Cancer
A source-by-source look at vegetarian and vegan cancer risk, the meat and dairy evidence, and the trials that did not pan out.
Ask whether a vegan diet prevents cancer and you will hear two confident answers. One side points to large studies in which vegetarians and vegans get less cancer. The other points to diet trials that changed what people ate and found no effect. Both sides are citing real research.
This post reads that research at the source and grades it. Results from cells and animals are labeled as such. Results from observational studies are described as associations, not proof. Where trials and observational studies disagree, both are shown, and every number comes from the paper it is attributed to.
The bottom line, in brief: in the best available data, the more plant-based the diet, the lower the cancer risk tends to be. The effect is moderate, part of it runs through body weight, the few findings that point the other way rest on small numbers and on low intakes of nutrients a well-planned vegan diet covers, and no diet guarantees anything. Later sections take on specific foods, the mechanisms proposed to explain the link, what the evidence says after a diagnosis, and what it cannot show.
The short answer
Does a vegan diet lower cancer risk?
Probably, by a moderate amount. A 2026 meta-analysis by Aune and colleagues combined seven prospective studies from the UK, the Netherlands, and the Adventist Health Studies [1]. Compared with non-vegetarians, vegetarians had a 13% lower risk of cancer overall (relative risk 0.87, 95% CI 0.84 to 0.91) and vegans a 23% lower risk (0.77, 0.70 to 0.85). Using World Cancer Research Fund criteria, the authors rated the evidence as “probable” for vegetarian diets and total cancer, and “limited-suggestive” for vegan diets [1].
Part of the gap reflects body weight. After adjusting for BMI, the vegetarian estimate moved from 0.87 to 0.92 and the vegan estimate from 0.77 to 0.82; both stayed statistically significant [1].
The largest pooled analysis so far, which reported results for specific cancers rather than for cancer overall, found that vegetarians had lower risks of cancers of the pancreas, breast, prostate, and kidney, and of multiple myeloma [2]. Its two higher estimates, discussed below, each rest on small numbers of cases, and the authors attribute both to low intakes of certain nutrients, such as calcium, which a well-planned vegan diet is designed to supply.
Does the type of plant-based diet matter for cancer risk?
It appears to. A 2023 meta-analysis of 76 publications, covering more than 2.2 million people, scored diets on plant-based indices rather than labels [3]. People with the most plant-based diets had a 12% lower cancer risk (0.88, 0.84 to 0.92), and 14% lower when the score emphasized vegetables, fruits, whole grains, and legumes (0.86, 0.80 to 0.92). Diets high in what the authors call unhealthy plant foods, “such as refined grains, starchy vegetables (e.g., white potatoes), or sugar,” were linked to a 7% higher cancer risk (95% CI 2% to 12%) [3]. The protective associations come from whole plant foods, not from a plant-based label.
How much does processed meat raise cancer risk?
In 2015, a working group of 22 scientists convened by the International Agency for Research on Cancer (IARC) classified processed meat as “carcinogenic to humans” (Group 1), on the basis of colorectal cancer, and red meat as “probably carcinogenic to humans” (Group 2A) [4]. A meta-analysis of ten cohort studies they reviewed found an 18% higher colorectal cancer risk per 50 g of processed meat a day, and 17% per 100 g of red meat. For red meat, the working group called the human evidence “limited” [4].
Not everyone reads that evidence the same way. A 2019 review by Han and colleagues in Annals of Internal Medicine rated the certainty of the evidence as low to very low, and the possible absolute effects of red and processed meat on cancer as “very small” [5]. Part of that rating comes from the method: under the GRADE system the authors used, “evidence from observational studies begins at low certainty.” After publication, the journal issued a correction because one author had not disclosed a grant from Texas A&M AgriLife Research; his updated disclosure also noted earlier funding from the International Life Sciences Institute (North America) [5].
Do dairy and eggs raise cancer risk?
The evidence is weaker than for processed meat and points in different directions by cancer type. Dairy has been linked to higher prostate cancer risk in the Adventist Health Study-2 [6], but to lower colorectal cancer risk in a large UK study whose authors attribute the benefit to calcium [7]. Pooled data on eggs show small associations that are borderline in places [8]. Each food is graded separately later in this post.
Why might animal foods raise cancer risk?
The best-supported pathway runs through IGF-1, a growth factor. Higher blood levels are linked to prostate and breast cancer in genetic as well as observational studies [9] [10], and vegans have lower levels than meat eaters [11]. Heme iron in red meat has consistent animal evidence for colorectal cancer, but the human evidence is mixed: one meta-analysis found a suggestive association [12], while a 2026 analysis with up to 36 years of follow-up found none [13]. Several other widely cited mechanisms rest mainly on cell or animal studies, as the section on mechanisms shows.
Which foods are linked to lower cancer risk?
Fiber and whole grains have the most consistent evidence, especially for colorectal cancer [14], although part of the whole-grain association may reflect other healthy habits [7]. Fruits and vegetables as a group show a smaller association with total cancer [15]. The evidence for cruciferous vegetables and mushrooms is weaker: suggestive at best for a few cancers, and for mushrooms driven largely by case-control studies [16] [17].
Does diet matter after a cancer diagnosis?
Less clearly than for prevention. The most comprehensive review of breast cancer survivors grades body fatness after diagnosis as the best-supported factor, with physical activity next and most dietary factors inconclusive [18]. Randomized trials of vegetable-rich diets did not reduce recurrence or progression [19] [20]. Specific findings, such as worse outcomes with whole milk and better ones with soy, come from observational studies.
Can a vegan diet guarantee you won’t get cancer?
No. Even the strongest favorable estimates describe lower risk, not no risk. And when researchers have tested dietary change directly, in randomized trials, the results have been mixed at best, as the next section shows.
How strong is the evidence?
Why do diet and cancer studies disagree?
Because they answer different questions. Four kinds of evidence appear in this post:
- Cell and animal studies show whether a mechanism is possible, often under conditions far from human eating. They are labeled every time they appear here.
- Prospective cohort studies follow large groups for years and record who develops cancer. They capture real diets, usually by questionnaire, but people who eat differently also live differently, so cohorts show associations rather than cause and effect.
- Mendelian randomization uses genetic variants as stand-ins for a risk factor. Because these variants “are randomly allocated and fixed at conception,” the method is less prone to confounding and reverse causation than ordinary studies of blood levels [9]. It can test something genes influence, such as the blood level of a hormone, but not a diet itself.
- Randomized trials assign the diet, which deals with confounding. But cancer develops over years, and long dietary trials are hard to run and hard to stick to, so several of the best-known diet trials test whether changing diet after a diagnosis prevents a recurrence.
Can changing your diet after a cancer diagnosis prevent recurrence?
Two major trials of vegetable-rich diets found no effect on cancer outcomes.
In the Women’s Healthy Eating and Living (WHEL) trial, 3,088 women previously treated for early-stage breast cancer were randomized to a telephone counseling program, with cooking classes and newsletters, that promoted vegetables, fruit, fiber, and lower fat, or to printed “5-A-Day” guidance [19]. The intervention changed diets: through four years, the intervention group ate 65% more vegetable servings, 25% more fruit, and 30% more fiber, and got 13% less of its energy from fat, than the comparison group. It did not change cancer outcomes. Over 7.3 years, 16.7% of the intervention group and 16.9% of the comparison group had an invasive breast cancer event (hazard ratio 0.96, 95% CI 0.80 to 1.14), and deaths did not differ either (0.91, 0.72 to 1.15) [19].
In the Men’s Eating and Living (MEAL) trial, 478 men with early-stage prostate cancer on active surveillance were randomized to telephone counseling promoting seven or more daily vegetable servings, or to written information about diet and prostate cancer [20]. Time to progression did not differ (adjusted hazard ratio 0.97, 0.76 to 1.25). The authors add that the study “may have been underpowered to identify a clinically important difference.” That makes MEAL a finding of no evidence of benefit, which is not the same as evidence of no benefit.
Two low-fat trials point the other way, with caveats. In the Women’s Intervention Nutrition Study (WINS), 2,437 women with resected breast cancer were randomized to a fat-reduction program or a control group [21]. At an interim analysis after about five years, run when funding for the intervention ended, the intervention group had a 24% lower risk of relapse (hazard ratio 0.76, 0.60 to 0.98), and its members weighed about 6 pounds less on average. In a secondary analysis of the Women’s Health Initiative, which randomized 48,835 postmenopausal women to a low-fat dietary pattern or their usual diet, the 1,764 women later diagnosed with breast cancer had better overall survival if they were in the low-fat group (hazard ratio 0.78, 0.65 to 0.94) [22]. Deaths from breast cancer itself were fewer, but that difference was not statistically significant (0.86, 0.64 to 1.17).
Taken together, these trials do not show that a healthier diet on its own prevents recurrence. The two that reported a benefit tested low-fat diets, not vegan ones, and in WINS the intervention group also weighed less than the control group.
Do antioxidant supplements prevent cancer?
The large trials say no, and some found harm. The Alpha-Tocopherol, Beta Carotene Cancer Prevention Study randomized 29,133 male smokers aged 50 to 69 in Finland to vitamin E (50 mg a day), beta carotene (20 mg a day), both, or placebo for five to eight years [23]. Men given beta carotene had an 18% higher incidence of lung cancer (95% CI 3% to 36%). The trial had been run because epidemiologic studies linked vegetables rich in beta carotene with lower cancer risk.
The Selenium and Vitamin E Cancer Prevention Trial (SELECT) randomized 35,533 men. Those given 400 IU of vitamin E a day had a 17% higher risk of prostate cancer than those given placebo (hazard ratio 1.17, 99% CI 1.004 to 1.36), an absolute increase of 1.6 cases per 1,000 person-years [24]. A Cochrane review of 78 randomized trials with 296,707 participants, healthy people and people with various diseases, concluded that beta carotene and vitamin E “seem to increase mortality, and so may higher doses of vitamin A” [25].
The lesson for the rest of this post: a food and an isolated compound from that food are not interchangeable. A protective association for vegetables is not evidence for a pill.
How should you read the numbers in this post?
Most results here are relative risks or hazard ratios. A value of 0.87 means 13% lower risk than the comparison group; a value of 1.18 means 18% higher. The range in parentheses is the 95% confidence interval. If it crosses 1.00, the data are compatible with no difference at all. Relative changes also say nothing about how common a cancer is to begin with, so the same relative change can mean very different numbers of actual cases.
Do vegetarians and vegans get less cancer?
The first two cohorts below, the Adventist Health Study-2 and EPIC-Oxford, are among the seven studies in Aune’s meta-analysis, so they are part of that result rather than independent confirmation of it [1].
What did the Adventist Health Study-2 find about vegetarians and cancer?
The Adventist Health Study-2 follows Seventh-day Adventists, a population with “a very low percentage of alcohol consumption or cigarette smoking” and a wide variety of diets [26]. Among 69,120 participants, 2,939 cancers were identified. Vegetarians as a group had an 8% lower risk of cancer overall than non-vegetarians (hazard ratio 0.92, 0.85 to 0.99), and vegans a 16% lower risk (0.84, 0.72 to 0.99). Vegans also had a lower risk of female-specific cancers (0.66, 0.47 to 0.92) [26].
For prostate cancer, among 26,346 men with 1,079 cases, vegans had a 35% lower risk (0.65, 0.49 to 0.85) [27]. By race, the association held for white men (0.63, 0.46 to 0.86); for Black men the estimate was similar (0.69, 0.41 to 1.18), based on 15 cases among Black vegans, but the interval included no effect, and the test for a difference between the two groups was not significant (P = 0.76). For advanced or high-grade cancers (237 cases), the vegan estimate was also similar, 0.70 (0.41 to 1.21), though again with a wider interval that included no effect.
