Systematic review of the impact of a plant-based diet on prostate cancer incidence and outcomes

Abstract

Background:

Plant-based diets are increasingly popular and have many well-established benefits for health and environmental sustainability. Our objective was to perform a systematic review of plant-based diets and prostate cancer.

Methods:

We performed a systematic database and citation search in February 2022. Studies were included if they reported primary data on plant-based dietary patterns (i.e., vegan, vegetarian, plant-based) and incidence among at-risk men for prostate cancer, or oncologic, general health/nutrition, or quality of life outcomes among patients with prostate cancer or caregivers.

Results:

A total of 32 publications were eligible for the qualitative synthesis, representing 5 interventional and 11 observational studies. Interventional studies primarily focused on lifestyle modification including plant-based diets for men on active surveillance for localized prostate cancer or with biochemical recurrence after treatment, showing improvements in short-term oncologic outcomes alongside improvements in general health and nutrition. Observational studies primarily focused on prostate cancer risk, showing either protective or null associations for plant-based dietary patterns. Studies of the vegan diet consistently showed favorable associations with risk and/or outcomes. Gaps in the current literature include impact for long-term disease-specific outcomes.

Conclusions:

Interventional studies showed generally favorable results of lifestyle modifications incorporating a plant-based diet with prostate cancer outcomes as well as improvements in nutrition and general health. Observational studies demonstrated either a lower risk of prostate cancer or no significant difference. These results are encouraging in light of the many benefits of plant-based diets for overall health, as well as environmental sustainability and animal welfare.

Introduction

Plant-based foods are increasingly widespread, with market data showing 29% growth in U.S. dollar sales of plant-based foods over the past two years.¹ The EAT-Lancet Commission concluded that universal adoption of a diet rich in plant-based foods with fewer animal source foods would reduce environmental degradation while preventing approximately 11 million human deaths per year.²

Plant-based diets reduce the risk of many major health conditions and increase survival.^3–6 By contrast, animal-based foods are associated with an increased risk of cardiovascular disease and cancer. The World Health Organization classifies processed meat as a Group 1 carcinogen and red meat as Group 2A.⁷ While this designation was primarily based on colorectal cancer risk, associations were also seen for pancreatic and prostate cancer.

There are many potential biological mechanisms through which plant-based diets might reduce prostate cancer risk, such as reduced exposure to hormones found in animal foods and heterocyclic amines formed during cooking as well as increased exposure to anti-cancer compounds found in plant foods.⁸ Moreover, obesity has been linked with an increased risk of aggressive prostate cancer ⁹, and studies suggest that people who adopt a plant-based diet gain less weight as they age.¹⁰ The importance of nutrition for men with prostate cancer is highlighted by cardiovascular disease being the leading competing cause of death.^11–13

Additionally, a study using animal models found a diet containing 20% plant protein instead of 20% animal protein was associated with a 37% reduction in tumor weight.¹⁴ Ecological studies suggest that countries with greater consumption of animal foods have higher prostate cancer mortality, and those with greater per capita consumption of vegetable-based foods have lower prostate cancer mortality.¹⁵ Importantly, many previous studies about nutrition and prostate cancer risk or outcomes have primarily focused on individual foods, rather than full dietary patterns. Our objective was to perform a systematic review of the literature regarding plant-based diets with prostate cancer risk and outcomes.

Methods

The study protocol was registered in the PROSPERO database (Record ID: 192654). A systematic literature search was conducted by an experienced medical librarian (CP) on February 28, 2022 and included the databases PubMed via PubMed.gov, Embase via Elsevier, the Cochrane Library via Wiley, Scopus via Elsevier, Web of Science via Clarivate, AMED via Ovid, Nursing and Allied Health via ProQuest, and CINAHL Plus via EBSCO. No limits were made for publication date or language. The search included prostate cancer and broad terminology for a vegetarian diet (Supplemental Table 1). Studies were included if they reported primary data on full plant-based dietary patterns (i.e. vegan, vegetarian, plant-based) and incidence among at-risk men for prostate cancer, or oncologic, general health/nutrition, or quality of life (QOL) outcomes among patients with prostate cancer or their caregivers. Animal studies, review articles, meta-analyses, abstracts, and articles not written in English were excluded. In addition, studies that pertained to single dietary components (e.g., dairy, soy, vegetables) instead of a full plant-based dietary pattern or studies that only focused on increasing consumption of plant-based foods without reduction or elimination of animal-based foods were excluded.^16,17 Use of supplements was not part of the selection criteria.

