Lead Exposure: No Safe Level, and It's Still Everywhere

Lead exposure is one of the oldest identified environmental hazards — and one of the least resolved. The Romans knew it was toxic. The CDCCenters for Disease Control and Prevention eliminated its 'level of concern' for blood lead in 2012 because the evidence showed there is no level at which lead exposure is safe. Not low risk. Not acceptable. No safe level. And yet over 170 million Americans alive today were exposed to high lead levels as children Americans alive today were exposed to harmful lead levels in early childhood — more than ninety percent of those born between 1950 and 1980 had blood lead above 5 µg/dL McFarland et al. 2022. Lead is part of the broader pattern of environmental chemical exposure this library documents, but it's not emerging science. It's established. The question isn't whether lead is harmful. It's why so much of it is still in the infrastructure — and, in some places, still being added on purpose.

Lead (Pb, atomic number 82, CASChemical Abstracts Service 7439-92-1) is a naturally occurring heavy metal — soft, dense, corrosion-resistant, and cheap to extract. Those properties made it one of the most useful industrial materials in history: pipes, paint, petrol additives, solder, ammunition, batteries, and radiation shielding. The same properties that make it useful — particularly its ability to mimic calcium in biological systems — make it toxic at exposure levels far below anything visible in a routine medical examination.

Lead damages the developing brain by impersonating calcium. Calcium is the molecular currency of neuronal communication — it triggers neurotransmitter release, gates ion channels, and switches enzymes on and off. Lead binds to the same sites with greater affinity, then either blocks the channel entirely or jams the signal on when nothing is happening. The brain fires at the wrong moments and stays quiet at the right ones.

Lead crosses the blood-brain barrierthe semipermeable membrane that controls what enters the brain from the bloodstream — lead crosses it by mimicking calcium ions by exploiting the same calcium transport channels the brain uses to maintain its own ion gradients. Once inside the developing nervous system, it does two contradictory things at the same synapse. At the presynaptic terminal it activates protein kinase Can enzyme switched on by calcium that drives neurotransmitter release and calmodulin in the absence of an actual signal, triggering spontaneous neurotransmitter release. At the same terminal, it blocks voltage-gated calcium channels, so when a signal does arrive, evoked release is dampened Bressler & Goldstein 1991. The presynapse becomes noisy and weak at the same time — a combination the developing cortex has no good way to compensate for.

The other receptor lead targets is the NMDAN-methyl-D-aspartate receptor — the channel that gates the long-term synaptic strengthening underlying learning and memory. Lead is a more potent inhibitor of NMDA function than zinc or magnesium, and the immature cortex is markedly more sensitive to this antagonism than the adult brain Guilarte & Miceli 1992. This receptor-level pharmacology is the proposed mechanistic reason the dose-response curve for childhood IQ gets steeper at lower exposures rather than flatter. At very low blood-lead concentrations every additional lead atom finds an unoccupied receptor to bind; at higher concentrations the receptors are already saturated and the marginal damage drops Bellinger 2008. The poison loses potency the more of it you have.

Whether the supralinear shape of the lead-IQ curve is purely biological or partly a statistical artifact of overlapping lognormal exposure distributions is still being argued — by Lanphear among others Hornung & Lanphear 2014. The honest answer is that mechanism research and population epidemiology arrive at the same place from opposite directions: the slope is steepest at the bottom. Whichever framing you accept, the implication is the same. There is no exposure floor below which the developing brain is unaffected.

The dominant sources have shifted over the decades. Leaded petrol — phased out globally by 2021, with Algeria the last country to stop selling it — was once the largest contributor to population-level exposure. Today, for children with elevated blood lead, deteriorating lead paint in pre-1978 (US) or pre-1992 (UK) housing remains the primary source. For the general population not living in old housing, dietary contributors lead the residual exposure: grains and baked goods, dairy, and fruit are the three largest categories in the FDAFood and Drug Administration's Total Diet Study, with grain-based foods contributing roughly a quarter of dietary lead in young children FDA Total Diet Study 2018-2020.

Then there are the sources people don't expect. Turner tested 72 decorated drinking glasses — the painted tumblers you'd pick up at a charity shop or a Christmas market — and found lead in 139 of 197 analyses, with migration tests exceeding regulatory limits in nearly every lip-rim painted product Turner 2017. Shearston and colleagues at UC Berkeley detected lead across all 30 tampon samples they tested, drawn from 14 brands and 18 product lines, though whether metals leach from tampons and absorb through vaginal tissue has not yet been studied Shearston et al. 2024. And in late 2023, a single cinnamon-flavoured applesauce product adulterated with lead chromate poisoned hundreds of US children before the FDAFood and Drug Administration identified the source — we'll come to that pattern below.

