Prostate cancer is the third leading cause of cancer death among men in the United States. Specific foods and nutritional supplements have been presumed to lower risk of recurrence, often based on extrapolating evidence for prevention of the primary diagnosis of prostate cancer. However, the newest data shows it is not that simple. Much of the advice regarding red meat, poultry, eggs, milk, fish, and coffee needs to be revamped. In addition, the role of vitamin E and selenium, long controversial topics, may need to be looked at with more individualization using genomic testing.
Prostate cancer is the most common non-skin malignancy and the third leading cause of cancer death among men in the United States. The American Cancer Society estimated in 2017 that 161,360 men would be diagnosed with, and 26,730 would die from, prostate cancer. The majority of these men, about 92%, were diagnosed with low-grade, localized disease. Globally prostate cancer incidence varies 60-fold, suggesting that lifestyle and dietary factors may have a large impact.1
In the last 10 years, significant strides have been made regarding the relationship between diet and prostate cancer. Recent research suggests a number of dietary factors are specifically associated with significant impacts on recurrence risk. This article will review some of the more important data that should inform the advice we give to men who have been diagnosed with prostate cancer.
Eggs and Chicken
In 2010, a paper by Erin Richman was among the first to inform us on the possible detrimental effects of egg and chicken consumption on prostate cancer recurrence. Richman prospectively tracked consumption of processed and unprocessed red meat, fish, poultry, and eggs to see whether risk of prostate cancer recurrence or progression was affected by eating these foods. Her initial assumption was that processed and red meat, then assumed to increase initial diagnosis, would be the culprits that increased risk of recurrence as well. Chicken and eggs were tracked only because it was thought they were safe and would have no effect. Richman and colleagues followed 1,294 men diagnosed with prostate cancer for an average of 2 years. Though a short trial, their data added up to 2,610 person-years during which time 127 events occurred, defined as either prostate cancer death, metastasis, elevated prostate-specific antigen (PSA), or secondary treatment.
The results were surprising. Men who ate the most eggs or poultry with skin were at double the risk of recurrence or progression during the trial compared to men who ate smaller amounts. For eggs, comparing the highest and lowest quartiles of consumption, or an average of 5.5 eggs per week against, 0.4 eggs, the hazard ratio (HR) was 2.02 (95% confidence interval [CI]: 1.10-3.72; P for trend=0.05). Comparing highest and lowest quartiles of poultry consumption yielded a nonsignificant HR of 1.55; however, a comparison of the upper and lower tertiles of those eating poultry with skin on yielded a statistically significant HR of 2.26. (95% CI: 1.36- 3.76; P for trend=0.003). Men with higher prognostic risk and a high poultry intake had a 4-fold increase in risk compared with men with low/intermediate prognostic risk and a low poultry intake (P for interaction=0.003). Eating processed or unprocessed red meat, fish, or skinless poultry was not associated with prostate cancer recurrence or progression; it was only consumption of eggs and poultry with skin that significantly increased risk.2
This was not what would have been expected, as the focus had been to advise against eating red meat and processed meat. In light of Richman’s data this recommendation seems misguided compared to reducing consumption of eggs and chicken skin. It also implies that factors that cause prostate cancer do not necessarily promote recurrence or progression.
Richman’s chicken and egg study was something of a turning point. Prior to its publication we had scant data about primary prevention and assumed that after diagnosis we should tell men pretty much the same things we told them before they were diagnosed. For prostate cancer prevention, we typically recommended that our patients follow a low-fat diet, rich in fruits and vegetables, and avoid high energy intake, excessive meat, dairy products, and calcium.
Men with prostate cancer, in their attempts to follow a “healthier” diet, often reduce their red meat consumption and increase egg and poultry consumption to compensate for a reduction in protein. Richman showed that this is ill-advised.
