Xenoestrogens #5: Cumulative endocrine effects.

23 October 2018 1 By Sco

Is not a single substance, but the interaction of a wide variety of hormonally active compounds, responsible for the increase in endocrine problems?

This post is an english adaption of my recent german post and part of a series on foreign substances that imitate the effect of the female sex hormone estradiol (“xenoestrogens”). I suggest to start with the basics::
Xenoestrogens #1: Natural Estrogens and Estrogen Receptor Signalling

Other posts I wrote on this topic include:
Prelude: Soy isoflavones and the “soy boy” story: 1, 2, 3
#2: Why are some compounds acting as endocrine disruptors, and where do they occur?
#3: Hormones in Plastics
#4: Hop polyphenols. Is beer consumption messing with our hormone balance?

We had this picture before. It shows different sources of xenoestrogens. all pics CC0, from pixabay


For the first time, a blog topic touches my own research. I admit that makes me a little nervous. But it is also exciting to let my readers participate in my professional life.
I’ve already written a lot about individual xenoestrogens here, and I could do that for years to come, there are so many different ones. Mould poisons, pesticides of all kinds, other natural ones… I feel like I haven’t even scratched the surface yet.

And I don’t know if it was noticed, but these posts all end with an (at least partial) attempt to ease your worries. In general, for these substances, the exposure of the average population (at least in well-protected Europe) is far below the toxicologically questionable limits.

And yet there are epidemiological effects that need to be explained. The average testosterone levels in men have indeed decreased over the last decades. Menstruation does start much earlier on average today. Breast cancer rates are higher, patients are younger than before.

So what is happening here? Genetic effects are ruled out, evolution is not so fast. They must actually be environmental or life-style influences. The exposure to xenoestrogens, which at present is certainly stronger than 50 years ago due to pesticides and plastics, would be an elegant explanation – but luckily, as I said, we do not achieve the necessary concentrations. Perhaps there are still undiscovered substances that make the difference? Or do we completely misjudge the toxicologic impact of a known chemical? Hmm… Possible, but not very likely. It is 2018, not 1970. Screening batteries that check thousands of substances at once have long been established.

But what if the problem is not a single substance, but the interaction of the many different endocrine active substances to which we are exposed?
Is this conceivable?

Good Question.

hey ho

Combinatory effects

Before we dive into the matter, perhaps we should clarify a few terms: The “combinatory effect” describes the effect of several substances on a toxicological “endpoint” (e.g. estrogenicity) compared to the effects of the individual substances.
In the simplest case those act in an additive manner. This means that the effects of the various substances add up. Strongly simplified example: 2 substances alone kill 10% of the cells, in combination then 20%.
If substances in a mixture are not additive, this means that they influence each other in their effect. One speaks then of an “interaction“, and we differentiate between:

  • synergististic interactions: the compounds potentiate each other’s impact, the combinatory effect is greater than additive
  • antagonistic interactions: the compounds negatively influence each other’s impact, the combinatory effect is less than additive

*of course, it is not that simple in real life, because this calculation ignores the dose-response relationship. In reality, mathematical models are needed to calculate the expected additivity. The “CI model”, for example. I could explain that, but it leads by far too far here. 😉

The scientific question

classical risk assessment

I have once described how the risk assessment of substances works in a previous post… briefly recapitulated: You take the amount of the substance per kg body weight that does not yet produce a toxic effect in animal experiments and divide it by 100. That is then the TDI – tolerable daily intake. Then one needs exposure data, that is in which food, of which on average how much is consumed, one finds the substance. From the combination of this information and the TDI, legal limits are calculated which are set in such a way that 90% of the population cannot reach the TDI. And the 10% that reach it (e.g. because they feed exclusively on potato chips contaminated with acrylamide) are still a factor of 100 below the toxicologically questionable dose at which negative effects begin.

cumulative effects of xenoestrogens

If we consider this, and now think about whether cumulative effects are responsible for the problems in the population described in the intro, one thing becomes clear: additive effects are not enough to not explain those.
Because it would mean that food/cosmetics/etc. to be contaminated with more than 100 xenoestrogens, all of which are at the limit of legality, on average (if the concentration is above the limit, the product would not be fit to be sold and therefore would not reach the consumer).

