Sugar does not wind children up: it was measured in 1994

Thirty-five mothers were certain that sugar drove their children wild. In 1994 two psychologists at the University of Kentucky, Daniel Hoover and Richard Milich, brought them in and gave every child the same drink, sweetened with aspartame, which contains no sugar. The only thing that changed was the label.

Half the mothers were told their son had just taken a large dose of sugar; the other half were told he had taken the sweetener. Then they were filmed playing with them. The mothers in the sugar group rated their children as far more hyperactive, and on the video you can see them correcting more, criticising more and never leaving them alone for a second. The children were the same and the drink was the same.

A year later the medical journal JAMA published a tally of twenty-three double-blind studies: sugar affects neither the behaviour nor the performance of children, not even those diagnosed as hyperactive. The idea grew out of the diet that the allergist Benjamin Feingold made fashionable in 1975 against additives, and it survives because sugar turns up exactly where children go wild: at parties.

At the next birthday party, watch them for twenty minutes before the cake comes out. You will hear the same racket with the cake still in its box.