We reviewed over 700 studies on toxins and their detoxification

We all have a toxin problem. No one is exempt.

The detection statistics leave little room for exceptions. Across 23 European countries, phthalates, PFAS and acrylamide were found in 100% of participants, pesticides and flame retardants in more than 99%, and bisphenols in at least 92% (Govarts et al. 2023). German nationwide testing of 107 substances found 31 of them in [almost] all participants (Umweltbundesamt, 2023). In Flemish teenagers, arsenic, cadmium, lead, PAHs, phthalates, pesticides, PCBs, flame retardants and PFAS were each detected in every subject tested (Schoeters et al. 2022).

Nor is this a matter of where one lives. PFAS were detected in over 99% of residents of St. Lawrence, Alaska (Byrne et al. 2018), and chlorinated pesticides at high concentrations in the liver, brain and fat of Greenland Inuit (Dewailly et al. 1999). Dioxins, furans and PCBs exceeded the limit of quantification in every serum sample in a Lausanne study, whose authors stressed that even in a country with stringent food-safety regulation these contaminants represent a significant public health concern (Oltramare et al. 2025). Some 99% of the global population lives where air pollution exceeds safe limits (Wiebe, Olayiwola, Adebiyi, 2025), and xenoestrogenic compounds have been detected even in rainwater (Thomson, Darwish, 2019).

Nor of age. The average newborn carries 287 chemicals in umbilical cord blood (Environmental Working Group, 2005), and all five PFAS tested were present in every mother and every fetal organ examined (Mamsen et al. 2017). At the other end, autopsy work found PFAS in all human tissues examined across 99 Spanish samples (Pérez et al. 2013) and in the blood, liver, kidney, brain and lung of every deceased individual studied in Denmark (Nielsen et al. 2024).

Nor, largely, of behaviour. The human diet is complex enough that food contaminants cannot be avoided, and exposure remains a permanent health risk (Cuevas-Gonzalez et al. 2020). Diet, housing and occupation shift the magnitude of exposure — sometimes considerably — but they do not produce a clean result.

None of which means everyone is equally affected. Body burden varies, and so does vulnerability: infants, pregnant women and the chronically exposed carry more risk than others. But the question worth asking is not whether you are contaminated. The biomonitoring answers that. It is what you carry, how much, and what follows from it.

It's not only the heavy metals

Lead, mercury, cadmium and arsenic dominate the conversation about toxic exposure, and they earn their place: three of them sit on the WHO's list of the ten chemicals of major public health concern (Bjørklund et al. 2019; Gorini, Tonacci, 2024). But they are a small fraction of what people actually carry.

Over 200 chemicals are neurotoxic to humans, among them PCBs, organophosphate pesticides, organic solvents and brominated flame retardants (Fuller et al. 2022). Across 23 European countries, phthalates, PFAS and acrylamide were detected in 100% of participants, pesticides and flame retardants in more than 99%, bisphenols in at least 92%, and the mycotoxin marker tDON in 96.5% (Govarts et al. 2023). German nationwide testing covering 107 substances found 31 of them in [almost] all participants (Umweltbundesamt, 2023). All 78 persistent organic pollutants tested — dioxins, PCBs, flame retardants, pesticides, furans — appeared in the adipose tissue of all 100 women examined (Ploteau et al. 2016), and 287 different chemicals have been identified in umbilical cord blood, at least 180 of them known carcinogens (Environmental Working Group, 2005).

Nor do substitutes solve the problem: the phthalate replacement DINCH was found in 99% of Europeans (Govarts et al. 2023), BPS shows a hundredfold higher oral bioavailability than BPA in a pig model (Amon, Kek, Klun, 2024), and BPB displays oestrogenic and anti-androgenic activity similar to or greater than BPA (Emanowicz et al., 2024).

This matters because exposures do not act alone. A review of metal, chemical and mycotoxin mixtures found synergistic effects dominating (Liu, Sayes, 2024), and chemicals individually judged non-carcinogenic may still produce carcinogenic synergies (Goodson III. et al. 2015). Removing the metals from the picture leaves most of the picture intact.

We are born pre-poisoned

Exposure does not begin with the first breath. The average newborn carries 287 chemicals in umbilical cord blood — mercury, PFAS, dioxins and furans, pesticides, flame retardants, PCBs and PAHs among them — of which at least 180 are known carcinogens, 217 are toxic to the brain and nervous system, and 208 cause birth defects or abnormal development in animal tests (Environmental Working Group, 2005; Yanev, Chaldakov, 2012).

The placenta is a poor filter. The majority of endocrine-disrupting chemicals cross it by active or passive transport and can be found in amniotic fluid or in placental tissue itself (Dalamaga et al. 2024). Lead crosses freely, with umbilical cord concentrations reaching 80–100% of maternal blood levels (Bjørklund et al., 2019). Transplacental transfer was estimated at roughly 70% for PFOS and 60% for PFOA, with PFOS traces present in every trimester and delivery sample and in almost all cord blood samples (Rovira et al., 2019). Mercury deposits readily in placenta and fetal tissue (Bernhoft, 2012), and polystyrene nanoparticles up to 240 nm have been shown to cross the placental barrier in a human placental perfusion model (Nair et al. 2025).

Fetal tissue itself carries the load. All five PFAS tested were found in every mother and in every fetal organ evaluated, indicating systemic fetal exposure (Mamsen et al. 2017). BPA has been detected in amniotic fluid, umbilical cord, fetal blood and fetal liver tissue (Anwar, Anjum, Ghayas, 2021), with its highest recorded concentrations in placenta and fetus (Jalal et al. 2018). Phthalates circulate through the placenta into fetal blood, where their half-life is extended relative to maternal serum (Genuis et al. 2012a). Microplastics have been found in placentas, meconium and infant faeces (Liu et al. 2023b). Cord blood measurements may in fact understate the burden, since cord blood contains a smaller percentage of lipids than adult blood and most of these compounds travel on serum lipids (Crinnon, 2009).

None of this is inert. Prenatal exposure to relatively low levels of a mixture of placental metals was associated with adverse newborn neurobehaviour, with cadmium carrying the largest weight in the mixture effect (Tung et al. 2022). Prenatal arsenic, cadmium and lead exposures were inversely associated with birth weight, birth length and head circumference (Bocca et al. 2019), and phthalate exposure during pregnancy with significantly higher rates of newborn genital anomalies (Sathyanarayana et al. 2016). Pregnant women exposed to high methylmercury concentrations can give birth to children with severe brain damage while showing no symptoms themselves (Bjørklund et al. 2019).

