Endocrine disruptors in cosmetics: parabens, UV filters and what of it is documented

You are standing in the bathroom. It is twenty to eight. Shower gel, shampoo, conditioner, body lotion, deodorant, face cream with sun protection, toothpaste, and perhaps foundation, mascara, a spray of perfume.

That takes seven minutes. You think about none of it. Which is exactly why it is interesting.

Many women know this moment from a different angle: the menstrual cycle is unsettled, the skin plays up, the breasts feel tight for two weeks a month. At the gynaecologist the ultrasound is unremarkable, the values sit within the reference range, and you go home with the feeling that something was not captured.

In situations like these I do not start with cosmetics. I get there at some point. Because you cannot change age, genetics and history. Your morning routine you can.

This article visibly separates four things: cell culture in the test tube, animal model, observation in humans and intervention in humans. With environmental topics that separation is the difference between information and panic, and it matters more to me than any recommendation that comes at the end.

The number everything starts with

The discussion about cosmetics almost always revolves around a single product. Is this cream all right. Is this deodorant a concern. That is the wrong question.

The right question is: how many products are there in total, and how often a day.

This association is not a chance finding. It also shows up where researchers looked especially closely.

On top of that comes a pattern described by a Dutch survey of 516 people: toothpaste, deodorant and day cream are typically used once or twice daily, body lotion and sun protection often on neighbouring body areas, and the majority of all products land on the skin in the morning.

That is the point. It is not a spread-out day, it is a block.

A number that is everywhere and still shaky

You may have come across the claim that a person puts twelve products with 168 ingredients on their skin each morning. The number comes from a 2004 survey by a consumer organisation. It is not peer reviewed and there is no methods publication for it.

I mention it here anyway, because you meet it everywhere. But I base nothing on it. The version that can be documented is above: up to nineteen products in 100 adolescents within 24 hours, clearly higher urinary paraben levels among the high users, and in the Boston study a monotonic association between product number and urine values.

Why is this a topic in a hormone consultation at all. Because some of these substances can dock onto receptors that are actually meant for the body's own hormones. This group is called xenoestrogens, and cosmetics is only one of several routes of entry. You will find the overview of all routes in the guide to xenoestrogens in everyday life, here it is exclusively about the skin.

And a word on gynaecology, because that question is often in the room. The gynaecological assessment is the foundation, not the competition. An environmental history covering home, workplace and bathroom shelf simply needs a consultation of its own, no matter who runs it. The world federation of gynaecology published its own paper on endocrine disruptors in 2022. The topic has arrived there. It just needs its own time, and I deliberately plan that time in for these questions.

Does it even arrive? The route through the skin

Before anyone may talk about hormonal effects, a prior question has to be settled. Does the substance arrive in the body at all, or does it stay lying on the surface.

The common picture is that the skin is a wall. That is only half right. The skin is more like a coffee filter. It holds back the coarse material and lets some of it through. How much depends on the size of the molecule, on fat solubility, on the area covered and on time.

Some of the parabens are already cleaved during passage through the skin. The skin contains esterases, enzymes that break the ester down into p-hydroxybenzoic acid. This cleavage is the main argument for the all-clear. But it is incomplete, otherwise intact esters would not be detectable in blood and tissue. Both things are true at the same time.

For sunscreens the absorption question has been answered particularly cleanly, because the American medicines agency investigated it itself. Twice.

The observation I consider the most practically important in this section comes from Korea. There, parabens were measured in the urine of 46 pregnant women and their newborns, in the children within 48 hours of birth.

Methylparaben was the highest, butylparaben the lowest, in mothers as in children. The values correlated with each other, and the child to mother ratio was 0.5 to 0.6 for methyl and propylparaben. What the mother applies is therefore measurable in the newborn. That is the reason I speak more plainly in pregnancy and breastfeeding than elsewhere.

Parabens, honestly sorted

Now to the substance group that causes the most trouble and has the weakest data.