As said, there are many xenoestrogens… But then again… not SO many. A very unlikely scenario, if you ask me (and you are, since you’re still reading^^).

No, what we are looking for are synergistic interactions!


A classical synergistic interaction. Don’t try this at home… CC0, pixabay

So the question is:

Do different xenoestrogens mutually strenghten each other’s effects?

This question has recently been asked by several renowned researchers of the field.ref
Our research group is partly working on it too. The results obtained so far are diverse and have to be considered in a differentiated way: Some estrogenically active substances do not influence each other, or even have an antagonistic effect on each other. Beer polyphenols, for example, seem to reduce the estrogenity of other substances. At least according to my work.ref

Nevertheless, there are indications that we might have a problem with combinatory effects in some cases.
The mould toxin Alternariol (which by the way is not regulated and could therefore be in completely unknown amounts in your tomato sauce) is only estrogenically active in very high, seemingly unrealistic concentrations. However, much smaller concentrations (50 times lower) are apparently sufficient to drastically increase the effect of other xenoestrogens. A colleague of mine – together with her working slaves… uhm, master students – has proven this in her investigations. In fact, Alternariol pushed the estrogenicity of both zearalenone and a-zearalenol, two other mycotoxins which can often be found in cereal products,ref as well as the endocrine impact of the soy ingredient genistein. Zearalenone and genistein also interact synergistically, albeit to a lesser extent.ref

Through these combinations, even the maximum stimulation of the estrogen receptor (ER) by the natural hormone 17b-estradiol could be exceeded, which none of the individual substances managed to do, even at high concentrations!

searching for a mechanism…

How Alternariol is doing this, we don’t know (yet). But we should expect other chemicals and xenoestrogens to interact in a similar way.

One problems we are constantly encountering is the extreme complexity of the processes that occur around the ER. There are interactions with various other signalling pathways. There are tens of phosphorylation sites at the ER itself, some of them activating, some supressing the response. There are co-factors en masse. And as if that were not enough, the two isoforms of the receptor, ER-a and ER-b, bind completely different substances and occur in different tissues. If our knowledge about the mechanisms here were greater, we would know what we are looking for.
But as it is, we are left with the possibility of “screening”, i.e. we select a few from a sea of xenoestrogens and test them against each other. And then you find something or not.

Limited validity

What we are also still far away from is thorough animal testing. All our studies are based on cell culture measurements. This – in all honesty – limits the informative value of our work enormously. We only provide hints that should be tested on animals. The problem with the required in vivo studies: money. Since one would have to test many different combinations, one would need a shitload of animals. And that is very expensive. In the pharmaceutical industry (e.g. for the clarification of interactions between drugs) such large-scale studies are made and demanded. But in its underfinanced little sister, toxicology, there are simply no resources for that.
Without saying, it is understood that food and cosmetic industry has little interest in financing studies which would eventually lead to stricter regulations.
And getting public money… well… let’s say we need much more and stronger data first.


Yeah I know. Believe me, nobody is enjoying that part of science. However, there’s sometimes no alternative. CC0, pixabay

One possibility, of course, could be offered by various computer models that simulate the behavior of substances in our bodies. They are still in their infancy, but once mature, I’m looking forward to revolution.


Now, I’ll finally get to the point:
There are indications that the danger from endocrine disruptors is not so much coming from individual xenoestrogens, but from the mixture of various such substances that occurs.
Accordingly, any reduction of these substances – regardless of whether they occur naturally or are carried in as contaminants – would be desirable.

A risk assessment that takes this into account would consequently have to apply the precautionary principle and, for example, not approve any new pesticides with an estrogenic effect at all, explicitly even if the concentrations attainable in the consumer were harmless to health, since synergistic interactions are not unlikely.

Yes, of course, I have just this tiny little junior post doc voice. So it’s important to say that this view is also shared by voices with considerably more scientific weight, like the endocrine society.ref

But of course that is a wish to the non-existent Santa Clause. We are obliged to continue to provide data after data until the issue is finally resolved and the burden of proof becomes so overwhelming that the authorities must act.

In my blog, I’m stating my honest opinion as a researcher, not less and not more. Sometimes I make errors. Discuss and disagree with me – if you are bringing the better arguments, I might rethink.