The consequences also arrive late. The "fetal basis of adult disease" hypothesis holds that in utero exposures can induce effects that only reveal themselves decades later — exposure to a reproductive toxicant such as dibutyl phthalate before birth may surface as infertility in adulthood (Barr, Bishop, Needham, 2007). Prenatal disruption of gene expression, chromatin accessibility and DNA methylation persists into adulthood as a form of silent reprogramming (Zhao et al. 2025), and birth cohort studies link prenatal BPA, phthalate and heavy metal exposure to differential DNA methylation in cord blood correlating with higher BMI and adiposity in childhood and adolescence (Alum et al. 2025b).

We do not acquire a chemical body burden over a lifetime. We inherit one, and then add to it.

The problem is not acute toxicity

When people picture "toxic," they think of acute poisoning — a large dose causing obvious, immediate harm. But that's not where the main threat lies today. On an individual level, everyday exposure to these chemicals rarely causes obvious illness and very rarely causes outright death (Rovira, Domingo, 2018). The problem is the opposite kind: small amounts, absorbed constantly, over years.

It builds up. Many toxins are stored in the body — in fat, bone, kidneys, brain — and cleared very slowly, so tiny daily doses accumulate over a lifetime rather than being flushed out (e.g. cadmium persists for decades; Liu et al., 2023). The body keeps recycling some of them rather than excreting them (Crinnon, 2009).

It's silent. Serious damage can happen without any warning symptoms. The clearest example: mothers exposed to mercury can have children with severe brain damage while feeling completely fine themselves (Bjørklund et al., 2019). More generally, symptoms may be absent despite significant poisoning (Blanuša et al., 2005), and chronic low-level poisoning often produces no obvious signs, which delays treatment (Sun et al., 2023).

It's delayed. Effects can surface years or even decades after exposure (Rozman et al., 1982) — or in the next generation rather than the exposed person.

"Safe" doses may not be safe. Because harm comes from long-term, low-level, mixed exposure, the traditional idea that "the dose makes the poison" breaks down. For some chemicals there's no threshold below which they're harmless (lead being the prime example; Fuller et al., 2022), and effects can even appear at low doses that don't show up at high ones (Mallozzi et al., 2016).

If the danger were acute poisoning, you'd simply avoid big doses. But because it's slow accumulation of many small exposures, the response has to be different: reducing ongoing exposure and helping the body clear its stored burden.

Invisible killers, acting at ultra-low doses

Invisible harm without perceptible symptoms or warning. Toxic damage can proceed silently: As we just learned, poisoning can be asymptomatic even when significant. But — unfortunately — there is more to the story.

Effects are subtle and hard to attribute. Many of the effects of contaminants on humans are subtle and difficult to quantify — particularly developmental, nervous-system, and hormonally-regulated effects, many of which are irreversible (Carpenter et al., 1998). POPs are described as having the potential to produce subtle toxic effects (Ahmed et al., 2021), and POP disruptions in early life can lead to subtle functional changes that may not emerge until later in life or even later, to the next generation (Purdel et al., 2014).

Even the "healthy" carry a burden. Even the non-exposed, asymptomatic population is exposed, as revealed by metal provocation urine testing (Blaurock-Busch, 2011), and standard blood/urine tests understate the accrued lifetime body burden (Sears, 2013).

Latency of years to decades. These chemicals may produce toxic effects years or decades after exposure (Rozman et al., 1982) — for PBDEs specifically, the identification of risk exposure might not be visible in the short-term (Renzelli et al., 2023). Mercury's low-grade chronic effects (fatigue, anxiety, depression, paresthesias, memory loss, poor concentration) are noted to be commonly missed in differential diagnosis (Bernhoft, 2012), and cadmium is linked to subclinical myocardial injury even in people with no atherosclerosis or chronic kidney disease (Scimeca et al., 2024).

"Killers" — associated with mortality

WHO estimated that about half of the 2 million deaths attributable to chemical exposures in 2019 were associated with lead (Gorini, Tonacci, 2024); the Lancet Commission puts lead-related deaths at ~900,000/year, likely an undercount (Fuller et al., 2022).

Cadmium is toxic at low levels, with systematic reviews documenting associations with increased all-cause mortality, cancer mortality, and cardiovascular disease (Minderoo-Monaco Commission, 2023); in a >6,000-participant cohort it was the metal with the strongest association with both cardiovascular disease and all-cause mortality (Martinez-Morata et al., 2024).

Air-pollution particulate matter contributes to millions of premature deaths annually (Anderson, Thundiyil, Stolbach, 2012; Fuller et al., 2022), and global chemical pollution is tied to over 1.8 million deaths from toxic chemical exposure (Zhao et al., 2025).

Mixture toxicity can produce devastating effects, including death, from pollutants at low or minute concentrations (Lagunas-Rangel et al., 2022).

"Ultra-low doses" — harm below conventional thresholds:

The WHO and independent research has determined that no level of lead in blood is safe (Minderoo-Monaco Commission, 2023)(Liu et al., 2023); and specifically there is no safe detectable level of lead in the blood of children, with blood lead <5 µg/dL accounting for 74% of the total IQ-loss cost burden (Boyle et al., 2021). The European Food Safety Authority, EFSA, concluded the provisional tolerable weekly intake for lead was no longer precautionary due to the lack of evidence for a threshold for Pb toxicity (Gorini, Tonacci, 2024), and lead even at very low doses raises blood pressure and cardiovascular risk (NTP, 2012; Li, Zhao, 2024; EFSA, 2010), with renal impairment already at the lowest levels (NTP, 2012; Harari et al., 2018).

Adverse effects documented at very low doses across toxin classes. Cadmium adverse effects appear with exposures as little as 15 µg/day, well within levels human populations encounter, and it disrupts homeostasis even at very low concentrations (Hafey et al., 2022; Vijiyakumar, Prince, 2024). Heavy metals generally (Pb, Cd, Hg, As, Cr) even at low doses are considered extremely harmful (Panaiotov et al., 2024); continued low-level exposure to As, Cd, Hg, Pb is tied to numerous adverse effects (Rovira, Domingo, 2018). Arsenic impairs memory and intellect even at low levels (Domingo-Relloso et al., 2024). Parabens can have adverse effects even at low concentrations, with DNA damage at concentrations an order of magnitude lower than required for transactivation (Wei et al., 2021). BPS, even at very low levels, can impact brain function (Emanowicz et al., 2024). POPs have disruptive effects even at the low exposure levels (Hu et al., 2021).