Breast cancer showed a weaker pattern. Among 50,404 women with 892 cases, vegetarians as a group had no lower risk than non-vegetarians (0.97, 0.84 to 1.11) [28]. Vegans had a lower point estimate, 0.78, but the confidence interval (0.58 to 1.05) included no effect. The authors’ summary: “lower risk in vegans is possible.” They also note that the non-vegetarian comparison group ate little meat, about 54 g a day on average, which “may cause an underestimation of the effect of a vegetarian diet” compared with a more typical non-vegetarian diet.
What did EPIC-Oxford find about vegetarians and cancer?
EPIC-Oxford recruited 63,550 men and women across the UK in the 1990s; its postal recruitment “was designed to recruit as many vegetarians and vegans as possible” [29]. The cohort as a whole had low cancer rates: incidence among all participants was 72% of national rates (standardized incidence ratio 72%, 95% CI 69% to 75%), and the authors note that rates for “both the vegetarians and the nonvegetarians” were low compared with national rates. That matters for interpretation. The comparison inside the study is between two low-risk groups, not between vegetarians and the general public.
A 2014 update pooled EPIC-Oxford with the earlier Oxford Vegetarian Study: 61,647 people, including 32,491 meat eaters, 8,612 fish eaters, and 20,544 vegetarians (2,246 of them vegans), followed for an average of 14.9 years, with 4,998 cancers [30]. Compared with meat eaters, vegetarians had a 12% lower risk of all cancers combined (0.88, 0.82 to 0.95), and so did fish eaters (0.88, 0.80 to 0.97). Vegans, analyzed separately, had a 19% lower risk (0.81, 0.66 to 0.98). That vegan result rested on 105 cancers, and after adjustment for BMI it was no longer statistically significant (0.82, 0.68 to 1.00).
What did the largest pooled study of vegetarians and cancer find?
A 2026 pooled analysis by Dunneram and colleagues combined nine prospective cohorts from the UK, the US, Taiwan, and India: 1,645,555 meat eaters, 57,016 poultry eaters, 42,910 pescatarians, 63,147 vegetarians, and 8,849 vegans, followed for a median of 16 years [2]. It reported results for 17 specific cancers rather than for cancer overall. Compared with meat eaters, a group defined as anyone eating any red or processed meat, vegetarians had lower risks of cancers of the pancreas (0.79, 0.65 to 0.97), breast (0.91, 0.86 to 0.97), prostate (0.88, 0.79 to 0.97), and kidney (0.72, 0.57 to 0.92), and of multiple myeloma (0.69, 0.51 to 0.93). The one higher estimate for vegetarians, for squamous cell carcinoma of the oesophagus (1.93, 1.30 to 2.87), rests on only 31 cases in three UK studies; in sensitivity analyses, it and the lower kidney cancer risk were the most consistent findings. The authors suggest this result, like the vegan colorectal result below, “might be due to a higher prevalence of inadequate intakes of some nutrients,” noting that intakes of protein, vitamin B12, and vitamin D were lower in vegetarians, and lower still in vegans, who also ate less calcium. Those are the nutrients a well-planned vegan diet deliberately covers, with fortified foods or supplements where needed.
Two cautions apply. Processed meat intake among the meat eaters was “moderately low,” with a median across cohorts of about 16 g a day, so the comparison group was not a heavy meat-eating population. And five of the nine cohorts, including the Adventist Health Study-2 and EPIC-Oxford, are also in Aune’s meta-analysis, so the two analyses are not independent [1]. The authors conclude that “the generalisability should be considered cautiously” [2].
Does a vegetarian or vegan diet affect colorectal cancer risk?
In the 2009 EPIC-Oxford analysis, vegetarians had a higher rate of colorectal cancer than meat eaters (incidence rate ratio 1.39, 95% CI 1.01 to 1.91) [29]. The authors called the finding “surprising” and wrote that it “might be partly due to chance.” In the larger 2014 pooled analysis, the vegetarian estimate was 1.03 (0.84 to 1.26), meaning no detectable difference, while fish eaters had a 34% lower risk (0.66, 0.48 to 0.92) [30].
Pooling six of the studies in Aune’s meta-analysis, vegetarians had a 14% lower risk of colorectal cancer (0.86, 0.76 to 0.97) [1]. For vegans, the pooled estimate from three studies was 1.02 (0.71 to 1.48), and the authors noted that the vegan studies “appeared to deviate in sensitivity analyses,” so the vegan results for colorectal and prostate cancer “may not have been representative for the overall evidence base.”
The largest pooled analysis reported a higher estimate for vegans, but it carries little weight on its own. In Dunneram and colleagues’ nine cohorts, vegetarians had no difference in colorectal cancer risk from meat eaters (1.03, 0.94 to 1.13) and pescatarians a lower risk (0.85, 0.77 to 0.93) [2]. The vegan estimate (1.40, 1.12 to 1.75) rests on 93 cases in seven studies, with fewer than 10 in five of them, and the authors write that it “should be interpreted with caution.” It was no longer statistically significant once the first four years of follow-up were excluded (1.25, 0.95 to 1.64), an analysis the authors ran to check for “reverse causality, where undiagnosed cancer might influence diet,” as when someone changes their diet because they already feel unwell. The authors also note possible misclassification of diet group at baseline and that “we did not assess duration of diet adherence before participants joined the cohorts,” so the analysis could not separate long-term vegans from recent ones, although in cohorts with such data, most vegetarians had followed their diet for at least several years. (Among never smokers, the estimate was 1.67, 1.28 to 2.18.)
The authors’ own explanation points to the fix. They suggest the result “might be due to their low average intake of calcium”: vegans had the lowest calcium intakes in every cohort with nutrient data, with a median of 590 mg a day against a UK reference intake of 700 mg [2]. They also note the World Cancer Research Fund’s conclusion that dairy products “and calcium supplements, probably protect against colorectal cancer,” so the protective calcium does not have to come from milk. A well-planned vegan diet that includes calcium-rich and fortified foods, or a supplement where needed, closes that gap; our colon cancer post lists the best plant sources and a daily target.
The fair summary: vegetarian diets are associated with lower or similar colorectal cancer risk. For vegans, the meta-analysis found no difference (1.02) [1], and the one higher estimate (1.40) [2] rests on few cases, lost statistical significance once the first four years of follow-up were excluded, and points to calcium, a nutrient a well-planned vegan diet can supply. The section on dairy below explains why calcium matters.
How much of the vegetarian cancer advantage comes from lower body weight?
Some of it, but not all. In Aune’s meta-analysis, adjusting for BMI shrank the vegetarian advantage for total cancer from 13% to 8% (0.87 to 0.92) and the vegan advantage from 23% to 18% (0.77 to 0.82), and both remained statistically significant [1]. The authors used the unadjusted figures for their main analysis because randomized trials show vegetarian and vegan diets can help with weight loss, so BMI “can therefore be considered a mediator”: lower weight may be one of the ways these diets work, not just a confounder.
Meat, dairy, and eggs, graded one at a time
“Animal foods” is not one exposure. The evidence differs sharply by food and by type of cancer, so each food gets its own grade here. All of the human evidence in this section is observational; the mechanisms proposed to explain it come in the next section.
Is processed meat worse than unprocessed red meat for cancer?
For colorectal cancer, the evidence says yes. IARC’s working group placed processed meat in Group 1 because its association with colorectal cancer was consistent “across studies in different populations, which make chance, bias, and confounding unlikely as explanations” [4]. It called the human evidence for unprocessed red meat “limited” and placed it a step lower, in Group 2A.
A 2021 meta-analysis of prospective studies by Farvid and colleagues, drawing on 148 published articles, gives the same ordering [31]. Comparing the highest with the lowest intakes, processed meat was associated with an 18% higher risk of colorectal cancer (1.18, 1.13 to 1.24) and unprocessed red meat with a 10% higher risk (1.10, 1.03 to 1.17). These are category comparisons, so they cannot be lined up directly with IARC’s per-gram figures, but they point the same way.
Does red meat raise the risk of cancers other than colorectal cancer?
Possibly, but the evidence is thinner. In Farvid’s analysis, the highest red meat intakes were associated with higher risk of breast cancer (1.09, 1.03 to 1.15), endometrial cancer (1.25, 1.01 to 1.56), lung cancer (1.26, 1.09 to 1.44), and hepatocellular carcinoma (1.22, 1.01 to 1.46) [31]. Processed meat was associated with higher risk of breast cancer (1.06, 1.01 to 1.12) and lung cancer (1.12, 1.05 to 1.20). For prostate cancer, neither red meat (1.07, 0.97 to 1.19) nor processed meat (1.04, 0.98 to 1.10) showed a significant association.
IARC’s working group also noted positive associations between red meat and cancers of the pancreas and the prostate, “mainly advanced prostate cancer,” and between processed meat and stomach cancer [4]. Some of these estimates are larger than the colorectal one, but they rest on less data: Farvid’s red meat estimates for lung and endometrial cancer pool 7 and 6 risk estimates, against 22 for colorectal cancer [31]. That is why they rank below the colorectal evidence here.
Does dairy increase the risk of prostate cancer?
Some large cohorts suggest it may. The Adventist Health Study-2 is well placed to test the question because it includes many men with no or very low dairy intake and “much calcium from nondairy sources” [6]. Among 28,737 men followed for an average of 7.8 years, with 1,254 prostate cancers, those at the 90th percentile of dairy intake (430 g a day) had a 27% higher risk than those at the 10th percentile (20.2 g a day) (hazard ratio 1.27, 1.12 to 1.43). The association held for advanced cancers (1.38, 1.02 to 1.88) and when vegans were excluded (1.22, 1.03 to 1.43), while calcium from non-dairy sources was not linked to higher risk. The authors found the steepest rise in risk at relatively low intakes.
The Physicians’ Health Study, which followed 21,660 U.S. male physicians for 28 years, found a weaker and more mixed picture [32]. Total dairy was not significantly associated with prostate cancer (1.12, 0.93 to 1.35, highest versus lowest intake). Skim and low-fat milk was associated with a 19% higher risk (1.19, 1.06 to 1.33), mainly for low-grade, early-stage, and screen-detected cancers. Whole milk was the only dairy food associated with fatal prostate cancer, and that estimate’s confidence interval crossed 1.00 (1.49, 0.97 to 2.28). Dairy intake was measured once, at baseline, an average of 14 years before diagnosis.
Whether dairy affects how prostate cancer progresses after a diagnosis is a separate question, covered later in this post.
Does dairy milk increase breast cancer risk?
One large cohort says it may; the wider literature has not agreed. In the Adventist Health Study-2, 52,795 North American women were followed for 7.9 years, with 1,057 breast cancers [33]. Comparing the 90th with the 10th percentile of intake, dairy milk was associated with a 50% higher risk (hazard ratio 1.50, 1.22 to 1.84). Full-fat and reduced-fat milk gave similar results, and cheese and yogurt showed no important associations. Substituting the median intake of dairy milk drinkers with the median intake of soy milk drinkers was associated with lower risk (0.68, 0.55 to 0.85). The authors disclose that three of them “are largely vegetarian.”
A commentary published alongside the study noted that the latest World Cancer Research Fund meta-analysis had found no association between dairy milk and breast cancer, and that the Adventist estimate was large compared with earlier prospective studies [34]. It flagged possible residual confounding: in this cohort, women who drink little or no dairy milk are thinner, and less likely to be screened for breast cancer or to take menopausal hormone therapy. It also estimated that the difference in blood IGF-1 between high and low milk drinkers would raise breast cancer risk by only “around 5%,” much less than the association observed, while adding that other growth factors and hormones in milk could play a role.
