Every One of 212 Plant Based Products Tested in Britain Contained a Fungal Toxin

The first is that current controls appear to be working. Every product came in under guideline levels, in a market where the guideline exists and is enforced. That is a pass, not a warning, and the authors point at the standards of the UK food industry when explaining it.

The second is about cumulative exposure. Guideline levels are set per food, on assumptions about how much of that food a person eats. Someone eating an entirely plant based diet is drawing a larger share of their total intake from the ingredient classes where these compounds concentrate: oats at breakfast, a soy drink in coffee, legume protein at lunch, a nut based product in the evening. Each item is compliant. The question the researchers raise is whether the sum of compliant items was ever the thing being assessed.

That is a real gap in how food safety limits are constructed, and it applies well beyond this category. It is also, importantly, a question rather than a finding. The study measured products. It did not measure people, and it did not measure harm.

What it does not justify

Going back to meat, buying an expensive detox product, or avoiding oat milk.

Nothing in this survey establishes that anybody was harmed, and the effect sizes at these concentrations are not known to be meaningful in a normal diet. Repeated low level exposure adding up over time is a plausible mechanism supported by earlier research, and plausible mechanisms are where nutrition science goes wrong most often. We wrote about that failure mode at length in the argument over ultra processed food, where a real signal, a contested definition and a marketing opportunity have been tangled together for years.

The practical version

Vary the base ingredient, which is advice that would be good even if this study had never been published.

A person whose plant protein comes from soy on Monday, chickpeas on Tuesday and wheat on Wednesday has a different exposure profile from a person eating the same soy product daily, and the same is true of milk alternatives. Variety is the only consumer level control that works here, because you cannot inspect a burger for mould that was on a bean in a silo last winter.

Beyond that, the useful shift is away from the category question and towards the ingredient question. We compared what the two protein sources actually deliver in plant protein versus animal protein, and looked at what happens when a health claim becomes a product line in the protein fortification trend. The through line in both is that a label describing what a food is not tells you very little about what it is.