Effect of prolonged sedation with dexmedetomidine, midazolam, propofol, and sevoflurane on sleep homeostasis in rats.

BACKGROUND:Sleep disruption is a common occurrence during medical care and is detrimental to patient recovery. Long-term sedation in the critical care setting is a modifiable factor that affects sleep, but the impact of different sedative-hypnotics on sleep homeostasis is not clear. METHODS:We conducted a systematic comparison of the effects of prolonged sedation (8 h) with i.v. and inhalational agents on sleep homeostasis. Adult Sprague-Dawley rats (n=10) received dexmedetomidine or midazolam on separate days. Another group (n=9) received propofol or sevoflurane on separate days. A third group (n=12) received coadministration of dexmedetomidine and sevoflurane. Wakefulness (wake), slow-wave sleep (SWS), and rapid eye movement (REM) sleep were quantified during the 48-h post-sedation period, during which we also assessed wake-associated neural dynamics using two electroencephalographic measures: theta-high gamma phase-amplitude coupling and high gamma weighted phase-lag index. RESULTS:Dexmedetomidine-, midazolam-, or propofol-induced sedation increased wake and decreased SWS and REM sleep (P<0.0001) during the 48-h post-sedation period. Sevoflurane produced no change in SWS, decreased wake for 3 h, and increased REM sleep for 6 h (P<0.02) post-sedation. Coadministration of dexmedetomidine and sevoflurane induced no change in wake (P>0.05), increased SWS for 3 h, and decreased REM sleep for 9 h (P<0.02) post-sedation. Dexmedetomidine, midazolam, and coadministration of dexmedetomidine with sevoflurane reduced wake-associated phase-amplitude coupling (P≤0.01). All sedatives except sevoflurane decreased wake-associated high gamma weighted phase-lag index (P<0.01). CONCLUSIONS:In contrast to i.v. drugs, prolonged sevoflurane sedation produced minimal changes in sleep homeostasis and neural dynamics. Further studies are warranted to assess inhalational agents for long-term sedation and sleep homeostasis.