Figure 1 shows the PRISMA flow chart. Database search results were independently screened for eligibility by two reviewers (NG, SL) using Covidence.¹⁸ The citation search involved screening references cited in all studies selected for inclusion from our database search. Disagreements were resolved by discussion and consensus of the study team. Full-text articles were independently reviewed by multiple study team members.

Figure 1.

Preferred reporting items for systematic reviews and meta-analyses (PRISMA) flowchart of the search process

Quality assessment was performed using the criteria by Hawker et al.¹⁹ The checklist assesses 9 components, each scored from “very poor” (1) to “good” (4), with 36 being the highest total score (Appendix 1). If a single study corresponded to multiple publications, information from all the publications were collated to tabulate a composite quality score.

Results

There were 5 interventional studies with 15 corresponding publications (Table 1, Supplemental Table 2, Supplemental Table 3), and 11 observational studies with 17 publications (Table 2, Supplemental Table 4). Overall, the median quality rating for included studies was 32.5 (range 20–36).

Table 1.

Interventional studies (n=5)

Study description, eligible population, sample size	Reported age	Follow-up	Key findings	Quality scorea
Gene Expression Modulation by Intervention with Nutrition and Lifestyle Trial among low-risk PCa patients on AS20,21 (n=31, single arm)	Mean age: 62.3 y Age range 49–80 y	3 months	1/31 patients progressed to surgery; Improved BMI*, systolic blood pressure,*, diastolic blood pressure*, LDL*, HDL*, total cholesterol*, LDL/HDL ratio*, waist circumference*, mental component summary (SF-36)*, intrusive thoughts and avoidance (impact of event scale)*. No significant difference in total PSA or %free PSA.	36
Prostate Cancer Lifestyle Trial- RCT of vegan diet/lifestyle intervention in low-risk PCa patients on AS22–28 (n=44 in intervention arm, n=49 in control arm)	Mean age intervention: 64.8 y Mean age control: 66.5 y	6–64 months	Reduced progression to treatment at 2 years*, reduced PSA at 1 year*, no significant difference in PSA at 2 years inhibited LNCaP cell growth*, increased intake of protective dietary factors and decreased intake of pathogenic dietary factors*, increase in IGFBP-1*, reduced total cholesterol/HDL/LDL at 2 years,* reduced weight at 1 year*, nutrient adequacy except vitamin D, positive reactions to intervention, improved physical health-related QOL, reduced perceived stress, high lifestyle adherence at 2 years	36
RCT of diet/lifestyle intervention for recurrent PCa after definitive primary treatment29,30 (n=17 in intervention arm, n=19 in control arm)	Mean age: 69.1 y	3 months after intervention	Reduced intake of saturated fat* animal protein*; Increased intake of vegetable protein*, total and insoluble fiber*, lycopene,* total carotenoids*, total phytoestrogens*; increased exercise*, improved QOL*; increased mean PSA doubling time	28
Non-randomized trial of plant-based diet/lifestyle intervention in recurrent PCa after surgery31 (n=10, single arm)	Age range: 56–78 y	4 months	Decrease in the slope of PSA change*, increase in median PSA doubling time	32
Non-randomized trial of plant-based diet/lifestyle intervention in recurrent PCa after primary therapy32–34 (n=14, single arm)	Median age: 70 y	6 months	Decrease in the rate of PSA rise*, increase in median PSA doubling time, reduced waist to hip ratio*, increased SHBG*, increase in median intake of whole grains*	36

RCT=randomized control trial, PCa=prostate cancer, AS=active surveillance, QOL=quality of life, SHBG=sex hormone binding globulin, IGFBP=insulin-like growth factor binding protein, BMI=body mass index

^*denotes statistical significance where test of significance performed

^aQuality assessment was performed using the criteria by Hawker et al.¹⁹ The checklist contains 9 questions (abstract/title, background/aims, methods, sampling strategy, data analysis description, ethics/bias, results, generalizability of results, and implications of findings). Each component is scored from “very poor” (1) to “good” (4), with 36 being the highest total score. If a single study corresponded to multiple publications, information from all the publications were collated to tabulate a composite quality score.

