Zolpidem in the Treatment of Transient Insomnia: A Double-Blind, Randomized Comparison With Placebo

Transcription

1 Sleep, 18(4): American Sleep Disorders Association and Sleep Research Society Zolpidem in the Treatment of Transient Insomnia: A Double-Blind, Randomized Comparison With Placebo *Thomas Roth, *Timothy Roehrs and tgerald Vogel *Sleep Disorders and Research Center, Henry Ford Hospital, Detroit, Michigan, U.S.A.; and tsleep Laboratory, Georgia Mental Health Institute, Atlanta, Georgia, U.S.A. Summary: Transient insomnia may be induced by stress, sleep in unfamiliar surroundings, jet lag and other factors. Zolpidem, a novel imidazopyridine hypnotic, has been shown to have hypnotic properties in most patients without significantly affecting next-day performance. Using the first-night effect in a sleep laboratory as a model of transient insomnia, this placebo-controlled, double-blind, parallel-group study evaluated the efficacy and safety of zolpidem in 462 normal volunteers. Zolpidem was tested at doses of 5, 7.5, 10, 15 and 20 mg, and statistical analysis of 7.5 mg and 10 mg was compared with placebo (unbalanced randomization). Compared with placebo, the 7.5 mg and 10 mg doses ofzolpidem decreased sleep latency and increased sleep duration and maintenance (i.e. reduced number of awakenings). Zolpidem (7.5 mg or 10 mg) had no significant effect on next-day psychomotor performance. No statistically significant differences in the overall side-effect profiles were found between 7.5-mg and IO-mg zolpidem dose groups and placebo. This study demonstrates that zolpidem at 7.5 mg and 10 mg is effective in the treatment of transient insomnia. Key Words: Hypnotics-Sedatives-Insomnia-Zolpidem-Imidazopyridines-Sleep Wakefulness-Sleep stages. Most individuals experience occasional episodes of acute sleep disturbance (transient insomnia) (1). Transient insomnia can be caused by numerous factors, including a change in sleep environment, travel across time zones or acute shifts in work schedule (1,2). The need for next-day alertness often makes transient insomnia a significant problem. When the sleep-disrupting situation can be anticipated or will last for only a few days, pharmacological intervention is often indicated (3,4). Ideal drug therapy should increase sleep duration and quality without causing serious side effects or impeding next-day psychomotor performance. Previous studies have demonstrated that zolpidem, a novel imidazopyridine, is effective in relieving the symptoms of chronic insomnia (5-8) and short-term insomnia (9). Zolpidem also increased total sleep time, sleep quality and ease of falling asleep in healthy subjects with normal sleep habits (10). Few serious side effects were reported in any study population, including the elderly (5,6,8,11-13), and next-day performance was not negatively affected (11-15). Using the first-night effect as a model of transient Accepted for publication October Address correspondence and reprint requests to Tom Roth, Ph.D., Henry Ford Hospital, Sleep Disorders Center, 2921 West Grand Boulevard, Detroit, MI 48202, U.S.A. 246 insomnia (16,17) the present study examined the effect of zolpidem on transient insomnia in a large subject population. The objectives of this study were to compare the efficacy of 7.5 mg and 10 mg zolpidem with placebo; to evaluate psychomotor function the morning after treatment; and to assess the dose-related effects ofzolpidem at doses of5, 7.5,10, IS and 20 mg. Subjects METHODS Healthy volunteers with normal sleep habits were recruited for this placebo-controlled, two-center, single-night, double-blind, unbalanced, randomized study. For inclusion in the study, subjects were required to report a routine sleep latency :s 30 minutes and sleep duration ::::6 hours. Additionally, all subjects were either males or sterilized females between the ages of 21 and 60 and were required to pass a physical examination and clinical laboratory tests. Because the study sought to evaluate the effects of zolpidem on the firstnight effect (caused by sleeping in an unfamiliar environment), volunteers who had previously slept in a sleep laboratory were excluded. Subjects were also excluded from the study if their histories included drug addiction or alcoholism within