Is dairy linked to lower colorectal cancer risk?
It is, in a very large UK study, and any fair account of dairy has to include it. Papier and colleagues tested 97 dietary factors in 542,778 women in the Million Women Study, 12,251 of whom developed colorectal cancer over 16.6 years [7]. Calcium had the strongest inverse association (0.83 per 300 mg a day, 0.77 to 0.89), and dairy milk was linked to a 14% lower risk per 200 g a day (0.86, 0.81 to 0.92). After adjusting for calcium, the milk association was no longer evident (0.94, 0.86 to 1.02). A genetic analysis using a variant in the lactase gene as a stand-in for milk intake gave a larger estimate (0.60 per 200 g a day, 0.46 to 0.74). The authors conclude that dairy’s protective association “is driven largely or wholly by calcium.”
The authors of the largest pooled study of vegetarians make the same point, suggesting that the higher colorectal cancer estimate among vegans in their data “might be due to their low average intake of calcium” [2].
If calcium is the active ingredient, the relevant question for people who avoid dairy is whether they get enough calcium from other foods, not whether they drink milk. Our colon cancer post covers calcium and the colon in more detail. Dairy, in other words, cuts both ways depending on the cancer.
Do eggs raise cancer risk?
The evidence is limited, and most of the associations are small. A 2015 dose-response meta-analysis of prospective studies by Keum and colleagues found that for breast cancer, each additional five eggs a week carried a non-significant 5% higher risk (1.05, 0.99 to 1.11) [8]. In a non-linear model, five or more eggs a week was associated with a small but significant increase compared with none: 1.04 (1.01 to 1.07) at five eggs a week and 1.09 (1.03 to 1.15) at about nine. For fatal prostate cancer, five more eggs a week carried a 47% higher risk (1.47, 1.01 to 2.14), based on 609 cases and with evidence of small-study effects; there was no association with total prostate cancer. The authors write that “potential for publication bias and confounding temper our conclusion.”
In the Health Professionals Follow-up Study, which followed 27,607 men from 1994 to 2008, men eating 2.5 or more eggs a week had an 81% higher risk of developing lethal prostate cancer than men eating fewer than half an egg a week (1.81, 1.13 to 2.89), based on 199 lethal cancers [35].
Which animal foods have the strongest evidence for raising cancer risk?
Our reading of the evidence above, from strongest to weakest:
- Processed meat: the strongest case. It is in IARC Group 1, with consistent associations with colorectal cancer across many populations.
- Unprocessed red meat: a weaker case. Associations with colorectal cancer are modest, IARC rates the human evidence “limited,” and associations with other cancers rest on less data.
- Dairy: mixed. It was linked to higher prostate cancer risk in the Adventist cohort, with weaker and mixed results in the Physicians’ Health Study, and to higher breast cancer risk in one cohort whose finding conflicts with the wider literature, set against lower colorectal cancer risk that appears to come from calcium.
- Eggs: weak. Pooled associations are small and borderline in places, with one larger estimate for lethal prostate cancer from a single cohort and an acknowledged risk of publication bias.
Proposed mechanisms, graded by evidence
Mechanisms explain why a diet might matter. They are also where the evidence is most uneven: some rest on large human studies, others on cells in a dish or tumors in mice. Each is graded below by the best human evidence behind it.
Does IGF-1 raise cancer risk?
IGF-1 (insulin-like growth factor 1) is a growth factor, made by the liver and other tissues, that helps regulate cell proliferation, differentiation, and programmed cell death [36]. Of all the mechanisms in this section, its link to cancer has the strongest human evidence.
In a collaborative analysis of up to 20 prospective studies, each standard deviation higher blood IGF-1 was associated with a 9% higher risk of prostate cancer (odds ratio 1.09, 1.07 to 1.11) and of aggressive prostate cancer (1.09, 1.03 to 1.16) [9]. Mendelian randomization gave similar estimates (1.07, 1.00 to 1.15 overall; 1.10, 1.01 to 1.20 for aggressive disease). In 206,263 women in UK Biobank, each 5 nmol/L higher IGF-1 was associated with an 11% higher risk of breast cancer (hazard ratio 1.11, 1.07 to 1.16), and the genetic analysis pointed the same way (odds ratio 1.05, 1.01 to 1.10), for estrogen-receptor-positive but not estrogen-receptor-negative tumors [10].
Does a plant-based diet lower IGF-1?
It appears to, modestly. In a cross-sectional study of 292 British women, vegans had 13% lower serum IGF-1 than meat eaters and vegetarians, and intake of protein rich in essential amino acids “explained most of the differences” [11]. A meta-analysis of milk studies estimated a 0.10 standard deviation increase in IGF-1 per standard deviation increase in milk intake [37], a small effect, which fits the commentary estimate in the dairy section above that milk’s effect on IGF-1 would explain only about 5% of the breast cancer association [34].
One human study of protein and cancer death used NHANES III, a U.S. survey, to follow 6,381 adults aged 50 and over [38]. Among those aged 50 to 65, people getting 20% or more of their calories from protein were more than four times as likely to die of cancer as those getting less than 10% (hazard ratio 4.33, 1.96 to 9.56), and the associations were “either abolished or attenuated if the proteins were plant derived.” In people over 65 the pattern reversed: high protein intake was associated with reduced cancer and overall mortality. Diet was measured with a single 24-hour recall, and the senior author has an equity interest in L-Nutra, a company that develops medical food.
What does Laron syndrome show about IGF-1 and cancer?
People with Laron syndrome have a genetic defect in the growth hormone receptor that leaves them with very low IGF-1, and they rarely develop cancer. A worldwide survey found that none of 230 patients had developed cancer, compared with cancers among their relatives, although the patients were significantly younger than their first-degree relatives [39]. In an Ecuadorian cohort monitored for 22 years, people with the receptor defect had only one non-lethal malignancy, against 17% cancer prevalence among unaffected relatives [40]. Two of that study’s authors have filed patents related to its subject. These are rare genetic conditions, not diets. They show what lifelong, very low IGF-1 does, not what a vegan diet does.
Do lab studies show that diet lowers IGF-1 enough to slow cancer cells?
A series of UCLA studies grew prostate cancer cells in serum taken from men before and after the Pritikin program. After 11 days on a diet with less than 10% of calories from fat, in which protein came from “nonfat milk and small amounts of fish or fowl,” plus an hour of aerobic exercise a day, serum IGF-1 fell 20% and serum-stimulated growth of the cancer cells fell 30% in 14 men [36]. Adding IGF-1 back to the post-program serum reversed the effect [41].
These are cell-culture experiments with no control group, and the program combined a non-vegan diet with daily exercise. A cross-sectional comparison found that men in a long-term exercise program with no dietary intervention had similar IGF-1 levels and similar reductions in cancer-cell growth [42]. In the one randomized trial in this series, 18 men with prostate cancer ate either a low-fat, soy-supplemented diet or a Western diet for four weeks; IGF-1 rose in both groups, and the reduction in cancer-cell growth tracked changes in omega-6 and omega-3 fatty acids, not IGF-1 [43]. That trial was stopped early after an interim analysis. R. James Barnard, an author on all of these studies, has disclosed a consulting relationship with the Pritikin Longevity Center [44].
Does cutting methionine prevent cancer?
Methionine is an essential amino acid. In mice, restricting it improved responses to chemotherapy in two models of chemotherapy-resistant colorectal cancer grown from patients’ tumors, and to radiation in a model of soft-tissue sarcoma [45]. The human part of that study put six healthy middle-aged people on a low-methionine diet for about three weeks and measured metabolism, not cancer. Two of the authors hold patents related to targeting amino acid metabolism in cancer therapy.
Human cohort data on methionine intake point the other way. A meta-analysis of eight prospective cohorts with 431,029 participants found that the highest intakes were associated with no significant change in colorectal cancer risk (0.89, 0.77 to 1.03) and a lower risk of colon cancer (0.77, 0.64 to 0.92) [46]. There is no human evidence that eating less methionine prevents cancer.
Do branched-chain amino acids raise cancer risk?
Not in the large studies that have looked. Among 196,161 women in the Nurses’ Health Studies, followed for a median of 20.8 years with 10,046 breast cancers, the top fifth of branched-chain amino acid intake had no higher risk than the bottom fifth (1.05, 0.98 to 1.12) [47]. For colorectal cancer, across three U.S. cohorts, the top fifth had a weak inverse association (0.89, 0.80 to 1.00) that disappeared after accounting for dairy calcium (0.96, 0.85 to 1.08) [48].
Does TMAO from meat and eggs cause cancer?
TMAO is made by gut bacteria from choline, betaine, and carnitine [49], and meat, milk, and eggs are sources of choline [50]. A meta-analysis of 16 studies with 5,930 participants found no significant overall association between TMAO levels and cancer (odds ratio 0.97, 0.64 to 1.46), with subgroup results that differed by whether TMAO was measured in blood or urine. In a case-control study nested in the PLCO cohort, with 761 colorectal cancer cases and matched controls, TMAO and its precursors were not associated with colorectal cancer overall, but TMAO and choline were associated with higher risk of distal colorectal cancer (TMAO 1.90, 1.24 to 2.92), and red meat intake correlated weakly with blood TMAO (rho 0.10) [51]. One of its authors holds patents relating to cardiovascular diagnostics and therapeutics.
Dietary choline gives conflicting results. In the Health Professionals Follow-up Study, men in the highest fifth of choline intake had a 70% higher risk of lethal prostate cancer (1.70, 1.18 to 2.45) [50]. In the ARIC cohort of 6,528 men, choline intake was not associated with total or lethal prostate cancer [52].
Does endotoxin in meat drive inflammation and cancer?
Some foods carry bacterial fragments that switch on inflammation through Toll-like receptors. In a lab screen of 40 food extracts applied to human immune cells, these stimulants “were present mainly in meat products and processed foods, but were minimal or undetectable in fresh fruit and vegetables” [53]. That is a cell-culture finding.
In people, the evidence comes from short meal studies that measured blood markers, not cancer. In 10 healthy adults, a 910-calorie fast-food breakfast of egg and sausage muffin sandwiches with hash browns raised plasma endotoxin by 47% within three hours; in another 10, an equal-calorie breakfast of oatmeal, milk, orange juice, raisins, peanut butter, and an English muffin did not, even though that meal contained more endotoxin (5,700 versus 4,200 endotoxin units) [54]. In a second study, cream raised plasma endotoxin but orange juice did not, although the orange juice drink carried more endotoxin than the cream; the authors judged the endotoxin in the drinks “unlikely” to explain the rise [55]. That study was supported in part by the Florida Department of Citrus. The two research teams read their results differently. The first thought the meal’s own endotoxin probably contributed, while calling it “remarkable” that the fiber-and-fruit meal did not raise plasma endotoxin despite similar content; the second pointed to saturated fat increasing the gut’s permeability. Either way, none of these studies measured cancer.
Does saturated fat fuel cancer spread?
In mouse models of human oral cancer, palmitic acid or a high-fat diet “specifically boosts the metastatic potential” of a subset of cancer cells carrying the fat receptor CD36, and blocking CD36 with antibodies “causes almost complete inhibition of metastasis” [56]. This is animal research, and the authors declared competing financial interests. Palmitic acid is a major fat in palm oil and is also found in meat and dairy fat, but a 2017 review notes that changes in its intake “do not influence significantly its tissue concentration because the exogenous source is counterbalanced by PA endogenous biosynthesis” [57]. There is no human evidence yet that dietary palmitic acid drives metastasis.
