have that evidence available. We do know that when compared to a U.S. Census population total deaths in vegetarian and nonvegetarian Adventists combined are estimated at 33 percent and 22 percent lower at younger and older ages, respectively.
The additional research to explain the neurological disease questions in the very elderly, will take time. What could be significant issues? A clear possibility is the lower intake of vitamin B12 in stricter vegetarians, both total vegetarians and many low-dairy lacto-ovo-vegetarians. We do have evidence, however, that with supplementation of B12, most Adventists are doing well in this regard. Then there is the fact that the brain, a fatty organ, has as its most common fatty acid constituent, a long chain omega-3 acid called docosahexaenoic acid (DHA for short). Yet, strict vegetarian diets provide none of this. DHA is found naturally in fatty fish (the fish obtain it from algae and other micro-organisms that they eat). Vegetarian diets do provide good amounts of a medium chain omega-3 fatty acid, alpha-linolenic acid (ALA for short) and it can be converted to DHA slowly, but possibly adequately. This conversion, though, may be partially prevented if there is too much omega-6 linoleic acid in the diet. Omega-6 linoleic acid is especially found in some vegetables, grains, walnuts, and soft plant-based margarines. It is very helpful in lowering blood cholesterol, and thus helping to prevent artery disease. So you see the balancing act that the best diet for elderly people may require! This presently is rather speculative, and there are also other ideas and possibly other factors involved.
It is of interest that these results “shine a spotlight” on elderly vegetarian Adventist (in their 70’s and older). Our bodies, their physiology and chemistry, are absolutely marvelously complex, but as we age, many bodily functions do tend to become less efficient. Vitamin B12 is absorbed less well by the stomach, the skin does less well when using sunlight to form vitamin D, and probably some of the chemical pathways mentioned above also become less effective. All total vegetarians, and many lacto-ovo-vegetarians, should definitely supplement with B12 as they grow older. Now, it seems possible that in later years (late 70’s and beyond), the usual recommendations just may benefit from further “fine-tuning.”
Summary
In summary, it is clearly true that Adventists as a group are doing better than others, and this includes the nonvegetarians. We are a fortunate people! The vegetarian Adventists are doing better yet. This is especially so for the pesco- and lacto-ovo-vegetarians, and male total vegetarians when looking at total mortality. More of us are living to older ages. In our latest, more comprehensive data, however, we do not find any clear overall mortality (longevity) advantage for total vegetarian women, though male total vegetarians are especially protected from deaths at earlier ages, as compared to nonvegetarians. The challenge is this: As compared to nonvegetarians, total vegetarians especially and also other vegetarians are doing particularly well as far as less heart disease, kidney disease, diabetes, and high blood pressure, thus “delivering” more of us to those older years. Despite this, it appears that for some in their later years the vegetarian advantage may then be offset by not doing quite so well from neurological diseases, when compared to low meat-and-fish-consuming elderly nonvegetarian Adventists. Perhaps this is natural for those vegetarians who would have otherwise died at younger ages, but surely it will be worth some effort to avoid this possible disadvantage in old age—remembering, however, that even the most adherent Adventists will eventually die!
What can we recommend as far as diet, based on all this evidence? Expressed “simply” our findings show that:
- Even though all Adventists as a group are doing well, a vegetarian way of eating is preferable for a long relatively healthy life. This is particularly helpful to prevent premature onset of many common diseases. Individual church members may well choose different ways and degrees of fulfilling this, according to their beliefs and life circumstances, and we do not presume to be critical of that. Here is how we see further details.
- It is best to avoid dairy milk, but then be sure to regularly consume calcium-containing vegetables (legumes, grains, green leafy vegetables). This will probably decrease risk of some common cancers. Moderate amounts of cheese and yogurt appear safe. There are many vegetable “milks” available. Possibly those with an adverse family history of breast or prostate cancer should particularly consider this.
