The so-called Southampton study (McCann et al, 2007) was funded by the UK's Food Standards Agency. It investigated the effects of two mixtures of commonly used food colours and sodium benzoate on the behaviour of two childrens' age groups. EFSA's panel on Food Additives, Flavourings, Processing Aids and Materials in Contact with Food (AFC) today said that the study provided "limited evidence that the mixtures of additives tested had a small effect on the activity and attention of some children". But it noted that the effects observed were not consistent for the two age groups and the two mixtures. When it was published in The Lancet last September, the conclusions - that artificial food colours and additives were seen to exacerbate hyperactive behaviour in children at least up to middle childhood - caused widespread concern amongst consumer groups and the media. As well as loud criticism directed at the FSA for the mild tone of its consumer advice, voices have been raised from some quarters calling for the additives to be banned outright. Such a ban would, however, have to come from the European Commission. As the food risk assessor for the bloc, it is EFSA's job to advise on scientific matters. And if EFSA sees no scientific grounds for a regulatory change based on the available evidence, no changes are likely to be forthcoming. Today's opinion from EFSA may cause dismay in the anti-additive lobby, but the book is far from closed on the safety of food additives. EFSA is currently conducting a review of all food colourings, including those used in the mixtures, and it expects to issued opinions on these by the end of the year. The study, which was published in The Lancet last year, was conducted in two phases. In stage one, 153 three-year olds and 144 eight- and nine-year olds were given one of two drink mixes containing artificial food colours and additives, or a placebo. The children were drawn from general population and across a range of hyperactivity and ADHD (attention deficit hyperactivity disorder) severities. Mix A contained sunset yellow (E110), tartrazine (E102), carmoisine (E122), ponceau 4R (E124) and sodium benzoate (E211). This same mix was used in an earlier study on a cohort of three-year-olds which was deemed inconclusive because the effects were not confirmed by clinicians. Mix B contained sunset yellow (E110), quinoline yellow (E104), carmoisine (E122), allura red (E129) and sodium benzoate (E211). Phase one lasted six weeks, and every child consuming the mixes and the placebo for one week each, with a one week wash-out period between each. Parents were asked to keep other sources of artificial out of the diet, and to keep a diary of violations. Phase two involved some of the children from the older group - responders and non-responders - during two half-day session a week apart, at which they were given either a placebo or an active drink similar to mix A or B, but the whole day's dose was given at once. Despite having received some addition information from the researchers that was not included in the write-up published in The Lancet, EFSA concluded that there is no basis for changing the ADIs because there were considerable uncertainties to the study. In particular, it said there was inconsistency with respect to age and gender of the children, the effects of the two mixtures of additives tested and the type of observer (parent, teacher, independent assessor). The clinical relevance of the effects measured was unknown, and there was a lack of information on dose-response. The relevance of the small effect size was also unknown. The fact that the study used mixtures of additives meant it was not possible to identify the effects of individual additives. Nor was any plausible biological mechanism given that might explain the possible link between the consumption of colours and behaviour. EFSA was also asked by the European Commission to look at other scientific studies in the same area of study. After searching the literature going back to the 1970s, the panel concluded that the Southampton study was the largest conducted to date on a suggested link between food additives and hyperactivity in the general population. It said that the majority of previous studies on the subject described the children involved as hyperactive, and therefore could not be seen as representative of the general population.