We’ve seen these parents in our practice: those that say their children get “hyper” from eating a red Popsicle or that the orange mac and cheese makes their ADHD somehow worse. But what’s the evidence that food dyes are actually related to behavior?

The safety of artificial food colors (AFCs) has been questioned since as far back as 1973, when Benjamin Feingold proposed that they were making kids hyperactive, along with a host of other side effects (see Feingold B. Why Your Child is Hyperactive. Random House;1985). But controlled studies have not all cleanly come to the same conclusion.

Artificial food colors, often called “food dyes,” are ubiquitous and can be found everywhere from cheese to cola to the peel of an orange, and especially in foods targeted to children. Variations of these additives have been used for hundreds of years, beginning with natural food colors and more recently artificial dyes. They have been on the radar of government agencies that regulate food and drugs in the US and Europe since the early twentieth century.

In the US, these dyes are regulated by the Food and Drug Administration (FDA), which has set acceptable daily limits and performs batch tests for safety of the nine “certified colors” that are synthetically created, although the same is not true for the 29 food dyes that are naturally derived. According to FDA statistics, the per capita use of AFCs has quadrupled in the last 50 years.

In 2011, the FDA convened an advisory committee to hear testimony on the evidence for AFCs’ relationship to ADHD. They ultimately decided (by an 8 to 6 vote) not to recommend banning AFCs or requiring a warning label. Here’s a breakdown of the evidence and recommendations for how you can approach questions about food dyes in your practice.

What the Research Shows

Food dyes have been examined in many studies dating back to the 1970s. Many studies have focused on comparing kids on a regular diet with kids on a specialized dye-free diet (such as the Kaiser Permanente diet, developed by the aforementioned Dr Feingold) to see how these diets effect measures of hyperactivity.

In 2004, a meta-analysis of 15 double-blind, placebo-controlled studies found a relatively small effect size of 0.28 between AFCs and placebo on ADHD symptoms and in parent rating measures (Schab DW & Trinh NH, J Dev Behav Pediatr 2004;25:423-434). (As a reminder, an effect size of 0.2 is considered small, 0.5 medium, and 0.8 large.)

Various studies of elimination diets, where most foods are eliminated before being slowly re-introduced in order to observe effects, have found that introduction of AFCs has negative results on behavior of both children with ADHD and those without (see for example Egger J et al, Lancet 1985;1:540–545). A 2012 review found that there was a >70% response rate to elimination diets and that most parents reported more hyperactivity when challenged with AFCs (Stevens L et al, Clin Pediatr 2012;50:279–293).

Two additional meta-analyses in 2012 (Nigg JT et al, JAACAP 2012;51:86–97) and 2010 (Stevenson J et al, Amer J Psychiatry 2010;167:108–1115) found small but significant effects “too substantial to ignore” (Nigg et al, opt.cit), but not conclusive.

Unfortunately, most studies analyzing the effects of AFCs are limited in some way. Many of these studies lump artificial food dyes together with each other or with preservatives, making it impossible to distinguish which colors may be more harmful. There is not a standard “dose” or duration of exposure to AFCs in the studies, further clouding the results.

Two of the better studies came from the Southampton group, which had excellent blinding for the challenges, but also had Na benzoate, a preservative, in the dye mix. They found a small but significant deleterious effect in a community sample of children not selected for having ADHD. The effect was noted across diagnostic groups, whether the children had ADHD, allergies, or neither. Importantly, the effect was linked to a common genotype that 60% of the children had (Bateman B et al, Ach Dis Child 2004;89:506–511; McCann D et al, Lancet 2007;370:1560–1567).

These results suggest that food dyes are a widespread public health problem, rather than just an ADHD problem. It may also shed some light on the heritability of ADHD symptoms: what is inherited may well be genetic vulnerabilities to environmental stresses, including chemicals in food.

Did the FDA Get it Right?

After the Southampton studies were released, there were some dramatic changes to the governance of food dyes in Europe. The government of the UK requested that food manufacturers remove synthetic dyes from food in favor of natural ones, and the EU asked manufacturers to either remove the dyes from foods or include a warning label stating that AFCs “may have an adverse effect on activity and attention in children.”

In the US, the Center for Science in the Public Interest led a charge to have the FDA look at this issue. During the food advisory committee’s two-day hearing, they reviewed available data and weighed a number of questions related to food dyes, ranging from whether further research was needed (93% voted yes), to whether foods containing dyes should carry a warning label, as is done in the EU (57% voted no), ultimately upholding a decision initially made in 2010.

Criticism of the FDA’s decision abounds. Among the flaws, critics have pointed out that the FDA focused on food dyes’ relationship to the clinical diagnosis of ADHD versus behavior problems and non-ADHD symptoms such as insomnia (Weiss B, Environ Health Perspec 2012;120:1–5).

But the research thus far has not been strong enough for the FDA to make any firm conclusions. One problem is that the dyes are already on the market and being used and, when that is the case, the burden of proof is to show harm.

If they just wanted to start putting AFCs in food, the burden of proof would be to show safety. It seems likely, in view of the 93% vote that more research is needed, that if the dyes were being newly proposed, the FDA would require more research before approving them. Randomized controlled trials with standard doses and durations of each specific AFC could paint a clearer picture of how they truly effect behavior.

CCPR’s Verdict: If only it were so easy to simply remove a food additive or avoid it altogether to make ADHD better or prevent the illness in the first place. However, ADHD is more complex and overdetermined than that. Many causes, both genetic and environmental, contribute, and the causes are not the same in all cases. Nevertheless, it seems clear that food dyes have a negative effect on behavior, whether in the context of a psychiatric diagnosis or not. We see no harm and perhaps some benefit from removing artificial food dyes from a child’s diet. And that goes for all kids—not just those with ADHD.