Parabens are esters of p-hydroxybenzoic acid. They preserve, meaning they prevent bacteria and mould from growing in a water-containing product. On the ingredient list you will find four names: methylparaben, ethylparaben, propylparaben, butylparaben. They usually appear at the end, because the list is sorted by descending amount.

Without preservation an opened moisturising cream would be a culture medium within a few weeks. That is the function at stake, and it is not decorative.

Besides the oestrogen receptor, further points of attack are under discussion. A 2018 review classifies parabens as a group of endocrine active substances and describes effects at nuclear receptors beyond the oestrogen receptor, among them the PPAR receptors of fat metabolism. An older review additionally names antiandrogenic activity and the inhibition of sulfotransferases, enzymes that break down the body's own hormones.

That is the mechanistic part. It is plausible, but it comes mainly from the test tube. What happens in humans.

The breast tissue paper and why it is often misquoted

There is one study that turns up in almost every text about parabens. It is over twenty years old, and it is the most frequently misquoted paper in this entire field.

This paper nevertheless became a headline, and the headline became an everyday fear. The idea that deodorant under the arm is a breast cancer risk is particularly persistent. To this day there is no robust epidemiological basis for that hypothesis. I mention it here because it is often asked about, not because I support it.

The follow-up work went into cell culture.

The same working group published a review in 2014 that maps in vitro findings onto the known hallmarks of cancer and calls for a regulatory review on that basis. That is the strongest formulation from the critical side. It is a position, not a consensus, and even it asks only for a review.

The cautious counter-voice comes from a 2022 review. It notes that systemic exposure is confirmed by detection in blood and urine, but explicitly describes the evidence from animal experiments and epidemiological studies for a link with breast cancer as limited. That is the wording I take my bearings from.

There is one more comparison that puts the order of magnitude in place. A 2024 umbrella review analysed 52 systematic reviews with 759 meta-analyses on chemicals from plastics. For phthalates and bisphenol A it describes numerous associations with human health endpoints. Parabens and UV filters do not appear in it as a substance group of their own.

What matters here is what this umbrella review did not do: it did not even search for parabens and UV filters, because the authors deliberately limited their scope to microplastics, polymers, plasticisers, flame retardants, bisphenols and PFAS. Their absence is therefore no evidence of anything. That the data on cosmetic preservatives are thinner than on plasticisers and bisphenols is my own assessment from the overall picture of the papers quoted above, not the result of this review.

What Europe has done with parabens

One sentence that keeps appearing in search engines runs: parabens are banned in the EU. The sentence is wrong. Reality is more differentiated and considerably more interesting.

What is interesting is how the responsible body words things. The Scientific Committee on Consumer Safety, SCCS for short, stated in 2021 for propylparaben that the substance is safe up to 0.14 percent. The same document states that the available data give indications of possible endocrine effects, but that the level of evidence is not sufficient to classify the substance as an endocrine disruptor.

That is neither an all-clear nor an alarm. It is an honest intermediate position, and those have become rare.

And it is the reason why my practical suggestion is not to avoid one particular product. It is to lower the number. That is exactly what the authority now calculates itself.

UV filters, and why sun protection stays anyway

This is where it gets delicate, so I will say the most important thing first.

Now the data, in order.

A second mechanism comes from the same working group. In a cell line with androgen and glucocorticoid receptors, benzophenone-3 and homosalate blocked the activation of the androgen receptor by dihydrotestosterone, with half-maximal inhibitory concentrations around 5 times 10 to the minus 6 mol. None of the eight filters tested showed an activating effect on its own. Antiandrogenic, then, and again from the test tube.

And now the question that actually gets asked in the consulting room: does this change my hormones.

There is one finding that points in the other direction, and I am not leaving it out.

How does the European side assess this. For benzophenone-3 the SCCS has differentiated the maximum amount: up to 6 percent in face cream, hand cream and lipstick is considered safe, in body cream and sprays only up to 2.2 percent. On the question of hormonal effects the committee writes that the evidence across in silico models, cell culture and animal studies is inconclusive and at best ambiguous.

I find this double finding instructive. The maximum amount falls, the hormone question stays open. That is precaution, not a verdict of guilt.