Table 2.

Observational studies (n=11)

Study description, eligible population, sample size	Reported Age	Follow-up	Key findings	Quality scorea
EPIC-OXFORD/Oxford Vegetarian study, Prospective study of men and women aged 20–89 in UK (n=61,566–64,234 across the different publications)35–38	Median age 46 y	Mean up to 14.9 y	Significantly higher standardized incidence ratio of PCa among nonvegetarians*, No significant difference in standardized mortality ratio, no significant reduction in relative risk of PCa among vegetarians and/or vegans	36
Prospective study of male and female participants in UK Biobank (n=217,937 men)39	Mean age 53.0–56.9 y by diet group	Mean 11.4 y	Significantly reduced risk of PCa among vegetarians/vegans*	36
Prospective study of California Seventh-Day Adventists (n=13, 857 men)40,41	Mean age 53.1 y	6 y	Significantly reduced risk of PCa among vegetarians*, No significant reduction in risk of PCa among those with vegetarian upbringing	33
Prospective study of US and Canadian Seventh-Day Adventists; Adventist-Health Study 2 42,43 (n=27,188–27,934 across the different publications)	Median age at overall PCa diagnosis 66 y, median age at advanced/high grade diagnosis 68 y, Age range 30–104 y	Mean 7.8–7.9 y	Significantly reduced risk of PCa among vegans*, No significant reduction in risk of advanced/high grade PCa for vegetarians/vegans	36
Collaborative analysis of 5 prospective studies (Adventist Mortality study, the Health Food Shoppers Study, the Adventist Health Study, the Heidelberg study, and the Oxford Vegetarian study)44,45 (n=19,406 nonvegetarian men, n=10,742 vegetarian men)	Range of median age nonvegetarians: 34–52 y Range of median age of vegetarians: 33–51 y	Mean 10.6 y	No significant difference in mortality from PCa	36
Prospective study of men and women in the Netherlands (n=1,133 vegetarians, n=9,949 non-vegetarians )46	Mean age PCa cases: 62.1 y	Mean 20.3 y	No significant difference in PCa risk or advanced PCa risk	20
Case-control study in North Indian Men47 (n=157 cases, n=170 BPH controls)	Mean age cases: 67.5 y, Mean age controls: 67.2 y	N/A	Significantly increased odds of PCa among nonvegetarians with A2 alleles*	26
Case-control study in North Indian Men48 (n=157 cases, n=170 BPH controls, n=370 healthy controls)	Mean age cases: 67.9 y, Mean age healthy controls: 66.3 y, Mean age BPH controls: 66.4 y	N/A	Significantly increased odds of PCa among non-vegetarian patients with variant Leu allele*	27
Case-control study in Taiwanese men49 (n=237 newly-diagnosed PCa cases, n=481 controls)	Mean age cases: 72.2 y, Mean age controls: 71.1 y	N/A	Significant protective association of moderate and greater consumption of local vegetarian food for PCa risk.* Stratified by BMI showed reduced risk among men with BMI ≤ 25 kg/m2* but not BMI >25 kg/m2	29
Prospective study of German men (n=521 strict vegetarians, n=337 moderate vegetarians )50	Not reported	11 y	No significant reduction in overall standardized mortality rate due to PCa or standardized mortality rate by time since study entry for vegetarians	32
Case-Control study in metastatic PCa51 (n=9 cases who adopted macrobiotic plant-based diet, n=9 controls)	Not reported	N/A	Improvements in bone lesions, longer progression-free survival, no significant difference in overall survival	29

PCa=prostate cancer, BMI=body mass index

^*denotes statistically significant finding

^aQuality assessment was performed using the criteria by Hawker et al.¹⁹ The checklist contains 9 questions (abstract/title, background/aims, methods, sampling strategy, data analysis description, ethics/bias, results, generalizability of results, and implications of findings). Each component is scored from “very poor” (1) to “good” (4), with 36 being the highest total score. If a single study corresponded to multiple publications, information from all the publications were collated to tabulate a composite quality score.

