The Pouch

In a 2024 paper based on supermarket data collected in 2021, University of Glasgow researchers counted the words in the baby-food aisles of three large UK supermarkets. Across 341 commercial baby foods, they recorded 1,666 marketing claims, a median of five per pack. Underneath those sat 1,003 separate emotional keywords, a median of three on every single product. The words clustered into four themes: contentment, happiness, pride, and, dominating all of them, love. "Smile from the inside." "Happy tummies." The authors concluded, carefully, that this kind of marketing "might exploit parental vulnerabilities and influence their purchasing."¹

Three emotional words per pack is not an accident, and it is not decoration. It is a dense emotional-marketing load in a category where parents are unusually vulnerable. And it is aimed at the single purchase where the buyer has the most love at stake and the least ability to check what she is being told.

Consider what a parent is actually asking when she lifts a pouch off the shelf. Not "does this sound nice." The real question, underneath, is harder: is this good nutrition for my baby, and has someone checked it? That question is close to unanswerable in the four seconds a busy person stands in an aisle. So the mind does what minds do under load. It quietly swaps the hard question for an easy one the pack has already answered. Does this look wholesome? Does it say the right words? The pouch says "no added sugar." It says "organic." It says "healthy," "no nasties," "only good stuff." It is arranged on a tidy numbered ladder: stage one, stage two, from four months, from six. And the feeling that arrives is closure. Someone designed this to be good for my baby, and someone checked it.

That feeling is doing enormous work. The question is whether the words underneath it can carry the weight.

What the words actually certify

Start with the most important thing, because the whole piece depends on it: the words are, very largely, true.

"No added sugar" means no sugar was added. That is a literal, checkable, accurate statement about the manufacturing step. Nobody tipped sucrose or syrup or fruit-juice concentrate into the pouch. "Organic" certifies that the farming inputs met an organic standard. "Stage one" and "from four months" describe a marketing age-ladder, printed on the pack. And the product is safe in the ordinary food-hygiene sense, governed by compositional standards (in England, the Processed Cereal-based Foods and Baby Foods for Infants and Young Children Regulations 2003) and overseen by the Food Standards Agency.² None of this is a lie. There is no villain here, and a parent who reaches for the reassuring pouch has not been fooled by a con.

The gap is not between the words and the truth. It is between what the words certify and what the parent reads into them.

Take the sugar. In 2015 the government's own Scientific Advisory Committee on Nutrition set out what counts as a "free sugar," the kind public-health guidance tells us to limit. The definition is precise, and it contains a clause most parents have never seen. Where the cellular structure of fruit or vegetable "had been broken down (such as in blended, pulped, pureed, extruded or powdered products) the sugars should be treated as free sugars."³ Canned, stewed, dried or pressed fruit keeps its sugar "intrinsic." But blend it, and the same sugar is reclassified, by the government's own standard, as free sugar.

A fruit pouch is, definitionally, a blended-fruit product. So a pouch can be entirely accurate in saying "no added sugar" and, by the government's own definition, still be a free-sugar food. The sugar behaves in the body more like juice than intact fruit. (Not identical: blended fruit keeps more fibre than strained juice, so this under-claims rather than over-claims. But close.) The label describes the manufacturing step. The definition describes the physical state of the sugar. Both are true at once. The parent reads the first and assumes it means the second.

This is the part worth slowing down on, because it is the whole mechanism in one place. The free-sugars definition is not secret. It has been public since 2015. A paediatric consultant, someone who knows the difference between "no sugar was added" and "low in free sugars," could know it exists. And yet the pack still produces the feeling of closure, because the wholesome word arrives first and answers the easy question before the hard one is ever consciously asked. The knowledge was available. The substitution happened anyway. That is not carelessness. It is how the swap works.

When the BBC's Panorama lab-tested baby-food pouches for its April 2025 investigation "The Truth about Baby Food Pouches," it examined 18 products across six brands. It reported that some fruit pouches promoted as "no added sugar" contained the equivalent of four teaspoons of sugar, with many products higher in sugar than the daily recommendation for a one-year-old.⁴ Those are free sugars from blended fruit, not added sugar. The distinction matters and the report holds it. But the parent buying the pouch cannot see it, because the only sugar word on the front of the pack is the reassuring one.

