Here's the title to a June 2009 article in the Harvard Mental Health Letter: "Diet and Attention Deficit Hyperactivity Disorder - Can some food additives or nutrients affect symptoms? The jury is still out."
Hmmm. That sounds very noncommittal. Let's start instead with the 2007 Southampton Study. Published in the Lancet, Britain's premier medical journal, this was a well-designed nutritional study! We're talking a randomized, placebo-controlled, crossover study. That's enough to make any biological scientist's heart go pitter patter. So what did the researchers do? They took 153 3 year-olds and 144 8/9 year-olds recruited from the community, and did a baseline measure of hyperactivity on the kids via a questionnaire filled out by teachers. All the kids were put on an additive and preservative free diet. At weeks 2, 4, and 6, half the kids were randomly/but in crossover fashion given a study juice drink containing either Mix A (with tartrazine (yellow #5), sodium benzoate, sunset yellow, carmoisine, ponceau 4R) or Mix B (sodium benzoate, sunset yellow, carmoisine, quinoline yellow, and allura red AC). Mix A was supposed to be equal to the additives and preservatives found in 2 bags of candy and was similar to a mix in previous studies, and Mix B to 4 bags of candy - meant to be a representation of the average amount of additives a kid on a normal diet might receive. The other half of the kids got a placebo drink that was the same color and flavor as the test drinks, but had no artificial colors or preservatives (I'm sure they managed it somehow). On weeks 3 and 5, everyone got a placebo (these were "washout weeks."
Then the researchers asked parents and teachers to assess the children's behavior using standard clinical instruments, and also asked independent reviewers to observe the kids at school. The investigators found a significant increase in hyperactivity during the weeks the kids (both 3 year olds and 8/9 year olds) consumed the drinks containing the artificial colors. Some kids appeared to be more vulnerable to the effect than others, but the overall effect was a 10% increase in hyperactivity. It was similar to the results found by an earlier meta analysis, and, in summary, the effect of removing additives from the kids' diet is about 1/3 the effect of giving kids Ritalin to calm them down.
The analysis of the study came down to this - some kids are very sensitive to additives, and their behavior will be significantly impacted. With other kids, it won't matter much. The results of the study, however, were impressive enough that the UK and several other European countries banned the use of the studied food additives, so Skittles sold in the UK have no sunset yellow or tartrazine, though as far as I know, Skittles in the US have the same old recipe of hyperactive family fun. Sodium benzoate was not banned.
So why is the Harvard Mental Health Letter so noncommittal? Well, they cite a well known study of 35 mother-son pairs where the mothers believed the boys, ages 5 to 7, were sugar sensitive. The researchers told the Moms that the boys would be randomly assigned to a dose of high sugar or a dose of aspartame. In reality, all the boys received aspartame. The mothers who thought their sons got sugar reported their child's behavior to be significantly more hyperactive afterward. "The researchers concluded that parental expectation may color perception when it comes to food related behaviors." Really? Well, doesn't a double-blind prospective cross-over design with independent examiners CONTROL for that sort of thing in the Southampton Study? I would think so.
The medical letter goes on to talk about Omega 6 to Omega 3 fatty acids (they recommended kids consume 12 ounces of low mercury shellfish or fish a week) and micronutrients (deficiencies in zinc, iron, magnesium, and vitamin B6 have been documented in children with ADHD, but there is no evidence that supplementation is helpful, and megadoses, which can be toxic, should be avoided). Eventually, they recommend "a healthful diet" for kids and minimizing processed and fast food. Well, I can get behind that. And, frankly, I don't recommend a steady diet of skittles to anyone, though it would be nice to have the option (as the European moms do) of having candy without the crappy additives. Since hyperactivity affects 10% of children (1), what they eat can have a large effect, overall, on kids and parents alike.