Dubreuil D, Jay T, Edeline J-M (2002):
In this study rats were exposed to 900 MHz electromagnetic fields, pulsed at 217 Hz. However, whereas in similar studies whole-body exposure was used, in this experiment only the rats' heads were exposed.
Forty-eight rats were used for each of two experiments. They were randomly assigned to one of four groups:
The rats adapted to restrained conditions for 7 days, and in the last 3 days of this adaptation period they were habituated to a radial-arm maze, which had a food reward at the end of each of the eight arms. In this elimination task, the animals had to learn to visit each arm only once. The rats' performance was evaluated by the number of errors and the rank of the first error (the number of consecutive good responses before the first error occurred). For all four groups the number of errors decreased significantly over the ten training sessions, and the rank of the first error increased. However, there was no difference between the groups in their performance.
In the other task, the navigation task, the rats had to learn to find and open a covered box containing a food reward. This was one of 13 boxes. It was always located in the same position in regards to the reference space formed by geometrical cues. After a pre-training period, the rats had four sessions on each of 14 training days. For these four daily trials, the rats were started from different locations - north, south, east, and west. From day to day the order of the starting positions was randomly changed. Evaluation of performance was based on the number of direct hits (opening the correct box first), the number of errors, and the time spent before opening the goal box. As in the first task, the rats' performance improved with time, but there was no difference between the groups.
The authors review factors that might account for the different results they obtained compared with similar studies of cognitive function in animals. In particular, they discuss the study of Lai (1994) in which rats were tested in a radial-arm maze, and suggest that the learning curves of Lai's study seem to indicate a performance effect rather than a learning deficit per se. In other words, factors unrelated to learning, such as changes in the animals' motivation or in the overall locomotor activity, contributed to produce a performance deficit.