Interventional studies

The Gene Expression Modulation by Intervention with Nutrition and Lifestyle (GEMINAL)

The GEMINAL study was a prospective, single-arm pilot clinical intervention trial of 31 men with biopsy-proven low-risk prostate cancer on active surveillance.²⁰ All patients underwent a 3-month comprehensive lifestyle program consisting of a low-fat, whole-foods, plant-based diet, with 10% of ingested calories from fat, and daily nutritional supplementation (including soy in the form of tofu and a soy protein beverage, fish oil, vitamin E, selenium, and vitamin C). All food was provided for participants during the study. The program also consisted of regular aerobic exercise, daily stress management, and a weekly support group. Each participant underwent a prostate biopsy at baseline and at 3 months. One patient was referred for surgery at the end of the trial due to upgrading to Gleason score 3+4, while the remainder of the cohort continued active surveillance. Total and % free PSA did not change significantly during the trial.²⁰ Participants also experienced significant improvements in body mass index (BMI), blood pressure, lipid profile, waist circumference, and mental health parameters.^20,21

Prostate Cancer Lifestyle Trial (PCLT)

The PCLT was a randomized prospective trial of 93 men with biopsy-proven low risk-prostate cancer on active surveillance.²² 44 patients were randomized to the experimental group and 49 to the control group. The experimental group participated in a 1-year intensive lifestyle program of a very low-fat vegan diet rich in complex carbohydrates, with approximately 10% of ingested calories from fat. The diet also included nutritional supplementation (including soy in the form of tofu and a soy protein beverage, fish oil, vitamin E, selenium, vitamin C, an iron-free multivitamin, and tomato-based vegetable juice), regular aerobic exercise, daily stress management, and a weekly support group. Patients in the control group followed usual care.

No patients in the intervention group underwent prostate cancer treatment during the study period, while 6 patients in the control group underwent conventional treatment due to concern for disease progression.²² Patients in the intervention group experienced a significant improvement in PSA compared to control patients at 1 year.²² Improvement in PSA was correlated with the extent of plant-based dietary change, and animal protein consumption was associated with higher PSA.^22,23

At two years follow-up, significantly more control group patients (n = 13) than experimental group patients (n =2) had undergone conventional prostate cancer treatment (p <.05). Changes in PSA did not differ significantly between the experimental and control groups at 2 years.²⁴

Additionally, LNCaP cell growth inhibition was significantly higher with serum from patients in the intervention group²². Dewell et al. examined the impact of the dietary intervention on oncogenic insulin-like growth factor (IGF) and protective IGF binding proteins (IGFBPs).²³ Participants in both groups demonstrated increases in IGF-1 and IGFBP-3 at 1 year, while only intervention patients experienced an increase in IGFBP-1. IGFBP-1 levels were correlated with vegetable protein consumption.²³

At one year, patients in the intervention group achieved multiple positive dietary changes including significantly less intake of pathogenic dietary factors (such as saturated fat and cholesterol) and increased intake of protective dietary factors (such as fiber, total and vegetable protein, and lycopene).²⁵ Nutritional adequacy was achieved for >20 nutrients including vitamin B12, with the exception of vitamin D, which required supplementation.²⁶ Adherence in the intervention group was 95% at 2 years, and there were no known adverse events attributable to the intervention.²⁴ Participants in the intervention group experienced a significant reduction in weight at 1 year compared to the control group.²² Intervention participants also experienced significant reductions in total cholesterol, high-density lipoprotein (HDL), and low-density lipoprotein (LDL) at 2 years, while control group participants did not.²⁴ There was no significant difference in serum testosterone, fasting insulin, or C-reactive protein between the groups,.^22,23

Additionally, patients in the intervention arm reported improvements in optimism, hope, fighting spirit, and physical health-related QOL (HRQOL), as well as reduced perceived stress.^27,28 Participants in the control group who made lifestyle changes also experienced improved physical HRQOL and reduced perceived stress.²⁸ Overall, Frattaroli et al. reported no difference in QOL among untreated men in either arm of the study; however, fewer controls (n=36) were free from treatment compared to the intervention group (n=42).²⁴