The low-dose / non-monotonic paradigm (the mechanistic backbone):

The classical toxicological paradigm — the dose makes the poison — fails for these compounds: harmful effects appear at doses lower than those usually studied, and EDCs can show non-monotonic dose-response, with the EPA endorsing a linear no-threshold model for carcinogens since cancer risk exists even after damage to a single cell (Costopoulou et al., 2025). EDCs cause effects at doses below the LOAEL, short for 'lowest observed adverse effect level', producing effects not predicted from high-dose studies (Gómez-Olarte et al., 2024; from the Palanza framework). The BPA demonstration is central: effects on metabolic-syndrome endpoints appeared at and below the predicted NOAEL, short for 'no observed adverse effect level', but were absent at the highest dose tested — without testing below the NOAEL, no effect would have been found (Mallozzi et al., 2016). For aluminum adjuvant, the smallest dose (200 µg/kg) — not the higher ones — was neurotoxic, leading the authors to conclude the dose does not always make the poison (Rubio-Casillas, Redwan, Uversky, 2022). PBDEs likewise show effects below the LOAEL and non-monotonic curves (Renzelli et al., 2023).

"Something from nothing" — mixtures push individually-safe doses into harm:

Because essentially everyone carries many compounds simultaneously, single compounds below the NOAEL in isolation can exert toxic effects at extremely low concentrations when combined, via additivity or synergy — as is usually the case in almost every human being (Thompson et al., 2009). Substances can display additive or synergistic effects even when each component is below the NOAEL (Gómez-Olarte et al., 2024; the "multi-headed dragon," Bloch et al., 2023). Cumulative effects at exposure levels beneath established thresholds can occur — the "something from nothing" phenomenon (Domingo, 2026). An 18-month study confirmed that exposure to very low doses (below the NOAEL) of a mixture induced histopathological lesions and cytotoxic effects (Dinca et al., 2023). And this may explain how pollutants with low or minute concentrations can promote disease and have devastating effects, including death (Lagunas-Rangel et al., 2022). One striking synergy example: a mycotoxin (OTA) combined with ABT — which alone had no effect — produced hepatotoxicity several orders of magnitude stronger than OTA alone (Bloch et al., 2023).

Pollution is among the leading causes of death worldwide

Global pollution is estimated to cause around 9 million premature deaths a year, with toxic chemical exposure specifically contributing over 1.8 million of those (Zhao et al., 2025).

The document frames pollution as the largest environmental cause of disease and premature death — responsible for far more deaths than war, terrorism, malaria, HIV, tuberculosis, drugs, or alcohol (the Lancet Commission framing; Fuller et al., 2022).

Roughly 2 million deaths in 2019 were attributed to chemical exposures specifically, and about half of those were linked to lead alone (Gorini, Tonacci, 2024).

Lead is tied to roughly 900,000 deaths a year, mostly in poorer countries — and the document notes this is probably an undercount, because newer research suggests lead harms the heart and kidneys at even lower levels than previously thought (Fuller et al., 2022).

Cadmium is associated with increased death from all causes, from cancer, and from heart disease (Minderoo-Monaco Commission, 2023; Martinez-Morata et al., 2024).

Air pollution (fine particles) is a major contributor to the overall toll, tied to millions of premature deaths and worse outcomes from infections (Fuller et al., 2022).

The true toll is almost certainly higher than the official figures, because only a tiny fraction of chemicals in use have been properly tested, so the deaths they cause simply can't be counted yet (Zhao et al., 2025; Trasande et al., 2016). In that sense the document treats the headline numbers as a floor, not a ceiling.

The honest caveat. These are modeled estimates, not body counts — they come from statistical models linking exposure levels to disease risk, so the exact figures carry real uncertainty (which is why sources give wide ranges, e.g. lead deaths estimated between roughly 550,000 and 1.3 million). Pollution also usually kills indirectly, by raising the risk of heart disease, stroke, and cancer, rather than being listed as the direct cause on a death certificate. So the accurate reading is that pollution is a top-tier contributing cause of global death — comparable to or exceeding several of the world's most-recognized killers — rather than a single, precisely-counted one.

Linked to virtually every form of modern disease

Read across the environmental health literature and a striking pattern emerges: there is barely a category of chronic modern illness without a pollutant somewhere in its causal literature.

Neurological and psychiatric. Over 200 chemicals are neurotoxic to humans (Fuller et al. 2022). Persistent organic pollutants are associated with Parkinson's disease, Alzheimer's disease, stroke, epileptic seizures, multiple sclerosis, dementia and ADHD (Iqubal et al. 2020). Environmental pollution has been linked to shortened lifespan, increased chronic disease, neurodevelopmental problems and increased dementia risk (Fahey et al. 2025), and air pollution is significantly associated with major depressive disorder (Borroni et al. 2024). Toxic causes of mental illness are, as one review puts it, overlooked (Genuis, 2008).

Metabolic and cardiovascular. A whole subclass of compounds is now designated metabolism-disrupting, implicated in obesity, insulin resistance and type 2 diabetes (Dalamaga et al. 2024). High POP levels are associated with cardiometabolic risk factors, obesity and type 2 diabetes (Rouhou et al. 2016; Mustieles, Arrebola, 2020), and POPs in air alongside particulate matter raise the risk of several chronic diseases (Scimeca et al. 2024). Low-level environmental cadmium is associated with all-cause mortality, cardiovascular mortality and cancer (Zwolak, 2025).

Renal, hepatic, haematological, skeletal, respiratory. Chronic heavy metal exposure manifests as neurological deterioration, cardiovascular disease, reproductive problems, nephropathy, respiratory dysfunction, bronchitis, pulmonary oedema, asthma, emphysema, hepatitis, anaemia, hyperpigmentation and cancer (Koyama, Kamogashira, Yamasoba, 2024). Chronic cadmium exposure produces renal dysfunction, anaemia and hepatic damage, along with urinary stone disease, hypertension and osteoporosis (Flora, Pachauri, 2010).

Immune and microbiome. Immunotoxicity ranks among the three most worrying and least charted consequences of chemical pollution (Fuller et al. 2022). PFAS are associated with allergies, susceptibility to infection, immunosuppression and autoimmune disease (Thoerig et al., 2025); early-life POP exposure reduces the capacity to fight infection and raises allergic asthma risk (Mallozzi et al. 2016). Particulates and xenobiotics significantly impact the respiratory and gastrointestinal microbiome through inflammation, oxidative stress and gut barrier dysfunction (Lang, Lipp, Wechselberger, 2025), and microplastic exposure markedly alters gut structure, giving rise to comorbidities and chronic conditions (Abbasi et al. 2025).

