REM sleep homeostasis: a matter of size?

In different species, REM sleep (REMS) occurrence appears to be finely regulated on either a short (Vivaldi E, 1994) or a long-term basis (Parmeggiani PL, 1980). However, the hypothesis that REMS amount is homeostatically regulated is challenged by data on humans which show that a scarce REMS rebound follows REMS deprivation (Horne J, 2000). Further analysis was carried out on data from an experiment in which 24 male Sprague-Dawley rats (250g) were exposed for 24h to different low ambient temperatures (Tas, ranging from –10°C to 10°C) and then allowed to recover for 4 days at normal laboratory Ta (Cerri M, 2005). REMS decreased proportionally with cold exposure, but a quick compensatory REMS rebound occurred during the first day of recovery when the previous loss went beyond an “alarm” threshold (AT) corresponding to 22% of the REMS daily need. By using data from literature we have calculated that AT for cats (Parmeggiani PL, 1980) and humans (Endo T, 1998) should correspond to 72% and 234% of the REMS daily need. Examining the three species together, AT appears to increase proportionally to the average duration of the REMS episode. It also appears to be positively related to their brain mass (Kg) according to a power function: y = 226.03x 0.36 , r2=0.986. This would suggest that, in analogy to what has been observed regarding the body’s energy needs, small mammals have a smaller capacity to buffer their REMS need than large ones.