It is worth noting that eggs and poultry do not increase risk of developing prostate cancer.3 In 2007 Harvard researchers Willard, Giovannucci, and Liu et al suggested that the factors that promote prostate cancer progression after diagnosis are different from the factors that initiate the cancer. Other risk factors may increase how aggressive the cancer is.4
Richman’s 2010 paper was not the first to associate poultry intake with prostate cancer progression. A 2001 paper by Michaud et al reported an association between poultry skin and metastatic prostate cancer risk.5
The explanation for how chicken skin and eggs influence recurrence seem obvious. Chicken skin contains large quantities of heterocyclic amines, the chemicals blamed for the increase in prostate cancer risk from high meat consumption.6 Cooked poultry skin contains more heterocyclic amines (HCA) than any other type of cooked meat; more HCA than barbecued beef or even fried bacon.7 Eggs contain the highest amount of choline found in any food; choline strongly promotes prostate cancer growth.8 Johansson et al reported in 2009 that high plasma choline is associated with greater risk of prostate cancer.9 Prostate cancer cells take up far more choline than healthy prostate cells.10 Radio-labeled choline is used for PET (positron emission tomography) scans of prostate tumors instead of radioactive glucose that is used in standard PET scans.11 Richman reported in 2012 that dietary choline was correlated with risk of prostate cancer death.8
Erin Richman, now Erin Van Blarigan, was also the lead author of a 2015 study that assessed risk of dying from prostate cancer with consumption of saturated fat vs vegetable fats after prostate cancer diagnosis. This was a prospective study of men diagnosed with prostate cancer (n=926) from the Physicians’ Health Study. During follow-up, 333 deaths (56 prostate cancer deaths) occurred. Even slight shifts in type of fat and percentage of calories sourced from fats and carbohydrates made significant differences. Men who obtained 5% more of their daily calories from saturated fat and 5% less of their daily calories from carbohydrate after diagnosis had a 1.8-fold increased risk of all-cause mortality (HR: 1.81; 95 % CI: 1.20- 2.74; P=0.005) and a 2.8-fold increased risk of prostate cancer–specific mortality (HR: 2.78; 95 % CI: 1.01-7.64; P=0.05). Men who obtained 10% more of their daily calories from vegetable fats and 10% less of their daily calories from carbohydrates had a 33% lower risk of all-cause mortality (HR: 0.67; 95 % CI: 0.47-0.96; P=0.03). The authors concluded that saturated fat intake may increase risk of death and vegetable fat intake may lower risk of death.12
Milk and Dairy Foods
In January 2018, a team that included Van Blarigan reported that consumption of dairy foods was associated with prostate cancer recurrence. They had followed 1,334 men with nonmetastatic prostate cancer who were part of the CAPSURE cohort for a mean of 8 years, starting about 2 years after diagnosis.13
Men who consumed >4 servings of whole milk per week had a 73% increased risk of recurrence of prostate cancer compared to those who consumed 0 to 3 servings per month (HR: 1.73; 95%CI: 1.00-2.98; P=0.04). Higher BMI tended to worsen the association (P=0.01). For men with a BMI ≥27, making the same comparison of >4 servings/week of whole milk vs 0-3 servings/month, the risk of recurrence tripled (HR: 2.96; 95% CI: 1.58-5.54; P<0.001). Whole milk consumption was not associated with increased risk in men with BMI <27. A number of earlier studies suggested that dairy products increased risk of primary diagnosis of prostate cancer as well.14-22
Data about milk consumption after diagnosis has been limited. Van Blarigan (as Richman) and her current coauthors reported in 2012 that among the 3,918 men in the Health Professionals Follow-up Study, those who consumed whole milk >4 times per week after diagnosis had double the risk of dying from prostate cancer and a 51% increased risk of prostate cancer recurrence compared to those who drank whole milk <3 times per month.23
These same authors also analyzed data from the 926 men in the Physicians Health Study and reported in 2015 that consuming >3 servings a day of dairy products was associated with a 2.4-fold increased risk of prostate cancer–specific mortality compared to consuming <1 serving per day.24
A 2017 Swedish study by Downer et al looked at dairy consumption among 230 men diagnosed with localized prostate cancer. Drinking ≥3 servings per day of high-fat milk increased prostate cancer mortality 6.1-fold over men who <1 serving per day.25
The bottom line is that consuming dairy, in particular whole milk, increases risk of recurrent prostate cancer.
Van Blarigan and colleagues had already reported that saturated fat intake is associated with increased risk of prostate cancer mortality. This data that whole milk has a greater effect suggests that the fat content of whole milk increases risk.12 This clearly should raise concern about other high-fat dairy products.
