Opioid effects on computer-derived sleep and EEG parameters in nondependent human addicts.

After one adaptation night, intramuscular doses of methadone (7.5, 15, 30 mg/70 kg), morphine (10 or 20 mg/70 kg), or placebo were given to seven male nondependent opiate addicts at weekly intervals in a randomized cross-over design. After three adaptation nights, heroin (3, 6, 12 mg/70 kg) was compared with morphine and placebo by means of a similar design in seven other subjects. Using electroencephalogram (EEG) bisector analysis, tape recordings of sleep were analyzed for two beta, three alpha, three theta, and two delta EEG patterns, as well as for detections of sleep spindles, K-complexes, eye movements, body movements, average electromyogram (EMG), and calculation of seven sleep-waking stages. All three opioids produce a dose-related arousal: they increase EMG and EEG measures of muscle activity, as well as body movements and EEG alpha, while decreasing EEG theta and spindling. These opioids also increase measures of waking state and decrease measures of spindle sleep and REM sleep. Although the 1974 version of the EEG bisector analysis is not exactly comparable to visual analysis, in this design it defined significant drug effects on sleep and EEG. Distinctive bisector analysis patterns are positively correlated with each sleep--waking stage.