Endocrine Disruptor Examples: The Common Ones, Sorted by Evidence

For the mechanism — receptor mimicry, receptor blocking, and why the low-dose end of the curve is the part conventional testing missed — start with the endocrine disruptors guide. This page is the index rather than the explanation: what the substances are, where you meet them in an ordinary week, and how firm the ground is under each one.

The WHOWorld Health Organization and UNEP review names close to 800 chemicals known or suspected of interfering with hormone systems. Almost none have been through an assay built to detect endocrine activity at the concentrations hormones themselves operate at. So any list is a list of the well-studied minority. That is not the same as the most harmful minority, and it's worth holding those two ideas apart while reading.

The evidence column below uses four words consistently. Strong means human data plus a reproduced mechanism, and usually a regulator that has already acted. Moderate means a solid mechanism with human evidence that is thinner, mixed, or restricted to particular exposure levels. Emerging means the mechanism looks real and the human epidemiology has barely started. Contested means credible researchers disagree, and the disagreement is about the finding itself rather than what to do about it.

Three things distort every ranking of this kind, and knowing them is most of the skill.

First, potency and exposure are different axes. Parabens are weak at the receptor — several orders of magnitude weaker than the body's own estradiol — but they sit on skin for sixteen hours a day and turn up in nearly everyone's urine. DEHP is a stronger anti-androgen but is largely a food-contact and dust exposure. A ranking by potency and a ranking by measured body burden give you different orders, and neither is wrong. That interaction is what the dose-response article covers, including the reason the low-dose end of the curve behaves unlike the textbook.

Second, nothing on the list arrives alone. Silva and colleagues took eight weak xenoestrogens, set each below its own no-effect concentration, and got a substantial estrogenic response from the mixture Silva et al. 2002. Every entry below was evaluated as if it were the only chemical in the room. It never is — see combination effects for what that does to the arithmetic.

Third, a chemical can be genuinely worth avoiding for reasons that have nothing to do with hormones. Formaldehyde is the clean example: an IARCInternational Agency for Research on Cancer Group 1 carcinogen with a solid occupational record, and a fairly poor candidate for the endocrine label. Filing it under 'endocrine disruptor' makes the list longer and the reasoning worse. The Endocrine SocietyThe main professional body for endocrinologists — its 2015 EDC-2 statement is the reference review for this field took the opposite approach in its second scientific statement, working domain by domain rather than assembling a roll call Gore et al. 2015.

Bisphenols. BPA is the most studied synthetic estrogen mimic in existence, and the regulatory record moved further on it than on anything else here: in 2023 EFSA re-evaluated the whole dossier on low-dose immune-effect data and cut the tolerable daily intake by a factor of 20,000×. At the new figure, ordinary dietary exposure exceeds the safe dose for every age group. Exposure itself is not in question — CDC biomonitoring detected BPA in 92.6% of Americans aged 6 and over of the US population Calafat et al. 2008.

The replacements are the interesting part. Rochester and Bolden reviewed the BPA substitutes systematically and found BPS and BPF showing hormonal activity comparable to the chemical they replaced Rochester and Bolden 2015. That is a mechanism-level finding rather than a population-level one, which is exactly why they sit at moderate rather than strong — the human epidemiology on the substitutes is roughly two decades behind the epidemiology on BPA. The BPA alternatives article goes through them one at a time.

Phthalates. The phthalate family splits cleanly. The high-molecular-weight plasticisers — DEHP above all — interfere with fetal testosterone synthesis, producing what toxicologists call the phthalate syndrome in rodents at doses that are not extreme. The low-molecular-weight fragrance carriers like DEP are much weaker but far more widely applied to skin.