A side aspect concerns the age of the bottle. A 2021 product analysis found a mean of 39 milligrams of benzophenone per kilogram in 16 octocrylene-containing sun protection products, and a mean of 75 milligrams per kilogram after an accelerated ageing procedure. In the product without octocrylene, benzophenone was not detectable. That is pure product chemistry, and there is no human study on the health consequences of this impurity. In practice it means: the opened bottle from the summer before last is not only a question of smell.

The counterweight that must not be missing from any of these texts

If you would still like to sidestep the filter question, there is a route that brings both together. Mineral filters made of zinc oxide or titanium dioxide stay on the surface and scatter the radiation instead of absorbing it.

Where sun protection weighs heavily in terms of quantity is shown by a modelling study from Switzerland. Aggregate exposure to octinoxate was calculated for 1196 people, linked with questionnaire data and the chemical analysis of the products actually used. In summer, 64 percent of exposure in women, 85 percent in men and 93 percent in children up to twelve years came from sunscreens. Overall the values lay below the derived limit, yet on an intense sunny day the 99.9th percentile exceeded this limit for thyroid-relevant effects, especially in children up to four years.

The authors themselves emphasise that the quantitative data on the skin permeability of this filter are insufficient. In practice the finding does not mean less sun protection, but additional shade, clothing and a hat for small children. How much sun the body actually needs and where the balance lies is covered in the guide How much sun does the body need.

Triclosan and the word perfume

Two topics that rarely appear together, but belong together here. One is an example that regulation can work. The other is the biggest gap on every label.

Triclosan: the case that is largely settled

Triclosan is an antibacterial substance. It used to be in deodorants, toothpaste, soaps, chopping boards and sportswear. Structurally it distantly resembles the thyroid hormones, and that is exactly where the discussed point of attack lies.

In humans the data are thin and inconsistent. In the NHANES analysis mentioned above there was a positive association between triclosan and total T3, but only in adolescents and in an exploratory analysis. In the Californian children's cohort, prenatal triclosan was associated with an earlier menarche. Both are associations.

And then there is the practical punchline. Since 30 July 2015 triclosan has been permitted in Europe in only a few product categories: toothpaste, hand soap, body soap and shower gel, non-spray deodorant, face powder, blemish concealers and nail cleaning products up to 0.3 percent, plus mouthwash up to 0.2 percent. In all other categories it is no longer permitted.

With that, triclosan is largely history in European cosmetics. I think it is worth saying so plainly for once. It is an example of a debated substance disappearing from the market without anyone having to buy a product to protect themselves.

Perfume: one word, many substances

Now to the biggest gap on the label. The European Cosmetics Regulation allows fragrance compositions and their starting materials to be grouped under the collective term perfume. The reason is recipe protection, that is, trade secrecy.

In practice that means: behind this single word there can be a mixture of many individual substances. That is also what the results of the 2012 product analysis point to, in which the highest concentrations and the most detections were found precisely in fragranced products and in sunscreens, and in which many substances detected were not on the label.

Among the substances that can occur in fragrance compositions are phthalates, in cosmetics mostly as diethyl phthalate. This substance group has its own and considerably more extensive body of data, which is not the topic here. For everyday life the association is enough: perfume users in the Boston study had 167 percent higher monoethyl phthalate values in the urine.

A note on the skin itself, because it is often asked about. If acne or blemished skin is your main topic, the cosmetics question is not the first place to start. That is about insulin, androgens and inflammation from the inside. It is covered in detail in the guide Hormonal acne from the inside and is deliberately not repeated here.

What matters practically, and what does not

Everything so far has been context. Now the question that comes at the end in the consulting room: does it achieve anything at all if I change my products, or is it symbolic politics.

There is a surprisingly clear answer to that question.

There is a European professional body that has already derived a graded recommendation from these data.

A second voice comes from the world federation of gynaecology and obstetrics. A 2022 FIGO paper included 16 intervention studies after a systematic search of five databases and assessed which measures can lower exposure to endocrine disruptors. On interventions that improve clinical endpoints there are, according to the authors, only limited data.

