The Wooden Spoon

Every few weeks, the careful kitchen keeper takes a bottle of mineral oil from under the sink, pours a measure onto a wooden spoon, and rubs it in with bare hands. The wood darkens. The grain drinks it in. The gesture feels ancient — tending a tool, preserving a surface, caring for something natural. It is one of the small rituals of a kitchen that has chosen well: wood instead of plastic, natural instead of synthetic, the conscious alternative.

The mineral oil is petroleum. The chemistry says it comes off in hot food. Nobody has tested whether it does.

That is not an accusation. It is a gap — between what we assume has been checked and what anyone has actually measured. And the chemistry behind that gap begins with a distinction no product label, care guide, or kitchen shop will ever explain.

The Finish You Never Asked About

Mineral oil saturated hydrocarbons have been found in human lymph nodes, liver, and fat tissue at concentrations that increase with every decade of life.¹⁵¹⁶ One of the daily pathways those hydrocarbons travel is through food. One of the surfaces food contacts is the finish on a wooden spoon. And the finish on most wooden spoons is mineral oil — a petroleum derivative that, unlike the plant-based oils it replaced, never solidifies.

That last distinction is the one that matters.

When you apply linseed oil to wood, a chemical transformation begins. The oil's unsaturated fatty acids react with oxygen, crosslinking into a three-dimensional polymer network — a process called autoxidative polymerisation.¹² Over twenty-four to seventy-two hours, the liquid becomes a solid film bonded to the wood fibre at the molecular level. It is no longer oil. It is a polymer. It does not migrate because it has become part of the wood.

Mineral oil — petroleum distillation products, technically saturated aliphatic and alicyclic hydrocarbons — has no unsaturated bonds available for this reaction.³ It cannot polymerise. It cannot form a film. It cannot bond to wood. It remains liquid within the wood's pore structure indefinitely — mobile, viscous, and available for transfer to anything that contacts it.

The marketing language collapses this distinction into a single verb. "Oil your spoon" describes two fundamentally different chemical processes. One creates a permanent solid barrier. The other replenishes a mobile petroleum reservoir. The care guides do not mention the difference. The product labels certainly do not. The word "oil" does the work of making them sound the same.

They are not the same. One is a finish. The other is a coating that never finishes.

The Study That Doesn't Exist

In September 2023, the European Food Safety Authority published its most comprehensive risk assessment of mineral oil hydrocarbons in food.⁴ The EFSA CONTAM Panel evaluated two families of compounds: MOSH (mineral oil saturated hydrocarbons) and MOAH (mineral oil aromatic hydrocarbons). Their conclusion on MOAH with three or more aromatic rings was unequivocal:

"The genotoxicity of MOH is associated with the presence of some MOAH with three or more aromatic rings."⁴

These compounds can damage DNA. They are carcinogenic. And because of this, EFSA concluded:

"Since the possible presence of genotoxic and carcinogenic components within MOAH prevents the setting of an HBGV, an MOE approach was applied."⁴

In regulatory terms, this means no threshold of safe exposure can be established using standard methods. EFSA applied a Margin of Exposure approach — the framework used when a health-based guidance value cannot be set — and found that for current dietary intake, the margins were troublingly thin. The reference point: a benchmark dose of 0.49 mg/kg body weight per day, derived from carcinogenicity data on polycyclic aromatic hydrocarbons as a surrogate.⁴

The opinion is two hundred pages long. It documents MOAH migration from recycled paperboard, machinery lubricants, processing aids, food additives, adhesives, printing inks, and environmental contamination.⁴ It is exhaustive.

Wood utensils are not mentioned. Not once.⁴

No published peer-reviewed migration study was identified, as of April 2026, testing mineral oil transfer from wood cooking utensils into food or food simulants under any conditions.⁵ Systematic searches across PubMed, the EFSA food contact materials database, the Food Packaging Forum's FCCmigex database of over 1,500 migration studies, the Dutch RIVM's 2019 review of mineral oil occurrence in food, and the UK Committee on Toxicity's 2023 discussion paper on EFSA's MOH opinion — all returned the same result: the pathway has never been investigated.⁵

The data that does exist for other food contact materials is instructive. Pan et al. (2021) analysed 159 food contact papers and raw materials from China: MOSH was detected in 82.6% of samples; MOAH in 50.4%.⁶ Mean MOSH migration in the highest category — long-term food storage papers — reached 91.2 mg/kg food.⁶ And paper sits passively around food. It does not move. It does not stir. It is not heated.

