Fifteen Thousand Chemicals Touch Food Packaging and Only 1,222 Have Known Health Risks

Nobody inspects the plastic around a block of cheddar with much suspicion. Food packaging is assumed to be inert, on the reasonable grounds that breakfast should not arrive with a side of chemicals.

The evidence says otherwise. Many food-contact materials leach chemicals into whatever they are holding. The best-known is bisphenol A, a plastics ingredient and a potent endocrine disrupter, which the European Union banned from food-contact materials in 2025 and which the United States prohibits in baby bottles, sippy cups and formula packaging.

Chemical whack-a-mole

The difficulty with banning one compound is what arrives to replace it.

BPA is frequently substituted with BPS, which turns out to have similar problems. Helene Weisinger, a chemist at the Swiss nonprofit Food Packaging Forum, calls the pattern chemical whack-a-mole, and it is the reason a ban can produce very little change in actual exposure.

Martin Wagner, a biologist at the Norwegian University of Science and Technology who studies the health effects of plastics and was not involved in the new research, has the data showing both halves of it. Stricter European rules have worked in one direction: exposure levels to phthalates and to regulated bisphenols are falling across European populations.

At the same time, he says, exposure to the alternative chemicals is rising.

The backlog is such that it would take hundreds, if not thousands, of years to test all of these chemicals

The size of the gap

The wider context is more alarming than the packaging question alone.

"There's 350,000 chemicals that are put on global markets, and for a lot of those, we have relatively little information about their toxicity," says Olwenn Martin, an associate professor in health and environment at University College London, who was also not involved in the research. "They've not been exhaustively tested."

She adds that this is well known within her field and would probably come as a shock to everybody else.

Food packaging is no exception. Of roughly 15,000 chemicals used in food-contact materials, the researchers report that 1,222 have known major health risks. For most of the remainder there is very little information at all.

That is the origin of the study. People kept asking Weisinger's team what to worry about, she says, on the reasonable basis that nobody can run measurements on 15,000 compounds.

Regulating by family

The proposal, published in Environmental Science and Technology, is to stop treating each chemical as an individual case.

Weisinger and her colleagues sorted known food-contact chemicals into structurally related groups and identified 38 groups in which many members carry clear health risks. Each of those groups also contains numerous compounds whose effects remain unknown, which the paper flags for future work.

The argument is that the unknowns do not have to be resolved first. Structural similarity is itself grounds for regulating a family together, which would close off the substitution route that makes single-compound bans so ineffective.

Martin supports the approach on grounds of pure arithmetic. The backlog, she says, is such that it would take hundreds if not thousands of years to test all of these chemicals at current rates, and regulatory bodies are already looking for alternatives. Grouping chemicals that are similar structurally, and making assumptions across the group, is one of them.

What it would change

The honest limitation is that grouping is an inference rather than a measurement, and some compounds would be restricted without anyone having demonstrated harm from that specific molecule.

Set against that is the status quo, in which a chemical stays in use because nobody has got round to testing it, and a banned one is replaced within months by a near-identical cousin that also has not been tested.

Banning entire families, as Wagner puts it, might stop that kind of unfortunate substitution. It is the difference between regulating chemistry and regulating names.