Last updated March 2018
For information on the types of studies and their pros and cons, see Basics of Nutrition Research.
- Disease Rate Statistics
- Vegetarian Cohorts
- Standardized Mortality Ratios
- Mortality Meta-Analysis (1999)
- Mortality Meta-Analysis (2012)
- EPIC-Oxford: Gallstones (2017)
- EPIC-Oxford: Mortality Rates (2015)
- EPIC-Oxford: Heart Disease Incidence (2013)
- EPIC-Oxford: Cataracts (2011)
- EPIC-Oxford: Diverticular Disease (2011)
- EPIC-Oxford: Heart Disease Mortality (2009)
- EPIC-Oxford: Bone Fractures (2007)
- EPIC-Oxford: Preliminary Mortality Rates (2003)
- British Mortality (2002)
- Adventist Health Study-2
- Adventist Health Study
- Heidelberg Study (German Vegetarians)
Disease Rate Statistics
When comparing incidence or mortality rates of two or more groups, one group is assigned the rate of 1.00 and the other groups are compared to it. These are called disease rate ratios or odds rate ratios.
Group A 1.00
Group B 1.35
Group B has a 35% higher rate of the disease than Group A.
Group A 1.00
Group B .85
Group B has a 15% lower rate of the disease than Group A.
In addition to the rates, there has to be a test to determine if the rates are different enough not to be due merely to random chance (also known as statistical significance). Statistical significance for a disease rate is usually expressed by way of a 95% confidence interval (CI). This is done by giving a lower and upper limit for the interval. If 1.00 does not fall between the two numbers (i.e., within the interval), then the finding is significant and not due to random chance.
Example 1: .85 (.75, .95)
The finding is statistically significant because 1.00 falls outside the 95% CI.
Example 2: .85 (.65, 1.05)
The finding is not statistically significant because 1.00 falls inside the 95% CI.
Sometimes, p-values are given rather than confidence intervals. In these cases, a p-value of less than .05 means the finding is statistically significant.
In the tables below, statistically significant findings are in red.
When disease rates are adjusted, it means they’re changed to account for variables that might affect them.
For example, a study finds that smoking is related to cancer and that drinking is also related to cancer. But many of the people in the study both smoke and drink, so you don’t know whether it was the smoking or the drinking (or both) that is actually related to the cancer. By adjusting, you can look at the different levels of drinking taking into account how much the subjects smoked, and get a number for drinking that isn’t influenced as much by smoking.
What often happens is that a variable loses its significance once the results are adjusted. For example, a study of people aged from 20 to 60 years old will likely correlate the likelihood of having a heart attack with having gray hair. But once you adjust for the age of the participants, the correlation with gray hair will fall away and we can then assume gray hair doesn’t cause heart attacks.
Most studies adjust for more than one variable at a time. They often adjust for all the variables that, in the non-adjusted analysis, had a significant relationship to the disease.
Well-designed studies allow researchers to consider adjusted results in their calculations. Frequently, these adjusted results are easier to draw conclusions from. The articles on VeganHealth.org use adjusted rates unless otherwise noted.
The following studies have compared disease rates between people with different diets but similar lifestyles:
|Population Studies with Large Numbers of Vegetarians|
|Study||Country||Years||Number of Vegans|
|Health Food Shoppers||UK||1976-88|
| Meta Analysis –
of the above studies
|Adventist Health Study-2||USA||2002 –||5,500|
Standardized Mortality Ratios
The following studies compared the death rates (deaths per 100,000 people per year, under age 90) of their entire study population to the greater population in that region, known as standardized mortality ratios (SMRs). The SMRs were:
- EPIC-Oxford – 52% (8)
- Adventist Mortality Study – 49% (3)
- Health Food Shoppers – 56% (3)
- Oxford Vegetarian Study – 46% (3)
- Heidelberg Study – 48% (3)
This means that people in these studies died at about half the rates of the overall populations in their region. The researchers considered most of this difference to be due to lower smoking rates in the study groups (3), and the healthy volunteer effect (3) (healthy people are more likely to volunteer for studies than unhealthy people) but some difference may have been due to a generally healthier diet overall than in the population at large (3, 8).