Reproductive and developmental. Sperm concentration in Western countries fell 50–60% between 1973 and 2011 (Levine et al. 2017; Stavros et al. 2025); about one in six US children now has a neurodevelopmental disorder, with toxic chemicals among the important causes (Minderoo-Monaco Commission, 2023).

Cancer. Cadmium, arsenic, benzene and aflatoxin B1 are IARC Group 1 human carcinogens (Liu et al. 2023; Umweltbundesamt, 2023; Fu et al. 2022), with further associations spanning colorectal (Lee et al. 2018), bile duct (Dzierzyński et al. 2024), prostate (Kahn et al. 2020) and thyroid cancer (Renzelli et al. 2023).

One caveat deserves stating plainly, because it is stated in the primary literature itself. Some effects — kidney disease following lead exposure, loss of specific neurons after methylmercury exposure — are clearly attributable to a particular exposure. Others, including cancer, birth defects and many endocrine and nervous system actions, are difficult to ascribe with certainty to any single exposure (Carpenter et al. 1998). "Linked to" is therefore doing real work in the heading above: for most of these conditions, pollutants are one contributing cause among several, not the cause. What the breadth of the evidence establishes is not that chemicals explain modern disease, but that there are very few modern diseases where they can safely be left out of the account — and that a system exposed continuously, from before birth, to mixtures acting synergistically (Liu, Sayes, 2024) is unlikely to have partitioned the damage neatly by organ.

Switzerland is not exempt

Switzerland — despite its reputation for clean environment and strict regulation — shows the same near-universal population burden documented elsewhere.

Heavy Metals

The national biomonitoring study (n = 789 blood samples) found essentially the whole population carrying a measurable metal burden:

Lead in 99% of blood samples, with 3% exceeding the safety threshold (BAG, 2023)

Cadmium in 97% of samples (BAG, 2023); a separate study of 1,400 participants detected Cd in all (100%) urine samples (Jenny-Burri et al., 2015)

Mercury in 85% of blood samples (BAG, 2023)

Arsenic in 76% of blood samples (BAG, 2023)

A separate Swiss population-based cohort of >900 participants (Perrais et al., 2024) corroborated this, quantifying cadmium in 100% of plasma and 98% of urine, lead in 100% of plasma, arsenic in 96% of plasma and 100% of urine, mercury in 98% of plasma, and aluminum in 89% of urine.

PFAS

The 2023 Swiss Human Biomonitoring study detected four PFAS (PFOA, PFNA, PFHxS, PFOS) in ~100% of blood samples and two more (PFDA, PFHpS) in ~90% (BAG, 2023). A pilot of the Swiss Health Study found PFAS in all 630 serum samples (100%), concluding that the ubiquitous and worldwide presence of PFASs currently represents a major and worrying public health concern (Jaus et al., 2025). Environmental contamination is documented too: a 2025 analysis found significant PFAS contamination across all fish species sampled from various Swiss lakes (Lucarini, Staedler, 2025), and in most cases a 70 kg person eating a 200 g filet of any of those fish would already exceed the EFSA tolerable weekly intake (Soudani et al., 2024; Lucarini, Staedler, 2025).

Bisphenols (BPA)

A Swiss biomonitoring study found bisphenols in the urine of infants in 47% of samples collected, concluding that toddlers and infants are extensively exposed to BP derivatives (Lucarini et al., 2020). Switzerland was also one of the 23 countries in the EU-wide HBM4EU survey (>10,000 participants) that detected bisphenols in ≥92% of participants, with more than 11% exceeding the health-based guidance values (Govarts et al., 2023).

Pesticides

Switzerland was likewise included in HBM4EU, which detected pesticides in >99% of all participants across the 23 countries, with children generally more exposed and reaching toxicologically relevant levels for one compound in 36% of cases (Govarts et al., 2023).

Taken together, the Swiss data span every major toxin class the document treats — metals, dioxins/furans/PCBs, PFAS, bisphenols, and pesticides — and consistently show detection rates at or near 100% in the general population, including infants, with a fraction of samples already crossing safety thresholds. The document's own conclusion, in the words of the cited Lausanne authors, is that a country with stringent food-safety and environmental regulation is precisely where these findings register as "a significant public health concern" (Oltramare et al., 2025). In other words, the document uses Switzerland as a demonstration that even best-case regulatory environments do not confer exemption — the burden is ubiquitous.

The problem with guidelines and regulation

Regulatory thresholds carry an implicit promise: stay below this number and you are safe. Several features of how those numbers are produced make the promise difficult to keep.

Most chemicals were never adequately assessed. Only a small fraction of the many thousands of manufactured chemicals in commerce have been adequately tested for safety or toxicity, which means the disease burdens attributable to them cannot be quantified and undercounting is probably substantial (Fuller et al. 2022). Some compounds remain unclassified even where concern is documented — there is currently no harmonised classification and labelling under the CLP Regulation for certain substances still undergoing evaluation for endocrine activity (Umweltbundesamt, 2023). In other words, ~350,000 widely-used chemicals may enter the environment, but only ~500–1,000 are regulated — under 1% (Zhao et al., 2025). In the EU, adequate safety/toxicity data exist for only ~14% of ~100,000 registered substances (Yanev, Chaldakov, 2012).

The numbers keep falling. In 2010 the EFSA CONTAM Panel concluded that the provisional tolerable weekly intake of 25 µg/kg bw for lead was no longer precautionary, because there is no evidence of a threshold for lead toxicity (Gorini, Tonacci, 2024). EFSA subsequently proposed lowering the tolerable daily intake for BPA by a factor of 10⁵ — from 4 µg/kg bw/day to 0.04 ng/kg bw/day — with the consequence that almost all adults exceed the correspondingly revised guidance value, where only 11% had exceeded the previous one (Govarts et al. 2023). A guidance value that moves by five orders of magnitude was never a measure of safety; it was a measure of what was known at the time.

Different bodies, different numbers. Reported no-observed-adverse-effect levels vary by two to three orders of magnitude across studies, reflecting differences in metabolic processing, genetic susceptibility and methodological approach (Han, Jin, 2025). For carcinogens, the US EPA holds that a linear no-threshold model should apply, since cancer risk exists even after damage to a single cell (Costopoulou et al., 2025) — a position that sits awkwardly beside threshold-based limits used elsewhere.