Water remains a critical route in specific situations. The most-cited figure for UK homes still served by lead pipes — eight to nine million — comes from a 1997 House of Commons research paper, and there is no current consolidated replacement of that estimate. The water industry has a collective ambition of being lead-pipe-free by 2050, but it is an industry target rather than a regulatory deadline. Fifty years. For a pipe. Our tap water guide covers the Flint crisis, pipe identification, and which filters actually remove lead — the short version is that running the tap for 30 seconds before first-draw morning water flushes overnight stagnation, and an NSF 53 certified carbon filter removes lead reliably.

The landmark study came in 1979. Herbert Needleman, a pediatrician at Harvard, had a problem that no existing study could solve: you can't deliberately expose children to lead. So he found a natural record. He collected baby teeth from 2,146 children in Chelsea and Somerville, Massachusetts, measured the lead deposited in the dentineThe hard tissue beneath tooth enamel — it accumulates lead over years, making shed baby teeth a biological archive of childhood exposure, and compared the cognitive performance of children at the extremes: 58 with the highest dentine lead against 100 with the lowest.

The high-lead children scored significantly lower on IQ tests and three measures of auditory processing, speech, and attention. Teachers, who didn't know which children were in which group, independently rated the high-lead children as more distractible, less organized, and less able to follow sequences Needleman et al. 1979. The study was attacked by the lead industry for years — Needleman was investigated, cleared, investigated again, and cleared again. The finding survived everything thrown at it. It took another 33 years for the CDC to formally eliminate the 'safe' level.

The quantification followed. Lanphear and colleagues pooled data from 1,333 children across seven international cohorts and reported a 3.9 IQ point decline (95% CI 2.4–5.3) associated with blood lead rising from 2.4 to 10 µg/dL — with the steepest losses occurring at the lowest concentrations IQ point decline (95% CI 2.4–5.3) associated with blood lead rising from 2.4 to 10 µg/dL — with the steepest losses at the lowest concentrations, not the highest Lanphear et al. 2005. Canfield's Rochester cohort confirmed the pattern: roughly half an IQ point lost for every 1 µg/dL increase in blood lead in linear modelling, with the nonlinear model showing IQ falling 7.4 points across the 1–10 µg/dL range — most of it at the bottom of the range Canfield et al. 2003. The dose-response curve for lead doesn't flatten out at low doses. It gets worse.

Here's the part most people don't know: lead isn't only a childhood issue. Between 1988 and 1994, the US National Health and Nutrition Examination Survey enrolled 14,289 adults — a nationally representative sample, chosen to reflect the country — and measured their blood lead. The levels were unremarkable. The 10th-to-90th percentile range ran from 1.0 to 6.7 µg/dL, well within what most American adults carried at the time. Nobody flagged them. Nobody intervened. Then Lanphear and colleagues tracked those 14,289 people for up to 20 years and counted who died.

The adults in the highest lead quintile had a 37% increase in all-cause mortality and a 70% increase in cardiovascular mortality compared to those in the lowest Lanphear et al. 2018. The team attributed an estimated 256,000 cardiovascular deaths per year in the United States alone are attributable to lead exposure — Lanphear et al. 2018 cardiovascular deaths per year in the United States to lead exposure. Not factory workers. Not people living next to a smelter. Adults with blood lead levels their doctors would have called normal. Globally, a 2023 analysis estimated that lead contributes to 5.5 million cardiovascular deaths and the loss of 765 million IQ points in children annually — an economic burden of roughly $6 trillion, or about seven percent of global GDP Larsen and Sanchez-Triana 2023.

The biology behind the adult mortality finding is well-mapped. Navas-Acien and colleagues' 2007 systematic review concluded that the evidence is sufficient to infer a causal relationship between lead exposure and hypertensionchronically elevated blood pressure, with cardiovascular outcomes 'suggestive but not yet sufficient' as of that year — a gap Lanphear's 2018 NHANES analysis has since narrowed considerably Navas-Acien et al. 2007. The mechanism is multi-step: lead promotes vascular oxidative stress, reduces nitric oxide bioavailability, impairs endothelial signalling, and amplifies adrenergic activity, producing the blood-pressure rise that drives the long-term cardiovascular risk Vaziri 2008.