Mortality Meta-Analysis (1999)
The 1999 Meta-Analysis (3) compared vegetarian death rates to non-vegetarians (click here for cancer rates). Although there have been more recent studies on vegetarian mortality rates, the 1999 Meta-Analysis remains the one study that included vegetarians from both North America and Europe and still probably deserves the most weight of anything that has been released to date.
|Table 1. Vegetarian vs. Non-vegetarian Death Rates.|
|Disease||Rates and Confidence Interval|
|Ischemic heart disease||
Table 1 shows that vegetarians had a 24% lower risk of ischemic heart disease, but there was no difference for stroke or all causes.
There was also a sub-section that separated the vegetarian group into lacto-ovo vegetarians and vegans. The Health Food Shoppers Study was left out of this portion of the analysis because it did not distinguish between lacto-ovo vegetarians and vegans (6). Compared to 31,766 people who ate meat at least once per week:
- Occasional meat eaters (8,135 people who ate meat less than once per week) had a 20% reduced rate of dying of heart disease and a 10% reduced rate of overall mortality.
- Those who ate no meat other than fish (2,375 people) had a 34% reduced rate of dying from heart disease and an 18% reduced rate of overall mortality.
- Lacto-ovo vegetarians (23,265 people) had a 38% reduced rate of dying from lung cancer, a 34% reduced rate of dying from heart disease, and a 15% reduced rate of mortality.
- Vegans (753 people) had a heart disease rate of .74 (.46, 1.21) and a mortality rate of 1.00 (.70, 1.44). There were no statistically significant differences between the vegans and the regular meat-eaters for any causes of death.
|Table 2. Causes of 68 Vegan Deaths|
There were only 68 vegan deaths in this population over the course of the studies, as distinct from 3,017 for regular meat-eaters. The breakdown of vegan deaths is shown in Table 2.
In personal communication, an author of the study said the “Other causes” category of death did not appear to be diet-related and so was not subdivided into smaller categories (4).
Could Vegans Have Fared Better?
It should be noted that when these studies began, the full importance of vegans’ getting a reliable supply of vitamin B12 was not known.
Mortality Meta-Analysis (2012)
Although the 2012 meta-analysis by Huang et al. (11) is more recent, it may not be as reliable as the 1999 meta-analysis because it includes a 1984 study on Zen priests (12) who were mostly semi-vegetarian and which used a standardized mortality ratio (comparing all the Zen priests to the greater population rather than comparing the “vegetarians” to non-vegetarians within the same group). The Heidelberg Study results were also included and its control group was semi-vegetarians, which means there were semi-vegetarians in both the “vegetarian” and “non-vegetarian” group in the 2012 meta-analysis; while this is not ideal, it should have biased the results against finding a beneficial effect of a vegetarian diet. In its favor, the 2012 meta-analysis includes data from EPIC-Oxford that was not available for the 1999 meta-analysis. The table shows the results.
|Mortality Rate of Vegetarians|
|All-cause mortality||0.91 (0.66, 1.16)|
|Ischemic heart disease mortality||0.71 (0.56, 0.87)|
|Cerebrovascular disease mortality||0.88 (0.70, 1.06)|
|Combined ischemic heart and circulatory disease mortality||0.84 (0.54, 1.14)|
|Cancer incidence||0.82 (0.67, 0.97)|
EPIC-Oxford: Gallstones (2017)
After a mean follow-up time of 13.8 years and after adjusting for a number of variables including smoking and alcohol intake, EPIC-Oxford found no significant difference in the risk of symptomatic gallstone disease in vegetarians compared to meat eaters (20). When a further adjustment was done for body mass index (BMI), vegetarians had a higher risk (1.22, 1.06-1.41, p-0.006) for symptomatic gallstone disease compared to non-vegetarians.
Among both vegetarians and non-vegetarians, those with the highest BMI had the highest risk of developing the disease.
EPIC-Oxford: Mortality (2015)
In their paper from 2015 (19), the mortality rate before age 90 in EPIC-Oxford was no different between vegetarians (including vegans) and regular meat-eaters (1.02, 0.94-1.10). Vegetarians had lower rates of mortality from pancreatic cancer (0.48, 0.28-0.82) and lymphatic cancer (0.50, 0.32-0.79). Semi-vegetarians had lower rates of death from pancreatic cancer (0.55, 0.36-.86). Pesco-vegetarians had lower death rates from all cancers (0.82, 0.70-0.97) but higher rates of cardiovascular disease (1.22, 1.02-1.46).
