Arsenic in Food: What's in Your Rice and What You Can Do About It

Arsenic in food is not contamination in the usual sense. Nobody adds it. It's a naturally occurring element in soil and groundwater that rice — uniquely among grains — concentrates through the way it's grown. Flooded paddy fields create the precise chemical conditions that make arsenic available to roots, and rice plants absorb it through the same transport pathway they use for silicon. The result is a staple food for half the world that carries a confirmed carcinogen at levels high enough that the EUEuropean Union set specific legal limits in 2016 — and tightened them in 2023. This is part of the broader picture of environmental chemical exposure — not a single product to avoid, but a dietary pattern to manage.

Arsenic (As, atomic number 33, CASChemical Abstracts Service 7440-38-2) is a metalloid — a naturally occurring element found in rock, soil, and water worldwide. It exists in two forms that matter for health. Inorganic arsenicThe toxic form — arsenic bonded to elements like oxygen, chlorine, or sulfur. Found in groundwater and concentrated in rice. IARC Group 1 carcinogen. is the toxic form: arsenic bonded to oxygen, chlorine, or sulfur, found in groundwater and absorbed by rice. Organic arsenicArsenic bonded to carbon — found mainly in seafood as arsenobetaine, which passes through the body harmlessly and is not considered a health concern is arsenic bonded to carbon, found mainly in seafood as arsenobetaine, and generally considered non-toxic. When this article says 'arsenic,' it means inorganic arsenic — the kind that matters.

IARCInternational Agency for Research on Cancer classifies inorganic arsenic as Group 1 — carcinogenic to humans — with sufficient evidence for lung, bladder, and skin cancers (Monograph Volume 100C, 2012). EFSAEuropean Food Safety Authority updated its risk assessment in January 2024 and set a BMDL05Benchmark Dose Lower Confidence Limit at 5% extra risk — the lowest exposure level associated with a measurable increase in cancer risk above background. of 0.06 µg/kg bw/dayinorganic arsenic — the 5%-extra-risk threshold based on US case-control data on skin cancer (squamous cell carcinoma) based on skin cancer, concluding that dietary exposure 'raises a health concern' and that children are covered by the same risk characterisation EFSA CONTAM Panel 2024. The same reference point is held to cover lung and bladder cancers, ischemic heart disease, chronic kidney disease, spontaneous abortion, stillbirth, infant mortality and neurodevelopmental effects. A confirmed carcinogen in a food that half the planet eats daily. That's the scope of this problem.

Rice is the dominant dietary source of inorganic arsenic for most consumers who don't live in areas with contaminated groundwater. The mechanism is agricultural, not industrial: rice paddies are flooded, and waterlogged soil shifts to reducing conditions where arseniteThe reduced form of inorganic arsenic (As(III)) — more mobile in flooded soils and more readily absorbed by rice roots than arsenate (As(V)) dominates. Rice roots take up arsenite through their silicon transport channels — the same pathway the plant uses for structural strength. No other major cereal crop is grown this way, which is why rice absorbs an order of magnitude more arsenic than wheat or barley from the same soil.

Menon and colleagues at the University of Sheffield tested 55 rice varieties sold in UK supermarkets and found that 28 out of 55 rice samples in UK shops exceeded the EU's 0.10 mg/kg infant-food limit — over half the products tested, with average iAs of 0.13 mg/kg in those 28 exceeded the EU's 0.10 mg/kg limit for rice destined for infant food production — more than half the products on the shelf, with an average inorganic-arsenic concentration of 0.13 mg/kg in those 28 Menon et al. 2020. Brown rice was significantly higher than white or wild; organically grown varieties skewed higher than conventional. The paper's authors recommended that infants under one limit consumption of those 28 high-arsenic rice types to no more than 20 grams per day. Organic doesn't mean lower arsenic — it often means higher, because organic paddy management can increase arsenite availability in soil.

