BHA (E320)

This is the main reason BHA keeps appearing on "ingredients to avoid" lists. IARC classed butylated hydroxyanisole as Group 2B, "possibly carcinogenic to humans," and the US National Toxicology Program lists it as reasonably anticipated to be a human carcinogen. Those judgments were driven mainly by repeated rodent studies showing forestomach tumors at fairly high dietary exposures. That is a real hazard signal and should not be waved away. But it is a hazard classification, not proof that ordinary food exposure in humans causes cancer at the levels people typically eat.

The cancer concern is real enough to justify caution, but the human evidence is still thin. The National Toxicology Program notes that epidemiological data are inadequate to evaluate BHA specifically in people, and the one cited Netherlands Cohort analysis did not find an increase in stomach-cancer risk at typical dietary intake levels. That does not prove zero risk, especially if future exposure data change. It does mean the stronger online claim - that eating normal amounts of BHA-containing foods is clearly causing cancer in humans - runs ahead of the current evidence base.

BHA is often described online as an endocrine disruptor, and there is some basis for that concern. EFSA's 2011 re-evaluation noted potential endocrine effects had been investigated in a number of studies, but those studies were either in vitro or used high dose levels rather than normal food exposure. That makes the concern scientifically plausible, not clinically settled. At the moment there is not strong human evidence showing that food-level BHA intake disrupts hormones in a meaningful way, so this is better framed as a watch-this-space issue than a proven everyday effect.

This is one place where the evidence is less alarming than social media often implies. In EFSA's review, a newer study suggesting reproductive and developmental effects was judged not biologically persuasive because the changes were small, lacked a clear dose-response, and did not follow standard guideline methods well. Other animal studies gave no-observed-adverse-effect levels around 100 mg/kg body weight per day in rats and 400 mg/kg per day in rabbits, and EFSA used those data when revising its acceptable daily intake. That is not proof of zero reproductive risk in humans, but it does weaken the stronger fertility-harm claims.

The regulatory picture is asymmetric, which is part of why BHA feels controversial. In the US, FDA still permits BHA under existing food-additive and GRAS-related regulations and, on February 10, 2026, began a post-market review of its safety. In the UK and EU, E320 remains authorized, but its permitted uses are narrower and more category-specific, with tighter limits and especially careful treatment in infant-food contexts. So the honest read is not "Europe banned it everywhere" and not "everyone agrees it is fine" either. Different regulators have drawn different practical boundaries around the same imperfect evidence.

Exposure estimates matter as much as hazard labels. EFSA's 2011 review concluded that adult exposure estimates were generally below the revised acceptable daily intake of 1.0 mg/kg body weight per day, and that children had higher body-weight-adjusted exposure than adults but were still usually below that limit even at high intake estimates. That does not mean exposure is irrelevant: children and frequent consumers of BHA-containing cereals, snack foods, sauces, or dehydrated products have less margin for error than occasional users. It does mean current evidence is more consistent with a dose-and-frequency issue than with universal high risk from trace exposure.

BHA usually shows up in exactly the kinds of foods LP readers are already suspicious of: shelf-stable, fat-containing, highly formulated products where preservatives help keep oils from going rancid. That context matters. Someone eating BHA occasionally in an otherwise minimally processed, high-fiber diet is in a different position from someone getting repeated exposure through a broader ultra-processed pattern. Nutrition evidence generally supports that bigger pattern-level factors - total diet quality, energy balance, fiber intake, and how much ultra-processed food you eat overall - drive more long-term risk than one preservative viewed in isolation.