In the main analyses in the above paragraph, some participants were recategorized based on a change in their diets. The researchers did a second analysis in which participants who changed their diets were removed and found an 8% reduced risk of early death in vegetarians that was just statistically significant (0.92, 0.84-0.99). Limiting the results further, to deaths before age 75, strengthened the finding (0.86, 0.77-0.97).
When vegans were separated from other vegetarians, there were no statistically significant differences in mortality rates for the six large categories of death. Eliminating participants who had changed diet categories didn’t significantly change the results. There were only 166 vegan deaths as distinct from 1,929 deaths in the entire cohort; meaning that reaching statistical significance was going to be unlikely.
EPIC-Oxford: Heart Disease Incidence (2013)
Heart disease rates of all vegetarians compared to all non-vegetarians were calculated for EPIC-Oxford from 1993 until 2009 (13). These participants were all thought to be free of heart disease at the beginning of the study. The results showed that vegetarians had a statistically significant, ~30% reduced risk of heart disease. This finding held even after adjusting for body mass index (BMI) and removing the first two years of follow-up.
The researchers believed the difference in heart disease rates to be due mainly to the lower non-HDL cholesterol levels and systolic blood pressure rates of the vegetarians.
|Risk of Heart Disease in Vegetarians Compared to Non-vegetarians in EPIC-Oxford|
|Model 1: age, smoking, alcohol, physical activity, education,
socioeconomic status, oral contraceptives, hormone replacement therapy
|0.68 (0.58, 0.81)|
|Model 1 plus Body Mass Index||0.72 (0.61, 0.85)|
|Model 1 excluding first 2 years of follow-up||0.69 (0.58, 0.82)|
EPIC-Oxford: Diverticular Disease (2011)
After a mean follow-up time of 11.6 years, EPIC Oxford found that vegetarians had a 31% lower risk (0.69, 0.55-0.86) of diverticular disease compared with meat eaters (14).
EPIC-Oxford: Cataracts (2011)
In an analysis of EPIC-Oxford participants ≥ 40 years old, a statistically significant, lower risk of cataracts was found in vegetarians than meat-eaters (9). Vegans had a 40% lower risk of cataracts than those eating more than 100 g/day of meat. More detailed results are in the table below. The results mainly applied to those aged ≥ 65 years at recruitment.
|Risk of Cataracts in EPIC-Oxford9|
|High meat eaters, ≥ 100 g/day||329||1.00|
|Moderate meat eaters, 50 – 99 g/day||489||0.96, 0.84-1.11|
|Low meat eaters, < 50 g/day||301||0.85, 0.72-0.99|
|Fish eaters||148||0.79, 0.65-0.97|
|Lacto-ovo vegetarians||198||0.70, 0.58-0.84|
|Results adjusted for: age, smoking, ethnicity, prior high blood pressure, receipt of long-term medical treatment, hormone replacement therapy.|
EPIC-Oxford: Heart Disease Mortality (2009)
Results of death from heart disease, stroke, and all causes through June 2007 were released from EPIC-Oxford in 2009 (8). There were no statistically significant differences between vegetarians and meat-eaters for deaths from any of these diseases. However, this was likely due to the relatively few deaths from heart disease overall, not allowing for statistical power.
There was also no difference between regular meat-eaters, fish-only meat eaters, and vegetarians. Vegans were not separated from the vegetarian group.
Results were adjusted for age, sex, smoking, and alcohol consumption.
The authors stated:
Average meat intake among the meat eaters was only moderate, at 79 g/d in men and 67 g/d in women; these intakes are much lower than those reported in the National Diet and Nutrition Survey for the United Kingdom. Consumption of vegetables and fruit was higher among vegetarians than among nonvegetarians, but the differences were not large (< 20%). Thus, if high intakes of meat had an adverse effect and high intakes of fruit and vegetables had a beneficial effect, the relatively low meat intake and high fruit and vegetable intake of the nonvegetarians in this cohort could reduce the chance of observing lower mortality rates in the vegetarians than in the nonvegetarians.
EPIC-Oxford: Bone Fractures (2007)
In a 2007 analysis of bone fractures from EPIC-Oxford, vegans had a 37% higher fracture rate than meat-eaters (10). When the results were adjusted for calcium intake, the vegans no longer had a higher fracture rate. And among the subjects who got 525 mg of calcium a day (only 55% of the vegans compared to about 95% of the other diet groups), vegans had the same fracture rates as the other diet groups.