The human anchor is Swan and colleagues, who measured prenatal phthalate metabolites in 134 pregnant women and then examined their sons. Mothers in the top quartile of monobutyl phthalate — ordinary consumer exposure, not an occupational cohort — had boys with markedly reduced anogenital distanceA developmental marker of how much testosterone signalling occurred in the womb; shorter distance indicates less masculinisation, an androgen-insufficiency marker Swan et al. 2005. Birth cohorts since have been mixed in size and consistency, which is normal for this kind of measurement, but the direction has held. The broader reproductive picture, including the 51.6% global decline in sperm concentration since 1973 decline in sperm concentration reported by Levine and colleagues Levine et al. 2023, is set out in the chemicals and fertility guide.

Brominated flame retardants. The PBDEs are the entry most people have never heard of and are most likely to be carrying. They were added by the tonne to furniture foam, mattresses and electronics casings from the 1970s until the penta and octa mixtures were withdrawn around 2004. Structurally they resemble thyroid hormone closely enough to displace it from its transport protein in blood, which alters how much reaches tissue without touching a receptor. They are lipophilic and persistent, so they accumulate in fat and in household dust rather than clearing. Two decades after the phase-out they remain measurable in blood and in the dust of homes with older furniture — a good illustration that a chemical leaving the market is not the same as a chemical leaving the population.

PFAS. The PFAS family earns its place on different grounds: not receptor mimicry so much as persistence. PFOS has a serum half-life measured in years rather than hours, so exposure accumulates rather than clearing overnight. Fenton and colleagues reviewed the human and animal literature and found reasonably consistent associations with thyroid hormone disturbance, altered lipid profiles, and reduced vaccine antibody response in children Fenton et al. 2021. The immune findings are arguably the strongest of the three.

The US EPA set enforceable drinking-water limits for PFOA, PFOS and several others in 2024 — the first federal limits for any of these compounds. The newer replacements, GenX among them, were introduced on the argument that shorter chains clear faster. They do. Whether faster clearance offsets similar liver and thyroid activity is not yet answered in humans, which is why they sit at emerging.

This is where the list gets crowded and the potencies get small — a combination that produces most of the overstatement in this subject.

Parabens bind estrogen receptors weakly; butylparaben is the strongest of the common ones and is still thousands of times weaker than estradiol. Darbre and colleagues detected intact parabens in human breast tissue, establishing that skin absorption delivers them into the body without complete metabolic breakdown Darbre et al. 2004 — a finding about absorption, frequently reported as a finding about tumours, which it was not. The EU restricted propyl- and butylparaben concentrations in 2014 and banned five of the longer-chain ones outright.

Triclosan has the clearest thyroid mechanism of the group: repeated rodent work shows circulating thyroxine falling after short exposures, through accelerated liver clearance rather than receptor blocking. The FDA removed it from consumer hand soap in 2016 for a different reason — manufacturers could not show it beat plain soap. Human thyroid evidence remains mixed, so it sits at moderate.

Formaldehyde and its releasers are the entries that most deserve unpicking. DMDM hydantoin is a preservative that slowly releases formaldehyde into the product over its shelf life. That is a real exposure and a real reason to choose differently — formaldehyde is a Group 1 carcinogen and a well-documented sensitiser. But it is not a hormone mimic, and calling it an endocrine disruptor to justify avoiding it weakens a case that stands perfectly well on its own evidence.

Oxybenzone belongs here too, with an honest caveat. FDA pharmacokinetic work confirmed it reaches plasma well above the agency's own threshold for requiring safety data, within hours of a normal application. Its estrogenic activity in cell assays is real. Whether that translates at the concentrations achieved in people is unsettled, and the risk of overstating it is that people skip sunscreen — a trade the evidence does not support.

Atrazine is the sharpest case of a strong mechanism and a contested human picture. Sanderson and colleagues showed that triazine herbicides induce aromatase — the enzyme converting testosterone to estrogen — in human adrenocortical cells, a mechanism that needs no receptor binding at all Sanderson et al. 2000. The amphibian work that followed found gonadal abnormalities in exposed frogs and was contested at length, partly on independent-replication grounds. The EU withdrew atrazine's approval in 2004 over groundwater persistence; the US still permits it and it remains among the most-detected pesticides in American drinking water. Human epidemiology is thin and confounded by everything else in an agricultural exposure — farm workers are not exposed to one compound, and the general population's dose arrives largely through water at concentrations orders of magnitude below anything tested in a laboratory. Strong in amphibians and contested in humans is not a fudge. It is two different questions with two different answers.

