The ingredient list that worries people on a carton of oat drink is almost always the least important part of the label. A randomised controlled trial published in 2026 in Clinical Gastroenterology and Hepatology gave healthy adults four weeks of either carboxymethyl cellulose, polysorbate-80, carrageenan, soy lecithin, native rice starch or nothing. No effect on faecal calprotectin, CRP, serum lipopolysaccharide-binding protein, cholesterol or the other metabolic markers.

Meanwhile, that same carton can contain twenty-five times less iodine than a glass of milk, and the figure appears nowhere on the packaging.

What the nutrition panel says, and what it leaves out

Here are the figures from the Swiss Food Composition Database, scaled to a 200 ml glass. The iodine column is the one you will almost never find in a comparison.

Drink, 200 ml glass	kcal	Protein g	Sugars g	Calcium mg	Vitamin B12 µg	Iodine µg
Whole milk UHT	136	6.4	9.2	240	0.32	19.0
Semi-skimmed milk 1.5% UHT	92	6.4	9.6	240	0.50	20.0
Soya, plain	80	7.2	2.8	24	0	2.6
Soya, fortified	80	7.2	2.8	240	0.56	2.6
Oat, plain	94	1.0	7.2	2	0	0.8
Oat, fortified	96	1.0	7.2	220	0.50	16.6
Almond, plain	66	2.2	2.6	24	0	0.8
Almond, fortified	66	2.2	2.6	220	0.72	0.8
Rice, plain	108	0.4	5.6	10	0	1.0
Rice, fortified	108	0.4	5.6	220	0	4.2
Soya, chocolate	152	7.6	16.6	30	0	2.6

Three things stand out. First, fortification works remarkably well for calcium: a fortified soya or oat drink delivers 220 to 240 mg per glass, which is 27.5 to 30% of the Swiss nutrient reference value of 800 mg, exactly like milk. The unfortified version of the same product delivers 2 to 24 mg. That is a factor of ten to a hundred, and the only difference between the two cartons is one line in the ingredient list.

Second, protein does not follow fortification at all. Soya gives 7.2 g per glass, more than milk’s 6.4 g. Oat gives 1.0 g and rice 0.4 g, six to sixteen times less. No additive fixes that, because it is not an additive question but a raw material question.

Third, iodine. Fortified soya has 2.6 µg per glass against 20 µg for semi-skimmed milk, nearly eight times less. Fortified almond has 0.8 µg, twenty-five times less. Fortified oat does better at 16.6 µg. Fortification puts back calcium, B12, vitamin D and riboflavin, but it does not consistently put back iodine, and nothing on the packaging flags this. The Swiss Federal Food Safety and Veterinary Office funds research specifically on the contribution of milk and dairy products to the population’s iodine supply, which indicates how much that source matters.

What the 2026 clinical trial actually measured

The trial in Clinical Gastroenterology and Hepatology deserves describing precisely, because it is the first to test several emulsifiers in humans under a controlled protocol.

Sixty healthy participants first followed an emulsifier-free diet for two weeks. They were then randomised, double-blind, for four further weeks, with the addition of carboxymethyl cellulose, polysorbate-80, carrageenan, soy lecithin, native rice starch or no additive, all delivered in brownies.

The results point in several directions. After the two weeks of emulsifier-free diet, cholesterol fell (P = .00006). Under supplementation, alpha diversity of the microbiota stayed stable, but microbial composition was affected by treatment. Concentrations of all short chain fatty acids were lower in those consuming carboxymethyl cellulose, a pattern mirrored by the other emulsifiers though not always reaching significance. In those consuming carrageenan, transcellular intestinal permeability increased relative to baseline (P = .04).

Against that, no differences were seen between placebo and emulsifiers in faecal calprotectin, CRP, serum lipopolysaccharide-binding protein, cholesterol or other metabolic markers at the end of the intervention. Serum inflammatory and cardiometabolic proteins were unchanged.

The authors call their trial exploratory themselves. Their conclusion is measured: emulsifier supplementation lowered short chain fatty acid concentrations compared with placebo, without affecting intestinal or systemic inflammation or metabolic endpoints, and these results point towards potential intestinal benefits of limiting dietary emulsifiers, which requires further investigation.

That is more nuanced than either story in circulation. The additives are not inert: they shift microbiota composition and lower short chain fatty acids, which are the useful product of fibre fermentation. And they do not produce the advertised intestinal inflammation either, at least not in healthy adults over four weeks.

Carrageenan, the special case

Carrageenan (E407) is the only additive on this list for which a specific signal emerged from the trial: an increase in transcellular intestinal permeability relative to baseline, at P = .04.