More information can be found in the article Calcium Part 2—Research.
Epic-Oxford: Preliminary Mortality Results (2003)
In 2003, preliminary results from EPIC-Oxford (Oxford component of the European Prospective Investigation into Cancer and Nutrition) comparing death rates of 46,562 subjects were reported (7). About 33% of the subjects were vegetarian (including many vegans). The results showed no statistically significant differences between the vegetarians and non-vegetarians in any of the mortality categories which included cancer, circulatory disease, ischemic heart disease, cerebrovascular disease (stroke), all other causes, and all causes combined.
British Mortality (2002)
In 2002, the final death rate ratios were published from the Health Food Shoppers Study and the Oxford Vegetarian Study (15). Because data on these populations have been published in the 1999 meta-analysis and many of the subjects went on to participate in EPIC-Oxford, I will not repeat the findings here.
However, there was one finding that has not been published elsewhere and it was the death rate ratio for “mental and neurological diseases.” The Health Food Shoppers Study showed no difference between vegetarians and non-vegetarians, but the Oxford Vegetarian study found vegetarians to have 2.5 times the number of deaths (2.46, 1.21-5.01). After excluding the first five years of follow-up, the death risk ratio went down to 2.21 (1.02 – 4.78). While the rate was statistically significant, and the difference was quite large, the confidence interval is very wide, indicating that it could possibly be due to random chance. Another likely explanation is that vitamin B12 deficiency was leading to neurological disorders. A less likely possibility, but one the that cannot be completely ruled out, is that vegetarians were suffering from low DHA levels.
Two analyses from the Adventist Health Study compared dementia in vegetarians to non-vegetarians, and were published in the same paper (16). The subjects had to have been vegetarian for at least 20 years. In the first, 68 vegans, 68 lacto-ovo vegetarians, and 136 omnivores were matched for age, gender, and zip code. The non-vegetarians were two to three times as likely to have dementia, though the finding was not statistically significant until previous meat-consumption was taken into account and even then it was just borderline. No difference in dementia was found between vegetarians and non-vegetarians among a second group of unmatched subjects. A 2013 report on mortality from Adventist Health Study-2 found no difference between vegetarians and non-vegetarians for death rates from neurologic diseases, though pesco- and semi-vegetarians were included in the “vegetarian” group (17).
Based on the fact that Seventh-day Adventists have long been aware that vitamin B12 can’t be found in plants, and are a tight community with strong support for vegetarianism, vitamin B12 deficiency might be less likely in the Adventist Health Study than in British vegetarians.
Adventist Health Study-2 (AHS2)
In 2013, death rates for the first 5.8 years of Adventist Health Study-2 (AHS2) were released (17). The findings worth mentioning are in the table below.
|Death Rates in Adventist Health Study-2|
|All Vegetarians||.88, .80-.97||.90, .82-.98|
|Vegetarian Men||.82, .72-.94||.83, .72-.96|
|Vegans||.85, .73-1.01||.84, .72-1.00|
|Lacto-Ovo||.91, .82-1.00||.92, .84-1.02|
|Pesco||.81, .69-.94||no impact|
|Semi||.92, .75-1.13||no impact|
|All Vegetarians||.87, .75-1.01||no impact|
|Vegetarian Men||.71, .51-.90||no impact|
|Vegan Men||.58, .38-.89||no impact|
|All Vegetarians||.93, .67-1.29||no impact|
|All Vegetarians||.48, .28-.82||no impact|
|Endocrine (primarily diabetes)|
|All Vegetarians||.61, .40-.92||.71, .46-1.09|
|Vegans||.74, .56-.99||no impact|
|Pesco||.71, .54-.94||.77, .60-1.00|
|All rates are compared to non-vegetarians.
Vegan – eat dairy, eggs, meat (including fish) < once per month (5,548 people)
Lacto-Ovo – meat (including fish) < once per month, dairy or eggs > once per month (21,177 people)
Pesco – non-fish meat < once per month, fish > once per month (7,194 people)
Semi – non-fish meat ≥ once per month, all meats > once per month but < once per week (4,031 people)
All Vegetarians – includes all the diet groups listed above.
Non-vegetarians – 35,359 people
AAdjusted for age, race (black, nonblack), smoking, exercise, income, education, marital status, alcohol, geographic region, and sleep.