The testing paradigm misses low-dose effects. Endocrine disruptors frequently exhibit non-monotonic dose-response, where low-dose chronic exposure produces more pronounced effects than acute high-dose exposure (Han, Jin, 2025). In one BPA study, metabolic effects were absent at the highest dose examined — 50,000 mg/kg/day — and appeared only below the established NOAEL; without testing beneath it, the investigators would have found no effect at all (Mallozzi et al. 2016). Health effects also depend not only on dose but on the life stage at which exposure occurs (Costopoulou et al., 2025), and exposure to a mixture at doses below the NOAEL for 18 months still induced histopathological lesions and cytotoxic effects (Dinca et al. 2023).

Limits are set one substance at a time. Real exposure is simultaneous. A review of metal, chemical and mycotoxin mixtures found synergistic effects dominating (Liu, Sayes, 2024); mixtures of five parabens suppressed androgen receptor activation at hundredfold lower concentrations than concentration-addition predicted (Martin et al. 2021); chemicals individually judged non-carcinogenic may produce carcinogenic synergies (Goodson III. et al. 2015); and an essentially no-response dose of a mercury salt combined with one-twentieth of the equivalent dose of a lead salt killed all animals tested (Schubert, Riley, Tyler, 1978). Investigators studying phthalate co-exposure concluded plainly that adverse health risks cannot be excluded (Govarts et al. 2023).

Restriction invites substitution rather than removal. While levels of legacy phthalates appear to be decreasing, those of their regrettable replacements are rising rapidly (Frederiksen et al. 2019) — DINCH now reaching 99% of Europeans (Govarts et al. 2023) — and BPS, a BPA substitute, itself exceeded its guidance value in 11% of adults (Govarts et al. 2023).

In practice, guidance values are exceeded routinely. Health-based guidance values were exceeded for at least 15 substances in EU-wide biomonitoring (Govarts et al. 2023); cadmium reached levels of concern in 16% of European adults, 42% in France and 36% in Germany (Tratnik et al. 2022); and breastfed infants exceeded the tolerable intakes for BPA, PFOS, PFOA and dioxin-like PCBs across all lactation periods (Rovira et al., 2022).

Perhaps the sharpest verdict on the framework comes from David Rall, former director of the US National Institute of Environmental Health Sciences, who observed that if thalidomide had caused a ten-point loss of IQ instead of obvious birth defects of the limbs, it would probably still be on the market (Minderoo-Monaco Commission, 2023).

None of this argues that thresholds are useless — they are how exposure is managed at population scale, and the downward revisions above are regulators doing their job as evidence accumulates. It argues something narrower: that a number below a limit describes compliance, not safety, and that the gap between the two is where most of the burden currently sits.

Mixtures create effects that regulation ignores entirely

Safety limits are usually set one chemical at a time, testing each in isolation. But real people carry dozens of pollutants at once. Chemicals which look harmless on their own can combine to cause real harm — sometimes just adding up, sometimes multiplying each other's effects (synergy) — even when each one sits below its own "safe" level (Thompson et al., 2009; Gómez-Olarte et al., 2024). Which leads us to the "something from nothing" problem: harm emerging from a mix of individually-safe doses (Domingo, 2026).

Between 2000 and 2020, about half of animal studies of low-dose chemical mixtures found bigger effects when the chemicals were combined than expected (Gómez-Olarte et al., 2024). The document argues this makes the old "safe threshold" approach inadequate, since meeting the limit for each chemical separately doesn't guarantee protection against the mixture (Lagunas-Rangel et al., 2022; Bloch et al., 2023).

Metals together are worse. Mice given lead, mercury, and cadmium as a mix showed worse memory and coordination problems than any single metal alone (Pyatha et al., 2023). A four-metal mix at levels each permitted in U.S. drinking water caused clear brain and behavior problems in young mice that the individual metals didn't (Chandra et al., 2026). Combining mercury with uranium, or lead with mercury, produced far more kidney or brain damage than adding the separate effects would predict (Domingo, 2026; Karri, Schuhmacher, Kumar, 2016).

In children: An Italian study of 300 schoolchildren found lead and arsenic together raised the risk of attention, aggression, and behavior problems beyond what each metal caused alone — the authors specifically warned the combination amplifies harm even at low levels (Renzetti et al., 2021).

Fertility and the sperm decline

Worldwide, infertility affects around 8–12% of couples, with male factors the primary cause in 50% of cases and around 7% of all men affected; the WHO has designated infertility a public health priority (Kumar, Singh, 2022; Stavros et al. 2025). The trend is what draws attention. A comprehensive meta-regression reported a significant decline in sperm counts between 1973 and 2011, driven by a 50–60% fall among men unselected by fertility from North America, Europe, Australia and New Zealand (Levine et al. 2017), and the most recent meta-analyses confirm that the decline has continued into the twenty-first century (Stavros et al. 2025).

Why environmental exposure is a prime suspect. Spermatogenesis and steroidogenesis are tightly regulated processes dependent on accurate hormonal signalling, which makes the male reproductive system unusually susceptible to endocrine-disrupting chemicals (Stavros et al. 2025). These compounds act through oxidative stress, apoptosis, receptor-mediated disruption of oestrogenic and androgenic signalling, alterations in the hypothalamic–pituitary–gonadal axis, and heritable epigenetic changes — impairing sperm concentration, motility, morphology, DNA integrity and steroidogenesis, with biomarkers of exposure correlating with reduced semen quality, hormonal imbalance and infertility (Stavros et al. 2025). Environmental pollution has emerged as a major cause of rising male infertility, unfavourably affecting semen quality by impairing spermatogenesis, steroidogenesis and Sertoli cell and sperm function (Kumar, Singh, 2022). Disruption of androgen activity impairs development of the male reproductive tract and external genitalia and may lead to incomplete masculinisation (Mallozzi et al. 2016), with epidemiological studies connecting exposure indicators to lower semen quality, hypospadias, cryptorchidism and hormonal abnormalities (Stavros et al. 2025). The sex-steroid disruptive effects of xenobiotics have been proposed as causes of the reported decline in sperm count and general fertility (Carpenter et al. 1998).