Does heme iron explain the red meat link to colon cancer?
Possibly, but the human evidence is mixed. A meta-analysis of prospective cohorts with 566,607 people and 4,734 colon cancers found a relative risk of 1.18 (1.06 to 1.32) for the highest versus lowest heme iron intake, which the authors described as a “suggestive association” [12]. In rats with chemically induced colon cancer, dietary hemoglobin and red meat consistently promoted aberrant crypt foci, a putative precancerous lesion, and the proposed mechanisms are that heme catalyzes the formation of carcinogenic N-nitroso compounds and of damaging aldehydes from fat oxidation.
A 2026 analysis of three US cohorts, the Nurses’ Health Study I and II and the Health Professionals Follow-up Study, followed 211,851 participants for up to 36 years and documented 2,983 colorectal cancers [13]. Comparing the highest with the lowest fifth of intake, it found no association for total heme iron (hazard ratio 1.03, 0.91 to 1.17) or for heme from red and processed meat (1.02, 0.91 to 1.16), and the authors conclude that the study “does not support a strong role of heme iron intake in CRC risk”; they note that heme intakes in these cohorts spanned narrower ranges than in cohorts reporting positive associations, and suggest that “there may be factors specific to processed meat that contribute more to CRC risk.” Higher calcium intake was linked to lower risk in the same cohorts (0.77, 0.68 to 0.86, highest versus lowest fifth), and the authors found that benefit “regardless of heme iron intake.” So the animal and mechanistic evidence points toward heme, while the human evidence is inconsistent, and this large, long-term analysis found no association. Our colon cancer post covers heme in more depth, including the heme in some plant-based burgers.
Which cancer mechanisms have the strongest evidence?
Our reading, ranked by the human evidence that each one raises cancer risk. Animal and cell findings are noted, but they do not move a mechanism up the list; a mechanism with no human data ranks above one that human studies have tested and found no harm from.
- IGF-1: the strongest. Blood IGF-1 is linked to prostate and breast cancer in cohorts and in genetic analyses, and plant-based diets are associated with modestly lower levels. How much of diet’s effect on cancer runs through IGF-1 has not been measured directly, and the serum-on-cancer-cells studies add little.
- Heme iron: consistent animal data and a plausible mechanism, but inconsistent human evidence: a suggestive association in one meta-analysis and none in a 2026 analysis with up to 36 years of follow-up.
- TMAO and choline: inconsistent human results: positive associations in single studies, but no overall association in a meta-analysis, and conflicting results for dietary choline.
- Endotoxin: short meal studies in people that measured blood markers, not cancer, plus cell studies.
- Saturated fat and CD36: striking animal results, but no human evidence yet.
- Branched-chain amino acids: no association in large human cohorts.
- Methionine: animal and metabolic evidence, but human cohorts point the other way, with the highest intakes linked to lower colon cancer risk.
Foods linked to lower cancer risk, graded by evidence
The protective side of the evidence is mostly observational too. Some associations are consistent and show a dose-response pattern; others shrink when you look only at the stronger study designs.
Does fiber lower colorectal cancer risk?
The evidence says yes, with whole grains and cereal fiber leading. A 2011 meta-analysis of 25 prospective studies by Aune and colleagues found that each 10 g a day of total dietary fiber was associated with a 10% lower risk of colorectal cancer (0.90, 0.86 to 0.94), as was each 10 g of cereal fiber (0.90, 0.83 to 0.97) [14]. Three daily servings of whole grains carried a 17% lower risk (0.83, 0.78 to 0.89). Fiber from fruit (0.93, 0.82 to 1.05) and from vegetables (0.98, 0.91 to 1.06) showed no significant association on its own. In the Million Women Study, though, the whole-grain association weakened as more lifestyle factors were adjusted for, and the authors suggest it may be partly due to residual confounding [7]. For food-by-food grades on colorectal cancer, from whole grains and beans to coffee, see our colon cancer post.
Does fiber lower breast cancer risk?
Modestly. A 2020 meta-analysis of 17 cohort studies, 2 nested case-control studies, and 1 clinical trial found the highest fiber intakes associated with an 8% lower risk of breast cancer (0.92, 0.88 to 0.95), and soluble fiber with a 10% lower risk (0.90, 0.84 to 0.96) [58].
What happens in the colon when people switch to a high-fiber diet?
A two-week experiment gives a glimpse. Middle-aged African Americans were fed a high-fiber, low-fat African-style diet, and rural South Africans a high-fat, low-fiber Western-style diet, under close supervision [59]. In colon biopsies, the switch “decreased proliferative rates in African Americans to levels lower than those of Africans at baseline,” while rates rose in the Africans. The African Americans also showed more butyrate production and less secondary bile acid production. The study enrolled 20 African Americans, lasted two weeks, and measured biomarkers of cancer risk rather than cancer.
Do fruits and vegetables lower cancer risk?
Yes, though less than you might expect. A 2017 dose-response meta-analysis of 95 prospective studies found each 200 g a day of fruits and vegetables associated with a 3% lower risk of total cancer (0.97, 0.95 to 0.99), with reductions continuing up to about 600 g a day [15]. Among specific types, cruciferous and green-yellow vegetables were linked to lower total cancer risk. The same analysis found larger associations for all-cause mortality (0.90, 0.87 to 0.93), so the case for fruits and vegetables does not rest on cancer alone.
Does eating broccoli and other cruciferous vegetables lower cancer risk?
Possibly, for some cancers. A 2022 umbrella review of 41 systematic reviews and meta-analyses covering 303 observational studies found no association between cruciferous vegetables and any health outcome that reached the level of “convincing or highly suggestive evidence” [16]. Six associations were rated “suggestive,” including lower risk of gastric, lung, and endometrial cancer.
Sulforaphane, a compound supplied by broccoli sprouts, has been tested in people, though not against cancer. In a 12-week randomized trial of 291 residents of Qidong, China, a broccoli sprout beverage increased urinary excretion of detoxified benzene by 61% and of detoxified acrolein by 23% compared with placebo [60]. A follow-up trial in 170 people found that only the full-strength beverage worked: the half and one-fifth doses “were not significantly different from placebo controls” [61]. These were short trials of detoxification markers, and the doses came from a broccoli sprout beverage, not from ordinary servings of broccoli.
Do mushrooms lower cancer risk?
The evidence is weaker than the headline. A 2021 meta-analysis of 17 observational studies found the highest mushroom intakes associated with a 34% lower total cancer risk (0.66, 0.55 to 0.78) [17]. But 11 of the 17 were case-control studies, which the authors note carry a potential for recall and selection bias. In the six cohort studies, the association was much smaller (0.89, 0.82 to 0.97), and the lower breast cancer risk appeared only in case-control studies.
A small trial offers a biological hint. Fifty-two healthy adults aged 21 to 41 ate 5 or 10 g of shiitake mushrooms a day for four weeks, and their immune cells showed greater proliferation in lab tests afterward (gamma-delta T cells 60% more, NK-T cells twice as much) [62]. Both groups ate mushrooms, so there was no mushroom-free comparison, and the study was funded by the U.S. Mushroom Council and the Australian Mushroom Growers Association.
Is soy safe after breast cancer?
The concern is that soy isoflavones have estrogen-like properties. In the After Breast Cancer Pooling Project, 9,514 breast cancer survivors from two U.S. cohorts and one Chinese cohort were followed for a mean of 7.4 years [63]. Women consuming 10 mg or more of isoflavones a day after diagnosis, compared with less than 4 mg, had a 25% lower risk of recurrence (0.75, 0.61 to 0.92) and a non-significant lower risk of death from breast cancer (0.83, 0.64 to 1.07). In the Adventist Health Study-2, among women without cancer at the start, “no clear associations were found between soy products and breast cancer, independently of dairy” [33]. The observational data show no sign of harm from soy foods, and some sign of benefit.
Should you take supplements instead of eating these foods?
No. As the section on evidence showed, beta carotene supplements raised lung cancer incidence in smokers [23], and vitamin E supplements raised prostate cancer risk [24]. The protective associations in this section are for foods and dietary patterns, not for isolated compounds.
Which protective foods have the strongest evidence?
Our reading, from strongest to weakest:
- Fiber and whole grains: the strongest. A consistent dose-response association with lower colorectal cancer risk in prospective studies, a modest association for breast cancer, and short-term biomarker changes in the colon when people switch diets, although residual confounding may explain part of the whole-grain association.
- Fruits and vegetables overall: a real but small association with lower total cancer risk.
- Soy: no sign of harm after breast cancer, and lower recurrence in pooled cohorts.
- Cruciferous vegetables: suggestive evidence for a few cancers; human trials of sulforaphane measure detoxification markers only.
- Mushrooms: weak. The association is driven largely by case-control studies.
Cellular senescence, an early-stage hypothesis
What is cellular senescence, and how might diet affect it?
Cellular senescence is “a process in which a cell exits the cell cycle in response to stressors,” and it is one of the hallmarks of aging [64]. For cancer it cuts both ways. Senescence helps protect against tumors early in life, but the review links the inflammatory signals released by senescent-like immune cells to “chronic inflammation, tumor development, and tissue dysfunction.”
A 2024 review in Cell Metabolism concludes that “a growing body of literature suggests that senescence is responsive to diet, both to calories and specific dietary macronutrients” [64]. The protein studies it describes are in mice: protein intake was associated with senescence markers in mouse liver, and protein restriction reduced inflammatory markers in the livers of old male mice. The human evidence it describes includes calorie restriction: in people who restricted calories by 30% for 10 years, the lining of the lower colon showed lower expression of senescence and inflammation markers.
How much weight should the senescence evidence carry?
Not much yet, and the review’s authors say so. They write that “there is still limited evidence for dietary interventions affecting senescence in humans,” that many of the papers they cite “use cell culture systems or are conducted using model organisms,” and that “interventions that delay or reduce senescence in one tissue could actually stimulate senescence in another” [64]. Two of the three authors have commercial interests in the field: one has received funding from, and advises, a company developing mTOR inhibitors, and another co-founded a company developing senotherapeutics. For now, senescence is a hypothesis about how diet might affect aging and cancer, not an established pathway.
After a diagnosis: what diet can and cannot do
People living with cancer ask the most practical question: does changing what I eat now change what happens next? The randomized trials covered earlier in this post found no effect from vegetable-rich diets [19] [20], and benefits only in trials of low-fat diets, one of which also produced a difference in body weight [21] [22]. The observational studies below fill in detail, with the usual caveats. This section covers breast and prostate cancer; our colon cancer post covers diet after a colorectal cancer diagnosis. It is also about recurrence and progression alongside standard treatment, not about diet as a treatment in itself; Does Sugar Feed Cancer? What the Evidence Actually Shows takes up the claim that diet can treat cancer, including the idea that cutting carbohydrates starves tumors.
What does the strongest evidence say about diet after breast cancer?
The World Cancer Research Fund’s Global Cancer Update Programme reviewed randomized trials and longitudinal studies of breast cancer survivors, and an independent expert panel graded the evidence [18]. Higher body mass index after diagnosis was graded “strong” (likelihood of causality: probable): each 5 kg/m² was associated with 7% higher all-cause mortality (1.07, 1.05 to 1.10) and 10% higher breast cancer mortality (1.10, 1.06 to 1.14). Recreational physical activity was graded limited-suggestive, with each 10 MET-hours a week associated with 15% lower all-cause mortality (0.85, 0.78 to 0.92). On diet, “no conclusions could be reached except for healthy dietary patterns, isoflavone and dietary fibre intake and serum 25(OH)D concentrations that were graded with limited-suggestive evidence for lower risk.”