2 ZOLPIDEM IN TRANSIENT INSOMNIA 247 the past year, significant mental or psychiatric disorders, seizures, serious head injury or a history of sleep apnea. Shift workers and others whose sleep schedules were expected to change by at least 6 hours within 7 days of the study were also excluded, as were those with a history of exaggerated response or hypersensitivity to benzodiazepines. A total of 462 subjects met the initial inclusion criteria and were randomly assigned to one of six groups: zolpidem 5 mg (n = 52), 7.5 mg (n = 102), 10 mg (n = 104), 15 mg (n = 51),20 mg (n = 51) or placebo (n = 102). Because the study focused on the efficacy of 7.5 mg and 10 mg zolpidem compared with placebo, an unbalanced randomization was employed, and twice as many subjects were placed into these groups as the others. All qualifying subjects gave written informed consent and were paid for their participation. Experimental methods conformed to the Declaration of He 1- sinki. Procedures Subjects were instructed not to nap or drink alcohol on the day of the study, or to consume food or caffeinated beverages after 7:00 p.m. Subjects reported to the sleep laboratory at least I hour before their habitual bedtime. A urine drug screen and an alcohol breathalyzer test were conducted, and vital signs were measured. Subjects completed a presleep questionnaire and electrodes for polysomnography (PSG) were applied according to standard procedures (10). Three capsules identical in appearance were administered to all study participants 30 minutes before bedtime. Eight hours later, all subjects were awakened, and polysomnography was discontinued. The polysomnographic records were scored without knowledge of treatment, according to standard criteria (18). Subjects were given I hour to shower and dress, and then they completed a postsleep questionnaire and psychomotor performance tests. Next-day performance was assessed by the Digit Symbol Substitution Tests (DSST) and Symbol Copying Tests (SCT), both of which have been shown to be sensitive in detecting residual effects of hypnotic benzodiazepines (19). Next-morning alertness and ability to concentrate were rated on Visual Analog Scales (V AS) on the Morning Questionnaire. Statistical analysis The six treatment groups were compared with respect to sleep histories, demography and vital signs. Chi-square tests were used for categorical data, and a two-way analysis of variance (ANOVA) procedure was applied to continuous data. The logit of sleep efficiency (i.e. the logarithm of p/l - p, where p = sleep time/ time in bed) and the logarithm oflatency to persistent sleep and subjective latency were used in all analyses oflatency and efficiency. The reason for this transformation was that the distributions of latency and efficiency of sleep were skewed and more variable in the placebo group than in the zolpidem groups. Thus, transformation of these two variables was used to lessen the inequality of cell variance and decrease the skewness ofthe distributions of the variables. Cochran Mantel-Haenszel tests were used to adjust for investigator effect in the comparison of dichotomous variables across treatment groups. Efficacy was assessed by a two-way ANOV A, incorporating effect for treatment (six levels), investigator (two levels) and treatment-by-investigator interaction. When a significant overall treatment difference (p < 0.05) was observed, pairwise comparisons restricted to the placebo versus 7.5-mg and I O-mg treatment groups were performed with t tests using the mean square error from the ANOV A. Confirmatory nonparametric analyses were also performed using the Kruskal-Wallis Test. ANOV A was employed for the other PSG parameters, the morning questionnaire