A wooden spoon is actively stirred through hot food — friction refreshing the contact surface, temperature reducing the oil's resistance to movement, the stirring action drawing fresh material from the wood's interior to its surface. It is the highest-contact use case for any food contact material. It has zero migration data.

The absence of study is not the same as the presence of safety.

The Regulatory Phantom

The European Union's framework regulation for food contact materials — Regulation 1935/2004 — lists seventeen categories of materials in Annex I.⁷ Wood is number seventeen. The framework requires that all food contact materials must not release constituents into food at levels harmful to human health, must not change food composition, taste, or odour, and must be traceable through the supply chain.⁷

The principle is clear. The mechanism is absent.

Of those seventeen categories, only four have harmonised specific measures — the detailed regulations that translate the general safety principle into actual migration limits, testing requirements, and mandatory Declarations of Compliance:⁷

Thirteen of seventeen categories — including wood — have no harmonised specific measures.⁷ No substance-specific migration limit. No overall migration limit. No mandatory Declaration of Compliance. No testing requirement.

A plastic spatula and a wooden spoon sit in the same drawer, touch the same food, enter the same mouth. The plastic spatula has been tested against over a thousand migration limits and carries documentation certifying compliance at every stage of its supply chain.⁸ The wooden spoon carries nothing. Its finish — the surface that actually contacts food — is not required to be disclosed, tested, or documented.

Wood is not merely absent from harmonised EU measures. It is absent from Council of Europe non-binding resolutions — which cover paper, metals, rubber, silicones, cork, and glass, among others — and absent from the BfR's national recommendations for Germany.⁹¹⁰ No layer of the European food safety framework has written a specific provision for wood. The proposed revision of Regulation 1935/2004, expected in 2023, has been repeatedly delayed; the European Parliament's most recent update states "it remains uncertain when this proposal will be tabled."¹¹

This is the architecture of what might be called a Regulatory Phantom: a product category that exists inside a regulatory framework — benefiting from the consumer's assumption that "EU regulated" means tested and verified — while having no material-specific regulation, no mandated testing, no required disclosure, and no migration data. The framework provides legitimacy without discipline. The consumer sees the framework and assumes the mechanism. The mechanism was never written.

Part of the reason it was never written is that nobody asks. Wood triggers what behavioural scientists call the appeal to nature heuristic — consumers substitute the question "Is this scientifically safe?" with the easier question "Does this feel natural?"¹² The material becomes the product. You are holding wood. You believe you are touching wood. The finish is invisible. And the act of choosing wood over plastic creates what Mazar and Zhong (2010) identified as moral licensing: the purchase IS the environmental act, and further scrutiny feels unnecessary.¹³ You already did the right thing. Why would you interrogate the finish?

The Thermal Pathway

Mineral oil viscosity is highly temperature-dependent. At 40°C, a food-grade white mineral oil (ISO VG12) has a kinematic viscosity of about 21 centistokes — roughly the consistency of a light cooking oil. At 100°C, simmering temperature, it drops to about 4 centistokes — thinner, closer to water.¹⁴ An approximately fivefold reduction in resistance to flow.

This matters because of where the mineral oil sits. It is a liquid inside the wood's pore structure. Heat makes it more mobile. The hotter the spoon, the more freely the oil can move toward the surface and into whatever the surface contacts. And the spoon's use case is defined by heat: stirring soup at 95–100°C, stirring risotto, stirring porridge, stirring sauces. The very thing a wooden spoon is for — contact with hot food — is the condition that maximises mineral oil mobility.

No study has tested migration at cooking temperatures for wood. But the physics is not ambiguous: a non-polymerising oil becomes more mobile with heat, the spoon provides sustained contact with hot food, and the stirring action continuously refreshes the contact surface — bringing new oil-laden wood into contact with food. This is not the passive contact of packaging sitting around food at refrigeration temperatures. This is active, prolonged, heated, mechanical contact.

The spoon use case is the worst-case migration scenario. It has never been tested.

