Artificial food colourings: do they affect children?

Yes, certain artificial food colours can nudge up activity and inattention in some children, and a landmark UK trial is the reason your child's squash may carry a warning sentence. But the honest version is narrower than the headlines suggest: the effect is real, modest, and varies a great deal from one child to the next, and it is not the cause of ADHD. The cancer worry that often gets folded into the same conversation belongs to different additives entirely, and even there the story is more about regulatory caution than proven human harm.

The Southampton study, in plain terms

The evidence behind almost everything that followed is a trial funded by the UK Food Standards Agency, run by a University of Southampton team and published in The Lancet in 2007.¹ Its design was unusually strong for a nutrition question: researchers recruited 297 children from the general community, not an ADHD clinic, in two age bands, 153 three-year-olds and 144 children aged 8 to 9. Each child went through periods of drinking a fruit drink spiked with one of two additive mixes, and a period drinking an identical-looking placebo. Neither the families nor the assessors knew which drink was which, and every child acted as their own comparison.

The active drinks contained six synthetic colours in two different combinations, plus sodium benzoate (E211), a common preservative. Behaviour was rated by parents, by teachers and by a trained independent observer, and these ratings were combined into a single "global hyperactivity aggregate." On average, children were rated as slightly more hyperactive during the additive periods than during the placebo periods. The effect was small, roughly a third of a standard deviation in the groups where it showed, and it was not seen in every child, in both age groups, or consistently across both mixes.

One detail is easy to miss: because these were ordinary children, not a diagnosed group, the finding implied colours can shift behaviour across the general population, not only in children already struggling with attention.

The "Southampton Six" and the warning label

The six colours used in the trial became known as the "Southampton Six," and you will still meet them on UK and European labels by name or E number.

Under EU Regulation 1333/2008, which applied from 20 July 2010 and was retained in UK law after Brexit, any food or drink containing one of these six colours must carry the statement that the colour "may have an adverse effect on activity and attention in children."⁵ Note what the label does not say: it names the colours, not the sodium benzoate that was also in the test drinks, though the preservative was part of the mixture that produced the effect. The warning is a prompt to choose, not a declaration that the product is unsafe.

What the regulators actually concluded

The UK and Europe did not ban these colours. Instead, in 2008 the Food Standards Agency called for a voluntary phase-out by manufacturers, ideally by the end of 2009, and pressed the EU for the mandatory warning label.⁴ Many UK makers reformulated rather than print a warning, which is why these colours are now uncommon in mainstream British products and more visible in some imported sweets.

Europe's scientific assessment was more cautious still. When EFSA examined the trial in 2008, it accepted limited evidence of a small effect in some children, but flagged real weaknesses: the additive mixtures meant no single colour could be blamed, the results were inconsistent across ages, sexes and the two mixes, and the clinical importance of the changes was unknown. On that basis it concluded the study could not be used to change the acceptable daily intake of any colour.² A year later EFSA did lower the acceptable daily intakes for three of the colours, Quinoline Yellow (E104), Sunset Yellow (E110) and Ponceau 4R (E124), but for separate toxicology reasons drawn from animal studies, not because of the behaviour findings.³

Why the US and California acted so much later

For more than a decade the United States took a different view. The FDA held that a causal link between colours and hyperactivity had not been established for children in general, while acknowledging that a susceptible subgroup might react.

The momentum shifted in California. In 2021 its Office of Environmental Health Hazard Assessment published a major review concluding that synthetic dyes can cause hyperactivity and other neurobehavioural effects in some children, and that decades-old federal safe levels may not adequately protect them.⁶ That assessment underpinned two first-in-the-nation laws: the California Food Safety Act of 2023, which banned Red 3 and three other additives statewide, and the California School Food Safety Act of 2024, which removes six dyes, Red 40, Yellow 5, Yellow 6, Blue 1, Blue 2 and Green 3, from public-school food, with makers given until the end of 2027 to reformulate.¹¹ The school measure is part of a wider unease about what children eat at scale, the same theme we explore in what school lunches do to children and in our piece on ultra-processed food.