The individual compounds. Phthalates produce the constellation known as phthalate syndrome — diminished anogenital distance, low sperm count, undescended testes, hypospadias and infertility (Kumar, Singh, 2022) — with decreased semen quality their most consistently observed effect (Chang et al. 2021). BPA is a reproductive toxicant that disrupts spermatogenesis, acts as an androgen receptor antagonist at high concentrations, affects Leydig cell steroidogenesis and interferes with LH receptor-ligand binding (Peretz et al. 2014; Matuszczak et al. 2019); among couples seeking infertility treatment, BPA was found in 98% of urine samples, with levels negatively correlated with sperm count and motility (Matuszczak et al. 2019). PFOA impairs sperm motility and penetration and is longitudinally associated with lower sperm concentration and count and higher LH and FSH in young men (Fenton et al. 2021), while PFAS levels correlate inversely with testosterone (Rahman, Stokey, Munson-McGee, 2026). All three studies examining organophosphate pesticides and semen quality, and all four examining PFAS, reported negative associations (Kahn et al. 2020). Prepubertal dioxin exposure was significantly associated with reduced sperm concentration and motility, dioxin and furan content ran 2.2–2.3 times higher in the ejaculate of infertile than fertile men, and male infants fed breast milk from women exposed to high dioxin concentrations around conception showed significantly decreased sperm concentration, total count and motility (Kumar, Singh, 2022). Chronic exposure to the household flame retardants TDCPP and TPP was associated with a 19% reduction in sperm concentration (Han, Jin, 2025), and PBDEs have been linked to cryptorchidism, testicular cancer, low testosterone and poor semen quality (Domingo et al. 2025). Microplastics negatively affect male fertility and sperm quality (Barceló, Picó, Alfarhan, 2023), with polystyrene particles inducing male reproductive dysfunction in mice (Jin et al. 2021) and interfering with testosterone synthesis (Liu et al., 2026).

Female fertility. The picture is equally broad: altered menstrual cyclicity, altered pubertal timing, disrupted ovarian folliculogenesis and steroidogenesis, primary ovarian insufficiency, polycystic ovary syndrome, endometriosis, leiomyomas, poor IVF outcomes and poor pregnancy outcomes (Land et al. 2025). Mean serum BPA was three times higher in women with a history of three or more consecutive first-trimester miscarriages than in non-parous women without fertility problems (Meeker, Sathyanarayana, Swan, 2009). Cadmium causes hormonal abnormalities and infertility in both sexes, and is associated with low birth weight, premature delivery and spontaneous miscarriage (Milancović et al. 2024). Women environmentally and occupationally exposed to mercury showed reduced fecundability, though some studies found no association (Henriques et al. 2019). Prenatal and early postnatal chemical exposure also appears to raise the incidence of reproductive disease later in life, including endometriosis and cancers of the breast, cervix, uterus and testis (Fuller et al. 2022).

The generational loop. After pregnant animals were exposed to endocrine disruptors found in common fungicides and pesticides, fertility alterations persisted through all four successive generations examined (Genuis, 2006), and epigenetic alterations within germ cells carry multigenerational consequences for the reproductive system (Thévenod, Lee, 2024). In utero exposure to a reproductive toxicant such as dibutyl phthalate may produce outcomes that only surface as adult infertility (Barr, Bishop, Needham, 2007).

One honest qualification: how much of the measured sperm decline is attributable to endocrine disruptors specifically remains unclear, even if the contribution appears substantial (Levine et al. 2017), and reviewers note that further work is needed to confirm the impact of these chemicals on fertility (Kumar, Singh, 2022). Obesity, smoking, heat, sedentary behaviour and methodological changes in semen analysis all sit in the same causal frame. What is not in dispute is the direction of travel, nor that compounds demonstrably capable of producing these effects are present in essentially everyone.

Epigenetic reprogramming that carries across generations

Pollution rewrites our genes — and the changes can pass to our children

The basic idea. Pollutants don't just poison the body directly. They can also flip "switches" that control which genes are turned on or off. The underlying DNA sequence stays the same, but the settings change — and some of those altered settings get copied into eggs, sperm, and offspring. So the exposed person isn't the only one affected; their children and grandchildren can inherit the damage too (Han, Jin, 2025; Zhao et al., 2025). One striking feature is that the change can stay hidden — it "primes" genes to misbehave later rather than causing obvious problems right away (Zhao et al., 2025).

Where the idea came from. It traces back to wildlife studies around the Great Lakes, where the reproductive problems showed up not in the polluted adults but in their offspring (Crinnon, 2009). In a much-cited animal experiment, pregnant animals exposed to common pesticides and fungicides produced offspring with fertility problems that continued through four generations (Genuis, 2006).

The clearest examples:

Fertility. In lab animals, sons of exposed parents showed sharply reduced sperm counts and testosterone, and the effects carried into grandchildren and great-grandchildren before slowly fading (Han, Jin, 2025). BPA can also damage the egg cells forming in a female fetus, meaning effects can reach her future children (Mallozzi et al., 2016).

Obesity and metabolism. This is the best-studied strand. Exposure in the womb to chemicals like BPA and arsenic can reset how the body handles fat and blood sugar, and children of more heavily exposed mothers tend to have higher body weight — with some of these effects passing to the next generation (Alum et al., 2025b; Minderoo-Monaco Commission, 2023). One estimate blamed prenatal BPA for about 42,400 new childhood-obesity cases a year (Trasande et al., 2015).

Why the document treats this as especially serious. The ability to harm future generations is one of the worst features of these chemicals, since it reaches people who were never directly exposed (Crinnon, 2009). That's why reducing exposure early is paramount — the damage can echo across generations (Sargis, Heindel, Padmanabhan, 2019).

The honest caveat. The strong, multi-generation proof comes mostly from animals. In people, the evidence is consistent but weaker — mainly because tracking families across several generations takes decades, so the human studies haven't caught up yet (Han, Jin, 2025; Alum et al., 2025b). So this is best read as well-established in animals and biologically plausible in humans, rather than fully proven in humans.

Children bear a disproportionate and irreversible burden

Childhood exposure is not a smaller version of adult exposure: children are both more exposed and more damaged by what reaches them.

Greater exposure. Infants in the womb and young children are at particularly high risk of plastic-related health effects, owing to the sensitivity of early development and to children's unique exposure patterns (Minderoo-Monaco Commission, 2023). Flame retardants leach from furniture and electronics into house dust, reaching young children who crawl on the floor and show age-appropriate hand-to-mouth behaviour (Minderoo-Monaco Commission, 2023), and babies and young children are the most vulnerable group with respect to phthalates because of their developmental status (Martínez et al. 2018). Breast milk adds a further route: PBDEs transfer during lactation, within critical developmental windows (Domingo et al. 2025), and breastfed infants exceeded tolerable intakes for BPA, PFOS, PFOA and dioxin-like PCBs across all lactation periods (Rovira et al., 2022). Even historical maternal exposure counts — lead stored in bone is released during pregnancy and lactation and transferred to the infant (Gulson et al. 2003; Peraza et al., 1998).

