Over the past decade, concerns about the health effects of bisphenol A (BPA), have forced food and beverage companies to largely abandon the use of the common plastic in many household items. In its place, they’ve turned to more than 50 “BPA-free” alternatives. Now, researchers report that some of these substitutes may cause the same ill effects in mice, particularly in reproductive cells. If the new results hold in further animal and human studies, they could upend efforts to mollify consumers’ health concerns over the plastics in food and beverage containers.

“It suggests these replacement bisphenols are not safe,” says Patrick Allard, a molecular biologist at the University of California (UC), Los Angeles, who was not involved with the study.

Concerns about BPA have been swirling since it was first introduced into commercial products in the 1940s. Since then, it has become ubiquitous in water bottles, toys, canned food linings, and even cash register receipts, as its clarity and toughness made it an essential component of polycarbonates and other common plastics. A 2003–04 study by the U.S. Centers for Disease Control and Prevention found that 93% of Americans have at least trace levels of BPA in their blood. Nevertheless, extensive studies by the U.S. Food and Drug Administration have not shown that BPA is dangerous to human health at normal exposure levels, though the conclusion remains controversial .

In 2003, while carrying out mouse studies unrelated to BPA, Patricia Hunt, a reproductive biologist at Washington State University in Pullman, and her colleagues found that the compound was leaching out of plastic cages housing female mice. The result was an increase in chromosomal abnormalities in the lab animals and their offspring. That finding, along with others in animals that suggested BPA “disrupts” estrogen hormone receptors, triggered an avalanche of studies that fingered the compound as interfering with meiosis, the process by which the number of chromosomes is cut in half and chromosomal segments are shuffled during the production of sperm and egg cells. The finding also led to new mouse cages, made of a more durable plastic called polysulfone.

But in recent studies, Hunt and her colleagues again noticed odd results in their mice. It was “a strange déjà vu experience,” Hunt says. “Our control studies started going wacko.” After months of work, Hunt and her colleagues traced the problem to contamination from cages damaged by washing and other normal wear and tear.

Hunt sent samples from damaged and undamaged cages to Roy Gerona, a chemist at UC San Francisco. Gerona and colleagues determined that the damaged cages were leaching out compounds manufacturers often use to replace BPA, such as bisphenol S (BPS) and diphenyl sulfone.

Gerona puzzled over an additional oddity: Polysulfone doesn’t contain BPS. After evaluating the starting material and leachates, Gerona says he believes the polysulfone degraded to produce BPS and other BPA-like compounds.

After getting the contamination under control, Hunt and her colleagues decided to test the effects of BPA alternatives directly. They fed pregnant female mice low doses of BPA, BPS, diphenyl sulfone, or a placebo. Compared with unexposed females, those exposed to BPA or its alternatives produced more protein markers of genetic damage during meiosis, they report today in Current Biology .

In previous studies, that kind of genetic damage has gone on to cause aneuploidy, an abnormal number of chromosomes that can trigger miscarriage in females and reduced sperm count in males. What’s more, in the current study Hunt and her colleagues showed that the effect lasts beyond the mothers and fetuses directly exposed to BPA and its alternatives. Genetic abnormalities persisted for two generations of male mice unexposed to BPA and its substitutes.

Just what this means for people is hard to say. “Nobody has ever proven it causes harm at the levels to which people are normally exposed to it,” says Oliver Jones, a chemist at RMIT University in Melbourne, Australia. However, Hunt and others suggest that the strong similarities in chemical structure between BPA and some of its alternatives mean that consumers may be wise to be wary of labels that tout “BPA-free” products.

The study also raises concerns about the reliability of widespread studies of BPA, says Monica Colaiacovo, a geneticist at Harvard Medical School in Boston. Ongoing studies of BPA’s effects commonly house animals in plastic cages previously thought not to expose them to bisphenollike compounds. Yet, Hunt’s cages were inadvertently subjecting animals to contamination.

“If you are already producing an effect in your control [animals], you might fail to see a significant difference” in your experimental animals, Colaiacovo says. This could make it even harder for scientists in the future to sort out any real dangers of BPA and its family of replacements.