The cancer question, honestly

The Southampton Six are a behaviour story, not a cancer story; at permitted intakes they are not considered carcinogenic. The online cancer worry actually blends two separate additives, and untangling them matters.

The first is Red 3 (erythrosine). In January 2025 the FDA revoked its authorisation for food and ingested drugs in the United States. The trigger was not new evidence of human harm but the Delaney Clause, a 1960 law that forbids any colour additive shown to cause cancer in people or animals, with no allowance for dose. High doses of Red 3 caused thyroid tumours in male rats through a hormonal mechanism, and the FDA states plainly that this mechanism does not occur in humans and that human risk at real exposures is not supported by the evidence.¹⁰ In the UK and EU, erythrosine (E127) has long been permitted only in very limited uses, mainly cocktail and candied cherries, so it was never a significant part of the British diet.

The second is titanium dioxide (E171), a brilliant white used to make products opaque rather than a colour in the usual sense. In 2021 EFSA concluded it could no longer be considered safe as a food additive, because a concern over genotoxicity, the potential to damage DNA or chromosomes, could not be ruled out. The EU acted on that uncertainty and banned it as a food additive from August 2022.⁸⁹ This is precaution under genuine scientific doubt, not a finding that the additive causes cancer. The UK's Food Standards Agency reviewed the same evidence, judged that the uncertainties did not justify a ban, and titanium dioxide remains permitted in UK food, a clear post-Brexit divergence.

The honest interpretation

Pulling the threads together, the fair conclusion is a measured one. The behavioural effect is real but modest, and it is concentrated in a minority of children. The most careful synthesis, a 2012 meta-analysis by Nigg and colleagues, found a small average effect of synthetic colours on attention and activity, with an effect size of roughly 0.2 to 0.3 across studies, larger in children chosen for known sensitivity, and estimated that around 8 per cent of children with ADHD may respond to colours.⁷ That is meaningful for those families and unremarkable for most.

Two cautions keep the picture honest. First, colours are not the cause of ADHD, which is a neurodevelopmental condition with a strong genetic basis; removing dyes will not treat it, which is why NICE does not recommend elimination diets as a general approach.¹² If attention and activity are a daily struggle, the colours question is a small side issue next to proper assessment, as we set out in our guide to ADHD diagnosis. Second, the trial mixed colours with sodium benzoate, so even its own signal cannot be credited to the colours alone. The truthful statement: a specific combination of additives slightly worsened behaviour on average, in a way some children feel and most do not.

Practical guidance for parents

Because so much UK reformulation has already happened, accidental high exposure is less likely than it was fifteen years ago. Affected products sit in the usual treat aisles: brightly coloured sweets, squashes and cordials, ice lollies, jellies and cake decorations. A calm approach looks like this.

• Notice before you act. If you suspect your child reacts, keep a simple food-and-behaviour diary for a week or two. Patterns are easy to imagine and easy to miss, and a written record beats memory.
• Run a fair test. Remove the six named colours for two to four weeks, watch carefully, then reintroduce them and see whether behaviour shifts back. A change that appears on removal and returns on reintroduction is far more convincing than a one-off.
• Read the label. Look for the colour names or E numbers in the table above and the warning sentence about activity and attention. Our guide to reading a nutrition label covers the rest of the small print.
• Get help before cutting food groups. A structured elimination diet should be done with a paediatric dietitian or paediatrician, not improvised, so your child's nutrition is protected.
• Keep perspective. An occasional coloured sweet is not a health emergency, and "natural" colourings are not automatically safer or better studied. The overall diet matters far more than any single additive.

This article is educational and does not constitute medical advice, diagnosis, or a treatment recommendation. Medication uses described as “off-label” are not licensed for that purpose in the UK and should only be considered under qualified clinical supervision. Always speak to your GP, pharmacist, or a registered specialist before starting, stopping, or changing any treatment. If you have severe or alarm symptoms - unintentional weight loss, blood in your stool, difficulty swallowing, persistent vomiting, a fever, or severe pain - seek urgent medical care.