Comment by barbazoo
3 days ago
> The correct (and scientifically valid) thing to do is to only take action when there is actual evidence and proof of harm being done.
Because we're talking about food I would actually like to see the opposite. Provide peer reviewed, gold standard studies showing that what you want to put in food is in fact safe.
There is no such thing as proving something "safe". Go back and re-read the parent comment. The important point you are missing is that basically anything can be "linked" to cancer, and if you adopt the argument you are making, there would be nothing left.
Proving something safe is logically equivalent to proving that it is not unsafe, which is the same thing as proving a negative, which cannot be done. I cannot prove there is not a teapot circling Mars, and I cannot prove that even the most inert ingredient, at some dose, will not harm you.
Anyone who has lived in California knows this absurdity more intuitively than most people, because California's stupid laws adopt the logic you are proposing, and basically everything in daily life is labeled as cancer-causing.
A lot of folks in child comments are echoing your sentiment that something “cannot be proved safe”. Your argument that proving something is “not unsafe” is proving a negative is fallacious; the same can literally be said about anything (proving something is X is the same as proving it is not not-X). Proving drugs are safe and effective is literally one of the jobs of the FDA. If you do not believe that is possible, then we may as well tear down the entire drug regulatory apparatus. I imagine you and many other commenters will sing a different tune when posed with that suggestion.
So, let’s stop pretending it’s not possible. We require drug companies show their products are safe and efficacious, and there is both a scientific and a legal framework by which we do this. Let’s debate whether or not the same framework should be applied to food additives (I would argue it should) rather than claim it is not possible.
We require drug companies show their products are safe and efficacious, and there is both a scientific and a legal framework by which we do this.
We don’t do this.
What the FDA requires is acceptable safety in light of the benefit provided.
The FDA approves highly toxic drugs all the time. Including ones with the risk of death. I don’t think anyone would call chemotherapy “safe”.
"Safe" for the FDA means the benefits outweigh the potential risks, not safe in absolute terms.
If the FDA actually required every drug to be proven safe at any dose for everyone, we'd have no modern drugs.
> There is no such thing as proving something "safe".
Is that not what NCAP does?
Or what NTSB and FAA do with aviation?
You can prove that some things are safe. Does not mean infallible, means safe.
>There is no such thing as proving something "safe". Go back and re-read the parent comment. The important point you are missing is that basically anything can be "linked" to cancer, and if you adopt the argument you are making, there would be nothing left.
Really? You have some studies linking wheat and whole grains to cancer? And I don't mean wheat crops sprayed with glyphosate, just straight up wheat? Raspberries? Strawberries?
The reality is, very little of the actual natural food in our food chain is directly linked to cancer. All the additives we pile on top, on the other hand, are.
I would argue if we can't show a direct benefit to the consumer, it shouldn't be in the food chain. So, what is the direct benefit to a human consuming red-5? "It looks better on store shelves" isn't a direct benefit.
A shelf stabilizer? Sure, plenty of instances that makes a lot of sense. Food coloring that happens to be cheaper than natural alternatives? Just... no.
Yes, whole grains cause cancer if you make them into bread and toast the bread. The evidence is much stronger than for Red dye No. 3.
https://www.fda.gov/food/process-contaminants-food/acrylamid...
Strawberries are also linked to cancer, because they contain sugar.
https://aacrjournals.org/cancerres/article/62/15/4339/508983...
Almost all natural foods are linked to cancer. The important question is, how large is the risk?
Dark toast is obviously much riskier than Red dye No. 3. We should think about that when we consider what to ban.
5 replies →
> The reality is, very little of the actual natural food in our food chain is directly linked to cancer.
Natural things aren't inherently safer. Are alcohol and red meat both considered natural? Alcohol is a group 1 carcinogen (same as tobacco and asbestos) and red meat is group 2A (probably linked to cancer). A cursory search shows some studies linking fish consumption to cancer, though I have no idea how accurate those studies are.
9 replies →
https://en.wikipedia.org/wiki/Appeal_to_nature
There is no way to establish a food as "safe".
Health outcomes are noisy, especially if taken over a long time. Peer reviewed studies are often flawed in various ways and most scientific studies lack the statistical power to be inconclusive.
The fear based approach to human diets can not work. We have to accept risks in our lives if we want to eat at all.
"lack the statistical power to be inconclusive" is that right?
Natural food is not safe. Many natural foods that people eat, specifically plants, contain natural toxins and carcinogens that would probably not pass your threshold for "safe".[1]
[1] https://pubmed.ncbi.nlm.nih.gov/2217210/
I don't know about "peer-reviewed gold-standard studies," but what you've described is basically how the EU does it – what goes in food must be proven safe.
It's the opposite of the US approach, which is to ban (only) proven-harmful ingredients.
I don't expect US food-safety laws to become more strict in the next four years, but who knows, maybe the dead-worm guy will surprise us.
It’s impossible to prove a negative.
and that's also completely and totally irrelevant to the problem at hand.
Proofs don't apply in biology. Nothing in biology is a truly logical system that can be proved or disproved. That's true for chemistry and physics too- the only system where anything can be proved is math.
In science, instead we gather evidence and evaluate it, and often come to the conclusion that it is so unlikely something is dangerous (given the data) that we presume it's safe. People use the term "scientific proof", but I'm not aware of any in biology that would truly be classified as proof.
Why wouldn't you be able to prove a negative link?
It is logically and practically impossible to prove things to be untrue. We can only prove things that are true.
The thing we could prove is “no detectable increase versus control, in our test data”. There is no way to prove “x does not cause cancer” any more than there is a way to prove “x does not cause meteors” or “x does not cause spontaneous human resurrections” or “x does not cause humans to turn into unicorns”.
4 replies →
It is certainly possible to show that there is no positive correlation with a certain statistical significance. Pretending we're talking about such high standards as "no human future or presently alive could ever be harmed by any quantity of x" completely misses the point and borders on bad faith.
Let's set workable standards for when something can be called safe and enforce them.