BFurther adjusted for body mass index (BMI)
When combining vegans, lacto-ovo-vegetarians, pesco-vegetarians, and semi-vegetarians into one group (see table footnote for definitions of these groups), vegetarians had a 12% lower risk of mortality. After adjusting for body mass index (BMI), the finding dropped to 10%; given that diet most likely effects BMI, adjusting for BMI might mask a true effect of diet.
Vegans had a 15% lower risk of death, but it was not quite statistically significant.
The difference in mortality rates can mostly be explained by a lower incidence of cardiovascular disease among vegetarian men. Vegetarian women had about the same rates as non-vegetarian women. This is similar to the findings from the first Adventist Health Study. There was also a benefit for all vegetarians for death from renal and endocrine (mostly diabetes) disease. No differences were found for all vegetarians for infectious or respiratory diseases.
No differences were found between diet groups for “all cancer.” Another study out of AHS-2 found lower rates for incidence of female-specific and gastrointestinal cancers between all vegetarians and non-vegetarians, discussed in the article Cancer Rates of Vegetarians. There were not enough deaths from cancer to break them down into types for this paper.
Adjusting for caloric intake resulted in only negligible changes for the results.
There are a couple caveats to this study. The diet was only measured at baseline, and not re-assessed again (as distinct from EPIC-Oxford which measured diet numerous times during the study). And only 5.7 years is not much follow-up which the researchers believed would bias the results towards finding no differences between diet groups.
In comparing their findings to British vegetarians, the researchers said:
The lack of similar findings in British vegetarians remains interesting, and this difference deserves careful study. In both cohorts, the non-vegetarians are a relatively healthy reference group. In both studies, the nutrient profiles of vegetarians differ in important ways from those of non-vegetarians, with vegetarians (especially vegans) consuming less saturated fat and more fiber. It appears that British vegetarians and US Adventist vegetarians eat somewhat differently. For instance, the vegetarians in our study consume more fiber and vitamin C than those of the EPIC-Oxford cohort: mean dietary fiber in EPIC-Oxford vegans was 27.7 g/d in men and 26.4 g/d in women compared with 45.6 g/d in men and 47.3 g/d in women in AHS-2 vegans; mean vitamin C in EPIC-Oxford vegans was 125 mg/d in men and 143 mg/d in women compared with 224 mg/d in men and 250 mg/d in women in AHS-2 vegans. Individuals electing vegetarian diets for ethical or environmental reasons may eat differently from those who choose vegetarian diets primarily for reasons of perceived superiority for health promotion. We believe that perceived healthfulness of vegetarian diets may be a major motivator of Adventist vegetarians.
In 2013, rates of hypothyroidism in AHS2 were published (18). Rates were reported both cross-sectionally at baseline and prospectively after about 4-6 years of follow-up (exact number not specified).
Results are in the table below. For the cross-sectional arm, lacto-ovo vegetarians were more likely to have been treated for hypothyroidism in the previous year, although the statistical significance was not strong. There were no other statistically significant findings, although vegans had a trend towards less hypothyroidism in the prospective arm.
According to the authors, “While vegan diets are associated with lower body weight, which may protect against hypothyroidism, the lower risk among vegans existed even after controlling for [body mass index] and potential demographic confounders.” It would have been interesting to see what the results were without adjusting for body mass index, but that information wasn’t reported.
Interestingly, adding salt to foods more often was associated with increased hypothyroidism among the entire population. Whether the salt was iodized was not determined (it likely was since most table salt in the U.S. is iodized).
|Hypothyroidism Rates in Adventist Healthy Study-2 (2013)18|
|semi-vegetarian||1.04 (0.91-1.19)||0.87 (0.65-1.17)|
|pesco-vegetarian||1.02 (0.90-1.15)||0.87 (0.65-1.17)|
|lacto-ovo vegetarian||1.09 (1.01-1.18)||1.07 (0.91-1.24)|
|vegan||0.89 (0.78-1.01)||0.78 (0.59-1.03)|
|Results adjusted for age, body mass index, gender, race, income, education, salt use.|
Adventist Health Study
- Data collected from 1976-1988
- 34,192 participants, members of the Seventh-day Adventist church
- 29% were vegetarian; 7-10% of the vegetarians were vegan.
