PARABENS ARE OESTROGENIC?
This is the first in a series of articles in each of which I will focus on one specific study that has contributed to the current concerns over the use of parabens in cosmetics. If you have not already done so, it may be useful for you to read the introductory article posted earlier before reading this one!
The paper I will critique in this article is:
Some Alkyl Hydroxy Benzoate Preservatives (Parabens) Are Estrogenic
Routledge, E.J, Parker, J, Odum, J, Ashby, J and Sumpter, J.P. – Tox & Appl Pharm 153, 12–19 (1998)
Oestrogenic activity is the measure of the ability of a substance to bind to oestrogen receptors, and is usually expressed as a value relative to the binding ability of oestradiol, a natural oestrogen.
This group of workers used two different methods to measure the oestrogenic activity of parabens, one in vitro (yeast recombinant assay) and one in vivo (rat uterotrophic assay). “In vitro” means, literally, “in glass”, and “in vivo” means “in life”, and the results of in vivo studies usually outweigh those of in vitro studies as they are much more likely to reflect the “in life” scenario. Additionally, the group used two methods of applying the dose in the in vivo studies – orally and by subcutaneous injection.
The results in the in vitro study were as follows:
Methylparaben – 2,500,000 times weaker than oestradiol
Ethylparaben – 150,000 times weaker than oestradiol
Propylparaben – 30,000 times weaker than oestradiol
Butylparaben – 10,000 times weaker than oestradiol
p-hydroxybenzoic acid (the common and major metabolite of all parabens) – no activity
For the in vivo study, the results were:
Methylparaben – no activity
Butylparaben – 100,000 times weaker than oestradiol
Ethylparaben and propylparaben were not tested in this assay but, given that the figure for butylparaben was a factor of 10 times weaker in vivo than in vitro, it may be suggested that the figures for ethylparaben and propylparaben could be 1,500,000 times and 300,000 times weaker than oestradiol, respectively, although this can only be claimed as a large approximation.
For completeness, I will report that the in vivo oral study found no activity for either methylparaben or butylparaben – the other parabens were not tested by this method.
Having said earlier that in vivo results are more relevant, I will now focus on that part of the study.
It is important to note that NO ACTIVITY was found for methylparaben. Therefore, on this basis alone, it is incorrect to make ANY blanket claims relating to oestrogenic activty for parabens as a group, as there are clearly significant differences in their properties, even in just this single study.
The first question that may spring to mind when seeing that butylparaben was measured to be 100,000 times weaker than oestradiol might be “that sounds extremely weak – surely too weak to be of any concern?” and, arguably, you could be correct.
100,000 times weaker than oestradiol is VERY weak, especially when this figure is compared with the oestrogenic activity of some of the phytoestrogens we eat every day that have activities only 2,000 times weaker than oestradiol – and who is concerned about these? But there is even more to consider . . . . . .
What dose was used to elicit the observed effect?
Several dose concentrations were used in the study, but the important (and relevant) ones are:
The lowest observed adverse effect level (LOAEL) was 200mg/kg body weight/day
The no observed adverse effect level (NOAEL) was 40mg/kg bodyweight/day
The units “mg/kg bodyweight/day” may seem a little complicated, but all it means is that the dose was based on the bodyweight of the test animal and it was applied daily. So the bottom line is that at a dose point somewhere between 40mg and 200mg an extremely weak effect occurs from exposure to butylparaben. On this basis, the authors concluded “…..the safety in use of these chemicals should be reassessed with particular attention being paid to the estimation/determination of the actual levels of systemic exposure of humans exposed to these chemicals in commercially available topical preparations”
So the next question is (or should be!) “how does this relate to cosmetics”? This requires, in effect, a risk assessment. We know the lowest dose that has no effect – all we need to know is the likely “dose” from using butylparaben in cosmetics. This is fairly simple to calculate.
The average daily exposure to cosmetics is estimated at 17g/day (the figure used by COLIPA – the EU equivalent of the PCPC).
A typical use concentration of butylparaben is 0.03% by weight of the cosmetic product. By no means all cosmetics contain butylparaben, so to assume a daily exposure to products containing butylparaben of 10g is not unreasonable. This means that, to achieve the same (lowest) dose at which an adverse effect was observed, a 40kg human would have to apply 60kg of cosmetics – EVERY DAY – and this assumes that all the butylparaben would be absorbed through the skin! Even is someone was fanatical enough to use 10 times the average amount of cosmetics, they would still need to be able to absorb 6kg of product!
Perhaps if the authors had taken the time to put their results into perspective, their study may have had a little less impact!
Even though the results should now look as though there has been a great deal of fuss over (almost) nothing, there’s more!
The dose in the in vivo study was administered by subcutaneous injection – in other words, it was not applied ON the skin, but UNDER the skin. This may not sound important, but it has huge ramifications for the relevance of the results!
1) Studies have shown that not all butylparaben applied to the skin in cosmetics is absorbed – one study found that only 4% was absorbed intact (Bando, H, Mori, S, Yamashita, F, Takakura, Y, Hashida, M, J. Pharmaceut. Sci. 86, 759 – 761 (1997) )
2) The skin contains enzymes called esterases, and these enzymes are present to start the metabolic process of a specific group of compounds – esters. Parabens are esters! Injecting below the skin bypasses part of the breakdown process and ensures that more intact butylparaben (in this instance) is available to reach oestrogen receptors in the body.
What this means is that even the high dose required to elicit the weak oestrogenic acivity measured in this study was exaggerated by the test conditions used.
A further assessment reveals an interesting scenario:
Given a typical exposure to butylparaben of 0.05mg/kg bodyweight/day (10g of cosmetic product containing 3mg butylparaben used by a 60kg human), and making the following assumptions:
1) Only 90% breakdown by skin enzymes
2) The same degree of binding to oestrogen receptors as determined by Routledge
3) Irreversible binding to the receptors
4) 100% stability of the receptors
5) Daily use
It would take 109 years to achieve the same dose as determined by Routledge to be 100,000 times weaker than oestradiol in this study – and many of my assumptions are worst-case and unlikely to ever occur (items 3 and 4, for example).
Whilst other studies have been carried out using similar methods, the results from Routledge’s work seem to give the “worst” results, and this is the study that is usually quoted in other work carried out on parabens. From these results, it is clear that methylparaben has no oestrogenic activity, and it is unlikely that either ethylparaben or propylparaben would display measurable oestrogenic activity under the test conditions used.
It is also clear that the oestrogenic activity of butylparaben is extremely weak and the dose used in the study is some 4,000 times that of typical daily exposure from the use of cosmetics. It seems reasonable to conclude that butylparaben poses no measurable risk in terms of potential hormone disruption by its weak oestrogenic activity.
As an aside, it concerns me that a group of scientists are able to produce a study and publish results which, by themselves, are totally out of context in terms of human exposure. Given the economic ramifications for the cosmetics industry from the headlines generated as a result of this study, and those further studies it inspired, I seriously question the morality of using research to identify a potential hazard, and then leaving it up to others to pick up the pieces and set the record straight. It is not a massive undertaking to carry out the risk assessment that I have included in this article.