Effects on Sleep-Wake States on Reaction Times and Priming Effects in a Semantic Priming Paradigm

Recently, Stickgold et al. (1999) demonstrated that subjects awakened from REM sleep showed a greater priming effect by weak primes than by strong primes, which is in contrast to the normal pattern of priming. Their interpretation was that REM sleep is involved in enhancing the strength of memory association. To replicate these findings a semantic priming experiment was designed. Subjects were exposed to prime-target pairs just before sleeping, just after awakening from slow wave sleep as well as from REM sleep, and after waking up in the morning. Reaction times on the target word were dependent on the level of association of the prime-target words, and on the sleep-wake state just before testing. In agreement with the literature was the finding that the stronger the association, the shorter the reaction time on the target was. Moreover, and in agreement with Stickgold et al.` findings, was the fact that reaction times after awakening from sleep and REM sleep were longer compared to those when wakefulness was preceding the test. This is in line with the phenomenon of ‘sleep inertia’. However, not in agreement with Stickgold et al.` results was that the priming effect after REM sleep was smaller than that after slow wave sleep. This effect was even comparable to the priming effect after waking. This finding jeopardises the interpretation and conclusions with respect to the positive function of REM sleep in memory consolidation