Greater vulnerability. Low-dose exposure to neurotoxic chemicals during key developmental periods has more serious effects than high-dose exposure in adults (Minderoo-Monaco Commission, 2023; Fuller et al. 2022). In adults, fat-soluble pollutants cross the blood–brain barrier; in a developing nervous system that barrier is not yet formed (Crinnon, 2009), and the fetal nervous system is estimated five times more susceptible to developmental neurotoxicity than the adult (Iqubal et al. 2020). Developing fetuses and neonates are the most vulnerable to endocrine disruption (Amon, Kek, Klun, 2024), and pregnancy is a critical window for lead, mercury, cadmium and arsenic (Irawati et al. 2025).

Irreversibility. Since 1979 it has been accepted that lead at remarkably low concentrations causes IQ decrements and behavioural problems in children exposed prenatally and in early postnatal years, with more recent work suggesting these actions are irreversible; prenatal PCB exposure produces cognitive and behavioural decrements that likewise appear irreversible, and developmental alterations are as a class irreversible, as are many effects on the nervous system and hormonally regulated organs (Carpenter et al. 1998). Brain damage from early-life exposure may be chronic, irreversible and difficult to treat, which is why preventing exposure is the most effective strategy for safeguarding the developing brain (Minderoo-Monaco Commission, 2023). German guidance values were themselves derived from irreversible effects in children: impaired cognitive test performance, hearing impairment, reduced antibody titres (Umweltbundesamt, 2023).

Delayed and inherited. Disruptions induced in early life can produce functional changes that emerge only much later, or in the next generation (Purdel et al. 2014) — in utero exposure to dibutyl phthalate may surface only as adult infertility (Barr, Bishop, Needham, 2007). Prenatal and perinatal disruption of gene expression, chromatin accessibility and DNA methylation persists into adulthood as silent reprogramming (Zhao et al. 2025), with birth cohort studies linking prenatal BPA, phthalate and metal exposure to differential cord-blood DNA methylation correlating with higher BMI in childhood and adolescence (Alum et al. 2025b). Epigenetic alterations in germ cells carry multigenerational consequences (Thévenod, Lee, 2024), and fertility alterations persisted through all four successive generations examined after pregnant animals were exposed to endocrine disruptors found in common fungicides and pesticides (Genuis, 2006).

What it adds up to. About one in six US children has a neurodevelopmental disorder, with toxic chemicals among the important causes; plastic-associated exposures are linked to prematurity, stillbirth, low birth weight, birth defects of the reproductive organs, neurodevelopmental impairment, impaired lung growth and childhood cancer (Minderoo-Monaco Commission, 2023). Prenatal organophosphate exposure alone costs Europe an estimated 13.0 million IQ points and 59,300 additional cases of intellectual disability annually (Trasande et al. 2015), and lead accounts for an estimated 30% of the global burden of idiopathic intellectual disability (Gorini, Tonacci, 2024).

The asymmetry is the point. An exposed adult may clear a pollutant; a child exposed during a developmental window may carry it permanently, and pass part of it on.

One qualification: breastfeeding remains recommended despite pollutants in breast milk, given its nutritional and immune benefits (Purdel et al. 2014). The case here is for reducing what enters the mother, not for changing how she feeds her child.

The economic costs are enormous, hidden and underestimated

Environmental exposure is usually framed as a health story. It is also a balance-sheet story, with numbers large enough to rival entire categories of public spending.

The headline figures. The burden of disease attributable to endocrine-disrupting chemicals in the European Union amounts to a median annual cost of €163 billion, or 1.28% of EU GDP, with a disease-contribution probability greater than 99% (Attina et al. 2016). In the United States the equivalent figure is $340 billion per year — 2.33% of GDP (Attina et al. 2016). For perspective, that is roughly one sixth of the €798 billion European cost of brain disorders in 2010 (Trasande et. al., 2015). Lead alone has been estimated to contribute $6 trillion to global health expenditures, 77% of it from cardiovascular morbidity and mortality and 23% from IQ loss (Salamanca-Fernández et al. 2025), with lead-related IQ losses driving global economic losses of almost $1 trillion annually (Fuller et al. 2022).

Individual compounds carry striking sums. In the US, BPA-attributable costs reach $165.9 billion for coronary heart disease and $62.4 billion for stroke; DEHP-attributable deaths imply a welfare cost of $490 billion; PBDE exposure in a single birth cohort cost $202 billion, or 1.1% of GDP (Minderoo-Monaco Commission, 2023). Phthalate-attributable adult obesity runs to €15.6 billion per year in Europe (Trasande et. al., 2015), and EDC-linked male reproductive disorders to nearly €15 billion (Hauser et al. 2015).

Why they are hidden. These costs appear in no product's price. They surface as medical spending, lost productivity, special education and diminished lifetime earnings, arriving decades after the exposure and attaching to no identifiable transaction: lead-related IQ losses register as school failure, behavioural disorders and reduced productivity (Fuller et al. 2022) — real expenditures that no accounting system traces to their source.

Why they are underestimated. The investigators say so themselves. The EU estimates represent only those endocrine disruptors with the highest probability of causation, and a broader analysis would have produced greater estimates of disease burden and cost (Trasande et al. 2016). Their own summary is blunter: the work surely represents a substantial underestimate of actual EDC-attributable disease, given its focus on fewer than 5% of endocrine disruptors, its examination of only a subset of health effects, and its exclusion of human suffering and other societal costs — while biomarker-based studies suffer from exposure imprecision that understates the true exposure-response relationships used in the modelling (Trasande et al. 2016). The same limitation runs through the wider literature: undercounting of the disease burden attributable to chemical pollution is probably substantial, because only a small fraction of the many thousands of manufactured chemicals in commerce have been adequately tested (Fuller et al. 2022).

So the figures above — hundreds of billions annually per economic bloc — are not estimates of the total. They are estimates of the small, well-studied corner of the problem where the evidence was strong enough to permit a calculation at all.

Different toxins, different compartments

Where a compound ends up depends on its chemistry. Distribution is governed by solubility in lipids or water (Yanev, Chaldakov, 2012).

