That's theoretically true, but wouldn't the fact that fasting is low-risk in healthy populations indicate that the risk is more likely to be low for populations

That's theoretically true, but wouldn't the fact that fasting is low-risk in healthy populations indicate that the risk is more likely to be low for populations with cancer? Also, from your example if you're one of the 100, then fasting is in fact advisable. I understand that purely for legal reasons we can't advise doctors to recommend fasting as a general guideline without gold-plated evidence. However it should be noted that just because the risk has not yet been perfectly quantified does not mean it is entirely unknown. For my part if I ever contract any sort of cancer, I will never reject a potential treatment on the grounds of "well we know it's low-risk in healthy populations, and given what we know of the biology it seems like there's some potential effectiveness. But we don't have a gold-plated-chocolate-coated mass population double-blind study for your specific cancer yet, so I'd advise against trying it". Taking minor, well-thought-out risks is the least one should do when fighting for one's life. It's a good thing doctors aren't financial advisors, or they'd tell you to avoid the S&P 500 index fund and stick to your 2% savings account.