Explainer: Why is the US eliminating synthetic food dyes?

US health secretary Robert Kennedy Jr has directed the country’s regulators to move against all synthetic food dyes, criticising food producers for feeding Americans petroleum-based ‘poisonous compounds’. Some large US food firms have begun switching towards alternatives and phasing out artificial colours voluntarily. But is this policy based on solid evidence, or politically motivated?

Why have synthetic food dyes been in the news?

The US Food and Drug Administration (FDA) said last year it wants to phase out all petroleum-based synthetic dyes from the country’s food supply, through a combination of regulation and voluntary elimination. More recently, the FDA said companies can claim products contain ‘no artificial food colors’ when there’s no petroleum-based dyes, whereas this label had previously not been permitted for foods containing naturally-derived dyes.

The FDA has begun the process of withdrawing two synthetic dyes: Citrus Red 2, approved to colour orange skins, and Orange B, approved for sausages and frankfurters, but now largely unused. These two colours are either banned or had never been approved in the EU and UK. The FDA is also working with industry to voluntarily eliminate the six remaining synthetic food colourings approved in the US over health concerns – Green 3, Red 40, Yellows 5 and 6 and Blues 1 and 2.

Why and where are these food colourings mostly used?

The dyes can give foods vibrant appealing colours associated with sweetness and ripe fruits. Dull grey foods are not visually appealing.

The dyes are in around a fifth of US food products, according to a recent US study. But disproportionately in foods such as cakes, cookies, pastries, ice cream and sugar-sweetened drinks. Sugary foods and products marketed to children are more likely to contain synthetic dyes.

Are dyes particularly problematic for children?

It can be tricky to separate the effects of synthetic dyes from other effects of the unhealthy foods that contain them. However, behavioural investigations in children – including controlled trials by Jim Stevenson and researchers at the university of Southampton, UK – present a consistent picture of a modest influence from artificial food dyes on behaviour.

Child behaviour was more unruly when ingesting artificial colourants, particularly synthetic azo dyes, with some children affected sufficiently to impact their school performance.

How are these dyes regulated elsewhere?

In 2010, the European Union required products containing the six dyes investigated in the Southampton studies to include the warning ‘may have adverse effect on activity and attention in children’ on their labels. Three of these are among the six now being voluntarily eliminated in the US, while the other three were either never approved or withdrawn from US use (see table). The UK similarly asked for a voluntary ban on their use in children’s food. Manufacturers mostly complied, avoiding the off-putting label.

Of the remaining three dyes being eliminated in the US, Green 3 (E143) is banned in the EU and UK, while the two blue dyes are approved for use in food and cosmetics. Red 3, now banned in the US, is approved in the EU for limited, specific uses.

In 2016, the European Food Safety Authority (EFSA) re-evaluated 41 food colours approved under older regulations. It withdrew a red dye (E128) because it is broken down into carcinogenic aniline in the body. The EFSA also lowered the acceptable dietary intakes of three other food dyes – E104 (quinoline yellow), E110 (sunset yellow, Yellow 6); and E124, a strawberry-red azo dye.

Have US policy makers taken action before?

Some US states have implemented their own bans on various colourings. In 2023, California banned certain food additives – including Red 3 (erythrosine), which gives food a bright, cherry-red colour. West Virginia, Virginia and other states more recently restricted synthetic dyes in foodstuffs. In January 2025, the FDA banned Red 3 because it could cause cancer in lab animals, although evidence for risk to humans is limited.

What’s the evidence for negative health effects?

While there are suspicions that synthetic dyes could encourage inflammation and heighten cancer risks, because the effects are relatively small, definitive evidence would require impractically large human trials. Most safety evidence is therefore inferred from rodent studies. The EFSA, for example, concluded in 2009 that DNA changes observed in mice fed tartrazine are not expected to result in cancer in humans.

While there are many differences between rodent and human metabolism, such studies are not entirely irrelevant. The FDA’s ban on Red 3 was triggered by studies linking cancer to feeding rats high levels. This triggered the Delaney clause, which says that if a food additive causes cancer in animals or people, its authorisation must be revoked. The FDA acknowledged that the levels of Red 3 people are exposed to are typically much lower than the rats ingested. An extensive evaluation of food additives by the World Health Organization in 2020 had concluded that there were no concerns over gene, reproductive or developmental toxicity from erythrosine (Red 3).

Have other dyes been tested in animals?

Some researchers, such as Lorne Hofseth at the University of South Carolina, US, aim to investigate food dyes in lab animals and – where relevant – trigger the Delaney clause to get them withdrawn. Hofseth warned about Allura red (Red 40) in a 2024 publication, citing a Canadian study linking it to colitis in mice; a US study of immune changes and colitis in mice; and his own findings of DNA damage, gut microbiome changes and intestinal inflammation in mice fed a high-fat diet.

The Canadian study led to questions in the European Parliament about Allura red, but the European Commission decided against restriction, given the available evidence.

There is a scattering of other published studies cited by critics, such as a 2021 report on rats drinking tartrazine-laced water developing breast tumours. However, the evidence is not strong enough to convince most health authorities. Health Canada has concluded that tartrazine can cause allergic reactions or sensitivities in some people, but that its levels in food do not present a significant health concern.

Why choose synthetic dyes?

Synthetic dye compounds are often intensely coloured due to their extended conjugated systems, which allows absorption of light in the visible spectrum. This allows them to be used in small amounts to produce strong colours. They are also generally stable under various food processing conditions, water soluble (or have variants soluble in spirits or fats, for example), and cheaper than naturally-derived alternatives.

However, azo dyes can also undergo reductive cleavage to form aromatic amines, either as impurities or breakdown products, which can be carcinogenic or genotoxic.

What does the food industry say?

The International Association of Color Manufacturers has said that colour additives are among the most widely studied food ingredients and fall under strict global regulatory requirements. It said colours are essential for consistency, visual appeal, and consumer trust in food products and have been rigorously reviewed by global health authorities with no safety concerns.

Other trade groups provide similar arguments. Industry has also published its own studies on dyes, in response to regulatory requests, such as for tartrazine (Yellow 5/E102).

Are there other options open to the food and beverage industry?

There are various naturally derived colourings available across global markets. The FDA last year approved three more food colourings from natural sources – a red algal extract, a butterfly pea flower extract for blue shades and calcium phosphate, a white dye suitable for chicken products and pastries. This year the FDA added beetroot red to the list of colour additives that are exempt from certification.

With contributions from Jim Stevenson, University of Southampton, UK; Lorne Hofseth, University of South Carolina, US; Alyson Mitchell, University of California, Davis, US; and Elizabeth Dunford, George Institute for Global Health and University of New South Wales in Sydney, Australia.