So what leads Dr Darbre to point the finger at cosmetic chemicals, particularly parabens? There are ways of assessing quantitatively how similar molecules are to oestrogen. Many many thousands have already been screened. What the data actually means is nowhere near as clear cut.
Lets look a bit more closely at the supposed oestrogenic effects of the parabens. Some work has been done in Reading by some researchers including Dr Darbre to quantify it. They set up an experiment that mimics the conditions in the body in a test tube. What they found was that you need around 100,000 times the quantity of butylparaben to get the same effect as oestrogen. And that was the strongest effect in the paraben family.
The weakness of this oestrogen mimicking power has been picked up on by the defenders of parabens. Indeed this really is the counterargument. Dene Godfrey, formerly the president of the Society of Cosmetic Scientists, wrote to the International Journal of Cosmetic Scientists responding to a paper that had quoted Darbre’s work. He pointed out that the weak oestrogenic activity had only been found in one of the parabens, and that even this was such a low level of activity that it barely counted as an oestrogen mimic.
The review counters this argument by calculating the total amount of paraben that might be absorbed from a pack of lotion and pointing out that it is high enough to match the concentration showing an effect in the test. Parabens might be 100,000 times less effective than oestrogen, but if the concentration is high enough, they could still have a marked effect.
It’s easy to get carried when you start discussing numbers. They give a completely false sense of certainty to an argument. Quantifying something in a test tube is a useful way to build up understanding of how the body works, but it is a very poor guide to predicting how it will work. The fact that a paraben has a particular effect in a test tube is far from strong evidence it would behave that way in the body.
But this cuts both ways. It might behave much more strongly in the body than you would imagine from the test tube behaviour.
No data on long term exposure to low levels of low potency endocrine disruptors
If you accept that parabens might mimic oestrogen and that this might have a harmful effect, the obvious next question is what does this mean in practice.
Unfortunately, there are no studies on the effects of long term exposure at low level to molecules with a low oestrogenic effect. We only have high dose short term studies in animals to go on. It is far from a good place to be when assessing cancer risk. You would for instance be very unlikely to pick up the risk of lung cancer from smoking by looking at the medical records of people who have inhaled a lot of smoke in a burning building.
We do know that high levels of oestrogenic compounds in polluted rivers have had severe effects on the reproductive systems of fish. But do you get anything remotely similar from very low levels of very weakly oestrogenic chemicals in humans? Humans are a lot bigger than fish and don’t live in water. It doesn’t sound like a very fair comparison.
It’s hard to see how a definitive study could be carried out. The weakly oestrogenic chemicals that we know about are widespread across the whole world. I can’t think of any practical way of finding a group of humans not exposed to these chemicals that we can use to compare with people who are exposed.
I think the only fair summary of the situation is that the risk that chemicals widely used in the modern world are causing damage to our reproductive systems, including increasing the risk of breast cancer is very much a theoretical one. There is no direct evidence supporting it, but none that allows it to be dismissed either.
Part V
Parabens are widely used according to cosmetic labels, and the labels understate the true extent of paraben use
I am pretty sure that this review paper was written after the controversy about parabens had become public, because many of the criticisms of the earlier paper are addressed. It emphasises that any personal care product containing parabens is equally risky, and that the particular kind of product is irrelevant. This is certainly true – if parabens do indeed have the ability to cause breast cancer then they should not be used in any personal care product. The reporting that concentrated on underarm antiperspirants was in this sense very misleading from whatever way you chose to look at it. It also gave industry critics a great stick with which to beat the original paper.
The argument is about parabens and nobody has ever disputed that they are widely used in cosmetics, although they are used at low levels. As a cosmetic formulator I can tell you that it is very rare that you need more than about 0.4% of parabens to adequately preserve a formulation. The legal limit is 0.8%. In any case they are usually used in combination with other preservatives in which case the levels are even lower. Nobody wants to use any more than they have to. Parabens are cost effective but they are still an expense.
In what looks suspiciously like cosmetic industry bashing for the sake of it, Dr Darbre has found some published data which analyses the paraben level of some marketed products and finds that some exceed the legal limit, and some that contain parabens that aren’t labelled as such.
I think that this point is wholly irrelevant to the question of whether or not parabens are safe. I doubt very much that the data showing levels of parabens higher than legal limits are accurate. Measuring the levels of parabens in samples accurately is not easy. A glance at Dr Darbre’s original paper will show that she herself isn’t very good at it. Remember that she detected parabens in her blank control samples where presumably there weren’t any either.
I have seen enough inaccurate chemical analyses in my career to have good grounds for scepticism about these figures.
But I don’t think anything is achieved by trying to knock the cosmetic industry’s compliance with legislation. It is a big and diverse industry and while there are no doubt a few rogues about if you look long and hard enough, the big companies all work to very high standards in their plants. They follow the rules. The debate should be about whether the regulations are protecting consumers. And to answer that, we need to know whether parabens actually pose a risk or not. If the link with breast cancer turns out to be true, then we shouldn’t be using them at all at any level.
Parabens have been used so long that when they were approved the requirements to prove their safety were much lower, and the gaps have never been filled in
Parabens were not covered by any legislation in Europe until 1976. In that year they were approved for use and the issue had never been looked at again since that time. Parabens had however been looked at by food regulators in the late nineties. This highlighted that there was little data available on whether they might cause cancer and none at all on effects on reproduction. So regulations have been set without any information about the risks society is running in this area.
That in itself is quite a scary thought.
Some work on rats was carried out in Japan and reported in 2001 and 2002. Male rats were fed parabens over a period of 8 weeks. At the end of the study it was found that their reproductive organs did show some abnormalities. The levels of parabens used were very high and the effects found were fairly small. But the study was very short. What is the significance of this? I have to say that it is far from clear that it means anything at all. But you could conclude that the results could be consistent with parabens being a long term low level risk.
There have been other studies as well, but I don’t think any of them changes the conclusion. There simply isn’t enough data to be sure one way or the other. The EU’s scientific committee did a very thorough review of the data in 2005. Their conclusion was frankly a bit Delphic. There were causes for concern and more data was needed, but nothing that justified lowering the acceptable levels.
The case against parabens is weak, but has not yet been answered.
Colin, firstly, thanks for quoting some of my work in this subject! I must correct a common misconception before I go into any further detail. There is NO evidence that any of the parabens are oestrogen mimics. Studies have measured oestrogenic activity (the one you quoted from me was carried out in 1998 by Routledge etal), and this is not the same as being an oestrogen mimic/hormone disruptor. Oestrogenic activity is the measure of the ability of a substance to bind to oestrogen receptors. When the substance is bound to the receptors, it is the effect (if any) on global gene expression (GGE) that determines whether or not, and to what extent the substance actually mimics oestrogen. Darbre herself co-authored a paper comparing the effects of several parabens with that of 17 beta oestradiol on GGE and found significant differences both between the individual parabens and oestradiol, indicating that parabens do not closely mimic oestrogen.
It is not correct to assert that 100,000 times the quantity of butylparaben is required to give the same effect as oestradiol. In the Routledge study quoted, the factor of 100,000 times weaker was observed using a dose approximately 4,000 times the normal daily exposure, and using subcutaneous injection to boot (thereby avoiding any breakdown by esterases in the skin/blood circulatory system and further exacerbating the already weak adverse effect).
The work you quote on rats in Japan was carried out by Oishi et al. This work was questioned by the SCCS in their latest review of parabens; questions were put to Oishi to try to clarify the detail, but no response was forthcoming, and her results, therefore, have been discounted in the assessment of the safety of parabens.