There’s a scene in The Simpsons where Homer loses a $20 note, and it flies out of the window and lands at the feet of his bored son Bart and his friend Milhouse. They spend it on incredible amounts of sugar, and go on a wild, hallucinogenic bender, like a Broadway version of a William Burroughs novel. When Bart wakes up — head pounding, eyes sunken — he’s wearing a boy scout uniform, having signed up in his delirium. His sister, Lisa, stands over him. ‘The remorse of the sugar junkie,’ she sighs.
I don’t know how to break it to you, but that scene is not 100 per cent scientifically accurate. Most of us, though, believe that, while drinking an all-syrup Squishee might not cause us to have a Fear and Loathing in Las Vegas experience, there is something in the idea of the ‘sugar rush’. It’s so much a part of our culture that there was a TV series a few years ago named after it. More importantly, most of us who are parents have been to small children’s birthday parties, and seen the damage that a gang of cake-fuelled four-year-olds can inflict on a new living room suite. But — you’ll find this hard to believe if you’ve had to battle those ravening, wild-eyed midget hordes yourself, but it appears to be true — the whole thing is a myth. Sugar does not drive children insane. It has hardly any effect on their behaviour at all.
The first time that a link between sugar and hyperactive children came to public attention was in 1978, in a study published in the journal Food and Cosmetics Toxicology. It looked at 265 children who had presented at a hospital with hyperkinesis — basically, running around too much and unable to concentrate. It found that they all had what looked like abnormally low blood-sugar levels, which, counterintuitively, can be caused by eating too much sugar, a condition known as ‘reactive hypoglycaemia’. In adults, hypoglycaemia can cause cognitive problems, including anxiety and emotional instability.
Around the same time, two cross-sectional studies (that is, studies looking at a group of people at a specific time) were carried out which found that hyperactive children who had eaten sugar tended to be more hyperactive than their unsweetened peers. Case, you might think, closed. But of course it wasn’t. Later research showed that the ‘abnormally low’ blood-sugar levels of the 1978 study were in fact within the normal range for children. And cross-sectional studies are limited: they can only show you that the kids who’ve eaten more sugar tend to be more hyperactive. It can’t show you that the sugar caused that hyperactivity. It could be that hyperactive kids hyperactively shovel sugar into their face, or there could be some entire other thing going on, some third factor causing both. (I don’t know what it could be, but the classic example of a third-factor problem is that ice-cream sales and drownings tend to go up at the same time. Not because ice cream drives us into the sea, but because on hot days, people are more likely to eat ice cream, and more likely to go swimming.)
In medical science the only way to show for definite that thing A causes thing B is the double-blind randomised controlled trial. You take a bunch of people; you divide them at random into two groups; you give one lot, your treatment group, some substance that you think will do something to them (in this case sugar); you give the other group, your control, a placebo. And you make sure no one — not you, not the patients, and not the patient’s parents — know which patient is in which group.
So someone did all that. In fact, lots of people did it, lots of times.
In 1995, the Journal of the American Medical Association (JAMA) published a meta-analysis. That’s one which takes all the studies on a certain topic and combines their data to make one mega-study, meaning that you can spot smaller effects. They looked at 23 studies published between 1983 and 1994, all of them looking at the effects of sugar on children compared with the effects of a placebo sweetener, saccharin or aspartame. The studies looked at 14 different outcomes, from academic performance to aggression to motor skills to mood. And, once you pooled all the data together, they could not find one single statistically significant effect. (Statistically significant meaning that there was a less than a one-in-20 chance of just getting that result by fluke.)
Well, you might say: ‘These eggheads with their tests and numbers don’t know my child. I would have been able to tell just by looking at him/her whether he/she’d been on the wine gums again.’ And that would be a reasonable hypothesis. Except that one of the things the studies measured was ‘parental rating’. They gave the kids the sugar or placebo, and then they asked the parents whether they seemed hyperactive. The parents were unable to do better than guessing. The same happened when teachers were asked to rate the children. If parents can’t tell whether their kids have had sugar or just a placebo, it’s probably time to admit that the sugar-rush effect is not real.
That’s not to say that there aren’t several very good reasons not to give your child too much sugar, most of which revolve around the standard ‘they’ll get fat and then get diabetes and also their teeth will fall out’ arguments, all of which are excellent and wise. But the widely held parental belief that sucrose has a similar effect on children as feeding after midnight has on gremlins simply has not stood up to inquiry.
My guess, though, is that most parents just won’t believe it. They’ve seen the swivelling eyes and the frothing mouth of one party-hat-wearing, pass-the-parcel-playing micro-lunatic too many.
You can show someone as much data as you like, but that sort of trauma overwhelms rational thought.