Heavy metals are generally hydro- and proteophilic — methylmercury being the clear exception — accumulating in non-fatty tissue such as organs and bone. Lead's maximal half-life is in bone, 20–30 years, against 35 days in blood and 40 in soft tissue (Aaseth et al. 2018). Cadmium accumulates mostly in liver and kidney, with a half-life up to 23.5 years (Milancović et al. 2024). Inorganic mercury salts settle in the kidney, while methylmercury crosses the blood–brain and blood–placenta barriers, accumulating in fetal brains more than maternal ones, with a brain half-life measured in years and decades (Gorini, Tonacci, 2024; Aaseth et al. 2018).

Persistent organic pollutants distribute between fatty and non-fatty tissue with a strong tendency toward the former: most lipophilic toxins are stored in adipose tissue (Jandacek, Tso, 2001), as are many endocrine disruptors, whose presence there further increases retention of other lipophilic chemicals (Dalamaga et al. 2024; Minderoo-Monaco Commission, 2023).

This is not academic. Only by knowing where a compound sits, and whether it is water- or fat-soluble, can the right approach to mobilisation and elimination be chosen — and it explains why blood testing misses what is stored elsewhere, as with lead in bone (Thomson, Darwish, 2019).

Some toxins stay in the body for decades

The word "exposure" suggests an event. For many compounds it is closer to an acquisition.

Cadmium is the clearest case: the body has no endogenous clearance mechanism for it, so it accumulates across a lifetime, with a half-life estimated at 15–30 years (Liu et al. 2023; Peraza et al., 1998; Milancović et al. 2024). That varies by compartment — 3–4 months in blood, up to 60 years in the kidneys (Echeverría et al. 2019), which hold over half the total body burden (Gil et al. 2011).

Lead behaves similarly: its half-life in blood is 28–35 days and in soft tissue about 40, but up to 95% resides in the skeleton with a half-life of 20–30 years, from which it is slowly released back into circulation (Gorini, Tonacci, 2024; Aaseth et al. 2018). Mercury persists up to 20 years in some forms, and its half-life in the human brain runs to years and decades, so neurotoxic effects outlast the exposure by a long margin (Gorini, Tonacci, 2024; Aaseth et al. 2018).

Organic pollutants are comparable: human elimination half-lives of PCBs and organochlorine pesticides run 6.4 to 30 years (Bu et al., 2015), dioxins 3 to 25 years (Sakurai et al. 2004), PFAS around 5 years in plasma (Genuis, Curtis, Birkholz, 2013). DDT and DDE remain detectable decades after application ceased (Han, Jin, 2025).

Which is why toxic effects may appear years or decades after the exposure causing them (Rozman et al. 1982).

Blood and urine poorly reflect the true body burden

Nearly all biomonitoring rests on blood and urine. Yet these are poor surrogates for toxins accrued over a lifetime, indicating exposures over recent days or months and, to a lesser extent, kidney burden (Sears, 2013). Blood concentrations are heavily affected by point exposures such as a meal in the preceding hours, and may not represent accumulated body burden (Mustieles, Arrebola, 2020).

The gap is compartment-specific. Blood lead reflects recent exposure while bone — holding up to 95% of the total — is the gold standard (Gorini, Tonacci, 2024); one tissue at one moment gives an incomplete picture of current or cumulative exposure (NTP, 2012). Blood, hair and urine mercury do not correlate with total body burden, nor hair mercury with brain content of metallic mercury (Bernhoft, 2012). For persistent organic pollutants, adipose tissue is the acknowledged gold standard, blood merely a surrogate assuming equilibrium (Moriceau et al. 2023), and remains most suitable for accumulated concentrations given its lower intra-individual variability (Mustieles, Arrebola, 2020). In 100 French women, some compounds present in adipose tissue were detected less often in blood, some not at all — most clearly the brominated flame retardants (Ploteau et al. 2016).

Sweat exposes the gap directly: BPA appeared in the sweat of 16 of 20 participants but the serum of only 2, leading the authors to conclude that blood and urine biomonitoring may underestimate total body burden (Genuis et al. 2012b).

The implication is uncomfortable: a "normal" result may mean low exposure — or a large stored burden that routine sampling cannot see.

What the evidence cannot settle

Everything above should be read with several constraints in view — most of them stated by the investigators themselves.

Detection is not the same as harm. Biomonitoring establishes that a compound is present, not that it is doing damage at the concentration measured. The distinction matters most where the finding is striking: micro- and nanoplastic concentrations in decedent brains were 7–30 times those in liver or kidney and higher still in dementia cases, but the authors are explicit that these data are associative and do not establish a causal role for such particles in health outcomes (Nihart et al. 2024).

Attribution is uneven. Some effects — kidney disease after lead exposure, loss of specific neurons after methylmercury exposure — are clearly attributable to a particular exposure. Others, including cancer, birth defects and many endocrine and nervous system actions, are difficult to ascribe with certainty to any single exposure (Carpenter et al. 1998). Alterations arising during development are rarely visible at birth, which makes the causal link harder to detect (Purdel et al. 2014).

Findings conflict. Some studies report no association between mercury concentrations and reduced fecundability, infertility or longer time to pregnancy (Henriques et al. 2019). A Spanish study of 215 young men found no significant negative impact of parabens on reproductive parameters (Adoamnei et al. 2018). Evidence on phthalates and endometriosis is mixed across three studies (Kahn et al. 2020), and data on the cognitive and neurological effects of some phthalates remain inconclusive (Umweltbundesamt, 2023). Reviews of endocrine disruptors and obesity acknowledge inconsistencies across the summarised literature (Amon, Kek, Klun, 2024).

Measurement is imprecise. Reported no-observed-adverse-effect levels vary by two to three orders of magnitude across studies (Han, Jin, 2025). Blood concentrations are affected by point exposures and may not represent accumulated burden (Mustieles, Arrebola, 2020), and biomarker-based studies suffer exposure imprecision that understates true exposure-response relationships (Trasande et al. 2016). Recent weight loss or disease-related mobilisation of stored pollutants can produce reverse-causality problems or spurious associations (Mustieles, Arrebola, 2020). Chemical analysis alone may be an unrealistic basis for assessing mixed exposures, given the number of residues and the structural diversity of the compounds involved (Arrebola et al. 2012).

Much of the mechanistic evidence is animal work. Human epidemiology and animal experiments are frequently cited together; the latter provide stronger causal evidence but require extrapolation across species and, often, across dose ranges.