[flagrant_taco]

Proving a link between cancer and one specific environmental factor is the physical impossibility you're looking for.

Epidemiological studies tracking humans over time aren't controlled and will at best show a correlation worth studying, lab animal studies aren't great analogs given all the confounding factors and environmental differences with humans in real life, and unless the cancer causing agent acts quickly a controlled human study is likely impossible as controlling variables over years simply won't happen at scale.

Also the idea that visible light should be banned because radiation is an asinine argument meant to just badger without actually making a point.

[boxed]

> Proving a link between cancer and one specific environmental factor is the physical impossibility you're looking for.

But.. that happens all the time? The link between excessive solar exposure ("tanning") and cancer is extremely well documented.

[flagrant_taco]

Even that is a strong correlation though, and a simplified one at that. Context matters even for sun exposure, if solar radiation was a direct cause of cancer you should reliably see consistent skin cancer rates among people with similar complexion and exposure rates.

I've seen some really compelling research looking into diet changes that may impact risk of skin cancer, specifically an increase in saturated fats helping to reduce risk. The idea is that saturated fats are envloved in the process of your body turning UV into vitamin D. With more saturated fats in your system the UV has effectively has a job and is processed by the body.

My point is simply that the body is complex, cancer generally develops over years, and trying to narrow down a common link strong enough to claim causation is tricky if not impossible.

[boxed]

> if solar radiation was a direct cause of cancer you should reliably see consistent skin cancer rates among people with similar complexion and exposure rates.

...which you do.

> My point is simply that the body is complex, cancer generally develops over years, and trying to narrow down a common link strong enough to claim causation is tricky if not impossible.

That sounds like weasel language to try to muddy the waters to me. If it's tricky or impossible to nail down the effect must be small, and probably not worth caring much about.