Richard Sharpe: Let common sense guide you in the saga of bisphenol A

The Independent has campaigned for restricted use of the chemical bisphenol A in the belief that it is likely to cause harm to humans, and babies in particular. Bisphenol A is used in the hard plastic of baby feeding bottles and small amounts can leach out into milk. Even after having spent around $50m on the question, scientists are split in their views on the health risks, and the Independent has challenged regulators and the UK Foods Standards Agency to explain why they continue to say that Bisphenol A is safe. To any sane reader this simply does not make sense; surely it's either safe or not? Perhaps after reading below, you may think, like me, that a lot of the $50m is money wasted because the original findings were deemed incorrect some time ago.

Let me start with a statement by one of the scientists (Fred vom Saal) who strongly advocates banning bisphenol A: "The science is clear and the findings are not just scary, they are horrific. Why [would] you feed a baby out of a clear, hard plastic bottle – it's like giving a baby a birth control pill?" Scary indeed, when you think how many mothers in Britain have done just this. Who would (knowingly) do such a thing and why on earth would the Food Standards Agency say this was OK? There is a simple explanation – this statement is not only wrong; it is a complete misrepresentation of what the scientific facts show.

Bisphenol A can act like an oestrogen, the primary female sex hormone. The contraceptive pill contains a very potent oestrogen, ethinyl oestradiol, and this is partly responsible for its effects (preventing ovulation). The difference between bisphenol A and ethinyl oestradiol is in their potency. You need at least 10,000 times as much bisphenol A as you do ethinyl oestradiol to have similar oestrogenic potency. There is accurate data on how much bisphenol A we ingest each day and it is at least 50,000 times less than the level needed to make this the equivalent of taking a contraceptive pill. In fact it is less still, because not much of the bisphenol A we ingest ever gets into the blood to cause any effect in the body.

So how can a sub-set of scientists make claims for bisphenol A being a dangerous oestrogen? The answer is that their own studies in laboratory animals showed it could cause potentially harmful effects. However, attempts to repeat these findings in much larger studies have failed. This is not unusual in scientific research. It is common for a small preliminary study to find something which cannot later be reproduced by another laboratory in a more detailed study. This is why one of the golden rules of scientific research is that results must be reproducible.

But once the results could not be reproduced, the authors of the original studies claimed this was because the follow-up studies had been paid for by the chemicals industry. But then, completely independent studies, by US and Japanese government laboratories, also failed to reproduce the original findings.

Most studies that reported finding effects of bisphenol A did not expose the animals to this chemical in the way that you and I are exposed (by mouth); they either injected it or used an implant under the skin. This effectively bypasses breakdown of bisphenol A in the gut and means that you get a much higher level of exposure than if the animals had been exposed orally. This may explain some of the differences in results from the contrasting studies.

End of story? No, the authors of the original studies assert these next studies did not use the right animal strain – it was not sensitive enough to oestrogens. But these differences are trivial and cannot rationally explain the difference in results. So on the one hand, small, preliminary studies; on the other, large, detailed studies designed explicitly to repeat the findings – which would you believe? To dismiss the more detailed studies in favour of the initial smaller studies not only defies common sense. It is bad science too.

So, can we all sleep free from worries about health risks of bisphenol A? Perhaps not quite yet. There is good evidence that obese people with heart and liver diseases also have higher levels of bisphenol A. This is probably because people with lots of bisphenol A are the ones drinking sugary drinks from cans – ie it is probably the diet causing the problem, not the chemical.

I should declare my credentials. I am funded by the UK Medical Research Council (an independent body). My research does not involve bisphenol A and I therefore have no vested interests in proving that bisphenol A is either safe or dangerous. From my research I do think that some environmental chemicals, to which we are exposed via our lifestyles and diets, probably do have harmful effects in the way claimed for bisphenol A. I am not a die-hard "chemicals are safe" man by any means, but I am convinced, based on present evidence, that bisphenol A poses minimal risk to people because of any oestrogenic effect. As to other possible effects, we will have to wait and see.

A final plea – use your common-sense when making your own decision about bisphenol A; in science, as in life, it is the most robust indicator of what is right and wrong.

Professor Richard Sharpe is a senior scientist at the Medical Research Council Human Reproductive Sciences Unit, Centre for Reproductive Biology, Edinburgh