Several other studies have looked at whole milk and prostate cancer recurrence. Pettersson et al reported that men who had the highest whole milk intake (>4 servings/week) had approximately a 2-fold increased risk of prostate cancer mortality (HR: 2.15; 95% CI: 1.28-3.60) compared to men with low intake (0-3 servings /month). In their study there was no association between total dairy consumption and risk of death from prostate cancer or biochemical or clinical recurrence.26
Whole milk and any other high-fat dairy products should be added to the list of foods to avoid for men with a history of prostate cancer.
The BMI data from Van Blarigan’s latest milk study reinforces how important BMI is in prostate cancer. Until this study the idea that the impact of foods would vary significantly with BMI was not considered. Van Blarigan was part of a team that reported in 2015 that prediagnosis obesity was associated with shorter telomere lengths in prostate stromal cells, and overweight patients had higher Gleason scores. Telomere length also varied with physical activity; more active men had significantly longer telomeres in their prostate cells and the least active men tended toward shorter lengths.27
Being overweight also affects prostate cancer prognosis. In a study of 2,546 men with localized prostate cancer in the Physician’s Health Study (PHS), a single-unit higher BMI before cancer diagnosis was associated with about a 10% increase in the risk of prostate cancer–specific mortality. A BMI of ≥30 was associated with nearly double the risk of prostate cancer death compared to normal BMI.28
Men who gain >2.2 kg between 5 years before to 1 year after prostate cancer surgery have a 94% increased risk of recurrence.29 Gaining weight after diagnosis is associated with increased risk of biochemical recurrence and prostate cancer–specific mortality.30 Metabolic syndrome is strongly associated with increased risk of high-grade and advanced prostate cancer.31,32
Prospective cohort studies suggest that vigorous exercise is associated with lower risk of prostate cancer–specific mortality. In terms of metabolic equivalent task (MET) values, vigorous exercise means MET>6. In simple terms, this means jogging, biking, swimming or exercise of similar intensity. After analyzing data from the Health Professionals Follow-up Study, Kenfield and other Harvard researchers reported in 2011 that men diagnosed with prostate cancer (n=2,705) who performed ≥3 hours per week of vigorous activity had a 61% lower risk of dying from prostate cancer when compared to men who exercised <1 hour per week.33
Men with prostate cancer, in their attempts to follow a ‘healthier’ diet, often reduce their red meat consumption and increase egg and poultry consumption to compensate for a reduction in protein. Richman showed that this is ill-advised.
Van Blarigan (as Richman), working with this same Harvard team, reported similar results in 2011 after analyzing data from the CAPSURE cohort: men who walked briskly (≥3 mph) 3 or more hours per week had a 57% lower risk of prostate cancer recurrence compared with men who walked <3 hours a week at an easy pace (<2 mph).34
Bonn reported similar effects in 2015 from a cohort of 4,623 men with localized prostate cancer. Men who either walked or biked ≥20 minutes per day had a 36% decrease in prostate cancer mortality.35
In a 2015 paper, again with this Harvard group, Van Blarigan explained in part why exercise might be so beneficial in prostate cancer. Exercise affects tumor morphology—it actually changes the architecture of prostate tumors, leading to more regularly shaped blood vessels in the tumors. Microvessel morphology was examined in men with prostate cancer (n=571) and compared to activity levels. The investigators concluded that vigorous walking was associated with “larger, more regularly shaped blood vessels compared with those of men who walked at a less than brisk pace.”36
While smoking cessation might seem to be wise, many men with prostate cancer find reasons to rationalize continuing to smoke. Smoking increases prostate cancer aggressiveness and the risk of dying from prostate cancer. For smokers, every aspect of prostate cancer prognosis is worse. They have a higher risk of progression, including biochemical recurrence, metastasis, and development of hormone-resistant disease.