The underlying review describes its dietary results as, “with few exceptions, … generally null” [65]. Fiber was one exception: each 10 g a day was associated with 13% lower all-cause mortality (0.87, 0.80 to 0.94), on evidence the review calls sparse but suggestive. Total fat was not associated with breast cancer death (1.00 per 10 g a day, 0.96 to 1.04), and only one of three studies of saturated fat found higher risk.
Do vegetables after breast cancer lower the risk of recurrence?
The evidence conflicts. In a secondary analysis of 3,080 women in the WHEL trial, among women taking tamoxifen, higher cruciferous vegetable intake was associated with lower recurrence (hazard ratio 0.65, 0.47 to 0.89), as was higher total vegetable intake (0.56, 0.41 to 0.77) [66]. A larger pooled analysis of 11,390 survivors from four cohorts in the After Breast Cancer Pooling Project, followed for a median of 9 years, found that cruciferous vegetable intake “was not associated with breast cancer outcomes” [67]. The pooled analysis included WHEL among its four cohorts, so its null result carries more weight than a subgroup finding from one of them.
Among 1,490 women in the comparison group of the WHEL trial, only the combination of five or more daily servings of vegetables and fruit with physical activity equivalent to walking 30 minutes six days a week “was associated with a significant survival advantage” (hazard ratio 0.56, 0.31 to 0.98) [68]. Neither habit alone reached significance in the multivariate model.
Soy, covered in the section on protective foods, was associated with lower recurrence and showed no sign of harm [63].
Does dairy affect survival after breast cancer?
High-fat dairy may. In the Life After Cancer Epidemiology study of 1,893 women followed for a median of 11.8 years, women eating one or more servings a day of high-fat dairy had higher breast cancer mortality than those eating less than half a serving (hazard ratio 1.49, 1.00 to 2.24), while low-fat dairy “was unrelated to recurrence or survival” [69]. The confidence interval touches 1.00, so this is a borderline finding from a single cohort.
Does following cancer-prevention guidelines help survivors live longer?
In 2,017 older women with a history of cancer in the Iowa Women’s Health Study, those who best followed the World Cancer Research Fund guidelines had 33% lower all-cause mortality than those who followed them least (hazard ratio 0.67, 0.50 to 0.94) [70]. But “adherence to the physical activity recommendation was the strongest factor associated with lower risk of each mortality outcome.” Following the dietary recommendations was associated with lower all-cause mortality but not with cancer-specific mortality, and adjusting for physical activity weakened even that association. Following the body weight recommendation was associated with higher all-cause mortality in these older women.
Can a plant-based lifestyle slow prostate cancer?
One randomized trial suggests it might, but it tested much more than diet. In 2005, Ornish and colleagues randomized 93 men with early prostate cancer who had chosen not to have conventional treatment [71]. The intervention was “a vegan diet supplemented with soy (1 daily serving of tofu plus 58 gm of a fortified soy protein powdered beverage), fish oil (3 gm daily), vitamin E (400 IU daily), selenium (200 mcg daily) and vitamin C (2 gm daily),” plus walking, daily stress management, and a weekly support group. After a year, PSA fell 4% in the intervention group and rose 6% in the control group, and none of the intervention group but six control patients went on to conventional treatment. Because the program combined diet, supplements, exercise, and stress management, the trial cannot say which part mattered. Its 400 IU daily dose of vitamin E is the same dose that raised prostate cancer risk in SELECT [24]. The lead author disclosed a financial relationship with the publishers Random House and HarperCollins.
A smaller randomized pilot tested a diet “emphasizing plant-based foods and fish” in 36 men whose PSA was rising after treatment, with cooking classes and mindfulness practice, against a wait-list control [72]. There was “no significant difference … in the log PSA slope between the two groups,” though PSA doubling time lengthened in the intervention group. The MEAL trial, described earlier, found that counseling to eat more vegetables did not slow progression [20].
Which foods eaten after a prostate cancer diagnosis are linked to progression?
In the CaPSURE registry, among 1,294 men with prostate cancer, intakes of processed and unprocessed red meat, fish, and skinless poultry after diagnosis were not associated with recurrence or progression [73]. Eggs (hazard ratio 2.02, 1.10 to 3.72) and poultry with skin (2.26, 1.36 to 3.76) were associated with roughly doubled risk. Among 1,560 men in the same registry, the highest quarter of cruciferous vegetable intake was associated with 59% lower risk of progression (0.41, 0.22 to 0.76), based on 134 events, only four of them prostate cancer deaths [74]. The CaPSURE registry behind both analyses has received pharmaceutical industry support, including founding sponsorship from TAP Pharmaceutical Products and an educational gift from Abbott Laboratories.
A 2024 CaPSURE analysis of 2,062 men looked at whole dietary patterns [75]. Men in the highest fifth of an overall plant-based diet index had a 47% lower risk of progression than those in the lowest fifth (0.53, 0.37 to 0.74). The middle fifths did not show a steady trend (1.02, 0.74, and 0.98), and a healthful plant-based index was not associated with progression overall (0.81, 0.54 to 1.20), though it was among men with higher-grade tumors (0.45, 0.25 to 0.81). Most of the 190 progression events were rises in PSA; 13 were prostate cancer deaths. And as the authors note, “even participants in the highest quintile of PDI consumed meat and dairy products; therefore we are unable to assess the associations of fully plant-based diets (eg, vegan, vegetarian).”
Does dairy affect prostate cancer survival?
The consistent signal is for whole milk, not dairy as a whole. In the Health Professionals Follow-up Study, among 3,918 men with localized prostate cancer, total milk and dairy after diagnosis were not associated with lethal prostate cancer, but men with the highest whole milk intake had more than twice the risk of progression (hazard ratio 2.15, 1.28 to 3.60), while the highest intake of low-fat dairy was associated with lower risk (0.62, 0.40 to 0.95) [76]. In the Physicians’ Health Study, whole milk measured years before diagnosis was associated with higher prostate cancer mortality among men who developed the disease (2.17, 1.34 to 3.51) [32]. A separate analysis of 926 men from the same study found three or more daily servings of total dairy after diagnosis associated with higher prostate cancer mortality, but the confidence interval included no effect (2.41, 0.96 to 6.02), based on 56 prostate cancer deaths [77].
In a Swedish cohort of 525 men, high-fat milk was not associated with prostate cancer death overall (1.28, 0.78 to 2.10); the six-fold figure from this study (6.10, 2.14 to 17.37) comes from the 230 men with localized disease [78]. In CaPSURE, among 1,334 men, drinking whole milk more than four times a week was associated with higher recurrence (1.73, 1.00 to 2.98), but only 110 men drank that much, and the association appeared only in men with a BMI of 27 or more, based on 14 events [79]. The authors note that “it is possible that this statistically significant association was due to chance.” Low-fat milk and other dairy foods were not associated with recurrence.
What does the after-diagnosis evidence add up to?
Our reading:
- Body fatness and physical activity have the strongest survivorship evidence in breast cancer, ahead of any specific food, although in one study of older women, meeting the body weight recommendation was linked to higher mortality.
- Diet changes tested in randomized trials have mostly not reduced recurrence on their own; the two positive breast cancer results came from low-fat diets, and in one of them the intervention group also weighed less.
- Whole milk and high-fat dairy show a repeated but fragile association with worse outcomes, mainly in prostate cancer, while low-fat dairy shows none or the reverse.
- Soy looks safe after breast cancer, with lower recurrence in pooled cohorts.
- Plant-based eating patterns are linked to slower prostate cancer progression, but only at the top of the range and in men who still ate meat and dairy.
- Cruciferous vegetables give conflicting results after breast cancer and a single positive registry result after prostate cancer.
What this evidence does not show
Has a vegan diet been tested against cancer in a randomized trial?
Not in any trial covered in this post. The randomized diet trials with cancer outcomes tested vegetable-rich or low-fat diets, not vegan ones [19] [20] [21] [22]. The one trial that included a vegan diet combined it with supplements, exercise, stress management, and a support group, and measured PSA and treatment decisions in 93 men [71]. Every vegetarian and vegan cancer estimate in this post comes from observational studies.
How reliable are the diet measurements behind these studies?
Less reliable than the precise-looking numbers suggest. Many of the cohorts measured diet with food frequency questionnaires, often only once. In the methionine meta-analysis, all but two of the included studies used a single baseline measurement of diet [46]. In the Physicians’ Health Study, dairy intake was measured once, an average of 14 years before diagnosis [32], and the NHANES protein study relied on a single 24-hour recall [38]. As the methionine authors note, this kind of misclassification in cohort studies “leads to null rather than to spurious associations,” so it tends to hide real effects rather than create false ones, but it also makes the size of any estimate uncertain [46].
Do plant-based diet studies actually study plant-based eaters?
Often not. In the 2024 prostate cancer study, “even participants in the highest quintile of PDI consumed meat and dairy products” [75]. In a related study of 47,239 men, “fewer than 1% of participants followed a strict vegetarian or vegan diet” [80]. Results for plant-based diet indices describe eating more plants and fewer animal foods within an omnivorous population. The comparison groups can also be low in meat themselves, as in the Adventist breast cancer analysis above, which blunts the contrast with vegetarian diets [28]. Vegan-specific estimates rest on far fewer people: in EPIC-Oxford, the vegan estimate for all cancers combined was based on 105 cancers [30]. Even the largest pooled analysis had 8,849 vegans, 0.5% of its participants, and its vegan colorectal estimate rested on 93 cases [2].
Are the people in these studies different from everyone else?
In some ways, yes. Adventists have “a very low percentage of alcohol consumption or cigarette smoking” [26], and in EPIC-Oxford, cancer rates for all participants, meat eaters included, were 72% of national rates [29]. Comparing people within such cohorts reduces some confounding, but the results may not carry over unchanged to the general population. Several of the prostate cancer cohorts were also mostly white: 95% of the men in the 2024 CaPSURE analysis identified as White [75].
Can statistical adjustment remove confounding completely?
No, and some authors show it directly. In the Million Women Study, adjusting step by step for lifestyle factors barely changed the associations for alcohol, calcium, and dairy milk, but weakened those for fruit, whole grains, and breakfast cereal, which the authors suggest “may have been at least partly due to residual confounding with lifestyle factors” [7]. The commentary on the Adventist dairy and breast cancer study raised the same concern about women who drink little dairy milk [34].
Is “no evidence of benefit” the same as “evidence of no benefit”?
No. The MEAL trial found no effect of a vegetable intervention on prostate cancer progression, but its authors note it “may have been underpowered” [20]. In the 2024 plant-based prostate cancer study, the confidence intervals for prostate cancer death “were too wide to draw meaningful conclusions from point estimates” [75]. Null results from small or short studies leave questions open; they do not close them.
Who funds this research?
Funding and conflicts run in every direction, which is why each reference in this post notes them. Meat industry observers attended the IARC meeting [4]. One author of the 2019 meat review initially left a research grant off his disclosure [5]. A shiitake trial was funded by mushroom growers [62], and an endotoxin study was supported by the Florida Department of Citrus [55]. An author of the UCLA serum studies consulted for the Pritikin center [44], the senior author of the protein study holds equity in a company that develops medical food [38], and the lead author of the EPIC-Oxford analyses disclosed membership in the Vegan Society [30]. None of these ties by itself invalidates a result, but each is worth knowing.
What should readers take from all this?