Beyond rice, inorganic arsenic appears in apple juice (the FDAFood and Drug Administration set a 10 ppb action level in 2023), hijiki seaweed (consistently high — the FSAFood Standards Agency advised against eating it on 5 August 2010), and drinking water in regions with naturally arsenic-rich geology. The drinking-water route reaches the US, too — just less visibly. USGSUnited States Geological Survey modelling built from 20,450 well measurements estimated that 2.1 million 95% CI 1.5-2.9 million Americans on conterminous-US domestic wells where modelled arsenic exceeds the WHO/EPA 10 µg/L threshold — Ayotte et al. 2017 Americans drinking from private wells in the conterminous US sit on aquifers where modelled arsenic exceeds the WHO/EPA 10 µg/L threshold (95% CI 1.5–2.9 million) Ayotte et al. 2017. Public water systems test for arsenic and meet the federal MCLMaximum Contaminant Level — the legally enforceable EPA standard for drinking water. The arsenic MCL is 10 µg/L, finalised in 2001 with a compliance deadline of 2006. either by treatment or blending; private wells aren't tested unless the homeowner pays for it, and roughly 44 million Americans drink from one.

The largest single example of the contaminated-aquifer route is Bangladesh, where tubewells drilled in the 1970s to provide clean water instead tapped into arsenic-laden aquifers. WHOWorld Health Organization called it 'the largest poisoning of a population in history' Smith Lingas Rahman 2000 — and Argos and colleagues subsequently put a number on the population at risk: between 35 and 77 million Bangladeshis exposed above the WHO guideline of 10 µg/L for decades before the contamination was identified in 1993. The wells drilled to save people from cholera gave them cancer instead.

What happened to those people got measured. The HEALS team — HEALSHealth Effects of Arsenic Longitudinal Study — a US/Bangladesh collaborative cohort recruited 2000-2002 in the Araihazar district, designed to track health outcomes from chronic arsenic-contaminated drinking water in adults exposed at low-to-moderate concentrations. recruited 11,746 adults aged 18-75 in the Araihazar district of Bangladesh, recruited 2000-2002 and followed biennially through February 2009 — Argos et al. 2010, Lancet adults aged 18 to 75 across the Araihazar district between 2000 and 2002, measured well-water arsenic at every household, validated exposure with urinary biomarkers, and followed each participant biennially. By February 2009, 407 of them had died. Adults whose well water carried 150 to 864 µg/L of arsenic — well above the WHO guideline but below the worst Bangladeshi exposures — had a 68% higher all-cause mortality rate than adults whose wells held 10 µg/L or less (HR 1.68 95% CI 1.26-2.23, all-cause mortality at 150.1-864 µg/L vs ≤10 µg/L well-water arsenic, after adjustment for age, sex, smoking, BMI, education, land ownership, 95% CI 1.26-2.23) Argos et al. 2010. Eleven thousand wells. One decade. A cohort large enough to see what arsenic does at the population scale, in a population that drank it.

The cancer evidence comes from high-exposure populations — communities drinking arsenic-contaminated water in Bangladesh, Taiwan, Chile, and Argentina — where decades of exposure at concentrations far above the WHO guideline produced measurable increases in lung, bladder, and skin cancer. At typical dietary exposure levels from rice, the individual risk is small. But because the entire population is exposed, the population-level impact is not.

The most striking demonstration of how long the damage lingers came out of northern Chile. The city of Antofagasta drew its drinking water from a single river system loaded with naturally high arsenic from 1958 until a treatment plant was built in 1970. Twelve years of exposure, then it stopped. Steinmaus and colleagues enrolled 232 bladder-cancer cases, 306 lung-cancer cases, and 640 controls between 2007 and 2010 — roughly 40 years after the water had been cleaned — and asked: did the cancer risk go away when the exposure stopped? It didn't. Bladder-cancer odds were 6.5× 95% CI 3.69-11.43, top exposure quartile vs lowest, in Chilean residents 40 years after their drinking water was cleaned — Steinmaus et al. 2013 higher (OR 6.50, 95% CI 3.69-11.43) and lung-cancer odds 4.3× higher in residents who'd been exposed during 1958-1970 — and the elevations persisted whether or not exposure continued after 1970 Steinmaus et al. 2013. The paper's title says it directly: 'high cancer risks 40 years after exposure cessation.'

