Nighttime versus daytime hypnotic self-administration.

RATIONALE AND OBJECTIVES: Previous studies have shown that insomniacs self-administer hypnotics at high nightly rates. This study assessed whether insomniacs' self-administration of hypnotics extended to the daytime. METHODS: Forty-four healthy men and women, 21-55 years old, with ( n=22) and without ( n=22) insomnia volunteered. They were randomized to one of two triazolam dose groups (0.125 or 0.25 mg) and their preference for placebo versus triazolam was assessed at night (2300 hours) and day (0900 hours) over 7 consecutive days in each phase. In both night and day phases of the study, subjects received triazolam or placebo in color-coded capsules on two sampling days or nights and then were forced to choose their preferred capsule on 5 subsequent days or nights. The order of day and night study phases and the placebo and triazolam sampling days was counterbalanced. In the night phase subjects went to bed from 2330 to 0730 hours and in the day phase they were tested for level of sleepiness-alertness at 1000, 1200, 1400, and 1600 hours by the multiple sleep latency test (MSLT) and mood and performance at 1100 and 1500 hours. RESULTS: More triazolam was chosen at night than during the day. No dose differences in preferences at night versus day or between insomniacs and normals were found. Insomniacs did not differ in their triazolam preferences between night and day, while the normals chose triazolam less frequently during the day. Among insomniacs, 40% chose triazolam on >3 of the 5 days. On both screening and placebo sampling days, those with a high (>60%) daytime triazolam preference had greater average daily sleep latencies on the MSLT than those with a low (<50%) daytime triazolam preference (i.e. with a placebo preference). In the triazolam preference group, triazolam reduced daily MSLT latencies to the level of the placebo preference group. CONCLUSIONS: This study shows that the minority of insomniacs who self-administer hypnotics during the day are physiologically aroused and the drug reduces their arousal suggesting that their daytime self-administration, like their night-time self-administration, is more consistent with therapy-seeking than drug-seeking behavior, at least for the short-term.

Disturbed sleep predicts hypnotic self-administration.

OBJECTIVES: To evaluative whether polysomnographically determined sleep variables in a large group of subjects reflecting a wide range of sleep disturbance would be predictive of the self-administration of capsules before sleep. METHODS: Sixty-four healthy men and women with and without insomnia (aged 21-55 years) were given an opportunity to self-administer placebo or triazolam (0.25 mg) capsules (single-choice method - available capsule or no capsule) before sleep in three separate studies. All qualified using the identical criteria based on a standard nocturnal polysomnogram. Screening sleep measures then were used to predict subsequent placebo and triazolam self-administration. RESULTS: The percent of placebo and triazolam choices did not differ between or within the three studies. Persons with persistent psychophysiologic insomnia self-administered more capsules than persons with sleep state misperception or normals, with the subject groups not differing in placebo vs. active drug preference. Screening polysomnographic measures predicted percent of capsule choices. The single best predictor was the ratio of minutes of stage 3-4 sleep to minutes of wake plus stage 1 sleep with R=0.44. The addition of % stage 3-4 sleep, wake before sleep and total sleep time increased R to 0.49. On morning mood ratings less ability to concentrate and greater fatigue (Profile of Mood States) predicted percent of capsule choices with R=0.36. CONCLUSIONS: These results show that the extent of sleep disturbance predicts the likelihood of self-administering a capsule before sleep regardless of whether it is placebo or active drug.