|Table 4. Adventist Health Study Results|
|Heart disease||38% Lower for Men||No Difference for Women|
|Overall mortality||Lived 3.21 more yrs||Lived 2.52 more yrs|
|aBMI—Body Mass Index. A measure of healthy body weight. Lower than 20 is underweight, while ≥ 25 is overweight.|
Compared to the non-vegetarians, vegetarians had about:
- 1/2 the high blood pressure and diabetes
- 2/3 the rheumatoid arthritis
Life expectancies in the Adventist Health Study were published in 2001 (2). They showed that white, non-Hispanic Seventh-day Adventists live longer than other white Californians (7.28 more years for men, 4.42 more years for women). According to the researchers, this group of Seventh-day Adventists was the longest-lived, formally studied population in the world (with an average lifespan of 78.5 years for men, 82.3 for women).
The following variables were shown to increase life expectancy:
- vegetarian diet
- eating nuts regularly
- physical activity
- lower body weight
- no smoking
Heidelberg Study (German Vegetarians)
In 2005, results from 21 years of follow-up of the Heidelberg Study were published (5). The 1999 Meta-Analysis included only 11 years of follow-up from this population.
There were only 60 vegans in the Heidelberg Study, which was 3% of the study group. The authors state, “Being a vegan was associated with a higher mortality risk (1.59; 95% CI (confidence interval), 0.98-2.59) than being a lacto-ovo vegetarian (1.08; 95% CI, 0.86-1.34), when compared with nonvegetarians with moderate meat/ fish consumption, accounting for all other variables (data not shown).”
|Table 5. Heidelberg Study Results|
|Activity||Rates and Confidence Interval||Notes|
|Ischemic heart disease|
|Ischemic heart disease|
As you can see, these findings were not statistically significant, but they were close for the vegans. Also note that these “nonvegetarians” were semi-vegetarians, eating very little meat. It wouldn’t surprise me to find out these vegans did not do much to supplement their diet with vitamin B12.
Other results of interest are shown in Table 5.
The authors had the following to say about physical activity:
Our findings corroborate epidemiologic evidence indicating that regular and vigorous physical activity is an effective means of preventing circulatory diseases and cancers at different sites.
In summary, not enough is yet known about vegan mortality to draw any conclusions other than that vegans do not have unusually high rates of mortality and they probably do better than the average person due either to diet or a healthier lifestyle.
1. Fraser GE. Associations between diet and cancer, ischemic heart disease, and all-cause mortality in non-Hispanic white California Seventh-day Adventists. Am J Clin Nutr. 1999 Sep;70(3 Suppl):532S-538S.
3. Key TJ, Fraser GE, Thorogood M, Appleby PN, Beral V, Reeves G, Burr ML, Chang-Claude J, Frentzel-Beyme R, Kuzma JW, Mann J, McPherson K. Mortality in vegetarians and nonvegetarians: detailed findings from a collaborative analysis of 5 prospective studies. Am J Clin Nutr. 1999 Sep;70(3 Suppl):516S-524S.
5. Chang-Claude J, Hermann S, Eilber U, Steindorf K. Lifestyle determinants and mortality in German vegetarians and health-conscious persons: results of a 21-year follow-up. Cancer Epidemiol Biomarkers Prev. 2005 Apr;14(4):963-8.
8. Key TJ, Appleby PN, Spencer EA, Travis RC, Roddam AW, Allen NE. Mortality in British vegetarians: results from the European Prospective Investigation into Cancer and Nutrition (EPIC-Oxford). Am J Clin Nutr 2009;89(suppl):1S-7S.
11. Huang T, Yang B, Zheng J, Li G, Wahlqvist ML, Li D. Cardiovascular Disease Mortality and Cancer Incidence in Vegetarians: A Meta-Analysis and Systematic Review. Ann Nutr Metab. 2012 Jun 1;60(4):233-240. (Link)
13. lCrowe FL, Appleby PN, Travis RC, Key TJ. Risk of hospitalization or death from ischemic heart disease among British vegetarians and nonvegetarians: results from the EPIC-Oxford cohort study. Am J Clin Nutr. 2013 Jan 30. [Epub ahead of print]
14. Crowe FL, Appleby PN, Allen NE, Key TJ. Diet and risk of diverticular disease in Oxford cohort of European Prospective Investigation into Cancer and Nutrition (EPIC): prospective study of British vegetarians and non-vegetarians. BMJ. 2011 Jul 19;343:d4131.