Food Dyes and Behavior: What the Science Says

Food dyes are one of the most visually obvious food additives and one of the most contested. They add no nutritional value, exist purely for appearance, and have a complicated regulatory history. Here’s what the science actually shows about their safety.

The Dyes in Use

The FDA currently certifies nine artificial color additives for use in food in the United States. The main ones you’ll encounter are:

Red 40, Yellow 5, and Yellow 6 account for the vast majority of food dye use in the United States. They’re petroleum-derived azo dyes, meaning they contain nitrogen-nitrogen double bonds. They’re extremely stable, bright, and cheap to produce.

The Southampton Study: What It Found and What It Didn’t

In 2007, a team led by Jim McCann published a randomized double-blind crossover trial in The Lancet (McCann et al., 2007). It’s the most important study in the food dye and behavior debate, and also one of the most frequently misrepresented.

The researchers tested two different drink mixtures in 153 three-year-olds and 144 eight-to-nine-year-olds. The mixtures contained:

• Mix A: Artificial dyes including Sunset Yellow, Tartrazine, Carmoisine, and Quinoline Yellow, plus 45 mg of sodium benzoate.
• Mix B: A different combination of dyes including Allura Red and others, plus 45 mg of sodium benzoate.
• Placebo: A juice drink with no additives.

The children received each drink for six weeks in a randomized order. Parents, teachers, and trained observers rated hyperactivity using standardized scales.

The results showed a statistically significant increase in hyperactivity scores in children who received both mix A and mix B compared to placebo. The effect was seen in both age groups. The magnitude was modest. It was not equivalent to a clinical ADHD diagnosis.

Two important limitations: First, the study used a mixture of dyes AND sodium benzoate. You can’t determine which ingredient was responsible, or whether the combination was needed. Second, “statistically significant” in a study with this design doesn’t tell you how large the effect was in individual children.

The Food Standards Agency (FSA) in the UK responded by recommending voluntary reformulation. The six dyes in the study became known as the “Southampton six.” By 2010, most UK manufacturers had replaced them with natural alternatives to avoid the EU warning label requirement.

The FDA’s Response

The FDA reviewed the same evidence in 2011 and reached a different conclusion. An FDA advisory committee examined the evidence and found:

• A statistical effect existed, but the size was modest.
• The study used a mixture, not individual dyes.
• They couldn’t establish that any specific dye was causing the effect.
• The evidence wasn’t sufficient to mandate label changes or a ban.

The FDA required no changes to labeling or permitted uses. It maintained that the dyes were safe at typical dietary exposure levels.

This wasn’t a cover-up or regulatory capture. The divergence between FDA and FSA/EU actions reflects a genuine difference in how regulators weigh precautionary action when evidence is mixed. The EU tends toward precaution. The FDA tends toward requiring more established harm before mandating changes.

Red 3: A Separate Concern

Red 3 (erythrosine) has a distinct concern from the hyperactivity issue. Studies showed it caused thyroid cancer in male rats at high doses. The FDA banned Red 3 from cosmetics in 1990 based on this evidence. However, it remained permitted in food under a legal provision called the Delaney Clause exemption (which complicated the regulatory situation).

In January 2025, the FDA finally revoked authorization for Red 3 in food and ingested drugs. Manufacturers have until 2027-2028 to reformulate. This doesn’t mean Red 3 was causing cancer in humans at food doses. The rat mechanism (very high doses causing thyroid hormone suppression leading to secondary cancer) likely doesn’t apply to humans at food-use levels. But the FDA decided to remove it given the availability of alternatives.

The Natural Alternative Trend

Many brands have moved to natural colorants voluntarily, driven by consumer demand rather than regulatory requirements. Whole Foods Market stopped stocking products with artificial dyes years ago. Several major candy and snack companies have reformulated specific products.

Natural colorants used as alternatives include:

• Beet juice/extract: Deep red, pink, and purple tones. pH-sensitive (turns blue-purple in alkaline conditions).
• Annatto: Orange-yellow, from the seeds of the achiote tree. Used in cheese, butter, cereals.
• Turmeric: Yellow, from dried turmeric root. Strong flavor at high concentrations, limiting its use.
• Spirulina extract: Blue-green, from a type of cyanobacteria. Used in some candies and beverages.
• Carmine/cochineal: Red, from dried female cochineal insects. Controversial for vegans and some religious groups, but well-studied safety profile.
• Paprika extract: Orange-red, from dried peppers.

Natural colorants have real drawbacks: they’re more expensive, often less stable to heat and light, and can introduce off-flavors at higher concentrations. Food technology is actively working to improve them.

Who Should Pay Attention

For most adults and most children, artificial food dyes in typical dietary amounts aren’t a meaningful health concern. The evidence doesn’t support avoiding them for cancer prevention or general health.

The group with the most evidence-based reason to limit them is children who already show signs of attention difficulties or hyperactivity. The McCann study and subsequent meta-analyses suggest a modest effect that could compound existing attention issues. The experiment of reducing dye consumption is free, low-risk, and can be done at home. If a parent sees a difference after removing dyes for a few weeks, that’s useful information about their specific child.

For adults concerned about gut health, some preliminary evidence suggests artificial dyes may affect gut bacteria, similar to other synthetic food chemicals. See our sodium benzoate article for more on the 2007 study, where the dyes and sodium benzoate were tested together. The gut microbiome connection to artificial dyes is even less established than the hyperactivity link, but it’s an active area of research.

The market is moving toward natural colorants regardless of where regulators land. In many product categories, making an artificial-dye-free choice is becoming as easy as reading the label.