- Regular nut consumption (in small amounts) is particularly healthy, helping to prevent cardiovascular disease.
- Avoiding large amounts of processed foods is a good choice to promote longevity, but this does not at all preclude the cooking of foods, or the preparation of interesting, tasty, combinations of whole plant-foods.
- Be sure to obtain a balance of protein (good plant proteins are quite adequate—nuts, whole grains, legumes), fats (mono-unsaturated oleic acid, in olive oil and many nuts are good choices), without too much carbohydrate (a little too easy to find in many plant foods). This will help with body weight, cardiovascular and bone health.
- Give consideration to a two-meal each day program, or at most a small-snack lunch, with a 14-16 hour food-free gap overnight between a mid to late afternoon meal and breakfast. This seems best for maintaining a healthy body weight, and is quite consistent with the recommendations of Ellen White.
Placed in conjunction with other work we also suggest that:
- For those who may consider eating fish, if obtained from clean waters it is probably beneficial (although large long-lived deep sea fish may accumulate mercury). Remember that Christ and his disciples ate fish. Ellen White did not clearly speak against fish, only expressing concern to avoid those sourced from polluted waters. Fish (especially salmon, herrings, and sardines) provide the best source of that long chain DHA fatty acid, and this just may be especially important in the elderly. Alternatively, you can supplement with fish oil capsules at perhaps 1,000–1,500 mg per day (too much may increase risk of a cardiac arrythmia called atrial fibrillation).
- It is probably true that good vegetable sources of medium chain omega-3 ALA acid can then allow satisfactory conversion to DHA in the body. If you prefer this stricter vegetarian route, it may be best to avoid large amounts of many commercial margarines that are very highly processed, though the evidence is not yet clear. These margarines are usually plant-based, often from corn, safflower, or sunflower oils. The omega-6 linoleic acid there is abnormally concentrated far above natural levels. Although more linoleic acid is good for blood cholesterol, large amounts are less clearly beneficial overall.
- Good vegetable sources of ALA include flaxseed (same as linseed) and chia (these two are best of all), but walnuts and soy are also good sources. An algae-based (hence vegetarian) source of DHA directly, is also available in capsule form.
- Total vegetarians and lacto-ovo-vegetarians (especially if you consume little dairy or eggs), please be sure to supplement with vitamin B12.
Clearly, diet is very complex. Medical and dietary research results in a constantly changing picture. What we will write in 10 years is likely to have some differences from what we write today, but hopefully these differences will be relatively minor, mainly filling gaps in our present understanding. What we can say with confidence is that the Adventist “health message” has resulted in many hundreds of thousands of extra good quality years of life among church members, this over five to six generations.
For those interested, a complete list of Adventist Health Study publications can be found at https://adventisthealthstudy.org/researchers.
The following information was included as a sidebar in the truncated version of this article printed in the June issue of Adventist Review and is here included for your reference.
If you are an Adventist Health Study-2 participant, we are most grateful for that.
Our study has so far focused on cancer, heart disease, risk of dying (longevity), but there are many other common medical conditions that may be influenced by diet. Some of these cannot be found by questionnaire with any accuracy (e.g., rheumatoid arthritis, congestive heart failure, dementia, kidney disease). In order to investigate them, it is essential that we identify study members who have developed these disorders. There is often an inexpensive way of doing this by matching study members to national databases of, say, kidney failure or heart bypass surgery. It is also extremely informative to match AHS-2 data with Medicare records (for those 44,000+ AHS-2 participants who are Medicare members).
So we are planning to do this, using your name, date of birth, and Social Security number (if you provided it) as personal identifiers. Please note that these databases always have very strict security protocols, and if you are not already on their rolls, your personal information will not be retained by them.
To decline the use of your data for these purposes, you can email us at [email protected] or send a note by mail to Adventist Health Study, Room 330, Nichol Hall, Loma Linda University, Loma Linda, CA 92350, if possible before July 31, 2024.