What the words don't even attempt

The same investigation found something quieter. The pouches it tested were low in iron, between 0.3 and 0.7 milligrams per serving, against an advised intake of around 7.8 milligrams a day — the UK reference nutrient intake for infants aged 7 to 12 months.^4,5 The investigation concluded such products should be used sparingly, not as a meal replacement.⁴ This is heterogeneous: savoury and fortified lines differ from fruit-led purées, and "all pouches are thin" would be false. But the pattern is real for the sweet fruit-led products that carry the wholesomest words.

"Healthy" and "organic" and "no nasties" are read as nutritionally optimised and checked for my baby. What they certify is marketing positioning and basic compliance. The Office for Health Improvement and Disparities, the public-health body inside the Department of Health, names this directly in its own guidance, calling it the "health halo effect" or "implied health claims."⁶ The regulator can see the mechanism. It is in the document.

And then there is the question the words don't touch at all.

The metals are in the soil, and the kitchen won't save you

Here is where most coverage of baby food goes wrong, and where this report parts company with it.

Lead, arsenic and cadmium are present in a great many infant foods. They are also present in a great many home-cooked ones, because they are not a manufacturing defect. They are in the soil. Arsenic, lead and cadmium are elements of the earth's crust, topped up by legacy pollution, and crops take them up through their roots from soil and water the same way they take up nutrients. Rice is the standout. Rice is unusually efficient at taking up arsenic, especially under flooded-paddy conditions, and brown rice tends to carry more than white. Root vegetables (carrots, potatoes, beetroot) draw up lead and cadmium directly from the soil they grow in.⁷ Organic certification regulates farming inputs. It does not change the metal already in the ground. Organic certification does not guarantee lower heavy-metal levels.

The cleanest demonstration of what this means came from US testing. The campaign group Healthy Babies Bright Futures tested 288 foods. It found detectable heavy metals in 94 per cent of store-bought baby food, and in 94 per cent of homemade purées and family foods. Its conclusion, in its words: levels "varied widely by food type, not by who made the food."⁸ ("Detectable" means a measurable trace, not a harmful dose. The distinction is load-bearing and the report keeps it.) This is US data, and it should not be read as a snapshot of the UK shelf. But the mechanism it reveals is not national. It is agronomy.

So before anything else, two things follow, and they are the most freeing facts in this entire subject.

First: cooking from scratch is not a contamination cure. The instinct that the homemade purée is the clean, safe, virtuous choice, that the pouch is the compromise a better parent wouldn't make, is, on the evidence, simply wrong. The metals come with the crop, not the convenience. A parent using pouches has not exposed her child to something a home cook avoids.

Second, and just as important: this is not a reason to panic. For most families, most of the time, the honest, evidence-warranted thing to do is to keep using pouches. Vary them, glance at the sugar, and otherwise carry on. The UK has binding maximum limits for these contaminants, and when the Food Standards Agency surveyed metals in commercial infant foods and formula, it found infant products generally within those limits, with stricter limits for infant formula than for general food.⁹ (The FSA has separately flagged inorganic arsenic as a concern for high consumers of rice-based products, which is one reason varying the grain matters.) Levels in most products are low. And the public-health consensus is unambiguous that milk plus a varied diet matters far more to a child's outcome than any single pouch ever could. Pouches are safe in the hygiene sense, affordable, portable, sometimes fortified, and a genuine lifeline for parents short of time and money. None of that is in dispute, and a report that left a parent feeling otherwise would have failed.

With that established, the guilt removed and the panic refused, we can look at the one thing that is actually missing, calmly.

The thing nobody has to show you

It would be easy, and wrong, to say "nobody checks." Somebody does. The UK's contaminant limits are legally binding. Under retained Regulation 1881/2006, a food that exceeds the maximum level for a contaminant cannot lawfully be placed on the market, and the same framework provides for official-control sampling and analysis by the competent authorities to enforce those limits.¹⁰ That is real checking: independent, statutory, backed by enforcement. The parent who feels "someone stands behind this shelf" is not feeling a phantom.

But look closely at what that checking is, and what it isn't. Official-control sampling is risk-based and partial. It tests some products, some of the time, to police a ceiling across the market. It does not test every pouch or every batch. And crucially, it produces no result the parent can see. Under the UK framework, there is no requirement for manufacturers to test each finished pouch for heavy metals or to disclose a finished-product result, to a regulator or to a parent. The limits are a ceiling with enforcement behind it. They are not a per-product, "this-pouch-was-tested-and-cleared" certificate.

