Propofol and Midazolam Inhibit Conscious Memory Processes Very Soon after Encoding: an Event-related Potential Study of Familiarity and Recollection in Volunteers

Robert A. Veselis,Kane O. Pryor,Ruth A. Reinsel,Yuelin Li,Meghana Mehta,Ray Johnson
DOI: https://doi.org/10.1097/aln.0b013e3181942ef0
IF: 8.986
2009-01-01
Anesthesiology
Abstract:Background Intravenous drugs active via gamma-aminobutyric acid receptors to produce memory impairment during conscious sedation. Memory function was assessed using event-related potentials (ERPs) while drug was present. Methods The continuous recognition task measured recognition of photographs from working (6 s) and long-term (27 s) memory while ERPs were recorded from Cz (familiarity recognition) and Pz electrodes (recollection recognition). Volunteer participants received sequential doses of one of placebo (n = 11), 0.45 and 0.9 microg/ml propofol (n = 10), 20 and 40 ng/ml midazolam (n = 12), 1.5 and 3 microg/ml thiopental (n = 11), or 0.25 and 0.4 ng/ml dexmedetomidine (n = 11). End-of-day yes/no recognition 225 min after the end of drug infusion tested memory retention of pictures encoded on the continuous recognition tasks. Results Active drugs increased reaction times and impaired memory on the continuous recognition task equally, except for a greater effect of midazolam (P < 0.04). Forgetting from continuous recognition tasks to end of day was similar for all drugs (P = 0.40), greater than placebo (P < 0.001). Propofol and midazolam decreased the area between first presentation (new) and recognized (old, 27 s later) ERP waveforms from long-term memory for familiarity (P = 0.03) and possibly for recollection processes (P = 0.12). Propofol shifted ERP amplitudes to smaller voltages (P < 0.002). Dexmedetomidine may have impaired familiarity more than recollection processes (P = 0.10). Thiopental had no effect on ERPs. Conclusion Propofol and midazolam impaired recognition ERPs from long-term memory but not working memory. ERP measures of memory revealed different pathways to end-of-day memory loss as early as 27 s after encoding.
What problem does this paper attempt to address?