In the first of the following 2 studies, 5,366 men with prostate cancer who smoked prior to diagnosis had a 61% higher risk of biochemical recurrence. Recurrence risk decreased in men who stopped smoking for 10 or more years to about the same level as those who had never smoked.37
In the second study (N=6,538) current smokers had 80% higher risk of biochemical recurrence compared with nonsmokers (HR: 1.80; 95% CI: 1.45-2.24; P<0.001). Former smokers had a 63% increased risk [HR: 1.63; 95% CI: 1.30-2.04; P<0.001)].38
In some studies men who eat more fish have lower risk of death from prostate cancer than men who eat little fish. While some meta-analyses have not found significant associations, others have. A 2010 pooled analysis by Szymanski of 4 cohort studies reported a significant 63% reduction in prostate cancer mortality among the fish eaters (n=49,661; relative risk [RR]: 0.37; 95% CI: 0.18-0.74).1
It is thought that oily fish that are high in omega-3 fatty acids, such as salmon, sardines, mackerel, and herring may have the most benefit. Yet there are some discrepancies. Brasky reported in 2013 that men with higher omega-3 blood levels are actually at higher risk of being diagnosed with prostate cancer by PSA testing.39 Yet this study has since been discredited.40-43
Several observational studies report that coffee consumption prediagnosis is associated with less prostate cancer mortality and reduced risk of recurrence.44
In 47,911 men in the Health Professionals Follow-up Study, 5,035 cases of prostate cancer were diagnosed, 642 of which proved fatal. Men who drank 6 or more cups of coffee per day compared with nondrinkers had an 18% lower risk of prostate cancer, but this did not reach statistical significance.(RR: 0.82; 95% CI: 0.68-0.98; P=0.10). The association was stronger and significant for prostate cancer death (for men who consumed ≥6 cups of coffee per day; RR: 0.40; 95% CI: 0.22-0.75: P=0.03).44
Geybels et al reported in 2013 that in a group of 630 men treated for prostate cancer, drinking >4 cups of regular or decaffeinated coffee per day prediagnosis vs <1 cup was associated with a 59% reduced risk of recurrence or progression (HR: 0.41; 95 % CI: 0.20–0.81; P=0.01).45 At the time of this review there is no published data examining coffee drinking postdiagnosis.
Cruciferous vegetables contain compounds whose metabolites stop cancer cells from growing and encourage apoptosis.46 Consuming these vegetables is associated with lower risk of developing advanced prostate cancer. Eating >1 serving of broccoli per week was significantly associated with a 45% decrease in risk (RR: 0.55; 95% CI: 0.34-0.89; P=0.02). The same frequency of consumption for cauliflower was associated with a 52% drop in risk (RR: 0.48; 95% CI: 0.25-0.8] P=0.03).47
A 2012 paper by Richman’s group looked at cruciferous consumption after diagnosis48 and examined whether intake of vegetables, tomato sauce, and beans would shift prostate cancer progression in 1,560 men diagnosed with localized prostate cancer. They reported that men in the highest quartile of cruciferous vegetable intake, who ate almost 6 servings per day after diagnosis, had a 59% lower risk of progression compared with men in the lowest quartile. Six servings a day is rather a lot of crucifers to ingest in 1 day.
Consumption of other vegetable groups was not associated with risk. Trends toward lower risk associated with consumption of other cruciferous vegetable types (coleslaw, cauliflower, and Brussels sprouts) did not reach statistical significance. No benefit was seen in this particular study for tomato sauce.
Laboratory and epidemiology research suggest soy isoflavones may inhibit prostate cancer growth. A 2009 meta-analysis looking at soy and prostate cancer risk reported findings that deserve close investigation. The studies on soy intake yielded a combined RR that was significantly lower (26%) in men who ate soy (RR: 0.74; 95% CI: 0.63-0.89; P=0.01). When the data were stratified into those eating nonfermented vs fermented soy foods, the benefit was solely from the nonfermented soy. Nonfermented soy foods lowered risk of prostate cancer by 30% (RR/OR of 0.70; 95% CI: 0.56-0.88; P=0.01). Fermented soy food consumption had no significant effect (RR: 1.02; 95% CI: 0.73, 1.42; P=0.92).49 This conclusion was corroborated by a meta-analysis in 2018.50
Tomato and Lycopene
Although studies have associated consumption of cooked tomatoes and tomato-based products with reduction of high-risk prostate cancer,51 it is unclear whether these same foods offer benefit postdiagnosis. Two studies have looked at the relationship between prostate cancer progression and tomato products—one investigated tomato sauce, the other lycopene, a carotenoid found in tomatoes—and results are not consistent. Chan et al reported an inverse linear relationship for tomato sauce and risk of progression for those consuming 2 servings per week,52 suggesting tomato sauce may have benefit.
