Britain's Royal Society announced this week that it has found no convincing evidence of adverse health effects from eating genetically modified (GM) potatoes in unpublished data from research at the Rowett Research Institute in Scotland.

The work was carried out by Arpad Pusztai, a senior researcher at the institute, who claimed to have detected a potential impact of such food on the immune system. This caused a nationwide controversy last year when it was described on a television programme (see Nature 394, 714; 1998 & 397, 547; 1999).

But a scientific panel set up by the society says that, on the basis of the information available to it, the work appears to be “flawed in many aspects of design, execution and analysis”. The panel adds that no conclusions should be drawn from it.

Pusztai told the panel that further data exists beyond that which was included in the review of his work. But he failed to produce such data or other evidence.

The panel's conclusions were published in a report on the Pusztai data and on toxicity in GM foods. The panel says that the slight differences that Pusztai claimed to have detected between rats fed predominantly on GM and non-GM potatoes could not be interpreted, because of the technical limitations of the experiment and incorrect use of statistical tests.

The panel also points out that, even if the experiment had been done skilfully, it would not be justifiable to draw conclusions about whether GM foods generally are harmful to human beings. “Each GM food must be assessed individually,” says the report.

Problems with the Pusztai data include a relative lack of information on how the GM and control diets differed in detailed composition. The GM potatoes used contained almost 20 per cent less protein than unmodified potatoes, and rats in the long-term feeding study were given additional protein to avoid starvation. The Royal Sociaty panel suggests that the observed effects could have been caused by the supplementary diet being inadequate or incomplete. Its report says that Pusztai's work attempted to cover too much ground with the information available.

The society says its review of internal Rowett institute documents was “entirely appropriate” given that these are now in the public domain.