Cassel JC, Cosquer B, Galani R, Kuster N (2004)
In 1994 Lai et al. published a paper that had a significant impact on the debate about whether radiofrequency radiation (RFR) in the cell phone range had the potential for adverse health effects. They reported that, after 45 minutes exposure to 2450 MHz microwaves, rats showed retarded learning while performing in a radial-arm maze to obtain food rewards. They claimed that this showed a deficit in spatial "working memory" function. Cobb et al. (2004) carried out a replication study and failed to confirm the findings of Lai et al. Now Cassel and colleagues have also performed a study to attempt to replicate the original study.
The strain, sex, and weight of the rats, the radial-maze, the microwave characteristics, and the exposure duration were the same as the two earlier studies. Like Cobb et al., Cassel used food pellets instead of the lab chow that had been used by Lai. Chow is crumbly and could lead to errors if residue was left in an arm of the maze.
Three groups of rats were used, with 12 in each group. Naive rats had no RFR exposure and no contact with the exposure system. Sham-exposed rats were placed in the exposure system but were not exposed to microwaves. The exposed group had 45 minute exposure to 2450 MHz microwaves (2µ pulse width, 500 pps, and whole-body SAR 0.6 W/kg). Prior to exposure the rats had been habituated to lab conditions for 10 days. They were placed on restricted food intake for another 10 days, the last 5 days of which were training days in the maze. The maze had 12 arms. During each of the 10 days of the actual experiments the exposed rats were exposed to the microwaves for 45 minutes before being tested. The rats were expected to enter each of the 12 arms to retrieve a food pellet placed near to the end of the arm. The errors recorded were of 3 types: The number of complete re-entries into arms that had already been completely visited; the number of arms visited before the first error was committed; and the number of errors calculated according to a method devised by Lai (i.e. the number of re-entries into already visited arms out of the first 12 visits).
There were no significant differences in the performance of exposed rats compared to the sham or naive animals. Cassel et al. postulate that the results of Lai and colleagues might have more to do with factors liable to performance bias than with spatial working memory per se.