Additional interventional studies

In a randomized pilot study of 36 men with PSA relapse after primary therapy for prostate cancer, patients were randomized to an 11-week intervention consisting of increased consumption of plant-based foods and oily fish, reducing or eliminating land-animal–based protein (including dairy), and mindfulness training versus usual care.²⁹ Mean PSA doubling time (in months) increased in the intervention arm (intervention group—baseline 21.5 (95% CI 12.8–66.8), 3-month 58.5 (95% CI 14.7-infinity) vs. control group—baseline 18.4 (95% CI 12.1–39.2), 3-month 18.7 (95% CI 10.6–81.0)). The intervention also led to a significant increase in vegetable protein consumption, reduction in animal protein consumption, reduction in saturated fat intake, and improvement in QOL. There was a significant relationship between increased vegetable protein consumption, other favorable nutritional changes, and frequency of mindfulness practice. This study has been reported in multiple publications,³⁰ including additional positive dietary changes from the intervention.

Saxe et al. initially reported a 4-month nonrandomized interventional trial of the impact of a plant-based diet combined with a mindfulness and stress reduction program on PSA kinetics in 10 men with prostate cancer and biochemical recurrence.³¹ Rates of PSA change were significantly decreased in 80% of men post-intervention. Estimated median PSA doubling time increased from 6.5 to 17.7 months.

In 2006, Saxe et al. expanded on this in the University of California, San Diego Healthy Men Study in which 14 men with biochemically recurrent prostate cancer underwent a plant-based diet and stress reduction intervention over a 6-month period.³² At the end of the study period, 9 of 10 patients showed a decrease in the rate of rise in PSA, and 4 had negative PSA rates indicating overall reductions in PSA levels. Median PSA doubling time increased in this study from 11.9 to 112.3 months post-intervention. Later work with this data suggested changes in central adiposity and hormonal milieu that corresponded temporally with the PSA rate decrease.³³ Nguyen et al. reported on adherence to the plant-based diet with respect to PSA changes.³⁴ They noted a significant decrease in the median rate of PSA rise during the first 3 months of the study, when increased intake of whole grains and vegetables was most pronounced (−0.002 vs. 0.059 in the pre-study period, p<0.01).

Observational studies

EPIC-OXFORD and Oxford Vegetarian cohorts

Key et al. recruited and surveyed 61,566 to 64,234 men and women ages 20–89 through the UK in 1990s. Cancer incidence was found through nationwide cancer registries.³⁵ Patients were recruited by general practitioners as well as mailing surveys through the use of Vegetarian Society of the UK, the Vegan Society, and Oxford Vegetarian study. Participants were categorized into 1) meat eaters, 2) fish eaters, 3) lacto-ovo-vegetarians, and 4) vegans. However, due to the small number of cancers in vegans, the vegan and vegetarian groups were combined in the analysis. The standardized incidence ratio (SIR) for prostate cancer was significantly higher in the overall study population (125%, 95% CI 110–142) and among non-vegetarians (129%, 95% CI 112–149), but not among vegetarians (106%, 95% CI 77–144). In multivariable analysis, the adjusted incidence rate ratio for overall prostate cancer was 0.90 (95% CI 0.61–1.33) among vegetarians compared to non-vegetarians. In an analysis of causes of mortality among 64,234 participants in the EPIC-Oxford cohort, mortality among participants was compared to contemporary mortality data in England and Wales to yield standardized mortality ratios (SMRs). There were 50 deaths from prostate cancer among nonvegetarians [SMR 92 (95% CI: 68, 121)] and 12 deaths from prostate cancer among vegetarians/vegans [SMR 75 (95% CI: 39, 130)].³⁶

Key et al. published a combined analysis of overall cancer and specific cancer types in the EPIC-Oxford and Oxford Vegetarian studies.³⁷ For prostate cancer, the relative risk was 0.76 (0.52–1.11) for fish eaters and 0.84 (0.66–1.07) for the combined vegetarian/vegan group compared to meat eaters (p=0.19). Results were similar after additional adjustment for BMI. After further stratification of dietary patterns, the relative risk was 0.87 (0.68–1.12) for vegetarians and 0.62 (0.31–1.22) for vegans compared to meat eaters; however, this analysis was limited by the small number of prostate cancer cases among vegans (n=9). In an additional pooled analysis of cancer incidence among the EPIC-Oxford and Oxford Vegetarian cohorts, the authors found that the relative risk of incident prostate cancer among 5,489 vegetarian men compared to 8,451 men who were meat eaters did not differ significantly [0.87 (0.64–1.18)].³⁸