responses and the psychomotor performance tests (DSST and SCT). Dose response was assessed with a linear regression analysis utilizing data from all dose groups. A model with linear and quadratic terms was initially fitted, and then terms were dropped in a stepwise fashion. The resulting model was then tested for goodness of fit. A p-value of <0.05 was considered statistically significant for all tests. All pairwise treatment comparisons utilized twotailed tests. Study population RESULTS No significant differences existed among the six treatment groups with respect to demography or baseline characteristics (Table I). A total of24 subjects out of 462 were excluded from the data set because of positive alcohol/drug tests (n = 11), consumption of caffeinated beverages (n = 10) and daytime napping (n = 3). Sleep induction and maintenance Compared to placebo, zolpidem 7.5 mg and 10 mg significantly reduced latency to persistent sleep, defined as the time from the beginning of PSG recording until the onset of the first 10 minutes of consecutive sleep. Mean latency to persistent sleep for the placebo group was 27 minutes (Table 2). Compared to the placebo group, mean sleep latency was significantly shorter for subjects treated with 7.5 mg or 10 mg zolpidem. Sleep, Vol. 18, No.4, 1995

3 248 T. ROTH ET AL. TABLE 1. Characteristics of subjects by randomized group Placebo 5 mg 7.5 mg Variable (n = 102) (n = 52) (n = 102) Gender [n (%)] Male 83 (81) 36 (69) 82 (80) Female 19 (19) 16 (31) 20 (20) Race [n (%)] White 84 (82) 43 (83) 86 (84) Black 15 (5) 6 (12) 12 (12) Other 3 (3) 3 (6) 4 (4) Age (years) Mean SD Weight (ks) Mean SD Height (cm) Mean SD " ns = No significant between-group differences. Zolpidem 10 mg 15 mg 20mg (n = 104) (n = 51) (n = 51} p-value 85 (82) 38 (75) 42 (82) (18) 13 (25) 9 (18) " 87 (84) 44 (86) 40 (78) (13) 4 (8) 10 (20) 4 (4) 3 (6) I (2) Subjects' perceptions of sleep latency were consistent with PSG results. Subjective sleep latency was significantly shorter than placebo for subjects taking 7.5 mg (p = 0.009) or 10 mg (p < 0.001) zolpidem (Table 2). Subjects treated with zolpidem 7.5 mg or 10 mg also reported significantly greater ease in falling asleep than did those taking placebo. On a scale of 0 to 100 (0 = "very easy" and 100 = "not at all easy"), the placebo group assessed average ease of falling asleep at 46, whereas the average was 30 for the 7.5-mg zolpidem group (p < 0.001) and 31 for the 10-mg group (p < 0.001). Sleep efficiency (total sleeping time/time in bed) was also significantly higher (p < 0.001) for zolpidem 7.5 mg and 10 mg than for placebo. Total bedtime was limited to 8 hours in order to reduce variability of sleep efficiency while providing adequate amount of sleep. Sleep efficiency for the placebo group was 88% and increased to 92% with both 7.5 mgand 10 mgzolpidem (Table 2). The mean number of awakenings was significantly reduced in subjects treated with 7.5 mg or 10 mg zolpidem compared to placebo (p = and p = 0.014, respectively). On the PSG recording, subjects in the placebo group woke an average of seven times during the night, compared to five awakenings with 7.5 mg or 10 mg zolpidem. Subjective assessments paralleled these findings. Subjects treated with 7.5 mg or 10 mg TABLE 2. Sleep parameters Zolpidem Placebo 5 mg (n = 102) (n = 52) Mean sleep latency (minutes) Polysomnography 27.1 ± 2.6 b 23.8 ± 3.0 Patient assessment 28.8 ± ± 2.0b Sleep efficiencyd (measured 87.8 ± ± 1.0 by polysomnography) Sleep maintenance (measured by polysomnography) No. of awakenings 6.7 ± S ± 0.6 Wake time during sleep (minutes) 31.4 ± ± 3.1 a Overall treatment comparison (all six treatment groups). b Data were unavailable for one patient. e Compared with placebo. d Sleep efficiency = total time asleep/time in bed. 7.5 mg (n = 102) 17.0 ± 1.4 (p < O.OOI)e 18.9 ± 1.2 (p = 0.009)b 91.7 ± 0.7 (p < O.OOl)e 5.0 ± 0.3 (p = 0.004)e 21.2 ± 2.4 (p = 0.004)e 10 mg 15 mg (n = 103) (n = 51) 17.4 ± 1.6 b IS.7 ± 2.5 (p < O.OOI)e 18.2 ± ± 2.7b (p < O.OOI)b 91.S ± ± 1.0 (p < O.OOI)e 5.3 ± ± 0.7 (p = 0.014) 21.4 ± ± 3.1 (p = 0.005)e 20 mg (n = 51) 20.6 ± ± ± ± ± 3.5 p-value a <0.001 < Sleep. Vol. 18. No

