EU Restricts 7 Additives in Baby Food: What Parents Should Know

On 29 January 2026, the European Union published two regulations that most parents will never hear about, and that change what may legally be put into food for the youngest and most medically vulnerable consumers in Europe.

Commission Regulations (EU) 2026/189 and 2026/196 amend the EU's food additives framework to restrict a specific group of thickeners, stabilisers and emulsifiers in products intended for infants, young children, and patients requiring foods for special medical purposes. The substances affected include carrageenan (E407), locust bean gum (E410), guar gum (E412), gum arabic (E414), xanthan gum (E415), pectins (E440) and starch sodium octenyl succinate (E1450).

These are not obscure chemicals. They appear across infant formula, follow-on formula, baby purees, thickened feeds for children with swallowing difficulties, and medical nutrition products. Several are also widely used in ordinary adult food, where they remain permitted.

That last point is the one worth sitting with. The EU has not concluded these substances are dangerous in general. It has concluded that the margin of safety that is adequate for an adult is not adequate for an infant — and the reasoning behind that distinction is what this article explains.

Every one of these additives exists to solve a texture problem. None of them contributes nutrition. Understanding their function makes it much easier to spot them on a label and to understand why they appear in infant products at all.

Several of these have entirely legitimate clinical uses. Locust bean gum and xanthan gum are used in thickened feeds for infants and children with dysphagia — difficulty swallowing — where thickening the liquid is the actual medical intervention. Carrageenan prevents liquid formula from separating on the shelf, which affects whether a feed delivers consistent nutrition from first bottle to last.

This is why the EU restricted rather than banned them. A blanket prohibition would remove products that some children clinically depend on. The regulations instead tighten the conditions of use — which categories, at what levels, for which ages.

The core of this decision is a principle that applies far beyond these seven additives, and it is worth understanding properly because it explains most infant-specific food regulation everywhere.

Additive safety is normally expressed as an Acceptable Daily Intake — an amount, expressed per kilogram of body weight, that can be consumed daily over a lifetime without appreciable risk. That framework quietly assumes a varied diet across many different foods, and it assumes a body that handles and excretes substances in a typical adult way.

Infants break both assumptions, in four distinct ways.

The first is dietary monotony. An exclusively formula-fed infant consumes one product, repeatedly, for months. There is no dietary variety to dilute exposure to any additive that product contains. An adult eating a food containing carrageenan is eating it among dozens of other foods; an infant may be receiving it in essentially every calorie consumed.

The second is intake relative to body weight. Infants consume a very high volume of food relative to their size — far higher, per kilogram, than at any later stage of life. Any per-kilogram exposure limit is therefore approached much faster.

The third is physiological immaturity. Infant kidney and liver function, gut barrier integrity and metabolic enzyme systems are still developing. The capacity to process and excrete a substance is not the same as an adult's, and the developing gut lining in particular is more permeable.

The fourth is developmental timing. Exposure during periods of rapid organ and immune system development may carry consequences that identical exposure in adulthood would not. This is the hardest factor to study and the one that most justifies precaution.

Together these mean an additive can have a perfectly reasonable safety margin for adults and a much narrower one for infants. That is the finding the January 2026 regulations act on.

Of the seven, carrageenan has attracted by far the most scientific and public argument, and it is worth separating the credible concerns from the overstated ones.

Carrageenan is extracted from red seaweed and has been used in food for a very long time. In infant formula it prevents the separation of fat and water, keeping the feed consistent.

The scientific debate centres on a distinction that is routinely lost in public discussion: the difference between food-grade carrageenan and degraded carrageenan, sometimes called poligeenan. Degraded carrageenan has a much lower molecular weight, is not permitted in food, and has been shown to cause intestinal inflammation in animal studies. Food-grade carrageenan is a different, higher molecular weight material.

The unresolved question is whether food-grade carrageenan can partially degrade — during processing, during storage, or in the digestive tract — and whether any resulting lower molecular weight fraction is biologically meaningful. Some studies suggest intestinal effects in animal models; others find no effect at realistic intakes. Reviews have reached different conclusions, and the disagreement is genuine rather than manufactured.

Regulators have landed in different places. The EU has now restricted its use in infant and young child categories. Some organic certification standards exclude it entirely. It remains permitted in general food across most jurisdictions, including in the United States.

The honest summary is that carrageenan is not established as harmful at food-grade quality and normal adult intakes, and that its use in a product consumed exclusively by infants is a different risk calculation — which is precisely the distinction the EU has now drawn.

Infant nutrition is the most tightly controlled food category in essentially every developed regulatory system, and comparing the approaches shows how unusual this category is.

Two features are common across all of them and are worth knowing.

Every system imposes compositional requirements rather than merely safety limits. Infant formula is one of very few foods where the law specifies what must be in it, in what quantity, not only what must not be. That reflects its status as a sole source of nutrition.

Every system restricts marketing of infant formula to some degree, following the World Health Organization's International Code of Marketing of Breast-milk Substitutes. The strictness varies considerably, but the principle — that this product should not be advertised like ordinary food — is near-universal.

The United States is currently modernising its side of this. The FDA's 2026 agenda includes beginning implementation of a plan to modernise infant formula nutrient requirements, and releasing exposure data on heavy metals and other contaminants in infant foods — lead, arsenic, cadmium, mercury and PFAS. That contaminant work runs in parallel to the additive question and is arguably the larger issue.