The evidence in this post supports eating mostly plants, with plenty of fiber-rich whole grains, vegetables, and fruit; limiting processed and red meat; keeping body fat in a healthy range and staying active, which matter most after a diagnosis; getting enough calcium, which on a vegan diet means choosing calcium-rich and fortified foods or a supplement; and not relying on high-dose antioxidant supplements. A well-planned vegan diet, built on a variety of legumes, greens, whole grains, fruits, vegetables, mushrooms, nuts, seeds, seaweed such as nori, herbs, and fortified plant milks, can meet all of these. On current evidence it lowers cancer risk moderately, not decisively, and nothing in this evidence suggests diet can replace screening or medical treatment.
Related reading
Our colon cancer post puts the 2026 pooled analysis’s vegan colorectal estimate in context, explains how to get enough calcium on a vegan diet, and grades individual foods for colorectal cancer risk. Does Sugar Feed Cancer? What the Evidence Actually Shows examines the claim that cutting carbohydrates starves cancer cells, and explains why no diet, plant-based or ketogenic, has been shown to treat cancer once it is diagnosed.
References
- Aune, D., Schlesinger, S., and Sobiecki, J. G. (2026). Vegetarian and vegan diets and cancer incidence: a systematic review and meta-analysis of prospective studies. European Journal of Epidemiology, 41, 531–545. (Seven prospective studies; total cancer RR 0.87 for vegetarians and 0.77 for vegans, 0.92 and 0.82 after BMI adjustment.) Funded by Raagholtstiftelsen; no competing interests declared. doi.org
- Dunneram, Y., Lee, J. Y., Watling, C. Z., et al. (2026). Vegetarian diets and cancer risk: pooled analysis of 1.8 million women and men in nine prospective studies on three continents. British Journal of Cancer, 134(8), 1218–1229. (Nine cohorts; 8,849 vegans; vegan colorectal cancer HR 1.40 from 93 cases.) Funded by the World Cancer Research Fund, with support for the Oxford authors from Cancer Research UK, the Medical Research Council, Wellcome, and UKRI; no competing interests declared. A published correction (April 2026; DOI 10.1038/s41416-026-03446-6) amended two passages of the text and changed no figures. doi.org
- Wang, Y., Liu, B., Han, H., et al. (2023). Associations between plant-based dietary patterns and risks of type 2 diabetes, cardiovascular disease, cancer, and mortality: a systematic review and meta-analysis. Nutrition Journal, 22, 46. (76 publications, 2,230,443 participants; cancer RR 0.88 for overall plant-based patterns and 0.86 for healthy ones; unhealthy patterns 7% higher. Figures reflect the published correction (Nutrition Journal, 2024, 23:6), which fixed a data-extraction error in some cancer results; the PubMed abstract still shows the original overall cancer estimate.) Funded by the U.S. National Institutes of Health; no competing interests declared. doi.org
- Bouvard, V., Loomis, D., Guyton, K. Z., et al., on behalf of the International Agency for Research on Cancer Monograph Working Group (2015). Carcinogenicity of consumption of red and processed meat. The Lancet Oncology, published online October 26, 2015. (Processed meat Group 1, red meat Group 2A.) Working Group members declared no competing interests apart from one member’s involvement in a World Cancer Research Fund project; meeting observers included representatives of the Beef Checkoff and the North American Meat Institute. doi.org
- Han, M. A., Zeraatkar, D., Guyatt, G. H., et al. (2019). Reduction of red and processed meat intake and cancer mortality and incidence: a systematic review and meta-analysis of cohort studies. Annals of Internal Medicine, 171, 711–720. (118 articles, 56 cohorts; low to very low certainty of evidence under GRADE.) No external funding. A published correction added one author’s previously undisclosed grant from Texas A&M AgriLife Research and earlier funding from the International Life Sciences Institute (North America). doi.org
- Orlich, M. J., Mashchak, A. D., Jaceldo-Siegl, K., et al. (2022). Dairy foods, calcium intakes, and risk of incident prostate cancer in Adventist Health Study–2. American Journal of Clinical Nutrition, 116, 314–324. (28,737 men; 90th versus 10th percentile of dairy intake HR 1.27.) Supported by the National Cancer Institute, the World Cancer Research Fund, and the Ardmore Institute of Health; no conflicts of interest reported. doi.org
- Papier, K., Bradbury, K. E., Balkwill, A., et al. (2025). Diet-wide analyses for risk of colorectal cancer: prospective study of 12,251 incident cases among 542,778 women in the UK. Nature Communications, 16, 375. (Million Women Study; calcium RR 0.83 per 300 mg a day, dairy milk 0.86 per 200 g a day.) Funded by Cancer Research UK and the UK Medical Research Council. One author reported consulting for AbbVie; the others declared no competing interests. doi.org
- Keum, N., Lee, D. H., Marchand, N., et al. (2015). Egg intake and cancers of the breast, ovary and prostate: a dose-response meta-analysis of prospective observational studies. British Journal of Nutrition, 114(7), 1099–1107. (Breast cancer RR 1.05 per five eggs a week; fatal prostate cancer 1.47 with evidence of small-study effects.) No financial support declared; no conflicts of interest declared. doi.org
- Watts, E. L., Perez-Cornago, A., Fensom, G. K., et al. (2023). Circulating insulin-like growth factors and risks of overall, aggressive and early-onset prostate cancer: a collaborative analysis of 20 prospective studies and Mendelian randomization analysis. International Journal of Epidemiology, 52(1), 71–86. (Cited here for its explanation of Mendelian randomization.) Funded by Cancer Research UK; no conflicts of interest declared. A 2025 correction fixed a typographic error in an author’s name; the findings were unchanged. doi.org
- Murphy, N., Knuppel, A., Papadimitriou, N., et al. (2020). Insulin-like growth factor-1, insulin-like growth factor-binding protein-3, and breast cancer risk: observational and Mendelian randomization analyses with approximately 430 000 women. Annals of Oncology, 31, 641–649. (UK Biobank, 206,263 women; HR 1.11 per 5 nmol/L; Mendelian randomization OR 1.05.) No funding and no conflicts of interest declared. doi.org
- Allen, N. E., Appleby, P. N., Davey, G. K., et al. (2002). The associations of diet with serum insulin-like growth factor I and its main binding proteins in 292 women meat-eaters, vegetarians, and vegans. Cancer Epidemiology, Biomarkers & Prevention, 11, 1441–1448. (Cross-sectional; IGF-1 13% lower in vegans.) Supported by Cancer Research UK and the Europe Against Cancer Programme of the European Commission. PubMed
- Bastide, N. M., Pierre, F. H. F., and Corpet, D. E. (2011). Heme iron from meat and risk of colorectal cancer: a meta-analysis and a review of the mechanisms involved. Cancer Prevention Research, 4(2), 177–184. (Meta-analysis of prospective cohorts, 566,607 people, plus a review of animal and mechanistic studies.) No potential conflicts of interest disclosed. doi.org
- Shi, W., Wang, P., and Giovannucci, E. L. (2026). Heme iron intake and the risk of colorectal cancer: a prospective analysis of the modifying role of calcium. Cancer Causes & Control, 37(9). (Nurses’ Health Study I and II and Health Professionals Follow-up Study; 211,851 participants, 2,983 colorectal cancers; total heme iron HR 1.03 and calcium HR 0.77, highest versus lowest fifth.) Supported by the U.S. National Institutes of Health and the American Cancer Society; no competing interests declared. doi.org
- Aune, D., Chan, D. S. M., Lau, R., et al. (2011). Dietary fibre, whole grains, and risk of colorectal cancer: systematic review and dose-response meta-analysis of prospective studies. BMJ, 343, d6617. (25 prospective studies; total fiber RR 0.90 per 10 g a day.) Funded by the World Cancer Research Fund as part of its Continuous Update Project; the authors declared no financial relationships with organizations that might have an interest in the work. doi.org
- Aune, D., Giovannucci, E., Boffetta, P., et al. (2017). Fruit and vegetable intake and the risk of cardiovascular disease, total cancer and all-cause mortality: a systematic review and dose-response meta-analysis of prospective studies. International Journal of Epidemiology, 46(3), 1029–1056. (95 studies; total cancer RR 0.97 per 200 g a day.) No conflicts of interest declared. doi.org
- Li, Y.-Z., Yang, Z.-Y., Gong, T.-T., et al. (2022). Cruciferous vegetable consumption and multiple health outcomes: an umbrella review of 41 systematic reviews and meta-analyses of 303 observational studies. Food & Function, 13, 4247–4259. Supported by the National Natural Science Foundation of China; no conflicts declared. doi.org
- Ba, D. M., Ssentongo, P., Beelman, R. B., Muscat, J., Gao, X., and Richie, J. P., Jr. (2021). Higher mushroom consumption is associated with lower risk of cancer: a systematic review and meta-analysis of observational studies. Advances in Nutrition, 12, 1691–1704. (17 studies, 11 of them case-control.) No funding received; no conflicts of interest. doi.org
- Tsilidis, K. K., Cariolou, M., Becerra-Tomás, N., et al. (2023). Postdiagnosis body fatness, recreational physical activity, dietary factors and breast cancer prognosis: Global Cancer Update Programme (CUP Global) summary of evidence grading. International Journal of Cancer, 152, 635–644. Funded by the World Cancer Research Fund network; no conflicts of interest declared. doi.org
- Pierce, J. P., Natarajan, L., Caan, B. J., et al. (2007). Influence of a diet very high in vegetables, fruit, and fiber and low in fat on prognosis following treatment for breast cancer: the Women’s Healthy Eating and Living (WHEL) randomized trial. JAMA, 298(3), 289–298. (Randomized trial, 3,088 women; invasive breast cancer events HR 0.96, 0.80 to 1.14.) Supported by the Walton Family Foundation and the National Cancer Institute; no financial disclosures reported. doi.org
- Parsons, J. K., Zahrieh, D., Mohler, J. L., et al. (2020). Effect of a behavioral intervention to increase vegetable consumption on cancer progression among men with early-stage prostate cancer: the MEAL randomized clinical trial. JAMA, 323(2), 140–148. (Randomized trial, 478 men; progression HR 0.97, 0.76 to 1.25.) The lead author reported grants from the National Cancer Institute, personal fees from Endocare and INSIGHTEC outside the submitted work, and stock in Pfizer and Johnson & Johnson. doi.org
- Chlebowski, R. T., Blackburn, G. L., Thomson, C. A., et al. (2006). Dietary fat reduction and breast cancer outcome: interim efficacy results from the Women’s Intervention Nutrition Study. Journal of the National Cancer Institute, 98, 1767–1776. (Randomized trial, 2,437 women; interim relapse HR 0.76, 0.60 to 0.98.) Supported by the National Cancer Institute, with supplemental projects funded by the Breast Cancer Research Foundation and the American Institute for Cancer Research. doi.org
- Chlebowski, R. T., Aragaki, A. K., Anderson, G. L., et al. (2018). Association of low-fat dietary pattern with breast cancer overall survival: a secondary analysis of the Women’s Health Initiative randomized clinical trial. JAMA Oncology, published online May 24, 2018. (Secondary analysis of 1,764 women diagnosed with breast cancer during the trial; overall survival HR 0.78, 0.65 to 0.94.) The Women’s Health Initiative is funded by the National Heart, Lung, and Blood Institute; the lead author reported consulting for Novartis, AstraZeneca, and Genentech. A published correction fixed the spelling of an author’s name; the findings were unchanged. doi.org