What Accumulates

MOSH — the saturated fraction of mineral oil — accumulates in human tissues over a lifetime. Barp et al. (2014) analysed autopsy samples from thirty-seven individuals aged twenty-five to ninety-one and measured MOSH concentrations across five tissue types [human biomonitoring]:¹⁵

TIMELINE OF PERSISTENCE

• Median MOSH in gut-lining lymph nodes (mesenteric): 166 mg/kg (range: 21–1,390 mg/kg)
• Median MOSH in subcutaneous fat: 87 mg/kg (range: 17–493 mg/kg)
• Median MOSH in liver: 71 mg/kg (range: 14–901 mg/kg)
• Median MOSH in spleen: 28 mg/kg (range: 6–1,400 mg/kg)
• Median MOSH in lung: 7 mg/kg

A re-evaluation by Pirow et al. (2024) — re-analysing the Barp dataset alongside biopsy data from 144 women — found that MOSH concentrations in gut-lining lymph nodes and fat tissue showed a 1.2–1.4-fold increase per decade of life in cross-sectional analysis.¹⁶ A quarter of subjects probably had more than 5 grams of MOSH in their body.¹⁶

EFSA's 2023 opinion acknowledged this tissue accumulation but concluded that current estimated dietary MOSH exposure "does not raise concern for human health."⁴ Two things about this conclusion must be noted. First: the exposure model on which it rests does not include migration from wood utensils — because no one has measured it. Every unmeasured source that contributes to actual MOSH intake means the real exposure is higher than the modelled exposure. Second: the "does not raise concern" conclusion applies to MOSH. It does not apply to MOAH.

MOAH — the aromatic fraction — is a different compound with a different risk profile. MOAH with three or more aromatic rings are genotoxic: they can damage DNA [in vitro, animal].⁴ Notably, Barp et al. detected no MOAH in any tissue sample.¹⁵ MOAH apparently does not accumulate in tissue. But this does not mean safe passage. Genotoxic carcinogens — as a class, not only MOAH — are understood to cause DNA damage during metabolic activation: the body's enzymes convert the parent compound into reactive intermediates that bind to DNA, a well-established mechanism in toxicology called bioactivation.⁴ A genotoxic substance that passes through and is cleared has still interacted with the cells it passed through. The absence of MOAH from tissue may reflect exactly this: metabolism into reactive forms, not harmless transit.

EFSA's assessment of toddler exposure to MOAH concluded:

"It is extremely likely (with 99–100% certainty) that MOEs lower than 10,000, raising concern for human health, are present for mean and high‐consuming toddlers."⁴

Margins of Exposure below 10,000 for a genotoxic carcinogen indicate that the safety margin between actual exposure and the dose that produced tumours in animal studies is uncomfortably thin. For toddlers, this margin is already inadequate — at current known exposure levels, from known sources, which do not include the spoon that stirs their porridge every morning.

The Counter-Position

The strongest defence of the current situation runs as follows: food-grade mineral oil is one of the most extensively studied food-contact substances in the world. EFSA's refining classification notes that white oils — the category that includes food-grade mineral oil — have "generally less than 5%" aromatics after processing.⁴ EFSA concluded that current MOSH dietary exposure does not raise concern. Migration from a wooden spoon, if it occurs, would contribute a fraction of total dietary mineral oil intake compared to the hours food spends in contact with recycled packaging, processing machinery, and environmental contamination. Regulators have not studied this pathway because the exposure arithmetic suggests it is trivially small compared to known sources.

This defence deserves to be stated fairly, because it is partially correct. It is true that packaging — which envelops the entire food mass for hours or days — almost certainly contributes more total mineral oil exposure than a spoon that contacts a fraction of the food for minutes. It is true that food-grade refining reduces aromatic content. It is true that EFSA's MOSH conclusion, at current modelled exposure, is reassuring.

Three things it does not explain.

First: "likely negligible" is an assumption, not a measurement. No migration study has quantified what transfers from a mineral-oil-finished spoon into hot food. The dose from this pathway is not known to be small. It is assumed to be small. These are different statements.

Second: the comparison between packaging and utensils is structurally flawed. Packaging sits passively around food at ambient or refrigeration temperatures. A spoon is stirred through food at simmering temperature, where mineral oil viscosity drops approximately fivefold, under continuous mechanical friction that refreshes the contact surface. The per-minute migration rate under these conditions could be substantially different from passive contact at 4°C. The analogy assumes equivalence where the physics diverges.