Additives are deliberately added. Contaminants arrive uninvited, and for infant food they are the subject of an equally active regulatory programme that parents often conflate with the additive debate.

Heavy metals are the primary concern. Lead, inorganic arsenic, cadmium and mercury occur naturally in soil and water and are taken up by crops. Rice is a particular issue for inorganic arsenic, which matters because rice cereal has traditionally been a common first food in many countries. Root vegetables take up cadmium. These are not manufacturing failures; they are agricultural realities requiring limits.

The FDA's Closer to Zero initiative addresses exactly this, and its 2026 commitments include establishing action levels for cadmium and inorganic arsenic in baby and toddler foods, alongside issuing guidance on preventive controls for chemical hazards. Action levels for lead in certain baby foods were set previously. The name of the programme reflects its logic: these substances cannot be eliminated from the food supply, so the objective is driving exposure progressively downward.

The EU sets maximum levels for contaminants in infant and young child foods that are typically stricter than those for general food, following the same reasoning that drives the additive restrictions — high intake relative to body weight, monotonous diet, developmental vulnerability.

Microplastics are the newest entry. The FDA's 2026 agenda includes research to identify methods for detecting, quantifying and characterising microplastics in food. This is at the measurement stage rather than the regulatory stage; you cannot set a limit for something you cannot reliably measure.

For a parent, the practical implication is that varying the grains a child eats — not relying exclusively on rice-based products — is a reasonable, evidence-aligned precaution, and one recommended by multiple health authorities independently of any specific additive question.

These two regulations are part of a broader tightening that has been running through 2026, and the context matters for anyone tracking where EU rules are heading.

On 20 January 2026, EFSA published revised guidance setting out the scientific data requirements for food additive applications under Regulation (EC) No 1333/2008. The revised requirements apply to all applications submitted from 20 July 2026 onward, covering both entirely new additives and proposed changes to the conditions of use of already-authorised ones.

This is a structural change rather than a substance-specific one. Raising the evidentiary bar for authorisation affects every future application, and it also affects the re-evaluation of existing additives — a programme the EU has been running for years, systematically reassessing substances approved decades ago under older and less demanding standards.

That re-evaluation programme is the reason additives keep appearing in EU news. Titanium dioxide, banned in 2022, came out of it. The restrictions on these seven thickeners and stabilisers came out of it. More will follow, because the programme is working through a large backlog of substances that were authorised before modern toxicological methods existed.

The direction is consistent: older authorisations are being re-examined against current science, and where the data is insufficient to confirm safety — particularly for vulnerable groups — use is narrowed rather than maintained by default. That is the precautionary principle operating as designed.

The EU's decisions do not bind anyone else, but they matter internationally for three reasons.

First, EFSA opinions are read globally. Regulators in many countries lack the resources to conduct independent toxicological reviews of every additive and routinely rely on EFSA and JECFA assessments. An EFSA conclusion tends to propagate, slowly, into other jurisdictions.

Second, manufacturing is global. A company producing infant formula for multiple markets faces a choice between maintaining separate formulations or adopting the strictest applicable standard across its output. For infant products specifically — where reputational risk is severe and production runs are long — harmonising upward is often the commercially rational decision. Parents outside the EU may therefore benefit from EU rules without their own regulator acting.

Third, these substances remain permitted in infant products in many countries, including in general use in the United States. A parent in a jurisdiction that has not restricted them is buying products that may contain additives the EU has now judged inappropriate for this age group.

That does not mean those products are unsafe. It means a specific regulator, applying a precautionary framework to a vulnerable population, reached a more restrictive conclusion than others have. Whether that conclusion propagates is a matter of time and regulatory capacity rather than scientific disagreement.

Knowing which products typically contain thickeners and stabilisers, and why, makes the label-reading exercise much faster.

One pattern is immediately useful: powdered formula generally requires fewer stabilisers than ready-to-feed liquid, because the suspension problem those additives solve does not exist until the powder is mixed. Parents concerned specifically about stabilisers, and without a clinical reason to use liquid, often find powder has a shorter ingredient list.

A second pattern concerns purees. Single-ingredient purees — just apple, just carrot — frequently contain nothing else at all. Multi-ingredient pouches with a smooth, uniform texture and long ambient shelf life are more likely to contain pectins or starches. Comparing two pouches on the same shelf usually shows the difference within seconds.

The exception that overrides all of this is clinical need. Where a thickener is present because a child has reflux or a swallowing difficulty, its presence is the reason the product works. That is not a case for seeking a simpler ingredient list.

Three things.

The regulations are narrow and specific. Seven additives, restricted in infant, young child and medical nutrition categories, effective from January 2026. They are not general bans, and the substances remain permitted in ordinary adult food across the EU.

The reasoning generalises further than the regulations do. The principle that infants require a separate safety assessment — because of monotonous diet, high intake relative to body weight, physiological immaturity and developmental timing — applies to every additive in every country, not only these seven in Europe. It is the most portable idea in this entire area of regulation.

And the practical response for a parent is modest. Read the ingredient list, understand what each thickener is doing in that specific product, prefer simpler formulations where a clinical reason for the additive does not exist, and treat any change to a prescribed feed as a medical decision rather than a label-reading one.

The EU has not told parents that these products were harming children. It has said that for this group of consumers, the evidence was not sufficient to justify continuing to allow these substances at previous levels — and in a precautionary system, insufficient evidence is itself a reason to narrow use.