Other observational and combination studies

Watling et al. reported on the risk of incident prostate cancer among 217, 937 male participants in the UK Biobank.³⁹ Using a food frequency questionnaire, participants were classified as regular meat-eaters, low meat-eaters, fish-eaters, and vegetarians/vegans. On multivariate analysis, vegetarians/vegans had a significantly reduced risk of prostate cancer compared to regular meat-eaters (HR=0.69, 95% CI: 0.54–0.89). This protective association remained unchanged after further adjusting for BMI and was not mediated by IGF-1 or free testosterone levels.

Fraser reported on the association of diet with chronic diseases in a cohort of 13,857 California Seventh-day Adventist men mailed a detailed lifestyle questionnaire in 1976.⁴⁰ Using a 51-item food frequency questionnaire, participants were classified as vegetarian (those who ate no fish, poultry, or meat, 29.5%), semi-vegetarian (those who ate fish and poultry but <1 time/wk, 21.2%), and non-vegetarian (the remaining subjects, 49.2%). Compared to vegetarians, non-vegetarians had a significant 54% greater risk of incident total prostate cancer. Men who had at least one Seventh-day Adventist parent and a vegetarian lifestyle in the home during the childhood and teenage years experienced a nonsignificant reduction in the relative risk of prostate cancer.⁴¹

In the Adventist Health Study-2, rates of incident prostate cancer were lower among vegans.^42,43 Tantamango-Bartley reported that on multivariate analysis adjusting for race, family history, education, screening and calorie intake, vegan diets had a significant protective association with prostate cancer risk (HR 0.65, 95% CI 0.49–0.85). ⁴³ Conversely, there was no significant difference in prostate cancer risk for lacto-ovo-vegetarian, pesco-vegetarian, and semi-vegetarian participants compared to non-vegetarians. There was no significant association between a plant-based dietary pattern and the development of advanced or high-grade prostate cancer. In a subsequent study of lycopene in this cohort, vegans had a significantly lower risk of prostate cancer in an age-adjusted model.⁴²

Key et al. also performed a collaborative analysis of 5 prospective cohort studies.^44,45 When adjusting for age, sex, and smoking, there was no significant difference in mortality due to prostate cancer (death rate ratio [DRR] 0.91, 95% CI 0.60–1.39), and no significant difference in prostate cancer mortality by age at death (<65 years, 65–79 years, 80–89 years) for vegetarians compared to nonvegetarians.

Gilsing et al. reported on dietary patterns and prostate cancer risk among men as part of the larger Netherlands Cohort Study—Meat Investigation Cohort (NLCS-MIC).⁴⁶ All participants in the study self-reported and completed a baseline food frequency questionnaire. Overall, there were 19 prostate cancer cases among 1,133 vegetarians and 377 prostate cancer cases among 9, 949 non-vegetarians. On multivariable analysis adjusting for age, total caloric intake, smoking, alcohol consumption, BMI, physical activity, education, and family history of prostate cancer there was no significant relationship between vegetarian versus non-vegetarian diet and overall prostate cancer risk. Similarly, there was no significant difference in the risk of advanced prostate cancer between vegetarians and non-vegetarians.