There is also a population-level finding that has stayed contested for two decades. Reyes 2007 attributed roughly half of the US violent-crime decline between 1992 and 2002 to the cohort effect of leaded-petrol phase-out two decades earlier — children with less prenatal and early-childhood lead exposure becoming young adults with measurably better impulse control Reyes 2007. Mielke and Zahran's six-cities study reported a 22-year lag between historical air-lead emissions and aggravated-assault rates that explained up to ninety percent of cross-city variance Mielke & Zahran 2012. A 2022 publication-bias-adjusted meta-analysis of 542 estimates from twenty-four studies puts the contribution at the lower end — perhaps 7 to 28 percent of the homicide decline — but does not eliminate it Higney, Hanley & Moro 2022. The individual-level mechanism is plausible. The population-level signal is robust. The size of the effect is what people argue about.

Lead is stored in bone and teeth for decades, then re-released into the bloodstream during any condition that mobilises calcium: pregnancy, lactation, menopause, osteoporosis, fracture. Over ninety-five percent of the body's lead burden sits in the skeleton, with a half-life of ten to thirty years. The exposure you had as a child can reach your own child decades later — through your bones.

The clearest evidence comes from a series of stable-isotope studies in Sydney in the 1990s. Brian Gulson and colleagues at Macquarie University recruited recent immigrants to Australia whose bones still carried the isotopic signature of European or Asian lead — distinct from the local Australian food and air the women were now exposed to. They tracked the isotope ratio in maternal blood across pregnancy and lactation. The bone-derived fraction was unmistakable: on average, 31% of maternal blood lead during pregnancy comes from bone laid down decades earlier — Gulson et al. 1997 of the lead in maternal blood during pregnancy came not from current diet but from skeleton built decades before Gulson et al. 1997. In a follow-on study tracking the same women postpartum, individual contributions of bone-stored lead to maternal blood ran from 26% to 99% of the postpartum bloodstream rise came from bone rather than diet Gulson et al. 1998. The skeleton is not an inert storage site. It is a slow-release reservoir.

What reaches maternal blood reaches the fetus almost immediately. There is no functional placental barrier to lead — maternal and umbilical cord concentrations are nearly identical, and lead is detectable in fetal brain tissue by the end of the first trimester Goyer 1990. The exposure continues after birth. Ettinger and colleagues measured lead in maternal blood, plasma, and breast milk one month postpartum, and in the infants three months later. A 1 µg/L rise in breast milk lead corresponded to a 1.8 µg/dL rise in the infant's blood lead, with the milk-to-plasma ratio averaging 7.7 times higher in breast milk than in maternal plasma — Ettinger et al. 2014 across 81 mother-infant pairs and reaching nearly 40 in one woman Ettinger et al. 2014. The biology that delivers calcium to the infant skeleton delivers lead alongside it.

The same mobilisation happens decades later. Postmenopausal women show significantly elevated blood and plasma lead compared with premenopausal women in the same cohort, as oestrogen withdrawal accelerates the bone demineralisation that releases stored lead Silbergeld et al. 1988. A grandmother's exposure to leaded petrol in 1965 and to flaking paint in a 1970s rental can show up in her own bloodstream four decades later, and in her granddaughter's umbilical cord nine months after that. Lead does not end with you.

Two intentional uses of lead are still reaching American and global consumers in 2026: leaded aviation gasoline burned by piston-engine aircraft, and lead chromate adulteration of imported turmeric, paprika, and cinnamon. Both are residues of the wider phase-out — car petrol stopped containing lead in 1996, but small aircraft kept their certification, and the same brilliant yellow-orange pigment that made paint dangerous still makes ground spices look more vivid in the bag.

Aviation gasoline sold as 100LL is the last transport fuel in the United States that still contains tetraethyllead. Roughly 170,000 piston-engine aircraft operate from approximately 20,000 American airports, burning a fuel specified at up to 2.12 grams of lead per US gallon under ASTMAmerican Society for Testing and Materials D910. The result, per the EPAEnvironmental Protection Agency, is ~470 tons of lead per year still emitted into US air by piston-engine aircraft — about 70% of all US air lead emissions of lead per year, around seventy percent of all remaining US air lead emissions EPA 2023 endangerment finding. In October 2023 the EPA finally issued a formal CAAClean Air Act §231 endangerment finding that lead emissions from piston aircraft contribute to air pollution dangerous to public health, obligating the agency to develop emission standards. Children living within 500 metres of a general aviation airport have measurably higher blood lead levels than other children of the same age, with the effect monotonic with distance — though the authors of the most-cited study note the magnitude is modest Miranda et al. 2011. Three decades after the science was settled on car gasoline, the agency that wrote the standard is still working on the version for Cessnas.