Third: EFSA's own framework identifies any unmeasured exposure pathway for a genotoxic carcinogen with no established health-based guidance value as a data gap that warrants investigation. You cannot dismiss a pathway you have never measured by comparing it to pathways you have. The dose argument holds only if you know the dose. Nobody has measured the dose from this pathway.

The decades defence — "mineral oil has been used on wooden utensils for a century without documented harm" — fails for the same reason. The absence of harm data exists because nobody has looked, not because someone looked and found nothing. This is the structure of the data vacuum: no regulation mandates the study, so no study is funded, so no data is generated, so no concern is raised, so no regulation is written. The vacuum sustains itself.

What Would Change This Analysis

Two studies would materially update this assessment. First: a peer-reviewed migration study measuring MOSH and MOAH transfer from mineral-oil-finished wooden spoons into a food simulant under conditions of intended use — stirring at simmering temperature for fifteen to thirty minutes. If migration proves negligible and 3+ ring MOAH are absent at the detection limit, the material concern weakens substantially. If migration is non-negligible or MOAH is detected, the regulatory gap becomes untenable. Second: independent analytical testing of commercially available "food-grade mineral oil" sold for cutting board conditioning, confirming or denying the presence of 3+ ring MOAH. EFSA's "generally less than 5%" aromatic content is a refining-class statistic, not a product-specific test.⁴

Either result — reassurance or confirmation — would be more valuable than the current state, which is neither.

The Levers

The alternative to mineral oil is not exotic, expensive, or new. It is older than petroleum. Linseed oil — pressed from flax seed — was used to finish wood in ancient China.¹⁷ Medieval European woodworkers combined beeswax with pine resin and linseed oil.¹⁷ The twelfth-century monk Theophilus Presbyter dissolved amber in heated linseed oil to create a polymerising varnish of remarkable durability.¹⁷ Drying oils were the standard finish for a thousand years. Mineral oil displaced them in the late nineteenth century — not because it performed better, but because it was cheaper: a petroleum refinery byproduct whose production cost was already embedded in gasoline processing, available at a price no agricultural oil could match.¹⁸

The irony is structural: the consumer who chooses a wooden spoon to "go natural" receives a product finished with the petroleum derivative that displaced the genuinely natural finishes centuries ago. The natural finish exists. The industrial substitute replaced it. The consumer does not know there were two bottles.

What you can do now (no cost):

Ask one question before your next wood utensil purchase: "What is the finish?" If the seller cannot answer — and most cannot — that is itself information. If you already own mineral-oil-finished utensils, avoid prolonged stirring in very hot food where possible, and when the time comes to re-oil, reach for a different bottle.

Choose a drying oil: food-grade raw linseed oil or tung oil. Not "boiled" linseed oil, which contains metallic driers (cobalt, manganese compounds) added to accelerate curing — these are not food-safe. Raw linseed oil polymerises more slowly (twenty-four to seventy-two hours) but produces a solid, crosslinked film that bonds to the wood and does not wash off or transfer to food.¹ It requires refreshing every few months rather than every few weeks — less maintenance, not more, because the finish stays where you put it.¹⁹

Note on walnut oil: it is an effective drying oil, but tree nuts are among the nine major food allergens classified by the FDA. For households with nut allergies, linseed or tung oil are the appropriate alternatives.

When replacing utensils:

Look for makers who disclose their finish composition. This is the minimum threshold of transparency: if you cannot determine what is on the surface that touches your food, the information gap has not been closed. Some makers already do this — Ziruma, for instance, specifies a plant-based finish of beeswax, lemon oil, and flaxseed oil, and explicitly states "no mineral oil."²⁰ The disclosure itself is the signal.

Unfinished wood — which you finish yourself with a drying oil — is another option. It requires one evening of application and a few days of curing. The result is a finish you chose, applied, and understand.

Artisan-made utensils with natural drying oil finishes are available from UK woodworkers. They cost more than a mass-market spoon (typically £8–25 versus £1–3) — a price gap that reflects, in part, the actual cost of a finish that cures over days rather than a petroleum derivative applied in seconds.

The gesture is the same. The bottle in your hand, the oil on the wood, the darkening grain. Same motion. Different chemistry. Different consequence. One finish bonds to the wood and stays. The other sits in the wood and waits to leave. Nobody told you there were two bottles.

Now you know.