Sobti et al. evaluated polymorphisms in the CYP17 gene, which encodes an enzyme involved in sex steroid synthesis, among Indian men with prostate cancer (n=157) and BPH (n=170) in a case-control study.⁴⁷ Overall, the majority of patients with prostate cancer had metastatic disease (81%), and the study suggested patients with a non-vegetarian diet had a nonsignificant increased odds of prostate cancer (OR 1.43, 95%CI 0.93–2.21). Upon stratifying subsets by allelic distributions, it appeared that the odds of prostate cancer was higher for men with A2 alleles if they had a non-vegetarian diet. For example, nonvegetarians with A2/A2 alleles had a nearly 4-fold increased risk of prostate cancer (OR 4.30 [1.27–14.53]). In 2011, Sobti et al. reported a similar case-control study evaluating Ser217Leu polymorphisms in HPC2/ELAC2, an important cancer-susceptibility gene.⁴⁸ They used the same cases of Indian men with prostate cancer and controls with BPH plus a comparison group of healthy controls (n=370). The findings suggested the proportion of participants with a vegetarian diet was slightly higher among BPH (54%) and healthy controls (51%) compared to prostate cancer cases (45%), but no statistical comparisons were performed. Non-vegetarian patients with the variant Leu allele had greater odds of prostate cancer relative to vegetarians with the Ser/Ser HPC2/ELAC2 wild type allele in the model with healthy controls (OR 2.11, 95%CI 1.08–4.12) or BPH controls (OR 2.49, 95%CI 1.27–4.90).

Using a case-control design, Chen et al. reported on 237 newly-identified prostate cancer cases and 481 controls in Taiwan.⁴⁹ The study assessed intake of local vegetarian food, which could include milk and/or eggs, eaten in the past 10 years. Moderate and greater consumption of local vegetarian food had a significant protective association (OR=0.67; 95% CI: 0.47, 0.94) for prostate cancer, compared to those who consumed this to a lesser amount or not at all. Upon stratification, a significant protective association was reported for men with BMI ≤25 kg/m² (OR=0.50, 95% CI=0.32, 0.76), but not for men with BMI >25 kg/m² (OR=1.25; 95% CI: 0.66, 2.37).

Chang-Claude et al. reported mortality patterns among 521 strict vegetarian and 337 moderate vegetarian men who occasionally ate meat and fish over an 11-year follow-up period.⁵⁰ Compared to the general contemporary population in Germany, vegetarians had a nearly 60% lower mortality due to prostate cancer although it was not statistically significant [SMR 0.41 (0.09–1.20)].⁵⁰

In a case-control study of 9 patients with metastatic prostate cancer who adopted a modified macrobiotic plant-based diet to varying degrees in addition to conventional treatment and 9 matched controls from tumor registries, 3 patients in the plant-based diet group experienced long-term improvement or regression of bone lesions compared to no patients in the control group.⁵¹ Additionally, all patients in the plant-based group had long-term survival free of progression prior to deviations from the recommended diet compared to 61% in the control group. Overall mean and median survival was 177 months and 228 months, respectively, in the plant-based diet group compared to 91 months and 72 months in the control group [OR 1.60 (−2.39–3.33)].

Discussion

Our systematic review identified 32 publications on plant-based diets and incidence or outcomes of prostate cancer. Observational studies primarily focused on incidence of total prostate cancer, showing either a lower risk of prostate cancer for men on a plant-based diet or no significant difference. Interventional studies focused on two main patient populations, those with low-risk prostate cancer on active surveillance and those with biochemical recurrence after initial therapy. These studies showed generally favorable results of lifestyle modifications incorporating a plant-based diet with oncologic outcomes such as PSA kinetics and need for additional treatment, as well as improvements in nutrition and general health.

Our findings align with and build upon previous literature on the topic. For example, Shin et al. performed a literature review of prostate cancer risk and dairy products, milk, yogurt, vegetarian or vegan diet, lacto-ovo-vegetarian, semi-vegetarian, pesco-vegetarian and plant-based diet from 2006 to 2017.⁸ In addition to reporting on individual plant and animal foods, their review identified 5 prospective cohort studies on vegetarian diet which found either a decreased or unchanged risk of prostate cancer. By contrast, among studies with vegan diet, all were associated with decreased prostate cancer risk. Unlike our study, this study only examined prostate cancer risk and did not evaluate other endpoints such as oncologic outcomes, QOL, or overall health among men with prostate cancer.

Exploring plant-based dietary patterns is particularly timely as a potential solution to current global issues such as emerging infectious diseases and climate change. The Centers for Disease Control estimates 3 of every 4 emerging infectious diseases in people come from animals.⁵² Antibiotic resistance is another major public health threat, and there is increasing concern about the contribution from use of antimicrobial drugs in food-producing animals.⁵³

In addition to considerations for human health, animal agriculture is also a significant contributor to global warming, pollution and extinction of wildlife.^54,55 A recent study found replacing half of all animal-based foods (e.g., dairy, meat and fish) would save 224 million metric tons of carbon per year, equivalent to 47.5 million passenger vehicles.⁵⁶ Given the strong rationale for a global shift toward more plant-based diets, the implications for prostate cancer are an important question.

