Jan 5, 2012
Meta-study on feeding experiments
GM plants in animal feed: No long-term consequences
The use of GM plants in food and feed is not likely to cause adverse long-term effects, according to the results of a new meta-study. The authors come to the conclusion that 90-day feeding trials are generally sufficient to ensure that a GM plant is safe. Longer and more elaborate studies would not normally provide any additional findings.
Feeding maize silage to cattle
The feeding experiments carried out as part of the risk assessment of GM plants are usually 90-day trials with rodents based on OECD standards. In the past, these experiments have often been criticised for being too short and not specific enough. Critics claimed that they would not identify long-term effects.
A team of scientists led by Agnes Ricroch of AgroParisTech, the Paris Institute of Technology for Life, Food and Environmental Sciences, has now conducted a literature review to investigate whether feeding trials that lasted longer than 90 days or were conducted across several generations revealed effects that would not have been discovered in standard 90-day studies. They reviewed 24 of these long-term and multi-generational studies that have been reported in the scientific literature since 1996. Of these, 22 were conducted by public research facilities. One multi-generational study with mice that was conducted at the University of Vienna in 2008 and caused a public sensation was not included because it was never published in a scientific peer-reviewed journal.
In addition, the scientists evaluated 90-day standard feeding studies of GM plants for which there are also long-term or multi-generational studies available.
Most of the 24 feeding trials selected involved Bt maize and herbicide-tolerant soya beans, although a few studies dealt with GM potatoes, triticale and rice. GM soya was fed primarily to mice and rats, while Bt maize tended to be fed to farm animals like cows, pigs and hens. A wide range of investigations was conducted on the trial animals in the different studies. They included measuring parameters like growth, food intake and organ weight, but also biochemical analyses of body fluids such as milk and blood, and histological examinations of various organs.
The authors write that the results of the studies analysed do not suggest any health hazards connected with the consumption of GM plants. In a few cases where there were statistically significant differences between the test and control groups, these fell within the normal biological variation range and had in most cases been regarded as irrelevant by those in charge of the trials.
However, the authors criticise the fact that a large number of the studies were based on flawed experimental protocols. In only six of the 24 studies was the number of trial animals in line with the OECD guidelines. All the other studies used fewer animals. The studies also frequently omitted to feed animals with an isogenic line corresponding to the GM plant in question as a control. Another criticism is that many of the studies did not repeat the experiments enough times, if at all.
In general, the authors interpret their results as showing that long-term and multi-generational studies do not usually produce any additional findings compared with 90-day studies. However, if such trials do need to be carried out, because questions still remain unanswered after completion of the 90-day trial, the authors recommend that the experimental protocols should be harmonised. In general though, they say that the meta-study shows that consumption of GM plants is not likely to have negative long-term effects.