I think that the best strategy is to be completely honest. We are pretty sure parabens are safe, but they are very widely used so we really have to be sure. The things are everywhere now after all.
Ideally a long term trial would be carried out using the very low levels of parabens that are used in cosmetics and comparing them with products without parabens at all. But the practical problems mean it is highly unlikely that it would ever be possible to actually do this. Following a large cohort of people for decade and monitoring their exposure to parabens, while looking for what is certainly going to be a very small effect is simply not feasible. Doing a large scale trial in animals wouldn’t be scientifically conclusive either way and in any case would just create another problem. We would then have to cope with animal lovers as well as paraben haters.
The only thing that might prove the case one way or the other would be if the industry commissioned a study to look at the level of parabens in breast cancer victims compared to the population as a whole. If there was a significantly higher level in the people with breast cancer then it might be reasonable to infer that it is the parabens that are causing the cancer. If there is no difference then it is hard to see how they could be playing a role. I can already hear objections to this logic on either side of the argument forming in my head, but we have to do something.
The way this study is done is as important as actually carrying it out. It should be announced in advance as the definitive study. The study design should be made public so that it is clear what is being done and why and is open to challenge from anyone interested. And above all it should be agreed that if parabens come out of it as a potential causative agent the industry will voluntarily withdraw them from use immediately.
I know very well that that is exactly what the industry would do in any case. The point is that most consumers don’t trust big companies. It doesn’t hugely matter how that situation has arisen, what matters is to win trust back. That can only be done by not only behaving in a way that is trustworthy, but in a way that is seen to be trustworthy.