For this, the penultimate article in my War & Peace-like series on parabens, I apologise for being slightly self-indulgent. The paper I would like to discuss has made very little impact in terms of adding to the body of information on parabens, and for very good reason (to be explained!).
The paper in question is:
Construction of simplified models to simulate estrogenic disruptions by esters of 4-hydroxybenzoic acid (parabens)
Patricia Guadarrama, Serguei Fomine, Roberto Salcedo, Ana Martinez, Biophys. Chem. 137 1 – 6 (2008)
Although not at all obvious from the title, this study also claimed to have developed a simplified model for determining the antimicrobial activity of the parabens.
In both cases, the authors devised purely theoretical ways (ie, without any actual testing) of measuring the oestrogenic and antimicrobial activity of parabens. After carefully studying the paper in detail, I decided to write to the editor of the publishing journal to draw attention to some points that I felt were important:
1) The authors’ theoretical model predicted the order of oestrogenic activity as being methylparaben > butylparaben > isobutylparaben > benzylparaben. In other words, methylparaben was predicted to be the most potently oestrogenic of the parabens tested. The only problem here is that EVERY animal-based study I have seen published has this order entirely the other way around, and usually with NO measurable oestrogenic activity detected for methylparaben!
2) The authors’ theoretical model predicted the order of antibacterial activity as being methylparaben > butylparaben > isobutylparaben > benzylparaben. In other words, methylparaben was predicted to be the most potently antibacterial of the parabens tested. The only problem here is that all antimicrobial activity data published for parabens show unequivocally that methylparaben is the least active, and that the antimicrobial activity increases with the increasing carbon chain length (methyl < ethyl < propyl < butyl etc.)
Whilst the all oestrogenic activity work on parabens has been relatively recent, the antimicrobial data has been generated over the past 80 years, and suddenly we have a theoretical model that disproves all that has gone before!
The journal duly published my letter and, as is right and fair, also published a response from the authors (and here, please bear in mind that I have just told them that their results are totally wrong and out of step with all previously published data), the first, and last, lines of which were:
“In response to this comment we should firstly point out that the level of theory used in our calculations is robust, validated with experimental data, and not challenged here. . . . . . The theoretical results that we obtained stand unchanged, exactly as reported after peer review, and are intended to be a constructive contribution to this ongoing debate”.
I am sure you can imagine my surprise at the suggestion that they were not being challenged – apart from telling them that they were totally wrong, I clearly totally agreed with their findings!
The following is from an e-mail sent to the journal editor:
Many thanks for the copy of the authors’ response to my comments. I must say, however, that I am completely bemused by their claim that I am not challenging the level of theory as I am saying that the results (in both models) are entirely the opposite of those generated by all the published practical studies of which I am aware. That is, surely, challenging the entire procedure employed in generating their results. I would appreciate the opportunity to respond to their reply, if that is acceptable and, if so, should I do this by e-mail to you, or more formally through the web site?
To which I received the following reply:
No – I must draw a line under this, else it will go on forever. Both sides have had their say and it will now be in the public domain for others to judge.
Surprise now turned into frustration, as the authors had completely ignored my criticism and, indeed, brushed it off as though it was not even at all critical! It beggars belief that their response was even published. Sadly, this is another example where the peer review process falls short of perfection (as with Darbre et al, 2004 as described in Part 2 of this series), albeit a rare example, but it serves to demonstrate that published scientific studies should always be treated with great caution and not taken at face value.
So, this is my little bit of self-indulgence, as it is the first time I have sounded off in public over what I perceive to be a gross injustice, and the publication of yet another totally misleading scientific paper. I am sure you can see that it is clear why this paper has had no impact in “parabens world”, but just consider the implications should this clearly flawed theoretical model be used to assess the oestrogenic activity of other substances . . . . . .
If you haven’t followed the other parts of this series, you may do so here.