A comprehensive review in Annals of Nutrition and Metabolism in 2026 puts that in context. Carrageenan is a sulfated polysaccharide extracted from red seaweeds. The evidence is contradictory: some studies report antioxidant, antiviral, antitumor and immunomodulatory effects, others suggest impaired glucose metabolism, intestinal inflammation and immune dysregulation. The authors conclude that these effects depend on carrageenan type, molecular weight, purity, dosage and host-related factors, and that carrageenan cannot be uniformly classified as beneficial or harmful. They note that the FDA and EFSA generally consider it safe, with lingering concerns about infant formulas and long-term exposure.

Work published in Clinical and Experimental Allergy in 2026 tested the kappa, iota and lambda forms on intestinal epithelial cells and a gut-on-a-chip model. All three showed dose-dependent cytotoxicity, with lambda-carrageenan producing the most severe effect at relatively lower doses. These are cells in culture, not people, and dose matters. But it fits the one human result that came out of the randomised trial.

My reading: if you want to drop one additive from your basket, that is the one, not the plant gums. The move is simple, carries no nutritional cost, and carrageenan-free products exist across essentially every range. It is not a health emergency either.

The gums: what they are, what they do

Thickeners worry people most while being the best understood.

Guar gum (E412) comes from the seed of Cyamopsis tetragonoloba, locust bean gum (E410) from carob seeds, xanthan gum (E415) from a bacterial fermentation and gellan gum (E418) likewise from fermentation. These are polysaccharides, which is to say soluble fibres in the chemical sense. They are not digested in the small intestine and are partly fermented by the colonic microbiota.

Their technical job is simple: keeping particles in suspension that would otherwise settle. Gellan gum is the one you meet most often in oat and almond drinks, at very low doses, typically around 0.01 to 0.05%, because it forms a network at minimal concentration. Without it, the carton would need shaking and the added calcium would settle at the bottom, making the fortification uneven from one glass to the next.

That is the point additive lists miss: in a fortified drink, the thickener is not cosmetic, it is what guarantees the advertised 220 mg of calcium actually ends up in your glass rather than in the bottom of the carton.

An in vitro study published in Nutrients in 2026 tested twelve food additives, including xanthan gum and kappa carrageenan, on faecal fermentations from six adults with Crohn’s disease in remission and six healthy controls. This kind of work measures fibre fermentation capacity and short chain fatty acid production outside the human body. It is useful for ranking hypotheses, not for concluding about clinical effects.

At the doses present in a plant drink, gums can cause bloating in sensitive people, as any soluble fibre would. That is not a toxicity signal.

Lecithin and emulsifiers

Lecithin (E322) is a mixture of phospholipids, most often extracted from soya or sunflower. Its job is holding together the fat phase and the water phase of a drink that would otherwise separate.

In the 2026 randomised trial, soy lecithin was among the five substances tested and stood out for no inflammatory or metabolic effect. It still matters for a different reason: people allergic to soya need to spot it. The law helps here, since the Swiss ordinance on food information requires that allergenic ingredients always be indicated and, under its article 11, be highlighted through typeface, style or background colour.

Carboxymethyl cellulose and polysorbate-80, the two emulsifiers behind most of the microbiome literature, are in fact uncommon in everyday plant drinks. They dominate the research because they are used in ice creams, sauces and many ultra-processed products. Transferring those findings straight to a carton of oat drink means changing the product mid-argument.

What the ingredient list actually tells you

A few Swiss rules make the label more readable than it looks.

Ingredients are listed in descending order of weight. On an oat drink, water therefore comes first, followed by oats, whose percentage is often declared. That percentage is far more informative than the presence or absence of gellan gum: a drink with 15% oats and one with 6% are not the same product.

Additives must be named by function category followed by the name or the E number. Seeing “stabiliser: gellan gum” or “stabiliser: E418” means exactly the same thing; the number is not a marker of chemical synthesis, it is an identifier.

Flavourings must be declared as such. The wording “natural flavouring” has a legal definition and does not mean an absence of processing.

The nutrition declaration only lists vitamins and minerals when present in significant amounts. For beverages that threshold is 7.5% of the reference value per 100 ml, against 15% for other foods. In practice that means 60 mg of calcium, 0.19 µg of B12 or 11.25 µg of iodine per 100 ml. That is precisely why iodine appears on none of these cartons: at 1.3 or 0.4 µg per 100 ml you are far below the threshold, and the law does not require declaring what is absent.

The most useful practical point is therefore the reverse of the usual reflex. A short ingredient list is no guarantee of nutritional quality: a plain almond drink with no additives gives 2.2 g of protein and 24 mg of calcium per glass, while a fortified soya drink with two stabilisers gives 7.2 g and 240 mg.

Sugars: where there really is a choice

Sugar is the one line on the label where a large difference reliably shows up, and it is not an additive question.