Cancer is not the only signal. Moon and colleagues followed 3,575 American Indian adults aged 45-74 in Arizona, Oklahoma, North Dakota, and South Dakota — Strong Heart Study cohort, baseline 1989-91, follow-up through 2008 American Indian adults across Arizona, Oklahoma, and the Dakotas for nearly two decades, with urinary-arsenic biomarkers and continuous cardiovascular outcome ascertainment. Cohort exposures were low to moderate — almost all below 100 µg/L in drinking water — but the gradient still mattered: cardiovascular mortality was 65% higher in the top urinary-arsenic quartile (above 15.7 µg/g creatinine) compared with the bottom (below 5.8 µg/g) (HR 1.65, 95% CI 1.20-2.27), with stroke mortality tripled and incident CVD up 32% Moon et al. 2013. Wasserman's group, running tests on 201 ten-year-olds in the same Araihazar district as the HEALS adults, found that children whose home wells carried more than 50 µg/L scored roughly 10 points lower on the WISC-III Performance subtest and 11 points lower on the Full-Scale IQ raw score than children below 5.5 µg/L Wasserman et al. 2004.

EFSA's 2024 margin-of-exposure analysis put numbers on what 'narrow' means at population scale. For average European consumers the margin between estimated dietary intake and the BMDL05 was narrow; for high consumers of rice and rice-based products — and for infants fed rice-based cereals — the margins were tighter still.

Children are disproportionately exposed because of body weight. A 10-kilogram infant eating a standard portion of rice cereal receives roughly three times more arsenic per kilogram of body weight than an adult eating the same food. Same cereal. Same serving. Three times the dose per kilo. This is why the EU's strictest limits apply to baby food: 0.020 mg/kg for ready-to-eat infant food and 0.10 mg/kg for rice destined for infant food production — compared to 0.15 mg/kg for general white rice.

The EU leads. Commission Regulation 2023/465, which updated and tightened the original 2016 limits, sets the most detailed framework: different limits for white rice, brown rice, rice cakes, rice milk, and baby food. The US has no binding limit on arsenic in general rice — only a guidance-level action level of 100 ppb for infant rice cereal (finalized 2020) and 10 ppb for apple juice (2023). The UK retained EU limits after Brexit. Codex Alimentarius — the international standard — is less strict than the EU at 0.2 mg/kg for polished rice.

The absence of a US general rice limit is notable. The FDAFood and Drug Administration conducted a risk assessment in 2016, found that dietary arsenic from rice was a concern, and chose to act only on infant cereal. The gap between what EFSA concluded in 2024 — that arsenic raises a health concern for all age groups — and the FDA's regulatory scope is significant. The same grain, the same arsenic. Different continents, different limits — or no limits at all.

The single most effective step is how you cook it. The standard absorption method — two cups of water to one cup of rice, all absorbed — retains almost all the arsenic. Rinsing alone has minimal effect Carey et al. 2015. But cooking rice like pasta — in a large pot of water and draining the excess — removes 40–60% of inorganic arsenic, depending on rice type and water ratio — Gray et al. 2016, Menon et al. 2021 of the inorganic arsenic Gray et al. 2016. The tradeoff is real: the same water that carries arsenic away also carries enriched B vitamins and iron. But if you're eating rice several times a week, the arsenic reduction outweighs the nutrient loss.

The eso-friendly approach to arsenic isn't avoidance — rice is nutritious, affordable, and central to most of the world's cuisines. It's preparation and variety. Cook it in excess water. Choose lower-arsenic varieties. Don't give babies rice cereal as the only first food. The arsenic is in the grain. How much stays in the grain is partly up to you.

Arsenic in rice is the quietest entry in this library — no concealed memo, no regulatory scandal, no banned chemical hiding behind a different name. It's geology meeting agriculture meeting diet, and the result is a confirmed carcinogen in a food that billions of people eat every day. The fix isn't dramatic. Cook the rice differently. Choose the right variety. Give infants something other than rice cereal as their only grain.

The deeper lesson is the one this library keeps finding: the testing model evaluates chemicals one at a time, in isolation. Your rice contains arsenic. Your tap water contains disinfection byproducts. Your plastic container sheds particles. The combined exposure is the one nobody measures.