So the parent's real question (has someone checked this one, and can I see it?) has a precise answer. A ceiling is enforced at the level of the market. The parent cannot tell, from public UK labelling, whether this finished pouch or batch was tested for heavy metals, or see any result if it was. The verification she assumes is sitting behind the wholesome word is not a per-product verification she can consult. There is no public, finished-product disclosure to read.

This is not a scandal and it is not bad faith. It is what happens when a test is costly to run and nobody is required to disclose one. Finished-product metal testing needs specialist laboratory analysis, the cost falls heavily up front, and publishing a result creates liability without reward. So the rational thing, in a competitive category with no disclosure requirement either way, is to test raw ingredients to an internal standard and publish nothing. No single actor's incentive ever owned a public, finished-product disclosure. It fell into the gap between all of them.

And here is the dot that makes the gap visible. In August 2025, the Office for Health Improvement and Disparities published voluntary guidelines for commercial baby food and drink, the most significant action on this category in years. They set targets for sugar and salt. They addressed labelling and marketing and the age at which products are promoted. They named the health-halo effect by name. Read the document end to end, and there is one subject it never raises: a search of its full text returns zero mentions of "contaminant," "heavy metal," "arsenic" or "cadmium."⁶ The 2025 instrument regulated the words and left the disclosure unowned. When some stakeholders proposed the guidelines be made mandatory, the document records, "Businesses did not comment on this point."⁶

What it costs to over-read a word

The market noticed the scrutiny. The Grocer, citing NIQ data, reported that in the year to the end of February 2026 branded baby-food volumes fell 14.7 per cent, 26.1 million fewer units and around £33 million of value, while supermarket own-label fell only 2.9 per cent.¹¹ Parents repriced the premium brands whose whole proposition is reassurance, and stuck with the cheaper option. It looks like the market correcting itself.

But look at what that correction did and didn't do. It punished reputation. It did not close the gap. The own-label pouch a parent traded down to is not guaranteed lower in free sugar, more nourishing, or any more disclosed as a finished product than the brand she left. No finished-product disclosure appeared. The sales crash is a verdict on trust, not a verdict on the pouch. A market can reprice a reputation overnight and leave the actual information a parent lacks exactly where it was.

There is a similar shape to the age ladder. For years, pouches were marketed "from four months," even as NHS advice has been to introduce solids at around six.¹² Some movement was already underway before the broadcast: Ella's Kitchen announced a relabelling in March 2025, rolling out from October 2025 and completing by March 2026; Piccolo and Aldi also realigned to the six-month advice during 2025.¹² A genuine clinical standard does not move in step with public scrutiny of a category. The age rung that felt to a parent like a developmental fact handed down by experts turns out to have been, substantially, a marketing decision, one the market itself is now quietly reversing. (The system does self-correct, which is real and should be said.) But notice the timing. The pouches were sold "from four months" for years. The closure parents felt (this is the right age, experts decided it) was felt long before the documented question surfaced. The feeling arrived first. The evidence that should have disturbed it arrived later, if at all.

The two ways to get this wrong are the same way

It is tempting to picture two kinds of parent here: the trusting one who believes the wholesome word, and the worried one who reads "arsenic in baby food" and resolves to cook everything from scratch in fear. They look like opposites. They are the same mistake wearing different clothes.

The trusting parent outsources the hard question to a wholesome word: "no added sugar," "organic." The worried parent outsources it to a wholesome category: "natural," "homemade," "nothing processed." Both are substituting a feeling for the evidence. And the same body of evidence dissolves both. The blended-fruit sugar is free sugar, so the reassuring word over-reads. The metals are soil-inherited, with homemade no cleaner than store-bought, so the panic over-reads too. Neither parent can see a finished-product disclosure. Both have read the pack. The error is not trusting, and it is not distrusting. The error is letting a signal, a word or a category, stand in for a question it was never built to answer.

This is also why this report cannot, structurally, slide into alarm. The moment it frightened the trusting parent, it would have become the panic pole it exists to refuse. The discipline isn't a tone held by willpower. It's built into the finding.