The second, a 2009 systematic review, concluded that there is insufficient evidence to support the usefulness of lycopene supplements in routine clinical practice for patients diagnosed with prostate cancer. The authors added that although the studies they reviewed indicate that lycopene is unlikely to be harmful, in their opinion none of the studies were conducted with adequately sound methodology.53
In a 2014 paper published with former associates from Harvard, Van Blarigan (as Richman) reported that, although closer adherence to a Mediterranean diet did not lower risk of dying from prostate cancer, it did lower overall mortality risk. Van Blarigan prospectively followed 47,867 men from 1986 to 2010. This included 4,538 men diagnosed with localized prostate cancer, followed from diagnosis to death or to January 2010. During that time, 6,220 prostate cancer cases were confirmed. The Mediterranean diet was not associated with risk of advanced cancer or prostate cancer mortality. However, there was a 22% lower risk of overall mortality (HR: 0.78; 95% CI: 0.67-0.90; P=0.0007) among men with greater adherence to the Mediterranean diet after diagnosis.54
Observational studies initially suggested an inverse association between vitamin E and risk of prostate cancer, which led to vitamin E frequently being recommended.55,56 Three large randomized controlled trials report conflicting results (ie, higher vitamin E was associated with worse disease risk).57-59 In 2013 Richman was part of the group that sorted out and possibly explained this vitamin E phenomenon.
They measured circulating α- and γ-tocopherol and genotyped 30 single nucleotide polymorphisms (SNPs) among 573 men with prostate cancer. They compared circulating vitamin E, genotypes, and risk of high-grade prostate cancer, and risk of recurrence (56 events; 3.7 years median follow-up). Circulating γ-tocopherol was associated with an 87% increase in risk of high-grade prostate cancer (Q4 v Q1 odds ratio [OR]: 1.87; 95% CI: 0.97-3.58; P=0.02). The less common allele in superoxide dismutase 3 (SOD3), 3rs699473, was associated with an increased risk of high-grade disease (OR: 1.40; 95% CI: 1.04-1.89). However, 2 independent SNPs in SOD1 were inversely associated with prostate cancer recurrence (HR: 0.49; 95% CI: 0.25-0.96 for rs17884057 and HR: 0.62; 95% CI: 0.40-0.95 for rs9967983). Genetic variation in SOD may be associated with risk of high-grade disease at diagnosis and disease recurrence. Circulating γ-tocopherol levels may also be associated with an increased risk of high-grade disease. This explanation should put an end to across-the-board recommendations that men with prostate cancer should take γ-tocopherol.60
In 2014 Van Blarigan and colleagues pursued these genetic variations a step further. Circulating prediagnostic α-tocopherol, γ-tocopherol, and lycopene levels were analyzed along with various SNPs and the risk of prostate cancer death in 2,439 men with prostate cancer in the Health Professionals Follow-up Study and Physicians' Health Study. They observed 223 events over 10 years of follow-up. Risks varied with different alleles. High α-tocopherol levels were in general associated with lower risk of prostate cancer mortality. Men homozygous for the less common glutathione peroxidase (GPX4) allele (rs3746165) had a 35% lower risk of prostate cancer death compared with men homozygous for the more common allele. However, men who were homozygous for the less common allele rs3746165, and who had high γ-tocopherol levels were at 3.5-fold increased risk of prostate cancer death. This implies that genetic SNP testing should be conducted prior to recommending vitamin E supplements.61
Selenium is another supplement commonly recommended to men with a history of prostate cancer that has been questioned by prospective human trials. The 2003 van den Brandt study associated low toenail selenium levels with higher risk of prostate cancer.62 Hurst’s 2012 meta-analysis perpetuated this belief.63 Yet this notion was countered by the results reported from the SELECT (Selenium and Vitamin E Cancer Prevention Trial) cohort that found supplementation with selenium had no significant impact on disease risk.58
After following 4,459 men diagnosed with prostate cancer for 22 years, Van Blarigan and colleagues reported in 2014 that those who took selenium supplements had a higher risk of dying from prostate cancer. During 7.8 years of follow-up, recurrence rates were 5.6 per 1,000 person-years for those not taking selenium supplements and 10.5 per 1,000 person-years for those who took 140 μg/day or more. When nonselenium users were compared with men who took selenium, rates of biochemical recurrence were 28.4 vs 29.3 per 1,000 person-years. Risk of dying from prostate cancer increased with selenium doses. Men consuming low-dose selenium (1-24 μg/day) had an 18% higher risk than nonusers. Those consuming 25 to 139 μg/day had a 33% increased risk of prostate cancer mortality and those taking 140 or more had a 2.60-fold greater risk. This clearly contradicted the general belief that every man who had prostate cancer should take 200 μg or more of selenium per day.64
In 2015 Van Blarigan and colleagues identified specific polymorphisms in selenoprotein coding genes that were associated with higher-grade disease that might affect prostate cancer recurrence.65
In 2014 Van Blarigan and colleagues published an important paper on folate. The group prospectively examined the association between postdiagnostic folate consumption and the risk of prostate cancer recurrence after radical prostatectomy, external beam radiation therapy, and brachytherapy. Prior to commencing this study, a randomized, placebo-controlled clinical trial of folic acid supplementation for the chemoprevention of colorectal adenoma had revealed an increased incidence of prostate cancer in the treatment group. The study was done with 1,153 men who had been treated with radical prostatectomy, external beam radiation therapy, and brachytherapy who participated in the CaPSURE (Cancer of the Prostate Strategic Urologic Research Endeavor) Diet and Lifestyle study.