Overall, the findings are encouraging in light of the many benefits of plant-based diets for overall health, environmental sustainability, and animal welfare and are in line with health and environmental recommendations to increase plant-based dietary consumption. Notably, other recent studies show that higher consumption of a healthy plant-based dietary index was associated with a significantly lower risk of having an elevated PSA,⁵⁷ and lower risk of total and lethal prostate cancer among men age <65.⁵⁸ While these studies included omnivorous participants consuming varying amounts of plant- and animal-based foods, they provide further supporting evidence about the health benefits of plant-based foods.

Prostate cancer diagnosis is a teachable moment, and previous studies have shown that providing counseling about nutrition can lead to sustained improvements in dietary habits among men with prostate cancer. For example, participants reported positive experiences with lifestyle changes, and lifestyle change adherence was 95% at 2-year follow-up in the intervention group of the Prostate Cancer Lifestyle Trial. ^24,27

In the Men’s Eating and Healthy Living (MEAL) Study, an intervention promoting consumption of more vegetables during active surveillance for prostate cancer led to sustained increases in intake of carotenoid, cruciferous-rich and leafy green vegetable intake through 2 years of follow-up.¹⁶ This study did not attempt to eliminate animal-based foods, and meat consumption did not differ between the intervention and control groups during follow-up. While not reporting on a plant-based dietary pattern, this study demonstrates that patients with prostate cancer are receptive to sustained lifestyle modification.

Providing guidance about healthy dietary patterns is a key component of holistic survivorship care.^59,60 Our results show that plant-based diets have the potential to improve prostate cancer outcomes in addition to well-established advantages for cardiovascular health and QOL.

Limitations and gaps in the published literature on this topic warrant further discussion. First, we did not identify any studies of the impact of plant-based diets in caregivers of patients with prostate cancer. Partners were included in some of the interventional studies, and whether lifestyle modification also led to positive health outcomes for family members is unknown and an important direction for future research. Most observational studies focused on incidence of total prostate cancer, rather than clinically significant disease, and there is potential for residual confounding from unmeasured factors (such as family history) or differences in access to care. Interventional studies showed a positive impact of plant-based diets among men with prostate cancer in multiple clinical scenarios, but most had a small sample size and short follow-up. Additionally, the interventions varied in their inclusion of exercise programs, mindfulness or stress-reduction practices, support services, and dietary supplements. The individual impact of the dietary changes in the interventional studies cannot be separated from the other components, which prevents the estimation of the diet-specific effect of the intervention and may affect the reported results. Some supplements (such as vitamin B12 or vitamin D) may be important for individuals following a plant-based diet. However, differential use of other supplements (such as selenium, Vitamin E and fish oil in PCLT) may have affected results. For example, selenium use after diagnosis of non-metastatic prostate cancer has been associated with adverse oncologic outcomes, such as increased prostate cancer mortality,⁶¹ and it is unknown whether a greater benefit from plant-based diets would have been observed without the inclusion of these supplements in the PCLT intervention.

Another important limitation of our study is not all plant-based diets are equivalent. In particular, dairy has been linked with an increased risk of prostate cancer,⁶² and is included in vegetarian diets but excluded from vegan diets. Meanwhile, other pesco-vegetarian diets excluded meat, chicken and dairy but included fish. These limitations notwithstanding, this article provides a comprehensive assessment of the current literature on plant-based diets in prostate cancer, which is important for patient counseling and identifies critical gaps for future research. In summary, interventional studies on the impact of plant-based diets on prostate cancer demonstrate improved oncologic and general health outcomes among men on active surveillance and with biochemical recurrence. Observational studies among at-risk men showed either reduced or equivalent risk of prostate cancer among those on plant-based diets compared to non-vegetarians. Notably, no study found an increased risk of prostate cancer with a plant-based diet. Therefore, men should be counseled about the potential benefits of plant-based diets for improving overall and prostate health.