What this is, exactly, is narrower than it first looks. The temptation is to file it under "marketing makes products look healthier than they are," which is a real pattern, and one worth decoding, but not this one. This is not, in general, a claim that performs better than it delivers. It is something stranger. A word stands in for a piece of knowledge that the parent cannot consult in any public form: the finished-product result no one is required to show her. And it stands in on the one purchase where she has the most love at stake and the least power to check, where the two ways of getting it wrong mirror each other exactly. The novelty isn't that a word over-promises. It is that the question underneath is unanswerable at the shelf, and the over-reading runs in both directions at once.

What this would take to be wrong

The honest landing here is calm, and it depends on a real claim about the world that could turn out otherwise.

The argument is that the gap is one of claim-scope, nutrition disclosure and shelf-invisibility, not of demonstrated harm. Suppose a comprehensive, recent UK survey tested finished pouches and found a meaningful share exceeding the binding limits for these metals at the amounts babies actually eat. Then the calm register would be wrong, the worried parent's instinct would have been right, and the words would be hiding a real danger rather than a narrow scope. That survey does not currently exist. The FSA's infant-metals survey is now over a decade old, and no one publishes a finished-product result on the specific pouch a parent buys, which is itself part of the point.⁹ The reverse case matters too. If a current survey confirmed UK pouches overwhelmingly within limits and showed the free-sugar issue confined to a small minority of fruit-led lines, the residual gap would shrink further toward "a low-stakes thing you can't see at the shelf," and the reassurance would be warranted almost entirely.

The behavioural reading rests on its own evidence, and that evidence should be held honestly. That parents substitute a wholesome signal for the verification they can't perform is the most economical explanation of several converging facts. There is the measured UK saturation of love-marketing. There is the established health-halo effect, by which a single positive attribute inflates a product's perceived healthiness. There is international choice-experiment evidence from Australia, measuring perception rather than purchase, in which regulated "no added sugar / no salt" claims made participants nearly fourteen times more likely to rate a product the healthiest.¹³ And there is the plain fact that the words moved markets. But the claim that the feeling of having checked is indistinguishable from having checked is an interpretation of that evidence, not a measured finding. If the halo effect replicates at a smaller size than the headline studies report, the mechanism described here is real but weaker than the strongest version of this argument implies. The reader is entitled to know which parts are measured and which part is the reading.

At the shelf, on a Saturday

None of this asks a parent to cook from scratch, fear the aisle, or feel she has done anything wrong. The single most useful thing is to know what the wholesome word certifies, which is marketing positioning and basic safety compliance, and what it does not: that the pouch is nutritionally optimised, or that a finished-product result for what the soil contains has been put anywhere she can see it.

So the one glance worth taking is on the back of the pack, not the front. The sugar-per-100g figure and the ingredients order (which manufacturers must already list by descending weight) tell you in a second whether you are holding mostly sweet fruit purée or something with real vegetable, protein and iron in it. Varying the grains and vegetables across the week, rather than rice-based every day, spreads the soil-inherited exposure without any fuss. That's it. One look, low-stakes, and nothing to feel guilty about.

Beyond that, the alternatives are ordinary. Plain mashed banana or avocado is a pouch's equal and costs nothing extra. A supermarket own-label pouch with a short, savoury-inclusive ingredients list, used as part of a varied diet alongside spoon-fed textured food, is a perfectly good choice. And the most honest thing to tell most families is the calmest: keep using pouches. Vary them, glance at the sugar, and otherwise relax, because on the current evidence that is the warranted read, not a hedge.

The thing a parent cannot fix at the shelf is the thing worth naming out loud. No front-of-pack signal tells her how much free sugar a "no added sugar" pouch holds, whether it is nourishing or merely sweet, or whether a finished-product result for the metals the soil puts in it has been disclosed anywhere. The fix for that is not parents squinting at ingredient lists in fear. The Commercial Baby Food Review partnership launched in March 2026 called for mandatory reform on composition, labelling, marketing, age-of-use claims and enforcement.¹⁴ That would tighten much of what the wholesome word currently outruns. It would not, on its own, close the contaminant-disclosure gap — which is the part a parent still cannot see, and the part still worth naming.

Until then, the word on the front of the pouch will keep answering the easy question. The hard one is still hers to ask.