- The Alpha-Tocopherol, Beta Carotene Cancer Prevention Study Group (1994). The effect of vitamin E and beta carotene on the incidence of lung cancer and other cancers in male smokers. New England Journal of Medicine, 330, 1029–1035. (Randomized trial, 29,133 male smokers; lung cancer incidence 18% higher with beta carotene.) Supported by a contract with the National Cancer Institute. doi.org
- Klein, E. A., Thompson, I. M., Jr., Tangen, C. M., et al. (2011). Vitamin E and the risk of prostate cancer: the Selenium and Vitamin E Cancer Prevention Trial (SELECT). JAMA, 306(14), 1549–1556. (Randomized trial, 35,533 men; vitamin E HR 1.17, 99% CI 1.004 to 1.36.) Supported by the National Cancer Institute and the National Center for Complementary and Alternative Medicine; study agents provided by Perrigo, Sabinsa, Tishcon, and DSM Nutritional Products. Several authors reported industry relationships, including vitamin and placebo pills supplied to one author’s institution by Wyeth (now Pfizer), pharmaceutical consulting fees, and a co-founder role at Cancer Prevention Pharmaceuticals. doi.org
- Bjelakovic, G., Nikolova, D., Gluud, L. L., Simonetti, R. G., and Gluud, C. (2012). Antioxidant supplements for prevention of mortality in healthy participants and patients with various diseases. Cochrane Database of Systematic Reviews, Issue 3, CD007176. (78 randomized trials, 296,707 participants.) Declarations of interest: none known. doi.org
- Tantamango-Bartley, Y., Jaceldo-Siegl, K., Fan, J., and Fraser, G. (2013). Vegetarian diets and the incidence of cancer in a low-risk population. Cancer Epidemiology, Biomarkers & Prevention, 22(2), 286–294. (Adventist Health Study-2, 69,120 participants; vegetarians HR 0.92, vegans 0.84.) Supported by a U.S. National Institutes of Health grant; no conflicts of interest declared. doi.org
- Tantamango-Bartley, Y., Knutsen, S. F., Knutsen, R., et al. (2016). Are strict vegetarians protected against prostate cancer? American Journal of Clinical Nutrition, 103, 153–160. (Adventist Health Study-2, 26,346 men; vegans HR 0.65.) Supported by U.S. National Institutes of Health, U.S. Department of Agriculture, and World Cancer Research Fund International grants to the Adventist Health Study-2; no conflicts of interest declared. doi.org
- Penniecook-Sawyers, J. A., Jaceldo-Siegl, K., Fan, J., et al. (2016). Vegetarian dietary patterns and the risk of breast cancer in a low-risk population. British Journal of Nutrition, 115(10), 1790–1797. (Adventist Health Study-2, 50,404 women; vegetarians HR 0.97, vegans 0.78, 0.58 to 1.05.) Funded by the U.S. National Institutes of Health; no conflicts of interest declared. doi.org
- Key, T. J., Appleby, P. N., Spencer, E. A., Travis, R. C., Roddam, A. W., and Allen, N. E. (2009). Cancer incidence in vegetarians: results from the European Prospective Investigation into Cancer and Nutrition (EPIC-Oxford). American Journal of Clinical Nutrition, 89(supplement). (63,550 participants; colorectal cancer IRR 1.39 for vegetarians vs. meat eaters.) Supported by Cancer Research UK and the Medical Research Council. The first author is a member of the Vegan Society. doi.org
- Key, T. J., Appleby, P. N., Crowe, F. L., Bradbury, K. E., Schmidt, J. A., and Travis, R. C. (2014). Cancer in British vegetarians: updated analyses of 4998 incident cancers in a cohort of 32,491 meat eaters, 8612 fish eaters, 18,298 vegetarians, and 2246 vegans. American Journal of Clinical Nutrition, 100(supplement), 378S–385S. (Pooled EPIC-Oxford and Oxford Vegetarian Study; all cancers RR 0.88 for vegetarians, 0.81 for vegans.) Funded by Cancer Research UK. The first author is a member of the Vegetarian Society and the Vegan Society. doi.org
- Farvid, M. S., Sidahmed, E., Spence, N. D., et al. (2021). Consumption of red meat and processed meat and cancer incidence: a systematic review and meta-analysis of prospective studies. European Journal of Epidemiology, 36, 937–951. (148 articles; highest versus lowest intake; colorectal cancer RR 1.18 for processed meat and 1.10 for red meat.) No funding source; no conflicts of interest declared. doi.org
- Song, Y., Chavarro, J. E., Cao, Y., et al. (2013). Whole milk intake is associated with prostate cancer-specific mortality among U.S. male physicians. Journal of Nutrition, 143, 189–196. (Physicians’ Health Study, 21,660 men; total dairy HR 1.12, 0.93 to 1.35; whole milk and fatal prostate cancer 1.49, 0.97 to 2.28.) Supported by the U.S. National Institutes of Health and the National Cancer Institute of Canada; no conflicts of interest. doi.org
- Fraser, G. E., Jaceldo-Siegl, K., Orlich, M., Mashchak, A., Sirirat, R., and Knutsen, S. (2020). Dairy, soy, and risk of breast cancer: those confounded milks. International Journal of Epidemiology, 49(5), 1526–1537. (52,795 women; dairy milk HR 1.50, 1.22 to 1.84.) Supported by the National Cancer Institute and the World Cancer Research Fund. The authors state that three of them are largely vegetarian and that all use dairy products. doi.org
- Perez-Cornago, A. (2020). Commentary: Dairy milk intake and breast cancer risk: does an association exist, and what might be the culprit? International Journal of Epidemiology, 49(5), 1537–1539. Supported by Cancer Research UK and the World Cancer Research Fund; no conflict of interest declared. doi.org
- Richman, E. L., Kenfield, S. A., Stampfer, M. J., Giovannucci, E. L., and Chan, J. M. (2011). Egg, red meat, and poultry intake and risk of lethal prostate cancer in the prostate-specific antigen-era: incidence and survival. Cancer Prevention Research, 4(12), 2110–2121. (Health Professionals Follow-up Study, 27,607 men; 2.5 or more eggs a week HR 1.81.) Supported by U.S. National Institutes of Health grants; no conflicts disclosed. doi.org
- Ngo, T. H., Barnard, R. J., Tymchuk, C. N., Cohen, P., and Aronson, W. J. (2002). Effect of diet and exercise on serum insulin, IGF-I, and IGFBP-1 levels and growth of LNCaP cells in vitro (United States). Cancer Causes and Control, 13, 929–935. (Serum from 14 men before and after an 11-day Pritikin diet and exercise program, tested on prostate cancer cells in culture.) Supported by the Cancer Research Fund, the Cancer Research Foundation of America, U.S. National Institutes of Health grants, and private foundations. doi.org
- Harrison, S., Lennon, R., Holly, J., et al. (2017). Does milk intake promote prostate cancer initiation or progression via effects on insulin-like growth factors (IGFs)? A systematic review and meta-analysis. Cancer Causes & Control, 28(6), 497–528. Funded by the World Cancer Research Fund. doi.org
- Levine, M. E., Suarez, J. A., Brandhorst, S., et al. (2014). Low protein intake is associated with a major reduction in IGF-1, cancer, and overall mortality in the 65 and younger but not older population. Cell Metabolism, 19(3), 407–417. (NHANES III, 6,381 adults aged 50 and over; diet from a single 24-hour recall.) The senior author has an equity interest in L-Nutra. doi.org
- Steuerman, R., Shevah, O., and Laron, Z. (2011). Congenital IGF1 deficiency tends to confer protection against post-natal development of malignancies. European Journal of Endocrinology, 164, 485–489. (Retrospective worldwide survey of 538 patients with congenital IGF-1 deficiency, 230 with Laron syndrome.) No conflicts of interest declared. doi.org
- Guevara-Aguirre, J., Balasubramanian, P., Guevara-Aguirre, M., et al. (2011). Growth hormone receptor deficiency is associated with a major reduction in pro-aging signaling, cancer, and diabetes in humans. Science Translational Medicine, 3(70), 70ra13. (Ecuadorian cohort monitored for 22 years.) Two authors have filed patents related to the study’s subject. doi.org
- Ngo, T. H., Barnard, R. J., Leung, P.-S., Cohen, P., and Aronson, W. J. (2003). Insulin-like growth factor I (IGF-I) and IGF binding protein-1 modulate prostate cancer cell growth and apoptosis: possible mediators for the effects of diet and exercise on cancer cell survival. Endocrinology, 144(6), 2319–2324. (Cell culture; IGF-1 add-back experiments.) doi.org
- Barnard, R. J., Ngo, T. H., Leung, P.-S., Aronson, W. J., and Golding, L. A. (2003). A low-fat diet and/or strenuous exercise alters the IGF axis in vivo and reduces prostate tumor cell growth in vitro. The Prostate, 56, 201–206. (Cross-sectional comparison of 14 controls, 8 long-term diet-and-exercise adherents, and 12 long-term exercisers; serum tested on cancer cells in culture.) doi.org
- Aronson, W. J., Barnard, R. J., Freedland, S. J., et al. (2010). Growth inhibitory effect of low fat diet on prostate cancer cells: results of a prospective, randomized dietary intervention trial in men with prostate cancer. Journal of Urology, 183(1), 345–350. (Randomized, 18 men, 4 weeks; enrollment stopped after an interim analysis.) One author reported a financial interest or other relationship with the Pritikin Longevity Center; soy protein was provided by Protein Technologies International. doi.org
- Soliman, S., Aronson, W. J., and Barnard, R. J. (2011). Analyzing serum-stimulated prostate cancer cell lines after low-fat, high-fiber diet and exercise intervention. Evidence-Based Complementary and Alternative Medicine, 2011, 529053. (Serum from five men before and after a 3-week Pritikin program.) Dr. Barnard declared a conflict of interest as a consultant to the Pritikin Longevity Center. doi.org
- Gao, X., Sanderson, S. M., Dai, Z., et al. (2019). Dietary methionine influences therapy in mouse cancer models and alters human metabolism. Nature, 572(7769), 397–401. (Mouse models, plus a feeding study in six healthy people that measured metabolism.) Two authors hold patents related to targeting amino acid metabolism in cancer therapy. doi.org
- Zhou, Z.-Y., Wan, X.-Y., and Cao, J.-W. (2013). Dietary methionine intake and risk of incident colorectal cancer: a meta-analysis of 8 prospective studies involving 431,029 participants. PLOS ONE, 8(12), e83588. Funded by the Natural Science Foundation of Hubei Province; no competing interests declared. doi.org
- Tobias, D. K., Chai, B., Tamimi, R. M., et al. (2021). Dietary intake of branched chain amino acids and breast cancer risk in the NHS and NHS II prospective cohorts. JNCI Cancer Spectrum, 5(3), pkab032. Funded by the National Cancer Institute and the Breast Cancer Research Foundation; no disclosures. doi.org
- Katagiri, R., Song, M., Zhang, X., et al. (2020). Dietary intake of branched-chain amino acids and risk of colorectal cancer. Cancer Prevention Research, 13(1), 65–72. (Nurses’ Health Study, Nurses’ Health Study II, and Health Professionals Follow-up Study.) One author reported consulting for pharmaceutical companies, which did not fund the study. doi.org
- Khodabakhshi, A., Monfared, V., Arabpour, Z., Vahid, F., and Hasani, M. (2022). Association between levels of trimethylamine N-oxide and cancer: a systematic review and meta-analysis. Nutrition and Cancer, published online October 10, 2022. (16 studies, 5,930 participants.) No conflicts of interest reported. doi.org
- Richman, E. L., Kenfield, S. A., Stampfer, M. J., Giovannucci, E. L., Zeisel, S. H., Willett, W. C., and Chan, J. M. (2012). Choline intake and risk of lethal prostate cancer: incidence and survival. American Journal of Clinical Nutrition, 96, 855–863. (Health Professionals Follow-up Study, 47,896 men.) Supported by the U.S. National Institutes of Health and the Prostate Cancer Foundation; no conflicts of interest. doi.org