I like quoting this paper because it dissolves the front line that some people draw around this topic. The world federation of gynaecology takes endocrine disruptors seriously. It is not a question of camp against camp, but a question of time and focus in a full day of consultations.

Reading the ingredient list, in seven words

You need no app and no training in chemistry. You need seven terms and thirty seconds per product, once.

And what does not count: no single seal, no brand, no price. A very expensive product with forty ingredients is no better in exposure terms than an inexpensive one with eight. That is why I recommend no products and no places to buy them. The decision lies with the ingredient list, not with the shelf.

The paragraph that belongs here

Now comes the part I consider the most important in the whole article.

This topic has a tipping point. It starts with a sensible question and sometimes ends in a state where every ingredient list is photographed, every restaurant visit means stress and every gift has to be checked first. I see this regularly in practice, and I also see how exhausted people become in the process.

Control over your own surroundings feels good at first. That is also why the topic is so appealing: finally something you can steer yourself. But when this steering starts eating up space, the net effect is negative. Chronic stress is a far better documented factor for your hormone system than methylparaben in a hand cream.

If you notice that purity thinking is increasingly governing your food, your daily life or your relationships, that is a topic of its own and not a cosmetics topic. The guide Understanding eating disorders between body and mind describes this transition more closely, and it describes it without assigning blame.

One last sentence for context, and it matters to me. Not every hormonal disturbance has an environmental cause. Menstrual cycle disturbances, trying to conceive, menopausal complaints and thyroid problems have a long list of possible reasons, and genetics, sleep, stress, weight, nutrition and illness sit right at the top of it. The environmental perspective is an additional lens, not an explanation for everything.

And for everything to do with medication the same principle applies. Anyone taking the pill, hormone replacement therapy, a bioidentical hormone preparation, metformin, an antiandrogen or a thyroid preparation does not change any of it because of a blog article. All of these are prescription-only medicines. Bioidentical hormones are not harmless simply because they resemble the body's own hormones, and they belong in a medical prescription with monitoring. Metformin is not approved for PCOS in Germany, so it is used off label there, which requires informed consent and follow-up. Antiandrogens have their own contraindications and regulatory warnings depending on the substance, and they require reliable contraception. I deliberately give no doses here, that belongs in a prescription and not in an article. Stopping them is a medical decision too, because an abrupt stop can trigger complaints of its own. Adjustments like these therefore belong in a conversation with medical supervision. And a recommended gynaecological or reproductive medicine assessment is not postponed because someone wants to change their cosmetics first. When you are trying to conceive, time is a real factor, and nothing in this article promises you a pregnancy.

Where cosmetics fit into the bigger picture

Cosmetics are one route of exposure among several. So to close, the context that widens the view instead of narrowing it.

The Endocrine Society, the large international professional body for hormone medicine, has published two detailed scientific statements on endocrine disruptors, in 2009 and 2015. The second paper considers the evidence strengthened in seven fields, among them female reproduction, thyroid and hormone-sensitive tumours. At the same time it explicitly urges caution about concluding causality in humans.

Two concepts come from these papers that explain why the usual limit-value logic does not work seamlessly in this field. One is non-monotonic dose-response relationships, the observation that a low dose sometimes behaves differently from a high one. The other is developmental windows, phases in which the same amount has an entirely different meaning. Both are well grounded, and neither is yet a statement about a single person in the consulting room.

Two further routes of exposure belong in this picture, without my explaining them here. Zearalenone is a mould toxin and at the same time the best documented mycoestrogen, it comes into play via grain and via damp indoor spaces and is described in detail in the guide Zearalenone as a mycoestrogen. And cadmium is considered the best studied metalloestrogen, where the decisive question first of all is which test says anything at all, and that is exactly the subject of the guide Measuring heavy metals in blood or urine.

That is the reason why I ask about these things when hormonal complaints are stubborn. Not because I assume they are the cause. But because they are among the few factors that can be changed once age and genetics are fixed.

Frequently asked questions