The other ongoing intentional source is harder to see because it is not legal — it is economically motivated adulteration of imported food. The brilliant yellow-orange of lead chromate costs less than turmeric does, and adding a small fraction to ground spice brightens the colour and inflates its market value. Forsyth and colleagues sampled 524 turmeric, pigment, dust, and soil samples across nine major Bangladeshi turmeric-producing districts and confirmed lead chromate adulteration in seven of them, with turmeric powder reaching 1,152 μg/g of lead in adulterated turmeric powder from Dhaka markets — Forsyth et al. 2019 of lead in market samples and the polishing-mill yellow pigments themselves running 2 to 10% lead by weight Forsyth et al. 2019. The practice began roughly thirty years ago, driven by consumer preference for brighter yellow.

The same pattern reached American children in late 2023 through a different spice. Cinnamon imported from Sri Lanka, processed in Ecuador, was contaminated with lead chromate at 2,270 to 5,110 parts per million — hundreds of times the FDA's action level — and ended up in pouches of cinnamon-flavoured applesauce on supermarket shelves across 44 US states, the District of Columbia, and Puerto Rico. By the time the CDCCenters for Disease Control and Prevention closed the outbreak investigation in April 2024, 566 cases of lead poisoning across 44 US states + DC + Puerto Rico from cinnamon-applesauce contaminated with lead chromate, 2023–2024 — Troeschel et al. 2025 cases had been reported, 542 of them children under six, with a median peak blood lead of 7.2 µg/dL and one child reaching 39.3 Troeschel et al. 2025. The FDAFood and Drug Administration's 2025 'Closer to Zero' action levels — 10 ppb for fruit purées, 20 ppb for root-vegetable baby foods — do not cover spices.

Lead is one of the most heavily regulated toxic substances in the world — and the regulation still hasn't caught up with the infrastructure. The US banned lead paint in 1978 via CPSCConsumer Product Safety Commission regulation 16 CFR 1303. The UK followed with SI 1992/31, which specifically banned the use of lead carbonate and lead sulphate as paint pigments — a narrower scope than most consumers realise, leaving certain industrial and restoration uses outside the prohibition. Leaded petrol was phased out globally by 2021, with Algeria the last country to stop selling it. The EPAEnvironmental Protection Agency finalised the Lead and Copper Rule Improvements in October 2024, requiring all lead service lines in the US to be replaced by November 2037 — the most significant federal action on lead pipes in decades.

The gap between what we know and what we've done is the story. The CDC says no level is safe. The US action level remains 15 µg/L. The EU is moving to 5 µg/L — but not until 2036. The UK's Drinking Water Inspectorate advisory group recommended the same reduction in December 2024; as of 2026 the recommendation has not been turned into law. Eight to nine million UK homes still drink water through pipes laid down before the bans, with the water industry's collective ambition to be lead-pipe-free by 2050 — three decades away, and not a regulatory deadline. No safe level — regulation decades behind

Lead exposure is largely a function of where you live and what your home is made of. The biggest controllable sources are drinking water (if you have lead pipes or solder), paint dust (if your home predates the ban), and soil (if you garden in urban areas near old buildings or roads). Unlike some of the chemicals in this library, lead has well-established, practical reduction steps — most of them free.

For water filtration details — which filters remove lead, which don't, and what to look for — see the full tap water guide. The eso-friendly approach to lead is straightforward: test if uncertain, filter if exposed, and prioritize children and pregnancy.

Lead is the chemical that proves the system can work — and how long it takes. Needleman published in 1979. The CDC eliminated the safe level in 2012. The EPA mandated pipe replacement in 2024. The UK's target for finishing the job is 2050. That's 71 years from the study that proved the harm to the deadline for removing the source. And the source is a pipe. A physical object. Something you could, in theory, just dig up.

The other chemicals in this library — the ones dissolved in products, embedded in coatings, hiding behind the word 'fragrance' — don't have pipes you can replace. Lead is the best-case scenario for how fast the system moves. Keep that in mind when you read about the ones that were discovered last decade.