Glyphosate is the entry most often placed higher than its evidence supports. The genuine controversy — IARC calling it probably carcinogenic in 2015 while EFSA and others disagreed — is about cancer, not hormones. The endocrine claim is separate and considerably weaker: some cell studies report estrogenic effects, regulatory reviews have not found the endocrine criteria met, and a recurring complication is that the formulated products contain surfactants which behave quite differently from the active ingredient. Contested is the accurate label, and importing certainty from the cancer argument into the hormone argument is a mistake worth avoiding.

Perchlorate is the quiet one. It blocks the sodium-iodide symporter, cutting the iodide supply the thyroid needs to make hormone at all. The mechanism is not in doubt. What varies is whether it matters for you, because the effect is much larger in people who are already iodine-deficient — a genuine interaction rather than a hedge.

Partly — and the endocrine label undersells two of them.

Lead and mercury are established developmental neurotoxicants. Grandjean and Landrigan's review placed both among the small set of industrial chemicals with confirmed developmental neurotoxicity in humans, alongside arsenic and PCBs Grandjean and Landrigan 2014. Lead does also have endocrine effects — delayed puberty in girls is the best-supported, along with disturbance of the growth hormone and thyroid axes — but that evidence is secondary to the neurological evidence, not a substitute for it. If you are reducing lead exposure, the reason is the brain. Filing it under hormones is a demotion.

Aluminium is the contested one, and the honest answer is that the science is not settled. The 'metalloestrogen' hypothesis proposes that aluminium salts interfere with estrogen receptor signalling; most of the supporting work is in cell culture, dermal absorption from antiperspirant is low single-digit percentages at most, and the large epidemiological studies have not found the association the hypothesis predicts. If you would rather not apply it daily, alternatives exist and nothing is lost by switching. Just don't file it beside BPA — the evidence is not in the same class.

Microplastics get listed as endocrine disruptors constantly. They are better understood as the vehicle: a particle carrying plasticisers, bisphenols and adsorbed pollutants across barriers, releasing them with heat and time. The payload is on this list. The particle is a delivery problem, and it deserves its own accounting rather than a slot in the middle of a chemical enumeration.

'Chemical-free' is the other recurring category error. Everything is a chemical, and several of the most potent endocrine-active compounds known are plant-derived — soy isoflavones bind estrogen receptors considerably more strongly than parabens do, which is a fact usually left out of the natural-versus-synthetic framing.

And a note on scale, because the cost of the whole class is easy to state and easy to overstate. Trasande and colleagues estimated the EU health burden attributable to endocrine-disrupting chemicals at roughly €157 billion per year, or about 1.23% of EU GDP annually Trasande et al. 2015. That estimate rests on expert probability judgements about causation, so it carries wide uncertainty — it is a serious calculation, not a measurement.

The useful version of this list is shorter than the list itself. A handful of entries — bisphenols, the anti-androgenic phthalates, PFAS, the brominated flame retardants — carry most of the well-established evidence and most of the measured human exposure. A second tier is mechanistically real and epidemiologically thin. A third tier is on the list mainly because it was on the last list.

Sorting them that way is not a way of letting anything off. It is what makes the strong entries survive an argument. When BPA, aluminium salts and glyphosate are presented as equivalently dangerous, the person who checks the aluminium evidence and finds it thin has every reason to discount the BPA evidence too — and that one is not thin at all.

The practical filter sits in the Eso-Friendly approach: what is in this, how long does it stay in contact, and what does the evidence actually say. For how any of these chemicals interfere in the first place, the mechanism guide is the place to go next.