Rung Five: cancer prevention and treatment, reviewed 30 October 2026
What actually helps, ranked by how well it's proven.
Every option here, from broccoli to radioligand therapy, sits on an evidence rung from 5 (established in large trials and guidelines) down to 1 (lab-only or untested). Ideas that were tested and failed are pulled out of the rankings and listed separately, so nothing is presented as more certain than it is. Our name comes from the top of the ladder: rung five means established in large trials and guidelines.
Start here
Six things that carry the most weight
The highest-graded steps you can take. They lower risk across many cancers at once, and most also help people already living with cancer. Tap any card for the evidence and sources.
Explore the site
Go deeper by topic
Every rating links to its studies. See all sources.
Diet
Eating to lower cancer risk
The strongest diet evidence is about overall patterns, not single superfoods. Four rules cover most of it; the food finder below rates individual foods.
Fill up on plants
Colorful vegetables and fruit, whole grains and beans. Together they probably lower risk of colorectal, mouth, throat and esophageal cancers.
Limit red and processed meat
Processed meat is a known cause of colorectal cancer, rising about 16% per 50 g a day. Red meat probably raises risk too.
Drink less alcohol
Linked to seven cancers. The American Cancer Society says not drinking is best.
Cut sugary drinks and highly processed food
The American Cancer Society advises limiting them, along with refined grains, mainly because of their link to weight gain.
Build your plate
- Half: vegetables and fruitGreens, crucifers, tomatoes, berries
- A quarter: whole grainsOats, barley, brown rice
- A quarter: plant or fish proteinBeans, lentils, tofu, oily fish
- Alongside: water, coffee or teaLet hot drinks cool first
- LimitProcessed and red meat, alcohol, sugary drinks
Food finder
Exercise
How much movement, and what it does
Exercise is one of the few self-care steps with randomized-trial proof in people who already have cancer. The targets below come from the American Cancer Society's prevention guideline and the 2019 international guideline for cancer survivors.
Check your week
Example week filled in. Change the numbers to your own; vigorous minutes count double.
What exercise is proven to do
| Goal | Evidence | What the studies found | What to do |
|---|
Myths and everyday exposures
What raises risk at home and work, and what doesn't
Claims about phones, deodorant, sugar and the like spread fast. Each one here gets a verdict on what the human evidence shows, plus an evidence rung from 1 to 5 for how sure that verdict is. They're the same rungs used everywhere on the site.
For hundreds more everyday claims, Harvard's Cancer FactFinder is an excellent companion resource.
Priority focus
The prostate cancer pathway, step by step
Prostate cancer is common and usually slow, so the biggest decisions are about finding the cancers that matter and not over-treating the ones that don't.
Decide about PSA testing, ideally in your mid-to-late 40s
The longest-running screening trial shows a lasting drop in prostate cancer deaths with regular PSA testing, along with real overdiagnosis. U.S. Task Force guidance (2018, under revision) calls screening an individual decision from 55 to 69 and advises against routine screening at 70 and older. The American Urological Association's 2026 guideline says doctors may offer a baseline PSA between 45 and 50. Men with a family history and Black men face higher risk and have more reason to start the conversation early.
If PSA is high, confirm before a biopsy
The 2026 AUA guideline advises repeating a newly raised PSA before further markers, imaging or biopsy. MRI before biopsy helps reduce detection of harmless cancers. Ask whether an MRI or a secondary blood or urine test fits your situation.
Low-risk cancer: active surveillance is a proven choice
In the ProtecT trial, men with PSA-detected localized prostate cancer were randomly assigned to monitoring, surgery or radiation. At 15 years about 97% in every group had not died of prostate cancer. Monitoring led to more progression, but not more deaths, and a quarter of monitored men never needed treatment. Today's MRI-based surveillance is more precise than the version tested.
While on surveillance: what men can do themselves
None of these has yet been shown to change survival in a large trial, but each is low-risk and several show measurable effects on the cancer's biology. In rough order of evidence: keep a healthy weight (strongly linked to advanced prostate cancer), exercise hard and regularly (lowered PSA and PSA velocity in a randomized trial), eat mostly plants (linked to 47% less progression in a large cohort), favor cooked tomatoes and unfermented soy, and ask your urologist before trying a high-omega-3, low-omega-6 diet with fish oil (lowered a tumor growth marker in a small trial).
Skip high-dose vitamin E and selenium pills: no benefit, and vitamin E raised risk. Watch dairy and calcium: the World Cancer Research Fund rates evidence of higher prostate risk as limited.
Higher-risk or advanced disease: ask about genetics and PSMA
Standard care is surgery or radiation, often with hormone therapy, and for advanced disease newer hormone drugs and chemotherapy. Three 2025–26 FDA approvals reward testing the tumor: the PSMA-targeted radioligand Pluvicto now covers metastatic hormone-sensitive as well as castration-resistant disease, capivasertib is approved for PTEN-deficient metastatic cancer, and talazoparib with enzalutamide is approved for HRR-mutated metastatic cancer that still responds to hormone therapy. Ask for a PSMA PET scan and tumor plus inherited genetic testing.
The register
Every prevention, treatment and support option we graded
Filter by what you need, then open any card for what the studies found, the catch, what to do, and the sources.