4 ZOLPIDEM IN TRANSIENT INSOMNIA 249 The Effects of Zolpidem on the Stages of Sleep (Figure 1) [ 0 Stage 1 0 Stage 2 0 Stage 34 REM I 70r ~ 60 ~ 50 ;:: g-40 - en " E 30 " e cf. 20 FIG Placebo 5 mg 7.5mg 10mg 15 mg 1< Significant difference trom placebo (p < 0.01) + Significant difte(ence trom placebo (p < 0.001) The effects of zolpidem on the stages of sleep. 20 mg zolpidem reported significantly less wake time during sleep than the 31 minutes recorded on PSG for the placebo group (21 minutes for 7.5 mg and 10 mg zolpidem). Subjective quality of sleep, measured by the Morning Questionnaire, was rated significantly (p < 0.001) better in the 7.S-mg and lo-mg treatment groups than in the placebo group. On a scale of 1 to 4 (1 = excellent, 2 = very good, 3 = fair, 4 = poor), subjects in both the zolpidem 7.5-mg and 1 O-mg groups rated the quality of their sleep as 2.2, whereas the mean rating of the placebo group was 2.7. Sleep stages Treatment with zolpidem had no effect on stage 1, stage 2 and stages 3-4 sleep. The percentage of slowwave (stages 3-4) sleep in subjects taking 10 mg zolpidem was comparable to placebo (14% vs. 12%), whereas stages 3-4 sleep in the 7.S-mg treatment group was 16% (p < 0.01). Both the 7.5-mgand lo-mggroups had significantly less rapid eye movement (REM) sleep than placebo: 21 % with placebo, 18% (p < 0.001) with 7.5 mgand 19% (p < 0.001) with 10 mg (Fig. 1). When zolpidem dose was increased, similar reductions in REM were observed, but there was no alteration in any other stage of the sleep architecture. Dose response Many study parameters, including sleep latency and sleep efficiency, showed a quadratic relationship relative to zolpidem dose. Although only 7.S-mg and 10- mg zolpidem groups were analyzed statistically in comparison with placebo, a quadratic relationship (defined by a polynomial equation of the second degree) between zolpidem dose and sleep latency was observed. Mean latency to persistent sleep decreased with increase in dosage up to the 7.S-10-mg range. Thereafter, the trend was toward a slight increase, to a mean of 19 minutes for the ls-mg treatment group and 21 minutes for the 20-mg group. Sleep efficiency increased up to the 10-mg dose and then decreased slightly with the ls-mg and 20-mg doses. Next-day performance and alertness Mean scores on both the DSST and SeT revealed no statistically significant differences between the 7.5- mg and lo-mg dose groups and placebo. Subjects in these groups also reported no significant differences in ability to concentrate, perception of sleepiness in the morning or during the day after treatment, or in drugged feeling (Table 3). TABLE 3. Next-day performance and alertness Zolpidem Variable Placebo 5 mg 7.5 mg 10mg 15 mg Digit Symbol 53.9 ± ± ± ± ± 1.9 Substitution Test Symbol Copying Test ± ± ± ± ± 4.2 Ability to 1.9± ± ± ± ± 0.1 concentrate b Morning sleepiness' 69.0 ± ± ± ± ± 3.7 Daytime sleepiness d 39.2% 34.6% 31.0% 31.7% 27.5% Drugged feeling' 85.6 ± ± ± ± ± 3.3 a Overall treatment comparison (all six treatment groups). ns = No significant between-group differences. h 1 = excellent, 2 = very good, 3 = fair, 4 = poor. 'Not at all sleepy, 100 = very sleepy. d Percent responding yes. e 0 = very drugged, loo = not at all drugged. 20 mg p-value a 52.1 ± ± ± ± % ± Sleep. Vol. 18. No