A plain oat drink contains 7.2 g of sugars per 200 ml glass without a single gram having been added. Those sugars come from the enzymatic hydrolysis of oat starch during manufacture, which is what gives these drinks their characteristic sweetness. Whole milk contains 9.2 g, as lactose. Plain soya contains 2.8 g.

The difference appears with flavoured versions. Chocolate soya climbs to 16.6 g of sugars per glass, six times the plain version, for 152 kcal against 80. That is the gap that counts, and it is read in the nutrition panel, not in the additive list.

Intense sweeteners are rare in basic plant drinks and more common in so-called light versions. They are subject to authorisation and to maximum levels.

What fortification demands technically

That calcium can be added at all is less obvious than the table suggests, and it explains some of the additives.

Calcium carbonate, the most common form, barely dissolves in water. It sits in the liquid as fine solid particles and settles under gravity. That is exactly where gellan gum comes in: it forms a weak gel holding the particles in suspension without noticeably thickening the drink. A fortified drink without a stabiliser would not be a better product, it would be a product where the first glass from the carton held too little calcium and the last far too much.

Vitamins raise a different problem. Vitamin D is fat-soluble and riboflavin is markedly light-sensitive, which is why these drinks are sold in opaque cartons rather than clear glass bottles. The packaging is therefore part of the fortification, not just a container.

Heat treatment, finally, explains why these products keep for months unrefrigerated. UHT processing briefly heats above 135°C and makes the drink microbiologically stable without any preservative being needed. Anyone hunting for preservatives in the ingredient list of a long-life plant drink generally finds none: the shelf life comes from the process, not from an additive. That stops applying once opened, hence the fridge consumption period printed on the carton.

This last point connects to the wider logic of the plant pantry: the date printed on an unopened carton and the period after opening answer two different questions.

What fortification changes across a whole diet

A modelling analysis published in Food Science and Nutrition in 2026 takes this past the single glass. It draws on the German DONALD cohort, with 7676 three-day weighed dietary records from 1072 participants aged 3 to 18, and simulates progressively replacing dairy foods with plant-based alternatives.

The findings are instructive. Protein intake exceeded reference intakes at every substitution level, with a smaller drop for soya than for oat. Vitamin B2 fell below the reference from 50% substitution onwards and vitamin B12 from 75%. Only full substitution with fortified products brought median B2 and B12 intakes back to the references. Calcium fell significantly and stayed well below the reference at all substitution levels. Iron, by contrast, rose and nearly reached the reference at 100% substitution. Greenhouse gas emissions and land use fell significantly as substitution increased.

Two lessons follow. Fortification is not a marketing detail, it is what decides whether a substitution works nutritionally. And on its own it was not enough for calcium in this population, a reminder that dairy was not the only source involved.

This study covers German children and adolescents, with eating habits and product fortification levels that are not Switzerland’s. It indicates a direction, not a transferable number.

How to choose in practice

Three questions are enough, in this order.

Which plant base? This is the choice that matters most. Soya is the only one that matches milk on protein, at 7.2 g per glass. Oat and rice are carbohydrate drinks, not protein sources. Almond sits between the two at 2.2 g. If these drinks make up a large share of your protein intake, the plant protein guide gives the reference portions.

Fortified or not? If the drink replaces milk in your diet rather than adding to it, fortification changes everything: calcium goes from 24 to 240 mg per glass, B12 from 0 to 0.56 µg. If you have a plant-based coffee now and then, the question is secondary. For calcium outside dairy, see calcium without dairy sources.

Plain or flavoured? Going from plain soya to chocolate soya adds 13.8 g of sugars and 72 kcal per glass. That is the one decision on this page that shows up in a daily intake.

Additives come after those three questions, not before. And if you want to drop one, carrageenan is the best-justified candidate on current data.

For full control over the composition, making it at home remains an option. It has a cost: without fortification, a homemade almond drink gives roughly 24 mg of calcium per glass and no B12, against 220 mg and 0.72 µg for the fortified shop equivalent. That is a genuine trade-off, not an automatic improvement.

Common mistakes

Judging a product by the length of its ingredient list. The shortest product can be the poorest. Fortified soya with two stabilisers beats additive-free plain almond on both calcium and protein.

Reading an E number as a marker of synthesis. E410 is locust bean gum, E407 comes from red seaweed. The number identifies, it does not qualify.

Transferring carboxymethyl cellulose studies to plant drinks. Those emulsifiers dominate the literature because they are elsewhere, in ice creams and sauces.

Believing fortified means equivalent to milk. On calcium, B12, vitamin D and riboflavin, yes. On iodine, often not, and on protein it depends entirely on the plant base.