Prostate cancer progressed in 101 men (8.76%) during a mean 34-month follow-up. Initially there was no association of folate intake and recurrence risk. However, on secondary analysis by treatment type, men who had a radical prostatectomy and were in the lowest decile of dietary folate intake had a 2.6-fold increase in the risk of recurrence (HR: 2.56; 95% CI: 1.23-5.29; P=0.01). In patients treated with external beam radiation and brachytherapy no evidence of an association between prostate cancer progression and folate intake was seen.
Though it is rare to see patients with low folate levels, this paper would suggest it is unlikely there is any harm with folate supplementation in men who have had prostate cancer. In fact, we should consider supplementation if a man’s circulating levels are low after prostatectomy.66
Taking a multivitamin is probably useful. The Physicians Health Study randomized trial of a regular multivitamin reported a modest but significant (8 %) reduction in total cancer incidence in men (HR: 0.92; 95% CI: 0.86-0.998; P=0.04). The men with a history of prior cancer had a 27% reduction in total cancer during the study (HR: 0.73; 95% CI: 0.56-0.96; P=0.02). There was no significant effect on risk of prostate cancer.67 There is still no strong evidence that any single nutrient offers protection against prostate cancer (neither development nor progression). However, there are hints that a "U-shape” relationship may exist between prostate cancer risk and folate, vitamin C, vitamin D, and calcium.68
In a 2016 paper, Nordström et al reported that circulating carotenoids were inversely associated with the risk of high-grade prostate cancer. The ORs for the highest vs lowest quartiles were 0.34 (95% CI: 0.18-0.66) for α-carotene, 0.31 (95% CI: 0.15-0.63) for β-carotene, 0.55 (95% CI: 0.28-1.08) for lycopene, and 0.37 (95% CI: 0.18-0.75) for total carotenoids. Once again, these effects were modified by various SNPs.69 Thus, the argument that genetic testing should be performed in prostate cancer patients continues to strengthen.
To summarize, based on the evidence for factors associated with a lower risk of recurrence, the following are the most important directives we can give men who have been diagnosed with prostate cancer:70,71
- Don’t smoke.
- Lose weight if BMI is ≥27.
- Exercise vigorously several days per week (do something to sweat).
- Eat lots of vegetables (particularly cruciferous vegetables).
- Eat more fish and vegetable fats (eg, fish, nuts, vegetable oils, soybeans, avocados, and flaxseed), less saturated fat, and less carbohydrates.
- Drink coffee (regular or decaffeinated, it doesn’t seem to matter).
- Limit consumption of eggs and poultry with skin on.
- Limit whole milk consumption, especially if your BMI is high.
- Limit consumption of refined grains, sugars, processed meat, and high-fat dairy products.
Providers should test SNPs before making suggestions for vitamin E, selenium, or lycopene supplementation.
Granted, at this point randomized clinical trials using the above as interventions are lacking. Still, this is more information than was available just a few years ago. A large amount of it is the result of Erin (Richman) Van Blarigan’s steady research.