We keep discovering new things, so don't cut out a whole food group
This is the strongest version of the health argument and the answer has to concede a great deal. Nutrition science really is unreliable and we really are still finding things. But the uncertainty runs in both directions, and a precautionary principle that only ever protects what people already do is not a precautionary principle.
Nutrition science is young, heavily confounded and reverses itself regularly. Compounds in animal foods are still being characterised, and a fully plant-based diet at population scale is new in a way that eating animals is not. Under real uncertainty, removing an entire category of food is a large bet with no hedge, and "nothing has gone wrong so far" is thin reassurance from a field that keeps discovering things it had missed.
This deserves to be taken seriously, because the premise is true. Nutrition research is mostly observational, confounding is severe, effect sizes are small and reversals are common. Anyone who tells you the science here is settled is overselling it, and the people making this argument are often better calibrated about the evidence than the people dismissing them.
The difficulty is that the argument is symmetric and is almost never applied symmetrically. If there are compounds in animal foods we have not characterised, there are also mechanisms in animal foods we have not finished characterising on the other side: Neu5Gc, TMAO, heme iron and the IGF-1 pathway are all live questions, and none of them is settled either. Uncertainty is a reason for caution about confident claims in general. It is not evidence for one diet.
The track record point is worth separating out. It is true that humans have eaten animals throughout history, and true that this tells you something: a practice sustained for millennia is unlikely to be acutely harmful. It is not evidence of optimality, and a great many things with long track records turned out to carry costs that were simply invisible at the time because people died of other things first.
Where the argument has real force is at the margins, and it should be conceded there. For pregnancy, infancy, old age and anyone with absorption problems, being conservative and monitoring properly is sensible, and a vegan diet in those cases needs more attention rather than less. That is an argument for care. It is not an argument for not asking the question.
What the argument actually does, structurally, is shift the burden so that any uncertainty counts in favour of carrying on. Used consistently it would freeze every dietary change anyone has ever made, including the ones this argument assumes are fine.
- We are still discovering compounds we did not know about
- Nutrition science keeps being wrong, so why risk it
- Animal foods have a track record that plant-based diets do not
- Why remove something when we do not fully understand it
Different wording, same argument. The response above answers all of them.
Treating the absence of proof against a claim as proof of it, or the absence of proof for a claim as proof against it.
How it applies here: Treats the possibility of undiscovered beneficial compounds as a reason to believe they exist and matter, when not yet knowing something is not evidence either way.
Demanding that others disprove a claim rather than supporting it. The burden sits with whoever makes the assertion.
How it applies here: Places the burden entirely on the change, so that uncertainty always counts in favour of current habits. Applied consistently it would rule out every dietary change, including the ones the arguer already accepts.