Choosing a rice drink as a child’s main milk substitute. At 0.4 g of protein per glass, it is not a nutritional equivalent.

Focusing on additives and ignoring the sugars line. The gap between plain and flavoured, six-fold on soya, is the only genuinely large gap on the label.

Adapting to your situation

You are replacing milk completely. Take fortified soya: it is the only base that holds up on protein and, once fortified, on calcium and B12. Keep an eye on iodine separately, since that source disappears.

You use a plant drink in coffee. Fortification matters little at that dose. Barista versions often contain acidity regulators, typically phosphates, to stop the drink curdling in hot coffee. That is a legitimate technical use.

You have a sensitive gut or irritable bowel syndrome. Carrageenan is the first to drop given the permeability signal from the 2026 trial. Gums can also cause bloating, as any soluble fibre can; try removing them one at a time rather than all at once.

You are allergic to soya. Soy lecithin must be highlighted in the ingredient list. Watch precautionary allergen labelling too, which is markedly more frequent on almond and coconut drinks than on oat or soya ones. For the rest of your diet, the page on cooking without soy covers the substitutions.

You are feeding a child. The authorities’ reservations about carrageenan concern infant formula specifically. A plant drink is not an infant formula in any case, and the choice belongs with medical advice.

Limits of this page

The composition figures come from the Swiss Food Composition Database and are averages for generic products. A specific product can differ substantially, and that is precisely the message of the work on market variability: an analysis of 68 drinks on the Ecuadorian market published in 2026 found fortification concentrations that frequently did not match those in cow’s milk, with wide dispersion even among products sharing a plant base.

The 2026 randomised trial covers 60 healthy adults over four weeks. The authors describe it as exploratory themselves. It says nothing about exposure over years, nor about people with inflammatory bowel disease. The work on carrageenan alone involves cell cultures and gut-on-a-chip models. The DONALD modelling study covers German children and adolescents.

The declaration thresholds and labelling rules are those of Swiss law as at 1 July 2025. Nothing on this page is medical advice.

Frequently asked questions

Are the additives in plant milks dangerous? The most complete controlled trial available found no effect on intestinal or systemic inflammation or on metabolic markers after four weeks in healthy adults. It did measure lower short chain fatty acids and a shift in microbiota composition.

Should I avoid carrageenan? It is the additive with the heaviest file, with increased transcellular permeability in the human trial (P = .04) and dose-dependent cytotoxicity in cells. Dropping it costs nothing nutritionally. Regulators consider it safe at dietary doses.

Is gellan gum chemical? It is produced by bacterial fermentation, like vinegar or yoghurt. It is a polysaccharide, so a soluble fibre in the chemical sense.

Is an additive-free drink better? Not nutritionally. Without a stabiliser, added calcium settles, making the fortification uneven. A plain additive-free drink is simply an unfortified drink.

Which plant drink comes closest to milk? Fortified soya: 7.2 g of protein and 240 mg of calcium per glass, against 6.4 g and 240 mg for milk. The remaining gap is iodine.

Why is iodine not on the packaging? Because there is not enough of it. The declaration threshold for a beverage is 7.5% of the reference value per 100 ml, which is 11.25 µg, and most of these drinks sit at 0.4 or 1.3 µg.

Do oat drinks contain added sugar? Usually not. Their 7.2 g of sugars per glass come from the enzymatic breakdown of oat starch during manufacture.

Is homemade plant milk healthier? It contains fewer additives and far less calcium and vitamin B12, unless you compensate elsewhere. See also the ideal beginner’s pantry for context on stocking these products.

The same reasoning applies to additives in meat substitutes, where the base and the salt matter more than the E list, and to the more controversial additives in that category.

Sources

• FSVO, Swiss Food Composition Database
• FSVO, Iodine: research projects on iodine supply in Switzerland
• FDHA Ordinance on Food Information, SR 817.022.16
• Clinical Gastroenterology and Hepatology 2026, randomised controlled trial of five dietary emulsifiers (PMID 40816342)
• Annals of Nutrition and Metabolism 2026, comprehensive review of carrageenan (PMID 42560937)
• Clinical and Experimental Allergy 2026, effects of kappa, iota and lambda carrageenans on the epithelial barrier (PMID 41942402)
• Nutrients 2026, in vitro effects of twelve additives on the microbiome and fibre fermentation (PMID 41754185)
• Food Science and Nutrition 2026, DONALD modelling of dairy substitution (PMID 42254422)
• Journal of the Science of Food and Agriculture 2026, declared composition of 68 plant-based drinks (PMID 42265545)
• Food Science and Nutrition 2026, allergens and nutrition information on 196 plant-based drinks (PMID 42553086)
• Nutrients 2026, food additive mixtures, glucose metabolism and type 2 diabetes risk (PMID 42588144)