Propofol and midazolam inhibit conscious memory processes very soon after encoding: An event-related potential study of familiarity and recollection in volunteers

Abstract

BACKGROUND: Intravenous drugs active via γ3-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 μg/ml propofol (n ≤ 10), 20 and 40 ng/ml midazolam (n ≤ 12), 1.5 and 3 μg/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. © 2009, the American Society of Anesthesiologists, Inc.