Ayurveda and integrative care
Traditional remedies, held to the same standard
We graded Ayurvedic herbs and practices exactly like everything else. Some have modest human evidence for easing treatment side effects or improving quality of life. None has been shown in human trials to treat or prevent cancer, and product quality is a real safety issue. Many other herbs, such as Sanjeevani and mangosteen, have only lab and animal results so far.
Buying Ayurvedic products safely
When researchers bought 193 Ayurvedic medicines online, about one in five contained lead, mercury or arsenic, and every one of those exceeded at least one daily safety limit. U.S.-made products were as likely to be contaminated as Indian-made ones. Rasa shastra formulas, which add metals on purpose, were more than twice as likely to contain them.
The risk shows up in people too: in a 2015 U.S. study of 115 people using Ayurvedic products, 40% had raised blood lead levels.
No U.S. state licenses Ayurvedic practitioners. In the UK, Cancer Research UK suggests choosing products registered under the Traditional Herbal Registration scheme.
Before taking any herb during treatment, tell your oncology team. Herbs can interact with chemotherapy, hormone therapy and surgery.
Prefer single-herb products with independent third-party testing. Avoid metal- or mineral-based formulas entirely, and products such as Carctol that claim to cure cancer.
Skip panchakarma cleansing during treatment. Its purging, enemas and bloodletting can cause infection and mineral imbalances, especially with low blood counts.
Yoga and meditation, from the same tradition, have better support for symptom relief than the herbs do.
What not to do
Proven to backfire, and tested but didn't hold up
Kept out of the rankings. The left column has evidence of harm. The right lists popular ideas that were put to a fair test and failed.
Evidence of harm
- Replacing standard treatment with alternative medicinePatients with curable cancers who chose alternative medicine alone had 2.5 times the risk of death within five years; more than 5 times for breast cancer and 4 times for colorectal cancer.
- High-dose vitamin E (400 IU a day)Raised prostate cancer risk by 17% in the 35,000-man SELECT trial.
- Beta-carotene pills for smokersTwo large trials (ATBC and CARET) found more lung cancer, not less, in smokers taking beta-carotene supplements. Food sources are fine.
- Apricot seeds, laetrile or amygdalinRelease cyanide in the body. A Cochrane review found no reliable evidence of benefit and called the balance of risk clearly negative.
- Metal-containing Ayurvedic formulasSome contained lead or mercury at levels 100 to 10,000 times acceptable limits.
- Ashwagandha with existing liver diseaseCase series describe herb-induced liver injury; three patients with prior liver disease died of liver failure.
- Giloy (Tinospora) with liver diseaseLinked to liver injury in 43 people at 13 Indian hospitals; four with existing liver disease died.
- Antioxidant pills during chemotherapyIn 1,134 women in a breast cancer chemotherapy trial, taking vitamins A, C or E, carotenoids or CoQ10 both before and during treatment was linked to about 41% higher recurrence risk (borderline significance). Iron and B12 were also linked to worse outcomes; multivitamins were not.
- St. John's wort during cancer treatmentIn a small crossover study it cut levels of the active form of the chemotherapy drug irinotecan by 42%. The National Cancer Institute warns it can make some cancer drugs work less well.
- Black salve and other corrosive skin pastesSold to "draw out" skin cancer, these bloodroot and zinc chloride pastes destroy healthy tissue, leave scars and can leave cancer behind.
- Gerson therapy and coffee enemasReviews found no anticancer effect. Repeated coffee enemas have caused deaths from fluid and mineral imbalance, and serious infections.
- Panchakarma purging and bloodletting during treatmentAyurvedic enemas, laxatives and induced vomiting can cause infection and dangerous mineral imbalances, and bloodletting is risky when blood counts are low.
Tested, didn't hold up
- Pomegranate extract for rising PSAA large placebo-controlled trial found no difference in PSA doubling time, because PSA slowed in the placebo group too.
- Selenium pills to prevent prostate cancerNo benefit in SELECT.
- Green tea extract (EGCG pills) for men at high riskA one-year randomized trial did not significantly reduce prostate cancer diagnoses. Drinking green tea is fine.
- Metformin added to hormone therapy in non-diabetic menNo overall survival benefit in the STAMPEDE trial of about 1,900 men, though it eased metabolic side effects.
- Extra vegetables alone to slow early prostate cancerThe MEAL trial raised intake to 7 or more servings a day for two years; progression was unchanged.
- Large monthly or yearly vitamin D dosesInfrequent big doses did not lower cancer deaths in pooled trials; daily dosing did.
How we score
What each rung means
Every rating on Rung Five answers one question: how strong is the human evidence? The answer is a rung from 1 to 5, and rung five is the strongest. It reflects the quality of the studies, not how popular, natural or expensive something is, and the same substance can sit on different rungs for different uses. Rungs measure certainty only, so there are no negative rungs. Whether something helps or harms is shown separately, as a verdict.
| Rung | Name | What it takes | How to read it |
|---|
Reading ratings on foods, myths and exposures
Treatments and habits elsewhere on the site are all things that might help, so a rung is all they need. Foods and everyday exposures can lower risk, raise it or make no difference, so each one gets two separate labels.
1Verdict
What the studies found: lowers risk, raises risk, no link, and so on. It's written in words, with green for lower risk and red for higher.
2Evidence rung
How sure we can be of that verdict, on the same 1 to 5 scale in the table above. On foods and exposures it's shown in grey, so color only ever means direction.
Three examples
Every verdict we use
Sources