- Byrd, D. A., Zouiouich, S., et al. (2024). Associations of serum trimethylamine N-oxide and its precursors with colorectal cancer risk in the Prostate, Lung, Colorectal, Ovarian Cancer Screening Trial Cohort. Cancer, 130(11), 1982–1990. (761 colorectal cancer cases and matched controls.) One author is a co-inventor on patents relating to cardiovascular diagnostics and therapeutics. doi.org
- Han, P., Bidulescu, A., Barber, J. R., et al. (2019). Dietary choline and betaine intakes and risk of total and lethal prostate cancer in the Atherosclerosis Risk in Communities (ARIC) Study. Cancer Causes & Control, 30(4), 343–354. (6,528 men.) Supported by the National Cancer Institute. doi.org
- Erridge, C. (2011). The capacity of foodstuffs to induce innate immune activation of human monocytes in vitro is dependent on food content of stimulants of Toll-like receptors 2 and 4. British Journal of Nutrition, 105, 15–23. (In vitro screen of 40 food extracts.) Supported by a University of Leicester research fellowship. doi.org
- Ghanim, H., Abuaysheh, S., Sia, C. L., et al. (2009). Increase in plasma endotoxin concentrations and the expression of Toll-like receptors and suppressor of cytokine signaling-3 in mononuclear cells after a high-fat, high-carbohydrate meal: implications for insulin resistance. Diabetes Care, 32, 2281–2287. (Two groups of 10 healthy adults, single meals.) No conflicts of interest reported. doi.org
- Deopurkar, R., Ghanim, H., Friedman, J., et al. (2010). Differential effects of cream, glucose, and orange juice on inflammation, endotoxin, and the expression of Toll-like receptor-4 and suppressor of cytokine signaling-3. Diabetes Care, 33, 991–997. (Four groups of 12, single drinks.) Supported in part by the Florida Department of Citrus, which also provided the orange juice, and by the U.S. National Institutes of Health and the American Diabetes Association; no conflicts of interest reported. doi.org
- Pascual, G., Avgustinova, A., Mejetta, S., et al. (2017). Targeting metastasis-initiating cells through the fatty acid receptor CD36. Nature, 541(7635), 41–45. (Mouse models of human oral cancer.) The authors declared competing financial interests, detailed in the online version of the paper. doi.org
- Carta, G., Murru, E., Banni, S., and Manca, C. (2017). Palmitic acid: physiological role, metabolism and nutritional implications. Frontiers in Physiology, 8, 902. (Review.) Supported by the University of Cagliari; no conflicts of interest. doi.org
- Farvid, M. S., Spence, N. D., Holmes, M. D., and Barnett, J. B. (2020). Fiber consumption and breast cancer incidence: a systematic review and meta-analysis of prospective studies. Cancer, published online 2020. (Highest versus lowest total fiber RR 0.92.) No specific funding disclosed. One author has received grants from Arla Foods for work outside this study. doi.org
- O’Keefe, S. J. D., Li, J. V., Lahti, L., et al. (2015). Fat, fibre and cancer risk in African Americans and rural Africans. Nature Communications, 6, 6342. (Two-week supervised diet exchange; colon biopsy and microbiome biomarkers.) Funded by the U.S. National Institutes of Health; no competing financial interests declared. doi.org
- Egner, P. A., Chen, J.-G., Zarth, A. T., et al. (2014). Rapid and sustainable detoxication of airborne pollutants by broccoli sprout beverage: results of a randomized clinical trial in China. Cancer Prevention Research, 7(8), 813–823. (Randomized, 291 participants, 12 weeks.) One author reported being a consultant or advisory board member for Pfizer. doi.org
- Chen, J.-G., Johnson, J., Egner, P., et al. (2019). Dose-dependent detoxication of the airborne pollutant benzene in a randomized trial of broccoli sprout beverage in Qidong, China. American Journal of Clinical Nutrition, 110, 675–684. (Randomized, 170 participants, 10 days, three doses and placebo.) No conflicts of interest reported. doi.org
- Dai, X., Stanilka, J. M., Rowe, C. A., et al. (2015). Consuming Lentinula edodes (shiitake) mushrooms daily improves human immunity: a randomized dietary intervention in healthy young adults. Journal of the American College of Nutrition, 34(6), 478–487. (52 adults, 5 or 10 g a day for four weeks, no mushroom-free group.) Funded by the U.S. Mushroom Council, the Australian Mushroom Growers Association, and the University of Florida Agriculture Experiment Station. doi.org
- Nechuta, S. J., Caan, B. J., Chen, W. Y., et al. (2012). Soy food intake after diagnosis of breast cancer and survival: an in-depth analysis of combined evidence from cohort studies of US and Chinese women. American Journal of Clinical Nutrition, 96, 123–132. (9,514 survivors; recurrence HR 0.75 for 10 mg or more of isoflavones a day.) Supported by the National Cancer Institute; no conflicts of interest. doi.org
- Calubag, M. F., Robbins, P. D., and Lamming, D. W. (2024). A nutrigeroscience approach: dietary macronutrients and cellular senescence. Cell Metabolism, 36(9), 1914–1944. (Review.) One author has received funding from, and is a scientific advisory board member of, Aeovian Pharmaceuticals, which is developing mTOR inhibitors; another is a co-founder of Itasca Therapeutics, which is developing senotherapeutics. doi.org
- Becerra-Tomás, N., Balducci, K., Abar, L., et al. (2023). Postdiagnosis dietary factors, supplement use and breast cancer prognosis: Global Cancer Update Programme (CUP Global) systematic literature review and meta-analysis. International Journal of Cancer, 152, 616–634. Funded by the World Cancer Research Fund network; no conflicts of interest declared. doi.org
- Thomson, C. A., Rock, C. L., Thompson, P. A., et al. (2011). Vegetable intake is associated with reduced breast cancer recurrence in tamoxifen users: a secondary analysis from the Women’s Healthy Eating and Living Study. Breast Cancer Research and Treatment, 125, 519–527. (Secondary analysis, 3,080 survivors.) Supported by the National Cancer Institute and the Walton Family Foundation. doi.org
- Nechuta, S., Caan, B. J., Chen, W. Y., et al. (2013). Post-diagnosis cruciferous vegetable consumption and breast cancer outcomes: a report from the After Breast Cancer Pooling Project. Cancer Epidemiology, Biomarkers & Prevention, 22(8), 1451–1456. (11,390 survivors from four cohorts.) Supported by the National Cancer Institute; no conflicts of interest. doi.org
- Pierce, J. P., Stefanick, M. L., Flatt, S. W., et al. (2007). Greater survival after breast cancer in physically active women with high vegetable-fruit intake regardless of obesity. Journal of Clinical Oncology, 25(17), 2345–2351. (1,490 women from the WHEL comparison group.) Supported by the Walton Family Foundation and the National Cancer Institute; no potential conflicts of interest. doi.org
- Kroenke, C. H., Kwan, M. L., Sweeney, C., Castillo, A., and Caan, B. J. (2013). High- and low-fat dairy intake, recurrence, and mortality after breast cancer diagnosis. Journal of the National Cancer Institute, 105, 616–623. (Life After Cancer Epidemiology study, 1,893 women.) Supported by the National Cancer Institute. doi.org
- Inoue-Choi, M., Robien, K., and Lazovich, D. (2013). Adherence to the WCRF/AICR guidelines for cancer prevention is associated with lower mortality among older female cancer survivors. Cancer Epidemiology, Biomarkers & Prevention, 22(5), 792–802. (Iowa Women’s Health Study, 2,017 survivors.) No conflicts of interest declared. doi.org
- Ornish, D., Weidner, G., Fair, W. R., et al. (2005). Intensive lifestyle changes may affect the progression of prostate cancer. Journal of Urology, 174(3), 1065–1069; discussion 1069–1070. (Randomized, 93 men, 1 year.) Supported by a U.S. Department of Defense grant. The lead author disclosed a financial interest or other relationship with Random House and HarperCollins; another author disclosed relationships with TAP Pharmaceutical Products, AstraZeneca, and Pfizer. Coauthor R. James Barnard disclosed a consulting relationship with the Pritikin Longevity Center in a later paper; no such disclosure appears in this one. doi.org
- Carmody, J., Olendzki, B., Reed, G., Andersen, V., and Rosenzweig, P. (2008). A dietary intervention for recurrent prostate cancer after definitive primary treatment: results of a randomized pilot trial. Urology, 72, 1324–1328. (Randomized pilot, 36 men, wait-list control.) Supported by a U.S. Department of Defense award. doi.org
- Richman, E. L., Stampfer, M. J., Paciorek, A., Broering, J. M., Carroll, P. R., and Chan, J. M. (2010). Intakes of meat, fish, poultry, and eggs and risk of prostate cancer progression. American Journal of Clinical Nutrition, 91, 712–721. (CaPSURE, 1,294 men.) Supported by the U.S. Department of Defense and the Prostate Cancer Foundation; the CaPSURE registry’s founding sponsor was TAP Pharmaceutical Products, which had no role in the study. doi.org
- Richman, E. L., Carroll, P. R., and Chan, J. M. (2012). Vegetable and fruit intake after diagnosis and risk of prostate cancer progression. International Journal of Cancer, 131(1), 201–210. (CaPSURE, 1,560 men, 134 progression events.) CaPSURE was supported by an unrestricted educational gift from Abbott Laboratories and by U.S. federal grants. doi.org
- Liu, V. N., Van Blarigan, E. L., Zhang, L., et al. (2024). Plant-based diets and disease progression in men with prostate cancer. JAMA Network Open, 7(5), e249053. (CaPSURE, 2,062 men.) CaPSURE has been supported by the U.S. Department of Defense, the Prostate Cancer Foundation, TAP Pharmaceuticals, and the National Cancer Institute; several authors reported relationships outside the submitted work. doi.org
- Pettersson, A., Kasperzyk, J. L., Kenfield, S. A., et al. (2012). Milk and dairy consumption among men with prostate cancer and risk of metastases and prostate cancer death. Cancer Epidemiology, Biomarkers & Prevention, 21(3), 428–436. (Health Professionals Follow-up Study, 3,918 men.) Supported by U.S. National Institutes of Health grants; no potential conflicts of interest disclosed. doi.org
- Yang, M., Kenfield, S. A., Van Blarigan, E. L., et al. (2015). Dairy intake after prostate cancer diagnosis in relation to disease-specific and total mortality. International Journal of Cancer, 137(10), 2462–2469. (Physicians’ Health Study, 926 men.) doi.org
- Downer, M. K., Batista, J. L., Mucci, L. A., et al. (2017). Dairy intake in relation to prostate cancer survival. International Journal of Cancer, 140(9), 2060–2069. (Swedish cohort, 525 men.) One author’s employment at a pharmaceutical company was declared unrelated to the work. doi.org
- Tat, D., Kenfield, S. A., Cowan, J. E., et al. (2018). Milk and other dairy foods in relation to prostate cancer recurrence: data from the Cancer of the Prostate Strategic Urologic Research Endeavor (CaPSURE). The Prostate, 78(1), 32–39. (1,334 men.) Funded by the U.S. Department of Defense and the National Institutes of Health; no conflicts of interest declared. doi.org
- Loeb, S., Fu, B. C., Bauer, S. R., et al. (2022). Association of plant-based diet index with prostate cancer risk. American Journal of Clinical Nutrition, 115, 662–670. (Health Professionals Follow-up Study, 47,239 men.) One author reports equity in Fellow Health; another was paid by Myovant Sciences for consulting on patient education materials. doi.org
Last updated: October 26, 2026 at 1:01 PM UTC