5 250 T. ROTH ET AL. Safety Forty-seven of the 462 randomized subjects spontaneously reported a total of71 adverse events, ranging from dry mouth to headache. The percentage of subjects who experienced adverse events with zolpidem doses at or below the recommended therapeutic dose (i.e. 10 mg) was not significantly different from the percentage in the placebo group (7.8% with placebo, 3.8% at 5 mg, 4.9% at 7.5 mg and 6.7% at 10 mg). At the higher dose of 15 mg zolpidem, 17.6% of patients experienced adverse events (p = vs. placebo) and at 20 mg, 31.4% reported adverse events (p < vs. placebo). DISCUSSION Disturbed sleep lasting only a few nights (transient insomnia) can result from many emotional and environmental factors, including acute stress, bereavement, shift work, travel across time zones or use of stimulants (1,2,20). Depending on age, personality and sleep habits, sleep disruption may take the form of prolonged sleep latency, increased number or duration of awakenings, and/or premature morning awakening and decreased total sleep time (1). Zolpidem (N,N,6-trimethyl 1-2(4-methylphenyl) imidazo-[ 1,2,-a]pyridine-3-acetamide hemitartrate) was developed to treat insomnia. Zolpidem binds to benzodiazepine (BZ) receptor sites in the brain. Although most benzodiazepines seem to lack selectivity for specific receptors, zolpidem appears to be a selective agonist at the alpha receptors (subunit of BZ 1 receptors) (21,22). This selective binding may explain the relative absence of my ore lax ant and anticonvulsant effects in animal studies as well as the preservation of deep sleep in human studies (23). According to a National Institute of Health Consensus Conference on drugs and insomnia, the most appropriate role for hypnotic therapy is in transient sleep disorders (3,4). It has been hypothesized that insomnia may begin with an untreated transient or short-term insomnia. It is believed that prompt evaluation and treatment of either condition will prevent the development of chronic insomnia (24). The present study was designed to elucidate the efficacy and secondary effects of zolpidem in the treatment of transient insomnia. To reproduce transient insomnia, the firstnight effect model was utilized (17). As indicated by numerous studies, subjects generally experience difficulty sleeping on their first night in a sleep laboratory (16,17). The resulting delay in sleep onset, increased number of awakenings and decreased sleep time and efficiency in the laboratory setting simulate transient insomnia and provide an excellent model for investigation (17). In this study of transient insomnia, zolpidem proved significantly more effective than placebo in all sleep induction and maintenance parameters. This trial, designed principally to evaluate the efficacy and safety of the 7.S-mg and lo-mg doses, demonstrated that these doses decreased sleep latency and improved sleep duration and quality. Although this study was not designed to statistically compare the zolpidem 15-mg and 20-mg doses to placebo, it is worthwhile to note that increasing the dose above 10 mg did not result in a corresponding increase in hypnotic efficacy. This observation confirms previous observations in trials with zolpidem in different patient populations (14,15). Thus, our findings in this sleep laboratory model of transient insomnia suggest that zolpidem may also be effective in treating acute sleep disturbance caused by stress, a change in sleep environment and travel across time zones. Importantly, using objective and subjective parameters, zolpidem did not have any residual effects the following morning. No subjects reported clinically important adverse events at any zolpidem dosage in this study. The percentage of patients experiencing side effects at doses of 10 mg or less was comparable to the placebo group. The results of this study with respect to efficacy, residual effects and adverse events corroborate the positive findings of earlier research on zolpidem (5-15). Both objective (PSG) and subjective measures confirmed zolpidem's efficacy. This correlation is important in practice because patients are likely to follow dosing recommendations if they perceive positive results (25). If they do not perceive that a hypnotic is improving their sleep, patients may take an additional dose before bedtime, increasing the potential for adverse events and residual effects. The perceived and measured effectiveness of zolpidem in decreasing sleep latency, increasing and maintaining sleep duration, and