Food dyes and ADHD have been studied for decades, and the short answer is: possibly a little, for some children. Trials found a small worsening in behavior after artificial food colors, and the effect wasn't limited to children with ADHD. The authors of one review don't consider dyes a main cause of ADHD.
What did studies of food dyes and ADHD find?
A broad summary comes from a review written as updated testimony to a US food safety advisory committee (Arnold, 2012). It pulled together more than 35 years of research on artificial food colors and children's behavior.
Its reading of the recent evidence was that artificial colors seem to have a small harmful effect on children's behavior, one unlikely to be down to chance. The important twist is who showed it. Some of the trials, run in Southampton, studied preschoolers and 8- and 9-year-olds who weren't a clinical group, not children picked because they had ADHD.
That's in line with where the authors land. They say artificial colors aren't a main cause of ADHD, though they may contribute in some cases. And they suggest dyes may be more a general public health question than an ADHD question.
The size of the effect matters too. In the recent trials the review covers, the effect was small, and the review says so directly of the Southampton trial in 8- and 9-year-olds. One pooled analysis found it in parents' ratings but not in teachers' or observers' ratings, and only in children picked out because they had already responded to a diet. Within that group, it showed up whether or not a child was hyperactive to begin with.
The review's own bottom line is careful. It calls the evidence inconclusive, "but too substantial to dismiss." So dyes are a fair thing to wonder about. They aren't a proven explanation for any one child's behavior, and the review doesn't treat them as one.
How big is the effect in children with ADHD?
A 2013 analysis looked only at young people with ADHD. Sonuga-Barke (2013) pooled randomized trials of several non-medical approaches in children and teens aged 3 to 18 who either had an ADHD diagnosis or scored above a validated cutoff for symptoms. Trials that compared children's behavior with and without artificial colors in their food were one of the groups.
The analysis asked a sharp question. Ratings from the people closest to the treatment were the least blinded. So the authors also checked ratings made by people who probably didn't know which condition the child was in. Many approaches looked promising on the less blinded ratings and then faded on the blinded ones. The artificial color trials were one of only two groups where the effect still held up, and on those blinded ratings it was between small and moderate.
But the limits change what these results can show. The food dye effects often came from children who'd been selected because they seemed sensitive to food, so they may not describe a typical child. When the analysis narrowed to four trials where few or no children were taking medication, the effect was no longer clear. The authors also say their analysis had little ability to spot publication bias, where trials that find nothing go unpublished, so it can't be ruled out.
So in children with ADHD, there may be a real effect, but it's modest, and it's clearest in those who were already thought to be sensitive to food. An earlier review of the dye research calls the evidence inconclusive, though too substantial to dismiss (Arnold, 2012).
Why the evidence is still unsettled
If the research has run for decades, why isn't there a firm answer? The 2012 review lists the reasons plainly (Arnold, 2012). The trials were flawed in different ways. Diagnosis wasn't standardized. How children were chosen was sometimes questionable. Blinding was imperfect or missing. Doses, how long children were exposed, and when they were tested afterward all varied.
There's a bigger problem too. Dyes were tested as mixtures, often combined with a preservative, rather than one color at a time. That makes it hard to say which ingredient did what. One pooled analysis in the review, covering studies in children, found the effect held in studies of colors without preservatives but wasn't reliable in studies that used only colors approved in the US.
What about bigger changes to what a child eats? That's a separate topic, covered in Food and ADHD: does what your child eats make a difference?, with more in the nutrition section. The strict diet one trial used, in children aged 4 to 8 with ADHD, was closely supervised by experts (Pelsser, 2011). A diet like that is a clinician's decision, never something to start on your own at home.
Try this today: write down the one question about food dyes you'd most like answered, and bring it to your child's doctor.
The flaws in the trials don't mean the effect is imaginary. They mean the size, the ingredients responsible and the children most affected are still open questions. An unanswered question isn't a settled no, and it isn't a settled yes either.
Only a clinician who knows your child can diagnose or advise on your child's situation, so talk to one about any question you have.
The numbers Stats and what the terms mean
The statistics behind the plain-language claims above.
- Schab and Trinh meta-analysis (cited in Arnold, 2012): parent ratings of ADHD symptoms on artificial food colors compared with a placebo, in children preselected because they had already responded to a diet, effect size 0.28.
- Nigg et al. meta-analysis (cited in Arnold, 2012): parent reports across 24 double-blind, placebo-controlled crossover studies of artificial food colors, effect size 0.18, reduced to 0.12 after adjustment for potential publication bias.
- Sonuga-Barke, 2013: ratings by probably blinded assessors in trials comparing behavior with and without artificial food colors, standardized mean difference=0.42.
What the terms mean:
- Effect size is a way of putting a difference on a common scale. Roughly, 0.2 is small, 0.5 is moderate and 0.8 is large.
- Standardized mean difference is one kind of effect size, read on the same rough scale.