Yellow 5 and children’s behaviour: what the studies actually tested

Tartrazine, also called Yellow 5 or E102, has been studied both on its own and in mixtures. A small 1994 trial found parent-rated behavioural changes in some children selected largely for suspected sensitivity.^[1] The widely cited 2007 Southampton trial tested colour mixtures with sodium benzoate, so it cannot isolate tartrazine’s contribution.^[2] A 2013 Hong Kong trial found no significant adverse behaviour difference with its colour mixture or a separate benzoate capsule.^[3] These studies concern behaviour scores, not proof that tartrazine causes an ADHD diagnosis. Their different designs and limitations matter.

What regulators concluded

Tartrazine is one of the most re-reviewed dyes on the market, and none of those reviews ended in a ban.

• EFSA re-evaluated tartrazine in 2009.
• JECFA, the joint WHO and FAO additive committee, re-evaluated it again in 2016.
• Both concluded it remains safe within its current acceptable daily intake.^[4]

That is a live regulatory approval, renewed twice by two different bodies. A body doesn’t set an acceptable daily intake for something it has decided to ban.

The 2007 trial behind the hyperactivity claim

Most “Yellow 5 causes hyperactivity” claims cite a 2007 randomized, double-blind, placebo-controlled crossover trial by McCann and colleagues, published in The Lancet.^[2] It enrolled 153 three-year-olds and 144 eight-to-nine-year-olds and gave them a challenge drink containing sodium benzoate plus one of two artificial food color mixes, compared against a placebo drink.^[2]

The trial never tested tartrazine on its own. Every dose the children drank combined multiple dyes with sodium benzoate, so the measured effect belongs to that combination as a whole, and the trial offers no way to isolate Yellow 5’s own contribution.^[2]

Within that limit, the trial did find a small but statistically significant increase in a combined teacher-and-parent hyperactivity rating in children given the mixtures compared with placebo, with effect sizes from 0.12 to 0.32.^[2]

Those are small effect sizes on a combined rating scale, and a rating-scale shift is a research signal rather than a clinical diagnosis. They came from a mixture that always included sodium benzoate. The same preservative shows up in its own additive record for exactly this reason, and our breakdown of what sodium benzoate reacts with covers the other half of that mixture in more depth.

Trials that tested tartrazine on its own, and a null result

In Rowe and Rowe’s 1994 blinded challenge study, 54 children received tartrazine or placebo at different doses. Most were selected because sensitivity was suspected or uncertain. Parents reported irritability, restlessness and sleep changes in some participants.^[1] That selected group cannot tell us the proportion of all children who would react, and the study does not establish a personal safe dose or lifelong harm.

A 2013 randomised, double-blind, placebo-controlled trial in 130 Hong Kong children aged 8 to 9 gave a capsule of artificial colours, a separate capsule of sodium benzoate, and a placebo. Neither capsule had a significant adverse effect on two behaviour scores compared with placebo.^[3] Like the 2007 trial, its colour capsule was a mixture, so it does not isolate tartrazine either.

What the FDA’s committee decided

The 2007 trial’s size and reach drew the FDA’s attention. It led the agency to review the evidence, hold a public hearing, and consult its own Food Advisory Committee.^[5] After that review, the FDA concluded the evidence did not warrant requiring a new warning or further agency action, effectively finding that a general-population causal link between synthetic dyes and hyperactivity had not been established.^[5]

The core fact people skip: the same evidence that fuels the “Yellow 5 causes hyperactivity” claim online is the evidence the FDA’s own committee reviewed and decided didn’t clear the bar for new restrictions.

The reaction that’s well documented

Tartrazine’s older, better characterized issue is a hypersensitivity reaction, and it shows up in a specific group of people rather than the population at large.

• Tartrazine sensitivity most often shows up as hives or, less often, asthma symptoms.^[6]
• Cross-sensitivity is common in people who are already sensitive to aspirin or other NSAIDs.^[6]
• Reported reactions to food colorants generally are rare across the population, but noticeably more common in people with an atopic history.^[7]

2 to 7% reported reaction rate to food colorants in atopic individuals, versus roughly 0.01 to 0.23% in the general population PMID 28407952

If you already know you react to aspirin, that’s the population this additive is worth watching for. Everyone else eating something yellow has little reason for concern.

What the evidence doesn’t establish

A 2012 meta-analysis pooling synthetic food-color trials found small effect sizes on parent and teacher hyperactivity ratings, somewhere around 0.12 to 0.21 for color additives specifically.^[8] That’s a shift in a symptom score, not evidence that any dye causes an ADHD diagnosis.^[8]

It also isn’t a settled dose picture. Current acceptable daily intakes for synthetic dyes rest on older studies that were never designed to catch behavioral effects in the first place, and a review of the dose-response data available found that, for the dyes with enough information to check, the ADI might not be low enough to protect every susceptible child.^[9] That is a real gap worth closing with better research. It stops well short of “no dose is safe,” and it stops just as far short of turning a lab hyperactivity score into a clinical verdict on a candy wrapper.

What’s established and what isn’t: a small shift in a combined hyperactivity rating from a color-and-preservative mixture is established. A causal link to ADHD, or to Yellow 5 in isolation, is not.

What this means for what’s in your cart

Two groups have an actual, evidence-backed reason to check for tartrazine on a label:

• Parents of a child already diagnosed with ADHD or a known dye reaction. The trial evidence, thin as it is, concentrates its clearest signal here, and a label-scanning app that flags tartrazine by name makes that check faster than reading fine print in a store aisle.
• Anyone with aspirin or salicylate sensitivity. Treat tartrazine like your other known triggers and watch for hives or breathing symptoms, the same way you would with any other cross-reactive additive.

The pattern isn’t unique to this dye. Sulfites in wine follow the same shape: not banned anywhere, labeled almost everywhere, and genuinely risky to a specific sensitive subgroup rather than to drinkers generally. Our roundup of additives with real trial evidence behind the concern walks through which ones clear that bar and which don’t.

Bottom line: The evidence is broader than one mixture trial, but it does not give a simple prediction for every child. Some studies report behavioural changes; others do not find a significant adverse difference. A mixture result is not an individual-dye result, and a symptom score is not an ADHD diagnosis.

If a particular food seems consistently linked to a child’s symptoms, discuss the pattern with their clinician. NICE supports a food diary and, when indicated, a dietitian referral rather than removing artificial colours and additives as a general ADHD treatment.

This article summarises publicly available regulatory assessments and peer-reviewed research about food ingredients. It is general information and is not intended to diagnose, treat, cure, or prevent any disease, and it is not personalised dietary or medical advice. Regulatory positions change; the EatMatrix app reflects the current assessments in our database. If you have an allergy, a medical condition, or are pregnant, talk to a qualified clinician or dietitian before changing what you eat.