improving sleep quality without causing significant side effects or affecting nextday performance suggests that zolpidem is important in the treatment of transient insomnia. REFERENCES I. Parkes JD. Insomnia. In: Sleep and its disorders. London: WB Saunders, 1985: Roth T, Roehrs T, ZorickF. Pharmacological treatment of sleep disorders. In: Williams RL, Karacan I, Moore CA, eds. Sleep disorders: diagnosis and treatment, 2nd ed. New York: John Wiley & Sons, 1988: Consensus Conference. Drugs and insomnia: the use of medications to promote sleep. lama 1984;251: Gillin Je. Relieffrom situational insomnia postgrad. Medicine 1992;92: Sleep, Vol. 18, No.4, 1995

6 ZOLPIDEM IN TRANSIENT INSOMNIA ScharfMB, Vogel GW, Pegram GV, et al. Dose response effects of zolpidem in patients with chronic insomnia. Sleep Res 1990; 19: Lahmeyer H, Doherty J, Fillingim J, et al. The effect ofzolpidem in outpatients with chronic insomnia. Sleep Res 1990; 19: Vogel G, Scharf M, Walsh J, Roth T. Effects of chronically administered zolpidem on the sleep of healthy insomniacs. Sleep Res 1989; 18: Scharf M, Vogel G, Kaffeman M, Ochs R. Dose-response of zolpidem in outpatients with chronic insomnia. Sleep Res 1991; 20: Wheatley D. Zolpidem and placebo: a study in general practice in patients suffering from insomnia. In: Sauvanet JP, Langer SZ, Morselli PL, cds. 10. Merlotti L, Roehrs T, Koshorek G, Zorick F, Lamphere J, Roth T. The dose effects ofzolpidem on the sleep of healthy normals. J Clin PsychopharmacoI1989;9:9-14. II. ScharfMB, Mayleben DW, Kaffeman M, Krall R, Ochs R. Dose response effects ofzolpidem in normal geriatric subjects. J Clin Psychiatry 1991;52: Shaw SH, Curson H, Coquelin JP. A double-blind, comparative study ofzolpidem and placebo in the treatment of insomnia in elderly psychiatric in-patients. J Int Med Res 1992;20: Maarek L, Cramer P, Attali P, Coquelin JP, Morselli PL. The safety and efficacy of zolpidem in insomniac patients: a longterm open study in general practice. J Int Med Res 1992;20: Nicholson AN, Pascoe PA. Hypnotic activity of an imidazopyridine (zolpidem). Br J Clin Pharmacol 1986;21 : Richens A, Mercer AJ, Jonens DM, Griffiths A, Marshall RW. Effects of zolpidem on saccadic eye movements and psychomotor performance: a double-blind, placebo controlled study in healthy volunteers. Br J Clin PharmacoI1993;36: Vogel G. Clinical uses and advantages of low doses of benzodiazepine hypnotics. J Clin Psychiatry 1992;53(Suppi): Roehrs T, Vogel G, Sterling W, Roth T. Dose effects oftemazepam in transient insomnia. Drug Res 1990;40(11): Rechtschaffen A, Kales A. A manual ojstandardized terminology, techniques, and scoring systems Jor sleep stages oj human subjects. Washington, DC: U.S. Government Printing Office, Roth T, Hauri P, Zorick F, Sateia M, Roehrs T, Kipp J. The effects ofmidazolam and temazepam on sleep and performance when administered in the middle of the night. J Clin PsychopharmacoI1985;5(2): Gillin JC, Byerley WF. The diagnosis and management of insomnia. N Engl J Med 1990;322: Benavides J, Peny B, Durant A, Arbilla S, Scatton B. Comparative in vivo and in vitro regional selectivity of central omega (benzodiazepine) site ligands in inhibiting ph]flumazenil binding in the rat central nervous system. J Pharmacol Exp Ther 1992;263: Benavides J, Peny B, Dubois A, et al. In vivo interaction of zolpidem with central benzodiazepine (BZD) binding sites (as labeled by [,H]Ro ) in the mouse brain. Preferential affinity ofzolpidem for the omega I (BZD I) subtype. J Pharmacol Exp Ther 1988;245: Perrault G, Morel E, Sanger DJ, Zivkovic B. Differences in pharmacological profiles of a new generation of benzodiazepine and non-benzodiazepine hypnotics. Eur J Pharmacol1990; 187: Nino-Murcia G. Diagnosis and treatment of insomnia and risks associated with lack of treatment. J Clin Psychiatry 1992; 53(Suppi): Roehrs T, Vogel G, Roth T. Rebound insomnia: its determinants and significance. Am J Med I 990;88(Suppl 3A):3A-395-3A-425. Sleep. Vol. 18. No.4, 1995